iPA-450/3-74-031
IUNE 1974
AIR POLLUTION
CONTROL ENGINEERING
AND COST STUDY
OF THE PAINT
AND VARNISH INDUSTRY
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Water Programs
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
-------
FOREWARD
This air pollution control engineering and cost study of the paint and varnish industry was conducted
under Contract No. 68-02-0259 for the office of Air Quality Planning and Standards, Emission
Standards and Engineering Division, Industrial Studies Branch. The study is being conducted by
the EPA under public law No. 91604, Clean Air Amendments of 1970. This law permits that industry
studies be conducted to obtain information as outlined below.
Section 111
This section of the law requires that future growth of industry should contribute a minimum impact
on the quality of air. It provides that the Federal Government will review the capabilities for con-
trolling air pollution in industry and that certain designed source categories will be required to
install the best emission control systems.
One source of information used to develop performance standards required to ensure best emission
control evolves from the type of industry study provided herein. Trade associations of the industry
are consulted during the establishment of proposed standards. Plants which utilize the best control
techniques are examined and source tested. Test data and economics of control are considered
in the setting of standards. Proposed standards are then reviewed at open meetings by industry,
governmental agencies and other interested parties.
Section 114
This section of the law provides to the Environmental Protection Agency the right to such infor-
mation as it deems necessary for establishing standards. Industry cannot withhold information
from the EPA on the grounds that it is considered confidential. In accordance with title 18 of
Section 1905 in the United States Code, if the EPA requires information which in a company's
opinion is a trade secret then the company may submit, along with the information, a request
that the data be maintained confidential. This request must include reasons why the data is
confidential. EPA Legal Council will review the request and will render a decision as to whether
in
-------
it should be held confidential and will so notify the company. If the decision is made that it will
not be held confidential, the company has 30 days to consider legal action.
The three general categories of information requested of industry are:
a. Emission data
b. Economic information
c. Process information
The Clean Air Act clearly states that emission data is public information and will be made available
upon request. Process data may be held confidential only if the information can be shown to
constitute a trade secret. The Office of General Council must rule on any process data considered
to be confidential by a company based on the reasons submitted with the request for confidentiality.
With a determination that certain information is a trade secret, such information will be used for
setting performance standards, but the data will not be released to the general public. Such infor-
mation may be used in summaries or in legal action if required by the Federal Government for
carrying out the purpose of the act.
Section 113 — Federal Enforcement Powers
Most state agencies will in all likelihood enforce new source performance standards; however,
the federal agency has the right to intervene under this section of the law. Where a state is
not prepared to enforce the standards, the Federal Government will have the responsibility.
The paint and varnish industry shall be considered to consist of all phases of operations normally
found located in plants engaged in the manufacture of paints, lacquers, varnishes, etc. (Standard
Industrial Classification 2851). The study shall include the storing, packaging, shipping, handling,
mixing, thinning, grinding, cooking, and other processing necessary in the manufacturing of paint
and varnish products from raw materials through finished products. The manufacture of resins by
both paint manufacturers and raw material suppliers shall be included. The manufacture and
processing of a number of pigments will also be included in this study. The study of the pigments
will not be as detailed as the paint and varnish study and it will be limited to those manufacturing
processes which contribute significantly to environmental air pollution. The study will develop the
following information on the paint and varnish industry:
1. An industry description which incorporates all industry information relevant to emission
problems, including comprehensive industry statistics, emission sources, types of emissions as well
as emission quantities as related to operational factors.
IV
-------
2. A study of technical and economic information on the best control systems. An assess-
ment of the economic impact on the entire industry if these best controls were applied to new
and existing plants.
3. Identification of technological and economic deficiencies in air pollution control technology
within the paint and varnish industry. Analysis of those deficiencies that restrict additional reduction
of emissions along with recommendations of R&D Programs that would produce the greatest
improvement in control technology.
Resulting reports will include past and projected production data, geographical location of paint
plants, types and amounts of air pollutant emissions, control techniques, performance and costs
of existing and best control technology, impact of emissions on air quality, inspection procedures
to determine compliance with air pollution control regulations, and areas of needed research and
development.
Techniques used in the study included literature searches, plant visits, source testing and compre-
hensive industry questionnaires. Two subcontractors were also employed. These subcontracts are
discussed below.
The Industrial Gas Cleaning Institute (IGCI) was subcontracted to supply detailed capital, instal-
lation and operating cost information for air pollution equipment used by the coating industry.
The Sherwin-Williams Company was subcontracted to supply capital and operating cost information
for the model plant developed for this study. These two subcontracts are discussed in more detail
in Chapter 7, "ECONOMICS OF EMISSION CONTROL".
-------
ACKNOWLEDGEMENTS
The work reported herein has been reviewed by an Industry Advisory Committee comprised of the
members listed below. The authors wish to gratefully acknowledge the technical review, advice
and assistance they provided throughout the period of this contract.
Committee Representatives
Ashland Chemical Company J. Blegen
Conchemco, Incorporated R. Radford
E. I. duPont de Nemours W. Zimmt
Eagle-Picher Industries, Inc. H. Stephenson
IGCI G. Brewer
Los Angeles County APCD W. Krentz
N. Schaeffer (alternate)
N. J. Department of Environmental Protection M. Polakovic
NL Industries G. Rodman
National Paint & Coatings Association R. Brown
R. Connor (alternate)
PPG Industries D. Bridge
T. Duvall (alternate)
Reichhold Chemicals, Incorporated S. Hewett
Sherwin-Williams F. Gaugush
A. Thomas (alternate)
Stresen-Reuter International/Lawter Chemical, Inc. A. Stresen-Reuter
R. Voedisch (alternate)
Union Carbide B. Duzy
Environmental Protection Agency
Chief, Industrial Studies Branch S. Cuffe
Industrial Survey Section J. Dale
Chief, Industrial Survey Section J. Sableski
VI
-------
National Paint and Coating Association
Executive Vice President R. Roland
Manufacturing Management Committee W. Bartelt
"The Design and Construction of a Modern Paint R. Brewster
Manufacturing Facility"
Air Correction Division of U.O.P. — Equipment Manufacturer
Hirt Combustion Engineers — Equipment Manufacturer
The assistance of other organizations and companies who supplied assistance in completion of
this study are too numerous to list and are also gratefully acknowledged, particularly those who
filled out questionnaires, supplied data, provided time for visits, and permitted source testing.
VII
-------
SUMMARY
Activity in this program was directed to an analysis of the technical and economic
aspects of emission controls in the paint and varnish manufacturing industry. The main objective
of the study was to provide an improved technical and economic basis which the Environmental
Protection Agency could use to evaluate emissions and control technology in the paint and
varnish industry. Industry structure and statistics, the nature and sources of emissions, technical
problems associated with emission controls, cost of emission control equipment and the overall
economic impact of air pollution control were among the major topics investigated in the program.
The salient results of the program are highlighted in the following sections.
Process and Emissions
Significant air pollution problems associated with this industry are localized odor nuisance
complaints and fugitive solvent vapors. The odor complaints result primarily from open kettles,
varnish cooks, closed reactor resin cooks and handling operations of low odor threshold raw
materials. The quantity of this type of emission is very small but, nevertheless, contains a wide
variety of highly odorous substances. Fugitive and non-fugitive solvent losses represent the
major quantity of emissions from the manufacturing operations but these amount to only a small
percent of the solvent losses which occur during the application and drying of coatings.
A brief history and description of the paint and varnish industry is presented in this
chapter. The various production processes and manufacturing steps are discussed. The chemistry
of varnishes and resin cooking is presented in some detail with the production of alkyd resins
given major emphasis. Other resins discussed are vinyl, acrylic, epoxy, urethane, cellulosic,
amino, hydrocarbon, phenolic and silicone.
Process information gathered from an industry questionnaire is summarized. A total of
452 questionnaires were mailed to 382 different paint companies. Of these, 338 or about 75
percent were used in the process information summaries presented. Included in these summaries
is a listing of major products, raw materials and production equipment.
The design basis and material balance of the model plant developed to evaluate the
economics of the industry are presented in detail. The design of the plant was based on a
study made by the Management Committee of the National Paint and Coating Association and
presented at the 1972 NPCA annual meeting in a paper by Mr. R. F. Brewster. The model
plant was modified in capacity to more closely approach the average size plant reported in
VIII
-------
the industry questionnaire. On-site resin production facilities were also added.
A description of the type, source and quantity of emissions that are encountered in
the production of paints and varnishes is discussed along with the influence of various process
operations on these emissions. Emission data from the industry questionnaire, source tests by
afterburner equipment manufacturers and source tests conducted by the Federal EPA are
summarized and presented. Primary emphasis was placed on emissions from varnish and resin
cooking. A detailed source by source calculation of emissions from the model plant is also
presented. Finally, there is a brief discussion of liquid and solid wastes.
Primary sources of gaseous emissions are: varnish cooking, resin cooking and thinning.
Secondary sources of gaseous emissions are: handling and storage, milling operations, blending
and finishing, and filling. The primary source of particulate emissions is the pigment dusting
encountered in the milling operation.
As a part of this study, the Federal EPA conducted 10 source tests on emissions from
the production of four types of resins in four different resin reactors. The work was carried
out by Scott Research Laboratories under the direction of the EPA source test group. Flow
rates, total hydrocarbons and gas chromatography data were obtained and are summarized
and presented.
Industry Statistics
The Paint and Allied Product Industry (SIC 2851) is made up of approximately 1,727
establishments operated by some 1,365 companies. In this chapter, the type, size and geographical
distribution of these plants are presented and analyzed. Past, present and projected industry
trends to the year 1985 are also presented and discussed. Data is presented on production
and production capacity relations, number of plants, size of plants, and typical plant and
equipment ages. A discussion of the influence on the future of this industry due to technical
innovation and other outside effects is also presented.
Measurements of Emissions
As yet, there is no tried and proven method for source measurements of hydrocarbons.
This problem is even further complicated by the large variety and highly cyclic nature of the
emissions emitted from paint and varnish kettles. In this chapter, these problems and the
current state-of-the-art for both source and continuous monitoring are discussed. Detailed descriptions
of the methods currently used by industry and regulatory agencies are discussed. A description
IX
-------
of the method used by the Federal EPA during the source testing conducted for this study
is also presented.
Inventory of Emissions
In this chapter, emission factors are developed for this industry as a function of production
capacity and type of product. Solvent emission factors were developed for three types of plants.
Plants Producing Resins Only 13,700 Ib/MM Ib Resin
Plants Producing Resins & Paints 45,000 Ib/MM gal Solvent Based Paint,
5,460 Ib/MM Ib Resin
Plants Producing Paints Only 56,000 Ib/MM gal Solvent Based Paint
Data collected from the industry questionnaire was used as the basis for the above emission
factors. The emission factors for plants producing both paint and resins compare well in total
with those calculated for the study's model plant. As might be expected, these emission factors
represent averages of widely differing individual plant emissions and care should be exercised
in applying them arbitrarily to specific plants.
By using an average gaseous emission factor of 50,000 pounds per million gallons
of solvent based paint produced and paint production for 1972, potential gaseous emissions
in the U.S.A. from paint production are estimated to be 17,900 tons for the year 1972. A
geographical distribution of organic emissions is presented in Figure 41 on page 198.
Emission Control Technology
In this chapter, a detailed discussion of the existing state-of-the-art of air pollution
control for the paint and varnish industry is presented. The discussion includes a description
of the best control techniques for each emission source, alternate control techniques, methods
of control other than add-on equipment, performance of currently used systems and capability
of best control systems to meet more stringent standards. Potential water and solid waste
disposal systems are also reviewed. Best control for pigment particulate emissions are fabric
filters. Best control for gaseous emissions including odors are catalytic and thermal afterburners.
Considerable discussion is devoted to the design and application of afterburner systems to
resin and varnish kettles. A discussion of the use of a refrigerated condenser as an alternate
control device is also included.
-------
Inspection Procedures
This section of the report was written for enforcement officers to assist them when
entering and inspecting paint, varnish and resin facilities and details the type and location of
emission problems and control equipment. It was written to provide the compliance inspector
with an understanding of the industry sufficient to conduct a proper inspection.
Economics of Emission Control
Specific among the goals of this study was the determination of the financial impact of
air pollution control on the paint and coating industry. To accomplish this, the investigation
was divided into the three following parts:
Cost of Best Control Equipment
Model Plant Economic Study
Projections of Present and Future Industry-wide Costs
Costs of best control equipment were obtained from equipment manufacturers through
the Industrial Gas Cleaning Institute which was retained as a subcontractor. These were obtained
in response to detailed process, operating and bid specifications prepared by Air Resources, Inc.
This information was then summarized and is graphically presented within this section. The
capital cost and total annual operating cost for turnkey systems, catalytic and thermal afterburners
plus auxiliaries and afterburners only are presented. These costs are plotted against size as
measured by gas flow. Separate plots are presented for afterburners with and without heat
exchange.
Model plant costs were developed by the Sherwin-Williams Company who acted as a
subcontractor for this specific purpose. A detailed capital cost of the model plant was made
by their engineering department. Balance sheets and operating statements for both the controlled
and uncontrolled plants were developed by their accounting department. Present total cost and
fifteen-year projected cost of the use of best control by the industry were also developed and
are summarized below:
Total Cost to Industry for Best Control
Capital Investment Cost $16,200,000
Annual Operating Cost 4,500,000
XI
-------
Projected Cost of Control (15-year period)
Capital Investment Cost $30,600,000
Annual Operating Cost 8,500,000
The application of best control technology to the model plant will result in a loss of
income ranging between $0.006 and $0.008 per gallon of paint produced. This is equivalent
to a decrease in stockholder equity of between 1.73 and 3.25 per cent.
Please note that all costs presented in this chapter are based on 1972 monetary
values and do not provide escalation factors for current requirements.
Pigment Industry
For purposes of this study, the pigment industry was included as part of the paint
and varnish industry. General statistics for the industry are presented followed by a review
of the pigments judged to have the most potential for significant air pollution emissions. Major
pigments studied were the cadmium pigments, zinc oxide, the chrome pigments, iron oxide and
titanium dioxide. Of these, the manufacture of titanium dioxide by both the sulfate process
and the chloride process was studied in further detail. A study of emission control technology
was also conducted and includes type and performance of currently used emission control
systems and other potential methods of control.
An industry questionnaire similar in content to the paint and varnish questionnaire was
prepared specifically for the titanium dioxide manufacturers. The results of the questionnaire
are summarized and presented. Estimated emission factors for various operations utilized by
both the sulfate and chloride process were developed and are also presented.
Recommended Research and Development Programs
The deficiencies in both pollution control technology and emission measurements are
treated in detail and recommendations are made for research and development programs which
can lead to improvements in these areas.
Recommended programs for improvement in air pollution control technology are as follows:
1. Design of a closed reactor for resin and varnish cooking.
2. Development of liquid handling of resin cook raw materials.
3. Development of inexpensive solvent scavenging system.
4. Development of transfer system for direct flame incinerator.
XII
-------
5. Determination of best existing control technology currently applied to the manufacture
of TiO2 by the sulfate process.
Recommended program for improvement in emission measurements is as follows:
1. Development of a simple and inexpensive instrument to detect hydrocarbon emission
and measure flow from kettle cooking operations.
XIII
-------
TABLE OF CONTENTS
LIST OF FIGURES xjx
LIST OF TABLES xxij
CHAPTER 1. PROCESS AND EMISSION 1
I. PROCESS 1
A. Detailed Process Description 1
1. History 1
2. Description of Manufacturing Process 2
B. Process Data from Questionnaires 42
1. Raw Materials 45
2. Products and Production 56
3. Process Equipment 58
a. Dispersion and Grinding Equipment 58
b. Solvent Storage Tanks 60
c. Resin Reactor Usage 60
C. Material Balance for Model Plant 63
1. Design Basis 63
2. Production and Inventory 65
3. Equipment Requirement 68
a. Paint Plant 68
b. Resin Plant 73
c. Tankage Requirements 77
4. Raw Materials 77
5. Labor Requirements 81
6. Plant Layout and Flow Sheet 81
II. EMISSIONS 81
A. Description of Emissions 81
1. Fugitive 90
2. Non-Fugitive 91
3. Chemical and Physical Properties 91
B. Sources of Emissions 92
1. Major 92
2. Minor 100
C. Quantities of Emission from Uncontrolled Plants 100
1. Model Plant 100
a. Solvent Emissions from Tanks 101
b. Manufacturing Area 101
c. Resin Production 106
d. Thin Tanks and Filter Presses 115
e. Summary 118
2. Varnish and Resins Production 118
D. Process Operations Influencing Emissions 133
1. Equipment and/or Process Characteristics 133
a. Handling and Storage 133
XIV
-------
b. Paint Operations 134
c. Resin Production 134
2. Raw Materials 135
a. Paniculate 135
b. Gaseous Emissions 135
3. Start-up and Shut Down 136
4. Operation Above and Below Capacity 136
5. Process Operation Upsets 136
E. Raw Data Tabulated 138
1. Questionnaires 138
F. By-Products 138
1. Liquid Wastes 143
2. Solid Waste 143
G. EPA Source Test Data 143
CHAPTER 2. INDUSTRY STATISTICS 151
I. TYPE, SIZE AND LOCATION OF PRESENT DAY PLANTS 151
II. PAST, PRESENT AND PROJECTED INDUSTRY TRENDS TO 1985 156
A. Production 156
B. Number of Plants 161
C. Size of Plants 163
D. Capacity — Production Relations 166
E. Typical Plant and Equipment Ages 166
F. Technological Revolutions and Outside Influences Causing Changes
in the Industry 168
1. Application Techniques 168
2. Pigment Industry 170
3. Environmental and Health Considerations 171
III. DISTRIBUTION OF CAPITAL EXPENDITURES 172
CHAPTER 3. MEASUREMENTS OF EMISSIONS 175
I. SAMPLING AND ANALYTICAL PROCEDURES 175
A. General Requirements for Source Testing 175
B. Description of Source Sampling and Analytical Procedures 178
1. Flow Measurement 178
2. Paniculate Measurements 179
a. General Considerations 179
b. Collection and Analysis Techniques 180
3. Hydrocarbon Analysis 181
a. Discontinuous Sampling 181
4. Analytical Techniques 184
a. Infrared Analyzers 184
b. Combustion Analyzers 184
c. Flame lonization Detection 185
d. Gas Chromatography 185
C. EPA Test Methods 185
xv
-------
II. CONTINUOUS SOURCE MONITORING TECHNIQUES USED BY INDUSTRY .. 191
CHAPTER 4. INVENTORY OF EMISSIONS 193
I. EMISSION FACTORS FOR EACH SOURCE 193
II. EMISSION INVENTORY FOR THE INDUSTRY 197
CHAPTER 5. EMISSION CONTROL TECHNOLOGY 199
I. DESCRIPTION OF BEST CONTROL SYSTEMS 199
A. Control of Participate Emissions 199
B. Control of Gaseous Emissions 202
1. Flame Incineration 202
2. Thermal Combustion 204
3. Catalytic Afterburners 215
II. DESCRIPTION OF EMISSION CONTROL OTHER THAN BEST CONTROL 224
A. Scrubbers 224
B. Vapor Condensation 224
III. METHODS OF CONTROL OTHER THAN ADD ON EQUIPMENT 233
A. Raw Material Substitution 233
1. Solvents 234
2. Pigments and Other Solids 235
B. Changes in Process or Operating Conditions 235
IV. PERFORMANCE OF CURRENTLY USED METHODS OF EMISSION
REDUCTION 235
A. Performance Data 235
1. Scrubbers 236
2. Afterburners 236
3. Fabric Filters 242
B. Operating Life and Maintenance Experience for Control Systems 242
V. CAPABILITY TO MEET MORE STRINGENT STANDARDS 244
A. Fabric Filters 249
B. Afterburners 249
C. Scrubbers 249
D. Refrigerated Condensers 250
VI. WATER AND SOLID WASTE PROBLEMS ASSOCIATED WITH
BEST CONTROL 250
CHAPTER 6. INSPECTION PROCEDURES 251
I. NATURE OF SOURCE PROBLEMS 251
II. PROCESS DESCRIPTION 252
A. Paint Manufacturing 252
B. Varnish Cooking 254
C. Resin Manufacturing 257
1. Oils or Fatty Acid 259
2. Polyols 259
3. Acids and Anhydrides 261
D. Air Pollution Control Techniques 265
III. INSPECTION POINTS 272
XVI
-------
A. Raw Material Handling and Storage 275
B. Manufacturing 275
C. Filling and Packaging 279
CHAPTER 7. ECONOMICS OF EMISSION CONTROL 281
I. COST OF BEST CONTROL EQUIPMENT 281
A. Thermal Afterburners 283
B. Catalytic Afterburners 284
C. Fabric Collectors 284
II. MODEL PLANT STUDY 285
A. Capital Cost of Plant 285
B. Balance Sheet and Operating Statement 285
C. Balance Sheet and Operating Statement for Controlled Plant 288
D. Cost of Control Other Than Add-on Equipment 289
1. Raw Material Substitution 289
2. Process Modification 290
E. Varying Types and Levels of Control 291
F. Impact on Income, Cash Flow and Investment 291
III. INDUSTRY WIDE STUDIES 294
A. Present Total Cost to Industry to Meet Best Control Requirements 294
B. Fifteen Year Projection of the Cost of Control 295
C. Sources of Capital for Pollution Control 296
D. Industry Structure 296
E. Product Elasticity — Production Substitution 297
CHAPTER 8. PIGMENT INDUSTRY 361
I. INTRODUCTION 361
A. Classification and Statistics 361
B. Purpose and Scope 365
II. REVIEW OF MAJOR PIGMENTS 366
A. Cadmium Pigments 366
B. Zinc Oxide 369
C. Chrome Pigments 376
1. Lead Chromes 376
2. Zinc Chromes 377
3. Chrome Green 377
4. Chromium Oxide Greens 377
5. Molybdate Orange 377
D. Iron Oxides 378
E. Titanium Dioxide Pigments 381
1. Sulfate Process 383
2. Chloride Process 393
3. Industry Statistics — Questionnaires 397
a. Products and Raw Materials 397
b. Process Equipment 399
III. EMISSIONS 399
XVII
-------
A. Description of Emission 399
B. Source of Emissions 404
1. Sulfate Process 404
2. Chloride Process 406
C. Measurement of Emissions 406
D. Raw Data Tabulated 407
1. Questionnaires 407
a. Digestors 407
b. Sulfate Calciners 408
c. Drying and Milling 408
d. Chlorinators 408
e. Summary of Emission Factors 415
2. Other Sources 415
IV. EMISSION CONTROL TECHNOLOGY 421
A. Description of Currently Used Control Systems 421
B. Other Methods of Control 421
C. Performance of Currently Used Control Systems 421
CHAPTER 9. RECOMMENDED RESEARCH AND DEVELOPMENT PROGRAMS 425
I. EMISSION CONTROL TECHNOLOGY 425
A. Technical Developments for Reduced Levels of Emission 425
1. Process Chemistry and Kinetics 425
2. Process Equipment and/or Operations 426
3. Control Equipment and/or Operations 426
B. Economic Deficiencies Preventing Reduced Levels of Emissions 427
C. R&D Priorities to Improve Control Technology 427
D. Recommended Programs for Achieving R&D Requirements 428
1. Closed Reactor Design Program 428
2. Investigation of Methods for Handling Liquid Resin Compounds 429
3. Development of Inexpensive Solvent Scavenging Systems 430
4. Development of Transfer System for Direct Flame Incineration 431
II. MEASUREMENT OF EMISSIONS 431
A. Deficiencies in Manual Methods of Source Sample Collection and Analysis .. 432
B. Deficiencies in Techniques and/or Equipment for Continuous Monitoring
of Source Emissions 432
C. R&D Priorities to Improve Measurement Techniques 432
D. Recommended Programs for Achieving R&D Requirements 433
III. PIGMENT INDUSTRY 435
REFERENCES 439
LIST OF STANDARD ABBREVIATIONS 442
APPENDICES
BIBLIOGRAPHIC DATA SHEET AND ABSTRACT
XVIII
-------
LIST OF FIGURES
Page No.
Figure 1 Paint Manufacturing Using Sand Mill for Grinding Operation 3
Figure 2 Typical Varnish Cooking Room 9
Figure 3 Modern Resin Production System of Solvent and Fusion Cooks 25
Figure 4 Materials Flow Sheet for Paint Manufacturing 41
Figure 5 Flow Sheet for Solvent Through a Paint-Resin Plant 62
Figure 6 Typical Sales and Production Curves — Trade Sales Products 69
Figure 7 Typical Inventory Curve — Trade Sales Products 70
Figure 8 Grinding Capacity — Ball and Pebble Mills 74
Figure 9 Modern Resin Production System 78
Figure 10 Model Plant — Factory Manning Chart 85
Figure 11 Materials Flow Sheet for Model Paint Plant 86
Figure 12 Model Paint Plant 87
Figure 13 Model Paint Plant 88
Figure 14 Model Plant — Area Plot Plan 89
Figure 15 Emission Points from Raw Material Storage Tanks of Model Plant 102
Figure 16 Emission Points from Production Areas of Model Plant 103
Figure 17 Emission Characteristics — Short Oil Fusion Cook 110
Figure 18 Emission Characteristics — Long Oil Solvent Cook 113
Figure 19 Hydrocarbon Emission for 500 Gallon, Closed Kettle 122
Figure 20 Hydrocarbon Emission for 500 Gallon, Closed Kettle 123
Figure 21 Hydrocarbon Emission for 1,000 Gallon, Closed Kettle 124
Figure 22 Hydrocarbon Emission for 1,000 Gallon, Closed Kettle 125
Figure 23 Hydrocarbon Emission — 2 Batches Monitored 126
Figure 24 Hydrocarbon Emission for 2,500 Gallon, Closed Kettle 127
Figure 25 Emissions from 100 Gallon Epoxy Reactor Solvent Pressure Cook
— 30 psig 129
Figure 26 Emission from 2,000 Gallon Reactor 130
Figure 27 Emission from 1,500 Gallon Polyester Reactor Fusion Cook 131
Figure 28 Emission from Two-1,000 Gallon Reactors 132
Figure 29 U.S. Paint & Varnish Industry Distribution by Product Class 152
Figure 30 U.S. Paint & Varnish Industry Distribution by Production Value Size Class 154
Figure 31 U.S. Paint & Varnish Industry 1971 Distribution of Plants & Employees by
Plant Size 155
Figure 32 U.S. Paint & Varnish Industry Distribution by State Estimated 1972
Production 158
Figure 33 U.S. Paint & Varnish Industry Shipments (Actual & Forecast) —
Million Gallons 159
Figure 34 U.S. Paint & Varnish Industry Shipments (Actual & Forecast) —
Million Dollars 160
Figure 35 U.S. Paint & Varnish Industry Distribution by State (Actual & Projected) 162
Figure 36 U.S. Paint & Varnish Industry Distribution of Plants by Size —
Number of Employees 164
XIX
-------
Page No.
Figure 37 U.S. Paint & Varnish Industry Capital Expenditures — Million Dollars 167
Figure 38 Particulate — Sampling Train 182
Figure 39 Schematic Flow Diagram of Flame lonization Detector 186
Figure 40 Hydrocarbon Monitoring System 187
Figure 41 Geographical Distribution of Organic Emissions from Paint Production 1972 ... 198
Figure 42 Pigment Emission Control System 200
Figure 43 Schematic Diagram of a Flame Incineration Unit 203
Figure 44 Schematic of Typical Thermal Afterburner and Control System 205
Figure 45 Residence Time vs. Temperature at 95% Conversion and 2 Btu/SCF 208
Figure 46 Conversion vs. Temperature at 0.6 sec Residence Time and 2 Btu/SCF 209
Figure 47 Conversion vs. Inlet Concentration at 0.6 sec Residence Time
and 1350°F Outlet 211
Figure 48 Schematic of Typical Thermal Afterburner and Control System 213
Figure 49 Schematic Diagram of a Catalytic Oxidation System for Varnish Kettles 216
Figure 50 Catalytic Oxidation Rates for Solvents 218
Figure 51 Reaction Rate Constants for Low, Intermediate, and High Temperatures 222
Figure 52 Comparison of Thermal and Catalytic Reaction Rates for Maleic Anhydride .... 223
Figure 53 Refrigerated Condenser System for Control of Resin Kettles 225
Figure 54 Refrigeration System 227
Figure 55 Two Stage Condenser System 228
Figure 56 Emission Characteristics — Kettle D 232
Figure 57 Paint Manufacturing Using Sand Mill for Grinding Operation 253
Figure 58 Materials Flow Sheet for Paint Manufacturing 255
Figure 59 Typical Varnish Cooking Room 258
Figure 60 Modern Resin Production System 266
Figure 61 Pigment Emission Control System 268
Figure 62 Thermal Afterburner System for Resin Reactor or Closed Kettle 270
Figure 63 Schematic Diagram of a Catalytic Afterburner System for Varnish Kettles 271
Figure 64 Capital Costs for Catalytic Afterburners Without Heat Exchange 327
Figure 65 Capital Costs for Catalytic Afterburners With (23% Efficient) Heat Exchange ... 328
Figure 66 Capital Costs for Thermal Afterburners Without Heat Exchange 329
Figure 67 Capital Costs for Thermal Afterburner With (42% Efficient) Heat Exchange 330
Figure 68 Total Installed Costs for Thermal and Catalytic Afterburners 331
Figure 69 Direct Annual Operating Costs for Thermal and Catalytic Afterburners
Without Heat Exchange 332
Figure 70 Total Annual Operating Cost for Thermal and Catalytic Afterburners
Without Heat Exchange 333
Figure 71 Direct Annual Operating Cost for Thermal and Catalytic Afterburners
With Heat Exchange 334
Figure 72 Total Annual Operating Cost for Thermal and Catalytic Afterburners
With Heat Exchange 335
Figure 73 Model Plant Factory Manning Chart 350
xx
-------
Page No.
Figure 74 Flow Diagram for Cadmium Sulfide Production 368
Figure 75 Flow Sheet for French Process 371
Figure 76 Flow Sheet for American Process 372
Figure 77A TiO2 Manufacture Sulfate Process 386
Figure 77B TiO2 Manufacture Sulfate Process 387
Figure 78 TiO2 Manufacture Chloride Process 394
Figure 79 Exhaust Rate from Sulfate Process Digestion Tanks 418
Figure 80 Calciner Emission Control System 419
XXI
-------
LIST OF TABLES
Page No.
Table 1 Varnish Raw Materials 8
Table 2 Questionnaire Distribution 44
Table 3 Geographical Distribution of Questionnaires 46
Table 4 Percentage Distribution of Questionnaires, Paint Plants, and Value of
Shipments 47
Table 5 Industry Questionnaire — Tabulation Summary 48
Table 6 Industry Questionnaire — Tabulation Summary 49
Table 7 Solvent Usage 54
Table 8 Drying Agents and Mercury Compounds 57
Table 9 Mills, Mixers, Etc 59
Table 10 Solvent Tanks Over 5,000 Gallons 61
Table 11 Filling Losses for Selected Solvents @ 20°C 61
Table 12 Reactor Usage, Gallons of Kettle Volume 64
Table 13 Resin Processing, Number of Plants 64
Table 14 Model Plant — Product Mix 66
Table 15 Model Plant — Trade Sales Color Distribution 66
Table 16 Model Plant — Product Type 67
Table 17 Model Plant — Finished Goods Inventory — Maximum Projection 71
Table 18 Model Plant — Equipment Specifications 72
Table 19 Model Plant — Paint Formulations for Alkyd Type Coatings 75
Table 20 Model Plant — Summary of Tankage Requirements 79
Table 21 Model Plant — Annual Raw Material Consumption 82
Table 22 Model Plant — Annual Package and Package Material Requirements 83
Table 23 Model Plant — Labor Requirements 84
Table 24 Composition of Oil and Varnish Emissions 93
Table 25 Odor and Composition (by Functional Groups) of Oil and Varnish Emissions ... 94
Table 26 Solvent Characteristics 96
Table 27 Maximum OSHA Allowable Concentration Limits 97
Table 28 Odor Thresholds of Some Organic Vapors 98
Table 29 Particle Size Range of Various Pigments and Extenders 99
Table 30 Operating Parameters for Model Plant Storage Tank Emission Sources 104
Table 31 Emissions During Filling for Model Plant Storage Tanks 105
Table 32 Contaminant Levels in Various Parts of Model Paint Plant 107
Table 33 Emissions from Model Plant Production Area Exhaust 108
Table 34 Summary of Emissions from Model Plant 500 Gallon Fusion Reactor
and 1,500 Gallon Solvent Reactor 114
Table 35 Emission Data Summary 116
Table 36 Emission Summary for Gaseous Contaminants — Model Plant 117
Table 37 Source Test Summary for Paint Plant 121
Table 38 Emission Data from Questionnaires — Paint and Resin Plants 139
Table 39 Emission Data from Questionnaires — Paint Plants 140
Table 40 Emission Data from Questionnaires — Resin Plants 141
XXII
-------
Page No.
Table 41 Questionnaire Summary of Waste Material Disposal — Odor Complaints
from Questionnaires 142
Table 42 Summary of Tested Batches 145
Table 43 U.S. Paint & Varnish Industry Distribution of Plants by Employee Size Class ... 165
Table 44 U.S. Paint, Varnish and Inorganic Pigments Industry 1967 Labor and
Finance Summary 173
Table 45 Common Raw Materials and Solvents Used in the Manufacture of Resins 176
Table 46 Solvent Boiling Ranges 177
Table 47 Emission Factors from Selected Paint Plants • 194
Table 48 Emission Factors from Selected Resin Plants 195
Table 49 Emission Factors from Selected Plants Producing Coatings and Resins 196
Table 50 Kettle A — Alkyd Cook 230
Table 51 Kettle D — Polyester 231
Table 52 Type 1 Plants — Air Pollution Control — Loading Mills, Etc. — Fabric
Filters — Scrubbers 237
Table 53 Type 1 Plants — Air Pollution Control — Reactors and Kettles — Scrubbers ... 239
Table 54 Type 3 Plants — Air Pollution Control — Reactors and Kettles —
Afterburners Scrubbers 240
Table 55 Air Pollution Control — Reactors and Kettles — Thermal Afterburners —
Catalytic Afterburners 241
Table 56 Type 2 Plants — Air Pollution Control — Loading, Mills, Etc. — Fabric
Filters 243
Table 57 Type 3 Plants — Air Pollution Control — Loading, Mills, Etc. — Mechanical
Fabric Filters 245
Table 58 Test Result of Control Equipment Efficiencies 246
Table 59 Thermal Afterburner Test Data 247
Table 60 Operating Life and Maintenance Requirement for Air Pollution Control
Equipment 248
Table 61 Varnish Raw Materials 256
Table 62 Classification of Typical Solvents 276
Table 63 Filling Losses for Selected Solvents @ 20°C 277
Table 64 Specifications for Abatement Equipment 299
Table 65 Instructions for Submitting Cost Data 302
Table 66 Thermal Afterburner Process Description for Resin Reactor Specification 304
Table 67 Thermal Afterburner Operating Conditions for Resin Reactor Specification
(Without Heat Exchange) 306
Table 68 Thermal Afterburner Operating Conditions for Resin Reactor Specification
(Without Heat Exchange) 307
Table 69 Thermal Afterburner Operating Conditions for Resin Reactor Specification
(With Heat Exchange) 308
Table 70 Thermal Afterburner Operating Conditions for Resin Reactor Specification
(With Heat Exchange) 309
XXIII
-------
Page No.
Table 71 Thermal Afterburner Process Description for Open Kettle Specification 310
Table 72 Thermal Afterburner Operating Conditions for Open Kettle Specification
(Without Heat Exchange) 312
Table 73 Thermal Afterburner Operating Conditions for Open Kettle Specification
(Without Heat Exchange) 313
Table 74 Catalytic Afterburner Process Description for Resin Reactor Specification 314
Table 75 Catalytic Afterburner Operating Conditions for Resin Reactor Specification
(Without Heat Exchange) 316
Table 76 Catalytic Afterburner Operating Conditions for Resin Reactor Specification
(Without Heat Exchange) 317
Table 77 Catalytic Afterburner Operating Conditions for Resin Reactor Specification
(With Heat Exchange) 318
Table 78 Catalytic Afterburner Operating Conditions for Resin Reactor Specification
(With Heat Exchange) 319
Table 79 Catalytic Afterburner Process Description for Open Kettle Specification 320
Table 80 Catalytic Afterburner Operating Conditions for Open Kettle Specification
(Without Heat Exchange) 322
Table 81 Catalytic Afterburner Operating Conditions for Open Kettle Specification
(Without Heat Exchange) 323
Table 82 City Cost Indices 324
Table 83 Average Hourly Labor Rates by Trade 325
Table 84 Installation and Operating Cost for Baghouse 326
Table 85 Model Plant Annual Raw Material Costs 336
Table 86 Model Plant Annual Package & Package Material Costs 338
Table 87 Model Plant Annual Wage & Salary Cost of Plant 339
Table 88 Model Plant Depreciation Schedule 341
Table 89 Model Plant Income Statement 342
Table 90 Model Plant Production Schedule 1.9 Million Gallons Output 344
Table 91 Model Plant Balance Sheet 346
Table 92 Model Plant Cash Flow Statement 347
Table 93 Model Plant Return on Investment 348
Table 94 Model Plant 1.9 Million Gallon Finished Output 349
Table 95 Model Paint Plant Cost, Dollars 351
Table 96 Model Paint Plant Cost, Dollars 352
Table 97 Model Controlled Plant Depreciation Schedule 353
Table 98 Model Controlled Plant Income Statement 354
Table 99 Model Controlled Plant Balance Sheet 356
Table 100 Model Controlled Plant Cash Flow Statement 357
Table 101 Model Controlled Plant Return on Investment 358
Table 102 Model Controlled Plant Air Emission Control Devices 359
Table 103 Pigment Production by Major Type 363
XXIV
-------
Page No.
Table 104 Major White Pigments 363
Table 105 Production of Organic Pigments 364
Table 106 Production of Major Inorganic Color Pigments 364
Table 107 Zinc Oxide Producers 375
Table 108 Analysis of llmenite Ores 384
Table 109 TiO2 Industry Questionnaire Tabulation Summary 1972 398
Table 110 TiO2 Industry Questionnaire Production — Raw Materials Inventory
Five TiO2 Plants 400
Table 111 TiO2 Industry Questionnaire Sulfate Process Equipment 401
Table 112 TiO2 Industry Questionnaire Chloride Process Equipment 402
Table 113 TiO2 Industry Questionnaire Mills, Etc 403
Table 114 Sulfate Process Emissions — Drying and Milling 409
Table 115 Chloride Process Emissions — Drying and Milling 410
Table 116 Chlorinator Emissions After Control 412
Table 117 Chlorinator Emissions Before Control 413
Table 118 Plant 6 — Emission Inventory 414
Table 119 Estimate of Emission Factors for Uncontrolled Processes 416
Table 120 TiO2 Industry Questionnaire Particulate Control Devices 422
Table 121 TiO2 Industry Questionnaire Emission Control Devices 423
xxv
-------
CHAPTER 1
PROCESS AND EMISSION
In this chapter a discussion of the process operations used in the manufacture of paint,
varnishes and resins will be presented. The type, quantity and source of emissions evolved from
these processes will also be presented.
I. PROCESS
This section will cover process operations only. The discussion will include a brief history of
the industry, a description of the various manufacturing processes, the influence of process operation
on emissions and detailed material balances. The material balance will be based on the model plant
developed for the industry.
A. Detailed Process Description
J^ History — The paint and varnish industry is one of the oldest manufacturing industries in the
United States. The industry is made up of about 1,500 companies operating about 1,700 plants.1 The
industry is well distributed geographically throughout the country and production volume is definitely
related to density of population. Even though about 36 companies account for about 64% of the
total sales, the industry is one of the few remaining which contains numerous small companies that
specialize in a limited product line to be marketed within a geographical region. There are fewer than
20 companies that sell paint nationwide.
The industry is now emerging as a scientific business from its beginning as an art, 50 years
ago. Even with rapid growth in technology, the industry process techniques still are not well defined
and not only vary from one producer to another but there is also variation in the techniques used by a
single producer. To add further complication, the industry is technically one of the most complex of
the chemical industry. A plant that produces a broad line of products might utilize over 600 different
raw materials and purchased intermediates. These materials can be generally classified in the following
categories: oils, metallic driers, resins, pigments, extenders, plasticizers, solvents, dyes, bleaching
agents, organic monomers for resins, and additives of many kinds.
-------
The industry produces an equally large number of finished products which are generally
classified as trade sale finishes, maintenance finishes, and industrial finishes.
Trade sale products are stock-type paints generally distributed through wholesale-retail
channels and packaged in sizes ranging from 1/2 pint to 1 gallon. A subdivision of trade sale products
are maintenance finishes which are used for the protection and upkeep of factories, buildings, and
structures such as bridges and storage tanks. Since they are usually stock type, they come under
the Department of Commerce definition of trade sales.
The other major type of paint products are industrial finishes which are generally defined as
those applied to manufactured products. These finishes, such as automotive, aircraft, furniture,
and electrical finishes are usually specifically formulated for the using industry. Within these major
product lines there are literally thousands of different products for many different applications and
types of customers. Trade sale finishes and industrial finishes are produced in almost equal volume
with the production for 1972 estimated at 465 million gallons for trade sales and 485 million gallons
for industrial finishes. Trade sales, however, are estimated to account for 55% of the dollar sales or
about $1,715 million dollars.1
2. Description of Manufacturing Process — Starting with all purchased raw material, the manu-
facturing process for pigmented products appears simple from a schematic viewpoint. Basically, it
consists of mixing or dispersing pigment and vehicle to give the final product. This process is
schematically illustrated in Figure 1.
The paint vehicle is defined as the liquid portion of the paint and consists of volatile solvent
or dispersing medium and non-volatile binder such as oils and resins. The non-volatile portion is
also called the vehicle solid or film former. The pigment portion of the paint consists of hiding pigments
such as titanium dioxide (TiO2), extenders such as talc or barium sulfate, colorants, and any mineral
matter used for flatting or other purposes.
Also included in the non-volatile portion of the coatings are various additives present in small
quantities for a variety of reasons. From an environmental standpoint the most important of these are
the preservatives and the driers. The preservatives are fungicides and mildewcides which sometimes
contain mercury compounds. While they are present in small quantities, they have nevertheless
become of interest to health and regulatory agencies. It is anticipated that the use of mercury will
be substantially eliminated in the next few years.
The driers are catalysts whose purpose is to promote the oxidative cross-linking of the resins
and/or oils used in coatings. Without such catalysts many of the coatings in use today would not be
-------
0
AZ?U
I®
2 3 Id
LJU?
u
It
0
s
££
Id
X
!
_l
N
C
r
\
0
0
flu
Dl
0°
zo
Jo
-------
possible since they would never dry completely, or at least would dry too slowly to be of practical value.
The common driers are organic acid salts of lead, cobalt, calcium, zinc, manganese, or zirconium.
Lead naphthenates have historically been the most often used but recent regulations limiting lead to
0.5% of the dried film and possibly more stringent future regulations have prompted some degree of
reformulation of the drier systems.
The incorporation of the pigment in the paint vehicle is accomplished by a combination of
grinding and dispersion or dispersion alone. When it is necessary to further grind the raw pigment,
pebble or steel ball mills are normally used. With the advent of fine particle grades of pigment and
extenders, as well as the wide spread use of wetting agents, the trend is toward milling methods
that are based on dispersion without grinding. Dispersion consists of breaking up of the pigment
clusters and agglomerates, followed by wetting of the individual particles with the binder or vehicle.
Some of the more popular methods currently being used are high-speed disc impellers, high speed
impingement mills and the sand mill.
Some typical paint formulas, both simple and complex, are listed below and on the following
page.
WHITE ACRYLIC BAKING ENAMEL*
Ib gal
Rutile TiC-2 220.3 6.64
Thermosetting acrylic resins (50% NVM) 444.1 54.20
Melamine resin (100% NVM) 95.2 9.52
Catalyst for melamine 6.63 0.87
Xylol 184.0 25.51
Ethylene glycol monoethyl ether acetate 26.4 3.26
976.3 100.00
'Formulas courtesy of Ashland Chemical Company.
-------
WHITE GLOSS INTERIOR ENAMEL*
Ib gal
Rutile TiO2 300.0 9.02
Suspension and flow agent 8.0 1.15
Long-oil, tall oil alkyd (70% NVM) 536.0 67.0
Mineral spirits 150.0 23.08
6% Co naphthenate 4.4 0.56
4% Ca naphthenate 9.4 1.25
6% Zr drier catalyst 12.5 1.75
Antiskinning agent 1.0 0.13
1,021.3 103.94
POLYVINYL ACETATE EMULSION WALL PAINT*
Grind (high speed mill) Ib gal
Water 275.0 33.05
KTPP, dispersant 2.0 0.10
R&R-551 soya lecithin dispersant 8.0 0.92
Tergitol® NPX, dispersant 2.0 0.23
Nopco® NOW, anti-foamer 0.5 0.07
Rutile titanium dioxide 200.0 5.94
Diatomaceous silica 45.0 2.34
Clay 130.0 6.05
Calcium carbonate 100.0 4.43
Sub-total 762.5 53.13
Reduction
Methocel® 65HG (31/2% solution) 120.0 14.42
Carbitol® acetate, coalescing agent 25.0 2.98
Nopco® NOW, anti-foamer 0.5 0.07
PVAc emulsion 224.0 24.62
Water 10.0 1.20
Pre-mix
PMS-30, mildewcide 0.3 —
Water 30.0 3.61
Total 1,172.3 100.03
'Formulas courtesy of Ashland Chemical Company.
5
-------
Aside from the dispersion step, pigment paint manufacturing involves handling of raw material
as well as handling and packaging of finished product.
Air pollutant emissions are primarily the fugitive type and consist of evaporation losses of
the volatile portion of the vehicle from the milling operation and from various product holding tanks
and packing stations. There are also some fugitive paniculate emissions that result from handling
and emptying of pigment or extender bags into the grinding and dispersion mills. In some plants
these loading areas are hooded and bags and pigment dusts are passed to a central collection
station. At this station bags are removed for refuse disposal and the pigment dust is collected in a
fabric filter and recycled into primer or other dark paint mixes.
Some of the larger and a few of the medium manufacturers make a significant amount
of their formulation ingredients, such as pigments, resins, and modified oils. Some manufacturers
produce these ingredients in an amount exceeding their requirements and sell the excess to other
manufacturers. A significant number also make only a portion of their resins and purchase the
remainder from their competitors or suppliers who specialize in resin manufacturing.
The manufacturing of resins and varnishes is by far the most complex process in a paint
plant, primarily as the result of the large variety of different raw materials, products and cooking
formulas utilized. One manufacturer has reported that he has over 9,300 active resin formulas. Of
these 600 to 700 are in general monthly use while the remainder are produced on an intermittent
basis. Furthermore, new formulas are added at the rate of 5 to 10 per month. While this plant is not
typical since it produces a more diverse product line than most, it nonetheless serves to illustrate
the wide diversity found in the industry. This diversity is the source of the difficulty that is encountered
in attempting to characterize the industry or in making generalized statements about emissions, control
technology, or any other subject for that matter.
The complexity begins with the nomenclature used in classification of the final product.
Originally, varnishes were all made from naturally occurring material and they were easily defined as
a homogeneous solution of drying oils and resins in organic solvents. As new synthetic resins were
developed, the resulting binders or varnishes were classified on the basis of the resins used.
Examples of this are alkyd, epoxy, and polyurethane resins. In an attempt to simplify nomenclature
the industry is attempting to adopt the term "clear coating" to cover some varnishes, resins and
lacquers.
There are two basic types of varnishes, spirit varnishes and oleoresinous varnishes.2 Spirit
varnishes are formed by dissolving a resin in a solvent and they dry by evaporation of the solvent.
-------
The dry film formed undergoes no substantial change in the process of drying and is classified as
non-convertible. A good example of this type of varnish is shellac which is a mixture of a natural
resin and alcohol.
Materials that might fall in the general category of spirit varnishes, that use nitro-cellulose
or similar compounds for their basic film former, have by common practice been termed lacquer. These
are defined as a colloidal dispersion or solutions of nitro-cellulose, or of similar film-forming compounds,
with resins and plasticizers in solvents and diluents which dry primarily by solvent evaporation. One
of the advantages and general characteristics of spirit varnishes is that upon recoating, the film
will soften and partially redissolve so that patched areas show no sharp line of demarcation.
Oleoresinous varnishes, as the name implies, are solutions of both oils and resins. These
varnishes dry by solvent evaporation and by reaction of the non-volatile liquid portion with oxygen in
the air to form a solid film. They are classified as oxygen convertible varnishes and the film formed
on drying is insoluble in the original solvent. A summary of the various types of material used in the
production of classical Oleoresinous varnishes is given in Table 1.
Varnish is cooked in both portable kettles and large reactors. Kettles are used only to a
limited extent and primarily by the smaller manufacturers. The very old, coke fired, 30 gallon capacity
copper kettles are no longer used. The varnish kettles which are used, have capacities of 150 to 375
gallons. These are fabricated of stainless steel, have straight sides and are equipped with three or
four-wheel trucks. Heating is done with natural gas or fuel oil for better temperature control. The
kettles are fitted with retractable hoods and exhaust pipes, some of which may incorporate solvent
condensers. Cooling and thinning is normally done in special rooms. A typical varnish production
operation is illustrated in Figure 2.
All of the resin manufacturers and most of the large and medium size paint manufacturers
utilize large indirectly heated closed reactors for varnish production. These reactors will be described
in more detail in later report sections covering modern resin manufacturing.
The manufacturing of Oleoresinous varnishes is somewhat more complex than for spirit
varnishes. Manufacture of this product consists of the heating or cooking of oil and resins together
for the purpose of obtaining compatability of resin and oil and solubility of the mixture in solvent as
well as for development of higher molecular weight molecules or polymers.
The time and temperature of the cook are the operating variables used to develop the desired
end product polymerization or "body". The chemical reactions that occur are not well defined. The
resin is a polymer before cooking and may or may not increase in molecular size during the cook.
-------
§
^
•a
<
c
a,
CO
O!
C
'E
c
-*
.i
<
,_:
0)
o
V]
J2
CO
4-*
0)
o
'>
h_
=>
CN
C/l
1
0)
c
'E
CD
iZ
00
CM
CO
•o
co
0>
C
CO
O)
c
CO
O
U
C*
I I
•| 8
O N
•* Lfi
DC
01
"•*.
«er (A
w ^ £
-155
5 < >
£ * &
CO
DC
>
v>
'L.
'a
CO
aj
c
5
»"
V)
C
'35
0)
DC
CO
D
CO
Z
CO
.c
a.
CO
z.
a_
68
>
CNI
1
8
c
0}
V)
O
Li-
ed
O
"o
1-
00
CO
C
CO
O
a
D
CO
^
0
X
•*'
^
(U
'tn
1/1
£
1
^i
c
1
0)
1
0)
1-
LO
D"
-0.
man! la, boe
eastindia
c
1
a
o
b
CO
§
0)
DC
o'
v>
0)
•a
'o
o
dammar, ac
'E
L^
0)
IU
~
sandaral ma
shellac
c
'in
O
E
i,
'wi
O
DC
-6
'6
1
c"
2
o
o
c
1
o
"o
s 1
eu c
c •» .E
8 oc iS
""- .- .y
C *^ QJ
'vt f ~ca
O ^ ^
»- c ^
>.
(O to
0
-o
(O
c
U
c
N
ja
oo
o
o
•D
CO
I
o>
c
3
H
co'
CO
U
'u
'4-"
b
.0"
CO
•D
>
CD
Q £
d CO
13
CD
0)
l/l
C
13
CO
0)
1
U
<4~
H—
c^
i!
CO
1
6
^
tn
IT
-o
S 1 £
"8
_>
1
E
CD
6
c
CD
E
CO
IU
Maleic
-------
0
0
a:
0
z
x:
8
o: i
3 (fl
? z
U. (t
1
J
0
Q.
>
h
oz
-------
The resin may react with the oil to produce copolymers of oil and resin, or it may exist as a homogeneous
mixture or solution of oil homopolymers and resin homopolymers.
In special cases it is possible to produce oleoresinous varnish without the application of heat
by simply dissolving or "cold cutting" the resin and heat bodied oil in the solvent. Heat bodying or
polymerization of an oil is done to increase its viscosity and is carried out in a kettle in a fashion
similar to varnish cooking. The fundamental reaction that occurs is polymerization of the oil monomers
to form dinners with a small portion of trimers.
It is possible to blend resins and heat bodied oil and obtain the same varnish that can be
produced by cooking the resin and the unbodied oils. This indicates that copolymerization is not
the fundamental reaction in varnish cooking.
A fundamental variable that determines the final property of a varnish is the proportion of
oil to resin. It is expressed as gallons of oil to 100 pounds of resin. On this basis, varnish is generally
classified as follows:
Short oil: 5 to 15 gallons (28 to 54% oil*)
Medium oil: 16 to 30 gallons (54 to 70% oil*)
Long oil: 30 to 60 gallons (70 to 86% oil*)
'Percentages based on an oil of approximately 7.8 Ib/gal.
A myriad of formulations are used in varnish manufacture. Typical varnish cooking formulations
are given below:2
20 GALLON CAN COATING VARNISH — STRAIGHT COOK
Ib gal
Amberol M-88 modified phenolic resin* 100 11.24
Castung® 403 Z3 dehydrated castor oil** 160 20.00
Kadox® 25, ultra-fine zinc oxide*** 3 —
Manganese naphthenate (6% Mn) 2 0.25
Mineral spirits 235 36.70
Theoretical yield 500 68.19
* Rohm and Haas Co.
** Baker Castor Oil Co.
*** New Jersey Zinc Co.
10
-------
Procedure
Heat resin and oil to 580°F (304°C) in 45 minutes. Add zinc oxide and hold 580°F about 30
minutes to body (50 seconds Ford No. 4 Cup at 52% solids in mineral spirits). Cool quickly by
water spray to 450°F (232°C) add thinner and drier and filter carefully.
Constants
Solids 52% minimum
Viscosity 50 to 60 sec. in Ford No. 4 Cup (about G-l)
Color, Gardner 1933 14 to 18
wt/gal 7.4 Ib
The above formula is made in portable kettles. High temperature is necessary to develop
maximum resistance, yet cooling must be fast to avoid overpolymerization. It is probably impossible
to handle batches larger than 300 gallons safely.
PHENOLIC RESIN SPAR VARNISH* — DISPERSION METHOD
Ib gal
Phenolic resin CKM 5254 100.0 10.0
Alkali refined linseed oil 85.0 11.0
Tung oil 164.0 21.0
Refined castor oil 8.0 1.0
Mineral spirits 202.0 31.0
Dipentene 26.9 3.8
Turpentine 26.9 3.8
n-Butyl alcohol 13.45 2.0
Theoretical yield 626.25 83.6
Procedure
Heat the resin and linseed oil to 560°F (293°C) for 30 minutes for body. Check with tung oil.
(Reduces to about 350°F.) Reheat to 450°F (232°C) in 15 minutes for body, add castor oil, thin and
filter.
Properties
Viscosity C to D
Solids 57%
Color, Gardner 1933 11
wt/gal 7.57 Ib
'Formula courtesy of Union Carbide Corporation, Chemicals and Plastics
11
-------
Driers (on oil)
Lead 0.1 %
Cobalt 0.01%
Manganese 0.005%
INSITU GENERAL PURPOSE VARNISH — UNIPHASE METHOD2
Ib gal
Alkali refined soya oil 23.7
Hydrated lime paste* grams 32
Pentaerythritol 19
Glycerol 21
Tall oil rosin 135
Maleic anhydride 38
Mineral spirits 372 58.0
Activ 8** fl oz 2
Calculated net yield, allowing for
loss in processing 100.0
Procedure
Maintain CO2 over batch throughout the cook. Heat soya oil to 400°F (204°C), add lime
paste, gain 450°F (232°C). Add the pentaerythritol slowly, keeping the temperature at 425 to 450°F
for 45 minutes. Open condenser, close stack, and heat to 470°F (243°C). Hold at 460 to 470°F. Add
the glycerol through the funnel, keeping the temperature above 425°F. Heat to 460°F, and hold
460 to 470°F for solubility of at least 2:1 in methanol. Cool to 400°F by circulating cold Dowtherm
through the jacket. Open stack and close the condenser. Add the rosin. After the rosin has melted,
add the maleic anhydride. Heat to 520°F and hold for a reduced viscosity of I-J at 50% solids in
mineral spirits. Drop to 480°F (249°C) and hold for a reduced viscosity of T-U. Cool, thin and filter.
The major problem in varnish cooking is control of cooking time and temperature to yield
* Hydrolysis catalyst, a paste of hydrated lime in soya oil.
** An organic drying catalyst — R. T. Vanderbilt Co. It supplements metallic driers; does not
replace them.
12
-------
the desired product and prevention of over bodying or gelation. This is done by a variety of cooking
methods and control techniques. The main methods for cooking modern oleoresinous varnishes are
listed below and the technique used is primarily a function of the type of ingredients employed.
(1) Regular or straight cook.
(2) Oil-checked method.
(3) Resin-checked method.
(4) Dispersion method of split-cook.
(5) Insitu or uniphase method.
These various methods are used to prevent gelation and/or improper dispersion. Methods (1) and
(4) apply when cooking varnishes from the material listed in Table 1. They are suitable for cooking
in the small portable kettles discussed earlier. Control of these cooks is done by examination of a
cold pill on glass. The criteria used are the clarity and the tack and string of the pill when touched.
The fifth method, listed above, for insitu cooking is carried out primarily in large set kettles
or reactors. As indicated in the insitu formula listed earlier, the principle of this type of cooking is
simultaneous reaction of basic raw material to form resin or both resin and oil which further react
to form the desired varnish in one continuous operation. This is a much more complicated cook
and more sophisticated control is required. The end point of the cook is controlled by periodic
viscosity measurements or using a remote viscosity instrument. The reaction is carried out in a closed
kettle blanketed with an inert atmosphere. An inert gas is bubbled through the mixture to aid in the
removal of water formed from esterification of the organic and/or rosin acids and polyols. The reactions
involved are illustrated on the following page.
Resins, along with drying oils, make up the backbone of surface coatings since they are
responsible for the formation of a film with physical integrity and adhesion. It is the ability to form
this film that differentiates paint from a mere mixture of pigment and vehicle. Film formers are such
a critical component of the finish product that the latter is generally classified by the predominant
resin associated with it (e.g. epoxy and polyurethane).
In their original usage resins were various natural solid and semi-solid organic substances
(such as rosin and shellac) exuded from certain plants and trees, however, since the mid 1920's
natural resins have been supplanted by synthetically manufactured chemicals. Manufactured phenolics
first became commercially available in 1924, alkyd vehicles in 1925, and alkyd resins modified by
urea-formaldehyde in 1925. Each of these introductions represented a major breakthrough in the
13
-------
FATTY ACID TO OIL
CH2OH CH2OOCR '
CHOH + 3RCOOH > CHOOCR + 3H20
CH2OH CH2OOCR
Plyol (Glycerol) + 3 Fatty Acid ->• Triglyceride + 3 water
MALEIC RESIN
R-CH2-CH=CH-CHa-R + HC = CH > R~CH=CH-CH-R
I I I
HC - (
OIL OR RESIN 0 0 0
C
MALEIC ANHYDRIDE
OIL OR RESIN
ADDUCT
R1 R1
I I
R MA + G > -G-MA-G-MA-G-
1,1 I
ADDUCT GLYCEROL 0 R -MA OH
H I
G-OH
R'-MA
MALEIC RESIN
14
-------
production and application of surface coatings. In 1970 resins accounted for approximately 18%
by weight and 40% by production value of all chemical raw materials consumed by the paint and
varnish industry (this was equivalent to about $0.32/lb for synthetic resins as opposed to $0.15/lb
for total raw materials). By comparison natural resins comprised less than 2% of total film formers
in 1970 where prior to World War II they had contributed more than 50%.3
There is a large variety of synthetic resins produced for use in the manufacture of surface
coatings. A listing of the more popular resins is given below. They are listed by order of 1970
consumption by the coatings industry.3
Consumption — 1970
Resin (million pounds)
Alkyd 680
Vinyl 300
Acrylic 220
Epoxy 80
Cellulosic 65
Amino 60
Urethane 55
Rosin ester 50
Styrene 40
Phenolic 35
Hydrocarbon 30
Other synthetic 45
By far the most widely used resins are the alkyds whose consumption is twice that of its
nearest rival, the vinyls. Oil free polyesters, included in the "other" group in 1970, are enjoying rapid
growth. Based on the questionnaire sample obtained for this study, 1972 production was at least
91 million pounds.
Resins are polymer molecules built up from simpler units, or building blocks. The union of
these building blocks is generally brought about by the action of heat and catalyst. When the
polymer molecule is a simple repetition of the same unit it is a homopolymer; when two or more
different type units are present a copolymer results. To be capable of polymerizing, a monomer must
contain chemical groups that have potential chemical reactivity. The major monomers used are those
containing one ethylene double bond which permits very long chain polymers.
There are two types of polymers used in coatings: condensation polymers and addition
15
-------
polymers. A condensation polymer is a polymer formed from the reaction of two (or more) different
types of molecules with the splitting out of some product, usually water. Addition polymers are
formed from the "addition" of molecules, or monomers, to form the larger molecule without any other
reaction product.
Examples of condensation polymers include polyesters, amino resins, and alkyds. Alkyds
are really oil modified polyesters and constitute the most important groups. Polyester type resins
are formed from acid-alcohol reactions. Polymer formation requires the use of polyfunctional reactants.
Difunctional reactants produce linear polymers while the use of trifunctional, or higher, constituents
permits branching and cross linking to take place.
Examples of addition polymers include the acrylics and vinyls. These proceed via a free
radical mechanism in the presence of a catalyst to produce linear polymers. Modifiers can be intro-
duced to produce cross linking where desired.
In addition to the kind of building blocks used, the average number units in (or total molecular
weight of) the product molecule determines properties of the resin. In general, increasing molecular
weight improves film hardness, strength, resistance to water, chemicals and solvents but also results
in reduced solubility and higher viscosity. Of course, ultimate film properties can also be determined
by appropriate combinations of two, or more, resin types. In fact, a significant portion of coating
formulations utilize such combinations.
