AP40
AIR POLLUTION
ENGINEERING MANUAL
AIR POLLUTION CONTROL DISTRICT
COUNTY OF LOS ANGELES
Compiled and Edited
by
John A- Danielson
U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
Public Health Service
Bureau of Disease Prevention Environmental Control
NATIONAL CENTER FOR AIR POLLUTION CONTROL
Cincinnati, Ohio
1967
-------
The ENVIRONMENTAL HEALTH SERIES of reports was established to re-
port the results of scientific and engineering studies of man's environment; the
community, whether urban, suburban, or rural, where he lives, works, and re-
laxes; the air, water, and earth he uses and re-uses; and the wastes he produces
and must dispose of in a way that preserves these natural resources. This SERIES
of reports provides for professional users a central source of information on the
intramural research activities of the Centers in the Bureau of Disease Prevention
and Environmental Control, and on their cooperative activities with state and local
agencies, research institutions, and industrial organizations. The general subject
area of each report is indicated by the letters that appear in the publication num-
ber; the indicators are
AP -- Air Pollution
RH -- Radiological Health
UIH -- Urban and Industrial Health
Reports in the SERIES will be distributed to requesters, as supplies permit.
Requests should be directed to the Center identified on the title page.
Public Health Service Publication No. 999-AP-40
?7i:il AC-DH3
-------
FOREWORD
As concern for the quality of the atmosphere has g
concern. Federal, State and local programs are a
sibility in the development and practice of the mar y
standing and resolution of the air pollution problem.
rown, so also has the response to that
ssuming increasingly greater respon-
disciplines that contribute to under-
Rapid program expansion imposes even greater
edge in the field of air pollution control. Much
tent scientists and engineers. However, in many in
been transcribed and organized into a form read:
info r mation.
demands for the dissemination of knowl-
has been accomplished by compe-
tances , the experience gained has not
ly accessible to those most in need of
to
We are pleased, therefore, to have the opportunity
Air Pollution Engineering Manual. Distilling as it d<
years of painstaking engineering innovation in the
cover, it should become a valuable--if not in<
idispensabl
knowledg
The manual is an outgrowth of the practical
of the Los Angeles County Air Pollution Control D
in the field. District personnel have worked clo
controls where none formerly existed.
It will be noted that there are categories of industr
The reason is that engineering control applications
tries located in Los Angeles County.
The manual was originated as a training aid for Dis
administrations of Mr, S. Smith Griswold, the f<
current, Air Pollution Control Officer for the Los
District. The editorial and technical content wer
the District. The staff, in turn, was supervised
Mr. Robert L. Chass, Chief Deputy Air Pollutioi
Lunche, Director of Engineering. Mr. John A. Dan
served as editor.
The U. S. Public Health Service, recognizing the
serve as publisher.
make available this new volume, the
es the equivalent of hundreds of man-
air pollution control field under one
e--tool.
e gained by the technical personnel
strict, long recognized as outstanding
ely with industry to develop emission
Lai emissions that are not discussed.
are described for only those indus-
:rict and industry engineers under the
rmer, and Mr. Louis J. Fuller, the
Angeles County Air Pollution Control
3 developed exclusively by the staff of
the development of the manual by
Control Officer and Mr. Robert G.
.elson, Senior Air Pollution Engineer,
during
need for such a manual, is pleased to
John T. Middle ton
Director
National Center for Air Pollution Control
This manual was prepared for publication by Mrs. Pauline Elliott, who did the
composition and makeup, and Mr. W. Robert Mobley, Consultant, under the di-
rection of Mr. Kenneth Cassel, Jr., at facilities of the U.S. Public Health
Service in Cincinnati, Ohio.
-------
PREFACE
This Air Pollution Engineering Manual deals with the control of air pollution at individual
sources. This approach is unique because it emphasizes the practical engineering prob-
lems of design and operation associated with the many sources of air pollution. These
sources reside in metallurgical, mechanical, incineration, combustion, petroleum, and
chemical processes. Although the literature contains excellent data on some of these pro-
cesses, no handbook or manual has ever been compiled to organize the data in this spe-
cialized branch of engineering until now. This manual should, therefore, fill a need.
That the air pollution problems of one area are different from those of another is well
known. The air pollution problems presented here originate in industrial and commercial
sources peculiar to the Los Angeles area. Consequently, some processes, such as the
burning of coal in combustion equipment, are not mentioned. Furthermore, the degree of
air pollution control strived for in this manual corresponds to the degree of control de-
manded by air pollution statutes of the Los Angeles County Air Pollution Control District.
Manyother areas require less stringent control and permit less efficient control devices.
This manual consists of 11 chapters, each by different authors," and 4 appendixes. The
first five chapters treat the history of air pollution in Los Angeles County, the types of
air contaminants, and the design of air pollution control devices. The remaining chapters
discuss the control of air pollution from specific sources. A reader interested in control-
ling air pollution from a specific source can gain the information needed by referring only
to the chapter of the manual dealing with that source. If he then desires more general in-
formation about an air pollution control device, he can refer to the chapters on control de-
vices. First, however, he should read Chapter 1 because it cites for Los Angeles County
the prohibitory rules that regulate the degree of control efficiency attained by the described
equipment.
Sole responsibility for the information is borne by the District, which presents the manual
as a contribution toward the advancement of national understanding of the control of air
pollution from stationary sources.
Louis J. Fuller
Air Pollution Control Officer
County of Los Angeles
-------
ACKNOWLEDGMENT
Under the provisions of the California law creating the Los Angeles County Air Pollution
Control District, the Board of Supervisors is empowered to act as the Air Pollution Con-
trol Board. Responsible for supervision and policy determination for the District, their
firm support of needed air pollution control measures has advanced engineering capability
in this field to a high degree. The information gained in Los Angeles County is applicable
to the improvement of air quality wherever air pollution is experienced. Without the sup-
port of this Board, the information presented here would not have been possible.
THE BOARD OF SUPERVISORS
OF LOS ANGELES COUNTY
BURTON W. CHACE, Chairman
Fourth District
FRANK G. BONELLI ERNEST E. DEBS
First District Third District
KENNETH HAHN WARREN M. DORN
Second District Fifth District
-------
EDITORIAL REVIEW COMMITTEE
Robert L. Chass
Robert G. Lunche
Eric E. Lemke
Robert J. Mac Knight
John L. Mills
TECHNICAL ASSISTANCE
Ivan S. Deckert
William F. Hammond
William B. Krenz
John L. Me Ginnity
Robert C. Murray
Robert T. Walsh
John E. Williamson
EDITORIAL ASSISTANCE
Jerome D. Beale
George Thomas
Edwin J. Vincent
Wayne E. Zwiacher
GRAPHIC ART
Lewis K. Smith
-------
CONTENTS
CHAPTER 1. INTRODUCTION
THE LOS ANGELES BASIN 3
RULES AND REGULATIONS IN LOS ANGELES COUNTY 3
Regulation II: Permits 4
Regulation IV: Prohibitions 4
Rule 50: The Ringelman Chart 4
Rule 51: Nuisance 4
Rule 52: Particulate Matter 5
Rule 53: Specific Contaminants 5
Rule 53. 1: Scavenger Plants 5
Rule 54: Dust and Fumes 5
Rule 56: Storage of Petroleum Products 5
Rules 57 and 58: Open Fires and Incinerators 5
Rule 59: Oil-Effluent Water Separators 5
Rule 61: Gasoline Loading Into Tank Trucks and Trailers 5
Rules 62 and 62. 1: Sulfur Content of Fuels 5
Rule 63: Gasoline Specifications 6
Rule 64: Reduction of Animal Matter 6
Rule 65: Gasoline Loading Into Tanks 6
Rule 66: Organic Solvents 6
Rule 66. 1: Architectural Coatings 6
Rule 66.2: Disposal and Evaporation of Solvents 6
ROLE OF THE AIR POLLUTION ENGINEER 6
Accomplishments of the Permit System 6
USE OF THIS MANUAL 6
General Design Problems ./^f- 7
Specific Air Pollution Sources < 7
CHAPTER 2. AIR CONTAMINANTS
INTRODUCTION 11
FACTORS IN AIR POLLUTION PROBLEMS 11
TYPES OF AIR CONTAMINANTS 12
Organic Gases 12
Current Sources in Los Angeles County 12
Hydrocarbons - 12 . . . Hydrocarbon derivatives - 12
Significance in Air Pollution Problem 14
Inorganic Gases 14
Current Sources in Los Angeles County 14
Oxides of nitrogen - 14 . . .Oxides of sulfur - 14. . . Carbon monoxide - 15
Significance in Air Pollution Problem 15
Oxides of nitrogen - 15 ... Oxides of sulfur - 15 ... Carbon monoxide - 16 ...
Miscellaneous inorganic gases - 16
Aerosols 16
Current Sources in Los Angeles County 16
Carbon or soot particles - 16. . . Metallic oxides and salts - 17. . Oily or tarry droplets - 17
Acid droplets - 17 . . Silicates and other inorganic dusts - 18. . . Metallic fumes - 18
Significance in Air Pollution Problem 18
AIR POLLUTION CONTROLS ALREADY IN EFFECT 18
CONTROL MEASURES STILL NEEDED 18
Motor Vehicle Emissions 20
Additional Controls Over Stationary Sources 20
Organic Gases 20
-------
CONTENTS
Oxides of Nitrogen 21
Oxides of Sulfur 21
Other Contaminants , 21
CHAPTER 3. DESIGN OF LOCAL EXHAUST SYSTEMS
FLUID FLOW FUNDAMENTALS 25
Bernoulli's Equation 25
Pitot Tube for Flow Measurement 25
Correction Factors 27
HOOD DESIGN 27
Continuity Equation 27
Air Flow Into a Duct 28
Null Point 28
Design of Hoods for Cold Processes 30
Spray Booths 32
Abrasive Blasting 32
Open-Surface Tanks 32
Design of Hoods for Hot Processes 34
Canopy Hoods 34
Circular high-canopy hoods - 34. . Rectangular high-canopy hoods - 38. . .
Circular low-canopy hoods - 39 . . Rectangular low-canopy hoods - 40. . .
Enclosures - 41
Specific Problems 42
Steaming tanks - 42 . . Preventing leakage - 42
Hood Construction 43
High-Temperature Materials 43
Corrosion-Resistant Materials 43
Design Proportions 43
Transition to Exhaust Duct 43
DUCT DESIGN 44
General Layout Considerations 44
Types of Losses 44
Inertia Losses 44
Orifice Losses 44
Straight-Duct Friction Losses 45
Elbow and Branch Entry Losses 45
Exhaust system calculator - 45 45
Contraction and Expansion Losses 47
Collection Equipment 47
Design Procedures 47
Methods of Calculation 47
Methods of Design 48
Calculation Procedures 49
Fan Static Pressure 50
Balanced-Duct Calculations 51
Blast Gate Method 52
Checking an Exhaust System 52
Illustrative Problem 52
Fan Curve Calculator 57
Corrections for Temperature and Elevation 57
Duct Construction 59
FAN DESIGN 60
Centrifugal Fans 60
Axial-Flow Fans 60
Fan Characteristics 61
Influence of Blade Shape 61
Geometrically Similar Fans 62
Multirating Tables 63
Fan Laws 63
Selecting a Fan From Multirating Tables 65
Construction Properties 65
-------
CONTENTS
Heat Resistance • 65
Explosive-Proof Fans and Motors 65
Fan Drives 67
VAPOR COMPRESSORS 67
Types of Compressors 67
Positive-Displacement Compressors 68
Dynamic Compressors 69
Reciprocating Compressors 69
Use in Air Pollution Control 72
CHECKING OF EXHAUST SYSTEM 72
Theory of Field Testing 72
Quantity Meters 72
Velocity Meters 72
Pitot Tubes 72
Pitot Tube Traversing Procedure 73
Altitude and Temperature Corrections for Pitot Tubes 73
Swinging-Vane Velocity Meter 73
Calibrating the Velocity Meter 74
Uses of the Velocity Meter 75
COOLING OF GASEOUS EFFLUENTS 76
Methods of Cooling Gases 76
Dilution With Ambient Air 76
Quenching With Water 79
Natural Convection and Radiation 81
Forced-Draft Cooling 86
Factors Determining Selection of Cooling Device 86
CHAPTER 4. AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
INERTIAL SEPARATORS 91
Single-Cyclone Separators 91
Theory of Operation 92
Separation Efficiency 93
Pressure Drop 93
Other Types of Cyclone Separators 94
High-Efficiency Cyclone Separators 94
Multiple-Cyclone Separators 94
Mechanical, Centrifugal Separators 94
Predicting Efficiency of Cyclones 95
Method of Solving a Problem 97
WET COLLECTION DEVICES 99
Theory of Collection 100
Mechanisms for Wetting the Particle 100
Types of Wet Collection Devices 101
Spray Chambers 101
Cyclone-Type Scrubbers 101
Orifice-Type Scrubbers 101
Mechanical Scrubbers 102
Mechanical, Centrifugal Collector With Water Sprays 102
High-Pressure Sprays . , 103
Venturi Scrubbers 104
Packed Towers 104
Wet Filters 105
The Role of Wet Collection Devices 105
BAGHOUSES 106
Filtration Process 106
Mechanisms 106
Direct interception - 108. . . Impingement - 109 . . .Diffusion - 110 . .Electrostatics - 110
Baghouse Resistance 110
Clean cloth resistance - 110 . . Resistance of dust mat - 111 ...
Effect of resistance on design - 115
Filtering Velocity 116
-------
CONTENTS
Filtering Media 118
Fibers 118
Cotton - 118 . . .Wool - 118 . . .Nylon - 118 . . Dynel - 118. . . Orion and Dacron - 118. . .
Teflon - 119 . . .Glass - 119
Yarn 119
Filament yarns - 119 . . Staple yarns - 119
Weave 120
Plain weave - 120. . . Twillweave - 1ZO. . . Satin weave - 1ZO
Finish 120
Size and Shape of Filters 121
Diameters of tubular filtering elements - 121 .. .Length of tubular bags - 121 . . .
Length-to-diameter ratio - 121. . . Multiple-tube bags - 122 . . .Envelope type - 122
Installation of Filters 122
Arrangement 122
Bag Spacing 123
Bag Attachment 123
Bottom attachments - 123 . . .Top support - 123
Cleaning of Filters 124
Methods 124
Manual cleaning - 124. . . Mechanical shakers - 125. . . Pneumatic shakers - 125. . .
Bag collapse - 127 . . .Sonic cleaning - 127. . . Reverse airflow - 128. . .
Reverse-air jets - 128
Cleaning Cycles 130
Manually initiated cycles - 130. . . Semiautomatic cycles - 131 ...
Fully automatic cycles - 131 . . .Continuous cleaning - 131
Disposal of Collected Dust 131
Baghouse Construction 132
Pushthrough versus Pullthrough 132
Structural Design 132
Hoppers 132
Size - 132. . . Slope of hopper sides - 133. . . Gage of metal - 133 . . .
Use of vibrators and rappers - 133 . . .Discharge - 134
Maintenance 134
Service 134
Bag Replacement 134
Precoating 135
SINGLE-STAGE ELECTRICAL PRECIPITATORS 135
History of Electrostatic Precipitation 135
Origins of Electrostatic Principles 135
Early Experiments With Electrostatics on Air Contaminants 137
Development of the First Successful Precipitator 137
Improvements in Design, and Acceptance by Industry 138
Advantages and Disadvantages of Electrical Precipitation 138
Mechanisms Involved in Electrical Precipitation 139
Diverse Applications of Electrical Precipitation 141
Construction Details of Electrical Precipitators 141
Discharge electrodes - 141 . . .Collecting electrodes - 141 . . .
Tubular collecting electrodes - 141. . . Removal of dust from collecting electrodes - 143. . .
Precipitator shells and hoppers - 144
High Voltage for Successful Operation 145
Tube-type rectifiers - 145. . . Solid-state rectifiers - 145
Effects of Wave Form 145
Controlled Sparking Rate 146
Operating Voltage 146
Uniform Gas Distribution 146
Cost of Electrical Precipitator Installations 146
Theoretical Analysis of Precipitator Performance 147
Particle charging - 147. . . Particle migration - 148
Theoretical Efficiency 149
Deficiencies in Theoretical Approach to Precipitator Efficiency. 150
Effects of Resistivity 150
-------
CONTENTS
Methods of Reducing Reentrainment 152
Practical Equations for Precipitator Design and Efficiency 153
Effects of Nonuniform Gas Velocity 154
Important Factors in the Design of a Precipitator 156
Summary and Conclusions 156
TWO-STAGE ELECTRICAL PRECIPITATORS 156
Theoretical Aspects 158
Theory of Dust Separation 158
Particle charging - 159
Drift Velocity 159
Efficiency 160
Design Factors 160
Electrical Requirements 160
Air Capacity 161
Air Distribution 161
Auxiliary Controls 162
Construction and Operation 163
Assembly 163
Maintenance 163
Safety 163
Application 163
Two-Stage Precipitators of Special Design 165
Equipment Selection 165
OTHER PARTICULATE-COLLECTING DEVICES 166
Settling Chambers 166
Impingement Separators 166
Panel Filters 167
Precleaners 168
CHAPTER 5. CONTROL EQUIPMENT FOR GASES AND VAPORS
AFTERBURNERS 171
Direct-Fired Afterburners 171
Specifications and Design Parameters 171
Operation 173
Efficiency 174
Design calculations - 174
Installation costs 178
Catalytic Afterburners 178
Specifications and Design Parameters 178
Operation 180
Efficiency 184
Design calculations - 185
Installation Costs 187
BOILERS USED AS AFTERBURNERS 187
Conditions for Use 187
Manner of Venting Contaminated Gases 188
Adaptable Types of Equipment 189
Boilers and Fired Heaters 189
Burners 190
Safety 190
Design Procedure 190
Test Data 192
ADSORPTION EQUIPMENT 192
Types of Adsorbents 194
Use of Activated Carbon in Air Pollution Control 194
Saturation 194
Retentivity 1 95
Breakpoint 1 95
Heat of Adsorption 195
Carbon Regeneration 1 96
Equipment Design 196
-------
CONTENTS
Fixed-Bed Adsorber 197
Conical fixed-bed adsorber - 199
Continuous Adsorber 200
Pressure Drop 200
Operational Problems 201
Particulate Matter 201
Corrosion 201
Polar and Nonpolar Compounds 201
VAPOR CONDENSERS 201
Types of Condensers 202
Surface and Contact Condensers 202
Typical Installations 204
Condensers in Control Systems 204
Subcooling Condensate 205
Contact Condensers 206
Sizing Contact Condensers 206
Surface Condensers 206
Characteristics of Condensation 206
Design of Surface Condensers 206
Applications 210
GAS ABSORPTION EQUIPMENT 210
General Types of Absorbers 211
Packed Tower Design 211
Packing Materials 212
Liquid Distribution 212
Tower Capacity 213
Tower Diameter 214
Number of Transfer Units (NTU) 216
Height of a Transfer Unit 217
Pressure Drop Through Packing 219
Illustrative Problem 219
Plate or Tray Towers 223
Types of Plates 223
Bubble Cap Plate Tower Design 224
Liquid Flow 224
Plate Design and Efficiency 224
Flooding 225
Liquid Gradient on Plate 226
Plate Spacing 227
Tower Diameter 227
Number of Theoretical Plates 228
Illustrative Problem 228
Comparison of Packed and Plate Towers 230
Vessels for Dispersion of Gas in Liquid 231
Spray Towers and Spray Chambers 231
Venturi Absorbers 231
CHAPTER 6. METALLURGICAL EQUIPMENT ^--^
FURNACE TYPES (^235
Reverberatory Furnace 235
Cupola Furnace 236
Combustion Air 236
Methods of Charging 237
Preheating Combustion Air 237
Electric Furnace 238
Direct-Arc Furnace 238
Indirect-Arc Furnace 238
Induction Furnace 239
Resistance Furnace 239
Crucible Furnace 239
-------
CONTENTS
Tilting Furnace 240
Pit Crucible 240
Stationary Crucible 240
Pot Furnace 240
STEEL-MANUFACTURING PROCESSES 241
Open-Hearth Furnaces 242
The Air Pollution Problem 243
Hooding and Ventilation Requirements 244
Air Pollution Control Equipment 246
Electric-Arc Furnaces f 247
The Air Pollution Problem 248
Hooding and Ventilation Requirements 248
Air Pollution Control Equipment 250
Baghouse dust collectors - 250. . Electrical precipitators - 252. . .Water scrubbers - 255
Electric-Induction Furnace 256
The Air Pollution Problem 256
Hooding and Ventilation Requirements 256
Air Pollution Control Equipment 257
IRON CASTING 258
Cupola Furnaces 258
The Air Pollution Problem 258
Hooding and Ventilation Requirements 258
Air Pollution Control Equipment 258
Afterburners - 258. . Baghouse dust collectors - 260. . Electrical precipitators - 262
Illustrative Problem 265
Electric-Arc Furnaces 267
The Air Pollution Problem 268
Hooding and Ventilation Requirements 268
Air Pollution Control Equipment 268
Baghouse dust collectors - 268. . Electrical precipitators - 268
Induction Furnaces 268
Reverberatory Furnaces 268
SECONDARY BRASS- AND BRONZE-MELTING PROCESSES 270
Furnace Types 270
The Air Pollution Problem 270
Characteristics of emissions - 270 . . Factors causing large concentrations of zinc fumes - 271
Crucible furnace--pit and tilt type - 273 . .Electric furnace--low-frequency induction type - 273
Cupola furnace - 274
Hooding and Ventilation Requirements 274
Reverberatory furnace--open-hearth type - 274. . Reverberatory furnace--cylindrical type - 276
Reverberatory furnace--tilting type - 276. . Reverberatory furnace--rotary tilting type - 276 . .
Crucible-type furnaces - 279. • Low-frequency induction furnace - 279. . Cupola furnace - 280
Air Pollution Control Equipment 280
Baghouses - 280 . .Electrical precipitators - 281 . .Scrubbers - 281
SECONDARY ALUMINUM-MEL TING PROCESSES 284
Types of Process 284
Crucible Furnaces 284
Reverberatory Furnaces 284
Fuel-Fired Furnaces 285
Electrically Heated Furnaces 285
Charging Practices 285
Pouring Practices 286
Fluxing 286
Cover fluxes - 287. . Solvent fluxes - 287 . .Degassing fluxes - 287. . .
Magnesium-reducing fluxes - 287
The Air Pollution Problem 288
Particle Size of Fumes From Fluxing 289
Hooding and Ventilation Requirements 289
Air Pollution Control Equipment 290
SECONDARY ZINC-MELTING PROCESSES 293
Zinc Melting 293
-------
CONTENTS
The Air Pollution Problem 294
Zinc Vaporization 294
Reduction Retort Furnaces 294
Reduction in Belgian Retorts 294
The Air Pollution Problem 296
Hooding and Ventilation Requirements 296
Distillation Retort Furnaces 296
The Air Pollution Problem 297
Hooding and Ventilation Requirements 297
Muffle Furnaces 297
The Air Pollution Problem 298
Hooding and Ventilation Requirements 299
Air Pollution Control Equipment 299
LEAD REFINING 300
Reverberatory Furnaces 300
The Air Pollution Problem 300
Hooding and Ventilation Requirements 301
Air Pollution Control Equipment 301
Lead Blast Furnaces 302
The Air Pollution Problem 303
Hooding and Ventilation Requirements 304
Air Pollution Control Equipment 304
Pot-Type Furnaces 304
The Air Pollution Problem 304
Hooding and Ventilation Requirements 305
Air Pollution Control Equipment 305
Barton Process 305
METAL SEPARATION PROCESSES 305
Aluminum Sweating 305
Zinc, Lead, Tin, Solder, and Low-Melting Alloy Sweating 306
The Air Pollution Problem 306
Contaminants from aluminum-separating processes - 306. . .
Contaminants from low-temperature sweating - 306
Hooding and Ventilation Requirements 307
Air Pollution Control Equipment 308
Aluminum-separating processes - 308. . . Low-temperature sweating - 308
CORE OVENS 309
Types of Ovens 309
Heating Core Ovens 312
Core Binders 312
The Air Pollution Problem 314
Hooding and Ventilation Requirements 315
Air Pollution Control Equipment 315
FOUNDRY SAND-HANDLING EQUIPMENT 315
Types of Equipment 315
The Air Pollution Problem 316
Hooding and Ventilation Requirements 317
Shakeout grates - 317. . .Other sand-handling equipment - 318
Air Pollution Control Equipment 318
HEAT TREATING SYSTEMS 320
Heat Treating Equipment 320
The Air Pollution Problem 320
Hooding and Ventilation Requirements 321
Air Pollution Control Equipment 321
CHAPTER 7. MECHANICAL EQUIPMENT
HOT-MIX ASPHALT PAVING BATCH PLANTS 325
Introduction 325
Raw Materials Used 325
Basic Equipment 325
Plant Operation 325
-------
CONTENTS
The Air Pollution Problem 326
Hooding and Ventilation Requirements 328
Air Pollution Control Equipment 328
Variables Affecting Scrubber Emissions 330
Collection Efficiencies Attained 332
Cost of Air Pollution Control Equipment 333
CONCRETE-BATCHING PLANTS 334
Wet-Concrete-Batching Plants 334
The Air Pollution Problem 334
Air Pollution Control Equipment 335
Cement-receiving and storage system - 335 . . .Cement weigh hopper - 336 . . .
Gathering hoppers - 336
Dry-Concrete-Batching Plants 336
The Air Pollution Problem 337
Hooding and Ventilation Requirements 337
Air Pollution Control Equipment 337
Dust created by truck movement - 337
Central Mix Plants 337
The Air Pollution Problem 338
Hooding and Ventilation Requirements 339
Air Pollution Control Equipment 339
CEMENT-HANDLING EQUIPMENT 339
The Air Pollution Problem 339
Hooding and Ventilation Requirements 339
Receiving Hoppers 339
Storage and Receiving Bins 339
Elevators and Screw Conveyors 340
Hopper Truck and Car Loading 340
Air Pollution Control Equipment 340
ROCK AND GRAVEL AGGREGATE PLANTS 340
The Air Pollution Problem 341
Hooding and Ventilation Requirements 341
Air Pollution Control Equipment 341
MINERAL WOOL FURNACES 342
Introduction 342
Types and Uses of Mineral Wool Products 342
Mineral Wool Production 342
The Air Pollution Problem 343
Hooding and Ventilation Requirements 344
Air Pollution Control Equipment 347
Baghouse Collection and Cupola Air Contaminants 347
Afterburner Control of Curing Oven Air Contaminants 347
Reducing Blowchamber Emissions 349
Controlling Asphalt Fumes 349
PERLITE-EXPANDING FURNACES 350
Introduction 350
Uses 350
Mining Sites 350
Perlite Expansion Plants 350
Expansion Furnaces 350
Gas and Product Cooling 350
Product Collectors and Classifiers 350
The Air Pollution Problem 351
Hooding and Ventilation Requirements 351
Air Pollution Control Equipment 351
FEED AND GRAIN MILLS 352
Introduction 352
Receiving, Handling, and Storing Operations 353
Feed-Manufacturing Processes 354
The Air Pollution Problem 355
Receiving, Handling, and Storing Operations 356
-------
xviii CONTENTS
Feed-Manufacturing Processes 357
Hooding and Ventilation Requirements 357
Receiving, Handling, and Storing Operations 357
Feed-Manufacturing Processes 358
Air Pollution Control Equipment 358
Receiving, Handling, and Storing Operations 359
Feed-Manufacturing Processes 359
Filter Vents 359
Cyclones 359
Baghouses 360
PNEUMATIC CONVEYING EQUIPMENT 362
Introduction 362
Types of Pneumatic Conveying Systems 362
Types of Air-Moving Used in Conveying 363
Preliminary Design Calculations 365
The Air Pollution Problem 367
Air Pollution Control Equipment 367
DRIERS 367
Introduction 367
Rotary Driers 367
Flash Driers 368
Spray Driers 369
Other Types of Driers 370
The Air Pollution Problem 371
Hooding and Ventilation Requirements 371
Air Pollution Control Equipment 371
Dust Control 371
Drying With Solvent Recovery 371
Smoke and Odor Emissions 372
WOODWORKING EQUIPMENT 372
Exhaust Systems 372
Construction of Exhaust Systems 372
The Air Pollution Problem 373
Hooding and Ventilation Requirements 373
Air Pollution Control Equipment 373
Disposal of Collected Wastes 374
RUBBER-COMPOUNDING EQUIPMENT 375
Introduction 375
Additives Employed in Rubber Compounding 375
The Air Pollution Problem 376
Hooding and Ventilation Requirements 377
Air Pollution Control Equipment 377
ASPHALT ROOFING FELT SATURATORS 378
Description and Operation 378
The Air Pollution Problem 378
Hooding and Ventilation Requirements 378
Air Pollution Control Equipment 378
Low-Voltage Electrical Precipitators 378
Design Considerations for Electrical Precipitators 379
Maintenance of Precipitators 381
Baghouses 382
Scrubbers 382
SOLVENT DEGREASERS 383
Introduction 383
Design and Operation 383
Types of Solvent 383
The Air Pollution Problem 383
Solvent Losses 383
Hooding and Ventilation Requirements 384
Air Pollution Control Equipment 384
Methods of Minimizing Solvent Emissions 384
-------
CONTENTS xix
Tank Covers 385
Controlling Vaporized Solvent 385
SURFACE-COATING OPERATIONS 387
Introduction 387
Spray Booths 3S&
Flowcoating Machines 388
Paint Dip Tanks 388
Roller Coating Machines 389
The Air Pollution Problem 389'
Air Contaminants From Paint Spray Booths 389
Air Contaminants From Other Devices 389
Hooding and Ventilation Requirements 389
Requirements for Paint Spray Booths 389
Requirements for Other Devices 389
Air Pollution Control Equipment 389
Control of Paint Spray Booth Particulates 389
Control of Organic Vapors From Surface Coatings 390
PIPE-COATING EQUIPMENT 390
Introduction 390
Methods of Application 390
Pipe Dipping 390
Pipe Spinning 390
Pipe Wrapping 391
Preparation of enamel - 391
The Air Pollution Problem 391
Hooding and Ventilation Requirements 391
Air Pollution Control Equipment 393
DRY CLEANING EQUIPMENT 393
The Air Pollution Problem 395
Solvents 395
Lint 396
Hooding and Ventilation Requirements 396
Air Pollution Control Equipment 396
ABRASIVE BLAST CLEANING 397
Introduction 397
Abrasive Materials 397
Method of Propelling the Abrasive 397
Equipment Used to Confine the Blast 398
The Air Pollution Problem 399
Hooding and Ventilation Requirements 400
Air Pollution Control Equipment 400
ZINC-GALVANIZING EQUIPMENT 401
Introduction 401
Cleaning 401
Cover Fluxes 402
Foaming Agents 402
Dusting Fluxes 402
The Air Pollution Problem 403
Physical and Chemical Composition of the Contaminants 403
Hooding and Ventilation Requirements 404
Air Pollution Control Equipment 405
Baghouses 406
Electrical Precipitators 408
CHAPTER 8. INCINERATION
DESIGN PRINCIPLES FOR MULTIPLE-CHAMBER INCINERATORS 413
Retort Type 413
In-Line Type 413
Description of the Process 415
Design Types and Limitations 416
Comparison of Types 416
-------
xx CONTENTS
Principles of Combustion 416
Design Factors 417
Design Precepts 417
GENERAL-REFUSE INCINERATORS
The Air Pollution Problem 420
Air Pollution Control Equipment 420
Design Procedure 421
General Construction 422
Refractories 423
Grates and Hearths 423
Air Inlets 423
Stack 425
Induced-Draft System 426
Operation 426
Illustrative Problem 426
MOBILE MULTIPLE-CHAMBER INCINERATORS
Design Procedure 4
Stack Requirements 428
Induced-Draft Fan System 428
Standards of Construction 430
Refractories 430
Grates 430
Air Inlets 430
Structure 430
Auxiliary Burners 430
Stack Emissions 430
Illustrative Problem 431
MULTIPLE-CHAMBER INCINERATORS FOR BURNING WOOD WASTE 436
Introduction 436
Description of the Refuse 436
The Air Pollution Problem 436
Air Pollution Control Equipment 436
Design Procedure 436
Incinerator Arrangements 439
Design Procedure for Mechanical Feed Systems 441
Surge Bin 441
Screw or Drag Conveying 442
Pneumatic Conveying 443
Standards for Construction 443
Refractories 444
Grates 444
Exterior Walls 444
Air Ports 444
Operation of Incinerators 445
Illustrative Problem 445
FLUE-FED APARTMENT INCINERATORS 447
Introduction 447
Description of Refuse 448
The Air Pollution Problem 448
Stack Emissions 448
Air Pollution Control Equipment 448
Installation of Afterburner on a Roof 449
Design Procedure 449
Draft control - 449. . . Chute door locks - 450 . . Design parameters - 450. . .
Limitations - 450. . . Typical installations - 451
Standards for Construction 452
Mounting and supports - 452 . . Metals - 452 . . . Castable refractories - 452 . . .
Firebrick - 452. . .Insulating firebrick - 452. . .Burners - 452 . . .
Draft control damper - 453 . . .Chute door locks - 453
Stack Emissions 453
Operation 453
-------
CONTENTS
Basement Afterburner 454
Design Procedure 454
Design parameters - 454. . . Typical installation
Standards for Construction 455
Hot-zone refractory - 455 . . .Draft control damper - 455
Stack Emissions 455
Operation 455
Multiple-Chamber Incinerator, Basement Installation 455
Design Procedure _ 455
Draft control - 456. . . Typical installation.- 456
Standards for Construction 457
Stack Emissions 457
Operation 457
Illustrative Problem 457
PATHOLOGICAL-WASTE INCINERATORS 460
The Air Pollution Problem 461
Air Pollution Control Equipment 461
Design Procedure 461
Ignition chamber - 463 . . .Secondary combustion zone - 463 . . .Stack design - 463. . .
Crematory design - 464 . . Incinerator design configuration - 464
Standards for Construction 465
Stack Emissions 466
Operation 466
Illustrative Problem 468
DEBONDING OF BRAKE SHOES AND RECLAMATION OF ELECTRICAL EQUIPMENT
WINDINGS 471
Debonding of Brake Shoes 471
Reclamation of Electrical Equipment Winding 472
The Air Pollution Problem 47Z
Air Pollution Control Equipment 473
Primary Ignition Chamber 473
Secondary Combustion Chamber 474
Stack 474
Emissions 474
Typical Reclamation Equipment 474
Standards for Construction 476
Illustrative Problem 476
DRUM RECLAMATION FURNACES 481
Introduction 481
Description of the Furnace Charge 482
Description of the Process 482
The Air Pollution Problem 482
Air Pollution Control Equipment 482
Primary Ignition Chamber, Batch Type 482
Primary Ignition Chamber, Continuous Type 483
Afterburner (Secondary Combustion Chamber) 486
Draft 486
Standards for Construction 486
Operation 487
Illustrative Problem 487
WIRE RECLAMATION 495
Description of the Process 495
Description of the Charge 496
The Air Pollution Problem 496
Air Pollution Control Equipment 496
Primary Ignition Chamber 496
Secondary Combustion 497
Emissions 497
Draft 498
Equipment Arrangement 498
General Construction 499
-------
CONTENTS
Refractories. 499
Charge Door 499
Combustion Air Ports 500
Gas Burners 500
Operation 500
Illustrative Problem 500
CHAPTER 9. COMBUSTION EQUIPMENT
GASEOUS AND LIQUID FUELS 507
Introduction 507
Gaseous Fuels 507
Oil Fuels 508
The Air Pollution Problem 509
Black Smoke 509
White Smoke 509
Particulate Emissions 510
Sulfur in Fuels 510
Sulfur Oxides 511
Oxides of Nitrogen 511
Air Pollution Control Methods • 511
Prohibitions Against Sulfur Emissions 511
Removal of Sulfur and Ash From Fuels 512
Illustrative Problem 512
GAS AND OIL BURNERS 514
Introduction 514
Draft Requirements 514
Gas Burners 514
Partially Aerated Burners 515
Multiple-Port Gas Burners 516
Forced-Draft Gas Burners 516
Gas Flow Rates 517
Oil Burners 517
Viscosity and Oil Preheaters 520
The Air Pollution Problem 521
Smoke and Unburned Contaminants 521
Ash and Sulfur Oxides 525
Oxides of Nitrogen 525
Air Pollution Control Equipment 525
BOILERS, HEATERS, AND STEAM GENERATORS 525
Introduction 525
Industrial Boilers and Water Heaters 525
Power Plant Steam Generators 526
Refinery Heaters 528
Hot Oil Heaters and Boilers 528
Fireboxes 529
Soot Blowing 531
The Air Pollution Problem 532
Solid Particulate Emission During Normal Oil Firing 534
Soot-Blowing Particulates 535
Sulfur Dioxide 535
Sulfur Trioxide 536
Excessive Visible Emissions 537
Oxides of Nitrogen 539
Estimating NOX Emissions 542
Air Pollution Control Equipment 543
Sulfur Compounds 544
Combustible Particulates 544
Soot Collectors 544
Sulfur Oxides Collection 545
Scrubbers for Sulfur Oxides 545
Baghouses and Precipitators 547
-------
CONTENTS
Alkaline Additives to Neutralize Sulfur Trioxide 548
Other Metal Oxides for Sulfur Dioxide Removal 549
Baghouses With Dolomite Addition for Sulfur Trioxide Removal 549
Electrical Precipitators With Additives 552
Carbon Adsorption of Sulfur Oxides 552
Oxidation of Sulfur Dioxide 553
Inhibiting Sulfur Trioxide Formation at Reduced Oxygen 553
Controlling Oxides of Nitrogen 554
Two-Stage Combustion 554
Corner-Fired Steam Generators 555
Lowering Excess Air 556
Eliminating Air Preheat 556
Other Means of Lowering Flame Temperature 557
Catalytic Decomposition of NO 557
Scrubbing NOX 558
Adsorption of NOX 558
CHAPTER 10. PETROLEUM EQUIPMENT
GENERAL INTRODUCTION 561
Crude Oil Production 561
Refining 561
Flares and Blowdown Systems 561
Pressure Relief Valves 561
Storage Vessels 561
Bulk-Loading Facilities 562
Catalyst Regenerators 562
Effluent-Waste Disposal 564
Pumps and Compressors 564
Air-Blowing Operations 564
Pipeline Valves and Flanges, Blind Changing, Process Drains 564
Cooling Towers 564
Vacuum Jets and Barometric Condensers 564
Effective Air Pollution Control Measures 565
Marketing 565
WASTE-GAS DISPOSAL SYSTEMS 565
Introduction 565
Design of Pressure Relief System 568
Safety Valves 569
Rupture Discs 570
Sizing rupture discs - 572 . . .Sizing liquid safety valves - 573 . . .
Sizing vapor and gas relief and safety valves - 574 . . .
Installing relief and safety valves and rupture discs - 575. . . Knockout vessels - 576 . . .
Sizing a blowdown line - 578
The Air Pollution Problem 584
Smoke From Flares 584
Other Air Contaminants From Flares , 584
Air Pollution Control Equipment 585
Types of Flares 585
Elevated flares - 585 . . .Ground level flares - 589 . . .Effect of steam injection - 593. . .
Design of a smokeless flare - 593 . . .Pilot ignition system - 595. . .
Instrumentation and control of steam and gas - 596 . . .Supply and control of steam - 598 . .
Design of water-injection-type ground flares - 603 . . .
Design of venturi-type ground flares - 604. . .Maintenance of flares - 606
STORAGE VESSELS 606
Types of Storage Vessels 606
Pressure Tanks and Fixed-Roof Tanks 606
Floating-Roof Tanks 607
Conservation Tanks 608
Open-Top Tanks, Reservoirs, Pits, and Ponds 611
The Air Pollution Problem 611
Factors Affecting Hydrocarbon Vapor Emissions 611
-------
CONTENTS
Hydrocarbon Emissions From Floating-Roof Tanks 612
Withdrawal emissions - 614 . . Application of results - 614
Hydrocarbon Emissions From Low-Pressure Tanks 614
Hydrocarbon Emissions From Fixed-Roof Tanks 618
Aerosol Emissions 622
Odors 622
Air Pollution Control Equipment 623
Seals for Floating-Roof Tanks 624
Floating Plastic Blankets 624
Plastic Microspheres 625
Vapor Balance Systems 627
Vapor Recovery Systems 627
Miscellaneous Control Measures 628
Masking Agents 629
Costs of Storage Vessels 629
LOADING FACILITIES 629
Introduction 629
Loading Racks 632
Marine Terminals 632
Loading Arm Assemblies 632
The Air Pollution Problem 633
Air Pollution Control Equipment 635
Types of Vapor Collection Devices for Overhead Loading 635
Collection of Vapors From Bottom Loading 638
Factors Affecting Design of Vapor Collection Apparatus 639
Methods of Vapor Disposal 640
CATALYST REGENERATION 642
Types of Catalysts 642
Loss of Catalyst Activity 644
Regeneration Processes 644
FCC Catalyst Regenerators 644
TCC Catalyst Regenerators 645
Catalyst Regeneration in Catalytic Reformer Units 645
The Air Pollution Problem 646
Air Pollution Control Equipment 647
Wet- and Dry-Type, Centrifugal Dust Collectors 647
Electrical Precipitators 648
Carbon Monoxide Waste-Heat Boilers 650
Economic Considerations 651
OIL-WATER EFFLUENT SYSTEMS 652
Functions of Systems 652
Handling of Crude-Oil Production Effluents 652
Handling of Refinery Effluents 653
Treatment of Effluents by Oil-Water Separators 653
Clarification of Final-Effluent Water Streams 653
Effluent Wastes From Marketing Operations 654
The Air Pollution Problem 654
Air Pollution Control Equipment 655
Hydrocarbons From Oil-Water Separators 655
Treatment of Refinery Liquid Wastes at Their Source 657
Oil-in-water emulsions - 657. . . Sulfur-bearing waters - 657. . .Acid sludge - 658 . . .
Spent caustic wastes - 659
PUMPS 659
Types of Pumps 659
Positive-Displacement Pumps 660
Centrifugal Pumps 660
The Air Pollution Problem 661
Air Pollution Control Equipment 661
Results of Study to Measure Losses From Pumps 664
AIRBLOWN ASPHALT 665
Recovery of Asphalt From Crude Oil 666
-------
CONTENTS xxv
Ai.rblowi.ng of Asphalt 666
The Air Pollution Problem 667
Air Pollution Control Equipment 667
VALVES 669
Types of Valves 669
Manual and Automatic Flow Control Valves 669
Pressure Relief and Safety Valves 670
The Air Pollution Problem 670
Total Emissions From Valves 671
Air Pollution Control Equipment 671
COOLING TOWERS 672
Characteristics of Cooling Tower Operation 673
The Air Pollution Problem 674
Air Pollution Control Equipment 675
MISCELLANEOUS SOURCES 675
Airblowing 675
Blind Changing 675
Equipment Turnarounds 676
Tank Cleaning 677
Use of Vacuum Jets 677
Use of Compressor Engine Exhausts 677
CHAPTER 11. CHEMICAL PROCESSING EQUIPMENT
RESIN KETTLES 681
Types of Resins 681
Phenolic Resins 681
Amino Resins 682
Polyester and Alkyd Resins 682
Polyurethane 683
Thermoplastic Resins 683
Polyvinyl Resins 683
Polystyrene 684
Petroleum and Coal Tar Resins 684
Resin-Manufacturing Equipment 684
The Air Pollution Problem 685
Air Pollution Control Equipment 686
VARNISH COOKERS 688
Introduction 688
Raw Materials for Varnish Making 689
Major Types of Manufacturing Equipment 690
Variations in Varnish Formulation 691
The Air Pollution Problem 691
Hooding and Ventilation Requirements 692
Air Pollution Control Equipment 692
Scrubbers 692
Adsorbers 693
Afterburners 693
SULFURIC ACID MANUFACTURING 695
Contact Process 695
The Air Pollution Problem 697
Air Pollution Control Equipment 698
Sulfur Dioxide Removal 698
Acid Mist Removal 698
Electrical precipitators - 698. . .Packed-bed separators - 699 . . .
Wire mesh mist eliminators - 700. . .Ceramic filters - 700 . , .Sonic agglomeration - 701. .
Miscellaneous devices - 701
PHOSPHORIC ACID MANUFACTURING 701
Phosphoric Acid Process 701
The Air Pollution Problem 702
Hooding and Ventilation Requirements 702
Air Pollution Control Equipment 703
-------
CONTENTS
PAINT-BAKING OVENS 704
Bake Oven Equipment 705
The Air Pollution Problem 706
Hooding and Ventilation Requirements 707
Air Pollution Control Equipment 708
Cost of Direct-Flame Afterburners 708
Illustrative Problem 708
SOAPS AND SYNTHETIC DETERGENTS 716
Soaps 716
Raw Materials 716
Fatty Acid Production 716
Soap Manufacture 717
Soap Finishing 717
Synthetic Detergents 718
The Air Pollution Problem 718
Soaps 718
Detergents 719
Air Pollution Control Equipment 719
Soaps 719
Detergents 720
GLASS MANUFACTURE 720
Types of Glass 720
Glass-Manufacturing Process 721
Handling, Mixing, and Storage Systems for Raw Materials 722
The Air Pollution Problem 723
Hooding and Ventilation Requirements 723
Air Pollution Control Equipment 724
Glass-Melting Furnaces 724
Continuous Soda-Lime Glass Furnaces 724
The Air Pollution Problem 726
Source test data - 727 . . .Opacity of stack emissions - 727
Hooding and Ventilation Requirements 728
Air Pollution Control Methods 729
Control of raw materials - 73 L . . Batch preparation - 732 . . .Checkers - 732 . . .
Preheaters - 733 . . Refractories and insulation - 733 . . Combustion of fuel - 734 . . .
Electric melting - 735. . . Baghouses and centrifugal scrubbers - 735
Glass-Forming Machines 736
The Air Pollution Problem 737
Air Pollution Control Methods 737
FRIT SMELTERS 738
Introduction 738
Raw Materials 738
Types of Smelters 738
Frit Manufacturing 740
Application, Firing, and Uses of Enamels 742
The Air Pollution Problem 743
Hooding and Ventilation Requirements 743
Air Pollution Control Equipment 744
FOOD PROCESSING EQUIPMENT 746
Coffee Processing 746
Batch Roasting 747
An Integrated Coffee Plant 747
The Air Pollution Problem 747
Hooding and Ventilation Requirements 749
Air Pollution Control Equipment 749
Smokehouses 750
The Smoking Process 750
Atmospheric Smokehouses 750
Recirculating Smokehouses 750
The Air Pollution Problem 751
Hooding and Ventilation Requirements 751
-------
CONTENTS xxvii
Bypassing control devices during nonsmoking periods - 752
Air Pollution Control Equipment <• 752
Afterburners - 752 . . Electrical precipitators - 752 • • Electrical precipitation versus
incineration - 753 . . .Why not immersion? - 754. • -Smoking through electrical precipitation-754
Deep Fat Frying • • 755
Batch or Continuous Operation 755
The Air Pollution Problem • 756
Hooding and Ventilation Requirements • 756
Air Pollution Control Equipment • 756
Oil collection - 757
Livestock Slaughtering 757
The Air Pollution Problem 757
Air Pollution Control Equipment • 757
Edible-Lard and Tallow Rendering 758
Dry Rendering 759
Low-Temperature, Continuous Rendering 759
Wet Rendering 759
The Air Pollution Problem 760
Hooding and Ventilation Requirements 760
Air Pollution Control Equipment 760
FISH CANNERIES AND FISH REDUCTION PLANTS 760
Wet-Fish Canning 761
Tuna Canning 761
Cannery Byproducts 762
Fish Meal Production 762
Fish Solubles and Fish Oil Production 763
Digestion Processes 765
The Air Pollution Problem 765
Odors From Meal Driers 765
Smoke From Driers 766
Dust From Driers and Conveyors 766
Odors From Reduction Cookers 766
Odors From Digesters 767
Odors From Evaporators 767
Odors From Edibles Cookers 767
Hooding and Ventilation Requirements 767
Air Pollution Control Equipment 768
Controlling Fish Meal Driers 768
Incinerating Drier Gases = 768
Chlorinating and Scrubbing Drier Gases 768
Controlling Reduction Cookers 770
Controlling Digesters 770
Controlling Evaporators 770
Collecting Dust 770
Controlling Edible-Fish Cookers 770
REDUCTION OF INEDIBLE ANIMAL MATTER 770
Dry Rendering 772
Wet Rendering 773
Refining Rendered Products 773
Drying Blood 775
Processing Feathers 775
Rotary Air Driers 775
The Air Pollution Problem 776
Cookers as Prominent Odor Sources 777
Odors From Air Driers 777
Odors and Dust From Rendered-Product Systems 778
Grease-Processing Odors 778
Raw-Materials Odors 778
Hooding and Ventilation Requirements 778
Emission Rates From Cookers 778
Emission Rates From Driers 779
-------
CONTENTS
Air Pollution Control Equipment 780
Controlling High-Moisture Streams 780
Subcooling Condensate 780
Condenser Tube Materials 781
Interceptors in Cooker Vent Lines 781
Vapor Incinerator 781
Condensation-Incineration Systems 782
Carbon Adsorption of Odors 783
Odor Scrubbers 783
Odor Masking and Counteraction 784
ELECTROPLATING 784
The Air Pollution Problem 785
Hooding and Ventilating Requirements 785
Air Pollution Control Equipment 786
Scrubbers 786
Mist Inhibitors 787
INSECTICIDE MANUFACTURE 787
Methods of Production 787
Solid-Insecticide Production Methods 787
Liquid-Insecticide Production Methods 791
The Air Pollution Problem 791
Hooding and Ventilation Requirements 791
Air Pollution Control Equipment 791
HAZARDOUS RADIOACTIVE MATERIAL 792
Hazards in the Handling of Radioisotopes 792
The Air Pollution Problem 793
Characteristics of Solid, Radioactive Waste 793
Characteristics of Liquid, Radioactive Waste 793
Problems in Control of Airborne, Radioactive Waste 794
Hooding and Ventilation Requirements 794
Hooding 794
Ventilation 794
Air Pollution Control Equipment 794
Reduction of Radioactive, Particulate Matter at Source 794
Design of Suitable Air-Cleaning Equipment 795
Reverse-jet baghouse - 796 . . Wet collectors - 796 . . Electrical precipitators - 796 . . .
Glass fiber filters - 797 . . .Paper filters - 797
Disposal and Control of Solid, Radioactive Waste 798
Disposal and Control of Liquid, Radioactive Waste 798
OIL AND SOLVENT RE-REFINING 799
Re-refining Process for Oils 799
Re-refining Process for Organic Solvents 800
The Air Pollution Problem 800
Air Pollution From Oil Re-refining 800
Air Pollution From Solvent Re-refining 800
Air Pollution Control Equipment 800
Oil Re-refining 800
Solvent Re-refining 801
CHEMICAL MILLING 801
Description of the Process 801
Etchant Solutions 802
The Air Pollution Problem 802
Mists 803
Vapors 804
Gases 804
Solvents 804
Hooding and Ventilation Requirements 804
Air Pollution Control Equipment 804
Corrosion Problems. 805
REFERENCES 807
-------
CONTENTS xxix
APPENDIX A: RULES AND REGULATIONS 831
Rules and Regulations of the Air Pollution Control District 831
APPENDIX B: ODOR-TESTING TECHNIQUES 861
The Odor Panel 861
The Odor Evaluation Room 861
Sampling Techniques 862
Evaluation of Odor Samples 863
Determination of Odor Concentration 864
APPENDIX C: HYPOTHETICAL AVAILABLE HEATS FROM NATURAL GAS 867
APPENDIX D: MISCELLANEOUS DATA 871
-------
-------
-------
CHAPTER 1
INTRODUCTION
JOHN A. DANIELSON, Senior Air Pollution Engineer
ROBERT L. CHASS, Chief Deputy Air Pollution Control Officer
ROBERT G. LUNCHE, Director of Engineering
-------
CHAPTER 1
INTRODUCTION
The past two decades have witnessed remarkable
progress in the development of reasonable engi-
neering solutions for controlling industrial and
commercial sources of air pollution. This man-
ual presents the practical technical knowledge ac-
quired through nearly 20 years of experience by
the Engineering Division of the Los Angeles County
Air Pollution Control District. With the rich back-
ground of experience attained by government and
industry, this engineering knowledge can now be
applied to solving specific community air pollution
problems throughout the world.
THE LOS ANGELES BASIN
Los Angeles and its environs have special prob-
lems peculiar to the area. Los Angeles County is
the largest heavily industrialized, semitropical
area in the world. It comprises 4, 083 square miles
and contains more than 75 incorporated cities and
large scattered unincorporated areas. Its popula-
tionhasmore than doubled since 1939, and indus-
tryhas expanded from approximately 6,000 estab-
lishments in 1939 to more than 17,000 in 1963
(Weisburd, 1962).
Topographical and meteorological conditions ag-
gravate the effects of the pollution produced by
this population and this industry in the Los Angeles
Basin. The average wind velocity there is less
than 6 miles per hour. The light winds that do
develop are relatively ineffective in carrying off
the polluted air because of the temperature in-
versions that prevail approximately 260 days of
the year. The height of the inversion base varies
from ground level to 3, 000 feet. These inversions
have been most noticeable during the summer
months, but in the last few years extreme inver-
sions have occurred in the November-December
period as well. Their effect is to limit vertical
distribution of atmospheric pollution while local
winds from the west are moving the air over the
area during the day.
In Los Angeles County, the complex mixture of
smoke, dusts, fumes, gases, and other solid and
liquid particles is called "smog. " This smog may
produce a single effect or a combination of effects,
such as irritation of eyes, irritation of throats,
reduction of visibility, damage to vegetation, crack-
ing of rubber, local nuisances , and a host of other
effects, real and fancied.
Any community suffering from an air pollution
problem must inevitably turn its attention to the
operations of industry, because these operations
have been most frequently associated with com-
munity air pollution problems. Accordingly, the
Los Angeles County control prog ram was first di-
rected to industrial operations.
Although the exact year when smog was first rec-
ognized in Los Angeles is not known, the first
public demands for relief from air pollution ap-
pear to have been made immediately after World
War II (Weisburd, 1962). Newspapers, in partic-
ular, began to expose the problem in the public in-
terest. As a consequence, air pollution control
groups were formed under health department juris -
diction--first by the city of Los Angeles, and then
by the county of Los Angeles in the unincorporated
areas. These control efforts failed, however, be-
cause of the multiplicity and inadequacy of the con-
trol jurisdictions. It was soon apparent that ade-
quate control could be exercised only by a single
authority with jurisdiction over the entire pollution
zone--the incorporated and unincorporated areas
of Los Angeles County. As a result, Assembly Bill
No. 1 was presented to the 1947 session of the
California Legislature. This Bill proposed con-
solidation of control measures. The Legislature
voted to add Chapter 2, "Air Pollution Control
District, " to Division 20 of the Health and Safety
Code relating to the control and suppression of air
pollution. Thus, the first statewide air pollution
control statute was enacted. This statute, The
California Act, is an enabling type of legislation
that legalizes the establishment of air pollution con-
trol districts on a local option basis by the coun-
ties of California.
RULES AND REGULATIONS
IN LOS ANGELES COUNTY
Under authority of the Health and Safety Code, the
Board of Supervisors of Los Angeles County en-
acted, on December 30, 1947, the first rules and
regulations guiding the conduct of the Los Angeles
County Air Pollution Control District. Additional
rules and regulations were enacted as the need a-
rose. The rules are contained in seven regula-
tions, as follows: (I) "General Provisions," (II)
"Permits," (III) "Fees ," (IV) "Prohibitions ,"
(V) "Procedure Before the Hearing Board, " (VI)
"Orchard Heaters, " (VII) "Emergencies."
The rules and regulations are extensive and are
shown in Appendix A for those desiring detailed
-------
INTRODUCTION
information. In using this manual, the reader
should be aware of certain provisions of these stat-
utes, and these will be summarized in this chap-
ter. Of most importance to the reader are Regu-
lations II "Permits" and IV "Prohibitions."
or denied the construction or operation of thou-
sands of industrial and commercial enterprises
in Los Angeles County. These 15 years of ex-
perience provide the background for much of the
data in this manual.
REGULATION II: PERMITS
The permit system of the Los Angeles County Air
Pollution Control District is one of the most im-
portant features of the air pollution control pro-
gram. A diagram of the permit system and how
it operates is given in Figure 1. In general, the
system requires owners, operators, or lessees
to apply for permits to construct and operate any
equipment capable of emitting air contaminants.
If the applicant's plans, specifications, and field
tests show that the equipment can operate within
the limits allowed by law, then a permit is granted.
If the equipment is capable of emitting contami-
nants that create a public nuisance or violate any
section of the State Health or Safety Code or the
rules and regulations of the Air Pollution Control
District, then a permit is denied.
This permit system is effective because it elimi-
nates use of equipment that emits excessive air
contaminants or reduces emissions to within al-
lowable limits by requiring that the design of the
equipment or of the process be modified or that
adequate control equipment be used. The con-
struction or operation of control equipment must
also be authorized by permit. Thus, the permit
system is a positive means of controlling air
pollution.
In using this regulation, the members of the En-
gineering Division of the Air Pollution Control
District have reviewed the design and approved
REGULATION IV: PROHIBITIONS
The rules in Regulation IV prohibit the emission
of certain air contaminants and regulate certain
types of equipment. Because these rules apply to
engineering problems and touch upon many sciences,
they require extraordinary care for their framing.
The prohibitions pertinent to readers of this manual
will now be discussed.
Rule 50: The Ringelman Chart
Rule 50 sets standards for reading densities and
opacities of visible emissions in determining vi-
olation of, or compliance with, the law. It lim-
its to 3 minutes in any hour the discharge, from
any single source, of any air contaminant that is
(1) as dark as or darker than that designated as
Number 2 on the Ringelmann Chart, or (2) of
such opacity as to obscure an observer's view to
a degree equal to or greater than that to which
smoke described in (1) does.
Rule 51: Nuisance
According to Rule 51, whatever tends to endan-
ger life or property or whatever effects the health
of the community is a public nuisance. The nui-
sance must, however, affect the community at
large and not merely one or a few persons. A
nuisance becomes a crime if it contributes seri-
ously to the discomfort of an area.
Enforcement iiiiiiiini
SUBMISSION Of CONSTRUCTION PLANS
AUTHORITY, TO CONSTRUCT ISSUED
PERMIT TO OPERATE GRANTED
INSPECT ON OF EQUIPMENT
A Cont
uinq Process
REVOCAT ON OF PERMIT
Or act
court
defects
n in criminaT or civil
f inspection discloses
or improper operation.
PERMIT TO OPERATE DENIED
APPEAL OF DENIAL
To hea r i nq board .
AUTHORITY TO CONSTRUCT DENIED
APPEAL OF DENIAL
To hean nq board, or
new pians submitted.
PETITION FOR VARIANCE
Submitted to hearing board to permit
operation for lirrited time while control
equipment is developed or installed.
Figure 1. The permit system and how it operates when industry seeks to install equipment that may
pol lute the ai r.
-------
Rules and Regulations
Rule 52: Particulate Matter
Rule 52 establishes the maximum allowable lim-
its for the discharge of particulate matter. It
limits the discharge of this contaminant from any
source to a maximum concentration of 0. 4 grain
per cubic foot of gas at standard conditions of 60° F
and 14. 7 psia. This rule does not, however, apply
when the particulate matter is a combustion
contaminant. *
Rule 53: Specific Contaminants
Rule 53 establishes the maximum allowable lim-
its for the discharge of sulfur compounds and com-
bustion contaminants as follows:
Rule 53a: Sulfur compounds calculated as sulfur
dioxide (802): 0.2 percent, by volume.
Rule 53b: Combustion contaminants: 0. 3 grain
per cubic foot of gas calculated to 12 percent
of carbon dioxide (CO;?) at standard conditions.
In measuring the combustion contaminants from
incinerators used to dispose of combustible ref-
use by burning, the CC>2 produced by combus-
tion of any liquid or gaseous fuels shall be ex-
cluded from the calculation to 12 percent of CO^.
Rule 53.1: Scavenger Plants
Rule 53. 1 sets forth the conditions under which
a. scavenger or recovery plant may operate un-
der permit. These plants are built in Los Angeles
County to recover sulfur products, which might
otherwise be emitted to the air.
Rule 56: Storage of Petroleum Products
Rule 56 sets forth the type of equipment that can
be used for the control of hydrocarbons arising
from the storage of gasoline and certain petrole-
um distillates. Rule 56 provides that any tank of
nore than 40, 000-gallon capacity used for storing
gasoline or any petroleum distillate having a vapor
pressure of 1-1 /2 psia or greater must be equipped
with a vapor loss control device such as a pontoon-
type or double-deck-type floating roof or a vapor
recovery system capable of collecting all emissions.
Rules 57 and 58: Open Fires and Incinerators
Rules 57 and 58 ban the burning of combustible
refuse in open fires and single-chamber incinera-
tors in the Los Angeles Basin.
Rule 59: Oil-Effluent Water Separators
Rule 59 regulates the type of equipment that can
be used for the control of hydrocarbons from oil-
water separators. Itprovides that this equipment
must either be covered, or provided with a float-
ing roof, or equipped with a vapor recovery sys-
tem, or fitted with other equipment of equal effi-
ciency if the effluent handled by the separator con-
tains a minimum of 200 or more gallons of petro-
leum products per day.
Rule 54: Dust and Fumes
Rule 54 establishes the maximum allowable lim-
its for the discharge of dusts and fumes according
to the process weightst of materials processed
per hour. The maximum allowable weight in
pounds per hour is graduated according to the
weights of materials processed per hour. The
maximsam emission allowed is 40 pounds per hour
where 60,000 or more pounds are processed in
the equipment in any given hour.
*Particulate matter is any material, except uncombined water,
that exists in a finely divided form as a liquid or solid at
standard conditions. A combustion contaminant is particulate
matter discharged into the atmosphere from the burning of any
kind of material containing carbon in a free or combined state.
Rule 61: Gasoline Loading Into Tank Trucks and Trailers
and I rollers
Rule 61 sets forth the type of control equipment
thatcan be used for the control of hydrocarbons
resulting from the loading of gasoline into tank
trucks. It provides for the installation of vapor
collection and disposal systems on bulk gasoline-
loading facilities where more than 20,000 gallons
of gasoline are loaded per day and requires that
the loading facilities be equipped with a vapor col-
lection and disposal system. The disposal system
employed musthave a minimum recovery efficiency
of 90 percent or a variable vapor space tank com-
pressor and fuel gas system of such capacity as
to handle all vapors and gases displaced from the
trucks being loaded.
tProcess weight is the total weight of all materials intro-
duced into any specific process that is capable of causing
any discharge into the atmosphere. Solid fuels charged are
considered part of the process weight, but liquid and gas-
eous fuels and combustion air are not. The ''process weight
per hour" is derived by dividing the total process weight
by the number of hours in one complete operation from the
beginning of any given process to the completion thereof, ex-
cluding any time during which the equipment is idle.
Rules 62 and 62.,. Sulfur Content of Fuels
Rules 62 and 62. 1 prohibit the burning in the Los
Angeles Basin of any gaseous fuel containing sul-
fur compounds in excess of 50 grains per 100 cubic
feet of gaseous fuel (calculated as hydrogen sulfide
-------
INTRODUCTION
at standard conditions) or any liquid or solid fuel
having a sulfur content in exces s of 0. 5 percent by
weight, except when natural gas or low-sulfur fuels
are not available.
Rule 63: Gasoline Specifications
Rule 63 prohibits the sale and use of fuel for motor
vehicles having a degree of unsaturation exceeding
a bromine number of 30.
Rule 64: Reduction of Animal Matter
Rule 64 requires thatmalodors -from all equipment
used for reduction of animal matter either be in-
cinerated at temperatures of not less than 1,200°F
for a period of not less than 0. 3 second or pro-
cessed in an odor-free manner under conditions
stated in the rule.
Rule 65:* Gasoline Loading into Tanks
Rule 65 prohibits the loading of gasoline into any
stationary tank with a capacity of 250 gallons or
more from any tank truck or trailer, except
through a permanent submerged fill pipe, unless
such tank is equipped with a vapor loss control
device as described in Rule 56, or is a pressure
tank as described in Rule 56.
Rule 66* Organic Solvents
Rule 66 requires that photochemically reactive
organic solvent emissions in excess of 40 pounds
per day (or 15 pounds per day from processes
involving contact with a flame or baking, heat-
curing or heat polymerizing, in the presence of
oxygen) shall not be emitted unless controlled by
incineration, adsorption, or in an equally effi-
cient manner.
Rule 66.1: Architectural Coatings
Rule 66. 1 prohibits the sale of architectural coat-
ings containing photochemically reactive solvents
in containers larger than 1 -quart capacity. It
also prohibits diluting any architectural coating
with a photochemically reactive solvent.
Rule 66.2: Disposal and Evaporation of Solvents
Rule 66. 2 prohibits the disposal of more than
1-1/2 gallons per day of any photochemically re-
active solvent by any means that will permit
the evaporation of such solvent into the atmo-
sphere.
ROLE OF THE AIR POLLUTION ENGINEER
Clearly, as indicated by this impressive list of
prohibitions, the rules and regulations affect the
operation of nearly every industry in Los Angeles
County. Through their enforcement, controls have
been applied to such diverse sources and opera-
tions as incinerators, open fires, rendering cook-
ers, coffee roasters, petroleum refineries, chem-
ical plants, rock crushers, and asphalt plants.
From the smelting of metals to the painting of man-
ufactured goods , industrial and commercial oper-
ations have been brought within the scope of the
control program. This control has been accom-
plished through the use of the permit system.
ACCOMPLISHMENTS OF THE PERMIT SYSTEM
Under the permit system every source capable of
emitting air contaminants and constructed since
February 1948 has needed an authorization to be
constructed and a permit to be operated from the
Engineering Division of the Air Pollution Control
District. From April 1948 throughDecember 1963,
56, 502 permits were issued by District engineers.
The estimated value of the basic equipment was
$651, 447, 000 and that of the control equipment was
an additional $107,507,000. During this same
period, 5, 075 applications for basic and control
equipment were denied.
This wealth of engineering experience is reflected
in the contents of this manual. Nearly all the data
presented were acquired through the experience
and work of the District's Air Pollution engineers
and of engineers in industry. Their pioneering ef-
forts to stay at least one pace ahead of the problem
have produced many engineering firsts in the con-
trol of air pollution.
USE OF THIS MANUAL
*Rules 65 and 66 were adopted just prior to the publication of
this manual. Consequently, control of equipment from sources
involving these specific contaminants has not been discussed tc
the degree specified by these rules.
Users of this manual should remember that the
degree of air pollution control discussed herein is
based upon the prohibitions as set forth by the rules
and regulations of the Los Angeles County Air Pol-
lution Control District. In many areas, air pollu-
tion regulations are less stringent, and control
devices of lower efficiency may be permitted.
-------
Use of This Manual
GENERAL DESIGN PROBLEMS
This manual consists of 11 chapters and 4 appen-
dixes. Chapters 2 through 5 present general de-
sign problems confronting air pollution engineers
in the development of air pollution control sys-
tems. Specifically, chapter Z describes the types
of air contaminants encountered and chapter 3 pre-
sents design problems of hoods and exhaust sys-
tems. Types of control devices, and their gen-
eral design features are discussed in chapters 4
and 5.
SPECIFIC AIR POLLUTION SOURCES
Chapters 6 through 11 discuss the control of air
pollution from specific sources. Each solution of
an air pollution problem represents a separate
section of the text. Many processes are discussed:
Metallurgical and mechanical processes, process-
es of incineration and combustion, and processes
associated with petroleum and chemical equipment,
each in a separate chapter and in that order. Usu-
ally the process is described and then the air pol-
lution problem associated with it is discussed, to-
gether with the characteristics of the air contami-
nants and the unique design features of the air pol-
lution control equipment.
Bythis arrangement, the reader can, if he wishes,
refer only to that section discussing the specific
process in which he is interested. If he wants to
know more about the general design features of the
air pollution control device serving that process,
he can refer to chapters 4 and 5 on control equip-
ment.
-------
CHAPTER 2
AIR CONTAMINANTS
JA.NET DICKINSON, Senior Air Pollution Analyst
ROBERT L. CHASS, Chief Deputy Air Pollution Control Officer
W. J. HAMMING, Chief Air Pollution Analyst
-------
CHAPTER 2
AIR CONTAMINANTS
INTRODUCTION
The purpose of this chapter is to describe briefly
the parameters of an air pollution problem, par-
ticularly the problem of Los Angeles County; the
measures taken to eliminate the problem; and those
still needed. Other chapters will delineate, in
detail, the methods and equipment successfully
used in the control of emissions of air contami-
nants from a variety of stationary sources.
The control program in the County of Los Angeles
during the past 15 years has been the most effec-
tive ever attempted anywhere. During the same
period, however, the county has had a phenomenal
population explosion that has caused the emissions
from motor vehicles to overtake and surpass the
gains made by control over stationary sources.
The net effect has been the more frequent occur-
rence of smog symptoms over an increasingly larg-
er area.
Since control over motor vehicle emissions is the
responsibility of the state rather than of the local
agency, substantial improvement in the situation
in Los Angeles will probably have to await the suc-
cessful accomplishment of the state's program.
In the interim, however, the Air Pollution Con-
trol District of Los Angeles County will continue
its efforts to reduce emissions of air contaminants
from stationary sources wherever possible within
its jurisdiction.
FACTORS IN AIR POLLUTION PROBLEMS
Literally, any substance not normally present in
the atmosphere , or measured there in greater than
normal concentrations, should be considered an
air contaminant. More practically, however, a
substance is not so labeled until its presence and
concentration produce or contribute to the produc-
tion of some deleterious effect.
Most foreign substances find their way into the at-
mosphere as the result of some human activity.
Under normal circumstances , they diffuse through-
out a rather large volume of air and do not accu-
mulate to potentially harmful concentrations.
Under less favorable conditions , however, the air
volume available for this diffusion becomes inade-
quate and materials dispersed in it concentrate
until an air pollution problem is created.
Air pollution problems may exist over a small
area as a result of just one emission source or
group of sources or they may be widespread and
cover a whole community or urban complex in-
volving a variety of sources. The effects thatcause
the situation to be regarded as a problem may be
limited in scope and associated with a single kind
of contaminant or they may be the variable results
of complexatmospheric interaction of a number of
contaminants.
The factors that contribute to the creation of an
air pollution problem are both natural and man
made. The natural factors are primarily mete-
orological, sometimes geographical, and are gen-
erally beyond man's sphere of control, whereas
the manmade factors involve the emission of air
contaminants in quantities sufficient to produce
deleterious effects and are within man's sphere
of control. The natural factors that restrict the
normal dilution of contaminant emissions include:
Temperature inversions, which prevent diffusion
upwards; very low wind speeds, which do little to
move emitted substances away from their points
of origin; and geographical terrain, which causes
the flow to follow certain patterns and carry from
one area to another whatever the air contains. The
manmade factors involve the contaminant emis-
sions resulting from some human activity.
The predominant kind of air pollution problem in-
volves simply the overloading of the atmosphere
with harmful or unpleasant materials. This is
the problem usually associated with an industrial
area, and the type that has been responsible for
all the killer air pollution incidents of the past.
Itis also the type of problem most readily solved,
if the need and desire to do so are great enough.
Contaminants frequently associated with this kind
of problem include: Sulfur compounds (sulfur
oxides, sulfates, sulfides, mercaptans); fluorides;
metallic oxides; odors; smoke; and all types of
dusts and fumes. The harmful effects may be such
as to cause illness and death to persons and ani-
mals, damage to vegetation, or just annoyance and
displeasure to persons in affected areas.
During the past 20 years, however, another kind
of air pollution problem has evolved--that pro-
duced by the photochemical reaction of organic
chemicals and oxides of nitrogen in the presence
of sunlight. The effects of this type of air pollu-
tion were first noted in the Los Angeles area in
the mid-1940"s , butthe cause was not then known,
11
-------
12
AIR .CONTAMINANTS
nor was there any apparent relationship between
the initial effects and air pollution.
The effect first noted was damage to vegetation.
This was followed by irritation of the eyes , marked
reduction in visibility not related to unusual quan-
tities of atmospheric moisture or dust, and later,
unexplained acceleration of the aging of rubber
products as evidenced by cracking. Careful re-
search over nearly 10 years demonstrated that all
these effects were produced by the reaction in the
atmosphere of organic compounds , principally hy-
drocarbons, and nitrogen dioxide, and that the
cracking of rubber products was caused specif-
ically by one of the products of these reactions,
ozone.
Initially, only a relatively small portion of the Los
Angeles Basin was affected, but a tremendous in-
flux of new residents and new industrial growth
during the past 15 years have caused a continuous
enlargementof the affected area. Within this area,
certain local problems related to single sources
and groups of sources also exist, but they are of
less significance than the over all problem of photo-
chemical smog.
TYPES OF AIR CONTAMINANTS
Substances considered air contaminants in Los
Angeles County fall into three classes on the basis
of their chemical composition and physical state.
Theseare (1) organic gases, (2) inorganic gases,
and (3) aerosols. Each class may include many
different compounds, emanate from several dif-
ferent sources, and contribute to the production
of a number of characteristic smog effects. A
brief summary of some of the contaminants, their
principal sources, and their significance is pre-
sented in Table 1.
ORGANIC GASES
The first group, organic gases, consists entirely
of compounds of carbon and hydrogen and their
derivatives. These include all classes of hydro-
carbons (olefins, paraffins, and aromatics) and
the compounds formed when some of the hydrogen
in the original compounds is replaced by oxygen,
halogens, nitro or other substituent groups. The
latter are the hydrocarbon derivatives.
The principal origin of hydrocarbons is petroleum,
and the principal sources of emissions of hydro-
carbons and their derivatives are those related to
the processing and use of petroleum and its prod-
ucts. Hydrocarbons are released to the atmo-
sphere during the refining of petroleum, during
the transfer and storage of petroleum products ,
and during the use of products such as fuels, lu-
bricants, and solvents. Derivatives of hydro-
carbons can also be released into the atmosphere
in connectionwith these processes and in connec-
tion with their manufacture and use. They can
evenbe formed in the atmosphere as the result of
certain photochemical reactions.
Current Sources in Los Angeles County
More specifically, the principal current sources
of organic gases in Los Angeles County are listed
in Table 1.
Hydrocarbons
The most important source, by far, of emission
of hydrocarbons is the use of gasoline for the oper-
ation of 3-1 / Z million mo tor vehicles. This source
alone accounts for approximately 1,930 tons per
day, or 70 percent of the total emissions. Of this
quantity, about 73 percent is attributed to exhaust
emissions; 10 percent, to crankcase emissions;
and 17 percent, to evaporation of fuel from car-
buretors and gasoline tanks. Except for about 2
percent of the total, the balance of the hydrocarbon
emissions are divided between the petroleum in-
dustryand industrial and commercial uses of or-
ganic solvents.
Kinds of hydrocarbons contributed by these sources
vary considerably. Auto exhaust, for example, is
the principal source of olefins , though other sources
connectedwith the operation of motor vehicles and
with the processing and handling of gasoline con-
tribute in direct proportion to the olefin content
of the gasoline marketed here. All these sources
contribute paraffins and aromatics , and emissions
of hydrocarbons from solvent usage are composed
almost entirely of these two classes.
Hydrocarbon derivatives
Of the 300 tons of hydrocarbon derivatives (or sub-
stituted hydrocarbons) emitted to the atmosphere
of Los Angeles County each day, about three-fourths
results from solvent uses such as surface coating,
degreasing, and dry cleaning, and other industrial
and commercial processes. The balance is in-
cluded in the products of combustion of various
petroleum fuels and of incineration of refuse. The
substituted hydrocarbons emitted to the atmosphere
by industrial and commercial use of organic sol-
vents include oxygenates, such as aldehydes, ke-
tones, and alcohols; organic acids; and chlorinated
hydrocarbons. Mosthydrocarbon derivatives as-
sociated with surface coating are oxygenates whose
presence canbe related either to the solvent itself
or to the products of the partial oxidation involved
in the drying of the coated objects. The hydro-
carbon derivatives associatedwith degreasing and
dry cleaning are mostly chlorinated hydrocarbons.
The derivatives associatedwith combustion, either
-------
Types of Contaminants
13
Table 1. AIR CONTAMINANTS IN LOS ANGELES COUNTY, THEIR PRINCIPAL
SOURCES AND SIGNIFICANCE (JANUARY 1964)
Emitted contaminant
Organic gases
Hydr oc arbons
Paraffins
Olefins
Aroma tics
Others
Oxygenated hydrocarbons
(Aldehydes, ketones ,
alcohols, acids)
Halogenated hydrocar-
bons (Carbon tetrachlo-
ride, perchloroethylene,
etc)
Inorganic gases
Oxides of nitrogen
(Nitric oxide, nitrogen
dioxide)
Oxides of sulfur
(Sulfur dioxide, sulfur
trioxide)
Carbon monoxide
Aerosols
Solid particles
Carbon or soot particles
Metal oxides and salts
Silicates and mineral
dusts
Metallic fumes
Liquid particles
Acid droplets
Oily or tarry droplets
Paints and surface
coatings
Principal sources
Processing and transfer of
petroleum products; use of
solvents; motor vehicles
Processing and transfer of
gasoline; motor vehicles
Same as for paraffins
Use of solvents; motor
vehicles
Use of solvents
Combustion of fuels; motor
vehicles
Combustion of fuels; chem-
ical industry
Motor vehicles; petroleum
refining; metals industry;
piston-driven aircraft
Combustion of fuels; motor
vehicles
Catalyst dusts from re-
fineries; motor vehicle
exhaust; combustion of
fuel oil; metals industry
Minerals industry; con-
struction
Metals industry
Combustion of fuels ;plating;
battery manufacture
Motor vehicles; asphalt
paving and roofing; asphalt
saturators; petroleum re-
fining
Various industries
Significant effects
Plant damage
X
X
(Atypical)
X
X
(Specific type)
Eye
irritation
X
X
X
X
X
X
X
Oxidant
formation
X
X
X
X
X
X
Visibility
reduction
X
X
X
X
X
X
X
X
X
X
X
X
X
Danger
to
health
X
X
X
(Occas-
sionally]
X
(Under
special
circum-
stances)
Other
Odors
Odors
Property
damage
-------
14
AIR CONTAMINANTS
of fuels or of refuse, are products of incomplete
combustion and are almost entirely oxygenates.
Thus, the composition of atmospheric emissions
of hydrocarbon derivatives is currently about one-
fourth to one-third chlorinated hydrocarbons and
two-thirds to three-fourths oxygenates.
In addition to the hydrocarbon derivatives contrib-
ted by direct emissions, the atmosphere of a photo-
chemical smog contains similar compounds formed
there as a result of the reactions that produced
the smog. These substitutedhydrocarbons include
oxygenates, such as aldehydes , ketones, alcohols,
and organic acids, and nitrogen-containing com-
pounds, suchasthe peroxyacyl nitrates and, per-
haps, nitro olefins. These compounds are the
products of partial oxidation of hydrocarbons and
some derivatives in the atmosphere and of atmo-
spheric reactions between oxides of nitrogen and
organic gases.
Significance in Air Pollution Problem
Hydrocarbons and their derivatives are important
factors in the air pollution problem in Los Angeles
County because of their ability to participate in the
atmospheric reactions thatproduce effects associ-
ated with photochemical smog. The most reactive
group, the olefins (unsaturated hydrocarbons),
can react with-nitrogen dioxide to produce plant
damage, eye irritation, visibility-reducing aero-
sols, and oxidants or ozone. Paraffins (saturated
hydrocarbons) can also reactwith nitrogen dioxide
to produce all these effects except plant damage.
Aromatic hydrocarbons, particularly those having
various substituent groups, can react with nitro-
gen dioxide to produce a type of plant damage dif-
ferent from that usually associated with smog and
produce all the other effects as well.
The hydrocarbon derivatives, particularly the al-
dehydes and ketones, and even some of the chlori-
nated hydrocarbons, can also react with nitrogen
dioxide in the atmosphere to produce eye irritation,
aerosols, and ozone. Further, some of the alde-
hydes and nitro derivatives are, themselves, lach-
rymators and some of the chlorinated hydrocarbons
are rather toxic. Except for the peroxyacyl nitrates ,
these compounds are not, however, generally as-
sociated with production of plant damage.
The hydrocarbons are further indicted because
photochemical reactions in which they participate
sometimes produce hydrocarbon derivatives such
as aldehydes, ketones, and nitro-substituted or-
ganics, which can in turn react to increase the
production of smog effects.
INORGANIC GASES
Inorganic gases constitute the second major group
of air contaminants in Los Angeles County. They
include oxides of nitrogen, oxides of sulfur, car-
bon monoxide, and much smaller quantities of am-
monia, hydrogen sulfide, and chlorine.
The principal source of all the oxides listed above
is the combustion of fuel for industrial, commer-
cial, and domestic uses; for transportation; for
space heating; and for generation of power. Ad-
ditionally, small quantities of sulfur oxides and
carbon monoxide, and the total of the minor con-
stituents, ammonia, hydrogen sulfide, and chlo-
rine, are emitted in connection with certain indus-
trial processes.
Current Sources in Los Angeles County
The principal sources currently responsible for
atmospheric emissions of each of the important
inorganic gaseous air contaminants will now be
discussed.
Oxides of nitrogen
A number of compounds must be classified as
oxides of nitrogen, but only two, nitric oxide (NO)
and nitrogen dioxide (NO2), are important as air
contaminants. The first, nitric oxide, is formed
through the direct combination of nitrogen and oxy-
gen from the air in the intense heat of any com-
bustion process. Nitric oxide in the atmosphere
is then able, in the presence of sunlight, to com-
bine with additional oxygen to form nitrogen dioxide.
Usually the concentrations of nitric oxide in the
combustion effluents are at least 5 to 10 times
greater than those of nitrogen dioxide. Nonethe-
less, since every mole of nitric oxide emitted to
the atmosphere has the potential to produce a mole
of nitrogen dioxide, one may not be considered
without the other. In fact, measurement of their
concentrations often provides only a sum of the
two reported as the dioxide.
Of the total quantities of these contaminants cur-
rently being emitted each day in Los Angeles
County, approximately 60 percent, or 490 tons,
must be attributed to the exhaust effluents from
gasoline-powered motor vehicles. Almostthe en-
tire balance is produced as the result of combus-
tion of fuel for space heating and power generation.
Oxides of sulfur
Air contaminants classified as oxides of sulfur
consist essentially of only two compounds , sulfur
dioxide (SO2> and sulfur trioxide (SOj). The pri-
mary source of both is the combination of atmo-
spheric oxygen with the sulfur in certain fuels
during their combustion. The total emitted quan-
tities of these substances are, therefore, directly
related to the sulfur content and total quantities of
-------
Types of Contaminants
15
the principal fuels used in a community. Normally,
the dioxide is emitted in much greater quantities
than the trioxide, the latter being formed only
under rather unusual conditions. In fact, the tri-
oxide is normally a finely divided aerosol rather
than a gas.
In Los Angeles County during the past few years,
the average daily emissions of sulfur oxides have
exhibited a marked seasonal variation as a result
of the promulgation of Rule 62. This rule places
a limitation on the sulfur content of the fuels that
may be burned during the period April 15 to
November 15. The effect of this rule is to cause
substitution of natural gas for fuel oil as the fuel
used for generation of electric power. Since nat-
ural gas contains no sulfur, the emissions of sul-
fur oxides are thus drastically reduced during this
period. Recent appraisals of total air pollution
for Los Angeles County must take this fact into
account.
During the period April 15 to November 15, there-
fore, emissions of sulfur oxides total approxi-
mately 60 tons per day, 10 percent of which is
attributable to combustion of fuels for power gen-
eration and space heating; 50 percent, to emis-
sions from sulfur and sulfuric acid plants; and the
remainder about evenly divided between emissions
from petroleum-refining operations and emissions
from automobile exhausts. During the balance of
the year, however, nearly 80 percent of the total
of 455 tons of emissions of sulfur oxides per day
must be attributed to combustion of fuel for power
generation and space heating; 10 percent, to emis-
sions from industrial sulfur recovery operations;
and the balance about evenly divided between emis-
sions from petroleum-refining operations and auto-
mobile exhausts.
Seasonal variations in emissions of sulfur oxides
are less likely to occur in the future because of
Rule 62. 1, adopted in 1964. This rule prohibits
the burning of fuels of high sulfur content at any
time when natural gas or low-sulfur fuels are
available. In the near future, however, natural
gas may not always be available during the winter
months.
Carbon monoxide
Carbon monoxide (CO) is a single contaminant
formed during incomplete oxidation of any carbo-
naceous fuel and currently has only one significant
source in Los Angeles County — the incomplete
combustion of gasoline in motor vehicles. Of a
total of nearly 10,660 tons of this contaminant
emitted per day, 97 percent is attributable to this
source. About 1. 5 percent is attributable to the
emissions from aircraft, and the balance, from
petroleum-refining operations.
Significance in Air Pollution Problem
The importance of the inorganic gases in an air
pollution problem varies with the gas in question.
Each will, therefore, be discussed separately.
Oxides of nitrogen
The oxides of nitrogen have far greater significance
in photochemical smog than any of the other in-
organic gaseous contaminants. Researchhas dem-
onstrated that nitrogen dioxide in the presence of
sunlight will undergo reactions with a number of
organic compounds to produce all the effects as-
sociated with photochemical smog. In fact, the
presence of the dioxide has been shown to be a nec-
essary condition for these reactions. This does
not, however, diminish the need for adequate con-
sideration of nitric oxide as an air contaminant,
since this is the form in which the oxides of nitro-
gen normally enter the atmosphere. If it were
possible to prevent the oxidation of nitric oxide
to nitrogen dioxide, there would, perhaps, be
little reason to consider the oxides of nitrogen as
air contaminants, at least in Los Angeles.
I n communities not affected by photochemical smog,
the oxides of nitrogen must be considered solely
for their inherent ability to produce deleterious
effects by themselves. The only effect that must
seriously be considered in this regard is their
toxicity, though the reddish-brown color of the
dioxide and its sharp odor could cause problems
in areas near a nylon plant, nitric acid plant, or
nitrate fertilizer plant. Nitric oxide is consider-
ably less toxic than the dioxide. It acts as an
asphyxiant when in concentrations great enough to
reduce the normal oxygen supply from the air.
Nitrogen dioxide, on the other hand, in concentra-
tions of approximately 5 ppm,can produce lung in-
jury and edema, and in greater concentrations,
fatal lung damage.
The dioxide, then, is heavily indicted as an un-
desirable constituent of the atmosphere, regard-
less of the type of air pollution problem under con-
sideration. Nitric oxide is indicted,too, because
of its ability to produce the dioxide by atmospheric
oxidation.
Fortunately, no linkbetween atmospheric concen-
trations of nitrogen oxides and actual injury or
illness in humans or animals has been reported
yet. Hopefully, recognition of the potential danger
will prevent any incident.
Oxides of sulfur
During the past few years, information in the lit-
erature has indicated that the presence of sulfur
-------
16
AIR CONTAMINANTS
dioxide in the photochemical-smog reaction en-
hances the formation of visibility-reducing aero-
sols. The mechanism responsible for this effect
has notbeendescribed, anditisnot known whether
sulfur dioxide enters into the organic photochemical
reactions or whether the additional aerosols ob-
served represent simply a combination of sulfur
dioxide and moisture.
Primarily, gaseous oxides of sulfur in the atmo-
sphere are significant because of their toxicity.
Both the dioxide and trioxide are capable of pro-
ducing illness and lung injury even at small con-
centrations, from 5 to 10 ppm. Further, each can
combine with water in the air to form toxic acid
aerosols that can corrode metal surfaces, fabrics,
and the leaves of plants. Sulfur dioxide by itself
also produces a characteristic type of damage to
vegetation whereby portions of the plants' leaves
are bleached in a specific pattern. In concentra-
tions as small as 5 ppm, sulfur dioxide is irri-
tating to the eyes and respiratory system.
Both the dioxide and trioxide can combine with
particles of soot and other aerosols to produce
contaminants more toxic than either alone. The
combination of the dioxide and trioxide with their
acid aerosols has also been found to exert a syn-
ergistic effect on their individual toxicities. These
mixtures were apparently responsible for the ill-
ness and death associated with the famous air pol-
lution incidents that occurred in the Meuse Valley,
Belgium; inDonora, Pennsylvania; and, more re-
cently, in London, England.
Carbon monoxide
Carbon monoxide plays no part in the formation of
photochemical smog though it is almost invariably
emitted to the atmosphere along with the most po-
tent of smog formers--hydrocarbons and oxides
of nitrogen. At concentrations of 200 ppm and
greater, itproduces illness and death by depriving
the blood of its oxygen-carrying capacity. It has
been detected in the atmosphere of various urban
centers of the world at concentrations from 10 to
100 ppm. Greater concentrations have occasion-
ally been measured in confined spaces such as
tunnels and large, poorly ventilated garages. At-
mospheric concentrations have not yet been linked
to fatalities but have sometimes been implicated
in short-term illnesses of traffic officers.
Miscellaneous inorganic gases
A few additional gases were listed among those
emitted to the atmosphere from various operations
in Los Angeles County. They include ammonia,
hydrogen sulfide, chlorine, and fluorine or fluo-
rides. Although none has been detected in greater
than trace quantities in the Los Angeles atmosphere
and none is known to have any significance in the
formation of photochemical smog, these contami-
nants can be important in other types of air pollu-
tion problems. Allare toxic in small to moderate
concentrations, and the first three have unpleasant
odors. Hydrogen sulfide can cause discoloration
of certain kinds of paint; ammonia and chlorine
can discolor certain fabric dyes; fluorine and fluo-
rides, especially hydrogen fluoride, are highly
toxic, corrosive, and capable of causing damage
to vegetation, and illness and injury to humans
and animals.
Many other inorganic gases maybe individually or
locally objectionable or toxic. These are of rela-
tivelyminor importance and will not be discussed
here.
AEROSOLS
Aerosols (also called particulate matter) present
in the atmosphere may be organic or inorganic in
composition, and in liquid or solid physical state.
By definition, they must be particles of very small
size or they will not remain dispersed in the atmo-
sphere. Among the most common are carbon or
soot particles; metallic oxides and salts; oily or
tarry droplets; acid droplets; silicates and other
inorganic dusts; and metallic fumes.
The quantities of aerosols emitted in Los Angeles
County, at present, are relatively small but in-
clude at least some amounts of all the types listed
above. Particles of larger than aerosol size are
also emitted but, because of their weight, do not
long remain airborne. Additionally, however,
vast quantities of aerosols are formed in the at-
mosphere as the result of photochemical reactions
among emitted contaminants. Total quantities of
these aerosols may easily exceed those of emitted
aerosols, at least in terms of particle numbers.
Current Sources in Los Angeles County
The most important current sources of aerosol
emissions in Los Angeles County, by type of aero-
sol, will now be discussed.
Carbon or soot particles
Probably the most commonly emitted kind of par-
ticle anywhere is carbon. Carbon particles are
nearly always present among the products of com-
bustion from all types of fuels, even from opera-
tions in which the combustion is apparently complete.
In Los Angeles County, the principal sources of
emissions containing carbon or soot particles are
the exhaust effluents from motor vehicles, and
the combustion of fuels for power generation and
-------
Types of Contaminants
17
space heating, though not all the particulates
emitted from these sources are carbon. Emis-
sions from the latter group of sources vary with
the Rule 62 period* since these particles occur in
greater quantities in the effluent from the burning
of fuel oil than in that from the burning of natural
gas. During Rule 62 periods, then, the combus-
tion of fuel for space heating and generation of
power accounts for about one-fourth of the carbon
particles emitted to the atmosphere, and during
non-Rule 62 periods, about one-half. The portion
contributed by auto exhaust varies, therefore,
during comparable periods from one-half to three-
fourths of the total.
The actual total of emitted carbon particles cannot
be estimated with much accuracy, but they probably
represent about one-third to one-half of the total
aerosol emissions.
The only other sources from which significant
quantities of carbon particles mightbe emitted are
incineration of refuse, operation of piston-driven
aircraft, and operation of ships and railroads.
Even the total of particulate emissions from these
sources does not, however, comprise 5 percent
of the total of carbon emissions.
Metallic oxides and salts
Metallic oxides and salts can be found in small
quantities in the emissions from many sources.
These sources include catalyst dusts from re-
finery operations, emissions from the metals in-
dustry, effluents from combustion of fuel oil, and
even exhaust from motor vehicles. The total quan-
tity of these emissions is, however, small and
probably does not constitute more than 5 to 10
percent of the total particulate emitted to the
atmosphere.
The materials emitted as catalyst dusts are mostly
oxides. Small quantities of metallic oxides may
also result from the combustion of fuel oil and
perhaps from metal-working operations. These
oxides mightinclude those ofvanadium, aluminum,
titanium, molybdenum, calcium, iron, barium,
lead, manganese, zinc, copper, nickel, magnesium,
chromium, and silver.
Metallic salts are emitted from essentially the
same sources—again, in small concentrations.
Most emissions of particulate lead in auto exhaust,
for example, are present as oxides and complex
salts, usually chlorides, bromides, and sulfates.
Metallic oxides are emitted from certain metals
operations, and small quantities of sulfates are
emitted from some industrial operations.
* April 15 to November 15.
Oily or tarry droplets
Small droplets of oily or tarry materials are fre-
quently found in combustion effluents from many
types of sources. The most common sources are
probably the emissions associated with the opera-
tion of motor vehicles, particularly crankcase
emissions; exhaust emissions from gasoline- and
dies el-powered vehicles; effluents from asphalt
manufacturing, saturating, paving, and roofing
operations; and effluents from inefficient combus-
tion of fuels in stationary sources. Small amounts
might also be found in the effluents from aircraft,
ships, and locomotives and from incineration of
refuse. Oily or tarry particles also appear to be
among the products of the photochemical reactions
that produce smog. Emitted quantities of these
materials probably comprise 10 to 20 percent of
the total particulate emissions.
Although the composition of these materials is not
well established, they appear to be predominantly
organic. They undoubtedly have relatively high
molecular weights and probably contain at least
some aromatics. The polycyclic hydrocarbons,
which currently cause so much concern, probably
occur in a liquid phase in the atmosphere.
Acid droplets
Small droplets of acid, both organic and inorganic,
are emitted from a number of sources in Los
Angeles County under certain conditions. These
sources include stack effluents from power plants,
especially during combustion of fuel oil; effluents
from industrial operations such as certain metal-
working and plating operations, and storage bat-
tery reclamation; effluents from waste rendering
and incineration; and even effluents from motor
vehicle exhaust. Under some circumstances,
these acid droplets are also formed in the atmo-
sphere. Like the other kinds of particulate matter,
the total emitted quantities of these droplets are
small, probably 5 to 10 percent of the total par-
ticulate emissions. Even the quantities of these
materials formed in the atmosphere are small
relative to the total.
The inorganic acids emitted to the atmosphere in-
clude, primarily, sulfuric and nitric acids; the
organic acids include probably acetic, propionic,
and butyric acids. The acids formed in the atmo-
sphere through combination of gases with water
include sulfurous, sulfuric, nitrous, and nitric
acids. Acid droplets formed through oxidation of
organic emissions may not include any but acetic
acid, if that.
-------
18
AIR CONTAMINANTS
Silicates and other inorganic dusts
Emissions of inorganic dusts in Los Angeles County
consist primarily of silicates, carbonates, and
oxides and are probably associated most commonly
with quarrying operations, sand and gravel plants,
and other phases of the minerals industry. They
can also result from highway construction and
landfill operations. Their quantity may represent
about 5 to 10 percent of the total particulate emissions.
Metallic fumes
The metals industry is probably responsible for
5 to 10 percent of the total aerosol emissions, and
metal fumes probably constitute less than half of
this portion. Metal fumes are generally considered
to be minute particles created by the condensation
of metals that have vaporized or sublimed from
the molten state.
Significance in Air Pollution Problem
The significance of aerosols, and of all airborne
particulate matter, varies with the type of air pol-
lution problem in which they are involved. In most
situations, particulate emissions represent a
major portion of the total quantity of air contami-
nants and would be important for their soiling and
and nuisance properties alone, if for no other.
Even in air pollution problems of the type pro-
duced by coal burning, which involves only carbon
particles, ash, and oxides of sulfur, there are in-
dications that the toxic effect of the sulfur dioxide
and trioxide is enhanced by the concomitant par-
ticulate matter. This kind of effect has been noted
in other cases involving aerosols and toxic gases
or liquids and has given rise to the theory that
other contaminants can adsorb on the surface of
the particles and thus come into contact with inner
surfaces of the lungs and mucous membranes in
much greater concentrations than would other-
wise be possible.
Particulate emissions are also associated with re-
duction of visibility. In some instances, this is
the simple physical phenomenon of obscuration of
visibility by the quantity of interfering material.
In those instances associated with photochemical
smog, however, the visibility reduction is due to
refraction and scattering of light, and the number
and size of the particles involved are much more
important than their identities. The smaller the
particles (maximum reduction of visibility at 0. 7 mi-
cron) and the larger their number, the greater
their collective effect on visibility.
It has also been suggested that the presence of
minute particles promotes the photochemical re-
actions that produce smog. Furthermore, small
aerosol particles are among the products of these
reactions and add to the visibility reduction pro-
duced by the emitted contaminants.
AIR POLLUTION CONTROLS
ALREADY IN EFFECT
When the air pollution problem in Los Angeles
County was recognized, an agency was immedi-
ately provided to study the problem and try to solve
it. The first Air Pollution Control District in
California was formed and charged with responsi-
bility for the elimination or, at least,significant
reduction of air pollution in Los Angeles County.
During its first 10 years, the District concentrated
its efforts on control of emissions from stationary
sources. Experience of other agencies in this field
had shown that certain kinds of industrial emis-
sions were most commonly responsible for air
pollution problems. Mobile equipment was ex-
empted from control and not at that time con-
sidered a serious source of contaminants.
Continuous study and diligence have since led to
the promulgation of the most stringent and com-
plete rules and regulations in force anywhere in
the world and to the most effective control program
currently feasible. Both are frequently copied and
studied. Table 2 concisely summarizes what has
been accomplished. Other sections of this man-
ual explain in detail the methods and equipment
used.
Perhaps the most graphic evidence of the success
of this control effort is the almost complete ab-
sence of emissions from stacks and chimneys any-
where in the Los Angeles Basin. Any source of
visible emissions immediately calls attention to
itself.
CONTROL MEASURES STILL NEEDED
Despite an almost incredibly successful program
of control over stationary-source emissions, the
persistence of unpleasant effects of air pollution
and the concentrations of atmospheric contami-
nants still being measured did not properly reflect
these dramatic reductions. A research program
undertaken concurrently with the programs of con-
trol and enforcementhad revealed that the air pol-
lution problem in the Los Angeles area was dif-
ferent from that usually encountered. It had shown
also that the hydrocarbons and oxides of nitrogen
primarily responsible for the effects associated
with smog in Los Angeles were likely to be emitted
only in connection with the processing and handling
of petroleum and the combustion of fuels.
As soon as hydrocarbons were recognized to be of
great significance in this kind of air pollution prob-
lem, measures were undertaken to control the
-------
Control Measures Needed
19
w
W
H
O
Q
J
o g
U <
§5?
H
2
t)
O
U
z
o
i—i
H
o
O
J
O
.0
id
-0
D-
£
5
t
f
U
1 £
at
.3
j;
J
s
U
g
t
s
T3
*•
>
O.
S
1
C
1
Fi
>f conta
to
«
£
0
0
1.
• jj
3 •"
Q
1 1
U o
u
"
S £
1*
5
« J
£ 3
0
C TJ
•2 s
E
=
U
..
o ;
u
4 «
) "
8 ?
o-
co .-
£ *
S
£
£
m
ri
in
S
3
00 _
^s
o
i;
.5
o
f- S
3
03 _
. •
In "
3
5 -
^
S
r"
S
o
Co
oo __,
•o
O v
J •-
s
x
x
X
x
S
0
^
5
"5
T3
U) M
C
:
h
111
;
rsl
x
X
o
o
£
1
i
k
i^
x
X
XX x»
X X
s
"o-
o
o
i
;,
>
* >
ti >-
III!
f£.!J.
|l!ljll
»u « » n
O
K
»
"_
O
c
jc c
i |l
i ? i ;
tl ffl ffl OT
0.
S
V
O v,"
! 1
a •
Iff
» « O
a.
g
X
X
JC "^ S
K
X X X X X X
in
J
E 3
i « s
|j|| 1 I
. | i s
« fe 0 W 00 O
S
x x
X
X X
X X
o
X
a
fi
S 3 5
c w G. «Z
J
S
in rg i-j -. -. o
X
X X X X X X
o" __
—
£
*
Mineral
Asphalt batching
A»ph»lt roofing sat
Cl»«a and frit furna
Concrete batching
Other
S
xxx
XX XXX
x - ^ x
"r M >, 2 £ £ S
g V h O I* (< «3
« Q Q « (X > O
O
*
x
« x x
(M M -
x
X X , X XX
X X X X X X
liliiii
- « U 2 U. 3e O
0
•t
x
x
X
x
GO
S
J<
x
»
„
X
M
0
^
-
In
1
o
S
m
2
M
O
0
V
o
o
NO
Total contaminants r
by APCD rules
j^
•*
in
CO
-
O
o
~
o
-
1
0
K
f,
* S
S E
-------
20
AIR CONTAMINANTS
emissions from their principal stationary sources,
the refineries. Although these measures failed to
eliminate smog effects throughout the basin, they
diminished these effects and reduced the atmo-
spheric hydrocarbon concentrations in the refinery
area of the county. At the same time, however,
damage to plants, irritation of eyes, and reduc-
tion in visibility were more widespread and in-
creasing in severity in suburban areas that had
previously been almost smog free. Obviously,
some important source of the contaminants that
produce photochemical smog had not been ade-
quately taken into account. This source proved
to be the most prevalent consumer of petroleum
products, the gasoline-powered motor vehicle.
Although no feasible means for control of motor
vehicle emissions were yet known and mobile
sources were then exempt from control, assess-
ment of the relative importance of this source of
emissions was clearly mandatory. Investigation
of exhaust emissions revealed that, although each
individual vehicle was negligible as a source, the
vast number of vehicles could have great signifi-
cance. Further study demonstrated how the phe-
nomenal postwar growth of the Los Angeles area
had so increased the sources of air contaminants
that the gains made by the control of stationary
sources had been almost nullified. From 1945 to
1955, for example, the pollution of Los Angeles
County increased by almost 50 percent; motor
vehicle registration and gasoline consumption in-
creased about 100 percent; and the number of in-
dustrial establishments increased by nearly 80 per-
cent. The district had had to "run at great speed
to stay in one place, " and the picture of what the
situation might have been without a control program
was almost unimaginable.
Estimating the total quantities of air pollution in
Los Angeles County, the district was able to de-
termine that hydrocarbon emissions from motor
vehicles as a fraction of the total for the county
had probably increased from about one-eighth in
1940 to one-third in 1950 and to one-half to two-
thirds in 1955. Oxides of nitrogen emissions from
motor vehicles probably constituted about 50 per-
cent of the total in 1940 and 1950, and 50 to 60
percent in 1955. During the entire period, motor
vehicle emissions probably accounted for 85 to 95
percent of the total of carbon monoxide emitted to
the atmosphere. These estimates represent the
net effect of both growth and control measures and
illustrate the change in emphasis that has grad-
ually taken place. Probably this would, however,
be less true of areas that had little or no control
over emissions from stationary sources.
MOTOR VEHICLE EMISSIONS
For Los Angeles County, this gradual change and
its effect on the solution of the air pollution prob-
lem had to be carefully evaluated, and the probable
necessity for control of motor vehicle emissions,
prudently considered. In 1957, though no controls
were available, appropriate steps were taken to
enable the district to encourage the development
of necessary control devices and require their use
when they became available. Additional study,
plus the increasing occurrence of the effects of
photochemical srnog in other areas of California,
suggested that control of mobile sources at the
local, or district, level probably would not be ade-
quate. In 1959, therefore, the state government
formally occupied this particular field of air pol-
lution control in California. Although the Los
Angeles County Air Pollution Control District con-
tinued its participation in research on vehicular
emissions, its primary responsibility reverted to
control of emissions from stationary sources.
ADDITIONAL CONTROLS OVER STATIONARY SOURCES
The principal areas in which additional controls
will be needed involve reduction of emissions of
organic gases and oxides of nitrogen. Present
control measures have so far brought about only
61 and 45 percent control, respectively, of these
emissions from stationary sources. Obviously,
control of these emissions from motor vehicles
is also necessary, but this is no longer the dis-
trict's responsibility.
Organic Gases
Judged from percent of control already achieved,
additional control of emissions of organic gases
could apparently be accomplished in four areas:
combustion of fuels, marketing of petroleum,
manufacture of chemicals , and use of organic sol-
vents. Onlytwo, use of organic solvents and mar-
keting of petroleum, present much opportunity for
significant reductions. Both have been studied by
the district to determine what control methods are
available and how much additional control could
be achieved. Possible control methods for or-
ganic solvent emissions include: substitution of
water-base paints and coatings; solvent absorp-
tion or adsorption and recovery; and incineration
of combustible emissions. Possible methods for
control of hydrocarbons emitted during marketing
of petroleum products include: absorption (or con-
densation) and recovery, and use of vapor return
equipment during transfer operations. The feasi-
bility and economics of each will be discussed in
more detail in other sections of this manual.
Control of these two sources during earlier pro-
grams of the district was not considered neces-
sary. With continued growth of the area and re-
duction of emissions from other sources, however,
the relative and absolute magnitudes of these emis-
sions have increased to the point where control
-------
Control Measures Needed
21
must be undertaken. Increased emissions from
petroleum marketing have simply come with in-
creased use of gasoline. Increased emissions
from solvent use reflect the increase in usage re-
lated to the growth of the area and the technolog-
ical advances illustrated by the growing demand
for dry cleaning solvents, especially since the
advent of coin-operated machines.
Oxides of Nitrogen
Additional reduction of oxides of nitrogen emis-
sions from stationary sources poses a diffifult
problem. Significant quantities are emitted only
in connection with combustion of fuels. The re-
duction that has already been achieved was ac-
complished through substitution of natural gas for
fuel oil during 7 months of the year. Rule 62.1,
which was adopted in 1964, will bring about addi-
tional reduction, but even combustion of natural
gas produces some oxides of nitrogen, and there
is apointbeyond -which control cannot be extended
by this means. Research into the nature of com-
bustionhas suggested that control of formation of
oxides of nitrogen may be possible through changes
in design of combustion equipment and through
rigorous control of combustion conditions.
Oxides of Sulfur
Rule 62. 1 will also bring about additional reduc-
tion in emissions of the oxides of sulfur. Further
control over emissions of these contaminants
would also appear to be possible in the chemical
industry. This is not, however, necessarily true,
since most of these emissions come from sulfur
recovery operations that represent excellent and
profitable elimination of sulfur oxides emissions
from refineries. Additional control in this area
is not feasible at this time.
Other Contaminants
There are no other areas in which significant ad-
ditional reduction of contaminant emissions can
be accomplished at present. Continued surveil-
lance of all contaminant emissions will, of course,
be maintained. If any emissions are found to have
increased to the point where more stringent con-
trol is necessary or if means are discovered to
make certain additional controls feasible, the prop-
er steps to put these into effect will be taken
immediately.
-------
CHAPTER 3
DESIGN OF LOCAL EXHAUST SYSTEMS
FLUID FLOW FUNDAMENTALS
HERBERT SIMON, Senior Air Pollution Engineer
JOHN L. McGINNITY, Intermediate Air Pollution Engineer5!
JOHN L. SPINKS, Air Pollution Engineer
HOOD DESIGN
HERBERT SIMON, Senior Air Pollution Engineer
DUCT DESIGN
EDWIN J0 VINCENT, Intermediate Air Pollution Engineer
FAN DESIGN
EDWIN J. VINCENT, Intermediate Air Pollution Engineer
LEWIS K. SMITH, Air Pollution Engineer
VAPOR COMPRESSORS
GEORGE THOMAS, Intermediate Air Pollution Engineer
CHECKING AN EXHAUST SYSTEM
JOSEPH D'IMPERIO, Air Pollution Engineer^
COOLING OF GASEOUS EFFLUENTS
GEORGE THOMAS, Intermediate Air Pollution Engineer
*Now with the National Center for Air Pollution Control, Public Health Service, U. S. Department of
Health, Education, and Welfare.
'Now deceased.
-------
CHAPTER 3
DESIGN OF LOCAL EXHAUST SYSTEMS
FLUID FLOW FUNDAMENTALS
Local exhaust systems are devices used to cap-
ture dusts and fumes or other contaminants at
their source and prevent the discharge of these
contaminants into the atmosphere. Close-fitting
hoods are used to capture the contaminants from
one or more locations so that the laden gases can
by conveyed through a system of ducts by one or
more exhaust fans. An air pollution control de-
vice can then be used to collect the air contami-
nants and discharge the cleansed air into the at-
mosphere.
In designing a local exhaxist system, sufficient air
must be provided for essentially complete pickup
of the contaminants. Conversely, too much air
can result in excessive construction and opera-
tion costs. It is, therefore, necessary for the de-
signer to understand certain physical principles
that are useful in analyzing the ventilation needs
and in selecting the hooding devices.
The nature of flow of a real fluid is very complex.
The basic laws describing the complete motion of
a fluid are, in general, unknown. Some simple
cases of laminar flow, however, maybe computed
analytically. For turbulent flow, on the other hand,
only a partial analysis can be made, by using the
principles of mechanics. The flow in exhaust sys-
tems is always turbulent; therefore, the final solu-
tion to these problems depends upon experimental
data .
BERNOULLI'S EQUATION
The basic energy equation of a frictionless, in-
compressible fluid for the case of steady flow along
a single streamline is given by Bernoulli as
D V
£— _i_ __
7 2g
(1)
where
h = elevation above any arbitrary datum, ft
p = pressure, Ib/ft
7 = specific -weight, Ib/ft
v = velocity, ft/sec
g - acceleration due to gravity, 32. 17 ft/sec
C = a constant, different for each streamline.
Each term in Bernoulli's equation has the units foot-
pounds per pound of fluid or feet of fluid. These
terms are frequently refer red to as elevation head,
pressurehead, and velocity head. They also rep-
resent the potential energy, pressure energy, and
velocity energy, respectively.
When Bernoulli's equation is applied to industrial
exhaust systems, the elevation term is usually
omitted, since only relatively small changes in el-
evation are involved. Since all streamlines orig-
inate from a reservoir of constant energy (the at-
mosphere), the constant is the same for all
streamlines, and the restriction of the equation to
a single streamline canbe removed. Furthermore,
since the pressure changes in nearly all exhaust
systems are at most only a few percent of the ab-
solute pressure, the assumption of incompressi-
bility may be made with negligible error. Although
steady-flow conditions do not always exist in ex-
haust systems, it is safe to make the assumption
of steady flow if the worst possible case is con-
sidered. Any error will then be on the safe side.
All real fluids have a property called viscosity.
Viscosity accounts for energy losses, which are
the result of shear stresses during flow. The mag-
nitude of the losses must be determined experi-
mentally, but once established, the values can be
applied to dynamically similar configurations.
Bernoulli's equation may be applied to a real fluid
by adding an energy loss term. Letting _1_ be an
upstream point and 2_ a downstream point, the
energy per unit weight at _1_ is equal to the energy
per unit weight at 2_ plus all energy losses between
point _1 and point 2_.
PITOT TUBE FOR FLOW MEASUREMENT
The velocity of a fluid (liquid) flowing in an open
channel may be measured by means of a simple
pitottube, as shown in Figure 2 (Streeter, 1951).
Although this instrument is simple, usually con-
sisting of a glass tube with a right-angle bend, it
is one of the most accurate means of measuring
velocity. When the tube opening is directed up-
stream, the fluid flows into the tube until the pres-
sure intensity builds up within the tube sufficiently
to withstand the impact of velocity against it. The
fluid at a point directly in front of the tube (stag-
nation point) is then at rest. The pressure at the
stagnation point is known from the height of the
liquid column in the tube. The velocity of the fluid
25
-------
26
DESIGN OF LOCAL EXHAUST SYSTEMS
lv;^s|-"'';*.^?';T' *#,;". "?Wf4
indicate the total pressure, but now the portion
of the total head caused by velocity cannot be dis-
tinguished. The static pressure in this case can
be measured by a piezometer or static tube, as
shown in Figure 3. The total pressure H consists
of the sum of the static pressure hg and the ve-
or
locity pressure hv,
H = h
+ h
(4)
Figure 2. Simple pi tot tube
(Streeter, 1951).
in the stream maybe evaluated by-writing Bernoulli's
equation between point 1_ upstream of the stagna-
tion point and point 2^ the stagnation point. Note
that hi = h2 and V2 = 0. Therefore
2g
solving for the velocity,
(2)
(3)
A simple pitottube measures the total head or total
pressure, which is composed of two parts, as
shown in Figure 2. These are the static pressure
h and the dynamic or velocity pressure hv. In
open-channel flow, hv is measured from the free
surface. When the fluid is in a pipe or conduit
in which it flows full, a simple pitot tube will again
FLOW
I STREAML!NES
hs
PIEZOMETER
OPEN INGS
The velocity can be determined, therefore, from
the difference between the total and static heads.
In practice, measurement of total pressure and
static pressure is combined into a single instru-
ment (pitot-static tube, Figure 4), which permits
direct measurement of velocity head since the
static head is automatically subtracted from the
totalhead. An inclined manometer (Ellison gauge)
is particularly useful when the heads are small as
in exhaust systems. Use of this device to mea-
sure the flow of a gas introduces, however, an
additional factor, which is the conversion of read-
ings in inches of manometer fluid into meaning-
ful velocity terms. This relationship, when water
is used as the manometer fluid to measure the ve-
locity of air, is
2
/ v \
= f a 1
I 4005 I
(5)
•where
velocity pressure or head, inches of
water
4005 = 1096.2
rVolume in ft of 1 Ib of air
at 70°F and 14. 7 psia
velocity of air, fpm.
P I ilZOMETER
OPEN INGS
Figure 3. Static tube (Streeter, 1951).
Figure 4. Pitot-static tube (Streeter, 1951).
-------
Hood Design
Correction Factors
The relationship expressed in equation 5 is exact
only for air at standard temperature and pressure,
70°F and 14. 7 psia, respectively. A correction
must be applied for other than standard condi-
tions. If the air in the duct departs from 70 °F by
more than about 50 °F, a correction is required:
\2
4005
r460
70
460 + t
(6)
•where
t = the temperature of the air,
For smaller temperature deviations, the error is
not significant and maybe neglected. If the gas is
other than air, a correction for the difference in
density may be applied:
in the path of the high-velocity dust particles.
Inertial forces carry the air contaminants into
the hood.
Exterior hoods must capture air contaminants
that are being generated from a point outside
the hood itself, sometimes some distance away.
Exterior hoods are the most difficult to design,
require the most air to control a given procese,
and are most sensitive to external conditions.
For example, a hood that works well in a still
atmosphere may be rendered completely inef-
fectual by even a slight draft through the area.
The best rule to follow in hood design is to
place the hood where the air contaminants are
generated. Since this is not always physically
possible, it is important to consider the de-
sign criteria for external hoods.
CONTINUITY EQUATION
h. =
v
_t
_____ density of gas
4005/ \density of air STP
(7)
where the density of the gas under conditions ac-
tually existing at the time of the measurement in-
cludes the effects of temperature, pressure, and
molecular weight.
The volume of air flow is dependent upon the cross-
sectional area and the average velocity of the air.
The relations hip may be represented by the familiar
equation
where
V = Av
(8)
HOOD DESIGN
Hoods are devices used to ventilate process equip-
ment by capturing emissions of heat or air con-
taminants, which are then conveyed through ex-
haust system ductwork to a more convenient
discharge point or to air pollution control equip-
ment. The quantity of air required to capture and
convey the air contaminants depends upon the size
and shape of the hood, its position relative to the
points of emission, and the nature and quantity of
the air contaminants.
Hoods can generallybe classified into three broad
groups: Enclosures, receiving hoods, and exterior
hoods. Enclosures usually surround the point of
emission, thoughsometim.es one face may be par-
tially or even completely open. Examples of this
type are paint spray booths, abrasive blasting
cabinets, totally enclosed bucket elevators, and
enclosures for conveyor belt transfer points, mul-
lers, vibrating screens, crushers, and so forth.
Receiving hoods are those wherein the air con-
taminants are injected into the hoods. For ex-
ample, the hood for a grinder is designed to be
V = total air volume, cfm
A = cross-sectional area, ft
v = velocity, fpm.
The continuity equation (equation 8) shows that,
for a given quantity of fluid, the velocity must in-
crease if the area decreases. Imagine that air
is being withdrawn from a point at the center of a
large room. Since an imaginary point has no di-
mensions, there will be no interference •with the
flow of air toward the point. The air will, there-
fore,approach this point radially and at a uniform
rate from all directions. The velocity of the air
must increase as it passes through a succession
of diminishing areas represented by spherical
surfaces in its approach to the imaginary point,
according to the relationship
V
4TTT
(9)
•where
= the distance from the imaginary point, ft.
-------
28
DESIGN OF LOCAL EXHAUST SYSTEMS
AIR FLOW INTO A DUCT
If a circular duct opening, representing a simple
hood, is substituted for the imaginary point, the
pattern of flow into the end of the duct, or hood,
will be modified as shown in Figure 5 because of
the interference from the duct. The velocity of
the air approaching a plain, circular opening along
the axis of the duct is given by Dalla Valle (1952)
as approximately
0. 1A
100 - Y
(10)
where
Y =
A =
the percent of the velocity at the open-
ing found at a point x on the axis
the distance outward along the axis from
the opening, ft
the area of the opening, ft .
The velocity at the opening is computed from the
continuity equation.
The actual flow pattern is found to be as shown
in Figure 5 from studies by Dalla Valle and others.
The lines of constant velocity are called contour
lines, while those perpendicular to them are
streamlines, which represent the direction of flow.
The addition of a flange improves the efficiency of
the duct as a hood for a distance of about one di-
ameter from the duct face. Beyond this point,
flanging the duct improves the efficiency only
slightly. Figure 6 illustrates flow patterns for
several sizes of square hoods. Because there is
little difference in the center line velocity of hoods
of equal air volume at a distance of one or two
hood diameters from the hood face, Hemeon (1955)
recommends using one equation for all shapes --
square, circular, and rectangular up to about 3:1
length-to-width ratio. He also does not distinguish
between flanged and unflanged hoods, which ap-
pears justified when these hoods are used only at
distances of one diameter or more from the hood
face. At close distances, flanged hoods are far
superior at the same volume. By rearranging
terms in equation 10 and combining with equation
8, the following is obtained:
V
v (lOx
x
Af)
(11)
where
V = the volume of air entering the hood, cfm
v = the velocity at point x, fpm
x = the distance to any point x on the axis or
center line of the hood measured from
the hood face, ft
A = the area of the hood face, ft .
Analysis of equation 11 shows that at the hood face
x = 0 and the equation becomes identical to equa-
tion 8. For large values of x, the Af term becomes
less significant, as the evidence shows it should.
To use equation 11, select a value of vx that is
sufficient to assure complete capture of the air
contaminants at point x. From the physical di-
mensions and location of the hood, Af and x are
determined. The volume required may then be
calculated.
While equation 1 1 applies to a freestanding or un-
obstructed hood, it can also be applied to a rec-
tangular hood bounded on one side by a plane sur-
face, as shown in Figure 7. The hood is considered
to be twice its actual size, the additional portion
being the mirror image of the actual hood and the
bounding plane being the bisector. Equation 11
then becomes
V == v
t x
lOx + 2 A,
(12)
where the terms have the same meaning as before.
NULL POINT
Air contaminants are often released into the at-
mosphere with considerable velocity at their point
of generation. Because the mass is essentially
small, however, the momentum is soon spent and
the particles are then easily captured. Hemeon
(1955) refers to a null point, shown in Figure 8,
as the distance \vithin which the initial energy of
an emitted air contaminant has been dissipated or
nullified in overcoming air resistance. If an ade-
quate velocity toward the hood is provided at the
farthestnull point from the hood, all the air con-
taminants released from the process "will be cap-
tured. What constitutes an adequate velocity to-
wards the hood depends upon drafts in the area and
cannot, therefore, be determined precisely.
Establishing the null point in advance for a new
proces s is not always easy or even possible. For
existing equipment, however, direct observation
•will usually establish a locus of null points. Ob-
viously, in the absence of external disturbances,
any positive velocity toward the hood at the farthest
null point will give assurance of complete capture.
When this is put into practice, however, the re-
sults are disappointing. Even closed rooms have
drafts and thermal currents that destroy the hood's
effectiveness unless a substantial velocity toward
the hood is created at the farthest null point. Ex-
perience has shown that a velocity of less than
100 fpm at a null point can seldom, if ever, be
tolerated without a loss in the hood's effectiveness.
Draft velocities in industrial situations may al-
most always be expected to be 200 to 300 fpm
-------
Hood Design
Z9
e 7
DISTANCE FROM OPENING, inches
4 5 f 7
DISTANCE FROM OPENING, inches
Figure 5. Actual flow contours and streamlines for flow into circular openings.
Contours are expressed as percentage of opening velocity (Dalla Valle, 1952).
-------
30
DESIGN OF LOCAL EXHAUST SYSTEMS
£34567
DISTANCE OUTWARD FROM OPENING, inches
Figure 6. Actual velocities for square openings
of different sizes. Air flow through each opening
is 500 cfm (Dalla Valle, 1952).
Figure 7. Rectangular hood bounded by a plane
surface (Hemeon, 1955).
or more periodically, and draft velocities of 500
to600fpmare not unusual in many cases. Drafts
such as these may prevent capture of air contami-
nants by exterior hoods, as illustrated in Figure 9
for the case of ahigh-canopy hood, unless adequate
baffling is provided or hood volume is increased
to unreasonable values. Baffling provides, in ef-
fect, an enclosure that is almost always the most
efficient hooding.
DESIGN OF HOODS FOR COLD PROCESSES
A large body of recommended ventilation rates has
be en built up over the years by various groups and
organizations who are concerned with the control
of air contaminants. This type of data is illustrated
in Table 3. The use of these recommended values
Figure 8. Location of null point and x-distance
(Hemeon, 1955),
Figure 9. Drafts divert the rising
column of air and prevent its capture
by the hood (Hemeon, 1955).
greatly simplifies hooding design for the control of
many common air pollution problems. Note that
almost all published recommendations have speci-
fied complete or nearly complete enclosure.
These published data provide a reliable guide for
the design engineer. The recommended values
must, however, be adjusted to specific applica-
tions that depart from the assumed normal
conditions.
-------
Hood Design
31
Table 3. EXHAUST REQUIREMENTS FOR VARIOUS OPERATIONS
Operation
Exhaust arrangement
Remarks
Abrasive blast
r o om s
Abrasive blast
cabinets
Bagging machines
Belt conveyors
Bucket elevator
Foundry screens
Tight enclosures with
air inlets (generally
in roof)
Tight enclosure
Booth or enclosure
Hoods at transfer
points enclosed as
much as possible
Tight casing
Enclosure
Foundry shakeout
Enclosure
Foundry shakeout Side hood (with side
shields when possible)
Grinders, disc
and portable
Grinders and
crushers
Mixer
Packaging
machines
Paint spray
Rubber rolls
(calendars)
Welding (arc)
Downdraft grilles in
bench or floor
Enclosure
Enclosure
Booth
Downdraft
Enclosure
Booth
Enclosure
Booth
For 60 to 100 fpm downdraft or 100 fpm
crossdraft in room
For 500 fpm through all openings, and
a minimum of 20 air changes per
minute
For 100 fpm through all openings for
paper bags; 200 fpm for cloth bags
For belt speeds less than 200 fpm,
V = 350 cfm/ft belt width with at least
150 fpm through openings. For belt
speeds greater than 200 fpm, V =
500 cfm/ft belt width with at least
200 fpm through remaining openings
For 100 cfm/ft of elevator casing
cross-section (exhaust near elevator
top and also vent at bottom if over
35 ft high)
Cylindrical--400 fpm through openings,
and not less than 100 cfm/ft2 of cross -
section; flat deck--200 fpm through
openings, and not less than 25 cfm/ft
of screen area
For 200 fpm through all openings, and
not less than 200 cfm/ft of grate area
with hot castings and 150 cfm/ft^ with
cool castings
For 400 to 500 cfm/ft2 grate area with
hot castings and 350 to 400 cfm/ft2 with
cool castings
For 200 to 400 fpm through open face,
but at least 150 cfm/ft of plan -working
area
For 200 fpm through openings
For 100 to 200 fpm through openings
For 50 to 100 fpm
For 75 to 150 fpm
For 100 to 400 fpm
For 100 to 200 fpm indraft, depending
upon size of "work, depth of booth, etc.
For 75 to 100 fpm through openings
For 100 fpm through openings
-------
32
DESIGN OF LOCAL EXHAUST SYSTEMS
Spray Booths
Spray booths of the open-face type are generally
designed to have a face indraft velocity of 100 to
200 fpm. This is usually adequate to assure com-
plete capture of alloverspray, provided the spray-
ing is done within the confines of the booth, and
the spray gun is always directed towards the in-
terior. It is a common practice, especially with
large work pieces, to place the work a short dis-
tance in front of the booth face. The overspray
deflected from.the workmay easily escape capture,
particularly with a careless or inexperienced op-
erator. If this situation is anticipated, the equip-
ment designer can provide a velocity of 100 fpm
at the farthest point to be controlled, as in the
following illustrative problem.
Abrasive Blasting
Abrasive blasting booths are similar to spray
booths except that a complete enclosure is always
required. In addition, particularly for small booths
(bench type), the ventilation rate must sometimes
be increased to accommodate the air used for blast-
ing. The volume of blasting air can be determined
from the manufacturer's specifications. For a
small blasting, booth, this will usually be about 50
to 150 cfm. The following illustrative problem
shows how the ventilation rate for this kind of
equipment is calculated.
Example 2
Given:
Example 1
Given:
A paint spray booth 10 feet wide by 7 feet high.
Work may be 5 feet in front of the booth face at
times. Nearly draftless area requires 100 fpm
at point of spraying.
Problem:
Determine the exhaust rate required.
Solution:
From equation 12, volume required =
lOx + 2 A,
V = v
t x
V = 100
From equation 8, face velocity =
V
= 19, 500 cfm
'f
f
19,500
,
fpm
When the spraying area is completely enclosed to
form a paint spray room, the ventilation require-
ments are not greatly reduced over those for
spraying inside an open-face booth. The reason
for this is that a velocity of approximately 100 fpm
must be provided through the room for the comfort
and health of the operator.
A small abrasive blasting enclosure 4 feet wide by
3 feet high by 3 feet deep. Total open area equals
1. 3 ft2.
Problem:
Determine the exhaust rate required.
Solution:
From Table 3, ventilation required = 500 fpm
through all openings but not less than 20 air
changes per minute.
Volume at 500 fpm through all openings:
= 650 cfm
Volume required ior 20 air changes per minute:
V = 500 x 1.3
V = 20 x volume of booth
V = 20x4x3x3
= 720 cfm
Open-Surface Tanks
Open-surface tanks may be controlled by canopy
hoods or by slot hoods, as illustrated in Figure 10.
The latter are more commonly employed. The
ventilation rates required for open-surface tanks
may be taken from Table 4, -which is a modifica-
tion of the American Standards Association code
Z9.1. These values should be considered as min-
imum under conditions where no significant drafts
will interfere with the operation of the hood. When
slot hoods are employed the usual practice is to
provide a slot along each long side of the tank. The
slots are designed for a velocity of 2, 000 fpm
through the sLot face at the required ventilation
rate. For a tank with two parallel slot hoods, the
ventilation rate required and the slot width bs may
be taken directly from Figure 11, which graphs
the American Standards Association code Z 9. 1.
-------
Hood Design
33
Figure 10. Slot hood for control
of emissions from open-surface tanks
(Adapted from industrial Ventilation.
1960).
Neither the code nor Figure 11 makes allowance
for drafts. The use of baffles is strongly recom-
mended wherever possible to minimize the effect
of drafts. If baffles cannot be used or arenot suf-
ficiently effective, the ventilation rate must be in-
creased. The slot width is also increased to hold
the slot face velocity in the range of 1,800 to
2, 000 fpm.
be used by assuming the tank to be half of a tank
twice as wide having slot hoods on both sides. This
procedure is illustrated below.
Example 4
Given:
The same tank as in Example 3, but a slot hood
is to be installed along one side only. The other
side is flush -with a vertical -wall.
Problem:
Determine the total exhaust rate and slot width
required.
Solution:
The ventilation rate in cfm per foot of tank length
is taken as half the rate for a tank twice as -wide
from Figure 11. Use width of 4 feet.
V
880
2
= 440 cfm per foot
Total exhaust volume required
The use of Figure 11 is illustrated in the follow-
ing problem:
Example 3
Given:
A chrome plating tank, 2 feet "wide by 3 feet long,
to be controlled by parallel slot hoods along each
of the 3-foot-long sides.
Problem:
Determine the total exhaust rate required and the
slot width.
Solution:
From Figure 11, the ventilation rate required is
390 cfm per foot of tank length.
V = 440 x 3
= 1, 320 cfm
V = 390 x 3
= 1, 170 cfm
From Figure 11, the slot width is 1~ inches.
8
If a slot hood is used on only one side of a tank to
capture emissions, and the opposite side of the
tank is bounded by a vertical wall, Figure 11 can
Slot width is read directly from Figure 11 for twice
the width bs = 2-5/8 inches.
,000
900
600
200
—
—
—
—
—
=-
^_
~
5
-, A
1
~\ \
/\
f
/
/
MM
/
s
III!
/
/
III!
j
/
/
iiiilini
/
/
Illllllll
/
|
3
2-1/2
1-3/4 u
=>
o
LU
oc.
1 t-
O
7,0 "*
O
3/4 ^
2
20 30 40 50 6C
WIDTH OF TANK (b), inches
Figure 11. Minimum ventilation rates
required for tanks.
-------
34
DESIGN OF LOCAL EXHAUST SYSTEMS
Table 4. VENTILATION RATES FOR OPEN-SURFACE TANKS
(American Air Filter Company, Inc., 1964)
Process
Plating
Chromium (chromic acid mist)
Arsenic (arsine)
Hydrogen cyanide
Cadmium
Anodizing
Metal cleaning (pickling)
Cold acid
Hot acid
Nitric and sulfuric acids
Nitric and hydrofluoric acids
Metal cleaning (degreasing)
Trichloroethylene
Ethylene dichloride
Carbon tetrachloride
Metal cleaning (caustic or electrolytic)
Not boiling
Boiling
Bright dip (nitric acid)
Stripping
Concentrated nitric acid
Concentrated nitric and sulfuric acids
Salt baths (molten salt)
Salt solution (Parkerise,Bonderise, etc.)
Not boiling
Boiling
Hot 'water (if vent, desired)
Not boiling
Boiling
Minimum ventilation rate,
cfm per ft of
hood opening
Enclosing
hood
One
open
side
75
65
75
75
75
65
75
75
75
75
75
75
65
75
75
75
75
50
90
75
50
75
Two
open
sides
100
90
100
100
100
90
100
100
100
100
100
100
90
100
100
100
100
75
90
100
75
100
Canopy
hood
Three
open
sides
125
100
125
125
125
100
125
125
125
125
125
125
100
125
125
125
125
75
100
125
75
125
Four
open
sides
175
150
175
175
175
150
175
175
175
175
175
175
150
175
175
175
175
125
150
175
125
175
Minimum ventilation rate, a
cfm per ft of tank area.
Lateral exhaust
W/T tank width
W/L _ ratio
tank length
W/L
00 to 0. 24
A B
125 175
90 130
125 175
125 175
125 175
90 L30
125 175
125 175
125 175
125 175
125 175
125 175
90 130
125 175
125 175
125 175
125 175
60 90
90 130
125 175
60 90
125 175
W/L
0. 25 to 0. 49
A B
150 200
110 150
150 200
150 200
150 200
110 150
150 200
150 200
150 200
150 200
150 200
150 200
110 150
150 200
150 200
150 200
150 200
75 100
110 150
150 200
75 100
150 200
W/L
0.50 to 1.0
A B
175 225
130 170
175 225
175 225
175 225
130 170
175 225
175 225
175 225
175 225
175 225
175 225
130 170
175 225
175 225
175 225
175 225
90 110
130 170
175 225
90 110
175 225
aColumn A refers to tank with hood along one side or two parallel sides when one hood is against a
wall or a baffle running length of tank and as high as tank is wide; also to tanks with exhaust mani-
fold along center line with W/2 becoming tank width in W/L ratio.
Column B refers to freestanding tank with hood along one side or two parallel sides.
DESIGN OF HOODS FOR HOT PROCESSES
Canopy Hoods
Circular high-canopy hoods
Hoodingfor hot processes requires the application
of different principles than that for cold processes
because of the thermal effect. When significant
quantities of heat are transferred to the surround-
ing air by conduction and convection,a thermal draft
is created that may cause a rising air current with
velocities sometimes over 400 fpm. The design
of the hood and the ventilation rate provided must
take this thermal draft into consideration.
As the heated air stream rising from a hot sur-
face moves upward, it mixes turbulently with the
surrounding air. The higher the air column rises
the larger it becomes and the more diluted -with
ambient air. Sutton (1950) investigated the turbu-
lent mixing of a rising column of hot air above a
heat source. Using data from experiments by
Schmidt published in Germany, and his own ex-
periments-with military smoke generators, Sutton
developed equations that describe the velocity and
-------
Hood Design
35
diameter of a hot rising jet at any height above a
hypothetical point source located a distance z be-
low the actual hot surface. Hemeon adapted Sutton' s
equations to the design of high-canopy hoods for
the control of air contaminants from hot sources.
The rising air column illustrated in Figure 12 ex-
pands approximately according to the empirical
formula
D
= 0.5
0. 88
(13)
where
D = the diameter of the hot column of air at
the level of the hood face, ft
xf = the distance from the hypothetical point
source to the hood face, ft.
From Figure 12 it is apparent that Xf is the sum
of y, the distance from hot source to the hood face,
and z the distance below the hot source to the hy-
pothetical point source. Values of z may be taken
from Figure 13. According to Hemeon, the ve-
locity of the rising column of air into the hood
may be calculated from
100
--
QJ
UJ
tt
3
o
CO
O
Q_
C_)
1- 5
° ,
>-
IT
t- ?
1
/
/
/
.
/
/
Jr*
/
/
/
s
/
/
/
j
i
jf
r
A
i = (2DS)'-38
1 2345 1C 3C
DIAMETER OF HOT SOURCE (Ds), feet
Figure 13. Value of z for use
with high-canopy hood equations.
37
0.29
,1/3
(14)
where
HYPOTHETICAL
POINT SOURCE
Figure 12. Dimensions used to
design high-canopy hoods for hot
sources (Hemeon, 1955).
v = the velocity of the hot air jet at the
level of the hood face, fpm
x = the height of the hood face above the
theoretical point source = y -I- z, ft
q = the rate at which heat is transferred
^•c
to the rising column of air, Btu/min.
The rate at which heat is absorbed by the rising
column may be calculated from the appropriate
natural convection heat loss coefficient q listed
in Table 5 and from the relationship
60
A At
s
(15)
•where
q = the total heat absorbed by the rising air
column, Btu/min
q.^ = the natural convection heat loss coeffi-
cient listed in Table 5, Btu/ft^ per hr per °F
A = the area of the hot source, ft^
s '
At = the temperature difference between the hot
source and the ambient air, °F.
-------
36
DESIGN OF LOCAL EXHAUST SYSTEMS
Table 5. COEFFICIENTS FOR CALCULATING SENSIBLE
HEAT LOSS BY NATURAL CONVECTION (Hemeon, 1955)
Shape or disposition of heat surface
Natural convection
heat loss (q) coefficienta
Vertical plates, over 2 ft high
Vertical plates, less than 2 ft high
(X = height in ft)
Horizontal plates, facing upward
Horizontal plates, facing down-
ward
Single horizontal cylinders
(where d is diameter in inches)1
Vertical cylinders, over 2 ft high
(same as horizontal)
0.3 (At)
,At
1/4
0.28
X
0.38 (At)
1/4
0.2 (At)
1/4
At
0.42(f)
1/4
Vertical cylinders less than 2 ft
high. Multiply q-^ from
formula above by appropriate
factor below:
Height, ft
Factor
0.
0.
0.
0.
0.
1.
1
2
3
4
5
0
3.5
2. 5
2.0
1.7
1. 5
1. 1
aHeat loss coefficient, q is related to q as follows:
i-j C
- - A At
60 s
Schmidt's experiments were conducted in a closed
laboratory environment designed to minimize drafts
and other disturbances. Nevertheless, Sutton re-
ports that there was a considerable amount of
waver and fluctuation in the rising air column. In
developing his equations, Sutton defined the hori-
zontal limits of the rising air column as the locus
of points having a temperature difference relative
to the ambient atmosphere equal to 10 percent of
that at the center of the column.
In view of the facts that this arbitrary definition
does not truly define the outer limits of the rising
air column and that greater effects of waver and
drafts may be expected in an industrial environ-
ment, a safety factor should be applied in calcu-
lating the size of the hood required and the mini-
mum ventilation rate to assure complete capture
of the emissions. Since high-canopy hoods usually
control emissions arising from horizontal-plane
surfaces, a simplification can be derived by com-
bining equations 14 and 15 with the heat transfer
coefficient for horizontal-plane surfaces and allow-
ing a 15 percent safety factor.
(16)
4
Although the mean diameter of the rising air col-
umn in the plane of the hood face is determined
from equation 13, the hood must be made some-
what larger in order to assure complete capture
of the rising column of contaminated air as it waver s
back and forth and is deflected by drafts. The
exact amount of allowance cannot be calculated
precisely, but factor s that must be considered in-
clude the horizontal velocity of the air currents in
Che area, the size and velocity of the rising air
jet, and the distance y of the hood above the hot
-------
Hood Design
37
source. Other factors being equal, it appears
most likely that the additional allowance for the
hood size must be a function of the distance y.
Increasing the diameter of the hood by a factor of
0. 8 y has been recommended (Industrial Ventila-
tion, 1960). The total volume for the hood can "be
calculated from
Diameter of rising air stream at the hood face
from Equation 13:
V,
v (A - A )
r f c
(17)
v/here
V
Vf
A
Af
= the total volume entering the hood,cfm
= the velocity of the rising air column at
the hood face, fpm
= the area of the rising column of con-
taminated air at the -hood face, ft
the required velocity through the re-
maining area of the hood, A - AC, f;
the total area of the hood face, ft^.
The value of vr selected will depend upon the
draftiness, height of the hood above the source,
and the seriousness of permitting some of the
contaminated air to escape capture. The value
of this velocity is usually taken in the range of
100 to 200 fpm. It is recommended that a value
less than 100 fpm not be used except under ex-
ceptional circumstances. The following problem
illustrates the use of this method to design a high-
canopy hood to control the emissions from a
rnetal-melting furnace.
Example 5
Given:
A zinc-melting pot 4 feet in diameter with metal
temperature 880°F. A high-canopy hood is to
be used to capture emissions. Because of inter-
ference, the hood must be located 10 feet above
the pot. Ambient air temperature is 80°F.
Problem:
Determine the size of hood and exhaust rate
required.
Solution:
z = 11 feet from Figure 13 for 4-foot-diameter
source
= 0.5 x
D = 0. 5 (21)
c
0.88
= 7. 3 feet
Area of rising air stream at the hood face:
TT ^ 2
A = — D
c 4 c
A = (0. 7854)(7.3)
c
= 42 square feet
Hood size required--including increase to allowfor
waver of jet and effect of drafts:
D = D + 0.8y
f c '
D = 7.3 + (0. 8)(10) = 15. 3
Use 15-foot-4-inch-diameter hood
Area of hood face:
A, - 7
-------
38
DESIGN OF LOCAL EXHAUST SYSTEMS
Example 6
Given:
The same furnace as in example problem No. 5,
but the hood is lowered to 6 feet above the pot.
Problem:
Determine the size of hood and exhaust rate
required.
Solution:
z = 11 feet from Figure 13 for 4-foot-diameter
source
xf = z + y
xf = 11 + 6
= 17 feet
Diameter of rising air stream at the hood face
from equation 13.
(800)
5/12
(17)
1/4
= 149 fpm
Total volume required for hood:
V = v A + 100 (A - A )
Vt = (149)(29.2) + (100)(93.7-29.2)
= 10,800 cfm
Rectangular High-Canopy Hoods
The control of emissions from sources with other
than circular shape may best be handled by hoods
of appropriate shape. Thus, a rectangular source
would require a rectangular hood in order to min-
imize the ventilation requirements. A circular
hoodusedto control a rectangular source of emis-
sion would require an excessive volume. The
method used to design a hood for a rectangular
source is illustrated in example 7.
D = 0. 5 (17)°'88 = 6. 1 feet
c
Area of the rising air stream at the hood face:
A = (0.7854)(6. 1) = 29. 2 square feet
c
Hood size required:
Df = D + 0.8 y
Df = (6. 1) + (0. 8)(6) = 10.9 feet
Use 1 0-foot-11-inch-diameter hood
Example 7
Given:
A rectangular lead-melting furnace 2 feet 6 inches
wide by 4 feet long. Metal temperature 700°F. A
high-canopy hood is to be used located 8 feet above
furnace. Assume 80°F ambient air.
Problem:
Determine the dimensions of the hood and the ex-
haust rate required.
Solution:
z = 6. 2 from Figure 13 for 2. 5-foot source
x = z + y = 6. 2 +
= 14. 2 feet
Area of the hood face:
The width of the rising air jet at the hood may be
calculated from
Af = (0.7854)00.92) = 93.7 square feet
D = 0. 5 x
c f
0. i
D =0.5 (14. 2)
0. 88
= 5. 2 feet
Velocity of rising air jet at hood face:
8(A)1/3(At)5/12
s
1/4
The length of the rising air jet may be assumed
to be increased over that of the source the same
amount as the width
D = (4) 4- (5.2 - 2.5) =6.7 feet
c
-------
Hood Design
39
The area of the rising air jet is
A = (5.2)(6.7) = 35 square feet
The hood must be larger than the rising air stream
to allow for waver and drafts. By allowing 0. 8 y
for both width and length, the hood size is
Width = (5.2) + (0.8)(8) = 11. 6 feet
Length = (6. 7) + (0. 8)(8) = 13. 1 feet
Use hood 11 feet 7 inches wide by 13 feet
1 inch long
Area of hood:
Af = (11.58M13. 083) = 152 square feet
Velocity of rising air jet:
Vf =
1/4
(620)
5/12
U4.2)
1/4
= 130 fpm
important distinction is that the hood is close
enough to the source that very little mixing be-
tween the rising air column and the surrounding
atmosphere occurs. The diameter of the air col-
umn may, therefore, be considered essentially
equal to the diameter of the hot source. The hood
need be larger by only a small amount than the hot
source to provide for the effects of waver and de-
flection due to drafts. When drafts are not a seri-
ous problem, extending the hood 6 inches on all
sides should be sufficient. This means that the
hood face diameter must be taken as 1 foot greater
than the diameter of the source. For rectangular
sources, a rectangular hood would be provided
with dimensions 1 foot wider and 1 foot longer than
the source. Under more severe conditions of draft
or toxic emissions, or both, a greater safety factor
is required, -which can be provided by increasing
the size of the hood an additional foot or more or
byproviding a complete enclosure. A solution to
the problem of designing low-canopy hoods for hot
sources has been proposed by Hemeon (1955).
Although the hood is usually larger than the source,
little error occurs if they are considered equal.
The total volume for the hood may then be deter-
mined from the following equation obtained by re-
arranging terms in Hemeon1 s equation and apply-
ing a 15 percent safety factor.
(18)
Total volume required for hood:
V = v A + v (A - A )
t f c r ± c
V = (130)(35) + (200)(152-35) = 28, 000 cfm
where
V
At =
total volume for the hood, cfm
the diameter of the hood, ±t
the difference Detween the temperature
ot tne not source and the ambient at-
mosphere, "F.
Note that in this problem a velocity of 200 fpm was
used through the area of the hood in excess of the
area of the rising air column. A larger value was
selected for this case because lead fumes must
be captured completely to protect the health of the
workers in the area.
A graphical solution to equation 18 is shown in
Figure 14. To use this graph, select a hood size
1 or 2 feet larger than the source. The total vol-
ume required for a hood Df feet in diameter may
then be read directly from the graph for the actual
temperature difference At between the hot source
and the surrounding atmosphere.
Circular low-canopy hoods
The design of low-canopy hoods is somewhat dif-
ferent from that for high-canopy hoods. A hood
may be considered a low-canopy hood when the
distance between the hood and the hot source does
not exceed approximately the diameter of the
source, or 3 feet, whichever is smaller. A rigid
distinction bet-ween low-canopy hoods and high-
canopy hoods is not intended or necessary. The
Example 8
Given:
A low-canopy hood is to be used to capture the
emissions during fluxing and slagging of brass
in a 20-inch-diameter ladle. The metal tem-
perature during this operation will not exceed
2, 350°F. The hood will be located 24 inches
above the metal surface. Ambient temperature
may be assumed to be 80°F.
-------
40
DESIGN OF LOCAL EXHAUST SYSTEMS
O
O
1DD
200
3,000 4,000 5,000
400 500 1,000 2,000
TOTAL VENTILATION RATE (Vt), cfm
Figure 14. Minimum ventilation rates required for circular low-canopy hoods.
10,000
Problem:
Determine the size of hood and exhaust rate
required.
Solution:
Temperature difference between hot source and
ambient air:
At = 2,350 - 80 = 2,270°F
Use a hood diameter 1 foot larger than the hot
source:
Df = 1.67 + 1.0 = 2.67 feet
Total exhaust rate required from Figure 14.
V =1, 150 cfm
Rectangular low-canopy hoods
In a similar manner, Hemeon's equations for
low-canopy hoods may be modified and simpli-
fied for application to rectangular hoods. With
a 15 percent safety factor, the equation then
becomes
t
L
where
4/3 5/12
5. 2 b At
(19)
V = the total volume for a low-canopy rec-
tangular hood, cfm
L = the length of the rectangular hood (usu-
ally 1 to 2 feet larger than the source),
ft
b = width of the rectangular hood (usually 1
to 2 feet larger than the source), ft
At = the temperature difference between the
hot source and the surrounding atmo-
sphere, °F.
Figure 15 is a graphical solution of equation 19.
The use of this graph to design a low-canopy rec-
tangular hood for a rectangular source is illus-
trated in example 9.
Example 9
Given:
A zinc die-casting machine with a 2-foot-wide by
3-foot-long holding pot for the molten zinc. A
low-canopy hood is to be provided 30 inches above
the pot. The metal temperature is 820°F. Am-
bient air temperature is 90°F.
-------
Hood Design
41
BO
100 150 200 3DO 400 500 BOO 800 1.000
MINIMUM VENTILATION RATE (Vt/L), cfm/ft of hood length
1,500
Figure 15. Minimum ventilation rates for rectangular low-canopy hoods.
Problem:
Determine hood size and exhaust rate required.
Total exhaust rate required for hood:
V = 430 x 4
= 1,720 cfm
Solution:
Use a hood 1 foot wider and 1 foot longer than
the source.
Hood size = 3 feet wide by 4 feet long.
Temperature difference between the hot source
and ambient air:
At = 820 - 90
= 730°F
Exhaust rate required per foot of hood length
from Figure 15.
430 cfm/ft
Enclosures
A low-canopy hood with baffles is essentially the
same as a complete enclosure. The exhaust rate
for an enclosure around a hot source must, there-
fore, be based on the same principles as that for
a low-canopy hood. Enclosures for hot processes
cannot, however, be designed in the same manner
as for cold processes. Here again, the thermal
draft must be accommodated by the hood. Fail-
ure to do so will certainly result in emissions
escaping from the hood openings. After deter-
mining the exhaust rate required to accommodate
the thermal draft, calculate the hood face velocity
or indraft through all openings. The indraft through
all openings in the hood should not be less than 100
fpm under any circumstances. When air contami-
nants are released with considerable force, a min-
imum indraft velocity of 200 fpm should be pro-
vided. When the air contaminants are released
with extremely great force as, for example, in a
-------
DESIGN OF LOCAL EXHAUST SYSTEMS
direct-arc electric steel-melting furnace, an in-
draft of 500 to 800 fpm through all openings in the
hood is required.
Specific Problems
Steaming tanks
When the hot source is a steaming tank of -water,
Hemeon (1955) develops a special equation by as-
suming a latent heat of 1, 000 Btu per pound of
water evaporated. He derives the following equa-
tion for the total volume required for a low-canopy
hood venting a tank of steaming hot water.
= 290 (W AD)
S I U
1/3
(20)
where
Vt
W
s
Af
D
the total hood exhaust rate, cfm
the rate at which steam is released,
Ib/mln
the area of the hood face, assumed
approximately equal to the tank area,
ft2
the diameter for circular tanks or the
width for rectangular tanks, ft.
q = the rate at -which heat is transferred
^c
to the air in the hood from the hot
source, Btu/min
A = the area of the orifice, ft
m
the average temperature of the air in-
side the hood, °F.
11 feet
Figure 16. Illustration of leakage from
top of hood (Hemeon, 1955).
Preventing leakage
Hoods for hot processes must be airtight. When
leaks or openings in the hood above the level of
the hood face occur, as illustrated in Figure 16,
they will be a source of leakage owing to a chim-
ney effect, unless the volume vented from the hood
is substantially increased. Since openings may
sometimes be unavoidable in the upper portions
of an enclosure or canopy hood, a means of de-
termining the amount of the leakage and the in-
crease in the volume required to eliminate the
leakage is necessary. Hemeon (1955) has devel-
oped an equation to determine the volume of leak-
age from a sharp-edge orifice in a hood at a point
above the hood face.
(21)
where
the velocity of escape through orifices
in the upper portions of a hood, fpm
the vertical distance above the hood
face to the location of the orifice, ft
A small amount of leakage can often be tolerated;
however, if the emissions are toxic or malodorous,
the leakage must be prevented completely. If all
the cracks or openings in the upper portion of the
hood cannot be eliminated, the volume vented from
the hood must be increased so that the minimum
indraft velocity through all openings including the
hoodface is in excess of the escape velocity through
the orifice calculated bymeans of equation 21. The
value of qc may be determined by using the appro-
priate heat transfer coefficient from Table 5 to-
gether with equation 15 or by any other appropriate
means. This method is illustrated in example 10.
Example 10
Given:
Several oil-fired crucible furnaces are hooded
and vented as illustrated in Figure 16. The en-
closure is 20 feet long. It is not possible to pre-
vent leakage at the top of the enclosure. Total
area of the leakage openings is 1 square foot. The
fuel rate is 30 gallons per hour and the heating
value is 140,000 Btu per gallon. Assume 80°F
ambient air and 150°F average temperature of
gases in the hood.
-------
Hood Design
43
Problem:
Determine the minimum face velocity and total
exhaust rate required to prevent leakage of con-
taminated air through the upper openings by as-
suming all openings are sharp-edge orifices.
Solution:
The rate of heat generation:
Btu 1
hr
x 140,000
x
gal 60
= 70,000^-
mm
Total open area:
A = (20 x I) + 1 = 141 ft
o
The escape velocity through the leakage orifice:
,A (460 + t
1 o m
HOOD CONSTRUCTION
If air temperature and corrosion problems are not
severe, hoods are usually constructed of galva-
nized sheetmetal. As with elbows and transitions,
the metal should be at least 2 gauges heavier than
the connecting duct. Reinforcement with angle
iron and other devices is required except for very
small hoods.
High-Temperature Materials
For elevated temperatures up to approximately
900°F, black iron may be employed, the thickness
of the metal being increased in proportion to the
temperature. For temperatures in the range of
400 to 500°F, 10-gauge metal is most commonly
employed. When the temperature of the hood is
as high as 900°F, the thickness of the metal may
be increased up to 1/4 inch. Over 900°F, up to
about 1,600 to 1,800°F, stainless steel must be
employed. If the hood temperature periodically
exceeds l,bOO°F or is in excess of 1,600°F for
a substantial amount of the time, refractory ma-
terials are required.
Corrosion-Resistant Materials
tl (460 + 150)
The required exhaust rate:
1/3
= 420 fpm
V = v A
t e o
V = (420)(141)
= 59, 000 cfm
Check mean hood air temperature:
Since q = V p c At:
^•c t p
•where
p = average density of mixture, 0. 075 Ib/ft
c = average specific heat of mixture, 0. 24
P Btu/lb per °F
At = average hood temperature minus ambient
air temperature.
At -
70,OOP
(59,000)(0.075)(0. 24)
At = 80 + 66
m
= 66°F
= 146°F
This adequately approximates the original assump-
tion.
A variety of materials are available for corrosive
conditions. Plywoodis sometimes employed for
relatively light duty or for temporary installa-
tions. A rubber or plastic coating may some-
times be applied on steel. Some of these coat-
ings can be applied like ordinary paint. If severe
corrosion problems exist, hoods must be con-
structed of sheets of PVC (polyvinyl chloride),
fiberglas, or transite.
Design Proportions
Although the items of primary importance in de-
signing hoods are the size, shape, and location of
the hood face, and the exhaust rate, the depth of
the hood and the transition to the connecting duct
must also be considered. A hood that is too shal-
low is nothing more than a flanged-duct opening.
On the other hand, excessive depth increases the
cost without serving a useful purpose.
Transition to Exhaust Duct
It is desirable to have a transition piece between
the hood and the exhaust system ductwork that is
cone shaped with an included angle of 60° or less.
This can often be made a part of the hood itself.
The exact shape of the transition is the most im-
portant factor in determining the hood orifice
losses. Examples of good practice in this regard
are illustrated in Figure 17.
-------
44
DESIGN OF LOCAL EXHAUST SYSTEMS
POURING STATION FOR SMALL MOLDS
TRANSITION
PI E
CE-^-l./
^4-
y ^
T
^MIN
•— 5/3 b ••
MIN
•— b— ••
MOLD
if
<— WIDE
FLANCE
TOP BAFFLE
V = 200(10X2 + A)
where V = minimum ventilation rate, cfm
X = distance between hood and ladle, ft
A = face area of hood, ft2
ENCLOSURE FOR FOUNDRY SHAKEOUT
TRANSITION
PIECE
Provide a minimum indraft of 200 cfm per square
foot of opening but not less than 200 cfm per
square foot of grate area for hot castings.
Figure 17. Examples of good hood design.
Note use of enclosure, flanges, and transi-
tions (Industrial Venti lation, 1960).
DUCT DESIGN
The design of hoods and the determination of ex-
haust volumes have been considered. Now the de-
sign of the ductwork required to conduct the con-
taminants to a collection device will be discussed.
Calculations of pressure drop, system resistance,
system balance, and duct construction will be
covered.
GENERAL LAYOUT CONSIDERATIONS
Before designing and installing an exhaust system,
try to group together the equipment to be served
in order to make the system as small and com-
pact as possible and thereby reduce the resistance
load and power required. Extending an exhaust
system to reach an isolated hood or enclosure is
usually costly in regard to power consumption, and
if the isolated hood cannot be located close to the
main exhaust system, the installation of a separate
systemto care for the isolated equipment is prob-
ably preferable, in terms of operating economy.
When long rows of equipment must be served, the
main header duct should be located as near as
possible to the center of the group of equipment in
order to equalize runs of branch duct. Where nec-
essary, the equipment should be divided into sub-
groups and subheaders located to provide good
distribution of airflow in the duct system, and
proper velocities at the hood and enclosure inlets.
Air flowing in ducts encounters resistance due to
frictionand dynamic losses. Friction losses oc-
cur from the rubbing of the air along the surface
of the duct, whereas dynamic losses occur from
air turbulence due to rapid changes in velocity or
direction. From Bernoulli's theorem, the sum
of the static and velocity pressure upstream is
equal to the sum of the static and velocity pres-
sure plus the friction and dynamic losses down-
stream. A fan is normally required to provide
sufficient static pressure to overcome the resis-
tance of the system.
TYPES OF LOSSES
The losses in an exhaust system may be expressed
as inertia losses, orifice losses, straight-duct
friction losses, elbow and branch entry losses,
and contraction and expansion losses. In addition
to losses from the ductwork, there are also pres-
sure losses through the air pollution collection
equipment.
Inertia Losses
Inertia losses may be defined as the energy re-
quired to accelerate the air from rest to the ve-
locity in the duct. In effect, they are the velocity
pressure. Many other losses are expressed in
terms of velocity pressure, but velocity pressure
itself represents the energy of acceleration. It is
calculated by equation 5, set forth earlier. By
this equation, values of velocity pressure versus
velocity have been calculated, as shown in Table t>.
Orifice Losses
The pressure or energy losses at the hood or duct
entrances vary widely depending on the shape of
the entrance. The losses are due mainly to the
vena contracta at the hood throat. They are usu-
ally expressed as a percentage of the velocity pres-
sure corresponding to the velocity at the hood
throat. The losses vary from 1. 8 hv for a sharp-
edge orifice to nearly zero for a well-rounded bell-
mouth entry. Losses for common shapes of en-
tries are given in Figure 18. Most complicated
entries can be broken down into two or more sim-
ple entries, and the total entry loss computed by
adding the individual losses.
-------
Duct Design
45
Table 6. TABLE FOR CONVERSION
OF VELOCITY (va)
TO VELOCITY PRESSURE (hv)
va> £Pm
400
500
600
700
800
900
1,000
1, 100
1, 200
1,300
1, 400
1,500
1,600
1,700
1,800
1,900
2,000
2, 100
2, 200
2, 300
2, 400
2, 500
2,600
2, 700
2,800
2,900
3,000
3, 100
3,200
3, 300
hv, in. WC
0. 010
0. 016
0.022
0. 031
0. 040
0. 051
0. 062
0. 075
0. 090
0. 105
0. 122
0. 140
0. 160
0. 180
0.202
0.225
0.249
0. 275
0. 301
0. 329
0. 359
0. 389
0.421
0.454
0.489
0. 524
0. 561
0. 599
0. 638
0. 678
va, fpm
3, 400
3, 500
3, 600
3, 700
3, 800
3, 900
4, 000
4, 100
4, 200
4, 300
4, 400
4, 500
4, 600
4, 700
4,800
4, 900
5, 000
5, 100
5,200
5, 300
5,400
5, 500
5, 600
5, 700
5, 800
5, 900
6, 000
6, 100
6, 200
hv, in.WC
0.720
0.764
0.808
0.853
0. 900
0.948
0. 998
1.049
1. 100
1. 152
1.208
1.262
1.319
1.377
1. 435
1.496
1.558
1.621
t.685
1.751
1.817
1.886
1. 955
2.026
2.098
2. 170
2. 244
2.320
2.397
Straight-Duct Friction Losses
Many charts have been developed that give the
friction losses in straight ducts. Most of these
charts are based on new, clean duct. A resis-
tance chart in which allowance has been made for
moderate roughness of the duct is shown in Figure 1 9.
Most exhaust systems collecting appreciable a-
mounts of air contaminants are believed to reach
at least this degree of roughness in a relatively
short time after being placed in operation. Fric-
tionlossin inches of water per 100 feet of duct is
plotted in terms of duct diameter, velocity, and
volume. If any two of these quantities are given,
the other two can be read from the chart.
Elbow and Branch Entry Losses
The simplest way to express resistance of elbows
and branch entries is in equivalent feet of straight
duct of the same diameter that will have the same
pressure los s as the fitting. The equivalent lengths
are added to the actual lengths of straight duct, and
the resistance for each run computed from Figure 19.
Equivalent lengths of elbows and entries .are given
in Table 7.
Exhaust system calculator
Most of the charts, tables, and equations have
been incorporated into a single sliderule device,
as illustrated in Figure 20. The upper scales on
the front side will give friction losses just as in
Figure 19. Velocity pressure can be read from
the same scales by setting 4, 000 on the velocity
scale opposite 1.00 on the friction scale. Then,
opposite any other velocity, the friction scale
will give the correct velocity pressure. The lower
scales perform volume and velocity calculations
fora given duct diameter. Temperature correc-
tion scales and a duct condition correction scale
are provided. On the reverse side (not shown in
Figure 20), equivalent lengths for elbows, branch
entries, and weather caps are given. On the low-
er portion of each side, hood entry losses are
given for all the usual entry shapes.
PLAIN DUCT FLANGED DUCT
END END
TRAP OR SETTLING CHAMBER
H = 0.9H
H, = 0.5 H,,
STANDARD GRINDER HOOD
ORIFICE PLUS
HE - t .8 Hy HE = 1.8 Hv ORIFICE
FLARED ENTRY DIRECT BRANCH BELL-MOUTH ENTRY
BOOTH
?~|
S
H, - 0.15H,,
Hr * 0.025 H.,
Hr = 0.5H.,
TAPERED HOODS
^
30
4 5
6 0
9 0
I 2 0
t 5 0
E
0.15 H
0.08 H
0 .06 H
0 0 B H
0.15 H
0.26 H
0 .4 0 H
0.25 H
0.16 H
0.15 H
0 I 7 H
0.25 H
0,35 H
0 . A 0 H
0
0 .
0 .
0 .
0 .
0 . 8
0 . B
~3 0~
0
0
0
0
0
0
V"
9
9
9
e
B
a
C 1 E N T
Figure 18. Hood entry losses (Adapted from
Industrial Ventilation, 1956).
-------
46
DESIGN OF LOCAL EXHAUST SYSTEMS
1 00,000
100
0
10 0 20 0 30 0.40 0 50 0.70 1 .
FRICTION, inches
0 2 3 45678910
of water/I 00 feet
Figure 19. Friction loss chart.
-------
Duct Design
47
Table 7. AIR FLOW RESISTANCE CAUSED BY ELBOWS AND BRANCH ENTRIES
EXPRESSED AS EQUIVALENT FEET OF STRAIGHT DUCT
(Adapted from Industrial Ventilation, 1956)
Diameter
of duct,
in.
3
4
5
6
7
8
C)
10
11
12
14
16
18
20
22
24
26
28
30
36
40
48
90° Elbow
/
/
-D.
Throat radius (R)
1. 0 D
5
7
9
11
12
14
17
20
23
25
30
36
41
46
53
59
64
71
75
92
105
130
1. 5 D
4
5
6
7
9
10
12
13
16
17
21
24
28
32
37
40
44
49
51
63
72
89
2. 0 D
3
4
5
6
7
8
10
11
13
14
17
20
23
26
30
33
36
40
42
52
59
73
i
i
x^
/
( R
\
60° Elbow
Throat radius(R)
0 D
4
5
7
8
9
1
12
14
17
20
23
27
32
36
39
44
48
52
55
6
8
75
91
1.5 D
3
4
5
5
6
7
9
10
12
13
16
18
22
24
27
30
33
35
38
46
51
62
2.0 D
2
3
4
4
5
6
7
8
10
11
13
15
18
20
22
25
27
29
31
38
42
51
«;
e^
— — —
X
s^ jf
45° Elbow
Throat radius (R)
r i. o D
2
4
5
6
7
8
9
10
11
12
14
17
20
23
27
30
32
35
37
46
52
64
1.5 D
1
3
4
4
5
5
6
7
8
9
10
12
13
16
18
20
22
24
26
32
35
44
2. 0 D
1
2
3
3
4
4
5
6
6
7
8
10
11
13
15
16
18
20
21
26
29
36
(i
Branch entry
Angle of entry (9)
45°
3
5
6
7
9
11
12
14
15
18
21
25
28
32
36
40
44
47
51
-
-
-
30°
2
3
4
5
6
7
8
9
10
11
13
15
18
20
23
25
28
30
32
-
-
-
15°
1
1
2
2
3
3
4
4
5
5
6
8
9
10
11
13
14
15
16
-
-
-
Contraction and Expansion Losses
When the cross -sectional area of a channel through
whicha gas is flowing contracts, a pressure loss
is encountered. The magnitude of the loss de-
pends upon the abruptness of the contraction.
When the cross-sectional area expands, a portion
of the decrease in velocity pressure may be con-
verted into static pressure. The increase or de-
crease in pressure from expansion and contraction
can be calculated from the diagrams and formulas
given in Tables 8 and 9. Losses from small
changes in velocity can be neglected.
data, resistances are usually estimated by com-
paring them with known values for similar equip-
ment. In collectors such as cyclones and scrub-
bers -where the velocities are high, pressure varies
approximately with the square of the velocity. If
the loss is known at one velocity, the loss at any
other velocity is computed by multiplying by the
square of the ratio of the velocities. In cloth fil-
ter dust collectors, however, the flow is laminar,
and pressure drop varies approximately as the
first power of the velocity ratios.
DESIGN PROCEDURES
Collection Equipment
Pressure through collection equipment varies
widely. Most manufacturers supply data on pres-
sure drop for their equipment. In the absence of
Methods of Calculation
The first step in designing an exhaust system is
to determine the volume of air required at each
hood or enclosure to ensure complete collection
-------
48
DESIGN OF LOCAL EXHAUST SYSTEMS
MI- C * **
90O243
HOOD ENTRY LOSSES
Figure 20. Exhaust system calculator.
of the air contaminants, by using the principles
given previously. The required conveying velocity
is then determined from the nature of the con-
taminant. Table 10 can be used in determining
conveying velocities.
The branch duct and header diameters are then
calculated to give the minimum conveying velocity.
When the calculated diameter lies between two
available diameters, the smaller diameter should
be chosen to ensure an adequate conveying velocity.
The duct layout is then completed, and the lengths
of ducts and number and kinds of fittings deter-
mined. The system resistance can then be com-
puted. The calculations can be most easily accom-
plished by using a tabular form such as those shown
later.
Methods of Design
In designing a system of ductwork with multiple
branches, the resistance of each branch must be
adjusted so that the static pressure balance, which
exists at the junction of two branches, •will give
the desired volume in each branch. In general,
two methods of accomplishing this result are used:
1. The balanced-duct or static pressure balance
method, in which duct sizes are chosen so that
the static-pressure balance at each junction will
achieve the desired air volume in each branch
duct.
2. The blast gate adjustment method, in which cal-
culations begin at the branch of greatest resis-
tance. The other branches are merely sized
-------
Duct Design
49
to give the minimum required velocity at the
desired volume. Blast gates are provided in
eachbranch, and after construction, the gates
are adjusted to give the desired volume in each
branch duct.
The balanced-duct system is less flexible and
more tedious to calculate, but it has no blast
gates that might collect deposits or be tampered
with by unauthorized persons. Layout must be
in complete detail and construction must follow
layout exactly.
The blast gate system has more flexibility for
future changes and is easier to calculate; vol-
umes can be adjusted within certain ranges,
and duct location is not so critical.
Calculation Procedures
The balanced-duct method: The calculations for a
balanced-duct system start at the branch of greatest
resistance. Using the duct size that will give the
required volume at the minimum conveying ve-
locity, calculate the static pressureup to the junc-
tion with the next branch. The static pressure is
then calculated along this next branch to the same
junction. If the two calculations agree within 5 per-
cent, the branches may be considered in balance.
Table 8. DUCTWORK DESIGN DATA SHOWING
STATIC PRESSURE LOSSES AND REGAINSa
THROUGH ENLARGING DUCT TRANSITIONS
(Industrial Ventilation, 1962)
Taper angle (0),
degrees
3-l/Z
5
10
15
20
25
30
over 30
X
8.13
5.73
2.84
1.86
1.38
1.07
0.87
Regain factor (R)
0.78
0.72
0.56
0.42
0.28
0.13
0.00
0.00
Loss factor (L)
0.22
0.28
0.44
0.58
0.72
0.87
1.00
1.00
aThe regain and loss factors are expressed as a fraction of
the velocity pressure difference between points (1) and (2).
In calculating the static pressure changes through an en-
larging duct transition, select R from the table and sub-
stitute in the equation
where
hy is
(+)
SP is (+) in discharge duct from fan
SP is (-) in inlet duct to fan
Table 9. DUCTWORK DESIGN DATA SHOWING CONTRACTION
PRESSURE LOSSES THROUGH DECREASING DUCT TRANSITIONS
(Industrial Ventilation, 1956)
Taper angle (9),
degrees
5
10
15
ZO
25
30
45
60
X
D1-D2
5. 73
2.84
1.86
1.38
1. 07
0.87
0. 50
0. 29
Loss fraction (L)
of hv' difference
0. 05
0. 06
0. 08
0. 10
0. 11
0. 13
0.20
0. 30
For abrupt contraction (6 > 60°)
Ratio D2/DL
0. 1
0.2
0. 3
0.4
0. 5
0. 6
0. 7
Factor K
0.48
0.46
0.42
0.37
0.32
0.26
0.20
SP change:
SP =SP -(h
2 1
- h ) - L (h - h )
: vi V2 vi
SP change:
SP =SP - (h
2 1
-h )-K(h
-------
50
DESIGN OF LOCAL EXHAUST SYSTEMS
Table 10. RECOMMENDED MINIMUM DUCT VELOCITIES
(Brandt, 1947)
Nature of contaminant
Example s
Duct velocity, fpm
Gases, vapors, smokes,
fumes, and very light
dusts
Medium-density dry dust
Average industrial dust
Heavy dusts
All vapors, gases,
and smokes; zinc
and aluminum ox-
ide fumes; wood,
flour, and cotton
lint
Buffing lint; saw-
dust; grain, rub-
ber, and plastic
dust
Sandblast and
grinding dust,
wood shavings,
cement dust
Lead, and foundry
shakeout dusts;
metal turnings
2, 000
3, 000
4,000
5,000
If the difference in static pressure is more than
20 percent, a smaller diameter duct should be
used in the branch with the lower pressure drop
to increase its resistance. When the difference
in pressure loss in the two branches is between
5 and 20 percent, balance can be obtained by in-
creasing the flow in the branch with the lower loss.
Since pressure losses increase as the square of
the volume, the increased volume can be readily
calculated as:
Corrected cfm =
/h larger
' s
h lower
s
(22)
The pressure loss in the header is then calculated
to the next branch. This branch is then sized to
achieve a static pressure balance at this junction
with the required volume (or slightly greater) in
the branch duct. This procedure is continued un-
til the discharge point of the system is reached.
The blast gate adjustment method: The calcula-
tions for a system to be balanced by blast gate ad-
justment also start at the branch of greatest resis-
tance and proceed to the header. Pressure losses
are then calculated only along the header. Pres-
sure drops in the remaining branches are not cal-
culated except when calculation is deemed advisable
in order to check a branch to be sure its pressure
drop does not exceed the static pressure at its
junction with the header.
Fan Static Pressure
The preceding calculations are based on static
pressure; that is, the balancing or governing pres-
sures at the duct junctions are static pressures.
Most fan-rating tables are given in terms of fan
static pressure. The National Association of Fan
Manufacturers defines the fan static pressure as
the total pressure diminished by the velocity pres-
sure at the fan outlet, or
fan h = fan H - h fan outlet (23)
s v
On the absolute pressure scale,
fan H = H outlet - H inlet (24)
Combining the two equations
fan h = H outlet - H inlet - h outlet
= h outlet + h outlet - (h inlet + h inlet)
s v s v
- h outlet
v
= h outlet - h inlet - h inlet
s s v
(25)
Static pressures are nearly always measured rel-
ative to atmospheric pressure, and static pressure
at the fan inlet is negative. In ordinary usage,
only the numerical values are considered, in which
case, equation 25 becomes
-------
Duct Design
51
fan h = h outlet + h inlet -
s s s
h inlet
v
(26)
In evaluating the performance of a fan, examine
the tables to determine whether they are based on
fain static pressure or on total pressure.
Balanced-Duct Calculations
A problem illustrating calculation by the'balanced-
duct method is worked out as follows. The given
operation involves the blending of dry powdered
materials. A sketch of the equipment is given in
Figure 21. The equipment and ventilation require-
ments are presented in Tables 11 and 12,
A minimum conveying velocity of 3, 500 fpm is to
be maintained in all ducts. Elbows have a throat
radius of 2 D. The balanced-duct method is to be
used in the duct design. The detailed calculations
are shown in Table 13. Calculations start at
branch A. A 6-inch-diameter duct gives the near-
est velocity to 3, 500 fpm at the required volume
of 750 cfm. The actual velocity of 3, 800 fpm is
entered in column 5, and the corresponding ve-
locity pressure, in column 6. From Figure 19,
the entry loss is 50 percent hv, which is entered
in column 7 left. The length of straight duct is
entered in column 8. The equivalent length for the
elbows is found in Table 7, and the sum is entered
in column 9 right. The total equivalent length is
then found by adding column 8 and column 9 right
and entering the sum in column 11. The resis-
tance per 100 feet of duct is then read from Fig-
ure 20 at 6-inch diameter and 3, 800 fpm, and is
entered in column 12. The resistance pressure
(hr) is calculated by multiplying column 11 by
column 12 and dividing by 100. This value is en-
tered in column 13. The static pressure is then
the sum of the velocity pressure and the hood loss
plus the resistance pressure, c olumn 6 -f column 7
right + c olumn 13.
In branch B, a volume of 200 cfm is required. A
3-1/2-inch duct would give a velocity of 3,000 fpm,
which is below the minimum. Hence the branch
was calculated with a 3-inch duct at 4, 000 fpm.
The resulting hg (column 14 left) was more than
20percent greater thanthatfor branch A. A 3-1/2-
inch duct must, therefore, be used and the volume
increased to 240 cfm to maintain the minimum ve-
locity. At these conditions, the hg values for the
two branches are within 5 percent and may be con-
sidered in balance.
In section C, a 7-inch duct will carry the combined
volume from branches A and B at the nearest ve-
locity above the minimum. The only pressure drop
Figure 21. Sketch of exhaust system used
in Table 11 showing duct design calcula-
tions by the balanced-duct method.
Table 11. EQUIPMENT AND VENTILATION REQUIREMENTS FOR
BLENDING DRY POWDERED MATERIALS
Equipment
Dump hopper (1),
2- by 3 -ft opening
Bucket elevator (2),
1 - by 2 -ft casing
Ribbon blender (3),
1- by 2 -ft opening
Drum-filling booth (4),
1- by 3 -ft opening
Cloth filter dust collector (5)
maximum resistance, 4 in.wg
Ventilation requirement
125-fpm indraft
through opening
100 cfm per ft2
of casing area
150-fpm indraft
through opening
200-fpm indraft
through opening
Volume, cfm
750
200
300
600
-------
52
DESIGN OF LOCAL EXHAUST SYSTEMS
Table 12. DUCT LENGTHS AND FITTINGS
REQUIRED IN BLENDING DRY
POWDERED MATERIALS
Branch
A
B
C
D
E
F
G
Length, ft
14
4
7
6
3
11
12
Elbows, No.
and degree
2, 90
1, 60
1, 90
1, 60
2, 90
1, 60
Entries, No.
and degree
1, 30
1, 30
1, 30
2, 90
is due to the friction in the 7 feet of straight duct.
This hr is added to the hg at the first junction. In
branch D, 300 cfm is required, and this volume
will give approximately 3, 450fpm in a 4-inch duct.
The resulting hg is, however, about 20 percent
lower than the h- at the main. The cfm must,
o
therefore, be increased by the ratio of the square
roots of the static pressures, or from 300 to 350 cfm.
In main duct E, an 8-inch diameter duct will handle
a volume of 1, 340 cfm at a velocity of 3, 800 fpm.
An hr of 0. 10 inch WC is recorded in column 13,
giving an hs of 2. 87 inches WC to the junction EF.
Calculation procedures for branch F are similar
to those for branch E, and the required 600 cfm
for the drum-filling booth must be increased to
640 cfm to obtain a static pressure balance at
junction EF.
The total volume of 1, 980 cfm gives a velocity of
3,600 fpm in a 10-inch-diameter main duct G to
thebaghouse. This run of duct and two 90° elbows
have a resistance pres sure of 0.88 inch WC, giv-
ing a total inlet static pressure of 7.75 inches WC,
after the given resistance of the baghouse is added.
The outlet static pressure is calculated similarly
by calculating the resistance of the straight run
of duct H. This static pressure of 0.21 inch WC
is added to the inlet static pressure. The velocity
pressure of 0. 81 inch WC (one hv at a velocity of
3, 600 fpm at the fan inlet) is subtracted from the
above total static pressure to yield a fan static
pressure of 7. 15 inches WC.
Blast Gate Method
The same system can be designed by the blast gate
adjustment method. The calculations are shown
in Table 14. Branch A is calculated as before.
Branches B, D, and F are calculated at or near
the minimum conveying velocity so that the hg
drop in each does not exceed the hg at the junction
with the main. No adjustments are made in the
volumes. Blast gates will be installed in each of
these branches to provide the required increase
in resistance.
CHECKING AN EXHAUST SYSTEM
The preceding example problem illustrates the
calculations for designing an exhaust system. In
checking plans for an exhaust system, use similar
calculations but take a different approach. A sys-
tem of duct-work with a specific exhauster is given
and the problem is to determine the flow conditions
that will exist.
The objectives of checking an exhaust system are:
1. To determine the exhaust volume and indraft
velocity at each pickup point and evaluate the
adequacy of contaminant pickup;
2. to determine the total exhaust volume and eval-
uate the size and performance of the collector
or control device;
3. to determine the system's static pressure and
evaluate the fan capacity, speed, and horse-
power required;
4. to determine the temperature at all points in
the system in order to evaluate the materials
of construction of the ductwork and the collector.
Illustrative Problem
To illustrate a method of checking an exhaust sys-
tem, another problem is -worked. A line drawing
of the duct-work is given in Figure 22. None of the
calculations used in designing the system are
given. Since no blast gates are shown, assume
that the system -was designed by the balanced-duct
method.
Resistance calculations are presented in Table 15.
This form was designed for maximum facility in
checking an exhaust system. Calculations start
at hood A -with an assumed velocity (or volume) of
3, 500 fpm. The static pressure drop is then com-
puted to Junction_C_. Branch B-C is then computed
•with an assumed velocity of 3, 500 fpm. Since the
hs from this branch does not match that from
branch A-C, the second velocity is corrected by
multiplying by the square root of hs A-C/hs B-C.
The corrected velocity is entered in column 14 and
is used to compute the cfm, which is entered in
column 1 5. The other branches are calculated in
the same manner, that is, assume a velocity at
the hood and correct it by the square root of the
hs ratio. Thus, all the calculations are related
to the original assumption of velocity in the first
branch.
-------
Duct Design
53
Table 13. EXHAUST SYSTEM CALCULATIONS BY BALANCED-DUCT METHOD
2
£
s
"
a
-
o
*
CO
r-
-
in
-
«
«,
-
Temp in
main,
"F
O M "Tj
U f> >
c
*
h
^
01 k
W o
S.c H
*S *
U rt
3 * i?
w o
C r-
j;
Resistance
per
100 ft
c **
o £ ti
H if 3
c £
>• 0)
£ 25
J5 If
c W "
rt . 01
M O T3 *«]
2 § ™
Elbows
Equivalent
length, ft
IS?
S S o"
J«.g
O
. *
5 .S
O X
2*
Jig
II
1 -3
0 S
o> j:
C U
1 s
3 £
> «
IP O.
| IS-
O'S g
™ N ^ r- t~ t
"> "] m r- t- r
^1 f'xl ro f\] IN] r
f\i co vo — H r
m o in o T
*\j m r^J INJ r
r- oo ^T in 0
— i o CT- rg CT-
-t -** O O O*
m co r*i *& -*o
T o^ r^ m m
xo -^ ro vO
(VI ^ ^H —1
rsj ro ro 6
h
01
a | '« "« [P:
° 1
c
r\j m n ^0 r-
"* «
w
(4
0>
...{;.. . .
OOO c°° OO
O^CT-vO S O^vO CT-xD
O
fM
C
v •* rt * r* so
£ t
S '
IT) O ^ ^H ^
O ^H" Ml CD 0
V
U
0 0 C 0 0
m o *" o ir>
~* V "*
^ ?
O 13 -H •* 1
Ox o oj °° ^
O -H -fi O C3
O O OOO
O O 0 O tfl
oo o so r- t
o
O4-
o^
o o o o <
mo •* o i
r- M IM to r
•o fo ro r» •*
< CQ CQ U Q
— r*- Is- r* in in
- oo co oo r- Is-
•j fM f\j CM fi r*-
- tn r-
m co
si (M* ^J
O r-'
30) ^
^ n i "
i a a _
ui « M V, S
6 5 ~ ' 5 "-
" « n 3 °
E S •
S . . f . 0 j." ^
OO° O C INI
4> 0^ xD ^ (^ « .
i ~- i - S t °
£ S » js +
s ^ 's P
=- " - C 2 S °° J=° r-'
4> y
3 4) u II II
M S •-*
i si s s
s o .= 5 :
V V T) *
U U , 0*
s s s 3 5
M * ii £
U *S ^ *
a 3 ^ «
3 "• *O •"» *J r1
» „ » ^ §
c o # o" y fc
o o o o
O O 0 O
CO xO xO xO
o o o
•^* oo oo
co cr- cr*
D O O
o o ^r
"> xO xO
-.foj
OO UI O O
W U* O X
^
6
II
c"
•5
4
*O
2
n
£
**"
1 '
«5
^ 1
A
-------
54
DESIGN OF LOCAL EXHAUST SYSTEMS
Q
O
ffi
H
H
W
H
<
o
i
H
W
W
W
ffi
H
W
w
2
O
f-H
H
P
U
O
S
W
H
w
H
w
ffi
X
W
•5
H
£
2
S
fNJ
-
0
o-
oo
r~
-0
«
*
*
-
o h° "a
U -° >
rt
V
IH
BQ
a
(U
D.
o
to
.3
s
U.
H
« h
o
n)
U
.3 * £
IH ^
Jd
Resistance
per 100
ft
a 41
^ 4)
j! "3 x
H||
0- V
4) ^
Branch entry
Elbowe
Equivalent
length, ft
o>
bo
Equivalent
length, ft
llf
tLs
"•s--
11 S 0
J *5 3
" « -o
V
V
o
?
o
K
U
.3
5"
j^ U.
1|
>
Volume, cfm
Branch Main
s IN
0 g •
ts s .
"S-g g
£
3
x
U
A
CJQ
J=
(B
V
§
-------
Duct Design
55
W
H
w
>H
w
H
w
ID
X
W
O
^
t-H
W
U
w
K
U
rt
o
h
w
Z
O
t— I
H
ID
O
2
5
2
-
-
0
CT>
CO
«,
»
m
«,
-
i
c
j
!-
g
U
Corrected
0)
LI
09
w
a
u
a
a
(.
£
i
1
i
a
a
1
h
a
01
43
U
J3
.S
E *
'rt
S
Branch
Ifl
•° S
.s
<4
s
H
C Ui
s?1
.s ^
j
•
L> *-•
0.0
o
-H
* J *
3 C
O1 4)
0) -^
Equivalent
length, ft
4)
CO
Equivalent
length, ft
o *2 *i
2 5 c
rt
D .60 _
SI
u
.S
*"
* b<
; *j «
* 0
£
'u c
•§!
s f h
J 5 *
i § gg
branch
or
main
ooo ooooo
rvJOO ^O-£)OOO
(*10000 (^(NJsD^^O
o o ooo o
o ooo o
o O O r\i O
5 S :• S ?• :• 5 5
,— t s£) fH. ^H |s_ (V^
rvj rsj M rsl c*i fl
^00-^1 o^r-co
«• - • _• _• o o- -; o
"• v °- "• *• v * *• "•
OVMOOO o
r- r- r- r- r- oo o
o" o o" o" o" o* o
ooooo ooooo
ooooo ooooo
.5
° '•O
c
1
.3
H
S
M
1
|
g
I)
.C
4
o
*J V
=
-------
56
DESIGN OF LOCAL EXHAUST SYSTEMS
E D
n, rr
4 in. I 3 in
*~v
ALL HOODS AftE STANDARD GRINDING HOODS.
ORIFICE LOSS = 0.65 hy
ELBOW THKOAl RADIUS - 2 D
BRANCH ENTRIES - 30°
NO DAMPERS OR BLAST GATES USED
Figure 22. Layout of ductwork used in
example showing procedure in checking
an exhaust system
Since assumed values of volume are used in the
calculations, the final result is the system's resis-
tance at a given total volume. The system's resis-
tance will increase with increase in volume by the
square of the ratio of the increase in volume. On
the other hand, the capacity of an exhauster de-
creases with increase in resistance. The one
point that satisfies both system and fan can best
be found by plotting the characteristic curves of
the system and the fan. The operating point is
the point of intersection of the two curves. The
system's characteristic curve is that curve
established by the static pressure losses through
the exhaust system for various air volumes. It
is computed by starting with the resistance and
volume from Table 15 and calculating the resis-
tance at other volumes by using the square of
the ratio of the volume change. The curve for
the sample problem is computed from Table 16.
Table 16. CALCULATIONS
FOR CHARACTERISTIC CURVE
Volume, cfm
1,600
1,800
2,000
1,400
1,200
1,000
500
Multiplying factor
(1, 800/1, 600)2 x 4.84
(2,000/1,600)2 x 4.84
(1, 400/1, 600)2 x 4.84
(1,200/1,600)2 x 4.84
(1, 000/1, 600)2 x 4.84
(500/1, 600)2 x 4. 84
New hs
4. 84
6. 12
7.56
3. 71
2. 72
1.89
0.47
Fan hs is used in computing this curve because
the fan, a Chicago No. 25 Steel Plate Exhauster,
is rated by the methods of the National Associa-
tion of Fan Manufacturers (NAFM). The fan
characteristic curves are families of curves at
different fan speeds defining static pressures de-
veloped for various volumes of air handled through
the fan. These data are available from fan manu-
facturers. Data for the single fan curve at 2, 600
rpmas specified by this example are obtained from
Chicago Blower Corporation Bulletin SPE-102.
Fan capacities at various static pressures and at
the given speed nearest to 2, 600 rpm are tabulated
on the left in Table 17; on the right the figures are
corrected to 2, 600 rpm by use of the fan laws, as
follows:
rpm
cfm = cfm x
rpm
.
S2
— li x
rrpm]
hp = hp x
2 rl | rpm.
A horsepower versus air volume curve can also
be plotted from Table 17.
Table 17. FAN CAPACITY AT VARIOUS
STATIC PRESSURES
From Chicag
rpm
2,630
2,615
2,605
2,625
2,620
cfm
2, 470
2,240
2, 005
1, 655
1, 065
lo Bxilletin
hs
5
4
6
7
8
hp
4. 44
3.89
3. 44
3.09
2.54
Corrected to 2, 600 rpm
cfm
2, 440
2, 230
2, 000
1, 640
1, 050
hs
4.9
5. 4
6.0
6.9
7.8
hp
4.3
3.8
3.4
3.0
2.5
The system curve, fan curve, and horsepower
curve are plotted in Figure 23. The fan and sys-
tem curves intersect at 1, 860 cfm and 6. 5 inches
h . The horsepower required is 3.2.
Since the volume obtained from the curves of Fig-
ure 23 is appreciably higher than the total volume
from Table 15, the volume at each hood must be
corrected. The correction factor is obtained by
dividing the vohime from the curve by the total
volume from Table 15. Corrections are made in
Table 18.
-------
Duct Design
57
500
1,000 1,500 2,000
VOLUME, cfm
2,500 3,000
Figure 23. Characteristic curves of an
exhaust system.
Fan Curve Calculator
The calculations required to
produce a fan curve from catalog
data have been incorporated in a
slide rule-type calculator (Fig-
ure 24). A calculator of this
type will reduce the time re-
quired to plot characteristic
curves such as shown in Figure
23.
CORRECTIONS FOR TEMPERATURE
AND ELEVATION
Fan tables, resistance charts,
and exhaust volume require-
ments are based on standard
atmospheric conditions of 70°F
and average barometric pres-
sure at sea level. Under these
conditions the density of air is
0. 075 pound per cubic foot.
Where conditions vary appre-
ciably from standard condi-
tions, the change in air den-
sity must be considered.
Figure 24. Fan curve calculator.
-------
58
DESIGN OF LOCAL EXHAUST SYSTEMS
Table 18. CORRECTIONS FOR HOOD VOLUME
Hood
A
B
D
E
H
Volume from
Table 15, cfm
320
480
170
290
340
Correction
factor
1,860/1,600
1,860/1, 600
1,860/1, 600
1,860/1,600
1,860/1, 600
Corrected
volume, cfm
370
560
200
340
390
The density of air varies inversely with absolute
temperature and directly with barometric pres-
sure. Both effects are combined in the density
correction factors given in Table 19.
Velocity pressure, static pressure, and resis-
tance pressure vary directly with gas density. In
calculating a system, if the temperatures in all
the ducts are approximately the same (within 25 °F),
compute the entire system's resistance as at stan-
dard conditions and correct the final system's
static pressure by multiplying by the density cor-
rection factor. If the temperatures in the different
branches vary, the static pressure in each branch
must be corrected.
A centrifugal fan connected to a given system will
exhaustthe same volume regardless of gas density.
The-weight of air exhausted will, however, be di-
rectly proportional to the density, and so will the
static pressure developed and the horsepower
consumed.
In selecting an exhauster from multirating tables
to move a given volume of air at a given static
pressure and at a given temperature and altitude,
proceed as follows:
1. Read the density correction factor from Table
19.
2. Divide the given hs by the correction factor.
3. Select the fan size and rpm based on the given
volume and the corrected static pressure.
4. Multiply the horsepower (given by the above se-
lection) by the density correction factor to ob-
tain the required horsepower.
Table 19. DENSITY CORRECTION FACTORSa
Temp,
°F
0
40
70
100
120
140
160
180
200
250
300
350
400
450
500
550
600
650
700
750
800
850
900
950
1,000
Altitude, ft above sea level
0
1. 15
1.06
1. 00
0.96
0.92
0.88
0.85
0.83
0.80
0. 75
0. 70
0.65
0.62
0.58
0. 55
0. 53
0. 50
0.48
0.46
0.44
0.42
0.40
0. 39
0. 38
0. 36
1, 000
1. 10
1.02
0. 96
0.91
0.88
0.85
0. 82
0.79
0. 77
0. 72
0. 67
0.62
0.60
0.56
0. 53
0.51
0.48
0.46
0. 44
0.42
0.40
0.38
0. 37
0.36
0.35
2,000
1.06
0.98
0. 93
0.88
0.85
0.82
0. 79
0. 77
0.74
0. 70
0. 65
0.60
0. 57
0. 54
0.51
0.49
0.46
0. 44
0.43
0.41
0.39
0.37
0. 36
0. 35
0.33
3,000
1.04
0.94
0.89
0.84
0.81
0.79
0.76
0. 74
0.71
0. 67
0.62
0.58
0.55
0.52
0.49
0. 47
0.45
0. 43
0.41
0.39
0.37
0.36
0.35
0. 34
0.32
4, 000
0.99
0.91
0.86
0.81
0.78
0. 76
0.74
0.71
0.68
0. 64
0.60
0.56
0.53
0. 50
0.47
0.45
0.43
0.41
0.39
0.38
0.36
0.34
0.33
0. 32
0. 31
5,000
0.95
0.88
0.83
0.78
0.75
6, 000
0. 92
0. 84
0.80
0.75
0.72
0.73 0.70
0.70 0.68
0.68 0.66
0.66
0.62
0.58
0.54
0.51
0.48
0.45
0.44
0.41
0.40
0.38
0.36
0.35
0.33
0.32
0.31
0.30
0.64
0. 60
0. 56
0. 52
0.49
0.46
0.44
0.42
0.40
0.38
0.37
0.35
0. 33
0. 32
0. 31
0. 30
0.29
7, 000
0.88
0.81
0. 78
0. 72
0. 70
0.68
0. 65
0.63
0.61
0. 58
0. 54
0.50
0.47
0. 44
0.42
0.40
0. 38
0. 36
0.35
0. 33
0.32
0. 31
0. 30
0.29
0.28
8, 000
0.85
0. 79
0. 74
0. 70
0.67
0.65
0.63
0. 61
0.59
0.55
0.52
0.48
0.44
0.42
0.40
0.38
0.36
0.34
0.33
0.32
0.31
0.30
0.29
0.28
0.27
aDensity in lb/ft3 = 0. 075 x density factor.
-------
Duct Design
59
The procedure for plotting the fan's characteristic
curve at other than standard conditions is as follows:
1. Correct the values of cfm, hg, and hp from the
tables to the given rpm.
2. Multiply the values of hs and hp by the density
correction factor.
3. Plot values of hs and hp from Step 2 against
values of cfm from Step 1.
DUCT CONSTRUCTION
Correct design and competent installation of sheet
steel ducts andhoods are necessary for the proper
functioning of an exhaust system. The folio-wing
construction and installation practices are recom-
mended (Industrial Ventilation, 1956):
1. All exhaust systems should be constructed of
new materials and installed in a permanent
and workmanlike manner. Interior of all ducts
should be smooth and free from obstructions,
with joints either -welded or soldered airtight.
2. Ducts should be constructed of galvanized sheet
steel riveted and soldered or black iron welded,
except where corrosive gases or mists or other
factors make, such materials impractical. Gal-
vanized construction is not recommended for
temperatures above 400°F. Welding of black
iron of 18 gauge and lighter is not recommend-
ed for field fabrication.
3. For average exhaust on noncorrosive applica-
tion, the folio-wing gauges should be used for
straight duct:
U. S, Standard gage
Class I Class II Class III
24
22
20
IS
22
20
18
16
20
18
16
14
Duct diameter
To 8 in.
8- to 18 in.
19 to 30 in.
Over 30 in.
Class I. Includes nonabrasive applications, such
as paint spraying, woodworking, food
products, and discharge ducts from dust
collectors.
Class II. Includes nonabrasive material in large
concentration, moderately abrasive ma-
terial in small to moderate concentra-
tions, and highly abrasive material in
small concentration.
Class III. Includes all highly abrasive material in
moderate to heavy concentrations and
moderately abrasive material in heavy
concentration.
Brown and Sharpe gage numbers are used
to indicate thickness of aluminum sheet as
compared with U. S. Standard gages for
steel sheet. When aluminum duct is indi-
cated, the following equivalent B and S
gages should be used:
Steel - U. S. Standard gage
26 24 22 20 18 16 14
Aluminum - B and S gage
24 22 20 18 16 14 12
4. Elbows and angles should be a minimum of
two gauges heavier than straight sections of
the same diameter.
5. Longitudinal joints of the ducts should be lappe
and riveted or spot-welded on 3-inch centers
or less.
6. Girth joints of ducts should be made -with the
lap in the direction of airflow. A 1-inch lap
should be used for ducts to 19-inch diameters
and 1-1/4-inch laps for diameters over 19
inches.
7. All bends should have an inside or throat ra-
dius of two pipe diameters -whenever possible,
but never less than one diameter. Large ra-
dii bends are recommended for heavy concen-
trations of highly abrasive dust. Ninety de-
gree elbows not over 6 inches in diameter
should be constructed of at least five sections,
and over 6-inch diameter of at least seven
sections.
8. The duct should be connected to the fan inlet
by means of a split-sleeve drawband at least
one pipe diameter long, but not less than 5
inches.
9. Transition in main and submains should be
tapered, -with a taper of about 5 inches
for each 1-inch change in diameter.
10. All branches should enter main at the large
end of transition at an angle not to exceed
45°, preferably 30° or less. Branches should
be connected only to the top or sides of main,
never to the bottom. Two branches should
never enter a main at diametrically opposite
points.
11. Dead-end caps should be provided on mains
and submains about 6 inches from the last
branch.
12. Cleanout openings should be provided every
10 or 12 feet and near each bend or duct
junction.
-------
60
DESIGN OF LOCAL EXHAUST SYSTEMS
13. The ducts should be supported sufficiently so
that no load is ever placed on connecting e-
quipment. Ducts 8 inches or smaller should
be supported at least every 12 feet, and larg-
er ducts, at least every 20 feet.
14. A minimum clearance of 6 inches should be
provided between the ducts and ceilings, walls,
or floors.
15. Blast gates used for adjusting a system should
be placed near the connection of branch to
main, and means provided for locking them
in place after the system has been balanced.
16. Round ducts should be used wherever possi-
ble. Where clearances prevent the use of
round ducts, rectangular ducts as nearly
square as possible may be used.
FAN DESIGN
Fans are used to move air from one point to anoth-
er. In the control of air pollution, the fan, blower,
or exhauster imparts movement to an air mass
and conveys the air contaminants from the source
of generation to a control device in which the air
contaminants are separated and collected, allow-
ing cleaned air to be exhausted to the atmosphere.
Fans are divided into two main classifications:
(1) radial -flow or centrifugal type, in which the
airflow is at right angles to the axis of rotation of
the rotor, and (2) axial-flow or propeller type, in
which the airflow is parallel to the axis of rota-
tion of the rotor.
CENTRIFUGAL FANS
A centrifugal fan consists of a -wheel or rotor
mounted on a shaft that rotates in a scroll-shaped
housing. Air enters at the eye of the rotor, makes
a right-angle turn, and is forced through the blades
of the rotor by centrifugal force into the scroll-
shaped housing. The centrifugal force imparts
static pressure to the air. The diverging shape
of the scroll also converts a portion of the velocity
pressure into static pressure.
Centrifugal fans may be divided into three main
classifications as follows:
1. Forward-curved-blade type. The rotor of the
forward-curved-blade fan is known as the
squirrel-cage rotor. A solid steel backplate
holds one end of the blade, and a shroud ring
supports the other end. The blades are shal-
low with the leading edge curved towards the
direction of rotation. The usual number of
blades is 20 to 64.
2. Back-ward-curved-blade type. In the back-
ward-curved-blade fan, the blades are in-
clined in a direction opposite to the direc-
tion of rotation, and the blades are larger
than those of the forward-curved-blade fan.
The usual number of blades is 14 to 24, and
they are supported by a solid steel backplate
and shroud ring.
3. Straight-blade type. The blades of the
straight-blade fan may be attached to the
rotor by a solid steel backplate or a spider
built up from the hub. The rotors are of
comparatively large diameter. The usual
number of blades is 5 to 12. This classifi-
cation includes a number of modified de-
signs whose characteristics are, in part,
similar to those of the forward- and back-
ward-curved blade types.
AXIAL-FLOW FANS
Axial-flow fans include all those wherein the air
flows through the impellers substantially parallel
•with the shaft upon -which the impeller is mounted.
Axial -flow fans depend upon the action of the re-
volving airfoil-type blades to pull the air in by the
leading edge and discharge it from the trailing
edge in a helical pattern of flow. Stationary vanes
may be installed on the suction side or the dis-
charge side of the rotor, or both. These vanes
convert the centrifugal force and the helical-flow
pattern to static pressure.
Axial fans may be divided into three main classi-
fications :
1. Propeller type. Propeller fans have large,
disc-like blades or narrow, airfoil-type
blades. The number of blades is 2 to 16.
The propeller fan blades may be mounted on
a large or small hub, depending upon the use
of the fan. The propeller fan is distinguished
from the tube-axial and vane-axial fans in that
it is equipped -with a mounting ring only.
2. Tube-axial type. The tube-axial fan is simi-
lar to the propeller fan except it is mounted
in a tube or cylinder. It is more efficient than
the propeller fan and, depending upon the de-
sign of the rotor and hub, may develop medi-
um pressures. A two-stage, tube-axial fan,
with one rotor revolving clockwise and the
second, counter-clockwise, -will recover a
large portion of the centrifugal force as static
pressure, -which would other-wise be lost in
turbulence. Two-stage, tube-axial fans ap-
proach vane-axial fans in efficiency.
3. Vane-axial type. The vane-axial fan is simi-
lar in design to a tube-axial fan except that
air-straightening vanes are installed on the
-------
Fan Design
suction side or discharge side of the rotor.
Vane-axial fans are readily adaptable to mul-
tistaging, and fans have been designed that
•will operate at a pressure of 16 inches water
column at high volume and efficiency.
FAN CHARACTERISTICS
The performance of a fan is characterized by the
volume of gas flow, the pressure at which this flow
is produced, the speed of rotation, the power re-
quired, and the efficiency. The relationships of
these quantities aremeasuredby the fan manufac-
turer with testing methods sponsored by the Na-
tional Association of Fan Manufacturers or the
American Society of Mechanical Engineers. Brief-
ly, the method consists of mounting a duct on the
fan outlet, operating the fan with various sized
orifices in the duct, and measuring the volume,
pressure, velocity, and power input. About 10
tests are run, with the duct opening varied from
wide open to completely closed. The test results
are then plotted against volume on the abscissa to
provide the characteristic curves of the fan, such
as those shown in Figure 25.
Figure 25. Typical fan characteristic curves
(Air Moving and Conditioning Assn., Inc.,1963).
From, the volume and pressure, the air horsepower
is computed, either the real power based on total
pressure or the fictitious static power based on
fan static pressure. The efficiency based on total
pressure is called mechanical efficiency.
INFLUENCE OF BLADE SHAPE
The size, shape, and number of blades in a cen-
trifugal fan have a considerable influence on the
operating characteristics of the fan. The general
effects are indicated by the curves in Figure 26,
FORWARD-CURVED-BLADE
VOLUME
TOTAL PP.
BACKWARD-CURVED-BLADE
VOLUME
STRAIGHT-BLADE FAN
VOLUME
Figure 26. Centrifugal fan typical
characteristic curves (Hicks, 1951).
These curves are shown for comparison purposes
only; they are not applicable for fan selection but
do indicate variations in the operating character-
istics of a specific type of fan.
1. Forward-Curved-Blade Fans. This type of
fan is normally referred to as a volume fan.
In this fan, the static pressure rises sharply
from free delivery to a point at approximate
maximum efficiency, then drops to point a.
shown in Figure 26, before rising to static
pressure at no delivery. Horsepower input
rises rapidlyfromno delivery to free delivery.
Sound level is least at maximum efficiency and
greatest at free delivery. Forward-curved-
blade fans are designed to handle large vol-
umes of air at low pressures. They rotate at
relatively low speeds, which results in quiet
operation. Initial cost of such a fan is low.
-------
62
DESIGN OF LOCAL EXHAUST SYSTEMS
Resistance of a system to be served by this
type of fan must be constant and must be deter-
mined accurately in advance because of sharp-
ly rising power demand. This type of fan should
notbe usedfor gases containing dusts or fumes
because deposits -will accumulate on the short
curved blades resulting in an unbalanced wheel
and excessive maintenance. The pressure
produced by a forward-curved-blade fan is
not normally sufficient to meet the pressure
requirements for the majority of air pollution
control devices. They are, however, used
extensively in heating, ventilating, and air
conditioning work. Also, they are commonly
used for exhausting air from one enclosed
space to another without the use of ductwork.
2. Backward-Curved-Blade Fans. The static
pressure of this type fan rises sharply from
free delivery almost to the point of no delivery.
Maximum efficiencies occur at maximum
horsepower input. The horsepower require-
ment is self-limiting; it rises to a maximum
as the capacity increases and then decreases
with additional capacity. Thus, when the re-
sistance of a complex exhaust system is fre-
quently changed because of production demands,
the self-limiting power requirements prevent
overloading the motor.
This type of fan develops higher pressure than
the forward-curved-blade type. Sound level
is least at maximum efficiency and increases
slightly at free delivery. The physical sizes
of back-ward-curved-blade fans for given duties
are large, but for most industrial work this
may be unimportant. The operating efficiency
is high, but initial cost is also high. Blade
shape is conducive to buildup of material and
should not be used on gases containing dusts
or fumes.
The backward-curved-multiblade fan is used
extensively in heating, ventilating, and air
conditioning work and for continuous service
where a large volume of air is to be handled.
Itis commonly found on forced-draft combus-
tion processes. It may be used on some air
pollution control devices, but must De installed
on the clean air discharge as an induced system.
This type fan is utilized for exhaust systems
handling gas streams that are contaminated
with dusts and fumes. Various blades and
scroll designs have been developedfor specific
dust-handling and pneumatic-conveying prob-
lems. This fan is too large for some duties,
but for most industrial work this may be un-
important. Initial cost of this type fan is less
than that of the back-ward-curved-blade type,
but efficiency is also less. Fan blades may
be made of an abrasive resistant alloy or
covered -with rubber to prevent high main-
tenance in systems handling abrasive or cor-
rosive materials.
A number of modified designs of straight-blade
fans have been specifically developed for hand-
ling contaminated air or gas streams.
Axial Fans. For this type fan, the horsepower
curve may be essentially flat and self-limiting,
dependingupon the design of the blades, or it
may fall from a maximum at no delivery as
capacity increases. The type of vanes in the
vane-axialfans measurably affects the horse-
power curve and efficiency. Maximum effi-
ciencies occur at a higher percent delivery
than with the centrifugal-type fan.
Space requirements for a specific fan duty are
exceptionally low. Available fans can be in-
stalled directly in circular ducts (vane-axial
or tube-axial type). Initial cost of the fan is
low.
The axial-type fan is best adapted for hand-
ling large volumes of air against low resis-
tance. The propeller type, which is equipped
only -with a mounting ring, is commonly used
for ventilation and is mounted directly in a
wall. Although the vane-axial and tube-axial
fans can deliver large volumes of air at rela-
tively high resistances , they are best suited for
handling clean air only. Any solid material
in the air being handled causes rapid erosion
of impellers, guide vanes, hubs, and the inner
wall of the cylindrical fan housing. This
results from the high tip speed of the fan and
the high air velocity through the fan housing.
3. Straight-Blade Fans. The static pressure of
this type fan rises sharply from free delivery
to a maximum point near no delivery, where
it falls off. Maximum static efficiency occurs
near maximum pressure. Mechanical effi-
ciency rises rapidly from no delivery to a
maximum near maximum pressure, then
drops slowly as the fan capacity approaches
free delivery.
Geometrically Similar Fans
Fan manufacturers customarily produce a series
of fans characterized by constant ratios of linear
dimensions and constant angles between various
fan parts. These fans are said to be geometrically
similar or of a homologous series. The drawings
of all the fans in the series are identical in all
views except for scale.
-------
Fan Design
63
It is usual for a manufacturer to produce homol-
ogous series of fans with diameters increasing by
a factor of about 1.10. Each is designated by the
impeller diameter or by an arbitrary symbol, often
a number proportional to the diameter.
Multirating Tables
The performance of each fan in a homologous
series is usually given in a series of tables called
multiratingtables. Values of static pressure are
usually arranged as headings of columns, which
contain the fan speed and horsepower required to
produce various volume flows. The point of maxi-
mum efficiency at each static pressure is usually
indicated.
FAN LAWS
Certain relationships have been established among
the variables affecting the performance of fans of
a homologous series or a single fan operating at
varying speed in a constant system. The quantity,
V^,and the power, p, are controlled by four inde-
pendent variables: (1) Fan size, wheel diameter,
D, (2) fan speed, N, (3) gas density, p, and
(4) system resistance, hr. Since all dimensions
of homologous fans are proportional, any dimen-
sion could be used to designate the size. The wheel
diameter is, however, nearly always used.
In order to develop these relationships, the effects
of system resistance must be fixed by limiting the
comparisons to the same points of rating. For
two fans of different size, the same point of rating
is obtained when the respective volumes are the
same percentage of wide open volume, and the
static pressure is the same percentage of shut-off
static pressure. For the same fan, the same point
of rating is obtained when the system is held con-
stant and the fan speed is varied.
equating exponents for like terms,
m: 0 = c
L: 3 = a - 3c
t : -1 = -b
and solving the equations simultaneously
a = 3; b = 1; c = 0
hence:
V = kD3N
(28)
Repeating for the system resistance developed
and noting that hr is fundamentally force per
unit area = mass X acceleration per area.,
m 1 = c
L -1 = a - 3c
t -2 = -b
a = 2; b = 2; c = 1
h = kD2N2p
(29)
And repeating again for the power required:
P =
For homologous fans (or the same fan) operating
at the same point of rating, the quantity (V^) and
the power (P) will depend upon the fan size (D),
fan speed (N), and gas density (p). The flow
through a fan is always in the turbulent region,
and the effect of viscosity is ignored. The form
of dependence can be derived from dimensional
analysis by the equation
V = k
(27)
By substituting fundamental dimensional units,
m 1 = c
L 2 = a - 3c
t -3 = -b
a = 5; D = 3; c = 1
P = kD5N3p
(30)
Equations 28, 29, and 30 for Vt, hr, and P define
the relationships among all the variables, within
the limitations originally stated. The equations
can be simplified, combined, or modified to yield
-------
64
DESIGN OF LOCAL EXHAUST SYSTEMS
a large number of relationships. The following
relationships derived from them are usually re-
ferred to as the Fan Laws.
1. Change in Fan Speed.
Fan size, gas density, and system constant.
a. Y£ varies as fan speed.
b. h varies as fan speed squared.
c. P varies as fan speed cubed.
2. Change in Fan Size.
Fan speed and gas density constant,
a. V varies as cube of wheel diameter.
b. h varies as square of wheel diameter.
c. P varies as fifth power of wheel diameter.
d. Tip speed varies as wheel diameter.
3. Change in Fan Size.
Tip speed and gas density constant.
a. V varies as square of -wheel diameter.
b. h remains constant.
r
c. P varies as square of wheel diameter.
d. rpm varies inversely as wheel diameter.
4. Change in Gas Density.
System, fan speed, and fan size constant.
a. V is constant.
b. h varies as density.
r
c, P varies as density.
5. Change in Gas Density.
Constant pressure and system, fixed fan size,
and variable fan speed.
a. V varies inversely as square root of density.
b. Fan speed varies inversely as square root of
density.
c. P varies inversely as square root of density.
6. Change in Gas Density.
Constant weight of gas, constant system, fixed
fan size, and variable fan speed.
a. V varies inversely as gas density.
b. h varies inversely as gas density.
c. Fan speed varies inversely as gas density.
d. P varies inversely as square gas density.
The fan laws enable a manufacturer to calculate
the operating characteristics for all the fans in a
homologous series from test data obtained from a
single fan in the series. The laws also enable
users of fans to make many needed computations.
A few of the more important cases are illustrated
as follows.
Example 11
A fan operating at 830 rpm delivers 8, 000 cfm at
6 inches static pressure and requires 11.5 horse-
power. It is desired to increase the output to
10, 000 cfm in the same system. What should be
the increased speed and -what will be the horse-
power required and the new static pressure?
Solution:
Use fan law la, b , c :
N1 = 830
fio,oool
|_8, 000 J
1, 037 rpm
P1 = 11.5
83°
], 037
830
= 9. 35 in. WC
= 22.4hp
Example 12
A fan is exhausting 12, 000 cfm of air at 600 °F.
(density = 0. 0375 pound per cubic foot at 4 inches
static pressure from a drier). Speed is 530 rpm,
and 13 horsepower is required. What will be the
required horsepower if air at 70°F (density 0. 075
pound per cubic foot) is pulled through the system?
Solution:
Use fan law 4 c :
P' = 13 h?"
. 375 J
= 26 hp
-------
Fan Design
65
If a 15-horsepower motor were used in this in-
stallation, it would be necessary to use a damper
•when starting up cold to prevent overloading the
fan motor.
Example 13
A 30-inch-diameter fan operating at 1, 050 rpm
delivers 4, 600 cfm at 5 inches static pressure.
What size fan of the same series •would deliver
11, 000 cfm at the same static pressure?
Solution:
Use fan law 2 a. .
D
ii.ooo\1/3
4, bOO )
(30) = 40. 0 in.
Selecting a Fan From Multirating Tables
A typical multirating table is given in Table 20.
The data in this table are for a paddle wheel-type
industrial exhauster. In using multirating tables,
use linear interpolation to find values between
those given in the table. For instance, from
Table 20 it is desired to find the fan speed that
will deliver 6, 300 cfm at 6-1/2 inches static
pressure. The nearest capacities are 6, 040
and 6, 550.
At 6 in. h the speed is
s
<^5-i.o88) = 1,092
At 7 in. h the speed is
s
1,160 +
6,300-6, 040
6, 550-6, 040
(1, 171 - 1, 160) = 1, 167 rpm.
The required speed at b-1/2 inches static pres-
sure and b, 300 cfm is halfway bet-ween 1, 092 and
1, 167 or 1, 129 rpm.
CONSTRUCTION PROPERTIES
Special materials of construction must be used ior
tans handling corrosive gases. Certain alloys that
have been used have proved very satisfactory.
Bronze alloys are used for handling sulfuric acid
tumes and other sulfates, halogen acids, various
organic gases, and mercury compounds. These
alloys are particularly applicable to low-tempera-
ture installations. Stainless steel is the most
commonly used metal for corrosion-resistant im-
pellers and fan housings. It has proved satisfac-
tory for exhausting the fumes of many acids. Pro-
tective coatings on standard fan housing and
impellers such as bisonite, cadmium, plating, hot
galvanizing, and rubber covering have proved
satisfactory. Cadmium plating and hot galvanizing
are often used in conjunction with a zinc chromate
primer, with which they form a chemical bond.
The zinc chromate primer may then be covered
with various types of paints. This combination
has proved favorable in atmospheres near the
ocean.
The increasing use of rubber for coating fan im-
pellers and housings deserves special mention.
Rubber is one of the least porous materials and,
when vulcanized to the metal, surrounds and pro-
tects the metal from corrosive gases or fumes.
Depending upon the particular application, soft,
medium, or firm rubber is bonded to the metal.
A good bond will yield an adhesive strength of
700pounds per square inch. When pure, live rub-
ber is so bonded, it is capable of withstanding the
high stresses set up in the fan and is sufficiently
flexible to resist cracking. Rubber-covered fans
have proved exceptionally durable and are found
throughout the chemical industry.
Heat Resistance
Standard construction fans with ball bearings can
withstand temperatures up to 250°F. Water-
cooled bearings, shaft coolers, and heat gaps per-
mit operation up to 800°F. A shaft cooler is a
separate, small, centrifugal fan that is mounted
between the fan housing and the inner bearing and
that circulates cool air over the bearing and shaft.
A heat gap, which is merely a space of 1-1/2 to
2 inches between the bearing pedestal and fan housing,
reduces heat transfer to the bearings by conduction.
Certain types of stainless steel will withstand the
high temperatures encountered in the induced-
draft fan from furnaces or combustion processes.
Stainless steel tans have been known to withstand
temperatures as high as 1, 100°F without excess
warping.
Expii sive-Proof Fans and Motors
Wheii an exhaust system is handling an explosive
mixture of air and gas or powder, a material to
be used in the construction of the fan must gen-
erally be specified to be one that will not produce
a spark if accidentally struck by another metal.
Normally, the fan impeller and housing are con-
structed of bronze or aluminum alloys, •whichpre-
cludes spark formation. Aluminum is frequently
used on some of the narrower or smaller fans,
especially those overhung on the motor shaft.
Aluminum reduces the weight and vibration of the
motor shaft and protects the motor bearing from
excessive wear.
-------
66
DESIGN OF LOCAL EXHAUST SYSTEMS
ft
w
.5
Ox
A
w
a
•H
CO
ft
w
fl"
•H
fx-
ft
OT
•H
xD
ft
w
.a
in
ft
w
fi
• H
•*
ft
w
_g
CO
A
w
.3
ro
ft
w
.a
ft
rD
ft
h
ft
,n
g
ft
fn
£
rd
a
a
h
ft
,d
,0
s
Pi
Vl
ft
,0
s
ft
VH
ft
£
ft
h
ft
£
ft
H
ft
£
a
&
H
a
,n
a
^
Velocity
pressure,
in. WC
Outlet
velocity,
fpm
Volume,
cfm
0
xO
•xf
co
•*
0
xO 0 fx-
xO ~H in
ro co co
tx- oo r-
co •* -xf
00 OO OO
o o oo
o co t—
00 00 00
OO in xO
00 CO •*
r- t^ r—
r- m r-
oo m co
~H r-l 1— t
m o xo
ON —< 00
in xo o
CO UTl 1^
xo co o
0 O — i
[x- O ro
ro m co
•xf \f •*<
co O oo
xQ ox oo
O O — i
O O O
O O O
0 O O
0 CM ^h
o o o
ro ox) co
in I-H in
oo ro ro
1-
xO
rH
1 — 1
O
f-
(M
m
0
i — i
00
o
(M
OO xD
r-< O
OO 00
in •*
IN] ro
Ox oo in
oo Ox tx-
XO XO t^
O O 00
co •xf m
0 O O
-i ro o
CM oo m
in in xo
0 -xf xO
in xo fx-
OX Ox Ox
ro ro o
•-< xD CO
•* -^ m
oo xo o
m r- o^
oo oo oo
t~- co oo
— 1 xO -H
ro co "xf
ox -xf r-
m r*- ox
r- t-- r-
oo xD t-
i—i in ox
ro rxl ro
ro xO oo
•xf xO 00
xO xO xO
ro *-* o
ro xD O
*-< '-I 00
ro 00 00
•-t ro t~-
in in in
O 00 0
xo o m
I-H ro oo
o o o
0 O 0
o o o
xO 00 O
o o o
ro ro •*
o m o
•xF -xf m
ro oo o
00 OX -H
oo ro LD
Ox oo CO
[x- 00 Ox
00 00 00
oo m o
^H oo in
r^ OO CO
o o m
I-H ro •*
00 00 00
ro oo oo
o oo ox
Ox O —I
m o r-,
•V xo r~
o o o
xo in in
OO Ox O
oo oo m
xo oo ox
o o o
co t- oo
CO —i 0
t- OO Ox
Ox tx- OO
OX c-H CO
Ox O O
CO CO CO
Ox tx- oo
m xo tx.
O xO •*
I-H oo m
0s Ox Ox
xo co m
xO ro O
•xf m XD
xD O xO
— H "* XO
00 OO OO
ro ox og
^ 0s xO
CO CO "^
OO xD 00
— 1 ^t1 xO
r- r~ r~
xO Ox [^
ro tx- ox)
oo oo ro
CO f- O
o co r-
xD xO xO
00 O 00
O xD 00
ro ro -xf
0 0 O
O 0 O
000
CXl Tf xO
00 00 00
O O 0
^ ^ m
m o m
m xo xo
00
•*
xO
O
1 — 1
ro
O
t^
•xf
r^
m
00
xO
o
ro
00
oo
0
xO
r— t
m
ro
T— 1
t-
Ox
Ox
0
m
o
oo
t~-
oo
ro
xO
OX
00
r~
xO
00
Ox
oo
00
CO
UO
m
Ox
t^
F-H
oo
CO
00
0
r-
Ox
oo
•*
o
o
0
oo
00
o
xO
o
(X-
o in
00 -4
[x- Ox
oo o
ro •*
ro ro
OO xO
Ox CO
m r-
r- oo
h- Ox
00 00
oo o
oo in
•xf m
o oo
i-l 00
oo ro
m oo
r- o
og ^
oo oo
^ xO
-—1 ^H
o o
0 —i
^H 00
oo r-
xD 0s
0 0
00 O
CO •*
Ox O
1 — 1
ro O
Ox 00
Ox O
O rH
r- r-
r- oo
O CO
rxl ^f
Ox Ox
in xD
0 Ox
xO xD
CO xD
CO xD
00 OO
cxl
•xf
•*
xO
•*
r-
o oo
xO CO
m xD
o o
o o
o o
0 OJ
ro rO
O O
xO xQ
m o
r~ oo
o o o
Ox xO ^
0 00 -xf
00 00 00
o o m
xO 00 O
ro co -xt1
o m in
o r^ co
Ox O 00
— 1 OJ 00
o m m
^H co in
co ro ro
CO OO xO
Ox ^ ^
xO OO Ox
00 O 00
•xf Ix- Ox
00 00 00
co h- oo
in xo co
00 CO O
00 <-H •*!
I-H -H
^ 0 00
xO ^H CO
Ox O O
i — i •— i
CO
Ox
c— "
o
0
Ox
—I 00 O
00 O O
r- oo ox
o o o
o o o
o o o
•* xD CO
CO CO CO
o o o
XD r- r-
m o m
oo ox ox
o
•xf
xO
00
o
CO
•xf
m
• — i
CO
00
O
oo
CO
o
r-
i — i
00
o
oo
CO
t-
o>
1 — 1
0
t-
OJ
m
-------
Vapor Compressors
67
Explosive-proof motors and fan wheels are re-
quired by law for installation in places where
an explosive mixture may be encountered. Ex-
haust systems such as those used in paint spray
booths usually consist of an aluminum or bronze
tube-axial fan and an explosive-proof motor that
drives the fan wheel by indirect drive.
Fan Drives
All types of fans may be obtained with either direct
drive or belt drive. Directly driven exhausters
offer the advantage of a more compact assembly
and ensure constant fan speed. They are not trou-
bled by the belt slippage that occurs -when belt-
driven fan drives are not properly maintained.
Fan speeds are, however, limited to the available
motor speeds, -which results in inflexibility ex-
cept in direct-current application. A quick change
in fan speed, which is possible "with belt-driven
fans, is a definite advantage in many applications.
VAPOR COMPRESSORS
Compressors are widely used in industry to in-
crease the pressure of gases or vapors for a
variety of reasons. They are used:
1. To provide the desired pressure for chemical
and physical reactions;
2. to control boiling points of fluids, as in gas
separation, refrigeration, and evaporation;
3. to evacuate enclosed volumes;
4. to transport gases or vapors;
5. to store compressible fluids as gases or
liquids under'pressure and assist in recov-
ering them from storage or tank cars;
6. to convert mechanical energy to fluid energy for
operating instruments, air agitation, fluidiza-
tion, solid transport, blowcases, air tools, and
motors.
Compressors normally take suction near atmo-
spheric pressure and deliver fluids at pressures
ranging upward to 40, 000 psig in commercial ap-
plications and even higher in experimental uses.
The capacity of commerically available compres-
sors ranges fromlow volumes up to 3 million cfm.
TYPES OF COMPRESSORS
Vapors or gases canbe compressed by either posi-
tive displacement or dynamic action. The positive-
displacement compressors produce pressure by
physically reducing the gas volume. The dynamic
compressors increase pressure by accelerating
the gas and converting the velocity into pressure
in a receiving chamber. Positive-displacement
compressors are of reciprocating- or rotary-dis-
placement types. The dynamic compressors are
centrifugal- or axial-flow machines. Figure 27
shows general compressor applications, and
Table 21 gives general limits of compressors.
I03 10"
COMPRESSOR INLET CAPACITY, cfm
Figure 27. General areas of compressor applications (Des Jardins, 1956).
-------
68
DESIGN OF LOCAL EXHAUST SYSTEMS
Table 21. GENERAL LIMITS OF COMPRESSORS
(Des Jardins, 1956)
Compressor
type
Reciprocating
Centrifugal
Rotary displace-
ment
Axial flow
Approximate maximum limits
of commercially
available compressors
Discharge
pressure,
psia
35, 000
4, 200
125
90
Compression
ratio per
stage
10
4
4
1.2
Inlet
capacity,
cfm
13,000
18, 000
7, 000
3, 000, 000
Positive-Displacement Compressors
A reciprocating compressor raises the pressure
of the air or gas by the forced reduction of its
volume through the movement of a piston within
the confines of a cylinder. These compressors
are commercially designed for volumes up to
15, 000 cfm and pressure sup to 40, OOOpsig. They
are by far the most common type in use both for
process systems and air pollution control sys-
tems (Cumiskey, 1956).
Rotary sliding-vane compressors have longitudinal
vanes that slide radially in a rotor mounted eccen-
trically in a cylinder. The rotor is supported at
each end by antifriction bearings mounted in the,
heads, which, in turn, are bolted and doweled to
the cylinder. Figure 28 shows a cross-sectional
view of a sliding-vane compressor.
In a sliding-vane unit, pressure is increased by
reducing the size of the compression cell while it
rotates from the suction to the discharge ports.
SLIDING VANE
ROTOR
CYLINDER
As the unit rotates, each compression cell reaches
maximum size when itpasses the inlet ports. Fur-
ther rotation of this cell reduces its size, and com-
pression is completed upon reaching the discharge
ports (Bruce and Schubert, 1956). In general,
single-stage, rotary, sliding-vane compressors
are suitable for pressures up to 50 psig. Multi-
staged machines are designed for pressures up to
250 psig, and booster units are available for pres-
sure up to 400 psig. These machines can deliver
up to 6, 000 cfm.
Rotary-lobe compressors have two mating, lobed
impellers that revolve within a cylinder. Timing
gears, mounted outside the cylinder, prevent the
impellers from contacting each other. The lobes
are mounted on shafts supported by antifriction
bearings. Figure 29 shows a cross-sectional view
of a rotary-lobe compressor. Flow through the
rotary-lobe compressor is accomplished by the
lobes' pushing the air or gas from the suction to
the discharge. Essentially, no compression takes
place -within the unit; rather, compression takes
place against system back pressure (Bruce and
Schubert, 1956). Rotary-lobe compressors are
available in sizes up to 50,000 cfm and pressures
up to 30 psig. Single-stage machines are usually
good for pressures up to 15 psig, and vacuums to
22 inches of mercury.
ROTARY LOBE
Figure 28. SIiding-vane
compressor (Bruce and
Schubert, 1956).
Figure 29. Rotary-lobe
compressor, (Bruce and
Schubert, 1956).
Rotary liquid-piston compressors use water or
other liquids, usually in a single rotating element
to displace the air or gas being handled. A ro-
tating element is mounted on a shaft and supportec
at each end by antifriction bearings. Figure 30
shows a cross-sectional view of a rotary liquid-
piston compressor. In the rotary liquid-piston
compressor, flow of compressed air or gas is
discharged in a uniform, nonpulsating stream.
Compression is obtained in this machine by ro-
tating a round, multiblade rotor freely in an
-------
Vapor Compressors
69
LIQUI D RING
INLET PORT
DISCHARGE PORT
Dl SCHARGE PORT
INLET PORT
Figure 30. Rotary Iiquid-giston compressor
(Bruce and Schubert, 1956).
tical casing partially filled with liquid. The ro-
tating force of the multiblade rotor causes the
liquid to follow the inside contour of the elliptical
casing. As the liquid recedes from the rotor blades
at the inlet port, the space between buckets fills
with the air or gas. As the liquid reaches the
narrow point of the elliptical casing, the air or
gas is compressed and forced out through the dis-
charge ports (Bruce and Schubert, 1956). Rotary
liquid-piston compressors are available in sizes
up to approximately 5, 000 cfm. Standard single-
stage units are usedfor pressures to 35 psig, and
special single-stage units, for pressures up to 75
psig. Units are staged above 75 psig.
Dynamic Compressors
Centrifugal compressors are similar to centrifugal
pumps and fans. An impeller rotates in a case,
imparting a high velocity and a centrifugal motion
to the gas being compressed. The impeller is
mounted on a shaft supported by bearings in each
end of the case. In multiple-stage compressors,
several impellers are mounted on a single shaft.
Passages conduct the gas from one stage to the
next. Guide vanes iri the passages direct the gas
flow from one impeller to the next at the proper
angle for efficient operation. Figure 31 shows a
cross-sectional view of a typical four-stage
compressor.
Since the flow of gas to the centrifugal compressor
is continuous, the fundamental concepts of fluid
flow apply. The gas enters at the eye of the im-
peller, passes through the impeller, changing in
velocity and direction, and exits into the diffuser
or volute, where the kinetic energy is converted to
pressure (Leonard, 1956).
The centrifugal compressor generally handles
a large volume of gas at relatively low pres-
sures, but some commercially used centrifugal
compressors have discharge pressures of up
to 4, 200 psig. These compressors are, of
course, multistaged. Generally, the single-
stage centrifugal compressor produces pres-
sures up to 35 psig.
The axial-flow compressor, shown in Figure 32
is another type of dynamic compressor. It is dis-
tinguished by the multiplicity of its rotor and stator
blades. These are either forged, machined, or
precision cast into airfoil shapes. The compres-
sor casing is made of cast iron, or fabricated out
of steel, depending upon inlet volume, pressure
ratio, and temperature conditions. Stator blades
are attached to the casing to direct the flow of gas
through the case.
The rotor is a drum with blades mounted around
its periphery. The drum is mounted on the shaft
supported by bearings in each end of the case. As
the rotor turns, the blades force air through the
compressor in an action similar to that of the pro-
peller fan. The stator blades control the direction
of the air as it leaves the rotor blades. Pressure
is increased owing to the kinetic energy given to
the gas, and the action of the gas on the stator
blades. Axial-flow compressors are high-speed
high-volume machines. The pressures attained
are relatively low, -with a maximum commercially
used discharge pressure of 90 psig. These com-
pressors are rarely used for inlet capacities be-
low 5, 000 cfm (Claude, 1956).
Reciprocating Compressors
Reciprocating compressors are positive-displace-
ment machines used to increase the pressure of a
definite volume of gas by volume reduction (Case,
1956). Most reciprocating compressors used in
heavy industrial production and continuous chem-
ical processing are stationary, water-cooled,
double-acting units (see Figure 33). The basic
running-gear mechanism is of the crank-and-fly-
wheel type enclosed in a cast-iron frame. The
crosshead construction permits complete separa-
tion of the compression cylinder from the crank-
case, an ideal feature for handling combustible,
toxic, or corrosive gases. Generally, the cylinder
is double acting, that is, compression occurs al-
ternately in the head and crank end of the cylinder.
The cylinder and its heads are usually water cooled
to reduce thermal stresses and dissipate part of
the heat developed during compression. Com-
pression rings on the pistons seal one end of the
cylinder from the other. The piston rod is sealed
in the cylinder by highly effective packing, and
any slight leakage may be collected in a vent gland
for return to suction or for venting to the atmosphere.
-------
70
DESIGN OF LOCAL EXHAUST SYSTEMS
Figure 31. Cross-section of a typical four-stage centrifugal compressor (Clark Bros. Co., Olean, N.Y.,
from Leonard, 1956).
Gas being corrroressed enters and leaves the
cylinder through the voluntary valves, which are
actuated entirely by the difference in pressure be-
tween the interior of the cylinder and the outside
system. Upon entering the cylinder, the gas may
be compressed from the initial to the desired final
pressure in one continuous step, that is, single-
stage compression. Alternatively, multistage com-
pression divides the compression into a series of
steps or stages, each occurring in an individual
cylinder. Here the gas is usually cooled between
the various stages of compression.
The compression process is fundamentally isen-
tropic (perfectly reversible adiabatic), with cer-
tain actual modifications or losses that may be
considered as efficiencies related to the isentropic
base. Thermal dynamic losses within the cylinder
including fluid friction losses through the valves,
heating of the gas on admission to the cylinder,
and ir reversibility of the process, maybe grouped
under the single term compression efficiency.
Mechanical friction losses encountered in the
piston rings, rod packings, and frame bearings
are grouped under the term mechanical efficiency.
Thus, the overall efficiency of the compressor is
the product of compression and mechanical
efficiency.
For given service, the actual brake horsepower
requirement of the compressor is normally about
18 to 33 percent greater than the calculated ideal
-------
Vapor Compressors
71
Figure 32. Axial-flow compressor (AlIis-Chalmers Manufacturing Company,
Claude, 1956).
Iwaukee, Wisconsin, from
Figure 33. Four-cylinder, horizontal, balanced, opposed, synchronous-motor compressor (Worthmaton
Corporation, Harrison, N.J., from Case, 1956).
-------
72
DESIGN OF LOCAL EXHAUST SYSTEMS
isentropic horsepower. Or, stated another way,
the overall efficiency of most compressors ie in
the range of 75 to 85 percent.
USE IN AIR POLLUTION CONTROL
Compressors are used to transport vapors or gases
from their source and deliver them to a control
device or system under pressure. In some cases,
the vapors or gases can be pressurized directly to a
holding vessel and then a compressor is used to
send the vapors to control equipment.
The vapors created from the refining, storing,
and bulk loading of volatile petroleum products are
being controlled by the use of compressors. The
compressors deliver the vapors under pressures
ranging from 5 to 200 psig to plant fuel systems,
process streams, or absorption systems.
Centrifugal, reciprocating, and rotary-lobe com-
pressors are being used for controlling air con-
taminants. Single-stage reciprocating machines
are the most common. Two-stage compressors
developing pressures up to 200 psig are in use.
CHECKING OF EXHAUST SYSTEM
Air flow measurements and test data are neces-
sary to determine whether an exhaust system is
functioning properly and in compliance with de-
sign specifications. Correct testing procedures
must be established to obtain measurements for
determining whether an exhaust system has suf-
ficient capacity for additional hoods, and also to
obtain operational data from existing installations
for designing future exhaust systems.
Velocity Meters
Velocity meters are more commonly used in the
field for determining air velocities. The most
accurate and widely accepted in engineering prac-
tices are the pitot tube and the swinging-vane ve-
locity meter.
PITOT TUBES
For determining air velocity, the standard pitot
tube, named for the man who discovered the prin-
ciple, is considered reliable and is generally ac-
cepted in engineering practice. It is the most
widely used field method for determining air
velocity.
A standard pitot tube (Figure 34) consists of two
concentric tubes: the inner tube measures the
impact pressure, which is the sum of the static
and kinetic pressures, while the outer tube mea-
sures only the static pressure. When the two tubes
are connected across a U-tube manometer or other
suitable pressure-measuring device, the static
pressure is nullified automatically and only the
velocity pressure (kinetic pressure) is registered.
The velocity is correlated to the velocity pressure
by the equation
1096.5
(si;
where
v = velocity, fpm
h = velocity pressure (manometer reading),
V in. WC
p = density of air, Ib/ft .
THEORY OF FIELD TESTING
For most purposes the most important factor is the
accurate measurement of air quantity. Most field
meters measure velocity rather than quantity.
This necessitates equating velocity to quantity. By
using equation 8 and a velocity meter, the quantity
of air flowing through an exhaust system can be
accurately measured.
Quantity Meters
Some examples of quantity meters are thin-plate
orifices, sharp-edged orifices , and venturimeter s .
These meters are used extensively in laboratory
studies, but infrequently in industrial exhaust
systems.
Clearly, below 1, 266 fpm, the velocity pressure
becomes extremely low and is, therefore, diffi-
cult to read accurately on a manometer. With a
IMPACT PRESSURE CONNECTION
TUBING ADAPTER
STANDARD BELL REDUCERS
STAINLESS STEEL TUBING
• STATIC PRESSURE CONNECTION
V STATIC PRESSURE HOLES STAINLESS STEEL PIPE NIPPLES
OUTER PIPE ONLY
IMPACT PRESSURE OPENING
Figure 34. Standard pitot tube (Western
Precipitation, Division of Joy Manufacturing
Co., Los Angeles, Calif., from ASHRAE Guide
and Data Book, 1963).
-------
Checking of Exhaust System
73
U-tube manometer, the accuracy is low for veloc-
ities below Z, 500 fpm. With a carefully made,
accurately leveled, inclined manometer, veloci-
ties as low as 600 fpm can be determined satis-
factorily, but field conditions ordinarily make
this procedure difficult (ASHRAE Guide and Data
Book, 1963).
Pitot Tube Traversing Procedure
Since the velocity in a duct is seldom uniform
across any cross section and since each pitot tube
reading determines the velocity at only one local-
ized point, a traverse of the duct is necessary in
order to compute the average velocity and thus
determine air flow accurately. Suggested pitot
tube locations for traversing round and rectangu-
lar ducts are shown in Figure 35.
The velocity in a duct varies greatly. It is gen-
erally lowest near the edges or corners and
greatest in the central portion. Because of this
fluctuation, a large number of readings must be
taken to determine the true average velocity. In
round ducts, not less than eight readings should
be taken along two diameters at centers of equal
annular areas. Additional readings are necessary
when ducts are larger than 1 foot in diameter. In
rectangular ducts, the readings should be taken in
the center of equal areas over the cross section of
the duct. The number of spaces should be taken
as depicted in Figure 35. In determining the av-
erage velocity in the duct, the velocity pressure
readings are converted to velocities; the veloci-
ties, not the velocity pressures, are averaged to
compute the average duct velocities.
Disturbed flow will give erroneous results; there-
fore, whenever possible, the pitot tube traverse
should be made at least 7. 5 duct diameters down-
stream, from any major air stream disturbances
such as a branch entry, fitting, or supply open-
ing (ASHRAE Guide and Data Book, 1963).
Altitude and Temperature Corrections for
Pitot Tubes
If the temperature of the air stream is more than
30° above or below the standard temperature of
70°F, or if the altitude is more than 1, 000 feet,
or if both conditions hold true, make a correction
for density change as follows:
Corrected velocity pressure = measured h x — |
p1 = relative density of air, at the mea-
sured condition, Ib/ft .
Cross section of a circular stack divided
into three concentric, equal areas, showing
location of traverse points. -The location
and number of these points for a stack of
given diameter can be determined from Tables
22 and 23.
Cross section of a rectangular stack divided into 12
equal areas, with traverse points located at the cen-
ter of each area. The minimum number of test points
i s shown i n Table 24.
Figure 35. Pitot tube traverse for round and
rectangular ducts.
where
(32)
h = velocity pressure, as determined by
pitot tube, in. WC
SWINGING-VANE VELOCITY METER
The factors that make the swinging-vane velocity
meter an extensivelyused field instrument are its
portability, instantaneous reading features, and
-------
74
DESIGN OF LOCAL EXHAUST SYSTEMS
Table 22. SUGGESTED NUMBER
OF EQUAL AREAS
FOR VELOCITY MEASUREMENT
IN CIRCULAR STACKS
Stack diameter,
ft
1 or less
1 to 2
2 to 4
4 to 6
over 6
Number of
equal areas
2
3
4
5
6 or more
Table 23. PERCENT OF CIRCULAR
STACK DIAMETER FROM INSIDE
WALL TO TRAVERSE POINT
Point
number
1
2
3
4
5
6
7
8
9
10
11
12
Number of areas selected
2
6.7
25. 0
75.0
93.3
-_
--
--
--
--
--
--
--
3
4.4
14.7
29.5
70. 5
85.3
95.6
--
--
--
-_
--
--
4
3.3
10. 5
19.4
32.3
67.7
80.6
89.5
96.7
--
--
--
--
5
2.5
8.2
14.6
22.6
34.2
65.8
77.4
85.4
91.8
97.5
--
--
6
2. 1
6.7
11.8
17.7
25. 0
35. 5
64.5
75.0
82.3
88.2
93.3
97. 9
Table 24. MINIMUM NUMBER
OF TEST POINTS
FOR RECTANGULAR DUCTS
Cross sectional area,
fl?
< 2.
2-25
> 25
Number of
test points
4
12
20
wide-range scale. The instrument is fairly rug-
ged, and its accuracy is suitable for most field
velocity determinations.
The meter consists of a pivoted vane enclosed in
a case, against which air exerts a pressure as it
passes through the instrument from an upstream
to a downstream opening; movement of the vane is
resisted by a hair spring and damping magnet.
The instrument gives instantaneous readings of
directional velocities on the indicating scale.
Calibrating the Velocity Meter
Before using a meter, check the zero setting. If the
pointer does not come to rest at the zero position,
turn the zero adjuster to make the necessary cor-
rections. The meter with its fittings is calibrated
as aunit; therefore, fittings cannotbe interchanged
from one meter to another. The serial number on
the fittings and on the meter must agree. If a
meter was originally calibrated for a filter, it
must always be used. Only connecting tubing of
the same length and inside diameter as that orig-
inally supplied with the meter should be used,
since changes in tubing affect the calibration of
the meter (Industrial Ventilation, 1956).
When the temperature of an air stream varies
more than 30° from the standard temperature of
70°F, or the altitude is more than 1, 000 feet, or
when both conditions are fulfilled, it is advisable
to make a correction for temperature and pres-
sure. Other correction factors, as shown in
Table 25 should also be used (Industrial Ventila-
tion, 1956).
Table 25. SOME CORRECTION FACTORS
FOR THE SWINGING-VANE VELOCITY METER
(Industrial Ventilation. I960)
Opening
Pressure opening3-
Hold meter jet against grille
(use gross area) more than
4 in. wide and up to 600 in.
area, free opening 70% or
more of gross area. Hold
meter jet against grille (use
free-open area)
Hold meter jet 1 inch in front
of grille (use gross area)
Suction opening
Square punched grille (use
free-open area)
Bar grille (use gross area)
Strip grille (use gross area)
Free open, no grille
Correction factor
0. 93
1. 00
0.88
0. 78
0.73
1.00
For pressure openings, it is advisable to use the
grille manufacturer's coefficient of discharge.
For suction openings, hold meter jet perpendicu-
lar to the opening, with the tip in the same plane
as the opening. This is very important because
velocities are changing very rapidly in front of a
suction opening.
Note: volume, cfm = area, ft x air velocity,
fpm x correction factor.
-------
Checking of Exhaust System
75
Uses of the Velocity Meter
Some uses of the meter and fittings are illustrated
in Figure 36. On large (at least 3 ft ) supply
openings, where the instrument itself 'will not seri-
ously block the opening and where the velocities
are low, hold the instrument itself in the air stream,
the air impinging directly in the left-hand port.
When the opening is smaller than 3 square feet,
or the velocities are above the no-jet scale, or
when both conditions hold true, appropriate fittings
mustbeused. On modern air-conditioning grilles,
the meter or fitting should be held between 1 and
2 inches in front of the grille.
If the exhaust opening is large (at least 3 ft2) and
the air velocities are low, as in spray booths,
chemical hoods, andsoforth, the meter itself can
be held in the air stream. The instrument should
be held so that the left-hand port of the meter is
flushwith the exhaust opening. When the opening
is smaller than 3 square feet, or the velocities
are above the no-jet scale, or-whenboth conditions
hold true, appropriate fittings must be used (In-
dustrial Ventilation, 1956).
2220 OR 3920
JET
3910. 2425,
OR 3290 JET
BLOWER
JET
SUPPLY SYSTEM
3930 JET
PLATING TANK
3910, 2425:
OR 3290
EXHAUST SYSTEM
GRINDER
EXHAUST
SPRAY BOOTH (NO-JET RANGE)
Figure 36. Some swinging-vane velocity meter applications (Industrial Ventilation
1960).
-------
76
DESIGN OF LOCAL EXHAUST SYSTEMS
COOLING OF GASEOUS EFFLUENTS
When designing an air pollution control system,
the designer must know the temperatures of the
gases to be handled before he can specify the
materials of construction for the system, the size
of ductwork, the size of the blower, and the type
of air pollution control device. Often, hot gases
must be cooled before being admitted to the con-
trol device. The cooling equipment will add to
the resistance of the flow of gases through the
exhaust system and may affect the volume and
composition of the gases. Since the gases must
pass through the cooling device, it must be de-
signed as an integral part of the exhaust system.
METHODS OF COOLING GASES
Although there are several methods of cooling hot
gases, those most commonly used in air pollution
control systems are: (1) Dilution with ambient air,
(2) quenching with water, and (3) natural convec-
tion and radiation from ductwork. In a. few cases,
forced-draftheat exchangers, air cooled and water
cooled, have been used.
With the dilution method, the hot gaseous effluent
from the process equipment is cooled by adding
sufficient ambient air to result in a mixture of
gases at the desired temperature. Natural con-
vection and radiation occur whenever there is a
temperature difference between the gases inside a
duct and the atmosphere surrounding it. Cooling
hot gases by this method requires only the pro-
vision of enough heat transfer area to obtain the
desired amount of cooling. The water quench
method uses the heat of vaporization of water to
cool the gases. Water is sprayed into the hot gas-
es under conditions conducive to evaporation, the
heat in the gases evaporates the water, and this
cools the gases. In forced-draft heat exchangers,
the hot gases are cooled by forcing cooling fluid
past the barrier separating the fluid from the hot
gases.
Dilution With Ambient Air
The cooling of gases by dilution with ambient air is
the most simple method that can be employed.
Essentially, it involves the mixing of ambient air
with a gas of known volume and temperature to
produce a low-temperature mixture that can be
admitted to an air pollution control device. In de-
signing such a system, first determine the volume
and temperature of air necessary to capture and
convey the air contaminants from a given source.
Then calculate the amount of ambient air required
to provide a mixture of the desired temperature.
The air pollution control device is then sized
to handle the combined mixture.
Although little instrumentation is required, a gas
temperature indicator •with a •warning device, at
the very least, should be used ahead of the air
pollution control device to ensure that no damage
occurs owing to sudden, unexpected surges of
temperature. The instrumentation may be ex-
panded to control either the fuel input to the
process or the volume of ambient air to the ex-
haust system.
This method of cooling hot gases is used exten-
sively where the hot gases are discharged from
process equipment in such a -way that an external
hood must be used to capture the air contaminants.
The amount of air needed to ensure complete cap-
ture of the air contaminants is generally sufficient
to cool the gases to approximately 500°F, which
permits the use of high-temperature air pollution
control devices. When the volume of hot gases is
small, this method may be used economically even
when much more air is needed to achieve the de-
sired cooling than that needed for adequate capture
of air contaminants.
When large volumes of hot gases require cooling,
the size of the exhaust system and control device
becomes excessively large for dilution cooling.
In any case, compare the costs of installation and
operation of the various cooling methods before
deciding -which method to use.
The following examples illustrate (1) a method of
determining the resultant temperature of the mix-
ture of the hot furnace gases and the ambient air
induced at the furnace hood, and (2) a method of
determining the volume of air needed to cool the
hot furnace gases to a selected temperature.
Example 14
Given:
Yellow-brass-melting crucible furnace. Fuel
burned: 1, 750 cfh natural gas with 20 percent
excess air.
Maximum gas temperature at furnace outlet:
2,500°F.
Volume of dilution air drawn in at the furnace
hood: 4, 000 cfm.
Maximum temperature of dilution air: 100°F.
For this problem, neglect the heat losses due
to radiation and natural convection from the
hood and ductwork. Assume complete combus-
tion of the fuel.
-------
Cooling of Gaseous Effluents
77
TO CONTROL DEVICE
PRODUCTS OF
COMBUSTION
ITEMP = f
2,500 F
h M
: :
V /
CRUCI BLE
' FURNACE
| HOOD
^v DILUTION AIR
4,000 cfm
FUEL
1,750 Cfh
Figure 37. Problem: Determine the tempera-
ture of the gases entering the control device.
Solution:
1. Determine the weight (W) and heat (Q) above
60 °F in the products of combustion (PC):
Q = ZQ.
W = Z W
1
Q. = W. h.
i 11
where
W. = weight of individual gas flowing, Ib/min
h. = enthalpy above 60 °F of each gas, Btu/lb
Q. = heat above 60°F in each gas, Btu/min
Convert fuel rate to cfm lj 75° = 29. 17 cfm
60
Referring to the calculation data of Table
26, Wi = 29. 14 Ib/min and Qj = 21,470
Btu/min
2. Heat above 60 °F in 4, 000 cfm of dilution air
entering hood at 100°F:
Density of air at 100°F = 0.0708 Ib/ft (from
Table Dl, Appendix D)
Enthalpy of air at 100 °F = 9. 6 Btu/lb (from
Table D3, Appendix D)
Weight of dilution air = (4, 000)(0. 0708) =
283. 2 Ib/min
Heat above 60°F in dilution air = (9. 6)(283.2)
= 2, 720 Btu/min
3. Enthalpy of mixture of PC and dilution air:
Total weight of mixture = 29. 14 + 283. 2
= 312.3 Ib/min
Heat above 60 °F in mixture 21, 470 + 2, 720
= 24, 190
24 190
Enthalpy of mixture =
= 77.4 Btu/lb
4. Temperature of mixture:
To determine the temperature of the mixture,
determine the enthalpy of the mixture at two
temperatures, preferably above and belowthe
calculated enthalpy of 77. 4 Btu/lb.
Table 26. CONVERSION VALUES
FOR ITEM 1, EXAMPLE 14
PC per cubic foot
of fuel
Component
C°2
H2°
N2
°2
Totals
Weight,
Ib/ft3
0. 132a
0. 099
0. 731
0. 037
PC from
furnace
Ib/min
3.85b
2.89
21.32
1.08
29. 14
Enthalpy
of
component
hi at 2, 500°F,
Btu/lb
690.2°
1, 318. 1
672.3
621.0
Heat above
60 °F in
component
Btu/min
2,660d
3,810
14,330
670
21,470
aFrom Table D7, Appendix D.
bW. = (0. 132 Ib/ft3)(29. 17 ft3/min) = 3.85 Ib/min
cFrom Table D3, Appendix D.
dQ. = (3.85 lb/min)(690.2 Btu/lb) = 2,660 Btu/min
-------
78
DESIGN OF LOCAL EXHAUST SYSTEMS
Then, by interpolation, the temperature cor-
responding to 77. 4 Btu/lb can be determined.
Since the mixture contains mostly nitrogen,
the enthalpies should be close to those of ni-
trogen. From Table D3 it appears that the
mixture temperature will be between 350°F
and 400°F. The enthalpy of the mixture, H ,
is:
Q
W
at desired temperatures
where
W = Z W.
m i
Q = Z W. h.
m 11
The O, and N, from the dilution air must be added
to the O2 and N2 from the PC.
Weight of dilution air = 283.2 Ib/min
O2 content = (283. 2)(0. 21) = 59. 5 Ib/min
N2 content = (283. 2)(0. 79) = 223. 7 Ib/min
Referring to the calculation data of Table 27:
h at 350°F = 22',53,0 = 72. 1 Btu/lb
m 312.3
h at 400°F = 2,6'15,° = 83. 6 Btu/lb.
m 312.3
By interpolation the mixture temperature = 373 °F
Therefore, the exhaust system and control device
must be designed to handle gases at 373 °F.
Example 15
Problem:
Using the same given data in Problem No. 1,de-
termine the amount of dilution air required to re-
duce the temperature of PC to 300°F.
Solution:
1. Heat above 60°F in PC at 2, 500°F = 21, 470
Btu/min (From Table 26)
2. Heat lost by PC in cooling from 2, 500° to
300°F: From. Table 26 obtain the weight (Wi)
of each component of PC discharged from the
furnace. From Table D3, obtain the enthalpy
(hj) of each component at 300 °F. Referring
to the calculation data of Table 28: Heat to
be lost = 21,470 - 1,847 = 19,623 Btu/min
3. Volume of air needed to cool PC to 300°F:
Air inlet temperature = 100°F (given)
Final air temperature = 300°F
h at 100°F = 9. 6 Btu/lb (from Table D3)
h at 300C'F = 57.8 Btu/lb (from Table D3)
Ah = 49. 2 Btu/lb
Weight of air needed = ——5—-— = 408 Ib/min
Table 27. CONVERSION VALUES
FOR ITEM 4, EXAMPLE 14
Component
co2
H2°
N2
°2
Totals
wi;
Ib/min
3.85
2.89
245. 02a
60.58C
312. 3b
>Md
at 350°E
Btu/lb
63.1
131.3
73. 3
64.8
v, d
hi
at 400°F,
Btu/lb
74. 9
154. 3
84.9
76.2
Qi = hiWi
at 350°F,
Btu/min
- 242. 9e
379.2
17, 980. 0
3, 925.0
22, 530
Qi = h^
at 400°F,
Btu/min
288.4
455.6
20, 800.0
4, 615.0
26, 150
a W^ of N2 is sum of N2 from PC and dilution air.
^Totals are rounded off to four significant figures.
CW- of O? is sum of O-> from PC and dilution air.
i £
From Table D3, Appendix D.
eQ = (3.85 Ib/min) (63. 1 Btu/lb) = 242.9 Btu/min,
i
-------
Cooling of Gaseous Effluents
79
Table 28. CONVERSION VALUES
Gaseous
components
C°2
H2°
N2
°2
Totals
Wif1
Ib/min
3.85
2. 89
21. 32
1. 08
29. 14
hj at 300°F,
Btu/lb
51. 3
108. 2
59.8
53. 4
Qi = hjWi.
at 300°F Btu/min
197. 7
312. 5
1, 279. 0
57. 5
1, 846. 7
Total heat above 60 °F in PC at 300 °F = 1, 847 Ib/min
Volume of dilution air at 100°F
p at 100°F = 0. 070S lb/ft3 (From Table Dl)
408
Volume =
0.0708
= 5, 760 cfm
The exhaust system must be designed to handle
a sufficient volume of gases at 300"F to pro-
vide an indraft of dilution air of 5, 760 cfm in ad-
dition to the products of combustion.
Quenching With Water
When a large volume of hot gas is to be cooled and
only a small quantity of dilution air is needed to
capture the air contaminants, some methods of
cooling other than dilution with ambient air should
be used. Since the evaporation of water requires
a large amount of heat, the gas can be cooled sim-
ply by spraying water into the hot gas.
For efficient evaporation of water in a gas stream,
it has been determined that the gas velocity should
be from 500 to 700 fpm and the entire cross sec-
tion of the stream should be covered with a fine
spray of water. If, however, water carryover is
undesirable, as in a baghouse, satisfactory set-
tling of the water droplets must be attained; hence,
lower velocities are employed. Eliminator plates
are seldom used in installations -where excessive
maintenance due to corrosion or fouling is expected.
To reduce further the likelihood of water droplet
carryover, place the water spray chamber as far
from the baghouse as practical.
Water spray pressures generally range from 50
to 150 jasigj however, to reduce the amount of
moisture collected, some installations have em-
ployed pressures as high as 400 psig. Since the
moisture collected in spray chambers readily cor-
rodes steel, the chambers are frequently lined with
materials resistant to corrosion.
If the gases discharged from the basic equipment
are exceptionally hot, as are those from the cupo-
la furnace, the first portion of the duct should be
refractory lined or made from stainless steel. In
some cases, stainless steel ducts -with water sprays
have been used between the furnace and the quench
chamber.
For controlling the gas temperature leaving the
quench chamber, a temperature controller is
generally used to regulate the amount of water
sprayed into the quench chamber. For emergen-
cy conditions, a second temperature controller
can be used to divert excessively hot gases away
from the air pollution control device.
Cooling hot gases with a water quench is relative-
ly simple and requires very little space. Figure
38 shows a quench chamber used to cool the gas-
Figure 38. A quench chamber in a baghouse
control system serving a cupola furnace
(Harsell Engineering Company, Inglewood,
Cali fornia).
-------
80
DESIGN OF LOCAL EXHAUST SYSTEMS
ecus effluent from a cupola furnace. O_uench cham-
bers are little more than enlarged portions of the
ductwork equipped with water sprays. They are
easy to operate and, with automatic temperature
controls, only that amount of water is used that
is needed to maintain the desired temperature of
the gases at the discharge. Their installation and
operating costs are generally considered to be less
than for other cooling methods. Quench chambers
should not be used when the gases to be cooled con-
tain a large amount of gases or fumes that become
highly corrosive when wet. This creates addition-
al maintenance problems, not only in the quench
chamber, but in the remainder of the ductwork,
the control device, and the blower.
The following example will illustrate some of the
factors that must be considered when designing a
quench chamber to cool the gaseous effluent from
a gray-iron-melting cupola.
Example 16
Given:
32-in.-I. D. cupola. Maximum temperature of
gaseous effluent at cupola outlet - 2, 000 °F.
Weight of gaseous effluent at cupola outlet -
216 Ib/min.
Volume of gaseous effluent at cupola outlet -
13, 280 cfm at 2, 000°F. This volume of effluent
includes indraft air at the charging door of the cu-
pola. The temperature of 2, 000°F is a maximum
temperature.
Assume the effluent gases have the same proper-
ties as air. Consideration of the enthalpies of the
gaseous constituents in the effluent gas stream will
show that this is a valid assumption. Any correc-
tions would introduce an insignificant refinement
to the calculations when considered with respect to
the accuracy of other design factors.
TEMP - 2,000^
I 3 , 280 cfm
EVAPORATIVE
COOLING WATER - ?
Y
WATER SPRAY
CONDITIONING CHAMBER
TO CONTROL DEVICE
TEMP - 225eF
COOL ING WATER
OUT
Figure 39. Problem: Determine the water
needed to cool the gaseous effluent to 225°F
and the total volume of gases discharged
from the quench chamber.
Solution:
1. Cooling required:
Enthalpy of gas at 2, 000°F = 509. 5 Btu/lb
Enthalpy of gas at 225°F = 39.6 Btu/lb
Ah = 469. 9 Btu/lb
(216)(469. 9) = 101, 300 Btu/min
2. Water to be evaporated:
Heat absorbed per Ib of water:
Q = h (225°F, 14. 7 psig)-hf (60°F)
O
= 1, 156.8 - 28.06 = 1, 128. 7 Btu/lb
Water required = —— ' " = 90 Ib/min
3. Volume of water evaporated at 225°F:
379 /460 + 225N
/460 + 225\
\ 460 + 60 /
18 V 460 + 60
(90) = 2, 510 cfm
4. Total volume vented from spray chamber at
225°F:
Cupola = (13, 280) f —
Water = 2, 510 cfm
Total = 6, 210 cfm
225 + 460\
000 + 460 /
= 3, 700 cfm
Problem Note: In this example, the calculated
amount of water required to cool the gases, 90
Ib/min or 10.8 gal/min, is only the water that
must be evaporated. Since all the water sprayed
into a quench chamber does not evaporate, the
pump and spray system should be sized to supply
more water than that calculated. The amount of
excess water needed will depend on factors such
as the inlet temperature of the gases, the tem-
perature drop required, the fineness of the water
spray, and the arrangement of spray heads. It
is not uncommon to size the pump to give 200 per-
cent of the water needed for evaporation. The ac-
tual amount of water used should be controlled by
the temperature of the gases discharged from the
quench chamber.
The loss of heat by radiation and convection from
the ducts was neglected. With long duct runs, how
ever, a considerable temperature drop in the gas-
-------
Cooling of Gaseous Effluents
81
eous effluent could occur, especially if the quench
chamber was installed near the downstream end of
the ductwork. If the quench chamber is placed near
the control device, adequate water entrainment sep-
arators must be employed.
Natural Convection and Radiation
When a hot gas flows through a duct, the duct be-
comes hot and heats the surrounding air. As the
air becomes heated, natural drafts are formed
carrying the heat away from the duct. This phe-
nomenon is called natural convection. Heat is also
discharged from the hot duct to its surroundings by
radiant energy.
The rate of heat transfer is a function of the resis-
tances to heat flow, the mean temperature differ-
ence between the hot gas and the air surrounding
the duct, and the surface area of the duct. It may
be expressed as:
where
Q
U
A
At
m
Q = UAAt (33)
= rate of heat transfer, Btu/hr
= overall heat transfer coefficient, Btu/
hr-°F-ft2
= heat transfer area, ft
= log-mean temperature difference, °F.
The rate of heat transfer is determined by the
amount of heat to be removed from the hot gas-
eous effluent entering the exhaust system. For
any particular basic process, the -weight of gas-
eous effluent and its maximum temperature are
fixed. The cooling system must, therefore, be
designed to dissipate sufficient heat to lower the
effluent temperature to the operating temperature
of the air pollution control device to be used.
The rate of heat transfer can be determined by the
enthalpy difference of the gas at the inlet and out-
let of the cooling system.
Q = WAh
(34)
of the gas. This term, too, is fixed for a partic-
ular process. It is calculated as follows:
At
vv -
l°sc
vv
(35)
where
t = gas temperature of inlet, °F
t = gas temperature at outlet, °F
t = air temperature, °F.
In many processes the temperature of the gaseous
effluent is not constant but varies during different
operational phases. The atmospheric tempera-
tures also vary a great deal. In such cases, the
cooling system must be designed for the worst con-
ditions that prevail to ensure adequate cooling at
all times. The inlet temperature (t,) chosenmust
be the maximum temperature of the gas entering
the system; t^ must be the maximum allowable
temperature of the gas discharged from the cool-
ing system; and ta must be the maximum expected
atmospheric temperature.
The overall coefficient of heat transfer, U, is the
reciprocal of the overall resistance to heat flow.
It is a function of the individual heat transfer coef-
ficients, which can be estimated by empirical equa-
tions. U must be based on either the inside or out-
side surface of the duct. For radiation-convection
cooling, it is generally based on the outside surface
and is denoted by Uo. Uo is defined by the following
equation (Kern, 1950):
U
h. h
10 o
h. +h
10 o
(36)
where
•where
h. = inside film coefficient based on the out-
side surface area, Btu/hr- °F-ft2
h = outside film coefficient, Btu/hr-°F-ft,
o
The inside film coefficient can be solved by the
formula (Kern, 1950):
W = weight of gas flowing, Ib/hr
Ah = enthalpy change between inlet and out-
let, Btu/lb.
The log-mean temperature difference is the dif-
ference in temperature bet-ween the air surround-
ing the duct, and the inlet and outlet temperature
. k /cA
\ = JnrJ (ifj
where
H
h.D
(37)
-l/3
and is plotted against
-------
82
DESIGN OF LOCAL EXHAUST SYSTEMS
Reynolds number (Re) as shown in.Fig-
ure 40
k = thermal conductivity, Btu/hr-ft-°F
D = inside diameter of duct, ft
C = heat capacity, Btu/lb-°F
(JL = viscosity, Ib/hr-ft.
oc
00
00
00
00
40
10
5
4
1
c = SPEC
D = INS
G = MASS
h - INS
k = THER
L = LENt
;» =»EI
To^S
"^^*$r
^"5rT
^^ffe'
^^s£
&*\
AREA THRO
IFIC HEAT,
OE DUMETE
VELOC TY,
DE FILM CO
MAL CON DUG
TH QF PftTH
•HT FLOW OF
/
*-£4
fl
'II
ft
1
/
W
I—
GH TUBES, f
Btu/lb-°F
R OF TUBES,
w/at, Ib/hr
2
t
ft2
tu/hr-ft!-°F
nv TV, Btu/hr-ft-°F
n
FLU D, Ib/hr
I I T T
4tt
1
Jr
f
^
S
s
r
/
/
/
/
J
/
r
i
1,000 2,000 4,000 10,000 20,000 50.000 100,000 200,000
REYNOLDS NUMBER (Re), &£
Figure 40. Tube-side heat-transfer curve
(Adapted from Sieder and Tate in Kern, 1950).
The Reynolds number is a function of the duct di-
ameter, the mass velocity, and viscosity of the
gas. It is calculated by the equation
Re = ^ (38)
where
G
W 2
= mass velocity = — Ib/hr-ft and
seen that an increase in Re will increase the rate
of heat transfer. Since the weight of gas flowing
is fixed, Re can be increased only by increasing
the velocity of the gas. It has already been shown
that an increase in velocity will increase the power
required to move the gases through the exhaust
system. Consequently, the optimum velocity for
good heat transfer at reasonable blower-operating
costs must be determined. It is known that a sac-
rifice in heat tramsfer rate to obtain lower blower
horsepower results in the most economical cooling
system. Owing to the many variables involved, how-
ever, each system must be calculated on its own
merits.
The outside film coefficient (hc) is the sum of the
coefficient due to natural convection (hc) and the
coefficient due to radiation (hr). An empirical
equation for hc for vertical pipes more than 1 foot
high and for horizontal pipes is (McAdams, 1942):
= 0. 27
0. 25
D
(39)
where
At = the temperature difference between the
outside duct wall and the air, t - t ,
°F
D = outside duct diameter, ft.
o
The radiation coefficient is computed from
(McAdams, 1942):
VT2
= 0.173 ^(Tl/lOO)4 - (T2/100)41 (40)
where
= flow area inside the duct =
TT D
e = emissivity of the duct surface, dimen-
sionless
a = Stefan-Boltzmann constant, 0. 173 x
JO'8 Btu/ft^-lv-'R4
T = absolute temperature of the duct sur-
face, °R
T = absolute temperature of the air, °R.
The inside film coefficient is a measure of the flow
of heat through the inside film. An increase in h^
will, therefore, increase the rate of heat trans-
ferred from the gas to the atmosphere. It can be
In Figure 41, Tj is plotted against hr for several
air temperatures; hr -was calculated for an emis-
sivity equal to 1.0. To obtain hr for a system,
multiply the hr found from Figure 41 by the emis-
sivity of the duct surface. Since the emissivity of
-------
Cooling of Gaseous Effluents
83
the surface is a function of the surface condition,
and SL black surface generally gives the highest
emissivity, the ductwork should be blackened.
T| - DUCT SURFACE TEMPERATURE, °R
T, -AIR TEMPERATURE, °R
THE DUCT SURFACE AND AIR TEMPERATURES
ARE PLOTTED IN °F.
200
400
600
800
1,000
1.200
DUCT SURFACE TEMPERATURE, °F
Figure 41. Coefficient of heat transfer
by radiation for e = 1.0 (Adapted from
McAdams, 1942).
When calculating ho, assume the temperature of
the duct wall (tw) and then check. The assumed
tw can then be checked with the following equa-
tion (Kern, 1950):
= t
h + h.
o 10
(t -t )
m a
(41)
•where
= the average gas temperature, °F.
If tw is not the same as assumed tw, estimate a
new tw and recalculate ho. When the assumed tw
and calculated t are the same, use the corre-
spending hQ to calculate Uo.
The heat transfer area (A) can now be calculated.
The length of duct needed to give the necessary
area is then calculated by using the outside di-
ameter used in determining the film coefficients.
If the length of duct needed is large, the ductwork
will probably be arranged in vertical columns to
conserve flo'or space. Figure 42 shows such an
installation serving a lead blast furnace and a lead
reverberatory furnace. The columns require sev-
eral 180° bends, which will offer a large resis-
tance to the flow of gas. To minimize these loss-
es, the gas velocity should be low, preferably
less than normal dust-conveying velocities. By
joining the bottoms of the columns with hoppers,
any dust settling out as a result of low velocities
can be collected without fouling the exhaust sys-
tem. If the cooling area is such that a single loop
around the plant or across a roof is sufficient, avoid
sharp bends and maintain carrying velocities. When
gases are cooled through a large temperature- range,
the volume will be reduced, so that smaller di-
ameter ductwork may be needed as the gases pro-
ceed through the cooling system. With cooling
columns, the diameter of the duct joining the last
column and the air pollution control device must
be sized properly to provide suitable conveying ve-
locities for the cooled effluent.
For most convection-radiation cooling systems,
the only equipment used is sufficient ductwork to
provide the required heat transfer area and, of
course, a blower of sufficient capacity to move the
gaseous effluent through the system. Unless the
temperature of the gases discharged from the ba-
sic process is exceptionally high, or there are
corrosive gases or fumes present, black iron duct-
work is generally satisfactory. The temperature
of the duct wall can be determined for any portion
of the ductwork by using the method previously
described for determining tw. If tw proves to be
greater than black iron can withstand, either use
a more heat resistant material for that portion of
the system or recirculate a portion of the cooled
gas to lower the gas temperature at the inlet to
the cooling system.
With this type of cooling, flexibility in control-
ling the gas temperature is limited. When either
the gas stream or air temperatures, or both, are
lower than design values, the gases discharged
from the cooling device will be less than that cal-
culated, and condensation of moisture from the
effluent within the control device might result.
Conversely, •when design temperatures are ex-
ceeded, the temperature of the gases discharged
from the cooling system could become too high.
To avoid damage to the air pollution control de-
vice, install a quick-response temperature con-
troller to warn the operator of the change in tem-
perature so that proper adjustments can be made.
The radiation-convection cooling system is in
operation whenever hot gases are being conduc-
ted through the exhaust system. The gases being
cooled are not diluted with any cooling fluid. The
exhaust system blower and the air pollution con-
trol device need not be sized for an extra volume
of gases due to dilution. Since no water is used,
there is no need for pumps, and corrosion prob-
lems are nonexistent. On the other hand, these
cooling systems require considerable space, and
blower horsepower requirements are high owing
to the additional resistance to gas flow.
-------
84
DESIGN OF LOCAL EXHAUST SYSTEMS
Figure 42. Radiation-convection cooling columns in an air pollution system serving a lead blast furnace
and a lead reverberatory furnace (Western Lead Products Company, City of Industry, California).
The following example illustrates a method of de-
termining the heat transfer area needed to cool
the gaseous effluent from the cupola of example 16
with a natural convection-radiation cooler.
Example 17
Given:
capacity of the gaseous constituents in the effluent
gas stream will show that this is a valid assump-
tion. Any correction would introduce an insignifi-
cant refinement to the calculations when considered
with respect to the accuracy of other design factors.
3Z-in.-I. D. cupola.
Gaseous effluent = 12, 960 Ib/hr.
Maximum temperature of effluent = 2, 000°F.
Volume of effluent at 2,000°F = 13,280 cfm. This
volume of effluent includes indraft air at the charg-
ing door of the cupola. The temperature of 2, 000
°F is a maximum.
The vertical cooling columns must be located a
minimum of 60 feet from the cupola.
Assume the effluent gases have the same physical
properties as air. Consideration of the enthalpy,
viscosity, thermal conductivity, density, andheat
TEMP • ?,000Cf
A A A
TO CONTROL DEVICE
TEMP - 225 F
V V V
COOLING COLUMNS
Figure 43. Problem: Determine the length
of duct needed to cool the gases to 225°F
by natural convection-radiation columns.
-------
Cooling of Gaseous Effluents
85
Solution:
1. Heat (Q) to be transferred:
Enthalpy of gas (2, 000°F) = 509. 5 Btu/lb
(from. Table D3, Appendix D)
Enthalpy of gas (225°F) = 39. 6
AH = 469.9
Q = (469. 9)(12,960) = 6, 078, 000 Btu/hr
2. Determine logarithmic mean temperature dif-
ference (At ):
m
Gas inlet temperature (t )
= 2,000°F
= 225°F
Cooling air temperature (t ) = 100°F
Gas outlet temperature (t,,)
log
(t.-t )
1 a
e (t -t )
2 a
j = (See Figure 40) = 215
(b) Obtain k, C, and—^ from Table Dl
k
k = 0.0297
C = 0.247
Cjj. = 0.775
k
(c) Substitute above data in formula, and
solve for h.:
i
h. = 215 [ —
'/3 = 2.66Btu/hr-ft2-°F
4. Convert h^ to inside film coefficient (h^0)based
on outside surface area:
Use a 10-gage duct wall, thickness = 0. 141
inch
(2,000-100) - (225-100)
1, 900
= 653°F
log
125
D =(2.2)
o
12
h.Q = (2.66) = 2.62 Btu/hr-ft-°F
3. Determine inside film coefficient (h ):
i
1/3
\k/
(a) Obtain j from Figure 40:
H
DG
Re =
Using a design velocity of 3, 500 fpm in the
horizontal section at the cupola discharge:
. 13,280 cfm , _0 ,2
Area = , ..., . = 3.79ft
3,500 fpm
Pipe diameter (D) =r3'7J)(4M =2.2
ft
12-960 Ib/hr = 3,4201b/hr-ft2
3. 79 ft
5. Determine the outside film coefficient (h ):
h = h + h
o c r
(a) h = 0.:
0.25
Assume a duct wall temperature of 525 °F
h = 0.
c
"•"(rnr)
0. 25
= 1. 00 Btu/hr-ft -°F
(b) Obtain h from Figure 41:
h =3.42 (Emissivity = 1.0)
Use an emissivity of 0. 736 for rusted black
iron duct
h =(3.42)(0.736) = 2. 52 Btu/hr-ft2-"F
Re =
=0. 094 Ib/hr (from Table Dl)
(2.2)(3,420)
0. 094
= 80,000
(c) h = h + h
o c r
= 1.00 + 2.52 = 3.52 Btu/hr-ft -°F
-------
86
DESIGN OF LOCAL EXHAUST SYSTEMS
(d) Since t was assumed, it must be checked
i w
as shown:
t =t - I, ,\ | (t - t )
w m I h0 + hio ' m a
t =2-°00/ 225 = 1.112-F
m t-
= 100°F
t = 1, 112
w
/ 3.5Z
I 3.52 + 2.
62
(1, 112 - 100) = 530°F
The assumed tw was 525 °F, which che.cks
closely with 530 °F
6. Determine the overall heat transfer coefficient
(U ) based on the outside surface area:
io o
7. Determine heat transfer area (A):
A =
Q
U A t
od m
6.078,000
(1.50)(653)
8. Determine length of duct (L) required:
L =
6, 210
(2.224)(7T)
= 886 ft
The duct from the cupola to the vertical column
is 60 feet long. The length of duct in the col-
umn section will, therefore, be 886 - 60 = 826
feet.
If columns are 50 feet high, then 826/50, or
16. 5 columns will be required. Since the con-
necting duct-work between columns will con-
sist of at least 2 feet of duct between each col-
umn, a total of 16 columns 50 feet high will be
required.
Problem Note: The example illustrates one meth-
od of determining the length of duct needed to cool
a given hot gaseous effluent. To determine the op-
timum duct diameter, it is necessary to make simi-
lar calculations for other duct diameters , and then
determine the pressure drop through each system.
By comparing the construction costs with the oper-
ating costs, the optimum duct diameter can be found.
Forced-Draft Cooling
Heat transfer by convection is due to fluid motion.
Cold fluid adjacent to a hot surface receives heat,
which is imparted to the bulk of the fluid by mixing.
With natural convection, the heated fluid adjacent
to the hot surface rises and is replaced by colder
fluid. By agitating the fluid, mixing occurs at a
much higher rate than with natural currents, and
heat is taken away from the hot surface at a much
higher rate. In most process applications, the agi-
tation is induced by circulating the fluid at a rapid
rate past the hot surface. This method of heat trans
fer is called forced convection. Since forced con-
vection transfers heat much faster than natural con-
vection, most process applications use forced-con-
vection heat exchangers. Whenever possible, heat
is exchanged between hot and cold streams to re-
duce the heat input to the process. There are, how-
ever, many industrial applications -where it is not
feasible to exchange heat, and so a cooling fluid such
as water or air is used, and the heat removed from
the stream is dissipated to the atmosphere. When
water is used, the heat is taken from the process
stream in a shell and tube cooler, and the heat
picked up by the water is dissipated to the atmo-
sphere in a cooling tower. When air is used as the
cooling medium Ln either shell and tube or fin tube
coolers, the heated air is discharged to the atmo-
sphere and is not recirculated through the cooler.
With forced-convection cooling, the temperature
of the cooled stream can be controlled -within nar-
row limits even -with widely varying atmospheric
or water temperatures. Heat transfer area is
greatly reduced from that needed with natural con-
vection. Power requirements to force the process
stream through the cooler are generally less. On
the other hand, either a pump or a blower is needed
to circulate the cooling fluid through the cooler.
With -water cooling, a cooling tower may be needed
and additional maintenance is required to clean
scale from the tubes.
FACTORS DETERMINING SELECTION OF COOLING DEVICE
Cooling by dilution air is commonly used where
conveying air volumes are low or -where there is
a large volume of dilution inherent in the hoods
required to capture the air contaminants. If large
gas volumes are necessary, and dilution air is not
economical, then direct cooling with water quench
chambers is generally favored over other cooling
devices. This is probably due to the small space
requirements, ease of operation, and low instal-
lation costs of the water quench chambers. When
the characteristics of the gaseous effluent and the
contaminants are such that -water cannot be used,
natural convection-radiation cooling is generally
employed. The ease of operation and low main-
tenance costs make these cooling systems more
-------
Cooling of Gaseous Effluents 87
attractive than forced-convection coolers. Infact, it has been used where the heat of the cooling air
forced-convection equipment has seldom been used can be utilized, for example as combustion air in
in air pollution control installations. In some cases the basic process being controlled.
-------
CHAPTER 4
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
INERTIAL SEPARATORS
HOWARD DEY, Air Pollution Engineer
JOHN MALONEY, Air Pollution Engineer*
JOSEPH D'IMPERIO, Air Pollution Engineert
WET COLLECTION DEVICES
EDWIN J. VINCENT, Intermediate Air Pollution Engineer
BAGHOUSES
HERBERT SIMON, Senior Air Pollution Engineer
SINGLE-STAGE ELECTRICAL PRECIPITATORS
HERBERT SIMON, Senior Air Pollution Engineer
TWO-STAGE ELECTRICAL PRECIPITATORS
ROBERT C. ADRIAN, Intermediate Air Pollution Engineer I
OTHER PARTICULATE-COLLECTING DEVICES
EDWIN J. VINCENT, Intermediate Air Pollution Engineer
*Now with the Air Pollution Control District of Riverside County, California.
'Now deceased.
+ Now with Aerojet-General Corporation, Azusa, California.
-------
CHAPTER 4
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
Air pollution control equipment may be classified
into two groups: (1) Equipment controlling partic-
ulate matter, and (2) equipment controlling gas-
eous emissions. From an air pollution viewpoint,
particulate matter is any material that exists as
a solid or liquid at standard conditions. Some ex-
amples of particulates are smoke, dusts, fumes,
mists, and sprays.
Devices for control of particulate matter are avail-
able in a "wide variety of designs using various prin-
ciples of operation and having a wide latitude in col-
lection efficiency, initial cost, operating and main-
tenance costs, space, arrangement, and materials
of construction. In selecting the optimum device
for a specific job, it is necessary to consider many
factors. Rose et al., (1958) consider the following
factors significant:
1. Particulate characteristics, such as particle
size range, particle shape, particle density,
and physico-chemical properties such as ag-
glomeration tendencies, corrosiveness, hygro-
scopic tendencies, stickiness, inflammability,
toxicity, electrical conductivity, and so forth.
2. Carrier gas characteristics, such as temper-
ature, pressure, humidity, density, viscosity,
dew points of condensable components, elec-
trical conductivity, corrosiveness, inflamma-
bility, toxicity, and so forth.
3. Process factors, such as volumetric gas rate,
particulate concentration, variability of ma-
terial flow rates, collection efficiency require-
ments, allowable pressure drop, product qual-
ity requirements, and so forth.
4. Operational factors, including structural lim-
itations such as head room, floor space, and
so forth, and equipment material limitations
such as pressure, temperature, corrosion ser-
vice requirements, and so forth.
In this chapter, devices for control of particulate
matter have been grouped into six classes: (1)
Inertial separators, (2) wet collection devices,
(3) baghouses, (4) single-stage electrical precip-
itators, (5) two-stage electrical precipitators,
and (6) other particulate-collecting devices.
INERTIAL SEPARATORS
Inertial separators are the most widely used de-
vices for collecting medium- and coarse-sized
particulates. The construction of inertial sep-
arators is usually relatively simple, and initial
costs and maintenance costs are generally lower
than for most other types of dust collectors. Col-
lection efficiencies, however, are usually not high.
Although suitable for medium-sized particulates
(15 to 40 ji), ordinary inertial separators are gen-
erally unsuitable for fine dusts or metallurgical
fumes. Dusts with a particle size ranging from
5 to 10 microns are normally too fine to be collec-
ted efficiently. In some cases, however, small-
diameter, high-efficiency cyclones can be effec-
tive in collecting particles in the 5-micron range.
Inertial separators operate by the principle of im-
parting centrifugal force to the particle to be re-
moved from the carrier gas stream. This force
is produced by directing the gas in a circular path
or effecting an abrupt change in direction.
SINGLE-CYCLONE SEPARATORS
A cyclone, which is an inertial separator without
moving parts, separates particulate matter from
a carrier gas by transforming the velocity of an
inlet stream into a double vortex confined -within
the cyclone. In the double vortex the entering gas
spirals downward at the outside and spirals up-
ward at the inside of the cyclone outlet. The par-
ticulates, because of their inertia, tend to move
toward the outside wall, from which they are led
to a receiver.
Cyclones can be designed to handle a wider range
of chemical and physical conditions of operation
than most other types of collection equipment can
handle. Any conditions for which structural ma-
terials are available can be met by a cyclone, if
the degree of collection falls within the operating
range of the cyclone, and physical characteristics
of the particulates are such that no fouling of the
cyclone or excessive wall buildup occurs.
Because of its versatility and low cost, the single-
cyclone separator is probably the most widelyused
of the dry centrifugal separators. These cyclones
are made in a wide variety of configurations. Al-
though many design factors must be considered,
91
-------
92
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
the degree of collection efficiency is most depen-
dent upon the horsepower expended. Hence, cy-
clones with high inlet velocities, small diameters,
and long cylinders are generally found most effi-
cient. They are commonly called pencil cyclones
or high-efficiency cyclones. Figure 44 shows a
single high-efficiency cyclone, with typical dimen-
sion ratios as follows:
combined into a dimensionless quantity called the
separation factor:
Major cylinder diameter
Major cylinder length
Cone length
Gas outlet diameter
Gas outlet length
Gas inlet height
Gas inlet width
Dust outlet
D
L = 2 D
c c
Z = 2 D
c c
D
D
H + S = 5/8 D
c c c
H
D
c
2
D
c
4
D
In Figure 44, this cyclone consists of a cylinder
with a tangential gas inlet, an axial gas outlet, and
a conical lower section with an axial dust outlet.
The gas inlet is a rectangular opening, with the
height of the opening equal to twice the width. The
gas outlet is a tube approximately one half the di-
ameter of the major cylinder, concentric with and
extending inside the major cylinder to slightly be-
low the lower edge of the gas inlet. The tangen-
tial, high-velocity gas entry imparts a circular
motion to the gas stream; the particulates, because
of their greater inertia, tend to concentrate on the
wall of the cyclone. The inlet gas follows a dou-
ble vortex path, spiraling downward at the outside
and spiraling upward at the inside to the gas out-
let. Figure 45 illustrates the double-vortex path
of the gas stream. The downward spiral, assis-
ted by gravity, carries the particulates downward
to the dust outlet where they drop into a dusttight
bin, or are removed by a rotary valve or screw
conveyor.
Theory of Operation
The centrifugal force applied to particulates varies
as the square of the inlet velocity and inversely as
the radius of the cyclone. These factors have been
S =
V
(42)
where
S =
V =
g =
separation factor
inlet velocity, ft/ sec
cyclone cylinder radius, ft
gravitational constant, 32. 2 ft/sec2.
It has not been possible to establish a definite cor-
relation between separation factor and collection
efficiency; yet, for cyclones of similar design and
use, collection efficiency generally varies directly
as a function of the separation factor.
Stern et al. (1956) discuss the variation of collec-
tion efficiency with inlet velocity. Several theo-
retical formulas are presented in which critical
particle size is shown to vary as 1/V1'2. Critical
particle size is defined as the largest sized par-
ticle not separated from the gas stream, all lar-
ger particles being separated, and critical-sized
and all smaller sizes being lost into the outlet duct.
The critical size varies inversely as the velocity,
and the greater the critical size, the less efficient
Figure 44. Single high-efficiency
cyclone with typical dimension
ratios.
-------
Inertial Separators
93
Figure 45. Double-vortex path
of the gas stream in a cyclone
(Montross, 1953).
is the cyclone collection. The collection efficien-
cy, therefore, varies as the inlet velocity. There
are, however, limits to the inlet velocity; if it is
too great, turbulence develops to such a degree at
the inlet that overall cyclone efficiency is reduced.
The velocity at -which excessive turbulence occurs
is dependent upon configuration of the inlet, de-
sign of the cyclone, and the characteristics of the
carrier gas.
Separation Efficiency
For high efficiency, the separating forces should
be large and the dust removal effective so that
separated dust is not reentrained. In general, cy-
clone efficiency increases with an increase in the
following: (1) Density of the particulate matter,
(2) inlet velocity into the cyclone, (3) cyclone body
length, (4) number of gas revolutions (experiments
indicate that the number of revolutions made bythe
gas stream in a typical simple cyclone ranges from
0. 5 to 3 and averages 1. 5 for cyclones of normal
configuration), (5) ratio of cyclone body diameter
to cyclone outlet diameter, (6) particle diameter,
(7) amount of dust entrained in carrier gas, and
(8) smoothness of inner cyclone wall.
An increase in the following will decrease the over-
all efficiency: (I) Carrier gas viscosity, (2) cy-
clone diameter, (3) gas outlet diameter, (4) gas
inlet duct width, (5) inlet area, and (6) gas density.
A common cause of poor cyclone performance is
leakage of air into the dust outlet. A small air leak
at this point can result in an appreciable decrease
in collection efficiency, particularly with fine dusts.
For continuous withdrawal of collected dust a ro-
tary star valve, a double-lock valve, or a screw
conveyor with a spring-loaded choke should be used.
Collection efficiency is noticeably reduced by the
installation of inlet vanes, probably because of in-
terference -with the normal flow pattern. In gen-
eral, all sorts of guide vanes, straightening vanes,
baffles, and so forth placed inside an other-wise
well-designed cyclone have been found of little val-
ue or actually detrimental. In some instances, for
poorly designed cyclones, these devices have im-
proved performance. Baffles designed to reduce
leakage of air into the dust outlet are sometimes
helpful. These consist of a horizontal, circular
device installed on the cyclone axis near the dust
outlet.
In practice, extensive agglomeration may be ex-
pected for dust concentrations greater than 100
grains per cubic foot and may be present at much
smaller concentrations, depending upon the phys-
ical properties of the particulates being collected.
Fibrous or tacky particles are especially apt to ag-
glomerate. Agglomeration produces a larger ef-
fective particle size and thereby increases the ef-
ficiency of separation. Nevertheless, extremely
sticky, hygroscopic, or similar material that could
possibly plug the dust outlet or accumulate on the
cyclone walls adversely affect cyclone operation.
In addition, the agglomeration effect is reduced
sharply when high inlet velocities are used. In
some cases where agglomeration -was significant,
an increase in cyclone inlet velocity actually re-
duced the collection efficiency. Conversely, the
efficiency was improved by reducing the inlet
velocity,
Pressure Drop
A satisfactory method of determining the pressure
drop of a given cyclone has not yet been developed.
Pressure drop, to be determined accurately, should
be determined experimentally on a geometrically
similar prototype. Lapple (1963) has suggested a
relationship that may be used to approximate the
pressure drop:
F =
KBH
2
D
(43)
where
F =
K =
H
D
cyclone friction loss, number of cyclone
inlet velocity heads, dimensionless
empirical proportionality constant, di-
mensionless
width of rectangular cyclone inlet duct, ft
height of rectangular cyclone inlet, ft
cyclone gas exit duct diameter, ft.
-------
94
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
In this equation, K varies from 7. 5 to 18.4. Pres-
sure drop values, for a value of K = 13. 0, have
been found to check with experimental data within
30 percent.
The Industrial Hygiene Codes Committee (1938)
states that the resistance pressure across pull-
through cyclones is approximately three inlet ve-
locity heads. For pushthrough cyclones venting
directly to the atmosphere, the resistance pres-
sure is approximately one and one-half velocity
heads. These values are valid for simple cyclones,
but a considerable variation may be expected for cy-
clones of unusual design.
OTHER TYPES OF CYCLONE SEPARATORS
High-Efficiency Cyclone Separators
When collection of particulates in the 5- to 10-
micron range is desired, long, small-diameter,
high-efficiency cyclones may sometimes be used.
Operation is, however, more expensive, since
pressure drop increases "with a decrease in cy-
clone diameter; the greater the pressure drop,
the greater the power cost.
High-efficiency cyclones are made more effective
than simple cyclones by increasing the body length
and decreasing the diameter. These two altera-
tions act both to increase retention time in the cy-
clone and exert greater centrifugal force on the par-
ticulates, -which results in greater separation.
Figure 46. Multiple-cyclone separa-
tor (Western Precipitation, Division
of Joy Manufacturing Company, Los
Angeles, Calif.).
Multiple-Cyclone Separators
A multiple-cyclone separator consists of a num-
ber of small-diameter cyclones operating in par-
allel, having a common gas inlet and outlet, as
shown in Figure 46. The flow pattern differs from
that in a conventional cyclone in that the gas,in-
stead of entering at the side to initiate the swirling
action, enters at the top of the collecting tube and
has a swirling action imparted to it by a station-
ary vane positioned in its path. The diameters of
the collecting tubes usually range from 1 foot to as
small as Z inches. Properly designed units can be
constructed that have a collection efficiency as high
as 90 percent for particulates in the 5- to 10-mi-
cron range.
Mechanical, Centrifugal Separators
Several types of collectors are readily available
in which centrifugal force is supplied by a rotat-
ing vane. Figure 47 illustrates this type of
collector, in which the unit serves both as
exhaust fan and dust collector. In operation,
the rotating fan blade exerts a large cen-
trifugal force on the particulates, ejecting
them from the tip of the blades to a skim-
mer bypass leading into a dust hopper.
Efficiencies of mechanical, centrifugal sep-
arators are somewhat higher than those
obtainable with simple cyclones. Mechanical,
centrifugal separators are compact and are
particularly useful where a large number of
individual collectors are required. These
units cannot, however, be generally used to
collect particulates that cake or tend to ac-
cumulate on the rotor blades since these
particulates cause clogging and unbalancing
of ';he impeller blades with resultant high
maintenance costs and shutdowns.
-------
Inertial Separators
95
Figure 47. Mechanical, centrifugal
separator (American Air Filter Com-
pany, Inc., Louisvi !le, Kentucky).
PREDICTING EFFICIENCY OF CYCLONES
Many investigations attempt to correlate cyclone
performance with various parameters. Lapple
(1951, 1963) treats the subject at length in sev-
eral publications, introducing the concept of cut
size (DpC), which is defined as the diameter of
those particles collected with 50 percent efficien-
cy. Collection efficiency for particles larger than
the cut size will be greater than 50 percent while
that for smaller particles will be less. Another
term used is the average particle size (DD), which
is simply the average of the size range. Tor ex-
ample, if the size range is 10 to 15 microns, D =
12.5 microns.
A separation efficiency correlation for typical cy-
clones of the type mentioned by Lapple is presen-
ted in Figure 48. Additional experimental data
have been used to check Lapple's ratios of D^/Dp,,.
All results compared favorably with the original
curve of Lapple. Manufacturers' efficiency curves
for cyclones and multiple cyclones converted to
Dp/Dpc cu-rves had slightly lower efficiencies than
Lapple's correlation for Dp/Dpc ratios greater
than 1. The maximum deviation noted -was 5 per-
cent for the cyclone curve at Dp/Dpc of 1-1/2 and
12 percent for the multiple-cyclone curve at Dp/
DpC of 2 to 3. Apparently, Lapple's correlation
is sufficiently accurate for an engineering estima-
tion of many cyclone applications. A size-efficien-
cy curve may be calculated from this correlation
after the actual size-of the cut size particle is es-
100
10
I/
I
0.3 0.4 0.5
PARTICLE SIZE RATIO, (Dp/DpcJ
Figure 48. Cyclone efficiency versus particle
size ratio (Lapple, 1951).
tablished. Particle cut size may be calculated by
equation 44:
D
pc
9M.b
where
D
pc
M- =
2 N V (p - p ) TT
e i p g
(44)
diameter cut size particle collected at
50 percent efficiency, ft
gas viscosity, Ib mass/sec-ft = centi-
poise x 0.672 x 10~3
b = cyclone inlet width, ft
N = effective number of turns within cyclone.
The number of turns are about five for a
high-efficiency cyclone but may vary from
1/2 to 10 for other cyclones (Freidlander
et al. , 1952)
V. = inlet gas velocity, ft/sec
p = true particle density, Ib/ft
p = gas density, lb/ft~.
Figure 49 presents a graphical solution of this
equation for typical cyclones having an inlet ve-
locity of 50 fps, gas viscosity of 0. 02 centipoise,
effective number of turns equal to five, and cy-
clone inlet width of Dc/4. From these curves,
the cut size may be approximated from the cy-
clone diameter and the dust's true specific grav-
ity. Corrections for viscosity, inlet gas velocity,
effective number of turns, and inlet width different
from those assumed in Figure 49 may be found
graphically by using Figure 50.
The calculated particle cut size may be used in
conjunction -with the general cyclone efficiency
curve of Lapple (1951, 1963) as shown in Fig-
ure 48 to calculate a. particle size efficiency
curve for the cyclone in question. A particle size
distribution of the feed must also be known or cal-
culable to continue the final efficiency determina-
-------
96
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
30
20
10 -
hereji = Viscosity, 0.20 centipoise
b = Inlet width, cyclone dia., D
= Number of turns, 5
V = Inlet velocity, 50 fps
= True specific weight of partlculate
" - • - 3
1 b/ft
P = Gas density, 1b/ft3
pc - Cut size, microns on ordinate
See Figure 50 for correction factors for othe
viscosities, inlet widths, inlet velocities,
- and number of turns.
5 10 1520
CYCLONE DIAMETER (D->, inches
40 50
I 00
Figure 49. Cyclone diameter versus cut size ([apple, 1951).
tion. Size distribution data should be plotted on
logarithmic-probability paper to check for reli-
ability. Drinker and Hatch (1954) state that
Epstein's work shows this plot is a straight line
for operations such as crushing and grinding. An
investigation of test results on samples from crys-
tallization, spray drying, calcining, and other
physical and chemical processes indicates that the
particle size distribution of these processes usu-
ally follows the laws of probability, and plots as
a straight line on logarithmic-probability paper.
The actual distribution used in the calculations
should be taken from the straight-line "smoothed
data." Methods of determining particle sizes have
an effect in determining the straight-line plot.
Most data from screen analyses plot as a curve
on logarithmic-probability paper if the values for
screens smaller than 150-mesh Tyler or 140-
mesh U. S. Screen Scale are used. Specifically,
minus 200 mesh and minus 325 mesh (both some-
times reported in screen analyses) give points
that are usually not in agreement with data ob-
tained when the minus-100-mesh material is sub-
jected to micromeragraph analysis.
A fractional-efficiency curve for a geometrically
similar cyclone may be constructed from a given
fractional-efficiency curve by the following pro-
cedure:
1. Determine Dpc from the fractional-efficiency
curve for a known cyclone.
2. Replot the fractional-efficiency curve as effi-
ciency versus the ratio Dp/DpC.
3. Calculate DpC for the unknown cyclone from
equation 44 or Figures 49 and 50.
4. Assume efficiency versus D_/DpC curve ap-
plies to the unknown cyclone.
Using the value of DpC for the unknown cyclone,
and the efficiency versus Dp/Dpc curve, cal-
culate new values of Dp versus efficiency and
plot as the fractional-efficiency curve of the
unknown, cyclone.
-------
Inertial Separators
97
VISCOSITY (ft) , centipoises
0 02 0 03 0 04 0 06 0 08 0 10
02 03 0405
MEET -MDTH DIANETER (b Dc )
An alternative method of obtaining total weight
of the fraction charged to the cyclone consists of
dividing each weight fraction by the fractional ef-
ficiency. The weight loss is then the difference
between the amount collected and the feed calcu-
lated from the efficiency.
The previously discussed method of predicting cy-
clone collection efficiencies is, of course, only
approximate. It can be useful if applied correctly.
Its utility will be increased once additional test in-
formation is obtained on various cyclones. As
more information is obtained, a family of curves
can be developed for various types of cyclones. The
resulting data should be similar to the data herein,
and the use of the illustrated curves could be ex-
tended to many different cyclone designs without
appreciable error.
INLET VELOCITY (Vc) fps
20 30 40
2 345
EFFECT'VE \LMBER OF TURNS <\.l
Figure 50. Correction factors for Figure 49
(Lapple, 1951).
5. In most cases, a range of Dpc for the unknown
cyclone should be selected instead of a single
value. Then, using the maximum and mini-
mum values for DpC,plot two size efficiency
curves. The overall efficiencies obtainedfrom
these curves serve as an engineering estimate
of the expected cyclone performance.
In some cases, size data are available only on the
materials already collected in a cyclone separator,
with no data on the cyclone loss rate and size dis-
tribution. The calculation procedure is identical
to the normal method except for the final loss rate
step. Here a loss factor must be determined from
the size range efficiency. If the efficiency is 50
percent, the loss factor is 1, and the cyclone loses
1 pound of material for every pound collected in
this size range. If the efficiency is 75 percent, the
loss rate is 1/3, and similarly, if the efficiency is
25 percent, the loss rate is 3. The loss rate for
each particle size range is the quantity collected
multiplied by the loss factor.
Method of Solving a Problem
Knowing the cyclone dimensions, the inlet gas ve-
locity, the viscosity, and the particle size distri-
bution of the dust, predict cyclone efficiencies as
shown in example 18.
Example 18
Given:
be
Cyclone diameter, D
Inlet width,
Inlet velocity,
Specific gravity
Gas viscosity
= 72 in.
= 17 in. = 0.235 D
V = 2, 400 fpm = 40 fps
= 1,5
- 0. 0185 centipoise
Particle size distribution of dust entering the cy-
clone (See curve in Figure 51).
Problem:
Determine the particle cut size DpC and use these
data to determine expected cyclone performance.
Solution:
1. Determine the particle cut size:
From Figure 49, the uncorrected DpC = 10. 5
microns.
The following correction factors are shown by
calculation and can also be obtained from Fig-
ure 50:
Inlet width factor -
).235
). 250
= 0. 97
-------
98
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
CUMULATIVE PERCENT WEIGHT LARGER THAN
99.99 99.9 99.8 99.5 99 98 95 90 80 70 60 50
90
/
/
f
J
J
f
/
/
/
PART 1 CLE
OF THE \
^
/
SIZE
INUS-
y
/
/
DISTRI
00-NES
s
f
r
BUT 10
H FRA
N
CTION
OF THE SAMPLE EIGHTY- FIVE
PERCENT OF THE MATERIAL WAS
GREATER THAN 200 MESH
1.01 0.050.1 0.2 0.5 I 2 5 10 20 30 HO 50
CUMULATIVE PERCENT WEIGHT SMALLER THA\
Figure 51. Particle size distribution
of dust in example problem.
Velocity factor
[W
= Vlb~
= 1. 12
Viscosity factor = \i ' Q2Q = 0.96
Number of turns factor = 1.0 (Number of turns
assumed to be 5)
Corrected cut size = (D )(correction factors)
. 96)(1.0)
= (10. 5)(0.
= 11.0 microns.
2. Calculate collection efficiencies by size incre-
ments:
Select size increments to obtain several values
of Dp less than DpC, and five or more values
between D_/D c ratios of 1 to 10. Calculate the
average size of each increment and tabulate as
Dp. Calculate the ratio Dp/DpC and tabulate for
the range at D . Particles of such size that the
ratio Dp/DpC is greater than 10 are considered
to be collected at 100 percent efficiency. From
Figure 48 obtain the collection efficiencies for
the size increments represented by the Dp/DpC
ratios and tabulate (Table 29).
Table 29. COLLECTION EFFICIENCIES
FOR SIZE INCREMENTS
Particle size, microns
Range
0 to 2
2 to 4
4 to 6
6 to 8
8 to 10
10 to 12
12 to 15
15 to 20
20 to 30
30 to 40
40 to 50
50 to 60
60 to 70
70 to 80
80 to 90
90 to 100
100+
Avg D
1
3
5
7
9
11
13. 5
17.5
25
35
45
55
65
75
85
95
100+
Ratio
D /D
P Pc
0. 09
0.27
0.45
0.64
0.82
1.00
1.23
1.59
2.28
3. 18
4. 1
5.0
5.9
6.8
7.7
8.6
10+
Efficiency
% "by wt
1
7
17
29
40
50
60
72
84
91
95
96
97
98
98.5
99
100
3. Plot the given particle size data of the inlet
dust to the cyclone:
Plot the particle size data on logarithmic-
probability paper and draw the best straight
line, giving maximum consideration to the
data that lie between 20 to 80 percent of the
extreme upper and lower values of the parti-
cle size range (Drinker and Hatch, 1954).
4. Tabulate the weight percentage of the dust cor-
responding to the micron size increments:
Using the smoothed data, as shown above, tab-
ulate the weight percentages corresponding to
the size increments in microns (Table 30).
5. Determine the overall efficiency:
When the particle size distribution is for the
cyclone feed, as given in this example, multi-
ply the percentage for each size increment by
its collection efficiency. The sum of these
products is the overall efficiency. This cal-
culation is presented in Table 31.
6. In some existing installations, it may be dif-
ficult or impossible to determine the particle
size analysis of the dust to the cyclone. In
-------
Wet Collection Devices
99
Table 30. WEIGHT PERCENTAGES
PER SIZE INCREMENTS
Particle size, microns
Range
0 to 2
2 to 4
4 to 6
6 to 8
8 to 10
10 to 12
12 to 15
15 to 20
20 to 30
30 to 40
40 to 50
50 to 60
60 to 70
70 to 80
80 to 90
90 to 100
100+
Avg D
P
1
3
5
7
9
11
14
18
25
35
45
55
65
75
85
95
100+
Wt % for
size
0.01
0.02
0.06
0. 11
0. 15
0.35
0.90
2.80
3.60
4.50
4.50
5.00
4.00
4.00
5.00
65.00
Table 31. CALCULATION OF
OVERALL EFFICIENCY
Particle size, microns
Range
0 to 2
2 to 4
4 to 6
6 to 8
8 to 10
10 to 12
12 to 15
15 to 20
20 to 30
30 to 40
40 to 50
50 to 60
60 to 70
70 to 80
80 to 90
90 to 100
100+
Total
Loss
Overall
efficiency
Loss, % of
feed
Avg D
P
1
3
5
7
9
11
14
18
25
35
45
55
65
75
85
95
100+
Efficiency
% by wt
< 1
7
17
29
40
50
60
72
84
91
95
96
97
98
98. 5
99
100
% by wt x
efficiency
0.001
0.003
0. 017
0. 044
0.075
0.210
0. 648
2. 35
3. 28
4.28
4. 32
4. 85
3.92
3. 94
4. 95
65. 00
97.89
2. 11
97.89
2. 11
the quotients gives the cyclone feed expressed
as percent of the cyclone catch. Divide 100 by
the cyclone catch to obtain the overall efficien-
cy. These calculations are presented in Table
32, with the particle size distribution curve for
this problem, but it is assumed that these data
are the particle size distribution curve for the
cyclone catch.
Table 32. CALCULATION OF OVERALL
EFFICIENCY FOR SPECIAL CASES
Particle size, microns
Range
0 to 2
2 to 4
4 to 6
6 to 8
8 to 10
10 to 12
12 to 15
15 to 20
20 to 30
30 to 40
40 to 50
50 to 60
60 to 70
70 to 80
80 to 90
90 to 100
100+
Total
Loss
Overall
efficiency
Loss, % of
feed
AvgDp
1
3
5
7
9
11
14
18
25
35
45
55
65
75
85
95
100+
% by wt
efficiency
< 1
7
17
29
40
50
60
72
84
91
95
96
97
98
98. 5
99
100
% by wt x
efficiency
0. 14
0. 11
0.20
0.27
0. 30
0. 58
1.25
3. 33
3.96
4. 74
4.69
5. 15
4.08
4. 06
5.05
65. 00
102. 91
2. 91
97. 16
2.84
these cases, a particle size analysis of the per-
centage for each size increment should be di-
vided by its collection efficiency. The sum of
From the preceding problem, the cyclone loss ex-
pressed as percent of feed is obviously 100 minus
the overall efficiency as calculated. If the loss of
any incremental size fraction is desired, this may
be calculated as follows:
1. Calculate the weight of each incremental frac-
tion of feed by using particle size distribution
data and total feed weight.
2. Multiply this weight by the percentage loss (100
minus the efficiency) for each increment to de-
termine the -weight loss.
WET COLLECTION DEVICES
Wet collection devices use a variety of methods to
wet the contaminant particles in order to remove
them from the gas stream. There is also a wide
-------
100
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
range in their cost, their collection efficiency, and
the amount of power they expend.
Wet collectors have the following advantages:
They have a constant pressure drop (at constant
volume), they present no secondary dust prob-
lem in disposing of the collected dust, and they
can handle high-temper ature or moisture-laden
gases. They can also handle corrosive gases or
aerosols, but corrosion-resistant construction
may add materially to their cost. Space require-
ments are reasonably small. Disposal of the waste
water or its clarification for reuse may, however,
be difficult or expensive.
Their collection efficiency varies widely with dif-
ferent designs. Most collectors decline rapidly
in efficiency for particles between 1 and 10 mi-
crons. Many investigators believe that collection
efficiency is directly related to the total power ex-
pended in forcing the gases through the collector
and in generating the water spray.
The process of contacting an air-contaminated gas
with a scrubbing liquid results in dissipation of me-
chanical energy in fluid turbulence and, ultimately,
in heat. The power dissipated is termed the con-
tacting power. Semrau (I960) made an exhaustive
literature survey to correlate scrubber efficiency
with contacting power. He states that contacting
power can be derived from (1) the kinetic energy
or pressure head of the gas stream, (Z) the kinet-
ic energy or pressure head of the liquid, or (3) en-
ergy supplied mechanically by a rotor. He con-
cludes: "Efficiency is found to have little relation
to scrubber design and geometry, but to be depen-
dent on the properties of the aerosol and on the
contacting power. "
THEORY OF COLLECTION
The principal mechanisms by which liquids may
be used to remove aerosols from gas streams
are as follows:
1. Wetting of the particles by contact with a liq-
uid droplet,
2. impingement of wetted or unwetted particles
on collecting surfaces followed by their re-
moval from the surfaces by a flush with a
liquid.
MECHANISMS FOR WETTING THE PARTICLE
The particles can be wetted by the following
mechanisms:
1. Impingement by spray droplets. A spray di-
rected across the path of the dust particles
impinges upon them with an efficiency propor-
tional to the number of droplets and to the
force imparted to them. Johnstone and
Roberts (1949) states that the optimum drop-
let particle size is about 100 microns. Above
100 microns there are too few droplets, and
below 100 microns, the droplets do not have
sufficient force. Fine spray is effective by
another mechanism, diffusion.
2. Diffusion. When liquid droplets are dis-
persed among dust particles, the dust
particles are deposited on the droplets by
Brownian movement or diffusion. This is
the principal! mechanism in the collection
of submicron particles. Diffusion as the
result of fluid turbulence may also be an
appreciable mechanism in the deposition
of dust particles on spray droplets.
3. Condensation (Lapple, 1963). If a gas is
cooled below the dewpoint in passing through
a wet collector, then condensation of mois-
ture occurs, the dust particles acting as
condensation nuclei. This effective increase
in the particle size makes subsequent collec-
tion easier. Condensation is an important
mechanism only for gases that are initially
hot. Condensation alone can remove only
relatively small amounts of dust, since the
amount of condensation required to remove
large concentrations is greater than can be
achieved.
4. Humidification and electrostatic precipita-
tion have been suggested as mechanisms
that facilitate collection of particles by caus-
ing them to agglomerate. These effects are
not, however, •well understood and cannot be
relied upon to play any significant role in the
collection mechanisms.
Several investigators have used wetting agents
for scrubbing water in an effort to improve col-
lection efficiency. In most cases, little or no
improvement has been found (Friedlander et al.,
1952). In order to be wetted, a particle must
either make contact with a spray droplet or im-
pinge upon a wetted surface. When either of
these occurs, the particle is apparently wetted
as adequately without the use of wetting agents
as it is with their use.
Particles that have been wetted must reach a
collection surface if the collecting process is
to be completed. They may be impinged against
surfaces placed in the path of the gas flow; or
centrifugal action may be used to throw them to
the outer walls of the collector; or simple grav-
ity settling may be employed.
In some devices impingement is the principal
collection mechanism, the water sprays being
-------
Wet Collection Devices
101
used merely to remove the dust from the col-
lection surfaces.
Centrifugal action may be provided by a vessel
that is essentially the same as a dry cyclone
separator. Helical vanes in a cylindrical ves-
sel are extensively used to supply centrifugal
action. In some devices, baffles are shaped
and placed so that they act both as impingement
and collection surfaces, and as imparters of
cyclonic motion to the gas stream.
TYPES OF WET COLLECTION DEVICES
Spray Chambers
The simplest type of scrubber is a chamber in
•which spray nozzles are placed. The gas stream
velocity decreases as it enters the chamber, and
the wetted particles settle and are collected at the
bottom of the chamber. The outlet of the chamber
is sometimes equipped with eliminator plates to
help prevent the liquid from being discharged with
the clean air stream. The spray chamber is ex-
tensively used as a gas cooler. Its efficiency as
a dust collector is low except for coarse dust.
Efficiency can be improved by baffle plates upon
-which particles can be impinged. Water rates
range from 3 to 8 gallons per minute (gpm) per
1, 000 cfm. Installed costs range from $0. 25 to
$0. 50 per cfm.
Cyclone-Type Scrubbers
Cyclone-type scrubbers range from simple dry
cyclones with spray nozzles to specially con-
structed multistage devices. All feature a tan-
gential inlet to a cylindrical body, and many fea-
ture additional vanes that accentuate the cyclonic
action and also act as impingement and collection
surfaces.
Figure 52 shows how a dry cyclone can be con-
verted to a scrubber. Some investigators dis-
agree on the most effective placement of spray
nozzles; however, the principal benefit is de-
rived from the wetted walls in preventing reen-
trainment of separated material. Figure 53
shows a standard type of cyclone scrubber. The
gas enters tangentially at the bottom of the scrub-
ber and pursues a spiral path upwards. Liquid
spray is introduced into the rotating gas from an
axially located manifold in the lower part of the
unit. The atomized fine-spray droplets are
caught in the rotating gas stream, and are, by
centrifugal force, swept across to the walls of
the cylinder, colliding with, absorbing, and col-
lecting the dust or fume particles en route. The
scrubbing liquid and particles run down the -walls
and out of the bottom of the unit; the clean gas
Figure 52. Conventional cyclone
converted to a scrubber.
leaves through the top. The scrubber in Figure
54 uses helical baffles to provide prolonged cen-
trifugal action, and multiple spray nozzles to in-
crease spray contact time.
Since centrifugal force is the principal collecting
mechanism, efficiency is promoted by compara-
tively high gas velocities. Pressure drop varies
from 2 to 8 inches water gage, and water rates
vary from 4 to 10 gpm per 1, 000 cfm gas handled.
The purchase cost for completed units varies
from $0. 50 to $1. 50 per cfm gas handled for stan-
dard construction. If corrosion-resistant materi-
als are required, costs may be much higher.
Orifice-Type Scrubbers
Orifice-type scrubbers are devices in -which the
velocity of the air is used to provide liquid contact.
The flow of air through a restricted passage (usually
curved) partially filled -with water causes the disper-
sion of the water. In turn, centrifugal forces, im-
pingement, and turbulence cause wetting of the parti-
cles and their collection. Water quantities in motion
are relatively large, but most of the -water can be
recirculated -without pumps or spray nozzles. Recir-
culation rates are as high as 20 gpm per 1, 000 cfm
gas. The degree of dispersion of the water is, how-
ever, not as great as is attained -with spray nozzles.
Pressure drop and purchase costs are comparable
-------
102
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
CLEANED GAS
CORE BUSTER DISC
TANGENTIAL CAS INLET
HATER «*TER
OUTLET INLET
CONTAMINATED
GAS INLET
Figure 53. Cyclone scrubber (Chemical
Construction Co., New York, N. Y.).
to those for cyclone-type scrubbers. Figure 55 il-
lustrates the action in an orifice-type scrubber.
Zig-zag plates remove spray droplets at the gas exil
Figure 56 illustrates a type in which the orifice is
formed by a cone inside the entrance ducts. Baffle
plates remove spray droplets at the gas exit.
Figure 55. Orifice scrubber (American
Air Fi I ter Co., Inc., LOUISVI I le, Ky. ).
Figure 54. Double-chamber cyclone scrubber
with helical baffles.
Mechanical Scrubbers
Mechanical scrubbers include those devices in which
the water spray is generated by a rotating element
such as a drum or disk. As with the orifice types,
the water is usually recirculated. In the scrubber
in Figure 57, the spray, because it is generated in
a restricted passage, promotes extreme turbulence
and increases chances for collision between dust
particles and spray droplets. Recirculation rates
and degree of dispersion vary widely with the dif-
ferent types of rotating elements. Installed costs
are around $1. 00 per cfm gas for standard con-
struction.
Mechanical, Centrifugal Collector With Water Sprays
A spray of water added to the inlet of a mechanical,
centrifugal collector increases its collection efficien
-------
Wet Collection Devices
103
Figure 56. Orifice scrubber (Western Precipi-
tation, Division of Joy Manufacturing Company,
Los Angeles, Cal i fornia).
cy. The mechanism is mainly one of impingement
of dust particles on the rotating blades. The spray
formed merely keeps the blades wet and flushes away
the collected dust (Figure 58). By the same mechan-
ism, good collection efficiencies can be achieved by
injecting a spray of water into the inlet of an ordi-
nary paddle-type centrifugal fan. This can substan-
tially increase the collection efficiency of a scrub-
bing installation. It also increases, however, the
wear and corrosion rate of the fan. Installed costs
for mechanical, centrifugal types are approximately
$1. 00 per cfm gas.
Figure 57. Mechanical scrubber (Schmieg Indus-
tries, Division of Aero-Flow Dynamics, Inc.,
Detroit, Michigan).
Figure 58. Mechanical, centrifugal
scrubber (American Air Filter Co.,
Inc., Louisvi I le, Ky.).
High-Pressure Sprays
Most scrubbers operate with water pressure of
from 100 to 150 psi. Increasing the pressure at
the spray nozzles has been found to increase col-
lection efficiency by creating more droplets and
giving them more force. A number of scrubbers
are now designed, therefore, to operate with -water
pressures at the spray nozzles of from 300 to 600
psi. Very small nozzle orifices are used, and in
most cases this precludes recirculation of water.
Nozzles must be located so that collision between
water droplets is minimized, and the design must
ensure maximum collision between water droplets
and the dust particles. Very high collection ef-
ficiencies have been reported. Water consumption
ranges from 5 to 10 gallons per 1, 000 cfm. In-
stalled costs are about the same as those for cy-
clone scrubbers. For a given water rate, oper-
ating costs are greater, but collection efficien-
cies are higher.
-------
104
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
Venturi Scrubbers
In the venturi scrubber, the gases are passed
through a venturi tube to •which low-pressure
water is added at the throat. Gas velocities at
the throat are from 15, 000 to 20, 000 fpm, and
pressure drops are from 10 to 30 inches water
gage. Recirculation of water is feasible. The
predominating mechanism is believed to be im-
paction. In spite of the relatively short contact
time, the extreme turbulence in the venturi pro-
motes very intimate contact. The wetted parti-
cles and droplets are collected in a cyclone spray
separator, as shown in Figure 59. Water rates
are about 3 gpm per 1, 000 cfm gas. Very high
collection efficiencies have been reported for
very fine dusts. Costs are from $0. 50 to $2. 00
per cfm for mild steel construction and $1. 00 to
$3.00 per cfm for stainless steel.