Alkyd Resin — Alkyds are widely used for a variety of reasons. They are ideal vehicles for pigmented
coatings and exhibit good wetting and dispersing properties. They use low cost solvents with minimum
odor and produce coatings having excellent durability, flexibility and gloss retention. Coatings can
be formulated with good solvent resistance, toughness, heat resistance, and color retention.
Alkyds can be combined with any of the following materials.5
Nitrocellulose Polyisocyanates
Urea-formaldehyde resins Silicone resins
Melamine-formaldehyde resins Polyamides
Phenolic resins Natural resins
Ethyl cellulose Cellulose acetatebutyrate
Chlorinated rubber Monomers (styrene, vinyl toluene, methyl methacrylate)
Chlorinated paraffin Synthetic latices (styrene-butadiene,
Epoxy resins polyvinyl acetate, acrylic)
Alkyd resins comprise a group of synthetic resins which can be described as oil-modified
16
-------
polyester resins. They are produced from the reaction of polyols or polyhydric alcohols, polybasic
acids and oils or fatty monobasic acids. A listing of commonly used raw material is given below:
Oils or fatty acids2
Linseed Castor
Soybean Coconut
Safflower Cottonseed
Tall oil fatty acid Laurie acid
Tall oil Pelargonic acid
Fish Isodecanoic acid
Tung (minor) Isooctanoic acid
Oiticica (minor)
Dehydrated castor (minor)
The materials in the first column are oxidizing or drying types. The materials in the second
column are non-oxidizing and yield soft non-drying alkyds which are used primarily as plasticizers
with hard resins. The acids shown in the last column are the only materials that are strictly synthetic
in origin.
Oils or fatty acids impart flexibility and drying to an alkyd. In general, the greater the unsatur-
ation (measured by iodine value), the greater the drying or hardening properties. The use and
characteristics of these oils was discussed earlier under varnish manufacturing.
Glycerol or glycerine was the first polyol used for alkyds. This material was first obtained
as a byproduct of the splitting of fat and oils in soap manufacturing. Also, glycerine is now produced
synthetically from petroleum sources and is supplied at a purity of 99+%. Glycerine has historically
been the most widely used polyol for alkyds.
At the present time, however, the leading polyol, based on usage, is pentaerythritol (PE),
which came into common use in the 1940's. PE is supplied as "technical grade" material and contains
mono, di, tri and polypentaerythritol. The material consists primarily of the mono form which is
illustrated on page 20.
The choice of which polyol to use is based on both technical and economic factors. For
some applications glycerine based alkyds are more suitable, while for others, pentaerythritol is
better. In many cases, however, a suitable alkyd can be formulated from either and for these a choice
can be made on the basis of cost or availability. Present economics favor pentaerythritol.
The important distinguishing feature of the various polyols is the number of potentially reactive
17
-------
hydroxyl groups in the molecule, known as functionality. The glycols with a functionality of two produce
only straight chain polymers and their alkyd resins are soft and flexible. The resultant products are
used primarily as plasticizers for hard resins. They are the least expensive polyols and are blended
with more reactive polyols such as PE. Glycerine has a functionality of three and is used primarily in
short and medium oil alkyds. Pentaerythritol, with a functionality of four, cross links to a greater
extent, forming harder polymers. It is ideal for use in long oil alkyds. However, due to its high reactivity,
it presents problems of end point control in processing of short or medium oil alkyds. Dilution with
ethylene glycol, as discussed earlier, reduces this latter problem and increases solubility in the
oils.
Acids and Anhydrides
Name
Formula
Form
Phthalic
anhydride
(ortho)
Isophthalic
acid (meta)
Terephthalic
acid (para)
Benzoic
acid
White solid
White needles
White crystals
White solid
18
-------
Name
Formula
Form
Maleic anhydride
H-C = C-H
C/VS
White solid
Maleic acid
c
^
OH
White solid
Ethylene glycol
Diethylene glycol
Propylene glycol
Glycerine
CP-95% glycerine
Super-98%
H
I
H-C-OH
I
H-C-OH
I
H
H H H H
II I I
HO-C-C-0-C-C-OH
I I II
H H H H
H H H
I i I
H-C-C-C-OH
I I I
H 0 H
H
H
I
H-C-OH
I
H-C-OH
I
H-C-OH
I
H
Liquid
Liquid
Liquid
Liquid
19
-------
Name
Formula
Form
Trimethylolepropane
Pentaerythritol
CH2OH
1 White solid
rie>-'-5 w L* ri -^ vJ r"i i
3 £. £L. '
, 1
1 CH2OH
wni-i r ru nui
nurioL.\. xuri->i'n
\ /
r White solid
/ \
unu r ^^LJ r\u
nUn2^ LrlpUn
Sorbitol
H H H H H H
I I I I I I
HO-C-C-C-C-C-C-OH
I I I I I I
H 0 0 0 OH
H H H H
0
White solid
Fumaric acid
HC —C —OH
II
HO-C-CH
White solid
0
The acidic material may be used in the form of an acid or an anhydride. An acid anhydride is
formed from the elimination of one molecule of water from two carboxylic acid groups and is pre-
ferred since it reacts faster and yields less water to be removed from the cook.
For many years, phthalic anhydride (ortho) was the only polybasic acid used in substantial
proportions in alkyds. It still remains the predominant dibasic acid. PA is produced from the
catalytic oxidation of naphthalene or ortho-xylene. Some resin manufacturers produce their
own PA. However, this process will not be studied here since it is more properly part of the
petrochemical industry.
Some of the acids listed above, such as benzole acid, are monobasic or mono-
functional, and cannot be used as the sole organic acid. They are used to terminate alkyd
polymers and to modify the properties of resins. Benzoic acid is substituted for part of the
fatty acids. It makes the resin harder and less flexible as well as enhancing the gloss.
20
-------
The chemistry of alkyd resin systems is very complex. So much so that theoretical
considerations offer only a good starting point. Final formulae and variations are developed
by trial and error changes, based on performance requirements and shortcomings of previous
batches.
Condensation is the reaction basic to all polyester resins, including alkyds. This
reaction follows the elementary equation for esterification as shown below:
.0 JQ
RC + R, OH J RC
OH OR,
I
Acid + Alcohol J Ester
For Alkyd Resins
PA + Glycerine J Glycerol phthalate + H2 0
H20
The ester monomer formed is very complex and further reacts to form large polymers
called resins. The polymers formed are low in molecular weight by comparison to other
resins. For example, alkyd resins have molecular weights ranging from 1,000 to 7,000 while
some vinyl and acrylic resins have average molecular weights in excess of 100,000 and in
some cases as high as 500,000.
The alkyd polymers also react with oil or fatty acid and are generally classified by the
amount of oil or PA used in the formulation, as described below:
% Oil _% PA
Short Oil
Medium Oil
Long Oil
Very Long Oil
The resulting reactants of the PA, polyol and oil may be represented in part as follows:
PA + G(OH)3
Phthalic Anhydride Glycerine
HQ-G-PA-G-PA-G-PA-G-PA- +
33 to 45
46 to 55
56 to 70
71 up
> 35
30 to 35
20 to 30
< 20
H20
OH PA PA
Glyceryl phthalate
OH
water
21
-------
This then will react with the long chain oil raw glyceride or fatty acid (~FA) to yield:
HO-G-PA-G-PA-G-PA-G-PA-
PA PA
FA FA
Alkyds can be manufactured directly from a fatty acid, polyol, and acid or from the
fatty acid oil, polyol, and acid. The second combination (oil, glycerine & PA) produces glyceryl
phthalate which is insoluble in the oil and precipitates. This problem can be overcome by first
converting the oil to a monoglyceride by heating with a polyol in the presence of a catalyst.
This process is called "alcoholysis" of the oil. The basic reaction is shown below:
CH-OOCR CH2OH CH2OH
I I '
C-HOOCR + 2CHOH + 3CHOH
I I I
C-H2OOCR C-H2OH CH2OOCR
Triglyceride Glycerine Monoglyceride
This is an ester interchange reaction with no loss of water.
When fatty acid rather than oil is used as the starting material, this is called the "one-
stage" process. In this process, the fatty acid and glycerine are added to the kettle, the agitator is
started and heat is introduced. Sparge gas is turned on to develop an inert atmosphere before the
temperature of the batch reaches 300°F where the reaction starts. When the batch reaches 400°F,
the PA is slowly added and cooking continued for another 3 to 4 hours until the desired body and
acid number are reached.
If the fusion process is being used, a continuous purge of inert gas is maintained to remove
the water formed in the reaction. This water may also be removed by what is known as the solvent
process. This latter process may also be referred to as the solution process or the azeotropic
process. It is similar to the fusion process except that about 10% aromatic solvent (usually xylene)
is added at the start. The vaporized solvent is passed into a condenser. The condensate then flows
to a decant receiver for separation of reaction water. Recovered solvent is returned to the reactor.
Most manufacturers maintain small inert gas flows throughout a solvent cook to prevent air leakage
into the reactor system.
A typical manufacturing formula for a 50% oil modified glyceryl phthalate alkyd using the
22
-------
one stage process is given below:5
Actual, Ibs Theoretical, Ibs
Phthalic anhydride 39.3 38.7
Glycerol (95% pure) 25.1 21.3
Linseed oil fatty acids 48.5 47.9
Catalyst
Methyl p-toluene sulfonate 0.2 0.2
113.2 108.1
Approx. loss 13.2 8.1
Yield 100.0 100.0
The above formulation is based on a 10% excess of glycerol and a manufacturing loss
of 5% more than theoretical.
As discussed earlier, when oil is used rather than fatty acid, the alkyds are produced in
a two stage process. In the first stage monoglyceride is first produced from the linseed oil and
glycerol. Part or all of the polyol is loaded into the reactor. The reactor is purged with inert gas
and the agitator and heat turned on. Before the catalyst and oil are added, the reactor is heated
to vaporize some polyol so that it can esterify any uncombined PA from the previous batch. If
this is not done, the uncombined PA will react with the catalyst and destroy its usefulness. The
catalyst and oil are then added and the alcoholysis of the polyol and oil is carried out between
450 and 500°F until the desired end point is reached. This end point is determined by solubility of
the batch in methanol. The solubility requirements vary with the type of alkyd being produced.
When the alcoholysis is completed, any additional polyol needed is added.
Following this, the required amounts of PA and esterification catalyst are slowly added. If
solvent cooling is to be used, the solvent is also added at this time. Cooking then proceeds as before.
A typical manufacturing formula for a 50% oil modified glyceryl phthalate alkyd using the
two stage process is given below:4
First stage Ib
Linseed oil 51.3
Glycerol (95%) 12.8
Catalyst, Ca(OH)2 0.026
23
-------
Second stage
Glycerol (95%) 6.2
Phthalic anhydride 39.7
Catalyst
Methyl p-toluene sulfonate 0.2
110.2
Approx. loss 10.2
Solids yield 100.0
Alkyd and other resins are cooked in closed kettles more properly called reactors. They
vary in size in commercial production from 500 to 10,000 gallons. A typical reactor system is shown in
Figure 3.
They are generally fabricated of Type 304 or 316 stainless steel with well polished surfaces
to assure easy cleaning. Design pressure is usually 50 psig. These reactors may be heated electrically,
direct fired with gas or oil, or indirectly heated using a heat transfer media such as Dowtherm®.
Reactors are equipped with a porthole sightglass, charging and sample line, condenser system,
weigh tanks, temperature measuring devices, and agitator. The porthole is used both for charging
solid material and for access to the kettle for cleaning and repair.
Good agitation is required for intimate mixing of the reactants and proper heat transfer. A
variety of different type blades are used but most are controlled with a variable speed drive. Too
rapid an agitation may cause discoloration and too slow an agitation can cause poor heat transfer
that will increase heat up time and may cause excessive polymerization due to localized hot spots
at the metal-liquid interfaces.
The reactor may be equipped with a variety of different condenser systems. The system
shown in Figure 3 includes a packed fractionating column, a reflux condenser and a main condenser.
The condensers are water cooled shell and tube type and may be either horizontally or vertically
inclined. Vapors are processed and condensed on the tube side and drain to a decant receiver for
separation and possible return of solvent to the reactor. A dual function aspirator Venturi scrubber
is often added to the system. It is used to ventilate the kettle during addition of solid materials and
may also remove entrained unreacted or vaporized solids and liquids from the venting gases.
It is not clear from the literature what kind of production rates, as a function of nominal
kettle volume, can be expected from an alkyd reactor. Rather than assume some maximum theoretical
production capacity, it was decided to rely on industry experience as determined from some of the
24
-------
- SPRAV TOWER
REFLUX
CONIDEMSER
VENT
FRACTIONATING
DISTILLATION
COLUN/IN
PORTHOLE
FOR SOLIDS
DIREdT FIRED OR
JACKETED FOR HIC3H
TEMPERATURE VAPOR
LIQUID
FIGURE
MODERN REISIN PRODUCTION SYSTEM
OF
SOLVENT AND FUSION COOKS
25
-------
questionnaires that were sent out as part of this study.
Questionnaire data from 18 alkyd resin producing plants with 45 alkyd kettles was analyzed
in an attempt to determine the effective production capacity of an alkyd resin reactor as a function
of kettle volume. Reactor volumes of 400 to 7,800 gallons were reported, with the majority of
kettles in the 500 to 2,000 gallon range. Cooking times from 61/2 to 29 hours were reported. In the
500 to 2,000 gallon size range, cooking times were well distributed in the 8 to 16 hour range.
In this same size range, production per batch worked out to about 5,000 pounds of alkyd
solids per 1,000 gallons of reactor volume. This number possessed a surprising degree of consistency,
within reasonable limits, from plant to plant. Assuming a typical density of about 9 pounds per
gallon for alkyds, this means that a 1,000 gallon kettle will produce about 550 gallons of alkyd
resin in each batch. The larger kettles tended to be filled to a larger fraction of their nominal volume.
Thinning tanks are always included as part of the reactor system. They are normally water
cooled and equipped with a condenser and agitator. The partially cooled finished alkyd is transferred
from the reactor to the partially filled thinning tank. Since most alkyd resins are thinned to 50%
solids, the capacities of these tanks are normally twice the capacity of the reactor. These tanks
are also frequently mounted on scales so that thinning solvents may be accurately added.
The final step in a reactor system is filtering of the thinned resin prior to final storage. This
is normally done while it is still hot. Filter presses are the most commonly used filtering device.
Centrifuges are also used. One manufacturer is currently using a continuous "in-line" blending and
filtering system for the majority of its resin production. It is not used in all cases, however, due to
limitations of the filters.
The manufacturing procedures and equipment used for the production of other resins listed
at the beginning of this discussion are quite similar. The major differences are the raw material and
process steps utilized. A brief discussion of the other important resin types follows. The technical
information contained in these discussions was obtained from Martens' book4 and the Federation
Series on Coating Technology.2 Economic and production information is from the SRI Chemical
Economics Handbook.3
Vinyl Resins — Vinyl polymers and copolymers were among the first of the synthetic polymers to gain
acceptance as binders for organic coatings. They are available in solution with organic solvents,
as high solids dispersions ("organosols" or "plastisols"), as dry powders or as water-borne latexes.
Discussion here is restricted to vinyl chloride, vinyl acetate, their related copolymers, and modified
types. Following conventional practice other materials like polyethylene, polystyrene etc. are not
26
-------
included although these also contain the basic vinyl building-block group R-CH=CHX.
In 1966 approximately 200 million pounds of vinyl resins were produced for use in surface
coatings. Four years later, in 1970 this number had increased by 50% to 300 million pounds. This
group of resins is highly versatile and relatively inexpensive. Next to the alkyds it enjoys the largest
surface coating use based on volume and sales value. One of the factors responsible for widespread
use of these resins is that manufacturers with alkyd kettles were able to adapt these for making
vinyl acetate latexes.
Vinyl resins are long molecular chains formed by the addition of vinyl chloride or vinyl
acetate monomers. No water or other reaction by-product is formed. The manufacturing chemistry
is outlined below.
CH2 = CMC CH2 = CHOOCCH3
Vinyl chloride »> i •* Vinyl acetate
Polymerize Copolymerize Polymerize
J 1
Polyvinyl Chloride Polycovinyl-Vinyl Chloride Acetate Polyvinyl acetate
-ChL-CH- -CH, -CH-CH9-CH- - CH, - CH -
2 | II I
Cl Cl O-C-CH3 0-C-CH3
O 0
Monomers comprising the final polymer retain their identity except for the double bonds
which largely disappear. The chain length, determined by the number of monomers, is controlled
by chain stopping agents or a combination of time, temperature and pressure. The polymerization
reaction is highly exothermic and, consequently, heat removal rather than addition is often the
problem once the reaction gets started.
There are basically four commercially available manufacturing processes for vinyl resins.
Combinations of more than one process, depending upon the end properties desired, is also viable.
These are:
1. Bulk polymerization — The monomers are polymerized by themselves (analogous to
the fusion process for alkyds). Due to poor heat transfer properties and difficulties in
controlling the reaction, this process has lost popularity.
2. Solution polymerization — Vinyl monomers are diluted in solvents containing catalysts.
Under controlled conditions this mixture is stirred in an autoclave until the increasing
27
-------
viscosity (by virtue of molecular size) reaches a predetermined level. When the desired
molecular weight range is attained the reaction is stopped. The polymer solution may be
used as is, or can be transferred to precipitation tanks where non-solvents precipitate the
resin. The several steps involved make this a costly, but nevertheless competitive, method.
3. Suspension polymerization — Small droplets of vinyl monomers containing catalysts
are suspended in water and polymerized. Polymerization is initiated within the droplet.
This method is used for the manufacture of a greater tonnage of vinyl resin than either
of the processes mentioned above. It has largely replaced the first and competes favorably
with the second in producing a less costly resin at some expense of quality. The autoclave
emulsion product is a suspension of submicron polymer particles and may be marketed in
the latex form or the dry resin may be recovered and sold as a powder. The separation of
water from submicron particles poses a difficult problem and invariably results in dusting and
loss of powder to the atmosphere as fugitive emissions. Resins sold as dry powder are used
to prepare organic coating dispersions ("plastisols" and "organosols").
4. Emulsion polymerization — This process differs from the preceding one in that an
emulsifier or surfactant is used causing the formation of a very stable fine dispersion of
monomer (and consequently of polymer) particles in water. The catalyst is dissolved in
the water phase and initiation takes place it the droplet interface. While suspension resins
are frequently filtered and dried prior to sale, emulsion polymers are rarely so treated.
Combination methods may employ one of the above processes followed by further chemical
treatment to impart additional desirable qualities such as alkyd compatability.
Solvent based vinyl systems require the use of relatively strong, polar solvents. Ketones,
nitroparaffins, esters, and some chlorinated solvents can be used. Aromatic hydrocarbons tend to
swell the resins but can successfully be used as diluents to reduce costs. Alcohols and aliphatic
hydrocarbons are not used.
Acrylic resins — Acrylic resins form a group of transparent thermoplastic (or thermosetting if suitably
modified) resins formed by polymerizing monomeric esters of acrylic acid or methacrylic acid.
Acrylic resins are addition polymers and produce no water or other reaction by-product.
Usage of the term "acrylic resins" is restricted here to resins having the following typical
acrylic or methacrylic structures:
28
-------
H
- CH2 ^C - CH2 - C -
COOC2H5
COOC2H5
or
CH,
1
r^ui /"" C1
— V_»rlrt """>*-' — ^
COOCH3
CH,
1
H2?C-
COOCH3
-1 N
(Poly) ethyl acrylate
(Poly) methyl methacrylate
Acrylic resins, available in several physical forms, are used in finishes for metal, wood,
leather, ceramic and plastic surfaces. Their coatings are characterized by resistance to degradation
by UV light, hydrolysis and corrosive chemicals, resistance to mechanical damage and a high
quality appearance with aesthetic appeal. Copolymerization with functional monomers is used to
control hardness and slip properties, adhesion, solvent resistance and to provide a means for
subsequent cross-linking if required. Acrylic polymerization is exothermic and presents a heat
removal problem.
The two basic types of acrylic emulsions available are thermoplastic and modified thermo-
setting. The former is comprised of copolymers of methyl methacrylate with varying amounts of
acrylate esters; these are for use in latex paints and in lacquer type finishes. Cross-linking reagents,
which react with the functional groups when the film is baked, can be added to produce a thermoset
polymer. Films formed in this manner display increased resistance to solvents, chemicals, and
greatly improved toughness and hardness at high temperatures.
Manufacturing processes for acrylic resins closely parallel those employed for vinyl resins.
These are:
1. Bulk polymerization — Polymerization of undiluted monomer.
2. Solution polymerization — Makes use of a solvent reaction medium in which both
monomer and polymer are soluble. This process is used for producing medium-molecular
weight resins.
3. Suspension polymerization — The liquid monomer is suspended as droplets in a water
medium by vigorous stirring. This technique is restricted almost exclusively to the manufacture
of a few powder resins for use in lacquers.
4. Emulsion polymerization — Liquid monomers and surfactant or emulsifying agents
are added directly to water. The rate of monomer addition (during the process), the pH and
the temperature of the emulsion determine final properties. Water acts as heat transfer
medium and a very high molecular weight polymer results. This method is used mostly
for water-base acrylic coating resins. In handling by consumers the hazard, odor and
29
-------
expense of flammable solvents are eliminated.
Monomers used in acrylic resin production include the methyl, ethyl, and n-butyl esters of
the acrylates and the methacrylates. These substances are characterized by a low to medium
boiling range and by very low odor thresholds. Indeed, odor nuisance is probably the primary air
pollution problem associated with acrylic resin production.
Solvent based acrylic systems utilize a variety of solvents. Thermoplastic solution resins
are supplied in solvents such as MEK, toluene, or mineral spirits depending on the particular
polymer and the intended application. Thermosetting acrylic resins are supplied as solutions in
solvents such as xylene, butanol, and ethylene glycol monoethyl ether acetate.
The following items of statistical interest3 concerning the market for acrylic resins, are offered:
1. In 1970, 220 million Ibs of acrylic resins were produced for a value of $85 million.
2. In 1970, these resins accounted for about 13.3% by weight and 16% by value of all
synthetic resins, placing them third after alkyds and vinyls.
3. Rohm and Haas Company is the largest producer of acrylics for latex paints (65 to
70% of this market in 1971).
4. Of the 14 other producers of acrylics for latex paints, only Union Carbide, Celanese
Resins, and Thibaut & Walker enjoy any significant sales volume. Most other suppliers in
the latex paint industry concentrate their efforts on vinyl resins (polyvinyl chlorides and
acetates).
5. Ten paint companies produce acrylic emulsions for captive use (including DuPont,
Celanese, Glidden, DeSoto, NL Industries, and Sherwin-Williams).
6. In 1971, over 60% of all acrylic lacquer was used by General Motors as topcoats
for automobiles (other domestic auto makers use thermosetting acrylic enamels).
7. More than 24 companies produce acrylic enamel resins (thermosetting solution polymers).
The major markets are served by Celanese, Cook Paint & Varnish, DeSoto, DuPont, Ford
Motor, Glidden, Inmont, Mobil, PPG, and Sherwin-Williams. These major markets are
almost entirely in captive use.
Epoxy resins — Industrial finishes make extensive use of epoxy surface coatings. This popularity is
is ascribed to their high resistance to corrosion, strong chemicals, physical abuse, and their
adhesion properties.
Disadvantages of epoxy coatings, which have curtailed their use in trade sales, are:
30
-------
1. The need for strong, toxic and/or pungent solvents
4
2. High cost (although cost per pound, per year of service can be lower than for other resins)
3. Tendency to chalk (loss of gloss) and
4. Appearance
The term "epoxy resin" refers broadly to polymers containing epoxide or oxirane groups:
H
H
R - C - C - R'
\/
O
The most commonly used epoxy resins in surface coatings are the diglycidyl ether type
derived from epichlorohydrin and bisphenol A. Manufacturing chemistry is outlined below:
CH3
fV-c
CH,
CH2 - CH - CH2 - Cl
0
alkaline
conditions
Epichlorohydrin
CH2 - CH - CH2 - O
0
'3
Bisphenol A
0 - CH,, - CH - CH,, - O
Epichlorohydrin is charged as a liquid to the reactor. The bisphenol A is a dry powder and
as such presents a possible dusting problem. The reaction is carried out at moderate temperatures
(up to 250°F) and produces HCI as a by-product.
The molecular weight of this entire group of resins is often too low to permit use without a
curing process (usually conducted at 140 to 400°F). This results in a two-package, "amine-cured"
coating, so named because the curing agents depend on the reaction between the terminal epoxide
groups of the epoxy resin and the amine hydrogen atoms of primary or secondary amines.
Among the first curing agents to be used for this application were the aliphatic polyamines
DTA (diethylenetriamine) and TETA (triethylenetetramine). These aliphatic amines can lead to
dermatitic or respiratory problems where proper precautions and housekeeping practices are not
followed during application.
Polyamine adducts provide comparable physical and chemical resistance properties and
exhibit lower dermatitic potential and lower vapor pressure than the parent aliphatic polyamine.
Amine terminated polyamide resins form another major group of curing agents with lower
31
-------
irritation potential. These resins produce a more flexible film with lower chemical resistance than
amine-cured films.
Solvents for most unmodified epoxy resins are ketones and esters. Aromatic hydrocarbons are
used as diluents. Solvent selection is based on method of application required: brush application requires
slow evaporating solvents (e.g. diacetone alcohol, methyl isobutyl ketone) and diluents (e.g. xylene),
while spray application calls for faster evaporating solvents (e.g. methyl ethyl ketone) and diluents
(e.g. toluene).
In 1970, epoxy coating resin consumption was approximately 80 million pounds; production
growth has exceeded a 12% average annual rate compared to a 2 to 3% average for total
surface coatings (based on data reported to the U.S. Department of Commerce, Bureau of the
Census). Classified as either unmodified or modified (by fatty acids), production can be broken
down as:
1. Unmodified
a. Shell — approximately 50%
b. Dow, Ciba and Celanese, jointly — approximately 40 to 45%
c. Union Carbide, Reichhold and Resyn — the remainder
2. Modified
a. For sale only — 21 companies produce modified epoxies for sale, this accounts
for 15% of total epoxy resin demand on weight basis.
b. For captive use — Celanese is the largest producer. Other companies with
significant captive consumption include Cook Paint & Varnish, DeSoto, duPont,
Inmont Chemical, Mobil, PPG, and Glidden Division of SCM.
Urethane resins — Unlike vinyls and acrylics, urethanes are not polymers made up of monomeric
urethane molecules or groups. They are reaction products of isocyanates with hydroxyl-containing
materials:
R-N=C = O + R'OH -> R-N-C-O-R'
I II
isocyanate hydroxyl H 0
The term "urethane" is used for protective finishes that incorporate a resin containing any
of the isocyanate-derived linkages as shown on the following page.
32
-------
0
II
- N - C - 0 -
C = 0
- N - H
allophanate
fi Y
- N - C - N
i O
C = O
- N - H
biuret
A
o=c c.o
N N
Y
y
isocyanurate
H O
H
fl _ C - N -
urea
H 0
-L'c'-o
urethane
Polymer formation is accomplished by the use of polyfunctional isocyanates and poly-
hydroxyl compounds. The primary raw material used for production of urethane coatings are
isocyanates, particularly TDI (toluene diisocyanate).
The commercial grade of TDI is usually a 80:20 mixture of the 2, 4 and 2, 6 isomers:
CH-
NCO
CH3
NCO
80%
20%
Handling of TDI poses special problems because it has a significant vapor pressure and
an irritant effect on the mucous membrane. It has been found4 that if the unreacted TDI in any
prepolymer is kept below 1% of the total product then the amount of TDI vapor released to the air
will be at a level considered "safe." Manufacturers are attempting to maintain the free or unreacted
isocyanate monomers below this safe level.
Another diisocyanate of commercial importance is 4, 4 — diphenyl-methane-diisocyanate
(MDI):
OCN
NCO
33
-------
MDI is a solid at ordinary temperatures.
Polyols commonly used included glycerine, glycols, IMP, PE, phenols, and polyethers.
Polyesters, alkyds, and castor oil can also be used to supply the hydroxyl groups. A drying oil
modified urethane (sometimes called "uralkyds") can be made in a manner similar to alkyds. In
this variation the diisocyanate performs the acid function of the phthalic anhydride in linking two
monoglycerides. Oil modified urethanes are finding increasing use, especially in trade sales products.
These avoid the elaborate curing mechanisms required for the moisture cure, blocked, and two-
package formulation types of urethanes.
These materials require strong solvents such as ketones, esters, and aromatics particularly
for reactive urethane resins. Adequate ventilation must be provided where these materials are to
be handled. Possible alternatives under investigation are coatings without solvents, water emulsion
systems and the use of less toxic aliphatic solvents in place of aromatics. This substitution by
aliphatics is practical only when total isocyanate content of the resin is below 25%. The oil modified
urethanes can utilize mineral spirits and naphtha type solvents.
In comparison to other resins (e.g. alkyds, epoxies, vinyls) urethanes can be formulated
to offer a better combination of curing rates, abrasion resistance and water resistance. For an
equivalent hardness they are more flexible. Their major drawbacks are cost, yellowing (of urethanes
based on aromatic isocyanates), need for strong solvents, and toxicity of some raw materials (as
previously discussed).
In the U.S. over 600 companies produce urethane coatings and about 80 companies manu-
facture urethane resins. Of these, only 22 produce and sell the resins but not the coatings — these
account for half of the resin produced. The largest single application for the coatings is in clear
wood finishes.
Cellulosic resin — Cellulosic resins are used in the production of about 57% of total lacquers which,
by definition, are surface coatings that dry solely by evaporation of an organic solvent (rather than
by reaction, such as oxidation or cross-linking, of resin components). Nitrocellulose accounts for
75% of this entire group and is used chiefly as a wood finish. It is easy to apply, low in cost, has
excellent intercoat adhesion and solvent release properties. It is also, however, flammable, it
yellows and has low chemical resistance.
Cellulosic lacquer resins are manufactured by the esterification of cellulose, a hydrolyl
containing compound. For example, nitrocellulose is prepared by nitration of cellulose with nitric
acid in the presence of sulfuric acid which removes water formed during the reaction. Manufacturing
34
-------
chemistry is:
Cellulose + HNO3 + H2SO4 -» Nitrocellulose + H2SO4 + H2O
Cellulose, in the form of cotton linters, is boiled and washed then treated with nitric acid.
The reaction product is then treated with ethyl, isopropyl or butyl alcohol (30% alcohol to 70% cellulose),
resulting in a highly flammable mixture. Even the recent use of water instead of alcohol as a
dampener does not appreciably reduce the hazard potential of handling nitrocellulose. Various
safeguards regarding this material are listed in "Handling of Nitrocellulose Manual Sheet N-1"
available from the Manufacturers Chemist Association. It may be pointed out that initial interest
in nitrocellulose was in the field of explosives.
This group of surface coatings represents a fair vapor emission potential since it is normally
used with several volatile materials such as ketones, esters, glycol ethers, alcohols, and other
hydrocarbon solvents.
Cellulosic lacquer resins are produced for sale only by Dow Chemical, duPont, Eastman
and Hercules. The lacquers themselves call for conventional grinding and mixing operations so
most coating producers are capable of manufacturing them.
Cellulosic lacquers can be reformulated with solvent systems considered exempt in many
states to accommodate their air pollution laws.
Markets for nitrocellulose lacquers are not expected to show substantial growth due to
competition from other coating systems. Cellulose acetate butyrate (CAB) is, however, expected
to display moderate growth because of its use in thermosetting acrylic automotive enamels to
improve their performance. CAB is produced by reaction of acetic and butyric acids and their
anhydrides with cellulose followed by partial hydrolysis.
Amino resins — Amino resins are used as a binding or cross-linking agent in the preparation of baking
or thermosetting coatings. They can be used in conjunction with alkyds, epoxies, thermosetting
acrylics, phenolics and other heat reactive resins. When used in this manner, these resins comprise
up to 30% of the total vehicle binding agents. Their use serves to improve baking speed, hardness,
solvent and chemical resistance of the film. Reaction occurs primarily with hydroxyl groups present
in the polymers to form a cross-linked, thermosetting polymer.
The most widely used resins in this group are condensation products of urea or melamine
with formaldehyde. Reaction with urea to form dimethylol urea (DMU) is illustrated:
NFL H - N - CH,OH
I ? I
C = 0 + 2 H CHO -> C = 0
I I
NH2 H - N - CH2OH
urea formaldehyde dimethylol urea
35
-------
Reaction of formaldehyde with melamine
NH
2
c
N
II
NH2 - C C - NH,
^^«~ *
N
proceeds such that the three —NHa groups convert to —N(CH2OH)2 groups to form hexamethylol
melamine. Recently methyl ethers of methylol melamine have received much attention with respect
to the elimination of organic solvents from baking finishes.
DMU and hexamethylol melamine undergo self-condensation reactions to form polymers
with the splitting out of water. The condensation is via methylol groups to yield on ether linkage
under alkaline conditions and via interaction between methylol and amide groups to form methylene
linkages under acid conditions. For use as coating resins, this self-condensation tendency is blocked
by etherification with reactive alcohols such as n-butanol:
H H H
I I I
R_N-CH2OH + C4HgOH -> R - N - C - 0 - C4Hg + H2O
H
methylol + n-butanol -> n-butyl ether + water
Some condensation takes place during etherification. The butylated aminoplast can also undergo
further polymerization but only at elevated temperatures. This is the basis for its use in baking
systems. The butylated amino resin will also cross-link with hydroxyl containing polymers when
acid catalyzed at elevated temperatures. This explains its use as a binding agent in conjunction
with other resins in baking finishes. The degree of polymerization allowed during etherification and
the amount of combined formaldehyde can be controlled to produce the desired properties.
An example of the preparation of a butylated ether of urea-formaldehyde is outlined by
Martens4 as follows:
"Heat one mole of urea with 2 to 3 moles of slightly alkaline aqueous formalin (pH 8 to 9)
to form the methylol derivatives. Add 2 to 3 moles of butanol and adjust with acid to a pH of 3 to 6.
Remove the bulk of the water by continuous azeotropic distillation and proceed to an atmospheric
distillation for removal of the final traces of water." Amino resins are usually supplied in butanol-xylene
solvent systems.
36
-------
Hydrocarbon resins5 — This group is loosely named and includes:
1. Coumarone — Indene Resins
2. Petroleum Resins
3. Styrene — Butadiene Resins
4. Terpene Resins
Together these materials account for a production value roughly equivalent to that of amino resins
($15 million in 1970). Discussion of individual resins is restricted to a brief description.
This group, broadly, covers low molecular weight (under 20,000), thermoplastic resins.
Their use is justified by ease of processing, by their neutral hydrocarbon nature, acid-alkali resistance,
solubility, and low cost.
1. Coumarone — Indene Resins: these were first produced commercially in the 1920's
by polymerization of the fraction of coal tar distillates containing coumarone and indene.
Polymerization is conducted in the presence of sulfuric acid (or sometimes aluminum,
stannic or antimony chloride) as catalyst. These resins are soluble in aromatic solvents
and moderately soluble in petroleum solvents. Structures of coumarone and indene are
illustrated:
H H
Indene
2. Petroleum resins: are derived from unsaturated hydrocarbons made available from
the cracking of petroleum. Because of a strong tendency to form gummy polymers, these
resins are undesirable in gasoline and are removed. This ease of polymerization is, however,
taken advantage of in the production of resins at low cost. They can be dissolved in solvents
and blended with bodied oils for aluminum paints. Monomers for these resins include
ethylene, acetylene, butylene, isobutylene, isoprene, and cyclopentadiene.
3. Styrene-butadiene resins: are copolymers containing at least 85% styrene. They are
soluble in aromatic hydrocarbons, ketones, and esters. They are not generally compatible
with alkyd resins or drying oils; but are compatible with rosin materials and coumarone-
indene resins.
4. Terpene resins are produced from compounds derived from the resin of the pine tree.
37
-------
The most important of these compounds in coating resins is /3-pinene:
CH,
CH,
These resins are soluble in aliphatic hydrocarbons.
All of the monomers used in the various types of hydrocarbon resins are considered
photochemically reactive. They contain olefinic and/or cyclo-olefinic types of unsaturation.
Phenolic resins — Among the oldest of synthetic resins, phenolics have been in use since 1910.
Although their use in coatings ranks them lower in consumption volume than other groups discussed
so far, their total production as a plastic exceeds a billion pounds annually.
Phenolics are condensation polymers obtained by the reaction of a phenol or a substituted
phenol with an aldehyde — usually formaldehyde. Depending upon ingredients used, their proportions,
catalysts employed, and on reaction conditions, the final form can be a viscous liquid or a brittle
solid. Further, these may or may not be heat-reactive.
Phenolics formed in alkaline systems have a high ratio of formaldehyde to phenol ranging
from 1.0 to 3.0. In acid systems the ratio lies between 0.7 and 0.9, and the product consists of a
thermoplastic or "novalak" (or heat-reactive and non-heat reactive, respectively). A high ratio alkaline
catalyst system produces thermosetting resins.
Resins prepared in alkali are of low molecular weight and their coatings on metal tend to
eyehole and crawl; phenolics formed in acid medium require the addition of curing agents which
limit their shelf life. Ammonia or amine catalyzed phenolics become the normal choice. These, when
cured, have an attractive gold color found in can linings. Most phenolic resin coatings comply with
Food and Drug Administration regulations concerning safe linings of food containers.
Manufacture of phenolics calls for ordinary stainless steel kettles designed to withstand
38
-------
both vacuum and moderate pressure and equipped with reflux condenser and dumping hatch.
The chemistry of phenolic resin production is illustrated below:
OH H OH
+ C - 0 -> f\T-—CH2°H + L U
H ^^ CH2OH
phenol formalciehyde
The —CH2OH groups can react with available ortho or para hydrogens on phenolic rings or with
other —CH2OH groups. Either way, a molecule of water is formed.
Most of the commercial phenolics use formaldehyde as the source of the aldehyde groups
even though, in theory, many aldehydes are suitable. Formaldehyde is a gas at ordinary temper-
atures but is usually supplied commercially in 37 to 50% aqueous solutions.
Besides phenol, a few substituted phenols are used. These include p-cresol, p-phenylphenol,
p-tert-butylphenol, and bisphenol A. Only the para forms of the above compounds are used since
the ortho and mefa forms produce polymers of poorer quality.
Solvents used for phenolics include alcohols, ketones, esters, and glycol ethers.
Silicone resins — Silicone resins are generally supplied as solutions in aromatic hydrocarbon solvents
which are included in manufacture to improve control of hydrolysis and prevent gelation. Aliphatic
hydrocarbons and alcohols may also be used; ketones and esters are seldom required.
Silicone resins exhibit excellent color and gloss retention even at 500°F. They can be
copolymerized to improve properties of other resins at elevated temperatures. They also exhibit
excellent external durability. These two properties are responsible for their position in the coatings
industry.
The manufacture of silicone resins begins with the production of chloro-substituted silanes.
Several methods are available, all of which produce mixtures of various alkyl-chloro substituted
products (such as (CH3)4Si, (CH3)3SiCI, (CH3)2SiCI2, CHsSiCU). These are separated by fractional
distillation.
The next step is the preparation of silanols as, for example,
RsSiCI + H2O -» R3SiOH + HCI
R2SiCI2 + 2H2O -» R2Si(OH)2 + 2HCI
RSiCI3 + 3H2O -» RSi(OH)3 + 3HCI
This is followed by the polymerization of the polyfunctional silanols:
39
-------
R
1
HO-Si-0-
i
1
R
R
1
Si-0-
1
1
R
R
I
Si
1
R
nR2Si(OH)2
n-2 +(n-1)H20
Monofunctional silanols are used to terminate the chain while trifunctional silanols can be used as
cross linking agents.
Silicone resins can be used as such but are often combined with other vehicles either as a
blend, or, better, as a copolymer. A commonly used copolymer is that made with alkyds. The
alkyd can either be cooked separately and then copolymerized with a methoxy or hydroxy functional
silicone intermediate; or the alkyd raw materials and the silicone intermediate can be charged
together and cooked in a single step.
Another important copolymer group is the silicone-acrylic copolymer emulsions. These are
used mainly for exterior house paint and maintenance coatings. This concludes the discussion of
the resin types.
As may be surmised from the preceding discussions, paint manufacturing can be summarized
as a raw material and finished products handling problem with a variety of intermediate batch
operations. The inter-relationship of all these operations is schematically illustrated in Figure 4.
As paint manufacturing has developed into a specialized branch of the chemical process industry,
the quantity and number of raw materials and finished products have increased significantly. The
industry, however, has been able to maintain the price of finished product by increased batch size,
by more efficient use of labor, and, most importantly, through improved material and product
handling.
Usually, plants that produce many of their own resins and manufacture a variety of trade
and industrial products are divided into the general areas indicated on Figure 4. This simplifies
handling of solid materials and solvents which vary for each type of operation. This is also true for
packaging of the finished products.
The early paint plants were normally built in four story buildings to allow for gravity process
flow. Vehicles and solid were stored on the top floor in tanks and bags. Mixing of pigments and
vehicles took place on the third floor; the grinding or dispersion occurred on the second floor; and
the filling, packaging, warehousing, and shipping took place on the first floor.
New paint plants are now designed differently from earlier facilities. They can be a single
story open building with or without a mezzanine or a three story plant laid out on a hillside. The
40
-------
tn
I
V)
cr
u
*
o
!
1
41
-------
major impetus for this change was handling and warehousing of raw material as well as finished
stock. This is done entirely on the ground floor using fork trucks, conveyors and storage racks. The
hillside three story building meets these requirements, since the top story is also a ground floor.
Processing is then done on the second floor and filling and warehousing on the lower level. Material
flow is also conducted by gravity and this arrangement is probably the best.
B. Process Data From Questionnaires
A primary source of information for this study will be the Paint and Varnish Industry Question-
naire which has been distributed to a sampling of the industry. The questionnaires were distributed
under authority granted the U.S. Environmental Protection Agency by Section 114 (a) of the
Clean Air Act.
Several criteria were considered in selecting those plants which would receive a questionnaire.
These included:
1. Geographical location
2. Type of products
3. Plant size
4. Presence of air pollution control equipment
5. Total output
6. Number of plants
The list of plants which were chosen represents a compromise with respect to the above-
mentioned criteria. The nature of the coatings industry is such that it is difficult, if not impossible,
to obtain a sample population which is representative of the industry according to all of these
criteria. If, for instance, one were to obtain a representative sample according to plant size, the
result would be a very poor sample with respect to production volume. The aim, then, was to obtain
a list of plants which would provide the desired information and yet provide at least "adequate"
representation with respect to the guidelines set forth above.
Table 2 contains a breakdown of the way the questionnaires were distributed. This table
reflects the groups as they were sent out.
The "Resin List" was taken from the NPCA Raw Materials Index - Resin Section.6 The
Index listed 59 companies which produced paint resins for sale. It includes some companies which
produce finished coatings as well as resins but does not include companies which produce resins
for internal use only. Of the 59 companies, 32 were chosen on the basis of diversity of product and
42
-------
amount of output. Plant locations were obtained from the companies themselves or, for those
companies which would not release that information, from the Chemical Economics Handbook
of the Stanford Research Institute3 or the Paint Red Book.7 A total of 55 plants were selected.
Where a company had more than a single plant, about half its plants were chosen.
A list and description of the 200 largest paint companies is contained in Kline's Marketing
Guide to the Paint Industry.1 These 200 companies account for about 88% of the industry output.
From this list 139 large companies were selected. These were further subdivided into single plant
companies and 31 multi-plant companies. As before, each of the 31 companies that had more than
one plant had questionnaires sent to about one half of its plants resulting in a total of 75 questionnaires.
Many of the plants in this group produce resin as well as paint.
The "small" plant list was obtained from a list of its "A" members supplied by the National
Paint and Coating Association. The companies were classified by the NPCA into very general
categories according to size, percent trade or industrial, and product type. This list was used to
obtain a sample of small producers. In the NPCA classification accompanying the membership
list, "small" was defined as having sales of less than 5 million dollars. A sample of 198 companies
classified as small was selected. Only two of these companies were included in the 200 largest
companies described in Kline's. The smallest of these two had sales of 1 to 2 million dollars.
Because of the nature of the paint industry, it is likely that these 198 companies average considerably
less than one million dollars in sales. As far as could be determined at the time, none of these
plants produced resins. This has been substantiated by the questionnaires received.
Finally, another 16 plants from 13 companies were selected. They were included on the
basis of reports that they contained air pollution control devices. Since one of the purposes of the
questionnaires was to obtain cost effectiveness information on such devices, effort was made
throughout to locate plants with control equipment. For this reason, a somewhat disproportionate
share of those plants in geographical areas known to have a history of vigorous pollution control
legislation received questionnaires.
Table 3 contains a geographical breakdown of the categories as listed in Table 2. Table 4
presents a percentage distribution by geographical regions for the entire questionnaire sample
compared to the distribution of all plants and of value of shipments for the entire industry.
Examination of Table 4 indicates that the mid-Atlantic region is somewhat under-represented.
A large part of this under-representation can be traced to the "small" plant list. It is felt that this
will not materially affect any conclusions which might be drawn from the questionnaires as a group.
43
-------
TABLE 2
QUESTIONNAIRE DISTRIBUTION
Type Companies Plants
Resin 32 55*
Marketing Guide
(a) Multi-plant 31 75
(b) Single plant 108 108
Small 198 198
Other 13 16
Total 382 452
"Ten of these plants also manufacture coatings.
44
-------
Questionnaire Returns
A total of 338 questionnaires have been returned in time to be used in this study. As each
one was received, it was placed into one of the three categories outlined below.
Plant Type Production Questionnaire Received
1 Coating and Resin 76
2 Coatings only 223
3 Resins only 39
In addition 22 questionnaires have been withheld by the Environmental Protection Agency due
to requests for confidentiality and are not included in the above. Another 13 plants, by mutual
agreement, were not required to return the questionnaire since it did not properly apply to their
situations. The 338 returns used amount to 75% of the total mailed.
1. Raw Materials
Table 6 lists raw materials used by the respondent plants. All quantities are in million
pounds per year. Broad categories covered are oils, polyols, acids, monomers, purchased resins,
pigments and extenders. Solvents, due to their importance in air pollution, are covered in a separate
table.
Among the oils, usage of linseed exceeded that of soybean oil, for responding plants, by
a much smaller margin than in industry-wide figures. Linseed is a drying oil but it yellows whereas
soybean oil is semi-drying and resists yellowing. Usage of these two oils, for the 338 plants was
about 95 million pounds and 82 million pounds respectively. Approximately one-fourth of the total
oil was used in production of resins by type 3 plants. Sixteen percent is used by type 2 plants for
production of coatings. The remaining 56% is used by type 1 plants for both resin production and
coating production.
The primary usage of polyhydric alcohols, or polyols, is in the production of resins by the
esterification reaction with an acid. Glycerol or glycerine, first obtained as a by-product of soap
manufacture, is used in short and medium oil alkyds. Pentaerythritol, with four hydroxyl groups, is
used primarily in long oil alkyds. These two polyols are used in quantities of 22.7 and 36.1 million
pounds, respectively. Ethylene and propylene glycols are the lowest cost polyols available and
their use in type 2 plants (which do not produce resins) is largely as an "anti-freeze" agent. "Other"
polyols listed include trimethylolethane and sorbitol.
45
-------
TABLE 3
GEOGRAPHICAL DISTRIBUTION OF QUESTIONNAIRES
Multi- Single
Location Resin Plant Plant Small Other Total
New England 3 2 3 15 2 25
Me., Vt., N.H.,
Mass., R.I., Conn.
Mid Atlantic 11 13 24 23 5 76
N.Y., Pa., N.J.
East No. Central 14 20 33 54 4 125
Wis., Mich., III.,
Ind., Ohio
West No. Central 1 3 13 13 0 30
N.D., S.D., Neb.,
Kan., Minn., la., Mo.
South Atlantic 3 8 12 22 0 45
Md., Del., W.Va.,
Va., N.C., S.C., Ga.
East So. Central 8 4 6 12 0 30
Ky., Tenn., Miss.,
Ala.
West So. Central 4 6 5 15 1 31
Tex., Okla., Ark.,
La.
Mountain 0 1 2 7 1 11
Mont., Id., Wyo.,
Nov., Utah, Colo.,
Ariz., N.Mex.
Pacific 11 18 10 37 3 79
Wash., Ore., Cal.
55 75 108 198 16 452
46
-------
Region
TABLE 4
PERCENTAGE DISTRIBUTION OF QUESTIONNAIRES,
PAINT PLANTS, AND VALUE OF SHIPMENTS
New England
Mid Atlantic
East No. Central
West No. Central
So. Atlantic
East So. Central
West So. Central
Mountain
Pacific
Questionnaires
5.5
16.8
27.7
6.6
10.0
6.6
6.9
2.4
17.5
Number of
Plants*
5.8
26.1
23.6
6.3
10.3
3.2
6.6
1.5
16.6
Value of
Shipments*
2.8
23.4
35.0
6.5
7.5
4.6
6.2
0.6
13.4
100.0
100.0
100.0
'Source: 1967 Census of Manufacturers
47
-------
TABLE 5
INDUSTRY QUESTIONNAIRE — TABULATION SUMMARY
Number of Plants in Sample
Products & Production
Major Products (MM gal)
Paints — oil/solvent base
Paints water base
Varnishes
Lacquers
Total
Trade sales
% of total sales
Major Resins and Varnishes
Produced (MM Ib)
Alkyd
Acrylic
Polyester
PVA
PVC
Epoxy
Urethane
Cellulosic
Ami no
Rosin
Styrene
Phenolic
HC
Other
Total
Varnish Total
Separately packaged solvents
Type 1 Type 2
76 223
98.32 77.01
49.28 43.86
21.43 1.10
12.07 23.16
181.10 145.13
108.66 77.45
60 53
299.15
51.07
28.76
48.72
0.03
49.88
4.21
2.18
19.42
0.65
15.79
3.16
0.08
29.67
552.77
18.68
17.07 8.41
Type 3
39
—
191.32
114.03
67.50
27.93
63.50
12.84
1.79
14.76
24.03
24.68
0.60
—
38.00
40.89
621.87
37.74
—
All
Types
338
175.33
93.14
22.53
35.23
326.23
186.11
57
490.47
165.10
96.26
76.65
63.53
62.72
6.00
16.94
43.45
25.33
16.39
3.16
38.08
70.56
1,174.64
56.42
25.48
1970
Industry
Total3
83
43
5
166
48
-------
TABLE 6
INDUSTRY QUESTIONNAIRE — TABULATION SUMMARY
Raw Materials Used (MM Ib)
Oils
Type 1
5.71
0.02
47.37
6.01
56.05
5.51
27.72
0.81
0.60
8.87
Type 2
2.15
0+
29.43
1.44
1.99
0.93
0.02
1.73
—
7.54
Type 3
0.85
0+
18.12
5.02
23.55
1.37
18.68
0.56
0.83
8.41
All
Types
8.71
0.02
94.92
12.47
81.59
7.81
46.42
3.10
1.43
24.82
Castor
Cotton Seed
Linseed
Safflower
Soya
Tung
Tall
Fish
Coconut
Other
Total 158.67 45.23 77.39 281.29
Polyols
Glycerol
Ethylene glycol
PE
TMP
PG
Other
Total 76.82 11.88 43.00 131.70
Acids
PA 61.69 42.86 104.55
IPA 7.38 4.82 12.20
MA 4.47 6.25 10.72
Benzoic 2.37 0.11 2.48
Adipic 0.77 0.49 1.26
Other 7.02 10.56 17.58
Total 83.70 — 65.09 148.79
16.07
12.53
21.80
2.20
9.38
14.84
0.02
7.95
0.04
—
2.52
1.35
6.64
4.54
14.22
0.57
12.87
4.16
22.73
25.02
36.06
2.77
24.77
20.35
49
-------
Type 1 Type 2
24.85
2.89
51.96
13.42
—
13.91
Type 3
43.68
10.66
21.63
16.23
60.39
95.21
All
Types
68.53
13.55
73.59
29.65
60.39
109.12
TABLE 6 (continued)
Monomers
Styrene
Vinyl Toluene
Acrylic Esters
Vinyl Acetate
Vinyl Chloride
Other
Total 107.03 — 247.80 354.83
Resins Purchased
Alkyd
Acrylic
Polyester
PVA
PVC
Epoxy
Epoxy
Urethane
Cellulosic
Amino
Rosin
Styrene
Phenolic
HC
Other
Total 235.76 360.52 24.86 621.14
Pigments
TiO2
Iron Oxide
Zinc Oxide
Zinc Chromate
Chromium
Metallic
Lead Chromate
Other Lead
Cadmium
Iron Blue
Other Inorganic
Total Inorganic 235.26 207.89 — 443.15
50
46.25
46.00
5.20
27.15
11.34
27.66
27.66
0.89
6.43
13.15
1.70
10.01
2.25
1.72
36.01
163.93
44.67
10.30
54.06
12.25
12.74
12.74
0.83
15.36
10.19
1.05
0.92
0.88
1.49
31.85
3.46
—
0.06
0.02
11.25
1.54
1.54
—
—
1.00
1.59
0.04
2.95
1.51
1.44
213.64
90.67
15.56
81.23
34.84
41.94
41.94
1.72
21.79
24.34
4.34
10.97
6.08
4.72
69.30
168.77
16.18
17.06
2.54
1.64
5.11
10.60
6.93
0.07
0.06
6.30
137.24
14.34
10.13
3.19
1.55
19.56
8.37
7.75
0.12
0.08
5.56
306.01
30.52
27.19
5.73
3.19
24.67
18.97
14.68
0.19
0.14
11.86
-------
TABLE 6 (continued)
All
Type 1 Type 2 Type 3 Types
Carbon blacks total 1.08 1.02 — 2.10
Organic total 7.42 2.58 — 10.00
Extenders total 247.36 245.88 — 493.24
51
-------
Acids are used in conjunction with polyols for resin production and are, therefore, not
purchased by type 2 plants. Phthalic anhydride is used in alkyd production, because being an
acid anhydride it has a fast reaction rate and produces minimum water. It accounts for 70% of the
usage or 104.5 million pounds. Isophthalic acid and maleic anhydride add another 10 to 12 million
pounds each. "Other" acids include fumaric and succinic acids.
Monomers provide individual molecules, acting as basic "building blocks", for polymers
which form unmodified resins. Significant quantities of monomers, other than those listed in the
questionnaire, were tabulated — the list has been expanded to include vinyl acetate and vinyl
chloride. In addition, "other" monomers are comprised mainly of the group indene, cyclopentadiene
and methyl cyclopentadiene. Usage of all monomers by type 3 plants is substantially higher than
for type 1 plants. This relationship is also reflected in the production of vinyl and acrylic resins,
in Table 5.
Approximately 34% of all purchased resin quantities are alkyds. This quantity obviously
does not represent the total consumption of alkyds by these plants. Since type 1 plants satisfy
some of their own needs for resins, while marketing other quantities and purchasing still others,
it is impossible to correlate production and consumption by resin types between Tables 5 and 6.
Quantities purchased by type 3 plants are intended for reformulation or merely to complete a
product line in resale. These purchases are small, with the exception of polyvinyl chloride which
accounts for better than half of their purchases.
Of the 31.8 million pounds "other" resins purchased by type 2 plants, almost 3 million
pounds were reported as asphalt, another 1.3 million pounds as chlorinated rubber and paraffin
and about 1 million pounds of vinyl acrylic. Similarly, 4 million pounds of vinyl acrylic contribute to
the 36.0 million pounds "other" resins purchased by type 1 plants.
The purposes of hiding, protection, and decorating a surface is served by pigments. Hiding
can be accomplished by white, opaque, pigments which reflect all wavelengths of incident light and
decorating by non-hiding pigments which impart a color to the coating by reflecting only a select
portion of the spectrum. In other cases, the color pigment has sufficient hiding power alone. It can
be seen from Table 6 that almost 71% of the inorganic pigments are titanium dioxide which has
an extremely high refractive index. Despite its higher cost per pound it is more economical per
unit hiding power.
Use of lead pigments, which at one time enjoyed great popularity, has been declining sub-
stantially due to the availability of higher hiding pigments such as TiO2. Recent legislation has also
52
-------
affected usage in trade sale finishes. A fair number of responding plants commented pertaining
to their use of these pigments. These comments included "... discontinued use. , .", "Other lead
... to be discontinued . . .", ". . . in traffic paint only.", and ". .. have phased out. ..". In all, planned
and implemented discontinuations in use of lead pigments accounted for 3% of this group; current
usage in traffic paints for 7.7%.
Raw materials listed under "extenders" include such items as talc, ground calcium carbonate,
silica, kaolin, barytes and mica. These materials, generally, have low refractive indices and little
hiding power by themselves but used in combination they provide important properties at low
cost. Extenders also permit the use of smaller quantities of higher cost pigments. Table 6 indicates
that usage of extenders actually exceeds that of pigments, by a small margin.
Table 7 presents a summary of solvent consumption by plant type as determined from
the questionnaire responses. Respondents were asked to report their solvents by group as follows:
1) Ethers 2) Olefinic ethers
3) Esters 4) Olefinic esters
5) Alcohol 6) Olefinic alcohol
7) Aromatic (Toluene & Xylene) 8) Branch chain ketone
9) Straight chain ketone 10) Olefinic ketone
11) Solvents containing a combination of hydrocarbons, alcohol, esters, ethers, or ketones
having an Olefinic or cyclo-olefinic type of unsaturation of 5% or more.
12) Solvent containing a combination of aromatic compounds with 8 or more carbon atoms
to the molecule, except ethylbenzene, of 8% or more.
13) Solvent containing a combination of ethylbenzene, ketones having branched hydro-
carbon structures, trichloroethylene, or toluene of 20% or more.
Grouped solvents which make up less than 5% of the total solvents coufd be lumped
together under "Other Solvents".
These classifications were based on the Los Angeles County Rule 66 type definition of
photochemically reactive solvents. Under this definition groups 2, 4, 6 to 8, and 10 to 13 are
considered photochemically reactive. This classification is used in a modified form in Table 7.
Groups 2, 4, 6, and 10 were reported in such minimal quantities that they have been simply
included in a separate category "Other Photochemical". Difference between "Rule 66" and the
classification system used here are found in categories 11, 12, and 13. Rule 66 considers solvents
to be reactive if they contain more than 5%, 8%, or 20%, respectively, of the objectionable substances.
53
-------
TABLE 7
SOLVENT USAGE, (MM Ib/yr)
(Questionnaire Data)
Ether
Ester
Alcohol
Aromatic
Ketone (Branch Chain)
Ketone (St. Chain)
Group 11
Group 12
Group 13
Mineral Spirits — Naphtha
Other Non-photochemical
Other Photochemical
Other
Photochemical
Non-photochemical
Unknown
Total
Summary:
Photochemical
Non-photochemical
Unknown
Total 1,268.31 100.0
Type 1
4.28
25.60
35.32
221.99
15.27
50.04
0.94
89.94
2.66
70.11
72.83
4.28
117.83
335.08
188.07
187.94
711.09
MMIb
583.11
400.83
284.37
4.59
23.29
25.59
105.72
9.53
28.38
11.36
27.60
6.23
49.25
39.73
0.95
19.38
161.39
121.58
68.63
351.60
%
46.0
31.6
22.4
Type 3
1.11
8.63
51.10
63.05
3.10
7.83
7.50
5.68
0.75
17.04
22.51
6.56
10.76
86.64
91.18
27.80
205.62
Industry
Total3
350
550
800
1 800
3390
54
-------
The present system includes the 5%, 8%, and 20% levels, respectively, in the reactive category.
The mineral spirits — naphtha group was added to accommodate the large quantities of
solvents reported as such. A qualification applies, however. If a solvent was reported as simply
mineral spirits it is included in this group. However, if a solvent is defined as "exempt" mineral
spirits (or naphtha, VMP, etc.) it is included in the category "Other Non-photochemical". The
category "Mineral Spirits — Naphtha" no doubt contains material which could be classified photo-
chemical according to the above definition as well as material which is non-photochemical.
Some companies reported solvents as simply "aliphatic". These were also included in the
group "Other Non-photochemical".
The "Cellosolve®" type solvents present some difficulty in classification. A series of glycol
ethers and esters of glycol ethers find extensive use as solvents. These are commonly known as
"Cellosolves®" due to their use as solvents for Cellulose® derivatives. One such compound, ethylene
glycol monoethylether has the following structure:
H H
I I
HO - C - C - 0 - C9H,
II 2 5
H H
An ester derivative of this, ethylene glycol monoethylether acetate has the structure:
H H
I I
CH.COOC - C - 0 - C9H,
3 I I 2 5
H H
The first of these compounds, Cellosolve®, could be classified as either an alcohol or as an
ether. For the purpose of this study Cellosolve® and its related compounds (e.g. methyl Cellosolve®,
butyl Cellosolve®, etc.) are classed as ethers.
The second of the above structure, Cellosolve® acetate, could be classed as an ether or as
an ester. For this study, Cellosolve® acetate and related compounds (e.g. butyl Cellosolve® acetate,
etc.) are considered esters.
Table 7 also contains totals with respect to photochemical properties. The "unknown"
group includes the "other" category as well as the mineral spirits — naphtha. Solvents that can
definitely be classed in photochemical groups make up 46% of the total. The largest single group
of solvents is the aromatics (toluene and xylene) which make up 31% of all solvents used.
It was often necessary to convert from data reported in gallons to a weight basis. Where a
specific compound was reported it was possible to use an actual density from the literature. Where
55
-------
mixtures or general classes were listed, an average value for solvents of the particular type was
used. Some representative conversion factors for mixtures and general classes are listed as follows:
Solvent Pounds/Gallon
Alcohols 6.7
Ketones 6.7
Esters 7.3
Aromatic 7.24
VMP Naphtha 6.3
Mineral Spirits 6.5
For solvents listed under "other" a value of 7.0 pounds/gallon was used.
Two other categories of material that have been listed in the questionnaire are drying
agents and mercury compounds. These are used in relatively small quantities, but may be important
from an environmental standpoint.
Drying agents are usually organic acid salts of lead, cobalt, zirconium, manganese, or
calcium. Lead based driers have historically been the most common. However, this is in the process
of changing due to increasingly stringent regulations on lead content. The questionnaire responses,
unfortunately do not list separate drier categories.
Mercury compounds, usually phenyl mercury based, are used as preservatives and fungicide.
Limitations on mercury content in paint are presently under consideration by various groups.
Usage of mercury compounds and drying agents as reported in the questionnaire is summarized
in Table 8.
2. Products and Production
In the Industry Questionnaire, production of coatings was broken down by oil/solvent base
paints, water base paints, varnishes and lacquers. Similarly resin production was reported by
fourteen major types. Production of coatings is in million gallons and that of resins in million pounds.
A summary of the production data tabulated is presented in Table 5. As described earlier, type 1
plants produce both coatings and resin, type 2 plants produce coatings only, and type 3 plants
produce resins only.
Total coatings production for the plants included in the table amounts to somewhat over
30% of the industry's 1972 production — total production estimated to be 930 million gallons based
on Current Industrial Reports (and estimates for December 1972). This 30+% of the total represents
a broad sample of plants since the emphasis in selecting was on diversity of plant type and size.
56
-------
TABLE 8
DRYING AGENTS AND MERCURY COMPOUNDS
(Questionnaire Data)
Drying Agents Hg Compounds
Type of Plant (Ib/yr) (Ib/yr)
1 6,427,484 695,977
2 4,399,769 637,878
3 46.892 1.898
Total 10,874,145 1,335,853
57
-------
Emphasis, in this sample, on plants with emission control devices skewed the distribution towards
larger plants which resulted in a sample consisting of 18% of all plants but 30% of all coatings
production. More than half of the coatings for type 1 and type 2 plants are oil/solvent base which
is fairly typical of the industry.
While trade sale finishes accounted for 49% of U.S. production in 1972, plants in Table 6
indicate that 57% of their products fall into this category. This results from including in the sample
a few very large type 1 plants specializing in sales to the consumer rather than to industry.
Table 5 also shows products and production under the category of resins and varnishes.
Although alkyd resins are facing increasing competition from acrylics, epoxies, polyvinyl acetate
and polyurethanes they still account for about 50% of all resins produced and consumed. Less
than 10% of the alkyds produced find uses in areas other than coatings. Production of alkyd resins
is distributed about equally in the industry between plants making resins only and those that produce
both resins and coatings. In our sample, however, distribution of alkyd production is weighted
towards those making both. Type 3 plants, on the other hand, produce substantially larger quantities
of acrylic, polyester, polyvinyl chloride, amino and hydrocarbon resins than do type 1 plants.
The largest single component in the "Other" resins category is vinyl acrylic amounting to
20 million pounds or almost one-third of the production listed under this subtitle. The remainder is
comprised of small volume resins such as chlorinated rubber, asphalt, silicones, and natural resins.
3. Process Equipment
a. Dispersion and Grinding Equipment — The principle manufacturing steps involved in producing
a finished coating from raw materials consist of various types of milling and dispersion steps.
Milling consists of a reduction in the size of the primary pigment particle. Dispersion consists of a
deagglomeration or separation of aggregates of individual particles, wetting of particles and agglom-
erates, and a uniform distributing of particles throughout the liquid phase. Pebble and ball mills
accomplish size reduction as well as dispersion. Roll and sand mills are primarily used for dispersion.
Various types of high speed dispersers, disc impellers, etc. are used for dispersing easily dispersed
pigments. A final category of simple mixers or blenders is employed for such purposes as thinning,
shading, and other finishing operations.
A portion of the questionnaire was devoted to a listing of such devices. A summary of
this information is presented in Table 9. A problem exists here of terminology. The terms "mills",
"mixers", "dispersers", etc. often have imprecise meanings as far as every day usage in the paint
58
-------
TABLE 9
MILLS, MIXERS, ETC.
(Questionnaire Data)
Number
Type 1 Type 2 Type 3
Roll 126 152 5
Pebble 336 506 0
Sand 299 289 5
Ball 382 282 1
Other 2741 2043 128
59
-------
industry is concerned. The roll, pebble, sand and ball mills have a definite meaning to most people
and it is felt that these have been reported in an unambiguous manner. The "other" category
apparently includes an undetermined, but perhaps large, degree of duplication. In many cases,
the same unit is apparently listed as, for instance, a "mixer" in one part of the questionnaire and
as some other type of device in another part. The numbers reported in the "other" category are
probably too high for reasons outlined above.
b. Solvent Storage Tanks — The questionnaires requested data on all solvent tanks over 5,000
gallons covering size, turnover, type of solvent, vapor pressure, and type of control if any. A summary
of total number of tanks, number of uncontrolled tanks, and turnover is given in Table 10. Uncontrolled
tanks comprise 41% of the total. Total quantity of solvent handled in 5,000 gallon or larger tank
can be calculated to equal 1,130 million pounds a year using a density of 7 pounds/gallon. This
represents 89% of the total consumption of 1,268 million pounds reported earlier in Section 1.
Controlled vents, as reported in the questionnaires, consisted almost entirely of conservation
vents. In general, tanks containing high vapor pressure solvents were more likely to have con-
servation vents though this was by no means universally observed. Filling losses are relatively
unaffected by use of conservation vents. Vapor losses during filling operations are proportional,
on a weight basis, to the product of vapor pressure and molecular weight. Some theoretical filling
losses have been calculated for some selected solvents and are given in Table 11. The calculation
assumes that the gas displaced from the tank being filled is saturated with the solvent in question
at a temperature of 20°C. A pumping rate in excess of 144 gal/min for toluene will give an instantaneous
emission rate in excess of the 8 Ib/hr limit often invoked for photochemically reactive solvents.
Whether this is in violation depends on local interpretation of local regulations and is beyond the
scope of this section. It can be pointed out that these losses represent a very small percentage
of the solvent being handled. On the other hand, a given quantity of solvent undergoes several
fill-empty cycles as it travels through a plant. These are shown schematically in Figure 5. A given
plant may have some or all of these cycles. Each is associated with a vapor loss due to pure dis-
placement of the gas above the container as it fills. The amount actually lost in each step will vary
as the vapor pressure varies due to temperature, solutes, and degree of saturation of the gas in
the vapor space.
c. Resin Reactor Usage — The amount of resin reactor volume attributable to various types of
operations is given in Table 12. A total of 371 kettles were reported by type 1 plants and 218 by
type 3. The average kettle volume is considerably higher for type 3 than for type 1. This is partly due
60
-------
TABLE 10
SOLVENT TANKS OVER 5,000 GALLONS
(Questionnaire Data)
Number of Tanks
Number Uncontrolled
Turnover (MM gal/yr)
Type 1
681
185
57.06
Type 2
497
278
44.01
Type 3
205
106
60.84
Total
1,383
569
161.91
TABLE 11
FILLING LOSSES FOR SELECTED SOLVENTS @ 20°C
(Questionnaire Data)
Solvent
Acetone
Ethyl Acetate
Toluene
Mineral Spirits
V.P.
@ 20°C, mm
186
74
22
2 (est.)
Molec. Wt.
58
88
92
160 (est.)
Filling Loss,
lb/100gal
0.494
0.299
0.0927
0.0147
61
-------
^
»-
VAPOR
L.OSS
*
1
1
>
HQ*
^m1"
4
UJ
III
1
1 '
z|
hi-
u
RESIN
REACTOR
u
SOLVENT
STORAGE
TANK
i
(T
y
— ?
0
u
,
1
h
u
T
Q
I
(/
in
w
J
H
,
f
t
}
]
h
z
oi
w 2
u/
IGURE
L
IT
it
EET FO
PAINT —
I
QO
It
I
62
-------
to the fact that type 1 plants have a large number of varnish kettles and these are generally
rather small.
Table 13 gives the number of plants which operate a particular type of reactor. The alkyd
kettles are divided into those which are used primarily for fusion cooking and those which are
primarily for solvent processing. Examination of the data indicates that, in terms of gallons, solvent
processing is strongly favored by type 1 plants while the opposite is true for type 3 plants. In terms of
numbers of plants, however, the preferences are much less pronounced.
The production of water based emulsions tends to be concentrated in the type 3 plants.
Varnishes, on the other hand, tend to be produced by type 1 plants. Varnishes in this context are
whatever the plants responding chose to call varnishes. Most of the time they consisted of the
classic oleoresinous type.
The tendency to operate on a 24 hour basis increased as the size of the resin plant increased.
Two shift operation was most common in the small to medium size plants. The large producers were
almost always 24 hour operations. A few of the small producers reported a single 8 to 10 hour shift.
C. Material Balance For Model Plant
One of the purposes of this industry study is to determine the financial impact of air pol-
lution control on the paint industry. To accomplish this purpose a model plant was developed and
has been used to:
1. Develop an operating statement for the uncontrolled plant.
2. Develop an operating statement for the plant using best control equipment and compare
with above.
3. Develop plant balance sheet to show the effect of capital investment for air pollution
control equipment upon assets, liabilities and equity.
Sherwin-Williams has been retained as the subcontractor to develop the model plant economics.
1. Design Basis
The major features for the model paint plant were based on the design contained in a
paper presented by Mr. R. F. Brewster at the National Paint and Coating Association meeting on
October 31, 1972.9 This design resulted from a study by the Management Committee of the NPCA.
By using this information, it was possible to take advantage of the expertise of those in the industry,
as well as provide a common ground with the industry. Copies of some of the slides and calculation
sheets on which the paper was based were provided by Mr. Brewster.
63
-------
TABLE 12
REACTOR USAGE, GALLONS OF KETTLE VOLUME
(Questionnaire Data)
Type 1 Type 3 Total
Alkyd
Fusion
Solvent
Varnish
Water Emulsion
Other Resin*
Other or Unspecified**
45,955
143,900
64,788
47,977
116,450
33,325
82,080
47,000
10,070
154,000
243,325
45,820
128,035
1 90,900
74,858
201 ,977
359,775
79,145
Total 452,395 582,295 1,034,690
*Such as polyester, epoxy, solvent based acrylic, etc.
"Includes heat bodying oils.
TABLE 13
RESIN PROCESSING, NUMBER OF PLANTS
(Questionnaire Data)
or Solvent)
3d**
Type 1
76
45
23
33
34
20
15
22
Type 3
39
16
10
9
9
9
13
8
Total
Alkyd (Fusion ai
Fusion
Solvent
Varnish
Water Emulsion
Other Resin*
Other or Unspecified**
*Such as polyester, epoxy, solvent based acrylic, etc.
"Includes heat bodying oils.
64
-------
The model plant developed for this study differs from the NPCA plant in two important
respects: (1) a resin plant has been added to produce trade sales and industrial alkyds; and (2) the
plant was modified to operate on a two-shift rather than single-shift basis. The first of these changes
necessitates additional facilities for handling the resin raw material as well as space and equipment
for resin production itself. The second change requires additional space for raw material and
finished product storage and handling as well as increased labor requirements. The paint manufactur-
ing equipment itself (mixers, dispersers, filling equipment, etc.) remains unchanged.
The NPCA plant was designed to produce one million gallons per year on a single-shift
basis. In projecting to two shifts, one should take into account the fact that some of the ball or
pebble milling processes require more than eight hours. Thus, going to two shifts will not necessarily
double the output of these operations. These, however, represent a small percentage of total
production. Of more importance is the question of labor efficiency. Industry personnel have indicated
that the efficiency of the second shift tends to be less than that of the first shift. Efficiency drops
even further if a third shift is used. Based on their comments, it will be assumed that the plant
produces at 90% efficiency on the second shift. The plant, then, will produce 1,900,000 gallons
per year. It is assumed that the product distribution is the same as that for the NPCA plant. This
distribution, with the revised gallon production, is given in Tables 14, 15, and 16. Table 16 contains,
in addition, the resin plant output, which will be discussed later.
It may be noted that the industrial coatings listed as "Other Solvent Based" were specified
to include the following:
Gallons Type
95,000 Monomer modified alkyd for fast dry coatings
95,000 Acrylic baking enamel
190,000 Alkyd urea baking enamel
In selecting these particular types, the size and expected level of technical sophistication of the
model plant were taken into consideration. These resins are to be purchased from outside suppliers
rather than manufactured in the plant.
2. Production and Inventory
The design of the NPCA plant is based on certain assumptions concerning production
rates and inventory. For consistency, these assumptions have been retained for this study. The
objective is to turn trade sales inventory six times a year and industrial inventory 18 times. That is,
the average inventory levels are 162/3% of annual trade sales shipments and 55/9% of annual
65
-------
TABLE 14
MODEL PLANT
PRODUCT MIX
60% Trade
70% Latex
30% Alkyd
40% Industrial
50% Alkyd
50% Other
Total
1,140,000 Gallons
800,000 Gallons
340,000 Gallons
760,000 Gallons
380,000 Gallons
380,000 Gallons
1,900,000 Gallons
TABLE 15
MODEL PLANT
TRADE SALES COLOR DISTRIBUTION
Whites and tint bases
Tints (12 shades)
Solid colors (7 colors)
Total
Latex
800,000
Alkyd
340,000
Total
673,000
94,000
33,000
283,000
33,000
24,000
956,000
127,000
57,000
1,140,000
66
-------
TABLE 16
MODEL PLANT
PRODUCT TYPE
Latex
Alkyd
Total
Trade sales
Outside H.P.
Outside trim
Inside flat
Inside semi-gloss
Inside gloss
Sub-total
% gal
14.4 275,000
27.6 525,000
42.0 800,000
%
0.83
11.78
5.39
18.00
gal
15,800
224,000
100,200
340,000
%
14.4
0.83
27.6
11.78
5.39
60.00
gal
275,000
15,800
525,000
225,000
100,200
1,140,000
Industrial sales % gal
Alkyd 20 380,000
Other solvent based 20 380,000
Sub-total 40 760,000
Alkyd resin
Long oil trade sales
Medium oil industrial
Short oil industrial
Total
Pounds (NVM)
1,025,000
285,000
856,000
2,166,000
67
-------
industrial shipments. It is further assumed that production rates are held uniform throughout the year.
Shipments of trade sales products will fluctuate through the year as shown in Figure 6. Industrial
shipments are uniform.
The consequences of these assumptions are that trade sales inventory must successively
increase and decrease to accommodate the annual sales curve. The typical trade sales inventory
curve for this plant is shown in Figure 7. Inventory peaks at 29% of annual sales in March and
declines to a low of 5% of annual sales in September. The finished goods inventory, including its
distribution by type of container, is shown in Table 17. Annual production and maximum inventory
levels by various categories are presented in this table. The distribution is as given for the NPCA
plant, whereas the actual numbers have been revised upwards to account for the increased yearly
production assumed for the present plant.
These assumptions may be somewhat unrealistic in some respects. First, the uniform
production rate may not be representative of the way many plants operate. Second, the minimum
trade sales inventory of 5% of annual shipments may be somewhat low. Both these points were
discussed in the NPCA report and it was felt by the authors that the plant, nonetheless, provided a
useful base from which to relate other situations.
3. Equipment Requirement
a. Paint Plant— Equipment requirements worked out for the NPCA plant are unchanged for this
study. The product distribution is the same, while increased production is accomplished solely by
the addition of another working shift. The equipment summary is presented in Table 18.
It is anticipated that high speed dispersion will be suitable for almost all the latex paints,
as well as for a relatively large portion of the architectural solvent finishes. The sand mills will
process those items having an intermediate difficulty of dispersion. The output of the sand mills is
assumed to be at an hourly rate five to ten times the size designation. Neither the high speed
dispersers nor the sand mills reduce particle size (i.e. grind).
The pebble mills will be used primarily for bright colored industrial finishes and for certain
trade sales products. Ball mills will see relatively limited use — primarily hard to disperse dark colors.
The ball and pebble mills were selected to provide a continuous capacity range rather than
to provide a given output. Mills of this type are assumed to have an effective working capacity of
20% to 50% of actual capacity. Outside this range, they do not operate properly. Figure 8 illustrates,
on a logarithmic scale, the grinding capacities of the ball and pebble mills. Since the scale is loga-
rithmic, each mill is represented by a bar of equal length. This chart portrays the degree of overlap
68
-------
0)
V
LJ
(/)
S
LJ
£
(D
IL
Q:
LJ'
u
n
0
(fl
§
tf)
8
6
u
0
1
LJ
-------
'N
0>
Q:
uj
0
T
u
E
(D
V
1C
U
.0.
(C
0
(0
^
j-
<
H
TI
\
(t
o^~
Zfl
hJ<
3?
i
(0
tf> 0N
a:
I-
^
£
70
-------
TABLE 17
MODEL PLANT
FINISHED GOODS INVENTORY — MAXIMUM PROJECTION
Annual
Maximum Inventory
Trade
Gallons
Drums
5
1
1/4
1/8
Industrial
Gallons
Drums
5
Total
Gallons
Drums
5
1
1/4
1/8
Production Packages
(gal) (number) Gallons
1,140,000 330,000
0
22,800
912,000
420,000
91,200
760,000 42,000
10,400
38,000
1,900,000 372,200
10,400
60,800
912,000
420,000
91,200
Number
0
6,600
264,000
119,000
26,400
575
2,110
575
8,710
264,000
119,000
26,400
71
-------
Pebble mills
1
1
1
1
1
1
Ball mills
1
1
Sand mills
1
1
1
High speed dispersers
TABLE 18
MODEL PLANT
EQUIPMENT SPECIFICATIONS
Diameter
21 in.
24 in.
32 in.
42 in.
60 in.
72 in.
32 in.
48 in.
10H.P.
25H.P.
50 H.P.
- 100 H.P.
28 in.
36 in.
36 in.
48 in.
72 in.
96 in.
32 in.
60 in.
125
470
3
8
16
Mixing tanks
Twin fillers
Associated filling equipment
1 — Labeller
1 - Bail-0-Matic
1 — Packing Station
1 - Case Sealer
Number
20
14
8
4
"Instantaneous" rate
24 one-gallon/min
40 quarts/min
60 pints/min
Size, gallons
220, portable
550
1,100
2,200
Filtering equipment
1 — Two cartridge
3 — Six cartridge
2 — Vibratory screens
72
-------
which exists between successive pebble mills. In this way, a continuous pebble milling capacity
ranging from 5.7 to 702 gallons is obtainable.
Mixing and finishing tanks are of two types, fixed and portable. The 220 gallon portable
tanks are to be used in conjunction with the high speed dispersers. The larger sizes are floor
mounted and equipped with belt driven turbine drives and electric motors.
Two twin filler machines are provided. A single unit operating at an 80% utilization rate can
theoretically handle all the filling operations, but it was felt that this represented a too narrow safety
factor. The filling machines can be moved back and forth under the tanks.
b. Resin Plant — Since the NPCA plant did not include resin production facilities, it was necessary
to start from scratch in designing this part. It was necessary to first estimate the amount of alkyd
resin required. The size of the resin plant and, finally, the specific equipment required were then
determined.
It has been assumed that the plant will produce all of its own alkyd resins. It will further be
assumed that one-half of the industrial output consists of alkyd based coatings. For the NPCA plant,
this means that 180,000 (trade) + 1/2 x 400,000 (industrial) or 380,000 gallons of alkyd based paints
will be produced. Scaling up to two shifts, we get 1.9 x 380,000 = 722,000 gallons.
Several workable, up-to-date, paint formulations for alkyd type coatings have been supplied
by Ashland Chemical Company. Based on these formulations (two of which are given in Table 19)
an average resin content of about three pounds solids per gallon of paint is representative for the
type of products this plant would produce. The required alkyd production is 3 x 722,000 = 2,166,000
pounds of resin solids per year.
The size of the resin reactors required for this amount of production was determined by
reference to material presented earlier. As discussed in Section A-2 of Chapter 1, industry practice
suggests that a medium size alkyd kettle produces about 5,000 pounds of resin solids per batch
per 1,000 gallons of reactor volume.
Even if the resin plant is operated on a two shift basis, only a single batch can normally be
processed per day due to the cooking times (8 to 16 hours) required. Assuming 250 working days
per year, 1,730 gallons of kettle volume are required as a minimum to produce the necessary resin
for the model plant.
In determining the exact configuration of the resin plant, three aspects were kept in mind;
efficiency, flexibility, and cooking times. In order to obtain the efficiency of large batches and still
retain the flexibility to handle a broad product line, it was decided that two reactors be installed. A
73
-------
ft
Q
ff
r U
In
s
o:
T ^
•*• L.'
fli T
o 1 o:
0) D
0
-
%
T
NX
J * (/)
3 « -, -1 T J g
^ t | X
IJ s* 00
S • T
-------
TABLE 19
MODEL PLANT
PAINT FORMULATIONS FOR ALKYD TYPE COATINGS
"Exempt" Air Dry Gloss Interior-Exterior Architectural Enamel
Ib gal
Rutile titanium dioxide1 306.75 8.99
Suspension and sag control agent2 5.11 0.35
Loss of dry inhibitor3 6.13 0.56
6% Calcium naphthenate 3.27 0.42
Long soya alkyd solution (70% NVM)4 221.90 27.70
"Exempt" mineral spirits 73.62 11.25
Disperse to 7+ Grind (Hegman)-Cowles or Pebble Mill
Long soya alkyd solution (70% NVM in "exempt"
mineral spirits)4 327.20 40.90
6% Cobalt naphthenate drier 3.07 0.39
6% Zirconium drier 10.22 1.43
Anti-skinning agent5 1.02 0.14
"Exempt" mineral spirits 51.12 7.87
1,009.41 100.00
1Titanox® 2060 Titanium Pigments Div. N.L. Industries; or equal
2Bentone® 38 Chemical Div. N.L. Industries; or equal
3LFD® Mooney Chemicals; or equal
"Aroplaz® 1266-M-70 Ashland Chemical Co.; or equal
5Exkin No. 2® Tenneco; or equal
75
-------
TABLE 19 (continued)
Air Dry Lead Free Yellow Enamel (Industrial)
Ib gal
Organic Yellow1 56.40 4.78
(Short tall oil fatty acids alkyd-50% NVM in xylol) 141.00 16.99
VM&P Naphtha 69.30 10.72
Toluene/VM&P 1/1 blend 100.00 14.58
Grind 24 hours porox balls
(Short tall oil fatty acids alkyd-50% NVM in xylol) 313.00 37.71
Butanol 32.60 4.82
Toluene 44.00 6.08
6% Cobalt octoate drier 1.52 0.21
6% Calcium naphthenate drier 30.63 3.94
6% Manganese naphthenate drier 0.76 0.11
Anti-skinning agent2 0.51 0.06
795.39 100.00
1DuPont's Dalamar YT-808D; or equal
2Exkin #2 Tenneco; or equal
76
-------
degree of excess capacity was built in to compensate for equipment malfunctions, etc. and to allow
increased production over and above the minimum without incorporating a third working shift.
In order to insure that all formulations can be processed in two working shifts, the largest
kettle should be less than 2,000 gallons in size. Data from the questionnaire sample indicates that,
as expected, process times tend to increase with the size of the batch. Up to 1,500 gallons kettle
size, processing times were almost always less than 16 hours. For the larger kettles, times in excess
of 16 hours were sometimes reported.
In view of these considerations, two reactors were specified. Their capacities are 1,500
gallons and 500 gallons, respectively. In summary, the resin plant description is as follows:
1. 1,500 Gallon Reactor — A reactor system similar to that shown in Figure 9 will be provided.
This system can be operated as either a fusion reactor or as a solvent process reactor.
Material: Stainless Steel
Heat Source: Dowtherm® (Typical)
Thin Tank: 3,000 Gallons Capacity
2. 500 Gallon Reactor — Also similar to Figure 9 except that the condensers and decanter-
receiver are omitted. This system operates only as a fusion reactor. The material of construction
and heat source are the same as for the larger reactor. The thin tank is 1,000 gallons capacity.
c. Tankage Requirements — Tankage requirements are based on the assumption that the plant
should maintain a minimum of one month's supply of raw materials. It was further assumed that
liquid raw materials would be delivered in 5,000 gallon quantities. The tank volume necessary to
maintain a minimum inventory of 30 days' supply can be calculated, then, by dividing the annual
requirements by twelve and adding 5,000 gallons to the result. For instance, 216,000 gallons per
year of odorless mineral spirits are required. The tank required is 216,000 4- 12 + 5,000 = 18,000
+ 5,000 = 23,000. A summary of the tankage requirements (sometimes rounded off to a convenient
size) is presented in Table 20. Industry sources have suggested that while this total tankage
ought to be adequate, in practice it would probably consist of a larger number of smaller tanks
4. Raw Materials
Raw material requirements were determined by simply scaling up the NPCA raw materials
sheet by a factor of 1.9 where appropriate. Additions and modifications have been made to accom-
modate the resin production. These changes and additions occur in the solvents, miscellaneous
industrial resins, oil, glycerine, and phthalic anhydride. The latter three quantities were not present
77
-------
REFLUX
COINDEirxlSElR
SPRAY TOWER
CONDENSER
FRACTIONATING
DISTILLATION!
COLUMN
DECAISITER
RECEIVER
SCRUBBER
REACTOR
PORTHOLE
FOR SOLIDS
OVERFLOW
CONDENSER
TMINNINO
TANK
DIRECT FIRED OR
JACKETED FOR HIC3H
TEMPERATURE VAPOR
R LIQUID
TO RESIN
STORAOE
RC3URE Q
MODERN RESIN PRODUCTION SYSTEM
78
-------
TABLE 20
MODEL PLANT
SUMMARY OF TANKAGE REQUIREMENTS
Size
(gallons)
25,000
25,000
12,500
12,500
7,500
17,500
22,500
10,000
5,000
10,000
7,500
1,000
1,000
1,000
1,000
1,000
1,000
1,000
Contents
OMS
Xylene
Oil
Oil
Glycerine
Exterior latex
Interior latex
Trade alkyd
Trade alkyd
Industrial alkyd resin
Industrial alkyd resin
Industrial alkyd
Industrial alkyd
Industrial alkyd
Industrial alkyd
Waste solvent
Waste solvent
Aqueous waste
Location
Outside
Outside
Inside, lower level
Inside, lower level
Inside, lower level
Inside, lower level
Inside, lower level
Inside, lower level
Inside, lower level
Inside, lower level
Inside, lower level
Inside, lower level
Inside, lower level
Inside, lower level
Inside, lower level
Outside
Outside
Inside, lower level
79
-------
at all in the NPCA plant.
The raw materials requirements for resin production are based on the following assumptions:
1. Trade sales alkyds consist of long oil types at an average of 65% oil.
2. 1/4 of the industrial alkyds are medium oil at an average of 50% oil.
3. 3/4 of the industrial alkyds are short oil types at an average of 40% oil.
A weighted average can be calculated as follows:
1.8/3.8 x 0.65 = 0.308
0.5/3.8 x 0.50 = 0.066
1.5/3.8 x 0.40 = 0.158
The average alkyd contains 53% oil. Assume also that, on the average, PA and glycerol
are present in a 2:1 weight ratio. This represents an excess hydroxy content (based on PA and
glycerine only) of about 20%. Resin raw materials can be calculated as follows:
Basis 1,140,000 pounds resin solids (one shift):
1,140,000 x 0.53 = 616,000 pounds oil
Correction for 5% loss (mostly water): 1.05 x 1,140,000 = 1,200,000 pounds
Glycerol + PA = 1,200,000 - 616,000 = 584,000 pounds.
Glycerol: 584,000 x 1/3 = 195,000 pounds
PA: 584,000 x 2/3 = 390,000 pounds
For two shift operation, assuming 90% efficiency on the second shift, the approximate
requirements are:
1,170,000 pounds oil
370,000 pounds glycerine
740,000 pounds PA
By volume the liquid requirements are:
1,170,000/7.64 = 153,000 gallons oil (as Soya)
370,000/10.5 = 35,000 gallons glycerine
In addition to these materials, solvent requirements over and above those in the NPCA
plant must be met. Since the alkyds are to be produced in the plant, the solvent that otherwise
would have been a part of the purchased resin solution must be supplied separately. For the trade
sales alkyds, this solvent will be an odorless mineral spirit. The industrial alkyds will require a
stronger solvent such as xylene or an equivalent exempt solvent system. These solvent require-
ments have been incorporated in the raw materials data sheet which is given in Table 21. This is
80
-------
followed by Table 22 which lists packaging material requirements. These tables contain some
revised entries prepared by Sherwin-Williams based on the NPCA report and on the resin require-
ment outlined earlier.
5. Labor Requirements
In revising and expanding the NPCA manning chart, the following assumptions were made:
1. The numbers and types of personnel in the factory and filling areas are the same for
the second shift as for the first shift.
2. All shipping and receiving is done on the first shift.
3. The resin plant requires two men on each shift.
4. Shipping and receiving personnel need to be increased by 50% to handle the increased
capacity.
These and other changes have been incorporated into the revised manpower requirements
shown in Table 23. A manning chart for the plant is given in Figure 10.
6. Plant Layout and Flow Sheet
A flow sheet for the plant is given in Figure 11. Prints of the plant layout incorporating
the design changes necessary to accommodate the resin plant and increased capacity are pre-
sented in Figures 12 and 13. In order to operate under the same storage and inventory assumptions
as the NPCA plant, it will be necessary to nearly double warehouse, raw material storage, and
shipping and receiving areas. This is in addition to any plant expansion needed to accommodate
the resin production facilities. The manufacturing and filling areas are essentially unchanged from
the NPCA plant. A site plan is given in Figure 14.
II. EMISSIONS
There are two major types of emissions from a paint plant. These are confined (or
non-fugitive) and fugitive. Non-fugitive emissions are those that are collected by and confined
within an exhaust system with or without an air pollution control device. Fugitive emissions are
those that escape into the plant atmosphere from various operations and exit the plant buildings
through the doors and windows in an unregulated fashion.
A. Description of Emissions
These two types of emissions can both be further subdivided into gaseous and particulate
emissions. Details of each type of fugitive and non-fugitive emissions are discussed on the following
page.
81
-------
TABLE 21
MODEL PLANT
ANNUAL RAW MATERIAL CONSUMPTION
(FINISHED OUTPUT OF 1.9 MILLION GALLONS)
Raw Material
Pigments & Extenders
Titanium Dioxide
Misc. Colors
Silica
Calcium Carbonate
Talc
Clay
Resins — Manufactured
Soya Oil
Phthalic Anhydride
Glycerine
Resins — Purchased
Exterior Latex (11.42 Ib/gal)
Interior Latex (10.6 Ib/gal)
Aropol 830-V-60
Aroplaz 7435-XM-50®
Resimene V-920®
Acryloid AT-5 1 ®
Epon 1001X75®
Solvents & Miscellaneous
Driers (9.8 Ib/gal)
Misc. Solids
Misc. Liquids (8 Ib/gal)
Water
Odorless Mineral Spirits
Anti-Skin (7.72 Ib/gal)
Toluene
Xylene
Aromatic 100 (Solvesso 100)
Aromatic 150 (Solvesso 150)
VM + P Naphtha
Butanol (6.74 Ib/gal)
Cellosolve® Acetate
Triethyl Amine (6.07 Ib/gal)
Trade Sales
Industrial Finishes
Pounds
2,039,000
27,500
1,179,000
893,500
475,850
199,700
714,000
276,000
138,000
1,644,500
2,141,200
Gallons
55,890
29,200
241,600
26,402
272,300
169,500
Pounds
760,000
380,000
232,750
522,500
492,000
470,000
247,000
400,000
870,000
152,000
510,000
60,600
81,340
30,400
22,774
Gallons
20,894
67,300
577
45,600
221,400
20,900
6,975
32,650
82
-------
TABLE 22
MODEL PLANT
ANNUAL PACKAGE & PACKAGE MATERIAL REQUIREMENTS
(FINISHED OUTPUT OF 1.9 MILLION GALLONS)
Trade Sales Industrial Sales
Package Type Quantity Quantity
Packages
Drums (55 gal) -- 10,364
Pails (5 gal) 22,800 12,350
Pails (5 gal) (Interior Coated) -- 25,650
Cans (gal) 273,600
Cans (gal) (Interior Coated) 638,400
Cans (qts) 205,200
Cans (qts) (Interior Coated) 205,200
Cans (pts) 91,200
Labels
5 gallon size 22,800 38,000
Gallon size 912,000
Quart size 410,400
Pint size 91,200
Cartons
Gallon size (4 gals) 228,000
Quart size (6 qts) 68,400
Pint size (12 pts) 7,600
Pallets (30" x 42" Size)
Drums 5,182
5 gal pails 1,425 2,375
Gallons 9,500
Quarts 977
Pints 136
83
-------
TABLE 23
MODEL PLANT
LABOR REQUIREMENTS
No. People
Salaried
1
1
1
1
1
1
2
3
1
Hourly Paid Employees
1st Shift Workers
15
2
2
2
2
12
2nd Shift Workers
10
2
2
2
2
5
Job Title
Plant Manager
Secretary
Supervisor 1st Shift
Supervisor 2nd Shift
Supervisor Warehouse & Shipping
Quality Control Chemist
Quality Control Technician
Clerks 1st Shift
Clerk 2nd Shift
Various Classifications
Resin Plant Operators
Fillers — Industrial
Fillers — Trade
Caser & Palletizer
Stockman
Various Classifications
Resin Plant Operators
Fillers — Industrial
Fillers — Trade
Caser & Palletizer
Stockman
84
-------
z<
< z
a. 2:
a:
UJ
co
cc
o
co a
cc <=
LU LU Q-
Q. CO i—
^ X X
i/> 2 co
X
CO
CVJ
X
CO
o
t—
CO
IO
a:
LU
_)
o
O
t-
co
>- — IH
I— O CO
< z u.
=) C X
I-O
CO LU
00
Z LU
LU a
_J C
(X LU
O O- —I
LU <
a. a H-
— o
a < H-
"— Q.
< z
a. D —
LU
>- ~ LU
_l CC _l
a: < a.
3 _l O
O < LU
X CO Q.
S* IJ2
aS
CO Ll_
CC CO
LU
a. c
O H*
CO U.
> x
o: co
LU
Q. h-
Z> CO
CO _
H
CO LU CO
a Q -
LU =3
s: a: CL
c t-
i— ^
o o
U- Q
LU
X
CO
85
-------
_
(
OLVENT
(flt
ff in
%
£o ••
I
J1
il
1
t
0
0
J
L
1
zffi
^
L
i
..7
INDING
=»ERSIOI
§§
I
I
1
y
IT
Q.
s ,
jf 1 I.I
PAINT RX
MATERlAI
STORAGE
1
in
\)
2
>
y
ti
o:
1
(
y
3
0
- I
y
o:
<
*
y
7®
i=<
RES
STOR
1$
s<
KIL
>ir>J RAW
CTERIAL
•ORAGE
'i^r
j/2«
to
7
0^
ho
1 IL
I
(fl
86
-------
I-
\l
§ I
°.< I
** ,
h
44,
2 0
f
1 •
1
ri I-H-I-H
hhhi^
FT^FI
H
H
H
-
--
H
-H
^ i
4
~] •
U
>^
j- J
(- -
h"
Li
h-
H
hj
H
hi
r
I
^fH
5
n
0
^
3
i
a
8
o
«
rr
i
It
K ~
B
o
h
j ^
i- j
Q. Ul
r!0
87
-------
' 2
'4
h
c^J
cSf
I5
n
8°^
-
\\
$
h
LS
ifi
J
^
E
LJ
I-
Z
Q. uj
Z
-------
?
V
5
o
8
n
VJ
-J
5
,(
89
Z
J
Q.
h
J
y
D
0
-------
1. Fugitive — In today's typical paint plant there are two types of fugitive emissions. These
are pigment particulate and paint solvents. In a small percentage of the plants an attempt is made
to collect these emissions. The incentive for doing so is based on insurance requirements as well
as occupational, health and safety rather than for air pollution considerations or regulations. The
newly passed Occupational Safety and Health Act (OSHA) will have a dramatic effect on the paint
industry practice and necessitate the collection of fugitive emissions in the future.
Fugitive particulate emissions consist primarily of the various dry material used such as
pigments and extender. Details of these pigments are discussed in Chapter 8 of this report. As a
general rule, the pigments are received and stored in 25 to 50 pound paper sacks or fiber drums.
Modern pigment manufacturers have developed finely sized pigments, 0.05 to 0.25 microns, for
ease of dispersion into the paint vehicle. Loading of these fine pigments into grinding equipment
results in fugitive particulate dust emissions into the surrounding plant areas. This dust is either
collected by a ventilation and exhaust system or allowed to settle and later collected as part of the
general housekeeping requirements. The pigment particles tend to agglomerate during shipment
and storage and the losses during loading are not as significant as might be associated with this
submicron particle size.
A variety of resins are received as granular or flaked solids which are of large size and do
not result in a fugitive dust emission. The manufacturer of these solid resins, however, does
encounter fugitive emission problems in his flaking or grinding operations.
Solvent emissions occur in almost every phase of paint and varnish manufacturing and in
numerous locations throughout individual plants. A listing of emission points is given below:
Location Operation Temperature Pressure, Atm
a.
b.
c.
d.
e.
f.
9-
h.
i.
j.
Resin Plant
Resin Plant
Resin Plant
Paint Plant
Paint Plant
Paint Plant
Paint Plant
Paint Plant
Paint Plant
Paint Plant
Thinning
Filtering
Storage Tanks
Blending Tanks
Grinding
Dispersion
Holding Tank
Filtering
Packaging
Storage Tanks
200 to SOOT
200 to 300° F
100°F
Ambient
Ambient
Ambient
Ambient
Ambient
Ambient
Ambient
1
1
1
1
1
1
1
1
1
1
90
-------
The extent of these emissions vary with the type of operation and the effort extended to control
atmospheric losses. The high temperature thinning and filtering results in the largest emissions,
while packaging in drums and cans contribute the smallest emission. Other operations contribute
intermediate emissions which vary depending on the degree of control exercised and the vapor
pressure of the solvent used. Simple good housekeeping rules, such as keeping loading hatches
closed, will significantly reduce these emissions.
In some cases, efforts are made to collect fugitive emissions by use of local exhaust
systems. More frequently, however, they are exhausted from the building by general building
exhaust fans which ventilate areas having the highest contaminant concentration.
The total quantities of solvent losses as fugitive emissions have not been extensively
measured to date and are not well-known. As a percentage of the total solvent used, these losses
are relatively small. As a percentage of total solvent loss, they are quite significant. Details are
presented in Part D of this Section.
The quantities of fugitive pigment emissions are also not well-known but represent a very
small percentage of the total pigments used in paint manufacturing. The quantity of emissions
cannot be easily measured directly but can be estimated using weight balance calculations on
plants with an efficient particulate collection system. This system will be described in more detail
in Chapter 5.
2. Non-fugitive — The number of regulated emissions emanating from a paint plant will vary
significantly with the type of operation involved. Some trade sales plants that manufacture none of
their own resins and make no effort to confine solvent losses during dispersion, filtering, or storage,
will have no non-fugitive emissions. On the other hand, resin plants or paint plants producing
resins and varnishes are likely to have a number of non-fugitive emissions. These emissions
consist primarily of organic vapors in air or inert gas streams.
3. Chemical and Physical Properties — There are two significant types of organic vapors
generated in paint manufacturing. These are varnish and resin kettle emissions which usually fall
into the non-fugitive category and solvent vapor emissions which are usually fugitive in nature.
The chemical and physical properties of each type will be discussed below.
Considerable effort has been expended to identify the various types of chemical compounds
emitted during a varnish cook. The majority of this work was done in the 1950's and is well
summarized by R. L. Stenburg in the H.E.W. Technical Report A58-4. Copies of his summaries
91
-------
are included here as Tables 24 and 25.
In general, one or more of the following compounds are emitted, depending upon the
ingredients in the cook and the cooking temperature; water vapor, fatty acids, glycerine, acrolein,
phenols, aldehydes, ketones, terpene oils, and terpene. These materials are mainly decomposition
products of the varnish ingredients.
Solvent vapors account for the majority of the gaseous emission from a paint plant. Some
of the relevant properties of the more widely used solvents are given in Table 26. None of these
materials are considered exceptionally hazardous. Several, however, have maximum allowable
concentration limits as set by OSHA. These are given in Table 27.
Odor thresholds for some of the organic vapors encountered in the paint plants as well as
raw materials and trace products which are encountered in resin manufacturing are given in Table
28. They include solvents listed earlier as well as some which might be found in the miscellaneous
solvent category.
The carrier gas for all contaminated streams except the resin reactor exhaust will be
essentially ambient air. This will be true also for the kettles during charging of reactants. During
most of the reaction cycle for the fusion kettle, the carrier gas consists of the sparge gas passed
through the reactor. This is produced in the inert gas generator and is approximately 10 to 12% CC>2
and 85 to 90% fxb, on a dry basis. Small quantities of CO may be present along with other trace
components normally found in air (argon, etc.) and combustion products. Gas leaving the scrubber-
ejectors and the solvent kettle condenser vent will be saturated in water vapor at the operating
temperature of the condenser or scrubber.
As discussed earlier, there are a variety of particulate emissions from paint and varnish
manufacturing. The major source of these emissions are the pigments and extenders. Representative
particle size ranges for the more widely used pigments and extenders are given in Table 29.
Maximum allowable concentration limits for these as set by OSHA are given in Table 27.
It can be noted that the thermal settling velocity of a 50 micron diameter particle of specific
gravity 2.0 is about 18 cm/sec or 0.6 ft/sec. Particles and agglomerates in this size range or larger
should not escape into the surrounding atmosphere during loading. On the other hand, particles
smaller than this size that do escape during loading have the potential to be swept out of the building
by the ventilation system.
B. Sources of Emissions
1. Major — The major sources of emission are listed on page 100.
92
-------
TABLE 24
COMPOSITION OF OIL AND VARNISH EMISSIONS
Bodying Oils
Water vapor
Fatty acids
Glycerine
Acrolein
Aldehydes
Ketones
Carbon dioxide
Manufacturing
Running Natural Oleo-Resinous
Gums Varnish
Water vapor Water vapor
Fatty acids Fatty acids
Terpenes Glycerine
Terpene oils Acrolein
Tar Phenols
Aldehydes
Ketones
Terpene oils
Terpenes
Carbon dioxide
Manufacturing
Alkyd Varnish
Water vapor
Fatty acids
Glycerine
Phthalic anhydride
Carbon dioxide
93
-------
TABLE 25
ODOR AND COMPOSITION (BY FUNCTIONAL GROUPS)
OF OIL AND VARNISH EMISSIONS
Process and Temperature Kettle Ingredients
Heat Polymerization of
Linseed Oil
575°F
3 Dark Linseed Oil
580° F
0-Pale Linseed Oil
580°F
1-Pale Linseed Oil
00-Pale Linseed Oil
580° F
Linseed Oil
Admoline (a catalyzed
Linseed Oil)
Raw Linseed Oil
Alkali Refined
Linseed Oil
Alkali Refined
Linseed Oil
Alkali Refined
Linseed Oil
Compounds Identified
Saturated and/or unsatur-
ated:
Aliphatic Fatty acids
Aldehydes
Aliphatic Esters
Paraffins
Olefins
Saturated and/or unsatur-
ated:
Carboxylic Acids
Aliphatic Esters
Alkyl Aldehydes
Paraffins
Olefins
Unsatu rated:
Acids
Esters
Aldehydes
Saturated and/or unsatur-
ated:
Aliphatic Fatty acids
Olefins
Saturated and/or unsatur-
ated:
Fatty Acids
Saturated Aldehydes
Olefins
Saturated and/or unsatur-
ated:
Aliphatic carboxylic
acids
Unsaturated Alkane Esters
Saturated Alkyl Aldehydes
Olefins
Odor
Harsh pungent odor.
Characteristic of short chain
aldehydes. (Cause eye
watering to varying degrees.)
Tall Oil Alkyd
475°F
Tall Oil
Glycerine
Fumaric Acid
Saturated and/or unsatur-
ated:
Carboxylic acids
Esters
Aldehydes
Alcohols
Paraffins
Olefins
Fumaric Acid
Hydrogen Sulfide
(Rotten egg odor).
94
-------
TABLE 25 (continued)
Process and Temperature Kettle Ingredients Compounds Identified
Odor
Tall Oil Alkyd
560°F
Soybean Oil Alkyd
475°F
Tall Oil
Glycerine
Pentaerythritol
Soybean Oil
Glycerine
Fumaric Acid
Rosin
Linseed Oil Alkyd
475°F
Alkali Refined
linseed oil
Glycerine
Phthalic Anhydride
Litharge
lemon-like odor.
Very Offensive
-Hydrogen Sulfide,
n-Butyl Mercaptan
(Skunk Odor).
Saturated and/or unsatur-
ated:
Aliphatic carboxylic Mild soapy and slightly
acids
Aliphatic Alcohols
Alkyl Aldehydes
Alkane Esters
Paraffins
Olefins
Aromatic Acids
Aromatic Esters
Aromatic Aldehydes
Aliphatic Carboxylic
acids
Aliphatic Aldehydes
Aliphatic Alcohols
Typical linseed oil
cooking odors
(harsh and stinging).
95
-------
TABLE 26
SOLVENT CHARACTERISTICS
Solvent
Mineral spirits
Xylene
Toluene
Methanol
Ethanol
n-Butanol
Acetone
MEK
Ethyl acetate
Molecular
weight
160(est.)
106
92
32
42
74
58
72
88
Vapor pressure
20°C
2 (est.)
7.1
22
92
43
4.3
186
80
56
, mm Hg
100°C
30 (est.)
250
580
-
-
400
-
-
—
Boiling poin
°C
—
-
-
64.7
78.4
-
56.5
79.6
77.1
*For solvents that boil below 100°C
96
-------
TABLE 27
MAXIMUM OSHA ALLOWABLE CONCENTRATION LIMITS
Maximum allowable exposure
Material (8 hour weighted average)
Ethyl Acrylate 25 ppm
n-Butanol 100 ppm
Phthalic Anhydride 12mg/m3
Toluene 200 ppm*
Xylene 100 ppm
Carbon black 3.5 mg/m3
Talc (non-asbestos form) 20 particles/cm3
Talc (asbestos form —
Tremolyte) 5 fibers/cm3**
Inert dust
Respirable 5 mg/m3
Total 15 mg/m3
'Will change to 100 ppm in near future
"Will change to 2 fibers/cm3 in 1976
Source: Fed. Reg., Vol. 37, No. 202, 1972. Additional information available from this source.
97
-------
TABLE 28*
ODOR THRESHOLDS OF SOME ORGANIC VAPORS
Chemical
Acetaldehyde
Acetone
Acrolein
Benzene
Ethanol
Ethyl acrylate
Formaldehyde
MEK
Methanol
Methyl methacrylate
Methylene chloride
Phenol
p-Xylene
Styrene (inhibited)
Styrene (uninhibited)
Toluene
Odor threshold, ppm
0.21
100.0
0.21
4.68
10.0
0.00047
1.0
10.0
100.0
0.21
214.0
0.0470
0.47
0.10
0.047
2.14
*Air Pollution Control Assoc. Journal, Volume 19, Number 2, Feb. 1969, pages 91 to 95
98
-------
TABLE 29
PARTICLE SIZE RANGE OF VARIOUS PIGMENTS AND EXTENDERS2,4
Pigment Size range, microns
TiO2 0.1 to 1
Extenders
Silica 0.1 to 20
CaCO3 0.03 to 8
Talc 0.2 to 10
Clay 0.5 to 10
Iron oxides 0.2 to 15
Carbon blacks 0.01 to 0.3
99
-------
1. Varnish cooking
2. Resin cooking
3. Thinning
They are all non-fugitive and consist primarily of organic vapors including phthalic anhydride, which
may or may not be in the vapor state. They are also normally the major source of potential odor
nuisance problems.
2. Minor— Minor sources of emissions included are tabulated below:
Location Fugitive Non-Fugitive
1. Handling & Storage Hydrocarbons
2. Milling Operation Hydrocarbons, Pigments Hydrocarbons, Pigments
3. Blending & Finishing Hydrocarbons
4. Filling Hydrocarbons
On an individual basis, each of these sources can be considered minor. If they were to be collected
and vented at a common point, they would constitute a major source. This is demonstrated in more
detail in the following section.
C. Quantities of Emission from Uncontrolled Plants
As indicated in Section I, the manufacturing of paint is a very non-standardized industry.
The type and quantity of emission will vary significantly from plant to plant. To provide a better
representation of the average situation, the model plant presented earlier will be used as a basis
for this discussion where applicable. Varnish and resin manufacturing are not well covered in the
model plant and the quantity of emissions from these sources will be discussed separately.
1. Model Plant — The model plant has been designed to reflect modern practices and technology
as generally applied, rather than the frontiers of technology as practiced by perhaps a few plants.
Likewise, the emissions calculations will assume commonly applied practice and observations.
It will further assume that equipment items, such as condensers, have been properly sized and are
properly maintained to insure good operation.
Common practice in this context is determined from questionnaire data, literature references,
information from equipment manufacturers, and personal experience. Wherever possible, data for
plants and equipment similar in size and function to that for the model plant will be used in preference
to overall industry averages. Where a system is sufficiently well defined, a theoretical calculation
is used to estimate emissions.
100
-------
A schematic diagram showing the emission points for this model plant are given in
Figures 15 and 16. Some tanks are tied in to common vent points as shown.
a. Solvent Emissions From Tanks — This group comprises emission points 1 through 7. It will
be assumed that the outside storage tanks (5, 6, 7) are equipped with conservation vents. Since
the inside storage tanks (1 to 4) are not subjected to significant daily temperature cycles, these
vents are left uncontrolled. Emissions, then, are confined to losses during filling operations. It
will further be assumed that xylene is the solvent used for industrial alkyds. Finally an average
ambient temperature of 20°C both inside and outside will be specified.
Table 30 summarizes the operating data for the storage tank emission sources. Table 31
lists the calculated emissions for these sources. For sources 1, 2 and 3 it is assumed that the
vapor pressure of the solvent over a resin solution is half that of the pure solvent. Similarly, the
waste solvent is assumed to be 50% xylene and the rest primarily mineral spirits. The vapor pressure
over this is taken as 50% that of pure xylene with the contribution of the mineral spirits neglected.
The emissions were calculated by converting the turnover in gallons to cubic feet and determining
the amount of solvent present in that number of cubic feet of gas saturated with solvent vapor at
the appropriate vapor pressure. Table 31 also gives the instantaneous emission rate assuming liquid
is pumped into the tank at 100 gallons per minute. Total emissions from these sources amount
to 184.8 pounds per year. Emission rates run as high as 2.06 pounds per hour at 100 gallons
per minute input filling rate.
b. Manufacturing Area — The manufacturing area emissions consist of sources 8 and 9. Upper
floor manufacturing area includes the mills, mixers, finishing tanks, etc. The lower floor includes
the filling area as well as the ambient air around the resin plant and indoor storage tanks. It does
not include the thin tank and resin reactor vents themselves as these will be considered separately.
The emissions from the manufacturing areas exit as part of the exhaust from the ventilation system.
The ventilation system for the model plant was designed to produce six air changes per
hour. This converts to an exhaust rate of 48,400 cfm for each of points 8 and 9. The Canadian
Paint Societies has discussed a model plant in which ventilation requirements were set at 2 cfm
per ft2 of floor area.10 The present situation compares well at 2.2 cfm per ft2. In order to determine
the emission rate for these sources, it is necessary to estimate the concentration of pollutants
in the exhaust gas stream.
An examination of the questionnaires reveals two plants which present data that is usable
101
-------
i
111
i
0
8
0
*
h
y
H) rn Z
U ^WJ 0 X
f 8^z r o P
8<< Otf)
P 0 hi . ,
2^ fij*
^ S<
:20 wff
^h< j 0
< (/)r> -;!UJp
^^n ^Qw
_ n y 0
-------
y
LL v/ i.i
H z 8
Z j< A
a h 8
z z Q)-^
8
-
I
j
u
0
EJECTOR -
SCRUBBER
SYSTEM
^-r
/K
a
y -*-
EJECTOR -
SCRUBBER
SYSTEM
-
j
^
0 tf\h_^"-
' v_5y^^^^^^^
\ x^
m
tr
* § |2
HS&
iD
Z
o:
D
U
^
z
1 ij /-N
< ^
0
y
Q.
Q.
D
0 zw
?8
D
U
^
Z
2 y
IT
f
„;
y
^
Q
J
-------
TABLE 30
OPERATING PARAMETERS FOR MODEL PLANT
STORAGE TANK EMISSION SOURCES
Emission point Tankage Material stored Turnover (gal/yr)
1 4,000 Industrial alkyd (50% NVM) 53,000
2 17,500 Industrial alkyd (50% NVM) 232,000
3 15,000 Trade alkyd (50% NVM) 257,000
4 32,500 Oil, glycerine 188,200
5 25,000 Mineral spirits 216,000
6 25,000 Xylene 234,000
7 2,000 Waste solvent (50% xylene) 30,900
104
-------
,_
CO
UJ
1
m
<
H
CO
^
2
<
P™~
LJJ
o
oc
o
CO
1-
2
£f
>
_J
Q-
_l
UJ
Q
O
5
tr
O
LL
0
2
_1
_l
LL
0
2
cc
D
Q
CO
O
CO
CO
s
LLJ
_^
c
03
4-*
O
•M
+->
a
c
c
[E
a
8
*
@
k_
JZ
^
CO
c
o
'co
CO
'E
UJ
5
•^
LL
O
<
(D
•*-<
CO
4-*
(A
-j
CO
^z
X
UJ
—
£
Q.
_Q.
C
O
'+J
CO
4-1
c
0)
o
o
o
U.
^^
^
'c
3
O
E
<
c
o
'ui
t/t
E
cu
CO
Q.
'o
c
i
•4-"
c
'o
Q.
C
o
'(/)
v»
'e
UJ
co co ^ r-- co co
p O •<* ; _• 00 CD O
»-' r-' CD C O CN r-'
^t <^ ^- ^ ^" ^- ^J-
CO CO CO CO CO CO CO
o o o o o o
o o co — : o LO o
p*- P^ co 'p co co r*«
^t" ^-" <— " en" ^-"
«-_ oq r^ LO ^- co oq
cri en 06 = «-' ca in ^-'
CO «— C CO CO 00
(/) tn
'Z c 'C
Q. V Q-
U) • ai a . >. .E - .= -^ >
X X S O S X X
i-CMCO^-LOCQP- ™
o
105
-------
for estimating exhaust gas contaminant levels. These companies reported the solvent vapor
concentrations in the air at various locations in their plants. The first plant (Plant A) produces almost
one million gallons per year of which 80% is solvent based.
The second plant (Plant B) is a 2.5 million gallons a year facility with 55% solvent based.
The results from this questionnaire are also summarized in Table 32.
Based on this data, an emission level of 15 ppm in the exhaust from the upper floor
production area (Source 8) should not be an unreasonable estimate. This allows for averaging and
dilution from parts of the area which show low solvent vapor levels. A level of 5 ppm will be used
to estimate the levels from the lower floor production area (Source 9). The calculated emissions
for those sources are given in Table 33. The total for these two sources is 63,500 pounds per year.
Only one plant reported sufficient data in the questionnaires to provide a basis by which
the model plant emissions can be compared to existing plant emissions. This plant produces 2.7
million gallons per year of coatings (all solvent based). The production areas corresponding to
points 8 and 9 of the model plant have ventilation rates which together total 35,800 cfm (versus
96,800 cfm for the model plant). This plant reported emissions from these areas of 62,660 pounds
per year. The emissions for the model plant, per million gallons of paint produced, in comparison
are higher. At least some of this difference can be attributed to the considerably higher ventilation
rate in the model plant.
No estimate has been made for particulate emissions from these sources. The absence of
ducts and hoods specifically for the purpose of dust collection means that a large part of the
pigment dust which escapes into the air will tend to settle out inside the plant. The ventilation
system will not provide air velocities sufficiently high to capture any but a portion of the particulate.
It would be very difficult to estimate the pigment emission from the plant, as presently designed,
by comparing data reported for hooded and locally exhausted systems.
c. Resin Production — Emissions from the 500 gallon fusion reactor are of two types: (1) Those
that occur during sparging; and (2) those that occur during loading when the hatch is opened.
The ejector-scrubber is in operation at all times during the cook. The emission levels that will
be estimated for this reactor are those at point 10, the reactor outlet vent and before the scrubber.
At the present time, insufficient information is available to estimate emissions downstream from
the scrubber.
It is expected that the small reactor will be used primarily to process small batches of
106
-------
CO
q
CD
CM
CO
LU
to
H
Q.
K
z
^
Q.
— 1
LU
Q
O
^
LL
O
0)
cc
Q-
(/)
O
oc
z
tyj
LU
LU
E
Q.
Q.
g
I
c
8
o
o
^ O CN O CD
!£ rf- CO CN" «
2 CN CN t- LT)
01
c q ^- •* q ro
•^ CN 1 o W • cxi I r^
> r- r- 10 UJ
CD
C
c
x
E
^
CD
c
-------
TABLE 33
EMISSIONS FROM MODEL PLANT PRODUCTION AREA EXHAUST
Source 8 Source 9
Exhaust rate, cfm 48,400 48,400
Principal contaminant Xylene Xylene
Concentration, ppm 15 5
Emissions, Ib/hr 11.9 4.0
Emissions, Ib/yr 47,500 16,000
108
-------
specialty industrial alkyds. For illustrative purposes, a "typical" batch will be assumed to consist
of a short oil formulation charge to the reactor at 2,500 pounds of raw materials. A total run time
of 8 hours is considered representative with 3 hours allowed for alcoholysis and 5 hours for
esterification.
A "typical" log for this process follows:
Time, hr Temperature, °F
0 Ambient Charge reactor with 1,000 pounds of oil, 600 pounds of glycerine.
Blanket with inert gas. Agitator and heat on.
1.5 450 Shut off agitator. Add Catalyst.
1.55 430 Agitator on.
1.75 450 Holdat450°F.
3.00 450 Alcoholysis complete. Agitator off. Cool slightly.
3.10 400 Add 900 pounds of phthalic anhydride. Agitator on. Sparge at 10 cfm.
3.30 450 Hold at 450°F for esterification.
5.10 450 Sparge at 5 cfm.
6.10 450 Sparge at 2.5 cfm.
7.50 450 Esterification complete. Cool.
8.00 400 Agitator, sparge off. Drop to thin tank.
This log is based on similar information contained in technical literature published by the
Brighton Corporation11 and on information contained in Martens' book.4 It is hypothetical but
reflects the general features of resin plant operation.
Figure 17 presents emission levels, exhaust rates for noncondensibles (air and/or inerts),
and water evolved as a function of time. Again, these are hypothetical but are intended to represent
general trends and orders of magnitude.
The emission curve is adapted from that presented later in Figure 23 of this report. This
represents hydrocarbon ppm (as methane) at the reactor vent for a fusion cook. Concentration is
reported on a dry basis. Generally speaking, the analytical method used to obtain this curve does
not measure phthalic anhydride, though small amounts may have entered the instrument and been
recorded. It represents that portion of the emissions which is virtually unaffected by the scrubber
and so provides an estimate of emissions from the scrubber exhaust. It was estimated by the
engineer who supervised the source test which produced Figure 23 that a typical scrubber will
remove, at most, 10% of these contaminants.
109
-------
IOO,
120
IOO
80
60
-4O
20
(fl
Q
Z
0
1
.
Q
Id
>
J
0
LJ
0
OJ
I
FIGURE: IT
EMISSION CHARACTERISTICS — SHORT OIL FUSION COOK
110
-------
Exhaust rates were based on 2 cfm inert gas blanket during alcoholysis, 0.04 cfm per gallon
of charge during the first two hours of esterification, 0.02 during the third hour, and 0.01 for the
remainder of the cook. A peak of 100 cfm during charging was assumed based on questionnaire
data. The water evolution curve is based on the Brighton11 information and Martens.4 It was con-
structed to reflect the characteristics of a short oil alkyd cook.
The 1,500 gallon reactor is set up for either fusion or solvent cooking. It is expected,
however, that solvent cooking will represent the normal mode of operation. A "typical" batch for
this reactor will be assumed to be a long oil alkyd formulation charged at 7,500 pounds of raw
materials. A total run time of 12 hours, 5 hours for alcoholysis and 7 for esterification will be assumed.
A "typical" log for this process is given on the following page.
Figure 18 presents contaminant levels, exhaust rates, and water evolution as a function
of time. The contaminant levels and exhaust rates are presented for point 12, the condenser
vent. The exhaust rates for noncondensibles is based on questionnaire data. Only a few plants
to date have reported exhaust rates from the condenser vent. Values of 2.5 cfm to 7 cfm have
been reported for the noncondensibles from kettles of 300 to 3,000 gallons capacity. A value of
0.25 cfm has been chosen to represent the inert gas blanket flow rate for this reactor.*
Maximum emission levels were calculated by assuming that the exhaust from the condenser
vent is saturated in xylene vapor at the operating temperature of the condenser. It was assumed
that the condenser vent temperature is at 100°F during the peak reaction period and drops slightly
towards the end of the cook. At 100°F the vapor pressure of xylene is taken as 18 mm Hg. At 1 atm
total pressure, this is equivalent to 24,000 ppm xylene (192,000 ppm as Ci).
The water evolution curve is representative of the shape often observed for long oil alkyds.
It reflects the observation that the reaction rate for long oil alkyds is very high initially and then
drops off for the remainder of the cook. This can be compared with the curve for the short oil alkyd
which shows a more moderate reaction rate which extends for a longer portion of the reaction time.
*For the two solvent cook source tests shown in Appendix D, noncondensible flow rates average
about 0.5 SCFM for one cook and about 0.25 SCFM for the other. These flows were observed
after the solvent was added and the condenser was turned on. It does not cover the loading and
heat-up phase during which the ejector was on.
111
-------
TYPICAL LOG OF 1,500 GALLON SOLVENT PROCESS REACTOR
Long Oil Alkyd — 60% Oil
Time, hr Temperature, °F
0 Ambient Charge reactor with 4,500 pounds oil. Blanket with inert gas. Agitator
and heat on.
2.0 500 Agitator off. Add 1,000 pounds glycerine. Add catalyst.
2.10 450 Agitator on. Inert gas blanket.
2.60 500 Hold at 500°F.
4.50 500 Alcoholysis complete. Cool to 400°F.
5.00 400 Start condenser. Add 350 pounds xylene, 2,000 pounds PA. Agitator
on. Inert gas blanket.