Packed Towers
In packed towers the contaminant-laden stream
is passed through a bed of a granular or fibrous
collection material, and a liquid is passed over
the collecting surface to keep it clean and pre-
vent reentrainment of deposited particles. Col-
lection of the contaminant depends upon the length
of contact time of the gas stream on the collecting
surfaces. This collecting surface material should
have a relatively large surface area, low "weight
per unit volume, and large free cross-section. Ir-
regularly shaped ceramic saddles are commonly
used as packing. Coke, broken rock, stoneware
shapes, Raschig rings, and spiral-shaped rings
are other materials and shapes often used. Bed
depths may vary from a fraction of an inch to
several feet depending upon the type of packing
and the application. Coarsely packed beds are
used for removing coarse dusts and mists that
are 10 microns or larger; velocities through the
bed should be about 400 fpm. Finely packed beds
may be used for removing contaminants in the 1-
to 5-micron range, but the velocity through the bed
must be kept very low, preferably below 50 fpm.
Finely packed beds tend to clog; their applications
are generally limited to dust-laden gases with rel-
atively low grain loadings or to liquid entrainment
collection.
Both costs and collection efficiency vary widely
with bed depths, design velocities, and types of
packing. For comparatively shallow beds, high
velocity, and coarse packing, the costs and col-
lection efficiency are comparable to those for a
simple spray chamber. For deep beds, fine pack-
ing, and low velocities, both the costs and collec-
tion efficiencies are about the same as those for
an electrical precipitator. Figure 60 illustrates
Figure 59. Venturi scrubber (Chemical Construction Co., New York, N. Y.).
-------
Wet Collection Devices
105
one type of thin-bed tower. The packing in this de-
vice consists of lightweight glass spheres kept in
motion by the air velocity.
Figure 60. Thin-bed packed tower
(National Dust Collector Corpora-
tion, Chicago, 111.).
Wei Fillers
A wet filter consists of a spray chamber with filter
pads composed of glass fibers, knitted wire mesh,
or other fibrous materials. The dust is collected
on the filter pads. The sprays are directed against
the pads to keep the dust washed off, as shown in
Figure 61. The pads are about 20 inches square
and 3 to 8 inches thick. The pads commonly used
contain coarse fibers and are not very efficient
for collecting fine dust. Fine glass wool fibers
are efficient, but their usefulness is limited be-
cause the pads mat and sag from their supports
when wetted.
Many wet collectors are a combination of the pre-
ceding types. One design consists of a spray cham-
ber followed by impingement screens, which are fol-
lowed by a centrifugal section, as in Figure 62. Sev-
eral other combinations are used. The device shown
in Figure 63 combines centrifugal and impingement
actions. In many devices, the wetting action and col-
lecting action take place in the same zone. Perfor-
mance data on a number of different kinds of wet col-
lectors are shown in Table 33.
THE ROLE OF WET COLLECTION DEVICES
The collection efficiency of wet collection devices
is proportional to the energy input to the device.
Since high-energy devices are expensive to install
1V2 inches
6 feet 8 inches
AIR FLO*
4 feet 2 inches OMITTING ACCESS DOOR
Figure 61. Wet filter (Buffalo Forge Company, Buffalo, N Y.).
-------
106
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
ys*
Figure 62. Multiple-action scrubber (Joy
Manufacturing Company, Pittsburgh, Pa.).
AIR
OUTLET
ENTRAPMENT'
STAGE
HATER
INLET
HASH 2
HASH I
Figure 63. Centrifugal and impingement
scrubber (Claude B. Schneible Company,
Detroit, Michigan).
and operate, there is a natural tendency to install
wet collectors of limited efficiency. In Los Angeles
County, in the early days of the Air Pollution Con-
trol District, many -wet collectors -were found to be
inadequate for meeting the emission standards. For
instance, many low-energy scrubbers -were installec
to collect the dust from asphaltic concrete-batch-
ing plants, but they -were not efficient enough to low-
er the emissions to the required level. A moderate
ly high-energy scrubber -was required to meet the
emission limits, but even these were inadequate to
reduce the emissions of grey iron cupolas to the re-
quired level. The high operating cost of high-ener-
gy scrubbers usually makes the total cost at least
as much as that of a high-temperature baghouse or
an electrical precipitator.
For collection of dusts and fumes, the baghouse is
to be preferred over a scrubber. The positive col-
lection mechanism of the baghouse ensures virtually
complete collection of almost any dust or fume,
whereas only the best scrubbers ensure good collec-
tion of very fine dusts and fumes. If, however,
mists or hygroscopic particles are present in the
effluent, a baghouse cannot be used. In many cases.
a scrubber is the only choice. Mists that form free>
running liquids when collected can be successfully
collected in an electrical precipitator. If, however,
sticky or gummy materials are formed, removing
the collected material is very difficult, and an elec-
trical precipitator is then impractical.
BAGHOUSES
Suspended dust and fumes may be removed from
an air stream by a number of different devices.
When high collection efficiency on small particle
size is required, however, the most widely used
method consists of separating the dust from the
air by means of a fabric filter. The fabric is usu-
ally made into bags of tubular or envelope shape.
The entire structure housing the bags is called a
baghouse. Typical baghouses are illustrated in
Figures 64, 65, 66, and 67.
FILTRATION PROCESS
Mechanisms
Filter fabrics normally used to remove dust and
fumes from airstreams are usually woven with
relatively large open spaces, sometimes 100 mi-
crons or larger in size (Environmental Sciences
and Engineering ; Drinker and Hatch, 1954;
Spaite et al. , 1 961 ; and Stairmand, 1956). Since
collection efficiencies for dust particles of 1 mi-
cron or less may exceed 90 percent (Environmental
Sciences and Engineering ), the filtering pro-
cess obviously cannot be simple sieving. Small
particles are initially captured and retained on
the fibers of the cloth by means of interception,
-------
Baghouses
107
Table 33. SCRUBBERS AND OTHER WET COLLECTORS
(Friedlander etal., 1952)
Device
Wet cell washer
(a 9-element washer
consisting of 3 units
' a
wet filter cells fol-
mator pad)
(31)
Wet cell washer
(an 8-element washer.
Stage 1 has one coun-
by one concurrent wet
cell. Stages 2 and 3
each have one coun-
ter-current and one
followed by a dry
eliminator pad)
(31)
Due on No. 5
ff.(\\
\OUJ
(63)
Cyclone scrubber
(62)
Centri-merge
(62)
Multi-wash
(62)
Manufacturer
Buffalo Forge Co.
Buffalo Forge Co.
Ducon Co.
ment Co.
Pease Anthony Equip-
ment Co.
Schmieg Industries,
Inc.
Claude B. Schneible
Test aerosol
No^rnxal air
Dust composed
of spheres of
copper sulfate
Dust composed
of UO} spheres
Dust from stone
and sand-drying
kiln
sulfunc acid
plant
SiO^ from silicon
ore furnace
from open hearth
furnace (oxygen
lanced)
Lime dust from
lime kiln
Iron ore and coke
dust from blast
furnace
Na2C03 fume
Foundry dust
Bilet
concentration
0.2 to 0. 5
ram /1,000ft3
1 to 2
>rains/l,000ft3
0.01 to 0.06
gram /1,000ft3
5. 8 grams/ft3
2 92 grams/ft3
3.98 grams/ft3
1 to 5. 99
grains /ft3
9.2 grams/ft3
3.0 to 24.0
grains/ft3
30 x 104
particles/ft3
>9. 7 x 104
particles/ft3
Particle size
at inlet, \t.
Most parti-
cles between
0. 3 and 0. 5
mlcr°" tes 1
2. 5 (mass
median)
0. 8 (mass
median)
1. 5
0.01 to 0. 35
0. 02 to 0. 5
2. 0 to 40. 0
0. 5 to 20.0
< i 5
Efficiency,
%
57
65
67
90
92
94
Cumulative
efficiency
stages
1 2 3
20 75 SO
13 74 80
74
(weight)
'
99. 7
86. 7
92 to 99
(weight)
99
99
(weight)
96.2
(count)
88. 9
(( ount)
Resistance
H20
0.19 to 0. 28
per wet cell)
0. 38 to 0. 56
0. 21 to 0. 23
per wot cell)
Same as A
Same as B
Same as C
0. 14 to 0. 17
(concurrent)
0. 2.1 to 0. 23
(counter-
current)
0.21 to 0. 23
(concurrent)
0. 32 to 0. 36
(counter
current) per
wet cell
9 7
29. 3
11.0
14 0
2 to 4
(rated)
2 to 4
(rated
5. 5
4
Velocity
(face of
*>et cell),
fpm
216
216
216
216
216
216
216
216
.
19,200
15,060
2,000
3, 300
Water rate
'per wet cell),
gal/1, 000 ft3
3. 3 to 5
5. 3 to 5
3. 3 to 5
3. 3 to 5
3. 3 to 5
3. 1 to 5
15. 3
15. 3
1
(ratod)
4
3.9
4.7
5 to 10
(rated)
5 to 10
(rated)
40
Remarks
Wet cells:
A. 3-in. cells of
150-ti fibers
7 9 lb/ft3
B. 6-m, cells of
1 50— u. fibers
random packed
7. 9 lb/ft3
C. 8-m. cells of
255-n fibers
partially ori-
ented, packed
4.9 lb/ft3
Same as A
Same as B
Same as C
Dry pads were 2 in.
thick and i om-
posed of J 0-n
fibers packed
0. 4 lb/ft3
thick and com-
posed of 255-|j.
fibers partially
oriented with
densities of
3. 3 lb/ft 3
4. 9 lb/ft3
Dry pads were 2 in.
thick and com-
posed of 10-n
fibers packed
0.46 lb/ft3
Flow rate was
9, 300 cfm.
Temperature
was 230"F in
and 120"F out.
effluent from
Sirocco No. 20
cyclones.
tion was 0. 0092
gram/ft3
Outlet concentra-
tion was 0. 01 1
gram/ft3
tion was 0. 0092
to 0. 069 gram/
ft3
Outlet concentra-
tion was 0, 08
grain/ft3
Outlet concentra-
tion was 0. 03
to 0. 08 grain/
ft3
aConcurrent means that water is sprayed in the direction of the air flow.
bCountercurrent means that water is sprayed against the air flow.
-------
108
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
SHAKER
MECHANISM
OUTLET
PIPE
INLET
PIPE
HOPPER
Figure 64. Typical simple baghouse with
mechanical shaking (Wheelabrator Corpora-
tion, Mishawaka, Indiana).
impingement, diffusion, gravitational settling,
and electrostatic attraction. Once a mat or cake
of dust is accumulated, further collection is ac-
complished by sieving as well as by the previous-
ly mentioned mechanisms. The cloth then serves
mainly as a supporting structure for the dust mat
responsible for the high collection efficiency. Per-
iodically the accumulated dust is removed for dis-
posal. Some residual dust remains and serves as
an aid to further filtering.
Direct interception
Under conditions normally existing in air filtra-
tion the flow is almost always laminar (Drinker
and Hatch, 1954; Hemeon, 1955, Rodebush, 1950;
and Underwood, 1962). For conditions of laminar
flow, a small inertialess particle will remain on
a single streamline. If the streamline passes
close to an obstruction, such as a fiber of the fil-
ter fabric, and within a distance equal to the radi-
us of the particle, the particle will contact the ob-
struction and will adhere because of the van der
Waals forces. While no real particle is complete-
ly inertialess, small particles of 1 micron or less
may be considered, without serious error, inertia-
less (Rodebush, 1950).
The shape of the streamlines is not affected by the
air stream velocity in laminar flow, so that collec-
tion by direct interception is independent of veloc-
SHAKER-
AIR REVERSAL
VALVE
CLEAN AIR MANIFOLD
WALKWAY
CLEAN AIR
TO FAN
SCREW
CONVEYOR
Figure 65. Fully automatic, compartmented baghouse with hopper discharge screw conveyor (Northern
Blower division. Buell Engineering Company, Inc., Cleveland, Ohio).
-------
Baghouses
109
CLEANING '/ ° AIR
FROM ATMOSPHERE
Figure 66. Envelope-type baghouse with
automatic reverse-air cleaning ( W. W.
Sly Manufacturing Company, Cleveland,
Ohio).
ity. The size of the obstruction is important since
streamlines pass closer to small obstructions than
they do to larger ones (Rodebush, 1950). Large
particles are also collected more easily since the
streamline need not pass as close, in the case of
a larger particle, for the particle to contact the
collecting surface. As the particle size increases,
however, inertial forces rapidly increase and pre-
dominate (Ranz, 1951).
Impingement
When a particle has an appreciable inertia, it will
not follow a streamline when the streamline is de-
flected from a straight path as it approaches an
obstruction. Whether or not the particle contacts
the surface of the obstruction depends upon the
size of the obstruction and the size and inertia of
the particle. As in the case of direct interception,
smaller obstructions are more effective collectors
for the mechanism of impingement or impaction
and for the same reason. Other factors being equal
a particle -with greater inertia is more likely to
strike a collecting surface.
The inertia of a particle may be measured by its
so-called stopping distance. This is the distance
that the particle would travel before coming to
rest if the streamline were to turn abruptly at
90 degrees.
CLEAN
AIR
TO FAN
Impaction is not a significant factor in
collecting particles of 1 micron di-
ameter or less. It is generally con-
sidered significant for collecting parti-
cles of 2-microns diameter or larger
(Rose et al. , 1958) and becomes the
predominant factor as particle size
increases (Rodebush, 1950).
For effective collection of particles
by inertial forces, the direction of
the aerosol stream must change
abruptly within a distance from the
collector or obstacle approximate-
ly equal to or less than the stopping
distance (Ranz, 1951). Effectively,
this requires a collector with a di-
mension perpendicular to the aero-
sol stream of the same magnitude
as the stopping distance (Ranz, 1951).
Theoretical considerations indicate
that the collection efficiency for a
given size particle decreases as the
collector size increases. Observa-
tions have shown that large fibers
do not collect small particles well.
In fact, for a given size fiber and
Figure 67. Reverse-jet baghouse (Western Pre-
cipitation Corporation, Los Angelas, Calif.).
-------
110
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
airstream velocity, there is a minimum par-
ticle size below -which virtually no collection
by inertial forces occurs (Ranz, 1951). On
the other hand, as fibers are made smaller,
collection continues to improve down to the
practical limits of fiber size (Rodebush, 1950).
The velocity of the airstream is important in im-
paction. Collection efficiency increases with in-
creasing velocity since the stopping distance also
increases with velocity. The underlying assump-
tion is that the particle velocity is the same as
that of the airstream, which is approximately true.
If the velocity becomes excessive, however, the
drag forces increase rapidly and may exceed the
adhesive forces so that collected particles are
blown off and collection efficiency decreases.
The fibers of filter fabrics are in general rela-
tively large compared with the size of the parti-
cles to be collected. Fibers of cotton and wool,
for example, are about 10 to 20 microns in di-
ameter (Rodebush, 1950). Fibers such as these
are too large to be effective collectors for parti-
cles a few microns or less in diameter. Collec-
tion efficiency for fine dusts and fumes can, there-
fore, be expected to be poor until a dust mat is
built up on the filter fabric. This has been veri-
fied by many field observations. For a short
time after new bags are installed or immediately
after the bags have been thoroughly cleaned,
visible emissions bleed through the fabric. In
most cases, bleeding ceases in a few seconds or
several minutes at the most (Rose et al. , 1958).
In some cases where bleeding has been a problem
after each cleaning cycle, reducing the cleaning
effectiveness has been found helpful.
Filter fabrics are sometimes woven from a mix-
ture of asbestos and wool fibers to take advantage
of the smaller diameter of the asbestos fibers and
to improve collection efficiency on fine dusts and
metallurgical fumes (Rodebush, 1950). Another
method reported successful is the use of a rela-
tively coarse dust as a precoat on the filter, which
then becomes highly efficient on very fine dusts
and fumes (Drinker and Hatch, 1954).
Diffusion
When particles are very small, of a dimension
about equal to the intermolecular distance, or
less than about 0. 1 to 0. 2 micron in diameter,
diffusion becomes the predominant mechanism of
deposition. Particles as small as these no longer
follow the streamlines because collisions with gas
molecules occur, resulting in a random Brownian
motion that increases the chance of contact be-
tween the particles and the collecting surfaces.
Once a few particles are collected, a concentra-
tion gradient is established that acts as a driving
force to increase the rate of deposition (Drinker
and Hatch, 1954). Lower air velocity increases
efficiency by increasing the time available and
hence the chance of contacting a collecting sur-
face. Smaller collectors or obstructions also
increase collection efficiency (Ranz, 1951).
Electrostatics
While electrostatics undoubtedly plays a role in
the capture and retention of dust particles by a
fabric filter, the evidence is inadequate to eval-
uate this mechanism quantitatively. Accord-
ing to Frederick (1961), electrostatics not only
may assist filtration by providing an attractive
force between the dust and fabric, but also may
affect particle agglomeration, fabric cleanability,
and collection efficiency. He attributes the gen-
eration of charge to frictional effects, stating
that the polarity, charge intensity, and charge
dissipation rate of both the dust and filter media,
and their relation to each other can enhance or
hinder the filtering process. He cites qualita-
tive differences only. For example, fabric A
may be better than fabric B on dust X, while
fabric B is better than A on dust Y. He gives
a "triboelectric" series for a number of filter
fabrics that may be useful as a guide to selecting
fabrics with desirable electrostatic properties.
This is a fertile field for further investigations.
Until more information is available, the relative
importance of electrostatics in determining the
best filter fabric for a particular installation
cannot be evaluated. Certainly, however, if one
fabric does not -work effectively, other fabrics
should be tried regardless of whether the dif-
ficulty is caused by the electrostatic properties
or the physical characteristics.
Baghouse Resistance
Clean cloth resistance
The resistance to airflow offered by clean filter
cloth is determined by the fibers of the cloth and
the manner in which they are woven together. Ob-
viously a tight weave offers more resistance than
a loose -weave at the same airflow rate. Since the
airflow is laminar, resistance -will vary directly
•with airflow. One of the characteristics of filter
fabrics frequently specified is the Frazier or
ASTM permeability, which is defined as the air
volume, in cfm, that -will pass through a square
foot of clean new cloth -with a pressure differen-
tial of 0. 50 inch WC. The usual range of values
varies from about 10 to 110 cfm per square foot.
The average airflow rate in use for industrial
filtration is about 3 cfm per square foot, and the
resistance of the clean cloth does not usually ex-
ceed about 0. 10 inch WC; often it is much less.
-------
Baghouses
111
Resistance of dust mat
Drinker and Hatch (1954), Hemeon (1955), Mum-
ford etal. (1940), Silverman (1950), Williams et
al. (1940), and others attempt to correlate the in-
crease in resistance of the dust mat or the com-
bination of dust mat and filter fabric with the
filtration velocity or filter ratio, gas viscosity
and density, dust concentration or absolute dust
load, elapsed time, and dust characteristics such
as particle size, true specific gravity, a particle
shape or specific surface factor, and a factor for
the percent of voids or the degree of packing.
The equations may approach the problem from the
theoretical point of view, building up relations
from basic considerations, or they may be com-
pletely practical, ignoring entirely the mechan-
isms involved and relating only the variables that
may be measured most easily. Regardless of
the approach, in the final analysis a measurement
must be made experimentally to determine a pro-
portionality constant or a "resistance factor" for
the particular dust under consideration. One meth-
od (Environmental Sciences and Engineering, )
of relating the variables follows:
Up)
(1 - £) v
mat
(45)
-where
(Ap)
(e)
mat
k
M-
d
e
= pressure drop through 1 square foot of
filtering area (force per unit area),
lb/ft2
= a constant, dimensionless
= gas viscosity, Ib sec/ft
= thickness of the mat of dust particles, ft
= fraction of voids in the mat of particles,
dimensionless
v = face velocity of the gas through the fab-
ric, ft/sec
V
= ratio of particle volume to particle sur-
P
face, ft3/ft2.
K
By substituting k =
•where
Ml - E) C
(46)
p = mass density of the particles, slugs
p = mass density of the gas, slugs
g = acceleration of gravity, ft/sec
C = dimensional constant.
and
d =
G v t
P g (1 -
(47)
where
G = concentration of dust in the gas streams,
lb/ft3
t = elapsed time, sec,
it is possible to solve for (Apt)mat> the pressure
loss through the mat of dust at the end of time
period t.
t
t mat
C
(48)
Values of K, the resistance coefficient, must be
determined experimentally. In practice it is com-
mon to express the pressure drop in inches of -water,
the dust concentration in grains per cubic foot, the
face velocity in feet per minute, and the time in
minutes. The dimensional constant C is adjusted as
required for the actual units used.
The K values are usually determined by using a
scale model unit either in a laboratory or in the
field, though care must be exercised in applying
these results to a full-scale unit (Stephan and
Walsh, I960). If a vertical bag is used, elutria-
tion of particles may occur, and the true value of
K may vary with time and position on the bag (En-
vironmental Sciences and Engineering ). The
measured value of K is an average value that may
not be the same when the scale or configuration is
changed. This is borne out by failure of some full-
scale units to function as anticipated from pilot
studies.
Williams etal. (1940) determined K values for a
number of dusts, as shown in Table 34. These
data were obtained by laboratory experiments by
using an airflow of 2 cubic feet per minute through
0. 2 square foot of cloth area or a filtering velocity
of 10 feet per minute. The tests were terminated
at 8 inches of -water column, maximum pressure
differential. Resistance coefficients were calcu-
lated from the relationship
K,
7, 000 (hf - h )
G t v
(49)
-------
112
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
•where
Kl
= resistance coefficient, in. WC; per Ib
of dust; per cloth area, ft^; per filter-
ing velocity, fpm
= final pressure drop across collected
dust and filter cloth, in. WC
= initial pressure drop across clean
cloth, in. WC
G = dust loading, grains/ft
t = elapsed time, min
v = filtering velocity, fpm.
The pressure loss through the collected dust mat
was found to increase uniformly with time, indi-
cating a linear relationship between resistance
and the thickness of the accumulated dust mat. The
data clearly show a trend of increasing resistance
with decreasing particle size. The test dust for
the data on particles 90 microns or less in diame-
ter was obtained by elutriation. In a full-scale bag-
house, particularly if relatively long vertical bags
are used, a substantial amount of elutriation can
be expected (Stephan etal., I960). The dust-laden
gas usually enters the filter bag at the bottom and
travels upward. As the gas filters through the cloth,
the upward velocity decreases so that only very
fine dust remains airborne to be deposited on the
upper portion of the bag. Since the actual pres-
sure loss through the bag must be the same through
all areas, the volume and filtering velocity through
some portions of the bag increase to excessively
high values. Stephan and Walsh (I960) found that
local filtering velocities vary by a factor of 4 or
more over a single filter bag. This, in turn, mai
lead to collapse or puncture of the filter cake
(Stephan et al. ,1960). Punctures are small holes
in the dust mat. They are usually self-repairing
because the increased airflow through the small
area of low resistance brings more dust with it.
Collapse of the filter cake, on the other hand,
is a shift in cake structure to a more compacted
condition -with a greater resistance. The collapse
may progress in several steps.
Both collapse and puncture of the filter cake are
phenomena caused by excessive filtering veloci-
ties. Some dust may be surmised to be embed-
ded in the interstices of the cloth when puncture
or collapse occur, so that normal cleaning will
not completely remove it. This may lead to
"blinding, " which is a plugging of the fabric
pores to such an extent that the resistance be-
comes excessively high permanently. Once it
starts, blinding tends to become worse rapid-
ly. For example, Stephan (I960) found tran-
sient local filtering velocities of about 100 fpm
through areas of puncture when the average
filtering velocity was only 0. 75 fpm. Further
evidence Ls cited by Lemke et al. (I960) who
note that, for fumes from galvanizing opera-
tions, filtering velocities must be kept below
approximately 2 feet per minute to avoid blind-
ing and cleaning difficulties. When higher fil-
tering velocities were employed, the residual
pressure loss after cleaning increased con-
tinuously from one cleaning cycle to the next
Table 34. FILTER RESISTANCE COEFFICIENTS, KI( FOR CERTAIN
INDUSTRIAL DUSTS ON CLOTH-TYPE AIR FILTERS (Williams et al. ,1940)
Dust
in. WC per Ib of dust per ft^ per minute of filtering velocity-
for particle size less than
Granite
Foundry
Gypsum
Feldspar
Stone
Lampblack
Zinc oxide
Wood
Resin (cold)
Oats
Corn
20 mesha
1.58
0.62
0.96
1.58
0.62
140 mesha
2. 20
1.58
0.62
375 mesha
3.78
6. 30
6.30
1.58
90 fib
6. 30
6. 30
9.60
3.78
45 yf>
11
8.80
20 \i.b
19.80
IS. 90
27. 30
25. 20
Z^
47.20
15.70C
aCoarse.
than 90 (j. or 45 |JL, medium; less than 20 |JL or 2 IJL, fine; theoretical size of
silica, no correction made for materials having other densities.
Flocculated material, not dispersed; size actually larger.
-------
Baghouses
113
until the volume was adversely affected. The
fume in this case -was largely ammonium
chloride. With lower filtering velocities the
equipment functioned well.
Hemeon (1955) takes a more practical approach
to the evaluation of pressure loss in cloth fil-
tration. He notes that the resistance of clean
new cloth can never again be attained once the
cloth has been used. He takes, therefore, the
resistance of the cloth-residual cake combina-
tion as the basic cloth resistance.
R = K V
o of
(50)
where
R = the basic cloth resistance, in. WC
o
K = resistance factor, in. WC/fpm
V = the filtering velocity, fpm.
The magnitude of the factor Ko depends upon
the nature and quantity of dust that remains
lodged in the interstices of the cloth. Thus,
it depends upon the effectiveness of the clean-
ing action as well as upon the dust and cloth
characteristics. Values of Ko are listed in
Table 35. The removable dust mat contrib-
utes a varying resistance according to the
relationship
Hemeon assumes that the basic dust resistance
depends only upon the physical properties of the
dust. Lunde and Lapple (1957) claim, however,
that the resistance coefficient of the dust cake
also depends upon the fabric. Too literal an ap-
plication of these data and equations should not be
attempted; rather they should be used as a guide
to be modified according to experience and the
particular situation.
Pring (1952) uses equation 49 to determine a
number of resistance factors, K2 as shown in
Table 37, which also lists typical filtering veloc-
ities for several dusts.
Mumford etal. (1940) investigated the resistance
of cotton filter cloth for coal dust. They per-
formed a series of bench-scale, laboratory-
type experiments using a minus-200-mesh coal
dust. The results confirmed a linear relation-
ship between the resistance and airflow rate
when the dust loading was held constant. As
shown in Figure 68, however, they report that
the resistance varies with the 1. 5 power of the
dust loading when the airflow rate is held con-
stant. Williams et al. (1940) and some other
investigators report that the resistance varies
linearly with dust loading.
Campbell and Fullerton (1962) also report a non-
linear relationship between resistance and filter-
ing velocity, as shown in Figures 69 and 70. These
Q W
R , = K Vr W =
d d f
(51)
where
R, = the basic dust resistance, in. WC
K = the resistance coefficient, in. WC/fpm/
oz of dust/ft2
V = the filtering velocity, fpm
W = dust loading, oz/ft2
Q = the air flow rate, cfm
A = the total cloth area, ft .
Values of the coefficient K^ are given in Table 36
for several different dusts and dust loadings. The
total pressure drop through the filter cloth may
then be calculated as the sum of the basic cloth
resistance and the basic dust resistance.
5
_^
'
0 02
^^
/
/
/ y
/
/
^^/
/
/
/
/
f
J
f
/
/
/
/
/
/
/
f
/
/
/
/
/
A
/
/
/
f
/
/
/
/ /
/ /
/
/
f
Fabric Cotton sateen, 96 x 64 thread
count 1 05 yd/lb, treated
with mixture of ammonium phos-
phate and boric acid solution
for flame proofing, clean
cloth permeabi 1 i ty 11 5 cfm @
0 50 inch water
Dus • Minus-200-mesh coal dust from
ba 1 1 mill
0 04 0 06 0 08 0 10 0 20
COAL DOST LOADING, In/ft2 FABRIC
R = KV +KVW
of d f
(52)
Figure 68. Pressure drop through cotton sateen
cloth versus coal dust loading for different
filtering velocities in a test unit (Mumford et
al., 1940).
-------
114
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
Table 35. VALUES OF THE BASIC CLOTH RESISTANCE
FACTOR, K0, OBSERVED IN SPECIFIED APPLICATIONS
(Hemeon, 1955)
Type
of
dust
Cloth area,
fl?
K
o
Flat bag collectors
Stone -crushing operations
Stone-crushing operations
Stone-crushing operations
Stone-crushing operations
Stone-crushing operations
Synthetic abrasive crushing operations
Clay crushing in dry pan
250
250
500
2,250
9,000
_.
500
0.83
0.49a
0.83
0.78
0.75
0. 74
0.79
1.01
0.80
1.60
Cloth tube collectors
Stone-crushing operations
Stone-crushing operations
Stone- crushing operations
Stone chiseling
Electric -welding fume
Iron cupola fumes
Foundry dust core knockout
Shot bla.st room ventilation
Same --pneumatic lift
Clay crushing in dry pan
2, 150
4,300
1, 500
400 to 1, 000
10
5, 200
2, 350
950
500
0. 47
0. 45
0. 60
0. 45
0. 37
0.40
0. 17 to 0.27
0. 70
2. 50
0. 28
0. 25
0. 58
0. 63
0. 39
0. 39
0. 34,0. 36,0.59
0. 60
aSame as first operation but after installation of pneumatic vibrator.
data were obtained by venting a portion of the ef-
fluent from a direct-arc steel-melting furnace to
a pilot model baghouse with glass fabric filtering
elements. No effort was made to control or cor-
relate the dust loading with the pressure loss.
Caplan (1954) states that pressure loss is linear-
ly related to gas flow if, and only if, the absolute
amount of dust remains constant when the gas
flow rate is varied. In practice this does some-
times happen. The amount of dust generated in
these cases is independent of the ventilation rate.
Increasing the volume vented above that required
to ensure 100 percent capture of emissions does
not, therefore, increase the total amount of dust
carried to the baghouse. In many cases, how-
ever, increasing the ventilation rate increases
the absolute amount of particulates, though the
increase in emissions may be less proportionate-
ly than the increase in gas rate. Thus, the effect
on the resistance of varying the filtering veloc-
ity depends upon factors that may easily be over-
looked or may be difficult to ascertain.
If the grain loading (dust concentration in the gas
stream in grains /ft 3 as distinguished from abso-
lute dust loading in Ib/min ) remains constant,
resistance is generally considered to vary as the
filtering velocity squared (Environmental Sciences
and Engineering ). The derivation of this
relationship is from the linear variation of resis-
tance with changes in volume when the absolute
dust load remains constant combined with the
linear variation of resistance "with changes in
absolute dust loading when the volume remains
constant. The latter condition may be restated
as a linear variation of resistance with changes
in grain loading when the volume remains con-
stant. If, however, the grain loading remains
-------
Baghouses
115
Table 36. VALUES OF BASIC DUST
RESISTANCE FACTOR (Kd) OBSERVED
ON SOME INDUSTRIAL INSTALLATIONS
(Hemeon, 1955)
Type of dust
Stone crushing (plant A)
Stone crushing (plant B)
Stone crushing (plant C)
Foundry, castings clean
Shot blasting
Pneumatic shot lift
Core knockout
Sandblasting (scale)
Cloth dust loading
(W),oz/ft2
5
12
14
17
22
25
as
7
8
8
1
0. 2
0. 3
1. 3
0. 2
2.4
0. 2
0. 1
7
Kd
0. 18
0. 12
0. 08
0. 12
0. 11
0. 02
0. 07
0. 16
0. 10
0.08
0.82
0.82
0.25
0.25
0.66
0.40
0. 55
0.68
0.20
Table 37. TYPICAL RESISTANCE FACTORS (K2)
AND COMMON FILTERING VELOCITIES (V)
FOR SELECTED UNSIZED DUSTS (Pring, 1952)
Dust
Nut shell dust
Asbestos
Titanium dioxide
White lead
Copper powder
Tobacco
Carbon black
Bismuth and cadmium
Insulating brick
Calcimine
Cement
Clay
Flour
Glass sand
Milk powder
Mixed pigments
Soap
Wood flour
Cloth
Cotton sateen
Napped orlon
Cotton sateen
E-21 wool
C-l 1 nylon
E-21 wool
Napped vinyon
Cotton sateen
Napped orlon
Cotton sateen
Resistance,
factor (K2)
0.2
2.18
94 to 206
34. 6 to 70
47 to 104
32.2
5. 1 to 10.6
36
22. 4 to 28.2
2.7
Velocity
(V),fpm
1. 5
6 to 8
4
2.6
1. 5 to 2. 7
2.9
2.7 to 3.0
1.2
4.5
2.3 to 2.9
1.6 to 3. 1
2. 8 to 4. 8
constant and the volume is increased, then
the absolute dust load must increase. The
result is that, for a constant grain loading, re-
sistance varies as the square of the volume or
filtering velocity (Brief et al., 1956).
Silverman (1950) states that, not-withstanding
the theoretical equations, an exponential re-
lationship exists in practice and that this has
been verified by Bloomfield and DallaValle.
1234
FILTER RATIO, cfm gas/ft2 fabric
Figure 69. Pressure drop versus filter ratio
for fabrics on 60-minute cleaning cycle (Camp-
bell and Fullerton, 1962). Note that A and C
are siliconized glass fabrics, B is a sili-
conized Dacron fabric.
The observation by Stephanet al., (I960) that
filter resistance coefficients actually vary
with time also supports an exponential rela-
tionship since the coefficients are based up-
on an assumed linearity.
Effect of resistance on design
In an actual installation the resistance of the
cloth filter and dust cake cannot be divorced
from the total exhaust system. The operating
characteristics of the exhaust blower and the
duct resistance will determine the way in-
creases in baghouse resistance affect the gas
rate. If the blower characteristic curve is
steep, the gas flow rate may be reduced only
slightly when the resistance of the filter bags
changes markedly. This occurs because, as
the volume decreases slightly, the pressure
delivered by the blower increases proportion-
ately more, while the duct resistance de-
creases, partially offsetting the increase in
resistance of the filter cloth. Some varia-
tion in resistance and air volume must normal-
ly occur, however, in all baghouse installa-
tions, even in the Hersey type to be discussed
later. Proper design requires the volume to
-------
116
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
90 mi n
30 m i n
12345
FILTER RATIO, cfm gas/ft2 fabric
Figure 70. Pressure drop versus filter ratio.
for glass fabric at various cleaning cycles
(Campbell and Fullerton, 1962).
be adequate to capture the emissions at the
source when the system resistance is a max-
imum and the gas volume a minimum. At
the same time, the filter ratio must not be
excessive immediately after cleaning •when
the system resistance is a minimum and the
gas volume a maximum.
Selecting or designing a baghouse requires
the following initial steps:
1. The minimum volume to be vented from
the basic equipment must be determined
according to the principles set forth else-
where in this manual.
2. A maximum desirable baghouse resistance
must be estimated.
3. The blower operating point is selected to
provide the minimum required volume at
the maximum baghouse resistance.
4. A minimum baghouse resistance is esti-
mated for the condition immediately after
the filter bags are thoroughly cleaned.
5. A second operating point on the blower charac-
teristic curve is determined for the clean bag
condition.
6. The minimum filtering area required is de-
termined by the maximum filtering velocity
permissible for the particular dust or fume
being collected.
7. The calculations are re checked,-with the fil-
tering area thus determined to ensure com-
patibility.
The most common deficiency in selecting and
designing baghouses and exhaust systems is
failure to take into consideration the normal
variation in air volume. The proper design
approach requires that the two extreme con-
ditions be considered separately rather than
on the basis of the average, because on the
average, conditions are not average.
Filtering Velocity
Filtering velocity or filter ratio is defined as
the ratio of gas filtered in cubic feet per min-
ute to the area of the filtering media in square
feet. The units of filter ratio are, therefore,
cfm/ft . By cancelling, the units of filter
ratio are reduced to feet per minute, and in
this form it Is often referred to as filtering
velocity. Physically, filter ratio, or filter-
ing velocity, represents the average velocity
with which the gas passes through the cloth
without regard to the fact that much of the
area is occupied by the fibers from -which the
cloth is woven. For this reason, the term
"superficial face velocity" is often used. Fil-
tering velocity is an important factor in fil-
tration. Too high a filter ratio results in
excessive pressure loss, reduced collection
efficiency, blinding, and rapid wear. Silver-
man (1950) recommends values of filtering
velocity from 0. 5 to 5. 0 fpm with an average
of 3.0 for common dusts. He states, how-
ever, that the velocity should be maintained
below 0. 5 fpm for fumes that tend to plug
fabrics. Watts and Higgins (1962) report that,
for control of emissions from brass smelter
operations, their experience indicates that
the filter ratio must be 1. 0 to 1.5 fpm or
even less -when spun Orion filter bags are used.
Adams (1964) recommends a maximum filter
ratio of 2. 0 for fumes from direct-arc steel-
melting furnaces with glass bags, or 3. 0 if
Orion bags are used. He estimates that aver-
age bag life imder these conditions is 18 months
for the glass and 5 years for the Orion. These
life figures generally appear to be too optimis-
tic in the case of the Orion and slightly pessi-
mistic in the case of the glass fiber bag.
Spaiteetal. (1961) recommend filter ratios of
1. 5 to 2. 0 fpm -when glass cloth is used athigh
temperature compared to 3. 0 fpm average
practice for low temperature filters. Drinker
-------
Baghouses
117
and Hatch (1954) also cite 3. 0 as a design fil-
ter ratio for typical dust and average concen-
trations. Stairmand (1956) gives a range of
1 to 6 feet per minute for normal fabric fil-
ters in actual practice but emphasizes the need
to operate with low filtering velocities since
higher velocities lead to compaction resulting
in excessive pressure drop or to breakdown of
the dust cake, which in turn results in reduced
collection efficiency. Roseetal. (1958) observe
that filter ratios range from 1 to 6 cubic feet per
minute per square foot of cloth area in practice
with 3. 0 as a common standard for normal dusts.
For metallurgical fumes, however, he recom-
mends that the filter ratio not exceed 1/2 to 1
cubic foot per minute per square foot of cloth
area. Brief et al. (1956) describe successful
baghouse installations serving direct-arc elec-
tric steel-melting furnaces using Orion bags at
filter ratios of 1.91 and 1.79.
Clement (1961) emphasizes that the filter ratio can-
not be too low from an operational viewpoint.
This is in conflict, however, with economic con-
siderations, which tend to prevent overdesign.
His recommended maximum filter ratios for
various dusts are shown in Table 38. These
values represent a compromise that experience
has shown optimum for minimizing total cost
Table 38. RECOMMENDED MAXIMUM FILTERING VELOCITIES AND MINIMUM DUST-CONVEYING
VELOCITIES FOR VARIOUS DUSTS AND FUMES (Clement, 1961)
Dust or fume
Alumina
Aluminum oxide
Abrasives
Asbestos
Buffing wheels
Bauxite
Baking powder
Bronze powder
Brunswick clay
Carbon
Coke
Charcoal
Cocoa
Chocolate
Cork
Ceramics
Clay
Chrome ore
Cotton
Cosmetics
Cleanser
Feeds and grain
Feldspar
Fertilizer
(bagging)
Fertilizer
(cooler, dryer]
Flour
Flint
Glass
Granite
Gypsum
Graphite
Maximum
filtering
velocity,
cfm/ft2
cloth area
2.25
2
3
2.75
3to3.25
2. 50
2.25to2.50
2
2.25
2
2. 25
2.25
2.25
2. 25
3
2.50
2. 25
2. 50
3. 50
Z
2. 25
3. 25
2. 50
2. 40
2
2. 50
2. 50
2. 50
2. 50
2. 50
2
Branch pipe
velocity,
fpm
44500c'f
4,500
4, 500
3,500 to 4,000
3,500 to 4, 000c'ord'b
4,500
4, 000 to 4, 500
5,000
4, 000 to 4, 500
4, 000 to 4, 500
4,000 to 4, 500a'8'h
4, 500a. g>h
4, 000a> e>g»h
4, 000a>e>g'h
3,000 to 3,500a'b. f
4,000 to 4,500
4, 000 to 4, 500
5, 000
3, 500a>b> c'f
4,000
4, 000a>b-g
3, 500a. h
4, 000 to 4, 500
4, 000
4, 500
3,500a»h
4, 500
4,000 to 4,500
4,500
4, 000
4, 500
Dust or fume
Iron ore
Iron oxide
Lampblack
Leather
Cement
crushing
Grinding (sep-
arators, cool-
ing, etc)
Conveying
Packers
Batch spouts
Limestone
Lead OKide
Lime
Manganese
Marble
Mica
Oyster shell
Paint pigments
Paper
Plastics
Quartz
Rock
Sanders
Silica
Soap
Starch
Sugar
Soapstone
Talc
Tobacco
Wood
Maximum
filtering
velocity,
cfm/ft2
cloth area
2
2
2
3.50
1. 50
2. 25
2.50
2. 75
3
2.75
2. 25
2
2. 25
3
2. 25
3
2
3. 50
2.50
2,75
3.25
3.25
2.75
2.25
2. 25
2.25
2. 25
2.25
3.50
3.50
Branch pipe
velocity,
fpm
4,500 to 5,000
4, 500
4,500
3, 500c>f
4, 500°' i
4, 000
4, 000
4, 000
4, 000
4, 500
4,500
4,000
5, 000
4, 500
4, 000
4, 500
4, 000,
3, 5001"
4,500a
4, 500
4, 500
4, 500b» d
4,500
3, 500a. b
3, 500a. b
4, 000a
4, 000
4,000
3, 500a>b'f
3, 500a,f
aPressure relief. bFlame-retardant cloth. cCyclone-type precleaner.
eSprinklers. £Special hoppers, gates, and valves.
1Insulate casing.
^Grounded bags.
dSpark arrester.
hSpecial electricals.
-------
118
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
when both maintenance and capital outlay are
considered.
Maximum recommended filter ratios should be
used as a guide only. Actual design values may
need to be reduced if grain loading is high or
particle size small, especially if the range of
particle sizes is also narrow as in a metallur-
gical fume. When compartmented baghouses
are used, the design filter ratio must be based
upon the available filter area remaining -with one
or two compartments offstream for cleaning or
servicing. Conventional baghouses for metallur-
gical fumes should in general be operated -with
filtering velocities in the range of 1. 5 to 2. 0. In
some cases, however, such as Lead fumes, the
experience of the Los Angeles County Air Pol-
lution Control District indicates that filter ratios
must not exceed 1. 0 and even less is recom-
mended. The values listed in Table 38 may be
used as a guide for other dusts.
Filtering Media
The filtering media selected for use in a baghouse
must be compatible with the temperature and pH
of the effluent. Maximum permissible tempera-
tures and chemical resistance are listed in Table
39 for the various fibers normally used for fil-
ter media in dust collectors. Each type of fiber
is also available in a wide range of cloth specifi-
cations, as illustrated by the data in Table 40,
-which lists specifications for only a few glass
fabrics.
Fibers
Cotton
Cotton has been for many years the standard
fiber for filter fabrics for common dusts. It
is inexpensive, readily available, an effective
filter media, and durable as long as the tempera-
ture is not excessive and no acid or strong alkali
Table 39. SUMMARY OF DATA ON
THE COMMON FILTER MEDIA USED IN
INDUSTRIAL BAGHOUSES
Fatifii-
done
Cotton
Dynel
Wool
Nylon
Orion
Dacron
Glass
Maximum temperature
at baghouse inlet
for continuous duty
Summary of
published data,
°F
160 to 190
150 to 180
180 to 235
200 to 290
200 to 350
250 to 350
500 to 700
Recommended
maximum,
°F
180
175
220
220
275
275
550
Chemical
resistance
Acid
Poor
Good
Good
Good
Good
Excellent
Excellent
Alkali
Fair
Good
Poor
Poor
Fair
Good
Excellent
is present. For applications such as abrasive
blasting, rock crushing, and conveying, cotton
•will probably continue to be the favored choice
for many years.
Wool
Before the development of the variety of synthetics
now available, wool was the only choice when the
temperature was around 200 °F or an acid condition
was present. Wool or a wool asbestos mixture is
still used in many metallurgical operations such as
secondary lead smelters though it has been sup-
planted to a great extent by Dae r on. In felted form,
wool has been the standard fabric foruse inHersey-
type reverse-jet baghouses.
Nylon
Nylon is a snythetic, organic fiber originally
developed by E.I. du Pont.de Nemours and Com-
pany and now produced by du Pont and other
manufacturers. It is available in both staple and
filament form. Nylon is relatively high in initial
cost, but it has many desirable physical proper-
ties. It has excellent resistance to abrasion
and flexing, toughness and elasticity, and resis-
tance to many chemicals (Filter Fabric Facts,
1954). Its heat resistance is not, however, as
good as that of Orion and Dacron. Because of
the slick surface, the filter cake may be removed
with a minimum of cleaning action. Nylon, how-
ever, is rarely used in baghouses, because other
synthetic fiber fabrics have higher heat resis-
tances and, in general, are equivalent in regard
to other properties.
Dynel
Acrylic fibers generally have low moisture
absorption, good strength, resilience, and
resistance to many chemicals and destructive
organisms such as mildew and bacteria. An
early acrylic-type fiber used for filter cloth was
Union Carbide and Carbon Corporation's Vinyon N,
a filament yarn. Vinyon N, a copolymer of an
acrylonitrile and vinyl chloride, was a modifica-
tion of the original Vinyon CF, which was a copoly-
mer of vinyl chloride and vinyl acetate. A modi-
fied version of this fiber in staple form is now
marketed under the name Dynel. Dynel has high
chemical resistance, particularly to strong alka-
lies and acids, and will not support combustion
(Filter Fabric Facts, 1954).
Orion and Dacron
Du Font's Orion, the first of the 100 percent
acrylics, is produced only in the staple form
at the present time. Originally both filament
and staple forms were available, but du Pont
discontinued manufacture of filament Orion
about 1957. Orion is light, strong, and resilient;
it has good heat resistance and excellent chemical
resistance, especially to acids (Filter Fabric Facts
-------
Baghouses
119
Table 40. TYPICAL SPECIFICATIONS FOR GLASS FILTER FABRICS
Fabric number
Average permeability
Mullen burst strength
(Avg PSI)
Weight, oz per yd
Thread count
Weave
Warp yarn
Fill yarn
501
17
588
9.36
54 x 52
Crowfoot
150's 1/2
ISO's 1/2
502
12
593
9.50
54 x 54
Crowfoot
ISO's 1/2
ISO's 1/2
600
81
485
8.27
64 x 34
3 x 1 Twill
150's 1/0
Bulked 1/4
601
75
595
10. 00
54 x 30
3 x 1 Twill
ISO's 1/2
Bulked 1/4
604
60
555
12.50
42 x 30
3 x 1 Twill
150's 2/2
Bulked 1 / 4
r 300
45 to 60
400
16. 30
48 x 22
2x2
Reverse
twill
150's 2/2
31/2 Staple
300A
30 to 40
450
17.67
48 x 24
2x2
Reverse
twill
150's 2/2
31/2 Staple
313A
33
540
13.50
34 x 42
Crowfoot
150's 2/2
Bulked 1/4
From: Menardi and Co. Bulletin
1954). At the time du Pont discontinued manufac-
turing filament Orion, Dacron was readily avail-
able. Since Dacron could be obtained in filament-
type yarn, felt by many to be superior to staple
yarn in cleanability, many users switched to Da-
cron at that time. Dacron, with similar physical
and chemical resistance properties, was also less
expensive than Orion.
Teflon
An experimental tetrafluoroethylene fiber, Teflon,
has been produced by du Pont but has received
only limited use in air filtration. It has exception-
al heat and chemical resistance but is also expen-
sive (Filter Fabric Facts, 1954). A Teflon-Orion
mixture called HT1 is used when fluorides are
present in the effluent in significant quantity.
Glass
Of all materials available for filtration, glass
fabrics have the highest resistance to high tem-
peratures and all chemicals (except fluorine).
Its physical weakness, however, particularly its
low abrasion and crushing resistance, requires
special precautions and design features. Care
must be taken to avoid damage by crushing in
packing, shipping, and storing (Underwood, 196Z).
Vigorous shaking must be avoided, though gentle
shaking with a period of about 50 cycles per min-
ute and amplitude about 5 percent of the bag length
is effective. The filtering velocity recommended,
to avoid blinding, is usually less than for other
fabrics on the same dust, since a more gentle
cleaning action is required.
Yorn
The characteristics of the filter cloth depend not
only on the material of -which the yarn is con-
structed, but also upon the construction of the
yarn, that is, weave, count, finish, and so forth.
Filament yarns
Filament yarns, available only in synthetic fibers,
are manufactured by extruding the material through
a perforated nozzle or spinneret. Individual fila-
ments may be twisted together to form a multi-
filament yarn. Filament yarns have a greater ten-
sile strength in relation to bulk and weight than
staple fiber yarns do. In addition, they have a
slicker surface (Filter Fabric Facts, 1954).
Staple yarns
Staple yarns of synthetic fibers are produced in
a similar manner, except that the filaments are
finer and shorter. One method of producing
staple fibers is to strike the filaments with a
blast of compressed air as they emerge from the
spinneret. The staple fibers are then caught on a
revolving drum from which they are gathered and
spun into a staple yarn. A variation of this pro-
cess is the production of a bulked filament. The
bulked or textured filament is produced by using
compressed air to rough up the surface of the
filament as it is extruded from the spinneret
(Marzocchi et al. , 1962).
Cotton staple fibers are cleaned and drawn into
parallel order by carding and other operations
and are eventually twisted into yarns by a spin-
ning process. Synthetic staple yarns are spun
in much the same manner. The properties of
spun yarn depend upon the amount of twist in spin-
ning. A highly twisted yarn tends to resist pene-
tration of particles into the interstices of the yarn
(Filter Fabric Facts. 1954).
Classification of yarns is different between cottons
and synthetics. In the cotton system, which is
used for spun yarns, yarns are measured in hanks
of 840 yards, and the yarn classification is the
number of hanks to the pound. Cotton yarns clas-
sified as 20's are, therefore, only half the weight
-------
120
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
of 10's. Filament yarns, on the other hand, are
classified in the European denier system. This
system, which originated with the old 450-meter
silk skein in 5/100 gram units, has a higher denier
to denote a heavier yarn. To convert, divide 5, 315
by the cotton yarn number to determine the filament
yarn denier (Filter Fabric Facts, 1954).
Weave
While the yarn and count are important, the weave
is also an important basic element in the construc-
tion and should not be neglected. There are three
basic variances of weave: Plain, twill, and satin.
The differences are the result of different systems
used in interlacing the lengthwise warp yarns with
the crosswise filling yarns.
Plain "weave
The plain -weave has a simple one up and one down
construction. It permits maximum yarn interlacing
per square inch and, in a tight weave, affords high
impermeability. II the count is lowered, this weave
may be made as open and porous as desired. The
plain weave is common in certain cotton ducks and
many synthetic constructions (Filter Fabric Facts,
1954).
Twill weave
The twill weave may be recognized by the sharp
diagonal twill line formed by the passage of a -warp
yarn over two or more filling yarns, the interlacing
moving one pick forward with each warp. In equiva-
lent constructions, twills have fewer interlacings
than the plain weave and, hence, greater porosity,
though this naturally depends on the count. Cotton
and synthetic filter twills are widely used (Filter
Fabric Facts, 1954).
Satin weave
The satin -weave, with even fewer interlacings
spaced widely but regularly, provides smooth
surface and increased porosity. These qualities
make them particularly valuable in gaseous filtra-
tion such as dust collection. Cotton fabrics in
this weave are commonly known as sateens. Cotton
sateen is probably more widely used than any other
fabric in baghouses for use at ambient temperature
(Filter Fabric Facts, 1954).
Finish
Dimensional stability is an important factor in fil-
ter fabrics. Cotton and wool fabrics must be pre-
shrunk. Synthetics are generally given a corre-
sponding treatment called heat-setting. This pro-
cess contributes to a more even balance of warp
and filling yarn tension, provides better surface
smoothness, reduces yarn slippage, controls poros-
ity, and virtually eliminates shrinkage, provided
the fabric is not subsequently exposed to exces-
sive temperature. The dimensional stability may
be lost if the fabric is subjected to temperatures
approaching that used in the original heat-setting
process. It is not unusual to observe bags that
have been subjected to excessive temperature with
shrinkage of 3 or 4 percent. This amounts to ap-
proximately 3 to 5 inches for a 6-inch-diameter
bag of average Length. As a result of the shrinkag
the bag may pull loose from its connection to the
floor plate or the upper support structure. In
some cases extensive damage to the baghouse stru
ture has occurred as a result of shrinkage.
Glass fabric bags are also given a treatment with
silicones derived from phenylmethyl silanes or
dimethyl silanes (Marzocchi et al. , 1962). Glass f
ter fabrics may be constructed of filament, sta-
ple or bulked (texturized) yarns, or a combina-
tion of these. An organic size or binder is ap-
plied to the glaiss fiber as it is extruded. This
later protects the fibers during the manufactur-
ing processes necessary to produce a fabric. After
weaving, the fabric is given a heat treatment. Dur
ing this treatment the organic size or binder is
burned off, and subsequently the silicone is appliec
which serves as a lubricant to protect the individ-
ual fibers from abrasion on each other.
Glass fabric is woven from multifilament yarns.
In one case investigated in Los Angeles County,
fumes from a gray iron cupola were found de-
posited among the fibers of the yarn. This ef-
fectively prevented relative motion of the individ-
ual fibers -when the cloth was flexed. The result
-was an apparent weakening of the cloth and a
greatly reduced bag life. This is thought to be a
result of the increased stress in the outer fibers
of each multifilament element because the yarn
•was forced to bend as though it -were a single sol-
id fiber instead of a bundle of individual fibers.
Other factors being equal, the maximum stress
introduced by flexure is proportional to the radi-
us of the liber. Washing the fabric with -water
and detergent removed the fume and restored
the cloth to its original strength. This illustrates
the importance of using the silicone coating as a
lubricant to permit the individual fibers to slide
upon one another as the cloth is flexed. Failure
of the silicone coating to function as intended re-
sults in rapid deterioration of the fabric. Laun-
dering glass filter bags periodically has become
routine in a number of plants in Los Angeles
County. A common practice is to maintain two
complete sets of bags. One set is laundered
-while the second is in use. The bags usually last
through several launderings.
Heat treatment relieves the stresses introduced
into the fibers because of the processes to which
they are subjected during fabrication of the yarn
and cloth. A permanent set is also put into the
glass fibers as a result of the heat treatment
-------
Baghouses
121
(Marzocchi et al. , 1962). During the heat treating
process, glass fabrics may be subjected to tem-
peratures of from 700 to 1, 200 °F. It is not, however,
recommended that these fabrics be exposed to
such high temperatures during use. The most im-
portant reason for this is that the silicone coating
undergoes a gradual deterioration at temperatures
approximately greater than 500°F (Spaite et al. ,
1961). The rate of deterioration of the silicones
increases with increasing temperatures. Thus,
short periods of operation at temperatures of
600 to 700 °F are permissible, but continuous oper-
ation at these temperatures will materially short-
en the life of the fabric. Tests have shown that
increased life can be attained by an additional
treatment with graphite (Spaite et al., 1963). At
present, the additional cost of the graphite treat-
ment does not appear to be warranted for most
high-temperature operations, but additional de-
velopments in this area may produce a superior
filter fabric for high-temperature operation.
Size and Shape of Filters
Diameters of tubular filtering elements
The most common shape of filter elements used
is a simple, circular cross-section tube. Most
standard commercial units employ tube diameters
of 5 or 6 inches. Filter cloth is provided in sev-
eral standard widths. One common size is approx-
imately 38 or 39 inches wide. Two 5- or 6-inch-
diameter bags can be obtained from a single
•width of cloth, the necessary seam being allowed.
For high-temperature applications, an 11-1/2- or
12-inch-diameter glass fiber bag is most com-
monly employed. Again, this is the most eco-
nomical size for the 38-inch-wide glass cloth that
is readily available. A few baghouses are de-
signed for use with 7- or 8-inch-diameter bags.
This size is probably based upon a 54-inch-wide
cloth from which two bags can be obtained from
a single width. Wool felts, which are used in
the Hersey reverse-air jet baghouses, are gen-
erally either 9 or 10 or 20 inches in diameter.
In general, bag diameters are determined main-
ly by the available "widths of yard goods.
The diameter of the filter bags used also influ-
ences the size of the baghouse. For example,
about 1, 750 square feet of filtering area can be
provided in about 80 square feet of floor area by
using 6-inch-diameter by 10-foot-long bags. If
12-inch-diameter bags were used instead, they
•would need to be about 14 feet long to provide the
same filtering area in the same floor space, though
12-inch-diameter bags can easily be made 20feet
long if there is adequate head room. This results
in a baghouse having about 2, 500 square feet of
filtering area in the same floor space.
Length of tubular bags
The length of cloth filter elements varies from
about 5 feet to approximately 30 feet. Most standard
baghouses employing 5- or 6-inch-diameter bags
use bag lengths from 5-1/2 feet to 10-1/2 feet.
The lengths for 11-1/2- or 12-inch-diameter bags
are generally about 15 to 25 feet.
Length-to-diameter ratio
Manufacturers have apparently not attempted to
establish a standard length-to-diameter ratio.
Indeed, from a theoretical point of view, the
length-to-diameter ratio should have no effect on
the collection efficiency of a bag except for the
influence of elutriation as previously discussed.
This ratio is, however, important from another
aspect. Assume an extreme case of a 30-foot-
long and 5-inch-diameter bag. When shaken,
such a bag will sway excessively. This could
easily result in one bag's rubbing upon the adja-
cent bag, which would be detrimental to good bag
life. Another aspect of the problem concerns the
cleaning of the bag by means of shaking. In order
to clean the bags adequately, sufficient force must
be applied to break up the dust cake and dislodge
some of the embedded dust from the fabric. Studies
have shown that, as the force applied is increased
(as measured by the acceleration given the bag by
the shaking mechanism), there is an increase in
the effectiveness of the cleaning up to a limiting
value (Walsh and Spaite, 1962). The studies
have also shown that the residual dust profile
varies along the length of the filter tube. This
is a result both of the manner in which the shaking
force is transmitted to the tube, and of the varia-
tion in dust cake properties. The efficiency of
cleaning by means of mechanical shaking varies
depending upon the length-to-diameter ratio,
though the manner of variation is not known. Ob-
viously, there is an optimum length-to-diameter
ratio that may differ for different cloths, dusts,
shaking intensities,and shaking frequencies
(Stephan et al., I960). Another factor that effec-
tively limits the length-to-diameter ratio is the
difficulty of fabrication. Sewing the longitudinal
seam becomes increasingly difficult as the length
of the bag increases. Continuous tube "weaving
could, however, be employed, if increased
lengths were advantageous.
Substantially more investigation is needed in this
field. At present, additional filtering area is
apparently frequently incorporated by increasing
the length of the filter tubes. When the length
appears to be unreasonably long or if there is a
limitation on head room, then the number of
filter tubes is increased.
An absolute limiting length of 30 times the di-
ameter has been suggested by Silverman (1950),
-------
122
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
based upon some unspecified experiments in
metallurgical baghouses. Some installations
using 11-1/2- or 12-inch-diameter bags operate
successfully with bag lengths up to about 30 feet.
These are, however, exceptional instances, and
a more practical limit appears to be about a 20-
to-1 length-to-diameter ratio. Most 11-1/2-inch-
diameter bags in Los Angeles County are made of
glass cloth and are 15 to 20 feet long (Crabaugh
et al., 1954). Greater length increases the stress
because of the greater weight that must be sup-
ported by the fabric. If mechanical shaking is
used, cleaning may be less effective unless more
energy is applied to each bag. This also in-
creases the stress. In high-temperature instal-
lations, dimensional instability may be increased.
Problems of stretching and shrinking may occur
at times, -which could be mitigated by using
shorter bags so that the same percent change in
length would not be excessive in absolute amount.
Capital outlay and floor space are both reduced
with an increase in bag length -while maintenance
is increased. Many 5- or 6-inch-diameter bags
are 8 to 10 feet long. Although no data establish
an optimum length-to-diameter ratio, a 20-to-l
ratio appears to be an approximate practical
limit. This is one of many areas in baghouse de-
sign that could benefit from further study.
maintenance are, however, not as easily accom-
plished as for the simple tube-type bag. Wear
is increased because of the friction bet-ween
the filter cloth and the -wire frame support struc-
ture. It would not be advisable, therefore, to
use this type of bag for application where rapid
wear of the filter media is anticipated. Applica-
tions for temperatures in excess of 300°F are,
therefore, ruled out completely because glass
cloth is not able to -withstand the abrasion. Only
a relatively few baghouses of this type are used
with synthetic filter fabrics such as Orion or
Dacron for intermediate temperatures. Dust
is collected on the outside of envelope bags as
opposed to the inside of tubular-type bags.
INSTALLATION OF FILTERS
Arrangement
The arrangement of tubular bags shown in Figure
71 can materially affect the number of bags that
can be installed in a given area. The staggered
arrangement is not as desirable as the straight,
even though it uses the area more efficiently, be-
cause access for maintenance, inspection, and
bag replacement is more difficult.
Multiple-tube bags
A variation of the tube-type bag is oval in cross-
section with vertical stitching that divides the
bag into several compartments. When inflated,
each compartment assumes a nearly circular
cross-section.
When the blower is turned off and the pressure
relieved, the bag returns to an oval shape,
•which helps to break up the filter cake. A bag
such as this requires a special mounting and is
somewhat more expensive for the same filtering
area than a standard round tubular bag. It has
an advantage in that a greater filtering area can
be accommodated in the same size housing.
There is, however, a disadvantage in that a hole
in one of the bags effectively destroys a greater
filtering area, and maintenance cost could thus
be substantially higher than for a conventional
baghouse in some cases.
Envelope type
Baghouses with envelope-shapedbags are second
only to the tubular-type bag in a number of units in
use. The filtering elements must be mounted on a
supporting structure usually made of wire. In
comparison to other designs, the envelope-type
baghouse permits a greater filtering area to be
installed in a given size volume. Inspection and
**•«»*•«€)
(id««*«
Figure 71. Arrangements of filter bags: 78 bags
arranged in line (good); 108 bags in staggered
arrangement in same size housing (poor) (Northern
Blower Division, Buell Engineering Co., Inc.,
Cleveland, Ohio).
-------
Baghouses
123
Bag Spacing
The clearance bet-ween bags is important for at
least two reasons. First, sufficient clearance
must be provided so that one bag does not rub
upon its neighbor. This is particularly impor-
tant for baghouses employing mechanical shaking
•where the vibration may cause the bags to oscil-
late. A minimum clearance of 2 inches is sug-
gested between bags of average length. Larger
clearances should be provided if the bags are
unusually long, for example, greater than 10 or
12 feet. Second, access for examination and
maintenance must be provided.
Walk-ways between banks of bags must also be
provided. The depth of banks should not be so
great that it is difficult or impossible to reach
to the farthest bag for maintenance and replace-
ment. This means that if a -walkway is provided
on one side only, each bank should be no more
than three or four bags deep if 6-inch-diameter
bags are used. Twelve -inch-diameter bags
should not be more than two bags deep if access
is provided on one side only. If access is pro-
vided on both sides, 6-inch-diameter bags must
not be more than eight bags deep, and 1 2-inch-
diameter bags must not be more than four bags
deep. The total number of bags in a bank de-
pends upon the shaking mechanism employed.
A single bank, in general, is operated by a
single shaking mechanism. A single compart-
ment may contain several banks of bags.
Walkways must be provided so that all portions
of the mechanism are easily accessible. Walk-
ways should be at least 18 inches wide: a 24-
inch width is recommended. When the bags are
longer than about 10 or 12 feet, a walkway should
be provided at two levels, one at the floor plate
and a second for access to the upper support
structure.
Bag Attachment
Bottom attachments
Tubular bags are most frequently attached to a
thimble on the tube sheet or floor plate, as il-
lustrated in Figure 72. A steel band is instal-
led around the bag bottom to effect a tight seal
between the cloth and the thimble. A cuff may
be sewn into the bottom of the bag, or the bot-
tom may be folded up once or twice to form a
self-cuff. This is the simplest, most trouble-
free arrangement and probably the most widely
used means of attaching tubular bags at the
bottom. The steel bands should be made of
stainless steel to avoid rust and corrosion prob-
lems. A simple screw-type closure mechanism
is usually employed, but quick-closing clamps,
as shown in Figure 72, are also available.
In a second method of attachment, a spring steel
band sewed into the bottom cuff of the bag is used.
The steel band is collapsed, inserted into a hole
in the floor plates, and allowed to expand. A
tight seal is required between the bag cuff and the
hole to prevent leakage; this requires a perfect fit.
A strip of padding is sewed into the cuff to help
adjust for size variations of the steel band and of
the hole. Because variations in size are inherent
in all manufacturing processes, it is difficult to
achieve sufficient uniformity of the bag and hole
to ensure a dustproof fit bet-ween them. This has
been tested by the Los Angeles County Air Pollu-
tion Control District on baghouses that serve
direct-arc steel-melting furnaces. Dust and
fume losses are usually 5 to 10 percent that in the
effluent stream entering the baghouse and not in-
frequently are greater. For comparison, dust
and fume losses from baghouses serving similar
furnaces, but with bags attached by other methods,
are usually 1 to 2 percent that in the effluent stream
entering the baghouse. Fitting a spring steel band
into a recessed hole in the cuff of the bag permits
more rapid installation. When relatively coarse
dusts are involved, adequate collection efficiency
can be attained, provided extra care is taken when
new bags are installed to make sure that the bags
fit well and are seated properly in the retaining
hole.
Top support
The top of the bag may also be installed over a
thimble by using a steel band in a manner similar
to that used with a thimble at the bottom. When
a thimble is used, the bag may have a cuff sewed
into it, or the end of the bag may be folded to
form a self-cuff. In most cases when mechani-
cal shaking is employed, this type of attachment
offers an advantage, since wear is usually most
severe near the bottom of the bag. The life of
the bag can usually be extended substantially by
making the bags extra long, folding the extra ma-
terial under the clamp at the top, and then lower-
ing the bag periodically about 3 to 4 inches at a
time. Bag life may be further extended by re-
versing the bags, top to bottom, provided this
is done before wear proceeds too far.
Another method consists of attaching the bag on-
to a steel disc or cap, which is supported at the
center, as illustrated in Figure 73. A common
method of attaching 5- or 6-inch-diameter bags
consists of sewing a loop at the end of the
bag; the loop is then placed over a hook as
shown in Figure 73. Another method involves
sewing the end of the bag into a flat strap, which
-------
124
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
Figure 72. Connection of filter bags to thimbles: Men attaching filter bags and two examples
of quick closing clamps (left and lower right, Northern Blower Division Buell Engineering Co.,
Inc., Cleveland, Ohio; upper right, Fuller Company, Dracco Division, Cleveland, Ohio).
is looped back and forth over a special hanger as
illustrated in Figure 74. This method is simple
and permits rapid installation. The length of the
bag is not critical, since adjustments are easily
made during installation or at any time there-
after.
Some bags of this type, which have a strap at the
upper end, were found to be developing small
holes near the top of the bag. The same situation
was found to be developing in several baghouses
using bags of the same design and manufacture.
Investigation revealed that the construction used
resulted in a stress concentration in a small area
of the bag. This problem was eliminated by using
a different sewing technique.
CLEANING OF FILTERS
Methods
As dust accumulates on the filtering elements,
the pressure loss increases until some maxi-
mum desirable value is reached. The filter
must then be cleaned to reduce the pressure
loss. Cleaning cycles may be manual, semi-
automatic, or fully automatic. Fully automatic
cycles may be initiated on a time cycle or -when
the pressure reaches a preset amount. Figure
75 shows a pressure switch with this function.
Some reverse-jet baghouses operate with con-
tinuous cleaning. Once a cleaning cycle is ini-
tiated, it should be carried through to comple-
tion v/ith sufficient cleaning intensity and time
duration to ensure thorough cleaning. Thorough
cleaning is also recommended each and every
time the blower is turned off (Stephan et al. ,
1960).
Manual cleaning
Small baghouses with up to about 500 or 600
square feet of filtering area are frequently
cleaned by hand levers. A manually operated
handle transmits a rap to the framework from
which the filtering elements are suspended.
This shakes the dust loose. Thorough clean-
ing is rarely achieved since a great amount of
-------
Baghouses
125
Figure 73. Connection of filter bags to top support: (a) hag loop and hook (Wheelabrator
Corporation, Mishawaka, Indiana), (b) caps for use with clamps (Northern Blower Division,
Buell Engineering Company, Inc., Cleveland, Ohio).
vigor must be applied continuously for several
minutes. Many workmen are not aware of the
amount of cleaning required or are not con-
scientious enough to clean the baghouse thor-
oughly each time. Since these small baghouses
rarely have manometers to indicate the pres-
sure, the operator cannot readily determine
when the baghouse has been adequately cleaned.
The use of a manometer appears to be almost
essential. One must, of course, shut the fan
off or otherwise deflate the bags before start-
ing to clean them.
Mechanical shakers
Most baghouses employ some type of mechan-
ical shaking. The electric motor shaker is
most common. A cam or eccentric translates
the rotary motion of the motor into an oscil-
lation. Bags may be shaken horizontally or
vertically.
It is essential that there be no pressure inside
a tubular filter bag during the shaking cycle.
A pressure too small to be measured with a
manometer may still be sufficient to interfere
•with adequate cleaning (Herrick, 1963). In
one investigation a pressure as small as 0. 02
inch of water column prevented effective clean-
ing (Mumford et al., 1940). Butterfly-type
dampers, unless they are positive seating, can-
not be used to close off a section for shaking
while the blower is operating. For this reason,
a small amount of reverse airflow is commonly
used to ensure complete bag collapse during shak-
ing unless the blower is off during the cleaning
cycle. When the baghouse serves a hot source
such as a furnace, the thermal drive may be suf-
ficient to interfere with cleaning even after the
blower is off.
Pneumatic shakers
Two types of pneumatic cleaning mechanisms
are used. In one type the air is used to operate
an air motor that imparts a high-frequency vi-
bration to the bag suspension framework. Al-
though the frequency is high, the amplitude is
low. This method is not effective for materials
difficult to shake loose from the bags, since the
total amount of energy imparted to the bag is low.
For dust from sandblasting operations, themeth-
-------
126
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
Figure 74. Method of hanging filter bags with strap top: (1) The end of the strap on the tube is brought
up between the two horizontal bars of the tube hook. (2) The strap end is folded over the bar, directly
under the vertical threaded spindle. (3) The remainder of the strap and the tube proper are brought up
and over to the left, with the strap wrapping around the offset horizontal bar of the tube hook, and lying
on, or over, the other end of the strap, originally threaded through the hook. The bag, as shown, can be
raised, if necessary, by pulling with the right hand on the other end of the strap. (4) The correctly in-
stalled tube. Note that the tube proper hangs directly under the vertical threaded spindle of the tube
hook (Wheelabrator Corporation, Mishawaka, Ind.).
-------
Baghouses
127
Figure 75. Differential-pressure switch used to
control cleaning of bags (F. W. Dwyer Manufact-
uring Company, Michigan City, Indiana).
od has been found adequate for small baghouses.
For larger units, two or more air motor shakers
must be used. This cleaning method is econom-
ically feasible only when compressed air is al-
ready available close to the baghouse.
Pneumatic cylinders are often used for cleaning
glass fiber filter bags. This method is used
on many of the baghouses serving gray iron
cupolas. The pneumatic cylinder gently oscil-
lates the framework from which the bags are
suspended. It is frequently used in conjunction
with reverse-air collapse of the bags. The am-
plitude is relatively large and the frequency low.
Bag collapse
Efficiency of cleaning can frequently be improved
by permitting a small volume of air to flow in
the reverse direction through the bags, causing
them to collapse completely. This method is
frequently used with glass fiber bags. The bags
may be collapsed and reinflated several times
for each cleaning. Usually a gentle action is ob-
tained by slowly opening and closing the control
valves. Sometimes, however, a stronger clean-
ing action is required, and the valves are opened
and closed quickly so that the bags "snap. " Bag
collapse may also be used with mechanical shak-
ing, sonic cleaning, or air pulses.
In one variation of this method, several rings
are installed on the inside of glass fiber bags.
When the air is reversed, the bags collapse in-
ward but the rings prevent the cloth from touch-
ing at the center. The flexing of the fabric breaks
the filter cake loose. This assertedly permits
the cake to fall free without interference. Air
pulses are sometimes used for the same reason.
During the gentle air reversal, before applica-
tion of the air pulse, the bags relax and have a
tendency towards collapsing. As the short air
pulses (generally three pulses of 1 second each)
sweep down the filter tube, they create a gentle
waving or shaking action, as shown in Figure 76.
Sonic cleaning
Sonic cleaning is relatively new and has not been
fully evaluated in the field. It is usually used
with bag collapse. The sonic horns employed
are relatively expensive, and it is doubtful that
the cleaning action is superior to that provided
by simple mechanical shaking. In addition, the
sound can be extremely annoying unless the bag-
house housing is insulated with sound-absorbing
materials.
-------
128
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
TUBE
COLLECTING
DUST
REVERSE
AIR ON
ONLY
PRESSURE JET
AND REVERSE
AIR ON
/\
WALLS COLLAPSE TOGETHER
PREVENT DUST FROM FALLING
SLUG OF AIR OPENS TUBE,-
ALLOWS OUST TO FALL FREELY
Figure 76. Illustration of method of cleaning
bags by collapse and air pulses (Pangborn Cor-
poration, Hagerstown, Indiana).
Reverse airflow
Some envelope-type baghouses use reverse air
for cleaning. The dust is collected on the out-
side of the filtering elements. A moving car-
riage seals off the outlet of one or several bags
at a time. Valves permit outside air to be
drawn through the bags in the reverse direction,
as shown in Figure 66. This permits continuous
cleaning, •with only a few bags, at the most, out
of service at any time. Sometimes a separate
air blower is used to provide the reverse air for
cleaning.
Reverse-air jets
Another reverse-air cleaning method is the re-
verse-jet mechanism developed by Hersey, about
1950. The Hersey-type baghouse, as shown in
Figure 67,uses a blow ring that travels up and
down the tubular bag. Air for cleaning is blown
through a narrow slot on the traveling ring through
the filter media in the reverse direction. The
filter fabric is also indented or flexed at the
point of contact with the ring. The combination
of flexing and reverse airflow thoroughly cleans
the accumulated dust from the bag. Filter bags
are usually made of felted -wool cloth. Felted
Orion or Dacroii are used for higher tempera-
tures or for better resistance to chemical at-
tack. Woven fabrics are sometimes used,but
they usually suffer from reduced collection ef-
ficiency because this cleaning method is too
thorough. A residual dust cake is essential to
the filtering process with -woven fabrics. Felted
cloths, however, do not require a residual dust
mat to filter effectively. The reverse-air jet
cleaning method sometimes results in a high
rate of -wear. Even though reports have been
published indicating Taag life of several years,
experience in Los Angeles County has varied
depending upon the application. When this
method has been applied to the collection of
metallurgical fumes, extremely high rates of
bag wear have ~been experienced. Mechanical
breakdowns of the reverse-air mechanisms
have also been encountered. All the units in-
stalled in Los Angeles County to serve metal-
lurgical operations have been abandoned after
a few months of operation or modified to me-
chanical shaker cleaning. A number of them
have, however, been operated successfully
for controlling dust from grain transfer and
other common dust operations.
There is a tendency to believe that the re-
verse-jet baghouse may be operated with fil-
tering velocities of about 20 to 30 fpm or even
more. This is not generally true, however,
as is shown by the Hersey data reproduced in
Figure 77. High filter ratios are permissible
in special cases only. For example, from
curve 9 in Figure 77, a filter ratio of 30-would
be permissible for leather-buffing dust (a very
coarse material with 30-mesh average size)
only if the grain loading were low, not over 3
or 4 grains per cubic foot. For higher grain
loadings the filter ratio should be reduced to
about 20 cfrn per square foot. For metallur-
gical fumes a maximum filter ratio of about
6 is often recommended, as shown in Table
41. If the grain loading is greater than aver-
age or the particle size is small, the filter-
ing velocity should be reduced to 5 fpm or
less. These recommendations are confirmed
by Hersey's curves, which show that, for
very fine dusts and fumes, the limiting filter
ratio should be approximately 6 as the grain
loading approaches zero. For normal condi-
tions, the filter ratio should be 3 to 5 for met-
allurgical fumes. Caplan (I960) states that the
nature of the dust is the most important variable.
Hersey-type baghouses should logically be oper-
ated with filter ratios that bear a fixed relation-
ship to those used with standard shaker-type bag-
houses. From experience with a variety of medi-
um and coarse dusts, one would expect that
-------
Baghouses
129
5 10 15 20 25
FILTERING VELOCITY dm ft2 FABRIC
Figure 77. Typical performance of reverse-jet
baghouses on a variety of dusts--dust load ver-
sus filtering velocity at 3*/2 inches water col-
umn pressure drop (Mersey, 1955). The key to
the numbers is as follows: (1) magnesium tri-
silicate, (2) carbon black, (3) starch dust, (4)
resinox, (5) diatomaceous earth, (6) kaolin, (7)
cement or limestone dust, (8) coal dust, and (9)
leather-buffing dust. For numbers 1 through 6,
99.94 to 99.99% pass 325 mesh; for numbers 7 and
8, 95% pass 200 mesh; number 9 is the 60-mesh
average.
reverse-air jet baghouses could be operated with
filtering velocities 5 or 6 times as great as those
for conventional baghouses. It has been well es-
tablished that, for most metallurgical dust and
fumes, filtration should be 1 to 2 fpm in conven-*
tional compartmented baghouses cleaned by shak-
ing, collapse, air pulses, or combinations of
these. When metallurgical or other problem
dusts and fumes are involved, the design of Hersey
baghouses should be more conservative than would
be indicated by strictly folio-wing any arbitrary
rule.
In order to avoid operating difficulties, the pres-
sure drop for a Hersey-type baghouse should usu-
ally be in the range of 3 to 5 inches water column
(Caplan, I960). Too low a resistance is undesirable,
since it prevents proper inflation of the bag. This
results in improper cleaning action. Too high
resistance is also undesirable, since it increases
the friction bet-ween the blow ring and the bag,
which increases -wear excessively. Hersey-type
baghouses should not be operated with pressure
drops in excess of 8 inches water column under
any circumstances (Caplan, I960). When the
cleaning cycle is pressure controlled, these lim-
its may be used as a guide. If, however, the fil-
tering velocity is excessive, some materials,
for example, metallurgical fumes, have a tenden-
cy to blind the bags so that even continuous clean-
ing fails to reduce the pressure as required. When
materials such as these are handled, filtering
velocities must be reduced. Using the values
recommended in Table 41 should provide trouble-
free operation in almost all cases. Pilot model
studies are useful when previous experience is
not available as a basis for determining filter
ratio. Sufficient time must be allowed for the
pilot unit to reach equilibrium before tests are
started. This may require several hundred
hours of continuous operation. Failure to allow
equilibrium to be attained can result in errone-
ous data and improper functioning of the full-
scale unit designed upon these data.
The speed or rate of travel of the blow ring up
and down the bag may be varied according to
the nature of the dust being filtered. In general,
speeds of from 20 to 50 fpm are employed. The
optimum rate of blow ring travel depends upon
the nature of the dust. As the blow ring travels,
the dust is blown off the inside surface of the
bag. This dust will tend to settle at a rate that
depends upon the particle size and the specific
gravity of the individual particles. It is prob-
ably desirable to adjust the blow ring rate of
travel so that it does not exceed the settling rate
of the dust. A ring speed of approximately 20
fpm has been found optimum for light materials
such as grain and flour dust. Speeds of 40 or
50 fpm can be tolerated by high-density dust
such as uranium. The volume of air blown
through the slot of the blow ring is usually 1. 0
to 1. 5 cubic feet per linear inch of slot. Slot
widths are generally 0. 03 to 0. 25 inch (Caplan,
I960). Newer designs employ wider slots and
centrifugal blowers to provide the reverse air.
The original design used a positive-displace-
ment blower. The reverse air must be pro-
vided at a pressure greater than the pressure
drop through the filter cloth. Furthermore,
since higher pressure drops generally indicate
finer dusts and fumes, -which tend to penetrate
the fabric to a greater extent, the differential
between the reverse air pressure and the pres-
sure inside the bag probably should be increased
somewhat as the pressure drop across the bags
increases.
When hot effluents -with a high moisture content
are handled, it may be necessary to preheat
the air used for reverse-jet cleaning. For ex-
ample, in one case encountered in Los Angeles
County, the effluent from a direct-fired dryer
was vented to a reverse-jet baghouse. When
ambient air was used for cleaning, condensa-
tion occurred "when the unit was started up early
in the morning. After a heat exchanger was in-
stalled to preheat the reverse air, the bags re-
mained dry. In some cases a portion of the hot,
clean exhaust may be used for reverse-jet clean-
ing. Care must be taken to remain at least 50° F
-------
130
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
Table 41. RECOMMENDED FABRIC AND MAXIMUM FILTERING VELOCITY
FOR DUST AND FUME COLLECTION IN REVERSE-J3T BAGHOUSES
(American Air Filter Co., Inc., Bulletin No. 279C, Louisville, Ky. )
Material
AruminTim oxide
Bauxite
Carbon, calcined
Carbon, green
Carbon, banbury mixer
Cement, raw
Cement, finished
Cement, milling
Chrome, (ferro) crushing
Clay, green
Clay, vitrified silicious
Enamel, (porcelain)
Flour
Grain
Graphite
Gypsum
Lead oxide fume
Lime
Limestone (crushing)
Metallurgical fumes
Mica
Paint pigments
Phenolic molding powders
Polyvinyl chloride (PVC)
Refractory brick sizing (after firing)
Sand scrubber
Silicon carbide
Soap and detergent powder
Soy bean
Starch
Sugar
Talc
Tantalum fluoride
Tobacco
Wood flour
Wood sawing
Zinc, metallic
Zinc, oxide
Zirconium oxide
Fabric
Cotton sateen
Cotton sateen
Cotton sateen, -wool felt
Orion felt
Wool felt
Cotton sateen
Cotton sateen
Cotton sateen
Cotton sateen
Cotton sateen
Cotton sateen
Cotton sateen
Cotton sateen
Wool felt, cotton sateen
Wool felt
Cotton sateen, Orion felt
Orion felt, -wool felt
Cotton sateen
Cotton sateen
Orion felt, wool felt
Cotton sateen
Cotton sateen
Cotton sateen
Wool felt
Cotton sateen
Cotton sateen, wool felt
Cotton sateen
Dacron felt, Orion felt
Cotton sateen
Cotton sateen
Cotton sateen, -wool felt
Cotton sateen
Orion felt
Cotton sateen
Cotton sateen
Cotton sateen
Orion felt, Dacron felt
Orion felt
Orion felt
Filtering
velocity, fpm
9
8
7a
5
?a
7
9
7
9
8
10
10
10a
12
5a
8
6a
8
9
6a
9
8
8
?a
10
7a
10
9a
10
8
8a
9
6a
9
8
9
8
6a
7
aDecrease 1 fpm if concentration is great or particle size small.
above the dew point at all times to avoid trouble
(Caplan, I960).
Metallurgical fumes may bleed through the filter
bags. In one case -where a synthetic felted bag
was used, a hard crust formed on the edges of
the blow ring slot. This crust rapidly wore out
the bags. Substitution of wool felted bags cured
the problem. Particle size is not, however, the
sole determining factor in leakage. Disturbance
of the dust deposit causes some particles to sift
through even dense wool felt. Hence, the less
reverse-jet activity, the higher the average col-
lection efficiency (Hersey, 1955). Collection
efficiency of fly ash (mass median size 16 |j.) was
found to be less than for either talc (mass medi-
an size 2. 5 fx) or vaporized silica (mass median
size 0,6 (j.) (Hersey, 1955), but the reason for
this was not determined.
Cleaning Cycles
Manually initiated cycles
Cleaning is most commonly initiated by manually
operating the required controls. Electrically or
-------
Baghouses
131
pneumatically operated shakers are activated by
pressing a control button or operating a valve.
Interlocks are frequently provided so that the fan
or blower must be shut off before the shaking
mechanism can be activated. This arrangement
is most suitable for operations that may be shut
•down whenever required for cleaning. It is also
suitable when cleaning once or twice a shift is
adequate. In such cases, the baghouse is cleaned
when the equipment is shut down for lunch or at
the end of the shift.
Semiautomatic cycles
Some installations use a semiautomatic clean-
ing cycle whereby, as the blower is turned off,
a timer is activated. After a delay to permit the
blower to come to rest, the shaking cycle is
initiated. An interlock prevents turning the blow-
er on again before the shaking cycle is completed.
This method has been used with success on melt-
ing furnaces where a heat does not last more than
about 2 hours and the baghouse is adequately sized,
so that shaking is not required more than once per
heat. At the end of the heat, when there are no
emissions from the furnace, the operator presses
the button that initiates the cycle. In about 5 or
6 minutes the baghouse has been cleaned and is
ready to control emissions from the next heat.
While the baghouse is being cleaned, the fur-
nace is empty and no air contaminants are re-
leased.
Fully automatic cycles
The most desirable method consists of cleaning
a fully automatic, compartmented baghouse on
a programmed cycle. The cycle may be initiated
at regular intervals or -when the pressure reaches
a predetermined value. When the cleaning cycle
is initiated, one compartment of the baghouse is
isolated by means of appropriate dampers. A
small volume of reverse air is usually used to
ensure collapse of the filter bags. The isolated
section is then cleaned by one of the methods
previously discussed. After the cleaning cycle
is completed, the compartment is again re-
turned to service. Each compartment, in turn,
is cleaned in the same manner. The advantage
of fully automatic cleaning is that it eliminates
the possibility of the operator's forgetting or
neglecting to clean the baghouse. Since, how-
ever, a greater amount of mechanism is re-
quired, the maintenance and the possibility of a
breakdown are increased slightly. In many
cases, fully automatic cleaning is essential
since the basic equipment served cannot be shut
down while the baghouse is cleaned. Equipment
that operates continuously or requires cleaning
during the cycle of operation requires the use
of a fully automatic, compartmented baghouse.
Compartmented baghouses must be designed to
provide adequate filtering area during all phases
of the operation. This means that, when one
section of the baghouse is out of service for
cleaning, the remaining sections must provide
sufficient filtering area. Frequently the design
permits two sections to be out of service at
one time and still provides sufficient filtering
area. This allows one section to be serviced
when bags need replacement while the remain-
ing sections continue to operate -without exceed-
ing the maximum permissible filtering velocity.
Compartmented baghouses are not, however,
suitable for very small units. A minimum of
five or six compartments is required for effi-
cient operation.
Continuous cleaning
Continuous cleaning is often used in Hersey-
type reverse-jet baghouses and in some envelope
types. It is suitable for installations that oper-
ate with a steady high dust load. If the dust load
is variable or light, continuous cleaning will
result in unnecessary operation of the carriage
and in excessive wear. Pressure control clean-
ing cycles allow an increase in the resistance of
the filter above what the same unit would have
for continuous cleaning, as illustrated in Figure
78. The curves show that, for a typical dust con-
centration of 0. 5 grain per cubic foot, operating
the cleaning mechanism 30 percent of the time
instead of 100 percent results in only a 10 per-
cent increase in filter resistance. If the filter-
ing area of the unit were increased 10 percent,
the pressure drop could be expected to be about
the same, but the filtering media would last
about 3 times as long. While the benefits are
not as great for heavier dust loading or fine
metallurgical fumes, pressure control cleaning
may still be advantageous since the cleaning
mechanism need not be operated as much during
periods of very light loading.
DISPOSAL OF COLLECTED DUST
Once the dust is collected in a baghouse, it must
be disposed of without creating a new dust prob-
lem. Occasionally one sees dust dropped on the
ground from the collecting hopper of a baghouse.
The wind then picks it up and blows it around the
neighborhood. The result is substantially the
same as if the dust had not been collected in the
first place.
The most common means of disposing of the col-
lected dust is to transfer it from the hopper of
the baghouse into a truck and then to a dump. In
order to minimize dust emissions during trans-
fer from the hopper to the truckj a sleeve or
-------
132
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
sock of canvas is frequently installed on the out-
let of the hopper. The sleeve should be suffi-
ciently long to reach to the floor of the truck
body. The dust must then be thoroughly -wetted
with water before it is transported to the dump.
This method is suitable for installations where-
in the quantity of dust collected is such that
emptying the hopper once a day is sufficient.
DUST AT 5 TO 10
gr/ft3 OR
METALLURGICAL FUME
AT 1 gr/ft3
TYPICAL DUST AT
0 5 gr/ft3
I i
20 40 60 80
REVERSE-JET OPERATION,
Figure 78. Effect, of pressure control on filter
resistance in a reverse-jet baghouse (Caplan,
1960).
When the quantity of dust collected is greater,
the hoppers must be emptied more frequently.
Some type of automatic or semiautomatic meth-
od is then advisable. One method consists of
using a trickle valve as illustrated in Figure 79.
The discharge may be to a completely enclosed
tote box. Another method consists of using a
rotary valve (Figure 79) that may be operated
continuously or intermittently. Both the trickle
and the rotary valve may be connected to dis-
charge to a screw conveyor that collects the
dust from several hoppers, sometimes even
from more than one baghouse, and discharges
into a covered tote box or other common coL-
lection point.
BAGHOUSE CONSTRUCTION
Pushthrough versus Pullthrough
The blower may be located on either side of the
baghouse. If it. is on the clean-air side, it is
referred to as a pullthrough baghouse. This
is desirable since it protects the blower from
the dust or fume being handled. On the other
hand, it does require a relatively airtight hous-
ing for the baghouse. The pushthrough type can
be operated with open sides as long as protection
from the weather is provided. This is advantageou
when handling hot gases, since it permits a greatei
degree of cooling. Thus, a higher inlet gas tem-
perature may be tolerated for the same tempera-
ture of the filtering media. For a pushthrough bag
house, however, the blower must handle the entire
dust load. This frequently amounts to several
hundred pounds of dust per hour, which may cause
substantial wear to the blower. These blowers
also require frequent dynamic balancing.
Structural Design
The gage of metal used to construct the bag-
house walls, hoppers, and so forth must be
adequate, and sufficient bracing must be pro-
vided to withstand the loads exerted. A pres-
sure differential of 8 inches water column
represents approximately 42 pounds per square
foot. The total air pressure exerted on a side
panel of a pullthrough baghouse may be in ex-
cess of 2 tons. Baghouses have been known to
collapse as a result of this air pressure when
inadequate bracing was provided. Pullthrough bag-
houses are more of a problem in this regard than
the pushthrough type for two reasons. First, iden-
tical baghouse structures can withstand more in-
ternal pressure than external pressure without
damage. Second, the pressure differential between
the inside and outside of the baghouse housing is
usually greater for a pullthrough installation than
for an otherwise identical pushthrough type.
Hoppers
Size
The size of the hoppers provided must be suffi-
cient to hold the collected dust until it is re-
moved for disposal. If the hopper is emptied
once per day, it must be large enough to hold the
total amount of dust collected in a full day's oper-
ation. Some reserve capacity should also be pro-
vided since the quantity of dust may vary from day
to day depending upon variations in the basic pro-
cess. If the hopper does not have adequate capac-
ity, dust already collected becomes reentrained
-------
Baghouses
133
Figure 79. Hopper discharge valves (Western Precipitation Corporation, Division of
Joy Manufacturing Co., Los Angeles, California).
increasing the total dust load on the filter cloth
and thereby the filter resistance. This is detri-
mental to the performance of the baghouse. De-
flectors are often installed to minimize or pre-
vent this reentrainment to some extent.
Slope of hopper sides
The slope of the sides of the hopper must be suf-
ficient to permit the dust to slide or flow freely.
The design must also consider the possibility of
bridging. Continuous emptying of hoppers will
help to prevent bridging of material that has a
strong tendency to do so. It will also prevent
operating difficulties with materials that tend to
become less fluid 'with time. For example, some
materials have a tendency to cake if permitted
to stand for a few hours or overnight. This is
especially true of hygroscopic materials that
absorb moisture from the air.
Gage of metal
The gage of metal required for constructing hop-
pers depends upon the size of the hopper and the
service. For small hoppers and light duty, 16-
gage metal may be used. The gage should be
increased as warranted by the size of the hopper
and the total weight of the dust to be held at any
one time. In addition, however, consideration
should be given to the fact that workers frequent-
ly hammer on the sides of hoppers to assist the
collected dust to flow freely from the discharge
gate. If materials tend to stick or cake or are
not freely flowing, some hammering on the sides
of the hoppers -will certainly result. Many hop-
pers have been badly dented as a result of rough
treatment.
Use of vibrators and rappers
A much better solution than hammering on the
sides of the hoppers is to provide mechanical
rappers or vibrators. The most frequently
used device is the electrically operated Syntron
vibrator. Air-operated vibrators are also used
extensively. A rapping device is highly desir-
able when a rotary discharge valve or screw con-
veyor is used. The rapper may be operated from
a cam attached to the shaft of the rotary valve. In
some cases the valve, rapper, and screw are all
operated from a single electric motor.
-------
134
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
Discharge
Many baghouses, especially small, simple ones,
use a slide gate at the bottom of the hopper to
control the discharge of the collected dust from
the hopper. Other valves commonly used are
shown in Figure 79. The rotary valve is usually
used on fully automatic units. The operation of
the gravity trickle valve may be affected by the
pressure in the baghouse.
MAINTENANCE
Service
Every mechanical device, no matter how well de-
signed and constructed, must be serviced peri-
odically if it is to continue to operate properly.
A baghouse, even the simplest, is no exception
to this rule. Maintenance is usually adequate
when the collected dust has sufficient economic
value. The self-interest of the operator then
requires that the equipment be kept in optimum
operating condition. In many cases, however,
baghouses are installed because local air pollu-
tion regulations require it. When the baghouse
is nonproductive, the operator has little motiva-
tion to maintain it in optimum condition; however,
this is'a foolish and shortsighted attitude. Unless
the baghouse is properly maintained, the invest-
ment, large or small, is wasted. In many cases
the additional expense required to recondition
equipment, which has long been neglected, is as
much or more than the expense of continually
maintaining the equipment in optimum condition
•would have been.
A proper maintenance program requires estab-
lishing a schedule for the various operations
that must be performed periodically. The hop-
pers should be emptied and the collected dust
disposed of at least once a day. Depending upon
the nature of the dust, the quantity collected,
and the general severity of the service, the equip-
ment should be thoroughly inspected at intervals
of a week, a month, or quarterly. Moving parts
such as the shaking mechanisms must be greased
and oiled at intervals specified by the manufac-
turer. For baghouses in daily use, all bags
should be examined at least once a week to de-
termine whether any are showing wear. Bags
having holes or rips should be replaced immedi-
ately. Frequently, trouble can be detected be-
fore it becomes fullblown. Large baghouses
benefit by the maintenance of a chart on which
the history of each bag is recorded. If bags in
one area show a history of more frequent re-
placement than those in other areas, this should
be investigated.
Bag Replacement
Some operators find it more economical to re-
place all the bags periodically before serious
trouble begins to develop. For example, one
operator in the Los Angeles area replaces all
the bags in a quarter of the baghouse every 3
months. Thus, every bag is replaced once a
year. A thorough inspection is, moreover,
made monthly. If an individual bag develops a
hole or a rip or shows any sign of wear, it is
replaced when detected. The advantage of this
maintenance schedule is that the overall cost
may be lower compared -with replacing bags
only when they fail. In this particular case,
experience with other, similar equipment in-
dicates that bag failures generally occur be-
tween 1 and 2 years after installation, with
an average life of 18 months. Thus, after a
year, frequent replacements would be required.
The labor required to replace a bag when one
bag is replaced at a time can be estimated to
be approximately 1/2 to 1 man-hour. If an
entire section (375 bags) is replaced at one
time, the greater efficiency reduces the labor
required to about 0. 086 man-hour per bag. In
either case, the cost of the bag itself is about
$10. While the labor and material cost of
group replacement is not necessarily less,
there are many other advantages. The bag-
house in this illustration serves a furnace
operated 24 hours per day, 7 days per week.
When a bag failure occurs, the baghouse must
be shut down while the bag is replaced. This
means that the furnace must shut down or a
citation will be received for excessive emis-
sions. Obviously, lost production time is ex-
pensive. When group replacement is used, ser-
vice is scheduled to coincide with furnace shut-
down for relining without loss of production.
Each operator must decide which method is best
in respect to his own operating experience, the
anticipated bag life, and the material and labor
cost. Also to be considered is whether or not
the equipment can easily be shut down when trou-
ble develops.
Replacement of one or several bags in a large bag-
house is not usually desirable though it is some-
times unavoidable if an individual bag becomes de-
fective. In this case, the resistance of the new
bags during the initial startup will be very low
compared with that of the older bags. As a result,
the filtering velocity through the new bags will be
many times in excess of the normal rate. This
could result in blinding of a new bag during the
first few minutes of operation. It would be de-
sirable to take the precaution of returning the
equipment to service gradually in such cases, but
baghouses are not normally designed and con-
structed in a manner that permits this to be done.
-------
Single-Stage Electrical Precipitators
135
Precoating
One solution to the problem of high filtering ve-
locities for new bags would be to precoat the bags
with dust to establish a cake immediately after
installation. Precoating is a very desirable pro-
cedure, and some authorities have recommended
that all bags should be precoated immediately
after each cleaning cycle. It has also been recom-
mended that compartmented baghouses have auto-
matic programming equipment so that each section,
after cleaning, is precoated before it is returned
to service. This was done in one case by instal-
ling a cyclone precleaner. The coarse dust col-
lected by the cyclone -was then automatically intro-
duced into the air stream immediately after each
cleaning cycle.
Precoating -with a relatively coarse dust is espe-
cially beneficial when a fine fume is being filtered
(Drinker and Hatch, 1954). The precoat ensures
a high efficiency immediately after the bags are
cleaned, increases the capacity of the unit, and
decreases the pressure loss. In many cases the
additional expense of equipment for automatically
precoating the bags would be repaid in additional
usable life of the filter media, improved collec-
tion efficiency, and reduced draft loss.
The design of some simple baghouses may un-
intentionally result in automatically precoating
the bags each time the unit is started. The in-
let duct usually enters the baghouse through the
dust-collecting hopper. At startup, some of the
previously collected dust in the hopper is dis-
turbed and serves as a precoat on the filter bags.
Since the collected dust is usually agglomerated
into relatively coarse particles, it is an effective
precoat material. If, however, an excessive
quantity of dust is deposited upon the filter media,
the capacity of the unit is reduced and the resis-
tance is increased unnecessarily.
SINGLE-STAGE ELECTRICAL PRECIPITATORS
Electrical precipitation is frequently called the
Cottrell process for Frederick Gardner Cottrell
(1877 to 1948), who designed and built the first
successful commercial precipitator. It is de-
fined as the use of an electrostatic field for pre-
cipitating or removing solid or liquid particles
from a gas in which the particles are carried
in suspension. The equipment used for this pro-
cess is called a precipitator or treater in the
United States. In Europe it is called an electro-
filter. A precipitator installation is shown in
Figure 80.
HISTORY OF ELECTROSTATIC PRECIPITATION
Origins of Electrostatic Principles
The first recorded reference to the phenomenon
of electrostatic attraction, which forms the basis
for the precipitating action in an electrical pre-
cipitator, is attributed (Priestley, 1958) to Thales
of Miletus about 600 B. C. He noted that a piece
of amber that has been rubbed attracts small, light
fibers. The word electricity came from elektron,
the Greek word for amber. Pliny wrote of the
attraction of chaff and other light objects to the
amber spindles of wheels in Syria.
It was not until William Gilbert published his his-
torical De Magnete in the year 1600 that serious
progress toward understanding electrical and
electrostatic phenomena commenced. Gilbert
compiled a list of "electrics, " materials posses-
sing the property of attraction -when rubbed, and
"nonelectrics, " materials not having this proper-
ty. In 1732 Stephen Gray succeeded in demon-
strating that the so-called nonelectrics could be
given an electrical charge if they were properly
insulated. Since some materials could be charged
positively and others negatively, two different
types of electricity were postulated. In 1754
John Canton demonstrated that materials could
be charged either positively or negatively, lead-
ing to the development of the single-fluid theory
of electricity proposed by Benjamin Franklin.
In 1832 Faraday proposed an atomic theory of
electricity. Faraday's theory resembled both
the one-fluid and two-fluid theories. He as-
sumed two kinds of charged particles, which
we now call protons and electrons. He as-
sumed that only the negative particles (elec-
trons) could be transferred from one body to
another.
Although the fact that charged particles at-
tract or repel each other, depending upon
whether the charges are unlike or like, had
been known for some time, it was not until
Coulomb devised a torsion balance of suffi-
cient sensitivity that the relationship between
the charge, separation, and force was deter-
mined. Coulomb demonstrated that the force
of attraction or repulsion between two static
charges is proportional to the product of the
charges and inversely proportional to the
square of the distance between them, as ex-
pressed in equation 53:
F =
DS
(53)
-------
136
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
Figure 80. An electrical precipitator control
steel-melting furnace (left) precipitator off.
Calif.).
ing the emissions from a 75-ton and a 50-ton electric-arc
(right) precipitator on (Bethlehem Steel Co., Los Angeles,
where
F
= force of attraction or repulsion be-
tween two particles, dynes
= charge on particles, s tat coulombs
D = dielectric constant of medium be-
tween the particles, dimensionless
S = distance between the particles, cm.
In a vacuum, for which the dielectric constant
D = 1, if the force is 1 dyne and the distance
between the (equal) charges is 1 centimeter,
the fundamental electrostatic unit of charge is
defined. Called a statcoulomb, it is the charge
associated with approximately 2. 08 x 10' elec-
trons .
The forces exerted by electrical charges are
dependent upon the medium through which they
are exerted. Thus, the force as defined by
Coulomb's law depends upon D, the dielectric
constant of the medium. Values of the dielec-
tric constant for a number of common materials
are given in Table 42. The dielectric constant
may be taken, with negligible error, as unity
for air at normal temperature and pressure.
In order to explain the phenomenon of attrac-
tion and repulsion between charges, a hypo-
thetical electric field is postulated. The
strength of an electric field at any point may
be expressed as the quotient of the force
exerted on a test charge placed at that point
divided by the magnitude of the charge. It
must be assumed, of course, that introducing
a charge into an electric field does not alter
the field, which is a reasonable assumption
only if the charge is very small compared with
the strength of the field. Field strength may
also be expressed as the potential difference
divided by the distance. Equation 54, defines
the strength of a uniform electric field:
F
q
V
(54)
-------
Single-Stage Electrical Precipitators
137
where
E = field strength or electrostatic potential
gradient, statvolt/cm
V = electrostatic potential difference, statvolt.
o
Table 42. DIELECTRIC CONSTANTS FOR SOME
COMMON MATERIALS
Material 1 Dielectric constant3
Air
Alumina
Ammonium chloride
Calcium carbonate
Dolomite
Ferrous oxide
Glass (pyrex)
Quartz (fused)
Sodium chloride
Steam
Sulfur
Titanium dioxide
Water
1.
4.
7
6.
6.
14.
3.
3.
6.
1.
4
14
80
0006
50
14
80
20
80
75
12
01
to
to
to
to
to
8.
8
6
4.
110
40
10
aThese values vary with the temperature,
humidity, pressure, and electrical fre-
quency at •which measured.
Early Experiments With Electrostatics on Air Contaminants
In 1824, Hohlfeld performed an experiment in which
he succeeded in clearing the air in a jar of fog by
means of an electrified point. Guitard performed
a similar experiment in 1850 in -which tobacco
smoke was cleared from the air in a glass cylinder
9 inches in diameter by 18 inches long. These ex-
periments were forgotten until Sir Oliver Lodge
uncovered them in 1905, more than 20 years after
he had independently demonstrated the same phe-
nomenon. Information in this field was also pub-
lished by Gaugain in 1862 on the disruptive dis-
charge between concentric cylindrical electrodes,
and by Nahrwold, who, in 1878, found that the
electric discharge from a sharp point in a tin
cylinder greatly increased the rate of settling or
collection of atmospheric dust. To make the col-
lected particles adhere, he coated the walls of
the cylinder with glycerin (White, 1957).
The first attempt to use the principles of elec-
trical precipitation commercially -was made by
Walker and Hutchings at a lead smelter works
at Baggilt, North Wales, in 1885. They were
inspired by the early work of Sir Oliver Lodge
in this field. This first attempt was not success-
ful, partly because lead fume is one of the most
difficult materials to collect by electrical pre-
cipitation and partly because they -were unable
to provide an adequate power supply with their
crude equipment (White, 1957).
Development of the First Successful Precipitator
The first successful commercial use of electri-
cal precipitation was developed by Cottrell in
1907 (Cameron, 1952). Cottrell, while an in-
structor at the University of California at
Berkeley, was approached by the management of
the recently constructed Du Pont Explosives and
Acids Manufacturing Plant near Pinole, California,
about 12 miles north of Berkeley on San Pablo Bay.
This plant was using the then new Mannheim pro-
cess or "contact" method in place of the chamber
process to manufacture sulfuric acid. In the
contact process, sulfur dioxide and oxygen are
passed through an iron oxide catalyst to form
sulfur trioxide from which the sulfuric acid is
made. Difficulty was experienced owing to ar-
senic, which -was poisoning the catalyst. Cottrell
first attempted a solution to the problem by
means of collecting the acid mist with a labora-
tory model centrifuge. Although the centrifuge
principle was moderately successful in the lab-
oratory, the first pilot plant model tried at
Pinole was a failure. Before Cottrell was able
to proceed further -with this -work, all his notes
and models -were destroyed in the fire that ac-
companied the San Francisco earthquake of 1906.
Discouraged but undaunted, Cottrell rejected an
appointment to head the Chemistry Department
at the Texas Agricultural and Mechanical Col-
lege in order to follow up an idea of collecting
the acid mist by electrical precipitation.
After demonstrating that electrical precipitation
•would collect smoke, Cottrell made a small con-
tact acid plant and passed the sulfuric acid mist
into a round glass jar. Inside the jar was a cyl-
inder of wire screening around which was wrapped
several turns of asbestos -wrapped sewing twine.
The walls of the jar became the collecting elec-
trode. Three factors contributed to the ultimate
success of his first electrostatic precipitator.
The first was the use of a pubescent electrode.
He also discovered that the use of negative
polarity resulted in a more stable and effi-
cient operation. The third factor was his use
of rectified alternating current. For this pur-
pose, he developed a mechanical rectifier. With
financial backing from friends, Cottrell organ-
ized two corporations and constructed a pilot
collector that handled 100 to 200 cubic feet of
gas per minute. This pilot unit was installed
at Pinole where it operated satisfactorily, han-
dling a gas current representing about 3 tons of
sulfuric acid per day and consuming less than
1/3 kilowatt. The apparatus is shown in Figure
81, taken from Cottrell's 1908 patent.
-------
138
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
Figure 81. Illustration from Cottrell's first
(1908) electrostatic precipitation patent, No.
895,729 (Research Cottrell, Bound Book, N. J.).
1. High efficiency can be attained. Efficiency
may exceed 99 percent in some cases.
2. Very small particles can be collected. There
is no theoretical lower limit to the size of a
particle that can be collected.
3. Dusts may be collected dry for recovery of
valuable: material.
4. Pressure and temperature drops are small.
The pressure drop through an electrical
precipitator seldom exceeds 0. 5 inch verti-
cal water column.
5. Precipitators are normally designed to oper-
ate continuously with little maintenance over
long periods of time.
6. There are very few, if any, moving parts,
which tends to reduce the maintenance
required.
7. Precipitators can be used at high tempera-
tures. Temperatures up to about 700 °F are
normal. Special designs have been used for
temperatures as high as 1,300°F, but ordi-
narily the temperature does not exceed
1, 000°F (Sproull, 1951).
Improvements in Design, and Acceptance by
Industry
After Cottrell proved that electrical precipita-
tion could be applied successfully to the collec-
tion of industrial air contaminants, the use of
electrical precipitation expanded into many di-
verse fields. Table 43 lists some of the pioneer
installations.
Table 44 summarizes the extent of the use of
electrical precipitation in the United States only
50 years after Cottrell first succeeded in demon-
strating the practicality of this principle for the
control of industrial air contaminants. Table 45
lists data that typify installations of modern elec-
trical precipitators. Obviously, precipitators
serve for a variety of industrial applications,
sizes, dust concentrations, particle sizes, and
efficiencies.
ADVANTAGES AND DISADVANTAGES OF ELECTRICAL
PRECIPITATION
The use of electrical precipitators for the collec-
tion of air contaminants has grown because of
many inherent advantages, some of which are
now listed.
8. Precipitators can be used to collect acid and
tar mists, which are difficult, if not impos-
sible, to collect by other methods.
9. Extremely corrosive materials can be collec-
ted with special construction.
10. Collection efficiency may be adjusted to suit
the application by increasing the unit size.
11. Very large gas flow rates can be handled.
12. The power requirements for flow handled are
low. For example, the actual power required
to clean 500, 000 cubic feet of gas per minute
at 95 percent efficiency, including the draft
loss, is only about 65 kilowatts (White, 1953).
Electrical precipitators are by no means a pan-
acea for air pollution problems. In many cases,
disadvantages far outweigh the advantages. Some
of the drawbacks are now listed.
1. Initial cost is high. In most cases the invest-
ment is greater than that required for any
other form of air pollution control.
-------
Single-Stage Electrical Precipitators
139
2. Precipitators are not easily adaptable to vari-
able conditions. Automatic voltage control
helps to a great extent, but precipitators are
most efficient -when operating conditions re-
main constant.
3. Some materials are extremely difficult to col-
lect in an electrical precipitator because of
extremely high or low resistivity or other
causes. In some cases, this factor alone
makes the use of electrical precipitation un-
economical, if not physically impossible.
4. Space requirements may sometimes be great-
er than those for a baghouse. In general, this
is true only when high collection efficiency is
required for materials difficult to collect by
precipitation.
5. Electrical precipitation is not applicable to
the removal of materials in the gaseous phase.
6. The use of a precleaner, generally of the cy-
clonic type, may be required to reduce the
dust load on a precipitator.
7. Special precautions are required to safeguard
personnel from the high voltage.
Table 43. PIONEER PRECIPITATOR
INSTALLATIONS, 1907 to 1920
(White, 1957)
Table 44. SUMMARY OF UNITED STATES
PRECIPITATOR INSTALLATIONS IN MAJOR
FIELDS OF APPLICATION,
1907 to 1957 (White, 1957)
Application
Sulfuric acid mist from contact acid
plant, 200 cfm
Smelter, zinc and lead fumes,
300, 000 cfm
Cement kiln d ist, 1 million cfm
Copper converter (lead fume),
200, 000 cfm
Gold and silver recovery from
furnace treatment of electrolytic
copper slimes
Absorption of chlorine gas by
powdered lime followed by precipi-
tator collection
Dwight-Lloyd sintering machine
lead fume, 20, 000 cfm
Tar removal from illuminating gas,
25,000 cfm
Cleaning ventilating air in factory,
air not recirculated, 55, 000 cfm
Paper pulp recovery of alkali salts
from waste liquor evaporated gases,
90, 000 cfm
Central gas cleaning plant,
2 million cfm
Date
1907
1910
1912
1912
1913
1913
1914
1915
1915
1916
1919
Location
Pinole, Calif.
Shasta Co. , Calif. ,
Balaklala
Riverside, Calif.
Garfield, Utah,
American Smelting and
Refining Co.
Perth Amboy, N. J. ,
Raritan Copper Works
Niagara Falls, N. Y. ,
Hooker Electro-
Chemical
Tooele, Utah, Inter-
national Smelting and
Refining Co.
Portland, Oregon
New Haven, Conn. ,
Winchester Arms
Canada
Anaconda, Mont. ,
Anaconda Copper
Smelting Co.
Application
Electrical power industry:
(fly ash)
Metallurgical :
Copper, lead, and zinc
Steel industry
Aluminum smelters
Cement industry:
Paper mills:
Chemical industry:
Detarring of fuel gases:
Carbon black:
Total
First
installation
1933
1910
1919
1949
1911
1916
1907
1915
1926
Number of
precipitators
730
200
312
88
215
160
500
600
50
2,855
Gas flow,
million cfm
15
22. 5
5.9
157
43. 4
29
18
9
4. 5
3. 3
264. 2
The decision -whether to use an electrical precipi-
tator, a baghouse, or some other type of collector
must be made after considering all the following
factors:
1. Initial investment;
Z. maintenance, including the cost of power to
operate the device;
3. space requirements;
4. collection efficiency, which must be evalu-
ated in terms of the value of the collected
material or restrictions placed on the dis-
charge of air contaminants by local regula-
tions, or both (sometimes good public rela-
tions require an even higher collection
efficiency than can be justified solely on
the basis of economics).
The cost of providing high efficiency is illus-
trated by the fact that the cost nearly doubles
when electrical precipitator collection efficiency
is increased from 80 to 96 percent and almost
triples from 80 to 99 percent.
MECHANISMS INVOLVED IN ELECTRICAL PRECIPITATION
The process of electrostatic precipitation con-
sists of a number of elements or mechanisms,
which are now listed.
1. Gas ions are formed by means of high-volt-
age corona discharge.
Z. The solid or liquid particles are charged by
bombardment by the gaseous ions or electrons.
3. The electrostatic field causes the charged
particles to migrate to a collecting electrode
of opposite polarity.
4. The charge on a particle must be neutralized
by the collecting electrode.
-------
140
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
w
£
O
1 — I
H
<;
u
i— t
j
&
OH
PH
M
O
I
Si
M .H
1 •"
iJ 3
H 3
§*
S3
<:
Q
J
-<
H
ni
H
XI
r
^
01
g t;
4J ^;
CO
O
U
£
U
3 S^
01 '[j
£ £
"u
«
°
X 1 1 =L
00 "-rj i— 1 O
•^ 1^ — l
OJ m o
5 O
C
o
• H
•M CO
rt . -M
h a) <.
CO fi 5? CO
3 OJ 5 C
-^ U ro -H
§ " &
u °*J
a a"
C ^0 r
CU 5 o
C™1 H
-
bo
g
a 2 S
°r
1*4
Application
£>,
^i
CO
T3
I-H
U^
00
0
0
o
o'
oo
ON
0
m
JN
in
s.
O
LT)
r\J
LO
o
o
o'
o
o
o
o
r~
tM
o
o
o
o"
m
t~
o
o
o
o
0
'
Ly ash-pulverized coal
k
.H
£
JH
H
o
in
rt' _ «, m
O O O O
0 0
sD LT)
r-H O 1— 1 f\J
+
QS Q*» Qs 0s
O^ O O O
0 0 O 0
in in in m
CO 0s 0s 0s
in o in
t*~ ^o r*~
2 2 2 §
m o in
CO ~H CO
"".
in in in o
— i -H rsj
O O O O
O CM
m o
o — « m o
o o m o
in in (N m
f- ro *— ' ""•*
o o o o
o m o o
o (NJ m o
ft ~H ^H
a
o o o o
"< O O 0
3 0 0 0
p o" o" o"
2
o o o o
o o o o
o o o o
o o o o
O O CM O
CO
bn
fired boilers
ust from kilns
ust from dryers
.ill ventilation
leaning blast furnace j
Q Q 2 0
T) 1*J
^ C ^
o "S £ ^
DH S
m
r»-
-
O
in
o'
o\
o
o
in
0
o
o
,_H
o
o
o'
o
in
~*
0
0
o
^
o
o
o
o*
o
fNJ
0
o
o
o
o
I
CU
^
for fuel
ollecting tars from co
0
n
o
-1
a-
o
o
m
m
O
in
o
o
o
o
[^
0
o
o
fl
o
o
o
in"
0
o
0
0
o
1
tu
a.
oven gases
ollecting fume from o;
hearth and electric
U
o
o
rj
00
O-
o
o
a-
o
o
o
4-J
O
^
O
m
o
in
o
o'
o
"~*
r-Z
O
O
m
^
R
0
• l-t
^J
1
—i ni
O -H
0 ^
m" CD
u}
to
rn CJ
-2j •*-*
furnaces
unic from kilns, roas
sintering machines,
aluminum pot lines, e
cid mist
Ut <
co
3
2 CO
E •-'
•S |
o ^
^
o
in
rj
O
-1
in
o
o
o
o
£
CM
O
O
m
o'
o
in
m
o
m
r~
(M
O
O
O
o"
O _i
t-J «J
^ ° '3
£ o 3
CO O JS
^ *•
CO 0
"i in s5 °
LJ
£ o o
CO O O
3 in o
C pg in
„
fVJ
cid mist
leaning hydrogen, CO
< O
rt
U
1
in in in
00 C"~ ^
4-> 4J 4->
O O O
ro in ro
-H m m
(VI fM
o o o
—I O O
o -H m
o' o' o'
o o o
o m o
NO in r-
O 00
o o o
o mo
in f"> ~-i
o o o
o o o
in o o
t^ o o
in o
i— i r\j
o o o
o o o
o o o
in o o
rg o m
i ^
^f 0)
0 >
> ») .
,-j " QJ
£ 4J 4J
2 3 ^ S
- s3 I
w CX rt
^
^J
>
(X
irt
CO
M
P
bo
m
tu
a
&1
^
""d
w
cu
u
d
0)
"y
^H
<-M
D
.
tU
^
"§
0)
b
o
h
CJ
00
?L
d
M
s
?
^
£
H
cti
S.S
CO CO
C M
Is
!°
P-. §
-------
Single-Stage Electrical Precipitators
141
5. Reentrainment of the collected particles must
be prevented.
6. The collected particles must be transferred
from the collecting electrode to storage for
subsequent disposal.
The accomplishment of these functions by an elec-
trical precipitator has required the development of
many specialized techniques for specific materials,
though the broad principles remain as enumerated.
DIVERSE APPLICATIONS OF ELECTRICAL PRECIPITATION
Table 46 illustrates the broad spectrum of ma-
terials that are collected by electrical precipita-
tion and the range of dust concentrations that may
be encountered in practice.
Dispersoids in gases may be a one-component
system, but two or more components are usual in
industrial air pollution control. The dispersed
phase may be a liquid, as in clouds, mists, or in
sprays, or may be a solid, as in a dust cloud or
metallurgical fume. Dispersed systems include
dusts, fogs, clouds, mists, hazes, fumes, or
smokes.
In general, the size of dust particles varies from
5 to 100 microns and fumes vary from 0. 1 to 5
microns. Table 47 lists typical particle sizes
encountered in industrial dusts and fumes.
Construction Details of Electrical Precipitators
Essential features of precipitator design, exem-
plified in Figure 82, include the following ele-
ments: Rappers, shell, cable from rectifier,
support frame, corona wires, collecting plates,
gas inlet, hoppers, wire-tensioning weights, and
hopper baffles.
Discharge electrodes
The discharge electrodes provide the corona,
without which the precipitator cannot function.
These may be round wire, square twisted rods,
ribbons, barbed -wire, and so forth. Steel al-
loys are commonly used, but other materials
that have been used include stainless steel,
fine silver, nichrome, aluminum, copper,
hastelloy, lead-covered steel wire, and titani-
um alloy. While the choice of material is usu-
ally dictated by the requirements of corrosion
resistance, the physical configuration must be
determined to meet the electrical characteris-
tics requirements. When round wires are
used, the diameter is usually about 3/32 inch,
though it may vary from about 1/16 to 1/8 inch.
Conventionally, 3/16-inch-square twisted wire
has been used for precipitators serving catalytic
cracking units. The use of barbs and various
special shapes is strongly advocated by some
authorities, but others equally competent dispute
these claims, pointing out that no decided ad-
vantage has ever been established for the use of
special discharge electrodes.
Collecting electrodes
The variety of collecting electrodes available
is even more diverse. Materials of construc-
tion and special shapes appear to be limited
only by the imagination of the designer. While
many of these special shapes have important
advantages, the use of smooth plates, with
fins to strengthen them and produce quiescent
zones, has become most common in recent
years. The preference between one special
shape and another frequently becomes one of
conjecture. Figure 83 illustrates some of the
special collecting electrode configurations
marketed. These include perforated or ex-
panded plates, rod curtains, and various hol-
low electrodes with pocket arrangements on
the outside surfaces for conducting the pre-
cipitated dust to the hopper in quiescent gas
zones. Concrete plates were used at one time
but "were abandoned about 1930 because of
excessive cost and weight. Smooth transite
plates are used occasionally because of their
excellent corrosion resistance. These are,
however, for unusual cases, because of the
severe reentrainment problem. For fly ash,
perforated or expanded metal plates provide
a multiplicity of closely spaced holes that
hold the ash while end baffles on the plates
shield the perforated surfaces from the di-
rect scouring action of the gas. Several vari-
ations of the V electrode have also been used
and have similar characteristics. The hollow
or pocket-type electrodes are attractive in
principle, but in practice, a large proportion
of the dust actually falls on the outside of the
plates. Furthermore, much of the dust col-
lected in the upper openings actually escapes
to the outside through the lower openings be-
cause of the piston action of the falling dust
(White, 1953).
Tubular collecting electrodes
Plate-type precipitators are usually preferred
because they can handle a larger volume of gas
in a smaller space for less investment than the
tube type. The tube type, often called "pipe
type, " lends itself more readily, however, to
wet collection and is, therefore, preferred for
acid mists and tars. In the case of detarring
precipitators, the tar collects on the inside
walls of the tubes and runs by gravity to col-
-------
142
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
Table 46. CONCENTRATIONS OF SUSPENDED MATTER
IN COMMERCIAL GASES IN TYPICAL ELECTRICAL
PRECIPITATOR INSTALLATIONS (Kirk and Othmer, 1947)
Acid mist--sulfuric acid (contact plant)
After roaster
Multiple-hearth roaster--zinc and pyrites
Flash roaster — zinc blend
After absorbers—tail gases
Acid mist—phosphoric acid — from burning
phosphorus (basis 100% H3PO4)
Assay offices and mines--ventilating gases
from furnaces and assay operations
Carbon black
From cracking natural gas
From oil cracking
Carbureted "water gas
Dry-tar basis
Wet-tar basis
Catalytic cracking units--oil
Atmospheric-pressure units--after me-
chanical collector
Natural catalyst
Synthetic catalyst
High-pressure units —after mechanical
collector
Natural catalyst
Synthetic catalys;
Cement-kiln gases (wet process)—dust con-
centrate entering stack
Wet-ga? basis
Dry-gas basis
Coke-fired producer gas
Coke-oven gas
Ahead of exhausters, dry-tar basis
After positive-displacement exhausters,
wet-tar basis
After centrifugal exhauster, wet-tar basis
Fly ash from boilers burning pulverized
soft coal
Gypsum-plant gas
From rotary calciners, wet-gas basis
From dryers, wet-gas basis
From gypsum kettles, dry-gas basis
Incinerators burning dry sewage sludge
Silica-rock treatment
Oil-fired rotary dryer
Preheater gases
Ventilating system
Tin smelting
Reverberatory furnaces
Calcining tin ores—rotary kilns
Zinc sintering machine — straight and
chloridized roast
Zinc-ore roasting
Flash roaster
Multiple-hearth roaster
Zinc oxide — Waelz plant
Concentrations,
grains/ft^
of gas, STP
1. 08
0.00475
0.722
48. 3
5. 1
19
0. 765
1. 08
19.45
16.5
7.19
4. 69
2.62
3. 34
0. 03
4. 51
3.14
66
32.82
64.52
6.
72
17
42
65
9. 37
2, 20
1. 44
0.311
3.92
3. 82
12.65
to 5.80
to 0.05550
to 2.310
to 66. 2
0.0028 to 0.0515
to 17
to 40
to
to
1. 590
2.26
to 85.60
to 52. 9
to 22.75
to 94. 60
to 3.80
to 4. 68
to 0.06
to 4.88
to 4. 58
to
to
3. 74
5
to 48. 07
to 26. 98
to 4.35
to 23. 50
to 15. 80
to 26. 20
to 3.12
to 4. 59
to 1.908
to 45. 05
to 7. 07
to 28. 62
-------
Single-Stage Electrical Prccipitators
143
Table 47. AVERAGE DIAMETER
OF PARTICLES IN VARIOUS INDUSTRIAL
OPERATIONS TYPICAL OF ELECTRICAL
PRECIPITATOR INSTALLATIONS
(Kirk and Othmer, 1947)
Particle
Coal dust
Powdered-coal ash
Tobacco smoke (tar mist)
Cement dust
Talc dust
Silica dust
Sprayed-zinc dust
Flour -mill dust
Alkali fume
Ammonium chloride fume
Zinc oxide fume
Condensed-zinc dust
Pigments
Sprayed dried milk
Average
diamster, |a
10
1 to 150
0.25
5 to 100
10
5
15
15
1 to 5
0. 1 to 1
0. 05
2
0. 2 to 5
0. 1 to 3
lecting troughs below. In the case of acid mist
collectors, a continuous film of water is main-
tained on the tube wall by means of weirs. The
tube-type precipitator is also commonly used
in the steel industry to clean combustible gas
from blast furnaces to prevent fouling of the gas
burners.
Removal of dust from collecting electrodes
Once the dust or fume has been precipitated
on the collecting electrode or plate, it must be
removed to a hopper or storage depository. In
order to do this, rappers are commonly em-
ployed. The plates are struck sharp, hammer-
like blows to dislodge the collected dust, which
then falls by gravity into the collecting hopper.
Reentrainment of a portion of the dust at this
point must be held to a minimum. Frequently,
satisfactory collection efficiency is completely
negated by improperly operated or adjusted
rappers.
For fly ash precipitation, the dust buildup on
the collecting plates should be allowed to reach
about 1/4 to 1/2 inch before it is rapped off.
Discharge electrode rappers are necessary
when treating ashes predominantly composed
of fine particles less than 10 microns in di-
ameter (White, 1953).
A satisfactory rapping system is characterized
by a high degree of reliability, by ability to
maintain uniform and closely controlled raps
over long periods of time without attention, and
by flexible and easily controlled rapping inten-
sity. The usual practice is to rap sufficiently
to dislodge all the dust layer at one time. Stack
puffs are prevented by rapping only a small
fraction of the electrodes at a time and using
proper sequence.
Rapping mechanisms include mechanical (elec-
tric motor operated) and pneumatic or air oper-
ated. Most new installations, however, now use
magnetic solenoid-operated rappers, which can
be adjusted more accurately to control both fre-
quency and intensity of the raps.
Rapping is usually done in zones, the number
and location of rappers being dictated by the
size and configuration of the precipitator. Rap-
pers are always adjusted in the field under oper-
ating conditions. Factors that influence the
intensity, frequency, and number of blows re-
quired per cycle include:
1. Agglomerating characteristics of the dust,
2. the rate at which the dust is accumulated on
the collecting electrode,
3. the tendency of the dust to become reen-
trained,
4. the effect of the accumulated dust on the
electrical operation of the precipitator, and
5. the cycle of operation of the equipment being
served.
In some cases where reentrainment is a severe
problem, precipitators may be designed so that
a number of sections may be closed in turn dur-
ing rapping by means of dampers. While this
may reduce the reentrainment loss during rap-
ping, the usual practice is to rap during normal
operation. When the equipment being served
operates in cycles, it may be possible to bypass
the precipitator for rapping during periods when
little or no air contaminants are being vented.
In some unusual cases it may be necessary to
deenergize the precipitator in order to obtain
effective removal of the collected dust during
rapping. In other cases deenergizing may suf-
fice to permit the collected dust to fall to the
hopper by its own weight -without the need to rap.
With tube-type precipitator s, when operated wet,
it is not necessary to use rapping. Some plate-
type precipitators are also operated without
rappers for various reasons. For example,
•when transite plates are used, rapping is unde-
sirable because these plates do not have ade-
quate mechanical strength to withstand repeated
blows. Hence, periodic water sprays are usu-
ally used in this case to wash the collected dust
off the plates. Cycling the water sprays prop-
erly makes possible keeping the plates wet be-
tween flushings, which is a great aid in improv-
-------
144
AIR POLLUTION CONTROL EQUIPMENT FOR P ARTICULATE MATTER
II RAPPERS
HT CABLE FROM
RECTIFIER
HOPPERS
Wl RE TENS I ON ING
WEIGHTS
HOPPER BAFFLES
Figure 82. Basic structure of a typical precipitator (Western Precipi-
tation, Division of Joy Manufacturing Co., Los Angeles, Calif.).
ing collection efficiency by minimizing reen-
trainment. The -water sprays, when used, tem-
porarily disrupt the electrical operation, so
that this method is employed only in unusual cases.
ment, poured concrete, carbon, tile, aluminum,
wood, wrought iron, alloys of steel, rubber-
coated steel, and vinyl, or other plastic coat-
ings on steel or other supporting structures.
Precipitator shells and hoppers
Precipitator shells may be made of a variety of
materials. These include ordinary mild steel,
lead-coated steel, 'acid-resisting brick and ce-
The collected dust is ordinarily stored in hop-
pers below the collecting electrodes for peri-
odic or continuous disposal. Adequate storage
must be provided to accommodate the collected
dust between hopper cleanouts. If the dust
-------
Single-Stage Electrical Precipitators
145
•••••••*•
GAS FLOW
GAS FLOW
ROD CURTAIN
AAAAA
ZIG-ZAG PLATE
The values of the potential difference used in single-
stage electrical precipitation are usually from
20, 000 to 100, 000 volts. Since unidirectional cur-
rent is required for electrical precipitation, it is
necessary to transform the available power to a
high voltage and then rectify the high voltage.
Early precipitators used mechanical rectification
exclusively, and many of them are still in use.
[ [3 3 [
GAS FLOW
C C3 3 [
COMMON PLATE
GAS FLOW
DUAL PLATES
VERTICAL GAS
FLOW PLATES
Figure 83. Some special collecting electrodes
used in electrical precipitators (Western Pre-
cipitation, Division of Joy Manufacturing Co.,
Los Angeles, Cal i f.).
Tube-type rectifiers
Electronic tube rectifiers were first used in elec-
trical precipitation around 1920. The early tubes
were unsatisfactory because of their short, un-
certain life. These tubes have now been devel-
oped to the point where the average life in elec-
trical precipitation service is in excess of
20, 000 hours. In some cases over 30, 000 hours
of service have been obtained.
Filament voltage adjustment on tube rectifiers
is a critical factor in tube life. As a rough
guide, increasing filament voltage by 5 percent
reduces filament life by a factor of 2 while re-
ducing voltage by 5 percent increases filament
life by a factor of 2. Thus, it is general prac-
tice to operate tube filaments in precipitator
service at 5 to 10 percent below rated values,
that is, at 18 to 19 volts rather than 20 volts,
which is the rated value for most precipitation
rectifier tubes (White, 1953).
Solid-state rectifiers
builds up to too high a level in the hopper, there
is danger of reentrainment or shorting the dis-
charge electrodes, or both. The sides of the
hoppers must have adequate slope to prevent
bridging and hangup. Vibrators maybe re-
quired if the dust or fume does not move free-
ly. Discharge from the hoppers may be by
means of slide gates, motor-operated rotary
valves, or screw conveyors. The latter two
are suitable for continuous operation.
High Voltage for Successful Operation
In order to achieve maximum collection effi-
ciency, electrical precipitators are operated as
close to the sparking voltage as practicable with-
out excessive sparking. The following gives the
order of magnitude of current and field strength
usually encountered in practice (Perry, 1950).
3 4
i = 3 x 10 to 3 x 10 statampere/cm (0. 03
to 0. 3 milliampere/ft)
E = 5 to 20 statvolt/cm (3.8 to 15.3 kilovolts/in.)
The development of solid-state rectifiers has
made mechanical rectification obsolete. Se-
lenium rectifiers provide reliable service with
long life; however, they are subject to damage
from excessively high temperatures. Silicon
rectifiers, which are even newer in precipita-
tion service, do not have the shortcoming of
being subject to temperature damage. Al-
though the solid-state rectifiers are somewhat
more expensive than the electronic-tube type,
their use is justified on the basis of a long,
useful life and troublefree operation. Life
expectancy of selenium rectifiers is estimated
to be about 100, 000 hours. Silicon rectifiers,
which are hermetically sealed, appear to have
unlimited life (Peach, 1959).
Effects of Wove Form
Rectifier connections are either half wave or
full wave. The half wave connection is pre-
ferred in some cases, since it permits a
greater degree of precipitator sectionaliza-
tion with a given number of electrical sets. In
-------
146
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
large precipitators, the corona electrodes are
always subdivided into several groups or sec-
tions, and the individual sections separately
energized by individual rectifier sets. This
arrangement permits each section to be operated
under optimum conditions and is necessary for
optimum performance. Although half wave con-
nection is sometimes preferred, full wave is
usually used on the outlet sections, to supply the
greater corona current demand required for these
sections.
Typical operating voltages for fly ash precipita-
tors of 8- or 9-inch plate-to-plate spacing range
between 40 and 55 kilovolts. Corona currents usu-
ally lie between 10 and 30 milliamperes per 1,000
feet of discharge wire. The average electric pow-
er supplied to the corona commonly ranges between
40 and 120 watts per 1, 000 cubic feet per minute uf
gas treated. In general, higher voltage and power
provides higher precipitator efficiency ana per-
formance (White, 1953).
Controlled Sparking Rate
Recent research has shown that, contrary to
earlier ideas, optimum collection efficiency is
usually obtained with precipitator voltages set
high enough to produce a substantial amount of
sparking (White, 1963). Some precipitators,
however, operate with practically no sparking.
The optimum degree of sparking depends upon
many factors, such as precipitator size, fume
characteristics, fume concentration, and so forth.
Maximum efficiency usually occurs from 50 to 100
sparks per minute. Figure 84 illustrates the vari-
ation of precipitator efficiency with sparking rate
for a particular combination of precipitator de-
sign and operating conditions, such as tempera-
ture, moisture content, and so forth.
100
SPARKS PER MINUTE
Figure 84. Variation of precipitator efficiency
with sparking rate for a representative fly-ash
precipitator (White, 1953).
Operating Voltage
The operating voltage of a precipitator cannot be
predicted precisely. Dust conditions have an
important bearing on the operating voltage. For
practical purposes, each manufacturer standard-
izes on a limited number of basic transformer
voltages. For example, one manufacturer (Cottr
Electrical Precipitators, 1952) designs all equip-
ment around transformer ratings of 30, 000; 60, 0
75,000; and 90, 000 volts secondary.
The use of automatic voltage control results in
increased collection efficiency from the same
size precipitator or permits the use of a smaller
precipitator for the same collection efficiency.
The precipitator voltage is maintained at the op-
timum value by a spark counter or current-sensi]
feedback circuit. Once the control has been set
for the desired spark rate, the precipitator is
held constantly at maximum efficiency regardless
of fluctuating conditions and without attention fron
an operator.
Uniform Gas Distribution
The average velocity of the gas in the duct up-
stream from a precipitator is usually 40 to 70
feet per second. In the treater, however, the
gas velocity is 2 to 8 feet per second. Because
maintaining uniform gas velocity and dust distri-
bution in the treater is important, much atten-
tion has been paid to the transition from a high
velocity in the duct to a low velocity in the pre-
cipitator. Splitters are almost universally used
in all bends or elbows in the approach to the pre-
cipitator. This also helps reduce the draft loss.
Distribution grids of many types have been devel-
oped, some of which are shown in Figure 85.
The choice of type to use in a particular instal-
lation can usually be made reliably only by means
of scale-model studies. Much of the -work in this
field is trial and error until a reasonably uni-
form gas velocity distribution is obtained in the
model. The percentage of open area has an im-
portant bearing on the performance of distribu-
tion grids. Experience may reduce the problem
to one of degree rather than of kind so that all
that need be determined is the optimum position
of the grid. In some cases, installing a perfor-
ated plate at the outlet of the precipitator has
been found as important as installing one at the
inlet. A very common type of design consists
of one or two flat perforated plates at the inlet
of the treater.
Cost of Electrical Precipitator Installations
Table 45 shows the variation of costs for elec-
trical precipitators depending upon the size,
type of dust or fume, and efficiency required.
Preliminary engineering studies and model
studies for gas distribution may add substan-
-------
Single-Stage Electrical Precipitators
147
PROTRUDING FACE
OF BASKET TOWARD
COLLECTOR INLET
POINT OF CONE TOWARD
COLLECTOR INLET
Figure 85. Examples of special perforated plate
gas distribution grids (Western Precipitation
Corporation, Division of Joy Manufacturing Co.,
Los Angeles, Calif.).
tially to costs shown. The costs of ductwork
to and from the precipitator, of foundations,
and of extending utility services to the area of
the precipitator are in addition to the installed
cost of the precipitator itself. Factors affect-
ing the cost of the precipitator include the pow-
er supply (rectifier, automatic voltage control,
number of sections individually energized, and
so forth), special plate design, electrical charac-
teristics of the dust or fume, collection efficien-
cy required, and special materials or type of
construction needed to resist corrosion or wear.
Theoretical Analysis of Precipitator Performance
A theoretical analysis of precipitator mechanisms
and performance involves two fundamental pro-
cesses, particle charging and particle migration.
Many factors affect both these mechanisms.
Particle charging
In order to derive an expression for the rate of
particle charging (White, 1951) and the maximum
charge attained by a particle, the following as-
sumptions are made:
1. The particles are spherical.
2. Particle spacing is much larger than particle
diameter.
3. The ion concentration and electric field in
the region of a particle are uniform.
These assumptions are reasonable approxima-
tions except for a few cases where the shape of
the particle may depart radically from the spheri-
cal.
A particle entering the charging field of an elec-
trical precipitator is bombarded by ions. Some
strike the particle and impart their charge to it.
As soon as a charge has been acquired by the
particle, an electric field is created that repels
similarly charged ions. Some ions continue to
strike the particle, but the rate at which they do
so continually diminishes until the charge ac-
quired by the particle is sufficient to prevent
further ions' striking it. This, then, is the lim-
iting charge that can be acquired by the particle.
The motion of gas ions in the electrostatic field
of an electrical precipitator constitutes an elec-
tric current
(55)
where
i = electrical current, statampere
j = current density, statampere/cm
2
A = area, cm .
The ion current density in the undistorted field
region outside the immediate influence of the
particle is
j = N £ M E
(56)
where
N = number of ions per cm
€ = elementary electrical charge = 4. 80
x 10"10 state oulomb
»,r • i •-,- cm/sec
M = ion mobility, :—;
statvolt/cm
The area of the ion stream that enters the
particle is determined by the total electric flux
as follows
A -
A "
kE
(57)
where
= total electric flue, statcoulombs
k = permittivity of free space, numerical-
ly equal to 1 in the cgs electrostatic
system of units.
-------
148
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
From electrostatic principles the electric flux
is found to be
/°
(
„
p h. cos
= p TT a E
p E a
a sin 9 d9
(58)
where
p = a factor defined in equation 61
n = number of elementary electrical charges
acquired by a particle
a = radius of a particle, cm.
Substituting and noting that the ion current is de-
termined by the number of elementary electronic
charges in a given time
dt
= p
N £ M E
pE a
(59)
Upon integration, the number of elementary elec-
tronic charges acquired by a particle of radius a
in time t is found to be
n = the limiting number of elementary elec-
trical charges acquired by a particle
at saturation
E = strength of charging field at the point
where the particle acquired its charge,
statvolt/cm.
For conditions normally existing in average pre-
cipitators, charging can be considered reason-
ably complete in about 0. 1 second. Since the
gas velocities are usually 2 to 8 fps, a particle
may travel only a few inches, or a foot at most,
before it has for all practical purposes attained
its limiting or saturation charge. Some charg-
ing also occurs by means of ion diffusion but
this can usually be neglected for particles larger
than about 0. 5 micron in diameter and does not
become a significant factor unless the particles
are smaller than about 0. 2 micron in diameter.
The usual practice is to assume that only ion
bombardment, charging occurs. Any error in-
troduced by this simplification is usually of less
magnitude than the effects of nonuniform gas
distribution, reentrainment, high dust resis-
tivity, rapping losses, and other practical prob-
lems that usually increase the actual losses by
a factor 2 or 3 times (White, 1953) the theoret-
ical value. The particle-charging time con-
stant t0 is defined as
1
N e M
(63)
t +
1
(60)
TT N E M
where
P =
a factor that depends upon the die-
electric constant D of the particle. The
numerical value of p ranges from a
value of 1 for materials with a dielectric
constant of 1 to 3 for some dielectric
materials, and is defined by
p = ] +
2 (D - 1)
(D + 2)
(61)
As the time t becomes large the value of the lim-
iting or saturation charge is
q = n £
s s
= p E a
(62)
where
= the limiting charge acquired by a par-
ticle, statcoulomh
and is the time required for 50 percent of the
limiting charge to be attained.
Particle migration
The force Fj (in dynes) exerted on a charged
particle in an electric field is proportional to
the charge q (in statcoulombs) on the particle
and the strength E of the electric field.
= qE
(64)
•where
E = strength of precipitating field, stat-
volt/cm.
This force accelerates the particle until the
viscous drag, or resistance of the gas in which
the particle is suspended, exactly equals the
force exerted by the electric field. Under con-
ditions normally existing in an electrical pre-
cipitator, the viscous drag FZ (in dynes) is
defined by Stokes' law
= 6 TT
(65)
-------
Single-Stage Electrical Precipitators
149
where
a = radius of particle, cm
u = viscosity of gas stream, poise
w = velocity of a particle relative to the
gas in which it is suspended, cm/sec.
Substituting the charge q acquired by the particle
2
Fn = p E E a
1 P
(66)
Since FI must be equal to F2 under equilibrium
conditions, the equations may be equated. Solv-
ing for the particle velocity
p E E a
P
6 7T U
(67)
For most common materials the dielectric con-
stant D is 2 to 8. Thus, the value of p varies
from 1. 50 to 2. 40, or the average is very near-
ly 2. The charging field E and the precipitating
field Ep are created by the same mechanism.
Tests have shown that the field strength is not
uniform, being highest in the vicinity of the dis-
charge electrode (White and Penny, 1961). It is a
common practice, however, in calculating the drift
velocity, to assume that these are approximately
equal. Making this assumption, and converting
the cgs units to those more convenient for prac-
tical application, we obtain for a particle in air
at 60°F
= 8.42 x 10
-3
dp
(68)
where
w = the particle drift velocity, ft/sec
E = the potential applied to the discharge
electrodes, KV/in.
d = the diameter of the particle, microns
p = a factor as before.
If the medium is a gas other than air or if the
air temperature departs from standard by more
than about 50°F, a multiplying factor of 0. 0178/u
must be used to correct for the effect of viscosity,
with u in centipoises.
In tests performed by White (1953) on an electri-
cal precipitator collecting fly ash from an elec-
tric steam power plant, the drift velocity was
calculated on the basis of actual measured effi-
ciency. The drift velocity was found to be con-
sistently about one-half that calculated from the
theoretical equations. The theoretical equations,
however, neglect such effects as nonuniform gas
velocity, erosion of dust, rapping losses, co-
rona quenching, high resistivity, half-wave rec-
tification, and so on.
Theoretical Efficiency
The trajectory of a particle in an electrical pre-
cipitator can be determined if the folio-wing as-
sumptions are made:
1. The strength of the precipitating field is
uniform. This is nearly true except in the
vicinity of the discharge electrode.
2. The migration or drift velocity w of the
particle is constant. This is true for a
particle -with a constant charge in a uni-
form field. Since the limiting charge is
closely approached within the first foot or
less of travel, this is a valid approxima-
tion for the conditions actually encountered
in precipitators.
3. The average forward velocity v of the par-
ticles suspended in the gas stream is uni-
form. Since precipitators almost always
operate with Reynolds numbers in the tur-
bulent range, the statistical mean velocity
of the particles may be considered uniform
in the direction of flow through the pre-
cipitator.
From the assumptions equation 69 is derived:
dv
dt
= C,
dw
~dT
= O
(69)
where
v = mean velocity of a particle in the gas
stream in the direction of gas flow,
ft/sec
w = mean velocity of a particle perpendic-
ular to the gas stream or in the direc-
tion of collecting electrodes, ft/sec.
By integration
(70)
The constants Cj and G£ depend upon dimensions
of the precipitator, the field strength, the point
at which the particle enters the electrical field,
-------
150
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
the mean velocity of the particle through the pre-
cipitator, and the electrical properties of the
particle and the particle diameter. Thus, the
trajectory is different for each particle, but it
is approximately a straight line. If all the par-
ticles have the same electrical characteristics,
and the worst case of a particle's entering mid-
way between the collecting electrodes is consid-
ered, the only variable is the particle diameter.
For each particle diameter there is some finite
length of collecting electrode required in order
to achieve theoretical 100 percent collection
efficiency. This length is defined by
L =
(71)
where
L
length of collecting electrode in di-
rection of gas flow, ft
separation of discharge and collec-
ting electrodes, ft.
The calculation of the theoretical length required
for 100 percent collection efficiency is illustrated
by the following example:
Example 19
Given:
A plate-type precipitator with 8-in. plate-to-
plate spacing and an applied voltage of 40, 000
volts. Mean gas velocity through the precipita-
tor is 5 fps and the minimum particle diameter
is 0. 5 (X.
Problem:
Find the minimum length of collecting electrode
in the direction of gas flow required for 100%
collection efficiency.
Solution:
1. Migration velocity from equation 68 using
p = 1
_
= (8.42 x 10 ) -^ (0.5) = 0.421 fps
2. Length of collecting electrode from equa-
tion 71
If charging is considered practically com-
plete in 0. 2 sec, an additional 1 ft must be
added to allow for the distance traveled by
the particle while being charged. The length
of collecting electrode in the direction of gas
flow required for 100% collection efficiency
is therefore 5 ft theoretically.
Deficiencies in Theoretical Approach to Precipitator
Efficiency
Example 19 illustrates that 100 percent collec-
tion efficiency should result theoretically from
a precipitator 5 feet long in the direction of gas
flow when the particles suspended in the gas are
0. 5 micron or larger. This particle size is
fairly typical of that encountered in practice.
The length of a precipitator is, however, gen-
erally bet-ween 8 and 24 feet in the direction of
gas flow. Yet no precipitator operates with
100 percent collection efficiency and, in fact,.
very few operate with collection efficiencies
much greater than 98 percent. The precipitator
in Example J 9, if only 5 feet long, might pos-
sibly fail to exceed 50 percent collection effi-
ciency in an actual case, depending upon the
electrical properties of the particles, opera-
tion of the rappers, and other factors not con-
sidered in the theoretical approach.
Effects of Resistivity
A dust such as carbon -with very low electrical
resistivity (Schmidt et al. ,1950) readily re-
linquishes its negative charge to the collecting
electrode and assumes a positive charge. Since
positive charges repel each other, the carbon
particle is repelled from the collecting elec-
trode into the gas stream -where it is bombarded
by negative ions and becomes negatively charged
again. The particles are thus alternately at-
tracted and repelled and so skip through the pre-
cipitator, knocking other particles, "which have
already been collected, off the collecting elec-
trode.
If the dust, for example powdered sulfur, has
a high electrical resistivity, it is unable to
give up its negative charge to the collecting
electrode. As the layer of dust builds up on
the electrode, it acts as an insulator. The po-
tential drop across this dust layer may build
up to high values, which may have an adverse
effect on the corona discharge and may set up
a secondary brush discharge at and within the
dust layer. This condition is called "back dis-
charge" or "back corona, " and may seriously
impair the performance of the precipitator.
When the dust, for example cement dust, has
medium resistivity, it can relinquish part of
-------
Single-Stage Electrical Precipitators
151
its charge to the collecting electrode. The
rate at which the charge leaks off increases as
the dust layer builds up and the potential drop
across the dust layer increases until a condi-
tion of equilibrium is achieved. Sufficient neg-
ative charge is retained by the particles to
maintain a force of attraction between the par-
ticles and the collecting electrode. When the
weight of the collected dust becomes sufficient-
ly great, particles fall off, of their own weight,
or are jarred loose when the electrodes are
rapped.
The electrical re.sistivity varies with tempera-
ture and moisture, as illustrated in Figure 86
for some representative dusts. Collection ef-
ficiency is adversely affected when the electrical
resistivity is as low as 10^ ohm-centimeter or
as high as 10 ohm-centimeter. Apparently
then, for many materials, collection efficiency
is adversely affected when the temperature is
250 to 400°F, the range in which it is normal-
ly desired to operate the precipitator. The ad-
verse effects of high resistivity may be avoided
by operating at a higher temperature, but this
is usually not desirable because of the addition-
al heat losses. Operation at lower temperatures
to the left of the peak of the resistivity curve is
frequently objectionable because of excessive
corrosion. An alternative is to increase the
moisture content or add other conditioning agents.
The addition of water vapor, acid, or other con-
ducting material increases the surface conductiv-
ity of high-resistivity dusts by adsorption on the
particle surfaces, "which reduces the apparent
electrical resistivity. Some materials used as
conditioning agents include water vapor, am-
monia, salt, acid, oil, sulfur dioxide, and tri-
ethylamine (Schmidt and Flodin, 1952).
In addition to the beneficial effects on the elec-
trical resistivity of the dust by the addition of
moisture, water vapor has a pronounced effect
on the sparking voltage in an electrical precipi-
tator. This effect is shown graphically in Figure
87, plotted from experimental data. In most
cases the effect of the moisture on the electri-
cal resistivity of the dust predominates when the
temperature is below 500°F, and the effect of
the moisture in increasing the sparking potential
predominates at temperatures above 500°F
(Sproull and Nakada, 1951).
300 400
I[«P!«I!UBES -F
Figure 86. Variation of apparent resistivity with temperature and moisture for some typical dusts and
fumes: (left) apparent resistivity of powdered lime rock used in making Portland cement; (right) ap-
parent resistivity of fume from open-hearth furnace (Sproull and Nakada, 1951).
-------
152
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
I I- ill GAP
1C 15
AIR MOISTURE, voI
20
at 1 atm
Figure 87. Sparking potential for negative pomt-
to-plane 0.5-inch gap as a function of the mois-
ture content of air at 1 atmosphere pressure for
various temperatures (Sproull and Nakada, 1951).
Although there is no doubt that the electrical prop-
erties of dusts and fumes may drastically affect
the operation of a precipitator, knowledge of
quantitative relationships in this respect is lim-
ited. It is known that performance is reduced as
the electrical resistivity becomes exceptionally
large or small. Some data have been published
on the variation of electrical resistivity with
temperature and humidity for a few dusts and
fumes. Sproull and Nakada (1951) analyze the
potential drop across a layer of collected dust.
Precipitation theory has not yet been developed
to the point where collection efficiency can be
accurately predicted •without reliance on empir-
ical data.
Methods of Reducing Reentrainment
Unless the dust collected by the precipitator can
be retained, the entire effort is -wasted. Once
it is collected by the collecting electrode, the
dust may be reentrained into the gas stream owing
to (1) low resistivity, which permits the negative
charge to leak off too rapidly and a positive
charge to be acquired; (Z) rapping; and (3) ero-
sion of the collected dust from the collecting
electrode. This may be because of nonuniform
gas velocity, which results in excessively high
velocity through some sections of the precipita-
tor or excessive turbulence.
The effects of low resistivity are not amenable to
correction. Fortunately, this problem does not
frequently arise. In the case of carbon black,
•which has too low a resistivity to permit pre-
cipitation, a practical solution has been found.
The electrical precipitator agglomerates the
particles of carbon that cannot be retained on the
collecting electrodes because of their low resis-
tivity. The agglomerated particles are collected
by a centrifugal collector that follows the pre-
cipitator.
To reduce erosion of dust from the collecting
electrodes, various special designs of elec-
trodes are used. The objective in all these de-
signs is to provide quiescent zones to prevent or
reduce erosion. The difficulty is reduced ma-
terially by good gas distribution in the precipita-
tor. The original design must take into considera-
tion the nature of the dust so that the maximum
velocity through the treater will be less than the
critical value at which erosion begins to increase
sharply. The critical velocity for any particular
dust can be determined only by actual test. Some
typical values for the gas velocity at which ero-
sion becomes significant are 2 fps for carbon
black, 8 fps for fly ash and 10 to 12 fps for ce-
ment kiln dust (Schmidt, 1949).
Dust reentrainrnent during rapping is controlled
by adjusting the rapping cycle and intensity to
minimize the degree of reentrainrnent. Rap-
ping cycles are determined experimentally after
the precipitator is placed in normal operation.
Rose and Wood (1956) analyze the theoretical col-
lection efficiency when reentrainrnent is con-
sidered to show that the equation for the loss takes
the form
Loss = C e
kt
(72)
where
C and k
t =
constants that depend upon the con-
figuration of the precipitator, prop-
erties of the dust, and many other
variables
the base for Naperian logarithms =
2.71828
the time a dust particle remains in
the precipitating field of the precipi-
tator, sec.
-------
Single-Stage Electrical Precipitators
153
Present knowledge of precipitation theory does
not permit an accurate evaluation of the con-
stants C and k. Their values must be deter-
mined empirically.
Practical Equations for Precipitator Design and
Efficiency
Empirical equations have been developed by
Anderson (1924), Walker and Coolidge (1953),
Schmidt (1928), Deutsch (1922), and others.
Deutsch published a proposed equation with a
form similar to
-wf
(73)
•where
r\ - weight fraction of dust collected
w = velocity of drift or migration of a
dust particle toward the collecting
electrode, fps
f = ratio of area of collecting electrodes
to the volume of gas passing through
the treater, (ft2/ft3/sec).
Anderson proposed an equation of the form
r\ = I - K (74)
Frequently the following modified forms are
used:
Plate-type precipitators
r\ = 1 - K
Ct
(76)
tube-type precipitators
f] - 1 - K
ZCt
(77).
K is a measure of the ease with which the dust or
fume can be precipitated, and C depends upon the
physical dimensions of the precipitator and the
voltage applied. For any particular installation
both K and C must be considered constants since
otherwise the equations are not useful. It is easy
to show that the last three equations are equivalent.
For plate-type precipitators
rj = 1 - K
Ct
C = -
t =
L
V
- 1 - K
cL
sv
where
K = an empirical constant
t = the time a dust particle remains in the
electrical field of the treater, sec.
Schmidt modified the Anderson equation to
c A
rj = 1 - K Q (75)
Q
•where
K and c = empirical constants
A = the area of the collecting electrodes
or plates, ft2
Q = the gas volume, cfs.
^here
W 2s
A = 2 L W
n = 1 - K
cA
E
Q
= a constant
s = separation or distance between dis-
charge electrode and collecting elec-
trode, ft
L = length of collecting electrode in the
direction of gas flow, ft
v = mean velocity of gas in the direction
of flow through the treater, fps
-------
154
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
W = width of collecting electrode or dimen-
sion perpendicular to direction of gas
flow, ft
A = area, ft
A = area of collecting electrode or plate,
For tube-type precipitators
2cL
rj = 1 - K
Q
A =
7T S
2 7T S L
TJ = 1 - K
Figure 88 graphically illustrates the relationship
among several common forms of these efficiency
equations. Note that in each equation there is at
least one arbitrary constant •whose value deter-
mines the efficiency. This is referred to as the
precipitation constant in the Anderson and Schmidt
= DMft velocity ft sec
= Volime of Eaj through prtcip
itator ft3 sec
- Base of Mapienln logarithms
I 1 03 a I 05
DRIFT VELOCITY (w). ft/sec
Figure 88. Relationship among various precipi
tation constants and drift velocity.
equations and the drift velocity in the Deutsch
equation. In reality, this is neither a constant
nor a true representation of the drift velocity.
Some typical values of the so-called drift or mi-
gration velocity are listed in Table 48.
Table 48. TYPICAL VALUES OF DRIFT
VELOCITY ENCOUNTERED IN PRACTICE FOR
USE WITH EFFICIENCY EQUATION
A
_B
Q
Application
Pulverized coal (fly ash)
Paper mills
Open-hearth furnace
Secondary blast furnace (80% foundry iron)
Gypsum
Hot phosphorous
Acid mist (H2SO4)
Acid -nist (TiO2)
Flash roaster
Multiple-hearth roaster
Portland cement manufacturing (wet process)
Portlant cement manufacturing (dry process)
Catalyst d ast
Gray iron cupola (iron-coke ratio =10)
velocity (w),
ft/sec
. 33 to 0. 44
0. 25
0. 19
0. 41
. 52 to 0.64
0. 09
. 19 to 0.25
0. 19 to 0. 25
0.25
0.26
0. 33 to 0. 37
0. 19 to 0.23
0. 25
0. 10 to 0. 12
The drift velocity w and precipitation constant K
are usually variables that are affected by the
electrical properties of the particles, -which in
turn, vary with temperature and humidity, and
by the applied voltage and the ionic current,
•which depend upon the temperature, humidity,
and dust load. They must also reflect the ef-
fects of reentrainment and rapping losses, as
well as nonuniform gas velocity distribution. In
general, the effects of none of these factors can
be predicted analytically -with any degree of
accuracy.
The design of electrical precipitators is today
almost entirely empirical. Designs are based
either upon previous experience with similar
processes or upon the results of pilot model
precipitator studies. Table 49 shows average
values for the major variables in precipitator
design. Precipitator manufacturers have ac-
cumulated considerable data through the years
upon which they can base the design of new
installations.
Effects of Nonuniform Gas Velocity
The importance of uniform gas velocity through
the treater cannot be overemphasized. In all
precipitator efficiency equations an increase in
the gas velocity or flow rate reduces the ef-
ficiency exponentially. Conversely, a decrease
-------
Single-Stage Electrical Precipitators
155
Table 49. TYPICAL VALUES OF SOME
DESIGN VARIABLES USED IN COMMERCIAL
ELECTRICAL PRE'CIPITATOR PRACTICE
Design variable
Plate spacing
Velocity through precipitator
Vertical height of plates
Horizontal length of plates
Applied voltage
Drift velocity w
Gas temperature
Treatment time
Draft loss
Efficiency
Corona current
Field strength
Normal range of values
8 to 11 in.
2 to 8 ft/sec
12 to 24 ft
0. 5 to 1. 0 x height
30 to 75 kv
0. 1 to 0. 7 ft/sec
up to 700°F standard
1,000°F high tempera-
ture 1,300°F special
2 to 10 sec
0. 1 to 0.5 in. WC
up to 99. 9 + % usually
90% to 98%
0.01 to 1. 0 ma/ft wire
7 to 15 kv/in.
in gas velocity or flow rate increases the effi-
ciency exponentially. For a constant volume
of gas through the precipitator, maximum ef-
ficiency is attained when the velocity is uni-
form. As the velocity increases through one
section of the precipitator, collection efficien-
cy decreases. At the same time the velocity
must decrease through other parts of the pre-
cipitator since the total flow rate remains the
same. The efficiency for the sections having
the lower velocity will increase. The increase
in efficiency through the low-velocity sections
of the precipitator can never compensate for
the loss in efficiency through the high-velocity
portions of the precipitator. This is illus-
trated by example 20:
Example 20
Given:
A horizontal-flow, single-stage electrical pre-
cipitator consisting of two ducts formed by plates
8 ft wide x 12 ft high on 10 in. centers, handling
3, 600 cfm with two grains of dust/ft3. The drift
velocity is 0. 38 fps.
Problem:
Find collection efficiency and dust loss in Ib/hr
for (1) Uniform gas velocity and (2) peak velocity
50% greater than average.
Solution:
For either case the loss is given by
- r7)(60)(60)(Q)(G)
Loss =
(7,000)
Ib/hi
For uniform gas velocity, collection efficiency
is given by
A
_E
Q
The plate area of each duct is
A = (2) (8) (12) =
The flow rate per duct is
192 ft
Q =
(3600)
(2)(60)
30 ft /sec
For uniform gas velocity
192
-0. 38
30
= 0. 912 or 91. 2%
Loss
- 0.912)(3600)(60)(2)
(7, 000)
5. 42 Ib/hr
For simplicity, assume that the velocity through
one of the ducts is 50% greater than average or
the volume is 2, 700 cfm and the volume through
the other duct is 50% less than average or 900
cfm. In an actual case where the velocity var-
ies continuously, it •would be necessary to di-
vide the precipitator into a great number of zones,
each having approximately constant velocity. The
procedure is illustrated by this simplified approach.
For the high-velocity duct
192
-0. 38
45
= 0.8025 or 80.25%
T (1-0. 8025)(2, 700)(60)(2)
Loss = (7,ooO) = 9. 15 Ib/hr
For the low-velocity duct
192
-0. 38
15
= 0. 9922 or 99. 22%
Loss
(1 - 0. 9922)(900)(60)(2)
(7,000)
= 0. 12 Ib/hr
where G = dust concentration, grains/ft .
The total loss for the two ducts with nonuniform
velocity is 9.15 + 0. 12 = 9. 27 Ib/hr. This is
71% greater than the 5. 42 Ib/hr loss with uni-
form gas velocity.
Figure 89 gives multiplying factors to correct
the loss for the effects of nonuniform gas ve-
locity. This graph was prepared by means of
-------
156
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
calculations similar to those in example 20.
In an actual case, since less than 50 percent
of the precipitator would be subjected to peak
gas velocities, the adverse effects of nonuni-
form gas distribution would be less severe.
On the other hand, in an actual precipitator,
reentrainment would be aggravated in the higher
velocity sections so that the actual losses in an
extreme case could be several times that pre-
dicted solely on the basis of velocity distribu-
tion.
3 5
12
PEAK
VE1. OCITY
16
AVERAGE
! 8
VELOCITY
2 0
Figure 89. Multiplying factors for loss from
electrical precipitator with nonumform gas
distribution. Loss equals FyLo.where LO equals
loss with uniform gas distribution.
Important Factors in the Design of a Precipitator
The following design factors are critical ele-
ments in an electrical precipitator (Schmidt
and Flodin, 1952): (1) Proportion, (2) capacity,
(3) cleaning of electrodes, (4) reliability of
components, (5) stability of electrical system,
(6) accessibility for maintenance, (7) control
of gas flow, (8) control of erosion of dust from
electrodes, and (9) power supply. This list is
not intended to be exhaustive or in order of im-
portance. All these items are interrelated,
and optimum performance cannot be achieved
if there are shortcomings in any of them. The
designer of an electrical precipitator is faced
with many decisions for which there is no clear-
cut solution.
Oftentimes, the most important factor in deter-
mining the length and width of a precipitator is
the available space. This factor also intro-
duces problems in the design of the ductwork
leading to and from the precipitator. Thus, it
may be necessary to increase the height of a
horizontal-::low precipitator because of a space
limitation on the length. Since the time in the
treater is reduced by restricting the length, an
additional increment of height is required to
compensate. Because this increases the dif-
ficulty of providing uniform gas distribution,
an additional increment of height is required
to compensate for nonuniform gas velocity dis-
tribution. The increased plate height intro-
duces additional problems in maintaining uni-
form plate-to-plate distance and in the discharge
electrode's suspension system. Optimum per-
formance requires uniform field strength
through all sections of the precipitator, which
in turn, depends upon near perfect alignment
of the electrode system. Even a small varia-
tion in spacing of discharge electrode to col-
lecting electrode can seriously reduce the
performance. Greater plate height may also
increase the dynamic instability of the discharge
electrode system, that is, it may increase the
tendency of the discharge electrodes to swing
or vibrate.
The tendency of the discharge electrodes to swing
or vibrate is overcome to some extent by guides
and heavy weights attached to the lower end of the
wires. Some sparking is desirable in a precipita-
tor, but with less than perfect alignment, the spark
ing will occur most frequently only at points where
the wire-to-plate spacing is the least, usually at
the lower edge of the plate. The difficulty is fre-
quently reduced or overcome by using shrouds on
the lower edge of the plate or on this section of
the discharge electrode.
SUMMARY AND CONCLUSIONS
Electrical precipitation is suitable for the col-
lection of a wide range of dusts and fumes. In
some cases, for example detarring, it is the
only feasible method; in other cases, it may be
the most economical choice. The design of an
electrical precipitator requires considerable
experience for successful application. The
fundamental theory of the mechanisms involved
in electrical precipitation is only partially under-
stood at present. Further research will tend to
make the design of electrical precipitators more
of a science and less of an art.
TWO-STAGE ELECTRICAL PRECIPITATORS
The Cottrell-type precipitator is usually de-
signed and custom built specifically for instal-
lations required to process large volumes of
-------
Two-Stage Electrical Precipitators
157
contaminated air. Since 1937 a somewhat dif-
ferent type has been marketed. This unit, de-
veloped by Penney (1937), is now called the low-
voltage, Penney, or more commonly, the two-
stage precipitator. It is also occasionally re-
ferred to as the air-conditioning precipitator or
"electronic air filter" (White, 1957).
The two-stage unit differs from the Cottrell
type in that the contaminated air is first passed
through a variable-strength ionizing field be-
fore being subjected to a separate uniform
field where the charged particles are collected.
Figure 90 shows the fundamental arrangement
of the active electrical components. Basic
operating principles are the same as those dis-
cussed for the Cottrell precipitator. A high-
voltage corona discharge ionizes gas molecules
that cause charging of particles passing through
the field. The charged particles then tend to
migrate toward electrically grounded or op-
positely charged surfaces where they are re-
moved from the airstream.
Most of the early applications of the two-stage
precipitators were for removal of tobacco
smoke, pollen, and similar air contaminants in
commercial air-conditioning installations. As
a result of mass production techniques, pre-
cipitators for air-conditioning installations are
now available in "building block" cells provid-
ing capacities up to a million cubic feet per
minute. Although these precipitators were de-
veloped principally for air-conditioning instal-
lations, their usefulness in the control of air
pollution soon became known. One of the first
units reported for the cleaning of process air
was for removal of ceramic overspray in pot-
tery glazing operations (Penney, 1937). Two-
stage precipitators are widely used for re-
COLLECTOR CELL
(TO COLLECT PARTICLES)
BAFFLE
(TO DISTRIBUTE
AIR UNIFORMLY)
-*
IONIZER
(TO CHARGE PARTICLES)
Figure 90. Components of standard two-stage precipitator (Westinghouse Electric
Corporation, Hyde Park, Boston, Mass.).
-------
158
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
moving oil mist created during operation of
high-speed cutting or grinding tools. Preci-
sion manufacturing and electronic assembly
areas are frequently equipped with precipita-
tors to remove small quantities of dust and
impurities from their environmental air.
Hotels, restaurants, food-processing compan-
ies, and pharmaceutical manufacturers often
use this method for cleaning circulating air.
Another installation, shown in Figure 91, is
designed to remove contaminants from the ex-
haust air of a meat-smoking operation.
a short circuit, and the precipitator' s efficiency
drops correspondingly.
Below the corona's starting voltage or critical
corona gradient, no ionization occurs and conse-
quently no charging of particles takes place. The
critical corona gradient, for round wires, is
basically a function of wire size and condition.
It may be determined by the equation
E = 30 M
s
(78)
•where
THEORETICAL ASPECTS
Theory of Dust Separation
The physics of dust separation in a two-stage pre-
cipitator may best be understood by examining the
stages separately. The function of the ionizing
stage is to induce an electrical charge upon the
particles in the airstream. When an electrical
potential is applied between a wire and a grounded
strut, as shown in Figure 92, an electric field
is created that varies from a high strength near
the •wire to a low at the strut. When the potential
is increased to the "critical corona voltage, "
local ionization of the airstream near the •wire
occurs and a blue corona is formed. Arcing or
"sparkover" results if the voltage is further in-
creased to a point -where total ionization of the
air between the electrodes occurs. This effects
E = critical corona gradient, kv/cm
M = roughness factor, usually between 0. 6
and 0. 9
r = wire radius, cm.
The required potential may then be determined by
V = E r In - -
s s 3 r
where
V = coiona starting voltage, kv
s = wire-to-strut spacing, cm.
(79)
Figure 91. Two-Stage precipitator controlling smokehouse emissions: (left) precipitator on, (right)
precipitator off (Luer Packing Co., Vernon, Calif.).
-------
Two-Stage Electrical Precipitators
159
ELECTROSTATIC FIELD
DUST PARTICLES
DISCHARGE ELECTRODES ^g J™J
or saturation charge, ns, for large values of
time (White, 1951).
Equation 80 applies to particles greater than 0. 5
micron in diameter, where charging is due pri-
marily to ion bombardment. Charging by ion
diffusion predominates for particles of about 0. 1
micron and smaller in diameter and requires a
somewhat different evaluation. Normally, neg-
lect of the ultrafine particles in determining
charging time introduces no significant errors
because these particles represent a small weight
fraction of the material being treated.
Figure 92. Schematic representation of two-stage
precipitator principle (Perry, 1950).
Particle Charging
The degree of electrical saturation of the dis-
persed particles may be given, for a spherical
conducting particle, by
3E a
(80)
t+
TT N k £
o
where
n = number of elementary electric charges
acquired by a particle
E = electric field strength, stat volts/cm
a = particle radius, cm
t = time interval that particle is exposed to
charging field, sec
N = ion density in charging zone, ions/cm
o
€ = charge on electron, 4. 8 x 10 stat-
coulomb
k = ion mobility, cm /sec-stat volt.
All units of this equation are expressed in the elec-
trostatic centimeter-gram-second system. White
has given the term
7T N
the notation of t ,
o
the particle-charging time constant, and states
that it ranges from 10-1 to 10~4 seconds with
charging normally effectively complete in about
10 seconds. The term : is the limiting
Drift Velocity
The charged particle reaching the collector sec-
tion is acted upon by two vector forces—its mo-
mentum and the electrical attraction for the
grounded or oppositely charged electrode. Ad-
ditionally, the motion of the particle toward the
electrode is retarded by viscous drag according
to Stokes' law. The net velocity component to-
ward the collecting electrode is termed the drift
velocity, and is described by the equation
p E E a
c p
6 7T (J.
where
w = drift velocity, cm/sec
E = electric field strength, stat volts/cm
M- =
viscosity, poises
p = a constant.
The subscripts p and c represent precipitating
and charging zones, respectively. Units are in
the electrostatic cgs system. The equation may
be modified by the Stokes-Cunningham correc-
tion factor for particles appreciably less than 1
micron in diameter, that is, approaching the
mean free path between molecular collisions in
air. For conducting particles, the constant p
equals 3, and for nonconductors, p is a function
of the dielectric constant and is usually between
1.5 and 2 (Rose and Wood, 1956).
Equation 81 illustrates the significance of the
electrical field's strength in collection effi-
ciency. The drift velocity varies with the
product of the charging field and collecting
field strengths. For this reason it is always
advantageous to operate a precipitator at the
maximum possible voltages without incur-
ring excessive sparkover. Field strength is
determined not only by impressed voltage
-------
160
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
but by electrode configuration, dust loading,
and other variables, so that a considerable
degree of experience is needed to evaluate
the drift velocity properly.
Efficiency
Determining the effectiveness of a device
for control of atmospheric pollution is fre-
quently difficult. When the airstream con-
tains ultrafine particles, the increase in
Hght transmittance may be important. In
some cases the reduction in number of par-
ticles or the reduction in darkness of filter
papers through which the air is passed may
be significant. The normal method of deter-
mining efficiency of precipitators is by
weight of material collected. The exponen-
tial Deutsch equation (Perry, 1950),
-wA
Q
(82)
where
F = efficiency, decimal equivalent
A = collecting area, ft
Q = volumetric flow rate, cfs,
with appropriate units for drift velocity, has
been developed primarily for application to
single-stage precipitators. Penney (1937)
presents the relationship for two-stage pre-
cipitators
F =
wL
(83)
•where
L = collector length, fps
d = distance between collector plates,
ft
for units with close plate spacing. The ex-
ponential-type equation is frequently found
applicable in practice. Walker and Coolidge
(1953) have found the expression
L
(84)
F = 1 - exp (-Kha) (V-V ) —
s v
where
h = relative humidity, decimal equivalent
a = particle radius, (j.
V = applied voltage, kv
K = a constant
to apply to both single- and two-stage precipita-
tors collecting gypsum dust under laboratory
conditions. The efficiency varies with the avail-
able voltage above the corona's starting voltage.
DESIGN FACTORS
Electrical Requirements
Normally, positive polarity in the ionizing sec-
tion is used in two-stage precipitators since it
is thought that less ozone and oxides of nitrogen
are thus produced. With positive polarity, spark-
over voltage is much closer to the critical co-
rona voltage than is found with negative polarity.
The practical operating voltage limit for stan-
dard units is about 18 kilovolts, with most oper-
ating at 1 0 to 13 kilovolts. Current flow under
these conditions is small, 4 to 10 milliamperes.
The collecting plates are usually activated at
5. 5 to 6. 5 kilovolts with precipitation's occur-
ring on the grounded spacing plates. The actual
current flow is very small since no corona ex-
ists between the plates.
In single-stage units recent developments have
made available rather elaborate automatic con-
trol devices to maintain the maximum practical
corona current. This type of control is not
feasible for two-stage units. For some applica-
tions, however, manually adjusted rheostats
have been used, and when a high degree of ef-
ficiency is required, the voltage can occasional-
ly be adjusted to compensate for buildup of col-
lected material.
Power consumption is a function chiefly of par-
ticle size, dust loading, voltage, and wire size.
The actual power required for removal of a
dust particle by precipitation is small com-
pared with that for mechanical collectors be-
cause the energy is applied primarily to the par-
ticle only and not to the total gas stream. In
practice, power requirements for standard two-
stage precipitators are 15 to 40 watts per thou-
sand cfm. The operating cost is, therefore,
low.
High voltage is obtained by vacuum tube recti-
fying power packs that operate from a 110- to
120-volt a-c primary circuit. On small units
one power pack may supply both ionizing and
collecting sections. For larger volumes two
or more power packs may be used in parallel
for various groupings of ionizing cells with
separate power packs for the collecting sections.
-------
Two-Stage Electrical Precipitators
161
Air Capacity
Manufacturers normally rate these units at 85 to
90 percent efficiency by tests based on discolora-
tion comparisons and at velocities between 300
to 600 fpm. For air-conditioning purposes these
values are usually adequate, but for cleaning
process air, a more thorough evaluation is neces-
sary. Efficiency of cleaning for this latter pur-
pose is usually based upon weight recovery and
will likely be lower than by discoloration com-
parison.
Equation 83 determines efficiency to be direct-
ly related to air velocity. Equation 82, for
units with constant collector area, relates ve-
locity to efficiency by an exponential function.
If dust particles move smoothly between the
plates, collection efficiency is a function only
of drift velocity and residence time. Penney
assumes streamline flow through the precipita-
tor, while recognizing that some turbulence
occurs, in arriving at the required collector
plate area for air-cleaning precipitators. Al-
though 600 fpm is the limiting velocity for
streamline flow in most two-stage units now
being manufactured, mechanical irregulari-
ties reduce the permissible velocity. Figure
93 includes a graphical representation of
equations 82 and 83 for data obtained on smoke-
houses.
It has been found that collection area is not
always controlling. At a velocity of 300 fpm
a dust particle is in the ionizing field only
about 0. 05 second, a very brief time when
compared with I. 0 to 10 seconds for single-
stage units. For some contaminants the in-
creased efficiency at low velocity is the ef-
fect of increased ionization time rather than
of streamline flow through the collector
plates.
The degree of ionization may be increased by
increasing the number of ionizing electrodes,
either by decreasing spacing or by installing
a second set of ionizing wires in series. Since
decreased spacing reduces the allowable volt-
age without sparkover, use of the series ar-
rangement appears advantageous. Decreased
spacing has the advantage of lower first cost
and lower space requirements.
Air Distribution
The distribution of the airstream entering the
precipitator is as critical for high-efficiency
two-stage operation as many other factors
normally receiving more attention. A super-
ficial velocity, the ratio of total airflow to
precipitator cross-sectional area, is useful
for equipment selection but may be misleading
for close design. For conditions of low overall
velocity of approximately 100 fpm, pressure
drop through a precipitator is insignificant,
and redistribution of high- or low-velocity
areas of the airstream will not occur. Var-
iations in airflow from 3 times average ve-
locity to actual reverse flow have been ob-
served in the vertical-velocity profile of these
units for hot gas streams. Figure 93 shows
that high velocity produces low efficiency
•while extremely low velocity does not result
in compensating improvement. Overall ef-
ficiency is thus lowered. Two-stage precipi-
tators are normally installed with horizontal
airflow and frequently in positions requiring
abrupt changes in direction of ductwork pre-
ceding the unit. Design such as this results
in turbulent, uneven airflow. If air enters the
precipitator plenum from an elbow or unsym-
metrical duct, the air tends to "pile up" on
the side of the precipitator opposite the entry.
Numerous methods are available for balancing
the flow. A straight section of duct upstream
eight duct diameters from the entry prevents
transverse unevenness if a gradually diverging
section precedes the precipitator. If this is
not possible, mechanical means must be used.
Turning vanes installed in an elbow or curve
maintain a uniform distribution and also re-
duce pressure drop across the elbow, but
do not balance flow satisfactorily. Baffles of
various types or egg crate straightening vanes
may be used in the transition duct. The most
effective air-balancing device found consists
of one or more perforated sheet metal plates
that fully cover the cross-section of the ple-
num preceding the ionizers.
The sheet metal plates introduce an additional
pressure drop that must be considered in the
initial exhaust system design. A study of
distribution in single-stage precipitators
found that each subsequent plate installed in
series added materially to flow uniformity
(Randolph, 1956). It has also been indicated
that, at low velocity, a perforated sheet im-
mediately following the collector plates may
in some cases be more effective than one
preceding the ionizers. A sheet in this loca-
tion also has the advantage of acting as an
additional collecting surface for charged par-
ticles, though this effect is usually minor.
An open area of 40 percent for the perforated
sheet has been found optimal, a range of 35
to 60 percent being generally adequate. In-
stallations handling heated airstreams, above
100°F , at low velocities, require baffling to
prevent high velocities at the tcp of the cham-
ber due to thermal effects. In these cases a
-------
162
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
100
90
» 80
I70
CJ
£ 60
z
5 50
LLJ
_l
5 40
30
20
E,
A. Smokehouses, extrapolated with Penney equation
B. Smokehouses, extrapolated with Deutsch equation
C. Asphalt saturator, operating test data
D. Galvanizing kettle, operating test data
F. Air conditioning", manufacturer's recommendation
1.5
2
GAS VELOCITY, fps
Figure 93. Efficiency of two-stage precipitator as function of velocity for several industrial
operations.
perforated sheet covering only the upper half
of the plenum preceding the ionizers may
suitably equalize the flow. Vertical-flow pre-
cipitators are not affected by thermal condi-
tions in this manner.
Auxiliary Controls
Two-stage precipitators have thus far not
been extended to process exhaust gases with
characteristics requiring unusual condition-
ing agents. Humidity adjustments, by -water
sprays or heating coils, are frequently used
to modify electrical properties of the con-
taminants to a suitably conductive condition.
For many materials, maintenance of greater
than 50 percent relative humidity is advanta-
geous. Under no condition, however, should
the gas stream to the ionizer be saturated;
reheating the airstream may be required to
avoid saturation. Water droplets should be
removed by mist eliminators preceding the
ionizer to prevent excessive sparkover.
The collector plates and housing must be cleaned
periodically. To keep this labor and downtime
to a minimum, the use of precleaners is fre-
quently recommended for the more easily re-
moved contaminants. For dry materials a cy-
clone is usually adequate, though a simple scrub-
ber is more commonly used, and the gas stream
is thereby humidified. If fibrous materials such
as cotton lint or synthetic fibres are entrained
in the exhaust air, they must be prevented from
reaching the precipitator where they may ac-
cumulate and bridge the plates, resulting in
arcing and possible duct fires. Scrubbers and
glass fiber filters have been successfully used
to prevent problems such as this.
Particles charged in the ionizing section may
sometimes have a drift velocity too low to be
completely removed in the collector section.
Operating the precipitator -without oil on the
collector plates and periodically blowing off
the flocculated material may also be desirable.
In either case the contaminants may precipitate
in the exhaust system or be collected by an af-
tercleaner following the precipitator. Inten-
tional use of this procedure is usually restric-
ted to dry dusts such as carbon black or normal
atmospheric dust. The aftercleaner may be a
filter, cyclone, or scrubber as required by the
specific process.
Air-conditioning installations are frequently
equipped -with automatic -washing and reoiling
devices. The aftercleaner then removes en-
trained water from the airstream and permits
uninterrupted air circulation through the system.
-------
Two-Stage Electrical Precipitators
163
CONSTRUCTION AND OPERATION
Assembly
Two-stage precipitators for capacities up to
20, 000 cfm may be supplied by the manufacturer
in completely preassembled package units re-
quiring only external wiring and duct connections.
For larger capacities or for heavy contaminant
loading, a field-assembled model is supplied.
This requires local fabrication of the precipita-
tor housing with necessary drains, doors, baf-
fles, and duct-work. Usually the ionizing and
collection sections are assembled on one frame
but they may be installed separately if desired.
The installed -weight of the precipitator is ap-
proximately 80 pounds per square foot for units
with over 5 square feet of cross-sectional area.
Usually, no additional foundation support is re-
quired for floor installations.
Smaller units may occasionally be adapted to fit
into existing ducts or transition chambers of an
exhaust system. If the precipitator vents to the
outside atmosphere, a shield must be provided
at the discharge side to protect it from weather
elements.
Maintenance
Process air may contain approximately 2 grains
of air contaminants per cubic foot in contrast to
air-conditioning loads of 2 grains per 1, 000
cubic feet. Because dusts and tars may not drain
or fall off, they may impose a limitation of hold-
ing capacity on the collector. Since no rapping
cycle is used on two-stage precipitators, the
collected materials are held to the plates for
relatively long periods of time and then -washed
down. The frequency of cleaning depends upon
the quantity of contaminant collected, though
cleaning cycles of 1 to 6 -weeks are typical. Some
installations are adaptable to automatic cleaning,
but in most, the collector plates must be -washed
down manually or removed and -washed in a tank.
If a dry dust is being collected, the plates are
usually recoated -with oil by either dipping or
spraying. When oils of low viscosity are col-
lected, the oil drips or runs off and hence only
occasional cleaning is needed to remove tars
or gummy deposits.
Ionizer wires do not require frequent cleaning.
These wires will, however, corrode slowly and
must occasionally be replaced. Stainless steel
wires rather than tungsten may be used if un-
usually corrosive conditions exist. The pre-
cipitator housing should be periodically washed
to remove deposited contaminants. Since most
standard precipitators are partially constructed
of aluminum, uninhibited caustic cleaning solu-
tion must not be used. Cleaning time varies
•with the nature of the collected contaminant.
Six to 12 man-hours per month may be consid-
ered average for a unit of 120 square feet
cross-sectional area.
Safety
Standard units are carefully constructed to
provide maximum electrical safety. At the low
current used, accidents are not common, but
normal high-voltage precautions must be ob-
served. Interlocks between access doors and
electrical elements should be used, and provi-
sions for delayed opening after deenergizing
are desirable to allow drainage of static charge.
The standard electrical systems are constructed
to shut off automatically if a direct arc occurs.
The inherent delay may, however, be sufficient
to ignite an excessive accumulation of combusti-
ble oils or tars. It is advisable, therefore, to
include automatic water sprinklers above the
collection unit. The fire hazard is minimized
by frequent cleaning if combustible contaminants
are being collected. A precipitator is obviously
not adaptable for use in exhaust systems hand-
ling vapors in explosive concentration.
APPLICATION
Among the types of operations that have been
successfully controlled by standard two-stage
precipitators are: (1) High-speed grinding
machines, (2) meat product smokehouses,
(3) continuous deep fat cookers, (4) asphalt
saturators, (5) galvanizing kettles, (6) rub-
ber-curing ovens, (7) carpeting dryers, and
(8) vacuum pumps.
The emissions from all these operations include
at least some oil mist. Oils, either mineral or
vegetable, have a relatively high drift velocity
and probably act as a conditioning medium for
less conductive emissions. In addition, the
oils deposited on the collector plates prevent
reentrainment of collected dust or fumes. Or-
ganic substances between Cg and C^^ have been
collected and, though showing some variation in
resistivity, are usually precipitated in the first
3 to 6 inches of a collector plate. Velocity and
ionization conditions that have been found ade-
quate for air pollution control purposes are
shown in Table 50.
Reentrainment of fumes from nonoiled plates
does not always occur. On a galvanizing in-
stallation, ammonium chloride used in the flux
is the largest constituent of the emissions.
-------
164
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
Table 50. INDUSTRIAL OPERATION OF TWO-STAGE PRECIPITATORS
Contaminant
source
Tool grinding
Meat smokehouse
Meat smokehouse
Galvanizing
Deep fat cooking
Asphalt saturator
(roofing paper)
Muller-type
mixer
C ontaminant
type
Oil aerosol
Wood smoke
Vaporized fats
Wood smoke
Vaporized fats
Oil aerosol
NH4C1 fume
Bacon fat
Aerosol
Oil aerosol
Phenol -formal-
dehyde resin
Ionizing
voltage
13, 000
13, 000
10, 000
14,300
13, 000
12, 000
13, 000
No. ot
ionizer
banks
1
2
1
2
2
1
1
Collector
voltage
6, 500
6,500
5,000
7,000
6,500
6,000
6,500
Efficiency,
wt %
90
(discoloration)
90
50
91
75
(light trans -
mittance)
85
87
Velocity,
fpm
333
60
50
60
68
145
75
Inlet
concentration,
grains/scf
--
0. 103
0. 181
0. 154
0. 384
0. 049
Remarks
Manufacturer1 s
recommendation
Humidified and
precleaned,
10-mil wire
Ionizer "wires at
1-1/2-in. spacing,
poor air distribu-
tion
Second ionizer,
1-1/2-in. spacing,
15-mil wire
Humidified and
precleaned
Humidified and
precleaned
Odor not suitably
reduced
During a test wherein fresh flux was added to
the galvanizing kettle though no galvanizing was
being done, ammonium chloride was found to
be flocculated in the precipitator and then re-
entrained in the exhaust air. During normal
galvanizing operations this did not occur. An
analysis of the materials collected in the pre-
cipitator showed that, during galvanizing, oil
from the metal being galvanized is vaporized.
Most of this is precipitated on the first few
inches of the collecting plates, but a small
quantity of the oil also precipitates with the
ammonium chloride over the balance of the
collecting area. The oil provides a medium
for holding the dry fume to the plates. This
effect is, of course, the reason for precoat-
ing the plates with oil in air-conditioning in-
stallations. The difference in precipitation
rate of oil mist and ammonium chloride fume
in the above example is illustrated in Figure 94.
Odors are frequently difficult control problems.
When the odor is due to particulate matter,
such as free fatty acids, the precipitator may
be entirely adequate. This is frequently found
to be the case with deep fat fryers. More com-
monly the odor source is both liquid aerosol
and vapor, and the degree of control by a pre-
cipitator depends upon the relative odor strength
of the two phases. For example, a precipita-
tor intended to eliminate both odors and visible
emissions from equipment blending hot phenolic
resins with other material is not suitable with-
Figure 94. Precipitator collector plates showing
rapid deposition of oil mist (dark area) compared
with (light area) ammonium chloride fume (Advance
Galvanizing Company, Los Angeles, Calif.).
-------
Two-Stage Electrical! Precipitators
165
out additional control equipment. Phenolic
resin dust is almost entirely removed from the
exhaust stream by the precipitator. Since free
phenol is present in both liquid and vapor form,
however, the odors are not eliminated unless
the temperature of the gas stream is low enough
to condense most of the phenol. The vapor pres-
sure of phenol at 220°F is about 5 times as great
as it is at 160°F.
Two-Stage Precipitators of Special Design
The foregoing discussion has been primarily
concerned with precipitators available from
manufacturers as standard units. The theory
is applicable, however, to less common units.
Under some conditions, dust of high resistivity,
above 10^ ohm-centimeters, causes ioniza-
tion at the collecting surface of single-stage pre-
cipitators. A decrease in the sparkover voltage
results, and the impressed high voltage may
have to be decreased to prevent excessive spark-
ing. The reduction may have to approach the
critical corona voltage, and if so, the corona
discharge and its resultant ionization diminish
with a corresponding drop in collection efficiency.
Sproull (1955) describes a two-stage unit designed
to circumvent this and other effects. For avoid-
ing back ionization at the grounded electrodes in
the ionization section, wider spacing between
ionizing electrodes -was used. Here negative
ionization -was used and at a correspondingly high-
er voltage owing to the -wide spacing. For pre-
venting reentrainment at the collector plates and
minimizing ionization and sparkover, electrodes
such as parallel sheets of expanded metal -were
found to perform more efficiently than the usual
flat plate electrodes. Optimum results with this
unit were obtained by using a 33-kilovolt reversing
polarity potential on the collector section.
For standard units the limiting air velocity is less
than 600 fpm. White and Cole (I960) described a
two-stage precipitator designed for high-efficiency
collection of oil aerosols at velocities between
2, 000 and 6, 000 fpm. The reentrainment of pre-
cipitated oil is prevented by use of a slotted tube
drain fitted over the trailing edge of the collecting
surface. Units such as these are designed and
manufactured to very close tolerances to permit
maximum electric field strength and the least
airflow disturbance. In the unit described, col-
lection voltage is held at 20 kilovolts while ioniz-
ing voltage is about 35 kilovolts. Negative ioniz-
ing polarity is used to provide a higher sparkover
level. Collection efficiencies as high as 99. 8 per-
cent by light diffusion standards are reported on
oil mists.
Ffcr continuous removal of collected contaminants,
-wetted film plates have been used in the collector
section. Installations have been made in which
the collection section has been replaced by a
•water scrubber, "which presumably acts as a more
efficient grounded electrode for some types of
contaminants. Located after the collecting plates,
a perforated plate on -which a flowing film of water
is maintained has been found to improve efficiency
slightly. The wetted baffle plate alone is not equiv-
alent to the effect of the normal collecting electrodes.
Sulfuric acid mist is efficiently collected by two-
stage precipitators constructed of corrosion-re-
sistant alloys. The Atomic Energy Commission
(1952) reports 94 percent efficiency, by radio-
activity-testing techniques, on acidic-cell ven-
tilation gases but adds a qualifying statement that
two-stage units are not recommended as the final
cleaner on exhaust gases containing radioactive
agents -without thorough trial in a pilot stage.
Self-cleaning precipitators are available in which
the collector plates are mounted on a chain belt,
as shown in Figure 95. The plates are slowly
passed through an oil bath that removes collected
solids and reapplies an oil coating to prevent re-
entrainment. Another somewhat similar unit
uses an automatically winding dry-filter medium
to trap the collected materials. Cleaning re-
quires only the occasional replacement of the
filter medium roll.
Equipment Selection
The validity of the theoretical expressions has
been well established for closely controlled
small precipitators. The design of industrial
units, however, invariably requires the use of
empirical correction factors and approxima-
tions. An analysis of equations 78 through 82
shows several physical properties on which in-
formation is not readily available to industrial
users. Particle size, for example, affects di-
rectly the limiting charge on the particle and
affects, therefore, the calculated drift velocity
and efficiency of separation. The actual size and
distribution of oil droplets in an industrial ex-
haust stream is rarely determined. Similarly,
the ionic current, which affects the field strength,
cannot be accurately measured by ammeters be-
cause the observed values also reflect current
leakage. For a small two-stage unit, ionic-cur-
rent flow is small, and any leakage affects the
total current very greatly.
Although operating costs of a precipitator are
low, the units are considered as high-efficiency,
high-cost control devices. A rule of thumb for
industrial installations is about $1. 00 per cfm
for the installed equipment. When no preclean-
er is required and a package precipitator is ade-
-------
166
AIR POLLUTION CONTROL EQUIPMENT FOR PARTICULATE MATTER
Figure 95. Automatic, self-cleaning,
two-stage precipitator. Dust parti-
cles charged in the ionizing section
are collected on aluminum plates.
These are mounted on a motor-driven
chain that automatically rinses and
reoils the plates every 24 hours
(American Air Filter Company, Louis-
ville, Ky.).
quate, the cost may be as low as $0. 25 per cfm
excluding installation and ductwork. A difficult
material to collect might require a precleaner,
two ionizer banks, steam coil reheater and
perforated baffle plates with an installed cost
of about $2. 00 per cfm.
Since many factors must be considered in de-
signing or selecting a two-stage unit for a given
process, some field experience with the charac-
teristics of the air contaminant is necessary.
In addition, a broad experience in precipitation
work is essential. If data on the specific pro-
cess to be controlled are lacking, then pilot or
laboratory-scale information must often be ob-
tained before a full-size unit is installed. Once
the pertinent data have been collected, the gen-
eral physical dimensions and electrical require-
ments of the precipitator can be determined by
the equations previously discussed.
OTHER PARTICULATE-COLLECTING DEVICES
In addition to the devices already mentioned for
the collection of particul'ate matter, there are
other devices of more simple designs that have
very limited application in the control of air
pollution. These include settling chambers,
impingement separators, and panel filters. Most
are used principally as precleaners, but some
are used as final collectors where the air con-
taminant is of large size or where the grain
loading is very small, for example, paper fil-
ters for paint spray booths.
SETTLING CHAMBERS
Settling chambers are one of the simplest and
earliest types of collection devices. The most
common form consists of a long, boxlike struc-
ture in the exhaust system. The velocity of the
dirty gas stream is reduced by the enlargement
in cross-sectional area, and particles with a
sufficiently high settling velocity are collected
by the action of gravity forces. A very long
chamber is required to collect small particles.
Structural limitations usually restrict the usage
of simple settling chambers to the collection of
particles 40 microns or greater. Their greatest
use is as a precleaner to remove coarse and
abrasive particles for the protection of the more
efficient collection equipment that follows the
chamber.
If horizontal shelves are closely spaced within a
settling chamber, the efficiency is greatly in-
creased because particle-settling distances are
reduced. A device such as this, known as a
Howard dust chamber, was patented in 1908. It
has a serious disadvantage in that the collected
material is very difficult to remove from the
shelves.
IMPINGEMENT SEPARATORS
When a gas stream carrying particulate matter
impinges on a body, the gas is deflected around
the body, while the particles, because of their
greater inertia, tend to strike the body and be
collected on its surface. A number of devices
use this principle. The bodies may be in the
form of plates, cylinders, ribbons, or spheres.
An impingement separator element with stag-
gered channels Ls shown in Figure 96.
Impingement separators are best used in the
collection of mists. The collected droplets form
a film on the surface and then gradually drip off
into a collection pan or tank. Conversely, col-
lected dry dust tends to become reentrained when
-------
Other Particulate-Collecting Devices
167
-STAGGERED CHANNELS
Figure 96. impingement separator elements.
it falls off the collecting surface. For this rea-
son, water sprays are sometimes used to wash
off the collected dust.
PANEL FILTERS
Panel filters are most commonly used in air con-
ditioning installations, though they do have several
important industrial applications. Filters are
supplied in units of convenient size, usually about
20 by 20 inches, to facilitate installation and clean-
ing. Each unit consists of a frame and a pad of
filter material, as shown in Figure 97. The frames
of similar units may be joined together to form a
panel. These filters are classified into two types,
viscous and dry.
Filters are called viscous because the filter medi-
um is coated with a viscous material to help catch
the dust and prevent reentrainment. The coating
material is usually an oil with a high flash point
Figure 97.
LouisviI!e,
Panel
Ky.).
fi Iter (American Air FiIter,
and low volatility. The filter pad consists of ma-
teriafs such as glass fibers, hemp fibers, ani-
mal hair, corrugated fiberboard, split wire, or
metal screening. When the maximum allowable
dust load has accumulated, the metal trays are re-
moved, cashed or steamed, reoiled, and put back
into service. Other pads are thrown away when they
become loaded with dust as shown by their increased
resistance to airflow. A common industrial applica-
tion of the wire screen-type filter is found in collec-
tion of mist generated from cutting oils used by
metal-cutting machines.
Dry filters are supplied in units similar to viscous
filters, except that the depth is usually greater.
The filter materials usually have smaller air
passages than the viscous filters do, and hence,
lower air velocities must be used to prevent ex-
cessive pressure drop. Dry filters are usually
operated at 30 to 60 fpm, as contrasted with 300
to 500 fpm for viscous filters. In order to in-
crease the filtering area per unit of frontal area,
the filter pads are often arranged in an accordian
form with pleats and pockets. When the pressure
drop becomes excessive because of accumulated
dust, the dry-type pads are discarded. Dry fil-
ters are frequently used to collect the overspray
from paint-spraying operation.
-------
168 AIR POLLUTION CONTROL EQUIPMENT FOR P ARTICULATE MATTER
PRECLEANERS on the more efficient (and more expensive) final
collector. If the collected material has value, a
Devices of limited efficiency are often used precleaner, for example, one ahead of a scrubber,
ahead of the final cleaner. If the gases contain can sometimes collect the bulk of it in a more
an appreciable amount of hard, coarse particles, usable form. Devices usually used as precleaners
a precleaner can materially reduce erosive wear are settling chambers and centrifugal separators.
-------
CHAPTER 5
CONTROL EQUIPMENT FOR GASES AND VAPORS
AFTERBURNERS
HOWARD F. DEY, Air Pollution Engineer
BOILERS USED AS AFTERBURNERS
WILLIAM L. POLGLASE, Air Pollution Engineer
ADSORPTION EQUIPMENT
MARC F. LeDUC, Air Pollution Engineer
VAPOR CONDENSERS
ROBERT T. WALSH, Senior Air Pollution Engineer*
ROBERT C. MURRAY, Senior Air Pollution Engineer
GAS ABSORPTION EQUIPMENT
HARRY E. CHATFIELD, Air Pollution Engineer
RAY M. INGELS, Air Pollution Engineer f
*Now with the National Center for Air Pollution Control, Public Health Service, U.S. Department of
Health, Education, and Welfare.
^ Now with State of California Vehicle Laboratories, Los Angeles, California.
-------
CHAPTER 5
CONTROL EQUIPMENT FOR GASES AND VAPORS
AFTERBURNERS
Afterburners are used as air pollution control
devices for a large variety of industrial and
commercial equipment. Whenever the equip-
ment releases combustible aerosols, gases, or
vapors into the atmosphere, an afterburner can
frequently be used to control these emissions
for compliance "with air pollution control regula-
tions. Opacities of visible emissions can be
reduced or eliminated, the concentrations of
particulate matter can be reduced, or nuisances
can be abated through their use. In some in-
stallations it is possible to reclaim a portion of
the sensible heat from the afterburner discharge
gases and thereby reduce the cost of operation.
Two types of afterburners are in current use:
Direct fired and catalytic. The direct-fired
type is much more commonly employed be-
cause of greater adaptability and lower first
cost as compared with the catalytic type. Never-
theless, catalytic afterburners appear to be use-
ful for control of some processes -whereby or-
ganic vapors are emitted. The use of a catalyst
promotes some combustion reactions at lower
temperatures than those normally required by
direct-fired afterburners. The operating tem-
peratures of catalytic afterburners may thus
be lower than those of the direct-fired type used
for the same service, and lower fuel costs are
sometimes possible through their use.
When incomplete combustion occurs in any after-
burner, odor concentrations are not reduced and,
at times, are actually increased across the after-
burner. To overcome this defect, incineration
at higher temperatures -with adequate oxygen is
required. In some cases, because of materials
of construction, catalytic afterburners cannot
be operated at temperatures necessary to pro-
duce complete combustion. In cases where
temperatures of 1200°F or more are required,
there appears to be little inducement to install
the higher priced catalytic unit since a well-
designed direct-fired unit will usually be adequate.
DIRECT-FIRED AFTERBURNERS
Direct-fired afterburners have proved capable
of controlling combustible emissions from
many kinds of industrial and commercial equip-
ment. Indeed, they are the most commonly
used air pollution control devices where emis-
sions of combustible aerosols, vapors, gases,
and odors are emitted. Equipment success-
fully controlled include aluminum chip driers,
animal blood driers, asphalt-blowing stills,
automotive brake shoe-debonding ovens, citrus
pulp driers, coffee roasters, electric insula-
tion burnoff ovens, flue-fed refuse incinerators,
foundry core ovens, meat smokehouses, paint-
baking ovens, rendering cookers, varnish cook-
ers, and other equipment operated within sim-
ilar temperature ranges.
Specifications and Design Parameters
The principal components of a direct-fired after-
burner are the combustion chamber, gas burner,
burner controls, and temperature indicator. The
afterburner chamber is generally cylindrical and
constructed of firebrick or castable refractory
with a sheet iron shell. For most afterburner
installations, Class 27 castable refractory and
high-duty firebrick are satisfactory since aver-
age gas temperatures seldom exceed 2, 000°F.
The afterburner chamber must be designed for
complete mixing of the contaminated gases with
the flames and the burner combustion gases.
One satisfactory method of achieving this ap-
pears to be the admission of the contaminated
gases into a throat where the burner is located.
Sufficiently high velocities may be obtained
here for thorough mixing of the gases with the
burner combustion products in the region of
highest temperature. Gas velocities in the
afterburner throat of 15 to 25 fps are found
adequate. A retention time of 0. 3 to 0. 5 sec-
ond for the gases within the afterburner, and
operating temperatures of 850°F to 1,500°F
have been found satisfactory for most instal-
lations. High efficiencies are normally achieved
at the higher operating temperatures. Figure
98 illustrates a typical direct-fired afterburner
in sectional view.
An exhaust fan is used to deliver the contam-
inated gases to the afterburner when natural
draft is inadequate. This fan may be an axial-
flow or squirrel cage type since the dust resis-
tance of the exhaust system installed is usually
low.
Several types of gas burners have been used
successfully. Among these are: Atmospheric,
nozzle mixing, pressure mixing, premixing,
and multijet gas burners. Many arrangements
171
-------
172
CONTROL EQUIPMENT FOR GASES AND VAPORS
REFRACTORY LINED
STEEL SHELL
REFRACTORY RING BAFFLE
INLET FOR CONTAMINATED
AIRSTREAH
Figure 98. Typical direct-fired afterburner
with tangential entries for both the fuel and
contaminated gases.
of burners in afterburner chambers are pos-
sible. The burner should be positioned to
provide complete mixing of the contaminated
gases with the burner combustion gases. In
addition, sufficient oxygen must be available
to burn all air contaminants completely. The
gas burner should be located near the entrance
into the afterburner for the contaminated gas-
es. One or more burners firing tangentially
into the base of a cylindrical afterburner cham-
ber have been used advantageously. Figures
99, 100, 101, and 102 illustrate afterburners
with this type of burner arrangement. Multi-
jet and multinozzle burners located in the base
of an afterburner chamber have also been used.
Figure 103 shows an afterburner with this
burner arrangement. The latter arrangement
appears to provide better mixing of contam-
inated gases with burner combustion gases
and consequently greater afterburner efficien-
cies. Afterburners have been constructed
that use multijet gas burners in which all com-
bustion air is supplied by the contaminated
gases, and thus minimum fuel requirements
for a specific temperature are provided. An
afterburner of this type is shown in Figure 104.
An afterburner using a single-inspirator gas
burner is shown in Figure 105. Burners having
Figure 99.Direct-fired afterburner venting three
varnish-cooking kettles (Standard Brand Paint
Co., Torrance, Calif.).
long, luminous flames appear to incinerate
contaminants more effectively than those having
short, nonluminous flames. This is probably
due to more effective mixing of the contam-
inated gases with the burner combustion gas-
es and greater heat transfer by radiation.
Long, luminous flames may be created if the
mixing of the gas and air is so slow that burning
is completed at a considerable distance from the
burner (North American Combustion Handbook,
1952).
Many direct-fired afterburners have only flame
failure controls. The installation of modulating
gas burner controls may effect considerable
savings in fuel where the volume of gases or the
amount of combustible material delivered to the
afterburner varies appreciably during the pro-
cess cycle, or where both vary. A constant
temperature in the afterburner chamber can be
maintained through a gas temperature sensing
element that actuates the burner input control.
When, however, the volume of contaminated gas-
es and the amount of combustible materials re-
-------
Afterburners
173
Figure 100. Direct-fired afterburner controlling emissions
Southgate, Calif.).
from a paint-baking oven (Rheem Mfg. Co.,
main relatively constant, the firing of the burn-
er at a fixed rate seems preferable.
An indicating- or recording-type temperature-
measuring device is usually installed to show
the afterburner's operating temperature at all
times. A bare thermocouple is normally used
because of low cost and rapid response to tem-
perature changes. The thermocouple should be
located near the top of the afterburner chamber
to avoid large errors produced by direct radia-
tion from the burner flames. The thermocouple
may be installed in a thermocouple well for pro-
tection.
A safety pilot is usually provided to shut off the
burner gas supply if the main burner malfunc-
tions or the flow of contaminated gases to the
afterburner is interrupted. It may also be ad-
visable to install a high-temperature-limiting
control to shut off the gas burner fuel supply
if the flow of contaminated gases to the after-
burner is interrupted.
Operation
Operation of direct-fired afterburners is rela-
tively simple. The contaminated gases are de-
livered to the afterburner from the process
equipment by the exhaust system. The con-
taminated gases are mixed thoroughly with the
flames and the burner combustion gases in the
afterburner throat. Next, the gases pass into
the main section of the afterburner where ve-
locity is reduced somewhat by the larger cross-
sectional area. Here the combustion reactions
are completed and the incinerated air contami-
nants and combustion gases are discharged to
the atmosphere. Although the mechanics of
most combustion reactions are doubtlessly
complicated, conditions must be provided that
will most nearly incinerate all the air con-
taminants to carbon dioxide and -water vapor.
-------
174
CONTROL EQUIPMENT FOR GASES AND VAPORS
Figure 101. A direct-fired afterburner control-
ling emissions from two recirculating-type meat
smokehouses (High Standard Meat Co., Los Angeles,
Calif.).
Figure 102. Direct-fired afterburner controlling
emissions from a plastic-curing oven (Industrie1
Rock, Glendaie, Calif.).
Efficiency
Efficiency of direct-fired afterburners depends
on several variables, namely, the degree of
mixing of the contaminated gases -with the flames
and burner combustion gases within the after-
burner cha-mb.er, operating temperature, re-
tention time of the gases -within the afterburner,
and concentration and types of contaminants to
be burned.
Design calculations
The following example illustrates some of the
factors that must be considered in the design
of a direct-flame afterburner.
Example 21
Given:
A direct-fired afterburner is to be installed to
incinerate the air contaminants discharged from
meat smokehouses and eliminate visible emis-
sions and odor. The maximum rate of discharge
has been determined to be 1, 000 scfm at 150°F.
Assume the effluent gases have the same prop-
erties as air. (Consideration of the enthalpies
and specific heats of the individual gaseous com-
ponents of the contaminated gas stream will show
this to be a reasonable assumption. Any correc-
tions would introduce an insignificant refinement
to the calculations when considered with respect
to the overall acciiracy required. Nevertheless,
there will be situations where the contaminated
gases contain a sufficient quantity of a. combusti-
ble gas or vapor to make necessary the calcula-
tion of its heat of combustion and consequent en-
thalpy and temperature increase of the after-
burner gases.)
Problem:
Determine the design features of a direct-flame
afterburner to incinerate the contaminated gases.
-------
Afterburners
175
Figure 103. Direct-firea afterburner control-
ling odorous emissions from animal matter-ren-
dering equipment (Peterson Manufacturing Co.,
Inc., Los Angeles. Calif.).
Figure 104. A direct-fired afterburner control-
ling emissions from five meat smokehouses (Pack-
ers Central Loading Co., Vernon, Calif.).
Solution:
1. Mass flow rate of contaminated gases:
(1,000 cfm)(60 min/hr)
13. 1 ft? air/lb air
= 4,580 Ib/hr
2g Heat required to increase the temperature
of the gases from 150° to 1, 200°F;
An operating temperature of 1, 200°F for an
afterburner has been found sufficient to in-
cinerate air contaminants emitted by meat
smokehous e s.
Enthalpy of gas at 1,200°F = 287.2 Btu/lb
(see Table D3, Appendix D)
Enthalpy of gas at 150°F = 21.6 Btu/lb
Ah = 265.6 Btu/lb
(4, 580)(265.6) = 1,216,000 Btu/hour
3. Heat losses from afterburner due to radia-
tion, convection, and conduction:
A quantity equal to 10 percent of item No. 2
will be assumed. This is a conservative
Figure 105. Direct-fired afterburner venting
three varnish-cooking kettles and a thinning
station (National Paint and Varnish Co., Los
Angeles, Cal i f.).
-------
176
CONTROL EQUIPMENT FOR GASES AND VAPORS
estimate for afterburners constructed, of
firebrick or castable refractory and operated
at normal temperatures.
(1,216, 000)(0. 10) = 121,600 Btu/hour
4. Total heat required by afterburner:
1,216,000 + 121,600 = 1, 338,000 Btu/hour
5. Required natural gas volume capacity of
burner:
The contaminated gases from many types of
equipment contain sufficient oxygen to furnish
at least the excess needed for proper com-
bustion of the natural gas supplied to the
burner. Smokehouses are considered to be
this type of equipment. The natural gas will,
therefore, be assumed to be supplied with
theoretical air at the burner and with ex-
cess air from the smokehouse gases.
Gross heating value of natural gas taken at
1, 100 Btu/ft3.
Net heat available at 1, 200 °F from the burn-
ing of 1 ft of natural gas with theoretical
air is 721.3 Btu/ft3 (see Table D7 in Appen-
dix D).
1, 338,000
721 '3 -
n ,,,. , 3,,
= 1,854 ftj /hour
6. Volume rate of gas burner combustion
products at 1, 200°F:
With theoretical air, 1 ft of natural gas
yields 11. 45 ft of products of combustion
(see Table D7 in Appendix D).
(1.854)(11.45)(1.200+ 460) _ 3
(3,600)(60+ 460) ~ ' /sec
7. Volume rate of contaminated gases at 1,200°F:
nated gases with flames of burner combustion
products. Use a design velocity of 20 ft/sec.
20
= 3.60 ft , .'. Use throat diameter of
25-1/2 in.
10. Diameter of afterburner combustion chamber:
Combustion chamber velocities of 10 to 15
ft /sec have been found high enough to provide
adequate turbulence for completing combus-
tion and allowing the construction of an after-
burner of reasonable length for recommended
residence time of 0. 3 sec. Use a design ve-
locity of 12 ft/sec.
72. 0
12
= 6. 00 ft , . . Use chamber diameter
of 33 in.
11. Length of afterburner combustion chamber:
Use a ratio of afterburner combustion cham-
ber length-to-diameter of 2. This ratio ap-
pears to be a reasonable minimum for the ve-
locities used and will provide adequate resi-
dence time of the gases in the combustion
chamber.
(2)(33) = 66 in.
12. Retention time of gases in afterburner com-
bustion chamber:
5.5 ft
~L2 ft/sec
= 0.46 sec
This is satisfactory. Experience indicates
that a retention time of 0. 3 sec at 1, 200°F
is sufficient for the incineration of nearly
all types of combustible air contaminants.
Results of several tests of direct-fired after-
burners are presented in Table 51. The process
equipment and the afterburners in each of these
tests are briefly discussed.
(1,000)(1.200 + 460)
(60)(60 + 460)
8. Total volume rate of gases in afterburner at
1, 200°F:
18.83 +53.2 = 72.0 ft3/sec
9. Diameter of afterburner throat:
Throat velocities of 15 to 25 ft/sec have been
found to provide adequate mixing of contami-
Test C-319
The afterburner in test C-319 and in test
C-725 was installed to reduce the discharge
of excessive particulate matter and elim-
inate odor and opacity of emissions from
the operation of varnish-cooking kettles.
Aerosols, vapors, and odors from the
operation of three varnish cooking kettles
and one kettle-cooling station are vented
by an exhaust system to a direct-fired
afterburner preceded by a water spray
leg. The water spray leg functions as a
-------
Afterburners
177
Table 51. STACK ANALYSES OF EMISSIONS FROM DIRECT-FIRED AFTERBURNERS
Test scries No.
Process equipment1*
Burner type
Afterburner combustion
chamber temp, PF
Inlet gas volume, sefm
Outlet gas volume, scfm
Particulate matter,
Ib/hr, inlet
Ib/hr, outlet
Control efficiency, %
Organic acids, Ib/hr,
inlet
Organic acids, Ib/hr,
outlet
Aldehydes, Ib/hr, inlet
Aldehydes, Ib/hr, outlet
C-319
Three varnish-
cooking kettles
Four no/./.lc
mixing
1, 220
950
1, 300
5. 20
n ^ i
U . J i
94
1. 41
0. 73
0. 30
0. 11
C-725
Three varnish-
cooking kettles
One inspirator
1, ZOO
200
920b
5. 70
0 20
96
0. 24
0. 00
0. 29
0. 02
C-462
Five meat
smokehouses
One pressure
mixing
630
305
500
0. 52
On 7
. O f
-10
0. 24
0. 00
0. 04
0. 06
C-566
Five meat
smokehouses
One multijet
850
1,600
2, 000
1.66
OcA
. DO
66
1. 88
0. 27
0. 49
0. 22
C-318
Paint bake oven
Three nozzle
mixing
1, 520
1, 400
1, 800
0. 40
On Q
. U7
78
--
0. 19
0. 03
C-729
Phthalic anhydride
production unit
One multijet
1, 200
3,800
6,400
20. 0
i n
i . u
95
29
0. 88
1. 75
0. 43
precleaner in which the contaminated gas
stream is cooled to provide some conden-
sation of vapors. The water sprays also
remove larger particles of mists and
solids and provide flashback control.
The afterburner is a vertical cylinder
tangentially fired at the base by four noz-
zle mixing gas burners as shown in Fig-
ure 99. The contaminated gas stream
enters the afterburner tangentially op-
posite the burners.
Test C-725
.The basic equipment and process in test C-725
are essentially the same as in test C-319. In
this installation, however, there is no water
spray leg. The afterburner consists of a hori-
zontal cylinder fired at one end by an inspira-
tor-type gas burner as shown in Figure 105.
The contaminated gas stream enters the after-
burner tangentially adjacent to the burner.
The entire system is placed under negative
pressure by an exhaust fan located at the
afterburner outlet.
The afterburner in this test and the one in
test C-319 show high efficiency in the incin-
eration of particulate matter.
Test C-462
Emissions of excessive opacity were abated
by the afterburner described in test C-462
and in test C-566. The srnoke and gases from
five natural-draft-type meat smokehouses are
vented through ductwork to a direct-fired af-
ter burner. The afterburner is a vertical
cylinder fired tangentially at the base by a
pressure mixing gas burner. The contam-
inated gases enter the afterburner axiallyat
the base.
Intermittent emissions were visible from
the afterburner during the test. Efficien-
cy in the incineration of particulate matter
•was negative. This appeared to be due
primarily to the unusually low afterburner
temperature. The visible component of
the smokehouse gases -was apparently
vaporized in the afterburner, which ren-
dered it invisible. Nevertheless, there
was no condensation downstream from the
afterburner to form a visible plume. The
afterburner was satisfactory for abating
the excessive opacity of emissions from.
the smokehouses.
Test C-566
The basic equipment and process in test
G-566 are virtually the same as those in
-------
178
CONTROL EQUIPMENT FOR GASES AND VAPORS
test C-462. The afterburner is a vertical
cylinder with an upshot multijet gas burner
located at the base as shown in Figure 104.
The smokehouse gases are conveyed to the
afterburner by an exhaust fan. One unusual
feature of this afterburner is that all oxy-
gen for combustion of the fuel gas is ob-
tained from the contaminated gas stream.
This afterburner was reasonably efficient
at a relatively low temperature, and no
visible emissions -were detected during the
test.
Test C-318
An afterburner was installed to reduce the
concentration of particulate matter in the
oven exhaust gases to allowable limits.
Metal drums are spray painted with epon-
phenolic and oleoresinous coatings and are
baked at 420°F in a conveyorized gas-fired,
re circulating-type paint bake oven. A
portion of the gases containing several sol-
vents is vented to an afterburner. The re-
mainder is recirculated from the heaters
to the oven. The afterburner consists of
a vertical cylinder with three nozzle mix-
ing gas burners located around its circum-
ference as shown in Figure 100. The con-
taminated gases enter the afterburner tan-
gentially and the burners fire the unit in a
similar manner.
The afterburner's operating temperature of
1, 520°F during the test was higher than usu-
al. Previously an operating temperature of
1, 410°F had appeared satisfactory for con-
trol of emissions from the oven.
Test C-729
An afterburner was installed to reduce the
excessive concentration of particulate mat-
ter and eliminate excessive opacity of emis-
sions from a phthalic anhydride production
unit. Phthalic anhydride is produced by a
catalytic oxidation process. The gases dis-
charged contain phthalic anhydride, naphtha-
lene, benzoic acid, naphthaquinone, and other
reaction products. These gases are vented
to a settling chamber and direct-fired after-
burner. The afterburner is a vertical cyl-
inder -with a multijet gas burner located in
the upper section as shown in Figure 106.
The contaminated gases enter the after-
burner tangentially near the burner and
pass downward through the unit. An ex-
haust fan is located at the afterburner out-
let.
Figure 106. Emissions from two phthalic anhydride
production units being controlled by direct-fired
afterburners (Reichhold Chemicals, Inc., Azusa,
Calif.).
Efficiency in the incineration of particulate
matter was high. Determination of the
odor threshold of afterburner inlet and out-
let gases indicates an odor reduction ef-
ficiency of 97 percent.
Installation costs
A survey of installation costs of direct-fired af-
terburners reveals a general range of $5.00 to
$10. 00 per scfm contaminated gas.
CATALYTIC AFTERBURNERS
Catalytic afterburners have found their greatest
use in the control of solvent and organic vapor
emissions exhausted from industrial ovens.
These emissions are created in metal-decorating
and metal -coating ovens, foundry core ovens,
wax burnout ovens, fabric-backing and fabric-
coating ovens, and ovens for the baking of the
binder in the production of rock wool batts.
Some typical catalytic afterburners are shown
in Figures 107, 108, 109, and 110.
Specifications and Design Parameters
Basically, a catalytic afterburner consists of
the afterburner housing containing a preheating
-------
Afterburners
179
Figure 107. Catalytic afterburners controlling
emissions from fabric-coating baking ovens
(Western Backing Corporation, Culver City, Calif.).
section (if one is necessary) and a catalyst sec-
tion. A gas burner preheats the contaminated
gases before they flow to the catalyst section.
Drawings of two catalytic afterburner installa-
tions are shown in Figures 111 and 112. Ar-
rangements for the recovery of heat from the
afterburner gases are illustrated.
Frequently, the contaminated gases are delivered
to the afterburner by the fan exhausting the pro-
cess equipment. In one type of catalytic after-
burner, the exhaust fan is located within the after-
burner housing between the preheat burner and
the catalyst bed. This fan also mixes the gases
and distributes them evenly over the catalyst.
Condensates do not occur in the fan since it oper-
ates above condensation temperature. Of course,
the fan must be constructed of materials that can
withstand the maximum temperature of the gases
being handled.
The interior chamber of the afterburner may be
constructed of 11- to 16-gage black iron, heat-
resisting steel, stainless steel, or refractory
materials. Heat-resisting steel should be used
for operating temperatures between 750° and
1, 100°F; stainless steel is recommended for
operating temperatures exceeding 1, 100°F.
Refractory materials are recommended for
temperatures exceeding 1, 300°F. A thickness
of 4 to 6 inches of similar thermal insulation
is required unless refractory materials are
used. The exterior sheet is usually fabricated
from 16- to 20-gage mild steel. The frame-
work is usually fabricated from standard struc-
tural steel. Gas velocities through the chamber
of about 20 fps have been found satisfactory.
The contaminated gases are preheated to the
reaction temperature by a gas burner before
passing through the catalyst. When the pre-
heat burner is on the discharge side of a fan,
a premix gas burner is normally used be-
cause of the positive pressure in the after-
burner chamber. When the fan is between
the preheat burner and the catalyst bed, an
atmospheric burner may be used since a neg-
ative pressure exists in the preheat section
of the afterburner chamber. Sizing the pre-
heat burner seems advisable to increase the
temperature of the contaminated gases to the
required catalyst discharge gas temperature
without regard to the heating value of the com-
bustible materials contained in the contaminated
gas stream, especially if considerable varia-
tion in concentration occurs. The concentra-
tion of combustibles from the process equip-
ment is normally 25 percent of the lower ex-
plosive limit or less to meet the requirements
of the National Board of Fire Underwriters.
Experience indicates that the preheat burner
should have sufficient capacity to heat the
contaminated gas stream to 950°F minimum
to obtain adequate catalytic combustion of the
compounds that are more difficult to burn.
The operating temperatures of catalytic after-
burners are usually about 650° to 1,000°F.
This is lower than that of most direct-fired
-------
180
CONTROL EQUIPMENT FOR GASES AND VAPORS
Figure 108. A catalytic afterburner used to control a foundry core baking oven (Catalytic
Combustion Corporation, Detroit, Michigan).
afterburners, but feasible for air contaminants
that are burned readily at these lower tempera-
tures. Direct-fired afterburners are ordinarily
installed to incinerate smoke and organic aero-
sols that are more difficult to burn.
An oxidizing catalyst is employed. Because of
its high activity the catalyst is usually platinum
with alloying metals. Other possible catalysts
include copper chrbmite and the oxides of cop-
per, chromium, manganese, nickel, and co-
balt (Krenz et al. , 1957). The catalyst has
a very porous, highly adsorptive surface.
It is deposited upon nickel alloy ribbon and
formed into filter-like mats (type A) or depos-
ited upon small, thin, ceramic rods for the
fabrication of small blocks or bricks (type B).
The two types of catalyst elements are shown
in Figure 113.
Catalytic afterburners possess an inherent
maintenance factor not present in direct-fired
afterburners, namely, that usage of the catalyst
produces a gradual loss of activity through
fouling and erosion of the catalyst surface. Oc-
casional cleaning and eventual replacement of the
catalyst are, therefore, required.
Modulating controls on the burner regulated by
the catalyst discharge gas temperature are
usually used. This allows the fuel gas input
to the preheat burner to be reduced as the
rate of heat released in the catalyst bed in-
creases as a result of larger concentrations
of combustible vapors. The sensing instru-
ment commonly used is a type employing a
fluid-filled bulb for detecting gas temperature
with capillary and bellows. Movement of the
bellows is amplified and transmitted to the
preheat burner gas valve and combustion air
blower blast gate. Electronic instruments
are used less frequently because of consid-
erably greater cost.
When operating conditions do not vary greatly,
an improved means of ensuring maximum after-
burner efficiency seems to be the firing of the
preheat burner at a fixed input capable of heat-
ing the contaminated gases to the temperature
required for complete oxidation at the maxi-
mum rate of influx. Installation of a high-tem-
perature-limiting control on the downstream
side of the catalyst may be necessary to pre-
vent overheating of the afterburner.
Operation
The contaminated gases are delivered to the
afterburner by the exhaust fan. The gases
pass into the preheat zone •where they are
heated to the temperature required to support
-------
Afterburners
181
CLEAN HOT EXHAUST
ENTRANCE TfO PASS OVERHEAD KONORAIL CORE BAKING OVEN
Figure 109. A catalytic afterburner venting a
foundry core baking oven (Catalytic Combustion
Corporation, Detroit, Michigan).
catalytic combustion. This temperature varies
with the nature and composition of the contam-
inants to be burned, generally being about 650°
to 1,000°F. Some burning of contaminants usu-
ally occurs in the preheating zone. The pre-
heated gases then flow through the catalyst bed
where the remaining combustible contaminants
are burned by catalysis.
The combustible materials may be present in any
concentration below the flammability limit. The
factors influencing their combustion are tempera-
ture, oxygen concentration, contact with the cata-
lyst surface, and nature of materials to be burned.
Since the combustion reaction is exothermic, an in-
crease in catalyst temperature is produced. The
greater the concentration of combustibles in the
entering gases, the greater the heat release rate
and the higher the catalyst temperature. When
this effect is appreciable, the initial preheat tem-
perature of the entering gas.es can be reduced
after the combustion reaction is established. This
may be accomplished by providing modulating con-
trol of the preheat burner input actuated by an
element sensing afterburner discharge gas tem-
perature. In this way essentially constant after-
burner discharge gas temperature and catalyst
temperature can be maintained. Preheat fuel
consumption is then theoretically inversely pro-
portional to the concentration of combustibles in
the entering gases.
The catalytic reaction depends upon the diffusion
of combustible vapor molecules to the porous
catalyst surface where they are adsorbed. Oxy-
gen in a highly active state is also adsorbed on
this surface. The combustion reaction takes
place, and the combustion products are desorbed.
Low-molecular-weight materials may react more
readily than those of high molecular "weight be-
cause of higher diffusion rates. The stability of
the molecule, however, must also be considered.
Methane is an example of a stable low-molecular-
weight compound requiring a high temperature
for catalytic combustion, about 760 °F (Suter,
1955), whereas hydrogen may be catalytically
burned at a temperature of about 500 °F (Oxycat
Technical Manual, 1956). The complete oxida-
tion of more stable compounds requires higher
temperatures and greater catalyst surface than
those for less stable compounds. Entrained
particles and liquid droplets are not likely to
contact the catalyst surface to any appreciable
degree because of their greater mass and lack
of diffusional movement (Suter, 1955).
A direct relationship may exist between the auto-
ignition temperature of an organic vapor and the
temperature at which catalytic oxidation will
occur. In other words, the higher the auto-igni-
tion temperature of a compound, the higher the
expected temperature required for catalytic
oxidation.
Flow through the catalyst bed should be turbu-
lent to promote contact of the contaminants
with the catalyst surface. For adequate incin-
eration, the combustible substances must be in
the vapor phase or must be capable of being
vaporized at a reasonably low temperature in
the preheat zone. Reaction products must be
sufficiently volatile for complete desorption
from the catalyst surface (Suter, 1955).
Substances that poison the catalyst must not be
present. Some of these are the vapors of metals
such as mercury, arsenic, zinc, and lead.
Substances that form solid oxides must also be
-------
182
CONTROL EQUIPMENT FOR GASES AND VAPORS
avoided since they coat the catalyst surface and
render it inactive. Some types of halogenated
hydrocarbons are also harmful to the catalyst.
Finally, the burned gases are discharged through
a stack to the atmosphere, to a process that may
use the sensible heat of the exhaust gases, such
as a bake oven or dryoff oven, or they may be
passed through an exchanger for heating the gas-
es entering the afterburner, which thereby re-
duces the amount of fuel required by the preheat
burners.
Figure 110. A catalytic afterburner controlling
emissions from a sheet metal-coating baking oven
(Advance Hetai Lithographing Inc., Ei Monte, Calif.
At least the theoretical quantity of oxygen re-
quired for complete oxidation of the combusti-
ble gases must be contained in the influent gas-
es. The efficiency of the catalyst is, however,
normally greater when excess oxygen is present.
Complete mixing of the contaminated gases with
the preheat burner combustion gases is required
for optimum performance of the afterburner.
This may be accomplished by baffles, elbows,
or a centrifugal fan. Moreover, the flow of
gases through the catalyst bed must be uniform.
This may be accomplished by means of splitting
vanes or a perforated baffle plate located up-
stream from the bed.
A maximum operating temperature of 1, 800 °F
is indicated for one make of catalytic after-
burner. This limit is apparently imposed by the
materials used in construction and by the cata-
lyst. Some solvent emissions from c'oating and
lithograph bake ovens are usually found adequate-
ly controlled at catalyst discharge temperatures
of 950° tol,OQO°F. For a properly designed
catalytic afterburner, an overall efficiency of
80 percent in the reduction of particulate mat-
ter may be expected. For solvents known to
burn at relatively low temperatures, the con-
-------
Afterburners
183
?—DISCHARGE TO ATMOSPHERE
RETURN TO OVEN
ISCHIRGE TO 1TIOSPHERE
Figure \\\. Typical catalytic afterburner
utilizing direct heat recovery.
Figure 1i2. Typical catalytic afterburner
utilizing indirect heat recovery.
Figure 113. Afterburner catalyst elements: Left, metal Iic-ribbon-type catalyst element,
type A (Catalytic Combustion Corporation, Detroit, Mich.); right, porcelain-rod-type
catalyst element, type B (Oxy-Catalyst, Inc., Berwyn, Pa.).
-------
184
CONTROL EQUIPMENT FOR GASES AND VAPORS
centration of particulate matter in the stack gas
has been reduced to as low as 0. 01 grain per scf.
Table 52 summarizes the operating conditions
and shows the coating applied for several coat-
ing ovens during tests of their catalytic after-
burners. The temperature range of the cata-
lyst is the range over which a series of tests
was made, the catalyst's discharge tempera-
ture being held constant for each test. The
catalyst's superficial gas velocities were con-
stant within +_ 15 percent for all runs. Figure
114 shows hydrocarbon combustion efficien-
cies as determined by tests of six different in-
stallations, all similar in operation though dif-
fering in type of coating, and type and quantity
of solvents, as gix^en in Table 52.
A hypothetical overall efficiency has been cal-
culated to show total air pollution reduction.
This is the total reduction of aldehydes as form-
aldehyde, hydrocarbons as hexane, and particu-
late matter. Figure 115 shows the resulting
curves over a range of 800 ° to 1, 200 °F for
coatings Numbers 1, 3, and 5 of Table 52.
In some instances the eye irritating charac-
teristic of gases discharged from equipment
has been more noticeable after a catalytic af-
terburner was installed. This is thought to be
due to the partial oxidation of organic substances
to aldehydes and organic acids. This condition
has been more apparent at operating tempera-
tures below 900°F. Little eye irritation or
odor is usually experienced at operating tem-
peratures above 900°F.
Efficiency
The efficiency of a catalytic afterburner depends
upon several variables, namely, contact of the
gases with the catalyst, uniform flow of the gases
through the catalyst bed, operating temperature,
catalyst surface area, nature of materials being
burned, and oxygen concentration. Catalyst
manufacturers specify a maximum s cfm per cata-
lyst unit.
r so
COATINGS
D OLEORESINOUS
4 EPOXY
• ALKYD
O PHENOLIC
* VINYL
* VINYL
SOLVENT
MINERAL SPIRITS
XYLOL BUTYL
MINERAL SPIRITS
MINERAL SPIRITS
HIBK
XYLOL ISOPHORONE
TEMPERATURE, °F
Figure 114. Hydrocarbon combustion efficiency
of catalytic afterburners (Krenz et al., 1957).
Table 52. COATING OVEN OPERATING CONDITIONS DURING TESTS
(Krenz, Adrian, and Ingels, 1957)
Coating
No.
1
2
3
4
5
6
Type
Vinyl
Vinyl
Epoxy
Phenolic
Oleoresinous
Alkyd
Quantity of
coating,
gal/hr
19
43
18.6
18. 5
17.5
8
Solvent
Xylol, isophorone
Methyl isobutyl ketone
Xylol, butyl cellosolve
Aromatic mineral
spirits
Aromatic mineral
spirits
Aromatic mineral
spirits
Quantity of
solvent,
Ib/hr
120
271
86
88
77
30
Oven temp
range, °F
Avg Max
350 370
340 370
350 390
414 430
425 475
290 300
Catalyst temp
range, °F
800 to 1, 200
800 to 1, 000
800 to 1, 200
730
800 to 1, 200
700 to 900
-------
Afterburners
185
900 1 000 I 100
TEMPERATURE •F
Solution:
1. Mass flow rate of contaminated gases:
(3,000 cfm)(60 min/hr) 74011, A,
———— -i—:—— ; = 13,7401b/hr
13. 1 ftj air/lb air
2. Heat required to increase the temperature of
the gases from 200° to 950°F:
Catalytic afterburner operating temperatures
of approximately 950°F have been found suf-
ficient to control emissions from most pro-
cess ovens.
Enthalpy of gas at 950°F = 222. 8 Btu/lb
(see Table D3, Appendix D)
Enthalpy of gas at 200°F = 33. 6 Btu/lb
Ah = 189. 2 Btu/lb
(13, 740)(189.2) = 2,600, 000 Btu/hour
Figure 115. Efficiency of a catalytic after-
burner as a function of catalyst temperature
(Krenz, Adrian, and Ingels, 1957).
The efficiency at various temperatures of several
catalytic afterburners serving paint bake ovens
processing a variety of coating materials is shown
graphically in Figure 115.
Design calculations
The following example illustrates some of the fac-
tors that must be considered in the design of a
catalytic afterburner.
Example 22
Given:
A catalytic afterburner is to be installed to incin-
erate the air contaminants discharged from a
direct-fired process oven. Visible emissions are
to be eliminated and odors are to be reduced. A
maximum gas discharge rate of 3,000 scfm at
200 °F has been determined. Assume the efflu-
ent gases have the same properties as air. (See
example for direct-fired afterburner for explana-
tion. )
Problem:
Determine the design features of a catalytic af-
terburner to incinerate the contaminants in the
exhaust gases.
3. Heat losses from afterburner due to radia-
tion, convection,and conduction:
Assume 10 percent of item 2. This appears
to be a reasonable estimate based on the
usual afterburner construction and operating
temperatures.
(2,600, 000)(0. 10) = 260, 000 Eta/hour
4. Total heat required by afterburner:
2,600,000 + 260,000 = 2,860,000 Btu/hour
5. Required natural gas volume capacity of
burner:
(See example for direct-fired afterburner for
explanation. )
Gross heating value of natural gas taken at
1, 100 Btu/ft^.
Net heat available at 950°F from the burning
of 1 ft of natural gas with theoretical air
is 785.2 Btu/ft3. (See Table D7, Appendix
7.)
2, 860,OOP
785.2
= 3,640 ft /hour
6. Maximum gross heat release rate required
of gas burner:
(3, 640)(1, 100) = 4,000,000 Btu/hour
7. Furnace combustion volume required for
gas burner:
-------
186
CONTROL EQUIPMENT FOR GASES AND VAPORS
Heat release rates in the range of 50, 000
Btu/hr-ft-* have been used for some cata-
lytic afterburner installations. This ap-
pears to be a reasonable value for design
purposes.
4.000,000 _ „ , 3
50,000 - 8°ft
Additional volume will be required owing to
displacement of catalyst elements.
8. Volume rate of gas burner combustion
products at 950°F:
With theoretical air, 1 ft of natural gas
yields 11. 45 ft"' of products of combustion
(see Table D7, Appendix D).
superficial gas velocity through the catalyst
bed of about 10 fps is satisfactory.
(3,640)(11.45)(950 + 460)
(60 + 460)(3,600)
/sec
9. Volume rate of contaminated gases at 950 °F:
(3,000)(95Q + 460)
(60)(60 + 460)
10. Total volume rate of gases in afterburner
at 950°F:
31.4 -1- 135.6 = 167. 0 ft3/sec
11. Number of catalyst elements required:
The type A catalyst element is 19 x 24 x
3-3/4 in. Experience has shown that a
(167. 0)(144)
= 5.27, .'. Use5 type A cata-
lyst elements
The type B catalyst element is 3 x 3-l/8x
5-1/2 in. and is rated by the manufacturer
at 5 to 15 scfm per element, depending upon
the application. Use 10 scfm per element.
3, OOP
10
= 300, . Use 300 type B catalyst
elements
Results of several tests of catalytic afterburners
are given in Table 53. The process equipment
and the afterburners in each of these tests are
briefly discussed below.
Test C-4IC
Test C-410 was conducted on a catalytic af-
terburner serving a conveyorized, gas-fired
paint bake oven. Sheets of tin plate.- were
roller coated with a vinyl coating and baked
at 350°F. The solvent-laden gas-^s were
vented to the preheat zone of the catalytic
afterburner where they were heated to 950 °F
by an atmospheric-type gas burner. A por-
tion of the solvent vapors waf burned by the
preheat burner. From the preheat zone,
the gases were drawn into an exhaust fan
located within the afterburner housing and
discharged through a bank of four catalyst
elements that use a metallic ribbon cata-
lyst support. A considerable portion of the
remaining solvent vapors was burned.
?able 53. STACK ANALYSES OF EMISSIONS FROM CATALYTIC AFTERBURNERS
Report series No.
Basic equipment
Burner type
Catalyst support
Afterburner combustion
chamber temp, °F
Inlet gas volume, scfm
Outlet gas volume, scfm
Particulate matter,
Ib/hr, inlet
Particulate matter,
Ib/hr, outlet
Control efficiency, %
Aldehydes, Ib/hr, inlet
Aldehydes, Ib/hr, outlet
C-410
Paint bake oven
Atmospheric
Metallic ribbon
1, 000
--
2, 800
6.7
1.4
79
0. 07
0. 31
C-374
Paint bake oven
Premix
Porcelain rod
800
5, 800
6, 100
4.4
2.6
41
0. 3
0.2
C-374
Paint bake oven
Premix
Porcelain rod
925
5,900
6,200
9.0
4. 5
50
0. 3
0. 4
C-374
Paint bake oven
Premix
Porcelain rod
1,050
5, 400
5, 800
9.9
3.6
64
0. 3
0. 2
C-374
Paint bake oven
Premix
Porcelain rod
1,200
5,400
5,900
7.7
2. 2
71
0.4
0. 5
-------
Boilers Used As Afterburners
187
It seems significant that the amount of alde-
hydes discharged from the afterburner is
approximately four times the amount enter-
ing. Large amounts of aldehydes are typi-
cal of incomplete combustion.
A series of tests was conducted on a second
catalytic afterburner serving a conveyorized
gas -fired paint bake oven. The afterburner
is a type wherein the catalyst elements are
small blocks of thin porcelain rods on which
the catalyst is deposited. Sheets of tin plate
were roller coated "with a vinyl coating and
then conveyed through the oven and baked
at about 320°F. An exhaust fan discharged
the gases and solvent vapors to a heat ex-
changer where the contaminated gases were
preheated before entering the afterburner.
The gases -were then further preheated by
gas burners before entering the catalyst
section. The sensible heat of the gases
discharged from the afterburner was used
to preheat the incoming contaminated gas-
es in the heat exchanger.
Test results show that an appreciable frac-
tion of the particulate matter was burned
by the preheat burners, an additional frac-
tion being burned in the catalyst section.
Moreover, the overall efficiency of the after-
burner in the burning of particulate matter
increased noticeably with increased after-
burnei temperature. There is, however,
no significant change in the concentration of
aldehydes in the inlet arid outlet gases.
Installation Costs
A survey of installation costs oi catalytic after-
burners reveals a general range from $5,00 to
$10. 00 per scfm contaminated gas.
BOILERS USED AS AFTERBURNERS
Fireboxes oi boilers and fired heaters can be
used, under proper conditions, as afterburners
to incinerate combustible air con tar. > inants .
This use is unique in that a basic source of
air contaminants, a boiler, is used to control
pollutants from another source. Boiler fire-
box conditions approximate those of a well-
designed afterburner, provided there are ade-
quate temperature, retention time, turbulence,
and flame. Oxidizatale contaminants, including
smoke and organic vapors and gases, can be
converted essentially to carbon dioxide and
water in boiler fireboxes.
The discussion of this section is limited to the
control of low-calorific-value gases and vapors
with common types of steam and hot water
boilers and heaters. When appreciable heat
is contained in the contaminated gases, the
firebox is usually of special design to take
advantage of the heat potential. These latter
units, commonly known as waste heat boilers,
are discussed in Chapter 9.
Completely satisfactory adaptations of boilers
for use as afterburners are not common. All
aspects of operation should be thoroughly eval-
uated before this method of air pollution con-
trol is used. The primary function of a boiler
is to supply steam or hot water, and whenever
its use as a control device conflicts with this
function, one or both of its purposes will
suffer. Some advantages and disadvantages of
boilers used as afterburners are shown in
Table 54.
CONDITIONS FOR USE
The determination to use a boiler as an after-
burner demands that the following conditions
exist:
1. The air contaminants to be controlled must
be almost -wholly combustible since a boiler
firebox cannot be expected to control non-
combustible pollutants. Inorganic dusts
and fumes deposit on heat transfer surfaces
and foul them with resulting losses in boiler
efficiency and steam-generating capacity.
Z. The volumes of contaminated gases must
not be excessive or they will reduce thermal
efficiencies in much the same way as ex-
cess combustion air does. The additional
volume of products of combustion will also
cause higher pressure drops through the
system, in some cases exceeding the draft
provided by existing boiler auxiliaries.
3. The oxygen content of the contaminated gases
when used as combustion air must be similar
to that of air to ensure adequate combustion.
Incomplete combustion can form tars or resins
that will deposit on heat transfer surfaces and
result in reduction of boiler efficiency. When
these contaminants exceed air pollution con-
trol standards for gas- or oil-fired boilers,
tube fouling will already have become a
major maintenance problem.
4. An adequate flame must be maintained con-
tinuously in the boiler firebox. High-low
or modulating burner controls are satis-
factory provided that the minimum firing
rate is sufficient to incinerate the maximum
volume of effluent that can be expected in
-------
188
CONTROL, EQUIPMENT FOR GASES AND VAPORS
Table 54. ADVANTAGES AND DISADVANTAGES OF USING A BOILER AS AN AFTER-
BURNER RATHER THAN A CONVENTIONAL DIRECT-FIRED AFTERBURNER
Advantages
Disadvantages
1. Large capital expenditure not required.
2. Boiler serves dual purpose as source of
process steam and as an air pollution
control device.
3. Auxiliary fuel not required for operation
of air pollution control device.
4. Operating and maintenance cost limited to
one piece of equipment.
5. Fuel saving, if effluent has some calorific
value (rare instances).
1. If air contaminant volumes are relatively
large, boiler fuel cost may be excessive.
2. High maintenance cost may be required
because of burner and boiler tube fouling.
3. Boiler must be fired at an adequate rate at
all times when effluent is vented to the fire-
box, regardless of steam requirements.
4. .Normally, two or more boilers must be
used, one as standby during shutdowns.
5. Pressure drop through boiler may be ex-
cessive if large volume of effluent intro-
duced into boiler causes back pressure on
exhaust system.
the boiler firebox. Obviously a burner
equipped with on-off controls would not be
feasible.
Boilers used as afterburners have successfully
controlled visible emissions from meat smoke-
houses and also obnoxious odors from rendering
cookers and from oil refinery processes involv-
ing cresylic and naphthenic acids, hydrogen sul-
fide, mercaptans, sour water strippers, ammonia
compounds, regeneration air from doctor treat-
ing plants, oil mists and vapors from process
columns, and so forth.
MANNER OF VENTING CONTAMINATED GASES
Like other types of controls, these units re-
quire a properly designed exhaust system to
convey air pollutants effectively from the point
of origin to the boiler firebox.
Contaminated gases may be introduced into the
boiler firebox in two ways: (1) Through the
burner, serving as combustion air, or (2)
downstream of the burner, serving as secon-
dary air.
Figures 116 and 117 show poor and good installa-
tions wherein the contaminated gases are intro-
duced through the burner. The oxygen content of
these gases must be nearly equivalent to that of
air to ensure good combustion. Excessive vol-
umes of nonoxidizing gases such as CO^, H£O,
and N£ can cause undesirable results ranging
from flame popping to complete outage of the burn-
er. Introducing contaminated gases through
the burner should promote good flame contact,
turburlence, temperature, and retention time.
CONTSMINAIED-MR DUCT-
Figure 116. Poor method of introducing contam-
inated air from diffuser to boiler firebox
through the burner air register. Oiffuser re-
stricts combustion air to burner. Moreover,
louver may partially close restricting flow of
contaminated air into boiler firebox.
Since the polluted gas stream furnishes a part
of the combustion air for the burner, less ad-
ditional air is required from the combustion air
system. Burner maintenance costs, however,
are higher. Contaminated gases should not
be introduced through the burner if a high
moisture content or corrosive gases and vapors
are present. In these cases, the gases should
be introduced into the boiler downstream of
the burner.
-------
Boilers Used As Afterburners
189
Figure 117. Good method of introducing contam-
inated air to boiler firebox through a custom-
made air register. There is good flame contact.
Contaminated air enters firebox through burner.
Note: type of burner is critical; contaminated
air is portion of combustion air; not applicable
where contaminated gases are corrosive.
Figures 118, 119, and 120 show both desirable
and undesirable methods of introducing con-
taminated gases downstream of the burner.
Gases must be carefully directed into the boil-
er firebox to ensure adequate flame contact.
An exhaust fan or a steam ejector is used to
convey the effluent through an exhaust system
from the source into the boiler firebox. Some-
times, a flame arrestor is installed to prevent
flashback. When gases of high moisture con-
tent must be incinerated, condensers are in-
stalled upstream of the boiler.
A reduction in boiler efficiency should be ex-
pected when gases are introduced directly in-
to the boiler firebox. In addition, incinera-
tion may not be complete, and partially oxi-
dized organics may be present in the products
of combustion. As a result, these particulates
deposit on boiler tubes and reduce heat trans-
fer. Because of these disadvantages, this
latter method should be used only when the
contaminated gases cannot be introduced di-
rectly through the burner.
ADAPTABLE TYPES OF EQUIPMENT
Boilers and Fired Heaters
Water-tube, locomotive, or HRT boilers and fired
heaters are the units most frequently used as after-
burners. Burners used with these units are usually
CONTAMINATED-AIR
DUCT ENTRANCE
Figure 118. Poor method shoeing entry of
contaminated air near boiler firebox rear
firewall. Flame contact is poor.
CONTAMINiTED-ftIR DUCT
Figure 119. Boiler firebox showing entry of
contaminated air through a diffuser in the
floor near the burner. The possibility for
flame contact is good. Note; Type of burner
is not critical; contaminated air is secondary
air for boiler; applicable where contaminated
gases are corrosive.
adaptable to incineration, and the fireboxes
are usually accessible. Thus, the contamina-
ted gases may be properly introduced either
through the burner or through the floor or
sides of the fireboxes.
Some types of boilers do not provide these fea-
tures. For example, polluted gases usually
-------
190
CONTROL EQUIPMENT FOR GASES AND VAPORS
EflLED OR OPEN TrPE.
CONTAMIMTED-iiR DUCT
Figure 120. Boiler firebox showing entry of
contaminated air through a duct at front of
boiler. Flame contact and mixing are poor.
cannot be introduced into the firing tube of a
Scotch marine boiler unless they are introduced
with the combustion air. Admitting contamina-
ted gases through the integral blower of a forced-
draft burner is normally not feasible.
Burners
Burner selection is greatly influenced by boiler or
heater firebox design, the method of introducing
polluted gases, and the characteristics of the pol-
lutants themselves.
Where gases are introduced as excess air through
the sides or floor of the firebox, any standard gas
or oil burner may be used. The gases must, how-
ever, be introduced near the burner end of the fire-
box to ensure adequate incineration.
Where contaminated gases are used as combus-
tion air, natural or induced draft is essential.
Multijet natural-gas burners and steam, pres-
sure, or air-atomizing oil burners are most
adaptable. The burner must be thoroughly main-
tained according to the character of contaminants
to be incinerated. Forced draft burners are not
recommended because of the probability of cor-
roding and fouling burner controls and blowers.
SAFETY
As with any afterburner or flare used to incin-
erate combustible gases, care must be taken to
prevent flashbacks and firebox explosions. This
problem is most acute when the contaminated
gas stream contains explosive hydrocarbon con-
centrations, for example, a. refinery flare. In
these instances, suitable flame arrestors are re
quired. Where continuing explosive concentra-
tions are likely, a control device other than a
boiler-afterburner is recommended. Flaring
or intermediate gas storage for use as fuel
might well be a more practical approach than
on-line incineration.
Contaminants in most exhaust gas streams are
normally 'well below explosive concentrations.
Ir. a few processes, however, combustible gas
concentrations can accumulate during shutdowns
with resultant explosion hazards on lightoff of
the boiler. For instance, a batch of raw or
partially cooked animal matter might be left
overnight in a rendering cooker ducted to a
boiler -in cinerator. This could generate enough
methane, hydrogen sulfide, and other organics
to produce an explosive mixture in the ductwork
leading to the boiler. If, subsequently, the
burner were ignited without first purging the
line, an explosion could occur. To avoid a
rare possibility such as this, both the boiler
firebox and the ductwork should be purged be-
fore igniting the burner.
Some fire hazard is created by the accumulation
of organic material in ductwork. Lines such as
these must usually be -washed periodically. The
degree oi org,anic accumulation can sometimes
be reduced by frequent steam purging or by
heating the dxictwork to prevent condensation.
DESIGN PROCEDURE
When evaluating a control system wherein a
boiler is to be used as an afterburner, one
should:
1. Determine the maximum volume, tempera-
ture, and characteristics of the polluted
gases to be vented to the boiler firebox;
2. ascertain that the exhaust system from the
source of the pollutant to the boiler firebox
is properly designed;
3. determine the manner in which the pollutants
are to be introduced into the boiler firebox;
4. calculate whether the boiler and burners
are of sufficient size and design to handle
the contaminated gases;
5. calculate the minimum firing rate at which
the boiler must be operated to ensure ade-
quate incineration;
-------
Boilers Used As Afterburners
191
6. provide that the firing rate does not fall
below the minimum rate determined in item 5.
The following example shows some of the factors
that must be considered in determining the feasi-
bility of using a boiler to incinerate exhaust gas-
es from four meat processing smokehouses.
Example 23
Given: Boiler data
Boiler, 150 hp, HRT type, multijet burner, gas
fired only--minimum instantaneous firing rate,
33. 2 cfm.
Automatic modulating controls, stack size, 30-
in. diameter x 40 ft high.
Assumptions:
Boiler operates at 80% efficiency, stack tem-
perature = 500°F, 20% excess air to burner.
Effluent data:
Maximum volume of effluent = 1,000 scfm,
(76. 4 lb/min) with smokehouse stacks damp-
ered. Minimum temperature of effluent =
100°F.
As sumptions:
Effluent gases have the same properties as
air. Exhaust system has been designed to con-
vey effluent gases properly from smokehouses
to boiler firebox. Effluent to enter boiler fire-
box as secondary air through diffuser in floor,
near burner end. Minimum incineration temper-
ature to be 1, 200°F.
Solution:
1. Btu input required to fire the 150-hp boiler
at rating:
1 boiler hp = 33,475 Btu/hr
(150 hp)(33,475 Btu/hr)
- Q g eff L = 6,277,000 Btu/hr
2. Natural gas flow required to fire boiler at
rating:
Gross heat standard of natural gas taken at
I, 100 Btu/ft3 at 60°F.
6,277,000 Btu/hr
= 5,707 cfh or 95. 1 cfm
1, 100 Btu/ft
3. Determine minimum firing rate for boiler:
Instantaneous minimum firing rate for boiler
determined by actual measurement = 38. 2 cfm
Therefore, minimum firing rate for boiler
38.2
= --•—- (100) = 40.2% of rating
4. Heat required from burner of boiler to raise
temperature of smokehouse effluent from
100° to 1, 200°F:
Enthalpy of gas (1,200°F) = 287.2 Btu/lb
(See Table D3 in Appendix D. )
Enthalpy of gas (100"F) =9.6 Btu/lb
(See Table D3 in Appendix D. )
(76. 4 lb/min)(277. 6 Btu/lb) = 21, 209 Btu/min
= 1,272,540
/—BOILER
/ HRT TYPE
/ 150 t.,
]
=5§
/• — s
t
i
°\
/— EFFLUENT GAS
/ MAX /Ol 1 ODD scfm
/ MIN TEHP 100 °F
] ,
*— rr1 > — DIFFUSED r y - <;
^ \ X-FAN ^~
^ BURNER
^-SMOKEHOUSE (TY
/
\
\
/
\
/
\
Figure 121. Sketch of proposed system.
Problem:
Determine whether use of a 150-hp HRT-type boiler
as an afterburner is feasible.
5. Natural gas flow required to supply 21,209
Btu/min:
Heat available at 1, 200 °F from the burning
3
of 1 ft of natural gas with 20% excess air
= 676.5 Btu/ft3 (see Table D7 in Appendix
D).
21,209 Btu/min ,, A .
— r— = 31.4 cfm
676. 5 Btu/ft
Since only 31.4 cfm is required to raise
temperature of smokehouse effluent from
100° to 1,200°F, minimum firing rate for
boiler is adequate.
6. Volume of products of combustion from boiler
firing at 150% rating with 20% excess air:
One ft of natural gas yields 13. 473 ft3 of
products of combustion (see Table D7 in
Appendix D).
-------
192
CONTROL EQUIPMENT FOR GASES AND VAPORS
Vol = '(142.7 scfm>(13.473) = 1,922.7 scfm
7. Total volume vented from boiler:
Volume of effluent (secondary air) = 1, 000 scfm
Volume of products of combustion = 1,922.7
Total volume vented = 1,000 + 1,922.7
= 2,922.7
8. Volume of gases vented at stack tempera-
ture of 500 °F:
IZ'922-7'
9. Stack velocity:
5, 396 cfm
Vel =
(60 sec/min)(4. 91 ft )
= 18. 32 ft/sec
Note: Stack: velocities not exceeding 30 ft/sec
are satisfactory.
10. Heat required to raise temperature of ef-
fluent from 100° to 500°F (stack tem-
perature):
Enthalpy of gas (500°F) = 106.7 Btu/lb
(See Table D3 in Appendix D. )
Enthalpy of gas (100 °F) = 9. 6 Btu/lb
(See Table D3 in Appendix D. )
Ah = 97. 1 Btu/lb
(76.4 lb/min)(97. 1 Btu/lb) = 7, 418 Btu/min
11. Natural gas flow required to supply 7, 418
Btu/min:
The net thermal energy per ft natural gas
above that required to bring the effluent to
the stack temperature of 500 °F = 878 Btu/ft3
(see Table D7 in Appendix D).
7, 418 Btu/min
Vol gas = ::—
878. 0 Btu/ft
1. 45 cfm
12. Incremental cost of natural gas (assume
rate of $0. 50 per 1, 000 ft3):
ft
S.45
60 min 24 hr $0. 50
x —; x
hr day
- $6. 08/day
1, 000 ft
13. Cost of operating a direct gas-fired after-
burner operating at same temperature
(neglecting initial capital expenditure):
3
31.4
ft
60
hr
24 hr
day
50
1, 000 ft"
= $22.61/day
Problem note: Calculations indicate that use of
this boiler as an afterburner is feasible. The
boiler is fired at an adequate rate at all times,
and excessive volumes of effluent are not vented
to this boiler firebox. Costs, including initial
capital expenditures, are nominal. Some ad-
ditional cost might be necessary to provide
more draft to offset increased pres sure drops
through the boiler.
TEST DATA
Tests have been conducted on several boilers
used as afterburners. The majority of tests
have been on boilers used to incinerate the ef-
fluent from meat smokehouses. One test, how-
ever, includes a boiler used to incinerate partial-
ly condensed vapors from rendering cookers.
Table 55 summarizes these test results and shows
the apparent efficiencies of boilers in control-
ling combustion contaminants, organic acids, and
aldehydes. Installations were such that tests
could not be conducted with the boilers operating
under identical conditions unless the contamina-
ted gases were vented to the boiler fireboxes.
ADSORPTION EQUIPMENT
Adsorption is the name for the phenomenon in
•which molecules of a fluid contact and adhere
to the surface of a solid. By this process,
gases, liquids, or solids, even at very small
concentrations, can be selectively captured
or removed from airstreams with specific
materials known as adsorbents. The material
adsorbed is called the adsorbate.
A change in the composition of the fluid con-
tacting the adsorbent results when one or
more of the components are adsorbed by the
adsorbent. The mechanism of this process
is complex, and while adsorption can occur at
all solid interfaces, it is usually small unless
the solid is highly porous and possesses fine
capillaries. The most important character-
istics of solid adsorbents are their large sur-
face-to-volume ratios and preferential affinity
for individual components.
-------
Adsorption Equipment
193
Table 55. TEST DATA ON BOILERS USED AS AFTERBURNERS
Equipment
tested
Volume of gases, scfm
Stack
Boiler inlet
Combustion contami-
nants, Ib/hr
Inlet
Outlet
Efficiency, %d
Organic acids, Ib/hr
Inlet
Outlet
Efficiency, %d
Aldehydes, Ib/hr
Inlet
Outlet
Efficiency, %d
426-hp boiler,
water -tube type,
gas fireda
8, 700
1, 600
2.4
0. 45
84
1. 5
0. 56
60
0.22
0. 09
59
Two 268-hp boilers,
common stack
water-tube type,
gas fireda
10, 300
2, 930
4. 6
0. 53
89
2. 7
0.64
78
0. 39
0. 40
0
200-hp boilers,
water-tube type,
gas firedb
4, 700
2, 400
2. 7
1. 6
41
2. 2
1. 4
36
0. 39
0. 30
23
Two 113-hp boilers,
common stack
locomotive type,
gas fireda
3,800
320
0. 19
0. 16
16
0. 12
0
100
0. 03
0
100
150-hp boiler,
HRT type,
gas firedc
3,400
470
0.74
0. 52
30
0.35
0. 14
60
0.012
0.09
0
3,600
750
0. 73
0. 71
3
0.44
0. 38
14
0. 03
0. 18
0
aMeat smokehouse effluent was admitted into boiler firebox through the multijet burner.
Meat smokehouse effluent was admitted into boiler firebox through diffuser located at front of firebox floor.
cRendering cooker effluent was admitted into boiler firebox through diffuser located at rear of firebox floor. Two tests were
run. The use of this boiler as an afterburner has been discontinued, primarily because the minimum firing rate of the boiler
was insufficient to incinerate air contaminants.
d
Efficiency shown is Apparent Efficiency. Boilers could not be tested unless air contaminant •were vented to it.
Many theories have been advanced to explain
the selective adsorption of certain vapors or
gases, the exact mechanism being still dis-
puted. In some cases, certainly, adsorption
is due to chemical combination of the gas with
the free valences of atoms on the surface of the
solid in the monomolecular layer, as was pro-
posed by Langmuir in 1916 (Glasstone, 1946).
Other investigators hold that the adsorbents
exert strong attractive forces, so that many
adsorbed layers form. These layers are under
pressure, partly because of layers on top and
because of the attractive force of the surface
of the adsorbent. In other cases, the evidence
indicates that adsorption is due to liquefaction
of the gas and its retention by capillary action
in the exceedingly fine pores of the adsorbing
solid. In many cases, the phenomena are
probably superimposed. The adsorptive power
of activated charcoal is due mainly to molecu-
lar capillary condensations "while the adsorp-
tive power of silica gel is due mainly to cap-
illary condensation. Note, however, that the
adsorptive power of any solid adsorbent may
vary appreciably -with the method of prepara-
tion as well as with the nature of the gas or
vapor adsorbed (Walker et al. , 1937).
In most processes involving adsorption, the
operation involves three steps. First, the
adsorbent is contacted with the fluid, and a
separation by adsorption results. Second,
the unadsorbed portion of the fluid is sepa-
rated from the adsorbent. In the case of gas-
es, this operation is completed on their pas-
sage through the adsorbent bed. Third, the
adsorbate is removed from the adsorbent,
which thereby regenerates the adsorbent. In
some cases the adsorbent is regenerated with-
out recovery of the adsorbate, as in the de-
colorizing of sugar solutions with bone char
and the treatment of lubricating oils "with
Fuller's earth. In the treatment of domestic
•water "with finely divided activated carbon,
both the adsorbent and the adsorbate are sep-
arated from the fluid and discarded.
Regeneration, -which involves raising the temper-
ature of the adsorbent, may be performed by
several methods, depending upon the adsorbate.
In the examples cited previously, where the ad-
sorbate has no economic value, the Fuller's
earth and bone char are heated directly with hot
gases. In the recovery of chlorine and sulfur
dioxide from silica gel, the adsorbent is heated
indirectly with a hot brine. In the recovery of
solvents, low-pressure steam is used and the
condensed vapors are separated from the water
by decantation or distillation, or both.
Adsorption can be specific and can, therefore,
be used to separate gases from gases, as in
the elimination of toxic materials such as sul-
fur dioxide or chlorine; the removal of vapor-
ized liquids from air, as in the capture of sol-
vents in surface coating operations; the re-
moval of colloids or suspended solids from
solutions, as in the decolorizing, clarification,
and purification of solutions; the removal of
-------
194
CONTROL EQUIPMENT FOR GASES AND VAPORS
ions from solutions, as in -water softening; and
the removal of dissolved gases in solution to
control odors or tastes,as in water treatment.
ly mixed. The amount added is sufficient only
to effect the purification; the separation is
made by settling or filtration.
TYPES OF ADSORBENTS
Solids possessing adsorptive properties exist
in great variety. Some of these solids and
their industrial uses are as follows:
Activated carbon
Alumina
Bauxite
Bone char
Decolorizing carbons
Fuller's earth
Magnesia
Silica gel
Strontium sulfate
Solvent recovery, elim-
ination of odors, purifica-
tion of gases
Drying of gases, air,
and liquids
Treatment of petrole-
um fractions; drying
of gases and liquids
Decolorizing of sugar
solutions
Decolorizing of oils, fats,
and waxes; deodorizing of
domestic water
Refining of lube oils and
vegetable and animal oils,
fats, and waxes
Treatment of gasoline and
solvents; removal of me-
tallic impurities from
caustic solutions
Drying and purification
of gases
Removal of iron from
caustic solutions
Activated carbon, silica gel, alumina, and
bauxite are used for selectively adsorbing
certain gaseous constituents from gas streams.
Activated carbon adsorbs organic gases and
vapors, even when \vater is present in the gas
stream. Silica gel, in the absence of water
vapor, adsorbs organic and inorganic gases;
however, in the presence of water vapor, it
adsorbs water vapor almost exclusively. Alu-
mina and bauxite are used chiefly in dehydra-
tion. Bone char, decolorizing carbon, Fuller's
earth, magnesia, and stiontium sulfate are
used mainly in removing impurities from solu-
tions. Bone char and Fuller's earth are normal-
ly used as beds through which the solutions are
allowed to percolate. Decolorizing carbon,
magnesia, and strontium sulfate are added to
the solution in finely divided form, and intimate -
USE OF ACTIVATED CARBON IN AIR POLLUTION CONTROL
Generally, the concentrations of the organic ma-
terials discharged to the atmosphere are relative
ly small and are usually governed by fire preven-
tion regulations and the health hazard standards
(Barry, I960). The latter is usually smaller
and in many cases is the governing concentration.
Concentrations may vary from 50 to 3, 000 ppm.
Activated carbon is the adsorbent most suitable
for removing organic vapors. Carbon adsorbs
substantially all the organic vapor from the air
at ambient temperature regardless of variation
in concentration and humidity conditions. Be-
cause the adsorbed compounds have practically
no vapor pressure at ambient temperatures,
the carbon system is particularly adapted to
the efficient recovery of solvents present in
air in small concentrations. This means the
system can always be designed for operation
without hazard because the vapor concentration
is always below the flammable range.
Since activated carbon adsorbs all the usual low-
boiling solvent vapors, it can be used to recover
practically any single solvent or any combina-
tion of low-boiling solvents. Turk and Bownes
(1951) state that the limitation for molecules
capable of removal by physical adsorption is
that they must be higher in molecular weight
than the normal components of air. In general,
removal of gaseous vapors by physical adsorp-
tion is practical for gases -with molecular weight
over 45. Probably the only solvent used -with a
molecular weight below 45 is methanol.
Saturation
Adsorption of a vapor by activated carbon ap-
parently occurs in two stages. Initially, ad-
sorption is rapid and complete, but a stage is
reached in which the carbon continues to re-
move the material but at a decreasing rate.
Eventually, the vapor concentration leaving
the carbon equals that of the inlet. At this
point the carbon is saturated, that is, it has
adsorbed the maximum amount of vapor that
it can adsorb at the specific temperature and
pressure. This saturation value is different
for each vapor and carbon. It is determined
experimentally by passing dry air saturated
with the g2.s or vapor, with temperature and
pressure maintained constant, through a
known amount of carbon until the carbon
ceases to increase in weight. Under these
conditions, the carbon is saturated with the
adsorbate.
-------
Adsorption Equipment
195
Retentivity
The retentive capacity of an activated carbon is
a more useful figure. It represents the amount
of adsorbate that a carbon, initially saturated,
can retain -when pure air is passed through the
carbon with the temperature and pressure main-
tained constant. This indicates the weight of the
particular gas or vapor that the carbon can com-
pletely retain. This is called the retentivity of
the carbon and is expressed as the ratio of the
weight of the adsorbate retained to the weight of
the carbon.
Breakpoint
When an air vapor mixture is passed over carbon,
adsorption is 100 percent at the beginning, but as
the retentive capacity of the carbon is reached,
traces of vapor appear in the exit air. This stage
of adsorption is called the breakpoint of the carbon,
"beyond which the efficiency of removal decreases
rapidly. As the flow of air is continued, addition-
al amounts of solvent are adsorbed, but the con-
centration of vapor in the exit air (Figure 122)
increases and eventually equals that in the inlet,
at which time the carbon is saturated at the
particular operating conditions.
Adsorption of Mixed Vapors
The adsorption phenomenon becomes somewhat
more complex if the gas or vapors to be adsorbed
consist of not one but several compounds. The
50
TIME, hours
Figure 122. Adsorption efficiency, single sol-
vent (Report No. 8, Experimental Program for
the Control of Organic Emissions from Protec-
tive Coating Operations, Los Angeles County
Air Pollution Control District, Los Angeles,
Cali-f., 1961).'
adsorption of the various components in a mix-
ture such as this is not uniform, and generally.
these components are adsorbed in an approxi-
mately inverse relationship to their relative
volatilities. Hence, when air containing a mix-
ture of organic vapors is passed through an
activated-carbon bed, the vapors are equally
adsorbed at the start; but as the amount of the
higher boiling constituent retained in the bed
increases, the more volatile vapor revaporizes.
After the breakpoint is reached, the exit vapor
consists largely of the more volatile material.
At this stage, the higher boiling component has
displaced the lower boiling component, and this
is repeated for each additional component, as
shown in Figure 123. This property of activated
carbon is the basis for hypersorption, a process
used for the separation of low-boiling hydro-
carbons. In the control of the discharge of or-
ganic vapors to the atmosphere, the adsorption
cycle should be stopped at the first breakpoint
as determined by the detection of vapors in
the discharge.
40
30
TIME hours
Figure 123. Adsorption efficiency, three-com-
ponent lacquer solvent (Report No. 8, Experi-
mental Program for the Control of Organic
Emissions from Protective Coating Operations,
Los Angeles County Air Pollution Control Dis-
trict, Los Angeles, Cal if., 1961).
Heat of Adsorption
The amount of organic vapors adsorbed by
activated carbon is a function of the boiling point,
molecular weight, concentration, pressure, and
temperature, Since adsorption is an exothermic
process, heat is liberated, which increases the
temperature of the carbon bed, and adsorption
-------
196
CONTROL EQUIPMENT FOR GASES AND VAPORS
may be necessary to provide cooling. The
same result can be obtained by diluting the gas
as it enters the adsorber. The vapor concentra-
tions encountered in paint spraying or coating
operations result in a temperature rise of about
15°F (Elliott et al. , 1961) and do not seriously
affect the capacity of the adsorbent. On the
other hand, the use of activated carbon to cap-
ture vaporized organic compounds at relative-
ly large concentrations, such as the discharge
from the filling of gasoline tanks, can result
in a temperature rise that can reach dangerous
levels.
sorbent. The steam consumption per pound of
solvent varies with time and the solvent adsorbed.
This is shown in Figure 124. The ratio of the
pounds of steam used per pound of perchloroethyle
recovered is plotted for 15-minute intervals. Thi;
reaches a minimum of about 4. 7 pounds after an
elapsed time of 90 minutes and then rises sharply.
The pounds of solvent recovered reaches a maxi-
mum at this time and then decreases. In Figure
124, the desorbing of toluene follows the same
pattern except that the steam consumption is high-
er. This is to be expected since its latent heat is
greater.
Carbon Regeneration
A desirable feature of using activated carbon
in the control of solvent emissions is its ability
to recover the adsorbed solvents on regenera-
tion. To remove the adsorbate from the car-
bon, the carbon must be heated to a tempera-
ture above that at which the solvents •were ad-
sorbed. Also essential to the process is a
carrier to remove the vapors released.
Regeneration is accomplished by passing a hot
gas through the carbon bed. Saturated steam
at low pressure, up to 5 psig, is the usual
source of heat and is sufficient to remove most
solvents. Steam superheated to as high as
650 °F may, however, be necessary to reactiv-
ate the carbon to its original condition (Barry,
I960). This is necessary when the solvent
adsorbed contains high-boiling constituents
such as are found in mineral spirits. Normal-
ly the flow of steam passes in a direction op-
posite to the flow of gases during adsorption.
With this arrangement, the steam passes up-
ward through the carbon. The steam through
the bed is only 1/5 to 1/10 of the air velocity
and is too low to initiate any boiling or crater-
ing of the bed. This counter cur rent flow is an
advantage in regeneration because a solvent
gradient exists across the adsorbent bed and,
depending on the concentration of adsorbate
and bed depth, the inlet side of the bed may
be saturated before the outlet reaches the
breakpoint. Thus, with countercurrent re-
generation, the solvent, driven out of the ad-
sorbent from the outlet side by the incoming
steam, will in turn start to remove vapor at
the inlet before it becomes heated, since it
is already saturated. This results in lower
steam consumption.
Steam requirements depend on external heat
losses as well as the nature of the solvent. The
heat liberated during adsorption is greater
(Mantell, 1961) than the heat of liquefaction, and
this difference may be large with an active ad-
\
60 80 100 120 140 160
ELAPSED TIME, minutes
Figure 124. Steam consumption per pound of sol-
vent recovered (Report No. 8, Experimental Pro-
gram for the Control of Organic Emissions from
Protective Coating Operations, Los Angeles
County Air Pollution Control District, Los
Angeles, Calif., 1961).
After the solvent is stripped, the carbon is not
only hot but is saturated with water. Cooling
and drying are usually done by blowing solvent-
free air through the carbon. The ensuing evap-
oration of the moisture is helpful in removing
the heat in the carbon. In surface-coating
operations, where the solvent vapors may con-
tain some relatively high-boiling constituents,
high-temperature stripping of the carbon is
periodically necessary to remove these com-
pounds. Superheated steam of about 650°F is
required (Elliott et al. , 1961), or the capacity
of the carbon is eventually reduced. Air must
not be used in cooling the carbon under these
conditions because of danger of a fire or an
explosion.
EQUIPMENT DESIGN
Barry (I960), reviewing the latest developments
on evaluating adsorption as a unit operation,
-------
Adsorption Equipment
197
concludes that adequate design and scaleup proce-
dures are not available in the chemical literature.
Manufacturer s of adsorbents have, however, ac-
cumulated much information on a confidential
basis with their clients. For the larger per-
centage of processes discharging organic vapors
to the atmosphere, such as dry cleaning, de-
greasing, paint spraying, tank dipping, and sol-
vent extracting, packaged equipment is available
that is suitable if certain precautions are taken.
These factors are discussed in the following
paragraphs.
A research program also was undertaken by
the Los Angeles County Air Pollution Control
District in conjunction with the United States
Public Health Service (Elliott et al. , 1961)
to develop some much-needed design data
and evaluate methods for the removal of
organic air contaminants.
In the capture and removal of organic com-
pounds, the vapor-laden air is passed through
a layer of activated carbon. The layer can be
either fixed or movable. The enclosure for a
simple fixed bed may be a vertical or a hori-
zontal cylindrical vessel. If more than one
carbon bed in a single vessel is used, the beds
are usually arranged as shown in Figure 125.
Multiple beds such as these are best suited to
a vertical vessel. Another type of fixed bed
is arranged in the shape of a cone, as shown
in Figures 126 and 127. It can be used in either
a vertical or horizontal enclosure and has cer-
tain advantages over the flat bed, as enumerated
later in this section.
A movable bed is shown in Figures 128 and 129.
In this design, the carbon bed is contained in a
drum, which rote^tes within an enclosiire.
Fixed-Bed Adsorber
The type of enclosure used 1o house an activated-
carbon adsorber with a fixed bed depends primarily
upon J:he volume oi gas fo be ha/idled and the allow-
abl< j L'essjre dr"p. The simnicFt equipment for
a Ixrd-bcd adsorber is a vertical, cylindrical
ve-;<-:el fitred with a perforated supporting screen
To.' .'i : caibon. Ibi gas stream, r ontairiing the
d.i-'- , pacers the vejj;] -=it the top -:..id flows
-!•.. - 11 '...i-ou:'/! the carbon bed. Dovi iilow allows
.- - '•- \.t>'.if.r g..',= ,'e1oci;,ieri. In upfli. o.r, the
.,-;_', >-_ ;y ': i - ''3 T.'d" i-i'^ed below a value that
f.reV" viT-3 ih'i -oiiijtij of rh' carbon, since this
resul!,'; i n cr.j.terLng and alirition of the adsorbent,
A c-in^le fixed be'! -unit is sa.tisi'actory if process
downtime is availaoie for regeneration of the
carbon. .he horizontal, cylindrical vessel with
a bed par ilit,! to the axis is normally used when
large volumes of gas muci be handled.
VAPOR TO CONDENSER
CARBON
Figure 125. Cross-section of adsorber with four
fixed beds of activated carbon (Report No. 8,
Experimental Program for the Control of Organic
Emissions from Protective Coating Operations,
Los Angeles County Air Pollution Control Dis-
trict, Los Angeles, Calif., 1961).
Figure !26, Top; Horizontal adsorber on the de-
scThing cycle with the superheated steam entering
at the apex (1). Condenser is located at the va-
por -iiitlei. ''2). Bottom: Horizontal carbon ad-
sorber. On the adsorption cycle the vapor-laden
air enters at the apex of the cone. The steam
enters eitner at the apex or at the bottom of the
cone for desorption (Report No. 3, Experimental
Program for the Control of Organic Emissions from
Protective Coatings, Los Angeles County Air Pol-
lution Control District, Los Angeles, Callf., 1959).
-------
198
CONTROL EQUIPMENT FOR GASES AND VAPORS
VAPOR-LADEN
AIR IN
Figure 127. Top: Diagrammatic sketch of verti-
cal adsorber with two cones, permitting studies
on different depths of carbon beds. Bottom:
Vertical cone adsorber in operation.
For the capture of vapors in a continuous oper-
ation, a minimum of two of these units is desir-
able. With this arrangement, one unit is ad-
sorbing v/hile the other is being stripped of sol-
vent and regenerated. Sufficient time or means
must be available to cool this unit to nearly
ambient temperature before it is returned to
service. A schematic diagram of this unit is
presented in Figure 130. The vapor-laden air
enters the first adsorber and passes down-
ward through the carbon bed, where it is di-
vested of its vapor, and then passes out to the
atmosphere. During this period, the second
adsorber is stripped of its adsorbate
Regeneration and cooling of the adsorbent usu-
ally determines the cycle time that may be
used. The stripping cycle must thus allow
sufficient time for the adsorbent to cool before
it is returned to the adsorption system. Re-
generation releases the bulk of the adsorbed
vapor rapidly, the rate reaching a maximum
early, then slowly trailing off as regeneration
is continued. No attempt is made to remove
all the adsorbate.
In Figure 131, a curve is shown in which the
pounds of toluene and perchloroethylene re-
covered are plotted against elapsed time, and
Figure 132 shows the pounds of steam per
pound of solvent for each 15-minute period
during stripping. The steam consumption is
approximately constant (Elliott et al. , 1961),
and to continue heating of the carbon bed until
all the solvent is removed would not be eco-
nomical in terms either of steam or time. It
is usually discontinued far short of this point.
This does, however, reduce the capacity of the
unit in the adsorption cycle.
Normally two adsorbing units are sufficient
if the regeneration and cooling of the second
bed can be completed before the first unit has
reached the breakpoint in the adsorbing cycle.
With three units it is possible to have one bed
adsorbing, one cooling, and one regenerating.
Vapor-free air from the adsorbing unit is used
to cool the unit just regenerated. An installa-
tion such as this is shown in Figure 133. By
operating two of the units in series, greater
adsorbing capa.city can be realized with the
same size bed. The air from the first bed,
after being stripped of vapor, is passed through
the second bed, which has been regenerated
but is still hot and wet. By using the vapor-
free air from the first unit to remove this
moisture, the ensuing evaporation of the water
effectively cools the carbon. After it cools,
it can more effectively adsorb and the first
bed can then be operated beyond its break-
point, -which increases its capacity. In the
meantime the third bed is regenerated. This
should be completed before the breakpoint
is reached in the second bed. A fourth bed
may also be used. One arrangement -would
be to have two units in parallel adsorbing and
both discharging to a. third unit, which is on
the cooling cycle while the fourth unit is be-
ing regenerated. This arrangement is com-
plex, and the increase in efficiency and capac-
ity may not justify the added cost.
-------
Adsorption Equipment
199
MUTING MSMMR
MOTOR
FMi
FILTER
COOLER
AIR AND SOLVENT
VAPOR IN
ACTIVE CARBON
STRIPPED AIR OUT
STEAM IN
ACTIVE CARSON
-STEAM AND SOLVENT
VAPOR OUT
Figure 128. Left: Diagrammatic sketch of a rotating fixed-bed continuous adsorber showing the
path of the vapor-laden air to the carbon bed. Right: Cut of continuous adsorber showing path
of steam during regeneration (Sutcliffe, Speaknan Canada, Ltd. Hamilton Ontario)
Figure 129. A continuous carbon adsorber serv-
ing a lithograph press. (Continental Can Co Inc
Robert Gair Div., Los Angeles, Calif.).
Figure 130. Diagrammatic sketch of a two-unit,
fixed-bed adsorber.
Conical fixed-bed adsorber
A cone-shaped bed is one bed configuration
that can be used where a relatively low pres-
sure drop is desired (Elliott et al. , 1961).
-------
200
CONTROL EQUIPMENT FOR GASES AND VAPORS
i on
ELAPSED TI»E minutes
Figure 131. Pounds of solvent recovered versus
time (Report No. 8, Experimental Program for
the Control of Organic Emissions from Protec-
tive Coating Operations, Los Angeles County
Air Pollution Control District, Los ftngeles,
Calif., 1961).
A comparison of this type of bed with a flat
bed is shown in Table 56. Both beds are
the same diameter and contain about the
same weight of carbon, yet the pressure
drop through the cone-shaped bed is less
than half that through the flat bed, even
when the volume of air passing through the
cone-shaped bed is more than twice that
through the flat bed. This cone carbon con-
tainer can be modified to a cylinder config-
uration with similar properties.
Continuous Adsorber
A continuous, activated-carbon, solvent recov-
ery unit is available. This unit consists of a.
totally enclosed, rotating drum carrying the
bed. Figure 128 shows the cutaway view of the
unit. The filtered air containing the solvent
vapor is delivered by the fan into the enclosure
and in turn enters ports to the carbon section.
These ports allow the solvent-laden air to enter
i
i^ *
r\
20 40 60 80 100 120 140 160
ELAPSED TIME minutes
Figure 132. Pounds of solvent recovered in 15-
minute intervals (Report No. 8, Experimental
Program for the Control of Organic Emissions
from Protective Coating Operations, Los Angeles
County Air Pollution Control District, Los
Angeles, Cali f., 1961).
the area above the carbon bed. From here it
passes through the bed and enters a similar
space on the inside of the cylindrical bed. It
then Leaves this enclosure through ports located
at the and o:" the drum opposite the entrance.
The vapor-free air travels axially to the drum
and is discharged to the atmosphere. The
steam, in the regeneration of trie carbon, enters
through a, row oi ports by means of a slide val ^e
as the cylinder rotates. The solvent and steam
leavt through a second row of ports, which is
served by a similar slide valve, and are sepa-
ratee continuousIv by decantation.
Pressure Drop
The pressure drop through the carbon bed is
a furc'don of the1 gas velocity, bed depth, and
the carbon particle size. Mantell (1961) pre-
Table 56. EXPERIMENTAL RESULTS OF FLAT- AND CONE-BED ADSORBERS
(Report No. 8, Experimental Program for the Control of Organic Eiriissions
from Protective Coating Operations, Los Angeles County Air Pollution Con-
trol District, Los Angeles, Calif., 1961).
Adsorber type
Enclosure
diameter,
in.
Air volume
cfm
Commercial flat bed
Vertical cone
36
36
5 50
1, 350
Air velocity; Pressure drop
Weight of j Carbon
'I through bed,! across adsorber. | carbon, bed depth,
fpm j in. H^O lb | in.
75 4.25 j 400
71 ! 1.81 352
-------
Vapor Condensers
201
VAPOR-LADEN AIR
CONDENSER
DECANTER
CARBON-^
VAPOR-FREE AIR
Figure 133. Diagrammatic sketch of a three-unit operation of a fixed-
bed adsorber showing No. 1 and No. 2 adsorbing in series and No. 3 re-
generating. Second cycle, No. 2 and No. 3 will be adsorbing with No.
1 regenerating. Final cycle, No. 3 and No. 1 will be adsorbing with
No. 2 regenerating.
sents three graphs in which pressure drop in
inches of water for different velocities is plot-
ted against bed densities in pounds per square
foot of bed area for several activated carbons
of different meshes. Carbon Products Division,
Union Carbide Corporation (1 955), presents an
empirical correlation representing the pres-
sure drop through a carbon bed at air veloci-
ties from 60 to 100 fpm against bed depth in
inches for two carbon mesh sizes. With this
empirical formula, Figure 134 was prepared
covering velocities from 60 to 140 fpm and for
bed depths of 10 to 50 inches. In the Report
No. 8 of the Experimental Program for the
Control of Organic Emissions (1961) pres-
sure drops for multiple-tray cone carbon
adsorbers are presented based on Union Car-
bide Corporation's empirical correlation, as
shown in Table 56. Note that, except for the
horizontal-cone and four-tray adsorber, the
pressure drop also includes that resulting
from the abrupt directional change of the air-
stream at both inlet and outlet.
OPERATIONAL PROBLEMS
Participate Matter
An activated-carbon adsorption bed should be
protected from particulate matter that can
coat the surface of the carbon. Without this
protection, the effective area and the ability
to adsorb will be impaired, and the life of
the carbon will be reduced if the material is
not removed by regeneration. In paint-spray-
ing operations (Elliott et al. , 1961) it -was
found that the carbon adsorbers could be
adequately protected from particulate
matter with efficient filters without exces-
sive increase in the total pressure drop.
Corrosion
Corrosion of adsorbers occurs when steam
is used in stripping solvents from activated
carbon. The amount of this corrosion is in-
tensified with increased steam temperature.
Corrosion can be controlled or reduced by
the use of stainless steel or by application of
a protective coating of a baked phenolic resin.
Polar and Nonpolar Compounds
Polar and nonpolar solvents are equally ad-
sorbed by activated carbon, but the recovery
of polar compounds on stripping with steam
requires an additional step of fractionation
by distillation to effect a separation from
the aqueous solution.
VAPOR CONDENSERS
Air contaminants can be discharged into the
atmosphere in the form of gases or vapors.
These gases or vapors can be controlled by
several different methods, for example, ab-
sorption, adsorption, condensation, or incin-
-------
20Z
CONTROL EQUIPMENT FOR GASES AND VAPORS
30
20
CARBON SIZE'. 4-6 MESH (TYLER)
„ , , V .1 56
A? = 9 370 (Too)
AP = PRESSURE DROP, inches cf water"
D = BED DEPTH, inches
V = VELOCITY, fpm
10
to
20
BED DEPTH, inches
30
40
50
Figure 134. Pressure drop versus carbon bed depth at various air velocities (Bulletin:
Solvent Recovery, 1955, Union Carbide Corporation, New York, N.Y.).
eration. In specific instances, control of
vapor-type discharges can best be accom-
plished by condensation. Other applications
require a condenser to be an integral part
of other air pollution control equipment. In
these cases, a condenser reduces the load
on a more expensive control device or re-
moves vapor components that may affect the
operation or cause corrosion of the main con-
trol element.
TYPES OF CONDENSERS
Surface and Contact Condensers
Vapors can be condensed either by increas-
ing pressure or extracting heat. In practice,
air pollution control condensers operate
through removal of heat from the vapor. Con-
densers differ principally in the means of
cooling. In surface condensers, the coolant
does not contact the vapors or condensate.
In contact condensers, coolant, vapors, and
condensate are intimately mixed.
Most surface condensers are of the tube and shell
type shown in Figure 135a. Water flows inside
the tubes, and vapors condense on the shell side.
Cooling water is normally chilled, as in a cooling
tower, and reused. Air-cooled surface conden-
sers, as shown in Figure 135b, and some water-
cooled units condense inside the tubes. Air-
cooled condensers are usually constructed "with
extended surface fins. Typical fin designs are
shown in Figures 135c and d. A section of an
atmospheric condenser is shown in Figure 135e.
Here vapors condense inside tubes cooled by
a falling curtain of water. The -water is cooled
by air circulated through the tube bundle. The
bundles can be mounted directly in a. cooling
tower or submerged in water. Contact con-
densers employ liquid coolants, usually water,
which come in direct contact with condensing
vapors. These devices are relatively uncom-
plicated, as shown by the typical designs of
Figure 135f, g, and h. Some contact con-
densers are simple spray chambers, usually
with baffles to ensure adequate contact. Others
are high-velocity jets designed to produce a
vacuum.
-------
Vapor Condensers
203
-*-
g
Figure 135. Types of condensers. Surface condensers: (a) Shell and tube, Schutte and Koerting Co.
Cornwell Heights, Penn.,(b) fin fan, Hudson Engineering Corp., Houston, Texas, (c) finned hairpin
section, Brown Fintube Co., Elyria, Ohio, (d) integral finned section, Calumet
Park, Mich.,and (e) tubular, Hudson Engineering Corp., Houston, Texas. Contact
(g) jet, Schutte and Koerting Co., Cornwell Heights, Penn., and (h) barometric
Co., CornwelI Heights, Penn.
& Hecla Inc., Al len
condensers: (f) Spray,
Schutte and Koerting
-------
204
CONTROL, EQUIPMENT FOR GASES AND VAPORS
In comparison with surface condensers, con-
tact condensers are more flexible, are simpler,
and considerably less expensive to install.
On the other hand, surface condensers re-
quire far less -water and produce 10 to 20tim.es
less condensate. Condensate from contact
units cannot be reused and may constitute a
waste disposal problem. Surface condensers
can be used to recover salable condensate,
if any. Surface condensers must be equipped
with more auxiliary equipment and generally
require a greater degree of maintenance.
Contact condensers normally afford a greater
degree of air pollution control than surface
condensers do because of condensate dilution.
With direct-contact units, about 15 pounds of
60°F water is required to condense 1 pound
of steam at 212°F and cool the conden-
sate to 140"F. The resultant 15:1 dilution
greatly reduces the concentration and vapor
pressure of volatile materials that are misci-
ble or soluble in water.
TYPICAL INSTALLATIONS
Condensers in Control Systems
Condensers collect condensable air contami-
jjants and materially reduce the volume of
contaminated gas streams containing conden-
sable vapors. To a degree condensers are
also scrubbers, contact units being generally
more effective as scrubbers than surface con-
densers are. Probably their most common
application is as an auxiliary to afterburners,
adsorbers, baghouses, and other control
devices. A number of possible combinations
are shown in Figures 136, 137, and 138. De-
signs depend on the particular air contami-
nants and condensable vapors and on their
concentrations in the contaminated stream.
The system shown in Figure 136 is designed
to control odorous gases contained in a high-
moisture gas stream, as from a rendering
cooker or blood cooker. The stream might
contain from 60 to 99 percent steam at tem-
peratures near 212°F. At the condenser,
vapors are liquefied at the boiling point. If
a strong vacuum is maintained, condensing
temperatures may be well below 212°F. Sub-
cooling may also occur. Uncondensed gases
are separated at the condenser and directed
to an afterburner through a vacuum pump.
A volume reduction of 95 percent and great-
er can be effected through use of either a
contact or surface condenser. Some air
contaminants may condense and others may
be dissolved in the condensate, A contact
condenser, because of greater condensate
10 HTHOSPHERE
FUEL
CONDENSUTE
TO SEWER
Figure 136. A condenser-afterburner air
pollution control system in which a vacu-
um pump is used to remove uncondensed
gases from condensate.
TO UHOSPHERE
Figure 137. A contact condenser-afterburner
air pollution control system in which mal-
odorous, uncondensed gases are separated
from condensate in a closed hot well.
dilution, generally removes more air contami-
nants than a surface condenser does.
The system shown in Figure 136 can be used
with a contact or surface condenser. In
either case a 32-foot barometric leg is re-
quired to pull a strong vacuum. Other vac-
uum devices, such as steam or water ejec-
-------
Vapor Condensers
205
.HATER OUT
KARM ORGANIC
LIQUID STREAM
COMPENSATE
RETURN
Figure 138. A surface condenser used to
prevent surge losses from an accumulator
tank handling warm, volatile, organic
liquid.
tors, might be used in lieu of a vacuum pump.
With steam ejectors, intercondensers and
aftercondensers are often required. The lat-
ter auxiliary condensers might require closed
hot 'wells to separate uncondensed gases from
condensate.
A variation is shown in Figure 137. Here a
contact condenser is used to control high-
moisture, odorous gases. Both condensate
and entrained gases drain to a closed hot well
"where malodorous gases separate by gravity.
Liquids are removed through a trap while
gases are vented to an afterburner or other
suitable control device. The system of
Figure 137 can be used "with surface conden-
sers but is more adaptable to contact units
where adequate subcooling can be readily
achieved.
The surface condenser arrangement shown
in Figure 138 is used to prevent the emis-
sion of condensable organics from blending
tanks, accumulator tanks, drying cleaning
equipment, and so forth. This arrangement
is adaptable to streams rich in condensable
vapors. The condenser is mounted in the
tank vent. Condensate is allowed to drain
to storage or to the original source. No
secondary controls are shown; however,
if further control is required, the saturated
gas stream from the condenser can be vented
to a carbon adsorber, afterburner, or flare
for final cleaning. The product recovered
often offsets the cost of the condenser.
Subcooling Condensate
When condensers are used as air pollution con-
trol devices, care should be taken to ensure
that there is no major evolution of volatiles
from the discharged condensate. Uncondensable
air contaminants should be either safely dissolved
in condensate or vented to further control equip-
ment. In most instances the condensate is merely
cooled to a temperature at which the vapor pres-
sure of contained air contaminants is satisfactorily
low. The required temperature varies with the
condensate. Most condensed aqueous solutions
should be cooled to 140°F or less before they
come into contact with the atmosphere. For vola-
tile organics, lower temperatures are required.
In general, subcooling requirements are more
stringent for surface units than for contact con-
densers where dilution is much greater. Never-
theless, many surface condenser designs do not
permit adequate condensate cooling. In the
typical water-cooled, horizontal, tube-and-shell
condenser of Figure 135a, the shell side tem-
perature is the same throughout the vessel.
Vapors condense, and condensate is removed
at the condensation temperature, which is gov-
erned by pressure. In a horizontal-tube unit
of this type, condensate temperature can be
lowered by: (1) Reducing the pressure on the
shell side, (2) adding a separate subcooler,
or (3) using the lower tubes for subcooling as
shown in Figure 139. Reducing the pressure
alters operating variables in the basic equip-
ment and is not feasible in many instances.
The arrangement of Figure 139 is adaptable to
most processes though it reduces the heat
VAPOR | i COOLANT
COOLANT
IN
Figure 139. Maintaining a condensate level
above the lower tubes to provide subcooling
in a horizontal tube-and-shell condenser.
-------
206
CONTROL EQUIPMENT FOR GASES AND VAPORS
transfer area available for condensation. Here
a level of condensate is maintained in the con-
denser shell. Condensate is chilled before
being discharged through the trap.
The latter arrangement can be used with
vertical-tube units, though it may not be nec-
essary. Vertical-tube condensers provide
some degree of subcooling even with conden-
sation on the shell side.
With condensation inside the tubes, subcooling
occurs in much the same manner whether tubes
are arranged vertically or horizontally. With
inside-the-tube condensation, both condensate
and uncondensed vapors pass through the full
tube length. A separate hot well is usually
provided to separate gases before condensate
is discharged.
CONTACT CONDENSERS
Sizing Contact Condensers
Water requirements for contact condensers
can be calculated directly from the conden-
sation rate, by assuming equilibrium con-
ditions. The cooling water (or other medium)
must absorb enough heat to balance the heat
of vaporization and condensate subcooling. Pip-
ing and hot wells must be sized on the maximum
condenser requirement. The following example
illustrates the method of calculating the quantity
of cooling water for a specific service.
Example 24
Given:
Malodorous exhaust vapors from a dry render-
ing cooker contain 95 percent steam at 200 "F
at 11.5 psia. The maximum evaporation rate
in the cooker is 2, 000 Ib per hour. Steam is
to be condensed at 200°F and cooled to 140°F
in a contact condenser. A vacuum pump re-
moves uncondensable vapors at the condenser
and maintains a slight vacuum on the cooker.
Problem:
Calculate the volume of 60°F fresh water re-
quired and the resultant condensate volume.
Solution:
Condensation: 2,000 x 977.9 Btu/hr = 1,960,000
Btu/hr
Subcooling: 2,000 (200-140) Btu/hr = 120,000
Btu/hr
Cooling load
2,080,000 Btu/hr
2, 080, 000 Btu/hr
Water requirement = (14Q.60) Btu/lb
= 26,000 Ib/hr
= 51.4 gpm
2, OOP Ib/hr
Total condensate = 51.4 +
60 x 8.33 Ib/gal
= 55. 4 gpm
SURFACE CONDENSERS
Characteristics of Condensation
Condensation occurs through two distinct physical
mechanisms, dropwise and filmwise condensation.
When a saturated pure vapor comes in contact with
a sufficiently cold horizontal surface, the vapor
condenses and forms liquid droplets on the surface.
These droplets fall from the surface, leaving bare
metal exposed on which successive condensate
drops may form. This is dropwise condensation.
Normally, a film occurs and coats the conden-
sing surface. Additional vapors must then con-
dense on this film rather than on the bare metal
surface. This is called filmwise condensation
and occurs in most condensation processes.
Heat transfer coefficients are one-fourth to one-
eighth the transfer units associated with dropwise
condensation (Kern, 1950).
Steam is the only pure vapor known to condense
in a dropwise manner. Dropwise condensation
has been found to take place at various times
when a mixture of vapors and gases is present.
Some degree of dropwise condensation may pos-
sibly be attained by using certain promoters.
Promoters such as oleic acid on nickel or
chrome plate, and benzyl mercaptan on copper
or brass become absorbed on the surface as a
very thin layer to prevent the metal surface
from being wetted by any condensate. Steel and
aluminum surfaces are difficult to treat to ac-
quire dropwise condensation. Use of these pro-
moters increases the heat transfer coefficient to
6 to 10 times the amount of filmwise coefficients
(Perry, 1950).
Design of Surface Condensers
Nearly all condenser design calculations are
based on heat transfer that is affected by an
overall transfer coefficient, temperatures,
and surface area. A mathematical solution to
the problem is usually achieved by the expres-
Q = UAT
(85)
-------
Vapor Condensers
207
where
Q =
U =
A =
equations are based only upon vapor entering the
condenser and are as follows:
heat transferred, Btu/hr
overall coefficient, Btu/hr per ft
per "F
heat transfer, ft
mean temperature difference, °F.
Condenser design is often more difficult than in-
dicated by the foregoing expression, and a sim-
plified or general overall heat transfer coeffi-
cient is not used. This is especially true when
a vapor is condensed in presence of a noncon-
densable gas (Donahue, 1956). Nusselt relations
were developed for streamlined flow of all vapor
entering vertical- or horizontal-tube exchangers.
These equations* account for the variation of the
film thickness (thinnest at top of the tube and tube
bundle of vertical and horizontal exchangers) by
expressing the vapor side mean heat-transfer
coefficient in terms of condensate loading. The
*3ymbol notations for these equations are defined at the end of
this chapter on page 232.
Kind of surface
Vertical-tube bundle
Horizontal-tube bundle 0.
Mean heat transfer
coefficient, h
2
1/3
1/3
In instances of streamlined flow of condensate,
the heat-transfer coefficient has been established
as inversely proportional to film thickness. Ob-
servations have, however, shown a decrease to
a certain point, and then a reverse effect "when
the coefficient increased. This reversal oc-
curred at a Reynolds number of approximately
1, 600, indicating that turbulence in liquid film
increases the heat transfer coefficient. Figure
140 shows the relationship between the coeffi-
cient and Reynolds number.
A temperature profile of vapor condensing in
the presence of a noneondensable gas on a tube
wall, as shown in Figure 141, indicates the
resistance to heat flow. Heat is transferred
in two ways from the vapor to the interface.
The sensible heat is removed in cooling the
vapor from t to t at the convection gas cool-
ing rate. The latent heat is removed only
0.1
0.05
100
1,000
10,000
100,000
REYNOLDS NUMBER =
Figure 140. Heat-transfer coefficient of condensation (Donahue, 1956).
-------
208
CONTROL EQUIPMENT FOR GASES AND VAPORS
For mass transfer:
Figure 141. Temperature profile showing
effect of vapor condensation on a tube
wall in presence of a noncondensable gas.
after the condensable vapor has been able to dif-
fuse through the noncondensable part to reach
the tube -wall. This means the latent heat trans-
fer is governed by mass transfer laws.
By using a heat balance around the interface,
the following equation is obtained:
h(t - t ) + KM A (p - p ) = U (t - t )
v c v v c c c w
(86)
When condensation of a pure vapor occurs, tc =
tv. When a condensable gas is present, however,
tc is lower than tv. In solving this equation, a
value of tc is selected by trial and error to satis-
fy the equilibrium condition. The calculation is
repeated for different points in the condenser.
The surface area, necessary is found by using
U and a mean temperature based on tc and tw
over the entire condensing range.
BM
JD
2/3
(89)
Flow inside tubes:
J = Jf - Jh JD
Flow across tube banks:
J = Jf - Jh - JD
0. 027
hc'l
i7~J
0. 2
(90)
0. 33
DG
0. 4
(91)
For solving equation 86, the following procedure
is recommended:
1. Using Raoult's law of partial pressures, cal-
culate the amount of vapor condensing at in-
let and outlet temperatures, and at least
three intermediate temperatures
2. Obtain the following physical properties at
the average of inlet and outlet temperatures
and pressures: (J., p, Dv, Mm, Mv, X, c, k,
(cu/k)2/3, and (u/p D )2/3
3. Choose trial unit
4. Calculate GC, Gj-,, and Ge
5. Calculate Apc + Ap-^ = Aps
6. Calculate h from equation:
Simultaneous heat and mass transfer must be
used to evaluate the equilibrium equation. The
following basic relations state the analogy
between friction, heat transfer , and mass transfer :
hD
= 0. 22
DG
0. 6
w
/3
For friction:
7. Calculate U
jf = l/2f
(87)
8. Calculate j, (for segmentally baffled shell),
For heat transfer:
Jh cG
,2/3
(88)
0.22
-------
Vapor Condensers
209
9. Calculate K from equation:
•^G
K =
(P -
MmPBML
- (P -
BM
- P
In
(P - Pc)
- Py)
In
(P-Pv)
10. Corresponding to the inlet tv, select by T
and E, tc to balance equation (87)
11. Find tc in same manner for other points
12. Calculate the heat removed between each
two successive temperature points, in-
cluding condensate cooling
13. Between each two successive temperature
points, calculate At based on the tempera-
ture difference between tc and tw
14. Using Uc, find the heat transfer surface re-
quired between two successive temperature
points, using At from step 13.
The preceding discussion pertains to the design
of a condenser for condensation of vapor in
presence of a noncondensable gas. ' The design
of the many types of condensers is a vast field
and much too lengthy to cover in this text.
Many technical reference books and articles
have been published containing condenser de-
sign and cost data (Chilton, 1949; Diehl, 1957;
How, 1956; Friedman, 1959; Kern, 1950; Nelson,
195S; Perry, 1950; Smith, 1958; and Thomas,
1959).
Some pertinent facts compiled from these and
other references that will assist in handling
condenser problems include (Kern, 1950):
1. Any saturated vapor can be condensed by
a direct spray of cold water under correct
temperature and pressure. If sufficient
contact is provided, coolant and vapor will
reach an equilibrium temperature. The
condensate created by the water should
not be objectionable in its liquid form.
2. Pure vapor or substantially pure vapor
can be considered condensed isothermally,
and during the condensate range the latent
heat of condensation is uniform.
4. In condensation of streams consisting
primarily of steam, the condenser size
ranges from 10, 000 to 60, 000 square feet
per shell (bundle), the tubes averaging 26
feet long.
5. In water-cooled tube-and-shell condensers
with shell side condensation, overall heat
transfer coefficients for*essentially pure
steam range from 200 to 800 Btu per hour
per square foot per °F.
6. With tube side condensation, coefficients
are generally lower than for comparable
shell side condensers. This phenomenon
is attributed to: (1) Lower coolant ve-
locities outside the tubes than are possible
with tube side cooling, and (2) increased
film thicknesses, namely, film resistances
inside the tubes.
7. Noncondensable gases at condenser tem-
perature blanket the condenser surface
and reduce the condenser capacity.
8. Condensation reduces the volume of the
vapor present and can be assumed to occur
at a constant pressure drop.
9. A balanced pressure drop may be assumed
in the horizontal condenser where partial
condensation is occurring.
10. Within low-pressure operating ranges, the
slight pressure loss due to friction in
vapor pipes may mean an appreciable loss
of total available temperature difference
(Perry, 1950).
11. Low-density steam under vacuum condi-
tions can cause a linear velocity to be
higher than is allowable with steam lines
(Perry, 1950).
12. Vapors should travel across the bundle
as fast as possible (Kern, 1950).
13. Air or inerts can cause up to 50 percent
reduction in condensation coefficients
(Kern, 1950; Perry, 1950).
14. Sources of air or inerts include: Dissolved
gas in the cooling water in cases of jet con-
densers, entrainment with steam, entrain-
ment with vapor, leaks, and noncondensable
gases (Perry, 1950).
3. If the temperature range of a mixture does
not exceed 10° to 20 °F, condensation of
this mixture may be treated as a pure com-
ponent.
15. In vertical-tube condensers, 60 percent
of the condensation occurs in the upper
half (Kern, 1950).
-------
310
CONTROL EQUIPMENT FOR GASES AND VAPORS
16. Horizontal position of a condenser dis-
tributes the vapor better and permits
easier removal of the condensate (Kern,
1950).
17. In the horizontal condenser, it is neces-
sary to prevent cooled condensate from
forming liquid pools and impeding the
flow of vapors (Kern, 1950).
18. Selection of which material should pass
through tubes cannot be decided by
fixed rules, because of factors at a vari-
ance with one another. When corrosive
condensate is encountered, condensation
within the tubes rather than the shell is
usually desirable (Nelson, 1958).
APPLICATIONS
Condensers have been used successfully
(either separately or with additional control
equipment) on the following processes or
equipment:
REFINERY AND PETROCHEMICAL
Alkylation unit accumulator vents
Amine stripper units
Butadiene accumulator vents
Coker blowdown
Ketone accumulator vents
Lube oil rerefining
Polyethylene gas preparation accumulator vents
Residium stripper unit accumulator vents
Storage equipment
Styrene-processing units
Toluene recovery accumulator vents
Udex extraction unit
CHEMICAL
Manufacture and storage of ammonia
Manufacture of Cooper naphthenates
Chlorine solution preparation
Manufacture of ethylene dibromide
Manufacture of detergent
Manufacture of insecticide
Manufacture of latex
Manufacture of nitric acid
Manufacture of phthalic anhydride
Resin reactors
Soil conditioner formulators
Solvent recovery
Thinning tanks
MISCE LL ANEOUS
Aluminum fluxing
Asphalt manufacturing
Blood meal driers
Coal tar-dipping operations
Degreasers
Dry cleaning units
Esterfication processes
Pectar preparation
Rendering cookers (animal waste)
Vitamin formulation
GAS ABSORPTION EQUIPMENT
Gas absorption is the mechanism whereby one
or more constituents are removed from a gas
stream by dissolving them in a selective liquid
solvent. This is one of the major chemical
engineering unit operations and is treated ex-
tensively in all basic chemical engineering text-
books. These texts deal with gas absorption as
a method of recovering valuable products from
gas streams, for example, in petroleum produc-
tion, natural gasoline is removed from wellhead
gas steams by absorption in a special hydro-
carbon oil. Absorption is also practiced in in-
dustrial chemical manufacturing as an important
operation in the production of a chemical com-
pound. For example, in the manufacture of
hydrochloric acid, one step in the process in-
volves the absorption of hydrogen chloride gas
in water.
From an air pollution standpoint, absorption is
useful as a method of reducing or eliminating
the discharge of air contaminants to the atmo-
sphere. Even in this application, absorption
can yield profits to the user. For example,
it can be employed to remove hydrogen sulfide
from process gas streams in a petroleum re-
finery to meet air pollution regulations. With
further processing, this hydrogen sulfide can
be converted to elemental sulfur, a valuable
product.
The gaseous air contaminants most commonly
controlled by absorption include sulfur dioxide,
hydrogen sulfide, hydrogen chloride, chlorine,
ammonia, oxides of nitrogen, and light hydro-
carbons.
-------
Gas Absorption Equipment
211
In other examples, such as solvent recovery,
desorption or stripping may be practiced after
absorption not only to recover a valuable ab-
sorbed constituent but also to recover valuable
solvent for reuse. Sometimes, after absorp-
tion, solute and solvent are not separated but
are used as a product or intermediate com-
pound in chemical manufacture.
Treybal (1955) lists some important aspects
that should be considered in selecting absorp-
tion solvents.
1. The gas solubility should be relatively high
so as to enhance the rate of absorption and
decrease the quantity of solvent required.
Solvents similar chemically to the solute
generally provide good solubility.
2. The solvents should have relatively low
volatilities so as to reduce solvent losses.
3. If possible, the solvents should be non-
corrosive so as to reduce construction
costs of the equipment.
4. The solvents should be inexpensive and
readily available.
5. The solvents should have relatively low
viscosities so as to increase absorption
and reduce flooding.
6. If possible, the solvents should be nontoxic,
nonflammable, chemically stable, and have
low freezing points.
GENERAL TYPES OF ABSORBERS
PACKED TOWER DESIGN
A packed tower is a tower that is filled "with
one of the many available packing materials,
as shown in Figure 142. The packing is de-
signed so as to expose a large surface area.
When this packing surface is wetted by the
solvent, it presents a large area of liquid
film for contacting the solute gas.
I GAS OUT
L I QU I D-
IN
LIQUID DISTRIBUTOR
LIQUID
RE-DISTRIBUTOR
PACKING SUPPORT
GAS IN
LIQUID OUT
Figure 142. Schematic diagram
of a packed tower (Treybal 1955
p. 134).
Gas absorption equipment is designed to provide
thorough contact between the gas and liquid sol-
vent in order to permit interphase diffusion of
the materials. The rate of mass transfer be-
tween the two phases is largely dependent upon
the surface exposed. Other factors governing
the absorption rate, such as solubility of the
gas in the particular solvent and degree of
chemical reaction, are characteristic of the
constituents involved and are more or less in-
dependent of the equipment used. This contact
between gas and liquid can be accomplished by
dispersing gas in liquid or vice versa.
Absorbers that disperse liquid include packed
towers, spray towers or spray chambers, and
venturi absorbers. Equipment that uses gas
dispersion includes tray towers and vessels
with sparging equipment.
Usually the flow through a packed column is
countercurrent, with the liquid introduced at
the top to trickle down through the packing
while gas is introduced at the bottom to pass
upward through the packing. This results
in highest possible efficiency, since, as the
solute concentration in the gas stream de-
creases as it rises through the tower, there
is constantly fresher solvent available for con-
tact. This gives maximum average driving
force for the diffusion process throughout the
entire column.
In concurrent flow, where the gas stream and
solvent enter at the top of the column, there
is initially a very high rate of absorption that
constantly decreases until, with an infinitely
tall tower, the gas and liquid would leave in
-------
212
CONTROL EQUIPMENT FOR GASES AND VAPORS
equilibrium. Concurrent flow is not often
used except in the case of a very tall column
built in two sections, both located on the
ground, the second section using concurrent
flow merely as an economy measure to ob-
viate the need for constructing the large gas
pipe from the top of the first section to the
bottom of the second. Moreover, for an
operation requiring an exceptionally high
solvent flow rate, concurrent flow might be
used to prevent flooding that could occur in
countercurrent flow.
Pocking Materials
The packing should provide a large surface
area and, for good fluid flow characteristics,
should be shaped to give large void space
when packed. It should likewise be strong
enough to handle and install without exces-
sive breakage, be chemically inert, and be
inexpensive.
Rock and gravel have been used but have
disadvantages of being too heavy, having
small surface areas, giving poor fluid flow
and, at times, not being chemically inert.
Coke lumps are also used sometimes and
here the weight disadvantage is not present.
Owing to its porosity, coke has a large
surface area per unit volume. The exposed
surface is not, however, as large as might
be expected since the pores are so small
that they become filled or filmed over by
the solvent, which considerably reduces
the effective surface.
Generally, packing in practice consists of
various manufactured shapes. Raschig
rings are the most common type, consisting
of hollow cylinders having an external di-
ameter equal to the length. Other shapes
include Berl saddles, Intalox saddles,
Lessing rings, cross-partition rings,
spiral-type rings, and drip-point grid
tiles. Figure 143 shows several common
shapes. Physical characteristics of these
various types of packings have been de-
termined experimentally and compiled in
tables by Leva (1953).
Packing may be dumped into the column ran-
domly, or regularly shaped packing may be
manually stacked in an orderly fashion. Ran-
domly dumped packing has a higher specific
surface contact area and a higher gas pressure
drop across the bed. The stacked packings have
an advantage of lower pressure drop and higher
possible liquid throughout, but the installation
cost is obviously higher. Table 57 and Figure
144 list typical packing costs and packed-tower
installed prices for 1959.
BERL SADDLE
RASCHIG RING
INTALOX SADDLE
PALL RING
TELLERETTE
Figure 143. Common tower packing materials
(Teller, I960, p. 122).
Table 57. COSTS OF REPRESENTATIVE
TOWER PACKINGS (Teller, I960)
Packing
Raschig rings, ceramic
Raschig rings, carbon
Berl saddles, ceramic
Intalox saddles, ceramic
Intalox saddles, carbon
Tellerettes, polyethylene
Low density
High density
Pall rings, ceramic (3 ASF)
Pall rings, polypropylene
Pall rings, stainless steel
Cost of packing, $/ft3 (1959basis)
1/2-in.
11.70
16. 90
24. 80
23. 55
-
-
-
-
41.00
186. 50
1-in.
6. 50
9.60
9.90
9.40
18.60
16.00
23.00
5.00
26.00
96.00
1-1/2-in.
5.05
8.00
7.50
7. 15
18.40
-
-
-
20.75
83.00
2-in.
4.85
6.60
7.70
7. 30
-
-
-
-
18.50
69.00
Liquid Distribution
Since the effectiveness of a packed tower de-
pends on the availability of a large, exposed,
liquid film, then obviously, if poor liquid dis-
tribution prevents a portion of the packing from
being irrigated, that portion of the tower is
ineffective. Poor distribution can be due to
improper introduction of the liquid at the top of
the tower and to channeling within the tower.
-------
Gas Absorption Equipment
213
10 000
20 30 40 50
DIMETER, inches
Figure 144. Packed-tower costs, 1959, with Raschig rings as
packing (Tel ler, 1960, p. 123).
At least five points of introduction of liquid per
square foot of tower cross-section must general-
ly be provided to ensure complete wetting. The
liquid rate must be sufficient to wet the packing
but not to flood the tower. Treybal (1955) states
that a superficial liquid velocity of at least 800
pounds of liquid per hour per square foot of
tower cross-section is desirable.
Solid-cone spray nozzles make excellent dis-
tributors but may plug if solid particles are
present in the solvent. In randomly packed
towers, the liquid tends to channel toward the
walls, because of the usually lower packing
density near the walls. In tall towers this
channeling is controlled by liquid redistribu-
tors at intervals of 10 to 15 feet. Moreover,
this effect is minimized if the packing pieces
are less than one-eighth the diameter of the
tower.
Tower Capacity
The terms used to indicate capacity of a
packed column or tower are load point and
flood point. For a given packing and liquid
rate, if gas pressure drop is plotted against
gas velocity on a logarithmic scale, there
are two distinct breakpoints "where the slope
of the curve increases. At low gas veloci-
ties the curve is almost parallel to that ob-
tained with dry packing, but above the break-
points, the pressure drop increases more
rapidly with increased gas velocity. The low-
er of these two breaks is the load point and the
higher one the flood point.
As gas velocity increases above the load
point, the liquid holdup in the bed increases
until, at the second breakpoint, the flood
point, most of the void space in the tower
is filled with liquid and there is liquid
entrainment in the gas stream. Of course,
at this point there is an excessive pressure
drop. Columns should seldom be operated
above the load point, but since the load point
is sometimes more difficult to establish than
the flood point, it is common practice to de-
sign for 40 to 70 percent of the flood point.
In general, flooding velocities are considerably
higher for stacked packing than for dumped
packing. The plot of Lobo (Figure 145) can
-------
214
CONTROL EQUIPMENT FOR GASES AND VAPORS
0 1
0 01
0 001
0
(L1 V)(oG
I 0
10 0
Figure 145. Correlation for flooding rate in randomly packed
towers (Lobo, 1945, p. 693).
be used to determine flow rates that will cause
flooding. This curve is based on measurements
with several liquids and gases on a variety of
packings.
For many years packed towers were designed
on the same basis as plate or tray towers.
The number of theoretical plates or trays re-
quired for a given degree of separation was
calculated and this quantity multiplied by a
figure called height equivalent to a theoretical
plate (HETP). This HETP was an experi-
mentally determined figure varying widely with
packing, flow rates of each fluid used, and
concentration of solute for any specific system.
Experimental evaluation of these variables
made use of this system too cumbersome and
it is now rarely used. Design procedures now
employ the concept of the transfer unit. The
major design items to be calculated are the
column diameter, number of transfer units,
the height of a transfer unit, and the system
pressure drop. These will be discussed in-
dividually.
Tower Diameter
As mentioned previously, gas velocity is limited
by flooding conditions in the tower. By use of
the design gas volume, design solvent flow rate,
and type of packing, the tower diameter can be
computed by using Lobe's correlation in Figure
145. Packing factors are obtained from Figure
146. The procedure is as follows:
1. Calculate the factor —
where
L' = liquid flow rate, Ib/hr
V = gas flow rate, Ib/hr
Pr = gas density, Ib/ft
p = liquid density, Ib/ft"
L
-------
Gas Absorption Equipment
215
10 000
000
100
PACKING FACTOR FOR 1 in INTALOX
SADDLES a «3 90
05 10 15 20 25 30 35
NQtMNM. P&CMNG SUE inches
Figure 146. Packing factors for Raschig rings and saddles
(Lobo, 1945, p. 693).
2. Using the calculated value in (1), obtain
from Figure 145 the value of
G'- Pr
gc PG
where
G' = gas flow rate, Ib/sec-ft of tower
cross-section
—-r = packing factor from Figure 146.
p = liquid density, Ib/ft
l_i
fi' = liquid viscosity, centipoises
g = gravitational constant, 32. 2 ft/sec .
3. Solve for G1, the superficial mass gas ve-
locity at flood point from the factor deter-
mined in (2).
4. Calculate S, the tower cross-section area
in ft^ for fraction of flooding velocity selec-
ted, f, by the equation
p = gas density, Ib/ft
G
S =
(G')(f){3,600)
(92)
-------
216
CONTROL EQUIPMENT FOR GASES AND VAPORS
5.
Calculate the tower's inside diameter, DC,
by the equation
DC =
0. 5
(93)
Tower diameter should be calculated for
conditions at both top and bottom of the
tower. The tower is designed to the larger
diameter.
Number of Transfer Units (NTU)
A transfer unit is a measure of the difficulty of
the mass transfer operation and is a function of
the solubility and concentrations of the solute
gas in the gas and liquid streams. It is ex-
pressed as NQQ or NQL, depending upon whether
the gas film or liquid film resistance controls
the absorption rate. The gas film resistance
usually controls when solubility of solute in sol-
vent is high and conversely, the liquid film con-
trols when the solubility is low.
In air pollution control work where, in general,
a relatively small concentration of solute is to
be removed from an airstream, a solvent in
which the solute gas is highly soluble is usually
selected in order to obtain the highest possible
economic separation. Thus, for the majority
of cases encountered, the gas film resistance
will be controlling.
One of the most widely used methods of determin-
ing the number of transfer units is that proposed
by Baker (1935), which is based upon an operating
diagram consisting of an equilibrium curve and
an operating line. For a given gas-liquid system,
if the temperature is constant and the gas partial
pressure is varied, the gas concentration in the
liquid changes to an equilibrium concentration
at each partial pressure. If the system consists
of a soluble gas to be removed, an insoluble
carrier gas, and a solvent, then, as the amount
of soluble gas in the system increases, the
equilibrium concentration of the soluble gas in
the liquid .increases but not proportionally.
These equilibrium conditions can exist for an
infinite number of concentration states and, when
plotted on X-Y coordinates, become the equilib-
rium curve. The operating line represents the
concentrations of solute in the gas stream and
in the liquid phase at various points in the tower.
When plotted as moles solute per mole solvent
versus moles solute per mole gas on X-Y co-
ordinates, the result is a straight line. Thus,
when the composition of the inlet gas and the
desired or required degree of absorption are
krown, the points on the operating line for each
end of the column can be calculated. The oper-
ating line is the straight line connecting the
two points. For absorption to occur, the oper-
ating line must lie above the equilibrium curve
on the diagram. The relative position of the
operating line and equilibrium curve indicates
how far the tower conditions are from equilib-
rium. The more widely separated the lines,
the further the tower conditions are from equi-
librium and the greater is the driving force
for the absorption operation.
Figure 147 illustrates the graphical method
of determining the number of transfer units
for a countercurrent packed tower with the
gas film controlling the absorption rate. The
equilibrium curve (line AB) for the particular
gas-liquid system is plotted from experimental
data, which, for most common systems, has
been determined. Much of these data can be
located in the International Critical Tables
and in Perry (1950). The operating line is
a straight line drawn between points D and C.
D is the point representing the concentra-
tions of solute in the gas stream and in the
liquid stream at the gas inlet and liquid out-
let (bottom of the tower). Point C corresponds
to these concentrations at the top of the column.
Line EF is di'awn so that all points on the line
are located midway on a vertical line between
the operating line and equilibrium curve.
Starting at point C on the operating line (con-
ditions at the top of the column), draw a hori-
zontal line CH so that CG = GH. Then draw a
vertical line HJ back to the operating line.
X = SOLUTE moles/SOLVENT mole
Figure 147. Graphical determination of
the number of transfer units.
-------
Gas Absorption Equipment
217
The step CHJ represents one gas transfer unit.
This stepwise procedure is continued to the
end of the operating line (conditions at the
bottom of the column). Two gas transfer units
(NOG) are shown in Figure 147.
If the liquid film resistance is the controlling
factor in the transfer of solute to solvent, draw
the line EF so that all points on the line are
located midway on the horizontal axis between
the operating line and equilibrium curve. Then,
starting at point D on the operating line, draw
a vertical line DK so that DL = LK. The step
is completed by drawing a line KJ back to the
operating line. This procedure is then con-
tinued to point C on the operating line. Figure
147 does not accurately indicate the number of
liquid transfer units since the line EF was
drawn for the case where the gas film resis-
tance controls.
Height of a Transfer Unit
Generalized correlations are available for
computing the height of a transfer unit and
are expressed as HQ and HL for heights of gas
and liquid transfer units respectively. These
use experimentally derived factors based on the
type of packing and the gas and liquid flow rates
as shown in equations 94 and 95.
H,
0.5
where
H = height of a gas transfer unit, ft
G
G = superficial gas rate, Ib/hr-ft
L = superficial liquid rate, Ib/hr-ft
a - a packing constant from Table 58
ft - a packing constant from Table 58
7 = a packing constant from Table 58
(j, = gas viscosity, Ib/hr-ft
G
p = gas density, Ib/ft
G
D = gas diffusivity, ft /hr.
G
(94)
The group I ;r—~ I is known as the Schmidt
\ G G /
V /
number as shown in Table 59*
Table 58. CONSTANTS FOR USE IN DETERMINING GAS FILM'S
HEIGHT OF TRANSFER UNITS (Treybal, 1955, p. 239)
Packing
Raschig rings
3/8 in.
1 in.
1-1/2 in.
2 in.
Berl saddles
1/2 in.
1 in.
1-1/2 in.
3 -in. partition rings
Spiral rings (stacked
staggered)
3-in. single spiral
3 -in. triple spiral
Drip-point grids
No. 6146
No. 6295
a
2. 32
7. 00
6. 41
17. 30
2. 58
3. 82
32. 40
0. 81
1. 97
5. 05'
650
2. 38
15.60
3.91
4. 56
ft
0. 45
0. 39
0. 32
0. 38
0. 38
0. 41
0. 30
0. 30
0. 36
0. 32
0. 53
0. 35
0. 38
0. 37
0. 17
7
0. 47
0. 58
0. 51
0. 66
0. 40
0. 45
0. 74
0. 24
0.40
0. 45
1 . 06
0. 29
0.60
0. 39
0. 27
Range of
G'
200 to 500
200 to 800
200 to 600
200 to 700
200 to 700
200 to 800
200 to 700
200 to 700
200 to 800
200 to 1, 000
150 to 900
130 to 700
200 to 1, 000
130 to 1, 000
100 to 1, 000
L
500 to 1, 500
400 to 500
500 to 4, 500
500 to 1, 500
1, 500 to 4, 500
500 to 4, 500
500 to 1, 500
1, 500 to 4, 500
400 to 4, 500
400 to 4, 500
3, 000 to 10, 000
3, 000 to 10, 000
500 to 3, 000
3, 000 to 6, 500
2, 000 to 11, 500
-------
218
CONTROL EQUIPMENT FOR GASES AND VAPORS
Table 59. DIFFUSION COEFFICIENTS OF
GASES AND VAPORS IN AIR AT 25 °C AND
1 ATM (Perry, 1950)
Substance
Ammonia
Carbon dioxide
Hydrogen
Oxygen
Water
Carbon disulfide
Ethyl ether
Methanol
Ethyl alcohol
Propyl alcohol
Butyl alcohol
Amyl alcohol
Hyxyl alcohol
Formic acid
Acetic acid
Propionic acid
i-Butyric acid
Valeric acid
i-Caproic acid
Diethyl amine
Butyl amine
Aniline
Chloro benzene
Chloro toluene
Propyl bromide
Propyl iodide
Benzene
Toluene
Ethyl benzene
Propyl benzene
Diphenyl
n -Octane
Mesitylene
D, cm /sec
0.236
0. 164
0. 410
0.206
0.256
0. 107
0.093
0. 159
0. 119
0. 100
0. 090
0. 070
0. 059
0. 159
0. 133
0.099
0. 081
0.067
0.060
0. 105
0. 101
0.072
0. 073
0. 065
0. 105
0. 096
0. 088
0. 084
0. 077
0. 059
0.068
0. 060
0. 067
_t_
pD
0.66
0.94
0.22
0.75
0.60
1.45
1.66
0.97
1.30
1. 55
1.72
2.21
2.60
0.97
1. 16
1.56
1.91
2. 31
2.58
1.47
1.53
2. 14
2. 12
2. 38
1.47
1.61
1.76
1.84
2.01
2.62
2.28
2.58
2.31
Table 60. CONSTANTS FOR USE IN
DETERMINING LIQUID FILM'S HEIGHT OF
TRANSFER UNITS (Treybal, 1955, p. 237)
Packing
Raschig rings
3/8 in.
1/2 in.
1 in.
1-1/2 in.
2 in.
Berl saddles
1/2 in.
1 in.
1-1/2 in.
3-in. partition rings
Spiral rings (stacked
staggered!
3-in. single spiral
3-m. triple spiral
Drip-point grids
No. 6146
No. 6295
0. 00182
0. 00357
0. 0100
0. 0111
0. 0125
0. 00666
0. 00588
0. 00625
0. 0625
0. 00909
0. 0116
0. 0154
0. 00725
n
0. 46
0. 35
0. 22
0. 22
0. 22
0. 28
0. 28
0. 28
0. 09
0. 28
0. 28
0. 23
0. 31
Range of L,
400 to 15, 000
400 to 15, 000
400 to 15, 000
400 to 15, 000
400 to 15, 000
400 to 15, 000
400 to 15, 000
400 to 15, 000
3, 000 to 14, 000
400 to 15, 000
3, 000 to 14, 000
3, 500 to 30, 000
2, 500 to 22,000
The group
is the Schmidt number
as shown in Table 61. Each of these empirical
equations neglects the effect of the other film's
resistance. Actually, however, even in the
case of absorbing highly soluble ammonia in
water, experimental results have shown that
the liquid film resistance is significant. The
height of an overall gas transfer unit, H_^,
is determined by the folio-wing equation, which
takes into account the liquid film resistance.
where
0. 5
(95)
H = height of a liquid transfer unit, ft
L = superficial liquid rate, Ib/hr-ft
fa = liquid viscosity, Ib/hr-ft
L
0 = a packing constant, Table 60
T] .= a packing constant, Table 60
p = liquid density, Ib/ft
L
D = liquid diffusivity, ft /hr.
L
H
OG
= H
-------
Gas Absorption Equipment
219
Table 61. DIFFUSION COEFFICIENTS IN
LIQUIDS AT 20°C (Perry, 1950)
mass flow rates, high-viscosity liquids cause
greater gas pressure drop than those of low vis-
cosity do.
9.
Solute
O2
CO2
N2O
CI2
Br2
H2
N?
£>
HC1
H2S
H2S04
HNO3
Acetylene
Acetic acid
M ethanol
Ethanol
Pr opanol
Butanol
Allyl alcohol
Phenol
Glycerol
Pyrogallol
Hydroquinone
Urea
Resorcinol
Ur ethane
Lactose
Maltose
Glucose
Mannitol
Raffinose
Sucrose
Sodium chloride
Sodium hydroxide
C02b
Phenolb
•t
Chloroform0
Phenol0
Chi or of or mc
Acetic acidc
Ethylene dichloridec
n x 105
(cm2/sec) x 105
I. 80
1. 50
1. 51
1.76
1.22
5. 13
1.64
2.64
1. 41
1.73
2. 60
1. 56
0.88
1.28
1. 00
0. 87
0. 77
0. 93
0. 84
0.72
0. 70
0. 77
1. 06
0. 80
0. 92
0.43
0. 43
0. 60
0. 58
0. 37
0. 45
1. 35
1. 51
3.40
0. 80
1. 23
1. 54
2. 11
LJL
pD
558
570
665
570
824
196
613
381
712
580
390
645
1., 140
785
1, 005
1, 150
1, 310
1,080
1, 200
1, 400
1,440
1, 300
946
1,260
1,090
2,340
2, 340
--
1,730
2, 720
2, 230
745
665
445
1, 900
1, 230
479
350
1.92 384
2.45
301
Leva's empirical relation applies below the load
point. This is as follows :
2
/1A~ W ^l^'/P A G'
Z ! PG
where
P = pressure drop, Ib/ft
Z = packed height of tower, ft
m = pressure drop constant from Table 62
n = pressure drop constant from Table 62
L1 = superficial mass liquid velocity, lb/
hr-ft2
G1 = superficial mass gas velocity, Ib/hr-
ft2
p = liquid density, Ib/ft
3
p_ = gas density, Ib/ft .
Illustrative Problem
The following example illustrates the preceding
principles of packed tower design. Knowing the
amount of solute in the gas stream, the total flow
rate of the gas stream, the most suitable solvent,
an acceptable packing, and the desired degree of
absorption, calculate the tower dimensions.
Given:
aSolvent is water except where indicated.
"Solvent is ethanol.
cSolvent is benzene.
Pressure Drop Through Pocking
Treybal (1955) states that pressure drop data of
various investigators varies widely even for the
same packing and flow rates. These discrep-
ancies were probably due to differences in pack-
ing density. Moreover, not enough work has
been done on liquids of high viscosity for proper
evaluation, though it is recognized that, at equal
Design a packed tower to remove 95% of the am-
monia from a gaseous mixture of 10% by volume
of ammonia and 90% by volume of air. The gas
mixture consists of 80 Ib-moles/hr at 68 °F and
1 atm. Water containing no ammonia is to be
used as solvent and the packing will be 1-inch
Raschig rings. The tower will be designed to
operate at 60% of the flood point, and isothermal
conditions at 68 °F will be assumed. The water
will not be recirculated.
Problem:
Determine water flow rate, tower diameter,
packed height, and tower pressure drop.
-------
220
CONTROL EQUIPMENT FOR GASES AND VAPORS
Table 62. PRESSURE DROP CONSTANTS FOR TOWER
PACKING (Treybal, 1955)
Packing
Raschig rings
Berl saddles
Intalox saddles
Drip-point grid
tiles
Nominal
size,
in.
1/2
3/4
1
1-1/2
2
1/2
3/4
1
1-1/2
1
1-1/2
No. 6146
Continuous
flue
Cross flue
No. 6295
Continuous
flue
Cross flue
m
139
32, 90
32. 10
12. 08
11. 13
60. 40
24. 10
16. 01
8. 01
12. 44
5.66
1. 045
1.218
1. 088
1. 435
n
0. 00720
0. 00450
0. 00434
0. 00398
0. 00295
0. 00340
0. 00295
0. 00295
0. 00225
0.00277
0.00225
0. 00214
0.00227
0. 00224
0. 00167
Range of L ,
lb/hr-ft2
300 to 8,600
1,800 to 10, 800
360 to 27, 000
720 to 18, 000
720 to 21, 000
300 to 14, 100
360 to 14, 400
720 to 78, 800
720 to 21, 600
2, 520 to 14, 400
2, 520 to 14, 400
3, 000 to 17, 000
300 to 17, 500
850 to 12, 500
900 to 12, 500
Range
of P/Z,
Ib/ft2-ft
0 to 2, 6
0 to 2. 6
0 to 2. 6
0 to 2. 6
0 to 2. 6
0 to 2. 6
0 to 2. 6
0 to 2.6
0 to 2.6
0 to 2.6
0 to 2. 6
0 to 0. 5
0 to 0. 5
0 to 0.5
0 to 0. 5
Solution:
1. Calculate the water rate: :
a. Equilibrium data for the system ammonia-
water are as follows:
X 0.0206 0.0310 0.0407 0.0502 0.0735 0.0962
Y 0.0158 0.0240 0.0329 0.0418 0.0660 0.0920
Plot the equilibrium curve as shown in Fig-
ure 148:
The curve is straight approximately to the
point P, with a slope of about 0. 75. Above
point P, the slope is variable and higher
than 0.75. Use 0.75 as the slope, m, of
the equilibrium curve.
b. When the temperature rise of the solvent
is negligible, apply the relation
G (m)
m
L
= 0. 70
G
L
L
gas rate = 80 Ib-moles/hr
liquid rate, Ib-moles/hr
(80)(0.75)
0.70
= 85. 8 Ib-moles/hr
2. From the given gas flow rate, the calculated
liquid rate, and the degree of absorption de-
sired (95% of ammonia), tabulate gas and
liquid flow rates at both ends of the tower:
Density,
Ib-moles/hr Ib/hr lb/ft3
Inlet gas (bottom) 80
Outlet gas (top) 72.4
Inlet water (top) . 85.8
Outlet liquor (bottom) 93.4
2,221 0.0720
2,092 0.0750
1,542 62.4
1,671 62.4
3. Calculate the tower diameter:
a. Use conditions at top of tower:
where
m = slope of equilibrium curve = 0. 75
°'5
1,542
°-5
= 0. 02f
-------
Gas Absorption Equipment
221
0 11
co
cc.
0
^
z\
>
y
^
y
/
' y
^
V
^/
Y
s
/
/^
/
/
/
A
/&
^
f
/
/
/
r&
1
/
f
(
/
t
= (0. 424)(0. 60} = 0, 254 Ib/sec-ft"
(2) Top of tower:
72
b. 'Jse conditions at bottom of tower:
0. 5
(!) !TTT-
X = 0 lei.U rirg v/a>er i.s NH
0. 0720 \ ' " b. Plot the operating line from the data ir, (a)
62, 4 / ' on the same graph used for the equilibrium
-------
222
CONTROL EQUIPMENT FOR GASES AND VAPORS
c. By the method of Baker (described previously)
graphically determine the number of transfer
units:
NTU = 6
5. Calculate the height of a transfer unit:
a. Gas transfer unit:
0.5
where
G = 2'2211b/2hr = 896lb/hr-ft2
2.48 ft
L =
I, 542
2.48
= 622 Ib/hr-ft
a - 7. 00 from Table 58
j3 = 0. 39 from Table 58
7 = 0. 58 from Table 58
3.66 from Table 59
H
(7.00H896)0'39 (0.66)0'5
"G
(622)
0. 58
= 1.92 ft
b. Liquid transfer unit:
•where
= 0. 01 from Table 60
0. 22 from Table 60
= 622 Ib/hr-ft2
= 1 centipoise = 2.42 Ib/hr-ft
I = 570 from Table 61
H
(570)°'5 = 0.79 ft
c. Overall gas transfer unit
HOG = HG
where
slope of equilibrium curve = 0. 75
gas rate = 80 Ib-moles/hr
liquid rate = 85. 8 Ib-moles/hr
6. Calculate the packed tower height (Z):
Z = NTU x HOG
Z = 6 x 2. 47 = 14. 8 ft
7. Calculate the tower pressure drop
Ap =
where
Ap = pressure drop, Ib/ft
Z = packed height = 14.8ft
m = 32. 10 from Table 62
n = 0. 00434 from Table 62
L = 622 Ib/hr-ft2
p = 62.4 Ib/ft3
G = 896 Ib/hr-ft2
PG= Avg gas density = 0.036 Ib/ft"
(32.1 x 10-8)(LO)(0-°0434)(622)/62-4(896)2(14.8)
0.0736
= 57.2 Ib/ft
-------
Gas Absorption Equipment
223
Ap =
57. 2 Ib/ft (1 in. WC)
5. 197 Ib/ft
= 11.0 in. WC
liquid. The liquid enters at the top of the
tower, flows across each plate and down-
ward from plate to plate through downspouts.
PLATE OR TRAY TOWERS
In contrast to packed towers, where gas and
solvent are in continuous contact throughout the
packed bed, plate towers employ stepwise con-
tact by means of a number of trays or plates
.that are arranged so that the gas is dispersed
through a layer of liquid on each plate. Each
plate is more or less a separate stage, and
the number of plates required is dependent
upon the difficulty of the mass transfer oper-
ation and the degree of separation desired.
Types of Plates
The bubble cap plate or tray is most common,
and most general references deal primarily
with it when discussing plate towers. Other
types of plates include perforated trays,
Turbogrid trays, and Flexitrays.
A schematic section of a bubble cap tray tower
is shown in Figure 149. Each plate is equipped
with openings (vapor risers) surmounted with
bubble caps. Typical bubble caps are illus-
trated in Figure 150. The gas rises through
the tower and passes through the openings in
the plate and through slots in the periphery
of the bubble caps, which are submerged in
3 CAP SUPPORTS
AT 120°F
HOLD-DOWN BAR
VAPOR RISER
CAST TRAY
-CAST CAP
SHEET METAL CAP
SHEET METAL TRAY
Figure 150. II lustration of
some typical bubble caps.
SHELL
TRAY
DOWSPOUT
TRAY
SUPPORT RING
TRAY
STIFFENER -
VAPOR
RISER
FROTH
^-LIQUID IN
BUBBLE CAP
INTERMEDIATE
FEED
-LIQUID OUT
Figure 149. Schematic diagram
of a bubble-cap tray tower
(Treybal, 1955, p. 111).
The depth of liquid on the plate, and liquid
flow patterns across the plate are controlled
by various weir arrangements, which will be
discussed in greater detail.
In perforated plates or sieve trays, the gas
passes upward through a pattern of holes
drilled or punched in the trays. Three-
sixteenth-inch-diameter holes spaced on a
3/4-inch triangular pitch are commonly used.
A disadvantage of this type is the tendency of
liquid to "weep" or leak down through the
holes instead of through the downspouts at
low gas velocities. Moreover, the trays
must be installed perfectly level, or chan-
neling, with resultant loss of efficiency,
will occur. On the other hand, a perforated
tray costs only 60 to 70 percent as much as
a bubble cap plate designed for the same
throughput. With towers of the same di-
ameter, perforated trays supposedly have
a capacity 10 to 40 percent greater than
that of bubble cap plate towers.
With Turbogrid trays, licensed by'Shell Develop-
ment Company, the vapor passes up through the
spaces between parallel rods or bars, and the
-------
224
CONTROL EQUIPMENT FOR GASES AND VAPORS
liquid level on the tray is maintained by the gas
pressure beneath the tray. There are no down-
spouts, and the liquid flows downward through
the same openings used by the upward flowing
gases. A Turbogrid tray is shown in Figure 151.
These are reputed to have high absorption effi-
ciencies even at high capacities with liquids con-
taining a small amount of suspended solids. For
example, a 50 percent increase in capacity has
been reported where bubble cap plates have been
replaced by Turbogrid trays in an existing tower.
Flexitrays, licensed by the Koch Engineering
Company, have floating caps that allow va.ria-
tions in the vapor openings with varying gas flow.
Different weights can be put on the caps so that
the slots will be only partially open at low gas
flow rates. This tray also has relatively low
resistance to liquid crossflow and supposedly
has advantages over bubble cap trays in large
TOP VIE*
CO' UKN
S^tLL
\
\
ic*
Fi gure 151. !! i usm ti on of a tyin ca I
Turbognd tr^y (Shell ileve I oomen t Co.,
Erneryv. Me, Ca i • i. ,.
columns or operations that require high liquid
rat-es. Flexitrays are claimed to have a capac-
ity 12 to 50 percent higher than that of bubble
cap plates and cost only 60 to 80 percent as
much.
Although the proponents of the various trays
make each sound attractive, it should be re-
membered that the bubble cap plate is still
the standard of the industry and presently
outnumbers all the other types. Thus further
discussion of plate towers will be devoted ex-
clusively to the design of bubble cap plates.
BUBBLE CAP PLATE TOWER DESIGN
Liquid Flow
Common variations in liquid flow across a
bubble cap plate include: (I) Crossflow in
opposite directions on alternate plates, (2)
crossflow in the same direction on all plates,
and (3) split-flow arrangements. There are
also variations in weir and downspout design.
Several liquid flow patterns are diagrammed
in Figure 152, and typical bubble cap tray
arrangements for different liquid flow paths
are shown in Figure 153.
The single-pass plate with a rectangular weir
shown in Figure 153a is the most common.
Much of its cross-sectional area is devoted
1o vapor flow, whereas, a split crossflow
plate, shovvn in Figure 153b, has more of its
cross-sec ional area devoted "-O liquid flow.
The .split-low tray also has greater down-
spout area, and the liquid flows a shorter
distance from the tray inlet to the overflow
"•eir. Thus split-flow !.ra/s handle higher
jiquir tlow rates and are suitable for large-
'liametcr towers.
CaFCdde tiay arrangements, shown in Fig-
ures 1 52d and 153c, ar-i used to keep the
Hquic! level at a more constar! depth over the
';nnro trav area despite rorsiderable liquid
h(-o<"i 'I'.ife i enti'il across tho iray. Tiie^e
EL.T- L'.cpd ior "•'>. f eptio r,;t 1! y la r p, e - a ia >''ie t e r
iG'A'.'is. I^aal-il -lovv. Figure1 15?f, i.. yl^o
a c >.")->:on a rr;',_0'e:t-i-*"t in ~:-: j ge-dir.r^ • ,-r
' ov. ^j s. "'•;" ' ••f.'.ui'i "-o\, .'~ii , be to a.j~>'! n'-ni
i.ic c< n -'-r c. ' ;' er^." "c- '. r-a j -, o .• it , :-a / ! < - :.i
t''«!3 Design and Efficient
For t.ie mi.st cfficJeut operation, bu'>ule Ctip
tray tcwor g '^iittt be desigin.d to 1,0.7: oror.i I ,c
opposing t( noc.'icies. High ]jc1.iid icvel.-j LII
tlit tr• hut
-------
Gas Absorption Equipment
225
LIQUID
DOKN
LIQUID
DOWN
INLET
»EIR
OVERFLOW
»EIR
T I
1 I J
VAPOR UP
1101) ID
DOWN
VftPOR UP
b.
u.
VAPOR
UP
Figure 152. Vapor and liquid flow patterns for
bubble cap tray towers: (a) One-pass tray bub-
ble plate column, liquid crossflow, opposite di-
rection on alternate plates; (b) one-pass tray
bubble plate column, liquid crossflow, same di-
rection all plates; (c) two-pass tray bubble
plate column, split liquid crossflow, opposite
directions on alternate plates; (d) one-pass
cascade tray bubble plate column, liquid cross-
flow, opposite direction on alternate plates
(Erjmister, 1948).
also give high pressure drop per tray. High
gas velocity, within limits, gives efficient
vapor-liquid contact by creating turbulent
conditions but also leads to high pressure
drop as well as high liquid entrainment.
Treybal (1955) lists recommended condi-
tions and dimensions for bubble cap trays
that have been found to be a useful com-
promise; these are listed in Table 63. In
this table, the liquid seal (hg) is the depth
of clear liquid over the top of the bubble
cap slots.
As stated before, each tray or plate is a sep-
arate stage and, for ultimate efficiency, the gas
and liquid would leave each tray in equilibrium
with each other at tray conditions. This would
be a theoretical plate. This theoretical condi-
tion does not normally exist in practice and
thus the actual number of trays required to
accomplish a specified degree of absorption
usually exceeds the number of theoretical
units required. The overall plate efficiency
of a tower is defined as the number of theoret-
ical equilibrium stages required for a given
Figure 153. Typical bubble cap
tray arrangements: (a) Single
crossflow, rectangular weirs;
(b) split crnssflow, rectangular
wei rs; (c) cascade crossflow ,
rectangular weirs; (d) reverse
flow, rectangular weir and di-
viding dam; (e) crossflow, cir-
cular wei rs; (f) radial flow,
ci rcular wei rs (Edmister, 1948).
degree of removal of solute from the gas
stream, or concentration of solute in solvent,
divided by the actual number of trays required
for this same operation. According to Clarke
(1947) an overall plate efficiency of 25 per-
cent is a conservative estimate for hydro-
carbon absorbers. O'Connell (1946) corre-
lates plate efficiency "with gas solubility and
liquid viscosity. This correlation is shown
in Figure 154. All such correlations are
empirically derived, and attempted theoret-
ical methods based on mas s-transfer prin-
ciples do not successfully predict overall
plate effic lency.
Flooding
When the liquid capacity of a plate absorber
is exceeded, the downspouts become filled.
Then, any slight increase in liquid or gas
flow increases the liquid level on the trays.
A further increase in pressure across the
trays causes more liquid to hack up through
the downspouts, resulting in still higher
liquid levels on the trays until, eventually,
the tower fills with liquid. This is known
as flooding, and at this point, the tray ef-
ficiency falls to a very low value, the gas
-------
226
CONTROL, EQUIPMENT FOR GASES AND VAPORS
Table 63. RECOMMENDED CONDITIONS AND DIMENSIONS FOR
BUBBLE CAP TRAYS (Treybal, 1955)
Tray spacing
Liquid seal
Liquid flow
Superficial slot velocity
Skirt clearance
Cap spacing
Downspout holdup
Downspout seal
Weir length
Liquid gradient
Pressure drop per tray
Tower diameter, ft
4 or less
4 or less
4 to 10
10 to 12
12 to 24
Pressure
Vacuum
Atm
500 lb/in2
a. Not over 0.22 ft3/sec-ft
diameter for single-pass
crossflow trays
b. Not over 0. 35 ft3/sec-ft
•weir length for others
3. 4/pp 0.5 ft/sec minimum
12/pE 0. 5 ft/sec maximum
0. 5 in. minimum; 1. 5 in. for
dirty liquids
1 in. minimum (low slot ve-
locities); 3 in. maximum
(high slot velocities)
Minimum of 0. 5 sec
0.5 in. minimum at no liquid
flow
Straight rectangular weirs for
crossflow trays, 0.6 to 0. 8 of
tower diameter
0. 5 in. (1 in. maximum)
Pressure
Atm
300 lb/in2
Tray spacing, in.
6 minimum
18 to 20
24
30
36
Liquid seal,
Hhs, in.
0. 5
1
3
Pressure drop
0. 07 to 0. 12 lb/in2
0. 15 lb/in2
flow is erratic, and liquid may be forced
out the gas exit pipe at the top of the tower.
Flooding occurs more rapidly with liquids
that tend to froth.
Tower design should allow sufficient down-
spout area and tray spacing to prevent flood-
ing under anticipated operation variations in
both gas and liquid flow. If there is any
question, it is better to over-design down-
spouts since they represent a relatively
small-cost item but are important from the
standpoint of potential flooding.
Liquid Gradient on Plate
The liquid gradient on a plate is the de-
creasing liquid depth from the liquid inlet
to outlet of the plate due to resistance to
fluid flow by the bubble caps and risers.
If this gradient is appreciable, more vapor
flows through the bubble caps where the
liquid depth is least. In extreme condi-
tions the caps near the liquid inlet may
become completely inoperative and liquid
may flow down through the risers. This
is called an unstable plate. Liquid gra-
GPO 806—6 I 4—9
-------
Gas Absorption Equipment
227
FACTOR, IHML
0 I
1 0
10
100 000
FACTOR
Figure 154. Correlation of plate efficiencies of gas absorbers
with gas solubility and liquid viscosity according to method of
O'Connell (Sherwood and Pigford, 1952, p. 301).
dient problems would naturally be more
likely in large towers, and in these cases,
the vapor distribution is controlled by two-
pass, split-flow/, cascade- or radial-type
trays.
nance and are not placed close together un-
less headroom limits the overall tower height.
Six inches is usually a minimum, even for
very small-diameter towers, and 18 to 24
inches is normally used for towers up to 4
feet in diameter.
Plcte Spacing
Operationally, the main consideration re-
garding tray spacing is to allow sufficient
space for the desired liquid level plus
space above the liquid for disengagement
of the gas and liquid phases without en-
trainment. Thus, in this respect, tray
spacing is closely related to gas velocity
through the tower. Spacing should also be
sufficient to provide insurance against
flooding. If flooding conditions exist even
for a short time, a tower with closely
spaced trays could become flooded. In
actual practice, however, trays are normal-
ly spaced for ease in cleaning and mainte-
Tower Diameter
The superficial linear gas velocity that will usu-
ally ensure against excessive entrainment is chosen
by the equation
where
PL
P,
G ~
K =
V = K
(98)
liquid density, Ib/ft
gas density, Ib/ft
an empirical constant.
-------
228
CONTROL EQUIPMENT FOR GASES AND VAPORS
The constant K can be determined by Figure 155,
which, is based on results of experimental study
and good commercial practice. The velocity cal-
culated in equation 98 is valid except for hydro-
carbon absorbers, which, according to Perry
(1950), should be designed for vapor velocities
65 to 80 percent that of the calculated values.
From this calculated velocity, if the volumetric
gas flow rate is known, the diameter can easily
be determined. In most cases the diameter
chosen in this manner is also adequate to han-
dle the normally expected liquid flow rate.
Treybal (1955) states that a well-designed single-
pass crossflow tray usually handles up to 100 gpm
per foot of diameter without excessive liquid
gradient.
= mole fraction of solute in liquid stream
at dilute end of countercurrent tower.
Illustrative Problem
The following example illustrates a method of
determining the number of plates or trays re-
quired and estimated diameter for a tray tower.
No attempt is made to design the bubble cap
plate itself for characteristics such as number
of caps, cap spacing, slot dimensions, and so
forth.
Problem:
Number of Theoretical Plates
The number of theoretical plates or trays is
usually determined graphically from an oper-
ating diagram composed of an operating line
and equilibrium curve constructed as previ-
ously described in the discussion of packed
towers. The actual procedure will be de-
scribed in the example problem that follows-.
If the solute concentrations in the gas and
liquid phases are low, as is frequently the
case in air pollution control, both the equi-
librium and operating curves can be con-
sidered as straight lines, and an analytical
solution may be used. The relationship as
taken from Sherwood and Pigford (1952) is:
N = log
P e
where
log
e\mG
\ r
(99)
Determine the number of actual plates and the
diameter of a bubble cap plate tower for re-
moving 90% of the ammonia from a gas stream
containing 600 Ib-moles/hr of gas at 68 °F and
1 atm composed of 10% by volume of ammonia
and 90% by volume of air.
Solvent rate expressed as moles solute/mole
solvent is obtained from an operating line dis-
placed substantially from the equilibrium curve
(Treybal, 1955) as shown in the illustration that
follows.
Solvent rate selected is 900 Ib-moles/hr of water
at 68 °F. The tower contains 24-inch tray spac-
ing and 1-inch liquid seal and operates at iso-
thermal conditions.
546 Ib-moles/hr
residue gas ^
900 Ib-moles/V
fresh solvent
N = number of theoretical plates
P
m = slope of equilibrium curve
G = superficial molar mass flow of gas,
lb-moles/hr-ft^ column cross-sec-
tion
L = liquor rate, Ib-moles/hr-ft2 column
cross-section
Y = mole fraction of solute in gas stream
at concentrated end of countercurrent
tower
Y = mole fraction of solute in gas stream
at dilute end of countercurrent tower
Bubble cap
Tray
Tower
600 Ib-moles/hr,
feed gas
Feed gas
Residue gas
Absorbent liquid
Rich liquid
Flow,
Ib-moles /hr
600
546
900
954
,954 Ib-moles/1
rich liquid
Flow, Densit
Ib/hr lb/ft3
16,680 0.07,
15,762 0.07!
16,200 62.3
17,118 62.3
-------
Gas Absorption Equipment
329
0 2t
10 12 14 IE 18 20 24
TRAY SPACING (t) inches
30 36 40
Figure 155. Tray-spacing constants to estimate
bubble cap tray tower's superficial vapor veloc-
ity (adapted from Perry, 1950).
Solution:
1. Calculate the mole ratios of solute in gas
and liquid streams at both ends of the tower:
(a) Mole ratios at bottom of tower:
Y = —- = 0. Ill mole NH /mole air
X = -~- = 0. 06 mole NH /mole HO
(b)Mole ratios at top of tower:
Y
= 0. 0111 mole NH /mole air
j
= o.o
2. The operating line is plotted as shown in Fig-
ure 156 from the conditions at top and bottom
of the column as determined in step 1. A
straight line is drawn between points Xj, Yj
and X2, Y2.
3. The curve of ammonia-water equilibrium is
plotted on the same graph from data taken
from Leva (1953) in terms of mole ratios.
4. Number of theoretical plates or trays:
A horizontal line AB is drawn from the oper-
ating line at the conditions at the top of the
column to the equilibrium curve. Line BC
is then drawn vertically from the equilibrium
line back to the operating line. The step ABC
is a theoretical plate. The stepwise proce-
0 12
0.10
O.OB
! 0 06
,,0 04
0 02
7
0 0.02 0 04 0 06 0 08 0.10
X = moles NH3/mole H20 at 68 °F
Figure 156. Plot of operating line from the
conditions at top and bottom of bubble cap
plate tower.
dure is repeated to the end of the operating
line. The solution shows 2. 45 theoretical
plates.
5. Number of actual plates or trays:
With a viscosity, (XL, of 1 centipoise for
water and a slope of the equilibrium curve,
m, of 0. 83, (this assumes the equilibrium
curve to be straight over the area covered
by the operating line), the value mp-L is
(1)(0.83) = 0.83. From Figure 155, the
overall plate efficiency is 72%.
Actual plates required:
2. 45
0.72
= 3.4 - use 4 bubble cap trays.
6. Tower diameter:
From Figure 157, with a 24-inch tray spac-
ing and 1-inch liquid seal, K = 0. 17
(a) Superficial linear gas velocity at bottom
of tower:
V =
0 ]7/62.3 - 0.0722
= °'17\ 576^
1/Z
= 5. 00 ft/sec
-------
Z30
CONTROL EQUIPMENT FOR GASES AND VAPORS
tion. Design procedures for multicomponent
absorption are more complicated than those
described previously and will not be attempted
here. Sherwood and Pigford (1952) devote an
entire chapter to these procedures.
COMPARISON OF PACKED AND PLATE TOWERS
While devices such as agitated vessels, spray
chambers, and venturi absorbers have lim-
ited application for gas absorption, the choice
of equipment is usually between a packed tower
and a plate tower. Both devices have advantages
and disadvantages for a given operation, de-
pending upon many factors, such as flow rates
for both gas and liquid, and degree of corrosive-
ness of the streams. Final selection should be
based upon the following comparative informa-
tion:
1. Packed towers are less expensive than
plate towers where materials of con-
struction must be corrosion resistant.
This is generally true for towers less
than 2 feet in diameter.
Figure 157. Venturi scrubber or absorber with
cyclone-type liquid separator (Chemical Con-
struction Corp., New York, N.Y.).
(b) Volumetric flow rate at bottom of tower:
2. Packed towers have smaller pressure
drops than plate towers designed for the
same throughput and, thus, are more
suitable for vacuum operation.
3. Packed towers are preferred for foamy
liquids.
4. The liquid holdup is usually less in a
packed tower.
(c) Tower cross-sectional area:
64.00
5.00
= 12. 80 ft
(d) Tower diameter:
D =
(4)(12.80)
3.14
= 4. 04 ft.
The principles just discussed are for absorp-
tion of a single component. Multicomponent
absorption is of great industrial importance in
the natural gasoline, petroleum, and petro-
chemical industries. Absorption of single
components such as H^S from multicomponent
gases will be discussed in Chapter 11. When
emissions consist of mixed-solvent vapors,
control by adsorption or incineration would
probably be more economical than by absorp-
5. Plate towers are preferable where the
liquid contains suspended solids since
they can be more easily cleaned. Packed
towers tend to plug more readily.
6. Plate towers are selected in larger sizes,
to minimize channeling and reduce weight.
Channeling is corrected in the larger di-
ameter and tall packed towers by instal-
lation o£ redistributor trays at given in-
tervals.
7. Plate towers are more suitable where the
operation involves appreciable tempera-
ture variation since expansion and con-
traction due to temperature change may
crush the packing in the tower.
8. In operations where there is heat of solu-
tion that must be removed, plate towers
are superior in performance since cool-
ing coils can be easily installed on the
plates.
-------
Gas Absorption Equipment
231
Most conditions being equal, economic con-
siderations favor packed towers for sizes
up to 2 feet in diameter.
VESSELS FOR DISPERSION OF GAS IN LIQUID
Probably the simplest method of dispersing a
gas in a liquid for absorption is by injecting
the gas through a perforated pipe or sparger
of some type into a vessel filled with the liq-
uid. Unless the sparger has minute perfora-
tions, the gas bubbles formed tend to be too
large and thus present a relatively small
interfacial surface for the absorption oper-
ation. If the sparger is designed to create the
necessary small bubbles, power requirements
to force the gas through the small openings
are high.
Increased dispersion may also be achieved
by injecting the gas just below a rotating
propeller, where the shearing action of the
blade breaks up the large bubbles. With a
single vessel, the advantage of true counter-
current flow cannot be fully realized since, if
there is good agitation, the concentration of
absorbed gas in the liquid is uniform through-
out the vessel. Thus, absorption equivalent to
only one theoretical plate can be achieved per
vessel. Although absorption with this equip-
ment is usually batchwise, continuous oper-
ation can be obtained with a series of vessels
wherein the gas and liquid pass from vessel
to vessel in opposite directions.
Vessels such as these have been used to re-
move highly odorous gaseous products from the
reaction of sulfur and sperm oil in the manu-
facture of specialty lubricants. Here the ef-
fluent gases, containing a considerable per-
centage of hydrogen sulfide, are forced by their
own pressure from the closed reactor, through
a vent pipe fitted with a sparger, into a tank
filled with caustic soda. This arrangement,
without auxiliary mechanical agitation of the
liquid, reduces the odor of the effluent gas to
an innocuous level. Control, however, is ef-
fected primarily by chemical reaction rather
than by true absorption.
Small tanks containing water or caustic soda are
used to eliminate visible emissions from vents
of hydrochloric acid storage tanks during tank
loading. Without any control device, these emis-
sions of hydrogen chloride vapor are dense enough
to violate most air pollution ordinances regard-
ing opacity. The opacity can be reduced to a
negligible amount by bubbling the displaced tank
vapors through a simple perforated pipe into the
water or caustic soda.
SPRAY TOWERS AND SPRAY CHAMBERS
Interphase contact in spray-type absorbers is
achieved by dispensing the liquid in the form
of a spray and passing the gas through this
spray. In order to present a large liquid sur-
face available for contact, sprays of droplets
ranging in size from 500 to 1, 000 microns are
necessary. Fine droplets require, however,
high pressure drop across the spray nozzles,
and there is danger of liquid entrainment at all
except very low gas velocities.
In a simple countercurrent spray tower -where
the liquid is sprayed down from the top and the
gas passes upward through the spray, absorp-
tion equivalent to one transfer unit is about
all that can be expected. Unless the diameter-
to-length ratio is very small, the gas will be
•well mixed with the spray, and true counter-
current flow will not be realized. Higher gas
velocities without excessive entrainment can
be obtained with a centrifugal-type spray cham-
ber, whereby the spray droplets are forced to
the chamber walls by the centrifugal action
of tangentially entering gas before they can
be carried out the top of the chamber. With
this arrangement, there is a crossflow type
of contact, and the degree of contact is lim-
ited to about one theoretical plate or transfer
unit.
Spray chambers or towers have been used
extensively for control of particulate matter
but, according to Sherwood and Pigford (1952),
their use for pure gas absorption seems to be
limited to air conditioning or deaeration of
water where very few transfer units are re-
quired. These chambers may also be used
for some highly soluble gases when the de-
gree of required removal is small, but, in
air pollution control work, this type of oper-
ation is not common. They have been used
as precleaners for particulate removal from
gas streams where other devices are used
for ultimate control of air pollution.
VENTURI ABSORBERS
Like spray towers and spray chambers, equip-
ment using the venturi principle is primarily
used for removing particulates from gas streams,
though it has some application to gas absorption.
In gas absorbers, the necessary interphase con-
tact is obtained by differences between the ve-
locity of gas and liquid particles, and by turbu-
lence created in the venturi throat. Dispersion
in venturi devices is achieved in two ways: By
injecting the liquid into the gas stream as it
passes through the venturi, as shown in Figure
157, or by admitting the gas to the liquid stream
-------
232
CONTROL EQUIPMENT FOR GASES AND VAPORS
as it passes through the venturi, as shown in
Figure 158. In the latter case, the venturi is
also a vacuum-producing device and inspirates
the gas into the venturi throat. With both types,
a gas-liquid separation chamber is necessary
to prevent entrainment. This can be a simple
tank, the stream from the venturi tube im-
pinging on the liquid surface, or, more effi-
ciently, a cyclone-type separator.
For the unit shown in Figure 157, the gas ve-
locities in the venturi throat range from 200 to
300 feet per second, and the liquid is injected
into the stream at a rate of about 3 gpm per
1,000 cfm of gas handled. These units are
designed specifically for collection of submi-
cron particulate matter, and utilize high horse-
power. For the liquid-jet eductor types, the
liquid consumption is 50 to 100 gpm per 1, 000
cfm of gas handled at a draft of 1 inch of water.
The liquid-jet eductor types are capable of de-
veloping drafts up to 8 inches of water at high-
er liquid flow rates. They find application
principally for the absorption of soluble gases,
but are also used for collection of particulate
matter larger than 1 or 2 microns in diameter.
Venturi units obtain a high degree of liquid-gas
mixing but have a disadvantage of a relatively
short contact time. Various literature sources
DISCHARGE
have indicated a high efficiency of absorption
for very soluble gases such as sulfur dioxide
and ammonia; however, for oxides of nitrogen
where contact time is of utmost importance,
Peters (1955) reports efficiencies of absorp-
tion of from 1 to 3 percent. Because of the
high degree of efficiency of venturi scrubbers
for particulate removal, they seem desirable
for use with a dirty gas stream that also con-
tains a highly soluble gas that must be removed.
A major disadvantage of venturi units is the
high pressure drop (often as high as 30 inches
of water) with attendant high power require-
ments for operation.
NOTATIONS
Figure 158. Venturi liquid-jet
eductor-type absorber (Schutte
and Koerting Company, Cornwells
Heights, Penna.).
A
C
D
D'
f
g
G
G
h
k
K
L
Mv =
Pc =
P =
•^v
PBM=
w
U
X
I1
M-f
ft2/hr
= surface separating hot and cold media,
ft2
= specific heat, Btu/lb-°F
= outside diameter tube, ft
- inside diameter tube, ft
= diffusion coefficient,
- friction factor
= acceleration of gravity, 64. 4 ft/sec-sec
= mass velocity of flow, Ib/hr-ft
= mass velocity through baffle opening,
Ib/hr-ft2
= maximum cross flow velocity, Ib/hr-ft
= weighted mass velocity, GC x G^, Ib/hr-
ft2
= mass velocity inside the tube, Ib/hr-ft
= coefficient of heat transfer, Btu/hr-ft2 - ° F
= thermal conductivity, Btu/hr-ft -°F
= coefficient of mass transfer, Ib moles/
hr-rt atmospheres
= tube length, ft
= molecular weight of mixture (vapor plus
inert gas)
= molecular \veight of vapor
= partial pressure of vapors at tc, atm
= partial pressure of vapors at tv, atm
logarithmic mean of the vapor pressures
at the interface and at the vapor stream,
atm
total pressure on system, atm
quantity of heat, Btu/hr
cordensate temperature, °F
vapor temperature, °F
water temperature, °F
condensing coefficient for pure vapor be-
tween tc and tw, Btu/hr-ft2-°F
rate of flow, Ib/hr
latent heat, Btu/lb
viscosity at average temperature, Ib/hr-ft
viscosity at average film temperature,
Ib/hr-ft
viscosity at tube wall temperature, Ib/hr-f
density at average fluid temperature, Ib/it
-------
CHAPTER 6
METALLURGICAL EQUIPMENT
FURNACE TYPES
JOHN A. DANIELSON, Senior Air Pollution Engineer
STEEL-MANUFACTURING PROCESSES
WILLIAM F. HAMMOND, Senior Air Pollution Engineer
JAMES T. NANCE, Intermediate Air Pollution Engineer
KARL D. LUEDTKE, Intermediate Air Pollution Engineer
IRON CASTING
WILLIAM F. HAMMOND, Senior Air Pollution Engineer
JAMES T. NANCE, Intermediate Air Pollution Engineer
SECONDARY BRASS- AND BRONZE-MELTING PROCESSES
WILLIAM F. HAMMOND, Senior Air Pollution Engineer
JAMES T. NANCE, Intermediate Air Pollution Engineer
EMMET F. SPENCER, Intermediate Air Pollution Engineer*
SECONDARY ALUMINUM-MELTING PROCESSES
WILLIAM F. HAMMOND, Senior Air Pollution Engineer
HERBERT SIMON, Senior Air Pollution Engineer
SECONDARY ZINC-MELTING PROCESSES
GEORGE THOMAS, Intermediate Air Pollution Engineer
LEAD REFINING
JAMES T. NANCE, Intermediate Air Pollution Engineer
KARL D. LUEDTKE, Intermediate Air Pollution Engineer
METAL SEPARATION PROCESSES
JAMES T. NANCE, Intermediate Air Pollution Engineer
EMMET F. SPENCER, Intermediate Air Pollution Engineer*
CORE OVENS
GEORGE THOMAS, Intermediate Air Pollution Engineer
FOUNDRY SAND-HANDLING EQUIPMENT
EDWIN J. VINCENT, Intermediate Air Pollution Engineer
HEAT TREATING SYSTEMS
JULIEN A. VERSSEN, Air Pollution Engineer
-'-Now with Inorganic Chemical Division, FMC Corporation, Newark, California.
-------
CHAPTER 6
METALLURGICAL EQUIPMENT
Efficient control of air contaminants from metal-
lurgical furnaces has been achieved only in re-
cent years. Since most of these furnaces discharge
high-temperature effluents containing submicron-
size dusts and fumes, these effluents must some-
times' be cooled an3 often further conditioned be-
fore ducting to a control device. The control device
must be one capable of high-efficiency collection
of submicron particles.
This chapter discusses these control devices and
the air pollution problems encountered in steel,
iron, brass, aluminum, zinc, lead, and metal
separationprocesses. Processes related to met-
allurgical operations such as manufacture of sand
cores, foundry sand-handling equipment, and heat
treating systems will be discussed near the end of
this chapter.
For those not acquainted "with the many types of
melting furnaces, the first part of this chapter
describes briefly the more common furnaces and
their principles of operation. The air pollution
aspects of these furnaces are not discussed im-
mediately since these problems are usually a func-
tion of the specific melting process and not of the
type of furnace used.
FURNACE TYPES
REVERBERATORY FURNACE
A reverberatory furnace operates by radiating heat
from its burner flame, roof, and -walls onto the
material heated. This type of furnace was devel-
oped particularly for melting solids and for refin-
ing and heating the resulting liquids. It is gen-
erally one of the least expensive methods for melt-
ing since the flame and products of combustion
come in direct contact with the solid and molten
metal. The reverberatory furnace usually con-
sists of a shallow, generally rectangular, refrac-
tory hearth for holding the metal charge. The fur-
nace is enclosed by vertical side walls and covered
with a low, arched, refractory-lined roof. Com-
bustion of fuel occurs directly above the molten
bath; the walls and roof receive radiant heat from
the hot combustion products and, in turn, reradiate
this heat to the surface of the bath. Transfer of
heat is accomplished almost entirely by radiation.
Reverberatory furnaces are available in many types
and designs, depending upon specific job require-
ments. Probablythe largest of the reverberatory
furnaces is the open-hearth furnace, widely used
in the manufacture of steel. This furnace oper-
ates in conjunction with two heat regenerators con-
sisting of brick checkerwork; these remove the
heat from the effluent and transfer it to the incom-
ing air (Figure 1 59). The transfer is accomplished
by a system of butterfly valves, which allows the
furnace gases to pass through one set of checker-
work, giving up heat, while the incoming combus-
tion air passes through the second set of checker-
work, taking up heat. Periodically the valves are
reversed, which allows incoming combustion air
to preheat in the first set of checkerwork while
the furnace gases are heating the second regen-
erator. The charge is introduced through refrac-
tory-lined doors in the front wall; finished steel
and slag are removed through a taphole in the rear
wall. Heat is provided by passing a luminous flame
with exces s air over the charged material. Details
of operation in the production of steel with the open-
hearthfurnace are described later in this chapter.
AIR PORT
GAS PORT
REGENERATIVE
CHAMBERS
AIR PORT
GAS PORT
REGENERATIVE
CHAMBERS
Figure 159. An open-hearth furnace (Begeman, 1947).
Another type of reverberatory furnace is the cy-
lindrical furnace, commonly used in the nonferrous
industries for melting and holding small heats of
aluminum, brass, and various alloys. Cylindrical
reverberatoryfurnaces are relatively small, usu-
ally rated at 500 pounds of aluminum. These fur-
naces (Figure 160) are fired through two tangential
nozzles that promote excellent combustion charac-
teristics and provide very rapid melting. The fur-
nace may be charged through a top opening or through
the end door. The end door also serves as an ac-
cess to the metal bath for adding alloying materials
or dressing.
Reverberatory furnace designs often use rotary
tilting mechanisms. A tilting furnace promotes
ease of metal distribution for all types of casting
23S
-------
236
METALLURGICAL EQUIPMENT
Figure 160. Gas-fired,
Heating Equipment Co.,
cylindrical reverberatory furnace (Bulletin No. 6011, Hevi-Duty
Watertown, Wise.).
operations--permanent mold, die casting, andsand
operations. Charging is accomplished by means
of a hopper that acts as a stack for the exhaust
gases; the metal charge lodges in the lower part of
the hopper where the melting takes place. The
furnace is end fired, and tilting of the furnace is
accomplished by means of an air or hydraulic ram.
Another type of tilting reverberatory furnace (Fig-
ure 161) normally finds application in nonferrous
metallurgical operations where large heats are re-
quired. In this installation, the furnace is gas
fired tangentially with three burners.
Many other variations and combinations of furnaces
using the reverberatory principle are manufactured
bymany firms throughout the United States and are
available commercially as prefabricated units.
CUPOLA FURNACE
For many years the cupola has been a standard
melting furnace for producing gray iron. It is also
used to melt or reduce copper, brasses, bronzes,
and lead. In addition to its high efficiency, the
cupola is simple in its construction and operation.
Unless carefully considered, however, its oper-
ation may lead to difficulties because of variations
in quantity and quality of raw metal, fuel, and air.
The basic equipment for a gray iron-melting oper-
ation consists of the cupola (Figure 162), which is
essentially a refractory-lined cylinder open at the
top and equipped with air ports (known as tuyeres)
at the bottom. Air is supplied from a forced-draft
blower. Alternate charges of metal, coke, and
limestone are placed on top of the burning coke bed
to fill the cupola. The heat generated melts the
metal, -which is drawn off through a tap hole. The
two principal dimensions of the cupola are its di-
ameter and operating height (charging door to tu-
yeres). The diameter determines the melting ca-
pacity, and the height affects the thermal efficiency.
Combustion Air
The control of air at the tuyeres influences produc-
tion rates, costs, metal losses, coke ratios, stack
-------
Furnace Types
237
Figure 161. Tangentially fired tilting reverberatory furnace (Bulletin No. 6011, Hevi-
Duty Heating Equipment Co., Watertown, Wise.).
temperature, physical properties of the metal, and
volume of stack emissions. Air is required, not
only to furnish oxygen for the combustion of coke,
which supplies the heat required for melting the
iron, but also to aid in the potential combustion of
the carbon, silicon, and manganese in the metal.
The latter function greatly influences the resultant
chemical and physical properties of the metal when
it is poured into the mold (Molcohy, 1950).
Combustion air may be provided by a positive-dis-
placement-type blower or a centrifugal blower.
The quantity of air theoretically required is deter-
mined primarily by the size of the cupola, the
melting rate, the metal-coke ratio, and the metal
temperature. The actual air supplied may be in-
creased as much as 15 percent to compensate for
leakage. Air pressure varies from 8 to 40 ounces
per square inch, depending upon design factors
such as ductwork layout, tuyere geometry, and the
height of the bed through which the air must be
forced. Automatic controls are frequently in-
stalled to maintain a constant-weight flow of air.
Methods of Charging
Various methods of charging materials into the
cupola are used. The smaller cupolas are fre-
quently charged by hand while larger units may be
charged with skip hoists with the various types of
cars, buckets, cranes, or trolleys. Charging and
melting is a continuous operation.
Preheating Combustion Air
In order to increase the efficiency of a cupola, three
methods are available for preheating combustion
-------
238
METALLURGICAL EQUIPMENT
Figure 162. A cupola furnace (American Foundrymen's
Association, 1949).
air. In the Moore system, a heat exchanger is
usedto transfer some of the waste heat of the stack
gases to the incoming combustion air. The Whiting
system uses a separate external heater for the
combustion air. The Griffin system passes the
stackgases through a chamber where air is intro-
duced and the CO is burned to CO2. The gases
then pass through a heat exchanger to preheat the
combustion air.
ELECTRIC FURNACE
Major advantages of the electric furnace over fuel-
fir ed furnaces are furnace atmosphere control and
high-temperature operation. Temperatures as
highas 6, 000°F are possible for special processes.
The electric furnace has three functions (Porter
1959):
1. Synthesis of compounds not available in the
natural statebyfusing selected raw materials,
2. purification of ores,
3. alteration of crystalline structure of ores hav-
ing a satisfactory chemical purity but an un-
desirable crystal structure.
There are four types of electric furnace: Direct-
arc, indirect-arc, resistance, and induction. Each
of these types will be discussed briefly.
Direct-Arc Furnace
In the direct-arc furnace, many and varied ar-
rangements are used to heat the metal charge, but
radiation between arc and the metal bath is the
principal method. Here, the heat is generated by
radiation from the arc as well as from the resis-
tance heat effect within the bath, as shown in Fig-
ure 163. Graphite and carbon electrodes are usu-
allyused and are spaced just below the surface of
the slag cover. The current passes from one elec-
trode through the slag, the metal charge, the slag,
and back to the other electrode. In some arrange-
ments, the current is carried from the metal
charge to the hearth. The slag serves a protective
function by shielding the metal charge from vapor-
ized carbon and the extremely high temperatures
at the arc.
Indirect-Arc Furnace
In the indirect-arc furnace, the metal charge is
placedbelowthe electrodes, and the arc is formed
between the electrodes and above the charge (Fig-
ure 163). Indirect-arc furnaces are used mainly
in the steel industry. One of the common smaller
furnaces is the indirect-arc rocking furnace, in
which an automatic rocking action of the furnace is
employed to ensure a homogeneous melt. This is
done by mounting the refractory-lined steel shell
on cog bearings so that the furnace may be rocked
through a 200° range. Radiated heat from the in-
direct arc, and conduction from the preheated re-
fractory lining initially melt small scrap, form-
ing a pool of molten metal at the bottom of the fur-
nace. Then the rocking action is initiated, and the
molten metal washes against the refractory, pick-
ing up additional heat, which is transferred by con-
vection and radiation to the larger pieces of metal.
During the heat, the rocking action is advanced
gradually to avoid a sudden tumbling of cold metal,
•which could fracture the graphite electrodes.
ELECTRODES'
DIRECT
^CHARGE
INDIRECT
Figure 163. Principles of operation
of two types of arc furnaces (Porter,
1959).
-------
Furnace Types
239
Induction Furnace
Resistance Furnace
The induction furnace consists of a crucible with-
in a water-cooled copper coil (Figure 164). An
alternating current in the coil around the crucible
induces eddy currents in the metal charge and thus
develops heat within the mass of the charge. The
furnace is used for the production of both ferrous
andnonferrous metals and alloys, generally from
scrapmetal. It provides good furnace atmosphere
control and can be used for large-volume produc-
tion of high-purity materials.
Three varieties of resistance furnaces are illus-
trated in Figure 165. The resistance furnace is
essentiallya refractory-lined chamber with elec-
trodes, movable or fixed, buried in the charge.
It is characterized by its simplicity of design and
operation. The charge itself acts as an electrical
resistance that generates heat.
The resistance furnace is used in the production
of ferroalloys (ferrochrome, ferrosilicon, and
others), cyanamide, silicon carbide, and graphite,
and in hardening and tempering tools and machine
parts.
CHARGE
CHARGE-
Figure 164. Principles of oper-
ation of an induction furnace
(Porter, 1959).
CRUCIBLE FURNACE
Crucible furnaces, used to melt metals having
melting points below 2,500°F are usually con-
structed •with a shell of •welded steel lined with re-
fractory materials. Their covers are constructed
of materials similar to the inner shell lining and
have a small hole over the crucible for charging
metal and exhausting the products of combustion.
The crucible rests on a pedestal in the center of
the furnace and is commonly constructed of a re-
fractory material such as clay-graphite mixtures
or silicon carbide. Crucibles are made in several
shapes and sizes for melting from 20 to 2, 000
pounds, rated in red brass.
Crucible furnaces are classified as tilting, pit, or
stationary furnaces. All types are provided with
one or more gas or oil burners mounted near the
•ELECTRODE
ELECTRODE-
Figure 165. Principles of operation of three types of resistance
furnace (Porter, 1959).
-------
240
METALLURGICAL EQUIPMENT
bottom of the unit. Flames ars directed tangen-
tially around the inside of the furnace. The cruci-
ble is heated both by radiation and by contact with
the hot gases.
Tilting Furnace
The tilting crucible furnace (Figure 166) is pro-
vided with devices for affixing the crucible to the
furnace so that the furnace may be tilted with the
crucible when the metal is poured. The entire
furnace is mounted on trunnions, around which the
furnace maybe tilted. The tilting mechanism can
be operated manually, hydraulically, or electrically.
Figure 166. TiI ting crucible furnace (Lind-
berg Engineering Co., Downey, Calif.).
Pit Crucible
The pit crucible furnace derives its name from
its location. The top of the furnace is near floor
level, which facilitates charging of the metal to
the furnace and removing of the crucible for pour-
ing. Pouring is usually accomplished by using
the same crucible as a ladle. The furnace cover
is provided with rollers or swinging mechanisms
for easy removal.
Stationary Crucible
The stationary crucible furnace is almost identical
to a pit furnace except that it is not sunk in a pit,
These furnaces are commonly used as holding fur-
naces, and the metal is poured by dipping with hand
ladles. Pouring may also be accomplished by re-
moving the crucible and using it as a ladle.
POT FURNACE
Pot furnaces are used to melt metals with melting
temperatures below 1, 400 "F. These furnaces may
be cylindrical or rectangular and consist of an
outer shell lined with refractory material, a com-
bustion chamber, and a pot. The pots are made
of pressed steel, cast steel, or cast iron with
flanged tops. The flange rests on the furnace
wall, holds the pot above the furnace floor, and
seals the contents of the pot from the products
of combustion of the fuel used. The shape of
the pot depends upon the operation to be con-
ducted. Large rectangular furnaces, general-
ly called kettles, are used to melt large amounts
of metal for dipping operations, such as galvaniz-
ing. For melting large castings , shallow, large-
diameter pots are used. When ingots or other
small pieces of metal are to be melted, deep pots
are used to promote better heat transfer. Pot
furnaces are usually emptied by tilting, dipping,
or pumping. A small pot furnace is shown in
Figure 167. Combustion equipment ranges from
Figure 167. A gas-fired small pot furnace
(Lindberg Engineering Co., Downey, Calif.).
-------
Steel Manufacturing Processes
241
simple atmospheric -type burners located directly
belowthepot to premix-type burners tangential-
ly fired as in crucible furnaces. The larger ket-
tles are generally provided with many small
burners along both sides of the pot.
STEEL-MANUFACTURING PROCESSES
Steel is a crystalline alloy, mainly of iron and
carbon, which attains greater hardness when
quenched from above its critical temperature than
•when cooled slowly. Carbon is the most important
constituent because of its effect on the strength of
the steel and its ability to harden. Other constitu-
ents that may be present as impurities or as
added alloying elements include manganese, sili-
con, phosphorus, sulfur, aluminum, nickel, chro-
mium, cobalt, molybdenum, vanadium, and copper
(Begeman, 1947).
Steel is made from pig iron and scrap steel by
oxidizing the impurities, reducing the iron oxides
to iron, and adding the desired alloying constitu-
ents. The two common steel-refining processes
are: (1) The basic process, wherein oxidation
takes place in combination with a strong base such
as lime; and (2) the acid process, wherein oxida-
tion takes place without the base addition. The
two processes have the common pur pose of remov-
ing the undesirable elements in the metal by the
chemical reaction of oxidation reduction. Depend-
ing upon the alloy being produced, the elements
removed from a melt may be silicon, sulfur, man-
ganese, phosphorus, or carbon. These elements
are not removed by direct chemical reaction but
by indirect reaction. Forabasic refining process,
limestone is added as a flux, and iron ore or mill
scale as an oxidizing agent. The reactions may
be shown as follows (Clapp and Clark, 1944):
CCX
C
Fe3C
3SiO2
Mn
MnO + SiO
+
+
-f
+
+
oaou<3
TT .-. r\
^ oa^» -t-
*• CU +
* U e(J)2
L>>^2
CO
2Fe
Fe
4Fe
(Si02)3(Slag)
Sulfur is partially removed in the following man-
ner, CaO + FeS—»• CaS + FeO. The resulting CaS
is taken up by the slag.
For an acid refining process the sequence of reac-
tions can be shown in a similar manner as follows
(Clapp and Clark, 1944):
2Fe
3Fe
Si
Si
Mn
Mn
4C
C
The metallic oxides and silicon then form slags
according to the equations:
^2
202
<*£ ^3^4
Fe3°4
Fe304
T~r\
Fe3°4
6FeO
2Fe
3FeO
Fe
3Fe
FeO
2FeO
MnO
2MnO
Si02
FeO
SiO,
3SiO_
— (FeO)2 •
-*• MnO • SiO,
i
-*• (MnO)^ •
2Fe3P
2
SFeO
-* MnO • SiO2 (Slag)
-- (FeO) • P205 + HFe
Steel-refining processes are usually accomplished
in the open hearth furnace, the electric furnace,
or the Bessemer converter.
Open-hearth furnaces have an approximate range
of 40 to 550 tons' capacity per heat with most
falling in the 100- to 200-ton range. Because of
the large capacities of these furnaces, they lend
themselves to large-volume steel production.
The three types of electric furnaces used are the
direct-arc, the indirect-arc, and the induction.
Electric furnaces are most often used where only
small quantities of pig iron are readily available
and where remelting of steel scrap, or small heats
of special alloys are required. Sometimes these
furnaces are used with open-hearth furnaces. In
such cases, the steel is first processed in an open-
hearth furnace and is then further refined or al-
loyed in an electric furnace.
Still in limited use today is the Bessemer con-
verter. It consists of a pear-shaped vessel or
converter, mounted on trunnions and easily tilted
for charging and pouring. Oxidation of manga-
nese, silicon, and carbon is accomplished by blow-
ing air through the molten metal. Converters have
been largely replaced owing to the increased pro-
duction rates achieved by the open-hearth and elec-
tric furnaces.
-------
242
METALLURGICAL EQUIPMENT
In I960 over 4 million tons of steel (2. 7 percent
of the total production) was produced by a recent-
ly developed process called the oxygen process.
This is similar to the Bessemer process in that
an oxidizing gas, oxygen instead of air in this case,
is blown through the molten metal. This oxygen-
blowing process can be used as a rapid source of
heat control to increase the temperature of the
furnace bath or may be used to refine the metal
by oxidizing the undesirable elements in the bath.
The principal advantage of this process is that it
shortens the refining time and thus reduces pro-
duction costs.
In the oxygen process, pure oxygen is immediate-
ly available to promote oxidation of the impurities
in the bath. If oxygen is used to reduce the carbon
content, then carbon monoxide and iron oxide are
formed, some oxygen remaining in the bath. Fig-
ure 168 shows this relationship for various bath
carbon percentages. In the oxygen process, the
oxygen also reacts at a slower rate with other ele-
ments such as silicon, manganese, and chromium
to reduce the content of these elements in the mol-
ten bath.
Steel-making capacity in the United States by type
of furnace is depicted in Table 64. In I960 over
85 percent of the steel-making operating capacity
100
80
60
40
S 20
EL>
a.
z:
4J
! 8
i 6
5 A
3=
^
^-«
•*^-
A
/
/
SLA(
SLAG
=55
— *.
y
/
^
BA:
VOL
•-,
/
1C
UME
v
/
•-
Tl
f
«,
1
^
s
'
fc
'
3.
70
, S^
x
/ \
\y
^
^V
X
\
"
0
Ib/ton
4*
r
\
\
|L
\
\
-e:
k
\
\
\
\
V
\
^
L_
\l
\
JL
OXYGEN
TO CO
OXYGEN TO
SLAG. FeO
OXYGEN TO
STEEL. 0
'0 02 0 04 0 06 01 02 04 06
BATH CARBON, percent
Figure 168. The oxygen reaction in molten steel
(Obrzut, 1958).
was represented by 906 open-hearth furnaces, 10
percent, by 301 electric furnaces, and 5 percent,
by 31 Bessemer converters and 12 oxygen process
furnaces. Total operating capacity was 143, 571, 000
tons.
Table 64. NUMBER AND CAPACITIES OF
STEEL FURNACES OPERATED IN
UNITED STATES, I960 (Steel Facts, American
Iron and Steel Institutes, New York, New York)
Furnace type
Open hearth
Electric
Bessemer
Oxygen process
Number
906
301
31
12
Capacity, tons
126,621, 630
14, 395, 940
3, 396, 000
4, 157, 400
The air contaminants vented from steel-melting
furnaces include gases , smoke, fumes, and dusts.
The quantities of these contaminants in the efflu-
ent gas stream depend upon the types of material
charged to the furnace. The gaseous emissions
result from the combustion of fuels and other com-
bustible contaminants in the furnace charge and
fromthe refining process. Smoke emissions re-
sult from incomplete combustion of the combusti-
bles in the fiirnace charge or of furnace fuels.
Particulate emissions originate partially from dirt
and impurities in the charge, but the major quantity
results from the refining process.
A study of the chemical reactions of the refining
processes reveals that a large portion of the par-
ticulate matter is emitted from steel furnaces in
the form of metallic oxides. These characteris-
tics are illustrated in Table 65, where the results
of a spectrographic analysis of the particulate dis -
charge from an open-hearth furnace are given, and
in Table 66, which gives a typical analysis of the
particulate discharge from an electric-arc fur-
nace. These fume emissions or metallic oxides
are very small, 65 to 70 percent falling into the
0- to 5 -micron range. Table 67 shows a size
analysis of the particulate emissions from an oper
hearth furnace and two electric -arc furnaces
along with other data. For a visual concept o
particle size and shapes, electron photomicro-
graphs of fumes from an electric-arc furnace aric
an open-hearth furnace are shown in Figures 16'
and 170.
OPEN-HEARTH FURNACES
The open-hearth furnace, which features the re
generative principle, was invented by William
Siemers in 1358. Although many improvement
-------
Steel Manufacturing Processes
243
Table 65. SPECTROGRAPHIC ANALYSIS OF
PARTICULATE DISCHARGE FROM AN OPEN
HEARTH FURNACEa
Element
Fe
Zn
Na
K
Al
Ca
Cr
Ni
Pb
Si
Sn
Cu
Mn
Mg
Li
Ba
Sr
Ag
Mo
Ti
V
Approximate amount, %
Remaining amount
10 to 15
1 to 2
1 to 2
5
5
2
2
5
5
1
0. 5
0. 5
0. 1
Trace
Trace
Trace
0. 05
Trace
Trace
0. 05
^These data are qualitative only and require
supplementary quantitative analysis for actual
amounts .
Table 66. TYPICAL EMISSIONS FROM
AN ELECTRIC-ARC FURNACE
(Coulter, 1954)
Component
Zinc oxide (ZnO)
Iron oxides
Lime (CaO)
Manganese oxide (MnO)
Alumina (A12O3)
Sulfur trioxide (SO3)
Silica (Si02)
Magnesium oxide (MgO)
Copper oxide (CuO)
Phosphorus pentoxide (P2O^)
Weight %
37
25
6
4
3
3
2
2
0.2
0. 2
and refinements have been made since then, the
process remains essentially the same. There are
roughly four methods of making basic open-hearth
steel in the United States. These are classed ac-
cording to the iron-bearing materials in the charge
as follows (Kirk and Othmer, 1947):
1. Hot metal (pig iron) and molten steel. By this
method, iron from the blast furnace, and steel
from the Bessemer converter are refined in
the open-hearth furnace.
2. Cold steel scrap and cold pig iron. This com-
bination is used by plants that have access to
supplies of inexpensive scrap and do not have
a blast furnace.
3. All steel scrap, This process is uncommon
in the American steel industry.
4. Steel scrap and molten pig iron. Most of the
integrated steel plants use this method, which
is the predominant process in the United States
and Canada.
In the last method, a typical initial charge consists
of 55 percent cold pig iron and 45 percent steel
scrap. Limestone and iron ore, equal in quanti-
ty to approximately 7 and 4percent, respectively,
of the total weight of the cold metal charged, are
also added. If molten pig iron cannot be obtained
in sufficient quantity to compl ete the initial charge,
more cold pig is charged with the scrap, and the
entire mass is heated in the furnace. The process
continues for approximately 2 hours until the scrap
has reached a temperature of about 2, 500°F and
has slightly fused. Molten pig is then added and
a lively action occur s in which almost all the sili-
con, manganese, and phosphorus, and part of the
carbon are oxidized. The first three elements
form compounds that slag with iron oxide and join
the iron and Hme silicates that are already melted.
The ore acts on the carbon for 3 or 4 hours long-
er while the limestone forms carbon dioxide and
completes the purification. The lime boil lasts for
another 2 or 3 hours and the heat is then ready to
be adjusted for final carbon content by adding pig
iron, ore, oroxygengas. Thedescribed operation
is commonly divided into three phases consisting
of the ore boil, the lime boil, and the working
period.
The heat for the process is provided by passing a
luminous flame with excess air over the charged
materials. The combustion air is alternately pre-
heated by two regenerating units, which, in turn,
are heatedby the products of combustion discharg-
ing from the furnace.
The Air Pollution Problem
Air contaminants are emitted from an open-hearth
furnace throughout the process, or heat, which
lasts from 8 to 10 hours. These contaminants can
be categorized as combustion contaminants and
refining contaminants. Combustion contaminants
result from steel scrap, which contains grease,
oil, or other combustible material, and from the
furnace fuel.
The particulate emissions that occur in greatest
quantities are the fumes, or oxides, of the vari-
ous metal constituents in the steel alloy being made.
These fumes are formed in accordance with the
-------
244
METALLURGICAL EQUIPMENT
Table 67. DUST AND FUME DISCHARGE FROM STEEL FURNACES
Test number
Furnace data
Type of furnace
Size of furnace
Process wt, Ib/hr
Stack gas data
Volume, scfm
Temperature, °F
Dust and fume data
Type of control equipment
Concentration, gr/scf
Dust emissions, Ib/hr
Particle size, wt %
0 to 5 (j.
5 to 10 u
10 to 20 n
20 to 44 (a-
> 44 [i
Specific gravity
1
Electric arc
2 ton and 5 tona
3, 755
7, 541
125
None
0. 1245
8. 05
67. 9
6.8
9. 8
9. 0
6. 5
--
2
Electric arc
50 ton
28, 823
23, 920
209
None
0. 5373
110. 16
71.9
8.3
6.0
7. 5
6.3
3.93
3
Open hearth
50 ton
13, 300
14, 150
1, 270
None
1. 13
137
64.7
6.79
11.9
8. 96
7.65
5
aBoth furnaces are vented by a common exhaust system and were tested
simultaneously.
refining chemistry previously discussed. The
concentration of the particulates in the gas stream
varies over a wide range during the heat, from
0. 10 to a maximum of 2. 0 grains per cubic foot
(Allen et al. , 1952). An average is 0. 7 grain per
cubic foot, or 16 pounds per ton of material charged.
The test results in Table 67 for the open-hearth
furnace show that 64. 7 percent of the emissions
are below 5 microns in size. The control device
selected must, therefore, be capable of high col-
lection efficiencies on small particles.
volume of gases to be vented from the furnace, the
maximum fuel input must be known:
Example 25
Given:
60-ton open-hearth furnace. Fuel input = 35 Ib
of U. S. Grade No. 6 fuel oil per min.
Another serious air pollution problem occurring
\vithopen-hearthfurnaceoperation is that of fluo-
ride emissions. These emissions have affected
plants, which in turn, have caused chronic poison-
ing of animals. Surveys have shown that fluorides
are contained in some iron ores such as those
mined in southern Utah. Control of fluoride emis-
sions presents a problembecause these emissions
are in both the gaseous and particulate state.
Hooding and Ventilation Requirements
The design parameters for an open-hearth furnace
control system for duct sizes, gas velocities, and
so forth are the same as those to be outlined for
the electric-arc furnace. In order to establish the
Problem:
Determine the volume of gases to be vented from
the furnace stack to the air pollution control sys-
tem.
Solution:
1. Volume of products of combustion from oil
burners:
One pound of U. S. Grade No. 6 fuel oil with
theoretical air produces 186. 1 scf gas (see
Table D6 in Appendix D).
V = 35 x 186.1 = 6,510 scfm
JPC
-------
Steel Manufacturing Processes
245
Figure 169. Electron photomicrographs of fumes from an electric furnace producing steel
for castings (Allen et al., 1952).
2.
Volume of air infiltrated through leaks owing
to reversing valves, stack dampers, cracks
in bricks, and so forth:
Assume the average 150 percent excess air
(combustion and infiltration) usually found in
the stacks of regenerative furnaces. Theo-
retical air for 1 pound of U. S. Grade No. 6
fuel oil is 177. Z scf (see Table D6 in Appen-
dix D).
V = 35 x 177. 2 x 1. 5 = 9, 320 scfm
E A
3. Total volume at 60°F to air pollution control
equipment:
V = V + V
T60 PC EA
= 6,510 + 9,320 = 15, 830 scfm
The temperature of the furnace gases leaving
the regenerator will be approximately 1, 300 °F.
In some installations, this heat source is used
to generate steam, by delivering the gases to
a waste heat boiler in which the temperature
would be reduced to about 500°F.
-------
246
METALLURGICAL EQUIPMENT
I
Figure 170. Electron photomicrographs of fumes from a cold-metal open-hearth steel
furnace (Allen et al., 1952).
4. Total volume of the air pollution control equip-
ment at 500°F operating temperature:
V
T500
= 15,830 x
460 + 500
460 + 60
= 29,200 cfm
Since the efficient operation of the open-hearth
furnace requires that all the products of combus-
tion, along with the air contaminants created in the
furnace, are to be conducted through the regener-
ator and then to a stack, it is necessary only to
direct the flow from the stack through suitable
ductwork to the control system. The size of the
blower must, of course, be increased to overcome
the additional resistance introduced by the control
system.
Air Pollution Control Equipment
Open-hearth furnaces have been successfully con-
trolled by electrical precipitators. On some in-
stallations, the control system has been refined
by installing a waste heat boiler between furnace
and control device. In this manner, heat is re-
claimed from the furnace exhaust gases, and at
the same time, the gases are reduced in tempera-
ture to within the design limits of the control de-
vice. In Table 68 are shown test results of a con-
trol system wherein the waste heat boiler and
electrical precipitator vent an open-hearth fur-
nace. This test was made on one of four control
systems installed to serve open-hearth furnaces.
These control systems are shown in Figure 171.
-------
Steel Manufacturing Processes
247
Table 68. DUST AND FUME EMISSIONS FROM
AN OPEN-HEARTH FURNACE SERVED BY AN
ELECTRICAL PRECIPITATOR
Furnace data:
Type of furnace (constructed
1916)
Size of furnace, tons
Test interval
Fuel input
Waste heat boiler data:
Gas volume, inlet, scfm
Gas temperature, inlet, °F
Gas temperature, outlet, °F
Water in waste gas, %
Steam production (average),
Ib/hr
Open hearth
63
1 hr during heat -working period
Natural gas, 21, 000 cfh
Fuel oil, 1.4 gpm
14,900
1, 330
460
12. 4
8,400
Precipitation data:
Gas volume, scfm
Dust and fume concentration
(dry volume)
Inlet, gr/scf
Outlet, gr/scf
Inlet, Ib/hr
Outlet, Ib/hr
Collection efficiency, %
14,900
0. 355
0. 004
39.6
0.406
98. 98
The factors to be considered in designing an elec-
trical precipitator to control the emissions from
an open-hearth furnace are the same as those that
will be described next.
Electric-Art Furnaces
The electric-arc furnace lends itself to accurate
control of temperature and time of reaction for
alloy composition. These advantages are achieved
because no harmful gases are emitted from an
electric arc that would otherwise produce an ad-
verse effect upon the metal being refined. Steel
may be produced in an arc furnace by either the
basic or the acid process. The furnace may be
charged with molten metal from an open-hearth
furnace (an operation known as duplexing), or it
may be charged with cold steel scrap. Owing to
the close control that can be achieved, low-grade
scrap can. be refined to meet close specifications
of the various steel alloys.
After the furnace has been charged with metal,
fluxes and other additions required to accomplish
the refining chemistry are charged according to
schedule. The additions vary depending upon the
steel being produced and the metal charged. Lime
is usually a basic addition along with others, such
as sand, fluorspar, iron ore, carbon, pig iron,
and other alloying elements. The operation then
continues in three phases: (1) The oxidizing peri-
od, in which the undesirable elements are oxidized
from the metal and removed as slag, (2) the re-
ducing period, in which oxygen is removed from
the metal mostly through the reaction with carbon,
and (3) the finishing period, in which additions are
made to bring the alloy within the desired specifi-
cations. The make-up of a typical charge to an
electric-arc furnace is shown in Table 69.
Table 69. TYPICAL CHARGE FOR AN
ELECTRIC-ARC FURNACE (Coulter, 1954)
Material
Fluxes, carbon, and ore
Turnings and borings
Home scrap
No. 2 baled scrap
Miscellaneous scrap (auto, etc)
Weight %
5
7
20
25
43
-------
248
METALLURGICAL EQUIPMENT
Figure 171. Electrical precipitators serving open-hearth furnaces.
The Air Pollution Problem
Hooding and Ventilation Requirements
The quantity and type of fumes emitted from an
electric-arc furnace depend upon several factors:
Furnace size, type of scrap, composition of scrap,
cleanliness of scrap, type of furnace process,
order of charging materials, melting rate, refin-
ing procedure, and tapping temperature. A large
portion of the fumes generated in a furnace is re-
tained in the slag; however, sizable quantities of
fumes escape and are discharged from the furnace
vent. Table 70 shows emission data for several
arc fur_naces, which vary from 4. 5 to 29. 4 pounds
per ton of metal melted. Most of the emissions
originate during the first half of the heat. Fig-
ure 172 shows a curve of emission rates during
the single heat.
Before the emissions can be collected they must
first be captured through some suitable hooding
arrangement at the furnace and must then be con-
veyed to a collection device that has a high collec-
tion efficiency on small particles.
Three types of hooding arrangements can be in-
stalled. The first is a canopy-type hood, which
is suspended directly over the furnace (Figure 173).
Ahood such as this has serious deficiencies in that
it must be mounted high enough above the furnace
to clear the electrodes and not interfere with the
crane when overhead charging is employed. As
the distance between the furnace and hood is in-
creased, the volume of air to be inspirated into
-------
Steel Manufacturing Processes
251
Figure 174. Close-fitting plenum-type hood serving an electric-arc furnace:
(left) furnace filled with hood in place, (right) furnace with hood removed
(Soule Steel Company, Los Angeles, Calif.).
-------
252
METALLURGICAL, EQUIPMENT
Figure 175. (Left) electric-arc furnace with plenum hood,
baghouse (National Supply, Torrance, Calif.).
(right) venting to a
Figure 176. Direct roof tap on an electric-arc
furnace (Alloy Steel and Metals Company, Los
Angeles, Calif.).
terial. Screw conveyors are frequently installed
on the hoppers as an aid for removing the materi-
al collected.
Control unit assembly must be constructed of ma-
terials that can withstand the temperatures of the
furnace and the effluent gas stream. Provision
should also be made to prevent sparks and burn-
ing material from entering the collector.
An outline of some of the design features of bag-
houses that serve electric steel-melting furnaces
is included in Table 71. Only one of these instal-
lations was equipped for reverse air cleaning.
This particular baghouse has been replaced with
a conventional shake cleaning unit because of the
high maintenance costs associated with the re-
verse air cleaning mechanism and because of the
excessive bag wear.
In Table 72 are shown test results of air pollution
control systems with baghouses serving electric-
arc steel-melting furnaces. The collection effi-
ciencies of the baghouses in tests 1, 2, and 3 are
within the range of expected efficiencies for in-
stallations of this type. In tests 4 and 5 the col-
lection efficiencies are subnormal, indicating mal-
function of the systems. This was evident at the
time of the tests from the visible discharge of
dust and fumes from the baghouse outlets. An in-
vestigation disclosed that those two baghouses had
many defective bags. The results are, however,
reported to emphasize the necessity of checking
the veracity of tests such as these.
Electrical precipitators
An electrical precipitator may be used to control
the emissions from an electric-arc furnace. The
fundamental design considerations for hooding, air
volume cooling, duct sizing, and fan selection are
the same as those outlined for baghouse control.
The one major difference pertains to the condition-
ing of the effluent gas stream. A baghouse sys-
tem should be designed so that the gas tempera-
ture remains below the maximum operating tem-
perature of the cloth bags and above the dewpoint.
For an electrical precipitator, control must be
much more accurate. The apparent resistivity
-------
Steel Manufacturing Processes
253
o
rt
o
H
U
w
J
J
o
o
w
w
t>
o
£
O
Q
o
w
w
H
w
o
o
w
J
w
X
o
fn
w
Q
U
n
ro o
-H r-
i
ro ro — «
ro o
""•-•- uTi
ro ro ^H
fM O
~-- O
— < O
1 „
fM rM -H
-M
fM O
---. 0
— " LT)
ro ro r— i
O
O
fM °°
^ ~~~- o
00 O
^ rsT " ^
1—1
^ O
o
!;
ro - ^ °
fM "^ 0 °
- -^, ^ o 0°
•^ ' _r ^
fM O
^~ \ ^ LD 00
^ f^1 _L t*~ "
-H -H O
r-
— » fNJ ^ O
-H — t O
0
NO
-H un r- o
— < — i 0
0
NO
o
o
m
— t
o
CO NO CO C-5
o
un un un
rg 'Nj
0
0
o
un
° •- » ^ ™d
o) £ £ 3 ^ oi i
,a 0 ° u C c >
g rt rt 2 u ^
3 S " oT "> - "> .
C S Usi 0(u u W>
( ll CU-H ™OJ) ^ rt
(tir- Cro £ A* £•"•1-'
-*->h ^ _i ^rt *-< -M
^ 3 2 (3 "^ ,-£ ^ oj
He. hd £HO [=4 h
:- ^ i
GO
O
fM M O
0) GO
fM >H
•~-^. (V1
1
O
o
, 0 ^
^M
fM
* O
^ en o
„ CJ O
fM
ro
fM
fM
—i
^5 §
f^ O
o
vD
fM
^
^ ° §
Z o
o
" 1 1
5-
H °
M" °
~~^ X^"'
-H
3
screen
i-.
ni
"3
Xl
3
•*_>
tubular
^
,0
3
^
nj
3
-§
*^
0)
_0 oS
0
0
o ^
^
0 ^
x
d °
2 t~
0 ~
ti M
0
U ^
§ S
i* ro
° vO
fM
C
0 "^
un
_5 o
o °°.
0
fM
c °,
0 £>
^ *~~
u
^ ^
Filter material
Filter area, ft2
01
o
^ 03 fl) rt
°. 3 G £
C "o "-1
•JH 0 4J
e -if u
o 2 v
U "* .H
Q
m
S g -
0 O
2 PH
ro
fM
^ ^ " >"
"° CO
0 0
2 «
un
CO |
§^
CO 5*
tM ^
^ ° t! °
2 S °
•O Pi
c
• H
O
a
r-1 nj
-1-1 *+-i
^0 M g
sO QJ
-0 OH
C
U
fM
~ s i §
" Z «
vD
CT-
-H O
c 'o
0 O
?, A
u
_0 o
« t h
u ^
CD C
C £ M rt
C P" c ^
i - ni .5 j;
tu " En E (L
t
0)
Vl
3
1/3
0
0
01
01
t,
3
0)
o
u
R
£5
n)
u
01
03
0
C
0)
IH
enclosu.
JH
3
D
"o
C
l;
o
t4
3
yi
o
o
"
0
a
•u
o
o
a
1 — '-
3
cx
1
ao
c
f
•*
0)
H
on
r*
Telescop:
(
be
S
escop
OJ
H
oc
'EL
o
01
O
'oj
H
4-J
fj
S
T!
T)
O
Q
X
C ^
(U •*->
GO y
S -S
G M
£ J
01 w
C1 3
C M
f^
•^ J
C/}
c
Permane
Water
"n
DC C3 CX
C nl rt
1 1 £
£3 "~U QJ
t^ § -^
QC C CH
C n) rt
fti U ^
•-J 3 In
3 -O rn
° ^
00
• S t)
1 ™
-^ O
rd
00 6
! " s
-H O >
3 o £
H ° "
^, ul
0! -H
connection
Cooling
method
(i
rt
* 4-J
!§ C
c ™
5 Ml
01 _rt
C ">
S, °"
>- I "°
. -O i t)
^ C ' 3
to rd
-------
254
METALLURGICAL EQUIPMENT
Table 72. DUST AND FUME EMISSIONS FROM ELECTRIC-ARC
STEEL FURNACES WITH BAGHOUSE CONTROLS
Test number
Furnace data
Type of furnace
Size of furnace,
tons
Process wt, Ib/hr
Baghouse data
Type of
baghouse
Filter material
Filter area, ft2
Filtering velocity,
fpm
Dust and fume data
Gas flow rate,scfm
Inlet
Outlet
Gas temperature, °F
Inlet average
Outlet average
Concentration,
gr /scf
Inlet
Outlet
Dust and fume
emission, Ib/hr
Inlet
Outlet
Control efficiency,%
Particle size, wt %
Inlet, 0 to 5 fj.
5 to 10
10 to 20
20 to 40
> 40
Outlet
1
3 -electrode
Direct arc
17
13, 700
Sectioned
tubular
Orion
20, 800
1.95
38, 400
40, 600
172
137
0. 507
0.003
166.9
1. 04
99.4
72. 0
10.5
2. 7
4.7
10. 1
100% < 2 p.
2
3 -electrode
Direct arc
3-1/2
4, 250
Compartmented
tubular
Orion
5, 540
1.78
10, 300
9, 900
135
106
0.346
0. 0067
30.5
0.57
98. 1
57. 2
37. 8
3.4
1.6
0
100% < 1 H-
3
Two, 3-electrode
Direct arc
4/4
3, 380/5, 131
Sectioned
tubular
Orion
11, 760
1.20
12, 960
14, 110
129
121
0. 398
0. 0065
44.2
0.79
98. 2
63. 3
17.7
8.0
8. 1
2. 9
100% < 2 p.
4
3-electrode
Direct arc
14
17, 650
Compartmented
tubular
Orion
25,760
1. 23
18, 700
31, 700
186
139
0. 370
0. 0158
59. 3
4. 3
92. 7a
59. 0
33. 1
4.9
3. 0
0. 0
72% < 5 (J.
5
3 -electrode
Direct arc
19
22, 300
Sectioned
tubular
Orion
26, 304
1. 75
42, 300
46, 100
167
153
0. 462
0. 047
167. 5
18. 5
88. 9a
43. 3
17. 7
6. 4
14. 60
18. 0
75% < 5 IJL
aAn investigation disclosed that poor efficiencies were due to defective bags in the baghouse.
of the material to be collected must first be de-
termined. After this is known, the condition of
the gas stream, and the temperature and humid-
ity that will result in the most efficient collec-
tion can be determined. Efficient collection usu-
allyfalls within a narrow temperature range, in
which case the conditioning system must be de-
signed to maintain the effluent gas stream with-
in thatrange. Figure 177 shows the relationship
between temperature, humidity, and collection
efficiency for an electrical precipitator serving
an electric-arc furnace in a specific installation.
For this particular installation an acceptable ef-
ficiency was not realized until the gas tempera-
ture was maintained below 127 °F and the humid-
ity above 49 percent. Table 73 shows operating
data for two installations of electrical precip-
itators serving electric-arc furnaces.
One general equation (Brief et al. , 1956) for ex-
pressing precipitator efficiency is
L_
E = 1 - KV
where
E = collection efficiency
(100)
-------
Steel Manufacturing Processes
255
K = precipitation constant (always less than
unity and dependent upon the resistivity
of the fume for a specific degree of gas
conditioning)
L = electrode length, ft
V = volumetric flow rate, cfm.
This equation shows some of the factors that must
be consideredbefore the control system can be de-
signed. Factors such as efficiency required, re-
sistivity of fume, gas conditioning, geometry of
precipitator, and others should all be discussed
with a manufacturer of electricalprecipitators be-
fore the design of the control system is formu-
lated. General design information on electrical
precipitators has been discussed in Chapter 4.
Water scrubbers
Water scrubbers have been used in many process-
es in which some contaminant must be removed
from a gas stream. These same scrubbing meth-
ods have been used to control the emissions from
electric-arc steel furnaces with varied results.
65
57
53
49
45
37
33
60
85 90 95
PRECIPITATOR EFFICIENCY, percent
145
143
141
139 '
137
135
133
131
129
127
125
100
Figure 177. Curves showing effects of variation
of the gas stream's temperature and humidity
upon efficiency of a specific electrical-pre-
cipitator installation (Coulter, 1954). -
Table 73. OPERATING DATA OF ELECTRICAL-PRECIPITATOR
CONTROL SYSTEMS SERVING ELECTRIC-ARC FURNACES
(Brief et al. , 1956)
Case
Operational data
Inlet gas volume, cfm
Inlet gas temperature, °F
Absolute humidity, Ib/lb dry gas
Inlet fume concentration, gr/ft^
Electrical-precipitator data
Type
Rectification
Size
Gas velocity, fps
Gas retention time, sec
Electrode length
, ,v .. electrode length 2
L/ v raLio . , sec/It
volumetric rate
Gas conditioner data
Type
Collection efficiency
A
105, 000
127
0.045
0. 68a to 1. 35b
High-eff plate
Mech, full wave
30 ducts, 10 in. x
1 8 ft x 1 8 ft
3.9
4.6
11, 880
6. 8:1
2-stage
evaporative
cooler
97 + %
B
33, 500
80
Ambient
0. 1 15
Exp metal plate
Mech, full wave
19 ducts, 8-3/4 in. x
17 ft 6 in. x 18 ft
2. 3
7. 8
7, 550
13.6:1
Radiation and
tempering air
cooler
92%
aAverage for one 50-ton and two 75-ton furnaces processing normal scrap.
Average for one 50-ton and two 75-ton furnaces processing dirty, subquality
scrap.
-------
256
METALLURGICAL EQUIPMENT
Table 74 shows the results of six tests on water
scrubbers serving electric-arc steel-melting fur-
naces. Wet collectors collect only the larger
particles and allow the submicr on par tides to pass
through and be discharged to the atmosphere.
These submicronparticles cause the greatest dif-
fusion of light and thus produce the greatest visual
opacity. A venturi scrubber can be operated at
greater efficiencies than those achieved by the
scrubbers depicted in Table 74. A basic disad-
vantage of many scrubbers is that their efficien-
cy of collection is proportional to their power in-
put; thus, if a scrubber has the feature of high col-
lection efficiency, the power input required to
realize this high efficiency is also large. In any
event, the decision to install a scrubber over some
other type of control device depends principally
upon the collection efficiency required and the
comparative costs of installation and operation.
ELECTRIC-INDUCTION FURNACE
The electric-induction furnace uses the material
to be heated as a secondary of a transformer.
When a high-frequency current is applied to the
furnace coils, an electromagnetic field is set up
in the core or space occupied by the metal to be
melted. This high-density electromagnetic field
induces currents in the metal, causing it to heat
and melt. These furnaces range in size from 30
pounds' to 8 tons' capacity. They are not well
adapted to a refining process and, for the most
part, are used for preparation of special alloys,
or for certified true heats , or for investment cast-
ings.
The Air Pollution Problem
The fume emissions from an electric-induction
furnace processing steel alloys have the same
characteristics as those from electric-arc fur-
naces. Since a high degree of control is exer-
cised in preparing alloys in this type of furnace,
metals contaminated •with combustible elements
such as rubber, grease, and so forth are not
charged to the furnace. This practice eliminates
the need for control of combustible contaminants.
The quantity of contaminants emitted from induc-
tion furnaces processing steel alloys varies. The
factors affecting the fume generation include com-
position of alloy, method of making the alloy ad-
dition, temperature of the melt, and size of the
furnace. When these factors are controlled, some
steel alloys can be made without the need of air
pollution control equipment.
Hooding and Ventilation Requirements
Since induction furnaces are relatively small, the
canopy-type hood is readily adaptable to capturing
the fumes. Recommended hood indraft velocities
vary from 200 to 500 fpm, depending upon the
hood, furnace geometry, cross-drafts, and tem-
peratures involved. The following example prob-
lem shows a method of calculating ventilation re-
quirements for a canopy-type hood serving an
induction furnace:
Example 26
Given:
1, 000-lb capacity electric-induction steel melt-
ing furnace
Pouring temperature = 3, 000°F
Diameter of crucible = 2 ft
Surface area of molten meta] = 3. 14 ft
Hood height above furnace = 3 ft
Room air temperature = 100 °F.
Problem:
Determine the minimum ventilation requirements
for the furnace.
Table 74. HYDROSTATIC SCRUBBER DATA
Test
Total number of furnaces
Furnace size, tons
Process wt, Ib/hr
Volume of gases inlet, scfm
Volume of gases outlet, scfm
Gas temperature inlet, °F
Gas temperature outlet, °F
Fume concentration inlet, gr/scf
Fume concentration outlet, gr/scf
Fume emission inlet, Ib/hr
Fume emission outlet, Ib/hr
Collection efficiency, /o
A
2
6 and 20
12, 444
17, 500
20, 600
132
89
0. 158
0. 055
23.7
9.71
59. 1
B
1
20
4, 720
22,700
24, 600
123
76
0. 0657
0. 0441
12.8
9. 3
27. 3
C
1
6
6,240
20,700
20,700
110
92
0. 167
0. 102
29.6
13. 2
55. 4
D
2
3 and 3
5,020
10, 140
10, 860
145
92.5
0. 329
0. 108
28. 7
10. 1
65
E
1
50
27,200
25, 900
29,800
297
99
0. 423
0. 109
94
27. 8
70. 4
F
1
75
43, 900
32, 400
35, 600
281
105
0. 966
0. 551
268
168
37. 3
-------
Steel Manufacturing Processes
257
Solution:
q = 5.4 A (m)1/3 (At)5/12 (from Chapter 3)
•where
q = rate of thermal air motion at top of heat
source, cfm
A = surface area of hot body and face area
S of hood, ft2
m = diameter of crucible, ft. For lack of
proved experimental values for m, the
diameter of the molten metal (heat source)
will be used in the operation
At = temperature differential bet-ween hot body
and room air, °F
q = (5.4)(3.14)(2)1/3(2,900)5/12
q = 590 cfm
The formula used in calculating the ventilation re-
quirements is accurate only for low-canopy hoods
having an area equal to that of the heat source and
having a maximum height of approximately 3 feet
above the furnace. For high-canopy hoods, the
hood area and ventilation volume mustbe increased
Air Pollution Control Equipment
The design considerations for the remainder of the
control system, including ductwork, type of col-
lector, and fan and motor selection, are the same
as outlined for electric-arc furnaces. Figure 178
is a. photograph of two induction furnaces served
by a canopy-type hood that vents to a baghouse.
Figure 178. Canopy-type hood serving two electric-induction furnaces (Centrifugal Casting,
Long Beach Calif.).
-------
258
METALLURGICAL EQUIPMENT
IRON CASTING
Control of the air pollution that results from the
melting and casting of iron may be conveniently
considered according to the type of furnace em-
ployed. The cupola, electric, and reverberatory
furnaces are the types most -widely encountered.
The air pollutants are similar, regardless of the
furnace used; the primary differences among the
air pollution control systems of the various fur-
nace types are to be found in the variations in hood-
ing, and the necessary preparation and treatment
of the contaminated gases from the furnaces. Es-
sentially, the air pollution problem becomes one
of entraining the smoke, dust, and fumes at the
furnace and transporting these contaminants to
suitable collectors.
CUPOLA FURNACES
The most widely encountered piece of equipment in
the gray iron industry is the cupola furnace. High
production rates are possible and production costs
per ton of metal are relatively low. Despite this,
where the product permits , some gray iron found-
ries have substituted reverberatory furnaces for
their cupolas rather than install the air pollution
control equipment that cupolas require. Table 75
shows one manufacturer's recommendations for
operating cupolas.
The Air Pollution Problem
Air contaminants emitted from cupola furnaces
are (1) gases, (2) dust and fumes, and (3) smoke
and oil vapor. The following is a typical cupola
combustion gas analysis: Carbon dioxide, 12. 2
percent; carbon monoxide, 11.2 percent; oxygen,
0. 4 percent; nitrogen, 76. 2 percent. Twenty to
thirty per cent "by weight of the fumes are less than
5 microns in size. A particle size analysis of the
dust and fumes collected from gray iron cupolas
is shown in Table 76, as are some emission rates.
Tables 77 and 73 show micromerograph and spec-
trographic particle size analysis of two sa.mples
taken from the hoppers of a bag filter serving a
gray iron cupola furnace. Dust in the discharge
gases arises from dirt on the metal charge and
from fines in the coke and lime stone charge. Smoke
and oil vapor jirise primarily from the partial com-
bustion and distillation of oil from greasy scrap
charged to the furnace.
Hooding and Ventilation Requirements
One way to capture the contaminants discharged
from a cupola furnace is to seal the cupola top
and vent all the gases to a control system. A
second method is to provide a vent in the side of
the cupola a few feet below the top of the burden
and vent the gases to a control system. The con-
trol system consists of an afterburner, a gas-
cooling device, and a dust collector, which is
either a baghouse or an electrical precipitator.
The system must be designed to exhaust enough
gas volume to remove all the products of combus-
tion from the cupola and to inspirate sufficient air
at the charge opening to prevent cupola gas dis-
charge at that point. In addition, the exhaust gas
volume must be sufficient to remove the products
of combustion from the afterburner section. In
cupolas of large diameter (over 36 in. ), enclosure
of the charge opening with refractory-lined or
water-cooled doors is usually necessary. These
doors are pneumatically operated to open only
during the actual dumping of a charge into the cu-
pola.
Even though a closed top cupola is equipped with a
door to cover the charge opening, it is common
practice to design the ventilation unit to provide
at least 250 fpm average indraft velocity across
the full open area of the charge opening.
Air Pollution Control Equipment
Collection efficiencies of several small-scale con-
trol devices on gray iron cupolas are shown in
Table 79. These tests indicate the superior ef-
ficiencies of baghouses and electrical precipita-
tors and, in practice, only these devices have been
found to operate satisfactorily in Los Angeles
County. As mentioned, these systems also include
auxiliary items such as afterburners, gas-cooling
devices, and settling chambers.
Afterburners
An afterburner is generally installed in a cupola
furnace control system for two reasons. The
high carbon monoxide content of the cupola ef-
fluent presents a definite explosion hazard; this
hazard can be avoided by burning the carbon
monoxide to carbon dioxide. Secondly, the after-
burner burns combustion particulates, such as
coke breeze and any smoke and oil vapors that
may be distilled from the furnace charge. This
combustion of oil vapors prevents later condensa-
tion on the surface of the filter bags and their re-
sultant blinding. While afterburners may be in-
stalled as separate units, the common practice is
to use the upper portion of the cupola between the
charging door and the cupola top as the afterburn-
er. When this is done, the height of the standard
cupola must usually be increased to give a vol-
ume sufficient to provide adequate residence time
to complete the combustion in the afterburner.
As described earlier, the pollution problem from
the various iron processes originates from emis-
sion of gases, dust, fumes, and smoke. The
ratios of the quantities of the contaminants emit-
ted from this equipment vary appreciably and
influence the selection of the control device or
devices to be employed.
GPO 80S—614-10
-------
Iron Casting
259
O >
^E
OH
s«
co r^
2 ^
o ^
I—I
H o
< Z
Q .
£H S
H .2
^ -*-1
Is
°6
W ^
_, o
J ft
rt "
W a
N
O
01
£ ^ £
M ^3
!ls f
T3 o *•"
5; u Q p.
M <" C
oc £
C/}
•^ in
- g£l
*s *w ^ u
£ 3
to
H S
_] o
en
fM
w — •
0) -^
"c 2
rfl o
•
fM
I.SS o
*j o ,£ —
U (fl
«T c H
"rt ° -a"
tH r- to
5 -^
vD
1M
* S
< 3 t
c
Diameter
inside
lining, in?"
c S | -
S^~ 5
•s ° S
^
~* v
"a S r
J= P S
m 2
•5
'o "
a 2
u "
"22
000
- _
O O O
-1
o o o
--< fM rn
O O O
— i rsj fM
O m m
CO CO ^f
•tf1 rg fM
0 O O
OO -i) ^D
-^ "-•
i— i —< f-g
f Tf
^ J<
rf in f\j
m — < r-
oo co r~
f-g "g f-g
i i i
5 S S
O -H fM
sO sO
0 0
rO O
- ^
0 O
-. — i
-^
o
0 O
0 0
in m
O O
sC 0
*
-^ --.
,i ^
Tf fM
CO i— t
fM ^
r- o
m oo
•
- -
5 *
fM
fM CO
0 ^ I^'
fM r\) rg
O O O
^, S fM
O O O
0 O 0
_
o o o
CO CO -*
sD 00 0
000
— i vD 00
un o o
CO — i CO
^D OO CO
o o o
0 rg r^
un r- co
--. -- -^
4 ^ ^
^ rg
7 "
m m o
r* co 0s
o ro un
r- fM un
— ' ^O co
in un o
rg
co 1* in
O 0 O
^ fM ^
0 O 0
* ^ t-
O O O
000
fM en "T
O O O
^f fM O
r- fM r-
o o o
— i — . fM
O O O
r-- oo o
— i rr co
000
r^ o >n
CO — < fO
T CO fM
O O O
un un m
0 m r-
o J. *
-~. ^. -.
ri J- -i
fM
^
o o r-
CO fNI CO
— < fM fM
OO ^f O
S £ £
-O r- co
000
fM eg rg
O 0 O
O 0 0
CO O — i
O O O
000
^ a, o
IT, xo r-
o o o
o o o
000
r- eg co
fM fO rO
o o o
O O 00
fM O O
rg -M m
O O O
m m o
t^ fM CO
in un in
0 o o
m r- (^
o ^* co
fM "g rM
r*- o ro
^r ^r
^ ^ a.
fM ^
7 7
-i ~i CO
tot-
^D fM OO
CO 0s CT1
rM
o o
o o
CO r\j
fM CO
0 0
0 0
-H ^H
0 0
O 0
2 ™
r- oo
0 0
0 0
o o
in in
co -*
co ^r
o o
o m
CO r-1
O O
o r~
ro xD
•M rg
CO T
m m
O O
r- r-
CO CO
rg co
rg rg
0s rg
CO rg
t- m
CO ^T
r- co
- -,
fM CO
0 0
^H rg
T3
en
« ^
~ rt
C H
^ *a
S o
to ao
*_§
u c
1 s.
-------
260
METALLURGICAL EQUIPMENT
Table 76. DUST AND FUME EMISSIONS FROM GRAY IRON CUPOLAS
Test- No.
Cupola data
Inside diameter, in.
Tuyere air, scfm
Iron- coke ratio
Process wt, Ib/hr
Stack gas data
Volume, scfm
Temperature, °F
coz, %
02, %
CO. %
N2> %
Dust and fume data
Type of control
equipment
Concentration, gr/scf
Inlet
Outlet
Dust emission, Ib/hr
Inlet
Outlet
Control efficiency, %
Particle size, wt %
0 to 5 \i
5 to 10 \i
10 to 20 H-
20 to 44 H-
> 44 [i
Specific gravity
1
60
-
7/1
8, 200
8, 300
1, 085
-
-
-
-
None
-
0. 913
-
65 •
-
18. 1
6.8
12.8
32.9
29. 3
3. 34
2
37
1, 950
6.66/1
8, 380
5, 520
1, 400
12. 3
-
-
-
None
-
1. 32
-
62.4
-
17. 2
8. 5
10. 1
17. 3
46. 9
2. 78
3
63
7, 500
10. 1/1
39, 100
30, 500
213
2.8
-
-
-
None
-
0. 413
-
108
-
23.6
4. 5
4.8
9.5
57.9
4
56
-
6. 5/1
24, 650
17, 700
210
4.7
12. 7
0
67. 5
Baghouse
1. 33
0. 051
197
7. 7
96
25. 8
6. 3
2. 2
10. Oa
55. 7b
5
42
-
9.2/1
14, 000
20, 300
430
5.2
11.8
0. 1
67. 3
Elec precip
afterburner
2.973
0. 0359
184. 7
6.24
96.6
-
-
-
_
-
6
60
-
9.6/1
36, 900
21, 000
222
-
-
-
-
Baghouse
0. 392
0. 0456
70. 6
8.2
88. 4
-
-
_
_
-
7
48
-
7.4/1
16,800
8,430
482.
-
-
-
-
Elec Precip
1. 522
0. 186
110
13. 2
87. 7
-
-
-
„
-
From 20 to 50 1i.
bGreater than 50 (a..
An afterburner should be designed with, heat ca-
pacity to raise the temperature of the combusti-
bles, inspirated air, and cupola gases to at least
1,200°F. The geometry of the secondary com-
bustion zone should be such that the products to
be incinerated have a retention time of at least
1/4 second. A luminous flame burner is desir-
able, since it presents more flame exposure.
Enough turbulence must be created in the gas
stream for thorough mixing of 9Ombustibles and
air. In large-diameter cupola furnaces, strati-
fication of the gas stream may make this a major
problem. One device, proved successful in pro-
moting mixing in large-diameter cupolas, is the
inverted cone shown in Figure 179. The combus-
tion air is inspirated through the charging door
and, if necessary, may also be inspirated through
openings strategically located in the cupola cir-
cumference, above the charging opening. The
rapid ignition of the combustible effluent by the
afterburner frequently results in a pulsating or
puffing emission discharge from the charging door.
This can be eliminated by the installation of an
ignition burner below the level of the charging
door, which ignites and partially burns the com-
bustible effluent.
A cupola afterburner need not be operated through
the entire furnace cycle. Even without an after-
burner, an active flame can be maintained in the
upper portion of the cupola. This requires con-
trol of the materials charged, and likewise, con-
trol of combustion air and mixing. The afterburn-
er must, however, be in operation during the fur-
nace light-off procedure. It is desirable to ignite
the coke bed •with gas torches, because consider-
able smoke may result if the light-off is done with
kindling wood.
Baghouse dust collectors
The temperature of the gas stream discharged
from the top of a cupola maybe as high as 2, 200°F.
If a baghouse is used as a control device, these
gases must be cooled to prevent burning or scorch-
ing of the cloth bags. Maximum temperatures
all owed vary from 180 °F for cotton bags to 500 °F
for glass fabric bags.
Cooling canbe effectedby radiant cooling columns,
evaporative water coolers, orbydilution with am-
bient air. Figure 180 shows an installation in
which the gas stream is cooled by dilution and ra-
-------
Iron Casting
261
Table 77. MICROMEROGRAPH PARTICLE SIZE
ANALYSIS OF TWO SAMPLES TAKEN FROM A
BAGHOUSE SERVING A GRAY
IRON CUPOLA FURNACE
Sample A
Equivalent
particle diameter,
r1
0. 9
1. 1
1. 4
1.8
I. 3
I. 8
3. 7
4.6
5. 5
6.4
6.9
7. 3
7. 8
8. 2
8. 7
9. 3
10. 1
11.0
12. 4
13. 7
16. 5
19. 3
22. 0
24. 7
27. 5
30. 2
34. 4
41. 3
55. 0
68. 7
82. 6
123
Cumulative
0. 0
1. 3
3. 4
7. 4
11. 6
15. 0
20. 4
24. 6
27. 3
29. 0
29. 8
30. 3
30. 7
31. 2
31.3
31.9
32. 1
33. 1
33. 5
33 6
33. 9
34. 2
34. 4
34. 7
35. 1
36. 0
37. 5
40. 6
46. 4
51. 1
55. 9
61. 4
Sample B
Equivalent
particle diameter
H
1. 0
1. 3
1. 6
2. 1
2.6
3. 0
4. 2
5. 2
6. 3
7. 3
7. 8
8. 4
8.9
9. 4
10. 1
10. 4
10. 9
12. 5
14. 1
15. 6
18. 8
21.9
25
28. 1
31.3
34. 4
39. 1
46. 9
62. 5
78. 1
93. 8
148
Cumulative
0. 0
1.7
3.6
7.0
10. 5
13. 3
19. 9
24. 8
29. 0
32. 5
34. 9
36. 3
38.6
39. 3
41. 1
42.0
43.2
45. 4
46.7
47. 0
47. 4
47.6
47. 7
48. 0
48. 4
48. 8
49.8
52. 3
56.7
63.4
69. 3
80. 5
Table 78. QUALITATIVE SPECTROGRAPHIC
ANALYSIS OF TWO SAMPLES TAKEN FROM
A BAGHOUSE SERVING A
GRAY IRON CUPOLA FURNACEa
Element
Aluminum
Antimony
Boron
Cadmium
Calcium
Chromium
Copper
Gallium
Germanium
Iron
Lead
Magnesium
Manganese
Molybdenum
Nickel
Potassium
Silicon
Silver
Tin
Titanium
Zinc
Approx amount,
%
Sample A
0. 81
0. 24
0. 050
0. 13
0. 16
0.022
0. 42
0.017
0. 018
6.0
17. 0
0. 29
1. 0
0. 0068
0. 023
1. 5
8.6
0. 0093
3. 41
0. 019
7. 1
Approx amount,
%
Sample B
1. 1
0. 24
0. 054
0. 064
0. 25
0. 019
0. 32
0. 019
0. 015
7. 5
17. 0
0. 30
0. 81
0. 0075
0. 022
1. 2
16. 0
0. 0089
0. 38
0. 034
5. 9
aThese data are qualitative only and require
supplementary quantitative analysis for actual
amounts of the elements found to be present.
These are the same samples as given in
Table 77.
Table 79. SOME COLLECTION EFFICIENCIES OF EXPERIMENTAL SMALL-
SCALE CONTROL DEVICES TESTED ON GRAY IRON CUPOLASa
Equipment tested
Controls for cupolas'11
High-efficiency cyclone
Dynamic water scrubber
Ventun-type scrubber
Dynamic --impingement
wet scrubber
Baghouse - - one silicone-
treated glass wool bag,
10 in. dia x 1 0 f t length
Evaporative cooler and
redwood pipe electrical
precipitator
Other basic equipment
Natural gas -fired
reverberatory furnace
Inlet
gas
volume,
scfm
330
1. 410
375
605
52. 7
1, 160
--
Outlet
gas
volume,
s cfm
384
1, 760
432
995
52 7
1, 330
5, 160
Inlet
dust
load,
gr/scf
1.225
1. 06
1. 17
0. °5
1. 32
1. 263
--
Outlet
dust
load,
gr/sci
0. 826
0. 522
0. 291
0 141
0. 046
0. 0289
0. 00288
Collection
efficiency,
aj
22. 5
38. 2
71. 3
75. 6
96. 5
97. 7
96.2°
Remarks
Water added before control unit for
cooling totalled 6 gpm
Two gpm-water introduced for cooling
gas stream, 3. 5 gpm added at ventun
throat, cyclonic scrubber operated dry
Water rate in excess of 10 gpm
Average temp, 372CF, average filter-
ing velocity, 3. 2.2. gpm
Water rate to cooler, 22 gpm; to
precipitator, 2 gpm
Melting rate, 546 Ib/hr, gas consump-
tion rate, 4,200 cfh, melting clean
scrap and pig iron
aln all cases, equipment was installed and operated according to the manufacturer's recommendations.
The six control devices were tested on the same cupola.
cThls is not an actual collection efficiency, but a percent reduction when compared with average cupola emissions.
-------
262
METALLURGICAL EQUIPMENT
GAS BURNERS
Figure 179. Integral afterburner with in-
verted cone installed in top part of cupola
to create turbulence to ensure complete
combusti on.
diation-convection cooling columns. Of the three
types of coolers, spraying is the most common.
All types have been discussed in Chapter 3.
For satisfactory baghouse operation, when metal-
lurgical fumes are to be collected, filtering ve-
locity should not exceed 2-1/2 fpm. Provisions
for cleaning collected material from the bags usu-
ally require compartmentation of the baghouse so
that one section of the baghouse may be isolated
and the bags shaken while the remainder of the
system is in operation. The gas temperature
throughthe baghouse should not be allowed to fall
below the dew point, because condensation within
the baghouse may cause the particles on the bag
surfaces to agglomerate, deteriorate the cloth,
and corrode the baghouse enclosure. A bypass
controlmust alsobe installed. If the cooling sys-
tem fails, the bypass is opened, which discharges
the effluent gas stream to the atmosphere and thus
prevents damage to the baigs from excessive tem-
peratures. Properly designed and maintatnedbag-
houses cannormally be expected to have efficien-
cies ranging upwards from 95 percent.
Electrical pr ecipitaitors
Electrical precipitators are an efficient control
device for collecting most metallurgical fumes
where steady-state conditions of temperature and
humidity can be maintained in the gases to be cleane
The procedures used in determining the effluent
gas volume and temperatures for a precipitator
control system are the same as those for a bag-
house control system. The collection efficiency
-------
Iron Casting
263
of an electrical precipitator depeads in part upon
the apparent resistivity of the material to be col-
lected. This, in turn, depends upon the charac-
teristics of the material, and the moisture con-
tent and temperature of the effluent gas stream.
After the condition of the gas stream under which
precipitation is to take place has been determined,
the system's conditioning units for controlling the
temperature and humidity of the effluent gas stream
can be designed. The large temperature fluctua-
tions of the effluent gas stream from a cupola re-
quire that the control system be designed to main-
Figure 180. Cupola controlled by radiation convection coolers and baghouse (Alhambra Foundry
Company, Alhambra, Cal i f.).
Cupola data
Size, 45-in. ID
Flue gas vol, 7,980 scfm
Tuyere air, 3,450 scfm
I ron - coke ratio, 8:1
Flue gas temp, 1 ,875° to 2,150°F
Charging rate, 20,200 Ib/hr
Gas conditloner data
Radiation and convection type
Cool ing area, 10,980 ft2
Log mean temp diff, 670°F
Heat trans coef, 1.59 Btu/hr-ft2 per°F
Gas vol (incl reci rculation), 16,100 scfm
Size, 16 col 42-in. dia x 42-in. H
Inlet gas temp, 1,030°F
Outlet gas temp, 404°F
Baghouse data
Tubular and compartmented type
Inlet gas volume, 13,100 cfm
Filter area, 4,835 ft2
FiIter media, si I icone glass
Shaking cycle, 90 mm (manual by
compartment)
Col lection efficiency,
Tube size, 11-in. dia x 180-in. L
Inlet gas temp, 404°F
Fi I termg velocity, 2.7 fpm
Pressure drop, 3 to 4 in. WC
-------
Z64
METALLURGICAL EQUIPMENT
tain proper levels of temperature and humidity.
Installation and operation of equipment to main-
tain these levels may be bulky and expensive, and
should be reviewed with the manufacturer. In
order to avoid corrosion in the precipitator unit,
the control system must be designed to prevent
water carryover or condensation. Figure 181
shows a cylindrical water spraiy conditioning charn-
Figure 181. Photograph of an electrical precipitator preceded by a water
vented cupola and afterburner not shown (Alabama Pipe Company, Southgate
Cupola data
Size, 42-in. ID Flue gas temp, 400° to 1,400°F
Flue gas vol, 8,700 scfm Iron - coke ratio, 9.2:1
Tuyere air, 3,000 cfm Charging rate, 14,000 Ib/hr
Afterburner data
Type of structure--an unused cupola furnace converted by the installation
of four premix gas burners with full modulatirig temperature controls to
maintain 1,100°F minimum outlet temperature. Fuel input, 10 million Btu/hr
maximum
spray conditioning chamber,
Calif.),
Evaporator cooler type
Water rate, 75 gpm (max)
Gas conditloner data
Gas temp inlet, 1,100°F mm.
Size, 10-ft 6-in. dia x 23-1t 6-m. length
Type, expanded metal
Col lectmg electrode, size, 17 ft
6 in. x 4 ft 6 i n.
Discharge electrode, 0.109-in. dia
Gas volume, 20,300 scfm
Outlet dust loss, 0.0359 gr/scf
Electrical precipitator data
No. of sections, 2 in series
Size, 23 ducts 8-3/4 in. x 17' ft 6 in.
x 9 ft
Average gas temp, 430°F
% Moisture in flue gas, 15%
Overall efficiency, 96.6
-------
Iron Casting
265
ber, upper left; electrical precipitator, center;
fan and discharge duct, upper right. These con-
trol units vent a cupola with a separate afterburner,
not shown in the photograph. The precipitator
rectifier is housed in the concrete block building
in the foreground.
Additional design information on electrical pre-
cipitators has been presented in Chapter 4.
Illustrative Problem
The following example shows some of the factors
that must be considered in designing a control
system for a gray iron cupola furnace.
Example 27
Given:
Problem:
Determine the design features of an evaporative
cooling system and a baghouse to serve the cupola.
Solution:
1. Volume of gases from tuyeres = 1,810 scfm
or 139. 3 Ib/min
2. Heat required from afterburner to raise tem-
perature of tuyere air products of combustion
from an assumed low of 500°F to a minimum
incineration temperature of 1,200°F:
A 32-in.-ID cupola
Charging door area, 4. 5 ft
Tuyere air, 1,810 scfm
Maximum gas temperature at cupola outlet,
2,000°F
Minimum incineration temperature to be main-
tained at cupola outlet, 1, 200 °F.
Assuine a closely coupled unit from the cupola to
the evaporative cooling chamber and an insulated
duct between the evaporative cooling chamber and
the baghouse.
Assume the effluent gases have the same proper-
ties as air. (Consideration of the enthalpies and
specific heats of the gaseous constituents in the
effluent gas stream •will show that this is an ac-
curate assumption. Any corrections would intro-
duce an insignificant refinement to the calculations
when considered with respect to the accuracy of
other design factors. )
Enthalpy of gas (1, 200°F) =
(see Table D3 in Appen-
dix D)
287.2 Btu/lb
Enthalpy of gas (500 ° F)
Ah
(139.3)(180.5)
= 106.7 Btu/lb
= 180.5
25. 150 Btu/min
3. Heat required from afterburner to raise charg-
ing door indraft air from 60° to 1,200°F:
Assume a charging door indraft velocity of
200 fpm, which will be adequate to ensure an
indraft of air at all times.
Charge door indraft volume = (4. 5)(200) =
900 scfm or 69. 3 Ib/min
EFFLUENT GAS TEMP
MAX - 2,000°F \
MIN = 1,200°F N
WATER SPRAY
CONDITIONING
CHAMBER
EVAPORATIVE
COOLING WATER : ?
^AFTERBURNER
MAX INPUT : •>
MAX INPUT - 500 cfh
^CHARGING DOOR
AREA - 4.5 ft2
INDRAFT VEL : 200 fpm
CUPOLA^TUYERE AIR - 1,810 scfm
FURNACE
FILTER
AREAr
FAN
BAGHOUSE INPUT
TEMP = 225°F
Figure 182. Control system for a gray iron
cupola furnace.
Enthalpy of gas (1,200°F) = 287.2 Btu/lb
Enthalpy of gas (60 °F) = 0
Ah = 287.2
(69. 3)(287. 2) = 19, 900 Btu/min
4. Total heat to be supplied by afterburner:
Heat to tuyere air
Heat to charge door
indraft volume
Total
= 25, 150 Btu/min
= 19, 900 Btu/min
45, 050 Btu/min
-------
266
METALLURGICAL EQUIPMENT
5. Requirednatural gas volume capacity of after-
burner to supply 45, 050 Btu/min:
Heating value of gas = 1, 100 Btu/ft
Heat available at 1,200°F, from the burning
of 1 ft3 of gas with theoretical air = 721. 3
Btu/ft3 (see Table D7 in Appendix D)
45,050
721.3
= 62. 4 cfm
6. Volume of products of combustion from after-
burner:
With theoretical air, 1 ft3 of gas yields 11.45
ft3 of products of combustion (see Table D7
in Appendix D)
(62.4)(11.45) = 715 cfm
7. Total volume of products to be vented from
cupola, scfm:
Volume from tuyere air = 1, 810
Volume for charge door
indraft = 900
Volume from afterburner = 715
3,425 scfm
or 264 Ib/min
8. Volume of vented gases at 1, 200°F:
I, 200 + 460^
(3
425) A. ZOO + 460\
'4"' V 60 + 460 /
= 10, 900 cfm
9. Duct diameter from cupola exit to evapora-
tive chamber:
Use design velocity of 3, 500 fpm
10, 900 , , ., , 2
Duct cross-sectional area = —-—rrr- = 3. 12 ft
3, D JU
• Use 24-in.-dia. duct
10. Cooling required to reduce temperature of
vented products from cupola from 1,200°
to 225°F:
Baghouse inlet design temperature taken as
225°F
Enthalpy of gas at 1,200°F = 287.2 Btu/lb
Enthalpy of gas at 225 °F = 39. 6 Btu/lb
Ah = 247. 6 Btu/lb
(264)(247.6) = 65, 300 Btu/min
11. Water to be evaporated to cool vented gas
products from 1,200° to 225°F:
Heat absorbed per Ib of water:
Q = h (225°F, 14. 7 psia) - h (60°F)
e l
= 1, 156. 8 - 28. 06 = 1, 128. 74 Btu/lb H2O
65,300
I 12g ?4 = 58. 0 Ib H20/min
12. Volume of evaporated cooling water at 225 °F:
v = 27. 36 ft /Ib HO (14. 7 psia, 225°F)
(58. 0)(27. 36) = 1,586 cfm
13. Total volume of products vented from spray
chamber:
Volume of products from
cupola = 3, 425 scfm
Volume of evaporated
cooling water = 1,586 cfm (225°F)
(3 425) f2" + 460N|
P,4
-------
Iron Casting
267
19. Gas velocity between cupola and spray cham-
ber when using 24-in. duct from calculation 9:
25. Filtering velocity using filter area from cal-
culation 15:
13,280
3. 142
= 4,230 fpm
Velocity is greater than necessary but not
excessive.
20. Cooling required to reduce temperature of
vented products from cupola from 2, 000°
to 225°F:
Enthalpy of gas at 2, 000°F = 509. 5 Btu/lb
Enthalpy of gas at 225 °F = 39. 6 Btu/lb
Ah = 469.9 Btu/lb
(216)(469. 9) = 101,300 Btu/min
21. Water to be evaporated to cool vented gas
products from 2, 000° to225°F:
Heat absorbed per Ib of water = 1,128.74
Btu/lb (see calculation 11)
= 9°
This is greater than that determined in cal-
culation 11 and must therefore be taken as
the design value.
22. Volume of evaporated cooling water at 225 °F:
v= 27.36 ft3/lb H20 (14. 7 psia, 225°F)
(90)(27. 36) = 2,460 cfm
23. Total volume of products vented from spray
chamber:
Volume of products from
cuP°la = 2, 805.5 scfm
Volume of evaporated
cooling water
= 2,460 cfm (225°F)
6, 160
3,053
= 2. 02 fpm
26.
This ratio is in agreement with a filtering
velocity of 2 fpm
The exhaust system and fan calculations are
made as outlined in Chapter 3. After a sys-
tem resistance curve has been calculated and
plotted, a fan is selected whose characteris-
tic curve will intersect the system curve at
the required air volume, which for this ex-
ample would be 6, 160 cfm.
Problemnote: This example problem illustrates
typical calculations that can be followed in de-
signing a cupola control system. Each installa-
tion must be evaluated separately, considering
expected maximum and minimum temperatures,
gas volumes, duct lengths, and so forth. For
example, this problem was patterned after a small
cupola where the charging door remains open.
For large cupolas, opening and closing the charg-
ing doors must be evaluated relative to its effect
upon gas volumes and temperatures. If duct runs
are long, the radiation-convection losses may be
worth considering. The sizing of the fan motor
depends upon the weight of gas moved per unit
time. This in turn depends upon the density (con-
sidering air, water vapor, and temperature) of
the gas stream. These items are taken into con-
sideration in making the exhaust system resis-
tance calculations. It may be necessary to re-
duce the system's airflow by dampering in order
to prevent overloading of the fan motor when mak-
ing a cold startup under ambient conditions. See
Chapter 3 for design parameters for cooling of
effluent gas stream with radiation-convection
cooling columns. Since the temperature of the
effluent gas stream from the cupola will fall in
the range of 1, 200° to 2, 000°F,the duct connect-
ing the cupola and water spray conditioning cham-
ber should be made of stainless steel or be re-
fractory lined.
ELECTRIC-ARC FURNACES
24. Gas velocity between spray chamber andbag-
houseusing 18-in. duct from calculation 14:
6, 160
1. 767
= 3, 480 cfm
Velocity is comparable with design value of
3,500 fpm
Electric-arc furnaces are commonly used in the
secondary melting of iron where special alloys
are to be made. These furnaces may be either
the direct- or indirect-arc type. Pig iron and
scrap iron are charged to the furnace and melted,
and alloying elements and fluxes are added at
specified intervals. These furnaces have the ad-
vantage of rapid and accurate heat control.
-------
268
METALLURGICAL EQUIPMENT
The Air Pollution Problem
Since no gases are used in the heating process,
some undesirable effects on the metal are elim-
inated. Since arc furnaces in the iron industry
are virtually always used to prepare special al-
loy irons, the quality of the material charged is
closely controlled. The charging of greasy scrap,
which would emit combustible air contaminants,
would only needlessly complicate the alloying
procedure. Afterburners are, therefore, rarely
required in conjunction with arc furnace opera-
tions. The emissions consist, primarily, of
metallurgical fumes and can be controlled by
either a baghouse or an electrical precipitator.
The emission rates from electric-arc furnaces
vary according to the process from 5 to 10 pounds
per ton of metal processed.
Hooding and Ventilation Requirements
Direct-arc furnaces for melting gray iron pre-
sent a unique and difficult problem of hooding.
The hood's geometry and the indraft velocity must
be designed to ensure virtually complete collec-
tion of the emissions from the furnace. Hood de-
sign varies considerably for different furnaces.
Furnaces are most successfully hooded by build-
ing the hood into the cover or top of the furnace.
This, of course,, means designing an air cham-
ber or compartment above the furnace roof and
providing a duct connection to the chamber so
that the collected contaminants may be vented to
the control device. Since direct-arc furnaces
receive only a limited use for melting cast iron,
generalizing about the ventilation requriements
is difficult; however, 5, 000 to 7, 500 cfm per ton
of production apparently yields a reasonable de-
gree of dust and fume capture. To be most ef-
fective, the ventilation air exhausted from the
furnace should also be available to the .furnace
hood during periods of tapping and charging the
furnace. This means that some type of telescop-
ing ductwork or slip-connection ductwork must
form the connection between the control device
and the hood. Figure 183 illustrates an adjust-
able-type hood used with a baghouse venting rock-
ing-arc furnaces. The hood is positioned by
means of a telescoping connection that is me-
chanically controlled. In the right foreground of
the photograph, the hood is shown lowered into
position -with the furnace in operation, while in
the left background, thehoodis shown raised from
the furnace to facilitate charging and furnace ac-
cess.
Air Pollution Control Equipment
Baghouse dust collectors
Elaborate facilities for cooling the effluent gas
streamfrom an electric furnace may not be nec-
essary for two reasons: (1) No products of com-
bustion result from the burning of fuel, and (2)
canopy-type or roof-type hoods are generally
used. Thus, the volume of the effluent gas stream
.'.slow, and the ventilation air drawn in at the hood
provides cooling. As with cupola baghouses, a
filtering velocity of 2-] /2 fpm should not be ex-
ceeded and a shaking mechanism and compart -
mentation must be provided.
Electrical precipitators
As in the case of baghouse dust collectors serv-
ing electric-arc furnaces, no elaborate facili-
ties are necessary for cooling the effluent gas
stream from an electric furnace vented to an
electrical precipitator, though the design humid-
ity and temperature of gases entering the elec-
trical precipitator must be met. This may re-
quire water spray sections or afterburner devices
toheatand humidify the gases vented to the pre-
cipitator.
INDUCTION FURNACES
Core-type electric-induction furnaces are also
used for melting cast iron. In this type of fur-
nace, alternating current is passed through a pri-
mary coil with a solid iron core. The molten
iron contained within a loop that surrounds the
primary coil acts as the secondary. The alter-
nating current that flows through the primary
indxices a current in the loop, and the electrical
resistance of the molten metal creates the heat
for melting. The heated metal circulates to the
main furnace chamber and is replaced by cooler
metal. This circulation results in uniform metal
temperature and alloy composition.
The electric-induction furnace generates con-
siderably smaller amounts of air contaminants
than the cupola or electric-airc furnace does; the
amount is mainly dependent upon the condition of
the metal charged. When pig iron and clean cast-
ing returns are charged, no air pollution control
equipment is necessary for ordinary melting.
Contaminated scrap or the addition of magnesium
for manufacturing ductile iron would, however,
necessitate air pollution control equipment. In
cases such as these, design requirements for a
baghouse control system with canopy-type hood-
ing are the same as later described in this chap-
ter for coreless induction furnaces for steel melt-
ing.
REVERBERATORY FURNACES
Small reverberatory furnaces are also used in
preparing gray and white cast iron alloys. If
clean metal is charged to these furnaces, no ex-
cessive air pollution results from their use. Fig-
ure 184 shows a small, gas-fired, reverberatory
-------
Iron Casting
269
Figure 183. Rocking-arc furnaces venting through adjustable hoods to a baghouse
(Centrifugal Casting Company, Long Beach, Calif.).
Figure 184. Gray iron reverberatory furnace (Pomona Foundry, Pomona, Calif.).
Reverberatory furnace data
Rated capacity, 1,000 Ib Typical charge, 300 Ib pig iron, 500 Ib
Fuel, natural gas scrap iron, 200 Ib foundry returns,
Furnace flue gases, calculated at 2 Ib soda ash
6,100 cfm at 2,850°F Melting rate, 750 Ib/hr
Pouring temp, 2,700°F Fuel rate, 4,200 ftVhr
Test data
Gas volume at hood, 5,160 scfm
Dust loss in gr/scf, 0.00278
Average gas temp, 775°F
Loss in Ib/hr, 0.13
-------
270
METALLURGICAL EQUIPMENT
furnace used in a gray iron foundry. Test results
made upon a. furnace of this type, rated at 1, 000-
pounds capacity, while it was melting clean scrap
iron and pig iron, showed a particulate loss to
the atmosphere of 0. 00278 grain per standard
cubic foot, or 0. 14 pound per hour. Because of
this low rate of particulate discharge, no air pol-
lution control devices have been found necessary
for the operations conducted in this type of fur-
nace melting iron.
to make commercial castings are usually melted
in low-frequency induction furnaces in the larger
foundries and in crucible-type, fuel-fired fur-
naces in the smaller job foundries. Electric fur-
naces, both arc and induction, are also used for
castings. Generalizing in. regard to the uses of
various rurnaces is difficult, since foundry prac-
tices are variable. A comparison of emissions
from various types of furnaces is given in Table
80.
SECONDARY BRASS- AND BRONZE-
MELTING PROCESSES
Copper when alloyed with zinz is usually termed
brass, and when alloyed with tin is termed bronze.
Other copper alloys are identified by the alloying
metals such as aluminum bronze and silicon
bronze. The true bronzes should not be con-
fused with some other common classifications of
bronzes, which are actually misnomers. For
example, "commercial bronze" is a wrought red
brass, and "manganese bronze" is a high-zinc
brass. Because of high strength, workability,
corrosion resistance, color, and other desirable
physical characteristics, the copper-base alloys
have found wide use for hardware, radiator cores,
condensers , jewelry, musical instruments , plumb-
ing fittings, electrical equipment, ship propel-
lers, and many other devices.
The remelting of nearly pure copper and bronze
does not have great interest from the standpoint
of air pollution since only small amounts of metal
are volatilized. This is due to the high boiling
points of copper and tin (above 4, 000 °F) and their
lownormal pouring temperatures of about 2, 000°
to 2,200°F. With good melting practice, total
emissions to the air should not exceed 0. 5 per-
cent of the process weight. The brasses contain-
ing 15 to 40 percent zinc, however, are poured
at temperatures near their boiling points (about
2, 200°F), and some vaporization or combustion
of desirable elements, particularly zinc, is in-
evitable. Emissions into the air may vary from
less than 0. 5 percent to 6 percent or more of the
total metal charge (St. John, 1955) and 2 to 15
percent of the zinc content through fuming (Allen
et al. , 1952), depending upon the composition of
the alloy, the type of furnace used, and the melt-
ing practice.
FURNACE TYPES
Brass and bronze shapes for working, such as
slabs and billets, are usually produced in large
ga,s-and oil-fired furnaces of the reverberatory
type. Most operators of secondary smelters also
use this type of furnace for reclaiming and re-
fining scrap metal, ordinarily casting the puri-
fied metal into rjies. Brasses and bronzes used
The Air Pollution Problem
Air contaminants emitted from brass furnaces
consist of products of combustion from the fuel,
and particulate matter in the form of dusts and
metallic fumes. The particulate matter com-
prising the dust and fume load varies according
tothefuel, alloy composition, melting tempera-
ture, type of furnace, arid many operating factors.
In addition to the ordinary solid particulate mat-
ter, such as fly ash, carbon, and mechanically
produced dust, the furnace emissions generally
contain fumes resulting from condensation and
oxidation of the more volatile elements, includ-
ing zinc, lead, and others.
As was previously mentioned, air pollution re-
sulting from the volatilization of metals during
the melting of nearly pure copper and bronze is
not too serious because of the high boiling-point
temperatures of copper, tin, nickel, aluminum,
and even lead commonly used in these alloys.
Alloys containing zinc ranging up to 7 percent can
be successfully processed \vith a minimum of
fume emission when a cohesive, inert slag cover
is used. This nominal figure is subject to some
variation depeading upon composition of alloy,
temperatures, operation procedures, and other
factors. Research is still necessary to deter-
mine the full range of effects these variables
have upon the emission rate.
Copper-base alloys containing 20 to 40 percent
zinc have low boiling points of approximately
2,100°F and melting temperatures of approxi-
mately 1, 700° to 1, 900 °F. These zinc-rich alloys
are poured at approximately 1, 900° to 2, 000°F,
which is only slightly below their boiling points.
Pure zinc melts at787°F and boils at 1,663°F.
Even within the pouring range, therefore, frac-
tions of high-zinc alloys usually boil and flash to
zinc oxide (Allen et al. , 1952). The zinc oxide
formed is submicron in size, and its escape to
the atmosphere canbe prevented only by collect-
ing the fumes and using highly efficient air pol-
lution control equipment.
Characteristics of emissions
Perhaps the best -way to understand che difficulty
of controlling metallic fumes from brass fur-
-------
Brass- and Bronze-Melting Processes
271
Table 80. DUST AND FUME DISCHARGE FROM BRASS FURNACES
Type of
iurnace
Rotary
Rotary
Rotary
Elcc ind
E ] c c ind
Elec ind
Cyl reverb
Cyl reverb
Cyl reverb
Cyl reverb
Crucible
Crucible
Crucible
Composition of alloy, %
Cu
85
76
85
60
71
71
87
77
80
80
65
60
77
Zn
5
14. 7
5
38
Z8
28
4
-
-
2
35
37
It
Pb
5
4. 7
5
2
-
-
0
18
13
10
-
1. 5
6
Sn
5
3. 4
5
-
1
1
8. 4
5
7
8
-
0. 5
3
Other
-
0. 67 Fe
-
-
-
-
0.6
-
-
-
-
1
2
Type of
control
None
None
Slag cover
None
None
None
None
None
Slag cover
None
None
None
Slag cover
Fuel
Oil
Oil
Oil
Elect
Elect
Elect
Oil
Oil
Oil
Oil
Gas
Gas
Gas
Pouring
temp, °F
No data
No data
No data
No data
No data
No data
No data
2, 100
2, 100
1, 900 to 2, 100
2, 100
1, 800
No data
Process wt,
Ib/hr
1, 104
3, 607
1, 165
1, 530
1, 600
1, 500
273
1, 267
1, 500
1, 250
470
108
500
Fume emission
Ib/hr
22. 5
25
2. 73
3. 47
0. 77
0. 54
2. 42
26. 1
a. i
10. 9
8. 67
0. 05
0. 822
nacesisto consider the physical characteristics
of these fumes. The particle sizes of zinc oxide
fumes vary from 0. 03 to 0. 3 micron. Electron
photomicrographs of these fumes are shown in
Figures 185 and 186. Lead oxide fumes, emitted
from many brass alloys, are within this same
range of particle sizes. The collection of these
very small particles requires high-efficiency
control devices. Thesemetallicfum.es also pro-
duce very opaque effluents, since particles of
0. 2- to 0. 6-micron diameter produce a maximum
scattering of light.
In copper-base alloy foundries, as much as 98
percent of the particulate matter contained in fur-
nace stack gases maybe zinc oxide and lead oxide,
depending upon the composition of the alloy. A
series of tests (Allenetal. , 1952) in Los Angeles
County indicated the zinc oxide content of fume
from representative red and yellow brass fur-
naces averaged 59 percent, while the lead oxide con-
tent averaged 15 percent. Other tests by the
same investigator s showed that the dust and fume
loading from red and yellow bras s furnaces varied
fromO. 022 to 0. 771 grain per cubic foot with an
average of 0.212 grain per cubic foot at stack
conditions.
Inhigh-lead alloys, these tests showed that lead
oxide constituted 56 percent of the particulate
matter in the exit gas. Lesser constituents of
fumes, such as tin, copper, cadmium, silicon,
and carbon, may also be present in varying
amounts, depending upon the composition of the
alloy and upon foundry practice.
Investigations prove conclusively that the most
troublesome fumes consist of particles of zinc
and lead compounds submicron in size, and that
air pollution control equipment capable of collect-
ing particulate matter from 1. 0 down to about
0. 03 micron is required. Photomicrographs of
samples taken when furnace emis sions 'were heavy
with smoke resulting from improper combustion
ormeltingof oily scrap indicated that the smoke
particles accompanying the fumes may be about
0.01 micron and smaller (Allen et al. , 1952).
Factors causing large concentrations of zinc
fumes
Four principal factors (Allen et al. , 1952) causing
relatively large concentrations of zinc fumes in
brass furnace gases are:
1. Alloy composition. The rate of loss of zinc
is approximately proportional to the zinc per-
centage in the alloy.
2. Pouring temperature. For a given percen-
tage of zinc, an increase of 100°F increases
the rate of loss of zinc about 3 times.
3. Type of furnace. Direct-fired furnaces pro-
duce larger fume concentrations than the cru-
cible type does, other conditions being con-
stant. The Los Angeles Nonferrous Found-
rymen's Committee, stated, "It is im-
probable that any open-flame furnace melting
alloys containing zinc and lead can be oper-
ated without creating excessive emissions.
It is conceded that anyone choosing to operate
that type of furnace will be required to in-
stall control equipment. "
4. Poor foundry practice. Excessive emissions
result from improper combustion, overheat-
ing of the charge, addition of zinc at maxi-
-------
272
METALLURGICAL EQUIPMENT
mum furnace temperature, flame impinge-
ment upon the metal charged, heating the
metal charged, heating the metal too fast,
and insufficient flux cover. Excessive super-
heating of the molten metal is to be avoided
for metallurgical and economic as well as
pollution control reasons. From an air pol-
lution viewpoint, the early addition of zinc is
preferable to gross additions at maximum
furnace temperatures.
In any fuel-fired furnace, the internal atmosphere
is of prime importance since there exists a con-
stantflowof combustion gases through the melt-
ing chamber, more or less in contact with the
metal. A reducing atmosphere is undesirable
from both the metallurgical and air pollution view-
points. With too little oxygen, the metal is ex-
posed to a. reducing atmosphere of'unburned fuel
and water vapor, which usually results in gassy
metal. Incomplete combustion, especially with
- 4-
Figure 185. Electron photomicrographs of fume from zinc smelter (Allen et al., 1952).
-------
Brass- and Bronze-Melting Processes
273
Aft- 4-
Figure 186. Electron photomicrographs of fume
from a yellow brass furnace (Allen et al., 1952).
oil firing, produces smoke and carbon. In one
case, a furnace was operated with a fuel mix-
ture so rich that incandescent carbon from the
fuel ignited the cloth filter bags in the baghouse
serving the furnace. To prevent these difficul-
ties, the atmosphere should be slightly oxidiz-
ing. Excess oxygen content should be greater
thanO. 1 percent; otherwise, castings will be af-
fected by gas porosity. Conversely, the excess
oxygen content must be less than 0. 5 percent to
prevent excessive metal oxidation (St. John, 1955).
The need for such close control of the internal
furnace atmosphere requires careful regulation
of the fuel and air input and frequent checking of
the combustion gases.
Crucible furnace--pit and tilt type
The indirect-fired, crucible-type furaace is used
extensively in foundries requiring small- and
medium-sized melts. The Hft-out-type crucible
is frequently employed in small furnaces. Tests
have demonstrated that, with careful practice
and use of slag covers, the crucible furnace is
capable of low-fume operation within the legal
limits for red brasses containing as much as 7
percent zinc. A slag cover does not sufficient-
ly suppress the emissions from alloys with a zinc
content in excess of 7 percent unless very low
pouring temperatures are used.
The slag cover, consisting mainly of crushed glass,
is not used as a true refining flux but as an inert,
cohesive slag of sufficient thickness to keep the
molten metal covered, [f the quantity of slag is
carefully controlled, a minimum of emissions
results from either melting or pouring. A slag
thickness of 1/4 to 3/8 inch is recommended.
Before any metal is added to the crucible, the
flux should be added so that, as melting takes
place, a cover is formed o£ sufficient thickness
to keep the molten metal divorced from the at-
mosphere.
When the crucible of molten metal has reached
the pouring floor, two holes are punched in the
slag cover on top of the metal, one through which
the metal is poured, the other to permit the en-
trance of air (Haley, 1949). Holding escaping
oxides to a minimum is possible either by using
patented attachments to hold back the slag at the
pouring sprue or using a hand-operated skimmer.
Electric furnace--low-frequency induction type
The low-frequency, indue ti on -type furnace has a
number of desirable characteristics for melting
brasses. The heating is rapid and uniform, and
the metal temperature can be accurately con-
trolled. Contamination from combustion gases
is completely eliminated. High-frequency induc-
tion furnaces are well adapted to copper- and
nickel-rich alloys but are not -widely used for
zinc-rich alloys. Low-frequency induction fur-
naces are more suitable for melting zinc-rich
alloys. During melting of clean metal, use of a
suitable flux cover over the metal prevents ex-
cessive fuming except during the back charging
and pouring phases of the heat. The usual flux
covers--borax, soda ash, and others—are de-
structive to furnace walls, but charcoal is used
-------
274
METALLURGICAL EQUIPMENT
with satisfactory results. During the test out-
lined in Table 8 1, case C, two-thirds of the total
fumes were released during the pouring and charg-
ing periods. A furnace, similar to that tested,
is shown in Figure 187.
Cupola furnace
The cupola furnace is used for reduction of cop-
per-base alloy slag and residues. The residues
charged have a recoverable metallic content of
25 to 30 percent. The balance of the unrecover-
able material consists of nonvolatile gangue,
mainly, silicates. In addition to the residues,
coke and flux are charged to the furnace. Period-
ically the recovered metal is tapped from the fur-
nace. The slag produced in the cupola is elim-
inated through a slag tap located slightly above
the metal tap.
In addition to the usual metallic fumes, the cupo-
la also discharges smoke and fly ash. Collec-
tion of these emissions is required at the cupola
stack, the charge door, and the metal tap spout.
With no control equipment, emissions of 60 to
100 percent opacity can be expected from the
charge door and stack. The opacity of the fumes
emitted from the metal tap varies from 60 to 80
percent.
The slag discharged from the cupola is rich in
zinc oxide. Although the slag leaves the furnace
at a temperature of approximately 1, 900 °F, the
zinc oxide is in solution and, at this tempera-
ture, does not volatilize to any extent. The dis-
charge slag is immediately cooled by water. The
emissions from the sla.g-tapiping operation rarely
exceed 5 percent opacity.
Hooding and Ventilation Requirements
Regardles s of the efficiency of the control device,
air pollution control is not complete unless all
the fumes generated by the furnace are captured.
Since different problems are encountered with the
various types of furnace, each will be discussed
separately.
Reverberatory furnace--open-hearth type
In a reverberatory open-hearth furnace, the prod-
ucts of combustion arid metallic fumes are nor-
mally vented directly from the furnace through
a cooling device to a baghouse. Auxiliary hoods
are required over the charge door, rabble (or
slag) door, and tap hole. These may vent to the
baghouse serving the furnace and hence cool the
hot combustion gases by dilution or may vent to
a smaller auxiliary baghouse.
Table 81. BRASS-MELTING FURNACE AND BAGHOUSE COLLECTOR DATA
Case
Furnace data
Type of furnace
Fuel used
Metal melted
Composition of metal melted, %
Copper
Zinc
Tin
Lead
Other
Melting rate, Ib/hr
Pouring temperature, °F
Slag cover thickness, in.
Slag cover material
Baghouse collector data
Volume of gases, cfm
Type of baghouse
Filter material
Filter area, ft2
Filter velocity, fpm
Inlet fume emission rate, Ib/hr
Outlet fume emission rate, Ib/hr
Collection efficiency, %
A
Crucible
Gas
Yellow brass
70.6
24.8
0. 5
3.3
0. 8
388
2, 160
1/2
Glass
9, 500
Sectional
tubular
Orion
3,836
2.47
2.55
0. 16
93.7
B
Crucible
Gas
Red brass
85. 9
3. 8
4.6
4. 4
1. 3
343
2, 350
1/2
Glass
9,700
Sectional
tubular
Orion
3,836
2. 53
1. 08
0. 04
96.2
C
Low-frequency induction
Electric
Red brass
82. 9
3. 5
4.6
8.4
0.6
1, 600
2, 300
3/4
Charcoal
1, 140
Sect: onal
tubular
Orion
400
2.85
2.2a
0.086
96. 0
^Includes pouring and charging operations.
-------
Brass- and Bronze-Melting Processes
275
Figure 187. Low-frequency induction furnace with fixed hood.
Whether the auxiliary hoods vent to the furnace
baghouse or to a separate filter, the furnace burn-
ers should be turned down or off during periods
when the furnace is opened for charging, rab-
bling, air lancing, removing slag, adding metal,
or pouring metal. Otherwise, the exhaust fan
may not have sufficient capacity to handle the
products of combustion and the additional air re-
quired to capture the fumes. Since no two of
these operations occur simultaneously, the re-
quired air volume for collection may be reduced
by the use of properly placed dampers within the
exhaust system.
If the entire furnace charge is made at the begin-
ning of the heat, the metal should be loaded in
such a way that the flame does not impinge di-
rectly upon the charge. If periodic charges are
made throughout the heat, the burners should be
turned off during charging operations. The opac-
ity of escaping fumes may vary from none to 15
percentwith the burners off and may be 60 to 70
percent with the burners ignited.
Well-designed hoods, properly located, with an
indraft velocity of 100 to 200 fpm, adequately cap-
ture the furnace emissions. If the hood is placed
too high for complete capture or is improperly
shaped and poorly fitted, higher indraft velocities
are required.
The rabble or slag door permits (1) mixing the
charge, (2) removing slag from the metal sur-
face, and (3) lancing the metal with compressed
air to eliminate iron from the metal when re-
quired by alloy specifications. Emissions from
the furnace may be of 50 to 90 percent opacity
during these operations, even with the burners
partially throttled. Again, 100 to 200 fpm in-
draft velocity is recommended for properly de-
signed hoods.
Generally, after the slag has been removed, metal,
usually zinc, must be added to bring the brass
within specifications. The furnace metal is at a
temperature well above the boiling point of zinc
and is no longer covered by the tenacious slag
-------
276
METALLURGICAL EQUIPMENT
cover. Hence, voluminous emissions of zinc
oxide result. The addition of slab zinc produces
100 percent opaque fumes in great quantity, -while
a brass addition may generate fumes of 50 per-
cent opacity. A well-designed hood is required
over the charge door or rabble door, through
which the metal is charged.
Perhaps the most critical operation from the
standpoint of air pollution occurs when the fur-
nace is tapped. Nearly continuous emissions of
90 to 100 percent opacity may be expected. Much
planning is required to design a hood that com-
pletely captures the emissions and yet permits
sufficient working room and visibility of the mol-
ten metal. Again, the burners should be turned
off or throttled as much as possible to reduce the
quantityoffum.es emitted.
The fluxes used in reverberatory furnaces nor-
mally present no air pollution problems. Gen-
erally, only nonvolatile fluxes such as borax,
soda ash, and iron oxide mill scale are used.
Reverberatory furnace — cylindrical type
Cylindrical-type reverberatory furnaces present
all the collection problems of the open-hearth
type with the additional complication of furnace
rotation. The cylindrical furnace may be ro-
tatedupto 90° for charging, slag removing, and
metal tapping. Withhoods installed infixed posi-
tion, the source of emissions may be several feet
from the hood, and thus no fumes would be col-
lected. Either a hood attached to the furnace and
venting to the control device through flexible duct-
work, or an oversized close-fitting hood covering
all possible locations of the emission source is
required. A close-fitting hood and high indraft
velocities are often necessary. For example,
an auxiliary hood over the combination charge
and slag door of a cylindrical brass furnace was
incapable of collecting all emissions, despite an
indraft velocity of 1,370 fpm. A similar hood
over the pouring spout was also inadequate, de-
spite an indraft velocity of 1, 540 fpm. Bothhoods
were improperly shaped and were located too high
above the source for adequate capture.
A cylindrical furnace rotates on its longitudinal
axis, and a tight breeching is mandatory at the
gas discharge end of the furnace. Adequate in-
draft velocity must be maintained through the
breeching connection to prevent the escape of
fumes.
The exhaust system for the cylindrical furnace, as
well as for all types of reverberatory furnaces,
must be designed to handle the products of com-
bustion at the maximum fuel rate. Any lesser
capacity results in a positive pressure within the
furnace during periods of maximum firing with
resultant emissions from all furnace openings.
Reverberatory furnace--tilting type
The tilting-type furnace differs from the rever-
beratory furnaces previously discussed in that
the exhaust stack is an integral part of the fur-
nace and rotates with the furnace during charg-
ing, skimming, and pouring. One type of tilting
furnace is charged through the stack, and skim-
ming and pouring are accomplished through a
small tap hole in the side of the furnace. Another
type has a closeable charge door, and a small
port through which the furnace gases escape.
These two furnace openings may describe a full
180° of arc during the various phases of a heat.
The wide range of position of the sources makes
complete capture of the fumes difficult. One suc-
cessful system utilizes a canopy hood, with side
panels that completely enclose the furnace. Clear-
ance for working around the furnace is provided
and a minimum indraft velocity of 125 fpm is re-
quired for this opening. This velocity provides
complete capture of the emissions unless a cross -
draft of 50 to 200 fpm prevails within the furnace
room, in which case an estimated 10 percent of
the fumes within the furnace hood escape from
beneath the hood. This condition is corrected by
suspending an asbestos curtain from the wind-
ward side of the hood to the floor.
Reverberatory furnace — rotary tilting type
The rotary tilting type of furnace not only tilts
for charging and pouring but rotates during the
melting period to improve heat transfer. Two
types are common. One is charged through the
burner end and is poxired from the exhaust port
of the furnace, opposite the burner. The other
has a side charge door at the center of the fur-
nace through which charging, slagging, and pour-
ing operations are conducted.
Because of the various movements of this type of
furnace, direct connection to the control device
is not feasible. The furnace is under positive
pressure throughout the heat, and fumes are emit-
ted through all furnace openings.
Hooding a rotary-tilting-type reverberatory fur-
nace for complete capture of fumes is difficult,
and complete collection is seldom achieved.
These furnaces are undoubtedly the most diffi-
cult type of brass furnace to control. To hood
them effectively requires a comprehensive de-
sign. The major source of emissions occurs at
the furnace discharge. Capture of fumes is ac-
complished by a hood or stack placed approxi-
-------
Brass- and Bronze-Melting Processes
277
mately 18 inches from the furnace. This clear-
ance is necessary to allow sufficient room for
tilting the furnace for pouring. A minimum in-
draft velocity of 1,750 fpm is usually required.
Although this method controls the emissions dur-
ing the melting phase, capturing the dense fumes
generated during the pour is difficult.
Hooding is sometimes installed at the burner end
of a furnace to capture emissions that may escape
from openings during melting, or particularly
during the time the furnace is tilted to pour. Be-
cause both ends of the furnace are open, a venting
action is created during the pour, causing fume
emissions to be dischargedfrom the elevated end
of the furnace. Close hooding is not practicable
because the operator must observe the conditions
within the furnace through the open ends. An
overhead canopy hood is usually installed. Fig-
ure 188 illustrates an installation in which a can-
opy hood is used to capture emissions from one
end of the furnace while, at the opposite end,
baffles have been extended from, around the stack
opening to minimize crossdrafts and aid in cap-
turing emissions from the ladle being filled from
the furnace.
Additional heavy emissions maybe expected dur-
ing charging, alloying, and slagging. High-over-
head canopy hoods are generally used. These
overheadhoods are, however, unsatisfactory un-
less they cover a large area, and a high indraft
velocity is provided.
Figure 188. Rotary-tiI ting-type reverberatory furnace venting to canopy hood and stack vent:
(top) Furnace during meltdown, (bottom) furnace during pour (Valley Brass, Inc. El Monte, Calif.).
-------
278
METALLURGICAL EQUIPMENT
The need for numerous "hoods and large air vol-
umes, with the resultant larger control device,
makes the tilting-type open-flame furnace expen-
sive to control. This type of furnace is being
gradually replaced by more easily controllable
equipment.
The following example illustrates the fundamental
design considerations of a side-draft hooa for a
rotary-tilting-type furnace.
Example 28
Given:
Rotary-tilting-type brass-melting furnace. Fuel
input, 17gal/hrU.S. Grade No. 5 fuel oil. Max-
imum temperature of products of combustion dis-
charged from furnace, 2,600°F.
Vol =
Wt =
60
(17)(8)(15.96)
60
= 468 scfm
= 36. 2 Ib/min
2. Volume of ambient air required to reduce
temperature of products of combustion from
2, 600° to 250°F:
Baghouse inlet design temperature selected,
250 °F. Ambient air temperature assumed to
• be 100°F.
Heat gained
ambient air
(Heat lost by products
| of combustion
MC At =MC At
a p a pc p pc
(M )(0. 25)(250-100) = (36. 2)(0. 27)(2, 600-250)
3-
TO BAGHOUSE
FURNACE
BURNER
REFRACTORY
37. 5 M = 23, 000
3,
M = 613 Ib/min
a
613
0. 071
!, 640 cfm at 100°F
3. Total volume of products to be vented through
hood:
Volume from furnace =
460
60 4- 460J
= 639 cfm
Volume from ambient air = (8,640)
Figure 189. Rotary-tiI ting-type brass-melting
furnace.
Problem:
Determine the design features of a side-draft hood
to vent the furnace.
Solution:
1. Volume and weight of products to be vented
from furnace:
With 10% excess air, 1 Ib of U. S. Grade No.
5 fuel oil yields 206. 6 ft3 or 15. 96 Ib of prod-
ucts of combustion. One gallon of fuel oil
weighs 8 Ib.
250 + 460\
100 + 460/
= 10, 950 cfm
Total = 11, 589 cfm at 250°F
4. Open area of hood: Design for a velocity of
2,000 ft/min. This is adequate to ensure
good pickup if the hood geometry is designed
properly.
11,589
2, 000
= 5. 78 ft
Problem note: Furnace gases should discharge
directly into center of hood opening. Position-
ing of the hood should be such that it picks up
emissions from the ladle during the furnace tilt
and pour. Sides extending to ground level may
be necessary to nullify crossdrafts. When the
-------
Brass- and Bronze-Melting Processes
279
furnace is tilted, emissions will escape from the
high side or the firing end opening. These may
be stopped by blowing a portion of the burner com-
bustion air through the furnace, which forces
emissions through the furnace discharge opening.
If this is not possible, an auxiliary hood should
be installed over the firing end of the furnace.
Crucible-type furnaces
One large-volume foundry, using tilting-type
crucible furnaces, installed an exhaust system
to control emissions during pouring. The ex-
haust system vents 14 furnaces to a baghouse.
The hooding collects all the fumes during pour-
ing without interfering with the furnace operation
in any way. The hood is equipped with a damper
that is closed when the furnace is in the normal
firing position. A linkage system opens the dam-
per when pouring begins. After the furnace is
tilted 40°, the damper is fully opened, remain-
ing there for the rest of the pour. It swings shut
automatically when the furnace is returned to the
firing position. The ductwork leading from the
hood pivots when the furnace is tilted. The en-
tire hood is fixed to the furnace with two bolts,
which permit its rapid removal for periodic re-
pairs to the furnace lining and crucible. Since
only one furnace is poured at a time and the sys-
tem operates only during the pour, only 1, 500
cfm is required to collect the fumes. Tests show
that the amount of particulate matter emitted to
the atmosphere with this system is 0. 1Z5 pound
per hour per furnace (Anonymous, 1950). This
contrasts with a loss of over 2 pounds per hour
uncontrolled.
Figure 190 shows an installation of a tilting-type
brass crucible furnace with a plenum roof-type
hood, which captures furnace emissions during
the meltdown and ladle emissions during the pour.
Emissions resulting from the pouring of molten
metal from a ladle into molds can also be con-
trolled by two other devices. The first is a fixed
pouring station that is hooded so that emissions
from the ladle and molds are captured during the
pouring (Figure 191). An installation of this type
requires a conveyorized mold line. The second
is a small hood attached to the pouring ladle and
vented to the control system through flexible duct-
work connections. Within the limits of the flexible
connection, the hood can travel with the ladle from
mold to mold as each is poured (Figure 192).
Low-frequency induction furnace
The control of the emissions from an induction
furnace is much more expensive and difficult if
oilyturnings are charged to the furnace. In ad-
Figure 190. Tilt-type crucible brass furnace with
a plenum roof-type hood.
Figure 191. Fixed-mold pouring station with
fume mold.
dition to the fumes common to brass melting,
great clouds of No. 5 Ringelmann black smoke
are generated -when the oily shavings contact the
molten heel within the furnace. Adequate hooding
enclosing the furnace is, therefore, required,
and a large volume of air is necessary to capture
the smoke and fumes. Where 900 cfm was suffi-
-------
280
METALLURGICAL EQUIPMENT
Figure 192. Pouring ladle with traveling fume
hood (Valley Brass, Inc., El Monte, Calif.).
cientto collect the pouring emissions from an in-
duction furnace using oil-free metal, 10, 000 cfm
was required throughout the heat for a similarly
sized furnace melting turnings with a 3 percent
oil content. Figure 193 shows an induction fur-
nace with an adjustable low-canopy hood that can
be positioned to cover both meltdown and pour-
ing operations. A baghouse collects the fume.
Another, smaller induction furnace is shown in
Figure 194. In addition to capturing furnace emis-
sions during meltdown, the hood captures emis-
sions during the pour into the ladle. Figure 195
shows the extent of emissions after the ladle is
removed from the hood area.
Cupola furnace
An exhaust system to control a cupola must have
sufficient capacity to remove the products of com-
bustion, collect the emissions from the metal tap
spout, and provide a minimum indraft velocity
of 250 fpm through the charge door. In addition,
side curtains may be required around the charge
doortoshield adverse crosscurrents. A canopy
hood is recommended for the metal tap spout.
The air requirement for this hood is a function
of its size and proximity to the source of emis-
sions.
Air Pollution Control Equipment
Baghouses
Baghouses with tubular filters are used to con-
trol the emissions from brass furnaces. This
type of collector is available in many useful and
effective forms. Wool, cotton, and synthetic fil-
ter media effectively separate submicron-sized
particulate matter from gases because of the
filtering action of the "mat" of particles previous-
ly collected.
The gas es leaving a reverberatory furnace may be
100° to 200°F hotter than the molten metal and
must be cooled before reaching the filter cloth.
Direct cooling, by spraying water into the hot com-
bustion gases, is not generally practiced because
(1) there is increased corrosion of the ductwork
and collection equipment, (2) the vaporized water
increases the exhaust gas volume, necessitating
a correspondingly larrger baghouse, and (3) the
temperature of the gases in the baghouse must be
kept above the dewpoint to prevent condensation
of water on the bags. The exhaust gases may be
cooled by dilution with cold air, but this increases
the size of the control equipment and the operating
costs of the exhaust system.
One cooling system employed consists of a water-
jacketed cooler followed by air-cooled radiation-
convection columns, as shown in Figure 195. The
water-jacketed section reduces the temperature
from approximately 2, 000 3 to 900 °F. The ra-
diation-convection coolers then reduce the tem-
perature to the degree required to protect the
fabric of the filter medium. Figure 196 depicts
an actual installation showing the cooling columns
and baghouse.
Treated orlon is gradually replacing glass cloth
as the most favored high-temperature fabric.
Although glass bags withstand higher tempera-
tures, the periodic shaking of the bags gradually
breaks the glass fibers and causes higher main-
tenance costs.
Probably the most critical design factor for a tu-
bular baghouse is the filtering velocity. A filter-
ing velocity of 2. 5 fpm is recommended for col-
lecting the fumes from brass furnaces with rela-
tively small concentrations of fume. Larger con-
centrations of fume require a lower filtering
velocity. A higher filtering velocity requires
more frequent shaking to maintain a pressure
drop through the baghouse-within reasonable lim-
its. Excessive bag v/ear results from frequent
shaking and higher filtering velocities. A pres-
sure drop of 2 to 5 inches of water column is
normal, and high pressure differentials across
the bags are to be avoided.
The baghouse should be completely enclosed to
protect the bags from inclement weather and
water condensation during the night when the
equipment is usually idle. The exhaust fan should
be placed downstream from the baghouse to pre-
vent blade abrasion. Moreover, problems with
fan balance due to materi£il's adhering to the
-------
Brass- and Bronze-Melting Processes
281
Figure 193. Electric induction tilting-type brass
furnace with adjustable canopy hood and baghouse
control device (American Brass Company, Paramount,
Calif.)-
Furnace data
Type, electric induction
Capacity, 3,000 Ib/hr
Electrical rating, 450 kw
Metal processed, brass
Control system data
Fan motor rating, 30 hp
Gas volume, 12,700 cfm
Baghouse type, compart-
mented, tubular
Fi Iter area, 7,896 ft2
Fi Iter medium, orlon
Shaking, automatic by
compartment
Fi Itering velocity, 1.6
Pressure drop, 1.8 to
4 in. WC
Gas stream cool ing, tem-
perature-control led
water sprays in duct
Hood indraft velocity,
560 ft/min
fpm
blades-willnot occur. Furthermore, brokenbags
are more easily detected when the exhaust system
discharges to the atmosphere through one opening.
In Table 81, the results of tests performed on
baghouses venting brass furnaces are shown. Note
that the melting rate of the induction furnace is
over 4 times that of the crucible gas-fired fur-
naces, yet the baghouse is only one-tenth as large.
Larger baghouses are necessary for crucible gas-
fired furnaces because of the heat and volume of
the products of combustion from the gas burners.
Electrical precipitators
Generally, electrical precipitators are extreme-
ly effective collectors for many substances in any
size range from 200 mesh (74 p.) to perhaps 0. 001
micron, -wet or dry, ambient or up to 1, 200°F.
This equipment has not, however, proved entire-
ly satisfactory on lead and zinc fumes. Lead ox-
ide in particular is difficult to collect because of
its relatively high resistivity, which can cause a
high potential to develop across the dust layer on the
collecting surface. This not only reduces the poten-
tial across the gas stream but may result in spark
discharge with resultant back ionization and re-
entrainment of dust. In addition, high-voltage
precipitators have not been available in small
units suitable for small nonferrous foundry use,
and the first costmay, moreover, be prohibitive.
Scrubbers
Dynamic scrubbers or mechanical washers have
proved in some applications to be effective from
-------
283
METALLURGICAL EQUIPMENT
Figure 194. Electric induction furnace with an extended hood over the pouring area: (left) Hood in place
during pouring operations, (right) ladle removed from the hood area (Western Brass Works, Los Angeles,
Calif.).
SUCTION/FAN
AND STACK
HIGH-TEMPERATURE
FUME COLLECTOR
RADIATION COOLERS WATER COOLING TOWER
AUTOMATIC DRAFT CONTROL
MAIN BRICK STACK
REVERBERATORY
FURNACE
Figure 195. Sketch of small baghouse for zinc fume (Allen et al., 1952).
-------
Brass- and Bronze-Melting Processes
283
Figure 196. Reverberatory open-hearth furnace whose slagging door and tap hole hoods vent to radiation
convection cooling columns and baghouse (H. Kramer and Company. El Segundo, Calif.).
Furnace data
Type, reverberatory
Capacity, 50 ton
Fuel input, 7,260,000 Btu/hr
Temperature of gas discharge, 2,300°F
Three baghouses in parallei
Fan motor rating, three 50 hp
Maximum gas volume, 54,400 cfm
Baghouse inlet temperature, 220°to 250°F
Fi Iter medium, or Ion
Baghouse type, compartmented, tubular
Control system data
serve three reverberatory furnaces and other smaller
furnaces.
Filter area (3 houses), 27,216 ft2
Maximum filter ratio, 2:l
Shaking, automatic by compartment
10 to 1 micron, but in addition to being ineffec-
tive in the submicron range, they have the dis-
advantage of high power consumption and mechan-
ical wear and usually require separation of the
metallic fumes and other particulate matter from
the circulating water.
A number of dynamic and static scrubbers have
been tested on brass furnaces and all have been
found unsatisfactory. The scrubbers not only
failed to reduce the particulate matter sufficient-
ly, but the opacity of the fumes escaping collec-
tion was excessive. The results of several scrub-
ber tests are summarized in Table 82. These
scrubbers have been replaced by baghouses.
Collectors depending upon centrifugal principles
alone are not adapted to brass furnace dust col-
lectionbecause of the low efficiency of these de-
vices on submicron-sized particulate matter. One
Los Angeles foundry operated a wet cyclone gas
conditioner venting to a wet entrainment separator
for recovering partially agglomerated zinc oxide
fume. The concentration of particulate matter
was relatively small, since tilting crucible fur-
naces with slag covers were used, and the device
was able to reduce the weight of the dust and fumes
emitted below the legal limits, but the number of
unagglomerated submicron-sized particles es-
caping collection was sufficient to cause period-
ic opacity violations. Consequently, this unit has
been replaced by a baghouse.
-------
284
METALLURGICAL EQUIPMENT
Table 82. EFFICIENCIES OF WET SCRUBBER CONTROL DEVICES
SERVING BRASS-MELTING FURNACES
Type
of
scrubber
Venturi with
wet cyclone
Dynamic wet
Dynamic wet
Water
rate,
gpm
7.6
20. 0
50. 0
Flue gas
volume,
scfm
860
770
1,870
Particulate
matter,
gr/scf
In
2. 71
0. 905
1.76
Out
0.704
0.367
0.598
Total dust,
Ib/hr
In
19.95
5.97
28.2
Out
7. 04
3.00
13.2
Efficiency,
%
65
50
53
SECONDARY ALUMINUM-MELTING
PROCESSES
TYPES OF PROCESS
Secondary aluminum melting is essentially the
process of remelting aluminum, but the term en-
compasses the following additional practices:
1. Fluxing. This term is applied to any pro-
cess in which materials areaddedto the melt
to aid in removal of gases, oxides, or other
impurities, but do not remain in the final
product.
2. Alloying. This term is applied to any pro-
cess in which mate rials are added to give de-
sired properties to the product and become
part of the final product.
3. Degassing. This includes any process used
to reduce or eliminate dissolved gases.
4. "Demagging. " This includes any process
used to reduce the magnesium content of the
alloy.
These terms are often used vaguely and overlap
to a great extent. For example, degassing and
demagging are usually accomplished by means
of fluxes. The use of zinc chloride and zinc flu-
oride fluxes increases the zinc content of alumi-
num alloys.
Aluminum for secondary melting comes from
three main sources:
1. Aluminum pigs. These maybe primary met-
al but mayalso be secondary aluminum pro-
duced by a large secondary smelter to meet
standard alloy specifications.
2. Foundry returns. These include gates, ris-
ers, runners, sprues, and rejected castings.
In foundries producing sand mold castings,
foundry returns may amount to 40 to 60 per-
cent of the metal poured.
3. Scrap. This category includes aluminum
contaminated with oil, grease, paint, rubber,
plastics,and other metals such as iron, mag-
nesium, zinc, and brass.
The melting of clean aluminum pigs and foundry
returns without the use of fluxes does not result
in the discharge of significant quantities of air
contaminants. The melting of aluminum scrap,
however, frequently requires air pollution con-
trol equipment to prevent the discharge of ex-
cessive air contaminants.
Crucible Furnaces
For melting small quantities of aluminum, up to
l.OOOpounds, crucible or pot-type furnaces are
used extensively. Almost all crucibles are made
of silicon carbide or similar refractory material.
Small crucibles are lifted out of the furnace and
used as ladles to pour into molds. The larger
crucibles are usually used with tilting-type fur-
naces. For die casting, molten metal is ladled
out with a small hand ladle or it can be fed auto-
matically to the die-casting machine.
Reverberator/ Furnaces
The reverberatory furnace is commonly used for
medium-and large-capacity heats. Small re-
verberatory furnaces up to approximately 3, 000
pounds' capacity may be of the tilting type. Some-
times a double-hearth construction is employed
in furnaces of 1,000 to 3,000 pounds' holding
-------
Aluminum-Melting Processes
285
capacity. This permits melting to take place in
one hearth, the second hearth serving only to
hold the molten metal at the appropriate temper-
ature. Advocates of this design stress that it re-
duces or eliminates the tendency of the metal to
absorb gas. The contention is that porosity re-
sults fromhydrogen gas, •which is liberated from
moisture trapped below the surface of molten
aluminum. The use of a double hearth permits
moisture to be driven off before the metal melts
and runs to the holding hearth. Sometimes the
melting hearth is also used as a sweat furnace to
separate the aluminum from contaminants such
as brass and steel. The use of double-hearth
furnaces for the larger capacity heats is not com-
mon.
A charging well is frequently used on aluminum
reverberatoryfurnaces. Figure 197 shows a 20-
ton reverberatory furnace with a charging well.
The well permits chips and other small aluminum
scrap to be introduced and immersed below the
liquid level. Chips and small scrap have an un-
usually high surface area-to-volume relation-
ship, and oxidation must be minimized. Large
quantities of flux are also added and stirred in to
dissolve the oxide coating and aid in the removal
of dirt and other impurities. The flux causes
the oxides and other impurities to rise to the sur-
face in the form of a dross that can be skimmed
off easily.
Reverberatory furnaces of 20- to 50-ton holding
capacity are common. Usually one heat is pro-
duced in a 24-hour period; however, the time
per heat in different shops varies from 4 hours
to as much as 72 hours. This type of furnace is
commonly used to melt a variety of scrap. The
materials charged, method of charging, size and
design of the furn'ace, heat input, and fluxing,
refining, and alloying procedures all have some
influence on the time required to complete a heat.
After the charge is completely melted, alloying
ingredients are added to adjust the composition
to required specifications. Large quantities of
fluxes are added when scrap of small size and low
grade is melted. The flux in some cases may
amount to as much as 30 percent of the weight of
scrap charged.
Fuel-fired furnaces used for aluminum melting
are extremely inefficient. Approximately 50 per-
cent of the gross heating value in the fuel is un-
available in the products of combustion. Radia-
tion and convection losses are high since little or
no insulation is used. Many small crucible fur-
naces probably do not achieve more than 5 per-
cent overall efficiency and some may not exceed
2 to 3 percent (Anderson, 1925). At the other
extreme a properly designed and operated fur-
nace may be able to use as much as 20 percent
of the gross heat in the fuel. Most furnaces can
be assumed to operate with efficiencies of 5 to
15 percent. This may become an important fac-
tor when air pollution control equipment must be
provided to handle the products of combustion.
Fortunately, this is seldom necessary. Controls,
if provided, are usually required only during the
degassing or demagging operations when the burn-
ers are off. Another possibility is to add fluxes
and scrap only to a charging •well that is vented
to control equipment.
Electrically Heated Furnaces
Electric induction furnaces are becoming increas-
ingly common for both melting and holding alumi-
num in spite of higher installation and operating
costs. Some of the advantages they offer over
other furnaces are higher efficiency, closer tem-
perature control, no contaminants from products
of combustion, less oxidation, and improved ho-
mogeneity of metal. Electric resistance heating
is sometimes used for holding but rarely for melt-
ing furnaces. Most electric furnaces for alumi-
num melting are relatively small though some
holding furnaces have capacities up to about 15
tons.
Charging Practices
Small crucible furnaces are usually charged by
hand with pigs and foundry returns . Many rever-
beratory furnaces are als o charged with the same
type of materials, but mechanical means are used
because of the larger quantity of materials in-
volved.
Fuel-Fired Furnaces
Both gas- and oil-fired furnaces are common,
though gas-firedfurnaces are usually preferred.
Frequently, combination burners are used so that
gas may be burned when available, with oil sub-
stituted during periods of gas curtailment.
When chips and light scrap are melted, it is a
common practice to melt some heavier scrap or
pigs first to form a molten "heel. '' The light
scrap is then added and immediately immersed
below the surface of the molten metal so that
further oxidation is prevented. The heel may
consist of 5, 000 to 20,000 pounds, depending up-
on the size of the furnace.
-------
286
METALLURGICAL EQUIPMENT
Figure 197. A 20-ton aluminum-melting reverbatory
furnace with charging well hood (Aaron Ferrer & Sons
Inc., Los Angeles, CaIi f.).
Pouring Practices
Tilting-type crucible furnaces are used when the
crucible is toolarge to be handled easily. These
furnaces are poured into smaller capacity ladles
for transfer to the molds. Larger reverberatory
furnaces are either tapped from a tap hole or si-
phoned into a ladle. Ladles vary up to 3 or 4 tons
capacity in some cases. Sometimes the ladles
are equipped with covers with electric resistance
heaters to prevent loss of temperature when the
ladle is not to be poured immediately or when the
pouring requires too long a time. Pouring mol-
ten aluminum does not usually result in the dis-
charge of air contaminants in significant quanti-
ties.
Fluxing
The objectives of fluxing generally fall into four
main categories:
1. Cover fluxes. These fluxes are used to cov-
er the surface of the metal to prevent further
oxidation and are usually liquid at the melting
point of aluminum. Some of these are also
effective in preventing gas absorption.
2. Solvent fluxes . These fluxes generally cause
the impurities and oxides to float on top of
the melt in the form of a dross that can be
skimmed off easily.
-------
Aluminum-Melting Processes
287
3. Degassing fluxes. These fluxes are used to
purge the melt of dissolved gases. The dis-
solved gas is assumed to be hydrogen, but
other gases are also highly soluble in alumi-
num. The solubility of gases in molten alu-
minum increases with temperature. The gas-
es most soluble in molten aluminum, in de-
creasing order of solubility, are hydrogen,
methane, carbon dioxide, sulfur dioxide, oxy-
gen, air, and carbon monoxide. The solu-
to ' *
bility of hydrogen is 6 or 7 times as great as
that of methane and over 10 times that of car-
bon dioxide. Elimination of hydrogen gas in
aluminum is a major problem.
4. Magnesium-reducing fluxes. These fluxes
are used to reduce the magnesium content
of the alloy (known as demagging). During
World War II it became necessary to recov-
er large quantities of aluminum scrap, much
of which had a magnesium content too high
for the intended use. It was found that the
magnesium could be selectively removed by
the use of appropriate fluxes.
The quantity and type of fluxing depend upon the
the type of furnace, the materials being melted,
and the specifications of the final product. A few
operators melting only pigs and returns find flux-
ing unnecessary. At the other extreme are large
secondary smelters that process very low-grade
scrap and sometimes use fluxes amounting to as
much as one-third of the weight of the aluminum
scrap charged. About 10 percent by weight is an
average figure for the amount of flux used for
medium- to low-grade scrap.
Fluxes for degassing or demagging may be either
solids or gases. The gaseous types are usually
preferred because they are easier to use, and the
rate of application is simpler to control. Some of
these, for example chlorine, may be used for
either degas sing or demagging, depending upon the
metal temperature. In general, any flux that is
effective in removing magnesium also removes
gas inclusions.
Cover fluxes
Cover fluxes are used to protect the metal from
contact with air and thereby prevent oxidation.
Most of these fluxes use sodium chloride as one
of the ingredients (Anderson, 1931). Various
proportions of sodium chloride are frequently
used with calcium chloride and calcium fluoride.
Sometimes cryolite or cryolite with aluminum
fluoride is added to dissolve oxides. Borax has
alsobeenused alone and in combination with so-
dium chloride.
Solvent fluxes
Solvent fluxes usually form a gas or vapor at the
temperature of the melt. Their action is largely
physical. The resulting agitation causes the ox-
ides and dirt to rise to the top of the molten metal
where they can be skimmed off. Included in this
group are aluminum chloride, ammonium chlo-
ride, and zinc chloride. Zinc chloride increases
the zinc content of the alloy probably according
to the equation
3 Z Cl,
n 2
+ 2 Al-
3 Z
+ 2 A1C1 (100)
Aluminum chloride, which is formed in this re-
action, is a vapor at temperatures above 352°F.
It bubbles out of the melt, forming a dense white
fume as it condenses in the atmosphere.
So-called chemical fluxes are solvents for alu-
minum oxide. Cryolite, other fluorides, or borax
is used for this purpose. Part of the action of
the fluorides is thought to be due to the libera-
tion of fluorine, •which attacks silicates and dirt.
Some chlorides are also used extensively, butthei;
action is not understood.
Degassing fluxes
There are many methods of removing dissolved
gas from molten aluminum, some of which do not
require the addition of a flux. Among the non-
flux methods are the use of vibration, high vac-
uum, and solidification with remelting. None is
as effective as the use of an active agent such as
chlorine gas. Helium, argon, and nitrogen gases
have also be en used successfully. Solid materials
that have been used include many metallic chlo-
rides. Some think that their action is physical
rather than chemical and that one gas is as good
as another. For this reason, nitrogen has been
used extensively. Nitrogen is not toxic, and vir-
tually no visible air contaminants are released
when it is used. In addition, it does not coarsen
the grain or remove sodium or magnesium from
the melt. The main objection to the use of nitro-
gen is that commercial nitrogen is usually con-
taminated with oxygen and water vapor (Eastwood,
1946).
Magnesium-reducing fluxes
The use of fluxes to reduce the magnesium con-
tent of aluminum alloys is a relatively new pro-
cedure. Certain fluxes have long been known to
tend to reduce the percent of magnesium in the
alloy, but this process did not become common-
place until the advent of World War II. Several
fluxes maybe used for this purpose. Aluminum
-------
288
METALLURGICAL, EQUIPMENT
fluoride and chlorine gas are perhaps the most
commonly used. The temperature of the melt
must be significantly greater in demagging than
in degassing, usually between 1, 400 ° and 1,500°F.
As much as 1 ton of aluminum fluoride is com-
monly used in reverberatory furnaces of 40- to
50-ton capacity. The aluminum fluoride is usu-
ally added to the molten metal •with smaller quan-
tities of other fluxes such as sodium chloride,
potassium chloride, and cryolite, and the entire
melt is stirred vigorously. Magnesium fluoride
is formed, which can then be skimmed off. Large
quantities of air contaminants are discharged
from this process.
Chlorine gas for this purpose is easier to regu-
late, but extra precautions must be taken because
of the extreme toxicity of this material. The
chlorine is fed under pressure through the tubes
or lances to the bottom of the melt and permitted
to bubble up through the molten aluminum. Fig-
ure 198 (left) shows a ladle of aluminum before
the lances are lowered into the metal. Figure
198 (right) shows the hood in position.
THE AIR POLLUTION PROBLEM
Frequently, a large part of the material charged
to a reverberatory furnace is low-grade scrap and
chips. Paint, dirt, oil, grease, and other con-
taminants from this scrap cause large quantities
of smoke and fumes to be discharged. Even if
the scrap is clean, large surface-to-volume ratios
require the use of more fluxes, -which can cause
serious air pollution problems.
In a study of the extent of visible emissions dis-
charged from degassing aluminum with chlorine
gas, the major parameters were found to be metal
temperature, chlorine flow rate, and magnesium
content of the alloy. Other factors affecting the
emissions to a lesser degree are the depth at
which the chlorine is released and the thickness
and composition of the dross on the metal surface.
Other factors remaining constant, the opacity of
the emissions at any time is an inverse function
of the percent magnesium in the metal at that
time.
When the magnesium content is reduced, either
by combining with chlorine to form magnesium
chloride (MgCl£) or by using an alloy containing
Figure 198. Ladle of molten aluminum with (left) lances in the raised position, and (right) hood in place and
lances lowered into aluminum.
-------
Aluminum-Melting Processes
289
less magnesium, a greater fraction of the chlo-
rine combines with the aluminum to form alu-
minum chloride (AlClj). The magnesium chlo-
ride melts at about 1, 312°F, so that it is a liquid
or solid at normal temperatures for this opera-
tion (about 1, 300° to 1, 350°F) and thus does not
contribute significantly to the emissions. A very
small amount may sometimes be released into
the atmosphere as a result of mechanical entrain-
ment. The aluminum chloride, on the other hand,
sublimes at about 352°F, so that it is a vapor at
the temperature of molten aluminum. As the
vapors cool in the atmosphere, submicron fumes
are formed, which have very great opacity in pro-
portion to the weight of material involved.
Chlorine has a much greater affinity for magne-
sium than it has for aluminum. This is shown by
the fact that alloys containing more than about
0. 5 percent magnesium (and 90 to 97 percent alu-
minum) usually produce only a moderate quantity
of fume in degassing with chlorine, while alloys
with more than about 0. 75 percent magnesium do
not usually produce a significant quantity of fume.
In alloys with greater magnesium content, not
only is less aluminum chloride formed, but also
a thick layer of dross (largely magnesium chlo-
ride) is built up on the surface, which further
suppresses the emission of fumes. Aluminum
chloride also reacts with magnesium to form mag-
nesium chloride and aluminum. The dross in-
creases the opportunities for this latter reaction.
When chlorine is used for demagging, it is added
so rapidly that large quantities of both aluminum
chloride and magnesium chloride are formed, the
molten bath is vigorously agitated, and not all of
the chlorine reacts 'withthe metals. As a result,
a large quantity of aluminum chloride is dis-
charged along with chlorine gas and some en-
trained magnesium, chloride. The aluminum chlo-
ride is extremely hygroscopic and absorbs mois-
ture from the air, •with which it reacts to form
hydrogen chloride. These air contaminants are
toxic, corrosive, and irritating.
of the fume from chlorinating aluminum to degas
revealed that 100 percent of the fume was smaller
than 2 microns and 90 to 95 percent smaller than
1 micron. Mean particle size appeared under a
microscope to be about 0. 7 micron.
HOODING AND VENTILATION REQUIREMENTS
When no charging well is provided, or when flux-
ing is done inside the furnace, or when dirty scrap
is charged directly into the furnace, then venting
the furnace may be necessary. In some cases,
the products of combustion must be vented to the
air pollution control equipment. The volume to
be vented to the collector, and the determination
of temperature may be found similarly to metal-
lurgical furnace calculation procedures described
elsewhere in this manual.
A canopy hood (as previously shown in Figure 197)
is usually used for capturing the emissions from
the charging well of an aluminum reverberatory
furnace. Calculation of the quantity of air re-
quired can be accomplished as shown in the fol-
lowing example.
Example 29
Given:
Metal surface, 2 ft 3 in. x 11 ft 3 in.
Temperature of molten metal, 1,350°F.
Hood opening dimensions, 3 ft 9 in. x 13 ft 9 in.
Height of hood face above metal surface, 2 ft 6 in.
Ambient air temperature, 80 °F.
Problem:
Determine the volume of air that must be vented
from a low-canopy hood over the charging well
of an aluminum-melting reverberatory furnace
to ensure complete capture of the air contami-
nants.
Particle Size of Fumes From Fluxing
One study (McCabe, 1952) found that the major
constituent in the fume from salt-cryolite flux-
ing in a furnace was sodium chloride with con-
siderable smaller quantities of compounds of alu-
minum and magnesium. Electron photomicro-
graphs of thermal precipitator samples indicated
that the particles of fume were all under 2 mi-
crons, most of thembeing 0. 1 micron. Thefumes
were somewhat corrosive when dry and, when
collected wet, formed a highly corrosive sludge
that tended to set up and harden if allowed to stand
for any appreciable time. Another study made
Solution:
As discussed in Chapter 3, the following equation
gives the total ventilation rate for low-canopy
hoods:
q = 5.4(A)(m
where
q =
A =
total ventilation rate required, cfm
2
area of the hood face, ft
-------
290
METALLURGICAL EQUIPMENT
m = the width of the hot metal surface at
the charging •well, ft
At = the difference in temperature between
the hot surface and the ambient air, °F.
q = (5.4)(3. 75)(13.75)(2. 25) (1,350-80) 2
= 7, 170 cfm
Problemnote: The volume calculated here is the
minimum ventilation required just to accommo-
date the rising column of air due to the thermal
drive. An additional allowance must be made to
take care of drafts. If volatile fluxes are used,
the volume of fumes generated must also be ac-
commodated. Inmost cases an allowance of about
25 percent additional volume is adequate to en-
sure complete pickup. The exhaust system should
therefore be designed to vent about 9, 000 cfm.
Although the gases vented from the charging well
are hot, sufficient air is drawn into the hood to
preclude any danger that the hot gas will damage
the exhaust system. The temperature of the
mixed gas stream is calculated in example 30.
Example 30
Given:
The furnace with charging "well and canopy hood
venting 9, 000 cfm as shown in Example 29.
Problem:
Determine the temperature of the air entering
the hood.
Solution:
1. Determine the heat transferred from the hot
metal surface to the air by natural convec-
tion:
From Chapter 3, H1 =
h A At
c s
60
where
H' =
heat transferred from hot metal sur-
face to the air by natural convection,
B tu / min
= coefficient of heat transfer from hori-
zontal plates by natural convection,
Btu/hr/ft2/°F
A = area of hot metal surface, ft
At = temperature difference between hot
metal surface and ambient air, °F.
By using hc = 0.38 (At)0'25 and substituting this
quantity into the equation,
H
0.38 (Ag)(At)
60
1. 25
H
(0. 38)(2. 25)(11. 25)(1,350-80)
1. 25
60
= 1,210 Btu/min
2. Solve for temperature of the air entering the
hood (assume specific volume of air = 13.8
ft3/lb):
q = We At
P
where c = specific heat of air at constant
pressure.
(1.210)(L3.8)
" (9, 000)(0.24)
Temperature of air entering the hood = 80
+ 7. 7 = 87.7°F.
The actual temperature of the air entering the
hood will be slightly higher than the value cal-
culated here, owing to radiation from the molten
metal surface, and radiation and convection from
the hood and the furnace. In some cases, when
the burners are operated at maximum capacity,
there may be a positive pressure in the furnace.
If the design of the furnace permits some of the
products of combustion to be vented into the hood,
the actual temperature may be substantially high-
er than shown here. This situation would also
require venting a greater volume to ensure cap-
turing the emissions.
AIR POLLUTION CONTROL EQUIPMENT
The emissions from aluminum, fluxing may con-
sist of hydrogen fluoride, hydrogen chloride, and
chlorine in a gaseous state, and aluminum chlo-
ride, magnesium chloride, aluminum fluoride,
magnesium fluoride, aluminum oxide, magne-
sium oxide, zinc chloride, zinc oxide, calcium
fluoride, calcium chloride, a.nd sodium chloride
in the solid state. Not all -will be present at one
time, and many other, minor contaminants may
be emitted in a specific case. Because of the
widely divergent properties of these various air
contaminants, the problem of control is compli-
cated.
GPO 8O6—614—1 I
-------
Aluminum-Melting Processes
291
Some type of scrubber is required to remove the
soluble gaseous fraction of the effluent, and either
abaghouse or an electricalprecipitator is needed
to control the solids. In order to obtain adequate
collection efficiency, the use of high-efficiency
scrubbers, with a caustic solution as the scrub-
bing medium, has been found necessary. This
is illustrated in Table 83, which shows typical
test data on collection efficiency for both ordi-
nary and high-efficiency scrubbers.
Table 83. SCRUBBER COLLECTION
EFFICIENCY FOR EMISSIONS FROM
CHLORINATING ALUMINUM
Scrubber collection efficiencies, %a
Contain in ants
HCL
CL2
Participates
Slot scrubber
Water
90 to 95
30 to 50
30 to 50
10% caustic
solution
95 to 99
50 to 60
50 to 60
Packed-column scrubber
Water
95 to 98
75 to 85
70 to 80
10% caustic
solution
99 to 100
90 to 95
80 to 90
Collection efficiency depends mainly upon scrubbing ratio
(gal per 1,000 ft }, velocity of gas in scrubber, and con-
tact time and to a lesser extent on other aspects of the
design. These values are typical efficiencies obtained by
actual tests but do not reflect the entire range of results.
Table 84 summarizes the results of a series of
200 tests made of control efficiencies of nine de-
vices by a major producer of aluminum (Jenny,
1951). These results represent the average range
of efficiencies for a number of tests but are not
necessarily the maximum or minimum values ob-
tained. In spite of the high efficiencies obtained
•with some of these devices, reducing the emis-
sions sufficiently to eliminate a visible plume
was very difficult. For the dry ultrasonic unit,
the opacity of the emissions exceeded 40 percent
when the outlet grain loading was greater than
0. 25 grain per cubic foot. The efficiency of this
unit varied widely with the inlet grain loading and
Table 84. AVERAGE COLLECTION
EFFICIENCY OBTAINED BY VARIOUS
DEVICES ON EMISSIONS FROM
CHLORINATING ALUMINUM (Jenny, 1951)
Type of device
Horizontal multipass wet cyclone
Single-pass wet dynamic collector
Packed-column water scrubber with
limestone packing
Ultrasonic agglomerator followed by
a multitube dry cyclone
Electrical precipitator
Efficiency, %
65 to 75
70 to 80
75 to 85
85 to 98
90 to 99
retention time, the efficiency increasing with in-
creasing values of either or both of these varia-
bles. Other tests by the same company on col-
lectors of a wet type revealed that the opacity
exceeded 40 percent periodically, even when the
average grain loading at the vent was as low as
0. 002 grain per cubic foot.
Figures 198, 199, and 200 show parts of a single
installation of air pollution control equipment
for the control of emissions from chlorinating
aluminum. One of the three stations where chlo-
rinating is performed is shown in Figure 198.
Note that the hooding closely encloses the source
so that a minimum volume of air is required to
attain 100 percent pickup of air contaminants. The
fumes are scrubbed in the packed-column scrubbers
showninFigure 199. Tnis system was designed
touse two of the three scrubbers in parallel, with
the third as a standby. The scrubbing medium
is a 10 percent caustic solution. After the scrub-
bing, the effluent is vented to a five-compartment
baghouse with a fully automatic shaking mechan-
ism to remove residual particulate matter. The
baghouse contains a total of 300 orlon bags with
a net filtering area of 12, 000 square feet. In ad-
dition to the fumes from chlorine fluxing, which
are vented through the scrubbers, two aluminum
dross-processing barrels (Figure 200) are vented
directlyto the baghouse. The total volume han-
dled by the baghouse is about 30, 000 cfm, of which
Figure 199. High-efficiency packed-column water scrubbers
used with a baghouse for control of emissions from chlorine
fluxing and dross processing.
-------
292
METALLURGICAL EQUIPMENT
approximately 6, 000 cfm is from the three chlo-
rine fluxing stations and the balance from the two
dross barrel hoods. The beneficial effect of the
bag precoating provided by the aluminum oxide
dust vented from the dross-processing barrels
permits a much higher filtering velocity than
would be advisable if only the fluxing stations
were being served by the baghouse.
Tests of the scrubber performance have shown
that virtually all the hydrogen chloride and more
than 90 percent of the chlorine are removed by
the caustic scrubbing solution. Since the efficien-
cy of aluminum chloride removal averages in ex-
cess of 80 percent, the loading of hygroscopic
and corrosive materials to the baghouse is rela-
tively light. The aluminum oxide dust from the
dross barrels acts as a. filter cake, which im-
proves the collection efficiency of the aluminum
chloride fume while simultaneously reducing or
eliminating the difficulties usually associated
with collecting hygroscopic materials. All ex-
posed metal parts are coated with polyvinyl chlo-
ride or other appropriate protective coatings.
The first year of operation indicates that no seri-
ous operational or maintenance problems -will de-
velop. This installation replaced an electrical
precipitator that was found extremely difficult
and expensive to maintain because of corrosion.
An electrical precipitator thathas been used suc-
cessfully to control the emissions from fluxing
aluminum is illustrated inFigure 201. At present
the trend in control equipment for aluminum-flux-
ing emissions appears to be Etway from electrical
precipitators and toward the scrubber-baghouse
combination.
Figure 200. Two aluminum dross-processing stations, one shown with hood door raised.
-------
Zinc-Melting Processes
293
Figure 201. Concrete shell-type electrical precipitator used for controlling emissions from fluoride fluxing
aluminum metal. The reverberatory furnace is shown in the left portion of the photograph (Apex Smelting Co
Long Beach, Calif.).
SECONDARY ZINC-MELTING PROCESSES
Zinc is melted in crucible, pot, kettle, rever-
beratory, or electric-induction furnaces for use
in alloying, casting, and galvanizing and is re-
claimed from higher melting point metals in sweat
furnaces. Secondary refining of zinc is conduc-
ted in retort furnaces, which can also be used to
manufacture zinc oxide by vaporizing and burn-
ing zinc in air. All these operations will be dis-
cussed in this section except the reclaiming of
zinc from other metals by use of a sweat furnace.
Information on this subject can be found in a fol-
lowing section entitled, "Metal Separation Pro-
cesses. "
ZINC MELTING
The melting operation is essentially the same in
all the different types of furnaces. In all but the
low-frequency induction furnace, solid metal can
be melted without the use of a molten heel. Orrce
a furnace is started, however, a molten heel is
generally retained after each tap for the begin-
ning of the next heat.
Zinc to be melted may be in the form of ingots,
reject castings, flashing, or scrap. Ingots, re-
jects, and heavy scrap are generally melted first
toprovide a molten bath to which light scrap and
flashing are added. After sufficient metal has
been melted, it is heated to the desired pouring
temperature, which may vary from 800° to
1, 100°F. Before the pouring, aflux is added and
the batch agitated to separate the dross accumu-
lated during the melting operation. Dross is
formed by the impurities charged with the metal
and from oxidation during the melting and heating
cycles. The flux tends to float any partially sub-
merged dross and conditions it so that it can be
skimmed from the surface. When only clean in-
got is melted, very little, if any, fluxing is nec-
essary. On the other hand, if dirty scrap is
melted, large amounts of fluxes are needed. Af-
ter the skimming, the melt is ready for pouring
-------
294
METALLURGICAL EQUIPMENT
into molds or ladles. No fluxing or special pro-
cedures are employed while the zinc is being
poured.
The Air Pollution Problem
The discharge of air contaminants from melting
furnaces is generally caused by excessive tem-
peratures and by the melting of metal contami-
nated with organic material. Fluxing can also
create excessive emissions, butfluxesare avail-
able that clean the metal without fuming.
Probably the first visible discharge noted from
a furnace is from organic material. Before the
melt is hot enough to vaporize any zinc, accom-
panying organic material is either partially ox-
idized or vaporized, causing smoke or oily mists
to be discharged. This portion of the emissions
can be controlled either by removing the organic
material before the charging to the furnace or by
completely burning the effluent in a suitable in-
cinerator or afterburner.
Normally, zinc is sufficiently fluid for pouring
attemperatures below 1, 100 "F. At that temper-
ature, its vapor pressure is 15.2 millimeters of
mercury, low enough that the amount of fumes
formed cannot be seen. If the metal is heated
above 1, 100°F, excessive vaporization can occur
and the resulting fumes need to be controlled with
an air pollution control device. Zinc can vapor-
ize and condense as metallic zinc if existing tem-
peratures and atmospheric conditions do not pro-
mote oxidation. Finely divided zinc so formed
is a definite fire hazard, and fires Have occurred
in baghouses collecting this material.
Many fluxes now in use do not tume, and air con-
taminants are not discharged. In some cases,
however, a specific fuming flux may be needed,
in which case a baghouse is required to collect
the emissions. An example of a fuming flux is
ammonium chloride, which, when heated to the
temperature of molten zinc, decomposes into
ammonia and hydrogen chloride gases. As the
gases rise into the atmosphere above the molten
metal, they recombine, forming a fume consisting
of very small particles of ammonium chloride.
Provided the temperature of the melt does not
exceed 1, 100°F, there should be no appreciable
amounts of air contaminants discharged when the
zinc is poured into molds. Some molds, how-
ever, especially in die casting, are coated with
mold release compounds containing oils or other
volatile material. The heat from the rnetal va-
porizes the oils, creating air contaminants. Re-
cently mold release compounds have been de-
veloped that do not contain oils, and this source
of air pollution is thereby eliminated.
ZINC VAPORIZATION
Retort furnaces are used for operations involving
the vaporization of zinc including (1) reclaiming
zinc from alloys, (2) refining by distillation,
(3) recQvering zinc from its oxide, (4) manufac-
turing zinc oxide, and (5) manufacturing pow-
dered zinc.
Three basic types of retort furnaces are used in
Los Angeles County: (1) Belgian retorts, (2) dis-
tillation retorts (sometimes called bottle retorts),
and (3) muffle furnaces. Belgian retorts are
used to reduce zinc oxide to metallic zinc. Dis-
tillation retorts, used for batch distillations , re-
claim zinc from alloys, refine zinc, make pow-
dered zinc, and make zinc oxide. Muffle fur-
naces, used for continuous distillation, reclaim
zinc from alloys, refine zinc, and make zinc ox-
ide.
Although zinc boils at 1,665°F, most retort fur-
naces are operated at temperatures ranging from
1,800° to2,280"F. Zinc vapor burns spontane-
ously in air; therefore, air must be excluded from
the retort and condenser when metallic zinc is
the desired product. Condensers are designed,
either for rapid cooling of the zinc vapors to a
temperature below the melting point to produce
powdered zinc, or for slower cooling to a tem-
perature above the melting point to produce liq-
uid zinc. When the desired product is zinc ox-
ide, the condenser is bypassed and the vapor is
discharged into a stream of air where spontane-
ous combustion converts the zinc to zinc oxide.
Excess air is used, not only to ensure sufficient
oxygen for the combustio>n, but also to cool the
products of combustion and convey the oxide to a
suitable collector.
REDUCTION RETORT FURNACES
Reduction in Belgian Retorts
The Belgian retort furnace is one of several hori-
zortal retort furnaces that have been for many
years the most common device for the reduction
of zinc. Although the horizontal retort process
is now being replaced by other methods capable
of handling larger volumes of metal per retort
and by the electrolytic process for the reduction
of zinc ore, only Belgian retorts are used in the
Los Angeles area. In this area, zinc ores are
not reduced; the reduction process is used to re-
claim zinc from the dross formed in zinc-melt-
ing operations, the zinc oxide collected by air
pollution control systems serving zinc alloy-melt--
ing operations, and the contaminated zinc oxide
from zinc oxide plants.
A typical Belgian retort (Figure 202) is about 8
inches in internal diameter and from 48 to 60 in-
ches long. One end is closed and a conical shaped
-------
Zinc-Melting Processes
295
FRONT KfALL
OF FURNACE
GROUT JOINT
CONDENSED METAL
VAPORS
FLAME FROM
COMBUSTIBLE GASES
METALLIC OXIDE CHARGE
WITH REDUCING MATERIALS
BURNER PORT
Figure 202. Diagram showing one bank of a Belgian retort furnace.
clay condenser from 18 to 24 inches long is at-
tached to the open end. The retorts are arranged
in banks with rows four to seven high and as many
retorts in a row as are needed to obtain the de-
sired production. The retorts are generally gas
fired.
The retorts are charged with s. mixture of zinc
oxide and powdered coke. Since these materials
are powdered, water is added to facilitate charg-
ing and allow the mixture to be packed tightly into
the retort. From three to four time s more carbon
is used than is needed for the reduction reaction.
After the charging, the condensers are replaced
and their mouths stuffed with a porous material.
A small hole is left through the stuffing to allow
moisture and unwanted volatile materials to es-
cape. About 3 hours are needed to expel all the
undesirable volatile materials from the retort.
About 6 hours after charging is completed, zinc
vapors appear. The charge in the retort is brought
up to 1, 832° to 2, 012°F for about 8 hours, af-
ter which it may rise slowly to a maximum of
2,280°F. The temperature 011 the outside of
the retorts ranges from 2,375° to 2,550°F.
The condensers are operated at from 780° to
1,020°F, a temperature range above the melting
point of zinc but where the vapor pressure is so
low that a minimum of zinc vapor is lost.
The reduction reaction of zinc oxide can be sum-
marized by the reaction:
Very little, if any, zinc oxide is, however, ac-
tually reduced by the solid carbon in the retort,
A series of reactions results in an atmosphere
rich in carbon monoxide, which does the actual
reducing. The reactions are reversible, but by
the use of an excess of carbon, they are forced
toward the right. The reactions probably get
started by the oxidation of a small portion of the
coke by the oxygen in the residual air in the re-
tort. The oxygen is quickly used, but the carbon
dioxide formed reacts with the carbon to form
carbon monoxide according to the equation:
CO +
C = 2CO
(102)
The carbon monoxide in turn reacts with zinc ox-
ide to produce zinc and carbon dioxide:
CO + ZnO = Zn
CO
2
(103)
ZnO + C = Zn + CO
(101)
Carbon monoxide is regenerated by use of equa-
tion 102, and the reduction ol the zinc oxide pro-
ceeds.
About 8 hours after the first zinc begins to be
discharged, the heat needed to maintain produc-
tion begins to increase and the amount of zinc
produced begins to decrease. Although zinc can
still be produced, the amount of heat absorbed by
the reduction reaction decreases and the tempera-
ture of the retort and its contents increases. Care
must be taken not to damage the retort or fuse
its charge. As a result, a 24-hour cycle has been
found to be an economical operation. The zinc
values still in the spent charge are recovered by
recycling with the fresh charges. A single-pass
-------
296
METALLURGICAL EQUIPMENT
recovery yields 65 to 70 percent of the zinc
charged, but, by recycling, an overall recovery
of 95 percent may be obtained.
The Air Pollution Problem
The air contaminants emitted vary in composi-
tion and concentration during the operating cycle
of Belgian retorts. During charging operation
very low concentrations are emitted. The feed
is moist and, therefore, not dusty. As the re-
torts are heated, steam is emitted. After zinc
begins to form, both carbon monoxide and zinc
vapors are discharged. These emissions burn
to form gaseous carbon dioxide and solid zinc
oxide. During the heating cycley- zinc is poured
from the condensers about three times at 6- to
7-hour intervals. The amount of zinc vapors dis-
charged increases during the tapping operation.
Before the spent charge is removed from the re-
torts, the temperature of the retorts is lowered,
but zinc fumes and dust from the spent charge
are discharged to the atmosphere.
Hooding and Ventilation Requirements
Air contaminants are discharged from each re-
tort. In one installation, a furnace has 240 re-
torts arranged in five horizontal rows with 48 re-
torts per row. The face of the furnace measures
70 fe«t long by 8 feet high; therefore, the air con-
taminants are discharged from 240 separate open-
ings and over an area of 560 square feet. A hood
2 feet wide by 70 feet long positioned immediate-
ly above the front of the furnace is used to collect
the air contaminants. The hood indraft is 175
fpm.
DISTILLATION RETORT FURNACES
The distillation retort furnace (Figure 203) con-
sists of a pear-shaped, gra.phite retort, which
may be 5 feet long by 2 feet in diameter at the
closed end by 1-1/2 feet in diameter at the open
end and 3 feet in diameter at its widest cross-
section. Normally, the retort is encased in a
brick furnace with only the open end protruding
and it is heated externally -with gas- or oil-fired
burners. The retorts are charged with molten,
impure zinc through the open end, and a condens-
er is attached to the opening to receive and con-
dense the zinc vapors. After the distillation is
completed, the condenser is moved away, the
residue is removed from the retort, and a new
batch is started.
SPEISE HOLE
Figure 203. Diagram of a disfiliation-tyne reton furnace.
-------
Zinc-Melting Processes
297
The vaporized zinc is conducted either to a con-
denser or discharged through an orifice into a
stream of air. Two types of condenser are used--
a brick-lined steel condenser operated at from
780° to 1, 012°F to condense the vapor to liquid
zinc, or a larger, unlined steel condenser that
cools the vapor to solid zinc. The latter con-
denser is used to manufacture powdered zinc.
The condensers must be operated at a slight pos-
itive pressure to keep air from entering them and
oxidizing the zinc. To ensure that there is a pos-
itive pressure, a small hole, called a "speise"
hole, is provided through which a small amount
of zinc vapor is allowed to escape continuously
into the atmosphere. The vapor burns with a
bright flame, indicating that there is a pressure
in the condenser. Iftheflame gets too large, the
pressure is too high. Ifitgoes out, the pressure
is too low. In either case, the proper adjust-
ments are made to obtain the desired condenser
pressure.
When it is desired to make zinc oxide, the vapor
from a retort is discharged through an orifice
into a stream of air where zinc oxide is formed
inside a refractory-lined chamber. The com-
bustion gases and air, which bear the oxide par-
ticles, are then carried to a baghouse collector
where the powdered oxide is collected.
The Air Pollution Problem
During the 24-hour cycle of the distillation re-
torts, zinc vapors escape irom the retort (l)when
the residue from the preceding batch is removed
from the retort and a new batch is charged, and
(2) when the second charge is added to the retort.
As the zinc vapors mix with air, they oxidize and
form a dense cloud of zinc oxide fumes. Air con-
taminants are discharged for about 1 hour each
time the charging hole is open. When the zinc is
actually being distilled, no fumes escape from
the retort; however, a small amount of zinc oxide
escapes from the speise hole in the condenser.
Although the emission rate is low, air contami-
nants are discharged for about 20 hours per day.
Hooding and Ventilation Requirements
To capture the emissions from a distillation re-
tort furnace, simple canopy hoods placed close
to and directly over the sources of emissions are
sufficient. In the only installation in Los Angeles
County, the charging end of the retort protrudes
a few inches through a 4-foot-wide, flat wall of
the furnace. The hood is 1 foot above the retort,
extends 1-1/4 feet out from the furnace wall, and
is4feetwide. The ventilation provided is 2,000
cfm, giving a hood indraft of 400 fpm. Fume
oickup is excellent. The speise hole is small
and all the fumes discharged are captured by a
1-foot-diameter hood provided with 200 cfm ven-
tilation. The hood indraft is 250 fpm.
The retorts are gas fired and the products of
combustion do not mix with the emissions from
the retort or the condenser. The exhausted gases
are heated slightly by the combustion of zinc and
from radiation and convection losses from the re-
tort, but the amount of heating is so low that no
cooling is necessary.
MUFFLE FURNACES
Muffle furnaces (Figure 204) are continuously fed
retort furnaces. They generally have a much
greater vaporizing capacity than either Belgian
retorts or bottle retorts do,and they are operated
continuously for several days at a time. Heat for
vaporization is supplied by gas- or oil-fired burn-
ers by conduction and radiation through a silicon
carbide arch that separates the zinc vapors and
the products of combustion. Molten zinc from
either a melting pot or sweat furnace is charged
through a feed well that also acts as an air lock.
The zinc vapors are conducted to a condenser
where purified liquid zinc is collected, or the
condenser is bypassed and the vapors are dis-
charged through an orifice into a stream of air
•where zinc oxide is formed.
A muffle furnace installation in Los Angeles
County consists of three identical furnaces, each
capable of vaporizing several tons of zinc per
day. These furnaces can produce zinc of 99. 99
percent purity and zinc oxide of 99. 95 percent
purity from zinc alloys. Each furnace has three
sections: (1) A vaporizing chamber, (2) a con-
denser, and (3) a sweating chamber. Figure
205 shows the feed ends of the furnaces, includ-
ing the sweating chambers, and some of the duct-
work and hoods serving the furnaces.
Each furnace, including the feed -well and sweat-
ing chamber, is heated indirectly with a combina-
tion gas- or oil-fired burner. The combustion
chamber, located directly over the vaporizing
chamber, is heated to about 2, 500°F. On leav-
ing the combustion chamber, the products of com-
bustion are conducted over the zinc feed well and
through the sweating chamber to supply the heat
needed for melting the zinc alloys from the scrap
charged and for heating the zinc in the feed well
to about 900°F.
Zinc vapors are conducted from the vaporizing
section into a multiple-chamber condenser. When
zinc oxide is the desired product, the vapors are
allowed to escape through an orifice at the top of
the first chamber of the condenser. Even when
maximum zinc oxide production is desired, some
molten zinc is nevertheless formed and collects
in the condenser.
-------
298
METALLURGICAL EQUIPMENT
MOLTEN 'METAL
UP 'HOLE
DUCT fOR OXIDE
COLLECTION
RISER CONDENSER
UNIT
figure 204. 'Diagram of a muffle furnace and condenser.
Figure 205. (Left) Zinc-vaporizing muffle furnaces,
(Pacific Smelting Co., Torrance, Calif.).
(right) baghouse for collecting the zinc oxide manufactured
When metallic zinc is the desired product, the
size of the orifice is greatly reduced, but not
entirely closed, so that most of the vapors enter
the second section of the condenser where they
condense to molten zinc. The molten zinc col-
lected in the condenser is held at about 900°F in
a well, fromwhichitis periodically tapped. The
well and the tap hole are so arranged that suffi-
cient molten zinc always remains in the well to
maintain an air lock.
The zinc that escapes from the orifice while mol-
ten zinc is being made burns to zinc oxide, -which
is conducted to the product baghouse.
The Air Pollution Problem
Dustandfumes are createdby the sweating oper-
tion. Scrap is charged into the sweating chamber
through the door shown in Figure 205. After the
-------
Zinc-Melting Processes
299
zinc alloys have been melted, the residue is
pushed out of the chamber through a second door
and onto a shaker screen where dross is sepa-
rated from solid metal. Excessive dust and fumes
are thereby created.
The zinc alloys charged into the vaporizing sec-
tion contain copper, aluminum, iron, lead, and
other impurities. As zinc is distilled from the
metals, the concentration of the impurities in-
creases until continued distillation becomes im-
practical. After 10 to 14 days of operation, the
residue, containing 10 to 50 percent zinc must
be removed. When tapped,' the temperature of
the residue is about 1,900°F, hot enough to re-
lease zinc oxide fumes. The molds collecting
the residue metal are so arranged that the metal
overflows from one mold to another; however,
the metal cools so rapidlythatfum.es are released
only from the pouring spout and the first two or
three molds. The fumes, almost entirely zinc
oxide, are 100 percent opaque from the pouring
spout and the first mold. At the third mold, the
opacity decreases to 10 percent.
Any discharge of zinc vapor from the condenser
forms zinc oxide of product purity; therefore, the
condenser vents into the intake hood of a product-
collecting exhaust system. Sinc'e some zinc oxide
is always produced, even when the condenser is
set to produce a maximum of liquid zinc, the
product-collecting exhaust system is always in
operation to prevent air contaminants from es-
caping from the condenser to the atmosphere.
Hooding and Ventilation Requirements
The dust and fumes created by the charging of
scrap and the sweating of zinc alloys from the
scrap originate inside the sweat chamber. The
thermal drafts cause the emissions to escape
from the upper portion of the sweat chamber
doors. Hoods are placed over the doors to col-
lect the emissions. The charging door hood ex-
tends 10 inches from the furnace wall and covers
a little more than the width of the door (see Fig-
ure Z05). With two furnaces in operation at the
same time, each of the charging door hoods is
supplied with 3, 200 cfm ventilation, which pro-
vides an indraft velocity of 700 fpm. All fumes
escaping from the charging doors are collected
by these hoods.
The unmelted scrap and dross are raked from a
sweating chamber onto a shaker screen. A hood
enclosing the discharge lip and the screen is pro-
vided •with 5, 500 cfm ventilation. The inlet ve-
locity is 250 fpm, sufficient to capture all of the
emissions escaping from both the furnace and the
screen.
A hood 3 feet square positioned over the residue
metal-tapping spout and the first mold is pro-
videdwithS, 700 cfm ventilation. During the tap-
ping, no metal is charged to either sweating
chamber, and the exhaust system dampers are
arranged so that approximately one-half of the
available volume is used at the tapping spout. The
indraft velocity is in excess of 900 fpm, and all
fumes released from the metal are collected, even
from the second and third molds up to 6 feet away
from the hood.
The ductwork joining the hoods to the control de-
vices is manifolded and dampered so that any or
all hoods can be opened or closed. The exhaust
system provides sufficient ventilation to control
the fumes createdby two furnaces in operation at
the same time. When residue metal is being
tapped from a furnace, no metal is being charged
to the other furnaces; therefore, all the ventila-
tion, or as much as is needed, can be used at the
tapping hood.
AIR POLLUTION CONTROL EQUIPMENT
For all the furnaces mentioned in this section,
that is reduction retort furnaces, distillation
retort furnaces, and muffle furnaces, air pollu-
tion control is achieved with a baghouse. In the
above-mentioned installation for a muffle furnace,
a low-efficiency cyclone and a baghouse are used
to control the emissions from the sweating cham-
bers and residue pouring operations of the three
muffle furnaces. Although the cyclone has a low
collection efficiency, it does collect from 5 to 10
per cent of the dust load and it is still used. The
cyclone was in existence before the taaghouse -was
installed.
The baghouse is a six-section, pull-through type
using 5, 616 square feet of glass cloth filtering
area. The filtering velocity is 3 fpm and the bags
are cleaned automatically at regular intervals by
shutting off one section, which allows the bags to
collapse. No shaking is required, and the col-
lected material merely drops into the hopper be-
low the bags.
Another exhaust system with a cyclone and bag-
house is used to collect the zinc oxide manufac-
tured by the muffle furnaces. The system has
three inlet hoods, one for each furnace, and each
is arranged to collect the zinc vapors discharged
from the orifice in the condenser. The ductwork
is manifolded into a single duct entering the cy-
clone, and dampers are provided so that any one
or any combination of the hoods can be used at
onetime. Since the exhausted gases and zinc ox-
ide are heated by the combustion of zinc and by
the sensible heat in the zinc, about 350 feet of
additional ductwork is provided to allow the ex-
-------
300
METALLURGICAL EQUIPMENT
hausted material to cool down to 180°F before
entering the baghouse.
The cyclone collects about 20 percent of the solid
materials in the exhaust gases, including all the
heavier particles such as vitrified zinc oxide and
solid zinc. The baghouse collects essentially all
the remaining 80 percent of the solids.
The baghouse collector is actually two standard
nine-section baghouses operating in parallel. In
this unit, orlonbags with a total of 16,848 square
feet of filtering area are used to filter the solids
from the gases. A 50-hp fan provides 30,500
cfm ventilation--15, 250 cfm for each furnace.
The filtering velocity is 1. 8 fpm. The bags are
cleaned at regular intervals by shutting off one
section and shaking the bags for a few seconds.
A screw conveyor in the bottom of each hopper
conveys the zinc oxide collected to a bagging ms.-
chine.
This system provides excellent ventilation for the
installation. None of the zinc oxide discharging
from the condensers escapes collection by the
hoods, and no visible emissions can be seen es-
caping from the baghouse.
Dust collectors for other zinc-melting and zinc-
vaporizing furnaces are very similar to the ones
already described. Glass bags have been found
adequate when gas temperatures exceed the Limits
of cotton or orlon. Filtering velocities of 3 fpm.
are generally employed and have been found ade-
quate.
LEAD REFINING
Control of the air pollution resulting from the
secondary smelting and reclaiming of lead scrap
maybe conveniently considered according to the
type of furnace employed. The reverberatory,
blast, and pot furnaces are the three types most
commonly used. In addition to refining lead,
most of the secondary refineries also produce
lead oxide by the Barton process.
Various grades of lead metal along with the oxides
are producedby the lead industry. The grade of
product desired determines the type of equipment
selected for its manufacture. The most common
grades of lead produced are soft, semisoft, and
hard. By starting with one of these grades and
using accepted refining and alloying techniques,
any special grade of lead or lead alloy can be
made.
Soft lead may be disignated as corroding, chem-
ical, acid copper, or common desilverized lead.
These four types are high-purity leads. Their
chemical requirements are presented in Table 85.
These leads are the products of the pot furnace after
a considerable amount of refining has been done.
Semisoft lead is the product of the reverberatory-
type furnace and usually contains from 0. 3 to 0. 4
percent antimony and up to 0. 05 percent copper.
Hard lead is made in the blast furnace. A typ-
ical composition for hard lead is 5 to 12 percent
antimony, 0. 2toO. 6 percent arsenic, 0. 5 to 1. 2
percent tin, 0.05 to 0. 15 percent copper, and
0. 001 to 0. 01 percent nickel,
REVERBERATORY FURNACES
Sweating operations are usually conducted in a
reverberatory-type furnace or tube. This type
of operation is discussed later in this chapter
in a section on "Metal SepEiration Processes. '
The reverberatory furnace is also used zo re-
claim lead from oxides and drosses. Very often
material for both sweating and reducing such as
lead scrap, battery plates, oxides, drosses, and
lead residues are charged to a reverberator^
furnace. The charges are made up of a mixture
of these materials and put into the furnace in
such a manner as to keep a very small mound
of unmeltec material on top of the bath. As the
mound becomes molten, more material is charged.
This type of furnace may be gas fired or oil fired,
or a combination of both. The temperature is main-
tained at approximately 2, 300°F. Only sufficient
draft is pulled to remove the smoke and fumes and
still allow the retention of as much heat as possible
over the hearth. The molten metal is tapped off
at intervals as a semisoft lead as the level of the
metal rises. This operation is continuous, and
recovery is generally about lOto 12 pounds of met-
al per hour per square foot of hearth area.
The Air Pollution Problem
Afairly high percentage of sulfur is usually pres-
ent in various forms in the charge to the rever-
beratory furnace. The temperature maintained
is sufficiently high to "kill" the sulfides and re-
sults in the formation of sulfur dioxide and sulfur
trioxide in the exit gases. Also present in the
smoke and fumes produced are oxides, sulfides,
and sulfates of lead, tin, arsenic, copper, and
antimony. An over all material balance shows on
the product side approximately 47 percent recov-
ery of metal, 46 percent recovery of slag some-
times called "litharge, " and 7 percent of smoke
and fumes.
The unagglomerated particulate matter emitted
from secondary lead-smelting operations has been
foundtohave aparticle size range from 0.07 to 0.4
micron with a mean of about 0. 3 micron (Allen
et al. , 1952). Figure 206 shows electron photo-
-------
Lead Refining
301
Table 85. CHEMICAL REQUIREMENTS FOR LEAD3
(ASTM Standards, Part 2, 1958)
Silver, max %
Silver, min. %
Copper, max %
Copper, min. %
Silver and copper together,
max %
Arsenic, antimony, and
tin together, max %
Zinc, max %
Iron, max %
Bismuth, max %
Lead (by difference),
min, %
Corroding
lead
0. 0015
0. 0015
0. 0025
0. 002
0.001
0. 002
0. 050
99.94
Chemical
lead
0. 020
0.002
0. 080
0.040
0.002
0.001
0. 002
0.005
99.90
Acid-
copper
lead
0. 002
0. 080
0. 040
0. 040
0. 002
0. 001
0. 002
0. 025
99.90
Common
desilverized
lead
0, 002
0. 0025
0. 005
0.002
0. 002
0. 150
99. 85
aCorroding lead is a designation used in the trade for many years to
describe lead refined to a high degree of purity.
Chemical lead is a term used in the trade to describe the undesilverized
lead produced from Southeastern Missouri ores.
Acid-copper lead is made by adding copper to fully refined lead.
Common desilverized lead is a designation used to describe fully
refined desilverized lead.
micrographs of lead fumes. The particles are
nearly spherical and have a distinct tendency to
agglomerate. The concentration of particulate
matter in stack gases ranges from 1.4 to 4. 5
grains per cubic foot.
Hooding and Ventilation Requirements
All the smoke and fumes produced by the rever-
beratory furnace must be collected and, since
they are combined •with the products of combus-
tion, the entire volume emitted from the furnace
must pass through the collector. It is not desir-
able to draw cool air into these furnaces through
the charge doors, inspection ports, or other open-
ings to keep air contaminants from escaping from
them; therefore, externalhoods are used to cap-
ture these emissions. The ventilating air for
these hoods as well as for the hoods venting slag
stations must also pass through the collector. In
large furnaces, this represents a considerable
volume of gases at fairly high temperatures.
Air Pollution Control Equipment
The only control systems found to operate satis-
factorily in>Los Angeles County have been those
employing abaghouse as a final collector. These
systems also include auxiliary items such as gas -
cooling devices and settling chambers.
A pull-through type of baghouse with compart-
ments that can be shut off one at a time is very
satisfactory. This allows atmospheric air to
enter one compartment and relieve any flow. The
bags may then be cleaned by a standard mechan-
ical shaking mechanism.
Provision should be made to prevent sparks and
burning materials from contacting the filtercloth,
and temperature must be controlled by preced-
ing the baghouse -with radiant cooling ducts, water-
jacketed cooling ducts, or other suitable devices
in order that the type of cloth used will have a
reasonable life. The type of cloth selected de-
pends upon parameters such as the temperature
and corrosivity of the entering gases, and the
permeability and abrasion- or stress-resisting
characteristics of the cloth. Dacron bags are
being successfully used in this service. The fil-
tering velocity should not exceed 2 fpm. Test
results of secondary lead-smelting furnaces vent-
ing to a baghouse control device are shown in
Table 86.
-------
30Z
METALLURGICAL EQUIPMENT
Figure 206. Electron photomicrographs of lead fumes (Allen et al., 1952).
The factors to be considered in designing these
control systems are similar to those discussed
previously in the sections on iron casting and
steel manufacturing.
LEAD BLAST FURNACES
The lead blast furnace or cupola is constructed
similarly to those used in the ferrous industry.
The materials forming the usual charge for the
blastfurnace, and a typical percentage composi-
tion are 4. 5 percent rerun slag, 4. 5 percent scrap
castiron, 3 percent limestone, 5. 5 percent coke,
and 82. 5 percent drosses, oxides, andreverbera-
tory slags. The rerun slag is the highly silicated
slag from previous blast furnace runs. The
drosses are miscellaneous drosses consisting
of copper drosses, caustic drosses, and dry
drosses obtained from refining processes in the
pot furnaces. The processes -will be described
inmore detail in the following paragraphs. The
coke is used as a source of heat, and combustion
air is introduced near the bottom of the furnace
through tuyeres at a gage pressure of about 8 to
12 ounces per square inch. Hard lead is charged
into the cupola at the start of the operation to
provide molten metal to fill the crucible. Normal
-------
Lead Refining
303
Table
DUST AND FUME EMISSIONS FROM A SECONDARY LEAD-SMELTING FURNACE
Test No.
Furnace data
Type of furnace
Fuel used
Material charged
Process weight, lb/hr
Control equipment data
Type of control equipment
Filter material
Filter area, ft2
Filter velocity, fpm at 327 °F
Dust and fume data
Gas flow rate, scfm
Furnace outlet
Baghouse outlet
Gas temperature, °F
Furnace outlet
Baghouse outlet
Concentration, gr/scf
Furnace outlet
Baghouse outlet
Dust and fume emission, lb/hr
Furnace outlet
Baghouse outlet
Baghouse efficiency, %
Baghouse catch, wt %
Particle size 0 to 1 jx
1 to 2
2 to 3
3 to 4
4 to 16
Sulfur compounds as SO->, vol %
Baghouse outlet
1
Reverberatory
Natural gas
Battery groups
2,500
Sectioned tubular baghousea
Dacron
16,000
0.98
3, 060
10,400b
951
327
4. 98
0. 013
130.5
1. 2
99. 1
13. 3
45.2
19. 1
14. 0
8.4
0. 104
2
Blast
Coke
Battery groups, dross, slag
2,670
Sectioned tubular baghouse
Dacron
16, 000
0.98
2, 170
13,000b
500
175
12.3
0.035
229
3.9
98.3
13.3
45. Z
19, 1
14. 0
8.4
0. 03
f"The same baghouse alternately serves the reverberatory furnace and the blast furnace.
Dilution air admitted to cool gas stream.
charges, as outlined previously, are then added
as the material melts down. The limestone and
iron form the flux that floats on top of the molten
lead and retards its oxidation.
As the level of molten material rises, the slag
is tapped at intervals while the molten lead flows
from the furnace at a more or less continuous
rate. The lead product is "hard" or "antimonial. "
Approximately 70 percent of the molten material
is tapped off as hard lead, and the remaining 30
percent, as slag. About 5 percent of the slag is
retained for rerun later.
typical material balance based upon the charge
to a blast furnace in which battery groups are
being processed is 70 percent recovery of lead,
8 percent slag, lOpercentmat (sulfur compounds
formed with slag), 5 percent -water (moisture
contained in charge), and 7 percent dust (lead ox-
ide and other particulates discharged from stack
of furnace with gaseous products of combustion).
Particulate matter loading in blast furnace gases
is exceedingly heavy, up to 4 grains per cubic
foot. The particle size distribution is very simi-
lar to that from gray iron cupolas, as described
previously in the section on "Iron Casting. "
The Air Pollution Problem
Combustion air from the tuyeres passing verti-
cally upward through the charge in a blast fur-
nace conveys oxides, smoke, bits of coke fuel,
and other particulates present in the charge. A
Blastfurnace stack gas temperatures range from
1,200° tol,350°F. In addition to the particu-
late matter, which consists of smoke, oil vapor,
fume, and dust, the blast furnace stack gases
contain carbon monoxide. An afterburner is nee-
-------
304
METALLURGICAL EQUIPMENT
essary to control the gaseous, liquid, and solid
combustible material in the effluent.
Hooding and Ventilation Requirements
The only practical way to capture the contami-
nants discharged from a lead blast furnace is to
seal the furnace and vent all the gases to a con-
trol system. The hooding and ventilation require-
ments are very similar to those for the gray
iron cupola, which are discussed in the section
on "Iron Casting. "
Air Pollution Control Equipment
The control system for a. lead blast furnace is
similar to that employed for gray iron cupola fur-
naces except that electrical precipitators are not
usedfor economic reasons. Moreover, difficul-
ties are encountered in conditioning the particles
to give them resistivity characteristics in the
range that -will allow efficient collection.
The factors to be considered in designing a con-
trol system for a blast furnace, including an af-
terburner and a baghouse, have been discussed
in the section on "Iron Casting. "
POT-TYPE FURNACES
Pot-type furnaces are used for remelting, alloy-
ing, and refining processes. Remelting is usually
done in small pot furnaces, and the materials
charged are usually alloys in the ingot form,
which do not require any further processing ex-
cept to be melted for casting operations.
The pots used in the secondary smelters range
from the smallest practical size of 1-ton capac-
ity up to 50 tons. Figure 207 is a photograph of
two pot furnaces utilizing a common ventilation
hood. These furnaces are usually gas fired.
Various refining and alloying operations are car-
ried on in these pots. Alloying usually begins
with a metal lower in the percentage of alloy-
ing materials than desired. The percent desired
is calculated and the amount is then added. An-
timony, tin, arsenic, copper, and nickel are the
most common alloying elements used.
The refining processes most commonly employed
are those for the removal of copper and antimony
to produce soft lead, and those for the removal of
arsenic, copper, and nickel to produce hard lead.
For copper removal, the temperature of the mol-
ten lead is allowed to drop to 620 °F and sulfur is
added. The mixture is agitated and copper sulfide
is skimmed off as dross. This is known as "cop-
per dross" and is charged into the blast furnace.
Figure 207. An installation used to capture emissions from
two lead pot furnaces. Hood serves either furnace alternately
(Morris P. Kirk & Son, fnc., Los Angeles, Calif.).
When aluminum is added to molten lead, it reacts
preferentially with copper, antimony, and nickel
to form complex compounds that can be skimmed
from the surface of the metal. The antimony con-
tent can also be reduced to about 0. 02 percent by
bubbling air through the molten lead. It can be
further reduced by adding a mixture of sodium
nitrate and sodium hydroxide and skimming the
resulting dross from the surface of the metal.
Another common refining procedure, "dry dross-
ing, " consists of introducing sawdust into the
agitated mass of molten metal. This forms car-
bon, which aids in separating the globules of lead
suspended in the dross, and reduces some of the
lead oxide to elemental lead.
In areas where there is no great concern about
air pollution, a mixture of sal ammoniac and rosin
may be used to clean the metal of impurities.
This method, however, produces copious quanti-
ties of dense, white fumes, and obnoxious odors.
In areas having air pollution laws, this method
is generally no longer used.
The Air Pollution Problem
Although the quantity of air contaminants dis-
chargedfrompotfurnaces as a result of remelt-
ing, alloying, and refining is much less than that
from reverberatory or blast furnaces, the cap-
ture and control of these contaminants is equally
important in order to prevent periodic violations
-------
Metal Separation Processes
305
of air pollution regulations and protect the health
of the employees.
Problems of industrial hygiene are inherent in
this industry. People working with this equip-
ment frequently inhale and ingest lead oxide fumes,
which are cumulative, systemic poisons. Fre-
quent medical examinations are necessary for all
employees, and a mandatory dosage of calcium
dis odium versenate maybe required daily in order
to keep the harmful effects to a minimum.
Hooding and Ventilation Requirements
Hood design procedures for pot furnaces are the
same as those outlined for electric-induction fur -
naces mentioned earlier in this chapter,
Air Pollution Control Equipmen,
The control systems for pot turnaces, as -with the
other lead furnaces, require the use of a baghouse
for the final collector. The temperature of the
gases is, however, generally much lower than
that from the other furnaces; therefore, the gas-
cooling devices, if needed, will be much smaller.
Afterburners are generally not required.
processes classified as metal separation that can
be troublesome from an air pollution standpoint.
In these, the metal desired is recovered from
scrap, usually a mixture of several metals.
Probably the most common of these processes,
aluminum sweating, is the recovery of aluminum
from aluminum drosses and other scrap. Other
examples of metal separation processes include
the recovery processes for zinc, lead, solder,
tin, and low-melting alloys from a host of scrap
materials.
ALUMINUM SWEATING
Open-flame, reverberatory-type furnaces are
used by secondary smelters to produce alumi-
num pigs for remeiting. These furnaces are con-
structed with the hearths sloping downward toward
the rear of the furnace. All. types 01 scrap alu-
minum are charged into one of these furnaces,
which operates at temperatures of ],250" to
I,400°F. In this temperature range, the alumi-
num melts, trickles down the hearth, and flows
from the furnace into a mold. The higher melt-
ing materials such as iron, brass, and dross
oxidation products formed during melting remain
within the furnace. This residual material is
periodically raked from the furnace hearth.
BARTON PROCESS
A rather specialized phase of the industry is the
production of lead oxide. Battery lead oxide,
containing about 20 percent finely divided free
lead, is usually produced by the Barton process.
Molten lead is run by gravity from, a melting pot
intoakettle equipped with paddles. The paddles
are rotated at about 150 rpm, rapidly agitating
the molten lead, which is at a temperature of 700°
to 900°F. Air is drawn through the kettles by
fans located on the air outlet side of a baghouse.
The lead oxide thus formed is conveyed pneu-
matically to the baghouse where it is collected
and delivered by screw conveyor to storage.
Other lead oxides requiring additional processing
but commonly made are red lead oxide (minium,
PbjO^j), used in the paint industry, and yellow
lead oxide (litharage or massicot, PbO), used in
the paint and ink industries.
Sincethe process requires the use of a baghouse
to collect the product, and no other contaminants
are discharged, no air pollution control system
as such is needed.
METAL SEPARATION PROCESSES
In addition to the metallurgical processes previ-
ouslymentioned in this chapter, there are other
Some large secondary aluminum smelters sepa-
rate the aluminum suspended in the dross by pro-
cessing the hot dross immediately after its re-
moval from the metal in the refining furnace. The
hot dross is raked into a refractory-lined barrel
to which a salt-cryolite flux is added. The bar~
rel is placed on a cradle and mechanically rotated
for several minutes. Periodically, the barrel is
stopped and the metal is tapped by removing a
clay plug in the base of the barrel. This process
continues until essentially all the free aluminum
has been drained and only dry dross remains.
The dross is then dumped and removed from the
premises. A hot dross-processing station has
been illustrated previously in Figure 200.
The aluminum globules suspended in the dross as
obtained from the hot dross process can also be
separated and reclaimed by a cold, dry, milling
process. In this process the large chunks of
dross are reduced in size by crushing and then
fed continuously to a ball mill where the oxides
and other nonmetallics are ground to a fine pow-
der, which allows separation from the larger
solid particles of aluminum. At the mill dis-
charge, the fine oxides are removed pneumatical-
ly and conveyed to a baghouse for ultimate dis-
posal. The remaining material passes over a
magnetic roll to remove tramp iron and is then
discharged into storage bins to await melting.
This process is used primarily to process dross-
es having a low aluminum content.
-------
306
METALLURGICAL EQUIPMENT
ZINC, LEAD, TIN, SOLDER, AND LOW-MELTING
ALLOY SWEATING
Although recovery of aluminum is the most com-
mon of the metal separation processes, others
that contribute to air pollution deserve mention.
These include zinc, lead, tin, solder, and low-
melting alloy sweating. Separation of these metals
by sweating is made possible by the differences
in their melting point temperatures. Some of
these melting temperatures are:
Tin
Lead
Zinc
Aluminum
Copper
Iron
450°F
621°F
787°F
1,220°F
1,981°F
2,795°F
When the material charged to a sweating furnace
contains a combination of two of these metals, it
canbe separated by carefully controlling the fur-
nace temperature so that the metal with the lower
melting point is sweated when the furnace tem-
perature is maintained slightly above its melting
point. After this metal has been melted and re-
moved, the furnace burners are extinguished and
the metal with the higher melting point is raked
from the hearth.
Zinc can be recovered by sweating in a rotary,
reverberatory, or muffle furnace. Zinc-bear-
ing materials fed to a sweating furnace usually
consist of scrap die-cast products such as auto-
mobile grilles, license plate frames, and zinc
skims and drosses.
The sweating of lead from scrap and dross is
widely practiced. Junk automobile storage bat-
teries supply most of the lead. In addition, lead-
sheathed cable and wire, aircraft tooling dies,
type metal drosses, and lead dross and skims
are also sweated. The rotary furnace, or sweat-
ing tube, is usually used when the material pro-
cessed has a low percent of metal to be recovered.
The reverberatory box-type furnace is usually
used when the percent of metal recovered is high.
Rotary and reverberatory furnaces are also used
to sweat solder and other low-melting alloys from
scrap metal. Automobile radiators and other
soldered articles such as gas meter boxes, radio
chassis, and so forth, make up the bulk of the
process metal, For this recovery, the furnace
is usually maintained between 650°F and 700°F.
Higher temperatures should be avoided in order
to prevent the possible loss of other recoverable
metals. For example, sweating automobile ra-
diators at 900°F causes excessive oxidation of
the copper.
The Air Pollution Problem
Contaminants From Aluminum-Separating
Processes
In theory, an aluminum-sweating furnace can be
operated with minor emissions of air contami-
nants if clean, carefully hand-picked metal free
of organic material is processed. In practice,
this selective operation does not occur and ex-
cessive emissions periodically result from un-
controlled furnaces. Stray magnesium pieces
scattered throughout the aluminum scrap are not
readily identified, and charging a small amount
of magnesium into a sweating furnace causes
large quantities of fumes to be emitted. Emis-
sions also result from the other materials charged,
suchas skims, drosses, scrap aluminum sheet,
p6tsandpans, aircraft engine s, and wrecked air-
planes containing oil, insulated wire, seats, in-
struments, plastic assemblies, magnesium and
zinc components, and so forth.
Smoke is caused by the incomplete combustion of
the organic constituents of rubber, oil and grease,
plastics, paint, cardboard, and paper. Fumes
result from the oxidation of stray magnesium or
zinc assemblies and from the volatilization of
fluxes in the dross. The sweating of dross and
skims is responsible for the high rates of emis-
sion of dust andfumes. Residual aluminum chlo-
ride flux in the dross is especially troublesome
because it sublimes at 352 °F and is very hygro-
scopic. In addition, it hydrolyzes and forms very
corrosive hydrogen chloride. In Table 87, test 1
shows results from an aluminum-sweating fur-
nace.
In the dry milling process, dust is generated at
the crusher, in the mill, at the shaker screens,
and at points of transfer. These locations must
be hooded to prevent the escape of fine dust to
the atmosphere.
When aluminum is reclaimed by the hot dross
process, some fumes are emitted from the flux
action; however, the main air pollution problem
is the collection of the mechanically generated
dust created by the rotation of the dross barrel.
Contaminants from low-temperature sweating
Air contaminants released from a zinc-sweating
furnace consist mainly of smoke and fumes. The
smoke is generated by the incomplete combustion
of the grease, rubber, plastics, and so forth
contained in the material. Zinc fumes are neg-
ligible at lowfurnace temperatures, for they have
a low vapor pressure even at 900°F. With ele-
vated furnace temperatures, however, heavy fum-
ing can result. In Table 87, test 2 shows results
from a zinc die-cast-sweating operation.
-------
Core Ovens
309
Figure 208. Aluminum-sweating furnace vented to an afterburner and baghouse
(Du-Pol Enterprises, Los Angeles, Calif.).
can be used to incinerate the contaminant, and a
baghouse maynotbe required. Conversely, only
a baghouse is required when the process scrap
is always free of oils or other combustible wa^te.
Water scrubbers have not proved satisfactory in
the collection of metallic fumes of this type.
CORE OVENS
In foundries, core ovens are used to bake the
cores used in sand molds. Most cores contain
binders that require baking to develop the strength
needed to resist erosion and deformation by metal
during the filling of the mold. Core ovens supply
the heat and, where necessary, the oxygen nec-
essary for the baking. Cores aremade in a large
variety of sizes and shapes and with a variety of
binders; therefore, a variety of types of core
ovens are needed to provide the space and heat
requirements for baking the cores.
Generally, emissions from core ovens are a mi-
nor source of air pollution when compared with
other metallurgical processes. If the ovens are
operated below 400 °F and are fire ~ jfcjth natural
gas, emissions are usually tolerah
less, there are instances, for
special core formulations are us .,,
sions can have opacities exceecc°
permitted in Los Angeles County, •ld-c
sions can be extremely irritating ^
cause of aldehydes and other oxidat^
In these cases, a control device is .
normally an afterburner.
TYPES OF OVENS
The various types of core ovens fall into the fol-
lowing five classes: Shelf ovens, drawer ovens,
portable-rack ovens, car ovens, conveyor ovens.
Shelf ovens are probably the simplest form of
core ovens. They are merely insulated steel
boxes, divided into sections by shelves. Core
plates carrying cores are placed directly on the
shelves. When a door is opened, all or at least
several shelves are exposed and a large amount
of heat escapes from the oven chamber. Figure
209 shows a gas-fired shelf oven. The hot gas-
es escaping during loading and unloading of the
shelves not only waste heat but also create unde-
sirable working conditions. Because of these un-
desirable characteristics, these ovens are gen-
erally limited to baking small cores, particular-
ly in a small-core department where the invest-
ment in oven equipment must be kept at a mini-
mum.
Shelf ovens have been replaced largely by the
more efficient drawer oven. One type of drawer
oven is shown in Figure 210. With these ovens,
one or more drawers can be withdrawn for load-
ing or unloading and, since the drawers are
equipped with rear-closing plates, hot gases do
not escape. Within the oven, the drawers are
supported on rollers and, when withdrawn, the
front end is supported by an overhead drawer-
selector with an operating arrangement to per-
mit engagement of any one or any combination of
v i -
-------
310
METALLURGICAL EQUIPMENT
Table 88. DUST AND FUME EMISSIONS FROM
AN ALUMINUM-SWEATING FURNACE CONTROLLED BY
AN AFTERBURNER AND BAGHOUSE
Furnace data
Type of furnace
Furnace hearth area
Process weight, Ib/hr
Material sweated
T
Reverberatory with integral afterburner
4 ft 7 in. W x 8 ft 10 in. L
2,870
Scrap aluminum
Baghouse data
Type of bags
Filter material
Filter area, ft2
Filter velocity, fpm
Precleaner
Tubular
Dacron
4,800
2.
Settling chamber
16
Dust and fume data
Gas flow rate, scfm
Average gas temperature, °F
Concentration, gr/scf
Dust and fume emission, Ib/hr
Particulate control efficiency, %
Settling
chamber inlet
1, 360
350
0. 505
5.89
Furnace charge
door hood
5, 580
204
0. 081
3. 88
Baghouse
outlet
8, 850a
150
0. 0077
0. 58
94. 1
Orsat analysis at settling
chamber inlet, volume %
CO2 6.
°2 8-
CO 0.
N2
H2°
77.
7.
6
02
33
25
Particle size analysis at bag-
house outlet, wt %
+ 60 mesh 85. 9
-60 mesh 14. 1
Particle size analysis of -60
mesh portion, wt %
0 to 2 (J. 6.9
2 to 5 |a 32. 4
5 to 10 fi 30. 9
10 to 20 fi 17. 7
20 to 40 |JL 7. 7
< 40 (J. 4. 4
Combustible carbon in particu-
lace discharge, dry wt %
Settling chamber
inlet 83.7
Furnace chamber
door hood exit 67. 3
aVolume is greater at the baghouse exit than at the 'nlet because of leakage.
These ovens are suitable for baking small- and
medium-sized cores, but they are limited in the
volume of cores that can be baked because of labor
involved in transporting the cores from the core
maker to the oven, placing them in the drawers,
removing them from the drawers, and taking
them to storage.
some of the handling of cores, por-
ens were developed. The core maker
cores directly onto a rack, •which,
is put into the oven. After the bak-
ack is removed and taken to storage.
, loaded rack can then be placed in the
gure 211 shows an empty rack oven.
-------
Metal Separation Processes
307
Table 87. DUST AND FUME EMISSIONS FROM AN ALUMINUM -
AND A ZINC-SWEATING FURNACE CONTROLLED BY A BAGHOUSE
Test No.
Furnace data
Type of furnace
Size of furnace
Process weight, Ib/hr
Material sweated
Baghouse data
Type of baghouse
Filter material
Filter area, ft
Filter velocity, fpm
Precleaner
Dust and fume data
Gas flow rate, scfm
Baghouse inlet
Baghouse outlet
Average gas temperature, °F
Baghouse inlet
Baghouse outlet
Concentration, gr/scf
Baghouse inlet
Baghouse outlet
Dust and fume emission, Ib/hr
Baghouse inlet
Baghouse outlet
Control efficiency, %
1
Reverberatory
5 ft 9 in.W x 6 ft
4 in. L x 4- ft H
760
Aluminum skims
'Sectioned tubular
Orion
5, 184
1.9
None
8,620
9,580
137
104
0. 124
0. 0138
9. 16
1. 13a
87. 7a
2
Reverberatory
5 ft 9 in. W x 6 ft
4 in. L x 4 ft H"
2, 080
Zinc castings
Sectioned tubular
Orion
5, 184
1.85
None
7, 680
7, 420
190
173
0.205
0.0078
13. 5
0.5
96.3
aVisible emissions released from the baghouse indicated that a
had broken during the latter part of the test.
The discharge from a lead-sweating furnace may
>e heavy with dust, fumes, smoke, sulfur com-
pounds, and fly ash. This is particularly true
when junk batteries are sweated. The battery
grourvs and plates removed from the cases con-
-Cts of asphaltic case, oil and grease around
' ""terminals, sulfuric acid, lead sulfate, lead
,ide, and wooden or glass fiber plate separators.
.he organic contaminants burn poorly and the
sulfur compounds release SO2 and 803. The sul-
fur trioxide is particularly troublesome; when
hydrolizedto sulfuric acid, the acid mist is dif-
ficult to collect and is extremely corrosive. The
lead oxide tumbles -within the rotating furnace
and the finer material is entrained in the vented
combustion gases.
Unagglomerated lead oxide fume particles vary
in diameter from about 0. 07 to 0. 4 micron, with
a mean of about 0. 3 micron (Allen et al, , 1952).
Uncontrolled rotary lead sweat furnaces emit ex-
cessively high quantities of air contaminants. Al-
though the other types of scrap lead and drosses
sweated in a reverberatory furnace are normally
muchless contaminated with organic matter and
acid, high emission rates occur periodically.
The contaminants generated during the sweating
of solder, tin, and other low-melting alloys con-
sist almost entirely of smoke and partially oxi-
dized organic material. The scrap metal charged
is usually contaminated with paint, oil, grease,
rust, and scale. Automobile radiators frequent-
ly contain residual antifreeze and sealing com-
pounds,
Hooding and Ventilation Requirements
The ventilation and hooding of .reverberatory fur-
naces and rotary furnaces used for the reclama-
tion processes just mentioned are similar to those
of furnaces of this type previously discussed in
this chapter. The exhaust system must have suf-
ficient capacity to remove the products of com-
bustion at the maximum firing rate and provide
adequate collection of the emissions from any
furnace opening.
In aluminum separation operations, raking the
residual metal and dross from the furnace is a
critical operation from an air pollution standpoint,
and hoods should be installed to capture emis-
-------
308
METALLURGICAL EQUIPMENT
sions at these locations. The required exhaust
volume may be effectively reduced by providing
a guillotine-type furnace door and opening it only
as needed to accomplish charging and raking. If
the burners are turned off during these opera-
tions, the indraft velocity through the charging
and raking opening is effectively increased and
the emissions from this location are reduced.
In low-temperature sweating operations, auxil-
iary hooding is usuallynecessary and varies with
the type of sweating furnace. For the convention-
al reverberatory-type furnace, a hood should be
installed above the furnace door so that escaping
fumes can be captured. The emissions occur
both during the normal melting process and dur-
ing the raking of the residual material from the
hearth. A rotary sweating furnace usually needs
only a hood over the high end of the tube. In cases
where the drosses are fine and dusty, however,
a hood Is necessary at the discharge end, too.
If the hoods are well designed and no unusual
crossdrafts are present, an indraft velocity of
100 to 200 fpm is adequate to prevent the escape
of the air contaminants.
Air Pollution Control Equipment
Aluminum-separating processes
Although air pollution control equipment is nec-
essary in aluminum reclamation processes, some
operating procedures reduce the quantity of emis-
sions. Whenever possible the stray magnesium
pieces and combustible material should be re-
moved from the aluminum scrap to be sweated.
The furnace burners should be operated so that
the flame does not impinge on the scrap metal,
particularly if the burners are oil fired.
An afterburner followed by a baghouse is recom-
mended as control equipment for an aluminum-
sweating furnace. Baghouse filtering velocities
should not exceed 3 fpm. The afterburner must
be so designed that the carbonaceous material
is intimately mixed with the exhaust air and held
at a suitable temperature for a sufficient length
of time to ensure complete incineration. For
this service, an afterburner temperature of
1, 200° to 1, 400"F is recommended with a re-
tention time of the gases in this hot zone of about
0.3 second. A luminous-flame afterburner is
generally the most desirable because of the great-
er flame area. Secondary air may have to be ad-
mitted to the afterburner to ensure complete com-
bustion. The afterburner may be constructed as
a separate unit from the furnace or may be con-
structed as an integral part of the furnace some-
what similar to a multiple-chamber incinerator.
General design features of afterburners have been
discussed in Chapter 5.
The hot gases must be cooled before enterii
baghouse, and radiant cooling or dilution -v
cold air is recommended in preference to eva
rative cooling with water. The sweating of a
minum drosses may result in severe corros
problems owing to the aluminum chloride f
contained in the dross. If the hot furnace gas
are cooled with water before entering the ba
house, the aluminum chloride hydrolyzes, pr
ducing hydrochloric acid. The ductwork and ba
are attacked, rapidly impairing the collection e
ficiency of the filter. Even the condensation fro
night air during shutdowns provides sufficier
moisture to corrode the equipment in the pres
ence of these chemicals.
Figure 208 shows an aluminum sweating furnac
with integral afterburner venting through hori.
zontally positioned radiation-convection coolinj
columns to a settling chamber andbaghouse. The
furnace charging door hood is vented directly tc
the baghouse. Table 88 shows test data acquired
while aluminum scrap heavily contaminated with
combustible material was being sweated in the
furnace. Combustible carbon -was present in the
particulate discharge and was coexistent in the
vent stream with excess oxygen as shown by the
Or sat analysis. This indicates that the rate of
combustible discharge from the scrap alumlnurr'
was in excess of the incinerating capacity of the
afterburner,
In thehot-dross process, the rotating barrel need
onlybe properly hooded and ducted to a baghouse.
INio afterburning is required, ana because of the
relatively large indraft air volume, no gas-cool-
ing facilities are required in the exhaust system.
In the dry milling process, the ball mill, crusher,
and all transfer points must be hooded and vented
to a baghouse in order to prevent the escape of
the dust created. The reqviired hood indi
locities vary from 150 to 500 fpm, depending""
on crossdrafts and the force -with which the dust
is generated, A baghouse filter velocity of 3 fpm
or less is recommended. No afterburning or gas -
cooling facilities are required in a dry-dross con-
trol system.
Low-temperature sweating
An afterburner should be provided to incinerate
the combustible matter discharged from a low-
temperature sweating furnace.
Since an afterburner cannot remove the noncom-
bustible portion of the effluent, a baghouse should
be used with the afterburner to capture the dust
and fumes. The maximum recommended bag-
house filter velocity is 3 fpm. In certain special
applications where the only emissions are oils
or other combustible material, an afterburner
-------
Core Ovens
311
Figure 209. Shelf oven (The Foundry Equipment Co.,
Cleveland, Ohio).
Racks are designed not only to fit the oven but
also to accommodate large or small cores. They
can be transported by an overhead monorail or
lift trucks, either manually or power operated.
For large cores, car ovens are generally used.
These ovens are similar to rack ovens but larger
and, instead of portable racks, cars riding on
rails are used. The cores, being large and heavy,
are generally loaded on the cars by crane. Tiered
pallets are frequently used to facilitate car load-
ing. Because of the size of the cores, most of a
day is usually needed to load a car; therefore,
baking is usually done overnight.
Conveyor ovens are used in foundries where a
large volume of cores of approximately the same
size arebaked. Of course, larger cores can also
be baked by allowing them to make two or more
passes through the oven.
Conveyor ovens have loading and unloading sta-
tions, a heated section, and a cooling section.
A horizontal-conveyor oven is shown in Figure
212. These ovens are generally located above
the floor level, in roof trusses, or above area-
ways between buildings. They have inclined en-
trances and exits to allow loading at the floor level
and, probably more important, to provide natural-
draft heat seals. The vertical-conveyor oven
shown in Figure 213 requires little floor space
for a large volume of baking. It is heated on the
Figure 210. Drawer oven (Despatch Oven Co.
Minneapolis, Minn.).
Figure 211. Rack oven (Despatch Oven Co.
Minneapolis, Minn.).
-------
312
METALLURGICAL EQUIPMENT
Figure 212. Horizontal, continuous oven (The Foundry Equipment Co., Cleveland, Ohio).
side where the cores enter the oven and through-
out the top of the oven. With the use of baffles
and a blower, the lower portion of the unloading
side of the oven cools the baked cores. Core
makers can be grouped around the loading side
of the oven to minimize the handling of cores.
HEATING CORE OVENS
Probably the simplest and crudest method of heat-
ing core ovens is to use burners along the floor
extending the entire length of the oven. These
burners cannot be regulated automatically and
they do not provide uniform heat throughout the
oven. They canbe dangerous, because of damping
out of the flame at the back of an oven, which al-
lows raw gas to accumulate resulting in explo-
sions. Although a few ovens are still heated in
this manner, most ovens use recirculating heater
units.
With recirculating heaters, a portion of the oven
gases is returned to the heater, and the rest is
vented through a dampered stack to the atmo-
sphere. Fresh air is mixed with the recirculated
gases, and the mixture is heated. The hot gases
and the products of combustion are blown into the
oven. The amount of fresh air admitted is con-
trolled by the amount of gases vented from the
oven and only enough is admitted to supply the
oxygen needed for the baking process.
CORE BINDERS
The primary reason for baking cores is to make
them strong enough so that they can be handled
while the mold is being made and so that they re-
sist erosion and deformation by metal when the
mold is being filled. The baking process drives
off water and other volatiles, which reduces the
total gas-forming material in the mold. Most of
the "volatiles" discharged can be considered air
contaminants. Their composition depends upon
the type of binder used in the core.
Numerous binders require baking, but they do
not all harden by the same chemical and physical
processes. Based on their method of hardening,
-------
Core Ovens
313
Figure 213. Vertical, continuous oven (The Foundry
Equipment Co., Cleveland, Ohio).
the binders can be subdivided into three types:
(1) Those that harden upon heating, (2) those that
harden upon cooling after being heated, and (3)
those that adhere upon heating. The binders of
the first type develop their strength by chemical
reaction, while those of the second and third types
function through physical phenomena.
Pitch, rosin, and similar materials of type 2 are
solids at room temperature, but upon heating,
they melt and flow around the sand grains. When
the mixture of sand and binder cools, the binder
solidifies and holds the grains together. Those
binders are frequently dissolved or dispersed in
solvent and, when baked, the solvent is driven
off, becoming an air contaminant.
The binders of type 3 are mixed with sand in the
dry state. Water is then added and the binder be-
comes gelatinous, which imparts green strength
to the mixture. Upon baking, these binders de-
hydrate, harden, and adhere to the sand grains
holding them together. Since baking only drives
water from the mixture, no air contaminants are
created.
Type-1 binders harden by chemical action, par-
tial oxidation, and polymerization. Drying oils,
of which linseed oil is typical, are made up of
unsaturated hydrocarbons that are liquid at room
temperature. Because they are unsaturated, the
molecules can react with other molecules or ele-
ments without producing side products. These
oils react with oxygen very slowly at room tem-
perature and faster at elevated temperatures, to
the extent that their unsaturation is partially sat-
isfied, and then they polymerize to form a solid
filmthatholds the sand grains together. If, how-
ever, toomuchheat is applied, the oxidation pro-
cess goes too far and some molecules break up
into lower molecular weight products instead of
polymerizing. The result is a weaker film, and
smoke, vapors, and gases are discharged.
The resin-type binders, such as phenol-formalde-
hyde, are intermediate, easily polymerized prod-
ucts of a phenol and formaldehyde condensation
reaction. When heateu, these compounds poly-
merize rapidly into a hard film. No side reac-
tions should, however, occur; these substances,
too, are organic and subject to burning if heated
excessively.
In actual practice, cores seldom contain only one
type of binder. A typical core mixture contains
930 pounds of sand, 7-1/2 pounds of core oil, 9
pounds of cereal binder, 3 pounds of kerosene,
and 38 pounds of water. The core oil contains
45 percent linseed oil, 28 percent gum rosin,
and 27 percent kerosene. All three types of
binder are present. The linseed oil in the core
oil is a type-1 binder and hardens by an oxida-
tion-polymerization process. The gum rosin of
the core oilis a type-2 binder and, after its sol-
vents are driven off., it melts and then hardens
whenthe cores are cooled. The cereal binder is
cornflour, a type-3 binder, which is used to im-
part green strength to the core by its gelatinous
reaction with -water before the core is baked.
During the baking of these cores, a series of
physical and chemical reactions occurs. First,
the moisture and light fractions of the oil are dis-
tilledoff. As the temperature rises, the heavier
fractions of the kerosene are vaporized and the
linseed oil begins to react with oxygen and to poly-
merize. At about 300 °F, the rosin melts, coat-
ing the grains with a thin film of rosin.
The polymerization of the linseed oil requires
more time than the physical changes that take
place do, and so the core is held at a. tempera-
ture of 375°Ffor 1-1/2 to 3-1/2 hours to develop
maximum strength. Ahigher temperature accel-
erates the polymerization, but the danger of over-
baking is alsomuch greater. For instance, when
linseed oil is baked at 375 °F, its maximum
strength is achieved in 1-1/2 hours, and its
-------
314
METALLURGICAL EQUIPMENT
strength does not deteriorate if it is baked for 3-
1/2 hours.. At400°F, a maximum strength, les.s
than that achieved at 375°F, is reached in 3/4
hour, but the strength begins to deteriorate if the
core is baked longer than 1-1/4 hours. And at
450 °F, the maximum strength is reached in a
little less than 3/4 hour and immediately begins
to deteriorate if overbaked. Of course, since the
entire body of the core cannot reach the oven tem-
perature at the same time, if high temperatures
are used the surface of the core is overbaked be-
fore the inner portions are completely baked.
Moreover, the high temperatures tend to create
smoke and objectionable gases that are discharged
from the oven as air contaminants.
The resinous-type binders also have kerosene
and cornflour added. Baking time and tempera-
ture requirements are, however, much less. In
fact, high-frequency dielectric ovens can be used
with the fast-setting synthetic resins. In these
ovens, the electrical field created causes noncon-
ductors within the field to become hot. The ovens
generally have a relatively small heating space,
through which a conveyor carries the cores. The
conveyor is one of the electrodes; therefore, only
the cores become heated. There are no hot gases
to contend with, and" only the small amount of
volatile materials in the cores are discharged.
Baking time generally runs 2-1/2 minutes.
Th« Air Pollution Problem
The air contaminants discharged from core ovens
consist of organic acids, aldehydes, hydrocarbon
vapors, and smoke. The vapors are the result
of the evaporation of hydrocarbon solvents, usu-
ally kerosene, and the light ends usually present
incoreoils. The organic acids, aldehydes, and
smoke are the result of partial oxidation of the
various organic materials in the cores. These
substances have obnoxious odors and are very
irritating to the eyes. The quantity and irritating
quality of the oxidation products generally in-
crease with an increase in baking temperature.
Emission rates, in general, are low, especially
from small- and medium-sized ovens operating
at 400°F or less. With some core binders, how-
ever, the emissions from small ovens operating
at low temperatures can be of sufficient quantity
to create a public nuisance. The emissions from
larger ovens are generally greater and are more
apt to create nuisances or be in excess of opacity
regulations. Table 89 shows the amounts of var-
ious contaminants discharged from three core
ovens. Test 1 shows the emissions from an un-
controlled oven, and tests 2 and 3 show the emis-
sions from two ovens as well as the afterburners
that control the emissions from them.
Table 89. AIR CONTAMINANT EMISSIONS FROM CORE OVENS
Test No.
Oven data
Size
Type
Operating temp, °F
Core binders
Weight of cores baked, Ib
Baking time, hr
Afterburner data
Size
Type
Burner capacity, Btu/hr
Air contaminants from:
Effluent gas volume, scfm
Effluent gas temperature, °F
Particulate matter, Ib/hr
Organic acids, Ib/hr
Aldehydes, ppm
Hydrocarbons, ppm
Opacity, %
Odor
1
6 ft 2 in. W x 7 ft 11 in.
H x 19 ft L
Direct gas -fired
380
1 to 1/2% phenolic resin
700
11
None
Oven
100
380
0. 13
0.068
52
124
0
Slight
2
3 ft 10 in. W x 5 ft 3
in H x 18 ft L
Direct gas-fired
400
3% linseed oil
1, 600
2-1/2 to 3
10 in. dia x 7 ft 6
in. H
Direct flame
200, 000
Oven
140
400
0. 2
0. 008
10
-
-
-
Afterburner
260
1,400
0. 013
0.000
10
< 10
0
Slight
3
4 ft 2 in. W x 6 ft 8
in. H x 5 ft 9 in. L
Indirect electric
400
1% linseed oil
600
6
3 ft dia x 4 ft H
Direct flame
600,000
Oven
250
400
0.27
0. 44
377
158
-
-
Afterburner
440
1, 780
0. 02
0. 087
4
< 19
0
None
-------
Foundry Sand-Handling Equipment
315
Excessive amounts of emissions can generally
be expected from ovens operated at 500 °F or
higher, and from ovens in which the cores baked
contain larger than normal amounts of kerosene,
fuel oil, or core oils. Visible emissions are
usually discharged from large conveyor ized ovens.
In many cases the opacity of these plumes has
been in excess of Los Angeles County's opacity
regulations.
Hooding and Ventilation Requirements
Most core ovens are vented directly to the atmo-
sphere through a stack. The ovens require suf-
ficient fresh air to be mixed with recirculated
gases and 'with the products of combustion from
the heater to keep the moisture content low and
to supply the oxygen necessary for proper bak-
ing of the drying oil-type core binders.
Generally, the excess gases and any contaminants
created are discharged from the oven through one
vent stack. Occasionally more than one vent is
used, but if the emissions are such that air pol-
lution controls are needed, then ducting the vents
to a control device is all that is necessary. The
use of hoods or of excess air is not necessary
to capture the emissions.
Air Pollution Control Equipment
As emphasized previously, when operated below
400 °F and when fired with natural gas, most core
ovens do not require air pollution control equip-
ment. There have been, however, several cases
•where excessive emissions have been discharged
and control equipment has been necessary.
Excessive emissions from core ovens have been
reduced to tolerable amounts by modifying the
composition of the core binders and lowering the
baking temperatures. For instance, smoke of
excessive opacity was discharging from an oven
baking cores containing 3 percent fuel oil and 1. 5
percent core oil at 500 °F. The core binder was
modified so that the cores contained 1. 5 percent
kerosene and 1. 5 percent core oil, and the bak-
ing temperature was reduced to 400°F. After
these modifications, no visible emissions were
discharged from the oven.
When it is not feasible or possible to reduce ex-
cessive emissions from an oven by modifying the
core mix or the baking temperature, afterburners
are the only control devices that have proved ef-
fective. Since the quantity and concentration of
the contaminants in the oven effluent are small,
no precleaners or flashback devices are needed.
Afterburners that have been used for controlling
the emissions from core ovens are predominantly
of the direct-flame type. The burners are nor-
mally designed to be capable of reaching a tem-
perature of at least 1, 200 °F under maximum load
conditions. For most operations, 1,200°F com-
pletely controls all visible emissions and prac-
tically all odors.
The afterburner shouldbe designed to have a max-
imum possible flame contact -with the gases to be
controlled and it should "be of sufficient size to
have a gas retention time of at least 0. 3 second.
Most authorities agree that the length-to-diameter
ratio should be in the range of 1-1/2 to 4.
In some instances, particularly on larger core
ovens, catalytic afterburners have been used to
control the emissions. With inlet temperatures
of from 600°to650°F, all visible emissions and
most of the odors were controlled. When cata-
lytic afterburners are used, however, care must
be taken to keep the catalyst in good condition;
otherwise, partial oxidation can result in the dis-
charge of combustion contaminants more objec-
tionable than the oven effluent.
FOUNDRY SAND-HANDLING EQUIPMENT
A foundry sand-handling system consists of a de-
vice for separating the casting from the mold,
and equipment for reconditioning the sand. The
separating device is usually a mechanically vi-
brated grate called a shakeout. For small cast-
ings a manual shakeout may be used.
TYPES OF EQUIPMENT
The minimum equipment required for recondi-
tioning the sand is a screen for removing over-
size particles, and a mixer-muller where clay
and water are combined with the sand to render
it ready for remolding. In addition, equipment
may be used to perform the following functions:
Sand cooling, oversize crushing, fines removal,
adherent coating removal, and conveying. A
typical sand-handling system is shown in Figure
214.
Both flat-deck screens and revolving, cylindrical
screens are used for coarse-particle removal.
Revolving screens can be ventilated at such a
rate as to remove excess fines.
Sand cooling can be accomplished in a number of
ways, depending upon the cooling requirements.
The amount of cooling required depends mainly
upon the ratio of metal to sand in the molds and
on the rate of re-use of the sand. With low metal-
to-sand and re-use ratios, no specific sand-cool-
ing equipment is required. When considerable
cooling is required, a rotary drum-type cooler
is usually used. A stream of air drawn through
the cascading sand both cools and removes fines.
-------
316
METALLURGICAL EQUIPMENT
TO BAGHOUSE
Figure 214. Typical foundry sand-handling system.
Oversize particles are hard agglomerates not
broken up by the handling operations from the
shakeout grate to the screen. Most of these are
portions of baked cores. Many foundries discard
the oversize particles, while others crush the
agglomerates to recover the sand. A hammer-
or screen-type mill is usually used for crushing.
Since molding sand is continuously reused, the
grains become coated with ahard, adherent layer
of clay and carbonaceous matter from the bonding
materials used. In time the sand becomes unus-
able unless the coating is removed or a certain
percentage of new sand is continuously added.
Pneumatic reclamation is the method most widely
usedfor coating removal. The sand is conveyed
in a high-velocity airstream from a turbine-type
blower and impinged on the inner surface of a
conical target. Abrasion removes a portion of
the coating mate rial in each pass. The fines thus
created are carried away in the airstream while
the sand grains settle in an expansion chamber,
as shown in Figure 215.
Foundry sand is usually conveyed by belt convey-
ors and bucket elevators, though pneumatic con-
veyors are used to some extent. Pneumatic con-
veying aids in cooling and fines removal.
The Air Pollution Problem
The air contaminants that may be emitted are
dust from sand breakdown, and smoke and organ-
ic vapors from the decomposition of the core
binders by the hot metal.
-------
Foundry Sand-Handling Equipment
317
FINES TO
DUST COLLECTOR
THROAT
NOZZLE GAP
NOZZLE
-AIR FROM
TURBO-BLOWER
Figure 215. Pneumatic sand scrubber (National Engineering Co., Chicago, III.).
Among the factors that influence emission rates
are size of casting, ratio of metal to sand, met-
al-pouring temperature, temperature of cast-
ing and sand at the shakeout, and handling meth-
ods. These factor shave a great influence on the
magnitude of the air pollution problem. For in-
stance, a steel foundry making large castings,
with a high metal-to-sand ratio requires a very
efficient control system to prevent excessive
emissions, A nonferrous foundry making small
castings with a low metal-to-sand ratio, on the
otherhand, may not require any controls, since
the bulk of the sand remains damp and emissions
are negligible.
Hooding and Ventilation Requirements
The ne°d for ventilation is determined bv the
sarn-p factor? that influence emission rate=. Min-
imum vnlum.es ot ventilation air require^ t~ en-
sure the adequate collection of the air contami-
nants are indicated in the discussion that follows
on the various emission sources.
Shakeout grates
The amount of ventilation air required for a
shakeout grate is determined largely by the type
of hood or enclosure. The more nearly complete
the enclosure, the less the required volume.
When large flasks are handled by an overhead
crane, an enclosing hood cannot be used, and a.
side or lateral hood is used instead. Recom-
mended types of hood are shown in Figure 216
and Figure 217 (upper). Downdraft hoods are not
recommended except for floor-dump type of opej •
ation? where sand and casting? are dropped from
a roller oonvpvor to s gathering conveyor below
the floor l»vel (Manual ot Exhaust Hood Designs.
)• An excessive exhaust volume i= require-i
to achieve adequate ,-on ••••»•<•>; ir> a downdrafi boon
because the indraft t-ploHfr i° working
-------
318
METALLURGICAL EQUIPMENT
.— MOVABLE PANELS 10 SECURE
/ DESIRED DISTRIBUTION
\ /'// I CHANNEL IRON GUABO OPTIONAL TOP
•—3/2 L-rV?—//H i -~
VELOCITY THROUGH OPENINGS
700-1.000 fpm (.
TAKE OFF
ARRANGE LENGTH
/ '} OF SLING CHAIN
TC CLEAR HOOD
MINIM PRACTICAL CLEARANCE -
SIDE-DRAFT HOOD
DUCT VELOCITY = 3,500 fpm MINIMUM
ENTRY LOSS =- I 78 SLOT »p»0 25 DUCT vp
^XT> 'DIKING OPENINGS
/, A**l KEEP AS SMIL AS
y0A^ POSSIBLE-^
/ A
MOLDS IN
HERE—^
SHAREOUT
•CASTINGS
OUT HERE
ENCLOSING HOOD
PROVIDES BEST CONTROL KITH LEAST VOLUME
DUCT VELOCITY = 3,500 fp«l MINIMUM
ENTRY LOSS = 0 25 vp
Figure 216. Foundry shakeout (Committee on In-
dustrial Ventilation, 1960),
BLANK WALL IN THIS POSITION IS
ALMOST AS GOOD AS DOUBLE HOOD
RIGIDLY BRACED
DOUBLE SIDE DRAFT
PROPORTIONS SAME AS SINGLE SIDE DRAFT HOOD EXCEPT FQR OVERHANG
PLENUM CHAMBER AND SLOTS FULL
LENGTH OF SHAKEOUT IN TUNNEL
PROVIDE PLENUM-
CLEANDUTS
DOHNDRAFT HOOD
SLATS SIZED FOR 1,000 TO 2,000 fp*
DUCT VELOCITY = 4,000 (pro MINIMUM
SIZE 0 FOR 1,00(1 fpm OR LESS
ENTRY LOSS = I Jfl SLOT vp PLUS FITTINGS
FOR COOL CASTINGS ONLY
DIFFICULT TO PREVENT PLUGGING OR EXCESS
FINES REMOVAL
SLOT
HOPPER
Figure 217. Foundry shakeout (Committee on In-
dustrial Ventilation, I960).
the thermal bouyancy caused by the hot sand and
casting. The indraft velocity is lowest where it
is needed most--at the center of the grate. The
exhaust volume requirements for the different
types of hood are shown in Table 90. Shakeout
hoppers should be exhausted with quantities of
about 10 percent of the total exhaust volume listed
in this table.
Other sand-handling equipment
Recommended ventilation volumes and hooding
procedure for bucket elevators and beli; convey-
ors are given in Figure 218; for sand screens, in
Figure 219; and for mixer-mullers, in Figure
220. The ventilation requirement for rotary
coolers is 400 cfm per square foot of open area.
For crushers the requirement varies from 500 to
1,000 cfm per square foot of enclosure opening.
Air Pollution Control Equipment
The most important contaminant to be collected
is dust, though smoke is sometimes intense enough
to constitute a problem. Organic vapors and gas-
es are usually not emitted in sufficient quantities
to be bothersome. The collectors usually used
are baghouses and scrubbers.
A "baghouse in good condition collects all the dust
and most of the smoke. A scrubber of moderate-
ly good efficiency collects the bulk of the dust, but
the very fine dust and the smoke are not collected
and in many cases leave a distinctly visible plume,
sufficient to violate some control regulations. A
baghouse, therefore, is the preferred collector
when the maximum control measures are desired
When only the shakeout is vented to a separate
collector, there maybe sufficient moisture in the
gases in some cases to cause condensation and
consequent blinding of the bags in a baghouse.
When, however, all the equipment in a sand-
handling system is served by a single exhaust
system, ample ambient air is drawn into the sys-
tem, to preclude any moisture problem in the bag-
house. The filtering velocity for this type of
service should not exceed 3 fpm. Cotton sateen
cloth is adequate for this service. Anoncompart-
mented-type baghouse is adequate for most job
shop foundries. For continuous-production found-
ries, a compartmented baghouse with automatic
bag-shaking mechanisms gives the most trouble-
free performance.
-------
Foundry Sand-Handling Equipment
319
Table 90. EXHAUST VOLUME REQUIREMENTS FOR
DIFFERENT TYPES OF HOOD
VENTILATING SHAKEOUT GRATES
Type of hood
Exhaust requirement
Hot castings
Cool castings
Enclosing
Enclosed two sides and
1/3 of top area
Side hood (as shown or
equivalent)
Double side hood
Downdraft
200 cfm/ft2 of open-
ing. At least 200
cfm/ft of grate a,rea
300 cfm/ft2 of grate
area
400 to 500 cfm/ft2 of
grate area
400 cfm/ft2 of grate
area
600 cfm/ft2 of grate
area
Not recommended
200 cfm/ft2 of open-
ing. At least 150
cfm/ft2 of grate area
275 cfm/ft2 of grate
area
350 to 400 cfm/ft2 of
grate area
300 cfm/ft2 of grate
area
200 to 250 cfm/ft2 of
grate area
aChoose higher values when (1) castings are very hot, (2) sand-to-
metal ratio is low, (3) crossdrafts are high.
ALTERNATE EXHAUST POINT
FOR ELEVATOR HEAD
ADDITIONAL VENTILATION
TO HOPPER, BIN, OR SCREEN
ADDITIONAL VENTILATION HERE
TO SUIT OPERATION
FOR CASING ONLY
0 , 100 Cfn/ft2 CASING CROSS
SECTION
DUC1 VELOCITY » 3 500 fpm MINIMUM
ENTRY LOSS - 1 0 vp OR CALCULATE
FROM INDIVIDUAL LOSSES
TAKE-OFF AT TOP FOR HOT MATE
RIALS AT TOP AND BOTTOM IF
ELEVATOR IS OVER 30 It HIGH
OTHERHISE OPTIONAL
ADDITIONAL VENTILATION HERE
AS PER BELT TRANSFER
BELT SPEED VOLUME
LESS THAN 200 fpm 350 cfm/ft OF BELT HIOTH
NOT LESS THAN 15(1 off/ft
OF OPENINt
OVER 200 fpm-
500 cfm/ft OF BELT IIDTH
NOT LESS THAN 200 cfm/ft
OF OPENING
Figure 218. Bucket elevator ventilation (Com
mittee on Industrial Ventilation, 1960).
KIN SLOPE
FEED
CANVAS CONNECTION IF DESIRED
COMPLETE ENCLOSURE
SCREEN
OVERSIZE
FUT DEU SCREEN
0 = 200 cfm !t2 THROUGH HOOD OPENINGS BUT NOT LESS THAN 50 cfm ft2
SCREEN *REA NO INCREASE FOR MULTIPLE DECKS
DUCT mOCITV - 3 500 fpm MINIMUM
ENTRY LOSS = 0 5 »p
COMPLETE
ENCLOSURE
SCREEN —
45° MM SLOPE
z__nv
FEED
HOPPER
ERSIZE
CYLINDRICAL SCREEN
0 = 100 cfn/ft2 CIRCULAR CROSS SECTION OF
SCREEN. AT LEAST «0 dm/It" OF EN-
CLOSURE OPENING
DUCT VELOCITY = 3,500 fpm MINIMUM
ENTRY LOSS = SEE "FLAT DECK SCREEN"
Figure 219. Screens (Committee on Industrial
Ventilation, 1960).
-------
320
METALLURGICAL, EQUIPMENT
INSULATION OR STRIP HEATERS
HAY BE REQUIRED TO -PREVENT
CONDENSATION IN DUCT IF STEAM
IS 6IVIN OFF
-BAFFLE
0 = 150 cfn/ft2 TOROU6H ALL OPENItlSS
BUT NOT LESS THAN
HIXER 014HETER
ft
4
6
8
10
EXHAUST
cfft
GSO
900
1,200
1.500
NOTE OTHER TYPES OF MIXERS. ENCLOSE
AS MUCH AS POSSIBLE AND PRO-
VIDE 150 cfm-ft? DF SEMAINING
OPENINGS
DUCT VELOCITY = 3,500 fpm MINIMUM
ENTRY LOSS : 0.25 vp
Figure 220. Mixer and muller hood (Committee on
Industrial Ventilation, 1960).
HEAT TREATING SYSTEMS
Heat treating involves the carefully controlled
heating and cooling of solid metals and alloys for
eff ectin-g certain desired changes in their physical
properties. At elevated temperatures, various
phase changes such as grain growth, recrystal-
lization, and diffusion or migration of atoms take
place in solid metals and alloys. If sufficient
time is allowed at the elevated temperature, the
process goes on until equilibrium is reached and
some stable form of the metal or alloy is obtained.
If, however, because of sudden and abrupt cooling,
time is not sxifficient to achieve equilibrium at the
elevated temperature, then some intermediate or
metastable form of the metal or alloys is obtained.
The tendency to assume a stable form is always
present and metals and alloys in a metastable form
can be made to approach their stable form as close-
ly as desired simply by reheating. The widely
differing properties that can be imparted to solid
metals and alloys in their stable and metastable
forms give purpose to the whole process of heat
treating.
In general, the methods used to heat treat both fer-
rous and nonf err ous metals are fundamentally sim-
ilar. These methods include hardening, quenching,
annealing, tempering, normalizing ferrous metals,
and refining g rain of nonferrous metals. Also in-
cluded in the caregory of heat treating a re the var -
IOUF methods of c.a se hardening steels by carburifc-
ing, rvaniding. nitriding, flame hardening, induc-
tion hardening, "ar-bonitriding, si!iconizing, and
so forth.
HEAT TREATING EQUIPMENT
Furnaces or ovens, atmospheTe generators, and
quench tanks or spray tanks are representative
of the equipment -used for heat treating.
Furnaces for heat treating are of all sizes and
shapes depending upon the temperature needed
andupon the dimensions and the number of pieces
tobetreated. A furnace maybe designed to oper-
ate continuously or "batchwise. The controls may
be either automatic or manual. These furnaces
are known by descriptive names such as box,
oven, pit, pot, rotary, tunnel, muffle, and others.
Regardless of the name, they all have the follow-
ing features in common: A steel outer shell, a
refractory lining, a combiastion or heating sys-
tem, and a heavy door (either cast iron or re-
inforced steel with refractory lining) that may be
opened from the top, the front, or from both the
front and the back.
Atmosphere generators are used to supply a con-
trolled environment inside the heat treating cham-
ber of the furnace. The atmosphere needed may
be either oxidizing, reducing, or neutral depend-
ing upon the particular metal or alloy undergoing
heat treatment and upon the final physical proper-
ties desired in the metal or alloy after treatment.
An atmosphere can be provided that will protect
the surface of the metal during heat treatment so
that subsequent cleaning and buffing of the part is
minimized, or one can be provided that will cause
the surface of the metal to be alloyed by diffusion
with certain selected elements in order to alter
the physical properties of the metal.
Quench tanks may be as simple as a tub of water
or as elaborate as a -well-engineered vessel
equipped-with properly designed means to circu-
late the quenching fluid and maintain the fluid at
the correct temperature. The part to be quenched
is either immersed into the fluid or is subjected
to a spray that is dashed against the part so that
no air or steam bubbles can remain attached to
the hot metal and thereby cause soft spots. The
fluid used for quenching may be •water, oil. mol-
ten salt, liquid air, brine solution, and so forth.
The purpose of quenching is to retain some meta-
stable form of an alloy (pure metals are not af-
fected by quenching) by rapidly cooling the alloy
to some temperature below the transformation
temperature.
The Air Pollution Problem
The heat treating process is currently regarded
as only a minor source of air pollution. Nonethe-
less, air pollutants that may be emitted from
-------
Heat Treating Systems
321
heat treating operations, and their origin, are
as follows:
1. Smoke and products of incomplete combus-
tion arising from, the improper operation of
a ga.s- or oil-fired combustion system;
2. vapors and fumes emanating from the
tilization of organic material on the metal
parts being heat treated;
3. oilmists andfumes issuing from oil quench-
ing baths (if water-soluble oils are used, the
fumes -will be a combination of steam and oil
mist);
4. saltfumes emittedfrom molten salt pot fur-
naces;
5. gases, produced by atmosphere generators,
used in the heat treating chamber of muffle
furnaces. (Insignificant amounts occasional-
ly leak out from some furnace openings that
cannot be sealed, but somewhat larger amounts
get into the surrounding air during purging
and also during loading and unloading oper-
ations. )
Hooding and Ventilation Requirements
Hooding and ventilation systems designed for heat
treating processes should be based on the rate
at which the hot, contaminated air is delivered
to the receiving face of the exhaust hood. To
prevent the hot, contaminated air from spilling
out around the edges of the exhaust hood, the rate
at which the exhaust system draws in air must in
all cases exceed the rate at which the hot, con-
taminated air is delivered to the exhaust system.
In the general case, a canopy hood mounted about
3 or 4 feet above a hot body has an excellent
chance of capturing all the hot, contaminated air
rising by convection from the hot body. The face
area of a canopy hood such as this should be
slightly larger than the maximum cross-sectional
area of the hot body. In order to avoid the need
for excessive exhaust capacity, it is advisable
not to oversize the canopy hood face area.
If a canopy hood is mounted too high above the
hot body, the column of hot, contaminated air is
influenced by turbulence, and the column becomes
more and more dilute by mixing with the surround-
ing air. Consequently the exhaust capacity must
be sufficient to handle this entire volume of diluted,
contaminated1 air. This is an inefficient way to
collect hot, contaminated air.
Many variations of canopy hoods are used because
of the many types of heat treating furnaces em-
ployed. Lateral-type hoods are also used. Gen-
eral features of design of hoods for these hot pro-
cesses are discussed in Chapter 3.
Air Pollution Control Equipment
The following methods effectively prevent and
control emissions resulting from heat treating
operations.
1. Proper selection of furnace burners andfuels
along -with observance of correct operating
procedures -will eliminate smoke and prod-
ucts of incomplete combustion as a source of
air pollution. (See Chapter 9.)
2. Removal of organic material adhering to metal
parts to be heat treated by either steam clean-
ing or solvent degreasing before heat treat-
ing -will eliminate this source of air pollu-
tion.
3. Mists and fumes issuing from oil quenching
baths canbe greatly reducedby selecting ap-
propriate oils and by adequate cooling of the
oil.
4. Baghouses are a satisfactory method of con-
trolling salt fumes from molten salt pots.
Particle sizes of fumes are usually between
0. 2 and 2 microns but may vary from this
range depending upon local factors such as
temperature, humidity, turbulence, and ag-
glomeration tendencies of the effluent. The
fumes are slightly hygroscopic and corro-
sive; therefore, operation of the baghouse
must be continuous to prevent blinding and
deterioration of the bag cloth and corrosion
of the metal structure. Acrylic-treated or-
lonis a satisfactory bag cloth because of its
chemical ana thermal resistance and its gen-
eral physical stability. Filtering velocities
should not be greater than 3 fpm. With these
design features, collection efficiencies ex-
ceeding 95 percent are normally achieved.
5. Flame curtains placed at the open ends of
continuous heat treating furnaces are effec-
tive in the control of any escaping, combus-
tible gases used for controlling the atmo-
sphere inside the furnace.
-------
CHAPTER 7
MECHANICAL EQUIPMENT
HOT-MIX ASPHALT PAVING BATCH PLANTS
JOHN A. DANIELSON, Senior Air
Pollution Engineer
ROY S. BROWN, JR. , Air Pollution Engineer
CONCRETE-BATCHING PLANTS
EDWIN J. VINCENT, Intermediate
Air Pollution Engineer
JOHN L. MC GINNITY, Intermediate
Air Pollution Engineer*
CEMENT-HANDLING EQUIPMENT
EDWIN J. VINCENT, Intermediate
Air Pollution Engineer
ROCK AND GRAVEL AGGREGATE PLANTS
EDWIN J. VINCENT, Intermediate
Air Pollution Engineer
MINERAL WOOL FURNACES
JOHN L. SPINKS, Air Pollution Engineer
PERLITE-EXPANDING FURNACES
EDWIN J. VINCENT, Intermediate
Air Pollution Engineer
FEED AND GRAIN MILLS
WILLIAM H. DONNELLY, Air Pollution Engineer
PNEUMATIC CONVEYING EQUIPMENT
EDWIN J. VINCENT, Intermediate
Air Pollution Engineer
DRIERS
EDWIN J. VINCENT, Intermediate
Air Pollution Engineer
JOHN L. MC GINNITY, Intermediate
Air Pollution Engineer*
WOODWORKING EQUIPMENT
ROBERT GOLDBERG, Air Pollution Engineer'
EDWARD HIGGINS, Air Pollution Engineer*
RUBBER-COMPOUNDING EQUIPMENT
JOSEPH D'IMPERIO, Air Pollution Engineer
ASPHALT ROOFING FELT-SATURATORS
SANFORD M. WEISS, Senior
Air Pollution Engineer
SOLVENT DEGREASERS
ROBERT T. WALSH, Senior
Air Pollution Engineer
SURFACE-COATING OPERATIONS
SANFORD M. WEISS, Senior
Air Pollution Engineer
PIPE-COATING EQUIPMENT
HARRY E. CHATFIELD, Air Pollution Engineer
DRY CLEANING EQUIPMENT
WILLIAM C. BAILOR, Air Pollution Engineer
ABRASIVE BLAST CLEANING
EDWIN J. VINCENT, Intermediate
Air Pollution Engineer
ZINC-GALVANIZING EQUIPMENT
GEORGE THOMAS, Intermediate
Air Pollution Engineer
*Now with National Center for Air Pollution Control, Public Health Service, U. S. Department of Health,
Education, and Welfare, Cincinnati, Ohio.
JNow with the Public Health Service, U.S. Department of Health, Education, and Welfare, St. Glenville,
Illinois.
$Now with New York-New Jersey Air Pollution Abatement Activity, National Center for Air Pollution Control,
Public Health Service, U.S. Department of Health, Education, and Welfare, Raritan Depot, Metuchen, New
Jersey.
-------
CHAPTER 7
MECHANICAL EQUIPMENT
HOT-MIX ASPHALT PAVING
BATCH PLANTS
INTRODUCTION
Hot-mix asphalt paving consists of a combina-
tion of aggregates* uniformly mixed and coated
•with asphalt cement. An asphalt batch plant is
usedto heat, mix, and combine the aggregate and
asphalt in the proper proportions to give the de-
sir-edpaving mix. After the material is mixed, it
is transported to the paving site and spread as a
loosely compacted layer with a uniformly smooth
surface. While still hot, the material is compacted
and densified by heavymotor-driven rollers to pro-
duce a smooth, we 11-compacted course.
Asphalt paving mixes may be produced from a -wide
range of aggregate combinations, each having par-
ticular characteristics and suited to specific de-
signandconstruction uses. Aside from.the amount
and grade of asphalt cement used, the principal
characteristics of the mix are determined by the
relative amounts of:
Coarse aggregate (retained on No. 8-mesh sieve),
fine aggregate (passing No. 8-mesh sieve), and
mineral dust (passing No. 200-mesh sieve).
The aggregate composition may vary from a coarse-
textured mix having a predominance of coarse ag-
gregate to a fine-textured mix having a predomi-
nance of fine aggregate. The Asphalt Institute
(1957)classifieslhot-mix asphalt paving according
to the relative amounts of coarse aggregate, fine
aggregate, and mineral dust. The general limits
for each mix type are shown in Table 91. The com-
positions used within each mix type are shown in
Tables 92 and 93.
Raw Materials Used
Aggregates of all sizes up to 2-1/2 inches are used
inhot-mix asphalt paving. The coarse aggregates
usually consist of crushed stone, crushed slag,
crushed gravel, or combinations thereof, or of
material such as decomposed granite naturally
occurring in a fractured condition, or of a highly
: Aggregate is a term used to describe the solid mineral load-bearing
constituents of asphalt paving such as sand particles and fragments
of stone, gravel, and so forth.
angular natural aggregate with a pitted or rough
surface texture. The fine aggregates usually con-
sist of natural sand and msiy contain added materi-^
als such as crushed stone, slag, or gravel. All
aggregates mustbe free from, coatings of clay, silt,
or other objectionable matter and should not con-
tain clay particles or other fine materials. The
aggregate must also meet tests for soundness
(ASTM designation C88) and wearability (ASTM
designation C131).
Mineral filler is used in some types of paving. It
usually consists of finely ground particles of crushed
rock, limestone, hydrated lime, Portland cement,
or other nonplastic mineral matter. A minimum
of 6 5 per cent of this material must pass a 200-mesh
sieve. Another name for mineral filler is mineral
dust.
Asphalt cement is used in amounts of 3 to 12 per-
cent by weight and is made from refined petroleum.
It is a solid at ambient temperature but is usually
usedas a liquid at 275° to 325°F. One property
measurement used in selecting an asphalt cement
is the "penetration" as determined by ASTM Method
D5. The most common penetration grades used in
asphalt paving are 60 to 70, 85 to 100, and 120 to
150. The grade used depends upon the type of ag-
gregate, the paving use, and the climatic condi-
tions.
Basic Equipment
A typical hot-mix asphalt paving batch plant usu-
ally consists of an oil- or gas-fired rotary drier,
a screening and classifying system, weigh boxes
for asphalt cement and aggregate, a mixer, and
the necessary conveying equipment consisting of
bucket elevators and belt conveyors. Equipment
for the storage of sand, gravel, asphalt cement,
and fuel oil is provided in most plants. Heaters
for the asphalt cement and fuel oil tanks are also
used.
Plant Operation
Plants varyin size. The majority in Los Angeles
County produce 4, 000-pound batches and have pro-
duction rates of 100 to 150 tons of asphalt paving
mixperhour. Some of the newer plants are 6, 000-
pound batch size and are capable of producing 150
to 250 tons per hour.
325
-------
326
MECHANICAL EQUIPMENT
Table 91. CLASSIFICATION OF HOT-MIX ASPHALT PAVING
(The Asphalt Institute, 1957)
Type
I
II
III
IV
V
VI
VII
VIII
Paving mix
designation
Description
Mac adam
Open graded
Coarse graded
Dense graded
Fine graded
Stone sheet
Sand sheet
Fine sheet
Ma ximum si /e
aggregate
normally used
Surlace and
leveling
mixes
3/8 to 3/4 in.
1/2 to 3/4 in.
1/2 to 1 in.
1/2 to 3/4 in.
1/2 to 3/4 in.
3/8 in.
No. 4
15a *, e, bin Je r,
and le\ ehng
mixes
2-1/2 in.
3/4 to l-l/2in. u 10
3/4 to l-l/2in. '• 30
y,
o
1 to 1-1/2 in. ^ 40
in
3/4 - H 60
H
,t* • £ 70
5/4 in. sj
O
on
Ul
Z
5/8 in. ^ 90
No. 4 100
Aggregate t ombinations
"n MINERAL DUS 1' (PASSING NO. 200SIEV1-
0 5 0 15 i
i milrt] 'iirii
u*'"* 'i
'^\
BASE, ' \^
»4t
\
BINDER.
AND
LEVELING
^ MIXES
.AGGREGATE PROP
..\
\
• ^
-^LR
\^Vj
AND ''"'f
|, LEVEL'li
/I ING
?/.% MIXEJ
•,
THIS ARE
1' NORMA1
ECOMMEN
— FOR PA1
°\
Lr<\\ CONS
' '",'• ' \
'''/ '''
,
.',
3RTIONS
A
T Y
CEMENT
1'RUC ITON
\
•— rv^^-j-o^J3c^ o
ooooocoooo o
To COARSE AGGREGATE (RETAINED NO. 3 SIEVE)
aCritical zone - Dust contents in this region should
not be used without a substantial background of ex-
perience with such mixes and/or suitable justifica-
tion by laboratory design tests.
^Intermediate zone - Dust contents in this region
sometimes used in surface and leveling mixes as
well as in base and binder mixes.
0 5 10 15
% MINERAL DUST (PASSING NO. 200 SIEVE)
Figure 221 is a flow diagram of a typical plant.
Aggregate is usually conveyed from the storage
bins to the rotary drier by means of a belt con-
veyor and bucket elevator. The drier is usually
either oil-or gas-fired and heats the aggregate to
temperatures ranging from 250° to 35~0°F. The
dried aggregate is conveyed by a bucket elevator
to the screening equipment where it is classified
and dumped into elevated storage bins. Selected
amounts of the proper size aggregate are dropped
from the storage bins to the weigh hopper. The
weighed aggregate is then dropped into the mixer
along with hot asphalt cement. The batch is mixed
and then dumped into waiting trucks for transporta-
tion to the paving site. Mineral filler can be added
directlytothe weigh hopper by means of an auxil-
iary bucket elevator and screw conveyor,,
Fine dust in the combustion gases from the rotary
drier is partially recovered in a precleaner and
discharged continuously into the hot dried aggre-
gate leaving the drier.
THE AIR POLLUTION PROBLEM
The largest source of dust emissions is the rotary
drier. Other sources are thehot aggregate bucket
elevator, the vibrating screens, the hot aggregate
bins, the aggregate weigh hopper, and the mixer.
Rotary drier emissions up to 6, 700 pounds per
hour have been measured, as shown in Table 94.
In one plant, 2, 000 pounds of dust per hour was
collected from the discharge of the secondary dust
sources, thatis, the vibrating screens, hot aggre-
gate bins, the aggregate weigh hopper, and the
mixer.
-------
Hot-Mix Asphalt Paving Batch Plants
327
Table 92. COMPILATION OF SUGGESTED MIX COMPOSITIONS (The Asphalt Institute, 1957)
Vflx ! Aggregate by si
type 1-1/2 in. + 1 in. 3/4 in 1/2 in. ; 3/8 in No 4
Mix | II a , 100 40
seal lib,' l 70 to 100 I 20
II b ; 00 70 to 1 00 ' 20
ii e i ; :oo ; 70 to 100 45 to i-,\ 20
S Ilia 100 75 to 100 13^
u , HI b ; 100 75 to 100 60 to 80 ' 35
r , Iv a ' 100 80 to 100 i 5s
, IV b 100 30to;oO'70to90!^0
a IV ( , 100 30 to 100 ' 60 to 80 48
i , V a ! iOO 85 to 100 c5
,, V b3 , 100 , 35 to 100 ! ro
o 85
o 40
o 40
.e in mix, %
No. 8 No 16 j No. 30
5
5
o 4o| 5
o 55
o ^5
o 75
20
20
35
o 7035
o 65
o 80
o 80
VI a , 100 ' 85 to 100
VI 1>* .00 8^ to 100
VII f ' ',00 8s to 100
.'III a ! IOO
35
50
50
65
65
o 20 ;
No. 50
o 20 ;
o 20
No. 100
o 35 10 to 22 6
o 35 10 to 22 6
o 50 18 to 29
o 50 18 to 29
13
13
o 50 19 to 30 13
o 65 37 o52 25 to 40 18
o 65 ' 37 o 52 25 to 40 I 1 8
o 80 • 50 o 70 : 3^ to 60
o 80 ' 47 o n8 30 to 5s
ft,
/O
80 to 95 70 o 89 55 to 80 30
95 to ,00 85 o 98 ^ 70 to 95 40
o 16 j 4
o 16 1 4
o 23
o 23
o 23
o 30
o 30
o 48
3
8
7
10
10
15
o 40 10
o 60 110
o 12
o 12
o 16
o 16
o 15
o 20
No. 200
0 to 4
0 to 4
0 to 4
0 to 4
2 to 8
2 to 8
4 to 10
4 to 10
0 to 8
Asphalt,
4. 0
4. 0
4.0
i. 0
3. 0
3. 5
3.5
3. 5
3 to 10 1 4. 0
o 20 ;3 to 10
o 30 16 to 12
o 25 13 to 8
o 35 14 to 14
o 76 ]20 to 40 is to 16
4. 0
4. 5
4. 5
6. 0
o 5 0
o 5.0
o 5.0
o 6 0
o 6. 0
o 6. 0
o 7. 0
o 7. 0
o 7. 0
o 7. 5
0 7. 5
o8.5
o8.5
o 11. 0
6. 5 to 12. 0
"May be used '"or base wnere coarse aggregate IP not < ronomicaliv available.
rable 93. COMPILATION OF SUGGESTED MIX COMPOSITIONS (The Asphalt Institute, 1957)
Mix
type
2-1/2 in.
1-1/2 in
1 in.
3/4 in.
1/2 in. ! 3/8 in
1
No. 4 No. 8
1
No. 16
No. 30
No. 50
No. 100
No. 200
Asphalt,
%
TIr
II d
III b
III c
III d
IV t
III b
V b->
VI ba
I a
II d
II e
III d
III e
IV d
IOO
35 to 70
100
100
100
100
100
100
100
70 to ,0(3
100
75 to :oo
80 to 100
100
70 to 100
100
100
75 to 100
80 to 100
100
100
100
0 to :5
70 to iOO
"0 .0 SO
- ;., 100
tl to 85
^0 to 90
70 to 100 j 45 to 75
1 35 o 60
75 to 100 60 o 85
75 to 100 60 o 85
I 45 o 70
- 60 o 80
75 to 100 | 60 to 85
85 to 100
8^ to 100
35 to 60
25 to 50
45 to 70
', 10 to 65
55 to 75
20 to 40
15 to 35
55 to 55
30 ti» 50
30 to 50
IS to 65
Beveling
3^ to 55
65 to 80
Base
15 to 35
10 to 30
30 to 50
30 to 50
45 to 62
5 to 20
5 to 20
20 to 35
20 to :>5
20 to 35
35 to 50
20 to 35
50 to 65
65 to 80
01) L
5 to c.0
5 t ) 20
20 15 3^
20 t > 35
35 to 50
37 to C2
47 to 68
10 to 22
5 to 20
5 to 20
19 to 30
10 to 22
25 to 40
30 to 55
5 to 20
5 to 20
19 to 30
6 to 16
3 to 12
3 to 12
13 to 23
6 to 16
18 to 30
20 to 40
3 to 12
3 to 12
13 to 23
4 to 12
2 to 8
2 to 8
7 to 15
4 to 12
10 to 20
10 to 25
2 to 8
2 to 8
7 to 15
0 to 4
0 to 4
2 to 8
0 to 4
0 to 4
0 to 8
2 to 8
3 to 10
3 to 8
0 to 3
0 to 4
0 to 4
0 to 4
0 to 4
0 to 8
3. 0 to 6. 0
3. 0 to 6. 0
3. 0 to 6. 0
3. 0 to fc. 0
3. 0 to 6. 0
3. 5 to 7. 0
3. 0 to 6. 0
4. 0 to 7. 5
4. 5 to 8.5
3. 0 to 4. 5
3. 0 to 6. 0
3. 0 to 6. 0
3. 0 to 6. 0
3. 0 to 6. 0
3. 5 to 7. 0
aMay he used ior base where
CYCLONE
COLO AGGREGATE
BUCKET ELEVATOR
SAND AND
AGGREGATE
BINS
HOT AGGREGATE-
BUCKET ELEVATOR
'guie 221. Flow diagram of a typical hot-mix asphalt paving batch plant.
-------
3Z8
MECHANICAL EQUIPMENT
Table 94. DUST AND FUME DISCHARGE FROM ASPHALT BATCH PLANTS
Test No.
Batch plant data
Mixer capacity, Ib
Process weight, Ib/hr
Drier fuel
Type of mix
Aggregate feed to drier^wt%
+ 10 mesh
-10 to +100 mesh
-100 to +200 mesh
-200 mesh
Dust and fume data
Gas volume, scfm
Gas temperature, °F
Dust loading, Ib/hr
Dust loading, grains/scf
Sieve analysis of dust, wt %
+ 100 mesh
-100 to +200 mesh
-200 mesh
Particle size of -200 mesh
0 to 5 |a, wt %
5 to 10 n, wt %
10 to 20 (i, wt %
20 to 50 |JL, wt %
> 50 \j., wt %
C-426
6,000
364, 000
Oil, PS300
City street, surface
70. 8
24.7
1.7
2.8
Vent linea
2,800
215
2, 000
81.8
4.3
6.5
89.2
19.3
20. 4
21.0
25. 1
14.2
Drier
21, 000
180
6,700
37. 2
17. 0
25.2
57.8
10. 1
11. 0
11. 0
21.4
46. 5
C-537
6,000
346,000
Oil, PS300
Highway, surface
68. 1
28.9
1.4
1.6
Vent linea
3,715
200
740
23. 29
0.5
4.6
94.9
18.8
27. 6
40.4
12. 1
1. 1
Drier
22, 050
430
4,720
24. 98
18.9
32. 2
48.9
9.2
12. 3
22.7
49.3
6.5
aVent line serves hot elevator, screens, bin, weigh hopper, and mixer.
Drier dust emissions increase with air mass ve-
locity, increasing rate of rotation,and feed rate,
but are independent of drier slope (Friedman and
Marshall, 1949). Particle size distribution of the
drier feed has an appreciable effect on the dis-
charge of dust. Tests show that about 55 percent
of the minus 200-mesh fraction in the drier feed
can be lost in processing. The dust emissions
from the secondary sources vary with the amount
of fine material in the feed and the mechanical con-
dition of the equipment. Table 94 and Figure 222
give results of source tests of two typical plants.
Particle size of the dust emissions and of the ag-
gregate feed to the drier are also shown.
HOODING AND VENTILATION REQUIREMENTS
Dust pickup must be provided at all the sources of
dust discharge. Total ventilation requirements
vary according to the size of the plant. For a
6, 000-pound-per-batchplant, 22,000 scfm is typ-
ical, of which 18,000 to 19,000 scfm is allotted
for use in controlling the drier emissions. The
top end of the drier must be closely hooded to pro-
vide for exhaust of the products of combustion and
entrained dxist. A ring-type hood located between
the sta.tiona.Ty portion of the burner housing and
the drier provides satisfactory pickup at the lower
end of the drier. An indraft velocity of 200 fpm
should be provided at the annular opening bet-ween
the circumference of the drier and the ring-type
hood.
The secondary dust sources, that is, the elevator,
vibrating screens, hot aggregate bins, weigh hop-
per, andmixer, are all totally enclosed, and hence,
no separate hooding is required. Dust collection
is provided by connecting this equipment through
branch ducting to the main exhaust system. Ap-
proximately 3, 000 to 3, 500 scfm will adequately
ventilate these secondary sources.
AIR POLLUTION CONTROL EQUIPMENT
Primary dust collection equipment usually consists
of a cyclone. Twin or multiple cyclones are also
used. The catch of the primary dust collector
is returned to the hot bucket elevator where it con-
tinues on with the main bulk of the drier aggregate.
The air discharge from the primary dust collector
is ducted to the final dust collection system.
Two principal types of final control equipment have
evolved from the many types employed over the
years: The multiple centrifugal-type spray cham-
ber (Figure 223) and the baffled-type spray tower
-------
Hot-Mix Asphalt Paving Batch Plants
.329
VENT LINE
2,000 Ib/hr
FROM DRYER
6,700 Ib/hr
TEST C-426
2,620 Ib/hr
CYCLONE
EFFICIENCY
= 90. 8*
TO ATMOSPHERE
! 25.5 Ib/hr *
RETURN TO HOT ELEVATOR I
6,080 Ib/hr
WATER AND MUD
2.595 Ib/hr
DRY DUST
VENT LINE
TEST C-537
118 Ib/h
MULTIPLE
CYCLONE
EFFICIENCY
= 92.2%
TO ATMOSPHERE
! 33 5 Ib/hr*
MULTIPLE
CENTRIFUGAL
SCRUBBER
EFFICIENCY
= 71.3*
WATER AND MUD
5,344 Ib/hr
84.5 Ib/hr
DRY DUST
Figure 222. Test data on air pollution control equipment serving two hot-mix asphalt
paving plants (vent line serves screens, hot bins, weigh hopper, and mixer).
Figure 223. Typical multiple centrifugal-type scrubber
serving a 4,000-pound-batch-capacity hot-mix asphalt
paving plant.
(Figure 224). The multiple centrifugal-type spray
chamber has proved the more efficient. It consists
of two or more internally fluted, cylindrical spray
chambers in which the dust-laden gases are ad-
mitted tangentially athigh velocities. These cham-
bers are each about the same size, that is, 6 feet
in diameter by 15 feet in length, if two chambers
are used, and 6 feet in diameter by 9 or 12 feet in
length if three chambers are used. Usually 7 to
12 spraynozzles are evenly spaced within each
chamber. The total water rate to the nozzles is
usually about 70 to 250 gpm at 50 to 100 psi. In
the baffled-type spray tower, there have been many
variations and designs, but fundamentally, each
consists of a chamber that is baffled to force the
gases to travel in a sinuous path, •which encourages
impingement of the dust particles against the sides
of the chamber and the baffles. Water spraynoz-
zles are located among the baffles, and the water
rate through the spray nozzles is usually between
100 to 300 gpm at 50 to 100 psi.
n both types of scrubber the water may be either
fresh or recirculated. Settling pits or concrete
tanks of sufficient capacity to allow most of the
collected dust to settle out of the water are re-
-------
330
MECHANICAL EQUIPMENT
Figure 224. Typical baffled-type spray tower serving
a 3,000-pound-batch-capacity hot-mix asphalt paving
plant (Griffith Company, Wilmington, Calif.).
The effect of aggregate fines feed rate on stack
emissions at constant water-gas ratio (an average
value for test considered) is shown in Figure 225
for multiple centrifugal-type scrubbers and baffled-
tower scrubbers. Stack emissions increase lin-
early with an increase in the amoun-t of minus 200-
mesh material processed. These losses can be
greatly reduced by using a clean or washed sand.
The required fines content of the hot-mix asphalt
paving is then obtained by adding mineral filler
directly to the plant -weigh hopper by means of an
auxiliary bucket elevator and screw conveyor.
Most asphalt paving batch plants burn natural gas.
When gas is not available, and if permitted by law,
a heavy fuel oil (U. S. Grade No. 6 or heavier) is
usually substituted. Dust emissions to the atmo-
sphere from plants with air pollution control de-
vices were found to be about 5. 1 pounds per hour
greater when the drier was fired with oil than they
were when the drier was fired with natural gas.
The difference is believed to represent particulate
matter residing in, or formed by, the fuel oil,
rather than additional dust from the drier. Simi-
larly, the burning of heavy fuel oils in other kinds
of combustion equipment results in greater emis-
sions of particulate matter.
The amount of water fad to the scrubber is a very
important consideration. The spray nozzles should
quired with a system using recirculated water.
The scrubber catch is usually hauled away and
discarded. It is usually unsuitable for use as min-
eral filler in the paving mix because it contains
organic matter and clay particles. The recircu-
lated water may become acidic and corrosive, de-
pending upon the amount of sulfur in the drier fuel,
and must then be treated with chemicals to protect
the scrubber and stack from corrosion. Caustic
soda and lime have been used successfully for this
purpose.
Variables Affecting Scrubber Emissions
In a recent study (Ingels et al. , I960), many source
tests (see Table 95) on asphalt paving plants in Los
Angeles County were used to correlate the major
variables affecting stack losses. Significant var-
iables include the aggregate fines feed rate (the
minus 200-mesh fraction), the type of fuel fired
inthe drier, the scrubber's •water-gas ratio,* and
the type of scrubber used. Other, less important
variables were also revealed in the study.
*The water-gas ratio is defined as the total quantity of water
sprayed in gallons per 1,000 scf of effluent gas.
10
0 2,000 4,000 6,000 8,000 10.000
QUANTITY Of FINES (MINUS 200 MESH) IN DRYER FEED, Ib/hr
Figure 225. Effect of aggregate fines feed rate on
stack emissions at average water-gas ratio (Ingeis
et al., 1960).
-------
Hot-Mix Asphalt Paving Batch Plants
331
Table 95. TEST DATA FROM HOT-MIX ASPHALT PAVING PLANTS CONTROLLED BY SCRUBBERS
Test No.
C-357
C-82
C-379
C-355
C-372B
C-372A
C-369
C-393
C-354
C-185
Scrubber
inlet dust
loading,
Ib/hr
940
427
4, 110
2, 170
Stack
emission,
Ib/hr
20. 7
35. 6
37. 1
47. 0
121 ! 19.2
76
352
4,260
1,640
10. 0
24, 4
26.9
27.8
21.3
C-173 -- 31.0
1
C-379
C-337
2
C-234
C-426
C-417
C-425
3
i 33.5
3,850
30. 3
305 13.6
Aggregate
fines rate, a
Ib / hr
9, 550
Water-gas
ratio,
gal/1, 000 scf
6. 62
4,460 3.94
8, 350
14, 000
2, 290
2, 840
4, 750
4, 050
6. 38
Overall
scrubber
efficiency,
wt %
97. 8
91.6
99. 1
6.81 97.8
10. 99
11.11
5.41
12. 01
6, 370 6. 10
' i
5, 220
8. 850
7, 520
6, 500
2, 510
!9. 40
20. 40
11. 01
5.92
11.11
Type
of
scrubber
C
C
C
C
84.2 C
86. 8
C
93. 0 C
99. 3 T
-p
98. 7
Type
of
drier
fuel
Oil
Oil
Oil
Oil
Oil
Gas
Oil
Oil
Production
rate,
tons/hr
183.9
96. 9
174. 0
209. 1
142. 9
158. 0
] 13. 0
Gas
effluent
volume,
scfm
23, 100
19,800
26, 200
25, 700
18,200
18, 000
16, 100
92. 3 | 19, 500
Oil 118.4 7,720
T Oil 137.8
T
T
99.2
95. 5
21.11 3, 730 j 7. 28
372
21.2
2,620 25.5
560 39.9
485
--
C-385 j 212
C-433
C-422(l)
C-422(2)
C-418
Averages
266
--
32.9
25.5
17. 5
11.0
26.6
37.0
3,400 30.8
1
26.7
2, 530
:o, 200
3, 050
2, 890
6,590
4, 890
5, 960
7, 140
3, 340
9, 350
5. 70 94. 3
^.75
2. 94
4. 26
6.60
4.56
3. 12
4. 90
j. 02
3. 90
C
C
T
Oil
184. 2
Oil 144.6
18, 700
17, 000
23, 700
Gas 191.3 28,300
Oil
114.6 24,300
Gas 124.4
T Gas 42. 0
99. 0 C
92.8
93. 2
--
91. 7
95. 8
--
__
(_•
C
C
C
C
C
C
99. 1 T
i
Oil
Oil
Oil
Gas
Oil
Gas
Oil
Oil
15,900
17,200
182. 0 1 22, 000
138.9
131. 4
131. 7
174. 3
114. 5
198. 0
152. 0
Oil H6.5
5, 900 j i 94. 9
24,600
18, 000
18, 200
20,000
19,600
21, 000
22,200
17, 100
Q_uantity of fines (minus 200 me
°C ~ Multiple centrifugal-type
T - Baffled tower scrubber.
sh) in dryer feed.
spray chamber.
oe located so as to cover the moving gas stream
adequately with fine spray. Sufficient water should
beused to cool the gases below the dew point. One
typical scrubber tested had an inlet gas at 200 °F
with 16. 8 percent water vapor content by volume,
and an outlet gas at 131 °F with 16. 3 percent water
vapor and saturated. The temperature at the gas
outlet of efficient scrubbers rareiy exceeds 140°F,,
and the gas is usually saturated -with -water vapor.