5.75 450 Hold at 450°F for esterification.
11.50 450 Esterification complete. Cool.
12.00 400 Agitator off. Drop to thin tank. Condenser off.
As before, this is an adaption of information published by Brighton Corporation and Martens.
It reflects the general features for this type of cook.
The reverse peaks in the curves of Figure 18 occur during loading. When the hatch is
opened, the scrubber-ejector is turned on. This draws all the exhaust to the scrubber vent and
prevents gas from escaping through the condenser. The flow rate to the scrubbers (Point 13)
will be expected to be in the 100 to 150 cfm range during loading. It is not possible, with available
information, to estimate the emission rates during this time. The rate will probably be rather high
for a short period of time and consist primarily of solvent vapors and phthalic anhydride.
The emissions, by weight, for each of the reactors can be determined from the information
given in Figures 17 and 18. For a given time period, the exhaust rate and the average hydrocarbon
concentration for that time period provides the information necessary to calculate the emissions.
This has been done for each reactor and the results summarized in Table 34. It should be noted
that the emissions during loading operations are not included since insufficient information is
available. It can be expected that the emission rate will be quite high for short periods of time
during charging.
Several comments can be made at this time. The emissions from the solvent reactor are
proportional to the assumed inert gas flow and strongly dependent on the assumed condenser
vent temperature. Source test personnel from air pollution equipment manufacturers report that
it is not uncommon for condensers to be in very poor operating condition due to fouling or other
112
-------
0
(I)
<
£
a.
Q.
200
0
TIME , hr.
10
TIME, hr.
0)
20O Q
Z
I
ISO
Q
y
ICO
Id
0
-------
TABLE 34
SUMMARY OF EMISSIONS FROM MODEL PLANT
500 GALLON FUSION REACTOR AND 1,500 GALLON SOLVENT REACTOR*
Fusion cook Solvent cook**
Time of cook, hr 8 12
Total emissions as methane,
pounds 1.15 0.7
Max. emission rate, Ib/hr 0.41 0.12
Emission per 1,000 pounds
charged, pounds 0.46 0.094
Max. emission rate per 1,000
pounds charged, Ib/hr 0.16 0.016
Emission per year based on
250 batches/year, pounds 288 175
*The emissions listed here for both fusion and solvent processes are for the "noncondensible"
organics. It does not include phthalic anhydride, very heavy organics, etc. that are emitted during
the baking phases of operation. No reliable data has been located for estimating these latter quantities.
**An average inert gas flow of 0.25 SCFM was used in estimating the solvent process emissions.
Many plants operate with essentially no inert gas flow while others may maintain several SCFM.
In the former case, emissions will be limited largely to a "breathing flow" situation and will be very
small. For the latter case emissions will be much higher than suggested in this table since emissions
are roughly proportional to inert gas flow rate.
114
-------
causes. If the condenser vent in the present instance were operating at 120°F instead of 100°F,
the emissions from this source will double.
The inert gas flow through solvent process reactors is also the subject of considerable
variation. Among other things, it varies as a function of reactor size and formulation. In some cases,
no inert gas flow is used while in others flows as high as 7 cfm have been reported in the questionnaires
and 16 cfm in data supplied by one company. Emissions, then, can range from essentially none
to several times that calculated for this case.
Finally, the emissions from these kettles were determined for specific cooking formulas
using specific process cycles and parameters. Consequently, one should not attempt to draw
generalized conclusions concerning the relative merits of solvent versus fusion cooking from the
results presented here. They are intended to be specific to the model plant under consideration
and are representative of the industry in general only to the extent that the model plant is repre-
sentative. It may be noted that the results fall within the range reported by one manufacturer as
listed in Table 3512.
d. Thin Tanks and Filter Presses — Emission source points 11, 14 and 15 (Figure 16) will be
covered in this category. It will be assumed that each resin is to be thinned to 50% NVM. The thin
tanks are equipped with condensers. It will be assumed that the condenser outlet vent operates at
100°F and that the gas is saturated in xylene at that temperature. The gas vented from the thin tank
consists of a noncondensible fraction which results from that displaced when the tank is filled.
The fusion reactor batch is about 250 gallons while that for the solvent reactor is about 750 gallons.
The vapor pressure of xylene at 100°F is about 18 mm Hg. A fairly straightforward calculation
gives emissions as xylene of 0.22 pounds per batch for point 11 and 0.65 pounds per batch for
point 14. Based on 250 batches per year, yearly emissions will be 55 pounds/year for point 11
and 163 pounds/year for point 14.
The filter presses, taken collectively as source 15, represent a significant source of vapor
emissions. It is not possible with presently available information to estimate the quantities involved,
however. Vaporized solvent escapes directly into the room air and leaves as part of the room
ventilation. Its contribution to emissions is included in the emissions estimated for the lower floor
manufacturing area since, as the model plant is conceived, the 48,400 cfm exhaust system for the
lower floor area includes the room in which the filter presses and thin tanks are located.
115
-------
0)
D>
oc
in
co
in
_i
00
2
&! .
£ S a
Il|o-
in
in
UJ
o
o
Cook, based upon 4
Sol
o
CO
O
4-1
O
CD
C
_Q)
X
on cook, based upon
Solven
o
o
CD
CO
O)
pounds/hou
issions exceeded
arging.
E
c
C
o
Q.
13
T3
S
o
o
o
o
'
€ £
CD
C
°co
v_
ra
jC
o
D)
C
u
TJ
3
O
.C
lo
•D
£Z
D
O
Q.
CO
73
CD
•D
CD
0)
O
X
CD
ssions
E
UJ
en
'5i
CO
Q.
CO
•a
c
CO
E
CD
.C
0
x
0)
t/)
^c
't/>
CD
CO
o
a
D
•a
CD
ro
_n
o
8
c
o
3
u.
CD «
»- at
CD
C
_CD
X
CO
c
O
o
CD
C
CD
O
L_
Q.
o
o
CD
C
_Q)
> O
Q. O
O >
rt 5
c
o
Q.
D
T3
CD
-^
O
O
U
4-t
C
CD
_>
O
in
o
in
CM
O
CM
CM
o
c
ro
4-f
1
CD
O
a
D
T3
Si
ro
JD
^"
O
O
CJ
•4-1
C
0)
_>
o
w
o
ro
+j
CO
I s_
|S£
E > J2
UJ < =.
CO O
d •-'
O5
CM
CO
O
>
E ro _
UJ DC Ci
in LO ^j-
"*. ^ "^
I-' O r-'
o o o
*J 4-> *J
r^ p~- •*
O i- ^t-
co
r-'
CM
CM
CO
co
CN
CM
O
4-t
ID
CO
CM
O
O
4-t
CD
O
O
O
ro o -
(U N CO
CC '(/:, &
§O O
O O
P. P. o
O
o
8
o o
8 8
c a)
•« a
CD >
DC H
8
I
"o
a.
„- 2
•^ E
>- ro
O CD
116
-------
EMISSION SUMMARY FOR
Emission point*
TABLE 36
GASEOUS CONTAMINANTS — MODEL PLANT
Number Description
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Indus, alkyd storage
Indus, alkyd storage
Trade alkyd storage
Oil, glycerine storage
Min. spirits storage
Xylene storage
Waste solvent storage
Upper floor vent
Lower floor vent
Fusion reactor
Thin tank
Solvent reactor
Solvent reactor
Thin tank
Principle emission
Xylene
Xylene
Mineral spirits
Oil, glycerine
Mineral spirits
Xylene
Xylene
Xylene
Xylene
Mixed organics
Xylene
Xylene
Xylene, PA
Xylene
Annual emissions
(Ib/yr)
9.1
39.8
18.7
nil
31.5
80.4
5.3
47,500
16,000
288
55
175
undetermined
163
*Refer to Figures 15 and 16 for locations of these points in model plant.
117
-------
e. Summary — Table 36 presents yearly emissions by point sources as defined in Figures 15
and 16. The total for these sources is 65,590 pounds per year.
' It should be realized, of course, that certain assumptions were made in calculating each of
the emissions. In order to understand the numbers, one must have a thorough understanding of
the nature of these assumptions and the methods of calculation. The effects of small changes
can sometimes be rather dramatic. For instance, a few ppm difference in the contaminant level
assumed for the manufacturing area exhaust systems (Point 8 and 9) can make a considerable
difference in the emissions from these sources. Likewise, as discussed earlier, emissions from the
resin plant are sensitive to the exact operating conditions.
The total gaseous emissions consist almost solely of solvents. They represent 1.7% of
total solvents consumed.
2. Varnish and Resins Production — No varnishes and a relatively small portion of resins
used in the model plant are produced in the model. Production of these materials is discussed in
more detail below.
Varnishes and oils are cooked or bodied at temperatures from 200 to 650°F. At about 350°F
decomposition begins and continues throughout the cooking cycle which normally runs between
8 and 12 hours. The quantity, composition and rate of emissions depend upon the ingredients
in the cook as well as the maximum temperature, the length, the method of introducing additive,
the degree of stirring and the use of inert gas blowing. In general the emissions will average between
one to three percent of the charge in oil bodying and three to six percent in varnish cooking17.
The exact amount of non-fugitive emissions for open kettle varnish cooking is not of great
significance for two reasons. First, the amount of varnish cooked in this fashion is quite small
and is declining. Secondly, modern varnish reactors are equipped with reflux and water cooled
condensers which provide better control of the extent of emission. Of more importance are the
characteristics of the emissions related to ease of removal by the applicable pollution control
devices.
Modern resin reactors and varnish cookers account for the majority of paint vehicle production
in the paint and varnish industry. As described earlier, these products are cooked in larger more
carefully controlled reactors equipped with product recovery devices which also help reduce
atmospheric emission. As with the old varnish kettles, the amount of emissions vary with the type
of cook, the cooking time, the maximum temperature, the initial ingredients as well as the type
and method of introducing additives.
118
-------
The basic methods of cooking used are solvent cooking and fusion cooking. The original
and still widely used method is fusion cooking.2 In this method the ingredients are heated
together without solvents at temperatures of 435 to 485°F. This type of process has the maximum
emission level caused primarily by the blowing of inert gas through the reactor to remove the water
of reaction. Refluxing is required when volatile monomers such as styrene are employed. Fusion
cooking is the fastest method for the production of polyester resins other than the alkyd resins.
Solvent cooking is the more modern and now popular cooking method. In this process a
small portion of aromatic solvent, usually 4 to 10% is added with the charge or after alcoholysis
in the two-stage procedure. The solvent is condensed and refluxed to a decanter for water separation
and then returned to the reactor. The main advantages of solvent cooking are faster removal of
reaction water, shorter cooking time for alkyd resins, better control of temperature, as well as
reduction of hydrocarbon emissions and phthalic sublimation. The condensers must be run warm
enough to prevent fouling by condensed phthalic anhydride or phthalic acid so some hydrocarbon
emissions are unavoidable.
A considerable amount of emission source testing has recently been completed by suppliers
of air pollution control devices on closed kettles or reactors. These efforts have been directed
primarily towards quantifying the emissions rather than qualifying them. Recent nationwide hydrocarbon
emission surveys of a large resin manufacturer indicated that better than 85% of the reactor
emissions were materials originally charged to the reactor.18 A similar conclusion without exact
percentages was indicated by W. G. Skelly in his report covering the analysis of potential sources
of odor and pollutants from the Bensenville, Illinois plant of Stresen-Reuter International.19
From an air pollution control engineering design standpoint, the conclusions above are the
major significant result of the above tests. The rest is mainly of academic interest. Assuming thermal
or catalytic incineration as the best control device, the type of information required is exhaust
rate, temperature, maximum hydrocarbon concentration or Btu loading, and particulate aerosol or
condensible hydrocarbon concentration. The hydrocarbon concentration is required to calculate the
heat released or system temperature rise from incineration of the fumes. This is required to
assure proper sizing of the system, burner, heat exchanger, and residence chamber. Particularly,
aerosol and condensible hydrocarbon concentrations are required to prevent condensation in the
connecting ductwork and heat exchanger or unit inlet, which will eventually lead to plugging, loss of
adequate ventilation and/or system fires or explosions.
For solvent cooking the quantity of emission does not vary significantly with the size of
119
-------
the reactor but is rather more a function of the volatility of the solvent being used and the size
and/or efficiency of the condenser. Since there is less sparge gas used in solvent cooking, exhaust
volumes are small (less than 1 SCFM) and consist primarily of noncondensibles. Emissions except
when charging will run from 0.1 to 0.5 pounds per hour and will be less cyclic in nature than for
fusion cooks.
Emissions during fusion cooking run much higher and vary with size of the reactor. The
total exhaust volume is dependent primarily on the sparge rate of inert gas. Dean H. Parker5
indicated typical sparge rates of 0.04 cfm/gal of charge during the first hour, 0.02 cfm/gal during
the second, and 0.01 cfm/gal during the remainder of the cook. The exhaust rate will average
from 2 cfm/100 gallons of capacity on small reactors to 1 cfm/100 gallons of capacity on large
reactors. A summary of source test results from a variety of resin reactors is presented in Table37.20
Since fusion cooking is a cyclic batch process, the concentration of emission will vary from
the start to finish of the cook. Hydrocarbon concentration will vary from 15,000 to 80,000 ppm as
methane equivalent, depending on the time of the cycle and the type of cook. There are at least
100 different emission curves that could be encountered if one tried to cover all of the different
cooking formulas. Particulate phthalic anhydride (PA) is also emitted from the kettle and concentration
levels vary depending on cycle time, type of cook, method of charging and type of PA used.
Charging of liquid PA rather than dry solid PA significantly reduces the emission rate. Maintaining
the linear velocity of the sparge gas below 150 ft/min will also reduce the carryover of PA. Entrained
and sublimed PA will run between 1 to 3 pounds per hour over a period of 50 to 70 minutes during
and following the charging period. When isophthalic acid is used these types of emissions are
reduced.
Details of emission levels from varnish and resin kettles are presented in Table 37 and
Figures 19 thru 24 and discussed below.
The 500 gallon reactor shown in Figures 19 and 20 had no reflux condenser and no
separate thinning tank. Thinning was done in the reactor. The combination of these two deviations
from normal practice accounted for the high emission rates. Corrective action has been taken.
The 1,000 gallon reactor shown in Figures 21 and 22 is equipped with a reflux condenser
and separate thinning tank. This kettle is used for both solvent and fusion cooks. As indicated
on Figure 22 the kettle is run blocked-in during most of the solvent cook, and there are no significant
emissions until the vent is opened after the batch has been dropped to the thin tank.
The 1,500 gallon kettle shown in Figure 23 has no reflux condenser. It is used for fusion
120
-------
c
cu
TJ
o
o
en
c
c
CO
1
k-
o
4-
•a
8!
-D
CO
O)
c
'a
Q.
o
en
c
0)
>
_cu
c
o
E
LU
4—
O
C
'c
en
CO
_Q
-t-J
CO
"E
o
CO
•M
c
CO
>
c
CO
•I-'
c
o
-t-J
en
c
'o
a
0
•a
en
c
°i_
D
_co
to
c
T3
CD
C
C
£
v>
O
O
O
condense:
2
0)
0
c/3
to
.c
cu
l".
cu
y
0
.c
cu
c
o
cu
E
'*""'
o
o
o
0
cu
0
cu
E
't"J
o
o
o
0
CO
c
0
CO
E
4-1
o
o
CJ
<
a.
d) «
O
in
O)
O)
CM
a
CD
CD
O3 CD
CM •
.t^
u
CO
a
CO
°
1
CO
Q.
1-
O ^-^
O CO
LT> CM
T-" 03
iz
g>
CO 03
O 03
o 52.
JO
CO
O3
O
0
LO
CM
T3
CO
O
O
f ^
T^-
CM
03
^
0)
LL
03
03
CO
CO
CO
O3
O
O
o
03
CO
O
0
CO
CO
„ ^^ C33 ^~. ^~.
?£
03
o
Q.
O
•j—
CM
03
^
O)
LL
03
03
CO
CM c
CM £
03 U
^
O3
-, O) O
3 T- CM
•3 03 03
^ ^
D3 O)
U- TJ LL LL
03 CO 03 03
03 O 03 03
a>. o 52. 52-
CO
CO
D3
O
(D
03
Q.
O
CO
CO
D3
O
CD
1
1
C
03
a.
O
u
Cl
c
c
C£
C
c
c
g
CO
CO
E
03
E
^
E
s
E
CD
CD
CU
CO
0)
121
-------
00
go
II
00
cc
LL
L
0
8
Q
0
t
8
8
0
8
8
(0
0
y'
tr
0
C_LN3-lVAinO3 H0>
Ol d d —
Nl
122
-------
8
o
o"
rt
o
o
o
o
(0
0
I
Id
F
U
h
0
UJ
3
uidd_aocdVA NI
123
-------
8
0
0'
0
8
0"
10
0
0
0
0
OJ
<0
•0
0)
<5
0
Id*
|
0 h
(VI
tvi
0
0
1
U
*
Q
HI
Q
0
0
(T
0
Q
1
UJ
-
a
5
0
I
UUd d - fcJOedVA Nl
124
-------
URS
u
— (\l IL
0
DC
5
Q
C-LN3nVAinO3
JdO
-------
IL
•
0
(*)
-------
—h i= 01
UJ
0
tr
<
o
0
(T
Q
HO)
UUdd->dOcdVA Nl
127
-------
cooks only. Thinning is done in a separate tank. Hydrocarbon emissions from the fusion cook
shown on Figure 23 are lower than previously plotted fusion cooks and more in line with what
might be considered normal emissions. It is felt that this kettle runs with lower emissions since
solvent is never added to the kettle or conversely that the other fusion cooks have higher than
normal emissions since solvent cooks are interdispersed with the fusion cooks.
The 2,500 gallon kettle shown in Figure 24 has a reflux condenser and thin down tank.
It is used for both solvent and fusion cooking and as indicated above has higher emission than
those shown on the kettle used for fusion cooking only.
Measurements of emission from three 60 gallon open varnish kettles were also completed.
The data is reported at the bottom of Table 37. In general, concentrations were below 220 ppm
except during the addition of lime to the 60 gallon lime cook. At this time there was a peak emission
of 1,800 ppm for about a one minute period.
It is also of interest that no increase in emission was measured during cooking of the wood
oil cook. Cooling was accomplished by the addition of one gpm of water for a three minute period.
The hydrocarbon concentration reported was calculated on a dry basis. A modified M—S—A
Total Combustibles Analyzer was used in measuring the hydrocarbon concentration. The instrument
was preceeded by a filter to prevent the cells from becoming fouled. This would tend to make the
readings presented low by the amount of PA and other heavy organics that might come out in the
filter. In clean applications and for measurement of low hydrocarbon concentrations, such as an
incinerator outlet, a portable Delphi Model C Flame lonization Detector was used. Vent flow rates
were measured with a precalibrated orifice and magnahelix gauge.
Emission measurements from other closed kettles are graphically presented in Figures 25,
26, 27 and 28. Figure 25 shows the emission from an epoxy cook. The graph is fairly self-explanatory.
Note that the kettle is closed-in and operated at 30 psig pressure for four hours.
Figure 26 shows emission rates from a 2,000 gallon resin reactor processing an alkyd
cook by the fusion method. Vent flow during cooking ran at a fairly steady rate of about 12 SCFM.
As indicated on the graph there was a decanter foam-over at hour seven when hot PA and solvent
were added to the batch. Although this is not normal operation, it is not an uncommon occurrence
in the industry.
Figure 27 shows emissions from a 1,500 gallon polyester fusion cook. The graph and
accompanying data are self-explanatory and show no unusual features.
Figure 28 shows emissions from two 1,000 gallon closed kettles. One kettle was processing
128
-------
Time
8:00 AM
8:45 AM
9:00 AM
9:30 AM
10:00 AM
10:30 AM
11:00 AM
12:30PM
2:45 PM
Mrs
0
3/4
1
1-1/2
2
2-1/2
3
4-1/2
6-3/4
Hydrocarbon
(% LEL)
225
270
325
340
300
275
Liquid
Temperature
100
450
450
440
450
440
305
260
IG
Flow
(Meter) Remarks
02
02
02
02
02
05
0 1
0 1
01
Start heat up
Exotherm
Close vent — add solvent
Pressure up to 24 psig
Vent remained closed until about 2 45 PM
450
400
UJ
300
200
IOO
I v
' EXOTHERM \
\
*
1^
OPEN VENT
VENT CLOSED
ADDING SOLVENT
I
I
450
400
300
200
UJ
or
01
UJ
Q_
S
UJ
100
456
TIME, HOURS
RGURE
8
EMISSIONS FROM IOO GALLON EIPOXY
SOLVENT PRESSURE COQK— -3O PSIG
129
-------
Time
10:20 AM
10:30 AM
10:45 AM
11:00 AM
11:30 AM
11:45 AM
12:00 noon
1:00 PM
1:30 PM
2:00 PM
2:30 PM
3:00 PM
3:20 PM
Hydrocarbon
(% LEL)
25
350
350
300
220
150
230
200
190
180
180
180
180
IG*
Flow
(SCFM)
10
20
30
30
10
30
30
60
60
60
80
80
Emissions
(Ib/hr)
4.4
8.8
11.3
83
1.9
8.6
7.5
14.3
13.5
135
18.0
18.0
Remarks
Prior to sparging
Start sparge
Increase sparge rate
Increase sparge rate
Make DPG correction addition
Sparge back to normal
Increase sparge rate
Increase sparge rate
*Flow rate based on rotometer reading.
'Emissions calculated as pounds/hour as methane (based on LEL measurements).
400
UJ
^ 30O
I
I
-------
300
• DECANTER
FOAM OVER
250
VENT FLOW - 12 SCFM
200
ISO
8
g
150
500
INCREASE I.G.
SPARGE
5O
HOLDING AT380°F
FOR PA. ADDITION
LINE PLUGGED
ADD RA. AND
SOLVENT
400
300
200
U
cr
UJ
a.
S
UJ
100
I
I
2345678
TIME , HOURS
FIGURE 2-7
EMISSION FROM ISOO GALLON
R EACTOR FUSION COOK
K)
131
-------
-ADD CATALYST THRU
HATCH REACTOR #2
150
-LOAD P.E. IN
THRU HATCH
125
IOO
X
I
O
u
O
i_
TJ
>»
X
50
LOAD RA. THRU HATCH
IN REACTOR #2
OPEN REACTOR #1
HATCH SAMPLE TAKEN
U
2 REACTORS
I- POLYURETHANE
ALKYD
2-INTERMEDIATE
BASE FISH OIL
ALKYD
25
SAMPLE DILUTED 10 TO I
FLOW- 138 TO 149
FIGURE
EMISSION FROM TWO-IOOO GALLON REIACTOF^S
132
-------
a polyurethane alkyd and the other kettle was processing an intermediate base fish oil alkyd.
Measurements were made on the inlet of a 500 SCFM thermal incinerator. The gases from the
two kettles were combined and had passed through one water scrubber and one water curtain
prior to measurement. It was also diluted about 10 to 1 with air prior to measurement. Flow at the
point of measurement was about 150 SCFM as measured by a pitot traverse. The exhaust was
further diluted prior to entry into the incinerator. Maximum hydrocarbon emissions ran around 31
Ib/hr. As indicated on the graph, there were a number of peak emissions throughout the period
of measurement. All were in excess of safe operating limits for the thermal afterburner.
In order to meet requirements of 1/4 LEL (13 Btu/SCF) and thermal incinerator inlet loading
requirements of 9 to 12 Btu/SCF emissions are diluted with air prior to being exhausted or sent
to a pollution control device. In most cases the streams involved are passed through some type
of crude water scrubber to remove heavy oils, resins and condensible phthalic anhydride. The
scrubber is normally retained as a pretreatment and safety device if an incinerator has been added
for final pollution control. Hydrocarbon concentration after air dilution will run from 1,000 to 4,000
ppm after an adequate water scrubber. Exhaust rates will average around 1,000 SCFM for a 2,500
gallon reactor capacity. This rate will vary significantly, however, if other vent streams from thinning
tanks, filter press and the like are included with the exhaust system.
D. Process Operations Influencing Emissions
1. Equipment and/or Process Characteristics
a. Handling and Storage— Materials handling activities are of two types: liquids and dry solids.
Bulk liquids are transferred wherever possible through pumps and meters. This confines emissions
in liquid transfer operations primarily to displacement of the atmosphere in the container being
filled. For a covered container it should be possible to estimate by engineering calculation what
these losses will be. If the container is open (for instance, an uncovered tub), the amount of solvent
lost through evaporation will be more difficult to determine. Breathing losses, due to changes in
temperature and atmospheric pressure, must also be considered. However, if the turnover of the
storage tanks is high compared to the size of the tanks and if the tank is not subject to large
short term temperature fluctuations, this breathing loss should be less than the filling losses.
The finished product filling area represents another source. Emissions from this area would
normally be part of the building ventilation exhaust. Determination of the quantities involved must
depend on operating experience and measurements.
133
-------
Liquids handled in drum size quantities are subject to evaporative losses during dumping
operations. Again these would normally show up in the building ventilation. The potential for emission
on a per gallon basis is higher for liquids handled in this form since the dumping tends to cause
more agitation of the liquid and tends to be performed into open tanks.
Dry solids handling is a source of particulate emissions. Materials of this sort are usually
received in 50 pound bags or in fiberboard drums. Until the bags are opened and dumped no
emissions will occur from this source. The bags are usually opened at the station at which they
will be used. These include mills, mixers and dispersers in the case of pigments and the resin
reactors in the case of acid anhydrides. Particulate emissions will also be part of the building
ventilation system.
b. Paint Operations — The production of paint products themselves consists of milling, dispersion
and mixing operations. Ball and pebble mills are free of gaseous emissions, except when loading
and unloading, since they are completely closed during operation. Sand mills often discharge into
open portable tanks. Since such tanks are not equipped with covers and vents, there exists a source
of solvent evaporation associated with the operation of the sand mill. Sand mills can be set up to
discharge into closed, vented containers. This will reduce evaporative losses. The high speed
dispersers sometimes operate in open tanks and provide a source of solvent emissions. Often,
however, finishing tanks are equipped with tops and top-mounted agitators. This should serve to
reduce emissions from these sources.
Most of the emissions from the paint manufacturing area will tend to be carried out with
the ventilation system. Hoods or local ductwork can be used to collect emissions from points
within the paint manufacturing area. In other cases, no attempt is made to contain emissions and
evaporation is directly to the plant ambient atmosphere. Particulate emissions from this area will
be minimal except when loading dry pigment into the equipment.
c. Resin Production — Resin processing represents a major source of emissions for the coating
industry.
Fusion cooking is characterized by a continuous sparge of inert gas. Intermittently during
the process, the hatch is opened for charging raw materials. During this time the volumetric flow
rate, and associated emissions, are increased considerably. The flow rate can increase during
charging to better than 100 cfm.
Since there is less inert gas used in solvent cooking, exhaust volumes are small and consist
primarily of noncondensibles. As in the fusion process, however, the rates increase during charging
134
-------
of raw materials. Condensers are used on solvent reactors but must be kept warm enough to prevent
accumulations of condensed phthalic. It is common practice in the resin industry to maintain a
small inert gas flow throughout a solvent cook to avoid the possibility of leaking air into the
system.
Thinning is usually done in separate tanks. The resin batch is "dropped" by gravity feed
or pumped into a thin tank which is equipped with a condenser to reduce solvent losses. In some
cases thinning is done in the resin kettle itself. This usually results in a higher solvent emission.
2. Raw Materials
The particular raw materials used in a given plant influence greatly both the type and
quantity of emissions produced. Emissions potential consists of two problem areas: (a) particulate
and (b) gaseous.
a. Particulate — Pigments and extenders account for a major portion of the particulate emissions.
The extenders, along with TiOz, represent the major groups of pigments. Others which find use
include iron oxides and carbon blacks. Representative particle size ranges for these products were
given earlier in Table 29. Particles outside these ranges are found, but the majority of particles of
a particular pigment material, as manufactured, fall within the ranges given.
Some pigments are now available already dispersed as a liquid slurry. The use of this type
of raw material will significantly reduce particulate emissions.
The other source of particulate emission consists of phthalic anhydride, PE, or other acids
and polyols. This is connected solely with the resin production operation, particularly during, and
immediately after charging. This type of emission can be reduced by the use of liquid phthalic
anhydride. Sublimed PA can also result in a particulate emission in that the vapor may condense
into a fine fume as it leaves the high temperature sections of the equipment. Phthalic also has
the unfortunate tendency to plug and foul condensers and similar equipment.
b. Gaseous Emissions — The primary source of gaseous emissions is the solvents. These are
potential air pollutants in all phases of the operations including handling, storage, resin production
and dispersion and mixing. Basic parameters influencing emissions of solvents include vapor pressure
and molecular weight. The rate of evaporation can also depend on such things as diffusivity,
surface tension and heat of vaporization. The influence of these properties is less, however, and
is rather difficult to predict. Table 26 gives vapor pressures and molecular weights for some of
the more common types of solvents encountered in the industry. The molecular weight is important
in that it affects the quantity of emissions on a weight basis.
135
-------
Oils and glycerine are relatively non-volatile and so do not contribute significantly to quantities
emitted except at very high temperature. They can, however, be the source of reaction by-products
which can present odor problems even in small quantities, or the source of aerosol mists.
3. Start-up and Shut Down
Start-up operations should present no unusual problems provided the appropriate condensers,
fans, scrubbers, etc., are turned on at or near the beginning of the process. If the process heat
and purge gas is allowed to commence operation before these items are turned on, a significant
pollution problem could exist during start-up. It is assumed that kettles, etc., would be purged
with inert gas prior to addition of any volatile material.
Shut down operations, particularly if done under emergency conditions, offer a greater
potential for air pollution. Good practice would dictate purging lines and vessels with inert gas to
eliminate residual solvent and PA. These materials would in all likelihood be vented to the atmos-
phere. A complete shut down might also involve solvent transfer operations with associated
displacement losses.
Shut down of resin reactor in many cases is accomplished by blowing the reactor materials
into the thinning tank. This is done by pressuring the tank with inert gas. This inert gas becomes
saturated with solvent and must eventually be vented to the atmosphere through the reactor
vent or thinning tank vents. This normally results in a highly concentrated emission for a short
period.
4. Operation Above and Below Capacity
Aside from the obvious statement that emissions are roughly proportional to production
output, operation at other than rated capacity should, in itself, offer no unique problems. Productivity
increases are accomplished primarily by increasing the number of batches. Since equipment is
sized on a per batch basis, the plant should operate smoothly.
The potential problem areas would more likely involve employee diligence in adhering to
proper procedures. As productivity is increased beyond normal levels there might be a tendency
for plant housekeeping and control to become sloppy. Operation significantly below capacity might
present the temptation to cut operating costs by turning down the operation of scrubbers, etc.
The effectiveness of supervision and management will determine the extent to which emissions
are adversely affected by production changes.
5. Process Operation Upsets
Several events can occur which could have an adverse effect on emissions. Loss of power
136
-------
would cause a shut down of pumps, stirring devices, controllers, etc. If the dowtherm heater is
at a higher elevation than the reactors, a pump will be required to return condensate to the heater.
Consequently, fail safe devices must activate which shut off heat to the reactors to prevent accumu-
lation of condensate. Condenser water flow will continue while the scrubber-ejector will cease to
function. The inert gas generator will shut down; though the inert gas storage tank will provide a
continued supply of sparge gas for a short length of time. The plant will revert to a "status quo"
situation in which air pollution effects should be small.
If power is returned soon enough it will be possible to resume operations at the point of
interruption. It may be desirable initially to increase the sparge rate in a fusion process reactor
until water which has accumulated is removed. Likewise, a period of sparging of the solvent process
reactor may be required for the same purpose. For some interval of time after power is resumed,
an increase in the emission rate may be observed due to increased sparging.
Condenser failure (due to loss of cooling water, leaks, or other causes) would have unfavor-
able consequences. The reactor vent would have to remain open until the reactor had cooled
sufficiently to lower the solvent vapor pressure to a safe value. Even if the sparge gas is turned
off immediately, a considerable amount of solvent would be lost. It is likely that the batch would
be lost. It is likely that the batch would have to be cooled immediately with cold dowtherm and,
unless there is reason to believe that condenser operation will resume soon, dropped to the thin
tank. If condenser failure is on the reactor condenser only, this should result in minimum emission.
If, however, there is a general loss of cooling water to the plant there seems no way to avoid a
significant emission problem whether the emissions came from the kettle or from the thin tank.
Partial failure of condenser (due to fouling, improperly operating controls, or other causes) is common.
Often processing is continued in spite of this with a significant increase in emission levels.
Finally, there are several other upsets which can occur to the detrement of air quality.
Any upset which results in a higher than anticipated temperature (e.g., failure of temperature
controller, improper charging of reactants, etc.) will tend to increase emissions. Any increase in
sparging rate will tend to increase emissions. Use of inert gas pressure to force resin from the
reactor into the thin tank (e.g., if the transfer line is tending to plug) will cause increased emissions.
As discussed earlier, when the last of the resin comes through, the pressure in the reactor will
be released through the thin tank condenser vent causing a surge in the amount of solvent vapors
emitted from the thin tank.
137
-------
E. Raw Data Tabulated
1. Questionnaires — Emission data has been reported by some of the plants responding to
the industry questionnaire that was distributed to a sampling of the industry. The data for those
plants having reasonably complete emission inventories is summarized in Tables 38 to 40.
There is wide diversity in the manner in which the data has been submitted. The numbers
as reported are difficult to interpret or compare as they stand. It is felt that the only meaningful
presentation of the data is to relate total emission rate to production. Emission data by itself contains
limited information unless the production rate is also known. This correlation will be presented in
Chapter 4.
The tabulation of emission rates do provide an indication of possible non-compliance
(depending on local regulations) by some plants. Also, it suggests the range of values that might
be encountered in source testing.
One of the paint industries more persistent problems has always been odor nuisance
complaints. Data on this subject has also been reported in the questionnaires. Of 76 Type 1 plants,
15 reported odor complaints from residential areas and six from commercial areas. Of the 39 Type
3 plants, nine reported complaints from residential sections. None of the Type 2 plants reported
odor nuisance complaints. This information is tabulated in Table 41.
The detection and measurement of odors is a very difficult area to predict. The odor levels
around a plant depend not only on the types and quantities of odorants emitted but also on local
atmospheric conditions, presence of competing odor sources, and proximity of neighboring residential
or commercial areas. Ultimately, subjective criterion must be relied upon.
The information reported by the questionnaire sample suggests that resin processing
represents the primary odor problem. This may be due to the presence of high temperatures which
tends to produce small quantities of particularly noxious substances.
Industry source test results have been listed in various parts of this report where it was
deemed most appropriate. Test results were obtained by air pollution control equipment manufacturers
as well as paint and resin producers. Information on reactor emissions can be found in Section ll-C-2
of Chapter 1. Information on performance of control equipment can be found in Section IV of Chapter
5. Test results obtained by the Federal EPA can be found in Section II-G of Chapter 1 and in Appendix D.
F. By-Products
The only reaction by-product present in any quantity is water formed in the esterification
reaction by which alkyds are produced at about 4 to 5% by weight of the final resin product solids.
138
-------
co
CO
TABLE
CO
LU
DC
Z
g
CO
LU
0
0
DC
LL
Q
O
CO
CO
^
LU
CO
h-
<
CL
Z
CO
LU
DC
Q
Z
z
CL
*0>
N
CO
c
co
CL
= 1
0
CO .
«!^ 1^—
_i .2
o— -
CO 1
CO
0)
JS
o
t
03
CL
C
.g
•J3
CO
o
Q.
co
>
0
03
E?
O
^
0
CL
C
'co
0)
DC
E
2
LL
.g
o
3
o
CL
C
'CO
CL
E
0
i_
CO
q
o'
C\J
co
LO
O
oo
o>
q
oo
co
co
co
co
(D
C5
00
LL
O
CO
8
o
co"
LL
O
CO
O)
if)
n —
o
i- O
T^ CO
to
£ o
B ro
— xi
o> ^
I? oo
£
o
CT)
o
o
03
CD
i; — x:
— 8 ^
CO O to
-g
CM
CO
o
CO CM
o
o
LO
0)
5?
T!
(O
03
co
•
co
c
o
18
CO
139
-------
TABLE 39
EMISSION DATA FROM QUESTIONNAIRES
PAINT PLANTS
Organic Paniculate Plant Size*, MM Gal
5.4lb/hr 3.1
2.09 Ib/hr 0.5
8.94 Ib/hr 0.6
30.0 Ib/hr 3.3
23,650 Ib/yr 0.2
1,300gal/yr 0.1
2 Ib/hr 1.3
10 Ib/hr 0.4
183 ton/yr —
3,425 gal/yr 0.3
4 Ib/hr 2.2
5 Ib/hr 0.2
48.9 Ib/hr 15.4
8.7 gr/SCF @ 800 SCFM —
2.78 Ib/hr —
68,000 Ib/yr 2.7
248 Ib/day 1.7
*Plant size represent solvent based production only and is based on 16 Hr operating day
6 day work week
140
-------
TABLE 40
EMISSION DATA FROM QUESTIONNAIRES
RESIN PLANTS
Organic Emissions Plant Size* (MM Lb)
360 Ib/hr 20.5
2 Ib/hr 3.8
3 Ib/hr 8.6
8.7 Ib/hr 25.5
15 Ib/hr 3.6
1.27 Ib/hr 39.1
66.6 Ib/hr 21.7
3.3 Ib/hr 14.7
4.5ib/hr 1.8
24 Ib/hr 11.7
7 Ib/hr 10.3
39.1 ton/yr 4.0
15.7 Ib/hr 60.3
668 Ib/hr** 47.0
'Based on 16 hour day — six day week
"Most of this emission is from "Resin Dryers"
141
-------
ro
^ o)
c „,
ll! T-
CN O) «-
in *$; q
O> «-' ' r^
CO C —
CD t-
cn co
co in
q oq
•*' d
CN ^-
CO
_l
CO
o
Q.
CO
Q
|
cr
LU
1-
s
LU
2
1
LL
O
cr
2
CO
LU
cr
^
2
O
CO
LU
ID
O
CD
•*-"
CD
>
~o
CD
^
5
3
CO
to
Q
c
o
o
T3
£
Q
*"^
C
ro
cu
C_>
O
*^
CO
o
£
>
o
c/)
0>
u
->
"ro
en
LO
»—
(~
6
s
C)
CO
en
CU
t!
CO
^f
CO
CO
^ «—
S^
"in
« 0
ro
g
«
c
L-
ro
o3
c
'«
cu
OC
O)
C
'.p
CJ
CO
4-1
O
s
8
to
CO
t!
0)
in
CO
^. — §
n 7
O —
CO
o
co^
in"
CN
-^
15
D5
"o
^~
0
CN
I
CM
CO
CO
*-
CO
-Q
"
E
z
CO
r—
CO
CN
""
CO
&
>
r-
§
CO
CO
1
1
CO
q
CO
1
1
en
CO
cn
iri
o
O5
en
o
CM
CM
in
en
CO
in
CM
in
^
CO
CM
>-'
CO
CM
in
00
CD
co^
co"
1
1
o>
CO
CO
00
px
5
en
03
LO
en
oq
en
ci
d
r-
CM'
o
r*^
^~
T—
,_
•*
CO
O)
CO
CO
l^%
CO
CM"
CN
CO
2
0
CO
LU
DC
O
&
LU
O
O
cr
LL
CO
o
o
cr
O
a
O
CO
"c
'co
Q.
E
8
o
o
c
CO
CO
3
Z
^_
O
§
co
'o
0)
E
E
o
O
~a
'c
CD
T3
'co
cr
fe
JO
a.
co
in
o>
(O
CD
CM
CM
O)
CO
co >?
a> c
*- o
t c
E '«
°co Q-
0.
o
'co
CD
cr
142
-------
In addition to water, various reaction by-products are formed in very small quantities. Often
these are detectable only by very sensitive analytical procedures. These products are important,
nevertheless, in that they constitute a primary source of odor problems even when present in
minimal quantities. These substances include aldehydes, esters and organic acids.
1. Liquid Wastes — Waste materials constitute a major source of potential liquid pollutants.
These include spoiled batches, residues and solvent and aqueous solutions for washing equip-
ment. The industry questionnaires have provided a source of information on this subject.
The questionnaires requested information on the amounts of resin and of paint disposed of
as well as solvent and water usage for clean up. The results have been tabulated and are given
in Table 41. Note that even plants which produce no water based coatings reported use of aqueous
solutions (probably caustic) for washing purposes. It would not be entirely accurate, then, to assign
all the aqueous waste to water based paint production. Nor, in the case of Type 1 plants, is it
possible to assign the portion of waste which is attributable to paint production as opposed to
resin production. It is apparent from the questionnaires that the term "kettle" means different things
to different people. To some, "kettle" is reserved for resin and varnish cooking vessels while in
other cases mixing and finishing tanks are also included in the term. One cannot necessarily
assign washing solutions for kettles to resin production in the case of Type 1 plants. Some generali-
zations can be made:
a. Aqueous waste far exceeds solvent wastes for all types of plants.
b. The major portion of solvent wastes can probably be attributed to coating production
as such.
c. Waste resin and paint account for less than 0.5% of shipments.
2. Solid Waste — Most solid waste, with the exception of that which can be considered part
of an air pollution emission, is incorporated into the liquid wastes described in the previous section.
These include pigment particulate and latex emulsion as well as the non-volatile portion of the
film former which would be left if the paint or resin were allowed to dry.
G. EPA Source Test Data
The U. S. Environmental Protection Agency retained Scott Research Laboratories, Plumstead-
ville, Pa., to source test several resin kettles. Flow rates, total hydrocarbons and gas chromatography
data were obtained from several processes. The test methods used are described in Chapter 3,
Section I-C. The results are presented in Appendix D. Portions of this section have been taken from
143
-------
the report prepared by Scott Laboratories for the E.P.A.
A summary of the batches tested follows:
Three basic types of varnish cooks were tested. These consisted of a polyester fusion cook,
an alkyd-fusion (oil modified polyester) cook, and an alkyd-solvent cook. In addition, some thinning
operations were tested. A detailed description of each process tested is given below.
Linseed Soya Alkyd Fusion Cook
This alkyd resin is a long oil alkyd produced from linseed oil, soybean oil, pentaerythritol,
and molten phthalic anhydride.
The two oils were added to the kettle and heating began. After the oils were added, a
blanket of inert gas was maintained over the batch. When the oils had heated sufficiently, the
pentaerytritol (PE) was added. The steam ejector was turned on in order to create a vacuum in the
stack while the PE was being dumped from bags into the reactor porthole. The temperature was
raised and a liquid catalyst was added. The batch was then heated to 440°F and kept at this temperature
until samples of the batch passed a clear test (test used to indicate completion of reaction). The
temperature was then reduced and molten phthalic anhydride was added. The batch was reheated
to 460°F; upon reaching 460°F, the inert gas blanket was removed and 20 SCFM of inert gas was
blown into the batch from the bottom of the kettle. This inert gas blow-through aids in removing
water produced in the process reaction. The inert gas flow as increased to 30 SCFM and maintained
until the proper acid value and viscosity were measured. Samples were obtained by opening a
porthole and removing some resin by means of a dipper. The 30 SCFM inert gas blow was removed
and a 5 to 10 SCFM blanket of inert gas maintained while the resin was cooled to 425°F. The batch
was then dropped into a thinning tank containing solvent.
A description of Kettle 4 may be found in Appendix D, Figure D-1.
Soya Alkyd-Fusion Cook
This alkyd resin is a medium oil alkyd produced from soybean oil, pentaerythritol (PE), and
molten phthalic anhydride. The oil was heated and then part of the PE added. A clear test was run
and then the remainder of the PE and molten phthalic anhydride was added. The steam ejector
was on only during the time that PE was being loaded into the reactor. A 5 to 8 SCFM inert gas blanket
was maintained during the initial period of heating and loading the reactor. A 10 SCFM inert gas blow
was started when the cook reached 480°F and was increased to 20 SCFM during the last part of the
cook. Resin was cooled to 430°F and dropped into solvent.
144
-------
o
'co
x
CO
o
o
o
cq
c\i
in
o
CD
o
o
o
o>
in
in
N.
cri
co co •—
co r- co
CO o CD
Q « C
I E 2
I- LU >
en
CM
co
d
C\J i-
m co
oo
p
o
CM
c\
oo
r--.
CM
CO
LU
Sc!
"D 'CO 3
c 0 -o
go: 2
D- Q-
0
o
m
0
o
m
m
0
co
m
O
o
co
m
o o o
m co o
co o o
m rf ^-
(M CM
co
r^
CD
CD
o
in
CD
CM
LU
m
o
00
Q
LU
£
LU
LL
O
0
CO
o
o
o
o
co
CM
m
c\i
c\j
m
m
CD T^
i- CM
oo oo
c n
3 c
CO
CM
(M
CO
CM i-
CM
CO
CO
(M CM
(M
oo
CO
CO
0!
§1
Q-i
C CO
.2 $
"-6
T3 0
>< CO
=£ C
< S-
c jo
.9 %
CO co
T> 0
^g
< =5-
c
.2
CO
T3
<
CO
c
o
C/5
LL ^_^
T3 CO
JT §"
< -2-
co
LL
T3 CO
-^ O
0
C
0
"o
CO
C
0
— — — c
o ~ -
co
c
'co
0
DC
0
oo
0
"o
0_
c
'co
0
DC
0
"co
0
"5
0-
145
-------
A description of reactor number three is also shown on Figure D-1* since reactors three and
four are identical.
Soya Alkyd-Solvent Cook
This resin is considered a long to medium oil alkyd made from soybean oil, pentaerythritol,
crotonic acid, and phthalic anhydride; it is used as a major ingredient in producing an acrylic resin.
Soybean oil, pentaerythritol and crotonic acid were placed into the reactor with the steam ejector
turned on; fumes were vented out the main stack (A) during this time. Upon completion of loading,
the ejector was turned off and a 10 SCFM blanket of inert gas was maintained over the batch while
it was heating. Upon reaching temperature molten phthalic anhydride was added, the main stack
was closed, and the kettle sealed. Xylene was added and the inert gas flow turned off. The condenser
was turned on and the batch was heated so that the solvent and water (reaction product) began
refluxing. During this period of the process the emissions were vented from the receiving tank stack
(B). The water driven off the kettle was drained from the receiving tank several times during the
cook; the solvent was returned to the batch. After the cook was completed, the batch was cooled
to 400°F and dropped into an empty thinning tank.
A description of reactor number two and associated equipment is presented in Figure D-2.*
Polyester Cook
This resin is a saturated polyester produced from propylene glycol, butylene glycol, glycerine,
and dimethyl terephthalate (DMT). This resin is used in producing powder coatings. The use of
DMT to fulfill the acid functionality results in the evolution of methanol rather than water as a reaction
product. The steam ejector was turned on while the raw materials were being added. During this
time emissions were vented to the main stack "A". The solvent was then added and the reactor
was sealed. Heating of the batch was begun; fumes were vented from stack "B" during this time.
As the temperature increased distillation began. The temperature was then held constant. When the
amount of distillate coming off decreased considerably, a 10 CFM inert gas blowthrough was begun
to step up the reaction.
All of the distillate collected in the receiver tank was drained into drums. When the reaction
stopped, the batch was placed under a vacuum. During this time all emissions to the atmosphere
were vented to stack "C". The batch was left under vacuum until the proper viscosity was reached.
Upon completion the resin was dropped into 55 gallon drums.
'Appendix D
146
-------
A description of reactor eight and its associated equipment is presented in Figure D-3.*
Thinning Operations
With the exception of one case, the thinning operations consisted of dropping the completed
resin directly from the reactor into a thinning tank containing the appropriate amount of solvent.
In one case, the resin was dropped into an empty tank and then the solvent was added to the
resin. In all cases, the emissions to the atmosphere simply consisted of fumes forced out of the
tank by positive displacement. All the tank vents were equipped with condensers to reduce emissions.
Figures D-1, D-2, and D-3* indicate the thinning tank, condensers and vent locations.
Discussion
The primary difficulty associated with measuring mass emissions from the various fume
stacks was the extremely low flow and "breathing" flow situations encountered. For most of the
larger stacks (8-9 in. diameters) where the pitot tube and hook gauge were used, the pressure
differentials were, in many cases, approaching the minimum sensitivity of the apparatus. In sampling
the smaller 2 inch stacks where the bag systems were more easily adapted, more confident data
were obtained. Where the "breathing" flow situations occurred, a high degree of care was required
to insure that the one-way flow valve was operating properly.
After having experienced the various flow situations associated with this type of process
operation, a more sophisticated flow measuring system could probably be engineered for future
investigations.
The kettle operator also had a direct effect on the flow rate at certain times during a particular
cook. He had control over the quantity and duration of inert gas and steam ejection, and as a
result, two different cookers processing similar batches of resin could create dissimilar mass emission
rates at the kettle outlet.
The results of the Orsat and chromatographic analyses for oxygen and carbon dioxide in
the bag samples yielded a large variation in stack gas molecular weight for the various stack gases
sampled. As a result an average molecular weight was determined using the extremes of 28.96
for ambient air and 29.88 for the inert gas stream that was used as a gas blanket in all the process
operations. This average molecular weight of 29.42 was used in all stack gas velocity measurements.
All the flowrate averages used to calculate mass emissions are included in Tables D-1
to D-16.*
'Appendix D
147
-------
The total hydrocarbon measurements were reasonably complete with the exception of the
reflux cycles on polyester resin cook #1 where the heated prefilter was not closed properly and
a leak occurred.
Upon examination of the data included in Tables D-1 to D-7*, it can be noted that relatively
good comparison exists between the average ppm-C levels for similar modes within process run
pairs. A similar comparison involving mass emissions, however, does not yield the same reasonable
duplication. The problem exists with the variations in flowrates experienced between similar modes
of the process run pairs. Some explanations for these variations have been discussed previously.
Tables D-1 to D-7* describe the process mode breakdown for each cook and includes
average flow and mass emission rates for each process mode. The flow and mass emission data
for the eight thinning tank tests are presented in Tables D-8 to D-9*.
All gas chromatography traces were reduced by first identifying specific hydrocarbons,
where possible, and then measuring peak areas and comparing them to standards. In an effort
to condense the data in a more useable form, the specific hydrocarbons were reduced to four main
groups: Aliphatics (Ce-Cio), Aromatics (Cs-Cs), identified oxygenates, and branched Aliphatics
(C8-Ci2). The data summaries (Tables D-10 to D-16)* present each process cook sampled with
average flow rates, total hydrocarbon emission rates, and the percentage of specific hydrocarbons
found in each of the above groups.
A summary of thinning tank emission data is presented in Table D-9* using the same format
described previously.
The GC data obtained for the alkyd solvent cooks were a great deal easier to interpret
than that of the fusion cooks due to the basic three component makeup of their effluents. One
difficulty arose when an attempt was made to compare the total carbon found with the hydrocarbon
analyzer, to that found with the gas chromatograph.
Even though the total hydrocarbon analyzer and the gas chromatograph were both adjusted
for carbon response, the data did not always coincide. The following sources of error could explain
this discrepancy.
1. The total hydrocarbon analyzer gives a continuous plot of the hydrocarbon concentrations, while
the G.C. gives only a point sample. With the fluctuating organic concentration, this point sample
may or may not be representative of the effluent.
2. Since the nature of the work did not permit duplicate injections into the chromatograph, an
* Appendix D
148
-------
injection error, perhaps sample pressurization, could have occurred.
3. The concentration of xylene was so high that even with the .25 cc sample loop, the column was
nearly overloaded. This could add another non-linearity aspect to the system.
149
-------
CHAPTER 2
INDUSTRY STATISTICS
I. TYPE, SIZE AND LOCATION OF PRESENT DAY PLANTS
The Paint and Allied Product Industry (SIC 2851) is made up of approximately 1,727
establishments operated by some 1,365 companies.7
These plants and companies primarily manufacture coatings such as paints, varnishes and
lacquers, along with such allied products as putty, caulking compounds, cleaners and other paint
sundries.
Primary to the Paint Industry are two broad categories of products, generally referred to
as "Trade Sales" and "Industrial Finishes".
Trade Sales products are usually "shelf items" to be used on the exterior and interior
surfaces of houses and buildings. These are manufactured in a wide range of colors and are
usually applied with brushes or rollers, although other application techniques are also employed.
Industrial Finishes are generally produced for and sold to other manufacturers for appli-
cation to durable customer products. The final application varies from spraying, dipping, etc., to
electrostatic methods.
Although some of the smaller manufacturers still tend to specialize in either Trade Sales
or Industrial Finish products, this tendency is not as strong as it has been in the past. Most of the
larger companies now produce both types.
A breakdown of the industry by product class is provided in Figure 29. The major categories
indicated are listed below in order and with product class codes:
(2851) 1. Exterior, oil-type sales paint products.
(2851) 2. Exterior, water-type paint products and tinting bases.
(2851) 3. Interior, oil-type trade sales paint products.
(2851) 4. Interior, water-type paint products and tinting bases.
(2851) 5. Trade sales lacquers.
(2851) 6. Industrial product finishes, except lacquers.
151
-------
10,000
5,000
1,000
500
U.S. PAINT S. VARNISH
INDUSTRY
DISTRIBUTION
BTV
PRODUCT CLASS
VALUE OF SHIPMENTS . IOSf/YR
NO. OF EMPLOYEES
NO. OF ESTABLISHMENTS
SOURCE : I96T CENSUS OF
MANUFACTURERS
FIGURE 29
PUTTY,
CAULKING,
ETC.
MISC.
PAINT
PRODUCTS
INDUS.
LACQ. INCL
ACRYLICS
INTERIOR
TYPEIWTR. TYPE
TRADE TINT. BASE
INDUS.
EXCEPT
LACQUERS
EXTERIOR
OIL TYPE WTR.
TRADE TINT. BASE
152
-------
(2851) 7. Industrial lacquers, including acrylics.
(2851) 8. Putty, caulking compounds and allied products.
(2851) 9. Miscellaneous paint products.
This histogram uses abbreviations of the above categories and is a plot of value of shipments,
number of establishments and number of employees for each of the (product) classes. Statistics
are based on the 1967 Census and no effort has been made to estimate corresponding figures
for other years. The number of establishments shown accounts for little more than 50% of the
total and provides a distribution of those showing some degree of specialization.
Figure 30 provides a distribution of paint plants by size as measured by production value.
It was based on estimates from data for individual companies reported by Kline's Marketing Guide
to the Paint Industry.1 This distribution follows the log-probability law rather well and is, consequently,
plotted on a probability x 2 log cycle grid. In practice, the small percentage of very large plants
show wide variations, partly because data on an individual plant basis is often withheld In addition,
it is common in statistical data of this nature to find that the extreme values do not follow the log-
normal distribution.
The example, shown by arrows in Figure 30, illustrates one of the useful aspects of a
log-probability plot. At the $6.4 million plant size lies the median production — half the paint is
produced by plants smaller than (or larger than) this size. Only 7% of the plants, however, are
larger than this size. The median plant produces somewhat less than $1 million worth of product.
It is of interest to note that the Model Plant discussed in Chapter 1 produces product
valued at about $6.6 million. This production value level is indicated bv an arrow in Figure 30.
Figure 31 provides a distribution of plants and industry employees by plant si?M as measured
by the number of people who work in the plant. It is based on figures taken from County Business
Patterns.''3 This Bureau of Census publication lists the number of plants in various size ranges
such as 1 to 3 employees, 4 to 7 employees, etc. From these tabulations, it is possible to obtain
the percentage of plants with fewer than n given number of employees Figure 31 plots percentage
of plants smaller than indicated plant size The total number of employees in any plant size can
also be computed and expressed as a percentage of total employment in the Paint and Varnish
Industry. For example, as shown by arrows on Figure 31, 30% of the plants in the industry employ
less than 8 people, 30% of the industry employees work in plants that have a plant employee
size of less than 50, and this plant size accounts for 78% of the industry plants.
Statistical measures of distribution: median, mean and mode have been included in this
153
-------
ssvno azis
154
-------
IUO9
mi
JMTI
Tnn
•on
u
u
Urn
.
50
K 40
T VI
y
N
ii
1-
ZM
<»
•
J-
/
f
f
x^
t a
. F
W7\
SOOf*
EMf
MED
MC»
MOC
x
K
^Jh
DIS
Etv
=E:
1_C*
.N>
•E >
/
J
/
) C
•J 1
• 1 I-
1PL
»71
PAH
EE!
1
/
X
/
5 2
|
*IBU
-OYE
ccx
-TER
FIG!
» £
, -100
'•lie
• 7\
/*
/
/
/
o a
\/A
rno
:ES
JMTV
NS
JRE
»
b
/
/
/
{
0 4(
>JRK
N <
BY
' Bt
: 3
/%j
7
s
/
/
) S<
HSr
3F"
Pl_
JSIN
I
/
^
9 *
PL_/
AN!
ESS
/
/
/
9 71
SJDL.
MSTTJ
r si2
/^
M3
/I
7
PI
Mi
IvK
9 *
IS
!•>
^E
/
X1
r/Kf
r>i>
PjAJ^
3OC
9 •
|R>
*•
S
/
t
/
/
JTS
,N 5
X
xl
5 91
< /
/
' :
/
o
- n.:
9 9!
/
7
*
•
"> 9t
RCENTA<3E-1_ESS THAM IMOCATED SIZE
155
-------
figure. These measures were similarly computed for past years and changes in them were used to
make projections for future plant sizes. Further discussion on projections will be found in part II
of this section under the subtitle "Size of Plants."
Geographical locations of areas of concentrated production are indicated in Figure 32. To
arrive at volume production for each state in 1972, state-by-state dollar production figures for 1967
(as reported by the Census of Manufacturers) were converted to gallons using average cost per
gallon, as computed from Current Industrial Reports. Computed production growth rates were used
to arrive at 1972 production volumes of each state. Since these volume production figures represent
a second generation estimate, they are indicated as ranges on the map. Estimated volumes of
the eight largest producers are shown in rectangular boxes.
Locations and number of plants by state for 1967 can be found later in Figure 35, a histogram
used for making projections in the following section.
II. PAST, PRESENT AND PROJECTED INDUSTRY TRENDS TO 1985
Several references, mostly published by Bureau of the Census, have been employed in
computing past and present production and in projecting future levels. Three main sources used are:
1. Current Industrial Reports, Series M28F8
2. Census of Manufacturers, 196714
3. County Business Patterns™
Various Industrial Reports were used to tally annual productions for 1965 to 1972 (Figures for 1972
are based on estimates for December 1972) in dollar value and gallon volume. Geographical
distribution on a state-by-state basis is based on the 1967 Census — the most recent available
at time of writing. Distribution of plants by employee size class are based on County Business
Patterns.
A. Production
Average annual growth of production rates were computed for the seven year base period
and applied to years in the 1972 to 1985 period. Projections for these years are plotted in Figure
33 as million gallons, and in Figure 34 as million dollars of shipments per year. Computed average
growth rates are indicated in terms of percent of current production per year for Trade Sale Finishes,
Industrial Finishes and total sales in each figure. Results are summarized on the following page.
156
-------
Trade Sales Industrial Total
Year
1965
1972
1985
\vg. Annual
Million
Gallons
411.0
456.0
550
1.50
Million
Dollars
1247.0
1663.7
2845
4.21
Million
Gallons
365.0
474.4
770
3.82
Million
Dollars
922.3
1355.6
2775
5.66
Million
Gallons
776.0
930.4
1320
2.63
Million
Dollars
2169.3
3019.3
5620
4.84
Rate %
The average growth rate was calculated in a manner similar to compound interest. For
example, let A be the production volume for the first year and B the volume for the nth year. The
average annual growth rate, a, is determined by solving the following equation for a:
A (1 + a)n = B
For total gallon production, this becomes
776.0 (1 + a)7 = 930.4
and a = 0.0263. The projected growth curve will plot as a straight line on semi-log coordinates.
The line, by definition, will pass through the first and last points.
It is apparent from these figures that in 1972 Trade Sale Finishes accounted for 49% of
the volume and 55% of the dollar value. In terms of gallons, this compares with 53% in 1965 pro-
jected to shrink to 41.5% in two decades. Over the same time period, however, revenues from
Trade Sales are expected to register a much smaller drop: from 57.5% of the total in 1965 to 51% in
1985. This comparison is readily apparent in the plots of Figures 33 and 34. The higher growth
rates (in terms of both volume and value) for Industrial finishes are attributed to the trend for pre-
coated materials such as siding for houses, paneling and certain plastics. Industrial finishes display
higher growth rates than Trade Sales. Their share of the market increases even more dramatically
when measured in gallons than in dollar value. This is attributable in part to the lower cost of
packaging and transportation required for industrial products.
Total production figures for 1967 as reported in the Census of Manufacturers differ from
those reported in the Current Industrial Reports by about 25% (variations between these sources
are somewhat smaller for the census years 1963 and 1958). The Current Industrial Reports are
published monthly and present a source of up-to-date information. They are estimates from a
sampling of about 310 plants and do not represent the accuracy that might be expected from the
larger sampling used in the Census of Manufacturers reports. Unfortunately, these reports are
prepared only about every four years.
157
-------
I
J
0
•0
2 2 • 2
•
f 7
0 0 0 0 0 jj
Q "« N - ni n
158
-------
1,000
1
j 900
J
3 800
7
I
J 700
• 600
)
>
500
r
J
\
F
j 400
300
250
U.S. PAINT \V <^1
^ + * TOTAl
^,XX
^^ INDUSTF
^ 1 SALE
^^ C
-.*»-*"*" TR>
-•^ SAL
. RATE I.S50/:
EPORTS
_ SALES
?IAL
S
•E 3.02*%,
>vDE
.ES
>
159
-------
10,000
9,000
•,000
U.S. PAINT & VARNISH INDUSTRY
SHIPMENTS (ACTUAL. &. FORECAST)
MILLION DOLLARS
SOURCE: \9OSt-7Z CURRENT INDUSTRIAL. REPORTS
7,004
A.R.I. ESTIMATES
FIC3URE
„••
.«•
1.000
(0
£
<
J
d
0
I
J
!
Irt
2
U 3,000
5
*)
TC TAL
S/< L.ES
AV<3. RATE
ft
U
J
> 2,000
TRADE
SALES
4^ xC.
1,300
AVC3. R*»TE -4-.2I -X.