Figure 226 shows the effect of the scrubber1 s water-
gas ratio on dust emissions-with the aggregate fines
feed rate held constant (an average value for the
test considered). Efficient scrubbers use -water
at rates of 6 to 10 gallons per 1,000 standard cubic
feet of gas. The efficiency falls off rapidly at -water
rates less than 6 gallons per 1,000 scf of gas. At
rates of more than 10 gallons per 1, 000 scf of gas,
the efficiency still increases, but at a lesser rate.
Curves are presented in Figures 227 and 228 from
which probable stack emissions can be predicted
for oil- and gas-fired plants with either multiple
centrifugal or baffled tower scrubbers. These
curves present emissions for various scrubbers'
water-gas ratios and aggregate fines rates. Emis-
sion predictions from these curves are accurate
onlyfor plants of the type and design already dis-
cussed.
The operation of the rotary drier is also an im-
portant variable. Dust emissions increase with an
increase of air mass velocity through the drier.
Obviously then, care should be taken to operate the
drier without a great amount of excess air. This
care effects fuel economy and reduces dust emis-
sions from the drier.
The firing rate of the drier is determined by the
amount of moisture in the aggregate and by the re-
quired hot aggregate temperature. The greater
the aggregate moisture content, the greater the
firing rate and the resulting dust emissions to the
atmosphere. In some plants, the increase inmois-
ture content of the flue gases may increase the ef-
ficiency of the scrubber sufficiently to offset the
increase in dust emissions from the drier.
Scrubber efficiencies also vary according to the
degree of precleaning done by the primary dust
collector. Tests (such as those presented in Table
95) have shown that overall efficiency of the pre-
-------
332
MECHANICAL EQUIPMENT
to
0 2 4 6 8 10 12 14 16 18 20
SCRUBBER HATER-GAS RATIO, gal/1,000 scf
Figure 226. Effect of scrubber's water-gas ratio on
stack emissions at average aggregate fines feed rate
in the drier feed (Ingels et al., 1960).
cleaner and final collector varies only slightly with
large variations in precleaner efficiency. Plants
•withless effective cyclone precleaning had, on the
average, larger particles entering the scrubber,
and consequently, show greater scrubber collec-
tion efficiencies. The principal advantage of an
efficient precleaner is that the valuable fines col-
lected can be discharged directly to the hot elevator
for use in the paving mix. Furthermore, less dust
is discharged to the scrubber, where more trouble-
some dust disposal problems are encountered.
Collection Efficiencies Attained
Collection efficiencies of cyclonic-type precleaners
vary from approximately 70 to 90 percent on an
overall weight basis. Scrubber efficiencies vary-
ing from 85 to nearly 100 percent have been found.
Overall collection efficiencies usually vary between
95 and 100 percent.
4,000 8,000 12,000 16,000
QUANTITY OF FINES (MINUS 200 MESH) IN DRYER FEED, Ib/hr
Figure 227. Emission prediction curves for multiple centrifugal
scrubbers serving asphaltic concrete plants (Ingels et al., 196
-------
Hot-Mix Asphalt Paving Batch Plants
333
60
- 40
20
10
4,000
30
20
__ __ 10
1,000 12,000 16,000 0 4,000
QUANTITY OF FINES (MINUS 200 MESH) IN DRYER FEED, Ib/hr
12,000
16,000
Figure 228. Emission prediction curves for baffled tower scrubbers serving asphaltic
concrete plants (Ingels et al., 1960).
Collection efficiencies of a simple cyclone and a
multiple cyclone for various particle sizes are
shown in Table 96. Multiple cyclones achieve high
efficiencies for particle sizes down to 5 microns,
whereas single cyclones are very inefficient for
particle sizes below 20 microns. The particle size
data from this table are plotted on log-probability
paper in Figure 229. This figure also shows the
particle size distribution of the scrubber outlet.
Other data on this installation have already been
presented in Figure 222, test C-537.
Cost of Air Pollution Control Equipment
The cost of control equipment varies according to
the manufacturer, location, and type of installa-
tion. A typical system consisting of a 12-foot-
diameter cyclone, atwin- or triple-tube scrubber
complete with ductwork, water pump, and fan will
cost about $25, 000 for a plant capable of handling
6,000-pound batches. If a multiple cyclone and
provisions for recirculating the -water are added,
the total cost may approach $45, 000.
Table 96. COLLECTION EFFICIENCY DATA FOR A CYCLONE AND
A MULTIPLE CYCLONE SERVING A HOT-MIX PAVING PLANT
Dust
particle
size, |JL
0 to 5
5 to 10
10 to 20
20 to 50
50+
Dust loading
Ib/hr
Test C-537
cyclone
Inlet,
6.2
9.4
13.8
22.9
47. 7
5,463
Outlet,
19. 3
31.9
31.6
15. 1
2. 1
1,525
Efficiency,
13.3
5.4
36. 1
81.6
98.8
72. 1%
Test C-537a
multiple cyclone
Inlet,
19.3
31.9
31.6
15. 1
2. 1
1,525
Outlet,
57. 0
34. 0
8.8
9.2
--
118.3
Efficiency,
77. 1
91. 7
97.8
99.9
100.0
92. 2%
See Table 94, test C-537 for plant operating data.
-------
334
MECHANICAL EQUIPMENT
O CYCLONE INLET
CYCLONE OUTLET-MULTIPLE CYCLONE INLET
* MULTIPLE CYCLONE OUTLET--SCRUBBER INLET
T SCRUBBER OUTLET
0.01
0.1 0.2 0.5 1
5 10 20 30 40 SO 60 70 BO 90 95 98 99 99.599.8 99.9 100
PERCENT LESS THAN GIVEN PARTICLE SIZE, microns
Figure 229. Plot of particle size of dust at the inlet and outlet of a cyclone and
multiple cyclone from test C-537.
CONCRETE-BATCHING PLANTS
Concrete-batching plants store, convey, measure,
and discharge the ingredients for making concrete
to mixing or transportation equipment. One type
is used to charge sand, aggregate, cement, and
water to transit-mix trucks, which mix the batch
en route to the site where the concrete is to be
poured; this operation is known as "wet batching. "
Another type is used to charge the sand, aggre-
gate, and cement to flat bed trucks, which trans-
port the batch to paving machines where water is
added and mixing takes place; this operation is
known as "dry batching. " A third type employs
the use of a central mix plant, from which wet con-
crete is delivered to the pouring site in open dump
trucks.
WET-CONCRETE-BATCHING PLANTS
In a typical-wet-concrete-batching plant, sand and
aggregates are elevated by belt conveyor or, clam
shell crane, or bucket elevator to overhead storage
bins. Cement from bottom-discharge hopper trucks
is conveyed to an elevated storage silo. Sand and
aggregates for a batch are weighed by successive
additions from the overhead bins to a weigh hopper.
Cement is delivered by a screw conveyor from the
silotoa separate weigh hopper. The -weighed ag-
gregates and cement are dropped into a gathering
hopper and flow into the receiving hopper to the
transit-mix truck. At the same time, the required
amount of water is injected into the flowing stream
of solids. Details and variations of this general
procedure will be discussed later.
The Air Pollution Problem
Dust, the air contaminant from, wet-concrete-batch-
ing, results from the material used. Sand and ag-
gregates for concrete production come directly
from a rock and gravel plant where they are washed
to remove silt and clay-like minerals. They thus
-------
Concrete-Batching Plants
335
arrive at the batch plant in a moist condition and
hence do not usually present a dust problem. When,
however, lightweight aggregates are used, they do
pose a problem. These materials are formed by
thermal expansion of certain minerals. They leave
the aggregate plant very dry and create consider-
able dust when handled. The simplest way to deal
•withthis problem is to wet each load of aggregate
thoroughly before it is dumped from the delivery
truck. Attempts to spray the aggregate as it is
being dumped have had very limited effectiveness.
If, therefore, wet or damp aggregate is used,
practically all the dust generated from concrete-
batching operations originates from the cement.
Particle size distribution and other characteristics
of the dust vary according to the grade of cement.
A range of 10 to 20 percent by weight of particles
of 5-micron size or less is typical for the various
grades of cement. Bulk density ranges from 50
to 65 pounds per cubic foot of cement. Table 97
shows additional characteristics of three common
grades of cement.
Air Pollution Control Equipment
Cement-receiving and storage system
Atypical cement-receiving and storage system is
shown in Figure Z30. The receiving hopper is at
or below ground level. If it is designed to fit the
canvas discharge tube of the hopper truck, little
or no dust is emitted at this point. After a brief
initial puff of dust, the hopper fills completely and
the cement flows from the truck without any free
fall. Cement elevators are either the vertical-
screw type or the enclosed-bucket type. Neither
emits any dust if in good condition. The cement
silo must be vented to allow the air displaced by
the cement to escape. Unless this vent is filtered,
a significant amount of dust escapes.
Figure 230 shows one type of filter. It consists
of a cloth tube with a stack and weather cap for pro-
tection. The pulley arrangement allows it to be
shaken from the ground so that the accumulated
layer of dust on the inside of the cloth tube can be
periodically removed. The cloth's area should be
sufficient to provide a filtering velocity of 3 fpm,
based upon the displaced air rate.
Table 97. CHARACTERISTICS OF THREE
GRADES OF CEMENT
Distribution, u
0 to 5
5 to 10
Cement, wt %
Grade I
13.2
15. 1
10 to 20 25.7
20 to 40
40 to 50
50 to 66
66 to 99
99 to 250
250 (60 mesh)
Bulk density,
lb/ft3
Specific gravity,
g/cm3 at 82°F
29. 0
7.0
5. 0
4. 0
1. 0
0
54. 0
3. 3
Grade II
9.6
16.6
18. 8
36. b
10.4
6.0
2. 0
0
0
51. 5
3. 3
Grade III
21. 8
22. 5
26.7
23.6
5.4
0
0
0
A
62.0
3. 3
Many concrete batch plants now receive cement
pneumatically from trucks equipped with compres-
sors and pneumatic delivery tubes. In these plants,
a single filtered vent used for the gravity filling of
cementhas proved inadequate, and other methods
of control are required. In this pneumatic delivery,
the volume of conveying air is approximately 350
cfm during most of the loading cycle and increases
to 700 cfm at the end of the cycle.
To control this volume of air, it is best to install
a small conventional cotton sateen baghouse with
a filtering area of 3 fpm (approximately 200 square
feet of cloth area) to vent the cement silo. The
baghouse should be equipped with a blower to re-
lieve the pressure built up within the silo. A
mechanical shaking mechanism also should be
provided to prevent cement from blinding the fil-
ter cloth of the baghouse.
Cement dust can be emitted from several points.
The receiving hopper, the elevator, and the silo
are the points of possible emission from the ce-
ment-receiving station. Other points of possible
dust emission are the cement weigh hopper, the
gathering hopper, and the mixer.
Another less expensive type of control device is
to mount a bank of approximately four simple fil-
tered vents atop the silo. The filtering area
should not exceed 7 fpm, giving an area of ap-
proximately 1 00 square feet for the 700 cfm of air
encountered at the end of the cycle. The filter
design must include a shaking mechanism to pre-
vent blinding of the filter cloth. The major dis-
-------
336
MECHANICAL EQUIPMENT
Figure 230. Cement-receiving and storage system.
advantage of using a bank of several simple filter
vents as just described is the possibility of pres-
sure build-up within the silo. If, for some reason,
the filter should become blinded, there is danger
of rupturing the silo. Therefore, proper mainte-
nance and regular inspection of the filter are
necessary.
weigh hopper is filled at a. fairly rapid rate, and
the displaced air entrains a significant amount of
dust. This dust may be controlled by venting the
displaced air back to the cernent silo or by install-
ing a filtered vent on the weigh hopper as described
for cement silos.
The vent should be of adequate size to provide a
filtering velocity of about 3 fpm, based upon the
cement's volumetric filling rate. For example,
if aweighhopper is filled at the rate of 1, 500 pounds
in 1 /2 minute, and the density of cement is 94 pounds
per cubic foot, the displaced air rate equals 1-, 500/
(94)(0. 5), or 32 cfrn. The required cloth area would
then be 32/3 or 10. 7 square feet.
Gathering hoppers
The dropping of a batch from the •weigh hopper to
the mixer can cause cement dust emissions from
several points. In the loading of transit-mix trucks,
a gathering hopper is usually used to control the
flow of the materials. Dust can be emitted from
the gathering hopper, the truck's receiving hopper,
andthemixer. The design and location of the gath-
ering hopper can do much to minimize dust emis-
sions. The hopper should make a good fit with the
truck receiving hopper, and its vertical position
should be adjustable. Figure 231 illustrates a de-
sign that has been used successfully in minimizing
dust emissions. Compressed-air cylinders raise
and lower the gathering hopper to accommodate
trucks of varying heights. A steel plate with a
foam rubber backing is attached to the bottom of
the gathering hopper and is lowered until it rests
on the top of the truck's receiving hopper. Water
for the mix is introduced through a jacket around
the discharge spout of the gathering hopper and
forms a dust-reducing curtain.
Where baghouses are used to control other larger
cement dust sources such as those existing in a
dry-concrete-batching plant or in a central mix
plant, then the cement silo can easily be vented
to the same baghouse.
Discharge of the cement hopper into the center of
the aggregate stream, and choke feed between the
weigh hopper and the gathering hopper suppress
dust emissions from the top of the gathering hopper.
Cement weigh hopper
The cement weigh hopper may be a compartment
in the aggregate weigh hopper or it may be a sep-
arate -weigh hopper. Cement is usually delivered
from the silo to the -weigh hopper by an enclosed
screw conveyor. To permit accurate weighing, a
flexible connectionbetweenthe screw conveyor and
weigh hopper is necessary. A canvas shroud is
usually used, and if properly installed and main-
tained, prevents dust emissions at this point. The
DRY-CONCRETE-BATCHING PLANTS
Dry-concrete-batching plants are used in road con-
struction -work. Because of advances in free-way
construction in recent years, plants such as these
are located in metropolitan areas, often in resi-
dential zones. The plants are portable, that is,
theymustbe designed to be moved easily from one
location to another. This is, of course, a factor
in the design of the air pollution control equipment.
-------
Concrete-Batching Plants
337
•HEIGH HOPPERS-
X
GATHERING
HOPPER
WATER
COMPRESSED-AIR
CYLINDERS
METAL PLATE
FOAM RUBBER
TRANSIT-MIX
TRUCK
Figure 231. An adjustable gathering hopper.
The Air Pollution Problem
Dry batching poses amuchmore difficult dust con-
trol problem than wet batching does. Since most
plants that do dry batching also do wet batching,
the gathering hopper must be set high enough to
accommodate transit-mix trucks. Since the re-
ceiving hopper of most transit-mix trucks is sev-
eral feet higher than the top of the flat-bed trucks
used in dry batching, there is a long free fall of
material •when a dry batch is dropped. This pro-
duces a considerable amount of dust, sufficient to
violate most codes that have an opacity limitation
applicable to this type of operation.
From an air pollution standpoint, the dust to be
collected has characteristics similar to those of
the cement dust already discussed for •wet-con-
crete-batching plants. In dry batching, however,
volumes of dust created are considerably greater
because: (1) The amount of concrete batched is
large, (2) no -water is used, and (3) the batches
are dropped rapidly into the waiting trucks to con-
serve time.
Hooding and Ventilation Requirements
A local exhaust system with an efficient dust col-
lector is required to control a dry batching plant
adequately. This is a difficult operation to hood
without interfering with the truck1 s movement or
the batch operator's view. The truckbed is usually
divided into several compartments, a batch being
dropped into each compartment. This necessitates
repeated spotting of each truck under the direction
of the batch operator; hence he must be able to see
the truck at the drop point. A canopy-type hood
just large enough to cover one compartment at a
time provides effective dust pickup and affords
adequate visibility. Figure Z32 shows a closeup
view of a hood of this type. The sides are made
of sheets of heavy rubber to permit contact with
the truckbed without damage. This hood is mounted
on rails to permit it to be withdrawn to allow wet
batching into transit-mix trucks.
The exhaust volume required to collect the dust
varies -with the shape and position of the hoods.
With reasonably goodhooding, the required volume
is approximately 6, 000 to 7, 000 cfm.
Air Pollution Control Equipment
Abaghouse is the most suitable type of dust collec-
tor for this service. Scrubbers have been used,but
they havebeen plagued with difficulties such as low
collection efficiency, plugged spray nozzles, cor-
rosion, and waste-water disposal problems. A
baghouse for this service should have a. filtering
velocity of 3 fpm. It may be of the intermittent
shaking type, since sufficient opportunities for
stopping the exhauster for bag shaking are usually
available. Figure 233 is an overall view of a typ-
ically controlled dry batching plant with the bag-
house shown on the left. The drop area tunnel is
enclosed on the sides and partially on the ends.
Dust created by truck movement
Inmany instances the greatest source of dust from
the operation of a concrete batch plant is that cre-
ated by the trucks entering and leaving the plant
area. If possible, the yard and access roads should
be paved or oiled, or if this is not feasible, they
should be watered frequently enough to suppress
the dust.
CENTRAL MIX PLANTS
The central mix plant, as shown in Figure 234, is
being used more and more extensively by the con-
crete industry in the Los Angeles area. In a cen-
tral batch operation, concrete is mixed in a sta-
tionary mixer, discharged into a dump truck, and
transported in a wet mixed condition to the pour-
ing site.
The handling of aggregate and cement at these plants
is similar to that at the other concrete batch plants.
-------
338
MECHANICAL EQUIPMENT
Figure 232. Closeup of hood for controlling dry hatching: (left) Hood in place, (right) hood
in retracted position (Graham Bros. E! Monte Caiif.).
Figure 234. Overall view of a central mix concrete-
batching plant controlled by a baghuuse (Griffith
Co., Los Angeles, Calif.).
Figure 233. Overall view of wet- and dry-concrete-
batching plant and baghouse located at a California
Freeway project (Guy F. Atkinson Co , Long Beach
Calif.).
Sand, aggregate, cement, and water are all weighed
or metered as in a wet-concrete-batching plant and
discharged through an enclosed system into the
mixer.
The Air Pollution Problem
From an air pollution control standpoint, this type
of .operation is preferable to dry batching. The
dust is more easily captured at the batch plant, and
further, there is no generation of dust at the pour-
ing site. The operation is also preferable to wet
batching because designing control equipment for
a stationary mixer is easier than it is for a transit-
mix truck-loading area.
-------
Cement-Handling Equipment
339
Hooding and Ventilation Requirements
Effective control at the discharge end of the mixer
is a function of good hood design and adequate ven-
tilation air. Ahydraulically operated, swing-away,
cone-shaped hood, as shown in Figure 235, is nor-
mally used with a 2-inch clearance between the
hood and the mixer. This installation employs a
mixer with a capacity of 8 cubic yards. The dis-
charge opening of the mixer is 40 inches in di-
ameter. Ventilation air was found to be 2,500 cfm,
For a hood of this type, indraft face velocities
should be between 1, 000 and 1, SOOfpm. Velocities
such as these are required for handling the air dis-
charged from the mixer, which is displaced air and
inspirated air from the aggregate and cement fall-
ing into the mixer.
Air Pollution Control Equipment
A baghouse, such as is shown in Figure 235, is
required to collect the dust emissions. A filter-
ing velocity of 3 fpm is adequate. Other baghouse
features are similar to those previously discussed
for dry-concrete-batching plants.
Figure 235. Hood for central mix plant: (top) In re-
tracted position, (bottom) in closed position (Griffith
Co., Los Angeles, Calif.).
CEMENT-HANDLING EQUIPMENT
Equipment used in handling cement includes hop-
pers, bins, screw conveyors, elevators., and pneu-
matic conveying equipment. The equipment to be
discussed in this section is that involved in the
operation of a bulk cement plant, which receives,
stores, transships, or bags cement. Its main
purpose is usually to transfer cement from one
type of carrier to another, such as from railway
cars to trucks or ships.
THE AIR POLLUTION PROBLEM
In the handling of cement, a dust problem can occur
if the proper equipment or hooding is not used. A
well-designed system should create little air pol-
lution. Sources of emissions include the storage
and receiving bins, elevators, screw conveyors,
a,nd the mobile conveyances.
Characteristics of cement dust have been discussed
in the section on wet-concrete-batching plants.
HOODING AND VENTILATION REQUIREMENTS
Receiving Hoppers
Railway cars are usually unloaded into an under-
ground hopper similar to the one described for
trucks in the preceding section. The canvas tube
is usually, however, permanently attached to the
receiving hopper and is attached by a flange to the
discharge spout of the hopper car. When flanges
fit properly, emissions from equipment such as
this are usually negligible.
Storage and Receiving Bins
Bins filled by bucket elevators must be ventilated
at a rate equal to the maximum volumetric filling
rate plus 200 fpm indraft at all openings. The area
of openings is usually very small. Since most bulk
-------
340
MECHANICAL EQUIPMENT
plants have a number of bins, a regular exhaust
system with a dust collector provides a more prac-
tical solution than the silo filter vents do that were
described for concrete batch plants. Bins filled by
pneumatic conveyors must, of course, use a dust
collector to filter the conveying air. Gravity-fed
bins and bins filled by bucket elevators can use
individual filter vents if desired.
Elevators and Screw Conveyors
Bucket elevators used for cement service are al-
ways totally enclosed. Ventilation must be pro-
vided for the bin into which it discharges. Since
elevators are nearly always fed by a screw con-
veyor that makes a dust-tight fit at the feed end, no
additional ventilation is usually required. Another
type of conveyor used for cement service is a ver-
tical screw conveyor. These, of course, cause no
dust emissions as long as they have no leaks. Hori-
zontal screw conveyors are frequently fed or dis-
charged through canvas tubes or shrouds. These
must be checked regularly for tears or leaks.
Hopper Truck and Car Loading
Hopper trucks and railroad cars are usually filled
from overhead bins and silos. The amount of dust
emitted is sufficient to cause a nuisance in almost
any location. Figure 236 shows a type of hood and
loading spout that permits these emissions to be
collected with a minimum amount of air. The ven-
tilation rate is the same as for bins, the displaced
air rate plus 200 fpm through all openings. If the
hood is designed to make a close fit with the hatch
AIR CONVEYOR
FROM CEMENT BIN
opening, the open spaces are very small and the
required exhaust volume is small. The hood is
attached to the telescoping cement discharge spout
in such a way that it can be raised and lowered
when hopper trucks are changed.
AIR POLLUTION CONTROL EQUIPMENT
Abaghouse has been found to be the most satisfac-
tory dust collector for handling the ventilation
points described. All sources are normally ducted
to a single baghouse. Cotton sateen cloth with a
filtering velocity of 3 fpm is adequate. Dacron
cloth, which provides longer wearing qualities but
is more expensive, can also be used.
ROCK AND GRAVEL AGGREGATE PLANTS
Rock and gravel plants supply sand and variously
sized aggregates for the construction and paving
industries. The sources of most aggregates used
in Los Angeles County are the gravel beds in the
San Fernando and San Gabriel valleys. The pro-
cessing of the gravel consists of screening out the
usable sizes and crushing the oversize into various
size ranges. Asimplified flow diagram for a typ-
ical plant is shown in Figure 237. Incoming mate-
rial is routed through a jaw crusher, which is set
to act upon rocks larger than about 6 inches and
to pass smaller sizes. The product from this
crusher is screened into sizes smaller and larger
than 1-1/2 to 2 inches, the undersize going to a
screening plant, and the oversize to the crushing
plant. These next crushers are of the cone or gy-
ratory type, as shown in Figure 238. In a large
plant, two or three primary crushers are used in
parallel followed by two to five secondary crush-
ers in parallel.
Figure 236. Hood for truck-loading station.
Figure 237. Simplified flow diagram of a typical
rock gravel plant.
-------
Rock and Gravel Aggregate Plants
341
SPIDEB ARK
SHIELD
RUGGED
Ttio uat
SPIDER
Figure 238. Gyratory crusher (AlMs-Chalmers
Manufacturing Company, Milwaukee, His.).
THE AIR POLLUTION PROBLEM
The sand and rock, as it comes from the pit, is
usually moist enough to remain nondusting through-
out the sand- anduncrushed-rock-screening stages.
When the pit material is not sufficiently moist, it
must be -wetted before it leaves the pit. As the
larger rocks are crushed, dry surfaces are ex-
posed and airborne dust can be created.
Aninventory of sources of dust emissions usually
begins with the first crusher and continues with
the conveyor transfer points to and including the
succeeding crushers. Here the rock is more fine-
ly ground, and dust emissions become greater. As
the process continues, dust emissions are again
prevalent from sources at conveyor transfer points
and at the final screens.
guide to the amount of ventilation air required
(Committee on Industrial Ventilation, I960):
1. Conveyor transfer points--350 cfm per foot of
belt width for speeds of less than 200 fpm; 500
cfm per foot of belt width for belt speeds over
200 fpmj
2. bucket elevators --tight casing required with
a ventilation rate of 100 cfm per square foot
of casing cross section;
3. vibrating screens --50 cfm per square foot of
screen area, no increase for multiple decks.
AIR POLLUTION CONTROL EQUIPMENT
One method of suppressing the dust emissions con-
sists of using water to keep the materials moist
at all stages of processing; the other, of using a
local exhaust system and a dust collector to collect
the dust from, all sources.
If the use of water can be tolerated, then water
can be added -with spray nozzles, usually at the
crusher locations and the shaker screens. Fig-
ure 239 shows nozzle arrangements for control of
emissions from the outlet of the crushers, Figure
240, nozzle arrangements at the inlet to the shak-
er screens. The amount of -water to be used can
best be determined by trial under normal operat-
ing conditions. Water quantities vary with crush-
er size, crusher setting, feed rate, type of feed,
and initial moisture content of the feed.
•HARD RUBBER SHIELD
T10 FLAT ATOIIZIHG TYPE
SPRAT NOZZLES ONE EACH
END OF RUBBER SHIELD
T»0 FLAT ATOMIZING TTPE
SPRAY NOZZLES ONE EACH
ENO OF RUBBER SHIELD
Figure 239. Nozzle arrangement for control of
dust emissions upon discharge of crusher.
HOODING AND VENTILATION REQUIREMENTS
The points that require hooding and ventilation are
the crusher discharge points, all elevator and belt
conveyor transfer points, and all screens.
All these dust sources should be enclosed as near-
ly completely as possible and a minimum indraft
velocity of 200 fpm should be maintained through
all open areas. The following rules are also a
Adding water in the described manner tends to
cause blinding of the finest size screens used in
the screening plants, which thereby reduces their
capacity. It also greatly reduces the amount of
rock dust that can be recovered, since most of the
finest particles adhere to larger particles. Since
rock dust is in considerable demand, some oper-
ators prefer to keep the crushed material dry and
collect the airborne dust with a local exhaust sys-
tem.
-------
342
MECHANICAL EQUIPMENT
DE-ANGLE CONE-TYPE
Figure 240. Nozzle arrangement for control of dust
emissions from the inlet to the shaker screens.
The preferred dust collector device is a baghouse.
Standard cotton sateenbags canbe used at a filter-
ing velocity of 3 fpm. For large plants that main-
tain continuous operation, compartmented collec-
tors are required to allow for bag shaking. Most
plants, however, have shutdown periods of suffi-
cient frequency to allow the use of a noncompart-
mented collector. Virtually 1 00 percent collection
canbe achieved, and as mentioned previously, the
dust is a salable product.
A combination of a dry centrifugal collector and a
•wet scrubber is sometimes used. In this case,
only the centrifugal device collects material in a
salable form. A centrifugal collector alone would
allow a considerable amount of very fine dust to be
emitted to the atmosphere. A scrubber of good
design is required, therefore, to prevent such
emissions.
MINERAL WOOL FURNACES
INTRODUCTION
The general product classification known as min-
eral wool -was formerly divided into three cate-
gories: Slag wool, rock "wool, and glass wool.
Slag "wool, which was made from iron slag or cop-
per slag, was first successfully manufactured in
England in 1885, after earlier attempts had failed
in the United States (Kirk and Othmer, 1947). The
first manufacture of rock wool (which was made
from natural rock) took place at Alexandria, In-
diana, in 1897. Glass wool (made from glass cul-
let or high silica sand, or both) was later pio-
neered in Newark, Ohio, in 1931.
Today, however, straight slag wool and rock wool
as sxich are no longer manufactured. A combina-
tion of slag and rock constitutes the cupola charge
materials inmore recent times, yielding a product
generally classified as mineral wool, as contrasted
with glass wool.
Mineral -wool is made today in Los Angeles County
with a cupola by using blast furnace slag, silica
rock, and coke (to serve as fuel). It has been pro-
duced here in the past by using a reverberatory
furnace charged with Borax ore tailings, dolomite,
and lime rock heated with natural gas.
Types and Uses of Mineral Wool Products
Mineral wool consists of silicate fibers 5 to 7 mi-
crons in diameter (Allen et al. , 1952) and about
1/2-inch long, and is used mainly for thermal and
acoustical insulation. It has a density of about 6
pounds per cubic foot and is collected initially as
a continuous loose blanket of fibers on a convey-
ingbelt. Itissold, however, as quilt, loose rolls,
industrial felt, batts, or in a granulated form.
Batts are rectangular sections of mineral wool ap-
proximately 4 by 15 by 48 to 60 inches in size.
These sections are covered on top and two sides
with paper, and the bottom is covered with either
an asphalt-coated paper or aluminum foil. Batts
are used for thermal insulation in residential homes
and for many other insulation needs.
Granulated mineral wool, which is handled pneu-
matically, isalsoused for home insulation. Quilt
is normally 60 inches wide and 2 inches thick and
contains the binder agent and paper cover. It is
used primarily for industrial insulation. Loose
rolls, which contain no binder agent and are some-
times enclosed in a fine mesh cover, are used for
applications such as'water heater s and house trail-
ers. Industrial felt consists of wool blanket with
binder agent but -without a paper covering and has
a slightly greater density than that of batts. It is
used for items such as walk-in refrigerators and
industrial ovens.
Mineral Wool Production
The cupola or furnace charge is heated to the mol-
ten state at about 3, 000°F, after which it is fed by
gravity into a device at the receiving end of a large
blowchamber. This device may be a trough-like
arrangement with several drains, or a cup-like
receiver on the end of a revolving arm. The mol-
tenmaterial is atomized by steam and blasted hor-
izontally towards the other end of the blow chamber.
When the cup or spinner device is used, the action
of the steam is assisted by centrifugal force. The
steam atomizes the molten rock into small globules
that develop and trail long, fibrous tails as they
-------
Mineral Wool Furnaces
343
travel towards the other end of the blowchamber.
These fibers reportedly can be drawn mechanically
or spun "without steam, but this process is loreign
to Los Angeles County.
Phenolic resin or a mixture of linseed oil and as-
phalt are examples of binding agents that can be
atomized at the center of the steam ring by a sep-
arate steam jet to act as a binder for the fibers.
Annealing oil can also be steam atomized near the
steam ring to incorporate a quality of resilience
to the fibers that prevents breakage.
Atemperature between 150° and 250°F is main-
tained in the blowchamber. Blowers, which take
suction beneath the wire mesh conveyor belt in the
blowchamber, aid the fibers in settling on the belt.
The mineral wool blanket of fibers is conveyed to
an oven for curing the binding agent. Normally
gas fired, the oven has a temperature of 300° to
500°F.
The mineral wool is next programmed through a
cooler, as shown in the flow diagram in Figure
241. Usually consisting of an enclosure housing
a blower, the cooler reduces the temperature of
the blanket to prevent the asphalt, which is applied
later to the paper cover, from melting.
To make batts, the blanket leaving the cooler is
processed through a multibladed, longitudinal cutter
to separate it into sections of desired widths. Brown
paper and either asphalt-coated paper or alumi-
num foil are then applied to the sections of blanket.
The asphalt-coated paper is passed through a bath
of hot asphalt just before its application to the un-
derside of each section. This asphalt film serves
as a moisture barrier as -well as a bonding agent
against "walls. The paper-covered sections are
cut to desired lengths by a transverse saw, after
which the finished product is packed for storage
and shipment. The two cutters, paper and asphalt
applicators, and conveyor systems are sometimes
referred to collectively as a batt machine.
A granulated-wool production line differs from that
just described in that the mineral wool blanket,
after leaving the blowchamber, is fed to a shredder
for granulation, then to a pelletizer. The pelletizer
serves two functions, namely, to form small 1-
inch-diameter wool pellets and to drop out small
black particles called shot, which form as the mol-
ten slag cools in the blowchamber. A bagging oper-
ation completes the process. Since no binding agent
is required, the curing oven is eliminated.
THE AIR POLLUTION PROBLEM
The major source of emissions is the cupola or
furnace stack. Its discharge consists primarily
of condensed fumes that have volatilized from the
HEAT
EXCHANGER I I
7\
PACKING
AND
STORAGE
T0_
ATMOSPHERE
Figure 241. Flow diagram of mineral wool process.
-------
344
MECHANICAL EQUIPMENT
molten charge, and gase^s such as sulfur oxides
andfluorides. Amounting to as much as 100 pounds
per hour and submicron in size, condensed fumes-
create a considerable amount of visible emissions
andean be a public nuisance. Table 98 shows the
weights of emissions discharged from uncontrolled
cupolas andfxirnaces. A particle size distribution
of the emissions is shown in Table 99.
Another source of air pollution is the blowchamber.
Its emissions (see Table 100) consist of fumes,
oil vapors, binding agent aerosols, and wool fibers.
In terms of weight, a blowchamber may also emit
as much as 100 pounds of particulate matter per
hour at a production rate of 2 tons per hour if the
blowchamber vent is uncontrolled. Approximately
90 percent of these emissions consists of mineral
•wool fibers.
Types of air contaminants from the curing oven
are identical to those from the blowchamber ex-
cept that no metallurgical fumes are involved.
These emissions amountto approximately 8 pounds
per hour at a production rate of 2 tons per hour,
as seen in Table 101, since the amount of wool
fibers discharged is muchless than that for a blow-
chamber. From a visible standpoint, however,
these pollutants may create opacities as high as
70 percent. Emissions from the cooler are only
4 or 5 pounds per hour at a production rate of 2
tons per hour (see Table 102). The asphalt appli-
cator can also be a source of air pollution if the
temperature of the melting or holding pot exceeds
400°F.
HOODING AND VENTILATION REQUIREMENTS
No special hooding arrangements as such are re-
quired in any of the exhaust systems employed in
the control of pollution from mineral wool process-
es. The one possible exception is that canopy hoods
may oe used over the asphalt tanks if the emissions
from these tanks are excessive and are vented to
an air pollution control device.
The ventilation requirements for the various indi-
vidual processes in a mineral wool system are
categorized as follows:
1. Cupolas. Based on test data, exhaust require-
ments can be estimated to be 5, 000 to 7, 000
scfm for a cupola -with a process -weight of
from 4,000 to 4,500 pounds per hour, on the
assumption that no outside cooling air is in-
troduced. The charge door should be kept in
the closed position to obtain maximum benefit
from the capacity of the exhaust fan. A ba-
rometric damper in the line bet-ween the cu-
pola and the blower can be used to control the
amount of gases pulled from the cupola. The
objective is to remove all tuyere air plus an
additional amount of air to maintain a slight
negative pressure above the burden.
2. Reverberatory furnaces. Ventilation require-
ments are about 15,000 to 20,000 cfm (at
600°F) for a furnace sized to produce 1, 500
to 3, 000 pounds of mineral wool per hour. The
heat in these furnace gases can be used in
making steam before filtration.
3. Blowchambers. For a blowchamber with a
size of about 4, 500 cubic feet and -with a ca^
pacity for processing 4, 000 pounds of wool an
hour, the minimum ventilation requirements
are 20, 000 to 25, 000 scfm. All duct takeoffs
must be located at the bottom of the blowcham-
ber beneath the conveyor to create downdraft,
Table 98. DUST AND FUME DISCHARGES FROM MINERAL WOOL
CUPOLAS AND FURNACES
Test No.
Test data
Process wt, Ib/hr
Stack volume, scfm
Stack gas temp, °Fb
Stack emissions, Ib/hr
gr/scf
SO2, mg/scf
Total SO2, %
SO,, mg/scf
CO, %
Cupola
I
3, 525
4, 550
309
49.7
1.28
32.6
0.04
18.5
0.9
3
4,429
4,545
295
45.6
0. 21
-
-
-
-
6A
_
4, 510
314
51. 1
1. 33
-
-
-
-
13
3,625
4, 760
338
29. 0
0. 71
-
-
-
-
Reverberatory
furnace
19a
3,050
2,740
625
7. 3
0. 31
-
-
-
-
aAn estimated 75 percent of the furnace gases -was used for -waste heat purposes
and -was not, therefore, included in the test.
^As measured after cooling, just upstream from control device.
-------
Mineral Wool Furnaces
345
which packs the newly formed wool fibers onto
the conveyor. From this viewpoint, 35, 000
scfm would be more desirable. In addition,
this increased ventilation holds the blowcham-
ber temperature down to tolerable limits, which
determine the type of air pollution control
equipment to be selected. If the plant is pro-
cessing granulated wool instead of batts, down-
draft is less important and satisfactory oper-
Table 99. PARTICLE SIZE ANALYSIS
BY MICROSCOPE OF TWO SAMPLES
TAKEN FROM THE DISCHARGE OF
A MINERAL WOOL CUPOLA FURNACE
Test No. 9A
Size
45
15
7.
1
1
range, \j.
to 75
to 45
5 to 15
to 7.5
Total
count
10
10
40
100
2, 000
Percent
by number
0. 5
0. 5
2. 0
5. 0
92.0
Percent
by wt
75.0
10.0
14.5
0.5
Nil
Tyler screen analysis: Retained on 200 mesh (74 (j.):
33.8%
Retained on 325 mesh (44 (jt):
20. 3%
Retained on pan (44 (i): 49. 9%
Ignition loss: 10%
Test No. 9B
Average particle size, p.
200.
60
40
10
5
1
Total
count
2
8
10
20
100
930
Percent
by number
0. 1
0.4
0. 5
1. 0
5. 0
93. 0
Percent
by wt
85.0
9.5
3. 5
1.08
0. 07
Nil
ation can be achieved •with a 25, 000-scfm ex-
haust system. If a lint cage is used to trap
wool fibers in the discharge gases, frequent
cleaning (four times an hour) of the cage is
imperative for proper ventilation.
Curing ovens. Exhaust requirements for a
2, 500-cubic-foot oven operating at 300° to
500°F and capable of processing 4, 000to6, 000
pounds of mineral wool an hour are about 5, 000
scfm. Sufficient oven gases must be removed
to prevent a pressure buildup so that leakage
does not occur. In sizing the fan, considera-
tion must be given to temperature rises and
possibly also to the introduction of outside
cooling air for proper fan operation, particu-
larly if the oven discharge gases are inciner-
ated.
Coolers. Coolers normally do not require air
pollution control devices. If outside ambient
air is used as the cooling medium, the ventila-.
tion requirements are 10,000 to 20,000 cfm
for a cooler whose area is about 70 square
feet.
Asphalt tanks. If temperature regulators are
successfully used to control emissions, the
ventilation requirements for melting, holding,
and dip tanks will be about 75 cfm for each
square foot of surface area. This value is for
open tanks and for hoods having one open side.
If the melting and holding tanks are closed,
natural-draft stacks may be used.
Table 100. EMISSIONS FROM MINERAL WOOL BLOWCHAMBERS
Test No.
Process wt, Ib/hr
Stack volume, scfm
Stack gas temp, °F
Blowchamber emissions, Ib/hr
Type of control equipment
Dust concentration, gr/scf
Inlet
Outlet
Dust emissions, Ib/hr
Inlet
Outlet
Control efficiency, %
SO2, mg/scf
Total SO2, %
Aldehydes, mg/scf
Total aldehydes, %
Combustibles, %
1
3,525
11, 100
196
9.20
None
0. 097
0.097
9. 20
9.20
-
1.04
0.0013
1.03
0.0036
-
6C
-
17, 200
196
5.02
a
0.034
0.011
5.02
1.62
67.90
-
-
-
-
-
13
3,625
15,760
160
7. 11
None
0. 0526
0.0526
7. 11
7. 11
-
-
-
-
-
-
14
3,525
28, 728
188
98. 21
None
0.399
0.399
98.21
98.21
-
-
-
-
.
-
17
3, 700
19,750
167
_
Lint cage
-
0. 012
-
2.03
_
_
-
.
.
-
25
4, 120
15,400
200
8. 3
Two wet centrifugal
water scrubbers in
parallel
0.063
0.028
8.30
3.60
57
_
_
_
_
-
aThis control equipment consisted of a water scrubber followed in series by an electrical precipitator.
-------
346
MECHANICAL EQUIPMENT
Table 101. EMISSIONS FROM MINERAL WOOL CURING OVENS
Test No.
Process wt, Ib/hr
Stack volume, scfm
Stack gas temp,°F
Oven emissions, Ib/hr
Type of control equipment
Dust concentration, gr/scf
Inlet
Outlet
Dust emissions, Ib/hr
Inlet
Outlet
Control efficiency, %
SO2, mg/scf
Total SO2, %
Aldehydes, mg/scf
Total aldehydes, Ib/hr
Inlet
Outlet
NO,, Ib/hr
Inlet
Outlet
Afterburner temp, °F
1
3, 525
4,740
326
8. 95
None
0. 22
0. 22
8. 95
8.95
-
3. 23
0. 0053
1. 24
-
._
6E
-
6, 130
314
22. 30
b
0.42
0. 083
22. 30
4. 36
81
-
-
-
-
-
_
_
-
13
3, 625
4, 862
353
5.20
None
0. 125
0. 325
5. 2.0
5. 2.0
-
-
-
-
-
_
-
18a
3, 050
1, 642
310
2.27
None
0. 161
0. 161
2. 27
2.27
-
-
-
-
-
-
_
-
-
Z2 !
5, 180 !
8,000 |
200
15. 20
Catalytic
afterburner
0. 221
0. 071
15. 20
4. 90
68
-
-
-
I. 90
0. 90
0. 60
0. 70
840 |
i
24
3, 500
4, 870
270
5
Direct- flame
afterburner
0. 119
0. 032
5
2. 50
50
-
-
-
2. 20
0. 94
0. 15
0. 45
1, 230
aDuring this test the oven was heated with waste heat from a reverberatory furnace. The quantity of
dust emissions appears low as a result of considerable leakage at the oven. Of the particulates col-
lected, 95,4% were volatile or combustibles.
DThis control equipment consisted of a water scrubber followed in series by an electrical precipitator.
Table 102. EMISSIONS FROM MINERAL WOOL COOLERS
Test No.
Process wt, Ib/hr
Stack volume, scfm
Stack gas temp, °F
Cooler emissions, Ib/hr
gr/scf
SO2, mg/scf
Total SO2, %
Aldehydes, mg/scf
Total aldehydes, %
~
3, 525
1, 350
128
0.75
0.047
0.49
0.0006
0.304
0.0009
j_ I
3, 700
8, 500
273
2.55
0. 035
i B
3, 050
16,696
170
3.58
0. 025
I
_
~
-
-
i ->
3, 050
8, 980
288
8. 39
0. 109
-
-
-
-
-------
Mineral Wool Furnaces
347
AIR POLLUTION CONTROL EQUIPMENT
Baghouse Collection and Cupola Air Contaminants
Baghouses have proved to be an effective and re-
liable means of controlling the discharge from
mineral wool cupolas. An installation of this type
is shown in Figure 242. Dacron or Orion bags,
\vhich can withstand temperatures up to 275 °F,
should be used. Of these two synthetic fabrics,
Dacron is now the more common, and features
several advantages over Orion, as discussed in
Chapter 4. Glass fabric bags cannot be used, owing
to the fluorides in the cupola effluent. (Results of
a stack test disclose fluorides in a concentration
9. 85 percent by 'weight in the particulate matter
discharged from a cupola. The life of glass bags
under these conditions is about 1 week. )
Provisions for automatic bag shaking should be in-
cluded in the baghouse design. Sufficient cloth
area should be provided so that the filtering veloc-
ity does not exceed 2. 5 fpm.
Since the discharge temperature of the gas is about
1, 000°F, heat-removing equipment must be used
to prevent damage to the cloth bags. This can be
accomplished with heat exchangers, evaporative
coolers, radiant cooling columns, or by dilution
with ambient air. The cooling device should not
permit the temperature in the baghouse to fall be-
low the dewpoint. Safety devices should be included
to divert the gas stream and thus protect the bag-
house from serious damage in the event of failure
of the cooling system. In some instances it may
also be desirable to include a cyclone or knockout
trap someplace upstream of the baghouse to re-
move large chunks of hot metal that can burn holes
in the bags even after passing through the cooling
system.
The solution to a typical design problem involving
a baghouse and an evaporative cooling system serv-
ing a cupola is described in Chapter 6.
Baghouses should be equally effective in controlling
emissions from reverberatory furnaces . The com-
ments made about cupolas are generally applicable
to these furnaces. Excelsior-packed water scrub-
bers have been tried in Los Angeles County but did
not comply with air pollution statutes relating to
opacity limitations.
Afterburner Control of Curing Oven Air Contominants
The effluent from the curing oven is composed
chiefly of oil and binder particles. These emis-
sions, while not a great contributor to air pollu-
tion in terms of weight, are severe in terms of
opacity. Since they are combustible, a possible
method of control is incineration. This method,
in fact, has proved practical for the mineral •wool
plant.
Generally, afterburners are divided into two cate-
gories, depending upon the method of oxidation.
These are direct-flame and catalytic. Important
considerations for the direct-flame type (see Table
103) are flame contact, residence times, and tem-
perature. The afterburner should be designed so
that a maximum of mixing is obtained with the flame.
The design should also provide sufficiently low gas
stream velocities to achieve a minimum retention
time of 0.3 second. An operating temperature of
1, 200 °F is the minimum requirement for efficient
incineration. Figure 243 shows the effectiveness
of the direct-flame type on curing oven emissions
at different operating temperatures.
Table 103. DATA FOR A MINERAL
WOOL CURING OVEN CONTROLLED BY
A DIRECT-FLAME AFTERBURNER
Oven data
gas fired, conveyorized
Operating temp, 350° to450°F
Heat input, 4 million Btu/hr
Afterburner data
Type, direct flame, gas fired, two-pass
Flame contact device, deflector plate
Heat input, 5 million Btu/hr
Size, 4 ft dia x 9 ft length with 3 ft dia x 1 0 f t length
Insulated retention tube
Gas temp inlet, 27CTF
Operating temp, 1,240°F
Gas velocity, 37 ft/sec
Retention time, 0. 3 sec
Collection efficiency (at 1,230°F)
On particulate matter, 50%
On aldehydes, 59%
On combustibles, 52%
On solvent soluble material, 68%
If a catalytic afterburner is used, the gas stream
must be preheated to about 1, 000°F. Some type
of pr ecleaner must be used to remove the mineral
•wool fibers and thus prevent fouling of the catalytic
elements. Because of this problem, catalytic af-
terburners have not proved very satisfactory for
this service.
Table 101 reflects a comparison of the effective-
ness of both afterburner types as a control device
on mineral wool curing ovens. Electrical precip-
itators have been used as an alternative means
of controlling emissions from mineral wool cur-
ing ovens. The precipitator is, however, preceded
by a water scrubber and high-velocity filter to re-
move the gummy material that would normally foul
the ionizer and plate sections.
-------
348
MECHANICAL EQUIPMENT
Figure 242. Baghouse controlling a mineral wool cupola.
No. 1
Tube side, gas
No. passes, 1
Shell side, cool ing air
No. passes, 4
Air vol, 2,840 scfm
Tube surface, 895 ft2
inlet temp (gas), 650°F
Outlet temp (gas), 440°F
Heat exchanger data:
No. 2
Tube side, gas
No. passes, 1
Shell side, cool ing ai r
No. passes, 3
Air vol, 10,500 cfm
Tube surface, 1,740 ft2
Inlet temp (gas), 400°F
Outlet temp (gas), 275°F
Baghouse data:
Type.pulI through, tubular
FiIter medium, orlon
Filter area. 5,232 ft2
Shaking cycle, 30 minutes
(Automatic, staggered by
compartment)
Tube size, 11^ in. dia x 15^ ft length
Gas temp inlet, 250°F
Collection efficiency, 97%
-------
Mineral Wool Furnaces
34'9
ooo
1 300
1 100 1 200
AFTERBURNER TEMPERATURE °F
Figure 243. Effectiveness of direct-flame afterburner
on curing oven emissions as a function of afterburner
temperature.
Reducing Blowchamber Emissions
If the blowchamber's temperature is maintained
below 175°F to preclude the formation of oil mist,
then the major air pollution problem is posed by
wool fibers. The most practical means of collect-
ing these fibers is an efficient water scrubber, as
shown in Figure 244. If, however, the blowcham-
ber's temperature rises above 250°F, the feasi-
bility of using a water scrubber is diminished. Test
25 shown in Table 100 gives the results of a stack
analysis of two "wet centrifugal water scrubbers
placed in parallel and venting a blowchamber. A
deflector plate at the blowchamber's entrance can
be used to deflect a large portion of the molten shot
and thereby reduce the blow chamber' s temperature
as well as reduce the chance for contact with oil
mists. Water injection at the receiving end of the
blowchamber combined with adequate ventilation
air can further reduce this temperature to 150 °F
or less.
A simple wire -mesh lint cage collects as much as
90 pounds of large pieces of fibrous material per
hour. Constant cleaning of the lint cage is, how-
ever, required; otherwise lack of ventilation results
in a temperature rise in the blowchamber.
Large water content in the blowchamber effluent
precludes classifying the baghouse as a practical
control device for the blowchamber. In addition,
the resin binder would plug the pores of the bags,
resulting in a severe maintenance problem.
Controlling Asphalt Fumes
Asphalt vapors emitted by the asphalt applicator
canbecome a serious source of air pollution if the
Figure 244. Mineral wool blowchamber controlled by an
mertial-type water scrubber.
asphalt's temperature is permitted to exceed 400 °F.
The simplest and most economical method of re-
ducing these emissions to the atmosphere is to
control the temperature. The temperature can
sometimes be held to a maximum of 325°F by
proper asphalt selection, thermostatic control,
and use of a holding pot separate from the melt-
ing pot. (Asphalts made from different crude oils
have different vaporizing points. )
If temperature control is used, best results can
be obtained by using three separate tanks: Melt-
ing tank, holding tank, and dip tank. All three
should be provided with individual heating facili-
ties, which thereby permits minimum temperature
differentials between tanks. In this manner, the
holding tank's temperature can be held to a mini-
mum (about 400°F) without regard to heat loss at
the dip tank. Automatic temperature controls are
necessary for the holding tank. An asphalt feed
control bar installed on the asphalt roller in the
dip tank permits the temperature to be reduced
even further. This feed control bar, which is ad-
justable against the roller, controls the thickness
of the asphalt film applied to the paper; otherwise
this thickness' would have to be controlled by con-
trolling temperature and asphalt viscosity.
If control of asphalt temperature proves imprac-
tical, then a collection device should be used to
preventthe fumes from escaping to the atmosphere.
This can be done effectively with a two-stage, low-
voltage electrical precipitator, and sometimes with
a high-efficiency water scrubber. If a scrubber
-------
350
MECHANICAL EQUIPMENT
is used, recirculation of the -water is not advised,
since plugging of the water nozzles may occur un-
less the asphalt particles are somehow removed,
say by flotation.
PERLITE-EXPANDING FURNACES
INTRODUCTION
Perlite is a glassy, volcanic rock of the composi-
tion of obsidian but divided into small, spherical
bodies by the tension developed during its con-
traction on cooling. It is grayish "with a soft, pearly
luster. Chemically, perlite consists chiefly of the
oxides of silicon and aluminum combined as a nat-
ural glass with water of hydration. Upon rapid
heating, the escaping water of hydration causes the
spherules to expand and form white, cellular, low-
density particles. This process is termed exfolia-
tion.
Uses
About 90 percent of expanded perlite is used as an
aggregate in plaster and concrete. When mixed
\vith gypsum and water, perlite creates a plaster
that can'be troweled or sprayed on lath to form a
lightweight, resilient wall or ceiling. Perlite in-
sulating concrete can be used in the form of pre-
cast slabs or poured on lath, formboard, or steel
decking. Loose perlite is also used extensively
as an insulating fill for concrete block walls, as
a cavity "wall insulation, and as an insulating fill
in attic floors. Other uses for perlite include: Oil
well cement; mineral filter aid; pipe, furnace, and
boiler insulation; foundry sand additive; packaging
medium; soi] conditioner; and ceramic and paint
additive.
Mining Sites
Several perlite ore deposits are in California, and
other deposits are in six of the Rocky Mountain
States. Perlite ore is surface mined or quarried
and is normally dried, crushed, and screened at
the mine. The normal size of crude perlite for
plaster aggregate ranges fromrninus 12 or 14 mesh
to plus 40 or 60 mesh. Some plants use a size
range with no limitations on the fines. Crude perlite
for concrete aggregate ranges from 1/8 inch, plus
16 mesh, to 1/2 inch, plus 100 mesh.
Perlite Expansion Plants
A plant for the expansion of perlite consists of ore-
unloading and storage facilities, a furnace-feeding
device, expanding furnace, provisions for gas and
product cooling, product-classifying and product-
collecting equipment, and dust collection equip-
ment. A schematic diagram ot a typical plant is
shown in .Figure 245. A plant producing a number
of products has several bins for the storage of dif-
ferent grades of crude perlite. If the minus 100-
mesh material is not removed from the perlite ore
at the mines, filtered vents are required on the
storaige bins to prevent dust emissions during ore-
unloa.ding operations.
Expansion Furnaces
Vertical furnaces, horizontal stationary furnaces,
and horizontal rotary furnaces are used for the
exfoliation of perlite, the vertical types being the
most numerous. Only a few of the furnaces are
refra.ctory lined.
Essentially all perlite furnac es are fired with nat-
ural gas. The natural gas rate, amount of excess
air, and ore feed rate are adjusted to give a fur-
nace temperature, an effluent gas flow rate, and
a material residence time that will yield a prod-
uct of the desired density. Product densities vary
from 2 to 15 pounds per cubic foot, and furnace
temperatures vary from 1, 450 ° to 1,800°F. The
relationships of temperature and residence time to
product density are, for the inost part, trade se-
crets. The expanded product is carried out the
top of the furnace by the combustion gases.
Gas and Product Cooling
Cooling by heat exchangers or by dilution with am-
bient air are the two common methods that have
been used. Combinations of the two are also used.
The final temperature to which the gases must be
cooled depends upon the type of dust collector used,
as will be discussed later.
Heat exchangers generally employed are of the
tubular type with forced-air convection. Large U-
tubes with natural convection would probably be
practical but have not been used extensively be-
cause: of the space requirements. Cooling by dilu-
tion greatly increases, of course, the volume ol
gases to be handled by the dust collector. Some
of the smaller plants, however, have used this
method satisfactorily.
Product Collectors and Classifiers
Cyclone separators are used to collect the product.
If only one product is made, a single cyclone sep-
arator is used. To make more than one product,
two cyclones in series are usually used, in which
case some means is often provided lor regulating
the collection efficiency of t.ie first cyclone so as
to allow a controlled amount of fines to pass through
to the second cyclone. The product collected in
the first cyclone is used as a plaster and cement
aggregate, and the fine product collected in the
second cyclone has uses such as filter aid, paint
-------
Perlite-Expanding Furnaces
351
F ILTERED
MR
COOLING AIR
n
Figure 245. Flow diagram of a typical perlite-expandmg plant.
additive, insecticide carrier, and others. The
products are packaged in 3- or 4-cubic-foot bags
by packing machines with little or no dust loss.
If a baghouse dust collector is used, an ultrafine
product is collected in the baghouse hopper.
THE AIR POLLUTION PROBLEM
A fine dust is emitted from the outlet of the last
product collector. The fineness of the dust varies
from one plant to another, depending upon the prod-
ucts desired. In any event, a baghouse is needed
to achieve complete control. For example, one
plant that was tested produced perlite for use in
manufacturing insulated wallboard. Only one prod-
uct cyclone was used. A particle size analysis of
the baghouse catch revealed that 64. 3 percent by
weight of the sample was minus 200 mesh; approx-
imately 20 percent by weight was less than 5 mi-
crons. Specific gravity was 2.69 at 69°F. Table
104showsa complete particle size analysis of the
cyclone and baghouse catches.
HOODING AND VENTILATION REQUIREMENTS
No hooding is required unless ventilation of the
sacking machines receiving product from the cy-
clones is necessary. For most plants, this is not
required, and only the air outlet of the last product
cyclone needs tobe ducted to a dust collector. The
volume of ventilation air required depends upon the
quantity of air needed to convey the product, the
amount of fuel burned, and the volume of dilution
air required to cool the effluent sufficiently for
admission to a dust collector. The first two factors
are fundamental to the basic design of the plant.
Once these are known, one can calculate the quan-
tity of dilution air required as a function of the tem-
perature limitation of the dust collector.
AIR POLLUTION CONTROL EQUIPMENT
Simple cyclones have been found inadequate for
collecting fine dust from perlite furnaces. Even
the relatively high-efficiency devices, such as
multiple small cyclones, have been deficient in com-
plying with air pollution prohibitions. Several
firms have attempted to use water scrubbers, but
most of these installations were unsuccessful.
Virtually all the perlite-expanding plants in the Los
Angeles area are now equipped with baghouses.
These efficient collectors, costing only slightly
more than a well-designed scrubber, are able to
collect a salable product.
Since the gases from the expanding furnace are
at a relatively high temperature, considerable
cooling is necessary in order to meet the temper-
ature limitations of any fabric used in a cloth filter
dust collector. When Dacron cloth is used, the
usual practice is to cool the gases to400°to 500°F
in a tubular heat exchanger. Further cooling takes
place in the cyclones, and sufficient dilution air
is admitted to cool the gases to200°to 250°F be-
fore they enter the baghouse. Siliconized glass
fabric has been used, the cooling accomplished
entirely by dilution. Other combinations are of
course possible, but these two are most popular.
In order to secure a uniform product from the ex-
raansiori furnace and classifying system, mainte-
nance of a constant flow rate chrdigh _nc" baghouse
-------
352
MECHANICAL EQUIPMENT
Table 104. PARTICLE SIZE ANALYSES
FROM THE PRIMARY CYCLONE AND
THE BAGHOUSE CATCH OF
A PER. LITE-EXPAND ING FURNACE
+ 10
-10+30
-30+60
-60+100
-100+200
-200
Particle size analysis,
wt %
Primary cyclone
catch
0. 4
26. fl
30.0
22. 2
14. 0
7. 4
Baghouse catch
0.0
0. 4
2. 1
9.5
23. 1
64. 3
Particle size analyses of
-200-mesh portion of samples
Diameter (D),^
45. 7
40.2
36.6
32.9
29.3
25.6
22.0
18. 3
16.5
14.6
12.8
12.2
11.6
11.0
10. 4
9.8
9.2
8.5
7. 3
6. 1
4. 9
3. 7
3. 0
2. 4
1.8
1. 5
1.2
Sample with diameter < D,
wt %
Primary cyclone
catch
100. 0
99.3
99. 0
96.3
93. 7
90. 2
85.4
80. 5
77. 1
71.2
63.2
60. 5
57. 5
55.6
52. 0
48.8
46.6
42. 0
35. 1
27. 3
19.0
11. 7
7.6
6. 1
3. 9
3. 7
3. 2
Baghouse catch
__
--
--
--
100. 0
99. 4
97.6
96.4
94. 5
93.6
91.5
88. 5
86. 1
82. 1
81. 2
74. 2
70. 6
66.4
55.2
43.0
29.4
16. 7
11.5
7. 0
2.4
1. 5
1. 2
is highly desirable. In general, the resistance of
a baghouse increases as the dust layer builds up.
This gives anonuniformflowrate unless measures
are taken to counteract this tendency. Three gen-
eral methods have been used to maintain relative-
ly uniform flow rates:
1. Use of a single-compartment baghouse with
an adjustable restriction in the inlet duct. The
restriction is set at a maximum value when
the bags are clean and is decreased as the
baghouse's resistance increases, and this
maintains a relatively constant total resistance.
This method requires frequent adjustment of
the restriction and res erve fan capacity. When
the restriction reaches its minimum value,
the process must be shut down.
2,. Use of compartmented baghouses, which per-
mits one compartment at a time to be shut off
for bag shaking. This produces a resistance
that varies cyclicly, but flow Variations can
"be kept within tolerable limits. The greater
the number of compartments, the smaller the
variations in flow.
3. Use of continuous-cleaning-type baghouses.
Included in this category are types using high-
pressure blow rings '(Hersey types), those
Tising traveling blow chambers on envelope-
type bags, and those using pulses of high-
pres sure air. These typies are capahle of main-
taining almost completely uniform flow rates,
but their costs are somewhat greater than
those of the other types.
Filtering velocities shouldbe 3 fpm or less for the
standard types using woven fabrics and about 10
fpm or less for the Hersey types.
FEED AND GRAIN MILLS
INTRODUCTION
Commercial development of feed mills, based up-
on scientific animal nutrition, has advanced rapid-
ly since 1930. Enriching feed with vitamins and
minerals has accelerated the growth rates of poul-
try and livestock to nearly double the average
growth rates of 1930.
With changes in feeding, the animals are increas-
ingly being moved from cattle range and rural
farm forage areas to confined pens and feed lots
near urban areas. This transition tends to locate
the feed and grain plants in congested areas -where
many conflicts about air pollution arise. The
handling and manufacture of feed and grain prod-
ucts generates many varieties and concentrations
of dust. These dusts are the sole air contaminants
from these plants.
To pinpoint the sources of dust, a simplified di-
agram of feed mill flow is presented in Figure 246.
The drawing delineates basic equipment in solid
lines and dust control equipment in dotted lines.
Solid-line arrows indicate the flow of basic mate-
rial from process to process. Dotted-line arrows
indicate the forced discharge of dusty air to col-
lectors.
-------
Feed and Grain Mills
353
•BAGHOUSES-
SHIPPING
Figure 246. Flow diagram of a simplified feed mill. Basic equipment shown in solid lines
dust control equipment, in dotted lines.
Receiving, Handling, and Storing Operations
Feed materials are shipped to feed and grain plants
in railroad cars and trucks. These carriers may
be classified according to the type of unloading
operation used.
One class includes hopper bottom railroad cars,
trucks and trailers, trucks with self-contained
conveyors, and hoist dump vehicles. The flow of
materials from these self-unloading shipping con-
tainers may be regulated so as to fill an inclined
chute or shallow hopper as rapidly as the material
is removed. This is the choked-feedmethod of un-
loading, in-which a solid stream of material moves
slowly into the receiving system with little or no
dust emissions. Figure 247 illustrates choked-
feed receiving from a hopper bottom railroad car.
Canvas boots or socks may be fastened to the spouts
and extend down within inches of the hopper grat-
ings, though they are not very frequently used.
Another class includes flat bed trucks and box cars
capable of being emptied into receiving hoppers
only by mechanical plows or shovels. The carrier
beds are about 3 feet above the hopper gratings,
which are located at track or ground level. The
flat bed carriers are usually unloaded into deep,
large-capacity receiving hoppers. The excess
surge-holding capacity allows enough time between
car unloadings for an empty car to be replaced by
a full car, while the handling system continues to
convey material out of the hopper. This method
provides for receiving the maximum number of
cars or trucks per day and may also effect some
savings in labor costs. Figure 248 shows the un-
loading of a boxcar into a deep hopper.
Feed materials are less commonly unloaded from
carriers by pneumatic conveyors. The material
may be fed manually to a flexible suction tube,
connectedto a pullthrough cyclone, which separates
the feed materials from the air conveying system
and drops them into a storage bin. Another pneu-
matic unloading system type uses specially con-
structed hopper bottom cars or trucks equipped
with air or mechanical agitation devices. These
devices feed the material through a rotary valve
to a pressure-type pneumatic conveyor. The air-
borne material from this type of conveyor is also
separated by a cyclone and dropped into storage.
Grain and feed storage bins maybe single or multi-
ple compartmented. They are usually constructed
of steel or concrete. Each bin or compartment is
enclosed by a dust-tight cover incorporating an ade-
-------
354
MECHANICAL EQUIPMENT
HOPPER
Figure 247. Hopper bottom railroad car unloading grain into a shallow hopper by the
choked-feed method (Koppel Bulk Terminal, Long Beach, Calif.).
Figure 248. Boxcar unloading grain into deep receiving
hopper.
quately sized vent. This vent provides an escape
for displaced air during filling and prevents the bin
from buckling under external atmospheric pressure
during the discharge operation.
Feed-Manufacturing Processes
From the storage bins, -whole grains are conveyed
to cleaning, rolling, grinding, and other plant pro-
cesses. The processed grains may be shipped to
consumers or held for feed formulation. Finished
"eed formulas are compounded from vitamins, anti-
aiotics, minerals, and all the processed materials.
These compounds may be prepared in the form of
:inely ground mash, pellets, or mixed mash and
pellets. The feeds may be shipped from the mill
in plant-owned delivery trucks, common carrier
;rucks, or by rail.
A certain amount of dockage is acceptable, by gov-
srnment grading standards, inallgrains. Dockage
is made up of dust, sticks, stones, stalks, stems,
vveed seeds, and other grains. A portion, if not
the majority of this undesirable material, must
be removed if the grain is to go into certain pro-
cesses. The degree of separation required depends
upon the actual process, for example, barley to
be ground in a hammer mill needs minimum clean-
ing whereas barley to be rolled requires a high de-
gree of cleaning. In some circumstances, received
grains may have been cleaned before elevator stor-
age or as preparation for export shipment, in order
to eliminate hazards of spontaneous heating, in-
sect infestation, and so forth.
Cleaning includes the several mechanical process-
es by which dockage is removed from grain. By
the nature .of its purposes, cleaning produces a
large amount of dust. The amount of dust varies
widely with the different field sources of grain and
its subsequent handling. Apreliminary step in the
cleaning process is termed scalping. In this pro-
cess, the grain is run through a coarse mesh screen
in shaker or reel form, to remove sticks, stones,
stalks, strings, and similar offal. The grain is
GPO 806*614—13
-------
Feed and Grain Mills
355
usually poured through the screen at low velocity
with little or no aeration; very little dust is gen-
erated. The shaker type of scalper maybe of dust-
tight design with no vent to the atmosphere. Another
step is called aspiration. Crosscurrents or coun-
tercurrents of air are directed through dispersed
falling grain. The process is designed to separate
field dust, fibers, chaff, and light trash from the
grain. The third step employs a stack of several
grading shaker screens to classify the grain. Mixed
grains are separated at this point. Noxious weed
seeds are also removed, to prevent them from be-
ing disseminated.
The three steps of cleaning may be accomplished
in separate devices or all in one piece of equip-
ment. A traditional type of cleaner, as shown in
Figure 249 combines all three of these steps in one
machine. This type commonly employs three inte-
gral blowers and has two exhaust airstreams that
carry away different types of separated materials.
Figure 249. Grain cleaner (Koppel Bulk Terminal,
Long Beach, Calif.).
Barley rolling is accomplished in equipment com-
monly called barley steamers and barley rollers.
Oats and milo may be processed in the same equip-
ment. Cleaned grain is conveyed and elevated from
storage to an open-coil-type steamer, which heats
and moistens the grain. It is then run through
steel rollers and dropped into a cooler through
which room air is pulled to cool the hot, moist
grain.
Many feed grains and some feeds such as copra or
cotton seed are ground in hammer mills. This
type • of mill is so constructed that it is also in-
herently a centrifugal blower. Granular materi-
al is fed into the center of a high-speed rotor, which
has pivoted or articulated hammers on the periph-
ery. The material is thrown centrifugally against
and through a perforated, peripheral plate or screen.
The proper flow of material through the mill re-
quires a strong stream of air. Supplemental air
capacity is generally supplied by a pullthrough
blower driven integrally from the mill shaft. The
ground product is then conveyed pneumatically to
a cyclone separator, which delivers the ground
meal to storage bins. Size reduction of feed is
sometimes accomplished in a burr mill or other
type of equipment that requires no airstream for
operation.
Pelleted dairy feed consists of several different
types of finely ground feed materials, combined
withmolasses and abinder material, steam condi-
tioned, and compressed into pellets by a pellet
mill. From the mill, pellets are dropped into a
cooler where a blower pulls room air through them.
After their cooling, dairy pellets are usually run
across a shaker screen for removal of any small
particles that occurred during the breaking of ex-
truded pellets away from the mill die. The parti-
cles are usually conveyed pneumatically from the
shaker back to the pellet mill feed.
Feed formulations are devised to suit all varia-
tions of creature appetites and conditions of live-
stock production, on a nationwide basis, or for in-
dividual flocks and herds. Component grains may
be steamrolled, or dusty feed material fines may
be pelleted to improve the texture and flavor.
A formulating equipment system consists of from
one to three scale hoppers, sized according to the
bulk class of products each weighs. Materials
may be measured into the scales by simple manual
operations or by elaborate pushbutton consoles that
operate remote conveyors from multiple storage
bins. After the scales there may be a single mixer
or a cascade of surge bins and parallel or tandem
mixers with oil and molasses sprayers. The batch-
es of finished feed may be conveyed to holding bins,
for later transfer to truck or railroad car, or they
maybe loaded directly to a carrier without holding.
THE AIR POLLUTION PROBLEM
Many feed and grain plants, originally located at
crossroads in sparsely settled farm areas, are
now surrounded by urban stores, offices, schools,
and modern residential developments. As a result
of frequent public complaints after community en-
circlement, the plants must either be relocated in
less sensitive industrial areas, or comprehensive
dust control programs must be initiated.
-------
356
MECHANICAL EQUIPMENT
There is now active medical research (McLouth
andPaulus, 1961) showing the deleterious or toxic
effects of feed grain dusts. Many individuals ex-
perience bronchial or allergic disturbances after
exposure to feed and grain process effluents. Per-
sons affected may be inside a grain-processing
plant or even some miles downwind (Cowan et al. ,
1963).
Pertinent to the control of dust inside plants is the
ever-present spectre of fire, sometimes sponta-
neous. Fire can run along dust deposits on mill
beams faster than a man can run to cut it off and
can thus envelop an entire building before fire equip-
ment can be used.
The destructive force of cereal dust explosions is
well known, especially the secondary type of ex-
plosion that occur s after a primary Shockwave has
lifted and mixed heavy dust deposits with air, cre-
ating a massive, explosive mixture.
The vacuum cleaning of mill interiors is, there-
fore, a constant, expensive chore. A likely answer
to the hazards of dust accumulation may be the con-
struction of unhoused feed process systems as is
nowfrequent practice in the power-generating, oil
refinery, and chemical process industries.
In undeveloped or farm areas, nopractical purpose
maybe served by preventing feed mill dust emis-
sions, but in urban areas, dust losses from feed
materials are likely to cause a nuisance. Basic
process equipment for either open or housed plants
will be increasingly required to effect dust-tight
enclosure by the use of sealants, gasketing, or
welded joints. Air vented from equipment will
need to be controlled either by filters attached to
basic equipment or by duct systems connected to
air pollution control equipment.
Feed materials and field run grains, received at
the mill, commonly contain much fine dust in ad-
dition to long, fiber-shaped dust particles. Fine
dust found in grain may include the actual soil in
which the grain was grown, owing to wind or rain
action in the field. Other fine particles may orig-
inate from weeds or insects or be produced from
the grain itself, by abrasion in handling and stor-
ing. For these reasons, no reliable prediction of
the kind and amount of dust in a shipment of field
run grain may be expected. The amount of dust
found in the many other miscellaneous feed mate-
rials varies far more widely than in grains.
The long-fibered dust particles, such as barley
beards and even weed seeds and other particles,
are much more an expected, characteristic part
of any particular grain shipment. These, however,
seldom present an air pollution problem.
Table 105 presents the particle size distribution
of dusts from a boxcar of barley received in a deep
hopper at a feed mill. Dust picked up by a control
hood was carried by a blower to a cyclone where
the larger particles dropped out and were collected
in a sack (sample No. 1). The cyclone then vented
to a baghouse, which collected the finer material
in a hopper (sample No. 2).
Table 105. PARTICLE SIZE ANALYSES OF
THE PRIMARY CYCLONE CATCH AND THE
SECONDARY BAGHOUSE CATCH OF DUST
FROM A RAILROAD RECEIVING HOPPER
HOOD CONTROLLING THE UNLOADING
OF A BOXCAR OF FEED-TYPE BARLEYa
Particle size distribution-by wt
Particle size, (i
0 to 5
5 to 10
10 to 20
20 to 44
44 to 74
74 to 149
!49 to 250
Over 250 (60 mesh)
Sample No, 1
cyclone bottoms, %
0.9
0. 9
3.9
9. 3
12.9
16.2
5. 4
50. 5
Sample No. 2
baghouse hopper, %
4
25
66
5
0
0
0
0
aSpecific gravity of both samples was 1. 8.
Receiving, Handling, and Storing Operations
The dusts that cause air pollution problems in re-
ceiving, handling, and storing operations are gen-
erally the fine dusts found in field run grains, or
in those feed materials from which much dust is
generated. When one of these materials is unload-
ed from flatbedtrucks or boxcars to deep hoppers,
it is dropped from a height of 3 to 15 feet in sudden
surges. The particles in the stream of free-falling
material disperse as they accelerate, and inspirate
a downward-moving column of air. When the mass
hits a hopper bottom, the energy expended causes
extreme air turbulence, abrasion, and deagglom-
eration of the particles. A violent generation of
dust occurs. It forms an ascending column that
boils out of the opposite end of the hopper. A dust
plume of lOOpercent opacity and of sufficient vol-
ume to envelop a boxcar completely may be formed
from the unloading of grain. Figure 250 shows how
dust is generated during the dumping of grain from
a boxcar into a deep hopper.
Conveying equipment does not usually present dif-
ficult dust problems; however, the rubbing fric-
tion of screw conveyors, drag conveyors, and buck-
et elevators on feed and grain abrades these mate-
rials, creating fine dust particles. Dust is gen-
erated at the transfer points of enclosed convey-
ing equipment, carried through bucket elevators,
and emitted at the discharge of the conveyed mate-
rials.
Belt conveyors are the most efficient type of han-
dling equipment, especially for large volumes of
-------
Feed and Grain Mills
357
/-DUST
/ PLUME
Figure 250. Unloading a boxcar into
a deep receiving hopper.
material and for long conveyances. They cause
less mechanical abrasion of the material and sep-
arate much less of the dusty fines from the grain
than screw conveyors do. Dusty air, however, is
usually generated at belt transfer points, result-
ing from aeration of material as it falls onto or
awayfrom a belt. A secondary problem with belt
conveyors results from materials' adhering to the
belt as it turns around the head pulley. These par-
ticles, usually coarse, drop from the returning
belt along its entire length.
Storage bins vent dust-laden air originating from
two sources. One is air displaced by incoming
material that falls freely from a spout at the top
of the bin, mixing dus t with the air in the bin. The
other is air inspiratedby the flow of incoming ma-
terial. This air may contain large quantities of
dust.
Shipping feed out of the plant, by spout loading in-
to cars or trucks, is similar to the storing opera-
tion. Most finished feeds are, however, some-
what agglomerated by molasses or oil additives so
that a minimum of dust is generated in the ship-
ping process. Dusty feeds, of course, require
special handling -when they are bulk loaded into
carriers.
When a large grain shipment is received, most car-
loads may contain a uniformly low content of fine
dust. The last several carloads, representing a
cleanup of fines thatbecame segregated in handling
and storage, may be excessively dusty.
Grain rolling and pelleting produce moist, agglom-
erated particles with no dust emissions from the
coolers.
In size reduction of whole grains or other feed ma-
terials, the amount of dust discharged from the
pneumatic conveyor cyclone may increase as the
materials are more finely ground. The character
of the material, however, is the chief determinant
of the dust generated.
During the formulating and mixing, some open-top
dump or cut-in hoppers, used to combine dust-
generating ingredients for mixing, require con-
trol. The methods of material handling such as
free fall, choke feed, and so forth determine the
character of the emissions in these .open systems.
Mixing systems now tend to be designed for dust-
tight enclosures of all conveying equipment, with
filter vents on surge bins and mixers. This plan
of dust control requires no other control equipment.
Poultry pellets are usually compounded with fish
oil or animal fats instead of molasses. If no fat
or pellet binder material is used, poultry pellets
that have been run through a shaker for removal
of fines may be moderately dusty. A totally dust-
enclosed type of shaker is recommended to pre-
vent dust loss to the air.
Care must be taken in returning collected dust to
a basic equipment system, or a heavy, recircu-
lating dust load may be created.
HOODING AND VENTILATION REQUIREMENTS
Hooding requirements in a feed mill are limited
to those for deep receiving hoppers, open convey-
ing equipment, and formulating hoppers in which
the material free falls without being enclosed. No
hooding is required for choke-feed hoppers, en-
closed conveying equipment, bins, or for any of
the manufacturing processes.
Feed-Manufacturing Processes
When grain is unloaded from carriers and conveyed
to storage, the granules flow in the form of a thick,
bulky stream that encloses and retains most of the
dust content. Thus the major proportion of dust
contained in the original grain shipment remains
to be removed by cleaning equipment that employs
large quantities of air. The dust must be separated
before this air is discharged to the atmosphere.
Receiving, Handling, and Storing Operations
A preferred method of hooding a deep receiving
hopper, to control dust emissions, is to exhaust
air from below the grating. As shown in Figure
251, a hopper with V-shapedbaffles belowthe grat-
ing is vented to control equipment. The baffles
reduce the area open to the atmosphere and also
reduce the air capacity required to vent the hopper
face. If the hopper is in a building, or complete-
ly sheltered from winds, an indraft velocity of 100
-------
358
MECHANICAL EQUIPMENT
fpm through the open area of the hopper, between
baffles, maybe effective. If moderate winds of 3
or 4 mph are to be encountered, an indraft veloc-
ity of 300 fpm may be required. For higher -winds,
fence-like baffles around the top of the hopper may
be required, to prevent the winnowing action of
strong wind currents across the hopper grating.
EXHAUST DUCT TO
DUST COLLECTOR
V-BAFFLES 7 P~~~^r ^GRATES
/ | I X
/\/X"/\/\i/\A/\/\/\/\/\
v y .. V J
^ PICK-UP SLOTS
\^_ /_
ELEVATION VIE*
Figure 251. Dust control hooding of deep receiving
hopper.
Belt conveyors are almost never fully enclosed.
They must, therefore, be hooded at both the point
where material is loaded onto the belt and the point
where it is discharged from the belt. The loading
and transfer chutes must be cleverly designed to
reduce dust generation at these locations. The
first objective is to design chutes so as to direct
the flowing material in the direction of belt travel.
The second objective is reduction of the open area
exposed to the atmosphere. The enclosing of the
transfer point maybe sealed right down to the belt,
with flexible rubber flaps. Moderate volumes of
pick-up air then suffice to control the dust. Indraft
velocities into the open-face areas of hoods, which
control belt transfer, should follow the same cri-
terion of 100 to 300 fpm recommended for receiv-
ing hoppers.
The secondary problem posed by mate rial that does
not fall cleanly away from the belt into the dis charge
chute may be remedied by the use of a rotary brush.
The brush is installed inside the combined dis-
charge chute and control hood, with a flexible rub-
ber wiper to close the hood up to the return belt.
The brush is usually driven by chain or V-belt from
the head pulley shaft at a speed 2 or 3 times that
of the pulley. This brush should be made of long-
fiber e
-------
Feed and Grain Mills
359
sensitive communities where nuisance complaints
and air pollution regulations take effect, baghou ses
are needed for final dust control of feed plants.
Table 106 shows the results of three tests for de-
termining the loss of grain dusts from cyclone out-
lets to the atmosphere.
Receiving, Handling, and Storing Operations
The deep free-fall type of receiving hopper is not
normally controlled in farm or nonsensitive areas.
In urban areas it maybe adequately controlled only
by a baghouse or cyclone-baghouse combination.
Dust emanating from pneumatic unloaders, pneu-
matic conveyors, belt conveyors, and elevators
need not be collected in nonsensitive areas. Other-
wise, baghouse control is needed in urban areas.
Storage bins and shipping containers need no con-
trol innonsensitive areas. Elsewhere the two ap-
plicable control methods are (1) to exhaust the
bins and containers by duct connection to baghouse
control systems, or (2) to employ some form of
a filter vent attached directly to each bin or ship-
ping container.
Feed-Manufacturing Processes
In urban or sensitive areas, grain cleaner and
hammer mill cyclones and cut-in hopper hoods need
to be controlled by baghouses. In undeveloped
areas, cleaner and hammer mill cyclones may be
vented to the atmosphere. If, however, much grain
is to be ground in a hammer mill, the use of a bag-
house to prevent economic loss may be feasible.
The,hot, moist, agglomerated particles in rolled-
grain cooler exhausts or in pellet cooler exhausts
are adequately controlled by a cyclone in any type
of area, though condensed water vapor plumes
from the cyclone are very noticeable under high-
moisture and cold-weather conditions.
Filter Vents
A filter vent consists of a filter cloth bag or sock,
usually made of cotton sateen, tightly fastened over
a vent. A sheet metal enclosure is added if the
vent is exposed to weather. The same control prin-
ciple can also be used in loading feed into trucks
or railroad cars, through down spouts inserted into
the hatches. A filter vent skirt is sealed around
both the spout rjipe and the hatch opening, as shown
in Figure 252.
The pneumatic loading of boxcars maybe controlled
by a flat filter cloth screen of cotton duck or cot-
ton drill across the door. In loading a. ship's hold,
at a high-volume rate with dusty material, effec-
tive control may be obtained with similar filter
cloth screens. A hatch opening, upto 25 by 30 feet
in size, can be enclosed by two 25- by 40-foot screens,
-with a wide center overlap around the downspout,
as shown in Figure 253.
Filter vents vary in size, from about 1 foot in di-
ameter by 2 feet in height, to perhaps 3 feet in di-
ameter by 5 feet in height. They may, however,
be of any size or shape. Filtering velocities should
not exceed 4 to 6 fpm for control of miscellaneous
feed material dusts. Higher velocities maybe used
in filtering coarse dust or when a filter is used for
short or intermittent periods of operation. Some
provision for shaking the bags shouldbe made when
necessary. Insect infestation should also be con-
sidered when filter bags are not cleaned or changed
from one bin filling to the next.
Cyclones
Cyclones are used with great versatility in feed
mills. They are an integral part of almost every
equipment system that handles air. In practice,
nearly all cyclones found in feed plants are of the
simple, low- or medium-efficiency types. High-
efficiency, multiple cyclones are subject to exces-
sive operational costs and maintenance problems.
Table 106. DUST LOSSES FROM CYCLONES
Grain
Basic equipment
Process wt, Ib/hr
Exhaust air volume, scfm
Dust concentration, gr/scf
Dust loss, Ib/hr
Malted barley
Grain cleaner
Test No. 1
53, 000
2, 970
0. 194
4. 95
Test No. 2
50, 000
2, 970
0. 160
4. 07
Feed barley
Hammer mill
10, 350
3, 790
0.488
15.8
Milo
Grain cleaner
11, 250
First cyclone
3, 680
0.058
1.83
Second cyclone
2, 610
0. 006
0. 13
-------
360
MECHANICAL EQUIPMENT
Figure 252. Counterweighted, telescoping downspouts
used to fill a hopper car. Loading is controlled
by filter vent skirts (Ralston Purina Company, Los
Angeles, Calif.).
Cyclones collect almost all grain dusts larger than
10 to 20 microns in diameter. They collect only
a very small proportion of the particles smaller
than 10 microns, as shown in Table 105. Thus,
their percentage efficiency, that is, the propor-
tion of the total material -weight caught to the total
material weight in the exhaust air stream, is very
high. Nevertheless, the proportion of fine dust
particles caught by a cyclone to the total number
of fine dust particles in an exhaust stream is in-
variably very low. These fines are the particles
that become airborne and constitute an air pollu-
tion problem. Special design information for cy-
clones is given in Chapter 4.
Doghouses
Baghouses for most mill operations tend to be of
the simplest and least expensive types, and use
cotton sateen in most cases. Hand shaking of the
filter bags is preferred, to avoid any risk of fire
from automatic shaking equipment. Filtering ve-
locities are from 2 to 3 fpm. for continuous oper-
ation, and up to 6 fpm for intermittent use. Cost
of the baghouse maybe as low as $1. 00 per square
foot of filter cloth.
Figure 253. Loading alfalfa pellets into a ship's hold, controlled by two 25- by 40-ft filter cloth
screens (Pacific Vegetable Oil Corporation, Long Beach, Calif.).
-------
Feed and Grain Mills
361
The static pressure drop through the baghouse is,
in most cases directly proportional to the filter -
ingvelocity. Where a dust cake is allowed to build
up for several hours before the shaking or a per-
manent low-porosity cake has developed, the pres -
sure drop in inches of water column may be esti-
mated as equal to the filtering velocity in fpm. Air-
streams with heavy dust or material loadings are
usually exhausted to a primary separator cyclone
and then to a baghouse. This method relieves the
baghouse of handling an excessive volume of bulk
material.
Larger feed mills and those operated in conjunc-
tion-with flour and cereal plants are usually equipped
•with the more sophisticated and expensive types
of baghouses. These use elaborate, mechanically
programmed bag shaking with filtering velocities
as high as 10 fpm. Reverse-jet and reverse-air-
blowing types are alsoused. One modern feed and
grain terminal, shown in Figure 254, makes very
extensive use of rever se-jet baghouses . It is prob-
ably the "world's most completely controlled feed
and grain terminal facility. Baghouses, as shown
in Figures 255 and 256, control dust from truck-
and railroad-receiving hoppers. Several other
baghouses, whichmaybe seen in Figure 254, con-
trol all the material-handling conveyors and ele-
vators, storage and "weighing facilities, and grain-
cleaning equipment. Another baghouse provides
ventilation to the hold of the ship, which is cov-
ered by filter cloth screens during the loading oper-
ation. The control equipment incorporated in this
facility prevents any visible emissions and is an
outstanding example of the control of air pollution
by this industry.
Figure 254. Modern bulK feed and gram terminal with
reverse-jet baghouses controlling all operations
(Koppel Bulk Terminal, Long Beach, Calif.).
Figure 255. Truck-receiving station with baghouse control
of the receiving hopper (Koppel Bulk Terminal, Long Beach,
Calif.; and Wunsch Harvesters, Phoenix, Ariz.).
-------
362
MECHANICAL EQUIPMENT
Figure 256. Automatic boxcar-unloading system with baghouse control of the receiving hopper (Koppel Bulk Terminal,
Long Beach, Calif.).
PNEUMATIC CONVEYING EQUIPMENT
INTRODUCTION
Pneumatic conveying involves the movement of
powdered, granular, or other free-flowing mate-
rial in a stream of air. The bulk of the material
is separated from the conveying air in a product
collector, usually a cyclone separator. If the air
discharge contains an appreciable amount of dust,
it must be passed through a dust collector before
being discharged to the atmosphere. A cloth filter
dust collector is almost invariably used for this
purpose. The weight of dust passing the product
collector is normally very small in proportion to
the weight of material conveyed, but it is usually
of very fine particle size, a relatively small amount
of which may result in excessive opacity.
Types of Pneumatic Conveying Systems
In general, there are two types of pneumatic con-
veying systems: Negative-pressure systems,
characterized by low capacity and low pressure
-------
Pneumatic Conveying Equipment
363
losses; and positive-pressure systems, charac-
terized by high capacity and high pressure losses
(Fischer, 1958). To convey from several points
to one point, a negative-pressure system is usu-
ally used. To convey from one point to several
points, a positive-pressure system is usually used.
In a negative system, the material is drawn into
the conveying line by suction created at the far end
of the system by a centrifugal fan or a rotary posi-
tive-displacement blower. The product is collected
in a cyclone separator, which has a rotary airlock
at its base enabling it to discharge material con-
tinuously while maintaining the vacuum. The fan
or blower is located on the air discharge side of
the cyclone to prevent excessive -wear from, product
abrasion. Narrow-blade centrifugal fans and cy-
clones are often made as integral units, as shown
in Figure 257. The filter is on the discharge side
of the fan. A rotary positive-displacement blower
can also be used in a negative system. The much
higher vacuum produced by this unit gives it a much
greater conveying capacity, but requires that the
cyclone collector be of heavier construction. The
close clearances -within these machines usually re-
quire that a filter be placed on the "inlet side of the
pump to prevent dust from being drawn through the
pump.
NARROI-BLADE-
CENTRIFUGAL FAlT
Figure 257. Negative-pressure conveying system
ROTARY POSITIVE
DISPLACEMENT
BlOIER-j
In positive-pressure systems, the air-moving unit
is at the head of the line instead of the end. Mate-
rial is fed into the airstream by a. rotary airlock
or feeder and is blown to its destination. Rotary
positive-displacement blowers or sliding-vane ro-
tary compressors are used in positive-pressure
systems. High pressures obtainable with these
units permit relatively large quantities of materi-
als to be conveyed with smaller volumes of air
than can be handled in a negative system. This
permits the use of smaller diameter conveying
lines and smaller dust filters since the filter unit
is generally rated on the amount of air it can han-
dle. The filter is placed at the end of the system
to filter the air discharging from the product col-
lector, as shown in Figure 258.
Types of Air-Moving Used in Conveying
The different devices used for moving air in con-
veying systems are characterized principally by
the'pressure that can be developed. The following
four groups (see Figure 259) include most of the
devices used:
1.
Industrial exhausters. These centrifugal fans
have a pressure limit of about 16 inches water
column. The weight of material conveyed is
only a f r action'of the weight of air moved. Their
use for conveying is usually limited to bulky,
Figure Z5U. Fos.'siye-jirsssure con-raying system.
-------
364
MECHANICAL EQUIPMENT
Figure 259. Types of air-moving devices used in pneumatic conveying systems: (top left) Industrial exhauster
(Chicago Blower Corp., Franklin Park, III.); (top right) narrow-blade centrifugal fan (Chicago Blower Corp V
(bottom left) rotary positive-displacement blower, (Sutorbilt Corp., Los Angeles, Calif )• (bottom right) slli
ing-vane rotary compressor, (Fuller Company, Catasauqua Pa.)
-------
Pneumatic Conveying Equipment
365
low-densitymaterials sucfh as sawdust, wood
shavings, cotton, and other fibrous materi-
als. They are used extensively to convey
materials from, cutting, shredding, and grind-
ing machines to storage or further processing.
2. Narrow-blade centrifugal fans. These fans
can developpressures of up to 50 or 60 inches
water column. Weights of material conveyed
are of the same magnitude as the weight of
conveying air. These fans are frequently
mounted as an integral part of a cyclone col-
lector. They are extensively used to unload
grains and other free-flowing materials. Their
use is confined almost exclusively to negative-
pressure systems. Conveying distance is lim-
ited to about 150 or 200 feet at any practical
conveying rate. Two of these fans are some-
times placed in serie.s to give additional ca-
pacity or extend the conveying distance.
3. Rotary positive-displacement blowers. These
units canproducepressures up to 15 psi. The
weight of material conveyed is several times
the -weight of the conveying air. They can con-
vey for distances of several hundred feet. They
are used in both positive- and negative-pres-
sure systems.
4. Sliding-vane rotary compressors. These ma-
chines operate in the pressure range of 15 to
50 psi for single stages and up to 100 psi for
double stages. They are water jacketed to
dissipate the heat of compression and can con-
vey for distances of several thousand feet at
very high ratios of solids to air.
Preliminary Design Calculations
The basic problem in design is to determine the
energy requirements. These can be expressed in
pressure and volume units, and from these units,
the size of the blower and the required horsepower
can be estimated. These procedures are useful,
for preliminary estimating purposes, to those con-
templating the installation of a pneumatic convey-
ing system, and would also be useful to an air pol-
lution control official in evaluating a proposed con-
veying system for permit requirements.
The first step in designing a conveying system is
to determine the required conveying velocity. Many
theoretical methods of making this determination
have been proposed. These methods, however,
give only the balancing or floating velocity, such
as the terminal velocity given by Stokes law. In
order to ensure sustained movement of solids, a
velocity considerably in excess of the floating ve-
locity must be used. Hence, reliance upon em-
pirically determined velocities is usually neces-
sary. Table 107 gives velocity ranges found satis-
factory for a number of materials.
Fischer (1957) divides the energy requirements
into two categories, one for overcoming material
losses and the other for overcoming air losses.
Air losses are those caused only by flow of the air.
Material losses are the additional losses due to
conveying the material. He subdivides the mate-
rial losses into four groups and estimates them by
the following empirical relationships:
1. Acceleration. Energy required to bring the
material from rest up to conveying velocity is
given by the formula
E = MV /2g (104)
where
E = energy, ft-lb/min
M = solids moved, Ib/min
V = velocity, ft/sec
g = acceleration due to gravity, ft/sec .
2. Lifting energy. Energy required to lift a given
amount of material a given distance can be ex-
pressed as
E = M (d ) (105)
v
where d = vertical distance, ft.
v
3. Horizontal requirements. The energy required
to move a material in a horizontal duct can be
estimated by the empirical formula
E = M
(dh)(f)
(106)
where
f = coefficient of friction (calculated as the
tangent of the angle of slide) between the
material being conveyed and the material
from which the duct is made
d = horizontal distance, ft.
h
4. Bends and elbows. The weight of solids mov-
ing around the bend is multiplied by the cen-
trifugal force imparted to it according to the
formula
E = (MV^gR)(d)(f)
(107)
where
R = radius of bend, ft
d = distance around bend, ft.
-------
366
MECHANICAL EQUIPMENT
Table 107. VELOCITIES FOR LOW-PRESSURE
PNEUMATIC CONVEYING SYSTEMS
(Alden, 1948)
Material
Ashes, clinkers, ground
Barley
Cement, Portland
Coal, powdered
Coffee beans, stoned
Coffee beans, unstoned
Cork, ground
Corn
Cotton
Cotton seed
Flour
Hemp
Hog waste
Jute
Lime
Metal turnings
Oats
Pulp chips
Rags
Rye
Salt
Sand
Sawdust
Sugar
Tanbark, dry
Tanbark, leached, damp
Wheat
Wood flour
Wool
Velocity,
6,
5,
6,
4,
3,
3,
3,
5,
4,
4,
3,
4,
4,
4,
5,
5,
4,
4,
4,
5,
5,
6,
4,
5,
4,
5,
5,
4,
4,
000
000
000
500
000
500
500
000
000
000
500
500
500
500
000
000
500
500
500
000
500
000
000
000
500
500
000
000
500
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
fpm
8,
6,
9,
6,
3,
4,
5,
7,
6,
6,
6,
6,
6,
6,
7,
7,
6,
7,
6,
7,
7,
9,
6,
6,
7,
7,
1,
6,
6,
500
500
000
000
500
000
500
000
000
000
000
000
500
000
000
000
000
000
500
000
500
000
000
000
000
500
000
000
000
Solution:
1. Mass rate:
With reference to Table 107, a conveying ve-
locity of 6, 500 fpm is selected (108. 3 ft/sec)
Mass rate = 15,
2. Material losses:
000/60 = 250 Ib/min
Acceleration loss = MV 1 2s =
250(108. 3)2
2(32.1)
Lifting energy =
Horizontal loss
250(300)(0.7)
Elbow losses =
3(250)(108.3)2
32.1(4)
54,500 ft-lb/min
= M(d ) = 250 x 70
V
= 17, 500 ft-lb/min
= M(d,)(f)
h
= 52, 500 ft-lb/min
3(Mv'Vg R)(d)(f)
2(3. 14)(4)(0.7)
4
= 301,000 ft -Ib/min
Total material loss = 54,500 + 17,500
+ 52, 500 + 301,
000 = 425, 500 ft-lb/min
Air losses are calculated by the methods given in
Chapter 3. Cyclone collector losses range from
2 to 4 inches of water column, and cloth filter
resistances range from 3 to 5 inches of water col-
umn.
To illustrate the calculation methods, a sample
problem will be worked.
Example 31
Given:
Material, salt
Conveying rate, 15, 000 Ib/hr
Horizontal distance, 300 ft
Vertical distance, 70 ft
Three 90° elbows of 40-ft radius
Angle of slide, 35° (tangent of 35° = 0. 7).
Problem:
Calculate the required power input.
Assume a 5-inch line:
3 14 5
Volume =—=- x — x 6,500 = 885 cfm
Convert material loss to pressure drop:
425, 000 ft-lb/min
885 ft /min
Ib/ft
Convert pressure drop to inches of water
column:
(481 Ib/ft )(12 in.)
= 92 in. WC.
62.4 Ib/ft
3. Air losses:
Total equivalent length of duct
Q nr\ i 7 A i \ M-* • ^ ^/ '**/ _ o o n f4.
= 3(JU -T 70 + ~ JO7 It
The friction loss cannot be read directly from
the Air Friction Chart (p. 46} because it is off the
-------
Driers
367
chart. Read the loss at 5, 000 fpm and multiply
by (6,500/5,000)2.
Friction loss = 8. 9 (6, 500/5, OOO)2
= 15 in. WC per 100 ft of duct
389
Total duct loss = (15) —j = 58 in. WC
Assume a cyclone loss of 3 in. and a filter loss
of 4 in. Total air loss = 58+ 3+4 = 65 in. WC
4. Total pressure loss:
Loss = 92 + 65 = 157 in. WC, or
(157 in. )
(62. 4 Ib/ft )
12 in.
(1 ft2)
144 in. '
= 5.7 Ib/in .
5. Required power input:
A rotarypositive-displacementblower will be
used in a positive-pressure system such as
shown in Figure 259.
Assume a blower efficiency of 60%. The re-
quired power input is:
(5. 7)(144)(885)
33,000(0.6)
= 37 hp
THE AIR POLLUTION PROBLEM
The tendency of dust to be emitted from the product
collector is determined largely by the amount of
fine material in the product. For finely pulverized
materials such as cement and flour, a dust filter
is absolutely necessary both from the point of view
of loss of product and creation of a dust nuisance.
For some materials, the amount of foreign mate-
rial determines the need for a dust filter. For
instance, whole grains do not require a filter if
they are completely clean; however, they usually
contain enough dirt to require a filter.
AIR POLLUTION CONTROL EQUIPMENT
A conventional baghouse is the usual dust filter
used, though reverse-air cleaning types are also
used. The dust filter for high solids-to-air sys-
tems may consist of cloth filter tubes mounted on
top of a storage bin, which is the product collector.
Cloth tubular filters are sometimes mounted in-
tegrally with cyclone product collectors. The fil-
ter tubes are mounted in a cylindrical housing •whose
lower part is a cyclone separator. The filter hous-
ing is divided into four compartments with auto-
matic shaking devices to allow continuous opera-
tion.
Filtering velocities commonly used range between
2 and 4 fpm. The optimum velocity varies with
particle size and the tendency of the dust to pack.
In general, the lower velocities tend to give more
trouble-free operation and it is seldom profitable
to economize by increasing the filtering velocity.
DRIERS
INTRODUCTION
A drier may be defined as a device for removing
water or other volatile material from a solid sub-
stance. Air contaminants emitted are dusts, va-
pors, and odors. Several driers for specific prod-
ucts and processes have been discussed in other
sections. Inthis section, some general character-
istics of driers and some details of a few specific
types will be considered.
Rotary Driers
A rotary drier consists of a rotating cylinder in-
clined to the horizontal with material fed to one
end and discharged at the opposite end. In the most
common type, heated air or combustion gases flow
through the cylinder in direct contact with the ma-
terial. Flow may be either parallel or counter-
current. This type is called a direct rotary drier.
In another type, called an indirect rotary drier,
heat is applied by combustion gases on the outside
of the cylinder or through steam tubes inside the
cylinder. Inthis type, a flow of air is maintained
through the drier to assist in the removal of water
or other vapors. In some cases, for example, in
heating of organic compounds for thermal decom-
position only, the process may be accomplished
without air movement through the drier.
The direct rotary drier has flights, which lift the
material and shower it down through the gas stream
as shown in Figure 260. Thus, it has a very high
potential for dust emissions. It cannot be used for
drying fine materials because loss of product would
be excessive. Indirect rotary driers have a much
lesser tendency to emit dust. They are the usual
choice when a continuous drier for powdery mate-
rial is required.
In I960, the Barber-Greene Company completed a
comprehensive testing program on full-scale rotary
drier? to evaluate the effects of the various de-
sign parameters. Over 600 individual test runs
were completed, and the company spent over
$175. 000 of research funds for the project.
-------
368
MECHANICAL EQUIPMENT
5.
Figure 260. Typical flights used in rotary driers.
The most important of the factors influencing drier
selection and performance that were varied or held
constant included: Tonnage rate, moisture content,
air flow rate through the drier, fuel oil rate to the
burner, air flow rate to the burner, drum slope,
drum diameter, drum length, and lifting flight de-
sign and arrangement. Some of the important re-
sults of this investigation are shown graphically in
Figures 261 through 264. While these results were
intended primarily for use in-the ajsphaltic concrete
industry, they may also be applied to similar rotary
driers for other materials in other industries.
Some of the conclusions drawn from this investiga-
tion are summarized and listed as follows.
1. Dust carryoutincreased proportionally to the
square of the gas exhaust volume as the vol-
ume was increased in the same drum.
2. On driers of the same length with the same
drum gas velocity and with other factors held
constant, the maximum production capacity
varied indirect ratio to drum cross-sectional
area.
3. An increase in drum gas velocity permitted
an increase in maximum production capacity,
but on a less than direct ratio.
4. Thermal efficiency-was a constant if the drier
was properly balanced and operated, regard-
less of drier size, diameter, length, or drum
gas velocity.
6.
In a conventionally designed drum, a particle
spent only a fraction of its time in the veil
suspension while in the drum --usually not over
3 to 5 percent. For the remaining time, the
particle cascaded at the bottom of the drum or
rode up in the flight pocket.
Flights in a drum usually retarded rather than
increased the flow of materials through the
drums.
o
6
/
0 2
00 700
/
0 4
INCREASE 1
800
DRUM
/
,21
0 E
N DRUM GAS V
900
GAS VELOCIT
,
/
0 i
ELOCITY. ',
1.000
Y fpm
/
/
0 1
.100 1
DO
00
Figure 261. Dust carryout versus drum gas velocity.
Example: -An increase of 501 in gas velocity from
600 to 900 fpm increases dust carryout by 125%
(Barber-Greene Company, 1960).
Flash Driers
In a flash drier, or pneunaatic conveying drier,
moisture is removed by dispersing the material
to be dried in a hot gas zone followed by convey-
ing at high velocities. The drier consists of a
furnace or other source of hot gases, a device for
dispersing the •wet material in the gases, a duct
through which the gases convey the material, and
a collection system for removing the dry product
from the gas stream. In the simplest type of sys-
tem, a screw conveyor drops the material directly
into a duct, as shown in Figure 265. Only free-
flowing materials can be handled this way. Some
recycled dry product often must be mixed "with the
wet material in order to achieve good dispersion.
-------
Driers
369
100
80
£ 60
- 40
20
0 20 40 60 60 100
INCREASE IN DRUM CROSS-SECTIONAL AREA, '/.
Figure 262. Drier production capacity versus drum cross-
sectional area. Example: A 50% increase in cross-
sectional area increases drying capacity by 50$ (Barber-
Greene Company, 1960).
150
125
20.5
T
-*50'/. I CREASE
15 20 25
DRUM LENGTH, ft
Figure 263. Drier production capacity versus drum length.
Example: A 50% increase in drier length, from 20 to 30 feet,
increases drying capacity by 20.5% (Barber-Greene Company
1960).
A cage mill is often used as the dispersing device.
Flash drying is often combined with fine grinding
as shown by the system in Figure 266.
Spray Driers
A spray drier is a device in which atomized par-
ticles of a solution, slurry, or gel are dispersed
in a hot gas zone (Marshall and Friedman, 1950).
The drier consists of a drying chamber, a source
of hot gases, a device for atomizing the feed, and
a means of separating the dry product from the ex-
haust gases. The last item is the one of concern
here.
Atomization is achieved by three devices: Centrif-
ugal discs, high-pressure nozzles, or two-fluid
nozzles. Centrifugal discs rotate at high speed in
a horizontal plane. The liquid is fed to the center
and discharged at the periphery as a fine spray.
High-pressure nozzles contain a. very small orifice
through which the liquid is forced at a very high
pressure. Particle size is controlled by amount
of pressure and size of orifice. Two-fluid nozzles
use air or steam under moderate pressure to atom-
ize the liquid. The fluids are fed by separate lines
to the nozzle where they impinge in a variety of
different ways to produce a spray.
The hot gases for spray driers are usually obtained
from a direct-fired air heater using natural gas or
fuel oil. In some cases waste flue gas from a boiler
is used. When carbon dioxide must be excluded
from the drying atmosphere, steam coils are used
to heat the air.
The drying chamber in some spray driers is shaped
like a cyclone separator and serves as a primary
product collector. In other types the drying cham-
ber acts as a settling chamber to collect the bulk
of the product. Sometimes, all the product is car-
ried out in the exhaust gases and collected in an
external product collector. The product collector
is nearly always a cyclone separator followed by a
secondary collector where needed.
- 60
29 1
^
^
/
X;
0 20 40 60 8
INCREASE IN DRUM GAS VELOCITY. '/,
00 700 800 900 1 000
DRUM GAS VELOCITY, tpm
^
0 100
100 1 200
Figure 264. Drier production capacity versus drum gas
velocity. Example: An increase of 50% in gas velocity,
from 600 to 900 fpm, increases drying capacity by 29. U
(Barber-Greene Company, 1960).
-------
370
MECHANICAL, EQUIPMENT
AIR FILTER
VENT FAN
EXPANSION JOINT
XCLEANOUT DOOR
Figure 266. Flash drying combined with size reduction
(Combustion Engineering, Inc., Windsor, Conn.).
Figure 265. Simplest type of flash drying system (Com-
bustion Engineering, Inc., Windsor, Conn.).
Other Types of Driers
The following types of driers usually emit negligible
amounts of dust. In some operations, however,
organic vapors and mists may constitute a problem.
Tray and compartment driers consist of a chamber
in which heated air circulates over the wet mate-
rial until the material reaches the desired mois-
ture content. Granular material, filter cakes,
pastes, and slurries are placed in trays, which are
put on stationary or movable racks, as shown in
Figure 267. Other materials are stacked or hung
on racks. The vertical turbodrier can be classi-
fied as a continuous tray drier. It consists of a
vertical, cylindrical housing with circular trays
mounted on a frameworkthat slowly revolves. Ma-
terial fed to the top tray is leveled by stationary
knives and, after about seven-eighths of a revolu-
tion, is pushed through a slot to the tray below,
where the procedure is repeated. Airflow across
the trays is produced by fans mounted on a central
shaft. Heating coils at the periphery of the housing
heats the air as it is recirculated.
Figure 267. Tray drier (j.p. Devme Mfg. Company,
Pittsburgh, Pa.).
Agitated pan driers consist of abowl-shaped vessel,
steam-jacketed on the bottom and part way up the
sides, with stirrer or scra.per blades to keep the
material agitated. The top may be open for atmo-
spheric drying or provided with a cover for vacuum
drying.
-------
Driers
371
Rotary vacuum driers are of two types. One type
consists of a stationary, jacketed cylinder mounted
horizontally with agitator blades mounted on a cen-
tral revolving shaft. Material is charged through
a manhole at the top and discharged through a man-
hole at the bottom. Another type of vacuum rotary
drier consists of a rotating, jacketed cylinder with
vacuum applied through hollow trunnions.
THE AIR POLLUTION PROBLEM
Air contaminants that may be emitted from driers
are dusts, vapors, smoke, and odors. The nature
of the emissions is determined by the material
being dried and by the operating conditions.
Dust can be a problem, in any drier in which the
material is agitated or stirred during the drying
process. Drier types that can be prolific dust pro-
ducers are direct-fired rotary driers, flash driers,
and spray driers. Types that produce less dust
are indirect-heated rotary driers, pan driers, and
cylinder driers. Other types that may emit no dust
include tray driers, sheeting driers, and driers
for products such as lumber, "bricks, ceramic
ware, and so forth.
When an organic liquid is to be removed from a
material, the emissions may include vapors, mists,
odors, and smoke.
HOODING AND VENTILATION REQUIREMENTS
Direct-fired rotary driers are usually equipped
•with an induced-draft fan or with a stack of suffi-
cient height to provide draft for the combustion
process. The ventilation requirement is equal to
the volume of the products of combustion, plus va-
pors driven off from the product, plus sufficient
excess to ensure an adequate indraft velocity through
all openings.
Flash driers and spray driers have no ventilation
requirement as such. The exhaust fan is usually
placed at the product discharge end of the system,
and the entire system is under negative pressure,
which precludes emissions, except for the final
collector.
AIR POLLUTION CONTROL EQUIPMENT
In general, three types of controls are used on
driers: Dust collectors, condensers, and after-
burners. The type of material being dried deter-
mines the kind of control device needed. Dust col-
lectors are the most frequently used type since dust
is usually the problem. All types of dust collec-
tors are used, depending upon the amount and par-
ticle size of the dust emitted. Condensers are
used -when a material'wet with an organic solvent
is dried. Afterburners are used to control smoke.
combustible particulate matter, vapors, and odors.
Dust Control
The types of dust collectors most commonly used
on driers are cyclones, scrubbers, and baghouses.
If there is only a. negligible amount of dust in the
effluentfiner than 20 microns, a cyclone is an ade-
quate collector; otherwise, it is not. Cyclones are
extensively used ahead of scrubbers in order to
collect product materials in the dry form. A bag-
houseisthe best collector if the exit gases can be
maintained above the dewpoint and the dust is not
sticky. In some cases a scrubber is the only fea-
sible control device.
The primary product collector for a flash drier is
nearly always a cyclone separator. When fine ma-
terials are dried or when grinding is incorporated
in the circuit, a baghouse following the cyclone is
normally required, both to prevent excessive loss
of product and to ensure control of air pollution.
The size of the baghouse is determined by the vol-
ume of the drying and conveying gases. The bag
material that can be used should be determined by
the temperature at the baghouse. In some cases
the temperature may be low enough to permit use
of cotton or wool, but in most cases Dacron or Or-
ion is better.
Baghouses and scrubbers are used as secondary
collectors for spray driers. A very efficient sec-
ondary collector is usually best in areas having a
strict limitation on particulate emissions. The
closeness of approach to the dewpoint determines
the suitability of a baghouse. When the feed liquid
is dilute and requires concentration, it can be used
as the scrubbing liquid in a wet collector and there-
by increase the concentration and recover the dust
in the exhaust gases at the same time.
Drying With Solvent Recovery
When a liquid other than water is to be removed
from a material, recovery of the solvent is fre-
quently desirable in order to lower costs, prevent
a safety hazard, and eliminate air pollution (Mar-
shall and Friedman, 1950). The value of the sol-
vent may require its recovery for economic oper-
tion. If the solvent is a toxic or flammable ma-
terial, health and safety considerations may dictate
its recovery.
Vacuum driers are well suited to recovery of sol-
ventvapors. The vapors are removed under slight
or high vacuum with only a small quantity of air,
which is originally present or leaks into the sys-
tem during operation. If dust is carried over, che
vapors are drawn through a dust collector to pre-
vent losses of product and fouling of condenser
surfaces. The collector is usually a scrubber in
order to preclude difficulties with condensed va-
por. In some cases, where condensation at the
collector can be prevented, bag lilters are used.
-------
372
MECHANICAL, EQUIPMENT
From the dust collector the vapors usually pass
through a surface condenser where the solvent is
collected in a barometric leg or a tank kept at a
low pressure. The gases leaving the condenser
consist of the inert gases that have leaked in plus
enough solvent vapor to form a saturated mixture
at the condenser's temperature and pressure.
Vacuum pumps, both rotary and reciprocating, and
steam jets are used as vacuum sources. The vac-
uum at the condenser must be adjusted so that the
boiling point of the solvent is well above the tem-
perature attainable in the condenser. Otherwise,
solvent recovery willbe poor or will be reduced to
zero if theboiling point is brought down to the con-
denser temperature. Sometimes, recovery can be
improved by placing an additional small condenser
on the outlet of the vacuum pump.
Drying with solvent recovery can be accomplished
with direct drying under certain circumstances.
Heated air or inert gases are used and the vapor-
ized solvent is recovered in a condenser. The non-
condensable gases are usually recirculated through
aheater. If air is used, the solvent concentration
must be kept well below the lower explosive limit.
Since the amount of inflammable solvents that could
be condensed at these concentrations and at fea-
sible condenser temperatures is negligible, this
method is restricted to noninflammable solvents
such as perchlorethylene, carbon tetrachloride,
and so forth. An inert gas atmosphere is needed
for recovering inflammable solvents from direct
driers. Since the cost of maintaining an inert at-
mosphere is considerable, this method is not wide-
ly used.
Smoke and Odor Emissions
Direct-fired rotary driers, when drying certain
organic materials, sometimes emit smoke and
odors. Cannery or brewery "wastes used to pro-
duce fertilizer or animal food are examples. Most
of these driers can be operated without excessive
air-contaminating emissions under the proper con-
ditions. If feed rate and temperature are properly
adjusted, a dry product results -without any local-
ized overheating. If, however,the feed rate is ex-
cessive, the required higher temperature causes
localized overheating and partial decomposition of
the product, resulting in the emission of smoke and
odors. Scrubbers are usually used to control dust
emissions from these driers, but are not adequate
for controlling smoke and odors.
Another drying operation that emits smoke is the
removal of cutting oils from metal turnings and
chips. This operation nearly always produces
enough smoke to violate smoke prohibitions. An
afterburner is the only feasible control. A tem-
perature of at least 1,200°F is required in the af-
terburner for complete smoke control. Tempera-
ture control in the drier is rather critical. The
temperature must be high enough to vaporize the
oil but not high enough to cause it to burn in the
drier since this would cause the chips to melt or
oxidize. A mechanical feeder is almost a neces-
sityto secure good control of the operation. Hand
feeding nearly always results in poor temperature
regulation and1 in undried and burned chips.
WOODWORKING EQUIPMENT
Woodworking machines produce large quantities
of waste sawdust, chips, and shavings that must
be removed from the equipment site. For this
purpose, exhaust systems are constructed that also
alleviate conditions tending to impair health of
operating personnel, collect wastes that may have
a resale value, and reduce fire hazards. The use
of an exhaust system, however, requires a dust
collector of some type to prevent an air pollution
problem.
EXHAUST SYSTEMS
Exhaust systems are used with many types of wood-
working machines capable of producing appreciable
sawdust, chips, or shavings by drilling, carving,
cutting, routing, turning, sawing, grinding, shred-
ding, planing, or sanding wood. Machines include
ripsaws, handsaws, resaws, trim saws, mitre
saws, panel saws, out-off saws, matchers, stick-
ers, grinders, moulders, planers, jointers, spin-
dle sanders, edge sanders, tenoners, mortisers,
wood hogs (hammer mills), groovers, borers,
dovetailers, and others. Exhaust systems serv-
ing wood hogs might more properly be termed pneu-
matic conveyors . Inpractice, however, woodhogs
are most often found connected to exhaust systems
that also serve other wood-working machines.
Exhaust systems serving various woodworking ma-
chinery are most frequently used at lumber mills,
furniture manufacturers, pi an ing mills, furniture-
refinishing shops, model shops, maintenance shops,
cabinet shops, sash and door manufacturers, and
carpenter shops. Many of the larger systems han-
dle several tons of -waste products per day. One
ofthe largest in the Los Angeles area burns 15 to
20 tons per day in a multiple-chamber incinerator.
One ton of waste sawdust, chips, and shavings oc-
cupies approximately 150 to 200 cubic feet of space.
Construction of Exhaust Systems
Atypical woodworking exhaust system consists of
hoods for the pickup of -wood dust and chips at the
machines, ductwork, a collection device (usually
a cyclone), a storage bin, and a fan blower to supply
air for conveying purposes. Almost all exhaust
systems are constructed of galvanized sheet metal.
-------
Woodworking Equipment
373
THE AIR POLLUTION PROBLEM
Wood-working exhaust systems are somewhat unique
in that they are almost always equipped with air
pollution control devices. If they were not so
equipped, the entrained sawdust -would result in
excessive opacities and dust loadings in exit gases
and could easily cause a local nuisance. Air con-
taminant emissions from systems such as these
are functions of the particular dust encountered
and the particular control device employed. The
dust particles are not excessively small in most
systems, and elaborate devices are not usually
required.
Particles emitted by woodworking machines vary
in size from less than 1 micron to chips and curls
several inches long. Hammer mill-type wood
hogs emit particles running the complete size
range, while sanders generate only very small
dustparticles. Wood waste particles from most
other machines are of larger size and greater
uniformity, seldom less than 10 microns. Other
factors determining particle size are the type of
wood processed and the sharpness of the cutting
tool. Hardwoods tend to splinter and break, yield-
ing smaller particles than soft woods do, which
tend to tear and shred. A dull cutting tool in-
creases tearing and shredding and produces larger
particle sizes.
Generally, the configuration of waste particles is
of little importance. There are, however, in-
stances where toothpick-like splinters and curls
have presented difficulties' in collection and stor-
age and in the emptying of storage bins.
HOODING AND VENTILATION REQUIREMENTS
Sawdust weighs from 7 to 15 pounds per cubic foot.
The minimum recommended air volume for each
pound of wood waste to be conveyed is 45 cubic
feet or, expressed differently, is 1, 500 cfm per
ton-hour of waste. In actual practice the air vol-
ume is usually much higher because of exhaust
velocity requirements.
Velocities recommended for conveying this mate-
rial range from 3, 500 to 4, 500 fpm, with most
ducts sized to give a velocity of 4, 000 fpm. In
practice, velocities of from 2, 000 to 6, 000 fpm
are encountered.
Table 108 lists recommended exhaust volumes for
average-sized woodworking machines. In each
case the duct is sized to give a conveying velocity
of 4, 000 fpm. Some modern high-speed or extra
large machines produce such large volumes of
wastes that greater exhaust volumes must be used.
Similarly, some small machines of the home wood-
shop or bench type may not require as large a
volume as that recommended.
Hooding devices vary somewhat, depending upon
the type of woodworking machine, and are of stan-
dard design throughout the industry. In most cases,
the hoods are merely scooped openings that catch
the wood waste as it is thrown from the saws or
blades of the machine. In design practice, no
problems should be encountered if air volumes
are chosen from those shown in Table 108,and if
hoods are shaped to cover the area assumed by
the thrown particles. Locating the hood as close
to the saw or blade as possible is advisable.
AIR POLLUTION CONTROL EQUIPMENT
The simple cyclone separator is the most common
device used for collecting wood dust and chips
from -woodworking exhaust systems. For these
exhaust systems, cyclones outnumber all other
devices by a large margin. Properly designed
cyclones have been found satisfactory for use -with
exhaust systems at cabinet shops, lumber yards,
planing mills, model shops, and most other wood-
processing plants.
Higher efficiency centrifugal collectors separate
smaller particles, but these devices are not com-
mon to wood-working systems. The main advan-
tage of simple cyclones over most other collec-
tion devices is simplicity of construction and ease
of operation. They are relatively inexpensive, re-
quire little maintenance, and have only moderate
power requirements.
The size and design of woodworking exhaust sys-
tem cyclones varies -with air volume and the type
of wood waste being handled. Where fine Sander
dust predominates, cyclones should be of high-
efficiency design -with diameters not greater than
3 feet. Coarse sawdust, curls, and chips, such
as are produced with ripsaws, moulders, and drills,
can be effectively collected with low-efficiency
cyclones up to 8 feet in diameter. Most wood-
working exhaust systems are employed to collect
a mixture of -wood -waste including both fine and
coarse particles. The exhaust system designer
must, therefore, carefully consider the quantities
of each type wood waste that will be handled. The
presence of appreciable percentages of coarse
particles in most systems allows the use of low-
and medium, efficiency cyclones , in which the pres-
sure drop does not normally exceed 2 inches of
"water column.
Baghouses are sometimes used -with wood-working
exhaust systems. Their use is relegated to those
systems handling fine dusts such as wood flour or
where small amounts of dust losses cannot be tol-
erated in the surrounding area. The efficiency of
baghouses on-woodworking exhaust systems is very
high--99 percent or more. They can be used to
filter particles as low as I/10 micron in size. In
some installations lower efficiency collectors such
-------
374
MECHANICAL, EQUIPMENT
Table 108. EXHAUST VOLUMES AND DUCT SIZES FOR
WOODWORKING EQUIPMENT
(Committee on Industrial Ventilation, I960)
Self-feed table ripsaw
Saw diameter, in.
Up to 16
Over 16
Self-feed, not on table
Gang ripsaws-
Saw diameter, in.
Up to 24
Over 24 up to 36
Over 36 up to 48
Over 48
ALL OTHER SAWS, includ-
ing table saws, mitre saws.
variety saws, and swing saws.
Saw diameter, in.
Up to 16
Over 16 up to 24
Over 24
Variety saw with Dado
head
Vertical belt sanders- (rear
belt and both pulleys enclosed)
and top run horizontal belt
sanders-
Belt width, in.
Up to 6
Over 9 up to 14
Over 14
Swing arm sancler:
Disc sanders' diameter, in.
Up to 12
Over 12 up to 18
Over 18 up to 25
Over 26 up to 32, 2 pipes
Over 32 up to 38, 2 pipes
Over 38 up to 48, 3 pipes
Triple-drum sanders:
Length, in.
Less than 30
Over 30 up to 36
Over 36 up to 42
Exhaust volume, cfm
Bottom
hood
440
550
800
550
800
1, 100
1, 400
350
440
550
550
440
550
800
1, 100
440
350
440
550
350
350
550
350
1, 100
1, 400
1,800
Over 42 up to 48 I 2, 200
Over 48
Horizontal belt sanders
Belt width, in.
Where bottom run of belt is
usec1
Up to 6
Over 6 up to 9
Over 9 up to 14
Over 14
3, 100
440
550
800
1, 100
Top
hood
350
350
550
350
440
550
550
each
and 550
350
350
440
550
Duct diameter, in.
Bottom Top
hood hood
4-1/2 4
5 4
6 5
5 4
6 4-1/2
7 5
8 5
4
4-1/2
5
5
4-1/2
5
6
7
4-1/2
4
4-1/2
5
4 each
4 and 5
5
4
7
8
9
10
12
4-1/2 4
5 4
6 4-1/2
7 5
Band saws and band
resaws:
Blade width, in.
Up to 2
Over 2 up to 3
Over 3 up to 4
Over 6 up to 8
Jointers:
Knife length, in.
Up to 6
Over 6 up to 12
Over 12 up to 20
Over 20
Single planers-
Knife length, in.
Up to 20
Over 20 up to 26
Over 26 up to 36
Over 36
Double planers
Knife length, in.
Up to 20
Over 20 up to 26
Over 26 up to 36
Over 36
Molders, matchers,
and sizers
Up to 7
Over 7 up to 12
Over 12 up to 18
Over IS up to 24
Over 24
Sash stickers
Tenoner
Automatic lathe
Forming lathe
Chain mortise
Dowel machine
Pane! raiser
Dovetail and lock
corner
Pulley pockets
Pulley stile
Glue jointer
Gainer
Router
Hogs
Up to 12 in. wide
Over 12 in. wide
Floorsweep
(6 to 8 in. dia)
£xhaust volume, cfm
Down run
350
550
800
1 , 100
1,400
350
440
550
600
500
800
1, 100
1, 400
Bottom
hood
550
550
800
1, 100
Bottom Top
hood hood
440 550
550 800
800 1, 100
1, 100 1, 400
1,400 1,770
550
440 to
Up rim
350
350
550
5 5C
550
Topi
hood
550
800
1 , 100
1 , 1 00
Right Left
hood hood
350 350
440 440
550 550
800 800
1,100 1, 100
1 A nr\
See moulder
800 to
350 to
350
350 to
550
550 to
550
550
800
350 to
350 to
1, 400
3, 100
800 to
5,000
1, 400
800
800
1, 400
800
1, 400
Duct diameter, in.
Down run Up run
4 4
5 4
6 5
7 5
8 5
4-1/2
5
6
5
6
7
8
Bottom Top
hood hood
5 5
5 6
6 7
7 7
Bottom Top Right Left
hood hood hood hood
4-1/2 54 4
5 6 4-12 4-1/2
6 755
7 866
8 977
5
6 to 15
4 to 8
4
4 to 6
4-1/2
4 1/2 to 6
4-1/2
4-1/2
6
4 to 8
4 to 6
8
12
6 to 8
as cyclones and impingement traps are installed
upstream to remove the bulk of entrained partic-
ulates before final filtering inabaghouse. Filter-
ing velocities of 3 fpm are satisfactory.
Disposal of Collected Wastes
Wood dust and chips collected with exhaust sys-
tems must be disposed of since they present a
storage problem and a fire hazard. Very often
a profit can be realized from this waste materi-
al. Wood wastes can be used productively for
things such as:
I. Plastics bulking agent for products such as
plastic wood, masonite, and so forth;
2, pressed woods such as firewood, fiberboard,
Firtex, and others:
3 , soil additives ;
-------
Rubber -Compounding Equipment
375
4. smokehouse fuel--hardwood sawdust is burned
to produce smoke in the processing of bacon,
ham, pastrami, and so forth;
5. floor sweep — sawdust with and -without oil is
spread on floors before they are swept to help
hold dust particles;
6. wood filler--sawdust canbe mixed -with -water
resins and other liquids and used as -wood
filler;
7. floor cover in butcher shops, restaurants,
and so forth;
8. waste heat boilers --heat can be recovered
from incinerator flue gases to generate steam,
hot water, and so forth.
When no productive disposal method can be used,
•wood waste is destroyed or removed in the most
convenient manner. Wood dust and chips collec-
ted by the woodworking exhaust systems can be
destroyed smokelessly by burning in a multiple-
chamber incinerator. Single -chamber incinera-
tors, for example, silo-type or teepee-type incin-
erators, cannot~be controlled adequately for satis-
factoryair pollution abatement. Generally, wood
-waste is conveyed from the collection device to the
incinerator by a pneumatic or a mechanical con-
veying system. In areas where the services of a
eut-and-cover dump are available, disposal by in-
cineration is usually not economical.
RUBBER-COMPOUNDING EQUIPMENT
INTRODUCTION
Rubber in its raw state is too plastic for most
commercial applications, and its use is, there-
fore, limited to a few items such as crepe rub-
ber shoe soles, rubber cements, adhesives, and
so forth (Shreve, 1945), Through a curing pro-
cess termed vulcanizing, raw rubber can be made
to lose plasticity and gain elasticity. By corn-
pounding the raw rubber with various types and
amounts-of additives before the vulcanizing, ten-
sile strength, a"brasion resistance, resiliency,
and other desirable properties can be imparted to
the rubber- The proportions and types of addi-
tives (including vulcanizing agents) compounded
into the raw rubber, and the vulcanizing temper-
ature, pressure, and time are varied in accor-
dance with the properties desired in the final prod-
uct. After the rubber is compounded, it is formed1
into the desired s~hape and then cured at the re-
quired temperature. In the forming steps, large
amounts of organic solvents are often used in the
form of rubber adhesives. Since the solvent emis-
sions are not controlled, they will not be dis-
cussed further ir, this section.
Additives Employed in Rubber Compounding
Types of additives that are compounded into the
rubber may be classified as vulcanizing agents,
vulcanizing accelerators, accelerator activators,
retarders, antioxidants, pigments, plasticizers
and softeners, and fillers. Examples of addi-
tives that maybe encountered in rubber compound-
ing are tabulated by type (Kirk and Othmer, 1947).
1. Vulcanizing agents. Sulfur -was originally
considered essential to vulcanizing and, though
vulcanizing is now possible -without it, sulfur
or sulfur compounds such as sulfur mono-
chloride are widely used. Selenium and tel-
lurium can also be used for this purpose.
2. Vulcanizing accelerators. Aldehyde-amines,
guanidines, and thiuram sulfides are used to
decrease the time and temperature required
for vulcanization.
3. Accelerator activators. Zinc oxide, stearic
acid, litharge, magnesium oxide, and amines
supplement the accelerators and, in addition,
modify finished product characteristics, for
example, they increase the modules of elas-
ticity.
4. Retarders, Salicylic acid, "benzole acid, and
phthalic anhydride retardthe rate of vulcaniz-
5. Antioxidants. Several organic compounds,
mostly alkylated amines, are used to retard
deterioration of the rubber caused by oxida-
tion and improve aging and flexing ability.
6. Pigments. Carbon black, zinc oxide, mag-
nesium carbonate, and certain clays are used
to increase tensile strength, abrasion resis-
tance, and tear resistance. Iron oxide, tita-
ni-um oxide, and organic dyestuffs are used to
color the rubber.
7. Plasticizers and softeners. Resins, vegetable
and mineral oils, and waxes are used to im-
prove resiliency, flexibility, and mixing and
processing characteristics.
8. Fillers. Whiting, slate flour, barytes, and
some of the pigments previously mentioned are
used to improve processing properties and
lower the cost of the finished product.
In the compounding of blends, the accelerators are
added first to the mass of raw rubber being milled
or mixed. Then a portion of the plasticizers (if
present in the blend recipe) are added, followed
by the reinforcing pigments, the remainder of the
plasticizers. the antioxidants, andanvinert fillers
or coloring agents. The vulcanizing agent is al-
ways introduced as the last ingredient.
-------
S76
MECHANICAL EQUIPMENT
In order to be effective in imparting various chosen
characteristics, all additives employed in a blend
must be homogeneously dispersed throughout the
blend. The two most commonly employed pieces
of equipment for blending rubber and additives are
rubber mills and Banbury mixers.
A typical rubber mill is shown in Figure 268. The
two rolls rotate toward each other at different
speeds, creating a shearing and mixing action. Raw
rubber is placed in the mill, and the additives are
introduced, generally one or two components at a
time. Additives may be finely divided solids or
liquids.
Another device commonly used for compounding
rubber stock is the Banbury mixer. Figure 269
shows cross-sections of two typical Banbury mix-
ers. Each consists of a completely enclosed mix-
ing chamber in which two spiral-shaped rotors, re-
volving in opposite directions and at different speeds,
operate to keep the stock in constant circulation.
A ridge between the two cylindrical chamber sec-
tions forces intermixing, and the close tolerances
of the rotors with the chamber walls results in a
shearing action. A floating weight in the feed neck
confines the batch within the sphere of mixing. This
combination of forces produces an ideally homo-
geneous batch.
THE AIR POLLUTION PROBLEM
Sources of air pollution from the mills are (1) fine-
ly ground dusts introduced as additives, (Z) fumes
generated by mechanical working of the batch by
the mill rollers, (3) oilmists from liquid additives,
and (4) odors. A major source of air pollution
from rubber mills occurs when the finely divided
dusts are introduced into the batch. Opacity of the
resultant dust cloud depends upon the character,
density, and particle size of the additive. Opacity
generally ranges from 5 to 50 percent, persisting
from a few seconds to several minutes.
Uncontrolled emissions vary from a negligible
amount to about 1 pound per hour, depending upon
the size of the mill, the size of the batch, and the
composition of the mix. Emissions average ap-
proximately 0. 5 pound per hour. Solvent vapors
emanating from the mix are ordinarily uncontrolled
and enter the atmosphere.
Introduction of ingredients into a Banbury mixer
is effected through the feed hopper. It is at this
point, during charging, that air contaminants may
enter the atmosphere. Emissions are similar to
those from the mills. In general, most of the dry
ingredients are added at the Banbury mixer, car-
bon black being the most troublesome.
Figure 268. Rubber mill (Parrel Corporation, Ansonia, Conn.).
-------
Rubber - Compounding Equipment
377
SPRAY SIDE
CONNECTION FOR
EXHAUST FAN
TO REMOVE DUST
BIN-TYPE HOPPER
FEED HOPPER DOOR //
AIR-OPERATEO
SINGLE-SLOPE FLOATING
•EIGHT IN DOIN POSITION
PLEXIGLASS COVER
CONNECTION FOR
POIDER FEED DUCT
ROTORS CORED FOR
CIRCULATION OF
COOLING MTER
OR STEAM
SLIDING DISCHARGE
AIR-OPERATED
FEED HOPPER DOOR
AIR-OPERATED
EXTENDED NECK
SIDES AND ROTORS
CORED FOR
CIRCULATION OF
COOLING WATER
OR STEAM
SLIDING DISCHARGE DOOR
AIR-OPERATED
SINGLE-SLOPE
FLOATING WEIGHT
HOODING AND VENTILATION REQUIREMENTS
Generally, rubber mills are provided with hoods,
as shown in Figure 270. The primary purpose of
ahood is to carry away heat generated by the mechan-
ical mixing action. As a secondary consideration,
the exhaust hood removes dust, fumes, and mists
emitted from the rolls. Sufficient volume should
be exhausted to give an indraft velocity of 100 fpm
through the openface of the enclosure. Figure 269
shows the exhaust provisions supplied with a stan-
dard Banbury mixer. If an unusual dust problem
is encountered, supplementary hooding can be added.
The minimum required exhaust volume is equal to
200 cfmper square foot of mixer charging opening.
Figure 270. Rubber mill with exhaust hood (National Seal
Division, Federal-Mogul-Bower Bearings, Inc., Downey,
Calif.).
AIR POLLUTION CONTROL EQUIPMENT
Figure 269. Two models of Banbury mixers (Farrel
Corporation, Ansonia, Conn.).
In general, emissions from Banbury mixers and
rubber mills are in a finely divided form and small-
er than 15 microns. Inertial separators are not,
therefore, effective control'devices for this ser-
vice. The most common control device employed
-------
378
MECHANICAL EQUIPMENT
is the baghouse; a we 11-designed baghouse can be
operated with 98 to 99. 5 percent efficiency.
Standard cotton sateen bags are adequate at a fil-
tering velocity of 3 fpm. In some cases scrubbers
have also proved satisfactory and advantageous in
scrubbing out some oil vapors and oil mists that
may be present in some blends.
ASPHALT ROOFING FELT SATURATORS
Asphalt saturators are used to prepare asphalt-
saturated felt in the manufacture of roofing paper
and roofing shingles. The roofings are made by
impregnating a vegetable felt base -with asphalt de-
rived from petroleum. The felt is made from fi-
brous vegetable matter and generally contains 5
to 10 percent water. The asphalts, known in the
industryas saturants, are derived as still bottoms
from petroleum crude oil, and are semisolids -with
softening points of 100° to 180°F.
DESCRIPTION AND OPERATION
Asphalt-saturated felt is manufactured in high-
speed, continuous-operating machines, referred
to as asphalt saturators. The asphalt saturator
consists of a dry looper, an asphalt spray section,
a saturating tank, and a wet looper. The felt is
continuously fed from rolls into the dry looper
where it is arranged over rollers into a series of
vertical loops used as live storage in the process
to permit maintenance of feed at a uniform rate to
the saturating process during roll changes. The
liquid asphalt at 400° to 450 °F may then be sprayed
on one side of the felt. This spray of hot asphalt
drives moisture in the felt out the unsprayed side
and prevents the moisture from forming blisters
when the felt is saturated. After being sprayed,
the felt passes through a tank of molten asphalt
that saturates the felt. The saturated felt then
enters the wet looper where the material is ar-
ranged over another set of rollers into long, ver-
tical loops to permit cooling of the asphalt. The
web of saturated felt is then rolled up from the
discharge end of the wet looper for use as roofing
felt or building paper, or a small quantity of bi-
tuminous material and mica schist or rock granules
are applied to the surface to make composition
roofing paper and shingles. Figure 271 is a sche-
matic drawing of an asphalt roofing saturator.
THE AIR POLLUTION PROBLEM
Asphalt is generally applied to the felt at 400° to
450°F. This relatively high temperature causes
the lower boiling components of the asphalt to va-
porize. Inaddition, moisture from, the felt vapor-
izes in the hot asphalt, resulting in steam distilla-
tion of the asphalt. These two vaporization mech-
anisms result in the creation of vaporous as well
as particulate air contaminants, the latter being
in the form of a highly opaque mist when the sat-
urator is in operation. When felt is not being pro-
cessed, the rate of air contaminant emissions de-
creases somewhat, but the opacity of the mist is
usually well above 50 percent over the saturation
tank. Figure 272 shows the mist emissions at the
asphalt saturator tank. Additional vapors andmists
are emitted from the saturated felt in the wet loop-
er. The mass emission rate is a function of felt
feed rate, feltmoisture content, number of sprays
used, and asphalt temperature, all of which are
highly variable. It has been found, however, that
the total contaminant emission rate is about 20 to
70 pounds per hour.
HOODING AND VENTILATION REQUIREMENTS
The points of air contaminant emission are the
asphalt presaturator sprays, the saturator tank,
the wet looper. Hoods for collecting the emissions
should be installed so that there is a single con-
tinuous enclosure around the points of emission,
extending down to the floor. Since operating per-
sonnel must have access to the saturator for oper-
ating adjustments, doorways or other provisions
for entrance in the hood must usually be supplied.
These should be kept as small as possible. In ad-
dition, openings in the hoods must be provided for
the entrance of felt and exit of the saturated mate-
rial. These openings should be as close to the
floor as possible. Experience indicates that a
minimum indraft velocity of 200 ipm is required
at all hood openings. Air volumes handled by the
exhaust system vary with hood design and saturator
size but are about 10, 000 to 20, O'OO scfm. Figures
273 and 274 illustrate hooding devices for an asphalt
roofing saturator.
AIR POLLUTION CONTROL EQUIPMENT
The large volume of air required in controlling the
saturator equipment generally makes incineration
impractical. Baghouses, spray scrubbers, and
two-stage electrical precipitators have been used
as air pollution control equipment for asphalt sat-
urators. Although more expensive in first cost,
electrical precipitators are the most efficient and
probably the most practical control device.
Low-Voltage Electrical Precipitators
The low-voltage, or two-sta.ge, electrical precip-
itator precededby a spray sc rubber as a per eel e an -
er gives relatively high collection efficiency as -well
as a substantial reduction in the opacity of the sat-
urator effluent. Fig-ures 275 and 276 show equip-
ment of this type. Table 10'9 shows the test results
on a scrubber precleaner followed by a two-stage
precipitator.
-------
Asphalt Roofing Felt Saturators
379
DRY
LOOPER
SPRAY
SECTION
ASPHALT
TANK
w
L
El
0(
)P
E
R
TO ROLL PRODUCT
OR SHINGLE
PRODUCT OPERATIONS
SATURANT
TO ASPHALT
—*-
HEATER
Figure 271. Schematic drawing of an asphalt roofing felt saturator.
Figure 272. Emissions from asphalt saturator tank
(Lloyd A. Fry Roofing Company, Los Angeles, Calif.;
Precleaners are usually wet collectors used to de-
crease the amount of oil mist handled by the pre-
cipitator. The precleaner removes particles
more than 1 micron in diameter. Simple spray
scrubbers or inertial scrubbers give adequate pre-
cleaner efficiency. The use of aprecleaner causes
evaporation of water into the warm airstream, ac-
companied by cooling of the airstream. The re-
sult is that the humidity at the lower temperature
approaches saturation, and water condensation in
the precipitator is a possibility.
Design Considerations for Electrical Precipitators
The design parameters of electrical precipitators
controlling asphalt saturators are particularly crit-
ical since many particles to be removed are less
than 1 micron in diameter. Particular attention
mustbe directed to air distribution within the pre-
cipitator and to temperature drop across the unit.
Examination of the theory of electrical precipita-
tors indicates that the time a contaminant particle
remains within the conveying and collector fields
has significant bearing on the precipitator' s effi-
ciency. Because the actual time in the electrical
field is somewhat inconvenient to calculate, this
parameter is usually expressed in terms of super-
ficial velocity, which is based upon the overall
area of the precipitator cells perpendicular to the
direction of airflow. A typical low-voltage, two-
stage precipitator has plate lengths of 8-1/2 inches
in the direction of airflow and plates spaced 5/16
inch apart. The precipitator is operated with 10
to 1 5 kilovolts of ionizing voltage and 6 to 7 kilo-
volts of collector voltage. With a precipitator of
this design, superficial velocities of less than 150
fpm usiaally provide adequate control of emissions
from asphalt saturators.
Air distribution within the precipitator is important
in this type of precipitator application. Poor air
distribution leads to high velocities in some sec-
tions of the precipitator, yielding low overall col-
-------
380
MECHANICAL EQUIPMENT
Table 109. EMISSIONS FROM A WATER SCRUBBER AND LOW-VOLTAGE,
TWO-STAGE ELECTRICAL PRECIPITATOR
VENTING AN ASPHALT SATURATOR
Volume, scfm
Temperature, °F
Emission rate,
gr/scf
Ib/scf
Water vapor, %
Collection efficiency
Scrubber inlet
20, 000
139
0. 416
71.4
3. 7
Scrubber, 71%
Precipitator inlet
20,234
85
0. 115
20
4.9
Precipitator, 50%
Precipitator outlet
20, 116
82
0. 058
10
4.8
Overall, 86%
Figure 273. Asphalt saturator hood at felt feed (Lloyd A. Fry Roofing Company
Los Angeles, Cal i f.), ' ''
lector efficiency. Uneven air distribution can be
caused by duct bends in the exhaust system just
ahead of the precipitator inlet, abrupt transitions
from inlet duct to precipitator housing, and buoy-
ancy effects of warm air. The effect of the buoy-
ancy of the warm air can be compensated for by
the installation of t>erf orated plates at the inlet and
discharge sections of the precipitator. The transi-
tion section from the duct of the exhaust system to
die housing of the precipitator should be carefully
designed to provide for smooth and gradual changes
indirections of airflow. Turbulence and poor ve-
locity distribution due to duct bends in the exhaust
system can be compensated for, in part, by the use
of straightening vanes or of sectioned, adjustable,
perforated plates.
The temperature drop across the precipitator is
another important design consideration. The re-
moval o± water from the felt in the saturator as
well as the water evaporated in the precleaner
increases the water content of the air entering the
-------
Asphalt Roofing Felt Saturators
381
Figure 274. .Asphalt saturator hood at saturator
discharge (Lloyd A. Fry Roofing Company, Los
Angeles, Calif.).
precipitator. The exit conditions of the scrubber
maybe suchthatthe air may be near the dew point
temperature, and the additional temperature drop
in the precipitator of about 5°F maybe sufficient
tc produce some condensation. The presence of
water droplets causes arcing between the elec-
trodes in the precipitator, resulting in a decrease
in collection efficiency. Water condensation can
be minimized by insulating the exhaust system ahead
of the precipitator and the precipitator housing or
byheating the airstream before it enters the pre-
cipitator. The heat added should be just sufficient
to stop the arcing, because excessive heat vaporizes
the oil and prevents its collection in the precipita-
tor.
Electric insulators exposed to the contaminated
airstream accumulate oil and -water. These de-
posits result in electrical leakage with accompany-
ing voltage drop and decrease in precipitator effi-
ciency. Insulators should, therefore, be isolated
from the contaminated airstream by being enclosed
in channels. The channels should be pressurized
slightly by small blowers to prevent infiltration of
contaminant.
Maintenance of Precipitators
The oils collected in the precipitators venting as-
phalt saturator s form tarry materials on the pre-
ASPHALT
StTURATOR
VISCOUS FILTER
INERT IAL SCRUBBER
75-hp EXHAUSTER
, 27,000 cfm
AT 8 in. SP
ELECTROSTATIC PRECIPITATOR
13,000-volt IONIZING SECTION
6,900-volt COLLECTION SECTION
Figure 275. Schematic drawing of electrical precipitator, precleaner, and exhaust system
for an asphalt saturator.
-------
382
MECHANICAL EQUIPMENT
Figure 276. Low-voltage, two-stage electrical precipitator venting an asphalt saturator
(Johns-ManviIle Products Corp., Los Angeles, Calif.).
cipitator's components. These deposits decrease
the efficiency of the precipitator, causing insu-
lating and arcing effects. Proper maintenance is
vital if the precipitator is to eliminate the emis-
sions from the saturator. Maintenance operations
should include the following:
1. The interior components of the precipitator
should be cleaned every 4 to 6 weeks by pres-
sure spraying "with water and detergent.
2. Regular checks should be made of the condi-
tion of all •wires and insulators. Cracked or
broken components should be replaced.
3. Components of the precipitator should be cleaned
down to bare metal about twice a year.
These procedures should not be interpreted as rigid
rules since individual installations vary considerably
in regard to operating time and quantity of material
collected.
Baghouses
Baghouse filters are occasionally used as air pollu-
tion control devices for asphalt saturators, but their
use is limited as a result of maintenance prob-
lems associated with filter bag upkeep and their
high power requirement. Oil collected by the fil-
ter fabric is oxidized and polymerized by the air-
stream, causing plugging of the fabric and increas-
ing of the pressure drop across the filter unit. The
air volume handled by the exhaust system then de-
creases because of increased pressure drop and re-
sults in loss of mist capture at the saturator1 s hood
openings. Another problem associated with bag
filters in this service is the reentrainment of col-
lected oil in the airstream. A cyclone separator
following the filter maybe partially successful as
an entrainment separator. Table 110 shows the re-
sults of a test on a bag filter unit followed by a cy-
clone separator,
Scrubbers
Spray-type scrubbers have met with limited suc-
cess as air pollution control devices for satura-
tors. Some spray scrubbers may have an effi-
ciency, based on weight removed, as high as 90
percent, but the scrubber's effluent may be from
50 to 100 percent opaque, and may thus be in ex-
cess of that allowed by law. This opaque discharge
is due to the extremely low collection efficiency of
spray scrubbers for particles loss than 1 micron in
diameter. These small-diameter particles, when
-------
Solvent Degreasers
383
Table 110. EMISSIONS FROM A BAG FILTER
AND CYCLONE SEPARATOR VENTING AN
ASPHALT SATURATOR
Volume, scfm
Temperature, "F
Emission rate,
gr/scf
Ib/hr
Water vapor, %
Collection efficiency, %
Control equipment
inlet
10, 300
217
0. 768
67. 7
6.4
Control equipment
discharge
10, 300
185
0.289
25. 5
6; 8
| 62. 3
emitted from the scrubber discharge, cause max-
imum light scattering and, therefore, high opaci-
ties. Table 111 shows the results of tests made
on a scrubbing system venting an asphalt saturator.
Theoretical evidence indicates that venturi-type
scrubbers remove contaminants with particle sizes
of less than 1 micron in diameter, but the high ini-
tial equipment cost and high energy requirements
oftheventuri scrubber make its use economically
unattractive compared with other forms of air pol-
lution control equipment.
Table 111. EMISSIONS FROM A WATER
SCRUBBER VENTING AN
ASPHALT SATURATOR
Volume, scfm
Temperature, °F
Emission rate,
gr/scf
Ib/hr
Water, %
Collection efficiency, %
Scrubber
inlet
12,000
138
0. 535
55.0
2. 7
Scrubber
discharge
12, 196
82
0.0737
7. 7
4. 2a
86
At 3. 7 volume % of water, vapor is saturated air.
Other qualitative tests run simultaneously showed
no particulate water.
SOLVENT DEGREASERS
INTRODUCTION
In many industries, metal-tabricated articles must
be washed or degreasedbefore their electroplating,
painting, or other surface finishing. Most de-
greasing operations of any size are carried out in
packaged units, termed degreasers, in which a
chlorinated organic solvent, either in the gaseous
or liquid state, is used to wash the parts free of
grease and oil. Some measurable solvent is emit-
ted as vapor from even the smallest degreaser,
and the sheer number of these units in large manu-
facturing areas makes their combined solvent emis-
sions significant to a community's air pollution.
Design and Operation
Designs of solvent degreasers run the gamut from
simple, unheated wash basins to large, heated,
conveyorized units in "which articles are washed
in hot solvent vapors. The vapor-spray unit de-
picted in Figure 277 is typical of the majority of
industrial degreasers. Solvent is vaporized in the
left portion of the tank either by electricity, steam,
or gas heat. Solvent vapors rise and fill that por-
tion of the tank below the •water-cooled condensers.
At the condensers there is a definite vapor line
that can be observed from the top of the tank. Con-
densed solvent runs through the collection trough
to the clean-solvent receptacle at the right of the
tank. Articles tobe degreasedare lowered in bas-
kets into the vapor space of the tank. Vapors con-
dense on the metal parts, andhot condensate rinses
oil and grease into the liquid receptacle. When
necessary, the flexible hose and spray pump are
used to rinse particularly dirty articles. Many
degreasers are equipped with lip-mounted exhaust
hoods that draw fumes from the top of the tank and
vent them outside the working area.
Types of Solvent
Nonflammable, chlorinated solvents are used al-
most exclusively with degreasers. An estimated
90 percent of the tonnage used for this purpose in
Los Angeles County is trichloroethylene, CHC1 =
CCl^, most of the remaining 10 percent being the
higher boiling perchloroethylene, CC12 = CCl->. Se -
lection of solvent is usually dictated by operators'
temperature requirements. Most greases and tars
dissolve readily at the 189°F boiling point of tri-
chloroethylene, and this is the apparent reason for
its wide use. Perchloroethylene, which boils at
249°F, is consequently used only when higher
temperatures or its slightly different chemical
properties are required. Freon also finds occa-
sional use in specialized degreaser applications.
THE AIR POLLUTION PROBLEM
The only air pollutant emitted from solvent-de-
greasing operations is the vapor of the organic
solvent. Both trichloroethylene and perchloro-
ethylene are considered slightly toxic. The Amer-
ican Conference of Governmental Hygienists rec-
ommends a maximum allowable concentration of
200 pprn for continuous 8-hour exposure to either
solvent. Acute exposure produces dizziness, se-
vere headaches, irritation of the mucous mem-
branes, and intoxication. Chronic exposure can
be fatal through damage to the liver and kidneys
(Sax, 1963).
Solvent Losses
Daily emissions of solvent from individual de-
greasers vary from a few pounds to as high as
-------
384
MECHANICAL EQUIPMENT
WATER JA.CKET—•>(
VAPO.R AREA-
WORK
BOILING LIQUID
DRAIN
FINNED COIL
CONDENSER
CGNDENSATE
COLLECTOR
WATER SEPARATOR
DRAIN
WATER SEPARATOR
STORAGE TANK
OVERFOLW LINE
PUMP SUMP
SPRAY PUMP
Figure 277. Vapor-spray degreaser (Catalog No. 10M359, Baron Industries, Los Angeles, Calif.).
l,300pounds (two 55-gallon drums). Total emis-
sions in large industrial areas are impressive.
For example, in Los Angeles County degreasing
operations are estimated to be responsible for
the emission of 45 tons of chlorinated solvents
per day. This represents some 70 percent of the
halogenated solvent usage and about 7. 5 percent
of the total organic solvent usage in that area
(Lunche et al. , 1957).
Solvent is lost from degreaser tanks in essen-
tially two -ways : Vaporization (including diffusion)
from the tank, and carryout with degreased arti-
cles. A leading degreaser manufacturer esti-
mates that about 0. 05 pound of solvent is lost by
vaporization per hour per square foot of open tank
area where there are no appreciable drafts across
the top of the tank. Obviously, a much higher
quantity of solvent is carried away when cross-
drafts are strong.
The quantity of solvent carried out with the prod-
uct (and later evaporated into the atmosphere) is
a function of product shape and the method in
which articles are distributed in the basket. In
many instances, these losses can be greatly re-
duced by proper alignment in the degreaser's
basket.
The cost of chlorinated solvents, currently about
$2 per gallon, often makes installation of special
equipment desirable to minimize vaporization and
carryout losses and to recover solvent from the
ventilator's exhaust gases.
HOODING AND VENTILATION REQUIREMENTS
Vapor spray degreaser s are not usually ventilated.
The condensing ring and the high density of the va-
por are relied upon to retain the vapor in the tank.
A small amount of vapor does escape, and general
room ventilation is used to remove the vapor from
the site. When a control device is used to collect
the vapors from the tank, a lateral slot hood may
be used, as shown in Figure 278. Slot hoods are
also used sometimes without control devices. In
both cases a minimum volume of air is used to
prevent excessive loss of valuable solvent or to
preclude overloading the control device. Slot hood
velocities should not exceed 1,000 fpm for this
service, and in many cases, by experimentation,
these velocities may be reduced. Size of tank, ob-
jects degreased, and drafts within the building all
influence slot velocities.
AIR POLLUTION CONTROL EQUIPMENT
Emission of solvent from degreasers can be min-
imized by location, operational methods, and tank
covers. In a few cases, surface condensers and
activated-carbon adsorbers have been used to col-
lect solvent vapors.
Methods of Minimizing Solvent Emissions
In a discussion of degreaser operation, The Met-
al Finishing-Guidebook-Directory (1957) recom-
-------
Solvent Degreasers
385
Figure 278. Vapor degreaser and hooding vented
to activated-carbon unit shown in Figure 279
(General Controls, Burbank, Calif.).
mends several techniques for reducing losses of
solvent and, consequently, air pollution:
1. A degreaser should always be located in a posi-
tion where it will not be subject to drafts from
open windows, doors, unit heaters, exhaust
fans, and so forth. If possible, a 12- to 18-
inch-high shield should be placed on the wind-
ward side^f the unit to eliminate drafts.
2. Work items should be placed in the basket in
such a way as to allow efficient drainage and
prevent dragout of solvent.
3. Metal construction should be used for all bas-
kets, hangers, separators, and so forth. Use
of rope and fabric that absorbs solvent should
be avoided.
4. The speed of-work entering and leaving the va-
por zone should be held to 12 fpm or less. The
rapid movement of work in the vapor zone
causes vapor to be lifted out of the machine.
5. Spraying above the vapor level should be avoid-
ed. The spray nozzle should be positioned in
the vapor space where it will not create dis-
turbances in the contents of the vapor.
6. Work should be held in the vapor until it reaches
the vapor temperature where all condensation
ceases. Removal before condensation has
ceased causes the work to come out wet with
liquid solvent.
7. When the metal articles are of such construc-
tion that liquid collects in pockets, the work
should be suspended in the free-board area
above the tank to allow further liquid drainage.
8. The degreaser tank should be kept covered
whenever possible.
Tank Covers
As operators have become more cognizant of the
costs of degreaser solvent and of the hazards to
worker health, the use of intricate and sometimes
costly tank closures has become popular. In ear-
lier times, most degreasers -were equipped with
relatively heavy, metal, one-piece covers. The
weight and unwieldy shape of these covers were
such that few operators could be depended upon to
place them over the tanks at the end of a working
day. Since modern tank closures are operated hy-
draulically or electrically with foot levers, but-
tons, andsoforth, workers can easily cover tanks,
even during short periods of work stoppage. There
are several varieties of automatically operated
closures, one of which is shown in Figure 279.
Most are fabricated of steel, screens, plastic, or
plastic-impregnated fabric. Closure is usually by
roll or guillotine action whereby vapor disturbance
is minimal. The use of solid hinged lids should
be avoided; however, hinged screen lids may be
used.
The solvent saving and air pollution control that can
be effected with automatic closures is a function of
prior operating technique. Where degreaser oper-
ation has been relatively haphazard, the use of
these covers has been shown to reduce emissions
of solvent well over 50 percent. When a degreaser
has been well located and operated, the savings
provided by these devices has been small. Because
of the high cost of chlorinated solvents, however,
automatic closures frequently pay for themselves
in short periods even at moderate usage of solvent.
Controlling Vaporized Solvent
While most solvent conservation efforts have been
directed toward prevention of emissions at the tank,
there are means by which these vapors can be re-
moved from a carrying air stream that would other -
wise be exhausted to the atmosphere. Practical
control methods are extremely limited and, indeed,
industrial application of chlorinated-solvent con-
trols have, to date, been uncommon. Adsorption
with activated carbon is, in fact, the only current-
ly feasible means that can he adapted to most de-
greasers. Activated carbon has a relatively high
-------
386
MECHANICAL EQUIPMENT
Figure 279. A hydraul ically operated screen-type closure: (left) Cover in closed position
(right) cover in open position (Baron Industries, Los Angeles, Calif.).
capacity for both trichloroethylene and perchloro-
ethylene, and adsorption units can be used to re-
cover up to 98 percent of the solvent vapors in ex-
haust gases from a degreaser.
An activated-carbon adsorber used to recover tri-
chloroethylene is shown in Figure 280. It consists
essentially of two parallel-flow carbon chambers
that can be operated either separately or simulta-
neously. Solvent-laden air is collected at spray
degreasing booths, as depicted in Figure 281, and
at the vapor degreaser, previously shown in Fig-
ure 278. The solvent-laden airstream is directed
to both carbon chambers except when one chamber
is being regenerated. A unit such as this must
necessarilybe designed to handle the required ex-
haust volume through only one chamber. The oper -
ator of this particular adsorber reports a 90 per-
cent reduction in usage of chlorinated solvent (1, 100
gallons per month) since its installation. Carbon
adsorption is especially suitable for spray degreas -
ing operations where the spray chamber must be
exhausted to protect the operator.
When solvent concentrations in exhaust gases are
relatively large, surface condensers can be used
to collect appreciable quantities of solvent. The
principal deterrent to the use of this type control
Figure 280. Two-chamber, activated-carbon adsorption
unit used to recover trichloroethylene from degreas-
ing exhaust gases (General Controls, Burbank, Calif.).
GPO 806—6 11
-------
Surf ace-Coating Operations
387
Figure 281. Spray degreasing table and hooding
vented to activated-carbon unit shown in Fig-
ure 280 (General Controls, Burbank, Calif.).
is the small concentration of chlorinated solvent
usually encountered in exhaust gases from degreas -
ers. At the 68 °F operating temperature of most
atmospheric, water-cooled condensers, the tri-
chloroethylene concentration can be held only to
7. 4 percent, and the perchloroethylene concentra-
tion, to 2. 4 percent. Chlorinated-solvent concen-
trations in exhaust gases from degreasersare usu-
ally well below these figures.
Since degreaser solvents are essentially noncom-
bustible, incineration is not a feasible method of
control. Moreover, the thermal decomposition of
chlorinated solvents can produce corrosive and
toxic compounds, such as hydrochloric acid and
phosgene, which are more objectionable air con-
taminants than the solvents.
SURFACE-COATING OPERATIONS
INTRODUCTION
Many devices are used in the painting and coating
of manufactured items. Basic coating operations
include dipping, spraying, flowcoating, and roller
coating. There are variations and combinations of
these operations, each designed for a specific task.
The coatings applied in these operations vary
widely as to composition and physical properties.
Table 112 gives some typical coating formulas.
Table 112. EXAMPLES OF SURFACE-COATING FORMULAS ON
AN AS-PURCHASED BASIS
Type of
surface
coating
Paint
Varnish
Enamel
Lacquer
Metal primer
Glaze
Resina
Sealer
Shellac
Stain
Zinc chromate
Composition of surface coating, %
Non-
volatile
portion
44
50
58
23
34
80
50
50
50
20
60
Hydrocarbons
Aliphatic
56
45
10
7
33
_
-
40
-
-
-
Aromatic
_
5
30
30
33
20
_
-
-
80
40
Alcohols
_
_
2
9
-
_
_
-
50
_
-
Ketones
_
_
_
22
-
_
_
_
_
_
Esters
and
ethers
_
_
_
9
_
_
_
10
_
_
-
lContains 50% solvent of an unspecified type.
-------
MECHANICAL EQUIPMENT
Spray Booths
In spraying operations, a spray gun, usually oper-
ated by compressed air, is used to spray the paint
on the object to be painted. A booth or enclosure,
ventilatedby a fan, provides a means of ventilating
the spray area to protect the health of the spray
gun operator and ensure that an explosive concen-
tration of solvent vapor does not develop. Table
113 shows threshold limit values of typical paint
solvents. These values are average concentrations
to which -workers may be safely exposed for an 8-
hour day without adverse effect on their health. The
sprayboothmay also be equipped to filter incom-
ing air as well as remove particulate mattef from
the exhausted air. A typical paint spray booth is
shown in Figure 282.
Table 113. THRESHOLD LIMIT VALUES OF
TYPICAL PAINT SOLVENTS
Lower
limit
%
Acetone 2. 15
Amyl acetate \ ' . 1
Methyl ethyl ketone
1.81
Butyl acetate j 1.7
Cellosolve | 2.6
Cellosolve acetate
Ethyl acetate
Ethanol
Naphtha (petroleum)
Toluene
Xyiene
1.71
2. 18
3.28
0. 92 to
1.27
!. 0
Mineral spirits | 3. 77
explosive |
(LED *
25% of
j LEL,
j PPm
22,000
ppm
5, 500
1 11, 100 j 2, 770
! 18,400
! 17, 300
j 26,700
1 17,400
1 22,300
j 33, 900
4, 600
Maximum
allowable
u
concentration,
ppm
1,000
200
250
4, 320 j 200
6,670
200
4, 350 j 100
5,570 400
8, 470 1 1.000
1. 1 j 9, 290 2, 320 ! 500
| 12,600
3,150 200
| 10, 100 2, 520
| 7,760
1, 940
200
500
i i I 1
aAdapted from. Factory Mutual Engineering Division, Handbook of
Industrial Loss Prevention, McGraw-Hill Book Co. , Inc. , New York,
1959.
^Adapted from American Medical Association Archives of Environ-
mental Healtn. 14.186-89, 1956.
Flowcoating Machines
Flowcoating consists of flowing the paints in a steady
stream over the work suspended from a. conveyor
line. Excess paint drains from the work to a basin
from where it is recirculated by a pump back to
the paint nozzles. Figures 283 and 234 show typ-
ical flowcoating machines.
Paint Dip Tanks
Paint dip tanks are simple paint containers, fre-
quently -with conical bottoms. The object to be
coated with paint is immersed and then removed.
Provision is made to drain the excess paint from
the work back to the tank, either by suspending
the work over the container or by using drain-
boards that drain back to the dip tank. Some meth-
od is usually provided for agitation of the paint in
the tank, in order to keep a uniform mixture. The
most frequently used method consists of pumping
paint from the tank bottom to a point near the tank
top but still under the liquid surface.
Figure 282. A typical water wash-type spray booth
(Binks Manufacturing Company, Los Angeles, Calif.).
Figure 283. Side view of a flowcoating machine
(Industrial Systems, !nc., Southgate, Calif.).
-------
Surface - Coating Operations
389
Figure 284. View of a flowcoating machine show-
ing drain decks and enclosures (Industrial
Systems, Inc., Southgate, Calif.).
Roller Cooling Machines
Paint roller coating machines are similar to print-
ing presses in construction. Themachines usually
have three or more power-driven rolls. One roll
runs partially immersed in the paint. This roll
transfers the paint to a second roll parallel to it.
The sheet work to be coated is run between the
second and a third roll and is coated by transfer
of paint from the second roll. The quantity of paint
applied is established by the distance between the
rolls through -which the sheet passes.
THE AIR POLLUTION PROBLEM
Air Contaminants From Paint Spray Booths
The discharge from a paint spray booth consists
of particulate matter and organic-solvent vapors.
The particulate matter consists of fine paint par-
ticles, whose concentration seldom exceeds 0.01
grain per scf of unfiltered exhaust. Despite this
small concentration, the location of the exhaust
stack must be carefully selected so as to prevent
paint spotting on neighboring property.
The solvent concentration in the spray booth ef-
fluent varies from 100 to 200 ppm. The solvent
emission out the spray booth stack varies widely
with extent of operation, from less than 1 pound
per day to over 3, 000 pounds per day. Paint sol-
vent vapors evidently take part in the photochemical
smog reactions leading to products that result in
eye irritation. Their odors may also cause local
nuisances. Essentially all the solvent in the coat-
ing mixture is eventually evaporated and emitted
to the atmosphere •
Air Contaminants From Other Devices
Air contaminants from paint dipping, flowcoating,
and roller coating exist only in the form of organic-
solvent vapors since no particulate matter is formed.
HOODING AND VENTILATION REQUIREMENTS
Requirements for Paint Spray Booths
The usual spray booth ventilation rate is 100 to 150
fpm per square foot of booth opening. Insurance
standards require that the enclosure for spraying
operations be designed and maintained so that the
average velocity over the face of the booth, during
spraying operations, is not less than 100 fpm.
Requirements for Other Devices
Dip tanks, flowcoaters, and roller coaters are fre-
quently operated without ventilation hoods. When
local ventilation at the unit is desirable, a canopy
hood may be installed,
AIR POLLUTION CONTROL EQUIPMENT
Control of Point Spray Booth Particulates
A considerable quantity of particulate matter re-
sults from the use of the common air atomization-
type spray gun. During painting of flat panels, a
minimum of 35 percent of the paint sprayed is not
deposited on the panels and is called overspray.
During the spraying of other articles, the over-
spray may be as high as 90 percent; however, 60
percent overspray is more common, Particulate
matter in paint spray booths is controlled by baffle
plates, filterpads, orwater spray curtains. Baf-
fle plates control particulates from enamel spray-
ing by adhesion, with removal efficiencies of 50 to
90 percent. Baffle plates have very low efficien-
cies in collecting lacquer spray particulates be-
cause of the rapid drying of the lacquer and con-
sequent slight'adhesion to the baffles.
Filterpads satisfactorily remove enamel and lac-
quer particulates with efficiencies as high as 90
percent. The filtering velocity should be less than
250 fpm.
Water curtains and sprays are satisfactory for re-
moving paint particulates, with efficiencies up to
-------
390
MECHANICAL EQUIPMENT
95 percent. A water circulation rate of 10 to 38
gallons per 1, 000 cubic feet of exhaust air is cus-
tomary. Surf ace-active agents are added to the
•water to aid in the removal of paint from the cir-
culating tank.
Control of Organic Vapors From Surface Coatings
Known solvent recovery processes make use of
condensation, compression, absorption, distilla-
tion, or adsorption principles. Organic solvents
used in coatings are not controllable by filters,
baffles, or water curtains. In view of the small
solvent vapor concentration in the airstream from
the spray booth or applicator hood, the only eco-
nomicall'y feasible solvent control method is ad-
sorption. Recent work (Elliott et al. , 1961) indi-
cates that adsorption by activated carbon can be
a feasible method for the control of paint solvents.
This-work indicates that control efficiencies of 90
percenter greater are possible, provided partic-
ulates are removed from the contaminated air-
stream by filtration before the airstream enters
the carbon bed. General design features of ad-
sorption-type devices have been discussed in Chap-
ter 5.
PIPE-COATING EQUIPMENT
mon qualities that make these materials excellent
for pipe coatings are as follows:
1. They resist moisture, and chemical and elec-
trolytic action.
2. Long-lasting adhesion canbe expected between
the coating and pipe,
3. They are stable over a wide temperature range
if properly compounded.
4. They are tough and resist mechanical abrasion.
5. They possess good ductility and can resist soil
contraction and expansion and underground
pipe movement
6. They resist aging over long periods of time.
METHODS OF APPLICATION
The three usual methods of applying asphalt or coal
tar coatings to pipe are dipping, wrapping, and
spinning (The Asphalt Institute, 1954; American
Water Works Association, 1951). These will be
discussed individually. With all application tech-
niques the pipe must be dry and rust free. Most
often a primer is applied before the final coating
is added.
INTRODUCTION
Iron and steel pipes are subject to corrosion and
oxidation, particularly in underground service. In
order to exclude the corrosive elements from con-
tact with the metal, many surface coatings have
been used. These include paints , lacquers, metal-
lic coatings, vitreous enamels, greases, cements,
and bituminous materials, both asphalt and coal
tar based. Only the bituminous materials will be
discussed in this section.
Asphalt, a residue derived from the distillation of
crude petroleum, becomes a dark brown to black
rubbery solid when air blown at elevated tempera-
tures and allowed to cool. Coal tar is a dark brown
to black, amorphous, solid residue resulting from
the destructive distillation of coal. Both materials
are compounded with mineral fillers and other in-
gredients to form the so-called enamel that is ap-
plied to the pipe. Both materials perform essen-
tially the same duty with some qualifications. With-
out the addition of plasticizers, the coal tar enam-
els tend to have a fairly narrow satisfactorily oper-
ating temperature range. Above or below this range,
they are too soft to stay in place or too brittle to
resist impact. The asphalts have a wider oper-
ating temperature range but have a disadvantage
of being slightly more permeable to moisture and
are affected more by soil minerals. Some corn-
Pipe Dipping
Pipe dipping involves applying the coating to both
the internal and external surfaces of the pipe by
completely immersing it in a large vat of molten
asphalt. Coal tar enamel cannot be applied by dip-
ping since it cannot be held in an open container
for long periods of time •without excessive changes
in its physical properties. The tank used is usu-
ally rectangular with dimensions to accommodate
the largest size pipe to be dipped. The asphalt is
kept at a specified temperature by heat-transfer
tubes submerged in the enamel. The pipe is low-
ered into the enamel until completely covered and
allowed to remain until the metal reaches the tem-
perature of the liquid. This is necessary for good
adhesion. It is then raised, tilted off horizontal
in order to drain off excess enamel, and allowed
to cool. Additional thickness may be obtained by
redipping. For the second and succeeding dips
however, the pipe must not remain in the tank long
enough to remelt the material already deposited.
Pipe Spinning
Pipe spinning is the name given to the procedure
•wherein molten asphalt or coal tar enamel is ap-
plied to the interior surface of a rotating pipe. The
spinning motion is given to the pipe by conveyor
-------
Pipe-Coating Equipment
391
wheels or endless chain slings. The enamel is ap-
plied by spray heads on a lance attached to a travel-
ing, heated, enamel kettle. The lance is inserted
the full length of the pipe and then the hot enamel
is sprayed as the lance is withdrawn. The spin-
ning of the pipe deposits the enamel in a uniform
layer and holds it in place until it hardens. The
spinning is continued with usually a cooling water
spray on the outside of the pipe until the enamel
temperature has cooled to about 100°F.
Pipe Wrapping
Pipe wrapping is the most complex of the common
pipe protection techniques involving asphalt and
coal tar because, in addition to the enamel, wrap-
pings of rag or asbestos felt, plastic film, fiber-
glas, metallic foil, kraft paper, or a combina-
tion of these are used. Two types of equipment are
used. One type consists of apparatus both to rotate
the pipe and move it longitudinally past a station-
ary enamel dispensing and -wrapping station, as
shown in Figure 285. In the other method, only
the pipe rotates, and the coal tar or asphalt kettle
and wrapping equipment travel on a track along the
length of the pipe (Figure 286).
The purpose of the wrapping is to make the pipe
covering more durable during handling and install-
ing as well as increase its aging and moisture ex-
clusion properties. The enamel has a dual pur-
pose—in addition to its corrosion-resisting func-
tion, it serves as an adhesive for the wrapping.
Preparation of enamel
Both coal tar and asphalt are shipped to the con-
sumer in solid, 100-pound, cylindrical or octa-
gonal castings or in 55-gallon fiber drums weigh-
ing about 650 pounds. Before being charged to the
melting equipment the material is manually chopped
into chunks weighing 20 pounds or less. The ma-
terial is melted and kept at application tempera-
ture in natural gas-, oil-, or LPG-fired kettles.
Lastly, the condensed vapors- and gaffes are toxic.
Prolonged breathing or' skin exposure can cause
itching, acne, eczema, psoriasis, loss of appe-
tite, nausea, diarrhea, headache, and other ail-
ments. Some medical researchers have stated
that the fumes may also have some cancer-pro-
ducing potential, but this has not been completely
substantiated.
Although the fumes are dense, the actual weight
of material emitted is relatively small. Tests con-
ducted on pipe-wrapping operations using both as-
phalt and coal tar enamels have shown emissions
ranging from a low of 1.8 pounds per hour to a
maximum of 17. 5 pounds per hour.
HOODING AND VENTILATION REQUIREMENTS
Because of the nature of all three of the methods
used to apply asphalt and coal tar enamels to pipe,
collection of the contaminants is difficult. Large
quantities of air are entrained because hoods usu-
ally cannot be placed close to the point of emis-
sion. In the pipe-dipping operation, after being
immersed, the pipe must be raised vertically above
the tank and allowed to drain. Although lip-type
hoods around the tank periphery may collect most
of the tank emissions, those from the pipe itself
cannot be collected by these hoods. In "wrapping,
especially for the traveling application type of
equipment, a hood as long as the pipe itself would
be necessary. A relatively small hood over the
wrapping and tar-dispensing equipment can be used
In the stationary kettle type of wrapper. In the
spinning operation, emissions come from both ends
of the pipe. Because of the need for working with
various pipe lengths, hoods at both pipe ends are not
practical. One solution is to install a stationary
hood at the end of the pipe where the lance is in-
serted. A portable fan or blower is used at the
other end to blow air through the pipe, conveying
the emissions to the hood at the other end.
THE AIR POLLUTION PROBLEM
By far the largest source of air pollution from as-
phalt or coal tar operations is the dense white emis -
sions caused by vaporization and subsequant con-
densation of volatile components in the enamel.
This cloud is composed of minute oil droplets and is
especially dense -whenever the surface of the molten
enamel is agitated. These emissions are objection-
able on three counts that include opacity, odor, and
toxicity- -those from coal tar being the more objec-
tionable. The visible emissions are intense enough
to violate most opacity regulations. The odor of the
emission is pungent and irritating with consider-
able nuisance-creating potential, and there maybe
the added nuisance caused by settling oil droplets.
Another solution of the fume collection problem is
to house ail the equipment and vent the building to
the air pollution control system selected. The
building itself then becomes the collection hood.
This, of course, dictates a large exhaust air vol-
ume to provide enough draft to prevent fume ac-
cumulation and maintain adequate room ventila-
tion for the -workers' comfort and safety. This
method may not be necessary for an isolated spin-
ner or wrapper, but for a dipping process or a pro-
cess using several coating operations, it is more
satisfactory than using local exhaust systems. As
adjuncts to the overall building exhaust system,
some local hoods at points of heavy emissions may
be desirable, especially if these points are in areas
frequented by operating personnel.
-------
392
MECHANICAL EQUIPMENT
.
Figure 285. Stationary kettle type of pipe-wrapping equipment and scrubber:
(top) Closeup, (bottom) overall view (Pacific Pipeline Construction Co.,
Montebello, Calif.).
Figure 286. Traveling kettle-type pipe-wrapping equipment (Southern Pipe and Casing Co.,
Azusa, Caiif,).
-------
Dry Cleaning Equipment
393
AIR POLLUTION CONTROL EQUIPMENT
Three basic types of devices can be considered for
control of the emissions from, asphalt and coal tar
application. These are (1) scrubbers, (2) incin-
erators (afterburners), and (3) electrical precip-
itators.
Abater scrubbers have been used most frequently
for controlling pipe-coating equipment opacity, drop-
lets, and odors. The baffled, water spray type of
scrubber has been employed almost exclusively.
These scrubbers have been operating satisfactorily
by employing 30 gprn water per 1, 000 cfm air to
be scrubbed, at a water pressure of 50 psig. A
typical scruober system of this type is shown in
Figure 287.
Figure 287. Scrubber system to control emissions
from a pipe-wrapping .and pipe-spinning operation
(Southern Pipe and Casing Co., Azusa, Calif.).
The efficiency of scrubbers can be affected not
only by their basic design, but by operational vari-
ables. Of most importance, the scrubber water
must be kept clean. If scum and oil are allowed
to collect for any extended period, and the dirty
water is recirculated, the spray heads begin to
plug, and this lowers the •water rate and reduces
the efficiency. An automatic skimming device is
helpful, but, even so, frequent water changes are
needed. In some instances, daily water changes
and thorough weekly cleaning, including spray heads,
have been necessary.
Properly designed and operated water scrubbers
serving pipe-dipping, pipe-wrapping, and pipe-
spinning operations have been shown by tests tp
have collection efficiencies of about 80 percent on
a "weight ba sis and to reduce visible emissions from
70 percent cpacitv to i 0 to 15 percent opacity.
Incineration is the most positive method of com-
plete control, but economic factors practically
eliminate its application. This is due to the large
quantity of air with a relatively small concentra-
tion of contaminants that must be heated to incin-
eration temperatures of 1,200° tol,400°F. For
example, atypical building housing pipe-wrapping
and pipe-spinning operations might require an ex-
haust volume as great as 40, 000 cfm for adequate
contaminant removal. Heating of this air from
80° tol,200°F would require about 50 million Btu
perhour. Thus, the operating cost as •well as the
initial cost of the relatively large unit required
makes an afterburner unfeasible.
Slectrical precipitator s can be used for controlling
emissions from pipe-coating operations, but, again,
their high initial cost, as compared "with that of
scrubber systems, has made them unattractive.
When, however, some of the maintenance and clean-
ing problems connected with scrubbers, as well as
the higher basic scrubber-ope rat ing costs are con-
sidered, the higher installation cost for precipita-
tors may be counterbalanced. Precipitators have
been used successfully for controlling the emis-
sions from roofing and building paper saturators.
In this operation the emissions are of the same
type as those from pipe coating, but are generally
much greater in concentration and quantity. In
practically all cases a precleaner, such as a wet
dynamic precipitator,is used to remove large par-
ticles and prevent excessive tar buildup on the pre-
cipitator parts . For pipe-coating operations, the
lower overall emissions may obviate the need for
the precieaner.
Although scrubbers have proved to be satisfactory
control devices for pipe-coating equipment, their
effectiveness cannot be described as excellent.
More research is needed on this air pollution con-
trol problem to achieve higher collection efficien-
cies and complete elimination of odors and visible
emissions.
DRY CLEANING EQUIPMENT
Dry cleaning is a process of cleaning soiled tex-
tiles, usually clothes, with organic solvents. The
textiles are cleaned by agitation in a solvent bath
and by rinsing with clean solvent. Excess solvent
is thrown off by centrifugal action in a rotating ex-
tractor, and the textiles are then tumbled to a dry
state in warm air. The solvent is reclaimed for re-
use by filtration and distillation. The filter cake
may be cooked as a further solvent recovery mea-
sure. Figures 288, 289, and 290 illustrate the
various cleaning equipment.
Dry cleaning equipment follows two basic designs.
One design is tailored for petroleum solvents, and
the other, for chlorinated hydrocarbon, or synthetic,
-------
394
MECHANICAL EQUIPMENT
figure 288. Synthetic-solvent dry cleaning unit with an
activated-carbon adsorber (Joseph's Cleaners and Dyers,
Los Angeles, Cali f.).
Figure 289. Petroleum solvent dry
Cleaners, Inglewood, Calif.).
cleaning unit (Century Park
-------
Dry Cleaning Equipment
395
Figure 290. Synthetic-solvent, coin-operated
dry cleaning unit (Norge Sales Corp. Los
Angeles, Calif.).
solvents. The increasingly popular coin-operated
machine, a totally enclosed and smaller automatic
version of a synthetic-solvent unit, incorporates
the features of the larger synthetic-solvent units.
In a petroleum solvent dry cleaning plant, the equip-
ment generally includes a washer, centrifuge, tum-
bler, filter, and, in many instances, a batch still.
The washer consists of a perforated, horizontal,
rotating drum enclosed in a vaportight housing.
Housing and drum are each equipped with a clo-
sure for loading and unloading. A tank in the bot-
tom of the housing serves as a reservoir for sol-
vent. The centrifuge,called an extractor, is used
to spin off solvent adhering to the clothes. Solvent
drains through perforations in the centrifuge bas-
ket and is piped to a tumbler, similar to the washer
but equipped with a blower and heater, and used to
circulate hot air through the clothes and exhaust
it to the atmosphere. Auxiliary equipment consists
of a filter to remove suspended material from the
solvent and a batch still to purify the solvent for
reuse.
With perchloroethylene, the washer and extractor
are combined in a single unit. The tumbler is
equipped with a condenser for recovery of solvent
vapor. The tumbler is a closed system while in
operation and is vented to the atmosphere only dur-
ing a short deodorizing period. A muck cooker is
often used to reclaim solvent from filter sludge.
THE AIR POLLUTION PROBLEM
Solvents
As previously mentioned, two types of solvents
are commonly used by the dry cleaning industry.
These are the petroleum solvents, of •which Stod-
dard solvent and 140°F solvent are most represen-
tative, and chlorinated solvents, hydrocarbon or
synthetic, of which perchloroethylene, also known
as tetrachloroethylene, is most representative.
Small quantities of proprietary compounds may be
added to the solvent by the dry cleaning operator
to aid in the cleaning action of the solvent and to
yield other beneficial effects. Table 114 lists some
properties of these solvents.
Table 114. PROPERTIES OF DRY CLEANING
SOLVENTS (Mellan, 1944, 1957)
Property
Distillation range, °F
API gravity
Specific gravity at 60°F
Lb/gal
Paraffins, %
Naphthenes, %
Aromatic s, %
Flash point (TCC), • F
Corrosiveness
Caution
Odor
Color
Cost, $/gal
140"F
358 to 396
47. 9
0. 789
6.57
45. 7
$1 2
12. 1
140
None
FUmmablc
Mild
Water white
0. i\
Stoddard
305 to 350
50. 1
0.779
6. 49
46. 5
41.9
1 1.6
100
None
Flammable
Sweet
Water white
0.20
Perchloroethylene
250 to Z54
-
1.61
13. 4
-
-
Extinguishes fire
Slight on metal
Toxic
Similar to ether's
Colorless
i. 00
The dry cleaning industry contributes to air pol-
lution by the release of organic-solvent vapors to
the atmosphere. A good dry cleaning solvent is
necessarily volatile, and this volatility can result
in emissions of solvent when storage tanks are
loaded, equipment doors are opened, ductwork or
equipment leaks, and textiles dripping solvent are
removed from equipment. The amount of solvent
emitted to the atmosphere from any one dry clean-
ing plant is dependent upon the equipment used, the
length of certain operations in the cleaning process,
the precautions used by the operating personnel,
and the quantity of clothes cleaned. The most im-
portant of these items are the precautions used and
the weight of clothes cleaned.
A typical synthetic-solvent plant processes 2, 000
pounds of textiles per 5-day •week and can clean
from 3, 000 to 14, 000 pounds, with an average of
5, 500 pounds of textiles per 55-gallon drum of
solvent (a consumption of 10 gallons of solvent for
each 1, 000 pounds of textiles). This average in-
cludes solvent recovered from filter sludge or
muck.
-------
396
MECHANICAL EQUIPMENT
The low cost of petroleum solvent provides little
economic incentive to the operator of a petroleum
solvent dry cleaning plant to conserve solvent and
prevent or control its emission to the atmosphere.
Emission of s olvent vapors occurs r>rimari..y from
the vent on the tumbler, and to a lesser extent,
during the transfer of wet textiles from one piece
of equipment to another and during disposal, of fil-
ter sludge. These emissions are increased by poor
operational practices. An average petroleum sol-
vent plant may process about 6, 000 pounds of tex-
tiles per 5-day week and usually cleans about b5
pounds of textiles per gallon of solvent (a consump-
tion of 1 5 gallons of solvent for each 1, 000 pounds
of textiles).
Obviously then, the use of petroleum solvents re-
sults in the emission to the atmosphere of, on an
average, 50 percent more solvent (by volume/ tnan
is emitted with the use of chlorinated solvent. Be-
cause a gallon of chlorinated solvent is much heav -
ierthana gallon of petroleum solvent, its use re-
sults in a 40 percent greater emission (by weight,!
of perchloroethylene.
Lini
The lint generated when fabrics are tumbled dry
must be removed before the air is discharged to
the atmosphere. This is a minor problem, easily
solved by devices not normally considered to be
air pollution control equipment. The synthetic-
solvent tumbler s are provided with a cloth bag to
filter the lint from the exhaust air. It is usually
cleaned out at the completion of the daily opera-
tion. The petroleum solvent tumblers are gener-
ally exhausted to a separate lint trap that is filled
•with -water and operates on the wet-impingement
principle. These lint traps must be cleanea regu-
larly to prevent the discharge of lint, wrich can
sometimes cause nuisance complaints, or the ac-
cumulation of lint, -which restricts airflow and in-
creases the hazards of fire and explosion.
HOODING AND VENTILATION REQUIREMENTS
Because of safety requirements, hooding and duct-
ing are an integral part of all dry cleaning equip-
ment. In synthetic-solvent plants, vents are pro-
vided near the doors of the washer-extractor and
the tumbler. An exhaust system is automatically
activated whenever these doors are opened, and
the system exhausts the vapors resulting from
transfer of the wet textiles. When a carbon ad-
eorptionunit is used to collect the perchloroethyl-
ene vapors, floor vents are also provided to cap-
ture vapors from other areas.
Ventilation requirements must meet the regula-
tions as delineated by the rules governing fire haz-
ards and toxicity. For Stoddard and 140°F solvent,
the concentration of vapor in and about the equip-
ment must not exceed 500 ppm by volume in the air,
to meet health requirements . This is considerably
les s tnan the Quantity permis sible to prevent a fire
hazard.
Perchloroethylene is not flammable but is toxic
and t'ne allowable concentration must not exceed
200 ppm by volume in air, -which represents the max-
*imum amount to which a person may be exposed
for 8 hours a day over a long period without en-
dangerment to health.
AIR POLLUTION CONTROL EQUIPMENT
The application of activated-carbon adsorption to
control solvent vapor emissions from dry clean-
ing equipment in which perchloroethylene is used
is dictated by economics. With other factors equa^.
the tenfold difference in cost per gallon between
chlorinated and petroleum solvents forces the user
of chlorinated solvents to obtain the maximum num-
ber of pounds of textiles cleaned per gallon of sol-
vent used in order to compete in terms of price
This means reducing solvent consumption to a min-
imum by efficient operation.
Packaged adsorption units using activated carbon
are available to the operator of a synthetic-sol-
vent cleaning plant for recovering perchloroethy]-
ene vapor that would normally be discharged to the
atmosphere. The adsorption unit is added to the
discharge of the ventilating system. Vapor laden
air collected from the -washer-extractor, tumbler,
and floor vents passes through a filter for removal
of entrained solids and then to the adsorber. Col-
lection of the solvent is 100 percent up to the break-
point of the carbon at the particular vapor concen-
tration and temperature. Good operation dictates
that the adsorption cycle stop short of this point.
Recovery of the solvent is effected by passing low-
pressure steam through the carbon. The steam-
vapor mixture is cooled and condensed, and the
solvent is separated from the water by decantation
and returned to the solvent storage tank for reuse.
The value of the solvent recovered makes possible
the amortization of the adsorption unit within 1_ to
2_ years.
Activated-carbon adsorption can also be adapted
to control the solvent emis sions from the petroleum
solvent dry cleaning plant, "but the lower value of
the recovered solvent requires a much longer peri-
od of time to pay the cost of the equipment.
Other methods of reducing emissions from dry
cleaning plants include good operational procedures
and equipment maintenance. In the petroleum sol-
vent plant, because of the low cost of the solvent,
minor leaks are likely to go unnoticed or unattended
in favor of uninterrupted production.
-------
Abrasive Blast Cleaning
397
Operational procedures that affect solvent emis-
sions include the transfer of solvent-wet textiles
from washer to extractor to tumbler. Where the
washer and extractor are combined, a considerable
saving in solvent is obtained, but the capacity of
the washer is reduced. This operation is usually
separated in petroleum solvent cleaning but com-
bined in the synthetic-solvent plant. Since the sol-
vent remaining in the fabric after extraction can
be discharged to the atmosphere by evaporation
during the tumbling and deodorizing operation, the
time period allowed controls the quantity of solvent
discharged.
In synthetic-solvent cleaning plants, tumblers are
equipped with a heater and fan to circulate warm
air through the clothes and are operated as a closed
system during the drying cycle. A water-cooled
condenser is provided to condense the solvent vapor,
but it cannot reduce the concentration below the
dewpoint. Upon completion of the drying cycle, the
discharge vent and inlet are opened and fresh air
is used to deodorize the clothes. Any remaining
solvent is discharged to the atmosphere. The length
of the extraction and tumbling cycles should be suf-
ficient to achieve maximum recovery of solvent.
When a separate tumbler is used, the extraction
cycle should dry the clothes sufficiently to mini-
mize the amount of solvent emitted by evaporation
during transfer.
Activated carbon can adsorb 100 percent of the sol-
vent up to the breakpoint; thus, overall efficiency
is dependent upon the effectiveness of the collec-
tion system, and operation and maintenance pro-
grams. Allowing ineffective adsorption techniques,
that is, placing the adsorbing unit on line without
proper drying and cooling; allowing floor vents to
become covered with lint; and allowing ductwork
to deteriorate are some examples of poor opera-
tion and maintenance.
ABRASIVE BLAST CLEANING
INTRODUCTION
Abrasive blast cleaning is the operation of clean-
ing or preparing a surface by forcibly propelling a
stream of abrasive material against the surface.
Blast cleaning operations may be classified accord-
ing to: (1) The abrasive material used, (2) the
method of propelling the abrasive, and (3) the equip-
ment used to control the abrasive stream or move
the articles being cleaned into the abrasive stream.
Abrasive Materials
Silica sand has been used longer than any other
material, principally because of its ready avail-
ability and low cost. It has a rather high breakdown
rate, but is still widely used where reclaiming the
abrasive is not feasible. Synthetic abrasives, such
as silicon carbide and aluminum oxide, are some-
times used as a substitute for sand in special appli-
cations. Extremely fine sand and talc are used in a
water suspensionf or fine finishing. Soft abrasives
such as ground corn cobs, cereal grains, and
cracked nut shells are used to clean without re-
moving any metal. Metallic abrasives are made
from cast iron and steel (Stine, 1955).
Cast irfm shot is made by spraying molten cast
iron into a water bath. The shot is hard and brittle,
but its breakdown rate is only 2, 5 percent that of
sand. Cast iron grit is made by crushing the over-
size and irregular particles formed when cast iron
shot is being made. The sharp edges of the grit
give it a very rapid cutting action. The breakdown
of the hard, brittle particles continually exposes
new cutting edges. Annealed shot is made from
special-alloy cast iron and is heat treated to re-
duce its brittleness. Its breakdown rate is only
one-third to one-half that of cast iron shot or grit.
Steel shot is produced by blowing molten steel. It
is not as hard as cast iron shot but is much tougher.
Its breakdown rate is only about one-fifth that of
cast iron shot.
Method of Propelling the Abrasive
Three means of propelling the abrasive are com-
pressed air, high-pressure water, and centrifugal
force.
Two types of compressed-air blasting machines
used are suction blast and direct-pressure blast.
The suction method uses two rubber hoses connect-
ed to a blasting gun. One of the hoses is connected
to the compressed-air supply, and the other is
connected to the bottom of the abrasive supply tank,
whose top is open. The gun, as shown in Figure
291 (topleft), consists of a casting with an air noz-
zle that discharges into a larger nozzle. The abra-
sive hose is attached to the chamber between the
nozzles. The high-velocity air jet, expanding into
the larger nozzle, creates a partial vacuum (12 to
17 inches mercury) in the chamber, and the abra-
sive is drawn in and expelled through the discharge
nozzle. In the direct-pressure types, as shown
in Figure 291 (bottom left), the abrasive supply
tank is a pressure vessel with the compressed-
air line connected to both the top and bottom dis-
charge line. This permits abrasive to flow by
gravity into the discharge hose -without loss of pres-
sure. Direct-pressure machines propel from 2 to
4times as much abrasive per cubic foot of air (at
equal pressures) as suction-type machines do, but
the cost of the suction machines is less. Com-
pressed air is also used in wet sandblasting. In
a specially designed direct-pressure machine, as
shown in Figure 291 (bottom right), the abrasive
supply tank is flooded with water, and a mixture
of sand and water is propelled by the compressed
-------
398
MECHANICAL EQUIPMENT
RUBBER TIP
ABRASIVE HOSE
CONNECTION
No. 6 HOSE CLAMP
AIR -
ABRASIVE HOSE
AIR SUPPLY VALVE
AIR
CHOKE
RELIEF
VALVE
EQUAL AIR PRESSURE
ABOVE AND BELOW
ABRASIVE
WATER-!
AIR SUPPLY VALVE
/
H;-AIR
CHOKE
RELIEF
VALVE
EQUAL AIR PRESSURE
ABOVE AND BELOW
ABRASIVE
/; -NSSS^ VK •
^ jj "tj:-~J ML
Figure 291. Types of compressed-air blasting machines: (top left) Suction gun, (top right) suction-
type blasting machine, (bottom left) direct-pressure blasting machine, (bottom right) wet blasting
machine (Bulletin No. 100B, Pangborn Corporation, Hagerstown, Md.).
air. Wet sandblasting can also be accomplished
by attaching a special nozzle head-with a water hose
to the nozzle of a direct-pressure machine as shown
in Figure 292.
Inhydraulic blasting, the propulsive force is high-
pressure water. A mixture of water and sand is
propelled through a nozzle with great force by a
pump that develops a pressure of 1, 000 to 2, 000
psi. Sand reclamation is usually practiced in these
systems. Figure 293 is a diagram of a complete
hydraulic blasting system. Equipment such as this
is used for core knockout and for cleaning very
large castings, heat exchanger tube bundles, and
other large pieces of equipment.
Centrifugal force is the third method of propelling
abrasive. Abrasive is fed to the center of a rotating
impeller, slides along spoke-like vanes, and is
discharged •with great force in a controlled pattern.
Figure 294 shows one type of abrasive impeller.
Metallic abrasives are used with this type of equip-
ment.
Equipment Used to Confine the Blast
The oldest and most widely used device to confine
and control the blast is the blasting room, which
consists of an enclosure •with the operator inside
manipulating the blast from a hose. Blasting rooms
vary widely in their construction. One popular
design is the all-steel, prefabricated room with
floor grating and a completely automatic abrasive
recovery system. These rooms usually use metal
grit or shot and often have monorail conveyors,
rail cars, or rotating tables to aid the operator in
handling the objects, which are usually large and
heavy. Less desirable designs, but sometimes
adequate, are makeshift rooms of wooden con-
struction used for infrequent sandblasting opera-
tions.
For cleaning small parts, the blasting cabinet is
frequently used. A blasting cabinet consists of a
relatively small enclosure with openings to -which
are attached long-sleeved rubber or canvas gloves
by which the operator, from outside the cabinet,
-------
Abrasive Blast Cleaning
399
Figure 292. Adapter nozzle that converts dry
sandblasting to wet sandblasting (Sanstorm
Manufacturing Co., Fresno, Calif.).
manipulates the blasting gun and objects to be abra-
sive blasted, as depicted in Figure 295. All types
of abrasives are used in cabinets--sand, metal-
lies, soft abrasives, and slurries.
Centrifugal impellers are usually incorporated in-
to a machine that handles the objects so as to ex-
pose all surfaces to the blast. The two most com-
mon types are those using tumbling action and those
containing rotating tables, as shown in Figures 296
and 297. Special machines are made for specific
jobs, such as cleaning sheet metal strip. The
housing of these machines confines the blast and
its effects. Automatic abrasive recovery and re-
cycle equipment are used.
Another machine consists of a perforated drum or
barrel rotating inside a cabinet. A blast gun is
mounted so as to project through one end of the
drum. Tumbling action exposes all parts of the
objects to the blast. Both sand and metallic abra-
sives are used. Abrasive-recycling equipment is
usually provided.
THE AIR POLLUTION PROBLEM
The amount of dust created by abrasive blast clean-
ing varies widely with the abrasive used, parts
being cleaned, and propelling medium. Dry sand-
blasting produces large dust concentrations as a re-
sult of breakdown of the sand. Metallic abrasives,
of course, produce less dust but can produce heavy
concentrations in cleaning such things as castings
with considerable amounts of adhering sand. The
CITY H,0 SUPPLY
NEW SAND ADDED THROUGH FLOOR OF ROOM
NOTE. SLOTS FOR GUN
III DOORS ALONG ENDS
OF ROOM AND ACROSS
FRONT OF ROOM
OPERATOR IS IN NORMAL
POSITION
SLOW-MOVING DEWATERING
DRAG CONVEVDR
TO SEWER
FLOW HIGH-PRESSURE HATER FROM PUMP TO GUN
SLURRY OF SAND AND WATER FROM BLAST SAND TANK TO GUN
SAND, WATER FLOWING THROUGH FLOOR ACROSS SCREEN
SAND, WATER, FOULING IN SLUDGE TANK PUMPED
TO CLASSIFIER SECTION
ACCEPTED SANO DROPS INTO BLAST SAND TANK FOR REUSE
FINES AND FOULING REJECTED BY CLASSIFIER
FLUMED TO SLUDGE COLLECTOR
SLUDGE COLLECTOR THICKENS AND REMOVES REJECTED
FINES AND FOULING AND THUS PROTECTS SEWERS
USED H,0 FLUMED TO SEWER
Figure 293. Hydraulic blasting system (Pangborn Corporation, Hagerstown, Md.).
-------
400
MECHANICAL EQUIPMENT
Figure 294. Centrifugal impeller for metallic
abrasives (Wheelabrator Corporation, Mishawaka
Ind.).
dust concentration is small during wet blasting or
when metallic abrasives are used for tasks such as
removing welding and for heat treating scale.
Figure 295. Blasting cabinet (Pangborn
Corporation, Hagerstown, Md.).
HOODING AND VENTILATION REQUIREMENTS
The structures previously described to control the
blast act as hoods, and exhaust ducts are attached
to them for ventilation.
Blast cleaning rooms are ventilated by baffled in-
let openings, usually in the roof, and exhausted
from near the floor. Recommended ventilation
rates vary from 60 to 100 fpm across the floor
area with 80 fpm the usual choice (Industrial Ven-
tilation, I960). These rates are based mainly on
the maintenance of visibility in the room. The
usual requirement for dust control is an indraft ve-
locity of at least 500 fpm through all openings (ibid).
By making the openings small, a small exhaust
volume suffices to meet the requirement, but visi-
bility is so poor during sandblasting as to impair
the operator's effectiveness seriously. Health
codes require that the operator wear an air-sup-
plied, Bureau of Mines approved abrasive blast-
ing helmet.
The ventilation requirement for blast cabinets is
similar to that for blasting rooms. Twenty air
changes per minute are usually recommended,
based primarily on the maintenance of visibility.
Even during wet sandblasting, this exhaust rate is
usually required for maintenance of visibility.
For blasting barrels, rotary tables, and tumbling -
type machines, the general rule of 500 fpm indraft
velocity at all openings is applicable. The total
area of openings of some machines is difficult to
measure; however, the manufacturer usually speci-
fies the required ventilation rate. This rate in-
cludes sufficient airflow to remove excess fines so
as to maintain the abrasive in an optimum condi-
tion.
AIR POLLUTION CONTROL EQUIPMENT
For dust of such widely varying concentration and
particle size as is produced in the various blast-
-------
-Galvanizing Equipment
401
Figure 296. Blast cleaning macnine that uses
tumbling action (Wheelabrator Corporation-,
Hishawaka, Ind.).
Figure 297. Blast cleaning machine containing
multiple rotary tables (Pangborn Corporation,
Hagerstown, Md.).
bags are adequate for this service. Since virtual-
ly all blast cleaning operations are intermittent, a
noncompartmented baghouse can be considered.
A scrubber of good design collects the bulk of the
dust, and wet collectors are used to some extent.
A scrubber of high power input is, however, re-
quired for collecting the very line dust.
Wet sandblasting does not require collection equip-
ment and provides a means of blast cleaning build-
ings, bridges, and other structures without cre-
ating a dust nuisance. Collecting the dust from dry
sandblasting of structures such as these would be
very difficult or impossible.
ZINC-GAlVANiZING EQUIPMENT
INTRODUCTION
Zinc galvanizing may be defined as the art of coat-
ing clean, oxide-free iron or steel with a thin layer
of zinc by immersion in molten zinc held at tem-
peratures of 840° to 860°F (Elliott et al. , 1961).
In order to achieve optimum results, the funda-
mental processing steps to be followed are:
1. Degreasing in a hot, alkaline solution;
Z. rinsing thoroughly in a water rinse;
3. pickling in an acid bath;
4. rinsing thoroughly in a water rinse;
5. prefluxing in zinc ammonium chloride solu-
tion;
6. immersing the article in the molten zinc
through a molten flux cover;
7. finishing (dusting with sal ammoniac to pro-
duce smooth finishes).
When considering the air pollution aspects of the
galvanizing operation, one might be inclined to
ornitthe first five steps because they do not normal-
ly produce excessive air contaminants. Improper
degreasing does, however, increase the genera-
tion of air contaminants when the article is im-
mersed in the hot zinc. Moreover, stripping pre-
vious zinc coatings in the pickling tanks causes
excessive acid mists to be generated.
ing operations, the baghouse is the most widely
used type of collector. The positive collection
mechanism of the baghouse ensures virtually 100
percent collection efficiency for an adequately sized
unit in good condition. The filtering velocity should
not exceed 3 fpm. Standard cotton sateen cloth
Cleaning
If an article is not thoroughly degreased, an oil
mist is discharged when the article is dipped into
the molten zinc. If the articles are not properly
pickled and rinsed, more flux must be used to
-------
402
MECHANICAL EQUIPMENT
achieve the desired coating, -which in turn creates
more fumes. It is important, therefore, to de-
grease, pickle, and rinse thoroughly the articles
being galvanized, not only to obtain a good zinc
coating, but alsoto reduce the generation of fumes
and facilitate the collection of unavoidable fumes.
ammonium chloride is used as a starting material,
the zinc chloride merely melts, trapping the bulk
of the gases formed, and retards the deposition
of the ammonium chloride. The flux cover is made
much more easily and -with less fuming when the
foaming-type zinc ammonium chloride is used in
place of ammonium chloride.
Cover Fluxes
On the assumption that the article was properly
prepared for dipping in the molten zinc, a flux
must still be used to remove the oxide film that
forms as the article is being transported from the
last rinse tank to the galvanizing kettle. To exclude
air from the part after fluxing, the flux is floated
on the zinc surface so that the article is fluxed as
it enters the zinc. Figure 298 shows the flux cover
on one end of a galvanizing kettle.
The flux cover has a number of important functions
in addition to the cleaning action already mentioned.
It serves as a preheating and drying medium to re-
duce spattering or explosions in the molten zinc,
and distortion of thin metal sections. It keeps the
zinc surface free of oxides, which, if occluded in
the coating, tend to dull it and retard drainage of
zinc from the work. Heat losses from the kettle
are also reduced.
The flux is thought to create most of the air con-
taminants from a galvanizing operation; therefore,
a description of fluxes , their composition and action
is of value. The theory is that, regardless of
whether ammonium chloride or zinc ammonium
chloride is used, the composition of the usable flux
cover is molten zinc chloride in -which ammonium
chloride is absorbed and ammonia and hydrogen
chloride gases are trapped. The active cleaning
agent is the hydrogen chloride gas formed by the
dissociation of ammonium chloride due to heat.
Zinc chloride is present either because it -was
placed there with the ammonium chloride or be-
cause of the reaction of hydrogen chloride with the
molten zinc. The zinc chloride is necessary to
maintain the active ingredient on the zinc surface.
Foaming Agents
If galvanizing is done with a thin layer of molten
flux, a higher temperature is reached throughout
the flux layer that induces fuming and loss of flux
ingredients. The flux becomes viscous and inac-
tive in a short period of time, requiring frequent
additions of fresh flux to keep it in prime condition.
The thin molten flux cover can be fluffed up by
additions of foaming agents such as glycerine, -wheat
bran, wood flour, sawdust, and others. The re-
sulting deep-foaming type of flux cover has the ad-
vantage of reducing the quantity of objectionable
fumes. Some other advantages are longer flux life,
greater ease of control in maintaining fluidity and
fluxing activity, reduction of zinc spattering, and
saving of flux, zinc, and heat.
To form a flux cover, either ammonium chloride
or, preferably, zinc ammonium chloride is placed
on the molten zinc surface. Usually a foaming
agent such as glycerine is added to the flux before
it is applied to the kettle. If ammonium chloride
is used, the heat from the zinc causes the salt to
decompose and form hydrogen chloride and am-
monia gases. Both gases tend to rise and escape
from the kettle where they cool and recombine to
forma fume of ammonium chloride. Because the
hydrogen chloride and zinc are very reactive, they
form zinc chloride, which remains on the zinc as
a liquid at the temperature of the zinc bath. Since
only part of the hydrogen chloride is used up in
this reaction, the fumes escaping contain an excess
of ammonia. As more ammonium chloride is add-
ed to the zinc surface, the zinc chloride that is
formed begins to absorb it. At the same time a
foam filled with hydrogen chloride and ammonia
gases is formed. The foaming agent increases the
depth and fluidity of the foam. If foaming-type zinc
To achieve the required foaming action, a small
but definite amount of foaming agent is added to the
flux. Too little or too much accelerates the rate
atwhichthe flux becomes too viscous. To reduce
the amount of fuming, the foaming agent should be
mixed with the flux before the flux is placed on the
surface of the zinc bath. Of the foaming agents
mentioned, glycerine seems to be the most efficient.
Observations of the fuming tendencies of various
proprietary foaming-type fluxes have shown that
some fume more than others. The compositions
of the proprietary foaming agents have not been re-
vealed by the manufacturers.
Dusting Fluxes
After an article to be galvanized has been charged
into the kettle through the flux cover, and while it
is still completely immersed1 in the zinc bath, it is
moved to a portion of the kettle -where it can be re-
-------
Zinc-Galvanizing Equipment
403
moved through a clean zinc surface (see Figure 298).
If the articles are small, suchas bolts, nuts, nails,
and so forth, they are usually dusted with powdered
ammonium chloride immediately upon removal
from the molten zinc. The dusting flux causes the
zinc to flow and results in a smooth, bright finish.
The dusting must be done before the work has time
to cool off, since the zinc coating must still be
molten in order to flow and drain properly from
the work. At the temperature of the molten zinc,
the flux decomposes generating fumes.
THE AIR POLLUTION PROBLEM
Observations of many galvanizing kettles have re-
vealed that air contaminants are discharged when-
ever the flux cover is disturbed, fresh flux is add-
ed, or galvanized objects are dusted with ammoni-
um chloride.
that volatilize at the temperature of the molten zinc.
In one case, the cleaning and pickling solutions did
not remove all the lubricant from chain link fence
material. The oil was vaporized and discharged
as an oil mist with the fumes from the flux cover.
The oil, in fact, formed about half of the fumes
discharged. In another case, sulfur was not re-
moved from an object before it -was charged to the
kettle. The resulting fumes were yellow and much
more opaque than would normally be expected.
To obtain brighter, smoother finishes, especially
on small items, the items are dusted -with finely
ground sal ammoniac immediately after being re-
moved from the molten zinc. The items dusted are
still at a temperature well above the decomposition
temperature of sal ammoniac. Nearly all the flux
is, therefore, converted to fumes by the operation.
Although only small amounts of dusting fluxes are
used, dense fumes are always created.
Flux agitation occurs to some extent each time an
object is immersed in the zinc through the flux
cover. If the objects are smooth and dry and the
agitation is not great, the amount of fuming is
small. When the agitation is severe a correspond-
ingly larger amount of fumes is discharged. Me-
chanical actions that break some of the bubbles
making up the flux cover release fume-forming
gases.
When fresh flux is placed on the molten metal in
a kettle, some time is required for it to form a
foaming cover and, during this interval, dense
fumes escape. Moreover, when fresh flux is
stirred into the existing flux cover, fumes are dis-
charged as a result of the agitation and the time
necessary for the fresh flux to be absorbed by and
become part of the foam. If the air contaminants
were due only to the volatile constituent in the flux,
the fumes would consist only of ammonium chlo-
ride. Zinc, zinc chloride, and oil, among other
materials, have, however, been identified in the
particulate matter discharged from galvanizing
kettles.
Zinc and zinc chloride have very low vapor pres-
sures at normal galvanizing temperatures, and
one would expect neither of them to vaporize to any
great extent. The emissions from these materials
are believed to be the result of mechanical entrain -
ment, which occurs when •wet articles are gal-
vanized. Frequently an object is immersed too
rapidly, which permits some of the steam to vent
into the molten zinc below the flux cover, the rapid-
ly escaping steam atomizing some zinc and flux
into the air.
Cases have been observed where the articles to be
galvanized are not cleaned thoroughly of materials
Physical and Chemical Composition
of the Contaminants
The appearance and. composition of the fumes dis-
charged from galvanizing operations vary accord-
ing to the operation being conducted. For example,
the galvanizing of nuts, bolts, and other small ar-
ticles does not create much agitation of the flux
cover, and emissions are slight. Some fumes are,
however, generated when the articles are dusted
with ammonium chloride upon removal from the
zinc bath. An analysis of these fumes revealed
that essentially only ammonium chloride was pres-
ent.
When many different articles are galvanized, some
disturb the flux and produce more fumes than others.
The fumes also contain substantial amounts of com-
pounds other than ammonium chloride. The gal-
vanizing of chain link fence material continuously
agitates the flux cover and results in a continuous
discharge of fumes from the kettle. The visual ap-
pearance of the fumes as they are discharged into
the air from the various operations is the same--
that of light gray smoke. Evenunder a microscope
the fumes from the various sources have the same
appearance. Figure 299 is a photomicrograph of
a sample of the fumes, indicating that the average
particle size is approximately 2 microns.
Under some circumstances the fumes may have
different characteristics, but these are attributed
to the influence of additional contaminants. For
example, Table 115 shows the comparison of the
catch from an electric precipitator serving a chain
link fencing process kettle with the catch of a bag-
house serving a job shop kettle.
-------
40-
MECHANICAL EQUIPMENT
Figure 298. Removing work through a clean zinc surface.
Fiux cover in foreground (LOS Angeies Galvanizing Co.,
Hunt ington ParK. Cal i f.;..
&M&sPSgmfE3s
Figure 298 Photomicrograph of fumes discharged from a
galvani zing kettle.
The material collected by the baghouse was dry
and powdery, but it did agglomerate and was diffi-
cult to shake from the bags with ordinary bag-
shaking procedures. The material taken from the
precipitator was sticky and had the general appear-
ance of thick grease. Table 115 shows that the
fumes are different chemically, which explains
their different appearance after being collected.
The oil in the fumes collected by the precipitator
undoubtedly came from a film of oil on the chain
link fence material that was vaporized as the fence
material was charged into the hot zinc.
HOODING AND VENTILATION REQUIREMENTS
In order to control the emissions from a galvaniz-
ing kettle, the fumes generated must be conducted
to an efficient control device. In job shops, the
headroom needed makes necessary the use of either
high-canopy or room-type hoods as shown in Fig-
ures 300 and 301. The amount of ventilation vol-
ume required with high-canopy hoods increases
considerably with the height of the hood; therefore,
the size of the collector must be large enough to
accommodate the large volumes required.
Slothoods are used only when the area of fume gen-
eration is small, such as the flux box of a chain
link ience-galvanizing kettle shown in Figure 302.
The slot velocities needed to overcome the thermal
draft for the entire surface of a large kettle are
high, and large air volumes cool the surface of a
zinc bath. This cooling effect creates problems in
applying a good zinc coating and increases fuel con-
sumption. When a slot hood can be used, the amount
-------
Zinc-Galvanizing Equipment
405
Table 115. CHEMICAL ANALYSES OF
THE FUMES COLLECTED BY
A BAGHOUSE AND BY AN ELECTRIC
PRECIPITATOR FROM ZINC-
GALVANIZING KETTLES
Component
NH4C1
ZnO
ZnC12
Zn
NH3
Oil
H20
C
Not identified
Fumes collected
in a baghouse
(job shop kettle),
wt %
68.0
15.8
3.6
4.9
1.0
1.4
2.5
2.8
-
Fumes collected
in a precipitator
(chain link galvanizing),
wtj%
23. 5
6. 5
15. 2
-
3. 0
41. 4
1.2
-
9. 2
Figure 300. High-canopy hood over a galvanizing
kettle (Superior Pacific Galvanizing Company, Inc.
Los Angeles, Cali f.).
Figure 301. Opening to a room-type hood over a
galvanizing kettle (Los Angeles Galvanizing Co.,
Huntington Park, Cal if.).
of ventilation required is smaller than that required
with high-canopy hoods, and control devices are
correspondingly smaller.
Low-canopy hoods can be used on a kettle when
headroom is not required. These hoods permit
lower ventilation rates for adequate fume capture,
and smaller control devices can be used.
AIR POLLUTION CONTROL EQUIPMENT
To collect fumes having particle sizes of 2 microns
or less requires a high-efficiency collector such
as a baghouse or an electrical precipitator. A
baghouse can be usedfor any galvanizing operation
where the air contaminants do not contain oil mists.
When an oil mist is present a precipitator should
be used.
Several scrubbers, similar to the one shown in
Figure 303, have been installed in attempts to con-
trol the emissions from galvanizing kettles, but
all have been unsatisfactory. Stack analyses dis-
closed that the amount of fumes collected by these
scrubbers was negligible. In each of the scrubbers
the contaminated gases were conducted around baf-
fles, through water sprays, and finally, through
-------
406
MECHANICAL EQUIPMENT
* * • t
* ^i&& ^ >• \%»**> \ o ^ ^ °
Figure 302. Slot-type hood serving a chain link fence-
galvanizing flux box (Anchor Post Products, Inc., of
California, Whittier, Calif.).
Figure 303. Water-wash scrubber serving
a continuous chain link galvanizing
kettle.
mist eliminators. Water was recirculated through
the scrubbers with only sufficient makeup to re-
place the amount lost due to evaporation and mist
discharge. The water pressure at the spray heads
•was approximately 25 psig in each scrubber.
Baghouses
Cotton cloth bags have been found to be an effective
filtering medium for baghouses serving the fumes
discharged from most galvanizing operations. Nei-
ther the fumes nor the gases discharged are del-
eterious to cotton, nor are they corrosive to the
baghouse shell. Because of the large volume of air
needed to capture the air contaminants, the tem-
perature of the gases is well below the 180°F limit
of cotton bags.
The fumes have a tendency to agglomerate, en-
hancing filtration; however, they also cling to the
bags, making difficult the cleaning of the bags by
mechanical shakers. The filtering velocity has a
marked effect on the tenacity of the fumes. Only
mechanical shaking was found necessary with a
filter velocity of 1 fpm. Withfrom 2- to 3-fpm ve-
locities, the bags had to be shaken mechanically at
2-hour intervals, and then every 2 weeks each bag
had to be vigorously shaken by hand. With filter
velocities in excess of 3 fpm, the fumes could not
be removed, even with vigorous shaking every 2 or
3 days.
Because low filtering velocities are required for
effective filtration, and large exhaust volumes, for
adequate fume capture, the baghouse will be large.
Figure 304 shows a baghouse with 13,200 square
feet of filter area being used to control the fumes
discharged from the kettle shown in Figure 300.
The following example shows some of the factors
that must be considered in designing a control sys-
tem for a galvanizing kettle.
Example 32
Given:
A galvanizing kettle, 4 feet wide by 25 feet long by
3 feet high contains molten zinc at a maximum tem-
perature of 860°F. The products of combustion
do not mix with the air contaminants from the ket-
tle. The oil and moisture content of the contami-
nants are assumed tobe negligible. The hood con-
figuration is such that one side -will be a part of the
building wall extending to the floor, and the oppo-
site side will be constructed of sheet metal, extend-
-------
Zinc-Galvanizing Equipment
407
Figure 304. Baghouse serving a galvanizing kettle
(Superior Pacific Galvanizing Co., Inc., Los
Angeles, Calif.).
The method involves calculating: (1) The heat loss
(H) from the process, (2) the hot air induction rate
(Qz)> P) the dimensions of the column of hot air
at the base of the hood, (4) the hood dimensions,
(5) the required exhaust rate (Q), and (6) the tem-
perature of the exhaust gases. The sizes of the
ductwork, baghouse, and fan can then be calculated.
1. Heat loss from kettle:
For horizontal hot surfaces
H
At
At
- A (At) (from Chapter 3)
60 s
= Hot surface area = (4)(25) = 100 ft
= Temperature difference between the hot
surface and the atmosphere
Assume air temperature to be 70 °F
Maximum zinc temperature = 860°F
= 860 - 70 = 790°F
H' =
60
(100)(790)5/4 = 2,660 Btu/min.
ing to within 8 feet from the floor. The ends of the
hood must be provided with crane-way access open-
ings 16 feet above the floor.
HOOD
HOOD
BAGHOUSE
VV
n
KETTLE
Figure 305. Design of problem presented in
example 32.
Problem:
Determine the design features of an air pollution
control system using a baghouse.
Solution:
By using the methods described in Chapter 3, the
required exhaust rate (Q) can be determined.
2. Hot air induction rate:
Qz = 7.4 (Z)3/2 (Hr) (from Chapter 3)
Z = effective height from the hypothetical point
source to the base of the hood = Y + 2B
Because of the configuration of the hood, the
value of Y is not clear. Although one side of
the hood extends to the floor and the other side
is 5 feet above the kettle, there will be open-
ings in each end extending to 1 3 feet above the
kettle. To ensure capturing the air contami-
nants, design for a hood height of Y = 13 feet
above the kettle.
The value for B also is not clear. In the der-
ivation of the equation, B is the diameter of
the hot surface and is used to calculate the ex-
pansion of the column of hot gases arising from
the hot surface. The hot surface in this case
is rectangular, 4 ft wide by 25 ft long. The
expansion of the column of hot gases is due to
mixing with cooler air. The cool air mixes
from all sides and is motivated by the tem-
perature differential. When the cool air pene-
trates halfway through the column it meets
cool air entering from the opposite side, and
thus cancels the driving force. From this it
is apparent that the short dimension of the hojt-
-------
4Q.8
MECHANICAL EQUIPMENT
air column must control the expansion of the
column. Therefore, B = 4 feet.
Z = 13. + (2),(4) = 21 ft
v = 7.4 (21)3/2 (
= 9,800cfm.
3. Dimensions of hot gas column at base of hood:
(z,0-1
(from Chapter 3)
D = B (See explanation in Item 2 above.)'
,0.88
D =
(21)
14.6
= 7. 3 ft
Assume that the length will expand the same
amount as the width.
Width expansion = 7.3 - 4 = 3.3ft
Length = 25 + 3.3 = 28. 3 ft
Dimensions of hot gas column = 7. 3-ft width
by 28. 3-ft length.
4. Hood dimensions:
Crossdrafts across each hood will be mini-
mized because the sides of the hood are low,
extending to the floor on one side and 5 feet
above the kettle on the other side. The high
openings on each end of the hood could, how-
ever, cause crossdrafts, blowing the fumes
away from the hood. The hood dimensions
should be larger than the dimensions of the
rising hot gas stream, the length being ex-
tended more than the width. A hood -with di-
mensions of 10 feet wide by 40 feet long should,
therefore, be provided.
5. Required exhaust rate:
Q = Q,7 + VA (from Chapter 3)
air
V = velocity of indraft required to keep
moving into all areas of hood.
A = hood area not occupied by the entering
hot gas current.
Design for V = 100 fpm
A = (10 x 40) - (7. 3 x 28. 3) = 400 - 206 = 194 ft
Q= 9,800 + (100)(194) = 29, 200 cfm
Design for 30, 000 cfm.
6. Exhaust gas temperature:
The temperature rise of the air is:
ST1 _ _ £.
Exhaust gas temperature = 75° F.
The temperature rise is not sufficient to affect
any of the following calculations, and is, there-
fore, neglected.
7. Duct diameter between hood and baghouse:
Use recommended velocity of 2, 000 ft/min
™ <- <.- 30,000 2
JJuct cross-section area = —— = 15 it
TT d
4
d
= 15 ft
= 4. 37 ft
Use a duct diameter of 52-1/2 inches
Note: By using a velocity greater than the min-
imum, the duct diameter can be decreased to
reduce construction costs. Horsepower re-
quirements will, however, be increased.
8. Required filter area of baghouse:
Provide a filtering velocity of 2 fpm
Filter area = 3°'2°00 = 15,000ft2.
9. The exhaust system and fan calculations are
made as outlined in Chapter 3. After a sys-
tem resistance curve is plotted and calculated,
a fan is selected whose characteristic curve
will intersect the system curve at the required
air volume of 30, 000 cfm.
Electrical Preci pitators
The use of a two-stage, low-voltage-type precip-
itator, as shown in Figure 306, has been investi-
gated for the control of galvanizing fumes in the
Los Angeles area. The investigation led to the use
of the precipitator to control fumes containing oil
from the flux box of a chain link fence-galvanizing
operation. The investigation also revealed that the
precipitator could not compete economically with
-------
Zinc-Galvanizing Equipment
409
a baghouse to control the dry and much more di-
lute fumes captured by a high-canopy hood serving
the entire galvanizing kettle.
Figure 306. Experimental electric precipitator
used in a galvanizing control study (Advance
Galvanizing Co., Los Angeles, Calif.).
serve the entire kettle would have to be operated.
at a velocity of at least 340 fpm to compete eco-
nomically with a baghouse.
The following example shows some of the factors
that must be considered in designing an exhaust
system with an electrical precipitator to control
the air contaminants discharged from a chain link
fence-galvanizing operation.
Example 33
Given:
A chain link fence-galvanizing kettle is provided
with a flux box, 10 inches wide by 10 feet long by
1 foot high. Zinc ammonium chloride is used as
a cover flux in the flux box. A slot hood is to be
used along one side of the flux box to capture the
fumes created.
SLOT HOOD
PRECIPITATOR
Figure 307. Design of problem presented in
example 33.
When only the flux box of a chain link fence-gal-
vanizing operation was vented, the air contami-
nants consisted of 41 percent by "weight oil mist
and 59 percent fumes. The concentration of the
air contaminants in the exhaust stream was 0. 154
grain per scf. With an exhaust gas velocity of 58
fpm through the precipitator, the collection effi-
ciency was 91 percent. With an air contaminant
concentration of 0. 072 grain per cubic foot and a
velocity of 330 fpm through the precipitator, the
collection efficiency was 79 percent.
When the entire kettle was vented with the aid of a
room-type hood, the air contaminants consisted
of 5 percent by weight oil and 95 percent fumes.
With an air contaminant concentration in the ex-
haust gases of 0. 0072 grain per scf and a gas ve-
locity of 340 fpm through the precipitator, the col-
lection efficiency was zero. Further tests of the
precipitator at lower velocities -were not •warranted,
because at this plant a full-scale precipitator to
Problem:
Determine the design features of an air pollution
control system using an electrical precipitator.
Solution:
1. Exhaust volume:
Design for 200 cfm per ft of flux box area
10 2
Area = -± (10) = 8.33 ft
(8.33)(200) = 1,666 cfm
Design for 1, 700 cfm.
2. Slot width:
Design for a slot velocity of 2, 000 fpm
-------
410
MECHANICAL, EQUIPMENT
Slot area = "7"^" = °- 85 ft = 122- 4 in-
Length - 10 ft = 120 in.
122.4
Width =
120
1. 02 in.
3. Diameter of duct from hood to precipitator:
Design for 2, 000 fpm
Use 12-1/2-in. -diameter duct.
4. Cross-sectional area of precipitator:
Design for 100 fpm
Area =
1, 700
100
= 17 ft
The exhaust system and fan calculations are
made as outlined in Chapter 3. After a sys-
tem resistance curve is plotted and calculated,
a fan is selected whose characteristic curve
intersects the system curve at the required
volume, which in this example, "will be 1, 700
elm.
-------
CHAPTER 8
INCINERATION
DESIGN PRINCIPLES FOR MULTIPLE-CHAMBER INCINERATORS
JOHN E. WILLIAMSON, Senior Air Pollution Engineer
GENERAL-REFUSE INCINERATORS
ROBERT J. MAC KNIGHT, Principal Air Pollution Engineer
JOHN E. WILLIAMSON, Senior Air Pollution Engineer
MOBILE MULTIPLE-CHAMBER INCINERATORS
ARTHUR B. NETZLEY, Intermediate Air Pollution Engineer
JOHN E. WILLIAMSON, Senior Air Pollution Engineer
MULTIPLE-CHAMBER INCINERATORS FOR BURNING WOOD WASTE
ARTHUR B. NETZLEY, Intermediate Air Pollution Engineer
JOHN E. WILLIAMSON, Senior Air Pollution Engineer
FLUE-FED APARTMENT INCINERATORS
JOSEPH J. SABLESKI, Air Pollution Engineer*
JOHN E. WILLIAMSON, Senior Air Pollution Engineer
PATHOLOGICAL-WASTE INCINERATORS
PAUL G. TALENS, Air Pollution Engineer
DEBONDING OF BRAKESHOES AND RECLAMATION OF ELECTRICAL EQUIPMENT WINDINGS
ARTHUR B. NETZLEY, Intermediate Air Pollution Engineer
DONALD F. WALTERS, Intermediate Air Pollution Engineer*
JOHN E. WILLIAMSON, Senior Air Pollution Engineer
DRUM RECLAMATION FURNACES
ROY S. BROWN, Air Pollution Engineer
ARTHUR B. NETZLEY, Intermediate Air Pollution Engineer
WIRE RECLAMATION
ARTHUR B. NETZLEY, Intermediate Air Pollution Engineer
*Now with National Center for Air Pollution Control, Public Health Service, U. S. Department of Health,
Education, and Welfare, Cincinnati, Ohio.
-------
CHAPTER 8
INCINERATION
DESIGN PRINCIPLES FOR
MULTIPLE-CHAMBER INCINERATORS
Disposal of combustible refuse and garbage is one
of the most perplexing problems facing urban so-
ciety tpday. The greater the population density the
more disturbing the problem. This refuse is cre-
ated by all elements of a community — industry,
commerce, and the public.
In the past, disposal of combustible -wastes was
lookeduponas anecessaryevil to be accomplished
as cheaply as possible. Industrial and commercial
installations used a box-like, single-chamber in-
cinerator to burn up to several tons a day. Refuse
from apartment houses was generally burned in a
chute-fed, single-chamber incinerator. In some
areas, especially southern California, each home-
owner disposed of his combustible refuse in a back-
yard incinerator.
During the past 1 5 years almost every large urban
area in the world has experienced a drastic in-
crease in the pollution of its atmosphere. As the
discomforts of air pollution became more notice-
able, public clamor for rigid regulation of air-
contaminating processes increased steadily. In
Los Angeles County this led to the banning of open
fires and single-chamber incinerators in Septem-
ber 1957. Since that date all incinerators con-
structed and put into operation in the county have
had to meet stringent criteria of performance as
well as definite minimum design requirements.
The standards presented in this chapter are tools
for creating designs for multiple-chamber incin-
erators that may be expected to burn rubbish with
a minimum discharge of air contaminants. Tab-
ular presentations alone are not sufficient for the
best application and understanding of the princi-
ples of design involved. Also essential is an un-
derstanding of the many factors that created the
need for a new approach to incineration and the
development of the multiple-chamber incinerator.
The design recommendations and supplementary
discussions provide answers to many of the ques-
tions that confront designers and operators of multi-
ple-chamber equipment. Caution is needed, howev-
er, in that only those qualified in combustion equip-
ment design and refractory construction should
try to apply the standards presented. Adequacy
of design, proper methods of construction, and
quality of materials are important to the satis-
factory completion of an incinerator that will meet
air pollution control requirements and have an av-
erage service life expectancy.
In this part of the chapter, the two basic types of
multiple-chamber incinerators are compared.
Moreover, the principles of combustion, the funda-
mental relationships for incinerator design, and
general design factors are discussed. These data
are, for the most part, applicable to other parts
of this chapter that discuss incinerators for spe-
cific uses.
In addition to discussing incinerators for burning
of combustible wood, paper refuse, and garbage,
this chapter includes the design of incinerators for
the reclamation of steel drums and wire and for
debonding of brakeshoes.
The configuration of modern multiple-chamber in-
cinerators falls into two general types as shown in
Figures 308 and 309. These are the retort type,
named for the return flow of gases through the "U"
arrangement of component chambers, and the in-
line type, so-called because the component cham-
bers follow one after the other in a line.
RETORT TYPE
Essential features that distinguish the retort type
of design are as follows.
1. The arrangement of the chambers causes the
combustion gases to flow through 90-degree
turns in both lateral and vertical directions.
2. The return flow of the gases permits the use
of a common wall between the primary and
secondary combustion stages.
3. Mixing chambers, flame ports, and curtain
wall ports have length -to -width ratios in the
range of 1:1 to 2. 4:1.
4. Bridge wall thickness under the flame TDort is
a function of dimensional requirements in the
mixing and combustion chambers. This re-
sults in construction that is somewhat unwieldy
in the size range above 500 raounds per hour.
IN-LINE TYPE
Distinguishing features of the in-line-type design
are as follows.
1. Flow of the combustion gases is straight through
the incinerator with 90-degree turns only in
the vertical direction.
413
-------
414
INCINERATION
CHAMBER y-FLME PORT
SECONDARY
COMBUSTION
CHMIBLR
CURTAIN1
WALL PORT
CLEANOUT DOOR
KITH UNDERGRATE
AIR PORT
CLEANOUT
DOOR
GRATES
IGNITION
CHAMBER
CHARGING DOOR
WITH OVERFIRE
A.IR PORT
IGNITION CHAMBER
GRATES
ASH PIT
STACK
SECONDARY
AIR PORT
MIXING CHAMBER
BURNER PORT
MIXING CHAMBER
CURTAIN HALL PORT
Figure 308. Cutaway of a retort multiple-chamber incinerator.
-------
Design Principles for Multiple-Chamber Incinerators
415
i-SECONDARY
COMBUSTION
CHAMBER
GRATES
-CLEANQUT DOORS KITH
UNDERGRATE AIR PORTS
MIXING CHAMBER
CURTAIN
WALL PORT
Figure 309. Cutaway of an in-line multiple-chamber incinerator.
2. The in-line arrangement is readily adaptable
to installations that require separated spacing
of the chambers for operating, maintenance,
or other reasons.
3. All ports and chambers extend across the full
width of the incinerator and are as wide as the
ignition chamber. Length-to-width ratios of
the flame port, mixing chamber, and curtain
•wall port flow cross sections range from 2:1
to 5:1.
DESCRIPTION OF THE PROCESS
The combustion process in a multiple-chamber in-
cinerator proceeds in two stages --primary or solid
fuel combustion in the ignition chamber, followed
by secondary or gaseous-phase combustion. The
secondary combustion zone is composed of two
parts, a downdraft or mixing chamber and an up-
pass expansion or combustion chamber.
The two-stage multiple-chamber incineration pro-
cess begins in the ignition chamber and includes
the drying, ignition, and combustion of the solid
refuse. As the burning proceeds, the moisture
and volatile components of the fuel are vaporized
andpartially oxidized in passing from the ignition
chamber through the flame port connecting the ig-
nition chamber with the mixing chamber. From the
flame port, the volatile components of the refuse
and the products of combustion flow down through
the mixing chamber into which secondary air is
introduced. The combination of adequate temper-
ature and additional air, augmented by mixing
chamber or secondary burners as necessary, as-
sists in initiating the second stage of the combus-
tion process. Turbulent mixing, resulting from
the restricted flow areas and abrupt changes in
flow direction, furthers the gaseous-phase reac-
tion. In passing through the curtain wall port
from the mixing chamber to the final combustion
chamber, the gases undergo additional changes in
direction accompanied by expansion and final ox-
idation of combustible components. Fly ash and
other solid particulate matter are collected in the
combustion chamber by v/all impingement and sim-
ple settling. The gases finally discharge through
a stack or a combination of a. gas cooler (for ex-
ample, a water spray chamber) and induced-draft
system. Either draft system must limit combus-
tion air to the quantity required at the nominal
capacity rating of the incinerator.
-------
416
INCINERATION
D'ESIGN TYPES AND LIMITATIONS
During the evaluation and development phases of
the multiple-chamber incinerator, different incin-
erator configurations with variations in the sizes
and shapes of the several chambers and ports were
tested. The results of these tests defined the op-
timum operating limits for the two basic styles of
multiple-chamber incinerators. Each style has
certain characteristics with regard to performance
and construction that limit its application.
Comparison of Types
The basic factors that tend to cause a difference
in performance in the two incinerators are (1)
proportioning of the flame port and mixing cham-
ber to maintain adequate gas velocities within di-
mensional limitations imposed "by the particular
type involved, (2) maintenance of proper flame
distribution over the flame port and across the mix-
ing chamber, and (3) flame travel through the mix-
ing chamber into the combustion chamber.
A retort incinerator in its optimum size range of-
fers the advantages of compactness and structural
economy because of its cubic shape and reduced
exterior wall length. It performs more efficiently
than its in-line counterpart in the capacity range
from 50 to 750 pounds per hour. In these small
sizes, the nearly square cross sections of the ports
and chambers function well because of the abrupt
turns in this design. In retort incinerators with a
capacity of 1, 000 pounds per hour or greater, the
increased size of the flow cross section reduces
the effective turbulence in the mixing chamber and
results in inadequate flame distribution and pene-
tration and in poor secondary air mixing.
No outstanding factors favor either the retort or
the in-line configurations in the capacity range of
750 to 1, 000 pounds per hour. The choice of re-
tort or in-line configuration in this range is influ-
enced by personal preference, space limitations,
the nature of the refuse, and charging conditions.
The in-line incinerator is well suited to high-ca-
pacity operation but is not very satisfactory for
service in small sizes. The smaller in-line in-
cinerator s are somewhat less efficient with regard
to secondary stage combustion than the retort type
is. In in-line incinerators with a capacity of less
than 750 pounds per hour, the shortness of the grate
length tends to inhibit flame propagation across the
width of the ignition chamber. This, coupled with
thin flame distribution over the bridge wall, may
result in the passage of smoke from smo'.dering
grate sections straight through the incinerator and
out of the stack without adequate mixing and secon-
dary combustion. In-line models in sizes of 750
pounds per hour or larger have grates long enough
to maintain burning across their width, resulting
in satisfactory flame distribution in the flame port
and mixing chamber. The shorter grates on the
smaller in-line incinerators also create a mainte-
nance problem. The bridge wall is very suscep-
tible to mechanical abuse since it is usually not
provided with a structural support or backing and
is thin where the secondary airlanes are located.
Careless stoking and grate cleaning in tile short-
grate in-line incinerators can break down the bridge
wall in a short time.
The upper limit for the use of the in-line inciner-
ator has not been established. Incinerators with
a capacity of less than 2,000 pounds per hour may
be standardized for construction purposes to a
great degree. Incinerators of larger capacity,
however, are not readily standardized since prob-
lems of construction, material usage, m-echanized
operation with stoking grate, induced-draft sys-
tems, and other factors make each installation
essentially one of custom design. Even so, the
design factors advocated herein are as applicable
tothe design of large incinerators as to the design
of smaller units.
PRINCIPLES OF COMBUSTION
Theoretical treatment of the complex reactions
taking place in combustion processes is as yet in-
complete, but the empirical art of combustion en-
gineering has developedto an advanced state. The
principles of solid-fuel combustion generally apply
to incineration processes and include the following.
1. Air and fuel must be in proper proportion.
2. Air and fuel, especially combustible gases,
must be mixed adequately.
3. Temperatures must be sufficient for ignition
of both the solid fuel and the gaseous compo-
nents .
4. Furnace volumes must be large enough to pro-
vide the retention time needed for complete
combustion.
5. Furnace proportions mast be such that igni-
tion temperatures are maintained and fly ash
entrainment is minimized.
Fluctuation in fuel quality makes satisfactory in-
cinerator design difficult. In addition to wide ranges
in composition, wetness , and volatility of fuel, there
is diversity in ash content, bulk density, heat of
combustion, burning rate, and component particle
size. All these affect, to some extent, the oper-
ating variables of flame propagation rate, flame
length, combustion air requirement, and the need
for auxiliary heat.
-------
Design Principles for Multiple-Chamber Incinerators
417
Fundamental relationships for incinerator design
were investigated by Rose and Crabaugh (1955)
and by the ASME Subcommittee on Incineration
Design Standards. The following were studied:
1. The relationship of combustion air distribu-
tion to the degree and rate of combustion at-
tained and to the discharge of air contami-
nants;
2. trie relationship of furnace proportions, that
is, chambers and ports, to the degree and rate
of combustion;
3. the effects of temperature and furnace design
on the percentage of acid, volatile organic,
and solid contaminants discharged and the per-
centage of combustibles in the solid contami-
nants discharged;
4. the relationship of combustion gas velocities
to the effects on turbulence and flame travel
and to the degree of combustion attained;
5. the relationship of the material "burned to the
formation of acid and volatile organic com-
pounds.
DESIGN FACTORS
Control of the combustion reaction, and reduction
in the amount of mechanically entrained fly ash are
most important in the efficient design of a multiple-
chamber incinerator. Ignition chamber parame-
ters are regarded as fundamental since solid con-
taminant discharges can be functions only of the
mechanical and chemical processes taking place
in the primary stage. Other important factors in-
clude the ratios of combustion air distribution, sup-
plementary draft and temperature criteria, and
the secondary-combustion-stage velocity and pro-
portion factors. Some of these factors are func-
tions of the desired hourly combustion rate and
are expressed in empirical formulas , while others
are assigned values that are independent of incin-
erator size.
Table 116 lists the basic parameters, evaluation
factors , and equations for designing multiple -cham-
ber incinerators and gives the minimum values
established for each. The allowable deviations
should be interpreted with discretion to avoid con-
sistently high or low deviation from the optimum
values . Application of these factors to design eval-
uation must be tempered by judgment and by an ap-
preciation of the practical limitations of construc-
tion and economy.
The values determined for the several parameters
are mean empirical values, accurate in the same
degree as the experimental accuracy of the eval-
uation tests. The significance of exact figures is
reduced further by the fluctuation of fuel composi-
tion and conditions . For purposes of design, per-
missible variations from the optimum mean are
plus or minus 10 percent, and velocities may de-
viate as much as 20 percent without serious con-
sequence.
The formulas governing ignition chamber design
were tentatively postulated from data available
through tests of units of varying proportions burn-
ing at maximum combustion rates. Optimum values
of the arch height and grate area maybe determined
by using the gross heating value of the refuse to
be burned and interpolating between the upper and
lower curves in Figures 310 and 311. An allow-
able deviation of these values of plus or minus 10
percent is considered reasonable. Rather than
establish formulas for both the upper and lower
curves of these figures, which represent 9,000
Btu per pound or more and 7, 500 Btu per pound
or less, respectively, a formula for the average
values of the two curves has been given. This
curve corresponds to a gross heating value of
8, 250 Btu per pound.
Design Precepts
The ignition mechanism should be one of fuel bed
surface combustion. This is attained by the pre-
dominant use of overfire combustion air and by
charging in such a manner as to attain concurrent
travel of both air and refuse with minimum admis-
sion of undertire combustion air. Limiting the
admission of underfire air and thereby maintaining
relatively low fuel bed temperatures is important.
With a relatively high air rate through the fuel bed,
the stack effluent contains appreciable quantities of
metallic salts and oxides in microcrystalline form.
A probable explanation is that vapor phase reac-
tions and vaporization of metals take place in high
fuel bed temperatures with resultant condensation
of particles in the effluent gases as they cool upon
leaving the stack.
To accomplish fuel bed surface combustion through
use of overfire air, the charging door should be
located at the end of the ignition chamber farthest
from the flame port, and the fuel moved through
the ignition chamber from front to rear. This way,
the volatiles from the fresh charge pass through
the flames of the stabilized and heated portion of
the burning fuel bed. Also, the rate of ignition of
unburned refuse is controlled, which prevents flash
volatilization with its resultant flame quenching and
smoke creation. Top or side charging is consid-
ered disadvantageous because of the suspension of
dust, disturbance of the stabilized fuel bed, and
the additional stoking required.
With good regulation of the burning rate through
proper charging, air port adjustment, and the use
of an ignition or '"primary" burner, the need for
-------
418
INCINERATION
Table 116. MULTIPLE-CHAMBER INCINERATOR DESIGN FACTORS
Item and symbol
Recommended value
Allowable
deviation
Primary combustion zone:
Grate loading, LQ
Grate area, AQ
Average a.rch height, H*
Length-to-width ratio (approx):
Retort
In-line
10 Log Rc, lb/hr-ft2 where Rc equals the
refuse combustion rate in Ib/hr (refer to
Fig. 310)
Rc * LG; ft2
4/3 (AG)4/11; ft (refer to Fig. 311)
Up to 500 Ib/hr,2:1; over 500 Ib/hr, 1. 75:1
Diminishing from-about 1.7:1 for 750 Ib/hr
to about 1:2 for 2, 000 Ib/hr capacity. Over-
square acceptable in units of more than 11 ft
ignition chamber length.
10%
10%
Secondary combustion zone.
Gas velocities:
Flame port at 1, 000°F, VFp
Mixing chamber at 1,000°F,
Curtain wall port at 950 °F, V
Combustion chamber at 900 °F,
Mixing chamber downpass length,
from top of ignition chamber arch to top
of curtain wall port.
Length-to-width ratios of flow cross
sections:
Retort, mixing chamber, and combus-
tion chamber
In-line
55 ft/sec
25 ft/sec
About 0, 7 of mixing chamber velocity
5 to 6 ft/sec; always less than 10 ft/sec
Average arch height, ft
Range - 1.3:1 to 1. 5:1
Fixed by gas velocities due to constant
incinerator width
± 20%
+ 20%
+ 20%
Combustion air:
Air requirement batch-charging opera-
tion
Combustion air distribution:
Overfire air ports
Underfire air ports
Mixing chamber air ports
Port sizing, nominal inlet velocity
pres sure
Air inlet ports oversize factors.
Primary air inlet
Underfire air inlet
Secondary air inlet
Basis: 300% excess air. 50% air require-
ment admitted through adjustable ports;
50% air requirement met by open charge
door and leakage
70% of total air required
10% of total £.ir required
20% of total air required
0. 1 inch water gage
1. 2
1. 5 for over 500 Ib/hr to 2. 5 for 50 Ib/hr
2. 0 for over 500 Ib/hr to 5. 0 for 50 Ib/hr
Furnace temperature.
Average temperature, combustion
products
1,000'F
+ 20°F
Ajuxiliary burners'
Normal duty requirements.
Primary burner
Secondary burner
3, 000 to 10, OOO^Btu per Ib of moisture in
4,000 to 12,000/the refuse
Draft requirements:
Theoretical stack draft, DT
Available primary air induction draft,
D^. (Assume equivalent to inlet ve-
locity pressure.)
Natural draft stack velocity, Vg
0. 15 to 0. 35 inch water gage
0. 1 inch water gage
Less than 30 ft/sec at 900°F
-------
General-Refuse Incinerators
419
10,000
, 000
4n nn
, uuu
, 000
o nnn
I , UUU
1 nnn
i»
^
_C2
^ 500!
u
ac Ann
^ 4UU
UJ
i— inn
5 J""
a:
z
o on n
3 C
3 C
•4 i-
usnawoo
30
?n
10
*
*
*
*
*
y >
/"
/ / j
9 f *
•** / •**
///
/ / /
ft/
^/
? A
* ^
/ /
/
/
/ / /
/ / /
/ //
* / •***
f *
^T »
*
*
.*
FHR npy
VALUES,
FOR MO IS
VALUES,
•• t
/
* f
» f
; /
/
/ / /
/^ **
r7
«*
*
.*
s
'
LG = 10 LOG Rc
REFUSE AND HIGH HE
USE +10% CURVE.
T REFUSE AND LOW H
USE -10% CURVE.
*
/
*
AT||qr;
EATING1
10
20
30
40
50
GRATE LOADING (LG), Ib/ft^-hr
Figure 310. Relationship of grate loading to combustion rate for multiple-chamber incinerators.
stoking can be reduced to that necessary for fuel
bed movement before the charging.
Application of the fundamental evaluation precepts
combined with admission of secondary air and with
trials of various proportions in both chamber and
port dimensions established parameters for the
mixing and combustion chamber portions of the
multiple-chamber incinerator. The primary ef-
fect of proper design has been attainment of a high-
er degree of completion of combustion of volatile
and solid combustible effluent components. De-
signing the combustion chamber as a settling cham-
ber has made possible a reduction in fly ash emis-
sions as well.
GENERAL-REFUSE INCINERATORS
The general refuse incinerators discussed here
are used for refuse originating from residences
and commercial and industrial establishments . Ex-
cluded, however, are the flue-fed, wood-burning,
and mobile incinerators, which are discussed in-
dividually in other parts of this chapter. General
refusemay be defined as combustible refuse such
as dry paper or a variable mixture of dry paper
and other combustible materials within the follow-
ing approximate maximum limits (percent of weight):
Drypaper (100); wood, scrap (50): shrubbery (30);
garbage (30); and sawdust, shavings (10).
-------
420
INCINERATION
CJ
ce
9
8
7
6
5
4
2
X^
*
••'
t#*
.•
,••
»•
^,sx-
IP
•V*
FOR OF
VALUES
FOR NIC
VALUES
y
X
XI
/•
X
X
,X
#*
*
..»
4
#
/
'"
HA = 4 (AQ) ' '
Y REF
, USE
1ST R
, USE
USE
+1C
EFUS
-1C
AN!
% (
E /
\% (
I H
:UR
\ND
:UR
It
VE
L
VE
H
oy
.•
.*'
H
*
,•
:/
HE
,X*
xX
XX*
X*
JING
.ATING
X
[•••'
2 345
10 20 30 40 50
GRATE AREA (AG), ft2
100
500 1,000
Figure 311. Relationship of arch height to grate area for multiple-chamber incinerators.
Basically, disposal of general refuse may be ac-
complished by incineration or disposal in a dump.
The burning dump that has been used for centuries
is rapidly becoming outdated as more and more
communities become conscious of air pollution.
Other types of incineration range from the use of
perforated 55-gallon drums, single-chamber in-
cinerators, and multiple-chamber incinerators to
large municipal incinerators. Where land is avail-
able, the cut and cover dump represents a more
desirable method of waste disposal than municipal
incineration from the standpoint both of economics
and air pollution control.
THE AIR POLLUTION PROBLEM
The incineration process in general refuse incin-
erators produces emissions of fly ash, smoke,
gases, and odors. Flyash and odors are undesir-
able primarily because of their nuisance potential
to the occupants of neighboring dwellings and busi-
nesses. Smoke and gases, which also have a nui-
sance potential, contribute to overall air pollution
through reduction in visibility and through their
ability to enter into smog-forming photochemical
reactions in the air.
Since single-chamber incinerators offer the advan-
tage of positive control of combustion air distribu-
tion and that of concentration of heat by virtue of
enclosing the fire within refractory walls, they are
believed to be considerably more effective than an
open fire. Even so, single-chamber incinerators
have been found to have particulate emissions of
from 14 to 35 pounds per ton of material burned.
By contrast, the particulate discharges from •well-
designed multiple-chamber incinerators average
4. 5 pounds per ton of refuse burned, which is one-
tenth to one-fourth the amount of solid and liquid
combustion contaminants emitted from single-
chamber units. Average amounts of particulate
emissions, as well as of the major gaseous con-
taminants, from single- and multiple-chamber in-
cinerators are given in Table 117.
AIR POLLUTION CONTROL EQUIPMENT
As always, the best methods of air pollution con-
trol are prevention of the creation of air pollutants
by disposal of the refuse in landfill projects. The
next best means of controlling air pollution from
the incineration of general re'fuse is complete com-
bustion in a multiple-chamber incinerator. The
-------
General-Refuse Incinerators
421
remainder of this part of the chapter is limited to
the design of multiple-chamber incinerators for
effective disposal of general refuse with a minimum
creation of air pollution.
Table 117. COMPARISON BETWEEN
AMOUNTS OF EMISSIONS FROM SINGLE- AND
MULTIPLE-CHAMBER INCINERATORS
Item
Particulate matter, gr/scf at 12% CO^
Volatile matter, gr/scf at 12% CO^
Total, g-n/scf at 12% CO^
Total, Ib/ton reiuse burned
Carbon monoxide, Ib/ton of refuse burned
Ammonia, Ib/ton of refuse burned
Organic acid (acetic), Ib/ton of refuse burned
Aldehydes {formaldehyde) , Ib/ton oi refuse burned
Nitrogen oxides, Ib/ton of refuse burned
Hydrocarbons (he\ane), Ib/ton Ql refuse burned
Multiple
chamber
0. 11
0.07
0 19
3. 50
2. 90
0
0 22
0.22
2.50
< 1
Single
chamber
0.9
0. 5
1. 4
23 8
197 to 991
0.9 to 4
< 3
5 to 64
•- 0 1
DESIGN PROCEDURE
The design factors itemized in Table.116^ are the
basis for the design of a multiple-chamber incin-
erator. .These facto_rs are used_to_dete.rrnine_the_
area of the grate, the average height of the arch,
theproportioningjo.fthe_ignition chamber, the siz-
ing of the ^asjjorts, the cross section of the mix-
ing ^chamber, the ..sizes of the inlet air ports, and
the other necessary dimensions and proportions.
Application of these factors, however, requires
that calculations be made to convert the data into
usable form. These calculations are illustrated
in the problem given at the end of this part of the
chapter.
These calculations fall into three general cate-,
gories: (1) Combustion calculations based upon
the refuse composition, assumed air requirements,
and estimated heat loss; (2) flow calculations based
upon the~ properties of the products of combustion
and assumed gas temperatures; and (3) dimension-
al calculations based upon simple mensuration and
empirical sizing equations. The calculations need-
ed to determine -weights, velocities, and average
temperatures of the products of combustion are de-
rived from standard calculation procedures for
combustion. Average gross heating values and
theoretical air quantities are used. Chemical prop-
erties and combustion data for the major compo-
nents of general refuse are given in Table 118. The
only omission is shrubbery, which may be safely
assumed to have the same composition as average
wood.
The average temperature of the combustion prod-
ucts is determined through normal calculations of
heat loss. The burning rate and average composi-
tion of the refuse are assumed to be constant. When
extremes in quality and composition of material
are encountered, the most difficult burning condi-
tion is assumed. Heat losses due to radiation, re-
fractory heat storage, and residue heat content are
assumed to average 20 to 30 percent of the gross
heating value of the refuse during the first hour of
operation. Readily available furnace data indicate
that the losses fall to approximately 10 to 15 per-
cent of the gros s heat after 4 to 5 hours of continu-
ous operation.
The calculated overall average gas temperature
should be about 1,000°F when calculations are
based on 300 percent excess combustion air and
the assumption of 20 to 30 percent heat loss given
previously. This calculated temperature is not
flame temperature and does not indicate the prob-
able maximum temperatures attained in the flame
port or mixing chamber. If the calculated tem-
perature is lower than 1,000°F, installation of
burners is indicated.
Only volume and temperature data for the prod-
ucts of combustion are required for determining
the cross-sectional flow areas of the respective
ports and chambers. The temperatures used are
approximations of the actual temperature gradient
in the incinerator as the products of combustion
cool while passing through the various ports and
chambers to the stack outlet.
Air ports are sized for admission of theoretical air
plus 100 percent excess air. The remaining air
enters the incinerator through the open charging
door during batch operation and through expansion
joints, cracks around doors, and so forth. Indraft
velocities in the combustion air ports (overfire,
underfire, and secondary) are assumed to be equal,
with a velocity pressure of 0. 1 inch water column
(equivalent to 1, 265 fpm). Designing the draft sys-
tem s o that available iirebox draft is about 0. 1 inch
water column, and oversizing the adjustable air-
ports ensure maintenance of proper air induction.
Calculations of draft characteristics follow stan-
dard stack design procedures common to all com-
bustion engineering. The stack velocity given for
natural draft systems accords with good practice
and ininimizes flow losses in the stack.
The remainder of the essential calculations needed
for designing an incinerator are based upon substi-
tution in the parametric equations and measure-
ment of the incinerator. Recommended grate load-
ing, grate area, and average arch height may be
calculated by equation or estimated from Figures
310 and 311. Proper length-to-"width ratios may
be determined and compared with proposed values.
Supplementary computations are usually required
in determining necessary auxiliary gas burner sizes
and auxiliary fuel supply line piping. Where the
moisture content of the refuse is less than 10 per-
cent by weight, burners are usually not required.
Moisture contents of from 10 to 20 percent normal-
-------
422
INCINERATION
Table 118. CHEMICAL PROPERTIES AND COMBUSTION DATA FOR
PAPER, WOOD, AND GARBAGE
A
n
a
1
y
s
i
s
Material
Carbon (C)
Hydrogen (H)
Nitrogen (N)
Oxygen (O)
Ash
Gross Btu/lb
Dry basis
Constituent
(Based on 1 Ib)
Theoretical air
(40% sat at 60 °F)
Flue gas with
theoretical air
co2
N2
H2Of ormed
H2O (air)
Total
Flue gas with
% excess air
as indicated
0%
50. 0
100. 0
150. 0
200. 0
300. 0
Sulfite paper, a
44. 34
6.27
48.39
1.00
7, 590
scfe
67. 58
68. 05
13.99
53.40
11. 78
0.47
79.65
79.65
113.44
147.23
181.26
215.28
283. 33
Ib
5. 16
5. 18
1.62
3. 94
0. 56
0. 02
6. 15
6. 16
8. 74
11. 32
13. 91
16. 51
21. 70
Average wood,'3
49. 56
6. 11
0. 07
43. S3
0. 42
8, 517
scf
77. 30
77.84
15.64
61. 10
11.48
0. 53
88.77
88.77
127. 42
166. 07
204. 99
243.91
321. 75
Ib
5. 90
5.93
1. 81
4. 51
0. 54
0. 02
6. 90
6. 91
9. 86
12. 81
15. 78
18. 75
24. 68
Douglas fir, c
52. 30
6. 30
0. 10
40. 50
0. 80
9,050
scf
84. 16
84. 75
16. 51
66.53
11. 84
0. 58
95. 46
95. 47
137. 55
179. 63
222. 01
264. 38
349. 13
Ib
6.43
6. 46
1.9]
4. 9J
0. 56
0. 02
7.42
7. 43
10. 64
13.86
17. 09
20. 12
26. 58
Garbage, d
52. 78
6.27
39.95
1. 00
8, 820
scf
85. 12
85. 72
16. 66
67. 23
11. 88
0. 59
96. 37
96. 38
139.. 24
182. 00
224. 86
267. 72
353. 44
Ib
6.50
6.53
1. 93
4. 97
0. 56
0. 02
7. 49
7. 50
10. 77
14. 04
17.21
20. 58
27. 12
aConstituents of sulfite paper, %
C6H10°5
C5H10°5
C6H10°5
Cellulose
Hemicellulose
Lignin
Resin
Ash
bKent, 1936.
cKent, 1961.
dEstimated.
eMeasured at 60°F and 14.7 psia.
C20H30°2
84
8
6
2
1
ly necessitate installation of mixing chamber burn-
ers, and moisture contents of over 20 percent usu-
ally necessitate inclusion of ignition chamber burn-
ers.
General Construction
The design arid construction of multiple-chamber
incinerators are regulated in several ways. Ordi-
nances and statutes that set forth basic building re-
quirements have been established by most, if not
all, municipalities. Air pollution control authori-
ties have also set some limitations in material and
construction that must be met, and manufacturers'
associations have established recommended mini-
mum standards to be followed.
The building codes governing incinerator construe -
tion adopted in the past have been based primarily
upon concepts of structural safety and fire preven-
tion by restriction of the rate of heat transfer
through the walls. Little or no attention was given
to the abrasion, erosion, spalling, and slagging
that are encountered in a high-temperature incin-
erator, and yet these conditions lead to equipment
failures that are comparable to structural or in-
sulation failures.
-------
General-Refuse Incinerators
423
The structural features and materials used in the
construction of multiple-chamber incinerators can
be discussed only in general terms. There are
as many methods of erecting the "walls of amultiple-
chamber incinerator as there are materials from
which to build them. Designs of multiple-cham-
ber incinerators are presented schematically in
Figures 312 and 313. The types of construction
and fabrication shown are typical of those in cur-
rent usage. The designs are shown with prefired
refractory brick linings and common brick ex-
terior walls. Structural details are not indicated
since the reinforcing and support of walls, arches,
and stack depend largely upon the size and type of
construction of the unit under construction. While
conventional "60° sprung arches" are shown for
the main arches and curtain wall port openings,
flat suspended arches and other standard types of
sprung arches may be substituted satisfactorily.
Air inlets have been shown both as circular and as
rectangular ports. Either may be used to provide
adequate inlet areas. The exterior of the incin-
erator may be of either brick or steel plate con-
struction, and the refractory lining may be of fire-
brick, castable refractory, or plastic firebrick,
or combinations thereof.
In accordance with standard practices, the exteri-
or walls are protected further from extreme tem-
perature conditions by providing a suitable periph-
eral airspace in brick construction, by providing
air-cooling lanes, or by using insulation in units
fabricated from steel.
Changes in the methods of construction of multiple-
chamber incinerators are typified in the portable
prefabricated units available today. Installation
of incinerators such as these is reduced simply to
placement of the unit on its foundation and attach-
ment of an auxiliary fuel supply where needed,
though transportation considerations of weight and
size limit their capacity to 500 pounds or less per
hour. Plastic and castable refractory linings in
steel exteriors are used widely for this type of
fabrication. All larger incinerators, regardless
of the type of construction, and those incinerators
for which brick is desired for an exterior are
erected on the site.
Refractories
The most important element in construction of
multiple-chamber incinerators, other than the de-
sign, is the proper installation and use of refrac-
tories. High-quality materials are absolutely nec-
essary if a reasonable and satisfactory service life
is to be expected. Manufacturers must use suit-
able materials of construction and be experienced
in high-temperature furnace fabrication and ref rac -
tory installation, since faulty construction may
well off s et the benefits of good design. In the choice
of one of the many available materials, maximum
service conditions should dictate the type of lining
for any furnace. Minimum specifications of mate-
rials in normal refuse service should include high-
heat-duty firebrick or 120 pounds per cubic foot
castable ref ractory. These materials, when prop-
erly installed, have proved capable of resisting the
abrasion, spalling, slagging, and erosion result-
ing from high-temperature incineration.
As the incinerator's capacity and severity of duty
increase, superior refractory materials such as
super duty firebrick and plastic firebrick should
be employed. A recent improvement in standard
construction has been the lining of all stacks -with
2, 000°F refractory of 2-inch minimum thickness.
4
Grates and Hearths
The grates commonly used in multiple -chamber
incinerators are made of cast iron in "Tee" or
channel cross section. As the size of the incin-
erator increases, the length of the ignition cham-
ber also increases. In the larger hand-charged
incinerators, keeping the rear section of the grates
completely covered is difficult because of the great-
er length of the ignition chamber. The substitution
of a hearth at the rear of the ignition chamber in
these units has been accepted as good practice,
since a hearth in this region prevents open areas
frombeing formed in the normally thin refuse pile.
This prevents excessive underfire air from enter-
ing in front of the bridge wall, which would increase
fly ash carryover and reduce combustion efficiency.
Since surface combustion is the primary combus-
tion principle, the use of a hearth has little effect
upon combustion rate.
Installation of a sloping grate, which slants down
from the front to the rear of the ignition chamber,
facilitates charging. A grate such as this also in-
creases the distance from the arch to the grates
atthe rear of the chamber and reduces the possi-
bility of fly ash entrainment, -which frequently oc-
cur s when the fuel bed surface approaches the level
of the flame port.
Air Inlets
Positive control for all combustion air inlets should
be provided by means of fully adjustable dampers.
The retort incinerator designs shown in Figure 312
incorporate round, spinner-type controls with ro-
tating shutters for both underfire and overfire air
openings, and rectangular ports with sliding or
hinged dampers for the secondary air openings.
The in-line incinerator designs shown in Figure
313 have rectangular ports for both overfire and
secondary air openings, and spinner-style ports
for the underfire air openings. Air ports may be
of any convenient shape, though the port arrange-
ment indicated in the in-line designs with rectan-
-------
424
INCINERATION
I. STACK
2. SECONDARY AIR PORT
3. GAS BURNERS
4. ASH PIT CLEANOUT DOOR
5. GRATES
6. CHARGING DOOR
7. FLAME PORT
8. UNDERFIRE AIR PORT
9. IGNITION CHAMBER
10. OVERFIRE AIR PORT
1 I . MIXING CHAMBER
12. COMBUSTION CHAMBER
13. CLEANOUT DOOR
14. CURTAIN WALL PORT
PLAN VIEW
SIDE ELEVATION
END ELEVATION
-C
-Q
CTL
O
h—
«a:
ce
LLJ
•x.
C_3
U_
O
LU
rsi
CO
LENGTH, inches
ABCDEFGH'I JKLMNOPQRSTUVWXYZ
50
100
150
250
500
750
1 000
314
404
»5
54
764
854
944
13i
18
224
27
36
494
54
22i
28i
334
37i
47 i
54
594
9
134
154
18
27
36
36
63
9
114
134
IB
224
27
204
27
29
36
494
54
584
134
18
224
27
36
45
45
18
19
20
22
28
32
35
a
12
14
18
24
30
34
184
23
27
30
364
40
45
20
28
354
40
484
514
544
3J
5
5
74
124
15
174
10
15
164
18
23
28
30
44
24
44
44
9
9
9
21
24
24
44
44
44
44
24
4
44
44
44
44
44
9
144
18
20
26
25
274
24
5
5
5
5
5
74
24
0
24
24
5
10
124
24
24
24
24
2i
24
24
44
44
44
44
9
9
9
24
24
24
24
44
44
44
tt4
«4
»4
44
9
9
9
44
44
44
44
9
9
9
6
8
9
12
16
18
22
4
5
6
6
8
8
10
'Dimension " H " given in feet.
Figure 312. Design standards for multiple-chamber, retort incinerators.
-------
General-Refuse Incinerators
425
PLAN VIEW
SIDE ELEVATION
1 .
2.
3.
4.
5.
STACK
SECONDARY AIR PORTS
ASH PIT CLEANOUT DOORS
GRATES
CHARGING DOOR
6. FLAME PORT
7. IGNITION CHAMBER
8. OVERFIRE AIR PORTS
9. MIXING CHAMBER
10. COMBUSTION CHAMBER
1 1.
12.
13.
1 4.
15.
CLEANOUT DOORS
UNDERFIRE AIR PORTS
CURTAIN WALL PORT
DAMPER
GAS BURNERS
OC
CD
h-
«E
OC
U-l
C_3 _CL
-C3
U_
CD
LLJ
f^J
CO
LENGTH, inches
ABCDEFGHIJKL*MNOPQRSTUVWXY
750
1000
1500
2000
854
944
99
108
494
54
764
90
514
54
65
69i
45
<7i
55
57i
15i
18
18
224
54
63
72
794
27
314
36
404
27
314
36
404
*D
94
11
124
15
24
29
32
36
18
224
27
314
32
35
3B
40
44
44
44
44
5
5
5
5
74
10
74
10
9
9
9
9
24
24
44
44
24
24
44
44
30
30
30
30
9
9
9
9
44
44
44
44
5
7
8
9
11
12
14
15
51
52
614
634
7
8
9
10
mension "L" given in feet.
Figure 313. Design standards for multiple-chamber, in-line incinerators.
gular overfire ports is preferred since the com-
bustion air is distributed more evenly across the
fuel bed.
Stack
Stacks for incinerators with a capacity of 500 pounds
or less per hour are usually constructed of a steel
shell lined with refractorv and mounted over the
combustion chamber. A refractory-lined rein-
forced, red brick stack is an alternative meth-
od of construction when appearance is deemed
important. Stacks for incinerators with a capacity
of more than 500 pounds per hour are normally
constructed in the same manner as those for
smaller units but are often free standing for struc-
tural stability, as indicated in Figure 313. Stack
linings should be increased in thickness as the in-
cinerator becomes larger in size.
-------
426
INCINERATION
Induced-Draft System
The replacement of a stack by an induced-draft
system introduces additional problems. Cooling
the effluent gases becomes necessary to reduce
their temperature to that for which the draft fan
is rated. Evaporative cooling-with water is a stan-
dard practice. The contact of the flue gas with
water forms a "weak acid solution that eventually
corrodes the evaporative cooler and accessory
equipment, making replacement necessary. To
overcome these problems, stainless steel or acid-
resistant brickmay be installed. The excess spray
water also creates a problem, requiring a sewer
outlet for its disposal or a recirculation system
for its reuse. Recirculation of acidic water not
only results in more rapid corrosion of the spray
chamber and fan, but also subjects the pump, pip-
ing, and spray nozzles to corrosion. The use of
an induced-draft system with a spray chamber ac-
complishes additional removal of large particulate
matter and "water-soluble gases.
Operation
The most important single aspect of operation of a
multiple-chamber incinerator is the method of
charging the refuse into the ignition chamber. A
multiple-chamber incinerator must be charged
properly at all times in order to reduce the forma-
tion of fly ash and maintain adequate flame cover-
age of the burning rubbish pile and the flame port.
A recommended charging cycle starts with the
placing of the initial charge of refuse in the incin-
erator. The ignition chamber should be filled to a
depth approximately two-thirds to three-fourths of
the distance between the grates and the arch before
lightoff. After approximately half of the refuse
has been burned, the remaining refuse should be
carefully stoked and pushed as far as possible to
the rear of the ignition chamber . New refuse should
be charged over the front section of the grates,
which have been emptied by the moving of the burn-
ing refuse. To prevent smothering the iire, no
material should be charged on top of the burning
refuse at the rear of the chamber. With this charg-
ing method, live flames cover the rear half of the
chamber, fill the flame port, and provide nearly
complete flame coverage in the mixing chamber.
The fire propagates over the surface of the newly
chargedmaterial, spreading evenly ana minimiz-
ing the possibility of smoke emissions. Since the
refuse pile need not be disturbed unduly, little or
no fly ash is emitted,
Characteristic of the multiple- chamber incinerator
is that control of air-polluting emissions is built
in, if the incinerator is operated with reasonable
care. The discharge of combustion contaminants
is almost entirely a function of ignition chamber
design and the actions of the operator, Control of
smoke is attained by proper admission of combus-
tion air and by use of secondary burners in cases
of incineration of refuse with a low heating value
or high moisture content. The use of secondary
burners is required at times since the efficiency
of the mixing chamber depends upon both luminous
flame and adequate temperatures for vapor phase
combustion. The need for supplementary burners
maybe determined readily by observing the nature
of the flame travel and coverage at both the flame
port and the curtain wall port.
The overfire and underfire air ports are usually
half-open at lightoff and are opened gradually to a
full open position as the incinerator reaches its
rated burning capacity. If black smoke is emitted,
the admission of more secondary air and reduction
of the capacity of other air ports are advisible. On
the other hand, white smoke is usually the result
of a too cold furnace and may be eliminated by re-
ducing or closing all air ports. After the final
charge or refuse, the air ports are closed gradually
s o that during the burndown period the only air in-
troduced into the furnace is provided through leaks
around door and port openings.
When ignition andmixing chamber burners are nec-
essary, the mixing chamber or secondary burner
is lighted before the incinerator is placed in to oper-
tion. The burner should remain in operation for
the first 15 to 20 minutes of operation and should
be used thereafter as needed. Under normal con-
ditions, the ignition chamber or primary burner
is used only when wet refuse is charged. At other
times, its use, too, maybe required when refuse
tobe burned contains high percentages of inorganic
compounds such as clay fillers used in quality paper.
Illustrative Problem
Problem:
Design a multiple-chamber incinerator to burn
paper with 15 percent moisture at a rate of 100 Ib/hr.
Solution:
1. Composition of refuse:
Dry combustibles (100 Ib/hr)(0. 85; = 85 rh/hr
Moisture (100 lb/hr)(0. 15) =151b/hr
2. Gross heat of combustion:
From Table 118, the gross heating value of
dry paper is 7,590 Btu/lb.
(85 lb/hr)(7, 590 Btu/lb) - 645, 200 Btu/hr
-------
General-Refuse Incinerators
427
3. Heat losses:
From Table 118, 0. 56 Ib of water is formed
from the combustion of 1 pound of dry paper.
Radiation, etc = (0. 20)
(645, 200 Btu/hr)
Evaporation of contained
moisture (15 Ib/hr)
,"(1, 060 Btu/lb) >-
v - - -^ (_.- u.,-V
Evaporation of water
from combustion
(0.56 lb/lb)(85 Ib/hr)
(1, 060 Btu/lb)
Total
4. Net heat:
= 129,040 Btu/hr
15,900 Btu/hr
= 50,400 Btu/hr
= 195,340 Btu/hr
645, 200 Btu/hr - 195,340 Btu/hr = 449,860
Btu/hr
5. Weight of products of combustion with 300 per-
cent excess air:
From Table 118, 21. 7 pounds of products of
combustion result from the combustion of 1
pound of paper with 300 percent excess air.-
Paper (85 Ib/hr)(21. 7 Ib/lb) = 1,844 Ib/hr
Water 15 Ib/hr = 15 Ib/hr
Total
6. Average gas temperature:
1,859 Ib/hr
The specific heat of the products of combus-
tion is 0. 26 Btu/lb-°F.
449,860 Btu/hr
= 930°F
(0. 26 Btu/lb-°F)(l, 859 Ib/hr)
T= 930°F + 60°F_ = 990°F
7. Combustion air requirements:
Basis:
Use 300 percent excess air; 200 percent ex-
cess air is admitted through open charging
door and leakage around doors, ports, expansion
joints, etc.
From Table 118, 68. 05 cf of air is theoretically
necessary to burn 1 pound of dry paper. J-—
(85 Ib/hr) (6 8. 05 cf/lb)(2) = ll,580cfh
or 192. 8 cfm
or 3.2 cfs
1
8. Air port opening requirements at 0. 1 in. WC:
From Table D8 in Appendix D, 1, 255 fpm is
equivalent to a velocity pressure of 0. 1 inch.
:- ' A
Total = (192.8 cfm)(144 in2/ft2) = ^2
1, 255 ft/mm
Overfire airport (0. 7)(22. 2 in?) = 15. 6 in?
Underfire airport (0. 1)(22. 2 in?) = 2.2 in?
Secondary airport (0. 2)(22. 2 in?) = 4. 4 in?
9. Volume of products of^combustion:
From Table 118, 283. 33 cf of products of com-
bustion are formed from the combustion of 1
pound of paper with 300 percent excess air.
Basis:
60 °F and 300 percent excess air
Paper (85 lb/hr)(283. 33 cf/Ib) = 24, 080 cfh
Water (15 Ib/hr)
Total
379 ft /lb-mol
18 Ib/mol
316 cfh
24,396 cfh
or 6. 8 cfs
10. Volume of products__pf combustion through flame
port:
Total volume minus secondary air
6. 8 cfs - (3.2 cfs)(0.20) = 6. 16 cfs
11. Flame port area:
From Table 116, velocity is 55 fps. - \
(6. 16 cfs)(l,560°R)
(55 fps)(520°R)
= 0.34ft
12. Mixing chamber area:
From Table 116, velocity is 25 fps.
(6. 8 cfs)(l, 460°R)
(25 fps)(520°R)
= 0. 76 ft
13. Curtain wall port area:
From Table 116, velocity is 20 fps.
(6.8 cfs)(l, 410°R)
(20 fps)(520°R)
= 0. 92 ft
14. Combustion chamber area:
From Table 116, velocity is 6 to 10 fps.
-------
428
INCINERATION
(6. 8 cfs)(l, 360°R)
(6 fps)(520°R)
= 2. 96 it
15. Stack area:
From Table 116, velocity is < 30 fps.
= 0. 71 ft
(6. 8 cfs)(l, 360°R)
(25 fps)(520°R)
16. Grate area:
From Figure 3 10, the grate loading for aver
age refuse is 18 lb/ft^-hr.
(100 Ib/hr)
2
= 5. 56 ft
18 Ib/ft -hr
17. Arch height:
From Figure 311, the arch height = 27 in.
18. Stack height:
From Table 116, D ~ 0. 17 in. WC.
= 0.52 PH (- - — )*
wh ere:
D
P
H
T
Tl
H
H
= draft, in. WC
= barometric pressure, psi
= height of stack above grates, ft
- ambient temperature, "Rankine
= average stack temperature, "Rankine.
D
(0.52)(P)( . _}
1
0. 17
(0.52)(14.7)(
1
1
= 18. 7 5 ft
520
1, 360
MOBILE MULTIPLE-CHAMBER
INCINERATORS
Mobile multiple-chamber incinerators provide a
unique method for on-the-site disposal of com-
bustible refuse. Limited numbers of these units
were constructed in Los Angeles County in the late
1950's and used successfully for land clearance,
housing tract construction, and other industrial
"•Kent, 1938.
activities where the permanent installation of an
incinerator or the hauling of refuse to another loca-
tion for disposal would have been less economical.
Although their technical efficiency was adequate,
mobile multiple-chamber incinerators never
achieved a popularity of any consequence because
of availability of more economical disposal meth-
ods .
At first glance, one may presume that a standard
multiple-chamber incinerator mounted on a trail-
er can serve as a mobile incinerator. This pre-
sumption is quickly dispelled when weight and size
limitations, draft, vibration, and other problems
inherent in mobile construction are more closely
examined. The discussion that follows provides
a designer with practical and economical answers
that facilitate the design and construction of suc-
cessful mobile multiple-chamber incinerators.
DESIGN PROCEDURE
Although mobile incinerators are designed with
parameters identical to those of multiple-chamber
incinerators, already described in this chapter,
they must be constructed of lightweight materials
and limited in size to comply with the State Vehicle
Code. Design configurations generally restrict
the maximum capacity of the retort style, as shown
in Figure 314, to 500 pounds per hour, and that of
the in-line style, as shown in Figure 315, to 1,000
pounds per hour.
Draft for mobile incinerators may be produced in
two ways. The first and most conventional -way is
the use of a stack, while the other incorporates an
induced-draft system that uses air to cool the ef-
fluent.
Stack Requirements
If a stack is used, it must be retractable to meet
the height requirements of the State Vehicle Code.
To accomplish this, it is usually hinged at the base
and, if necessary, folds in the middle, permitting
it to lie horizontally on the top of the incinerator.
The stack is unlined to reduce not only its weight
but the size and weight of the elevating equipment,
which consists of a frame, steel cables, and pul-
leys, operatedby a hand crank or geared to a small
gasoline engine.
Induced-Draft Fan System
A typical induced-draft system consists of an un-
insulated breeching of 10-gage steel plate where
products of combustion from the incinerator are
cooled by mixing with air to a temperature that can
be safely handled by an induced-draft fan. Cooling
air is introduced through manually adjustable and
barometric dampers located in the breeching.
-------
Mobile Multiple-Chamber Incinerators
429
r~S
Figure 314. A 500-pound-per-hour mobile incinerator
with retractable stack.
Figure 315. A 1,000-pound-per-hour mobile incinerator
with retractable stack.
Heat and material balances, necessary for design-
ing the breeching, are computed by the methods
shown in the illustrative problem on page 431. The
breeching should be sized to give an average ve-
locity through its cross-sectional area of about
40fps. At this velocity, adequate mixing of cool-
ing air with the products of combustion occurs with-
in 0. 4 second, producing a relatively uniform tem-
perature without excessive frictional losses. De-
signing the breeching for low frictional losses per-
mits use of inexpensive axial-flow or propeller-type
fans.
Manually adjustable dampers allow for introduction
of dilution air into the breeching to cool the prod-
ucts of combustion. These dampers must be sized
to provide sufficient air at the maximum burning
rate of the incinerator to cool the gases to the de-
sign temperature of the fan. Barometric dampers
balance the induced-draft system by sustaining an
adequate and uniform draft in the incinerator. Their
use comes into play primarily at lightoff and burn-
down when the charging door is closed and the air-
ports are partially opened, restricting the gas flow
through the incinerator. Under these conditions,
the barometric dampers open more widely, allow-
ing additional air to be induced into the breeching.
This prevents an increased draft from developing
in the ignition chamber. During capacity opera-
tion, when the gas flow through the incinerator is
maximum, the balancing effect of barometric damp-
ers is not required.
A major problem in the design of the induced-draft
fan system is the proper selection of the fan. Fans
capable of operating from ambient temperatures
to temperatures in excess of 1,200°F are avail-
able. Fans designed to operate in excess of 800°F
must be constructed of stainless steel and should
be equipped with water-cooled bearings. These
fans are costly; their bearing-cooling requirements
virtually eliminate them from use on portable equip-
ment.
Low-temperature fans with mild steel blades are
capable of operating up to 300 °F. The maximum
operating temperature of these fans can be in-
creased to 800 °F by the addition of simple and in-
expensive bearing coolers commonly called heat
slingers. As the temperature increases above
300 °F, the maximum permissible rpm is reduced
for any class or duty of a specific fan. This capac-
ity reduction ranges from about 1 0 percent at 600 °F
to 30 percent at800°F. Therefore, it is necessary
to install a larger fan of the same class or the
same size fan of higher class when operating tem-
peratures exceed 600°F.
If dilution air is introduced in excess of that neces -
saryto cool the effluent to a temperature that can
be handled by an inexpensive fan, this excess air
will require an increase, not only in fan size, but
also in horsepower and operating cost. All factors
considered, the apparently optimum operating tem-
perature of the fan is 600 °F. At this temperature
-------
430
INCINERATION
an inexpensive and minimum sized fan constructed
of mild steel can be used. Since propeller-type
fans have the advantages of compactness, low cost,
and light weight, they are usually selected over
centrifugal types. Bronze blades available at a
nominal increase in cost over steel blades and
capable of operation up to 800°F are usually in-
stalled to provide a safety factor.
The induced-draft fan is powered by a small gaso-
line engine through a chain or belt drive. The en-
gine is sized for maximum power requirements,
which occur at lightoff when the air handled by the
fan is at ambient temperature. As the tempera-
ture of the exhaust gases rises, the fan horsepower
at a constant rpm decreases in proportion to the
change in density of the gases. The draft of the
incinerator can also be regulated by changing the
speed of the gasoline engine driving the fan.
STANDARDS OF CONSTRUCTION
The mechanical design and construction of a mobile
incinerator mustnot only meet the dimensional and
•weight requirements of the Vehicle Code but also
provide a rigid frame and refractories of sufficient
quality to provide a satisfactory service life.
Refractories
Since refractories constitute 60 to 75 percent of
the total weight of a mobile incinerator, low-den-
sity refractory materials must be selected. These
materials should have a minimum pyrometric cone
equivalent (PCE) of 15 and be relatively resistant
toabrasion, spalling, and physical shock. Thermal
conductivity should be about 5.4 at 2,000°F by
ASTM C-201 so that backing -with insulation will
not be necessary.
Because shaped firebricks are not suitable ior mo-
bile installations because of excessive weight and
problems in anchoring firmly to walls, other re-
fractories must be investigated. A number of
standard castable refractories manufactured today
meetthese specifications. They are composed of
approximately equal portions of alumina and silica,
are easy to cast, and have a density of about 80
pounds per cubic foot.
Exterior'walls and arches are secured against thin
corrugated steel sheets with stainless steel anchors
arranged on 12-to 15-inch center s, while interior
walls are self-supporting. Walls and arches are
usually 4-1/2 inches thick for incinerators with
capacities of less than 600 pounds per hour and 6
inches thick for units of larger capacity. It is of
the utmost importance that castable refractories
be installed strictly in accord •with the information
and directions provided by the manufacturer.
The bridge-wall is susceptible to damage by care-
less operation, and the curtain wall is subjected to
high-temperature flame impingement accompanied
by high velocities, which tend to erode its surface.
At these locations, the use of heavier castable re-
fractories, which are more resistant to abrasion
and erosion, is advantageous . A number of mate-
rials with densities of about 120 pounds per cubic
foot have the special qualities to fill this need.
Grates
Cast iron grates, available today in many sizes,
shapes, and patterns, are satisfactory for burn-
ing general refuse, as described previously in this
chapter. Castable refractory grates, described
later in this chapter, should be installed in incin-
erators designed to burn large quantities of wood.
Air Inlets
Combustion air may be controlled by providing ad-
justable dampers in the throats of all air ports.
Dampers used for controlling overfire and under-
fire air are subject to warpage from high tem-
peratures and should be constructed of stainless
steel or cast iron. The secondary air port damp-
er is not subjected to much heat and may be con-
structed of 10-gage mild steel plate.
Structure
The trailer and frame for supporting the incin-
erator should be designed by qualified structural
engineers. A trailer of welded steel construction
must be rigid enough to prevent the transmission
of stresses and strains to the refractory walls dur-
ing travel over rough terrain. The external frame
should also be engineered to cope not only with
mechanical stresses imposed during transporta-
tion but also with thermal stresses produced dur-
ing the operation of the incinerator.
Auxiliary Burners
Mobile incinerators usually burn refuse varying
widely in composition, requiring auxiliary burn-
ers sized in accordance with the information pre-
sented in Table 116. These burners are fired with
LPG supplied from tanks mounted upon the incin-
erator trailer.
STACK EMISSIONS
The quality and composition of emissions from
mobile multiple-chamber incinerators are similar
to those from stationary multiple-chamber incin-
erator s in burning general refuse. The air pollu-
tants in pounds per ton of refuse burned are given
in Table 117.
-------
Mobile Multiple-Chamber Incinerators
431
Illustrative Problem
Problem:
Design an induced-draft fan system for a mobile
multiple-chamber incinerator.
3. Flow.of dilution air at 60°F:
(11, 100 lb/hr)l
'379 ft'/lb mole
29 Ib/lb mole
= I, 420 cfm or 40. 3 cfs
V 1 hr \
/\60 min/
Given:
Refuse to be burned is 1, 000 pounds of wood per
hour with 20 percent by weight moisture.
Solution:
1. Weight of products of combustion with 300 per-
cent excess air:
From Table 118, there are 24.68 Ib of com-
bustion products from the combustion of 1 Ib
of average dry 'wood with 300 percent excess
air, 40 percent saturated.
Wood (800 lb/hr)(24.
Moisture
Total
Ib/lb) = 19,750 Ib/hr
200 Ib/hr
19, 950 Ib/hr
2. Weight of dilution air required to reduce prod-
ucts of combustion from 900° to 600 °F:
Assume combustion products are equivalent
to air in composition. Average specific heat
of air is 0. 26 Btu/lb-°F.
(w )(c )(t -t ) = (w )(c )(t -t )
a p2 2 a pc pi 1 2
•where:
pc
"P2
4. Gas flow through breeching:
From Table 118, there are 321.7 ft3 of com-
bustion products at 60 °F from combustion of
1 Ib of average dry wood with 300 percent ex-
cess air.
Wood (800 lb/hr)(321.7 ft3/lb) = 257, 000 cfh
^379 ft3/lb
Moisture (200 Ib/hr)!
Total
18 Ib/lb mole
L\
4, 200 cfh
261,200 cfh
Dilution air at 60 °F (2, 420 cfm)(6° min)
V 1 hr /
Total
= 145, OOP cfh
406, 200 cfh
or 6, 770 cfm
or 113 cfs
Total gas flow at 600°F
(1,^n°iy (6' 77° cfm) = 13' °°° cfm
(52° R) or 230 cfs
5. Cross section of breeching:
Design breeching for an average gas flow rate
of 40 fps at 600 °F:
Area =
= weight of dilution air, Ib/hr
= "weight of combustion products; Ib/hr
= average specific heat of products of
combustion, Btu/lb-°F
= average specific heat of air, Btu/lb-°F
= final temperature, °F
= initial temperature of combustion prod-
ucts, °F
= air temperature, °F
(19, 950 lb/hr)(0. 26 Btu/lb-°F)(900°F - 600°F)
(0.26 Btu/lb-°F)(600°F - 60°F)
= 11,100 Ib/hr
(230 cfs)
(40 fps)
5. 75 ft"
Dimensions: 18 in. high x 46 in. wide.
6. Length of breeching:
Design breeching for a residence time of 0.45
sec at 40 fps
Length = (0.45 sec)(40 fps) = 18 ft
Use a double-pass breeching 9 ft long to fit on
top of the incinerator.
7. Static pressure behind adjustable dampers and
barometric dampers at capacity operation:
Assume static pressure behind the adjustable
dampers and barometric dampers is essen-
tially the same.
-------
432
INCINERATION
( a) Assume static pressure in combustion cham-
ber, SP = 0. 30 in. WC
(b) Contraction loss from combustion chamber
into duct leading to breeching:
Ratio r -
cross-sectional area of duct
horizontal cross-sectional area
combustion chamber
= o.33
17.5 ft
Contraction loss is 0. 38 VP* (velocity pres -
sure head) at the velocity through the duct.
Velocity through 5. 75 ft2 port at 900 °F
(261, ZOO cfh) (1,360"R) (1)
(3,600 scfs/hr) (520°R (5. 75 ft2 pS
Assume composition of combustion prod-
ucts is equivalent to air.
Velocity head of 32. 9 fps at 900°F
v = 2. 9Vth '
"where:
v = gas velocity, fps
t = absolute gas temperature, °R
h = velocity pressure (head), in. WC
h =
--
2.9/ (t)
/32.9fps\2 / 1 \
\ 2.9 / \1,360°R/
h = 0. 090 in. WC
Contraction loss
(0. 38 VP)
(0. 090 in. WC)
1 VP
= 0. 04 in. WC
(c) Right-angle bend into breeching.
Assume 1 VP loss for right-angle bend.
1 VP at 32. 9 fps and 900°F = 0. 09 in. WC
(d) Total static pressure
a + b + c = total static pressure
(0.30 in. WC) + (0.04 in. WC) +
(0. 09 in. WC) = 0. 43 in. WC
8. Indraft velocity through dampers:
Design breeching for a gas velocity of 40 fps
at 600°F. At a velocity pressure of 40 fps
and 600°F,
h =
iii
(t)
= °- 18 in- wc
h =
Total pressure = velocity pressure + static
pressure.
Total pressure = 0. 18 in. WC + 0. 43 in. WC
= 0. 61 in. WC.
Assume static friction loss through dampers
is 0. 65 VP.
Total pressure = velocity pressure + static
pressure.
0.61 in. WC = 1 VP + 0.65 VP
VP = 0. 37 in. WC
From Table D8, Appendix D, the velocity
at 60°F and 0. 37 in. WC is 2, 410 fpm.
9. Size of adjustable dampers (assume barometric
dampers closed):
Design dampers 100% oversize to allow for
operation of the incinerator in excess of de-
sign capacity.
Dilution air = 2,420 cfm
cfm) (2) = 2.03ft2
(2,410 fpm)
Badger and McCase, 1936.
10. Static-pressure drop through induced-draft
system at capacity operation with a 600 °F out-
let temperature:
tResearch-Cottrell, Inc.
(a) Static pressure at dampers, SP = 0.43 in. W
-------
Mobile Multiple-Chamber Incinerators
433
(b) Double pass breeching 18 feet long:
2*
f =
0.002 hv
mt
where:
f = friction, in. WC
h = duct length, ft
v = gas velocity, fps
t = absolute gas temperature, °R
m = hydraulic radius
11. Calculate points on system static-pressure
curve based upon capacity operation at 600 °F:
SP2 = (sPj
where:
SP2
cfm
2
cfm =
SP,
unknown static pressure
proposed cfm
known cfm
known static pressure
cross-sectional area of breeching, ft
perimeter of breeching, ft
= (.002)(18) (40)2 = wc
(0.54)(1, 060)
(c) 180° bend at one end of breeching.
Assume 2-VP loss at 40 fps and 600°F
h =
(2 VP)
(0. 18 in. WC)
(1 VP)
= 0. 36 in. WC
(d) 90° bend at fan discharge.
Use 9 ft opening to reduce pressure drop.
Assume 1-VP loss for 90° bend at 600°F.
Velocity = (40 fps)^7 ] = 25. 6 fps
h =
h =
, 060/
= 0.07 in. WC
'^(°-07(invpTC) - 0.07 in. WC
(e) Total static pressure for system:
(a) + (b) + (c) + (d) = total static pressure
(0. 43) + (0. 10) + (0. 36) + (0. 07) = 0. 96 in. WC
«Griswold,
Assume cfm = 10,000
sp2 = (0.
Assume cfm = 20,000
= (0.
). 57 in. WC
- Z. 28 in. WC
12. Fan specifications:
Select fan that will deliver, as near as possi-
ble, 13, 000 cfm at 0. 96 in. WC and 600°F.
Fan performance given for 60 °F operation:
1, 160 rpm 60°F
1.4 in. WC
21, 000 cfm
1 0 bhp
2.0 in. WC
13, 800 cfm
7 . 5 bhp
2. 2 in. WC
10, 000 cfm
6 . 7 bhp
Calculate points for 600 °F fan performance
curve: With rpm and cfm held constant, static
pressure and bhp vary directly with gas density
or inversely with absolute temperature.
Ratio =
520°F
1,060°F
= 0.49
1, 160 rpm 600°F
0. 7 in. WC 1.0 in. WC 1.1 in. WC
21, 000 cfm 13, 800 cfm 10, 000 cfm
4.9 bhp 3.7 bhp 3. 3 bhp
-------
434
INCINERATION
13. Operating point at 600°F:
Intersection of 600 °F system curve with 600 °F
fan curve is shown in Figure 316.
13, 400 cfm
1. 02 in. WC at 600°F
1, 160 rpm
3.7 bhp
14. Static-pressure drop.for induced-draft system
at 60°F:
This condition occurs at lightoff before igni-
tion. Assume negligible airflow through in-
cinerator and static pressure 0. 3 in. WC in
combustion chamber and behind barometric
and adjustable dampers.
Assume total airflow through fan is 13,000 cfm.
(a) Behind dampers static pressure = 0. 30 in. WC.
(b) Friction through 18-foot-long breeching:
Cross-sectional area = 5.75 ft
2
f =
0.002 hv
mt
f =
(0.002)(18)(40)
(0.54)(520) ~ "•""*
(c) 180° bend at end of breeching:
Assume 2-VP loss at 40 fps and 60°F
From Table D8, Appendix D, VP =
0. 36 in. WC
25
5,000
10,000 15,000
VOLUME, cfm
20,000
25,1
Figure 316. Fan and system curves at 60°F and 600°F.
-------
Mobile Multiple-Chamber Incinerators
435
(d) 90° bend fan discharge through 9 ft outlet:
Assume 1 VP at 26.5 fps at 60°F
From Table D8, Appendix D, 1 VP =
0. 16 in. WC
(1 VP) ^~
(e) Total static pressure:
a + b + c + d = 1. 39 in. WC
15. System static-pressure curve development at
60"F:
cfm, \ ^
= (sp
Assume cfm = 10,000
=P2 = '-"(irS)2 - »•
82 in. WC
Assume cfm = 16,000
sp2 = (1.
s,ooo\z
ooo/
= 2. 10 in. WC
16. Operating point at lightoff where the 60 °F sys-
tem curve intersects the 60°F fan curve (see
Figure 316):
15, 200 cfm
1. 90 in. WC at 60°F
1, 160 rpm
8. 0 bhp
Select a 1 0-hp gasoline engine to drive the fan.
17. Total system pressure behind dampers:
Assume negligible airflow through incinerator
at lightoff.
Static pressure behind adjustable and baromet-
ric dampers at 60°F:
= 0. 41 in. WC
Total air velocity at 60°F in breeching:
15, 200 cfm -, , .„ ,
— — = 2, 640 fpm
5.75 ft
From Table D8, Appendix D, VP - 0.41
in. WC
Total pressure = velocity pressure + static
pressure
Total pressure = 0.44 in. WC + 0. 41 in. WC
Total pressure = 0.85 in. WC
18. Air velocity through adjustable and baromet-
ric dampers:
Assume friction loss through dampers at 0.65
VP inlet.
Total pressure = velocity pressure + static
pressure
0.85 in. WC = 1 VP + 0.65 VP
VP = 0. 52 in. WC
From Table D8, Appendix D, inlet velocity
is 2, 860 fpm.
19. Airflow through adjustable dampers:
(2. 03 ft2)(2, 860 fpm) = 5, 800 cfm
20. Airflow through barometric dampers:
Assume negligible airflow through incinerator
Total airflow through fan 15, 200 cfm
Adjustable dampers 5, SOO cfm.
Barometric dampers 9, 400 cfm
21. Selection of barometric dampers:
Minimum damper area.
(9, 400 cfm)
Area =
(2, 860 fpm)
= 3.28 ft
Select four 15-in. -diameter barometric damp-
ers with total area about 40% in excess of min-
imum area to allow for operating flexibility.
Total open area of 4 dampers:
4 dampers
7-= -
(damper)
= 4.9ft
-------
436
INCINERATION
MULTIPLE-CHAMBER INCINERATORS FOR
BURNING WOOD WASTE
INTRODUCTION
Although a small part of the -wood waste produced
from lumber mills and wood-working industries
can be processed into useful products such as chip
board, fireplace logs, and paper, the bulk of this
waste is disposed of by incineration, open burning,
or hauling to a dump. The most satisfactory air
pollution solution is, of course, landfill disposal.
The final choice of the method of disposal is pri-
marily determined by economics and by the air
pollution regulations existing in the locale.
There are, in general, three methods of burning
wood waste. These are (1) open burning, that is,
burning in a pile without any surrounding structure;
(2) burning in single-chamber incinerators, includ-
ing the tepee and silo structures; and (3) burning
in multiple-chamber incinerators. Of these, the
latter is the most satisfactory from an air pollu-
tion standpoint.
Open burning with no control over combustion air
produces more air contaminants than single-cham-
ber incinerators do with regulated air supply. The
tepee and silo single-chamber incinerators also
differ in combustion efficiency and emission of air
contaminants.
Tepee incinerators are simple structures consist-
ing usually of nothing more than a sheet metal shell
supportedby structural steel members in a shape
similar to that of an Indian tepee. They are usu-
ally located at lumber mills and have limited con-
trol of primary combustion air. Many units em-
ploy blowers to supply air to the base of the burn-
ing pile to increase the burning rate. The metal
shell is cooled by peripheral air, which flows up-
ward and over the inside surfaces. Excessive
combustion air admitted in this manner prevents
good control of the combustion process and results
in excessive smoke and other air contaminants.
A silo incinerator consists of a steel cylindrical
chamber lined with high-duty refractory materials.
The top of the cylindrical chamber usually tapers
to a smaller diameter and extends upward, form-
ing a stack to promote draft. Air is admitted
through louverslocated near the base of the struc-
ture. High temperatures can be maintained in the
refractory-lined chamber, resulting in higher com-
bustion efficiencies than in the tepee units. Single-
chamber silo incinerators are not, however, satis-
factory where air pollution is a serious problem,
and have been found to emit particulate matter in
excess of 12 pounds per ton of -wood waste burned.
Description of the Refuse
Wood waste is produced by industry in a great many
sizes and shapes ranging from fine sander dust to
large pieces of lumber. Physical properties and
combustion data for several common woods are
given in Table 118. Green lumber at the mill varies
widely in moisture content. For example, green
redwood may contain over 50 percent moisture by
"weight, while construction-grade lumber such as
Douglas fir contains from 10 to 25 percent moisture
depending upon its age. Kiln-dried wood may con-
tain as little as 5 or 6 percent moisture.
THE AIR POLLUTION PROBLEM
Burning of wood waste in open areas and at dump
sites or in single-chamber incinerators is accom-
panied by dense clouds of smoke, fly ash, and dis-
agreeable odors. Basically, these air contaminants
are caused by incomplete combustion and are dis-
charged in the form of particulate matter, alde-
hydes, hydrocarbons and organic acids,as -well as
smoke and fly ash. They are usually present in
the greatest concentrations after the lightoff peri-
od or during times of heavy charging.
While single-chamber silo incinerators have been
found to have particulate emissions in excess of
12 pounds per ton of wood waste, the particulate
discharge from multiple-chamber incinerators de-
signed to burn small wood particles ranges from
1-1/2 to 6-1/2 pounds per ton of wood waste burned,
as shown in Table 119. Smoke is visible from a
well-designed multiple-chamber incinerator only
for a few minutes after lightoff and is occasionally
accompanied by minute amounts of fly ash.
AIR POLLUTION CONTROL EQUIPMENT
Air pollution from the burning of wood waste can
be reduced to a minimum through the use of multi-
ple-chamber incinerators. By promoting com-
plete combustion, multiple-chamber incinerators
produce considerably less air pollution than is
emitted from single-chamber incinerators or by
openburning. Multiple-chamber incinerators dis-
cussed in the remainder of this part of the chapter
are designed to burn all forms of wood waste--
from large pieces of lumber to sawdust particles
that may comprise from 10 to 100 percent of the
total weight of the charge. The designs of me-
chanical feed systems are also included since the
feed system must be properly integrated with the
design of the incinerator to pr omote maximum com-
bustion.
DESIGN PROCEDURE
The fundamental principles of combustion discussed
in the first part of this chapter are applicable to
designing these incinerators. Where 10 percent
-------
Multiple-Chamber Incinerators for Burning Wood Waste
437
Table 119. SOURCE TEST DATA FOR MULTIPLE-CHAMBER
INCINERATORS BURNING WOOD
Item
Incinerator capacity
Normal burning rate
Moisture content of refuse
Stack volume
Secondary chamber temperature
Particulate matter
Particulate matter
Sulfur dioxide
Carbon monoxide
Organic acid--as acetic
Aldehydes --as formaldehyde
Hydrocarbons — as hexane
Units
Ib/hr
lb/hra
wt %
scfm
°F
gr/scf
at 12%
CO2
lb/tonb
lb/tonb
lb/tonb
lb/tonb
lb/tonb
ppm
Test No.
1
150
170
10
420
1, 600
0. 058
2.0
0
0
0.8
2.0
9
2
350
300
5
557
1,400
0. 038
1.4
0
0
1.2
1.9
9
3
750
740
10
3,260
1, 500
0. 095
3. 2
0
0
0. 54
0.8
9
4
1, 000
1, 055
25
3, 300
1,850
0. 23
6.6
0
0
0. 85
3.0
9
5
3,000
2, 910
10
15, 300
1,600
0. 11
3.6
0
0
1.2
6.0
9
Burning rate based on stack analysis.
bPounds of contaminants per ton of wood burned.
or more of the wood waste is in the form of saw-
dust and shavings, it must be fed at a continuous
rate by a mechanical feed system. Differences in
some design factors from those given at the be-
ginning of this chapter for hand-charged general-
refuse incinerators generally reflect the higher
temperatures developed from the exclusive and
continuous mechanical charging of wood, and dif-
ferences in the distribution of combustion air.
The gross heating value of kiln-dried wood is 9., 000
Btu per pound and is represented by the upper
curves of Figures 310 and 311. These curves can
be used to determine grate loading and average
arch height, respectively. Other design factors
differing from those for general-refuse incinera-
tors are given in Table 120. These design factors
include secondary chamber cross-sectional areas,
inlet air port sizes, and other values and propor-
tions.
An illustrative problem at the end of this part of
the chapter shows how these factors are used to
designa multiple-chamber incinerator with a me-
chanical feed system. The calculations in this
problem fall into three general categories : (1) Com-
bustion calculations based upon refuse composi-
tion, projected air requirements, and heat trans-
fer; (2) gaseous flow calculations based upon the
products of combustion at elevated temperatures;
and (3) dimensional calculations based upon equa-
tions determined empirically from source testing.
Chemical properties and combustion data for av-
erage wood and Douglas fir, given in Table 118,
and similar values for other kinds of wood can be
used to determine the weights, velocities, and av-
erage temperatures of the products of combustion
For calculation purposes, the burning rate and wood
waste composition are assumed constant, and the
incinerator is considered to be under relatively
steady-state conditions. Calculations are always
based upon refuse that is the most difficult to burn.
Heat losses by radiation, heat stored in refractory,
and heat content of the residue are assumed to av-
erage 20 to 30 percent of the gross heating value
of the refuse during the first hour of operation.
These heat losses drop to 10 to 15 percent after
4 or 5 hours of operation.
To determine the cross-sectional flow areas of the
secondary ports and chambers, only volumes and
temperature levels of the products of combustion
are required. The temperature gradient in which
the products of combustion cool as they pass from
the flame port to the stack are averages based upon
source tests of similar incinerators.
The calculated overall average gas temperature
shouldbe about 1, 300°F based on 200 percent ex-
cess combustion air and the 20 to 30 percent heat
losses. The calculated temperatures are not flame
temperatures and do not indicate temperatures at-
tained in the flame port or mixing chamber.
Indraft velocities through the ignition chamber air
ports are assumed to average 900 fpm, equivalent
to a velocity pressure of 0. 05 inch WC, while in-
draft velocities through the secondary air ports av-
-------
438
INCINERATION
Table 120. DESIGN FACTORS FOR MULTIPLE-CHAMBER
INCINERATORS FOR BURNING WOOD WASTE
Item and symbol
Recommended value and units
Allowable
deviation
Primary combustion zone.
Grate loading, LQ
Grate area, AQ
Average arch height, HA
Length-to-width ratio (approx):
Retort
10 Log Rc; lb/hr-ft2 where Rc equals
the refuse combustion rate in Ib/hr
(refer to Figure 310)
Rc - LG; ft2
4/3 (AG)4/11; ft (refer to Figure 311
and + 10% curve)
Up to 500 Ib/hr, 2-1; over 500 Ib/hr,
1. 75.1
Diminishing from about 1, 7:1 for
7SO-lb/hr to about 1:2 for 2, 000-
I Ib/hr capacity. Oversqaure ac-
| ceptable in units of more than 11-ft
I ignition chamber length
10%
+ 10%
Secondary combustion zone.
Gas velocities
Flame port at 1,900°F, Vpp
Mixing chamber at 1, 550C'F,
Curtain wall port at 1, 500°F,
VCWP
Combustion chamber at 1,200°F,
vcc
Mixing chamber downpass length,
^MC' ^ronl toP °f ignitl°n chamber
arch to top of curtain wall port
Length-to-width ratios of flow
cross sections-
Retort, mixing chamber, and
combustion chamber
In-line
50 ft/sec
15 ft/sec
20 ft/sec
5 to 10 ft/sec; always less than
10 ft/sec
Average arch height, ft
Range: 1. 3:1 to 1. 5:1
Fixed by gas velocities due to
constant incinerator width
+ 20%
+_ 20%
+ 20%
20%
Combustion a^r •
Air requirement, batch, or con-
tinuous charging
Combustion air distribution, % of
total air required-
Overfire air ports
Underfire air ports
Mixing chamber air ports
Curtain wall port or side ports
Port sizing, nominal inlet,
velocity pressure, and velocity
(•without oversize factors), in. WC
or fpm:
Overfire port
Underfire port
Mixing chamber port
Curtain wall port or side port
Basis 200% excess air. 100%
excess air admitted into ignition
chamber, 50% theoretical air
through mixing chamber air ports
and 50T,i theoretical air through
curtain wall air port or side
air ports.
60%
6%
17%
17%
0. 051 or 900
0.051 or 900
0.062 or 1, 000
0.062 or 1, 000
Furnace tempera_tur_e
Average temperature, combus-
tion products at 200% excess air
1, 300°F
+ 20°F
Auxiliary burners:
Secondary burner {if required)
2, 500 to 5, 000 Btu per Ib of
moisture in the refuse
Draft requirements-
Theoretical stack draft, DT
Available primary air induction
draft, D. (assume equivalent to
inlet velocity pressure)
Natural draft stack velocity, Vg
0. 15 to 0. 35 in. WC
0.05 to 0. 10 in.WC
Less than 25 ft/sec at 1, 100°F
-------
Multiple-Chamber Incinerators for Burning Wood Waste
439
erage 1, 000 fpm (0. 06 in. WC). The incinerator
draft system should be designed to produce a nega-
tive static pressure of at least 0.05 inch WC in the
ignition chamber.
Primary air ports for continuously fed incinerators
are sized for induction of theoretical plus 100 per-
cent excess air. Ten percent of this air is admitted
through ports located below the grates, and 90 per-
cent, above the grates. Additional primary air
canbe admitted by opening the charging door when
necessary. Air is induced into the mixing cham-
ber not only to support combustion but also to cool
the combustion gases and prolong the service life
of the refractories . Mixing chamber air ports lo-
cated in the bridge wall are sized to admit 50 to
100 percent of theoretical air. Air is sometimes
admitted to the combustion chamber through air
ports located in the curtain wall and sized to ad-
mit an additional 50 percent of theoretical air.
Although some combustion air enters the ignition
chamber along with the sawdust from the pneumatic
conveying system, this air usually amounts to less
than 7 percent of the total combustion air and can
be neglected in determining the size of the primary
airports. Airports are designed with the factors
given in Table 120.
Unless the wood refuse is extremely wet, auxiliary
gas burner s are not required in the ignition cham-
ber to initiate and sustain combustion. If products
such as rubber, oily rags, and plastics are present
inappreciable quantities in the -wood -wastes, they
produce partially oxidized compounds that require
high temperatures for complete secondary com-
bustion. Thus, secondary burners should be in-
stalled in the mixing chamber with automatic con-
trols to maintain the required high temperatures
under all phases of operation.
Incinerator Arrangements
Incinerators for burning -wood use both in-line and
retort styles as shown in Figures 317 and 318. In-
cinerators-with capacities of less than 500 pounds
per hour are usually constructed as retorts. Units
ranging from 500 to 1,000 pounds per hour may,
however, follow either the in-line or retort style
for the arrangement of chambers. In-line styles
are recommended for incinerators with capacities
in excess of 1, 000 pounds per hour because of not
only the inherent higher costs of the retort but also
the difficulties in cooling the internal walls. A
retort-type incinerator with a prefabricated steel
shell is shown in Figure 318. A single-chamber,
silo-type incinerator can be converted to multiple
chamber by attaching a dutch oven consisting of an
ignition chamber and a mixing chamber as depicted
in Figures 319 and 320.
Figure 317. A 2,000-lb-per-hour, in-line multiple-chamber
incinerator (Metro Goldwyn Mayer, Inc., Culver City, Calif.).
Figure 318. A 150-lb-per-hour, retort multiple-chamber
incinerator (Acme Woodcraft, Los Angeles, Calif.).
-------
440
INCINERATION
Figure 319. A 1,000-Ib-per-hour, in-line
multiple-chamber incinerator (silo con-
version) (O'rban Lumber Co., Pasadena
Cal if.).
CHARGING
DOOR WITH
OVERFIRE
AIR PORT
SHO<
•ASH PIT-
CLEANOUT
DOORS
CLEANOUT DOORS-
Figure 320. Schematic diagram of an in-line multiple-chamber
incinerator (silo conversion).
-------
Multiple-Chamber Incinerators for Burning Wood Waste
441
In the design of the silo conversion, the size of the
ignition chamber and mixing chamber attached to
an existing silo is limited by a maximum allowable
gas velocity of 10 fps through the horizontal cross -
sectionalbase of the silo, or by the effective draft
developed by the stack. Effective draft, in turn,
is limited by the height of the silo and its internal
dimensions.
If the attachment of an ignition and mixing cham-
ber to a silo results in a gas velocity through the
base of the silo of less than 5 fps, a refractory
tunnel •with a cross -sectional area equal to the cur-
tain wall port area should extend from the curtain
wallhalfway across the base of the silo. The tunnel
acts as an extension of the mixing chamber and
provides additional flame residence time and turbu-
lence necessary to complete the combustion pro-
cess.
DESIGN PROCEDURE FOR MECHANICAL FEED SYSTEMS
During the development of the multiple-chamber
incinerator, hand charging of sawdust and inter-
mittent delivery of sawdust from local exhaust sys -
terns serving woodworking equipment were found
to smother the flames periodically in the ignition
chamber and thus cause excessive smoke. To
overcome this problem, a feed system was devel-
oped for delivering small wood particles to the ig-
nition chamber at a constant rate and thus sustain
continuous burning over the entire surface of the
pile. This system, illustrated in Figure 3Z1, con-
sists basically of a surge bin for holding sawdust
and wood chips from local exhaust systems serv-
ing woodworking equipment. Screw or drag con-
veyors in the bottom of the surge bin move the wood
waste at a uniform rate to the pickup point of a
pneumatic conveying system. The pneumatic con-
veyor transfers the waste to a cyclone where the
waste drops into the ignition chamber.
Surge Bin
Bins usually fabricated of sheet metal are designed
in such a way as to augment gravity flow of saw-
dust and wood chips to the conveyor at the bottom
of the bin. Waste is produced in a wide variety of
sizes and shapes, ranging from fine sander dust
to large chips fromhoggers . Gravity flow of mate-
rial is a complex function of the composition, size,
shape, density, packing pressure, adhesive quali-
ties, and moisture content. For example, pine
wood shavings do not flow as easily as hardwood
shavings of identical size and shape do because the
resin content of the pine wood imparts an adhesive-
ness hindering the flow. The flow characteristics
of a particular -wood waste are, therefore, of pri-
mary importance in the final selection of the shape
of the bin.
WOODWORKING
EX'HAUST SVST.EM
CYCL(M*E
•Figure 321. Diagram of a mechanical feed system.
Wood wastes that exhibit poor flow characteristics
shouldbe handled in bins constructed with vertical
sides and screw or drag conveyors covering the
entire flat bottom of the bin, as shown in Figure
322. If the wood waste has fairly free flow charac-
teristics, a bin with four vertical sides and a slop-
ing bottom may be used, as shown in Figure 323.
The included angle between the vertical side and
sloping bottom should not exceed 45 degrees. Wood
waste that exhibits ideal flow characteristics may
use a vee-bottom bin, as depicted in Figure 324.
The included angle between sloping sections should
not exceed 60 degrees for most efficient operation.
Although good bin design assists the flow of saw-
dust to the conveyors, bins with sloping bottoms
require mechanical agitators or vibrators to pre-
vent bridging. Vibrators are generally superior
for this purpose since reciprocating and rotating
bar agitators tend to shear and bend out of shape
under heavy loads. To be most effective, vibrators
are usually mounted about one -fourth of the distance
from the base of the sloping bottom of the bin, which
is usually constructed of a large, unsupported sec-
tion of sheet metal. This method of construction
permits transmis sion of vibration more easily than
-------
442
INCINERATION
transmission from sloping sections rigidly sup-
ported with stiffened angles or steel structural
members. If the bottom is so large as to require
some type of external cross-sectional support, the
support members should be attached only at the
edges of the section.
\
Figure 322. Vertical-sided feed bin with four parallel
screw conveyors (Brown Saltman Furniture Co., Los
Angeles, Calif.).
Vibrators operating continuously may cause the
wood waste to pack and bridge in the bottom of the
bin. To remedy this condition, the frequency of
vibration or the amplitude of the vibratory stroke
maybe changed, or a mechanical timer maybe in-
stalled to actuate the vibrators at desired intervals.
Screw or Drag Conveying
Screw or drag conveyors are placed in the bottom
of a feed bin to remove sawdust and shavings from
the bin at a regulated rate. Screw conveyors are
preferred, except where long, tough, fibrous shav-
ings are to be conveyed. Since material such as
this would bind in conveyor flights, the more ex-
pensive drag conveyor must be used.
Screw conveyors with variable pitch are recom-
mended over screws with uniform pitch because
they permit more even loading of the screw along
the entire length and thus minimize the compress-
ing of sawdust and shavings causing bridging above
the discharge end of the screw. Because rela-
tively large pieces of wood may enter the convey-
ing system, screw conveyors should be at least 6
inches in diameter to ensure their passage.
Regulation of the flow of wood waste is dependent
upon the bulk density of the waste to be handled as
well as upon the number, diameter, and speed of
the screw conveyors. The bulk density of most
wood wastes varies from 4 to 12 pounds per cubic
foot, depending upon the kind of wood processed
and the shape of the particles. Determination of
the density must be based upon sawdust in its com-
pressed form at the bottom of the bin, rather than
in aloose form. Once the density has been estab-
lished, the type of bin selected, and number of
screws determined, the diameter and speed of the
screws can be calculated. Provisions should also
be made for a gear head or varidrive to regulate
the speed of the conveyors so that they supply wood
waste over a range of 33 to 100 percent of the burn-
ing capacity of the incinerator.
To prevent sawdust from being aspirated into the
pneumatic conveying system faster than the normal
delivery rate of the screw, conveyors should ex-
tend at least three screw diameters beyond the end
Figure 323. Feed bin with sloping bottom
(California Moulding and Trim Mfg. Co.,
Los Angeles Calif.).
-------
Multiple - Chamber Incinerators for Burning Wood Waste
443
Figure 324. Feed bin with vee bottom (Orban
Lumber Co., Pasadena, Calif.).
of the bin, and the shrouds should be installed over
the extended section, as shown in Figure 325.
Shrouds are usually adjusted after the unit is in
operation, to provide optimum clearance over the
flights and prevent binding.
Pneumatic Conveying
While general design features for pneumatic con-
veying systems are discussed in the preceding
chapter, a number of specific features should be
considered in designing pneumatic conveying sys-
tems for wood waste incineration.
Pneumatic conveying systems are generally de-
signed for a ratio of 2/3 cfm of conveying air per
hour per pound of sawdust to be burned. About 10
percent of this conveying air should be admitted to
the incinerator along with the wood waste to assist
in spreading the particles evenly over the entire
grate area and to maintain active flame over the
surface of the burning pile. The amount of convey-
ing air entering the ignition chamber rnay be regu-
lated by ins tailing either a butterfly damper in the
10-5 outlet duct of the cyclone separator, or spiral
vanes withri '.he cone o- the cyclone separator.
Sawdust pickup and conveying velocities should be
at least 3, 000 fpm to prevent sawdust blockage in
the ductwork. Blower motors should be oversized
to accommodate occasional surges of sawdust
through the pneumatic conveying system.
Figure 325. Screw conveyor with shroud (Acme Woodcraft,
Los Angeles, Calif.).
Cyclone separators used in conjunction with the
blower are of the small-diameter, high-efficien-
cy type with separation factor s that exceed 100, as
described in Chapter 4.