INDUSTRIAt.
AVC3. RATE
160
-------
Statistics presented by the Census are, nevertheless, the most detailed available and Figure
35 is based on them. This histogram provides a state-by-state breakdown, for 1967, of value of
shipments and number of establishments and employees. States with less than $6 million in
shipment value are not indicated. Projected values for 1972 and 1985 are also shown for each
state; these values are based on growth rates computed from the Current Industrial Reports.
It can be assumed that, despite the difference of total production as stated by the two
sources, growth rates derived from one can be applied to geographical distribution obtained from
the other. The second simplification assumed in estimating 1972 and 1985 productions was that
each of the 30 states would increase at the same industry-wide rate. This accounts for equal
increments on the logarithmic scale used in Figure 35.
Whereas projections based on average annual growth rates over preceding years do not
directly take into account several variables that could affect production (such as projected growth
rates of all industry, GNP, population, per capita consumption, technological breakthroughs, etc.),
it is expected that a similar number and nature of variables existed and determined output over the
base period. Consequently, projections based on increasingly complex and seemingly sophisticated
indices may be no better than those obtained above. Factors used here and elsewhere represent,
at best, an approximation of future trends.
B. Number of Plants
The number and size of the plants operating is somewhat vague, with the Census of
Manufacturers reports perhaps being the best source of information available. According to the
Census reports, the number of plants and companies operating them are shown below:
Year
1958
1963
1967
No. of
Plants
1,709
1,788
1,701
No. of
Companies
N/A
1,579
1,459
Plants w/20 or
More Employees
600
654
680
The number of establishments (plants) is only an approximation because those plants
with less than ten employees were not required to complete the census report in 1967, resulting
in estimates being calculated and incorporated into the figures. The count on those plants with 20
or more employees is far more reliable.
In terms of geographical location, Figure 35 plots the number of plants for each of the top
30 paint-producing states. There is, apparently, very little direct correlation between paint produced
161
-------
162
-------
and number of plants in a given state indicating emphasis, in some areas, on smaller plants.
C. Size of Plants
As indicated in Figure 31, the median plant, when ranked by number of employees, in
1971 had 16 employees whereas the median employee, ranked by the size of the plant in which
he worked, was employed in a plant having 100 employees. This merely implies that 50% of the
plants employed more than 16 people, but 50% of industry work force were working in plants
having 100 or more workers. The plot provides a quick and easy method for finding the percentage
of establishments in any selected size range.
The only plant size projections attempted are in terms of number of employees. As previously
discussed in Part I of this chapter, the number of plants in each size category (1 to 3 employees,
4 to 7 employees, etc.) was expressed as a percentage of total plants. These figures were obtained
for the years 1965 to 1970 from the County Business Patterns of corresponding years. Since
successive years shows small changes, the total change was measured in terms of the statistical
distribution parameters: geometric standard deviation, mean, median and mode. As illustrated
earlier in Figure 31, a log-probability plot of the distribution greatly facilitates computation of these
parameters from which a "bell curve" can be derived, if desired.
Figure 36 plots percentages of plants smaller than indicated employment size — actual
distributions for 1965 and 1970 and projected distributions (broken lines) in five-year increments
to 1985. Actual distributions are completely defined by:21
1. Geometric mean, or median, which is plant size class at 50%, and
2. Geometric standard deviation which is equal to
plant size class at 84.1%
plant size class at 50%
There was found to be a negligible change in the latter quantity over the five-year base period.
Had there been a noticeable difference this would have resulted in different "slopes" for 1965 and
1970, and the projected distributions would need to take this into account. Instead, since the plots
are "parallel," projections can be based on a change of median size alone.
The increase in median plant size appears minor (16 employees in 1965 to 23 employees
twenty years later). At the 98% point, however, Figure 36 indicates that the number of employees
will have increased from 230 to almost 400. Conversely, the largest 2% of the plants will have
increased in size to employing more than 400. This increase in size is in spite of technological
improvements and increased productivity. Table 43 lists percentages of plants in indicated employee
163
-------
300
200
ER OF EMPLOYEES
fc If m *J • 10 1
> 5 o o o 5 <
NUMBI
5 *
J
N
PLAN
, ~j • • 5 !
5
4
i
U
S
S. F
DIS
OURC
=>AI
>TR
E: is
P
)S
p
/
N1
IB)
-Nl
(>
»es
AT-
»7&
RO
/
/,
/M/
' &
JTIC
JMB
A.CTU
, 197
TERN
, 198
JECT
FIG
//
'//}
'///
//
/
\#
3N <
ER
IAL
0 -C
S.
0 , 1
IONI«
URI
'A
yb
xRr
OF"
OF
& F
our-
98S
5.
E o
t
f
MIS
Pl_,
EN/
»RO.
JTY
- Af
56
4
/ ^
^
/a
7
H
A,KI
IPLjC
JECT
BU
^.1.
x
//
r
INO
"S B
3YEE
ED)
SINE5
/;
'^/y
/
*> — 11
^ 1!
"^ 11
US
pf «
s-
>s
^
f J
Vf
^/
/
»es
>7C
>75
>8C
»e*
rrp
SIZE
/
'V
'//<
'//
tY
>
//
'
/f
$/
^
> 5 10 T5"~20 30 40 50 60 70 ao »5 M »5 •
PERCENTAGE PLANTS LESS THAN INDICATED SIZE
164
-------
TABLE 43
U.S. PAINT AND VARNISH INDUSTRY
DISTRIBUTION OF PLANTS BY EMPLOYEE SIZE CLASS
Percent of Establishments in Indicated Employee Size Class
Year 1971 1970 1969 1968 1967 1966 1965
Employee size
Class
1 to 3
4 to 7
8 to 19
20 to 49
50 to 99
100 to 249
250 to 499
500+
Total units reporting
Total employees
15.2
15.4
26.3
22.0
11.3
7.3
2.1
0.4
1,568
63,865
15.2
15.5
25.1
23.5
11.1
7.1
2.3
0.3
1,597
65,601
15.4
15.8
25.1
23.0
11.1
6.6
2.5
0.5
1,607
66,474
16.3
16.5
25.3
22.2
11.0
6.3
2.0
0.5
1,624
63,623
17.8
15.5
25.7
21.6
11.1
5.9
1.9
0.6
1,654
64,959
17.8
17.4
25.9
20.7
10.2
5.6
2.0
0.4
1,690
62,359
17.9
18.0
25.4
21.3
9.5
5.7
1.7
0.5
1,734
63,255
Source: "County Business Patterns", U.S. Summaries for 1971, 1970, 1969, 1968, 1967, 1966
and 1965
165
-------
size classes. It is noteworthy that the total number of plants has registered a decline for every
year from 1965 through 1971.
This increase in average size is not surprising since the paint industry is labor intensive.
It is not feasible to operate on a continuous production basis as in chemical manufacturing. Larger
equipment can be used to increase batch size and productivity, but the wide range of products
and colors still precludes the use of any other technique except batch processing.
In Figure 35 a comparison between the bar indicating number of plants and the line giving
number of employees will yield an average plant size for each state.
D. Capacity — Production Relations
There are no statistics available on theoretical capacity for the industry. One can "pick"
a reasonable number, such as 80% or 90% and assume that all plants, on an average, operate
their equipment at this fraction of full capacity. If necessary, one could calculate industry capacity
by dividing production data previously presented by 0.80 to 0.90.
E. Typical Plant and Equipment Ages
It is suspected that a vast majority of the some 1,700 establishments will have quite old
structures and equipment. There are several reasons for arriving at this conclusion:
1. To avoid high shipping cost, most of the plants are established in geographical sections
of the country where the paint demand is high (see Figure 32).
2. Most have been established for quite a number of years, having evolved from single
ownership to corporate structure in these geographical areas where land values and
construction cost have increased at a very rapid pace.
3. Paint manufacturing does not cause extensive damage to buildings, nor does it have
a deteriorating effect on a majority of the equipment used.
4. Inspection of capital expenditures (see Figure 37 and the accompanying discussion on
capital expenditure, Section III) shows us that for the period of 1963 to 1970 an average
of 35% has been expended for buildings and structures, while 65% has gone for new
machinery and equipment.
Within the Paint Industry the following types of equipment will be found:
1. Storage tanks
2. Reaction kettles
3. Pumps and motors
4. Filters and strainers
166
-------
1.000
U.S. RAINT &. VARNISH INDUSTRY
CAPITAL EXPENDITURES
MILLION DOLLARS
SOURCE: STATISTICAL ABSTRACTS
-1- OF THE. U.S. 1903 - I97O
500
FIGURE 37
^ /-
s /
DOLLARS
LLI
(/)
y
=jioo
TOTAL
Viy-G. RATE 15.30*^.
\ NEW MACHINERY
ZJ *•
EQUIPMENT
AVG. RATE
I
\
X
y
^
D.'
<
0
70
AVG
STRUCTURES
RATE: \&^*>°/~
20
10
167
-------
5. Filling and capping equipment
6. Packaging equipment
7. Mixing and dispersing machinery
8. Grinding machinery
9. Electronic instruments of various types
10. Material handling equipment.
Much of this type of equipment is common to other industries and requires little change
to adapt to paint manufacturing.
In general, we would expect to find quite old structures, many of which may be 50 years old
or older. However, the equipment should be relatively newer due to changes in manufacturing and
increasing batch size. The average age of equipment may run in the range of ten to twenty years.
F. Technological Revolutions and Outside Influences Causing Changes in the Industry
The paint and varnish industry as a whole is not a large investor in research and development.
Less than 10% of the research and development dollar goes for basic research. The vast majority
goes for the support of existing products, solution of manufacturing difficulties, and customer
assistance. Certain selected areas of research are being intensively studied, however. These are
discussed below.
Many of the smaller firms cannot afford sizable research and development expenditures
and depend upon raw material suppliers and group research firms who pool their efforts for the
member companies. The most prominent group research firms are the Paint Research Association
in Chicago and the Coatings Research Group in Cleveland.
1. Application Techniques — Perhaps the most dynamic changes in the industry are being
brought about by advances in the state-of-the-art; advances which could well be termed techno-
logical breakthroughs, in retrospect, if current developmental work is successful and results in wide
acceptance. Such advances include:
1. Powder coatings
2. Electrostatic spray
3. Electrodeposition
4. Radiation curing
Powdered coatings represent one significant advancement, in recent years, for the industry.
In addition to its several technical advantages this method eliminates the use of solvents, thereby
reducing atmospheric pollution as well as fire hazards at points of production and application.
168
-------
Although some powders present a dust explosion hazard, this problem can be taken into account
during design of the process.
It is expected that the industry will concentrate much attention on development and use of
powder coatings. Powder coatings can be applied with flocking guns, electrostatic spray equipment,
or electrostatic fluidizing chambers. The best known method uses a preheated object immersed
into a fluidized chamber where the powder is kept in motion by ascending gas flow. Powder
coming in contact with the preheated surface melts and fuses to the surface. The coating is cured
in an oven. A patent was recently issued to Grow Chemical Company covering the production of
powder coatings in water slurry form. It is said that these can be applied with existing solvent based
equipment.22
Much recent interest in this field has been expressed by automobile manufacturers who
claim that recycling permits material utilization of 98%.23 Ford has been powder coating truck wheels
for a few months and is expected to announce in late 1973 a pilot line for applying topcoats. General
Motors has a one-color automatic spray and recovery unit for topcoats. This unit employs eight guns.
Estimates for powder coating consumption range as high as 200 million pounds/year for
metal coating by 1980. In addition, the powder coating of glass containers is said by some to
offer an equal potential consumption by 1979,24 for a total of 400 million pounds per year by 1980.
These estimates are admittedly somewhat optimistic but more stringent air pollution regulations
could tip the scales in their direction.
Electrostatic forces have been employed in spraying methods. Basically, a charged (paint)
particle will follow an electric field. If an electrical potential difference exists between a particle and
a surface, the particle will tend to flow to the surface and be deposited there. Various available
methods differ only in the method whereby the particles are charged and brought to the proximity
of a surface. Any material loss occurs only for those particles whose momentum in a direction
away from the piece to be coated overcome the electrostatic force of attraction. A reduction of
spraying velocity will, therefore, minimize losses.
Electrodeposition differs from the above in that electric forces are used to deposit liquid
paint particles much like in an electroplating operation. The coating material contains resins, additives
and pigments. Complete and uniform coverage of all exposed surfaces can be obtained. Film thick-
ness is uniform since the deposition ceases when the film is thick enough to act as an insulator
at the applied voltage. Materials used are generally water-based thus reducing solvents.
Radiation curing utilizes free radicals and ions to initiate polymerization or other reactions
169
-------
required for a complete cure. A thin beam of electrons is generated by applying 300 kilo-volts to
a high vacuum accelerator tube; this beam can be used for curing with very little heat build-up
making it particularly desirable for coating wood, rubber and fabric materials. Although this method
is suitable for continuous curing its use has been curtailed by high costs and technical problems.
Ultraviolet and high intensity visible radiation have also been utilized for curing.
2. Pigment Industry — Another important area where technology is expected to set the trend
is in the pigment industry, in the production of titanium dioxide. Estimates25 indicate that the
production by the chloride process will surpass that using the sulfate process by 1975. Currently
the chloride process accounts for about 47% of the production, its share is expected to grow to
60% while total production advances by 18%. Listed below are estimates for totals of seven leading
producers of the pigment:
MM Ibs
Process Current 1975 1973*
Sulfate 918 818 737
Chloride 820 1,230 837
Total 1,738 2,048 1,574
The sulfate process uses a lower grade, iron containing, ore known as ilmenite or high
TiO2 slag. The ore is dissolved in concentrated sulfuric acid to form sulfates of both iron and titanium,
and titanium dioxide is precipitated as a hydrate. The chloride process, on the other hand, starts
with Rutile, a high grade ore or upgraded ilmenite. Technology for using ilmenite directly has
recently been developed by duPont.26 Also, anatase production by the chloride process is presently
under investigation. The sulfate process results in a significant potential for emissions of sulfur
oxides. The chloride process is inherently cleaner. It not only has a lower potential contribution to
air pollution but also has fewer liquid wastes.
Foreign trade is generally not expected to be a major factor in the coatings industry. However,
imports have been increasing recently for titanium dioxide with a significant increase in 1972.
Aided by the general economic upturn in industry that year, demand so far outstripped supply of
TiC-2 that suppliers reduced inventories by more than 50% of the previous year. Imports of this
vital pigment almost doubled, with West Germany and Canada providing most of the difference
(75% of the increase over 1971 and 48% of the total imports in 1972).
While demand grew, domestic production remained essentially static. Two leading suppliers
*E. I. duPont estimate
170
-------
of TiO2 have either plants or an interest in plants in West Germany and Canada.
The same companies have also shut down some plants in the U.S. Analysts feel that the
surge in imports was caused by their efforts to satisfy market demand with imported material.
Production capacity is expected to head upward again over the 1972 to 1975 period with
several suppliers planning expansions and new plants.26
DuPont is planning chloride process expansions at Edge Moor, Delaware, and New John-
sonville, Tennessee, which combined with phasing out of sulfate capacity at Edge Moor will result
in a net increase of 188,000 tons/year. New Jersey Zinc and Kerr-McGee are planning additional
chloride process capacity, also. No new sulfate process expansions have been announced.
3. Environmental & Health Considerations — A significant influence on the industry in recent
years has come from the laws and regulations at every level of government. In general, these
regulations have been concerned with the environmental effect of heavy metals, such as
lead and mercury, the pollution of water and waterways, and the pollution of the atmosphere with
organic emissions. Some others have been concerned with the safe use and shipment of paints.
In January, 1971, the "Lead-based Paint Poisoning Prevention Act" was enacted. This law
prohibited the use of paints containing more than 1% lead by weight in the non-volatile portion of
liquid paints or in the dried film on all surfaces accessible to children in residential structures
constructed or rehabilitated by the Federal Government or with Federal assistance. The surfaces
include all household interiors and such exteriors as stairs, porches, windows, and doors.
In March, 1972, the Food and Drug Administration ordered a reduction of the lead content
in paints used in and around households to a maximum of 0.5% by January 1, 1973, and possibly
to 0.06%. Household surfaces were extended to include such manufactured products as toys.
The proposed enforcement of a 0.06% lead content level by December 31, 1974, will not
only require the use of higher priced substitutes but also time consuming and cbstly reformulation of
many paint products. This lower limitation may not be enforced if the Consumer Products Safety
Commission determines a higher level to a maximum of 0.5% is safe.*
Mercury is also toxic to humans at some level of exposure. Organic compounds of mercury
are used in water-based paints as preservatives to prevent bacterial action. Mercury fungicides
are used in many exterior paints to prevent fungi from attacking the dried film.
The curtailment by the EPA of mercury usage has again required a search for substitutes
and reformulation.
'Public Law 93-151
171
-------
Air pollution legislation has also had its effect on the Paint Industry, particularly those
regulations patterned after "Rule 66" which seeks to control the emissions of photochemically
reactive hydrocarbons which react with nitric oxide in the presence of ultraviolet radiation to form
oxidants.
The Water Pollution Control Act as amended in 1965 and 1972 has, and will cause many
paint manufacturers to design alternate ways of discharging waters. The industry contributes to
water pollution primarily through the discharge of slurries.
III. DISTRIBUTION OF CAPITAL EXPENDITURES
The Paint and Varnish Industry has shown a fairly steady increase in capital expenditures
over the 1963 to 1970 period. Figure 37 is a plot of expenditures for new buildings and structures
and for new machinery and equipment. Totals are also plotted for the seven-year period. Although
this industry does not spend as large a percentage of its revenues in this category as does the
entire Chemical Industry, it has maintained a steady input of funds in response to growing demands.
Since 1963, approximately 65% of all capital expenditures have been for new machinery and
equipment with the remainder going for new buildings and structures.
Average annual growth rates for capital spending were computed for the base period, for
each category, and for the total. These rates were used in projecting to 1985. The rate for total
spending averages 15.3% — a rate that seems unrealistically high, especially when compared to
the rate of increase in value of shipments. This is further emphasized by selecting the six-year
base period 1963 to 1969 which reduces the growth rate substantially resulting in capital expendi-
tures less than 50% of those obtained by using the 1963 to 1970 period (to illustrate, the reader
may draw an imaginary line through the corresponding points for 1963 and 1969 and extend this
line to 1985). This apparent discrepancy seems unavoidable where data is subject to very large
year to year fluctuations as in the case here.
While Figure 37 attempts to provide a long range picture of the capital spending pattern,
Table 44 provides a comprehensive distribution of labor and finance within the industry. Statistics
here include a listing for Inorganic Pigments.
172
-------
TABLE 44
U.S. PAINT, VARNISH AND INORGANIC PIGMENTS INDUSTRY
1967 LABOR AND FINANCE SUMMARY
Establishments
Establishments with 1 to 19 Employees
Establishments with 20 to 99 Employees
Establishments with > 100 Employees
Employees
Payroll
Production Workers
Wages
Man Hours
% Man Hours
% Man Hours
% Man Hours
% Man Hours
Cost of Material
Materials, Containers, etc.. Consumed
Cost of Resales
Fuels Consumed
Purchased Electricity
Contract Work
Value of Shipments
Value of Resales
Value Added by Manufacture
Manufacturers Inventories
Total
Beginning 1967
Finished Products
Work in Process
Material, Suppliers, Fuel, etc.
Total
Millions of Dollars
Millions of Dollars
Millions of Hours
January to March
April to June
July to September
October to December
Millions of Dollars
Total
(Including Resales)
Millions of Dollars
Total
Paint &
Allied
Products
1,701
1,701
1,021
521
159
66,100
492
36,300
223.4
73.1
24.2
25.45
25.5
24.9
1 ,606.4
1,461.0
125.9
7.8
9.8
1.9
Inorganic
Pigments
98
98
38
33
27
12,600
97.2
8,900
63.2
17.7
26.0
25.4
23.7
24.9
235.0
197.3
15.2
15.4
6.3
0.9
2,911.4
173.2
1,318.5
406.1
224.8
23.9
157.4
549.3
20.1
316.3
104.3
45.0
9.6
49.7
173
-------
TABLE 44 (continued)
U.S. PAINT, VARNISH AND INORGANIC PIGMENTS INDUSTRY
1967 LABOR AND FINANCE SUMMARY
Paint &
Allied Inorganic
Products Pigments
End of 1967 Total 426.7 108.6
Finished Products 234.6
Work in Process 27.7 9.6
Material, Supplies, Fuel, etc. 164.4 51.9
Expenditures for Plant and Equipment 73.5 21.7
New Plant and Equipment Total 70.7 20.8
New Structures and Additions to Plants 28.8 3.1
New Machinery and Equipment 41.9 17.7
Used Plant and Equipment 2.8 0.9
Source: 1967 Census of Manufacturers, Bureau of the Census
174
-------
CHAPTER 3
MEASUREMENTS OF EMISSIONS
I. SAMPLING AND ANALYTICAL PROCEDURES
A. General Requirements for Source Testing
Several comments can be made concerning source testing for the Paint and Varnish
Industry.
1. The testing personnel should become as familiar as possible with the process under
investigation before testing begins. Products produced, raw materials used, conditions of operation,
location of sampling points, etc., should be ascertained as much as possible.
2. The chemical identity of the probable emissions species should be anticipated in
advance and their chemical and physical properties determined. A brief description of the species
that may be encountered in various types of operations follows.
Some of the emissions characteristics of resin and varnish cooks are covered in Chapter 1.
The major part of the emissions from these operations consists of solvents, steam, and some of
the more volatile reactants. Also present will be various reaction and degradation products that
can be formed during processing such as aldehydes and organic acid. The latter products are
usually present in relatively small amounts. However, they often constitute the most noxious
components of the fumes. Table 45 lists the more common raw materials and solvents used in
the manufacture of various resins. This list is not intended to be comprehensive, but should be
used only as a guide as to what to expect when source testing a particular resin operation.
Since thinning is sometimes done in the resin cooker, solvents can represent a major
emission even from fusion processing. The properties of the solvents used should be kept in mind
when designing an analysis train for source testing. The chemical composition and vapor pressure
characteristics are of particular importance. The boiling range of selected solvents is given in
Table 46. A boiling range is given even for specific compounds since the grades usually used in
the paint industry are of varying degrees of purity.
175
-------
Resin
Alkyd
Urethane
Acrylic
(solution)
Phenolic
Vinyl
(solution)
Amino
TABLE 45
COMMON RAW MATERIALS AND SOLVENTS
USED IN THE MANUFACTURE OF RESINS
Raw Materials Solvents
Polybasic Acids
Phthalic Anhydride
Maleic Anhydride
Fumaric Acid
Polyols
Glycerol
Pentaerythritol
Oils
Soya
Tall Oil Fatty Acids
Benzoic Acid
Toluene Diisocyanate
Polypropylene Glycol
Linseed Oil
Acrylic Monomers
Methacyrlic Monomers
Formaldehyde
Phenols
Cresol
Vinyl Monomers
Dimethylol Urea
Urea
Formaldehyde
Butanol
Melamine
Xylene
Mineral Spirits
Toluene
Naphtha
Xylene
Mineral Spirits
MEK
Ethyl Acetate
Aromatic Hydrocarbons
Ketones
Esters
Alcohols
Ketones
Esters
Ketones
Aromatic Hydrocarbons
Butanol
Xylene
176
-------
TABLE 46
SOLVENT BOILING RANGES
Solvent
Hexanes
Naphtha (light)
Xylene
Toluene
VMP Naphtha
Regular Mineral Spirits
Aromatics (medium)
Ethyl Alcohol
n-Butyl Alcohol
Acetone
Ethyl Acetate
Methyl Ethyl Ketone
Ethylene Glycol Monoethyl Ether
Boiling Range, °F
140 to 160
205 to 250
275 to 290
230 to 232
210 to 300
31010395
315to390
170 to 174
241 to 246
133 to 135
162 to 176
172 to 178
270 to 279
Source: Martens4
177
-------
In addition to the raw materials and solvents, the source tester should be aware of the
possible side reactions that may occur in processing. Certain of these reaction products have a
very low odor threshold and their presence in even very low concentrations may be undesirable.
For example, polyols can undergo oxidation or dehydrogenation to form aldehydes and ketones.
These can undergo further oxidation to acids. The breaking of the carbon chain can occur on
oxidation to give an aldehyde and ketone each of which has fewer carbon atoms than the original
alcohol. Some of these reactions may occur in the kettle while others may require more severe
conditions such as would be encountered in a fume incinerator.
3. The sampling team should have the capability to perform continuous, on-line measure-
ments as well as discrete sampling. The highly cyclical emission curves of such operations as
resin cooking makes continuous measurements virtually mandatory.
4. Cross checking of the measurements should be made whenever possible. This should
preferably be through on-site analysis to minimize the necessity of return trips. Wherever feasible,
an over-all material balance on the sample should be used to check the consistency of the analytical
results.
5. Equipment used should be as simple and as portable as possible and require a minimum
of utility hook-ups.
6. Explosion proof equipment is required when in close proximity to volatile hydrocarbons.
7. A complete log should be made, if possible, correlating emissions levels with kettle
temperature, addition of raw materials and solvents, etc.
B. Description of Source Sampling and Analytical Procedures
1. Flow Measurement — The most widely accepted technique presently consists of flow measure-
ment per ASTM/PTC-27 using a reverse pilot tube (S type). Care must be exercised to see that
a representative traverse is made and that the measurement points are sufficiently far from any
flow disturbances. Duct temperature and pressure should also be recorded.
Pilot tubes, however, cannot be used with much accuracy below linear gas velocities of
10 to 15 feet/second. For lower gas velocities, various types of anemometers are available including
vane, hot-wire, swinging vane and heated thermocouple types. These instruments often have
limitations as to the types of environments in which they can be used so care should be exercised
in selecting them. Hot-wire types may present an explosion hazard in some atmospheres.
Resin kettle vents are sometimes quite small in diameter (four inches or less). For small
178
-------
diameter vents, the use of a pitot tube can be undesirable in that its very presence in the vent may
disturb the flow pattern so as to give a faulty reading. This problem can be overcome in some
cases by the use of very small L shaped pitot tubes. These, however, are more prone to clog by,
for instance, phthalic anhydride.
An alternate method of measuring low flows in small ducts is through the use of a calibrated
orifice. These have been used with some degree of success in the field. A low pressure drop,
pre-calibrated orifice is mounted in an assembly which can then be fitted to the end of a reactor
vent. All of the gas flow is made to pass through the orifice.
A variation on this scheme, which has met with less success, is the use of a plastic bag
provided with a hole. The bag is slipped over the end of a vent and the pressure in the bag during
gas flow is measured and correlated with gas flow.
For very low flow situations, the use of a bag without an orifice can be used. A completely
evacuated bag can be slipped over the end of the vent and allowed to fill with the effluent gas. In
this* case, the bag collects all of the gas emitted. The bag can then be sealed and the contents
pumped out through a meter. The total quantity of gas is then recorded.
A further complication, particularly in resin cooking, is flow rate fluctuation. For instance,
when the kettle is opened for raw materials addition, a vacuum is drawn in the kettle causing large
volumes of air to pass through the open port, into the kettle, and through the vent. Likewise sparging
rate may vary during a cook. If a measure of the total quantity of emissions is to be obtained, flow
rate as well as emissions concentration must be known at all times during the process cycle.
For some ducts, such as those associated with baghouses, a pitot tube is suitable. For any
device which measures linear gas velocity, the flow profile across the duct should be known at all
times. This is often impractical, however, in many process situations. A useful compromise might
be to take a traverse during a period of constant flow and monitor the centerline velocity continuously
during the rest of the run. If the velocity profile is well shaped, and if the Reynolds' number does
not pass through the transistion region between laminar and turbulent flow during velocity fluctu-
ations, then the centerline velocity can be related to the average velocity with reasonable accuracy.
In extreme cases, it may be necessary to continuously monitor the velocity at more than one point.
2. Particulate Measurements
a. General Considerations — Here again, considerable variations will be encountered in the
amount and nature of paniculate emissions. Some of the situations which may be encountered
are listed below.
179
-------
1. Solid, inorganic participate — An example of this type is pigment participate. These
are essentially non-volatile and have little, or no, solubility in pure water. This type is the easiest
to handle from a collection standpoint.
2. Liquid organic aerosol — This represents a more difficult situation. The aerosol will
usually be present in a gas stream which is saturated in the vapor phase with the same hydrocarbons
present in the liquid droplets. The amount and composition of the aerosol may vary with location.
When the gas stream leaves the process unit it is at high temperature and only small amounts of
entrained liquid should be present; whereas downstream where the temperature is lower, significantly
more may have formed.
The physical properties of some aerosols can significantly complicate source sampling
techniques. Where very high boiling components are involved, if the sampling is done by one of the
common methods of drying and weighing the collected material, the aerosol will be included in the
total particulate measurement. Where a more volatile aerosol has been collected, it may evaporate
before weighing and be lost. The intermediate situation is more troublesome in that part may be lost
and part retained.
3. Solid organic particulate — A good example of this type is phthalic particulate. Some of
the same comments apply here that were stated in the preceding case. The organic particulate
may be volatile under conditions of subsequent sample processing.
4. Organic vapor — This can be a relatively easy situation to handle providing that the
vapor is kept above its dew point in the sampling equipment. Problems can arise due to the large
number of chemical species that may be present in a given sample.
b. Collection and Analysis Techniques — The usual procedure for particulate sampling is to sample
a portion of the effluent over a period of time using some kind of sampling train. Instantaneous, or
"grab" samples, are not well suited to particulate collection. The sampling train consists of a sampling
probe, collection device or devices, gas flow meter and a gas pump for drawing the sample.
Sampling of streams containing particulate should always be done "isokinetically" whenever
possible. That is, the gas velocity in the probe tip should be the same as that in the bulk gas
stream in the vicinity of the sample point. In the case of very small particles, say 5 microns or less,
some deviation from isokinetic sampling conditions is acceptable.
Sampling under unsteady state conditions presents special problems. A given stream may
be in an unsteady state with respect to flow rate, particulate loading, or both. If the flow rate fluctuates,
gas velocity in the vicinity of the probe tip must be monitored and corrections made to the sample
180
-------
flow rate as necessary to maintain isokinetic conditions. If participate loading changes, then samples
must be taken at several times in the process cycle. Unsteady state conditions can be expected
to occur often in the Paint and Varnish Industry.
A particulate sampling train in use by the EPA is described in the Federal Register (Vol.
36, No. 247, Dec. 23, 1971). A sketch of the sampling train is presented in Figure 38. The probe
and filter holder are heated to prevent unwanted condensation. The first two impingers are filled
with water, while the third is left empty and the fourth is filled with pre-weighed silica gel. Details
of the operation of this train may be found in the Federal Register referenced above.
The Los Angeles Air Pollution Control District Source Test Manual describes an apparatus
similar to the EPA train. The major exceptions are that a small cyclone is placed before the filter
and three impingers are used instead of four.
A third alternative is to place the filter holder assembly into the stack itself. In this way,
condensation on the filter is avoided since the filter temperature will tend to approach that of the
gas stream. This technique also reduces the amount of particulate that could settle on the surfaces
of tubing between the probe tip and the filter. This can be particularly important in the case of
sticky particulate material.
All of the above methods work well on dry inorganic particulate such as pigment dust.
When sticky or tarry material is measured, all methods are subject to problems of clogging in the
probe and filter, and deposition of material in the lines. In such cases, it is sometimes helpful to pass
the gas sample into the impingers first with the filter placed downstream (but before the silica gel).
with a volatile solvent, evaporating to dryness and weighing the residue. The disadvantage here
weighing the residue. Material deposited in lines can sometimes be collected by washing the lines
with a volatile solvent, evaporating to dryness, and weighing the residue. The disadvantage here
is that it is not possible to distinguish between aerosol at stack conditions and other condensible
materials.
3. Hydrocarbon Analysis
a. Discontinuous Sampling — Included under this classification are any samples drawn over an
interval of time comprising only a portion of the process cycle and analyzed on a batch basis to
give an average concentration for that time interval. For instance, "grab" samples taken instantane-
ously as well as semicontinuous samples drawn at constant flow rate for a specified period of
time (on the order of 20 to 30 minutes) are considered in this section.
181
-------
id
(T
i
L
0
3
182
-------
1. Los Angeles County APCD Methods — The Los Angeles County APCD manual describes
several methods of total hydrocarbon analysis. Each method utilizes the same sampling technique
which consists of "grab" sampling in evacuated containers. If fitted with an orifice or other control
device, the evacuated container can be used to draw semicontinuous samples. Two* liter round
bottomed flasks are the containers normally used. Protective wrapping should be applied to minimize
breakage where glass containers are used.
Prior to evacuation, the container should be thoroughly cleaned and dried. The sampling
line should be equipped with a filter to exclude paniculate and provided with a means of flushing
with stack gas prior to sampling.
The advantage of this technique is that samples can be taken very quickly. In principal, it
is possible to draw a sufficient number of samples over a period of time to catch fairly rapid fluctu-
ations in concentration. For a typical resin cook, however, the number of samples required may be
unwieldy.
Potential problems with this approach include clogging of the inlet components, deposition
by condensation or adsorption on the walls of the container, and the inability to obtain results in
the field. These difficulties make this method unsuitable for testing the effluent from alkyd type
cooks and, in fact, the LAAPCD does not recommend it for this type of process. It may, nevertheless,
be useful for processes which do not produce high boiling organics, particulates, or aerosols.
Various approaches can be used to minimize, at least partially, some of the objections.
Dilution of the sample can eliminate dew point problems in some cases. Adsorption can be minimized
by using teflon bags. Heated syringes are available for "grab" sampling. Used with sample dilution
techniques these can provide good accuracy and reproducability. Analysis of the sample can be by
infrared, combustion analysis, flame ionization techniques, chromatographic, or combinations of
these.
2. Semicontinuous Sampling Trains — The Graphic Arts Technical Foundation has devised
a method for sampling hydrocarbon emissions in the printing industry. The sample is drawn by
an evacuated cylinder through a cold trap immersed in dry ice. This separates the hydrocarbons
into "condensible" and "noncondensible" fractions which are then analyzed separately. The
noncondensibles in the cylinder are analyzed using gas chromatographic techniques on a Poropak
Q column. The contents of the cold trap are analyzed by heating the trap at a controlled rate and
analyzing the vapors with a flame ionization detector. If the sample line is heated up to the cold
trap and fitted with a filter, this technique should work reasonably well. Its chief disadvantages are
'Currently using 8 liters
183
-------
the time required for analysis and the inability to analyze the samples in the field.
Another variation involves collection of a "condensible" fraction in ice water impingers
accompanied by continuous, on site measure of noncondensibles using a portable flame ionization
detector. The condensibles are measured in the laboratory by evaporating the water and weighing
the residue. This is subject to error in some situations due to loss of some of the samples during
drying. An alternative to evaporating the water is solvent extraction of the organics in the impingers
followed by fractional distillation to remove the solvent. The condensible organics can be further
characterized in the laboratory by chromatographic or other means.
Deposition of phthalic or heavy organics in the lines upstream from the impingers can be a
problem with this technique. Also, the use of a FID in the field can be difficult and cumbersome.
4. Analytical Techniques
a. Infrared Analyzers — Infrared analyzers are well suited to continuous measurements. Normally,
they are set up to monitor a single component. This is the source of their chief disadvantage. One
must know in advance what components will be present and the instrument modified to detect a
specific chemical. The level of the chosen component must be representative of the level of total
hydrocarbon emissions. The primary use of IR analysis may be as a detector in some other form
of analysis system.
b. Combustion Analyzers — Two types of instruments of this type are available. In the first, the
sample stream, mixed with air if necessary, is passed over a heated platinum wire which catalytically
oxidizes the hydrocarbons. The heat of combustion causes the wire to heat up which changes its
resistance. The wire forms one leg of a bridge circuit. The change of resistance unbalances the
bridge and this is used as a measure of total combustibles.
The disadvantage of this system is that the platinum is easily poisoned and so cannot be
depended upon to have constant activity over a period of time. Also, for simple hydrocarbons the
heat of combustion follows a regular pattern with the number of carbon atoms. For more complex
molecules, the heat of combustion follows a more variable pattern so that one must know the principle
components present to relate instrument output to total hydrocarbons. Finally, other combustibles,
such as CO, interfere with the analysis.
In another type of combustion analyzer, the organics are burned completely over a catalyst
such as copper oxide. The CO2 produced is measured, typically by infrared, and is used as a
measure of total hydrocarbons. Poisoning is not a problem here and response depends only on
184
-------
the number of carbon atoms and not on the particular molecules present. The presence of CCb
and CO in the original sample interferes with the analysis and must be determined separately.
Dual column chromatographic techniques, with backflushing, have been used with this type of
system, but are not applicable to continuous monitoring.
c. Flame lonization Detection — This technique is probably the most promising at the present
time. When hydrocarbons pass through a hydrogen flame they are ionized. This enables the flame
to carry an electrical current. By measuring this current, it is possible to get a measure of the carbon
atoms present. The method is not sensitive to CO2, CO or water vapor. It adapts easily to continuous
measurements and is the most sensitive to low concentrations. Furthermore, instruments are
commercially available which are entirely enclosed in a temperature controlled oven which allows
operation at high temperatures.
The chief disadvantage is that the principal components of the stack must be known.
Response for simple hydrocarbons is regular with carbon number. However, correction factors must
be applied if the stream contains significant amounts of hydrocarbons containing oxygen or other
groups. In extreme cases, calibration mixtures will be required. Also, the instrument requires a supply
of hydrogen and pure air or oxygen. A schematic of a typical flame ionization detector system set
up for continuous measurement is shown in Figure 39. Particulate and aerosol must be excluded
from the sample stream prior to the hydrogen flame.
d. Gas Chromatography — The previous methods attempt to measure total organics without
regard to identification of individual components. (Infrared can be used for the latter purpose but
it requires a far more elaborate and sophisticated instrument than is usually used in air pollution
analysis.) Gas Chromatography, on the other hand, permits a separation of a sample into its com-
ponents. A detector, usually thermal conductivity or flame ionization, is then used to identify and
measure the fractions. A wide variety of techniques and columns are available to permit identification
of practically any material that may be encountered. The requirements can be quite elaborate,
however, and do not lend themselves to field use. Nor is GC suitable for continuous monitoring.
Also, the multiplicity of components usually involved may make column selection difficult. It is useful
as a support device for the previously described devices.
C. E.P.A. Test Methods*
The location of the instrumentation relative to the sample source, necessitated transporting
the sample more than 100 feet through a heated sample transport system (Figure 40), consisting
This section has been adapted from a report prepared for the E.P.A. by Scott Laboratories,
Plumsteadville, Pa. The technique described herein was devised for sampling a particular plant and
may not be completely general in its applicability.
185
-------
0
u.
3
3
L.
U
y
I
U
186
-------
^y
CO Q.LU
x2z
LU^-J
CO
LU
Q
LU
LU
X
UJ
Z
_J
UJ
1
— J
Q.
S
ti
LU
fc
CO
CO
LU
1
jj
>
1
— — -
I
\^-
8 i
LU Q.
N CO
_l
L_ O
O CD CC
H < N
Q rr <
H ><
LU
X
1
J
ct
CD
CO O
< h-
(D <
o
o:
X
0
X- ^z<
Q;
Q
UJ
or
LU
-------
of a 1/4" O.D. x 18" stainless steel sample probe connected to six feet of externally heated 1/4"
O.D. teflon sample line. This flexible section of sample line was connected to a heated 7 cm clamshell
prefilter joined to a heated stainless steel bellows pump. From this point the sample was transported
approximately 100 feet through resistance heated 3/8" O. D. stainless steel tubing. With the exception
of the sample probe all components of this system were maintained at approximately 250°F to
prevent hydrocarbon losses due to condensation. The hydrocarbon analyzer and gas chromatograph
were connected in parallel at the end of this section of sample line.
Total hydrocarbons were continuously monitored with a Scott Model 215 heated total hydro-
carbon analyzer, utilizing a flame ionization detector. The detector bench and sample pump were
maintained at 300°F to prevent condensation losses of high molecular weight hydrocarbons. The
analyzer was optimized to yield an oxygen response of less than 0.2%. The carbon response
linearity was checked by comparing hexane, toluene and propylene standards against propane.
The average carbon response for the three classes of hydrocarbons was 98.9%. The instrument
was spanned with "Close Tolerance" (±2.0% analysis) blends of propane in air and zeroed with
hydrocarbon free air (<0.1 ppm Ci). The fuel and oxidant for the flame ionization detector were
40% hydrogen in helium and blended air respectively. The continuous total hydrocarbon trace was
recorded on a Texas Instrument Servo/Riter II recorder.
For all tests the sample backpressure was maintained at 2.0 psig with a bypass flowrate
of 6 SCFH. The analyzer was zeroed and spanned before and after each test and at regular intervals
during each test. In order to correlate total HC emissions with process conditions, notations were
made on the charts of process mode changes, time checks, gas chromatograph injections,
reactor temperatures and other pertinent information.
Each total hydrocarbon strip chart was divided into specific intervals based on the various
modes of operation for each process cook. An average hydrocarbon level in ppm-Ci was then
calculated for each process interval. Mass emission rates in Ib/hr were then calculated based on the
average flowrates measured for each process interval. The mass emissions per tons of resin produced
was calculated by first finding tons of resin produced per hour and comparing that with the mass
emissions rate (Ibs/hr).
While the T.H.C. analyzer was continuously monitoring hydrocarbon concentrations, point
samples were injected periodically into a gas chromatograph. The gas chromatograph, a Varian
Model 1200 equipped with a flame ionization detector and a (1 mv.) Texas Instrument Recorder
was used to provide a qualitative and semi-quantitative analysis of the effluents from their kettles.
188
-------
Ideally, for a quantitative analysis, it would be necessary to prepare a standard to analyze
each component of the effluent. Since this was impractical, an alternate procedure, to give a semi-
quantitative analysis, was chosen. Through proper manipulation of the fuel to oxidant ratio, the flame
ionization detector of the gas chromatograph was rendered linear for carbon response. Linearity
was checked by comparing cylinders of known concentrations of m-xylene, benzene, and n-hexane.
The variation in response factor for these compounds was approximately 6%. Thus, a cylinder
containing known carbon concentration (59.7 ppm m-xylene or 477.6 ppm Ci) could be used to
determine carbon concentration of each component in the sample. Operating parameters for the
gas chromatograph with the flame ionization detector are summarized below:
GAS CHROMATOGRAPH PARAMETERS
VARIAN MODEL 1200 GAS CHROMATOGRAPH
Parameters
Detector Temp.
Gaseous
Injection
Sample Loop
Column
Temp.
Program
Fuel (H2)
Press (psi)
Flow Rate
(cc/min.)
Oxidant (02)
Press (psi)
Flow Rate
(cc/min.)
Alkyd Solvent
Cook
200°C
.25 cm3
IGEPAL
Capillary
(50' x .02" I. D.)
40°C-110°C
@ 4°/min.
20
62.5
35
600
Polyester
Cook
250°C
1 cm3
Poropak Q
(3V2')
100°C-230°C
@ 6°/min.
20
62.5
35
600
Alkyd Fusion
Cook (soya)
200°C
1 cm3
IGEPAL
40°C-130°C
@ 4°/min.
20
62.5
35
600
Thinning
200°C
1 cm3
IGEPAL
40°C-130°C
@ 4°/min.
20
62.5
35
600
Carrier gas (He)
Flow Rate
(cc/min.)
50
Expecting to see solvent emissions from the kettles, an IGEPAL (Nonyl Phenoxypoloxyethylene
Ethanol) capillary column, (50' x .02" I. D.) was chosen to perform the analysis. With a quick
qualitative analysis in mind, sample chromatograms were prepared from the vapors of various paints.
189
-------
Then, having a previous knowledge of the paint solvents and holding chromatographic conditions
constant, each component was identified by comparing its retention time to that of standards.
Duplicating these chromatographic conditions in the field would enable a reasonable analysis to be
performed.
After the initial testing of the stacks began, however, it became obvious that light solvents
were involved in only the alkyd solvent cook, where xylene and ethyl benzene vapors were the
major components of the effluent. The IGEPAL column, as field chromatographic conditions were
designed, was not capable of handling the heavier constituents (high boilers) or the aliphatic hydro-
carbons in the effluent of the alkyd fusion or polyester resin cooks. It was necessary therefore to
obtain additional chromatographic data by taking five liter Tedlar® bag samples during the fusion
and polyester cooks. These bags were returned to a laboratory where other columns and operating
conditions could be utilized in an effort to identify and quantify the major hydrocarbons. Bag samples
were also taken of the thinning tank emissions.
Several techniques were employed in an effort to obtain accurate stack gas flow rates at
the various sampling locations. Most of the measurements were made with an S-type pilot tube
and hook gauge capable of measuring pressure differentials of 0.002 to 2.000 inches of water.
Where pressure differentials greater than 2.0 inches of water were found, a 10-inch water manometer
was used. When the pilot tube was calibrated in the laboratory, a pilol factor of 0.850 was calculated
for pressure readings of 0.01 inches of HaO or greater. For pressure readings of less lhan 0.010,
a pilol faclor of 0.758 was calculated. For lower flow conditions where the pilot lube could nol be
used, Iwo differenl bag samples were configured. The first of Ihese consisted of a 5 cubic fool
Tedlar® bag filled wilh a lapered silicone rubber bool, large enough lo fil over a Ihree inch diameter
slack. The flow measuremenl was made by attaching ihe bag lo the desired stack opening and
recording ihe lime required lo colled a sufficienl volume of sample. The bag was Ihen attached lo
a pump and evacuated ihrough a dry gas meter to measure the volume collected.
This melhod proved to be adequate until it was noticed that some of the stacks were
"breathing" i.e., drawing in ambient air and exhausting stack gases in a cyclic manner. This meanl
that a new system had to be configured lhal would permil only Ihe measuremenl of Ihe posilive
portion of Ihe flow and nol yield a nel flow over Ihe brealhing cycle. This was accomplished by
adding a silicone rubber one way flapper valve al Ihe exit of the stack. Since Ihe flow rale under
Ihese condilions was exlremely low, a smaller 2.5 cubic fool bag was used. The inlel lo ihe bag
was filled wilh the same silicone rubber boot used on the original bag, and Ihe oullel connections
190
-------
used for evacuating the bag consisted of a short piece of 1/4" O.D. Teflon® tubing fitted with a
syringe cap to prevent leaks during the sampling period. The bag was placed inside a cardboard
box to protect it from the wind while sampling.
Flow measurements were made at half hour intervals during each process, when possible,
and at more frequent intervals when process changes required. Stack gas temperatures were also
measured in conjunction with each flow measurement.
II. CONTINUOUS SOURCE MONITORING TECHNIQUES USED BY INDUSTRY
Industry has generally not used continuous monitoring techniques on any kind of a routine
basis. Sampling has been done on a "grab" or semicontinuous basis and then only as needed
for a specific situation, such as design information or to gather emission data for operating permits.
The grab samples have not proven to be too reliable and have further decreased the desire of
industry to monitor emissions in any fashion.
A continuous hydrocarbon analyzer has been used by one manufacturer for monitoring
emissions from resin manufacturing. The system uses an explosion proof flame ionization detector.
The sampling line is fitted with an entrainment separator. This was supplemented with a gas chromato-
graph for determining the composition of the sample stream. Sampling was done at the condenser
vents of the reactors. Their procedure did not include methods for determining paniculate or aerosol
loadings.
Some thermal and catalytic incinerator manufacturers also use continuous flame ionization
monitoring of a kettle cook for determining design parameters. They usually measure only for the
two or three cooks that represent the plant's maximum emissions.
The continuous monitoring of emissions from an incinerator outlet could be very successfully
applied. All problems of condensation, line plugging and hold up of heavy materials in the sample
train should be eliminated after the gases have been incinerated.
How well the flame ionization detector would stand up under continuous use is not known;
however, it is anticipated it would require more service and operator time than the incinerator
itself.
This flame ionization detector could also be used for control of hydrocarbon concentration
at an incinerator outlet. This would be especially helpful and adaptable for catalytic incineration.
Control would be accomplished by adjustment of the catalyst temperature to compensate for decrease
in catalyst activity as the catalyst ages with time.
191
-------
192
-------
CHAPTER 4
INVENTORY OF EMISSIONS
I. EMISSION FACTORS FOR EACH SOURCE
The emission data reported in the questionnaires has been examined in an effort to
correlate quantity of emission with production volume. A primary reason for this effort was to obtain
an approximate geographical distribution of air pollution "potential" for the industry. Those sections
of the questionnaires dealing with emissions were on the whole the least comprehensively filled
out. Fewer than 10% of the questionnaires reported sufficient information to enable an estimated
emission factor to be computed. The extreme variations in the calculated values and the lack of
any discernable pattern casts some doubt on the realiability of the estimates so obtained.
Three types of emission factors were sought. From plants which make coatings only (no
resins or varnish), pounds of gaseous emissions per million gallons solvent based coatings and
pounds of particulate emission per million gallons of all types of coatings were correlated. These
distinctions were made on the assumption that most gaseous emissions are attributable to solvent
based coating production and that particulate emission, mainly pigment, are attributable to most
surface coating production. Table 47 presents the emission factors calculated from these plants.
Plants which manufacture only resins were used to obtain pounds of gaseous emissions per
million pounds of resin solids. This data is summarized in Table 48.
Plants which produce both finished coatings and resins permitted the calculation of all three
types of emission factors. Data from plants in this category were used only where particular emissions
reported could be assigned to either paint or resin production with some degree of confidence.
Information from these plants is summarized in Table 49.
Only those plants whose emission inventory was reasonably inclusive of most of the major
sources were used. Major sources in this context include raw material handling and storage operations,
reactor vents, filling operations, ventilation hood exhaust vents, and general building exhaust
systems. Most emission data was reported in pounds per hour. This was multiplied by the number
193
-------
TABLE 47
EMISSION FACTORS FROM SELECTED PAINT PLANTS
Plant
2A
28
2C
2D
2E
2F
2G
2H
21
2J
2K
2L
2M
2N
20
2P
2Q
iuorano
Gaseous
(Ib/MM gal solvent based)
8,800
21,000
71,500
45,800
102,800
89,300
7,700
141,200
95,900
9,200
104,000
15,900
25,200
46,100
e:fi nnn
Participate
(Ib/MM gal)
295,000
58,100
3,600
Source: Questionnaire data
194
-------
TABLE 48
EMISSION FACTORS FROM SELECTED RESIN PLANTS
Plant
3A
3B
3C
3D
3E
3F
3G
3H
31
3J
3K
3L
3M
3N
Average
Gaseous
(Ib/MM Ib resin solids) Principal resin types
87,600 Polyester, alkyd
2,600 Alkyd, varnish
1,750 Rosin types
1,700 Water based acrylic
21,100 Water based vinyl
162 Acrylic (water & solvents)
15,350 Hydrocarbon
1,120 Alkyd, water based vinyl
12,600 Varnish
10,500 Amino, polyester
3,400 Epoxy
19,500 Epoxy
1,300 Alkyd, polyester
73,150* Vinyl (solvent and water)
13,700
*Most of this emission is from "resin dryer" operation.
Not included in average.
Source: Questionnaire data
195
-------
TABLE 49
EMISSION FACTORS FROM SELECTED PLANTS
PRODUCING COATINGS AND RESINS
Plant
1A
1B
1C
1D
1E
1F
1G
1H
Average
Resin Production
Gaseous
(Ib/MM Ib)
824
560
1,100
1,800
22,000
6,500
Coating Production
Gaseous
(Ib/MM gal solvent based)
36,800
5,300
22,000*
113,000
46,400
35,100
37,700
65,000
Particulate
(Ib/MM gal)
19,400
5,460
45,000
*ln addition to this, the plant also emits 235,000 pounds/year from spray booths and drying ovens.
Source: Questionnaire data
196
-------
of plant operating hours per year on the assumption that the source in question operates at the
stated emission rate for the same number of hours. In some cases, at least, this assumption is
probably not correct. Emissions were correlated for the uncontrolled process assuming no emission
control devices.
No particular pattern has emerged to help interpret the results summarized in Tables 47
to 49. The magnitudes of the emission factors calculated for coating production, for instance, show
no dependence on solvents used, production volume, per cent industrial sales, etc. Similarly, emissions
from resin production do not seem to correlate with types of cooks, types of resins, or solvents
used. In any event, simple averages were calculated and are also presented in the Tables. No
attempt was made to average the particulate emissions due to wide scatter and few data points.
II. EMISSION INVENTORY FOR THE INDUSTRY
The emission factors obtained above permit one to estimate potential emissions for a
geographical area for which production figures are available. Production by states has been estimated
for 1972 in Chapter 2. Using these outputs and assuming a uniform ratio of solvent based production
(77% of total gallons produced), gaseous pollution potential from uncontrolled paint manufacturing
operations has been estimated for each state. The results are presented in Figure 41. An average
emission factor of 50,000 pounds gaseous emissions per million gallons of solvent based paint
was used in the calculation. Total gaseous emissions for the U.S.A. from paint production are
estimated to be 23,200 tons for the year 1972.
The emission factors presented in this section can be applied to the model plant discussed
in Chapter 1. The model plant produces 1.1 million gallons of solvent based paint and 2.1 million
pounds of resin solids. Based on the average emission factor for this type of plant given in Table
49, the model plant should emit (1.1 x 45,000) + (2.1 x 5,460) = 60,966 pounds per year of gaseous
emissions. The source-by-source calculation in Chapter 1 totaled 65,590 pounds. The two calculation
methods agree quite well as far as totals are concerned. The relative amounts of emissions attributable
to resin production as opposed to paint production show wider variations between the two methods.
The emission factors predict less from paint production and more from resin production.
The degree of consistency between the two methods is remarkable considering the extreme
scatter present in the emission factor data and the assumptions used in the model plant calculation.
It would be unwise, however, to draw any strong conclusions concerning the applicability of either
approach to coatings manufacturing operations selected at random.
197
-------
198
-------
CHAPTER 5
EMISSION CONTROL TECHNOLOGY
In this chapter a detailed discussion of the existing state of the art of air pollution control
for this industry will be presented. The discussion will include a description of the best control
technique for each emission source, alternate control techniques, methods of control other than
add on equipment, types of performance of currently used systems and capability of best control
systems to meet more stringent standards. Potential water and solid waste disposal systems will
also be reviewed.
I. DESCRIPTION OF BEST CONTROL SYSTEMS
As discussed in Chapter 1, the two mam types of non-fugitive emissions are gaseous
organic pollutants and pigment and resin particulate. The best control techniques for these pollutants
are discussed below.
A. Control of Particulate Emissions
The best control device for pigment and resin particulate is a fabric collector. The system
for collecting and controlling pigment dust during the loading operation of mixers, ball mills and
the like is shown in Figure 42. Depending on the plant layout there may be one or more of these
systems. The principal part of the system is the baghouse or fabric filter. It is by far the best control
device and is ideally suited for this application. Collection efficiency of the submicron pigment
particulate (0.05 to 0.25/x) is very high, in the range of 99.9%. The gas stream is low temperature
and the grain loading is low. The collected product can be recycled for use in dark primer paints.
The installed cost for this system can be quite expensive, depending upon the collection system
used. Basically there are two types. A fixed collection hood for each loading station that collects
both dust and empty pigment bags or a movable flexible hose that can be placed at one or more
loading stations while the pigment bags are being emptied. In some cases the hose is connected
directly to the mixing tank and ventilation air is drawn through the loading hatch with the pigment
particulate. This is a very efficient method of capturing the fugitive dust. The disadvantage occurs
due to increased solvent losses from the mixing tank.
199
-------
200
-------
There is a large variety of baghouses currently available as "off the shelf" items from over
a dozen manufacturers. They all operate in essentially the same manner. The pigment laden air
is filtered by the cloth tube or bags contained in the house. The house itself is divided into four
sections — the dirty air plenum, the bag area, the clean air plenum and the particulate collection
hoppers. The function of the two plenums is to properly distribute the air flow to and from the bags.
Many types of bag materials may be used, but woven cotton bags are not only adequate but also
the least expensive and, therefore, the most commonly used for this service. The collected particulate
needs to be periodically cleaned from the bags. This can be done either intermittently or continuously.
Intermittent baghouses are designed for periodic cleaning, such as once a day, and are easily adapted
to the low dust loadings and batch operation encountered in the paint industry. They are normally
the lowest cost baghouse. Continuous baghouses are cleaned automatically and are able to operate
24 hours per day. They can also handle high dust loadings.
Cleaning of the bags can be accomplished by a variety of methods which includes shaking
the bags, reversing the air flow through the bags, blowing a jet of air on the bags from a reciprocating
manifold, or rapidly expanding the bags by a pulse of compressed air. The method most commonly
used with the intermittent baghouse is shaking which can be done manually at the end of a days
operation. The bags in a shaker-type baghouse are supported by a structural framework which is
free to oscillate. If this type of baghouse is used on a continuous basis the bags are cleaned auto-
matically. Periodically, by use of a timer, a damper isolates a compartment of the shaker baghouse
so that no air flows. The bags in the isolated compartment are then shaken for a minute or so,
during which time, the collected pigment or resin is dislodged and falls into the baghouse hoppers.
Another type of continuous baghouse commonly used in this industry, as characterized by its cleaning
method, is the reverse pulse baghouse. This type utilizes a short (100 millisecond or less) pulse
of compressed air through a venturi or diffuser. The primary air pulse is directed from the top to
the bottom of the bag and aspirates secondary air as it passes through the venturi. The resulting
high flow air pulse travels the length of the bag rapidly expanding the cloth and dislodging the
collected dust cake.
The primary variable used in applying a baghouse to this or any application is the air-to-
cloth ratio as defined below:
R = Q/A
where:- R = air-to-cloth ratio, feet/min
Q = volumetric air flow, ACFM
A = net cloth area, ft2
201
-------
Air-to-cloth ratios normally run from 1 to 3 for shaker and reverse air baghouses while reverse
pulse units will run between 5 and 18. The basis for selection will vary with dust loading, dust
type, desired bag life and allowable pressure drop. Typical air-to-cloth ratio utilized in the paint
industry are 8.4/1 for a reverse pulse jet baghouse with polyester bags and 2/1 for a shaker baghouse
with cotton satin bags.
In most cases, pigment emissions are of a minor nature and do not violate existing air
pollution control regulations. Because of this and the extra cost involved, the use of a baghouse
collection system is not a wide spread practice. Those in existence have been motivated by occupational
health, good housekeeping and product recovery.
A similar problem exists for resin manufacturing plants in the grinding or flaking of the
hardened resin. The same approach is normally used. The air from the baghouse may still contain
a significant odor and further treatment may be required if a local nuisance exists.
B. Control of Gaseous Emissions
The best control technique for gaseous emission from the paint and varnish industry is
oxidation or combustion of the organic pollutants to CC>2 and HkO. This is the only control technique
currently being used that has proven effective for all cases. Three general methods are employed
to oxidize waste gases, as follows:
Flame Incineration or Direct Combustion
Thermal Combustion or Oxidation
Catalytic Combustion or Oxidation
All of the above methods are oxidation processes. Ordinarily, each requires that the gaseous effluents
be heated to the point where oxidation of the combustible will take place. The three methods
differ basically in the temperature to which the gas stream must be heated. These methods will
be briefly described for better understanding of their application to the paint and varnish industry.
1. Flame Incineration
Flame incineration is the easiest of the three to understand, as it comes the closest to
every day experience. When a gas stream is contaminated with combustibles at a concentration
approaching the lower flammable limit, it is frequently practical to add a small amount of natural
gas as an auxiliary fuel and sufficient air for combustion when necessary, and then pass the resulting
mixture through a burner. The contaminants in the mixture serve as a part of the fuel. Flame inciner-
ators of this type are most often used for closed chemical reactors. Figure 43 is a schematic illus-
tration of a flame incineration unit. This unit resembles a flare operated within a combustion chamber
202
-------
LADEN
PROCESS
STEAKS
COMBUSTION
AIR
NATURAL
GAS
FIGURE:
SCHEMATIC DIAGRAM OF A FLAME INCINERATION UNIT
203
-------
where the combustion conditions may be controlled carefully.
Flame incineration cannot be used for open kettle cooking because of the high volume and
low combustible concentration of the exhaust. It would be ideally suited for closed reactor kettles
if these could be run at a high enough pressure to supply the driving force through the system.
Unfortunately most kettles must be opened periodically for additions of various materials such as
pentaerythritol and phthalic anhydride. Also, many resin cooks must be run under a vacuum. Aside
from these limitations, a suitable transfer line would have to be developed to be both functional
and comply with insurance safety requirements.
Flame incineration systems are generally difficult to handle with regard to either self-recuper-
ative or makeup types of heat exchangers because of the extremely high temperature of the
effluent from the combustion chamber. For this reason, it is usually most economical to provide a
steam generator to recover the waste heat rather than a gas-to-gas heat exchanger. With water
as the cooling fluid, the hot side of the heat exchanger may be exposed to the temperatures generated
in the combustion zone.
2. Thermal Combustion
It is far more likely that the concentration of combustible contaminants in an air stream
will be well below the lower limit of flammability. When this is the case, thermal oxidation is considerably
more economical than flame incineration. Thermal oxidation is carried out by equipment such as
that illustrated schematically in Figure 44. In this equipment, a gas burner is used to raise the
temperature of the flowing stream sufficiently to cause a slow thermal reaction to occur in a residence
chamber.
Whereas flame temperatures bring about oxidation by free radical mechanisms at temperatures
of 2500°F,27 28 29 high conversions are produced by thermal afterburners at temperatures in the order
of 1400°F with 1/2 second residence time.27 28 29 Thermal afterburners can operate at conversion
efficiencies in excess of 97% depending upon a number of operating variables. These conditions can
be summarized as follows:
a. Operating Temperature
b. Quantity of Hydrocarbon
c. Residence Chamber Size or Residence Time
d. Type of Fume
e. Uniformity of Temperature — Good Mixing
The theoretical aspects of the reaction kinetic for thermal oxidation are discussed on the
following page.