A flap-type damper equipped with a counterbalance
weight should be installed at the bottom outlet of
the cyclone separator. This damper is adjusted
to close automatically when the blower is not in
operation, which prevents the hot gases of the ig-
nition chamber from damaging the sheet metal of
the cyclone separator and also prevents smoke from
being emitted to the atmosphere from the top of
the cyclone. This damper should be constructed of
1/4-inch stainless steel plate since it is subject to
intense radiation from the burning pile. By con-
struction of a square -shaped damper with a square
duct extending below, the damper is able to swing
out of the wa-y of the falling wood waste. The
damper should be large enough to form a tight,
overlapping seal with a smooth, stainless steel
flange located below the round duct at the bottom
of the cyclone separator.
To ensure proper operation, the equipment should
be electrically interlocked to start simultaneously
or in the following order : (1) Blower, (2) convey-
ors, (3| vibrators or agitators.
STANDARDS FOR CONSTRUCTION
While structural features of wood-burning incin-
erators are similar to those o>" general-refuse in-
cinerators, wood-burning incinerators must be de-
signedfor greater stresses and strains caused by
increased thermal expansion resulting from higher
temperature operation. Refractories, therefore,
are selected to resist not only normal abrasion
-------
444
INCINERATION
from charging but also erosion, spalling, and slag-
ging resulting from high-temperature, high-veloc-
ity flame impingement.
Refractories
Super-duty plastic refractory or super-duty fire
clay firebrick are recommended for the interior
walls and arches that come into direct contact with
flames and hot gases, since temperatures usually
exceed 2,000°F. Expansion joints -with 1/2-inch
minimum width should be installed for every 6-foot
section of refractory. These joints must be sealed
completely with high-duty ceramic packing with
minimum service temperatures of 2, 500°F, Pack-
ing of this type is necessary to prevent ashes from
collecting in the open joints and fusing in such a
•way as to render the joint useless.
The first 10 feet of stack must be lined with high-
duty firebrick or an equivalent castable refractory.
The remainder of the stack may be lined with a
lower duty material since flame impingement in
this area does not normally occur. The charging
door and other access doors, "with the exception
of the ash pit cleanout doors, should be lined with
120-pound-per-cubic-foot, ASTM Class 24, cast-
able refractory.
The minimum heights for free-standing firebrick
walls of given thickness are as follows.
Thickness of walls, in. Unsupported height, ft
4-1/2
9
13-1/2
4
10
14
Firebrick walls extending above these heights should
be held to exterior supports •with stainless steel
anchors that permit a differential rate of expan-
sion. Walls constructed of plastic refractory should
be anchored to exterior structural steel members
on 18-inch centers.
Arches may be constructed of firebrick or plastic
refractory. Firebrick arranged to form 60-degree
arches should be limited to a maximum span of 5
feet 10 inches for 4-1/2-inch thickness and 8 feet
for 9-inch thickness. Arches with spans greater
than 8 feet should be constructed of suspended,
super-duty, fireclay shapes or super-daty, plas-
tic refractory. Plastic refractory used for this
purpose must be suspended from refractory cones
or stainless steel anchors spaced not more than
15 inches apart.
Grates
Two materials satisfactory for construction of
grates are cast iron and castable refractory. Cast
iron grates are available in a -wide variety of sizes
and shapes. They are of much heavier construc-
tion than those used in comparable general-refuse
incinerators, to minimize deformation at high tem-
peratures. Where blocks or scraps of wood are to
be burned, bar- or channel-shaped grates should
be employed, but when wood -waste accumulated
from wood-working equipment is to be burned, pin-
hole grates should be installed. Typical pinhole
grates consist of 6-inch-•'wide by 24-inch-long by
3/4-inch-thick slab sections of cast iron -with 1/2-
inch holes on 2-inch centers. Grates of this type
are capable of retaining small wood particles that
might otherwise fall unburned into the ashpit.
Refractory grates are nearly always constructed
in the form of 60-degree sprung arches. On incin-
erators of 250-pound-per-hour capacity or less,
grates are constructed of ASTM Class 24 refrac-
tory 5 to 6 inches thick, with 1-inch holes on 5- to
6-inch centers. ASTM Class 27 castable refrac-
tory 6 to 8 inches thick, -with 1-inch holes on 6- to
9-inch centers is used in larger incinerators.
Caution is required in operating incinerators with
cast iron grates. Underfire air must not be unduly
restricted nor should the ash pit be allowed to fill
within 1 foot of the grates. Heat buildup in the ash
pit from either condition can cause the grates to
warp and sag. Misoperation of this type does not
deform grates constructed of castable refractory.
These grates are, however, susceptible to damage
from careless stoking and cleaning.
When most of the charge consists of sawdust or
similar small wood particles delivered by a uni-
form feed system, a solid hearth should be installed
at the rear of the ignition chamber to prevent the
introduction of excessive underfire air at this lo-
cation. As the size of the incinerator increases,
hearths are sometimes installed along the sidewalls
also to prevent excessive underfire air. In any
event, the hearth area should not exceed 30 per-
cent of the total horizontal area of the primary
ignition chamber.
Exterior Walls
Incinerators can be constructed with exterior walls
of red brick or steel plate. Red brick exteriors
are usually constructed of two layers of red brick
bondedbya reinforced concrete center. Exterior
steel plate may be of the thin, corrugated type used
to back plastic refractory, or as heavy as 10 gauge
to support interior brick construction.
Air Ports
Combustion air port controls shouldbe constructed
of cast iron not less than 1/2 inch thick. These
ports should fit tightly to reduce air leakage to a
minimum.
-------
Multiple-Chamber Incinerators for Burning Wood Waste
445
OPERATION OF INCINERATORS
Certain differences exist between the operation of
•wood-burning incinerators and general-refuse in-
cinerators. The operator of a general-refuse in-
cinerator generally relies on auxiliary burners to
maintain temperatures for maximum combustion
in the secondary chamber. The operator of a wood-
burning incinerator, without provisions for auxil-
iaryburners, is able to maintain adequate secon-
dary chamber temperatures by proper charging
and control of combustion air.
Generous amounts of clean dry paper are mixed
with the wood for the initial charge. After the igni-
tion chamber is one-half to two-thirds full, addi-
tional paper is placed on top of the pile to ensure
quick flame coverage at the surface. It is impor-
tant, in keeping smoke to a minimum, that only
clean dry paper and dry scrap wood comprise the
initial charge. After charging is completed, the
paper is ignited near the front of the chamber and
the charge door is closed. All combustion air ports
are almost completely closed to restrict combus-
tion air.
As burning proceeds, the incinerator passes through
the most critical period of its operation. By ob-
serving the emissions from the stack, the neces-
sary adjustments can be made promptly to reduce
or eliminate smoke. Gray or white smoke emit-
ted after lightoff indicates that the incinerator is
cold. This smoke can be minimized or eliminated
by closing all air ports. Smoke of this color usu-
ally ceases within a fewminutes after lightoff when
flames completely cover the refuse pile and fill
the flame port. A few minutes later, black smoke
may appear, resulting from a lack of adequate
combustion air. These emissions can usually be
eliminated by opening primary air ports and then
the secondary air ports. If additional combustion
air is required, it may be supplied by opening the
charge door.
Although each incinerator has its own operating
characteristics, the overfire and underfire air
ports can usually be opened 5 to 10 minutes after
lightoff, and the secondary port, 20 to 30 minutes
later. If opening of the secondary ports results in
gray or white smoke emissions, the ports should
be closed immediately since the incinerator has
not yet reached its normal operating temperature.
After attaining normal operating temperatures,
maximum combustion is maintained by placing the
mechanical feed system in operation or by hand
charging at regular intervals.
Illustrative Problem
Problem:
Design a multiple-chamber incinerator to burn
1, 000 pounds of Douglas fir waste per hour with a
maximum moisture content of 10 percent.
Solution:
1. Composition of the refuse:
Dry cornbus -
titles (1,000 lb/lir)(0. 90) = 900 Ib/hr
Moisture (1,000 lb/hr)(0. 10) = 100 Ih/hr
Total 1-000
2. Gross heat input:
From Table 118, the gross heat of combustion
of 1 pound of dry Douglas fir is 9, 050 Btu/lb
(900 lb/hr)(9, 050 Btu/lb) = 8, 140, 000 Btu/hr
3. Heat losses:
(a) Assume radiation, convection, and storage
heat losses are 20 percent of gross heat
input:
(0.20)(8, 140, 000 Btu/hr) = 1, 625, 000 Btu/h:
(b) Evaporation of contained moisture:
The gross heat of vaporization of water at
60°F is 1, 060 Btu/lb
(100 lb/hr)(l, 060 Btu/lb) = 106, 000 Btu/hr
(c) Evaporation of water formed by combus-
tion:
From Table 118, there is 0. 563 Ib of water
formed from the combustion of 1 pound of
dry Douglas fir.
JO. 563 Ib H O\
(900 lb/hr)( — ) (1, 060 Btu/lb)
= 537,000 Btu/hr
(d) Total heat losses:
a + b + c = total losses
1,625,000 Btu/hr + 106,000 Btu/hr +
537, 000 Btu/hr = 2, 268,000 Btu/hr
4. Net heat available:
8, 140,000 Btu/hr - 2, 268,000 Btu/hr =
5,872,000 Btu/hr
5. Weight of products of combustion:
From Table 118, there is 13. 86 Ib of com-
bustion products from 1 pound dry Douglas fir
-------
446
INCINERATION
with 100 percent excess air, and 20, 30 Ib of
combustion products from 1 pound dry Doug-
las fir with ZOO percent excess air. Weight
of products of combustion at 100 percent ex-
cess air:
Wood (900 lb/hr)(13. 86 Ib/lb) = 12,4501b/hr
Moisture (100 Ib/hr) 100 Ib/hr
12,550 Ib/hr
Weight of products of combustion at 200 per-
cent excess air:
Wood (900 lb/hr){20. 30 Ib/lb) = 18, 200 Ib/hr
Moisture (100 Ib/hr) 100 Ib/hr
18,300 Ib/hr
6. Average gas temperature at 200 percent ex-
cess air:
Q = w c (T - T )
p p 2 1
where:
Q - net heat available, Btu/hr
w = weight of products of combustion,
P Ib/hr
c = specific heat of products of combus-
P tion, Btu/lb-°F
T = average gas temperature, °F
T = initial temperature, °F
T = T + —
2 1 (wp)(cp)
. 2, 700 scfrn 2
Area = —.,. . = 3- 00 ft
900 fpm
or 432 in.2
= 60°F
= 1,300°F
5, 872, OOP Btu/hr
(18, 300 lb/hr)(0. 26 Btu/lb-°F)
Overfire air port area:
Assume overfire air port area is 90 per-
cent of total.
(0. 90)(432 in.2) = 388 in.2
Underfire air port area:
Assume underfire air port area is 10 per-
cent of total.
(0. 10)(432 in.2) = 43 in?
(b) Secondary air ports located in the bridge
wall and curtain wall:
Assume 50 percent theoretical air through
each port.
From Table 118, 42.37 scf of air is re-
quired per pound of dry Douglas fir.
(900 lb/hr)(42.37 scf/lb) = 38, 100 scfh
or 634 scfm
Average air velocity through secondary
ports is assumed to be 1, 000 fpm or 0. 063
in. WC velocity pressure.
634 scfm . 2
Area = = 0.63ft
1, 000 fpm
or 91 in.
j. Volume of products of combustion:
(a) Volume through flame port:
Assume 100 percent excess air through
flame port.
From Table 118, there is 179. 6 scf of prod-
ucts of combustion from 1 pound dry Doug-
las fir.
7. Combustion air port areas:
Wood
(900 lb/hr)(179. 6 scf/lb) = 161,000 scfh
(a) Primary air port sizes:
Assume primary air at 100 percent excess.
From Table 118, 179. 6 scf of air is re-
quired per pound of dry Douglas fir
(900 lb/hr)(179.6 scf/lb) = 161, 300 scfh
or 2, 700 scfm
Moisture
(100 Ib/hr)
2- 100
,i -
18 Ib/lb mole ! 63 ,100 scfh
or 2, 720 scfm
or 45. 4 scfs
(b) Volume through mixing chamber:
Assume the average air velocity through
the primary air ports is 900 fpm or 0. 052
in. WC velocity pressure.
Assume 50 percent theoretical air is add-
ed through secondary port to combustion
products from primary chamber.
-------
Flue-Fed Apartment Incinerators
447
(900 lb/hr)(42. 37 scf/lb) = 38, 100 scfh
or 10.6 scfs
Totalvolume = 45.4scfs
+ 10. 6 scfs
56.0 scfs
(c) Volume through combustion chamber:
Assume 50 percent theoretical air is added
through cooling air ports in curtain wall.
Total volume = 56.0 scfs + 10.6 = 66.6 scfs
9. Incinerator cross-sectional areas :
(a^ Flame port area:
From Table 120, designfor an average ve-
locity of 50 fps and 1, 900 °F gas tempera-
ture:
10. Grate area:
From Figure 310, the grate loading is 33 lb/
hr-ft for the upper +10 percent curve used
for 9,000 Btu/lb refuse.
(1, OOP Ib/hr)
Total grate area 33 Ib/hr.ft2
= 30.3 ft
11. Horizontal dimensions of ignition chamber:
From Table 120, length-to-width ratio = 1.5
Let W = width and L = 1. 5W
Then (W)(l. 5W) = 30. 3 ft2
1. 5W = 30. 3
W2 =-20. 1
Width = 4. 5 fi
Length = (1.5)(4. 5) = 6.75
(45.4 scfs)(2,360°R)
- (50fps)<520-R)
(b) Mixing chamber cross -sectional rea:
From Table 120, designfor an a ^rage ve-
locity of 25 fps and 1, 550 °F b_ . tempera-
ture:
(56.0 scfs)(2,010°R) _ 2
Area = — .,,--,.. OT,. = o. 65 ft
(25 ±ps)(520 R)
(c) Curtain wall port area:
From Table 120, designfor an average ve-
locity of 20. 0 fps and 1, 500°F gas tempera-
ture :
(56.0 scfs)(l,960°R) 2
Area = 7———•— —— — = 10. 5 it
(20. 0 fps)(520°R)
(d) Combustion chamber cross-sectional area:
From Table 120, designfor an average ve-
locity of 7. 5 fps and 1,200°F gas tempera-
ture:
Area =
(66. 6 scfs)(l, 660°R)
(7. 5 fps)(520°R) = 28.4
(e) Stack area:
From Table 120, designfor a velocity of 20
fps and 1, 100°F gas temperature:
(66. 6 scfs)(l. 560°R) 2
<20fps)(520'R) = 10'°ft
12. Arch height:
From Figure 311, the arch height for the up-
per + 10 percent curve is 5 ft.
13. Stack height:
From Figure 313, stack height above grade is
35 ft.
FLUE-FED APARTMENT INCINERATORS
INTRODUCTION
An incinerator in which the chimney also serves
as a chute for refuse charging, as shown in Fig-
ure 326, is known as a flue-fed incinerator (Mac-
Knight et al. , 1960). For some 40 years the sin-
gle-chamber, flue-fed incinerator has been built
as an integral part of apartment buildings. The
incinerator is usually located centrally in the build-
ing to minimize the distance from the apartments
to the charging chutes located on each floor. Oc-
casionally the incinerator may be located on an
outside wall-with charging chutes outside the build-
ing and adjacent to a balcony or fire escape plat-
form. The flue-fed incinerator is also used to
some extent in schools, hospitals, and office build-
ings.
The single-chamber, flue-fed incinerator has a
box-like combustion chamber separated by dump
grates from an ashpit below. Atmospheric gas
burners located below the grates are used primari-
ly for dehydration of garbage and other wet mate-
rial. A cleanout door is provided for removal of
ashes from the ash pit. A charging door above the
grates is used for igniting the refuse and for stok-
-------
448
INCINERATION
CHARGING DOOR
'OVERFIRE
/AIR PORT
COMBUSTION CHAMBER
GRATES
BASEMENT
FLOOR
CIEANOUT DOOR
UNDERFIRE AIR PORT
Figure 326. Unmodified flue-fed incinerator
(MacKnight et al., 1960).
ing the burning pile near the end of the burning
period. In most instances both doors are provided
with spinners to allow admission of underfire (under-
grate) and overfire (overgrate) air. The walls of
the incinerator are customarily constructed of two
layers of brick. The inner layer consists of 4-1/2
inches of firebrick separated by a 1/2-inch air-
space from a 9-inch common brick exterior. The
flue is normally constructed of 9 inches of common
brick with a 1-inch flue tile lining. The inside di-
mensions of the flue are usually 16 by 16 inches
for apartment buildings 3 to 4 stories in height.
Description of Refuse
The composition (% by wt, minimum to maximum)
of apartment house refuse usually falls v/ithin the
following limits: Dry paper, 50 to 100; garbage,
0 to 30; plastics, 0 to 3; noncombustibles, 2 to 10;
and other (including rags, waxed cartons, green-
ery, and so forth), 0 to 10. If food is prepared in
the apartments the percent of garbage, plastics,
•waxed cartons , and noncombustibles in refuse ap-
proaches the upper limits . If food is not prepared
on the premises then the refuse is more likely to
have ahigher percentage of drypaper. An average
value taken by incinerator designers for the produc-
tion of refuse by apartment dwellers is 1 pound per
person per day.
THE AIR POLLUTION PROBLEM
When first ignited, refuse in a flue-fed inciner-
ator burns very rapidly. Air inspiration during
this initial flash burning period is usually insuffi-
cient to meet the combustion requirements of the
rapidly burning dry refuse, resulting in incomplete
combustion and black smoke emissions. The con-
current extreme gas turbulence results in the en-
trainment of large quantities of fly ash (ash and
charred paper).
Once the initial flash burning period has passed,
an excessive draft develops as a result of the high
flue gas temperature and the long flue. The amount
of air admitted through the air ports becomes great-
er than the demand for combustion air, and the
temperature in the combustion chamber gradually
diminishes as the excess air increases. As this
process continues, the combustible gases, oils,
tars, and fats, produced by low-temperature com-
bustion atthe surface of the refuse pile and by de-
structive distillation within the pile, pass out the
stack incompletely burned in the form of white or
light gray smoke.
The use of undergrate burners tends to entrain fly
ashin the hot gases passing through and around the
fuel bed. This problem is further aggravated by
stoking of the burning refuse pile under excessive
draft conditions, resulting in the discharge of large
quantities of fly ash from the stack. The problem
is compounded by the charging of refuse down the
flue during the burning period, which smothers and
scatters the burning pile and results in severe fly
ash emissions and smoke production.
Stack Emissions
The range of particulate emissions found by a
series of tests, in pounds per ton of refuse burned,
is shown in Table 121. Associated data are in-
cluded in the table as a matter of general interest.
Other, less plentiful data indicate that emissions
(in pounds per ton) are as follows: organic acids,
9. 5; aldehydes, 1. 5; hydrocarbons, 2; and nitro-
gen oxides, 6.
AIR POLLUTION CONTROL EQUIPMENT
There are three basic methods of altering a flue-
fed incinerator to prevent the discharge of air con-
-------
Flue-Fed Apartment Incinerators
449
Table 121. PARTICULATE EMISSIONS FROM
TY. ^"A.L FLUE-FED INCINERATORS
Test
designation
C-95
651
881
C-116
D-3
D-2
C-499-1
650
D-l
C-43
C-50
C-44
Particulate matter,
Ib/ton
76
52
48
37
37
34
25
23
23
19
17
7
gr/scf
at 12% CO^
2.27
1.40
1.60
1. 40
1. 18
1. 06
0. 94
0.99
0.75
0.75
0.60
0.27
gr/scf
0. 61
0. 13
0. 21
0. 20
0. 21
0. 18
0. 10
0. 2
0.26
0. 09
0. 08
Average
stack
volume,
scfm
458
1, 190
326
213
820
930
500
1, 120
860
530
441
817
Grate
area,
ft2
1. 5
16
9
8
12
12
6
12
12
4
20
8
Stack
height,
ft
25
35
68
32
80
80
54
56
80
25
56
46
taminants. Two of these methods involve the ad-
dition of an afterburner to the existing incinerator.
The third method involves the installation of a well-
designed multiple-chamber incinerator'. . Appur-
tenances for regulating stack draft and effective-
ly controlling the charging during the burning peri-
od are essential to these methods.
INSTALLATION OF AFTERBURNER ON A ROOF
A typical installation of an afterburner on a roof
includes the use of a damper located at the base of
the stack to control the excessive draft and burn-
ing rate, and an afterburner mounted on top of the
existing stack to control the smoke and combustible
gases. Chute door locks are installed to prevent
damage to the draft control damper from the charg-
ing of refuse during the burning period.
Design Procedure
Draft control
The excessive draft conditions prevailing in flue -fed
incinerators mustbe reduced before an afterburn-
er will function successfully. A swinging, counter-
weighted damper can be used for draft control by
pivoting it on a rod along one edge so that it can be
swung flush with one wall of the flue to permit
charging of refuse. Swung back into a horizontal
position, the damper can maintain the draft in the
basement combustion chamber within suitable lim-
its. This damper is equipped with adjustable open-
ings in its surface since the exact restriction re-
quired for a specific unit cannot be determined.
By changing the location of the adjustable plates
fastened to the surface of the damper, the size of
the openings can be decreased until the desired
draft, usually from 0. 10 to 0. 20 inch WC, is at-
tained. These dampers are located in the flue be-
neath the first-floor chute door to ensure a nega-
tive pressure at each door and thus prevent smoke
and sparks from blowing by the doors into the
building.
The effect of a damper on combustion chamber
draft, burning rate, and flue gas velocity is shown
in Figures 327, 328, and 329. These graphs were
obtained by testing a 6-story, flue-fed incinerator
equipped with a draft control damper having a 6-
1/2-inch-diameter orifice and afterburner. The
curves in the graphs designated "uncontrolled flue-
fed incinerator" were obtained by operating the
incinerator with the damper open and the after-
burner -off.
Figure 327 shows the draft in the combustion cham-
ber of the incinerator to be lower and more stable
"when the damper is in use. Figure 328 shows that
the initial peak burning is considerably reduced
when the damper is used. Figure 329 shows that
the flue gas velocities are lower when the draft
control damper is used.
The lower draft condition in the combustion cham-
ber, attained from the use of a draft control damp-
er, minimizes the entrainment of fly ash in the flue
0 35
0 25
_ 0 20
0 10
UNMODIFIED FLUE-FED INClNEBATDR
I
FLUE-FED INCINERATOR «ITH AFTERBURNER
AND DRAFT CONTROL DAMPER
I I I
LEGENC
B OVERGRATE BURNERS IGNITED
S REFUSE STOKED
20 30
TIME OF OPERATION -i i
Figure 327. Draft in combustion chamber of
modified and unmodified flue-fed incinera-
tors (MacKmght et ai., i960).
-------
450
INCINERATION
BOO
LEGEND
UNMODIFIED FLUE-IED INCINERATOR
FLOE-FED INCINERATOR MODIFIED
HITH DRAFT CONTROL DAMPER
ONDERGRATE BURNERS IGNITED
REFOSE STOKED
15 20 25
TIME FROM LIGHTOFF m n
Figure 328. Burning rate versus time for modified and unmodified flue-fed incinerators.
gases. This condition also reduces the burning
rate, permitting the use of a smaller aiterburner
than otherwise wouldbe required. Installation and
operating costs of the afterburner are according-
ly reduced,
Chute door locks
The charging of refuse during the burning period
can be prevented easily and economically by in-
stalling solenoid locks on each of the chiite-charg-
ing doors. The use of this type of lock permits
their actuation from a single switch in the base-
ment before the damper is closed and the burning
cycle begins.
If refuse is charged down the flue during the burn-
ing period when the draft-controlling damper is
closed, several undesirable events may occur.
The damper may be bent, or even detached from
its supports, or the refuse may pile up on the
damper and block the flue, causing the gases from
the refuse burning in the combustion chamber to
vent into the basement. Chute door IOCKS prevent
these problems in addition to preventing smother-
ing of the refuse pile and the subsequent creation
of smoke and lly ash.
Design parameters
Parameters for roof afterburners are essentially
the same as the parameters employed in design-
ing afterburners for smokehouses, ovens, and so
forth. For a discussion of appropriate parameters
relating to retention time, mixing of gases, gas
velocities, temperatiire levels, flame character-
istics, and burner types and arrangements, see
the first part of Chapter 5,
Limitations
Most flues are not airtight since cracks develop
with age and use. In particular, relatively large
openings occur around the chute doors.
Air inspirated in this manner mixes with the flue
gases and passes through the afterburner. Addi-
tional air entering the afterburner lowers exit gas
temperatures, increases gas velocities, and re-
duces residence times. Thus, this overall effect
reduces the efficiency of the afterburner.
As the height of the building increases, the air in-
duced into the flue also increases. No definite
building height limitation can be given since air
GPO 805—614-16
-------
Flue-Fed Apartment Incinerators
451
FLUE-FED INCINERATOR IITH AFTERBURNER
AND DRAFT CONTROL DAMPER
LEGEND
6 - OVERGRATE BURNERS IGNITED
S REFUSE STOKED
I
20 30
TIKE OF OPERATION mm
Figure 329. Flue gas velocity at inlet of afterburner
of modified and unmodified flue-fed incinerators (Mac-
Knight et a)., 1960).
leakage increases in importance with increasing
height. As of 1963, however, a 9-story building
is the tallest building in Los Angeles County on
which a roof afterburner has been successfully
employed.
Typical installations
Figures 330 and 331 are cutaway drawings of two
typical afterburners mounted on flue-fed inciner-
ators. The inner passage of the afterburner in
Figure 330 is built in the shape of a lazy L. A
premix gas burner fires horizontally into the pass-
age |ust below the L. The inclined section above
the burner provides an impingement surface for
the burner flames and also deflects the effluent
from its vertical path. Mixing of the flames and
Hue gases beneath the inclined section has proved
adequate to burn the contaminants in the inciner-
ator gases.
The afterburner shown in Figure 331 consists basic-
ally of a ring burner followed by a venturi throat,
a baffle to ensure contact between burner flames
and flue gases, and a combustion chamber. The
ilue gases enter the afterburner through the ring
burner. The cross section of the burner ring is
ELECTRIC LOCK
/CHUTE DOOR
CHARGING
DOOR
OVERFIRE
AIR PORT
CLiANOUT DOOR
UNDERFIRE
AIR PORT
Figure 330. Flue-fed incinerator modified by a roof
afterburner and a draft control damper.
\-BAFFLE
BURNER
Figure 331. Flue-fed incinerator modified by a roof
afterburner, and a draft control damper (Sargent
Afterburner, Kearney, N.J.).
-------
452
INCINERATION
that of an isosceles right triangle, the hypotenuse
connecting the two legs of the triangle forming the
inside surface of the burner. Equally spaced ori-
fices are located in this surface of the burner to
create a conical flame pattern and yet prevent the
flames frombeing extinguished by the rush of flue
gases over its inner face.
The premix burner, venturi throat, and baffle are
empirically sized to cause maximum mixing between
the flue gases and burner flames consistent -with
minimum pressure loss through the afterburner.
All smoke and unburned volatiles passing through
the ring burner are brought into intimate contact
with the flames. Additional combustion air may
be supplied through openings below the burner in
the wall of the afterburner. Because of the remote
location of the afterburner, automatic spark igni-
tion and complete flame failure controls are usu-
ally installed.
Standards for Construction
There are several reasons why the maintenance of
a roof afterburner is likely to be inadequate. First,
it operates in an unfrequented location. Second,
responsibility for its operation and service is usu-
ally assigned to unskilled janitorial personnel.
Third, its installation and use stem strictly from
legal compulsion, and little attention is given over
and above the minimum necessary to meet the re-
quirements of the law. Consequently, an after-
burner should be constructed of durable materials
that require as little maintenance as possible.
Mounting and supports
The flue is dismantled to a height of 2 feet above
the roof, and the afterburner is constructed on the
flue above this point. Shortening the flue facili-
tates -work on the unit, reduces windloading and
earthquake stresses, andmakes the completed unit
less prominent.
One method successfully used to fasten the after-
burner on the flue consists of bolting a 1/4-inch-
thick steel plate to the flue and welding the after-
burner shell to the plate. A central hole the size
of the flue opening is, of course, first cut in the
plate.
Additional support is usually provided by three 1/4-
inch guy cables evenly spaced around the after-
burner. The guys can be welded to the afterburner
shell near its top and attached to the building by
bolts going through the roof. In buildings con-
structed of wood, the bolts should enter the roof
joists.
Metals
The following metals are recommended for use. in
afterburners because experience has shown that
they resist deterioration under the conditions of
their use. Sheet steel with a minimum thickness
of 12-gage is recommended for use in fabricating
afterburner shells. Stainless steel, type 321, 1/8-
inch thick, is recommended for use in all after-
burner baffles receiving direct flame impingement.
Support rods for baffles should have a minimum
diameter of 1/2 inch and should be made of type 321
stainless steel.
Castable refractories
Castable refractories used near the burner are
invariably subjected to direct flame impingement.
Thes e linings should be at least ASTM Class 27 re-
fractories with a minimum thickness of 4 inches.
(Class 27 refractories are those castable refrac-
tories capable of withstanding temperatures of
2,700°F.) In addition, a 1-inch-thick backing of
1,000°F castable insulation or equivalent should
be placed between the refractory and the metal
shell. (The castable venturi throat in the after-
burner of Figure 331 is so constructed.)
Castable refractory used to line the afterburner
shell above the actual burning zone should have a
minimum thickness of 2 inches. Class 24 cast-
ables, that is, those castable refractories capable
of withstanding temperatures of 2,400°F, can be
used in this area. A popular method of installing
these linings consists of casting them in a ring
shape and slipping them into the shell. Support foi
each castable section is derived from metal clips
welded to the inner wall just below the first sec-
tion of the shell. These clips fit into recesses pro-
vided in this castable section so that they are pro-
tected from the flames of the afterburner.
Firebrick
Of the four types of firebrick, only high-duty brie
is normally used in afterburners. When used, i
is generally limited to areas receiving direct flam
impingement or high-temperature flame radiatio
Lower duty castable refractories or insulating fire
brick are normally used instead of firebrick i
places of less severe duty.
Insulating firebrick
In areas of the afterburner without direct flam.'
impingement, 2, 000 °F insulating firebrick may b<
successfullyused. This type of brick is frequent
ly used in 2-1 ,/2-inch-thick rings to line the uppe
section of afterburners. (The stack above the ven
turi throat in the afterburner of Figure 331 is linec
with this type of brick. ) When cut into wedge shapes
and arranged around the inner shell, the individual
bricks lock into place and mortar need not be used
Burners
A forced-draft gas burner such as that described
in the second part of Chapter 9 should be used.
-------
Flue-Fed Apartment Incinerators
453
This type of burner supplies much of its own air
needed for combustion. Since the amount of oxy-
gen in the flue gas is less than that in a normal
atmosphere, the ability to supply a significant por-
tion of its own oxygen requirements is an impor-
tant factor in burner selection.
Draft control damper
The draft control damper receives direct flame
impingement from the refuse burning in the com-
bustion chamber. For this reason, the damper
must b,e constructed of stainless steel. Dampers
of 20-gage type 302 stainless steel with 3/4 inch
of each edge bent down to increase stiffness have
proved satisfactory.
Chute door locks
Figure 332 shows a typical chute door lock instal-
lation. The lock shown is of the type that allows
the door to be closed at all times without breaking
the latch. This type of latch is recommended for
use because it permits the door to be closed dur-
ing the incinerator's operation.
Stack Emissions
Data obtained from stack tests on a typical flue-fed
incinerator modified with a draft control clamper
and roof afterburner are given in Table 122. Data
obtained from the test designated in the table as
C-546 also give emis sions of aldehydes as formal-
dehyde as 2 pounds per ton, emissions of organic
acids as acetic acid as 2. 1 pounds per ton, and
emissions of nitrogen oxides as 7 pounds per ton.
Operation
The sequence of operations performed in using a
flue-fed incinerator, modified as discussed here-
in, starts with the locking of the chute doors from
the main switch in the basement. The draft con-
trol damper is closed and the afterburner ignited
by remote control from another switch also lo-
cated in the basement.
Figure 332. Typical chute door lock installation.
The refuse is then ignited and, if the refuse is
moist, the grate burners are also lighted. The
refuse may be stoked frequently to uncover fresh
material without fear of creating excessive fly ash
emissions because of the draft-limiting action of
the damper.
When the refuse has been destroyed, the grate
burners are turned off and the grates are cleaned
by dumping the ashes into the ash pit. After a brief
period of time is allowed, to permit smoke from.
the smoldering ashes to clear, the afterburner is
turned off and the draft control damper opened.
The final step, that of unlocking the chute doors,
should not be performed until about 1 0 minutes after
the grates have been cleaned. This delay allows
Table 122. PARTICULATE EMISSIONS FROM A TYPICAL FLUE-FED INCINERATOR
MODIFIED WITH A DRAFT CONTROL DAMPER AND A ROOF AFTERBURNER
Test
designa-
tion
C-586-A1
C-586-A2
C-586-A3
C-546
Burning
rate,
Ib/hr
100
80
68
49
Particulate matter
Ib/ton
5. 9
5. 2
5.6
1.2
gr/scf
at 12% CO^
2
0. 20
0. 18
0. 20
0. 15
gr/scf
0.004
0. 035
0. 034
0. 027
Afterburner
efficiency,
%
80
82
80
85
Average
oxygen
content,
%
12. 1
11.6
12. 7
9. 5
Average
stack
volume,
scfm
760
690
710
590
Average
outlet
temperature,
0 F
1, 130
1, 240
1, 130
1, 560
-------
454
INCINERATION
the incinerator to cool so that newly charged refuse
is not ignitedby the residualheat in the incinerator.
BASEMENT AFTERBURNER
A typical basement afterburner installation uses
a damper located at the base of the stack to control
excessive draft and burning rate, and an afterburn-
er located directly above the damper to consume
smoke and combustible gases. Cooling air is ducted
from the basement and admitted into the flue just
below the first-floor charging chute to lower the
temperature of the flue gas, thereby protecting the
low-duty refractory lining of the flue and preventing
the chute doors from becoming excessively hot.
Chute door locks are also used on this unit to pre-
vent damage to the damper from the charging of
refuse during the burning period.
Design Procedure
The solenoid chute door locks and the draft control
damper described in the discussion of roof after-
burners are equally applicable to a basement after-
burner installation. When used in a basement af-
terburner, the draft control damper is positioned
in the flue as closely as possible to the combustion
chamber. If the damper in this location is allowed
to swing downward into the combustion chamber
to permit charging, its upward swing may be ob-
structed by the accumulated refuse. To overcome
this problem, the damper must be hinged to per-
mit it to swing upward and lie against the flue "wall.
Design parameters
The parameters, such as retention time, temper-
ature level, and so forth, employed in designing a
basement afterburner are the same as those em-
ployed in designing afterburners for smokehouses,
ovens, and so forth as given in the first part of
Chapter 5. The following specific features not
encountered in the design of industrial afterburn-
ers must be considered in designing a basement
afterburner.
1. The burners themselves must be located so
that they do not obstruct the fall of refuse
through the flue. Relatively inexpensive at-
mospheric or venturi burners are used in this
installation since they can be arranged to fire
across the flue. While forced-draft burners
maybe used, their higher cost usually makes
them impractical.
2. The shape of the flue cannot be modified to
produce a desired gas velocity, induce turbu-
lence, or promote flame coverage, as the com-
bustion chamber of an industrial afterburner
can. The desired flame contact and mixing
are promoted in basement afterburners by the
proper location of the orifices in the damper.
This damper is designed with an orifice located
directlybelow the mouth of each burner. This
arrangement provides the necessary contact
between the afterburner flames and the prod-
ucts of combustion.
It it also necessary to provide for the admission
of outside air into the flue to lower the temperature
of the gases leaving the afterburner. Cooling the
flue gas to 500 °F protects the low-duty refractory
inner lining of the flue from deterioration and pre-
vents the outer walls of the ilue, including the chute
doors, from becoming excessively hot.
This air is supplied from the basement through a
duct installed through the first floor and is intro-
duced into the flue just below the first-floor charg-
ing door. This arrangement is used to prevent
flue gases from venting directly into the living area
of the building if the flue becomes accidentally
blocked. A uniform draft on the downstream, side
of the afterburner is maintained by a barometric
damper placed in the duct's entrance. It also pro-
vides the advantage of closing if any unexpected
back pressure occurs in the flue.
Typical installation
In the basement afterburner, as shown in Figure
333, the burners are mounted in a rectangular
hole located in the flue a short distance above the
basement combustion chamber. A steel frame in-
serted in the hole supports the flue and the compo-
nents of the afterburner. The afterburner unit con-
sists of four venturi burners equally spaced across
the opening. The area of the opening not occupied
by the burners is covered by a steel plate to pre-
vent the entrance of dilution air. An adequate
amount of secondary air is admitted through this
plate by holes equally spaced around each burner
The burners are provided with a continuously oper-
ating pilot.
The draft control damper is located just below the
burners. Four slots in the surface of the damper
are located directly under each burner and as near
as practicable to the wall in which the burners are
mounted.
Because temperatures in the flue in the afterburn-
er zone may be approximately 1, 200 °F or more
the flue tile lining has been replaced with firebrick
The firebrick extends from the damper, past the
burners, and ends just below the first-floor chute
door.
Cooling air is admitted to the flue above the after-
burner zone through a duct fitted with a barometric
damper.
-------
Flue-Fed Apartment Incinerators
455
ELECTRIC LOCK IN OPEN
POSITION FOR CHARGING
CHUTE DOOR
COOLING AIR DUCT
FIRST-FLOOR LEVEL
BAROMETRIC DAMPER
STEEL FRAME
AIR HOLES
PORTS FOR VENTURI
GAS BURNERS
DAMPER WITH ORIFICES
(SHOWN IN POSITION FOR
CHARGING OF REFUSE -
NOTE DURING THE BURNING
CYCLE THE CHUTE DOORS ARE
LOCKED AND THE DAMPER WITH
ORIFICES IS PLACED IN A
HORIZONTAL POSITION
Figure 333. Flue-fed incinerator modified by an
afterburner at the base of the flue.
duty firebrick in the area between the hinged damp-
er and the first-floor chute door. High-duty fire-
brick is recommended instead of lower duty firebrick
or insulating firebrick because the refractory in
this area must withstand both heat and compres-
sive load.
Draft control damper
Since the orifices of the draft control damper are
used to direct the combustion products from the
refuse into the afterburner flames, the damper
should be installed s o as to minimize leakage around
its edges. A small ledge approximately 1/2 inch
wide is built into the refractory lining of the flue
when the refractory is installed. The damper,
when in place, rests against this ledge, prevent-
ing excessive leakage.
Stack Emissions
Emission data, in pounds per ton of refuse burned,
obtained from tests on two typical flue-fed incin-
erators modified with basement afterburners and
draft control dampers are presented in Table 123.
Organic acids are reported as acetic acid, and
aldehydes as formaldehyde.
Operation
The sequence of operation in using a flue-fed in-
cinerator modified with a basement afterburner is
the same as that described in the corresponding
section under roof afterburners.
Ad vantages
Compared with the roof afterburner, the base-
ment afterburner has the advantages of shorter
gas lines, a less expensive ignition system,
and greater accessibility.
Disadv anfages
The basement unit has the disadvantage of cre-
ating a hotter than normal flue, and may re-
quire expensive rebricking in the area near
the afterburner.
Standards for Construction
The construction standards applicable to the draft
con 'ol damper and chute door locks have been
covt ed in the discussion of roof afterburners.
Othe standards are given in what follows.
Hot-zi. e refractory
The flue le lining, which is usually a low-refrac-
tory-dut erracotta, should be replaced with high-
MULTIPLE-CHAMBER INCINERATOR, BASEMENT INSTALLATION
A flue-fed incinerator modified by the installation
of a multiple -chamber incinerator in the basement
includes the conversion of the combustion chamber
of the flue-fed incinerator into a storage chamber
for refuse. The refuse is manually transferred
from storage to the multiple-chamber incinerator
where it is burned. The products of combustion
are ducted back into the flue above a sliding damper
that seals off the refuse chamber, preventing un-
controlled dilution air leakage. As -with other modi-
fications, chute door locks are used to prevent the
charging of refuse during the burning period.
Design Procedure
The second part of this chapter may be consulted
for design procedures used for the multiple-cham-
ber incinerator. Other design features embodied
in the completed assembly follow.
1. The distance between the multiple-chamber
incinerator and the storage bin should facilitate
the transfer of rubbish.
-------
466
INCINERATION
Table 123. EMISSIONS FROM FLUE-FED INCINERATORS MODIFIED WITH A BASEMENT
AFTERBURNER AND DRAFT CONTROL DAMPER
Test
designa -
tion
C-619
C-822
Number
of
stories
4
6
Burning
rate,
Ib/hr
32
104
Particulate matter
Ib/ton
6. 1
6.5
gr/scf
at 12% CO^
0. 22
0. 23
gr/scf
0. Oil
0. 028
Organic
acids ,
Ib/ton
5. i
5. 9
Nitrogen
oxides,
Ib/ton
16. 0
4. 2
Alde-
hydes ,
Ib/ton
3. 1
1. 8
Average
stack
volume,
scfm
970
1, 400
Average
temperature
at stack
outlet, °F
640
450
2. As a further convenience in transferring the
rubbish, the multiple-chamber incinerator
should be constructed with the ignition cham-
ber side close to the storage bin.
3. The multiple-chamber incinerator installed
should be large enough to allow all the refuse
normally collected per day to be consumed
within 1 hour.
4. The draft provided for the multiple-chamber
incinerator should be limited to its design value.
Draft control
The draft furnished for incinerators of the size
commonly used in apartment houses, that is, in-
cinerator s burning between 50 and 250 pounds per
hour, should not exceed approximately 0. ZO inch
of water column. Since the existing flue of the
former single chamber is usually excessively high
for the new installation, some provision ior draft
control must be furnished. A barometric damper,
as shown in Figure 334, is installed at the end of
the breeching be twee i the multiple-chamber incin-
erator and the flue to maintain the correct draft.
Typical installation
The multiple-chamber installation is depicted in
Figure 334. Conversion of the combustion chamber
has been accomplished by removing the grates and
smoothing the interior walls with plaster. To fa-
cilitate the removal of refuse for charging into the
multiple-chamber incinerator, a large section of
the front \\all has been removed and replaced by a
steel door. A breec hing with a barometric damper
has been installed from the top of the secondary
combustion chamber oi the multiple-chamber in-
cinerator to the existing flue. A steel damper has
been installed in the flue below the breeching to
prevent dilution air from entering the flue through
the refuse storage chamber.
Advanfages
The multiple-chamber incinerator installation
has two advantages relative to roof afterburn-
ers.
SLIDING DAMPER-f-.
REFUSE COLLECTION
CHAMBER
BASEMENT FLOOR
MULTIPLE CHAMBER INCINERATOR"
Figure 334. Flue-fed incinerator modified by the installa-
tion of a multiple-chamber incinerator (MacKnight et al.
1960).
i. The cost of installation is lower as com-
pared with that of a roof afterburner on
buildings over 2 stories high. The first
cost of a roof afterburner increases with
building height because of the additional size,
and length of gas line required.
2, It has no height limit. As explained in the
section on draft control for roof afterburn-
ers, the height of a building on which an
afterburner can be installed is limited by
the amount of air leaking into the flue above
the draft control darnper.
-------
Flue-Fed Apartment Incinerators
457
The only advantage a multiple-chamber base-
ment installation has relative to a basement
afterburner is that the flue gases from the
multiple-chamber installation are about 200 °F
cooler, than those from abasement afterburn-
er. Because the flue linings and walls are
correspondingly cooler, theyare subjected to
less thermal stress, and also are less likely
to cause painful burns to apartment tenants.
the refuse is of low heating value or high moisture
content. The charging and operation of the incin-
erator are as described in the second part of this
chapter. Burning is usually carried out once a
day, since the bin does not normally provide stor-
age for much more than that length of time. When
burning is completed, the incinerator burners are
turned off, the doors to the bin are closed, the flue
damper is opened, and the chute doors unlocked.
Di sac/vantages
The multiple -chamber installation has the dis -
advantage of requiring hand transfer of all ref-
use from the storage bin into the multiple -
chamber incinerator--a distasteful and time-
consuming task. A second disadvantage is the
amount of valuable basement space occupied
by the multiple-chamber incinerator, "which
otherwise "would be available for tenants' use.
Standards for Construction
Standards for constructing a multiple-chamber
incinerator may be found in the second part of this
chapter.
Stack Emissions
Emissions from typical flue-fed multiple-chamber
installations are given in Table 124. Associated
data have been included as a matter of interest.
I llustrati ve Problem
Problem:
Calculate the size of a barometric damper to be
installed in the breeching between a basement 100-
pound-per-hour multiple-chamber incinerator and
the flue to limit the draft for the multiple-chamber
incinerator to 0. 2 in. WC.
Given:
The flue is 18 inches square and has a cross-sec-
tional area of 2.25 ft . The flue extends 92 feet
above the breeching. The breeching itself is a 12-
inch-diameter, insulated, straight duct 10 feet
long.
Solution:
1. Compute the theoretical draft in the breeching
at various average gas temperatures:
Operation
Before burning is begun, the solenoid locks on the
charging chute doors are actuated and the damper
below the breeching is closed. The mixing cham-
ber burners of the incinerator are then ignited.
The ignition chamber burners are also ignited if
D
= 0.52 PH (- -
o
"Kent, 1338.
Table 124. EMISSIONS FROM MULTIPLE-CHAMBER INCINERATOR, BASEMENT INSTALLATION
Test
No.
C-511
C-514
C-515
C-513
C-512
Number
stones
4
4
5
11
11
Size of
erator,
Ib/hr
100
50
100
250
150
Burning
rate
during
test,
Ib/hr
65
38
77
217
140
Particulate matter
gr/scf
at 12% CO
0.2
0.2
0.3
0.5
0.3
gr/scf
0.016
0. 035
0. 023
0. 020
0. 016
Emissions, Ib per ton of refuse burned
Partic -
ulate
matter
1. 7
8 4
5. 2
4. 3
4. 5
Organic
acids as
acetic
acid
1. 2
10 5
1. 0
2.6
4.3
Nitrogen
oxides
0.8
2. 3
3. 1
1.7
2.8
Alde-
hydes
as
formal-
dehyde
0. 14
0.47
0.52
0. 37
0.85
Hydro-
carbons
as
hexane
0. 14
3. 16
3 10
No data
4. 20
Avg
stack
scfm
860
510
1, 000
2,700
2, 300
Temp at
top floor
chute
door,
•F
310
310
310
230
190
Draft at
ignition
chamber,
in. WC
Maximum
0.04
Maximum
0.06a
Maximum
0.03b
Not
recorded
Maximum
0.09
Dia of
baro-
metric
damper,
in.
12
10
12
14C
14C
Airflow
through
baro-
metric
damper,
scfm
26Z
None
Not
recorded
Not
recorded
Estimate
850
aA 1/^-in. x 20-in. air leak around sliding damper
Air leaks around sliding damper.
cTwo barometric dampers installed.
-------
458
INCINERATION
where:
D = theoretical draft, in. WC
P = barometric pressure, psi
H = height of flue above breeching, ft
T = ambient temperature, degrees Rankine
o
T = average stack temperature, degrees
Rankine.
For an average flue gas temperature of 100° F:
= (0.52)(14. 7)(92)(
1
•) = 0. 092 in. WC
1
520 560
Theoretical draft (calculated by the same for-
mula) versus temperature is given in the fol-
lowing tabulation:
Temp, °F Dt, in. WC Temp, °F Dt, in. WC
100 0.09 400 0.53
200 0.29 500 0.62
300 0.43 600 0.69
2. Compute the weight of air that must enter
through the barometric damper to cool the
products of combustion from the multiple -
chamber incinerator to 300°F, heat losses being
neglected:
Although neglecting losses causes the damper
to be somewhat oversized, the draft can still
be regulated with the weights on the damper.
With the damper undersized, however, the
draft cannot always be controlled.
(WA)(V
-------
Flue-Fed Apartment Incinerators
459
where:
F = friction loss, in. WC
H = length of breeching, ft
V = velocity, fps
D = duct diameter, ft
T = temperature, degrees Rankine.
,2
B
(0.008)(10)(51.9)
(1)(760)
= 0.283 in. WC
8. Velocity through the flue:
Area of flue = 2. 25 ft
Velocity (300°F) =
4°" 73
= 18. 3 fps
2. 25 ft
9. Friction loss in the flue:
2*
= 0.002(H)(V)
~C^ [m \ I T"1 \
where:
F = friction loss, in. WC
H = height of flue above breeching, ft
V = velocity, fps
m = hydraulic radius, ft
T = temperature, degrees Rankine.
For rectangular cross section, the hydraulic
radius is:
wetted perimeter
For the given flue, the hydraulic radius is:
= (2.25ft )(12in./ft) =
/ A \ / ~i a • \ w.-'i—'j.L.
(4)(18 in. )
(0.002)(92)(18.3)
= 0.216 in. WC.
11. Frictional losses (calculatedby the same meth-
od) for assumed flue gas temperatures of 400 °
and 500 °F are given in the following tabulation:
Temp, °F
400
500
Friction loss, in. WC
0.28
0. 18
12. Determine the flue gas temperature:
A flue gas temperature of 380 °F, representing
a difference of 0.2 in. WC between the the-
oretical draft and the frictional losses, is ob-
tained from a plot of the data derived herein,
as shown in Figure 335.
0 100 200 300 400 500 BOO 700
TEMPERATURE °F
Figure 335. Draft at breeching of a multiple-chamber
basement installation versus average flue gas tem-
perature.
13. Weight of air entering through the barometric
damper at 375°F:
F (0.375)(760)
10. Total friction losses in breeching and flue:
Total friction losses = 0. 283 in. WC + 0. 216 in. WC
= 0.499 in. WC
|:Gr i swol d, 1 %6.
(WA)(V
-------
460
INCINERATION
C - average specific beat of products of com-
bustion from multiple-chamber incinera-
tor over temperature range of Tj to T£,
Btu/lb-°F
C = average specific heat of air over tem-
P perature range T2 to TA> Btu/lb-°F
T = final temperature of flue gases, °F
T = average temperature of gases from mul-
tiple-chamber incinerator, °F
T = temperature of air, °F.
(W lb/sec)(0. 24 Btu/lb-°F)(380 °F - 60°F) =
(0. 517 lb/sec)(0.26 Btu/lb-°F)(990°F - 380°F)
WA = 1. 07 Ib/sec
14. Volume of air entering through the barometric
damper:
Volume (60°F) =
(1. 07 lb/sec)(60 sec/min)(379 ft /mol)
29 Ib/mol
= 840cfm
15. Area of barometric damper:
The effective open area of a barometric damp-
er is about 70 percent of its cross-sectional
area. The areabasedonthe calculated amount
of air to be inspirated must, therefore, be
increased accordingly.
From Table D8, Appendix D, one velocity
head at 0. 2 in. WC and 60 °F is 1, 780 fpm.
. (840 cfm)(144 in. Z/ft2)
Area = ~ (l,780fpm)(0.70) = 9? ln"
16. Diameter of barometric damper:
,2
Area =
Diameter =
(7T)(Diameter)
,1/2
= 12 in. (sized
to the nearest
inch)
PATHOLOGICAL-WASTE INCINERATORS
Pathological waste is defined to include all, or parts
of, organs, bones, muscles, other tissues,and or-
ganic wastes of human or animal origin. This sec-
tion is limited to those incinerators used for the
burning of pathological wastes and to crematory
furnaces thathave design standards similar to those
of pathological waste incinerators.
Chemically, pathological waste is composed prin-
cipally of carbon, hydrogen, and oxygen. Slight
amounts of many minerals, along with a trace of
nitrogen, are also present. Physically, this waste
consists of cellular structuredmaterials and fluids.
Each cell contains water, along with the elements
and compounds forming the cell. The cells com-
prise thehair, fatty tissue, proteinaceous tissue,
andbone in proportions varying 'with different ani-
mals. Blood and various other fluids in the organs
are almost completely water.
The average chemical composition of whole ani-
mals, except for the proportion of water present,
is very similar in all animals. The proportion of
water present, compared with the total weight of
the animals , varies widely among different animals ,
and among various conditions of freshness or de-
composition of the animal material. Average chem-
ical properties of pathological waste and combus-
tion data are given in Table 125. These combus-
tion data have been found to provide good results
when used in design calculations for pathological-
waste incinerators. The cremation of human re-
mains differs from other pathological incineration
only in that the body is usually contained in a wood-
Table 125. CHEMICAL COMPOSITION OF
PATHOLOGICAL WASTE AND
COMBUSTION DATA
Ultimate analysis
(whole dead animal)
Constituent
Carbon
Hydrogen
Oxygen
Water
Nitrogen
Mineral (ash)
As charged
% by weight
14.7
2. 7
11. S
62. 1
Trace
9
Ash-free combustible
% by weight
50.80
9. 35
39.85
-
-
-
Dry combustible empirical formula -
Combustion data
(based on 1 Ib of dry ash-free combustible)
Theoretical air
40% sat
Flue gas with
theoretical
air 40%
saturated
at 60°F
CO2
^2
H^O formed
H^O air
Products of combustion total
Gross heat of combustion
Quantity
Ib
7. 028
7. 059
1. 858
5. 402
0. 763
0. 031
8. 054
Volume
scf
92. 40
93
16. 06
73. 24
15.99
0.63
105. 92
8, 820 Btu per Ib
-------
Pathological-Waste Incinerators
461
en casket. The casket must be considered when
designing these units and is usually assumed, for
design calculations, to weigh 75 pounds and to have
the chemical analyses and combustion properties
of average wood given in Table 118.
THE AIR POLLUTION PROBLEM
Pathological-waste incinerators can produce emis-
sions of fly ash, smoke, gases, and odors that
would be highly objectionable. Fly ash emission
is usually inconsequential in this type of incinera-
tor, but odor emissions may be very great. Vis-
ible smoke from this type of incinerator is highly
repugnant on esthetic grounds to most people and
is especially undesirable from crematory furnaces.
Poorly designed incinerators , with inadequate mix-
ing, temperatures, and residence times emit high-
ly objectionable air contaminants . Table 126 pre-
sents emission values measured for two separate
multiple -chamber pathological-waste incinerators
operating without secondary burners. These data
show emissions that are similar to a single cham-
ber design without a secondary chamber or after-
burner.
AIR POLLUTION CONTROL EQUIPMENT
The prevention of air contaminant emissions by
good equipment design is the best air pollution con-
trol procedure to follow. Inadequate equipment
may be compensated for by use of an afterburner
designed according to precepts setforthin the first
part of this chapter. New equipment employing
good design concepts can produce maximum com-
bustion of pathological-waste material with a min-
imum of air contaminant emissions.
Design Procedure
A principal consideration in the design of patho-
logical-waste incinerators is provision for the re-
lease of fluids as the material is destroyed. These
Table 126. EMISSIONS FROM TWO PATHOLOGICAL-WASTE
INCINERATORS WITHOUT SECONDARY BURNERS
(SOURCE TESTS OF TWO PATHOLOGICAL-
WASTE INCINERATORS)
Test No.
Rate of destruction to
powdery ash, Ib/hr
Type of waste
Combustion contaminants,
gr/scfa at 12% CO2
gr/scf
Ib/hr
Ib/ton charged
Organic acids,
gr/scf
Ib/hr
Ib/ton charged
Aldehydes,
gr/scf
Ib/hr
Ib/ton charged
Nitrogen oxides,
ppm
Ib/hr
Ib/ton charged
Hydrocarbons
549
Mixing chamber burner
not operating
26.4
Placental tissue in
newspaper at 40 °F
0. 500
0. 017
0. 030
2. 270
0.010
0. 020
1. 514
0. 007
0. 013
0. 985
14. 700
0. 016
1.210
Nil
563
Mixing chamber burner
not operating
107
Dogs freshly killed
0. 300
0. 128
0.430
8. 040
0.034
0. 110
2. 050
0.010
0. 033
0.617
95
0.082
1.550
Nil
aCO-, from burning of waste only used to convert to basis of 12% CO2-
See Rule 53b, incinerators, in Appendix A.
-------
462
INCINERATION
fluids are frequently released in such quantities
that they do not immediately evaporate and, hence,
require the use of a solid hearth rather than grates
in the ignition chamber. Pathological waste can-
not be considered as forming a. fuel bed when being
incinerated, and the passage of air through the
burning material is not a requirement in these
units.
The presence of a relatively high percentage of
moisture throughout each individual cell comprising
the pathological waste presents a difficult evapora-
tion problem. Evaporation of the moisture is nec-
essary before the combustible animal tissue can be
ignited. Moisture, however, evaporates only from
those cells upon and near the surface of the mate-
rial exposed to heat. Deeper lying tissue is al-
most completely insulated from the heat in the
chamber and is heated only slowly. Evaporation
of moisture from deeper cells cannot take place
until the destruction of the cellular material above
them causes them to be near the surface receiv-
ing heat. While the heat of combustion of the dry
cellular material is considerable, the relatively
small proportion of this material to the large amount
of moisture present makes it ineffectual in initiating
the evaporation processes. Auxiliary fuel inust be
burned to accomplish the necessary dehydration.
As with other incinerator design calculations, those
for pathological-waste incinerators also fall into
three general categories: (1) Combustion calcula-
tions, based upon the heat input of auxiliary fuel,
the composition of waste, the assumed require-
ments for air, and heat losses; (2) flow calcula-
tions based upon the products of combustion and
the expected gas temperatures; and (3) dimen-
sional calculations based upon simple mensura-
tion and empirical sizing equations. The factors
to be used in these calculations for pathological
incinerator design are given in Tables 127 and 128.
Simplifying assumptions may be made as follows:
1. The evaporation and burning rates, auxiliary-
fuel burning rate, and average waste composi-
tion are taken as constant. Design parameters
should be based upon that waste containing the
highest percentage of moisture that may be
expected to be destroyed in the unit.
2. The average temperature of the combustion
products is determined through calculation of
heat loss by using radiation and storage loss-
es as determined in Table 129.
3. The overall average gas temperature should
be about 1, 500°F when calculations are based
upon air for the combustible waste at 100 per-
cent in excess of theoretical, and upon air for
the primary burner at 20 percent in excess of
theoretical. The minimum temperature of the
gases leaving the ignition chamber should be
1,600°F.
4. Indraft velocity in the air ports is assumed to
beatO. 1 inch water column velocity pressure
(1, 255 fpm).
5. The secondary air port is sized to provide 100
per cent of the theoretical air for the combus-
tible material in the waste charged.
Table 127. DESIGN FACTORS FOR
PATHOLOGICAL IGNITION CHAMBER
(INCINERATOR CAPACITY,
25 Ib/hr TO 250 Ib/hr)
Item
Hearth loading
Hearth length-to-width ratio
Primary burner design
Arch height
Gross neat release--ignition chamber
Specific heat of the products of com-
bustion including combustion of
waste and natural gas
Recommended value
10 Ib/hr ft2
2
10 cf natural gas
Ib waste burned
See Figure 336
See Figure 336
0. 29 Btu/lb-°F
Allowable
deviation, %
+ 10
+ 20
i 10
+ 20
+ 20
-
Table 128. GAS VELOCITIES AND DRAFT
(PATHOLOGICAL INCINERATORS WITH HOT
GAS PASSAGE BELOW A SOLID HEARTH)
Item
Gas velocities,
Flame port at 1, 600°F, fps
Mixing chamber at 1,600°F, fps
Port at bottom of mixing chamber at
1, 550"F, fps
Chamber below hearth at 1, 500"F, fps
Port at bottom of combustion chamber
at 1, 500°F, fps
Combustion chamber at 1,400°F, fps
Stack at 1, 400°F, fps
Draft,
Combustion chamber, in. WC,
Ignition chamber, in. WC.
Recommended
values
20
20
20
10
20
5
20
0.25a
0. 05 to 0. 10
Allowable
deviation, %
-f 20
1 20
± 20
+ 100
+ 20
+ 100
+ 25
{^5
+ 0
aDraft can be 0. 20 in. WC for incinerators with a cold hearth.
A primary air port is not normally necessary in
these units. Sufficient combustion air in the pri-
mary chamber is normally provided for both the
fuel and waste material by the burner and by leak-
age at the charge door and other points in this
chamber. When a primary air port is desired, its
sizing should be based on the admission of 200 per-
cent of the theoretical air necessary for the com-
bustible waste material.
-------
Pathological-Waste Incinerators
463
Table 129. HEAT LOSSES FROM IGNITION
CHAMBER (STORAGE, CONVECTION, AND
RADIATION LOSSES DURING INITIAL
90 MINUTES OF PATHOLOGICAL
INCINERATION OPERATION)
Incinerator capacity,
Ib/hr
25
50
100
200
250
Loss expressed as
% of gross heat input
36. 3
32.8
29.75
25. 3
23. 6
The combustion calculations needed to determine
weights and velocities of the products of combus-
tion along with average temperatures may be de-
rived from standard calculation procedures when
the preceding assumptions, are followed. The siz-
ing requirements for inlet air areas are minimum;
these areas should be oversized in practice to pro-
vide for operational latitude.
Ignition chamber
Dimensions of the ignition chamber are determined
by deriving hearth loading and area, average arch
height, and chamber volume from Figure 336 and
from the factors given in Table 127. The input
capacity of the ignition chamber burner is also de-
termined from the factors given in Table 127.
36
32
28
24
20
50
100 150 200
CAPACITY, Ib/hr
250
300
Figure 336. Arch height of pathological-waste
ignition chamber.
Length-to-width ratios for the hearth are not crit-
ical. To provide for single-layer disposition of the
material upon the hearth, however, with the re-
sultant maximum exposure of the material to the
burner flame pattern, a length-to-width ratio of
2 to 1 is most practical.
Secondary combustion zone
The velocity parameters stated in Table 128 are
nottoo critical in these units. The relatively small
amount of combustible material in the waste does
not provide a problem too severe for achieving
complete combustion. Particulate discharge from
these incinerators has been found to be very light.
The principal design consideration is an effective
rate of destruction of the waste. Design consider-
ation must, however, be given to one peculiar
problem in the burning of this waste material:
Whenever a deposit of fatty tissue or hair is ex-
posed to flame or high-temperature gases.it quick-
ly volatilizes. The sudden volatilization of these
parts causes a flooding of gases and vapors that
would be beyond the capacity of the secondary com-
bustion zone designed on the basis of an average
rate of operation. These periods of sudden vola-
tilization then result in considerable amounts of un-
burned gases and vapors, which issue from the
stack as dense, visible smoke.
Design of the secondary combustion zone for low-
velocity gas movement at average volumes will
provide for complete combustion, even during the
periods of abnormally high combustion rates.
An auxiliary burner in the secondary combustion
zone is necessary for these incinerators. This
burner capacityneed only be sufficient to maintain
a 1, 600°F temperature in the gases. To do this,
the burner should be so located that the gases flow-
ing from the ignition chamber can first mix with
secondaryair before passing through the flame of
the secondary burner. Its location should also,
however, be such that sufficient residence time of
combustion gases is provided in the mixing cham-
ber, after passage through the flame, for secon-
dary combustion to occur.
Stack design
Calculations for stack design should be based up-
on a gas temperature of 1,400°F. Because design
calculations are based upon an average rate of
operation and because there will be periods when
this rate will be exceeded, stack design velocity
should be at or below 20 fps. Stack height should
be determined so as to provide a minimum avail-
able draft of 0, 20 inch water column at the breech-
ing. This is an absolute minimum draft provision
for all pathological-waste incinerators. When pas-
-------
464
INCINERATION
sage of hot gas beneath the hearth is to be provided,
the minimum, available stack draft at the breech-
ing should be designed for 0. 25 inch water column.
This higher draft will compensate for the addi-
tional resistance to gas flow in its passage beneath
the hot hearth.
Supplementary calculations
Piping requirements for the gas fuel supply
line should be determined. This sizing should
provide for supplying the total maximum capac-
ity of the burners used in both the ignition and
secondary combustion chambers.
Crematory design
The shape and size of the ignition chamber in cre-
matory units is dictated by the dimensions of a
casket. The same factors influencing the design
of other pathological-waste units should, however,
be used for all other parameters of the crematory
ignition chamber. In calculating the volume and
weight of products of combustion, consideration
mustbe given to the admission of somewhat larg-
er amounts of excess air when the design includes
a charge door at one end and a cleanout door at the
other end of the ignition chamber. Increasing the
burner capacity in these installations may be nec-
essary. Parameters for the secondary combus-
tion zone of crematory furnaces will be based upon
the same factors as those given for normal patho-
logical-waste incinerators. The volume and weight
of products of combustion will include those from
the burning of the casket. These units cannot, how-
ever, be designed on the basis of assuming that the
burning rate is constant. There will be some period
of time during the total operation in which a higher
rate of production of combustion products occurs.
Table 130 sets forth two possible operating pro-
cedures with arbitrary but representative grouping
of periods of operation that will produce varying
combustion rates. The factors for the parameters
of the secondary combustion zone should be used
for the period of operation that produces the great-
est flow of combustion products.
Incinerator design configuration
There are several possible configurations that
might be used in the construction of pathological-
waste incinerators. Several are illustrated in Fig-
ures 337, 338, and 339.
Table 130. OPERATING PROCEDURES FOR CREMATORY
Phase
Charging3"
Ignition
Full combustion
Final combustion
Calcining
Duration,
1-1/2 hr
operation,
rnin
-
15
30
45
1 to 12 hr
Burner settings
Secondary zone on
All on
All on
All on
All off (or small
primary on)
Casket
20% burns
80% burns
-
-
Moisture
-
20% evap
80% evap
-
Body
Tissue
-
10% burns
90% burns
-
Bone
Calcined
-
-
50%
50%
OR
Ignition
Full combustion
Final combustion
Calcining
Duration,
2-1/2 hr
operation,
min
15
30
15
90
1 to 12 hr
All on
Primary off
All on
All on
All off (small primary
may be on)
20% burns
60% burns
20% burns
-
-
20% evap
20% evap
60% evap
-
-
20% burns
80% burns
-
-
50%
-
Charge: Casket
Body
75 Ib wood
180 Ib
Moisture - 108 Ib
Tissue - 50 Ib
Bone - 22 Ib
-------
Pathological-Waste Incinerators
465
CHUHGIMG'
0001!
IGNITION CHAMBER
Figure 337. Multiple-chamber pathological-waste incinerator.
Illustrated in Figure 337 is an adaptation of the
designfor the retort-type multiple-chamber incin-
erator for destruction of pathological waste. In
this adaptation, three configuration differences
from the cornparable'unit illustrated in Figure 308
are immediately visible:
1. The use of a solid hearth instead of grates;
2. the provision for heating the hearth by passing
the products of combustion from the mixing
chamber through a chamber beneath the hearth
before they exit to the combustion chamber,
permitting both transfer of heat to the unex-
posed portions of the material lying on the
hearth, and more rapid evaporation of fluids
that spill upon the hearth or seep through it; and
3. the use of a side charging door.
Individual components in pathological waste are
frequently large. In addition, the charge must be
disposed over the hearth in a single layer of com-
ponents to provide for maximum exposure of sur-
face area to the burner flame. These two factors
make necessary the designing of the charge open-
ing with width and height dimensions close to the
maximum dimensions of the ignition chamber. The
side charging door will not, as -with the incinera-
tion of general refuse, cause the emission of ex-
cessive particulate matter from these incinerators.
Figure 338 illustrates a retort for the burning of
pathological waste added to a standard multiple-
chamber incinerator. When these retorts are used,
the gases from the retort should pass across the
rear of the ignition chamber of the standard in-
cinerator. The design of the retort incorporates
the factors given for the design of the ignition cham-
ber of a pathological-waste incinerator. The de-
sign of the remainder of the combination incinerator
is only slightly influenced by the addition of this
retort under most circumstances. This design
concept may be used only where the pathological-
waste load occurs periodically and in small amounts.
Figure 339 illustrates a human crematory retort.
This illustration is but one design, and many varia-
tions are found. Characteristically, these retorts
provide for a flame along the length of a shallow,
narrow, long charging chamber. The design illus-
trated employs a "hothearth. " Other designs pro-
vide for flame passage on all sides of the charge
including the underside. The hot hearth principle
is not always employed in crematory retorts. The
unit illustrated was not originally designed with
secondary burners; these burners were added in
the gas passageway below from the primary igni-
tion chamber at a later date to eliminate smoke.
Standards for Construction
The general discussions for the construction of
multiple-chamber incinerators given in previous
parts of this chapter cover most of the problems
found in constructing pathological-waste inciner-
ators. The use of extra-high-grade refractories
(super-duty fire brick or its equivalent plastic or
castable refractory) in these units is imperative.
Hearth construction must provide physical strength
at elevated temperature to sustain the maximum
loading possible. Characteristically, in patho-
-------
466
INCINERATION
MIXING-
CHAMBER \ FLAME PORT-
SECONDARY
COMBUSTION
CHAMBER
CLEANOUT
DOOR
IGNITION
CHAMBER
PATHOLOGICAL REFUSE
CHARGING DOOR
CLEANOUT DOOR KITH
UNDERGRATE AIR PORT
"-GENERAL REFUSE CHARGING DOOR
»ITH OVERFIRE AIR PORT
PATHOLOGICAL-
PRIMARY
BURNER PORT
Figure 338. Multiple-chamber incinerator with a pathological-waste retort.
logical-waste incinerators, the initial charge of
material on the hearth can have a total weight in
excess of the hourly capacity of the unit. When
making calculations for the strength of the hearth,
calculate the hearth loading at twice the combus-
tion rate, or more.
Stack Emissions
Visual emissions of fly ash are not evident from
pathological-waste incinerators. Air contami-
nants, as solid, liquid, and gaseous emissions , de-
termined for two typical units, are given in Table
131.
Operation
Operation of pathological-waste incinerators is, in
general, more simple than that for other types of
refuse incinerators. Preheating the secondary
combustion zone before charging and operating
these units is good practice. The primary burner
or burners should not be ignited until charging has
been completed and the charge door closed. The
material to be destroyed should be disposed on the
hearth in a manner that provides for maximum ex-
posure to the flame of the primary burner. Fur-
ther overcharging the unit by placing one compo-
nent of the charge on top of another is not good
practice. Careshouldbe exercised to ensure that
the primary burner port is not blocked by any ele-
ment of the charge.
When the amount of material to be destroyed ex-
ceeds what can be normally charged, stoking and
additional charging should be practiced only after
considerable reduction of the initial charge has oc-
curred. The primary burner should be shut off
before the charge door is opened and stoking or
additional charging takes place. Before an addi-
tional charge is made, the material remaining on
the hearth should be gently pushed towards the end
of the hearth nearest the flame port. The fresh
charge should then be disposed on the exposed
-------
Pathological-Waste Incinerators
467
hearth toward the primary burner. When recharg-
ing is complete, the charge door should first be
closedbefore the primary burner is once again ig-
nited.
Air port adjustment normally has only a minor role
in the regulation of the operation of these incinera-
tors. Making further adjustments to the secondary
air port after it has beenadjusted to provide proper
operation under normal burning conditions is usu-
ally not necessary. The only operating difficulty
to be encountered occurs when large deposits of
fatty tissue or hair are exposed to the burner flame.
As previously stated, the sudden volatilization of
this material occasions a sudden rush of gases and
vapors into the secondary chamber. On these oc-
casions some black smoke may issue from the stack.
This surge of gas volume, if very large, could even
result in pressurizing the ignition chamber, caus-
ing smoke to be forced out around the charge door.
Operational control, when this occurs, is obtained
by reducing the burner rate in the ignition chamber.
Under exceptional conditions, shutting this burner
off for a few minutes may even be necessary. White
PRIMARY
BURNER
IGNITION
CHAMBER
CLEANOU
DOORS
CIEANOUT
DOOR
Figure 339. Crematory retort.
Table 131. EMISSIONS FROM TWO PATHOLOGICAL-WASTE
INCINERATORS WITH SECONDARY BURNERS
Test No.
Rate of destruction to
powdery ash, Ib/hr
Type of \vaste
Combustion contaminants,
gr/scfa at 12% CO2
gr/scf
Ib/hr
Ib/ton charged
Organic acids,
gr/scf
Ib/hr
Ib/ton charged
Aldehydes ,
gr/scf
Ib/hr
Ib/ton charged
Nitrogen oxides,
ppm
Ib/hr
Ib/ton charged
Hydrocarbons
549
Mixing chamber burner
Operating
19.2
Placental tissue in
newspaper at 40°F
0. 200
0. 014
0. 030
3. 120
0. 006
0. 010
1. 040
N.A.b
N.A.b
N.A.b
42. 70
0. 08
8.84
Nil
563
Mixing chamber burner
Operating
99
Dogs freshly killed
0. 300
0. 936
0. 360
7. 260
0. 013
0. 050
1. 010
0, 006
0. 020
0. 400
131
0. 099
2
Nil
from burning of waste
Rule 53b, incinerators, in
bNot available.
used only to convert to basis of 12% CO2-
Appendix A.
See
-------
468
INCINERATION
smoke issuing from the stack usually indicates that
air is entering the unit in an amount exceeding the
ability of the burners to heat sufficiently. This is
best overcome by increasing secondary or primary
burner fuel rates. Very rarely is it necessary to
adjust the secondary air port to lower the admis-
sion of air when white smoke persists.
Automatic temperature control may be used to
operate these units. Temperature control should
be achieved by using the primary burner only. The
secondary burner should not be shut off or modu-
lated to a lower operating rate by these controls.
The temperature-sensing element may be placed
in the combustion chamber, breeching, or stack.
Precise temperature control at any of these points
is then achieved by modulating or shutting off the
primary burner. This operation to control tem-
peratures does not affect the emission of air con-
taminants. When temperature control is attempted
by control of the secondary burner, provision of
the response desired will be found difficult, and
the emissions of air contaminants will be increased
when the burner's rate of fire is reduced or shut
off by control action.
Solution:
1. Design features of ignition chamber:
From Table 127, hearth loading is 10 Ib/hr
per ft2.
Hearth area
(100 Ib/hr) -r (10 lb/hr-ft2) = 10 ft2
Hearth dimensions
Length-to-width ratio = 2
Let w = width of hearth in ft
(w)(2w) = hearth area
2w2 = 10 ft2
Length = 2w
2. 24 ft
4. 48 ft
From Figure 336, arch height = 25. 8 in.
3
Total ignition chamber volume = 21. 5 ft
There is no burndown period in the operation of
pathological-waste incinerators. The degree of
destruction desired for the waste material dictates
the length of time the primary burner is left in
operation. Some operations are normally ceased
whenthe material has been reduced to clean, white
bone. When reduction of the bone to powdery ash
is desired, the primary burners are continued in
ope ration until this is achieved. After the shutoff
of the primaryburner, the secondary burner should
not be shut off until smoldering from the residual
material on the hearth in the primary chamber has
ceased.
The hearth should be frequently cleaned to pre-
vent buildup of ash residue and slag-like deposits.
Frequency of cleanout of the combustion or settling
chamber depends upon incinerator use. Deposits
in this chamber shouldbe removed to avoid re-en-
trainment in the exhaust gases.
Illustrative Problem
Problem:
Design an incinerator to dispose of 100 pounds of
dog bodies per hour.
2. Capacity of primary burner:
From Table 127, primary burner consumptior
is 10 scf natural gas per Ib waste burned.
/IOscf natural gas\/Mp_lb_\ = c£h
\ lb waste burned ) \ hx )
3. Composition by weight of refuse:
Dry combustibles (100 lb/hr)(0. 29) = 29 Ib/h
Contained moisture (100 lb/hr)(0. 62) = 62 Ib/h
Ash (1001b/hr)(0. 09)= 9 Ib/hi
Total
4. Gross heat input:
100 Ib/hi
From Table 125, the gross heating value of
waste is 8, 820 Btu/lb and from Table D7 in
AppendixD, the gross heating value of natural
gas is 1, 100 Btu/scf.
Waste
(29 lb/hr)(8, 820 Btu/lb) =
256,000 Btu/hr
Select a multiple-chamber retort-type incinerator
with a hot-gas passage below a solid hearth.
Natural gas
(1, 000 cfh)(l, 100Btu/scf)= 1, 100,000 Btu/hr
Total 1, 356,000 Btu/hr
-------
Pathological-Waste Incinerators
469
5. Heat losses:
(a) From Table 129, gross heat losses by stor-
age, conduction, and radiation are 29, 75%
of gross heat input.
(0.2975)(1, 356, 000 Btu/hr) = 404, 000 Btu/hr
(b) Evaporation of contained moisture at 60 °F
The heat of vaporization of water at 60°F
is 1,060 Btu/lb
(62 lb/hr)(l, 060 Btu/lb) = 65, 700 Btu/hr
(c) Evaporation of water formed from combus-
tion of waste at 60°F
From Table 125, combustion of 1 Ib of waste
yields 0. 763 Ib of water.
(0. 763 lb/lb)(29 lb/hr)(l, 060 Btu/lb) =
23,450 Btu/hr
(d) Evaporation of water formed from combus-
tion of natural gas at 60°F
From Table D7, Appendix D, 0.099 Ib of
water is formed from combustion of 1 scf
of natural gas.
. 099 Ib water\
1 scf
(1,000 scfh)(l,060 Btu/lb)
105,000 Btu/hr
H = (9 lb/hr)(0.217 Btu/lb- °F)(1 , 600 °F - 60°F)
= 3, 000 Btu/hr
(f) Total heat losses
(a) + (b) + (c) + (d) + (e) = total heat losses
404, 000 Btu/hr + 65, 700 Btu/hr + 23, 450 Btu/hr
105, 000 Btu/hr + 3, 000 Btu/hr = 601, 150 Btu/hr
6. Net heat available to raise products of com-
bustion:
Gross heat input - heat losses = net heat avail-
able
1, 356, 000 Btu/hr - 601, 150 Btu/hr = 754, 850
Btu/hr
7. Weight of products of combustion:
From Table 125, combustion of 1 Ib waste with
100% excess air will yield 15. 113 Ib of com-
bustion products.
From Table D7, Appendix D, combustion of
1 scf natural gas with 20% excess air will yield
0, 999 Ib of combustion products.
Waste (29 lb/hr)(15. 133 Ib/lb) = 438 Ib/hr
Contained moisture = 62 Ib/hr
Natural gas
(1,000 cfh)(0. 999 Ib/scf) = 999 Ib/hr
Total weight of combustion
products 1,499 Ib/hr
(e) Sensible heat in ash
Assume ash is equivalent in composition to
calcium carbonate.
Average specific heat is 0.217 Btu/lb-°F
H = W
A
' V
where
H = sensible heat, Btu/hr
W = weight of ash, Ib/hr
C = average specific heat of ash,
P Btu/lb- °F
T = final temperature, °F
T = initial temperature, °F
Average gas temperature:
Assurnethe average specific heat of combus-
tion products is 0.29 Btu/lb-°F
Q = (W)(C )(T - T )
p 2 1
where
Q = net heat available, Btu/hr
W = weight of combustion products, Ib/hr
C = average specific heat of combustion
products, Btu/lb-°F
_T = average gas temperature, °F
T = initial temperature, °F.
T
Q
= T + -
1 (W )(C )
-------
470
INCINERATION
= 60 +
754,850
(1,499)(0.29)
= 1,740°F
This average temperature exceeds minimum
design temperature of 1, 600 °F. The primary
burner has, therefore, adequate capacity.
9. Secondary air port size:
Design secondary air port to supply theoretical
air for combustion of waste.
Design secondary air port 100% oversize with
adraftof 0.10 in. WC. From Table D8, Ap-
pendix D, 0. 10 in. WC is 1, 255 fpm.
From Table 125, lib of waste requires 93 scf
of air.
(29 lb/hr)(93 scf/lb) = 2, 697 cfh
or 44. 93 cfm
or 0. 749 cfs
(44.93 c£m)(144 in.2/ft2J_
1, 255 fpm
(2) = 10. 3 in.
10. Weight of maximum air through secondary port:
From Table Dl, Appendix D, the density of air
is 0. 0763 Ib/scf.
(2)(2, 697 cfh)(0. 0763 Ib/scf) = 4ll.51b/hr
11. Heat required to raise maximum secondary
air from 60° tol,600°F:
From Table D4, Appendix D, 399.6 Btu is
required toraisel Ib air from 60° to 1,600°F.
(411.5 lb/hr)(399.6 Btu/lb) = 164, 400 Btu/hr
Waste (29 lb/hr)(l 98. 92 scf/lb) = 5, 769 scfh
SZ 1WM
>h
Natural gas (1,000 scfh)(^' 5J Scf 1= 13,530 scfh
Total volume of gases
20, 604 scfh
or 344 scfm
or 5.75 scfs
(b) Through exit from mixing chamber
Design secondary burner for combustion
at 20% excess air.
Products of combustior
through flame port =
Products of combustion
from secondary burner
(298 cfh)(13. 53 scf/scf) =
Maximum air through
secondary air port
(2)(2, 697 scfh)
or
or
20, 604 scfh
4,030 scfh
5, 394 scfh
30,028 scfh
500 scfm
8.33 scfs
14. Incinerator cross -sectional areas:
(a) Flame port area
Design flame port for 20 fps velocity at
1, 600 °F
12. Natural gas required by secondary burner:
Design for combustion of natural gas with 20%
excess air. From Table D7, Appendix D,
the calorific value of natural gas is 552.9
Btu/scf at 1, 600°F.
(164,400 Btu/hr) -r (552.9 Btu/scf) = 298 cfh
13. Volume of products of combustion:
(a) Through flame port
From Table 125, combustion of 1 Ib waste
with 100% excess air will yield 198. 92 scf
of combustion products.
From Table D7, Appendix D, combustion
of 1 scf natural gas with 20% excess air will
yield 13. 477 scf of combustion products.
(b) Mixing chamber area
Design mixing chamber for 20 fps velocity
at 1, 600°F
3. 33 scfs)(2, 060°F)
Area =
(20 fps)(520°R)
= 1.65ft
(c) Port area at bottom of mixing chamber
Design port for 20 fps velocity at 1, 550°F
(20 fps)(520°R)
(8.33 scfs)(2,010°R) ,
Area = ,„ „ ,—;',„^ „ „„- = 1. 61 ft
(d) Chamber area beneath hearth
Design chamber for 10 fps velocity at
1, 500°F
-------
Debonding of Brakeshoes and Reclamation of Electrical Equipment Windings
471
(8.338cfs)(1.960°R) _ 2
(10fps)(520°R) ~ >14 "
( e) Port at bottom of combustion chamber
Design port for 20 fps velocity at 1, 500°F
2
(8.33 scfa)(1.960°R)
- (^fps)(520°R)
ft
(f) Combustion chamber
Design combustion chamber for 5 fps ve-
locity at 1, 400 °F
(8.33 scfs)(l,8bO°R) 2
Area = - wC7n°p\ - = 5- 95 ft
(5 fps)(520 R)
(g) Stack
Design stack for 20 fps velocity at 1,400°F
Area _ (8.33Scfs)(l,860°F) _ 2
Area ~ (20fps)(520°R) ' 1>49ft
15. Stack height:
Design stack for an effective draft of 0.25 in.
WC in the combustion chamber. Assume fric-
tion losses are 12. 5% of theoretical draft at
20 fps and 1, 400 °F
Theoretical draft required
0. 25 in. WC
1 - 0.125
= 0.286 in. WC
Stack height
D = 0. 42 PH
where
HJ
D = theoretical draft, in. WC
T = ambient air temperature, °R
T = average stack gas temperature, °R
P = atmospheric pressure, Ib/in.
H = stack height, ft
0. 286
H =
(0.42)(14. 7),
-J_ 1 \
520 " 1,860/
= 33.4ft
Kent, 193b.
OEBONDING OF BRAKE SHOES AND RECLAMA-
TION OF ELECTRICAL EQUIPMENT WINDINGS
Brake shoe debonding and reclamation of electrical
equipment windings are similar combustion pro-
cesses, both using equipment nearly identical in
design. These processes differ from incineration
and other combustion reclamation processes in that
the combustible contents of the charge are usually
less than 10 percent by weight, and high tempera-
tures must be avoided to prevent damaging the sal-
vageable parts.
DEBONDING OF BRAKE SHOES
Bonded brake linings contain asbestos mixed with
binders consisting of phenolic resins, synthetic
rubber, or bodied oils such as dehydrated linseed
oil. Carbon black, graphite, metallic lead, thin
brass strips, and cashew nut shell oil added in
small amounts act as friction-modifying agents
(Kirk and Othmer, 1947). These materials are
blended and extruded into curved lining to fit the
brake shoe. The lining is then heated to produce
a. hard surface.
Adhesives for bonding the lining are composed
mostly of rubber or phenolic resins. Small amounts
of vinyl are sometimes combined with phenolic
resins. The linings are originally bonded to steel
shoes with adhesive, of a thickness of 0. 008 to 0.01
inch, by subjecting them to pressure and a temper-
ature of 400°F for a specific time to develop max-
imum bond strength.
In the brake-debonding process, brake shoes are
charged to an oven, called a debonder, to which
external heat is applied carefully to minimize warp-
age of the shoes. Adhesive portions of the lining
start to melt, and destructive distillation begins
at about 600°F (Kirk and Othmer, 1947). In the
absence of flame, the melting of the adhesive pro-
ceeds until enough organic material is volatilized
to initiate burning. At 800"F, thermal debonding
results in the adhesive's being burned or charred.
Burning continues at temperatures less than 1,000 °F
until all combustibles have been consumed. Once
combustion has been initiated, the heating value of
the adhesive is usually sufficient to maintain burn-
ing without external heat.
After the brake shoes are removed from the de-
bonder, the brittle linings either fall from the shoes
or are knocked loose by light tapping. Carbonized
material adhering to the shoes is removed by abra-
sive blasting, and the clean shoes are ready for
bonding with new linings.
-------
472
INCINERATION
RECLAMATION OF ELECTRICAL EQUIPMENT WINDING
A major portion of the reclamation of direct-cur-
rent electrical equipment involves automotive start-
ers and generators. An average-size starter or
generator weighs 20 pounds and contains approxi-
mately 2 or 3 pounds of salvageable copper wire.
Reclamation of alternating-current electrical equip-
ment usually involves squirrel cage motors . Rotors
removed from squirrel cage motors contain no or-
ganic material and, therefore, require no process-
ing. The starters of these motors contain 5 to 10
percent combustible organic materials.
Table 132 gives the average composition and the
amount of combustibles in major components of
electrical equipment. While these data still hold
true today, trends in new construction point to the
use of greater quantities of noncombustible glass
cloth in place of cambric. Acrylic resin, epoxy
resin, silicone elastomers, and polyvinyl chloride
are replacing cambric installation and varnish
coatings.
In rebuilding electrical equipment and reclaiming
copper windings, the insulation is burned from the
windings of motors, generators, and transformers.
After combustion is completed, the copper wire
windings are separated and sold for scrap. Pole
pieces, shafts, frames, and other parts are cleaned
of char and rewound with new wire. During the
reclamation process, combustible organic com-
pounds used to insulate copper wire begin to vola-
tilize upon application of heat. Ignition occurs
above 600°F, and combustion is virtually completed
at 900°F. Since the combustible contents of the
charge are usually insufficient to sustain burning,
auxiliary heat is usually supplied by primary burn-
ers during the complete operation. By restricting
the combustion air, the burning insulation may pro-
vide over 50 percent of the total process heat re-
quirements.
The temperature in the furnace is kept below
1, 000°F to minimize warpage of metal parts and
oxidation of the copper wire. The larger the in-
dividual item, the longer the preheat time, to pre-
vent warping of the steel components. For example,
a 100-hp motor requires a preheat time of over
1-1/2 hours.
THE AIR POLLUTION PROBLEM
The practice of reclaiming electrical windings and
debonding brakeshoes by open burning or burning
Table 132. COMBUSTIBLE CONTENT OF ELECTRICAL EQUIPMENT
Components
A-C industrial
Motor and generators
Casing
Stator
Squirrel cage rotor
Wound rotor
D-C industrial
Motors and generators
Casing
Armature
Field rings
Automotive
Starters and generators
Casing
Armatures with shaft
Generator field coils
Starter field coils
Transformers
Casing
Windings
Average wt %
combustible
Nil
5 to 10
Nil
5 to 10
Nil
5 to 10
5 to 7
Nil
1 to 2
5 to 10
5 to 10
Nil
7 to 10
Combustible description
Cambric and varnish
Cambric and varnish
Cambric and varnish
Cambric and varnish, or
acrylic resin, epoxy resin,
silicones, PVC
Wood strips
Fish paper
Cambric and varnish
A'arnish
Fish paper
Cambric and varnish
Oila
lOil-filled transformers only.
-------
Debonding ot Brakeshoes and Reclamation of Electrical Equipment Windings
473
in a single-chamber device results in the emission
of large quantities of smoke, odors, and other
combustion contaminants. Emissions of air con-
taminants from these processes are summarized
in Tables 133 and 134.
Table 133. STACK EMISSIONS FROM
BRAKESHOE DEBONDING IN SINGLE
CHAMBERS WITHOUT CONTROLS
Composition of charge
Charge weight, Ib
Duration of toil, mm
Combustible content of charge, wt %
Sta k gas flow rate, sefm
Av rage #as temperature, CF
Pa iculatt matter, gr/scf at IZ1^, CO,
Pa iculate matter, Ib/hr
Sul ur dioxide, !b/hr
Ca )on moiiovide, Ib/hr
Or anic acids as aietu acid Ib/hr
Alt hydc-s as lorma] debydc. Ib/hr
\il ogen oxides as \Oi, Ib/hr
Hyr roc arbons as hexane, Ib/hr
Smoke emissions, opacity range
Ringelmann chart
C-606
175 shoes
Z65
38
5
380
ZOO
Z. 6
0 70
0. Z4
0.54
0 ZZ
0 10
0. OZ
0. 05
0 to 80% brown-white
C-651
60 shoes
--
8. 75
5
150
600
0. 5
0. 75
a
°
<*
a
a
0 to 10% bro\\n
nplt-d.
AIR POLLUTION CONTROL EQUIPMENT
Debonding of brake shoes and reclamation of elec-
trical windings conducted in a single-chamber unit
can be easily controlled by using an afterburner
as described in the first part of Chapter 5.
Two basic configurations of equipment using after-
burners effectively accomplish these reclamation
processes with a minimum discharge of air con-
taminants. One is a single structure housing the
primary and secondary combustion chambers, while
the other consists of two separate pieces of equip-
ment, a primary chamber and an afterburner or
secondary chamber. Variations in the design of
these two configurations are many, and the final
selection of a particular design is based upon con-
siderations such as space limitation, process con-
ditions, maintenance, capital investment, and
operating expenses. In designing an effective af-
terburner, the size and appurtenances of the pri-
mary ignition chamber must be known or be initial-
ly designed.
Primary Ignition Chamber
The size of the primary chamber is determined
from the production rate or volume of the batch
charge desired. On the average, 1 cubic foot of
space holds in random arrangement 27 brake shoes,
or 34 automotive field coils, or 10 automotive ar-
matures. In siting the primary chamber, addi-
tional space is provided over that space occupied
by the charge, to make it easier to load and un-
load. For example, in a batch process, 350 auto-
motive generator field coils or 200 average-size
brake shoes canbe randomly placed in a 55-gallon
drum.
Table 134. STACK EMISSIONS FROM RECLAIMING ELECTRICAL WINDINGS IN
SINGLE CHAMBERS WITHOUT CONTROLS
Item
Composition of charge
Charge weight, Ib
Duration of test, min
Combustible content of charge, wt %
Stack gas flow rate, scfm
Average gas temperature, °F
Particulate matter, gr/scf at 12% CO2
Particulate matter, Ib/hr
Sulfur dioxide, Ib/hr
Carbon monoxide, Ib/hr
Organic acids as acetic acid, Ib/hr
Aldehydes as formaldehyde, Ib/hr
Nitrogen oxides as N©2 , Ib/hr
Hydrocarbons as hexane, Ib/hr
Smoke emissions, opacity range
Ringelmann chart
Odors
Test No.
C-342
100-hp
generator
stator
--
22. 5a
5
320
680
1.9
2.43
0. 13
1. 90
0.35
0. 08
3. 07
Nil
15 to 30%
C-497
14 pole
pieces
3, 825
60
5
400
350
1. 1
0. 65
--
0. 35
0. 33
0. 079
--
--
--
C-542
200 auto
armatures
161
55
1.7
210
360
3. 3
1. 64
0
0. 50
0. 62
0. 29
0. 03
0. 16
0 to 100%
C-541-1
Auto
armatures
1, 034
45. 4
1.2
790
470
0. 54
1.04
0. 02
1.39
0. 42
0. 13
0. 12
0. 09
0 to 30%
white
C-541-3
Auto field
coils
356
16
5.9
950
290
1. 33
2. 51
0. 13
4.72
1. 01
0. 49
0. 08
0. 11
0 to 80%
gray
aTest duration does not include preheat period.
-------
474
INCINERATION
Primary burner capacity is computed by conven-
tional heat and material balances to determine the
amount of heat necessary to raise the tempera-
ture of the mate rial being processed to 850 °F. This
temperature ensures ignition of combustibles, and
maintenance of the temperature necessary for com-
plete combustion. Gas burners must supply suf-
ficient heat not only for ignition, but also to sus-
tain burning. The lower the combustible content
of the charge, the more heat that must be supplied
by the prirruiry burners. Consequently, primary
burners are sized for minimum combustible con-
tent of the charge.
Adjustable air ports near the bottom of the primary
chamber should be large enough to provide theoret-
ical air plus 100 percent excess air. These ports
should be sized to provide this quantity of air for
the maximum combustible content of the charge.
Secondary Combustion Chamber
The mixing chamber or afterburner is designed for
maximum effluent from the primary chamber using
conventional heat and material balances. For a
given charge, maximum effluent occurs when the
combustible content of the charge is at a maximum.
The mixing chamber burner or afterburner must
be capable of raising the temperature of the max-
imum quantity of effluent expected from a tempera-
ture of 850° tol,400°F. These burners are posi-
tioned to blanket the cross-sectional area of the
afterburner completely with flame.
The cross-sectional area of the mixing chamber is
based upon an average gas velocity ranging from
20 to 30 fps for the total effluent. Gas velocities
in this range promote turbulent mixing of the gas-
eous effluent from the ignition chamber with the
flames from the mixing chamber burner. Baffles
and abrupt changes in direction of gas flow also
promote turbulent mixing, which is essential lor
complete combustion. The mixing chamber or af-
terburner should be of sufficient length to allow a
residence time of at least 0. 15 to 0. 2 second.
Secondary air ports should provide theoretical air
for maximum combustible content of the charge.
Stack
In designing a stack for minimum height, stack
gas velocities should not exceed 20 fps at maximum
temperatures to minimize the effects of friction.
Effective draft is computed as theoretical stack
draft minus friction losses at design flow condi-
tions. An effective draft or negative static pres-
sure of from 0. 05 to 0. 10 inch WC should be avail-
able in the ignition chamber when the unit is oper-
ating at rated capacity.
Emissions
Stack emissions from brake-debonding and recla-
mation equipment using secondary combustion are
listed in Table 135. Note that, in all cases, the
carbon monoxide has been eliminated, and the par-
ticulate matter reduced by approximately 90 per-
cent 'when compared with emissions from uncon-
trolled units cited in Table 134.
Typical Reclamation Equipment
The multiple-chamber incinerator previously dis-
cussed in the first two parts of this chapter can be
adapted for these processes. Figure 340 shows
an incinerator of this kind that differs from a stan-
dard multiple-chamber incinerator by its oversize
ignition chamber and the absence of the grates and
ashpits. The third chamber (the combustion cham-
ber) is less useful because there is little or no fly
ash to be removed from the gas stream. The com-
bustion chamber does, however, complete the sec-
ondary combustion process and protect the stack
lining from direct flame impingements.
Primary burners are of the atmospheric type and
canbe mounted through the sides and at the bottom
of the primary chamber. An alternative arrange-
ment consists of dual-pipe burners placed across
the base of the primary chamber, which results in
more even distribution of heat over the cross sec-
tion of this chamber. These burners must, of
course, be positioned so that there will be no inter-
ference when the racks containing the charge of
material are inserted or removed. Mixing cham-
ber burners are located in the same position as
shown for a standard multiple-chamber incinerator.
Air ports are also similar in construction and loca-
tion to thos e mounted on a standard multiple-cham-
ber incinerator.
Another satisfactory single-structure design con-
sists of only two refractory-lined chambers, as
illustrated in Figure 341. It differs from the con-
ventional three-chamber unit already described
only in that the third chamber has been eliminated.
A relatively simple design using a separate primary
chamber and afterburner is shown in Figure 342.
The primary chamber consists of a tubular frame
with sheet metal siding. Drilied-pipe gas burners
are installed in the bottom of the chamber. The
material to be reclaimed is placed in a 55-gallon
drum with a perforated bottom, and the drum is
placed in the primary chamber. The contents of
the drum are heated and ignited by the pipe burn-
ers, and the hot gases and smoke flow to an after-
burner mounted on top of the primary chamber.
Heat is supplied to the afterburner by a fan-air
burner firing tangentially into the refractory-lined
chamber. This equipment is usually equipped with
-------
Debonding of Brakeshoes and Reclamation of Electrical Equipment Windings
475
Table 135. STACK EMISSIONS FROM DEBONDERS AND RECLAMATION EQUIPMENT
USING SECONDARY COMBUSTION
Item
Equipment description
Composition of charge
Charge weight, Ib
Combustible content, wt %
Duration of test, min
Stack gas flow rate, scfm
Stack gas temperature, °F
Secondary afterburner temperature, °F
Participate matter, gr/scf at 12% CO^
Particulate matter, Ib/hr
Sulfur dioxide, Ib/hr
Carbon monoxide, Ib/hr
Organic acids as acetic acid, Ib/hr
Aldehydes as formaldehyde, Ib/hr
Nitrogen oxides as NO^>, Ib/hr
Hydrocarbons as hexane, Ib/hr
Smoke emissions, opacity range
Rineelmanii chart
Test No.
C-286
Dual -chamber
brake debonder
480 brake shoes
--
--
30
181
999
0.24
0. 12
0. 12
0
0. 08
--
--
--
0
C-541-4
M-C incinerator
Auto field coils
386
6. 7
8
990
1, 340
0. 04
0. 37
0
0
0. 90
0. 08
0. 30
0.23
0
C-497
Oven with afterburner
(14 generator pole
pieces)
3, 825
60
950
1, 250
2, 000
0. 016
0. 059
0
0. 079
0
StCONQHR
COMBUSTION
CHAMBER
IGNITION CH1BBE
CLEANOUT 0001
CURTAIN IALL PORT
Figure 340. Multiple-chamber incinerator adapted for use in reclamation processes (see Figure 308).
a stack (hat is 16 to 20 Icet above ground level.
Since good heat control is difficult to maintain, this
equipment is more suitable for brakcshoe deboiid-
ing than for electrical-winding and armature core
reclamation.
In some cases, an oven vented to an afterburner
is used. This oven differs from the refractory-
lined primary chamber in that there is no direct
flame contact with the charge, and the hot combus-
tion gases arc' recirculated within the chambers for
-------
476
INCINERATION
Standards for Construction
Materials and methods of construction are similar
to those used for multiple-chamber incinerators,
as described in the second part of this chapter.
Exterior shells are constructed of 12-gage-min-
imum-thickness steel plates properly placed and
supportedby external structure members. Block
insulation with a minimum thickness of 2 inches
anda service temperature of 1, 000°F is normally
used between the steel shell cind the refractory lin-
ing to conserve heat and protect the operator. High-
heat-duty firebrick with adequate expansion joints
is used for lining the primary chamber as well as
the secondary chamber or afterburner. Stacks
are constructed of 10-gage steel plates and are
lined with 2 inches of insulating firebrick or cast-
able refractory having a minimum service tem-
perature of 2,000°F.
Figure 341. Dual-chamber reclamation furnace (Auto
Parts Exchange, City of Industry, Calif.).
more precise heat conservation and control. The
installation of a cam-ope rated temperature con-
troller makes possible a gradual elevation of
primary-chamber temperature and an exact con-
trol of temperature over extended periods of time.
This type of control is widely used for processing
electrical windings from motors and generators
where warpage of the laminations is to be avoided.
This type of reclamation equipment lends itself to
either the batch or continuous process.
A continuous-process device is shown in Figure
343; it consists of an endless-chain conveyor that
transports material into a tunnel-like chamber.
The products of combustion, smoke, and volatile
components are collected near the center of the
tunnel and vented to an afterburner. Asbestos cur-
tains are installed where the parts enter and leave
the chamber; they conserve heat by reducing the
induction of air. Continuous-process equipment of
this type usually has a higher heat requirement than
corresponding batch equipment does because of the
induction of excessive air at the openings to the
primary chamber.
Illustrative Problem
Problem:
Designbatch equipment to debond 200 average-size
brake shoes or 175 average-size automobile gener-
ator field coils --each batch -will require a 30 -minute
period.
Solution:
1. Ignition cham.be r dimensions:
200 average-size brake shoes -weigh 350 Ib
175 average-size field coils weigh 350 Ib
Average bulk density of brake shoes is 27
units /ft3
Average bulk density of field coils is 34
units /ft3
Brake shoes with 25% free volume
= 10ft
Field coils with 50% free volume
(175 coils) (-AJL-Y-L-) = loft
V 34 coils/\0. 5/
Another unique design, which can be used for semi-
continuous operation, consists of two refractory-
lined compartments connected back to back. While
material is being processed in one of the compart-
ments, the other compartment is being unloaded
and reloaded. Again, the smoke and gaseous ef-
fluents are vented to a vertical afterburner and
stack.
Primary-chamber dimensions
2 ft wide x 3 ft high x 1 ft 8 in. deep
2. Design capacity of primary gas burner:
Design for minimum combustibles content of
3% by weight.
-------
Debonding of Brakeshoes and Reclamation of Electrical Equipment Windings
477
Figure 342, Brake debonding in a 55-gallon drum venting to an afterburner: (1) Drum holding
brakeshoes, (2) secondary combustion chamber, (3) secondary burner (afterburner), (4) pri-
mary burner--pipe type, (5) stack (Gnggs Specialty Products, Huntington Park, Calif.).
(a) Heat required to raise temperature of charge
from 60° to 900°F. Neglect moisture in
charge:
Average specific heat of brakeshoes or auto-
mobile generators is 0.21 Btu/lb-°F
Q = (W)(Cp)(T2 - Tj)
where
Q = heat required, Btu/hr
W =
Cp =
weight of charge, Ib/hr
average specific heat of charge, Btu/lb-
°F
final temperature, °F
initial temperature, °F
(0.97)
/Z charges\ / 350 lb\
\ hr / \ charge J
hr ) \ charge
(0.21 Btu/lb-0F)(900°F - 60°F)
-------
478
INCINERATION
n., , ••«
Figure 343. Continuous brakeshoe debonder (Wagner Electric Corp., El Segundo, Calif.).
Q = 119,800 Btu/hr
(b) Heat required to raise products of combus-
tion from 60° to 900°F:
Assume combustibles have a composition
equivalent to PS-400 fuel oil. Design for
200% excess air, 40% saturated. From
Table D6, Appendix D, products of com-
bustion weigh 41. 47 Ib from combustion of
1 Ib combustible (PS-400 fuel). Average
specific heat of products of combustion is
0. 26 Btu/lb-°F.
Weight of products of combustion, W:
Heat required
W = (0.
^charge
= 870 Ib/hr
2 charges\/41.47 lb
~
Q = (W)(Cp)(T2 -
"where
Q = heat required, Btu/hr
W = weight of products of combustion,
Ib/hr
Cp = average specific heat of products of
combustion, Btu/lb-°F
T9 = final temperature, °F
T = initial temperature, "F
Q = (870 lb/hr)(0. 26 Btu/lb-°F)(900 °F -60°F)
= 190,000 Btu/hr
-------
Debonding of Brakeshoes and Reclamation of Electrical Equipment Windings
479
(c) Net heat required for process:
(a) + (b) = Total net heat
119,800 Btu/hr -t- 190, 000 Btu/hr = 309,800 Btu/hr
(d) Gross heat required for process:
Assume radiation, convection, and storage
heat losses are 30% of gross heat input. Net
heat available for process is 70% of gross
heat input.
309, 800 Btu/hr
0. 70
= 442, 000 Btu/hr
(a) Primary air port:
Maximum airflow.
Port size
/ 212 cfm \
ize = I • II
\1, 255 fpm /
or
212 cfm
212 Cfm \(2) = 0.338ft2
(b) Secondary air port:
Maximum airflow.
or 48.6 in. '
(e) Heat supplied by combustibles in charge:
From Table D5, Appendix D, the gross
heat of combustion from 1 Ib combustible
(PS-400 fuel oil) is 18,000 Btu/lb
(0.
\
Hchar geM hr
= 378, 000 Btu/hr
fitu/
(f) Net heat required in primary chamber:
442,000 Btu/hr - 378,000 Btu/hr = 64,000 Btu/hr
(g) Primary burner capacity:
From Table D7, Appendix D, the calorific
value of natural gas is 765. 3 Btu/scf at
900°F.
64,OOP Btu/hr
765. 3 Btu/ft3 =
3. Size of combustion air ports:
Design all port areas 100% oversize.
Assume 100% excess air through the primary
air port and theoretical air through the secon-
dary port.
hr
Ib
J
Port size =
103
cfm \
fpm/
(2) = 0. 1640 ft
23.5 in.
or 103 cfm
2
4. Design capacity of secondary burner (after-
burner):
Design for a maximum combustible content of
charge of 5% by weight.
(a) Maximum products of combustion with no
secondary air:
Weight of products of combustion of natural
gas with 20% excess air is 0. 999 Ib/scf.
hr
(84 cfh natural gas)(0. 999 Ib/scf)
= 1,460 Ib/hr
__ 84 Ib/hr
1, 544 Ib/hr
(b) Heat required to raise products of combus-
tion from 900° to 1,400°F:
Q = W C (T_ - T) see item 2(b).
p 2 1
Designfora maximum combustible content of
the charge of 5% by weight. Assume the draft Q= (i, 544 lb/hr)(0. 26 Btu/lb-°F)(1, 400 °F - 900 °F)
at all ports is 0. 10 in. WC. From Table D8,
Appendix D, 0.10 in. WC is 1,255 fpm Q = 201, 000 Btu/hr
From Table D6, Appendix D, 363 scf of com-
bustion air is required for combustion of 1 Ib
(PS-400 fuel oil) at 100% excess air, and 177
scf of combustion air is required for combustion
of 1 Ib (PS-400 fuel oil) at theoretical air.
(c) Burner capacity:
From Table D7, Appendix D, the calorific
value of natural gas at 1,400 °F is 616 Btu/scf.
-------
480
INCINERATION
201, OOP Btu/hr
616 Btu/scf
= 326 cfh
5. Size of mixing chamber (afterburner):
From Tables D6 and D7, Appendix D, there
are 540 scf of products from combustion of
1 Ib combustible (PS-400 fuel oil) at 200% ex-
cess air, and 13. 53 scf of products of com-
bustion from 1 scf natural gas at 20% excess
air.
(a) Cross-sectional area of inlet duct:
Design for gas flow of 20 fps at 900°F
Gas flow at 60 °F
Combustibles at 200% excess air
(0.'
hr
= 18,900 cfh
Natural gas at 20% excess air
(84 cfh.)(13.53 ft3/ft3) = 1, 133 cfh
20, 023 cfh
or 333 cfm
or 5. 55 cfs
Cross-sectional area
= 0.728 ft
2
(b) Cross-sectional area of mixing chamber
(afterburner):
Design for gas flow of 25 fps at 1, 400°F
Gas flow at 60"F
Combustibles from primary
chamber 20, 055 cfh
Secondary gas burner
(326 cfh)(13.53 ft3/ft3)
or
or
4,410 cfh
24,465 cfh
408 cfm
6.8 cfs
Cross-sectional area
(c) Length of mixing chamber (afte-burner):
Design for residence time of 0. 15 second
Length = (25 fps x 0. 15 second) = 3.75 ft
6. Stack diameter:
Design for a gas velocity of 20 fps at 1, 200°F
Cros s -sectional area
Stack diameter 13. 9 in.
Select 14-inch diameter.
7. Stack height:
(a) Theoretical draft for a 10-ft section at
1,ZOO°F:
= 0.52 PH /—
\ **
" V
where
D = theoretical draft, in. WC
P = atmospheric pressure, Ib/in. absolute
H = stack height, ft
T = temperature of stack gases, °R
T = temperature of air, °R
= (0.52)(14.
= (76.5)
D = 0. 101 in. WC
(— l \
\E>20 ~ l,66oy
(0. 00192 - 0. 00060)
(b) Stack friction for a 10-ft section at 1, 200°F:
(0. 008)(H)(V)2 '
_b =
where
F = friction, in. WC
H = stack height, ft
Kent,
!rjM5Wold,
-------
Drum Reclamation Furnaces
481
V = velocity, fps
D = stack diameter, ft
T = absolute stack temperature, °R
_ (0.008)(10)(20)2
(1.25)(1,660)
(c) Net effective draft for a 10-ft section:
(a) - (b) = Net draft
0. 101 in. WC - 0. 015 in. WC = 0. 086 in. WC
(d) Ignition chamber:
Assume static pressure of 0. 05 in. WC
(e) Friction loss in secondary-combustion
zone:
(1) Contraction loss into secondary zone:
Assume 0. 5 VP loss at 20 fps and 900 °F
Gas velocity 20 fps at 900 °F
where
V
T
h
gas velocity, fps
absolute temperature, °R
static pressure, in. WC
2
h = 0. 035 in. WC
Contraction lossf-°' °35'™' WCj (0. 5 VP)
= 0.017 in. WC
(2) Design for two 90-degree bends in sec-
ondary zone:
Assume 1-VP loss for each 90-degree
bend and that the products of combus-
tion have a composition equivalent to
that of air.
Gas velocity 25 fps and 1,400°F
h =
3te)
h = 0. 04 in. WC
Los s
/O. 04 in. WC\
V VP /
(2 VP) = 0. 08 in, WC
(3) Friction loss through secondary zone:
0.008 (H)(V)2
= see ltem
F =
(0. 008)(3.75)(25)
(1.25)(1,860)
1Research-Cottrel1, Inc.
F = 0.008 in. WC
(f) Total effective draft required from stack:
(d) + (e)(l) + (e)(2) + (e)(3) = total
0. 050 in. WC -f 0. 017 in. WC +
0. 080 in. WC + 0. 008 in. WC
= 0. 155 in. WC
(g) Stack height:
Let H = stack height, ft
/O. 086 in. WC\ , ,
( 10-ft stack j(H) = 0.155 in. WC
H = 18.0ft
DRUM RECLAMATION FURNACES
INTRODUCTION
Drum reclamation constitutes an important seg-
ment of the salvage industry. In this operation,
steel drums used in transporting and storing chem-
icals and other industrial materials are cleaned,
repaired, and repainted for reuse. Although steel
drums are made in many sizes, 30-gallon and 55-
gallon sizes are the two most common.
-------
482
INCINERATION
Drum construction, closed-top or open-top, de-
termines the process selected for the cleaning
phase of reclamation. Closed-top drums are cleaned
with solvents, hot caustic, or other chemical solu-
tions; open-top drums can be cleaned not only with
chemicals but by burning the combustible materials
adhering to the drum surfaces. Since cleaning
open-top drums by incineration can usually be done
at a cost lower than that of chemical cleaning, it
has been widely adopted by industry. This incin-
eration process and its related equipment are dis-
cussed here.
Description of the Furnace Charge
Typical materials to be burned from open-top steel
drums include asphalt compounds, sealants, paints,
lacquers, resins, plastics, lard, foodstuffs, grease,
solvents, and numerous other industrial liquid and
solid materials. Of course, the variable amount
of residue remaining in the drums results riot only
from the nature of the contained material but also
from the unpredictable degree of thoroughness with
which the drum is emptied. Although a few 55-
gall on drums received for processing may contain
as much as 20 pounds of combustible material,
over 90 percent normally contain less than 3 or 4
pounds; most of the 30-gallon drums contain corre-
spondingly less. In current plant operations in Los
Angeles County, 55-gallon drums constitute 75 to
80 percent of total open-top drums reclaimed by
incineration, with 30-gallon drums making up the
balance.
Description of the Process
Open-top steel drums may be cleaned by burning
out the residual materials in the open or in refrac-
tory-lined chambers. The drums are generally
in an inverted position, with the open top down so
that residual materials have a chance to melt and
flow free of the drum as well as burn. In the fur-
nace, flame applied to the exterior surface to burn
off grease, paint, and other coatings is also carried
into the interior of the drum by ignition of molten
material dripping from the interior surfaces.
After the combustibles are consumed, the drums
are allowed to cool. They are then shot peened
to remove all ash and char. Dents or surface ir-
regularities are removed by special rolling ma-
chines; finally, the drums are tested hydraulically
and protective coatings applied.
As expected, burning residue from drums in re-
fractory-lined furnaces is more efficient than burn-
ing in the open since heat is conserved within the
furnace, and combustion air can be controlled.
Refractory-lined furnaces can be classified as to
type of process--batch or continuous. A batch-
type single-chamber furnace, as shown in Figure
344, is designed to accommodate one drum at a
time; its capacity is usually limited to less than
30 drums per hour. Continuous-type furnaces,
depicted in Figure 345, are constructed in the form
of a tunnel and are usually designed to burn about
150 drums per hour.
Drums are supported upside down upon a drag con-
veyor, the drum covers sometimes resting across
adjacent drum bottoms. They move through the
tunnel where burner flames impinge on the exterior
surfaces. Exterior coatings burn and peel off while
residual materials inside catch fire, melt, and drip
onto a flat surface at the base of the conveyor. Al-
though melted materials may burn upon the flat
surface, they are scraped along and carried from
the furnace by the returning flights of the conveyor.
Water sprays are used to quench any burning mate-
rials before they leave the furnace.
Drums are spaced at least 3 or 4 inches apart on
the drag conveyor to allow flames from the primary
burners to cover the drum surface completely.
THE AIR POLLUTION PROBLEM
The practice of burning off organic residues, paint,
and other materials from drums either in the open
or in a single refractory-lined chamber results
in the emission of large quantities of smoke, odor,
and combustion contaminants. These emissions
can occur not only from the fan discharge or stack
but also from the furnace ports and other openings.
AIR POLLUTION CONTROL EQUIPMENT
There is no feasible way of controlling emissions
from open burning. However, emissions can be
controlled from properly designed single -chamber
furnaces by venting to an afterburner or a secon-
dary combustion chamber similar in arrangement
to the mixing chamber of a multiple-chamber in-
cinerator as described in the first two parts of this
chapter. Information on the design of afterburners
is given in the first part of Chapter 5. In design-
ing an effective afterburner or an equally effective
secondary combustion chamber, however, the size
and appurtenances of the primary chamber must
be selected first.
Primary Ignition Chamber, Batch Type
A batch-type chamber, shown in Figure 344, is
designed to hold one 55-gallon drum with a space
of 6 inches or more between the drum and refrac-
tory walls. Obviously, this same chamber can
alsobe used to process the smaller 30-gallon drum.
Since the drum is burned upside down to allow re-
SPO 806—614—17
-------
Drum. Reclamation Furnaces
483
Figure 344. Batch-type drum reclamation furnace with
an afterburner (Apex Drum Co., Los Angeles, Calif.).
sidual materials to melt and flow from the drum,
removing the products of combustion from the bot-
tom of the chamber rather than the top is advan-
tageous in order to promote the carryover of flames
into the drum interior.
Several gas burners are strategically arranged
around the chamber so as to cover the exterior
drum surface completely with flame. These gas
burners usually operate at 20 percent excess com-
bustion air. Air is supplied for combustion of drum
residue through air ports in the sides of the cham-
ber.
For design purposes, air ports should permit the
induction of 200 percent excess air for combustion
of 4 pounds of combustible materials within a nom-
inal 4-minute period. The composition of the com-
bustible is considered equivalent to US Grade 6
fuel oil. The primary burners should be capable
of raising the temperature of the induced air to
1,000°F and of the steel drum to at least 900°F,
based upon the most severe operating condition--
that of maximum air induction and negligible com-
bustible materials on the drum. Yet, excessive
drum temperatures must be avoided to prevent
drum warpage and scale formation.
Primary Ignition Chamber, Continuous Type
Although the design of a. continuous-type ignition
chamber for reclaiming 55- and 30-gallon drums
involves the same basic factors of combustion as
those for the batch-type chamber, certain factors
such as combustion volume, burner capacity, and
combustion air differ markedly for this dynamic
process.
The process requires sustained temperatures for
removal by melting, and virtually complete com-
bustion of all residue and surface coatings on the
drum during its period of conveyance through the
furnace. As shown in Figure 346, the furnace is
constructed in the form of a tunnel that can be con-
veniently divided into three zones. After entering
the tunnel, the drums pass through the preheat zone
where they are heated.by radiation from the igni-
tion zone; they then pass through the ignition zone
where combustibles in direct contact -with burner
flames ignite and burn; lastly, they pass through
the cooling zone where a small amount of burning
continues until combustion is complete and the
drums are cooled by induced air.
Of necessity, various dimensions of the tunnel are
established by the size of a standard 55-gallon drum
whichaverages 24 inches in diameter and 35 inches
in height. With only minor adjustments, this tun-
nel can also serve in processing the smaller 30-
gallon drum, whichaverages 19 inches in diameter
and 29 inches in height.
Although combustible content of the combustible
materials on each drum can vary drastically, over
90 percent of all drums as received for processing
contain from nearly zerotoabout 4 pounds of com-
bustible materials. Fortunately, it is possible to
design a continuous furnace that will process drums
containing this range of combustible without re-
quiring extensive and continual adjustments.
Combustion dynamics do, however, require a fur-
nace of an optimum size to accommodate the vari-
ations in burning rates among the drums as they
move along the tunnel so that all products of com-
bustion are retained for admission to the afterburn-
er or secondary combustion chamber. To process
an average drum containing anywhere from zero to
4pounds of combustibles requires an average of 4
minutes. The 55-gallon drums must be spaced on
the conveyor not less than 3 or 4 inches apart to
allow complete flame coverage of the exterior sur-
face by flame passage among the drums. In pro-
cessing 150 drums per hour or 2. 5 drums per
minute, with a design space of 5 inches between
drums, the conveyor must move at the rate of 6 fpm;
therefore, the combined length of the ignition zone
and the cooling zone in which all burning takes place
is 24 feet. Most drums are allowed to reach about
900°Fin the furnace whereupon they begin to glow
-------
484
INCINERATION
Figure 345. Continuous-type drum reclamation furnace with an afterburner
(D and M Drum Company, South El Monte, Calif.).
a dull red, but the temperature of the drum must
not exceed a bright orange color of 1,000°F; other-
wise excessive drum warpage and scaling occurs
with a subsequent loss in the strength of the steel.
Of course, drum temperatures do not represent
the temperature of the exhaust gases leaving the
ignition zone.
Optimum furnace performance requires that the
furnace be adjustable in conveyor speed and burn-
er setting. If drums containing negligible combus-
tibles are processed exclusively, the speed of the
conveyor and production rate can be increased.
Conversely, if so-called difficult drums, drums
containing more than 4 pounds of highly combus-
tible asphaltic and adhesive compounds , are burned
exclusively, they must be spaced 6 feet or more
apart on the conveyor moving at a normal speed of
6 fpm in order to retain the same residence time
but prevent overloading the afterburner.
Air for combustion of combustible materials on the
drums is supplied through minimum size drum in-
let and outlet openings on the ends of the tunnel,
in order to maximize indraft velocities.
A practical clearance of about 1 to 2 inches is pro-
vided between the 55-gallon drum, and the walls and
arch of the refractory-lined opening. The area
required for the protruding conveyor through which
air can be induced should also be kept as small as
possible. The internal dimensions of the opening
are 26 inches wide by 36-1/2 inches high. A space
for the conveyor of about 14 square inches is pro-
vided at each end. The openings should extend at
least 30 inches, which exceeds the minimum 27-
or 29-inch space allowance for 55-gallon drums
upon the conveyor. At least one drum should al-
ways be in position to blank off most of the area of
the opening and thereby create high indraft veloc-
ities .
Air curtains may also be installed at the ends of
the tunnel to help prevent the escape of smoke caused
by air currents or wind across the face of the tun-
nel. They consist of drilled pipe located around
the inside edge of the tunnel opening through which
air is injected across the face of the opening to
flow inward to the center of the tunnel.
An average indraft velocity of 200 fpm through the
tunnel openings without drums on the conveyor sup-
-------
Drum Reclamation Furnaces
485
-STACK
SECONDARY
AIR PORT
SECONDARY
BURNER
PRIMARY
BURNER
AFTERBURNER
(SECONDARY
COMBUSTION
CHAMBER)
-BAfFtt
Figure 346. Diagram of a continuous-type drum reclamation furnace with an afterburner.
plies approximately 50 percent in excess of theo-
retical air for burning a maximum of 4 pounds of
combustible per drum. In this case, the combus-
tibles are considered equivalent in composition to
US Grade 6 fuel oil. Nevertheless, in addition to
combustion air through the tunnel openings, up to
100 percent of theoretical air should be supplied
through a secondary air port for operating flexi-
bility.
As shown in Figure 346, the ignition zone is located
at the central part of the tunnel. Primary burners
are designed to attain drum temperatures of 900°F
and average effluent temperatures of about 1, 000 °F,
based upon the drums' containing no appreciable
combustible residue. The volume of the ignition
zone may be determined from a heat release factor
of about 22,000 Btu per hour per cubic foot with
primaryburners at maximum design capacity and
drums containing negligible combustible materials.
This factor is in line with the heat release factors
for oil-fired furnace fireboxes operating at tem-
peratures of less than 1,800°F.
Since flames must effectively cover the exterior
surface of the drum, the burners are mounted in
refractory-walls 6 inches from the sides of the 55-
gallon drum. Thus, internal width of the zone is
36 inches. The arch rises about 78 inches, an ar-
bitrary design figure, above the base of the con-
veyor to provide volume for collecting the products
of combustion.
As shown in Figure 346, only the ignition zone con-
tains primary burners. These burners are ar-
ranged in eight vertical rows of two or three burn-
ers -with four rows on each side of the chamber
spaced about 2 to 2-1/2 feet apart.
The rows are offset 1 to 1-1/4 feet from opposite
sides of the chamber to prevent the flames from
the burner on one side of the chamber from direct-
ly opposing flames from burners on the opposite
side.
Burners are mounted on at least two levels to cover
the surface of the drum completely with flame. If
-------
486
INCINERATION
each row contains three burners, the burners are
mounted 12 inches apart vertically, and the bottom
burner is mounted 6 inches above the top of the con-
veyor. The first row of burners on each side is
usually set for operation at maximum capacity, its
flame travel extending about three-fourths of the
width of the zone. The burners in the rows that
follow are adjusted manually, usually at reduced
capacity, or controlled automatically by a signal
from a thermocouple at the inlet to the afterburner.
The cooling zone provides for completion of the
combustion process within the nominal 4-mlnute
design residence time. Usually only a small per-
centage of the total combustion occurs within the
zone.
After the length of the ignition zone is computed,
the length of the cooling zone is determined in feet
by subtracting from, the combined length of the igni-
tion zone and cooling zone as described. Internal
cross section dimensions for the cooling zone match
those of the ignition zone.
The refractory-lined preheat zone of 10 feet has
been found to conserve heat adequately within the
tunnel and protect the operator from excessive
heat if he is stationed at the inlet opening. In-
ternal cross-section dimensions of the cooling zone
also match those of the ignition zone.
Afterburner (Secondary Combustion Chamber)
To meet air pollution regulations, afterburners or
secondary combustion chambers should be designed
to raise the maximum volume of effluent from the
ignition zone to at least 1, 400 °F for a minimum of
0. 5 second. These conditions ensure essentially
complete combustion of elemental carbon and most
organic combustion contaminants in the primary
effluent.
For turbulent mixing of the gaseous effluent with
flames from, natural gas- or oil-fired secondary
burners, the cross section of the secondary com-
bustion zone should be designed for average gas
velocities of 20 to 30 fps and contain baffles or
abrupt changes in gas flow. Secondary air ports
shouldprovide 100 percent theoretical combustion
air for the combustible materials based upon a total
processing time of 4 minutes per drum containing
a maximum of 4 pounds of combustible materials.
Draft
Although draft is usually produced by a natural-
draft stack or an induced-draft fan, the fan is pre-
ferred since it provides more nearly uniform draft
under all phases of operation. Studies of various
induced-draft fan systems show lowest costs for a
system designed around a steel fan with heat sling -
ers where combustion gases to the fan are coole
to 800 °F or less by either air dilution or evapora
tive cooling. Cooling by air dilution is, however
preferred over evaporative cooling for several
reasons. There is less corrosion of fan and duct•
work with air dilution and there is very little de-
posit of fly ash and other combustion particles up-
on the blades of the induced-draft fan as commonl-;
occurs when this fan follows a water spray cham-
ber. In fact, with spray chambers, scraping de-
posits from the fan blades may be necessary every
few days to keep the fan from becoming unbalanced
A satisfactory air dilution system should consist
of a chamber with a cross section such that a mix-
ture of dilution air and combustion gases has an
average velocity of 20 to 30 fps for a residence
time of about 0. 2 or 0. 3 second. Turbulent mix-
ing is further enhanced by adding baffles or right-
angle bends. The induced-draft fan can be pro-
tected from excessively high temperatures by
motor-driven dilution air dampers set to respond
to a signal from a thermocouple located at the fan
inlet.
Standards for Construction
Mechanical design and structural features of drum
reclamation furnaces are discussed in general
terms since most municipalities have their own
specific building requirements. While these codes
are written primarily to provide safe structures
and prevent fires, designers should not hesitate to
gobeyond the codes in specifying refractories that
will give a reasonably long service life and resist
abrasion, erosion, spalling, and slagging.
The exterior walls of the furnaces are usually con-
structed of bonded brick or steel plate. These ex-
terior walls are separated from the inner refrac-
tory lining by an airspace for cooling or by 2-1/2
inches or more of insulating materials with a ser-
vice temperature of at least 2, 000°F.
Furnace parts encountering the most severe heat,
such as the single-batch chamber, the ignition zone
of the continuous furnace, and the afterburner or
secondary combustion chamber,should be lined with
at least 9 inches of superduty firebrick or plastic
refractory. Other parts of the continuous furnace
under less severe heat conditions, preheat zone,
cooling zone, and tunnel openings may be lined
with 9 inches of high-heat-duty firebrick or ASTM
Class 27 castable refractory.
Natural-draft stacks are usually constructed of 10-
gage steel plate and lined with at least 2-1/2 inches
of insulating castable with a minimum service tem-
perature of 2, 000°F.
Induced-draft fans may be constructed of low-carbon
steel if gases are cooled by dilution air, but if water
-------
Drum Reclamation Furnaces
487
sprays are used to cool exhaust gases, then the fan
blades and the casing should be constructed of stain-
less steel or other corrosion- and heat-resistant
metals.
Drag conveyors in continuous furnaces are driven
by gearhead motors with bearings constructed of
heat-resistant alloy or with bearings cooled by
water.
Primary and secondary burners are usually nozzle
mix type to provide luminous flame. Combustion
air to the burners maybe suppliedby a singleblow-
er, but burner controls should allow for high turn-
down ratios.
Operation
Control of emissions from reclamation furnaces
•withafterburners or secondary combustion cham-
bers still depends to a great extent upon the skill
and vigilance of the operator. If visible emissions
occur as a result of overloading the afterburner,
the primary ignition burners should be cut back to
reduce the burning rate.
While black exhaust smoke may indicate a lack of
combustion air as well as an overloading of the after-
burner, white smoke usually indicates insufficient
temperature in the afterburner. White smoke can
usually be reduced or eliminated by reducing the
combustion air or by increasing the fuel consump-
tion of the secondary burners.
Drum temperatures should be kept below 1,000°F
to minimize drum warpage and scaling.
conveyor = 36-l/2in. ; length = 30 in. Total
opening area (2){6. 8 ft2) = 13.6ft2. Induced
air 13. 6 ft2 x 200 fpm = 2, 720 scfm.
b. Size of ignition burners to raise effluent to
1,000°F.
Design ignition burners for most severe oper-
ation, that of negligible combustible per drum.
Burners must raise temperature of drums to
900°F. From Table Dl, Appendix D, density
of air at 60°F and 14. 7 psia is 0. 076 lb/ft3.
Average specific heat of products of combus-
tion is 0.26 Btu/lb-°F.
(1) Heat required to raise induced air from
60° to 1, OOO'F:
Q = W C (T, - T )
1 a pa 2 a
where
Q = heat required, Btu/hr
W = weight of air, Ib/hr
3.
C = average specific heat over tern-
pa ^
perature range
T = final temperature, °F
Lj
T = ambient air temperature, °F
Illustrative Problem
Problem:
QI = (2, 720 scfm)
60min\/0.
( — - - II
X r ' ^
076 Ib
f t
,
(0.26 Btu/lb-°F)
Design a continuous-tunnel-type furnace for pro-
cessing 150 standard 55-gallon, steel, open-top
drums per hour.
(1, 000°F - 60°F) = 3, 030,000 Btu/hr
(2) Heat required to raise temperature of
drums from 60° to 900°F:
Given:
Combustible material attached to each drum varies
from near zero to 4 pounds (typical of range of com-
bxistibles onmost drums as received for process-
ing).
The specific heat of steel for this tem-
perature range is 0. 12 Btu/lb-°F.
Q
2
~ WJ c
d
,
pd
(T, -
All combustible material on the drums is con-
sidered to have a composition equivalent to that of
US Grade 6 fuel oil.
Solution:
1. Primary ignition chamber:
a. Induced air through openings at 200 fpm and
60°F. Opening width = 26 in., height above
Q^ - heat required, Btu/hr
W = weight of drums, Ib/hr
C = specific heat of steel, Btu/lb-°F
pd
T = final temperature
Tj = initial temperature
-------
488
INCINERATION
/ISO
hr
v
= 833.,'000 Btu/hr
(3) Total heat required in ignition zone:
Assume heat losses through radiation,
storage, and so on are 10 percent of
total cgroas heat input.
3,030,OOP Btu/hr + 833/000 Btu/hr
'0..'90
= 4, 300,'000 Btu/hr
(4). Natural— gas 'capacity of primary burners:
From Table D,7, Appendix D, the calorific
value -of 1 .scf natural gas -with 20 percent
exoesis ,air is ,7 56.. 2 Btu .at 1,000°F.
4,300, WO Btu/hr „,„
Total capacity = ' ^ 2 Btu/8cf = 5, 830 scfh
(5) Individual burner capacity:
Install -eight rows of three burners each
(four rows on each side of the zone,)
„ .„ 5,830 scfh
Burner capacity = — = 243 scfh
c. Excess primary combustion air:
Assume all air for burning materials on drums
is induced through tunnel openings (including
air supplied by air curtains).
(1) Maximum design burning rate:
hr
10 Ib/min
(2) Total combustion air available through tun-
nel openings:
2, 720 scfm ..„., ,
Air = ' „ .—: = 272 scf/lb
10 Ib/min
From Table D6, Appendix D, 1 Ib US
'Grade .6 fuel oil requires 177 scf air
40 percent saturated at 60°F.
% -excess .air available =
272 scf - 177 scf
177 scf
= 54%
d. Average gas temperature in ignition zone when
burning a maximum of 4 pounds combustibles
per drum:
As sume first rows of burners on opposite sides
of zone are operating at 1 million Btu/hr
(910 scfh) to ignite combustibles on drums.
Design gas burners to operate with 20 percent
excess air. Assume radiation, storage, and
otherheat losses are 35 percent of gross heat
input at furnace temperatures near 2, 000°F.
From Table D5, Appendix D, the gross heat
of combustion of 1 pound US Grade 6 fuel oil
is 18, 000 Btu.
(1) Gross heat:
Primary burner = 1 million Btu/hr
•Combustibles
Total
3/000 7^1= 10, 800, 000 Btu/hr
= 11, 800, 000 Btu/hr
(2.) Heat losses, radiation, storage, and
so on:
'(0. 35)'(11, 8-0'0, 000 Btu/hr) = 4, 130, 000 Btu/1
(3) Evaporation of moisture contained in
drums:
Assume anaverage of 0. 5 Ib -water per drum.
The heat of vaporization of 1 pound of water
at 60°F and 14. 1 psia is 1, 060 Btu.
Ib
ISO drums
Vl,
)\
060 Btu
Ib
79, 500 Btu/hr
(4) Evaporation of water formed by combus-
tion:
From Tables D7 and D6, Appendix D,
0. 099 Ib water is formed by burning 1 scf
natural gas with 20 percent excess air while
0. 91 Ib water is formed from burning 1 Ib
US Grade 6 fuel oil -with 54 percent excess
air.
Natural gas:
(910 scfh)
IQ. 099 Ib HO
k scf
= 95, 500 Btu/hr
hr
Combustibles:
Total
~) = 578, 000 Btu/h
= 673,500 Btu/h
-------
Drum Reclamation Furnaces
489
(5) Total heat losses:
(2) + (3) + (4) = 4, 883, 000 Btu/hr
(6) Net heat available to raise temperature of
products of combustion:
11, 800, 000 Btu/hr - 4, 883,000 Btu/hr
= 6,917,000 Btu/hr
(7) Weight of products of combustion:
From Tables D7 and D6, Appendix D,
there is 0. 999 lb products of com-
bustion from 1 scf natural gas with 20
percent excess air and there is 21. 71
lb products of combustion from 1 pound
US Grade 6 fuel oil with 54 percent ex-
cess air.
(910 scfh)(0.999 Ib/scf) =
909 Ib/hr
Total
= 13, 000 Ib/hr
= 13, 909 Ib/hr
(8) Average gas temperature:
Average specific heat of products of com-
bustion (equivalent to air) is taken to be
0. 26 Btu/lb-°F for the given temperature
range.
Primary burners:
(5,830 scfh)(13.53 scf/scf) = 79,000 scfh
Total = 242,000 scfh
4, 040 scfm
67.3 scfs
(2) With 4 pounds combustibles per drum
Assume primary burners are operating at
910 scfh. From Table D6, Appendix D,
there is 281.9 ft-* products of combustion
from 1 pound US Grade 6 fuel oil with 54
percent excess air.
Combustibles:
(600 lb/hr)(281. 9 ft3/lb)
Primary burners:
(910 scfh)(13.53 scf/scf)
Total
= 169, 000 scfh
= 12,300 scfh
= 181, 300 scfh
3,020 scfm
50. 4 scfs
The most severe operating conditions exist,
therefore, in the ignition chamber when
drums with negligible combustible material
are processed.
AT =
W C
t pc
where
AT = temperature rise, °F above 60°F
Q = heat available, Btu/hr
W = weight of products of combustion,
t Ib/hr
C = average specific heat, Btu/lb-°F
pc
6, 917, OOP Btu/hr
(13, 909 lb/hr)(0. 26
= 1,910°F
Final Temp = 60 + 1,910 = 1,970°F
e. Volume of products of combustion at bO°F
(1) With negligible combustibles on drums:
Induced air:
(2, 720 scfm)(60 min/hr) - 163, 000 scfh
f. Volume of ignition zone:
Assume aheat release factor of 22, 000 Btu/hr-
ft , •which is similar to heat release factors
for oil-fired furnace fireboxes operating at
less than 1, 800 °F. Assume drums contain
negligible combustible materials.
Volume =
(5,830 scfh)(l, 100 Btu/scf)
22, 000 Btu/hr-ft
= 292 ft
g. Length of ignition zone'.
Assume width = 36 in.; height = 84 in. , in-
cluding the conveyor
Length =
volume
_ 292 ft
(height)(width) ~~ (3 ft)(7 ft)
= 14 ft
h. Cooling zone length:
Assume width = 36 in. , height = 84 in. , in-
cluding the conveyor. Design ignition zone and
cooling zone for a total residence time of 4
-------
490
INCINERATION
min. Assume a drum spacing of 29 in. (5 in.
between drums). Internal cross-sectional di-
mensions match those of ignition zone.
Conveyor speed
/ISO drumsV 1 hr \(z. 41 ft\
\ hr A 60 min/\ drum /
hr
= 6 fpm
Length of ignition and cooling zones
Total length = (6 ft/min) (4 min) = 24 ft
Cooling zone length, 24 ft - 14 ft = 10 ft
i. Preheat zone length:
Design this zone to minimize radiation losses
and to protect operator. Internal cross-sec-
tional dimensions match those of ignition zone.
Design preheat zone length = 10 ft
Evaluation of existing design shows that a pre-
heat zone length of 10 ft will be adequate.
2. Secondary-combustion chamber (afterburner):
a. Design gas burners for most severe operation
•(drums contain negligible combustibles). Af-
terburner will raise temperature of products
of combustion from ignition zone from 1,000°
to 1,400°F.
(1) Weight of products of combustion:
From ignition zone:
Induced air:
where
T = initial temperature, °F
T = final temperature, °F
I,400°F - 1, 000°F)
= 1,900, 000 Btu/hr
(2, 720 scfm)(0. 076 —\
V ft/
= 12,400 Ib/hr
(3) Total heat required in afterburner:
Assume heat losses by radiation, convec-
tion, andsoonarelO percent of gross heat
supply at 1, 400°F
Total heat = L 900^000 Btu/hr = 2,n0>000 Btu/hr
(4) Total capacity of secondary burners:
From Table D7, Appendix D, the calorific
value of 1 scf natural gas is 615. 4 Btu at
1,400°F with 20 percent excess air.
2,110,OOP Btu/hr A_n
Natural gas = - ' ". ; = 3,430 scfh
615.4 Btu/scf
(5) Individual secondary burner capacity:
Install four burners--two on each side of
the horizontal section.
3, 430 scfh _-, „
Capacity of burner = ——; = 856 scfh
4 burners
b. Cross-sectional area:
Design afterburner for a cross-section veloc-
ity of 30 fps maximum at 1,400°F. From
Table in Appendix, there are 13. 53 scf prod-
ucts of combustion from 1 scf natural gas with
2.0 percent excess air.
Natural gas:
(5, 830 scfh)
Total
(0. 999 lb)
scf
= 5,870 Ib/hr
= 18,270 Ib/hr
(2) Heat required to raise temperature of prod-
ucts of combustion to 1, 400 °F:
Average specific heat of products of com-
bustion is 0.26 Btu/lb-"F over the given
temperature range.
Wt Cpc (T2 - V
(1) Volume of products of combustion when
drums are burned with negligible combus-
tibles :
Induced air:
(2, 720 scfm)(
in\
1
hr /
Primary burners:
= 163, 000 scfh
scf
79, 000 scfh
-------
Drum Reclamation Furnaces
491
Secondary burners:
. 53 scfv
(3, 430 scfh)
Total
scf
46,500 scfh
288,500 scfh
4, 810 scfm
80. 2 scfs
(2) Volume of products of combustion through
afterburner when drums are burned with
4 Ib combustibles:
Assume primary burners are operating at
910 scfh and that secondary burner s operate
at 20 percent full capacity. Assume after-
burner outlet temperature is 2, 100°F.
Products of combustion:
Combustibles:
(600 lb/hr)(281. 9 cf/lb) = 169, 000 scfh
Primary burners:
(910 sc
/13.53 sc£\
scfh)| - I
\ Scf /
12,300 scfh
Secondary burners:
(0.20)(3,430 scfh)
/13.53 scf\
V scf /
= 9, 300 scfh
Total
= 190, 600 scfh
3, 180 scfm
53. 0 scfs
(3) Internal cross -sectional area and dimen-
sions :
Arca =
80. 2 scfs
.
= 9-6 ft
2
Dimensions = 3 ft 2 in. wide x 3 ft high
c. Length of afterburner:
Design for a minimum residence time of 0. 5
second.
Length = (30 fps)(0. 5 second) - 15 ft
d. Afterburner arrangement:
Design for 2 right-angle bends and add dilution
air at third right-angle bend,
e. Secondary air port
Designa secondary air port to supply up to 100
percent theoretical air for drums containing
4 Ib combustibles. From Table D6, Appen-
dix D, 177 scf air is required to burn 1 pound
of US Grade 6 fuel oil.
(1) Volume of combustion air at 60°F:
(600 Ib/hr)
= 106,000 scfh
1,770 scfm
(2) Pressure drop through opening at end of
tunnel:
Drum cross-sectional area:
24 in. in diameter x 34 in. high =5.7 ft^
T , . , .. 2,720 scfm
Inlet velocity = • ;— • •—-
(2)(opening area - drum area)
Inlet velocity =
2,720 scfm
(2)(6. 8 ft2 - 5.7 ft2)
= 1, 190 fpm
Total pressure behind opening:
TP = VP + SP
where
TP = total pressure, in. WC
SP = static pressure, in. WC
VP = velocity pressure, in. WC
From Table D8, Appendix D, velocity
pressure is 0.090 in. WC for a velocity
of 1, 190 fpm at 60 °F. Assume static
pressure drop through sharp-edge orifice
opening is 0. 5 VP and negligible friction
loss in a 30-in. length of opening.
TP = 0.090 in. WC + 0. 5 (0. 090 in. WC)
= 0. 135 in. WC
(3) Pressure drop through preheat zone:
Cross-sectional area for air flow with drums
upon conveyor. Assume half of total com-
bustion air through preheat zone.
22 2
Area = 21 ft - 5. 7 ft = 15. 3 ft
-------
492
INCINERATION
Velocity = (0.5X2. 7ZO.cfm) __
15.3 ft
Because of low velocity, the pressure drop
is negligible.
(4) Pressure drop through ignition zone:
Assume friction is negligible and pressure
drop is 1 VP for 90-degree bend (into af-
terburner).
Assume flow conditions for negligible com-
bustibles on drums. Cross section inlet
duct of afterburner 9.6 ft2.
Induced air:
Primary burners:
= 2, 720 scfm
(5,830 scfh)
Total
[hr) (13.53 scf)
—*r- J ~ = 1,320 scfm
min
scf
= 4, 040 scfm
67.4 scfs
Average velocity into afterburner at 1, 000 °F:
. 4
17 i •
y =
. . _
'
VP = 0. 032 in. WC at 1, 000°F.
Pressure drop through one half ignition
zone = 0. 032 in. WC
(5) Total pressure at inlet to afterburner (be
hind secondary air port):
Tunnel opening
Preheat zone
Ignition zone
= 0. 045 in. WC
= - 0
= 0. 032 in. WC
Total static pressure = 0. 077 in. WC
TP = VP + SP
TP = 0.032 + 0.077 = 0. 109 in WC
1. 5 VP = 0. 109 in. WC
VP = 0.073 in. WC
From Table D8 in Appendix D:
Velocity = 1, 070 fpm at 60°F
(7) Secondary air port area:
,,. . 1, 770 scfm , .c . 2
Minimum area = ——n_n = 1.65 ft
1, 070 fprn
Install oversize secondary port -with area
= 2 ft2
3. Dimensions of dilution air chamber:
Design dilution air port to reduce temperature
of products of combustion from afterburner
to 700°F for safe fan operation.
a. Dilution air required to lower products of com-
bustion from 1, 400° to 700°F:
Density of products of combustion 0. 076 Ib/scf
at 14. 7 psia -and ,6.0 °F
W C (T - T ) = W C (T-T)
pc pc 2 1 a pa 1 a
(4,810
_ 700.F)
W = 400 Ib/min
3.
_.., .. . 400 Ib/min
Dilution air = • -- - = 5, 260 scfm
0. 076 Ib/scf
b. Dilution air r equired to lower products *of com-
bustion from 2, 100° to 700 °F;
) 180
(6) Velocity through secondary air port:
Assume 0.5 VP static pressure drop through
sharp-edge opening of secondary port.
TP = VP + SP
0.109 in. WC = VP + 0. 5 VP
W = 529 Ib/hr
a
Dilution air =
529 Ib/hr
0.076 Ib/scf
= -6, 950 scfm
-------
Drum Reclamation Furnaces
493
c. Cross section of dilution air chamber;
Design for a velo-city of 30 fps- at 700 °F to en-
sure turbulent flow for good mixing.
Total flow to fan at 700° F:
Condition 1: (no combustibles on drams, max-
imum primary bmmer capacity),
(4-, 8-10 srfm + 5,26^ scfm),
("1'
= 26, 400 cfm
Condition 2: (41b comhuatiMes per dram, pri-
mary bramersf. 910 scfm)
(6, 950 scfm Hr 3, 180, sciraaj,
R }
= 26, 500 cfm
5-20 R
Maximum cross-sectional area = - ' - — cm
= 14.7ft
Cross-sectional diamemsioma 4 ft 8 in. wide
x 3- ft 2 MB. high
d. Leng,tfe of dilution, air chamber:
Desigp, fofr a. residence time of 0. 3 sec
Length = (30-fpa),(Qv 3 sec) = 9 ft
4. Static pre-ssiEE® d'r-op through system:
Design aysterniw-iihiiiniduced-draft fan mounted
at ground level and- a, vertical stack on the fan
outlet 50'in. in.d'iametcr x 20 ft high. The 50-
in. diameter- will keep stack velocity near 30
fps through afterburner.
a. Static pr-essure at afterburner inlet = 0. 077
in. WC, see item (5), page 492.
b. Static pressure. d'rop> through afterburner:
(1) Velocity pressure at 30 fps and 1, 000°F:
Aa.sume combustion products are equiva-
lent in composition to air.
*
V = 2. 9VtiT
where
V = gas velocity, fps
t = absolute temperature, °R
h = velocity pressure (head), in. WC
h =
= °-073in-wc
(2) Pressure drop from contraction at inlet to
afterburner:
Assume 0. 5 VP drop for abrupt contraction
(0.5/0.07 ln'Vp C\ = 0.035 in.
WC
(3) Pressure drop for three right-angle bends:
Assume 1 VP for each right-angle bend.
= 0.2iin. WC
(4) Friction loss through 15 ft of ductwork hav-
ing dimensions 3 ft 2 in. wide x 3 ft
high.
r
0.002 h v
mt
zt
where
f = friction, in. WC
h = duct length, ft
v = gas velocity, fps
t = absolute temperature, °R
m = hydraulic radius
cross-sectional area of duct, ft
m = •
perimeter of duct, ft
(0.002)(15 ft)(30£psr n n, . . w_
f = (0.778ft)(l,460-R) = °' °24 ln" WC
(5) Total drop through afterburner:
(2) + (3) + (4) = 0.269 in. WC
*Research-Co«3tel 1, inc.
tGriswold, 19U6.
-------
494
INCINERATION
c. Static pressure drop through dilution air cham-
ber having dimensions 4 ft 8 in. wide x
3 ft 2 in. high.
(1) Friction loss through ductwork at 700°F:
2
f =
0.002 h v
mt
- «•«'""»<=
d. Static pressure drop through 50-in. -diameter
x 20-ft-high stack on discharge side of fan:
Stack velocity at 700 °F =
26' 5°°
13. 64 ft
= 1, 940 fpm
= 32.4fps
where
r
D
4
0.002 h v
mt
4. 16
= 1. 04 ft
0. 002 (20 ft)(32. 4 fps)_
(1.04 ft)(l, 160°F)
= 0. 035 in. WC
Total static pressure drop through system:
Tunnel
Afterburner
= 0. 077 in. WC
= 0. 269 in. WC
Dilution air chamber = 0, 015 in. WC
Fan outlet duct
= 0. 035 in. W
Total static pressure = 0. 396 in. WC
5. Dilution air port size:
a. Total pressure behind dilution air port:
Velocity pressure 30 fps at 700°F, VP = 0. 10
in. WC
TP = VP + SP
TP = 0.10 + (0.077 + 0. 269) = 0.446 in. WC
b. Inlet velocity through dilution air port:
Assume 0. 5 VP static pressure drop for sharp-
edge orifice air port.
TP = VP + SP
0. 446 in. WC = VP + 0. 5 VP; VP = 0. 298 in.
WC at 60°F
Inlet velocity = 2, 165 fpm
c. Size of dilution air port:
,,. . . 6, 950 scfm , _, . 2
Minimum size = ——— = 3.21 ft
2, 165 fpm
Select a port with area = 4 ft
6. System static pressure curve development at
700°F:
2
= SP.
where
SP = static pressure, final conditions, in. W<
SP = static pressure, initial conditions, in. M
cfm = gas flow, final conditions, cfm
cfm = gas flow, initial conditions, cfm
Assume cfm = 30, 000
SP2 = (0.396)
so,
= 0.507 in. WC
Assume cfm = 20,000
= (0. 396)
= 0. 226 in. WC
7. Fan specifications:
Select a fan that will deliver about 26, 500 cfm
at 700°F and 0. 4 in. WC static pressure.
a. Fan performance at 60 °F operation:
329 rpm
0. 75 in. WC
26, 775 cfm
14.9 bhp
60°F
1.0 in. WC
25, 245 cfm
12.66 bhp
1. 25 in. WC
22,185 cfm
10.79 bhp
b. Calculate points for 700°F fan performance
curve:
With rpm and cfm held constant, static pres-
sure and bhp vary directly with gas density or
inversely with absolute temperature.
-------
Wire Reclamation
495
Correction ratio =
329 rpm 700°F
520
1, 160
= 0. 448
0. 448 in. WC
25, 245 cfm
5. 7 bhp
0. 57 in. WC
22,185 cfm
4.9 bhp
0. 336 in. WC
26, 775 cfm
6.7 bhp
c. Operating point at 700°F:
The intersection of the 700 °F system curve
with the 700°F fan curve, as shown in Figure
347, yields data indicating that this system
will handle a volume of 26, 000 cfm at 0. 38 in.
WCat700°F. The fan will operate at 329 rpm
with 6. 3 bhp.
Fan Selection:
Select a 20-hp motor to drive fan since about
14 bhp will be required when starting from a
cold lightoff.
Select a fan with a capacity and static pressure
10 to 20 percent in excess of the operating
point shown in Figure 347 as a safety factor
for overload capability.
WIRE RECLAMATION
Scrap-insulated electrical wire from construction
sites and factories, and worn-out insulated wire
from utility companies and other industrial opera-
tions constitute the bulk of the insulated wire pro-
cessedforthe recovery of copper scrap. Several
methods are employed for removing insulation from
the copper core. The method selected depends not
only upon the size of the wire but also upon the
composition of the insulation. This process differs
from other reclamation in that the combustible con-
tent of the charge is always more than 10 percent
by weight and usually exceeds 20 percent. This
one distinguishing feature is reflected in the spe-
cialized designs of combustion equipment used ex-
clusively for reclaiming electrical insulated wire.
DESCRIPTION OF THE PROCESS
Inorganic insulating materials such as fiber glass
and ceramics cannot be burned and must be re-
10.000
15,000
20,000 25,000
VOLUME, cfm
30,000
35,000
Figure 347. Performance curve of 700°F fan.
-------
496
INCINERATION
moved mechanically. Much insulation is composed
of organic compounds that will burn; however, not
all combustible insulation is removed by this meth-
od. Because of excessive oxidation of copper, wire
smaller than 14 gage is not burned; it is actually
thrown away because of the lack of a satisfactory
economical method of removing insulation. On the
other hand, communication cable, 1 inch in di-
ameter or greater, is usually cut into pieces about
1 foot long and the insulation is hand stripped. This
method has proved more satisfactory than burning
since the copper scrap is clean and free of the sur-
face oxides and foreign matter associated with the
burning process.
Wire of the intermediate sizes -was formerly burned
in the open or in single-chamber furnaces in Los
Angeles County. When burned in the open, the wire
was spread in thin piles less than 1 foot high and
sprinkled with some type of petroleum distillate to
initiate combustion. The combustible content of
the wire was usually sufficient to maintain active
burning until the insulation was consumed.
In a single-chamber furnace, the wire was ignited
with a hand torch or a gas burner mounted through
the side of the chamber. After ignition, the burn-
ing process was also self-sustaining in this equip-
ment. After burning was complete, the wire was
allowed to cool and the char adhering to the bare
copper wire was removedby rapping or by high ve-
locity jets of -water.
DESCRIPTION OF THE CHARGE
A great variety of materials composes the com-
bustible insulation: Rubber, paper, cotton, silk,
and plastics such as polyethylene and polyvinyl
chloride. Moreover, the wire itself may have a
baked-on coating of plastics, paint, or varnish.
As received for burning, the total combustible con-
tent of the insulated wire may vary widely from
several percentto over 50 percent by weight. Most
commercial wire contains from 20 to 35 percent
insulation.
THE AIR POLLUTION PROBLEM
Burning in the open is accompanied by copious
quantities of dense smoke, disagreeable odors, in-
organic materials, and oxygenated hydrocarbons.
Burning in single-chamber incinerators produces
somewhat less smoke, odors, and other air con-
taminants than open burning does, since combus -
tion air can be regulated. Results of Source Tests
C-624-1 and C-543-1, givenin Table 136, are con-
sidered representative of emissions from single-
chamber incinerators since these tests were con-
ducted on two multiple-chamber retort furnaces
operating without their secondary burners. In
these two tests, particulate-matter concentrations
in the stack effluent averaged 356 and 190 pounds
per ton of insulation burned, respectively. Smoke
emissions were a constant 100 percent black dur-
ing one entire test and varied continuously from 20
to 90 percent gray during the other test.
AIR POLLUTION CONTROL EQUIPMENT
The only practical industrial equipment available
today for controlling emis sions from single -cham-
ber insulation-burning incinerators is an after-
burner or secondary combustion chamber. For
the design of afterburners see general informa-
tion in the first part of Chapter 5.
Final selection of equipment designed to burn in-
sulation and control emissions is based upon con-
siderations such as space limitations, charge com-
position, process conditions, maintenance, capital
investment, and operating expenses. To design
an effective afterburner or secondary combustion
chamber, one must determine or initially design
the size and appurtenances of the primary ignition
chamber. Recommended values for designing a
complete furnace are presented in Table 137.
During recent years, reclamation of wire coated
with PVC or other plastics containing inorganic
filler of organic oxides or clay has been on the in-
crease. During combustion of the wire, the in-
organic materials are volatilized in the form of
fine particles. These particles are entrained by
the combustion products from the primary cham-
ber and pass through the secondary combustion
zone without burning. The concentration of these
inorganic particles can result in emissions of ex-
cessive opacity. If sucha problem develops, sec-
ondary combustion must be followed by a baghouse
or possibly a venturi scrubber operating at a high
pressure drop. Designs for baghouses or venturi
scrubbers are discussed in Chapter 4.
Primary Ignition Chamber
The size of the primary chamber is based upon the
density, volume, and burning rate of a typical
charge. There is nothing critical about the shape
of this chamber. Any reasonable box shape will
suffice for a given batch charge provided addition-
al space is provided to facilitate loading and un-
loading.
Control of primary combustion air is critical since
not only must high temperatures be prevented from
excessively oxidizing copper, but also the burning
rate must be restricted to prevent overloading of
the secondary-combustion chamber. Precise con-
trol of combustion air is import ant be cause it makes
possible the use of an afterburner or secondary
combustion chamber of reasonable size.
-------
Wire Reclamation
497
Table 136. SOURCE TESTS: WIRE RECLAMATION WITH MULTIPLE-CHAMBER RETORT FURNACES
Test No.
Operation of secondary burners
Incinerator number
Charge composition
Test duration, min
Charge weight, Ib
Combustibles in charge, wt %
Ash in charge, % by wt
Combustion rate, Ib/hr
Smoke opacities, %
Participates, gr/scf at 12% CO2
Ib particulates /ton combustible
Mixing chamber, °F
Mixing chamber velocity, fpm
Aldehydes, ppm
Hydrocarbons, ppm
Nitrogen oxides, ppm
Sulfur compounds as SO^, % by vol
C-624-1
Burners off
1
5/8 in. OD typical
rubber-covered wire
24
220
35
6
195
Constant 100% black
29.0
356
780
11
105
640
11
0.012
C-624-2
Burners on
1
5/8 in. OD typical
rubber -covered
40
233
16
6
56
0 to 25% white
0,26
35
l,880a
45. 0
5
8
25
0.0039
C-543-1
Burners off
2
3/8 to 5/8 in. OD
cotton-rubber
wire
20
100
19
4
57
20 to 90% gray
3.5
190
300 est
9. 1
9 to 36
9 to 31
2.9 to 8.5
0.0014
C-543-3
Burners on
2
3/8 to 5/8 m. OD
cotton-rubber
wire
17
147
34.7
4
180
0 to 10% white
0.32
21
l,880a
31.2
4
8
10.4
0. 0027
C-696
Burners on
3
1/4 to 3/4 in. OD
cotton - rubbe r -
83
960
22.6
4
160
0 to 10% white
0. 16
20
2,000a
42.5
Not available
Not available
Not available
0.0094
aTemperature measured by chromel alumel thermocouple in flame contact.
To minimize the size of the afterburner or secon-
dary combustion chamber, the primary chamber
should be equipped with a tightly fitting air port
and a side swing charge door. Although the pri-
mary air ports are designed to supply 100 percent
excess theoretical air for operating flexibility, air
leakage around the edges of the charge door and
air ports inmost cases supplies the required com-
bustion air so that primary air ports are usually
kept in a closed position. For design purposes,
indraft velocities through the primary air ports
should average 900 fpm, equivalent to a velocity
pressure of 0. 05 inch of water column.
Because the combustion process is self-sustaining,
only a small-capacity primary-chamber burner,
that is, one capable of 50, 000 Btu per hour, is re-
quired for igniting the refuse. After ignition, emis-
sions from the primary chamber usually consist
of smoke and gases without flame and vary in tem-
perature from 900° to 1, 300°F upon entering the
secondary combustion chamber.
Ducts or ports connecting the secondary chamber
or afterburner with the primary chamber are de-
signed for a velocity of 30 fps or less, at maxi-
mum combustion rates, to prevent excessive re-
striction to the flow of gases. Undue restriction
may result in emission of smoke and flames from
the primary air ports or around the charging door.
Secondary Combustion
As the gaseous emissions enter the secondary com-
bustion chamber or afterburner, combustion air
is added (up to 100 percent of theoretical) through
the secondary air port. The effluent then passes
through the luminous flames of the secondary burn-
er, which is designed to attain an average gas tem-
perature of 1,6QQ"F. This temperature is main-
tained for a minimum of 0. 5 second with average
gas velocities of 25 to 40 fps. Baffles and abrupt
changes in direction provide additional turbulence
for mixing burner flames with the air and combus-
tion gases.
Secondary combustion air ports are also designed
for 900-fpm inlet velocities. Additional air may
be induced through a port in the base of the stack
to cool the stack gases and protect the refractory
lining of the stack.
Emissions
Table 136 shows significant reductions of particu-
lates, aldehydes, hydrocarbons, and smoke through
secondary combustion. Source tests C-624 and
C-543 were conducted on two multiple-chamber
retort furnaces with and without their secondary
burners in operation. Tests conducted on a third
furnace with secondary combustion show a partic-
ulate discharge of only 20 pounds per ton of in-
sulation burned. This lower rate is the result of
havingahightemperature of 2, 000°F in the secon-
dary chamber and a long residence time of 0. 55
second.
Not all combustible insulation can be burned in
multiple-chamber retort furnaces without produc-
ing emissions of excessive opacity. Clays and in-
organic oxides used as fillers in polyvinyl chloride
-------
498
INCINERATION
Table 137. EQUIPMENT DESIGN FACTORS
Item
Gas velocities
Primary-chamber outlet duct or
port at 1,300°F
Afterburner or secondary mixing
chamber at 1,600°F
Extended secondary mixing
chamber curtain wall port tunnel
at 1, 600°F
Stack
Residence time
Maximum flow at 1, 600°F
Combustion air
Air requirements
Primary air
Secondary air
Combustion air distribution
Primary ports
Secondary ports
Airport inlet velocity
Primary airport
Secondary airport
Auxiliary burners
Primary burner or torch capacity
Secondary burner capacity
Draft requirements
Ignition chamber
Outlet from secondary chamber
(afterburner)
Recommended
value and units
30 ft/sec
30 ft/sec
30 ft/sec
30 ft/sec
0.50 sec
100% excess
100% theoretical
66%
35%
900 fpm or 0. 051 in. WC
900 fpm or 0. 051 in. WC
50 cfh
15,600 Btu/lb combustible
0.05 to 0. 10 in. WC
0. 20 in. WC
Allowable
deviation
+ 20%
+ 20%
+ 20%
+ 20%
+ 20%
and other insulation plastics produce objectionable
emissions consisting of micron and submicron,
noncombustible, inorganic particles of clay and
metallic oxides that are vaporized as the plastic
burns.
Draft
Draft is usually produced by a natural-draft stack
that canbe designed by standard calculations. Re-
gardless of the method employed to generate draft,
at least 0. 05 inch water column negative static
pressure should be available in the ignition cham-
ber, and a minimum of 0. 20 inch water column at
the outlet from the secondary combustion chamber.
Equipment Arrangement
Batch equipment is usually constructed in one of
two configurations--a dual structure consisting of
a primary chamber venting through an afterburner
or a single structure containing a primary chamber
and one or two secondary combustion chambers
arranged similarly to a multiple-chamber incin-
erator.
A typical multiple-chamber retort wire reclama-
tion furnace, shown in Figure 348, differs from a
multiple-chamber retort incinerator in that the
primary chamber has no grates and the charge
rests upon the floor of the chamber. To increase
the residence time in the secondary combustion
zone, the curtain wall port is extended across the
bottom of the combustion chamber, forming a tun-
nel.
-------
Wire Reclamation
499
MIXING CHAMBER-
FLftME PORT
SECONDARY
COMBUSTION
CHAMBER
CLEANOUT DOOR
TUNNEL
CHARGING DOOR
KITH AIR PORT
Figure 348. Multiple-chamber retort furnace.
Secondary combustion can actually be initiated in
the primary chamber by installing an auxiliary
burner with a capacity of about 300, 000 Btu per
hour through the outside wall of the primary cham-
ber directly opposite the flame port. Flames from
this burner start secondary combustion of the ef-
fluent from the burning pile before this effluent
enters the flame port. Thus, residence time in
the secondary combustion zone is increased.
Since fly ash is not present in appreciable amounts,
the third chamber can either be eliminated or de-
signed to maintain gas velocities equal to or less
than those in the mixing chamber. Figure 349
shows a three-chamber retort furnace designed to
burn 1, 000 pounds of insulation-covered wire per
hour.
General Construction
Construction, in general, follows many practices
given for multiple-chamber incinerators described
in the first part of this chapter. Only those skilled
in installing high-temperature refractories should
be employed in constructing this specialized equip-
ment.
Refractories
Although primary ignition chambers can be lined
with high-duty fire clay firebrick, secondary mix-
ing chambers, curtain wall port tunnels, and af-
terburners should be lined with superduty firebrick
or superduty plastic refractory.
Since flames may extend into them, stacks must be
fully lined with insulating brick or castable refrac-
tory with a service temperature of at least 2, 500 °F.
Expansion joints must be provided as specified by
the refractory manufacturer.
Charge Door
A side swing charge door is installed in contrast to
the guillotine-type door found on multiple-chamber
incinerators. Mating surface s of the door and door
jambs are grooved or recessed. The door is pro-
vided with a positive locking device, such as a cam
or wedge lock, to hold the mating surfaces in close
contact. High-heat-duty ASTM Class 24 castable
refractory is used to line the charge door.
-------
500
INCINERATION
Figure 349. A 1,000-pound-per-hour, multiple-chamber
retort furnace (Amana Scrap Metals, Compton, Calif.).
Combustion Air Ports
Air ports in the primary chamber should, be con-
structed of cast iron at least 1/2 inch thick to min-
imize warpage. Swing-type ports should be used
•with positive locking devices. Since the exterior
surface around the secondary air port is relative-
ly cool, materials of construction used for secon-
dary air ports are not critical. Ten-gage steel
plate can be used and snug fits are easily attained.
Gas Burners
To ignite the charge, hand-held natural gas torches
or low-capacity, permanently mounted, atmospheric
gas burners with flame safety controls may be in-
stalled in the primary chamber.
Secondary burners can be of several types--at-
mospheric, premix, or nozzle mix. They should
have flame safety controls and be adjusted to give
along, luminous flame for maximum effectiveness
in promoting secondary combustion. Secondary
burners should be mounted through the side of the
mixing chamber opposite the flame port, and flames*
from these burners should completely blanket the
cross section of the mixing chamber.
Operation
One of the most important factors concerning oper-
ation is to restrict the combustion rate in the ig-
nition chamber by tightly closing all primary air
ports and sealing the charge; door to prevent gas-
eous overloading of the secondary combustion
chamber. If overloading does occur •with all open-
ings closed in the ignition chamber, the combus-
tion rate canbe further reduced by spraying -water
onto the burning charge, being extremely careful
not to spray directly against the hot refractory
walls.
Although primary burners are used simply to ig-
nite the charge, secondary burners are operated
through out the burning period. In fact, the secon-
dary chamber or afterburner should be preheated
10 minutes before a cold lightoff to minimize smoke.
Materials must not be removed from the primary
chamber before the reclamation process is com-
plete since excessive smoke will be emitted to the
atmosphere. Wire is removed from the chamber
while hot with only traces of smoke present, and
it must be immediately quenched with water to stop
the smoke as well as to clean char and residual
materials from the reclaimed copper metal.
Secondary air ports should be adjusted to maintain
high temperatures in the secondary combustion
zone without emissions of black or white smoke
from the stack. Black smoke, may indicate a lack
of combustion air, which may be eliminated by
opening the secondary air ports.
Since the inorganic materials in vinyl-coated wire
are emitted to the atmosphere as submicron-size
particles even after passing through the secondary
combustion zone, the percentage of vinyl-coated
wire in a given charge may need to be restricted
in order to prevent excessive emissions.
Illustrative Problem
Problem:
Design equipment to process a 250-pound batch of
commercial insulated electrical wire containing
25 weightpercent combustibles. One batch charge
will require 30 min.
Given:
Design calculations apply equally to a single cham-
ber venting to an afterburner or to a multiple-
chamber retort-type furnace.
-------
Wire Reclamation
501
Solution:
1. Primary ignition chamber:
Assume bulk density of randomly packed -wire
charge at 4 lb/ft . Design ignition chamber
50 percent oversize.
(1.50)
(250 Ib \
4 lb/ft3/
94. 0 ft"
Use dimensions of 2. 75 ft x 5. 25 ft x 3. 25
ft high
2. Ignition chamber gas burners:
Install minimum size gas burners for lightoff.
Burner capacity = 50 cfh
3. Primary air ports:
Assume inlet velocity through port is 900 fpm
at0.052in. WC. Assume 100 percent excess
air in ignition chamber and composition of
combustibles equivalent to U.S. Grade 6 (P.S.
400) fuel oil. From Table D6, Appendix D,
354. 4 scf of air is required for 1 pound of
combustibles.
/250 Ib wire\/0.
\ 30 min )\
25 Ib combustibles\(354. 4 scf
Ib wire
\/354. 4 scf\
A ^ / =
737 scfm
Port area:
<737 cfm
/737 cfm\ /144 in. 2
\900fpm; ^ ft2
118 in.'
4. Port or duct connecting single chamber to an
afterburner (equivalent to a flame port in a
multiple-chamber incinerator):
Design for 30 fps at 1,300°F. Assume 100
percent excess air in ignition chamber. From
Table D6, Appendix D, there are 363. 3 scf
of products from combustion of 1 pound of
c ombus tible s.
(250 Ib wire\/0.
30 min M
25 Ib combustibles\/60
Ib wire
= 45, 300 scfh
ir 755 scfm
363. 3 scf
or
Area
12.6 scfs
Jl
12.6 scfs\/rl,760°R\144 in
30 fps
520 °R
ft
= 204 in.
5. Secondary air port size:
Design for 100 percent theoretical air through
secondary air ports. Inlet velocity is 900 fpm
or 0. 051 in. WC. From Table D6, Appendix
D, 177.2 scf of air is required per pound of
combustibles.
^250 Ib wireX /O. 25 Ib combustibles
I 30 min / V Ib wire
A/I77.2 scf \
A «> / =
368 scfm
Port area
/368 scfm\/144 in.
= ^900 fpm J\ {Z
= 59 in.
6. Equilibrium temperature between products of
combustion from ignition chamber at 1, 300 °F
and secondary dilution air at 60°F:
Weight of secondary dilution air:
, 25 Ib combustibles
(250 Ib wire\/0. ,
30 min )\
Ib wire
;\/13.51 lb\
A *
28. 1 Ib/min
Weight of products of combustion from igni-
tion chamber:
(250 Ib wire\/0. 25 Ib combustibles\/27. 96 Ib \ _
30 min j( Ib wire /( Ib / ~
30 min /I Ib wire
58. 2 Ib/min
(W )(C )[T - T ] = W (C )[T - Tj
a pa 2 a c pc 1 2J
where
W = -weight of secondary dilution air, Ib/min
a
C = specific heat of air, Btu/lb-°F
pa
T = final gas temperature, °F
T = initial gas temperature, °F
T = inlet air temperature, °F
W = weight of products of combustion, Ib/mir
-------
502
INCINERATION
C = specific heat of products of combustion,
P° Btu/lb-°F
(Z8. 1 lb/min)(0. 26 Btu/lb- °F)(T2 - 60°F) =
(58.2 lb/min)(0.26 Btu/lb- °F)(1, 300°F - TZ>
T = 870°F
7. Secondary burner (afterburner) capacity:
Design secondary burner to raise temperature of
products of combustion from ignition chamber
and secondary air from 870° to 1,600°F.
Assumed specific heat of products of combus-
tion is 0, 26 Btu/lb-°F.
and secondary burners for 20 percent excess
air. Total volume through mixing chamber:
Q = W C [T - T i
c p i. 1"
where
Q = heat required, Btu/min
W = weight of products of combustion from
ignition chamber and dilution air,
Ib/min
C = specific heat of products of combustion
from ignition chamber and dilution air,
Btu/lb-°F
T, = final temperature
T = initial temperature
Q = (28. 1 Ib/min + 58. 2 lb/min)(0. 26 Btu/lb- °F)
(1, 600°F - 870°F)
Q = 16, 400 Btu/min or 985, 000 Btu/hr
Design secondary burners for 20 percent ex-
cess air. From Table D7, Appendix D, the
calorific value of 1 scf of natural gas at
1, 600°F is 552. 9 Btu.
985,000 Btu/hr
Burner capacity = 552.9Btu/sCr = L
Gross secondary heat _ (1,775 scfh)(l , 100 Btu/scf)
Ib combustibles 125 Ib/hr
= 15, 600 Btu/lb
8. Mixing chamber (afterburner) cross-section-
al area.:
This area is also equivalent to the cross sec-
tion of the curtain wall port tunnel of a multi-
ple-chamber unit. Design for 30 fps at 1, 600°F
Products of combustion
from ignition chamber
45, 300 scfh
Secondary gas burners
(1,775 cfh)(13. 53 cf/cf) = 24, 000 scfh
Secondary air
j?t.o c ,
{368 scfrn
hr
Cross-section area
22,100 scfh
91, 400 scfh
or 1, 520 scfm
or 25.4 scfs
_ I25- 4 scfs\/2, 060°R\/144 in. 2 \
"( 30 fps }\ 520°R )\ 2 )
= 483 in.
9. Total length of secondary combustion chamber
(afterburner):
Assume cross-section area of curtain wall port
or tunnel is equal to cross section of mixing
chamber. Design for a residence time of 0.50
second.
Length of secondary zone = (30 fps)(0. 50 sec)
= 15 ft
10. Dilution air port at base of stack:
Design dilution air port to reduce temperature
of gases from mixing chamber from 1, 600°
tol,200°F. Specific heat of products of com-
bustion and dilution air is 0.26 Btu/lb-°F.
From Table D7, Appendix: D, there is 0. 999
Ib products of combustion of 1 scf natural gas
with 20 percent excess air.
Products of combustion
from ignition chamber
Secondary air through port
Secondary burner
(1, 775 scfh)(0. 999 lb/scf)/,--r .• •}
\o(J mini
Total gases:
= 58.2 Ib/min
= 28. 1 Ib/min
= 29.6 Ib/min
115. 9 Ib/min
WC [T - T ] = W (C ) [T - T 1
a pa 2 aj c pc L 1 2J
(See calculation No. 7.)
-------
Wire Reclamation
503
W (0.26 Btu/Tar-0F)(l,2000F - 60°F) =
cL
(115. 9 lb/min)(0.26 Btu/lb- °F)(1, 400°F - 1,200°F)
1, 140 W = 23,200
3,
W =20.4 Ib/min
a
Volume of dilution air:
^379 scf/lb mole
29 Ib/lb mole
/20. 4 lb\/:
\ min A
= 267 scfm
or 4. 45 scfs
Assume velocity through dilution air ports is
1, 255 fpm or 0. 10 in. WC velocity pressure.
Port area
_ /267 scfm\/1
>255fpmA
267 scfm\/144 in. \ _A , . 2
= 30. 6 in.
,255 fpm
11. Stack cross-sectional area:
Design for a velocity of 30 fps at 1, 200°F.
Total volume of flow:
Gases from mixing chamber
Dilution air at "base of stack
Stack cross-sectional area:
25.5 scfs
4. 4 scfs
29. 9 scfs
Select a 24-in.-ID stack
12. Stack height above grade:
Design for a static pressure of 0.20 in. WC
at base of stack.
(a) Theoretical draft of a 10-foot section at
1,200°F:
D = 0. 52 PHJ-f: -
f ' ^
where
D = theoretical draft, in. WC
P = absolute atmospheric pressure, Ib/in.
H = stack height, ft
T = temperature of stack gases, °R
T = temperature of air, °F
- (0.
52X14.7X10^ -7^o) = 0-
101 in. WC
(b) Stack friction for 10 -foot section:
.zt
F =
0. 008 H(V)
D T
where
F = friction, in. WC
H = stack height, ft
V = velocity, fps
D = stack diameter, ft
T = absolute stack temperature, °F
)2
(2.0)(1,660)
__ 0 008 (10)(30)
(c) Net effective draft for 10-foot section:
Net draft = a - b = 0. 101 - 0. 022
= 0. 079 in. WC
(d) Stack height, H, above grade:
H =
0.20 in. WC
0.079 in. WC/10 ft
= 25.4ft
*Kent, 193b.
tGriswold, 19U6.
-------
CHAPTER 9
COMBUSTION EQUIPMENT
GASfOUS AND LIQUID FUELS
ROBERT T. WALSH, Senior Air Pollution Engineer*
•GAS .AND OIL BURNERS
ROBERT T. WALSH, Seniox Air Pollution Engineer*
•BOILERS, 'H.EAT.ERS, AND ST'EAM GENERATORS
ROBERT T. WALSH, Senior Air Pollution Engineer*
*Now with New York-New Jersey Air Poll-ution Abatement Actwity, National Center for Air Pollution
Control, Public Health Service, U.S. Department of Health, Education, and Welfare, Raritan Depot,
Metuchen., New Jersey.
-------
CHAPTER 9
COMBUSTION EQUIPMENT
GASEOUS AND LIQUID FUELS
INTRODUCTION
For centuries, combustible materials containing
carbon and hydrogen have furnished man with his
most versatile source of heat and convertible
energy. Recent years have seen him, to a large
degree, weaned from the conventional solid fuels --
coal, wood, peat, and lignite--in favor of more
convenient gaseous and liquid hydrocarbons. Al-
though nuclear power and sunlight will probably
become increasingly prominent, hydrocarbons will
surely continue to provide a significant portion of
our domestic heat and power supply and our vehi-
cle fuels.
The burning of gaseous and liquid fuels is so com-
monplace that it enters directly into a vast num-
ber of air-polluting processes. Most boilers,
heaters, ovens, and driers are heated by the com-
bustion of hydrocarbon fuels. Many other process-
es use steam, hot water, or electrical energy gen-
erated from the burning of hydrocarbons.
Whenever hydrocarbon fuels are burned, gaseous
oxidation products are formed and, in almost ev-
ery case, vented to the atmosphere. Optimum
combustion of "clean" fuels, for example, natural
gas and lightweight oils, results in gases contain-
ing essentially water vapor, carbon dioxide, nitro-
gen, and oxygen—all normal constituents of the
atmosphere--as well as some oxides of nitrogen,
which are air contaminants. The burning of any
fuel under less than optimum conditions produces
some quantities of carbon, ash, and unburned
and partially burned hydrocarbons. In addition,
many fuels contain sulfur and metallic compounds
that are, even in the oxidized state, air pollutants.
The fuel picture is changing. Coal, a principal
solid fuel in some areas, but not in Los Angeles,
has less acceptance than it once enjoyed, because
of inherent drawbacks in material handling and
combustion, as well as to its tendency to create
greater quantities of air pollution. In many in-
stances where coal is employed on a large scale,
it is pulverized to facilitate handling and burning.
Moreover, treating coal to lower its ash and
sulfur contents has become commonplace. The
trend is away from high-sulfur, high-ash coals
and fuel oils and toward "cleaner" gaseous and
liquid fuels. In all fairness it must be reported
that coal producers are working vigorously to
regain their market^ by new techniques, such as
pipelining coal slurry, to eliminate certain pres-
ent disadvantages.
Gaseous Fuels
Most of the fuel gas consumed in the United States
is a naturally occurring mixture of low-molecular-
weight hydrocarbons, of which methane and ethane
predominate. Some natural gases from the well
contain hydrogen sulfide and other gaseous sulfur
compounds. Natural gas as marketed is, however,
extremely pure, so much so that sulfur compounds
are usually added to distribution lines (about 0. 15
grain, calculated as sulfur, per 100 scf) to impart
a detectable odor to the fuel. Because available
natural gas supplies often contain small quantities
of carbon dioxide and nitrogen and a varying ratio
of methane to ethane and higher hydrocarbons,
gross heating values range from 900 to 1,200 Btu
per scf in different localities. Analyses of some
natural gases are presented in Tables 138 and 139.
Table 138. COMBUSTION DATA SUMMARY FOR
A TEXAS NATURAL GAS
Analysis
Component
CO,
^2
°2
CH,
C2"6
C3Hg
L-C4H10
N-C4H1Q
CSHU
C^H| , and higher
% by volume
0
5
0
81
9
3
0
0
0
0
100
15
1 1
665
505
19
14
09
05
00
Gross heating value, 1, 100 Btu/scf
Combustion air requirement, scf/scf
Theoretical
2.0% excess
100% excess
10. 36
12. 43
.20. 72
Products of combustion, p»r 1 scf
Component At theoretical air
CO; 1. 134 set
HZ0
X ,
°2
Total
i . 0 8 3 scf
K. Z3b scf
0. 132 Ib
0. 0^9 Ib
0. 609 Ib
i
1 1. 453 sci 1 0. 340 Ib
At 20% excess air
1. 134 scf
2. 083 sit
9. 821 scf
0. 435 scf
13. 473 scf
0. 132 Ib
0. 099 Ib
0. 72b Ib
0. 037 Ib
0. 9Q-J Ib
507
-------
508
COMBUSTION EQUIPMENT
Table 139. COMBUSTION CHARACTERISTICS OF GASEOUS FUELS
Material
k*us-* hydrocarbons*
Hydrogen
Methane
Ethane
Ethyl ene
Propane
Propylene
felt an e
Natural jjases
Los Angeles, Calif
Birmingham, Ala.
Kansas City, Mo.b
Pittsburgh, Pa.b
Cracked, dry
Coking, dry
Reforming, dry
Cracked, dry
Fluid cat. , dry
Thermafor cat. . dry
Refinery, dry
Miscellaneous eases5
Coke oven
Density,
lb/ft3
at 60"F
0.0053
0.0422
0.0792
0.0746
0. 1162
0.1110
0.1530
0.0460
0. 0460
0.0483
0.0467
0.0572
0.0628
0.0795
0.0755
0.0776
0.0663
0.0740
0.0306
0.0414
Analysis. % by volume
H2
100
9.5
4.9
3.8
5. 5
19.5
3. 3
51.9
49. 6
TH
100
81. 1
90
84.1
83.4
64.5
44.6
27.5
40.2
31.7
24.6
36
32.3
10.9
C H
100
3.6
7.4
3. 3
7
8.2
5.4
7. 9
— -
C H
100
9.7
5
6.7
15.8
In
24.3
27.6
21.2
8.7
9.6
18.2
2.5
C1HA
100
1.9
1. 5
3
1. 1
15. 1
10
7. 5
C H
3. 2
6. I
",H
100
3. 5
6.7
14
22.4
23.8
24.7
20.6
19.7
CO
5. 5
27 5
21.9
C4HR
1.3
CO
L
3.6
C4Hin
100
0.4
2.9
i. 5
7.2
6.6
0.4
1.9
O
0.0!
0.4
C5H12,
0. 1
0.6
N
2
4.8
5
Inert
5.2
5
9.2
0.8
6.5
7.5
8
Heating value,
Btu/ft3
at 60T
Gross
325
1, 010
1,770
1,614
2,520
2,336
3,265
1, 100
1, 002
974
1, 129
1,316
1,463
1,745
1,617
1,609
1,384
1,540
569
536
Net
275
910
1, 619
1, 513
2,319
2, 186
3,014
990
904
879
1,021
1,200
1, 340
,592
,475
,470
,264
, 407
509
461
Theoretical
air require-
ment, ft3
dry air/ft3
fuel
2. 38
9.57
16. 75
14.29
23.90
21 44
31. 10
10. 36
9.44
9. 17
10.62
12. 34
14
16. 90
15.20
15.90
13.70
15.70
5.45
5. 05
CO£ in dry products
of combustion at
theoretical air,
% by volume
0
11.6
13. 1
15
13.7
15
14
11.9
11.8
11.8
12
11.5
--
--
--
10.8
25 5
14
kelson, 1958.
The North American Manufacturing Co., 1952.
In addition to natural gas, several other gases,
some mixtures, some pure compounds, are used
as combustion fuels. These range from by-prod-
uct and manufactured gases to liquefied petrole-
um gas (LPG). Typical analyses of available gas-
eous fuels are listed in Table 139. Some by-prod-
uct gases such as refinery "make gas" contain ap-
preciable percentages of higher molecular weight
hydrocarbons so that their heating values are
somewhat greater than those of natural gases.
Most by-product and manufactured gases contain
significant quantities of carbon monoxide and
inerts such as nitrogen and carbon dioxide, re-
sulting in heating values ranging from 100 to 600
Btu per scf.
Bottled liquefied petroleum gas consists of one or
a mixture of the following: Propane, propylene,
butane, and butylene. Because of its ease of
liquefaction and relatively high gross heating val-
ue--2, 520 to 3, 265 Btu per scf--the use of LPG
has been steadily increasing over the past few
decades. It finds its greatest application as nat-
ural gas standby fuel, as vehicular fuel, in porta-
ble equipment, and for general use in remote areas
to which piping less expensive fuels, such as nat-
ural gas, is not practical.
Oil Fuels
The term fuel oil applies to a wide range of liq-
uid petroleum products including crude oil, distil-
lates, and residuals. Most products marketed as
fuel oils have been refined to some degree to re-
move impurities and to fix upper and lower limits
of gravity, flash point, viscosity, and heating val-
ue. The sulfur and ash contents and the viscosity
are the major characteristics that affect air con-
taminant emissions.
Table 140 provides United States Bureau of Stan-
dards specifications for fuel oils. These stan-
dards often serve as guides in fuel selection rather
than as rigid limitations. Suppliers are likely to
market fuels that meet the needs of their locali-
ties and that are normal products of their partic-
ular crude oil stocks and refining processes.
These fuels frequently do not fit into any one of
the classifications listed in Table 140. Products
such as these are commonly sold under a com-
pany name such as Diesel Furnace Oil, Low-Sulfur
Stove Oil, or Light Crack Residual Oil.
In Table 140, Numbers 1 and 2 are distillate oils,
while Numbers 5 and 6 are residuals or "bottoms"
from refinery processes. Number 4 oils are like-
ly to be distillates or blends containing appreciable
distillate stock. The Bureau of Standards does
not list a Number 3 oil.
In general, the distillate oils contain appreciably
lesser concentrations of the potential air con-
taminants --sulfur and ash--than the more viscous
residuals do. This can be seen from the recom-
mended specifications in Table 140. It is a result
of the fact that most of the sulfur and ash in crude
-------
Gaseous and Liquid Fuels
509
Table 140. COMMERCIAL, STANDARDS FOR FUEL OILSa (Commercial Standard CS 12-48)
Grade of fuel oil
Number
1
2
4
5
6
Description
Distillate oil intended for vaporizing
pot-type burners and other burners
requiring this grade**
Distillate oil for general purpose do-
mestic heating for use in burners not
requiring No, 1
Oil tor burner installations not equip-
ped with preheating facilities
Residual-type oil for burner installa-
tions equipped with preheating facili-
ties
Oil for use in burners equipped with
preheaters permitting a high-viscosity
fuel
Flash
point,
"F
min
100
or legal
100
or legal
130
or legal
130
or legal
150
or legal
Pour
point,
'F
max
0
zoe
£0
Water
and
sedi-
ment,
%
max
Trace
0. JO
0.50
2 00#
Carbon
residue
on IO%
residuum,
%
max
0, 15
0. 35
Ash,
%
max
Gravity ,
"API
min
35
26
Distillation
temperatures, "F
10%
point
max
420
f
90%
point
max
475
End
point
max
625
Kinematic viscosity
Secoads.Sayibc'lt Cemxistofces at
Universal
at IQO'F
max
40
125
min
45
150
Furol
at 122'F
max
40
300
min
—
45
100 T
max
(26.4
—
min
(5-8)
(32.1)
(92)
122'
max
(81)
(638)
F
man .
—
aRecogmzing the necessity for low-sulfur fuel oils used in connection with heat treatment, nonferrous metal, glass, and ceramic furnaces, and other special uses,
a sulfur requirement may be specified in accordance with the following table:
Grade of fuel oil Sulfur max, %
No. 1 0. 5
No. i 1.0
Nos.4, 5, and 6 No limit
Other sulfur limits may be specified only by mutual agreement between the buyer and seller.
It is the intent of these classifications that failure to meet any requirement of a given grade does not automatically place an oil in the next lower grade unless in
fact it meets all requirements of the lower grade.
cGrade No. 3 became obsolete with the issuance of the 1948 commercial standard for fuel oils.
dNo. 1 oil shall be tested for corrosion in accordance with ASTM Designation D130-30 for 3 hours at 122'F. The exposed copper strip shall show no erav or
black deposit. G
eLower or higher pour points may be specified whenever required by conditions of storage or use. These specifications shall not, however, require a pour
point lower than 0°F under any conditions.
fThe 10% point may be specified at 440°F maximum for use in other than atomizing burners.
«The amount of water by distillation plus the sediment by extraction shall not exceed 2. 00%. The amount of sediment by extraction shall not exceed 0. 50. A
deduction in quantity shall be made for all water and sediment in excess of 1. 0%.
oil is tied up in long-chain, high-boiling-point or-
ganic compounds, which tend to concentrate in
residuals from refinery processes. Moreover,
most effective sulfur-removing processes are
adaptable only to low-viscosity distillate oils.
Table 141 provides combustion data for a U.S.
Grade 6 residual fuel oil. Residual fuel oils are
markedly less expensive than distillate oils but
require more elaborate burner equipment for
proper combustion. "Heavy crack" residual fuels
are normally burned, therefore, only in large
combustion sources. Most small operators, par-
ticularly those who burn natural gas on a curtail-
able basis, prefer to use cleaner, more trouble-
free distillate oils as stand-by.
THE AIR POLLUTION PROBLEM
Air contaminants generated from fuel burning fall
into three categories: (1) Carbon and the unburned
and partially oxidized organic materials that re-
sult from incomplete combustion, (2) sulfur ox-
ides and ash directly attributable to fuel composi-
tion, and (3) oxides of nitrogen formed at firebox
temperatures from oxygen and nitrogen of the air.
Incomplete combustion products can usually be
held to tolerable minimums with proper operation
of modern burner equipment. Sulfur and ash emis-
sions are governed by the fuel makeup. Nitrogen
oxide concentrations are primarily functions of
firebox design and temperature.
Black Smoke
When hydrocarbon fuels are burned in a deficien-
cy of oxygen, some carbon particles can be found
in the products of combustion. With poor fuel
atomization, inadequate mixing, or marked oxy-
gen shortage, the concentration of carbon in-
creases to the point where a visible blackness is
imparted to exit gases. Black smoke, -when it
occurs, is usually connected with the burning of
viscous, heavy-crack residual oils and of solid
fuels. Creating black smoke by burning gaseous
fuels is difficult, though not impossible. Other
products of incomplete combustion, such as
carbon monoxide, usually accompany black smoke
emissions. The degree of blackness is historic-
ally significant, since the Ringelmann Chart was
developed for this type of smoke. Heavy, carbo-
naceous accumulations in exhaust stacks, com-
monly termed soot, are attributable to the same
cause as black smoke, namely, poor combustion.
White Smoke
Visible emissions ranging from grey through
brown to white can also be created in the com-
bustion of hydrocarbon fuels, particularly liq-
-------
510
COMBUSTION EQUIPMENT
Table 141. COMBUSTION DATA SUMMARY FOR
A NUMBER SIX FUEL OIL
Analysis
Component
Carbon
Hydrogen
Sulfur
Water
Ash
% by weight
88. 3
9. 5
1. 6
0. 05
0. 10
Gross heating value
152, 000 Btu/gal
18,000 Btu/lb
Combustion air requirement (dry)
Theoretical
10% excess
20% excess
100% excess
scf/lb
176. 3
193. 9
211.6
352. 6
Ib/lb
13. 4
14. 8
16. 1
26. 9
Products of combustion, per Ib of fuel oil
Assume air at 40% RH, 60°F
Component
C°2
H2°
so2
°2
Total
At theoretical air
27. 9 scf
19. 3
0. 2
139. 3
---
186. 7
3. 24 Ib
0. 92
0. 03
10. 30
---
14. 49
At 20% excess air
427. 9 scf
19. 5
0. 2
167. 5
7. 4
222. 5
3. 24 Ib
0. 93
0. 03
12. 38
0. 64
17. 22
uid fuels. White or opaque smoke, that is, non-
black smoke, is the result of finely divided par-
ticulates--usually liquid particulates--in the gas
stream. These visible pollutants are most often
caused by vaporization of hydrocarbons in the
firebox, sometimes accompanied by cracking, and
subsequent condensation of droplets at 300° to
500°F stack temperatures. White smoke is fre-
quently attributed to excessive combustion air
(cold fire) or loss of flame (gassing). Visible
contaminants can also exist where combustion is
optimum and the concentration of oxidizable ma-
terials is at a minimum. This situation is ap-
parently limited to large power plant boilers where
there is measurable sulfur trioxide in exhaust
gases.
Participate Emissions
Combustion gases can contain particulate matter
in the form of unburned carbon and hydrocarbon as
well as inorganic ash. With the proper use of ade-
quate burner equipment, oxidizable particulates,
both solids and liquids, can usually be kept well
below typical emission standards, for example,
Rule 53b of the Los Angeles County Air Pollution
Control District allows 0. 3 grain of particulate
matter per scf of exhaust gases calculated on a
12 percent carbon dioxide basis. Where unburne
particulate concentrations approach allowed limii
the Ringelmann number or opacity of the exhaust
gases is usually high and may exceed legal stan-
dards for visible contaminants. The operator of
combustion equipment malfunctioning in this "way
can almost always correct the combustion condi-
tions to control these emissions unless the grade
of fuel is improper for the combustion equipment
or vice versa. Ash collected at large, efficient
power plant boilers during oil burning normally
contains less than 10 percent carbon and other
combustibles.
The quantity of inorganic solid particulate s in ex-
haust gases is entirely dependent upon the charac-
teristics of the fuel. There is no measurable in-
organic ash in exhaust gases from the combustion
of natural gas or other clean gaseous hydrocarbon
except for that small quantity attributable to the
dust usually present to some degree in all air usec
for combustion. Distillate fuel oils do not contain
appreciable amounts of ash. Typical analyses
show variations from a trace to about 0. 03 per-
cent by weight. In residual oils, however, inor-
ganic ash-forming materials are found in quantities
up to 0. 1 percent by weight. Most of this material
is held in long-chain organo-metallic compounds.
The strong oxidation conditions present in most
fireboxes convert these materials to metallic ox-
ides, sulfates, and chlorides. As would be ex-
pected, the compounds show up as finely divided
particulates in exhaust gases. Table 142 provides
a spectrographic analysis of the inorganic fuel oil
ash collected at a large power plant boiler.
The combined ash and unburned particulates in
exhaust gases from gaseous or liquid fuel com-
bustion are not likely to exceed local air pollution
control statutes. For instance, the efficient burn-
ing of a common heavy residual oil of 0. 1 percent
ash results in a stack gas concentration of only
0. 030 grain per scf at 12 percent carbon dioxide.
Sulfur in Fuels
In liquid hydrocarbon fuels, sulfur occurs in con-
centrations ranging from a trace to more than 5
percent by weight. Much of this sulfur is present
as malodorous sulfides and mercaptans. Natural
gas fuels contain very little sulfur as marketed,
usually only enough to impart a detectable odor to
the gas. Some by-product gases, however, contain
appreciable sulfides and mercaptans. Distillate
oils may contain as much as 1 percent sulfur,
though most distillates have less than 0. 3 percent.
There is normally much more sulfur in heavy resid-
ual oils than in gaseous fuels and distillate oils.
-------
Gaseous and Liquid Fuels
511
Most of these oils contain more than 1 percent
sulfur by weight. In the Los Angeles area the heavy
residual oil commonly burned in power plant boil-
ers averages 1. 6 percent sulfur. The only low-
sulfur (less than 0. 5 percent) residual oils avail-
able are those resulting from the refining of low-
sulfur crude oils, which are relatively rare.
Table 142. TYPICAL FUEL OIL ASH ANALYSIS
Constituent
Iron
Aluminum
Vanadium
Silicon
Nickel
Magnesium
Chromium
Calcium
Sodium-
Cobalt
Titanium
Molybdenum
Lead
Copper
Silver
Total
Weight %
22.99
21. 90
19. 60
16.42
11. 86
1. 78
1. 37
1. 14
1
0. 91
0. 55
0. 23
0. 17
0. 05
0. 03
100
Sulfur Oxides
Most of the sulfur present in fuels is converted to
sulfur dioxide on combustion. A typical residual
fuel oil of 1. 6 percent sulfur yields a concentra-
tion of 1, 000 ppm sulfur dioxide when burned with
the theoretical amount of combustion air. As
shown in the sample calculations, this is equiva-
lent to 832 ppm at 20 percent excess combustion
air, a point at which many industrial boilers are
operated.
In some combustion processes, a small portion
of the sulfur--usually no more than 5 percent of
the total — is converted to sulfur trioxide, the
anhydride of sulfuric acid. Sulfur trioxide is
highly reactive and extremely hygroscopic as
compared with sulfur dioxide. It is considered
a chief cause of the visible plume created by
burning high-sulfur fuel oils in large power plant
boilers. Besides obscuring visibility, these con-
taminants can result in acid damage to vegetation
and property in downwind areas. The factors
governing firebox formation of sulfur trioxide are
not fully understood, but it is recognized to occur
principally in large combustion installations oper-
ated at high firebox temperatures.
Oxides of Nitrogen
In every combustion process the high temperatures
at the burner result in the fixation of some oxides
of nitrogen. These compounds are found in stack
gases mainly as nitric oxide (NO) with lesser
amounts of nitrogen dioxide (NO£) and only traces
of other oxides. Since NO continues to oxidize to
NO-; in the air at ordinary temperatures, there is
no way to predict -with accuracy the amounts of
each separately in vented gases at a given time.
The total amount of NO + NO2 in a sample is de-
termined and referred to as "oxides of nitrogen"
or NOX (Los Angeles County Air Pollution Control
District, 1960a).
AIR POLLUTION CONTROL METHODS
An operator can take only two options to reduce
air contaminant emissions from a combustion
source, namely, remove the sulfur compounds
and ash from, his combustion gases or switch to a
cleaner and usually more expensive fuel. Only
limited progress has been made in removing air
pollutants from combustion products. These meth-
ods are discussed later in this chapter in the sec-
tion, "Boilers, Heaters, and Steam Generators, "
inasmuch as they are employed only at large com-
bustion sources.
Prohibitions Against Sulfur Emissions
Two types of regulations have been used to limit
the concentration of sulfur contaminants at com-
bustion sources and thus outlaw the burning of
certain fuels. One sets a ceiling on the fuel's sul-
fur content while the other fixes a maximum allow-
able flue gas concentration. Both types of pro-
hibitions have been enacted in Los Angeles County.
Rule 53a limits the concentration of sulfur com-
pounds in exhaust gases from any combustion pro-
cess to 0. 2 percent by volume calculated as sul-
fur dioxide. Rules 62 and 62. 1 prohibit the burn-
ing of gaseous fuels containing more than 50 grains
of sulfur compounds per 100 cubic feet, calculated
as hydrogen sulfide, and liquid fuels containing
more than 0. 5 percent sulfur by weight. Rules
62 and 62, 1 are considerably more stringent than
Rule 53a, limiting sulfur concentrations to rough-
ly one-eighth of that allowed by Rule 53a. Until
practical stack control methods are developed,
either type of regulation will be effective in pro-
hibiting the burning of high-sulfur fuels, on the
assumption that the chosen concentrations are
comparable.
Supplementary provisions are sometimes used
wherein allowable stack sulfur emissions are
based upon ground level concentrations, usually
measured as sulfur dioxide. The regulations of
the San Francisco Bay Area Air Pollution Control
District, for instance, set an effluent limit ot" 0. 2
-------
512
COMBUSTION EQUIPMENT
percent as sulfur dioxide but allow heavier dis-
charges of sulfur compounds, provided .ground level
conceatratiDiis do not exceed specified limits. The
allowed concentrations are on a sliding time basis
ranging from 3 minutes at 1. 5 pprn to 8 hours at
Q. 3 ppm, sulfur compounds being measured as sul-
fur dioxide. Obviously, statutes such as these allots
heavier emissions where dispersion conditions are
favorable and where the source is isolated from
similar sulfur-emitting plants.
Removal of Sulfur and Ash from Fueli
To •whatever degree is economically feasible,
hydrocarbon fuels are treated to remove sulfur
compounds as well as inorganic ash. The prac-
tical sulfur removal methods are essentially
restricted to scrubbing or liquid-liquid extrac-
tion, sometimes accompanied by catalytic de-
composition. Natural gas is commonly scrubbed
at the natural gas plant before its introduction
into transmission pipe lines. Higher molecular
•weight hydrocarbon gases and distillate oils are
treated at the refinery before they are marketed.
There is at present no economical method of re-
moving sulfur from heavy residual oils. As was
previously mentioned, most of the sulfur in these
viscous oils is tied up in large molecules. To
remove the undesirable sulfur, one must remove
it from the molecule, as with hydrocracking or
thermal cracking processes, and thereby cre-
ate hydrocarbons of considerably lower molecular
weights. Both these processes add appreciably
to fuel costs and are now used to yield products
such as gasoline and distillate oils, which com-
mand markedly higher prices than residuals.
The ash tends to concentrate in the residuals.
Apparently, on the basis of present technological
trends, yields of residual oils from refineries
will be steadily reduced in coming years. New
processes being developed and perfected are
aimed primarily at greater yields of gasoline and
diesel and aircraft fuels. Much more •work is
being done on the development of these processes
than on methods of removing sulfur from highly
cracked residual oils.
Figure 350 shows production trends for residual
and distillate oils over the period 1943 through
1962. Obviously, residual yields dropped off
sharply while distillate production increased in
inverse proportion. The yield curve for resid-
ual oil is expected to continue to decline as ex-
isting refinery equipment is replaced. Much of
the oil listed as distillate in Figure 350 is used
as motor fuel, "while almost all the residual oil
is burned in boilers and heaters. Over the 20-
year period from 1943 to 1962, refinery crude
oil input increased approximately twofold. Resid-
ual production decreased, therefore, in terms
both of volume and yield in that interval. This
situation has been offset somewhat by the in-
creased importation of residual oil.
Illustrative Problem
The following example illustrates calculations
used in determining sulfur oxides and ash con-
tent in flue gases formed txy burning a heavy
residual fuel.
Example 34
Given:
A heavy residual fuel is to be burne-d in a com-
bustion process -with 20 percent excess air. The
oil analyses, % by weight, is as follows:
Carbon
Hydrogen
Sulfur
Water
Ash
88. 3
9.5
1.6
•0. 05
0. 10
Problem:
Determine the combustion air requirement and
the concentration of sulfur oxides in £hie gases
while assuming 3 percent of the sulfur is con-
verted to sulfur trioxide. Determine the ash con-
centration in flue gases at 12 percent carbon di-
oxide while assuming complete combustion. Use
as a basis 1 pound of fuel oil.
Solution:
1. Theoretical combustion air requirement:
Carbon C + O.,
CO,
(0.883 lb){ —J = 2.351bO = 132. 5 scf of air
Hydrogen
+ 1/2
(0. 095 lb)|—J = 0. 76 Ib O = 42. 9 scf of air
Sulfur S + O_
SO,
(0.016){ —) = 0. 016 Ib O = 0.9 scf of air
Total
176. 3 scf of dry air/lb oil
177.6 scf air at 40% RH, 60°F per Ib oil
-------
Gaseous and Liquid Fuels
513
1943
1945
1947
1949
1951
1953 (955
YEAR
(957
1959
1961
1963
Figure 350. Production trends, U.S. refineries, 1943-1962.
2. Air requirement at 20% excess air:
(176.3)(1.20) = 212 scf dry air/lb oil
(177. 6)(1. 20) = 213 scf moist air/lb oil
3. Products of combustion (assume complete
combustion and neglect 803):
Carbon dioxide
379 scf
-\
(44 lb/mol/
Water from combustion
(0.0951b)(18^3?9y
Water in fuel:
0.0005 Ib
= 18. 0 scf
0.011 scf
Nitrogen
(212 scf)(0.79) = 167.5 scf
Water in air, 40% RH, 60°F
(0. 0072 scf/scf air)(213 scf) = 1.5 scf
Sulfur oxides as sulfur dioxide
0.2 scf
0.0161 —
Oxygen
(176.3 scf)(1.20 - 1.00)(0. 21) = 7. 4 scf
Total
4. Sulfur dioxide concentration:
222.5 scf/
Ib oil
(10 )(0. 97) = 827 ppm
-------
514
COMBUSTION EQUIPMENT
5. Sulfur trioxide concentration:
(0. 016)
6. Inorganic ash concentration:
(0.001 lb)(7, 000 gr/lb)/~|-j\ = 0. 0314 gr/scf
GAS AND OIL BURNERS
INTRODUCTION
A burner is essentially a triggering mechanism
used to ignite and oxidize hydrocarbon fuels. In
general, burners are designed and operated to
push the oxidation reactions as close as possible
to completion with the maximum production of
carbon dioxide and water, leaving a minimum of
unburned and partially oxidized compounds in ex-
haust gases. Burner efficiency can be measured
by the water and carbon dioxide contents of com-
bustion gases or, conversely, by the concentrations
of carbon monoxide, carbon, aldehydes, and other
oxidizable compounds. Insofar as hydrocarbon-
derived pollutants are concerned, optimum burn-
er operation goes hand in hand with minimum air
pollution.
The purpose of this part of the chapter is to pre-
sent general burner principles with emphasis on
major design and operation variables that affect
air pollution. There are so many variations in
burner design that discussing each separately
would not be practical. Specific operating instruc-
tions for any given burner should be obtained from
the manufacturer or agent.
Burners and the combustion equipment in which
they are located are commonly thought of as
sources of air pollution. Burners, however, are
also used frequently as air pollution control equip-
ment. Their; most common control application is
in vapor incineration, but many are used for pur-
poses such as refuse incinerator auxiliaries and
as tempering heaters with baghouses, precipita-
tors, and centrifugal collectors. Almost all
burners used in air pollution control devices are
designed to handle gaseous fuels exclusively.
A burner consists primarily of a means of meter-
ing the two reactants, oxygen and fuel, and a
means of mixing the reactants before and con-
currently with ignition. Many burners also in-
clude flame safety devices and auxiliaries to condi-
tion the temperature and viscosity of the fuel,
as well as fans and pumps to move or pressurize
air or fuel. The simplest burners are employed
•with gaseous combustion fuels while the most
complex units are used with heavy oils and with
solid fuels.
Draft Requirements
In all combustion equipment, some energy is re-
quired to push or pull the combustion air, fuel,
or products of combustion through the burner and
also through the heat exchange portion of the com-
bustion equipment. With small gas-fired appliances
the line gas pressure together •with the bouyancy of
warm oxidation products are sufficient to provide
the necessary draft. With larger equipment, either
extended natural-draft stacks or blowers must be
used. Blowers may be positioned ahead of or be-
hind the firebox. When located ahead of the fire-
box, a blower is sometimes constructed as an in-
tegral part of the burner and is driven by a motor
common to a fuel pump or atomizing device. Forced
draft burners provide greater flexibility and can
be used in situations where the firebox itself is
under pressure.
Gas Burners
Owing principally to the low viscosity of gaseous
fuels, gas burners are considerably simpler than
those used with liquid and solid fuels. Gases can
be transmitted and mixed •with combustion air much
more easily than other fuels can. This is not to
say that all gas burners are necessarily uncompli-
cated mechanisms. Many are equipped with elabor-
ate combustion air auxiliaries and flame control
features. For a specific installation, however, a
gas burner is almost always less complicated than
its liquid or solid fuel-burning counterpart de-
signed for the same application.
In most gas burners, only a portion of the air re-
quirement—termed primary combustion air--is
mixed with fuel before ignition. These burners
constitute the large majority of equipment in use
today, ranging from small appliances to large
power plant gas burners. Two other types in rea-
sonably wide use do not fall into this category--
totally aerated burners and nonprimary aerated
burners.
With totally aerated burners, all combustion air
is mixed with fuel before ignition. These units
are employed at installations such as metallurgi-
cal furnaces, -where operation within narrow oxy-
gen concentration limits or even in reducing at-
mospheres is desirable.
In nonprimary aerated burners, no combustion air
is mixed with fuel ahead of the burner port. The
gaseous fuel is merely allowed to jet through an
orifice in such a pattern or manner as to provide
adequate mixing with oxygen. .Vlost of these units
GPO 806—614-18
-------
Gas and Oil Burners
515
employ narrow slotted ports, giving the flame a
thin fan shape. In other nonprimary aerated burn-
er.s, a circular orifice is employed, and the jet-
ted fuel is allowed to impinge on a target surface
in such a manner as to provide turbulence and mix-
ing. Many nonprimary aerated burners are of
multiport design, employing a number of slots or
orifices in order to provide maximum interface
surface between fuel and combustion air.
Partially Aerated Burners
The venturi-shaped burner in Figure 351 can be
used to illustrate the basic operation of partially
aerated atmospheric gas burners. Gaseous fuel
is introduced through the control valve into the
burner head and allowed to flow through the fixed
orifice into the throat. The jetted gas stream in-
duces combustion air to flow through the primary
airport and creates enough turbulence to mix fuel
and air between the orifice and the burner tip.
The quantity of primary air induced is governed
by the airport setting, the specific gravity of the
gas, and the gas pressure. Ignition starts at the
"burner tip where additional air--termed secondary
combustion air--contacts the mixture. Combus-
tion is completed off the burner tip as additional
secondary air reacts with the burning mixture.
yellow tip curve results in a smoky flame with
possible flashback. Natural gases are relative-
ly slow burning and are not likely to flash back
unless conditions are severe.
NATURAL GAS ANALYSIS
CH4 -- 84 50-4
C2H6 = 14 50
C02 = 0 20
Op = 0 20
Figure 351. Typical atmospheric gas burner.
0 10 20 30 40 50 60 "70
INPUT RATE, 1,000's of Btu/hr per in.2 of port area
Figure 352. Flame stability limits burning
a natural gas in an atmospheric burner
(American Gas Association Laboratories, 1940).
For a given fuel, the combustion efficiency and
the stability, shape, and luminosity of the flame
are dependent upon the primary and secondary air
rates and the degree of turbulence. A high pri-
mary air rate produces a short, blue flame, while
a low primary air rate results in a long, luminous
flame. If primary air is reduced too greatly, the
flame becomes smoky with yellow tips, and flash-
back may occur out the primary combustion air-
port. If the primary air rate is increased too
much, the flame becomes unstable and lifts from
the burner port. These limits are plotted in Fig-
ure 352 for an 1, 100 Btu per cubic foot natural
gas. The cross-hatched area between the two
curves represents the stable range of burner oper-
ation for a typical partially aerated burner. Oper-
ation above the lift curve results in the flame's
lifting from the burner, while operation below the
The effect of primary air at the same gas input
is illustrated in Figure 353 for the same natural
gas described in Figure 352. At the maximum
primary air rate shown, 66. 8 percent of the the-
oretical combustion requirement, the inner blue
cone of the flame is sharply defined while the out-
er luminous cone is almost indistinguishable at
the tip. At the lowest primary air rate, 49. 1 per-
cent, the flame becomes extremely luminous, the
inner blue cone blending into the luminous outer
cone.
The burner characteristics of different fuel gases
are dependent to a large degree upon speeds of
flame propagation. Gases such as hydrogen, car-
bon monoxide, ethylene, benzene, and propylene,
with high ignition velocities, are prone to flash-
back through the burner at low primary air rates.
-------
516
COMBUSTION EQUIPMENT
66.8
63.4
60.4 57.1
'i PRIMARY AIR
53.3
Figure 353. Natural gas flames with varying primary air (American Gas Association
Laboratories, 1940).
Nevertheless, the latter fast-burning gases do not
tend to blow off or lift from the burner tip as read-
ily as the slower burning fuels, methane, ethane,
and butane. Gases with high ignition velocities are,
therefore, normally operated at somewhat higher
primary air rates than natural gas and liquefied
petroleum gas are. This can be seen by compar-
ing the stability range of the fast-burning manu-
factured gas of Figure 354 with that of the natural
gas of Figure 352. The lift curve for the manufac-
tured gas is considerably higher than for natural
gas. For example, at 70 percent primary air and
30, 000 Btu per hour per square inch of port area,
the manufactured gas flame is stable, while that
of natural gas is unstable. The yellow tip curve
for this gas is also higher. Its marked propensity
to burn back out the airport is shown by the flash-
back limit curve.
Other factors, such as port size and shape, also
influence bvirner operation. The reader should con-
sult a burner handbook and publications of the Amer-
ican Gas Association for detailed discussions of the
subject.
Multiple-Port Gas Burners
Burners with multiple orifices are widely used
in boilers, heaters, and vapor incinerators. The
individual ports are usually of partial-aeration or
nonprimary-aeration design. Over a given cross-
section, a multiple-port burner provides better
distribution of flame and heat than a single-port
unit does. For this reason, multiple-port burners
have an inherent advantage in vapor incineration.
Forced-Draft Gas Burners
The availability of a combustion air blower pro-
vides greater flexibility and often better combus-
tion than an atmospheric gas burner affords. The
simplest forced-draft units consist merely of low-
pressure fans with gaseous fuel orifices located
in the discharges. In some cases, the fuel is in-
troduced ahead of the blower and allowed to mix
in the fan housing. One of the more complex de-
signs is the low-pressure premix unit, shown in
Figure 355. Here, a blower is used to force
combustion air through a venturi at pressures up
-------
Gas and Oil Burners
517
110
= 3d
= 23 75
= 17 45
= 3 35
= 0 85
= 15 20
ILIUM* = 8 45
100 00
IUUMINANTS-OLEFINS AND
AROMATICS
DAMPER
10 20 30 40 50 60
INPUT RATE, MBTU'hr-in2 of port area
Figure 354. Flame stability limits burning a
manufactured gas in an atmospheric burner
(American Gas Association Laboratories, 1940).
to 3 psig. Gaseous fuel is drawn into the system
at the throat of the venturi and mixes in fixed
proportion with combustion air ahead of the burn-
er nozzle. With an arrangement such as this, the
shape, makeup, and luminosity of the flame can
be precisely controlled. Moreover, the flame
has appreciable velocity. These burners are em-
ployed in metallurgical processes where precise
atmospheric control is desired, in some vapor
incinerators, and in crematories and pathologi-
cal-waste incinerators, where a strong flame
must be directed on animal tissue.
Gas Flow Rates
Gaseous fuel is commonly introduced through one
or more fixed orifices at the burner. These ori-
fices constitute the principal pressure drop in the
gas-piping system and govern the flow of fuel to
the burner. Flow through an orifice is propor-
tional to the square root of gas pressure so that
minor upstream pressure fluctuations do not have
AIR-
BIOKER
GAS IN
FIREBOX
ZEDS-PRESSURE
REGULATOR
Figure 355. A multiple-port burner (nonpnmary
aerated) installed in a vapor incinerator.
a great effect on flow rate. The nomographs of
Figures 356 and 357 provide flow rates for 0. 65
specific gravity {referred to air) natural gas
through standard orifices at various gas pressures.
Oil Burners
Inasmuch as liquid fuels must be vaporized be-
fore combustion can take place, an oil burner
must accomplish an additional function not re-
quired of a gas burner. Light oils can be vapor-
ized from a static vessel or wick. This princi-
ple is used with items such as kerosene lamps
and blow torches but is not practical for most
burners. In almost all industrial applications,
the fuel is first atomized then allowed to vapor-
ize on absorbing heat from the flame. The effi-
ciency of an oil burner depends largely upon
atomization and fuel-air mixing.
There are four basic types of oil burners, differ-
ing principally in the methods of atomization: Low-
pressure air-atomizing; high-pressure steam- or
air-atomizing; high-pressure oil-atomizing; and
centrifugal or rotary cup burners. A fifth type,
the mechanical atomizing burner, employs both
high-pressure oil and centrifugal action.
With low-pressure, air-atomizing burners, such
as that shown in Figure 358, a major portion of
the combustion air requirement is supplied near
the oil orifice at 1/2 to 5 psig. This air abrades
and atomizes the jetted oil stream in an area of
high turbulence. Secondary combustion air is
admitted around the periphery of the mixture. In
comparison with other types of oil burners, these
units provide a greater volume of air in close
proximity to the atomized oil--from 10 to 60 per-
cent of the theoretical combustion requirement.
For this reason, the flame is reasonably short.
-------
518
COMBUSTION EQUIPMENT
Alt
TZST/Mf
<•*
~" — ^/Mrf*
— ___ OTf*w
** 4 / us, •***».
I n
Ml y ~ 1_^^
f 4_:E-?-^
0/v
Cl/B/C
§ E =
«*-J
S F
^
=5$
5 «1 =
«*/-=
8
i
K
— fOO
-If
-20
-2S
-So
!
, Si
-- ^N
± ^
Figure 356. Natural gas flow through standard orifices (Southern California Gas Co.).
-------
Gas and Oil Burners
519
*
j|-/»
=—#
§
&V
Ct/S/C
CO.
CO.
as
*
5
,
/
-------
520
COMBUSTION EQUIPMENT
-AIR
OIL
Figure 358. Low-pressure, air-atomizing oil burner
(Hauck Manufacturing Co., 1953).
With the high-pressure, steam- or air-atomiz-
ing burners of Figure 359, an auxiliary fluid--
steam or air--is used to break up the fuel oil
stream at the burner tip. The auxiliary fluid,
moving at high velocity, atomizes the slower
moving oil stream as the mixture passes into
the burner tile. The atomizing fluid is provided
at pressures ranging from 30 to 150 psig. The
volume of atomizing air, when used, is normally
much smaller than that encountered with low-
pressure, air-atomizing burners. Compressed-
air consumption ranges from 30 to 200 cubic
feet of free air per gallon of oil, that is, from
2 to 14 percent of the theoretical combustion
requirement. These burners are reasonably in-
expensive and are likely to be employed where oil
is burned only infrequently, as on a standby
basis. Steam-atomizing burners perform satis-
factorily at viscosities of 150 to 200 Sabolt Sec-
onds Universal (hereafter referred to as SSU).
Air-atomizing burners require lower viscosi-
ties, usually 30 to 100 SSU.
Oil pressure atomizing burners depend upon high
fuel pressure (75 to 150 psig) to cause the oil to
break up into small droplets upon passing through
an orifice. The fixed orifices of these units are
considerably smaller than those used with other
HOLE FOR PILOT TIP
TILE
CLEAN-OUT PLUG
OIL VALVE / PACKING
OIL INLET
STEAM OR COMPRESSED
AIR INLET
REGISTER
Figure 359. High-pressure, steam- or air-
atomizing oil burner (North American Manu-
facturing Co., 1952).
types of oil burners. An inherent disadvantage is
that the burner atomizes properly only over a fair-
ly narrow pressure range.
Mechanical atomizing burners are the most com-
mon oil burners found at la.rge power plant steam
generators. In the wide-range mechanical atom-
izing assembly shown in Figure 360, the fuel oil
is given a strong whirling action before it is re-
leased through the orifice. Proper atomization
is dependent upon centrifugal velocities, which in
turn require high pressures, that is, 100 to 200
psig. The wide-range unit of Figure 361 over-
comes a principal disadvantage of this type burn-
er, namely a narrow turndown ratio. In the burn-
er, some of the whirling oil flows through the ori-
fice while excess oil is drawn off through the cen-
tral oil return line.
Figure 360. Wide-range mechanical atomizing
burner (de Lorenzi, 1947).
Rotary cup burners, such as that shown in Figure
362, provide atomization by centrifugally throw-
ing the fuel from a rotating cup or plate. Oil is
distributed on the cup in a thin film. As with oil
pressure atomizing burners, no air is mixed with
the oil before atomization. Combustion air is ad-
mitted through an annular port around the rotary
cup. These burners are usually constructed with
integral forced-draft blowers. A common motor
often drives the oil pump, rotating cup, and blower.
Rotary cup burners can be used to burn oils of
widely varying viscosity, ranging from distillates
to residuals of greater than 300 SSU.
Viscosity and Oil Preheaters
The key to optimum oil burner operation is care-
ful control of fuel viscosity. A given burner func-
-------
Gas and Oil Burners
521
TIP
FERRULE VGASKET
Figure 361. A wide-range mechanical-atomizing
assembly with central oil return line(de Lorenzi
1947).
MOUNTING
HINGE
MOTOR
OIL
AIR
Figure 362. Rotary cup oil burner (Hauck
Manufacturing Co., 1953).
tions properly only if the viscosity at the "burner
orifice is held between fairly narrow limits. If
the viscosity is too high, effective atomization
does not take place. If the viscosity is too low,
oil flow through the orifice is too great, upsetting
the balance between combustion air and fuel.
There are several viscosity measurement scales.
At viscosities of less than 100 SSU, fuel oils can
be burned efficiently in almost any burner. Most
burners are designed for optimum performance
at 150 SSU or lower. All distillate oils and some
blends are of less than 100 SSU at 60°F, as shown
in Figure 363. Where fuel oil viscosity at am-
bient temperature is not compatible with the burn-
er, preheaters are used. With the chart provided
in Figure 363, fuel oil viscosities can be estimated
at different temperatures. The sloped lines rep-
resent fuels with average viscosity-temperature
relationships. When the viscosity at a given tem-
perature is known, viscosities at other tempera-
tures can be predicted by extending lines of paral-
lel slope. The chart also allows conversion from
different viscosity scales.
Oil preheaters may be mounted directly on the
burner, at the supply tank, or just about any
place in between. Preheater selection is depen-
dent to a large degree upon the fuel itself. Most
heavy residual oil must be warm to allow pump-
ing. A preheater for such oil is likely, therefore,
to be located at or near the supply tank. With
lower viscosity oils, preheaters are often located
at the burner, preheat temperatures are lower,
and the heaters are normally smaller and less
complicated.
Oil preheaters are operated with either electricity
or steam. Electrical heaters allow a greater de-
gree of flexibility. They can be used at times when
the combustion equipment is cold and no steam is
available. Where only steam preheat is used, an
auxiliary source of steam independent of the com-
bustion equipment on which the burner is located
should be available. If an oil burner is ignited
from a cold start, and the oil is not preheated to
its normal temperature, igniting the burner is
often difficult or impossible. Excessive air con-
taminants can be expected from this practice.
THE AIR POLLUTION PROBLEM
The burning of combustion fuels can produce sul-
fur oxides, inorganic ash, oxides of nitrogen,
carbon, and unburned and partially oxidized hy-
drocarbons. Most of these contaminants, notably
sulfur oxides and inorganic ash, are attributable
directly to the fuel and are independent of equip-
ment design or operation. The principal air con-
taminants affected by burner design and operation
are oxidizable materials --carbon, carbon monox-
ide, aldehydes, organic acids, and unburned hy-
drocarbons. To a lesser degree, burner design
can also affect oxides of nitrogen, but these emis-
sions are dependent largely upon the design of the
furnace and other combustion equipment.
Smoke and Unburned Contaminants
Modern burner equipment has been perfected to
the point where all common fuels can be burned
without causing excessive discharges of oxidiz-
able materials in exhaust gases. If the proper
combination of burner and fuel has been selected,
and if the burner is operated properly, no visible
emissions should be caused by oxidizable air con-
taminants, and the concentrations of items such
as aldehydes and carbon monoxide should be neg-
ligible. Nevertheless, smoke and oxidizable mate-
rials are often found in burner exhaust products.
-------
522
COMBUSTION EQUIPMENT
= -8
(SQN033S) ~IQUnj 1TQBAVS oooooo oo.fc .06
-------
Gas and Oil Burners
523
VISCOSITY TEMPERATURE RELATION FOR FUELS
Description of the Chart
The horizontal scales at the top and bottom of the chart
are identical and represent temperature, both in degrees
Fahrenheit and Centigrade.
The vertical scales represent viscosity in terms of the
several methods of measurement now in common use. These
scales appear opposite the temperatures at which each cus-
tomarily is standardized for measuring liquid fuels; namely—
Viscosity Measurement Temperature
Sayholt Universal
Saybolt Furol
Engler Degrees
Redwood No. 1
Redwood No. 2
(Admiralty)
Kinematic
ut
at 100°F.
210°F.
at 77°F.
122°F.
20°C.
50°C.
irxrc.
at 70°F.
IOO°F.
140"F.
200°F.
at 77°F.
at both edges
Conversion of viscosity from one unit to another by means
of the chart is reasonably accurate for all practical purposes.
Caution should be used, however, when using the chart to
convert a given viscosity unit to kinematic viscosity at vari-
ous temperatures. The reason for this lies in the fact that
the conversion factors to kinematic vary slightly with tem-
perature and therefore, a single kinematic scale cannot be
precise at all temperatures. The approximate values ob-
tained from the chart, however, should be sufficiently accu-
rate for practical purposes, such as finding the proper
atomizing temperature or limits of pumpubiltty in this
viscosity unit. For a more precise conversion from Redwood,
Sayholt or Kngler units to kinematic viscosity, reference
should be made to the specific conversion factors usually
found in technical handbooks covering flow of fluids.
Viscosity, of course, decreases with increase in tempera-
ture. The diagonal lines, accordingly, represent the average
slopes in viscosity encountered with bunker and diesel fuels,
respectively. As the chart is prepared logarithmically, the
slopes appear as two groups of straight parallel lines. While
these particular slopes will not hold with all oils, they do
serve as a good index in the majority of cases, and should
therefore prove sufficiently accurate for most practical
purposes.
The dotted horizontal lines in the right section of the
chart indicate the ranges of viscosity recommended for best
atomizing fluidity by American burner manufacturers. In
Saybolt Universal measurement, the top line represents 200
seconds; the center line 150 seconds; the bottom line 100
seconds. To obtain proper atomization with most installa-
tions, viscosity at the burners should fall within the upper
range for forced draft, and within the lower range for
natural draft. This rule, however, has certain important
exceptions in the case of European burner practice as ex-
plained on the chart.
The dotted horizontal lines in the upper left section indi-
cate the maximum range of viscosity which will assure free
and efficient pumping. These lines represent, respectively,
400 seconds and 500 seconds Saybolt Furol. When equipped
for heavy-duty transfer, in which suction head is not a prob-
lem, it is possible to pump without difficulty at the upper
limit, or even above in some instances. However, it is pre-
ferable not to allow viscosity to exceed the lower limit.
To Find Viscosity at
Different Temperatures
Knowing the viscosity of an orl in one scale at one tem-
perature, to determine its viscosity in the same or a different
scale at a different temperature the procedure given in the
following example is used.
Let us assume a diesel fuel having a known viscosity Red-
wood No 1 (at IOO°F) of 44.7 seconds. This viscosity is
indicated by point "O".
Through "O" we draw the line E-E parallel to the nearest
diagonal. The line E-E intercepts Redwood No. 1 again at
"Q" (70°F.) showing 64 seconds, and at "S" (140°F.) show-
ing 35.2 seconds It also intercepts Engler at "P" (20°C.)
showing 226 degrees, and at "R" (5()°C.) showing 1.382
degrees. In this particular case, the only interception of
Saybolt Universal which would fall within the chart is
likewise at "O" ( IOO°F) showing 50 seconds.
In a like manner, viscosity conversions between different
scales and standardized temperatures can quickly be found
from any other known viscosity.
To Find the Proper
Atomizing Temperature
Let us assume that you have a bunker fuel oil having a
viscosity of 150 seconds Saybolt Furol (at 122°F.). Through
this point draw the parallel "A-A". The same line, of course,
would apply if we assumed the same oil, knowing its vis-
cosity only in terms of one of the other scales, such as 3500
seconds Saybolt Universal (at 100°F.).
The line "A-A" intercepts the upper and lower ranges
of atomizing viscosity between "Y-X" and "X-Z", respec-
tively Temperature ranges corresponding to these viscosity
ranges are then readily found by laying a straight-edge ver-
tically on the chart and noting the points "C", "B" and
"D" on the top and bottom temperature scales. In this case,
the temperature for 200 seconds Saybolt Universal viscosity
is 193°F; for 150 seconds it is 207.5°F. and for 100 seconds
232.8°F The atomizing range would be 193-207.5°F. for
forced draft or 207 5-232.8°F. for natural draft.
As a general rule, the lower the viscosity, the better the
atomization: hence, where difficulty is experienced in ob-
taining complete combustion, as evidenced by excessive
smoke, or by dry soot which sometimes is noticeable even
with a clear stack, it may prove advisable to operate in the
higher temperature (lower viscosity) range with forced as
well as natural draft. Severe cases may require raising the
oil temperature to correspond to the point of lowest prac-
ticable atomizing viscosity (the interception of the lowest
dotted line). However, engineers must be guided by pre-
vailing conditions. No fixed rule will apply in all instances.
To Find the Limit
of Pumping Temperature
Let us assume a Grade "C" bunker fuel oil having a
known viscosity Saybolt Furol (at 122°F.) of 150 seconds.
Through this point we draw the parallel "A-A". The tem-
peratures "F" and "G" corresponding to the points of inter-
ception "T" and "U" are found in the same manner as
described above.
Figure 363. Viscosity-temperature relation for fuel oils (Reprinted by permission of the
copyright owner, Esso Research and Engineering Co., Linden, N.J.).
-------
524
COMBUSTION EQUIPMENT
The problem is almost always traceable to tte
same origins, that is, the hurner and fuel are
not compatible, or the burner is not properly ad-
justed or operated.
Oxidizable emissions depend upon the degree to
which performance falls below the optimum, capa-
bilities of combustion equipment. Actual per-
formance is, however, a difficult thing to predict.
A survey (Chass and George, I960) of gas- or oil-
fired equipment in the Los Angeles area was made
on this subject. Some 27 representative equip-
ment items, ranging from a small water heater
to an 870-hp boiler, were tested for combustion
characteristics as well as air pollutants. No at-
tempts were made to adjust the equipment before
the tests; the data reflected, therefore, what can
be considered normal operation.
Thirteen of the equipment items tested were fired
alternately with both gas and oil, oil being the
standby fuel. Four items were fired only with
fuel oil and 10 only with natural gas. Natural gas
is the predominant fuel in the test area. Curtail -
able gas users do not normally burn standby fuel
more than 20 days in a given calendar year. Dur-
ing some -winter seasons, small users have not
been curtailed at all. Thus, oil burning was not
an everyday occurrence in most of the gas- or
oil-fired equipment tested.
The survey disclosed some points that would have
been anticipated and others that would not. Table
143 summarizes emission factors developed from
the data.
The most surprising indication -was that the fine-
ness of combustion control was much less when
natural gas rather than fuel oil was used for fir-
ing. This "was shown by the prevalence of car-
bon monoxide and the wide variation in fuel-air
ratios during gas firing. In contrast, only negli-
gible carbon monoxide was measured during fuel
oil burning, and fuel-air ratios were held to much
more constant figures. The same equipment found
to emit appreciable carbon monoxide on gas firing
discharged essentially no carbon monoxide (0. 003%)
when burning heavy fuel oil. In addition, combus-
tion efficiencies were better when high-viscosity
fuels (less than 17° API) rather than low-viscosity
fuels (greater than 28° AP[) -were burned. Appar-
ently then, surveillance by the operator is a func-
tion of the complexity of burning the particular
fuel. With a relatively easy-to-burn fuel such as
natural gas, attentiveness can be expected to be
minimal; -while with high-viscosity oils, burner
control will be most favorable. This phenomenon
is probably peculiar to areas where natural gas is
the predominant fuel and would be difficult to pre-
dict for other areas.
This situation may be due in part to the fact that
smoke serves as a better alarm on oil firing.
Smoke is likely to be emitted on oil firing -when
combustion is only moderately inefficient. An
operator would be expected to notice visible emis-
sions from the stack and make corrections at the
burner. During gas firing, smoke does not occur
unless combustion is markedly incomplete. A
gas-lired burner can emit appreciable carbon
monoxide without imparting perceptible opacity
to products of combustion. Thus, a gas burner
operaitor can well be ignorant of the fact that his
equipment is not functioning efficiently.
As would be expected, the survey showed that
emissions of particulate matter were appreciably
higher during oil burning. Oil burning produced
almost 10 times more particulates than natural
gas burning did. There was little measured dif-
ference in particulate emissions between distillate
and residual oil burning, even though the residual
oils contained appreciably more inorganic ash.
In addition, the data showed, surprisingly, that
distillate oils produced slightly greater quantities
of aldehydes than residual oils did, probably be-
cause of the poorer combustion efficiencies en-
countered with light oils. Natural gas produced
appreciably lesser aldehydes, even though com-
bustion efficiencies in general were lower, as
measured by the presence of carbon monoxide.
During gas firing, high carbon monoxide values
were generally accompanied by greater aldehyde
Table 143. EMISSIONS FROM GAS-FIRED AND OIL-FIRED EQUIPMENT (Chass and George, 1960)
Fuel
burnc-cl
Natural gas
Light oilb
Heavy oil'
Hums
tested
IT,
10
7
Carbon monoxide
Maximum, %
6.400
O.OZO
0. 003
No. > 0. 9%
3
0
0
No. > 0. 09%
5
0
0
lb/ equivalent barrel of fuel oila
Particulate matter
Range
0. 013 to 0. 353
0. 126 to 1. 720
0. 420 to 1. 220
Average
0. 077
0. 735
0. 750
Aldehydes as formaldehydes
Range
0. 017 to 0. 191
0. 042 to 1. 008
0. 042 to 0. 462
Average
0. 068
0. 185
0. 160
'if 6,000 it-' of natural gas is considered equivalent to 1 barrel of fuel oil.
bLight oils ranged from ZS. 7 to 45. 1 API gravity.
1 Heavy oils ranged from 8. 0 to 16.5 API gravity.
-------
Boilers, Heaters, and Steam Generators
525
concentrations. In no case, however, did alde-
hyde concentrations exceed 25 ppm (as formal-
dehyde) -when natural gas was burned.
Ash and Sulfur Oxides
Stack discharges of sulfur oxides and ash are
functions of fuel composition. During gas firing,
both contaminants are well below nominal air
pollution control standards. During oil firing, the
inorganic ash content of combustion gases is nor-
mally less than 0. 1 grain per scf, but sulfur oxide
concentrations can be appreciable. Regulations
limiting stack emissions of sulfur and the sulfur
content of fuels have been enacted in several
areas of the United States, as noted in the pre-
ceding part of this chapter. Regulations on par-
ticulate matter are aimed collectively at both in-
organic ash and combustible solids. When exces-
sive emissions are encountered during oil firing,
carbon and other oxidizable participates usually
predominate.
Oxides of Nitrogen
Combustion processes as a group represent the
major stationary source of oxides of nitrogen in
most communities. Concentrations In products
of combustion range from less than 10 to over
1, 000 ppm by volume, measured as nitrogen di-
oxide. Concentrations appear to be a. function of
temperature and firebox design. The smallest
concentrations are found at small appliances in
which there is appreciable excess air at the burn-
er. The largest concentrations are found in gas-
es from the largest combustion sources --steam
power plants, which are operated at high firebox
temperatures. _Combustion equipment of less
than 20 million Btu per hour gross input do"es not
normally emit NOX in concentrations greater than
100 ppm. This subject is covered more complete-
ly in the next part of this chapter.
AIR POLLUTION CONTROL EQUIPMENT
Wherever control equipment is considered for
combustion processes, it is almost always for
controlling nonoxidizable materials, notably ash
and sulfur oxides. If unburned and partially burned
hydrocarbons or carbon particulates arc the prin-
cipal contaminants, the normal procedure is to in-
crease combustion efficiency rather than collect
these materials at the stack. An efficient burner
is, therefore, the best and most inexpensive means
of controlling combustible air contaminants. Con-
trol equipment has been proposed only for large
combustion sources. To date, no satisfactory full-
scale control devices have been installed on the
principal type of objectionable equipment, steam
power plant boilers, as indicated in the next part
of this chapter.
BOILERS, HEATERS, AND STEAM GENERATORS
INTRODUCTION
Boilers, heaters, steam generators, and similar
combustion equipment fired with fossil fuels arc
used in commerce and industry to transfer heat
from combustion gases to water or other fluids.
The only significant emissions to the atmosphere
from this equipment in normal operation, regard-
less of the fluid being heated or vaporized, are
those resulting from the burning of fossil fuels.
Differences in design and operation of this equip-
ment can, however, affect production of air con-
taminants.
A boiler or heater consists essentially of a burn-
er, firebox, heat exchanger, and a means of cre-
ating and directing a flow of gases through the unit.
All combustion equipment--from the smallest
domestic water heater to the largest power plant
steam generator--includes these essentials. Most
also include some auxiliaries. The number and
complexity of auxiliaries tend to increase with
size. Larger combustion equipment often includes
flame safety devices, soot blowers, air prcb.ca.tors,
fuel heaters, and automatic flue gas analyzers.
Inasmuch as coal is not used as boiler fuel in Los
Angeles County, this discussion is limited to boil-
ers, heaters, and power plant steam generators
fired with gas or fuel oil.
Industrial Boilers and Water Heaters
The vast majority of combustion equipment is used
to heat or vaporize water, or both. For conve-
nience, industrial water heaters are considered
together with boilers inasmuch as identical equip-
ment is frequently used for both purposes. These
boilers and heaters fall into three genera! classi-
fications: Fire tube, water tube, and sectional.
Fire cube boilers constitiite the largest share of
small and medium-size industrial units, im md-
ing the Scotch marine and firebox types, as shown
in Figures 364 and 365. InJjLre_tub_o_boilers , the
Products of combustion pass through the heat ex~^~
changer tubes, whjle_j,yjitj^,_stea£ri^ £r_P-ili£X_QiLi£L_
is contained outside the tubes. Many boilers such
as these are sold as packaged units, with burners,
blowers, pumps, arid other auxiliaries all mounted
on the same framework.
Water tube boilers are constructed in a wide range
of sizes. Both the smallest and largest industrial
units are likely to be of water tube design and,
in fact, all large boilers (steam generators) are
of this type. The smallest units are of simple
box construction, commonly using tubing to cir-
culate water and steam. In the water tube design
-------
5Z6
COMBUSTION EQUIPMENT
Figure 364. A three-pass, Scotch-marine boiler (Ray Burner Co., Boiler Division,
San Franci sco, Cali f.).
shown in Figure 366, fluid is heated under pres-
sure in a coil heat exchanger and flashed into
steam in an external chamber. These relatively
small, controlled-circulation boilers are capable
of producing steam within minutes after a cold
start. Industrial water tube boilers, such as that
shown in Figure 367, are usually constructed -with
comparatively larger fireboxes than fire tube
boilers have. In all water tube boilers, the water,
steam, or heat transfer medium is circulated
through the tubes while hot products of combus-
tion pass outside the tubes.
Sectional boilers employ irregularly shaped heat
exchangers and cannot be classed as either water
tube or fire tube. Hot combustion gases are di-
rected through some of these passages, transfer-
ring heat through metal walls to water or steam
in the other passages. These units are manu-
factured in identical sections, such as those shown
in Figure 368, which can be joined together accord-
ing to the needs of the operator. A sectional boil-
er consists of one or more sections and can be en-
larged or reduced by adding or removing sections.
The heat exchanger assemblies are usually fabri-
cated of cast iron. For this reason these boilers
are not suitable for pressures greatly exceeding
15 psig. Cast iron sectional boilers find frequent
use as water heaters and steam generators used
in conjunction with space heating and laundries.
Power Plant Steam Generators
The largest boilers are located at steam power
plants where high-pressure, superheated steam
is used to drive turbo-electric generators. These
water tube units are commonly termed steam
generators. Nevertheless, there is no definite
size limitation for equipment such as this, and
steam generator designs do not differ markedly
from those of many smaller industrial boilers.
-------
Boilers, Heaters, and Steam Generators
527
Figure 365. A fire-tube boiler
Works, Erie, Pa.).
with a refractory-1 ined firebox (Erie,City Iron
Power plant steam generators produce from
50, 000 to 5 million pounds of steam per hour at
up to 2, 500 psig and 1, 000°F.
A typical large-city power plant steam generator
consumes 2, 500, 000 cubic feet of natural gas per
hour or 450 barrels of Number 6 fuel oil per hour,
exhausts some 700, 000 scfm combustion products
and furnishes all the steam required to drive a
310, 000-kilowatt electric generator. The trend
toward large steam generators is illustrated in
Figure 369.
A conventional front-fired power plant steam gen-
erator is shown in Figure 370. It is equipped with
the full line of boiler auxiliaries: Air preheater,
oil heater, economizer, superheater, and so forth.
As much heat as is practical is extracted from com-
bustion products. Stack temperatures are normal-
ly maintained at 225° to 320°F. Condensation
and resultant corrosion are the principal deterrents
to lower power plant temperatures. When exhaust
gas temperatures approach the dew point, conden-
sation and visible stack plumes are encountered.
Steam generators operate with thermal efficiencies
of about 90 percent, and operating variables are
more carefully controlled than in any other type of
combustion equipment. Of prime concern is the
excess air rate. Any air above the theoretical re-
quirement represents a thermal loss, but the fire-
box oxygen concentration must nevertheless be
sufficiently high to provide near perfect combus-
-------
528
COMBUSTION EQUIPMENT
Figure 366. A forced-circulation boiler with a coil water tube heat exchanger and an
external flash chamber-accumulator (The Clayton Manufacturing Co., El Monte, Calif.).
tion. Power plant operators hold excess air rates
during fuel oil firing as low as feasible by provid-
ing strong mixing conditions and optimum fuel oil
atomization at the burner. During gas firing, ex-
cess air rates are about 10 percent above the the-
oretical requirement. When fuel oil is burned,
excess air is usually held below 15 percent. At-
tempts have been made to operate with excess air
rates as low as 1 percent (about 0. 2 percent oxy-
gen) on oil firing (Glaubitz, 1963). The benefits
from this practice are reduced corrosion, less
air contaminants, anil increased thermal efficien-
cies.
Refinery Heaters
Refinery oil heaters are noteworthy inasmuch as
they usually comprise large combustion units and
are likely to be fired with a wide variety of re-
finery by-product fuels, both gaseous and liquid.
These fuels can be the least saleable refinery
products, notably heavy residual oils and high-
sulfur-bearing gas streams. The gaseous fuels
are usually mixtures that for one reason or another,
are not marketed or further processed. Typical
analyses of refinery make gases are included in
Table 139 on page 508. Note that they can contain
appreciable amounts of sulfur, hydrogen, carbon
monoxide, and higher molecular weight hydrocar-
bons. The latter are responsible for the relative-
ly high heating values of refinery make gases.
Petroleum process heaters are apt to be fired with
the highest viscosity oil fuels produced at a re-
finery. Residual fuel oils have traditionally been
difficult to market; conseatently, operators pre-
fer to bxirn as much of these as possible in their
own equipment.
In most refinery heaters, such as those shown in
Figures 371 and 372, an oil or other petroleum
product flows inside the heat exchange tubes. Fire
tube oil heaters find only occasional use. These
heaters, like all other refinery equipment, are
normally operated 24 hours a day, 7 days a week.
They are not likely to be shut down, except dur-
ing periods of inspection and repair. Hot fuel
oils are almost always available, and there is
little likelihood of having to start a cold heater
with unheated, high-viscosity fuel oil.
Hot Oil Heaters and Boilers
In a number of industrial combustion equipment
units, a stable heat transfer oil is heated or
-------
Boilers, Heaters, and Steam Generators
529
Figure 367. An industrial water tube boiler (The Babcock and Wilcox Co., New York).
vaporized. Some of these units are simple water
tube boilers in which the heat transfer oil mere-
ly replaces water. Others are custom designed
for the particular oil and application. These
boilers, most often found in the chemical pro-
cess industries, are used to transfer heat to
another fluid in a heat exchanger device. Their
principal advantage is the lower vapor pressures
(higher boiling points) of the stable organic oils
as compared with that of water. Most have boil-
ing points between 300° and 800°F. In this
range, the compounds, whether gases or liquids,
exhibit markedly less vapor pressure than steam-
does at the same temperature.
The most common heat transfer medium is Dow-
therm A, * a mixture of diphenyl and diphenyl ox-
ide, with a boiling point of 495°F at 14. 7 psia
(Dow Chemical Co. , 1963). A number of other
*;Tegi stered Trademark of the Dow Chemical Company.
oils are also marketed. Almost any liquid that
is stable at the required elevated temperatures
and has a stiitable vapor pressure curve would be
satisfactory for this use. Most of these materials
are not highly toxic. Moreover, they are not
emitted to the atmosphere in quantities sufficient
to cause an odor nuisance or health hazard ex-
cept for instances of equipment failure or gross
disrepair. Some oils have sharp, penetrating
odors that can be detected in the boiler room.
These odors can be an annoyance to plant person-
nel.
Fireboxes
Stack emissions from heaters and boilers are in-
herently tied to the fuels and burners, as noted in
the preceding parts of this chapter. Of prime air
pollution concern in the combustion equipment is
the firebox in which the burners are located. Most
fireboxes are constructed of such a shape and size
that the burner flames are contained within the
-------
530
COMBUSTION EQUIPMENT
Figure 368. A cast iron sectional boiler
(Crane Co., Johnstown, Pa.).
firebox and do not impinge upon the firebox walls
or the heat exchange equipment. Flame impinge-
ment on either heat transfer surfaces or firebox
walls usually results in incomplete combustion
and a marked increase in air contaminant emis-
The volume of the firebox is governed by the type
of flame and the heat release rate. Where flame
are luminous and relatively long, allowable heat
release rates are low as compared with those of
short, non-luminous flames. Clean, gaseous fue
can be burned at rates ranging up to 1 million Bt\
per hour per cubic foot of firebox volume. The
latter rate is possible only with strong mixing coi
ditions and necessarily high pressure drops acroi
the burner. In practice, natural gas heat release
rates of 100, 000 Btu per hour per cubic foot and
lower are more common. When oil is burned,
even on a stand-by basis, heat release rates are
always below the latter figure. The upper limit
for burning low-viscosity fuel oils is about
100, 000 Btu per hour per cubic foot of firebox
volume. Heavy residual oils require greater
combustion space. Design rates for residual-oil-
fired combustion fireboxes range from 20, 000 to
40, 000 Btu per hour per cubic foot (The North
American Manufacturing Company, 1952). Oil-
burning heat release rates often govern firebox
design, even though gaseous fuels may be burned
in the equipment most of the time.
Most fireboxes of small and intermediate-size
boilers and heaters are constructed of firebrick
or refractory cement. Some are of metal con-
struction, usually where firebox temperatures
are relatively low. In large installations, the
firebox, which is often termed here a furnace,
is lined with water tubes or water walls through
which cooling water is circulated. In these de-
signs, water must be circulated at a. sufficient
uu-
aU
UU1
4UH
0
T
\
46
-\
V
n
\
p
49 1 50
.
N
:
N
\
P,
51 1 52
i
*%*
j
j
SIZE
0 UNDER MILLION Ib/hr STEAM
[] BETWEEN 1 AND 2 MILLION Ib/hr STEAM
| OVER 2 MILLION Ib/hr STEAM
mLINES REPRESENT TRENDS ONLY
—
"4,
^
^
p
*:
*•
»
=S=5
i
^
*
%
^\
n
=»S5
I
*=*f
/
^
^
^
Ir
\
6!
'^
I
53 1 54 1 55 56 57 58 ' 59
YEAR
-^
%
\.
<
*
p
"I
^*f
60 61 62
Figure 369. Percent of boilers of each size group purchased in the
year ending April 30 (Frankenburg, 1963).
-------
Boilers, Heaters, and Steam Generators
531
INDUCED D«»FT
FAN
Figure 370. A front-fired power plant steam generator (The Babcock
Wilcox Co., New York).
and
rate to prevent heat damage to the metal walls
or tubes. Much of the heat transfer in the fur-
nace is by radiation rather than convection. Al-
most all large boilers, for example, steam gen-
erators at power plants, are constructed with
water-tube-lined fireboxes.
Soot Blowing
Whenever fuels of measurable ash content are
burned, some solids, including both carbon and
inorganic ash, adhere to heat transfer surfaces in
the combustion equipment. These deposits must
be periodically removed to maintain adequate heat
transfer rates. It is common practice to remove
these deposits with jets of air or steam while the
combustion equipment is in operation. The re-
moved soot particulates are entrained in combus-
tion gases. During periods of soot blowing, par-
ticulate concentrations are, as would be expected,
considerably greater than during normal opera-
tion. Instantaneous particulate concentrations
vary greatly during soot blowing because of the
inherent operating characteristics of the lances.
A typical long retractable soot blower is shown
in Figure 373. During operation the lance con-
taining the air or steam jets rotates and moves
horizontally across the tube surface. On the in-
stroke, most of the particulates are removed.
Consequently, stack emissions are heavier on the
instroke than on the outstroke for a given lance.
Normally, there are from 8 to 15 of these blow-
ers on a large power plant's water tube boiler.
The blowers are usually operated in sequence, by
starting at the front or upstream tube surfaces and
working downstream, finally cleaning the air pre-
heater.
-------
532
COMBUSTION EQUIPMENT
o o o o <
o o o o o I
O O O O D OJ
00000
o o o o o o
o o o o o ,
o o o o o o]
O O O O O i_
n
r
r
n
Figure 371. A large box-type refinery heater.
Whenever residual fuel oils or solid fuels are
burned in large steam generators, tube clean-
ing is usually conducted at least once during
every 24 hours of operation. When clean nat-
ural gas fuels are burned, the same boiler or
heater can be operated indefinitely without soot
blowing, except possibly for the air preheater.
In fact, the burning of natural gas gradually re-
moves materials deposited during oil firing. At
many highly integrated power plant boilers, soot
blowers are operated automatically, at 2 - to 4-
hour intervals. At many older installations,
soot-blowing equipment is likely to be manual,
the operation is time-consuming, and intervals
between blowings comparatively longer. In the
latter cases, particulates are somewhat larger
and emissions during any one blowing are like-
ly to be heavier in comparison with automated
lancing operations. Where soot blowers are
manually operated, the tubes are not usually
cleaned more than once per 24 hours of oil fir-
ing-
air contaminants in combustion equipment
are formed in the firebox-.and are definitely influ-
enced by firebox and burner design. Heat exchange
surfaces are generally considered to have some
catalytic effect on sulfur oxides and possibly ox-
ides of nitrogen, but there are few data or little
agreement on the extent of this influence. Tube
surfaces, -without question, affect pollutants in
that they collect enough particulates to require
lancing during periods of oil firing.
For the purpose of this discussion, the assump-
tion is that the material being heated or vaporized
in the heat exchanger tubes is not a significant
source of air pollution. This is, in fact, the situa-
tion with common combustion equipment. A major
odor nuisance or health hazard can, however, be
caused by venting organic vapors to the atmosphere,
though this -would be expected to occur only during
equipment breakdown. That an operator would
purposely jettison significant quantities of his prod-
uct or of expensive heat transfer fluids is not likely.
THE AIR POLLUTION PROBLEM
Air contaminants, emitted from combustion equip-
ment are described in the preceding parts of this
chapter, covering fuels and burners. Nevertheless,
the size and design of combustion equipment greatly
affect the quality and quantity of stack emissions.
Combustion equipment emits both visible and non-
visible air contaminants. Visible contaminants are
principally liquid and solid particulates. Nonvisible
contaminants include nitrogen oxides, carbon mon-
oxide, and sulfur dioxide. A material that strad-
dles both categories is sulfur trioxide, the extreme-
ly hygroscopic anhydride of sulfuric acid.
-------
Boilers, Heaters, and Steam Generators
533
TO STUCK
OUT
o o o o
o o o o o o
o o o o o
o o o o o o
o o o o
Figure 372. A vertical, cylindrical refinery heater (Union Oil Co.,
Los Angeles, Cal if.).
AIR PREHEAT
Figure 373. A long-travel retracting soot blower with an air motor drive
(Diamond Power Specialty Corp., Lancaster, Ohio).
-------
534
COMBUSTION EQUIPMENT
Solid Porticulate Emission During Normal Oil Firing
Where combustion is most nearly complete, in-
organic ash constitutes the principal particulate
emission. The inorganic ash contents of most
fuel oils and all gaseous fuels are normally well
below the concentrations that would cause exces-
sive particulate emissions. As noted in the first
part of this chapter, an inorganic ash ccntent of
0. 1 percent in typical residual fuel oil results in a
stack concentration of only 0. 03 grain per scf at
12 percent carbon dioxide. Particulates from re-
sidual-oil firing are considerably larger than those
emitted during gas firing, as can be seen in Fig-
ures 374 and 375. Nevertheless, participates
from oil burning are still principally in the sub-
micron range and are in sufficiently large con-
centration to cause perceptible light scattering.
Finely divided ash is considered a contributor to
visible stack plumes at a large power plant's
steam generators. Most of this material is in
the form of metal oxides, sulfates, and chlorides.
A spectrographic analysis of a typical residual
oil ash is presented in Table 142.
** S\ K? ^ .&J*»<«* T?t f >"..,., fa
-------
Boilers, Heaters, and Steam Generators
535
Table 144. PARTICLE SIZE DISTRIBUTION OF
TYPICAL MATERIAL COLLECTED FROM
A STEAM GENERATOR STACK DURING
THE BURNING OF RESIDUAL FUEL OIL
Larut-sl
Control District Rules and Regulations). This in-
cludes combustible particulates as well as inorgan-
ic ash. This limit may be exceeded when common
hydrocarbon gases or fuel oils are burned if ap-
preciable amounts of carbon or carbonized high
molecular weight hydrocarbon materials,or both
are present. The latter situation results from
either poor operation or incorrect selection of
burner and fuels. In these instances, the result-
ing visible contaminants at the stack are black,
and are apt to exceed allowable limits, most of
which are based on the Ringelrnann Chart.
The shapes of carbon or combustible particulates
vary somewhat with fuels and operating conditions.
If a light fuel oil or gaseous fuel is burned in a de-
ficiency of oxygen, the resulting carbon particles
are likely to be exceedingly fine. If, on the other
hand, these contaminants are the result of burning
heavy fuel oil with improper atomization, the car-
bon particles emitted are likely to be in the form
of cenospheres, as depicted in Figure 376. Ceno-
Figure 376. Photograph with light microscope
of cenospheres found in breeching of large
oil-fired steam generator (MacPhee, 1957).
spheres are spherical, hollow particles, essen-
tially the same as those produced during spray
drying. Cenospheres have appreciably smaller
bulk densities than solid particulates do (MacPhee
et al. , 1957).
An operator is not likely to \varit to discharge
carbon in either form, or to have to control these
particulates at the stack. When there is evidence
of appreciably unburned particulates in combus-
tion gases, steps should be taken to improve com-
bustion at the burner. With high-viscosity oils,
these steps can consist of using lower viscosi-
ties or increasing pressure drops across the burn-
er to provide proper atomization.
Soot-Blowing Particulates
At times •when soot blowers are in operation, par-
ticulate matter concentrations in exit gases in-
crease markedly. Instantaneous concentrations
depend upon the dirtiness of tube surface and up-
on the rate at which the lance moves across the
tubes. Soot-blown air contaminants can impart
excessive opacities to stack gases and cause
damage by acidified particulate deposition in im-
mediately adjoining areas. The air pollution po-
tential, in terms both of opacity and nuisance,
increases with the time interval between soot-
blowing operations. Where tubes are blown at
Z- to 4-hour intervals, as is done on many mod-
ern combustion devices, there is little increase
in the opacity of stack emissions, and the small
sizes of particulates as well as the relatively
small concentration reduce the possibility of fall-
out damage. Intervals of 8 hours and longer be-
tween tube lancings can result in excessive visi-
ble opacities as well as fallout damage.
Soot-blowing air contaminants are not considered
to be highly significant in the overall air pollution
of a. given area, inasmuch as they are. emitted only
for relatively short intervals and tend to settle
close to the source. They represent less than 10
percent of the total particulates emitted from an
oil-fired boiler. Many operators avoid technical
opacity violations by special scheduling of soot-
blowing operations. This involves either more
frequent lancing or the lengthening of the total
operation. Neither course reduces overall air
pollution, but both can allow technical compliance
with air pollution control regulations involving
permitted opacities.
Sulfur Dioxide
As pointed out in the first part of this chapter,
practically all fuel-contained sulfur--upwards of
95 percent — shows up in exhaust gases as sulfur
dioxide, a colorless gas. There is no way of pre-
venting the formation of sulfur dioxide, and con-
-------
536
COMBUSTION EQUIPMENT
centrations are functions of the fuel's sulfur con-
tent. As undesirable as sulfur dioxide is, it is,
nevertheless, generally considered less obnoxious
than sulfur trioxide and the odorous sulfidcs and
mercaptans contained in the fuel.
Sulfur Trioxide
Up to 5 percent of the total fuel's sulfur is con-
verted to the higher oxide, sulfur trioxide, in
large combustion equipment. The volume of sul-
fur trioxide found in gases from power plant steam
generators (5 to 50 ppm) is considered to be a
principal cause of the visible plume often present
during oi] firing. It readily combines with water
to form sulfuric acid and, as such, can cause
acid damage in downwind areas.
The oxidation of sulfur is considered to proceed
in two steps as follows:
S
S0_
1/2 O
As shown in Figure 377, equilibrium at ambient
temperatures strongly favors sulfur trioxide
rather than the dioxide. At elevated tempera-
tures, the dioxide predominates. The reaction
rate falls off rapidly, however, below 700°F; as
a result, the major portion of the fuel's sulfur is
still in the dioxide form when discharged from
combustion equipment.
As might be expected, the degree of sulfur tri-
oxide formation in combustion equipment varies
widely. Concentrations are negligible in small
equipment, even when fired with high-sulfur fuel
oils. As equipment sizes and firebox tempera-
tures increase, sulfur trioxide concentrations in-
crease appreciably though seldom exceeding 35
ppm. Heaviest emissions are found at the larg-
est combustion sources--power plant steam
generators.
Formation of sulfur trioxide appears to depend
upon several factors. Concentrations tend to
increase with increases in firebox temperatures
and oxygen concentrations. In addition, oxida-
3,500
3,000
2,500
2,000
!,500
!,000
o to
0 20
0 30
0.40 0 50
VOLUME RATIO, S03/S03
0 SO
SO 2
0 70
0 BO
0.90
1 00
Figure 377. Equilibrium concentrations of 803-802 at various oxygen concentrations
as per the reaction S09(g) + 1/2 {big) = Slh(g). (Adapted from Hougen and Watson,
1945).
-------
Boilers, Heaters, and Steam Generators
537
tion catalysts such as vanadium, iron, and nickel
oxides tend to increase SO-^ production. Par-
ticulates that adhere to tube surfaces usually
contain appreciable quantities of all three of these
catalytic materials.
Crumley and Fletcher (1956) ran a series of experi-
ments on a small kerosene fuel furnace from which
they concluded that, for a given total sulfur oxide
+ SO) concentration:
1. SO, formation increases as flame tempera-
tures are increased up to about 3, 150°F;
2. above 3, 150°F, SO3 formation does not in-
crease, that is, the SO,/SO2 rate remains
constant;
3. when flame temperatures are held constant,
SO? formation decreases as the excess air
rate is reduced:
4. SO^ formation decreases with coarser atom-
ization. This phenomenon may be a result of
lower flame temperature.
The work of Glaubitz (1963) generally agrees with
these conclusions regarding small oxygen concen-
trations at the burner. It is discussed later in this
chapter.
Sulfur trioxide is considered the principal cause
of the visible plumes emitted from power plant
steam plant generators. It apparently unites with
moisture in the air and with flue gases to form a
finely divided sulfuric acid aerosol. Droplet con-
densation may be enhanced by the presence of par-
ticulate matter, which provides condensation nuclei.
These visible emissions are interrelated with so-
called dew/ point raising. The presence of sulfur
trioxide and sulfuric acid effectively results in a
gaseous mixture that appears to have a dew point
higher than would be predicted solely on the basis
of the moisture content. These dew point eleva-
tions can exceed 200°F. Figure 378 shows typi-
cal dew points and sulfur trioxide concentrations
measured at an experimental oil-fired furnace
(Rendle and Wilsdon, 1956). Note that sulfur tri-
oxide, in concentrations ranging from 5 to 25 ppm
by volume, increases the dew point (about 115°F
based upon IH^O alone) by increments of 20°F and
170°F respectively. There are noticeable dif-
ferences in published values of dew point eleva-
tion. These are attributable in part at least to
the difficulties encountered in SO^ analysis.
Sulfur trioxide has a further disadvantage in that
it tends to acidify particulate matter discharged
from combustion equipment. This is commonly
evidenced by acid spotting of painted and metallic
surfaces, as well as of vegetation in the down-
wind area Acid damage is usually the result of
discharge of particulates during soot blowing.
150 200 250
ElEVHTED DEI POINT °r
Figure 378. Dew point elevation as a function
of sulfur trioxide concentration (Adapted
from Rendle and Wilsdon, 1956).
Excessive Visible Emissions
Combustion equipment has traditionally been as-
sociated with visible smoke plumes caused by un -
burned carbon and organics. With modern steam
generators, markedly incomplete combustion is
a relative rarity. Combustible air contaminants
are seldom present in sufficient concentrations to
obscure visibility. Nevertheless, visible plumes
of greater than 40 percent opacity are common at
large oil-fired steam generators where there are
only minimal quantities of unburned materials in
exhaust gases. These opaque emissions are com-
monly attributed to finely divided inorganic mate-
rials, notably sulfur trioxide and inorganic partic-
ulates.
The formation of visible plumes in stack gases
that are practically devoid of unburned carbo-
naceous materials is not fully understood. The
phenomenon is known to occur only when there
is appreciable sulfur in the fuel and when the
steam generator is of relatively large capacity
greater than about 60, 000 pounds of steam per
hour. The plumes do not occur during the burn-
ing of "clean" natural gas, that is, gas with only
sufficient sulfur to impart a detectable odor--
about 0. 15 grain per 100 cubic feet of gas. With
fuel oils of 0. 3 to 0. 5 percent sulfur, some vis-
ible emissions can occur at the stack, but opaci-
ties do not usually exceed 40 percent. At small-
er power plants, that is, those with a capacity
of 50, 000 to 500, 000 pounds of steam per hour,
-------
538
COMBUSTION EQUIPMENT
the opacity of exhaust gases does not normally
exceed 30 percent when the plant is Tired with
residual oil of average sulfur content, namely
1. 4 to 2.0 percent. Steam generators of
750, 000 pounds per hour and greater ratings
can be expected to emit gases of heavier than 40
percent ooacity when fired with oil of more than
about 1. 0 percent sulfur.
Figure 379 illustrates the difference in visible
emissions from two identical sido-by-side steam
generators on oil and gas firing. Both are rated
at 1, 200, 000 pounds of steam per hour and are
of conventional front-fired design. Stack tem-
peratures were approximately 300 °F. The unit
on the left was being fired with natural gas, and
there was no detectable opacity in the exhaust
gases. The identical unit on the right was being
fired with fuel oil of approximately 1. 6 percent
sulfur and was discharging gases of approximate-
ly 80 percent opacity. The visible plume irom an
oil-fired unit such as this normally varies from
white to brown, depending upon weather conditions
and the makeup of particulate matter. In some
cases, the visible plume appears to be detached
from the stack. The gas stream immediately
above the stack outlet is clear or at least of low
opacity but becomes opaque further downstream.
Apparently, cooling in the immediate stack dis-
charge area lowers temperatures below the dew
point, causing formation of extremely fine sulfur
trioxide and acid droplets.
There is some difference of opinion as to the cause
of this plume, but all evidence points to sulfur tri-
oxide as the principal determinant, with partic-
ulate matter as a possible contributor. Observa
tions have been made with residual fuel oils of
varying sulfur content. In general, the fuels coi
taining greater percentages of sulfur were found
to produce heavier opacities. , Since low-sulfur
oils also have lower ash contents, there is less
particulate matter in stack gases during the burr
ing of low-sulfur fuels as evidenced when opaci-
ties are lowest.
Stack opacities could conceivably be reduced if
the inorganic particulate sizes were increased
above the submicron level. With an arrange-
ment such as this, there would be an appreciably
smaller number of submicron particles in a given
volume of stack gases. No satisfactory method o,
increasing particle sizes has, however, been de-
veloped for residual oil. The most obvious meth-
of--coarser atomization--would result in increase
combustible particulates that would possibly be
more undesirable than the visible plume.
Some trials have been made by injecting sulfur
trioxide into the relatively clean stack gases
from natural gas firing. These experiments in-
dicate that definite visible opacity can be imparted
at concentrati_ons__of 5_pp_m SO^Jay^vplume and
greater. Stack gases with 5 ppm SO3 liaxTari~opac-
ity of approximately 20 percent. An opacity of ap-
proximately 50 percent resulted when the SO^ con-
centration was increased to 15 ppm. The test unit
was a 1, 200, 000-pound-per-hour steam generator
from which there was no visible plume during nor-
mal gas firing. The stack gases during SO3 addi-
tion appeared white when viewed with the sun at the
rear of the observer and were not unlike plumes
discharged from the same unit during oil firing.
Figure 379. Exhaust gases from identical steam
generators showing visible plume from oil-fired
unit (right) as compared with clear stack of
gas-fired unit (left).
There is evidence that "dirtiness, " that is, ac-
cumulation of deposits on tube surfaces, also con-
tributes to opacity. Identical side-by-side steam
generators have been observed to emit gases of
markedly different opacity "when fired at the same
rate with the same fuel oil. Invariably, the unit
with the thicker tube deposits is found to emit
heavier visible emissions and to contain consider-
ably larger 803 concentrations. This phenomenon
indicates that tube deposits are effective in catalyt-
ically oxidizing SO^ to SOj. The deposits that
contribute to the "dirtiness" apparently are not
sufficiently removed by normal soot-blowing proce-
dures. To lower opacities and SO^ emissions ef-
fectively, one should wash the tube surfaces with
an aqueous solution. The cleaning of heat exchang-
er surfaces in this manner requires that the unit
be shut down and allowed to cool beforehand. At
most electric power stations, steam generators
cannot be taken out of service often. Consequent-
ly, tube washing is a relatively infrequent occur-
rence.
-------
Boilers, Heaters, and Steam Generators
539
Oxides of Nitrogen
Combustion equipment collectively represents the
largest nonvehicular source of oxides of nitrogen
air contaminants in most industrial areas. In Los
Angeles County, where there is a high motor ve-
hicle density, boilers and heaters are still respon-
sible for more than 30 percent of the total oxides
of nitrogen discharged to the atmosphere and for
more than 90 percent of the total from all station-
ary sources (Los Angeles County Air Pollution
Control District, 1963).
Exhaust gas concentrations of oxides of nitrogen
range from less than 10 ppm by volume for small
gas-fired water heaters to over 1, 000 ppm for
large power plant steam generators. Since both
concentrations and gas volumes increase with
size, power plant steam generators are always
large sources of NO . These generators are
much more significant in the overall air pollution
picture than the markedly larger number of small-
er domestic and industrial heaters and boilers are.
One of these generators can emit over i, 000 pounds
of oxides of nitrogen per hour. As shown in Table
145, there is a wide variation in NOX emissions,
even from equipment of the same general type and
size.
Emissions of oxides of nitrogen from combustion
equipment result from fixation of atmospheric
nitrogen in the fireboxes. The principal high-
temperature reaction is the formation of nitric
oxide as follows:
C>
NO +
Z
NO
Z'
The latter reaction reaches a maximum at about
600°F and is extremely slow at ambient temper-
ature. Nitrogen dioxide is considerably more
reactive than nitric oxide, and is a more obnox-
ious air contaminant.
A number of other oxides of nitrogen are also
formed to lesser degrees. These include N7O,
N-,0^, N^Oj., and NOj and are not considered to
be emitted in significant amounts. For purposes
of this discussion, all oxides of nitrogen are con-
sidered collectively under the term NOX. In the
quantitative analysis of oxides of nitrogen, all
oxides are commonly oxidized to the dioxide. Re-
sults are reported in concentrations of NOX as NO
At ordinary temperatures, fixation does not pro-
ceed to any measurable extent, but the reaction
rate and equilibrium concentrations increase
markedly with temperature. In Figure 330, equi-
librium concentrations of nitric oxide are plotted
over ranges of temperature and oxygen concentra-
tions found in combustion equipment. Obviously,
100 ppm NO corresponds to equilibrium at about
1, 800 "F for the oxygen contents foxmd in high-
efficiency combustion gases, namely, 2 to 3 per-
cent oxygen. At 3, 000 °F, equilibrium concentra-
tions of NO are well over 1, 000 ppm, even at 1
percent oxygen. Calculated flame temperatures
are in excess of 4, 000°F at 10 percent excess air
(about Z percent oxygen) for both oil and gas fir-
ing when air preheated at 600°F is used.
Table 145. EMISSIONS OF OXIDES OF NITROGEN FROM INDUSTRIAL BOILERS AND HEATERS
(Mills et al. , 1961)
Source
Small oil heaters
Natural gas
Fuel oil
Largo refinery heaters
Natural gas
Fuel oil
Small boilers (less than 500 hp)
Natural gas
Fiu-1 oil
Large boilers (5QO hp and larger)
Natural gas
Fuel oil
Power plant steam generators
Natural gas-
Fuel oil
Heat input range,
millions of Dtu/hr
Less than 60
90 to 200
Less than 20
20 to 90
ZOO to Z. 000
Range of NOK com
in flue gases,
ppm by vol
ZO to 100
25 to 137
5 to n.i
15 lo 387
45 to 149
Z14 to Z8Z
75 to 320
275 to 600
Avg NOX i one ,
ppm by vol
47
59
5 i
1Z2
91
258
205
420
NOX emission iaetors,
avg Ib per
million Btu
0. 06
0. 33
0. 25
0. 52
0. 14
0. 49
0. 28
0. 62
0. 56
0 78
-------
540
COMBUSTION EQUIPMENT
I (00 2 600
TEMPERATURE
Figure 380. Equilibrium concentrations of nitric
oxide in the range 1,800° to 3,200°F as qer the
reaction: N2 + 02 ^—^ 2ND at 78% NT (calculated
from data in Hougen and Watson, 1945).
Emission concentrations of oxides of nitrogen
from combustion equipment are apparently gov-
erned by the formation rate of nitric oxide.
Measured NO^ concentrations from large boilers,
as shown in Table 145 and Figure 381, are well
below? equilibrium concentrations at maximum
firebox temperatures. For oil-fired steam gen-
erators, the exit NO concentrations of Table 145
(175 to 600 ppm) correspond to equilibrium at
2,000° to Z,450°F, respectively, and 3 percent
oxygen. Theoretical flame temperatures are well
in excess of 3, 600°F where the equilibrium con-
centration is 4, 150 ppm at 3 percent oxygen.
Measuring firebox temperatures or residence
times in the hottest zone is not feasible. Never-
theless, the gases can be assumed to be subject
to temperatures in excess of 3, 000°F in large
fireboxes, at least momentarily. On the basis
of measured exit concentrations, the residence
time at maximum temperature appears to be ex-
tremely short.
The rate of formation of NO increases markedly
above 3,000°F', as shown in Table 146. Times
of formation of 500 ppm nitric oxide were calcu-
lated at 3 percent oxygen and 75 percent nitroger
These calculated values may be somewhat low in
that all the nitrogen fixation is assumed to occur
the exit oxygen concentration. Some fixation woi
probably take place at larger oxygen concentra-
tions before combustion is completed. The time)
of formation illustrate the rapid change in rate
between Z, 800° and3,600°F. At 2, 800 "F, the
time for formation of 500 ppm NO is 16. 2 second:
at 3, 200°F, the time is 1.10 seconds; a-nd at 3, 60
0. 117 second.
At any given temperature, the decomposition
(2NO •
N,
rate constant is much greater
than the formation rate constant. This fact may
lend hope that NO could be decomposed back to
the elements before other more stable oxides are
foz-med. This latter possibility appears blocked
by the marked slowness, if not stagnation, of the
decomposition rate in the necessary temperature
range--2, 000°F and lower. Decomposition be-
comes negligible below about 3, 200 "F, according
to Ermenc (1956), and no known data indicate that
measurable decomposition occurs below 2, 800°F.
For the residence times possible in boiler fire-
boxes, any decomposition taking place below
3, 000 °F would appear to be insignificant. In any
case, measured NO emissions are well below
the 2,800°F equilibrium concentration of 1,380
ppm. Decomposition cannot occur when NO con-
centrations are less than equilibrium concentra-
tions.
Since NO formation rate constants are extremely
high in the range of 3, 500° to 4, 000°F, a frac-
tion of a second's residence time more or less
can make a significant difference in NO and NOX
concentrations. Measurements of oxides of nitro-
gen at large steam generators, in fact, bear this
out. NO concentrations at these sources are ex-
tremely variable, indicating that there are small,
almost imperceptible changes in operating condi-
tions that greatly increase or decrease NOx emis-
Table 146. EQUILIBRIUM CONCENTRATIONS
AND TIMES OF FORMATION OF NITRIC
OXIDE AT ELEVATED TEMPERATURES AT
75 PERCENT NITROGEN AND
3 PERCENT OXYGEN
Temperature ,
"F
2, 000
2, 400
2, 800
3, ZOO
3, 600
Equilibrium
concentration
of nitric oxide,
ppma
180
550
1, 380
2, 600
4, 150
Time of formation
of 500 ppm NO,
seconds'3
1, 370
16. 200
1. 100
0. 117
aHougen and Watson, 1945.
bDaniels and Gilbert, 1948.
-------
Boilers, Heaters, and Steam Generators
541
1,400
1,200
1,000
| | • GAS FIRING
' FUEL FIRING
400
200
EL SEGUNOO
Figure 381. Oxides of nitrogen concentrations in gases from various gas-fired,
oil-fired, and coal-fired steam generators (Barnhart and Oiehl, I960).
sions. In studying effects of any specific operating
condition on NOX, great care must be taken to see
that other variables are not inadvertently changed
in the process.
Emissions of oxides of nitrogen are functions of
nitrogen and oxygen concentrations in the firebox.
Since there is an abundance of nitrogen in combus-
tion air streams, oxygen, the reactant in short
supply, governs the rate of production of nitric
oxide. Thus, at a given firebox temperature, NO
formation increases with oxygen concentration.
The data in Figure 332 illustrate this point. NO,
concentrations were measured at oxygen concen-
trations from 1.5 to 3.5 percent from an oil-fired
steam generator operating at full load. The listed
NOX concentrations are comparatively small for
equipment such as this, but show the elfect of re-
duced oxygen. The figures were extrapolated to
smaller and larger oxygen concentrations, prin-
cipally for purposes ol illustration. Note that a
reduction from 3. 5 to 1.5 percent oxygen produces
a definite--approximately 20 percent — reduction
in NOX.
Other tests have been made, both on gas and oil
firing, with oxygen concentrations approaching
the theoretical combustion requirement. These
studies indicate that reductions of greater than
30 percent can probably be effected at reasonable
oxygen concentrations, that is, 0. 5 percent or
slightly higher. As will be discussed later in this
chapter, somewhat greater reductions of NO can
be attained by manipulating the points of entry of
combustion air.
As would be expected,__fuels producing higher
_f 1 amejtempera tures aj_sj2_m^od_ucc_ ^r_eaiter_Npx
emissions. The data in Table 145 and Figures
381 and 383 show that average NOX emissions are
some 35 to 50 percent higher during oil firing
than during gas firing. Theoretical flame tem-
peratures are 200° to 300°F higher on oil firing
than on gas firing. Barnhart and Dieh] (1960)
found that the same phenomenon was true for the
burning of c oal, also a hotter fuel than oil. They
report NOX concentrations at several different
installations burning oil, gas, and coal. The val-
ues shown in Figure 381 emphasize the wide varia-
tions in NOX that can be expected from similar
equipment. In general, fuels \\ith higher carbon/
hydrogen ratios produce higher flame tempera-
tures and greater NOX concentrations.
Analyses have shown that NOX emissions vary with
lirebox and burner designs and \\ith operations
that tend to allei t maximum temperature, oxygen
conccnl ration, and resident e time at maximum
temperature. Multiple-port burners, in general,
are associated with larger NO concentrations
-------
542
COMBUSTION EQUIPMENT
20 30 40 i
01YGEN ill 'LUC G1SES % 01 10'. l«f
Figure 382. Effect of residual oxygen or
NOX emissions.
than single -port burners are. This is attributed
to the greater amount of flame cooling by radia-
tion in the case of single-port burners. With
multiple -port burners, a large part of the burn-
ing gases is surrounded by the flames from other
burners. Thus, the inside burners cannot "see"
radiant heat exchange surfaces and cannot, thei'e-
fore, give up their heat rapidly.
A further indicator of the effect of tlame cooling
is the reduction in NOX at reduced firing rates
noted by Barnhart and Diehl (I960). On a test
iurnace at full load, the NO r concentration was
measured as 300 ppm. As the fuel input was re-
duced to one-half and one-quarter load, NO val-
ues dropped to 185 and 145 ppm, respectively.
Maximum measured firebox temperature de-
creased from 3, 165 °F, to 3, 010 °F, to 2, 110° F.
Lowering the burner input apparently allows the
flame to be cooled faster as heat is radiated to
heat transfer surfaces in the walls of the firebox.
Possibly the greatest contributor to greater nitric
oxide formation in large steam, generators is the
preheating of combustion air. In most instances,
combustion air is preheated to about 600 °F be-
fore introduction into the firebox. As a result,
maximum firebox temperatures are 500° to 60l
greater than would be the case if ambient-tempe
ture air were used. There are few good data re
garding the effect of temperatures of combustion
air on NOX formation. Barnhart and Diehl (I960
report only a 17 percent NOX reduction when the
combustion air temperature is reduced from 560
to 97 °F. Another unpublished source reports a
better than 75 p.e_rcent reduction of NO when air
preheat is eliminated.
The nitrogen content of the fuel is generally con-
sidered to have little effect on NOX formation.
There is so much nitrogen in the combustion gas-
es that fuel nitrogen is relatively inconsequential,
Moreover, fuel sulfur does not appear to affect
NOX formation. Sulfur has been injected experi-
mentally into fuel oils. There was no significant
difference in NOX concentrations, even though the
sulfur content was increased 4 times.
Estimating NOX Emissions
Mills et al. (1961) measured NOX emissions fron
a v/ide variety of combustion equipment ranging
from small kilns to large steam generators. Fror
these data, they were able to establish a general
relationship between gross heat input and NOX,
which is shown in Figure 333. Data cover both
gas and oil firing; gross heat inputs range from
less than 10, 000 Btu per hour (9 scfh natural gas)
to 2 million Btu per hour (a 220-megawatt power
plant steam generator). The data for both fuels
plotted to straight lines on log-log coordinates,
even though there are decided differences in fire-
box design, excess air, and flame temperature
over the range of equipment tested. As would be
expected, NOX values are tower (about 50 percent)
for ,giis firing than for oil firing. The carbon/hy-
drogen ratios of most oils are about twice those
of common natural gases, and oil flame tempera-
tures are 200° to 300°F higher.
emissions from almost any combustion de-
NO.
vice can be estimated with the curves of Figure
383. For instance, a 200-horsepower oil-fired
boiler operating at 80 percent overall efficiency
would have a gross heat input of 8, 36 0, 000 Btx1
per hour. From Figure 383, emissions are 1. 1
pounds of NO per hour.
When combustion air is preieated, preheat must
be added to the gross input. For example, a
1, 100, 000-pound-per-hour steam generator has
a rated fuel input of 1 . 6 x 10" Btu per hour. In
addition, combustion air is preheated from 60°
to 600 CF. The difference in combustion air tem-
peratures represents a 14 percent increase in
gross heat input. The adjusted gross input is,
therefore, ] . 82 x 10 Btuperhour, which, the
curve shows, is equivalent to a discharge of 1, 030
pounds of NOX per hour.
-------
Boilers, Heaters, and Steam Generators
543
10,000
/
/ /
^
/
/
/
y
f
t
r
f
t
A
>
/
/
\/\
>—?
j
f
/
~f-
/
,
f
{
f
j~
J
f
f
/
f
f
y
f
^
j
f
<
'
4-
s
t
«v
/^
f
r
j
'{-7»
•3
•• LEGEND
BTU
THE
PRE
THI
BUS
OR
PRE
/
f
.^
INP
FUE
HEAT
S GR
TION
NEAR
SSUR
,
f
,
0 + N
UT IN
L * H
ED CO
APH A
PROC
ONE
E
f
t
f
'
/
f
?
/
f
/
f
f
/
^
f
/
W - -
r H
%
'
EE:
2 (CALCULATED AS N02)^ - •
•LUOES GROSS HEAT IN L_ _ .
AT CONTAINED IN THE
HBUSriON AIR <•
'PLIES ONLY TO COM- i
.SSES TAKING PLACE AT
UMOSPHERE OF ABSOLUTE
t .
' 1
^
1,000
too
2 10
1 0
0 1
0 01
|05 106 10' 10s 13s
AVERAGF RATE DF HEM INPUT TO A dNIT IN A GIVEN CLASS OF COMBUSTION EQUIPMENT Btu li r
Figure 383. Estimation of average unit NOX emissions from similar pieces of
combustion equipment (Mills et al., 1961).
The gross heat input as used in Figure 383 is an
indirect indicator of residence time in the fire-
box, as •well as volume rate oi the effluent. The
curves represent typical industrial equipment
designed neither to decrease nor increase NOX
emissions. For any given piece of equipment,
wide deviations may be expected if excess air
rates or fuel compositions are changed greatly
from normal.
AIR POLLUTION CONTROL EQUIPMENT
A discxission of control equipment serving
large gas- or oil-fired boilers and heaters
must necessarily be theoretical at present
inasmuch as most combustion air contami-
nants have not been controlled, except or'
a pilot plant or experimental basis. Ex-
cept for soot-blo\ving participate collec-
-------
544
COMBUSTION EQUIPMENT
tors, few installations of control equipment
are serving gas- or oil-fired steam generators.
Air pollution control equipment would necessarily
be limited to power plant steam generators or
other fired combustion equipment of comparable
size. Small and intermediate-size boilers and
heaters are not likely to need any control devices
unless fuels are highly contaminated. Normally
no attempt is made to control even those relative-
ly heavy particulate concentrations emitted from
intermediate-size boilers during soot blowing.
Where optimum air pollution control is desired
in smaller equipment, it is normal practice to
burn only clean fuels such as natural gas and low-
sulfur distillate oils and to employ high-efficien-
cy burners.
The control of air contaminants from power plant
steam generators must be considered from a num-
ber of aspects. Some contaminants are amenable
to some degree of control through firebox and
burner modifications, namely, sulfur trioxide,
oxides of nitrogen, and combustible particulates.
Others, such as sulfur dioxide and inorganic par-
ticulates, can be removed only by treatment of
effluent gas. No one control method now under
study is capable of removing all types of con-
taminants emitted to the atmosphere from large
combustion sources.
Sulfur Compounds
Four major methods of controlling sulfur dioxide
and trioxide are potentially feasible: Scrubbers,
cloth filters, dry adsorbers, and electrical pre-
cipitators. Scrubbers and possibly adsorbers are
the only ones that could be used without auxiliary
control mechanisms. With both cloth filters and
precipitators, converting the sulfur dioxide to a
collectible particulate form would first be neces-
sary. This would necessitate either oxidation to
sulfur trioxide and sulfuric acid, or reaction with
an alkaline additive.
Conceivably the problem could be attacked from
the standpoint of sulfur trioxide elimination only.
The trioxide is more objectionable than the di-
oxide because it is highly acidic and imparts
visible opacity to flue gases. Nevertheless, in
most areas, sulfur trioxide collection would be
only an intermediate step in that it would remove
no more than 5 percent of the total sulfur oxides
emitted from a given combustion source. Emis-
sions of sulfur trioxide could be prevented by
(1) inhibiting trioxide formation,. (2) neutraliz-
ing with dry alkaline additives, or (3) selective-
ly controlling the effluent.
Combustible Particulates
As explained in the second part of this chapter,
combustible particulates are not common in
exit gases from small and medium-size com-
bustion equipment except during startup periods.
Generally, combustibles almost always result from
poor maintenance or operation, or improper se-
lection of burner or fuel. In large boilers, fuel
usage and combustion air are carefully monitored
to provide nearly perfect combustion. Combus-
tible contaminants are seldom emitted in concen-
trations sufficient to impart perceptible opacity
or blackness to stack gases. Conceivably, com-
bustibles could be incinerated in an afterburner.
An installation such as this would, however, be
highly unlikely and might occur only where the
fuel burned is essentially a waste product. Ex-
amples might be wood pulp or petroleum products
contaminated with inorganic sludge.
Soot Collectors
The only air pollution control devices that have
found ready acceptance on oil-fired power plant
boilers are dust collectors used to control par-
ticulates during soot blowing. This equipment
serves principally to collect large particulate
matter—greater than 10 microns--that would
otherwise settle in the immediate area. Soot
collectors are used only during periods of soot
blowing. They are not designed to control the
extremely fine particulates emitted during nor-
mal oil firing, particularly the submicron par-
ticulates responsible for opaque plumes.
Dry, small-diameter, multiple cyclones are the
most common soot control devices installed.
This equipment is reasonably inexpensive, and
pressure drops do not usually exceed 4 inches
of water column. Nevertheless, any dry or wet
collector capable of 90 percent or greater col-
lection above 10 microns would be satisfactory in
many locales. More efficient controls, such as
cloth filters may need to be installed to collect
soot-blown air contaminants where fallout is caus-
ing a public nuisance.
No good data are available regarding the soot
collection afforded by centrifugal collectors on
large steam generators. This is due principal-
ly to the difficulty of obtaining representative test
samples during soot blowing. As can be appre-
ciated, dust loadings during tube-lancing oper-
ations are extremely variable. When a given
section of tubes is lanced, the resultant dust
concentration is heaviest when steam or air is
first injected through the lance. Inasmuch as
the efficiency of particulate collection cannot be
measured accurately, the common yardstick for
acceptability of a soot collector is its observed
ability to prevent fallout of large particulates in
downwind areas.
The soot collectors normally encountered at steam
generators are not designed to collect the submi-
cron particles emitted during normal firing. It is
-------
Boilers, Heaters, and Steam Generators
545
doubtful that operators of large oil-fired boilers
•would install more efficient particulate collection
devices unless they also served to remove sulfur
or nitrogen oxides. Most devices that show prom-
ise of SO, removal would also, however, collect
solid particulate matter. Scrubbers, cloth filters,
and precipitators--all possible controls of sulfur
oxides—would remove a major portion of the par-
ticulate matter emitted from oil-fired combustion
equipment.
Sulfur Oxides Collection
The removal of sulfur oxides has been the subject
of considerable investigation, more often in con-
nection •with metal-smelting and coal-firing oper-
ations than with oil-fired steam generators. In
a few installations, these oxides are being collected
from coal- and oil-fired combustion gases. Sev-
eral control methods show definite promise, but
to date, none are sufficiently inexpensive to war-
rant widespread installation.
Scrubbers for Sulfur Oxides
The only full-scale sulfur oxide controls installed
at steam power plants have been scrubbers using
•water or basic aqueous solutions. The few in-
stallations of this sort existing today are in Eng-
land. They can provide 90 percent and greater
removal of both sulfur dioxide and sulfur trioxide.
Three principal scrubbing methods (Bienstock et
al, , 1958) using water solutions have been studied
for power plant use: The nonregenerative lime-
stone process, the Fulham-Simon-Carves process,
and the regenerative sodium suifite process. They
differ mainly in costs, scrubbing vehicles, by-
products, and quantities of wastes produced. In
no case are the by-products of sufficient value to
offset fully the costs of installing and operating
the scrubber.
In the nonregenerative limestone process shown
in Figure 384, a slurry of 5 to 10 percent calcium
carbonate is circulated through a packed tower.
Removal of about 90 percent of the sulfur pro-
duces calcium suifite and calcium sulfate. Slurry
from the tower is crystallized, settled, and
mixed with more limestone before it is recircu-
lated to the scrubber. Calcium hydroxide is more
reactive than the carbonate, but the added cost
makes it unattractive. The calcium sulfite-sul-
fate sludge produced in the process has no by-
product value and presents a waste disposal prob-
lem. If disposal as suifite is not feasible, the
suifite is oxidized to sulfate.
The Fulham-Simon-Carves process employs an
ammoniacal liquor to remove sulfur oxides, also
in a packed tower. Ammonium sulfate fertilizer
and sulfur are by-products. Ammonia reacts
with sulfur oxides to form ammonium sulfate
LIME
CALCIUM SULFATE
Figure 384. The nonregenerative limestone process
for the scrubbing of sulfur oxides (Rees, 1955).
principally, with smaller amounts of suifite, bi-
sulfite, and thiosulfate. After treatment with
sulfuric acid, the spent scrubber liquor is auto-
claved at 200 psig and 360 "F to produce sulfur
and ammonium sulfate. The system is described
in the flow diagram of Figure 385.
The regenerative sodium suifite process of Figure
386 uses a scrubbing solution of sodium suifite
and sodium bisulfite. Sulfur dioxide reacts with
sodium suifite to form sodium bisulfite in the
scrubber as follows:
HO +
SO,
Na2 S°3 '
2 Na H SO
3'
The spent scrubber solution is desulfated and
treated with zinc oxide, which converts the bi-
sulfite back to suifite and precipitates zinc sui-
fite. Zinc suifite is calcined, driving off sulfur
dioxide and regenerating zinc oxide. Sulfur di-
oxide is a by-product of the sodium suifite pro-
cess. The principal waste is a sludge of cal-
cium sulfate.
In any scrubbing process, combustion gases are
cooled to the point where they are no longer mark-
edly buoyant. Auxiliary blowers must be employed
to direct the gases through stacks. Even with ad-
ditional blowers, the scrubbed gases could settle
to ground level in immediately adjacent areas be-
-------
546
COMBUSTION EQUIPMENT
STACK
SLUDGE WASHINGS
DISCARD DISCARD
Figure 385. The FuI ham-Simon-Carves ammoniacal
liquor process for the scrubbing of sulfur oxides
(Rees, 1955).
SODA
ASH
Figure 386. The regenerative sodium sulfite
process for scrubbing sulfur oxides (Rees,
1955).
fore residual sulfur dioxide could be sufficiently
diluted. Localized sulfur dioxide buildups have
in fact occurred at a power station in England
where scrubbing is employed.
Scrubbing by any of the described methods requires
an appreciable investment in equipment and material-
ly increases operating costs at a steam power plant.
Field et al. (1957), made an extensive comparative-
cost study of the three processes using as a base
a coal-fired steam generator exhausting 330, 000
scfm combustion gases. A high-sulfur coal and
a low-sulfur coal were considered. Their find-
ings are summarized in Table 147 along -with a
similar comparison based upon oil burning. The
heat input to an oil-fired steam generator produc-
ing the same volume and sulfur oxide concentra-
tion would be slightly higher than that of the coal-
fired unit. Burning 60 tons of coal per hour
•would produce about the same volume of gases
as burning 283 barrels of U.S. Grade No. 6 fuel
oil per hour.
It can be seen from Table 147 that 90 percent re-
moval by alkaline scrubbing represents an initial
investment of $1, 646, 750 to $4, 945, 400, depend-
ing upon the process and sulfur content of the fuel.
When fuel oil of about 1.6 percent is burned, the
limestone process,-with no by-products, is seen
to be the least expensive of the three to install
and operate, even if credit is allowed for by-
products. At greater sulfur contents, both the
ammonia and sulfite processes become compara-
tively more attractive by reason of increased by-
product value. At fuel sulfur concentrations of
5 percent and greater, the lower operating costs
of the ammonia and sulfite processes are some-
what offset by the greater initial investment re-
quired in comparison with the limestone process.
Under existing economics, most operators con-
sider scrubbing costs prohibitive. Operating
expenses alone range from 10 to 25 percent of
fuel costs on oil firing and are comparatively
higher for coal burning.
GPO 8O6—614—19
-------
Boilers, Heaters, and Steam Generators
547
Table 147. COSTS OF SCRUBBING SOX FROM 20 MILLION SCFM FLUE GASES AT COAL-FIRED'
AND OIL-FIREDb POWER PLANTS (Calculated from data in Field et al. , 1957)
Total investment0
Limestone process
Ammonia process
Sulfite process
Annual operating
costs
$/ton coal burned
Limestone process
Ammonia process
Sulfite process
$/bbl of oil burned
Limestone process
Ammonia process
Sulfite process
0. 083%
$1
3
2
No credit
for products
1.24
2.99
2. 10
0.263
0.643
0.445
SOX by volume
,646,750
,221, 100
, 433,000
Credit for
products
1.62
1.97
0.343
0.418
0. 30%
No credit
for products
1.93
6.54
3.21
0.409
1.380
0.704
SOX by volume
$1, 922, 200
4,945,400
3, 105, 800
Credit for
products
1.43
2. 17
0.303
0.460
aAt 60 tons/hr coal burned, sulfur content 1.5 and 5.0%, respectively.
'•'At 283 barrels/hr No. 6 fuel, sulfur content 1.6 and 5.5%, respectively.
"-Includes working capital at 10% on fired capital.
No provision for interest or return on investment.
eAmmonium sulfate at $32/ton, sulfur at $28/ton, SO2 at $14/ton.
* Anhydrous ammonia at $100/ton delivered.
In this analysis no consideration was made for
waste disposal. If sludge from the processes
could not be dumped into existing facilities, dis-
posal costs could be appreciable. Far greater
quantities of sludge are produced in the limestone
process than in either of the other two methods.
In the sulfite form, limestone process sludge
represents a greater hazard to marine plant and
animal life than calcium sulfate does. Even cal-
cium sulfate is undesirable inasmuch as it set-
tles at the point of discharge, and continuous
operation results in a significant buildup in ponds,
streams, and so forth.
Doghouses and Precipitators
No particulate-matter collector by itself is a
satisfactory control for all sulfur oxides from
oil-fired power plants. This includes centrifu-
gal collectors as well as baghouses and electrical
precipitators. At best these devices can be ex-
pected to remove only sulfur trioxide. Sulfur di-
oxide is a gas at temperatures well below normal
stack conditions and is unaffected by these collec
tors. Nevertheless, attempts have been made to
use baghouses and precipitators principally to
minimize acid fallout damage and visible emis-
sions. Even for these limited purposes, the par-
ticulate collectors have not been completely suc-
cessful.
Cloth filters are effective only as long as the
fabric remains reasonably clean and permeable.
Experiments at oil-fired steam generators have
shown that the collected materials adhere to
fabrics and tend to hydrate upon cooling. The
resulting crust-like formations are almost im-
possible to remove and render the cloth imperme-
able to airflow. Moreover, many fabrics are not
resistant to acids and rapidly disintegrate -with
use.
Several pilot plant studies have been made cover-
ing the feasibility of single-stage precipitators
for controlling visible emissions. In one instance,
a full-size electrical precipitator was installed to
control the 350, 000 scfm discharged from an oil-
fired 1, 200, 000-pound-per-hour power plant steam
generator. The precipitator was on the downstream
side of the air heater where temperatures ranged
from 280° to 350°F. At these temperatures,
much of the sulfur trioxide was in the gaseous
state and passed through the precipitator. A pilot
precipitator showed considerably better SO3 re-
moval and no visible plume when gases were
cooled to about 90°F, well below the water dew
point (based upon moisture content only). Oper-
ation at less than 200°F stack temperatures is
not considered practical by power plant opera-
tors because of resultant corrosion and conden-
sation.
-------
548
Boilers, Heaters, and Steam Generators
The one full-scale electrical precipitaitor was not
consistently effective in controlling plume opacity
below 40 percent during oil firing. Data cover-
ing its operation are included in Table 148. Dur-
ing part of the test period, dolomite or ammonia
additives were injected into the combustion gases
ahead of the precipitator. Table 148 shows that,
with no additives, efficiencies of SO, and par-
ticulate matter removal were less than 50 per-
cent during most trials. Most of the particulate
concentrations shown in Table 148 are abnormal-
ly high for an oil-fired steam generator and indi-
cate a large percentage of combustible particu-
lates. Normal particulate concentrations are be-
tween 0. G2 and 0. 04 grain per scf. The 87 to 90
percent maximum efficiencies were achieved at
loadings of approximately 0. 3 grain per scf. The
26 to 46 percent particulate removal of Runs 3,
4, 6, and 7 are probably more indicative of expec-
ted precipitator performance than the highest ef-
ficiencies are. The reported particulates remov-
al of 0. 04 grain per scf corresponds to about 110
pounds of precipitator catch per hour. Moreover,
the 10 to 20 ppm concentrations of 803 in precip-
itator discharge gases were appreciable. Under
optimum conditions, the opacity of full-scale pre-
cipitator exhaust gases was about 20 percent when
oil of 1. 6 percent sulfur was burned. When the
precipitator was turned off, the plume was of ap-
proximately 60 percent opacity.
Alkaline Additives to Neutralize Sulfur Trioxide
Inexpensive alkaline materials have been injected
into flue gases to neutralize sulfur trioxide and
sulfuric acid. The purposes are to inhibit corro-
sion of boiler tube surfaces and prevent visible
emissions as well as acid fallout damage. The
most common materials used for selective 803
neutralization are calciiim and magnesium oxide,
hydroxides, and carbonates (dolomite). Some at
tention has also been directed to the injection of
gaseous ammonia.
The purpose of injecting calcium and magnesium
compounds is to form neutral sulfates, which are
markedly less corrosive and less hygroscopic
than sulfuric acid and contribute little to visible
opacity. These materials have been mixed with
oil fuels and have been injected at fireboxes and
various downstream points. The hydroxide is
probably the most reactive form. When, how-
ever, hydroxides or carbonates are injected into
extremely hot areas such as the firebox, they are
probably converted to the oxide form before re-
acting with sulfur trioxide.
Some operators report mild success with these
additives in controlling tube corrosion, but their
effect on sulfur trioxide concentrations and visi-
ble plume formation is questionable. Injection
of any calcium or magnesium additive by itself
apparently cannot be expected to lower SOg in
stack gases appreciably. This is indicated by
data in Table 148, which show that the reduction
in sulfur trioxide concentration across the pre-
cipitator is almost the same with and without
dolomite. Some operators have reported a mild
reduction in visible emissions, but there are few
quantitative data on the subject.
The stoichiometric equivalent of 30 ppm by vol-
ume SO-j is 0. 055 grain per scf as calcium carbo-
nate, Ca COj. The additive would not usually be
injected in concentrations greater than 3 times
the stoichiometric equivalent, that is, 0. I6t> grain
per scf as Ca COj. Particles of additive or cal-
cium, or both, or magnesium sulfate do not in
themselves appreciably affect plume opacity when
Table 148. EFFICIENCIES OF SULFUR TRIOXIDE AND PARTICULATE REMOVAL BY A FULL-SCALE
ELECTRICAL PRECIPITATOR SERVING AN OIL-FIRED STEAM GENERATOR
Run No.
Fuel oil rate,
l.OOOlb/hr
Temperature,
'F
Additive,
(tolomite.lb/hr
ammonia, ppm
Participates,
m, gr/3cf
out, gr/scf
i fficiency, %
SO,, in, ppm
out, ppm
S03
in, ppm
out, pprn
i fficiency, %
Drw point,
outlet, "F
1
-
-
:
-
870
871
25.8
13.4
48
-
2
89
295
_
o. 1745
0. 1119
36
835
357
23. 2
16.4
29
-
3
89
292
_
0.0511
0.0382
26
832
860
21.3
12.6
41
-
4
89
308
-
0.0609
0. 0327
46
802
891
ii. 5
16. 7
26
-
5
88
308
_
-
802
843
25.7
15.5
40
-
6
88
-
_
0. 0818
0.0463
43
808
833
22. 8
13.2
42
285
7
86
303
_
0.0650
0.0451
30
817
843
27.4
13. 9
49
8
86
309
0.0928
0.0586
36
858
781
27. 3
13. 5
51
-
9
86
309
817
834
28.0
11.9
58
-
10
87
305
90
0.0947
0.0658
22
786
715
27.5
13.0
53
120
11
87
309
180
0. 1587
0.0381
76
777
809
30.2
10. 3
64
120
12
37
HO
270
0.0554
0.0226
59
767
821
30.7
13.2
57
UO
13
47
-
50
0. 3573
0. 0322
90
758
781
1 1.7
'
14
37
318
270
0. 1814
0. 1585
1 3
725
SOS
25.3
10. 0
61
120
15
87
-
270
0.2964
0.0386
37
7S1
814
40. 5
15.2
62
-
-------
Boilers, Heaters, and Steam Generators
549
additive particles are appreciably larger than 1
micron. The dolomite powders used for this pur-
pose do not normally contain any appreciable per-
centage of material smaller than 10 microns.
Gaseous ammonia reacts more readily with SO?
than dolomite does but is not a great improve-
ment in terms of air pollution. When injected
in proper concentrations, ammonia is reported
by Rendle and Wilsdon (1956) to decrease plume
opacity and SO-> presumably by the formation of
ammonium sulfate. As greater quantities of NHj
are added, it begins to react with SO2, forming
sulfites and bisulfites, which increase rather
than lessen opacity. These latter materials ap-
parently sublime in the range of normal stack
temperatures to form extremely small particles.
The precise control of ammonia concentrations
necessary for plume reduction is not considered
practical for most power plant operations.
The mere addition of ammonia, dolomite, or any
other additive does not remove air contaminants
from a gas stream but converts undesirable sulfur
trioxide and sulfuric acid to a less noxious form.
The addition of any additive to fuel gases without
subsequent control increases particulate emissions.
As -will be noted later in this chapter, stack gas
additives offer some promise -when used in con-
junction with cloth filters. This arrangement
allows better contact of additives with stack gas-
es and also eliminates the problem of increased
particulates.
Other Metal Oxides for Sulfur Dioxide Removal
The use of ammonia, calcium, and magnesium
solids has generally been considered only for
SO-} control and only on a throwaway basis, that
is, the resultant sulfates would either be allowed
to discharge from the stack or would be thrown away
if collection devices were used. Bienstock and
Field (I960) investigated the use of several more
costly metal oxides for the removal of sulfur di-
oxide in flue gases. These investigations were
experimental and used materials considerably
more expensive than limestone. Most of the
additives could not be used on a throwaway basis.
The test materials reacted with sulfur dioxide
to form stable compounds, principally sulfides
and sulfates. Any usable process for sulfur con-
trol in a power plant would require that these
materials be regenerated for continuous use.
This would normally necessitate electrolytic or
thermal decomposition, usually with the produc-
tion of sulfur dioxide as a by-product.
The adsorptive or reactive capacities of several
metal oxides are included in Tables 149 and 150
for the two temperatures of 265° and 625°F,
respectively. Note that manganese oxides and
copper oxides are among those having the great-
est adsorptive capacities. Note also that cal-
cium and magnesium compounds have relatively
low capacities for sulfur dioxide in comparison
with the other materials tested.
If metal oxide adsorption were to be employed at
a power plant, it would probably necessitate a
floating bed absorber with special consideration
taken to minimize the pressure drop across the
unit. In the bench-scale experiments of Bienstock
and Field, a fixed-bed arrangement was used. A
design such as this would probably result in ex-
cessive pressure drops in a steam power plant
and would also require a significant amount of
extra labor.
Doghouses With Dolomite Addition for
Sulfur Trioxide Removol
The principal objection to baghouse operation,
that is, encrustation, blinding, and deteriora-
tion of tLe cloth, can be overcome by injecting
dry dolomite dust into the gas stream ahead of
the collection device. Before startup, the bag
filters are precoated with dolomite. During
operations, the additive is continuously injected
into the gas stream at 2 or 3 times the stoichio-
metric equivalent of the sulfur trioxide content.
Sulfur trioxide reacts in the gas stream and on
the surface of the bags to form calcium sulfate,
a collectible solid.
The data in Table 151 were taken from a pilot in-
stallation used to control gases from an oil-fired
steam generator. The gases -were cooled in a
surface heat exchanger from the normal stack
temperature (290° to 305°F) to less than 185°F.
Dry powdered dolomite was added just ahead of
the centrifugal fan preceding the baghouse. The
dolomite addition rate was approximately 0. 1
grain per scf, about 3 times the stoichiometric
803 equivalent at 20 ppm. Pressure drops
across the unit ranged from 2 to 3 inches water
column. Table 151 shows that the baghouse ef-
fected an SO3 removal of greater than 90 percent
in most instances, as well as reduction of 100°
to 150°F in dew point. The resulting dew points,
that is, 110° to 120°F, were close to the water
dew points of the gases. No visible emissions
were reported from the unit even though the gas-
es were discharged at 165° to 185°F.
This method can be used only to remove sulfur
trioxide and particulate matter and prevent
visible emissions. It has essentially no effect
on the less reactive gas, sulfur dioxide. It
provides considerably greater contact between
sulfur trioxide and dolomite than is afforded by
additive injection without the baghouse. The
method shows promise primarily in that it could
-------
550
COMBUSTION EQUIPMENT
Table 149. EFFECTIVENESS OF VARIOUS COMPOUNDS IN ADSORBING 90% OF 3, 000 ppm
OF SULFUR DIOXIDE FROM GASES AT 265°Fa'b (Bienstock and Field, I960)
Adsorbent
Manganese oxide
Cobalt oxide
Manganese oxide
Manganese oxide
Aluminum -sodium
oxide
Hopcalite
Cobalt oxide
Chromium -sodium
oxides
Nickel oxide
Aluminum -potassium
oxides
Nickel oxide
Sodium carbonate
Sodium stannate
Iron oxide
Sodium aluminate
Cadmium oxide
Copper oxide
Potassium carbonate
Crystalline phase
X-ray analysis)
Mn°1.88
C°3°4
Mn°1.88
7-Mn203
C°3°4
NiO
7-Al203
NiO
Na CO
2 3
o-Fe2O3
NaAlO
CdO
CuO
A1203
Na20
CuO
MnO2
Cr2Oj
Na2°
A1203
K2°
Purity,
wt %
90
97
96
73
25
11
79
00
70
26
91
73
21
90
99
95
93
96
97
99
98
Bulk
density,
g/Co
0 14
0.46
0.50
0.67
0.54
0.93
0.66
0.91
0.74
0.61
1.49
0.98
0.91
0.98
0.90
1. 13
0.89
0.89
02 adsorbed,
g/100 g
adsorbent
3 3
25
23
19
18
13
12
12
9
6
6
5
4
3
3
1
1
1
Preparation
(NH4)2S2Og
^4 jyjH
3
Ppt washed and dried at 130°C
Na2CO3
CoSO4 -
Ppt washed, dried at 130°C, and heated in vacuo at 300 to
340 *C for 20 hr
Electrolysis
4
3pt washed, dried at 130°C, and heated in vacuo at 300 to
40°C for 20 hr
Na COj
MnSO »
Ppt washed, dried at 130"C, and heated in vacuo at 300 to
40'C for 20 hr
Na CO
A123
Ppt washed, dried at 130"C, and heated with H at 600 to
40'C for 20 hr
Dried at 130°C
NaOCl
CoS04 NaOH"
Ppt washed, dried at 130°C, and heated in vacuo at 300 to
40°C for 20 hr
N*2C°3 .
Ppt washed, dried at 130"C, and heated with H at 600 to
64Q°C for 20 hr
Na,CO,
7 3
NiS04 '
3pt washed, dried at 130°C, and heated in vacuo at 300 to
340"C for 20 hr
K CO
Al (SO ) -
Ppt washed, dried at 130°C, and heated with H at 600 to
640"C for 20 hr
NaOCl
NiSO •
Ppt washed, dried at 130°C, and heated in vacuo at 300 to
340°C for 20 hr
Solution of sodium carbonate dried at 130°C and heated in
vacuo at 300 to 340DC for 20 hr
Sodium stannate, dried at 130°C and heated in vacuo at
300 to 340°C for 20 hr
Na2C°3 u
Ppt washed, dried at 130°C, and heated at 300 to 340°C in
a stream of nitrogen for 3 hr
Solution of sodium aluminate, dried .at 130"C, and heated
in vacuo at 300 to 340°C for 20 hr
Na CO
CdS04 '
Ppt washed, dried at 130°C, and heated in vacuo at 370 to
400°C for 20 hr
Na CO
CuSO. ••
4
Ppt washed, compressed at 4,000 psi, dried at 130°C,
and heated in vacuo at 300 to 340°C for 20 hr
Solution of potassium carbonate dried at 130°C and heated
in vacuo at 300 to 340 °C for 20 hr
Bismuth oxide (aBi^C^), molybdenum oxide (MoOj), lead oxide (PbO),
less than 1 g of sulfur dioxide for 100 g of charge.
zinc oxide (ZnO), and calcium hydroxide (Ca{OH)2) adsorbed
aThe 265 °F is close to the stack discharge temperature of power plant steam generators.
^Hourly space velocity of gas, 1, 050 hr~*; mesh size of adsorbent, 8-24.
-------
Boilers, Heaters, and Steam Generators
551
Table 150. EFFECTIVENESS OF VARIOUS COMPOUNDS IN ADSORBING
90% OF 3,000 ppm OF SULFUR DIOXIDE FROM
GASES AT 6Z5°Fa'b (Bienstock and Field, 1960)
Adsorbent
Manga.ne*e oxide
Hopcalite
Copper oxide
Manganese oxide
Crystal I im phase
(X-ray analyse}
Mn°1.88
CuO
Mn°1.88
Cobalt oxide
Cobalt oxide
Lead oxide
Aluminum- souium
oxides
Chr omium - s od i urn
oxides
Co,04
Co O
3 4
PbO
A12°3
1
Sodium aluminate ' NaAlO,
MnO,;
CuO
Na O
Cr.,0,
2
I
I ,
Nickel oxide NiO
Nickel oxide | NiO
i
Aluminum-potassium i >-Al O Al O
oxides 1 K t>
1 2
Cadmium oxide ] CdO [
1 j
Sodium stannate
1
Sodium carbonate | Na CO !
Iron oxide
'a~Fe2°3
Calcium hydroxide |ca(OH);
1
Ca(OH)
CaO
Jurily.
wt %
91
')(
79
11
99
97
100
99
73
25
70
26
96
91
90
73
97
95
99
93
81
19
!
hulk
ei.sity.
0. 1 i
0.67
0.92
0.89
0 50
0 46
0.66
1.23
0. 54
0.91
0 90
0 74
1 49
0.61
1 13
0 91
0 98
0 98
0. 36
U/IOO B
.idsoriii ill
(.1
58
57
56
S!
47
44
18
17
16
10
9
7
6
Preparation
(Ml .l.h.O
MnSO(_._ .
l'|it w.islu .1. ilm '1 .it 1 10 •< .' and hriiti-'l ,n vai uo al JOO to !4(TC
lor 20 hr
MnSO ^~-
1
1'pl u.i-lu-il, drlid .it 1 10"C. and healed in v.u mi al iOO to 340'C
lor 20 hr
Dried in vat no ,it 500 to i!(KC
Na CO
2 1
Ppl w.ishid, >oni|>ribt.fil at 4.000 p,i, dried al HO'C, and healed
111 V.HUO at 300 to t40"C for 20 hr
l-Tleilrolysis
MnSO ^
Ppt u.ishi d, dried at 1 10°C. and healed in vacuo at 300 to 340°C
Mr 20 hr
c .,so4
for 20 hr
NaOCl
CoSO — — »
4 NaOH
for 20 hr
Na.CO,
2 3
Ppt washed, dried at 130nC, and heated in vacuo at 300 to 340aC
for 20 hr
Na.CO,
2 3
Ppt washed, dried at i30DC, and heated with H at 600 to 640°C
for 20 hr
Na CO
2 3
Ppt washed, dried at 130°C, and heated with H^ at 600 to 640°C
for 20 hr
Solution of sodium alunnnate, dried at 130"C, and heated in vacuo
at 300 to 340'C for 20 hr
Na^CO3
MiSO^ *•
for 20 hr
NaOCl
N'S°4 .\aOH~"
Ppt washer!, dried at 130°C, and heated in vacuo at 300 to 340°C
for 20 hr
K CO
AI2'S04»3
Ppt washed, dried at }30°C, and heated with H? at 600 to 640°C
5
5
4
_s
2
for 20 hr
Na^CO
CdSO = —V
Ppt washed, dried at i 30°C, and heated in vacuo at 370 to 400 °C
for 20 hr
Sodium stannato dried at 130°C and heated in vacuo at 300 to
340-C for 20 hr
Solution of sodium carbonate dried at 130"C and heated in vacuo
at 300 to 340°C for 20 hr
iMa CO
Fe(NOJ)? -
Ppt washed, drie'l at 130°C, and heated at 300 to 340°C in a
stream of nitrogen for 3 hr
NaOH
Ca(NO?)2 «.
Ppt washed, dried at 130°C, and heated m vacuo at 300 to 340"C
for 20 hr
Aluminum oxide ("|A1^O3), bismuth oxide (o-Bi^Oj), calcium oxide (CaO), maynesmn
carbonate (K2CO^( adsorbed less than 1 y of sulfur dioxide for 100 g of charge
xide (MgO) molybdenum oxide (MoO5), zinc oxide (ZnO), and potassiurr
»7Ti« 625°F is the approximate temperature of flue gases at the inlet of the air prehe
^Hourly space velocity of gas, 1, 050 hr"1; mesh size of adsorbent, 8-^4.
-------
552
COMBUSTION EQUIPMENT
Table 151. FILTERING SO3 AND DOLOMITE
ADDITIVE IN A PILOT BAGHOUSE
DOLOMITE
STORAGE
TO ATMOSPHERE
GASES FROM OIL-FIRED
STEAM GENERATOR
290°-350°F
BAGHOUSE
HEAT
EXCHANGER
i
Baghouse
Temp,
°F
185
186
185
173
165
175
183
AP,
in. WC
2
2.2
2. 2
2
2
3
3
In
S03,
ppm by vol
15
12.6
19.4
12.5
20.4
10. 5
16.4
Dew point,
o jrb
250
230
260
220
230
250
250
Out
S03,
ppm by vol
3. 1
2.7
1. 1
0.9
0.6
1.5
1.3
Dew point,
°pb
150
120
120
120
120
110
120
aDolomite added at about 3 times stoichiometric SO^ equivalent at 20 ppm
SO-j. Bags precoated with dolomite before startup.
"Apparent dew point. True dew point about 115°F.
be used for total sulfur oxides collection if all
oxides could be converted to the trioxide form.
That an operator would install this control equip-
ment merely to remove the 5 percent or less of
the total sulfur represented by SOj is doubtful.
To be completely successful as a sulfur oxides
control, the baghouse process would necessarily
require a companion process in which sulfur
dioxide would be oxidized to sulfur trioxide pre-
sumably with the aid of a catalyst. Catalytic
oxidation of sulfur dioxide has not yet been
proved economical for oil-fired combustion gas-
es. The point will be discussed later in this
chapter.
If more reactive alkaline additives were used
rather than dolomite, a substantial SO2 removal
might be afforded. A number of possible mate-
rials are listed in Tables 149 and 150; however,
most of these are considerably more expensive
than dolomite.
Some trials have been made with ammonia addi-
tives in conjunction with a pilot baghouse. The
resultant ammonium compounds tended to blind
the filter cloth, causing excessive pressure
drops. The collected material could not be re-
moved from the cloth by normal shaking methods.
Electrical Precipitalors With Additives
The use of dolomite with electrical precipitators
is not as effective as it is with baghouses. This
is probably true for other alkaline materials as
well. A baghouse apparently provides the more
intimate contact required to push the reaction of
basic additives with small concentrations of sul-
fur trioxide. For effective SO^ removal in a
precipitator, the additive would have to be mark-
edly more reactive than dolomite. Moreover,
the resultant neutralized compound would nec-
essarily have to be a partlculate collectible in
the precipitator.