204
-------
L
0
o
IL
D
8
(VI
Id
> It
J b hi
6^:<
0 u
x\
0 o
5 (t 0
i_ 0
7 ifl
0 li
u a:
> " 2
U UJ
1- h
>^
to:
?3
*d
Id It
Ih
2 z
Id 0 Q.1
HU 5L
Ido
I- o
It (\j
Id —
it in
CD
-
Oidtf)
NL_E:T
IL
0
0
0
0)
(L
U
t-
y j
Z LJ
it D
L
i£-
\iy
P
in
Id
Q
LJ
!c
Z
P
0
u
FIGURE
DC
y
j
8
Q
b^
y u
P
i S
u y
205
-------
If a gas containing species A and B which react to produce C, such that there is substantially
no reverse reaction of C decomposing into A and B, we can write the equation
A + B-» C
or for the reaction of a hydrocarbon in air
He + O2 -» CO2 + H2O
For flow processes, let us assume steady state conditions of a gas stream passing through
a tube with good uniformity across any given cross-section. The concentration of He, O2, COz,
and h^O will vary with time as an element of the flowing fluid moves through the reactor, or with
distance, in that time and distance into the reactor are interchangeable for steady-state conditions.
We can expect the concentration of He and O2 to drop off and can define the rate of
disappearance of He as follows:
R = - k (Hc)m (O2)n
where:
R = Rate of disappearance of He, mol/hr
k = Reaction Rate Constant, function of temperature
(He) = Concentration of He, mol/fraction
(O2) = Concentration of O2, mol/fraction
For almost all cases, including varnish cooking, the O2 content of the gas stream is high
and the He content is low. Because of this, the O2 concentration is substantially constant and can
be omitted. The exponent of (He) can be taken as 1.
R =-k (He)
Using x as the concentration of combustible:
dx dx
= —kx or — = —kt
Integrating this gives x = Ae
Using initial conditions x = XQ at t = o
dt x
-kt
x = x at t = t-i
x0 = Ae° or A = x0
=e-kt
If conversion is defined as: c =
Then 1 - c = e ~kt
XQ-X
206
-------
RE
LU
- (0
DC -fc
-I F
!'
ho
0
i 2
o u
205
-------
If a gas containing species A and B which react to produce C, such that there is substantially
no reverse reaction of C decomposing into A and B, we can write the equation
A + B^ C
or for the reaction of a hydrocarbon in air
He + O2 -» CO2 + H2O
For flow processes, let us assume steady state conditions of a gas stream passing through
a tube with good uniformity across any given cross-section. The concentration of He, 02, CC>2,
and hbO will vary with time as an element of the flowing fluid moves through the reactor, or with
distance, in that time and distance into the reactor are interchangeable for steady-state conditions.
We can expect the concentration of He and Oa to drop off and can define the rate of
disappearance of He as follows:
R = - k (Hc)m (O2)n
where:
R = Rate of disappearance of He, mol/hr
k = Reaction Rate Constant, function of temperature
(He) = Concentration of He, mol/fraction
(62) = Concentration of O2, mol/fraction
For almost all cases, including varnish cooking, the C-2 content of the gas stream is high
and the He content is low. Because of this, the 62 concentration is substantially constant and can
be omitted. The exponent of (He) can be taken as 1.
R =_k (He)
Using x as the concentration of combustible:
dx dx
— = —kx or — = —kt
dt x
—kt
Integrating this gives x = Ae
Using initial conditions x = xo at t = o
x = x at t = t-i
x0 = Ae° or A =
— =e"kt
x,.
XQ -X
If conversion is defined as: c = —5
xo
Then 1 - c = e ~kt
206
-------
From this, knowing the reaction rate constant and residence time, the conversion can be
easily predicted. This formula applies directly only for an isothermal system. However, for most
afterburner applications of contaminated air, k is a function of temperature which varies along the
length of the afterburner. Calculation of conversion, then, requires a more lengthy solution of the
rate equation over differential elements of the incinerator.
Past research indicates that most combustibles follow the pseudo first order reaction pattern
described above and rate constants can be correlated by plotting In k vs. —.
The few points required to plot these curves are experimentally determined in the laboratory
by measuring conversion rates for a known combustible material at two or more operating temperatures
in a reactor of fixed residence time.
This type of experimental work, to our knowledge, has not been done on open or closed
varnish or resin kettle fumes. Experimental work on other hydrocarbon vapors indicates, however,
that general guidelines can be applied to this type of contaminant. For example, 0.3 to 0.6 seconds
residence appears to be the optimum range of reaction time and has been adopted as a standard
by most afterburner manufacturers and air pollution regulatory bodies.
Figure 45 shows a relationship between temperature and residence time at a fixed conversion
level of 95% and fume concentration of 2 Btu/SCF.
Figure 46 shows the relationship between conversion and temperature at a fixed residence
time of 0.6 seconds.
Most afterburners are capable of operating at a temperature high enough to oxidize any
organic material, except carbon particulate, and the actual operating temperature can be developed
for a given type of afterburner after installation by trial and error. It is important when this approach
is taken to insure the afterburner has at least 0.6 seconds residence time and the ability to operate
at 1500°F.
Afterburners are designed to operate at a controlled outlet temperature and heat is added
by the burner to the contaminated air stream to be incinerated until this temperature is obtained.
Assuming the burner outlet is set high enough for the oxidation reaction to initiate, the heat of
reaction of the combustibles will also be added to the air stream in the afterburner. This, in turn.
will reduce the heat requirements of the burner. Less fuel will be added to the burner. Temperature
rise across the afterburner will increase and the afterburner inlet temperature will drop. The gas
temperature in the residence chamber now varies from inlet to outlet. The higher the fume loading
the greater the temperature variation across the afterburner. If the afterburner outlet temperature is
207
-------
FIGURE 4.5
RESIDENCE TIME VS. TEMPERATURE
AT
9.5% C.ONVERSION & 2BTU/SC.F
ii
0)
N,
y
h
y
o
i/)
y
1.7
I.S
0.9
O.T
0.5
ISOO
I35O
I4-OO
OUTLET TEMPERATURE, °F
208
-------
From this, knowing the reaction rate constant and residence time, the conversion can be
easily predicted. This formula applies directly only for an isothermal system. However, for most
afterburner applications of contaminated air, k is a function of temperature which vanes along the
length of the afterburner. Calculation of conversion, then, requires a more lengthy solution of the
rate equation over differential elements of the incinerator.
Past research indicates that most combustibles follow the pseudo first order reaction pattern
described above and rate constants can be correlated by plotting In k vs. —.
The few points required to plot these curves are experimentally determined in the laboratory
by measuring conversion rates for a known combustible material at two or more operating temperatures
in a reactor of fixed residence time.
This type of experimental work, to our knowledge, has not been done on open or closed
varnish or resin kettle fumes. Experimental work on other hydrocarbon vapors indicates, however,
that general guidelines can be applied to this type of contaminant. For example, 0.3 to 0.6 seconds
residence appears to be the optimum range of reaction time and has been adopted as a standard
by most afterburner manufacturers and air pollution regulatory bodies.
Figure 45 shows a relationship between temperature and residence time at a fixed conversion
level of 95% and fume concentration of 2 Btu/SCF.
Figure 46 shows the relationship between conversion and temperature at a fixed residence
time of 0.6 seconds.
Most afterburners are capable of operating at a temperature high enough to oxidize any
organic material, except carbon particulate, and the actual operating temperature can be developed
for a given type of afterburner after installation by trial and error. It is important when this approach
is taken to insure the afterburner has at least 0.6 seconds residence time and the ability to operate
at 1500°F.
Afterburners are designed to operate at a controlled outlet temperature and heat is added
by the burner to the contaminated air stream to be incinerated until this temperature is obtained.
Assuming the burner outlet is set high enough for the oxidation reaction to initiate, the heat of
reaction of the combustibles will also be added to the air stream in the afterburner. This, in turn.
will reduce the heat requirements of the burner. Less fuel will be added to the burner. Temperature
rise across the afterburner will increase and the afterburner inlet temperature will drop. The gas
temperature in the residence chamber now varies from inlet to outlet. The higher the fume loading
the greater the temperature variation across the afterburner. If the afterburner outlet temperature is
207
-------
FIGURE
RESIDENCE TIME VS. TEMPERATURE
AT
CONVERSION &. 2BTU/SCF
0
y
0)
s
y
5
h
y
0
y
D
0)
y
1.7
0.9
0.7
0.&
I3.OO
I4OO
OUTL.ET TEMPERATURE, °r
208
-------
FIGURE "4-e
CONVERSION VS. TEMPERATURE
AT"
0.6 SEC. RESIDENCE TtME & 2BTU/SCF
100
90
z
0
vi
C
u
>
o
o
ao
TO
eo
14-00
OUTLET TEMPERATURE,*F
209
-------
held constant, the average temperature in the afterburner will decrease as the fume load increases.
Since the reaction rate is a function of temperature, it will also decrease.
The theoretical relationship between conversion and fume loading at a fixed outlet temperature
and residence time is described in Figure 47. This relationship is applicable only to systems having
a fixed outlet temperature. If the inlet temperature is held constant, the opposite effect will occur and
conversion efficiency could increase with higher inlet concentrations. This is due to the first order
dependence of oxidation rate on initial concentration, the effect of local heat of reaction, the effect
of increased concentration of free radicals and the higher reaction temperatures reached. These
effects are discussed in more detail in the Afterburner Systems Study, EPA Contract EHS-D-71-3.27
From the above discussion, it can be seen that for a given conversion level, the higher the
fume load the higher the outlet temperature must run to assure the required conversion. In fact,
for very high fume loads, if the outlet temperature is not set high enough, the temperature out of
the burner can drop below the reaction initiation temperature and the reaction will cycle on and off.
Field experience indicates the above discussion is not applicable at low fume concentrations.
This is caused by difficulties of measurement, generation of combustible material from the burner,
and generation of small quantities of partially reacted material that are normally insiginificant at
high fume concentration levels.
Frequently it is possible to identify specific materials which must be oxidized in order to
conform to the prevailing air correction ordinances, or to good practice. For example, for closed
kettles emitting a mixture of xylene and phthalic anhydride, one should consider the residence
time-temperature curve for P.A. and select conditions which will give a 90% or better conversion
of this material. As the xylene is less refractory than the P.A., this selection is "safe".
Often, however, it is not possible to chemically identify the specific compounds involved.
In such cases, the emission must be treated on the basis of past experience or by using pilot after-
burner equipment for field testing. Because the oxidation rates given are influenced by burner
design, distribution in the combustion chamber, etc., it is wise to follow the recommendations of
the manufacturer supplying the equipment rather than transfer data of this sort from one design
to another.29
In general terms, a thermal afterburner consists of a preheat burner, a mixing device, and
a residence chamber. The gas to be disposed of is first passed by and mixed with burner combustion
products to preheat it to at least the reaction initiation temperature. The type of mixing device used
is the single most important design feature in that it affects the amount of direct flame incineration
210
-------
FIGURE
CONVERSION VS. INL-EIT CONCENTRATION
AT
0.6 SEC.. RESIDENCE TIME & I3£>O~F OUTLET
90
ao
0
(A
or
y
>
z
0
0
7O
60
O 2 4 6
INLET CONCENTRATION,
211
-------
that occurs and the temperature uniformity of the gases in the residence chamber.
As high as 20% of the gas can be mixed with and incinerated by the high temperature
flame. This is a desirable effect in that it decreases the burner preheat fuel requirement and decreases
the combustible load on the afterburner.
Gas flow and temperature uniformity out of a mixing device and into the residence chamber
are important to assure that all the combustible is held at the proper temperature for the design
time period. This effect is sometimes called turbulence, which is a misleading word. Uniform flow
and temperature with or without turbulence is the important criteria. The importance of this variable
is explicitly outlined in the Afterburner Systems Study, EPA Contract EHS-D-71-3,27 and is quoted
below.
"Mixing of bypassed fume and hot combustion gases is the most crucial step
in attaining good afterburner performance. Typically, afterburners are designed with
-.5 second total residence time. This time is nearly all required for this mixing step
and many designs fail to complete the mixing in the distance (time) available. Some
fume escapes without being raised to a sufficiently high temperature. To meet a
performance specification (if possible at all) more fuel must be burned than would
be needed if mixing were complete. Distributed burners are placed directly in the
fume stream and divide the flame into many individual jets or lines of flame sur-
rounded by fume. This subdivision greatly speeds the mixing process, and these
burners are well suited to use oxygen from the fume for combustion. The use of
outside air requires an additional 30 to 50% of the fuel to be burned to heat it to
1400°F. Distributed burners are subject to fouling, have somewhat limited turndown,
aren't available for use with oil fuel, may be difficult to use with outside air, and have
a few other potential drawbacks. Therefore, many afterburners employ discrete
burners which give either long or short point sources of flame. The mixing problem
is much more difficult since there is no subdivision at the burner. Internal baffles
are required in the relatively short afterburner chambers utilized in available designs.
Many designs stress "flame contact" in an attempt to mix fume and flame as
rapidly as possible. This often leads to flame quenching and an increase in pol-
lutants in the fume stream since, as was mentioned above, complete fume/fuel
mixing gives a noncombustible mixture. Fuel should be burned as rapidly as possible
and the hot gases should be mixed with bypassed fume. Pressure drop must be
expended in achieving good mixing through baffles and/or a long chamber. Mixing
will be faster when there is initial fume/flame subdivision."
A schematic illustration of a typical thermal afterburner and control system has been presented
in Figure 48. As indicated on this Figure, most afterburners are equipped with heat exchangers
to decrease the fuel requirements of the preheat burner. The burner fuel consumption is the major
operating cost of thermal afterburners, and self-recuperative heat exchangers, except for very small
flow rates or unusual circumstances, are an economic necessity with rapid payout. The use of
212
-------
(VI
J
,2
h
0
U
„ it
1 U
V|
-w
. U '
U. £ r£
0 *™\
0 P ^
o s H
S ^0
II Id DC
o; Q. h
u id 0
H HO
nJ
X
Q.1
SL
ido
C]
zli
OC 12
Dh
OQ
INL-ET
CHAMBER
^
LJ J
Z U
D
L
Z
<
bJ
D
L
LJ
oc >•
UJ ^
fjj
IB
fc*
0 ^
y u
UJ CD
i cr
0 u
-------
additional heat exchange, where possible, has become more attractive in recent times with the advent
of fuel shortages and increased fuel costs.
Another technique commonly used to reduce operating cost through reduced fuel savings
is preheating of the contaminated air with a raw gas or secondary air burner. This type burner
utilizes the oxygen in the contaminated air for combustion of the natural gas.
This type of operation saves the cost of heating the combustion air and fuel consumption
can be reduced about 30%. These burners, however, do not operate well at oxygen concentrations
below 16% which would be the case for the emissions from a closed varnish or resin kettle or
reactor. The maximum allowable fume concentration to the afterburner as set by most insurance
companies is 1/4 of the lower explosive limit, or 13 Btu/SCF. This requires that the kettle exhaust
be diluted sufficiently with air to raise the oxygen concentration above 16%.
As mentioned earlier, heat exchange is the most common type of fuel saving applied to
thermal afterburners.
The type most often used is the self-recuperative type of exchanger which allows part of
the heat in the effluent gas to be used for preheating the inlet gas before it reaches the combustion
chamber. Gas-to-gas heat exchangers of this kind are generally supplied as a part of the thermal
afterburner unit and are ordinarily available only from the manufacturers of equipment. Inherently,
such exchangers are costly, and have a low heat transfer coefficient, 4 to 6 Btu/ft2 — °F. Aside
from the refractory regenerative type heat exchanger, most exchangers are of metal construction
and have a hot gas or afterburner outlet limitation of 1 SOOT. The most common type in use is the
shell and tube exchanger, and they are available in either parallel flow or countercurrent flow.
Parallel flow is most commonly supplied as a standard, especially where the fume load and
outlet temperature are high. At high fume loads, there is a danger of temperature spiraling. At
fume loads approaching 1/4 LEL, a 7.5% change in concentration will result in a temperature
change in the residence chamber of SOT. It is possible with sudden increase in fume loads to
raise the afterburner outlet temperature above the control point which, via the heat exchanger,
raises the inlet temperature above the burner's low fire control ability. The increased inlet, in turn,
raises the outlet some more which, in turn, raises the inlet, and so on. Parallel exchangers having
equal thermal flows on shell and tube side have a maximum thermal efficiency of 50% and are
self-limiting with respect to temperature spiraling. They also have the advantage of running the lowest
mean material temperatures, since the hottest gas is always exchanged with the coolest gas.
Countercurrent heat exchangers are much more thermally efficient for the same amount of
214
-------
heat exchanger area. They do run at higher material temperatures, however, and require more
expensive materials of construction. They are also more susceptible to temperature spiraling when
running at high fume rates. Because of this, a system designed around this type of exchanger
should use a burner duty of at least 50°F. An afterburner heat exchanger system using a parallel
flow heat exchanger is designed for a burner duty of at least 25°F. Burner duty in this context,
means the temperature rise added to the fume stream by the burner between the heat exchanger
outlet and the residence chamber inlet.
Rotary heat exchangers are available and will run at thermal efficiencies of 80%. They are
of metal construction (321 SS) and are still limited to a maximum operating temperature of 1500°F.
Their disadvantages are high initial and maintenance cost, as well as cross leakage. The cost of
an 80% rotary ($84,000) is about double a 60% cross flow shell and tube ($47,000) for a flow of
12,000 SCFM. The rotary exchanger is still in the development stage and, along with its many
operating problems, is limited to smaller sizes.
3. Catalytic Afterburners
While thermal afterburners bring about oxidation at concentrations below the limits of flame
combustion, catalytic afterburners operate both below the limits of flammability and below the normal
oxidation temperatures of the contaminants as well. The reaction is instantaneous by comparison
to thermal oxidation and no residence chamber is required. Catalytic oxidation is carried out by
equipment such as that illustrated in Figure 49. This particular arrangement provides an ideal
location for the fan used in the system. Generally speaking, catalytic afterburner systems are the
least costly when comparisons are made at the optimum level of heat recovery. Detailed installation
and operating and maintenance costs for both catalytic and thermal afterburners are presented
in Chapter 7.
While catalytic oxidation is attractive from an economic standpoint, several factors must be
considered in selecting this form of waste gas treatment. Catalysts require regular maintenance
in the form of periodic washing to remove atmospheric dust and dirt and, in the case of higher
temperature cooking, to remove traces of particulates and other ash-like residue originating in
kettle materials. In addition, it is necessary to reactivate the catalysts periodically. There are a
number of materials, such as phosphorus, silicon and lead, known to shorten the active life of
these catalysts. However, when the characteristics of the gas to be treated are suitable, catalytic
oxidation is a highly satisfactory method for air pollution control.27
215
-------
216
-------
Because catalytic units are not automatically functional when operated at design temperatures,
they may not be approved for installation until means for ensuring adequate performance of the
catalyst on a long term basis can be demonstrated.
As is the case for thermal afterburners, catalytic units can also operate at efficiencies in
excess of 97% depending upon a number of operating variables. These conditions can be summarized
as follows:
a. Type of Hydrocarbon — Reaction Rate
b. Quantity of Hydrocarbon
c. Type of Catalyst — Activity
d. Quantity of Catalyst — Space Velocity
e. Operating Temperature
f. Presence of Catalyst Suppressants or Poisons
Some hydrocarbons react faster or oxidize at lower temperatures than others. Relative
reaction rate constants for various hydrocarbons in the presence of a precious metal catalyst are
given in Figure 50.
As discussed earlier, the higher the hydrocarbon content of the exhaust stream, the greater
the heat release, and the higher the temperature rise for an equal conversion efficiency. This is
less of a problem than with thermal incinerators for two reasons. First, catalytic afterburners con-
structed of aluminized steel can sustain a temperature rise of 700°F, and units constructed of
stainless steel can sustain temperature rises of 1,000°F. High temperature rises do not result in
a lower average catalyst temperature or reaction rate since catalytic units are normally controlled
on the catalyst inlet side.
The catalysts in common use for oxidation reactions involved in stationary source control
are metals of the platinum-palladium group. The only catalysts used in significant amounts for oxidation
of trace hydrocarbons other than the platinum-palladium metals are merchandised in Europe and
consist of cobalt, copper, and chromium oxides incorporated into a ceramic substrate.
Oxidation, which is of primary concern in air pollution control, generally takes place only
over metal catalysts, and is inhibited by acidity in the support. The kinds of metals that are capable
of bringing about oxidation reactions are listed in order of the frequency of their utilization.
a. Platinum and Palladium
b. Nickel, Cobalt, Iron, Vanadium
c. Silver and Copper
d. Tungsten, Chromium, Molybdenum
217
-------
FIGURE
GATAL.YTIC OXIDATION RATES FOR SOLVENTS
1.0 -
ct
Q
0
<
y
or
iu
j
u
or
O.I
IOOO »SO «00 8SO SOO
rso
7CXJ
6OO
TEMPERATURE °F
218
-------
Because the reactions take place only on the surface of the active metals, very small
crystallites are generally preferred as they provide a much larger surface area per unit weight of
precious metal.
The available surface area of the substrate may be a critical factor, but is more likely to
be of marginal importance in oxidation reactions. The substrate surface area must be sufficiently
adequate that the active metal crystallites do not pile up one above the other and thus limit the
surface area exposure to the bulk gas stream.
In general, the metal ribbon catalysts have very little substrate surface area. The spherical
catalysts ordinarily have extremely large surface areas associated with a "micropore structure".
Generally the area is measured in square meters per gram of material with values frequently as
high as 250 square meters per gram. The ceramic honeycomb materials manufactured by duPont,
Minnesota Mining and Manufacturing and Dow Corning have somewhat greater superficial surface
area than do the metal ribbon catalysts, but have substantially no micropore area as the spherical
supports have.
Generally, catalysts of higher activity can be produced from a porous catalyst
support of high surface area than from a metal substrate. However, for ordinary applications at high
temperatures the precious metal located away from the outside surface of the spherical catalysts
is not used effectively, and the advantages of the porous bases seem to lie only in their having
more superficial surface area than it is convenient to pack into the same volume with the metallic
ribbons or ceramic honeycombs.
Even though the oxidation reaction over a catalyst is instantaneous in comparison to thermal
oxidation reaction time, the more catalyst it contacts the greater the conversion rate. The disad-
vantages of adding additional catalyst in an afterburner are increased pressure drop across the
catalyst bed and higher capital cost. For this reason, the expected operating temperature and type
of hydrocarbon must be considered when sizing the catalyst volume for a catalytic afterburner.
The effect of operating temperature on catalyst activity is illustrated on the previously
presented Figure 50. Basically, catalysts function by altering the rate of a reaction. If a gas contains
species A and B which react to produce C, such that there is substantially no reverse reaction
of C decomposing into A and B, we write the equation:
A + B^ C
For flow processes, if we assume steady conditions of a gas stream passing through a
container with good uniformity across any given cross-section, the concentration of A, B, and C
219
-------
will vary with time as an element of the flowing fluid moves through the reactor, or with distance,
in that time and distance into the reactor are interchangeable for steady-state conditions. We will
expect the concentration of A or B to drop off and can define the rate of disappearance of A as
follows:
dA
dA dt
where, —~ = rate of disappearance of A, mol/sec.
dt
For a normal gas phase reaction, the rate of disappearance of A will vary with the concen-
tration of A, the concentration of B, and the temperature. If we assume that the reaction is controlled
primarily by the concentration of A, as in the case of burning traces of hydrocarbons in an air
stream, the equation becomes:
dA
— = -ka
where, dt
k = reaction rate constant
a = concentration of A in mols/liter
The solution to this equation is:
where:
ao = initial concentration of A
which simply says that the concentration of A will drop off exponentially with time, or with distance
as the flowing stream moves down the reactor. If it does so, the reaction is defined as a first
order reaction with respect to A. In order to predict the course of a reaction such as this, all
we need to do is know what k will be. Generally speaking, for any given reaction, k is a function
only of temperature.
For most purposes, we may use:
-Eo
k = (A) (e) RT
where
A = collision coefficient
Eo = energy of activation, Btu/lb mol
220
-------
R = universal gas constant = 1.92 Btu/lb mol °R
T = temperature, °Rankine
The constant A is a measure of the likelihood of a molecule of A bumping into one of the
more plentiful B molecules. The exponential function measures the probability that a collision of
A and B will have sufficient energy to bring about a reaction. The energy of activation determines
the temperature level to which the mixture must be raised before the reaction will proceed at a
significant rate.
The function of the catalyst is to reduce the activation energy, and bring about reactions
at a lower temperature than can be accomplished by thermal means alone.
The reaction rate constant plot is ordinarily straight only over a relatively narrow temperature
range. At low temperatures, the catalyst surface tends to be fully occupied by gases and the reaction
rate may be limited by the availability of unoccupied surface. For example, at low temperatures,
oxygen adsorbs on platinum surfaces almost to the exclusion of methane or other light hydrocarbons.
In the middle range, the rate at which the reaction occurs on the metal surface is limiting, and
adsorption and desorption occur with relative ease.
At high temperatures the catalyst surface is likely to remain relatively free of adsorbed gases,
and the difficulty of adsorbing the reactants onto the surface limits the reaction curve. Figure 51
indicates the shape of the reaction rate curve over these three ranges.
As mentioned earlier, oxidation reaction rates of hydrocarbons in the presence of catalysts
are significantly higher than those encountered in thermal oxidation. A comparison of thermal and
catalytic reaction rates for maleic anhydride is presented in Figure 52. From this comparison it can
be seen that catalytic oxidation is instantaneous by comparison at a significantly reduced temperature.
Since catalytic units can operate at relatively low temperatures, they can be designed with
inner walls of aluminized carbon steel. This low construction cost offsets the high cost of the
catalyst contained in the afterburner and overall capital costs of catalytic afterburners run about
the same as other types of afterburners.
Heat exchangers are also frequently used as heat recovery devices with catalytic units.
They may also be designed of carbon steel or aluminized carbon steel. Being relatively less expensive,
they have payouts equivalent to those exchangers used with other types of afterburners. The design
of heat exchangers for catalytic units should follow the same rules outlined earlier for thermal
afterburners.
While the scientific basis of catalysis is not as clear as that of many of the other industrial
221
-------
FIGURE 51
REAC.TION RATE C.ONSTANTS
FOR
LOW, INTERMEDIATE, AND HIGH TEMPERATURES
Id
0
Id
C
y
\
y
o:
1.0
O.I
.01
COMTROL
RATE:
LED
ON
-SURFACE REACTION
RATE CONTROLLED
DESCRPTION
coNT ROLLED;
IOOO 9SO 9OO SSO SCO
TOO 6SO
eoo
TEIMPEIRATURE, * F
222
-------
FIGURE
OOMPARISON OF THERMAL AND C-ATALJlfTJC
REAOTION RATES FOR MAL.EIC ANHYDRIDE
UJ
!c
Z
0
h
0
<
y
it
UJ
j
y
o:
O.I
O.OI
O.OOI
k
\
D C
f) c
<) ff
\
^
) C
) tf
) 0
) C
1 C
J n
L
\
\
>
) C
) I/
J -
—
TI-
RE
^
\
) C
) C
(EIR
AC,
^
C
If
c
r
(VIA
hric
) C
) C
) C
L_
N
) C
) if
) (J
) C
) C
> 0
**,
) (
) u
^ 0
*->»
) C
T C
0 (C
CA"
RE:>
-***-.
*•
) C
> ir
h
rALY'
\CTI
,
^J
) C
) C
^
nc
DN
— -.
^*^>,
C
> d
d
»
i
)
1
D
TEMPERATURE
223
-------
arts, the thermodynamics are sufficiently well known that catalysts can be characterized with good
success. Catalytic afterburners have been widely used in the past for the oxidation of paint solvents,
and emissions from open and closed varnish kettles. Most new installations, however, are using
thermal afterburners. It is anticipated that this trend will reverse itself in the near future due to the
ever increasing shortage of fuel.
II. DESCRIPTION OF EMISSION CONTROL OTHER THAN BEST CONTROL
A. Scrubbers
Wet scrubbers can be relatively effective for the control of phthalic anhydride particulate
and very heavy organics if properly designed. Efficiencies will run between 85 to 95% at pressure
drops ranging from 4 to 8 in. w.c. They have little effect, however, on solvents and other light
and medium weight organics. For the type of material which was measured in the emission curves
given in Chapter 1, scrubbers are not suitable control devices. Source test engineers from air
pollution equipment manufacturers have reported that the typical scrubber installation removes at
most 10% of these contaminants.
Water treatment and/or disposal of waste can be expensive in many cases. With increasing
emphasis on water pollution control, treatment of effluent streams will be required in more and
more localities. The use of chemicals, e.g. permanganate, in the scrubbing solution further increases
cost.
Relatively low cost, as well as safety considerations, have traditionally been the primary
advantages of scrubbers in the resin industry. While it will probably cost more to operate scrubbers
in the future than in the past, their use as pre-treatment devices to remove phthalic and add a
measure of safety before incineration may still be desirable, particularly in fusion cooking.
B. Vapor Condensation
The use of refrigerated condenser systems has sometimes been proposed as an air pollution
control device. A system of this type is in use in at least one location, the Sherwin-Williams plant
in Oakland, California. Sherwin-Williams has supplied design and operating information for their
system so that it can be included in this report.
A schematic of the system is shown in Figure 53. Four kettles are controlled by this system.
Kettles A and B are conventional Dowtherm® heated kettles used for alkyds and polyesters. Kettle A
contains a reflux condenser as well as a final water cooled condenser while Kettle B has only
the final condenser. Kettle C is steam heated and is used only for cutting resin into solvent. Kettle D
224
-------
I
u
f
u
vy
8
2 *
j8h
fcs
^
i
g
h
LJ
^
Z
a
(T
L
0
0
U
y
i
tn
o:
Q
Q
LJ
Ic
o:
ff
L
J!
225
-------
is Dowtherm® heated and used for alkyds and polyesters. While the mechanical details of D are
considered proprietary, it does contain a water cooled final condenser.
The exhaust streams from Kettles A, B and D are controlled by identical two stage refriger-
ated condenser systems. Kettle C requires only a single stage. A separate condenser system is
used for each reactor but all condensers are supplied by a single refrigeration system. The refriger-
ated condensers are used only during solvent cooking.
The refrigeration system is shown in Figure 54. All temperatures shown are to be considered
nominal only. They can vary from run to run over a range of several degrees. Two separate glycol
circulation systems are employed. The "sub-zero" loop is cooled directly by the refrigeration system.
A branch of this loop is used in the cold side of a heat exchanger to cool the "40°F" circuit.
The design basis for this system requires a refrigeration duty of 54,800 Btu/hr. Almost 1/3
of this is due to heat gained in the transfer lines. Two 5 ton refrigeration units are provided. Only
one operates at a time except when cooling down the entire system at start-up. The design was
based on the assumption that all kettles are operating simultaneously. It was further assumed that
each kettle exhaust rate can run as high as 15 cfm noncondensibles and that this gas is saturated
with solvent at the operating temperature of the final condenser (~100°F). The refrigeration unit
has sufficient capacity to handle a continuous discharge exhaust at that level even though the
kettles are operated in such a way that the inert gas flow is intermittent. Furthermore, measurements
by Sherwin-Williams have shown that the final condenser vent averages only 50 to 60% saturation
for these kettles. The 54,800 Btu/hr, then, represents a worst case situation.
A schematic of one of the two stage condenser systems is shown in Figure 55. The exhaust
gas from the water cooled condenser flows to the shell side of the first stage condenser which is
cooled by the "40°F" glycol. From there it flows to the second stage condenser which is operated
at sub-zero. The majority of the condensate is collected in the first stage.
On occasion, the second stage condenser accumulates an excessive amount of frost. To
alleviate this, a system of four-way valves is provided which enables the streams to be switched.
In this case, the kettle exhaust flows first to the second stage which is now cooled by the "40°F"
glycol, and so on.
Kettles A, B and D use systems .identical to that shown in Figure 55. Kettle C uses a single
stage condenser cooled by the sub-zero circuit. Frost formation is not a problem here since no
water is formed in this kettle.
The results of two tests, conducted by an outside consultant, have been supplied by Sherwin-
226
-------
LJ
0
id
h
H 3
D tt
Sc u
U 0
a:
: y
IL N
A
o m
? D
J (/)
1
1
1
u ^
K Z
0 §
Z U
0
F J
RIGERA
C-TRIC/^
IL U
y J
(T y
U
u
QC
0
y
227
-------
u_
0
K)
1
U.I
ol
1
—
'1
L.
1 u.
°o
1 1
-^
i
1
u.
?J
_
228
-------
Williams. A tall oil fatty acid alkyd cook was conducted in Kettle A by the solvent process. Testing
and observation was done over the entire cooking cycle even though no inert gas flow was used
until the last 80 minutes. The results are presented in Table 50. No condensate was collected
until the stripping operation was started. Inert gas rate during stripping was about 8 SCFM.
Light organics were determined by GC analysis of gas samples collected in sample bulbs
at the second stage condenser vents. Heavy organics were determined by GC analysis of gas
samples absorbed in silica gel at the second stage condenser vent and by GC analysis of the
condensate collected in the two condensers.
A solvent process polyester cook was conducted in Kettle D. The results of this batch are
presented in Table 51. The inert gas flow was turned on for a short period of time during the
cook and then turned off. Figure 56 shows cumulative gas vented and cumulative condensate formed
during this run. The results from this run are not typical of those for Kettle D. The large volume of
condensate collected during the fourth hour was due to improper operation of the water cooled
condenser during this time. Consequently, a higher than normal amount of material passed the
water cooled condenser. Inert gas flow rates of about 6 SCFM were observed in the early part
of the run.
Examination of Tables 50 and 51 show that for the alkyd cook on Kettle A, the condenser
system was 84.6% effective on total organics and 85.7% effective on reactives, while for the run
on Kettle D, the efficiencies were 98.4% and 98.9% respectively.
The extremely high efficiencies in the second case are probably due to the fact that the
majority of the condensate was collected while the inert gas flow was negligible. During this period
the vapor entering the first stage condenser was almost 100% condensible. This represents an
optimum situation from the standpoint of condenser efficiency provided adequate refrigeration duty
is available.
In any event, the systems performed well enough in both cases to bring the kettles into
compliance with Bay Area APCD Regulation 3 which requires 85% reduction in reactive organics.
During the test period while both the previously described cooks were taking place, power
consumption by the compressor and glycol circulation pumps ranged from a low of 3,720 watts
to a maximum of 9,150 watts with a time average value of 7,790 watts. The reason for the variation
is that the compressor cycles on and off as required to maintain the coolant temperature. This
represents the primary operating cost of the system. No additional manpower is required as this
system can be attended to by normal staff. The system has not been in operation for a sufficient
229
-------
§ £
o
o
p
d
in r~ co oo en
r^ co r- o O
«-;«-. p p p
ci o cd ci c!
o
r--
cu
4-1
CO
c
cu
•o
8-g
II
T3
C
oo
S
d
p
d
CO
I §
d
c
CD
>
li? to
| c?
a 2
2 co
r s
S
LU
_l
CO
O
O
O
Q
c
cu
ts
o o
CO ^~
P 05
d ^
8 8 °
05 "-. T
,-' O CN
CD
O
CO
CO
CO
S>
o
.c
o>
CO
CO
O)
fi ^
en
I
in
c\i
CO
•*
oo
in
oo
(Q (0
ts =s
« -a
p
d
in in CD
in CN T-
CN «- p
d CN d
S §
d d
ao
o
o
CM
oo
CM
in
c\i
(O
•(-»
4-1
QJ
C
CD
V)
TO
/n
CD
t»
c
o
cu
cu
O)
CO
cu
>
**f
o
4-(
c
8
cu
L_
I
230
-------
TABLE 51
KETTLE D — POLYESTER
Light organics
(Cs, and less)
Toluene
Ethylbenzene
Hexane
Benzene
Light carbonyls
(as HCHO)
Total organic
Water
Total
First Stage
Inlet, Ib/Batch
0.001
7.838
5.375
1.196
0.339
Tr.
14.749
1.663
16.412
First Stage
Condensate, Ib/Batch
6.95
4.80
0.85
Tr.
—
12.60
1.63
14.23
Second Stage
Condensate, Ib/Batch
0.785
0.565
0.240
0.328
—
1.918
0.033
1.951
Total Emission
to Atmosphere
Ib/Batch
0.001
0.103
0.010
0.106
0.011
—
0.231
Gas vented total
Average emission rate,
before control
Peak emission rate,
before control
Average emission rate,
after control
Test period
% reduction, total organics
% reduction, reactives
= 125 SCF inerts
= 1.55 Ib/hr organics (first stage inlet)
= 52 Ib/hr organics (first stage inlet)
= 0.0243 Ib/hr (second stage outlet)
= 9.5 hr
= 98.4%
= 98.9%
231
-------
Q
(0
111
DC
2
L
i
0)
0
UJ
0
g
1
UJ
9 R
if) O
nvj_oj_
232
-------
length of time to determine equipment repair and replacement requirements.
The capital cost of the entire emission control system in 1971 was about $90,000. Estimated
capital cost for a thermal afterburner system, without heat exchange, is $32,000. With a 42% heat
exchange, the capital cost would be about $50,000.
Thermal oxidation has a substantial advantage as far as initial cost is concerned. The operating
cost advantage probably lies with the condenser system, though insufficient information is available
to determine the extent of this at present.
The condenser possesses some features which are attractive in spite of its higher initial
cost. It is an inherently safe control device since no flame or high temperatures are involved. This
is demonstrated by its performance on the polyester cook in Kettle D described earlier. As mentioned
before, this cook was not typical in that the water cooled condenser failed to operate properly for
a time. During this period, instantaneous emissions ran as high as 52 Ib/hr as compared to an
overall average of 1.55 Ib/hr. Such a sudden increase over the expected rate could be disastrous
for an afterburner installation unless elaborate safety devices were installed. The condenser system
handled the increased emission rate without problem.
On the other hand, the condensers are not suitable for use when fusion cooking. Also,
the efficiencies are high only when the gas stream is very concentrated in condensibles. This means
that condensers are impractical for controlling emissions from other parts of a paint plant such as
mixing tanks, filter presses, etc. An afterburner installation can be designed to handle both resin
kettles as well as other sources simultaneously. Perhaps the ideal solution would be to control the
kettles with refrigerated condensers, where safety is a prime consideration, and to control the other
sources by collecting their emissions in hoods and ducts and then incinerating the fumes.
III. METHODS OF CONTROL OTHER THAN ADD ON EQUIPMENT
As is the case with most industrial operation, it is possible to reduce and sometimes eliminate
emissions of pollutants by modification of the manufacturing operation. Various approaches of this
type, applicable to this industry, are listed below.
A. Raw Material Substitution
Raw material substitution in many cases can provide an excellent and inexpensive method
of emission reduction for the manufacturer. Unfortunately, this may not also be true for the user
since raw material substitution can lead to both higher prices and poorer quality. These potential
problems should be remembered when considering this method of emission control. An example,
233
-------
unrelated to air pollution, of product degradation in this industry caused by raw material substitution
is the removal of mercury fungicides and lead drying agents. Industry spokesmen have indicated
it may be as long as three years before an equal raw material substitution may be found and the
quality of paint returned to its former position.
1. Solvents — As discussed earlier, the major emission in the paint manufacturing industry
comes from the solvents used in the various production operations. The quantity of emission is
small, however, when compared to that emitted by the final user.
Current regulation in most states allow for significant emissions of the non-photochemically re-
active solvents described earlier in Section I. Because of this there has been significant pressure to
develop paint products with a non-solvent system, or with a non-photochemically reactive solvent system.
Significant progress has been made by the paint industry in the use of exempt systems through
the development of the water emulsion paints. These products have been developed without sacrifice
of quality or increased cost.
The current development of dry powder systems also appears to offer these same advantages
and may result in a significant reduction of solvent emission from both production and application of paint.
Another technological development which has great potential in reducing solvent emissions,
particularly by the user, is the formulaticn of water reducible industrial coatings. Heretofore, the use of
water based coatings has been restricted almost exclusively to trade sales products. In the opinion of
some experts, use of such industrial coating systems will come to fruition before powder coatings.
Use of high solids systems is a further area for potential improvement. The net result of this
approach is to reduce the amount of solvent present per unit coverage of the applied coating. Conse-
quently, solvent emissions by the user will be reduced.
There has also been a significant switch in the industry to non-photochemically reactive
solvents; and, as might be suspected, this has lead to a reduction of objectionable emission by both
manufacturer and user.
It appears that in many cases, either singly or in combination, these changes in the solvent
system paints have resulted in either or both a higher price product or a poorer quality product.
Some paint manufacturers feel it would be less expensive, overall, for their customer to apply
pollution abatement to existing solvent emissions than it would be to pay the additional product
cost of special paint solvent systems that are currently exempt. This, coupled with possible poor
product quality and the likelihood that these solvents may not be exempt in the future, may make
this raw material substitution a poor choice.
234
-------
2. Pigments and Other Solids — The quantity of pigment emission can be reduced by the use
of liquid slurries, which eliminates the dusting problem associated with handling of the bagged
solid products. The use of water slurries of TiC>2, currently being practiced by many manufacturers,
is a good example of such a raw material substitution This substitution, as well as can be determined,
will not result in additional cost, providing the usage level approaches 500 tons/year
The use of liquid phthalic anhydride in resin manufacturing can also result in reduced kettle
emissions. Introduction of PA into the kettle (as a liquid) eliminates the necessity of evacuating the
kettle when this material is loaded as a solid through the hatch door As described earlier, significant
PA and solvent emission occur during this loading period
The disadvantage to the use of liquid PA is the potential for emissions from the heated
PA storage tank. This can be easily avoided, however, by venting the tank through a water jacketed
vertical condenser with provision for admitting steam to the jacket. The tanks are blanketed with
inert gas during storage with the appropriate provision for pressure and relief control. During filling
the condensers remove PA vapors which can later be melted off and run back into the tanks.
B. Changes in Process or Operating Conditions
As discussed many times, there are two major types of resin cooking, solvent and fusion.
In the solvent cooking process it is possible to maintain a substantially closed system by:
1. Not using sparge gas
2. Use of adequate condensers
3. Use of liquid polyols and acids
4. Use of proper unloading or transfer methods
Operations in this manner can almost completely eliminate emissions from this process.
Continued development of closed kettle cooks is probably the most significant process and operating
change that can be effected.
IV. PERFORMANCE OF CURRENTLY USED METHODS OF EMISSION REDUCTION
A. Performance Data
Data describing the control efficiency of the various air pollution control devices currently
being used by the paint and varnish industry is presented and discussed below. The data was
gathered from the questionnaires used in this study, industry measurement not reported in the
questionnaires and source testing conducted by the EPA.
235
-------
1. Scrubbers — Scrubbers are frequently encountered in the paint and varnish industry. Simple
spray towers are usually installed as part of closed resin kettle systems. As discussed in Section I,
these are not considered pollution control equipment for the purpose of this report. Some perform-
ance data for these devices has been reported in the questionnaires and will be discussed below.
Scrubbers which qualify as air pollution control devices in the present context have also
been reported in the questionnaires. The reported data for these devices has been described in
Tables 52, 53, and 54.
These are predominately spray type scrubbers. Either plain water or water to which some
chemical (e.g., caustic) has been added is used as the scrubbing liquid.
The efficiencies of these devices are reported in the questionnaires to range from 50% to
99% with practically all at 90% or better. Those who report efficiencies for spray towers which
serve as process equipment generally claim efficiencies in excess of 90%.
The accuracy of these reports would seem to be questionable. The efficiency of a well
designed scrubber could be good on relatively large particulate or on readily condensible organics.
On small particulates or relatively volatile organics, the types of scrubbers generally reported would
probably be ineffective. They are inherently limited where volatile, non-soluble contaminants are
concerned. Aside from this problem, they also frequently violate the opacity regulations of most
states.
Furthermore, it appears that in many cases an arbitrary number was picked for efficiency
rather than a number based on actual measurements. Often, the inlet and outlet loadings were
unknown yet a very high efficiency was claimed which seems contradictory.
If large amounts of condensible materials are emitted from a reactor, actual tests could
indicate good efficiency. This could lead to a false sense of security, however, if the analytical
technique ignores the noncondensibles. As reported in an earlier section, actual measurements
have been performed on the noncondensible fraction of the emissions. Levels as high as several
hundred thousand parts per million (CHU equivalent) have been measured. It was further found
that the typical scrubber was at best 10% effective on these emissions.
2. Afterburners — Tables 54 and 55 summarize the questionnaire data reported for thermal and
catalytic afterburners. Efficiencies for thermal afterburner are reported from "poor" to "excellent".
Where a numerical value is given, it is always reported in excess of 90%. On the whole, these
values are probably somewhat more reliable than those given for scrubbers. The inlet loading must
generally be known for design and safety purposes and it is possible to estimate the effectiveness
236
-------
TABLE 52*
TYPE 1 PLANTS
AIR POLLUTION CONTROL — LOADING MILLS, ETC
Paint
Production
MM gal/yr
6.4
2.6
1.0
5.6
1.2
4.3
2.6
3.0
2.4
0.1
1.9
14.9
8.6
8.2
5.6
1.0
0.7
1.0
2.4
2.7
8.3
8.7
1.1
1.0
FABRIC
Number
of Devices
4
1
1
4
1
7
3
1
14
1
1
1
4
1
3
1
3
1
2
1
1
1
1
4
FILTERS
Total Gas
Flow. SCFM
1,000
—
—
6,200
3,600
18,000
5.362
3,250
18,620
—
1,071
—
12,700
5,370
—
—
1,155
2,000
2,600
2,058
—
15.000
6,000
Air to
Cloth Ratio
6.1
—
—
3.5
4.5
3.0
7.7
42
7.0
—
7 1
4,0
62
26
—
—
7.7
67
30
18.3
—
2.9
2.6
'Questionnaire Data
237
-------
Resin
Production
MM Ib/yr
0.4
2.8
3.0
1.2
14.9
2.4
3.6
Number
of Devices
TABLE 52 (Continued)
SCRUBBERS
Total Gas
Row, SCFM
1
1
1
1
1
2
1
Efficiency
2,500
4,500
9,610
6,000
99%
95%
Type
Wetted baffle
Water spray
Roto-clone
Water spray
Spray
238
-------
TABLE 53*
TYPE 1 PLANTS
AIR POLLUTION CONTROL — REACTORS AND KETTLES
SCRUBBERS
Resin
Production
MM Ib/yr
1.9
11.1
0.0
20.0
0.5
5.7
33.7
20.7
5.4
16.4
0.3
17.6
2.9
16.0
3.5
Number
of
Devices
1
4
1
1
3
1
2
1
1
3
3
5
2
3
2
Gas Flow
(cfm) Efficiency" Type
— — —
— — —
— — —
— — KMnO4
— 90% Spray
134 50% —
— — —
— — —
15 85% —
— — Caustic
600 99% Caustic
500 98% —
— — —
2,500 90% —
1,180 — Cyclone
Operating
Cost
($/year)
—
2,000
—
—
15,340
—
—
—
2,600
15,400
175,000
13,300
—
—
1,1700
*Questionnaire Data
**See comments in text concerning reported efficiencies.
239
-------
O)
$
o
co"
>
o
c
co
'o
o
O3
LO
OJ
LO
O)
s
UJ
m
CO
Q.
CO
UJ
Q.
CO
LLJ
t
LLJ
^
Q
Z
CO
tr
0
REACT
1 CONTROL -
O
ID
__J
O
Q.
CC
CO
$ E Ti
CC H 1
CM
—J. f-\
^" ^J
/"™\ —
C/D
CC d_
UJ E u.
Z £°~
DC
8 8
*. °M.
~">
CD
DC
UJ
t -1-
< ^P
O O
0 0
^j ^D
«-" LO"
CD CD
O>
Q.
1-
E E
CO CO
K f
>-
CD CD
CD CD
O O
O)
.E ^
•H ^_| CD
CD (/) 03
* 8 ^
"
§O
CM
O «-
LO
00
o
o
LO
O
O
cn
g 2
CO
DC
UJ
m
m
z>
DC
O
cn
o>
E
D
Z
c
8
DC
4—
O
c
g
'4-1
D
T3
O
ol
|o
LLJ
S3 LI E
0 S
Q
O t.
.E C
If
D_
o
'•p >.
S 52 _ ,
(0 Q. ~ '
O C/5 O
o
o
o
co"
CM
O)
O
O
CM
O
LO
O
O)
o
CO
LO
o
O
O
•^ CO «-
O
CM
CO
CO
CO
CO
00
O
CM
LO
CM
O
CM
O
OO
CD
'o
CD
O
Q.
CD
cn
c
'c
CD
o
o
o
•g
CD
CD
CO
i!
II
CD (0
3. ®
O co
240
-------
TABLE 55*
AIR POLLUTION CONTROL — REACTORS AND KETTLES
THERMAL AFTERBURNERS
Resin
Production
MM Ibs/yr
8.2
38.5
1.9
10.6
0.3
0.2
0.0
3.5
20.7
Number
of Devices
1
1
(flare)
1
1
1
1
1
1
2
Gas Flow
(CFM)
2,000
4,200
2,000
7,000
1,200
500
920
10,000
—
Temp
(°F)
1,500
1,800
1,200
1,100
1,500
1,400
1,200
1,300
1,200
Res. Time
(Sec)
0.5
2.4**
2.1
2.0
—
0.3
—
0.5
—
Efficiency
98%
—
Excellent
Fair
99%
Excellent
—
Poor
—
Operating
Cost
($/Year)
10,000
—
—
—
1,800
370
267
30,600
—
CATALYTIC AFTERBURNERS
Resin
Production
MM Ibs/yr
1.2
0.2
33.0
17.6
16.0
Number
of Devices
1
2
2
1
1
Gas Flow
(CFM)
1,200
—
7,500
2,980
3,000
Temp.
(°F)
1,200
1,000
750
500-1,000
900
Efficiency
Excellent
Operating Cost
($/Year)
4,800
Good —
90% 6,300
Good —
'Questionnaire Data
"This device is a true flare and, as such, residence time is not a measurable parameter.
241
-------
of well designed afterburners by noting the outlet temperature and residence time. Furthermore,
afterburners do not possess the inherent limitations to which scrubbers are subject. Any combustible
material can be incinerated if the temperature and residence time are sufficiently high. The temper-
atures and residence times reported are usually consistent, in a general way at least, with the
efficiencies reported.Beyond that, it is not possible to either affirm or deny the reported performance.
Similar comments apply to the questionnaire returns on catalytic afterburners. Again, if the
temperature is high enough (up to the point at which the catalyst is damaged), enough catalyst is
used and the catalyst is maintained properly, most materials can be oxidized efficiently. The
efficiencies reported ("good" to "excellent", or 90% to 95%) are at least feasible in the light of
the operating conditions reported. Since no information is given on catalyst maintenance or replace-
ment, it is not possible to determine which, if any, of the respondents is optimistic in his reports.
3. Fabric Filters — Fabric filters, usually bag type filters, were the most commonly employed
pollution control device reported in the questionnaires. Tables 52, 56 and 57 summarize the
questionnaire results obtained. They are used for control of dry particulate from loading, mills, etc.
Bag filters are known to attain very high efficiencies (99% + ) even on sub-micron particulate in a
properly designed system. Where reported, efficiencies usually were given as 90% or better. Since
the normal application of fabric filters in the paint industry is for clean, dry particulate at ambient
temperatures, there is no reason to doubt that very high efficiencies can be obtained.
Some additional source test data other than that contained in the questionnaires has been
collected from other plants and source test groups. This data is presented on Table 58.
The Environmental Protection Agency retained a subcontractor to perform tests to determine
the efficiency of thermal afterburners. The test methods were discussed earlier in Chapter 3. The
results of their testing is presented in Table 59.
B. Operating Life and Maintenance Experience for Control Systems
This subject was not well covered by the respondents to the study's questionnaire and the
results presented in Table 60 are based on data supplied from other sources. The Normal Life
listed are based on trouble-free operation and do not consider the past losses caused by explosions
and fires.
In general maintenance is not a significant operating cost. Cost for thermal afterburners
with heat exchangers will run higher than those without. This is caused by tube fouling, tube burn-
out and metal failure due to thermal expansion inherent with the use of metal exchangers. Good
design will keep these maintenance costs to a minimum.
242
-------
TABLE 56*
TYPE 2 PLANTS
AIR POLLUTION CONTROL — LOADING, MILLS, ETC.
FABRIC FILTERS
Paint
Production
MM gal/yr
1.7
2.4
0.5
0.7
1.1
2.5
0.8
1.2
0.5
0.2
0.2
6.0
0.1
1.8
0.6
0.7
2.5
1.3
0.1
0.7
0.6
Number
of Devices
1
11
1
1
2
3
1
1
1
1
2
3
1
1
1
1
1
1
1
1
1
Total Gas
Flow, SCFM
5,000
7,000
—
5,000
3,850
10,500
—
—
2,000
83
1,231
—
—
5,600
—
—
2,376
2,700
875
1,231
—
Air to
Cloth Ratio
2.5
2.0
—
6.3
4.0
2.9
—
—
25.0
25.5
8.2
—
—
7.0
—
—
2.0
9.0
41.7
8.2
—
*Questionnaire Data
243
-------
TABLE 56 (Continued)
Paint
Production Number Total Gas Air to
MM gal/yr of Devices Flow, SCFM Cloth Ratio
2.4 1 1,200 8.0
2.7 1 200 1.8
0.8 2 924 3.5
0.1 1 — —
2.4 4 11,500 —
1.3 1 2,100 —
0.6 2 2,788 —
0.6 2 5,600 20.9
1.2 1 1,200 10.7
1.2 3 18,000 —
244
-------
TABLE 57*
TYPE 3 PLANTS
AIR POLLUTION CONTROL — LOADING, MILLS, ETC.
FABRIC FILTERS
Resin Production, MM pounds/yr 38.0 80.0
Number of Devices 1 3
Total Gas Flow, SCFM 1,900 12,000
Air to Cloth Ratio 2.8 3.9
MECHANICAL
Resin Production, MM pounds/yr 2.6 80.0
Number of Devices 4 4
Total Gas Flow, SCFM 11,000 —
% Efficiency 90 —
'Questionnaire Data
245
-------
g
O
O
T- CN in co
O5 O) O O5
o
CO
CO
in
cb
o"- a)
2 s
CO
o
CN
«- O5 CM |-~ LO ^J-
in in co co T- «-
O
00
CM
CM
O
CO
LU
O
UJ
o
LL
LL
UJ
00
co"
o o
CM CM
co CM
<* oo «- in
CM in in co
o
CM
LU
200%
LU
200
CO
in
LU
m
LLJ
2
fL
O
UJ
O
DC
O
o
CO
UJ
cc
LU
o
OT
LU
co
0
| |
1 I
CD O
DC CO
CD
>;
I—
CO
E
CD
j_
CO
'E
c
O
15
u
o
_i
^r
§
"CO
O
-r*
_co
CO
0
CO
E
CD
H
CD
CO
CO
~3
>
®
^.
C
CD
D
CO
00
c
c
CD
Q.
°0
o
co
a. o
O o
CO
CO
c
'o
c
b
^^
s
u.
CJ
CO
o
T—
^
S:
LL.
U
CO
in
n
^
.c
Q.
O)
in
CM
2
LL
^
§
^
Q.
O)
in
CM
2
LL
CJ
CO
O
00
CD
CO
CD
1
Water cooled
T3
CD
_C
I
8
o
c
o
'co
condenser
CD
1o
CD
"o
Q.
._
Double ventu
•o
CD
.C
i
o
8
0
'co
o
o
CD
'oT
3
5
o
"o
Q.
Recirculating
CO
'c
"CO
O
CD"
o
CD
E
o
O
o
t3
CO
CD
CD
water scrubb
CO
to
D
CO
O
to
O)
o
246
-------
>,
o
O q T- o) i—
•^ d d i-1 T-: d W
.c
^
0
_c
^
X
O
X
o
z
o
z
o
CO
o
CO
o
d
in
o
o
d
1
CM
O
0)
1
1
00
in
o
o
d
oo
CO
0
o
d
O)
co
co
05
CM
0
0
CO
CO
o
o
d
CM
8
d
co
co
CD
00
0
o
in
O)
8
d
in
in
d
0)
o
CM
in
o
0
co
in
d
$
d
m
00
s
o
0
o>
CO
o
o
d
m
CM
d
CM
o
CO
CM
in
o
o
o
CM
in
o
o
d
8
o
d
CO
o
Q
-CO _
-Q co
a
CD
| CO
tr w
S c
Q. 'o>
CO CD
co co
o fo
0 ^
« 0
<
co
in
in N
d r
in in
o o CM CM m
co
in co oo
E «
o £
o co
247
-------
TABLE 60
OPERATING LIFE AND MAINTENANCE REQUIREMENT
FOR AIR POLLUTION CONTROL EQUIPMENT
TYPE
SCRUBBERS
Carbon Steel
Stainless Steel
THERMAL AFTERBURNERS
Without Heat Exchanger
With Heat Exchanger
NORMAL LIFE
YEARS
5
10
15
10
CATALYTIC AFTERBURNERS
Unit Only With & Without Heat Exchanger 15
Catalysts Only 3
FABRIC FILTERS
Unit Only
Bags Only
20
2 to 4
— MAINTENANCE EXPERIENCE —
AMOUNT ANNUAL COST*
High
Medium
Medium
Medium
Low
High
Low
Medium
10%
5%
2%
3.5%
2%
10%
2%
10%
*Percent of initial equipment cost—does not include replacement cost.
248
-------
For catalytic afterburners the maintenance cost for units with or without heat exchangers
are about the same. This is true since these units operate at much lower temperatures and in
either case are built of all metal construction.
Maintenance costs for scrubbers are primarily dependent on the degree of corrosion
experienced. This is primarily a function of the acidity of the scrubbing liquid which will vary sig-
nificantly with the amount of fresh water used. Scrubbers utilizing once-through water will have a
longer life than that listed on Table 60.
V. CAPABILITY TO MEET MORE STRINGENT STANDARDS
A. Fabric Filters
Removal of pigment and other dry particulate is accomplished by use of fabric filters. As
discussed earlier, this control device is capable of efficiencies approaching 99.9 and should have
no difficulty meeting more stringent standards.
B. Afterburners
Either catalytic or thermal afterburners can be operated at increased efficiencies to meet
more stringent air pollution control standards. For thermal afterburners this can be accomplished
by increased operating temperature assuming additional fuel is available. For reactions that are
time limited, efficiencies can be increased by increasing the residence time. This can be accomplished
by reducing gas flow or adding length to the existing unit.
The efficiency of catalytic afterburners can also be increased by increasing the operating
temperature and/or increasing the amount of catalyst used. To maintain higher efficiency it will
probably also be necessary to shorten the catalyst maintenance cycle and the reactivation or
replacement period.
C. Scrubbers
There is little that can be done to improve the efficiencies of the type of scrubbers that
are currently used in this industry. They are fairly effective for removal of large particulate and
heavy condensibles. They are very ineffective on noncondensible organic materials and small
particulate. Changes in operating conditions that would normally improve scrubber efficiency (i.e.
pressure drop, gas to liquid ratio) should have little effect on these emissions. This is typical of
scrubber operation on emissions which have a wide size range of particulate. The large material
is easily collected with any type of scrubber while small particulate and gaseous emission require
a special scrubber operated at very high pressure drops.
249
-------
D. • Refrigerated Condensers
There has been only limited experience in the operation of this device. It appears that
its average maximum efficiency will run around 90%. The possibility of this device meeting high
efficiency standards that might develop in the future is remote.
VI. WATER AND SOLID WASTE PROBLEMS ASSOCIATED WITH BEST CONTROL
One of the many advantages of the use of afterburners for best control is the elimination
of all water and solid waste problems associated with other control methods. In fact, in many
cases the liquid waste solvent discussed earlier in Section III B-6 can be used as a fuel for the
afterburner. This not only solves a waste disposal problem but also reduces afterburner operating
cost.
The pigment dust collected in fabric filters presents no serious solid waste disposal problem.
The quantities are small, non-hazardous, and may be disposed of by scavenger services in land
fill operations. In many cases, they are also recycled back into paint production for use in such
items as dark primers.
250
-------
CHAPTER 6
INSPECTION PROCEDURES
This section is written for a compliance inspector who is familiar with inspection procedures
and testing procedures in general. The best single source of background information for the com-
pliance inspector is the Field Operations and Enforcement Manuals for Air Pollution Control,
Volume I, Organization and Basic Procedures — APTD-1100, and Volume II, Control Technology
and General Source Inspection — APTD-1101. These two volumes are published by the Federal
Environmental Protection Agency and may be obtained from the National Technical Information
Center, Springfield, Virginia 22151. The purpose of this chapter will be to detail the particular
emission problems and control techniques specific to the paint and varnish industry and to help
enforcement officers when entering and inspecting these facilities.
I. NATURE OF SOURCE PROBLEMS
The paint and varnish industry is one of the more complex of the chemical industries.
The three basic manufacturing operations in this industry are varnish cooking, resin cooking and
paint blending. The air pollution regulations most often applied to this industry are the nuisance
regulations covering odor and visible emissions. The other significant problem of this industry
is the emission of fugitive solvent vapors which as yet is not covered by law in most states.
The quantity of emissions released from varnish cooking are small but still present a signifi-
cant local air pollution problem because of the wide variety of highly odorous substances. The
amount of varnish cooking carried out by the industry has been steadily declining and is expected
to continue to decline in the future. This is especially true of the small open kettle batch which
represents the more difficult to control emissions.
Resin manufacturing emissions tend to be highly cyclical in nature and consist of solvent
vapors, phthalic (or other acid) anhydride, polyols (either solid or liquid) and a variety of partial
reaction products. The last category is usually present in the smallest quantities but may represent
the most noxious component from an odor standpoint.
251
-------
Exhaust volumes for a given kettle can peak at up to 100 to 200 SCFM for short periods
of time and may be essentially zero during other portions of the cook. Concentration can vary
from zero up to several hundred thousand ppm.
Every cooking formula will exhibit its own particular emission characteristics. The greatest
quantity of emission occurs from the paint manufacturers operation and consists primarily of solvent
vapors and pigment particles. They are usually fugitive in nature and exit to the atmosphere through
the general building ventilation. Their concentrations are usually quite low (<50 ppm) but the overall
quantity is high due to the large quantity of ventilation air. Future OSHA and air pollution regulations
will probably require the capture of these emissions at their source. This should result in low volume
»
higher concentration emissions which can be more easily controlled.
II. PROCESS DESCRIPTION
The process narratives that follow will be specific to oil based paints and varnishes. The
manufacture of water based paints is of minor importance from an air pollution point of view. This
is also true of the manufacture of the resins used in the blending of water based paints since they
are normally provided outside this industry in chemical process plants.
A. Paint Manufacturing
Starting with all purchased raw material, the manufacturing process for pigmented products
is deceptively simple from a process viewpoint. Basically it consists of mixing or dispersing pigment
and vehicle to give the final product. This is schematically illustrated in Figure 57.
The paint vehicle is defined as the liquid portion of the paint and consists of volatile solvent
and non-volatile binder such as oils and resins. The non-volatile portion is also called the vehicle
solid or film former. The pigment portion of the paint consists of hiding pigments such as titanium
dioxide (TiOa), extenders or fillers such as talc or barium sulfate, and any mineral matter used
for flatting or other purposes.
The incorporation of the pigment in the paint vehicle is accomplished by a combination of
grinding and dispersion or dispersion alone. When it is necessary to further grind the raw pigment,
the pebble or steel ball mills are normally used. With the advent of fine particle grades of pigment
and extenders, as well as the widespread use of wetting agents, the trend is toward milling methods
that are based on dispersion without grinding. This dispersion consists of breakup of the pigment
clusters and agglomerates, followed by wetting of the individual particles with the binder or vehicle.
Some of the more popular methods currently being used are high speed disc impellers, high speed
impingement mills and the sand mill.
252
-------
0
z?u
<-!?
y
tr
DING HATCH
J
r
•^
X
u <
IK h
QL
A
\
\
n
(T l-
1
u
oc
D
0
z
0
(AU
DO.
o°
5?
%
fa
1*
253
-------
Aside from this dispersion step, pigment paint manufacturing involves handling of raw material
as well as handling and packaging of finished product. Operations of a typical plant may be sum-
marized as a raw material and finished product handling problem with a variety of interdispersed
batch operations. The interrelationship of all these operations is schematically illustrated in Figure
58. The operations depicted are those of a plant that makes its own resins and produces both
trade sale and industrial finishes.
Some of the larger and a few of the medium size manufacturers produce a significant
amount of their formulation ingredients, including pigments, resins and modified oils. Certain manu-
facturers produce these ingredients in an amount exceeding their requirements and sell the excess
to other manufacturers. A significant number also produce only a portion of their resins and purchase
the remainder from their competitors or suppliers who specialize in resin manufacturing.
B. Varnish Cooking
The manufacturing of resins and varnishes is by far the most complex process in a paint
plant, primarily as the result of the large variety of different raw materials, products and cooking
formulas utilized. The complexity begins with the nomenclature used in classification of the final
product. Originally, varnishes were all made from naturally occurring material and they were easily
defined as a homogeneous solution of drying oils and resins in organic solvents. As new synthetic
resins were developed, the resulting binders or varnishes were classified on the basis of the resins
used. Examples of this are alkyd, epoxy and polyurethane resin varnishes.
There are two basic types of varnishes, spirit varnishes and oleoresinous varnishes. Spirit
varnishes are formed by dissolving a resin in a solvent. They dry by solvent evaporation. Shellac
is a good example of a spirit varnish. Another material that might fall in this category is lacquer.
Technically, lacquers are defined as a colloidal dispersion or solution of nitro-cellulose, or of similar
film-forming compounds, with resins and plasticizers, in solvent and diluents which dry primarily
by solvent evaporation. Oleoresinous varnishes, as the name implies, are solutions of both oils
and resins. These varnishes dry by solvent evaporation and by reaction of the non-volatile liquid
portion with oxygen in the air to form a solid film. They are classified as oxygen convertible varnishes
and the film formed on drying is insoluble in the original solvent. A summary of the various types
of material used in the production of classical varnishes is given in Table 61.
Varnish is cooked in both portable kettles and large reactors. Kettles are used only to a
limited extent and primarily by the smaller manufacturers. The very old, coke fired, 30 gallon capacity
copper kettles are no longer used. The varnish kettles which are used, have capacities of 150 to
254
-------
i
S
I
255
-------
C
tU
o>
to
01
c
'c
c
!*
to
o
c/>
J3
CO
•(-•
03
O
'>
c J±
< D
c
CD
O)
nj
01
C
c
0)
CO
LU
CO
1—
0)
Q
CM
CD
V3
(/>
c
'tn
CD
CC
CO
in r-
'^ "1
CO |
CO Q.
cu 63
— ^E
s >
i-' csi
0
D)
§
C
cu
CO c/3
3 ^
CO
Z to'
to
O
O
co'
~o
"5
h-
co
to
c
CO
'+J
0
a.
CO
E
'o
to
O
^:
"o
>*
X
*
c
0)
V)
g
tu
CO
.d
E
'E
8
^~
N
iri
t/t
c
cu
"o
0)
1
CD
H
LO
D
CO
CO
manila, boe
eastindia
c
cu
Q.
O
O
-a
c
CO
c
8
cu
CC
o
V)
0>
'o
o
dammar, ac
1
"tu c- '5
o" "tfl —
(/) ^- "*"'
I E .S
c =3
tn u) EC S
~6
C
1
0
"o
" CU
trt Q.
.E 1 -i a -a
8 cc e § o
^ <-> 0 CO T;,
.E S !s> .a g
« £ (0 = t
O *J -g co ~ -
i- r- ^ 4-1 — '
> - .=
CO co' !S co
cu
Zinc resinat
J3
T3
L.
CO
I
D)
C
1-
to"
Oiticica
.0
o
o
•(->
S
1
CO
T3
0
Q
cu
6
C
-------
375 gallons. These are fabricated of stainless steel, have straight sides and are equipped with
three or four-wheel trucks. Heating is done with natural gas or fuel oil for better temperature control.
The kettles are fitted with retractable hoods and exhaust pipes, some of which may incorporate
solvent condensers. Cooling and thinning are normally done in special rooms. A typical varnish
production operation is illustrated in Figure 59.
The manufacturing of oleoresinous varnishes is somewhat more complex than for spirit
varnishes. This manufacture consists of the heating or cooking oil and resins together for the purpose
of obtaining compatability of resin and oil and solubility of the mixture in solvent, as well as for
development of higher molecular weight molecules or polymers.
The time and temperature of the cook are the operating variables used to develop the desired
end product polymerization or "body". The chemical reactions which occur are not well defined.
The resin is a polymer before cooking and may or may not increase in molecular size during the
cook. This resin may react with the oil to produce copolymers of oil and resin or it may exist
as a homogeneous mixture or solution of oil homopolymers and resin homopolymers.
It is possible to blend resins and heat-bodied oil and obtain the same varnish that can be
produced by cooking the resin and the unbodied oils. This indicates that copolymerization is not
the fundamental reaction in varnish cooking.
Heat bodying or polymerization of an oil is done to increase its viscosity and is carried out
in a kettle in a fashion similar to varnish cooking. The fundamental reaction that occurs is poly-
merization of the oil monomers to form dimers with a small portion of trimers.
C. Resin Manufacturing
There is a large variety of synthetic resins produced for use in the manufacture of surface
coatings. A listing of the more popular resins is given below. They are listed by order of consumption
by the coatings industry:3
Alkyd Amino
Vinyl Urethane
Acrylic Rosin Ester
Epoxy Styrene Butadiene
Cellulosic Phenolic
Hydrocarbon
By far the most widely used of these resins are the alkyds and the vinyls. Alkyd consumption
is approximately twice that of the vinyl. Further discussion will concentrate on alkyd resins.
257
-------
0
bJ
tr
D
0
0
0
0
8
0)
0
E
oz
OP
Uin
OP
0
zo
29
I*
258
-------
Alkyd resins comprise a group of synthetic resins which can be described as oil-modified
polyester resins. They are produced from the reaction of polyols or polyhydric alcohol, polybasic
acid and oil or fatty monobasic acid. A listing and discussion of commonly used raw materials
will follow.2
1. Oils or fatty acid
Linseed Castor
Soybean Coconut
Safflower Cottonseed
Tall Oil Fatty Acid Laurie Acid
Tall Oil Pelargonic Acid
Fish Isodecanoic Acid
Tung (minor)
Oiticica (minor)
Dehydrated Castor (minor)
The materials in the first column are oxidizing or drying types. The materials in the second
column are non-oxidizing and yield soft non-drying alkyds which are used primarily as plasticizers
for hard resins. The acids shown in this column are the only materials that are strictly synthetic in
origin.
2. Polyols
Name Formula Form
Ethylene glycol
Liquid
H
HC - OH
HC - OH
I
H
H H H H
Diethylene glycol I I | | Liquid
HO-C-C-0-C-C-OH
II II
H H H H
259
-------
H H H
Propylene glycol | | | Liquid
H-C-C-C-OH
I I I
H 0 H
H
H
Glycerine I
HC - OH
CP-95% glycerine | Liquid
HC - OH
Super-98% |
HC - OH
H
v .2
Pentaerythritol .C White Solid
HOH2C \H2OH
Glycerol or glycerine was the first polyol used for alkyds and is still widely used.
The first polyol, based on usage, is pentaerythritol (PE), which came into common use in
the 1940's. PE is supplied as "technical grade" material and contains mono, di, tri and polypen-
taerythritol. The material consists primarily of the mono form which was illustrated previously in
the list of polyols.
The important distinguishing feature of the various polyols is the number of potentially reactive
hydroxyl groups in the molecule, known as functionality. The glycols with a functionality of two
produce only straight chain polymers and their resins are soft and flexible. The resultant products
are used primarily as plasticizers for hard resins. Glycerine has a functionality of three and is used
primarily in short and medium oil alkyds. Pentaerythritol, with a functionality of four, cross-links to
a greater extent, forming harder polymers. It is ideal for use in long oil alkyds.
260
-------
3. Acids and Anhydrides
Name
Formula
Form
Phthalic
anhydride
(ortho)
White Solid
Isophthalic acid
(meta)
White Needles
Terephthalic
acid (para)
White Crystals
261
-------
Maleic H • j White Solid
anhydride ^Cx /C^
The acidic material can be used as an acid or anhydride. The anhydride is formed from two molecules
of acid minus a molecule of water or removal of one molecule of water from a diacid. It is preferred,
since it reacts faster and yields less water for removal from the cook.
For many years, phthalic anhydride (ortho) (PA) was the only polybasic acid used in sub-
stantial proportions in alkyds. It still remains the predominant dibasic acid. PA is produced from
the catalytic oxidation of naphthalene or ortho-xylene.
The chemistry of alkyd resin systems is very complex. So much so that theoretical con-
siderations offer only a good starting point. Final formula and variations are developed by trial
and error changes, based on performance requirements and shortcomings of previous batches.
Condensation is the reaction basic to all polyester resins, including alkyds. This reaction
follows the elementary equation for esterification as shown below:
.0 o
RC + R1 OH £ R(T + H20
OH \)Rl
Acid + Alcohol = Ester + Water
For Alkyd Resins
PA + Glycerine ?± Ester + H2O
The ester monomer formed is very complex and further reacts to form large polymers
called resins. The polymers formed are low in molecular weight by comparison to other resins. For
example alkyd resins have molecular weights ranging from 1,000 to 7,000 while some vinyl and
acrylic resins have average molecular weights in excess of 100,000 and in some cases as high
as 500,000.
262
-------
The alkyd polymers also react with oil or fatty acid and are generally classified by the
amount of oil or PA used in the formulation, as described below:
% Oil % PA
Short Oil 33 to 45 35
Medium Oil 46 to 55 30 to 35
Long Oil 56 to 70 20 to 30
Very Long Oil 71 up 20
The resulting reactants of the PA, polyol and oil may be represented in part as shown below:
PA G(OH)3
Phthalic Anhydride + Glycerine ->
HO-G-PA-G-PA-G-PA-G-PA-
I I I I
0 PA PA 0 + H90
H H 2
Glyceryl phthalate + water
This will then react with the long chain oil monoglyceride or fatty acid (FA) to yield:
HO - G - PA - G -A - G - PA - G - PA -
I I
PA PA
FA FA
Alkyds can be manufactured directly from a fatty acid, polyol and acid or from oil, polyol
and acid. The second combination (oil, glycerine and PA) produces glyceryl phthalate which is
insoluble in the oil and precipitates. This problem can be overcome by first converting the oil to
a monoglyceride by heating with a polyol in the presence of a catalyst. This process is called
alcoholysis of the oil. The basic reaction is shown below:
H2COOCR H2COH H2COH
HCOOCR -f 2 HCOH >3 HCOH
I I I
H2COOCR H2COH H2COOCR
Triglyceride Glycerine Monoglyceride
This is an ester interchange reaction with no loss of water.
When fatty acid rather than oil is used as the starting material, this is called the "one-stage"
process. In this process, the fatty acid and glycerine are added to the kettle, the agitator is started
263
-------
and heat is introduced. When the batch reaches 440°F, the PA is slowly added and cooking con-
tinued for another 3 to 4 hours until the desired body and acid number are reached.
If the fusion process is being used, a continuous purge of inert gas is maintained to remove
the water formed in the reaction. This water may also be removed by what is known as the solvent
process. It is similar to the fusion process except that about 10% aromatic solvent (usually xylene)
is added to the start. The vaporized solvent is passed into a condenser. The condensate then
flows to a decant receiver for separation of reaction water. Recovered solvent is returned to the
reactor. Inert gas flow in solvent cooking is very low and in some cases is not used.
As discussed earlier, when oil is used rather than fatty acid, the alkyds are produced in a
two-stage process. In the first stage the monoglyceride is first produced from the linseed oil and
glycerol. Catalyst and oil are added and the alcoholysis of the polyol and oil is carried out between
450 and SOOT until the desired end point is reached. When the alcoholysis is completed, any
additional polyol needed is added.
Following this, the required amount of PA and esterification catalyst are slowly added. If
solvent cooking is to be used, the solvent is also added at this time. Cooking then proceeds as before.
A typical manufacturing formula for a 50% oil-modified glyceryl phthalate alkyd using the
two-stage process is given below.
Ib
First stage
Linseed oil 51.3
Glycerol (95%) 12.8
Catalyst, Ca(OH)2 0.026
Second stage
Glycerol (95%) 6.2
Phthalic Anhydride 39.7
Catalyst
Methyl p-Toluene Sulfonate 0.2
110.2
Approx. Loss 10.2
Solids Yield 100.0
Alkyd and other resins are cooked in closed kettles, more properly called reactors. They
vary in size in commercial production from 500 to 10,000 gallons. A typical reactor system is
264
-------
shown in Figure 60. They are generally fabricated of Type 304 or 316 stainless steel with well
polished surfaces to assure easy cleaning. Design pressure is usually 50 psig. These reactors
may be heated electrically, direct fired with gas or oil, or indirectly heated using a heat transfer
media such as Dowtherm®. They are also equipped with a manway, sight-glass, charging and
sampling line, condenser system, weigh tanks, temperature measuring devices and agitator. The
manway is used both for charging solid material and for access to the kettle for cleaning and repair.
The reactor may be equipped with a variety of different condenser systems. The system
shown in Figure 60 includes a packed fractionating column, a reflux condenser and a main condenser.
The condensers are water cooled shell and tube type and may be either horizontally or vertically
inclined. Vapors are processed and condensed on the tube side and drain to a decant receiver
for separation and possible return of solvent to the reactor. A dual function aspirator venturi scrubber
is often added to the system. It is used to ventilate the kettle during addition of solid materials and
may also remove entrained unreacted or vaporized solids and liquids from the venting gases.
Thinning tanks are always included as part of the reactor system. They are normally water
cooled and equipped with a condenser and agitator. The partially cooled finished alkyd is transferred
from the reactor to the partially filled thinning tank. Since most alkyd resins are thinned to 50%
solids, the capacities of these tanks are normally twice the capacity of the reactors. These tanks
are also frequently mounted on scales so that thinning solvents may be accurately added.
The final step in a reactor system is filtering of the thinned resin prior to final storage.
This is normally done while it is still hot. Filter presses are the most commonly used filtering device.
The manufacturing procedures and equipment used for the production of other resins listed
at the beginning of this discussion are quite similar. The major differences are the raw materials
and the process steps utilized. A detailed discussion of these other resins is beyond the scope of
this narrative.
D. Air Pollution Control Techniques
Collection of particulate pigment or resin emission is a simple straightforward job. The
only practical control device is a fabric filter, and it is ideally suited for this application. Collection
efficiency for the submicron pigment dust (0.05 to 0.25 microns) is in the range of 99.9%. There
are no temperature problems since the exhaust system runs at ambient temperatures. The grain
loading is very low and baglife is extensive. Approximately 0.01% of the loaded pigments are lost
and collected. Grain loadings to the fabric filter run around 0.19 grain/SCF. A typical collection
265
-------
-SPRAY TOWER
REFLUX VEMT
/CONDEIXISETR
F RAC T ION AT i NG
DISTILLATION
PORTHOLE
FOR SOLIDS
DIRECT FIRED OR
JACKETED FOR HIGH
TEMPERATURE VAPOR
R LIQUID
TO RES INI
STORAGE
PUMI
FIGURE 6O
MODERN REISIN PRODUCTION SYSTEM
266
-------
system is shown in Figure 61. The collection system can be a fixed hood which can handle both
dust and pigment bags or a flexible hose positioned above the loading hatch or attached to the top
of the tank. The tank attachment provides the most positive control of fugitive dust emission but
also increases pigment and solvent losses slightly.
The application of control equipment to this problem is quite simple and can be solved
with standard off-the-shelf equipment from a host of suppliers.
The control of hydrocarbon and odors from the various emission sources listed earlier is
not quite as straightforward as the dust emission. There are three types of control equipment that
have been applied to this problem. They are catalytic and thermal combustion devices and wet
scrubbers.
As a general rule, wet scrubbing does not provide a satisfactory solution for the following
reasons:
1. Removal efficiency of fine hydrocarbon aerosol is not good at economically practical
pressure drops.
2. Noncondensible hydrocarbon solvent vapors will not be removed.
3. Odor removal without the addition of an oxidizing agent such as potassium permanganate
or sodium hypochorite is unsatisfactory. If an oxidizing agent is used, operating cost
will be quite high due to the high concentration of other oxidizable material such as
phthalic anhydride, resins and oil.
4. Mobile packing and high make-up water rates are required to prevent plugging of the
scrubber beds and spray nozzles.
5. Correction of the air pollution problem with wet scrubbing causes an equivalent water
pollution problem.
The only control technique currently being used that has proven effective for all cases is
combustion. Three general methods are employed to combust waste gases, as shown below.
1. Flame Incineration
2. Thermal Combustion
3. Catalytic Combustion
All of the above methods are oxidation processes. Ordinarily, each requires that the gaseous
effluents be heated to the point where oxidation of the combustible will take place. The three methods
differ basically in the temperature to which the gas stream must be heated.
267
-------
RECOVERED RIOMEMT
RECYCLE FOR DARK
PRIMER PAINTS
(0
LJ
C
3
I
8
0
(f)
u
h
UJ
S3
0.
268
-------
Flame incineration is the easiest of the three to understand, as it comes the closest to
everyday experience. When a gas stream is contaminated with combustibles at a concentration
approaching the lower flammable limit, it is frequently practical to add a small amount of natural
gas as an auxiliary fuel and sufficient air for combustion when necessary, and then pass the resulting
mixture through a burner. The contaminants in the mixture serve as a part of the fuel. Flame inciner-
ators of this type are most often used for closed chemical reactors. They are not used on resin
reactors at present. They may be an ideal solution some day, however, when methods of operating
a closed, pressurized resin reactor are developed.
It is far more likely that the concentration of combustible contaminants in an air stream will
be well below the lower limit of flammability. When this is the case, direct thermal combustion is
considerably more economical than flame combustion. Direct thermal combustion is carried out by
equipment such as that illustrated in Figure 62. In this equipment, a gas burner is used to raise
the temperature of the flowing stream sufficiently to cause a slow thermal reaction to occur in a
residence chamber. Whereas flame temperatures bring about oxidation by free radical mechanisms
at temperature of 2500°F and higher, thermal combustion of ordinary hydrocarbon compounds
begins to take place at temperatures as low as 900 to 1000°F. Good conversion efficiencies are
produced at temperatures in the order of 1400°F with a residence time of 0.3 to 0.6 seconds.
Catalytic combustion is carried out by bringing the gas stream into intimate contact with a
bed of catalyst. In this system, the reaction takes place directly upon the surface of the catalyst
which is usually composed of precious metals such as platinum and palladium. While thermal
combustion equipment brings about oxidation at concentrations below the limits of flame combustion,
catalytic combustion operated below the limits of flammability and below the normal oxidation
temperatures of the contaminants. The reaction is instantaneous by comparison to thermal combustion
and no residence chamber is required. Catalytic combustion is carried out by equipment such as
that illustrated in Figure 63.
In general, catalytic afterburners are less expensive to operate, however, they depend
directly on the performance of the catalyst for their effectiveness. It will not function properly if
the catalyst becomes deactivated. Because of this, catalytic units are not inherently functional when
operated at design temperature. In many areas, means for ensuring adequate performance of the
catalyst on a long term basis will be required by environmental control offices.
The basis for design of either catalytic or thermal combustion is the hydrocarbon concentration
of the exhaust gases handled by the afterburner. The maximum hydrocarbon level is set by most
269
-------
I]
270
-------
271
-------
insurance companies at one-quarter of the lower explosive limit (LE.L.) which is equivalent to 13
Btu/SCF of exhaust gas. Once the rate of emission is determined, it is then necessary to calculate
the dilution air required to meet 1/4 L.E.L. and set up the ductwork system to provide for this dilution.
When possible, dilution air should be utilized to help capture as many fugitive fume emissions as
possible. For example, this can be accomplished by taking the dilution air from a hood positioned
over the resin filter press and venting the thinning tanks and product run-down tanks into the same
system.
The major problem with catalytic or thermal afterburners as applied to open or closed
resin and varnish kettles is the danger of fires and/or explosions. This has happened in numerous
occasions in the past due primarily to excessive hydrocarbon emission from kettles. These problems
have been all but eliminated on newer units by assuring that the design was based on actual
emission measurements of the highest emitting cook and the addition of some of the following
system safety features:
1. High limit temperature alarm to shut off burner and activate a diversion system.
2. High velocity duct section to assure gas flow to afterburner substantially exceeds flame
propogation velocity of hydrocarbons being burned.
3. Double manifolding or hot gas recycle to prevent condensation of heavy hydrocarbons
or phthalic anhydride.
4. Diversion system to block off hydrocarbon emissions to the unit, by-passing them directly
out a separate exhaust, and introduction of fresh air to purge the unit.
5. Pneumatic operation of the diversion system to assure fast positive action and provide
a fail-safe system in the event of either air or electrical failure.
6. Purging with inert gas in the event of power failure.
III. INSPECTION POINTS
There are few, if any, federal or state air pollution control regulations dealing specifically
with the paint and varnish industry. General regulations that affect the industry are the opacity
regulation and odor nuisance laws. Solvent emissions is the other area of potential problems.
Most plants are quite likely to be affected by regulations such as Rule 66 of the LOS ANGELES
COUNTY AIR POLLUTION CONTROL DISTRICT or Regulation 3 of the SAN FRANCISCO BAY
AREA AIR POLLUTION CONTROL DISTRICT.
In preparing for an inspection of a paint and varnish plant, the air pollution inspector should
272
-------
be familiar with:
1. The conduct of an odor survey
2. Opacity observation
3. Classification of photochemically reactive solvents
His main task during the inspection will be:
1. Determination of visible plumes and boundary line odor levels
2. Review of process flow sheet
3. Inspection of production facilities
4. Determination of source testing requirements
Determination of compliance of the odor nuisance regulation is difficult. Consistant previous
odor complaints are a good indication that a plant may be out of compliance in this area. If possible,
the inspection should be arranged when the wind is blowing toward the area from which the complaints
have occurred in the past. In any event, determination of odor nuisance should be made down
wind from the plant before the on-site visit, since exposure to high levels inside the plant will dull
sensitivity to lower odor levels. A listing of odor thresholds for material likely to be used by a paint
and varnish plant was given previously in Table 28 and is presented again on the following page.
Inspection for visible stack plumes should also be made at this time. Any violation can
then be further detailed during the plant inspection.
Prior to the plant inspection, an attempt should be made to meet with the plant engineer
and obtain and/or develop a process flow sheet listing all potential emission points. This will aid in
conducting a proper on-site inspection and a good emission inventory.
The next step is the on-site inspection of the facilities. This can be divided into the following
three areas:
1. Raw material handling and storage
a. Liquids
b. Dry materials
2. Manufacturing
a. Resin manufacturing
b. Varnish cooking
c. Paint blending
3. Filling, packaging and product storage
273
-------
TABLE 28*
ODOR THRESHOLDS OF SOME ORGANIC VAPORS
Chemical
Acetaldehyde
Acetone
Acrolein
Benzene
Ethanol
Ethyl acrylate
Formaldehyde
MEK
Methanol
Methyl methacrylate
Methylene chloride
Phenol
p-Xylene
Styrene (inhibited)
Styrene (uninhibited)
Toluene
Odor threshold, ppm
0.21
100.0
0.21
4.68
10.0
0.00047
1.0
10.0
100.0
0.21
214.0
0.0470
. 0.47
0.10
0.047
2.14
*Air Pollution Control Assoc. Journal, Volume 19, Number 2, Feb. 1969, pages 91 to 95
274
-------
A. Raw Material Handling and Storage
Liquid raw materials handled in bulk quantities will be considered first. Oils and polyols
are not very volatile and can sometimes be neglected as emission sources at this point in the
process. Solvents (and materials, such as resin solutions, that contain solvents) are the major
potential vapor emission sources in the handling and storage stage of the plant operation. Even
in this category many substances can be neglected. Many parts of the country have no regulations
concerning non-photochemically reactive organics. In those localities, tanks containing such material
can often be neglected if there is no apparent odor nuisance. Likewise, many localities exempt
solvents having vapor pressures below some specified value. These solvents can usually be
neglected on this basis. The size (capacity), contents, venting arrangement and frequency of filling
of all other tanks should be noted. Particular attention should be given to the solvents known to
have a high odor potential, such as isophorone, styrene, ethyl acrylate, etc.
The most important solvents that must be considered when determining compliance in most
parts of the country are the relatively volatile, photochemically reactive organics. A listing of solvents
commonly used in paint and varnish manufacturing is given in Table 62. Of these, the ones most often
used in sufficient quantities to justify bulk storage are toluene, xylene and the lower boiling aromatic
naphthas. The pumping rates for these solvents should be determined so that displacement losses
can be calculated. For example, for toluene at 20°C, a pumping rate of 144 gal/min or more will
give emission rates in excess of 8 pounds/hr which would violate allowable emission rates in some
states. Filling losses for selected solvents is given in Table 63. Vapor pressure is strongly temper-
ature dependent so that a solvent that can be neglected in a cold climate may have to be considered
where ambient temperatures tend to be higher.
Dry materials are usually received and stored in bags or drums so they normally present
no pollution potential until dumped during the manufacturing process. The same is true of liquid
materials received and stored in cans and drums. Any area where mechanical transfer or handling
(such as sifting, blending, etc.) of solid powders takes place in contact with air should be noted
as a potential emission source.
B. Manufacturing
Attention can now be directed towards the manufacturing area. It is necessary to locate
and identify all vents serving the equipment in this area. As before, in many localities, a distinction is
made between photochemically reactive and non-photochemically reactive organics with the latter
often exempt from the regulation. The potential for odor nuisances for non-reactive materials should
275
-------
TABLE 62
CLASSIFICATION OF TYPICAL SOLVENTS
Vapor Pressure @ 20°C
Solvents (mm Hg)
Status*
Alcohols
Methyl
Ethyl
Isopropyl
n-Butyl
sec-Butyl
Ketones
Acetone
MEK
MIBK
Isophorone
Esters
Ethyl acetate
Isopropyl acetate
n-Butyl acetate
Amyl acetate (mixed)
Ethers (and derivatives)
Ethylene glycol monoethylether
Ethylene glycol monoethyletheracetate
Aromatics**
Xylene
Toluene
Other
2-Nitropropane
Methylene chloride
Trichloroethylene
97.3
43.9 '
32.8
4.3
12.5
186.0
70.2
15.0
0.18
74.4
43.2
10.0
3.8
0.1
2.0
7.1
22.0
12.9
360.0
65.0
exempt
exempt
exempt
exempt
exempt
exempt
exempt
not exempt
not exempt
exempt
exempt
exempt
exempt
exempt
exempt
not exempt
not exempt
exempt
exempt
not exempt
These solvents are classified according to a LAAPCD Rule 66 type of definition.
"Large quantities of various mineral spirits and napthas are consumed. Some of these contain
sufficient quantities of aromatics to be classified as "reactive". Each case must be considered
individually, however. This is also true of vapor pressures.
276
-------
TABLE 63
FILLING LOSSES FOR SELECTED SOLVENTS @ 20°C
Solvent
Acetone
Ethyl Acetate
Toluene
Mineral Spirits
V.P. @ 20°C, mm
186
74
22
2 (est.)
Mol. Wt.
58
88
92
160 (est.)
Filling Loss,
lb/100gal
0.494
0.299
0.0927
0.0147
SAMPLE CALCULATION
Material: Acetone
Filling loss = 186mm x
760 mm
1
Molar volume @ 20°C=24 liter/g-mole
x 100 gal x 3.79 liter/gal
24 liter/g-mole
x 58 gram/g-mole
1
454 g/lb
= 0.494 Ib
277
-------
not be overlooked. Where this is true, attention can be concentrated on those vents having photo-
chemically reactive or odorous emissions. Regulations can change, of course, so that in the future,
regulation of all organic emissions may become typical, or more refined measurements of reactivity
may lead to different classifications.
The resin plant represents by far the most difficult area in which to determine compliance.
The processes are so cyclic in nature and industry practice so diversified that there seems no
alternative other than a continuous (or, at least, a semi-continuous) emission monitor of an entire
cook. It should be noted that two vents are involved in a solvent cook and one in a fusion cook.
The thin tank condenser vent and, where applicable, hood vents over the resin filter presses are
other points to be considered in the resin plant. A portable anemometer capable of measuring very
low velocities and a portable total hydrocarbon detector (both explosion proof) are helpful in determin-
ing compliance for these operations. Where separate standards exist for toxic materials (such as
isocyanates from polyurethane production) it may be necessary to use special techniques for
measuring the concentration of such materials. There is no adequate substitute for a competently
conducted source test over the period of the cook to establish compliance for a resin operation.
Operations in the paint plant are relatively easier to monitor. If the only ventilation in these
areas is the general building ventilation system then, in many localities, no control regulations
apply. Likewise, if no reactive organic solvents or solvents with pronounced odor characteristics are
used, regulations may not apply. The use of a weight balance calculation to determine emissions
is not recommended in that a small difference between large numbers is usually involved. Where
emissions are collected in ducts and exhausted at one or more point sources, they should be investi-
gated for compliance. In paint plants, particulate as well as organic emissions must be considered.
One system often employed in paint plants is to collect emissions (solvent vapor as well as particulate)
in ductwork, pass the gas stream through a particulate collector (usually a fabric filter type) in
which pigment solids are removed, and exhaust through a stack to the atmosphere. The inlet and
outlet loadings for this particulate collector ordinarily need not be measured if the filter appears to
be intact and the organic concentration at the stack determined. Much higher flows are encountered
here than in the resin plant, and conventional source velocity and concentration measurement tech-
niques are likely to be satisfactory.
Types of emission control devices encountered in the paint and resin industry include scrubbers,
fabric filters and thermal and catalytic afterburners. The typical scrubber installation of the type
usually found on resin kettles set up for fusion cooking tends to be rather ineffective on all but large
278
-------
particulate and heavy condensible materials. Furthermore, such scrubbers can sometimes be the
source of visible plumes. Where these devices represent the only pollution control equipment in
a given resin plant, some consideration might be given to source testing the plant, particularly if
severe odor or visible emission problems exist.
Fabric filters probably represent the only practical reliable means of collecting pigment dust
emissions applicable to most paint manufacturing facilities. Electrostatic precipitators would also be
effective but so far as is known they find little if any use in the paint industry due to cost considerations.
They are sometimes used in pigment manufacturing but in that case their primary purpose is to
collect acid mist.
Unless there is some evidence to the contrary, fabric filters operating at air to cloth ratios
of 2 to 3 can probably be assumed to comply with most existing air pollution regulations (based
on weight only) provided bag integrity is maintained (this is not to imply that higher air to cloth
ratios may not also be effective). Regulations aimed specifically at the submicron respirable particulate
size range may require further investigation of filter performance, however.
Thermal and catalytic afterburners are the most effective control devices for organic emissions.
Thermal afterburners can also be effective for combustible particulate, such as phthalic anhydride,
though the proper operating conditions may be considerably different than those for vapors.
Where afterburners are used as emission control devices, it is often possible to determine
compliance by inspecting the system. It must first be determined whether all possible sources of
organic vapors are vented into the afterburner inlet. For thermal afterburners, a determination that
both temperature and residence time are acceptable is sometimes sufficient to establish compliance.
Temperatures of 1250 to 1450°F and residence times from 0.3 sec to 0.6 sec are commonly
acceptable. For catalytic afterburners, compliance is somewhat more difficult to ascertain. Proof that
a regular catalyst maintenance schedule has been adhered to as well as a measurement of operating
temperature, or temperature rise across the catalyst bed, can suggest that a unit is in compliance,
but only a measurement of discharge hydrocarbon concentration can confirm adequate performance.
C. Filling and Packaging
Some of the same comments apply to the filling and packaging areas as apply to the paint
manufacturing area. Solvent vapors represent the principle emissions from this area. Any exposure
of solvent-containing materials to air should be viewed as a potential source. Low reactivity, lack
of odor nuisance potential or measurement of low emission rate are the grounds for demonstrating
279
-------
compliance.
A complete compliance survey should identify as potential sources all of those points at which
process equipment is vented to the atmosphere, or at which the solid or liquid materials being
processed are exposed to the air. These points are most conveniently noted on a flow diagram
of the process in question. Each such point should be noted as:
1. Likely to be in compliance by virtue of the materials being handled having little tendency
to become airborn, or
2. Being exempt from regulation as non-reactive, non-odor-causing vapors, or
3. Having measured emissions low enough to be in compliance or having adequate
collection and/or disposal equipment for the particular materials involved.
Many potential sources may be found to be in compliance by virtue of face-value evidence
of the types indicated. Other potential sources should be considered suspect and accepted source
testing methods used to establish compliance.
280
-------
CHAPTER 7
ECONOMICS OF EMISSION CONTROL
Specific among the goals of this study was the determination of financial impact of air
pollution control on the paint industry. In order to accomplish this goal, the effort was subdivided
into three parts as follows:
I. Cost of Best Control Equipment
II. Model Plant Study
III. Industry Wide Studies
These elements of the study are covered in the following pages.
I. COST OF BEST CONTROL EQUIPMENT
In order to develop current and accurate costs for best control as applied to the paint and
varnish industry, a number of equipment manufacturers and the Industrial Gas Cleaning Institute
were retained as subcontractors to furnish cost information.
The Industrial Gas Cleaning Institute (IGCI) is an association of air pollution equipment
manufacturers and represents a majority of the larger suppliers to the marketplace. The IGCI
has had considerable experience in supplying information of the type required in this study to the
Federal Environmental Protection Agency in a number of "Air Pollution Control Technology and
Costs" contracts. The IGCI was retained as a consultant in order to select the member companies
most qualified to provide the required cost information. In addition to selecting companies, the
Standards Committee of the IGCI was used to review and approve all technical information supplied
in the study.
The IGCI member companies and equipment manufacturers selected were each supplied
the following information:
1. Process narrative for the paint and varnish industry
2. Specifications for abatement equipment
3. Instructions for submitting cost data
281
-------
4. Process descriptions for thermal and catalytic afterburners for a resin reactor and for
an open kettle
5. Operating conditions for thermal or catalytic afterburners (with and without heat
exchange) for large and small closed kettles; and (without heat exchange) for large
and small open kettles
6. Data forms for estimated capital and annual operating cost
The process narrative and cost data forms (items 1 and 6 above) are included in this report as
Appendix B and Appendix C, respectively. Other information supplied above (items 2 thru 5) follow
as Tables 64 to 81 which can be found at the end of this chapter.
Based on information supplied, participating IGCI member companies furnished equipment
bid information and estimates of erection or installation of the system required. According to instruction,
the erection or installation estimates were based on a Milwaukee, Wisconsin location or an alternate
city with a construction cost index near the national average. Therefore, the installation cost infor-
mation presented was adjusted based on the cost indices presented in Table 82. This information
was taken from Building Construction Cost Data 1970* which presents a construction cost index
for 90 cities using 100 to represent the national average. The indices presented are for the building
trades only and are used as representative for general rates on field construction. Since the indices
do not account for differences in labor productivity, the cost variations among cities may be
understated.
Direct operating cost information was also provided for the best control equipment by the
participating IGCI member companies. These costs were developed using estimated requirements
and the following basis:
Units Unit Price
Operator
Supervisor
Labor
Maintenance materials
Replacement parts
Electric power
Fuel
Water (process)
Water (cooling)
Chemicals
Dollars/hour
Dollars/year
Dollars/year
Dollars/kwh
Dollars/MM Btu
Dollars/M gal
Dollars/M gal
Dollars/year
$6.00
$8.00
Cost**
Cost**
$0.011
$0.80
$0.25
$0.05
'Published by Robert Snow Means Company.
"See Appendix C. 282
-------
With this information, annual operating costs were developed using direct costs estimated on the
preceding plus an annual capital charge.
Cost information for fabric collectors was developed using equipment quotations and
operating conditions obtained directly from equipment suppliers. In this manner, information was
developed using a single air rate for the pulse test and shaker type fabric collectors. With this
information, Air Resources, Inc., developed installation and operation cost estimates using the same
bases as for the IGCI work. The estimate information is presented on the following page.
Capital and operating costs presented herein are based on 1972 dollar values. Caution
should be exercised in using the cost information presented as a basis for preliminary estimates
of air pollution control cost. Dollar values may require adjustment due to inflation and labor rates
and productivity may require adjustment to prevailing conditions.
A. Thermal Afterburners
Process descriptions were developed as follows:
1. Thermal afterburner process description for resin reactor specification
2. Thermal afterburner process description for open kettle specification
Operating conditions were developed for the resin reactor case with and without the use of heat
exchange. Operating conditions for the open kettle specification were developed without heat
exchange.
The process description for Resin Reactor Specification is presented in Table 66. Related
operating conditions of afterburners without heat exchange are presented in Tables 67 and 68.
Similar conditions with heat exchange are presented in Tables 69 and 70.
The Process Description for Open Kettle Specification is presented in Table 71. Related
operating conditions of thermal afterburners without heat exchange are presented in Tables 72
and 73.
Capital cost information shown in Figures at the end of this chapter was based upon the
specification bid information as previously described. Capital cost information for thermal afterburners
without heat exchange is presented in Figure 66. Similar cost information with heat exchange (42%
efficient) is presented in Figure 67. Total installed costs for thermal afterburners are presented in
Figure 68 and a comparison is presented with catalytic afterburners. This figure also compares
systems with and without heat exchange.
Annual operating costs were developed using methods described previously. Figure 69
presents direct annual operating costs for thermal afterburners, without heat exchange, and also
283
-------
includes a comparison with catalytic afterburners. Similar information with heat exchange included
is presented in Figure 71. Figure 70 presents total annual operating cost for thermal afterburners
without heat exchangers and also includes a comparison with catalytic afterburners. Similar infor-
mation with heat exchangers is presented in Figure 72.
B. Catalytic Afterburners
Process descriptions were developed for catalytic afterburners as shown:
1. Catalytic afterburner process description for resin reactor specification
2. Catalytic afterburner process description for open kettle specification
Operating conditions were developed for the resin reactor case with and without heat exchange.
Operating conditions were developed for the open kettle without heat exchange.
The process description for resin reactor specification is presented in Table 74. Related
operating conditions without heat exchange are presented in Tables 75 and 76. Similar conditions
with heat exchange are presented in Tables 77 and 78.
The process description for open kettle specification is presented in Table 79. Related
operating conditions without heat exchange are presented in Tables 80 and 81.
Capital cost information was developed based upon specification bid information as
previously described. Capital cost information for catalytic afterburners without heat exchange are
presented in Figure 64. Similar cost information with heat exchange (23% efficient) is presented
in Figure 65. Total installed costs for catalytic afterburners are presented in Figure 68 and a compari-
son is presented with thermal afterburners. This figure also compares systems with and without
heat exchange.
Annual operating costs were developed using methods described previously. Figure 69
presents direct annual operating costs for catalytic afterburners, without heat exchange, as well
as comparative data for thermal afterburners. Similar information, with heat exchange, is presented
in Figure 71.
C. Fabric Collectors
Cost of fabric collectors was estimated using a single case for pigment recovery. This
was based on the following operation conditions:
Air handling rate, CFM 3,800
Pigment recovered, Ib/yr 2,850
Using these conditions, air to cloth ratios were selected and cost information developed. This infor-
mation is presented in Table 84.
284
-------
II. MODEL PLANT STUDY
In order to develop current and accurate cost information for the paint and varnish industry,
Air Resources, Inc. enlisted the assistance of the Sherwin-Williams Company as a subcontractor.
The Sherwin-Williams Company is a large multiplant company in the paint and varnish industry
and possesses a thorough knowledge of investment and operating costs for plants of the type
considered herein.
As a basis for estimating costs for the Model Plant the following were established:
1. The plant site consists of a hillside location approximately 37 miles northwest of Chicago
and in or around the city of Elgin, Illinois.
2. Land costs are not included in the estimate.
3. All required utilities are available at the property line.
4. Uniform Building Code, OSHA and National Fire Protection Association regulations
are applied to the plant.
5. Construction costs are based on December, 1972 for the Chicago, Illinois, metropolitan
area.
Using the above bases, capital costs were developed for the model plant along with balance sheets,
operating statements and financial impact of control. These are discussed below.
A. Capital Cost of Plant
Using bases described previously, an estimate of capital investment was developed for the
Model Plant. The estimate was based on an item-by-item cost breakdown which is presented in
Appendix A of this report. The specific equipment selected for inclusion in the estimate is believed
to be representative of that which might be installed in a plant of the model type. Certain of the
equipment was selected based on accessability of information and does not necessarily represent
the optimum or the lowest cost equipment that could be utilized. Inclusion of specific equipment
items should not be considered to be an endorsement of any brand or model.
The total estimated cost for the Model Plant is $3,755,000. A summary of the cost items
is presented in Table 95. In addition, Table 96 presents a summary of equipment and utility costs.
B. Balance Sheet and Operating Statement
In developing financial information for the uncontrolled plant, an attempt was made to
utilize a realistic approach while at the same time trying to avoid unnecessarily complex handling
and accounting situations. In order to simplify accounting, the model plant was assumed to be
fully operational with start-up costs and investment credits having expired in a prior period. Labor
285
-------
availability, specific geographical raw material supply and geographical competitive market forces
were not considered. All data is based upon the model with operations on a two-shift basis.
With these bases, the following information was developed:
Table 85 — Annual Raw Material Requirements and Estimated Annual Costs
Table 86 — Annual Package and Package Material Requirements and Estimated Annual Costs
Table 87 — Annual Wage and Salary Estimated Costs
Table 88 — Annual Depreciation Costs
Table 89 — Income Statement Year Ending December 31, 19 2
Table 90 — Annual Production Schedule Detailed by Month Showing Sales and Inventory Levels
Table 91 — Balance Sheet at Year-end, December 31, 19_2
Table 92 — Cash Flow Statement for Year 19_2
Table 93 — Return on Investment Year 19 2
Table 94 — Annual Product Mix
Figure 73 — Factory Manning and Organization Chart
Discussion of the schedules are presented on the following pages.
Raw Materials (Table 85)
This schedule details the raw material requirements necessary to produce the product mix
included in Table 94. The raw material prices used in this schedule are based upon vendor
quotation and include delivery charges.
Packages and Associated Materials (Table 86)
Table 86 details the packages and associated materials to satisfy the product mix and the
packaging schedule for the model plant.
Wages and Salaries (Table 87)
Table 87 establishes the wage and salaries estimates for the Model. These requirements
were based upon the organization shown in Figure 72. While the salaries assumed may represent
a very subjective viewpoint, these are thought to be representative. The hourly wage rate was
calculated by using the average rate for SIC 2851 for 1971 from the Statistical Abstract of the
United States, with a factor of 10 per cent added. For second shift requirements a shift differential
of 100/hr was added. FICA taxes are calculated using current employer percentages. Fringes exclusive
of FICA were estimated at 5 per cent of wage and salary cost.
Depreciation (Table 88)
Table 88 presents the depreciation costs associated with the model plant. The straight line
286
-------
method was used for calculating depreciation without deducting any salvage costs. The buildings
were depreciated over an 80 year life and the equipment over a 10 year life.
Income Statement (Table 89)
This table was prepared based on several assumptions and estimates which include the
following:
1. Selling price was developed using an average from the Current Industrial Reports M28F
for 1971. The average from this report was increased 4 per cent for industrial shipments
to adjust for growth rise.
2. Estimates were made for "other manufacturing costs" of 500/gal on trade shipments and
300/gal on industrial shipments. Industrial shipment costs are estimated lower since
less operating cost is associated with the equipment and filling of large industrial packages.
3. The estimate used for selling, general and administrative expense are directed primarily
at selling expense. For trade sales an estimate of 15 per cent of sales was used,
and for the industrial products only 5%. The differences in these percentages reflect
alternate levels of effort required by a small industrial manufacturer to service a specific
industry (or at best several small manufacturers) compared to servicing a large number
of retailers.
In addition, taxes, both federal and local, were estimated and these percentages are shown in
the schedule.
Production Schedule (Table 90)
The table presents a detailed estimate breakdown by months of production, shipments and
inventory. Also included is a safety reserve amounting to approximately 5 per cent of annual
shipments.
Balance Sheet (Table 91)
In preparing the balance sheet, liabilities were recognized for taxes and accounts payable,
but such items as prepaid expenses were specifically avoided.
Inventory accounts were developed using an average cost basis and the gallonage shown
in ending inventory in Table 90 for December 31, 19_2.
Plant financing was assumed to consist of a capitalization of one million dollars and a
bank loan of four million dollars over a ten-year term at 7 per cent interest.
The balance sheet shows negative cash which would indicate the need for at least seasonal
borrowing currently. This solution would become less acute as interest expense decreases with
287
-------
long term debt.
Cash Flow Statement (Table 92)
A simplified cash flow statement is presented which assumes that no change occurs in
accounts receivable, accounts payable and in inventory levels between years.
Return on Investment (Table 93)
Table 93 presents returns on investment calculations using four separate calculation
techniques. These include the total gross assets available method, the total net assets available
method, stockholders equity plus long term debt method and the stockholders equity method.
Product Mix (Table 94)
Table 94 shows the product mix assumed for model plant. Fifty per cent of the industrial
output was selected by the subcontractors in preparing the table. Considering the size of the plant
and the associated difficulties of manufacturing industrial finishes, the following mix was established:
Gallons Name
1. 95,000 Monomer modified alkyd for fast dry coatings
2. 95,000 Acrylic baking enamel
3. 190,000 Alkyd urea baking enamels
380,000 50% of industrial output
C. Balance Sheet and Operating Statement for Controlled Plant
The balance sheet and operating statement for the controlled plant were prepared using
the same assumptions and basis which were documented previously for the controlled plant. That
financial information which is changed for the controlled plant solution is as follows:
Table 97 — Annual Depreciation Costs Including Air Emission Control Devices
Table 98 — Income Statement Year Ending December 31, 19—2 adjusted to show the
effect of air emission control devices
Table 99 — Balance sheet at year-end, December 31, 19 2 adjusted to show the effect
of air emission control devices
Table 100 — Cash flow statement year 19 2 adjusted to show the effect of air emission
control devices
Table 101 — Return on investment year 19 2 adjusted to show the effect of air emission
control devices
Table 102 — Financial information on recommended air emission control devices
Each table above is presented in respective order at the end of this chapter.
288
-------
D. Cost of Control Other Than Add-on Equipment
The high costs associated with air pollution control in the paint industry invites consideration
of alternative means for reduction of emission levels. Significant among these alternatives is the
potential for raw material substitution and/or process modification.
1^ Raw Material Substitution — Appreciable reduction in emissions of air pollutants can be
achieved when it is possible to substitute raw materials with compounds which from an air pollution
standpoint offer more desirable physical and chemical properties. The most beneficial choice for
substitution is the solvent base of the paint since such materials represent the major source of
emissions from paint plants. Paramount among the possibilities for solvent substitution is the use of
water or the conversion to water base coatings. Where water based formulations can be developed
to satisfy any of a variety of end uses, an order of magnitude of emission reduction is possible in
the paint industry. This is not always easily accomplished since product quality must be maintained.
Significant progress has been made in the paint industry through the development of water
emulsion trade sales paints. Many of these developments have been achieved without increase in
cost or sacrifice in quality.
Current development work is underway on the use of dry powder paint formulations. Where
such formulations will fulfill the end-use requirements, significant emission reduction is possible in
not only production, but in paint application as well.
Another technology development which has great potential in reducing solvent emissions,
particularly by the user, is the formulation of water reducible industrial coatings. Heretofore, the use
of water based coatings has been restricted almost exclusively to trade sales products. In the opinion
of some experts, use of such industrial coating systems will come to fruition before powder coatings.
Use of high solids systems is a further area for potential improvement. The net result of
this approach is to reduce the amount of solvent present per unit coverage of the applied coating.
Consequently, solvent emissions by the user will be reduced.
Good potential exists for reduction in the quantity of undesirable emissions using either
alternative solvents or preparations which entirely eliminate the use of solvents. While considerable
development work remains to be accomplished in the "no-solvent" area, a significant effort has
been expended toward the substitution of alternative solvent materials. Much of this effort has been
devoted to substitution with "non-photochemically" reactive solvents. This latter work has been very
successful and a significant switch is now underway or being contemplated by the industry. Such
conversions are resulting in sizeable reductions in photochemically reactive emissions by both the
289
-------
manufacturer and the user.
Aside from consideration of solvents, potential exists for emission level reduction through
substitution of other constituents in the paint formulation. Use of the liquid instead of the solid
form of certain raw materials, e.g., phthalic anhydride (PA) in resin manufacture, can result in reduced
emissions. This is economically feasible when large quantities are being used. Introduction of liquid
materials eliminates the necessity for evacuation of the equipment during loading operations. In the
case of phthalic anhydride the potential for emission is transferred to another point in the process
when the liquid form is used. Emission potential exists at the heated PA storage tank vent. Such
emission can be easily controlled, however, by venting the storage tank through a water jacketed
vertical condenser with provisions for admitting steam to the jacket. The tank is blanketed with
inert gas and appropriate pressure relief control provided. During loading the water cooled condenser
collects PA vapors as a solid which is later melted and returned to the storage tank.
Raw material substitution as outlined above can provide an excellent and sometimes in-
expensive method for emission reduction by the manufacturer. Unfortunately, these benefits may
not necessarily accrue to the users since raw material substitution can lead to both higher costs
and lower quality. An example of this which is unrelated to air pollution is the product deterioration
which has occurred as the result of the elimination of mercury fungicides and lead drying agents.
Industry spokesmen point out in this latter case that a period as long as several years may be
required to develop substitute raw materials which will restore paint to its former quality.
In many cases to date where raw material substitution has been made, the net result
has been either a higher product cost and/or a lower quality product. It is the conclusion of some
paint manufacturers that it would be less expensive overall for their customers to add control
equipment for existing solvent emissions than to absorb the cost for special solvent systems. This
factor coupled with the potential for degradation in product quality as well as the possibility for
future regulatory changes, e.g., the exempt solvent definition, suggest that raw material substitution
should be carefully evaluated as an alternative means for control.
2. Process Modification — Some potential exists for reduction in emission through process
modification. Possible modification includes:
a. Reduction in use of sparge gases
b. Use of adequate or refrigerated condensers
c. Use of proper unloading and transfer methods
d. Use of liquid slurries in pigment handling
290
-------
Taking maximum advantage of these modifications can substantially reduce emissions in
some operations such as resin cooking by the solvent method. The actual benefits to be derived
depend on the specific operation involved and should be subjected to a cost/benefit analysis in
relation to corresponding requirements and costs for air pollution control.
E. Varying Types and Levels of Control
Varying types and levels of control can be achieved with the thermal or catalytic afterburner
segments of the air pollution control system for the paint plant. However, significant cost variations
will be encountered which are dependent on the variation in control desired.
For example, suppose a hypothetical thermal afterburner is operating at an outlet temper-
ature of 1370°F obtained from a burner AT from fuel gas of 550°F and a fume combustion efficiency
of 90%. If the burner AT is increased by SOT, the efficiency might, for some fumes, increase to
95% and give an outlet temperature of 1450°F. For a situation with these parameters, this variation
in control level would result in an increase in fuel costs of around 9% without heat exchange.
For the same afterburner to which 42% heat exchange has been added, the same increase in
control level would result in an increase in fuel cost that would be less in actual dollars than the
increase for the first case but could amount to about 30% of the fuel cost of the heat exchanged
afterburner before increasing the temperature.
In the increased efficiency case above, additional maintenance costs would be encountered.
This arises from increased maintenance resulting from higher temperature This maintenance in-
crease is not an easily assignable cost on afterburners since the number of start-ups and shut-
downs influence this cost markedly. For the same operation cycle, howpver, maintenance cost
will be significantly higher for the heat exchanger case and will partially offset the lower fuel cost
indicated above.
Other operating costs for higher efficiency operation in the thermal afterburner will not be
significant. Slight increase in fan horsepower requirements, however, might be involved. Operating
labor requirements should be identical for either the high or low efficiency case.
Higher efficiency levels in baghouse operations are not readily obtainable since these systems
already operate at 99+ per cent efficiency (weight basis) when the system is in good working
order. Regular maintenance and observation will maintain these systems at very high levels of
efficiency.
F. Impact on Income, Cash Flow and Investment
The impact of air pollution control costs for the paint industry will be significant. Basing an
291
-------
assessment of this impact on the model plant described previously, the following comparison may
be made between the controlled and uncontrolled plants:
ANNUAL INCOME STATEMENT
Uncontrolled
plant
$6
5
$1
$
$
,604,400
,020,374
,584,026
761,141
822,885
252,000
570,885
342,531
Controlled
plant
$6,604,400
5,029,602
$1,574,798
761,141
$ 813,657
254,101
$ 559,556
335,733
Change for
controlled
plant
—
+ 9,228
$ - 9,228
—
$- 9,228
+ 2,101
$-11,329
- 6,798
Revenue (net)
Cost of goods sold
Gross profit
Selling, general and admin, expense
Operating income
Interest expense
Income before taxes*
Income taxes
Net income $ 228,354 $ 223,823 $- 4,531
* Includes local, state and federal taxes.
This information indicates a reduction in annual income after tax of $4,531 for the addition
of air pollution control equipment to the model plant. This loss in income amounts to 1.98 per
cent of the total net income of the model plant or an average of $0.006 per gallon of paint produced.
The information above regarding loss of income is based on the assumption that the model
plant would not encounter a price increase as a result of the effect of control on its supplier of
purchased resins. Of the 8,115,300 pounds of resin used by the model plant, 1,992,600 pounds of
solvent based resins are purchased from outside suppliers. A significant portion of the control
cost is directly attributable to resin manufacture. The design basis for manufactured resin incorpor-
ates a production capacity of 2,166,000 pounds annually. The average capital cost of control for
the resin plant has been shown previously as $26,000. The operating cost attributable to control
on the model resin plant at design capacity is as follows:
292
-------
INCREASED ANNUAL MANUFACTURING COST OF RESIN RESULTING FROM CONTROL
Depreciation (installed cost) $2,600
Annual operating cost 5,595
Decrease in operating income $8,195
Interest expense (assumed 7%) 1,638
Decrease in pretax income $9,833
Tax reductions (Federal & Local) 5,900
Decrease in net income $3,933
Applying this reduction in net income to the 2,166,000 pounds of resin used as the design basis, the
model plant will show increased production and interest costs of $0.00182 per pound of resin. With the
assumption that the outside resin supplier would incur the same increase in per pound cost and
pass this along as a cost increase, the loss in income to the model plant would be increased
significantly. Applying the $0.00182 per pound cost increase to 1,992,600 pounds of resin purchased
would increase income loss by $3,626.53 annually or to $0.00191 per gallon of paint produced.
This assumes that production of water based latex resins do not require equivalent control devices.
While the assumption made above for purchased resin cost increase is rather general, the
order of magnitude of increase deserves consideration in all of the financial calculations relating to
best control. Therefore, in the following financial data the influence of this factor will be pointed as an
effect of purchased resin cost increase.
Effect of control on cash flow also is significant. Based on information previously presented
for the model plant, the following comparison can be made:
NET CASH FLOW (ANNUALLY)
Net cash flow
Uncontrolled plant ($14,512)
Controlled plant ( 19,043)
Change in cash flow ($ 4,531)
While a cash out flow is indicated for each case, the cash out flow will be increased by
$4,531 for the addition of control equipment. The negative cash flow calculated suggests that some
seasonal short-term financing would be necessary to furnish cash requirements and that such
debt would be increased for the controlled plant case.
Each of several methods for assessment of return on investment illustrates the impact of
addition of air pollution control. These methods of calculation and the resultant conclusions have
293
-------
been shown previously for the model plant. A summary of the calculations is as follows:
RETURN ON INVESTMENT (PER CENT)
Change for
Uncontrolled Controlled controlled
plant plant plant
Total gross assets 6.53% 6.42% -1.68%
Total net assets 6.96% 6.85% -1.58%
Stockholders' equity plus
long-term debt 7.55% 7.43% -1.59%
Stockholders'equity 19.70% 19.36% -1.73%
The largest impact indicated from this summary is on the owners of the business or
the stockholders. Based on stockholders' equity, approximately a 1.73% reduction in annual return
results from the control additions.
Assuming a purchased resin cost increase as outlined previously, a reduction in income of
$3,627 would be encountered with a resultant $220,196 annual income. With this reduced income,
the loss of return on stockholders' equity for the controlled situation would be 19.04%. On this
basis, the net reduction in return based on stockholders' equity would be 3.25%.
III. INDUSTRY WIDE STUDIES
An assessement of the impact of best control on an industry wide basis can be made
using the following assumptions:
1. "Best Control" will be a requirement for all plants.
2. Cost of control is directly proportioned to overall production rates, i.e., the model plant
represents the average plant.
While these assumptions present certain inaccuracies, they should permit a reasonable estimate
of the order of magnitude of control cost. With these above assumptions, the industry wide influences
are reviewed below:
A. Present Total Cost to Industry to Meet Best Control Requirements
Paint production during 1972 was at a level of 930 million gallons per year. A reasonable
basis for estimating the order of magnitude of total cost to industry to meet best control might use
the capital and operating costs developed for the model plant above on the basis of cost per
gallon. With this basis, the approximate costs can be calculated as shown on the following page.
294
-------
TOTAL COST TO INDUSTRY FOR BEST CONTROL
Total Cost
Capital Investment Cost $16,200,000
Annual Operating Cost* $ 4,500,000
OPERATING COSTS PLUS DEPRECIATION
In the case of capital investment cost,
the above does not reflect a credit for
any existing air pollution control equip-
ment. The cost information presented is
based on 1972 monetary values, costs
and productivity and does not provide
escalation factors.
B. Fifteen Year Projection of the Cost of Control
Indicated above was a 1972 paint production rate of 930 million gallons per year. Most
of this production was accounted for by a relatively small number of plants. Paint production is
projected to increase to 1,320 million gallons per year in 1985. Extrapolating beyond 1985, total
annual paint production could exceed 1,700 million gallons per year.
Approximately 20% of the total operating plants produce about 85% of all coatings. The
industry trend is toward larger plants combined with a tendency to operate these plants on a three-
shift or 24-hour basis. This trend is expected to continue with the result that most increases in
production will be due to the installation of new plants of large capacity or by full utilization of
presently idle capacity rather than by increasing the operating hours.
Whether new plants or idle capacity are utilized, air pollution control requirements would
need to increase beyond the minimum necessary for today's production. A reasonable basis for
estimating the order of magnitude of future control costs might include an extrapolation of cumulative
capital cost and annual operating cost as a function of the projected industry shipment. This
extrapolation could be based on a proportionate cost per gallon of paint equal to that for the model
plant based on 1972 cost factors. This extrapolation provided the following estimate of projected
costs:
PROJECTED COST OF CONTROL (15 YEAR PERIOD)
Total Cost
Capital Investment Cost $30,600,000
Annual Operating Cost* $ 8,500,000
*BASED ON 1,700 MILLION GALLONS
PAINT PRODUCTION EACH YEAR
These estimates are based upon 1972
dollars. Changes in cost due to inflation,
labor productivity, increased taxes, and
other factors could easily swell the above
total by a factor of two or more.
295
-------
C. Sources of Capital For Pollution Control
Projections of capital investment based on the model plant suggest that, at present, on
the order of 16 million dollars in expenditures is required to provide best air pollution control. Further
projections suggest that at least 30.6 million will need to be expended in the 15 year period after
1972. This represents a substantial capital requirement on the particular industry and the optimum
method of financing will be dependent on some of the following variables:
1. Business condition of the specific paint manufacturer
2. Money market or interest rates
3. Strength of the equity market
4. National monetary valuation and controls
In the reasonable future, those manufacturers representing the majority of production will
all probably utilize the following sources of capital:
1. Retained earnings
2. Long-term borrowing or bonds
3. Portions of new equity capital
The financing technique used will in all likelihood result in an increase in the debt/equity ratios
for the industry.
Accelerated tax credits and accelerated depreciation allowances are expected to be important
factors in offsetting a portion of the capital requirements. It is anticipated, however, that eventually
the cost of control will be absorbed in price increases to the extent that competition from other
coatings or construction materials will allow.
D. Industry Structure
The paint industry has an oligopolistic structure. That is to say, the majority of production
is controlled by relatively few companies. Approximately 1,727 paint plants were being operated
in the United States during 1972. These plants were either directly owned or controlled by 1,365
U.S. companies. This does not include those plants which make paint components only, such as
resins, but not finished coatings.