Carbon Adsorption of Sulfur Oxides
Activated carbon is known to adsorb both sulfur
oxides and oxides of nitrogen. Haagen-Smit
(1958) measured the quantity of sulfur dioxide
adsorbed on various grades of high-quality acti-
vated carbon between 77° and 300°F. The data
included in Table 152 indicate that a process such
as this might not be attractive unless a suitably
rapid means of regenerating the carbon were
developed. About 13 parts of sulfur dioxide by
weight are adsorbed per hundred parts of the
best carbons at 150°F. To be practical for a
-------
Boilers, Heaters, and Steam Generators
553
Table 152. ADSORPTION OF SULFUR DIOXIDE ON
ACTIVATED CHARCOAL (Haagen-Srnit, 1958)
Charcoala
A
B
B
C
D
E
F
G
Adsorption
temperature,
oF
77
300
77
150
300
77
300
77
150
220
77
77
77
Flow
rate,
cfm
0. 1
0. 1
0. 1
0. 1
0. 1
0. 1
0. 1
0. 1
0. 1
0. 1
0. 1
0. 1
0. 1
Quantity of SO2,
mg SO2/g
charcoal
153
19
285
127
29
54
30
244
82
62
143
253
225
aA, B, C, and so on are code letters for the different
charcoals tested.
large volume of flue gases, a process such as
this would probably require a floating bed with
continuous carbon reactivation. In view of the
relatively high cost of these carbons, losses on
reactivation would have to be minimal to make
the process attractive economically.
A German process has been developed by Reinluft
that uses a low-grade activated carbon to adsorb
both sulfur dioxide and trioxide as well as NOX.
The process is reported to provide 85 percent or
better removal of both oxides of nitrogen and sul-
fur oxides. Sulfur trioxide is first removed at
600° to 800°F. The gases are then cooled to
200°F where sulfur dioxide and oxides of nitrogen
are adsorbed. SO2 and NO are catalytically oxi-
dized to SO3 and NO2 on the surface of the carbon.
Upon picking up further moisture, the anhydrides
are converted to sulfuric and nitric acids.
Oxidation of Sulfur Dioxide
Considerable attention has been given to the oxi-
dation of sulfur dioxide to the trioxide. There
are workable processes by which 803 could be re-
moved from stack gases; however, most-operators
would not consider it practical to operate this con-
trol equipment unless it "would remove essentially
all sulfur compounds. The oxidation process is
thus a major stumbling point in power plant air
pollution control.
As noted earlier, sulfur oxidizes in two steps,
the first forming the dioxide, the second, the
trioxide. In the manufacture of sulfuric acid,
approximately 97 percent conversion to SO? is
achieved with the use of oxidation catalysts at
carefully controlled temperatures. The optimum
temperature range for SOj formation is between
800° and 840°F. Maximum conversion is ac-
complished at these temperatures with the aid of
vanadium, nickel, and platinum catalysts. Equi-
librium at lower temperatures definitely favors
the higher oxide, SO,. This is generally offset
by decreased reaction rates.
Attempts have been made to adapt the catalytic
process to power plant flue gases with 500 to
1, 000 ppm sulfur dioxide. These trials have not
been successful in obtaining good conversion at
tolerable pressure drops. In most instances, a
fixed catalyst bed has had to be used to obtain 90
percent or greater conversion to sulfur trioxide.
These fixed beds result in pressure drops of 15
inches of water column and greater and would
represent appreciable power expenditures for
the volume handled at power plants. To be prac-
tical, catalytic oxidation would probably have
to be accomplished in a floating bed with a pres-
sure drop of less than 4 inches of water column.
Trials with catalysts injected into the gas stream
have been disappointing in that conversion to SOj
is normally below 50 percent.
If the oxidation process were perfected, it could
be'used in conjunction with baghouse particulate
collectors or possibly with concentrated sulfuric
acid scrubbers. An arrangement such as this
would be expected to remove 90 percent or more
of the sulfur oxides, depending principally upon
the degree of oxidation. Any oxidation process
would have to be preceded by a cloth filter or by
a precipitator to remove materials that might
poison the catalyst or contaminate the acid by-
product. Vanadium and iron oxide catalysts would
probably be preferred over platinum, which is
readily poisoned by arsenic and halogens. Cat-
alysts that are resistant to fouling would have an
obvious advantage for this purpose.
Inhibiting Sulfur Trioxide Formation at
Reduced Oxygen
Some operators report a definite reduction of sul-
fur trioxide and lowering of the dew point when
combustion air is reduced almost to the stoichio-
metric fuel requirement. Crumley and Fletcher
(1956) ran a series of laboratory tests that showed
a reduction of sulfur trioxide as excess air was
reduced from approximately 70 to 9 percent of the
theoretical combustion requirement. More recent-
ly Glaubitz (1963) reported definite dew point low-
ering at 1 to 3 percent excess air, that is, 0. 2 to
0. 6 percent oxygen in flue gas. A significant fact
-------
554
COMBUSTION EQUIPMENT
derived from these investigations is that large
steam generators can be operated at flue gas oxy-
gen concentrations as low as 0. 2 percent without
a marked decrease in combustion efficiency.
In Figure 387 oxygen concentrations in flue gases
from oil-fired equipment are plotted against ap-
parent dew points for fuels of 1.3, 2. 4, and 3. Z
percent sulfur. Note that dew point is a function
of the fuel's sulfur content. This phenomenon is
more pronounced at higher excess air rates. As
excess air is reduced toward zero, the effect of
fuel sulfur is diminished. At 0. 2 percent oxygen,
there is essentially no difference in dew points of
any of the fuels. The resultant common dew point
of 127°F at 0. 2 percent oxygen, is approximately
200°F lower than the reported value for the high-
est sulfur fuel at 3. 0 percent oxygen (15 percent
excess air).
0 15 20
OXYGEN IN FLUE GUSES '.
Figure 387. Dew point raising in an oil-fired
boiler at varying oxygen concentrations (Glaubitz,
1963).
Others have confirmed that operation at 0. 2 to 0. (
percent oxygen is feasible and materially lowers
plume opacity. This operation also results in
heavier concentrations of particulate matter, mos
of which is carbon. One operator reported that
operation at 0. 3 percent oxygen resulted in an in-
crease in thermal efficiency of 1 to 2 percent over
normal operation at 2 percent oxygen.
Operation at small oxygen concentrations warrants
further investigation inasmuch as it would appear
to provide benefits in terms of control of air pollu-
tion as -well as economy of plant operation. Wheth
the increase of combustible contaminants is offset
by the decrease in sulfur trioxide and visible emis
sions remains to be determined. This method
also offers some promise of reducing nitrogen
oxide concentrations in flue gases.
Controlling Oxides of Nitrogen
It is theoretically possible to reduce NOX con-
centrations in combustion process flue gases by
(1) modifying the burners or firebox, or both, to
prevent its formation; (2) decomposing nitric
oxide and possibly nitrogen dioxide back to the
elements oxygen and nitrogen; or (3) scrubbing
the effluent gases. Of the three possibilities,
modifications of the combustion equipment have
been shown to be the most effective and probably
offer the most promise of further NOX reduction
at combustion sources. No practical methods of
decomposition or scrubbing are presently avail-
able.
A furnace modification has been developed by
which steam generator NOX emissions can be
reduced by 40 to 50 percent. These designs are
relatively new, and further refinements can rea-
sonably be expected to be made by which NOX
concentrations can be reduced to a significantly
larger degree. Successful methods tend to low-
er maximum firebox temperatures, promote
faster flame cooling, and reduce oxygen concen-
trations in the highest temperature zones.
Two-Stage Combustion
One of the most effective NO -reducing methods
applied to steam generators has been the splitting
of combustion air in the manner described by
Barnhart and Diehl (I960) as two-stage combus-
tion. With two-stage combustion, only 90 to 95
percent of the theoretical combustion air require-
ment is injected at the burner. The remaining
air is introduced a few feet downstream of the
burner to complete combustion over a somewhat
longer zone. With this arrangement, the total
excess air rate is held to the same figure used
during normal firing, that is, about 10 percent.
This delayed air introduction was found to reduce
-------
Boilers, Heaters, and Steam Generators
555
NOx concentrations in flue gases by 40 to 50 per-
cent. Large steam generators have been ob-
served to operate on two-stage combustion with-
out any measurable loss in combustion efficiency
on either gas or oil firing.
A typical front-fired boiler using two-stage com-
bustion is shown in Figure 388. The major por-
tion of the combustion air requirement is intro-
duced at the burners in the normal manner, while
auxiliary air is introduced downstream after the
last (upper) row of burners.
Figure 388. A front-fired boiler modified
to provide two-stage combustion.
The data in Table 153 show the effect of two-stage
combustion when natural gas and residual fuel oil
are fired. The maximum NOX reduction was
achieved on oil firing with only 90 percent of the
theoretical air requirement introduced through
the burner. On a percentage basis, the reduction
that can be achieved on oil firing is usually some-
what higher than the comparable reduction on gas
firing. This would appear to be due to the flame
temperature differential between the two fuels.
Corner-Fired Steam Generators
A furnace design that provides an NOX reduction
comparable to that provided by two-stage combus-
tion is the so-called corner-fired or tangentially
fired steam generator. A boiler incorporating
this design is shown in Figures 389 and 390. The
corner-fired boiler is a major deviation from con-
ventional front-fired units in that the number of
burner assemblies is considerably less than •with
front-fired units, in which multiple-burner assem-
blies are used. Single-burner assemblies are
mounted in the four corners of the furnace, and
there are usually three or four burner ports in
a vertical line in each assembly. The burners
are designed to provide a relatively long, luminous
flame. Thus, the flames from each burner can
"see" a considerably larger area of wall heat trans-
fer surfaces than those from burners in front-fired
units can. As a result, maximum flame tempera-
tures are apparently lowered enough to reduce
nitric oxide formation considerably. Corner-fired
boilers also employ somewhat higher water circula-
tion rates through furnace tubes. This probably
provides faster cooling of gases in the furnace.
Tangentially or corner-fired steam generators are
reported by Sensenbaugh and Jonakin (I960) to
result in NOX concentrations roughly equivalent
to those of units equipped -with two-stage combus-
tion. Table 154 lists NOX concentrations from
corner-fired boilers and from front-fired units
of approximately the same size. It can be seen
that tangential firing resulted in 53 percent less
NOX on gas firing and 48 percent less NOX on oil
firing than conventional steam generators without
NOx-reducing features could achieve. From both
Table 153. EFFECT OF TWO-STAGE COMBUSTION ON NOX
CONCENTRATIONS FROM A LARGE STEAM GENERATOR AT
NORMAL FULL LOAD (Barnhart and Diehl, I960)
Fuel
Oil and gas
combined
Oil
NOX concentration
All air through
burners,
ppm by vol
525
580
Two- stage combustion
Air through
burners,
% of theoretical3-
95
90
NOX,
ppm by vol
385
305
Reduction,
%
27
47
aThe remaining 15 to 20 percent of combustion air was injected
a few feet downstream of the burner to provide an excess air
ratio of 7 to 10 percent.
-------
556
COMBUSTION EQUIPMENT
Figure 389. A corner-fired steam generator with tilting burners positioned for
varying load and superheat (Combustion Engineering, inc., New York, N.Y.).
Figure 390. Cross-section of a corner-fired
boiler firebox (Combustion Engineering, Inc.,
New York, N.Y.).
corner-fired and front-fired units NOX
emissions were lower on gas firing than
on oil firing, the differences ranging
from 24 to 59 percent.
A corner-fired furnace represents a
major portion of the steam generator
design. It cannot be built into existing
equipment as two-stage combustion can.
Lowering Excess Air
As pointed out earlier, NO concentra-
tions can be lowered 20 to 30 percent by
reducing excess air rates. The mini-
mum possible oxygen concentration is
about 0.2 percent (approximately 1 per-
cent excess air). Nevertheless, simple
lowering of excess air does not appear
to be as effective nor as explosion proof
as two-stage combustion is. It is doubt-
ful, therefore, that this practice will
find much acceptance solely for NOX re-
duction. If, however, low-oxygen com-
bustion were used to inhibit sulfur tri-
oxide formation, some lowering of ox-
ides of nitrogen could be expected as a
bonus.
Eliminating Air Preheat
An obvious method of reducing NOX at
large steam generators is the elimina-
-------
Boilers, Heaters, and Steam Generators
557
Table 154. A COMPARISON OF NOX CONCENTRATIONS IN PPM
BY VO-LUME FROM CORNER-FIRED AND FROM FRONT-
FIRED STEAM GENERATORS21 AT NORMAL FULL LOAD
(Sensenbaugh and Jonakin, I960)
Unit
A
B
C
D
E
F
G
H
Firing
Front
Front
Front
Front
Front
Corner
Corner
Corner
Fuel
Gas
520
290
319
226
164
157
Oil
685
567
505
482
362
309
209
aThe front-fired steam generators were not designed or
operated to minimize NO .
tion of combustion air preheaters. Reducing air
temperatures by 400° to 600°F would lower max-
imum firebox temperatures by an almost equal in-
crement. The lowering of flame temperature
would t>e expected to reduce nitric oxide formation
considerably, though there are few data to support
or refute the theory. As noted earlier, NOX re-
ductions of 12 to 75 percent have been reported.
The only available data have been measured at
pilot equipment. What NOX reduction could be ex-
pected from operating large combustion equipment
without air preheat is not known.
Combustion air preheaters are found on almost
all boilers and heaters of 100 million Btu per
hour and greater gross input. They are located
on the discharge of the equipment just ahead of
the stack. Combustion gas temperatures are
reduced from about 850° to about 300°F at the
point of discharge. Air temperatures are in-
creased by a comparable increment of 450°
to 600°F. Obviously, combustion products
could not be discharged at 800° to 900°F with-
out a gross sacrifice of thermal efficiency. This
heat can, however, be recovered in water or low-
pressure steam rather than air. If the residual
enthalpy were used to heat or vaporize water,
combustion air could be introduced to the furnace
at ambient temperature, and thermal efficiency
•would not suffer.
Other Means of Lowering Flame Temperature
Several other methods of lowering flame tem-
perature are possible. Most of these have defi-
nite economic drawbacks and have not been in-
corporated into large combustion equipment.
Firebox temperatures can be lowered by inject-
ing water, steam, or dilution gases at the burn-
ers. All these possibilities would reduce ther-
mal efficiencies to a greater degree than could be
tolerated at power plants.
NO could theoretically be reduced by adding
water-cooled heat transfer surfaces in the im-
mediate vicinity of the burners. An arrange-
ment such as this would probably be most ef-
fective if the cooling surfaces were spaced be-
tween burners so that flames could "see" more
cooling surfaces. Any arrangement of this sort
would necessarily require rapid circulation
through the heat exchangers, more so than that
provided in common waterwalls of furnaces. No
data are available on the effectiveness of this
technique.
Catalytic Decomposition of NOX
Attempts have been made to decompose NO and
N©2 back to the elements, nitrogen and oxygen.
As noted earlier in this chapter, NO equilibrium
concentrations are still appreciable (100 ppm)
in combustion gases at 1, 800°F. To be accept-
able, decomposition would necessarily have to
be accomplished well below 2,000°F. In this
range, decomposition is extremely slow without
catalysis.
Faith et al. (1957), tested the effectiveness of a
series of commercial and specially prepared
catalysts in decomposing NO-NO£ mixtures.
None of the catalysts were judged sufficiently
active to produce more than a slight decomposi-
tion.
Catalytic decomposition as a control method
would appear to depend solely upon the develop-
ment of a suitable catalyst. When and if this
catalyst is found, it will probably have to be
used at temperatures above 1,000°F. Thus,
it would have to be installed ahead of the air pre-
heater, and possibly ahead of the tube sections
as well.
-------
558
COMBUSTION EQUIPMENT
Scrubbing NO x
Nitrogen dioxide can be absorbed in water and
alkaline solutions though removal efficiencies
are generally low. NO2 reacts with water to
form nitric acid and nitrous acid or nitric ox-
ide. Peters (1955b) reports removal efficiencies
as high as 50 percent at combined NO^ and ^O^
concentrations of 2 percent (20, 000 ppm). NO2
removal decreases greatly, however, at lesser
concentrations, dropping to 10 percent at 2, 000
ppm (Peters, 1955b).
Nitric oxide is much less reactive than NO^ is,
and scrubbing methods are even less successful.
Even if more efficient scrubbing solutions were
found, NO would probably have to be oxidized
first to NO2 to accomplish an adequate cleanup.
Scrubbing of oxides of nitrogen appears to have
most of the economic disadvantages of scrubbing
of sulfur dioxide plus an inherent low removal
efficiency.
Adsorption of NOX
Some laboratory investigations have been made
into the adsorption of NO2 on activated carbon
and silica gel. Both of these media adsorb mea-
surable amounts of NO2, but carbon appears to
offer more promise in the control of emissions
from combustion equipment.
Silica gel is reported to provide efficiencies of
close to 90 percent in adsorbing NO2 and N2O4
at 70°F in large concentrations, that is, about 2
percent by volume. At smaller concentrations,
efficiencies decrease, becoming only about 30
percent at 0. 20 percent (2, 000 ppm). Moreover,
available silica gels have extremely low capaci-
ties for NO2 at these conditions (Peters, 1955b).
In short, adsorption on silica gel appears im-
practical in light of existing data.
Haagen-Smit (1958) noted that oxides of nitrogen
are adsorbed on activated carbon ahead of sulfur
dioxide. Collected NO is then displaced from
the carbon as it becomes saturated with SO2.
The Reinluft process, noted earlier in the chapter,
is used to adsorb both SO-> and NO in a moving
bed of low-grade activated charcoal. It is re-
ported to provide up to 85 percent removal of both
SO2 and NOX, producing sulfuric and nitric acids
as by-products. This process is still in the ex-
perimental stage and has not yet been proved
satisfactory for large steam generators. Never-
theless, it is one of the few methods that offers
hope of high-efficiency control at combustion
sources.
-------
CHAPTER 10
PETROLEUM EQUIPMENT
GENERAL INTRODUCTION
ROBERT C. MURRAY
Senior Air Pollution Engineer
PUMPS
ROBERT H. KINSEY
Air Pollution Engineer
WASTE-GAS DISPOSAL SYSTEMS
DONALD F. WALTERS*
Intermediate Air Pollution Engineer
AIRBLOWN ASPHALT
ROBERT H. KINSEY
Air Pollution Engineer
STORAGE VESSELS
ROBERT C. MURRAY
Senior Air Pollution Engineer
LOADING FACILITIES
ROBERT H. KINSEYt
Air Pollution Engineer
VALVES
ROBERT H. KINSEY
Air Pollution Engineer
CATALYST REGENERATION
STANLEY T. CUFFE*
Air Pollution Engineer
COOLING TOWERS
ROBERT C. MURRAY
Senior Air Pollution Engineer
OIL-WATER EFFLUENT SYSTEMS
ROBERT H. KINSEY
Air Pollution Engineer
MISCELLANEOUS SOURCES
ROBERT H. KINSEY
Air Pollution Engineer
*Now with National Center for Air Pollution Control, Public Health Service, U.S. Department of Health,
Education, and Welfare, Cincinnati, Ohio.
(Now with Lockheed Missiles Systems Company, 1111 Lockheed Way, Sunnyvale, California.
-------
CHAPTER 10
PETROLEUM EQUIPMENT
GENERAL INTRODUCTION
Operations of the petroleum industry can logically
be divided into production, refining, and market-
ing. Production includes locating and drilling oil
•wells, pumping and pretreating the crude oil, re-
covering gas condensate, and shipping these raw
products to the refinery or, in the case'of gas, to
commercial sales outlets. Refining, which ex-
tends to the conversion of crude to a finished sal-
able product, includes oil refining and the manu-
facture of various chemicals derived from petro-
leum. This chemical manufacture is often re-
ferred to as the petrochemical industry. Market-
ing involves the distribution and the actual sale
of the finished products. These activities and
their sources of air pollution are briefly discussed
in this introduction. In the remainder of the chap-
ter, they are discussed much more thoroughly,
and adequate air pollution controls are recom-
mended.
CRUDE OIL PRODUCTION
The air contaminants emitted from crude oil pro-
duction consist chiefly of the lighter saturated
hydrocarbons. The main sources are process
equipment and storage vessels. Hydrogen sul-
fide gas may be an additional contaminant in
some production areas. Internal combustion
equipment, mostly natural gas-fired compres-
sors, contributes relatively negligible quantities
of sulfur dioxide, nitrogen oxides, and particulate
matter. Potential individual sources of air con-
taminants are shown in Table 155.
Contribution of air contaminants from crude-oil
production varies widely with location and con-
centration of producing facilities. In isolated or
scattered locations, many of the sources cannot
be controlled feasibly. Control and pretreat-
ment facilities such as natural gasoline plants
are more likely to be located in more developed
or highly productive areas. These factors are
significant in determining where air contami-
nant emissions from production equipment must
be minimized by proper use of air pollution con-
trol equipment. Control equipment for the vari-
ous air pollution sources associated with crude-
oil production are listed in Table 155. Their ap-
plication can usually result in economic savings.
REFINING
Oil companies have installed or modified equip-
ment not only to prevent economic losses but
also to try to improve community relations, pre-
vent fire hazards, and comply with air pollution
laws. The air contaminants emitted from equip-
ment associated with oil refining include hydro-
carbons, carbon monoxide, sulfur and nitrogen
compounds, malodorous materials, particulate
matter, aldehydes, organic acids, and ammonia.
The potential sources of these pollutants are
shown in Table 156.
Flares and Slowdown Systems
To prevent unsafe operating pressures in process
units during shutdowns and startups and to handle
miscellaneous hydrocarbon leaks, the refinery
must provide a means of venting hydrocarbon vapors
safely. Either a properly sized elevated flare
using steam injection or a series of venturi burn-
ers actuated by pressure increases is satisfactory.
Good instrumentation and properly balanced steam-
to-hydrocarbon ratios are prime factors in the de-
sign of a safe, smokeless flare.
Pressure Relief Valves
In refinery operations, process vessels are pro-
tected from overpressure by relief valves. These
pressure-relieving devices are normally spring-
loaded valves. Corrosion or improper reseat-
ing of the valve seat results in leakage. Prop-
er maintenance through routine inspections, or
use of rupture discs, or manifolding the discharge
side to vapor recovery or to a flare minimizes
air contamination from this source.
Storage Vessels
Tanks used to store crude oil and volatile petro-
leum distillates are a large potential source of
hydrocarbon emissions. Hydrocarbons can be
discharged to the atmosphere from a storage tank
as a result of diurnal temperature changes, fill-
ing operations, and volatilization. Control effi-
ciencies of 85 to 100 percent can be realized by
using properly designed vapor recovery or dis-
posal systems, floating-roof tanks, or pressure
tanks.
561
-------
562
PETROLEUM EQUIPMENT
Table 155. SOURCES AND CONTROL OF AIR CONTAMINANTS FROM
CRUDE-OIL PRODUCTION FACILITIES
Phase of operation
Source
Contaminant
Acceptable control
Well drilling, pumping
Gas venting for production
rate test
Oil well pumping
Effluent sumps
Methane
Light hydrocarbon vapors
Hydrocarbon vapors,
Smokeless flares, wet-gas-
gathering system
Proper maintenance
Replacement with closed vessels
connected to vapor recovery
Storage, shipment
Gas-oil separators
Storage tanks
Dehydrating tanks
Tank truck loading
Effluent sumps
Heaters, boilers
Light hydrocarbon vapors
Light hydrocarbon vapors,
H2S
Hydrocarbon vapors, H,S
Hydrocarbon vapors
Hydrocarbon vapors
H2S, HC, SO2, NOX,
particulate matter
Relief to wet-gas-gathering
system
Vapor recovery, floating roofs,
pressure tanks, white paint
Closed vessels, connected to
vapor recovery
Vapor return, vapor recovery,
vapor incineration, bottom loading
Replacement with closed vessels
connected to vapor recovery
Proper operation, use of gas fuel
Compression, absorption,
dehydrating, water treating
Compressors, pumps
Scrubbers, KO pots
Absorbers, fractionators,
strippers
Tank truck loading
Gas odorizing
Waste-effluent treating
Storage vessels
Heaters, boilers
Hydrocarbon vapors, H2S
Hydrocarbon vapors, H2S
Hydrocarbon vapors
Hydrocarbon vapors, H2S
mercaptans
Hydrocarbon vapors
Hydrocarbon vapors, H;>S
Hydrocarbon, SO2, NOX,
particulate matter
Mechanical seals, packing glands
vented to vapor recovery
Relief to flare or vapor recovery
Relief to flare or vapor recovery
Vapor return, vapor recovery,
vapor incineration, bottom loading
Positive pumping, adsorption
Enclosed separators, vapor re-
covery or incineration
Vapor recovery, vapor balance,
floating roofs
Proper operation, substitute gas
as fuel
Bulk-Loading Facilities
The filling of vessels used for transport of petro-
leum products is potentially a large source of hy-
drocarbon emissions. As the product is loaded,
it displaces gases containing hydrocarbons to the
atmosphere. An adequate method of preventing
these emissions consists of collecting the vapors
by enclosing the filling hatch and piping the cap-
tured vapors to recovery or disposal equipment.
Submerged filling and bottom loading also reduces
the amount of displaced hydrocarbon vapors.
Catalyst Regenerators
Modern refining processes include many opera-
tions using solid-type catalysts. These catalysts
become contaminated with coke buildup during
operation and must be regenerated or discarded.
For certain processes to be economically feasible,
for example, catalytic cracking, regeneration of
the catalyst is a necessity and is achieved by burn-
ing off the coke under controlled combustion con-
ditions. The resulting flue gases may contain
catalyst dust, hydrocarbons, and other impurities
originating in the charging stock, as well as the
products of combustion.
The dust problem encountered in regeneration of
moving-bed-type catalysts requires control by
water scrubbers and cyclones, cyclones and pre-
cipitators, or high-efficiency cyclones, depend-
ing upon the type of catalyst, the process, and the
regenerator conditions. Hydrocarbons, carbon
monoxide, ammonia, and organic acids can be
controlled effectively by incineration in carbon
monoxide waste-heat boilers. The -waste-heat
boiler offers a secondary control feature for
plumes emitted from fluid catalytic cracking
units. This type of visible plume, shown in Fig-
ure 391, whose degree of opacity is dependent
upon atmospheric humidity, can be eliminated by
using the carbon monoxide waste-heat boiler.
-------
General Introduction
563
Table 156. POTENTIAL SOURCES OF EMISSIONS FROM OIL REFINING
Type of emission
Potential source
Hydrocarbons
Sulfur oxides
Carbon monoxide
Nitrogen oxides
Particulate matter
Odors
Aldehydes
Ammonia
Air blowing, barometric condensers, blind changing, blowdown systems, boilers,
catalyst regenerators, compressors, cooling towers, decoking operations, flares,
heaters, incinerators, loading facilities, processing vessels, pumps, sampling
operations, tanks, turnaround operations, vacuum jets, waste-effluent-handling
equipment
Boilers, catalyst regenerators, decoking operations, flares, heaters, incinerators,
treaters, acid sludge disposal
Catalyst regenerators, compressor engines, coking operations, incinerators
Boilers, catalyst regenerators, compressor engines, flares
Boilers, catalyst regenerators, coking operations, heaters, incinerators
Air blowing, barometric condensers, drains, process vessels, steam blowing,
tanks, treaters, waste-effluent-handling equipment
Catalyst regenerators, compressor engines
Catalyst regenerators
Figure 391. A fluid catalytic cracking unit as a source of a visible plume. Use of a carbon monoxide
waste-heat boiler eliminates this plume formation.
-------
564
PETROLEUM EQUIPMENT
Other processes in refining operations employ
liquid or solid catalysts. Regenerating some of
these catalysts at the unit is feasible. Other
catalysts are consumed or require special treat-
ment by their manufacturer. Where regenera-
tion is possible, a closed system can be effected
to minimize the release of any air contaminants
by venting the regenerator effluent to the firebox
of a heater.
Effluent-Waste Disposal
Waste water, spent acids, spent caustic and
other waste liquid materials are generated by
refining operations and present disposal prob-
lems. The waste water is processed through
clarification units or gravity separators. Un-
less adequate control measures are taken, hy-
drocarbons contained in the waste water are
emitted to the atmosphere. Acceptable control
is achieved by venting the clarifier to vapor re-
covery and enclosing the separator with a float-
ing roof or a vapor-tight cover. In the latter
case, the vapor section should be gas blanketed
to prevent explosive mixtures and fires. Spent
waste materials can be recovered as acids or
phenolic compounds, or hauled to an acceptable
disposal site (ocean or desert).
Pipeline Valves and Flanges, Blind Changing,
Process Drains
Liquid and vapor leaks can develop at valve stems
as a result of heat, pressure, friction, corro-
sion, and vibration. Regular equipment inspec-
tions, followed by adequate maintenance can keep
losses at a minimum. Leaks at flange connec-
tions are negligible if the connections are proper-
ly installed and maintained. Installation or re-
moval of pipeline blinds can result in spillage of
some product. A certain amount of this spilled
product evaporates regardless of drainage and
flushing facilities. Special pipeline blinds have,
however, been developed to reduce the amount of
spillage.
In refinery operation, condensate water and
flushing water must be drained from process
equipment. These drains also remove liquid
leakage or spills and water used to cool pump
glands. Modern refining designs provide waste-
water-effluent systems with running-liquid-sealed
traps and liquid-sealed and covered junction
boxes. These seals keep the amount of liquid
hydrocarbons exposed to the air at a minimum
and thereby reduce hydrocarbon losses.
Pumps and Compressors
Pumps and compressors required to move liq-
uids and gases in the refinery can leak product
at the point of contact between the moving shaft
and stationary casing. Properly maintained pack-
ing glands or mechanical seals minimize the emis-
sions from pumps. Compressor glands can be
vented to a vapor recovery system or smokeless
flare.
The internal combustion engines normally used to
drive the compressors are fueled by natural or
refinery process gas. Even with relatively high
combustion efficiency and steady load conditions,
some fuel can pass through the engine unburned.
Nitrogen oxides, aldehydes, and sulfur oxides
can also be found in the exhaust gases. Control
methods for reducing these contaminants are
being studied.
Air-Blowing Operations
Venting the air used for "brightening" and agita-
tion of petrole-um products or oxidation of asphalt
results in a discharge of entrained hydrocarbon
vapors and mists, and malodorous compounds.
Mechanical agitators that replace air agitation
can reduce the volumes of these emissions. For
the effluent fumes from asphalt oxidation,incin-
eration gives effective control of the hydrocar-
bons and malodors.
Cooling Towers
The large amounts of water used for cooling are
conserved by recooling the water in wooden towers.
Cooling is accomplished by evaporating part of
I his water. Any hydrocarbons that might be en-
trained or dissolved in the water as a result of
leaking heat exchange equipment are readily dis-
charged to the atmosphere. Proper design and
maintenance of heat exchange equipment mini-
mizes this loss. Advancement of the fin-fan cool-
ing equipment has also replaced the need of the
conventional cooling tower in many instances.
Process water that has come into contact with a
hydrocarbon stream or has otherwise been con-
taminated with odorous material should not be
piped to a cooling tower.
Vacuum Jets and Barometric Condensers
Some process equipment is operated at less than
atmospheric pressure. Steam-driven vacuum
jets and barometric condensers are used to ob-
tain the desired vacuum. The lighter hydrocar-
bons that are not condensed are discharged to
the atmosphere unless controlled. These hydro-
carbons can be completely controlled by incin-
erating the discharge. The barometric hot well
can also b'e enclosed and vented to a vapor dis-
posal system. The water of the hot well should
not be turned to a cooling tower.
-------
Waste-Gas Disposal Systems
565
EFFECTIVE AIR POLLUTION CONTROL MEASURES
Control of air contaminants can be accomplished
by process change, installation of control equip-
ment, improved housekeeping, and better equip-
ment maintenance. Some combination of these
often proves the most effective solution. Table
157 indicates various methods of controlling
most air pollution sources encountered in the
oil refinery. These techniques are also applicable
to petrochemical operations. Most of these con-
trols result in some form of economic saving.
MARKETING
An extensive network of pipelines, terminals,
truck fleets, marine tankers, and storage and
loading equipment must be used to deliver the
finished petro^um product to the user. Hydro-
carbon emissions from the distribution of prod-
ucts derive principally from storage vessels and
filling operations. Additional hydrocarbon emis-
sions may occur from pump seals, spillage, and
effluent-water separators. Table 158 lists prac-
tical methods of minimizing these emissions
from this section of the industry.
WASTE-GAS DISPOSAL SYSTEMS
INTRODUCTION
Large volumes of hydrocarbon gases are pro-
duced in modern refinery and petrochemical
plants. Generally, these gases are used as fuel
or as raw material for further processing. In
the past, however, large quantities of these gases
Table 157. SUGGESTED CONTROL MEASURES FOR REDUCTION OF
AIR CONTAMINANTS FROM PETROLEUM REFINING
Source
Control method
Storage vessels
Catalyst regenerators
Accumulator vents
Slowdown systems
Pumps and compressors
Vacuum jets
Equipment valves
Pressure relief valves
Effluent-waste disposal
Bulk-loading facilities
Acid treating
Acid sludge storage and
shipping
Spent-caustic handling
Doctor treating
Sour-water treating
Mercaptan disposal
Asphalt blowing
Shutdowns, turnarounds
Vapor recovery systems; floating-roof tanks; pressure tanks; vapor balance;
painting tanks white
Cyclones - precipitator - CO boiler; cyclones - water scrubber; multiple cyclones
Vapor recovery; vapor incineration
Smokeless flares - gas recovery
Mechanical seals; vapor recovery; sealing glands by oil pressure; maintenance
Vapor incineration
Inspection and maintenance
Vapor recovery; vapor incineration; rupture discs; inspection and maintenance
Enclosing separators; covering sewer boxes and using liquid seal; liquid seals
on drains •>
Vapor collection with recovery or incineration; submerged or bottom loading
Continuous-type agitators 'with mechanical mixing; replace -with catalytic
hydrogenation units; incinerate all vented cases; stop sludge burning
Caustic scrubbing; incineration; vapor return system; disposal at sea
Incineration; scrubbing |
Steam strip spent doctor solution to hydrocarbon recovery before air regen-
eration; replace treating unit with other, less objectionable units (Merox)
Use sour-water oxidizers and gas incineration; conversion to ammonium
sulfate
Conversion to disulfides; adding to catalytic cracking charge stock; incin-
eration; using material in organic synthesis
Incineration; water scrubbing (nonrecirculating type)
Depressure and purge to vapor recovery
-------
566
PETROLEUM EQUIPMENT
Table 158. SOURCES AND CONTROL OF HYDROCARBON
LOSSES FROM PETROLEUM MARKETING
Source
Control method
Storage vessels
Bulk-loading facilities
Service station delivery
Automotive fueling
Pumps
Separators
Spills, leaks
Floating-roof tanks; vapor recovery; vapor disposal; vapor balance;
pressure tanks; painting tanks white
Vapor collection with recovery or incineration; submerged loading,
bottom loading
Vapor return; vapor incineration
Vapor return
Mechanical seals; maintenance
Covers; use of fixed-roof tanks
Maintenance; proper housekeeping
were considered waste gases, and along with waste
liquids, were dumped to open pits and burned,
producing large volumes of black smoke. With
modernization of processing units, this method of
-waste-gas disposal, even for emergency gas re-
leases, has become less acceptable to the indus-
try. Moreover, many local governments have
adopted or are contemplating ordinances limit-
ing the opacity of smoke from combustion process-
es.
Nevertheless, petroleum refineries are still faced
with the problem of safe disposal of volatile liq-
uids and gases resulting from scheduled shut-
downs and sudden or unexpected upsets in process
units. Emergencies that can cause the sudde'n
venting of excessive amounts of gases and vapors
include fires, compressor failures, overpres-
sures in process vessels, line breaks, leaks, and
power failures. Uncontrolled releases of large
volumes of gases also constitute a serious safety
hazard to personnel and equipment.
A system for disposal of emergency and waste
refinery gases consists of a manifolded pres-
sure-relieving or blowdown system, and a blow-
down recovery system or a. system of flares for
the combustion of the excess gases, or both. Many
refineries, however, do not operate blowdown
recovery systems. In addition to disposing of
emergency and excess gas flows, these systems
are used in the evacuation of units during shut-
downs and turnarounds. Normally a unit is shut
down by depressuring into a fuel gas or vapor
recovery system, with further depressuring to
essentially atmospheric pressure by venting to
a low-pressure flare system. Thus, overall
emissions of refinery hydrocarbons are sub-
stantially reduced.
Refinery pressure-relieving systems, common-
ly called blowdown systems, are used primarily
to ensure the safety of personnel and protect
equipment in the event of emergencies such as
process upset, equipment failure, and fire. In
addition, a properly designed pressure relief
system permits substantial reduction of hydro-
carbon emissions to the atmosphere.
The equipment in a refinery can operate at pres-
sures ranging from less than atmospheric to
1, 000 psig and higher. This equipment must be
designed to permit safe disposal of excess gases
and liquids in case operational difficulties or
fires occur. These materials are usually re-
moved from the process area by automatic safety
and relief valves, as well as by manually con-
trolled valves, manifolded to a header that con-
ducts the material away from the unit involved.
The preferred method of disposing of the waste
gases that cannot be recovered in a blowdown
recovery system is by burning in a smokeless
flare. Liquid blowdowns are usually conducted
to appropriately designed holding vessels and
reclaimed.
A blowdown or pressure-relieving system con-
sists of relief valves, safety valves, manual
bypass valves, blowdown headers, knockout ves-
sels, and holding tanks. A blowdown recovery
system also includes compressors and vapor surge
vessels such as gas holders or vapor spheres.
Flares are usually considered as part of the blow-
down system in a modern refinery.
The pressure-relieving system can be used for
liquids or vapors or both. For reasons of
economy and safety, vessels and equipment dis-
-------
Waste-Gas Disposal Systems
567
charging to blowdown systems are usually segre-
gated according to their operating pressure. In
other words, there is a high-pressure blowdown
system for equipment working, for example,
above 100 psig, and low-pressure systems for
those vessels with working pressures below 100
psig. Butane and propane are usually discharged
to a separate blowdown drum, which is operated
above atmospheric pressure to increase recov-
ery of liquids. Usually a direct-contact type of
condenser is used to permit recovery of as much
hydrocarbon liquid as possible from the blow-
down vapors. The noncondensables are burned
in a flare.
A pressure-relieving system used in one modern
petroleum refinery is shown in Figure 392, This
system is used not only as a safety measure but
also as a means of reducing the emission of hy-
drocarbons to the atmosphere. This installation
actually includes four separate collecting systems
as follows: (1) The low-pressure blowdown sys-
tem for vapors from equipment with working
pressure below 100 psig, (2) the high-pressure
blowdown system for vapors from equipment
with working pressures above 100 psig, (3) the
liquid blowdown system for liquids at all pres-
sures, and (4) the light-ends blowdown for butanes
and lighter hydrocarbon blowdown products.
The liquid portion of light hydrocarbon products
released through the light-ends blowdown sys-
tem is recovered in a drum near the flare. A
backpressure of 50 psig is maintained on the
drum, which minimizes the amount of vapor that
vents through a backpressure regulator to the
high-pressure blowdown line. The high-pres-
sure, low-pressure, and liquid -blowdown sys-
tems all discharge into the main blowdown ves-
sel, Any entrained liquid is dropped out and
pumped to a storage tank for recovery. Offgas
from this blowdown drum flows to a vertical
vessel with baffle trays in which the gases are
contacted directly with water, which condenses
some of the hydrocarbons and permits their re-
covery. The overhead vapors from this so-
called sump tank flow to the flare system mani-
fold for disposal by burning in a smokeless flare
system.
The unique blowdown system shown in Figure 393
was installed primarily as an air pollution con-
trol measure. The system serves a delayed cok-
TO FLARE STACK
#«
LIGHT-ENDS CONDENSATE RECOVERY
Figure 392. Typical modern refinery blowdown system.
-------
568
PETROLEUM EQUIPMENT
GAS OIL
STEAM, HATER,
AND HYDROCARBONS
GAS OIL
SCRUBBER
AIR
CONDENSERS
X
X
AIR
SUB-COOLER
ACCUMULATOR
OIL
SEPARATION
TANK
FLARE
HATER SEAL
DRUM
SKIMMEO OIL TO STORAGE
HATER TO TREATING UNIT
Figure 393. Coke drum blowdown system.
ing unit. In this process, each drum is taken off
the line as it is filled with coke. The drum is
then purged with steam and cooled with water.
The steam-water-hydrocarbon mixture flows to
a gas oil scrubber whose primary purpose is to
remove entrained coke fines. At the same time
some heavier hydrocarbons are condensed, and
the mixture is pumped to a settling tank. The
scrubbed gases flow to an air-cooled condenser
and then through an air-cooled subcooler to an
accumulator drum.
The air condenser sections are controlled by
temperature and used as needed. The design
outlet temperature range of the condensers is
212° to 270°F, and about 20(TF for the sub-
cooler.
The oil layer in the accumulator is skimmed
off and pressured to the oil-settling tank white
the water phase is sewered. Offgas flows through
a water seal to a smokeless elevated flare. The
oil-settling tank is a 3, 000-barrel fixed-roof tank
equipped with an oil skimmer. The oil phase is
pumped to storage, and the water is sewered
for further treatment at a central waste-water
facility.
This installation has eliminated a previous nui-
sance from heavy oil mist and the daily emission
of approximately 5-1/2 tons of hydrocarbons.
Design of Pressure Relief System
The design of a pressure relief system is one of
the most important problems in the planning of a
refinery or petrochemical plant. The safety of
personnel and equipment depends upon the prop-
er design and functioning of this type of system.
The consequences of poor design can be disastrous.
A pressure relief system can consist of one re-
lief valve, safety valve, or rupture disc, or of
several relief devices manifolded to a common
header. Usually the systems are segregated
according to the type of material handled, that
is, liquid or vapor, as well as to the operating
pressures involved.
The several factors that must be considered in
designing a pressure relief system are (1) the
governing code, such as that of ASME (American
Society of Mechanical Engineers, 1962); (2) char-
acteristics of the pressure relief devices; (3) the
design pressure of the equipment protected by
-------
Waste-Gas Disposal Systems
569
the pressure relief devices, (4) line sizes and
lengths, and (5) physical properties of the mate-
rial to be relieved to the system.
In discussing pressure relief systems, the
terms commonly used should be defined. The
following definitions are taken from the API
Manual (I960).
1. A relief valve is an automatic pressure-
relieving device actuated by the static pres-
sure upstream of the valve. It opens further
with increase of pressure over the set pres-
sure. It is used primarily for liquid service.
2. A safety valve is an automatic relieving de-
vice actuated by the static pressure upstream
of the valve and characterized by full opening
or pop action upon opening. It is used for
gas or vapor service.
3. A rupture disc consists of a thin metal di-
aphragm held between flanges.
4. The maximum allowable working pressure
(that is, design pressure), as defined in the
construction codes for unfired pressure ves-
sels, depends upon the type of material, its
thickness, and the service condition set as
the basis for design. The vessel may not be
operated above this pressure or its equivalent
at any metal temperature higher than that
used in its design; consequently, for that
metal temperature, it is the highest pressure
at which the primary safety or relief valve
may be set to open.
5, The operating pressure of a. vessel is the
pressure, in psig, to which the vessel is
usually subjected in service. A processing
vessel is usually designed to a maximum
allowable working pressure, in psig, that
•will provide a suitable margin above the
operating pressure in order to prevent any
undesirable operation of the relief valves.
(It is suggested that this margin be approxi-
mately 10 percent higher, or 25 psi, which-
ever is greater. )
6. The set pressure, in psig, is the inlet pres-
sure at which the safety or relief valve is
adjusted to open.
7. Accumulation is the pressure increase over
the maximum allowable working pressure of
the vessel during discharge to the safety or
relief valve expressed as a percent of that
pressure or pounds per square inch.
8. Over pressure is the pressure increase over
the set pressure of the primary relieving de-
vice. It is the same as accumulation when
the relieving device is set at the maximum
allowable working pressure of the vessel.
(From this definition note that when the set
pressure of the first safety or relief valve
to open is less than the maximum allowable
working pressure of the vessel the over-
pressure may be greater than 10 percent of
the set pressure of the first safety or relief
valve. )
9. Blowdown is the difference between the set
pressure and the reseating pressure of a.
safety or relief valve, expressed as a per-
cent of a set pressure or pounds per square
inch.
10. Lift is the rise of the disc in a safety or re-
lief valve.
11. Backpressure is the pressure developed on
the discharge side of the safety valves.
12. Superimposed backpressure is the pressure
in the discharge header before the safety valve
opens (discharged from other valves).
13. Built-up backpressure is the pressure in the
discharge header after the safety valve opens.
Safety Valves
Nozzle-type safety valves are available in the con-
ventional or balanced-bellows configurations.
These two types of valves are shown schematic-
ally in Figures 394 and 395. Backpressure in the
piping downstream of the standard-type valve
affects its set pressure, but theoretically, this
backpressure does not affect the set pressure of
the balanced-type valve. Owing, however, to
imperfections in manufacture and limitations of
practical design, the balanced valves available
vary in relieving pressure when the backpres-
sure reaches approximately 40 percent of the set
pressure. The actual accumulation depends up-
on the manufacturer.
Untj.1 the advent of balanced valves, the general
practice in the industry was to select safety valves
that start relieving at the design pressure of the
vessel and reach full capacity at 3 to 10 percent
above the design pressure. This overpressure
was defined as accumulation. With the balanced
safety valves, the allowable accumulation can be
retained with smaller pipe size.
Each safety valve installation is an individual
problem. The required capacity of the valve
depends upon the condition producing the over-
pressure. Some of the conditions that can cause
overpressure in refinery process vessels, and
the required relief capacity for each condition
are given in Table 159.
-------
570
PETROLEUM EQUIPMENT
SPRING
TO VENT
LINE
D
TO VENT
LINE
FROM PRESSURE VESSEL
(BACK PRESSURE DECREASES SET PRESSURE)
FROM PRESSURE VESSEL
(BACK PRESSURE INCREASES SET PRESSURE)
Figure 394. Schematic diagram of standard safety valves (Samans, 1955).
SPRING
BONNET VENT
TO VENT
LINE
FROM PRESSURE VESSEL FROM PRESSURE VESSEL
(BACK PRESSURE HAS VERY LITTLE EFFECT ON SET PRESSURE)
Figure 395. Schematic diagram of balanced safety valves (Samans, 1955)
Rupture Discs
A rupture disc is an emergency relief device
consisting of a thin metal diaphragm carefully
designed to rupture at a predetermined pressure.
The obvious difference bet-ween a relief or safety
valve and a rupture disc is tha.t the valve reseats
and the disc does not. Rupture discs may be in-
stalled in parallel or series with a relief valve.
To prevent an incorrect pressure differential
-------
Waste-Gas Disposal Systems
571
from existing, the space between the disc and
the valve must be maintained at atmospheric
pressure. The arrangement of a rupture disc
to supplement a relief or safety valve is shown
in Figure 396. In an installation such as this,
the relief or safety valve is sized by convention-
al methods, presented later, and the rupture
disc is usually designed to relieve at 1. 5 times
the maximum allowable working pressure of the
vessel (Bingham, 1958).
Table 159. OPERATIONAL DIFFICULTIES OF A REFINERY AND REQUIRED RELIEF CAPACITIES
(American Petroleum Institute, I960)
Condition
Required relief capacity
Relief valve
for liquid relief
Safety relief valve for vapor relief
Closed outlets on vessels
Cooling-water failure to condenser
Maximum liquid
pump-in rate
Top-tower reflux failure
Sidestream reflux failure
Lean-oil failure to absorber
Accumulation of noncondensables
Entrance of highly volatile
material:
Water into hot oil
Light hydrocarbons into hot oil
Overfilling storage or surge vessel
Failure of automatic controls:
Tower pressure controller,
to closed position
All valves, to closed position,
except water and reflux valves
Abnormal heat or vapor input:
Fired heaters or steam reboilers
Split reboiler tube
Internal explosions
Chemical reaction
Hydraulic expansion:
Cold fluid shut in
Lines outside process area
shut in
Exterior fire
Maximum liquid
pump-in rate
No operational
requirement
Nominal size
Nominal size
Total incoming steam and vapor, plus that generated
therein under normal operation
Total incoming steam and vapor, plus that generated
therein under normal operation, less vapor condensed
by sidestream reflux. Consideration may be given to
the suppression of vapor production as the result of the
valve's relieving pressure being above operating pres-
sure, with the assumption of constant heat input
Total vapor to condenser
Difference between vapor entering and leaving section
None
Same effect in towers as for cooling-water failure or
overfilling in other vessels
For towers--usually not predictable
For heat exchangers—assume an area twice the
internal cross-sectional area of one tube so as to
provide for the vapor generated by the entrance of
the volatile fluid
Total normally uncondensed vapor
No operational requirement
Estimated maximum vapor generation including non-
condensable from overheating
Steam entering from twice the cross-sectional area
of one tube
Not controlled by conventional relief devices, but by
avoidance of circumstances
Estimated vapor generation from both normal and un-
controlled conditions
Estimate by the method given in Sect 6 of API Manual,
RP 520
-------
572
PETROLEUM EQUIPMENT
TO VENT RELIEF VALVE
ATTACHES HERE
RUPTURE DISC
CONNECTION FOR
PRESSURE GAGE
*• TO VESSEL
In determining the size of a disc, three important
effects that must be evaluated are low rupture
pressure, elevated temperatures, and corrosion.
Minimum rupture pressures with maximum
recommended temperatures are given in Table
160. Manufacturers can supply discs that are
guaranteed to burst at plus or minus 5 percent
of their rated pressures.
The corrosive effects of a system determine
the type of material used in a disc. Even a
slight amount of corrosion can drastically short-
en disc life. Discs are available with plastic
linings, or they can be made from pure carbon
materials.
pci i cc VALVE
PROCESS GAS LINE ATTACHES HERE
Sizing rupture discs
Th.e causes of overpressure, and the required
capacity for a disc can be determined by meth-
ods previously discussed.
The first estimate of the required rupture disc
area can be made by using the formula (Bingham,
1958):
Q
A =
11.4 P
(108)
Figure 396. Rupture disc and relief valve installation:
(top) How rupture disc gives secondary protection,
(bottom) assembly protects relief valve from disc
fragments (Bingham, 1958).
where
A = area of disc, in.
Table 160. MINIMUM RUPTURE PRESSURES, psig (Puleo, I960;
Copyrighted by Gulf Publishing Co. , Houston, Texas)
Disc size,
in.
1/4
1/2
1
1-1/2
Z
3
4
6
8
10
12
16
20
24
Aluminum
310
100
55
40
33
23
15
12
9
7
6
5
3
3
Aluminum
lead lined
405
160
84
60
44
31
21
17
19
16
10
8
8
8
Copper
500
250
120
85
50
35
28
25
35
42
55
55
70
60
Copper
lead lined
650
330
175
120
65
50
40
25
35
42
55
55
70
60
Silver
485
250
125
85
50
35
28
24
27
--
--
--
--
Platinum
500
250
140
120
65
45
35
26
--
--
--
--
--
--
Nickel
950
450
230
150
95
63
51
37
30
47
--
--
--
--
Monel
1,085
530
265
180
105
74
58
43
34
28
360
270
215
178
Inconel
1, 550
775
410
260
150
105
82
61
48
_-
--
_-
--
--
321 or 347
stainless
1, 600
820
435
280
160
115
90
70
55
45
45
33
27
65
Maximum 250°F 250T
recommended 120°F 120°C
temperature
(base temperature, 72°F[20°C])
250°F
120°C
250°F
120°C
250°F
120°C
600°F
320°C
750°F 800°F
400°C 430°C
900°F
480°C
600°F
320°C
-------
Waste-Gas Disposal Systems
573
Q = required capacity, cfm air
3.
P = relieving pressure, psia.
When the overpressure is caused by an explosion,
a method of sizing discs has been presented by
Lowenstein (1958). In an explosion, a relief or
safety valve does not respond fast enough and a
rupture disc is required.
The maximum allowable backpressure in an in-
dividual discharge line from a disc is 10 per-
cent of the disc's bursting pressure. The max-
imum allowable backpressure for a manifolded
blowdown header serving rupture discs and re-
lief or safety valves should not exceed the in-
dividual allowable backpressure for the lowest
rupture pressure, or 25 percent of the lowest
set pressure of the included valves, -whichever
is less.
Sizing liquid safety valves
To calculate the required area for a relief
valve handling liquid and with constant back-
pressure, the following formula may be used:
A =
0. 5
(109)
C = constant for relief valve and percent
accumulation
Q = required liquid flow at flowing tem-
perature, gpm
P = relieving pressure at inlet, psia
P = discharge pressure at outlet, psia
S = specific gravity of fluid at flowing
conditions.
For one manufacturer, the valve constant is
27. 2. The overpressure factor for 10 percent
accumulation, or overpressure, is determined
from Figure 397 to be 0. 6. Equation 109 be-
comes, therefore, for this particular type of
valve with a 10 percent accumulation:
A =
Q.
II
^- \ S
•32 Ipi-
,0.5
(110)
The use of a balanced relief valve such as the
bellows type permits a variable percent back-
pressure but introduces another variable into
the valve-sizing equation. Equation 110 now
becomes:
where
A = effective opening of valve, in.
A =
16.32 L
f
0.5
(111)
»*T£»
SET PRESSURE - 100 psig
RUTED CHPHCIT1 »T 25 OP = 83 5 epm
FUCTOR = 0 6 (FRO* CUKVF'
c««cinr HMO ar = a t dis s> = 50 i
IUO»»BLE OVERPRESSURE. %
Figure 397. Overpressure sizing factor for liquid relief valves (Consolidated Safety Relief
Valves, Manning, Maxwell,and Moore, Inc., Catalog 1900, Tulsa, Okla.).
-------
574
PETROLEUM EQUIPMENT
•where nomenclature is as before and L£ is the
variable backpressure flow factor. This factor
is supplied by the particular manufacturer, typ-
ified by Figure 398.
Sizing vapor and gas relief
and safety valves
The theoretical area required to vent a given
amount of gas or vapor can be calculated by
assuming adiabatic reversible flow of an ideal
gas through a nozzle. Based upon these as-
sumptions, the following equation can be de-
rived:
A =
where
A
W
C
CD
W
CCDP1
T "I
"M j
0.5
(112)
area, in.
flow capacity, Ib/hr
nozzle gas constant, which varies as
the ratio of specific heats, as shown
in Figure 399
coefficient of discharge for nozzle or
orifice
P = inlet pressure, psia
T = inlet temperature, "R
M = average molecular -weight of gas.
k =
c /c
P
specific heat at
constant pressure
specific heat at
consta.nt volume
For hydrocarbon vapors where the actual value
of k is not known, the conservative value of
k = 1. 001 has been commonly used (C = 315).
The nozzle discharge coefficient for a well-
designed relief valve is about 0. 97. Hydro-
carbon gases can be corrected for nonideality
by use of a compressibility factor. With these
assumptions, equation 112 reduces to:
m -0.5
A =
W
306 P,
M
(113)
where
A = area, in.
Z = compressibility factor. For hydro-
carbons, Z may be determined from
Figure 400 or is usually taken as 1.0
if unknown.
0.4
0.3
0.2
0.1
D
EXAM
^
CAP
»cin c
'LE
SET PRESSURE - 100 psig
FLO*IKG PRESSURE AT 10'. OVERPRESSURE - 100 + 10
CONSTANT BACK PRESSURE - 75 psig OR 89 7 psia
JRYE
= 110 psig OR
^
124 7 psia
BACK PRESSURE PERCENTAGE = 89 7/124.7 - 71. W
FOLLOW DOTTED LINE FROM BACK PRESSURE PERCENTAGE
""SCALE TO FLO* CORRECTION FACTOR SCALE AND FIND
THAT FACTOR EQUALS 0 93
1
~
^^
Sv
i
0*
U4
,
USE ONLY Fl
»ITN tami
X
s
R STANDARD VALVES
HI BACKPRESSURE
\
\
\
>
I
\
\
\
]
20 30 40
BACK PRESSURE PERCENTAGE (ABSOLUTE) =
50 60 70
BACK PRESSURE - (ABSOLUTE)
100
FLOKING PRESSURE (SET PRESSURE + OVERPRESSUBE)-(ABSOLUTE)
x 100
Figure 398. Overpressure sizing factor for standard vapor safety valves (Consolidated Safety
Relief Valves, Manning, Maxwell, and Moore, Inc., Tulsa, Okfa.).
-------
Waste-Gas Disposal Systems
575
1
350
340
330
3ZO
310
1
Figut
Uonhl
'
/
/
/"
'
/
/
/
/
/
/
/
FLO) FOMUIA
C = 5
zajk _
i *
,x
^
CI1CDLITIONS
\
;
1 K"'
,*
An approximation of the absolute temperature at
the valve outlet can be calculated under critical
flow conditions from the folio-wing equation:
T = T j- -1 (115)
where
T - temperature at valve outlet, R
T = temperature at valve inlet, °R
k = ratio of specific heats, c /c .
Before 1957, capacity conversion formulas for
valve sizing in petroleum service were given in
k = c/cv formulas have been incorporated in Section
VIII of the ASME Unfired Pressure Vessel Code
e 399. Nozzte gas constant (American Society of naA?i
nir-ll Cnrrinoorc fOI^1! \,L7O^I.
Where the critical pressure ratio is such that
subsonic fluid velocities are obtained, a correc-
tion factor Kj-,p as shown in Figure 401 may be
applied. For more precise calculations, the
following formula may be used:
A =
W
2,370 P
: E
ZT/M
0.5
(114)
The catalogs of relief valve manufacturers are
also sources of valve-sizing methods and spe-
cific details about various types of valves.
Installing relief and safety valves
and rupture discs
The same general rules for discharge piping
apply equally to relief and safety valves and
rupture discs. Inlet piping should be such that
0.1
REDUCED PRESSURE PR=-jr
2.0
2.5
3.0
Figure 400. Compressibility constants for hydrocarbons (American
Petroleum Institute, 1960).
-------
576
PETROLEUM EQUIPMENT
0.82
0.66
0 70
5 0.78
UJ
» 0 62
0 86
0 90
0.94
06
i
0.1 08
CQMCCTIW FACTOR, ftbp
09
Figure 401. Correction factor
(Conison, 1960).
for subsonic flow
there is direct and unobstructed flow between the
vessel and the relief device. A conservative
limit for the total pressure drop between the
vessel and the safety valve is 2 percent of the
absolute relieving pressure.
The discharge piping for relief and safety valves
and rupture discs should have a minimum of
fittings and bends. There should be minimum
loading on the valve, and piping should be used
with adequate supports and expansion joints.
Suitable drains should be used to prevent liq-
uid accumulation in the piping and valves.
Figures 402, 403, 404, and 405 illustrate good
design of relief device piping (for further de-
tails on Figures 403 and 405, see Tables 161
and 162, respectively).
Knockout vessels
In a vapor blowdown system, a knockout drum
is used to remove entrained liquids from the
gas stream. This is particularly important if
the gas is to be burned in a smokeless flare. A
knockout drum can be quickly sized or checked
by the use of a graphical calculation (see Fig-
ure 406; Kerns, I960). The diameter of the
drum is based on the allowable vapor velocity,
which can be determined by the well-known
equation:
= *
0.5
(116)
where
u = maximum allowable vapor velocity,
ft/sec
p = liquid density, Ib/ft
p = vapor density, Ib/ft
4> ~ a constant. Use - 0. 2 to 0.3.
= 0. 22.1 is often used for light liq-
uid loading.
DESICN BEND TO
TAKE CMC OF
VESSEL EXPANSION
Figure 402. Inlet piping for safety valves:
(left) Horizontal vessel nozzles, when used
for safety valve mounting can be connected
in manner illustrated; (right) valve can be
isolated from process fluid in manner illus-
trated (Driskell, I960; copyrighted by Gulf
Publishing Co., Houston, Texas).
The maximum design velocity should be 0. 5 umax
to allow for gas surges.
Light liquid loads indicate the use of a vertical
vessel, and heavy liquid loads, a horizontal
vessel. The optimum dimensions of the vessel
will have a length-to-diameter ratio (L/D) of 3
for larger drums and 4 for smaller drums, and
never less than 4 feet between tangents (Kerns,
1960).
When "wire mesh is used in the drum as an added
precaution against mist entrainment, the selected
diameter should be multiplied by 0. 65 for con-
ventional mesh and 0. 62 for high-capacity mesh
(Neimeyer, 1961).
Surge time for most designs is 5 to 10 minutes.
The graoliical sizing method of Figure 406 is
based on a surge time of 7-1/2 minutes.
The preliminary sizing of a knockout drum is
illustrated by the following example:
Example 35
Given:
Gas flow 100 ft /sec (under flow conditions)
Vapor density, p , 0.1 Ib/ft3
•J
Liquid density, p-p 50 Ib/ft .
-------
Waste-Gas Disposal Systems
577
J- — -
-{
A.
C.
LONG-RADIUS ELBOW
PROVIDE HORIZONTAL RUN HERE IF
NECESSARY BECAUSE OF EXPANSION
PLAN
DRAIN TO
MANIFOLD
STACK
ENTRANCE ANGLED TO
REDUCE FRICTION
ELEVATION
N
DISCHARGE MANIFOLD
DRAIN
CAP
E.
PURGE GAS INERT
TO PROCESS FLUID
Figure 403. Discharge piping for relief and safety valves: (A) For
air or gas service, (B) for air, gas, or steam service, (C) for
liquid service, (0) for steam or vapor service, (E) for steam or
vapor service to 3-inch pipe, (F) closed system for hazardous
service, (G) open system for pyrophoric gases (Oriskell, I960;
copyrighted by Gulf Publishing Co., Houston, Texas; for further
details, see Table 161).
-------
578
PETROLEUM EQUIPMENT
3. Diameter of vessel:
TOP OF VESSEL
Figure 404. Discharge piping for relief and
safety valves: (top) A cap like one illus-
trated protects discharge pipe from being
plugged with snow, (bottom) piping must be
adequately anchored to prevent sway or vi-
bration while the valve is discharging
(Driskell, 1960; copyrighted by Gulf Publish-
ing Co., Houston, Texas).
Problem:
Determine dimensions of knockout drum.
Solution:
1. Maximum allowable vapor velocity, u
0. 5
,0.5
rt ? 9 "7
max ' L 0. 1
u = 5. 06 ft/sec.
max
2. Design vapor velocity, u :
u = u x 0. 50
D max
u = 5.06 x 0.50 = 2. 5 ft/sec.
f(4)(100)-|
" 100(2.5) J
0.5
D = 7. 12 ft.
Use 7-ft diameter.
4. Height of vessel:
Assume low liquid loading.
Use vertical drum, L/D = 3.
Height = 3 x 7 ft = 21 ft.
Alternative solution:
The same problem can be solved graphically as
follows:
. PI/PV
50
0. 1
= 500
2. Enter Figure 406 at 100 cfs and proceed
vertically to
Pl/Pv
Proceed horizontally arid read drum di-
ameter as 7 feet.
3. Again assume L/D ratio = 3.
4. Therefore, drum dimensions are 7 ft in di-
ameter x 21 ft high.
Sizing a blowdown line
As previously stated, the selection of a par-
ticular line capacity depends upon the folio-wing
considerations: (1) Maximum expected vapor
flow, (2) maximum allowable backpressure in
the system, (3) type of relief device to be used,
and (4) governing code.
The maximum design capacity of a blowdown line
is generally based upon the operation of a group
of relief and safety valves. Selection of a de-
sign capacity is based upon upsets in the process
or by exterior fire. Table 159 indicates the re-
lief requirements for various conditions.
The maximum allowable backpressure in the re-
lieving system depends upon the vessel with the
lowest operating or working pressure, the type
of valve used, and the code used. In the past,
the pressure drop in the relief manifold was
customarily limited to 10 percent of the set
GPO 806—614—20
-------
Waste-Gas Disposal Systems
579
Table 161. SUPPLEMENTARY INFORMATION TO FIGURE 403
(Driskell, I960; copyrighted by Gulf Publishing Co. , Houston, Texas)
Service
Letters keyed to
caption for Figure 403
Valve indoors Valve outdoors
A,b B,b E
C
D
B
E
A, B
C
A, B, D
G
A,bBb
D
B
B
A, B
C
A, B, D
G
Nonhazardous service3-
Air or gas
Liquid
Steam or vapor
Discharge pipe size to 1 in.
Discharge pipe size to 1-1/2 to 2-1/2 in.
Discharge pipe size to 3 in. and over
Hazardous service3-
Closed system (to vent stack, burning
stack, or scrubber)
Open system (to atmosphere)
Gasc
Liquidd
Vaporc> d
Pyrophoric gases or vaporc
Low-temperature service
At or below ambient--design discharge pipe so that snow or ice accumulate
at any point in the line where the temperature may be at or below freezing.
Use A, if possible. Where necessary, B may be used with a cover.
Below 32°F--locate safety valve to avoid need for discharge piping, if
possible. Discharge opening and exposed spring must be protected from
the weather. A housing or local heating may be required. The discharge,
if properly designed, may be sealed with a low-viscosity oil and covered
with plastic to prevent the entrance of moisture.
aFlammable or toxic fluids are considered hazardous.
bDischarge pipe not required if outlet over 7 feet above walkway, or directed
away from personnel, or both.
°Carry discharge outdoors to a safe elevation.
Carry to an appropriate drain.
pressure. As previously stated, however, the
development of balanced relief and safety valves
has removed this restriction. In the usual re-
finery application, there can be considerable
savings in piping and valves with balanced valves
and about a 40 percent backpressure.
Where several valves discharge to a common
header, the use of two separate relieving sys-
tems--high- and low-pressure--may be econom-
ically advantageous. Otherwise, a single mani-
fold design will be limited by the lowest pres-
sured vessel.
A reduction in the size of the manifold line may
be achieved if the operating pressure of a vessel
is less than the maximum working, or design,
pressure. The set pressure of the relief or safety
valve can be made less than the design pressure,
permitting a greater backpressure in the relief
line.
Another method that can be used with standard
safety valves is to plug the guide and vent the
bonnet, as shown in Figure 394. An increase in
backpressure lowers the relieving pressure and
yet does not overpressure the vessel. The ar-
rangement can, however, upset the process if
the valve setting is too close to the operating
pressure. Thus, in a manifold system, an up-
set in one section of a process could cause ad-
ditional relief or safety valves to vent.
In determining the size of a vapor relief line,
the pressure drop is usually large, and this pre-
cludes the direct use of a Fanning equation. In
calculations of compressible fluid flow, the follow-
ing criteria are used (Crane Company, 1957):
1. If the pressure drop is less than 10 percent
of the inlet pressure, reasonable accuracy
is obtained if the density of the gas is based
upon either inlet or outlet conditions.
-------
580
PETROLEUM EQUIPMENT
Site
1B-gage SHEET STACK
A.
STACK HEIGHT OVER
iOO Ib INCLUDING
FLANGE
3 JACK SCREWS
SPACED FOR
REMOVAL OF
DISC ASSEMBLY
INSERTION TYPE
ASSEMBLY
STACK
INDEPENDENTLY
SUPPORTED
~ .HOOK OR
i/'DRAWBDLT
C.
100-Ib MAX. WEIGHT
INCLUDING FLANGE
100-lb MAX.-
VENT PIPE
ft
STACK
-INDEPENDENTLY
SUPPORTED
-DRAW BOLTS
f
E.
SUPPORT
INDEPENDENT OF
3ISC ASSEMBLY
\
ALLOW CLEARANCE
FOR EASY REMOVAL
--CONSIDER CROWN
OF DISC
Figure 405. Discharge piping for rupture discs: (A) For lightweight assembly, (B) for heavy assembly
with short stack, (C) for heavy assembly with long stack, (D) double disc with lightweight assembly,
(E) double disc with heavy assembly, (F) closed system (Oriskell, I960; copyrighted by Gulf Publish-
ing Co., Houston Texas; for further details, see Table 162).
Table 162. SUPPLEMENTARY INFORMATION
TO FIGURE 405 (Driskell, I960;
copyrighted by Gulf Publishing Co. ,
Houston, Texas)
Service
Discharge to atmosphere
Outdoors, lightweight assemblya
Outdoors, heavy assembly
Indoorsc
Closed system
Letters keyed to
caption for Figure 405
Single disc
A
B, C
C
F
Double disc
D
E
E
F
aParts of assembly 100 Ib or less for ease of handling.
Parts of assembly exceed 100 Ib and require mechanical
lifting.
cVent stack through roof.
2. If the pressure drop is greater than 10 per-
cent but less than about. 40 percent of inlet
pressure, the Fanning equation maybe used
with reasonable accuracy if an average den-
sity is used. Otherwise a method with a
kinetic energy correction can be used.
3. For greater pressure drops, empirical equa-
tions can be used.
API Manual RP520 presents kinetic-energy cor
rection factors, as shown in Figure 407, that
may be applied to the Fanning equation.
Another method generally used involves dividin
the line into increments having pressure drops
-------
Waste-Gas Disposal Systems
581
100
10
1.0
O.I
10 100
DRUM VAPOR CAPACITY (AT FUMING TEMPERATURE AND PRESSURE), cfs
1,000
Figure 406. Knockout drum-sizing chart (Kerns, 1960; copyrighted
by Gulf Publishing Co., Houston, Texas)
10 percent or less and working from the line
terminus back to the relief device.
With the greater availability of computers more
exact methods of calculation can be used. Machine
computers can handle the tedious equations for
calculating pressure drop of compressible fluids
where the velocity is subsonic and the density of
the vapor or gas is constantly changing.
For hand calculations, a simplified method has
been proposed (Conison, I960) that gives con-
servative results. The maximum carrying capac-
ity of any line is limited by the acoustic velocity
at the outlet of the pipe and in turn sets the out-
let pressure. The equation developed by Crocker
for solving the maximum pipe capacity for flow-
ing gas and vapors is as follows:
where
/_RT_
W_ Vk (k + 1)
11,400
(117)
^ = outlet pressure, psia
d = ID of pipe line, in.
R =
k =
W =
1, 544
mol wt of gas
ratio of specific heats, c /c
P ^
vapor or gas, Ib/hr
T = outlet temperature, °R,
Equation 117 is used to determine the pressure
at the pipe line outlet with W pounds of gas or
vapor flowing per hour. If the vapors are dis-
charged to the atmosphere, the outlet pressure
must be equal to or greater than atmospheric
pressure. If P2 calculated is less than 1 atmo-
sphere, then W can be increased before any ef-
-------
582
PETROLEUM EQUIPMENT
C = KINETIC-ENERGY CORRECTION FACTOR
= INTERNAL DIAMETER OF PIPE, in
= INLET PRESSURE psia
= PRESSURE OBOP IN LINE BASED ON
INLET CONDITIONS (P, UNO P,)
= CORRECTED PRESSURE DROP psi
= Ib hr GAS
= INLET DENSITY pd
0 (6 0 20 0.24 0 28 0 32
PRESSURE DROP (BASED OK P, AND P, I
INLET PRESSURE psia
Figure 407. Kinetic energy correction for pressure drop for
isothermal flow (American Petroleum Institute, 1960).
feet is made on backpressure in the line. If P£
calculated is equal to atmospheric pressure,
then any increase in W increases the discharge
pressure at the pipe outlet. If P2 calculated is
greater than atmosphere, then it must be added
to the line friction loss calculated from the re-
lief device to the pipe outlet in order to determine
the total backpressure at the relief device.
To simplify the calculation of the line pres-
sure drop, the following equation can be used
when the line lengths are approximately 100 feet
or more or velocity change is small:
gD 144
(118)
where
P = inlet pressure, psig
-------
Waste-Gas Disposal Systems
583
vt
f =
1 =
D =
outlet pressure, psig (equal to values
in equation 117 when Pj = atmospheric
pressure or greater)
vapor density, Ib/ft , at line terminus
a friction factor
line length, ft
31.2 ft/sec2
line ID, ft
V- = velocity at line outlet, fps.
Inspection of equation 118 reveals that the quantity
is the Fanning equation for determin-
ing pressure drop in a line in pounds per square
foot. This quantity is readily determined with the
aid of conventional charts in handbooks and other
publications.
All gas or vapor terms in the final or line outlet
conditions are based on the inlet temperature T,
calculated from equation 115 and f^ from equa-
tion 118. Where the line lengths are less than
100 feet, equation 118 is modified as follows:
Pl =
(119)
where: (J.. = inlet velocity, fps.
Equation 119 can be rearranged to facilitate trial
and error solutions:
P2 = 210 + f20.
, 017)(600)
4+2 log
+ 2 log
460 \ (460) (14. 5 (.077)
^i
50.8
/ (32.2)(144)
1/2
1/2
-------
584
PETROLEUM EQUIPMENT
As a first approximation, ignore loge
460
1/2
Then:
P = [210 + (20. 4)(50. 8)]
= [1,249]1/2
= 35. 3 psia.
3. Correct P for change in velocity:
W v
Velocity =
35.3
1 (25, 000)(5. 35)
A (3,600)(0. 2006)
u. = 185 fps
i
101. 6 log 2.49 = (101. 6)(. 912) = 92.6
And applying the correction for the log term:
PL = [1,249 + 92.6]1/2 = [1,342]1/2
P = 36. 7 psia.
Smoke From Flares
Smoke is the result of incomplete combustion.
Smokeless combustion can be achieved by:
(1) Adequate heat values to obtain the minimum
theoretical combustion temperatures, (2) ade-
quate combustion air, and (3) adequate mixing
of the air and fuel.
An insufficient supply of a:.r results in a smoky
flame. Combustion begins around the periphery
of the gas stream where the air and fuel mix,
and within this flame envelope the supply of air
is limited. Hydrocarbon side reactions occur
with the production of smo'ke. In this reducing
atmosphere, hydrocarbons crack to elemental
hydrogen and carbon, or polymerize to form
hydrocarbons. Since the carbon particles are
difficult to burn, large volumes of carbon parti-
cles appear as smoke upon cooling. Side reac-
tions become more pronounced as molecular
weight and unsaturation of the fuel gas increase.
Olefins, diolefins, and aromatics characteristic-
ally burn with smoky, sooty flames as compared
with paraffins and naphthenes (Rupp, 1956).
A smokeless flame can be obtained when an ade-
quate amount of combustion air is mixed suffi-
ciently with the fuel so that it burns completely
and rapidly before any side reactions can take
place.
Other Air Contaminants From Flares
Combustion of hydrocarbons in the steam-in-
spirated-type elevated flare appears to be com-
plete. The results of a field test (Sussman et al.
1958) on a flare unit such as this were reported
in the form of ratios as follows:
THE AIR POLLUTION PROBLEM
The air pollution problem associated with the un-
controlled disposal of waste gases is the venting
of large volumes of hydrocarbons and other odor-
ous gases and aerosols. The preferred control
method for excess gases and vapors is to re-
cover them in a blowdown recovery system and,
failing that, to incinerate them in an elevated-
type flare. Such flares introduce the possibility
of smoke and other objectionable gases such as
carbon monoxide, sulfur dioxide, and nitrogen
oxides. Flares have been further developed to
ensure that this combustion is smokeless and
in some cases nonluminous. Luminosity, while
not an air pollution problem, does attract atten-
tion to the refinery operation and in certain cases
can cause bad public relations. There is also the
consideration of military security in which non-
luminous emergency gas flares would be desirable.
CO : hydrocarbons 2,100:1
CO : CO 243:1
These results indicate that the hydrocarbon and
carbon monoxide emissions from a flare can be
much greater than those from a properly oper-
ated refinery boiler or furnace. Calculations
based on these data, with the assumption of a
gas with two carbon atoms and a molecular weight
of 30, indicate that the flares in Los Angeles County
cause an average daily emission of approximately
100 pounds of hydrocarbons per day and 840 pounds
of carbon monoxide per day.
Other combustion contaminants from a flare in-
clude nitrogen oxides. The importance of these
compounds to the total air pollution problem de-
pends upon the particular conditions in a partic-
ular locality. The total emission of nitrogen oxides
from the approximately 40 flares in Los Angeles
County has been estimated (Chass and George,
I960) at 110 pounds per day.
-------
Waste-Gas Disposal Systems
585
Other air contaminants that can be emitted from
flares depend upon the composition of the gases
burned. The most commonly detected emission
is sulfur dioxide, resulting from the combustion
of various sulfur compounds (usually hydrogen
sulfide) in the flared gas, Toxicity, combined
with low odor threshold, make venting of hydro-
gen sulfide to a flare an unsuitable and some-
times dangerous method of disposal. In addition,
burning relatively small amounts of hydrogen sul-
fide can create enough sulfur dioxide to cause
crop damage or local nuisance.
Materials that tend to cause health hazards or
nuisances should not be disposed of in flares.
Compounds such as mercaptans or chlorinated
hydrocarbons require special combustion devices
with chemical treatment of the gas or its prod-
ucts of combustion.
AIR POLLUTION CONTROL EQUIPMENT
The ideal refinery flare, according to the Amer-
ican Petroleum Institute, is a simple device for
safe and inconspicuous disposal of waste gases by
combustion. From an air pollution viewpoint, the
ideal flare is a combustion device that burns waste
gases completely and smokelessly.
Types of Flares
There are, in general, three types of flares for
the disposal of waste gases: Elevated flares,
ground-level flares, and burning pits.
The burning pits are reserved for extremely
large gas flows caused by catastrophic emergen-
cies in which the capacity of the primary smoke-
less flares is exceeded. Ordinarily, the main
gas header to the flare system has a water seal
bypass to a burning pit. Excessive pressure in
the header blows the water seal and permits the
vapors and gases to vent a burning pit where
combustion occurs.
The essential parts of a flare are the burner,
stack, seal, liquid trap, controls, pilot burner,
and ignition system. In some cases, vented gas-
es flow through chemical solutions to receive
treatment before combustion. As an example,
gases vented from an isomerization unit that may
contain small amounts of hydrochloric acid are
scrubbed with caustic before being vented to the
flare.
Elevated flares
Smokeless combustion can be obtained in an ele-
vated flare by the injection of an inert gas to the
combustion zone to provide turbulence and inspi-
rateair. A mechanical air-mixing system would
be ideal but is not economical in view of the large
volume of gases handled. The most commonly
encountered air-inspirating material for an ele-
vated flare is steam. Three main types of steam-
injected elevated flares are in use. These types
vary in the manner in which the steam is injected
into the combustion zone.
In the first type, there is a commercially avail-
able multiple nozzle, as shown in Figure 408,
which consists of an alloy steel tip mounted on
the top of an elevated stack (Brumbaugh, 1947;
Hannaman and Etingen, 1956). Steam injection
is accomplished by several small jets placed
concentrically around the flare tip. These jets
PATENTED
Figure 408. View of John Zink smokeless
flare burner (John Zink Co., Tulsa, Okla.)
-------
586
PETROLEUM EQUIPMENT
are installed at an angle, causing the steam to
discharge in a converging pattern immediately
above the flare tip.
A second type of elevated flare has a flare tip
with no obstruction to flow, that is, the flare tip
is the same diameter as the stack. The steam
is iniected by a single nozzle located concen-
trically within the burner tip. In this type of
flare, the steam is premixed with the gas before
ignition and discharge.
A third type of elevated flare has been used by
the Sinclair Oil Company (Decker, 1950). It
is equipped with a flare tip constructed to cause
the gases to flow through several tangential open-
ings to promote turbulence. A steam ring at the
top of the stack has numerous equally spaced
holes about .1/8 inch in diameter for discharging
steam into the gas stream.
The injection of steam in this latter flare may be
automatically or manually controlled. All the
flares of this type located in Los Angeles County
are instrumented to the extent that steam, is auto-
matically supplied when there is a measurable
gas flow. In most cases, the steam is propor-
tioned automatically to the rate of gas flow; how-
ever, in some installations, the steam is auto-
matically supplied at maximum rates, and manual
throttling of a steam valve is required for adjust-
ing the steam flow to the particular gas flow rate.
There are many variations of instrumentation
among various flares, some designs being more
desirable than others. For economic reasons,
all designs attempt to proportion steam flow to
the gas flow rate.
Steam injection is generally believed to result
in the following benefits: (1) Energy available
at relatively low cost can be used to inspirate
air and provide turbulence within the flame,
(Z) steam reacts with the fuel to form oxygen-
ated compounds that burn readily at relatively
low temperatures, (3) water-gas reactions
also occur with this same end result, and (4)
steam reduces the partial pressure of the fuel
and retards polymerization. (Inert gases such
as nitrogen have also been found effective for
this purpose; however, the expense of providing
a diluent such as this is prohibitive. )
The effectiveness of steam injection in an ele-
vated flare is graphically illustrated by compar-
ing Figures 409 and 410.
Multisfeam-jet-fype elevated flare
A multisteam -jet-type elevated flare
(Cleveland, 1952) is shown in Figure 411.
All relief headers from process units com-
Figure 409. Refinery flare with steam
inject!on in operation.
Figure 410. Refinery flare with steam
injection not in operation.
-------
Waste-Gas Disposal Systems
587
STEAM
3-m STEAM RING
PILOT
MAIN COLLECTION SYSTEM
HYDROGEN REACTOR
DROPOUT
PETROCHEMICAL
SYSTEM
DRIP
TANK
CONOENSATE
•*•
BLINDS
BY-PASS
(5-in WATER \
SEAL TANK 1
20-i n x
40- ft
MAIN
FLARE
DRAIN
3-in NOZZLE
STEAM
PILOT
-o-
CATALYTIC CRACKING COMPRESSORS
14-in x
15-ft
AUXILIARY
FLARE
DRAIN
Figure 411. Waste-^as flare system usmj multiple-steam-jet
burner (Cleveland, 1952).
bine into a common header that conducts the
hydrocarbon gases and vapors to a large
knockout drum. Any entrained liquid is
dropped out and pumped to storage. The
gases then flow in one of two ways. For
emergency gas releases that are smaller
than or equal to the design rate, the flow is
directed to the main flare stack. Hydro-
carbons are ignited by continuous pilot
burners, and steam is injected by means
of small jet fingers placed concentrically
about the stack tip. The steam is injected
in proportion to the gas flow. The steam
control system consists of a pressure con-
troller, having a range of 0 to 20 inches
water column, that senses the pressure in
the vent line and sends an air signal to a
valve operator mounted on a 2-inch V-Port
control valve in the steam line. If the emer-
gency gas flow exceeds the designed capacity
of the main flare, backpressure in the vent
line increases, displacing the water seal
and permitting gas flow to the auxiliary
flare. Steam consumption of the burner at
a peak flow is about 0. 2 to 0. 5 pound of
steam per pound of gas, depending upon the
amount and composition of hydrocarbon gas-
es being vented. In general, the amount of
steam required increases with increases in
molecular weight and the degree of unsatura-
tion of the gas.
A small amount of steam (300 to 400 pounds
per hour) is allowed to flow through the jet
fingers at all times. This steam not only
permits smokeless combustion of gas flows
too small to actuate the steam control valves
but also keeps the jet fingers cooled and
open.
-------
588
PETROLEUM EQUIPMENT
Esso-fype elevated flare
A second type of elevated, smokeless,
steam-injected flare is the Esso type. The
design is based upon the original installation
in the Bayway Refinery of the Standard Oil
Company of New Jersey (Smolen, 1951 and
1952). A typical flare system serving a
petrochemical plant using this type burner
is shown in Figure 412. The type of hydro-
carbon gases vented can range from a sat-
urated to a completely unsaturated material.
The injection of steam is not only propor-
tioned by the pressure in the blowdown lines
but is also regulated according to the type
of material being flared. This is accom-
plished by the use of a ratio relay that is
manually controlled. The relay is located
in a central control room where the operator
has an unobstructed view of the flare tip.
In normal operation the relay is set to han-
dle feed gas, which is most common to this
installation.
In this installation, a blowdown header con-
ducts the gases to a water seal drum, as
shown in Figure 413. The end of the blow-
down line is equipped with two slotted ori-
fices. The flow transmitter senses the
pressure differential across the seal drum
and transmits an air signal to the ratio re-
lay. The signal to this relay is either ampli-
fied or attenuated, depending upon its setting.
An air signal is then transmitted to a flow
controller that operates two parallel steam
valves. The 1-inch steam valve begins to
open at an air pressure of 3 psig and is fully
open at 5 psig. The 3-inch valve starts to
open at 5 psig and is fully open at 15 psig
air pressure. As the gas flow increases,
the water level in the pipe becomes lower
than the water level in the drum, and more
of the slot is uncovered. Thus, the difference
in pressure between the line and the seal drun
increases. This information is transmitted
as an air signal to actuate the steam valves.
The slotted orifice senses flows that are too
small to be indicated by a Pitot-tube-type
flow meter. The water level is maintained
1-1/2 inches above the top of the orifice to
take care of sudden surges of gas to the systen
A 3-inch steam nozzle is so positioned with-
in the stack that the expansion of the steam
just fills the stack and mixes with the gas to
provide smokeless combustion. This type of
flare is probably less efficient in the use of
steam than some of the commercially avail-
able flares but is desirable from the stand-
points of simpler construction and lower
maintenance costs.
X™3 IGNITORS ,
/ (TfPICAL 3 PLACES) ^J
•3 2-1 n PILOT BURNERS
120° APART)
SHALL FLOW
FLO, PURGE GAS
jf III
^SLOTTED III
ORIFICE III
Figure 412. Waste-gas flare system using Esso-type burner.
-------
Waste-Gas Disposal Systems
589
l-in. MOTOR VALVE
KNOCK OUT VESSEL
Figure 413., Water seal drum with slotted orifice for measuring
gas flow to flare.
Sinclair-type elevated flare
A. diagram (Decker, 1950) of an installation
using a Sinclair-type elevated flare is shown
in Figure 414. A detail of the burner used
for this flare is shown in Figure 415,
The flow of steam from the ring inspirates air
into the combustion area, and the shroud pro-
tects the burner from wind currents and pro-
vides a partial mixing chamber for the air and
gas. Steam is automatically supplied when
there is gas flow. A pressure-sensing ele-
ment actuates a control valve in the steam
supply line. A small bypass valve permits a
small, continuous flow of steam to the ring,
keeping the steam ring holes open and per-
mitting smokeless burning of small gas flows.
Ground level flares
There are four principal types of ground level
flare: Horizontal venturi, water injection, multi-
jet, and vertical venturi.
Horizontal, venturi-type ground flare
A typical horizontal, venturi-type ground
flare system is shown in Figure 416. In this
system, the refinery flare header discharges
to a knockout drum -where any entrained liq-
uid is separated and pumped to storage. The
gas flows to the burner header, which is con-
nected to three separate banks of standard
gas burners through automatic valves of the
snap-action type that open at predetermined
pressures. If any or all of the pressure
valves fail, a bypass line with a liquid seal
is provided (with no valves in the circuit),
which discharges to the largest bank of burn-
ers.
The automatic-valve operation schedule is
determined by the quantity of gas most likely
to be relieved to the system. The allowable
back-pressure in the refinery flare header
determines the minimum pressure for the con-
trol valve on the No. 1 burner bank. On the
assumption that the first valve was set at 3
psig, then the second valve for the No. 2 burn-
-------
590
PETROLEUM EQUIPMENT
Figure 414. Diagram of waste-gas flare system
using a Sinclai r burner.
er bank would be set for some higher pres-
sure, say 5 psig. The quantity of gas most
likely to be released then determines the size
and the number of burners for this section.
Again, the third most likely quantity of gas
determines the pressure setting and the size
of the third control valve. Together, the burn-
er capacity should equal the maximum expected
flow rate.
The valve-operating schedule for the system
pictured in Figure 416 is set up as follows:
1. When the relief header pressure reaches
3 psig, the first control valve opens and
the four small venturi burners go into
operation. The controller setting keeps
the valve open until the pressure decreases
to about 1-1/2 psig.
2. When the header pressure reaches 5 psig,
the second valve opens and remains open
until the pressure drops to about 3 psig.
3. When the pressure reaches 6 psig, the
third valve opens and remains open until
the pressure decreases to 4 psig.
4. At about 7 psig, the gas blows the liquid
seal.
A small flare unit of this design, with a capac-
ity of 2 million scf per day, reportedly cost
FUSTIC
INSULATION
SECTION «-A
ESS STANDPIPE
PROTECTING SHROUD
STEAK SUPPLY PIPES
FLAME ARRESTER
Figure 415. Detail of Sinclair flare burner, plan
and elevation (Decker, 1950).
less than $5, 000. 00 in 1953 (Beychok, 1953).
Another large, horizontal, venturi-type flare
that has a capacity of 14 million scfh and re-
quires specially constructed venturi burners
(throat diameter ranges f:rom 5 to 18 inches)
cost $63,000.
Wafer- injection-type ground flare
Another type of ground flare used in petroleum
refineries has a water spray to inspirate air
and provide water vapor for the smokeless
combustion of gases (Figure 417). This flare
requires an adequate supply of water and a
reasonable amount of open space.
The structure of the flare consists of three
concentric stacks. The combustion chamber
contains the burner, the pilot burner, the end
of the ignitor tube, and the water spray dis -
-------
Waste-Gas Disposal Systems
591
BURNER BANKS
STEEL, CEMENT.OR
REFRACTORY WUL
Figure 415. Typical venturi ground flare. The ignitors for pilot
burners and the warning element for pilot operation are not shown
(American Petroleum Institute, 1957).
BOTTLED GAS
VENTURI BURNER
GAS TO PILOT
IGNITOR TUBE
OIL TO PILOT \
\
SPARK IGNITOR
FLAME ARRESTER
WATER SUPPLY
WATER STRAINERS
Figure 417. Typical water-spray-type ground flare.
Six water sprays are shown. Two pilots and two
ignitors are recommended (American Petroleum
Institute, 1957).
tributor ring. The primary purpose of the
intermediate stack is to confine the water spray
so that it will be mixed intimately with burn-
ing gases. The outer stack confines the flame
and directs it upward.
Water sprays in elevated flares are not too
practical for several reasons. Difficulty is
experienced in keeping the water spray in the
flame zone, and scale formed in the waterline
tends to plug the nozzles. In one case it was
-------
592
PETROLEUM EQUIPMENT
necessary to install a return system that per-
mitted continuous waterflow to bypass the
spray nozzle. Water main pressure dictates
the height to which water can be injected with-
out the use of a booster pump. For a 100- to
250-foot stack, a booster pump would undoubted-
ly be required. Rain created by the spray from
the flare stack is objectionable from the stand-
point of corrosion of nearby structures and
other equipment.
Water is not as effective as steam for control-
ling smoke with high gas flow rates, unsatu-
rated materials, or wet gases. The water
spray flare is economical when venting rates
are not too high and slight smoking can be
tolerated. In Los Angeles County, where re-
strictions on the emission of smoke from
flares are very strict, a. water spray smoke-
less flare is not acceptable.
1 jet-type ground flare
A recent type of flare developed by the refin-
ing industry is known as a multijet (Miller et
al. , 1956). This type of flare was designed
to burn excess hydrocarbons without smoke,
noise, or visible flame. It is claimed to
be less expensive than the steam-injected
type, on the assumption that new steam
facilities must be installed to serve a
steam-injected flare unit. Where the steam
can be diverted from noncritical operations
such as tank heating, the cost of the multijet
flare and the steam-inspirating elevated flare
may be similar.
A sketch of an installation of a multijet flare
is shown in Figure 418., The flare uses two
sets of burners; the smaller group handles
normal gas leakage and small gas releases,
while both burner groups are used at higher
flaring rates. This sequential operation is
controlled by two water-sealed drums set to
release at different pressures. In extreme
emergencies, the multijet burners are by-
passed by means of a water seal that directs
the gases to the center of the stack. This
seal blows at flaring rates higher than the
design capacity of the flare. At such an ex-
cessive rate, the combustion is both luminous
and smoky, but the unit is usually sized so
that an overcapacity flow would be a rare
occurrence. The overcapacity line may also
be designed to discharge through a water seal
to a nearby elevated flare rather than to the
center of a multijet stack. Similar staging
could be accomplished with automatic valves
or backpressure regulators; however, in this
FLARED GASES
FLOW BALANCING
, VALVE
SEAL DAM
SEAL NATER
VENT
FIRST-STAGE
SEAL DRUM
SEAL DAM
(J
SECQND-ST.AGE
DURNERS
SHELL
FIRST-STAGE BURNERS
i /r
OVER- !
CAPACITY
SEAL
SECOND-STAGE
SEAL DRUM
SEAL WATER
TO SEWER
Figure 418. Flow diagram of multijet-flare system (Miller et al., 1956).
-------
Waste-Gas Disposal Systems
593
case, the water seal drums are used because
of reliability and ease of maintenance. The
staging system is balanced by adjusting the
hand control butterfly valve leading to the
first-stage drum. After its initial setting,
this valve is locked into position.
Design details of this installation are given
in the literature reference (Miller et al. ,
1956).
Vertical, venturi-type ground flare
Another type of flare based upon the use of
commercial-type venturi burners is shown
in Figure 419. This type of flare has been
used to handle vapors from gas-blanketed
tanks, and vapors displaced from the depres-
suring of butane and propane tank trucks.
Since the commercial venturi burner requires
a certain minimum pressure to operate effi-
ciently, a gas blower must be provided. In
the installation shown in Figure 420, two
burners operate at a pressure of 1/2 to 8 psig.
A compressor takes vapors from tankage and
discharges them at a rate of 6, 000 cfh and 7
psig through a water seal tank and a flame
arrestor to the flare. This type of arrange-
ment can readily be modified to handle dif-
ferent volumes of vapors by the installation
of the necessary number of burners.
This type of flare is suitable for relatively
small flows of gas of a constant rate. Its
main application is in situations where other
means of disposing of gases and vapors are
not available.
Effect of steam injection
A flare installation that does not inspirate an ade-
quate amount of air or does not mix the air and
hydrocarbons properly emits dense, black clouds
of smoke that obscure the flame. The injection
of steam into the zone of combustion causes a
gradual decrease in the amount of smoke, and the
flame becomes more visible. When trailing smoke
has been eliminated, the flame is very luminous
and orange with a few wisps of black smoke around
the periphery. The minimum amount of steam re-
quired produces a yellowish-orange, luminous
flame with no smoke. Increasing the amount of
steam injection further decreases the luminosity
of the flame. As the steam rate increases, the
flame becomes colorless and finally invisible
during the day. At night this flame appears blue.
An injection of an excessive amount of steam
causes the flame to disappear completely and be
replaced with a steam plume. An excessive
amount of steam may extinguish the burning gases
STEEL SHELL
^ REFRACTORY
STA_
3 ft DIAMETER X ID ft HIGH
PILOT GAS
WASTE GAS
Figure 419. Vertical, venturi -type flare.
and permit unburned hydrocarbons to discharge to
the atmosphere. When the flame is out, there is
a change in the sound of the flare because a stean
hiss replaces the roar of combustion. The com-
mercially available pilot burners are usually not
extinguished by excessive amounts of steam, and
the flame reappears as the steam injection rate is
reduced. As the use of automatic instrumentation
becomes more prevalent in flare installations, the
use of excessive amounts of steam and the emis-
sion of unburned hydrocarbons decrease and great-
er steam economies can be achieved. In evaluat-
ing flare installations from an air pollution stand-
point, controlling the volume of steam is important.
Too little steam results in black smoke, which,
obviously, is objectionable. Conversely, ex-
cessive use of steam, produces a white steam
plume and an invisible emission of unburned
hydrocarbons. A condition such as this can also
be a serious air pollution problem.
Design of a smokeless flare
The choice of a flare is dictated by the particular
requirements of the installation. A flare may be
-------
594
PETROLEUM EQUIPMENT
STUCK
BURNERS
Figure 420. Flow diagram of tank-gas-blanketing system
venting to a vertical, venturi flare.
located either at ground level or on an elevated
structure. Ground flares are less expensive,
but locations must be based upon considerations
such as proximity of combustible materials,
tanks, and refinery processing equipment. In a
congested refinery area, there may be no choice
but to use an elevated flare,
A method of determining the distance a stack
should be from surrounding equipment and per-
sonnel has been developed (Hajek and Ludwig,
I960). The recommended equation is
D =
(-121)
where
D
F =
minimum distance, ft from the flame
to the object
a dimensionless constant equal to 0. 20
for methane, which has a hydrogen-to-
carbon weight ratio of 0. 333, and equal
to 0. 33 for propane, which has a hydro-
gen-to-carbon weight ratio of 0. 222 .
(Use 0. 40 when in doubt. )
K. = heat release, Btu/hr
?
K = a constant, Btu/hr-ft":
K = 1,000 for objects exposed 20
minutes or more
K = 1, 500 for objects exposed less
than 20 minutes.
The asual flare system includes gas collection
equipment, the liquid knockout tank preceding the
flare stack. A water seal tank is usually located
between the knockout pot and the flare stack to
prevent flashbacks into the system. Flame ar-
restors are sometimes used in place of or in con-
junction with a water seal pot. Pressure-tem-
perature-actuated check valves have been used
in small ground flares to prevent flash-back. The
-------
Waste-Gas Disposal Systems
595
flare stack should be continuously purged with
steam or refinery gas to prevent the formation of
a combustible mixture that could cause an explo-
sion in the stack. A purge steam flow of 10 cfm
is recommended for a commercial-size burner
section (Hajek and Ludwig, I960).
The preferred method of inspirating air is inject-
ing steam either into the stack or into the combus-
tion zone. Water has sometimes been used in
ground flares where there is an abundant supply.
There is, however, less assurance of complete
combustion when water is used, because the flare
is limited in its operation by the type and composi-
tion of gases it can handle efficiently.
The diameter of the flare stack depends upon the
expected emergency gas flow rate and the per-
missible backpressure in the vapor relief mani-
fold system. The stack diameter is usually the
same or greater than that of the vapor header
discharging to the stack and should be the same
diameter as or greater than that of the burner
section. The velocity of the gas in the stack
should be as high as possible to permit use of
lower stack heights, promote turbulent flow with
resultant improved combustion, and prevent
flashback. Stack gas velocity is limited to about
500 fps in order to pre/ent extinction of the
flame by blowout. A discharge velocity of 300
to 400 fps based upon pressure drop considera-
tions is the optimum design figure of a patented
flare tip manufactured by the John Zink Company.
The nature of the gas determines optimum dis-
charge velocity (John Zink Company).
Adequate stack heights must he provided to per-
mit safe dispersion of toxic or combustible mate-
rial in the event of pilot burner failure. Tech-
niques are available for calculating adequate stack
heights to obtain certain ground concentrations at
various distances from the stack, depending upon
atmospheric conditions (Bodurtha, 1958; Gosline
et al. , 1956). These methods of calculation
should not be generally applied to any one loca-
tion, and meteorological data should be obtained
for the particular location involved.
The structural support of an elevated-flare stack
over 40 to 50 feet high requires the use of guy
wires. A self-supporting stack over 50 feet high
requires a large and expensive foundation. Stacks
over 100 feet high are usually supported by a
steel structure such as is shown in Figure 421.
Three burner designs for elevated flares have
been discussed—the multisteam-jet, or Zink,
and the Esso and Sinclair types. The choice of
burner is a matter of personal preference. The
Zink burner provides more efficient use of steam,
which is important in a flare that is in constant
use. On the other hand, the simplicity, ease of
Figure 421. A 200-ft flare stack supported
by a steel tower (Atlantic-Richfleld Co.
Wilmington, Calif.).
maintenance, and large capacity of the Esso burn-
er might be important considerations in another
installation.
As previously mentioned, the amount of steam
required for smokeless combustion varies accord-
ing to the maximum expected gas flow, the molec-
ular weight, and the percent of unsaturated hydro-
carbons in the gas. Data for steam requirements
for elevated flares are shown in Figure 42^. Actu-
al tests should be run on the various materials to
be flared in order to determine a suitable steam-
to-hydrocarbon ratio. In the typical refinery,
the ratio of steam to hydrocarbon varies from
0. L. to 0. 5 pound of steam per pound of hydrocar-
bon. The John Zink Company's recommendation
for their burner is 5 to 6 pounds per 1, 000 cubic
feet of 3. 30-moieoular-werght gas at a pressure
drop of 0. 65 psig.
Pilot ignition system
The ignition of flare gases is normally accom-
plished with one of three pilot burners. A sepa-
-------
596
PETROLEUM EQUIPMENT
*****
40 50 60 70
UHSATURATES, S by weight
90
100
Figure 422. Steam requirements for smokeless burning of
unsaturated hydrocarbon vapor (American Petroleum Insti-
tute, 1957).
rate system must be provided for the ignition of
the pilot burner to safeguard against flame fail-
ure. In this system, an easily ignited flame with
stable combustion and low fuel usage must be pro-
vided. In addition, the system must be protected
from the weather.
One good arrangement for a pilot igniter is shown
in Figure 423. To obtain the proper fuel-air ratio
for ignition in this system, the two plug valves
are opened and adjustments are made with the
globe valves, or pressure regulator valves. After
the mixing, the fuel-air mixture is lit in an igni-
tion chamber by an automotive spark plug con-
trolled by a momentary-contact switch. The igni-
tion chamber is equipped with a heavy Pyrex glass
window through which both the spark and ignition
flame can be observed. The flame front travels
through the ignitor pipe to the top of the pilot
burner. The mixing of fuel gas and air in the
supply lines is prevented by the use of double
check valves in both the fuel and air line. The
collection of -water in the ignitor tube can be pre-
vented by the installation of an automatic drain
in the lower end of the tube at the base of the
flare. After the pilot burner has beer lit, the
flame front generator is turned off by closing
the plug cocks in the fuel and air lines. This
prevents the collection of condensate and the
overheating of the ignitor tube.
On elevated flares, the pilot flame is usually
not visible, and an alarm system to indicate
flame failure is desirable. This is usually ac-
complished by installing thermocouples in the
pilot burner flame. In the event of flame fail-
ure, the temperature drops to a preset level,
and an ^Llarm sounds.
Instrumentation and control of steam and gas
For adequate prevention of smoke emission
and possible violations of air pollution regula
tions, an elevated, smokeless flare should be
-------
Waste-Gas Disposal Systems
597
Figure 423. Remote-control system for igniting
flare pilot burners (American Petroleum Insti-
tute, 1957).
VHPOR I HUT
DRUM
equipped to provide steam automatically and in
proportion to the emergency gas flow.
Basically, the instrumentation required for a
flare is a flow-sensing element, such as a Pitot
tube, and a flow transmitter that sends a signal
(usually pneumatic) to a control valve in the
steam line. Although the Pitot tube has been
used extensively in flare systems, it is limited
by the minimum linear velocity required to pro-
duce a measurable velocity head. Thus, small
gas flows will not actuate the steam control
valves. This problem is usually overcome by
installing a small bypass valve to permit a
constant flow of steam to the flame burner.
A more sensitive type of flow-measuring device
is the inverted weir. A typical installation is
shown in Figure 424. A variation of the inverted
weir is the slotted orifice previously shown in
Figure 413. The operation of this installation
has already been described.
The hot-wire flow meter has also been used in
flare systems (Huebner, 1959). The sensing
element is basically a heat loss anemometer
consisting of an electrically heated wire ex-
posed to the gas stream to measure the velocity.
The gas flow is perpendicular to the axis of the
hot wire. A conventional recorder is used with
this probe, modified for the resistance bridge
circuit of the gas flow meter. As the flow of
i
/
^
FLO*
RECORDER
FL»RE
n
HIHTER LEVEL
if \
SEALING WATER INLET
TO SENER
Figure 424. Inverted weir for measuring gas flow to a flare.
The end of the low-pressure line to the How recorder should
be at the same level as the tops of the slots in the inverted
weir. The end of the high-pressure line to the recorder should
be at the same level as the bottoms of the weir slots (American
Petroleum Institute, 1957).
-------
598
PETROLEUM EQUIPMENT
gas past the probe varies, the heat loss irom
the hot wire varies and causes an imbalance of
the bridge circuit. The recorder then adjusts
for the imbalance in the bridge and indicates the
gas flow. This type of installation provides sen-
sitivity at low velocities, and the gas flow mea-
surement can be made without causing an appre-
ciable pressure drop. This is an important ad-
vantage in 3. system using constant backpressure-
type relief valves. One flow meter of this type
in use has a velocity range of 0 to 6, 000 fpm.
The hot-wire flow meter can be used as a primary
flow-sensing element or as a leak detector in
laterals connected to the main flare header.
Another system using a venturi tube as the
primary element for measuring the rate of gas
flow to a flare is shown in Figure 425.
Supply and control of steam
After the amount of steam required for maxi-
mum design gas flow rates is determined, the
size of the steam supply line can be estimated
by conventional methods of calculation, such
as shown in Figure 426. The following example
illustrates the calculations for sizing the steam
supply line.
FLOW RECORDER
IN BYPASS LINE
Figure 425. System for measuring flare gas. Small flows of gas
are measured by the flow recorder in the bypass line. When a
blow occurs that is large enough to overcome the static head of
the sealing liquid in the seal pot, the liquid is blown to the slop
and blowdown drum. The gas flow is measured by the venturi in the
main line to the flare (American Petroleum Institute, 1957).
-------
Waste-Gas Disposal Systems
599
index
200 300 400 500 600 700 800 900 1000 1100 1200'
t - Temperature in Degrees Fahrenheit
=• 30-
oi 20 —
— 1000
— 800
— 600
— 500
— 400
—200
— 150
2.5
~ 6
7-
~2f-
Schedule Numbei
Figure 426. Steam pipe sizing chart. Establish the steam pressure and temperature
intersection. Draw a horizontal line to specific volume scale V Draw a line
rom V to the expected rate of flow, V. Mark the intersect with the index line
Using either known quantity, pipe size, d; or velocity, V; find the unknown by
drawing a line from the index to the known quantity (Crane Company 1957)
Example 37
Given:
200 psig (215 psia) saturated steam
9,000 Ib/hr propane
1, 000 Ib/hr propylene
10% (by weight) unsaturated material.
Problem:
Determine the size of the steam supply line re-
quired.
Solution:
From Figure 422, the steam-to-hydrocarbon
ratio should be 0. 55.
Steam required = (10, 000 lb/hr)(0. 55) = 5, 500
Ib/hr
With allowance for a future increase in steam re-
quirements, the steam line should be designed
to provide 7, 000 Ib/hr at a velocity of 6, 000 fpm.
From Figure 426, the pipe diameter is found
to be 3 inches .
The number and size of steam jets can be esti-
mated by the following empirical equation (Marks,
1951) for steam flow through a small nozzle:
0. 0165 AP
0.97
1
(122)
where
W = steam flow, Ib/sec
A = nozzle area, in.
-------
600
PETROLEUM EQUIPMENT
P = upstream pressure, psia.
Commercial burners use 1/8- to 1/2-inch-di-
ameter stainless steel pipe for the steam jets
with orifices of 1/8 to 7/16 inch in diameter.
The number of jets depends upon the gas flow
rates and the steam to be delivered into the com-
bustion zone.
Figure 427 is a plot of steam flow versus up-
stream pressure for various sizes of jet orifices,
This chart may be used for preliminary design
or for checking an existing installation as shown
in Example 38.
Example 38
Given:
Steam flow, 5, 500 Ib/hr
Available pressure upstream of jets, 80 psia
Assume jet orifice diameter, 3/8 in.
Problem:
Determine minimum number of steam jets re-
quired.
Solution:
From Figure 427, the steam, flow per jet =
460 Ib/hr
Number of jets required = './„,-, ,-, = 11.97
460 Ib/hr
Use 12 steam jets -with 3/8-inch orifices.
As shown in Figure 428, a jet located at an acute
angle to the direction of a gas flow improves the
mixing of the gas with air or steam. Commer-
cial flare burners usually have steam jets placed
at angles of 15 to 60 degrees with the gas flow.
A steam control system is provided to ensure
correct proportions of gas and steam flow. A
control valve with equal percentage characteris-
tics is often used in this application. A diagram
of this type of valve is shown in Figure 429. Flow
curves for valves with various characteristics
are shown in Figure 430. The manufacturer's
literature should be consulted for specific valves.
Accurate selection of the size of steam, control
valve requires a full knowledge of the actual
flowing conditions. Inmost cases, the pressure
1,400
1,200
1,000
200
40
60
80 100 120
JET UPSTREAM PRESSURE, psia
Figure 427. Jet upstream pressure versus jet capacity (based on equation
* -- 0.&165 AP where P < 0.575 Patm)-
-------
Waste-Gas Disposal Systems
601
0 6
30 45 60
IET ANGLE, degrees
Figure 428. Relationship between flame length
and jet angle (Gumz, 1950).
ACTUATOR
INLET
LOWER GUIDE
BRUSHING
BLIND HEAD
Figure 429. Diagram of double-seated, V-port
control valve and valve power unit (Holzbock
1959).
100
4 5 6 7 B 9 10 20
FLO* THRU VALVE, * of maximum
30 40 SO 60 80 100
Figure 430. Flow curves for control valves with characterized
plugs (Lieblich, 1953).
-------
602
PETROLEUM EQUIPMENT
across the valve must be estimated. A con-
servative working rule is that one-third of the
total system pressure drop, including all fittings
and equipment, should be absorbed by the con-
trol valve. The pressure drop across valves in
long lines or high-pressure drop lines may be
somewhat lower. In these cases the pressure
drop should be at least 15 to 25 percent of the
total system drop, provided the variations in
flow rates are small. A control valve can reg-
ulate flow only by absorbing energy and giving
a pressure drop to the system.
The most convenient method of sizing control
valves involves the use of the flow coefficient,
Cv. This flow coefficient is essentially a capac-
ity index of the valve and can be obtained from
the manufacturer's literature.
Pj, upstream pressure, J60 psig
P2i downstream pressure, 100 psig
Steam density, downstream, 0. 226 Ib/ft .
Problem:
Select a control valve for this system.
Solution:
Determine GV from the formula as shown in
'Table 163:
C
W
2. 1 [(P1 - P2)(P1 - P.,)]
1/2
(123)
By using the basic conversion formulas shown
in Table 163, the flow coefficient for any re-
striction can be determined. Under special
conditions, such as a high pressure drop or use
of special designs, deviation from the simple
fundamental law can be substantial. For most
practical valve-sizing problems, the use of the
simplified formulas is adequate.
A brief method of selection of a control valve
is explained in the following example.
Example 39
Given:
Gas flow, 10,0001b/hr
Steam-to-hydrocarbon ratio, 0.55 (by wtl
Maximum steam flow, 5, 500 Ib/hr
5, 500
2.1 [(174.7 - 114. 7)(174. 7 + 114.7)]
1/2
= 19. 8
A valve with a Cv of 19. 8 is indicated. Since
an equal percentage characteristic is desired
in this application, a correction factor should
be applied. This adjustment is necessary be-
cause' of the flow characteristics of this type of
valve. It is suggested by the manufacturer that
a 20 percent increase in the C be taken to corn-
Table 163. EQUATIONS FOR CONTROL VALVE SIZING
(Mason-Neilan Division, 1963)
NOMENCLATURE POR Cv FORMULAS
V — flow in U.S gpm
Q = cfh @ 14 7 psia and 60°F
W = Ib/hr
Pi r= inlet pressure — psia (14.7 -+- psi guage)
Py = outlet pressure — psia (14.7 -j- psi gauge)
G = gas s.p gr. (air = 1 0)
G( — sp. gr. @ flowing temperature
T, = flowing temperature — °F abs (460 + °F)
T.I, = superheat in "F
When P2 is less than the expression
i-1- Pj) becomes 0.8
(2)
2) On gas flow the effect of (lowing temperatures may be neglected for all temperatures between 30°F and 150°F.
For higher or lower temperatures a correction should be included.
-------
Waste-Gas Disposal Systems
603
pensate for this adjustment. Thus the Cy for
the equal percentage valve would be 23. 8.
Other items to consider in the selection of con-
trol valves are the valve actuator, valve posi-
tioners, and future steam requirements. The
control valve actuator supplies the power for
operating the valve. In flare applications the
power unit is usually a pneumatic-spring-dia-
phragm-type actuator of the type shown in Fig-
ure 423, operated by 3 to 15 psig air pressure.
These units are designed to open the valve if
the air pressure fails. Selection of valve actua-
tors can be made by referring to manufacturer's
literature.
Control valve positioners are devices that posi-
tion the valve stem to match the controller's
output signal. The valve plug is thus moved to
the desired position, overcoming the packing
friction and unbalanced forces in the system.
Positioners are also used when split-range
valves are operated by one controller. In most
elevated-flare installations, the range of gas
flow is such that one steam valve does not prop-
erly proportion the steam flow, and two parallel
steam valves are required. This means that
one valve moves from closed to open when the
controller output changes from 3 to 9 psig, and
the other valve is operated when the controller
output is 9 to 15 psig.
Where future steam requirements may be expected
to increase, the steam supply line and control
valves can be sized for the expected larger capac-
ity. Smaller inner-valve plugs can then be used
in the control valves to supply the necessary re-
duced capacity while the larger body size is re-
tained. The smaller plugs have the same flow
characteristics as the standard size plugs, but
flow can be reduced as low as 40 percent of nom-
inal capacity.
Design of water-injection-type ground flares
Designing a typical water injection flare consists
of sizing the stack structure and combustion air-
ports, and determining the water required to ob-
tain smokeless combustion (American Petroleum
Institute, 1957).
With the three-concentric-stack-type flare as
previously discussed, the air ports in the inter-
mediate and outer stacks should be designed to
admit 150 percent of the theoretical air for com-
plete combustion. The draft for these stacks
can be calculated from the equation
h = H
P P
a v
where
h = draft, in. of water
H = height of stack, ft
p = density of cold air, Ib/ft
cL
p = density of hot gas in stack, lb/ft~
p = density of water, Ib/ft .
w
The area of the stack's arches can be calculated
according to the formula
A =
W
457 CY (h )(p )
w a
(125)
•where
A = area of orifice, ft
W = rate of flow, Ib/hr
C = orifice discharge coefficient, dimen-
sionless
Y = expansion factor, dimensionles s
h = differential pressure across orifice,
in. of -water at 60 °F
p = density of air at upstream tap condi-
a tions, Ib/ft3.
In this case pa is the density of air at 60°F, Y
is assumed to be 1. 0, and C is assigned the value
of 0. 6. Equation 125 can now be reduced to
A =
W
20.9 h
(126)
(124)
Test data indicate that water pressure is more
important in achieving smokeless burning than
the amount of water delivered to the flare. In
general, a high water pressure results in better
mixing of gas. Higher water pressure is re-
quired as the molecular weight and unsaturated
content of the gas increase. Table 164 lists
water spray pressures required for smokeless
burning.
Satisfactory proportioning of the flow of water
to the flow of gas is difficult to achieve because
the pressure drop required for proper spray
nozzle operation is high. Where the opacity of
smoke emission is limited, some type of re-
mote manual or automatic control is necessary.
-------
604
PETROLEUM EQUIPMENT
Table 164. WATER SPRAY PRESSURE'S
REQUIRED FOR SMOKELESS BURNINGa
('American Petroleum Institute, 1957)
Table 165. DESIGN DATA FOR A FLARE SYSTEM
USING SPECIAL VENTURI BURNERS
(Brumbaugh, 1947)
—
scfh
300,000
150, 000
_, !
% by vol
0 to 20
30
125,000 j 40
weight
psig
ZS | 30 to 40
33 SO
37
120
' 1
1 :[pm
31 to 35
45
- I
The data in this table were obtained with a 1 - 1/2 -inch-diameter
spray nozzle in a ground flare with the following dimensions:
Height, ft Diameter, ft
Outer stack
Intermediate- stac
Inner stack
3-0
12
4
14
6
Design of venturi-type ground flares
The venturi-type ground flare, as previously
discussed, consists of burners, pilots, ignitors,
and control valves.
The total pressure drop permitted in a given in-
stallation depends upon the characteristics of
the particular blowdown system. In general,
the allowable pressure drop through the relief
valve headers, liquid traps, burners, and so
forth, must riot exceed one-half the internal
unit's relieving pressure. The burner cut-in
schedule is based upon a knowledge of the source,
frequency, and quantity of the release gases.
Pressure downstream of the control valves must
be adequate to provide stable burner operation.
Flare installations designed for relatively small
gas flows can use clusters of commercially avail-
able venturi burners. For large gas releases,
special venturi burners must be constructed. The
venturi (air-inspirating) burners are installed
in clusters 'with a small venturi-type pilot burn-
er in the center. This burner should be connected
to an independent gas source. The burners may
be mounted vertically or horizontally. The burn-
ers should fire through a refractory wall to pro-
vide protection for personnel and equipment. Con-
trols can be installed to give remote indication of
the pilot burner's operation.
For large-capacity venturi burners, field tests
are necessary to obtain the proper throat-to-
orifice ratio and the minimum pressure for stable
burner operation. The design of one flare sys-
tem using special venturi burners has been re-
ported (Brumbaugh, 1947). An analysis of the
burner limitations and the pressure relief sys-
tem in this installation yielded the design data
set forth in Table 165.
No
Cut-m
pres&ure,
psi
Cut-out
pressure,
psi
2-1/4 !/4
i 2-3/4 j 1/2
3 ! 3-1/4 ! 3/4
4 ! 3-3/4
1
' Gas orifice
j diameter,
1 '"•
1.61
1 2.9C
1 4. 03
1
Venturi
throat dia.
in.
5
a
11.5
18
Rates of throat-
to-orifice
area
9.6
7.8
8. I
6.6
After 5 years, this flare was reported to be
satisiactory and had required relatively few
changes (Green, 1952).
The selection of the control valves and burners
for a small-capacity ground flare is indicated
by the following example:
Example 40
Given:
Range of gas flow, 2, 000 to 30, 000 cfh
Most, frequently expected gas flow, 12, 000 cfh
Blowdown line size, 4-in. dia
Specific gravity of gas, 1. 2
Calorific value of gas, 1,300 Btu/ft3
Flowing temperature of gas, 100°F.
Problem:
Select control valves and determine the number
and size of standard air-inspirating burners to
permit smokeless burning of all expected gas
flows.
Solution:
On the basis of the range of expected gas flow,
try three banks of burners with a water seal
bypass to the largest bank to handle gas flows
in excess of flare capacity. The maximum
allowable pressure at the burners has been set
at 5 psig. Various intermediate pressures for
the control valves "will be arbitrarily selected.
The intermediate pressures, which indicate
stable operations of the different burner banks
relative to the gas flows, will be used as the
operating points for the valves.
1. Valve selection and capacity data:
Try two 1-inch and one 2-inch single-seated,
quick-opening valves.
-------
Waste-Gas Disposal Systems
605
Valve
size,
in.
1
2
Capacity
index,
c,r
14
46
Pressure,
psi
0. 5
1. 0
3. 0
5. 0
3. 0
5. 0
Capacity,
cfh
2, 070
2,940
5, 080
6, 580
15, 000
20, 000
Range of No. 1 bank burners is 2, 000 to
6, 000 cfh, with valve capacity range from
2, 000 to 7, 000 cfh.
3. Burner selection--No. 2 bank:
No. 2 bank of burners to be sized such that
capacity of the 1 and 2 banks will equal the
most frequently expected flow of 12, 000 cfh
Burner selection--No. 1 bank:
No. 1 bank of burners to handle a minimum
flow of 2, 000 cfh at 0. 5 psig.
Try a No. 16X NGE burner with a 1/2-in.
orifice.
From Table 166, capacity of a No. 16X burn-
er at 0. 5 psig (1, 000 Btu/ft3 gas) is 1, 360 cfh.
Table 166. VENTURI BURNER CAPACITIES,
ft3/hr (Natural Gas Equipment, Inc., 1955)a
Gas pressure,
in. H2O
2
4
6
8
10
1/2 ps g
1 Ps g
2 ps g
3 Ps g
4 ps g
5 ps g
6 ps g
7 Ps g
8 ps g
Type 14
3/16-in. orifice
70
100
123
142
160
210
273
385
.
Type 16
7/ 16-in. orifice
1, 042
1, 483
2, 157
2,654
3, 065
3, 407
3, 742
4, 040
4, 320
Type 16X
1/2-in. orifice
1, 360
1, 900
2, 640
3,200
3, 680
4, 080
4, 480
4, 800
5, 160
Basis- 1,000 Btu/ft3 natural gas.
Capacity of 1, 300 Btu/ft3 gas:
1, 000
1, 300
x 1,360 cfh = 1, 047 cfh/burner
Number of burners required:
2,000 cfh
-—r—i—;r~rr = 1. 91 burners
1, 047 cfh/burner
Use two burners.
No. 1 bank capacity at other operating pres-
sures :
No. of Capacity, cfh
burners 0.5 psig 1. 0 psig 3. 0 psig 5 psig
2 2,094 2,930 4,920 6,270
Use 6,000 cfh as approximate capacity of
No. 1 bank.
12,000 cfh - 6,000 cfh = 6,000 cfh
Size and capacity of No. 2 bank burners and
valves will be the same as those of No. 1
bank.
4. Burner selection--No. 3 bank:
No. 3 bank capacity must equal the difference
between 30,000 cfh and 12,000 cfh.
30,000 cfh - 12,000 cfh = 18, 000 cfh
From Table 166, capacity of No. 16X burn-
er at 5 psig is 4, 080 cfh (1, 000 Btu/ft3 gas).
Capacity for 1, 300 Btu/ft gas:
1, 000
1, 300
x 4,080 = 3, 140 cfh/burner
Number of burners required:
18,000/cfh
3, 140 cfh/burner
= 5. 7 burners
Use six No. 16X NGE burners.
No. 3 bank capacity at other pressures:
No. of
burners
Capacity, cfh
3.0 psig 5.0 psig
14,760 18,830
Range of No. 3 bank burners, 14, 760 to
18, 300 cfh, with 2-inch valve range of
15,500 to 20,000 cfh.
-------
606
PETROLEUM EQUIPMENT
5. Safety seal:
Basis: Seal pressure
Sealing liquid
Temperature
ft of water =
fa psig
Water
70°F
(6 + 14.7) Ib/in. 2 (144) in. 2/ft2
62.3 lb/ft3
= 47.8 ft
6. Summary of flare operation:
Valve action
Burner Valve
Valve Open, Closed, Range, capacity capacity
No. psig psig psi at 5 psig at 5 psig
1
2
3
1.0
3.0
5. 0
0.5
1.0
3. 0
0. 5-5
1-5
5-6
6, 270
6, 270
18, 830
31, 370
6,580
6, 580
20, 000
33, 160
The bypass seal is set to open to No. 3 burner
bank at 6 psig.
7. Sketch of flare:
Gas flow
Bank No. 1 Bank No. i Bank No. 3
M It 11II
I Seal
Maintenance of flares
Most refineries and petrochemical plants have
a fixed schedule for inspection and maintenance
of processing units and their auxiliaries. The
flare system should not be exempted from this
practice. Removal of a flare from service for
maintenance requires some type of standby equip-
ment to disperse emergency gas vents during the
shutdown. A simple stack with pilot burner should
suffice for a standby. Coordinating this inspec-
tion to take place at time when the major process-
ing units are also shut down is good practice.
Flare instrumentation requires scheduled main-
tenance to ensure proper operation. Most of the
costs and problems of flare maintenance arise
from the instrumentation.
Maintenance expenses for flare burners can be
reduced by constructing them of chrome-nickel
alloy. Because of the inaccessibility of elevated
flares, the use of alloy construction is recom-
mended.
STORAGE VESSELS
TYPES OF STORAGE VESSELS
Even in the most modern petroleum refineries and
petrochemical plants, storage facilities must be
provided for large volumes of liquids and gases.
These facilities can be classified as closed-stor-
age or open-storage vessels. Closed-storage
vessels include fixed-roof .tanks, pressure tanks,
floating-roof tanks and conservation tanks. Open-
storage vessels include ope'n tanks, reservoirs,
pits, and ponds.
Closed-storage vessels are constructed in a vari-
ety of shapes, but most commonly as cylinders,
spheres, or spheroids. Steel plate is the usual
material of construction though concrete, wood,
and other materials are sometimes used. Before
modern welding methods, the sections of the tank
shell were joined by rivets or bolts. Welded joints
are now used almost universally except for the
small bolted tank found in production fields. The
definition of a welded shell tank is given by API
Standard 12 C entitled "Welded Oil Storage Tanks. "
Capacities of storage vessels range from a few
gallons up to 500, 000 barrels, but tanks with
capacities in excess of 150, 000 barrels are rel-
atively rare.
Open-storage vessels are also found in a variety
of shapes and materials of construction. Open
tanks generally have cylindrical or rectangular
shells of steel, wood, or concrete. Reservoirs,
pits, ponds, and sumps are usually oval, circu-
lar, or rectangular depressions in the ground. The
sides and bottom may be the earth itself or may
be covered with an asphalt-like material or con-
crete. Any roofs or covers are usually of wood
with asphalt or tar protection. Capacities of the
larger reservoirs may be as much as 3 million
barrels.
Vapors, gases, aerosols, and odors are exam-
ples of air contaminants emitted from storage
facilities. In most cases, practical and feasible
air pollution control measures are available to
reduce the emissions.
Pressure Tanks and Fixed-Roof Tanks
Pressure tanks and fixed-roof tanks are grouped
together because, in a sense, pressure tanks
are special examples of fixed-roof tanks de-
signed to operate at greater than atmospheric
pressure. A horizontal, cylindrical (bullet)
pressure tank is shown in Figure 431. Other
-------
Storage Vessels
607
Table 167. ROOF PROPERTIES OF STEEL TANKS
(Bussard, 1956)
Figure 431. Horizontal, cylindrical pressure
tank (Graver Tank and Manufacturing Company,
Division Union Tank Car Co., East Chicago,
Indiana).
types of pressure tanks--spheres, plain and
noded spheroids, and noded hemispheroids--are
illustrated in Figure 432. Maximum capacities
of these pressure tanks are as much as 30,000
barrels for spheres and hemispheroids, and
120, 000 barrels for noded spheroids. Spheres
can be operated at pressures up to 217 psi;
spheroids, up to 50 psi; noded spheroids, up
to 20 psi; and plain or noded hemispheroids, up
to 15 and 2-1/2 psi respectively. Horizontal,
cylindrical pressure tanks are constructed with
various capacities and pressures.
The ordinary vertical, cylindrical, fixed-roof
tank is shown in Figure 433. This type of storage
facility operates at or within a few ounces of pres-
sure and may have a flat, recessed flat, conical,
or domed roof. The term gastight, often applied
to welded tanks, is misleading. Many of the roofs
of the welded tanks have free vents open to the
atmosphere. Others are equipped with conserva-
tion vents that open at very slight positive pres-
sures. A tank also has many standard appurte-
nances including gaging hatches, sample hatches,
relief vents, and foam mixers. Any of these acces-
sories may fail in service and result in vapor leaks.
The operating pressure of a tank is limited by the
thickness (weight) of the roof, as noted in Table 167.
A cone roof tank may be operated at higher pres-
sures, if necessary, by structural reinforcement
or -weighting of the roof. Safe operating pressures
up to 4 ounces can be realized by this added ex-
pense. Use of unsupported dome-shaped roofs is
another method of increasing the allowable operat-
ing pressure of the fixed-roof tank.
Floating-Roof Tanks
Floating-roof storage tanks are used for storing
volatile material with vapor pressures in the low-
Thickness, in.
(gage)
1/16 (16)
5/64 (14)
7/64 (12)
1/8 (11)
9/64 (10)
5/32 (9)
11/64 (8)
3/16 (7)a
1/4 (3)
Wt, lb/ft2
2.553
3. 187
4. 473
5. 107
5. 740
6. 374
7. 000
7. 650
10.200
Operating pressure,
oz/in.
0.284
0.354
0. 497
0.568
0.638
0. 708
0.778
0.850
1.333
aMinimum thickness specified by API Std 12C.
er explosive range, to minimize potential fire or
explosion hazards. These vessels also economic-
ally store volatile products that do not boil at at-
mospheric pressures or less and at storage tem-
peratures or below. These tanks are subclassi-
fied by the type of floating-roof section as pan,
pontoon, or double-deck floating-roof tanks (Fig-
ure 434).
Pan-type floating-roof tanks were placed in ser-
vice more than 40 years ago. These roofs re-
quire considerable support or trussing to prevent
the flat metal plate used as the roof from buck-
ling (Figure 434, lower right). These roofs are
seldom used on new tanks because extreme tilting
and holes in the roof have caused more than one-
fifth of installed pan roofs to sink, and because
their use results in high vaporization losses.
Solar heat falling on the metal roof in contact
with the liquid surface results in higher than
normal liquid surface temperatures. Hydrocar-
bons boil away more rapidly at the higher tem-
peratures and escape from the opening around
the periphery of the roof.
To overcome these disadvantages, pontoon sec-
tions 'were added to the top of the exposed deck.
Better stability of the roof was obtained, and a
center drain with hinged or flexible connections
solved the drainage problem. Center-weighted
pontoons, double pontoons, and high- and low-
deck-pontoon floating-roof tanks are available
today. Current practice is to use the pontoon
roof on tanks with very large diameters. In-
cluded with some pontoon roof designs is a, vapor
trap or dam installed on the underside of the roof.
This trap helps retain any vapors formed as a
result of localized boiling and converts the dead
vapor space into an insulation medium. This dead
vapor space tends to retard additional boiling.
The more expensive double-deck floating roof -was
eventually introduced to reduce the effect of solar
boiling and to gain roof rigidity. The final design
-------
608
PETROLEUM EQUIPMENT
it
Figure 432. Types of pressure vessels: (upper left) 51-foot-diameter spheres (Bu-
tane is stored in these spheres at a petroleum refinery in California. Capacity of
each is 15,000 barrels, diameter is 54 feet 9 inches and design working pressure is
35 pounds per square inch.); (upper right) two 5,000-barrel spheroids designed for
20-psi pressure; (lower left) large noded spheroids, each designed for 100,000-bar-
rel capacity and 15-psi pressure; (lower right) a 20,000-barrel noded hemispheroid
designed for 2-1/2-psi pressure (Chicago Bridge and Iron Company (1959).
kHA ',
generally incorporates compartmented dead-air
spaces more than 12 inches deep over the entire
liquid surface. The top deck is generally sloped
toward the center or to a drainage area. Any
liquid forming or falling on the roof top is drained
away through a flexible roof drain to prevent the
roof from sinking. The bottom deck is normally
coned upwards. This traps under the roof any
vapors entrained with incoming liquid or any va-
pors that might form in storage. A vertical dam
similar to those used on pan or pontoon floating
roofs can also be added to retain these vapors.
Conservation Tanks
Storage vessels classified a.s conservation
tanks include lifter-roof tanks and tanks
with internal, flexible diaphragms or in-
ternal, plastic, floating blankets. The
lifter roof or, as more commonly known,
gas holder, is used for low-pressure gas-
eous products or for low-volatility liquids.
This type of vessel can be employed as a
vapor surge tank when manifolded to vapor
spaces of fixed-roof tanks.
-------
Storage Vessels
609
Figure 433. Vertical, cylindrical, fixed-roof
storage tank.
Two types of lifter-roof tanks are available, as
shown in Figure 435. One type has a dry seal
consisting of a gastight, flexible fabric; the other
type employs a liquid seal. The sealing liquid
can be fuel oil, kerosene, or water. Water should
not be employed as a sealing liquid where there
is danger of freezing.
The physical weight of the roof itself floating on
vapor maintains a slight positive pressure in the
lifter-roof tank. When the roof has reached its
maximum height, the vapor is vented to prevent
overpressure and damage to tank.
The conservation tank classification also includes
fixed-roof tanks with an internal coated-fabric
diaphragm, as shown in Figure 436. The dia-
phragm is flexible and rises and falls to balance
"
Figure 434. Types of floating-roof tanks: (upper left) Sectional view of single-deck
center-weighted (pan-type) floating roof; (upper right) sectional view of pontoon
deck floating roof; (lower left) cutaway view of double-deck floating roof- (lower
right) cutaway view of trussed-pan floating roof (Graver Tank and Manufacturing
Company, East Chicago, Ind.).
-------
610
PETROLEUM EQUIPMENT
Figure 435. Types of lifter-roof tanks: (left) Sectional view of
expansion roof tank with a liquid seal, (right) closeup view of
liquid seal and vapor piping (Graver Tank and Manufacturing Co.,
Division of Union Tank Car Co., East Chicago, Indiana).
f 1
Figure 436. Conservation tanks; (left) Sectional view of inte-
grated conservation tank with internal, flexible diaphragm;
(right) cutaway view of a vapor conservation tank showing
flexible membrane (Chicago Bridge and Iron Co., Chicago, III.).
GPO 8O6—614—2t
-------
Storage Vessels
611
pressure is 1/2 ounce per square inch, which is
approximately one-eighth the operating pressure
possible with most gas holders. Two basic types
of diaphragm tanks are the integrated tank, which
stores both liquid and vapor, and the separate
tank, which stores only vapor. Common trade
names for integrated tanks are "diaflote, " "dia-
lift, " and "vapor-mizer" tanks (Bussard, 1956),
or they may be referred to as vapor spheres or
vapor tanks. The separate type of tank offers
more flexibility and does not require extensive
alteration of existing tanks.
Open-Top Tanks, Reservoirs, Pits, and Ponds
The open-top tank is not used as extensively as
in the past. Safety, conservation, and house-
keeping are factors effecting the elimination of
open vessels. Even tanks that require full access
can and should be equipped with removable covers.
The open vessels generally have a cylindrical
shell, but some have a rectangular shell.
Reservoirs were devised to store the large quanti-
ties of residual oils, fuel oils, and, sometimes,
crude oils resulting from petroleum production
and refining. Safety considerations, larger fixed-
roof tanks, and controlled crude oil production
have reduced the number of reservoirs in use to-
day. Even when covered, reservoirs have open
vents, which maintain atmospheric pressures in
the reservoir. Windbreaks divert the windflow
pattern over a large roof area and prevent the
roof from raising and buckling.
Open ponds or earthen pits were created by diking
low areas or by excavation. These storage facili-
ties served for holding waste products, refinery
effluent water, or inexpensive oil products for
considerable periods of time. In these, oils
"weathered" extensively, leaving viscous, tar-
like materials, and water seeped into the lower
ground levels. As the pond filled -with solids and
semisolids, the contents were removed by me-
chanical means, covered in place, or the pond
was simply abandoned. The use of these ponds
has diminished, and the remaining ponds are usu-
ally reserved for emergency service.
Smaller ponds or sumps were once used extensive-
ly in the crude oil production fields. This use was
primarily for drilling muds though oil-water emul-
sions and crude oil were also stored by this method.
Their use is gradually disappearing because unat-
tended or abandoned sumps cause nuisance problems
to a community.
THE AIR POLLUTION PROBLEM
Different types and quantities of air pollution can be
associated with the storage vessel. The types of
pollution can be separated into three categories--
vapors, aerosols or mists, and odors. Of these
pollutants, the largest in quantity and concentra-
tion are hydrocarbon vapors.
Factors Affecting Hydrocarbon Vapor Emissions
Emissions of hydrocarbon vapors result from the
volatility of the materials being stored. They are
effected by physical actions on the material stored
or on the storage itself. Changes in heat or pres-
sure change the rate of evaporation. Heat is a
prime factor and can cause unlimited vaporization
of a volatile liquid. Heat is received from direct
solar radiation or contact with the warm ambient
air, or is introduced during processing. The rate
of evaporation is correlated with atmospheric tem-
perature, weather conditions, tank shell tempera-
ture, vapor space temperature, and liquid body
and surface temperatures.
The vapor space of a tank can contain any degree
of saturation of air •with vapor of the liquid up to
the degree corresponding to the total vapor pres-
sure exerted by the liquid at storage temperatures.
Since the pressure in this vapor space increases
with temperature increase, some of the air-vapor
mixture may have to be discharged or "breathed
out" to prevent the safe operating pressure of the
tank from being exceeded. These emissions are
continually promoted by the diurnal change in at-
mospheric temperatures, referred to as the tank's
breathing cycle.
When the air temperature cools, as at night, the
vapor space "within the tank cools and the vapors
contract. Fresh air is drawn in through tank vac-
uum vents to compensate for the decrease in vapor
volume. As this fresh air upsets any existing
equilibrium of saturation by diluting the vapor con-
centration, more volatile hydrocarbons evaporate
from the liquid to restore the equilibrium. When
the atmospheric temperature increases, as occurs
with daylight, the vapor space warms, and the
volume of rich vapors and the pressure in the tank
increase. In freely vented tanks, or -when the
pressure settings of the relief vents have been ex-
ceeded, the vapors are forced out of the tank. This
cycle is repeated each day and night. Variation
in vapor space temperature also results from
cloudiness, wind, or rain.
Filling operations also result in expulsion of part
or all of the vapors from the tank. The rate and
quantity of vapor emissions from filling are di-
rectly proportional to the amount and the rate at
which liquid is charged to the vessel. Moreover,
as the liquid contents are withdrawn from the
tank, air replaces the empty space. This fresh
air allows more evaporation to take place.
-------
612
PETROLEUM EQUIPMENT
Another emission of vapors caused by atmospheric
conditions is termed a windage emission. This
emission results from wind's blowing through a
free-vented tank and entraining or educting some of
the saturated vapors. The windage emission is
not as large as that occurring during breathing
or filling cycles. Other variables affecting emis-
sions include: Volume of vapor space, frequency
of filling, and vapor tightness of the vessel. Tanks
that can be kept completely full of liquid limit the
volume of the vapor space into which volatile hy-
drocarbons can vaporize and eventually be emitted
to the atmosphere. The frequency of filling and
emptying a tank influences the overall vapor emis-
sions. When extensive periods of time elapse
between pumping operations, the vapor space of a
tank becomes more nearly fully saturated with
vapor from the liquid. Then, during filling of the
tank or during breathing cycles, a larger concen-
tration of vapors exists in the air-vapor mixture
vented to the atmosphere. Vapor tightness of the
tank can influence the evaporation rate. The mov-
ing molecule in the vapor state tends to keep going
if there is no restraining force such as a tight shell
or roof.
Different causes of emissions are associated with
a floating-roof tank. These causes are known as
wicking and wetting. Wicking emissions are caused
by the capillary flow of the liquid between the outer
side of the sealing ring and the inner side of the
tank wall. The wetting emission results when the
floating roof moves towards the bottom of the tank
during emptying and leaves the inner tank shell
covered with a film of liquid, which evaporates
when exposed to the atmosphere.
Hydrocarbon Emissions From Floating-Roof Tanks
The American Petroleum. Institute (1962b) has
published a method of determining the standing
(wicking) and withdrawal (wetting) evaporation
emissions associated with floating roof-tanks.
The method is applicable to tanks in crude oil as
•well as gasoline service. It is based upon field
test data for the standing emission, and labora-
tory data for the withdrawal emission. The corre-
lation presents factors under many combinations
of tank construction, type and condition of roof
seal, and color of tank paint. Parameters in-
clude range of vapor pressure from 2 to 11 psia
true vapor pressure, 4 to 16 mph average wind
velocity, and 20- to 200-foot-diameter tanks.
The standing storage emission is determined
from Table 168. It is the product of emission
factor L£ obtained from the graph and correspond-
ing factors obtained from the table. One must
know the following factors to find the value of the
standing storage emissions: (1) Type of product
stored, (2)Reid vapor pressure, (3) average
storage temperature, (4) type of shell construc-
tion, (5) tank diameter, (6) color of tank paint,
(7) type of floating roof, (8) type and condition
of seal, and (9) average wind velocity in area.
The standing storage emission formula is given a
k k k
s c p
(127)
where Ly = standing storage evaporation emis-
sion, bbl/yr
k = tank factor -with values as follows:
0. 045 for welded tank -with pan or pontoon ro
single or double seal;
0. 11 for riveted tank with pontoon roof, doub
seal;
0. 13 for riveted tank with pontoon roof, sing^
seal;
0.13 for riveted tank •with pan roof, double si
0. 14 for riveted tank, •with pan roof, single se
(double deck roof is similar to a pontoon roof
D = tank diameter, ft [for tanks larger than 150ft
in diameter use ISO1- 5 (I)/150)]
P = true vapor pressure of stock at its average
storage temperature, psia
V = average wind velocity, mph
k = seal factor:
1. 00 for tight-fitting seals (typical of modern
metallic or tube seals)
1. 33 for loose-fitting seals
k = stock factor:
c
1. 00 for gasoline stocks
0. 75 for crude oils
k = paint factor for color of shell and roof:
1.00 for aluminum or light grey
0. 90 for white.
Actual standing storage emissions of petroleum
hydrocarbons from tanks equipped with seals in
good operation should not deviate from the esti-
mated emissions determined by this equation by
more than + 25 percent. The actual emissions,
however, can exceed the calculated amount by
two or three times for a seal in poor condition.
The seal length can be expressed in terms of tank
diameter because the two are directly proportion-
al to each other. The actual emission is not di-
-------
Storage Vessels
613
Table 168. STANDING STORAGE EVAPORATION EMISSIONS FROM FLOATING-ROOF TANKS:
Ly (LOSS IN bbl/yr) = Lf (LOSS FACTOR FROM FIGURE 437) TIMES MULTIPLYING FACTOR
(FROM THIS TABLE; American Petroleum Institute, 1962b)
Multiplying
factors
apply to
Gasoline
Crude oil
Welded tanks
Pan or pontoon roof
Single or double seal
Modern
Tank
paint*1
Lt
grey
1.0
0.75
White
0.90
0.68
Olda
Tank
paint
Lt
grey
1. 33
1.0
White
1.20
0.90
Riveted tanks
Pan roof
Single seal
Modern
Tank
paint
Lt
grey
3.2
2.4
White
2.9
2.2
Olda
Tank
paint
Lt
grey
4.2
3. 1
White
3.8
2.8
Double seal
Modern
Tank
paint
Lt
grey White
2.8 2.5
2.1 1.9
Olda
Tank
paint
Lt
grey
3.8
2.8
White
3.4
2. 5
Pontoon roof
Single seal
Modern
Tank
paint
Lt
grey
2.8
2. 1
White
l.S
1.9
Olda
Tank
paint
Lt
grey
3.8
2.8
White
3.4
2.5
Double seal
Modern
Tank
paint
Lt
grey
2. 5
1.9
White
2.2
1. 7
Olda
Tank
paint
Lt
grey
3. 3
2.5
White
3.0
2.2
aSeals installed before 1942 are classed as old seals.
Aluminum paint is considered light grey in loss estimation.
\
FOR AVERAGE HIND VELOCITY
REFER TO API BULLETIN
2513, EVAPORATION LOSS IN
THE PETROLEUM INDUSTRY-
CAUSES AND CONTROL, OR
LOCAL HEATHER BUREAU DATA
/I / I/ /I / I / \7V
30 4050 60 70 80 90 100 ttO 120 130 140 150
[ILI/J//Z.
FOR TANKS LARGER THAN ISO ft
DIAMETER, MULTIPLY LOSS FOR
150-ft-DIAMETER TANK BY RATIO
8 7 6 5 4 3 2 1
TRUE VAPOR PRESSURE, psia
100 20O 300 400 50O 600 700 800 900 IOOO
LOSS FACTOR, L,
(MULTIPLY BY VALUE FROM TABLE TO OBTAIN ADJUSTED LOSS)
Figure 437. Calculation of emission factor, Lf, for standing storage evaporation
emissions from floating-roof tanks (see Table 168).
rectly proportional to the diameter because sev-
eral other variables are involved. Items such as
wind velocity and the decreased shading effect
of the shell on the roof of large-diameter tanks
are examples.
Emissions increase, but not directly, as the vapor
pressure increases. The relationship P/(14. 7 - P)
correctly identifies this phenomenon, and no sub-
stantial error exists within the valid range of
this correlation.
-------
614
PETROLEUM EQUIPMENT
Standing storage emissions increase but do not
double when the average "wind velocity doubles.
The 0. 7 exponent applied to the wind factor fits
data for average wind velocities exceeding 4 mph.
No localities were recorded as having less than
this 4 mph average wind velocity.
This chart is intended for stabilized crudes that
have not been subjected to extreme weathering
or mixed with light oils.
The average stock temperatures should be used
in these vapor pressure determinations.
Withdrawal emissions
As product is withdrawn from a floating-roof tank,
the wetted inner shell is exposed to the atmosphere.
Part of the stock clinging to the inner surface drains
down the shell. The remainder evaporates to the
atmosphere. Tests made determined the amount
of gasoline clinging to a rusty steel surface as
ranging from 0. 02 to 0. 10 barrel of gasoline per
1, 000 square feet of surface.
The withdrawal emissions are represented by the
equation
32, 400 §
(128)
where
Withdrawal emissions should be added to the stani
ing storage emissions when gunited tanks are en-
countered.
Hydrocarbon Emissions From Low-Pressure Tanks
Low-pressure tanks are used to store petroleum
stocks of up to 30 pounds RVP* with relief valve
settings of 15 psig. The American Petroleum In-
stitute's Evaporation Loss Committee (1962c)
recommends a theoretical approach to emission
calculations from tanks such as these. Insufficien
data are available to establish any accurate corre-
lation with actual field conditions.
Application of the following equation indicates the
theoretical pressure (P?) required to prevent
breathing losses:
W = -withdrawal emissions, bbl per million
bbl throughout
C = 0.02 (based on barrels of clingage per
1,000 ft2 of shell surface)
D = tank diameter, ft
Withdrawal emissions for gunited tanks can be
significant. Laboratory data indicated a factor
of C = 2. 0. Since withdrawal emissions counter
standing storage emissions, a factor C = 1. 0 is
recommended for gunite-lined tanks storing gaso-
line.
Application of results
The emissions from floating-roof tanks can be
estimated from Table 168. Necessary data in-
clude: Tank diameter; color of tank paint; type
of tank shell, roof, and seal; Reid vapor pres-
sure and average temperature of stored product;
and the average wind velocity at tank site.
The true vapor pressure, P, can be obtained
from vapor pressure charts by the use of data
in Figures 438 and 439. To use these charts,
one must know the Reid vapor pressure of the
stock. Figure 438 is used for gasoline or other
finished stocks. The value of S (slope of the
ASTM distillation curve at 10 percent evaporated)
can be estimated by using suggested values given
in a note of the chart. The value of S is zero for
a single component stock. The vapor pressure
chart, Figure 439, should be used for crude oils.
= 1.1 (P +
£i
(129)
where
P? - gage pressure at which pressure vent
opens, psig
P = atmospheric pressure
3-
P. = gage pressure at which vacuum vent
opens, psig
p = true vapor pressure at 90°F minimum
liquid surface temperature, psia
p = true vapor pressure at 100°F maximum
liquid surface temperature, psia.
This equation is applicable only when the vapor
pressure at minimum surface temperature (P})
is less than the absolute pressure (Pj + Pa) at
which the vacuum vent opens. Air always exists
in the vapor space under a condition such as this.
Figure 440 is a plot of equation 129. The pres-
sure required to eliminate breathing emissions
from products ranging up to 17. 5 psia TVPt at
100°F storage temperature and 14. 7 atmospheric
pressure can be determined from this curve. The
gage pressure at which the vacuum vent opens
*RVP refers to Reid vapor pressure as measured by ASTM D
323-58 Standard Method of Test for Vapor Pressure of
Petroleum Products (kejd Method).
fTVP refers to true vapor pressure.
-------
Storage Vessels
615
I— 0.20
0.30
0.40
0.50
0.60
0.7O
0.80
0.90
1.00
— 1.50
— 2.00
2.50
3.00
3.50
4.00
5.00
6.00
7.00
— 8.00
— 9.00
— I 0.0
— I 1.0
— 12.0
— 13.0
14.0
15.0
I 6.0
17.0
I B.O
I 9.0
20.0
21.0
22.O
23.0
24.0
120—T
S = SLOPE OF THE astm DISTILLATION
CURVE AT 10% EVAPORATED=
°F AT 15% MINUS °F AT 5%
10
IN THE ABSENCE OF DISTILLATION
DATA THE FOLLOWING AVERAGE VALUE
OF S MAY BE USED
100 —
90 —
60 —
70 ~
60 —
50-
40 —
30 —
20 -E
10 —
0—'
MOTOR GASOLINE
AVIATION GASOLINE
LIGHT NAPHTHA (9 TO U Ib
NAPHTHA (2 TO 8 Ib rvp)
rvp)
3
2
3.5
2.5
Figure 438. Vapor pressures of gasolines and finished petroleum
products, 1 Ib to 20 Ib RVP. Nomograph drawn from data of the
National Bureau of Standards (American Petroleum Institute, 19B2b).
(Pj) is zero for this curve. The values of p^ and
p^ were obtained from Figure 438. Since higher
vapor pressure stocks have a smaller distillation
slope (s), a range of distillation slopes was used.
The altitude of the storage vessel's location af-
fects the required storage pressure. Proper ad-
justments for various altitudes can be made by
substituting the proper atmospheric pressure
(P ) in equation 129. Table 169 lists atmospher-
ic pressures at various altitudes.
Some pressure tanks must be operated at relative-
ly low pressures — some by design, others because
of corroded tank conditions. Pressure settings
from zero to 2. 5 psig are believed to decrease the
breathing emissions from 100 percent to zero per-
cent, depending upon the vapor pressure of the
material stored. This is shown in Figure 441.
Each additional increment of pressure reduces
the breathing emissions by a progressively smaller
amount. Boiling emissions occur when the true
vapor pressure of the liquid exceeds the pressure
-------
616
PETROLEUM EQUIPMENT
9
10
I I
12
13
14
15
20
25
, — 2
— 3
— 10
140 —^3
130 —E
120 —E
I 10 -E
100 —E
90 -=
80 -=
70 —E
60
50 —E
30 —E
20
10 —E
Figure 439. Vapor pressures of crude oil (American Petroleum
Institute, 1962b).
Table 169. ATMOSPHERIC PRESSURE AT
ALTITUDES ABOVE SEA LEVEL
(American Petroleum Institute, 1962c)
Altitude, ft
1, 000
2, 000
3, 000
4, 000
5, 000
Pressure, psia
14. 17
13.66
13. 17
12.69
12. 23
vent setting. If this vapor pressure equals or
exceeds the absolute pressure (Pj 4- PR) at which
the tank vent opens, air is kept out of the tank.
The absolute tank pressure then equals the vapor
pressure of the liquid at the liquid surface tem-
perature. The storage pressure required to pre-
vent boiling is
- P
(130)
This equation is also indicated in Figure 440.
These minimum pressure requirements have
-------
Storage Vessels
617
14
12
7
BOILING CURVE
z
7
NOTE FOR VALUES OF P2 BETWEEN 20 AND 30 psia,
MULTIPLY THE REID VAPOR PRESSURE AT
100°F BY 1 07
••^••••••••••••H •••• • • •••••••• •• •
•• • •••••••• • • •••••••B mm I
• •••••••• • m ••••••*• MI I
• •••••••• • • •••••••B •< I
» •••••••• • • •••••••• Ml •
• •••••••• • • ••••••«* VI |
• •••••••• • • •••••••• •• I
• •••••••• • • •••••••• •• I
• •••••••• • • ••••••*• mm I
• •••••••• • • •••••••• •• •
10
TRUE
15
VAPOR PRESSURE AT 100°F
20
psia
25
30
Figure 440. Storage pressure required to eliminate breathing and boiling losses.
For values of po between 20 and 30 psia, multiply the Reid vapor pressure at
100°F by 1.07 (American Petroleum Institute, 1962c).
proved adequate to prevent boiling emissions
under usual storage conditions. The true vapor
pressure at 100°F can be obtained from Figure
438 up to 20 pounds RVP. In the range of 20 to
30 RVP, P^ is approximately 7 percent higher
than the RVP at 100°F.
A filling or working emission occurs if the tank
pressure exceeds the vent setting. During the
initial stage of filling, compression of the air-
hydrocarbon mixture with some condensation of
vapor takes place if the tank pressure is less
than the pressure vent setting. This condensa-
tion maintains a fairly constant hydrocarbon
partial pressure. Thus, a certain fraction of
the vapor space can be filled with a liquid be-
fore the tank pressure increases above the vent
setting. As filling continues, the total pressure
increases to the pressure at which the relief
valve opens. Venting to the atmosphere occurs
beyond this point. If there is no change in tem-
perature of the liquid or vapor during the filling
-------
618
PETROLEUM EQUIPMENT
100
60
eo
:r 20
10
20
30
40
OPERATING PRESSURE RANGE, oi/in.
Figure 441. Relationship for estimating motor
gasoline breathing emissions from tanks oper-
ating at less than the recommended 2.5-psig
vent setting (American Petroleum Institute,
1962C).
period, the liquid entering the tank displaces
to the atmosphere an equal volume of vapors.
The total emissions depend upon the capacity of
the vapor space of the tank. Since the tempera-
ture changes as condensation occurs, the rates
of filling and emptying can also affect the vapor
emissions. These variables increase the diffi-
culty of determining the actual emissions. In
order that theoretical emissions can be calcu-
lated, two assumptions are made:
1. Equilibrium exists between the hydrocarbon
content in the vapor and liquid phases under
given temperature and pressure conditions.
2. Filling begins at slightly below atmospheric
pressure.
The following equation can then be derived:
3 p (P - P - p >
V
100 (P + P - p )
£L £. V
(131)
where
F = working emissions, % of volume pump
p = true vapor pressure at liquid tempera-
ture, psia
P = gage pressure at. which vacuum vent
opens, psig
P = gage pressure at which pressure vent
opens, psig.
This calculated emission is correct on the assump-
tion that the vapor pressure of the liquid at its sur-
face temperature and the vapor space temperature
are the same at the start and end of filling. The
emissions, expressed as a percentage, are re-
duced to the extent that the tank is not completely
filled.
Obtaining the true liquid-surface temperature is
difficult. Thus, the value of pv is based upon the
average main body temperature of the liquid. As
a result of possible variables, the required pres-
sure to prevent breathing emissions from low-
pressure tanks, as found by Figure 440, should
be considered to have no pressure rise available
to decrease the working emissions. The working
emissions can be found in tie same manner as
for an atmospheric tank.
Figure 442 is based upon equation 131, except
that the emission values are plotted for various
vapor pressures and pressure vent settings
greater than atmospheric pressure. The straight
line gives theoretical filling emissions from tanks
with vents set at only slightly greater than at-
mospheric pressure. The values are representa-
tive for 12 turnovers per year normally experi-
enced with this type of low-pressure storage.
Hydrocarbon Emissions From Fixed-Roof Tanks
A revised method of determining hydrocarbon
emissions from fixed-roof tanks has been pub-
lished by the American Petroleum Institute
(1962a). Various test data were evaluated and
correlated to obtain methods of estimating breath-
ing emissions and filling and emptying (working)
emissions from fixed-roof storage tanks. The
method is applicable to the full range of petroleum
products, from crude oil to finished gasoline.
Data were considered only for tanks with tight
bottoms, shells, and roofs. All tank connections
were assumed to be vapor tight and liquid tight.
Of 256 separate tests recorded and screened, 178
were found acceptable for correlation. A limited
number of factors were definitely found to estab-
lish a correlation. The following factors were
applied in the correlation:
1. True vapor pressure, P, at storage condi-
tions, in pounds per square inch absolute
(if temperature of the liquid was not available,
a temperature 5°F above average atmospheric
temperature was selected);
P = atmospheric pressure, psia
tank diameter, D, in feet;
-------
Storage Vessels
619
0 5
-------
630
PETROLEUM EQUIPMENT
Table 170. PAINT FACTORS FOR DETERMINING
EVAPORATION EMISSION FROM FIXED-ROOF TANKS
(American Petroleum Institute, 1962a)
Tank color
Roof
White
Alumlnuma
White
Aluminuma
White
Aluminum
White
Light gray
Medium gray
Shell
White
White
Aluminum3-
Aluminum*
Aluminum^
Aluminum
Gray
Light gray
Medium gray
Paint factor
Paint in
good condition
1.00
1.04
1. 16
1.20
1.30
1.39
1.30
1.33
1.46
Paint in
poor condition
1. 15
1.18
1.24
1.29
1.38
1.46
1.38
1.38
1.38
aSpecular.
bD if fuse.
0.8
0.6
0.4
0.2
20
T»KK DIMETER, ft
30
Figure 443. Adjustment factor for smaI[-diameter tanks
(American Petroleum Institute, 1962a).
tanks 30 feet or more in diameter, use
a factor = 1. The breathing emissions
from fixed-roof tanks can also be esti-
mated from Figure 444, as well as from
equation 133.
The working emissions include two phases of stor-
age: (1) The filling emissions under which vapors
are displaced by incoming liquid, and (2) the
emptying emissions, which draw in fresh air and
thus allow additional vaporization to take place.
Variables considered in determining this loss are
true vapor pressure, throughput, and tank turn-
overs, which yield the equation:
F _ /3 PV \
110,ooo y
K
(134)
where
F = -working loss, bbl
-------
Storage Vessels
621
"STT
n.
s
«
t
£
:°
t
I
c
(.
ff
c
c
1
c
1
c
'
j
*n « •« o> to •» n J
83288SSS?
|||| ^
^iliPH
J4« =l
III .r
s&s*g1&£l»
I I I =1
i i 1
8^
7
y
y
y
y
y
y
V
~7
71
y
y
y
0
J 1
S a
.00001
• 0006
• 0008
• '0001
0009
ooos
ooor
oooc
OOfZ
ooo:
ocii
ooci
OCIl
0001
006
008
001
009
ooc
ooc
ocz
002
Cll
Ofl
CZ1
001
06
08
at
09
OS
•OOCZI
.00001
• 0006
•0008
•ooo/
0009
• oooc
oooc
.ooc:
oooz
Oil I
oosi
OSl I
0001
006
008
oat
009
OOi
OOf
ooc
os:
oo:
5^1
osi
szi
ooi
06
08
01
09
oc
or
01
6
I
I
9 3
o
o
CM
•*- CO
o en
O CD
— 4-*
to =3
CD to
c=
DC
-C =3
•*-• £0
CO
CL> O
I— w
CO +-•
-------
622
PETROLEUM EQUIPMENT
P = true vapor pressure at storage temper-
atures, psia (if these temperature data
are not available, estimation of 5 °F
a~bove average ambient temperature is
satisfactory)
V = volume of liquid pumped into tank, bbl
K = turnover factor determined from Figure
445..
a. 8
0.4
0.2
NOTE: FOR 36 TURNOVERS PER YEAR
0 36
too
200 300
TURNOVERS PER YEAR
400
500
Figure 445. Effect of turnover on working
emission's. For 36 or less turnovers per
year, 1U = 1.'0 (American Petroleum Insti-
tute, 1'9B2a).
By using equation 134, a nomograph has been de-
veloped in Figure 446 showing the working emis-
sions of gasoline and crude oil from fixed-roof
tanks. Limited data resulted in the committee's
using the same formula for crude oil breathing
emissions as for gasoline breathing emissions
with an applied adjustment factor Kc. This ap-
proach is based upon an assumption that the emis-
sions from crude oil storage vary in the same
manner as the emissions from gasoline storage,
calculated from variables in equation 132. The
adjustment factor, KC, represents the ratio be-
t-ween the respective emissions. The true vapor
pressure of crude oil must be determined from
Figure 439. This figure applies to stabilized
crude oil only. The breathing emission factor of
0. 58 results in part from, slower convective move-
ment. This is true in the case of a liquid surface
less volatile than the body of the liquid. In con-
sidering the working emissions from crude oil
storage, however, filling cycles are normally
less frequent than daily breathing cycles are. Thus
more crude oil -evaporates between cycles, creat-
ing a more saturated vapor space. The action of
filling causes fresh liquid to move to the surface.
A factor somewhere between 0. 58 and unity ap-
pears feasible. A review of the scattered data
available supports a factor of 0. 75. Equation 134
then becomes
F
= f2-Z5PV\ K
I 10,000 / t
(135)
where
F
CO
P
V
K
= working emissions for crude oil, bbl
= true vapor pressure, psia, determined
from Figure 439 (again this may be
estimated at 5°F higher than average
ambient temperature in lieu of better
data)
= volume pumped into tank, bbl
= turnover factor, determined from
Figure 445.
Aerosol Emissions
Storage equipment can also cause air pollution in
the form of aerosols or mists. An aerosol-type
discharge is associated -with storage of heated
asphalt. This discharge is more predominant
during filling operations. The reasons for this
emission, other than basic displacement, are not
thoroughly understood. Continued oxidation of
the asphalt followed by condensation, or conden-
sation of any moisture in the hot gases upon their
entering the cooler atmosphere, are believed to
be the primary causes of the; mists. An analysis
conducted during the filling operation found essen-
tially air and water as the main components of the
displaced vapor. Table 171 shows the results of
this analysis. These vapors are frequently highly
odoriferous.
Whenever live steam or air is added to a vessel
for mixing, heating, oxidizing, or brightening,
droplets or aerosols can be entrained with the
discharge gases. Visible discharges, product
loss, and odors can result.
Odors
The release of odors is closely related to evapo-
ration and filling operations associated with the
storage vessel. The concentration of odors is
not, however, directly proportional to the amount
of material released. Some relatively heavy
compounds are very noticeable at dilutions of 1
to 5 ppm. These compounds are often toxic or
highly malodorous and generally contain sulfur or
nitrogen compounds.
-------
Storage Vessels
623
Agitation, especially by means of air or live
steam, will increase the release of odors to
the atmosphere.
AIR POLLUTION CONTROL EQUIPMENT
Control of air pollution originating from storage
vessels serves a three-fold purpose: (1) Elimina-
tion or reduction of air contaminants, (2) elimina-
tion or reduction of fire hazards, and (3) economic
savings through recovery of valuable products.
Methods of control include use of floating roofs,
plastic blankets, spheres, variable vapor space
systems, various recovery systems, and altered
pumping and storage operations.
J- 30
20-
EXAMPLE:
56000 Barrel Tank
Throughputs 560000 Barred per Year
Turnov«ri=10
True Vapor Preuure= 5.8 psia
Working lots = 975 Barrels per year
1000 -
-
™
,
1500 —
.
•
2000-
•
.
3000 —
-
4000-
sooo -
6000 -
7000 -
8000 -
9000 -
10000 -
>
JS
t
o
£
I
1
j£
^
10.0-
9.0'-
8.0-
7.0-
6.0 -:
.
s.o-
4.0-
:
•
3.0-
•
''u'-
13 ~
1.0-
___
fc "— •
•= -o ^
iii
5 ~ |.
£ "* •
ft. 4- 1-
?'''
20
X
X*
s''
_0 to 36
-40
-50
-60
T -70
• -80
^ -90
8. -100
• -'»
1 -150
-5 -175
Z -200
..- - 250
> -300
| ^400
c
O
*"
10 -
90-
80-
7.0
-------
624
PETROLEUM EQUIPMENT
Table 171. ANALYSIS OF VAPORS DISPLACED
DURING FILLING 85/100 PAVING-GRADE
ASPHALT INTO A FIXED-ROOF TANKa
Component
Methane
Ethane
Heavy hydrocarbons (28° API gravity)
Nitrogen
Oxygen
Carbon dioxide
Water
Argon
Volume %
Trace
Trace
0. 1
67. 3
13. 0
1. 4
18.2
Trace
aSample was collected over 3-1/2-hour filling
period, the noncondensables were analyzed by
mass spectrometer. Condensable hydrocarbons
were separated from the steam, and gravity and
distillation curves were determined.
Seals for Floating-Roof Tanks
The principle by which a floating roof controls
emissions from a volatile liquid is that of elim-
inating the vapor space so that the liquid cannot
evaporate and later be vented. To be successful
the floating roof must completely seal off the liq-
uid surface from the atmosphere (Chicago Bridge
and Iron Company, 1959). The seal for the float-
ing roof is therefore very important. A sectional
view of the sealing mechanism is shown in Figure
447. The floating section is customarily construc-
ted about 8 inches less in diameter than the tank
shell. A sealing mechanism must be provided for
the remaining open annular gap. The seal also
helps keep the roof centered.
Conventional seals generally consist of vertical
metal plates or shoes connected by braces or
pantograph devices to the floating roof. The shoes
are suspended in such a way that they are forced
outward against the inner tank wall. An impervi-
ous fabric bridges the annular area between the
tops of the shoes contacting the tank wall and the
circumference of the floating roof. To reduce
emissions, a secondary seal or wiper blade has
been added to the floating-roof design by extend-
ing the fabric seal or by adding a second section
of fabric as shown in Figure 448. This seal re-
mains in contact with the tank wall. Its flexibility
allows it to make contact even in rivet head areas
of the inner shell or in places where the shell
might be slightly out of round. This improvement
lowers hydrocarbon emissions further by reducing
the effect of wetting and wicking associated with
floating-roof tanks.
Recently, other types of sealing devices to close
the annular gap have been marketed, as shown in
Figure 449. These devices consist of a fabric
tube that rests on the surface of liquid exposed
FLEXURE CLOSURE
FLEXURE
UPPER SNSULATOK
LOWER INSULATOR
PANTAGRAPH HANGER
SEALING RING
•OTTOM DECK
Figure 447. Sectional view of double-deck floating-
roof's sealing mechanism (Chicago Bridge and Iron
Co., Chicago, III.).
in the annular space. The fabric tube is filled
with air, liquid or plastic material. The pneu-
matic, inflated seal is provided with uniform air
pressure by means of a small expansion chamber
and control valves. The sides of the tube remain
in contact with the roof and inner shell. The liq-
uid-filled tube holds a ribbed scuff band against
the tank -wall. The ribbed band acts as a series
of wiper blades as well as a closure. All tubes
are protected by some type of weather covering.
A weather covering can also be added to protect
the sealing fabric of the conventional seals. The
covering includes flat metal sections held in place
by a metal band. The metal protects the fabric
seal from the elements. When floating-roof sec-
tions are added to older tanks constructed of
riveted sections, better contact of the shoes with
the shell can be ensured by guniting or plastic
coating the inner shell. The wetting condition of
gunited walls may, however, offset the gain of
better contact.
Floating Plastic Blankets
A floating plastic blanket, operates on the same
principle of control as a floating roof. It is also
-------
Storage Vessels
625
EXPANSION JOINT FABRIC
SECONDARY SEAL
FABRIC
SOLE PLATE
SEALING RING
FIEX1 ""
STANDARD HORTON
SEALING RING
WITH
PANTAGRAPH
HANGERS
Figure 448. Secondary seals stop vapor loss from
high winds on riveted tanks by sealing off the
space between the tank shell and the sealing ring
sole plate (Chicago Bridge and Iron Co., Chicago,
III.).
available as a surface cover, as depicted in Fig-
ure 450. It was developed in France and has been
tried principally in foreign markets (Laroche
Bouvier and Company). Recent applications have
been made in the United States. The blanket is
usually made of polyvinyl chloride but can be
made of other plastics such as polyvinyl alcohol,
superpolyamides, polyesters, fluoride hydrocar-
bons, and so forth. The blanket's underside is
constructed of a large number of floats of the
same plastic material. The blanket is custom
manufactured so that only a 1-inch gap remains
around the periphery. A vertical raised skirt
is provided at the edge of the blanket to serve as
a vapor seal over the annular area. Once this
area is saturated, further evaporation diminish-
es. The only remaining loss is gaseous diffusion.
The seal is made as effective as possible by using
an elastic, Z-shaped skirt.
top of the blanket. Another feature includes a
stainless steel cable grid to prevent a buildup
of static charges. The grid is closely attached
just under the blanket in parallel lines and con-
nected to the tank shell by a flexible conductor
cable. Installation of a plastic blanket is con-
venient for both new and existing tanks. The
blanket is made in sections and can be introduced
into a tank through a manhole.
A rigid foam-plastic cover constructed of poly-
is ocyanate foam is also available to equip small
fixed-roof tanks with a floating cover. The cov-
er is manufactured in radial sections, each
equipped with a flexible neoprene seal attached
on the outer edge. The sections are easily in-
stalled through roof manholes and assembled
with slip-fit joints.
Plastic Microsph eres
An outgrowth of application of plastic material
provides another type of control mechanism.
This type of control is also similar to the float-
ing roof. A phenolic or urea resin in the shape
of tiny, hollow, spherical particles has been de-
veloped by Standard Oil Company of Ohio (Ameri-
can Petroleum Institute, 1962d). This material
has the physical properties necessary to form a
foam covering over the denser petroleum prod-
ucts. The fluidity of the layer enables it to flow
around any internal tank parts -while keeping the
liquid surface sealed throughout any level changes.
These plastic spheres are known under their trade-
mark names of microballoons or Microspheres.
These coverings have proved to be effective con-
trols for fixed-roof crude oil tanks. Excessive
amounts of condensation or high turbulence should
be avoided. The plastic foam has not proved as
satisfactory for one-component liquid or gasoline
products.
A 1/2-inch layer of foam has been found sufficient
for crude oil where pumping rates do not exceed
4, 000 barrels per hour. A layer 1 inch thick is
recommended for pumping rates up to 10, 000
barrels per hour. In order to overcome wall
holdup in smaller tanks, it is suggested that a
1-inch layer be used regardless of pumping rates.
For tanks storing gasoline, the recommended
foam thickness is 2 inches for tanks up to 40 feet
in diameter, and 1 inch for all larger diameter
vessels.
Provisions are made in the blanket for openings
fitted -with vertical sleeves for measuring and
sampling operations. These openings have a
crosscut, flexible inner diaphragm to minimize
exposure of the liquid surface. Small holes
with downspouts to effect a liquid seal are used
to provide drainage of any condensate from the
Various methods can be used to put the foam cov-
ering on the crude oil. One method is to inject
the plastic spheres with the crude oil as it is
charged to the tank. Spheres are added by means
of an aspirator and hopper similar to equipment
used in fire-fighting foam systems. The spheres
can also be added by placing the desired quantity
-------
626
PETROLEUM EQUIPMENT
WEATHER SHIELD
SEALING BANB
SEALING LIQUID
WEATHER SHIELD
HANGBR BAR
CURTAIN 5*Al
SBAl ENVELOPS
»\ V*^ V- c ,. •» *
,fSv\\ \ v.«' \ , -^
Figure 449. Sealing devices for floating-roof
tanks: (upper left) Liquid-filled tube seal,
(upper right) inflated tube seal, (lower left)
foam-filled tube seal (Chicago Bridge and Iron
Co., Chicago, III.)
-------
Storage Vessels
627
DETMl OF PAKEIS «SSE«6L» OEEASSINE TRAP
fALOKINUM ALLOY PANEL L|fT1NG CABLE -
BUTT STRAP -
PERIPHERAL ANGLE
DETAIL OF PERIPHERAL SEAL
Figure 450. Fixed-roof tank with internal plastic
floating blanket (Laroche Bouvier and Company,
5. Boulevard Edgar-Quinet, Colombes (Seine),France).
in the water or sediment. At high temperatures,
the the r mo-setting resins soften, liquefy, and
mix "with the fuel oil, asphalt, or coke.
Vapor Balance Systems
Variable vapor space or vapor balance systems
are designed to contain the vapors produced in
storage. They do not achieve as great a re-
duction in emissions as an appropriately de-
signed vapor recovery system does. A well-
planned unit includes storage of similar or
related products, and uses the advantage of
in-balance pumping situations. Only the vapor
space of the tanks is manifolded together in
these systems. Other systems include a vapor
reservoir tank that is either a lifter-roof type
or a vessel with an internal diaphragm. The
latter vessel can be an integrated vapor-liquid
tank or a separate vaporsphere. The manifold
system includes various sizes of lightweight
lines installed to effect a balanced pressure
drop in all the branches while not exceeding
allowable pressure drops. Providing isolating
valves for each tank so that each tank can be
removed from the vapor balance system dur-
ing gaging or sampling operations is also good
practice. Excessive vapors that exceed the
capacity of the balance system should be incin-
erated in a smokeless flare or used as fuel.
on the clean, dry floor of the tank just beiore the
crude oil is charged. A wetting agent must be
used when the foam covering is to be used on gas-
oline products. This is accomplished by slurry-
ing the plastic spheres, wetting agent, and gaso-
line in a separate container. The slurry is then
injected into the tank. Changes in tank operation
are not necessary except for gaging or sampling.
A floating-type well attached to a common-type
gaging tape allows accurate measurement of the
tank's contents. A sample thief with a piercing-
type bottom is needed for sampling.
Protection against excessive loss of the plastic
spheres is necessary because of the relative
value of the foam covering. Precaution must be
taken against overfilling and pumping the tank too
low. Standard precautions against air entrainment
in pipelines normally safeguard against the latter.
Overfilling can be prevented by automatic shutoff
valves or preset shutoff operations. Low-level
shutoff should prevent vortices created during
tank emptying. Other than loss of the foam, no
trouble should be encountered if the spheres
escape into process lines. The plastic material
is not as abrasive as the sand particles normally
found entrained in crude oil. Excessive pres-
sures crush the spheres and the plastic settles
Vapor Recovery Systems
The vapor recovery system is in many ways
similar to and yet superior to a vapor balance
system in terms of emissions prevented. The
service of this type of vapor recovery system
is more flexible as to the number of tanks and
products being stored. The recovery unit is
designed to handle vapors originating from, fill-
ing operations as well as from breathing. The
recovered vapors are compressed and charged
to an absorption unit for recovery of condensable
hydrocarbons. Noncondensable vapors are piped
to the fuel gas system or to a smokeless flare.
When absorption of the condensable vapors is
not practical from an economic standpoint, these
vapors, too, are sent directly to the fuel system
or incinerated in a smokeless flare.
The recovery system, like the vapor balance
system, includes vapor lines interconnecting
the vapor space of the tanks that the system
serves. Each tank should be capable of being
isolated from the system. This enables the
tanks to be sampled or gaged without a result-
ing loss of vapors from the entire system. The
branches are usually isolated by providing a
butterfly-type valve, a regulator, or a check
valve. Since the valves offer more line resis-
tance, their use is sometimes restricted. Small
-------
628
PETROLEUM EQUIPMENT
vessels or knockout pots should be installed at
low points on the vapor manifold lines to remove
any condensate.
In some vapor recovery systems, certain tanks
must be blanketed with an inert atmosphere in
order to prevent explosive mixtures and product
contamination. In other, larger systems, the
entire manifolded section is maintained under a
vacuum. Each tank is isolated by a regulator-
control valve. The valves operate from pressure
changes occurring in the tank vapor space.
Because the vapor-gathering system is based
upon positive net vapor flow to the terminus
(suction of compressors), the proper size of
the vapor lines is important. Sizing of the line,
as well as that of the compressors, absorption
unit, or flare, is based upon the anticipated
amount of vapors. These vapors are the result
of filling operations and breathing. The distance
through which the vapors must be moved is also
important.
INTERNED I ATE
LOCATING FLANGE
POSITIONING ROD
BAFFLE PLATE
Miscellaneous Control Measures
Recent tests have shown that breathing emissions
from fixed-roof tanks can be reduced by increas-
ing the storage pressure. An increase of 1 ounce
per square inch was found to result in an 8 per-
cent decrease in emissions due to breathing.
Tanks operated at 2-1/2 psig or higher were found
to have little or no breathing emissions. The
pressure setting, however, should not exceed the
weight of the roof.
A major supplier (Shand and Jurs Co. ) of tank
accessories offers another method of reducing
breathing losses. The method is based upon the
degree of saturation in the vapor space. A baffle
located in a horizontal position immediately below
the vent, as shown in Figure 451, directs enter-
ing atmospheric air into a stratified layer next to
the top of the tank. Since this air is lighter, it
tends to remain in the top area.; thus, there is
less mixing of the free air and any of the rich
vapor immediately above the liquid surface. The
top stratified layer is first expelled during the
outbreathing cycle. Test data indicate a reduced
surface evaporation of 25 to 50 percent.
Hydrocarbon emissions can be minimized further
by the proper selection of paint for the tank shell
and roof. The protective coating applied to the
outside of shell and roof influences the vapor
space and liquid temperatures. Reflectivity and
glossiness of a paint determine the quantity of
heat a vessel can receive via radiation. A cooler
roof and shell also allows any heat retained in the
stored material to dissipate. Weathering of the
paint also influences its effectiveness. The rela-
Figure 451. Air baffle (Shand and Jurs Co., Berkeley,
Calif.).
tionship of paints in keeping tanks from warming
in the sun is indicated in Table 172. Vapor space
temperature reductions of 60°F have been reported.
Similarly, liquid-surface temperature reductions
of 3 to 11 degrees have been achieved. Data
gathered by the American Petroleum Institute on
hydrocarbon emissions indicate breathing emis-
sion reductions of 25 percent for aluminum over
black paint and 25 percent for white over aluminum
paint. All paints revert to "black body" heat ab-
sorption media in a corrosive or dirt-laden atmo-
sphere.
-------
Loading Facilities
629
Table 172. RELATIVE EFFECTIVENESS OF
PAINTS IN KEEPING TANKS FROM WARMING IN
THE SUN (Nelson, 1953)
Color
Black
No paint
Red (bright)
Red (dark)
Green (dark)
Red
Aluminum (weathered)
Green (dark chrome)
Green
Blue
Gray
Blue (dark Prussian)
Yellow
Gray (light)
Aluminum
Tan
Aluminum (new)
Red iron oxide
Cream or pale blue
Green (light)
Gray (glossy)
Blue (light)
Pink (light)
Cream (light)
White
Tin plate
Mirror or sun shaded
Relative effectiveness
as reflector or
rejector of heat, %
0
10. 0
17.2
21.3
21.3
27.6
35.5
40.4
40.8
45.5
47. 0
49. 5
56.5
57.0
59.2
64.5
67.0
69.5
72.8
78.5
81.0
85.0
86.5
88.5
90.0
97.5
100.0
Insulation applied to the outside of the tank is one
method of reducing the heat energy normally con-
ducted through the wall and roof of the vessel.
Another method of controlling tank temperatures
is the use of water. The water can be sprayed
or retained on the roof surface. The evaporation
of the water results in cooling of the tank vapors.
Increased maintenance and corrosion problems
may, however, be encountered.
Storage temperatures may be reduced by external
refrigeration or autorefrigeration. External re-
frigeration units require the circulation of the re-
frigerant or of the tank contents. Autorefrigera-
tion is practical in one-component liquid hydro-
carbon storage where high vapor pressure mate-
rial is involved. The pressure in the tank is re-
duced by removing a portion of the vapor. Addi-
tional vapor is immediately formed. This flash
vaporization results in lowering the temperature
of the main liquid body.
Routine operations can be conducted in such a
manner as to minimize other emissions associ-
ated with storage tanks. Use of remote-level
reading gages and sampling devices reduces
emissions by eliminating the need to open tank
gage hatches. Emissions can be further re-
duced by proper production scheduling to (1)
maintain a minimum of vapor space, (2) pump
liquid _to the storage tank during cool hours and
withdraw during hotter periods, and (3) main-
tain short periods bet-ween pumping operations.
Using wet scrubbers as control equipment for
certain stored materials that are sufficiently
soluble in the scrubbing media employed is both
possible and practical. The scrubbers can be
located over the vent when the scrubbing medi-
um, for example, a water scrubber for aqua
ammonia storage, can be tolerated in the product
In other cases, the vent of one or more tanks
can be manifolded so that any displaced gas is
passed through a scrubbing unit before being
discharged to the atmosphere. A typical ex-
ample is a scrubber packed with plastic spirals
that serves ketone storage vessels. The scrub-
bing liquid is water, which is drained to a
closed waste effluent disposal system.
Properly designed condensers can be used to
reduce the vapor load from tank vents in order
that smaller control devices can be employed.
Masking Agents
Masking agents do not afford any degree of con-
trol of the emissions from storage equipment.
The agent is employed to make the vapor or gas
less objectionable. On the basis of local experi-
ence, the use of these agents is impractical, and
in the long run, proper control equipment is nec-
essary.
Costs of Storage Vessels
The installed costs of various storage vessels
are indicated in Figures 452 through 459. In-
cluded are standard tank accessories such as
manholes, vents, ladders, stairways, drains,
gage hatches, and flanged connections.
LOADING FACILITIES
INTRODUCTION
Gasoline and other petroleum products are
distributed from the manufacturing facility to
the consumer by a network of pipelines, tank
vehicle routes, railroad tank cars, and ocean-
going tankers, as shown in Figure 460.
As integral parts of the network, intermediate
storage and loading stations receive products
from refineries by either pipelines or tank ve-
-------
630
PETROLEUM EQUIPMENT
loop
80
•- 40
/
X
0 20 40 BO 80 100
CAPACITY, millions of barrels
Figure 452. Installed costs of cone roof
tanks.*
l.DOD
100
10
~^?L
0 20 40 60 80 100
CAPACITY, millions of barrels
Figure 453. Installed costs of double-deck
floating roof tanks.*
100
60
60
.- 40
20
7
1,000
20 40 60 80 100
CAPACITY, mi I I ions of barrels
Figure 454. Installed costs of pontoon
floating-roof tanks,*
100
10 20 30 40 50
CAPACITY, millions of barrels
Figure 455. Installed costs of spherical
pressure storage tanks.*
"Including accessories, delivered and erected (Prater and Mvlo
1961; copyrighted by Gulf Publishing Co. w
-------
Loading Facilities
631
1,001
0 10 20 30 40 50
CAPACITY, mi 11 ions of barrels
100
80
- 60
- 40
0 10 20 30 40 50
CAPACITY, mi 11 ions of barrels
Figure 456. Installed costs of. spheroids.*
Figure 457. Installed costs of basic
hemispheroids.*
i.ooop
100
10
100
0 20 40 60 80 100
CAPACITY, millions of barrels
Figure 458. Installed costs of 5-ft lift
expansion roof storage tanks.*
1,000
5 100
10
1,000
100
-110
20 40 60 80 100
CAPACITY, millions ot barrels
Figure 459. Installed costs of 10-ft lift
expansion roof storage tanks.*
•Including accessories, delivered and erected (Prater and Mylo,
1961; copyrighted by Gulf Publishing Co., Houston, Texas).
-------
632
PETROLEUM EQUIPMENT
Figure 460. Representation of gasoline distribution
system in Los Angeles County, showing flow of gaso-
line from refinery to consumer.
hides. If the intermediate station is supplied
by pipeline, it is called a bulk terminal, to
distinguish it from the station supplied by tank
vehicle, which is called a bulk plant. Retail
service stations fueling motor vehicles for the
public are, as a general rule, supplied by tank
vehicle from bulk terminals or bulk plants.
Consumer accounts, which are privately owned
facilities operated, for example, to fuel vehicles
of a company fleet, are supplied by tank vehicles
from intermediate bulk installations or directly
from refineries.
Gasoline and other petroleum products are
loaded into tank trucks, trailers, or tank cars
at bulk installations and refineries by means of
loading racks. Bulk products are also delivered
into tankers at bulk marine terminals.
Loading Racks
Loading racks are facilities containing equip-
ment to meter and deliver the various products
into tank vehicles from storage. Sizes of
loading racks vary in accordance with the
number of products to be loaded and the num-
ber of trucks or railroad tank cars to be
accommodated. The loading platform may
be an elevated structure for overhead filling
of vehicles, that is, through the top hatches in
the tank vehicle, or a ground-level facility for
bottom filling. The elevated-platform structure
employed for overhead filling, shown in Figures
461 and 462, is generally constructed with hinged
side platforms attached to the sides of a central
walkway in such a way that they can be raised
when not in use. Thus, when a vehicle is
positioned adjacent to the central walkway for
loading, the hinged side platforms can be low-
ered to rest upon the top of the vehicle to pro-
vide an access to the compartment hatches. The
meters, valves, loading tubes or spouts, motoi
switches, and similar necessary loading equip-
ment are located on the central walkway. Bottom-
loading installations are less elaborate, since
the tank vehicle is filled through easily accessi-
ble fittings on the underside of the vehicles.
Marine Terminals
Marine terminals have storage facilities for
crude oil, gasoline, and other petroleum prod-
ucts, and facilities for loading and unloading
these products to and from oceangoing tankers
or barges. The loading equipment is on the dock
and, in modern terminals, is similar to elevated-
tank vehicle-loading facilities except for size
(see Figure 463). A pipeline manifold with
flexible hoses is used for loading at older
terminals. Marine installations are consider-
ably larger and operate at much greater loading
rates than inland loading installations.
Loading Arm Assemblies
The term loading arm assembly refers to the
equipment and appurtenances at the discharge
end of a product pipeline thai are necessary to
the filling of an individual tank vehicle or tanker
compartment. Component parts may include
piping, valves, meters, swivel joints, fill spouts,
and vapor collection adapters. These installa-
tions are commonly called loading arms. A
loading arm without provisions to control vapors
displaced from the compartment during filling is
shown in Figure 464.
Overhead loading arms employed for filling of
tank trucks or railroad tank cars maybe classi-
fied in accordance with the manner in which ver-
tical movement of the arm is achieved, such as
pneumatic, counterweighted, or torsion spring.
The pneumatically operated arm is a successor
to the common spring-loaded., automatic-locking
arm in which the spring-loaded cylinder has been
replaced by an air cylinder (see Figure 465).
Bottom loading employs a flexible hose or a non-
flexible, swing-type arm connected to the vehicle
from ground-level pipeline termini.
Loading arms at modern marine terminals are
similar in design to those used for overhead
loading of tank vehicles. The tanker loading
-------
Loading Facilities
633
Figure 461. An overhead-controlled loading rack (Phillips
Petroleum, LOS Angeles, Calif.).
arms are too large for manual operation, re-
quiring a hydraulic system to effect arm motion.
Older installations use reinforced, flexible hoses
to convey products from pipeline discharge mani-
folds to the tanker. The hoses are positioned by
means of a winch or crane.
THE AIR POLLUTION PROBLEM
When a compartment of a tank vehicle or tanker
is filled through an open overhead hatch or bot-
tom connection, the incoming liquid displaces
the vapors in the compartment to the atmosphere.
Except in rare instances, where a tank vehicle
or tanker is free of hydrocarbon vapor, as when
being used for the first time, the displaced va-
pors consist of a mixture of air and hydro-
carbon concentration, depending upon the prod-
uct being loaded, the temperature of the prod-
uct and of the tank compartment, and the type
of loading. Ordinarily, but not always, when
gasoline is loaded, the hydrocarbon concen-
tration of the vapors is from 30 to 50 percent
by volume and consists of gasoline fractions
ranging from methane through hexane (Deckert
et al. , 1958). Table 173 shows a typical analy-
sis of the vapors emitted during the loading of
motor gasoline into tank vehicles.
The volume of vapors produced during the load-
ing operation, as well as their composition, is
greatly influenced by the type of loading or fill-
ing employed. The types in use throughout the
industry may be classified under two general
headings, overhead loading and bottom loading.
Overhead loading, presently the most widely
used method, may be further divided into
splash and submerged filling. In splash fill-
ing, the outlet of the delivery tube is above the
liquid surface during all or most of the loading.
In submerged filling the outlet of the delivery
tube is extended to within 6 inches of the bottom
and is submerged beneath the liquid during most of
the loading. Splash filling generates more turbu-
lence and therefore more hydrocarbon vapors
-------
634
PETROLEUM EQUIPMENT
than submerged filling does, other conditions
being equal. On the basis of a typical 50 percent
splash filling operation, vapor losses from the
overhead filling of tank vehicles -with gasoline
have been determined empirically to amount to
0. 1 to 0. 3 percent of the volume loaded (Deckert
et al. , 1958). Figure 466 presents a correlation
of loading losses with gasoline vapor pressures.
Figure 462. A closeup view of a controlled loading
arm with the access platform in a lowered position
(Phillips Petroleum, Los Angeles, Calif.).
Figure 464. View of uncontrolled loading arm.
Figure 463. Marine terminal loading station
(Chiksan Company, Brea, Calif.).
Figure 465. View of a pneumatically operated loading
arm (Union Oil Company of California, Los Angeles,
Calif.).
-------
Loading Facilities
635
Table 173. TYPICAL ANALYSIS OF
VAPORS FROM THE BULK LOADING OF
GASOLINE INTO TANK TRUCKS
(Deckert et al. , 1958)
Fraction
Air
Hydrocarbon
Propane
Iso-Butane
Butene
N- Butane
Iso-Pentane
Pentene
N-Pentane
Hexane
Vol %
58. I
0.6}
2.9
3.2
17.4
7.7
5. 1
2.0
3.0y
> 41.9
100.0
wt %
37.6
Q.b
3.8
4. 0
22.5
12.4
8.0
3. 1
8.0J
> 62.4
100. 0
o.w
70 80
GASOLINE LIQUID TEMPERATURE DF
Figure 466. Correlation of tank vehicle- loading
losses (50% submerged filling) with Reid vapor
pressure and liquid temperatures of the motor
gasol ine.
Bottom, loading has been introduced by a few oil
companies and found practical for loading trucks
{Hunter, 1959). The equipment required is
simpler than that used for overhead loading.
Loading by this method is accomplished by connect-
ing 3. swing-type loading arm or hose at ground level,
as shown in Figure 467, to a matching fitting on
the underside of the tank vehicles. Aircraft-
type, quick-coupling valves are used to ensure
a fast, positive shutoff and prevent liquid spills.
Several companies experienced in aircraft-fuel-
ing operations have developed fully automatic
hottom-loading systems. All the loading is sub-
merged and under a slight pressure; thus, turbu-
lence and resultant production of vapors are
minimized.
The method employed for loading marine tankers
is essentially a bottom-loading operation. Liquid
is delivered to the various compartments through
lines that discharge at the bottom of each com-
partment. The vapors displaced during loading
are vented through a manifold line to the top of
the ship's mast for discharge to the atmosphere.
In addition to the emissions resulting from the
displacement of hydrocarbon vapors from the
tank vehicles, additional emissions during load-
ing result from evaporation of spillage, drain-
age, and leakage of product.
AIR POLLUTION CONTROL EQUIPMENT
An effective system for control of vapor emis-
sions from loading must include a device to col-
lect the vapors at the tank vehicle hatch and a
means for disposal of these vapors.
Types of Vapor Collection Devices for Overhead Loading
Four types of vapor collectors or closures, fitting
the loading tube, have been developed for use dur-
ing overhead-loading operations of trucks: The
General Petroleum Corporation unit, the Vernon
Tool Company or Greenwood unit, the SOCO unit,
and the Chiksan unit. All are essentially plug-
shaped devices that fit into the hatch openings
and have a central channel through which gasoline
can flow into the tank vehicle compartment. This
central channel, actually a section of the loading
tube, is surrounded by an annular vapor space.
Entry into this vapor space is achieved through
openings on the bottom of the closure that are
below the point of contact of the external closure
surface with the sides of the hatch opening. Thus,
vapors are prevented from passing around the
closure and out of the hatch, and must flow in-
-------
636
PETROLEUM EQUIPMENT
CHEVRON
Figure 467. View of a bottom-loading station (Standard Oil Company of California,
Western Operations Inc., Los Angeles, Calif.).
stead into the annular space, which in turn, is
connected to a hose or pipe leading to a vapor
disposal system.
The vapor closure device developed by the Gen-
eral Petroleum Corporation (now Mobil Oil Corp.)
has the annular vapor space connected to an
auxiliary, transparent, plexiglas vapor chamber
section above the closure to allow the operator
to observe the calibrated capacity markers. * A
typical Mobil Oil Corporation vapor closure is
shown in Figure 468. A neoprene rubber bellows
above the plexiglas chamber compensates for ver-
tical misalignment of the closure in the hatch open-
ing. The closure is aluminum and is cast in the
shape of a truncated cone. The lateral surface of
the closure is faced with a neoprene rubber gasket
in the shape of a spherical section so as to give a
vaportight seal between the closure and the hatch
when the closure is positioned in the hatch for
loading. The top of the closure has openings for
the loading tube and the vapor takeoff line. An
adjustable slipring serves as a positioner enabling
the loading operators to slide the closure to the
proper height on the loading tube for various
depths of tank vehicle compartments. This
closure requires a constant downward force to
keep it in contact with the hatch opening's sides
at all times during filling and is built to fit only
hatches 8 to 10 inches in diameter.
*These markers are gages located within the tank compartment
and positioned at a calibrated volume to indicate visually
the amount of liquid loaded.
The second type of closure, the Greenwood Unit,
(Figures 469 and 470), which also requires a
downward force during the filling operation, was
developed by the Vernon Tool Company. This
closure is also cast aluminum in the shape of
a plug similar to the Mobil Oil Corporation closure
and with a neoprene rubber gasket. This closure
has no auxiliary, transparent, vapor chamber
section, though some versions of this closure do
have auxiliary, metal vapor chambers or a trans-
parent, light well. The closure fits tank truck
compartments with hatches from 8 to 10 inches
in diameter. Since compartments with hatches
of larger diameters are sometimes encountered,
an adapter has been provided. The adapter con-
sists of a flat, gasketed plate with an 8-inch-di-
ameter hole in the center through which the
closure can be inserted.
The third type of vapor closure, referred to as
SOCO, was developed by Standard Oil Company
of California (Figures 471, 472, and 473). It
consists of an aluminum cast plug of more com-
plicated design. This closure is locked into the
hatch opening by a cam lever that forces a float-
ing, internal, cylindrical section to move upward
and squeeze a neoprene rubber collar out against
the sides of the hatch opening, which effects a
vaportight seal during all phases of loading. As
the floating, internal, cylindrical section is
rolled upward by the action of the cam lever de-
vice, it exposes the vapor entry opening, A pis-
ton-type, internal filling valve, similar to an
aircraft-fueling valve, was developed for this
closure. A safety shutoff float operates a needle
-------
Loading Facilities
637
Figure 469. View of the Greenwood vapor closure
(Atlantic-Richfield Oil Corporation, Los Angeles,
Calif.).
Figure 468. View of General Petroleum Corporation
Vapor closure (Mobil Oil Corporation, Los Angeles,
Calif.).
pilot valve that controls the internal valve to pre-
vent overfilling. The cam lever must be released
to remove the vapor closure. The floating cyl-
inder is returned to the closed position at the
same time. Thus, the vapor side is sealed off
to prevent any leakage from the vapor-gathering
lines. At the same time the internal valve is
locked in the closed position. SOCO closures
fit only hatches 8 inches in diameter, though
adapters have been developed for hatches of
greater diameter. This adapter is a circular
casting with an 8-inch opening and is placed over
the hatch opening. When the SOCO unit is insert-
ed, spring-loaded arms act to clamp and seal the
adapter against the top of the hatch.
Figure 470. Closeup view of Greenwood vapor closure
(Atlantic-Richield Oil Corporation, Los Angeles, Calif.).
The Chiksan Company has recently offered a
fourth system, a modern loading arm that in-
corporates the hatch closure, the vapor return
line, and the fill line as an assembled unit (Fig-
ure 474). This unit incorporates features to
prevent overfills, topping off, or filling unless
-------
638
PET-R OLEUM EQUIPMENT
Figure 471. Closeup view of SOCO vapor closure,
withdrawn position (American Airlines, Los Angeles,
Calif.).
Figure 472. Closeup view of SOCO vapor closure,
filling position (American Airlines, Los Angeles,
Calif.).
the assembly is properly seated in the truck
hatch. A pneumatic system ensures contact with
the tank truck as the gasoline is added and pro-
vides a delay at the end of the loading cycle to
achieve adequate drainage of the arm before it is
withdrawn from the truck hatch.
The slide positioner of the Mobil Oil Corporation
vapor closure, though permitting adjustments
for submerged loading, can be a source of
vapor leaks and requires proper attention by the
operator. SOCO closures with inner valves are
considerably heavier than other types, and the
inner valve involves added pressure drops, which
slow the loading rates. Both the Greenwood and
the Mobil Oil Corporation closures require vapor
check valves in the vapor-gathering lines to pre-
vent the vapor from discharging back to the at-
mosphere when the loading assembly is with-
drawn. In addition, inspections have shown that
the Mobil Oil Corporation and Greenwood closures
require nearly vertical entry of the loading tube
into the compartment hatch opening in order to
provide a tight seal against vapor leaks. A
connecting rod between the riser and filling stem
has been added to some assemblies, as shown
in Figure 475, to form a pantograph arrange-
ment to maintain the filling stem of the loading
arm in the vertical position at. all times. The
loading operator is thus able to obtain good seal-
ing contact more quickly bet-ween the vapor col-
lector and the hatch opening.
Collection of Vapors From Bottom Loading
Vapors displaced from tank vehicles during the
bottom-loading operation are more easily col-
lected than those are that result from overhead
loading. The filling line and the vapor collection
line are independent of each other with resultant
simplification of the design (see Figure 476).
The vapor collection line is usually similar to
the loading line, consisting of a flexible hose or
swing-type arm connected to a quick-acting
valve fitting on the dome of the vehicle. This
fitting could be placed at ground level to simplify
the operation further.
A check valve must be installed on the vapor col-
lection line to prevent backflow of vapors to the
atmosphere -when the connection to the tank ve-
hicle is broken.
-------
Loading Facilities
639
DEUD HAN CONTROL
VHVE
VAPOR-oEALINC
RING
FMEPtENt./ OVERFI
SHUT-OFF FLOAT
Figure 473. Schematic drawing of SOCO vapor
closure used to collect displaced vapors
during loading (Standard Oil Company of
California, Western Operations, Inc., Los
Angeles, Calif.).
Factors Affecting Design of Vapor Collection Apparatus
In designing for complete vapor pickup at the
tank vehicle hatch, several factors, including
tank settling, liquid drainage, and topping off
must be considered.
The settling of a tank vehicle due to the weight
of product being added requires that provision
be made for vertical travel of the leading arm
to follow the motion of the vehicle so that the
vapor collector remains sealed in the tank hatch
during the entire loading cycle. Two solutions
to the problem of settling have been used. The
first, applicable to pneumatically operated arms,
includes the continuous application of air pres-
sure to the piston in the air cylinder acting on
the arm. The arm is thus forced to follow the
motion of the vehicle without need for clamping
or fastening the vapor collector to the tank ve-
hicle. The second solution, employed on coun-
terweighted and torsion spring loading arms,
provides for locking the vapor collector to the
tank vehicle hatch. The arm then necessarily
follows the motion of the vehicle. The second
solution is also applicable to vapor collection
arms or hoses that are connected to the top of a
tank vehicle during bottom loading.
The second problem, that of preventing consid-
erable liquid drainage from a loading arm as it
is withdrawn after completion of filling opera-
tions, has been adequately solved. The air valve
that operates the air cylinder of pneumatically
operated loading arms may be modified by addi-
tion of an orifice on the discharge side of the
valve. The orifice allows 30 to 45 seconds to
elapse before the loading assembly clears the
hatch compartment. This time interval is suffi-
r
Figure 474. View of a pneumatically operated
loading assembly with an integrated vapor
closure and return line (Chiksan Co., Brea,
Calif.).
-------
640
PETROLEUM EQUIPMENT
Figure 475. View of a pneumatically operated
loading arm showing pantograph linkage (Atlan-
tic-Richfield Oil Corporation, Los Angeles,
Cal if.).
Figure 476. Bottom loading of tank trucks pro-
vides one way to collect vapor during loading
in conjunction with the use of return line to
storage tanks (Standard Oil Company of Cali-
fornia, Western Operations, Inc., Los Angeles,
Calif.).
cient to permit complete draining of liquid into
tank compartments from arms fitted with loading
valves located in an outboard position. Loading
arms with inboard valves require additional drain-
age time and present the problem of gasoline re-
tention in the horizontal section of the arm. To
prevent drainage the SOCO vapor collection clo-
sure is fitted with an internal shutoff valve that
is closed before the loading arm is withdrawn
from the tank hatch. Providing for thermal ex-
pansion has been found necessary when an in-
board valve and a SOCO vapor closure are used.
This has been accomplished by installing a small
expansion chamber at the normal position of the
loading arm's vacuum breaker. In bottom load-
ing, the valve coupling at the end of the loading
arm or hose, as well as the mating portion of the
valve on the trucks, is self-sealing to prevent
drainage of product when the connection is made
or broken.
The third factor to be considered in the design
of an effective vapor collection system is top-
ping off. Topping off is the term applied to the
loading operation during which the liquid level
is adjusted to the capacity marker inside the
tank vehicle compartment. Since the loading
arm is out of the compartment hatch during the
topping operation, vapor pickup by the collector
is nil. Metering the desired volumes during
loading is one solution to the problem. Metered
loading must, however, be restricted to empty
trucks or to trucks prechecked for loading
volume available. Accuracy of certain totaliz-
ing meters or preset stop meters is satisfactory
for loading without the need for subsequent open
topping. An interlock device for the pneumatic-
type loading arms, consisting of pneumatic con-
trol or mechanical linkage, prevents opening of
the loading valve unless the air cylinder valve
is in the down position. Thus, open topping is
theoretically impossible.
Topping off is not a problem when bottom load-
ing is employed. Metered loading, or installa-
tion of a sensing device in the vehicle compart-
ments that actuates a shutoff valve located either
on the truck or the loading island, eliminates the
need for topping off.
Methods of Vapor Disposal
The methods of disposing of vapors collected
during loading operations may be considered
under three headings: Using the vapors as fuel,
processing the vapors for recovery of hydro-
carbons, or effecting a vapor balance system in
conjunction with submerged loading.
The first method of disposal, using the vapors
directly as fuel, may be employed when the load-
ing facilities are located in or near a facility
that includes fired heaters or boilers. In a typ-
ical disposal system, the displaced vapors flow
through a drip pot to a small vapor holder that
is gas blanketed to prevent forming of explosive
mixtures. The vapors are drawn from the holder
by a compressor and are discharged to the fuel
gas system.
-------
Loading Facilities
641
The second method of disposal uses equipment
designed to recover the hydrocarbon vapors.
Vapors have been successfully absorbed in a
liquid such as gasoline or kerosine. If the loading
facility is located near a refinery or gas absorp-
tion plant, the vapor line can be connected from
the loading facility to an existing vapor recovery
system through a regulator valve.
Vapors are recovered from loading installations
distant from existing processing facilities by
use of package units. One such unit (Figures 477
and 478) that absorbs hydrocarbon vapors in gaso-
line has been developed by the Superior Tank and
Construction Company. This unit includes a va-
porsphere or tank equipped with flexible mem-
brane diaphragm, saturator, absorber, compres-
sor, pumps, and instrumentation. Units are
available to fit any size operation at any desired
loading location since they use the gasoline prod-
uct as the absorbent.
Explosive mixtures must be prevented from ex-
isting in this unit. This is accomplished by pass-
ing the vapors displaced at the loading' rack through
a saturator countercurrently to gasoline pumped
from storage. The saturated vapors then flow to
the vaporsphere. Position of the diaphragm in
the vaporsphere automatically actuates a com-
pressor that draws the vapors from the sphere
and injects them at about 200 psig into the ab-
sorber. Countercurrent flow of stripped gasoline
from the saturator or of fresh gasoline from stor-
age is used to absorb the hydrocarbon vapors.
Gasoline from the absorber bottoms is returned
to storage while the tail gases, essentially air,
are released to the atmosphere through a back-
pressure regulator. Some difficulty has been
experienced -with air entrained or dissolved in
the sponge gasoline returning to storage. Any
air released in the storage tank is discharged to
the atmosphere saturated with hydrocarbon vapors.
A considerable portion of the air can be removed
by flashing the liquid gasoline from the absorber
in one or more additional vessels operating at
successively lower pressures.
Another type of package unit adsorbs the hydro-
carbon vapors on activated carbon, but no in-
stallation of this kind has been observed in Los
Angeles County. The application of this type of
unit is presently restricted to loading installations
Figure 477. View of smalI-capacity vaporsaver gasoline absorption unit
(American Airlines, Los Angeles, Calif.).
-------
642
PETROLEUM EQUIPMENT
TANK GAGE
SWITCH
TO COMPRESSOR
STARTER
TANK
TRUCK
PUMP GASOLINE FEED
TO SATURATION POT
GASOLINE TO LOADING RACK
T>^
LOADING RACK FEED PUMP
Figure 478. Schematic flow diagram of a vaporsaver unit used for recovery
of loading rack vapors at a bulk terminal.
that have low throughputs of gasoline, since the
adsorbing capacity and the life of the carbon are
limited. Units of this type find application in con-
trol of vapors resulting from fueling of jet aircraft.
The vapors displaced during bottom filling are
minimal. Data indicate a volume displacement
ratio of vapor to liquid of nearly 1:1. A closed
system can then be employed by returning all the
displaced vapors to a storage tank. The storage
tank should be connected to a vapor recovery
system.
CATALYST REGENERATION
Modern petroleum processes of cracking, re-
forming, hydrotreating, alkylation, polymeriza-
tion, isomerization, and hydrocracking are com-
mercially feasible because of materials called
catalysts. Catalysts have the ability, when in
contact with a reactant or mixture of reactants,
to accelerate preferentially or retard the rate
of specific reactions and to do this, with few
exceptions, without being chemically altered
themselves. Different catalysts vary in their
effects. One might, for example, increase oxi-
dation rates while another might change the rate
of dehydrogenation or alkylation.
Contact between the catalyst and reactants is
achieved in some processes by passing the reac
tants through fixed beds or layers of catalysts
contained in a reactor vessel. Contact in other
processes involves simultaneous charging of
catalyst and reactants to a reactor vessel and
withdrawal of used catalyst in one stream, and
products and unreacted materials in another
stream. The first process may be termed a
fixed-bed system and the latter a moving-bed
systein. Moving-bed systems may be further
subclassified by the type of catalyst and meth-
od of.transporting it through the process. Ex-
amples are the use of vaporized charge material
to fluidize powdered catalyst, as in fluid catalyt-
ic cracking units (FCC), and the use of bucket
elevators, screws, airlifts, and so forth, to
move the catalyst pellets or beads, as in Thermo-
for catalytic cracking units (TCC) (see Figures
479, 480, and 481).
TYPES OF CATALYSTS
Generally, the catalysts are used in the form of
solids at process temperatures, though some
liquid catalysts are used alone or impregnated
into inert solid carriers. Pellets, beads, and
powders are the common physical shapes. Crack-
ing catalysts are usually beads or powders of
synthetic silica-alumina compositions, includ-
ing acid-treated bentonite clay, Fuller's earth,
aluminum hydrosilicates, and bauxite. Little-
used synthetic catalysts include silica-magnesia,
alumina-boria, and silica-zirconia (Nelson, 1958).
Bead or pelleted catalyst, noted for ease of han-
dling and freedom from plugging, is used in TCC
units while powdered catalyst is used in FCC
units. Natural catalysts are softer and fail more
rapidly at high temperatures than most synthetic
GPO 8OG—614—22
-------
Catalyst Regeneration
643
Figure 479. Simplified flow diagram of a Model IV
fluid catalytic cracking unit (Oil and Gas
Journal, 1957).
catalysts do. The cost of natural catalysts, how-
ever, is under $100 per ton while synthetic types
cost $300 or more per ton.
Catalysts employed in catalytic reforming include
the platinum-containing catalysts used in modern
fixed-bed reformers, except for the bauxite pellet
catalyst for Cycloversion used at 950° to 1,000°F
and 50 to 57 psig, and the molybdena-alumina
catalysts used for fluid hydroformlng. Fixed-bed
reactors operate at 825° to 1,000°F and 200 to
1, 000 psig with catalyst pellets about 1/8 inch in
diameter. These catalysts contain less than 1
percent platinum and are supported on a base of
either alumina or silica-alumina. Acid-type
catalyst required for reforming processes may
be provided by one of the oxides as the catalyst
base. The acid may be a halogen compound add-
Figure 480. Thermofor catalytic cracking unit
(Union Oil Company of California, Los Angeles,
Calif.).
Figure 481. Simplified flow diagram of Thermofor
catalytic cracking unit with modern catalyst air-
lift (Oil and Gas Journal, 1957).
ed to the catalyst, or may be directly added to
the reformer charge. The flow diagram of a
platforming process is shown in Figure 482.
The major desulfurization processes-Autofining,
Dies elf orming, HDS, Hydrofining, Ultrafining,
Unifining, and so forth—employ a cobalt-molyb-
denum catalyst supported on bauxite and operate
within a range of 450° to 850°F and 50 to 1, 500
psig.
Commercial alkylation processes employ as
catalysts either sulfuric acid, hydrogen fluoride,
or aluminum chloride with a hydrogen chloride
promoter.
Commercial polymerization catalysts consist
of a thin film of phosphoric acid on fine-mesh
quartz, copper pyrophosphate, or a calcined
mixture of phosphoric acid.
-------
644
PETROLEUM EQUIPMENT
STABILIZER 6*S |
Figure 482. Simplified flow diagram of platforming
process (Oil and Gas Journal, 1957).
Isomerization processes such as JButamer, Iso-
kel, Isomerate, Penex, and Pentafining employ
a noble metal, usually platinum, as the catalyst
in a hydrogen atmosphere. Liquid-phase iso-
merization is accomplished with aluminum chlo-
ride in molten antimony chloride with a hydrogen
chloride activator.
Loss of Catalyst Activity
The activity of a catalyst, or its effectiveness
in changing rates of specific reactions decreases
with on-stream time. The rate of decrease is
related to composition of reactants contacted,
throughput rate, and operating conditions. Loss
of activity results from metal contamination and
poisoning or deposits that coat the catalyst sur-
faces and thus reduce the catalytic area available
for contact with the reactants. Frequently car-
bon from the coking of organic materials is the
main deposit. To continue in successful opera-
tion, catalyst activity must be restored. One
procedure consists of replacing the spent cata-
lyst with fresh catalyst. A second procedure
consists of treating the spent catalyst for remov-
al of contaminants. This latter procedure, called
catalyst regeneration, is the more significant
from the standpoint of air pollution, since com-
bustion is frequently the method of regeneration.
In fixed-bed systems, catalysts are regenerated
periodically in the reactor or removed and re-
turned to the manufacturer for regeneration. In
moving-bed systems, catalysts are continuously
removed from the reactor, regenerated in a spe-
cial regenerator vessel, and returned to the re-
actor.
REGENERATION PROCESSES
Catalysts for the catalytic cracking and reform-
ing processes are regenerated to restore activity
by burning off the carbon (coke) and other deposits
from the catalyst surface at controlled tempera-
ture and regeneration air rates. Actually, the
so-called "carbon" on the catalyst is not all pure
carbon but contains other compounds. Moreover,
the catalyst is not entirely freed of the carbon
deposits during regeneration, though an effort is
made to keep the residual carbon below 0. 9 per-
cent by weight on the regenerated catalyst. FCC
units, all of which have continuous catalyst regen-
eration, have a coke burnoff rate 5 to 10 times
higher than TCC unit regenerators have. Since
fixed-bed reformer units, which incorporate cata-
lyst regeneration, have a very small coke laydown
on the catalyst surface, they require regeneration
only once or twice a year, as the desulfurizer
reactors do, which have both a coke and sulfur
laydown.
FCC Catalyst Regenerators
Catalyst regenerators for FCC units may be
located alongside, above, or below the reactor.
Regenerators normally have a vertical, cylin-
drical shape with a domed top. The inside shell
of the regenerator is insulated with 4 to 6 inches
of refractory lining. This lining may also be ex-
tended into the regenerator's discharge line and
the regenerator's catalyst charge line.
The upper section of the regenerator is equipped
with internal cyclone separators to separate the
catalyst dust from the regeneration "combustion
gases. The number of cyclone separators varies
from a single-stage or two-stage separator to
as many as 12 sets of three-stage cyclone sepa-
rators. External size of the regenerator varies
from 20 feet in diameter by 40 feet high to 50
feet in diameter by 85 feet high. In Los Angeles
County, regenerator flue gases pass through
additional equipment, consisting of electrical
precipitators or cyclone separators and elec-
trical precipitators for final dust removal, be-
fore discharging to the atmosphere. Carbon
monoxide waste-heat boilers are employed be-
fore or following the electrical precipitators.
In a typical FCC unit, as shown in Figure 479,
the spent catalyst from the base of the reactor
is steam stripped to remove residual hydrocar-
bons and then transferred to the regenerator by
injecting preheated air into the transfer line.
Burning off of the carbon starts when the hot,
spent catalyst contacts the air, and continues as
the catalyst flows up through the regenerator to
the overflow well. Additional combustion air is
furnished by the main blower. The amount of air
supplied is controlled to prevent glazing the cata-
lyst. This results in the formation of consider-
able amounts of carbon monoxide. The depth of
the fluidized catalyst bed is usually limited to
-------
Catalyst Regeneration
645
15 feet to prevent the load on the cyclones from
being excessive. Regenerated catalyst flows
down through the overflow well to the reactor
as a result of a slight pressure differential.
The flue gases pass through the regenerator's
cyclone separators, for removal of most of the
catalyst more than 10 microns in size; through
a steam generator, where process steam is
made; through a pressure-reducing chamber to
air pollution control units; and then to the atmo-
sphere. The pressure-reducing chamber serves
as a noise suppressor. Final dust cleanup is
accomplished by passing the effluent gases from
the cyclone separators through an electric pre-
cipitator. The gases from the precipitator are
introduced into a carbon monoxide boiler where
the sensible heat and the heat content of the CO
is used to produce steam in some flow schemes.
Other operations place the waste-heat boiler be-
fore the precipitator.
According to Brown and Wainwright (1952), the
weight of dust per cubic foot of exit gas remains
constant at about 0. 002 pound at bed velocities
up to a critical velocity of 1. 5 fps, whereupon
it rises rapidly with higher velocities, for exam-
ple, to 0. 01 pound at 1. 8 fps. The pressures in
FCC unit regenerators are always low, between
1 and 10 psig. Regeneration temperatures are
usually between 1, 050° andl,150°F. Other
general operating data for large and small FCC
unit regenerators are as follows:
Catalyst circulation
rate, tons/min
Coke burnoff rate,
Ib/hr
Small unit Large unit
10
60
5,000 34,000
Regeneration air rate,
scfm 13,000 102, 000 •
TCC Catalyst Regenerators
TCC (and Houdry unit) catalyst regenerators,
referred to as kilns, are usually vertical struc-
tures with horizontal, rectangular, or square
cross sections. A regenerator that has a cata-
lyst circulation rate of 150 tons per hour would
have an outside dimension of about 11 feet square
by 43 feet high. This size regenerator, or kiln,
has approximately 10 regeneration zones and a
topside kiln hopper. Each zone is equipped with
a flue gas duct, air distributors, and steam- or
water-cooling coils. The carbon steel shell of
the regenerator is lined with about 4 inches of
firebrick, which is, in turn, covered with alloy
steel. The discharge flue gases from the regen-
eration kilns are usually vented through dry^type,
centrifugal dust collectors.
In a TCC unit, Figure 480, spent catalyst (beads)
from the base of the reactor is steam purged for
removal of hydrocarbons and lifted by a bucket
elevator to a hopper above the regeneration kiln.
Catalyst fines at this point in the process are
separated from catalyst beads in an elutriator
vessel using up-flowing gases and are collected
from these gases in a cyclone separator dis-
charging to a fines bin. Spent catalyst beads
drop through a series of combustion zones, each
of which contains flue gas collectors, combustion
air distributors, and cooling coils. The cata-
lyst is regenerated as it flows downward through
the kiln zones counter cur rent to preheated air
(400° to 900°F). The pressure is essentially
atmospheric in the kiln. Water is circulated
through cooling coils in each kiln zone to control
the rate of coke combustion. The regeneration
temperatures at the top of the kiln are between
800° and 900°F, while the bottom section of the
kiln operates between 1, 000" and 1, 100°F. A
minimum temperature of 900°F is required for
catalyst regeneration. An average-size TCC unit
regenerator with a catalyst circulation of 2. 5 tons
per minute has a coke burnoff rate of 3, 500 pounds
per hour and a regeneration air rate of 24, 000
scfm.
Regenerated catalyst from the bottom of the kiln
is then transferred by bucket elevator to the cata-
lyst bin for reuse in the reactor. The more
modern TCC units use a catalyst airlift (Figure
481) rather than bucket elevators for returning
regenerated catalyst to the reactor, and gravity
flow for moving spent catalyst to the regenerator.
The elevators of those units must be vented
through wet centrifugal collectors or scrubbers
to the atmosphere.
Catalyst Regeneration in Catalytic Reformer
Units
Some types of catalytic reformer units are
shut down once or twice each year for re-
generation of the catalyst in the desulfurizer re-
actor. Reforming units using Sinclair-Baker
catalyst are in this category. Before the regen-
erating, the reformer system is depressured,
first to the fuel gas system and then to vapor re-
covery. A steam jet discharging to vapor recov-
ery is then used to evacuate the reformer furth'er
to 100 millimeters of mercury absolute pressure.
An inert gas such as nitrogen is introduced to
purge and then repressure the system to 50 psig.
The nitrogen is circulated by the recycle gas
compressor through the heaters, reactors, heat
exchangers, flash drum, and regeneration gas
drier. Inert gas circulation is continued while
combustion air for burning off the coke is intro-
duced into the top of the first reactor by the re-
generation air compressor. The rate of air is
controlled to maintain catalyst bed temperatures
below 850°F. Pressure is controlled to 150 psig
by releasing products of combustion to the fire-
-------
646
PETROLEUM EQUIPMENT
box of the reformer heater. After burning is
completed tin the first reactor, as indicated by
the rise in oxygen content in the effluent, the air
supply is then switched to the second reactor.
The same procedure is repeated for the other
reactors.
In the regeneration cycle, circulation of approx-
imately 15, 000 scfm flue gas is maintained by
using the reformer recycle gas compressor, and
approximately 500 scfm regeneration air is added
for burning off the coke. About 24 to 30 hours
is required for regeneration, based upon a coke
content of 5 percent by weight in the catalyst. The
coke may run about 90 percent carbon and 10 per-
cent hydrogen.
Desulfurization reactors are depressured in the
same manner as the catalytic reformer described.
During catalyst regeneration, however, super-
heated steam is added along with inert air con-
taining about 1. 4 mol percent oxygen to effect
temperature control. In addition to coke, there
are also sulfur deposits that are burned to sulfur
dioxide. In some installations the regeneration
gases are passed through packed scrubbers that
use water or caustic for partial absorption of
sulfur dioxide. These reactors are also regen-
erated for a period of approximately 24 hours
about once or twice a year.
Regeneration of fluid hydroforming catalyst, a
white powder consisting of molybdena-coated
alumina, is accomplished by continuously with-
drawing a portion of the catalyst recirculating
in the reactor and burning the carbon off in a
separate regenerator using fresh air with no pre-
heat. The regeneration temperature is 1, 100°
to 1, 150°F at 200 to 250 psig with 100 percent
carbon removal. Molybdenum sulfide, formed by
the reaction of catalyst molybdenum oxide and
feed stock sulfur, is reoxidized to molybdenum
oxide with the release of sulfur dioxide during
regeneration.
In alkylation units using hydrogen fluoride as
catalyst, the acid strength is restored by remov-
ing the water of dilution by distillation. The ef-
fectiveness of alkylation units using sulfuric acid
as the catalyst is maintained by adding fresh acid
as spent acid is withdrawn. The spent acid may
be reconcentrated or used as is for other purposes.
Phosphoric acid catalyst used in polymerization
units is regenerated by water washing, steaming,
and drying the fine-mesh quartz carrier, and
adding fresh phosphoric acid. After the excess acid
is drained, the reactor is ready to go back on
stream.
Many of the remaining catalytic processes re-
quire only infrequent catalyst replacement or
regeneration (Unicracking and Isomax). In the
H-Oil process, however, catalyst is continuously
replaced.
THE AIR POLLUTION PROBLEM
Air contaminants are invariably released to the
atmosphere from regeneration operations,
especially from operations involving combus-
tion. The variety of air contaminants released
is broad and may include catalyst dust and other
particulate matter, oil mists, hydrocarbons,
ammonia, sulfur oxides, chlorides, cyanides,
nitrogen oxides, carbon monoxide, and aerosols.
The contaminants evolved by any one type of re-
generator are a function of the compositions of
the catalyst and reactant, and operating conditions.
Tables 174 through 179 show stack emissions
for regeneration of both FCC and TCC units.
The data in these tables are the results of a
testing program (Sussman, 1957) to establish
the magnitude of the listed components in the
catalyst regeneration gases.
The largest quantities of air pollution from, cat-
alyst-regenerating operations are experienced in
FCC units. The pollutants include carbon mon-
oxide, hydrocarbons, catalyst fines dust, oxides
of nitrogen and sulfur, ammonia, aldehydes, and
cyanide. Typical losses from fluid catalytic crack
ing regenerators, based upon Tables 175 through
178, include:
Loss to
Pollutant atmosphere, Ib/hr
Carbon monoxide 24, 300
Sulfur dioxide 545
Hydrocarbons 231
NO as nitrogen dioxide 80. 2
X
Particulate matter 65. 5
Ammonia 57. 4
Sulfur trioxide 32.7
Aldehydes as formaldehyde 21.6
Cyanides as hydrogen cyanide 0. 27
TCC catalyst regeneration produces air contami-
nants similar to those from FCC catalyst regen-
eration. Quantities produced, however, are con-
siderably less, as can be seen from Tables 175
through 178. The bead-type catalyst used in TCC
units does not result in the large amount of cata-
lyst fines that are encountered in FCC units.
Air pollution problems are not as severe from
catalyst regeneration of reforming and desulfuriza-
tion reactors as those from FCC and TCC units.
These reactors are regenerated only once or
twice a year for a period of about 24 hours. The
burning-off of the coke and sulfur deposits on the
-------
Catalyst Regeneration
647
Table 174. OPERATING CHARACTERISTICS OF FLUID AND THERMOFOR
CATALYTIC CRACKING UNITS (Sussman, 1957)
Typea
FCC
FCC
FCC
FCC
TCC
TCC
FCC
TCC
TCC
TCC
TCC
TCC
TCC
FCC
TCC
Date
tested,
1956
10/4
12/4
8/30
11/27
11/1
11/1
10/9
10/18
10/18
9/19
9/19
9/12
9/12
11/8
12/19
Feed rate
Fresh,
bpd
40,000
29,500
24, 000
32, 610
9,525
8,525
25,000
10, 000
8, 000
7,071
6,506
7,099
6, 053
6,462
8,000
Recycle,
bpd
10,000
2,045
0
13,680
1,500
7,400
9,000
0
3,000
5,538
5,602
6,004
6,013
606
3,000
Catalyst
circulation
rate,tons/hr
4,500
1,560
1,380
2,532
180
150
3,240
165
150
150
150
150
120
390
200
Regenerator
air rate,
scfm
112,000
28,000
22,200
97,500
27,000
27, 000
64, 000
22, 000
27,600
24,000
25, 000
27,000
23,000
13,300
16,800
Coke burn-
off rate,
Ib/hr
38,000
23,000
21, 300
36,416
4,715
2,610
21, 600
5,655
4, 620
4,410
5,020
3,420
3,000
5,400
3, 760
Avg gas
temp,
°F
820
510
520
485
840
700
530
660
610
850
740
810
710
610
680
aAll fluid catalytic cracking units
all Thermofor catalytic cracking
are equipped with electrical precipitators;
units are equipped with cyclone collectors.
Table 175. PARTICULATE LOSS FROM
FLUID AND THERMOFOR CATALYTIC
CRACKING UNIT STACKS
(Sussman, 1957)
Type
Total particulate, a
Ib/hr
FCC
FCC
FCC
FCC
TCC
TCC
FCC
TCC
TCC
TCC
TCC
TCC
TCC
FCC
TCC
57. 50
61. 00
181. 00
58.70
1.36
1.64
28.30
13.80
8.06
3.44
2.22
9.53
10. 01
6.42
4.30
aThe total particulate loss includes
weight of insoluble solids in the water,
and HCL impinger solution added to the
alundum thimble catch.
AIR POLLUTION CONTROL EQUIPMENT
Dust from FCC catalyst regenerators is collected
by centrifugal collectors or centrifugal collectors
and electrical precipitators. General design fea-
tures of these collectors are discussed in Chapter
4. Carbon monoxide waste-heat boilers eliminate
carbon monoxide and hydrocarbon emissions in
FCC regeneration gases. Dry-type, centrifugal
dust collectors are used to collect the catalyst
fines from TCC regeneration gas. Dust emis-
sions from TCC unit reactors and regenerator
catalyst elevators can be adequately controlled
by wet- or dry-type, centrifugal collectors.
Presently, no TCC units are equipped \vith car-
bon monoxide waste-heat boilers. Manifolding
several TCC units could possibly result in a
quantity of flue gas large enough to improve
economic justification for a CO boiler.
The carbon monoxide and hydrocarbons in re-
forming and desulfurization catalyst regeneration
gases can be efficiently incinerated by passing
the regeneration gases through a heater firebox.
In some installations the sulfur dioxide is scrubbed
bypassing the regeneration gases through a packed
water or caustic tower.
catalyst surface produces hydrocarbons, sulfur
dioxide, and carbon monoxide, in addition to
carbon dioxide and water.
Wet- and Dry-Type, Centrifugal Dust Collectors
Cyclone separators are widely used for catalyst
dust collection systems in refineries. They are
located in the upper sections of both FCC unit
-------
643
PETROLEUM EQUIPMENT
Table 176. TOTAL HYDROCARBON EMISSIONS FROM FLUID AND THERMOFOR CATALYTIC
CRACKING UNIT STACKSa (Sussman, 1957)
Type
FCC
FCC
FCC
FCCC
TCCe
TCCe
FCCd-e
TCCd
TCCd
TCCb'c
r^ ^x-sb, C
TCC '
TCC
TCC
FCC
TCC
Mass spectrometer
Hydrocarbons
7.4
3. 1
2. 1
1
_
_
-
0. 4
0. 5
0. 1
0. 5
0. 3
1.4
Hydrocarbons
1,213
1, ISO
760
98
_.
_
_,
308
4,484
87
121
328
1,655
Wt % C and C
67.7
84. 1
68. 3
42. 3
_
_
_
40. 9
55. 1
79.5
67.4
51.2
61.9
Vol % C and C
87. 4
94.6
85. 5
54. 1
_
_
„
70.8
81.4
77
67. 8
75. 3
18. 8
Infrared spectrophotometer
Hydrocarbons
(as hexane),
tons /day
2. 80
0. 89
0. 60
0. 30
0. 02
0. 02h
1. 20
0. 04
0. 15
g
0.02
f
0. 01
-
0. 30
Hydrocarbons
(as hexane),
ppm
142
78
65
12
8
116
13
43
_
14
_
9
Trace
108
aAll concentrations are reported on a dry basis.
bOnly the mass spectrometer results for Units F-2T and F-4T were reliable. Since Units F-1T and F-2T
and Units F-3T and F-4T are twin units, the data shown result from combining the twin units.
cNo methane present as determined by mass spectrometer.
Mass spectrometer determinations include oxygenated C^ and Cg hydrocarbons.
eThe mass spectrometer results were not reliable.
*The infrared spectrophotometer results were not reliable.
^Concentrations of hydrocarbons are below limit of accuracy of the infrared spectrophotometer.
Infrared spectrophotometric determinations were made on Unit D-1T only. The results shown were
obtained by assuming that twin Unit D-1T and D-2T emit the same quantity of hydrocarbons.
reactors and regenerators for collecting en-
trained catalyst. Some TCC units also use cy-
clones for catalyst fines collection from kiln re-
generation gases. The cyclones are employed
as a single unit or in multiple two-stage or three-
stage series systems. Large FCC unit regen-
erators may have as many as 12 three-stage cy-
clones, while smaller units may have only 1 two-
stage cyclone. In general, high-efficiency cy-
clones have dust collection efficiencies of over
90 percent for particle sizes of more than 15
microns. The efficiency drops off rapidly for
particles of less than 10 microns.
Multiple cyclones are used in some cases for
catalyst fines collection catalyst regeneration
gases in TCC units. Dust collection efficiencies
are in the same range as those for high-efficien-
cy cyclones. Wet-type, centrifugal collectors or
scrubbers adequately clean the gas streams from
the catalyst elevators, and part of the regenera-
tion gases from the kilns. Untreated water in the
wet collector, however, can cause a carbonate
deposit on the impeller, which is responsible
for excessive wear on the collector bearings.
This can and has resulted in excessive shutdown
time for repairs. Table 180 shows particulate
emissions from two wet-type, centrifugal cata-
lyst dust collectors.
tlectri cal Precipitators
Many FCC units incorporate electrical precipita-
tors for final collection of cs.talyst dust from
catalyst regeneration gases. Electrical precipi-
tators (see Figure 483) are rormally installed in
parallel systems because of the large volume of
regeneration gases involved in FCC unit regerj-
eratois. Power requirements for these precip-
itators may range from 35 kva for small FCC
units to 140 kva for the larger installations. The
hot gases from the regenerator must be cooled
from approximately 1, 100" to below 500°F be-
fore entering the precipitator. This is accom-
-------
Catalyst Regeneration
649
Table 177. EMISSIONS OF SULFUR OXIDES, AMMONIA, AND CYANIDES FROM STACKS OF
FLUID AND THERMOFOR CATALYTIC CRACKING UNITSa (Sussman, 1957)
Type
FCC
FCC
FCC
FCC
TCC
TCC
FCC
TCC
TCC
TCC
TCC
TCC
TCC
FCC
TCC
so3
Ib/hr
164
12.0
1.20
8. 90
1.25
-
6. 90
5. 10
2. 0
1. 60
2. 70
5. 74
7. 77
3. 07
0.62
S02>
Chemical anal.
Ib/hr
535
362
1,260
453
17.5
-
648
15. 1
14. 0
18.7
13.2
13.0
11. 1
205
24.4
ppm
438
512
2, 190
308
114
-
984
86
65
151
136
105
97
1, 310
141
MS,b
ppm
47
220
1,850
20
-
-
-
15
10
-
91
-
60
360
15
Totals
as SO2,
vol %
0. 055
0.540
0.220
0.031
0. Oil
-
0.098
0.011
0.008
0.016
0.016
0. 015
0.015
0. 130
0. 014
Wt % SO3
in total
oxides
of sulfur
23.5
3. 2
0. 1
1.8
6.7
-
1. 1
25. 0
13. 0
7.9
17.0
30.6
41.2
1.4
2.5
NH3,
Ib/hr
130
27.0
20.5
26. 0
1.20
-
118
4. 60
3.40
2.20
1. 90
1.56
3. 12
23.0
2.80
ppm
401
140
134
67
29
-
675
99
60
67
74
47
103
550
61
Cyanides as HCN,
Chemical anal.
Ib/hr
0.250
0.280
Trace
0.291
0.010
-
0. 054
0. 005
0.060
Trace
Trace
Trace
Trace
0. 018
0.039
ppm
0.48
0.94
Trace
0. 47
0. 15
-
0. 19
0.07
0.70
Trace
Trace
Trace
Trace
0.27
0.54
MS,b
ppm
430
360
240
170
-
-
-
370
230
-
90
-
180
190
220
aAll concentrations are reported on a dry basis.
MS = mass spectrophotometer.
Table 178. EMISSIONS OF ALDEHYDES, OXIDES OF NITROGEN, CO2, O2, CO, AND N2 FROM
STACKS OF FLUID AND THERMOFOR CATALYTIC CRACKING UNITSa (Sussman, 1957)
Type
FCC
FCC
FCC
FCC
TCC
TCC
FCC
TCC
TCC
TCC
TCC
TCC
TCC
FCC
TCC
Aldehydes as HCHO,
Ib/hr
77.0
18.0
25.9
4.0
3.5
-
0.9
2. 2
1.2
0.6
0. 4
2.6
3. 4
1.5
14. 3
ppm
130
53
96
5
49
-
3
26
12
12
9
44
63
20
177
NOX as NO2,
Ib/hr
26.0
4.2
163
202
5.7
-
5.9
0
0
3. 1
2.2
2. 7
0.6
-
7. 7
ppm
29
8
394
191
51
-
12
0
0
34
32
30
7
-
62
NO by
MS,
ppm
250
-
160
11
-
-
-
200
170
-
190
-
130
310
230
CO2, vol %,
ORSAT
8. 7
8.5
10. 0
13.4
8.2
-
9.5
9.2
4. 7
9.6
12.8
8.4
8.8
7.8
9.0
MS
11. 1
8.8
11.8
13.4
-
-
_
12. 1
9.0
-
13.3
-
9.2
7.8
9.0
O2, vol %,
ORSAT
5. 1
3.5
2. 3
2.0
7.9
-
2. 7
6.6
13. 5
8.3
2.5
9.8
7.8
5. 1
6.9
MS
2.2
4. 1
2. 3
2.3
-
-
_
-
_
-
2.5
-
11. 1
5. 5
7.3
CO, vol %,
ORSAT
4.9
7.8
6. 1
0
1.4
-
6.8
3.2
0. 7
1.5
3.6
0
2.6
6. 1
4. 1
N2, vol %
by diff,
ORSAT
81. 0
80.2
81.6
84.6
82.5
_
81. 0
81. 0
81. 1
80.6
81. 1
81.6
80.8
81. 0
80. 0
All concentrations are reported on a dry basis.
plished by a waste-heat boiler. The electrical
conductivity of the gas stream may be increased
by injecting ammonia upstream of the precipitator.
The inlet ducting is designed to effect a uniform
gas distribution through the precipitator cross
section. A perforated-plate inlet or vane sec-
tion assists in accomplishing the desired dis-
tribution.
The precipitators usually employ either a con-
tinuous-type electrode-rapping and plate-vibrating
sequence or an intermittent hourly rapping cycle.
A dust plume up to 90 percent opacity arises for
a period of 1 to 2 minutes from the precipita-
tor's discharge stack during the intermittent hour-
ly rapping cycle. This high-opacity, short-in-
terval plume is not normally encountered with
the continuous rapping sequence.
-------
650
PETROLEUM EQUIPMENT
Table 179. MOISTURE AND FLUE GAS VOLUMES, %, FROM
STACKS OF FLUID AND THERMOFOR CATALYTIC CRACKING UNITS
(Sussman, 1957)
Type
FCC
FCC
FCC
FCC
TCC
TCC
FCC
TCC
TCC
TCC
TCC
TCC
TCC
FCC
TCC
Vol % H2O
as determined
from sampling
trains
19.7
19.2
26.3
18. 7
12. 1
-
18
16.5
11. 1
12.2
19
11
11
25. 3
7. 5
Vol % H2O
in
MSa sample
0.480
0.470
0. 186
0.229
-
-
-
0.626
2.448
-
0.885
-
0.600
0.458
1.762
Rate of flow of
flue gases (wet
basis), scfm
151,000
86,300
77,200
178, 800
17,300
20, 800
80,900
20,700
23,600
13,970
11,660
13,800
12,700
20, 800
18,400
Rate of flow of
flue gases (dry
basis), scfm
121,300
69,700
56, 900
145,400
15,200
-
65, 000
17,280
20,980
12.Z70
9,600
12, 300
11,300
15,540
17,000
MS = mass spectrophotometer.
Figure 483. Top of fluid catalytic cracking
unit's Cottrell precipitator. Electrode
terminals and 36-inch-diameter flue gasline
between precipitator and silencer are shown
(Union Oil Company of California, Los Angeles,
Calif.).
Carbon Monoxide Waste-Heat Boilers
Large amounts of carbon monoxide gases are
discharged to the atmosphere with the regenera-
tion flue gases of an FCC unit. The carbon mon-
oxide waste-heat boiler is a means of using the
heat of combustion of carbon monoxide and other
combustible, and the sensible heat of the regen-
eration gases. From the air pollution viewpoint,
the CO boiler oxidizes the carbon monoxide and
other combustibles, mainly hydrocarbons, to
ca.rbon dioxide and water.
In most cases, auxiliary fuel is required in addi-
tion to the carbon monoxide and may be either
fuel oil, refinery process gas, or natural gas.
The CO boiler may be a vertical structure with
either a rectangular or circular cross section
with water-cooled walls, as shown in Figure 484.
The outer dimensions of a typical rectangular
boiler are 32 feet wide by 44 feet deep by 64 feet
high, with a 200-foot-high stack. The boiler is
equipped with a forced-draft fan and four sets of
fixed, tangential-type burners (one set for each
corner). A typical set of burners includes two
carbon monoxide gas compartments, four fuel
gas nozzles, and two steam-atomized oil burners,
as shown in Figure 485. The burners are approxi-
mately 1-1/2 feet wide by 6 feet high. A tangen-
tial-type mixing of the gases for more nearly
complete combustion is achieved by arranging
the burners slightly off center.
-------
Catalyst Regeneration
651
Table 180. EMISSIONS FROM WET-TYPE, CENTRIFUGAL CATALYST
DUST COLLECTORS (THERMOFOR CATALYTIC CRACKING UNIT)a
Inlet gas volume, scfm
Inlet gas temperature, °F
Inlet gas I-^O content, vol %
Particulate matter, Ib/hr
Outlet gas volume, scfm
Outlet gas temperature, °F
Outlet gas H2O content, vol %
Particulate loss, Ib/hr
Collection efficiency, %
Collector No. 1
with two inlet streams
TCC No. 1
1, 780
720
38.8
31.7
TCC No. 2
2,090
690
39.3
40. 1
Collector No. 1 discharge
4, 230
210
41.2
10.2
85.8
Collector No. 2
with two inlet streams
TCC No. 3
2, 350
740
27.6
23.2
TCC No. 4
1, 680
650
22. 1
52.0
Collector No. 2 discharge
5,090
210
30. 4
8.6
88.6
aThe inlet of each collector is connected by ductwork to the reactor elevator and
the Thermofor kiln of two Thermofor catalytic cracking units.
Figure 484. Cylindrical, water-cooled, carbon
monoxide waste-heat boiler (Combustion En-
gineering, Inc., Windsor, Conn.).
elude oxidation of the sulfur compounds in the fuel
oil or refinery gas to sulfur dioxide. The small
amount of ammonia in the regeneration flue gas
is primarily converted to oxides of nitrogen at
the firebox temperature of between 1, 800° and
2,000°F, Table 181 shows the emissions from
an FCC unit's CO boiler.
Economic Considerations
The economics of a CO boiler installation can
be generalized sufficiently to determine a range
of catalytic cracking unit sizes that can pay out
a boiler (Alexander and Bradley, 1958). The
main variable used in determining the size of the
catalytic cracking unit is coke-burning rate.
Other variables that affect payout include the
following in the order of decreasing importance:
(1) Fuel value, (2) CO2/CO ratio, (3) flue gas
temperatures, (4) excess oxygen in CO gas,
(5) hydrogen content of regenerator coke.
Regeneration gases from the FCC unit are nor-
mally delivered to the inlet of the CO boiler
ductwork at about 1, 100°F and 2 psig. When-
ever the overhead regenerator gases first pass
through an electrical precipitator, the inlet gas
to the precipitator must be cooled below 500°F.
The CO boiler would then receive regeneration
flue gas between 450° and 500°F.
The main reactions of the CO boiler's firebox in-
clude burning the refinery gas or fuel oil to car-
bon dioxide and water and completing the oxida-
tion of the carbon monoxide. Other reactions in-
On the assumption that additional steam is re-
quired in the refinery, a coke burnoff rate of
10, 000 pounds per hour or more can be econom-
ically attractive for installation of a CO boiler
when the fuel has a value of 20 cents per million
Btu. If, however, additional steam is not re-
quired, the minimum coke-burning rate to pro-
vide a reasonable payout for a CO boiler is
about 18, 000 pounds per hour. A payout of 6
years after taxes is assumed to be an attractive
investment. In some areas, the reduction in
the air contaminants is sufficiently important to
justify a payout longer than 6 years.
-------
652
PETROLEUM EQUIPMENT
Figure 485. Corner-fired burners of a carbon monoxide waste-heat
boiler: (left) Elevation view showing a typical set of burners
for one corner; (right) plan view of firebox showing location of
the four sets of burners (Combustion Engineering, Inc., Windsor,
Conn.).
Table 181. EMISSIONS FROM THE STACKS
OF FLUID CATALYTIC CRACKING UNITS'
CARBON MONOXIDE WASTE-
HEAT BOILERSa
Gas volume, scfm
Gas temperature, °F
Dust loss, Ib/hr
NOX as NO2, ppm
Aldehydes as HCHO, ppm
NH3, Ib/hr
SO2, Ib/hr
SO3> Ib/hr
Organic acids as acetic, ppm
Hydrocarbons as C^, ppm
CO2, vol % dry basis
CO, vol % dry basis
O, vol % dry basis
H2O, vol %
Unit I
East
stack
96,800
470
44
173
15
19.8
269
0. 16
-
None
14
0
3
22.4
West
stack
97,200
450
33
190
11
22.5
282
0.4
-
None
14. 4
0
2.6
22. 7
Unit II
60,700
570
34.9
67
5
Noneb
265
1.61
11. 7
< 8
8.8
0
3.5
23.9
aBoth FCC Units I and II are equipped with electrostatic
precipitators.
FCC Unit II does not use NH, injection for precipitator
conditioning.
OIL-WATER EFFLUENT SYSTEMS
FUNCTIONS OF SYSTEMS
Oil-water effluent systems are found in the three
phases of the petroleum industry--production,
refining, and marketing. The systems vary in
size and complexity though their basic function
remains the same, that is, to collect and sep-
arate wastes, to recover valuable oils, and to
remove undesirable contaminants before dis-
charge of the water to ocean, rivers, or channels.
Handling of Crude-Oil Production Effluents
In the production of crude oil, wastes such as
oily brine, drilling muds, tank bottoms, and
free oil are generated. Of these, the oilfield
brines present the most difficult disposal prob-
lem because of the large volume encountered
-------
Oil-Water Effluent Systems
653
(Rudolfs, 1953). Community disposal facilities
capable of processing the brines to meet local
•water pollution standards are often set up to
handle the treatment of brines. The most effec-
tive method of disposal of brines has been in-
jection into underground formations.
A typical collection system associated with the
crude-oil production phase of the industry usu-
ally includes a number of small gathering lines
or channels transmitting waste -water from wash
tanks, leaky equipment, and treaters to an earthen
pit, a concrete-lined sump, or a steel waste-water
tank. A pump decants waste water from these
containers to water-treating facilities before in-
jection into underground formations or disposal
to sewer systems. Any oil accumulating on the
surface of the water is skimmed off to storage
tanks.
Handling of Refinery Effluents
The effluent disposal systems found in refineries
are larger and more elaborate than those in the
production phase. A typical modern refinery
gathering system usually includes gathering
lines, drain seals, junction boxes, and channels
of vitrified clay or concrete for transmitting
waste water from processing units to large
basins or ponds used as oil-water separators.
These basins are sized to receive all effluent
water, sometimes even including rain runoff,
and may be earthen pits, concrete-lined basins,
or steel tanks.
Liquid wastes discharging to these systems
originate at a wide variety of sources such as
pump glands, accumulators, spills, cleanouts,
sampling lines, relief valves, and many others.
The types of liquid wastes may be classified as
waste water with:
1. Oil present as free oil, emulsified oil, or
as oil coating on suspended matter;
2. chemicals present as suspensoids, emulsoids,
or solutes. These chemicals include acids,
alkalies, phenols, sulfur compounds, clay,
and others.
Emissions from these varied liquid wastes can
best be controlled by properly maintaining, iso-
lating, and treating the wastes at their source; by
using efficient oil-water separators; and by
minimizing the formation of emulsions. Recov-
ery of some of the wastes as valuable byproducts
is growing in importance.
Treatment of Effluents by Oil-Water Separators
The waste water from the process facilities and
treating units just discussed flows to the oil-
water separator for recovery of free oil and
settleable solids.
The American Petroleum Institute is recognized
as an authoritative source of information on the
design of oil-water separators, and its recom-
mended methods are used generally by refineries
in Los Angeles County. The basis for design
of a separator is the difference in gravity of oil
and water. A drawing of a typical separator is
shown in Figure 486.
Factors affecting the efficiency of separation
include temperature of -water, particle size,
density, and amounts and characteristics of sus-
pended matter. Stable emulsions are not affected
by gravity-type separators and must be treated
separately.
The oil-water separator design must provide for
efficient inlet and outlet construction, sediment
collection mechanisms, and oil skimmers. Re-
inforced concrete construction has been found
most desirable for reasons of economy, mainte-
nance, and efficiency.
Clarification of Final-Effluent Water Streams
The effluent -water from the oil-water separator
may require further treatment before final dis-
charge to municipal sewer systems, channels,
rivers, or streams. The type and extent of
treatment depend upon the nature of the contami-
nants present, and on the local water pollution
ordinances governing the concentration and
amounts of contaminants to be discharged in re-
finery effluent waters. The methods of final-ef-
fluent clarification to be briefly discussed here
include (1) filtration, (2) chemical flocculation,
and (3) biological treatment.
Several different types of filters may be used to
clarify the separator effluent. Hay-type filters,
sand filters, and vacuum precoat filters are the
most common. The selection of any one type de-
pends upon the properties of the effluent stream
and upon economic considerations.
The application of chemical flocculation to the
treatment of separator effluent water is a rela-
tively recent development (Reno et al. , 1958;
Castler et al. , 1956). Methods of treatment are
either by sedimentation or flotation. In sedimen-
tation processes, chemicals such as copper sul-
fate, activated silica, alum, and lime are added
to the waste-water stream before it is fed to the
clarifiers. The chemicals cause the suspended
particles to agglomerate and settle out. Sedi-
ment is removed from the bottom of the ciarifiers
by mechanical scrapers.
-------
654
PETROLEUM EQUIPMENT
PUN
TO «ATER DISPOSAL
TRANSVERSE OPENINGS
ELEVATION
Figure 486. A modern oil-water separator.
Effectiveness of the sedimentation techniques in
the treatment of separator effluents is limited by
the small oil particles contained in the waste
water. These particles, being lighter than water,
do not settle out easily. They may also become
attached to particles of suspended solids and
thereby increase in buoyancy.
In the flotation process a colloidal floe and air
under pressure are injected into the waste water.
The stream is then fed to a clarifier through a
backpressure valve that reduces the pressure to
atmospheric. The dissolved air is suddenly re-
leased in the form of tiny bubbles that carry the
particles of oil and coalesced solids to the surface
where they are skimmed off by mechanical flight
scrapers. Of the two, the flotation process has
the potential to become the more efficient and
economical.
Biological treating units such as trickling filters
activated-sludge basins, and stabilization basins
have been incorporated into modern refinery
waste disposal systems. By combining adsorp-
tion and oxidation, these units are capable of re-
ducing oil, biological oxygen demand, and pheno-
lic content from effluent water streams. To pre-
vent the release of air pollutants to the atmosphere,
certain pieces of equipment, such as clarifiers, di-
gesters, and filters, used in biological treatment
should be covered and vented to recovery facili-
ties or incinerated.
Effluent Wastes From Marketing Operations
In the marketing and transportation phase of the
industry, waste water containing oil maybe dis-
charged during the cleaning of ballast tanks or
ships, tank trucks, and tank cars. Leaky valves
and connections and flushing of pipelines are
other sources of effluents. The methods used
for treatment and disposal of these waters are
similar to those used in the1 other phases of the
industry.
THE AIR POLLUTION PROBLEM
From an air pollution standpoint the most objec-
tionable contaminants emitted from liquid waste
streams are hydrocarbons, sulfur compounds,
and other malodorous materials.
The effect of hydrocarbons in smog-producing
reactions is well known, and sulfur compounds
such as mercaptans and sulfides produce very
objectionable odors, even in high dilution. These
contaminants can escape to the atmosphere from
openings in the sewer system, open channels,
open vessels, and open oil-water separators.
The large exposed surface area of these sep-
arators requires that effective means of control
be instituted to minimize hydrocarbon losses to
the atmosphere from this source. A method
(Jones and Viles, 1952) developed by personnel
of Humble Oil and Refining Company may be
used to estimate the hydrocarbon losses from
-------
Oil-Water Effluent Systems
655
open oil-water separators. In the development
of this method the principal variables that in-
fluence evaporation rates were assumed to be
vapor pressure of the oil, and wind velocity.
Experimental work was done to observe and
correlate the effects of these factofs on evapo-
ration rates. From the data compiled, a proce-
dure for calculation of losses •was devised. Es-
sentially, this procedure is as follows:
1. Obtain a representative sample of oil at the
surface of the separator.
2. Obtain the vapor pressure of the sample and
the average wind velocity at the surface of
the separator.
3. Using Figure 487, find the loss in bbl/day
per ft2.
4. Since the data compiled were collected
under ideal conditions, a correlation
factor is needed to correct the value ob
tained from Figure 487 to actual separator
conditions. This correlation factor may
be found by measuring the evaporation rate
of a weighted sample of a constantly boiling
hydrocarbon from a shallow vessel placed
on the surface of the separator. The cor-
relation factor is then calculated as the
ratio of the measured rate of evaporation
to the theoretical evaporation rate from
Figure 487.
5. The product of the theoretical separator loss,
the correlation factor, and the separator area
represents the total evaporation loss.
AIR POLLUTION CONTROL EQUIPMENT
Hydrocarbons From Oil-Water Separators
The most effective means of control of hydrocar-
bon emissions from oil-water separators has been
the covering of forebays or primary separator
sections. Either fixed roofs or floating roofs
(Brown and Sublett, 1957) are acceptable covers.
Separation and skimming of over 80 percent of
the flotable oil layer is effected in the covered
sections. Thus, only a minimum of oil is con-
tained in the effluent water, which flows under
concrete curtains to the open afterbays or secon-
dary separator sections.
Satisfactory fixed roofs have been constructed
by using wooden beams for structural support,
and asbestos paper as a cover. A mastic-type
sealing compound is then used to seal all joints
and cracks. Although this form of roof is ac-
ceptable for the control of pollutants, in practice,
making the roof completely vaportight is difficult.
The resultant leakage of air into the vapor space,
Figure 487. Relationship of laboratory evapora-
tion rates for various wind velocities to vapor
pressure of oil (Jones and Viles, 1952).
and vapor leakage into the atmosphere are not
desirable from standpoints of air pollution or
safety. For example, an explosive mixture re-
sulting from leakage of air from gaging opera-
tions into the vapor space of a fixed-roof sep-
arator at a Los Angeles refinery was ignited by
a static electric spark. The destruction of the
wooden roof has emphasized the need for elim-
ination of the vapor space. Another type of en-
closed separator with a concrete cover and gas
blanketing of the vapor space has proved satis-
factory. The effluent vapors from this system
are vented to vapor recovery.
The explosion hazard associated with fixed roofs
is not present in a floating-roof installation. These
roofs are similar to those developed for storage
tanks. The floating covers are built to fit into
bays with about 1 inch of clearance around the
perimeter. Fabric or rubber may be used to seal
the gap between the roof edge and the container
wall. The roofs are fitted with access manholes,
skimmers, gage hatches, and supporting legs.
Floating roofs on refinery separators are shown
in Figures 488 and 489. In operation, skimmed
oil flows through lines from the skimmers to
a covered tank (floating roof or connected to
vapor recovery) or sump and then is pumped to de-
-------
656
PETROLEUM EQUIPMENT
Figure 488. Floating-roof cover on refinery oil-water separator
(Union Oil Company of California, Los Angeles, Calif.).
emulsifying processing facilities. Effluent water
from the oil-water separator is handled in the
manner described previously.
In addition to covering the separator, open sewer
lines that may carry volatile products are con-
verted to closed, underground lines with water-
seal-type vents. Junction boxes are vented to
vapor recovery facilities, and steam is used to
blanket the sewer lines to inhibit formation of ex-
plosive mixtures.
Accurate calculation of the hydrocarbon losses
from separators fitted with fixed roofs is difficult
because of the many variables of weather and re-
finery operations involved. One empirical equa-
tion that has been used with reasonable success
to calculate losses from separators is
AdHm
(12)(379)
(136)
where
w
A
d
weight of hydrocarbon loss, Ib/hr
area of covered separator, ft
depth of vapor space, in.
H = vol % of hydrocarbons as hexane in
the vapor space
rn = molecular weight of hexane.
In using this equation, assume that the density of
condensed vapors (C^Hj^) equals 5.5 pounds per
gallon and that the vapor in the separator is dis-
placed once per hour. The vapor concentration i;
determined by using the a,verage of readings from
a calibrated explosion meter over the entire cov-
ered area. The assumption that the vapors are
displaced once every hour was determined by us-
ing data from work done by the Pacific Coast Gas
Association (Powell, 1950).
The previously discussed methods of obtaining
emissions from uncovered separators may also
be applied to sections covered with fixed roofs.
Use of more than one method and a number of
tests of one source over a considerable period
of time are necessary to ensure an acceptable
estimate of emissions.
Emissions from separators fitted with floating-
roof covers may be assumed to be almost negli-
gible. A rough approximation of the magnitude of
the emission can be made by assuming the emis-
sion to be from a floating-roof storage tank of
-------
Oil-Water Effluent Systems
657
Figure 489. Floating roof on refinery oil-water
separator (Atlantic-Ricnfield Oil Company, Los
Angeles, Calif.).
particles in water that cannot be divided effec-
tively by means of gravity alone. Gravity-type
oil-water separators are, in most cases, inef-
fective in breaking the emulsions, and means
are provided for separate treatment where the
problem is serious.
Oil-in-water emulsions are objectionable in the
drainage system since the separation of other-
wise recoverable oil may be impaired by their
presence. Moreover, when emulsions of this
type are discharged into large bodies of water,
the oil is released by the effect of dilution, and
serious pollution of the water may result.
Formation of emulsions may be minimized by
proper design of process equipment and piping.
Several methods, both physical and chemical,
are available for use in breaking emulsions.
Physical methods of separation include direct
application of heat, distillation, centrifuging,
filtration, and use of an electric field. The ef-
fectiveness of any one method depends upon the
type of emulsion to be treated. Chemical meth-
ods of separation are many and varied. During
recent years the treatment of waste water con-
taining emulsions with coagulating chemicals has
become increasingly popular.
Variations of this form of treatment include air
flotation systems, and biological treatment of the
waste water, as discussed previously in this sec-
tion.
equivalent perimeter. The API method of calcu-
lating losses from storage tanks can then be ap-
plied.
Treatment of Refinery Liquid Wastes at Their Source
Isolation of certain odor- and chemical-bear ing
liquid wastes at their source for treatment be-
fore discharge of the water to the refinery 'waste-
water-gathering system has been found to be the
most effective and economical means of minimiz-
ing odor and chemicals problems. The unit that
is the source of wastes must be studied for possi-
ble changes in the operating process to reduce
wastes. In some cases the wastes from one pro-
cess may be used to treat the wastes from anoth-
er. Among the principal streams that are treat-
ed separately are oil-in-water emulsions, sulfur-
bearing waters, acid sludge, and spent caustic
wastes.
Oil-in-Water Emulsions
Oil-in-water emulsions are types of wastes that
can be treated at their source. An oil-in-water
emulsion may be defined as a suspension of oil
Sulfur-bearing waters
Sulfides and mercaptans are removed from waste-
water streams by various methods. Some refin-
eries strip the waste -water in a column with live
steam. The overhead vapors from the column
are condensed and collected in an accumulator
from which the noncondensables flow to sulfur-
recovery facilities or are incinerated; One Los
Angeles refinery removes all the hydrogen sul-
fide and about 90 percent of the ammonia from a
waste stream by this method. Flue gas has also
been used successfully as the stripping medium
in pilot-plant studies. Bottoms water from steam-
stripping towers, being essentially sulfide free,
can then be drained to the refinery's sewer sys-
tem.
Oxidation of sulfides in waste water is also an
effective means of treatment (Smith, 1956a). Air
and heat are used to convert sulfides and mer-
captans to thiosulfates, which are water soluble
and not objectionable. Figure 490 depicts the
flow through an air oxidation unit. Experience
has shown that, under certain conditions, the
thiosulfates may be reduced by the action of
Vibrio desulfuricans bacteria, which results in
-------
658
PETROLEUM EQUIPMENT
STEAM
AIR
SULFIDE HATER1
TO FURNACE FIREBOX
OR INCINERATOR
COOLER
OXIDIZER COLUMN
TO SEWER
Figure 490. Flow diagram of air oxidation process (Smith, 1956b).
the release of hydrogen sulfide. The reduction
takes place only in the absence of dissolved oxy-
gen. Care must be used to keep this water from
entering retention sumps or pits subject to this
bacterial attack.
2. processing to produce byproducts such as
ammonium sulfate, metallic sulfates, oils,
tars, and other materials;
3. processing for recovery of acid.
Chlorine is also used as an oxidizing agent for
sulfides. It is added in stoichiometric quanti-
ties proportional to the waste water. This meth-
od is limited by the high cost of chlorine. Water
containing dissolved sulfur dioxide has been used
to reduce sulfide concentration in waste waters.
For removing small amounts of hydrogen sulfide,
copper sulfate and zinc chloride have been used
to react and precipitate the sulfur as copper and
zinc sulfides. Hydrogen sulfide maybe released,
however, if the water treated with these com-
pounds contacts an acid stream.
Acid sludge
The acid sludge produced from treating opera-
tions varies with the stock treated and the con-
ditions of treatment. The sludge may vary from
a low-viscosity liquid to a solid. Methods of dis-
posal of this sludge are many and varied. Basic-
ally, they may be considered under three general
headings:
1. Disposal by burning as fuel, or dumping in
the ground or at sea;
The burning of sludge results in discharge to the
atmosphere of excessive amounts of sulfur dioxid<
and sulfur trioxide from furnace stacks. This
latter consideration has caused the discontinuance
of this method of disposal in Los Angeles County.
If sludge is solid or semis olid it may be buried
in specially constructed pits. This method of
disposal, however, creates the problem of acid
leaching out to adjacent waters. Dumping in
designated sea areas eliminates pollution of the
potable waters and atmosphere of populated areas
Recovery of sulfuric acid from sludge is accom-
plished essentially by either hydrolysis or thermal
decomposition processes. Sulfuric acid sludge is
hydrolyzed by heating it with live steam in the
presence of water. The resulting product sepa-
rates into two distinct phases. One phase con-
sists of diluted sulfuric acid -with a small amount
of suspended carbonaceous material, and the sec-
ond phase, of a viscous acid-oil layer. The dilute
sulfuric acid may be (1) neutralized by alkaline
wastes, (2) reacted chemically with ammonia-
water solution to produce ammonium sulfate for
fertilizer, or (3) concentrated by heating.
-------
Pumps
659
Acid sludge may be decomposed by heating to
300°F to form coke, sulfur dioxide, oil, water,
and lighter boiling hydrocarbons as a gas. Sev-
eral commercial decomposition processes have
been developed to use the sulfur present in the
sludge. In all these processes a kiln is used
•wherein the sludge-is mixed with hot coke or
some other carrying agent and heated to the re-
quired temperature. Another process allows
the acid sludge to be burned directly. The sul-
fur dioxide gases from the reaction are purified
and then either converted to sulfuric acid (con-
tact process) or to free elemental sulfur. The
tail gases emitted from these decomposition pro-
cesses may create an odor nuisance as well as
cause damage to vegetation in the surrounding
area. Because of this, the tail gases may re-
quire additional treatment to preclude the possi-
bility of a nuisance.
Of all the methods discussed, hydrolysis and de-
composition are the most desirable from the
standpoint of air pollution control, though they
are not the most economical when the volume of
acid is small.
Spent caustic wastes
Caustic soda is widely used in the industry to
neutralize acidic materials found in crude oil
and its fractions. It is also used to remove
mercaptans, naphthenates, or cresols from gas,
gasoline, kerosene, and other product streams.
The resulting spent caustic is imbued with the
odors of the compounds that have been extracted
in the various treating processes (American
Petroleum Institute, I960). This spent caustic
can be a source of intense objectionable odors
and can result in nuisance complaints.
Spent caustic is treated by direct methods or
chemical processing, or both. Direct methods
of disposal include ponding, dilution, disposal
wells, and sale. Of these, ponding is not recom-
mended, since the pond could become a source
of air pollutants as well as a possible source of
contamination of underground water through seep-
age. Dilution of spent caustic in large bodies of
water is a commonly used method of disposal.
The ocean and brackish waters are the only desir-
able areas for this disposal, to preclude pollu-
tion of fresh-water streams.
Disposal •wells afford another convenient means
of disposing of spent-caustic solutions, provided
that local conditions are favorable. The method
consists of pumping the liquid -wastes into under-
ground formations that contain saline or nonpotable
water. Spent caustics that contain phenolates,
cresolates, and sulfides may be sold outside the
industry for recovery of these materials.
In addition to these direct methods of disposal,
chemical processing methods are available.
These include neutralization, combination of
neutralization and oxidation, and combination of
oxidation and chemical separation.
Neutralization of high-alkaline caustic -wastes may
be effected by means of spent acids from other re-
finery operations. After neutralization, the result-
ing salt solution may be suitable for discharge into
the refinery's drainage system. In some cases
odorous or oily materials may have to be stripped
from the product before discharge. In these in-
stances effluent gases should be incinerated.
Spent-caustic solutions can also be neutralized
with acid gases such as flue gases (Fisher and
Moriarty, 1953). Oxygen contained in the flue
gas tends to oxidize sulfides and mercaptides as
a secondary reaction. Effluent gases from this
reaction must be properly incinerated to prevent
odor problems. The resulting treated solution
contains carbonates, bicarbonates, thiosulfates,
sulfates, and sulfites and may be suitable for dis-
charge into the drainage system.
A recently developed method of treating caustic
•wastes involves the addition of pickling acid. The
acid is mixed with caustic and is airblown. The
resulting solution is filtered and naphtha is added
to extract organic acids for recovery. Fumes
from the airblowing operation must be incinerated.
The treated salt solution is discharged to a drain-
age system.
PUMPS
TYPES OF PUMPS
Pumps are used in every phase of the petroleum
industry. Their applications range from the lift-
ing of crude oil from the depths of a well to the
dispensing of fuel to automobile engines. Leakage
from pumps can cause air pollution wherever or-
ganic liquids are handled.
Pumps are available in a wide variety of models
and sizes. Their capacities may range from
several milliliters per hour, required for some
laboratory pumps, to 3/4 million gallons per min-
ute, required of each of the new pumps at Grand
Coulee Dam (Dolman, 1952).
Materials used for construction of pumps are also
many and varied. All the common machinable
metals and alloys, as well as plastics, rubber,
-------
660
PETROLEUM EQUIPMENT
and ceramics, are used. Pumps may be classi-
fied under two general headings, positive displace-
ment and centrifugal.
pict some typical pumps of each type. When a
positive-displacement pump is stopped, it serves
as a check valve to prevent backflow.
Positive-Displacement Pumps
Positive-displacement pumps have as their prin-
ciple of operation the displacement of the liquid
from the pump case by reciprocating action of a
piston or diaphragm, or rotating action of a gear,
cam, vane, or screw. The type of action may be
used to classify positive-displacement pumps as
reciprocating or rotary. Figures 491 and 492 de-
Centrifugal Pumps
Centrifugal pumps operate by the principle of con-
verting velocity pressure generated by centrifugal
force to static pressure. Velocity is imparted to
the fluid by an impeller that is rotated at high
speeds. The fluid enters at the center of the im-
peller and is discharged from its periphery. Un-
like positive-displacement pumps, when the cen-
S-PRCCESS .
~ LIQU I D
SUCTION POSIT ION
DISCHARGE DISCHARGE PIPE
ION
LD I S C H A R G E.(T~
*\ VALVES,
[ S U C T I 0 Nj
tVALVEi'
D ISCHARGE POSI T ION
CYLINDER LI QUIDCYL INDER
COUNTER-
BORES '
SUCTION- PI PE CONNECT ION''
C
DISCHARGE VALVES
MINION onrTinu
GEAR co!i^*^>c\OSSHEA v!A!!y^
GEAR,
CONNECTING
RODS
CONNECTING
RODS
Figure 491. Reciprocating pumps: (a) Principle of reciprocating pump,
(b) principle of fluid-operated diaphragm pump, (c) direct-acting steam
pump, (d) principle of mechanical diaphragm pump, (e) piston-type power
pump, (f) plunger-type power pump with adjustable stroke, (g) inverted,
vertical, triplex power pump (Dolman, 1952).
-------
Pumps
661
SUCTION DISCHARGE
-SEAR
D ISCHARGE
DISCHARGE
L3BEE" ^SUCTION
...OTOR
d
a b c
INLET DISCHARGE
t
DISCHARGE, DRIVING GEAR 0 I SC HA RG E SEAL
,__WK=t35B_*Jw .1. II n Ji ' u n *tr7777rr*- » „ c u
l\t 1 -
DISCHARGE
SHAFT
SUCTION
ECCENTRI
SUCTION
D ISCHARGE
OTOR
ROLLER-' ECCENTRIC
h
FLEXIBLE RUBBER
t _T
I D L E'R 0 T 0 R S
SUCTION
ECCEJMR I~C*T
SQUEEZE RING
Figure 492. Rotary pumps: (a) External-geai pump, (b) internal-gear
pump, (c) three-lobe oump, (d) four-lobe pump, (e) sliding-vane pump,
(f) single-screw pump, (g) swinging-vane puinp, (h) cam or roller pump,
(i) cam-and-piston pump, (j) three-screw pump;(k) shuttle-block pump,
(I) squeegee pump, (m) neoprene vane pump (Dolman, 1952).
trifugal type of pump is stopped there is a tenden-
cy for the fluid to backflow. Figures 493 and 494
depict some centrifugal pumps.
Other specialized types of pumps are available,
but, generally, the pumps used by the petroleum
industry fall into the two categories discussed.
Power for driving the various types of pumps is
usually derived from electric motors, internal
combustion engines, or steam drives. Any one
of these sources may be adapted for use with
either reciprocating pumps or centrifugal pumps.
Most rotary pumps are driven by electric motor.
The opening in the cylinder or fluid end through
which the connecting rod actuates the piston is
the major potential source of contaminants from
a reciprocating pump. In centrifugal pumps,
normally the only potential source of leakage
occurs where the drive shaft passes through the
impeller casing.
AIR POLLUTION CONTROL EQUIPMENT
Several means have been devised for sealing the
annular clearance between pump shafts and fluid
casings to retard leakage. For most refinery ap
plications, packed seals and mechanical seals an
widely used,,
THE AIR POLLUTION PROBLEM
Operation of various pumps in the handling of fluids
in petroleum process units can result in the re-
lease of air contaminants. Volatile materials such
as hydrocarbons, and odorous substances such as
hydrogen sulfide or mercaptans are of particular
concern because of the large volumes handled. Both
reciprocating and centrifugal pumps can be sources
of emissions.
Packed seals can be used on both positive dis-
placement and centrifugal type pumps (Elonka,
1956). Typical packed seals, as shown in Fig-
ure 495, generally consist of a stuffing box filled
with sealing material that encases the moving
shaft. The stuffing box is fitted with a takeup
ring that is made to compress the packing and
cause it to tighten around the shaft. Materials
used for packing vary with the product temper-
ature, physical and chemical properties, pres-
-------
662
PETROLEUM EQUIPMENT
DISCHARGE
' NOZZLE
IMPELLER
DISCHARGE
VOLUTE DIFFUSION
IMPELLER VANES VANES
DISCHARGE
VANES
VANES
f
Figure 493. Centrifugal pumps: (a) Principle of centrifugal-type
pump, (b) radial section through volute-type pump, (c) radial sec-
tion through diffuser-type pump, (d) open impeller (e) semi-en-
closed impeller, (f) closed impeller, (g) nonclog impeller (Dolman,
1952).
sure, and pump type. Some commonly used
materials are metal, rubber, leather, wood, and
plastics.
Lubrication of the contact surfaces of the pack-
ing and shaft is effected by a controlled amount
of product leakage to the atmosphere. This fea-
ture makes packing seals undesirable in applica-
tions where the product can cause a pollution prob-
lem. The packing itself may also be saturated witl
some material such as graphite or oil that acts as
a lubricant. IB some cases cooling or quench
water is used to cool the impeller shaft and the
bearings.
The second commonly used means of sealing is
the mechanical seal (Elonka, 1956), which was
developed over a period of years as a means of
reducing leakage from pump glands. This type
of seal can be used only in pumps that have a
rotary shaft motion. A simple mechanical seal
consists of two rings with wearing surfaces at
right angles to the shaft (see Figure 496). One
ring is stationary while the other is attached to
the shaft and rotates with it. A spring and the
action of fluid pressure keep the two faces in
contact. Lubrication of the wearing faces is ef-
fected by a thin film, of the material being pumped
The wearing faces are precisely finished to en-
sure perfectly flat surfaces. Materials used in
the manufacture of the sealing rings are many
and varied. Choice of materials depends pri-
marily upon properties of fluid being pumped,
pressure, temperature, and speed of rotation.
The vast majority of rotating faces in com-
mercial use are made of carbon (Woodhouse,
1957).
Emissions to the atmosphere from centrifugal
pumps may be controlled in some cases by use
of the described mechanical-type seals instead
of packing glands. For cases not feasible to
control "with mechanical seals, specialized types
of pumps, such as canned, diaphragm, or elec-
tromagnetic, are required.
The canned-type pump is totally enclosed, with
its motor built as an integral part of the pump.
Seals and attendant leakage are eliminated. The
diaphragm pump is another type devoid of seals.
A diaphragm is actuated hydraulically, me-
chanically, or pneumatically to effect a pump-
ing action. The electromagnetic pumps use an
electric current passed through the fluid, which
is in the presence of a strong magnetic field,
to cause motion.
A pressure-seal-type application can reduce
packing gland leakage. A liquid, less volatile
or dangerous than the product being pumped, is
-------
Pumps
663
''BALL BEARING
.SUCTION CHAMBER
FOR WATER
/COOLING
PRESSURE '
RELIEF CHAMBER SUC"ON->P|PING AND
VALVE FOR CONTROL
OF STUFFING-BOX PRESSURE
j
AR I NG
COOLING
JACKET
MERCURY
SEAL"
FFUSEBS CABLE
TERMINAL
PUMP
KBOWLS
MOTOR CASE
MOTOR ROTOR
WINDINGS
(INTEGRAL
Wl TH
IMPELLER)
•MOTOR STA,-
OR, WINDING
-0 IAPHRAGM
^PROTECTS
STATOR)
^BEARING
STRAINER'
Figure 494. Centrifugal pumps: (h) single-stage, double-suction,
split-case, centrifugal pump; (i) close-coupled water pump; (j)
four-stage pump with opposed impellers; (k) turbine-type, deep-wel
pump; (I) submersible-motor, deep-well pump; (m) close-coupled,
vertical, turret-type pump; (n) pump with integral motor (Dolman,
1952).
PRODUCT
PRODUCT
PACKING
Figure 495. Diagram of simple uncooled packed seal
introduced between two sets of packing. This
sealing liquid must also be compatible with the
product. Since this liquid is maintained at a
higher pressure than the product, some of it
passes by the packing into the product. The
pressure differential prevents the product from
leaking out-ward, and the sealing liquid pro-
vides the necessary lubricant for the packing
gland. Some of the sealing liquid passes the
outer packing (hence the necessity of low vola-
tility), and a means should be provided for its
disposal.
This application is also adaptable to pumps
with mechanical seals. A dual set of mechan-
ical seals similar to the two sets of packing is
used.
-------
664
PETROLEUM EQUIPMENT
SET SCRE1S
FLANGE
LOCK PIN
SPRING HOLDER
U-CUP F3LLOKER
STATIONARY FACE
SEAT GASKET
SEAL FLANGE GASKET
Figure 496. Diagram of simple mechanical seal
(Borg-Warner Mechanical Seals. A Division of
Borg-Warner Corporation, Vernon, Calif.).
Volatile vapors that leak past a main seal may-
be vented to vapor recovery by using dual seals
and a shaft housing.
Other than the direct methods used to control
leakage, operational changes may minimize re-
lease of contaminants to the atmosphere. One
desirable change is to bleed off pump casings
during shutdown to the fuel gas system, vapor
recovery facilities, or a flare instead of di-
rectly to the atmosphere.
Results of Study to Measure Losses From Pumps
The results of a testing program (Steigerwald,
1958) to establish the magnitude of hydrocarbon
losses from pumps are presented in Tables 182
through 185. The data collected during the study
are presented in Table 184 as a comparison of
the effectiveness of packing glands and mechani-
cal se-als in preventing leakage.
Table 182. SCOPE AND RESULTS OF FIELD TESTS ON PUMP SEALS (Steigerwald, 1958)
Group No. a
1
2
3
4
5
6
7
8
9
Subtotal
10
11
Total
number
oi' seals
76
82
66
127
266
56
163
191
150
), 177
92
78
12 68
13
14
15
16
17
18
Subtotal
19
20
21
22
23
24
25
26
27
Subtotal
Totals
49
179
103
100
175
124
968
26
32
38
72
173
150
60
40
50
641
2,786
Seals inspected
Number
14
13
12
21
59
16
34
35
19
223
15
9
9
0
21
18
15
26
25
138
6
5
8
13
29
17
°7o of total
18
16
18
17
22
28
21
18
13
19
16
12
13
0
12
18
15
16
20
14
23
Measured leaks
Number
2
0
0
6
3
0
13
2
2
28
5
1
0
0
12
Hydrocarbon
los s ,
Ib/day
60
0
0
294
19
0
262
23
7
665
26
4
0
0
83
0 0
6 | 280
6
2
?Z
226
16
635
1 23
16 1 0
21
18
17
11
7 12
7
20
112
473
18
40
17
17
0
D
1
0
4
3
0
15
75
0
0
383
71
0
19
82
0
578
1,878
Small leaks
Number
2
2
2
7
13
5
6
4
2
43
2
2
3
0
5
3
4
11
0
30
3
0
3
3
7
0
2
3
1
22
95
Hydrocarbon
loss,
lb/dayb
2
2
2
7
13
5
6
4
2
43
2
2
3
0
5
3
4
11
0
30
3
0
3
3
7
0
2
3
1
22
95
Hydrocarbon loss
from inspected pumps,
Ib/day
62
2
2
301
32
5
26S
27
9
708
28
6
3
0
88
3
284
237
16
665
26
0
3
386
78
0
21
85
1
600
1, 973
aGroup numbers represent a specific combination oi pump type, sea] type, pump operation, and product.
A value of 1 pound per day was assigned to a small leak on a pump seal.
-------
Airblown Asphalt
665
Table 183. EXTRAPOLATION OF FIELD DATA BY SAMPLING GROUPS TO
OBTAIN A TOTAL LOSS FIGURE (Steigerwald, 1958)
1
Group No.a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Totals
2
Total
number
of seals
76
82
66
127
266
56
163
191
150
92
78
68
49
179
103
100
175
124
26
32
38
72
173
150
60
40
50
2,786
3
Number
of seals
inspected
14
13
12
21
59
16
34
35
19
15
9
9
0
21
18
15
26
25
6
5
8
13
29
17
7
7
20
473
4
Hydrocarbon
loss from
inspected seal,
lh/Hay
62
2
2
301
32
5
268
27
9
28
6
3
_-
88
3
284
237
16
26
0
3
386
78
0
21
85
1
1,973
5
Avg hydrocarbon
loss per inspected
seal, lb/dayb
4. 4
0.2
0.2
14.4
0.6
0.3
7.9
0.8
0.5
1.8
0. 7
0.3
4.2
0.2
18.8
9. 1
0.6
4.3
0
0.4
29.6
2.7
0
3.0
12. 1
0. 1
4.2
6
Total
hydrocarbon
loss, lb/dayc
335
16
13
1,830
160
17
1,289
153
75
166
55
20
752
21
1,880
1,592
74
112
0
15
2, 131
467
0
180
484
5
11,842 or
6 tons per
day
aGroup numbers represent a specific combination of pump type, seal type, pump
operation, and product.
Divide hydrocarbon loss from inspected seal, Ib/day, by number of seals inspected.
cMultiply average hydrocarbon loss per inspected seal, Ib/day, by total number of
seals.
The slight difference between the average losses
from mechanical seals and packed glands during
handling of highly volatile hydrocarbons needs
further clarification. Pumps in continuous ser-
vice show an average loss per seal of 18. 3 and
7. 9 pounds per day for packed and mechanical
seals, respectively, indicating that mechanical
seals are far more efficient when running con-
tinuously. On spare or standby service the
packed seals are more effective, losing 1.8
pounds per day to an average loss of 4. 4 pounds
from mechanical seals. Reciprocating pumps
handling light products are the worst offenders
both in incidence of leak and magnitude of av-
erage emissions. The largest leak encountered
in the study, 266 pounds per day, was from a
reciprocating pump on intermittent service han-
dling liquefied petroleum gas.
AIRBLOWN ASPHALT
Asphalt is a dark brown to black, solid or semi-
solid material found in naturally occurring de-
posits or as a colloidal suspension in crude oil.
Analytical methods have been used to separate
asphalt into three component groups--asphaltenes,
resins, and oils. A particular grade of asphalt
may be characterized by the amounts of each
group it contains. The asphaltene particle pro-
vides a nucleus about which the resin forms a
protective coating. The particles are suspended
-------
666
PETROLEUM EQUIPMENT
Table 184. EFFECTIVENESS OF MECHANICAL AND PACKED SEALS ON
VARIOUS TYPES OF HYDROCARBONS (Steigerwald, 1958)
Seal type
Mechanical
Avg
Packed
Avg
Packed
Avg
Pump type
Centrifugal
Centrifugal
Reciprocating
Type
hydrocarbon
being pumped,
Ib Reid
> 26
5 to 26
0.5 to 5
> 0.5
> 26
5 to 26
0.5 to 5
> 0. 5
26
5 to 26
0. 5 to 5
> 0.5
Avg hydrocarbon
loss per
inspected seal,
Ib/day
9.2
0.6
0.3
3.2
10.3
5.9
0.4
4.8
16.6
4.0
0. 1
5.4
Leak incidence
Small leaks, a
% of total
inspected
19
18
19
19
20
32
12
22
31
24
9
20
Large leaks,
% of total
inspected
21
5
4
13
37
34
4
23
42
10
0
13
Small leaks lose less than 1 pound of hydrocarbon per day.
Table 185. AVERAGE PUMP SEAL LOSSES BY
VOLATILITY OF PRODUCT BEING PUMPED
(Steigerwald, 1958)
Product,
Ib Reid
26
5 to Ib
0. 5 to 5
Total number
of seals
reported
765
1,216
805
Number of
seals
inspected
125
204
144
Avg hydrocarbon
loss per inspected
seal, Ib/day
11.1
2.7
0. 3
in an oil that is usually paraffinic but can be
naphthenic or naptheno-aromatic.
RECOVERY OF ASPHALT FROM CRUDE OIL
Over 90 percent of all asphalt used in the United
States is recovered from crude oil (Kirk and Othmer,
1947). The method of recovery depends upon the
type of crude oil being processed. Practically
all types of crudes are first distilled at atmospher-
ic pressure to remove the lower boiling materials
such as gasoline, kerosene, diesel oil, and others.
Recovery of nondistillable asphalt from selected
topped crudes may then be accomplished by vac-
uum distillation, solvent extraction, or a com-
bination of both.
A typical vacuum distillation unit is depicted in
Figure 497. A unit such as this uses a heater,
preflash tower, vacuum vessel, and appurtenances
for processing topped crudes. Distillation of topped
crude under a high vacuum removes oils and wax
as distillate products, leaving the asphalt as a
residue. The amount of oil distilled from the resi-
due asphalt controls its properties; the more oil
and resin or oily constituents removed by dis-
tillation, the harder the residual asphalt. Resid-
ual asphalt can be used as paving material or it
can be further refined by airblowing.
Asphalt is also produced as a secondary product
in solvent extraction processes. As shown in
Figure 498, this process separates the asphalt
from remaining constituents of topped crudes by
differences in chemical types and molecular
weights rather than boiling points as in vacuum
distillation processes. The solvent, usually a
light hydrocarbon such as propane or butane, is
used to remove selectively a gas-oil fraction
from the asphalt residue.
AIRBLOWING OF ASPHALT
Economical removal of the gas-oil fraction from
topped crude, leaving an asphaltic product, is
occasionally feasible only by airblowing the crude
residue at elevated temperatures. Excellent pav-
ing-grade asphalts are produced by this method.
Another important application of airblowing is in
the production of high-quality specialty asphalts
for roofing, pipe coating, and similar uses.
These asphalts require certain plastic proper-
ties imparted by reacting with air.
Airblowing is mainly a dehydrogenation process.
Oxygen in the air combines v/ith hydrogen in the
oil molecules to form water vapor. The pro-
gressive loss of hydrogen results in polymeriza-
tion or condensation of the asphalt to the desired
consistency. Blowing is usually carried out batch-
wise in horizontal or vertical stills equipped to
blanket the charge with steam, but it may also be
-------
Airblown Asphalt
667
PREFLASH
TO«ER
NONCONDENSABLE
GAS TO TREATER UNIT
INCINERATED IN
FURNACE FIREBOX
ACCUMULATOR
MS Oil
ASPHALT
HEATER
Figure 497. Flow diagram of vacuum distillation
unit.
done continuously. Vertical stills are more ef-
ficient because of longer air-asphalt contact time.
The asphalt is heated by an internal fire-tube
heater or by circulating the charge material
through a separate tubestill. A temperature of
300° to 400°F is reached before the airblow-
ing cycle begins. Air quantities used range from
5 to 20 cubic feet per minute per ton of charge
(Earth, 1958). Little additional heat is then
needed since the reaction becomes exothermic.
Figure 499 depicts the flow through a typical
batch-type unit.
THE AIR POLLUTION PROBLEM
Effluents from the asphalt airblowing stills in-
clude oxygen, nitrogen and its compounds, water
vapor, sulfur compounds, and hydrocarbons as
gases, odors, and aerosols. Discharge of these
vapors directly to the atmosphere is objectionable
from an air pollution control standpoint. The dis-
agreeable odors and airborne oil particles en-
trained with the gases result in nuisance com-
plaints. Disposal methods are available that can
satisfactorily eliminate the pollution potential of
the effluents.
AIR POLLUTION CONTROL EQUIPMENT
Control of effluent vapors from asphalt airblow-
ing stills has been accomplished by scrubbing and
incineration, singly or in combination. Most in-
stallations use the combination. Potential air
pollutants can be removed from asphalt still gas-
es by scrubbing alone. One effective control in-
stallation in Los Angeles County uses sea water
for one-pass scrubbing of effluent gases from foui
asphalt airblowing stills. The fume scrubber is
a standard venturi-type unit. The scrubber ef-
fluent is discharged into an enclosed oil-water
gravity-type separator for recovery of oil, which
is reprocessed or used as fuel. Effluent gases
from the covered separator that collects the
scrubber discharge are not incinerated but flow
through a steam-blanketed stack to the atmosphere
The system, shown in Figure 500, removes es-
sentially all potential air pollutants from the ef-
fluent stream. A limiting factor in the applica-
tion of this method of control is the water supply.
Since a high water-to-vapor scrubbing ratio
(100 gallons/1, 000 scf) is necessary, an econom-
ical source of water should be readily available
to supply the large volume required for one-pass
operation.
-------
668
PETROLEUM EQUIPMENT
(PROPANE \
ACCUMULATOR^
REDUCED CRUDE
OIL
DEASPHALTING
TOKER
STEAM
EVAPORATORS
STEAM
OIL
STRIPPER
PROPANE
COMPRESSOR
DEASPHALTED
OIL
HEATER
FLASH
TOWER
Figure 498. Flow diagram of propane deasphalting
unit.
TO COVERED EFFLUENT
HATER SEPARATOR
STEAM
ASPHALT
ASPHALT
STRIPPER
STEAM
OFF GAS TO
HEATER
BLOWING STILL
SCRUBBER KNOCKOUT DfiUM
Figure 499. Flow diagram of airblown asphalt manufacture (batch process).
-------
Valves
669
OCEAN HATER
(80 psi)
FUME
SCRUBBER
EXHAUST GASES
TO ATMOSPHERE
SlEAM
KNOCKOUT
DRUM
AIR BLOW
ASPHALT STILLS
(BATCH OPERATION)
BLANKET
HIST ELIMINATOR
COVERED SEPARATOR
CONDENSATE
TO STORAGE
I !
SKIMMED OIL EFFLUENT KATER TO
TO STORAGE COVERED SEPARATOR
Figure 500. Flow diagram of scrubbing system.
Where removal of most of the potential air pollu-
tants is not feasible by scrubbing alone, the non-
condensables must be incinerated. Essential to
effective incineration is direct-flame contact
with the vapors, a minimum retention time of
0.3 second in the combustion zone, and mainte-
nance of a minimum combustion chamber tem-
perature of 1,200°F. Other desirable features
include turbulent mixing of vapors in the combus-
tion chamber, tangential flame entry, and ade-
quate instrumentation. Primary condensation
of any steam or water vapor allows use of small-
er incinerators and results in fuel savings. Some
of the heat released by incineration of the waste
gases may be recovered and used for generation
of steam. General design features of waste
gas afterburners and boilers are discussed else-
where in this manual.
Catalytic fume burners a^-e not recommended
for the disposal of vapors from the air blowing
of asphalt because the matter entrained in the
vapors would quickly clog the catalyst bed.
VALVES
TYPES OF VALVES
Valves are employed in every phase of the petro-
leum industry where petroleum or petroleum
product is transferred by piping from one point
to another. There is a great variety of valve
designs, but, generally, valves may be classi-
fied by their application as flow control or pres-
sure relief.
Manual and Automatic Flow Control Valves
Manual and automatic1 flow control valves are
used to regulate the flow of fluids through a sys-
tem. Included under this classification are the
gate, globe, angle, plug, and other common
types of valves. These valves are subject to
product leakage from the valve stem as a result
of the action of vibration, heat, pressure, cor-
rosion,or improper maintenance of valve stem
packing (see Figure 501).
-------
670
PETROLEUM EQUIPMENT
BODY
PLUG
BUNDHEAD
Figure 501. Typical valve showing various parts
and potential source of hydrocarbon emission
from the valve stem (Mason-Nei Ian, Division of
Worthington Corporation, Norwood, Mass.).
Pressure Relief and Safety Valves
Pressure relief and safety valves are used to
prevent excessive pressures from developing
in process vessels and lines. The relief valve
designates liquid flow while the safety valve
designates vapor or gas flow. These valves
may develop leaks because of the corrosive
action of the product or because of failure of the
valve to reseat properly after blowoff. Rup-
ture discs are sometimes used in place of
pressure relief valves. Their use is restricted
to equipment in batch-type processes. The
maintenance and operational difficulties caused
by the inaccessibility of many pressure relief
valves may allow leakage to become substantial.
THE AIR POLLUTION PROBLEM
Quantitative data as to actual extent of emissions
to the atmosphere from this Leakage are some-
what limited, but available data indicate that
emissions vary over a wide range. Liquid leak-
age results in emissions from evaporation of liq-
uid while gas leakage results in immediate emis-
sions. The results of a test program (Kanter et
al. , 1958) conducted to establish the magnitude
of hydrocarbon emissions from, valves are pre-
sented in Table 186. In this program, valves in
a group of 11 Los Angeles County refineries
were surveyed. Both liquid and gaseous leaks
were measured or estimatec in the survey. Leaks
•were detected by visible means for liquid leaks,
and by spraying with soap solution followed by
inspection for bubble formation for gaseous leaks.
Liquid leakage rates were measured by collect-
ing liquid over a period of time. Flow rates for
gaseous leaks were determined by enclosing the
valve in polyethylene bags and venting the vapor
through a wet test meter.
Apparent from Table 186 is that 70 percent of the
measurable leaks in gas service average less
than 9. 1 pounds of emissions per day. In liquid
service, 90 percent of the measurable leaks av-
erage less than 8. 8 pounds of emissions per day.
Consideration of remaining data shows that the
frequency distribution of leaks is extremely
skewed.
An example of low leakage rate was observed
in one refinery where over 3, 500 valves han-
dling a wide variety of products under all con-
ditions of temperature and pressure were in-
spected. The average leak rate was 0. 038
pound per day per valve.
Examples of high leakage rates were found in
two refineries where all 440 valves inspected in
gas service had an average leak rate of 1. 6 pounds
per day per valve, and in one other refinery "where
all 1, 335 valves inspected in liquid service had an
average leak rate of 0. 32 pound per day per valve.
These examples illustrate the wide divergence
from the average valve leak rate that can exist
among refineries in a single area, all subject
to the same obligations to restrict their emis-
sions to the greatest possible extent. These re-
sults could not be applied, even approximately,
to refineries in other areas "where standards
may be different.
These testing programs -were also conducted on
pressure relief valves in the same oil refineries.
The results of this phase of the program are shown
in Table 187. As can be seen from the data, re-
lief valves on operational units have a slightly
lower leak incidence but a much higher average
-------
Valves
671
Table 186. LEAKAGE OF HYDROCARBONS FROM VALVES OF
REFINERIES IN LOS ANGELES COUNTY (Kanter et al. , 1958)
Total number of valves
Number of valves inspected
Small leaksa
Large leaks
Leaks measured
Total measured leakage, Ib /day
Average leak rate—large
leaks, Ib/day
Total from, all large leaks,
Ib/day
Estimated total from small
leaks, lb/dayb
Total estimated leakage from
all inspected valves, Ib/day
Average leakage per inspected
valve, Ib/day
Valves in
gaseous service
31, 000
2,258
256
118
24
218
9. 1
1, 072
26
1,098
0.486
Valves in
liquid service
101, 000
7,263
768
79
76
670
8.8
708
77
785
0. 108
All valves
132, 000
9,521
1,024
197
100
888
8.9
1, 780
103
1,883
0. 198
aSmall leaks are defined as leaks too small to be measured--those estimated to
be less than 0. 2 pound per day.
"Leaks too small to be measured were estimated to have an average rate of 0. 1
pound per day. This is one-half the smallest measured rate.
leakage rate than valves on pressure storage ves-
sels do. Moreover, dual-type valves (two single
relief valves connected in parallel to ensure ef-
fective release of abnormal pressures) on pres-
sure storage vessels have a greater leak inci-
dence and a larger average leakage rate than
single-type valves on similar service do. For
valves on operational vessels, the average for
all refineries was 2. 9 pounds of hydrocarbons
per day per valve. Average losses from spe-
cific refineries, however, varied from 0 to 9. 1
pounds per day per valve. Under diverse con-
ditions of operation and maintenance, emissions
can vary greatly from one refinery to another.
Total Emissions From Valves
Since emissions to the atmosphere from valves
are highly dependent upon maintenance, total
valve losses cannot be estimated accurately.
From the testing program mentioned, emis-
sions from, valves averaged 12 percent of the
total emissions from all refineries in Los An-
geles County. As of 1963, hydrocarbon emis-
sions from valves in Los Angeles County refin-
eries are estimated at about 11 tons per day. As
stated previously, however, these emissions
varied greatly from one refinery to another, and
average percentage figures should not be used in
predicting emissions from a given refinery.
Table 187. LEAKAGE OF HYDROCARBONS
FROM PRESSURE RELIEF AND SAFETY
VALVES OF REFINERIES IN LOS ANGELES
COUNTY (Kanter et al. , 1958)
Valve
group
Operational
units
Pressure
storage:
Single
Dual
Number
of valves
reported
1, 113
237
115
Number
of valves
tested
165
174
79
Hydrocarbon
emis sion,
Ib/day
480
56
98
Emission per
tested valve,
lb/daya
2. 90
0. 32
1.24
Total
emission ,
Ib/day
3, 230
80
140
AIR POLLUTION CONTROL EQUIPMENT
Obviously, the controlling factor in preventing
leakage from valves is maintenance. An effec-
tive schedule of inspection and preventive mainte-
nance can keep leakage at a minimum. Minor
leaks that might not be detected by casual obser-
vation can be located and eliminated by thorough
periodic inspections. New blind designs are
being incorporated in refinery pipeline systems
in conjunction with flow valves (see Figure 502).
This is done to ensure against normal leakage
that can occur through a closed valve.
-------
672
PETROLEUM EQUIPMENT
Figure 502. Bar-operated line blind that is ideal
for installation ahead of shutoff valve to ensure
against valve leaks and vapor emissions from valve
stem (Hamer Oil Tool Co. Catalog Sheet., Long
Beach, Calif.).
Emissions from pressure relief valves are
sometimes controlled by manifolding to a
vapor control device, such as described in
Chapter 5. Normally, these disposal systems
are not designed exclusively to collect vapors
from relief valves. The primary function of
the system may be to collect off gases produced
by a process iinit, or vapors released from
storage facilities, or those released by depres-
surizing equipment during shutdowns.
Another method of control to prevent excessive
emissions from relief valve leakage is the use of
a dual valve with a shutoff interlock. A means
of removing and repairing a detected leaking
valve without waiting until the equipment can be
taken out of service is thus provided. The prac-
tice of allowing a valve with a minor leak to
continue in service without correction until the
operating unit is shut down for general inspection
is common in many refineries. This practice
should be kept at a minimum.
A rupture disc is sometimes used to protect
against relief valve leakage caused by excessive
corrosion. The disc is installed on the pressure
side of the relief valve. The space between the
rupture disc and relief valve seat should be pro-
tected from pinhole leaks that could occur in
the rupture disc. Otherwise, an incorrect pres-
sure differential could keep the rupture disc from
breaking at its specified pressure. This, in
turn, could keep the relief valve from opening,
and excessive pressures could occur in the oper-
ating equipment.
One method of ensuring against these small leaks
in rupture discs is to install a pressure gage and
a small manually operated purge valve in the
system. The pressure gage would easily detect
any pressure increases from even small leaks.
In the event of leaks, the vessel would be re-
moved from service, and the faulty rupture disc
would then be replaced. A second, but less
satisfactory method from an air pollution con-
trol standpoint, is to maintain the space at at-
mospheric pressure by installing a small vent
opening. Any minute leaks would then be vent-
ed directly to the atmosphere, and a pressure
increase could not exist.
COOLING TOWERS
Cooling towers are major items of heat-transfer
equipment in the petroleum a.nd petrochemical
industries. They are designed to cool, by air,
the water used to cool industrial processes.
Cooling of the water by air involves evapora-
tion of a portion of the water into the air so that
the remaining water is cooled by furnishing heat
for this evaporation process. This cooled water
is used, in turn, in heat-exchange equipment to
cool other liquids and gases.
There are two styles of cooling towers—classified
by means of air movement. In one style, the
earliest developed, the prevailing wind is used
for the required ventilation. It has become known
as the natural draft or atmospheric type of cool-
ing tower (see Figure 503).
The other type of cooling tower employs fans to
move the air and is known as a mechanical-draft
cooling tower (see Figure 504). Fan location is
used in further classifying the tower as a forced-
or induced-draft cooling tov/er. The forced-draft
cooling tower has not proved very satisfactory,
since it has a tendency to recirculate its hot,
humid exhaust vapor in place of fresh air, and
its air distribution is poor because of the 90-degree
turn the air must make at high velocity (Kern,
1950).
Spray ponds, once used extensively for cooling
of water, have been abandoned in favor of cool-
ing towers. Spray ponds are limited in their per-
formance and suffer from high water, losses.
-------
Cooling Towers
673
Figure 503. Natural-draft cooling tower (Shell Oil Company,
Los Angeles, Cali f.).
CHARACTERISTICS OF COOLING TOWER OPERATION
Petroleum and petrochemical operations require
large quantities of water for temperature con-
trol purposes. The water is normally circulated
by pump between the heat-exchange equipment and
the cooling tower. The hydrocarbon stream to
be cooled can also be circulated directly through
the cooling tower. Approximately 1, 000 Btu is
required to evaporate 1 pound of water. This
is equivalent to cooling 100 pounds of water 10°F.
Thus, 1 percent of water is lost through evapora-
tion for every 10 degrees of cooling accomplished.
Additionally, a spray loss amounting to no more
than 0. 2 percent must be included for properly
designed atmospheric or mechanical-draft
towers. Water cannot be cooled below the wet
bulb temperature of the air entering the cooling
tower.
The performance of an individual cooling tower
is governed by the ratio of weights of air to water
and the time of contact between the air and water.
Commercially, the variation in the ratio of air
to water is first obtained by maintaining the air
velocity constant at approximately 350 fpm per
square foot of active tower area and by varying
the-water concentration (Perry, 1950). A secon-
dary operation calls for varying the air velocity to
meet the cooling requirements. The contact time
between water and air is a function of the time re -
quired for the water to be discharged from distribu-
tion nozzles and fall through a series of gridded decks
to the tower basin. Thus, the contact time is gov-
erned by the tower height. If the contact time is in-
sufficient, the ratio of air to water cannot be increased
to obtain the required cooling. A minimum cooling
tower height must be maintained. Where a wide ap-
proach (difference between the cold water tempera-
ture and the wet bulb temperature of the inlet air) of
15° to 20°F to the wet bulb temperature, and a
25° to 35°F cooling range (difference between the
temperature of the hot and cold water) are required,
a relatively low cooling tower is adequate (15 to 20
feet). Other ranges are shown in Table 188,
The cooling performance of a tower with a set
depth of packing varies with water concentration.
Maximum contact and performance have been
found with a water concentration of 2 to 3 gallons
of water per ininute per square foot of ground
area. The problem in designing a cooling tower
is one of determining the proper concentration
-------
674
PETROLEUM EQUIPMENT
Table 188. COOLING TOWER APPROACH
VERSUS WATER TRAVEL
Figure 504. Cutaway view of a mechanical-draft
cooling tower (Fluor Products Company, Inc.,
Santa Rosa, Calif.).
of water to obtain desired cooling. A high cool-
ing tower must be used if the water concentration
is less than 1. 6 gallons per square foot. Low
towers can be employed if the water concentration
exceeds 3 gallons per square foot. If the required
'water concentration is known, the tower area can
be found by dividing the water circulation rate
(gallons per minute) by the water concentration
(gallons per minute per square foot).
The required tower size (Perry, 1950) is thereby
dependent upon: (1) cooling range (hot water
minus cold water temperature); (2) approach
(cold water minus -wet bulb temperature); (3)
amount of liquid to be cooled; (4) wet bulb tem-
perature; (5) air velocity through cell; and (6)
tower height.
Various technical articles are available by which
a cooling tower may be designed for a specific
duty (Natural Gas Processors Suppliers Associa-
tion, 1957; Perry, 1950).
THE AIR POLLUTION PROBLEM
Cooling towers used in conjunction with equip-
ment processing hydrocarbons and their deriva-
tives are potential sources of air pollution be-
cause of possible contamination of the water. The
cooling water may be contaminated by leaks from
the process side of heat-exchange equipment, di-
Approach, °F
15 to 20
8 to 15
4 to 8
4a
Cooling range, °F
25 to 35
25 to 35
25 to 35
Water travel, ft
15 to 20
25 to 30
35 to 40
35 to 40
Designing cooling towers with an approach of less
than 4°F is not economical.
rect and intentional contact with process streams,
or improper process unit operation. As this
water is passed over a cooling tower, volatile
hydrocarbons and other materials accumulated
in the water readily evaporate into the atmosphere.
When odorous materials are contained in the water,
a nuisance is easily created.
Inhibitors or additives used in the cooling tower
to combat corrosion or algae growth should not
cause any significant air pollution emissions, nor
should the water-softening facilities common to
many cooling towers be a problem.
A survey (Bonamassa and Yee, 1957) of the oil
refineries operating in Los Angeles County in-
dicated hydrocarbon concentrations of approxi-
mately 20 percent in the cooling water of the
cooling towers (see Table 189). Cooling towers
in which hydrocarbons were detected were tested
quantitatively. Three tons of hydrocarbons per
day were found being discharged into the atmo-
sphere from these sources. Individually the emis-
sions varied from 4 to 1, 500 pounds per cooling
Table 189. HYDROCARBON EMISSIONS FROM
COOLING TOWERS (Bonamassa and Yee, 1957)
Cooling
tower
1
2
3
4
5
6
7
3
9
10
11
12
13
14
15
16
17
Water circulation,
gpm
14, 000
3, 120
28, 000
3, 000
1, 000
14, 000
14, 000
12, 000
13, 000
1, 000
15, 000
10, 000
3, 000
1, 800
700
1, 000
400
Hydrocarbon emissions,
Ib/day (as hexane)
1, 570
1, 400
700
616
532
31?
239
239
136
147
129
56
22
10
10
8
4
Total 6,236
GPO 806—614—23
-------
Miscellaneous Sources
675
tower per day. A study of operating variables
failed to indicate any correlation among the emis-
sions, the size of the tower, the water circulation
rate, or the particular duty of the tower. Apparent-
ly the amount of hydrocarbon present in the water
depends upon the state of maintenance of the pro-
cess equipment, particularly the heat-exchange
equipment, condensers, and coolers through which
the water is circulated. The quantity and type of
emissions should be determined by observing and
testing each tower individually.
One survey of the cooling towers in a designated
area is felt to be representative of the emissions
under existing operating conditions and mainte-
nance practices. The actual emission rate of any
specific tower and the degree of odor nuisance
vary as leaks develop, are detected, and repaired.
Overall leakage probably remains constant in view
of the large number of potential sources that can
cause new leaks even as the old ones are repaired.
product. The exhaust air is saturated with hydro-
carbon vapors or aerosols, and, if discharged
directly to the atmosphere, is a source of air
pollution. The extent of airblowing operations
and the magnitude of emissions from the equip-
ment vary widely among refineries. Results of
a survey (Kanter et al. , 1958) on the magnitude
of hydrocarbon emissions from airblowing of
petroleum fractions in Los Angeles County re-
fineries, presented in Table 190, show emis-
sions of less than 1/2 ton per day. These re-
fineries operated a total of seven airblowing
units "with a combined capacity of 25, 000 barrels
per day and a total airflow rate of 3, 300 cfm.
The tabulated results do not include airblowing
of asphalt, which has been discussed-elsewhere
in this chapter. Emissions from airblowing for
removal of moisture, or for agitation of products
may be minimized by replacing the airblowing
equipment with mechanical agitators and incin-
erating the exhaust vapors.
AIR POLLUTION CONTROL EQUIPMENT
The control of hydrocarbon discharges or of re-
lease of odoriferous compounds at the cooling
tower is not practical. Instead, the control must
be at the point where the contaminant enters the
cooling water. Hence, systems of detection of
contamination in water, proper maintenance,
speedy repair of leakage from process equipment
and piping, and good housekeeping programs in
general are necessary to minimize the air pollu-
tion occurring at the cooling tower. "Water that
has been used in contact with process streams,
as in direct-contact or barometric-type con-
densers, should be eliminated from the cooling
tower if this air pollution source is to be com-
pletely controlled. Greater use of fin-fan cool-
ers can also control the emissions indirectly by
reducing or eliminating the volume of cooling
•water to be aerated in a cooling tower.
MISCELLANEOUS SOURCES
A number of relatively minor sources of air pollu-
tion contribute approximately 10 percent of the
total hydrocarbon emissions to the atmosphere
from refineries (Kanter et al. , 1958). Six of
these sources, not discussed elsewhere in this
manual, include airblowing, blind changing, equip-
ment turnaround, tank cleaning, use of vacuum
jets, and use of compressor engine exhausts.
AIRBLOWING
In certain refining operations, air is blown through
heavier petroleum fractions (see Figure 505) for
the purpose of removing moisture or agitating the
AIR SATURATED VITH —^
HYDROCARBONS AND IATER
o o
O O
Figure 505. Improvement of product color by means
of air agitation, a source of air pollution.
BLIND CHANGING
Refinery operations frequently require that a
pipeline be used for more than one product. To
prevent leakage and contamination of a particular
product, other product-connecting and product-
feeding lines are customarily "blinded off. " "Blind-
ing a line" is the term commonly used for the in-
serting of a flat, solid plate between two flanges
-------
676
PETROLEUM EQUIPMENT
of a pipe connection. Blinds are normally used
instead of valves to isolate pipelines because a
more positive shutoff can be secured and because
of generally lower costs. In opening, or break-
ing, the flanged connection to insert the blind,
spillage of product in that portion of the pipeline can
occur. The magnitude of emissions to the atmosphere
from this spillage is a function of the vapor pres sur e
of the product, type of ground surface beneath the
blind, distance to the nearest drain, and amount of
liquid holdup in the pipeline.
Table 190. HYDROCARBON EMISSIONS
FROM AIRBLOWTNG OPERATIONS OF
REFINERIES IN LOS ANGELES
COUNTY (Kanter et al. , 1958)
Number of units
Refinery A (one unit)
Refinery B (five units)
Refinery C (one unit)
Total
Emissions,
905
35
2
942
Ib/day
Results of a survey (Kanter et al. , 1958) conducted
to evaluate the emissions from blind changing in
Los Angeles County refineries indicated that a
wide variation exists in the number of pipeline
service and corresponding blind changes and in
the amount of spillage for different refineries
of comparable size. The average emission from
blind changing in Los Angeles County refineries
was calculated at 0. 1 ton per day.
Emissions to the atmosphere from the changing
of blinds can be minimized by pumping out the
pipeline and then flushing the line with water be-
fore breaking the flange. In the case of highly
volatile hydrocarbons, a slight vacuum may be
maintained in the line. Spillage resulting from
blind changing can also be minimized by use of
"line" blinds in place of the common "slip" blinds.
Line blinds, depicted in Figure 506, do not re-
quire a complete break of the flange connection
during the changing operation. These blinds use
a gear mechanism to release the spectacle plate
without actually breaking the line. Combinations
of this device in conjunction with gate valves are
available to allow changing of the line blind while
the line is under pressure from either direction.
The line blind is finding many applications in
new process equipment where frequent changes
in services of pipelines occur. Data compiled
during the survey (Kanter et al. , 1958) indicate
that slip blinds spilled an average of 5 gallons
per change compared with line blind valves,
which spilled an average of 2 gallons per change.
EQUIPMENT TURNAROUNDS
Periodic maintenance and repair of process
equipment are essential to refinery operations.
A major phase of the maintenance program is
the shutting down and starting up of the various
units, usually called a turnaround.
The procedure for shutting down a unit varies
from refinery to refinery and between units in
a refinery. In general, shutdowns are effected
by first shutting off the "heat supply to the unit
and circulating the feed stock through the unit
Figure 506. Typical line blind valve (Haraer Oi
Tool Company, Long Beach, Calif.).
-------
Miscellaneous Sources
677
as it cools. Gas oil may be blended into the
feedstock to prevent its solidification as the
temperature drops. The cooled liquid is then
pumped out to storage facilities, leaving hydro-
carbon vapors in the unit. The pressure of the
hydrocarbon vapors in the unit is reduced by
evacuating the various items of equipment to a
disposal facility such as a fuel gas system, a va-
por recovery system, a flare, or in some cases,
to the atmosphere. Discharging vapors to the
atmosphere is undesirable from the standpoint
of air pollution control since as much as sev-
eral thousand pounds of hydrocarbons or other
objectionable vapors or odors can be released
during a shutdown. The residual hydrocarbons
remaining in the unit after depressuring are
purged out with steam., nitrogen, or water. Any
purged gases should be discharged to the afore-
mentioned disposal facilities. Condensed steam
and water effluent that may be contaminated with
hydrocarbons or malodorous compounds during
p.urging should be handled by closed water-treat-
ing systems.
Results of a survey (Kanter et al. , 1958) to de-
termine the magnitude of hydrocarbon emissions
from turnarounds in Los Angeles County refin-
eries showed emissions totaling a maximum of
254 tons per year or 0. 7 ton per day. Sixty per-
cent of all shutdowns were found to occur on
Sunday and Monday. On this basis, the 2-day
emissions totaled 3 tons or 152 tons per year.
TANK CLEANING
Storage tanks in a refinery require periodic clean-
ing .and repair. For this purpose, the contents of
a tank are removed and residual vapors are purged
until the tank is considered safe for entry by main-
tenance crews. Purging can result in the release
of hydrocarbon or odorous material in the form
of vapors to the atmosphere. These vapors should
be discharged to a vapor recovery system or flare.
Data obtained from the refinery survey (Kanter
et al. , 1958) were used to estimate the quantity
of hydrocarbon emissions to the atmosphere
from tank cleaning as follows:
1. When the vapors in the tank were released
to a recovery or disposal system before the
tank was opened for maintenance, the emis-
sions were considered negligible.
2. When the stored liquid was transferred to
another tank, and the emptied vessel was
opened for maintenance without purging to
a recovery or disposal system, the emis-
sion to the atmosphere was considered to be
equal to the weight of hydrocarbon vapor
occupying the total volume of the tank at the
reported pressure. (For floating-roof tanks,
the minimum volume was used. )
3. For vapor storage, when tanks were not purged
to a recovery or disposal system, estimates
were made as described in item 2.
The calculated emissions, for an average of 174
tanks cleaned per year, were 1. 3 tons of hydro-
carbons per day.
Steam cleaning of railroad tank cars used for
transporting petroleum products can similarly be
a source of emissions if the injected steam and
entrained hydrocarbons are vented directly to
the atmosphere. Although no quantitative data
are available to determine the magnitude of these
emissions, the main objection to this type of
operation is its nuisance-causing potential. Some
measure of control of these emissions may be
effected by condensing the effluent steam and
vapors. The condensate can then be separated
into hydrocarbon and water phases for recovery.
Noncondensable vapors should be incinerated.
USE OF VACUUM JETS
Certain refinery processes are conducted under
vacuum conditions. The most practical way to
create and maintain the necessary vacuum is to
use steam-actuated vacuum jets, singly or in
series (see Figure 507). Barometric condensers
are often used after each vacuum jet to remove
steam and condensable hydrocarbons.
The effluent stream from the last stage of the
vacuum jet system should be controlled by con-
densing as much of the effluent as is practical
and incinerating the noncondensables in an after-
burner or heater firebox. Condensate should be
handled by a closed treating system for recovery
of hydrocarbons. The hot -well that receives
water from the barometric condensers may also
have to be enclosed and any off gases incinerated.
USE OF COMPRESSOR ENGINE EXHAUSTS
Refining operations- require the use of various
types of gas compressors. These machines are
often driven by internal combustion engines that
exhaust air contaminants to the atmosphere. Al-
though these engines are normally fired with nat-
ural gas and operate at essentially constant loads,
some unburned fuel passes through the engine.
Oxides of nitrogen are also found in the exhaust
gases as a result of nitrogen fixation in the com-
bustion cylinders.
Results of a survey (Kanter et al. , 1958) con-
ducted to determine the contribution made by
compressor engine exhausts to overall emis-
-------
678
PETROLEUM EQUIPMENT
sions from refineries are presented in Table
191. The composition of the hydrocarbons shown
was generally over 90 percent methane.
In addition to the compounds listed in the table,
aldehydes and ammonia may also be present in
engine exhausts. Test data on these components
were, however, inconclusive.
Table 191. EMISSIONS FROM COMPRESSOR
INTERNAL COMBUSTION ENGINES IN
LOS ANGELES COUNTY REFINERIES
(Kanter et al. , 1958)
Number of compressor engines
Fuel gas burned, mcfd
Exhaust gas, scfm
Contaminants in exhaust gases, ppm
Hydrocarbons
Oxides of nitrogen, as NC>2
130
10,500
165,000
1, 240
315
STEAM
STEAM
SUCTION
T
NONCONOENSABLES TO
FUME INCINERATOR
WATER AND CONDENSABLES
Figure 507. Schematic drawing of a two-stage,
steam-actuated vacuum jet.
-------
CHAPTER 11
CHEMICAL PROCESSING EQUIPMENT
RESIN KETTLES
HARRY E0 CHATFIELD, Air Pollution Engineer
VARNISH COOKERS
HARRY E. CHATFIELD, Air Pollution Engineer
SULFURIC ACID MANUFACTURING
ROBERT J. MAC KNIGHT
Principal Air Pollution Engineer
STANLEY T. CUFFE, Air Pollution Engineer*
PHOSPHORIC ACID MANUFACTURING
EMMET F. SPENCER, JR.
Intermediate Air Pollution Engineer'
RAY M. INGELS, Air Pollution Engineer!
PAINT-BAKING OVENS
JULIEN A. VERSSEN, Air Pollution Engineer
SOAPS AND SYNTHETIC DETERGENTS
ROBERT C. MURRAY, Senior Air Pollution Engineer
EDWIN J. VINCENT
Intermediate Air Pollution Engineer
GLASS MANUFACTURE
ARTHUR B. NETZLEY
Intermediate Air Pollution Engineer
JOHN L. MC GINNITY
Intermediate Air Pollution Engineer*
FRIT SMELTERS
JOHN L. SPINKS, Air Pollution Engineer
FOOD-PROCESSING EQUIPMENT
W. L. POLGLASE, Air Pollution Engineer
H. F. DEY, Air Pollution Engineer
ROBERT T. WALSH, Senior Air Pollution Engineer
FISH CANNERIES AND FISH REDUCTION PLANTS
ROBERT T, WALSH, Senior Air Pollution Engineer!
KARL D. LUEDTKE
Intermediate Air Pollution Engineer
LEWIS K. SMITH, Air Pollution Engineer
REDUCTION OF INEDIBLE ANIMAL MATTER
ROBERT T. WALSH, Senior Air Pollution Engineer
PAUL G. TALENS, Air Pollution Engineer
ELECTROPLATING
EMMET F. SPENCER, JR.
Intermediate Air Pollution Engineer
GEORGE THOMAS
Intermediate Air Pollution Engineer
INSECTICIDE MANUFACTURE
WILLIAM C. BAILOR, Air Pollution Engineer
JOSEPH D'IMPERIO, Air Pollution Engineerll
HAZARDOUS RADIOACTIVE MATERIAL
WILLIAM C. BAILOR, Air Pollution Engineer
OIL AND SOLVENT RE-REFINING
JOSEPH D'IMPERIO, Air Pollution Engineer
CHEMICAL MILLING
GEORGE THOMAS
Intermediate Air Pollution Engineer
*Now with National Center for Air Pollution Control, Public Health Service, U. S. Department of Health,
Education, and Welfare, Cincinnati, Ohio.
fNow with FMC Corporation, 633 Third Ave. , New York, N. Y.
jNow with State of California Vehicle Laboratories, 434 S. San Pedro St. , Los Angeles, Calif.
§Now with New York-New Jersey Air Pollution Abatement Activity, National Center for Air Pollution
Control, Public Health Service, U. S. Department of Health, Education, and Welfare, Raritan Depot,
Metuchen, N. J.
llNow deceased.
-------
CHAPTER II
CHEMICAL PROCESSING EQUIPMENT
RESIN KETTLES
resisting qualities to cross-linked molecular
structures.
TYPES OF RESINS
A resin is defined by the American Society for
Testing Materials (ASTM) as a solid or semi-
solid, water-insoluble, organic substance, with
little or no tendency to crystallize. Resins are
the basic components of plastics and are impor-
tant components of surf ace-coating formulations.
For both uses, growth in recent years has been
phenomenal; more than 5, 000 companies in the
United States now produce plastics.
There are two types of resins--natural and syn-
thetic. The natural resins are obtained directly
from sources such as fossil remains and tree sap.
These include Congo, Batu, and East India resins
from fossils; lac from insects; and damar and
rosin from tree sap. Synthetic resins are those
made by man.
Synthetic resins can be classified by physical
properties as thermoplastic or thermosetting.
Thermoplastic resins undergo no permanent
change upon heating. They can be softened,
melted, and molded into shapes they retain
upon cooling, without change in their physical
properties. Thermosetting resins, on the other
hand, can be softened, melted, and molded upon
heating, but upon continued heating, they harden
or set to a permanent, rigid state and cannot be
remolded.
In this section, several synthetic resins are
discussed briefly (Kirk and Othmer, 1947;
Plastics Catalog Corporation, 1959; Shreve,
1956). For each, an example of ingredients is
given and a typical manufacturing operation is
discussed. Each basic resin type requires many
modifications both in ingredients and techniques
of synthesis in order to satisfy proposed uses
and provide desired properties. Not all these
variations, however, will be discussed since
not all present individual air pollution problems.
Thermosetting resins are obtained from fusible
ingredients that undergo condensation and poly-
merization reactions under the influence of heat,
pressure, and a catalyst and form rigid shapes
that resist the actions of heat and solvents.
These resins, including phenolic, amino, poly-
ester, and polyurethane resins, owe their heat-
Phenolic Resins
Phenolic resins can be made from almost any
phenolic compound and an aldehyde. Phenol and
formaldehyde are by far the most common in-
gredients used, but others include phenol-fur-
fural, resorcinol-formaldehyde, and many simi-
lar combinations. Since a large proportion of
phenolic-resin production goes into the manu-
facture of molding materials, the most desirable
process for this manufacture will be described.
Phenol and formaldehyde, along with an acid
catalyst (usually sulfuric, hydrochloric, or
phosphoric acid), are charged to a steam-
jacketed or otherwise indirectly heated resin
kettle that is provided with a reflux condenser
and is capable of being operated under vacuum.
The following formula shows the basic reaction:
OH
H2S04
PHENOL
OH
FORMALDEHYDE
HO
H20
TYPICAL (INTERMEDIATE) CONDENSATION PRODUCT
Heat is applied to start the reaction, and then
the exothermic reaction sustains itself for a
while without additional heat. Water formed
during the reaction is totally refluxed to the
kettle. After the reaction is complete, the upper
layer of water in the kettle is removed by draw-
ing a vacuum on the kettle. The warm., dehy-
drated resin is poured onto a cooling floor or
into shallow trays and then ground to powder
after it hardens. This powder is mixed with
other ingredients to make the final plastic mate-
rial. Characteristics of the molding powder,
as well as the time and rate of reaction, depend
upon the concentration of catalyst used, the
phenol-formaldehyde ratio used, and the reac-
tion temperature maintained.
681
-------
682
CHEMICAL PROCESSING EQUIPMENT
Amino Resins
The most important amino resins are the urea-
formaldehyde and melamine-formaldehyde resins.
The urea-formaldehyde reaction is simple: 1
mole of urea is mixed with 2 moles of formalde-
hyde as 38 percent solution. The mixture is
kept alkaline with ammonia pH 7.6 to 8. The
reaction is carried out at 77° F for 2 days at
atmospheric pressure without any reflux.
The melamine resins are made in much the same
manner except that the reactants must be heated
to about 176°F initially, in order to dissolve the
melamine. The solution is then cooled to 77 °F
for 2 days to complete the reaction.
The equipment needed for the synthesis of the
amino resins consists of kettles for the conden-
sation reaction (usually nickel or nickel-clad
steel), evaporators for concentrating the resin,
and some type of dryer.
The amino resins are used as molding compounds,
adhesives, and protective coatings, and for treat-
ing textiles and paper.
Polyester and Alkyd Resins
There is much confusion concerning the mean-
ing of the two terms polyester and alkyd. Ap-
parently, by chemical definition, the product
obtained by the condensation reaction between
a polyhydric alcohol and a polybasic acid, whether
or not it is modified by other materials, is prop-
erly called a polyester. All polyesters can then
be divided into three basic classes: Unsaturated
polyesters, saturated polyesters, and alkyds.
1. Unsaturated polyesters are formed when
either of the reactants (alcohol and acid)
contains, or both contain, a double-bonded
pair of carbon atoms. The materials usu-
ally used are glycols of ethylene, propylene,
and butylene and unsaturated dibasic acids
such as maleic anhydride and fumaric acid.
A typical reaction is as follows:
HC - Cx
HC - C'"0
8
MALEIC ANHYDRIDE
H
HC - OH
HC - OH
ETHYLENE GLYCOL
0 0 u H
ii n n "
— C - C = C - C - 0 - C - C - 0 ---*•
H H H H
REPRESENTATIVE SEGMENT OF CHAIN-FORMED
2.
The resulting polyester is capable of cross-
linking and is usually blended with a poly-
merizable material such as styrene. Under
heat or a peroxide catalyst, or both, this
blend copolymerizes into a thermosetting
resin. It has recently found extensive use
in the reinforced-plasti.cs field where it is
Laminated with fibrous glass. It is also
molded into many forms for a variety of uses.
Saturated polyesters are made from saturated
acids and alcohols, as indicated by the follow-
ing reaction:
00
II / - \ II
H - 0 - C -( >- C
\ - /
TEREPHTHALIC ACID
u u
" "
i.° " H
0
\ II
+ H1,-
H
0 - C - C - 0 - H
ETHYLENE GLYCOL
H
ON " "
- C - 0 - C - C ---•
H H
POLYESTER (REPEATING UNIT)
H20
The polyesters formed are long-chain, sat-
urated materials not capable of cross-linking.
Several of these are used as plasticizers. A
special type made from ethylene glycol and
terephthalic acid has been made into fiber
(Dacron) and fiJm (MylarvB^. Still others of
this type with lower molecular "weights are
being used with di-isocyanates to form poly-
urethane resins.
Alkyd resins differ from, other polyesters
as a result of modification by additions of
fatty, monobasic acids. This is known as oil
modification since the fatty acids are usu-
ally in the form of naturally occurring oils
such as linseed, tung, soya, cottonseed, and,
at times, fish oil. The alkyds, thinned with
organic solvents, are used predominantly in
the protective coating industry in varnishes,
paints, and enamels.
The most widely used base ingredients are
phthalic anhydride and glycerol. Smaller
quantities of other acids such as maleic,
fumaric, and others and alcohols such as
pentaerythritol, sorbitol, mannitol, ethylene
glycol, and others are used. These are re-
acted with the oils already mentioned to
form the resin.
The oils, as they exist naturally, are pre-
dominantly in the form of triglycerides and
do not react with the polybasic acid. They
are changed to the reactive monoglyceride
by reaction with a portion of the glycerol or
other alcohol to be used. Heat and a cata-
lyst are needed to promote this reaction,
-------
Resin Kettles
683
which is known as alcoholysis. The resin
is then formed by reacting this monoglyceride
•with the acid by agitation and sparging with
inert gas until the condensation reaction prod-
uct has reached the proper viscosity. The
reaction takes place in an enclosed resin ket-
tle equipped with a condenser and usually a
scrubber, at temperatures slightly below
500 °F. The alcoholysis can be accomplished
first and then the acid and more alcohol can
be added to the kettle, or all the ingredients
can be added simultaneously.
An example of an alcoholysis reaction followed
by reaction of the monoglyceride formed with
phthalic anhydride is shown in the following:
C3 H5 (C17 H33 C00)3 + C3 H5 (OH)3
GLYCEROL TRIESTER OF OLEIC ACID GLYCERIN
H
HC - OH 0
•HC - 0 - C - C17 H33
HC - OH
H
a
MONOGLVCERIDE OF OLEIC ACID PHTHALIC ANHYDRIDE
H H H
i 0 - C - C - C - 0 -
0
C17 H33 ~ C = °
REPEATING UNIT FOR OIL-MODIFIED ALKYD
Polyurethane
The manufacture of the finished polyurethane
resin differs from the others described in that
no heated reaction in a kettle is involved. One
of the reactants, however, is a saturated poly-
ester resin, as already mentioned, or, more
recently, a polyether resin. To form a flexible
foam product, the resin, typically a polyether
such as polyoxypropylenetriol, is reacted with
tolylene diisocyanate and water in an approximate
100: 42: 3 ratio by -weight, along with small quanti-
ties of an emulsifying agent, a polymerization
catalyst, and a silicone lubricant. The ingredi-
ents are metered to a mixing head that deposits
the mixture onto a moving conveyor. The resin
and tolylene diisocyanate (TDI) polymerize and
cross-link to form the urethane resin. The TDI
also reacts with the water, yielding urea and
carbon dioxide. The evolved gas forms a foam-
like structure. The product forms as a contin-
uous loaf. After room temperature curing for
about a day, the loaf can be cut into desired
sizes and shapes, depending upon required use.
The flexible foams have found -wide use in auto-
mobile and furniture upholstery and in many
other specialty items.
By varying the ingredients and adding other blow-
ing agents such as Freon 11, rigid foams with
fine, close-cell structure can be formed. These
can be formed in place by spraying techniques
and are used extensively as insulating materials.
Thermoplastic Resins
As already stated, thermoplastic resins are
capable of being reworked after they have been
formed into rigid shapes. The subdivisions in
this group that are discussed here are the vinyls,
styrenes, and the coal tar and petroleum base
Polyvinyl Resins
The polyvinyl resins are those having a vinyl
(CH=CH2) group. The most important of these
are made from the polymerization of vinyl ace-
tate and vinyl chloride. Other associated resins
are also discussed briefly.
Vinyl acetate monomer is a clear liquid made
from the reaction between acetylene and acetic
acid. The monomer can be polymerized in bulk,
in solution, or in beads or emulsion. In the bulk
reaction, only small batches can be safely han-
dled because of the almost explosive violence of
the reaction once it has been catalyzed by a small
amount of peroxide. Probably the most common
method of preparation is in solution. In this
process, a mixture of 60 volumes vinyl acetate
and 40 volumes benzene is fed to a jacketed,
stirred resin kettle equipped -with a reflux con-
denser. A small amount of peroxide catalyst
is added and the mixture is heated until gentle
refluxing is obtained. After about 3 hours, ap-
proximately 70 percent is polymerized, and the
run is transferred to another kettle where the
solvent and unreacted monomer are removed by
steam distillation. The wet polymer is then
dried. Polyvinyl acetate is used extensively in
water-based paints, and for adhesives, textile
finishes, and production of polyvinyl butyral.
Vinyl chloride monomer under normal conditions
is a gas that boils at -14°C. It is usually stored
and reacted as a liquid under pressure. It is
made by the catalytic combination of acetylene
and hydrogen chloride gas or by the chlorination
of ethylene followed by the catalytic removal of hy-
drogen chloride. It is polymerized in a. jacketed,
stirred autoclave. Since the reaction is highly exo-
thermic and can result in local overheating and poor
quality, it is usually carried out as a water emul-
sion to facilitate more precise control. To ensure
-------
684
CHEMICAL PROCESSING EQUIPMENT
quality and a properly controlled reaction, several
additives are used. These include an emulsifying
agent such as soap, a protective colloid such as glue,
a pH control such as acetic acid or other moderate-
ly weak acid (2. 5 is common), oxidation and re-
duction agents such as ammonium persulfate and
sodium bisulfite, respectively, to control the oxi-
dation-reduction atmosphere, a catalyst or initia-
tor like benzoyl peroxide, and a chain length-con-
trolling agent such as carbon tetrachloride. The
reaction is carried out in a completely enclosed
vessel with the pressure controlled to maintain
the unreacted vinyl chloride in the liquid state.
As the reaction progresses, a suspension of latex
or polymer is formed. This raw latex is removed
from the kettle, and the unreacted monomer is
removed by evaporation and recovered by com-
pression and condensation.
A modification of the emulsion reaction is known
as suspension polymerization. In this process,
droplets of monomer are kept dispersed by rapid
agitation in a -water solution of sodium sulfate or
in a colloidal suspension such as gelatin in water.
During the reaction, the droplets of monomer are
converted to beads of polymer that are easily re-
covered and cleaned. This process is more
troublesome and exacting than the emulsion reac-
tion but eliminates the contaminating effects of
the emulsifying agent and other additives.
Other vinyl-type resins are polyvinylidene chloride
(Saran®), polytetrafluoroethylene (fluoroethene),
polyvinyl alcohol, polyvinyl butyral, and others.
The first two of these are made by controlled poly-
merization of the monomers in a manner similar to
that previously described for polyvinyl chloride.
Polyvinyl alcohol has no existing monomer and
is prepared from polyvinyl acetate by hydrolysis.
Polyvinyl alcohol is unique among resins in that
it is completely soluble in both hot and cold water.
Polyvinyl butyral is made by the condensation
reaction of butyraldehyde and polyvinyl alcohol.
All have specific properties that make them super-
ior for certain applications.
Polystyrene
Polystyrene, discovered in 1831, is one of the
oldest resins known. Because of its transparent,
glasslike properties, its practical application
was recognized even then. Two major obstacles
prevented its commercial development--prepara-
tion of styrene monomer itself, and some means
of preventing premature polymerization. These
obstacles were not overcome until nearly 100
years later.
Styrene is a colorless liquid that boils at 145°C.
It is prepared commercially from ethylbenzene,
which, in turn, is made by reaction of benzene
with ethylene in presence of a Fridel-Crafts cata-
lyst such as aluminum chloride. During storage
or shipment the styrene must contain a polymeriza-
tion Inhibitor such as hydroquinone and must be
kept under a protective atmosphere of nitrogen
or natural gas.
Styrene can be polymerized in bulk, emulsion,
or suspension by using techniques similar to
those previously described. The reaction is
exothermic and has a runaway tendency unless
the temperature is carefully controlled. Oxygen
must be excluded from the reaction since it causes
a yellowing of the product and affects the rate of
polymerization.
Polystyrene is used in tremendous quantities for
many purposes. Because of its ease of handling,
dimensional stability, and unlimited color possi-
bilities, it is used widely for toys, novelties,
toilet articles, houseware parts, radio and tele-
vision parts, wall tile, and other products. Dis-
advantages include limited heat resistance, brit-
tleness, and vulnerability to attack by organic
solvents such as kerosine and carbon tetrachloride
Petroleum and Coal Tar Resins
Petroleum and coal tar resins are the least ex-
pensive of the synthetic resins. They are made
from the polymerization of unsaturated hydrocarbons
found in crude distillate from coal tar in coke ovens
or from cracking of petroleum. The exact chemical
nature of these hydrocarbons has not been deter-
mined, but the unsaturates of coal .tar origin are
known to be primarily cyclic -while petroleum deriva-
tives are both straight- and close-chain types.
Most typical of the coal tar resins are those
called Coumarone-Indene resin because these
two compounds constitute a large portion of the
distillate used for the reaction. The polymeriza-
tion is initiated by a catalyst (usually sulfuric
acid). After the reaction has proceeded as far as
is desired, the unreacted monomer is removed
by distillation. By controlling time, temperature,
and proportions, many modifications of color and
physical characteristics can be produced. The
petroleum base distillate is polymerized in the
same manner, yielding resins of slightly lower
specific gravity than that of the coal tar resins.
These resins are used in coating adhesives, in
oleoresinous varnishes, and in floor coverings
(the so-called asphalt tile).
Resin-Manufacturing Equipment
Most resins are polymerized or otherwise reacted
in a stainless steel, jacketed, indirectly heated
vessel, -which is completely enclosed, equipped
-------
Resin Kettles
685
with a stirring mechanism, and generally contains
an integral reflux condenser (Figure 508). Since
most of the reactions previously described are
exothermic, cooling coils are usually required.
Some resins, such as the phenolics, require
that the kettle be under vacuum during part of
the cycle. This can be supplied either by a vac-
uum pump or by a steam or water jet ejector.
Moreover, for some reactions, that of polyvinyl
chloride for example, the vessel must be capable
of being operated under pressure. This is nec-
essary to keep the normally gaseous monomer in
a liquid state. The size of resin-processing ket-
tles varies from a few hundred to several thou-
sand gallons' capacity.
Because of the many types of raw materials,
ranging from gases to solids, storage facilities
vary accordingly—ethylene, a gas, is handled
as such; vinyl chloride, a gas at standard condi-
tions, is liquefied easily under pressure. It is
stored, therefore, as a liquid in a pressurized
vessel. Most of the other liquid monomers do
not present any particular storage problems.
Some, such as styrene, must be stored under an
inert atmosphere to prevent premature poly-
merization. Some of the more volatile mate-
rials are stored in cooled tanks to prevent ex-
cessive vapor loss. Some of the materials have
strong odors, and care must be taken to prevent
emission of odors to the atmosphere. Solids,
such as phthalic anhydride, are usually packaged
and stored in bags or fiber drums.
Treatment of the resin after polymerization varies
with the proposed use. Resins for moldings are
dried and crushed or ground into molding powder.
Resins, such as the alkyd resins, to be used for
protective coatings are normally transferred to
an agitated thinning tank, as shown in Figure 509,
•where they are thinned with some type of solvent
and then stored in large steel tanks equipped
with -water-cooled condensers to prevent loss of
solvent to the atmosphere (Figure 510). Still
other resins are stored in latex form as they
come from the kettle.
Figure 508. Typical resin-manufacturing unit
showing process kettle and liquid feed tanks
(Silmar Chemical Company, Hawthorne, Calif.).
THE AIR POLLUTION PROBLEM
The major sources of possible air contamination
in resin manufacturing are the emissions of raw
materials or monomer to the atmosphere, emis-
sions of solvent or other volatile liquids during
the reaction, emissions of sublimed solids such
as phthalic anhydride in alkyd production, emis-
sions of solvents during thinning of some resins,
and emissions of solvents during storage and
handling of thinned resins. Table 192 lists the
most probable types and sources of air contami-
nants from various resin-manufacturing opera-
tions.
In the formulation of polyurethane foam, a slight
excess of tolylene diisocyanate is usually added.
Some of this is vaporized and emitted along
with carbon dioxide during the reaction. The
TDI fumes are extremely irritating to the eyes
and respiratory system and are a source of local
air pollution. Since the vapor pressure of TDI
is small, the fumes are minute in quantity and,
if exhausted from the immediate work area and
discharged to the outside atmosphere, are soon
diluted to a nondetectible concentration. No
specific controls have been needed to prevent
emission of TDI fumes to the atmosphere.
The finished solid resin represents a very small
problem--chiefly some dust from crushing and
grinding operations for molding powders. Gen-
erally the material is pneumatically conveyed
from the grinder or pulverizer through a cyclone
separator to a storage hopper. The fines escap-
ing the cyclone outlet are collected by a baghouse-
type dust collector. The collector should be de-
signed for a filter velocity of about 4 fpm or less.
-------
686
CHEMICAL PROCESSING EQUIPMENT
Figure 509. Resin-thinning tanks with water-cooled condensers
Lynnwood, Calif.).
(Allied Chemical Corp., Plastics Div.
Figure 510. Resin storage tanks with condensers
(Allied Chemical Corp., Plastics Division, Lynn-
wood, Calif.).
Most of the contaminants are readily condensable.
In addition to these, however, small quantities
of noncondensable, odorous gases similar to those
from varnish cooking may be emitted. These are
more prevalent in the manufacture of oil-modi-
fied alkyds where a drying oil such as tung, lin-
seed, or soya is reacted with glycerin and phtha-
lic anhydride. When a drying oil is heated,
acrolein and other odorous materials are emitted
at temperatures exceeding about 350°F (see
further discussion under Varnish Cookers). The
intensity of these emissions Is directly propor-
tional to maximum reaction temperatures. Thus,
the intensity of noncondensable gases from resin
formulation should be considerably less than
that of gases from varnish cooking since the re-
action temperature is approximately 100°F lower.
AIR POLLUTION CONTROL EQUIPMENT
Control of monomer and volatile solvent emis-
sions during storage before the reaction and of
-------
Resin Kettles
687
Table 192. PRINCIPAL AIR CONTAMINANTS AND SOURCES OF EMISSION FROM
RESIN-MANUFACTURING OPERATIONS
Resin
Air contaminant
Possible sources
of emission
Phenolic
Amino
Polyester and alkyds
Polyvinyl acetate
Polyvinyl chloride
Polystyrene
Petroleum and coal
tar resins
Polyurethane resins
Aldehyde odor
Aldehyde odor
Oil-cooking odors
Phthalic anhydride fumes
Solvent
Vinyl acetate odor
Solvent
Vinyl chloride odor
Styrene odor
Monomer odors
Tolylene diisocyanate
Storage, leaks, condenser outlet,
vacuum pump discharge
Storage, leaks
Uncontrolled resin kettle discharge
Kettle or condenser discharge
Storage, condenser outlet during
reaction, condenser outlet during
steam distillation to recover sol-
vent and unreacted monomer
Leaks in pressurized system
Leaks in storage and reaction
equipment
Leaks in storage and reaction
equipment
Emission from finished foam result-
ing from excess TDI in formulation
solvent emissions during thinning and storage
after the polymerization of the resin is relatively
simple. It involves care in maintaining gastight
containers for gases or liquefied gases stored
under pressure, and condensers or cooling coils
on other vessels handling liquids that might vapor-
ize. Since most resins are thinned at elevated
temperatures near the boiling point of the thinner,
resin-thinning tanks, especially, require ade-
quate condensers. Aside from the necessity for
control of air pollution, these steps are needed
to prevent the loss of valuable products.
Heated tanks used for storage of liquid phthalic
and maleic anhydrides should be equipped with
condensation devices to prevent losses of sub-
limed material. An excellent device is a water -
jacketed, vertical condenser with provisions for
admitting steam to the jacket and provisions for
a pressure relief valve at the condenser outlet
set at perhaps 4 ounces' pressure. During stor-
age the tank is kept under a slight pressure of
about 2 ounces, an inert gas making the tank
completely closed. During filling, the displaced
gas, with any sublimed phthalic anhydride, is forced
through the cooled condenser where the phthalic is
deposited on the condenser walls. After filling is
completed, the condensed phthalic is remelted by
passing steam through the condenser jacket.
Addition of solids such as phthalic anhydride to
other ingredients that are above the sublimation
temperature of the phthalic anhydride causes
temporary emissions that violate most air pollu-
tion standards regarding opacity of smoke or
fumes. These emissions subside somewhat as
soon as the solid is completely dissolved but re-
main in evidence at a reduced opacity until the
reaction has been completed. The emissions
can be controlled fairly easily with simple scrub-
bing devices. Various types of scrubbers can
be used. A common system that has been proved
effective consists of a settling chamber, com-
monly called a resin slop tank, followed by an
exhaust stack equipped with water sprays. The
spray system should provide for at least 2 gallons
per 1, 000 scf at a velocity of 5 fps. The settling
chamber can consist of an enclosed vessel par-
tially filled with water capable of being circulated
with gas connections from the reaction vessel and
to the exhaust stack. Some solids and water of
reaction are collected in the settling tank, the
remainder being knocked down by the water sprays
in the stack. Another example is shown in Fig-
ure 511. Here the vapors from a polyester resin
process kettle are first passed through a spray
chamber-type precleaner folio-wed by a venturi
scrubber. This system effectively reduces visi-
ble emissions. Scrubber water may be recircu-
lated or used on a once-through basis, depend-
ing primarily upon the available waste-water dis-
posal system. The scrubber water can be odor-
ous and should be discharged to a sanitary sewer.
Many resin polymerization reactions, for example,
polyvinyl acetate by the solution method, require
-------
688
CHEMICAL PROCESSING EQUIPMENT
Figure 511. Venturi scrubber venting resin-man-
ufacturing equipment (Silmar Chemical Corporation,
Hawthorne, Calif.).
refluxing of ingredients during the reaction. Thus,
all reactors for this or other reactions involving
the vaporization of portions of the reactor con-
tents must be equipped with suitable reflux- or
horizontal-type condensers or a combination of
both. The only problems involved here are prop-
er sizing of the condensers and maintaining the
cooling medium at the temperature necessary to
effect complete condensation.
When noncondensable, odor-bearing gases are
emitted during the reaction, especially with alkyd
resin production as already mentioned, and these
gases are in sufficient concentration to create
a public nuisance, more extensive air pollution
control equipment is necessary. These are
discussed thoroughly under other sections con-
cerning odors (Varnish Cookers and Reduction of
Inedible Animal Matter) and include equipment for
absorption and chemical oxidation, adsorption, and
combustion, both catalytic and direct-flame type.
VARNISH COOKERS
INTRODUCTION
Varnish as a finished product is defined as a
transparent, homogeneous, heat-processed
blend of drying oil, resin, drier, and solvent.
When the varnish is applied as a thin film, the
solvent evaporates, and the remaining mate-
rials oxidize and polymerize to form a hard,
solid, continuous, transparent coating (Kirk
and Othmer, 1947). In the protective coating
industry, the term varnish is also used to
describe a base or vehicle for pigmented coat-
ings. In this form, the drier and most of the
solvent are usually omitted until the final for-
mulation of the pigmented coating.
The uses of varnish are many and varied. It is
commonly used where a transparent coating is
desired for visible surfaces such as furniture
and floor coatings and overprint for paper labels.
In other applications, the surface is not visible,
but varnish is more economical and gives better
protection than pigmented coatings do. These
include metal-container coatings, insulating
varnishes, and bottle cap liners.
Historically, varnish meant one type of prod-
uct, the oleoresinous varnish (oil plus resin).
This product, by definition, was also required
to dry to a transparent film. Recently, many
other products have been developed that are
called varnishes but do not meet the require-
ments of this definition. Some of the most im-
portant ones are now listed.
1. Spirit varnish is a solution of a resin with
little or no oil in a volatile solvent. It
normally dries to a hard, brittle finish. The
most common varnish of this type is shellac,
which is a solution of the natural lac resin in
denatured alcohol.
2. ATkyd resin varnish is a. solution of alkyd
resin in a volatile solvent with added drier.
It is similar to conventional varnish in that
an oil-modified alkyd resin is used. There
is a marked difference "between alkyd resin
varnish and oleoresinous varnish in their
raw materials and in their manufacturing
processes. The alkyd varnishes, however,
have properties similar to those of oleo-
resinous varnishes and are used primarily
for the same purposes.
3. Asphalt varnish is a solution of asphalt in
volatile solvent. It is formed at high tem-
peratures of 300° to 500 °F and is used as
black enamel where low cost and excellent
chemical resistance are desired.
-------
Varnish Cookers
689
4. Lithograph varnish is manufactured from
linseed oil, -which is oxidized and poly-
merized to higher viscosity and then blended
with drier and resin. It is used as a vehicle
for pigmented lithographing printing ink.
Raw Materials for Varnish Making
Oleoresinous varnishes, in their final form ready
for application as surface coatings, are composed
of four groups of materials: Resins, oils, sol-
vents, and driers.
The varnish-making resins are hard, brittle,
noncrystalline solids that usually melt and dis-
solve in organic solvent. Their functions in the
formulation are to contribute to the drying speed,
hardness, toughness, and gloss. The resins
most commonly used are the natural resins, such
as Rosin, Congo, Batu, East India, Dammar,
and Lac; the semisynthetic or modified natural
resins, which include the metallic resins and
ester gum; and the synthetic resins such as phe-
nolics, modified phenolic, maleic, terpene, and
the coal tar and petroleum-based resins.
The drying oils are liquid substances that oxi-
dize and polymerize in the atmosphere to form
hard, dry, resinous solids. They help give the
final varnish product its flexibility, adhesion,
and durability. The most extensively used are
the naturally occurring fatty oils such as linseed,
tung or China wood, safflower, soybean, and cotton-
seed. Also used in fair quantities are fish oil
and tall oil, which is a blend of oil and resin acid
recovered from the black liquor in the manufac-
ture of pulp by the sulfate process.
Solvents are used in varnish formulation to re-
duce the viscosity of the material so that it can
be applied as a thin film. The solvent evaporates
upon application arid is not a permanent part of
the finish. Solvents vary widely both in drying
time and in ability to dissolve various resins. In
general, aliphatic hydrocarbons such as kerosine
and mineral spirits are classed as low-power
solvents -while the aromatics such as toluene and
xylene are high-power solvents. The types and
quantities of solvent to be used for a specific
formulation can be determined only after con-
sideration of the type of resin used, the percent
solids and viscosity of the finished product, and
the characteristics required.
Driers are added to catalyze the oxidation and
promote polymerization of the film after applica-
tion. They may be added to the other ingredients
during cooking or, more commonly, to the finished
product. The driers used are soaps of heavy met-
als such as lead, cobalt, or manganese. In order
of importance are naphthenates, tallates, octoates,
linoleates, and resinates of these heavy metals.
Manufacturing Processes Involving Heat
In the manufacture of varnish products, the ap-
plication of heat to a single ingredient or to a
mixture of ingredients is the most important
single operation. Heating performs many func-
tions, depending upon the raw materials used
and the point in the formulation cycle. Several
of the most important of these functions follow:
1, Polymerization. Probably the most impor-
tant purpose in heating the ingredients in a
varnish is to polymerize the oil. The pres-
ence of the resins has essentially no effect
except to slow the polymerization reaction.
For a fast-polymerizing oil such as China
wood, the resin may be added early in order
to make the reaction more controllable.
When a slower oil such as linseed is used
the resin may be added after the polymeriza-
tion (or bodying, as it is called) has nearly
progressed to the desired viscosity point.
This method is preferable for slow oils
since prolonged contact of oil and resins in
bodying tends to result in a darker product.
2. Depolymerization. Some natural resins are
so high in molecular weight that they are in-
soluble in drying oil. By heating these resins
to a relatively high temperature, 600° to
650°F, the resin structure breaks down with
a resultant loss of 10 to 30 percent of the
original resin. The remaining resin is then
readily soluble in the drying oils and can be
processed to a finished varnish product.
This heating procedure is commonly known
as gum running.
3. Melting and accelerated solution. The tem-
peratures used in varnish cooking are high
enough to make the viscous and solid ingredi-
ents fluid and easily blendable. Moreover,
the solvent is easier to incorporate into the
cooked products at an elevated temperature.
4. Esterfication. In varnish making, rosin or
tall oil is treated "with a polyhydric alcohol
to form an ester. Glycerol and pentaerythritol
are typical alcohols used. Reaction temper-
atures during this operation are 450° to
525°F.
5. Gas checking or isomerization. Some of the
more active oils, such as China wood oil,
dry to partially crystalline films. This fault
is eliminated by heating the oil to 450° to
580°F. Apparently, this heating changes
the location of the double bond in the mole-
cule to a less reactive position, which there-
by eliminates extreme reactivity during oxi-
dation but does not drastically affect the poly-
merization rate.
-------
690
CHEMICAL PROCESSING EQUIPMENT
6. Distillation and evaporation. Distillation and
evaporation are used to remove from formula-
tion some of the undesirable constituents such
as volatile constituents of resins that have
been subjected to gum running, excess gly-
cerol, and "water. These materials may be
condensed and recovered or removed with
the remainder of the varnish fumes.
Major Types of Manufacturing Equipment
Varnish cooking is accomplished in two types of
vessels--the old open kettle and the newer, total-
ly enclosed, stationary kettle.
The portable open kettle is cylindrical, has a
flat bottom, rests upon a three- or four-wheel
truck, and is heated over an open flame. This
type kettle usually varies in capacity from 150
to 370 gallons and is made of steel, copper,
monel, aluminum, or stainless steel. Under
normal operating conditions, the kettle is
charged in the loading room, moved to the fire
pit, heated, then transferred to another location
for cooling, and finally to still another location
for addition of thinner and' drier. In some oper-
ations involving open-kettle cooking, a portable
electric mixer is used or, even more crudely,
the mixing is done manually. Figure 512 shows a
kettle of this type.
The more modern equipment consists of a totally
enclosed, autoclave-type kettle set over or with-
in a totally enclosed source of heat. The kettle
is usually in the shape of a cylinder with dome-
shaped top and bottom, is normally constructed
of stainless steel, and has completely automatic
controls. Heat is supplied by natural gas, oil,
electric coils, circulating Dowtherm, or hot
oil. Liquid raw material:; are pumped directly
and solids are added through a manhole at the
top, which can be sealed. During the operation,
the unit is completely enclosed and usually sup-
plied with an inert atmosphere such as nitrogen
or carbon dioxide. The kettle is also usually
equipped with cooling coils to cool the end prod-
*>•
j^;a:<
Figure 512. Uncontrolled open Kettle for varnish cooking.
-------
Varnish Cookers
691
uct quickly or prevent the exothermic poly-
merization reaction from ruining a batch. Fin-
ished material is pumped to other tanks equipped
with agitators and integral condensers for thin-
ning.
Variations in Varnish Formulation
There appear to be as many different varnish
formulations as there are enterprises making
varnish. Minor deviations in formulation ex-
ist among varnishes to be used for the same
purpose, in addition to the major deviations in
formulation that exist among varnishes accord-
ing to their ultimate uses.
A time-honored varnish formula is described
as follows (Von Fisher, 1948): 40 gallons of
China wood oil are heated rapidly to about
575 °F to gasproof the oil and body it slightly.
One-hundred pounds of ester gum are then add-
ed, "which cools the oil and slows and controls
its tendency to gel rapidly. The varnish base
is then cooled to about 400°F, at "which tempera-
ture mineral spirits can be added to bring the
finished varnish to a viscosity of about 1 to 2
poises and a nonvolatile content of 50 to 60 per-
cent. To this finished varnish, some antiskin-
ning agents and driers, usually lead and cobalt,
are added.
Various modifications of this basic recipe have
been used since the advent of synthetic resins.
Replacement of the ester gum with a pure phenol-
formaldehyde resin adds substantially to the
durability of the finished varnish. Other modifi-
cations include replacement of the glycerin ester
of rosin with pentaerythritol, which not only aids
drying but also improves the durability of the
varnish; and addition of maleic anhydride and
fumaric acid, "which improve durability even
more. A typical varnish formulation, described
as a general-purpose utility varnish, is shown
in Table 193.
Table 193. TYPICAL MODERN
VARNISH FORMULATION
Constituent
Tall oil
Dehydrated castor oil
Pentaerythritol
Maleic anhydride
Mineral spirits
Lead naphthenate
Cobalt naphthenate
Manganese naphthenate
Total
Quantity, Ib
1, 000
405
136
40
1, 530
30
10
5
3, 156
THE AIR POLLUTION PROBLEM
Varnishes are cooked and oils bodied at temper-
atures of from 200° to 600°F. At about 350°F
the products begin to decompose, resulting in the
emission of decomposition products from the cook-
ing vessel. As long as the cooking is continued
above this temperature, the emissions continue,
the maximum rate occurring shortly after the max-
imum temperature has been reached. These cooks
average 8 to 12 hours. The quantity, composition,
and rate of emissions depend upon the ingredients
in the cook, the maximum temperature, method of
introducing additives, degree of stirring, cooking
time, and extent of air or inert gas blowing (Sten-
burg, 1958). Total emissions for oleoresinous
varnishes average from 3 to 6 percent, and those
from oil cooking and blowing, 1 to 3 percent.
Cooker emissions vary in composition, depending
upon the ingredients in the cook. Mattiello (1943)
states that compounds emitted from cooking of
oleoresinous varnish include water vapor, fatty
acids, glycerine, acrolein, phenols, aldehydes,
ketones, terpene oils, terpenes, and carbon di-
oxide. Bodying oils emit these same compounds
less the phenols, terpene oils, and terpenes.
Gum running yields water vapor, fatty acids, ter-
penes, terpene oils, and tar. Besides the air con-
taminants listed by Mattiello, some highly offen-
sive sulfur compounds such as hydrogen sulfide,
allyl sulfide, butyl mercaptan, and thiophene are
formed when tall oil is esterfied with glycerine
and pentaerythritol, and these compounds are
emitted as a result of small amounts of sulfur in
the tall oil. Attempts to alleviate this problem in-
volve further refining tall oil to remove as much
sulfur as possible. At times, rapid cooling of
the cook is necessary because of lack of tempera-
ture control or to the nature of the particular
product. This rapid cooling is sometimes done
by injecting water directly into the mixture. In
this case, unless extreme care is taken, the
rapid emission of water vapor can entrain siz-
able'amounts of liquid droplets.
Of all these compounds emitted, acrolein is the
one most generally associated with oil cooking
because of its pungent, disagreeable odor and
irritating characteristics. Some of the more
odorous compounds have very low odor thres-
holds; acrolein, for example, has a threshold
at 1. 8 ppm and some of the sulfur compounds
have a threshold at about 0.001 ppm.
Although they are not from the varnish cook,
emissions of thinning solvent are important air
contaminants to be considered from varnish
manufacturing. In most of the newer installa-
tions, the cooked varnish is pumped to a thin-
ning tank that is equipped with integral conden-
-------
692
CHEMICAL PROCESSING EQUIPMENT
sers. In this equipment the emissions of sol-
vent are kept to a minimum. In the older open-
kettle operations, however, the thinning opera-
tion is carried out near the boiling point of the
solvent, and emissions of vapor can be consid-
erable. In this case, the thinning tank can be
hooded and the vapors ducted to the same ex-
haust system that removes the fumes from the
cooking kettles. In the design of air pollution con-
trol equipment for these operations, the emis-
sions of solvent as well as the emissions from
the cooker should be considered. From the
viewpoint of amount of air pollution, the emis-
sions of solvent are more important. Emis-
sions from the cooker constitute essentially a
local nuisance problem because of odors while
the emissions of solvent contribute to the over-
all hydrocarbon concentration in the atmosphere.
In general, the emissions of solvent in varnish-
cooking operations amount to 1 to Z percent by
weight of the solvent used.
In addition to the emissions of vapor and fume,
the addition of solid material to a varnish cook
may cause short bursts of dust emissions. For
example, when phthalic anhydride is added to a
cook that is above the compound's sublimation
temperature, considerable dust is created.
HOODING AND VENTILATION REQUIREMENTS
For the control methods to succeed, the fumes
must be captured and conveyed to the control
device. The exhaust system should be designed
to remove the fumes from the kettles under all
operating conditions without hindering kettle
operation. For open kettles, the hood should
fit closely enough to prevent excess entraining
air. Bidlack and Fasig (1951) state that a 1, 000-
gallon closed kettle should be provided with an
exhaust capacity of 300 to 400 cfm, and a stan-
dard open kettle, with 800 to 1, 000 cfm. Air
volumes for open kettles have been shown to be
reduced to approximately 200 cfm if the hooding
is properly designed. Figure 513 shows re-
tractable hoods in place over open, portable
kettles. These hoods have openings for addi-
tion of material and for thermometers and agita-
tors. The large opening for adding material
should have a hinged cover that can be kept closed
except when additions are actually being made.
An indraft velocity of about 150 fpm must be
maintained through the hood openings to prevent
leakage of contaminants. Condensate that col-
lects on the inner surface of the hood should
be collected in an outer trough and drained to
a container.
The main ductwork problems are corrosion and
fouling. The corrosion can be overcome by in-
stalling stainless steel ductwork. In installa-
Figure 513. Open varnish-cooking kettles with
exhaust hoods (Standard Brand Paints Company,
Torrance, Calif.).
tions processing alkyd resins, especially where
solid phthalic or maleic anhydride or penta-
erythritol is added, deposition of material in
the ductwork is heavy and results in plugging
in a relatively short time. For these systems,
rectangular ducts are preferable, with one side
hinged or removable to facilitate cleaning.
Conveying velocities of 1,500 to 2,000 fpm in the
ductwork are usually satisfactory.
AIR POLLUTION CONTROL EQUIPMENT
All operations in which varnish is cooked or dry-
ing oils are bodied by application of heat, with
or without blowing with air or inert gases, should
have air pollution control devices. Of the total
material charged to an uncontrolled kettle, 1 to
5 percent is discharged to the atmosphere dur-
ing the cook. This material includes the odorous,
irritating compounds previously mentioned. The
control devices applicable to varnish cooking are
the same as those used for controlling other
odor sources, with some modifications to meet
situations unique to this operation. In addition
to odor sources, visible emissions must also
be eliminated.
Scrubbers
Scrubbing and condensing equipment is not capa-
ble of controlling varnish odors adequately because
most of the objectionable material is in the form of
noncondensable or insoluble gas or vapor, or is
particulate matter of very small size. Scrubbers
are, however, valuable adjuncts when used as
precleaners.
Microscopic examination of particles deposited
on glass slides held in the path of emissions
-------
Varnish Cookers
693
from a varnish-cooking operation showed a par-
ticle size range from 2 to ZO microns' diameter.
The most frequent sizes were 8 to 10 microns.
Scrubbers have little effect on most of these
small particles. Scrubbers do, however, con-
trol emissions of sublimed solids such as phtha-
lic anhydride. This material solidifies into
relatively large flakes that are fairly easy to
control.
The scrubbing equipment used upstream from
the final collection device is generally a spray
tower, a plate tower, a chamber or tower with
a series of baffles and water curtains, an agi-
tated tank, or a water jet scrubber. The spray
tower is probably the most efficient because the
high degree of atomization that can be obtained
in the scrubbing liquid by the sprays allows for
maximum contact between the scrubbing water
and particulate matter. A major disadvantage
of the spray tower, however, is the excessive
maintenance required to keep the spray nozzle
free from clogging and in proper operation if
the scrubbing media are recirculated. From an
economic point of view, the baffled water cur-
tain scrubber is better but is less efficient.
Packed towers are not usually used in varnish
operations, because of the gummy condensed
fumes that rapidly plug the tower. In practical-
ly all the scrubbing devices, the flow of vapor
is counter cur rent to that of the scrubbing liq-
uid. Although various liquids have been used
as scrubbing media, for example, acids, bases,
various oils, and solvents, all are too expen-
sive to be used in large-scale operations. Water
is generally used and is usually not recirculated.
Wetting agents are at times added to the water
to increase its efficiency as an absorbent.
Adsorbers
Adsorbing equipment, especially activated-car-
bon filters, are very efficient in removing sol-
vents and odors from gas streams. To main-
tain this efficiency, however, the gas stream
entering the carbon filters must be almost com-
pletely free of solids and entrained oil droplets.
Unfortunately, varnish cooker effluent is not
free of these materials. Without some highly
efficient precleaning device, an activated-car-
bon filter serving a varnish cooker would rapid-
ly become clogged and inoperative. If used
downstream, from an efficient filter or precip-
itator, an activated-carbon unit could control
solvents and odors effectively. The economy
of this combined system is, however, question-
able compared to that of combustion in terms
of both the original cost and the operating cost.
Activated carbon has recently received a fair
amount of attention in operations involving the
control of emissions of solvents. Here, the
economics of the operation are greatly enhanced
by the possibility of solvent recovery. Thus, a
relatively high initial installation cost could be
recovered in a short period. This economic lever
does not exist for cooking operations, however,
and as far as can be determined, little, if any,
emissions from varnish cooking have been con-
trolled by adsorption.
Afterburners
At present, combustion is the only control method
that has proved effective. The other methods
listed individually remove varying amounts of the
contaminants from varnish cooking, but a properly
designed afterburner can do the job alone. In
some instances, using a scrubber as a precleaner
may be desirable from an operational point of view.
Afterburners that have been used for controlling
emissions from varnish cookers have been pre-
dominantly of the direct-flame type, effluent
gases and flame entering tangentially, as shown
in Figure 514, or the axially fired type shown
^sspp^w^^p^
Figure 514. Tangentially fired vertical after-
burner (Major Paint Co., Division of Standard
Brands Paint Co., Torrance, Calif.).
-------
694
CHEMICAL PROCESSING EQUIPMENT
in Figure 515. The burners are normally de-
signed to be capable of reaching a temperature
of 1,400°F under maximum load conditions. For
most operations, however, 1,200°F completely
controls all visible emissions and practically all
odors. At temperatures appreciably below 1,200°F,
incomplete combustion results in intermediate
products of combustion, and highly odorous mate-
rials are emitted. Table 194 summarizes re-
sults of stack tests on two types of direct-flame
afterburners.
The afterburner should be designed to have the
maximum possible flame contact with the gases
to be controlled and it should be of sufficient size
to have a gas retention time of at least 0. 3 second.
Most authorities agree that the length-to-diameter
ratio should be about 2. 5 to 4:1. In order to pre-
vent flashback and serious fire hazard, the fumes
must enter the afterburner at a velocity faster
than the flame propagation rate in the reverse
direction. An even more positive fire control
is a flame trap or barrier bet-ween the afterburn-
er and the kettle. This could be a simple scrub-
ber, as shown in the schematic plan of a control
system in Figure 516. Mills et al. (1960) de-
scribe the optimum design features of direct-
flame afterburners used specifically for varnish
cookers.
Table 194. SUMMARY OF RESULTS OF
STACK DISCHARGE TESTS OF
TWO TYPES OF AFTERBURNERS
SERVING VARNISH-COOKING KETTLES
Firing method
Air entry
Mixing method
Number of kettles exhausted
during test
Material processed
Exhaust system volume, scfm
Gas fuel use, cfm fuel per
cfm exhaust air
Combustion chamber
Average velocity, fps
Length-to-diameter ratio
Mixing velocity, fps
Average temperature, °F
Residence time, sec
Efficiency of removal, %
Particulate matter
Organic acids
Hydrocarbons
Unit A
Tangential
Tangential
Orifice
2
Alkyd
Spar varnish
950
0. 32
15
4
29
1, 220
0. 7
94
50
99+
Unit B
Axial
Tangential
Baffle
2
Linseed oil
320
0.027
12
2
17. 5
1, 100
0. 3
88
No data
96
Catalytic afterburners have been used to control
emissions from varnish cookers in some sec-
iilllllllllllli
Figure 515. Axially gas-fired afterburner with induced-draft fan (McCloskey Varnish Co. of the West,
Los Angeles, Cali f.).
-------
Sulfuric Acid Manufacturing
695
TANGENTIALLY FIRED AFTERBURNER
PORTABLE COOKERS
WITH HOODS
RECIRCULATING WATER BASIN
WATER-RECIRCULATING PUMP
Figure 516. Schematic plan for varnish-cooking control system.
ions of the United States. Some references to
tiese installations have indicated satisfactory
erformance. They have not been used in the
outhern California area. Catalytic units have,
owever, been used considerably for controlling
•missions of solvent from sources such as tin
>late-coating operations. Temperatures near
)50°F at the catalyst outlet have been adequate
,o control visible emissions and some odors
rom these operations. The effluent from var-
lish cookers could probably be incinerated to
about the same degree at the same temperature
range as that of the solvent from metal coating.
Although few analytical results have been lo-
cated, Selheimer and Lance (1954) report that
laboratory-scale, catalytic-combustion tests
on varnish-cooking effluent was considered
satisfactory at about 700 °F. They report
800°F as necessary to burn phthalic anhydride
fumes, and their tests indicate an inability to
incinerate pentaerythritol properly. This lat-
ter result has been refuted by manufacturers
of catalytic units, but no specific data have been
provided.
SULFURIC ACID MANUFACTURING
Sulfuric acid is used as a basic raw material
in an extremely wide range of industrial pro-
cesses and manufacturing operations. Because
of its widespread usage, sulfuric acid plants are
scattered throughout the nation near every indus-
trial complex.
Basically, the production of sulfuric acid involves
the generation of sulfur dioxide (802), its oxida-
tion to sulfur trioxide (SO,), and the hydration of
the SOg to form sulfuric acid. The two main pro-
cesses are the chamber process and the contact
process. The chamber process uses the reduc-
tion of nitrogen dioxide to nitric oxide as the oxi-
dizing mechanism to convert the SC>2 to SO^. The
contact process, using a catalyst to oxidize the
SO to SO,, is the more modern and the more
commonly encountered. For these reasons fur-
ther discussion will be restricted to the contact
process of sulfuric acid manufacture.
CONTACT PROCESS
A flow diagram of a "Type S" (sulfur-burning,
hot-gas purification type) contact sulfuric acid
plant is shown in Figure 517. Combustion air is
drawn through a silencer, or a filter when the air
is dust laden, by either a single-stage, centrif-
ugal blower or a positive-pressure-type blower.
Since the blower is located on the upstream side,
the entire plant is under a slight pressure, vary-
ing from 1. 5 to 3. 0 psig. The combustion air
is passed through a drying tower before it enters
-------
696
CHEMICAL, PROCESSING EQUIPMENT
FILTERED AIR
STACK
»EAK ACID ^
FILTERED
AIR
DRYING
TOKER
U1LTEN
SULFUR
SULFUR
BURNER
BOILER
WEAK IMPURE ACID
Figure 517. Flow diagram of a typical " Type S"
sulfur-burning contact sulfuric acid plant.
the sulfur burner. In the drying tower, moisture
is removed from the air by counter cur rent scrub-
bing with 98 to 99 percent sulfuric acid at tem-
peratures from 90° to 120°F. The drying tower
has a topside internal-spray eliminator located
just below the air outlet to minimize acid mist
carryover to the sulfur burner.
Molten sulfur is pumped to the burner where it
is burned with the dried combustion air to form
SO-,. Normally a gas containing approximately
9 percent SO;? is produced in a Type S plant. The
combustion gases together with excess air leave
the burner at about 1, 600 °F and are cooled to
approximately SOOT in a water tube-type waste -
heat boiler. The combustion gases then pass
through a hot-gas filter into the first stage or
"pass" of the catalytic converter at between
750° and 800 °F to begin the oxidation of the SO2
to SO3. If the molten sulfur feed has been fil-
tered at the start of the process, the hot-gas
filter may be omitted. Because the conversion
reaction is exothermic, the gas mixture from
the first stage of the converter is cooled in a
smaller waste-heat boiler. Gas cooling after
the second and third converter stages is achieved
by steam superheaters. Gas leaving the fourth
stage of the converter is partially cooled to ap-
proximately 450°F in an economizer. Further
cooling takes place in the gas duct before the
gas enters the absorber. The extent of cooling
required depends largely upon whether or not
oleum is to be produced. The total equivalent
conversion from SO2 to SO3 in the four con-
version stages is about 98 percent. Table 195
shows typical temperatures and conversions
at each stage of the four-stage converter. These
figures vary somewhat with variations in gas
composition, operating rate, and catalyst con-
dition.
The cooled 803 combustion gas mixture enters
the lower section of the absorbing tower, which
is similar to the drying tower. The SO^ is ab-
Table 195. TEMPERATURES AND CONVERSIONS
IN EACH STAGE OF A FOUR-STAGE CONVERTER
FOR A "TYPE S" SULFUR-BURNING CONTACT
SULFURIC ACID PLANT
Location of gas
Entering 1st pass
Leaving 1st pass
Entering 2d pass
Leaving 2d pass
Entering 3d pass
Leaving 3d pass
Entering 4th pass
Leaving 4th pass
Total rise
Tempe
°C
410
601.8
191.8
438
485. 3
47. 3
432
443
11
427
430.3
3. 3
253. 4
ratures,
°F
770
1, 115
345
820
906
86
810
830
20
800
806
6
457
Equivalent
conversion, %
74. 0
18.4
4. 3
1.3
98. 0
sorbed in a circulating stream, of 99 percent
sulfuric acid. The nonabsorbed tail gases pass
overhead through mist removal equipment to
the exit gas stack (Duecker and West, 1959).
A contact process plant intended mainly for use
with various concentrations of hydrogen sulfide
(H£S) as a feed material is known as a wet-gas
plant, as shown in Figure 518. The wet-gas
plant's combustion furnace is also used for burn-
ing sulfur or dissociating spent sulfuric acid. A
common procedure for wet-gas plants located near
petroleum refineries is to burn simultaneously
H2S, molten sulfur, and spent sulfuric acid from
the alkylation processes at the refineries. In
some instances a plant of this type may produce
sulfuric acid by using only H-,3 or spent acid.
-------
Sulfuric Acid Manufacturing
697
ABSORBER
STACK
MIST
PRECIPITATOR
DRYING
TOHER
ACID TO STORAGE
Figure 518. Flow diagram of a contact-type wet-gas sulfur plant.
In a wet-gas plant, the EUS gas, saturated with
water vapor, is charged to the combustion fur-
nace along with atmospheric air. The SO^
formed, together with the other combustion
products, is then cooled and treated for mist
removal. Gas may be cooled by a waste-heat
boiler or by a quench tower folio-wed by Karbate
and updraft coolers. Mist formed is removed
by an electrical precipitator. Moisture is re-
moved from the SC>2 and airstream with con-
centrated sulfuric acid in a drying tower. A
centrifugal blower takes suction on the drying
tower and discharges the dried SC>2 and air to
the converters. The balance of the wet-gas pro-
cess is essentially the same as that of the pre-
viously discussed sulfur-burning process.
Table 196. SULFUR TRIOXIDE AND SULFUR
DIOXIDE EMISSIONS FROM TWO ABSORBERS IN
CONTACT SULFURIC ACID PLANTS
Gas flow rate,
scfm
Sulfur trioxide,
gr/scf
% by vol as SO2
Ib/hr
Sulfur dioxide,
gr/scf
% by vol
Ib/hr
Outlet of
absorber No. 1
9, 600
0. 033
0. 002
2.73
2.63
0.22
216
Outlet of
absorber No. 2
7, 200
0. 39
2. 4
2. 45
151.2
THE AIR POLLUTION PROBLEM
The only significant source of air contaminant
discharge from a contact sulfuric acid plant
is the tail gas discharge from the SOj absorber.
While these tail gases consist primarily of in-
nocuous nitrogen, oxygen, and some carbon di-
oxide, they also contain small concentrations
of SO2 and smaller amounts of 803 and sulfuric
acid mist. Table 196 shows the SO2 and SO3
discharged from two wet-gas sulfuric plant ab-
sorbers.
A well-designed contact process sulfuric acid
plant operates at 90 to 95 percent conversion of
the sulfur feed into product sulfuric acid. Thus
a 250-ton-per-day plant can discharge 1. 25 to 2.5
tons of SO2 and 803 per day. When present in
sufficient concentration, SO£ is irritating to
throat and nasal passages and injurious to vege-
tation. SO2 concentrations greater than 0.25 ppm
cause injury to plants on long exposure. The
permissible limit for humans for prolonged ex-
posure is 10 ppm.
-------
698
CHEMICAL PROCESSING EQUIPMENT
Tail gases that contain SOj, owing to incomplete
absorption in the absorber stack, hydrate and
form a finely divided mist upon contact with at-
mospheric moisture. According to Fairlie
(1936) the process temperature of gas going to
the absorber should be on the lower side of a
temperature range between 150° and 230°C.
The optimum acid concentration in the absorb-
ing tower is 98. 5 percent. This concentration
has the lowest SOj vapor pressure. The partial
pressure of 803 increases if the absorbing acid
is too strong, and SOj passes out with the tail
gases. If a concentration of absorbing acid
less than 98. 5 percent is used, the beta phase
of SC>3, which is less easily absorbed, is pro-
duced. A mist may also form when the pro-
cess gases are cooled before final absorption,
as in the manufacture of oleum.
Water-based mists can form as a result of the
presence of water vapor in the process gases fed
to the converter. This condition is often caused
by poor performance of the drying tower. Effi-
cient performance should result in a moisture
loading of 5 milligrams or less per cubic foot.
In sulfur-burning plants, mists may be formed
from water resulting from the combustion of
hydrocarbon impurities in the sulfur. Mists
formed in the wet-purification systems of an
acid sludge regeneration plant are not complete-
ly removed by electrostatic precipitation. The
mists pass through the drying tower and are
volatilized in the converter. The mist reforms,
however, when the gases are cooled in the ab-
sorption tower. Water-based mists can also form
from any steam or water leaks into the system.
AIR POLLUTION CONTROL EQUIPMENT
Sulfur Dioxide Removal
Water scrubbing of the SO^ absorber tail gases
can remove 50 to 75 percent of the SO2 content.
Scrubbing towers using 3-inch or larger stacked
rings or redwood slats are often employed. On
startups, when SO2 concentrations are large,
soda ash solution is usually used in place of
straight water. Water scrubbing is feasible where
disposal of the acidic waste water does not present
a problem.
Tail gases may be scrubbed with soda ash solution
to produce marketable sodium bisulfite. A cyclic
process using sodium sulfite-bisulfite has also
been reported. Steam regeneration costs in the
cyclic process are, however, relatively high, and
the capacity of the scrubbing solution is limited
by the low solubility of sodium bisulfite. The
dilute scrubber solution has, moreover, little
economic value.
The most widely known process for removal of
SO, from a gas stream is scrubbing with am-
monia solution. It was developed by Consolidated
Mining and Smelting Company and installed at its
Trail, British Columbia, plant (Duecker and West,
1959). Single- and two-stage absorber systems
reportedly reduce SO2 concentrations in tail gases
to 0. 08 and 0. 03 percent respectively. Two-
stage systems are designed to handle SO2 gas
concentrations as great as 0. 9 percent. Large
SO2 concentrations resulting from acid plant
startups and upsets could be handled adequate-
ly by a system such as this.
The 803 mist presents the most difficult prob-
lem of air pollution control since it is generally
of the smallest particle size. The particle size
of these acid mists ranges from submicron to
10 microns and larger. Acid mist composed of
particles of less than 10 microns in size is visible
in the absorber tail gases if present in amounts
greater than 1 milligram of sulfuric acid per
cubic foot of gas. As the particle size decreases,
the plume becomes more dense because of the
greater light-scattering effect of the smaller
particles. Maximum light scattering occurs
•when the particle size approximates the wave
length of light. Thus, the predominant factor
in the visibility of an acid plant's plume is
particle size of the acid mists rather than the
weight of mist discharged. Acid particles larg-
er than 10 microns are probably present as a
result of mechanical entrainment. These larg-
er particles deposit readily on duct and stack
walls and contribute little to the opacity of the
plume.
Acid Mist Removal
Electrical precipitators
Electrical precipitators are widely used for re-
moval of sulfuric acid mist from the cold SO,
gas stream of •wet-purification systems. The
wet-lead-tube type is used extensively in this
servic e.
Tube-type precipitators have also been used for
treating tail gases from SOj absorber towers.
More recently, however, two-stage, plate-type
precipitators have been used successfully. One
such unit, lead lined throughout to prevent corro-
sion, is designed to handle approximately 20, 000
cfm tail gas from a 300-ton-per-day contact
sulfuric acid plant. This wet-gas plant process-
es H^S, sulfur, and spent alkylation acid. Dry
gas containing SO2, carbon dioxide, oxygen,
nitrogen, and 5 to 10 milligrams of acid mist
per cubic foot enters two inlet ducts to the pre-
cipitator. The gas flows upward through dis -
-------
Sulfuric Acid Manufacturing
699
tribution tiles to the humidifying section. This
section contains 5 feet of 3-inch single-spiral
tile irrigated by 800 gpm weak sulfuric acid.
The conditioned gas then flows to the ionizing
section, which consists of about 75 grounded
curtain electrodes and 100 electrode wire ex-
tensions .
Ionized gas then flows to the precipitator section
where charged acid particles migrate to the col-
lector plate electrodes. There are twelve 14-
by 14-foot lead plates and 375 electrode wires.
The negative wire voltage is 75, 000. Acid mi-
grating to the plates flows down through the pre-
cipitator and is collected in the humidifying sec-
tion. The gas from the precipitator section flows
to a 5-foot-diameter, lead-lined stack that dis-
charges to the atmosphere 150 feet above grade.
The high-voltage electrode wires are suspended
vertically by three sets of insulators. Horizontal
motion is eliminated by four diagonally placed in-
sulators, which are isolated from the gas stream
by oil seals. All structural material in contact
with the acid mist is lead clad. Electrical wires
are stainless steel cores with lead cladding. Volt-
age is supplied from a generator with a maximum
capacity of 30 kilovolt-amperes. A battery of
silicon rectifiers supplies 75, 000 volts of direct
current to the electrode wires.
Table 197 shows the sulfur trioxide and sulfur
dioxide emissions from the previously described
two-stage electrical precipitator. The acid mist
collection efficiency was only 93 percent. A
mechanical rectifier was, however, supplying
only 36, 000 volts to the precipitator during this
test. During normal operation, silicon rectifiers
supply 75, 000 volts to the electrode wires.
Packed-bed separators
Packed-bed separators employ sand, coke, or
glass or metal fibers to intercept acid mist par-
ticles. The packing also causes the particles to
coalesce by reason of high turbulence in the small
spaces between packing. Moderate-sized particles
of mist have been effectively removed in a 12-inch-
deep bed of 1-inch Berl saddles with gas veloc-
ities of approximately 10 fps.
Glass fiber filters have not been very effective
in mist removal because of a tendency on the
part of the fiber to sag and mat. Nevertheless,
experimental reports by Fairs (1958) on acid
mist removal by silicone-treated glass wool are
encouraging. A special fine-glass wool with a
fiber diameter between 5 and 30 microns was
used. The coarser fibers allowed adequate pene-
tration of the bed by the mist particles to ensure
a reasonable long life and provided sufficient
support for the finer fibers in their trapping of
the small acid mist particles.
The glass "wool was treated by compressing it
into a filter 2 inches thick to a density of 10
pounds per cubic foot. It was then placed in
a sheet metal container and heated at 500°C
for 1 hour. By this treatment, the stresses
in the compressed fibers were relieved, and
the fiber mass could be removed from the mold
without losing shape or compression. The
fibers were then treated with a solution of meth-
yl chlorosilane.
Table 197. SULFUR TRIOXIDE AND SULFUR
DIOXIDE EMISSIONS FROM A TWO-STAGE
ELECTRICAL PRECIPITATOR SERVING
A CONTACT SULFURIC ACID PLANT
Gas flow rate, scfm
Gas temperature, °F
Average gas velocity, fps
Collection efficiency, a %
Moisture in gas, %
CO;?, % (stack conditions)
O2 , %(stack conditions)
CO, % (stack conditions)
N£, % (stack conditions)
Sulfur trioxide,
gr/scf
Ib/hr
% by volume
Sulfur dioxide,
gr/scf
Ib/hr
% by volume
Inlet of
precipitator
13,400
160
36.5
0. 8
5. 9
9.6
0
83. 4
0. 062
7. 1
0. 0042
4. 1
470
0. 345
Outlet of
precipitator
13,100
80
20.6
93
4. 1
6
8. 4
0
81.2
0. 0048
0. 54
0. 00032
4. 1
460
0. 345b
aA mechanical rectifier was supplying only 36, 000
volts to the precipitator. During normal operation,
silicon rectifiers supply 75, 000 volts to the electrode
wires. This should increase the acid mist collection
efficiency appreciably.
DRule 53. 1 for "scavenger plants" is applicable to
this plant rather than Rule 53a, which limits emis-
sions' of SO2 to 0. 2 percent by volume. This plant
recovers SO2 that would otherwise be emitted to the
atmosphere.
The threshold concentration for mist visibility
after scrubbing has been found experimentally
by Fairs (1958) to be about 3. 6 x lO"4 gram
SOj per cubic foot. The discharge gases from
the silicone-treated filter had an SO, concen-
tration of 1. 8 to 2. 5 x 10~4 gram per cubic
foot and no appreciable acid mist plume. A
faint plume became perceptible at approximate-
ly 'weekly intervals but was eliminated by flush-
ing the filter bed "with water. The average tail
gas-filtering rate for the treated filter was
15. 6 cfm per square foot of filtering area for
-------
700
CHEMICAL PROCESSING EQUIPMENT
a pressure drop of 9-1/2 to 10 inches water
column. According to Fairs, the effective
life of the silicone fiber should be at least
5, 000 hours. Garnetted terylene was also
used but was not as efficient as silicone-treated
glass wool. It should, however, prove ade-
quate for less stringent duties. Its life should
be long since it does not require silicone pre-
treatment. The use of untreated glass wool
fiber proved unsatisfactory in reducing the
opacity of the acid mist plume.
Table 198 shows the SC>2 and acid mist emis-
sions from the outlet of a typical silicone-
treated, glass fiber mist eliminator. This
control unit processes absorber discharge gas
from a contact sulfuric acid plant. The acid
mist collection efficiency for the fiber glass
mist eliminator was 98. 9 percent. A success-
ful application of a mist eliminator using treat-
ed fiber (Figure 519) has been made by the
Monsanto Chemical Company (Brink, 1959). The
exact treatment given to the fiber is not available
since it is the property of the inventor, J. A.
Brink, Jr.
Table 198. EMISSIONS OF SULFUR DIOXIDE
AND ACID MIST FROM THE OUTLET OF
A SILICONE-TREATED, GLASS FIBER MIST
ELIMINATOR SERVING A CONTACT
SULFURIC ACID PLANT
RETAINER PLATE
rfil
Concentration, gr/scf
Concentration, ppm
Weight, Ib/hr
Collection efficiency, %
Gas flow rate, scfm
Avg gas velocity, fps
Gas temperature, °F
Mist
elimmat or
inlet
Acid
mist
0. 30
200
45
Mist eliminator outlet
Acid
mist
0. 035
25
0. 5
Sulfur
dioxide
1. 50
1, 300
160
98. 9
14, 000
19
160
Wire mesh mist eliminators
Wire mesh mist eliminators are usually con-
structed in two stages. The lower stage of
wire mesh may have a bulk density of about
14 pounds per cubic foot, while the upper stage
is less dense. The two stages are separated
by several feet in a vertical duct. The high-
density lower stage acts as a coalescer. The
re-entrained coalesced particles are removed
in the upper stage. Typical gas velocities for
these units range from 11 to 18 fps. The kinet-
ic energy of the mist particle is apparently too
low to promote coalescence at velocities less
than 11 fps, and re-entrainment becomes a
problem at velocities greater than 18 fps. The
L
SUPPORT
PLATE
V.
SUPPORT PLATE
WIRE MESH
FIBER PACKING
-BRINK
ELEMENT
SEAL
POT
Figure 519. Brink fiber mist eliminator (Brink,
1959).
tail gas pressure drop through a wire mesh
mist installation is approximately 3 inches
'water column.
Exit sulfuric acid mist loadings of less than 5
milligrams per cubic foot of gas are normally
obtained from wire mesh units serving plants
making 98 percent acid. No type of mechan-
ical coalescer, however, has satisfactorily
controlled acid mists from oleum-producing
plants. Corrosion possibilities from concen-
trated sulfuric acid must be considered in se-
lecting wire mesh material. The initial cost of
wire mesh equipment is modest. The value of
recovered sulfuric acid is usually sufficient to
pay the first investment in 1 or 2 years (Duecker
and West, 1959).
Ceramic filters
Porous ceramic filter tubes have proved success-
ful in removing acid mist. The filter tubes are
usually several feet in length and several inches
in diameter with a wall thickness of about 3/8 inch.
-------
Phosphoric Acid Manufacturing
701
The tubes are mounted in a horizontal tube sheet,
with the tops open and the bottoms closed. The
tail gases flow downward into the tubes and pass
out through porous walls. Appreciably more fil-
tering area is required for the ceramic filter
than for the wire mesh type. The porous ceramic
filter is composed of small particles of alumina
or similar refractory material fused with a binder.
The maintenance costs for ceramic tubes is con-
siderably higher than those for wire mesh filters
because of tube breakage. Initial installation
costs are also considerably higher than those for
wire mesh. A pressure drop of 8 to 10 inches
water column is required to effect mist removal
equivalent to that of a wire mesh filter. Thus,
operating costs would also be appreciable (Duecker
and West, 1959).
Sonic agglomeration
The principle of sonic agglomeration is also
used to remove acid particles from waste-gas
streams. Sound waves cause smaller particles
in an aerosol to vibrate and thereby coalesce
into larger particles. Conventional cyclone sep-
arators can then be used for removal of these
larger particles. One installation treating exit
stack gases from a contact acid plant has been
reported to remove 90 percent by "weight of acid
in the gas stream. The tail gases leaving the
sonic collector contained 2 to 3 milligrams of
100 percent sulfuric acid mist per cubic foot. A
nuisance factor must be taken into consideration,
however, since some of the sound frequencies
are in the audible range (Duecker and West, 1959).
Miscellaneous devices
Simple baffles and cyclone separators are not
effective in collecting particles smaller than
5 microns in size. A considerable amount of
the larger size acid mist particles may be re-
moved; however, the visibility of the stack
plume is not greatly affected, since the smallest
particle size contributes most to visibility. Vane-
type separators operate at relatively high gas
velocities and thus make better use of the parti-
cles' kinetic energy. They have been found to be
moderately effective for contact plants having
"wet-purification systems in reducing stack plume
opacities (Duecker and West, 1959).
PHOSPHORIC ACID MANUFACTURING
During the past 20 years, the use of phosphorus-
containing chemical fertilizers, phosphoric acid,
and phosphate salts and derivatives has increased
greatly. In addition to their very large use in
fertilizers, phosphorus derivatives are widely
used in food and medicine, and for treating water,
plasticizing in the plastic and lacquer industries,
flameproofing cloth and paper, refining petroleum,
rustproofing metal, and for a large number of
miscellaneous purposes. Most of the phosphate
salts are produced for detergents in "washing
compounds.
With the exception of the fertilizer products,
most phosphorus compounds are derived from
orthophosphoric acid, produced by the oxidation
of elemental phosphorus. At present, elemental
phosphorus is manufactured on a large enough
scale to be classed as a heavy chemical and is
shipped in tank cars from the point of initial
manufacture, where the raw materials are inex-
pensive, to distant plants for its conversion to
phosphoric acid, phosphates, and other compounds.
PHOSPHORIC ACID PROCESS
Generally, phosphoric acid is made by burning
phosphorus to form the pentoxide and reacting
the pentoxide with water to form the acid. Spe-
cifically, liquid phosphorus (melting point 112°F)
is pumped into a refractory-lined tower where it
is burned to form phosphoric oxide, P4O10, which
is equivalent algebraicly to two molecules of the
theoretical pentoxide, P2Oc, and is, therefore,
commonly termed phosphorus pentoxide:
SO,
P O
4 10
An excess of air is provided to ensure complete
oxidation so that no phosphorus trioxide (P2O^)
or yellow phosphorus is coproduced. The reac-
tion is exothermic, and considerable heat must
be removed to reduce corrosion. Generally,
water is sprayed into the hot gases to reduce
their temperature before they enter the hydrating
section.
Additional water is sprayed counter currently to
the gas stream, hydrating the phosphorus pent-
oxide to orthophosphoric acid and diluting the
acid to about 75 to 85 percent:
P O
410
6H2°
4H PO
3 4
The hot phosphoric acid discharges continuously
into a tank, from which it is periodically re-
moved for storage or purification. The tail gas
from the hydrator is discharged to a final col-
lector where most of the residual acid mist is
removed before the tail gas is vented to the air.
A general flow diagram for a phosphoric acid
plant is shown in Figure 520.
-------
702
CHEMICAL PROCESSING EQUIPMENT
The raw acid contains arsenic and other heavy
metals. These impurities are precipitated as
sulfides. A slight excess of hydrogen sulfide,
sodium hydrosulfide, or sodium sulfide is add-
ed and the treated acid is filtered. The excess
hydrogen sulfide is removed from the acid by
air blowing.
The entire process is very corrosive, and
special materials of construction are required.
Stainless steel, carbon, and graphite are com-
monly used for this severe service.
Special facilities are required for handling
elemental yellow phosphorus since it ignites
spontaneously on contact with air at atmospher-
ic temperatures and is highly toxic. Phosphorus
is always shipped and stored under water to pre-
vent combustion. The tank car of phosphorus is
heated by steam coils to melt the water-covered
phosphorus. Heated water at about 135 °F is
then pumped into the tank car and displaces the
phosphorus, which flows into a storage tank.
A similar system using hot displacement water
is frequently used to feed phosphorus to the
burning tower.
THE AIR POLLUTION PROBLEM
A number of air contaminants, such as phosphine,
phosphorus pentoxide, hydrogen sulfide, and phos-
phoric acid mist, may be released by the phosphor-
ic acid process.
Phosphine (PH^), a very toxic gas, may be formed
by the hydrolysis of metallic phosphides that exist
as impurities in the phosphorus. When the tank
car is opened, the phosphine usually ignites spon-
taneously but only momentarily.
Phosphorus pentoxide (P^O^Q), created when
phosphorus is burned •with excess air, forms
an extremely dense fume. Our military forces
take advantage of this property by using this com-
pound to form smoke screens. The fumes are
submicron in size and are 100 percent opaque.
Except for this military use, phosphorus pent-
oxide is never released to the atmosphere unless
phosphorus is accidentally spilled and exposed
to air. Since handling elemental phosphorus is
extremely hazardous, stringent safety precau-
tions are mandatory, and phosphorus spills are
very infrequent.
Hydrogen sulfide (^S) is released from the acid
during treatment with NaHS to precipitate sulfides
of antimony and arsenic and othrar heavy metals.
Removal of these heavy metals is necessary for
manufacture of food grade acid. ^S is highly
toxic and flammable. Health authorities recom-
mend a maximum allowable concentration of this
gas of 20 ppm for an 8-hour exposure. The odor
threshold is 0. 19 ppm (Gillespie and Johnstone,
1955). In practice, however, H^S is blown from
the treating tank and piped to the phosphorous-
burning tower where it is burned to SO2. Source
test information indicates that the concentration
of SO2 in the gaseous effluent from the acid tower
scrubberwillnot exceed 0. 03 volume percent. Evo-
lution of HzS is also minimized by restricting the
amount of NaHS in excess of that needed to precipitate
arsenic and antimony and other heavy metals.
The manufacture of phosphoric acid cannot be
accomplished in a practical way by burning
phosphorus and bubbling the resultant products
through either water or dilute phosphoric acid
(Slaik and Turk, 1953). When water vapor comes
into contact with a gas stream that contains a
volatile anhydride, such as phosphorus pentoxide,
an acid mist consisting of liquid particles of var-
ious sizes is formed almost instantly. An investi-
gation (Brink, 1959) indicates that the particle
size of the phosphoric acid aerosol is small,
about 2 microns or less, and that it has a median
diameter of 1. 6 microns, with a range of 0. 4 to
2. 6 microns.
The tail gas discharged from the phosphoric acid
plant is saturated with water vapor and produces
a 100 percent opaque plume. The concentration
of phosphoric acid in this plume may be kept
small with a well-designed plant. This loss
amounts to 0.2 percent or less of the phosphorus
charged to the combustion chamber as phosphorus
pentoxide.
HOODING AND VENTILATION REQUIREMENTS
All the reactions involved take place in closed
vessels. The phosphorus-burning chamber and
the hydrator vessel are kept under a slight neg-
ative pressure by the fan that handles the effluent
gases, as shown in Figure 520. This is neces-
sary to prevent loss of product as well as to pre-
vent air pollution.
MOLTEN PHOSPHORUS
PHOSPHORUS
BURNER
Figure 520. General flow diagram for phosphoric
acid production.
-------
Phosphoric Acid Manufacturing
703
The hydrogen sulfide generated during the acid
purification treatment must be captured and col-
lected, and sufficient ventilation must be pro-
vided to prevent an explosive concentration, for
hydrogen sulfide has a lower explosive limit of
4. 3 percent. The sulfiding agent must be care-
fully metered into the acid to prevent excessive-
ly rapid evolution of hydrogen sulfide.
AIR POLLUTION CONTROL EQUIPMENT
The hydrogen sulfide can be removed by chem-
ical absorption or by combustion. Weak solu-
tions of caustic soda or soda ash sprayed
countercurrently to the gas stream react with
the hydrogen sulfide and neutralize it:
Na2C°3
2NaOH + H S
NaHCO + NaHS
Na S + 2H O
L-. L*
The hydrogen sulfide may also be oxidized in
a suitable afterburner:
2H S +
30,
2H O
+ 2SCL
The phosphoric acid mist in the tail gas is
commonly removed by an electrical precip-
itator, a venturi scrubber, or a Brink fiber
mist eliminator (Brink, 1959). All are very
effective in this service.
Table 199. STACK ANALYSES OF EMISSIONS
FROM A PHOSPHORIC ACID PLANT
WITH A VENTURI SCRUBBER
Phosphorus burning rate, Ib/hr 2, 650
Temperature, °F
Vaporizer outlet 1,650
Burner outlet 880
Venturi scrubber outlet 195
Stack gas 175
Pressure drop, in. WC
Across venturi scrubber 25.2
Across entrainment separator 1.9
Emissions as % of phosphorous burned 0. 2
In 1962, the TVA constructed a stainless steel
phosphoric acid unit that has an adjustable ven-
turi scrubber, followed by a packed scrubber,
and a wire mesh mist eliminator. When the
venturi scrubber is adjusted to give a pressure
drop of 37 inches of water column or higher,
losses of PTO5 from the unit amount to only
about 5 pounds per hour at phosphorus-burn-
ing rates up to 6, 000 pounds per hour.
Considerable research and development work
by the TVA demonstrated that good recovery of
phosphoric acid mist could be achieved by intro-
ducing water vapor into the hot gases from, the
combustion of phosphorus, passing the mixture
through a packed tower, and condensing it
(Slaik and Turk, 1953).
The Tennessee Valley Authority has used elec-
trical precipitators for many years to reduce
the emission of phosphoric acid mist (Striplin,
1948). Severe corrosion has always been a
problem with these precipitators. Published
data (Slaik and Turk, 1953) indicate that the
problem has been partially solved by reducing
the tail gas temperature to 135° to 185°F.
The acid discharged amounts to about 0. 15 per-
cent of the phosphorus pentoxide charged to the
combustion chamber as phosphorus. The rela-
tively low gas temperatures and consequently
infrequent failure of the wires are given as the
reason for the high mist recovery from the gas
stream.
The TVA replaced one of the electrical precip-
itators with a venturi scrubber in 1954. The
venturi scrubber is constructed of stainless
steel and is 14 feet 6 inches high, with a 30-
inch-diameter inlet and outlet and a 11-1/2-
inch-diameter throat (Barber, 1958). The
scrubber is followed by a centrifugal entrain-
ment separator. Stack analyses of emissions
from this production unit are summarized in
Table 199.
A large-scale plant using a Raschig ring-packed
tower followed by three gas coolers was built.
Overall phosphorus pentoxide recovery exceeded
99. 9 percent, but the process -was eventually
abandoned because of the excessive rate of corro-
sion of the gas coolers.
This same process, with a second packed scrubber
or glass fiber-packed filter unit for acid mist re-
moval replacing the gas cooler, is used by a number
of phosphoric acid producers throughout the country.
These plants routinely operate with phosphorus
pentoxide recovery efficiencies in excess of 99. 8
percent. A visible phosphoric acid plume still
remains, though the phosphorus content has been
reduced to less than 0. 1 grain per scf. A plant
such as this is in operation in Los Angeles County
and is shown in Figure 521. The plume contains
a large percentage of water vapor and does not
violate local air pollution prohibitions. Stack
analyses of emissions from this plant are shown
in Table 200.
The packed scrubber must be thoroughly and uni-
formly wetted with either water or weak acid and
must have uniform gas distribution to achieve high
-------
704
CHEMICAL PROCESSING EQUIPMENT
Table 200. STACK ANALYSES OF EMISSIONS FROM A PHOSPHORIC ACID
PLANT WITH TWO RASCHIG RING-PACKED SCRUBBERS
Phosphorus burning rate, Ib/hr
Gas rate, stack outlet, scfm
Gas temperature, stack outlet, °F
Diameter of first packed scrubber, ft
Height of first scrubber's Raschig ring packing, ft
Diameter of final packed scrubber, ft
Height of final scrubber's Raschig ring packing, ft
Final scrubber's superficial velocity, fpm
P2O5 emitted, gr/scf
P2O5 emitted, Ib/hr
Emissions as % of phosphorus burned
Report series No.
C-167 A
1,875
12,200
175
8.5
12
20
3
47
0.095
9.9
0.23
C-167 B
895
3,420
162
8.5
12
20
3
13
0. 108
3.2
0. 16
collection efficiency. Good gas distribution is also
mandatory for glass fiber filter units, and a super-
ficial gas velocity of less than 100 fpm is recom-
mended.
Figure 521. Phosphoric acid plant with a Raschig
ring-packed scrubber.
The Brink (1959) fiber mist eliminator is a rela-
tively new type of collector that has been used
successfully on sulfuric acid mist, oleum, phos-
phoric acid, ammonium chloride fume, and various
organics. Collectors of this type have been dis-
cussed in the preceding section of this chapter.
At one plant owned by Monsa.nto Chemical Company,
the stack plume was very persistent and visible.
Thirty milligrams of fine sulfuric acid mists per
standard cubic foot and 80 to 200 milligrams of
phosphoric acid particles per standard cubic foot
•were emitted from the stack. To correct the
situation, a gas absorption apparatus followed
by a fiber mist collector was installed. Collec-
tion efficiencies of 99 percent on particles less
than 3 microns in diameter and of 100 percent
on larger particles were achieved. The stack
plume, which consists of 15 percent •water vapor,
disappears within 40 to 50 feet of the stack on
dry days and within 150 feet on wet days. No
maintenance problems or changes in pressure drop
through the apparatus have been encountered.
PAINT-BAKING OVENS
Although bake ovens have extensive industrial
applications, this section is limited to those
used to dry or harden surface coatings con-
currently -with the removal of organic solvents
by evaporation. Moreover, the word paint is
used throughout this section as a general term
for any of the many and varied types of surface
coatings, for example, inks, varnishes, paints,
enamels, lacquers, shellacs, and resins.
Paint baking causes not only the evaporation of
the organic solvents used as diluents and thinners
but also the drying and hardening of a surface
coating. The essential requirement in paint bak-
ing is that the paint be exposed to the proper de-
-------
Paint-Baking Ovens
705
gree and amount of heat, but there are other re-
quirements, too, as follows:
1. Within the oven, the atmosphere resulting
from the vaporization of organic solvents
from the paint must be maintained below
the lower explosive limit (LEL).
2. When the doors are open and employees
are loading or unloading the oven, the at-
mosphere within the oven must be kept well
below the toxic level.
3. The atmosphere in which the painted surface
is baked must, in some cases but not all, be
kept free from the products of incomplete
combustion of the oil or gas used for firing.
4. The atmosphere within the oven must be
free from dust.
5. The nuisance and air pollution potentials
of emissions from the oven to the outside
atmosphere must be evaluated.
BAKE OVEN EQUIPMENT
A bake oven is a heated enclosure used indus-
trially to dry and bake materials at elevated
temperatures. In paint drying and baking,
these temperatures may range from 100° to
600°F. The typical construction of a bake
oven consists of a frame-work of heavy struc-
tural steel that supports an inner and outer
shell of heavy-gage steel sheet metal. The
space between the inner and outer shell is filled
with insulation that should be supported to pre-
vent separation and settling. Allowance should
be made for expansion and contraction due to
temperature changes, and the amount of steel
in contact with both hot and cold sides should
be kept to a minimum to reduce heat loss.
Heavy, insulated double doors with approved
explosion-type catches are characteristic of
industrial bake ovens.
In some ovens, the products of combustion enter
and come into direct contact with the work in
process; in other ovens, the heating is indirect,
and the products of combustion do not enter the
oven nor do they at any time come into contact
with the -work in process. The source of heat
may be gas, electricity, oil, steam, or infra-
red lamps, whichever is available and appro-
priate to the process. In all bake ovens, ac-
curate, dependable temperature control and
uniform heating are requisites. All three
methods of heat transfer are used in any paint
bake oven. The heat radiates from the hot
oven walls. The movement of the heated air
by means of circulating fans applies heat by
mechanical convection. The interior of the
paint film is heated by conduction through the
article upon which the paint has been applied
as well as by internal conduction through the
paint itself.
An exhaust fan should be provided for all but
the smaller ovens. The exhaust duct's intake
openings should be located in the area of the
greatest concentration of vapors. In general,
the organic vapors from the volatile organic
solvents customarily used in paints are heavier
than air. For this reason, bottom ventilation
of a paint-baking oven is indicated. The prod-
ucts of combustion from the burning of fuel are,
however, lighter than air and should, there-
fore, be vented from the top. The products
of combustion from indirectly gas- or oil-
fired air heaters are preferably exhausted by
a separate draft fan not connected to the oven's
ventilation.
In addition, every properly designed bake oven
has a number of automatic safety features to
meet the recommendations of insurance inspec-
tion services, but these features will not be dis-
cussed here since they can be learned from the
National Fire Protection Association (1963).
Batch process paint-baking ovens have an in-
sulated chamber, some form of air circula-
tion, a combustion system, a heat exchanger,
a variety of safety controls, fresh air filters,
and either a natural-draft or an induced-draft
exhaust system (Figure 522). The painted
Figure 522. An indirectly heated, gas-fired, re
circulating, batch-type paint-baking oven.
-------
706
CHEMICAL PROCESSING EQUIPMENT
products to be baked may rest on permament
racks or hooks inside the oven or may be placed
on trucks that can be moved in and out of the
oven for loading or unloading. Batch paint-bak-
ing ovens offer the advantage of low investment
and are completely adequate for many jobs. Be-
cause the items to be baked are all put into the
oven at one time, the organic solvents do not
evaporate at a constant rate. Since a peak
evaporation rate is reached -within a few min-
utes after loading, all the organic solvents will
have evaporated long before the work load is
removed from the oven.
For large-volume production, continuous-type
paint-baking ovens are usually used. These
are high-production machines that permit a
precise control of baking conditions. They
can be designed and built in units to meet any
production requirements. A continuous bake
oven, as shown in Figure 523, consists prin-
cipally of an insulated cabinet with positively
controlled circulation of heated air, combus-
tion systems, safety controls, fresh air fil-
ters, induced-draft exhaust, and a moving con-
veyor or belt by which the painted product is
carried into, through, and out of the oven.
Automatic control devices maintain any de-
sired baking conditions. Because the workload
is introduced into a continuous oven in a steady
stream by means of an endless belt or convey-
or, the evaporation of organic solvents approach-
es a constant rate.
THE AIR POLLUTION PROBLEM
The air pollutants emitted from paint-baking
ovens are as follows:
1. Smoke and products of incomplete combus-
tion arise from the improper operation of a
gas- or oil-fired combustion system used
for heating the oven.
2. Organic-solvent vapors arise from the
evaporation of the thinners and diluents
used in the surface coa.tings. A classifica-
tion of the organic solvents used in surface-
coating formulations, giving general formu-
lae and examples, is shown in Tables 201
and 202. The composition of the organic-
solvent vapors emitted from a paint-baking
oven might be expected to have the same
composition as that of the organic solvents
used in the formulation and thinning of the
surface coating, but they do not. Partially
oxidized and polymerized compounds are
produced at bake oven temperatures. When
effluent from paint-baking ovens is irradi-
ated in the presence oi NO, it can produce
eye irritation as severe as that produced by
automobile exhaust.
3. The aerosols resulting from the partial
oxidation and polymerization of the organ-
ic solvents and resins used in the paint
formulation are obnoxious from the stand-
CUKIAIN
Figure 523. A direct-heated, gas-fired, recirculating, continuous
paint-baking oven, as used in Example 41 (Zone 1 is 4 ft wide; zone 2,
5 ft 4 in. wide; zone 3, 4 ft wide.).
GPO 8O6—614—24
-------
Paint-Baking Ovens
707
Table 201. CLASSIFICATION OF ORGANIC SOLVENTS
USED IN SURFACE COATING (Lunche et al. , 1957)
Class name
Aliphatic hydrocarbons
Aromatic hydrocarbons
Ketones
Alcohols (and glycols)
Ethers
Esters
Miscellaneous
General formulaa
R-H
4>-H
0
II ,
R-C-R
R-OH
R-O-R'
ft ,
R-C-O-R
Examples
Hexane, Stoddard solvent, naphtha,
mineral spirits
Benzene, toluene, xylene
Methyl, ethyl ketone, acetone,
methyl isobutyl ketone
Methanol, isopropanol, sec-
butonol, ethanol
Ethyl ether
Ethyl acetate, butyl acetate,
n-butyl acetate
Turpentine, carbon disulfide,
nitromethane
aR or R = any straight- or branched-chain hydrocarbon radical.
= any benzene ring-type hydrocarbon radical.
O = oxygen atom.
C = carbon atom.
H = hydrogen atom.
Aldehydes, terpenes, sulfur compounds, nitrogen compounds, mixtures.
points of odor nuisance. Moreover, these
emissions are extremely irritating to eyes
even without irradiation.
Table 202. CLASSIFICATION OF ORGANIC
SOLVENTS AND EXAMPLES
(Lunche et al. , 1957)
Mineral -spirits and terpenes
Aliphatic hydrocarbons and mixtures
Turpentine
D i|x.ntcno
Hexane
Cyt lohexane
Naphtha
Aromatic hydrocarbons
B< n^ol
Xylols
Aromatic mixtures
Kctones
Acetone
Mfthyl ethyl ketone
Dusobutyl ketone
Cyi lohexanone
Ethyl amyl ketone
Diacetone alcohol
Isophorone
Mesityl oxide
Mixed ketones
Esters
Ethyl acetate
Amyl acetate
N-butyl acetate
Cellosolve acetate
N, isopropyl acetate
Methyl amyl acetate
Mixed esters
Alcohols and ^lycol ethers
Methanol
Ethanol
Isopropyl alcohol
N-butanols
Sec-butanols
Cellosolvi-s (not including esters)
Methyl isobulyl carbmol
Mixed alcohols
Chlorinated hydrocarbons
Trie hloroethylen
Mono-chloroben ene
Di-chlorobenzen
Ethylene chchlor ie
Carbon tetrachlt ide
Chloroform
Chloroiethene
Mixed chlorinated solvents
Others
Carbon bisulfide
Dimethyl formamide
Nitro benzene
HOODING AND VENTILATION REQUIREMENTS
Fire underwriters' standards demand that suffi-
cient fresh air be adequately mixed with the or-
ganic-solvent vapors inside the oven so that the
concentrations of flammable vapor in all parts of
the oven are safely less than the lower explosive
limit (LEL) at all times. The LEL of gas in air
is the minimum volume at which it will burn,
expressed in percent by volume. As an approxi-
mate rule, the vapors produced by 1 gallon of
most organic solvents, when diffused in 2, 500
cubic feet of air at 70°F, form the leanest mix-
ture that still explodes or flashes in the presence
of a flame or spark. A factor of safety four
times the LEL is customarily provided. For
each gallon of organic solvent evaporated in a
paint-baking oven, therefore, at least 10, 000
cubic feet of fresh air (computed at 70 °F) must
be supplied to the oven. Ovens vented to solvent
recovery systems, however, are exempted from
this requirement.
Additional requirements of the fire underwriters'
standards are:
1. The exhaust duct openings shall be located
in the area of the greatest concentration of
vapors.
2. The oven must be mechanically ventilated
with power-driven fans.
-------
708
CHEMICAL PROCESSING EQUIPMENT
3. Each oven shall have its own individual ex-
haust system (there are some exceptions
for very small ovens), which is, moreover,
not connected with any exhaust system used
to vent the products of combustion from in-
directly gas- or oil-fired heaters.
4. The fresh air supplied shall be thoroughly
circulated to all parts of the oven.
5. Dampers must be so designed that, even
when fully closed, they permit the entire
volume of fresh air needed for meeting the
demands of safe ventilation to pass through
the oven,
6. A volume of air equal to that of the fresh
outside air supplied must be exhausted
from the oven in order to keep the system
in balance.
7. If a shutdown occurs during which vapors
could accumulate in the oven, the oven
shall be purged for a length of time suffi-
cient to permit four complete oven volume
air changes.
The designer of air pollution control equipment
must be concerned with the fire underwriters'
standards for paint-baking oven installations.
It is pointless to design a control system for
ovens that does not meet the LEL standards of
the fire underwriters. The large volumes of
air used to dilute the organic-solvent vapors
in these ovens have a major effect on the de-
sign and operation of the air pollution control
equipment.
AIR POLLUTION CONTROL EQUIPMENT
Smoke and products of incomplete combustion
can be eliminated as a source of air pollution
by proper selection of burners and fuels along
with observance of correct operating proce-
dures (see Chapter 9, Combustion Equipment).
The air pollution problem can be diminished
by decreasing the operating temperature in-
side the paint-baking oven. Eye irritation has
repeatedly been observed from oven emis-
sions when the operating temperature of the
oven was above 375 °F. Markedly less eye ir-
ritation was noted when the operating tempera-
ture of the oven was kept below this tempera-
ture. In one case, the reduction of excessive
oven temperature resulted in an improvement
of product quality with no loss of production.
This indicates that there undoubtedly are op-
timum temperature and time relationships in
bake oven operations that could be exploited
in the interests of reducing air pollution.
The use of water-based surface coatings offer
another possible opportunity to reduce organic-
solvent emissions from paint-baking ovens.
Organic-solvent vapors, odors, and aerosols
emitted from paint-baking ovens can best be
controlled by being vented to direct-fired
vapor combustion devices operated at tem-
peratures of 1,400°F or higher. Catalytic
afterburners have not provg'd as satisfactory
as the direct-fired afterburners for this use.
Tables 203 and Z04 show efficiencies of direct-
flame afterburners and cata.lytic afterburners,
respectively. In some cases, as shown in Table
204, the organic emissions from a paint-baking
oven have been more offensive and irritating to
eyes after passing through a catalytic afterburner
than they were on entering the afterburner. This
is true even at temperatures of 900° to 1,000°F,
considered to be high for catalytic units.
COST OF DIRECT-FLAME AFTERBURNERS
The cost of installing a direct-flame afterburn-
er for the control of organic-solvent vapor emis-
sions from a paint-baking oven depends upon the
capacity of the afterburner, the amount of in-
strumentation required, arid the degree of cor-
rosion resistance needed. The costs listed in
Table 205 are for complete installations only
and include the costs of the afterburner, the
foundation work, the ductwork, the utilities,
(installed and connected), complete instrumen-
tation, the furnace and burner controls, the fan,
the stack, and also a test involving all neces-
sary adjustements.
Illustrative Problem
Example 41
Given:
A continuous oven (similar to that shown in Fig-
ure 523) is to be used to bake steel parts of
various shapes for metal furniture. The parts
to be baked enter the oven at one end on an over-
head conveyor at 3 to 7 fpm conveyor speed and
leave at the other end. The parts are to be
baked at 375"F. They are coated with enamel
mixed with thinner in the proportion of 4 gallons
of enamel to 1 gallon of thinner. The total con-
sumption of thinned enamel does not exceed 40
gallons per 8-hour day.
The enamel weighs 9. 2 pounds per gallon and
49 percent is nonvolatiles (see Table DID in
Appendix D). The thinner weighs 7 pounds per
gallon.
A direct-fired heater and a 5-hp blower pro-
vide 4, 800 cfm heated air. This air circulates
through the oven and the direct-fired heater
where a portion of the fumes from, the baking
-------
Paint-Baking Ovens
709
Table 203. EFFICIENCY OF DIRECT-FLAME AFTERBURNERS IN THE CONTROL OF
CONTAMINANTS EMITTED FROM PAINT-BAKING OVENS
Source
test
No.
C-722
C-722
C-767
C-776
C-795
C-820
Type of
coating
applied
Vinyl
Enamel
(sanitary)
Vinyl
No. 53
Varnish
No. 127
(oleoresin)
Varnish
(alkyd resin)
Varnish
(alkyd amine)
Class of
solvent
useda
Ketones,
aroma tics
Mineral
spirits ,
aromatics
Aromatic s,
ketones
Aromatics,
aliphatic s,
mineral
spirits
Alcohols,
aromatics ,
aliphatics
Mineral
spirits,
aromatics
Afterburner
temp,
°F
1, 100
1,200
1,400
1, 100
1,200
1,400
1, 100
1,200
1,300
1,400
1, 100
1, 200
1,300
1,400
1, 100
1,200
1,300
1,400
1,200
1,300
1,400
Reduction obtained, %
Odorb
15
40
98.8
2.5 xf
42
94.2
0
90
86
98.3
-
1. 7 xf
40
98. 3
98.6
-
Partic-
ulates0
58
70
87
1.2 xf
1. 5
74
-
-
-
;
Combustible
gases'^
.:
—
1.2 xf
1.5
34
95
29
38
39
69
41
48
75
95
60
79
92
Aldehydes6
1.3 xf
18
57
1. 6 xf
1. 1 xf
81
-
-
-
-
Refer to Table 202.
"Odor concentrations were determined by the Los Angeles County Air Pollution Control
District's Standard Method (see Appendix C).
°Particulate matter was determined by the APCD Source Test Method.
Combustible gases were determined by the CCIR Method.
eAldehydes were determined by the APCD 5-46 Method.
This notation represents an increase in concentration. The number is a multiplying factor to
be applied to the afterburner's inlet concentration.
process is burned on each air change. Since
the inside volume of the oven is 2, 252 cubic
feet, there are 2. 13.air changes per minute
in the oven (4, 800 cfm/2, 252 cf = 2. 13).
Two 3-hp blowers provide air seals, one at
each end of the oven. A 3/4-hp exhauster re-
moves a portion of the contaminated air and
provides one fresh air change in the oven each
1. 5 minutes in order to remove the fumes aris-
ing from the paint-baking process. The burn-
er supplied with this oven is rated at 1, 200, 000
Btu per hour. Assume the overall heat trans-
fer coefficient for this oven is 0. 66 Btu per
hour per square foot per °F.
Problem:
Solution:
1. Total weight of solvent emitted per day:
The maximum daily usage of thinned enamel
is 40 gallons, and this is a blend of 4 gallons
of unthinned enamel plus 1 gallon of thinner.
There are, therefore, (40/5)(4) = 32 gal-
lons of unthinned enamel used per day plus
(40/5)(1) = 8 gallons of thinner per day.
The volatiles in the unthinned enamel are
(32)(9.2)(1. 00-0.49) = 150 pounds, and the
volatiles in the thinner are (8)(7) = 56 pounds
per day.
Determine the design features of an air pollu-
tion control system incorporating an afterburn-
er to serve this oven.
150 + 56 = 206 Ib solvent emitted per day
-------
710
CHEMICAL PROCESSING EQUIPMENT
Table 204. EFFICIENCY OF CATALYTIC AFTERBURNERS IN THE CONTROL OF
CONTAMINANTS EMITTED FROM PAINT-BAKING OVENS
Source
test
No.
C-239
C-258
C-276
C-374
C-374
C-374
C-374
C-375
C-375
C-375
C-375
C-391
C-410
C-579
O-12-62
Type of
coating
applied
C-Enamel
Phenolic
Varnish
(alkyd resin)
Vinyl
(No. 52 Al.)
Varnish
Vinyl
(No. 53 Gold)
Enamel
(No. 127)
Vinyl
(No. 52 Al.)
Enamel
(No. 127)
oleoresin
Varnish
(No. 10-304)
Vinyl
(No. 53 Gold)
Vinyl
Vinyl
Varnish
(No. 2201)
(alkyd amine)
Enamel
(No. 1106 B)
oleoresin
Class of
solvent
useda
Aromatics,
aliphatic s
Aromatics,
aliphatics ,
alcohols,
ketones
Alcohols,
aromatics ,
aliphatics
Aromatics,
ketones
Aromatics ,
aliphatics,
alcohols
Aromatics,
ketones
Aromatics,
aliphatics,
mineral
spirits
Aromatics,
ketones
Aromatics ,
aliphatics ,
mineral
spirits
Aromatics ,
aliphatics,
mineral
spirits
Aromatics ,
ketones
Aromatics ,
mineral
spirits,
ketones
Aromatics ,
mineral
spirits,
ketones
Mineral
spirits,
aromatics
Afterburner
temp,
°F
740
760
760
700
300
900
800
925
1, 200
800
925
1, 050
970
1, 060
800
925
1, 050
700
800
900
700
800
900
700
800
900
700
800
900
950
950
1, 010
700
800
900
1,000
850
950
1, 000
Reduction obtained, %
Odorb
-
-
-
-
-
-
-
-
-
-
-
-
1.2*1
1.2 xf
1.2 x£
1.2 x£
2. 9 xf
1. 4 xf
36
Particulatesc
1.1 xf
1.1 x£
1. 4 x£
3. 5 x£
4. 3 xf
65
41
50
71
25
21
60
63
60
11
0
13
33
57
62
15
1. 9 x£
1. 4 xf
1. 1 xf
46
39
61
58
66
55
48
79
-
_
Combustible
gases
-
-
-
-
-
_
-
-
-
-
":
-
77
85
82
89
_
Aldehydes6
100
59
33
";
33
1.3 x£
1 25 x£
38
1. 4 xf
29
1. 8 xf
1. 8 xf
0
50
33
*.7xJ
3.2 xf
2.2 xf
89
75
33
70
24
1. 4 x£
13
1. 4 x£
1.3 x£
7.5 x£
10 xf
4. 7 x£
-
_
aRefer to Table 202.
^Odor concentrations were determined by the L.OS Angeles County Air Pollution Control District's
Standard Method (see Appendix C).
cParticulate matter was determined by the APCD Source Test Method.
"Combustible gases were determined by the CCIR Method.
eAldehydes were determined by the APCD Method 5-46.
This notation represents an increase in concentration. The number is a multiplying factor to be
applied to the afterburner's inlet concentration.
-------
Paint-Baking Ovens
711
Table 205. COST OF DIRECT-FLAME
AFTERBURNER, $/scfm CONTAMINATED
GAS TO BE INCINERATED
Volume of contaminated gases,
scfm
500
1, 000
2, 000
4, 000
Cost, a
$/scfm
15
10
7
6
These costs are for complete installations and include:
the afterburner, the foundation work, the ductwork, the
utilities installed, complete instrumentation, the furnace
and burner controls, the fan, the stack, and also a test
involving all necessary adjustments
2.
3.
Quantity of solvent emitted from the oven
per day:
Of the 206 pounds of solvent emitted to the
atmosphere per day, 60 percent is charge-
able to the spray coating operation, and 40
percent to the paint-baking oven (see Table
206). The selection of the percent loss to
be charged to the oven depends upon the
time that elapses from the application of
surface coating until the item enters the
oven, the item being coated, the type sur-
face coating applied, the type and amount
of thinner used, the ambient temperature,
and the effect of stray air currents.
(206)(0.40)
7
11.8 gal solvent emitted
per day
Volume of contaminated air exhausted from
the paint-baking oven per 8-hour work day:
diluents in this enamel are not specified,
assume the thinner to be a mixture of
xylol and mineral spirits. At 70°F, the
LEL of xylol is 1.0, and that of mineral
spirits is 0.77 (refer to Table 207). When-
ever multicomponent solvents are used, the
individual solvent chosen should be that
whose data result in the largest volume of
air required. In this case it will be mineral
spirits. The volume of air at 70°F that is
rendered barely explosive per gallon of
solvent is given in column J, Table 207.
It can also be computed, if necessary, by
the following formula:
(8. 33 (G)(100-LEL)
(0.075)(/os)(LEL)
volume in ft"
70°F
at
Table 206. SOLVENT LOSS, % TO BE CHARGED
TO A PAINT-BAKING OVEN OPERATED IN
CONJUNCTION WITH VARIOUS METHODS OF
APPLYING THE SURFACE COATING
Method of applying
the surface coating
Spray coating
Spray coating large flat surfaces
Dip coating
Flow coating
Roller coating
L,ossa charged to
the oven, %
10 to 30
20 to 40
30 to 60
30 to 60
50 to 80
aThe selection of % of loss depends upon the time
that elapses from the application of the surface coat-
ing until the item enters the oven, the item being
coated, the type of surface coating being applied,
the type and amount of thinner used, the ambient
temperature, and the effect of stray air currents.
2,252 ft
1. 5 min
1, 500 cfm contaminated air at
375 °F exhausted from the paint-
baking oven
8.33 = weight in pounds of 1 gallon of
water at 70°F
(1,500)(60)(8) =
720, 000 ft /day of contam-
inated air exhausted from the
paint-baking oven
4. Volume of contaminated air exhausted per
gallon of solvent emitted:
720,OOP
11.8
= 61, 017 ft air/gallon solvent
5. Safety factor of oven versus that required
by the fire underwriters' standard:
The safety factor required by the under-
writers is 4 x LEL (in order to prevent
the vapor concentration from exceeding
15% LEL). Since the solvents used as
G = specific gravity of the solvent
(water = 1.0)
LEL = lower explosive limit of the sol-
vent
0. 075 = weight, Ib of 1 ft of air at 70°F
and 29. 9 inches of mercury pres-
sure
ps = vapor density of the solvent (air =
1.0).
For mineral spirits, therefore:
(8.33)(0.8)(100-0.8)
(0.075)(3.9)(0.8)
_
- 2'825
at 70°F
^ 70 F
-------
712
CHEMICAL PROCESSING EQUIPMENT
Table 207. PROPERTIES OF COMMONLY USED FLAMMABLE LIQUIDSa
(National Fire Protection Association, 1963)
Acetone
Amyl acetate n
Amyl acetate iso
Amyl alcohol n
Benzol {benzene)
Botyl acetate n
Butyl alcohol n
Butyl propionate
Carbon disulfide
Cellosolve ^ethyl cellosolve)
Cellosolve acetate
Cresol m or p
Cyclohexane
Cyclohexanone
Cymene-para
Denatured alcohol
Dibutylphthalate o
Dichlorohenzene ortho
Diethyl ketone
Dimethyl lormarrude
Dioxane--l,4
(diethylene dioxide)
Ethyl acetate
Ethyl alcohol
Ethyl ether
Ethyl lactate
Ethyl methyl ether
Ethyl propionate
G.,0J,ne
Methyl acetate
Methyl alcohol
Methyl carbitol
Methyl cellosolve acetate
Methyl ether
Methyl ethyl ketone
Methyl lactate
Mineral spirits No. 10
Naphtha (V.M. and P. Regular)
Naphthalene
Nitrobenzol
Nitroethane
Nitromethanee
Nitropropane--!
Nitropropane--2
Paraffin oil
Propyl acetate--iso
Propyl alcohol n
Propyl alcohol--iso
Propyl ether--iao
Pyridine
Rosin oil
Soybean Oil
Toluol
Turpentine
Vinyl acetate
Xylene (xylol)
A
Wlb
58
130
88
88
78
116
74
130
152
76
90
132
I 13
108
84
312
134
278
147
36
73
88
88
46
74
118
60
102
74
32
120
1 18
46
72
104
128
123
75
61
89
89
102
60
60
102
79
92
86
106
B
paint,
°F
0
77
77
91
11
72
84
90
150
-12.
104
124
90
202
-4
ill
117
60
315
1S1
136
54
24
55
-49
115
-35
54
-45
100
H
52
200
132
Gas
21
121
104
20
174
190
82
95 (o.c.)
120 (o c.)
103 (o.c.)
444
< 0
40
59
53
-18
68
266
540
40
95
18
63
C
°F
1,000
714
715
572
1,044
790
650
300
371
211
460
715
1, 180
1,038
500
788
817
750
757
1 193
833
356
800
793
356
752
374
890
495
444
935
867
662
960
725
47J
450
979
900
773
785
789
802
550
860
700
750
830
900
648
8i3
997
488
800
867
Lower
2.6
1. 1
1.0
at 212"F
1. 2
1. 4
at 212° F
1 7
1. 4
1. 3
2.6
1. 7
1. 3
at 212°F
1 I
at 102 'F
1. 3
i 1
at 212'F
0. 70
at 212 T
2.2
2 2
at 2I2°F
I, 0
2. 5
4, 3
1. 9
1 5
at 212°F
2.0
1. 9
1. 4
0. 7
3. 1
7 3
3.4
1.8
2 2
t 212*F
0. 3
at 212T
0. 92
0 9
1 8
at 200"F
4 0
7 3
2, 6
2.6
1. 1
1 8
2. 1
2.0
1 4
1.8
i. 4
0,8
2.6
1 0
D
Upper
12.8
7. 5
7 5
10.0
at 212°F
7. 1
7.6
11.2
44
15.7
7, 1
at 302°F
8
9.-Z
11
9
19
48
10 1
1 1
7.6
5
16
36
18
10
6.0
5 9
5 9
8
n. 5
12
21
12. 4
6 7
13. 4
6.0
E
Watcr-l)
0 8
0. 9
0. 9
0.8
0. 9
0 9
0. 8
0. 9
1. 0
1. 3
0.931
0. 975
1 \
1, 0
0.8
0, 9
0.9
0.8
! 04
1. 3
0 816a
0 9
1.0+
0.9
0 «
0,7
i.O-f
0.7
0 9
O.S
< 1.0
0.9
0. 9
0,B
1,035
1.005
0 8
0.8
0.75
1. 1
1.2
1. 1
1 1
1 0
1.0
0.6
0. 9
0 8
0.8
0.7
1 0
1 0
0.9
1 6b
0. 9
•- 1
0. 9
0.9
Pamphlet
F
(Air 1
2. 00
4. 5
4 5
3.0
2. 8
4. 00
2. 6
4. 5
5. 2
2 6
3. 10
4.72
3. 9
2. 9
3, 4
4.6
1 6
5 J
2 5
3.0
3. 0
1.6
2 6
4. i
2. 1
3. 5
3. 4
£ 6
L. 1
4. 14
4.07
1.6
2 5
3. 6
3. «
3. 73
4, 4
4. 5
2. 6
2 1
3 1
3. 1
2 5
3 5
2 1
2. 1
3. 5
2 7
i 1
). 0
3 7
G
Hoilinj:
°F
1J4
300
290
2SO
176
260
243
29^
408
1 15
27S
313
270
395
179
313
349
175
690
3S6
_,lb Qb
307
214
371
173
^5
309
50
210
100 to 400
504 lo 574
600
HO
147
37°
289
-11
176
293
300
212 to 320
424
412
237
214
263
248
95 to 140
194
207
181
1 Sfc
2 (9
> 680
231
300
161
292
H
6 66
7 50
7. 50
6 66
7. 50
7 50
6. 66
7 s
8. 33
10.83
7. 75
8. 13
9. 16
8. 33
6.66
7. 50
7.50
6.66
8. 33
10. 33
6.79b
7, 50
B. 33
7.50
6 66
5.83
8.33
5 83
7. SO
6 66
7, 50
6.66
8.62
8.38
6 66
9. 16
6.66
6.24
9. 16
10.0
9. 16
9. 16
8. J3
8. 33
S.OO
7. S
6.66
6.66
5.83
8. 33
8. 33
7. SO
13 3b
7. 5
7. 5
7 5
I
Vapor
liquid
44. 4
22. 2
22 I
29. 6
35 7
24. 9
34. 2
22. S
21.4
55. 5
33. 3
23. 0
31. 3
29.8
30.6
29 1
21 8
55.5
28. 3
39, 9
36.6
32 9
55.5
29 9
27.0
37 0
48. 5
29.6
38
3.02
3. 10
6. 34
4.29
4 29
5. 37
3.80
6.35
6.35
3.81
4. 93
•4 29
4. 44
3.6
Approximate ft3
bar*ly explosive
per gal of solvent^ d
1, 663
1, 97*,
2, 198
2,437
2, 503
1.435
2,408
4.ZM
i,24&
1,32?
2,37f,
2,67^
2, 323
£,*!*>
3,094
1,258
1,774
1,795
1.281
1.^35
1,545
1,775
1,811
1,474
2.0BO
1. IBS
1,024
1, 762
1,507
2, 802
3,052
1,689
738
1,341
1, 34)
2, 392
1,555
1, 972
2.073
1, 563
2,259
2.26B
1,248
2,673
purity
and ai
on]y
arge.'
applu
1.4, 4, The ma
tiply by
-------
Paint-Baking Ovens
713
2,825(460 + 375) 3 .
(460 + 70) = 4'450ft /gala
The safety factor is —' = 13. 7
4,450
Thus, the volume exhausted is satisfactory,
being well above the required safety factor
of 4.
6. Quantity of particulate matter per cubic
foot of contaminated air exhausted from
the oven:
Assume that 20 percent of the solvent
evaporated in the paint-baking oven is in-
cinerated and 80 percent is exhausted.
Hence, 206 pounds of solvent used per day
x 40 percent charged to the oven x 80 per-
cent exhausted equals 65.9 pounds of sol-
vent exhausted from the oven per day.
The amount of particulate matter that may
be formed from solvents evaporated in a
paint-baking oven can be predicted from
Figure 524. The graph indicates that 385
grains of particulate matter is formed
per pound of solvent evaporated at 375°F.
7. Check the oven heat load against the rated
capacity of the burner supplied with the
oven:
(a') Heat loss to the surrounding air from
the oven surfaces:
Q1 = UAAT
where
U = overall heat transfer coefficient,
Btu/hr per ft^ per °F
A = outside oven area, ft
AT = temperature difference between in-
side of oven and ambient air.
Total outside oven area:
9 ft x 27 ft x 2 = 486 it'
4 ft x 27 ft =108
q
fl
7
g
•i
4
1
?
1
X'
X
4
>
(/
/
<
\^>
f
{/
-------
714
CHEMICAL PROCESSING EQUIPMENT
9 ft x 17 ft x 2 = 306
4 ft x 17 ft = 68
9ft x 14ft x 2 = 252
5 ft - 4 in. x 14 ft = 75
u
al
1,295 ft2
0. 66 Btu/hr per ft2 per °F (given)
(0. 66)(1,295)(375-60)
269,230 Btu/hr
(b) Heat required to raise the material
being processed to the baking tem-
perature:
Q, = W C . At. + W, C At + W h
2 Ipll 2p22 3v
where
W
At =
and W =
wt of steel parts = 20, 000 lb/8 hr
2,500 Ib/hr
specific heat of steel = 0. 12
Btu/lb per °F
(T - T ) = 375 - 60°F = 315°F
the wt of thinned, partially dried
enamel adhered to the metal parts
entering the oven per hour. This
is composed of the 40 percent of
the solvent chargeable to the oven
(see step 2) plus the nonvolatile
portion of the enamel.
Q = (2,500)(0. 12)(315) + (29.
(10. 3)(150) = 100,626 Btu/hr
(c) Heat required for oven makeup air:
Q, = WC At
3 4 p4
where
(11.8 gal /day)(7 Ib/gal
8 hr/day
= 10. 3 Ib solvent per hr
W = "weight of fresh air
= (0.0755) -^-(1,500) 4- (60)
,3 mm hr
460 + 60
460 + 375
= 4,233 Ib/hr
C = average specific heat of air = 0. 24
P Btu/lb - °F
At = 315°F
Q = (4, 233)(0.24)(315) = 320, 015 Btu/hr
(d) Total oven heat load:
Q = Q + Q + Q =269,230+ 100,626 +
t J, L, J
320,015 = 689,870 Btu/hr
The burner supplied is rated at
1, 200, 000 Btu/hr and is ample for
this application.
Mass flow rate of the contaminated gases
from the oven to the direct-flame after-
burner:
(32 gal /day)(9.Z Ib/gal )(51%) _18. 8 Ib nonvolatile
8 hr/day per hr
10. 3 Ib/hr + 18 Ib/hr = 29. 1 Ib/hr enters the oven
C = the specific heat of thinned, partially
P dried enamel, which is assumed to
be 0. 5 Btu/lb per °F
At, = 315°F
and W = wt of solvent per hr that is evaporated
in the oven is 10. 3 Ib/hr
h = the heat of vaporization of the solvent,
which is assumed to be 150 Btu/lb
=56>050ft/hr
Heat required to increase the temperature
of the gases from 375° to 1,400°F:
An afterburner operating temperature of
1, 400 °F is usually sufficient to incinerate
the air contaminants emitted by most paint-
baking ovens.
Use air enthalpies as shown in Table D4
in Appendix D.
Enthalpy of air at 1, 400 °F = 26. 13 Btu/ft
Enthalpy of air at 375 °F = 5. 83 Btu/ft3
Ah = 20. 30 Btu/ft3
(56,050)(20. 3) = 1, 137,815 Btu/hr
-------
Paint-Baking Ovens
715
10. Heat losses from afterburner due to radia-
tion, convection, and conduction:
Losses equal to 15 percent of the total heat
input are assumed. This is a conservative
estimate for afterburners constructed of
firebrick or castable refractory and oper-
ated at 1, 400°F.
(1, 137,815)(0. 15) = 170, 670 Btu/hr
11. Total heat required by the afterburner:
1,137,815+ 170,670= 1, 308,485 Btu/hr
12. Required natural gas volume capacity of
the burner:
Because of the large volume of air used to
dilute the organic solvent vapor in the oven,
the contaminated gases contain sufficient
oxygen to furnish all the combustion air
needed for the proper combustion of the
natural gas supplied to the burner. The
natural gas is, therefore, supplied with
theoretical air from the contaminated gas
stream from the oven. This is introduced
through a premix combustion air blower.
Not only does this provide excellent mixing,
but it has an added advantage in that the
effluent that goes through the burner itself
also goes through the hottest area in the
afterburner, even though only momentarily.
The gross heating value of the natural gas
is assumed to be 1, 100 Btu/ft3. The net
heat available at 1, 400°F from the burning
of 1 cubic foot of natural gas under these
conditions is 939 Btu (see Table Cl in
Appendix C).
(see Table D7 in Appendix D).
1,308,485
939
1, 395 ft /hr
13. Volume of natural gas at 1, 400°F:
With theoretical air, 1 ft3 of natural gas
yields 11.45 ft3 of products of combustion
(see Table D7 in Appendix D).
(1,395)(11.45)(1,400 + 460) _
(3,600)(60 + 460) " b' V "
14. Volume of contaminated gas at 1, 400°F:
Mass flow rate = 56, 050 ft3/hr
Effluent used for combustion air:
Theoretical air required - 10. 36
ft"
ft"
Total effluent used = (1, 395) — (10. 36) :
14, 452 ft /hr
Volume of contaminated gas = 56, 050 -
14, 452 = 41, 598 ft3/hr at 60°F
41,598 (1,400 + 460) ., c , 3
• Tisr „,„ + «„, •«•,»„'.•;'
15. Total volume of gases in the afterburner:
3
15. 9 + 41. 5 = 57. 4 ft /sec
16. Diameter of afterburner orifice:
The effluent from paint-baking ovens de-
mands vigorous treatment if it is to be
rendered innocuous. Orifice velocities of
40 to 60 ft/sec are, therefore, recom-
mended in order to provide adequate mix-
ing of contaminated gases with the burn-
er's combustion products. Use a design
velocity of 50 ft/sec:
Cross-sectional area = —rr— = 1. 15 ft
Diameter = 2
L. 15)(144)'
3.1416
= 14. 5 in.
17. Diameter of the afterburner's combustion
chamber:
Combustion chamber velocities of 20 to 30
ft/sec have been found sufficient to provide
adequate turbulence for completing com-
bustion and to allow the construction of an
afterburner of reasonable length in order
to obtain the required residence time. Use
a design velocity of 25 ft/sec:
57. 4 2
Cross-sectional area = —r-r— = 2, 3 ft
Diameter = 2
>.3)(144)
5. 1416
= 20.5 in.
ft of gas
18. Length of afterburner's combustion^cham-
ber:
Use a design retention time of 0. 5 second:
Length = (25)(0.5) = 12.5ft
-------
716
CHEMICAL PROCESSING EQUIPMENT
Problem note: A comparison of the total
oven heat load (689, 870 Btu/hr) with the
total heat required to operate the after-
burner (1,308,485 Btu/hr) indicates the
possibility of salvaging heat from the
afterburner to provide part or all of the
oven's heat demand. This may be accom-
plished in several ways or in a combina-
tion oFthe following ways:
(a) The fresh makeup air for the oven can
be directed through a heat exchanger
placed downstream from the afterburn-
er's exhaust.
(b) The oven's effluent can be preheated
before entering the afterburner by
providing a heat exchanger downstream
from the afterburner.
(c) A portion of the hot discharge stream
from the afterburner can be diverted,
mixed with the fresh makeup air, and
recirculated through the oven to pro-
vide all the heat required by the oven.
This is a desirable method of conserv-
ing heat energy since it has a high ef-
ficiency and a low equipment cost.
(d) The heat from the afterburner can also
be made available for a variety of plant
processes other than the paint-baking
oven.
SOAPS AND SYNTHETIC DETERGENTS
(lower grade) fats than toilet soap are and con-
tains a slight amount of free alkali. The amount
of alkali should be limited to prevent skin irrita-
tion. Laundry soaps comprise the third grade.
They are prepared from the darker fats and
contain relatively large amounts of free alkali.
These soaps are also available in cake, flake,
granule, or powder form. Laundry soaps also
contain "builders, " which lower costs and aid
in the detergent action. Builders include soda
ash, sodium silicate, sodium tripolyphosphate,
and tetrasodium pyrophosphate. These build-
ers are added for optimum soil removal and
act as water softeners as well as cleaning agents.
Raw Materials
Tallow constitutes about half of the fats and
oils consumed by the soap industry (Molos,
1961). Tallow is a mixture of glycerides ob-
tained by steam rendering cattle fat and, to a
lesser extent, sheep fat. Greases are second
in volume of fatty material used, comprising
about ZO percent of the totaL. Greases are
generally obtained from hogs, small domestic
animals, and garbage. These greases are
obtained by steam rendering or solvent ex-
traction, Usually the greases, tallows, and
other fats are blended (Shreve, 1956).
Coconut oil is the third-ranking source of
fatty acids. It is usually blended with tallow
to increase the solubility of the soap. Other
oils used include palm, palm-kernel, Babassu
nut, cottonseed, soya bean, and peanut (Shreve,
1956).
SOAPS
Soaps consist principally of sodium or potassium
salts of fatty acids containing 12 to 18 carbon
atoms (Kirk and Othmer, 1947). The soaps are
made by reacting sodium or potassium hydroxide
with fats or oils (saponification). They can also
be prepared by neutralizing fatty acids with
sodium or potassium hydroxide or sodium car-
bonate. Generally, sodium soaps are referred
to as hard soaps, and potassium soaps, as soft
or liquid soaps. Technical developments in the
industry now enable sodium soaps to be made
with all the properties of a soft-type soap.
Soaps are produced in a number of different,
but not sharply defined, grades. The best
grades are toilet or castile soaps. These are
made in a bar, paste, or liquid form and contain
little or no alkali. The next grade is made in
bars, flakes, granules, or powders. This type
of soap is used for dishwashing and laundering
woolens or fine fabrics. This soap is also
essentially pure. It is prepared from darker
Hatty Acid Production
Fats and oils may be hydrolyzed or "split" to
obtain fatty acids and glycerol. Separated fatty
acids can then be used for soap or other prod-
ucts. Three general methods are available to
hydrolyze the fats and oils: Twitchell, batch
autoclave, and continuous high-pressure pro-
cess.
The Twitchell process consists of boiling the
fats and oils batchwise in an open tank for 20
to 48 hours with 0. 75 to 1.25 percent Twitchell's
reagent and 0.5 percent sulfuric acid. Twitchell';
reagent consists of alkyl-aryl sulfonic acid or
cycloaliphatic sulfonic acids. Enough water is
used to yield a 5 to 15 percent glycerin solution.
The reaction is usually completed in two counter-
current stages. The aqueous solution contain-
ing the glycerine is withdrawn after settling,
neutralized with slaked lime, and filtered to
remove the calcium sulfate. The glycerine
liquor is concentrated by evaporation. The fatty
acid fraction is decanted from the upper phases.
-------
Soaps and Synthetic Detergents
717
The batch autoclave process operates under
pressures ranging from 75 to 150 psi and tem-
peratures of 300° to 350°F. An oxide catalyst
of zinc, calcium, or manganese is used in the
amount of 1 to 2 percent of the batch by weight.
Higher pressures (425 to 450 psi) and tempera-
ture (450°F) are required if no catalyst is used.
Each batch requires from 5 to 10 hours.
The continuous high-pressure process for split-
ting fats and oils is done in a vertical column.
After the fats and oils are vacuum deaerated to
prevent darkening, they are charged to the bot-
tom of the column through a sparge ring. De-
aerated, demineralized water is charged to the
top of the tower. High-pressure live steam is
injected into the reaction zone, approximately
the midpoint of the tower, where the fats and
oils are split into fatty acids and glycerine. A
pressure of 600 to 700 psi and a temperature
of approximately 485 °F are required. The
oil droplets, entering the bottom of the tower
through the sparge ring, rise up through the
water -glycerine solution because of lower
density. The water-glycerine is drawn off
at the bottom of the tower. As the oil droplets
rise, the glycerides split, and the freed fatty acids
are separated from, the glycerine, which dis-
solves in the incoming water. The fatty acids
pass overhead to a decanter where any en-
trained water is removed.
The mixed fatty acids obtained by these pro-
cesses can be used directly in soap manufacture
or they can be separated into more refined
fractions (palmitic acid, stearic acid, oleic acid,
and others). Steam distillation, pressing, or
fractional-crystallization methods are practical.
Soop Manufacture
The kettle or full-boiled batch process is the
most widely used and oldest method of soap
manufacture. First, the molten fats are boiled
in a caustic solution by using live steam. After
saponification of the fats and oils to soap and
glycerine is essentially completed, salt is add-
ed to separate the soap from the aqueous phase.
The soap is settled by gravity, and the glycerine
phase is drawn off. After the batch is diluted
with water, the mixture is reboiled. The aque-
ous layer is again drawn off. Strong caustic
can then be added if required. The mixture
is again boiled and settled, and the aqueous
layer is removed. This aqueous layer contains
excess caustic and can be used in later batches.
The soap mixture is again diluted with water,
reboiled, and allowed to settle. Three layers
settle out. The upper layer is the crude prod-
uct sometimes called neat soap. The middle
layer, called nigre, is dark colored and strong-
ly alkaline. The bottom layer is primarily
caustic though it may contain some soap. The
nigre can be used in cheaper, darker grades
of soap or used in the next batch.
The Sharpies process, an adaptation of the full-
boiled process, makes soap continuously. Cen-
trifuges are used to separate the soap from the
aqueous phases. The total process time is re-
duced from several days to a few hours. Four
stages corresponding to the four stages of the
kettle process are used. The flow of the caustic
is counter cur rent to the flow of fats and oils.
The semi-boiled process involves the boiling
of fats and oils with the theoretically required
caustic. No washing of glycerine is done. The
cold-make process is similar to the semi-boiled
operation, except that the components are mixed
in a crutcher (a type of mixer) and run into
frames, where saponification is completed over
several days at room temperatures up to 110°F.
Any free fatty acids must then be neutralized
by caustic soda or soda ash. The amount of
caustic added must be carefully controlled to
limit free-caustic concentration in the final
product to 0. 02 to 0. 10 percent.
Soap Finishing
Soap is finished in many forms—bars, flakes,
granules, liquids, or powder.
Neat soap contains approximately 30 percent
moisture and must be dried in hot-air driers,
spray driers, or steam-heated tubes. The hot-
air process is used to make bar soap, soap
chips, and soap flakes. Liquid soap is charged
through a chilled roller. The thin film formed
is stripped off in ribbons by a serrated scraper
knife. The ribbons fall to a wire screen con-
veyor that carries them through the drier in
several passes. The ribbons are then air
cooled and carried to storage by screw con-
veyors, which break the soap ribbons into chips.
When bar or flake form is desired, the soap
chips are charged to mixers. Here perfumes,
dyes, pigments, and preservatives are added.
Next, the material is rolled in granite- or steel-
roll mills. The milled chips are made into bars
by extrusion through a die. As it emerges in a
continuous bar, the soap is cut or stamped into
cakes. If soap flakes are desired, the chips
are taken from the finishing roll as a very thin
film and cut into flakes.
Spray drying is a widely used method of finish-
ing soap. After alkalinity builders are mixed with
the neat soap, special high-pressure pumps force
-------
718
CHEMICAL PROCESSING EQUIPMENT
it through high spray towers. Hot air, 500°F, is
blown countercurrently or concurrently to the fall-
ing droplets. The sprayed soap dries into a pow-
der.
SYNTHETIC DETERGENTS
The surf ace-active agents most commonly known
as detergents can be grouped into five main
chemical classes, as shown in Table 208. The
classification is based upon the ionic properties.
Anionic detergents ionize in water to give a nega-
tively charged organic ion and can be subclassi-
fied as sulfated fatty alcohols, alkyl-aryl sul-
fonates, and miscellaneous sulfates and sulfonates
Another class is the cationic synthetics,which
yield positively charged ions in the presence of
water. Detergents that do not ionize comprise
the final group.
Detergents contain from 20 to 40 percent active
agent. The remainder includes builders, fillers,
dye, and other compounds. The most common
builders are again sodium tripolyphosphate and
tetrasodium pyrophosphate. Sodium carboxy-
methyl cellulose can be added to obtain better
dirt suspension properties. Sodium silicate is
added to counter corrosion in aluminum wash-
ing machines.
The largest selling type of detergent base, alkyl-
aryl sulfonate, is prepared from an aromatic
hydrocarbon and polymerized straight-chain
hydrocarbon. Necessary water solubility is
obtained by sulfonating the water-insoluble hy-
drocarbon product, usually an alkylated benzene.
A flow diagram of this process is shown in Fig-
ure 525, It involves reacting dodecene and
benzene in the presence of an aluminum chloride
catalyst. The desired boiling range of dodecyl-
benzene is obtained by fractionation. This prod-
uct is sulfonated and neutralized with caustic.
Builders are added to the slurry, and the prod-
uct is dried. A flake finish is obtained by drum
drying, and a bead-shaped product is obtained
by spray drying.
Biodegradable linear alkylbenzene sulfonate can
be prepared from kerosine and benzene. The
kerosine is chlorinated in the presence of an
iodine catalyst. The mono chlorinated kerosine
is then reacted with benzene by using aluminum
chloride catalyst. Fractionation, sulfonation, and
neutralization cycles then complete the process.
These softer linear detergents have less foaming
tendencies than do the hard synthetic detergents.
Detergents can also be prepared by a hydrogena-
tion process from base stocks derived from
coconut oil and alcohol (Niven, 1955). The reac-
tion rate is controlled to maintain the unreacted
ester in slight excess. The next step is hydroly-
sis of the reacted product. The top layer, con-
taining product and reducing alcohol, is separated
by distillation. The product is then sulfonated
with concentrated sulfuric acid or liquid 803.
The sulfate produced is then converted to a salt
with alkali. Excess sodium sulfate can be re-
moved or left as a filler. The product is then
dried.
THE AIR POLLUTION PROBLEM
Soops
The principal air pollution problem in the prep-
aration of soap is odors. The extent of the odor
problem depends upon the type of charge stock.
Low-grade stocks obtained from rendered grease
and fats tend to be more odoriferous. In an at-
Table 208. SYNTHETIC DETERGENTS
Classification
lonization
with water
Trade names
Main uses
Sulfated fatty alcohols
Alkyl-aryl sulfonates
Miscellaneous sulfates and sulfonates
Cationic agents
Nonionic agents
Negative
Negative
Negative
Positive
Do not ionize
Orvus, Dreft,
Duponal, etc
Oronite, Ultrawet,
Santomerse, Tide,
Fab, Surf, Cheer,
etc
Merpols, MP-189,
Arctic Snytex, Vel,
Igepon A, Igepon T,
etc
Spans, Tweens,
Glim, Triton, etc
Strong soap competitors
Low-price cleaners for home
and industry
Cleaners with hard water, ex-
cellent soil removers
Germicidals
Soil removers and grease emul-
sifiers (low foam)
-------
Soaps and Synthetic Detergents
719
ALKYLATION
REACTOR
BENZENE
FRACTfONATOR
INTERMEDIATE
FRACTIONATOR
DODECYLBENZENE
FRACTIONATOR
DODECYLBENZENE SULFONATOR SPENT ACID
STORAGE SETTLER
in
SU L FU R ) C AC I 0
N«OH SOLUT)ON
NEUTRALIZER CRUTCHER
•*BU i L oe R s "
SPENT Ac I D
SODIUM
DO 0 E CYL BENZENE
SUL FON I C ACID
Figure 525. Flow sheet for alkyl-aryl sulfonate (Anon.,
1961. Copyrighted by Gulf Publishing Co., Houston,
Texas).
tempt to obtain a better smelling product, soap
manufacturers sometimes employ a deodorizer.
The undesirable components are removed by
means of live steam or a vacuum.
Blending, mixing, and packaging the finished
soap can cause local dust problems.
and hydrocaroon vapors can be released dur-
ing this period. Air is normally agitated
when neutralization is necessary, and the vented
air can be a source of odors.
AIR POLLUTION CONTROL EQUIPMENT
Detergents
The air pollution problems encountered in pre-
paring the base stocks are similar to those as-
sociated with a petroleum refinery. Relief
valves, storage vessels, and pump seals can
allow volatile hydrocarbons to escape to the at-
mosphere. Some of the fractionating equipment
is operated at atmospheric or under vacuum con-
ditions. The vents from accumulators or vacu-
um-producing equipment are a source of pollu-
tion. When kerosine is chlorinated, chlorine and
hydrogen chloride gas are vented from the chlo-
rinator.
Equipment to handle the spent aluminum chlo-
ride complexes must be provided. The com-
plex is first hydrolyzed and then the acid water
may have to be neutralized. The equipment
used in this operation can be a source of air
pollution. The hydrolysis phase requires agita-
tion and is exothermic. Hydrogen chloride
Soaps
Odors can be most successfully controlled by
incineration. Condensation can be employed
as an auxiliary to incineration. The effluent
from soap kettles, Twitchell tanks, and high-
pressure splitters contains large amounts of
steam. By condensing part or all of the steam,
the volume of effluent to be incinerated is great-
ly reduced. Part of the odorous material is
condensed or knocked down by the condensing
steam. Contact-type condensers should be
avoided unless the contaminated water can be
directly sewered. In any event the contaminated
water should not be cooled in a cooling tower or
spray pond.
The most economical method of incineration
consists of venting the noncondensable efflu-
ent into the firebox of a continuously operat-
ing boiler. Complete destruction of the odors
is achieved if the effluent is injected as part
-------
720
CHEMICAL PROCESSING EQUIPMENT
of the combustion air. A temperature of
1, 200°F should be maintained at the exit of
the combustion chamber.
Absoprtion of odors in scrubbing liquids has
been tried with varying success. One fatty
acid plant tried an alkaline scrubber on the
assumption that the odorous materials were
acid; benefits were negligible. In a soap
plant using the Sharpies continuous soap-
making process, the odorous substances
were identified as low-boiling amines (Molos,
1961). A scrubber using a sulfuric acid
solution at a pH of 2 was successful in re-
moving odors from the centrifuge room and
from the spray drier effluent. Apparently,
the success of odor removal by scrubbing
depends upon identifying the odorous sub-
stances and finding a scrubbing liquid that
reacts rapidly and completely with these
substances.
In the finishing operations, dust is the prin-
cipal air polluti.on problem. Of these opera-
tions, spray drying has the greatest air pollu-
tion potential. Granulating, screening, con-
veying, and mixing dry soap and other dry
ingredients create sufficient dust to require
an extensive dust collection and control sys-
tem.
Spray driers are usually controllable by scrub-
bers. A high-efficiency scrubber, such as a
venturi scrubber, is required to ensure the
collection of the very fine particles. Cen-
trifugal collectors are usually used ahead of
the scrubber in order to collect as much mate-
rial as possible in the dry form.
Reject soap bars and chips are sometimes
ground in a screen-type hammer mill. The
ground soap is used in industrial and laundry
blends. The mill discharge is screened, and
the oversize particles are recycled to the mill.
Since soap dust is irritating to the nasal pas-
sages, this operation requires very good dust
control. The mill and the screen must be kept
under suction, and any open transfer points
must be hooded. A baghouse is the usual col-
lector for this service. The collected dust can
be combined -with the product.
Detergents
Relief valves can be vented to a smokeless
flare. Atmospheric vents on condensers and
accumulators can be controlled by connecting
the vent to a vapor recovery system or by
using a heliflow water condenser. Vacuum
jets are vented to the firebox of heaters.
Barometric legs should also be sealed off and
the vapors incinerated. Centrifugal pumps
should be equipped with mechanical seals and
properly maintained.
The hydrolyzing and neutralizing equipment
should be vented to a properly designed water
scrubber.
GLASS MANUFACTURE
Glass has been made for over 3, 500 years, but
only in the last 75 years have engineering and
science been able to exploit its basic properties
of hardness, smoothness, and transparency so
that it can now be made into thousands of diverse
products.
The economics and techniques connected •with
mass production of glass articles have led to
the construction of glass-manufacturing plants
near or •within highly populated areas. Un-
fortunately, airborne contaminants generated
by these glass plants can contribute substantial-
ly to the air pollution problem of the surround-
ing community. Control of dust and fumes has,
therefore, been, and must continue to be, in-
herent to the progress of this expanding industry.
Air pollution control is necessary, not only to
eliminate nuisances, but also to bring substan-
tial savings by extending the service life of
the equipment and by reducing operating ex-
penses and down time for repair. Reduction
in plant source emissions can be accomplished
by several methods, including control of raw
materials, batch formulation, efficient com-
bustion of fuel, proper design of glass-melt-
ing furnaces, and the installation of control
equipment.
TYPES OF GLASS
Nearly all glass produced commercially is one
of five basic and broad types: Soda-lime, lead,
fused silica, borosilicate, and 96 percent silica.
Of these, modern soda-lime glass is well suited
for melting and shaping into window glass, plate
glass, containers, inexpensive tableware, elec-
tric light bulbs, and many other inexpensive,
mass-produced articles. It presently consti-
tutes 90 percent of the total production of com-
mercial glass (Kirk and Othmer, 1947).
Typical compositions of soda-lime glass and
the four other major types of commercial glass
are shown on Table 209. Major ingredients of
soda-lime glass are sand, limestone, soda ash,
and cullet. Minor ingredients include salt cake,
aluminum oxide, barium oxide, and boron oxide.
Minor ingredients may be included as impuri-
ties in one or more of the major raw ingredients.
-------
Glass Manufacture
721
Table 209. COMPOSITIONS OF COMMERCIAL GLASSES (Kirk and Othmer, 1947)
Composition,
Component
Si02
Na2O
K2°
cao
PbO
B2°3
A12O3
MO
o
Soda-lime
70 to 75 (72)
12 to 18 (15)
0 to 1
5 to 14 (9)
-
-
0.5 to 2.5 (1)
0 to 4 (3)
Lead
53 to 68 (68)
5 to 10 (10)
1 to 10 (6)
0 to 6 (1)
15 to 40 (15)
-
0 to 2
—
Borosilicate
73 to 82 (80)
3 to 10 (4)
0.4 to 1
0 to 1
0 to 10
5 to 20 (14)
2 to 3 (2)
""
96% silica
96
-
-
-
-
3
-
-
Silica glass
99.8
-
-
-
-
-
-
-
The figures in parentheses give the approximate composition of a typical member.
Soda-lime glasses are colored by adding a
small percentage of oxides of nickel, iron,
manganese, copper, and cobalt, and elemen-
tal carbon as solutions or colloidal particles
(Tooley, 1953).
Although glass production results in tens of
thousands of different articles, it can be divid-
ed into the following general types (Kirk and
Othmer, 1947):
Flat glass 25
Containers 50
Tableware 8
Miscellaneous instruments, scientific
equipment, and others 17
GLASS-MANUFACTURING PROCESS
Soda-lime glass is produced on a massive scale
in large, direct-fired, continuous melting fur-
naces. Other types of glass are melted in small
batch furnaces having capacities ranging from
only a few pounds to several tons per day. Air
pollution from the batch furnaces is minor, but
the production of soda-lime glass creates major
problems of air pollution control.
A complete process flow diagram for the con-
tinuous production of soda-lime glass is shown
on Figure 526. Silica sand, dry powders,
granular oxides, carbonates, cullet (broken
glass), and other raw materials are transferred
from railroad hopper cars and trucks to storage
bins. These materials are withdrawn from the
storage bins, batch weighed, and blended in a
mixer. The mixed batch is then conveyed to
the feeders attached to the side of the furnace.
Although dust emissions are created during
these operations, control can be accomplished
by totally enclosing the equipment, and install-
ing filter vents, exhaust systems, and bag-
houses.
Screw- or reciprocating-type feeders contin-
uously supply batch-blended materials to the
direct-fired, regenerative furnace. These dry
materials float upon the molten glass within
the furnace until they xnelt. Carbonates de-
compose releasing carbon dioxide in the form
of bubbles. Volatilized particulates, com-
posed mostly of alkali oxides and sulfates,
are captured by the flame and hot gases pass-
ing across the molten surface. The particu-
lates are either deposited in the checkers and
refractory-lined passages or expelled to the
atmosphere.
The mixture of materials is held around 2, 700°F
in a molten state until it acquires the homogeneous
character of glass. Then it is gradually cooled
to about 2, 200 °F to make it viscous enough to
form. In a matter of seconds, while at a yellow-
orange hot temperature, the glass is drawn
from the furnace and worked on forming machines
by a variety of methods including pressing, blow-
ing in molds, drawing, rolling, and casting.
One source of air pollution results from, the use
of hydrocarbon greases and oils to lubricate the
hot delivery systems and molds of glass-forming
machines. The smoking from these greases and
oils creates a significant source of air pollution
separate from furnace emissions.
Immediately after being shaped in the machines,
the glass articles are conveyed to continuous
annealing ovens, where they are heat treated
to remove strains that have developed during
the molding or shaping operations and then
subjected to slow, controlled cooling. Gas-
fired or electrically heated annealing ovens
are not emitters of air contaminants in any
-------
122
CHEMICAL PROCESSING EQUIPMENT
Materials dry, or nearly dry
Continuous tank furnace
down through top (crown)
Submerged throat in bndgewall
At about I . 472 - 2,01 2° p_
depending on article and process
LIMESTONE
or burnt lime
to yield lime, CaO
MgO also results
if raw material
contains MgC03
Approx 20-120 mesh
FELDSPAR
R2O.AI203.6Si02
to yield alumina, AIZ03
Also yields Si02,
and NaaO or KaO
Pulverized or granular
Borax or boric acid
to yield B203, and
other additions to
yield K20, MgO,
ZnO, BaO, and PbO
fining, oxidizing,
decolorizing, and
coloring agents
Melting
about 2 . 70o°V
-^ '
Refining^
fining and
homogenizing
Fabrication
Hot, viscous liquid glass
shaped by pressing,
blowing, pressing and
blowing, drawing, or rolling
Crushed cullet
of same composition
as that to be melted
Hot zone about 930 f
60-90 minutes in —
continuous belt tunnel (eh
shing J
Figure 526. Flow diagram for soda-lime glass manufacture (Kirk
and Othmer, 1947).
significant quantity. After leaving the anneal-
ing ovens, the glass articles are inspected and
packed or subjected to further finishing opera-
tions.
Glass-forming machines for mass production
of other articles such as rod, tube, and sheet
usually do not emit contaminants in significant
amounts.
HANDLING, MIXING, AND STORAGE SYSTEMS FOR
RAW MATERIALS
Material-handling systems for batch mixing
and conveying materials for making soda-
lime glass normally use commercial equip-
ment of standard design. This equipment is
usually housed in a structure separate from
the glass-melting furnace and is commonly
referred to as a "batch plant. " A flow dia-
gram of a typical batch plant is shown in Figure
527. In most batch plants, the storage bins are
located on top, and the weigh hoppers and mixers
are below them to make use of the gravity flow.
Major raw materials and cullet (broken scrap
glass) are conveyed from railroad hopper cars
or hopper trucks by a combination of screw
conveyors, belt conveyors, and bucket eleva-
tors, or by pneumatic conveyors (not shown in
Figure 527) to the elevated storage bins. Mino
ingredients are usually delivered to the plant
in paper bags or cardboard drums and trans-
ferred by hand to small bins.
Ingredients comprising a batch of glass are
dropped by gravity from the storage bins into
weigh hoppers and then released to fall into
the mixer. Cullet is ground and then mixed
with the dry ingredients in the mixer. Ground
cullet may also bypass the mixer and be mixed
instead with the other blended materials in the
bottom of a bucket elevator. A typical batch
charge for making soda-lime flint glass in a
mixer with a capacity of 55 cubic feet consists
of:
Ib
Silica sand
Cullet
Soda ash
2,300
650
690
-------
Glass Manufacture
723
Limestone
Niter
Salt cake
Arsenic
Decolorizer
570
7
12
2
1_
4,232
Raw materials are blended in the mixer for peri-
ods of 3 to 5 minutes and then conveyed to a
charge bin located alongside the melting furnace.
At the bottom of the charge bin, rotary valves
feed the blended materials into reciprocating- or
screw-type furnace feeders.
In a slightly different arrangement of equipment
to permit closer control of batch composition,
blended materials are discharged from the mixer
into batch cans that have a capacity of one mixer
load each. Loaded cans are then conveyed by
monorail to the furnace feeders. Trends in
batch plant design are toward single reinforced-
concrete structures in which outer walls and
partitions constitute the storage bins. Complete
automation is provided so that the batch plant is
under direct and instant control of the furnace
foreman.
THE AIR POLLUTION PROBLEM
The major raw materials for making soda-lime
glass — sand, soda-ash, and limestone—usually
contain particles averaging about 300 microns
in size. A small percentage of these particles,
however, is less than 50 microns but present
in sufficient quantities to cause dust emissions
during conveying, mixing, and storage opera-
tions. Moreover, minor raw materials such as
salt cake and sulfur can create dust emissions
during handling. Dust is the only air contaminant
from batch plants, and control of dust emissions
poses problems similar to those in industrial
plants handling similar dusty powder or granular
materials.
HOODING AND VENTILATION REQUIREMENTS
Dust control equipment can be installed on con-
veying systems that use open conveyor belts.
A considerable reduction in the size of the dust
control equipment can be realized by totally en-
closing all conveying equipment and sealing all
covers and access openings with gaskets of
polyurethane foam. In fact, by totally enclos-
ing all conveying equipment, exhaust systems
CULLET
RAW MATERIALS
RECEIVING
HOPPER
V
SCREW
CONVEYOR
STORAGE BINS
MAJOR RAW MATERIALS
FILTER
VENTS
MINOR
INGREDIENT
STORAGE
BINS
BELT CONVEYOR
BATCH
STORAGE
BIN
FURNACE
FEEDER
Figure 527. Process flow diagram of a batch plant.
-------
724
CHEMICAL PROCESSING EQUIPMENT
become unnecessary, and relatively small filter
vents or dust cabinets can be attached directly
to the conveying equipment and storage bins.
On the other hand, exhaust systems are re-
quired for ventilating the weigh hoppers and
mixers. For example, a 60-cubic-foot-capacity
mixer and a 4, 500-pound-capacity mixer each
require about 600 cfm ventilation air. Seals of
polyvinylchloride should be installed between
the rotating body of the mixer and its frame to
reduce ventilation to a minimum.
Railroad hopper cars and hopper bottom trucks
must be connected to sealed receiving hoppers
by fabric sleeves so that dust generated in the
hoppers during the loading operation is either
filtered through the sleeves or exhausted through
a baghouse.
Local exhaust systems for dust pickup are de-
signed by using the recommended practice of the
Committee on Industrial Ventilation (I960). For
example, the ventilation rate at the transfer
point between two open belt conveyors is 350
cfm per foot of belt width, with 200 fpm minimum
velocity through the hood openings.
AIR POLLUTION CONTROL EQUIPMENT
Because dust emissions contain particles only
a few microns in diameter, cyclones and cen-
trifugal scrubbers are not as effective as bag-
houses or filters are in collecting these small
particles; consequently, simple cloth filters and
baghouses are used almost exclusively in con-
trolling dust emissions from batch plants.
Filter socks or simple baghouses •with inter-
mittent shaking mechanisms are usually de-
signed for a filter velocity of 3 fpm, but bag-
houses with continuous cleaning devices such as
pulse jets or reverse air systems can be de-
signed for filter velocities as high as 10 fpm.
Filtration cloths are usually cotton, though ny-
lon, orlon, and dacron are sometimes used.
Dusts collected are generally noncorrosive.
Filters or baghouses for storage bins are de-
signed to accommodate not only displaced air
from the filling operation but also air induced
by falling materials. Filtration of air exhaust
from pneumatic conveyors used in filling the
bins must also be provided. Filters •with at least
a 1-square-foot area should be mounted on the
hand-filled minor-ingredient bins.
Transfer chutes of special design are used for
hand filling the minor ingredient bins. They are
first attached securely with gaskets to the top of
the bins. The bags are dropped into a chute
containing knives across the bottom. The knives
split the bag, and as the materials fall into the
bin, the broken bag seals off the escape of dust
from the top of the chute.
GLASS-MELTING FURNACES
While Limited quantities of special glasses such
as lead or borosilicate are melted in electrically
heated pots or in small-batch, regenerative fur-
naces with capacities up to 10 tons per day, the
bulk of production, soda-lime glass, is melted
in direct-fired, continuous, regenerative furnaces.
Many of these furnaces have added electric induc-
tion systems called "boosters" to increase capac-
ity. Continuous, regenerative furnaces usually
range in capacity from 50 to 300 tons of glass
per day; 100 tons is the most common capac-
ity found in the United States.
Continuous Soda-Lime Glass Furnaces
Continuous, regenerative, tank furnaces differ
in design according to the type of glass products
manufactured. All have two compartments. In
the first compartment, called the melter, the dry
ingredients are mixed in correct proportions and
are continuously fed onto a molten mass of glass
having a temperature near 2, 700°F. The dry mate-
rials rnelt after floating a third to one-half of the
way across the compartment and disappearing into
the surface of a clear, viscous-liquid glass. Glass
flows from the melter into the second compartment,
commonly referred to as the refiner, where it is
mixed for homogeneity and heat conditioned to
eliminate bubbles and stones. The temperature
is gradually lowered to about 2,200°F. The
amount of glass circulating within the melter
and refiner is about 10 times the amount with-
drawn for production (Sharp, 1954).
Regenerative furnaces for container and tableware
manufacture have a submerged opening or throat
separating the refiner from the melter. The throat
prevents undissolved materials and scum on the
surface from entering the refiner. Glass flows
from the semicircular refining compartment into
long, refractory-lined chambers called forehearths.
Oil or gas burners and ventilating dampers ac-
curately control the temperature and viscosity of
the glass that is fed from the end of the forehearth
to glass-forming machines.
Continuous furnaces for manufacturing rod, tube,
and sheet glass differ from furnaces for container
and tableware manufacture in that they have no
throat between the melter and refiner. The com-
partments are separated from each other by float-
ing refractory beams riding in a drop arch across
the entire width of the furnace. Glass flows from
-------
Glass Manufacture
725
the rectangular-shaped refiner directly into the
forming machines.
Regenerative firing systems for continuous glass
furnaces were first devised by Siemens in 1852,
and since then, nearly all continuous glass fur-
naces in the United States have used them. In
Europe, continuous glass furnaces employ both
recuperative and regenerative systems.
Regenerative firing systems consist of dual
chambers filled with brick checkerwork. While
the products of combustion from the melter pass
through and heat one chamber, combustion air
is preheated in the opposite chamber. The func-
tions of each chamber are interchanged during
the reverse flow of air and combustion products.
Reversals occur every 15 to 20 minutes as re-
quired for maximum conservation of heat.
Two basic configurations are used in designing
continuous, regenerative furnaces--the end port
shown in Figure 528, and the side port in Figure
529. In the side port furnace, combustion prod-
ucts and flames pass in one direction across
the melter during one-half of the cycle. The
flow is reversed during the other half cycle.
The side port design is commonly used in large
furnaces with melter areas in excess of 300
square feet (Tooley, 1953).
In the end port configuration, combustion products
and flames travel in a horizontal U-shaped path
across the surface of the glass within the melter.
Fuel and air mix and ignite at one port and dis-
charge through a second port adjacent to the first
on the same end wall of the furnace. While the
end port design has been used extensively in small-
REFINER SIDE WALL
GLASS SURFACE IN REFINER
FOREHEARTH
INDUCED DRAFT FAN
FEEDER
PARTING WALL '
SECONDARY CHECKERS'
CURTAIN WALL
RIDER ARCHES
Figure 528. Regenerative end port glass-melting furnace.
-------
726
CHEMICAL PROCESSING EQUIPMENT
er furnaces with melter areas from 50 to 300
square feet, it has also been used in furnaces with
melter areas up to 800 square feet.
Continuous furnaces are usually operated slightly
above atmospheric pressure within the melter to
prevent air induction at the feeders and an over-
all loss in combustion efficiency. Furnace draft
can be produced by several methods: Induced-
draft fans, natural-draft stacks, and ejectors.
cooled liquid. It has nondirectional properties,
fracture characteristics of an amorphous solid,
and no freezing or melting point. To account for
the wide range of properties, glass is considered
to be a configuration of atoms: rather than an ag-
gregate of molecules. Zachariasen (1932) pro-
posed the theory that glass consists of an extended,
continuous, three-dimensional network of ions with
a certain amount of short-distance-ordered ar-
rangement similar to that of a polyhedral crystal.
The Air Pollution Problem
Particulates expelled from the melter are the re-
sult of complex physical and chemical reactions
that occur during the melting process.
Glass has properties akin to those of crystalline
solids, including rigidity, cold flow, and hard-
ness. At the same time, it behaves like a super-
These dissimilar properties explain in part -why
predictions of particulate losses from the melter
based solely upon known temperatures and vapor
pressures of pure compounds; have been inaccurate
Other phenomena affect the generation of par-
ticulates. During the melting process, carbon
dioxide bubbles and propels particulates from the
melting batch. Particulates are entrained by the
fast-moving stream of flames and combustion
REFINER SIDE
MELTER SIDE WALL THROAT^
MELTER BOTTOM ,
GLASS SURFACE IN REFINER
FOREHEARTH
GLASS SURFACE IN MELTER
NATURAL DRAFT STACK
BACK WALL
COMBUSTION AIR BLOWER
MOVABLE REFRACTORY BAFFLE
RIDER ARCHES
BURNER
Figure 529. Regenerative side port glass-melting furnace.
-------
Glass Manufacture
727
gases. As consumption of fuel and refractory tem-
peratures of the furnace increase with glass ton-
nage, particulates also increase in quantity. Par-
ticulates, swept from the melter, are either col-
lected 5ii the checkerwork and gas passages or
exhausted to the atmosphere.
Source test data
In a recent study, many source tests of glass
furnaces in Los Angeles County were used for
determining the major variables influencing stack
emissions. As summarized in Table 210, data
include: Particulate emissions, opacities, pro-
cess variables, and furnace design factors.
Particle size distribution of two typical stack
samples is shown in Table 211. These particu-
late samples were obtained from the catch of a
pilot baghouse venting part of the effluent from
a large soda-lime container furnace.
Chemical composition of the particulates was
determined by microquantitative methods or by
spectrographic analysis. Five separate samples,
four from a pilot baghouse, and one from the
stack of a soda-lime regenerative furnace, are
given in Table 212. They were found to be com-
posed mostly of alkali sulfates although alkalies
are reported as oxides. The chemical composi-
tion of sample 5 was also checked by X-ray
crystallography. In this analysis, the only
crystalline material present in identifiable
amounts was two polymorphic forms of sodium
sulfate.
Opacity of stack emissions
From the source test data available, particu-
late emissions did not correlate with the opacity
of the stack emissions. Some generalizations on
opacity can, however, be made. Opacities usu-
Table 210. SOURCE TEST DATA FOR GLASS-MELTING FURNACES
Test No.
C-339b
C-339
C-382-1
C-382-2
C-536
C-383
Pri Lab
Pri Lab
Pri Lab
Pri Lab
Pri Lab
Pri Lab
Pri Lab
C-101
C-120
C-577
C-278-1
C-278-2
C-653
C-244-1
C-244-2
C-420-1
C-420-2
C-743
C-471
Type
of
furnacea
EP
EP
EP
EP
EP
EP
EP
EP
EP
EP
EP
EP
EP
EP
EP
SP
SP
SP
SP
SP
SP
SP
SP
SP
SP
Type
of
fuelb
0-300
G
G
G
0-200
G
G
G
G
G
G
0-300
G
G
G
0-300
G
0-300
G
G
G
G
G
G
G
xl
(particulate
emissions),
Ib/hr
7.00
3.00
4. 60
6.40
4. 70
8.40
3.86
4.76
4.26
6.84
4. 62
3.96
7. 16
9. 54
9. 90
12. 70
3. 97
8. 44
8. 90
6. 30
3.00
6.30
6.60
10.20
6. 70
X2 *
process wt ratio),
lb/hr-ft2
of melter area
16.7
13.8
16.5
18.2
17. 5
17.9
10. 9
14.6
17. 1
17.4
18. 5
14.6
20.2
15.2
14.2
24.2
18.3
18.5
22.0
7.5
5.4
10.7
13.2
26.2
11.6
X3-
wt fraction
of cullet in
charge0
0.300
0.300
0. 300
0. 300
0. 199
0. 300
0. 094
0. 094
0. 157
0. 094
0.365
0.269
0. 175
0.300
0. 320
0. 134
0. 361
0. 360
0. 131
0. 182
0. 100
0. 100
0. 100
0.047
0.276
x4
(checker volume),
ft3/ftZ of melter
5. 40
5. 40
5. 40
5. 40
5. 40
6.50
8.00
8.00
8.00
8. 00
9. 00
9. 00
9.00
5.00
5. 00
6. 90
6. 93
6.93
8. 74
7. 60
7. 60
7. 60
7. 60
8.25
5.60
Maximum
opacity
of stack
emissions, %
50
10
10
10
10
20
25
25
25
25
-_
45
20
20
20
35
20
20
40
25
25
10
5
25
30
aEP = end port, regenerative furnace; SP = side port, regenerative furnace.
bG = natural gas; 0-200 = U.S. Grade 3 fuel oil; 0-300 = U.S. Grade 5 fuel oil.
cConstants: Sulfate content of charge 0. 18 to 0. 34 wt %.
Fines (-325 mesh) content of charge 0. 2 to 0. 3 wt %.
-------
728
CHEMICAL PROCESSING EQUIPMENT
ally increase as particulate emissions increase.
More often than not, furnaces burning U. S. Grade
5 fuel oil have plumes exceeding 40 percent white
opacity while operating at a maximum pull rate,
which is the glass industry's common term for
production rate. Plumes from these same fur-
naces were only 15 to 30 percent white opacity
while burning natural gas or U. S. Grade 3
(P. S. 200) fuel oil. Somewhat lower opacities
may be expected from furnaces with ejector
draft systems as compared with furnaces with
natural-draft stacks or induced-draft fans.
Hooding and Ventilation Requirements
In order to determine the correct size of air
pollution control equipment, the volume of dirty
exhaust gas from a furnace must be known. Some
of the more important factors affecting exhaust
volumes include: Furnace size, pull rate, com-
bustion efficiency, checker volume, and furnace
condition.
Exhaust volumes can be determined from fuel
requirements for container furnaces given by
the formula of Cressey and Lyle (1956).
F = [50 + 0. 6A] + 4. 8T
where
F = total heat, 106 Btu/day
(137)
A := melter area, ftz
T =: pull rate, tons/day.
This straight-line formula includes minimum
heat to sustain an idle condition plus additional
Table 211. SIZE DISTRIBUTION OF
PARTICULATE EMISSIONS
(MICROMEROGRAPH ANALYSES)
Furnace 1 Flint glass
Diameter (D),
H
36. 60
22.00
18. 30
16. 50
14.60
12.80
12.20
11. 60
11. 00
10.40
9.80
9. 20
8. 50
7. 30
6. 10
4.88
3. 66
3. 05
2, 44
1.83
1. 52
1.22
% (by wt)
less than D
100
99. 5
98.6
97. 7
94.0
84. 6
80. 7
76.6
72. 7
67. 7
62.4
58. 3
51.8
43. 1
34. 4
28. 0
21. 3
18. 6
14. 9
11. 0
8. 3
4. 1
Furnace 2 Amber glass
Diameter (D),
V
17.40
15. 70
14. 00
12.20
11.60
11. 00
10.50
9.90
9. 30
8. 80
8. 10
7. 00
5.80
4.65
3.49
2.91
2.33
1.74
1.45
1. 16
% (by wt)
less than D
100
99.8
99. 4
96.8
92.5
89.5
87.2
83.4
78. 7
75.0
73. 4
60. 3
47. 6
35.6
25.4
20. 5
16.4
10. 9
8.9
5.3
Table 212. CHEMICAL COMPOSITION OF PARTICULATE EMISSIONS
(QUANTITATIVE ANALYSES), METALLIC IONS REPORTED AS OXIDES
Sample source
Test
type of glass
components
Silica (Si02)
Calcium oxide (C_O)
Bi
Sulfuric anhydride (SO,)
Boric anhydride (B2Oj)
Arsenic oxide (As2Oj)
Chloride (Cl)
Lead oxide (PbO)
K2O + Na2O
AL2O3
Fluoride
Fe203
MgO
ZnO
Unknown metallic oxide (R2O^)
Loss on ignition
Baghouse
catch
No. 1
amber,
wt %
0. 03
1. 70
46. 92
3.67
7. 71
0. 01
0. 39
29.47
10. 10
Baghouse
catch
No. 2
flint,
wt %
0. 3
2. 3
25. 1
1. 3
28. 1
3.5
8.6
30.8
Baghouse
catch
No. 3
amber,
wt %
0. 1
0. 8
46. 7
26. 1
0. 1
0.5
25. 7
Baghouse
catch
No. 4
flint,
wt %
4. 1
19.2
30.5
36.5
0.2
0.6
1.4
7. 5
Millipore
filter
No. 5 '
flint,
wt %
3.3
39.4
39.2
6.5
11. 6
-------
Glass Manufacture
729
heat for a specified pull rate. Fuel require-
ments for bridgewall-type, regenerative fur-
naces are also given by Sharp (1955) and are
shown in Figure 530. The melter rating para-
meter of 4 square feet of melter surface area
per daily ton of glass should be used to estimate
the fuel requirements of container furnaces at
maximum pull rates, but 8 square feet per ton
can be used for. estimating fuel requirements
for non-bridgewall furnaces supplying glass
for tableware and for sheet, rod, and tube
manufacture. Fuel requirements given are
averages for furnaces constructed before 1955;
consequently, these furnaces generally require
more fuel per ton of glass than do furnaces con-
structed since 1955. After the fuel require-
ments are determined, exhaust volumes are com-
puted on the basis of combustion with 40 percent
12 000.
10 000
200 300
400 500
MELTER AREA,
Figure 530. Natural gas for bridgewall-type
regenerative furnaces (Sharp, 1955).
excess combustion air. Forty percent excess
combustion air is chosen as representing av-
erage combustion conditions near the end of the
campaign (a total period of operation without shut-
ting down for repairs to the furnace).
Exhaust volumes determined from fuel require-
ments are for furnaces with induced-draft sys-
tems or natural-draft stacks. Exhaust volumes
for ejector systems can be estimated by increas-
ing the exhaust volume by 30 to 40 percent to
account for ejector air mixed with the furnace
effluent.
Exhaust gases from furnaces with natural-draft
stacks or induced-draft fan systems usually range
in temperature from 600" to 850°F, but exhaust
gas temperatures -from furnaces containing ejec-
tors are lower and vary from 400° to 600°F.
In Table 213 are found chemical analyses of gas-
eous components of exhaust gases from large,
regenerating, gas-fired furnaces melting three
kinds of soda-lime glass.
Air Pollution Control Methods
As the furnace campaign progresses, dust carry-
over speeds destruction of the checkers. Upper
courses of the firebrick checker glaze when sub-
jected to high temperatures. Dust and condensate
collect on the brick surface and form slag that
drips downward into the lower courses -where it
solidifies at the lower temperature and plugs the
checkers. Slag may also act somewhat like fly-
paper, tenaciously clinging to the upper courses
and eventually sealing off upper gas passages.
Hot spots develop around clogged checkers and
intensify the destructive forces, which are re-
flected by a drop in regenerator efficiency and a
rise in fuel consumption and horsepower required
to overcome additional gas flow resistance througl
the checkers. Checker damage can finally reach
Table 213. CHEMICAL COMPOSITION OF GASEOUS EMISSIONS
FROM GAS-FIRED, REGENERATIVE FURNACES
Gaseous components
Nitrogen, vol %
Oxygen, vol %
Water vapor, vol %
Carbon dioxide, vol %
Carbon monoxide, vol %
Sulfur dioxide (802), ppm
Sulfur trioxide (SO^), ppm
Nitrogen oxides (NO,NO2), ppm
Organic acids, ppm
Aldehydes, ppm
Flint glass
71.9
9.3
12.4
6.4
0
0
0
724
NAa
NA
Amber glass
81. 8
10.2
7. 7
8. 0
0.007
61
12
137
50
7
Georgia green
72.5
8.0
12. 1
7.4
0
14
15
NA
NA
NA
= not available.
-------
730
CHEMICAL PROCESSING EQUIPMENT
a point where operation is no longer economical or
is physically impossible because of collapse. Thus,
successful operation of modern regenerative fur-
naces requires keeping dust carryover from the
melter to an absolute minimum, 'which also coin-
cides with air pollution control objectives by prevent-
ing air contaminants from entering the atmosphere.
Aside from reducing air contaminants, benefits de-
rived from reducing dust carryover are many and
include longer furnace campaigns, lower mainte-
nance costs, and savings on fuel.
In order to determine which design and operating
variables have the greatest effect upon dust carry-
over and particulate emissions, statistical analysis
was performed on source test data previously men-
tioned in Table 210.
By the method of Brandon (1959), particulate emis-
sions, the dependent variable was found to corre-
late with the following independent variables and
nonquantitative factors:
1. Process weight, Ib/hr-ft ;
2. cullet, wt % of charge;
•j ^
3. checker volume, ft /ft melter;
4. type of furnace, side port or end port;
5. type of fuel, U.S. Grade 5 (PS300) oil or
natural gas;
6. melter area, ft .
Several simplifying assumptions are made so that
furnaces of different sizes can be compared. Pro-
cess weight per square foot of melter describes a
unit process occurring in each furnace regardless
of size. Cubic feet of checkers per square foot
of melter not only defines the unit's dust-collect-
ing capability but is also a measure of fuel economy.
Source tests C-382 and C-536 in Table 210, and
other source tests show no appreciable difference
in particulate emissions from burning natural gas
or U. S. Grade 3 fuel oil.
Correlation of particulate emissions with weight
percent sulfate (SOo) and minus 325-mesh fines
in the charge was not possible because of insuffi-
cient test data. Limited data available indicate
that particulate emissions may double when total
sulfate (SOg) content of the batch charge is in-
creased from 0. 3 to 1.0 weight percent. Total
sulfates (SOj) include equivalent amounts of ele-
mental sulfur and all compounds containing sulfur.
Sulfates usually comprise over 50 percent of the
particulate emissions. They act as fluxing agents
preventing the melting dry-batch charge from
forming a crust that interferes with heat trans-
fer and melting (Tooley, 1953). Compounds of
arsenic, boron, fluorine, and metallic selenium
are also expected to be found along -with sodium
sulfate in the particulate emissions because of
their high vapor pressures.
Data roughly indicate that particulate emissions
increase severalfold when the quantity of minus
325-mesh fines increases from 0. 3 weight per-
cent to 1 or 2 weight percent.
Statistical analysis using the method of curvilinear
multiple correlation by Ezekiel (1941) results in the
following equation, -which describes particulate
emissions, the dependent variable, as a function
of four independent variables and two nonquanti-
tative independent factors. This equation is
valid only when two other independent variables--
sulfate content and content of minus 325-mesh fines
of the batch—lie between 0. 1 to 0. 3 weight percent
and, also, when fluorine, boron, and lead com-
pounds are either absent from the batch charge or
present only in trace amounts.
X = a + 0. 0226(X ) - 0. 329 X - 4. 412 X,
•*• £ £ J
0. 9379 X - 0. 635 (X^) + 6. 170 X
(138)
where
X = particulate emissions, Ib/hr
X = process wt, Ib/hr-ft melter
X
= wt fraction of cullet in charge
= checker volume, ft /ft melter
X = melter area, ft /1 00
a = constant involving two nonquantitative
independent factors relating the type
of furnace (side port or end port) and
the type of fuel (U. S. Grade 5 fuel or
natural gas) .
a = -0.493 end port - U.S. Grade 5
fuel oil
a = -0.623 side port - U.S. Grade 5
fuel oil
3. = -1.286 end port - natural gas
a = -1.416 side port - natural gas.
Particulate emissions computed by this equation
for 25 source tests show a standard deviation
from measured particulate emissions of +_ 1.4
pounds per hour. Further statistical refine-
ment failed to yield a lower standard deviation.
-------
Glass Manufacture
731
Emissions to the atmosphere can be predicted
by using equation 138 or Figures 531 through
534, which are based upon this equation. The
curves should be used only within the limits in-
dicated for the variables. The curves should
not be extrapolated in either direction with the
expectation of any degree of accuracy, even
though they appear as straight lines. Particu-
late emissions are first determined from Fig-
ure 531, then positive or negative corrections
obtained from Figures 53Z through 534 are add-
ed to the emissions obtained from Figure 531.
Design and operation of soda-lime, continuous,
regenerative furnaces to alleviate dust carry-
over and minimize particulate emissions are
discussed in succeeding paragraphs. Advantages
of all-electric, continuous furnaces for melting
glass are also cited.
Control of raw materials
Although glassmakers have traditionally sought
fine-particle materials for easier melting, these
materials have intensified dust carryover in re-
X4, ft3 checkers/ft2 melter
Figure 531. Particulate emissions versus checker
volume per ft2 of melter.
X2 PROCESS HEIGHT, Ib/hr per ft' melter
Figure 532. Correction to particulate emissions
for process weight per ft2 melter.
-2
0 05
010 0 15 0 20
X,. HEIGHT FRACTION OF COLLET IN CHARGE
0 25
0 30
0 35
Figure 533. Correction to particulate emissions for cullet content
of the batch charge.
-------
732
CHEMICAL PROCESSING EQUIPMENT
500
. ft' meltsr
Figure 534. Correction to particulate emissions
for melter area.
generative furnaces. A compromise must be
reached. Major raw materials should be in the
form of small particles, many of them passing
U. S. 30-mesh screen, but not more than 0. 3
weight percent passing U.S. 325-mesh screen.
Because crystals of soda ash, limestone, and
other materials may be friable and crush in the
mixer, producing excessive amounts of fines,
screen analyses of individual raw materials
should not be combined for estimating the screen
analyses of the batch charge. Crystalline shape
and density of raw materials should be thoroughly
investigated before raw material suppliers are
selected.
Since particulate emissions from soda-lime re-
generative furnaces increase -with an increase in
equivalent sulfate (SOj) present in the batch
charge, sulfate content should be reduced to an
absolute minimum consistent -with good glass-
making. Preferably, it should be below 0. 3
weight percent. Equivalent sulfate (SO,) content
of the batch includes all sulfur compounds and
elemental sulfur. Compounds of fluorine, boron,
lead, and arsenic are also known to promote dust
carryover (Tooley, 1953), but the magnitude of
their effect upon emissions is still unknown. In
soda-lime glass manufacture, these materials
should be eliminated or should be present in only
trace amounts.
From the standpoint of suppressing stack emis-
sions, cullet content of the batch charge should
be kept as high as possible. Plant economics may,
nevertheless, dictate reduction in cullet where fuel
or cullet is high in cost or where cullet is in short
supply. Some manufacturing plants are able to
supply all their cullet requirements from scrap and
reject glass-ware.
Batch preparation
There are a number of ways to condition a batch
charge and reduce dust carryover. Some soda-
lime glass manufacturers add moisture to the
dry batch, but the relative merits of this process
are debatable. Moisture is sprayed into the dry-
batch charge at the mixer as a solution containing
1 gallon of surface-active -wetting agent to 750
gallons of water. Surface tension of the water is
reduced by the wetting agent so that the water
wets the finest particles and is evenly distributed
throughout the batch (Wilson, I960). Fluxing
materials such as salt cake appear more effec-
tive, since the unmelted batch does not usually
travel so far in the melter tank before it melts.
Moisture content of the batch is normally in-
creased to about 2 percent by weight. If the
moisture content exceeds 3 percent, batch in-
gredients adhere to materials-handling equip-
ment and may cake in storage bins or batch cans.
Other batch preparation'methods have been em-
ployed on a limited-production or experimental
basis to reduce dust carryover from soda-ash
glass manufacture. One method involves pre-
sintering the batch to form cullet and then charg-
ing only this cullet to the furnace. Advantages
claimed are faster melting, better batch con-
trol, less seed formation, reduced clogging in
the checkers, and lower stack losses (Arrandale,
1962). A Dutch oven doghouse cover also reduces
dust carryover by sintering the top of the floating
dry batch before it enters the melter. This meth-
od is probably not as efficient as is complete pre-
sintering in reducing dust carryover.
Other methods include: (1) Charging briquets,
-which are made from regular batch ingredients
by adding up to 10 percent by weight of water;
(2) charging wet batches containing 6 percent
moisture, which are made by first dissolving
soda-ash to form a saturated solution and mix-
ing this solution with sand and the other dry
materials; (3) charging the dry batch (Submerged)
in the melter; (4) enclosing batch feeders (Fabri-
anio, 1961); and (5) installing batch feeders on
opposite sides of end port, regenerative furnaces
and charging alternately on the side under fire.
Checkers
The design concept of modern regenerative fur-
naces, with its emphasis on maximum use of
fuel, is also indirectly committed to reducing
dust carryover. All things being equal, less
fuel burned per ton of glass means less dust
entrainment by hot combustion gases and flames
flowing across the surface of the melting glass.
Although container furnaces constructed over
15 years ago required over 7,000 cubic feet of
natural gas per ton of glass at maximum pull
-------
Glass Manufacture
733
rates, container furnaces built today can melt
a ton of glass -with less than 5, 000 cubic feet
of natural gas.
While several design changes are responsible
for this improvement, one of the most important
is the increase in checker volume. The ratio
of checker volume, cubic feet per square foot of
melter, has been rising during the years from
about 5 to 9 today. Enlarged checkers not only
reduce fuel consumption and particulate forma-
tion but also present a more effective trap for
dust particles that are expelled from the melter.
Source tests conducted by a large glass-manu-
facturing company indicated that over 50 percent
of the dust carryover from the melter is collected
by the checkers and gas passages instead of en-
tering the atmosphere.
Of course, the economics connected with regen-
erative furnace operation dictates the checker
volume. The law of diminishing returns oper-
ates where capital outlay for an added volume
of checkers will no longer be paid within a spe-
cified period by an incremental reduction in fuel
costs. Checkers have been designed in double-
pass arrangements to recover as much as 55 per-
cent of the heat from the waste gases (Sharp, 1954).
Although dust collects within checkers by mechan-
isms of impingement and settling, the relationship
among various factors influencing dust collection
is unknown. These factors include: Gas velocity,
brick size, flue spacing, brick setting, and brick
composition. Checkers designed for maximum
fuel economy may not necessarily have the high-
est collection efficiency. Further testing -will
be necessary in order to evaluate checker de-
signs. Checkers designed for maximum heat
exchange contain maximum heat transfer surface
per unit volume, a condition met only by smaller
refractories with tighter spacing. Heat transfer
surfaces can be computed by the method given in
Trinks (1955). Since gas velocities are also
highest for maximum heat transfer, less dust
collects by simple settling than by impingement.
Dust collection is further complicated in that
smaller brick increases the potential for clogging.
To prevent clogging in the checkers and ensure
a reasonable level of heat transfer, checkers
should be cleaned once per month or more often;
an adequate number of access doors should be
provided for this purpose (Spain, 1956b). Com-
pressed air, water, or steam may be used to
flush fine particles from the checkers. Virtual-
ly nothing can be done to remove slag after it
has formed. Checkers can be arranged in a
double vertical pass to reduce overall furnace
height and make cleaning easier. Access doors
should also be provided for removing dust de-
posits from the flues.
Preheaters
Further reductions in fuel consumption to re-
duce dust emissions may be realized by install-
ing rotary, regenerative air preheaters in series
with the checkers. Additional benefits include
less checker plugging, reduced maintenance, and
increased checker life. Rotating elements of
the preheater are constructed of mild steel, low-
alloy steel, or ceramic materials. Preheaters
raise the temperature of the air to over 1, 000°F,
and the increased velocity of this preheated air
aids in purging dust deposits that block gas pas-
sages of the checkers. Exhaust gases passing
through the opposite side of the preheater are
cooled below 800°F before being exhausted to
the atmosphere. A heat balance study of a plate
glass, regenerative furnace shows a 9 percent
increase in heat use by the installation of a
rotary, regenerative air preheater (Waitkus,
1962). To maintain heat transfer and prevent
re-entrainment, dust deposits on the preheater
elements must be removed by periodic cleaning.
Ductwork and valves should be installed for by-
passing rotary air preheaters during the cleaning
stage.
Refractories and insulation
Slagging of the upper courses of checkerwork
can be alleviated in most cases by installing
basic (high alumina content) brick in place of
superduty firebrick (Robertson et al. , 1957).
Basic brick courses extend from the topdown-
ward to positions where checker temperatures
arebelow 1,500°F. At this temperature, fire-
brickno longer "wets" and forms slagwithdust
particles. Dust usually collects in the lower
courses of firebrick in the form of fine particles
that are easily removed by cleaning. Although
basic brick costs 3 or 4 times as much as super-
duty firebrick, some glass manufacturers are
constructing entire checkerworks of basic brick
where slagging and clogging are most severe.
In some instances, basic refractories are re-
placing fireclay rider tiles and rider arches
in checker supports (Van Dreser, 1962). A
word of caution, basic brick is no panacea for
all ills of checkers. Chemical composition of
the dust should be known, to determine com-
patibility with the checkers (Fabrianio, 1961).
Regenerative furnaces can be designed to con-
sume less fuel and emit less dust by proper
selection and application of insulating refrac-
tories. A heat balance study of a side port, re-
generative furnace shows that, in the melting
process, glass receives 10 percent of heat
transfer from convection and 90 percent from
radiation. Of the radiation portion of heat
transferred, the crown accounts for 33 percent
-------
734
CHEMICAL PROCESSING EQUIPMENT
(Merritt, 1958). Since heat losses through the un-
insulated crown can run as high as 1 0 percent of
the total heat input, there is need for insulation
at this spot.
Most crowns are constructed of silica brick
with a maximum furnace capacity restricted to an
operating temperature of 2,850°F (Sharp, 1955).
Insulation usually consists of insulating silica
brick backed with high-duty plastic refractory.
Furnaces are first operated without insulation,
so that cracks can be observed. Then the cracks
are sealed with silica cement, and the insulation
is applied.
Insulation is needed on the melter sidewall and
at the port necks to prevent glassy buildup caused
by condensation of vapors. Condensate buildup
flows across port sills into the melter and can
become a major source of stones.
While insulation of sidewalls shows negligible
fuel reduction for flint glass manufacture, it
does show substantial fuel reduction for colored
glasses. The problem in manufacturing colored
glass is to maintain a high enough temperature
below the surface to speed the solution of stones
and prevent stagnation. Insulation on sidewalls
raises the mean temperature to a point where
stones dissolve and glass circulates freely.
Six inches or more of electrofusion cast block
laid over a clay bottom in a bed of mortar (Baque,
1954) not only saves fuel but is also less subject
to erosion than is fireclay block.
Insulation is seldom needed on the refining end
of the furnace since refiners have become cool-
ing chambers at today's high pull rates. Nose
crowns, however, are insulated to minimize con-
densation and drip (Bailey, 1957). Checkers are
sometimes encased in steel to prevent air infiltra-
tion through cracks and holes that develop in the
refractory regenerator walls during the campaign.
Combustion of fuel
Furnace size also has an effect upon use of fuel,
with a corresponding effect on the emissions of
dust. Large furnaces are more economical than
are small furnaces because the radiating surface
or heat loss per unit volume of glass is greater
for small furnaces.
Slightly greater fuel economy may be expected
from end port furnaces as compared with side
port furnaces of equal capacity. Here again, the
end port furnace has a heat loss advantage over
the side port furnace because it has less exposed
exterior surface area for radiating heat. Side
port furnaces can, however, be operated at great-
er percentages in excess of capacity since mixing
of fuel with air is more efficient through several
smaller inlet ports than it is through only one
large inlet port. In fact, end port furnaces are
limited in design to the amount of fuel that can
be efficiently mixed with air and burned through
this one inlet port (Spain, 1955). As far as dust
losses are concerned, there are only negligible
differences between end port and side port furnaces
of equal size. Reduced fuel consumption to re-
duce dust carryover can also be realized by in-
creasing the depth of the melter to the maximum
consistent with good-quality glass. Maximum
depths for container furnaces are 42 inches for
flint glass (Tooley, 1953) and about 36 inches for
amber glass and emerald green glass.
Dust emissions as well as fuel consumption can
also be reduced by firing practice. Rapid changes
in pull rates are wasteful of fuel and increase
stack emissions. Hence, charge rates and glass
pull rates for continuous furnaces should remain
as constant as possible by balancing loads be-
tween the glass-forming machines. If possible,
furnaces should be fired on natural gas or U.S.
Grade 3 or lighter fuel oil. Particulate emis-
sions increase an average of about 1 pound per
hour when U.S. Grade 5 fuel oil is used instead
of natural gas or U.S. Grade 3 fuel oil, and
opacities may exceed 40 percent white.
Combustion air should be thoroughly mixed -with
fuel with only enough excess air present to en-
sure complete combustion without smoke. Ex-
cess air robs the furnace of process heat by
dilution, and this heat loss must be overcome
by burning additional fuel. Volume of the melter
should be designed for a maximum fuel heat re-
lease of about 13, 000 Btu per hour per cubic foot.
Furnace reversals should be performed by an
automatic control system to ensure optimum
combustion. Only automatic systems can pro-
vide the exact timing required for opening and
closing the dampers and valves and for co-
ordinating fuel and combustion airflow (Bulcraig
and Haigh, 1961). For instance, fuel flow and
ignition must be delayed until combustion air
travels through the checkers after reversal to
mix with fuel at the inlet port to the melter. Fur-
nace reversals are usually performed in fixed
periods of 15 to 20 minutes, but an improvement
in regenerator efficiency can be realized by pro-
gramming reversal periods to checker tempera-
tures measured optically. Reversals can then
occur when checker temperatures reach preset
values consistent with maximum heat transfer
(Robertson et al. , 1957).
An excellent system for controlling air-to-fuel
ratios incorporates continuous flue gas analyzers
-------
Glass Manufacture
735
for oxygen and combustible hydrocarbons. With
this system, the most efficient combustion and
best flame shape and coverage occur at optimum
oxygen with a trace of combustible hydrocarbons
present in the flue gas. Sample gas is cleaned
for the analyzers through water-cooled probes
containing sprays. The system automatically
adjusts to compensate for changes in ambient
air density. Fuel savings of 6 to 8 percent
can be accomplished on furnaces -with analyzers
over furnaces not so equipped (Gunsaulus, 1958).
Combustion of natural gas in new furnaces occurs
efficiently "when the oxygen content of the flue
gases in the exhaust ports is less than 2 percent
by volume. As the campaign progresses, air
infiltration through cracks and pores in the
brick-work, air leakage through valves and damp-
ers, increased pressure drop through the regen-
erators, and other effects combine to make
combustion less efficient. To maintain maxi-
mum combustion throughout the campaign, pres-
sure checks with draft gages should be run peri-
odically at specified locations (Spain, 1956a).
Fuel savings can also be expedited by placing
furnace operators on an incentive plan to keep
combustion air to a minimum.
Electric melting
Although melting glass by electricity is a more
costly process than melting glass by natural
gas or fuel oil, melting electrically is a more
thermally efficient process since heat can be
applied directly to the body of the glass.
Electric induction systems installed on regen-
erative furnaces are designed to increase max-
imum pull rates by as much as 50 percent. These
systems are called boosters and consist of sev-
eral water-cooled graphite or molybdenum elec-
trodes equally spaced along the sides of the melt-
er 18 to 32 inches below the surface of the glass.
Source test results indicate that pull rates can
be increased without any appreciable increase in
dust carryover or particulate emissions. Fur-
nace temperatures may also be reduced by
boosters, preventing refractory damage at peak
operations.
Furnace capacity increase is nearly proportional
to the amount of electrical energy expended. A
56-ton-per-day regenerative furnace requires
480 kilowatt-hours in the booster to melt an addi-
tional ton of glass, which is close to the theoret-
ical amount of heat needed to melt a ton of glass
(Tooley, 1953).
Electric induction can also be used exclusively
for melting glass on a large scale. Design of
this type of furnace is simplified since regen-
erative checkerworks and large ductwork are no
longer required (Tooley, 1953). One recently
constructed 10-ton-per-day, all-electric furnace
consists of a simple tank with molybdenum elec-
trodes. A small vent leads directly to the at-
mosphere, and dust emissions through this vent
are very small. The furnace operates -with a
crown temperature below 600°F and with a
thermal efficiency of over 60 percent. Glass
quality is excellent, with homogeneity nearly
that of optical glass. After the first 11 months
of operation, there was no apparent wear on the
refractories (Peckham, 1962). First costs and
maintenance expenses are substantially lower
than for a comparable-size regenerative furnace.
An electric furnace may prove competitive with
regenerative furnaces in areas with low-cost
electrical power.
Baghouses and centrifugal scrubbers
Air pollution control equipment can be installed
on regenerative furnaces where particulate
emissions or opacities cannot be reduced to
required amounts through changes in furnace de-
sign, control of raw materials, and operating
procedures. Regenerative furnaces maybe
vented by two types of common industrial con-
trol devices--wet centrifugal scrubbers and
baghouses.
Figure 535 shows a low-pressure, wet, cen-
trifugal scrubber containing two separate con-
Figure 535. Wet, centrifugal-type scrubber con-
trolling emissions from a glass-melting furnace
(Thatcher Glass Co., Sangus, Calif.).
-------
736
CHEMICAL PROCESSING EQUIPMENT
tacting sections within a single casing. Sep-
arate 50-horsepower, circulating fans force
dirty gas through each section containing two
to three impingement elements similar to fixed
blades of a turbine. Although the collection
efficiency of this device is considered about the
highest for its type, source tests show an over-
all efficiency of only 52 percent. This low ef-
ficiency demonstrates the inherent inability of
the low-pressure, wet, centrifugal scrubbers
to collect particulates of submicron size.
On the other hand, baghouses show collection
efficiencies of over 99 percent. Although
baghouses have not as yet been installed on
large continuous, regenerative furnaces, they
have been installed on small regenerative fur-
naces. One baghouse alternately vents a 1, 800-
pound- and a 5, 000-pound-batch regenerative
furnace used for melting optical and special
glasses used in scientific instruments. Bags
are made of silicone-treated glass fiber. Off-
gases are tempered by ambient air to reduce the
temperature to 400°F, a safe operating temper-
ature for this fabric.
Another baghouse, although no longer in operation,
venteda 10-ton-per-day regenerative furnace for
melting soda-lirne flint glass. Stack gases •were
cooled to 250°F by radiation and convection from
an uninsulated steel duct before entering the bag-
house containing orlon bags.
To determine the feasibility of using a cloth fil-
tering device on large continuous, regenerative
furnaces, a pilot baghouse was used -with bags
made of various commercial fabrics. An air-
to-gas heat exchanger containing 38 tubes, each
1-1/2 inches in outer diameter by 120 inches in
length, cooled furnace exhaust gases before the
gases entered the pilot baghouse. The baghouse
contained 36 bags, each 6 inches in diameter by
111 inches in length, with a 432-net-square-foot
filter area. A 3-horsepower exhaust fan was
mounted on the discharge duct of the baghouse.
When subjected to exhaust gases from amber
glass manufacture, bags made of cotton, orlon,
dynel, and dacron showed rapid deterioration
and stiffening. Only orlon and dacron bags ap-
peared in satisfactory condition when controlling
dirty gas from flint glass manufacture and when the
dirty gas was held well above its dew point. This
difference in corrosion between amber and flint
glass was found to be caused by the difference in
concentrations of sulfur trioxide (803) present in
the flue gas.
To reduce the concentration of SOj from amber
glass manufacture, iron pyrites •were substituted
for elemental sulfur in the batch, but this change
met with no marked success. Stoichiometric
amounts of ammonia gas were also injected to
remove SO-^ as ammonium sulfate. Ammonia in-
jection not only failed to lessen bag deterioration
but also caused the heat exchanger tubes to foul
more rapidly.
In all cases, the baghouse temperature had to be
kept above the dew point of the furnace effluent
to prevent condensation from blinding the bags
and promoting rapid chemical attack. At times,
the baghouse had to be operated with an inlet
temperature as high as 280°F to stay above the
elevated dew point caused by the presence of SOj.
Additional pilot baghouse studies are needed to
evaluate orlon and dacron properly for flint glass
manufacture. Experiments are also required for
evaluating silicone-treated glass fiber bags in con
trolling exhaust gases from regenerative furnaces
melting all types of glass.
Information now available indicates that glass
fiber bags can perform at temperatures as
high as 500 °F, well above the elevated dew
points. They are virtually unaffected by rela-
tively large concentrations of SO^ and SO^, and
there is less danger from condensation. One
advantage of glass fiber is that less precooling
of exhaust gases is required because of the high-
er allowable operating temperatures. Reverse
air collapse is generally conceded to be the best
method of cleaning glass fiber bags, since this
material is fragile and easily breaks when regu-
lar shakers are installed.
Furnace effluent can be cooled by several meth-
ods: Air dilution, radiation cooling columns,
air-gas heat exchangers, and water spray
chambers. Regardless of the cooling method se-
lected, automatic controls should be installed to
ensure proper temperatures during the complete
firing cycle. Each cooling method has its ad-
vantages and disadvantages. Dilution of offgases
•with air is the simplest and most troublefree
way to reduce temperature but requires the larg-
est baghouse. Air-to-gas heat exchangers and
radiation and convection ductwork are subject to
rapid fouling from dust in the effluent. Automatic
surface-cleaning devices should be provided, or
access openings installed for frequent manual
cleaning to maintain clean surfaces for adequate
heat transfer. If spray chambers are used, se-
vere problems in condensation and temperature
control are anticipated.
GLASS-FORMING MACHINES
From ancient times, bottles and tableware were
made by handblowing until mechanical production
began in the decade preceding the turn of the cen-
tury with the discovery of the "press and blow"
-------
Glass Manufacture
737
and the "blow and blow" processes. At first,
machines were semiautomatic in operation.
Machine feeding was done by hand. Fully auto-
matic machines made their appearance during
World War I and completely replaced the semi-
automatic machines by 1925. Two types of auto-
matic feeders were developed and are in use today.
The first type consists of a device for dipping and
evacuating the blank mold in a revolving pot of
glass. The second type, called a gob feeder,
consists of an orifice in the forehearth combined
with shears and gathering chutes (Tooley, 1953).
Glass container-forming machines are of two
general types. The first type is a rotating ma-
chine in which glass is processed through a
sequence of stations involving pressing, blowing,
or both. An example of this type of machine is
a Lynch machine. A second type is used in con-
junction "with a gob feeder and consists of inde-
pendent sections in which each section is a com-
plete manufacturing unit. There is no rotation,
and the molds have only to open and close. An
example of this type is the Hartford-Empire
Individual Section (I. S.) six-section machine
shown in Figure 536. Mechanical details and
operations of various glass-forming machines
Figure 536. Hartford-Empire i.S. six-section glass-
forming machine. (Thatcher Glass Co., Sangus, Calif.).
for manufacturing containers, flat glass, and
tableware are found in the Handbook of Glass
Manufacture (Tooley, 1953).
The Air Pollution Problem
Dense smoke is generated by flash vaporization
of hydrocarbon greases and oils from contact
lubrication of hot gob shears and gob delivery
systems. This smoke emission can exceed 40
percent white opacity.
Molds are lubricated with mixtures of greases
and oils and graphite applied to the hot internal
surfaces once during 10- to 20-minute periods.
This smoke is usually 100 percent white in
opacity and exists for 1 or 2 seconds. It rapid-
ly loses its opacity and is completely dissipated
within several seconds.
Air Pollution Control Methods
During the past decade, grease and oil lubri-
cants for gob shears and gob delivery systems
have been replaced by silicone emulsions and
water-soluble oils at ratios of 90 to 150 parts
of water to 1 part oil or silicone. The effect
has been the virtual elimination of smoke. The
emulsions and solutions are applied by intermit'ten
sprays to the delivery system and shears only when
the shears are in an opened position.
Lubricating properties of silicone-based emul-
sions appear in some respects superior to those
of soluble oil solutions. Gob drop speeds are
increased by 20 to 25 percent. Apparently, the
gob rides down the delivery chute on a cushion
of steam. Heat from the gob breaks the silicone
emulsion, forming an extremely stable resin, a
condensation product of siloxane, which acts as
a smooth base for the cushioning effect of steam.
This resin is degraded in a matter of seconds
and must be reformed continuously by reapply-
ing the silicone emulsion.
While graphite gives no apparent advantages to
emulsions, a combination of water soluble oil and
silicone emulsion appears to be most effective
(Singer, 1956). Oil aids the wetting of metal
surfaces with silicone and coats metal surfaces,
retarding rust formation. Sodium nitrite is also
helpful in inhibiting rust when added to silicone
emulsion. Water for mixtures must be pure, and
in most cases, requires treatment in ion ex-
changers or demineralizers.
Water treatment is most critical for soluble oil to
prevent growth of algae and bacteria. Oil solutions
form gelatinous, icicle-like deposits upon drying
on the surfaces of pipes and arms of the I. S. ma-
-------
738
CHEMICAL PROCESSING EQUIPMENT
chine. These particles should not be allowed to
fall into the mold. Optimum results are obtained
by flood lubrication of the delivery system to the
maximum amount that can be handled by a runoff
wire or blown off by air. Dry lubrication of
delivery systems has been tried on an experi-
mental basis by coating the metal contact sur-
faces with molybdenum disulfide or graphite.
Although future developments in the application
of emulsions to molds look promising, present
practice still relies upon mixtures of hydro-
carbon greases, oils, and graphite. Silicone
emulsions and soluble oils eliminate smoke,
but several difficulties must be overcome be-
fore they can be widely used for mold lubrica-
tion. Water emulsions with their high specific
heat cause excessive cooling if they are not ap-
plied evenly to the mold surfaces by proper
atomization. Fine sprays meet with wind re-
sistance, and these sprays cannot be effectively
directed to cover the shoulder sections of some
molds. Because of the low viscosity of water
emulsions, the emulsions are very difficult to
meter through existing sight oil feeders. One
company has equipped its machine with individual
positive-displacement pumps for each nozzle.
Invert-post cross-spraying is found to be most
effective in giving a uniform coating to the molds
of I. S. machines (Bailey, 1957).
Rotating machines are much easier to lubricate
than are individual section machines. Emulsion
sprays are most effective on rotating machines
when mounted at the point of transfer of gobs
from the blank mold to the blow mold.
FRIT SMELTERS
INTRODUCTION
Ceramic coatings are generally divided into two
classes, depending upon whether they are applied
to metal or to glass and pottery. In the case of
metal, the coating is widely referred to in this
country as porcelain enamel. The use of the
term vitreous enamel seems to be preferred in
Europe. Glass enamel is sometimes used inter-
changeably with both terms. On the other hand,
the coating applied to glass or pottery is known
as ceramic glaze.
Ceramic coatings are essentially water suspen-
sions of ground frit and clay. Frit is prepared
by fusing various minerals in a smelter. The
molten material is then quenched with air or
water. This quenching operation causes the
melt to solidify rapidly and shatter into numerous
small glass particles, called frit. After a drying
process, the frit is finely ground in a ball mill,
where other materials are added. When suspend-
ed in a solution of water and clay, the resulting
mixture is known as a ceramic slip. Enamel
slip is applied to metals and fired at high tem-
peratures in a furnace. Glaze slip is applied to
pottery or glass and fired in a kiln.
Raw Materials
The raw materials that go into the manufacture
of various frits are similar to each other whether
the frit is for enameling on steel or aluminum or
for glazing. The basic difference is in the chem-
ical composition.
The raw materials used in enamels and glazes
may be divided into the following six groups:
Refractories, fluxes, opacifiers, colors, float-
ing agents, and electrolytes (Andrews, 1961).
The refractories include materials such as
quartz, feldspar, and clay, which contribute to
the acidic part of the melt a.nd give body to the
glass. The fluxes include minerals such as
borax, soda ash, cryolite, fluorspar, and litharge,
which are basic in character and react with the
acidic refractories to form the glass and, more-
over, tend to lower the fusion temperatures of
the glasses. These refractory and flux materials
chiefly comprise the ingredients that go into the
raw batch that is charged to the smelter.
Materials falling into the other four groups are
introduced later as mill additions and rarely ex-
ceed 15 percent of the total frit composition.
They include opacifiers, which are compounds
added to the glass to give it an opaque appear-
ance such as the characteristic white of porce-
lain enamels. Examples are tin oxide, anti-
mony oxide, sodium antimonate, and zirconium
oxide. The color materials include compounds
such as the oxides of cobalt, copper, iron, and
nickel. The floating agents consist of clay and
gums and are used to suspend the enamel or
glaze in water. Electrolytes such as bos:ax,
soda ash, magnesium sulfate, and magnesium
carbonate are added to flocculate the clay and
further aid the clay in keeping the enamel or
glaze in suspension (Parmelee, 1951).
Types of Smelters
Smelters used in frit making, -whether for
enamel or glaze, may be grouped into three
classes: Rotary, hearth, and crucible. The
rotary smelter is cylindrical and can be rotated
in either direction to facilitate fusing, as shown
in Figure 537. It can also be tilted vertically
for the pouring operation, as demonstrated in
Figure 538. The smelter is open at one end for
the introduction of fuel anc combustion air. It
is similarly open at the opposite end for the dis-
charge of flue gases and for charging raw mate-
GPO 806—614—25
-------
Frit Smelters
739
Figure 537. Rotary-type frit
Los Angeles, Cal if.).
smelter (Ferro Corp.
Figure 538. Rotary-type frit smelter in pouring position (Ferro Corp.,
Los Angeles, Cali f.)
-------
740
CHEMICAL PROCESSING EQUIPMENT
rials. Operated solely as a batch-type smelter,
it is normally charged by means of a screw con-
veyor, which is inserted through the opening.
Rotary smelters are normally sized to take
batches varying from approximately 100 to 3, 000
pounds. Fired -with either gas or oil, the smelter
is lined with high-alumina, refractory firebrick
with an average life of from 400 to 600 melts.
Firing cycles vary from 1 to 4 hours.
The hearth smelter consists of a brick floor, on
which the raw materials are melted, surrounded
by a boxlike enclosure. This type of smelter
can be either continuous, as illustrated in Figure
539, or batch type, as shown in Figure 540. In
either case, the hearth (or bottom) is sloped
from one side to a point on the opposite side
where the molten material is tapped. The con-
tinuous type is usually screw fed. A flue stack
is located on the opposite end. Oil or gas is
normally used as fuel for the one or more burn-
ers. The walls and floor are lined with a first-
quality, refractory firebrick. The batch type is
sized to take batches ranging from 100 to several
thousand pounds. About 30 pounds of batch can
be smelted for each square foot of hearth area.
The typical continuous-hearth smelter can pro-
cess 1, 000 to 1, 500 pounds of raw materials
per hour.
The crucible smelter consists of a high-refrac-
tory, fireclay, removable crucible mounted
v/ithin a circular, insulated, steel shell lined
with high-grade firebrick, as shown in Figure
541. Heating is usually accomplished with oil
or gas burners, though electricity can be used.
The combustion chamber surrounds the crucible,
occupying the space between the crucible and the
shell lining. Because the heat must be trans-
mitted through the crucible to the batch, re-
fractory and fuel costs are high. Crucibles can
be sized to smelt a 5-pound batch for laboratory
purposes, but the commercial crucibles are
sized to take bitches from 100 pounds to 3, 000
pounds. Smeltiug cycles vary from 2 to 3 hours
at 'v
-------
Frit Smelters
741
Figure 539. Continuous-hearth-type frit smelter (Ferro Corporation
Los Angeles, Calif.).
Figure 540. Batch-hearth-type frit smelter (Ferro Corporation,
Los Angeles, Cal if.).
-------
742
CHEMICAL PROCESSING EQUIPMENT
Figure 541. Crucible-type frit smelter (California
Metal Enameling Company, Los Angeles, Calif.).
water to flow out through an overflow. Rapid
cooling is somewhat impeded by this method
owing to a layer of steam that forms over the
glass. Air--water quenching appears to be the
most economical and effective method since a
more thorough shattering of the glass results.
In this method the molten material is poured
from the smelter and passed through a blast of
air and water. Quenching causes the molten ma-
terial to solidify and shatter into numerous small
glass particles (called frit) ranging from 1/4 inch
in diameter down to submicron sizes. Its main
purpose is to facilitate grinding.
After draining, die frit contains 5 to I 5 percent
water and inay be milled in this condition or
may first be dried. Three types of dryers are
employed: The drying table, the stationary
dryer, and the rotary dryer. The drying table
is a flat hearth on which the frit is placed. Heat
is applied beneath the hearth, and the frit is
raked manually. The stationary dryer consists
of a sheet iron chamber in which a basket of
frit is placed. Heated air from an exchanger on
the smelter flue is passed through the basket of
frit. The rotary dryer consists of a porcelain-
lined rotating cylinder that is inclined slightly,
causing the frit to move through continuously.
The typical size is approximately 2 feet in di-
ameter and 20 feet long, though larger cylinders
are used. The rotary dryer, which is economical
and efficient, can be heated by waste heat or by
oil or gas. The frit can be further refined by
using magnetic separation to remove small iron
particles, which would otherwise cause black
specks in the enamel.
The final step in frit making is size reduction,
which is normally done with a ball mill. Frits
used in porcelain enamel are required to pass
a No. 100 sieve (150 \i), though a certain per-
cent of fines must remain as residue on a finer
sieve. In the case of ceramic glaze frits, a
finer grind is necessary. About one-half of a
batch must be less than 2. 5 microns with the
remainder no greater than 1C microns. Effi-
cient milling is best obtained when the speed of
rotation is such that the balls ride three-fourths
of the way up one side of the cylinder, and the inner
most balls slide back down over the outermost
balls. This is achieved, for example, at a speed
of 25 rpm for a 4-foot-diameter cylinder. Porce-
lain balls or flint pebbles are used in the mill. The
diameter of the balls ranges from 1 to 3 inches,
and the charge should be maintained at about 55
percent of the mill volume. Ball wear amounts to
5 to 10 pounds in milling 1, 000 pounds of frit.
Colors, opacifiers, floating agents, and electro-
lytes are mixed with the frit before it is charged
to the ball mill. After the milling operation be-
gins, water is added at a constant rate to keep
the specific gravity of the slurry (referred to as
slip) at the correct value at all times. After the
milling operation, the ceramic slip is screened
to remove large particles. A 1 - to 2-day aging
process then takes place at a temperature close
to that at -which the enamel or glaze is to be ap-
plied. Aging is necessary to set up an equilibrium
among the clay, frit, and solution. The enamel
or glaze slip is now ready for application.
Application, Firing, and Uses of Enamels
Enamels and glazes may be applied to ware
blanks by immersion or spraying (Hansen,
1932). The pouring and brushing methods are
seldom employed today. In the dipping opera-
tion, the blank is immersed in the slip and then
withdrawn and allowed to drain. If the slip is
thick, the excess enamel must be shaken from
the ware, a process called slushing. Spraying
is the application of enamel or glaze slip to
ware by atomizing it through an air gun.
After the enamel or glaze has been applied, it
must then be burned or fired on the ware to
fuse the coating to a smooth, continuous, glassy
layer. The firing temperatures and cycles for
porcelain enamel on steel and aluminum are
approximately 1,500°F (Shreve, 1945) for 5
minutes and 1,000°F for 5 minutes, respective-
ly. Ceramic glaze, however, is fired on pottery
at about 2, 300 °F for several hours or even days.
-------
Frit Smelters
743
The firing is accomplished in what is called a
furnace in the porcelain enamel industry, and a
kiln in the ceramic glaze industry.
Porcelain enamel is used as a protective coating
for metals--primarily steel, cast iron, and
aluminum. Familiar items are bathtubs, water
heater tanks, refrigerators, washing machines,
and cooking ranges. Coated aluminum is being
used more and more in recent times for signs
such as those installed on highways. Ceramic
glazes are used as a decorative or protective
coating on a wide variety of pottery and glass
articles. Examples are lavatory basins, water
closets, closet bowls, chinaware, and figurines.
THE AIR POLLUTION PROBLEM
Significant dust and fume emissions are created
by the frit-smelting operation. These emissions
consist primarily of condensed metallic oxide
fumes that have volatilized from the molten charge.
They also contain mineral dust carryover and
sometimes contain noxious gases such as hydro-
gen fluoride. In addition, products of combus-
tion, and glass fibers are released. The quanti-
ty of these air contaminants can be reduced by
following good smelter-operating procedures.
This can be accomplished by not rotating the
smelter too rapidly, to prevent excessive dust
carryover, and by not heating the batch too rapid-
ly or too long, to prevent volatilizing the more
fusible elements before they react with the more
refractory materials. A typical rotary smelter,
for example, discharges to the atmosphere, 10 to
15 pounds of dust and fumes per hour per ton of
material charged. In some cases, where ingredi-
ents require high melting temperatures (1, 500°F
or higher), emissions as great as 50 pounds per
hour per ton of material have been observed.
Depending upon the composition of the batch, a
significant visible plume may or may not be
present. Tables 214 through 217 indicate the
extent of emissions from uncontrolled, rotary
frit smelters for various-sized batches and com-
positions.
HOODING AND VENTILATION REQUIREMENTS
Rotary smelters require a detached canopy-type
hood suspended from the lower end of a vertical
stack as shown in Figures 537 and 538. It is
suspended far enough above the floor to trap the
discharge gases from the smelter when in the
horizontal position. Refractory-lined, it is of
sufficient size to prevent gases from escaping
into the room, its size varying with the size of
the smelter. The typical hood opening area
ranges from 3 to 5 square feet. The stack should
be of sufficient height to obtain good draft--about
20 feet--if it is not vented to air pollution control
Table 214. DUST AND FUME DISCHARGE FROM
A 1, 000-POUND, ROTARY FRIT SMELTER
Test data
Process wt, Ib/hr
Stack vol. scfm
Stack gas temp, °F
Concentration, gr/scf
Stack emissions, Ib/hr
CO, vol % (stack condition)
N^, vol % (stack condition)
Process wt, Ib/hr
Stack vol, scfm
Stack gas temp, °F
Concentration, gr/scf
Stack emissions, Ib/hr
CO, vol % (stack condition)
N2, vol % (stack condition)
1
174a
1, 390
450
0. 118
1.41
0. 002
76.9
4
292b
1, 310
960
0. Ill
1.25
0
73
2
174a
1,540
750
0. 387
5. 11
0.001
75. 10
5
292b
1,400
950
0. 141
1. 79
0
72. 60
3
174a
1,630
900
0. 381
5. 32
0.002
73. 50
6
292b
1,480
930
0. 124
1.57
0
73. 30
aThese three tests represent approximately the 1st, 2d, and 3d
hours of a 248-minute smelting cycle. The total charge amounted
to 717 pounds of material consisting of borax, feldspar, sodium
fluoride, soda ash, and zinc oxide.
These three tests represent approximately the 1st, 2d, and 3d
hours of a 195-minute smelting cycle. The total charge amounted
to 949 pounds of material consisting of litharge, silica, boric
acid, feldspar, fluorspar, borax, and zircon.
Table 215. DUST AND FUME DISCHARGE FROM
A 3, 000-POUND, ROTARY FRIT SMELTER
Test data
Process wt, Ib/hr
Stack vol, scfm
Stack gas temp, °F
Concentration, gr/scf
Stack emissions, Ib/hr
CO, vol % (stack condition)
N2> vol % (stack condition)
Test No.
7
472a
2,240
630
0. 143
2.70
0.02
75. 30
8
472a
2,270
800
0. 114
2.20
0. 02
75. 60
9
472a
I, 260
840
0. 172
3.30
0.02
76.30
aThese three tests represent approximately the 1st, 2d, and 3d
hours of a 248-minute smelting cycle. The total charge
amounted to 1, 951 pounds of material consisting of litharge,
silica, boric acid, feldspar, whiting, borax, and zircon.
Table 216. FLUORIDE DISCHARGE FROM
A ROTARY FRIT SMELTER
Process wt, Ib/hr
Stack vol, scfm
Stack gas temp, °F
Concentration, gr/scf
Stack emissions, Ib/hr
Test No.
10
174a
1,400
530
0. 061
0. 73
11
174a
1, 600
840
0.035
0.48
12
162b
1, 000
480
0. 196
1.68
13
162b
1, 000
480
0.058
0.50
aThese two tests were of 90 minutes' duration each and represented
approximately the first half and the second half of a 248-minute
smelting cycle. The total charge amounted to 7 1 7 pounds of material
consisting of borax, feldsparj sodium fluoride, soda ash, and zinc
oxide.
These two 6t)-minute tests represented approximately the 1st and the
4th hours of a 450-minute smelting cycle. The total charge amounted
to l,iJ13 pounds of material consisting of sodium carbonate, calcium
carbonate, pyrobar, and silica. The test was specifically conducted
for a batch containing maximum carbonates (19%) and no litharge.
-------
744
CHEMICAL PROCESSING EQUIPMENT
Table 217. DUST AND FUME DISCHARGE FROM
A 2, 000-POUND ROTARY FRIT SMELTER
Process wt, Ib/hr
Stack vol, scfm
Stack gas temp, °
Concentration, gr
Stack emissions,
F
,'scf
Ib/hr
Test No.
14
857a
2, 430
600
0. 130
2. 710
15
857a
2, 430
600
0. 112
2. 340
16
890b
4, S47
340
0. Ill
4. 150
17
890b
4, 347
340
0. 103
5.820
aThesc two 60-minute tests represent the 1st and 2d hours of a 140-
minute smelting cycle. The total charge amounted to 2,000 pounds
of material containing silica, litharge, and whiting.
bThese two 60-minute tests represent the 1st hour and 37 minutes of
a 135-minutc smelting cycle. The total charge amounted to 2,000
pounds of material containing silica, litharge, and whiting.
equipment. If it is vented to control equipment,
ventilation requirements are approximately 3, 000
scfm for a 2, 000-pound batch smelter as an ex-
ample. Hood indraft velocity should be about 500
fpm.
Crucible and hearth smelters do not require hoods
but do require a 20- or 25-foot stack to conform
with good chimney design practice if not vented to
air pollution control equipment. Some crucible
smelters are vented directly into the room. If
vented to air pollution control equipment, a canopy
hood must be used on the crucible smelter. Hood
indraft velocities should be approximately 200 fpm.
The requirement for a hearth- (box-) type smelter
is approximately 4, 000 scfm for a 3, 000-pound
batch smelter. As a general rule, about 70 scfm
is required for each square foot of hearth area.
AIR POLLUTION CONTROL EQUIPMENT
The two most feasible control devices for frit
smelters are baghouses arid venturi water scrub-
bers. Of these devices, baghouses are more ef-
fective. Glass bags cannot be used, however,
owing to the occasional presence of fluorides in
the effluent. The discharge gases must be cooled
by heat exchangers, quench chambers, cooling
columns, or by some other device to a tempera-
ture compatible with the fabric material selected.
Filtering velocities should not exceed 2. 5 fpm.
A venturi-type water scrubber is satisfactory if
at least 20 to 25 inches of pressure drop is main-
tained across the venturi throat. The throat veloc
ity should be between 15, 000 and 20, 000 fpm. Th<
water requirement at the throat is about 6 gpm for
each 1, 000 cubic feet of gas treated. Power con-
sumption is high owing to the high pressure drop.
The venturi scrubber shown in Figure 542 was
installed to serve one rotary and two hearth smelt
ers simultaneously. Table 218 includes data
indicating the collection efficiency of this scrub-
ber when venting a frit srnelter.
A baghouse installation venting four rotary, gas-
fired frit smelters is shown in Figure 543. The
production capacity of one of the smelters is
3, 000 pounds while that of the other three is 1, 00
pounds each. Maximum gas temperatures encour
tered in the discharge stack at a point 20 feet dovv
stream of the smelters are approximately 950°F
while the average temperature is 780"F.
Table 218. EFFICIENCY OF VENTURI WATER SCRUBBER ON PARTICULATE
MATTER AND FLUORIDES WHEN VENTING THREE FRIT SMELTERS
Test No.'
Test data
Process wt, Ib/hr
Stack vol, scfm
Stack gas temp, °F
Dust concentration, gr/scf
Inlet
Outlet
Dust emissions, Ib/hr
Inlet
Outlet
Control efficiency, %
18
19
20
Dust and fumes
1,360
4,280
570
0.228
0. 074
8. 37
2. 72
67. 50
1, 360
4, 280
552
0.234
0. 077
8.60
2.85
67.20
1, 360
4, 280
564
0. 127
0. 088
1. 78
1. 35
30. 70
21
22
23
Fluorides
1, 360
4, 280
570
0.092
0.006
3.38
0.22
93.20
1, 360
4, 280
552
0. 137
0.008
5. 03
0. 29
94
1, 360
4,280
564
0. 034
0.017
0.48
0.26
50
aTests No. 18 and 21 represent the first 54 minutes of the 107-minute smeltirg cycle, tests No. 19
and 22 represent the last 54 minutes, and tests No. 20 and 23 represent the 23-minute tapping
period. Total process weight was 3, 000 pounds of material consisting of borax, potassium car-
bonate, potassium nitrate, zinc oxide, titanium, oxide, ammonium phosphate, lithium carbonate,
sodium silico-fluoride, fluorspar, silica, and talc. Pressure drop across throat was 21 in. WC.
Water flow rate to throat was 50 gpm.
-------
Frit Smelters
745
Figure 542. Ventun water scrubber venting
three frit smelters (Ferro Corporation, Los
Angeles, Calif.)-
TfTra
lllllli'lU
Figure 543. Baghouse with radiant cooling
columns venting four rotary frit smelters
(Glostex Chemicals, Inc., Vernon, Calif.).
-------
746
CHEMICAL PROCESSING EQUIPMENT
The baghouse is a cloth-tubular, pullthrough
type, containing 4, 400 square feet of cloth area.
It is equipped with an exhaust fan that delivers
9, 300 cfm at approximately l?0°F. The filtering
velocity is 2. 2 fpm.
Radiation cooling columns are used to reduce the
effluent gas temperature from 585 °F at the inlet
to the cooling columns to 185°F at the baghouse
inlet. Approximately 1, 300 lineal feet of 30-
inch-diameter, heavy-gage steel duct with a sur-
face area of 10, 000 square feet is used. The av-
erage overall heat transfer coefficient is 1. 35 Btu
per hour per square foot per °F, as calculated
from actual test data. The cooling columns are not
one continuous run, but consist of single, double,
and triple runs. Thus, the gas mass velocity
varies considerably throughout the unit, with
resulting changes in heat transfer coefficients.
Additional cooling is accomplished with dilution
air at the detached hoods, which are suspended
about 1 foot away from the discharge end of each
smelter. The baghouse inlet temperature of
185°F is satisfactory for the dacron cloth mate-
rial used, and excellent bag life can be expected.
FOOD PROCESSING EQUIPMENT
Most foods consumed in the United States today,
whether of animal or vegetable origin, are pro-
cessed to some degree before marketing. His-
torically, certain foods have been subjected to
various preserving processes. More recently
we find food purveyors increasingly concerned
with processes that render foods more flavorful
and easier to prepare. The trend toward great-
er presale food preparation has possibly caused
a shift of at least some air pollutants from many
domestic kitchens to a significantly smaller num-
ber of food-processing plants.
Food processing includes operations such as
slaughtering, smoking, drying, cooking, bak-
ing, frying, boiling, dehydrating, hydrogenating,
fermenting, distilling, curing, ripening, roast-
ing, broiling, barbecuing, canning, freezing,
enriching, and packaging. Some produce large
volumes of air contaminants, others, only in-
significant amounts. Equipment used to process
food is legion. Some of the unit operations in-
volved are the following (Kirk and Othmer, 1947):
Material handling: Conveying, elevating, pump-
ing, packing and shipping.
Separating: Centrifuging, draining, evacuating,
filtering, percolating, fitting, pressing, skimming,
sorting, and trimming (drying, screening, sifting,
and washing fall into this category).
Heat exchanging: Chilling, freezing, and refrig-
erating; heating, cooking, broiling, roasting, bak-
ing, and so forth.
Mixing: Agitating, beating, blending, diffusing,
dispersing, emulsifying, homogenizing, kneading,
stirring, whipping, working, and so forth.
Disintegrating: Breaking, chipping, chopping,
crushing, cutting, grinding, milling, maturating,
pulverizing, refining, (as by punching, rolling,
and so forth), shredding, slicing, and spraying.
Forming: Casting, extruding, flaking, molding,
pelletizing, rolling, shaping, stamping, and die
casting.
Coating: Dipping, enrobing, glazing, icing, pan-
ning, and so forth.
Decorating: Embossing, imprinting, sugaring,
topping, and so forth.
Controlling: Controlling air humidity, temperatu:
pressure, and velocity; inspecting, measuring, tei
pering, weighing, and so forth.
Packaging: Capping, cloi=ing, filling, labeling,
packing, wrapping, and so forth.
Storing: Piling, stacking, •warehousing, and so
forth.
A description and discussion of each type of equip-
ment used for food processing is not within the
scope of this manual. The following discussion
will be limited to food processes in -which air polli
tion problems are inherent and in which typical
food-processing air contaminants are encountered
This section is not concerned with the production
of pet foods or livestock feeds, though in some
instances, these materials are byproducts of food
processes.
COFFEE PROCESSING
Most coffee is grown in Central and South Americ
After harvesting and drying at or near the coffee
plantation, most "green" coffee beans are exporte
and further processed before sale to the consume
Coffee processing in the United States consists es
sentially of cleaning, roasting, grinding, and pac
ing.
Roasting is the key operation and produces most
of the air contaminants associated with the indus-
try. Roasting reduces the sugar and moisture
contents of green coffee and also renders the bull
density of the beans about 50 percent lighter. Ar
apparently desired result is the production of
•water-soluble degradation products that impart
-------
Food-Processing Equipment
747
most of the flavor to the brewed coffee. Roasting
also causes the beans to expand and split into
halves, releasing small quantities of chaff.
Batch Roasting
The oldest and simplest coffee roasters are direct-
fired (usually by natural gas), rotary, cylindrical
chambers. These units are designed to handle
from 200 to 500 pounds of green beans per 15- to
20-minute cycle and are normally operated at
about 400°F. A calculated quantity of water is
added at the completion of the roast to quench the
beans before discharge from the roaster. After
they are dumped, the beans are further cooled
with air and run through a "stoner" air classifier
to remove metal and other heavy objects before
the grinding and packaging. The roaster and
cooler and all air-cleaning devices are normally
equipped with cyclone separators to remove dust
and chaff from exhaust gases. Most present-day
coffee roasters are of batch design, though the
newer and larger installations tend to favor con-
tinuous roasters.
In the batch roaster shown in Figure 544, some
of the gases are recirculated. A portion of the
gases is bled off at a point between the burner
and the roaster. Thus, the burner incinerates
combustible contaminants and becomes both an
air pollution control device and a heat source for
the roaster.
I PERIOD I - BEFORE SMOKE APPEARS Heating medium is
I circulated normally for about '4 of the roasting
[cycle. Damper C is open to vent excess gases. 1)m~
I per B ts closed Cut-off slide A is open for normal
I circuiation.
I PERIOD II - WHEN FIRST SHORE APPEftRS, Damper C is
I closed- forcing excess gases through the name (ts
I burn smoke). Damper B is open to vent excess gases
| after smoke ts burned. Cut-off SI ids A ts open
I normal circulation. Roasting period completed B
I cept for application of water,
I PERIOD 111 - WHEN iATER IS APPLIED. Cut-off Si
I A is closed to prevent return of water to the c
ff««.
lean
through trie flame. Damper 8 remains open to vent
al! gases after smoke is burned and steam is reduced
ta ftmsicfe vapor
Figure 544. A recirculating-batch coffee
roaster (Jabez Burns - Gump Division,Blaw-
Knox Company, New York, N.Y.).
An Integrated Coffee Plant
A process flow sheet of a typical large, integrated
coffee plant is shown in Figure 545. Green beans
are first run through mechanical cleaning equip-
ment to remove any remaining hulls and foreign
matter before the roasting. This system is seen
to include a dump tank, scalper, weigh hopper,
mixer, and several bins, elevators, and convey-
ors. Cleaning systems such as this commonly
include one or more centrifugal separators from
which process air is exhausted.
The direct, gas-fired roasters depicted in Fig-
ures 545 and 546 are of continuous rather than
batch design. Temperatures of 400°F to 500°F
are maintained in the roaster, and the residence
time is adjusted by controlling the drum speed.
Roaster exhaust products are drawn off through
a cyclone separator and afterburner, with some
recirculation from the cyclone to the roaster.
Chaff and other particulates from the cyclone
are fed to a chaff collection system. Hot beans
are continuously conveyed through the air cooler
and stoner sections. Both the cooler and the
stoner are equipped with cyclones to collect par-
ticulates.
The equipment following the stoner is used only
to blend, grind, and package roasted coffee.
Normally, there are no points in these systems
where process air is emitted to the atmosphere.
At the plant shown on the flow sheet, chaff is
collected from several points and run to a hold-
ing bin from which it is fed at a uniform rate to
an incinerator. Conveyors in the chaff system
may be of almost any type, though pneumatic con-
veyors are most common. The design of the in-
cinerator depicted is similar to that of the saw-
dust burners described in Chapter 8 but the
incinerator is much smaller.
The Air Pollution Problem
Dust, chaff, coffee bean oils (as mists), smoke,
and odors are the principal air contaminants
emitted from coffee processing. In addition,
combustion contaminants are discharged if chaff
is incinerated. Dust is exhausted from several
points in the process, while smoke and odors are
confined to the roaster, chaff incinerator, and, in
some cases, to the cooler.
Coffee chaff is the main source of particulates,
but green beans, as received, also contain ap-
preciable quantities of sand and miscellaneous
dirt. The major portion of this dirt is removed
by air washing in the green coffee-cleaning sys-
tem. Some chaff (about 1 percent of the green
weight) is released from the bean on roasting and
-------
748
CHEMICAL PROCESSING EQUIPMENT
GREEN
COFFEE
DUMP
TO
ATMOSPHERE
I
TO TO
ATMOSPHERE ATMOSPHERE
i i
AFTERBURNER ;
T"
GAS
SURGE
BIN
CYCLONE
I I
BURNER Ft-*1 1
CONTINUOUS
ROASTER
TO GRINDING
BLENDING AND
AND PACKAGING
Figure 545. Typical flow sheet for a coffee-roasting plant.
Figure 546. A continuous coffee roaster and cooler:(left) continuous
roaster, showing course of the heated gases as they are drawn through
the coffee beans in the perforated, helical-flanged cylinder and then
into the recirculation system; (right) left-side elevation of contin-
uous roaster, showing relationship of recirculating and cooler fans
and the respective collectors on the roof (Jabez Burns - Gump Division,
Blaw-Knox Company, New York, N.Y.).
-------
Food-Processing Equipment
749
is removed with roaster exhaust gases. A small
amount of chaff carries through to the cooler and
stoner. After the roasting, coffee chaff is light
and flaky, particle sizes usually exceeding 100
microns. As shown in Table 219, particulate-
matter emissions from coffee processing are well
below the limits permitted by typical dust and
fume prohibitions.
Table 219. ANALYSIS OF COFFEE
ROASTER EXHAUST GASES
Contaminant concentration
Particulate matter, gr/scf
Aldehydes
(as formaldehyde), ppm
Organic acids
(as acetic acid), pprn
Oxides of nitrogen
(as NO^), ppm
Continuous roaster
Roaster
0. 189
139
223
26.8
Cooler
0. 006
--
--
--
Batch roaster
0. 160
42
175
21.4
Coffee roaster odors are attributed to alcohols,
aldehydes, organic acids, and nitrogen and sulfur
eompounds, which are all probably breakdown
products of sugars and oils. Roasted coffee odors
are considered pleasant by many people, and in-
deed, they may often be pleasant under certain
conditions. Nevertheless, continual exposure to
uncontrolled roaster exhaust gases usually elicits
widespread complaints from adjacent residents.
The pleasant aroma of a short sniff apparently
develops into an annoyance upon long exposure.
Visible bluish-white smoke emissions from coffee
roasters are caused by distilled oils and organic
breakdown products. The moisture content of
green coffee is only 6 to 14 percent, and thus
there is not sufficient water vapor in the 400 °F
to 500°F exhaust gases to form a visible steam
plume. From uncontrolled, continuous roasters,
the opacity of exhaust gases exceeds 40 percent
almost continuously. From batch roasters, ex-
haust opacities normally exceed 40 percent only
during the last 10 to 15 minutes of a 20-minute
roast. Smoke opacity appears to be a function
of the oil content, the more oily coffee producing
the heavier smoke. The water quenching of
batch-roasted coffee causes visible steam emis-
sions that seldom persist longer than 30 seconds
per batch.
Hooding and Ventilation Requirements
Exhaust volumes from coffee-processing sys-
tems do not vary greatly from one plant to
another insofar as roasting, cooling, and stoning
are concerned. Roasters equipped with gas re-
circulation systems exhaust about 24 scf per
pound of finished coffee. Volumes from nonre-
circulation roasters average about 40 scf per
pound. A 10,000-pound-per-hour, continuous
roaster with a recirculation system exhausts
about 4, 000 scfm. A 500-pound-per-batch, non-
recirculation roaster exhausts about 1,000 scfm.
Each batch cycle lasts about 20 minutes.
Coolers of the continuous type exhaust about
120 scf per pound of coffee. Batch-type cool-
ers are operated at ratios of about 10 scfm per
pound. The time required for batch cooling
varies somewhat with the operator. Batch-cool-
ing requirements are inversely related to the
degree of water quenching employed.
Continuous-type stoners use about 40 scf air per
pound of coffee. Batch-stoning processes require
from 4 to 10 scfm per pound, depending upon duct-
work size and batch time.
Air Pollution Control Equipment
Air contaminants from coffee-processing plants
have been successfully controlled with afterburn-
ers and cyclone separators, and combinations
thereof. Incineration is necessary only with roaster
exhaust gases. There is little smoke in other coffee
plant exit gas streams where only dust collectors
are required to comply with air pollution control
regulations.
Separate afterburners are preferable to the com-
bination heater-incinerator of the batch roaster
shown in Figure 544. When the afterburner serves
as the roaster's heat source, its maximum operat-
ing temperature is limited to about 1, 000 °F. A
temperature of 1,200°F or greater is necessary
to provide good particulate incineration and odor
removal.
A roaster afterburner should always be preceded
by an efficient cyclone separator in which most
of the particulates are removed. A residence
time of 0. 3 second is sufficient to incinerate
most vapors and small-diameter particles at
1,200°F. Higher temperatures and longer resi-
dences are, however, required to burn large-
diameter, solid particles. Afterburner design
is discussed in Chapter 5.
Properly designed centrifugal separators are
required on essentially all process air streams
up to and including the stoner and chaff collec-
tion system. With the plant shown, cyclones
are required at the roaster, cooler, stoner,
chaff storage bin, and chaff incinerator. In
addition, the scalper is a centrifugal classifier
venting process air. Some plants also vent the
-------
750
CHEMICAL PROCESSING EQUIPMENT
green coffee dump tank and several conveyors and
elevators to centrifugal dust collectors.
For best results the chaff incinerator should be
of the design discussed in Chapter 8 in which
combustible material is fed at a uniform rate.
It is, however, considerably smaller and has
burning rates usually below 100 pounds per hour.
The inorganic ash content of the chaff, at approxi-
mately 5 percent by weight, is considerably great-
er than that of most combustible refuse fed to
incinerators. Provisions should be made in the
incinerator design so that this material does
not become entrained in the exhaust gases. If
most of the noncombustible material is dis-
charged with products of combustion from the
incinerator, the combustion contaminants then
exceed 0. 3 grain per cubic foot calculated to
12 percent carbon dioxide.
SMOKEHOUSES
Smoking has been used for centuries to preserve
meat and fish products. Modern smoking opera-
tions do not differ greatly from those used by our
forefathers, though the prime purposes of smoking
today appear to be the imparting of flavor, color,
and "customer appeal" to the food product. Cur-
ing and storage processes have been improved
to the point where preservation is no longer the
principal objective.
The vast majority of smoked products are meats
of porcine and bovine origin. Some fish and
poultry and, in rare instances, vegetable prod-
ucts are also smoked as gourmet items.
Table 220. ANALYSIS OF WOOD SMOKE
USED IN MEAT SMOKEHOUSES
(Jensen, 1945)
Contaminant Concentration, ppm
Formaldehyde
Higher aldehydes
Formic acid
Acetic and higher acids
Phenols
Ketones
Resins
20 to 40
140 to 180
90 to 125
460 to 500
20 to 30
190 to 200
1,000
Atmospheric Smokehouses
The oldest smokehouses are of atmospheric or
natural-draft design. These boxlike structures
are usually heated directly with natural gas or
wood. Smoke is often generated by heating
sawdust on a steel plate. These smoke gener-
ators are normally heated with natural gas pipe
burners located in the bottom of the house. Hot,
smoky gases are allowed to rise by natural con-
vection through racks of meat. Large atmospheric
houses are often built with two or three levels of
meat racks. One or more stacks are provided
to exhaust spent gases at the top of the house. In
some instances the vents are equipped with ex-
haust fans. During the smoking and drying cy-
cles, exhaust gas temperatures range from
120° to 150°F. Slightly higher temperatures
are sometimes encountered during the cooking
cycle.
The Smoking Process
Smoking is a diffusion process in which food
products are exposed to an atmosphere of hard-
wood smoke. Table 220 lists an analysis of
smoke produced through the destructive distilla-
tion of a hardwood. As smoke is circulated over
the food, aldehydes, organic acids, and other
organics are adsorbed onto its outer surface.
Smoking usually darkens the food's natural color,
and in some cases, glazes the outer surface.
Regardless of smokehouse design, some spent
gases are always exhausted to the atmosphere.
These contain odorous, eye-irritating gases and
finely divided, organic particulates, often in
sufficient concentration to exceed local opacity
restrictions.
Smokehouses are also used to cook and dry food
products either before or after smoking. Air
contaminants emitted during cooking and drying
are normally well below allowable control limits.
Recirculating Smokehouses
Most large, modern, production meat smoke-
houses are of the recirculating type (Figure 547)
wherein smoke is circulated at reasonably high
velocities over the surface of the product. The
purpose is to provide faster and more nearly
uniform diffusion of organics onto the product,
and more uniform temperatures throughout the
house. These units are usually of stainless
steel construction and are heated by steam or
gas. Smoke is piped to the house from external
smoke generators. Each unit is equipped with
a large circulating fan and, in some instances, a
smaller exhaust fan. During smoking and cooking,
exhaust volumes of 1 to 4 cfm per square foot of
floor area are maintained. The exhaust rate is
increased to 5 to 10 cfm per square foot during
the drying cycle. Recirculating smokehouses
are usually equipped with temperature and hu-
midity controls, and the opacity and makeup of
exhaust gas are usually more constant than those
from atmospheric units.
-------
Food-Processing Equipment
751
AUTOMATIC
ALTERNATING
DAMPERS
HOT -I
AIR
AND
SMOKE
SUPPLY
HOT
AIR
AND
SMOKE
SUPPLY
ftmmt
Figure 547. A modern recirculating smokehouse
(Atmos Corp., Chicago, III.).
The Air Pollution Problem
Smokehouse exhaust products include organic
gases, liquids, and solids, all of -which must
be considered air contaminants. Many of the
gaseous compounds are irritating to the eyes
and reasonably odorous. A large portion of the
particulates is in the submicron size range where
light scattering is maximum. These air con-
taminants are attributable to smoke, that is, to
smoke generated from hardwood, rather than
from the cooked product itself.
Exhaust gases from both atmospheric and re-
circulating smokehouses can be periodically
expected to exceed 40 percent opacity, the
maximum allowable under many local air pollu-
tion control regulations. With the possible ex-
ception of public nuisance, smokehouse exhaust
gases are not likely to exceed other local air
quality standards. As shown in Table 220, con-
centrations of particulate matter average only
0. 14 grain per scf.
Hooding and Ventilation Requirements
Atmospheric smokehouses are designed with ex-
haust volumes of about 3 cubic feet per square
foot of floor area. Somewhat higher volumes are
used with atmospheric houses of two or more
stories. Inasmuch as there are no air recircula-
tion and normally little provision for forced draft,
the exhaust rate for an atmospheric house is es-
sentially constant over the drying, cooking, and
smoking cycles. Moreover, there is often some
smoke in the house even during the cooking and
drying cycles. This is particularly true where
smoke is generated in the house rather than in
an external smoke generator. If gases are to be
ducted to air pollution control equipment, an ex-
haust fan should be employed to offset the added
pressure drop. When an afterburner is used, it
can often be positioned to provide additional nat-
ural draft.
Recirculation smokehouses have a considerably
"wider range of exhaust rates. During smoking
and cooking cycles, volumes of 1 to 4 cubic
feet per square foot of floor area are exhausted.
The rate increases to 5 to 10 cubic feet per square
foot during the drying cycle. Recirculation houses
are almost always equipped with external smoke
generators, and a control of smoke flow is much
more positive. There is essentially no smoke in
the houses during the cooking and drying cycles.
Most smokehouses do not require hooding. Ex-
haust gases are normally ducted directly to the
atmosphere or to control equipment. Some at-
mospheric houses are, however, equipped with
hoods over the loading doors to gather smoke that
might escape during the shifting of meat racks. The
latter situation is due to the inherently poor dis-
tribution of smoke and heat in an atmospheric house.
To maintain product uniformity, the meat racks
must often be shifted while there is smoke in the
house. Most atmospheric houses do not have ex-
haust systems adequate to prevent appreciable
smoke emissions from the door during these in-
stances. Hoods and exhaust systems are some-
times installed principally for worker comfort. The
hoods or fans, or both, may be located in corridor
ceilings immediately above the doors. These ven-
tilators are often operated automatically whenever
the doors are opened. Volumes can be appreciable,
in some instances exceeding the smokehouse's ex-
haust rate.
There are normally no appreciable smoke emis-
sions from doors of recirculation-type smoke-
houses. Temperature and smoke distribution
are sufficient so that there is no need to shift
meat in the houses. Moreover, the doors are
designed to provide tighter closures. Recircula-
-------
752
CHEMICAL PROCESSING EQUIPMENT
tion houses are operated under positive pressure,
and any small opening causes large emissions of
smoke.
Bypassing control devices during nonsmoking
periods
Many operators of recirculation smokehouses
find it desirable to bypass air pollution control
devices during nonsmoking periods. From the
standpoint of air pollution control, this practice
is not unreasonable. The major smokehouse air
contaminant is smoke. Concentrations of air con-
taminants during cooking and drying are relative-
ly small, comparable to those of ordinary meat-
cooking ovens. Drying-cycle exhaust gases are
2 to 4 times more voluminous than those vented
during the smoking cycle. The size of control
equipment is materially increased if drying
gases are ducted to it. The initial cost and oper-
ating cost of a smokehouse's air pollution control
system can, therefore, be considerably reduced
if exhaust gases are bypassed during drying and
cooking cycles when no smoke is introduced into
the house.
If houses are to be bypassed during nonsmoking
periods, the ductwork and valving should be de-
signed to provide automatic or nearly automatic
operation. Water seal dampers (Figure 548) are
preferable. Mechanical dampers demand optimum
maintenance for satisfactory closure. They are
considerably more likely to malfunction owing to
corrosion and contamination with greases and tars.
Moreover, mechanical dampers are more suscepti-
ble to physical damage than water dampers are.
Ideally, damper operation should be keyed to other
GASES FROM
SMOKEHOUSE
TO CONTROL
-*—
DEVICE
Figure 548. Diagram of a water-operated damper
used to bypass the air pollution control device
during nonsmoking periods.
smokehouse auxiliaries such as fans and smoke
generators. Where controls are manually oper-
ated, there is a strong possibility that dampers
will not be opened or closed at proper times,
causing either overloading of the control device
or the discharge of untreated air contaminants
directly to the atmosphere.
Air Pollution Control Equipment
Afterburners
Smoke, odors, eye irritants,, and organic partic-
ulate matter can be controlled with afterburners,
provided temperature and design are adequate.
Most of these contaminants can be eliminated at
temperatures of 1, 000°F to 1, 200°F in well-de-
signed units. Larger diameter particulate matter
is somewhat more difficult to burn at these tem-
peratures; however, since concentrations of par-
ticulate matter from smokehouses are reasonably
small, this limitation is not critical.
Electrical precipitators
Low-voltage, two-stage electrical precipitators
were installed in the Los Angeles area as early
as 1957 to control visible smokehouse air con-
taminants. They have since been used at many
other locations in the United States. Before
1957, their use had been confined principally to
air-conditioning applications.
Electrical precipitators are, of course, effective
only in the collection of particulate matter. They
cannot be used to control gases or vapors. At
smokehouse installations, their purpose is to
collect the submicron smoke particles responsi-
ble for visible opacity. Two-stage precipitators
have been shown capable of reducing smoke opaci-
ties to less than 10 percent under ideal conditions.
A typical two-stage precipitator control system
with a wet, centrifugal collector is shown in Fig-
ure 549. The wet collector is used to control
temperature and humidity zmd also remove a
small amount of particulates. This is followed
by a heater in which gas temperatures are regu-
lated before the gases enter the ionizer. Voltages
of 6, 000 to 15, 000 volts are applied to the ionizer
and plate sections. Particulate matter collects
on the plates and drains, as a gummy liquid, to
the collection pan below.
For satisfactory control of visible emissions, it
has been found that superficial gas velocities
through the plate collector section should not ex-
ceed 100 fpm. Some difficulty has been experience
owing to channeling in the collector. For best opei
ation, vanes or other means of ensuring uniform
flow should be used ahead of the plate section.
-------
Food-Processing Equipment
753
Figure 549. A two-stage precipitator and wet centrifugal
collector venting smokehouses (The Rath Packing Co., Vernon, Calif.).
Even under optimum conditions, a slight trace of
smoke can be expected from the precipitator' s
outlet. At the discharge of the unit, eye irrita-
tion is usually severe, and odors are strong
though not overpowering. These odors and eye
irritants can constitute a public nuisance, de-
pending upon plant location.
Electrical precipitation versus incineration
Both electrical precipitation and incineration
offer the classical choice of high initial cost
versus high operating cost, but in addition, they
differ markedly from the standpoint of air pollu-
tion control.
Electrical precipitators are capable of collect-
ing particulate matter and thereby reducing
visible emissions to tolerable amounts. They
have no effect on nitrogen oxides and little
effect, if any, on gaseous eye irritants and
odors. If arcing occurs, some small and prob-
ably insignificant quantity of ozone is also pro-
duced. The initial cost of precipitators is high,
and their operating cost low in comparison with
that of afterburners. Smokehouse precipita-
tors do, however, require a relatively high
degree of maintenance. If they are not proper-
ly maintained, poor control efficiency and fire
damage are probable. Fire damage can result
in extended outage periods during which uncon-
trolled exhaust gases may vent directly to the
atmosphere.
Incineration is much more effective than elec-
trical precipitation is in controlling gaseous
organics and finely divided particulates. Large
particles are, however, relatively difficult to
burn at the normal operating temperatures and
residence times of smokehouse afterburners.
Under average conditions, collection efficiency
for particulate matter (about 65 percent) is
roughly the same as that of a two-stage elec-
trical precipitator. Fuel costs make the oper-
ation of an incineration device more expensive
than that of a precipitator. Nevertheless,
maintenance is much less a problem. There
is no buildup of tars and resins in the afterburner
or stack to impede its operation. As with any
smokehouse control device, tars accumulate in
the ductwork between the house and afterburner,
necessitating periodic cleaning. As shown in
Table 221, incineration creates additional nitro-
gen oxides, increasing concentrations from about
4 ppm to approximately 12 ppm on the average.
Comparative test data on smokehouse afterburners
and electrical precipitators, as shown in Tables
221 and 222, indicate that collection efficiencies
for particulate matter, aldehydes, and organic
acids are of the same magnitude for both types
of control dequipment. These data fail to re-
flect larger concentrations of odors and eye
irritants from electrical precipitators that are
readily apparent upon personal inspection of the
devices.
-------
754
CHEMICAL PROCESSING EQUIPMENT
Table 221. ANALYSES OF MEAT SMOKEHOUSE EXHAUST GASES BEFORE AND
AFTER INCINERATION IN NATURAL GAS-FIRED AFTERBURNERS
Particulate matter,
gr /scf
Aldehydes (as form-
aldehyde), ppm
Organic acids (as
acetic acid)
Oxides of nitrogen
(as NO2), ppm
Contaminant concentration
Smokehouse
Range
0.016 to 0.234
8 to 74
30 to 156
1. 2 to 7. 2
Average
0. 141
40
87
3.9
Afterburner
Range
0. Oil to 0. 070
5 to 61
0 to 76
3.7 to 33.8
Average
0. 048
25
33. 5
11. 7
Control
efficiency,
66
38
62
Negative
Table 222. ANALYSES OF MEAT SMOKEHOUSE EXHAUST GASES BEFORE AND AFTER
CONTROL IN TWO-STAGE ELECTRICAL PRECIPITATION SYSTEMS
Particulate matter,
gr/scf
Aldehydes (as formalde-
hyde), ppm
Organic acids (as acetic
acid), ppm
Contaminant concentration
Smokehouse
Range
0. 33 to 0. 181
Average
0. 090
74
91
Control systema
Range
0. 016 to 0. 051
Average
0. 032
47
48
Control
efficiency, %
65
37
47
aEach control system is equipped with a wet centrifugal collector upstream irom the
precipitator.
Why not immersion?
Conventional smoking operations can be seen
as an extremely devious method of coating food
products with a myriad of hardwood distillation
products. One might wonder why this coating
is not applied by simple immersion. Unfortu-
nately, many of the compounds present in smoke
are highly toxic. If these were deposited heavily
on the food product, results could be fatal. Smok-
ing, therefore, provides a reasonably foolproof,
if quaint, means of assuring that these toxic com-
pounds do not accumulate in lethal concentrations.
Many states have laws prohibiting the smoking of
meats by liquid immersion.
Smoking through electrical precipitation
Some attempts have been made to precipitate
smoke particles electrically onto food products
in the smokehouse, and a few smokehouses so
designed are in operation today. From the opera-
tors1 point of view, this arrangement offers the
advantages of faster smoking and greater use
of generated smoke. From the standpoint of air
pollution control, it is desirable inasmuch as
considerably lesser quantities of air contami-
nants are vented to the atmosphere than are
vented from a conventional, uncontrolled smoke-
house.
These units normally consist of a conveyorized
enclosure equipped with an ionizer section simi-
lar to those used with two-stage precipitator s.
The food product is usually passed 2 to 3 inches
below the ionizing wires, which are charged with
about 15, 000 volts. No electrical charge is ap-
plied to the food products or the conveyor. These
smokers are operated at ambient temperatures
and do not lend themselves to use for either cook-
ing or drying food products,, As would be expected,
spacing is a critical factor.
-------
Food-Processing Equipment
755
There are very few precipitation smokehouses
in the Unites States today, and for this reason,
little reliable data about the operating charac-
teristics or the air pollutants emitted are avail-
able. Smokehouses of this design have been
reported to operate with visible emissions of
only 5 to 10 percent opacity. Concentrations of
air contaminants in gases from precipitation-
type smokehouses would, under optimum condi-
tions, be expected to be approximately equiva-
lent to those from conventional smokehouses
equipped with two-stage electrical precipitators.
These units offer the potential of markedly re-
duced smoking times. Indeed, the few operating
units have residence times of less than 5 minutes.
If equipment such as this were perfected for a
•wider range of operation, residence times would
not be expected to exceed 10 minutes.
The application of precipitation smokehouses
is today limited by a number of inherent problems,
the foremost of which is the irregular shape of
many smoked products, that is, hams, ham hocks,
and salami. The degree of smoke deposition in
a unit such as this is governed by the distance
between the ionizer and the food product. Irregu-
lar spacing results, therefore, in irregular smok-
ing of round and odd-shaped products that cannot
be positioned so that all surfaces are equidistant
from ionizer wires. The few existing installa-
tions are used to impart a light smoke to regular-
shaped, flat items such as fish fillets and sliced
meat products.
DEEP FAT FRYING
Deep fat or "French" frying involves the cooking
of foods in hot oils or greases. Deep-fried prod-
ucts include doughnuts, fritters, croquettes, vari-
ous potato shapes, and breaded and batter-dipped
fish and meat. Most of these foods contain some
moisture, a large portion of which is volatilized
out as steam during frying. Some cooking oils,
as well as animal or vegetable oils from the prod-
uct, are usually steam distilled during the pro-
cess.
Batch or Continuous Operation
Deep fat frying is in common usage in homes,
restaurants, and frozen food plants. In the home
and in smaller commercial establishments, batch-
type operation is more common. The principal
equipment is an externally heated cooking oil vat.
Oil temperatures are usually controlled to be-
tween 325° and 400 °F. Almost any type of
heating is possible. Where combustion fuels
are used, burner gases are vented separately.
The product to be fried is either manually or
mechanically inserted into the hot grease and
removed after a definite time interval.
In large commercial establishments, highly
mechanized, conveyorized fryers, such as that
shown in Figure 550, are used. The raw food
product is loaded onto an endless conveyor belt
and passed through hot grease at a rate adjusted
to provide the proper cook time. Almost all
Figure 550.
(right) end
A continuous deep
view (J.C. Pitman
fat fryer; (left) Interior view,
& Sons, Inc., Concord, N.H.).
-------
756
CHEMICAL PROCESSING EQUIPMENT
fryers are of one-pass design. Frequently, cook-
ing units are followed by product coolers and pack-
aging and freezing equipment.
The Air Pollution Problem
In a typical large industrial operation of this
type, the cooking vat constitutes the principal
source of air contaminants. Uncooked materials
are usually wet or pasty, and the feed system
produces little or no air pollution. Most cooked-
product-handling systems are also innocuous,
except in rare instances where fine, dusty mate-
rials are encountered.
Odors, visible smoke, and entrained fat particles
are emitted from the cooking vats. Depending
upon operating conditions and the surrounding
area, these contaminants may or may not be in
sufficient concentration to exceed the limits of
local opacity or nuisance regulations.
From the standpoint of air pollution control, the
most objectionable operations involve foods con-
taining appreciable fats and oils. Light ends oi
these oils are distilled during cooking. In gen-
eral, the deep frying of vegetable products is
less troublesome than that of fish and meat prod-
ucts, which contain higher percentages of iats
and oils.
Most food products cooked in this manner con-
tain between 30 and 75 percent moisture before
the cooking. Almost all moisture is driven off
in the cooking vat and appears as steam in ex-
haust gases. Moisture concentrations in stack
gases are usually between 5 and 20 percent, de-
pending upon the volume of air drawn into the
cooker hood and exhaust system. In highly
mechanized installations, very little air enters
under the cooker hood. As a result, the \varm
air-stream from a fryer such as this is often
saturated, and downstream cooling causes visi-
ble condensation at or near the stack exit.
Moisture has two effects: (1) It causes fats and
oils to be steam distilled from the cooking vat,
and (2) it masks visible stack emissions. Smoke
observations of equipment such as this must be
made at the point in the stack plume where water
vapor has disappeared. This is best accomplished
when the weather is warm and dry. On a cold,
moist day, the vapor plume may extend as far
as the smoke,
Excessive smoking is most often due either to
overheating or to the characteristics of the
material being cooked. When, for instance,
potato chip or corn chip fryers are operated
in normal temperature ranges, there is usually
no more than a trace of smoke in exhaust gases.
On the other hand, several meat product fryers
have been found to exhaust gases of high opacity,
and control equipment was needed to bring them
into compliance with local regulations. These
visible emissions appear to be finely divided
fat and oil particles distilled either from the
product or the cooking oil. Cooking oils are
usually compounded within reasonably narrow
boiling ranges, and when fresh, very little of
the oils is steam distilled. Most objectionable
air contaminants probably originate, therefore,
in the product or in spent cooking oil.
The carryover of oil droplets can also cause a
nuisance by spotting fabrics, painted surfaces,
and other property in the surrounding area.
This problem is most likely to occur when the
raw food contains relatively large concentrations
of moisture, a situation in -which steam distilla-
tion is proportionally higher.
Hooding and Ventilation Requirements
Deep fat fryers should always be hooded and
ventec through a fan. Axial-flow fans are pre-
ferred. Exhaust volumes are governed by the
open area under the hood. Where there is open
area around the full hood periphery, the indraft
velocity should be at least 100 fpm. In many
modern units, the dryer sides are completely
enclosed, and the only open areas are at the con-
veyor's inlet and outlet. At these installations,
exhaust volumes are considerably lesser, even
though indraft velocities are well above 100 fpm.
If control equipment is to be employed, exhaust
volumes become an important factor. In these
instarces, redesigning the existing hoods to low-
er the exhaust rates is often desirable.
Air Pollution Control Equipment
Incineration, low-voltage electrical precipita-
tion, and entrainment separation have been used
to control air contaminants from deep fat fryers.
Since practically all air contaminants from fry-
ers are combustible, a well-designed afterburn-
er provides adequate control if the operating tem-
perature is sufficiently high. Temperatures from
1,000° to 1,200°F are often sufficient to eliminate
smoke-causing particulates and to incinerate odors
and eye irritants. The combustion of larger par-
ticles usually requires higher temperatures, some-
times as high as 1,600°F. The concentration of
particulates in fryer exit gases is, however, nor-
mally less than 0. 1 grain per scf, which is -well
below common limits for particulate emissions.
Two-stage, low-voltage electrical precipitators
(6, 000 to 15, 000 volts) can be used to collect a
substantial portion of the particulates responsible
-------
Food-Processing Equipment
757
for visible air contamination. These devices,
unfortunately, do not remove the gaseous con-
taminants that are usually responsible for odors
and eye irritation. As would be expected, the
effectiveness of a precipitator depends upon the
particular fryer it is serving. If particulates
are the only significant contaminants in the ex-
haust gases, a precipitator can provide an ade-
quate means of control. If, on the other hand, the
problem is due to odors of overheated oil or prod-
uct, a device such as this is of little benefit. For
optimum performance, the temperature, humidity,
and volume of gases vented to a two-stage pre-
cipitator must be controlled within reasonably
narrow limits. The oils collected are usually
free flowing and readily drain from collector
plates. A collection trough should be provided
to prevent plate fouling and damage to the roof
or other supporting structure on which the pre-
cipitator is located.
Oil collection
Entrainment separators have been employed
with varying success to remove entrained oils
in fryer exhaust stacks. These are most use-
ful "where the concentration of oils is relatively
large. The material collected can represent a
savings in oil and can prevent damage to ad-
jacent roofing. Because of the inherently low
collection efficiency of these devices, their use
•would not be recommended where smoke or
odors constitute the major air pollution prob-
lem. Some cooking oils usually collect on the
inner surfaces of uninsulated exhaust stacks
and drain back towards the cooker. Most com-
mercial fryers are equipped "with Dans to col-
lect this drainage at the bottom of the stack.
LIVESTOCK SLAUGHTERING
Slaughtering operations have traditionally
been associated with odorous air contaminants,
though much of these odors is due to byproduct
operations rather than to slaughtering and meat
dressing itself. Slaughtering is considered to
include only the killing of the animal and the
separation of the carcass into humanly edible
meat and inedible byproducts. The smoking
of edible meat products, and reduction of edi-
ble materials are discussed in this subsection,
while the reduction of inedible materials is
covered in another part of this chapter.
Cattle-, sheep-, and hog-killing operations are
necessarily more extensive than those concerned
with poultry, though poultry houses usually han-
dle appreciably larger numbers of animals.
A flow diagram of a typical cattle-slaughtering
operation is shown in Figure 551. The animal
is stunned, bled, skinned, eviscerated, and
trimmed as shown. Blood is drained and col-
lected in a holding tank. After removal, en-
trails are sliced in a "gut hasher, " then washed
to separate the partially digested food termed
"paunch manure. " Many slaughterers have
heated reduction facilities in which blood, in-
testines, bones, and other inedible materials
are processed to recover tallow, fertilizer,
and animal feeds. The firms that do not oper-
ate this equipment usually sell their offal to
scavenger plants that deal exclusively in by-
products. Hides are almost always shipped to
leather-processing firms. Dressed beef, nor-
mally about 56 percent of the live weight, is
refrigerated before it is shipped.
The Air Pollution Problem
Odors represent the only air contaminants
emitted from slaughtering operations. The
odors could be differentiated as (1) those
released from the animal upon the killing and
cutting, and upon the exposure of blood and
flesh to air; and (2) those resulting from the
decay of animal matter spilled on exposed sur-
faces or otherwise exposed to the atmosphere.
Odors from the first source are not appreciable
when healthy livestock is used. Where nuisance-
causing odors are encountered from slaughter-
ing, they are almost always attributable to in-
adequate sanitary measures . These odors are
probably breakdown products of proteins. Amines
and sulfur compounds are considered to be the
most disagreeably odorous breakdown products.
In addition to these sources, there are odors
at slaughterhouse stockyards and from the stor-
age of blood, intestines, hides, and paunch
manure before their shipping or further process-
ing.
Air Pollution Control Equipment
As has been explained, odorous air contami-
nants are emitted from several points in a
slaughtering operation. Installing control equip-
ment at each source would be difficult if not im-
possible. Methods of odor control available in-
clude: (1) Rigid sanitation measures to prevent
the decomposition of animal matter, and (2) com-
plete enclosure of the operation to capture the
effluent and exhaust it through a control device.
Where slaughtering is government inspected,
the operators are required to wash their kill
rooms constantly, clean manure from stock
pens, and dispose of all byproducts as rapidly
as possible. These measures normally hold
plant odors to a tolerable minimum.
-------
758
CHEMICAL PROCESSING EQUIPMENT
Figure 551. Typical livestock-slaughtering and processing area
(The Globe Company, Chicago, III.).
When a slaughterer is located in a residential
area, the odor reduction afforded by strict
sanitation may not be sufficient. In these in-
stances, full-plant air conditioning might be
necessary. Filtration with activated carbon
would appear to be the only practical means
of controlling the large volume of exhaust
gases from a plant of this type. The latter
method has not yet been employed at slaughter-
houses in the United States. Nevertheless,
activated-carbon filtration of the entire plant
has been employed to control similar odors
at animal matter byproduct plants. With in-
creasing urbanization, this method of control
may, conceivably, be used in the near future.
EDIBLE-LARD AND TALLOW RENDERING
Methods used to produce edible lard and tallow
are similar to those described later in this
chapter for rendering of inedibles. As -with
processes for inedibles, feedstocks are heated
either directly or indirectly -with steam to ef-
fect a phase separation yielding fats, water,
and solids. Moisture is removed either by
vaporization or by mechanical means. Tallow
and solids are mechanically separated from
one another in presses, centrifuges, and filters.
The only major process differences between
rendering edibles and rendering inedibles are
due to the composition and freshness of the
materials handled. Edible feedstocks contain
80 to 90 percent lard or tallow, 10 to ZO per-
cent moisture, and less than 5 percent muscle
tissue. Inedible feedstocks contain appreciably
higher precentages of both moisture and solids.
Edible feedstocks, in addition to being more
select portions of the animal, are generally
much fresher than inedible cooker materials
are.
Whenever its products are intended for human
consumption, the process is much more stringent-
ly supervised and regulated by Federal and local
agencies. There are numerous government regu-
lations concerning the freshness of edible-render-
ing feedstocks, the cleanliness of processing
equipment, and the handling of rendered fats.
For instance, paragraph 15. 1 of the United States
Department of Agriculture's Meat Inspection
Regulations specifies that inspected feed mate-
rial must be heated to a temperature not lower
than 170°F for a period of not less than 30
minutes -when edible lard or tallow is being pro-
duced.
-------
Food-Processing Equipment
759
Dry Rendering
Most of the high-quality edible lard and tallow
are produced in indirectly steam-heated cookers.
These processes are frequently carried out at
temperatures of less than 212°K. The lower
operating temperatures are afforded either by
vacuum cooking or by finely grinding the feed-
stocks. The vacuum process is usually per-
formed batchwise in a horizontal, steam-jac-
keted cooker very similar to those used for
rendering of inedibles. The vacuum is usually
created through the use of steam- or water-
operated ejectors. Variations in dry, edible-
rendering processes are usually concerned
with temperatures and the degree of comminu-
tion of fats. Where raw materials are ground
into fine particles, operation at lower tem-
peratures is usually possible, even without
a vacuum-producing device.
Low-Temperature, Continuous Rendering
ous basis from high-fat feedstocks. A typical
process is shown in Figure 552. Feedstocks
are first introduced to a grinder, where they
are finely shredded at 120 °F, and then heated
to approximately 185°F before being passed
through a desludging centrifuge in which solids
are removed from the water and tallow. Liq-
uids are then reheated to about 200 °F in a
steam jet heater. The remaining moisture is
removed from the hot tallow in a second cen-
trifuge from which edible lard or tallow is
run to storage. The separated water is piped
to a skimming pond where it is cooled before
being sewered. Vapors from the several
vessels are vented to a fume scrubber (con-
tact condenser).
Wet Rendering
The wet rendering process involves rendering
of fats in a vertical, closed tank with the feed
material in direct contact with live steam.
Dry rendering processes have been developed
to produce edible lard and tallow on a continu-
The principal advantage of this type of render-
ing is that large quantities of lard or tallow
CUTTING AND/DR
KILLING FATS
HEADER SLOPE TO SCRUBBER, NO POCKETS OR TRAPS t~
rr* . . prv-prx—-T 1
IME t—J
IBBER\/
DIAPHRAGM
•* iVALVE
V r-M TEMP CONTROLLER
V V
r*
-------
760
CHEMICAL PROCESSING EQUIPMENT
can be produced without finely grinding the feed
material. Low-cost equipment and labor can
be used.
Wet rendering, however, necessarily requires
higher temperatures (280° to 300 °F) and in-
ternal pressures of 40 to 38 psig. The quality
of the lard or tallow produced is relatively low,
owing to the high temperatures to which it is
subjected.
The Air Pollution Problem
The only noteworthy air contaminants generated
from edible-rendering processes are odors. In
comparison with odors generated from inedible-
rendering processes, however, those from
edible-rendering processes are -relatively minor.
In Los Angeles County, rendering of edibles ac-
counts for only about 10 percent of the total ani-
mal matter rendered. Rendering of inedibles
at packing houses constitutes approximately 32
percent and that at scavenger plants accounts
for the remaining 58 percent of the tonnage.
In addition, rates of odor emissions from ren-
dering of edibles are low compared with those
from inedible-rendering processes. Inasmuch
as edible feedstocks contain relatively low per-
centages of water, the resultant steam generated
from cookers is not appreciable, 6,300 scf per
ton. Feedstocks contain approximately 15 per-
cent moisture, as compared with 50 percent
from inedible cooker materials. Odor concen-
trations in exhaust gases from the rendering
of edibles are significant at 3, 000 odor units
per scf but not excessive. Equipment at plants
rendering edibles is kept scrupulously clean,
which substantially reduces odors from inplarit
handling operations.
Hooding and Ventilation Requirements
Almost always, cooker gases from rendering of
edibles can be piped directly to air pollution con-
trol devices. "Where condenser odor control de-
vices are used, there is usually enough vacuum,
that is, pressure differential, in the ductwork
to cause vapors to flow from the cooker at a
sufficiently high rate. Steam or water ejectors
are sometimes employed to lower operating
temperatures or to remove water vapor more
rapidly. Uncondensible gases do not exceed
5 percent of cooker gases unless there is ap-
preciable leakage into the system, as through
seals on shafts, doors, and so forth.
Where cooking is performed at pressures
greater than 1 atmosphere, piping must usually
be arranged in a manner that prevents surging
when high-pressure gases are released. If
the main valve is released quickly, the high-
pressure vapors usually cause slugs of grease
and solids to be carried over into the control
system. Severe surging can cause siphoning
of all the material from cooker to the control
system. To prevent this, the piping is often
arranged with a small pipe, 1 to 2 inches in
diameter, that bypasses the main cooker's
exhaust line. High pressures are reduced by
venting first through the small pipe to the con-
trol device. Once the high pressure is relieved,
the large valve can be opened to provide great-
er flow.
Air Pollution Control Equipement
Water spray contact condensers are the simplest
devices used for controlling odorous air con-
taminants from rendering of edibles. These con-
dense ai major portion of the steam-laden effluent
vapors and dissolve much of the odorous materi-
als. Water requirements of the contact condenser
for edible-rendering operations are considerably
lower than those for the conta.ct condenser used
to control cooker gases from rendering of in-
edibles. This is due primarily to the lower
moisture content of feedstocks and the resultant
lower volume of steam exhausted from the cook-
er. Exit -water temperatures should be held
below 140°F to prevent the release of volatile,
odorous materials from down stream piping
and sewers.
Surface condensers are also satisfactory con-
trol devices for edible-rendering processes.
At the same condensate volume and tempera-
ture, however, surface condensers by them-
selves are not as effective as contact con-
dens 63'S. This is due to the inherently lower
condensate volume and larger concentration
of odorous materials in the condensate of sur-
face condensers.
That an edible-rendering process -would require
more extensive odor control than would be af-
forded by an adequate condenser is unlikely.
Nevertheless, uncondensed offgases from con-
densers could be further controlled by incin-
eration or carbon adsorption, as outlined for
processing of inedibles later in this chapter.
FISH CANNERIES AND FISH
REDUCTION PLANTS
Canning is the principal method of preserving
highly perishable fish foodstuffs . Canneries
for this purpose are usually located near har-
bors where fish can be unloaded directly from
boats. Byproduct reduction plants are operated
at or near fish canneries to process scrap ma-
-------
Fish Canneries and Fish Reduction Plants
761
terials, and much of the odorous air contami-
nants generally attributed to canneries emanate
from byproduct processes. Only choice por-
tions of sound fish are canned for human con-
sumption. The remainder is converted into by-
products, notably fish oil and high-protein
animal feed supplements.
Basically there are two types of fish-canning
operations in use today. In the older, so-called
"wet-fish" method, trimmed fish are cooked
directly in the can. The more popular "pre-
cooked" process is used primarily to can tuna.
The latter method is characterized by the cook-
ing of whole, eviscerated fish, and the hand
sorting of choice parts before canning.
into open cans that are conveyed through a 100-
to ZOO-foot-long hot-exhaust box. Here live
steam is employed to cook the fish. Hot-ex-
haust boxes are vented through several stacks
located along their lengths (Figure 553). At
the discharge end, cans may be mechanically
upended so that "stick water" is decanted from
the cans while the cooked fish remains. Stick
water consists of condensed steam, juices, and
oils that have cooked out of the fish. This liq-
uid is collected and retained for byproduct
processing as described later in this section.
The cans of drained fish are filled with tomato
sauce, olive oil, or other suitable liquid before
being sealed. Sealed cans are pressure cooked
before their labeling, packing, and shipping.
WET-FISH CANNING
Wet-fish canning is used to preserve salmon,
anchovies, mackerel, sardines, and similar
species that can be obtained locally and brought
to the cannery quickly. The distinctive feature
of the wet-fish process is the complete removal
of heads,tails, and entrails before the cooking.
Trimmed ana eviscerated raw fish is packed
TUNA CANNING
The precooked canning method was developed
to improve the physical appearance of canned
fish. It is confined to the commercial canning
of larger fishes, principally tuna. Whole,
eviscerated fish are placed in wire baskets and
charged to live-steam-heated cookers such as
those of Figure 554. The cookers are operated
Figure 553. Unsealed cans of cooked mackerel being conveyed from
the hot-exhaust box cooker of a wet-fish process (Star-Kist Foods,
Inc., Terminal Island, Calif.).
-------
762
CHEMICAL PROCESSING EQUIPMENT
Figure 554. A bank of Iive-steam-heated cookers
used to process raw, whole tuna (Star-Kist Foods
Inc., Terminal Island, Calif.).
at about 5 psig pressure, condensate being dis-
charged through steam traps. Air, steam, and
any uncondensed, odorous gases are bled from
the cookers through one or more small vents in
the ceiling.
As the fish are cooked, juices, condensed steam,
and oils are collected, centrifuged, and pumped
to stick water and oil storage tanks. Cooking
reduces the weight of a fish by about one-third.
After the cooking, the flesh is cooled so that it
becomes firm before it is handled. It is then
placed on a conveyorized picking line. Operators
stationed along the conveyor select the portions
to be canned for human consumption. After being
packed and sealed in cans, the fish is pressure
cooked for sterilization before its labeling, pack-
ing, and shipping. Much of the dark meat is
canned for pet food. Only about one-third of the
raw tuna weight is canned as food for humans and
pets. The remaining skin, bone, and other scrap,
roughly amounting to one-third of the raw weight,
is fed to the fish meal reduction system.
CANNERY BYPRODUCTS
A large fraction of the fish received in a cannery
is processed into byproducts. In the precook
process, about two-thirds of the raw fish weight
is directed to byproduct reduction systems as
stick water or solid scrap. The wet-fish process
usually produces somewhat less offal, depending
principally upon the size of fish. Typical head-
and-tail mackerel scrap is pictured in Figure 555.
In addition, whole fish may be rejected at the can-
ning line because of spoilage, freezer burns, bad
color, and so forth. Any fish or portions of fish
Figure 555. Typical raw head-and-tai I mackerel
scrap awaiting processing in a fish meal reduction
system (Star-Kist Foods, Inc., Terminal Island,
Calif.).
not suitable for human consumption or for pet
food are handled in the reduction plant. In order
of volume and relative importance, the byproducts
are: Fish meal, used almost exclusively as an
animal feed supplement; fish oil, used in the
paint industry and in vitamin manufacture; and
"liquid fish" and "fish solubles," high-protein
concentrates. The latter are manufactured
somewhat differently, but both are used as ani-
mal feed supplements and as fertilizers.
FISH MEAL PRODUCTION
Fish scrap from the canning lines, including
any rejected -whole fish, is charged to contin-
uous live-steam cookers in the meal plant.
Flow through a. typical fish meal plant is dia-
grammed in Figure 556. Cookers of the type
shown in Figure 557 are operated at bet-ween
2 and 5 psig steam pressure. Material charged
to the cookers normally contains 20 30 percent
solids. Cooked scrap has a slightly smaller
solids content owing to the condensed steam
picked up in cooking. After the material
leaves the cooker it is pressed to remove
oil and water, and this pressing lowers the
moisture content of the press cake to approxi-
mately 50 percent. The press cake is broken up,
usually in a hammer mill, and dried in a direct-
fired rotary drier or in a steam-tube rotary
drier. Typical fish meal driers yield 2 to 10
tons of meal per hour with a moisture content
of 4 to 10 percent. Both types of driers em-
-------
Fish Canneries and Fish Reduction Plants
763
FISH SCRAP
VAPORS TO CONDENSER
LIVE STEAM COOKER
STEAM 5 psig
WATER AND SOLUBLES
FISH OIL
PRESS
PRESS / \ PRESS
CAKE/ \WATER
/ X
DRIER GASES TO
Figure 556. Flow diagram of a fish meal reduction system including
oil-separating and oil-clarifying equipment.
Figure 557. A live-steam reduction cooker and a
continuous press (Standard Steel Corp., Los Angeles,
Cal if.).
ploy air as the drying medium. Moisture is
removed with exhaust gases, which are volu-
minous.
Direct-fired driers include stationary fireboxes
ahead of the rotating section, as shown in Fig-
ure 558. They are normally fired with natural
gas or fuel oil. Combustion is completed in
the firebox. Hot products of combustion are
mixed with air to provide a temperature of 400°
to 1, 000°F at the point where wet meal is initial-
ly contacted. Hot, moist exhaust gases from
the drier contain appreciable fine meal, which
is commonly collected in a cyclone separator.
The essential feature of steamtube driers is
a bank of longitudinal, rotating steamtubes
arranged in a cylindrical pattern, as shown in
Figure 559. Steam pressures range from 50
to 100 psig in the tubes. Heat is transferred
both to the meal and air. As with direct-fired
units, gases pass parallel to meal along the
axis of the drier and are vented through a cy-
clone separator. Meal produced in steamtube
driers is less likely to be over-heated and is
generally of higher quality than that from direct-
fired units.
FISH SOLUBLES AND FISH OIL PRODUCTION
Fish solubles is the term used to designate the
molasses-like concentrate containing soluble
proteins and vitamins that have been extracted
from fish flesh by cooking processes. The
flow diagram of Figure 556 includes the sep-
aration of press water and fish oil. The sources
of solubles and oils are the juices and conden-
sate collected as press water and stick water.
-------
764
CHEMICAL PROCESSING EQUIPMENT
EXHAUST
1 V
VSPOR AND PRODUCT
SEPARATOR
Figure 558. A parailei-flow, direct-fired, rotary, fish meal drier
(Standard Steel Corp., Los Angeles, Calif.).
Figure 559. A steamtube, rotary, fish meal drier (Standard Steel Corp.,
Los Angeles, Calif.).
-------
Fish Canneries and Fish Reduction Plants
765
These two liquids may be processed separately
or blended before their processing. The liq-
uids are first acidified to prevent bacterial de-
composition. Some protein is flocculated up-
on the addition of acid. The floe and other
suspended solids are removed in a centrifuge
and recycled to the fish meal reduction process.
Liquids pass through a second centrifuge, where
the fish oils are removed. The water layer is
pumped to multiple-effect evaporators where
the solids content is increased from approxi-
mately 6 to 50 percent by weight. Uncondensed
gases are removed from the process at one of
the evaporator effects, which is operated under
high vacuum. The vacuum is held by a water
or steam ejector. "Where steam ejectors are
used they are equipped with barometric-leg
aftercondensers.
DIGESTION PROCESSES
Fish viscera are usually digested by enzymatic
and bacterial action rather than by thermal re-
duction. The product is a liquid that is concen-
trated by evaporation and marketed as a high-
protein livestock feed supplement very similar
to fish solubles.
Most cannery-operated digestion processes are
of the enzymatic type and are used only to pro-
cess viscera. Stomach enzymes, under con-
trolled pH and temperature, reduce the viscera
to a liquid. The process is usually carried out
in a simple tank at atmospheric pressure, near-
ambient temperature, and an acid pH. Essen-
tially no moisture is evaporated during digestion.
Before concentration, the digested liquid is fil-
tered and centrifuged to remove small quantities
of scales, bones, and oil. The evaporation pro-
cess is identical to that used for fish solubles,
yielding a liquid of 50 percent solids.
Bacterial digestion is used to reduce all types
of fish flesh. It is carried out at an alkaline
pH in equipment similar to that used for en-
zymatic processes. Again, there is no appre-
ciable moisture evaporation, but odors evolved
are considerably stronger and more likely to
elicit nuisance complaints.
THE AIR POLLUTION PROBLEM
Air contaminants emanate from a number of
sources in fish canneries and fish reduction
plants, including both edible-rendering and
byproduct processes. Odors are the most ob-
jectionable of these contaminants, though dust
and smoke can be a major problem. In a fish
cannery, some odor is unavoidable owing to
the nature of the species. Heavy odor emis-
sions that cause nuisance complaints can usu-
ally, however, be traced to poor sanitation or
inadequate control of air contaminants. Tri-
methyl amine, (CH^J^N, is the principal com-
pound identified with fish odors.
Reduction processes produce more odors than
cannery operations do. Materials fed to re-
duction processes are generally in a greater
state of decay than the fish are that are pro-
cessed for human consumption. Edible por-
tions of the fish are always handled first, and
great care is maintained to guarantee the qual-
ity of edible products. The portions that are
unsuitable for human consumption have much
less value, and it is not uncommon for opera-
tors to allow reduction plant feedstocks to
decompose markedly before the processing.
The largest sources of reduction plant odors
are fish meal driers. Lesser quantities of
odors are emitted from cookers preceding
meal driers, from digestion processes, oil-
water separators, and evaporators. Dust
emissions are limited to driers and the pneu-
matic conveyors and grinders following them.
Smoke can be created by overheating or burn-
ing meal in the drier.
Odors From Meal Driers
Fish meal driers exhaust large volumes of gases
at significantly large odor concentrations. Dur-
ing the processing of fresh fish scrap, odor con-
centrations in exhaust gases range from 1, 000 to
5, 000 odor units per scf (see Appendix for defini-
tion of odor units and method of measuring odor
concentrations). If the feedstocks are highly de-
cayed, much greater odor concentrations can
be expected. The result is an extremely heavy
rate of odor emission, even when fresh fish
scrap is processed. For example, a direct-
fired drier producing 5 tons of dried fish meal
per hour exhausts about 44 million odor units
per minute if the concentration is 2, 000 odor
units per scf, and the exhaust rate is 22, 000
scfm. Drier exit temperatures average about
200°F, and moisture content normally ranges
between 15 and 25 percent by volume.
Emissions from steamtube driers are less
voluminous and can be less odorous than those
from direct-fired units. With steamtube driers,
there is less likelihood of burning or overheat-
ing the meal and, therefore, excessively heavy
odor concentrations are encountered less often.
Moisture contents are comparatively greater
in gases from steamtube driers. Typical gases
from emitted steamtube driers during tuna scrap
processing contain about 25 percent moisture as
compared with approximately 15 percent from
-------
766
CHEMICAL PROCESSING EQUIPMENT
a direct-fired unit processing the same material.
As a result, volumes from steamtube driers are
30 to 45 percent lower than those from compar-
able direct-fired units. Odor concentrations
from steamtube driers are generally in the same
range as those from direct-fired units when
fresh fish scrap is being processed under prop-
er operating conditions, that is, when meal is
not overheated.
Smoke From Driers
Excessive visible air contaminants can be cre-
ated in fish meal driers by the overheating of
meal and volatilization of low-boiling oils and
other organic compounds. Smoke is more likely
to be emitted from direct-fired driers than from
steamtube units, particularly if flames are allowed
to impinge directly on the meal. All driers have
limits for gas discharge temperature above which
excessive visible contaminants appear in the exit
gas stream. For direct-fired units, this limit
is about 190°F for tuna scrap and about 215°F
for wet-fish scrap. The smoking limit is a
function of drier design as well as of feedstocks
and varies somewhat from unit to unit.
The addition of certain low-boiling materials to
drier feedstocks can also create visible emis-
sions when there is essentially no overheating
of meal in the drier. One such material is di-
gested fish concentrate. Some operators add
this high-protein liquid to drier feedstocks to
upgrade the protein content of meal. Digested
fish concentrate can contain low-boiling com-
pounds that are vaporized into exhaust gases and
condense upon discharge to the atmosphere.
These finely divided, organic, liquid particulates
can impart greater than 40 percent opacities to
drier gases. Scrubbing the drier gases with
water aggravates the problem by lowering the
temperature, which increases condensation and,
thereby, the opacity.
Dust From Driers and Conveyors
The only major points of dust emission in can-
neries and reduction plants are the driers them-
selves and the grinders and conveyors used to
handle dried fish meal. Driers and pneumatic
conveyors are equipped with cyclone separators,
and emissions are functions of collection effi-
ciencies.
Fish meal does not usually contain a large frac-
tion of fines. A particle size analysis of a typ-
ical meal is provided in Table 223. This meal
sample was collected in a pneumatic conveyor
handling ground fish meal. It can be seen that
the sample contains only 0. 6 percent by weight
less than 5 microns in diameter, and 1.4 per-
Table 223. PARTICLE SIZE ANALYSIS OF
A TYPICAL GROUND,
DRIED FISH MEALa
Range of particle diameter,
0
5
10
20
44
74
149
246
590
1, 651
more
to 5
to 10
to 20
to 44
to 74
to 149
to 246
to 590
to 1,651
to 2, 450
than 2, 450
wt %
0.6
0.8
2.6
7.5
11. 5
29.9
16.4
22. 8
7.4
0.4
0. 1
aSample drawn from a pneumatic conveyor
following a direct-fired drier and hammer
mill.
Size determination by micromerograph.
cent less than 10 microns in diameter. Ninety-
six percent is larger than 20 microns.
Concentrations of fines in exit gases are usually
less than 0.4 grain per scf. The pneumatic con-
veyor cyclone handling the meal of Table 223 was
found to be better than 99. 9 percent efficient,
with an exit dust concentration of less than 0. 01
grain per scf. This efficiency is much greater
than would be predicted on the basis of cyclone
design and particle size. It indicates that ap-
preciable agglomeration probably takes place
in the cyclone.
Odors From Reduction Cookers
The cookers preceding fish meal driers exhaust
gases of heavy odor concentration. Nevertheless,
the volumes of these offgases are appreciably less
than those from driers. Cooker gases are simi-
lar to those from indirectly heated rendering
cookers. They consist almost entirely of water
vapor but contain significant quantities of ex-
tremely odorous organic gases and vapors. Odor
concentrations from live-steam-heated cookers
range from 5, 000 to over 100,000 odor units per
scf, depending to a large degree upon the state
of feedstocks. Any malodorous gases contained
in the cellular flesh structure are usually liberated
when the material is first heated in the cooker.
Essentially no solids are in the effluent from the
cookers, though some entrained oil particulates
are usually present. The volumes of exhaust
vapors depend upon the degree of sealing provided
in the cooker. All the steam can be contained in
the cooker with no leakage. Most cookers, how-
-------
Fish Canneries and Fish Reduction Plants
767
ever, are designed to bleed off 100 to 1, 000 cfm
through one or more stacks. The latter arrange-
ment is recommended, since it provides a posi-
tive exhaust point at which air contaminants can
be controlled. Otherwise the malodorous gases
would be liberated at the press and grinder where
they are difficult to contain.
Odors From Digesters
The digestion of fish scrap produces only small
volumes of exhaust gases, though these gases
can have a large odor concentration. The en-
zymatic, acid-pH decomposition of viscera
does not normally produce odor concentrations
greater than 20, 000 odor units per scf, depend-
ing again upon the quality of feedstocks. Alka-
line digestion of fish scrap, on the other hand,
is productive of strong odors that are likely to
create a public nuisance.
Odors From Evaporators
The evaporation of the water-soluble extracts--
stick water and press water—does not generally
result in heavy odor emissions. This is pri-
marily due to the use of water ejector-condensers.
Odors could be considerably heavier if different
types of vacuum-producing equipment were em-
ployed. Most fish canneries are located near
large bodies of water,and it is common to use
water jet ejectors to maintain a vacuum on the
evaporator system. Alluncondensed gases and
vapors from the evaporators are vented to the
ejectors, which act as contact condensers. Most
of the odorous compounds are condensed or dis-
solved in the effluent water. If steam ejectors
and surface condensers, rather than contact
condensers, are used to produce rhe vacuum,
odor emissions to the atmosphere are much
greater. Contact condensers (water ejectors)
provide a dilution of condensate 10 to 20 times
greater than that produced by surface-type con-
densers used with steam ejectors or vacuum
pumps.
The hot-exhaust boxes of wet-fish production
systems are commonly vented directly to the
atmosphere. These offgases consist mostly of
steam with some noncondensible air and mal-
odorous gases entrained. Hot-exhaust boxes
are the points of initial cooking of wet fish, and
are, therefore, origins of large quantities of
gases and vapors.
HOODING AND VENTILATION REQUIREMENTS
When air pollution control is employed, most
fish cannery and reduction processes are vented
directly to the control device. The only equip-
ment requiring hooding are the presses and
grinders intermediate between cookers and
driers in a fish meal system. Hot material
from the cooker evolves appreciable steam and
odors when the oil and water are pressed from
it and when the resultant press cake is broken
up before the drying. The vapors liberated at
these points consist principally of steam. When
the gases are vented to a condenser, hooding
should be as tight as possible to prevent dilu-
tion with air. Indraft velocities of 100 fpm
across the open area under the hood are nor-
mally satisfactory. Where possible, the source
itself should be totally enclosed and ducted to
control equipment. Unfortunately, the designs
of many presses and grinders are not conducive
to complete enclosure, and hoods must be em-
ployed.
The largest contaminated gas streams are ex-
hausted from fish meal driers. As shown in
Table 224, volume rates are lower from steam-
tube driers than from direct-fired units. For
the hypothetical comparison made in this table,
the fired drier exhausts 70 percent more gases
than the steamtube drier does and the moisture
content is comparatively lesser, 16. 1 against
25 percent. A 10-ton-per-hour fired drier
would exhaust 22,830 scfm at about 200°F,
while a steamtube unit of the same size would
exhaust only 13,500 scfm at about 180°F.
Odors From Edibles Cookers
While most odorous air contaminants are con-
sidered to emanate from fish reduction processes,
the handling and cooking of edible fish also pro-
duce measurable odors. The largest single sources
are the cookers described earlier in this section.
The precooked process is less productive of odors
than the wet-fish process is. When tuna is cooked
in the live-steam cookers of Figure 554, much of
the odorous gases and vapors is condensed in the
cooker and the steam trap. Only the volatile,
albeit highly odorous compounds are vented through
the steamtrap.
Table 224. CHARACTERISTICS OF EXHAUST
GASES FROM TYPICAL DIRECT-FIRED AND
STEAMTUBE FISH MEAL DRIERSa
j Steamtube
Moisture evaporated from meal, scfm^
Natural gas fuel, scfmc
Moisture in products of combustion, scfm
Total moisture in exhaust gases, scfm
Dry exhaust gases, scfm
Total exhaust gases, scfm
Moisture content, % by volume
Temperature of exhaust gases, °F
drier
333
-
-
338
1,012
1, 350
25
ISO
Direct-fired
drier
338
14
31
369
1, 914
2,283
16. 1
205
Basis 1 ton of feed per hour to drier. Moisture content of press
cake to drier, 50% by weight
"Moisture content of dried meal, 4% by weight.
GNatural gas of i, 100 Btu per scf gross heating value.
-------
768
CHEMICAL PROCESSING EQUIPMENT
Exhaust volumes from live-steam-heated cook-
ers range from 100 to 1, 000 cfm and depend
to a large degree upon cooker design. Inlet
and exit seals should be tight to prevent leakage.
Most cookers are vented through a single stack.
Digestion tanks with a capacity of 2,000 gallons
or less seldom exhaust more than 50 scfm. Ex-
haust volumes from digesters vary appreciably
during the processing of a batch, exit rates being
negligible much of the time.
Where water ejector contact condensers are em-
ployed on evaporators, exhaust rates are well
below 50 scfm. If surface condensers or vacuum
pumps are employed instead of contact condensers,
exhaust volumes can exceed 100 cfm.
Fish meal pneumatic conveyors are designed to
provide from 45 to 70 cubic feet of air per pound
of meal conveyed. A pneumatic conveyor handling
5 tons of dried meal per hour exhausts about
10, 000 cfm.
Exhaust gases from cookers used in the precooked
tuna process are relatively small in volume and
include only those gases that are not condensed or
dissolved at the steamtrap or the cooker itself.
Gases evolved from the hot-exhaust boxes of the
wet-fish lines are considerably more voluminous.
AIR POLLUTION CONTROL EQUIPMENT
Fish cannery and fish reduction equipment are
controlled principally with condensers, scrub-
bers, afterburners, and centrifugal dust col-
lectors. Where odors are concerned, incinera-
tion is preferable if it can be adapted to the pro-
cess. Incineration provides the most positive
control of nuisance-causing odorous compounds.
Condensers are effective 'where exhaust gases
contain appreciable moisture, while centrifugal
collectors are usually satisfactory to prevent
excessive dust emissions. Scrubber-chlorina-
tors find particular use in the control of odors
from fish meal driers.
Incinerating Drier Gases
Incineration of odorous air contaminants from
fish meal driers is possible, though costly. A
properly designed afterburner control system
requires a dust collector ahead of the afterburner
to remove solids that cannot readily be burned.
The incineration of solid particulates at 1, 200°F
or lov/er can result in partial oxidation of partic-
ulates, which tends to increase rather than de-
crease odor concentrations. A contact condenser-
scrubber removes much of the difficult-to-burn
particulates and materially reduces the volume
rate by condensing the moisture. If the partic-
ulate matter concentration in gases to the after-
burner is sufficiently small, incineration at
1,200°F reduces odor concentrations to about
50 odor units per scf. Owing to the high cost
of fuel in such an arrangement, few large in-
stallations of afterburners serve fish meal
driers. To make incineration economically at-
tractive, heat from the afterburner should be
reclaimed in some manner. The most likely
arrangement is the preheating of air to the drier.
An afterburner operating at 1,200°F provides
all the heat necessary to operate the drier, which
thus eliminates the need for a firebox.
Chlorinating and Scrubbing Drier Gases
A unique scrubber-chlorinator design has been
developed to control satisfactorily the odors
from fish meal driers. This unit is demon-
strated in the flow diagram of Figure 560 and
pictured in Figure 561. The process depends
largely upon the reaction of chlorine gas with
odorous compounds at drier exit temperatures.
As shown in Figure 560, gases from the drier
are first directed through a cyclone separator
to remove fine particulates. Chlorine is then
added at a rate calculated to provide a concen-
tration of 20 ppm by volume in the gas stream.
The reaction is allowed to proceed at about
200°F--the drier exit temperature--in the duct-
work for approximately 0. 6 second before being
chilled and scrubbed with sea water in a packed
tower. Gases pass up through the packing coun-
tercurrently to the sea water.
Controlling Fish Meal Driers
Because of the exceedingly large volume of mal-
odorous exhaust products from driers, they
constitute the most costly air pollution control
problem in a reduction plant. Drier gases nor-
mally contain only 15 to 25 percent moisture.
Thus, even after condensation, the volume is
great. Moreover, there are enough entrained
solids in drier exit gases to make incineration
difficult.
In Figure 562, odor concentrations from the
scrubber exit are plotted against the chlorine
addition rate at constant g£.s and sea water
throughput. As can be seen from the curve,
odors reach a minimum at about 20 ppm chlo-
rine. When more than 20 ppm are added, chlo-
rine odors become readily detectable in treated
gases, and odor concentrations tend to increase.
All the odor measurements used to draw this
curve were made on drier gas samples taken
between 170° and 205 °F, when there was es-
sentially no overheating of meal in the drier.
-------
Fish Canneries and Fish Reduction Plants
769
CHLORINE GAS
VENTURI—*
TC3k
TO ATMOSPHERE ^_
Figure 560. A chlorinator-scrubber odor control system venting a fish
meal drier.
Figure 561. A cnlorinator-scrubber odor control
system venting a fish meal drier (Star-Kist
Foods, Inc., Terminal Island, Calif.).
350
SAMPLES WERE COLLECTED WHEN DRIER DISCHARGE
TEWEftftTURES *ERE BEIO" 205°F AT HIGHEB
TEMPERATURES ODOR LEVELS INCREASE MARKEDLY
REGARDLESS OF CHLORINE CONCENTRATION
5 10 '5 3 K
CHLORINE GAS ADDITION RATE, ppm by volume
Figure 562. Exit odor concentrations from a chlo-
rinator-scrubner as a function of the chlorine gas
addition rate. Temperatures of gas discharged from
drier are less than 205°F.
-------
770
CHEMICAL PROCESSING EQUIPMENT
This method provides an overall odor reduction
of 95 to 99 percent when fresh fish scrap is being
processed in the drier. Chlorination itself pro-
vides a 50 to 80 percent reduction in odor con-
centration. Scrubbing reduces the remaining
odor concentration by another 50 to 80 percent.
Condensation provides a 12 to 22 percent re-
duction in volume, depending upon the original
moisture content of the gases.
The exact mechanism of the chlorination reac-
tion is uncertain, but it is assumed that chlorine
reacts with odorous compounds, probably amines,
to form additional products that are less odorous
than the original compounds. Chlorine is not
considered to be a sufficiently strong oxidizing
agent to oxidize fully the odorous organic mate-
rials present in drier gases.
Controlling Reduction Cookers
Inasmuch as cooker emissions consist primarily
of steam, they can be controlled with condensers
and secondary controls if necessary. A contact
condenser operating at 100°F or a lower effluent
temperature can remove a major portion of cook-
er odors. If noncondensable gases from the con-
denser are large in volume, they can be directed
to an afterburner, a carbon adsorber, or a chlo-
rinator-scrubber. Normally, there is little
entrained air or other noncondensable gases in
cooker vapors.
Controlling Digesters
Digester gases are most easily controlled with
afterburners. These gases are small in volume
and require only minimal fuel for incineration.
Digester offgases contain no appreciable moisture
or particulates. Odor concentrations can normal-
ly be reduced by 99 percent or more at 1, 200°F
in a properly designed afterburner.
Controlling Evaporators
Evaporators for stick water, press water, and
digested liquor can be controlled with condensers
and afterburners and combinations thereof. Most
evaporators are equipped with water ejector con-
tact condensers to provide the necessary vacuum
in the one effect of the multiple evaporator effects.
Condensate temperatures from these ejectors are
usually less than 80°F. As a result, they con-
dense and dissolve most of the odorous compounds
that would otherwise be discharged to the atmo-
sphere. Condensate cannot be circulated through
cooling towers without causing the emission of
strong odors. Ideally, sea water or harbor
water is used for this purpose, -with no recircu-
lation. The entrained air contaminants do not
add enough material to tail waters to create a
water pollution problem.
If water ejectors are not used, odorous air con-
taminants are emitted in much heavier concen-
tration. The most likely alternative is a steam
ejector and surface-type aftercondenser, possi-
bly with multiple ejector stages. Noxious odors
from an operation such as this are stronger and
more voluminous than those emitted from contact
condensers. An afterburner operating at 1,200°F
or greater is usually the most practical means
of controlling these processes. Activated carbon
can be used in lieu of an afterburner.
Collecting Dust
As previously noted, fish meal does not contain
a large amount of extremely fine particles, that
is, those less than 10 microns. For this reason,
cyclone separators are normally sufficient to
prevent excessive emissions from the drier and
subsequent pneumatic conveyors. If the meal
from a particular plant were to contain appre-
ciably more fine material than the sample shown
in Table 223, more efficient dust collectors,
such as small-diameter, multiple cyclones or
baghouses, would have to be used.
Controlling Edible-Fish Cookers
Exhaust gases from both precooked and wet-fish
process cookers consist essentially of water
vapor. At tuna cookers, most of this vapor is
condensed in the steamtraps on the cookers. If
further control is desired, an afterburner,
carbon adsorber, or low-temperature contact
condenser is recommended.
The hot-exhaust boxes of wet-fish processes
represent large odor sources that can be con-
trolled 'with contact condensers, often at little
expense to the operator. Most canneries are
located near large bodies of water. Sea water
or harbor water can be directed to contact con-
densers at little cost in these instances. Since
exhaust box gases are principally water, there
is a marked reduction in volume across a con-
denser such as this, in addition to a decrease
in odor concentration.
REDUCTION OF INEDIBLE
ANIMAL MATTER
Animal matter not suitable as food for humans
or pets is converted into salable byproducts
through various reduction processes. Animal
matter reduction is the principal waste disposal
outlet for slaughterhouses, butcher shops,
poultry dressers, and other processors of
GPO 8O6—614—26
-------
Reduction of Inedible Animal Matter
771
flesh foods. In addition, it is used to dispose
of whole animals such as cows, horses, sheep,
poultry, dogs, and cats that have died through
natural or accidental causes. If it were not
for reduction facilities, these remains would
have to be buried to prevent a serious health
hazard. The principal products of reduction
processes are proteinaceous meals, which find
primary use as poultry and livestock feeds, and
tallow.
Much reduction equipment is operated in meat-
packing plants to handle only the "captive" blood,
meat, and bone scrap offal produced on the
premises. Other reduction cookers and driers
are located in scavenger rendering plants,
which are operated solely for the byproducts.
In Figure 563, "dead stock" is shown awaiting
dismemberment at a scavenger plant. Com-
mon rendering cooker feedstocks are pictured
in Figure 564. In general, the materials pro-
cessed in captive packing kouse systems are
fresher than those handled at scavenger plants
where feedstocks can be highly decayed. Typ-
ical slaughterhouse yields of inedible offal,
bone, and blood are listed in Table 225.
The animal matter reduction industry has been
traditionally considered one of the "offensive
trades." The reputation is not undeserved.
Raw materials and process exhaust gases are
highly malodorous and capable of eliciting nui-
sance complaints in surrounding areas. In
Figure 564. Inedible animal matter in the receiving
pit of a rendering system (California Rendering Co.,
Ltd., Los Angeles, Cal if.).
Table 225. INEDIBLE, REDUCTION PROCESS
RAW MATERIALS ORIGINATING
FROM SLAUGHTERHOUSES
(The Globe Co., Chicago, 111.)
Source, Ib live wt
Steers, 1, 000
Cows
Calves, 200
Sheep, 80
Hogs, 200
Inedible offal and bone,
Ib/head
90 to 100
110 to 125
15 to 20
8 to 10
10 to 15
Blood,
Ib/head
55
--
5
4
7
Figure 563. Dead stock awaiting skinning and dis-
memberment at a scavenger rendering plant (Califor-
nia Rendering Co..Ltd., Los Angeles, Calif.).
recognition of these facts, specific air pollution
control regulations have been enacted requiring
the control of odorous process vapors.
Rendering, itself, is a specific, heated reduc-
tion process wherein fat-containing materials
are reduced to tallow and proteinaceous meal.
Blood drying, feather cooking, and grease re-
claiming are other reduction operations usually
performed as companion processes in render-
ing plants.
Reduction processes are influenced largely by
the makeup of feedstocks. As can be seen from
Table 226, some materials, such as blood and
feathers, are essentially grease free, while
others contain more than 30 percent tallow.
-------
772
CHEMICAL PROCESSING EQUIPMENT
Table 226. COMPOSITION OF TYPICAL INEDIBLE RAW MATERIALS
CHARGED TO REDUCTION PROCESSES
(The Globe Co., Chicago, 111.)
Source
Packing house offal and bone
Steers
Cows
Calves
Sheep
Hogs
Dead stock (whole animals)
Cattle
Cows
Sheep
Hogs
Blood
Feathers (from poultry houses)
Butcher shop scrap
Tallow or grease,
wt %
15 to 20
10 to 20
8 to 12
25 to 35
15 to 20
12
8 to 10
22
30
-
-
37
Solids ,
wt %
30 to 35
20 to 30
20 to 25
20 to 25
18 to 25
25
23
25
25 to 30
12 to 13
20 to 30
25
Moisture,
wt %
45 to 55
50 to 70
60 to 70
45 to 55
55 to 67
63
67 to 69
53
40 to 45
87 to 88
70 to 80
38
Where no tallow is present, the reduction pro-
cess becomes primarily evaporation with, possi-
bly, some thermal digestion.
DRY RENDERING
The most widely used reduction process is dry
rendering, wherein materials containing tallow-
are heated indirectly, usually in a steam-jac-
keted vessel. Heat breaks down the flesh and
bone structure, allowing tallow to separate
from solids and water. In the process, most
of the moisture is evaporated. Emissions con-
sist essentially of steam with small quantities
of entrained tallow, solids, and gases.
Dry rendering may be performed batchwise or
continuously and may be accomplished at pres-
sures greater or less than atmospheric. A
typical batch-type, steam-jacketed, dry render-
ing cooker is shown in Figure 565. These ves-
sels are normally charged with 3, 000 to 10, 000
pounds of animal matter per batch. The cookers
are equipped with longitudinal agitators that are
driven at 25 to 65 rpm. Each batch is cooked
for 3/4 to 4 hours.
Pressures of 50 psig and greater are used to
digest bones, hooves, hides, and hair. At the
resulting temperature (about 300°F), these
materials are reduced to a pulpy mass. In typ-
ical dry-pressure-rendering cycles, the cooker
vent is initially closed to cause pressure and
temperature to increase. Some materials are
cooked as long as 2 hours at elevated pressure
to obtain the necessary digestion. After pres-
sures are reduced, the batch is cooked or dried
to remove additional moisture and to complete
tallow-solids separation.
Some dry rendering operations are carried out
under vacuum to remove moisture rapidly at
temperatures sufficiently low to inhibit degrada-
tion of products. Vacuum, rendering processes
are essentially all of the batch type. The vacu-
um is usually produced with a precondenser,
steam ejector, and aftercondenser. Cooker
pressures are close to atmospheric at the start,
then diminish markedly as the moisture content
of the charge decreases. Vacuum rendering
produces high-quality tallow but has a disadvan-
tage in that temperatures are low and incomplete
cooking of bones, hair, and so forth, may occur.
Highly mechanized, continuous, dry rendering
processes are in use in some parts of the United
States. Many processes consist essentially of
a series of grinders, steam-jacketed conveyor-
cookers, and presses. Animal matter is ground
before it is fed to a precooker. After the initial
cook, the material is again ground before its
final processing in the second-stage cooker.
Tallow and steam vapors are removed from
solids at various points in the system. Cooked
material from the second stage is pressed to
remove residual tallow. The continuous sys-
tem of Figure 566 is unique in that it uses re-
cycled tallow, and a vertical-tube vacuum cooker.
Selected meat and bone scrap is ground and
slurried with hot tallow before being charged to
the cooker. Slurry is circulated through the
tubes, and vapors are vented to a contact con-
denser. Steam is condensed ahead of the ejec-
-------
Reduction of Inedible Animal Matter
773
p. y
-CZ3
L
^^
b L:
u c
«#^
-i rl
n 0
^ v-
u
tor, and a barometric leg is employed. Tallow
and solids are continuously drawn from the
bottom of the cooker.
WET RENDERING
One of the oldest reduction methods is the
process, wherein animal matter is cooked in a
closed vessel with live steam. There is little
evolution of steam. Most of the contained mois-
ture is removed as a liquid. Live steam is fed
to a charge in a closed, vertical kettle until the
internal pressure reaches approximately 60 psig
(about 307°F). Heat causes a phase separation
of water, tallow, and solids. After initial cook-
ing, the pressure is released, and some steam
is flashed from the system. The charge is then
cooked at atmospheric pressure until tallow
Figure 565. A horizontal, batch-type, dry-rendering
cooker equipped with a charging elevator (Standard
Steel Corp., Los Angeles, Calif.).
separation is complete. Water, tallow, and
solids are separated by settling, pressing, and
centrifuging.
The water layer from a wet rendering process
contains 6 to 7 percent solids. Soluble proteins
can be recovered by evaporation, as in the pro-
cessing of stick water at fish reduction plants.
Wet rendering finds some use today in the han-
dling of dead stock, namely whole animals that
have died through accidents or natural causes.
It has given way to dry rendering at most pack-
ing houses and scavenger plants. Wet rendering
is used to a limited degree in the production of
edible fats and oils, as noted previously in this
chapter.
REFINING RENDERED PRODUCTS
At the completion of the cook cycle, tallow and
solids are run through a series of separation
equipment as in the integrated plant of Figure
567. Some systems are more complex than
others, but the essential purpose is to produce
dry, proteinaceous cracklings and clear, mois-
ture-free tallow. In almost all cases, the cook-
ers are discharged into perforated percolator
-------
774
CHEMICAL PROCESSING EQUIPMENT
HATER SUPPLY
BAROMETRIC CONDENSER
DISINTEGRATOR
RAH MATERIAL
TRAMP METAL DISCHARGE
EXPELLED
FAT PUMP
WATER DISCHARGE
Figure 566. A continuous, vacuum rendering system employing tallow recycl ing (Carver-Greenfield
Process, The V.D. Anderson Co., Cleveland, Ohio).
EXHAUST VAPORS TO CONTROL EQUIPMENT
MEAL
CRACKLINGS TROUGH »ITH SCREtf CONVEYOR
Figure 567 An integrated dry rendering plant equipped with batch
cookers, percolators, a cracklings press, and a tallow-settling tank.
-------
Reduction of Inedible Animal Matter
775
pans that allow free-running tallow to drain from
hot solids. The remaining solids are pressed to
remove residual tallow. Dry cracklings are usu-
ally ground to a meal before being marketed. In
Figure 568, grease-laden cracklings are being
dumped from a percolator pan after free tallow
has been drained.
Figure 568. Tallow-laden cracKlings being dumped
from a percolator after free tallow has been allowed
to drain (California Rendering Co., Ltd., Los Angeles,
Ca I i fornia).
Tallow from the percolators and presses is
further treated to remove minor quantities of
solids and water. Solids may be removed in
desludging centrifuges, filters, or settling tanks.
Traces of moisture are often removed from it
by boiling or blowing air through heated tallow.
Some operators remove moisture by settling in
cone-bottom tanks, often with the aid of soda
ash or sulfuric acid to provide better phase
separation.
In some instances, solvents are used to extract
tallow from rendered solids. Solvent extraction
allows extremely fine control of products. The
Belgian De Smet process, in which hexane is
employed, has been adopted by some Tenderers
in the United States and Canada. The entire
process is enclosed in a vaportight building to
minimize the explosion hazard. After extrac-
tion, hexane is stripped from tallow and solids.
The only measurable air contaminants, solvent
vapors, are vented at one or more condensers.
DRYING BLOOD
Animal blood is evaporated and thermally di-
gested to produce a dry meal used as a fertiliz-
er, as a livestock feed supplement, and, to a
limited degree, as a glue. Blood contains only
10 to 15 percent solids and essentially no fat.
At most packing houses, it is dried in horizontal,
dry rendering cookers. In typical slaughtering
operations, blood is continually drained from.
the kill floor to one or more cookers, throughout
the day. Initially, while there is appreciable
moisture in the blood, heat transfer through the
jacket is reasonably rapid. As the moisture
content decreases, however, heat transfer be-
comes slower. During the final portion of the
cycle, drying is extremely slow, and dusty meal
can be entrained in exit gases.
In some instances, a tubular evaporator is used
to remove the initial portion of the water. When
the moisture content decreases to about 65 per-
cent, the material is transferred to a dry ren-
dering cooker for final evaporation.
Some animal blood is spray dried to produce
a plywood glue that commands a price con-
siderably higher than that of fertilizer or live-
stock feed. This is an air-drying process, and
exhaust gases a-re markedly more voluminous
than those of rendering equipment. Feedstocks
are usually concentrated in an evaporator be-
fore the spray drying.
PROCESSING FEATHERS
Poultry feathers are pressure cooked and sub-
sequently dried to produce a high-protein meal
used principally as a poultry feed supplement.
Feathers, like blood, contain practically no fat,
and meal is the only product of the system.
Feathers are pressure cooked at about 50 psig
to hydrolyze the protein keratin, their principal
constituent. Initial cooking is usually carried
out in a dry rendering cooker. Final moisture
removal may be accomplished in the cooker at
ambient pressure or in separate air-drying
equipment. Rotary steamtube air driers, such
as that shown in Figure 569, are frequently
used for this purpose. If separate driers are
employed, the material is transferred from
cooker to drier at a moisture content of about
50 percent.
ROTARY AIR DRIERS
Direct-fired rotary driers are seldom used in
the reduction of inedible packing house •waste
or dead stock. As noted previously in this
chapter, they find wide use in the reduction of
fish scrap. Fired driers have been used to a
-------
776
CHEMICAL PROCESSING EQUIPMENT
SIDE ELEVATION SHOWING
ARRANGEMENT OF TUBES
MATERIAL DISCHARGE
FRONT ELEVATION SHOWING STEAM FLOW
VARIABLE FEEDER-OPTIONAL - USED
-WITH PARALLEL DRYER SYSTEMS
\
MATERIAL INLET
SIR EXHAUST
CONOENSATE
OUTLET
MATERIAL DISCHARGE
INNER PIPE
.___ OUTER PIPE
CUTAWAY OF PIPES
Figure 569. A rotary steamtube air drier of the type commonly
used for the continuous drying of cooked feathers (The V.D.
Anderson Co., Cleveland, Ohio)
limited degree to dry wet rendering tankage
and some materials of low tallow content.
Where air driers are required, steamtube
units are generally more satisfactory from the
standpoint of both product quality and odor emis-
THE AIR POLLUTION PROBLEM
Malodors are the principal air contaminants
emitted from inedible-rendering equipment and
from other heated animal matter reduction pro-
cesses. Reduction plant odors emanate from
the handling and storage of raw materials and
products as well as from heated reduction pro-
cesses. Some feed materials are highly decayed,
even before delivery to sca.veiiger rendering
plants, and the grinding, conveying, and storage
of these materials cannot help but generate some
malodors. Cooking and drying processes are,
nevertheless, considered the largest odor sources,
and most odor control programs have been di-
rected at them. Handling and storage odors can
usually be kept to a tolerable minimum by fre-
quently washing working surfaces and by pro-
cessing uncooked feedstocks as rapidly as possible
McCord and Witheridge (1949), who discuss
the "offensive trades" at length, attribute
rendering plant malodors to a variety of com-
-------
Reduction of Inedible Animal Matter
777
pounds. Ronald (1935) identifies rendering
odors as principally ammonia, ethylamines,
and hydrogen sulfide, all decomposition products
of proteins. Skatole, other amines, sulfides,
and mercaptans are also usually present. Tallow
and fats do not generate as great quantities of
odorous materials. Aldehydes, organic acids,
and other partial oxidation products are, the
principal odorous breakdown products of fats.
Putrescine, NH£ (CP^^Nr^, and cadaverine,
NH2(CH2)5NH2, are two extremely malodorous
diamines associated with decaying flesh and
rendering plants. Several specific compounds
have extremely low odor thresholds and are de-
tectable in concentrations as small as 10 parts
per billion (ppb). Odor threshold concentra-
tions of some pertinent compounds are listed
in Table 227. Many suspected compounds
have not been positively identified nor have
their odor thresholds been determined.
Cookers As Prominent Odor Sources
When animal matter is subjected to heat, the
cell structure breaks down liberating volatile
gases and vapors. Further heating causes some
chemical decomposition, and the resultant prod-
ucts are often highly odorous. All these mal-
odorous gases and vapors are entrained in ex-
haust gases.
Exhaust products from cooking processes con-
sist essentially of steam. Entrained gases and
vapors are, nevertheless, highly odorous and
apt to elicit nuisance complaints in areas sur-
Table 227. ODOR THRESHOLD CONCENTRA-
TIONS OF SELECTED COMPOUNDS
(Dalla Valle and Dudley, 1939)
Substance
Acrolein
Allyl aznine
Ally! mercaptana
Ammonia
Dibutyl sulfide
Ethyl mercaptana
Hydrogen sulfide
Oxidized oils
Skatole
Sulfur dioxide
Formula
CH2:CH- CHO
CH2:CH-CH2-NH2
CH2:CH-CH2-SH
NH3
(C4H9)2S
C2H5 SH
H2S
C?H8NH
so2
Threshold concentration,
mg/llter
0.038
0. 067
0. 00005
0. 037
0, 0011
0. 00019
0. 0011
0. OOJ1
0. 0012
0. 009
ppm by volume
16
28
0. 016
52
0. 180
0.072
0. 770
-.
0.220
3. 3
Average value obtained with material of varying purity.
rounding animal matter reduction plants. Odor
concentrations measured in exhaust gases of
typical reduction processes are listed in Table
228. Evidently there is a wide variation in
odor concentrations from similar equipment.
For instance, dry-batch rendering processes
range from 5, 000 to 500, 000 odor units per scf,
depending principally upon the type and "ripe-
ness" of feedstocks. Blood drying can be even
more odorous, with concentrations as great as
1 million odor units per scf if the blood is allowed
to age for only 24 hours before processing.
Odors From Air Driers
As can be seen from Table 228, feather drier
odor concentrations, though generally smaller,
are more variable than those from rendering
Table 228. ODOR CONCENTRATIONS AND EMISSION RATES FROM
INEDIBLE REDUCTION PROCESSES
Source
Rendering cooker,
dry-batch type
Blood cooker,
dry-batch type
Feather drier,
steamtubec
Blood spray
drier0' d
Grease-drying tank,
air blowing
156°F
170°F
225°F
Odor concentration,
odor unit/scf
Range
5, 000 to
500, 000
10, 000 to
1 million
600 to
25, 000
600 to
1, 000
Typical average
50, 000
100, 000
2, 000
800
4, 500
15,000
60, 000
Typical moisture
content of
feeding stocks, %
50
90
50
60'
< 5
Exhaust products,
scf /ton of feeda
20, 000
38, 000
77, 000
100, 000
100 scfm
per tank
Odor emission
rate, odor unit/
ton of feed
1, 000 x 106
3, 800 x 106
153 x 106
80 x 106
aAssuming 5 percent moisture in solid products of system.
"Noncondensable gases are neglected in determining emission rates.
cExhaust gases are assumed to contain 25 percent moisture.
Blood handled in spray drier before any appreciable decomposition occurs.
-------
778
CHEMICAL PROCESSING EQUIPMENT
cookers. Their largest odor concentrations--
25, 000 odor units per scf--are associated with
operations where feedstocks are putrefied or
not completely cooked beforehand or where the
meal is overheated in the drier. Under optimum
conditions, odor concentrations from these driers
should not exceed 2, 000 odor units per scf. With
blood spray driers, where extreme care is main-
tained to ensure freshness of feedstocks, con-
centrations can be less than 1, 000 odor units per
scf. In general, air drier odor concentrations
are less than those of cookers for the following
reasons: (1) In most instances feedstocks are
cooked or partially evaporated before the air
drying; (2) odorous gases are more dilute in
drier exit gases; (3) feedstocks are often fresher.
Odors and Dust From Rendered-Product Systems
Some odors and dust are emitted from cooked
animal matter as it is separated and refined.
The heaviest points of odor emission are the
percolators into which hot cooker contents are
dumped. Steam and odors evolve from the hot
material, particularly during times of cooker
unloading. Cookers are normally dumped at or
near 212°F. Lesser volumes of steam and odors
are generated at presses, centrifuges, and settling
tanks where meal and tallow are heated slightly
to effect the desired separation.
The grinding of pressed solids, and subsequent
meal conveying are the only points of dust emis-
sion from rendering systems. These particulates
are reasonably coarse, and dust is usually not
excessive.
Grease-Processing Odors
Some odors are generated at processing tanks
when moisture is removed from grease or tallow
by boiling or by air blowing or both. If air is
used for this purpose, exhaust volumes seldom
exceed 100 scfm, but odor concentrations are
measurable. Odor concentration is a function
of operating temperature. As shown in Table
228, measured concentrations have been found
to range from 4, 500 odor units per scf at 150°F,
to 60,000 odor units per scf at 225 °F. Odor
concentrations vary greatly with the type of
grease processed and the air rate, as well as
with temperature.
Row-Materials Odors
Some malodors emanate from the cutting and
handling of raw materials. In most instances
these emissions are not great. Odors usually
originate at the point where raw material is
first sliced, ground, or otherwise broken into
smaller parts. Most feedstocks are ground in
a hammer mill before the cooking. Large,
whole animals (dead stock) must be skinned,
eviscerated, and at least partially dismembered
before being fed to rendering equipment. If the
animal is badly decomposed, this skinning and
cutting operation can evolve strong odors.
HOODING AND VENTILATION REQUIREMENTS
All heated animal matter reduction processes
should be vented directly to control equipment.
Hooding is used in some instances to collect
malodors generated in the processing of raw
materials and cooked products.
If highly decayed dead stock is being processed,
the entire dead stock room should be ventilated
at a rate of 40 or more air changes per hour for
worker comfort. Areas should also be ventilated
where raw materials are stored unrefrigerated
for any appreciable time before processing.
Hooding may be employed on raw-material
grinders preceding cookers and percolator
pans and expeller presses used to handle
cooked products. Although the volume of
steam and odors evolved at any of these points
does not exceed 100 cfm, greater volumes are
normally required to offset crossdrafts. In-
draft velocities of 100 fpm under hoods are usu-
ally satisfactory.
Emission Rotes From Cookers
The ventilation rates of cookers can be esti-
mated directly from the quantity of moisture
removed and the time of removal. Maximum
emission rates from dry cookers are approxi-
mately twice the average moisture evaporation
rates. In the determination of exhaust volumes,
noncondensable gases can normally be neglected.
Consider a batch cooker that removes 3, 000
pounds of moisture from 6, 000 pounds of animal
matter in 3 hours, a relatively long cook cycle.
The average rate of emission is 16.7 pounds
per minute or 450 cfm steam at about 212°F.
The instantaneous evaporation rate and cumula-
tive moisture removal are- plotted in Figure
570. The maximum evolution rate apparently
occurs near the initial portion of the cook at
29 pounds per minute or 790 cfm at 212°F. As
moisture is removed from a batch cooker, the
heat transfer rate decreases, the temperatures
rise, and the evaporation rate falls off. The
general shapes of the curves in Figure 570 are
typical of batch-cooking cycles. Where cook
times are appreciably shorter, evaporation
rates are greater; nevertheless, the ratio of
maximum to average evaporation rate is main-
tained at approximately 2 to 1.
-------
Reduction of Inedible Animal Matter
779
10 15 20 25 30 35 4
COOKING TIME, hours
Figure 570. Steam emission pattern from a
hatch-type, dry rendering cooker operated
at ambient pressure.
The length of a cooking cycle, and the evapo-
ration rate are dependent upon the temperature
in the steam jacket, and the rotational speed
of the agitator. The highest permissible agita-
tor speeds (about 65 rpm) can result in cooking
times of 45 minutes to 1 hour. Many operators,
particularly at packing houses, use slower
agitator speeds, and cycles are as long as 4
hours.
If vacuum cooking is employed, volume rates
and temperatures decrease as the batch pro-
gresses. With these systems, the vacuum-
producing devices largely govern cooking times.
The evaporation rate in a vacuum, system is lim-
ited by the rate at which steam can be removed,
usually by condensation. If vapor cannot be
condensed as fast as it is evaporated, the cycle
is merely lengthened.
Pressure cookers have a slightly different emis-
sion pattern, but maximum emission rates are
again twice the average. During the initial
portion of the cycle, there are no emissions
while pressures are increasing to the desired
maximum. The cooker is vented at elevated
pressure, usually about 50 psig. High-pres-
sure vapors are relieved through small bypass
lines so that the surge of steam is not more than
the control system can handle. Most of the con-
tained moisture is evaporated after pressures
are reduced to ambient levels.
Vapor emission rates from wet rendering cook-
ers are considerably lower than those from dry
cookers, comparable to initial volumes during
pressure cooking. Only enough steam is flash
evaporated to reduce the pressure to 1 atmo-
sphere. The large percentage of moisture in
a wet rendering process is removed as water
by physical separation rather than by evapora-
tion.
Emission rates from continuous, dry rendering
processes are steady and can be calculated
directly from the moisture content of feedstocks
and products. To lower the moisture content
from 50 to 5 percent in typical meat and bone
scrap, 1, 670 scfm or 79 pounds of steam per
minute would be evaporated if the charge rate
to the cooker were 10, 000 pounds per hour.
Emission rates from blood cookers are general-
ly lower than those from dry rendering cookers
owing to the longer cook cycles employed. Blood
is continually added to an operating cooker during
a typical packinghouse workday. The emission
rate fluctuates as a function of the moisture con-
tent in the cooker. A cook cycle may extend
over 8 or 10 hours, and charging patterns can
vary tremendously. Emission rates do not
normally exceed 500 cfm, and at times, are con-
siderably lower.
Emission rates from feather cookers follow the
same pattern as those from other dry pressure
cookers though rates are lower and cooking
times usually longer. Inasmuch as feathers
contain no appreciable tallow, heat transfer is
relatively slow. At some plants, batches of
feathers are cooked as long as 8 hours. Where
separate driers are used, feathers are still
cooked 2 to 4 hours, which reduces the moisture
content to 50 percent before the charging to a
drier.
Emission Rates From Driers
Most air-drying processes are operated on a
continuous basis with no measurable fluctua-
tions in exhaust rates. Enough air and, in some
instances, products of combustion are added to
yield a moisture content of 10 to 30 percent by
volume in the exit gas stream. To dry 2, 000
pounds of cooked feathers per hour from 50 to
5 percent moisture requires a drier (steamtube)
exhaust volume of 1, 660 scfm at 20 percent
moisture in the gases. Volumes from air driers
are always much greater than those from cook-
ing processes, and they contain far greater
quantities of noncondensable gases.
-------
780
CHEMICAL PROCESSING EQUIPMENT
AIR POLLUTION CONTROL EQUIPMENT
The principal devices used to control reduction
plant odors are afterburners and condensers,
installed separately and in combination. Ad-
sorbers and scrubbers also find use. Dust is
not a major problem at animal matter reduction
plants, and simple cyclones are usually suffi-
cient to prevent excessive emissions.
Selection of odor control equipment is influenced
greatly by the moisture content of the malodorous
stream, or conversely, by the percentage of non-
condensable gases. It is usually more costly to
control noncondensable gases than moisture.
Reduction plant exhaust streams fall into two
general types: (1) Those consisting almost en-
tirely (95 percent or greater) of water vapor,
as from rendering cookers and blood cookers,
and (2) air drier exhaust gases, which seldom
contain more than 30 percent moisture by volume.
Controlling High-Moisture Streams
Condensing moisture from wet cooker gases is
almost always economically attractive. Some
malodors are usually condensed or dissolved in
the condensate. In any case the volume is re-
duced by a factor of 10 or more. The remaining
noxious gases can be directed to a further control
device such as an afterburner or carbon adsorber
before being vented to the atmosphere.
Selection of the condenser depends upon the par-
ticular facilities of the operator. The principal
types of condensers noted in Chapter 5 are adapt-
able to reduction cooker exhaust streams. Con-
tact condensers and air-cooled and water-cooled
surface condensers have been successfully used
for this purpose.
Contact condensers are more efficient control
devices than surface condensers are, though
both types are highly effective -when coupled -with
an afterburner or carbon adsorber. This is
illustrated by data in Table 229. Odor concen-
trations are seen to be considerably greater in
gases from surface condensers than in those
from contact condensers. With condensate at
80°F, a contact condenser reduces odor concen-
trations by about 80 percent and odor emission
rates by 99 percent. At the same condensate
temperature, odor concentrations increase across
a surface condenser. Either type of condenser,
however, reduces the volume of cooker vapors
by 95 percent or more. Thus, even a surface
condenser lowers the odor emission rate by about
50 percent.
Contact condensers are relatively inexpensive
to install but require large quantities of one-
pass cooling water. From 15 to 20 pounds of
cooling water is necessary to condense and sub-
cool adequately 1 pound of steam. Since cooling
water and condensate are intimately mixed, the
resultant liquid cannot be cooled in an atmospheric
cooling tower without emission of malodors to the
atmosphere. The large condensate volume that
must be disposed of can overload sewer facili-
ties in reduction plant areas.
Subcooling Condensate
Surface condensers, whether air cooled or water
cooled, should be designed to provide subcooling
Table 229. ODOR REMOVAL EFFICIENCIES OF CONDENSERS OR AFTERBURNERS,
OR BOTH, VENTING A TYPICAL DRY RENDERING COOKER3
(Calculated from Mills et al. , 1963)
Odors from cookers
Concentration,
odor units/scf
50, 000
Emission rate,
odor units /min
25,000,000
Condenser
type
None
Surface
Surface
Contact
Contact
Condensate
temperature,
"F
--
80
140
80
140
Afterburner
temperature,
°F
1,200
None
1,200
None
1,200
Odors from control system
Concentration,
odors
units /scf
100 to 150
(Mode 120)
100, 000 to
10 million
(Mode 500,000)
50 to 100
(Mode 75)
2,000 to
20, 000
(Mode 10, 000)
20 to 50
(Mode 25)
Modal emission
rate, odor
units /mm
90, 000
12, 500,000
6, 000
250,000
2,000
Odor removal
efficiency,
99. 40
50
99.93
99
99.99
Based on a hypothetical cooker that emits 500 scfm of vapor containing 5 per cent noncondensable gases.
-------
Reduction of Inedible Animal Matter
781
of condensate to 140°F or lower. This maybe
accomplished in several ways, as noted in Chapter
5. The need for subcooling is negated when high
vacuum is employed. With vacuum operation,
volatile, malodorous gases are drawn off through
the ejector or vacuum pump, and condensation
temperatures are often less than 140°F. At a
vacuum of 24 inches of mercury (2. 9 psia), the
condensation temperature of steam is 140°F.
Condenser Tube Materials
Reduction process vapors can be highly corro-
sive to the metals commonly used in surface
condenser tubes. Both acid and alkaline vapors
can be present, sometimes alternately in the
same equipment. Vapors from relatively fresh
meat and bone scrap rendering are mildly acidic,
and some brasses are satisfactory. Brasses
fail rapidly, however, under alkaline conditions.
Mild steel tubes are adequate where the pH is
greater than 7. 0 but quickly corrode under acid
conditions.
Some operations, such as dead stock rendering,
can produce alkaline and acid gases alternately
during the cook cycle. Here neither brass nor
mild steel is satisfactory. In these cases, stain-
less steels have been successfully employed.
With a relatively constant pH condition, less ex-
pensive metals could be used.
Where acid-base conditions are uncertain, a
pH determination should be made. The vapors
should be sampled over the complete process
cycle with all representative feedstocks in the
cookers.
Interceptors in Cooker Vent Lines
Air pollution control systems venting cookers
should be equipped with interceptor traps to pre-
vent fouling of condensers and other control de-
vices. So-called wild blows are relatively com-
mon in dry rendering operations. They result
from momentary plugging of the cooker vent.
Steam pressures increase until they are suffi-
cient to unblock the line. In the unblocking, a
measurable quantity of animal matter is forced
through the vent line at high velocity. If there
is no interceptor, this material fouls condensers,
hot wells, afterburners, and other connected con-
trol devices. Although a wild blow is an opera-
tional problem, it greatly affects the efficiency
of odor control equipment.
The systems shown in Figures 571 and 572 in-
clude interceptors in the vent lines between the
cookers and condensers. The installation de-
picted in Figure 572 uses an air-cooled con-
Figure 571. A condenser-afterburner control
system with an interceptor located between
the rendering cooker and condenser.
denser and afterburner. Most tanks are of suffi-
cient size to hold approximately one-half of a full
cooker charge. They are designed so that col-
lected materials can be drained while the cooker
and control system are in operation.
Vapor Incineration
For animal matter reduction processes, as with
most odor sources, flame incineration is the most
positive control method. Afterburners have been
used individually and in combination with other de-
vices, principally condensers. Rule 64 of the
Los Angeles County Air Pollution Control District
(see Appendix A), -which specifically governs heat-
ed animal matter reduction processes, uses incin-
eration at 1,200°F as an odor control standard.
Any control method or device as effective as
flame incineration at 1,200°F is acceptable under
the regulation.
Total incineration is used to control low-mois-
ture reduction process streams, as from driers,
and various other streams of small volume. At
reduction plants, steamtube driers are normal-
ly the largest equipment controlled in this manner.
Gases from the driers are vented directly to after-
burners, which are operated at temperatures of
1, 200°F or higher. Dust is usually not in sufficient
concentration to impede incineration. If there is
appreciable particulate matter in the gas stream,
auxiliary dust collectors must be installed or the
afterburner must be operated at 1, 600°F or high-
er. At 1,200°F, solids are only partially incin-
erated.
Flame incineration at 1,200°F reduces odor con-
centrations from steamtube driers to 100 to 150
-------
782
CHEMICAL PROCESSING EQUIPMENT
Figure 572. A cooker control system including an interceptor,
air-cooled condenser, and afterburner (California protein Products,
Los Angeles, Cali f.).
odor units per scf where dust loading is not ex-
cessive. Some variation can be expected when
concentrations are greatly in excess of the nom-
inal 2, 000 odor units per scf usually encountered
in drier gases.
Because of the large volumes exhausted from
driers, afterburner fuel requirements are a
major consideration. A drier emitting 3,000
scfm requires about 4, 800 scfh natural gas for
1,200°F incineration. Several means of recov-
ering waste heat from large afterburner streams
have been used. The most common are the
generation of steam and preheating of drier
inlet gases.
In the control of spray driers, dust collectors
must often be employed ahead of the afterburner.
High-efficiency centrifugal collectors, baghouses,
or precipitators maybe required as precleaners,
depending upon the size and concentration of par-
ticulates.
Condensotion—Incineration Systems
As noted earlier, wet cooker vapors are seldom
incinerated in toto. While 100 percent incinera-
tion is feasible, operating costs are much great-
er than for condenser-afterburner combinations.
Both types of control systems provide better
than 99 percent odor removal, but the combina-
tion system results in a much lower odor emis-
sion rate.
The cooker control systems shown in Figures
571 and 572 and in Chapter 5, illustrate typical
combinations of condensers and afterburners.
Uncondensed gases are separated from con-
densate at either the condenser or hot -well.
Gases enter the afterburner near ambient tem-
perature. Either contact or surface condensers
serve to remove essentially all particulates. The
remaining "clean" uncondensed gases can be
readily incinerated at 1,200"F. In some instances
there are minor concentrations of methane and
other fuel gases in the stream. Uncondensed
gases from surface condensers are richer in com-
bustibles than are those from contact condensers.
As shown in Table 229, odor removal efficiencies
greater than 99. 9 percent are possible with con-
denser-afterburner systems serving dry render-
ing cookers.
-------
Reduction of Inedible Animal Matter
783
When the moisture content of the contaminated
stream is from 15 to 40 percent, the use of
condensers may or may not be advantageous.
In these cases, a number of factors must be
weighed including volumes, exit temperatures,
fuel costs, water availability, and equipment
costs among others.
Incineration is not required 'with some condenser
installations. An afterburner or equally effective
device should always be employed to eliminate the
heavy concentrations of malodors vented from
surface condensers. With contact condensers,
however, adequate control can be effected if con-
densate temperatures are sufficiently low. The
data in Table 229 show that a contact condenser
operating at 80°F provides about 99 percent con-
trol of cooker odors. When there are appreciable
uncondensed gases or when condensate tempera-
tures are greater than approximately 140°F, the
additional control and insurance afforded by in-
cineration may be mandatory.
Carbon Adsorption of Odors
Most of the malodorous gases emitted from re-
duction processes can be adsorbed on activated
carbon to some degree. The capacities of
activated carbons for hydrogen sulfide, uric
acid, skatole, putrescine, and several other
specific compounds found in reduction plant gas-
es are considered "satisfactory" to "high. " For
ammonia and low-molecular-weight amines,
they have somewhat lower capacities. The latter
compounds tend to be desorbed as the carbon be-
comes saturated with high-molecular-weight
compounds (Barnebey-Cheney Co. , Bulletin
T-642). For the mixture of malodorous mate-
rials encountered at reduction plants, a high-
quality carbon would be expected to adsorb
from 10 to 25 percent of its weight before the
breakthrough point is reached.
Carbon adsorbers are as efficient as afterburn-
ers but have limitations that often make them
unattractive for cooker control. Their most
useful application is the control of large volumes
of relatively cool and dry gases. Adsorbers
usually cannot be employed in reduction process
streams without auxiliary dust collectors, con-
densers, or coolers.
Carbon adsorbers cannot be used to control
emissions from wet cookers unless the adsorb-
ers are preceded by condensers. Activated
carbon does not adsorb satisfactorily at tem-
peratures greater than 120°F. To cool cooker
vapors, which are predominantly steam, to
this temperature, most of the moisture must
be recovered. At 120°F, saturated air contains
only 11.5 percent water vapor by volume. Con-
denser-adsorber systems are reported to re-
move odors as efficiently as condenser-after-
burner systems. No comparative odor con-
centration data are available.
Drier exhaust streams can be controlled with
adsorbers if inlet temperatures and dust con-
centrations can be held sufficiently low and
small, respectively. Many driers are exhausted
at temperatures higher than 200°F and contain
enough fine particulates to foul adsorbers. A
scrubber-contact condenser is often a satis-
factory means of removing particulates and low-
ering temperatures before adsorption. If, how-
ever, there are appreciable particulates of less
than 10 microns diameter, more efficient dust
control devices are necessary.
Regeneration of activated carbon is a major con-
sideration at animal matter reduction plants.
Carbon life between regenerations can be as
short as 24 hours, particularly where malodors
are in heavy concentration, and the carbon has
a low capacity for the compounds being adsorbed.
Regeneration frequency is a function of many
factors, including malodor concentration, the
quality and quantity of carbon, and the kind of
compounds that must be adsorbed.
Some means must be employed to contain or
destroy the desorbed gases; otherwise, mal-
odors are vented to the atmosphere in essentially
the same form that they were collected. Incin-
eration at 1,200°F or higher is the most common
method of controlling these gases. For streams
of low volume, afterburners used during regen-
eration can be as large and as costly as those
used to incinerate odors from the basic reduc-
tion equipment. The need for incineration of
desorbed gases usually offsets the advantages
of carbon adsorption for streams of low volume.
If the exhaust rate is sufficiently small, incin-
erating vapors directly, as they are evolved
from the reduction equipment or condenser, is
considerably simpler.
Odor Scrubbers
Conventional scrubbers are seldom used to
control reduction process odors. Of course,
contact condensers provide some scrubbing
of cooker gases; nevertheless, these devices
are principally condensers, and tail waters
cannot be recirculated. It is conceivable that
alkaline or acid scrubbers would be effective
for drier gases if all the odorous compounds
reacted in the same manner. Unfortunately,
the malodorous mixtures encountered in typical
reduction processes are not homogenous from
the acid-base standpoint.
-------
784
CHEMICAL PROCESSING EQUIPMENT
Strong oxidizing solutions, such as chlorine
dioxide, are reported to destroy many of the
odorous organic materials (Woodward and
Fenrich, 1952). With any type of recirculating
chemical scrubber, the contaminated stream
would first have to be cooled to ambient tem-
perature, by condensation if necessary.
Odor Masking and Counteraction
Masking agents and odor counteractants have
been used with some success to offset iii-plant
odors. These materials are added to cooker
feedstocks and sprayed in processing and storage
areas. They are reported to provide a degree
of nuisance elimination and worker comfort,
particularly in high-odor areas such as dead
stock skinning rooms. Masking agents and
counteractants, however, are not recommended
for the control of odors from heated animal
matter reduction equipment.
ELECTROPLATING
Electroplating is a process used to deposit,
or plate, a coating of metal upon the surface
of another metal by electrochemical reactions.
In variations of this process, nonmetallic sur-
faces have been plated with metals, and a non-
metal such as rubber has been used as a plating
material. Industrial and commercial applica-
tions of electroplating are numerous, ranging
from manufactured parts for automobiles, tools,
other hardware, and furniture to toys. Brass,
bronze, chromium (chrome), copper, cadmium,
iron, lead, nickel, tin, zinc, and the precious
metals are most commonly electroplated.
Platings are applied to decorate, to reduce
corrosion, to improve wearing qualities and
other mechanical properties, or to serve as
a base for subsequent plating with another metal.
The purpose and type of plating determine the de-
tails of the process employed and, indirectly,
the air pollution potential, which is a function of
the type and rate of "gassing, " or release of gas
bubbles from plating solutions with entrainment
of droplets of solution as a mist. The degree
of severity of air pollution from these process-
es may vary from being an insignificant problem
to a nuisance.
An electroplating system consists of two elec-
trodes--an anode and a cathode--immersed in
an electrolyte and connected to an external
source of direct-current electricity. The base
material upon which the plating is to be deposited
makes up the cathode. In most electroplating
systems, a bar of the metal to be deposited is
used as the anode. The electrolyte is a solution
containing: (1) Ions of the metal to be deposited
and (2) additional dissolved materials to aid in
electrical conductivity and produce desirable
characteristics in the deposited plating.
When an electric current is passed through the
electrolyte, ions from the electrolyte are re-
duced, or deposited, at the cathode, and an
equivalent amount of either the same or a dif-
ferent element is oxidized or dissolved at the
anode. In some systems, for example, chrome
plating, the deposited metal does not dissolve
at the anode, and hence, insoluble anodes are
used, the source of the deposited metal being
ions formed from salts of that metal previously
dissolved in the electrolyte.
The character of the deposited metal is affected
by many factors, including the pH of the electro-
lyte, the metallic ion concentration, the sim-
plicity or complexity of the metallic ion (includ-
ing its primary and secondary ionization prod-
ucts), the anodic and cathodic current densities,
the temperature of the electrolyte, and the
presence of modifying or "addition agents." By
varying these factors, the deposit can be varied
from a rough, granular, loosely adherent plat-
ing to a strong, adherent, mirror-finish plating.
If the electromotive force used is greater than
that needed to deposit the metal, hydrogen is
also formed at the cathode, and oxygen forms
at the anode. When insoluble anodes are used,
oxygen or a halogen (if halide salts are used in
the electrolyte) is formed at the anode. Both of
these situations produce gassing.
A potential air pollution problem, can also occur
in the preparation of articles for plating. These
procedures, primarily cleaning processes, are
as important as the plating operation itself for
the production of high-quality finishes of im-
pervious, adherent metal coatings. The clean-
ing of metals before electroplating generally
requires a multistage procedure as follows:
1. Precleaning by vapor degreasing or by soak-
ing in a solvent, an emulsifiable solvent, or
an emulsion (used for heavily soiled items);
2. intermediate cleaning with an alkaline bath
soak treatment;
3. electrocleaning with an alkaline anodic or
cathodic bath treatment, or both (the chem-
ical and mechanical [gassing] action created
by passing a current through the bath between
the immersed article and an electrode pro-
duces the cleaning);
-------
Electroplating
785
4. pickling with an acid bath soak treatment,
with or without electricity.
The selection of an appropriate cleaning method
in any given case depends upon three important
factors: The type and quantity of the soil, com-
position and surface texture of the base metal,
and the degree of cleanliness required. In gen-
eral, oil, grease, and loose dirt are removed
first; then scale is removed, and, just before
the plating, the pickling process is employed.
The articles to be plated are thoroughly rinsed
after each treatment to keep them from contami-
nating succeeding baths. A cold rinse is usually
used after the pickling to keep the articles from
drying before their immersion in the plating bath.
Electrocleaning and electropickling are generally
faster than similar soak procedures; however,
the electroprocesses always produce more gas-
sing (hydrogen at the cathode and oxygen at the
anode) than the nonelectroprocesses. The gas-
sing from cleaning solutions tends to create
mists that may but usually do not, cause signi-
ficant air pollution problems.
THE AIR POLLUTION PROBLEM
The electrolytic processes do not operate with
100 percent efficiency, and some of the current
decomposes water in the bath, evolving hydrogen
and oxygen gases. In fact, the chief advantage
of electrocleaning is the mechanical action pro-
duced by the vigorous evolution of hydrogen at
the cathode, which tends to lift off films of oil,
grease, paint, and dirt. The rate of gassing
varies -widely with the individual process. If
the gassing rate is high, entrained mists of
acids, alkaline materials, or other bath con-
stituents are discharged to the atmosphere.
Most of the electrolytic plating and cleaning pro-
cesses are of little interest from a standpoint of
air pollution because the emissions are inoffen-
sive and of negligible volume, owing to low gas-
sing rates. Generally, air pollution control
equipment is not required for any of these pro-
cesses except the chromium-plating process.
In this process, large volumes of hydrogen and
oxygen gases are evolved. The bubbles rise
and break the surface with considerable energy,
entraining chromic acid mist, which is dis-
charged to the atmosphere. Chromic acid mist
is very toxic and corrosive and its discharge
to the atmosphere should be prevented.
Chromic acid emissions have caused numerous
nuisance complaints and frequently cause prop-
erty damage. Particularly vulnerable are auto-
mobiles parked downwind of chrome-plating in-
stallations. The acid mist spots car finishes
severely. The amounts of acid involved are
relatively small but are sufficient to cause
damage. In a typical decorative chromium-
plating installation with an exhaust system but
without mist control equipment, a stack test
disclosed that 0. 45 pound of chromic acid per
hour was being discharged from a 1, 300-gallon
tank.
Chromium-plating processes can be divided into
two general classes, one of which offers a con-
siderably greater air pollution problem than the
other. "Hard chrome" plating, which causes
the more severe problem, produces a thick,
hard, smooth, corrosion-resistant coating.
This plating process requires a current density
of about 250 amperes per square foot, which
results in a high rate of gassing and a heavy
evolution of acid mist. The less severe problem
is presented by the process called "decorative
chrome" plating, which requires a current
density of only about 100 amperes per square
foot and results in a definitely lower gassing
rate.
HOODING AND VENTILATING REQUIREMENTS
Local exhaust systems are installed on many
electroplating tanks to reduce the concentra-
tions of steam, gases, and mists to what are
commonly accepted as safe amounts for person-
nel in the plating room. In the past, these ex-
haust systems were often omitted altogether,
and the resulting working conditions were often
unhealthful.
In 1951, the American Standards Association
introduced Code Z9. 1 for Ventilation and Oper-
ation of Open Surface Tanks. This code is an
organized engineering approach designed to re-
place the rule-of-thumb methods applied in the
past. The use of this code in designing plating
tank exhaust systems is recommended by public
health officials and industrial hygienists.
Most exhaust systems use slot hoods to capture
the mists discharged from the plating solutions.
These hoods have been found satisfactory when
properly designed. To obtain adequate distribu-
tion of ventilation along the entire length of the
slot hood, the slot velocity should be high, 2,000
fpm or more, and the plenum velocity should
be one-half of the slot velocity or less. With
hoods over 10 feet in length, either multiple
takeoffs or splitter vanes are needed. Enough
takeoffs or splitter vanes should be used to re-
duce the length of the slot to sections not more
than 10 feet long.
Ventilation rates for tanks, as previously dis-
cussed in Chapter 3, are for tanks located in
areas having no crossdrafts. In drafty areas,
-------
786
CHEMICAL PROCESSING EQUIPMENT
ventilation rates must be increased and baffles
should be used to shield the tank.
AIR POLLUTION CONTROL EQUIPMENT
Scrubbers
The device most commonly used to control air
contaminants in hard-chrome-plating tank ex-
haust gases is a wet collector. This type of
equipment is also suitable for controlling mists
from any other type of plating or cleaning tank
that may cause a problem. Figure 573 shows
a ventilation system •with a. spray-type scrubber
used to control mists from two 18-foot chrome-
plating tanks. Many other types of commercial
wet collectors are available, constructed of
various corrosion-resistant materials. Water
circulation rates are usually 10 to 12 gpm per
1, 000 cfm. If the water is recirculated, the
makeup rate is about 2.5 to 4 gph per 1, 000 cfm.
The scrubber water, of course, becomes con-
taminated with the acid discharged from the plat-
ing tank; therefore, efficient mist eliminators
must be used in the scrubber to prevent a con-
taminated water mist from discharging to the
atmosphere.
The scrubber water is commonly used for plat-
ing tank makeup. This procedure not only re-
moves the acid from the scrubber but also re-
duces the amount of makeup acid needed for the
plating solution. In some scrubbers, a very
small quantity of fresh water is used to collect
the acid mist; the resulting solution is continu-
ously drained from the scrubber either into the
plating tank or into a holding tank, from which
it can be taken for plating solution makeup.
The mists collected by the air pollution control
system are corrosive to iron ' r steel; therefore,
hood, ducts, and scrubbers .•; these materials
must be lined with, or repl. ' d by, corrosion-
resistant materials. Steel icts and scrubbers
lined with materials such at, polyvinyl chloride
have been found to resist r J.equately the corro-
sive action of the mists, in recent years, hoods,
ducts, and scrubbers m de entirely of polyester
resins reinforced with glass fibers have been
used in air pollution c /ntrol systems handling
acid or alkaline solutions. These systems have
been found to be ve- y resistant to the corrosive
effects of plating solutions.
The scrubber r'moves chromic acid mist with
high efficiency. V commonly used field method
of determining ciiromic acid mist evolution
consists of holding a sheet of white paper over
the surface of the tank or scrubber discharge.
Any mist contacting the paper immediately
stains it. A piece of paper held in the discharge
of a well-designed scrubber shows no signs of
staining.
Figure 573. Two control systems with scrubbers venting four chrome-
plating tanks. Each scrubber vents two tanks (Industrial Systems, Inc.
South Gate, Calif.).
-------
Insecticide Manufacture
787
Mist Inhibitors
The mist emissions from a decorative-chrome-
plating tank and from other tanks with lesser
mist problems can be substantially eliminated
by adding a suitable surf ace-active agent to the
plating solutions. The action of the surface -
active agent reduces the surface tension, which,
in turn, reduces the size of the hydrogen bubbles.
Their rates of rise, and the energy of their evolu-
tion are greatly reduced, and the amount of mist
is also greatly reduced. Several of these mist
inhibitors are commercially available.
If the proper concentration of mist inhibitor is
maintained, a sheet of paper placed 1 inch above
the bath surface shows no spotting.
INSECTICIDE MANUFACTURE
The innumerable substances used commercially
as insecticides can be conveniently classified
according to method of action, namely: (1) Stom-
ach poisons, which act in the digestive system;
(2) contact poisons, which act by direct external
contact with the insect at some stage of its life
cycle; and (3) fumigants, which attack the
respiratory system.
A few of the commonly used 3 secticides, clas-
sified according to method r action, are shown
in Table 230. The classifii ition is somewhat
arbitrary in that many poisons, such as nicotine,
possess the characteristics of two or three classes.
Table 230. SOME COMMON INSECTICIDES
CLASSIFIED ACCORDING TO
METHOD OF ACTION
Stomach poisons
Paris green
Lead arsenate
Calcium arsenate
Sodium fluoride
Cryolite
Rotenone
Contact poisons
DDT
Pyrethrum
Sulfur
Lime -sulfur
Nicotine sulfate
Methoxychlor
Fumigants
Sulfur dioxide
Nicotine
Hydrocyanic acid
Naphthalene
P-dichloro -benzene
Ethylene oxide
Human threshold limit values of various insecti-
cides are shown in Table 231. They represent
conditions under which it is believed that nearly
all workers may be repeatedly exposed day after
day, without adverse effect. The amount by which
these figures may be exceeded for short periods
without injury to health depends upon factors such
as (1) the nature of the contaminant, (2) whether
large concentrations over short periods produce
acute poisoning, (3) -whether the effects are
cumulative, (4) the frequency with which large
concentrations occur, and (5) the duration of
these periods.
METHODS OF PRODUCTION
Production of the toxic substances used in in-
secticides involves the same operations employed
for general chemical processing. Similarly,
chemical-processing equipment, that is, reac-
tion kettles, filters, heat exchangers, and so
forth, are the same as discussed in other sec-
tions of this chapter. Emphasis is given, there-
fore, to the equipment and techniques encoun-
tered in the compounding and blending of com-
mercial insecticides to achieve specific chemi-
cal and physical properties.
Most commercial insecticides are used as either
dusts or sprays. Insecticides employed as dusts
are in the solid state in the 0. 5- to 10-micron
size range. Insecticides employed as sprays
may be manufactured and sold as either solids
or liquids. The solids are designed to go into
solution in an appropriate solvent or to form a
colloidal suspension; liquids may be either solu-
tions or water base emulsions. No matter what
physical state or form is involved, insecticides
are usually a blend of several ingredients in
order to achieve desirable characteristics. A
convenient means of classifying equipment and
their related processing techniques is to differ-
entiate them by the state of the end product.
Equipment used to process insecticides where
the end product is a solid is designated solid-
insecticide-processing equipment. Equipment
used to process insecticides where the end
product is a liquid is designated as liquid-in-
secticide-processing equipment.
Solid-Insecticide Production Methods
Solid mixtures of insecticides may be com-
pounded by either (1) adding the toxicant in
liquid state to a dust mixture or (2) adding s.
solid toxicant to the dust mixture.
Figure 574 illustrates equipment used if the
toxicant in liquid state is sprayed into a dust
mixture during the blending process. After
leaving the rotary sifter, the solid raw mate-
rials are carried by elevator to the upper
mixer where the liquid toxicant is introduced
by means of spray nozzles. This particular
unit has discharge gates at each end of the up-
per mixer, which permit the wetted mixture
to be introduced either directly into the second
mixer or into the high-speed fine-grinding pul-
verizer and then into the second mixer. From
the second mixer, a discharge gate with a built-
in feeder screw conveys the mixture to a second
elevator for transfer to the holding bin where the
finished batch is available for packaging. Al-
though as much as 50 percent by weight of liq-
uid toxicant may be added to the blend, the di-
-------
788
CHEMICAL PROCESSING EQUIPMENT
Table 231. THRESHOLD LIMIT VALUES OF VARIOUS INSECTICIDES
Substance
Threshold limit value,
mg/meter
Aldrin (1,2,3,4,10,10-hexachloro-
1,4,4a,5,8,8a-hexahydro-l,4,5,8-
dimethanonaphthalene)
Arsenic
Calcium arsenate
Chlordane (1, 2, 4,5,6,7,8,8-octachloro-3a, 4, 7,
7a-tetrahydro-4, 7-methanoindane)
Chlorinated camphene, 60%
2, 4-D (2, 4-dichlorophenoxyacetic acid
DDT (2, 2-bis(p-chlorophenyl)
-1,1, 1 -trichloroethane)
Dieldrin (1,2, 3, 4, 10, 1 O-hexachloro-6, 7,
epoxy-1,4,4a,5,6,7,8,8a-octahydro-
1, 4, 5, 8-dimethano-naphthalene)
Dinitro-o-cresol
EPN (O-ethyl O-p-nitrophenyl thionobenzenephos-
phonate)
Ferbam (ferric dimethyl dithiocarbamate)
Lead arsenate
Lindane (hexachlorocyclohexane gamma isomer)
Malathion (O, O-dimethyl dithiophosphate of
diethyl mercaptosuccinate)
Methoxychlor (2, 2-di-p-methoxyphenyl-l, 1, 1 ••
trichloroethane)
Nicotine
Parathion (O, O-diethyl-O-p-nitrophenyl
thiophosphate)
Pentachlorophenol
Phosphorus pentasulfide
Picric acid
Pyrethrum
Rotenone
TEDP (tetraethyl dithionopyrophosphate)
TEPP (tetraethyl pyrophosphate)
Thiram (tetramethyl thiuram disulfide)
Warfarin (3-(a-acetonylbenzyl) 4-
hydroxycoumarin)
0.25
0. 5
1
0.5
0. 5
10
1
0. 25
0. 2
0. 5
15
0. 15
0. 5
15
15
0.5
0. 1
0.5
1
0. 1
5
5
0.2
0. 05
5
0. 1
luent clays are porous and absorb the liquid
to such a degree that the ingredients of the mix
are essentially solids and act as such. In in-
secticide processing, the type of mixer general-
ly employed to blend liquids with dusts is the
ribbon blender.
Figure 575 is an illustration of a ribbon blender
screw. This screw consists of two or more
ribbon flights of different diameters and opposite
hand, mounted one within the other on the same
shaft by rigid supporting lugs. Ingredients of
the mix are moved forward by one flight and
backward by the other, which thereby induces
positive and thorough mixing -with a gradual
propulsion of the mixed material to the discharge.
An example of an insecticide compound produced
by this method is toxaphene dust. A commonly
used formulation is:
Toxaphene (chlorinated
camphene) 40
Kerosine 4. 5
Finely divided porous clay 55. 5
The toxaphene is melted and mixed with the
kerosine, then sprayed into the clay and thor-
oughly blended.
-------
Insecticide Manufacture
789
Figure 574. Sol id-insecticide-processing unit (Poulsen Company, Los Angeles, Calif.).
Figure 575. Ribbon blender screw (Link-Belt Company, Los Angeles, Calif.).
When the toxicant is in the solid state, the in-
gredients of the blend are intimately ground, usu-
ally in stages, and blended by mechanical mixing
operations. The equipment employed consists of
standard grinding and size reduction machines
such as ball mills, hammer mills, air mills,
disc mills, roller mills, and others. A spe-
cific example of a grinding and blending facility
for solid insecticide is shown in Figures 576 and
577. This installation is used for compounding
DDT dust. The grinding and blending operations
are done in two stages. First, the material is
processed in the premix grinding unit and then
transferred to the final grinding and blending unit.
-------
790
CHEMICAL PROCESSING EQUIPMENT
H
f*
3IRREI
Figure 576. Premix grinding
unit.
Figure 576 illustrates the equipment comprising
the premix grinding unit. This unit is used for
the initial grading and blending of DDT and silica
mixtures. DDT flakes, 75 percent of which have a
particle size of 1-centimeter diameter, are emp-
tied from sacks into a hopper. A conveyor takes
this DDT to a crusher from which it is conveyed
to a pulverizer. Finely ground silica (0. 2- to
2-micron size) is introduced to the pulverizer.
Silica is added because DDT becomes waxy at
temperatures approaching its melting point and
has a tendency to cake and resist grinding.
Silica acts as a stabilizing agent. The coarsely
ground silica-DDT mixture is then discharged
into a ribbon blender for thorough mixing be-
fore being conveyed to a barrel-filling unit,
which packs the mixture for aging before its
further grinding.
The final grinding unit shown in Figure 577 takes
the coarsely ground DDT-silica mixture and sub-
jects it to fine grinding and blending. The aged
DDT-silica mixture is fed into a ribbon blender
where additional silica and wetting agents are
added to the mix. The mix is then screw con-
veyed to a high-speed grinding mill that uses
rotating blades to shear the insecticidal mix-
ture. A pneumatic conveying system carries
the material to a cyclone separator from which
it drops into another blender. After this mix-
ing operation, the blend is finely ground by high-
pressure air in an airmill. The blend is air
conveyed to a reverse-jet baghouse that dis-
charges into another blender. Additional air
grinding is then repeated before the barrel
filling and packing.
( , r1 TO • TKOS'PHER
FINISHED PRODUCT
Figure 577. Final grinding and blending unit.
-------
Insecticide Manufacture
791
Liquid-Insecticide Production Methods
Liquid insecticides may be produced as either
solutions, emulsions, or suspensions. The
most common means of production consists of
introducing a solid toxicant into a liquid carrier,
which results in either a solution, emulsion,
or suspension.
Figure 578 shows equipment employed in a liq-
uid-emulsion insecticide plant that makes the
emulsion by introducing a solid toxicant into a
liquid carrier in the presence of an emulsifying
agent. A typical formulation is:
DDT (technical)
Emulsifying agent No. 1
Emulsifying agent No. 2
Organic solvent
Ib
200
12
12
569 (79.5 gal)
TO ATMOSPHERE
Figure 578. Liquid-insecticide-formulating unit.
addition of the liquids, until the desired emulsi-
fied state is achieved. The finished product is
then pumped to the drum-filling station for pack-
aging.
THE AIR POLLUTION PROBLEM
As can be seen from the installations just de-
scribed, air pollutants generated by the insecti-
cide industry are of two types--dusts and or-
ganic solvent vapors.
To collect insecticide dusts, high-efficiency
collectors are mandatory, since in many in-
stances, the dust is extremely toxic and cannot
be allowed to escape into the atmosphere, even
in small amounts. The moderate fineness, 0. 5
to 10 microns, of the dust necessitates using
collectors that are effective in these particle
size ranges. For the most part, the dusts en-
countered are noncorrosive.
Organic solvent vapors emitted from liquid-in-
secticide production processes ordinarily orig-
inate from relatively nonvolatile solvents. These
vapors are of such concentration, nature, and
quantity as to be inoffensive from a viewpoint of
air pollution.
HOODING AND VENTILATION REQUIREMENTS
Because of the toxicity of the dusts used in the
manufacture of insecticides, it is important
that all sources of dust be enclosed or tightly
hooded to prevent exposure of this dust to per-
sonnel in the working area. Wherever possible,
the sources should be completely enclosed and
ventilated to an air pollution control device.
Some of the sources emitting dust are bag pack-
ers, barrel fillers, hoppers, crushers, con-
veyors, blenders, mixing tanks, and grinding
mills. Of these, the crushing and grinding
operations are the largest sources of emission.
In most cases, these are not conducive to com-
plete enclosure, and hoods must be employed.
Indraft velocities through openings in hoods
around crushers and mills should be 400 fpm
or higher. Velocities through hood openings
for the other operations, where dust is re-
leased with low velocities, should be 200 to
300 fpm.
The operation consists of adding the DDT to the
mixing tank, the DDT being held on a horizontal
wire screen located at the vertical midpoint of
the tank. Organic solvent and emulsifying agents
are then pumped into the mixing tank at the ap-
proximate level of the dry DDT. The mixture is
continually agitated, both during and after the
AIR POLLUTION CONTROL EQUIPMENT
Baghouses employing cotton sateen bags are. the
most common means of controlling emissions
from the insecticide-manufacturing industry.
In some applications, water scrubbers, of both
the spray chamber and the packed-tower types,
are used to control dust emissions. Inertial
-------
792
CHEMICAL PROCESSING EQUIPMENT
separators such as cyclones and mechanical
centrifugal separators are not used because
collection efficiencies are not high enough to
prevent the smaller size toxic particles from
being emitted into the atmosphere.
In the solid-insecticide-processing unit previ-
ously discussed and illustrated in Figure 574,
air pollution control is achieved by dust pickup
hoods located at the inlet rotary sifter and at
the automatic bag packer. The dust picked up
at these points is filtered by the use of cloth
bags. Most units of this type are entirely
enclosed, air contaminants being discharged
only at the inlet to the unit and at the outlet.
The contaminants emitted are extremely
fine dust and, since no elevated temperatures
are encountered and the materials handled are
not particularly corrosive to cloth, can be
easily collected by simple cloth bag filters.
If extremely large throughputs are encountered,
a conventional baghouse may be required. Since
no extreme conditions of operation are general-
ly involved, the most widely used filter material
is a cotton sateen cloth.
In larger installations, such as those illustrated
in Figures 576 and 577 for compounding DDT
dust, several baghouses are usually used. In
the premix grinding unit shown in Figure 576,
the air pollution control equipment consists of
an exhaust system discharging into a baghouse,
which is equipped with a pullthrough exhaust
fan. The exhaust ducting connects to both the
DDT and the silica hoppers, the DDT crusher,
the blender, and the barrel-filling unit. Dust
collected in the baghouse is conveyed to the
barrel-filling unit for packaging.
The final grinding unit, shown in Figure 577,
uses air pollution control equipment consisting
of a baghouse that serves the receiving hopper,
the blenders, and the cyclone air discharge of
the high-speed grinding mill. Dust collected
in this baghouse is recycled to the feed blender.
The final blenders and the barrel filler and
packer are vented to one of the reverse-jet
baghouses serving the fluid energy mill. As
in the case of mixing liquid toxicant with dust,
the material collected is not corrosive to
cotton cloth, and no elevated temperatures are
encountered. In this installation, cotton sateen
bags of 1. 12 to 1. 24 pounds weight per yard
with an average pore size of 0. 004 inch are
employed as the filtering medium.
In liquid-insecticide manufacturing, air pollu-
tion control problems usually entail collection
of dusts in a wet airstream. Baghouses can-
not, therefore, be used, and some type of
scrubber must be employed. For the liquid-
emulsion insecticide plant shown in Figure 578,
the air pollution control equipment for the solid
and liquid aerosols consists of a packed tower
that vents the mixing tank. The tower is packed
with 1-inch Intalox saddles, the packing being
4-1/2 feet high, which equals a volume of 14
cubic feet. The water rate through the tower
is approximately 20 gpm. The tower is used
to control dust emissions from the mixing tank,
which occur when dry material is charged to the
tank, and also occur during the first stages of
agitation. Solvent vapors are not effectively
prevented by the tower from entering the atmo-
sphere since the solvent is insoluble in water.
Solvent emissions originate from the storage
tank and drum-filling unit. In the installation
described, no provision is made to prevent the
solvent from escaping to the atmosphere since
total solvent emissions are calculated to be
only 5. 4 pounds per day.
HAZARDOUS RADIOACTIVE MATERIAL
Although the responsibility for overseeing the
control of radioactive materials is predomi-
nantly that of the Federal government, more
and more responsibility is expected to be placed
at state and local levels. For this reason,
those concerned with air pollution must become
acquainted with the problems associated with
this new field, particularly those problems
arising as more and smaller industries make
use of radioactive materials.
HAZARDS IN THE HANDLING OF RADIOISOTOPES
The hazards encountered in handling of radio-
isotopes may be classified in order of impor-
tance as follows: (1) Deposition of radio-
isotope in the body, (2) exposure of the whole
body to gamma radiation, (3) exposure of the
body to beta radiation, and (4) exposure of the
hands or other limited parts to beta or gamma
radiation. Deposition of a radioisotope in the
body occurs by ingestion, inhalation, or ab-
sorption through either the Intact or injured
body surface. Inhalation of a radioactive gas,
vapor, spray, or dust may occur. Spray or
dust is particularly hazardous because of the
large fraction of contamination retained by the
lungs (National Bureau of Standards, 1949).
Types of radiation are listed in Table 232. The
ranges of activity may be defined as: (1) Tracer
level, less than 1 x 10~° curie; (2) low level,
1 x 10-6 to 1 x ID'3 curie; (3) medium level,
1 x 10-3 curie to 1 curie; and (4) high level,
1 curie and over. The handling of tracer quan-
tities of radioisotopes usually presents no ex-
ternal hazard. Ordinary laboratory manipula-
tions are performed -with special precautions
-------
Hazardous Radioactive Material
793
Table 232. TYPES OF RADIATION
Type of
radiation
Alpha (a)
Beta (j3)
Gamma (7)
Neutron (n)
Physical nature
Heavy particle,
helium nucleus,
double positive
charge
Light -particle
electron, single
negative charge
Ray, similar to
X-ray
Moderately heavy
particle, neutral
charge
Distance of
travel in air
Few inches maximum
Few yards maximum
Very long
Very long
Effective shielding
Skin or thin layer
of any solid mate-
rial
One -half inch of
any solid material
Lead, other heavy
metals, concrete,
tightly packed soil
Water, paraffin
Usual means of
detection
Proportional counter,
ion chamber, scin-
tillation counter
Geiger counter, film
badge, dosimeter
Geiger counter, ion
chamber, film badge,
dosimeter
Proportional counter
containing boric com-
pound, ion chamber
with cadmium shield
to prevent absorption of radioactive material
by the body.
THE AIR POLLUTION PROBLEM
Radioactive materials used in industry are a
definite hazard today and will become an in-
creasing rather than a diminishing hazard in
the future. In industry, the maximum per-
missible dose of direct, whole-body radiation
of persons from all radioactive materials,
airborne or nonairborne, is 5,000 millirem
per year. There is greater likelihood that this
limit will be reduced than that it -will be in-
creased. Airborne radiological hazards can
result from routine or accidental venting of
radioactive mists, dusts, metallurgical fumes,
and gases and from spillages of liquids or
solids. Presently existing governmental regu-
lation of the rate of venting airborne, radio-
active materials consists primarily of spe-
cific limitations based upon individual chemi-
cal compounds or upon concentrations of ra-
dioactivity from single vents. No concepts
have been promulgated concerning methods
of controlling total radioactive air pollution
from all sources in an entire area. Whether
it -will be either desirable or necessary to
find a solution or solutions to these problems
is an unanswered question.
The characteristics of radioactive, gaseous
or airborne, particulate wastes vary widely
depending upon the nature of the operation from
which they originate. In gaseous form they may
range from rare gases, such as argon (A^l)
from air-cooled reactors, to highly corrosive
gases, such as hydrogen fluoride from chemi-
cal and metallurgical processes. Particulate
matter or aerosols may be organic or inorganic
and range in size from less than 0. 05 micron
to 20 microns. The smaller particles originate
from metallurgical fumes caused by oxidation
or vaporization. The larger particles may be
acid mist droplets, which are low in specific
gravity and remain suspended in air or gas
streams for longer periods (Liberman, 1957).
Characteristics of Solid, Radioactive Waste
Solid, radioactive wastes are of two general
classes—combustible and noncombustible.
Typical combustible solid wastes are paper,
clothes, filters, and wood. Noncombustible,
solid wastes may include nonrecoverable scrap,
evaporator bottoms, contaminated process equip
ment, floor sweepings, and broken glassware.
If inadequate provisions are made for proper
handling and disposal of these wastes, a distinct
nuisance, and, under certain circumstances,
even a hazard, could result.
Characteristics of Liquid, Radioactive Waste
Liquid, radioactive wastes are evolved in all
nuclear energy operations-from laboratory
research to full-scale production* Liquid
wastes with relatively small concentrations
of radioactivity originate in laboratory oper-
ations where relatively small quantities of
radioactive materials' are involved. Other
sources are the processing of uranium ore and
feed material; the normal operation of essen-
tially all reactors, particularly water-cooled
types; and the routine chemical processing of
reactor fuels. High-activity liquid wastes
-------
794
CHEMICAL PROCESSING EQUIPMENT
are produced by the chemical processing of
reactor fuels.
Problems in Control of Airborne, Radioactive Waste
Removal of radioactive suspended particles,
vapors, and gases from "hot" (radioactive)
exhaust systems before discharge to the at-
mosphere is a serious problem confronting
all nuclear energy and radiochemistry instal-
lations. Removal is necessary in order to pre-
vent dangerous contamination of the immediate
and neighboring areas. Air pollution brought
about through discharge of radioactive stack
gas wastes from ventilation systems is only
partially avoided by filter devices, no matter
how efficient they may be, if the discharge
contains radioactive gases. In systems using
filter media such as paper, cloth, glass fiber,
and so forth, activity eventually builds up in
the filter media through dust loading; the same
situation applies to electrical precipitators.
Another problem in the control of airborne,
radioactive -waste is the low dust loading of
exhaust streams. The dust concentration of
ambient air is usually about 1 grain per 1, 000
cubic feet. At installations handling radio-
active material, owing to precleaning of the
entering air, aerosols may have concentra-
tions as small as 10 to 10~3 grain per 1, 000
cubic feet. In contrast, loadings of some in-
dustrial gases may reach several hundred
grains per cubic foot, though values of 20
grains or less per cubic foot are more com-
mon.
An outstanding feature to consider with air-
cleaning requirements for many nuclear oper-
ations is the extremely small permissible con-
centrations of various radioisotopes in the at-
mosphere (see Table 233). Often, removal ef-
ficiencies of about 99. 9 percent or greater for
particles less than 1 micron in diameter are
necessary. This high removal efficiency lim-
its the selection of control equipment for ra-
dioactive applications.
HOODING AND VENTILATION REQUIREMENTS
Hooding
Hooding for radiochemical processes must pre-
vent radioactive contaminants, such as dust
and fumes, from escaping into the work area
and must deliver them to suitable control de-
vices. Radioactive sources require proper
shielding to prevent the escape of radiation
and are not considered in this section. The
materials used for construction for hoods depend
upon the type and quantities of radioactivity and
the nature of the process. Stainless steel,
masonite, transite, or sheet steel, surfaced
•with a washable or strippable paint, can be
used (Ward, 1952). Where it is necessary in
a process to handle material that may cause
dusts or fumes to form, a completely enclosed
hood should be used, equipped with a glove box
or dry box. Any tools used for manipulation
should not be removed from the hood.
Ventilation
The recommended airflow for toxic material
across the face of a hood is 150 fpm (Manufac-
turing Chemists' Association, 1954). Turbu-
lence of air entering a. hood can be reduced by
the addition of picture frame airfoils to the edges.
Hoods should not be located where drafts •will
affect their operation. When more than one hood
is located in a room, fan motors should be oper-
ated by a single switch. The fan should freely
discharge to the atmosphere and be connected
to the outlet side of any control device, the
motors being located outside the air ducts to
prevent their contamination. Hood and ducts
should be equipped with manometers to indi-
cate that they are operating under a negative
pressure.
AIR POLLUTION CONTROL EQUIPMENT
Reduction of Radioactive, Particulate Matter at Source
Reduction at the source has been defined as the
design of processes so as.to minimize the initial
release of particulate matter at its source. The
principle is not new; it is applied, for example,
in the ceramics industry where dry powders
are wetted and mixed as a slurry to minimize
the production of dust. But Its application to
radioactive aerosols is particularly worthwhile
since it (1) provides a cleaner effluent, (2) re-
duces radiation hazards involved in the mainte-
nance of air-cleaning equipment or those re-
sulting from the buildup of dust activity, (3) per-
mits the use of simpler and less expensive air-
cleaning equipment, and (4) becomes a part of
the process once reduction has been established.
In general, preventing the formation of highly
toxic aerosols is preferable to cleaning by
secondary equipment.
The design or redesign of processes for reduc-
tion at the source should be based upon a study
of the quantity and physical characteristics of
the contaminant, and the manner in which it is
released. Examples of this concept are instal-
-------
Hazardous Radioactive Material
795
Table 233. PROPERTIES OF RADIOISOTOPES (Benedict and Pigford, 1957)
Isotope
H3
Be?
ci4
Na24
p32
Cl36
K42
Fe55
Fe^g
Ni59
Co°0
Ni63
Cu°4
Zn"
Ge?l
Ga?2
A'Iz
£s
Y90
Y91
Nb95
Tc96
Mo99
Pd103
Rhl°3
Rhl°5
Agl°5
R 109
AgH°
Agin
Sr?13
In114
Half-life
12.5 yr
52.9 days
5, 568. yr
112 mm
15 hr
14. 3 days
87. 1 days
4. 4 x 105 yr
109 min
12.4 hr
152 days
85 days
16 days
27. 8 days
2.9 yr
2.6 hr
45. 1 days
8 x 104 yr
5. 3 yr
85 yr
12.8 hr
250 days
11.4 days
14.3 hr
26.8 hr
35.9 hr
9. 4 yr
19.5 days
53 days
19.9 yr
61 hr
61 days
35 days
4. 2 days
67 hr
17 days
57 min
36.5 hr
40 days
i yr
470 days
270 days
7. 6 days
112 days
49 days
Type of
decay
Beta
Beta
Beta
Beta
Beta
Beta
Beta
Beta
Beta
Beta
Beta
Beta, EC
EC, no beta
EC, no beta
Beta
Beta
EC
'Beta
Beta
EC, beta
EC, beta
EC, no beta
Beta
Beta
Beta, no EC
Beta
Beta, no EC
Beta
Beta
Beta
Beta
Beta
EC, no beta
Beta
EC
ITC
Beta
EC
Beta
EC, no beta
Beta.ITno EC
Beta
EC, no beta
IT, no EC
Maximum permissible
concentration in air,
microcuries /ml
Soluble
2 x 10-?
2 x 10-?
1 x 10-?
2 x 10-?
4 x 10-8
2 x 10-9
9 x 10-9
1 x 10-8
7 x 10-8
1 x ID'9
8 x 10-9
6 x 10-9
4 x 10"?
3 x lO-8
3 x lO-8
5 x 10-9
2 x lO-8
1 x lO-8
2 x 10-9
7 x 10~8
4x 10-9
4x lO'7
8x 10-9
4x ID"9
4 x 10"8
1 x 10-8
3 x 10'10
3 x 10-H
4x 10-9
1 x 10"9
2 x ID"8
2 x ID"8
3 x ID"8
5 x ID"8
3 x 10"6
a
3 x 10"8
2 x lO-8
3 x 10'9
2 x 10"9
7x 1C'9
1 x ID"8
1 x ID"8
4x 10"9
Insoluble
4 x 10-5 Suba
4 x 10-8
9 x ID'8
5 x 10-9
3 x 10-9
9 x 10-9
8 x 10-10
4x10-8 Sub
4 x 10"9
4 x ID'9
8x ID'10
2 x ID"9
8 x ID'8
3 x ID"8
2 x lO"8
2 x 10-9
3 x 10-8
3x 10-1°
1 x ID'8
4 x ID"8
2 x 10-9
2 x 10-?
6 x 10-9
3x 10-9
6 x 10-9
3 x 10-? Sub
2 x 10-9
1 x 10-9
2 x 10-1°
3x 10-9
1 x 10-9
3x10-9
8x 10-9
7 x 10-9
3 x ID'8
2 x 10"6
2 x lO-8
3x10-9
2x 10-1°
3x ID'9
3 x 10-10
8 x 10'9
2 x 10'9
7x10-1°
Isotope
Sb122
sbi24
Sb125
Te127
Tui29
Cs134
xei35
Csl37
Ba140
Lal40
p^!43
Cel44
Pml47
Sm^l
EU154
Ho"*
Tml?°
Rel83
Ir190
Ir192
Aul99
Tf204
Po210
At2 H
Ac227
Th232
Pa233
U233
Th234
U238
Pu239
Am241
Cm242
Half-life
2. 8 days
60 days
~2. 7 yr
115 days
33.5 days
8. 1 days
5. 3 days
2. 3 yr
7
9. 1 hr
33 yr
12.8 days
40 hr
33. 1 days
13.7 days
282 days
2.6 yr
73 yr
16 yr
>30 yr
129 days
6. 8 days
155 days
12.6 days
74. 4 days
2.7 days
3. 1 days
47.9 days
3. 5 yr
138. 3 days
7.5 hr
22 yr
1.39 x 1010
27.4 days
1.62 x 105
y
24. 1 days
4.49 x 109
2.44 x 104
yr
470 yr
162. 5 days
Type of
decay
Beta
Beta, no EC
Beta
IT
IT
Beta
Beta
Beta, no EC
Beta
Beta
Beta
Beta
Beta
Beta
Beta
Beta
Beta
Beta
Beta
Beta, no EC
Beta
EC
EC
EC, beta
Beta, no EC
Beta
Beta
Beta, EC
Alpha, beta
stable
Alpha, EC
Alpha, beta
Alpha, beta
stable
Beta
Alpha, beta
stable
Beta
Alpha, beta
stable
Alpha, beta
stable
Alpha, beta
stable
Alpha, beta
stable
Maximum permissible
concentration in air,
microcuries /ml
Soluble
6 x 10'9
5 x ID"9
2 x ID"8
6 x 10"8
2 x 10-?
1 x 10"10
1 x ID'9
2 x 10-9
4x ID'9
5 x ID"9
2 x lO-8
1 x ID"8
3x10-1°
2 x ID'9
2 x ID"9
Ix 10-1°
7 x ID"9
1 x 10"9
2 x ID"8
9 x ID'8
4x ID"8
4x 10~9
1 x ID"8
4 x 1C"8
2 x 10"9
2 x ID'8
2 x 10"11
2x10-1°
8 x 10-14
2 x 10-8
2x 10-H
2x 10-9
3xlO-12
6 x IO-I4
2 x 10"13
4 x jo-12
Insoluble
5 x 1C'9
7x 10-10
9x10-1°
3 x ID'8
1 x 10'7
1 x ID"8
3 x lO'7 Sub
4x10-1°
1 x 10'7 Sub
5x 10-1°
IxlO"9
4 x ID'9
5 x ID"9
6x 10-9
2x10-1°
3 x ID'9
5 x 10"9
2x10-1°
6xlO'9
Ix ID'9
2 x 10-8
5x10-9
1 x ID"8
9x 10-1°
8 x 10-9
3 x ID"8
4x ID'9
9 x 10-1°
7 x ID"12
1 x ID'9
9 x lO'l3
io-i2
6 x 10-9
io-9
4xlO-12
6 x IO-I2
aValues given are for submersion in an infinite cloud of
gaseous material.
"Orbital-electron capture.
clsomeric transition.
lation of glass fiber filters on the inlet of ven-
tilating or cooling air to minimize the irradia-
tion of ambient dust particles, and treatment
of ducts to minimize corrosion and flaking
(Friedlander et al., 1952).
Design of Suitable Air-Cleaning Equipment
The most satisfactory control of particulate
contamination -with air-cleaning equipment re-
sults from using combinations of the various
collectors. These installations should be de-
signed to terminate with the most efficient
separator possible, the nature of the gases
being considered. To reduce maintenance,
less efficient cleaners capable of holding or
disposing of most of the weight load should be
placed before the final stage. It is good prac-
tice to arrange the equipment in order of in-
creasing efficiency. A typical example of such
an arrangement is a wet collector such as a
centrifugal scrubber to cool the gases and re-
move most of the larger particles, an efficient
dry filter such as a glass fiber filter to remove
most of the remaining particulate matter, and
a highly efficient paper filter to perform the
-------
796
CHEMICAL PROCESSING EQUIPMENT
final cleaning, If the gases are moist, as in
this example, the paper filter should be pre-
ceded by a preheater to dry the gases (Fried-
lander, 1952).
An air-cleaning installation for highly toxic
aerosols should fulfill the following require-
ments (Friedlander et al. , 1952):
1. "It should discharge innocuous air.
2. "The equipment should require only occa-
sional replacement and should be designed
for easy maintenance. Frequent replace-
ment or cleaning entails excessive exposure
to radiation and the danger of redispersing
the collected material.
3. "The particulate matter should be separated
in a form allowing easy disposal. The use
of wet collectors, for example, poses the
additional problem of disposing of volumes
of contaminated liquid. Wet collection does,
however, reduce considerably the danger of
redispersion.
4. "Initial and maintenance costs, as well as
operating costs, should be as low as possi-
ble while fulfilling the preceding three con-
ditions. In this respect, pressure drop is
generally an important consideration. "
Reverse-jet baghouse
One type of commercially available dust collector
that meets the requirements of filtering airborne,
radioactive particles from ventilation exhaust
streams is a bag filter employing what is called
reverse-jet cleaning. This type of baghouse
(described in Chapter 4) has an efficiency as
high as the conventional cloth bag or cloth screen
collector and is particularly adapted to an in-
stallation where the grain loading of the effluent
is low. The bag material is a hard wool felt of
the pressed type, about 1/16-inch thick, or a
cloth woven of glass fibers. The gas flow is
likely to be around 10 to 40 cfm per square foot
of bag area when the pressure drop is maintained
at usual values such as 2 to 7 inches water col-
umn (Anderson, 1958).
The conventional cloth bag or cloth screen col-
lectors, which are cleaned periodically by auto-
matic shaking devices, may allow a puff of dust
to escape after the shaking operation. The prob-
lem of maintenance in this instance presents a
contamination and radiation hazard. For this
reason, the reverse-jet baghouse is generally
preferred.
Wet collectors
Another method of treating contaminated ex-
haust air before discharge to the atmosphere
involves the use of wet collectors of various
types. These collectors are relatively effec-
tive on gases. Investigation covering changing
of water supply or recirculating has shown the
latter procedure useful for considerable periods
of time without apparent adverse effect. Evapo-
ration is compensated for by fresh supply. In-
soluble radioactive salts, soluble salts, and
other radioactive particles that may form a
solution, suspension, or sludge in the reser-
voir result in fairly high radioactivity of the
scrubbing media. Precautions must be taken
during maintenance to avoid carryover of the
scrubbing media since the radioactive con-
tamination of entrained liquid would be trans-
ferred to the preheater or filter, resulting in
high radiation levels at those points.
Disadvantages of wet collectors
Some important disadvantages of wet col-
lectors make them less attractive than
other types of collectors,, Wet collectors
present the difficult problem of separating
the radioactive, solid ma.terial from the
water in which it is suspended. Mainte-
nance and corrosion are serious problems.
Considerable quantities of water are re-
quired, and, if the radioactive solids are
not separated from the -water, this in turn
leads to a final storage and disposal problem.
Electrical precipitators
Radioactive, airborne particles, when given an
electrical charge, can be collected on grounded
surfaces. The fact that the particles are radio-
active has very little to do -with their behavior
in an electrical precipitator. Experiments con-
ducted with precipitators using the alpha emitter
polonium and the beta emitter sulfur 35 indicate
that neither material behaves in a way different
from nonradioactive material.
Water-flushed-type, single-stage, industrial
precipitators, and air-conditioning-type, two-
stage precipitators are used for separating ra-
dioactive dusts and fumes from gases at atomic
energy plants and laboratories. A small elec-
trical precipitator of the water-flushed type with
a design capacity of 200 cfm was installed to
test efficiency of collecting and removing par-
ticulate radioactivity from the offgas system
of an isotope recovery operation. This precip-
itator consists of 23 vertical collecting p-'pes
with an ionizing wire centered in each pupe. The
inside surfaces of the! pipes serve as collecting
-------
Hazardous Radioactive Material
797
walls. For wet operation, the collecting walls
are water flushed by means of spray nozzles in-
stalled at the top of each pipe. This water is re-
cycled continuously at a rate of 35 gpm over the
collecting walls while high voltage is applied to
the electrodes. This unit reportedly collects
more than 99. 99 percent of the particulate radio-
activity in the offgas at 50 to 55 kilovolts when
the concentration of radioactivity as solids is
greater than 5. 0 x 10"^ microcuries per cubic
centimeter of offgas (Anderson, 1958).
Based upon tests made at the Oak Ridge Nation-
al Laboratory, Anderson (1958) makes the follow-
ing evaluation of precipitators used in radio-
active applications:
1. "Electrical precipitators are not intended
to collect the ultra fine particles which
may be discharged from radiochemistry
installations.
2. "With uneven airflow, the air velocity
through some of the collector cells may
be sufficiently above velocity limits to blow
off collected "wastes which would then be
discharged to the atmosphere.
3. "Efficient operation depends a great deal
on the regularity with which the unit is
cleaned. At best the electrical precip-
itator is only approximately 90 percent
efficient. This may be demonstrated by
the fact that dense clouds of tobacco smoke
fed into the precipitator will escape from
it in concentrations great enough so that the
escaping smoke can be seen. The blue
color of tobacco smoke is evidence that
most of its particles have a diameter less
than the wavelength of light, which is
roughly 0. 5 micron.
Glass fiber filters
Glass fiber or glass fiber paper is often used
as a filter medium and is effective in the oper-
ation of radiochemistry hoods, canopies, and
gloved boxes. One of the most efficient light-
weight, inorganic filters developed to date is
made with a continuous, pleated sheet of micro-
glass fiber paper. The pleats of the glass paper
are separated by a corrugated material (paper,
glass paper, aluminum foil, plastic, or as-
bestos paper) for easy passage of air to the deep
pleats of the filter paper. The assembly of the
filter paper and corrugated separators is sealed
in a frame of wood, cadmium plated steel, stain-
less steel, or aluminum. This construction per-
mits a large area of filter paper to be presented
to the airstream of a correspondingly low re-
sistance (Flanders Filters, Inc. , Riverhead,
N. Y.).
Glass fiber, from which filters are made, with-
stands temperatures up to 1, 000°F. It is non-
combustible and has extremely low thermal
conductivity and low heat capacity. The fibers
are noncellular, are like minute rods of glass,
and do not absorb moisture; however, water
can enter the interstices. The material is
relatively nonsettling, noncorrosive, and durable.
It is resistant to acid fumes and vapors, except
hydrogen fluoride.
The installation and replacement costs of glass
fiber filters are low. Final disposal of used
filters may be accomplished by incinerating at
over 1, 000°F with provisions for decontaminating
the stack gases. This melts the glass fibers,
reducing the physical mass to the size of a glass
bead. Thus, glass fiber filters provide, in part,
a very good answer to the problem of control
and final disposal of radioactive contaminants.
4. "For absolute efficiency an after-filter of
the Cambridge or MSA Ultra-Aire type is
necessary to catch the dirt should the pre-
cipitator short circuit.
5. "Difficulty may be experienced if the dust-
load builds up faster than it can be removed
eventually becoming so heavy that arcing
occurs between the dirt bridges resulting
in a fire hazard.
6. "Devices such as the single-stage indus-
trial precipitator and the air-conditioning
type two-stage precipitator accomplish
only one phase of the problem. The final
disposal of radioactive wastes collected
and accumulated during operation and main-
tenance still remains."
Paper filters
A highly efficient paper filter medium can be
used with adequate effectiveness on incoming
ventilating air and as a final cleaner in many
instances. This type filter is composed of as-
bestos cellulose paper. A more recently de-
veloped filter has a glass fiber web. It is de-
signed and manufactured in corrugated form
to increase the available filter area and load-
ing capacity and to reduce initial resistance.
The filter units are tested at rated capacity
with standard U. S. Army Chemical Corps test
equipment for resistance and initial penetra-
tion and are unconditionally guaranteed to be
at least 99. 95 percent effective against 0. 3-
micron-diameter dioctyl phthalate particles.
This filter performs as well as, or better
-------
798
CHEMICAL PROCESSING EQUIPMENT
than, ..the earlier paper types and under tem-
peratures up to 1, 000°F.
Airborne, radioactive wastes are only part of
the control and disposal problem of nuclear
energy and radiochemistry installations. Solid
and liquid, radioactive wastes are subject to
the same limitations on disposal to the environ-
ment.
The methods of disposing of the final waste from
the collection systems present additional prob-
lems, as follows (Anderson, 1958):
1. "Incineration results in stack gas and par-
ticle discharge which is a cycle of the en-
tire problem repeated over again.
2. "Direct burial results in redispersal and
ground contamination with associated prob-
lems related to the ground water table.
3. "High dust or particle loading capacity re-
sults in high radioactivity of the collecting
media.
4. "Vapors, acid fumes and unfilterable gases
may cause rapid deterioration and disinte-
gration of filter media resulting in a main-
tenance and health hazard problem.
5. "Mechanical replacement costs are high
because of the remote handling involved.
6. "An auxiliary unit for emergency or main-
tenance shutdown must be available to pre-
vent the possibility of reverse flow of the
air stream out of "hot" equipment into
controlled rooms and areas."
Disposal and Control of Solid, Radioactive Waste
The most common method of disposal of solid,
radioactive wastes is land burial at isolated
and controlled areas. The earth cover over
these burial pits is usually about 12 feet, and
the surface is monitored regularly. A method
used for disposal of low-level, radioactive,
solid wastes consists of putting the wastes in
concrete and dumping it at sea. Incineration
of combustible, solid wastes is practiced, with
provisions for decontaminating the flue gases
(Shamos and Roth, 1950).
Disposal and Control of Liquid, Radioactive Waste
Low-level, radioactive, liquid wastes, under
proper environmental conditions, are suscepti-
ble to either direct disposal to nature or dis-
posal after minimum treatment. Treatment
processes used include coprecipitation, ion ex-
change, biological systems similar to sewage
treatment methods, and others. Only to the
extent that it is absolutely safe, maximum use
is made of the dilution factors that may be avail-
able in the environment and that can be assessed
quantitatively.
High-activity, liquid wastes associated with the
chemical processing of reactor fuels constitute
the butk of the engineering problem of disposal
of radioactive wastes. Highly radioactive, liq-
uid wastes are currently stored in specially de-
signed tanks. Since the effective life of the fis-
sion products constituting the wastes may be
measured in terms of hundreds of years, tank
storage is not a permanent solution to the dis-
posal problem. Evaporation before storage is
generally practiced to reduce storage volume
and cost. The degree to which evaporation is
carried out is limited in some instances by the
percentage of solids present in the waste or by
considerations of corrosion.
There are several practical approaches to
ultimate, safe disposal of high-activity, liquid
wastes. The actual fission products in radio-
active waste material may be fixed in an inert,
solid carrier so that the possibility of migra-
tion of the radioactivity into the environment
is eliminated or reduced to acceptable and safe
limits. The carrier containing the radioactive
material could then be permanently stored or
buried in selected locations. Fixation on clay,
incorporation in feldspars, conversion to oxide,
elutriation of the oxide, and fixation of the
elutriant are examples of systems under devel-
opme nt.
Because of the particular radiotoxicity and long
half-life of strontium-90 and cesium-137, the
removal and separate fixation and handling of
these two isotopes would substantially reduce
the effective life and activity of the waste and
facilitate its final disposal. With cesium and
strontium removed, the possibilities of safe
disposal into the environment under controlled
conditions are greatly increased.
It rrmy be practical to dispose of the wastes
underground in some cases without any treat-
ment, into formations such as (1) spaces pre-
pared by dissolution in salt beds or salt domes,
(2) deep basins containing connate brines and
with no hydraulic or hydrologic connection to
potable waters or other potentially valuable
natural resources, and (3) special excavations
in selected shale formations (Liberman, 1957).
-------
Oil and Solvent Re-Refining
799
OIL AND SOLVENT RE-REFINING
Many millions of gallons of oils and solvents
are used annually for lubricating vehicle en-
gines and other machinery, transmitting pres-
sure hydraulically, cleaning manufactured arti-
cles and textiles, and dissolving or extracting
soluble materials. In the course of their usage,
these oils and solvents accumulate impurities,
decompose, and lose effectiveness. The im-
purities include dirt, scale, water, acids, de-
composition products, and other foreign mate-
rials. Reclaiming some of these oils and sol-
vents for reuse by removal of the impurities
can be effected in many instances by re-refining
processes.
Most re-refiners must practice stringent econ-
omies to survive, and for this reason, second-
hand, cannibalized, or makeshift equipment is
often employed. Many re-refiners also neglect
maintenance, repairs, and general housekeeping
in order to keep operating costs low. As a result,
air pollution control is minimal or lacking unless
made mandatory by legislation.
RE-REFINING PROCESS FOR OILS
Lubricating oils collected from service stations
are the main source of supply. A typical scheme
for re-refining lubricating oil is shown in Figure
579. Re-refining is normally a batch process.
Treating clay, for example, Fuller's earth, is
added to the contaminated oil at ambient tem-
perature to aid in the removal of carbon mate-
rials. The mixture is next circulated through a
fired heater, usually a pipe or tube still, to a flash
tower for removal of diluent hydrocarbons and
water. The oil being reclaimed and the products
desired determine the final temperature (300°
to 600°F). Live steam, introduced at the base
of the flash tower, is used to assist in this phase
of the operation. Besides distilling off the light
fractions contained in the oil, the steam pre-
vents excessive cracking of the oil at the higher
temperatures.
A barometric condenser maintains a vacuum on
the tower. The overhead vapors containing
steam, low-boiling organic materials, and en-
trained hydrocarbons are aspirated through the
condenser to a separator tank. The condensate,
consisting of light gas, oil, and water, is col-
lected and separated in the separator tank. Non-
condensible gases are usually incinerated in
fireboxes of adjacent combustion equipment. The
light oil condensate is decanted from the water
and is suitable as liquid fuel. The contaminated
water is piped to a skimming pond where it is
cooled and either reused or disposed of by drain-
ing to a sewer. The oil-clay mixture is with-
drawn from the tower and filtered. The oil is
NaOH
H2S04
Na4SiQ4
FLASHBACK
"T1 ARRESTER
+ (~\ KNOCKOUT
'XV-j-V DRUM
NON-CONDEN-
SABLE GASES
Figure 579. Composite flow sheet for re-refining process.
-------
800
CHEMICAL PROCESSING EQUIPMENT
blended with additives and is canned or drummed.
The clay is usually hauled to a dump.
In some re-refineries, the process is preceded
by a dehydration operation. Water is removed
from the oil by using sodium silicate, sodium
hydroxide, and heat. Dehydrated oil is decanted
from the mixture and charged to the still. Sul-
furic acid treatment is also employed at some
re-refineries before the refining process. The
acid-treated oil is settled, decanted from the
acid sludge, and neutralized 'with caustic. Be-
fore the clay is added, sulfuric acid treatment
or air blowing may also be used to improve
color of the re-refined oil.
RE-REFINING PROCESS FOR ORGANIC SOLVENTS
The typical organic solvent re-refining process
is similar to that described for oil re-refining.
The prime difference between the processes is
that the volatilities of the organic solvents re-
refined are much greater than those of lubrica-
ting oils. Mineral spirits, benzene, toluene,
xylene, ketones, esters, alcohols, trichloro-
ethylene, and tetrachloroethylene from paint,
lacquer, degreasers, and dry cleaners are ex-
amples of solvents reclaimed by re-refining.
Figure 580 illustrates a typical solvent recov-
ery system. The mixture to be processed is
introduced into a settling tank to permit the
solids to settle out. The supernatant liquid is
then preheated and charged to a pot still topped
by a fractionating section, which may be under
vacuum. Vapors from the still are condensed
in a water-cooled surface condenser. Reflux-
ing may or may not be done, depending upon the
product, the degree of purity desired, and the
contaminants present. The condensate is ac-
cumulated in a holding tank, where a salt such
as sodium carbonate is added to "break" the
water from the solvent. After the water settles
out, it is removed, and the solvent is drummed
off as product.
THE AIR POLLUTION PROBLEM
Air Pollution From Oil Re-refining
The two primary air pollution problems connected
with oil re-refining are odors and hydrocarbon
vapors.
Chief odor sources are the contaminated water
and the noncondensible gases from the separator
tank £ind dehydration tank. Obnoxious odors
emanate from the skimming pond. Odors also
occur from the barometric condenser leg. If
the process water is aerated in a cooling tower
or spray pond, a serious odor problem occurs.
Other odors can originate from the dehydration
operation and from sulfuric acid sludges and
clay filter cakes.
In addition to air pollution from odors, oil re-
refining processes can emit some hydrocarbons
into the atmosphere. These originate from the
noncondensible gases and the layers of light,
volatile hydrocarbons on the surface of the sep-
arator tank and the skimming pond.
Air Pollution From Solvent Re-refining
As in oil re-refining, the chief air pollution
problems are odors but these are less severe
than those occurring from re-refining of lubri-
cating oil. Sources of ernissions are the settling
tanks during filling and sludge drawoff, the draw-
off of bottoms from the still, the product receiv-
ers, and the water jet reservoir (if vacuum is
produced by a barometric water jet). By creating
a vacuum, the water jet entraps the solvent va-
pors from the still.
FRACTIONATING
SECTION
RECC
MIX
VERABLE
TURE 1
SETTLING
TUNK
PREHEATER f
S~^ J
.(\) \
««TER PROOOCT
TO SE«ER
Figure 580. Typical solvent re-refining installation.
AIR POLLUTION CONTROL EQUIPMENT
Oil Re-refining
The most acceptable method of controlling emis-
sions from re-refining is incineration. Usually
the firebox of a boiler or heater provides ade-
quate incineration. The separator tank must be
covered and vented to a firebox. The vent line
should be equipped with a knockout drum and a
flashback arrester. Additional safety protection
can be achieved by introducing live steam into
the vent line upstream from the firebox. Other
vessels, for example, dehydrating tanks and
mixing tanks, may be tied into this system.
Emissions from the barometric, or contact,
-------
Chemical Milling
801
condenser can be controlled by maintaining a
closed recycle -water system or by modifying
the operation by substituting a shell-and-tube-
type condenser.
Recycle water, highly odorous from contact
with the oil and heated by contact with the hot
vapors, must be allowed to cool before reuse.
It can be controlled by cooling in a covered
settling tank that is properly vented to an
operating boiler or heater firebox. Con-
taminated recycle water must not be cooled
by aerating in a spray pond or cooling tower.
Solvent Re-refining
Usually, in the solvent re-refining industry, air
pollution control is lacking without enforcement,
and solvent vapors are allowed to escape into
the atmosphere. If, however, control is re-
quired, it can easily be accomplished by venting
the barometric -water jet vacuum system to a
boiler firebox, provided appropriate flashback
prevention measures have been taken. Emis-
sions from the bottom drawoff of the still are
slight since most of the volatiles have been
flashed off. Emissions from the settling tank
and the product receivers are normally too
small to create any problems, but they can be
controlled by being vented also to a boiler fire-
box.
CHEMICAL MILLING
The chemical milling process was developed
by the aircraft industry as a solution to the
problem of making lightweight parts of intri-
cate shapes for missiles. These parts could
not be formed if mechanically milled first, and
no machines were available that could mill them
after they -were formed. Chemical milling is
based upon the theory that an appropriate etch
solution dissolves equal quantities of metal per
given time from either flat or curved surfaces.
The process was quickly adopted by the aircraft
industry, and etchants were developed for
chemically milling many metals used in aircraft
and missiles, including aluminum, titanium,
stainless steel, and magnesium.
DESCRIPTION OF THE PROCESS
to protect the surface from oxidation in air and
provide a surface that will accept and hold a
masking agent or material.
Maskings are either tapes with pressure-sensi-
tive adhesives or paint-like substances that are
applied by brushing, dipping, spraying, or flow-
coating. Figure 581 shows a sheet of stainless
steel being flow-coated with a rubber base mask-
ing material. These paint-like maskings must
be cured, usually in a bake oven. After curing,
the masking is removed or stripped from those
areas to be milled. Figure 582 shows one meth-
od of scribing the masking by use of a template.
Figure 581. An 18- by 6-foot, stainless steel
sheet being masked by flow-coating with a rub-
ber-based masking (U.S. Chemical Milling Corp.
Manhatten Beach, Cali f.).
Milling is accomplished by submerging the pre-
pared article in an appropriate etching solution.
The depth of the cut is controlled by the length
of time the article is held in the etching solu-
tion. To stop the milling action, remove the
article from the etchant and rinse off the adher-
ing solution with water. During the milling
step, some metals are discolored by their etch-
ing solutions. The smutty discoloration is re-
moved in a brightening solution such as cold,
dilute nitric acid. A flow diagram of the pro-
cess is shown in Figure 583.
Before an article can be chemically milled, the
surface of the metal must be clean. The usual
metal surface preparation includes (1) degreas-
ing, (2) alkaline cleaning, (3) pickling, and
(4) surface passivation. The cleaning is needed
to provide a clean surface in order to insure uni-
form dissolving of the metal when it is submerged
in the milling solution. The passivation is needed
After the milling, the paint-like masking is
softened in a solution consisting, for example,
of 80 percent chlorinated hydrocarbons and 20
percent high-boiling alcohols, and is then
stripped off by hand. Figure 584 shows the in-
spection of a part. The metal thickness is mea-
sured before the masking is removed. Figure
585 shows the masking being removed from a
-------
802
CHEMICAL PROCESSING EQUIPMENT
Figure 582. The masking on a titanium part is being
scribed by use of a template. After scribing, the
masking will be stripped from those areas shown by
the holes in the templates. The stripped areas will
then be milled. The black part in the foreground and
those in the background have not yet been scribed
(U.S. Chemical Milling Corporation, Manhatten Beach,
Calif.).
section of a •wing skin. The entire side shown
was masked,, and some areas of the other side
were etched. In Figure 586, the masking is
being stripped from a milled part.
ETCHANT SOLUTIONS
Etchants range from sodium hydroxide solution
for aluminum to aqua regia for stainless steel.
For milling a specific metal, the concentration
of the chemical in the solution may vary widely
between different operators; however, each
operator controls the concentration of his solu-
tion to within very close limits. The concen-
tration of the solution affects the milling rate;
therefore, it must be closely controlled to ob-
tain the desired rate. For milling aluminum,
the solutions in use contain from 7 to 30 per-
cent sodium hydroxide. For milling magnesi-
um, dilute sulfuric acid solutions are adequate.
Stainless steels require strong solutions, usu-
ally aqua regia fortified with sulfuric acid. In
most of the milling solutions, surface-active
agents are used to ensure smooth, even cuts.
The surface-active agents also reduce the ten-
dency toward mist formation by reducing the
surface tension of the solution. The solutions,
during milling operations, are generally main-
tained at constant temperatures ranging from
105°F to 190°F.
INCOMING PARTS
1 SOLVENT DEGREASER 180 F
2 HOT ALMLIHE CtHNW ISO
3 COLD WATER RINSE
4 CHROMIC ACID
5 HOT WA7ER RINSE ISO f
ETCHING AREA
9 MASKING FOR SPRAY COAT 13 ETCHING TANK
TEMPLATE MASKING
14 CQNTKH. PANEL
10 PAINT BOOTH
11 DRYING OVEN 15 COLD WATER RINSE
12 TAPER ETCHING TANK 16 SMUT REMOVAL
17 MASKING REMOVAL
13 INSPECTION
19 CENTRIFUGE
20 BYPRODUCT
This process is patented and licensed by
Turco Products Co., Wilmington, Calif.
Figure 583. A flow diagram showing the typical
steps necessary to the chem-mi11 ing process
(Scheer, 1956).
Figure 584. Inspection of milled parts. The in-
strument measures the metal thickness before the
masking is removed (U.S. Chemical Milling Corpora-
tion, Manhatten Beach, Calif.).
THE AIR POLLUTION PROBLEM
The air contaminants emitted in the prepara-
tion of metals by chemical milling consist of
mists, vapors, gases, and organic solvents.
GPO 806—614—27
-------
Chemical Milling
803
Figure 585. Stripping masking from a section of a wing skin of a
B-58. The entire side shown was masked. Some areas of the other
side were milled (U.S. Chemical Milling Corporation, Manhatten Beach,
Cal if.).
Figure 586. Masking being stripped, milled parts with masking still
in place, and milled parts with masking removed (U.S. Chemical Mil-
ling Corporation, Manhatten Beach, Calif.).
Mists
A mist of the etching solution used in a milling
process is discharged from the milling tank
owing to entrainment of droplets of the solution
by the gas bubbles formed by the chemical
action of the etchant on the metal. The amount
of mist generated depends upon factors such as
the nature of the chemical reaction, the solu-
tion temperature, and the surface tension of the
solution. Since the solutions from which the
mists are formed are very corrosive, the mists,
too, are very corrosive and are capable of caus-
ing annoyance, or a nuisance, or a health hazard
to persons, or damage to property.
-------
804
CHEMICAL, PROCESSING EQUIPMENT
Vapors
Some of the acid solutions used, such as hydro-
chloric and nitric, have high vapor pressures
at the temperatures used for the milling pro-
cess; therefore, appreciable amounts of acid
vapors are discharged. Unlike the discharge
of mists, which occurs only during the milling,
the vapors are discharged continuously from the
hot solution. Under certain atmospheric condi-
tions, the vapors condense, forming acid mists
in the atmosphere.
Gases
Since hydrogen is formed in chemical milling,
proper ventilation must be provided to prevent
the accumulation of dangerous concentrations of
this gas.
Solvents
Organic solvent vapors may be emitted from the
vapor degreaser, the maskant area, and the
curing station in the cleaning and masking pro-
cesses. This type of air contaminant, and the
method of controlling it are described elsewhere
in this manual. Alkaline cleaning, pickling, and
passivating tanks from the other phases of the
cleaning processes have been found to be minor
sources of air pollution.
HOODING AND VENTILATION REQUIREMENTS
The air contaminants released from chemical
milling tanks can be captured by local exhaust
systems. Since open tanks are used to provide
unobstructed working area, most exhaust sys-
tems employ slotted hoods to capture the mists
and vapors. In designing slot hoods for chem-
ical milling equipment, it is particularly im-
portant to provide for the elimination of exces-
sive cross-drafts as well as for adequate dis-
tribution of ventilation along the entire length
of the hoods. The minimum ventilation rates
previously mentioned in Chapter 3 are for tanks
located in an area having no cross-drafts. If
the tank is to be located outside or in a very
drafty building, either the ventilation rate will
have to be greatly increased or baffles must be
used to shield the tank from winds or drafts.
In some instances, both baffles and increased
ventilation are needed.
Adequate distribution of ventilation along the
entire length of a slot can be attained by pro-
viding a high slot velocity and a relatively low
plenum velocity. The slot velocity should be at
least 2, 000 fpm, and the plenum velocity should
be not more than half of the slot velocity. With
hoods more than 10 feet in length, either multi-
ple takeoffs or splitter vanes are needed. Enough
takeoffs or splitters should be used to reduce
the length of the slot to sections not more than
10 feet long.
Under excessively drafty conditions, a hood en-
closing the tank can be used to advantage. The
hood should cover the entire tank and have suf-
ficient height to accommodate the largest metal
sections that can be handled in the tank. Vari-
ous methods have been used to get work into and
out of the tank. In one installation, the hood has
doors on one end, and a monorail, suspended
below the hood roof, that runs out through the
doors. The work is carried on the monorail
into the hood and above the solution. After the
work is lowered into the solution, the doors are
closed, when necessary, to ensure complete
capture of the air contaminants created. In
another installation, the hood is left open on one
end, and a slot hood placed across the opening.
The top of the hood is slotted to provide for the
movement of the crane cable. This slot is
nominally closed with rubber strips, which are
pushed aside by the cable during movement of
the crane.
AIR POLLUTION CONTROL EQUIPMENT
Many types of wet collectors that can control
the emissions from chemical milling tanks are
commercially available. The one most common-
ly used is the spray and baffle type, owing prob-
ably to its low cost and ease of coating with cor-
rosion-inhibiting materials. Moreover, the oper-
ation and maintenance of this type are simple
and inexpensive compared with those of other
types of scrubbers.
Figure 587 shows an exhaust and mist control
system employing two scrubbers, one for each
side of a 24-foot-long by 6-foot-wide tank used
for chemically milling stainless steel and ti-
tanium. The etching solution is a mixture of
hydrochloric, nitric, and sulfuric acids and is
heated to 150 °F. Acid vapors discharged from
the solution are captured by slot hoods, one on
each side of the tank. The ducts from each
hood exit downward from the center. Each hood
has four splitter vanes, which divide it into four
sections. The overall hood length is 24 feet,
the end-sections and those adjacent being 4 feet
long each, and the center section being 8 feet
long. Distribution of ventilation is excellent.
Each hood is supplied with 18, 000 cfm ventila-
tion, and the slot is sized to give an intake ve-
locity of 2, 000 fpm. The plenum velocity is
less than 1,000 fpm. It is estimated that this
system provides sufficient ventilation to capture
-------
Chemical Milling
805
Figure 587. A tank used for the chemical milling of
stainless steel, and part of its air pollution
control system. The hoods, ductwork, and scrubbers
shown are made entirely of polyester resin rein-
forced with fiberglas. The fans and discharge ducts,
not shown, are steel-coated with polyester resin.
(U.S. Chemical Mi 11 ing Corporation, Manhatten Beach,
Calif.).
at least 95 percent of the vapors emerging from
the process.
The scrubbers are of the spray and baffle type,
as shown in Figure 588. They are cylindrical,
two baffles forming three concentric chambers.
Gases enter at the top and flow down through
the center cylindrical section. Water from a
bank of sprays scrubs the gases as they enter
this section. The bottom of the scrubber is
filled with water to a depth of 1 foot. The gases
and scrubbing water flow downward through the
center section and impinge on the water. The
gases turn 180 degrees and flow upward through the
second chamber. Most of the scrubbing water
remains in the sump. The depth of water in the
sump is maintained at a uniform level with a
float valve and an overflow line. The scrubber
is equipped with a pump to circulate the sump
water to the sprays. In this installation, how-
ever, only fresh water is used, the sump being
kept full and overflowing all the time.
The gases flow upward through the second section and
over the second baffle. They turn 180 degrees to enter
the third section. In the third section, the gas-
es flow down and around to the outlet port. Most
of the entrained moisture entering the second
section is removed either by impingement on
the walls of that section or by centrifugal im-
pingement during the 180-degree change of direction
into the third section. The last of the entrained
water is deposited on the walls of the third sec-
tion. The gases then flow from the scrubber to
the fan, from 'which they are discharged to the
atmosphere through ducts.
The hoods, the scrubber, and the ductwork con-
necting the hoods to the scrubbers and the scrub-
bers to the fans are made entirely of polyester
resin reinforced with glass fibers. The fans and
discharge ducts are made of steel coated with
polyester resin.
The existing system provides satisfactory con-
trol of the vapors. It captures an estimated 95
percent of the vapors at the tank, and the gases
discharged have only a slight acid odor.
Corrosion Problems
Whenever moisture is present in an exhaust
system, the iron or steel surfaces should be
coated to prevent corrosion. Since, however,
zinc is soluble in both acid and alkaline solu-
tions, galvanized iron cannot be used when
chemical milling tanks are vented. A coating
such as polyvinylchloride (PVC), which is not
attacked by either dilute acids or dilute alkalies,
should be used. It has been found, however,
that the PVC linings in ducts and scrubbers can-
not withstand the strongly oxidizing acids used
for stainless steel and titanium milling. These
highly corrosive acids have been successfully
handled in exhaust systems made of polyester
resins reinforced with fiberglas. Hoods, ducts,
and scrubbers are available made entirely of
polyester-fiberglas material. Figure 587 shows
an air pollution control system venting a 24-foot -
long tank for stainless steel chemical milling.
The hoods, ductwork up to the blowers, and the
scrubbers are made entirely of polyester-fiber-
glas material. The steel blowers and discharge
ducts are coated with polyester resin. Some
blower manufacturers are now advertising blow-
ers with scrolls made entirely of polyester-
fiberglas and with steel wheels coated with the
same material.
-------
806
CHEMICAL PROCESSING EQUIPMENT
AUTOMATIC
FLOAT
VALVE
WATER OVERFLOW
SUPPLYV
MOTOR AND
PUMP ASS'Y
DRAIN
ACCESS ,
DOOR\
INNER -*
SPRAYT
NOZZLES
-I GAGE
j ,0 7sTANDPIPE
AND OVERFLOW
DRAIN
Model
24
30
42
60
72
Motor3
hp
1
1
14
14
2
Pump
gpm
14
14
18
20
22
Drain h
weight"
110
260
380
600
963
A
24
30
42
BO
72
B
12
18
24
32
42
0
22
324
43
514
844
D
12
18
24
32
42
E
12J
12
12
12
12
F
694
80
92
102
132
G
9
101
12
14
24
H
9
9
8
12
18
J
224
38
42
50
60
K
35!i
40
46
55
764
L
6
9
(2
18
24
Spray
nozzl es
7
7
9
10
1 1
Drain
size, in.
1
1
1
1
1
Min.-range-max.
fpm
1,000
1,000
1,000
1,000
1,000
cfm
800
1,700
3,000
5,200
9,500
fpm
3,000
3,000
3,000
3,000
3,000
cfm
2,300
5,200
9,000
16,000
28,000
"Motor is 440/220 volts, 3 phase, 60 cycle. Exhaust fan and motor furnished upon request.
Does not include recirculat ing motor and pump.
Figure 588. A scrubber used to control the acid vapors discharged
from a tank used to mill stainless steel (Lin-0-Coat scrubber,
manufactured by Diversified Plastics, Inc., Paramount, Calif.).
-------
REFERENCES
Adams, R. L. 1964.
Application of Baghouses to Electric Furnace Fume Control. JAPCA. 14(8):299-302 (Aug).
Air Moving and Conditioning Assn. , Inc. 1963.
Bulletin 210. 205 W. Touhy Ave. , Park Ridge, 111.
Air Pollution Handbook.
See Magill et al.
Alden, J. L. 1948.
Design of Industrial Exhaust Systems. The Industrial Press, New York, N. Y.
Alexander, W. H. , andR.L. Bradley. 1958.
Can You Justify a CO Boiler? Petrol. Refiner. 37:107-12 (Aug).
Allen, G. L. , F.H. Viets, and L. C. McCabe. 1952.
Control of Metallurgical and Mineral Dusts and Fumes in Los Angeles County, California. Bureau
of Mines Information Circular 7627, U.S. Department of Interior, Washington, D. C. (Apr).
American Air Filter Company, Inc. 1958.
American Filter Handbook. Louisville, Ky. 40208.
American Air Filter Company, Inc. 1964.
Calendar-Memo-Handbook. Louisville, Ky. 40208.
American Conference of Governmental Industrial Hygienists. I960.
Threshold Limit Values for 1960. Arch. Environ. Health. 1:140-44 (Aug).
American Foundrymen's Association. 1949.
Handbook of Cupola Operation. American Foundrymen's Association, Chicago, 111. (Des Plaines,
111.)
American Gas Association Laboratories. 1940.
Research in Fundamentals of Atmospheric Gas Burner Design. Research Bulletin No. 10. Clevelai
Ohio (Mar).
American Petroleum Institute. I960.
Recommended Practice for the Design and Installation of Pressure-Relieving Systems in Refineries
Part I. 2d ed. (Sept).
American Petroleum Institute. 1962a.
Evaporation Loss From Fixed-Roof Tanks. Bulletin 2518 (June).
American Petroleum Institute. 1962b.
Evaporation Loss From Floating-Roof Tanks. Bulletin 2517 (Feb).
American Petroleum Institute. 1962c.
Evaporation Loss From Low-Pressure Tanks. Bulletin 2516 (Mar).
American Petroleum Institute. 1962d.
Use of Plastic Foam to Reduce Evaporation Loss. Bulletin 2515 (Jan).
American Petroleum Institute - American Society of Mechanical Engineers. 1951.
Unfired Pressure Vessel Code. 5th ed.
American Petroleum Institute. Division of Refining. 1951.
Manual on Disposal of Refinery Wastes. Vol III. Chemical Wastes. 2d ed. New York, N. Y.
807
-------
808 References - American
American Petroleum Institute. Division of Refining. 1957.
Manual on Disposal of Refinery Wastes. Vol II. Waste Gases and Particulate Matter. 5th ed.
New York, N. Y.
American Society of Heat-ing, Refrigerating, and Air-Conditioning Engineers, Inc. 1963.
ASHRAE Guide and Data Book. New York, N. Y.
American Society of Mechanical Engineers. 1962.
ASME Unfired Pressure Vessel Code. Section VIII. New York, N. Y.
American Society for Testing Materials. 1958.
Specifications for Pig Lead, B29-55 In: 1958 Book of ASTM Standards, Part 2, Non-Ferrous
Metals. Philadelphia, Pa.
American Society for Testing Materials. 1959.
Standard Method for Measurement of Odor in Atmospheres (Dilution Method). Designation D1391-57.
In: ASTM Standards on Methods of Atmospheric Sampling and Analysis. Prepared by ASTM Com-
mittee D-22. Philadelphia, Pa.
American Water Works Association. 1951.
Standard Specifications for Coal-Tar Enamel Protective Coatings for Steel Water Pipe. New York,
N. Y.
Anderson, E. lv^-±.
Some Factors and Principles Involved in the Separation and Collection of Dust, Mist and Fume From
Gases. Trans. Am. Inst. Chem. Engrs., Part 1. 16:69-86.
Anderson, H. E. 1958.
Filtering Radioactive Particles From Stack Gas, Air Conditioning, Heating, a.nd Ventilating, 55:71-
77 (Feb).
Anderson, R. J. 1925.
The Metallurgy of Aluminium and Aluminium Alloys. Henry Gary Baird and Co. , Inc. , New York,
N. Y.
Anderson, R. J. 1931.
Secondary Aluminum. The Sherwood Press, Inc. , Cleveland, Ohio.
Andrews, A.I. 1961.
Porcelain Enamels. Garrard Press, Champaign, 111.
Anonymous. 1950.
Repeal Brass Installs Unique Fume Control System. Western Metalworking. 8:32 (Mar).
Anonymous. 1957.
Metal Finishing-Guidebook-Directory. Metals and Plastics Publications, Inc., Westwood, N, J.
Anonymous. 1961.
1961 Petrochemical Handbook Indexes. Alkyl Aryl Sulfonates. Hydrocarbon Processing and
Petroleum Refiner. 40:217 (Nov).
Arrandale, R. S. 1962.
Can Presintering Solve Glass Batch Problems? Ceram. Ind. 78:82, 83, 127, 138 (Apr).
ASHRAE Guide and Data Book.
See American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc. 1963.
The Asphalt Institute. 1954.
Asphalt Protective Coatings for Pipelines. College Park, Md.
The Asphalt Institute. 1957.
Specifications and Construction Methods for Hot-Mix Asphalt Paving for Streets and Highways.
College Park, Md. (May).
Atomic Energy Commission.
See Ward, 1952.
-------
References - Bonamassa 809
Badger, W. L. , and W. L. McCase. 1936.
Elements of Chemical Engineering. 2d ed. McGraw-Hill Book Co. , Inc. , New York, N. Y.
Bailey, T. W. 1957.
A Report on the Use of Silicone Lubricants at the Wheaton Glass Company. Glass Ind. 38:433-40
(Aug).
Baker, T. C. 1935.
Distillation and Absorption in Packed Columns. Ind. Eng. Chem. 27:977 (Aug).
Baque, H. W. 1954.
Cut Heating Costs 10% With Proper Insulation. Ceram. Ind. 62:77, 111 (Mar).
Barber, J. C. 1958.
Corrosion Problems in the Manufacture of Phosphoric Acid From Elemental Phosphorus. Corro-
sion. 14:21-26 (Aug).
Barker-Greene Company. I960.
Drier Principles. Aurora, 111.
Barnebey-Cheney Co.
Bulletin T-642. Columbus 19, Ohio.
Barnhart, D.H., andE.K. Diehl. I960.
Control of Nitrogen Oxides in Boiler Flue Gases by Two-Stage Combustion. JAPCA. 10:397-406
(Oct).
Barry, H.M. I960.
Fixed-Bed Adsorption. Chem. Eng. 67:105-07 (Feb 8).
Earth, E.J. 1958.
How to Make Roofing Asphalts. Petrol. Refiner. 37:172 (Mar).
Begeman, M. L. 1947.
Manufacturing Processes. 2d ed. , John Wiley and Sons, Inc., New York, N. Y.
Benedict, M. , and T. H. Pigford. 1957.
Nuclear Chemical Engineering. McGraw-Hill Book Co. , Inc. , New York, N. Y.
Beychok, M. 1953.
Build a Flare for Under $5000. Petrol. Processing. 8:1162-63 (Aug).
Bidlack, V. C. , and E. W. Fasig. 1951.
Paint and Varnish Production Manual. John Wiley and Sons, New York, N. Y.
Bienstock, D. , L. W. Brunn, E.M. Murphy, andH.E. Benson. 1958.
Sulfur Dioxide--Its Chemistry and Removal From Industrial Waste Gases. Bureau of Mines In-
formation Circular 7836. U. S. Department of Interior, Washington, D. C.
Bienstock, D. , andF.J. Field. I960.
Bench-Scale Investigation on Removing Sulfur Dioxide From Flue Gases. JAPCA. 10:121-25 (Apr).
Bingham, J. E. 1958.
Check Your Procedures on Rupture Disc Installation. Chem. Eng. 65:143-45 (Apr 17).
Bodurtha, F. T. , Jr. 1958.
Flare Stacks - How Tall? Chem. Eng. 65:177-80 (Dec 15).
Bonamassa, F. , and Y. S. Yee. 1957.
Emission of Hydrocarbons to the Atmosphere From Cooling Towers. Report No. 5. Joint District,
Federal, and State Project for the Evaluation of Refinery Emissions. Los Angeles County Air
Pollution Control District. Los Angeles, Calif. (Aug).
-------
810 References - Brandon
Brandon, D. B. 1959.
Developing Mathematical Models for Computer Control. ISA Journal 6:70-73.
Brandt, A.D. . 1947.
Industrial Health Engineering. John Wiley and Sons, Inc. , New York, N. Y.
Brief, R. S. , A.H. Rose, and D. G. Stephan. 1956.
Properties and Control of Electric-Arc Steel Furnace Fumes. JAPCA. 6:220-24 (Feb).
Brink, J. A. , Jr. 1959.
Monsanto Solves Air Pollution Problems With New Fiber Mist Eliminator. Chem. Eng. 66:183-86
(Nov 19).
Brown, C. O. , and R. B. Wainright. 1952.
Synthetic Fluid Cracking Catalyst; Their Application and Utilization. Oil Gas J. 51(30): 133-37.
Brown, G. W, , and J. E. Sublett. 1957.
Union Oil Company Builds New Waste-Water Facilities. Ind. Water and Wastes. 2:6-8 (Jan-Feb).
Bruce, W. L. , andW.A. Schubert. 1956.
Rotary Compressors and Vacuum Pumps. Chem. Eng. 63:224-26 (June).
Brumbaugh, A. K. , Jr. 1947.
Smokeless Burning of Refinery Vent Gases. Petrol. Processing. 2:181.
Bulcraig, W.R., and J, B. Haigh. 1961.
The Reduction of Air Pollution During the Reversal of a Producer Gas-Fired Regenerative Glass
Tank Furnace. J. Inst. Fuel. 34:73-78 (Feb).
Bussard, W. A. 1956.
Evaporation Losses and Their Control in Storage. Petrol. Processing. 11:104-26 (July).
Cameron, F. 1952.
Cottrell, Samaritan of Science. Doubleday and Co. , Inc. , Garden City, New York.
Campbell, W.W. , and R. W. Fullerton. 1962.
Development of an Electric-Furnace Dust-Control System. JAPCA. 12:574-77, 590 (Dec).
Caplan, K.J. 1954.
A Self-Cleaning Air Filter. Chem. Eng. Progr. 50:409-14 (Aug).
Caplan, K.J. I960.
Predicting the Performance of Reverse-Jet Filters. Air Conditioning, Heating, and Ventilating.
57:62-65 (Oct).
Carbon Products Division, Union Carbide Corporation. 1955.
Solvent Recovery by the "Columbia" Activated Carbon System. 270 Park Ave. , New York, N. Y.
Case, E.L. 1956.
Reciprocating Compressors. Chem. Eng. 63:216-23 (June).
Castler, L. A. , E.K. Daniels, andJ.R. Lutz. 1956.
Pollution Control at Ferndale, Washington. Presented at 21st American Petroleum Institute
Division of Refining Midyear Meeting, May 14-17, 1956.
Chass, R.L. 1959.
Engineering Control of Air Pollution in Los Angeles County. Presented at the Sanitary Engineer-
ing Division Program of the American Society of Civil Engineers, Los Angeles, Calif. , Feb. 1959.
Chass, R.L., andR.E. George. I960.
Contaminant Emissions From the Combustion of Fuels. JAPCA. 10:34-43 (Feb).
-------
References - Dalla Valle 811
Chicago Bridge and Iron Company. 1959.
The Storage of Volatile Liquids. Technical Bulletin No. 20. Oakbrook, Hinsdale P. O. , 111.
Chilton, C. H. 1949.
Cost Data Correlated. Chem. Eng. 46:97-106 (June).
Clapp, W. H. , and S. Clark. 1944.
Engineering Materials and Processes. International Textbook Co. , Scranton, Pa.
Clarke, L. 1947.
Manual for Process Engineering Calculations. McGraw-Hill Book Co. , Inc. , New York, N. Y.
Claude, R.E. 1956.
Axial Compressors. Chem. Eng. 63:212-15 (June).
Clement, R. L. 1961.
Selection, Application and Maintenance of Cloth Dust Filters. Plant Eng. 15:92-97 (Aug).
Cleveland, D. L. 1952.
Design and Operation of a Steam Inspirating Flare. Presented at 17th American Petroleum Insti-
tute Division of Refining Midyear Meeting, May 13, 1952.
Committee on Industrial Ventilation, I960.
Industrial Ventilation. 6th ed. American Conference of Governmental Industrial Hygienists,
Lansing, Mich. (4th ed. , 1956; 7th ed., 1962)
Conison, J. I960.
MOW to Design a Pressure Relief System. Chem. Eng. 67:109-14 (July 25).
Cornell, D. , W. G. Knapp, and J. R. Fair. I960.
Mass Transfer Efficiency - Packed Columns. Pts. 1 and 2. Chem. Eng. Progr. 56:68-74 (July);
48-53 (Aug).
Cottrell Electrical Precipitators.
See Western Precipitation Corporation, 1952.
Coulter, R.S. 1954.
Smoke, Dust, Fumes Closely Controlled in Electric Furnaces. Iron Age. 173:107-10 (Jan 14).
Cowan, D.W. , H. J. Thompson, H. J. Paulus, and P. W. Mielke, Jr. 1963.
Bronchial Asthma Associated With Air Pollutants From the Grain Industry. JAPCA. 13:546-52 (Nov).
Crabaugh, H. R. , A. H. Rose, Jr., and R. L. Chass. 1954.
Dust and Fumes From Gray Iron Cupolas--How They Are Controlled in Los Angeles County. Air Re-
pair. 4:125-30 (Nov).
Crane Company. 1957.
Flow of Fluids. Technical Paper No. 410. Chicago, 111.
Creesey, M.S., andA.K. Lyle. 1956.
What's New in Container Plant Design. Ceram. Ind. 67:98-101 (Sept).
Crumley, P. H. , and A. W. Fletcher. 1956.
The Formation of Sulphur Trioxide in Flue Gases. J. Inst. Fuel. 29:322-27 (Aug).
Cumiskey, J.E. 1956.
Reciprocating Compressor Special Report, Introduction to Principles. Petrol. Refiner. 35:120-23
(Nov).
Dalla Valle, J. M. 1952.
Exhaust Hoods. 2d ed. The Industrial Press, New York, N. Y.
-------
812 References - Dalla VaLle
Dalla Valle, J.M., and H. C. Dudley. 1939.
Evaluation of Odor Nuisances in Manufacturing of Kraft Paper. Public Health Repts. 54:35-43
(Jan 13).
Decker, W. H. 1950.
Safe, Smokeless Combustion Features Waste Gas Burner at Sinclair Refinery. Petrol. Process-
ing. 5:965-66, 968 (Sept).
Deckert, I. S. , R. G. Lunche, and R. C. Murray. 1958.
Control of Vapors From Bulk Gasoline Loading. JAPCA. 8:223-33 (Nov).
de Lorenzi, O. 1947.
Combustion Engineering. Combustion Engineering--Superheater, Inc. , New York, N. Y.
(de Lorenzi, O. , 1947. Courtesy Peabody Engineering Corp.).
Des Jardins, P. R. 1956.
Handling Compressible Fluids in Chemical Processing. Chem. Eng. 63:178-87 (June).
Deutsch, W. 1922.
Movement and Charge of Electrical Carriers in a Cylindrical Condenser. Ann. Physik, 68,
4th Ser. (373, orig. ser.):335-44.
Diehl, J.E. 1957.
Calculate Condenser Pressure Drop. Petrol. Refiner. 36:147-53. (Oct).
Dolman, R.E. 1952.
Pumps. Chem. Eng. 59:155-69 (Mar).
Donahue, D.A. 1956.
Heat Exchangers. Petrol. Processing. 11:102-32 (Mar).
Dow Chemical Co. 1963.
Dowtherm Handbook. Midland, Mich.
Drinker, P., and T. Hatch. 1954.
Industrial Dust. 2d ed. McGraw-Hill Book Co. , Inc., New York, N. Y.
Driskell, L. R. I960.
Design Tips for Piping of Pressure Relief Devices. Petrol. Refiner. 39:127-32 (July).
Duecker, W. W. , andJ.R. West. 1959.
Manufacture of Sulfuric Acid. 1st ed. Reinhold Publishing Corp. , New York, N. Y.
Eastwood, L. W. 1946.
Gas in Light Alloys. John Wiley and Sons, New York, N. Y.
Edmister, W. C. 1948.
Hydrocarbon Absorption and Fractionation Process Design Methods. Petrol. Engr. 20(3): 193-98,
200 (Dec).
Elliott, J., N. Kayne, and M. LeDuc. 1961.
Experimental Program for the Control of Organic Emissions From Protective Coating Operations.
Report No. 8. Los Angeles County Air Pollution Control District, Los Angeles, Calif. (Jan).
Elonka, S. 1955.
Manual on Packing. Power. 99:107-30 (Mar).
Elonka, S. 1956.
Manual on Mechanical Seal. Power. 100:109-32 (Mar).
-------
References - Gilbert 813
Environmental Sciences and Engineering, Division of Air Pollution.
Control of Particulate Emissions (Training Course Manual). Robert A. Taft Sanitary Engineer-
ing Center, Cincinnati, Ohio.
Ermenc, E. D. 1956.
Wisconsin Process Pebble Furnace Fixes Atmospheric Nitrogen. Chem. Eng. Progr. 52:149 (Apr).
Ezekiel, M. 1941.
Methods of Correlation Analysis. 2d ed. John Wiley and Sons, Inc. , New York, N. Y.
Fabrianio, W. L. 1961.
How to Combat Checker Clogging and Carry-Over. Ceram. Ind. 77:102, 103, 127 (Sept).
Fairlie, A.M. 1936.
Sulfuric Acid Manufacture. 5th ed. Reinhold Publishing Corp. , New York, N. Y.
Fairs, G. L. 1953.
High Efficiency Fibre Filters for the Treatment of Fine Mists. Trans. Inst. Chem. Engrs.
36:476-85.
Faith, W. L. , N. A. Renzetti, andL.H. Rogers. 1957.
Third Technical Progress Report, Report No. 17. Air Pollution Foundation, Los Angeles, Calif.
(Mar).
Field, J. H. , L.M. Brunn, W. P. Haynes, and H. E. Benson. 1957.
Cost Estimates of Liquid-Scrubbing Processes for Removing Sulfur Dioxide From Flue Gases.
JAPCA. 7:109-15 (Aug).
Filter Fabric Facts.
See Willington Sears Co., 1954.
Fischer, J. 1957.
Air Conveying of Dry Materials. Northwest. Miller (Minneapolis, Minn.). Dec 31, 1957.
Fischer, J. 1958.
Practical Pneumatic Conveyor Design. Chem. Eng. 65:114-18 (June 2).
Fisher, M. M. , and E.G. Moriarty. 1953.
Waste Disposal in Urban Areas. Industry and Power (St. Joseph, Mich.). 64:83-85 (Apr).
Fox, E.A., and V.E. Gex. 1957.
Procedure for Measuring Odor Concentration in Air and Gases. JAPCA. 7:60-61 (May).
Frankenburg, T. T. 1963.
Air Pollution From Power Plants and Its Control. Combustion. 34(8):28-31 (Feb).
Frederick, E. R. 1961.
How Dust Filter Selection Depends Upon Electrostatics. Chem. Eng. 68:107 (June 26).
Friedlander, S.K., L. Silverman, P. Drinker, and M. W. First. 1952.
Handbook on Air Cleaning. U.S. Atomic Energy Commission, Washington, D. C. (Sept). (AECD-
3361; NYO-1572).
Friedman, S. H. 1959.
Use These Computation Shortcuts. Chem. Cng. 66:149-50 (Sept 21).
Friedman, S.J., and W. R. Marshall, Jr. 1949.
Studies in Rotary Drying. Chem. Eng. Progr. 45:482 (Aug).
Gilbert, N. , and F. Daniels. 1948.
Fixation of Atmospheric Nitrogen in a Gas Heated Furnace. Ind. Eng. Chem. 40:1719-23 (Sept).
-------
814 References - Gillespie
Gillespie, G.R., and H. F. Johnstone. 1955.
Particle Size Distribution in Some Hygroscopic Aerosols. Chem. Eng. Prog. 51:78 (Feb).
Glasstone, S. 1946.
Textbook of Physical Chemistry. D. Van Nostrand Co. , Inc. , Princeton, N. J.
Glaubitz, F. 1963.
The 2conomic Combustion of Sulfur-Containing Heating Oil. Parts land II. Combustion. 34(7):31-
35 (Jan); 34(9):25-32 (Mar).
Gosline, C. A. , L. L. Falk, andE.N. Helmers. 1956.
Evaluation of Weather Effects Section 5. In: Air Pollution Handbook. McGraw-Hill Book Co. , Inc.
New York, N. Y.
Green, W.G. 1952.
Smokeless Burning of Refinery Vent Gases. Presented at 17th American Petroleum Institute
Division of Refining Midyear Meeting, May 13, 1952.
Griswold, J. 1946.
Fuels, Combustion, andFurnaces. Isted. McGraw-Hill Book Co., Inc., New York, N.Y.
Griswold, S. S. , and I. Weisburd, eds. 1962.
Air Pollution Control Field Operations Manual Guide for Inspection and Enforcement. Public
Health Service Publication No. 937. Division of Air Pollution, Public Health Service, U.S.
Department of Health, Education, and Welfare, Washington, D. C.
Gumz, W. 1950.
Overfire Air Jets in European Practice. Combustion. 22:39-48 (Apr).
Gunsaulus, R.K. 1958.
How to Record and Control Furnace Flue Gas Automatically. Ceram. Ind. 70:70-73, 93, 95 (Mar).
Haagen-Smit, A. J. 1958.
Studies of Air Pollution Control by Southern California Edison Company. JAPCA. 7:251-55 (Feb).
Hajek, J. D. , and E. E. Ludwig. I960.
How to Design Safe Flare Stacks. Parts I and II. Petrol. Engr. 32;C31-C38; C44-C51 (June,
July).
Haley, R. H. 1949.
How to Control Fumes in Non-Ferrous Melting. Foundry. 77:118, 121 (Sept).
Hannaman, J.R., andA.J. Etingen. 1956.
Here's a Workable Smokeless Flare System. Petrol. Processing. 11:66-69 (Mar).
Hauck Manufacturing Company. 1953.
Industrial Combustion Data. Brooklyn, N. Y.
Hemeon, W. C. L. 1955.
Plant and Process Ventilation. Isted. The Industrial Press, New York, N.Y. (2d ed, 1963).
Hansen, J.E. 1932.
The Advanced Technique of Porcelain Enameling. Enamelist Publishing Co. . Cleveland, Ohio.
Herrick, R. A. 1963.
A Baghouse Test Program for Oxygen Lanced Open Hearth Fume Control. JAPCA. 13:28-32
•(Jan).
Hersey, H.J., Jr. 1955.
Reverse-Jet Filters. Ind. Chemist. 31:138 (Mar).
-------
References - Kerns 815
Hicks, T. 1951.
Power's Handbook on Fans. Power. 95:87-102 (Oct).
Holzbock, W. G. 1959.
Control Valve Construction. Chem. Eng. 66:135-38 (Apr 6).
Hougen, O.A., andK.M. Watson. 1945.
Industrial Chemical Calculations. 2d ed. John Wiley and Sons, Inc., New York, N. Y.
How, H. 1956.
How to Design Barometric Condensers. Chem. Eng. 63:174-82 (Feb).
Huebner, W. O. 1959.
How Esso Cuts Flare Gas Losses. Air Eng. 1:38-39 (Apr).
Hunter, W. L. 1959.
Bottom-Loading Tank Trucks Successful. Petrol. Refiner. 38:171-72. (Oct).
Industrial Hygiene Codes Committee. 1938.
Fundamentals of Design, Construction, Operation and Maintenance of Exhaust Systems. American
Foundrymen's Association, Chicago, 111. (Des Plaines, 111.)
Industrial Ventilation.
See Committee on Industrial Ventilation, I960.
Ingels, R.M., N. R. Shaffer, andJ.A. Danielson. I960.
Control of Asphaltic Concrete Plants in Los Angeles County. JAPCA. 10:29-33 (Feb).
Jenny, J. P. 1951.
Smelter Fume Control. Proceedings of Forty-Fourth Annual Convention, Air Pollution and Smoke
Prevention Association of America (Air Pollution Control Association), Roanoke, Va. , May 7-10,
1951.
Jensen, L. B. 1945.
Microbiology of Meats. 2d ed. Garrard Press, Champaign, 111.
Johnstone, H.F., andM.H. Roberts. 1949.
Deposition of Aerosol Particles From Moving Gas Streams. Ind. Eng. Chem. 41:2417-23 (Nov).
Jones, C. T. , and P. S. Viles. 1952.
Estimating Oil Losses by Atmospheric Evaporation From Refinery Separator Surfaces. Petrol.
Refiner. 31:117-20 (Jan).
Kane, John M.
Manual of Exhaust Hood Designs. Vol. 2. Bull. No. 270-E2B. American Air Filter Co., Inc.,
Louisville, Ky. 40208.
Kanter, C. V. , R. G. Lunche, F. Bonamassa, B.J. Steigerwald, andR.K. Palmer. 1958.
Emissions to the Atmosphere From Petroleum Refineries in Los Angeles County. Report No. 9.
Joint District, Federal, and State Project for the Evaluation of Refinery Emissions. Los Angeles
County Air Pollution Control District, Los Angeles, Calif.
Kent, R. T. ed. 1938.
Mechanical Engineers Handbook, llth ed. John Wiley and Sons, Inc, New York, N. Y. (12th ed. ,
1950).
Kern, D. Q. 1950.
Process Heat Transfer. McGraw-Hill Book Co. , Inc. , New York, N. Y.
Kerns, G. D. I960.
New Charts Speed Drum Sizing. Petrol. Refiner. 39:168-70 (July).
-------
816 References - Kirk
Kirk, R.E., and D. F. Othmer. eds. 1947.
Encyclopedia of Chemical Technology. 1st ed. Interscience Encyclopedia, Inc. , New York, N. Y.
(Volumes 2, 5, 7, 8, 9, 10, and 14).
Krenz, W. B. , R.C. Adrian, and R. M. Ingels. 1957.
Control of Solvent Losses in Los Angeles County. Proceedings of the Golden Jubilee Meeting of
Air Pollution Control Association, St. Louis, Mo., June 2-6, 1957.
Lapple, C.E. 1951.
Processes Use Many Collection Types. Chem. Eng. 58:145-51 (May).
Lapple, C.E. 1963.
Dust and Mist Collection. In: Chemical Engineers' Handbook. Perry, J. H. , ed. McGraw-Hill
Book Co. , Inc. See also, Air Pollution Abatement Manual. C. A. Gosline, ed. Manufacturing
Chemists' Association, Inc., Washington, D. C. 1952.
Lemke, E.E., W.F. Hammond, and G. Thomas. I960.
Air Pollution Control Measures for Hot Dip Galvanizing Kettles. JAPCA. 10(L):70-76 (Feb).
Leonard, E. S. 1956.
Centrifugal Compressors. Chem. Eng. 63:206-11 (June).
Leva, M. 1953.
Tower Packings and Packed Tower Design. The United States Stoneware Co. Akron, Ohio.
Liberman, J.A. 1957.
Engineering Aspects of the Disposal of Radioactive Wastes From the Peace-Time Applications of
Nuclear Technology. Am. J. Public Health. 47:345-51 (Mar).
Lieblich, N. 1953.
Selection of Diaphragm Control Valves and Regulators. Proceedings of the Southwestern Gas
Measurement Short Course, University of Oklahoma.
Lobo, W. E., L. Friend, F. Hashmall, and F. Zenz. 1945.
The Limiting Capacity of Dumped Tower Packings. Trans. Am. Inst. Chem. Engrs. 41:693-710.
Los Angeles County Air Pollution Control District. 1960a.
Organization and Program. Report No. 1. Emissions of Oxides of Nitrogen From Stationary
Sources in Los Angeles County. Los Angeles, Calif.
Los Angeles County Air Pollution Control District. 1960b.
Oxides,of Nitrogen Emitted by Small Sources. Report No. 2. Emissions of Oxides of Nitrogen
From Stationary Sources in Los Angeles County. Los Angeles, Calif. (Sept).
Los Angeles County Air Pollution Control District. 1963.
Summary of Air Pollution Statistics. Los Angeles, Calif. (Jan).
Lowenstein, J. G. 1958.
Calculate Adequate Rupture Disc Size. Chem. Eng. 65:157-58 (Jan 13).
Lunche, R. G. , A. Stein, C.J. Seymour, and R. L. Weimer. 1957.
Distribution Survey of Products Emitting Organic Vapors in Los Angeles County. Chem. Eng. Prog
53:371-76 (Aug).
Lunde, K.E., and C.E. Lapple. 1957.
Dust and Mist Collection--A Critique on the State of the Art. Chem. Eng. Prog. 53:385-91
(Aug).
MacKnight, R. J. , J.E. Williamson, J. J. Sableski, Jr., andJ.O. Dealy. I960.
Controlling the Flue-Fed Incinerator. JAPCA. 10:103-09 (Apr).
-------
References - Mills 817
MacPhec, R.D., J.R. Taylor, and A. L. Chaney. 1957.
Some Data on Participates From Fuel Oil Burning. Proceedings of Semi-Annual Technical Con-
ference, Air Pollution Control Association, San Francisco, Calif., Nov. 18-19, 1957.
Magill, P. L. , F.R. Holden, and C. Ackley, eds. 1956.
Air Pollution Handbook. McGraw-Hill Book Co. , Inc. , New York, N. Y.
Mantell, C. L. 1961.
Adsorption. 2d ed. McGraw-Hill Book Co. , Inc. , New York, N. Y.
Manual of Exhaust Hood Designs.
See Kane, J. M.
Manufacturing Chemists' Association, General Safety Committee. 1954.
Guide for Safety in the Chemical Laboratory. Van Nostrand, New York, N. Y.
Marks, L. S. , ed. 1951.
Mechanical Engineers' Handbook. 5th ed. McGraw-Hill Book Co. , Inc. , New York, N. Y.
Marshall, W. R. , Jr., and S. J. Friedman. 1950.
Drying. In: Chemical Engineers' Handbook, 3d ed. Perry, J. H. , ed. McGraw-Hill Book Co. ,
Inc. , New York, N. Y.
Marzocchi, A., F. Lachut, and W. H. Willis, Jr. 1962.
Glass Fibers and Their Use as Filter Media. JAPCA. 12:38-42 (Jan).
Mason-Neilan Division. 1963.
Valve Slide Rule Instructions. Worthington Corp., Norwood, Mass.
Mattiello, J.J. 1943.
Protective and Decorative Coatings. Vol. III. John Wiley and Sons, New York, N. Y.
McAdams, W. H. 1942.
Heat Transmission. 2d ed. McGraw-Hill Book Co. , Inc. , New York, N. Y.
McCabe, L. C. 1952.
Secondary Processing of Aluminum. Ind. Eng. Chem. 44:121A-122A (May).
McCord, C. P. , and W. N. Witheridge. 1949.
Odors, Physiology and Control. McGraw-Hill Book Co. , Inc. , New York, N. Y.
McLouth, M.E., andH.J. Paulus. 1961.
Air Pollution From the Grain Industry. JAPCA. 11:313-17 (July).
Mellan, I. 1944.
Industrial Solvents. Reinhold Publishing Corp. , New York, N. Y.
Mellan, I. 1957.
Handbook of Solvents. Reinhold Publishing Corp. , New York, N. Y.
Merritt, F. H. 1958.
Side Port Furnaces. Glass Ind. 39:375-96 (July).
Miller, P.O., Jr., E. J. Hibshman, and J. R. Connell. 1956.
The Design of Smokeless, Nonluminous Flares. Presented at 21st American Petroleum Institute
Division of Refining Midyear Meeting, May 14-17, 1956.
Mills, J. L. , W.F. Hammond, andR.C. Adrian. I960.
Design of Afterburners for Varnish Cookers. JAPCA. 10:161-68 (Apr).
-------
818 References - Mills
Mills, J. L. , K.D. Luedtke, P. F. Woolrich, andL.B. Perry. 1961.
A Summary of Data on Air Pollution by Oxides of Nitrogen Vented From Stationary Sources. Final
Report. Report No. 4. Emissions of Oxides of Nitrogen From Stationary Sources in Los Angeles
County. Los Angeles, Calif. (July).
Mills, J.L., R.T. Walsh, K.D. Luedtke, andL.K. Smith. 1963.
Quantitative Odor Measurement. JAPCA. 13:467-75 (Oct).
Molcohy, B.D. 1950.
The Cupola--Its Raw Materials and Operation. Foundry. 78:75-76 (Mar).
Molos, J.E. 1961.
Control of Odors From a Continuous Soap Making Process. JAPCA. 11:9-13, 44 (Jan).
Monstross, C.F. 1953.
Entrainment Separation. Chem. Eng. 60:213-36 (Oct).
Mumford, A. R. , et al. 1940.
Characteristics of Cloth Filters on Coal Dust-Air Mixtures. Trans. Am. Soc. Mech. Engrs.
62:271-81.
National Bureau of Standards. 1949.
Safe Handling of Radioactive Isotopes. Handbook No. 42.
National Fire Protection Association. 1963.
Standard for Ovens and Furnaces (NFPA No. 86A). 60 Batterymarch Street, Boston, Mass.
Natural Gas Equipment, Inc. 1955.
Venturi-Type Burner Catalog. 1810 Fair Oaks Ave. , South Pasadena, Calif.
Natural Gas Processors Suppliers Association. 1957.
Engineering Data Book. 7th ed. Tulsa, Okla.
Neimeyer, E.R. 1961.
Check These Points When Designing Knockout Drums. Hydrocarbon Process. Petrol. Refiner.
40:155-56 (June).
Nelson, W. L. 1953.
How Painting Affects Storage Tank Losses. Oil Gas J. 52:130 (Nov 2).
Nelson, W. L. 1958.
Petroleum Refinery Engineering. 4th ed. McGraw-Hill Book Co. , Inc., New York, N. Y.
New York Blower Company. 1948.
Bulletin 482, La Porte, Ind.
Niven, W. W. , Jr. 1955.
Industrial Detergency. Remhold Publishing Corp. , New York, N. Y.
Nonferrous Foundrymen's Smog Committee. .
Furnace Practice Manual, Los Angeles, California.
The North American Manufacturing Co. 1952.
Combustion Handbook. Cleveland, Ohio.
Obrzut, J. J. 1958.
Will the Big Demand for Oxygen Come From Open Hearths9 Iron Age. 182(Pt. 2):172-74 (Sept 11).
-------
References - Randolph 819
O'Connell, H.E. 1946.
Plate Efficiency of Fractionating Columns and Absorbers. Trans. Am. Inst. Chem. Engrs.
42:741-55.
Oil and Gas Journal. 1957.
Special Processing Section: Processing in Today's Refineries. Oil Gas J. 55:121-68 (Mar 25).
Oxy-Catalyst, Inc. 1956.
Oxycat Technical Manual. Berwyn, Pa. (Jan).
Parmelee, C. W. 1951.
Ceramic Glazes. 2d ed. Industrial Publications, Inc. , Chicago 3, 111.
Peach, N. 1959.
Electrical Conversion. Power. 103:67-90 (Dec).
Peckham, G.W. 1962.
New Electric Furnace Reduces Melting Costs. Glass Ind. 43:552, 568, 573 (Oct).
Penney, G.W. 1937.
A New Electrostatic Precipitator. Elec. Eng. 56:159-63 (Jan).
Perry, J. H. , ed. 1950.
Chemical Engineers' Handbook. 3d ed. McGraw-Hill Book Co., Inc., New York, N. Y.
Peters, M. S. 1955a.
Principles and Processes for Removing Nitrogen Oxides From Gases. Engineering Experiment
Station, University of Illinois. Technical Report No. 14 (Aug).
Peters, M.S. 1955b.
Stop Pollution by Nitrogen Oxides. Chem. Eng. 62:197 (May).
Plastics Catalog Corporation. 1959.
Modern Plastics Encyclopedia Issue for 1960. 37( 1 A): 68- 1 80. Bristol, Conn.
Porter, E. D. 1959.
Electric Furnaces. Chem. Eng. 66:133-36 (Mar 9).
Powell, J.S. 1950.
Selection of an Odorant for Natural Gas Odorization. Proceedings of Pacific Coast Gas Associa-
tion. 41:134-38.
Prater, N. H. , and J. Mylo. 1961.
Equipment Cost Data File. Hydrocarbon Proces s. Petrol. Refiner. 40:209-10 (May); 173-74
(June); 163-64 (July); 131-32 (Aug).
Priestley, H. 1958.
Introductory Physics. Allyn and Bacon, Inc. , Boston, Mass.
Pring, R. T. 1952.
Bag-Type Cloth Dust and Fume Collectors. In: United States Technical Conference on Air Pollu-
tion, May 3-5, 1950, Washington, D. C. , L.C. McCabe, chairman. McGraw-Hill Book Co., Inc.,
New York, N. Y.
Puleo, P. A. I960.
Which to Use--Relief Valve or Rupture Disc? Petrol. Refiner. 39:157-62 (Oct).
Randolph, W. W. 1956.
Gas Distribution in Electrostatic Precipitators . Proceedings for Engineering Seminar on Electro-
static Precipitation. Pennsylvania State University, University Park, Pa.
-------
820 References - Ranz
Ranz, W.E. 1951.
The Impaction of Aerosol Particles on Cylindrical and Spherical Collectors. Engineering Ex-
periment Station, University of Illinois, Technical Report No. 3. (March 31).
Rees, R. L. 1955.
The Removal of Sulfur Dioxide From Power-Plant Stack Gases. In: International Congress on
Air Pollution. Problems and Control of Air Pollution. F. S. Mallette, ed. Reinhold Publishing
Corp. , New York, N. Y.
Rendle, L.K., andR.D. Wilsdon. 1956.
The Prevention of Acid Condensation in Oil-Fired Boilers. J. Inst. Fuel. 29:372-80 (Sept).
Reno, G. J. , E.D. Neumann, and L. C. Burroughs, 1958.
How Shell Treats Refinery Wastes. Petrol. Refiner. 37:153-60 (May).
Report No. 3, Experimental Program for the Control of Organic Emissions From Protective Coat-
ing Operations. See Spencer et al. , 1959.
Report No. 8, Experimental Program for the Control of Organic Emissions From Protective Coat-
ing Operations. See Elliott et al. , 1961.
Research-Cottrell, Inc.
Velocity Tables. Bound Brook, N. J. , (p. 1).
Robertson, L. F. , F. V. Halvonik, andJ.J. McMackin. 1957.
Regenerator Efficiency and Basic Brick in a Glass Furnace. Am. Cerarn. Soc. Bull. 36:381-84
(Nov).
Rodebush, W. H. 1950.
Filtration of Aerosols. Chapter 9. In: Handbook on Aerosols. U.S. Atomic Energy Commission,
Washington, D.C., 1950. Reprinted 1963 (Available From U.S. Department of Commerce, Office
of Technical Services).
Ronald, D. 1935.
Handbook of Offensive Trades. William Hodge and Co. , Ltd. , London, Eng.
Rose, A. H., andH.R. Crabaugh. 1955.
Incinerator Design Standards: Research Findings. Publication 60. Los Angeles County Air
Pollution Control District, Los Angeles, Calif.
Rose, A. H. , Jr., D. G. Stephan, andR.L. Stenburg. 1958.
Prevention and Control of Air Pollution by Process Changes or Equipment. Tech. Rept. A58-11.
Robert A. Taft Sanitary Engineering Center, Cincinnati, Ohio; See also, Air Pollution World
Health Organization Monograph Series No. 46, Columbia University Press, New York, N. Y. , 1961.
Rose, H.E., andA.J. Wood. 1956.
An Introduction to Electrostatic Precipitation in Theory and Practice. Essential Books, Fairlawn,
N. J.
Rudolfs, W. 1953.
Industrial Water and Wastes, Their Disposal and Treatment. Reinhold Publishing Corp., New
York, N. Y.
Rupp, W. H. 1956.
Air Pollution Sources and Their Control. Section 1. In: Air Pollution Handbook. McGraw-Hill
Book Co. , Inc. , New York, N. Y.
St. John, H.M. 1955.
Melting Practice in the Brass Foundry. Foundry. 83:107-08 (Nov).
-------
References - Singer 821
Samans, W. 1955.
Pressure Relief Devices. Petrol. Processing. 10:849-53 (June).
Sax, N.R. 1963.
Dangerous Properties of Industrial Materials. 2d ed. Reinhold Publishing Corp. New York, N. Y.
Scheer, J.F. 1956.
The Big Etch. Skyline (North American Aviation, Inc.). 14:23 (May).
Schmidt, A. 1928.
Progress in the Art of Electrical Precipitation. Trans. Am. Inst. Chem. Engrs. 21:11-33.
Schmidt, W. A. 1949.
Electrical Precipitation and Mechanical Dust Collection. Ind. Eng. Chem. 41:2428-32 (Nov).
Schmidt, W. A. , and C. R. Flodin. 1952.
Fundamental Principles, Design, Application, Performance, and Limitations of Electrical Pre-
cipitation Equipment. In: United States Technical Conference on Air Pollution, May 3-5, 1950,
Washington, D. C. L. C. McCabe, chairman. McGraw-Hill Book Co. , New York, N. Y.
Schmidt, W. A. , W. T. Sproull, and Y. Nakada. 1950.
How the Operation of a Cottrell Precipitator is Affected by the Resistivity of the Collected Mate-
rial. Presented at 170th General Meeting, American Institute of Mining and Metallurgical En-
gineers, New York, N. Y. , Feb. 15, 1950.
Selheimer, C.W. , and R. Lance. 1954.
Analysis of Fumes Leaving Resin Kettles and Fume Abatement Equipment. Offic. Dig. Federation
Paint and Varnish Production Clubs. 27:711-68 (Aug).
Semrau, K. T. I960.
Correlation of Dust Scrubber Efficiency. JAPCA. 10:200-07 (June).
Sensenbaugh, J.D., and J. Jonakin. 1960.
Effects of Combustion Conditions on Nitrogen Oxide Formation in Boiler Furnaces. Paper No. 60-
WA-334, American Society of Mechanical Engineers, New York, N. Y.
Shamos, M. H. , andS.G. Roth. 1950.
Industrial and Safety Problems of Nuclear Technology. Harper Bros. , New York, N. Y.
Shand and Jurs Co.
Bulletin 52-D. Berkeley, Calif.
Sharp, D. E. 1954.
Fuel Utilization in Glass Melting. Ceram. Ind. 63:45-51 (Dec).
Sharp, D. E. 1955.
Tank Design Aids in Melting and Fining. Ceram. Ind. 64:49-51, 88(Feb).
Sherwood, T.K., and R. L. Pigford. 1952.
Absorption and Extraction. McGraw-Hill Book Co. , Inc. , New York, N. Y.
Shreve, R.N. 1945.
The Chemical Process Industries. 1st ed. McGraw-Hill Book Co. , Inc. , New York, N. Y.
(2d ed. , 1956).
Silverman, L. 1950.
Filtration Through Porous Materials. Am. Ind. Hyg. Assoc. Quart. 11:11-20.
Singer, S.J. 1956.
Silicones Open New Era in Glass Making. Ceram. Ind. 57:82, 83, 117 (Nov).
-------
822 References - Slaik
Slaik, H. , and A. Turk. 1953.
Air Conservation Engineering. 2d ed. Connor Engineering Corp. , Danbury, Conn.
Smith, A. G. 1956a.
Air Oxidation of Sulfides in Process Waters and Caustic Solutions. Presented at 21st American
Petroleum Institute Division of Refining Midyear Meeting, May 14-17, 1956.
Smith, A. G. 1956b.
Ridding Process Waters and Caustic Solutions of Sulfides. Oil Gas J. 54:95-96, 98-99 (July 9).
Smith, E. C. 1958.
Air-Cooled Heat Exchangers. Chem. Eng. 65:145-50 (Nov 17).
Smolen, W. H. 1951.
Smokeless Flare Stacks. Petrol. Processing. 6:978-82 (Sept).
Smolen, W. H. 1952.
Design of Smokeless Flares. Presented at 17th American Petroleum Institute Division of Re-
fining Midyear Meeting, May 13, 1952.
Spain, R. W. 1955.
Glass Furnaces and How They Operate. Ceram. Ind. 65:71-74 (Aug).
Spain, R. W. 1956a.
How to Get Better Results From Long Campaigns. Ceram. Ind. 67:84-85, 87 (Nov).
Spain, R.W. 1956b.
How to Control Poor Operating Conditions. Ceram. Ind. 67:80-83 (Dec).
Spaite, P.W., I.E. Hagan, and W. F. Todd. 1963.
A Protective Finish for Glass Fiber Fabrics. Chem. Eng. Progr. 59:54-57 (Apr).
Spaite, P. W. , D. G. Stephan, andA.H. Rose, Jr. 1961.
High Temperature Fabric Filtration of Industrial Gases. JAPCA. 11:243-47 (May).
Spencer, E. F. , Jr., N. Kayne, M. F. Le Due, and J. H. Elliott. 1959.
Experimental Program for the Control of Organic Emissions From Protective Coating Opera-
tions. Report No. 3. Los Angeles County Air Pollution Control District, Los Angeles, Calif.
(July).
Sproull, W. T. 1951.
Precipitators Stop Dust and Fumes. Chem. Eng. 58:151-54 (May).
Sproull, W. T. , and Y. Nakada. 1951.
Operation of Cottrell Precipitators--Effects of Moisture and Temperature. Ind. Eng. Chem.
43:1350-58 (June).
Sproull, W. T. 1955.
Collecting High Resistivity Dusts and Fumes. Ind. Eng. Chem. 47:940-44 (Apr).
Stairmand, C. J. 1956.
The Design and Performance of Modern Gas-Cleaning Equipment. J. Inst. Fuel. 29:58-76 (Feb).
Steigerwald, B. J. 1958.
Emissions of Hydrocarbons to the Atmosphere From Seals on Pumps and Compressors. Report
No. 6. Joint District, Federal, and State Project for the Evaluation of Refinery Emissions.
Los Angeles County Air Pollution Control District, Los Angeles, Calif. (Apr).
-------
References - Underwood 823
Stenburg, R.L. 1958.
Control of Atmospheric Emissions From Paint and Varnish Manufacturing Operations. U. S. Depart-
ment of Health, Education, and Welfare, Robert A. Taft Sanitary Engineering Center, Cincinnati,
Ohio. Technical Report A58-4 (June). Also in: Paint and Varnish Production. 49:61-65, 111-14.
(1959).
Stephan, D.G. , and G. W. Walsh. I960.
Residual Dust Profiles in -- Air Filtration. Ind. Eng. Chem. 52:999-1002 (Dec).
Stephan, D. G. , G. W. Walsh, and R. A. Herrick. I960.
Concepts in Fabric Air Filtration. Am. Ind. Hyg. Assoc. J. 21:1-14 (Feb).
Stern, A. C. , K.J. Caplan, and P.O. Bush. 1956.
Removal of Particulate Matter From Gaseous Wastes: Cyclone Dust Collectors. Division of
Refining, American Petroleum Institute, New York, N. Y.
Stine, V. F. 1955.
Blast Cleaning in Industry, Bulletin No. 1500. Pangborn Corporation, Hagerstown, Md.
Streeter, V. L. 1951.
Fluid Mechanics. McGraw-Hill Book Co. , Inc., New York, N. Y.
Striplin, M. M. , Jr. 1948.
Development of Processes and Equipment for Production of Phosphoric Acid. Chemical Engineer-
ing Report No. 2. Tennessee Valley Authority.
Sussrnan, V. H. 1957.
Atmospheric Emissions From Catalytic Cracking Unit Regenerator Stacks. Report No. 4. Joint
District, Federal, and State Project for Evaluation of Refinery Emissions. Los Angeles County
Air Pollution Control District, Los Angeles, Calif. (June).
Sussman, V. H. , R.K. Palmer, F. Bonamassa, B. J. Steigerwald, and R. G. Lunche. 1958.
Emissions to the Atmosphere From Eight Miscellaneous Sources in Oil Refineries. Report No. 8.
Joint District, Federal, and State Project for the Evaluation of Refinery Emissions. Los Angeles
County Air Pollution Control District, Los Angeles, Calif. (June).
Suter, H. R. 1955.
Range of Applicability of Catalytic Fume Burners. JAPCA. 5(3):173-75, 184 (Nov).
Sutton, O.G. 1950.
The Dispersion of Hot Gases in the Atmosphere. J. Meteorol. 7:307-12 (Oct).
Teller, A. J. I960.
Absorption With Chemical Reaction. Chem. Eng. 67:111-24 (July 11).
Thomas, J. W. 1959.
Air vs. Water Cooling, Cost Comparison. Chem. Eng. Progr. 55:38-41 (Apr).
Tooley, F. V. 1953.
Handbook of Glass Manufacture. Volumes I and II. Ogden Publishing Co. , New York, N. Y.
Treybal, R. E. 1955.
Mass-Transfer Operations. McGraw-Hill Book Co. , Inc. , New York, N. Y.
Trinks, W. 1955.
Industrial Furnaces. Volume I. 3d ed. John Wiley and Sons, New York, N. Y.
Turk, A., andK.A. Bownes. 1951.
Adsorption Can Control Odors. Chem. Eng. 58:156-58 (May).
Underwood, G. 1962.
Removal of Sub-Micron Particles From Industrial Gases, Particularly in the Steel and Electricity
Industries. Intern. J. Air Water Pollution. 6:229-63 (May-Aug).
-------
824 References - United
U.S. National Bureau of Standards.
Fuel Oils. Commercial Standard CS-48. Clearinghouse for Federal Scientific and Technical
Information, Springfield, Va. 22151.
U.S. National Bureau of Standards. 1949.
Handbook No. 42. Safe Handling of Radioactive Isotopes. Washington, D. C.
Van Dreser, M.L. 1962.
Basic Refractories for the Glass Industry. Glass Ind. 43:18-21 (Jan).
Von Fisher, W. 1948.
Paint and Varnish Technology. Reinhold Publishing Corp. , New York, N. Y.
Waitkus, J. 1962.
Recover Waste Heat to Reduce Glass Tank Operating Cost. Ceram. Ind. 79:38-42, 68-70 (Dec).
Walker, E.A. , andJ.E. Coolidge. 1953.
Semiempirical Equation of Electrostatic Precipitation. Ind. Eng. Chem. 45:2417-22 (Nov).
Walker, W. H. , W.K. Lewis, W. H. McAdams, and E. R. Gilliland. 1937.
Principles of Chemical Engineering. 3d ed. McGraw-Hill Publishing Co., Inc., New York, N. Y.
Walsh, G. W. , and P. W. Spaite. 1962.
An Analysis of Mechanical Shaking in Air Filtration, JAPCA. 12:57-61 (Feb).
Ward, D.R. 1952.
Design of Laboratories for Safe Use of Radioisotopes. AECU-2226. U.S. Atomic Energy Com-
mission Advisory Field Service Branch, Isotopes Division, Oak Ridge, Tenn. (Nov).
Watts, D.L. , andJ.F. Higgins. 1962.
The New Baghouse Installation for Cleaning Smelter Gases at Phelps Dodge Refining Corporation.
JAPCA. 12:217-20 (May).
Weisburd.
See, Griswold, 1962.
Western Precipitation Corporation, 1952.
Cottrell Electrical Precipitators. 3d ed. Los Angeles, Calif.
White, H. J. 1951.
Particle Charging in Electrostatic Precipitation. Trans, Am. Inst. Elec. Engrs. 70(11): 1186-9 1.
White, H. J. 1953.
Electrostatic Precipitators for Electric Generating Stations. Trans. Am. Inst. Elec. Engrs.
72(III):229-41.
White, H. J. 1957.
Fifty Years of Electrostatic Precipitation. JAPCA. 7:166-77 (Nov).
White, H. J. 1963.
Industrial Electrostatic Precipitation. Addison-Wesley Publication Co., Reading, Mass.
White, H. J. , and W. H. Cole. I960.
Design and Performance Characteristics of High-Velocity, High-Efficiency Air Cleaning Pre-
cipitators. JAPCA. 10:239-45 (June).
White, H.J., and G. W. Penney. 1961.
Basic Concepts. In: Electrical Precipitation Fundamentals. Proceedings for Engineering Seminar
on Electrostatic Precipitation, June 17-21, 1957. Pennsylvania State University, Department of
Electrical Engineering and General Extension, University Park, Pa.
-------
References - Zink 825
Williams, C.E., et al. 1940.
Determination of Cloth Area for Industrial Air Filters. Heating, Piping, Air Conditioning.
12:259-63 (Apr).
Williams Patent Crusher and Pulverizer Co. , Inc.
Bulletin 696. St. Louis 6, Mo.
Willington Sears Co. 1954.
Filter Fabric Facts. New York, N. Y.
Wilson, E. F. I960.
Dust Control in Glass Manufacturing. Glass Ind. 41:202-03, 236, 237 (Apr).
Woodhouse, H. 1957.
Centrifugal Pump Packings and Seals. Pt. 3. Mechanical Seals. Petrol. Refiner. 36:207-11
(Apr).
Woodward, E. R. , andE.R. Fenrich. 1952.
Odor Control With Chlorine Dioxide. Chem. Eng. 59:174-75 (Apr).
Zachariasen, W. H. 1932.
The Atomic Arrangement of Glass. J. Am. Chem. Soc. 54:3841-51.
John Zink Company.
Flare Bulletin. Tulsa, Okla.
-------
APPENDIX A: RULES AND REGULATIONS
APPENDIX B: ODOR-TESTING TECHNIQUES
KARL D. LUEDTKE, Intermediate Air Pollution Engineer
APPENDIX C: HYPOTHETICAL AVAILABLE HEATS FROM NATURAL GAS
SANFORD M. WEISS, Senior Air Pollution Engineer
APPENDIX 0: MISCELLANEOUS DATA
-------
APPENDIX A
Except for format and the Contents, Appendix A was set in type exactly
as it appears in Los Angeles APCD RULES AND REGULATIONS manual.
-------
APPENDIX A: RULES AND REGULATIONS
RULES AND REGULATIONS OF THE
AIR POLLUTION CONTROL DISTRICT
REGULATION I. GENERAL PROVISIONS
RULE 1. TITLE
These rules and regulations shall be known as the
rules of the Air Pollution Control District.
RULE 2. (Amended 1-16-58) DEFINITIONS
a. Except as otherwise specifically provided in
these rules and except where the context other-
wise indicates, words used in these rules are
used in exactly the same sense as the same
words are used in Chapter 2, Division 20 of
the Health and Safety Code.
b. (Amended 1-16-58) Person. "Person" means
any person, firm, association, organiza-
tion, partnership, business trust, corpora-
tion, company, contractor, supplier, install-
er, user or owner, or any state or local gov-
ernmental agency or public district or any
officer or employee thereof.
c. Board. "Board" means the Air Pollution Con-
trol Board of the Air Pollution Control District
of Los Angeles County.
e. Section. "Section" means section of the Health
and Safety Code of the State of California un-
less some other statute is specifically men-
tioned.
f. Rule. "Rule" means a rule of the Air Pollu-
tion Control District of Los Angeles County.
g. (Amended 3-14-63) Los Angeles Basin. "Los
Angeles Basin" is defined as being within the
following described boundaries:
Beginning at the intersection of the southerly
boundary of the Angeles National Forest with 1.
the easterly boundary of the County of Los
Angeles; thence along said easterly boundary
in a general southwesterly direction to the
mean high tide line of the Pacific Ocean; thence
continuing along the boundary of the County of
Los Angeles (in the Pacific Ocean) in a gen-
eral southwesterly, westerly and northwester-
ly direction to its most westerly intersection m.
with the boundary of the City of Los Angeles
(in the Pacific Ocean); thence in a general
831
k.
northerly direction along the generally west-
erly boundary of the City of Los Angeles to its
most northerly intersection with the westerly
boundary of the County of Los Angeles; thence
in a general easterly direction along the north-
erly boundary of said City of Los Angeles to
the southwesterly corner of Section 16, Town-
ship 2 North, Range 13 West, S. B. B. & M. ;
thence in a general easterly direction along
said southerlyboundary of the Angeles Nation-
al Forest to said easterly boundary of the
County of Los Angeles.
Regulation. "Regulation" means one of the
major subdivisions of the Rules of the Air Pol-
lution Control District of Los Angeles County.
(Amended 1-16-58) Particulate Matter. "Par-
ticulate Matter" is any material, except un-
combined water, which exists in a finely di-
vided form as a liquid or solid at standard con-
ditions .
Process Weight Per Hour. "Process Weight"
is the total weight of all materials introduced
into any specific process which process may
cause any discharge into the atmosphere. Sol-
id fuels charged will be considered as part of
the process weight, but liquid and gaseous
fuels and combustion air will not. "The Pro-
cess Weight Per Hour" will be derived by di-
viding the total process weight by the number
of hours in one complete operation from the
beginning of any given process to the comple-
tion thereof, excluding any time during which
the equipment is idle.
Dusts . "Dusts" are minute solid particles re-
leased into the air by natural forces or by me-
chanical processes such as crushing, grind-
ing, milling, drilling, demolishing, shovel-
ing, conveying, covering, bagging, sweeping,
etc.
Condensed Fumes. "Condensed Fumes" are
minute solid particles generated by the con-
densation of vapors from solid matter after
volatilization from the molten state, or may
be generated by sublimation, distillation, cal-
cination, or chemical reaction, when these
processes create air-borne particles.
Combustion Contaminants. "Combustion Con-
taminants " are particulate matter discharged
into the atmosphere from the burning of any
-------
832
RULES AND REGULATIONS
kind of material containing carbon in a free
or combined state.
n. Atmosphere. "Atmosphere" means the air
that envelops or surrounds the earth. Where,
air pollutants are emitted into a building not
designed specifically as a piece of air pollu-
tion control equipment, such emission into the
building shall be considered an emission into
the atmosphere.
o. Combustible Refuse (Amended 3-2-67) ''Com-
bustible Refuse" is any solid or liquid com-
bustible waste material containing carbon in
a free or combined state.
p. Multiple —Chamber Incinerator. "Multiple -
chamber Incinerator" is any article, machine,
equipment, contrivance, structure or part of
a structure, used to dispose of combustible
refuse by burning, consisting of three or more
refractory lined combustion furnaces in se-
ries, physically separated by refractory walls ,
interconnected by gas passage ports or ducts
and employing adequate design parameters
necessary for maximum combustion of the ma-
terial to be burned. The refractories shall
have a Pyrometric Cone Equivalent of at least
17, tested according to the method described
inthe American Society for Testing Materials,
Method C-24.
q. Oil-Effluent Wa_te_r_ Separator. "Oil-effluent
water separator" is any tank, box, sump or
other container in which any petroleum or
product thereof, floating on or entrained or
contained in water entering such tank, box,
sump or other container, is physically sep-
arated and removed from such water prior to
outfall, drainage, or recovery of such water.
RULE 3. STANDARD CONDITIONS
Standard conditions are a gas temperature of 60
degrees Fahrenheit and a gas pressure of 14.7
pounds per square inch absolute. Results of all
analyses and tests shall be calculated or reported
at this gas temperature and pressure.
tain authorization for such construction from
the Air Pollution Control Officer. An author-
ity to construct shall remain in effect until the
permit to operate the equipment for which the
application was filed is granted or denied or
the application is canceled.
b. Permit to Operate. (Amended 11-16-54) Be-
fore any article, machine, equipment or other
contrivance described in Rule 10(a) maybe
operated or used, a written permit shall be
obtained from the Air Pollution Control Offi-
cer. No permit to operette or use shall be
granted either by the Air Pollution Control
Officer or the Hearing Board for any article,
machine, equipment or contrivance described
in Rule 10(a), constructed or installed without
authorization as required by Rule 10(a), until
the information required is presented to the
Air Pollution Control Officer and such article,
machine, equipment or contrivance is altered,
if necessary, and made to conform to the stan-
dards set forth in Rule 20 and elsewhere in
these Rules and Regulations.
c. Posting of Permit to Operate. (Amended 3-
2-67) A person -who has been granted under
Rule 10 a permit to operate any article, ma-
chine, equipment, or other contrivance de-
scribed in Rule 10(b), shall firmly affix such
permit to operate, an approved facsimile, or
other approved identification bearing the per-
mit number upon the article, machine, equip-
ment, or other contrivance in such a manner
as tobe clearly visible and accessible. In the
event that the article, machine, equipment,
or other contrivance -is so constructed or
operated that the permit to operate cannot be
so placed, the permit to operate shall be
mounted so as to be clearly visible in an ac-
cessible place within 25 feet of the article,
machine, equipment or other contrivance, or
maintained readily available at all times on
the operating premises.
d. (Adopted 3-28-57) A person shall not wilfully
deface, alter, forge, counterfeit, or falsify a
permit to operate any article, machine, equip-
ment, or other contrivance.
REGULATION II. PERMITS
RULE 10. PERMITS REQUIRED
a. Authority to Construct. (Amended 4-2-64) Any
person building, erecting, altering or replac-
ing any article, machine, equipment or other
contrivance, the use of which may cause the
issuance of air contaminants or the use of
which may eliminate or reduce or control the
issuance of air contaminants, shall first ob-
f. Permit to Sell or Rent. Adopted 1-16-58)
Any person who sells or rents to another per-
son an incinerator which may be used to dis-
pose of combustible refuse by burning within
the Los Angeles Basin and which incinerator
is to be used exclusively in connection with
any structure, which structure is designed for
and used exclusively as a dwelling for not more
than four families, shall first obtain a permit
from the Air Pollution Control Officer to sell
or rent such incinerator.
-------
Rules and Regulations of the Air Pollution Control District
833
RULE 11. EXEMPTIONS
An authority to construct or a permit to operate
shall not be required for:
a. (Amended 3-2-67) Vehicles as defined by the
Vehicle Code of the State of California but not
including any article, machine, equipment or
other contrivance mounted on such vehicle that
would otherwise require a permit under the
provisions of these Rules and Regulations.
b. Vehicles used to transport passengers or
freight.
c. Equipment utilized exclusively in connection
•withany structure, which structure is designed
for and used exclusively as a dwelling for not
more than four families.
d. The following equipment:
1. Comfort air conditioning or comfort ven-
tilating systems which are not designed
to remove air contaminants generated by
or released from specific units or equip-
ment.
2. Refrigeration units except those used as,
or in conjunction with, air pollution con-
trol equipment.
3. (Amended 3-2-67) Piston type internal
combustion engines.
Water cooling towers and water cooling
ponds not used for evaporative cooling of
process water or not used for evaporative
cooling of water from barometric jets or
from barometric condensers.
6. Equipment used exclusively for steam
cleaning.
7. Presses used exclusively for extruding
metals, minerals, plastics or wood.
8. Porcelain enameling furnaces, porcelain
enameling drying ovens , vitreous enamel-
ing furnaces or vitreous enameling drying
ovens.
9. Presses used for the curing of rubber
products and plastic products.
10. Equipment used exclusively for space
heating, other than boilers.
13. Equipment used for hydraulic or hydro-
static testing.
14. (Amended 7-28-66) All sheet-fed print-
ing presses and all other printing presses
using exclusively inks containing less than
10% organic solvents, diluents or thin-
ners .
17. Tanks, vessels and pumping equipment
used exclusively for the storage or dis-
pensing of fresh commercial or purer
grades of:
a. Sulfuric acid with an acid strength of
99 per cent or less by weight.
b. Phosphoric acid with an acid strength
of 99 per cent or less by weight.
c. Nitric acid with an acid strength of 70
per cent or less by weight.
18. Ovens used exclusively for the curing of
plastics which are concurrently being
vacuum held to a mold or for the soften-
ing or annealing of plastics.
19. (Amended 6-1-65) Equipment used exclu-
sively for the dyeing or stripping (bleach-
ing) of textiles where no organic solvents,
diluents or thinners are used.
20. (Amended 7-28-66) Equipment used ex-
clusively to mill or grind coatings and
molding compounds where all materials
charged are in a paste form.
21. Crucible type or pot type furnaces with a
brimful capacity of less than 450 cubic
inches of any molten metal.
22. (Amended 6-1-65) Equipment used exclu-
sively for the melting or applying of wax
where no organic solvents, diluents or
thinners are used.
23. Equipment used exclusively for bonding
lining to brake shoes.
24. Lint traps used exclusively in conjunction
with dry cleaning tumblers.
25. Equipment used in eating establishments
for the purpose of preparing food for hu-
man consumption.
-------
834
RULES A.ND REGULATIONS
26, Equipment used exclusively to compress
or hold dry natural gas.
27. Tumblers used for the cleaning or de-
burring of metal products without abra-
sive blasting.
28. Shell core and shell-mold manufacturing
machines.
29. Molds used for the casting of metals.
30. (Amended 3-2-67) Abrasive blast cab-
inet-dust filter integral combination
units where the total internal volume
of the blast section is 50 cubic feet or less .
31. Batch mixers of 5 cubic feet rated work-
ing capacity or less.
32. Equipment used exclusively for the pack-
aging of lubricants or greases.
33. (Amended 3-2-67) Equipment used exclu-
sively for the manufacture of water emul-
sions of asphalt, greases, oils or waxes.
34. Ovens used exclusively for the curing of
vinyl plastisols by the closed mold curing
process.
35. Equipment used exclusively for conveying
and storing plastic pellets.
36. Equipment used exclusively for the mix-
ing and blending of materials at ambient
temperature to make water based adhe-
sive s .
37. Smokehouses in which the maximum hori-
zontal inside cross-sectional area does
not exceed 20 square feet.
38. Platen presses used for laminating.
e. The following equipment or any exhaust sys-
tem or collector serving exclusively such
equipment:
1. Blast cleaning equipment using a suspen-
sion of abrasive in water.
2. Ovens, mixers and blenders used in bak-
eries where the products are edible and
intended for human consumption.
3, Kilns used for firing ceramic \vare, heat-
ed exclusively by natural gas, liquefied
petroleum gas, electricity or any com-
bination thereof.
4. Laboratory equipment used exclusively
for chemical or physical analyses and
bench scale laboratory equipment.
5. Equipment used for inspection of metal
products.
6. Confection cookers where the products
are edible and intended for human con-
sumption.
7. Equipment used exclusively for forging,
pressing, rolling or drawing of metals or
for heating metals immediately prior to
forging, pressing, rolling or drawing.
8. Die casting machines,
9. Atmosphere generators used in connec-
tion with metal heat treating processes.
10. Photographic process equipment by which
an image is reproduced upon material
sensitized to radiant energy.
11. Brazing, soldering or welding equipment.
12. Equipment used exclusively for the sin-
tering of glass or metals.
1 3. (Amended 3 -2 -67) Equipment used for buff-
ing (except automatic or semi-automatic
tire buffers) or polishing, carving, cut-
ting, drilling, machining, routing, sand-
ing, sawing, surface grinding or turning
of ceramic artwork, ceramic precision
parts, leather, metals, plastics, fiber-
board, masonry, asbestos, carbon or gra-
phite .
14. (Amended 3-2-67) Equipment used for
carving, cutting, drilling, surface grind-
ing, planing, routing, sanding, shredding
or turning of wood or the pressing or stor-
ing of sawdust, "wood chips or wood shav-
ings.
15. (Amended 3-2-67) Equipment using aque-
ous solutions for surface preparation,
cleaning, stripping, etching (does not in-
clude chemical milling) or the electrolyt-
ic plating with, electrolytic polishing of,
or the electrolytic stripping of brass,
bronze, cadmium, copper, iron, lead,
nickel, tin, zinc, and precious metals.
16. Equipment used for washing or drying
products fabricated from metal or glass,
provided that no volatile organic mate-
rials are used in the process and that no
oil or solid fuel is burned.
17. Laundry dryers, extractors or tumblers
used for fabrics cleaned only with water
solutions of bleach or detergents.
GPO 806—614—28
-------
Rules and Regulations of the Air Pollution Control District
835
19. Foundry sand mold forming equipment to
which no heat is applied.
20. Ovens used exclusively for curing potting
materials or castings made with epoxy
resins.
21. Equipment used to liquefy or separate
oxygen, nitrogen or the rare gases from
the air.
22. Equipment used for compression molding
and injection molding of plastics.
23. (Amended 6-1-65) Mixers for rubber or
plastics where no material in powder form
is added and no organic solvents, diluents
or thinners are used.
24. Equipment used exclusively to package
Pharmaceuticals and cosmetics or to coat
pharmaceutical tablets.
25. (Amended 6-1-65) Equipment used exclu-
sively to grind, blend or package tea, co-
coa, spices or roasted coffee.
26. (Amended 7-28-66) Roll mills or calen-
ders for rubber or plastics where no or-
ganic solvents, diluents or thinners are
used.
27. (Adopted 3-2-67) Vacuum producing de-
vices used in laboratory operations or in
connection with other equipment which is
exempt by Rule 11.
f. Steam generators, steam superheaters, water
boilers, water heaters, and closed heat trans-
fer systems that are fired exclusively with one
of the following:
1. Natural gas.
2. Liquefied petroleum gas.
3. A combination of natural gas and liquefied
petroleum gas.
g. Natural draft hoods, natural draft stacks or
natural draft ventilators.
h. Containers, reservoirs, or tanks used exclu-
sively for:
1. (Amended 6-1-65) Dipping operations for
coating objects \vith oils, waxes or greas-
es where no organic solvents, diluents or
thinners are used.
2. Dipping operations for applying coatings
of natural or synthetic resins which con-
tain no organic solvents.
3. Storage of liquefied gases.
5. Unheated storage of organic materials
with an initial boiling point of 300 °F. or
greater.
6. The storage of fuel oils with a gravity of
25° API or lower.
7. The storage of lubricating oils.
8. The storage of fuel oils with a gravity of
40° API or lower and having a capacity of
10,000 gallons or less.
9. (Amended 3-2-67) The storage of organic
liquids, except gasoline, normally used as
solvents, diluents or thinners, inks, col-
orants, paints, lacquers, enamels, var-
ishes, liquid resins or other surface coat-
ings, and having a capacity of 6, 000 gal-
lons or less.
10. (Amended 3-2-67) The storage of liquid
soaps, liquid detergents, vegetable oils,
waxes or wax emulsions.
11. The storage of asphalt.
12. (Amended 6-1-65) Unheated solvent dis-
pensing containers, unheated non-convey-
orized solvent rinsing containers or un-
heated non-conveyorized coating dip tanks
of 100 gallons capacity or less.
14. (Adopted 6-1-65) The storage of gasoline
having a capacity of less than 250 gallons.
15. (Adopted 3-2-67) Transporting materials
on streets or highways.
Equipment used exclusively for heat treating
glass or metals, or used exclusively for case
hardening, carburizing, cyaniding, nitriding,
carbonitriding, siliconizing or diffusion treat-
ing of metal objects.
Crucible furnaces, pot furnaces or induction
furnaces, with a capacity of 1000 pounds or
less each, in which no sweating or distilling
is conducted and from which only the follow-
ing metals are poured or in which only the
following metals are held in a molten state:
1. Aluminum or any alloy containing over
50 per cent aluminum.
2. Magnesium or any alloy containing over
50 per cent magnesium.
3. Lead or any alloy containing over 50 per
cent lead.
-------
836
RULES AND REGULATIONS
4. Tin or any alloy containing over 50 per
cent tin.
5. Zinc or any alloy containing over 50 per
cent zinc.
6. Copper.
7. Precious metals.
k. Vacuum cleaning systems used exclusively for
industrial, commercial or residential house-
keeping purposes.
1. Structural changes which cannot change the
quality, nature or quantity of air contaminant
emissions.
m. Repairs or maintenance not involving struc-
tural changes to any equipment for which a
permit has been granted.
n. Identical replacements in whole or in part of
any article, machine, equipment or other con-
trivance where a permit to operate had pre-
viously been granted for such equipment under
Rule 10.
RULE 12. (Amended 1-16-58) TRANSFER
An authority to construct, permit to operate or
permit to sell or rent shall not be transferable,
whether by operation of law or otherwise, either
from one location to another, from one piece of
equipment to another, or from one person to an-
other.
RULE 13. BLANKET PERMITS. (Deleted 4-2-64)
RULE 14. (Amended 1-16-58) APPLICATIONS
Every application for an authority to construct,
permit to operate or permit to sell or rent required
under Rule 10 shall be filed in the manner and form
prescribed by the Air Pollution Control Officer,
and shall give all the information necessary to en-
able the Air Pollution Control Officer to make the
determination required by Rule ZO hereof.
RULE 17. (Amended 6-1-65) CANCELLATION OF
APPLICATIONS
a.
b.
(Amended 6-1-65) An authority to construct
shall expire and the application shall be can-
celed two years from the date of issuance of
the authority to construct.
(Amended 6-1-65) An application for permit
to operate existing equipment shall be canceled
two years from the date of filing of the appli-
cation.
RULE 18. (Amended 1-16-58) ACTION ON AP-
PLICATIONS
The Air Pollution Control Officer shall act, with-
in a reasonable time, on an application for author-
ity to construct, permit to operate or permit to
sell or rent, and shall notify the applicant in writ-
ing of his approval, conditional approval or denial.
RULE 19. (Adopted 3-28-57) PROVISION OF SAM-
PLING AND TESTING FACILITIES
A person operating or using any article, machine,
equipment or other contrivance for which these
rules require a permit shall provide and maintain
such sampling and testing facilities as specified in
the authority to construct or permit to operate.
RULE 20. (Amended 3-14-63) STANDARDS FOR
GRANTING APPLICATIONS
a. The Air Pollution Control Officer shall deny
an authority to construct, permit to operate
or permit to sell or rent, except as provided
in Rule 21, if the applicant does not show that
every article, machine, equipment or other
contrivance, the use of which may cause the
issuance of air contaminants, or the use of
which may eliminate or reduce or control the
issuance of air contaminants, is so designed,
controlled, or equipped with such air pollu-
tion control equipment, that it may be expec-
ted to operate without emitting or without caus -
ing to be emitted air contaminants in violation
of Sections 24242 or 24243, Health and Safety
Code, or of these Rules and Regulations.
b. (Adopted 3-28-57) Before an authority to con-
struct or a permit to operate is granted, the
Air Pollution Control Officer may require the
applicant to provide and maintain such facili-
ties as are necessary for sampling and test-
ing purposes in order to secure information
that will disclose the nature, extent, quantity
or degree of air contaminants discharged into
the atmosphere from the article, machine,
equipment or other contrivance described in
the authority to construct or permit to oper-
ate. In the event of such a requirement, the
Air Pollution Control Officer shall notify the
applicant in writing of the required size, num-
ber and location of sampling holes; the size
and location of the sampling platform; the ac-
cess to the sampling platform; and the utili-
ties for operating the sampling and testing
equipment. The platform and access shall be
constructed in accordance -with the General
Industry Safety Orders of the State of Cali-
fornia.
-------
Rules and Regulations of the Air Pollution Control District
837
c. (Adopted 6-25-59) In acting upon a Permit to
Operate, if the Air Pollution Control Officer
finds that the article, machine, equipment or
other contrivance has been constructed not in
accordance with the Authority to Construct, he
shall deny the Permit to Operate. The Air
Pollution Control Officer shall not accept any
further application for Permit to Operate the
article, machine, equipment or other contriv-
ance so constructed until he finds that the arti-
cle, machine, equipment or other contrivance
has been reconstructed in accordance with the
Authority to Construct.
RULE 21. (Amended 12-4-58) CONDITIONAL AP-
PROVAL
a. The Air Pollution Control Officer may issue
an authority to construct or a permit to oper-
ate, subject to conditions which will bring the
operation of any article, machine, equipment
or other contrivance within the standards of
Rule 20, in which case the conditions shall be
specified in writing. Commencing wop^k under
such an authority to construct or operation
under such a permit to operate shall be deemed
acceptance of all the conditions so specified.
The Air Pollution Control Officer shall issue
an authority to construct or a permit to oper-
ate with revised conditions upon receipt of a
new application, if the applicant demonstrates
that the article, machine, equipment or other
contrivance can operate within the standards
of Rule 20 under the revised conditions.
b. The Air Pollution Control Officer may issue
a permit to sell or rent, subject to conditions
which will bring the operation of any article,
machine, equipment or other contrivance with-
in the standards of Rule 20, in which case the
conditions shall be specified in writing. Sell-
ing or renting under such a permit to sell or
rent shall be deemed acceptance of all the con-
ditions so specified. The Air Pollution Con-
trol Officer shall issue a permit to sell or
rent with revised conditions upon receipt of a
new application, if the applicant demonstrates
that the article, machine, equipment or other
contrivance can operate within the standards
of Rule 20 under the revised conditions.
RULE 22. (Amended 1-16-58) DENIAL OF AP-
PLICATIONS
In the event of denial of an authority to construct,
permit to operate or permit to sell or rent, the
Air Pollution Control Officer shall notify the ap-
plicant in writing of the reasons therefor. Ser-
vice of this notification may be made in person or
by mail, and such service may be proved by the
written acknowledgment of the persons served or
affidavit of the person making the service. The
Air Pollution Control Officer shall not accept a
further application unless the applicant has com-
plied with the objections specified by the Air Pol-
lution Control Officer as his reasons for denial of
the authority to construct, the permit to operate
or the permit to sell or rent.
RULE 23. (Amended 1-16-58) FURTHER INFOR-
MATION
Before acting on an application for authority to con-
struct, permit to operate or permit to sell or rent,
the Air Pollution Control Officer may require the
applicant to furnish further information or further
plans or specifications.
RULE 24. (Amended 1-16-58) APPLICATIONS
DEEMED DENIED
The applicant may at his option deem the authority
to construct, permit to operate or permit to sell
or rent denied if the Air Pollution Control Officer
fails to act on the application within 30 days after
filing, or within 30 days after applicant furnishes
the further information, plans and specifications
requested by the Air Pollution Control Officer,
whichever is later.
RULE 25. (Amended 1-16-58) APPEALS
Within 10 days after notice, by the Air Pollution
Control Officer, of denial or conditional approval
of an authority to construct, permit to operate or
permit to sell or rent, the applicant may petition
the Hearing Board, in writing, for a public hear-
ing. The Hearing Board, after notice and a public
hearing held within 30 days after filing the petition,
may sustain or reverse the action of the Air Pol-
lution Control Officer; such order may be made
subject to specified conditions.
REGULATION III. FEES
RULE 40. PERMIT FEES
Every applicant, except any state or local govern-
mental agency or public district, for an authority
to construct or a permit to operate any article,
machine, equipment or other contrivance, for which
an authority to construct or permit to operate is
requiredby the State law or the Rules and Regula-
tions of the Air Pollution Control District, shall
pay a filing fee of $40. 00. Where an application
is filed for a permit to operate any article, ma-
chine, equipment or other contrivance by reason
of transfer from one person to another, and where
a permit to operate had previously been granted
under Rule 10 and no alteration, addition or trans-
-------
838
RULES AND REGULATIONS
fer of location has been made, the applicant shall
pay only a $10.00 filing fee.
(Amended 6-L-65) Every applicant, except any
state or local governmental agency or public dis-
trict, for a permit to operate,, who files an appli-
cation with the Air Pollution Control Officer, shall,
in addition to the filing fee prescribed herein, pay
the fee for the issuance of a permit to operate in
the amount prescribed in the following schedules,
provided, however, that the filing fee shall be ap-
plied to the fee prescribed for the issuance of the
permit to operate.
(Amended 6-1-65) If an application for an authority
to construct or a permit to operate is canceled, or
if an authority to construct or a permit to operate
is denied and such denial becomes final, the filing
fee required herein shall not be refunded nor ap-
plied to any subsequent application,
(Amended 6-1-65) Where an application is filed for
a permit to operate any article, machine, equip-
ment or other contrivance by reason of transfer
of location or transfer from one person to another,
or both, and where a permit to operate had pre-
viously been granted for such equipment under Rule
10 and an alteration or addition has been made,
the applicant shall be assessed a fee based upon
the increase in total horsepower rating, the in-
crease in total fuel consumption expressed in thou-
sands of British Thermal Units (BTU) per hour,
the increase in total electrical energy rating, the
increase in maximum horizontal inside cross sec-
tional area or the increase in total stationary con-
tainer capacity resulting from such alterations or
additions, as described in the fee schedules con-
tained herein. Where the application is for trans-
fer of location and no alteration or addition has
been made, the applicant shall pay only a filing fee
of $40.
(Amended 6- 1 -65) Where an application is filed for
an authority to construct or a permit to operate
exclusively involving revisions to the conditions
of an existing permit to operate or involving alter-
ations or additions resulting in a change to any ex-
isting article, machine, equipment or other con-
trivance holding a permit under the provisions of
Rule 10 of these Rules and Regulations, the appli-
cant shall be assessed a fee based upon the in-
crease intotal horsepower rating, the increase in
total fuel consumption expressed in thousands of
British Thermal Units (BTU) per hour, the in-
crease in total electrical energy rating, the in-
crease in maximum horizontal inside cross sec-
tional area or the increase in total stationary con-
tainer capacity resulting from such alterations
or additions, as described in the fee schedules
contained herein. Where there is no change or is
a decrease in such ratings, the applicant shall pay
only the amount of the filing fee required herein.
After the provisions for granting permits as set
forth in Chapter 2, Division 20, of the Health and
Safety Code and the Rules and Regulations have
been complied with, the applicant shall be notified
by the Air Pollution Control Officer, in writing, of
the fee to be paid for issuance of the permit to
operate. Such notice may be given by personal
service or by deposit, postpaid, in the United
States mail and shall serve as a temporaryperm.it
to operate for 30 days from the date of personal
service or mailing. Nonpayment of the fee within
this period of time shall result in the automatic
cancellation of the application.
In the event that more than one fee schedule is ap-
plicable to a permit to operate, the governing
schedule shall be that which results in the higher
fee.
(Adopted 6-1-65) Where a single permit to operate
has been granted under Rule 10 prior to July 1,
1957, and where the Air Pollution Control Officer
would, since that date, have issued separate or
revised permits for each permit unit included in
the original application, the Air Pollution Control
Officer may issue such separate or revised per-
mits without fees.
In the event that a permit to operate is granted by
the Hearing Board after denial by the Air Pollu-
tion Control Officer or after the applicant deems
his application denied, the applicant shall pay the
fee prescribed in the following schedules within 30
days after the date of the decision of the Hearing
Board. Nonpayment of the fee within this period
of time shall result in automatic cancellation of
the permit and the application. Such a fee shall
not be charged for a permit to operate granted by
the Hearing Board for the duration of a variance.
(Amended 7-1-64) A request for a duplicate per-
mit to operate shall be made in writing to the Air
Pollution Control Officer within 10 days after the
destruction, loss or defacement of a permit to
operate. A fee of $2. 00 shall be charged, except
to any state or local governmental agency or public
district, for issuing a duplicate permit to operate.
It is hereby determined that the cost of issuing
permits and of inspections pertaining to such issu-
ance exceeds the fees prescribed.
Schedule 1 (Amended 7-1-64)
ELECTRIC MOTOR HORSEPOWER SCHEDULE
Any article, machine, equipment, or other con-
trivance where an electric motor is used as the
power supply shall be assessed a permit fee based
on the total rated motor horsepower of all electric
-------
Rules and Regulations of the Air Pollution Control District
839
motors included in any article, machine, equip-
ment or other contrivance, in accordance with the
following schedule:
Horsepower
Fee
a)
b)
c)
d)
e)
f)
g)
h)
up to and including 2-1/2 $ 40. 00
greater than 2-1/2 but less than 5 . . 100. 00
5 or greater but less than 15 200. 00
15 or greater but less than 45 300. 00
45 or greater but less than 65 400. 00
65 or greater but less than 125 . . . 500. 00
125 or greater but less than 200 . . . 600. 00
200 or greater 800. 00
Schedule 2 (Amended 7-1-64)
FUEL BURNING EQUIPMENT SCHEDULE
Any article, machine, equipment or other contriv-
ance in which fuel is burned, with the exception of
incinerators which are covered in Schedule 4, shall
be assessed a permit fee based upon the design
fuel consumption of the article, machine, equip-
ment or other contrivance expressed in thousands
of British thermal units (BTU) per hour, using
gross heating values of the fuel, in accordance
with the following schedule:
1000 British Thermal Units per Hour Fee
a) up to and including 150 $ 40. 00
b) greater than 150 but less than 400 . . 100. 00
c) 400 or greater but less than 650 . . . 200. 00
d) 650 or greater but less than 1500 . . 300. 00
e) 1500 or greater but less than 2500. . 400. 00
f) 2500 or greater but less than 5000. . 500.00
g) 5000 or greater but less than 15000 . 600. 00
h) 15000 or greater 800. 00
Schedule 3 (Amended 7-1-64)
ELECTRICAL ENERGY SCHEDULE
Any article, machine, equipment or other contriv-
ance which uses electrical energy, with the excep-
tion of electric motors covered in Schedule 1, shall
be assessed a permit fee based on the total kilo-
volt ampere (KVA) ratings, in accordance with the
following schedule:
Kilovolt Amperes
Fee
a)
b)
c)
a)
e)
f)
g)
up to and including 20
greater than 20 but less than 40 .
40 or greater but less than 145 . .
145 or greater but less than 450 .
450 or greater but less than 4500
$ 40.00
100.00
200.00
300.00
400.00
00
00
45000 or greater 800. 00
4500 or greater but less than 14500 . . 500,
14500 or greater but less than 45000 . 600,
Schedule 4 (Amended 7-1-64)
INCINERATOR SCHEDULE
Any article, machine, equipment or other contriv-
ance designed and used primarily to dispose of
combustible refuse by wholly consuming the mate-
rial charged leaving only the ashes or residue shall
be assessed a permit fee based on the following
schedule of the maximum horizontal inside cross
sectional area, in square feet, of the primary com-
bustion chamber:
Area, in Square Feet
Fee
a) up to and including 3 $ 40. 00
b)
c)
d)
e)
f)
g)
h)
greater than 3 but less than 4.
4 or greater but less than 7 . .
7 or greater but less than 10 .
10 or greater but less than 15.
15 or greater but less than 23.
23 or greater but less than 40.
40 or greater
100.
200.
300.
400.
500.
600.
00
00
00
00
00
00
800.00
Schedule 5 (Amended 7-1-64)
STATIONARY CONTAINER SCHEDULE
Any stationary tank, reservoir, or other contain-
er shall be assessed a permit fee based on the fol-
lowing schedule of capacities in gallons or cubic
equivalent:
Gallons
Fee
a) up to and including 4000 ......... $ 40. 00
b) greater than 4000 but less than
10000 .................... 60. 00
c) 10000 or greater but less than
40000 .................... 100. 00
d) 40000 or greater but less than
100000 ................... 200. 00
e) 100000 or greater but less than
400000 ................... 300. 00
400000 or greater but less than
1000000 .................. 400. 00
1000000 or greater but less than
4000000 .................. 500. 00
4000000 or greater ............ 600. 00
f)
g)
h)
Schedule 6 (Amended 7-1-64)
MISCELLANEOUS SCHEDULE
Anyarticle, machine, equipment or other contriv-
ance which is not included in the preceding sched-
ules shall be assessed a permit fee of $40. 00.
-------
840
RULES AND REGULATIONS
RULE 42.
FEES
(Amended 3-14-63) HEARING BOARD
a. Every applicant or petitioner for variance, or
for the extension, revocation or modification
of a variance, or for an appeal from a denial
or conditional approval of an authority to con-
struct, permit to operate or permit to sell or
rent, except any state or local governmental
agency or public district, shall pay to the
Clerk ol" the Hearing Board, on filing, a fee
in the sum of $16. 50. It is hereby determined
that the cost of administration of Article 5,
ChapterZ, Division20, Health and Safety Code,
or Rule 25 of these Rules and Regulations, ex-
ceeds $16. 50 per petition.
b. (Amended 3-14-63) Any person requesting a
transcript of the hearing shall pay the cost of
such transcript.
c. This rule shall not apply to petitions filed by
the Air Pollution Control Officer.
air contaminant for a period or periods aggregating
more than three minutes in any one hour which is:
a. As dark or darker in shade as that designated
as No. 2 on the Ringelmann Chart, as pub-
lished by the United States Bureau of Mines,
or
b. Of such opacity as to obscure an observer's
view to a degree equal to or greater than does
smoke described in subsection (a) of this Rule.
RULE 51. NUISANCE
A person shall not discharge iromany source what-
soever such quantities of air contaminants or other
material which cause injury, detriment, nuisance
or annoyance to any considerable number of per-
sons or to the public or which endanger the com-
fort, repose, health or safety of any such persons
or the public or which cause or have a natural ten-
dency to cause injury or damage to business or
property.
RULE 43. ANALYSIS FEES
Whenever the Air Pollution Control Officer finds
that an analysis of the emission from any source
is necessary to determine the extent and amount
of pollutants being discharged into the atmosphere
which cannot be determined by visual observation,
he may order the collection of samples and the
analysis made by qualified personnel of the Air
Pollution Control District. The time required for
collecting samples, making the analysis and pre-
paring the necessary reports, but excluding time
required in going to and from such premises shall
be charged against the owner or operator of said
premises in a reasonable sum to be determined by
the Air Pollution Control Officer, which said sum
is not to exceed the actual cost of such work.
RULE 52.
MATTER
(Amended 3-2-6?) PARTICULATE
RULE 44.
FOR:
TECHNICAL REPORTS - CHARGES
Information, circulars, reports of technical work,
and other reports prepared by the Air Pollution
Control District when supplied to othei govern-
mental agencies or individuals or groups request-
ing copies of the same may be charged for by the
District in a sum not to exceed the cost of prepara-
tion and distribution of such documents. All such
monies collected shall be turned into the general
funds of the said District.
REGULATION IV. PROHIBITIONS
RULE 50. RINGELMANN CHART
A person shall not discharge into the atmosphere
from any single source of emission whatsoever any
Except as otherwise provided in Rules 53 and 54,
a person shall not discharge into the atmosphere
from any source particulate matter in excess of
0. 3 grain per cubic foot of gas at standard condi-
tions.
RULE 53. (Amended 1-16-58) SPECIFIC CON-
TAMINANTS
A person shall not discharge into the atmosphere
from any single s our ce of emis sion whats oever any
one or more of the following contaminants, in any
state or combination thereof, exceeding in concen-
tration at the point of discharge:
a. Sulphur Compounds calculated as sulphur di-
oxide (SO2): 0. 2 per cent, by volume.
b. (Amended 1-16-58) Combustion Contaminants:
0. 3 grain per cubic foot of gas calculated to
12 per cent of carbon dioxide (002) at stan-
dard conditions. In measuring the combus-
tion contaminants from incinerators used to
dispose of combustible refuse by burning, the
carbon dioxide (CO^) produced by combustion
of any liquid or gaseous fuels shall be excluded
from the calculation to 12 per cent of carbon
dioxide (CO2).
RULE 53. 1. SCAVENGES PLANTS
Where a separate source of air pollution is a. scav-
enger or recovery plant, recovering pollutants
which would other wise be emitted to the atmosphere,
-------
Rules and Regulations of the Air Pollution Control District
841
the Air Pollution Control Officer may grant a per-
mit to operate -where the total emission of pollu-
tants is substantially less with the plant in opera-
tion than when closed, even though the concentra-
tion exceeds that permitted by Rule 53 (a). The Air
Pollution Control Officer shall report immediate-
ly in writing to the Air Pollution Control Board the
grantingofanysuchperm.it, together with the facts
and reasons therefor.
RULE 54. DUST AND FUMES
A person shall not discharge in any one hour from
any sour ce whatsoever dust or fumes in total quan-
tities in excess of the amount shown in the follow-
ing table: (see next page)
To use the following table, take the process weight
per hour as such is defined in Rule 2(j). Then find
this figure on the table, opposite which is the max-
imum number of pounds of contaminants which may
be discharged into the atmosphere in any one hour.
As an example, if A has a process which emits con-
taminants into the atmosphere and which process
takes 3 hours to complete, he "will divide the weight
of all materials in the specific process, in this
example, 1,500 Ibs. by 3 giving a process weight
per hour of 500 Ibs. The table shows that A may
not discharge more than 1. 77 Ibs. in any one hour
during the process. Where the process weight per
hour falls between figures in the left hand column,
the exact weight of permitted discharge may be in-
terpolated.
RULE 55. (Amended 1-16-58) EXCEPTIONS
The provisions of Rule 50 do not apply to:
a. Smoke from fires set by or permitted by any
public officer if such fire is set or permission
given in the performance of the official duty
of such officer, and such fire in the opinion
of such officer is necessary:
(1) For the purpose of the prevention of a fire
hazard which cannot be abated by any other
means, or
(2) The instruction of public employees in the
methods of fighting fire.
b. Smoke from fires set pursuant to permit on
property used for industrial purposes for the
purpose of instruction of employees in methods
of fighting fire.
c. Agricultural operations in the growing of crops,
or raising of fowls or animals.
d. The use of an orchard or citrus grove heater
which does notproduce unconsumed solid car-
bonaceous matter at a rate in excess of one
(1) gram per minute.
The use of other equipment in agricultural
operations in the growing crops, or raising
of fowls or animals.
RULE 56. (Amended 1-16-58) STORAGE OF PE-
TROLEUM PRODUCTS
A person shall not place, store or hold in any sta-
tionary tank, reservoir or other container of more
than 40, 000 gallons capacity any gasoline or any
petroleum distillate having a vapor pressure of
1.5 pounds per square inch absolute or greater
under actual storage conditions, unless such tank,
reservoir or other container is a pressure tank
maintaining working pressures sufficient at all
times to prevent hydrocarbon vapor or gas loss to
the atmosphere, or is designed and equipped with
one of the following vapor loss control devices,
properly installed, in good working order and in
operation:
a. A floating roof, consisting of a pontoon type
or double-deck type roof, resting on the sur-
face of the liquid contents and equipped with
a closure seal, or seals, to close the space
between the roof edge and tank wall. The con-
trol equipment provided for in this paragraph
shall not be used if the gasoline or petroleum
distillate has a vapor pressure of 11.0 pounds
per square inch absolute or greater under ac-
tual storage conditions. All tank gauging and
sampling devices shall be gas-tight except
when gauging or sampling is taking place.
b. A vapor recovery system, consisting of a
vapor gathering system capable of collecting
the hydrocarbon vapors and gases discharged
and a vapor disposal system capable of pro-
cessing such hydrocarbon vapors and gases so
as to prevent their emission to the atmosphere
and with all tank gauging and sampling devices
gas-tight except when gauging or sampling is
taking place.
c. Other equipment of equal efficiency, provided
such equipment is submitted to and approved
by the Air Pollution Control Officer.
RULE 57. (Amended 1-16-58) OPEN FIRES
A person shall not burn any combustible refuse in
any open outdoor fire within the Los Angeles Basin,
except:
a. When such fire is set or permission for such
fire is given in the performance of the official
-------
842
RULES AMD REGULATIONS
TABLE
*Process
Wt/hr(lbs)
50
100
150
200
250
300
350
400
450
500
550
600
650
700
750
800
850
900
950
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2100
2200
2300
2400
2500
2600
2700
2800
2900
3000
3100
3200
3300
Maximum Weight
Disch/hr(lbs)
.24
.46
.66
.85
1.03
1.20
1. 35
1.50
1. 63
1.77
1.89
2.01
2. 12
2.24
2. 34
2.43
2.53
2.62
2.72
2.80
2.97
3. 12
3.26
3.40
3.54
3.66
3. 79
3.91
4. 03
4. 14
4. 24
4. 34
4.44
4. 55
4. 64
4.74
4.84
4. 92
5.02
5. 10
5. 18
5.27
5.36
*Process
Wt/hr(lbs)
3400
3500
3600
3700
3800
3900
4000
4100
4200
4300
4400
4500
4600
4700
4800
4900
5000
5500
6000
6500
7000
7500
8000
8500
9000
9500
10000
11000
12000
13000
14000
15000
16000
17000
18000
19000
20000
30000
40000
50000
60000
or
more
Maximum Weight
Disch/hr(lbs)
5.44
5.52
5.61
5.69
5. 77
5.85
5.93
6.01
6.08
6. 15
6.22
6. 30
6.37
6.45
6.52
6.60
6.67
7.03
7.37
7.71
8.05
8.39
8.71
9. 03
9.36
9.67
10. 0
10.63
11.28
11.89
12. 50
13. 13
13. 74
14. 36
14.97
15.58
16. 19
22.22
28.3
34.3
40.0
*See Definition in Rule 2(j).
duty of any public officer, and such fire in the
opinion of such officer is necessary:
1. For the purpose of ilie o * jvention of a fire
hazard which cannot be abated by any oth-
er means, or
2. The instruction of public employees in the
methods of fighting fire.
b. When such fire is set pursuant to permit on
property used for industrial purposes for the
purpose of instruction of employees in meth-
ods of fighting fire.
When suchfire is set inthe course of any agri-
cultural operation in the growing of crops, or
raising of fowls or animals.
-------
Rules and Regulations of the Air Pollution Control District
843
These exceptions shall not be effective on any cal-
endar day on which the Air Pollution Control Offi-
cer determines that:
1. The inversion base at 4:00 A. M. , Pacific
Standard Time, will be lower than one
thousand five hundred feet above mean sea
level, and
2. The maximum mixing height will not be
above three thousand five hundred feet,
and
3. The average surface "wind speed between
6:00 A.M. and 12:00 noon, Pacific Stan-
dard Time, will not exceed five miles per
hour.
RULE 58.
BURNING
(Amended 3-14-63) INCINERATOR
A person shall not burn any combustible refuse in
any incinerator -within the Los Angeles Basin ex-
cept in a multiple-chamber incinerator as de-
scribed in Rule 2(p), or in equipment found by the
Air Pollution Control Officer in advance of such
use to be equally effective for the purpose of air
pollution control as an approved multiple-chamber
incinerator.
RULE 59. (Amended 1-16-58) OIL-EFFLUENT
WATER SEPARATOR
A person shall not use any compartment of any
single or multiple compartment oil-effluent water
separator which compartment receives effluent
water containing 200 gallons a day or more of any
petroleum product or mixture of petroleum prod-
ucts from any equipment processing, refining,
treating, storing or handling kerosine or other
petroleum product of equal or greater volatility
than kerosine, unless such compartment is equip-
ped with one of the following vapor loss control de-
vices, properly installed, in good working order
and in operation:
a. A solid cover with all openings sealed and to-
tally enclosing the liquid contents. All gaug-
ing and sampling devices shall be gas -tight ex-
cept when gauging or sampling is taking place.
b. A floating roof, consisting of a pontoon type
or double-deck type roof, resting on the sur-
face of the liquid contents and equipped with
a closure seal, or seals, to close the space
between the roof edge and container wall. All
gauging and sampling devices shall be gas-
tight except when gauging or sampling is tak-
ing place.
c. A vapor recovery system, consisting of a va-
por gathering system capable of collecting the
hydrocarbon vapors and gases discharged and
a vapor disposal system capable of processing
such hydrocarbon vapors and gases so as to
prevent their emission to the atmosphere and
with all tank gauging and sampling devices
gas-tight except when gauging or sampling is
taking place.
d. Other equipment of equal efficiency, provided
such equipment is submitted to and approved
by the Air Pollution Control Officer.
This rule shall not apply to any oil-effluent water
separator used exclusively in conjunction with the
production of crude oil.
For the purpose of this rule, "kerosine" is defined
as any petroleum product •which, when distilled by
ASTM standard test Method D 86-56, will give a
temperature of 401 °F. or less at the 10 per cent
point recovered.
RULE 60. (Adopted 12-15-55) CIRCUMVENTION
A person shall not build, erect, install, or use
any article, machine, equipment or other contriv-
ance, the use of which, without resulting in a re-
duction in the total release of air contaminants to
the atmosphere, reduces or conceals an emission
which would otherwise constitute a violation of
Division 20, Chapter 2 of the Health and Safety Code
of the State of California or of these Rules and Reg-
ulations. This Rule shall not apply to cases in
which the only violation involved is of Section 24243
of the Health and Safety Code of the State of Cali-
fornia, or of Rule 51 of these Rules and Regula-
tions.
RULE 61. (Amended 3-14-63) GASOLINE LOAD-
ING INTO TANK TRUCKS AND TRAILERS
A person shall not load gasoline into any tank truck
or trailer from any loading facility unless such
loading facility is equipped with a vapor collection
and disposal system or its equivalent, properly in-
stalled, in good working order and in operation.
When loading is effected through the hatches of a
tank truck or trailer with a loading arm equipped
with a vapor collecting adaptor, a pneumatic, hy-
draulic or other mechanical means shall be pro-
vided to for ce a vapor-tight seal between the adapt-
or and the hatch. A means shall be provided to
prevent liquid gasoline drainage from the loading
device when it is removed from the hatch of any
tank truck or trailer, or to accomplish complete
drainage before such removal.
-------
844
RULES AND REGULATIONS
When loading is effected through means other than
hatches, all loading and vapor lines shall be equip-
ped with fittings which make vapor-tight connec-
tions and which close automatically when discon-
nected.
The vapor disposal portion of the system shall con-
sist of one of the following:
a. A vapor-liquid absorber system with a mini-
mum recovery efficiency of 90 per cent by
weight of all the hydrocarbon vapors and gas-
es entering such disposal system.
b. A variable vapor space tank, compressor, and
fuel gas system of sufficient capacity to re-
ceive all hydrocarbon vapors and gases dis-
placed from the tank trucks and trailers being
loaded.
c. (Amended 3-14-63) Other equipment of at least
90 per cent efficiency, provided such equip-
ment is submitted to and approved by the Air
Pollution Control Officer.
This rule shall not apply to the loading of gasoline
into tank trucks and trailers from any loading fa-
cility from which not more than 20, 000 gallons of
gasoline are loaded in any one day.
For the purpose of this rule, any petroleum dis-
tillate having a Reid vapor pressure of four pounds
or greater shall be included by the term "gasoline".
(Amended 12-4-58) For the purpos-e of this rule,
"loading facility" means any aggregation or com-
bination of gasoline loading equipment which is both
(1) possessed by one person, and (2) located so
that all the gasoline loading outlets for such ag-
gregation or combination of loading equipment can
be encompassed within any circle of 300 feet in di-
ameter.
RULE 62. (Amended 3-16-61) SULFUR CON-
TENTS OF FUELS
A person shall not burn within the Los Angeles
Basin at any time between May 1 and September
30, both dates inclusive, during the calendar year
1959, and each year thereafter between April 15
and November 15 both inclusive, of the same cal-
endar year, any gaseous fuel containing sulfur
compounds in excess of 50 grains per 100 cubic
feet of gaseous fuel, calculated as hydrogen sul-
fide at standard conditions, or any liquid fuel or
solid fuel having a sulfur content in excess of 0. 5
per cent by weight.
The provisions of this rule shall not apply to:
a. The burning of sulfur, hydrogen sulfide, acid
sludge or other sulfur compounds in the manu-
facturing of sulfur or sulfur compounds.
b. The incinerating of waste gases provided that
the gross heating value of such gases is less
than 300 British thermal units per cubic foot
at standard conditions and the fuel used to in-
cinerate such waste gases does not contain
sulfur or sulfur compounds in excess of the
amount specified in this rule.
c. The use of solid fuels in any metallurgical
process.
d. The use of fuels where the gaseous products
of combustion are used as raw materials for
other processes.
e. The use of liquid or solid fuel to propel or test
any vehicle, aircraft, missile, locomotive,
boat or ship.
f. The use of liquid fuel whenever the supply of
gaseous fuel, the burning of which is permit-
ted by this rule, is not physically available to
the user due to accident, act of God, act of
war, act of the public enemy, or failure of the
supplier.
RULE 62. 1 (Adopted 1-14-64)
a. A per son shall not burn within the Los Angeles
Basin at any time between the days of Novem-
ber 16 of any year arid April 14 of the next
succeeding calendar year, both dates inclusive,
any fuel described in the first paragraph of
Rule 62 of these Rules and Regulations.
b. The provisions of this Rule do not apply to:
(1) Any use of fuel described in Subsections
a, b, c, d, e, and f of said Rule 62 under
the conditions and for the uses set forth
in said Subsections.
(2) The use of liquid fuel during a period for
which the supplier of gaseous fuel, the
burning of which is not prohibited by this
Rule, interrupts the delivery of gaseous
fuel to the user.
c. Every holder of, and every applicant for a per-
mit to operate fuel-burning equipment under
these Rules and Regulations shall notify the
air pollution control officer in the manner and
form prescribed by him, of each interruption
in and resumption of delivery of gaseous fuel
to his equipment.
-------
Rules and Regulations of the Air Pollution Control District
845
RULE 63. (Amended 1-25-62) GASOLINE SPECI-
FICATIONS
a. A person shall not, after June 30, I960, sell
or supply for use within the District as a fuel
for motor vehicles as defined by the Vehicle
Code of the State of California, gasoline hav-
ing a degree of unsaturation greater than that
indicated by a Bromine Number of 30 as de-
termined by ASTM Method D1159-57T modi-
fied by omis sion of the mercuric chloride cat-
alyst.
b. For the purpose of this rule, the term "gaso-
line" means any petroleum distillate having a
Reid vapor pressure of more than four pounds.
RULE 65. (Amended 6-1-65) GASOLINE LOADING
INTO TANKS
A person shall not after January 1, 1965, load or
permit the loading of gasoline into any stationary
tank with a capacity of 250 gallons or more from
any tank truck or trailer, except through a perma-
nent submerged fill pipe, unless such tank is
equipped with a vapor loss control device as de-
scribed in Rule 56, or is a pressure tank as de-
scribed in Rule 56.
The provisions of the first paragraph of this rule
shall not apply to the loading of gasoline into any
tank having a capacity of less than 2, 000 gallons
which was installed prior to the date of adoption of
this rule nor to any underground tank installed pri-
or to the date of adoption of this rule where the fill
line between the fill connection and tank is offset.
RULE 64. (Amended 3-2-67) REDUCTION OF
ANIMAL MATTER
A person shall not operate or use any article, ma-
chine, equipment or other contrivance for the re-
duction of animal matter unless all gases, vapors
and gas-entrained effluents from such an article,
machine, equipment or other contrivance are:
a. Incinerated at temperatures of not less than
1200 degrees Fahrenheit for a period of not
less than 0. 3 second, or
b. Processed in such a manner determined by
the Air Pollution Control Officer to be equally,
or more, effective for the purpose of air pol-
lution control than (a) above.
A per son incinerating or processing gases, vapors
or gas-entrained effluents pursuant to this rule
shall provide, properly install and maintain in cal-
ibration, in good working order and in operation
devices, as specified in the Authority to Construct
or Permit to Operate or as specified by the Air
Pollution Control Officer, for indicating tempera-
ture, pressure or other operating conditions.
For the purpose of this rule, "reduction" is de-
fined as any heated process, including rendering,
cooking, drying, dehydrating, digesting, evapo-
rating and protein concentrating.
The provisions of this rule shall not apply to any
article, machine, equipment or other contrivance
used exclusively for the processing of food for hu-
man consumption.
Any person operating or using any gasoline tank
with a capacity of 250 gallons or more installed
prior to the date of adoption of this rule shall apply
for a permit to operate such tank before January 1,
1965. The provisions of Rule 40 shall not apply
during the period between the date of adoption of
this rule and January 1, 1965, to any gasoline tank
installed prior to the date of adoption of this rule
provided an application for permit to operate is
filed before January 1, 1965.
A person shall not install any gasoline tank with a
capacity of 250 gallons or more unless such tank
is equipped as described in the first paragraph of
this rule.
For the purpose of this rule, the term "gasoline"
is defined as any petroleum distillate having a Reid
vapor pressure of 4 pounds or greater.
For the purpose of this rule, the term "submerged
fill pipe" is defined as any fill pipe the discharge
opening of which is entirely submerged when the
liquid level is 6 inches above the bottom of the
tank. "Submerged fill pipe" when applied to a tank
which is loaded from the side is defined as any fill
pipe the discharge opening of which is entirely sub-
merged when the liquid level is 18 inches above the
bottom of the tank.
(Adopted 6-1-65) The provisions of this rule do not
apply to any stationary tank which is used primar-
ily for the fueling of implements of husbandry, as
such vehicles are defined in Division 16 (Section
36000, et seq.) of the Vehicle Code.
RULE 66. (Adopted 7-28-66) ORGANIC SOLVENTS
a. A person shall not discharge more than 15
pounds of organic materials into the atmo-
sphere in any one day from any article, ma-
chine, equipment or other contrivance in which
-------
846
RULES AND REGULATIONS
any organic solvent or any material containing
organic s olvent comes into contact with flame
or is baked, heat-cured or heat-polymerized,
in the presence of oxygen, unless all organic
materials discharged from such article, ma-
chine, equipment or other contrivance have
been reduced either by at least 85 per cent
overall or to not more than 15 pounds in any
one day.
b. A person shall not discharge more than 40
pounds of organic material into the atmosphere
in any one day from any article, machine,
equipment or other contrivance used under
conditions other than described in section (a),
for employing, applying, evaporating or dry-
ing any photochemically reactive solvent, as
defined in section (k), or material containing
such solvent, unless all organic materials dis -
charged from such article, machine, equip-
ment or other contrivance have been reduced
either by at least 85 per cent overall or to not
more than 40 pounds in any one day.
c. Any series of articles, machines, equipment
or other contrivances designed for processing
a continuously moving sheet, web, strip or
wire which is subjected to any combination of
operations described in sections (a) or (b) in-
volving any photochemically reactive solvent,
as defined in section (k), or material contain-
ing such solvent, shall be subject to compli-
ance with section (b). Where only non-photo-
chemically reactive solvents or material con-
taining only non-photochemically reactive sol-
vents are employed or applied, and where any
portion or portions of said series of articles,
machines, equipment or other contj-ivances
involves operations described in section (a),
said portions shall be collectively subject to
compliance with section (a).
d. Emissions of organic materials to the atmo-
sphere from the clean-up with photochemically
reactive solvent, as defined in section (k), of
any article, machine, equipment or other con-
trivance described in sections (a), (b) or (c),
shall be included with the other emissions of
organic materials from that article, machine,
equipment or other contrivance for determin-
ing compliance with this rule.
e. Emissions of organic materials to the atmo-
sphere as a result of spontaneously continu-
ing drying of products for the first 12 hours
after their removal from any article, machine,
equipment or other contrivance described in
sections (a), (b) or (c), shall be included with
other emissions of organic materials from
that article, machine, equipment or other con-
trivance for determining compliance with this
rule.
f. Emissions of organic materials into the at-
mosphere required to be controlled by sections
(a), (b) or (c), shall be reduced by:
(1) Incineration, provided that 90 per cent or
more of the carbon in the organic mate-
rial being incinerated is oxidized to car-
bon dioxide, or
(Z) Adsorption, or
(3) Processing in a manner determined by the
Air Pollution Control Officer to be not
less effective than (1) or (Z) above.
g. A person incinerating, adsorbing, or other-
wise processing organic materials pursuant
to this rule shall provide, properly install and
maintain in calibration, in good working order
and in operation, devices as specified in the
authority to construct or the permit to operate,
or as specified by the Air Pollution Control
Officer, for indicating temperatures, pres-
sures, rates of flow or other operating condi-
tions necessary to determine the degree and
effectiveness of air pollution control.
h. Any person using organic solvents or any ma-
terials containing organic solvents shall supply
the Air Pollution Control Officer, upon request
and in the manner and form prescribed by him,
written evidence of the chemical composition,
physical properties and amount consumed for
each organic solvent used.
i. The provisions of this rule shall not apply to:
(1) The manufacture of organic solvents, or
the transport or storage of organic sol-
vents or materials containing organic sol-
vents .
(Z) The use of equipment for which other re-
quirements arespecified by Rules 56, 59,
61 or 65 or which are exempt from air
pollution control requirements by said
rules.
(3) The spraying or other employment of in-
secticides, pesticides or herbicides.
(4) The employment, application, evapora-
tion or drying of saturated halogenated hy-
drocarbons or perchloroethylene.
j. For the purposes of this rule, organic sol-
vents include diluents andthinners and are de-
fined as organic materials which are liquids
at standard conditions and which are used as
dissolvers, viscosity reducers or cleaning
agents.
-------
Rules and Regulations of the Air Pollution Control District
847
k. For the purposes of this rule, a photochem-
ically reactive solvent is any solvent with an
aggregate of more than 20 per cent of its total
volume composed of the chemical compounds
classified below or which exceeds any of the
following individual percentage composition
limitations, referred to the total volume of
solvent:
(1) A combination of hydrocarbons, alcohols,
aldehydes, esters, ethers or ketones hav-
ing an olefinic or cycloolefinic type of un-
saturation: 5 per cent;
(2) A combination of aromatic compounds with
eight or more carbon atoms to the mole-
cule except ethylbenzene: 8 per cent;
(3) A combination of ethylbenzene, ketones
having branched hydrocarbon structures,
trichloroethylene or toluene: 20 per cent.
Whenever any organic solvent or any constituent
of an organic solvent may be classified from its
chemical structure into more than one of the above
groups of organic compounds, it shall be consid-
ered as a member of the most reactive chemical
group, that is, that group having the least allow-
able per cent of the total volume of solvents.
1. For the purposes of this rule, organic mate-
rials are defined as chemical compounds of
carbon excluding carbon monoxide, carbon di-
oxide, carbonic acid, metallic carbides, me-
tallic carbonates and ammonium carbonate.
m. This rule shall be effective on the date of its
adoption as to any article, machine, equip-
ment or other contrivance, not then completed
and put into service. As to all other articles,
machines, equipment or other contrivances,
this rule shall be effective:
(1) On July 1, 1967, for those emitting 500
pounds or more of organic materials in
any one day.
(2) On October 1, 1967, for those emitting
100 pounds or more but less than 500
pounds of organic materials in any one
day.
(3) On March 1, 1968, for those subject to
compliance with section (a), and emitting
15 pounds or more but less than 100 pounds
of organic materials in any one day, and
for those subject to compliance with sec-
tion (b), and emitting 40 pounds or more
but less than 100 pounds in any one day.
RULE 66. 1 (Adopted 7-28-66) ARCHITECTURAL
COATINGS
a. After July 1, 1967, a person shall not sell or
offer for sale for use in Los Angeles County,
in containers of one quart capacity or larger,
any architectural coating containing photo-
chemically reactive solvent, as defined in Rule
66(k).
b. After July 1, 1967, a person shall not employ,
apply, evaporate or dry in Los Angeles County
any architectural coating, purchased in con-
tainers of one quart capacity or larger, con-
taining photochemically reactive solvent, as
defined in Rule 66(k).
c. After July 1, 1967, a person shall not thin or
dilute any architectural coating with a photo-
chemically reactive s olvent, as defined in Rule
66(k).
d. For the purposes of this rule, an architectural
coating is defined as a coating used for resi-
dential or commercial buildings and their ap-
purtenances; or industrial buildings.
RULE 66.2 (Adopted 7-28-66) DISPOSAL AND
EVAPORATION OF SOLVENTS
A person shall not during any one day dispose of a
total of more than 1-1/2 gallons of any photochem-
ically reactive solvent, as defined in Rule 66(k), or
of any material containing more than 1-1/2 gallons
of any such photochemically reactive solvent by
any means which will permit the evaporation of
such solvent into the atmosphere.
REGULATION V. PROCEDURE BEFORE THE
HEARING BOARD
RULE 75. GENERAL
This regulation shall apply to all hearings before
the Hearing Board of the Air Pollution Control Dis-
trict.
RULE 76. (Revised 8-25-64) FILING PETITIONS
Requests for hearing shall be initiated by the fil-
ing of a petition in triplicate with the Clerk of the
Hearing Board at Room 601B, 220 North Broadway,
Los Angeles, California, 90012, and the payment
of the fee of $16. 50 provided for in Rule 42 of these
Rules and Regulations, after service of a copy of
the petition has been made on the Air Pollution
Control Officer at 434 South San Pedro Street, Los
Angeles, California, 90013, and one copy on the
holder of the permit or variance, if any, involved.
Service may be made in person or by mail, and
service may be proved by written acknowledgment
-------
848
RULES AND REGULATIONS
c.
of the person served or by the affidavit of the per-
son making the service.
RULE 77. (Amended 1-16-58) CONTENTS OF
PETITIONS
Every petition shall state:
a. The name, address and telephone number of
the petitioner, or other person authorized to
receive service of notices.
b. Whether the petitioner is an individual, co-
partnership, corporation or other entity, and
names and address of the partners if a. co-
partnership, names and address of the offi-
cers, if a corporation, and the names and ad-
dress of the persons in control, if other en-
tity.
The type of business or activity involved in the
application and the street address at which it
is conducted.
d. A brief description of the article, machine,
equipment or other contrivance, if any, in-
volved in the application.
e. The section or rule under which the petition
is filed; that is, whether petitioner desires a
hearing:
(1) to determine whether a permit shall be
revoked or suspended permit reinstated
under Section 24274, Health and Safety
Code of the State of California;
(2) for a variance under Section 24292, Health
and Safety Code;
(3) to revoke or modify a variance under Sec-
tion 24298, Health and Safety Code;
(4) (Amended 1-16-68) to review the denial
or conditional granting of an authority to
construct, permit to operate or permit to
sell or rent under Rule 25 of these Rules
and Regulations.
f. Each petition shall be signed by the petitioner,
or by some person on his behalf, and where
the person signing is not the petitioner it shall
set forth his authority to sign.
g. Petitions for revocation of permits shall allege
in addition the rule under which permit was
granted, the rule or section which is alleged
to have: been violated, together •with a brief
statement of the facts constituting such alleged
violation.
h. Petitions for reinstatement of suspended per-
mits shall allege in addition the rule under
which the permit was granted, the request and
alleged refusal which formed the basis for
such suspension, together with a brief state-
ment as to why information requested, if any,
was not furnished, whether such information
is believed by petitioner to be pertinent, and,
if so, when it will be furnished.
i. All petitions shall be typewritten, double
spaced, on legal or letter size paper, on one
side of the paper only, leaving a margin of at
least one inch at the top and left side of each
sheet.
RULE 78. (Amended 4-2-64). PETITIONS FOR
VARIANCES
In addition to the matters required by Rule 77, pe-
titions for variances shall state briefly:
a. The section, rule or order complained of.
b. The facts showing why compliance with the
section, rule, or order is unreasonable.
c. For what period of time the variance is sought
and why.
d. The damage or harm resulting or which would
result to petitioner from a compliance with
such section, rule or order,
e. The requirements which, petitioner can meet
and the date when petitioner can comply with
such requirements.
f. The advantages and disadvantages to the resi-
dents of the district resulting from requiring
compliance or resulting from granting a vari-
ance.
g. Whether or not operations under such variance,
if granted, would constitute a nuisance.
h. Whether or not any case involving the same
identical equipment or process is pending in
any court, civil or criminal.
i. (Amended 4-2-64) Whether or not the subject
equipment or process is covered by a permit
to operate issued by the Air Pollution Control
Officer.
RULE 79.
DENIAL
(Amended 1-16-58) APPEAL FROM
A petition to review a denial or conditional approval
of an authority to construct, permit to operate or
permit to sell or rent shall, in addition to the mat-
-------
Rules and Regulations of the Air Pollution Control District
849
ters required by Rule 77, set forth a summary of
the application or a copy thereof and the alleged
reasons for the denial or conditional approval and
the reasons for appeal.
RULE 80. FAILURE TO COMPLY WITH RULES
The Clerk of the Hearing Board shall not accept
for filing any petition which does not comply with
these Rules relating to the form, filing and ser-
vice of petitions unless the chairman or any two
members of the Hearing Board direct otherwise
and confirm such direction in writing. Such di-
rection need not be made at a meeting of the Hear-
ing Board. The chairman or any two members,
without a meeting, may require the petitioner to
state further facts or reframe a petition so as to
disclose clearly the issues involved.
RULE 82. ANSWERS
Any per son may file an answer within 10 days after
service. All answers shall be served the same as
petitions under Rule 76.
RULE 83. DISMISSAL OF PETITION
The petitioner may dismiss his petition at any time
before submission of the case to the Hearing Board,
without a hearing or meeting of the Hearing Board.
The Clerk of the Hearing Board shall notify all
interested persons of such dismissal.
RULE 84. (Revised 8-25-64) PLACE OF HEARING
All hearings shall be held at Room 601B, 220 North
Broadway, Los Angeles, California, 90012, unless
some other place is designated by the Hearing
Board.
RULE 85. NOTICE OF HEARING
ered in the direct examination; to impeach
any •witness regardless of which party
first called him to testify; and to rebut the
evidence against him. If respondent does
not testify in his own behalf he may be
called and examined as if under cross-
examination.
(c) The hearing need not be conducted accord-
ing to technical rules relating to evidence
and witnesses. Any relevant evidence
shall be admitted if it is the sort of evi-
dence on which responsible persons are
accustomed to rely in the conduct of ser-
ious affairs, regardless of the existence
of any common law or statutory rule which
might make improper the admission of
such evidence over objection in civil ac-
tions. Hearsay evidence may be used for
the purpose of supplementing or explain-
ing any direct evidence but shall not be
sufficient in itself to support a finding un-
less it would be admissible over objection
in civil actions. The rules of privilege
shall be effective to the same extent that
they are now or hereafter may be recog-
nized in civil actions, and irrelevant and
unduly repetitious evidence shall be ex-
cluded.
RULE 87. PRELIMINARY MATTERS
Preliminary matters such as setting a date for
hearing, granting continuances, approving peti-
tions for filing, allowing amendments and other
preliminary rulings not determinative of the mer-
its of the case may be made by the chairman or
any two members of the Hearing Board without a
hearing or meeting of the Hearing Board and with-
out notice.
The Clerk of the Hearing Board shall mail or de-
liver a notice of hearing to the petitioner, the Air
Pollution Control Officer, the holder of the permit
or variance involved, if any, and to any person en-
titled to notice under Sections 24275, 24295 or
24299, Health and Safety Code.
RULE
OFFICIAL NOTICE
The Hearing Board may take official notice of any
matter which may be judicially noticed by the courts
of this State.
RULE 86. EVIDENCE
(a) Oral evidence shall be taken only on oath
or affirmation.
(b) Each party shall have these rights: to call
and examine witnesses; to introduce ex-
hibits; to cross-examine opposing wit-
nesses on any matter relevant to the is-
sues even though that matter was not cov-
RULE 89. CONTINUANCES
The chairman or any two members of the Hearing
Board shall grant any continuance of 15 days or
less, concurred in by petitioner, the Air Pollution
Control Officer and by every person who has filed
an answer in the action and may grant any reason-
able continuance; in either case such action may
beexparte, without a meeting of the Hearing Board
and without prior notice.
-------
850
RULES AND REGULATIONS
RULE 90. DECISION
The decision shall be in writing, served and filed
within 15 days after submission of the cause by the
parties thereto and shall contain a brief statement
of facts found to be true, the determination of the
issues presented and the order of the Hearing
Board. A copy shall be mailed or delivered to the
Air Pollution Control Officer, the petitioner and to
every person who has filed an answer or who has
appeared as a partyinperson or by counsel at the
hearing.
RULE 91. EFFECTIVE DATE OF DECISION
The decision shall become effective 15 days after
delivering or mailing a copy of the decision, as
provided in Rule 90, or the Hearing Board may
order that the decision shall become effective
sooner.
RULE 95. LACK OF PERMIT
The Hearing Board shall not receive or accept a
petition for a variance for the operation or use of
anyarticle, machine, equipment or other contriv-
ance until a permit to operate has been granted or
deniedby the Air Pollution Control Officer; except
that an appeal from a denial of a permit to operate
and a petition for a variance may be filed with the
Hearing Board in a single petition.. A variance
granted by the Hearing Board after a denial of a
permit to operate by the Air Pollution Control Of-
ficer may include a permit to operate for the dur-
ation of the variance.
shall first obtain a permit from the Air Pollution
Control Officer to do so.
RULE 103. (Amended 1-16-58) TRANSFER
A permit to operate shall not be transferable,
whether by operation of law or otherwise, either
from one location to another, from one piece of
equipment to another, or from one person to
another.
RULE 105. (Amended 1-16-58) APPLICATION
FOR PERMITS
Every application for a permit required under Rule
102 shall be filed in the manner and form required
by the Air Pollution Control Officer. Incomplete
applications will not be accepted.
RULE 106. ACTION ON APPLICATIONS
The Air Pollution Control Officer shall act on all
applications within a reasonable time and shall
notify the applicant in writing of the approval, con-
ditional approval or denial of the application.
RULE 107. (Amended 1-16-58) STANDARDS FOR
GRANTING PERMITS
The Air Pollution Control Officer shall deny a per-
mit if the applicant does not show that equipment
described in Rules 100 and 102 is so designed or
controlled that it will not produce unconsumed solid
carbonaceous matter at the rate in excess of one
(1) gram per minute except as prescribed under
Rule 108.
REGULATION VI. ORCHARD OR CITRUS GROVE
HEATERS
RULE 100. DEFINITION
"Orchard or citrus grove heater" means anyarticle,
machine, equipment or other contrivance, burning
any type of fuel, capable of emitting air contami-
nants, used or capable of being used for the pur-
pose of giving protection from frost damage.
RULE 101. (Amended 3-2-67) EXCEPTIONS
Rules 10, 14, 20, 21, 24, 40, 62 and 62. 1 do not
apply to orchard or citrus grove heaters.
RULE 102. (Amended 1-16-58) PERMITS RE-
QUIRED
Anyperson erecting, altering, replacing, operat-
ing or using any orchard or citrus grove heater
RULE 108. (Amended 1-16-58) CONDITIONAL
APPROVAL
a. The Air Pollution Control Officer may issue
a permit subject to conditions which will bring
the orchard or citrus grove heater within the
standards of Rule 107 in •which case the con-
ditions shall be specified in writing.
b. Erecting, altering, operating or using under
conditional permit shall be deemed acceptance
of all conditions so specified.
RULE 109. (Amended 1-16-58) DENIAL OF AP-
PLICATIONS
In the event of denial of a permit, the Air Pollution
Control Officer shall notify the applicant in writing
of the reasons therefor. Service of this notifica-
tion may be made in person or by mail, and such
service may be proved by the written acknowledg-
ment of the person served or affidavit of the per-
-------
Rules and Regulations of the Air Pollution Control District
851
son making the service. The Air Pollution Con-
trol Officer shall not accept a further application
unless the applicant has complied with the objec-
tions specified by the Air Pollution Control Officer
as his reasons for denial.
RULE 110. (Amended 1-16-58) APPEALS
Within 10 days after notice of denial or conditional
approval of a permit by the Air Pollution Control
Officer, the applicant may petition the Hearing
Board, in writing, for a public hearing. The Hear-
ing Board, after notice and a public hearing held
within 30 days after filing the petition, may sus-
tain or reverse the action of the Air Pollution Con-
trol Officer; such order may be made subject to
specified conditions.
RULE 120. (Effective 1-16-58) FEES
A request for a duplicate permit for orchard or
citrus grove heaters shall be made in writing to
the Air Pollution Control Officer within 10 days
after the destruction, loss or defacement of a per-
mit. The fee for issuing a duplicate permit shall
be $1.00.
RULE 130. (Amended 1-16-58) PROHIBITIONS
a. These rules prohibit the erecting, altering,
replacing, operating or using any orchard or
citrus grove heater which produces uncon-
sumed solid carbonaceous matter at the rate
of more than one (1) gram per minute, except
under the conditions as set forth in Rule 108.
b. Open fires for orchard or citrus grove heat-
ing are prohibited.
c. The use of rubber tires or any rubber products
in any combustion process in connection with
any orchard or citrus grove heating is hereby
prohibited.
d. (Amended 1-16-58) All types of orchard or cit-
rus grove heating equipment commonly known
or designated as follows:
1. Garbage pail
2. Smith Evans
3. Citrus with Olsen Stack
4. Canco 5 gallon
5. Dunn
6. Hamilton Bread Pan
7. Wheeling
8. Canco 3 gallon
9. Chinn
10. Baby Cone
11. Citrus Regular
e.
12. Stub Stack
13. Citrus 15-inch stack
14. Exchange Model 5-1/2-inch diameter stack
15. Exchange Model 6-inch diameter stack
16. Hy-Lo Drum
17. Hy-Lo Hot Blast
18. Pheysey Beacon
may not be used or operated for the purpose
of giving protection from frost damage.
(Amended 1-16-58) All types of orchard or cit-
rus grove heating equipment commonly known
or designated as follows:
Name
1. Hy-Lo 1929
2. Hy-Lo 148
3. Hy-Lo Double
Stack
4. Jumbo Cone
5. Lemora
6. National Double
Stack
7. Surplus Chemical
Warfare Service
Smoke Generator
Maximum Primary Air
Orifice in Square Inches
0. 606(equivalent
hole of 7/8 in.
0. 606(equivalent
hole of 7/8 in.
0. 606(equivalent
hole of 7/8 in.
0. 196(equivalent
hole of 1/2 in.
0. 606(equivalent
hole of 7/8 in.
0. 802(equivalent
hole of 7/8 in.
and one hole of
diameter)
0. 802 (equivalent
hole of 7/8 in.
and one hole of
diameter)
to one
diameter)
to one
diameter)
to one
diameter)
to one
diameter)
to one
diameter)
to one
diameter
1/2 in.
to one
diameter
1/2 in.
f.
may not be used or operated for the purpose
of giving protection from frost damage unless
the primary air orifice(s) contain(s) not more
than the maximum area designated above.
(Amended 1 -16-58) All types of orchard or cit-
rus grove heating equipment commonly known
or designated as follows:
Name
Maximum Primary Air
Orifice in Square Inches
Exchange Model
7 in. dia. stack
Hy-Lo 148 Special
Hy-Lo 230
Lazy Flame 24 in.
stack
Lazy Flame 18 in.
stack
National Junior
0. 606(equivalent
hole of 7/8 in.
0. 606(equivalent
hole of 7/8 in.
0, 606(equivalent
hole of 7/8 in.
0. 606(equivalent
hole of 7/8 in.
1. 212(equivalent
holes of 7/8 in.
1. 212 (e quivalent
holes of 7/8 in.
to one
diameter)
to one
diameter)
to one
diameter)
to one
diameter)
to two
diameter)
to two
diameter)
-------
852
RULES AND REGULATIONS
may not be used or operated for the purpose
of giving protection from frost damage unless
the primary air orifice(s) is (are) so adjusted
or regulated to a maximum opening of not
greater than the area designated above.
g. (Amended 1-16-58) Any new complete orchard
or citrus grove heating equipment of the dis-
tilling type notlisted in subsection "e" and "f"
of this rule must contain a primary air orifice
of such design that not more than one (1) gram
per minute of unconsumed solid carbonaceous
matter is emitted.
h. (Amended 1-16-58) No heater may be placed,
be permitted to be placed or be permitted to
remain in any orchard or citrus grove or in
any other place where heaters may be fired to
furnish protection from frost damage unless
a permit or conditional permit has been issued.
i. (Amended 1-16-58) The use or operation of
any partial assembly of any type heater for the
purpose of giving protection from frost damage
is hereby prohibited. A permit or conditional
permit issued for the use or operation of any
type orchard or citrus grove heater is for the
use or operation of a complete heater assem-
bly.
sampling stations shall be continuously maintained
at locations des ignated by the Air Pollution Control
Officer after consultation with the Scientific Com-
mittee. The Air Pollution Control Officer may
maintain such additional sampling stations as may
be necessary. These additional stations may be
permanent, temporary, fixed, or mobile, and may
be activated upon orders of the Air Pollution Con-
trol Officer.
RULE 152. (Revised 7-26-56) AIR SAMPLING
The Air Pollution Control Officer shall establish
procedures •whereby adequate samplings and anal-
yses of air contaminants will be taken at each of
the stations established under RULE 151.
RULE ]53. (Adopted 6-20-55) REPORTS
Tne Air Pollution Control Officer shall make daily
summaries of the readings required by Rule 152.
The summaries shall be in such form as to be un-
derstandable by the public. These summaries
shall be public records and immediately after prep-
aration shall be filed at the main office of the Air
Pollution Control District and be available to the
public, press, radio, television, and other mass
media of communication.
REGULATION VII. EMERGENCIES
(Revised 7-26-56) This emergency regulation is
designed to prevent the excessive buildup of air
contaminants and to avoid any possibility of a ca-
tastrophe caused by toxic concentrations of air con-
taminants. Past history indicates that the possi-
bility of such a catastrophe is extremely remote.
The Air Foliation Control Board deems it desir-
able to have ready an adequate plan to prevent such
an occurrence, and in case of the happening of this
unforeseen event, to provide for adequate actions
to protect the health of the citizens in the Air Pol-
lution Control District.
RULE 150. (Adopted 6-20-55) GENERAL
Notwithstanding any other provisions of these rules
and regulations, the provisions of this regulation
shall apply within the Los Angeles Basin to the
control of emissions of air contaminants during any
"alert" stage as provided herein.
RULE 151. {Amended 1-16-58) SAMPLING STA-
TIONS
The Air Pollution Control Officer shall maintain
at least six (6) permanently located atmospheric
sampling stations adequately equipped. These
RULE 154. (Revised 7-26-56) CONTINUING PRO-
GRAM OE VOLUNTARY COOPERATION
Upon the adoption of this regulation the Air Pollu-
tion Control Officer shall inform the public of ways
in which air pollution can be reduced and shall re-
quest voluntary cooperation from all persons in
all activities which contribute to air pollution. Civ-
ic groups shall be encouraged to undertake cam-
paigns of education and voluntary air pollution re-
duction in theii respective communities. Public
officials shall be urged to take promptly such steps
as may be helpful to reduce air contamination to a.
minimum within the areas of their authority. Em-
ployers shall be requested to establish car pools.
Users of automotive vehicles shall be urged to keep
motors in good condition anc to plan routes and
schedules -which will contribute minimum contami-
nation to critical areas of pollution. All industrial,
commercial and business establishments which
emit hydrocarbons or the air contaminants named
in RULE 156 should critically study their opera-
tions from the standpoint of air contamination and
should take appropriate action voluntarily to re-
duce air pollution.
RULE 154. 1. PLANS
a. (Revised 7-26-56) If the Air Pollution Control
Officer finds that any industrial, business or
commercial establishment or activity emits
-------
Rules and Regulations of the Air Pollution Control District
853
hydrocarbons or any of the contaminants named
inRule 156, he may give written notice to the
owner or operator of such industrial, business
or commercial establishment or activity to
submit to the Air Pollution Control Officer
plans for immediate shutdown or curtailment,
in the event of an air pollution emergency, all
of the sources of hydrocarbons or any of the
contaminants named in RULE 156, including
vehicles owned or operated by such person,
his agents or employees in the scope of the
business or operation of such establishment or
activity. Such plans shall include, in addition
to the other matters set forth in this rule, a
list of all such sources of hydrocarbons and
and any of the contaminants named in RULE
156, and a statement of the minimum time and
the recommended time to effect a complete
shutdown of each source in the event of an air
pollution emergency. Such notice may be
served in the manner prescribed by law for
the service of summons, or by registered or
certified mail. Each such person shall, with-
in sixty (60) days after the receipt of such no-
tice, or within such additional time as the Air
Pollution Control Officer may specify in writ-
ing, submit to the Air Pollution Control Offi-
cer the plans and information described in the
notice.
b. (Revised 7-26-56) The Air Pollution Control
Officer shall prepare appropriate plans to be
made effective and action to be taken in re-
spect to a First or Second Alert as follows:
(Revised 7-26-56) In respect to a First Alert,
the Air Pollution Control Officer shall develop
plans calling for the operation of all private-
ly owned vehicles on a pool basis as may be
arranged by persons and employers of persons
operating vehicles from home to work and in
the business of such employer.
(Revised 7-26-56) In respect to a Second Alert,
the Control Officer shall prepare a program
of action and steps to be taken under the pro-
visions of RULE 158, paragraph c. The gen-
eral nature of the plans to be made effective
upon a Second Alert shall be reported to and
subject to review and approval by the Air Pol-
lution Control Board.
(Revised 7-26-56) It shall be the objective of
such program to result in bringing about a di-
minution of air contaminants -which occasioned
the Second Alert and to prevent any increase
thereof in order to protect the health of all
persons within the area affected by the alert.
It shall also be the objective of such plans that
they may be effective to curtail the operations
of industrial, business, commercial and other
activities -within the basin, but without undue
interference with the operations of public utili-
ties or other productive, industrial, business
and other activities, the conduct of which is
is essential to the health and -welfare of the
community. It is further intended that any
said plan of action shall not jeopardize the
welfare of the public or result in irreparable
injury to any means of production or distribu-
tion or the rendering of public utility services.
(Adopted 6-20-55) The Air Pollution Control
Officer shall further, by cooperative agree-
ments or in addition to cooperative agree-
ments, prepare plans for action in respect to
industry, business, transportation, hospitals,
schools and other appropriate public and pri-
vate institutions, and the public generally, to
accomplish the purposes of the Second Alert
action as set forth in Rule 158d. The general
nature of the plans to be made effective upon
a Second Alert shall be reported to and sub-
ject to review and approval by the Air Pollu-
tion Control Board.
(Adopted 6-20-55) All plans and programs of
action to make effective the procedures pre-
scribed in Rule 158, paragraphs c. , and d. ,
shall be consistent with and designed to accom-
plish the purposes , and shall be subject to the
conditions and limitations, set forth in said
paragraphs c., and d.
(Adopted 6-20-55) The Air Pollution Control
Officer shall give, or cause to be given, wide
publicity in regard to plans for action to be ap-
plicable under Rule 158, paragraphs c. , and
d. , in order that all persons within the dis-
trict shall be able to understand and be pre-
pared to render compliance therewith in the
event of the sounding of a Second Alert.
RULE 155. (Revised7-26-56)DECLARATION
OF ALERTS
The Air Pollution Control Officer shall declare the
appropriate "alert" whenever the concentration of
any air pollution contaminant has been verified to
have reached the standards set forth in Rule 156.
RULE 155.1. (Adopted 7-26-56) NOTIFICATION
OF ALERTS
Following the declaration of the appropriate "alert",
the Air Pollution Control Officer shall communi-
cate notification of the declaration of the alert to:
a. The Los Angeles County Sheriff and the Sheriff
shall broadcast the declaration of the "alert"
by the Sheriff's teletype and radio system to:
1. All Sheriff's substations.
2. All city police departments.
3. California Highway Patrol.
-------
854
RULES AND REGULATIONS
b. Local public officials and public safety per-
sonnel, whohave responsibilities or interests
in air pollution alerts.
c. Air polluting industrial plants and processes
which require "alert" data in order to effect
pre-arranged plans designed to reduce the out-
put of air contaminants.
d. The general public.
e. All Air Pollution Control District personnel.
RULE 155.2. (Amended 3-2-67)
MUNICATION SYSTEM
RADIO COM-
The Air Pollution Control Officer shall install and
maintain, in continuous operation, a radio trans -
mitter with selective calling facilities for the pur-
pose of broadcasting the declaration of alerts and
information and instructions which may be appro-
priate to carry out the provisions of this regula-
tion.
Radio receiving equipment with decoding device
capable of receiving broadcasts from the Air Pol-
lution Control Officer of the declaration of alerts
and information and instructions thereto shall be
installed and properly maintained and operated
during all hours of plant operation by any person
who operates or uses any:
a. Petroleum refinery.
b. Bulk gasoline loading facility for tank vehicles,
tank cars, or marine vessels, from which fa-
cility 20, 000 gallons or more of gasoline are
loaded per day. For purposes of this para-
graph, "gasoline" means any petroleum dis-
tillate having a Reid vapor pressure of four
pounds or greater, and "facility" means all
gasoline loading equipment "which is both: (1)
possessed by one person, and (2) located so
that all the gasoline loading outlets for such
aggregation or combination of loading equip-
ment can be encompassed within any circle of
300 feet in diameter.
c. Asphalt saturator.
d. Asphalt paving manufacturing plant.
e. Asphalt manufacturing plant.
f. Chemical plant which:
(1) Reacts or produces any organic liquids
or gases.
(2) Produces sulfuric acid, nitric acid, phos-
phoric acid, or sulfur.
g. Paint, enamel, lacquer, or varnish manufac-
turing plant in which 10, 000 gallons or more
per month of organic solvents, diluents or
thinners, or any combination thereof are com-
bined or manufactured into paint, enamel, lac-
quer, or varnish.
h. Rubber tire manufacturing or rubber reclaim
ing plant.
i. Automobile assembly or automobile body plant.
j. Metal melting, refining or smelting plant in
whicha total of 2, 500pouncls or more of metal
are in a molten state at any one time or are
poured in any one hour.
k. Rock wool manufacturing plant.
1. Glass or frit manufacturing plant in which a
total of 4, 000 pounds or more of glass or frit
or both are in a molten state at any one time
or are poured in any one hour.
RULE 156. (Adopted 6-20-55) ALERT STAGES
FOR TOXIC AIR POLLUTANTS. (In parts per
million of air)
FIRST
ALERT
SECOND
ALERT
THIRD
ALERT
CARBON MON-
OXIDE*
NITROGEN
OXIDES*
SULFUR
OXIDES*
OZONE*
100
3
3
0. 5
200
5
5
1. 0
300
10
10
1.5
FIRST ALERT: Close approach to maximum al-
lowable concentration for the population at
large. Still safe but approaching a point where
preventive action is required.
SECOND ALERT: Air contamination level at which
a healthmenace exists in a preliminary stage.
THIRD ALERT: Air contamination level at which
a dangerous health menace exists.
*How measured: The concentrations of air con-
taminants shall be measured in accordance -with
the procedures and recommendations established
by the Scientific Committee,
RULE 157.
ACTION
(Amended 12-4-58) FIRST ALERT
This is a warning alert and shall be called declared
whenever the concentration of any contaminant has
beenverifiedtohave reached the standards for the
"first alert" set forth in Rule 156. The following
-------
Rules and Regulations of the Air Pollution Control District
855
action shall be taken upon the calling of the First
Alert:
a. A per1" on shall not burn any combustible refuse
at any location within the basin in an open fire.
b. Any person operating or maintaining any in-
dustrial, commercial or business establish-
ment other than power plants or heating plants
essential to health or safety, which establish-
ments emit hydrocarbons or any of the con-
taminants named in Rule 156, and any person
operating any private noncommercial vehicle,
shall, during the First Alert period, take the
neces sary preliminary steps to the action re-
quired should a Second Alert be declared.
c. The Air Pollution Control Officer shall, by the
use of all appropriate mass media of commu-
nication, request the public to stop all unes-
sential use of vehicles in the basin and to oper-
ate all privately owned vehicles on a pool basis ,
and shall request all employers to activate
employee car pools.
d. When, after the declaration of the First Alert
it appears to the Air Pollution Control Officer
that the concentration of any contaminants in
all or any portion of the basin is increasing in
such a manner that a Second Alert is likely to
be called, he shall take the following actions:
(1) Notify the Emergency Action Committee
and request advice on actions to be taken.
(2) Give all possible notice tothe public by all
mass media of communication that a Sec-
ond Alert may be called.
RULE 158. (Revised 7-26-56) SECOND ALERT
ACTION
This is a preliminary health hazard alert and shall
be declared when an air contaminant has been veri-
fied to have reached the standards set forth for the
"Second Alert" in Rule 156.
The following action shall be taken upon the call-
ing of the Second Alert:
a. (Adopted 6-20-55) The action set forth in Rule
157, and
b. (Adopted 6-20-55) The Emergency Action
Committee and the Air Pollution Control Board,
if not already activated, shall be called into
session and shall remain in session or recon-
vene from time to time as directed by the Air
Pollution Control Officer to study all pertinent
information relating to the emergency and to
recommend to the Air Pollution Control Offi-
cer actions to be taken from time to time as
conditions change.
c. (Revised 7-26-56) The Air Pollution Control
Officer shall make effective, upon notice as
provided in Rule 155. 1, the program of action
to be taken as previously developed pursuant
to Rule 154. 1, paragraph b. , and to carry out
the policy stated therein.
(Revised 7-26-56) Pursuant to this alert, the
Air Pollution Control Officer may impose lim-
itations as to the general operation of vehicles
as provided in Rule 154. 1, permitting limited
operation essential to accommodate industry,
business, public utility and other services as
may be necessary in the public welfare.
d. (Adopted 6-20-55) In the event the control mea-
sures made effective under paragraph c. above
prove tobe inadequate to control the increase
in the concentration of air contaminants, the
Air Pollution Control Officer, with the advice
of the Emergency Action Committee and with
the concurrence of the Air Pollution Control
Board shall take such steps as he may deem
necessary to assure adequate control of ex-
isting air contaminants and to protect the
health and safety of the public, but, if possible,
without employing such drastic remedial mea-
sures as to completely disrupt the economic
life of the community or to result in irrepar-
able injury to any form of production, manu-
facture or business.
(Revised 7-26-56) The Air Pollution Control
Officer may, with the concurrence of the Air
Pollution Control Board, order the closing of
any industrial, commercial or business estab-
lishment and stop all vehicular traffic, except
authorized emergency vehicles as defined in
the California Vehicle Code, vehicles used in
public transportation and vehicles the opera-
tion of which is necessary for the protection
of the health and welfare of the public, if in
the opinion of the Air Pollution Control Offi-
cer, the continued operation of such establish-
ment or vehicle contributes to the further con-
centration of any air contaminant, the concen-
tration of which caused the declaration of the
"alert".
(Revised 7-26-56) The Air Pollution Control
Officer, during a Second Alert, shall keep the
public suitably informed of all significant
changes in the concentration of toxic air con-
taminants.
e. (Adopted 9-28-61) In the event that the Air
Pollution Control Officer determines that the
public health and safety is in danger, the
-------
856
RULES AND REGULATIONS
Emergency Action Committee and the Air Pol-
lution Control Board may take any action au-
thorized by this rule with less than a quorum
present. A majority vote of the members
present is required for any such action.
RULE 159. (Revised 7-26-56) THIRD ALERT
This is a dangerous health hazard alert and shall
be declared when an air contaminant has been veri-
fied to have reached the standards set forth for the
"Third Alert" in Rule 156.
The following auction shall be taken upon the call-
ing of the Third Alert:
a. (Adopted 6-20-55) The actions set forth in
Rules 157 and 158, and
b. (Adopted 6-20-55) If it appears that the steps
taken by the Air Pollution Control Officer will
be inadequate to cope with the emergency, the
Air Pollution Control Board shall request the
Governor to declare that a state of emergency
exists and to take appropriate actions as set
forth in the California Disaster Act.
RULE 160. (Revised 7-26-56) END OF ALERT
The Air Pollution Control Officer shall declare the
termination of the appropriate alert whenever the
concentration of an air contaminant which caused
the declaration of such alert has been verified to
have fallen below the standards set forth in Rule
156 for the calling of such alert and the available
scientific and meteorological data indicates that
the concentration of such air contaminant will not
immediately increase again so as to reach the stan-
dards set forth for such alert in Rule 156. The Air
Pollution Control Officer shall immediately com-
municate the declaration of the termination of the
alert in the mariner provided in Rule 155. 1 for the
declaration of alerts. The Sheriff shall broadcast
the termination of the alert in the same manner as
provided in Rule 155. 1 for the declaration of alerts.
RULE'161. (Revised 7-26-56) ENFORCEMENT
When an "alert" has been called the Air Pollution
Control Officer, the Sheriff, their deputies, and
all other peace officers within the Basin shall en-
force the appropriate provisions of this regulation
and all orders of the Air Pollution Control Board
or the Air Pollution Control Officer made pursuant
to this regulation against any person who, having
knowledge of the declaration of an alert, refuses
to comply with the rules set forth in this regula-
tion or any order of the Air Pollution Control Board
or the Air Pollution Control Officer made pursuant
to this regulation.
RULE 163. (Revised 7-26-56) SCIENTIFIC COM-
MITTEE
A Scientific Committee shall be appointed by the
Air Pollution Control Board. Members shall be
licensee physicians , medical scientists , biologists,
chemists, engineers, or meteorologists, each of
whom has had experience in air pollution control
work, or other experts with scientific training.
(Adopted 6-20-55) The Air Pollution Control Offi-
cer and the County Counsel shall be ex-officio
members of the Scientific Committee.
(Adopted 6-20-55) The term of appointment of all
members except the ex-officio members shall be
two (2) years. The Scientific Committee shall act
through a majority. There shall be at least fifteen
(15) members on the Committee.
The Scientific Committee shall have the following
duties:
a. (Revised 7-26-56) Study and recommend. The
Scientific Committee shall study and make
recommendations to the Air Pollution Control
Board of the most suitable methods for mea-
surement of air contaminants and on any chang-
es recommended for the concentrations set
forthinRULE 156. The Air Pollution Control
Board may adopt such recommended changes
for the concentrations of toxic air contami-
nants for each alert staged by amendment to
RULE 156.
b. (Adopted 6-20-55) Consult. The Scientific
Committee shall serve in a consultant advisory
capacity to the Air Pollution Control Officer
concerning any air pollution health problem
which may arise. The Scientific Committee
shall also advise the Air Pollution Board on
any recommended changes in this emergency
regulation which will provide greater protec-
tion of the health and welfare of all persons
within the Air Pollution Control District.
RULE 164. (Revised 9-28-61) EMERGENCY AC-
TION COMMITTEE
An Emergency Action Committee shall be appointed
by the Air Pollution Control Board. The com-
mittee shall be composed of ten (10) appointed
members and of these members two shall be ex-
perts with scientific training or knowledge in air
pollution matters, two shall be licensed physicians,
two shall be representatives of industry, two shall
be representatives of law enforcement, and two
shall be members of the public at large.
The County Health Officer, the Sheriff, and the
County Counsel shall be ex-officio members of the
-------
Rules and Regulations of the Air Pollution Control District 857
Committee. In the absence of an ex-officio mem- and to advise the Air Pollution Control Officer as
ber, his deputy may act for him. to the appropriate action to be taken when the con-
centration of any of the contaminants set forth in
The term of appointment of appointed members Rule 156 has been verified to be approaching the
shall be two years. standards set forth in Rule 156 for a Second Alert.
The duties of the Emergency Action Committee
shall be to meet with the Air Pollution Control The Committee shall meet when called into session
Officer when called into session, to evaluate data, and not less than every three months.
-------
APPENDIX 8
-------
APPENDIX B: ODOR-TESTING TECHNIQUES
Modern technology has not yet produced a precise
method of analyzing odor concentration or odor
quality. In some instances, it has been possible
to measure concentrations of specific odorous
compounds through chemical or spectroscopic
analyses. The odors of concern to air pollution
engineers, however, are usually mixtures of sev-
eral odorous compounds (McCord and Witheridge,
1949). Identification and measurement of each
constituent is usually a tedious, if not impossible,
task. For this reason, it is more practical to
measure the aggregate odor concentration or
detectability of a gas stream in terms of odor
units. An odor unit is defined as the quantity of
any odor or mixture of odors that, when dispersed
in one cubic foot of odor-free air, produces a
median threshold odor detection response. The
overall odor measurement techniques to determine
odor units require that human olfactory organs
serve as analytical tools. Inasmuch as olfactory
responses are somewhat transitory, particular
care must be taken to eliminate extraneous odors
and false olfactory responses.
A dilution method has been developed (Fox and
Gex, 1957) that uses the human nose to measure
odor concentration. It generally follows the
American Society for Testing Materials Method
D1391-57 (Standard Method for Measurement of
Odor in Atmospheres [Dilution Method]) and in-
corporates some refinements. The method con-
sists, in essence, of successively diluting a gas
sample with odor-free air until a threshold dilu-
tion is reached, that is, at further dilution no
odor is detectable by the human nose. To mini-
mize the effect of variations in olfactory sys-
tems, a panel of several persons is used. The
odor concentration is determined by plotting di-
lution response data on log-probability coordi-
nates.
This dilution method serves principally to mea-
sure odor concentration. It is a valuable tool
with which to evaluate the performance of odor
control equipment, and the quantitative odor
nuisance potential of a source. The quality or
objectionability of an odor cannot be evaluated
with the same assurance. While the dilution
method can be used to measure objectionability
thresholds, results are not as reproducible as
detectability measurements are. This is due
principally to the subjective nature of human
olfactory responses. The average subject can
report the presence or absence of an odor with
more certainty than he can determine objection-
ability.
Odor testing is a comparatively recent develop-
ment. Certain modifications (Mills et al. , 1963)
of the American Society for Testing Materials
static test procedure were developed to accom-
modate the method to field problems and to ac-
celerate the testing procedure, at the same time
maintaining or improving the reproducibility and
reliability of results.
For employment of this method for odor evalua-
tion, a selected group of individuals must be used
as odor panel members, and an air-conditioned,
odor-free room must be used for the test.
THE ODOR PANEL
The ASTM procedure describes a suitable meth-
od of screening and selecting members of the
odor panel. The selectees should be persons who
are neither the most sensitive to odors nor the
most insensitive of those screened. The choice
of panel members should be limited to those with
the most generally reliable olfactory perception.
Consistent and reproducible results have been
found to be obtained with a panel consisting of at
least eight persons. Although a panel of six per-
sons is adequate at times, eight is preferred,
because the probabilities of inconclusive results
(with the resultant necessity of rerunning the test)
are thereby reduced.
If possible, the panel members should be allowed
to relax in the odor-free room for 10 to 15 min-
utes before the test. This ensures that their
olfactory senses are not fatigued or dulled by ex-
traneous odors. Test periods should be limited
to 30 minutes or less. If testing is required over
a longer period, adequate rest periods should be
scheduled to preclude fatigue.
THE ODOR EVALUATION ROOM
A typical plan for an odor evaluation room is
shown in Figure Bl. Essential features are:
(1) Separation of the work area from the evalua-
tion area, (2) provision for relatively odor-free
air at room temperature with moderate humidity
by use of an air-conditioning unit, and (3) an
activated-carbon adsorption unit to provide and
circulate odor-free air to the evaluation area.
An odor evaluation room should be designed to
minimize the possibility of extraneous odors in
the vicinity of the panelists. It should be devoid
of fabrics, such as carpeting, draperies, or up-
holstery, that might hold odorous materials. The
room should be so located in the building that
861
-------
862
ODOR-TESTING TECHNIQUES
there is no introduction of odors into the air con-
ditioner inlet or through doors, cracks, and so
forth. Air circulation should be such that the
activated-carbon unit discharges air near the
panelists. All air from the work area should be
filtered before it comes into contact with panelists.
SAMPLING TECHNIQUES
Representative sampling points are chosen ac-
cording to standard air-sampling techniques. In
most instances, 250-milliliter grab samples are
sufficient. These are collected in gas-sampling
tubes such as those shown in Figure B2. Possible
sources of error are foreign odors from the sam-
pling train, improperly cleaned glassware, and
condensation in the sample tube.
The use of rubber or plastic tubing and other
heat-sensitive materials in the sample probe
should be avoided, particularly if the gas stream
is at an elevated temperature. The apparatus of
Figure B2 is recommended wherein all tubing
and joints upstream of the sampling tube are con-
structed of glass. The rubber bulb evacuator is
on the downstream side of the tube and does not
contaminate the sample.
The problems of condensation and adsorption of
odorous material on the inner walls of the sam-
pling apparatus are the most difficult to over-
come or even to evaluate. Odor adsorption can
be minimized by flushing the sampling equipment
with enough of the gas stream to allow tempera-
ture and humidity to reach equilibrium. The area
of ground glass in contact with the sample should be
held to a minimum.
Condensation in the tube can introduce a large
error when the moisture content is much more
than 20 percent by volume. When the gas stream
bears a high moisture content, a second sampling
technique has been devised in which the sample
is diluted in the sample tube with dry, odor-free
air. Dilution air is drawn through a cartridge
charged with activated carbon and a suitable des-
sicant. This sampling technique provides a di-
lution of 10:1 or greater in the tube. Equipment
used for dilution sampling is diagrammed in Fig-
ure B3. The 1 -millimeter-outside-diameter
capillary tube used as a probe is inserted through
a new, size 000, cork stopper with the aid of an
18-gage hypodermic needle as a sleeve. The
sample is obtained by placing the free end of the
capillary into the gas stream and withdrawing
the required 5 to 10 milliliters of air from the
sample tube with the 10-milliliter syringe. The
volume withdrawn is replaced by an equal volume,
which enters through the capillary tube. The
small diameter of the capillary minimizes diffu-
sion across the tube.
In both techniques, the stopcock nearest the
squeeze bulb is closed first. When equilibrium
conditions are established, the other stopcock
is closed and the probe removed from the gas
stream.
I
AIR
EVALUATION AREA
ACTIVATED CARBON UNIT
6OO TO 12OO CFM
QQQQQQQQQ
CHAIRS
O- ^> -O- •O O •£>
-
O
ODOR-FREE AIR PLENUM
f=l
I
LJ
ENTRANCE
(CLOSED)
WORK AREA
AIR
INTAKE
BENCH
Figure B1. Odor-free room.
-------
Evaluation of Odor Samples
863
SAMPLE TUBE (250ml.)
Figure B2. Odor sampling equipment for dry gases.
> MEDICAL SWINGE
-------
864
ODOR-TESTING TECHNIQUES
a positive or negative detection of odors on a
tally sheet together with the number of the sample.
Each panelist purges his syringe with air between
samples.
Some compounds such as aldehydes deaden the
sense of smell and cause erratic results, that
is, the dilution response data do not plot to a
straight line on log-probability coordinates.
While there is no entirely satisfactory method
of overcoming this effect, it can be at least
partially offset by allowing more time between
samples for panelists' olfactory systems to re-
cover.
DETERMINATION OF ODOR CONCENTRATION
The odor responses of the panel are quantified by
calculating the percent of the panel members de-
tecting odors at each dilution, as shown in Table
Bl. The ratio of the diluted volume to the orig-
inal sample is termed the dilution factor. Odor
responses are plotted against dilution factors to
determine odor concentration.
Dilution response data follow a cumulative normal
distribution curve. If plotted on rectilinear co-
ordinates, these data produce an s-shaped curve.
The points at the extremes of the curve would
represent panelists who are the most and the
least sensitive to the particular odors. The area
in the middle of the curve would represent average
olfactory responses.
When dilution response data are plotted on log-
arithmic-probability coordinates, they tend to
follow a straight line. This phenomenon is shown
in Figure B5, where the test data of Table Bl are
plotted. The subject gases evaluated were repli-
cate samples of discharge gases from a fish meal
drier. The data plot to a reasonably straight line.
Maximum deviation from a straight line is prin-
cipally a function of the number of panelists.
ODOR RESPONSE CHART
- —
^••x
__
\
~~K
— •.
'>»*
^ **
3,000 o u /scf
*
^
>
OFF-GASES FROM FISH MEAL DRIER
EVALUATED BY THE STATIC METHOD
13 r=z]r— \ 1 -: 1 | -
.
SAMPL
SAMPL
No
No
•
*
0 90 95 9
"— - —
~~^
— -,
•
*
— ~,
.
1,400 ou/stl
L
—
^
OFF-GASES FROM FISH MEAL DRIER
EVALUATED BY THE DYNAMIC METHOD
.
SAMPLE Mo J •
SAMPLE No ( *
5 10 20 30 40 50 60 70
PERCENT OF PANEL REPORTING POSITIVE RESPONSES
Figure B5. Plot of dilution response data.
The point at which the plotted line crosses the
50 percent panel response line is the threshold
concentration. The dilution factor at the thres-
hold is the odor concentration, usually stated in
terms of odor units per scf. The total rate of
odor emission in odor units per minute may
then be calculated by multiplying the concen-
tration by the total volume of the effluent.
Table Bl. DATA FROM A TYPICAL DILUTION TEST
Sample
No.
1
Z
2
Dilution
designation
A
B
C
D
A
B
C
Dilution
factor3-
1, 000
2, 500
10, 000
5, 000
2, 500
5, 000
10, 000
No. of
panel
members
8
8
8
8
8
8
8
No. of
panel members
detecting odor
6
4
3
2
5
3
1
% of
panel members
detecting odor
75
50
38
25
63
38
13
aThe dilution factor is the volume of the diluted sample evaluated by the
panel members, divided by the volume of the original undiluted sample
contained therein.
Zero and 100 percent responses are considered indeterminate.
-------
APPENDIX C
GPO 806-614—29
-------
APPENDIX C: HYPOTHETICAL AVAILABLE HEATS FROM NATURAL GAS
Burners for combustion devices such as after-
burners frequently use the oxygen present in the
contaminated effluent stream. An example would
be a natural gas-fired afterburner that takes in
60 percent of its combustion air from the atmo-
sphere, and 40 percent from an air containing
contaminated effluent stream.
One step in checking afterburner design is the cal-
culation of the natural gas flow rate required to
raise an effluent stream to a given temperature.
A calculation such as this normally makes use of
the available heat from natural gas. Available
heat is the amount of heat remaining after the
products of combustion from a cubic foot of natural
gas are raised to the afterburner temperature.
Available heat from natural gas is shown in Table
D7.
If the afterburner gas burner takes a portion of
the combustion air from the effluent stream,
then the calculation of the gas flow rate becomes
a trial-and-error procedure. By the method of
hypothetical available heats given here, the trial-
and-error solution is eliminated.
Let the heat content of an effluent stream, at the
desired final temperature, be H Btu/lb. Since
10. 36 cubic feet of air is required for combus-
tion of 1 cubic foot of natural gas, the weight of
air taken from the effluent would be
W = (10.36)(l-X)(p)
(Cl)
The heat contents of this secondary combustion
air would be
Q = WH = (10.36)(l-X)(p)(H)
where
(C2)
W = weight of combustion air from the ef-
fluent per cubic foot of natural gas,
Ib/ft natural gas
H = heat content of the effluent at the re-
quired temperature, Btu/lb
X = fraction of theoretical combustion air,
furnished as primary air through burn-
er
The natural gas used in illustrating this calcula-
tion procedure requires 10.36 cubic feet of air
for theoretical combustion of 1 cubic foot of gas
(Los Angeles area natural gas). Products of
complete combustion evolved from this process
are carbon dioxide, water, and nitrogen. If
the combustion of 1 cubic foot of natural gas is
thought of as taking place at 60 °F, then a portion
of the heat released by combustion must be used
to raise the products of combustion from 60 °F to
the temperature of the device. The remaining
heat is called available heat. This quantity repre-
sents the heat from natural gas that can be used
to do useful work in the combustion device, such
as heating an effluent stream in an afterburner.
Consider a gas-fired aite.rburner adjusted to pro-
vide a fraction, X, of theoretical air through the
burner. If the contaminated effluent contains air,
then the remaining air for combustion, 1-X, is
taken from the effluent stream. This means that
a smaller quantity of effluent has to be heated by
the natural gas, since a portion of the effluent is
involved in the combustion reaction. Thus, a
burner taking combustion air from an effluent
stream can be fired to raise the temperature of
the effluent at a natural gas input lower than that
of a burner firing with all combustion air taken
from the atmosphere.
p - density of air at 60 °F
= 0.0764 Ib/ft3.
Since Q Btu per cubic foot of natural gas is not
required to heat the effluent, it can be added to
the available heat, A, at the afterburner tem-
perature, or
= A
Q
(C3)
•where
A = hypothetical available heat, Btu/ft
natural gas
Q = heat content of secondary combustion air
from equation C2.
Equation C3 is given in terms of temperature in the
following equations:
Temperature, °F
600
700
800
900
1,000
1, 100
Hypothetical available heat,
Btu/ft natural gas
871 + 104 (1-X)
846 + 124 (1-X)
S^l + 144 (1-X)
798 + 167 (1-X)
773 + 185 (1-X)
747 + 206 (1-X)
867
-------
HYPOTHETICAL AVAILABLE HEATS FROM NATURAL GAS
1,
1,
1,
1,
1,
1,
1,
200
300
400
500
600
700
800
721 H
693 J
669 H
643 H
615 4
590 -}
562 H
- 227
h 249
f- 270
- 292
- 314
- 336
- 358
(1-X)
(1-X)
(1-X)
(1-X)
(1-X)
(1-X)
(1-X)
3. The gases in the afterburner will consist of:
a. Products of combustion from
1,230 cfh
natural gas with theoretical air - 1, 230
x 11. 45 scfh,
b. the portion of the effluent not
<"* r~»i-vi T"\iicf"i<"»'n a i >* — j^f-fln^-nt- TT-I-I!
used for
X = fraction of theoretical air furnished
as the burner's primary air.
Hypothetical available heats are given in Table
Cl for varying temperatures and percentages
of primary air.
The use of this concept is illustrated in the follow-
ing examples.
Example Cl:
An afterburner is used to heat an effluent stream
to 1,200°F by using 1 x 10& Btu/hr. The burner
is installed and adjusted so that 60% of the theoret-
ical combustion air is furnished through the burner,
and the remainder is taken from the effluent. De-
termine the required natural gas rate.
1. The percent primary air is 60%, the required
temperature is 1,200°F, the hypothetical
available heat from Table Cl is 812 Btu/ft3
of gas.
2. Burner flow rate = 106/812 = 1,230 cfh gas.
(1, 230)(10.36)(1-X).
Example C2:
An afterburner is used to heat an effluent stream
to 1,200°F by using 1 x 10& Btu/hr. The burner
is installed and adjusted so that all the combustion
air is taken from the effluent stream. Determine
the natural gas rate.
This is equivalent to the burner's operating at 0%
primary air.
1. At 1,2 00° F the hypothetical available heat is
948 Btu/ft3 for 0% primary air.
2. Burner flow rate = 106/948 = 1,058 cfh.
3. Gases in afterburner will consist of:
a. Combustion products from 1, 058 cfh nat-
ural gas with theoretical air = 1, 05& x
11.45,
b. the portion of the effluent not used for
secondary combustion air = effluent
volume - (1,058)(10. 36)(1-X).
Table Cl. HYPOTHETICAL AVAILABLE HEATS
Hypothetical available heats, Btu/ft gas
l emp,
°F
600
700
800
900
1, 000
1, 100
1,200
1, 300
1, 400
1, 500
1, 600
1,700
1, 800
% primary air through the burner
0
975
970
965
965
958
953
948
942
939
935
929
926
920
10
965
958
950
948
939
933
926
917
912
906
897
892
885
20
954
945
936
931
921
912
903
892
885
976
866
859
949
30
944
933
922
915
902
391
380
867
858
847
834
825
813
40
933
921
907
898
884
871
858
842
831
818
803
791
777
50
923
908
893
881
865
850
835
818
804
789
772
758
741
60
913
896
878
365
847
830
812
793
777
760
740
724
706
70
902
883
864
848
328
809
789
763
750
730
709
691
670
80
892
971
850
831
310
788
767
743
723
701
677
657
634
90
881
359
835
314
791
768
744
718
696
672
646
623
598
-------
APPENDIX 0
-------
-------
APPENDIX D: MISCELLANEOUS DATA
Table Dl. PROPERTIES OF AIR
Temp,
°F
0
20
40
60
80
100
120
140
160
180
200
250
300
350
400
450
500
600
700
800
900
1, 000
1,200
1, 400
1, 600
1, 800
2, 000
Specific heat
at constant
pressure (Cp),
Btu/lb-°F
0.240
0.240
0. 240
0. 240
0.240
0. 240
0. 240
0. 240
0. 240
0.240
0.240
0. 241
0.241
0.241
0. 241
0.242
0.242
0.242
0. 243
0. 244
0.245
0.246
0. 248
0.251
0.254
0.257
0.260
Absolute
viscosity (|a),
Ib/hr-ft
0.040
0.041
0.042
0.043
0.045
0. 047
0. 047
0. 048
0.050
0.051
0.052
0. 055
0.058
0.060
0. 063
0. 065
0.067
0. 072
0.076
0.080
0.085
0. 089
0.097
0. 105
0. 112
0. 120
0. 127
Thermal
conductivity
(k),
Btu/hr-ft-T
0. 0124
0.0128
0. 0132
0. 0136
0. 0140
0.0145
0. 0149
0.0153
0.0158
0. 0162
0. 0166
0. 0174
0. 0182
0.0191
0. 0200
0. 0207
0.0214
0. 0229
0.0243
0.0257
0.0270
0. 0283
0. 0308
0.0328
0. 0346
0. 0360
0.0370
Prandtl No.
(Qi/k),
(dimensionless)
0. 77
0. 77
0. 77
0.76
0. 77
0.76
0. 76
0. 76
0. 76
0.76
0. 76
0. 76
0.76
0. 76
0. 76
0. 76
0. 76
0. 76
0.76
0. 76
0. 77
0.77
0. 78
0. 80
0.82
0.85
0.83
Density
(P),
Ib/ft3a
0. 0863
0. 0827
0. 0794
0. 0763
0. 0734
0. 0708
0. 0684
0. 0662
0. 0639
0. 0619
0. 0601
0. 0558
0. 0521
0. 0489
0. 0460
0. 0435
0. 0412
0. 0373
0. 0341
0. 0314
0. 0295
0. 0275
0. 0238
0. 0212
0. 0192
0. 0175
0. 0161
p taken at pressure of 29.92 inches of mercury.
871
-------
872
MISCELLANEOUS DATA
Table D2. THRESHOLD LIMIT VALUES (Copyright, 1966, American Conference of
Governmental Industrial Hygienists)*
Recommended Values
Substance
Acetaldehyde
Acetic acid
Acetic anhydride
Acetone J
Acetonitrile
Acetylene dichloride, see 1,
2 Dichloroethylene
Acetylene tetrabromide
Acrolein
Acrylonitrile -skin
Aldrin-skin
Allyl alcohol -skin
Allyl chloride
c Allyl glycidyl ether (AGE)
Allyl propyl disulfide
2 Aminoethanol, see
Ethanolamine
Ammonia
Ammonium sulfamate (Am-
mate)
n-Amyl acetate
Aniline -skin
"Anisidine (o, p-isomers)-
skin
Antimony and compounds
(as Sb)
ANTU (alpha naphthyl thio-
urea)
Arsenic and compounds
(as As)
Arsine
Barium (soluble compounds)
cBenzene (benzol) -skin
Benzidine-skin
p-Benzoquinone, see Quinone
Benzoyl peroxide
Benzyl chloride
Beryllium
eBiphenyl, see Diphenyl
Boron oxide
cBoron trifluoride
Bromine
Butadiene (1, 3 -butadiene) 1
Butanethiol, see Butyl
mercaptan
2-Butanone
2-Butoxy ethanol (Butyl
Cellos olve ) - skin
eButyl acetate (n-butyl acetate)
Butyl alcohol
tert. Butyl alcohol
°Butylamine-skin
ctert. Butyl chromate (as
CrO3)-skin
n-Butyl glycidyl ether (BGE)
ppm
200
10
5
,000
40
1
0,1
20
—
2
1
10
2
50
__
100
5
--
--
--
--
0.05
--
25
—
--
1
--
—
1
0. 1
,000
200
50
--
100
100
5
__
50
rng/m3 b
360
25
20
2,400
70
14
0.25
45
0.25
5
3
45
12
35
15
525
19
0.5
0.5
0.3
0.5
0.2
0.5
80
A1
5
5
0. 002
15
3
0.7
2,200
590
240
--
300
300
15
0. 1
270
Substance
Butyl mercaptan
p-tert. Butyltoluene
Cadmium oxide fume
Calcium arsenate
Calcium oxide
Camphor
dCarbaryl (Sevin) (R)
Carbon dioxide
Carbon disulfide-skin
e Carbon monoxide
Carbon tetracliloride-skdm
C hlo r dane - s kin
Chlorinated camphene, -skin
Chlorinated diphenyl oxide
Chlorine
Chlorine dioxide
c Chlorine trifluoride
cChloroacetaldehyde
Chlorobenzene (mono-
chlorobenzene)
Chlorobromome thane
2-Chloro-l, 3 butadiene,
see Chloroprene
Chlorodiphenyl (42% chlo-
rine)-skin
Chlorodiphenyl (54% chlo-
rine)-skin
1, Chloro, 2,3 epoxypropane,
see Epichlorhydrin
2, Chloroethanol, see
Ethylene chlorohydrin
Chloroethylene, see Vinyl
chloride
c Chloroform (trichloro-
methane)
1 -Chloro- 1 -nitropr opane
Chloropicrin
Chloroprene (2-chloro-l, 3-
butadiene)-skin
Chromic acid and chromates
(as CrO3)
e Cobalt
Copper fume
Dusts and mists
"Cotton dust (raw)
Crag (R) herbicide
Cresol (all isomers)-skin
Cyanide (as CN)-skin
Cyclohexane
Cyclohexanol
Cyclohexanone
eCyclohexene
d C y cl ope nta die ne
2, 4-D
DDT-skin
ppm
10
10
--
--
--
__
...
5,000
20
_..
10
...
...
__
_..
0.1
0. 1
1
75
200
—
--
50
20
0. 1
25
_..
...
...
__
...
_..
5
__.
--
50
50
__
75
...
...
mg/m^ b
35
60
0. 1
1
5
2
5
9,000
60
--
65
0.5
0.5
0.5
--
0.3
0.4
3
350
1,050
1
0.5
240
100
0.7
90
0. 1
--
0. 1
1.0
1
15
22
5
--
200
200
--
200
10
1
*See Table D2 Footnotes, pages
878.
-------
Threshold Limit Values
873
Substance
ppm mg/m3 b
Substance
ppm
mg/i
,3 b
DDVP-skin -- 1
Decaborane-skin 0.05 0.3
Derneton (R)-skin -- 0. 1
Diacetone alcohol (4-hy-
droxy-4-methyl-2-penta-
none) 50 240
1,2 Diaminoethane, see
Ethylenediamine Diborane
el, 2-Dibromoethane (ethylene
dibromide)-skin
Co-Dichlorobenzene 50 300
p-Dichlorobenzene 75 450
Dichlorodifluoromethane 1,000 4,950
dl, 3-Dichloro-5-dimethyl
hydantoin -- 0.2
1, 1,-Dichloroethane 100 400
1, Z-Dichloroethane 50 200
1,2-Dichloroethylene 200 790
cDichloroethyl ether-skin 15 90
Dichloromethane, see
Methylenechloride
Dichloromonofluoromethane 1,000 4.200
CI, 1-Dichloro-l-nitroethane 10 60
1,2-Dichloropropane, see
Propylenedichloride
Dichlorotetrafluoroethane 1,000 7,000
Dieldrin-skin -- 0.25
Diethylamine 25 75
Diethylether, see Ethyl ether
Difluorodibromomethane 100 860
cDiglycidyl ether (DGE) 0.5 2.8
Dihydroxybenzene, see
Hyd r oquinone
Diisobutyl ketone 50 290
Dimethoxymethane, see
Methylal
Dimethyl acetamide-skin 10 35
dDimethylamine 10 18
Dimethylaminobenzene, see
Xylidene
Dimethylaniline (N-di-
methylaniline)-skin 5 25
Dimethylbenzene, see Xylene
dDimethyl 1,2-dibro-2, 2-
dichloroethyl phosphate,
(Dibrom) (R) -- 3
dDimethylformanide-skin 10 30
2, 6 Dimethylheptanone, see
Diisobutyl ketone
1, 1-Dimethylhydrazine-skin 0.5 1 .
Dimethylsulfate-skin 1 5
Dinitrobenzene (all isomers)-
skin -- 1
Dinitro-o-cresol-skin -- 0.2
Dinitrotoluene-skin -- 1.5
Dioxane (Diethylene dioxide) -
skin 100 360
Dipropylene glycol methyl
ether-skin 100 600
Di-sec, octyl phthalate (Di-
2 -ethylhexylphthalate -- 5
Endrin-skin -- 0. 1
Epichlorhydrin-skin 5 19
EPN-skin -- 0.5
1, 2-Epoxypropane, see
Propylene oxide
2, 3-Epoxy-l-propanol see
Glycidol
Ethanethiol, see Ethyl -
mercaptan
Ethanolamine 3 6
2 Ethoxyethanol-skin 200 740
2 Ethoxyethylacetate (Cello-
solve acetate)-skin 100 540
Ethyl acetate 400 1,400
Ethyl acrylate-skin 25 100
Ethyl alcohol (ethanol) 1,000 1,900
eEthylamine
'eEthylbenzene — —
Ethyl bromide 200 890
Ethyl chloride 1,000 2,600
Ethyl ether 400 1, 200
Ethyl formate 100 300
'eEthyl mercaptan
Ethyl silicate 100 850
Ethylene chlorohydrin-skin 5 16
Ethylenediamine 10 25
Ethylene dibromide, see
1, 2 -Dibromoethane
Ethylene dichloride, see
1, 2-Dichloroethane
cEthylene glycol dinitrate-
skin 0.2 1.2
Ethylene glycol monomethyl
ether acetate, see Methyl
cellos olve acetate
eEthylene imine-skin
Ethylene oxide 50 90
Ethylidine chloride, see
1, 1-Dichloroethane
Ferbam -- 15
Ferrovanadium dust -- 1
Fluoride (as F) -- 2.5
Fluorine 0.1 0.2
Fluorotrichloromethane 1,000 5,600
"-Formaldehyde 5 6
Freon 11, see Fluorotri-
chloromethane
Freon 12, see Dichlorodi-
fluoromethane
Freon 13B1, see Trifluoro-
monobrome thane
Freon 21, see Dichloromono-
fluoromethane
Freon 112, see 1, 1, 2, 2-
Tetrachloro-1, 2 difluoro-
e thane
-------
874
MISCELLANEOUS DATA
Substance
Freon 113, see 1, 1, 2-Tri-
chloro, 1,2, 2-trifluoro-
ethane
Freon 114, see Dichloro-
tetrafluoroethane
Furfural -skin
Furfuryl alcohol
Gasoline
Glycidol (2, 3-Epoxy-l -pro-
panol)
Glycol monoethyl ether, see
2 -Ethoxyethanol
eGuthion, see Azinphosmethyl
Hafnium
Heptachlor-skin
Heptane (n-heptane)
Hexachloroethane-skin
Hexane (n-hexane)
2-Hexanone
Hexone
sec-Hexyl acetate
Hydrazine-skin
Hydrogen bromide
cHydrogen chloride
Hydrogen cyanide -skin
Hydrogen fluoride
Hydrogen peroxide, 90%
Hydrogen selenide
i
Hydrogen sulfide
Hydr oquinone
clodine
elron oxide fume
Isoamyl alcohol
Isophorone
Isopropyl alcohol
Is opr opylamine
Isopropylether
Isopropyl glycidyl ether (IGE)
Ketene
Lead
Lead arsenate
Lindane-skin
Lithium hydride
L, P. G. (Liquified petroleum
gas) 1
Magnesium oxide fume
Mala thion- skin
cManganese
Mercury- skin
Mercury (organic compounds)-
skin
Mesityl oxide
Methanethiol, see Methyl
mercaptan
Methoxychlor
2-Methoxyethanol, see
Methyl cellosolve
Methyl acetate
Methyl acetylene (propyne) 1
ppm
5
50
--
50
__
--
500
1
500
100
100
50
1
3
5
10
3
1
0.05
10
_-
0. 1
100
25
400
5
500
50
0. 5
__
__
__
,000
--
_ «.
_ _
--
_ _
25
--
200
,000
mg/m
20
200
A6
150
0. 5
0.5
2,000
10
1, 800
410
410
295
1. 3
10
7
11
9
1.4
0.2
15
">
1
360
140
980
12
2, 100
240
0.9
0. 2
0. 15
0. 5
0. 025
1,800
15
15
5
0. 1
0.01
100
15
610
1, 650
a
Substance PPm
Methyl acetylene -propadiene
mixture (MAPP) 1,
Methyl acrylate-skin
Methylal (dimethoxyme thane) 1,
Methyl alcohol (methanol)
Methyl amyl alcohol, see
Methyl isobutyl carbinol
°Methyl bromide -skin
Methyl butyl ketone, see
2-Hexanone
Methyl cellosolve-skin
Methyl cellosolve acetate -
skin
Methyl chloride
Methyl chloroform
Methylcyclohexane
Methylcyclohexanol
o-Methylcyclohexanone-skiri
Methyl ethyl ketone (MEK),
see 2-Butanone
Methyl formate
Methyl isobutyl carbinol- skin
Methyl isobutyl ketone, see
Hexone
j
' Methyl mercaptan
Methyl methacrylate
Methyl propyl ketone, see
2 -Pentanone
"-ffMethyl styrene
cMethylene bisphenyl iso-
cyanate (MDI)
Methylene chloride (dichlo-
romethane)
Molybdenum (soluble corn-
pounds)
(insoluble compounds)
Monomethyl aniline -skin
Morpholine-skin
Naphtha (coal tar)
Naphtha (petroleum.)
dNaphthalene
/3-Naphthylamine
Nickel carbonyl
Nickel, metal and soluble
compounds
Nicotine -skin
j
Nitric acid
p -Nit roaniline- skin
Nitrobenzene -skin
"p - Nit rochloro -benzene -skin
Nitroethane
cNitrogen dioxide
"Nitrogen trifluoride
cNitroglycerin- -f EGDN-skin
Nitrome thane
1 -Nitropropane
2-Nitropropane
N-Nitrosodimethyl-amine
(Di -methyl -nitr os oamine)-
skin
mg/m
000 1,
10
000 3,
200
20
25
25
100
350 1,
500 2,
100
100
100
25
10
100
100
0.02
500 1,
--
--
2
20
200
500 2,
10
--
0.001
--
—
2
1
1
--
100
5
10
0.2
100
25
25
—
b
800
35
100
260
80
80
120
210
900
000
470
460
250
100
20
410
480
0. 2
740
5
15
9
70
800
000
50
A2
0. 007
1
0.5
5
6
5
1
310
9
29
2
250
90
90
0
A6
-------
Threshold Limit Values
875
Substance
ppma mg/m "
Substance
ppm°
mg/m
3 b
0. 05
0.1
0.005
0. 1
0.3
0.5
2
0. 01
5
5
Nitrotoluene-skin 5 30
Nitrotrichloromethane, see
Chloropicrin
Octane 500 2, 350
Oil mist (mineral) -- 5
Osmium tetroxide -- 0. 002
Oxygen difluoride
Ozone
Parathion-skin
Pentaborane
Pentachloronaphthalene-skin —
Pentachlorophenol-skin
Pentane 1,000
2-Pentanone 200
Perchloroethylene 100
Perchloromethyl mercaptan 0. 1
Perchloryl fluoride 3
Phenol-skin 5
"p-Phenylene diamine-skin
Phenylethylene, see Styrene
Phenyl glycidyl ether (PGE) 50
Phenylhydrazine-skin 5
Phosdrin (Mevinphos) (R)-
skin
^Phosgene (carbonyl chloride)
Phosphine
Phosphoric acid
Phosphorus (yellow)
Phosphorus pentachloride
Phosphorus pentasulfide
Phosphorus trichloride
dPhthalic anhydride
Picric acid-skin
Platinum (Soluble salts)
Polytetrafluoroethylene de-
composition products -- A
^Propane 1,000 1,800
Propyne, see Methyl-
acetylene
jSPropiolactone — A
n-Propyl acetate 200 840
n-Propyl nitrate 25 110
Propylene dichloride 75 350
ePropylene imine-skin -- —
Propylene oxide 100 240
Pyrethrum -- 5
Pyridine 5 15
Quinone 0.1 0.4
Rotenone (commercial) -- 5
"Selenium compounds (as Se) -- 0.2
Silver, metal and soluble
compounds — 0.01
Sodium fluoroacetate (1080) -
skin -- 0.05
Sodium hydroxide -- 2
Stibine 0.1 0.5
Stoddard solvent 500 2,900
Strychnine -- 0.15
0
0
, 950
700
670
0.8
13.5
19
0. 1
310
22
0. 1
0. 4
0.4
1
0. 1
1
1
3
12
0. 1
0,002
cStyrene monomer (phenyl-
ethylene) 100 420
Sulfur dioxide 5 13
Sulfur hexafluoride 1,000 6,000
Sulfuric acid -- 1
Sulfur monochloride 1 6
Sulfur pentafluoride 0. 025 0. 25
Sulfuryl fluoride 5 20
Systox, see Demeton
2,4,5 T -_ 10
Tantalum - - 5
TEDP - skin -- 0. 2
Teflon (R) decomposition
products -- A4
Tellurium -- 0. 1
TEPP - skin -- 0. 05
dl, 1, 1, 2-Tetrachloro-2, 2-
difluoroethane 500 4, 170
1, 1,2, 2-Tetrachloro-l,2-
difluoroethane 500 4, 170
1, 1, 2, 2-Tetrachloroethane-
skin 5 35
Tetrachloroethylene, see
Perchloroethylene
Tetrachloromethane, see
Carbon tetrachloride
Tetraethyl lead (as Pb)-skin -- 0.075
Tetrahydrofuran 200 590
Tetranitromethane 1 8
Tetryl (2, 4, 6-trinitrophenyl-
methylnitramine)-skin -- 1.5
Thallium (soluble compounds)-
skin -- 0. 1
Thiram -- 5
Tin (inorganic compounds,
except oxide) -- 2
Tin (organic compounds) -- 0. 1
Titanium dioxide -- 15
Toluene (toluol) 200 750
cToluene-2, 4-diisocyanate 0.02 0.14
o-Toluidine-skin 5 22
Toxaphene, see Chlorinated
camphene
1, 1, 1-Trichloroethane, see
Methyl chloroform
Trichloroethylene 100 535
Trichloromethane, see
Chloroform
Trichloronaphthalene-skin -- 5
1, 2, 3-Trichloropropane 50 300
1, 1,2-Trichloro 1,2,2-tri-
fluoroethane 1,000 7,600
Triethylamine 25 100
Trifluoromonobromomethane 1,000 6,100
2, 4, 6-Trinitrophenol see
Picric acid
2, 4, 6-Trinitrophenylmethyl-
nitramine, see Tetryl
-------
876
MISCELLANEOUS DATA
Substance
Trinitrotoluene -skin
Triorthocresyl phosphate
Triphenyl phosphate
Turpentine
Uranium {soluble compounds)
(insoluble compounds)
GVanadium {V^C^ dust)
(V2O5 fume)
Vinyl benzene, see Styrene
cVinyl chloride
ppma
--
--
--
100
--
--
--
--
500
mg/m3 b
1.5
0. 1
3
560
0.05
0. 25
0. 5
0. 1
1,300
Substance ppma
Vinylcyanide, see Acrylo-
nitrile
Vinyl toluene 100
Warfarin
eXylene (xylol)
Xylidine-skin 5
d Yttrium
Zinc oxide fume
Zirconium, compounds (as Zr)
mg/m3 b
480
0. 1
--
25
1
5
5
Radioactivity: For permissible concentrations of radioisotopes in air, see U. S. Department of Commerce
National Bureau of Standards, Handbook 69, "Maximum Permissible Body Burdens and Maximum Permis-
sible Concentrations of Radionuclides in Air and in Water for Occupational Exposure, " June 5, 1959. Also
see U. S. Department of Commerce National Bureau of Standards, Handbook 59, "Permissible Dose from
External Sources of Ionizing Radiation," September 24, 1954, and addendum of April 15, 1958.
Note: Footnotes to Recommended Values.
aParts of vapor or gas per million parts of air plus vapor by volume at 25°C and 760 mm. Hg
pressure.
Approximate milligrams of particulate per cubic meter of air.
""Indicates a vahae that should not be exceeded.
d!966 addition.
eSee tentative limits.
See A values on page 878.
Substance
Respirable Dusts Evaluated by Count
mp/ft3 a Substance
mp/ft
3 a
Silica
Crystalline
Quartz, threshold limit calculated
from the formula
25 0
%SiO
Cristobalite formula calculated "
Amorphous, including natural
diatomaceous earth 20
Silicates {less than 1% crystalline silica)
Asbestos 5
Mica 20
Soapstone 20
Talc
Portland Cement
Miscellaneous (less than 1% crystalline
silica)d
Graphite (natural)
"Inert" or Nuisance Par-
ticulate s
see Appendix D
10
50
50
50 (or 15 mg/m -which-
ever is the smaller)
Conversion factors
mppcf x 35. 3 = million particles per cubic meter
= particles per c.c.
Note: Footnotes to Respirable Dusts Evaluated by Count.
aMillions of particles per cubic foot of air, based on impinger samples counted by light-field technics.
The percentage of crystalline silica in the 'formula is the amount determined from air-borne samples,
except in those instances in which other methods have been shown to be applicable.
-------
Threshold Limit Values
877
Tentative Values
These substances, with their corresponding tentative limits, comprise those for which a limit has- been
assigned for the first time or for which a change in the 'Recommended' listing has. been made. In both
cases, the assigned limits should be considered trial values that will remain in the tentative listing for a
period of at least two years, during which time difinitive evidence and experience is. sought. If accept-
able at the end of two years, these substances and values will be moved to the RECOMMENDED'list.
Documentation for tentative values are available for each of these substances,..
Substance
Acrylamide-skin
2 - Aminopyridine
sec-Amyl acetate
Azinphos -methyl -skin
Bromoform-skin
n- Butyl acetate
sec-Butyl acetate
tert-Butyl acetate
esec-Butyl alcohol
Cadmium (metal dust and
soluble salts)
Carbon black
Carbon monoxide
6 ex- Chlor oacetophenone
(phenacychloride)
o-Chlorobenzylidene malono-
nitrile (OCBM)
cChlorine
eChromium, sol. chromic,
chromous salts, as Cr
metallic and insoluble salts
Coal tar pitch volatiles (ben-
zene soluble fraction)
(anthracene, BaP, phenan-
threne, acridine, chrysene,
pyrene)
eCobalt, metal fume and dust
Crotonaldehyde
Cumene-skin
Cyclohexane
Cyclohexene
Diaz omethane
1, 2-Dibromo-ethane-skin
Dibutyl phosphate
dD ibutylphthalate
Diethylamino ethanol-skin
eDiisopropylamine-skin
Dimethylphthalate
eDiphenyl
Ethylamine
Ethyl sec-amyl ketone (5-
methyl-3-heptanone)
Ethyl benzene
Ethyl butyl ketone (3-Heptan-
one)
Ethylene glycol dinitrate and/
or nitroglycerin-skin
Ethylene imine-skin
ppma
--
0.5
125
--
0.5
150.
200
200.
150.
--
--
50.
0.05
.05
1.
--
—
--
__
2.
50.
300.
300.
0.2
25.
1.
--
10.
5.
0.2
10.
25.
100.
50.
0. 02f
0.5
mg/m3 b
0.3
2
650
0.2
5
710.
950
950.
450.
0.2
3.5
55.
0.3
0.4
3.
0.5
1.
0.2
0. 1
6.
245.
1,050.
1,015.
0.4
190.
5.
5.
50.
20.
5.
1.
18.
130.
435.
230.
1
O.lf
1.
Substance
""Ethyl mercaptan
N-EthylmorphoIine-skin.
Fibrous glass
Formic acid
eGasoline
sec-Hexyl acetate
eHexachloronaphtha!ene -skin-
Iron oxide fume
Isoamyl acetate
Isobutyl acetate
elsobutyl alcohol
Isopropyl acetate
Maleic anhydride
Methylamine
Methyl (n-amyl) ketone ("2.-
Heptanone)
Methyl iodide-skin
Methyl isocyanate-skin
cMonomethyl hydrazine-skin
dNaphtha (coal tar)
eNitric oxide
eOctachloronaphthalene-sk±n
Oxalic acid
eParquat-skin
Phenyl ether (vapor)
Phenyl ether-Biphenyl mix-
ture (vapor)
dPhenyl glycidyl ether (PGE>
Pival (2-Pivalyl-l, 3-in-
dandione)
ePropyl alcohol
Propylene imine-skin
Rhodium, metal fume and
dusts
Soluble salts
eRonnel
Selenium hexafluoride
Tellurium hexafluoride
c. eTerphenyls
eTetrachloronaphthalene-skin
Tetramethyl lead (TML> (as-
lead)-skin
Tetramethyl succinonitrile —
skin
Tremolite
eTributyl phosphate
1,1, 2-Trichloroethane-skin
Xylene
Zinc chloride
ppma
10.
20.
--
5.
A6
50.
--
__
100.
150.
100.
250.
0.25
10.
100.
5.
0.02
0.2
100.
25.
--
--
--
1.
1.
LO.
--
200.
2.
—
--
--
0. 05
0. 0-2
1.
mg/m3 b
25.
94.
5.
9.
300.
0.2
10.
525.
700.
300.
950.
1.
12.
465.
28.
0. 05
0.35
400.
30.
0. 1
1.
0.5
7.
7.
62.
0. 1
450
5.
0. 1
0. 001
15.
0.4
0.2
9.4
2.
0.075
0-. 5 3.
5 mppcf
10.
100.
—
5.
45.
435.
1.
-------
878 MISCELLANEOUS DATA
Note: Footnotes to Tentative Values.
aParts of vapor or gas per million parts of air plus vapor by volume at 25°C and 760 mm Hg pres-
sure.
Approximate milligrams of particulate per cubic meter of air.
clndicates a. value that should not be exceeded.
1966 revision.
e!966 additions.
For intermittent exposures only.
"A" Values
A Benzidine. Because of high incidence of bladder tumors in man, any exposure, including skin, is
extremely hazardous.
A /3-Naphthylamine. Because of the extremely high incidence of bladder tumors in workers handling
this compound and the inability to control exposures, /3-naphthylamine has been prohibited from
manufacture, use and other activities that involve human contact by the State of Pennsylvania.
A N-Nitrosodimethylamine. Because of extremely high toxicity and presumed carcinogenic potential
of this compound, contact by any route should not be permitted.
4
A Polytetrafluoroethylene* decomposition products. Thermal decomposition of the fluorocarbon
chain in air leads to the formation of oxidized products containing carbon, fluorine,and oxygen.
Because these products decompose by hydrolysis in alkaline solution, they can be quantitatively
determined in air as fluoride to provide an index of exposure. No TLV is recommended pending
determination of the toxicity of the products, but air concentrations should be minimal.
A /SPropiolactone. Because of high acute toxicity and demonstrated skin tumor production in animals,
contact by any route should be avoided.
A Gasoline. The composition of gasoline varies greatly and thus a single TLV for all types of gaso-
line is no longer applicable. In s/ leral, the aromatic hydrocarbon content will determine what
TLV applies. Consequently the content of benzene, other aromatics and additives should be de-
termined to arrive at the appropriate TLV (Elkins, et al. , A. I. H. A. J. Z4, 99, 1963).
*Trade Names: Algoflon, Fluon, Halon, Teflon, Tetran
-------
Enthalpies of Various Gases Expressed in Btu/lb of Gas
879
Table D3. ENTHALPIES OF VARIOUS GASES
EXPRESSED IN Btu/lb OF GAS
Temp,
DF
100
150
200
250
300
350
400
450
500
550
600
700
800
900
1,000
1, 100
1,200
1,300
1,400
1,500
1,600
1,700
1,800
1,900
2,000
2, 100
2,200
2,300
2,400
2,500
3,000
3, 500
co2
5.8
17.6
29.3
40.3
51.3
63. 1
74. 9
87.0
99. 1
111.8
124.5
150.2
176.8
204. 1
231.9
260.2
289. 0
318. 0
347.6
377.6
407.8
438.2
469. 1
500. 1
531.4
562.8
594.3
626.2
658.2
690.2
852.3
1,017.4
N2
6.4
20.6
34.8
47.7
59.8
73.3
84.9
97.5
110. 1
122.9
135.6
161.4
187.4
213.8
240.5
267.5
294.9
326. 1
350.5
378.7
407. 3
435. 9
464.8
493.7
523.0
552.7
582.0
612.3
642.3
672.3
823.8
978.0
H20"
17.8
40.3
62.7
85.5
108.2
131.3
154.3
177. 7
201. 0
224.8
248. 7
297. 1
346.4
396. 7
447,7
499.7
552.9
606.8
661.3
717.6
774.2
831.4
889.8
948.7
1,003. 1
1,069.2
1, 130.3
1, 192.6
1,256.8
1,318. 1
1,640.2
1,975.4
°2
8.8
19.8
30.9
42. 1
53.4
64.8
76.2
87.8
99.5
111.3
123.2
147.2
171.7
196.5
221. 6
247.0
272.7
298.5
324.6
350.8
377.3
403. 7
430. 4
457. 3
484. 5
511.4
538.6
566. 1
593.5
621.0
760. 1
901. 7
Air
9.6
21.6
33.6
45.7
57.8
70.0
82. 1
94. 4
106. 7
119.2
131.6
156.7
182.2
211.4
234. 1
260.5
287.2
314.2
341.5
369.0
396.8
424.6
452.9
481.2
509.5
538. 1
567. 1
596. 1
625.0
654. 3
802.3
950.3
The enthalpies tabulated for E^O represent a gaseous system, and
the enthalpies do not include the latent heat of vaporization. It is
recommended that the latent heat of vaporization at 60°F (1, 059- 9
Btu/lb) be used where necessary.
-------
880
MISCELLANEOUS DATA
Table D4. ENTHALPIES OF GASES EXPRESSED IN Btu/scf OF GAS, REFERENCE 60°F
°F
60
77
100
200
300
400
500
600
700
800
900
1,000
1, 100
1,200
1,300
1, 400
1,500
1,600
1,700
1,800
1,900
2, 000
2, 100
2,200
2,300
2, 400
2,500
3,000
3,500
4,000
4, 500
5,000
5,500
6,000
6,500
N2
-
0.31
0. 74
2. 58
4. 42
6.27
8. 14
10.02
11. 93
13.85
15.80
17.77
19.78
21. 79
23.84
25. 90
27. 98
30. 10
32.21
34. 34
36.48
38. 65
40.84
43. 00
45.24
47. 46
49. 67
60.91
72. 31
83. 79
95. 37
107.04
118. 78
132.54
142. 37
°2
-
0.31
0.74
2.61
4. 50
6.43
8.40
10.40
12.43
14.49
16. 59
18.71
20.85
23.02
25.20
27. 40
29.62
31.85
34. 10
36. 34
38.61
40. 90
43. 17
45.47
47.79
50. 11
52. 43
64. 18
76. 13
88.29
100. 64
113.20
125.89
139. 74
151. 72
Air
-
0.32
0.74
2.58
4.42
6.29
8. 17
10.07
12.00
13.95
15.92
17.92
19.94
21.98
24. 05
26. 13
28.24
30.38
32.50
34.66
36.82
38.99
41. 18
43.39
45.61
47.83
50.07
61.39
72.87
84.42
96. 11
107.91
119.78
131.73
143.76
H2
-
0. 31
0. 73
2.55
4.40
6.24
8.09
9.89
11.77
13.61
15.47
17.36
19.20
21. 08
22. 95
24.87
26.80
28. 70
30.62
32.52
34. 45
36.43
38.49
40.57
42.66
44. 71
46.82
57.22
68. 14
79. 38
90. 68
102.42
114.21
126. 16
138. 35
CO
-
0. 32
0. 74
2.58
4. 43
6.29
8. 18
10. 08
12.01
13.96
15.94
17.94
19.97
22. 02
24. 10
26. 19
28.31
30.44
32.58
34. 74
36. 93
39. 12
41.31
43. 53
45. 74
47.99
50. 23
61.55
73. 00
84.56
96.21
107.93
119. 70
131. 52
143.37
CO2
-
0. 39
0. 94
3.39
5. 98
8.69
11.52
14. 44
17. 45
20.54
23.70
26.92
30.21
33. 55
36. 93
40. 36
43.85
47.35
50.89
54.48
58. 07
61. 71
65.35
69. 02
72. 71
76.43
80. 15
98. 96
118. 15
137. 62
157. 20
176. 93
196. 77
216. 77
236.88
H20a
-
0. 36
0.85
2.98
5. 14
7. 33
9.52
11.81
14. 11
16.45
18.84
21.27
23.74
26. 26
28.82
31.42
34.08
36.77
39.49
42.26
45. 06
47. 91
50. 78
53. 68
56.64
59.59
62.60
77.98
93.92
110.28
126.96
143.92
161. 07
178.41
195.82
aEnthalpies are for a gaseous system, and do not include latent heat
L = 1, 059. 9 Btu/lb or 50. 34 Btu/scf of H2O vapor at 60°F and 14.
of vaporization.
696 psia.
-------
Combustion Data Based on 1 Pound of Fuel Oil
881
Table D5. TYPICAL PHYSICAL PROPERTIES OF FUEL OILS
Pacific standard No.
Grade
Common name
T
y a
P n
i a
c 1
a y
1 s
i
s
S
P c
e a
c t
i i
f 0
i n
c s
Carbon '(C)
Hydrogen (H)
Sulfur (S)d
Water (H2O)
Other
("Be')
Ib/gal
Sp gr 60°/60°
Approximate
Btu/gal
Appr oximate
Btu/lb
Max viscosity
Flash) Min
point) Max
Max water and
sediment
Max 10% point
Max 90% point
Max endpoint
PS No. 100
1
Kerosine
2
Distillate
84. 7%
15.3%
0. 02%
-
-
41.8°
6.83
0.82
136,000
19,910
1
-
110°F
165°Fa
0. 05%
420°F
-
600°F
2
.
125°F
190°Fa
0. 05%
440 "F
620°F
-
PS No. 200
3
Straight -run fuel oil
85.8%
12. 1%
1.2%
-
0.9%
26.2°
7.50
0.90
142,000
18, 950
45 sec (100°F)b
150°F
200°Fa
0. 1%
460°F
675"F
-
PS No. 300
5
Low-crack fuel oil
87. 5%
10.2%
1.1%
0. 05%
1.1%
16.5°
8
0.96
146,000
18,250
40 sec (122°F)C
150°F
-
1.0%
-
-
-
PS No. 400
6
Heavy-crack fuel oil
88.3%
9.5%
1.2%
0. 05%
1.0%
8.9°
8.33
1
152,000
18,000
300 sec (122°F)C
150°F
-
2.0%
-
-
-
Or legal maximum.
^Saybolt Universal.
cSaybolt Furol.
"Sulfur contents are only typical and will vary in different locales.
Table D6. COMBUSTION DATA BASED ON 1 POUND OF FUEL OILa> b> c
Constituent
Theoretical air
(40% sat'd at 60 °F)
Flue gas
constitu-
ents with
theoret-
ical air
co2
SO2
N2
H,O formed
H2O (fuel)
H2O (air)
Total
Amount of
flue gas
with %
excess
air as
indicated;
0
7.5
10
12.5
15
17.5
20
30
40
50
75
100
SO^ % by vol and wt
with theoretical air
PS No. 100
ft3
197.3
26.73
0.002
154.8
28. 76
1.367
211.659
211. 7
226.5
231.4
236.4
241.3
246.2
251.2
270. 9
290.6
310.4
359.7
409. 0
0.0011
Ib
15.04
3. 104
0. 0004
11.44
1.368
0.0662
15. 9786
15.98
17. 11
17,48
17.86
18.24
18.61
18.99
20.49
22. 00
23.50
27. 26
31.02
0. 0025
PS No. 200
ft3
185. 1
27.08
0. 142
145.2
22. 75
1.283
196.455
196.5
210.4
215.0
219.6
224.3
228.9
233.5
252.0
270.5
289.1
335.3
381.6
0.072
Ib
14. 11
3. 144
0. 0240
10. 74
1. 082
0. 0621
15. 0521
15.05
16. 11
16.46
16.81
17. 17
17.52
17.87
19.28
20.69
22. 11
25.63
29.16
0.16
PS No. 300
ft3
179. 1
27.61
0. 130
140.5
19. 18
0.011
1.242
188.673
188.7
202. 1
206.6
211. 1
215.6
220.0
224.5
242.4
260.3
278.3
323.0
367.8
0.069
Ib
13.66
3.207
0. 0220
10.39
0. 9118
0.0005
0. 0601
14. 5914
14.59
15. 62
15.96
16.30
16.64
16.98
17.32
18.69
20. 05
21.42
24.84
28.25
0. 15
PS No. 400
ft3
177.2
27.86
0. 142
139.0
17.86
0. Oil
1.228
186. 101
186.0
199.4
203.8
208.3
212.7
217. 1
221.5
239.3
257,0
274.7
319.0
363.3
0.076
Ib
13.51
3.236
0. 0240
10.28
0. 8491
0. 0005
0.0595
14, 4491
14.45
15, 46
15. 80
16. 14
16.48
16.81
17. 15
18.50
19.85
21.21
24.58
27.96
0.17
aCombustion products calculated for combustion with air 40% saturated at 60°F. All volumes
measured as gases at 60°F. Moisture in fuel included where indicated.
^Maximum accuracy of calculations: 1:1000.
cBased on physical properties in Table D5.
-------
882
MISCELLANEOUS DATA
Table D7. COMBUSTION CHARACTERISTICS OF NATURAL GAS
Average analysis,
CO2
N2
O^
CH4
C2H6
C,Hn
i-C H.Q
n-C4H1Q
c5+
volume %a
0
5. 15
0
81. 11
9.665
3.505
0. 19
0. 24
0.09
0.05
100.00
Average gross heat, 1, 100 Btu/ft
Air required for combustion
3 3
Theoretical - 10.360ft /ft gas
20% excess ^ir - 12.432 ft3/ft3 gas
Products of combustion/ft of gas
Theoretical air
CO2
H2°
N2
°2
Total
Vol
1. 134 ft3
2.083
8.236
-
11.453 ft^
Wt
0. 132 Ib
0.099
0. 609
0.840 Ib
20% excess air
Vol
1. 134 ft
2.083
9.873
0.435
13.525 ft3
Wt
0. 132 Ib
0. 099
0. 731
0. 037
0. 999" Ib
Available heat, Btu/ft gas,a based on latent heat of vaporization of water at 60°F
Temp, "F
Available heat, Btu,
with theoretical air
Available heat, Btu, 20% excess air
100
150
200
250
300
350
400
450
500
550
600
700
800
900
1,000
1, 100
1,200
1,300
1,400
1,500
1,600
1,700
1,800
1,900
2,000
2, 100
2,200
2,300
2,400
2,500
3,000
3,500
988.6
976. 1
963.7
952. 1
941.0
928.8
917.8
906.2
894.6
882.7
870.9
846.2
820.7
797.7
772.6
747.2
721.3
693.0
668.6
642.7
614.6
589.8
562.3
534.8
507.5
478.7
450.7
421.9
393.0
364.6
219. 1
70.4
992.2
973.0
958.5
949.9
932.0
917.8
905. 1
891.5
878. 0
864. 1
850.4
821.8
792. 3
765. 3
736. 2
706. 6
676.5
643.6
615. 4
584.5
552.9
523. 7
491. 7
459.9
438.2
394.9
362.5
329. 1
295.6
262.6
94.2
--
aAverage of two samples analyzed by Southern Calif. Gas Co. , 1956.
-------
Conversion Table of Velocity (V) to Velocity Pressure (VP)
883
Table D8. CONVERSION TABLE OF VELOCITY (V) TO VELOCITY PRESSURE (VP)
Velocity,
fpm
800
850
900
950
1, 000
1,050
1, 100
1, 150
1, 200
1,250
1, 300
1, 350
1, 400
1,450
1,500
1,550
1,600
1,650
1, 700
1,750
1,800
1,850
1,900
1,950
2, 000
2, 050
2, 100
2, 150
2,200
2,250
2, 300
2, 350
2, 400
2, 450
2,500
VP at 70 °F
in. WC
0. 040
0.045
0.051
0.056
0. 062
0.069
0.075
0.082
0.090
0. 097
0. 105
0. 114
0. 122
0. 131
0. 140
0. 150
0. 160
0. 170
0. 180
0. 191
0.202
0. 213
0.225
0.237
0.249
0.262
0.275
0.288
0.301
0.316
0.329
0.344
0.359
0.375
0.389
VPat 60 °F
in. WC
0.041
0.046
0.052
0.057
0.063
0. 070
0. 077
0. 084
0. 092
0.099
0. 107
0. 116
0. 124
0. 134
0. 143
0. 153
0. 163
0. 173
0. 184
0. 195
0. 206
0.217
0.229
0. 242
0.254
0.267
0. 280
0.294
0.307
0. 322
0.335
0.351
0.366
0.382
0.396
Velocity,
fpm
2,550
2, 600
2, 650
2. 700
2, 750
2,800
2, 850
2, 900
2, 950
3, 000
3, 050
3, 100
3, 150
3, 200
3,250
3, 300
3, 350
3, 400
3,450
3,500
3,550
3, 600
3, 650
3, 700
3, 750
3, 800
3,850
3, 900
3, 950
4, 000
4, 050
4, 100
4, 150
4, 200
4,250
VP at 70°F
in. WC
0. 406
0.421
0. 438
0. 454
0. 472
0.489
0.507
0.524
0. 543
0.561
0.581
0.599
0.618
0. 638
0.658
0.678
0.699
0. 730
0. 742
0. 764
0. 785
0.808
0.830
0.853
0.876
0. 900
0. 924
0. 948
0. 973
0. 998
1. 022
1.049
1. 073
1. 100
1. 126
VP at 60°F
in. WC
0. 414
0.429
0. 446
0. 463
0. 481
0.498
0.517
0.534
0.553
0. 572
0.592
0.611
0.630
0.650
0.671
0.691
0.712
0. 734
0.756
0.779
0. 800
0.824
0.846
0.869
0.893
0. 917
0. 942
0. 966
0.992
1'. 017
1.042
1.069
1.094
1. 122
1. 148
Velocity,
fpm
4,300
4,350
4, 400
4, 450
4, 500
4, 550
4, 600
4,650
4,700
4,750
4,800
4,850
4,900
4,950
5,000
5, 050
5, 100
5, 150
5,200
5,250
5,300
5,350
5,400
5, 450
5,500
5,550
5,600
5, 650
5,700
5, 750
5, 800
5,850
5,900
5,950
6,000
VPat 70°F
in. WC
1. 152
1. 179
1.208
1.235
1.262
1.291
1.319
1. 348
1. 377
1.407
1. 435
1.466
1.496
1.527
1.558
1.590
1. 621
1.654
1. 685
1.718
1. 751
1. 784
1.817
1.851
1.886
1.919
1.955
1.991
2.026
2.061
2. 098
2. 134
2. 170
2.207
2. 244
VP at 60°F
in. WC
1. 174
1. 202
1. 231
1. 259
1.286
1.316
1.344
1.374
1.403
1.434
1.463
1.494
1.525
1. 556
1.588
1.621
1.652
1.686
1. 717
1. 751
1.785
1.818
1.852
1.887
1. 922
1.956
1.993
2.029
2.065
2. 101
2. 138
2. 175
2.212
2.249
2.287
-------
884
MISCELLANEOUS DATA
Table D9. DENSITIES OF TYPICAL SOLID MATERIALS AS THEY OCCUR IN
MATE RIAL-HAND LING AND PROCESSING OPERATIONS
Material
Densities,
lb/ft3
Ashes, dry, loose 38
Ashes, wet, loose 47
Baking powder „ 56
Bone, ground, dry 75
Borax 105 to 110
Calcium carbide, crushed
3-1/2 in. x 2 in. , loose 77
2 in. x 1/2 in. , loose 75
1/2 in. x 1/8 in. , loose 80
1/8 in. x 0 in. , loose 82
Carbon, activated, very fine, dry 8 to 20
Cement, Portland, loose 94
Cement, Portland, clinker 95
Charcoal, broken, all sizes 15 to 30
Charcoal, broken, 1-1/2 in. x 0 in 14
Charcoal, ground ., . 10
Chips, wood 18
Cinders, blast furnace , 57
Cinders, coal, ashes, and clinker 40
Clay, dry in lumps, loose 63
Coal, anthracite, broken, loose 55 to 60
Coal, bituminous, broken, loose 50 to 54
Coal, bituminous, 5 in. x 0 in. , dry 54
Coal, bituminous, 1/2 in. x 0 in. , dry 45
Coal, bituminous, 1/8 in. x 0 in. , dry 43
Coke, lump, average 28 to 32
Coke, breeze 30 to 34
Coke, petroleum, lump 40 to 50
Cork, solid 15
Cork, in bales 8 to 9
Cork, ground, 10 in. rnesh x 0 in 4 to 5
Gullet, glass, average 85 to 100
Gullet, glass, 3/4 in. x 0 in 80 to 90
Dolomite, crushed, 2 in. x 1/2 in 94
Dolomite, crushed, 1/2 in. x 0 in 98
Earth, common loam, dry, loose 76
Earth, common loam, moist, loose 73
Feldspar, broken, in loose piles 90 to 100
Fluorspar, broken, in loose piles 110 to 125
Fluorspar, ground, 100 mesh x 0 in 90 to 100
Flint, pebbles 105
Fullers Earth, dry 30 to 35
Glass batch 90 to 110
Gniess, broken, in loose piles 96
Granite, broken, in loose piles 96
Granite, crushed, 1-1/4 in. x 10 mesh 98
Gravel, mixed sizes, loose 96 to 100
Gravel, 2 in. x 1/4 in. , loose , 105 to 110
Gravel, 3/4 in. x 1/8 in. , loose 98 to 100
Greenstone, broken, in loose piles 107
-------
Densities of Typical Solid Materials
885
Material
Densities,
lb/ft3
Gypsum, broken, in loose piles 90 to 94
Gypsum, crushed, 1 in. x 0 in. , loose 90
Gypsum, ground, loose 50 to 56
Iron (cast) borings, fine 120 to 155
Iron ore, loose 125 to 150
Lime, hydrated, -200 mesh 20 to 25
Lime, quick, lump, 1-1/2 in. x 0 in 70 to 80
Lime, quick, lump, 1/2 in. x 0 in 70
Lime, quick, ground 60 to 65
Lime, quick, from oyster shells, loose . 45 to 50
Limestone, broken, in loose piles 95
Limestone, sized 3 in. x 2 in. , loose 95
Limestone, sized, 2 in. x 1/2 in. , loose 92
Limestone, sized, 1/2 in. x 0 in. , loose 96
Limestone, ground, -50 mesh, loose 84
Limestone, ground, -200 mesh, loose 65
Marble, crushed 95
Oyster shells, piled 60
Phosphate rock, broken, in loose piles 75 to 85
Phosphate rock, pebble 90 to 100
Quartz, broken, in loose piles 94
Rubber, shredded scrap 46
Salt, coarse 45 to 52
Salt, fine 42 to 50
Salt, table 42 to 45
Salt, rock, broken, in loose piles 50
Salt, cake, coarse 55 to 60
Salt, cake, fine 45 to 50
Sand, dry, loose 90 to 95
Sand, wet, loose 105 to 110
Sand and gravel, dry 90 to 105
Sand and gravel, wet 105 to 125
Sand, molding, prepared and loose 77 to 80
Sand, molding, rammed 90 to 100
Sand, molding, shaken out or new 100
Sandstone, broken, in loose piles 82 to 86
Sawdust, dry 7 to 12
Scale, rolling mill 105
Shale, crushed, in loose piles 92
Slag, bank, crushed 80
Slag, furnace, granulated 60
Soda ash, dense 60 to 62
Soda ash, light 28 to 32
Soda, bicarbonate, loose 50 to 58
Starch, granular 22 to 25
Stone, crushed, 1 in. x 0 in 85 to 105
Sugar, granulated, loose 42 to BO
Sugar, brown 45 to 55
Sulphur, ground, -100 mesh 75 to 85
Sulphur, ground, -200 mesh 50 to 55
Trap rock, broken, in loose piles 107
Trap rock, crushed 95 to 1 05
-------
886
MISCELLANEOUS DATA
Q
a
u
<3
^
ID
O
U
co W
§g
Q ^
^ O
CO
H
a
S
<
z
w
Q
ret
H
tt) ^H
013 O
a >
5 -°
•^ ^
I*
CO
a)
, — i
'-^ 1>
tt >
o >>•
> •£
Z
^
nl
M
""Q
^
S
K^
"i
ni
cu
r-j
^j
"rt
ID
^
0)
S
fi
•H
H
i — I
g
rt
fl
W
IH
3
cr
U
j
0)
d
n)
W
y--^.
irt 'o
T U
^ "
H-i •— '
r-H
s
a!
fl
H
0 ~
•] ^,
CJ -^~^
H
, — 1
I
a)
H
IH
0)
cr
n)
^
fi
ni
ft
O
U
m
IN]
O
+-1
0
C\J
0
o
—"
0
o
00
o
m
O
4J
o
*t
m
^
o
•^
o
m
0
r-H
0
0
oo
m
r-'
o
^
v£3
in
r~
o
m
0
0)
n)
f-H
ft
"M
Pittsbur
Glass
0
tNI
O
-*-1
o
— 1
o
0
tM
O
0
in
0
m
o
•^
o
-M
s
o
o
-^
in
(M
0
i— i
O
0s
2
0
+J
0
o
rH
O
0
m
t~
0
(N]
jj
M
r-'
o
^
t-'
i — 1
a
ni
fi
W
2 |
en
•H
"o
to
0
v£>
1
&
0 G
N ^
in
tM
0
•^
0
tM
0
i — i
0
o
o
o
m
o
m
o
-p
o
in
^sl
m
0^
0
^
CO
oo
o
in
r-
m
00
0
oo
o
r~'
o
m
xD
m
00
13
ni
0)
i-l ^
National
Compa
to
•H
"o
m
o
^ ^
1 ^H
0*2
O c
n -H
N £
in
ro
0
•^
0
r\l
O
O
i — i
in
o
40
in
•*
o
^
o
-*°
s
m
0
•4-J
0
f\]
o
o
in
00
1
1
o
f-
o
v£5
O
oo
J
fl
o
cq a'
Andrew
Compa
1X1
INJ
O
, — i
C^
•*
0
^r
m
M
(M
tT-
00
CO
oo
o
•"'
•xT
r-
CJ
60
fn
0)
>
.
.s
ft
nj
•H
OJO
• H
0
"o
g
"o
fi
o
"O
a
w
rt
CQ
m
-------
Concentrations of Materials in the Air
887
ro n
CO -I-1
cu ro
= a.
CO CD
ca Q-
cu
-C -C
-(-• tiD
c ^
— J3
CO
CO -I-1
-M
ca —
o -o
— TO
-------
888
MISCELLANEOUS DATA
...
X-R
ay Dif
6,2
Ultramicrosc
10
E
Ult
Ultracentrifu
22
raction
)
ope
ectro
afiltr.
8
?e
Microscope
6,23
tion
Cent
Sec
Tu
Ll
Microsco
4
ifuge
6
imentation &
13,19
Permeability
9,11,17
rbidimetry
2,5
;ht Scattering
15,21
pe
Gas
Sravit
21
Elutnation
,7,18
y Settling
^Maj
nine '
Visi
ools (Microme
iievm
lie to
12,24
ter, V
Eye
srnier Caliper,
etc)
•- PARTICLE SIZE LIMITS
UNDER AVERAGE CONDITIONS.
1. STATED METHOD IS OF
DOUBTFUL UTILITY IN THESE
SIZE RANGES
VARIATIONS IN THE LIMITS OF EACH
METHOD ARE POSSIBLE DEPENDING
ON THE QUALITY OF THE INSTRUMENT,
SKILL Of THE OPERATOR, ETC
00005 0001 0005 0.01
005 0.1 05 1 5 10
PARTICLE SIZE (Microns)
Note: The numbers in Figures
02 and D3 represent bibl iog-
so 100 500 1,000 5,000 10,000 raphy references that can be
CONVENTIONS furnished upon request.
Figure D2. Limits of particle size-measuring equipment (Issued as a public service
by Mines Safety Appliances Co., 201 N. Braddock Ave., Pittsburgh, Pa. Prepared by
Southwest Research Institute.).
Figure D3. Sizes of airborne contaminants (Issued as a public service by Mines
Safety Appliances Co., 201 N. Braddock Ave., Pittsburgh, Pa. Prepared by South-
west Research Insti tute.).
-------
The Frank Chart
889
DIAM
OF
PAR-
TICLES
IN
riiciww
8000
6000
4000
2000
1000
800
600
400
200
100
80
60
40
20
10
6
4
2
I
.8
.6
.4
.2
ni
.001
US.
ST'D
MESH
10-
20-
60-
100-
150-
?00-
iiiO-
32S-
1000-
SCALE OF
ATMOSPHERIC IMPURITIES
TT
9
15 —
I*'
64 ui
1" £f
-V-
^5
a.
?•!«
r-I*
1 -
1JJ
Tc s
S I
F
< c
E
i. JL.
&*
- 25]
?
e
3
)
S
128
I
XI
-is
[3
Of
o
-z-
e
u.
tn
O
a.
^
2
Ifr
-^-
sl
1)
UJ
£
&
?
g
1,
1
^z^gfc
fffffl
iHtigyr
fe "* S "* S "'
is Iff
j5
,
F=
ft(
,
2
S
•4r
LL
III
Z
a
c!
-!*
-pr
u}
£
Z
S
^.
UJ
O
<
=c
vn
§
s
^ f-
•*-* L_
-£. ee-
•5 t^ lo
UJ ?? 3
szi
t
"L .
^JoS1
3|§
S o o
Egg-^
? •
1 '
""'•flrj' '" "
"o — r
~f —
is
m S
g 8
5 ts
— s-
fti £
« 3
^ ,
u>
1
-U-
u)
z
y
S
o
H
a
UJ
a
T,
V
^
•z.
o
5
nr"
^
$
X
y_i
k
a
o
1>
5:
>
,
—
Uyv
^
5
X
a.
-+--
:•
Ul
ti
LL
i •
o -m-
1
— pr
- pi
fl Ei
^3 -2
— st^— ^f
8s o u
S fri f
Zp ""
d.
RATE OF
SETTLING
IN F.P.M
FOR
SPHERES
WtCRAVI
AT70-F.
790
59.2
14.8
.592
.14-8
.007=
5"P£R HR.
It'pfBHR
.00007«
o
0
o
DUST PARTICLES
CONTAINED IN
1 CUB. FT. OF AIR
(See Foot Note)
NUMBER
.075
.6
75
600
75000
600.000
75XI06
60X|0T
75*IO»
60X10'°
75XIO'1
75X|0'5
SURFACE
AREA
IN SO,. IN
.OOOJ6S
.00073
.00365S
fclN.SQ.
007i
03652
ilN.SQ
.073
.365 £
.73
3.65S
.9 IN. SO-
73
36.5 S
365 S
2.5»QFr.
LAWS OF SETTLING
IN RELATION TO
PARTICLE SIZE
(Lines of Demarcation approx.)
H
| PARTICLES SETTLE WITH CONSTANT VELOC
E
r
A
PARTICLES FALL WITH
INCREASING VELOCITY
•v /IpH
STOKES
LAW
C=? ~f)
FORAIRAT70'F.
C'300,460s,dl
C». 00592 s,0a
CUNNINGHAM'S
FACTOR
C«C'(|+K£)
C'=CofW)KtHAW
K= .8 TO .86
^ARTICLES M
GAS MOLE
JROWNIAN
MOVEMENT
•ntir
OVelocitiJ. cm/sec.
C'Velocity ft/win
d'Diom.ofpor-
ticle in cm.
D'Diam.of par-
ticle in Microns
r» Radius of par-
ticle in cm.
g»98l cm /sec?-
occelerotion
S^« Density of
particle
S/Densiti) of Air
(Very Small
relative to S,)
?7=Viscositi|of
air in poises
*l8l4X|0'7for
oirot70°H
*= IQ-* cm.
(Mean free
poth of QCS
molecules)
DVE LIKE
.CULES
A* Distance of
motion intim«t
*8.3I$XI07
T» Absolute
Temperature
N -Number of das
molecules in
onemol=606XlO"
FigLre 04. The Frank Chart: Size and characteristics of airborne solids.
It is assumed that the particles are of uniform spherical shape having
specific gravity 1 and that the dust concentration is 0.6 grain per
1,000 ft3 of air, the average of metropolitan districts. (Compiled by
>I.G. Frank, Copyrighted by American Air Filter Co., Inc., Dust Control
Products, Louisville, Ky.).
-------
890
MISCELLANEOUS DATA
Figure D5. Range of particle sizes, concentrations, and
collector performance (Compiled by S. Sylvan, April 1952-
Copynght, 1952, American Air Filter Co., Inc., Louisville,
Ky.).
Figure D6. Psychrometric chart for humid air based on 1 Ib
dry weight. (Copyright, 1951, American Air Filter Co., Inc.,
Louisville, Ky.).
-------
High-Temperature Psychrometric Chart
891
o
o
CM
•Jo 3Hfuva3d«3i sine IBM
o o o o
o> o> N
-------
892
MISCELLANEOUS DATA
+±
s
I
i
AS iN33)iad
GPO 8O6—614—30
------- |