As indicated previously, about 20% of the total number of plants, or 345, produce about
75% of the total paint production. This 75% of total production is controlled by fewer than 200
companies which, in turn, control some of the smaller plants. The distribution of paint production
among controlling companies is approximately as follows:
296
-------
DISTRIBUTION OF PAINT PRODUCTION
1.7
Approximate
Number of Percentage of
Companies Total Production
Major Companies 36 64
Intermediate Companies 164 24
Small Companies 1,165 12
1,365 100.0
The above confirms that the paint industry has an oligopolistic structure.
In spite of this structure, the coatings business is probably unique among major manufactur-
ing industries in that the very small producer is able to survive in the face of competition from
some of the largest corporations in the world. While the trend is towards larger plants, it is a slow
process and the continued existence of the small producer is assured for the foreseeable future.
There are several reasons for this:
1. The diversity of products is so large that there will always be a place for the producer
of low volume specialty items.
2. Finished coatings are expensive to ship due to their weight and volume. Consequently,
the economy benefits of large operations are diminished as the shipping distance is
increased.
3. A large amount of technical assistance is available from raw material suppliers which
enables the small company to remain technologically up-to-date.
4. The nature of the manufacturing operations (batch processing of relatively small
quantities) limits the economy that can be attained in large scale operations. Indeed,
with the possible exception of filling and warehousing operations, there is little difference
in the manner in which the small plant does things as opposed to the way large plants
operate.
E. Product Elasticity — Production Substitution
The paint industry continues to demonstrate an upward growth pattern. This growth
continues even during periods of rapidly increasing prices. The industry exhibits a strong indication
of price inelasticity, i.e., demand and total receipts do not significantly change with rising prices.
In the case of increasing prices, the separation of "real" and inflationary related contributions is
difficult. In the case of price increases associated with air emission control devices, little if any
effect on demand inelasticity can be projected.
297
-------
Future increases in price will be influenced by a number of cost factors only one of which
is projected air pollution control costs. Perhaps more significant to the paint industry will be cost
increases from the following areas:
1. Inflation
2. Labor productivity reduction and cost increase
3. Energy cost increases
4. Water treatment and solid waste disposal costs
5. OSHA regulations
Also to be considered in price increase is product reformulation in terms of quality or of pollution
control requirements of the user.
Product substitution is also a significant factor in the present era of rapid technological
growth. Product substitution has been significant in recent years in the following areas:
1. Wall coverings
2. Paneling
3. Plastics
It is anticipated that a high level of product substitution will continue in the above areas and extend
into other areas. However, for the foreseeable future, it is not anticipated that air pollution control
costs alone will cause a substantial increase in product substitution.
298
-------
TABLE 64
SPECIFICATIONS FOR ABATEMENT EQUIPMENT
1. SCOPE
A. This specification covers vendor requirements for air pollution control equipment for the
subject process. The intent of the specification is to describe the service as thoroughly as possible
so as to secure vendor's proposal for equipment which is suitable in every respect for the service
intended. Basic information is tabulated in sections 2 and 3. The vendor should specify any of the
performance characteristics which cannot be guaranteed without samples of process effluent.
B. The vendor shall submit a bid showing three separate prices as described below.
1. All labor, materials, equipment and services to furnish one pollution abatement device
together with the following:
a. All ladders, platforms and other accessways to provide convenient access to all
points requiring observation or maintenance
b. Foundation bolts as required
c. Six (6) sets of drawings, instructions, spare parts list, etc., pertinent to the above
2. Auxiliaries including:
a. Fan(s)
b. Pump(s)
c. Damper(s)
d. Conditioning Equipment
e. Dust Disposal Equipment
3. A turnkey installation of the entire system including the following installation costs:
a. Engineering
b. Foundations & Support
c. Ductwork
d. Stack
e. Electrical
f. Piping
g. Insulation
h. Painting
i. Startup
299
-------
TABLE 64 (continued)
k. Performance Test
I. Other (including general tradework such as erection, rigging, etc.)
C. For the "pollution abatement device only" quotation, the vendor shall furnish the equipment
FOB point of manufacture, and shall furnish as a part of this project competent supervision of the
erection, which shall be by others.
D. Vendor shall furnish* the following drawings, etc., as a minimum:
1. With his proposal:
a. Plan and elevation showing general arrangement
b. Typical details of collector internals proposed
c. Data relating projected performance with respect to pressure drop, gas absorption
efficiency and particulate removal efficiency to operating parameters such as gas flow
2. Upon receipt of order:
a. Proposed schedule of design and delivery
3. Within 60 days of order:
a. Complete drawings of equipment for approval by customer
4. 30 days prior to shipment:
a. Certified drawings of equipment, six sets
b. Installation instructions, six sets
c. Starting and operating instructions, six sets
d. Maintenance instructions and recommended spare parts lists, six sets
E. The design and construction of the collector and auxiliaries shall conform to the general
conditions given in Section 5, and to good engineering practice.
2. PROCESS PERFORMANCE GUARANTEE
A. The equipment will be guaranteed to reduce the particulate and/or gas contaminant loadings
as indicated in the service description.
B. Performance test will be conducted in accordance with IGCI test methods where applicable.
C. Testing shall be conducted at a time mutually agreeable to the customer and the vendor.
D. The cost of the performance test is to be included in vendor's turnkey proposal.
*This is a typical request. The member companies are NOT to furnish this material under the
present project.
300
-------
TABLE 64 (continued)
E. In the event the equipment fails to comply with the guarantee at the specified design
conditions, the vendor shall make every effort to correct any defect expeditiously at his own expense.
Subsequent retesting to obtain a satisfactory result shall be at the vendor's expense.
3. GENERAL CONDITIONS
A. Materials and Workmanship
Only new materials of the best quality shall be used in the manufacture of items covered
by this specification. Workmanship shall be of high quality and performed by competent workmen.
B. Equipment
Equipment not of vendor's manufacture furnished as a part of this collector shall be regarded
in every respect as though it were of vendor's original manufacture.
C. Compliance with Applicable Work Standards and Codes
It shall be the responsibility of the vendor to design and manufacture the equipment specified
in compliance with the practice specified by applicable codes.
D. Delivery Schedules
The vendor shall arrange delivery of equipment under this contract so as to provide for
unloading at the job site within a time period specified by the customer. Vendor shall provide for
expediting and following shipment of materials to the extent required to comply with delivery specified.
301
-------
TABLE 65
INSTRUCTIONS FOR SUBMITTING COST DATA
Two forms (two copies each) are enclosed with each specification. These are for submitting:
(A) Estimated Capital Cost Data
(B) Annual Operating Cost Data
These forms will also be used to exhibit averages of the three cost estimates for each process and
equipment type. Because your costs will be averaged with those of other IGCI members, it is
necessary to prepare them in accordance with instructions given in the following paragraphs.
(A) Estimated Capital Cost Data
The upper part of this form should already be filled out for the particular application when you
receive it. This information on operating conditions should be identical to that in the specification
and is repeated here only for the convenience of those reading the form.
You should fill in the dollar amounts estimated in the appropriate spaces on the bottom half of
the form. It should not be necessary to add any information other than the dollar amounts. If
you wish to provide a description of the equipment proposed, please do so on one or more separate
sheets of paper, and attach it to the form. If any item is not involved in the equipment you are
proposing, please indicate this by writing "none" in the space rather than leaving it blank or using
a zero.
(1) The "gas cleaning device" cost should be reported just as you would report a flange-
to-flange equipment sale to the IGCI. That is, a complete device including necessary auxiliaries
such as power supplies, mist eliminators, etc. Do NOT include such items as fans, solids handling
equipment, etc., unless these are an integral part of your gas cleaning device.
(2) "Auxiliaries" are those items of equipment which are frequently supplied with the gas
cleaning device. There is a purely arbitrary definition of those items included here and those included
in the "Installation" Costs. Do NOT include any of the cost of erecting or installing auxiliaries in
this category.
(3) "Installation Cost" should include all of the material not in (1) or (2) and the field
labor required to complete a turnkey installation. In cases where the equipment supplier ordinarily
erects the equipment but does not supply labor for foundations, etc., it is necessary to include an
estimated cost for these items. General tradework, including rigging, erection, etc. should be included
in the "Other" category.
302
-------
TABLE 65 (continued)
The installation should be estimated for a new plant, or one in which there are no limitations
imposed by the arrangement of existing equipment. Installation labor should be estimated on the
basis that the erection will take place in an area where labor rates are near the U.S. average,
and the distance from your plant is no more than 500 miles. Milwaukee, Wisconsin is an example
of a city with near-average labor rates.
(B) Annual Operating Cost Data
Some of the information will be supplied by Air Resources, such as unit costs for labor and utilities,
and annualized capital charges. You should fill in the usage figures for the complete abatement
system in the units indicated below. Please include the unit price.
Labor hr/year
Maintenance Materials $/year
Replacement Parts $/year
Electric Power kw-hr/year
Fuel MMBtu/year
Water (Process) MM gal/year
Water (Cooling) MM gal/year
Chemicals $/year
Air Resources will average the consumption figures reported, and convert them to dollar values for
inclusion in the final report.
Be sure that the operating factor, in hours per year, supplied by ARI, is used for estimating the
utility and labor requirements.
303
-------
TABLE 66
THERMAL AFTERBURNER PROCESS DESCRIPTION
FOR RESIN REACTOR SPECIFICATION
This specification describes the requirements for a thermal combustion system for abatement of
the hydrocarbon emissions from the resin production facility of a paint and varnish plant. The system will
be similar to that shown in Figure 9. All reactor thinning tanks and product rundown tanks will be vented
to the collection system. Dilution air will be supplied through a hood over the resin filter press(es). A
minimum flow damper is to be supplied in this part of the ventilation system. It is to be sized to allow fora
maximum fume concentration of40%LEL in the fully closed position. A high velocity Venturi section is to
be located at the afterburner inlet to assure gas flow is in excess of flame propagation velocity at 112
design flow rate.
The afterburner is to be natural gas fired. Sufficient gas, having a specific gravity of 0.60 and an
upper heating value of 1040 Btu/SCF, is available at a pressure of 1.0 psig.
The exhaust gas contains sufficient oxygen, greater than 16% Oa fo allow firing of the afterburner
with a raw gas or process air burner. A combustion air system is not required. Fume load to the
incinerator is composed of 75% kettle emission and 25% from other sources. Average emissions are
60% of peak. The system is to be designed on peak emission but operating costs are to be based on
average emissions. Operating conditions listed are based on peak emissions. Fume destruction in
burner is to be calculated as follows:
10% for catalytic units with or without heat exchange
10% for thermal units with heat exchange
20% for thermal units without heat exchange
Please fill in estimated efficiency of afterburner and burner duty.
The afterburner is to be supplied with a suitable control panel and all equipment is to be designed
for outdoor operation. Incinerator operating and safety controls are to be designed to meet F.I.A.
(Factory Insurance Association) requirement. All dampers are to be pneumatically operated and
contain an integral fail safe air reservoir. The system is to automatically divert in the event of low flow,
high afterburner temperature, high reactor pressure, and afterburner preheat burner failure. The
exhaust system should also be purged with inert gas on fan failure or loss of flow. Damper operating is
to be sequential with the position switches mounted on the damper arm and not the operator. The
system fan shall be located after the preheat burner or the incinerator outlet and shall be constructed to
withstand 200°F higher than design operating temperature. The fan shall have a vee belt drive and fan
and motor shall have the capacity to overcome the pressure drop of the ductwork, afterburner and any
heat exchanger that may be used. System ductwork should be sized for a maximum &Pof2in. w.c.hot.
The fan motor may be sized for restricted flow cold start. The ductwork to the incinerator should be
heated either by the use of a double manifold or hot gas recycle or a combination of both.
The heat exchanger when included is to be the parallel flow shell and tube type and designed to
operate at an afterburner outlet temperature of 1500°F. Maximum exchanger pressure drop, both sides,
should not exceed a total of 6 in. w.c. hot. Dirty gas is to flow through the tube side.
304
-------
TABLE 66 (Continued)
INSTALLATION
A complete turnkey proposal including ductwork, structural steel, fuel and inert gas piping, etc., is
requested. For the purpose of this proposal fan and damper including operators are to be considered
as auxiliary. The controls and control cabinet are to be included with the afterburner price. The
afterburner will be assumed to be located on a structural steel base on the resin plant roof. No
modification to the building structural steel is required. A tie through the roof from the base to the
building steel is required. The base and tie in are part of the installed structural cost. All utilities are
available within 30 ft of the control cabinet, motor, and burner. Plant air is available at 100 psig and
located within 30 ft. Inert gas is available and will require 30 ft of piping from supply to ductwork. The
existing stack can be used for the diversion stack. A 10 ft exhaust stack will be mounted on the
afterburner, double manifold or exhaust fan. A total of thirty-five (35) feet of new exhaust ductwork and
twenty-five (25) feet of double manifold will be required.
305
-------
TABLE 67
THERMAL AFTERBURNER OPERATING CONDITIONS
FOR RESIN REACTOR SPECIFICATION
(Without Heat Exchange)
Small Large
Process Conditions
Reactors, Number 1 2
Size each, gal 1,700 5,000
Hcbn Emission Max, Ib/hr 42.4 254
Hcbn Emission Ave, Ib/hr 24.7 148
Total Exhaust Rate, SCFM 1,000 6,000
Exhaust Temperature, °F 110 110
Heat of Combustion of
Reactor Fume, Btu/lb 17,000 17,000
Hydrocarbon Concentration
Maximum, Btu/SCF 12 12
Average, Btu/SCF 7 7
Unit Residence Time, sec @ 1500°F 0.6 0.6
Afterburner Without Heat Exchange
Unit Inlet, °F 325 325
Burner AT from Fuel Gas, °F 550 550
* Burner AT from Flame Combustion, °F 125 125
Burner Outlet Temperature, °F 1,000 1,000
** Unit A7 from Thermal Combustion, °F 435 435
Unit Outlet Temperature, °F 1,435 1,435
Burner Duty, MM Btu/hr
Estimated Unit Removal Efficiency
* Assumes 20% fume combustion in burner flame
** Assumes 95% overall fume combustion
306
-------
TABLE 68
THERMAL AFTERBURNER OPERATING CONDITIONS
FOR RESIN REACTOR SPECIFICATION
(Without Heat Exchange)
Small Large
Process Conditions
Reactors, Number 1 3
Size each, gal 5,000 5,000
Hcbn Emission Max, Ib/hr 125 423
Hcbn Emission Ave, Iblhr 74 246
Total Exhaust Rate, SCFM 3,000 10,000
Exhaust Temperature, °F 110 110
Heat of Combustion of
Reactor Fume, Btulib 17,000 17,000
Hydrocarbon Concentration
Maximum, Btu/SCF 12 12
Average, Btu/SCF 7 7
Unit Residence Time, sec @ 1500°F 0.6 0.6
Afterburner Without Heat Exchange
Unit Inlet, °F 325 325
Burner AT from Fuel Gas, °F 550 550
"Burner AT" from Flame Combustion, °F 125 125
Burner Outlet Temperature, °F 1,000 1,000
**Unit AT from Thermal Combustion, °F 435 435
Unit Outlet Temperature, °F 1,435 1,435
Burner Duty, MM Btu/hr
Estimated Unit Removal Efficiency
* Assumes 20% fume combustion in burner flame
** Assumes 95% overall fume combustion
307
-------
TABLE 69
THERMAL AFTERBURNER OPERATING CONDITIONS
FOR RESIN REACTOR SPECIFICATION
(With Heat Exchange)
Small Large
Process Conditions
Reactors, Number 2
Size each, gal 5,000
Hcbn Emission Max, Ib/hr 254
Hcbn Emission Ave, Ib/hr 148
Total Exhaust Rate, SCFM 6,000
Exhaust Temperature, °F 110
Heat of Combustion of
Reactor Fume, Btu/lb 17,000
Hydrocarbon Concentration
Maximum, Btu/SCF 12
Average, Btu/SCF 7
Unit Residence Time, sec @ 1500°F 0.6
Afterburner with Heat Exchange
Inlet Tube Side, °F 325
Unit Inlet, °F 825
Burner AT from Fuel Gas, °F 115
* Burner AT from Flame Combustion, °F 60
Burner Outlet Temperature, °F 1,000
** Unit AT from Thermal Combustion, °F 500
Unit Outlet Temperature, °F 1,500
Outlet Shell Side, °F 1,030
Burner Duty, MM Btu/hr
Exchanger Duty, MM Btu/hr
Thermal Efficiency -42%
Overall Heat Trans. Coef., U
Tube Surface Area, ft2
Estimated Afterburner Removal Efficiency
"Assumes 10% fume combustion in burner flame
"Assumes 95% overall fume combustion
308
-------
TABLE 70
THERMAL AFTERBURNER OPERATING CONDITIONS
FOR RESIN REACTOR SPECIFICATION
(With Heat Exchange)
Small Large
Process Conditions
Reactors, Number 1 3
Size each, gal 5,000 5,000
Hcbn Emission Max, Ib/hr 127 423
Hcbn Emission Ave, Ib/hr 74 246
Total Exhaust Rate, SCFM 3,000 10,000
Exhaust Temperature, °F 110 110
Heat of Combustion of
Reactor Fume, Btulib 17,000 17,000
Hydrocarbon Concentration
Maximum, BtulSCF 12 12
Average, BtulSCF 7 7
Unit Residence Time, sec @ 1500°F 0.6 0.6
Afterburner with Heat Exchange
Inlet Tube Side, °F 325 325
Unit Inlet, °F 825 825
Burner AT from Fuel Gas, °F 115 115
* Burner A7 from Flame Combustion, °F 60 60
Burner Outlet Temperature, °F 1,000 1,000
** Unit AT from Thermal Combustion, °F 500 500
Unit Outlet Temperature, °F 1,500 1,500
Outlet Shell Side, °F 1,030 1,030
Burner Duty, MM Btu/hr
Exchanger Duty, MM Btu/hr
Thermal Efficiency -42% -42%
Overall Heat Trans. Coef., U
Tube Surface Area, ft2
Estimated Afterburner Removal Efficiency
"Assumes 10% fume combustion in burner flame
**Assumes 95% overall fume combustion
309
-------
TABLE 71
THERMAL AFTERBURNER PROCESS DESCRIPTION
FOR OPEN KETTLE SPECIFICATION
This specification describes the requirements for a thermal combustion system for abatement of
the hydrocarbon emissions from the resin production facility of a paint and varnish plant. The system will
be similar to that shown in Figure 9. All reactor thinning tanks and product rundown tanks will be vented
to the collection system. Dilution air will be supplied through a hood over the resin filter press(es). A
minimum flow damper is to be supplied in this part of the ventilation system. It is to be sized to allow fora
maximum fume concentration of 40 % LEL in the fully closed position. A high velocity Venturi section is to
be located at the afterburner inlet to assure gas flow is in excess of flame propagation velocity at 1/2
design flow rate.
The afterburner is to be natural gas fired. Sufficient gas, having a specific gravity of 0.60 and an
upper heating value of 1040 Btu/SCF, is available at a pressure of 1.0 psig.
The exhaust gas contains sufficient oxygen, greater than 16% 02, to allow firing of the afterburner
with a raw gas or process air burner. A combustion air system is not required. Fume load to the
afterburner is composed of 75% kettle emission and 25% from other sources. Average emissions are
60% of peak. The system is to be designed on peak emission but operating costs are to be based on
average emissions. Operating conditions listed are based on peak emissions. Fume destruction in
burner is to be calculated as follows:
10% for catalytic units with or without heat exchange
10% for thermal units with heat exchange
20% for thermal units without heat exchange
Please fill in estimated efficiency of afterburner and burner duty.
The afterburner is to be supplied with a suitable control panel and all equipment is to be designed
for outdoor operation. Afterburner operating and safety controls are to be designed to meet F.I.A.
(Factory Insurance Association) requirement. All dampers are to be pneumatically operated and
contain an integral fail safe air reservoir. The system is to automatically divert in the event of low flow,
high afterburner temperature, high reactor pressure, and afterburner preheat burner failure. The
exhaust system should also be purged with inert gas on fan failure or loss of flow. Damper operating is
to be sequential with the position switches mounted on the damper arm and not the operator. The
system fan shall be located after the preheat burner or the afterburner outlet and shall be constructed to
withstand 200°F higher than design operating temperature. The fan shall have a vee belt drive and fan
and motor shall have the capacity to overcome the pressure drop of the ductwork, afterburner and any
heat exchanger that may be used. System ductwork should be sized for a maximum AP of 2 in. w.c. hot.
The fan motor may be sized for restricted flow cold start. The ductwork to the afterburner should be
heated either by the use of a double manifold or hot gas recycle or a combination of both.
The heat exchanger when included is to be the parallel flow shell and tube type and designed to
operate at an afterburner outlet temperature of 1500°F. Maximum exchanger pressure drop, both sides,
should not exceed a total of 6 in. w.c. hot. Dirty gas is to flow through the tube side.
310
-------
TABLE 71 (Continued)
INSTALLATION
A complete turnkey proposal including ductwork, structural steel, fuel and inert gas piping, etc., is
requested. For the purpose of this proposal fan and damper including operators are to be considered
as auxiliary. The controls and control cabinet are to be included with the afterburner price. The
afterburner will be assumed to be located on a structural steel base on the resin plant roof. No
modification to the building structural steel is required. A tie through the roof from the base to the
building steel is required. The base and tie in are part of the installed structural cost. All utilities are
available within 30 ft of the control cabinet, motor, and burner. Plant air is available at 100 psig and
located within 30 ft. Inert gas is available and will require 30 ft of piping from supply to ductwork. The
existing stack can be used for the diversion stack. A 10 ft exhaust stack will be mounted on the
afterburner, double manifold or exhaust fan. A total of thirty-five (35) feet of new exhaust ductwork and
twenty-five (25) feet of double manifold will be required.
311
-------
TABLE 72
THERMAL AFTERBURNER OPERATING CONDITIONS
FOR OPEN KETTLE SPECIFICATION
(Without Heat Exchange)
Small Large
Process Conditions
Kettles, Number 2 4
Size each, gal 300 300
Hcbn Emission Max, Ib/hr 45 90
Hcbn Emission Ave, Ib/hr 26.2 52
Total Exhaust Rate, SCFM 1,000 2,000
Exhaust Temperature, °F 80 80
Heat of Combustion of
Kettle Fume, Btu/lb 16,000 16,000
Hydrocarbon Concentration
Maximum, Btu/SCF 12 12
Average, Btu/SCF 7 7
Unit Residence Time, sec @ 1500°F 0.6 0.6
Afterburner Without Heat Exchange
Unit Inlet, °F 300 300
Burner A7 from Fuel Gas, °F 575 575
* Burner AT from Flame Combustion, °F 125 125
Burner Outlet Temperature, °F 1,000 1,000
** Unit AT from Thermal Combustion, °F 435 435
Unit Outlet Temperature, °F .1,435 1,435
Burner Duty, MM Btu/hr
Estimated Unit Removal Efficiency
* Assumes 20% fume combustion in burner flame
** Assumes 95% overall fume combustion
312
-------
TABLE 73
THERMAL AFTERBURNER OPERATING CONDITIONS
FOR OPEN KETTLE SPECIFICATION
(Without Heat Exchange)
Small Large
Process Conditions
Kettles, Number 1 6
Size each, gal 300 300
Hcbn Emission Max, Ib/hr 22.5 135
Hcbn Emission Ave, Ib/hr 13.1 78.8
Total Exhaust Rate, SCFM 500 3,000
Exhaust Temperature, °F 80 80
Heat of Combustion of
Kettle Fume, Btu/lb 16,000 16,000
Hydrocarbon Concentration
Maximum, Btu/SCF 12 12
Average, Btu/SCF 7 7
Unit Residence Time, sec @ 1500°F 0.6 0.6
Afterburner Without Heat Exchange
Unit Inlet, °F 300 300
Burner AT from Fuel Gas, °F 575 575
* Burner AT from Flame Combustion, °F 125 125
Burner Outlet Temperature, °F 1,000 1,000
** Unit AT from Thermal Combustion, °F 435 435
Unit Outlet Temperature, °F 1,435 1,435
Burner Duty, MM Btu/hr
Estimated Unit Removal Efficiency
"Assumes 20% fume combustion in burner flame
** Assumes 95% overall fume combustion
313
-------
TABLE 74
CATALYTIC AFTERBURNER PROCESS DESCRIPTION
FOR RESIN REACTOR SPECIFICATION
This specification describes the requirements for a catalytic combustion system for abatement of
the hydrocarbon emissions from the resin production facility of a paint and varnish plant. The system will
be similar to that shown in Figure 9.'All reactor thinning tanks and product rundown tanks will be vented
to the collection system. Dilution air will be supplied through a hood over the resin filter press(es). A
minimum flow damper is to be supplied in this part of the ventilation system. It is to be sized to allow for a
maximum fume concentration of40%LELin the fully closed position. A high velocity Venturi section is to
be located at the afterburner inlet to assure gas flow is in excess of flame propagation velocity at 112
design flow rate.
The afterburner is to be natural gas fired. Sufficient gas, having a specific gravity of 0.60 and an
upper heating value of 1040 Btu/SCF, is available at a pressure of 1.0 psig.
The exhaust gas contains sufficient oxygen, greater than 16% Oz to allow firing of the afterburner
with a raw gas or process air burner. A combustion air system is not required. Fume load to the
afterburner is composed of 75% kettle emission and 25% from other sources. Average emissions are
60% of peak. The system is to be designed on peak emission but operating costs are to be based on
average emissions. Operating conditions listed are based on peak emissions. Fume destruction in
burner is to be calculated as follows:
10% for catalytic units with or without heat exchange
10% for thermal units with heat exchange
20% for thermal units without heat exchange
Please fill in estimated efficiency of afterburner and burner duty, catalyst face velocity, and
catalyst volume.
The afterburner is to be supplied with a suitable control panel and all equipment is to be designed
for outdoor operation. Afterburner operating and safety controls are to be designed to meet F.I.A.
(Factory Insurance Association) requirement. All dampers are to be pneumatically operated and
contain an integral fail safe air reservoir. The system is to automatically divert in the event of low flow,
high afterburner temperature, high reactor pressure, and afterburner preheat burner failure. The
exhaust system should also be purged with inert gas on fan failure or loss of flow. Damper operating is
to be sequential with the position switches mounted on the damper arm and not the operator. The
system fan shall be located after the preheat burner or the afterburner outlet and shall be constructed to
withstand 200°Fhigher than design operating temperature. The fan shall have a vee belt drive and fan
and motor shall have the capacity to overcome the pressure drop of the ductwork, afterburner and any
heat exchanger that may be used. System ductwork should be sized for a maximum APof2/n. w.c.hot.
The fan motor may be sized for restricted flow cold start. The ductwork to the afterburner should be
heated either by the use of a double manifold or hot gas recycle or a combination of both.
The heat exchanger when included is to be the parallel flow shell and tube type and designed to
operate at an afterburner outlet temperature of 1200°F. Maximum exchanger pressure drop, both sides,
should not exceed a total of 6 in. w.c. hot. Dirty gas is to flow through the tube side.
'Appendix B
314
-------
TABLE 74 (Continued)
INSTALLATION
A complete turnkey proposal including ductwork, structural steel, fuel and inert gas piping, etc., is
requested. For the purpose of this proposal fan and damper including operators are to be considered
as auxiliary. The controls and control cabinet are to be included with the afterburner price. The
afterburner will be assumed to be located on a structural steel base on the resin plant roof. No
modification to the building structural steel is required. A tie through the roof from the base to the
building steel is required. The base and tie in are part of the installed structural cost. All utilities are
available within 30 ft of the control cabinet, motor, and burner. Plant air is available at 100 psig and
located within 30 ft. Inert gas is available and will require 30 ft of piping from supply to ductwork. The
existing stack can be used for the diversion stack. A 10 ft exhaust stack will be mounted on the
afterburner, double manifold or exhaust fan. A total of thirty-five (35) feet of new exhaust ductwork and
twenty-five (25) feet of double manifold will be required.
315
-------
TABLE 75
CATALYTIC AFTERBURNER OPERATING CONDITIONS
FOR RESIN REACTOR SPECIFICATION
(Without Heat Exchange)
Small Large
Process Conditions
Reactors, Number 1 2
Size each, gal 1,700 5,000
Hcbn Emission Max, Ib/hr 42.4 254
Hcbn Emission Ave, Ib/hr 24.7 148
Total Exhaust Rate, SCFM 1,000 6,000
Exhaust Temperature, °F 110 110
Heat of Combustion of
Reactor Fume, Btulib 17,000 17,000
Hydrocarbon Concentration
Maximum, Btu/SCF 12 12
Average, Btu/SCF 7 7
Afterburner Without Heat Exchange
Unit Inlet, °F 300 300
Burner A7 from Fuel Gas, °F 240 240
* Burner AT from Flame Combustion, °F 60 60
Burner Outlet Temperature, °F 600 600
** Unit AT" from Thermal Combustion, °F 500 500
Unit Outlet Temperature, °F 1,100 1,100
Burner Duty, MM Btu/hr
Estimated Unit Removal Efficiency
***Catalyst Face Velocity, SF/min
Catalyst Volume, ft3
''Assumes 10% fume combustion in burner flame
** Assumes 95% overall fume combustion
*** Basis 70°F exhaust temperature
316
-------
TABLE 76
CATALYTIC AFTERBURNER OPERATING CONDITIONS
FOR RESIN REACTOR SPECIFICATION
(Without Heat Exchange)
Small Large
Process Conditions
Reactors, Number 1 3
Size each, gal 5,000 5,000
Hcbn Emission Max, Ib/hr 127 423
Hcbn Emission Ave, Ib/hr 74 246
Total Exhaust Rate, SCFM 3,000 10,000
Exhaust Temperature, °F 110 110
Heat of Combustion of
Reactor Fume, Btulib 17,000 17,000
Hydrocarbon Concentration
Maximum, Btu/SCF 12 12
Average, Btu/SCF 7 7
Afterburner Without Heat Exchange
Unit Inlet, °F 300 300
Burner AT from Fuel Gas, °F 240 240
* Burner AT from Flame Combustion, °F 60 60
Burner Outlet Temperature, °F 600 600
** Unit AT from Thermal Combustion, °F 500 500
Unit Outlet Temperature, °F 1,100 1,100
Burner Duty, MM Btu/hr
Estimated Unit Removal Efficiency
""Catalyst Face Velocity, SF/min
Catalyst Volume, ft3
"Assumes 10% fume combustion in burner flame
** Assumes 95% overall fume combustion
*** Basis 70°F exhaust temperature
317
-------
TABLE 77
CATALYTIC AFTERBURNER OPERATING CONDITIONS
FOR RESIN REACTOR SPECIFICATION
(With Heat Exchange)
Small Large
Process Conditions
Reactors, Number 2
Size each, gal 5,000
Hcbn Emission Max, Ib/hr 254
Hcbn Emission Ave, Ib/hr 148
Total Exhaust Rate, SCFM 6,000
Exhaust Temperature, °F 110
Heat of Combustion of
Reactor Fume, Btulib 17,000
Hydrocarbon Concentration
Maximum, BtulSCF 12
Average, Btu/SCF 7
Afterburner with Heat Exchange
Inlet Tube Side, °F 300
Unit Inlet, °F 500
Burner AT from Fuel Gas, °F 40
* Burner A7 from Flame Combustion, °F 60
Burner Outlet Temperature, °F 600
** Unit AT from Thermal Combustion, °F 500
Unit Outlet Temperature, °F 1,100
Outlet Shell Side, °F 915
Burner Duty, MM Btu/hr
Exchanger Duty, MM Btu/hr
Thermal Efficiency -23%
Overall Heat Trans. Coef., U
Tube Surface Area, ft2
Estimated Unit Removal Efficiency
"""Catalyst Face Velocity, SF/min
Catalyst Volume, ft3
"Assumes 10% fume combustion in burner flame
"" Assumes 95% overall fume combustion
""Basis 70°F exhaust temperature
318
-------
TABLE 78
CATALYTIC AFTERBURNER OPERATING CONDITIONS
FOR RESIN REACTOR SPECIFICATION
(With Heat Exchange)
Small Large
Process Conditions
Reactors, Number 1 3
Size each, gal 5,000 5,000
Hcbn Emission Max, Ib/hr 127 423
Hcbn Emission Ave, Iblhr 74 246
Total Exhaust Rate, SCFM 3,000 10,000
Exhaust Temperature, °F 110 110
Heat of Combustion of
Reactor Fume, Btu/lb 17,000 17,000
Hydrocarbon Concentration
Maximum, Btu/SCF 12 12
Average, Btu/SCF 7 7
Afterburner with Heat Exchange
Inlet Tube Side, °F 300 300
Unit Inlet, °F 500 500
Burner AT from Fuel Gas, °F 40 40
* Burner AT from Flame Combustion, °F 60 60
Burner Outlet Temperature, °F 600 600
** Unit AT from Thermal Combustion, °F 500 500
Unit Outlet Temperature, °F 1,100 1,100
Outlet Shell Side, °F 915 915
Burner Duty, MM Btu/hr
Exchanger Duty, MM Btu/hr
Thermal Efficiency -23% -23%
Overall Heat Trans. Coef., U
Tube Surface Area, ft2
Estimated Unit Removal Efficiency
"Catalyst Face Velocity, SF/min
Catalyst Volume, ft3
* Assumes 10% fume combustion in burner flame
** Assumes 95% overall fume combustion
*** Basis 70°F exhaust temperature
319
-------
TABLE 79
CATALYTIC AFTERBURNER PROCESS DESCRIPTION
FOR OPEN KETTLE SPECIFICATION
This specification describes the requirements for a catalytic combustion system for abatement of
the hydrocarbon emissions from the resin production facility of a paint and varnish plant. The system will
be similar to that shown in Figure 9. * All reactor thinning tanks and product rundown tanks will be vented
to the collection system. Dilution air will be supplied through a hood over the resin filter press(es). A
minimum flow damper is to be supplied in this part of the ventilation system. It is to be sized to allow for a
maximum fume concentration of 40 %LEL in the fully closed position. A high velocity Venturi section is to
be located at the afterburner inlet to assure gas flow is in excess of flame propagation velocity at 112
design flow rate.
The afterburner is to be natural gas fired. Sufficient gas, having a specific gravity of 0.60 and an
upper heating value of 1040 Btu/SCF, is available at a pressure of 1.0 psig.
The exhaust gas contains sufficient oxygen, greater than 16% 0% to allow firing of the afterburner
with a raw gas or process air burner. A combustion air system is not required. Fume load to the
afterburner is composed of 75% kettle emission and 25% from other sources. Average emissions are
60% of peak. The system is to be designed on peak emission but operating costs are to be based on
average emissions. Operating conditions listed are based on peak emissions. Fume destruction in
burner is to be calculated as follows:
10% for catalytic units with or without heat exchange
10% for thermal units with heat exchange
20% for thermal units without heat exchange
Please fill in estimated efficiency of afterburner and burner duty, catalyst face velocity, and
catalyst volume.
The afterburner is to be supplied with a suitable control panel and all equipment is to be designed
for outdoor operation. Afterburner operating and safety controls are to be designed to meet F.I.A.
(Factory Insurance Association) requirement. All dampers are to be pneumatically operated and
contain an integral fail safe air reservoir. The system is to automatically divert in the event of low flow,
high afterburner temperature, high reactor pressure, and afterburner preheat burner failure. The
exhaust system should also be purged with inert gas on fan failure or loss of flow. Damper operating is
to be sequential with the position switches mounted on the damper arm and not the operator. The
system fan shall be located after the preheat burner or the afterburner outlet and shall be constructed to
withstand 200°F higher than design operating temperature. The fan shall have a vee belt drive and fan
and motor shall have the capacity to overcome the pressure drop of the ductwork, afterburner and any
heat exchanger that may be used. System ductwork should be sized for a maximum AP of 2 in. w.c.hot.
The fan motor may be sized for restricted flow cold start. The ductwork to the afterburner should be
heated either by the use of a double manifold or hot gas recycle or a combination of both.
The heat exchanger when included is to be the parallel flow shell and tube type and designed to
operate at an afterburner outlet temperature of 1200°F. Maximum exchanger pressure drop, both sides,
should not exceed a total of 6 in. w.c. hot. Dirty gas is to flow through the tube side.
"Appendix B
320
-------
TABLE 79 (Continued)
INSTALLATION
A complete turnkey proposal Including ductwork, structural steel, fuel and inert gas piping, etc., is
requested. For the purpose of this proposal fan and damper including operators are to be considered
as auxiliary. The controls and control cabinet are to be included with the afterburner price. The
afterburner will be assumed to be located on a structural steel base on the resin plant roof. No
modification to the building structural steel is required. A tie through the roof from the base to the
building steel is required. The base and tie in are part of the installed structural cost. All utilities are
available within 30 ft of the control cabinet, motor, and burner. Plant air is available at 100 psig and
located within 30 ft. Inert gas is available and will require 30 ft of piping from supply to ductwork. The
existing stack can be used for the diversion stack. A 10 ft exhaust stack will be mounted on the
afterburner, double manifold or exhaust fan. A total of thirty-five (35) feet of new exhaust ductwork and
twenty-five (25) feet of double manifold will be required.
321
-------
TABLE 80
CATALYTIC AFTERBURNER OPERATING CONDITIONS
FOR OPEN KETTLE SPECIFICATION
(Without Heat Exchange)
Small Large
Process Conditions
Kettles, Number 1 6
Size each, gal 300 300
Hcbn Emission Max, Ib/hr 22.5 135
Hcbn Emission Ave, Ib/hr 13.1 78.8
Total Exhaust Rate, SCFM 500 3,000
Exhaust Temperature, °F 80 80
Heat of Combustion of
Kettle Fume, Btu/lb 16,000 16,000
Hydrocarbon Concentration
Maximum, Btu/SCF 12 12
Average, Btu/SCF 7 7
Afterburner Without Heat Exchange
Unit Inlet, °F 280 280
Burner AT" from Fuel Gas, °F 260 260
* Burner AT from Flame Combustion, °F 60 60
Burner Outlet Temperature, °F 600 600
** Unit A7 from Thermal Combustion, °F 500 500
Unit Outlet Temperature, °F 1,100 1,100
Burner Duty, MM Btu/hr
Estimated Unit Removal Efficiency
"''Catalyst Face Velocity, SF/min
Catalyst Volume, ft3
* Assumes 10% fume combustion in burner flame
"Assumes 95% overall fume combustion
*** Basis 70°F exhaust temperature
322
-------
TABLE 81
CATALYTIC AFTERBURNER OPERATING CONDITIONS
FOR OPEN KETTLE SPECIFICATION
(Without Heat Exchange)
Small Large
Process Conditions
Kettles, Number 2 4
Size each, gal 300 300
Hcbn Emission Max, Ib/hr 45 90
Hcbn Emission Ave, Ib/hr 262 52
Total Exhaust Rate, SCFM 1,000 2,000
Exhaust Temperature, °F 80 80
Heat of Combustion of
Kettle Fume, Btu/lb 16,000 16,000
Hydrocarbon Concentration
Maximum, Btu/SCF 12 12
Average, Btu/SCF 7 7
Afterburner Without Heat Exchange
Unit Inlet, °F 280 280
Burner AT from Fuel Gas, °F 260 260
* Burner AT from Flame Combustion, °F 60 60
Burner Outlet Temperature, °F 600 600
** Unit AT from Thermal Combustion, °F 500 500
Unit Outlet Temperature, °F 1,100 1,100
Burner Duty, MM Btu/hr
Estimated Unit Removal Efficiency
***Cate/ysf Face Velocity, SF/min
Catalyst Volume, ft3
* Assumes 10% fume combustion in burner flame
** Assumes 95% overall fume combustion
*** Basis 70°F exhaust temperature
323
-------
TABLE 82
CITY COST INDICES
Average 1969 Construction Cost & Labor Indices
City
Albany, N.Y.
Albuquerque, N.M.
Amarillo, Tx.
Anchorage, Ak.
Atlanta, Ga.
Baltimore, Md.
Baton Rouge, La.
Birmingham, Al.
Boston, Ma.
Bridgeport, Ct.
Buffalo, N.Y.
Burlington, Vt.
Charlotte, N.C.
Chattanooga, Tn.
Chicago, III.
Cincinnati, Oh.
Cleveland, Oh.
Columbus, Oh.
Dallas, Tx.
Dayton, Oh.
Denver, Co.
Des Moines, la.
Detroit, Mi.
Edmonton, Cn.
El Paso, Tx.
Erie, Pa.
Evansville, In.
Grand Rapids, Mi.
Harrisburg, Pa.
Hartford, Ct.
Honolulu, Hi.
Houston, Tx.
Indianapolis, In.
Jackson, Ms.
Jacksonville, Fl.
Kansas City, Mo.
Knoxville, Tn.
Las Vegas, Nv.
Little Rock, Ar.
Los Angeles, Ca.
Louisville, Ky.
Madison, Wi.
Manchester, N.H.
Memphis, Tn.
Miami, Fl.
Index
Labor
98
86
87
131
88
90
83
79
106
104
104
86
70
81
107
108
121
106
86
100
94
93
117
80
77
98
93
103
90
104
99
92
97
73
78
94
82
115
78
113
92
95
89
83
98
Total
100
95
84
148
94
93
88
86
103
102
107
90
75
84
103
104
112
99
89
103
91
96
111
83
83
99
97
99
92
100
109
89
98
75
79
93
82
107
81
102
93
98
92
82
94
City
Milwaukee, Wi.
Minneapolis, Mn.
Mobile, Al.
Montreal, Can.
Nashville, Tn.
Newark, N.J.
New Haven, Ct.
New Orleans, La.
New York, N.Y.
Norfolk, Va.
Oklahoma City, Ok.
Omaha, Nb.
Philadelphia, Pa.
Phoenix, Az.
Pittsburgh, Pa.
Portland, Me.
Portland, Or.
Providence, R.I.
Richmond, Va.
Rochester, N.Y.
Rockford, III.
Sacramento, Ca.
St. Louis, Mo.
Salt Lake City, Ut.
San Antonio, Tx.
San Diego, Ca.
San Francisco, Ca.
Savannah, Ga.
Scranton, Pa.
Seattle, Wa.
Shreveport, La.
South Bend, In.
Spokane, Wa.
Springfield, Ma.
Syracuse, N.Y.
Tampa, Fl.
Toledo, Oh.
Toronto, Cn.
Trenton, N.J.
Tulsa, Ok.
Vancouver, Cn.
Washington, D.C.
Wichita, Ks.
Winnipeg, Cn.
Youngstown, Oh.
Index
Labor
103
99
94
77
79
122
102
89
132
73
82
90
106
101
110
82
102
98
76
110
109
117
110
93
82
111
124
72
94
104
82
99
101
99
105
81
105
84
114
85
81
98
85
62
107
Total
108
98
90
89
82
109
100
95
118
77
88
93
101
97
106
87
103
97
79
107
109
110
103
95
82
107
109
77
96
99
89
97
100
97
103
84
105
93
103
89
91
94
90
82
106
Historical Average
Year
1969
1968
1967
1966
1965
1964
1963
1962
1961
1960
1959
1958
1957
1956
1955
1954
1953
1952
1951
1950
1949
1948
1947
1946
1945
1944
1943
1942
1941
1940
1939
1938
1937
1936
1935
1934
1933
1932
1931
1930
1929
1928
1927
1926
1925
1924
Index
100
91
86
83
79
78
76
74
72
71
69
67
65
63
59
58
57
55
53
49
48
48
43
35
30
29
29
28
25
24
23
23
23
20
20
20
18
17
20
22
23
23
23
23
23
23
324
-------
TABLE 83
AVERAGE HOURLY LABOR RATES BY TRADE
Trade
Common Building Labor
Skilled Average
Helpers Average
Foremen (usually 35<^ over trade)
Bricklayers
Bricklayers Helpers
Carpenters
Cement Finishers
Electricians
Glaziers
Hoist Engineers
Lathers
Marble & Terrazzo Workers
Painters, Ordinary
Painters, Structural Steel
Paperhangers
Plasterers
Plasterers Helpers
Plumbers
Power Shovel or Crane Operator
Rodmen (Reinforcing)
Roofers, Composition
Roofers, Tile & Slate
Roofers Helpers (Composition)
Steamfitters
Sprinkler Installers
Structural Steel Workers
Tile Layers (Floor)
Tile Layers Helpers
Truck Drivers
Welders, Structural Steel
1970
$5.00
6.85
5.15
7.20
7.15
5.20
6.95
6.75
7.50
6.25
7.05
6.60
6.45
6.20
6.50
6.30
6.60
5.30
7.75
7.20
7.30
6.30
6.35
4.75
7.70
7.70
7.45
6.50
5.25
5.15
7.15
1969
$4.55
6.05
4.65
6.40
6.40
4.70
6.15
5.90
6.45
5.50
5.90
5.95
5.60
5.45
5.80
5.60
5.95
4.85
6.90
6.20
6.35
5.55
5.60
4.45
6.90
6.90
6.45
5.60
4.80
4.60
6.35
1968
$4.10
5.50
4.20
5.85
5.85
4.30
5.40
5.30
5.95
5.10
5.40
5.45
5.25
5.05
5.30
5.15
5.50
4.45
6.15
5.65
5.80
5.05
5.10
4.00
6.10
6.10
5.90
5.20
4.35
4.30
5.80
1967
$3.85
5.15
4.00
5.50
5.55
4.05
5.10
5.05
5.60
4.75
5.10
5.20
5.05
4.75
4.95
4.75
5.15
4.15
5.75
5.35
5.45
4.75
4.85
3.75
5.70
5.70
5.55
4.90
4.15
3.95
5.45
1966
$3.65
4.90
3.85
5.25
5.35
3.95
4.90
4.85
5.45
4.60
4.85
5.05
4.90
4.50
4.80
4.55
5.00
4.00
5.55
5.05
5.15
4.65
4.80
3.55
5.50
5.50
5.25
4.80
4.05
3.65
5.10
325
-------
TABLE 84
INSTALLATION AND OPERATING COST FOR BAGHOUSE
Type Collector
Type Pulse
Shaker
Bag type
Cloth, ft2
Air, CFM
Air/cloth ratio CFM/ft2
Capital cost, $
Installation cost, $
Total cost, $
Operating & Maintenance
Cost - 8 hr day, $
Pigment recovered, Ib/yr
Compressed air required
@ 100 psig
Polyester felt
452
3,800
8.4/1
3,900
2,900
6,800
350/yr
2,850
4.5 SCFM Ave.
9.0 SCFM Max.
Cotton sateen
1,900
3,800
2/1
4,500
3,400
7,900
245/yr
2,850
None
326
-------
FIGURE
CAPITAL COSTS
FOR
CATALYTIC AFTERBURNERS
WITHOUT HEAT EXCHANGE
500000
100000
CO
DC
o
Q
to*
O
O
Q.
<
o
10000
1000
AFTERBURNER
PLUS AUXILIARIES
AFTERBURNER ONLY
BASED ON DATA FROM
EPA CONTRACT NO. 68-02-0289
BASED ON DATA FROM
EPA CONTRACT NO. 68-02-0259
I I I I I I I I I
100
1000
GAS FLOW, SCFM
10000
30000
327
-------
FIGURE 65
CAPITAL COSTS
FOR
CATALYTIC AFTERBURNERS
WITH (23% EFFICIENT) HEAT EXCHANGE
500000
CO
cc
o
Q
k
O
O
a.
<
O
100000
10000
1000
sfTERBURh
ER WIT
AFTER
H E
BUR
XCH
•JET!
ANC
W
JEI
IT>
TURNKEY SYSTEM — "
*
1
P
E
LU
X<
3 AUXILIAF
CHANGER 0
NES-H
NLY-*
r-=
P=
***•
4?
**• •
^
>
^
^
&
^
x
r
x*
x
x
/
/
-?
e
r
>
^
X
r*
f
100 1000 10000 3000C
GAS FLOW,SCFM
328
-------
FIGURE
CAPITAL COSTS
FOR
THERMAL AFTERBURNERS
WITHOUT HEAT EXCHANGE
500000
100000
V)
cc
o
G
te
o
o
Q.
<
O
10000
1000
AFTERBURNER
PLUS AUXILIARIES
AFTERBURNER ONLY
100
1000
GAS FLOW, SCFM
10000
30000
329
-------
FIGURE 67
CAPITAL COSTS
FOR
THERMAL AFTERBURNER
WITH (42% EFFICIENT) HEAT EXCHANGE
500000
V)
cc
o
Q
to"
o
o
100000
10000
1000
-AFTERBU
I
RNER~]
WIT
AFTERE
^ E
JURN
XCh
ER
AN
Wl
3E
TH
TURNKEY SYS'
R
PL
•>
.U
(C
S AUXILIAR
HANGER 0
•EM--4
IE8-^
^JLY-^
^
r"
^
***
**
^
•^
^^
^
>
^
r^
^
jt
~*>
j*
s
•*
^
t
A
1
^
r
£
^
100 1000 10000 3000C
GAS FLOW, SCFM
330
-------
FIGURE 6S
TOTAL INSTALLED COSTS
FOR
THERMAL AND CATALYTIC AFTERBURNERS
500000
100000
GO
cc
o
Q
CO
O
O
0.
<
o
10000
1000
THERMAL WITH
42% HEAT EXCHANGE
CATALYTIC WITHOUT
- HEAT EXCHANGE -
THERMAL WITHOUT
HEAT EXCHANGE
CATALYTIC WITH
23 % HEAT EXCHANGE
100
1000
GAS FLOW.SCFM
10000
30000
331
-------
a:
_
o
o
in
o
o
ce
UJ
a.
o
FIGURE 69
DIRECT ANNUAL OPERATING COSTS
FOR
THERMAL AND CATALYTIC AFTERBURNERS
WITHOUT HEAT EXCHANGE
500000
100000
10000
1000
THERMAL
CATALYTIC
100
1000
GAS FLOW, SCFM
10000
30000
332
-------
FIGURE 70
TOTAL ANNUAL OPERATING COST
FOR
THERMAL AND CATALYTIC AFTERBURNERS WITHOUT HEAT EXCHANGE
500000
en
tr
o
Q
CO
o
o
CL
UJ
Q.
O
100000
10000
1000
THERMAL
CATALYTIC
100
1000
GAS FLOW, SCFM
10000
30000
333
-------
c/i
cc
o
Q
O
O
o
UJ
CL
O
FIGURE 71
DIRECT ANNUAL OPERATING COST
FOR
THERMAL AND CATALYTIC AFTERBURNERS WITH HEAT EXCHANGE
500000
100000
10000
1000
THERMAL
CATALYTIC
/"
-^
1000
GAS FLOW, SCFM
10000
30000
334
-------
FIGURE
TOTAL ANNUAL OPERATING COST
FOR
THERMAL AND CATALYTIC AFTERBURNERS WITH HEAT EXCHANGE
500000
a:
o
Q
V)
o
o
o
o:
LJ
Q.
o
100000
10000
1000
THERMAL
:ATALYTI
<
j(
^~
c
S
^
4
t
<*
/
\
i
/
s
>
>
(j
/
I
/
f
s
100 1000 10000 3000C
GAS FLOW, SCFM
335
-------
00
LU
_1
CD
H
B
T3
2
Q-
cn
.c
(0
'c
LL
"co
•^
to
T3
C
to _ —
O o
O Ij
_J -J
"^ r-T
> y
D "••
8 S
tt
V)
C
_o
"co
d
•a
^
o
a.
o
o
o
CO
p»*
in
r-
o
0
d
o
CM
in
o
CM
O
q
p-'
CM
•CO
1
O
o
o"
CD
O
o
R
in
o"
in
in
V)
o
q
CM
W
1
g
O
a>
CO
o
O 00 LO CM CD IT
in co co CN ^ .
r-^ d 06 in d CJ
oo ^- CM ^- in S
^" O> O> * — CO -
O5 00 ^f CO CD £2
CO CM CM CO g
' *
O 00 O 00
O CO O CO
d «-' P-: oo
C^ P**" ^* ^~
^ ^ 1 ^l 1 °.
co ^f r~ ^-
p**1 ^~ ^D
s
o in CM
in q | co |
CD CN CO
1 1 1 1 1
o o o
o in o
°. "^ 1 ^ 1
O CM CM
CO CO CM
co CM in
o o in CM CD co
in in CD CM •«*• co
P~ £35 00 00 O »t
00 CD CM O) LO CD
CN" ^f~ *t" in" CD" •*"
«- CM CN <- CO
CO
o in 05 CM oo
in q r^ co «-_
CD CN CN CO CO
1 1 1 1 1
§o o o o
o o in o
in o in co P~
p^ C3^ co in o)
CN p"* O5 P^ C7>
«- 00 If »-
OOOO LOOOOO
qqoq CMOOOO
«— co d cn T— crJoind
ooomin cocooco^
^o cyj rt- ^- p^ (^ CO O) *
COr^p^OO CDCOO5CDCO
«— CM CN CO «—
OOOO O O O
OOOO O O O
CM O O CN O LO O
•* CO t- «- O CD <*
co o oo r^ | | oo CD in
in CD CD oo oo o «—
in ^t in in cn cn co
III 1 1 1 1 1
O CO C^ CO CO
o o o o o
O_ O_ O^ LO_ (\ | | |
T CD" co" f" «—
T~" r^ co ^f ^f
r~ CM •— CD «—
o
o
d
in
in
in
*
o
o
o
in
in
in
in
*~
o
in
0
CO
1
o
o
o
o
LO
1
1
1
1
o
o
05
* —
CN
CN
O
0
O)
^~
T—
CN
CN
0
in
CD
CO
1
o
o
CO
o
CD
1
1
1
|
in
CM
^f
^~
*""..
CM
•
r-"
O
o
^
px
CO
Pv."
o
in
in
CN
d
f
s
»
LU
° «
° 0
0
-CL
5. -^ '~ ."^ fD CD — —
Ei-Soooi-O S
CT O
"S
5
_ <
°
~ "
cc
a>
CC
a ^s
o
0)
CC
2 ^
ii Hill
tu
CC
co
^
o
D
O.
"ca
S
336
-------
cu
•^
c
O
~™^
in
00
LU
i
_J
CQ
^
i™~
•4-1
o
-o
o
Cu
en
CU
'E
LL
CO
V)
T3
C
^~
$
3
Tl
O
£
U3
CU
"co
s
CO
H
CO
0
(—
cu
"co
"D • — •
Q 1
O QJ
i —
w
c
_o
"CD
V)
"O
D
0
Q.
CD
_
CO
"^
- 8
Q °
CU V.
u cu
'C Q-
Q. ' '
c
o
"CD
(D
(A
"D
§
0
CL
OOOC3>OCDOOOCOincOO^t O S
<-OOCOO'-OOO'-CNCNOCO O3 °P
csicicJ'-cscNiodcD'^odcDodincN r-' £2
inooooomcooococo<-^frCNO)^j- o t:
OCDOO CO*-'3-*-«3-C31OOOOCN 00 ^4.
r^ «- oo" o" o" o" co" in <- CD" t- o r-. f2
co»-o incN«-in r- »- in
«- CO ^
V> W (N
o o ^•ooocoincoO'd'O) r^
00 p COpppi-CNCNOCOO <*
^ ci r^cdco^-cdcicoiricN^ in
CO CO COOOCOCO'-'tfCNCn^t-^t ^~
O>|<- |CO'*'-'^-a)OOOOCNCO O
«- CN ciocom*-co*-o CN co
CN «- •- in <- co o
«— CN
fn ft* ~
ff.
So oooooomoo
|P jppppininr^oo
r-^ ci «-' co rr CD r-^ oo oo in CN
CN •* ^-CNCNCNCNT- >-•*
<&
o o o in o
o o o P~ in
1 1 1 1 1 1 co^ ^ en en to | |
in T- o CD CN
•3" CN CN CO
CN
O o «* ^ o r~~
^" O I^ O) O 1^
co_ •* | | ^ | | | | oq co, in
«- o CN o r~
OO CO CN CN CD
o o o o CN <- a>
COOOC33OOO COCO
O O O" ^ O •*' CD 00
C500 o o o
00, CN CD, | «* | | | | | | | |
in en «- CD
in CN <*• CN
CN
cu
•*-<
CO
2
CO
DC
liscellaneous
3 Ib/gal)
ds
jids (8 Ib/gal)
oi — 0"
c3 ~ to !_i
*2 o3 o o j
> Q S H <
o
c/3
•(->
c/>
"CD
us
vvdiei
Odorles
CO
01
CN
r-
.c
t^ 03
C/3 jjj
C "5
< h-
cu
c
_CU
X
100 (Solvesso 100;
150 (Solvesso 150;
Uj.U~
O O «
Aromati
Aromati
\/l\/1 l D
CD
-S5
^ q
CJ> OJ tQ
^ _Q CD """"'
E ~ QJ ^
1^1
£ £ co <
l_ c "> -C
; 3 i.S
£ I! 0) ~
> CO O 1-
ts & Miscellaneous
c
CI)
"o
C/3
"CD
•4-1
0
1-
CO
•4-J
o
•a
c
CD
(3
337
-------
CD
00
LJJ
CO
H
co
CO
0
CJ
_i
oc
LLJ
1-
g
LLJ
C3
^
^
o
Q.
68
LU
<
^
O
^
°~
<
z
z
-^
<
n
CO
O
_i
(3
1
1
—
5
O>
LL
O
1-
^
Q_
1-
0
Q
LLJ
CO
•^
LL
' '
CO
U
D
•a
o
it
CO
a>
.c
CO
'E
LL
CO
1
c
C/l
3
-§
o
CL
0>
CO
CO
^
"co
>
•a
>
Q
CD
^CJ
cL
£
c
CO
o
Oi
D
"co
>
-o'
>
Q
8
Cu
£
J-
D
a
o
(s^
CO
r-."
O
^/5
@
in
CO
d
-^5.
CD
CO
o"
|
|
|
00
o
ai
O5
°l
j"
0
LO
<*
CM
O
in
00
CM"
O
CO
O
CD
CD_
r-«"
CM
u
LO
^
CM
o
o
OD
CM"
CM
CO
ai
a>
0?
u
in
•*r
CD
00
O
in
CD
in"
CM
1
1
1
1 |
1 1
1 1
o o
CO 00
a> oo
«* o
CM O
00 CO
r- oo
r—
^ 5
O O
o in
co •^•
co en
CM CM
O O
o o
co ^r
00 00
r- oo
CM CO
1 1
1 1
1 1
^ ^3*
CM Tf
r^ *—
in o
00 <-
^- co
CM CM
2 ;§
o o
P~- CM
oo r-.
»- CM
^ ^
o o
o o
CM CM
in LO
o o
CM CM
I
1
1
00
o
CM
CO
o
00
2
o
00
00
o
o
CM
a>
,_
CD
in
CD
CM_
in
^/J
CD
1^.
ai
CO
CN"
in
a
o
CM
in
oo
o
CM
in
co
CD
O
O
co
co
I I I
1 1
CM O O O
oo qi CN|p
«*' o co •.
«~ CO 00 00
CM CD CM CO
r~r co"
o in o o
*)• r- p in
CD oo co r^-
o o o o
o o o o
°°. °. ^". ^N.
CM" CM o" T-
CM T- r- CD
CD "*
CM
in
5
CM"
v>
o o o
p CM •
o in •
CD CO •
°. °°. '
co" in"
co
o o o
p p p
in oo •^:
*d" r^ a)
O)
o
CO
o o
CM O
in o
in in
«l "1 I 1 1
00 00
o
CM
in
o
f^
CM
VI
,_
CO
00
CM
CO
^>
00
w
CO CD
CO CO
CM in
oo r-.
t- CO
I I I I
o
o
R
t—
Iff
8
d
o
CM
•*f"
CO
O
CM
r^
in
co"
o
CD
05
00
^f
O
CO
CO
CO
CO
co"
•^-
@ @) (g) ©I
o o o o
I CD CD CD CO
CO CO CO CO
in o r^ co
CM o r^- co
«* in o >-
«-' ro"
LLJ
Q
O
S
cu
O)
CO
J^
U
CO
a.
CO
o
O
i_
^O
'C
0)
— •—
o
... ^ — — ccccc
Sfi^mcnTOCDCOCDtD
§,0(20.00000
U
CO
o.
I S
-Lo
o
c
o
=s o t
c/5 *
in
O Q-
—
c.
$,
CO
O
i_ij
CD
N
CO
c
o
"io
CT
CD
(U
N
(75
*-t
CO
a
a.
CM
-^
CU
N
co
c
Q.
CO
O
t
CO
O
"co
o
h-
cu
N
(75
c
CM
x
c
o
co
CO CO
co r> r- in
E - o t «
3 CO — CO +•>
d O) [0 D C
Q in O O b.
o
o
QJ
O)
CO
O
CO
Q.
CO —
CO
o
-D
C
CO
O
338
-------
m
1-
z
_l
o_
LL
o
1—
in
O
O
>
*<.
<
03
d
^
g
_l
<
Z
Z
_J
Q_
|
LU
O
^
CO
0
_l
_l
<
Z
]
_l
OT
LL
,_
Q.
o
D
| | |
I
2
LL
""•"*
Z
C
u
-o
o
ct
OJ
SI
vt
'c
II
Is
1
c
G
-o
£
(/5
CU
•D
1 —
>
f
•4—
c
C
2
CO
S .£
ro JS
"CD
CO O
"co — -
W <
r cu
o £
"co ">
M
< ^
CD
C >
< g
> CJ
CO —
$ <
H CD
13 '*"'
° tr
1 1
C ___
<
S£
>• r^ r^
w
1 1 1 1
1 1 1 1
o o o o o
o in o o o
in co o_ o_ p.
w
D)
c
'o.
D.
w
c ^ §
o
o
T±
t
1
o
CO
in
1
I
I
oo"
1
1
o
o
CN
E
CU
r~
O O O
o o o
CM 00 O)
05 0) CO
1 1 1
o o o
00 CM CD
1^ r^ CM
1 1
1 1
| |
O O O
CM 00 «*
^ ^ '*'
| |
1 1
§o in
in r~-
oo in in
[ro
'E
-5
o
o
CM"
CM
v>
o
CO
oo"
£
o
CD
CM
co"
w
g,
s
OOtDCDCDCDO)
"3-OOOOCO
OOOCOOOOOCO
CO CO CD CD CD
CN «-«-«- T-
OJ «- CO
T- CO O
•*_ CO CO | I
o" co" CD"
in T-
o
o
CM
in"
CM
I I I I I I
O O O
O CM O
00 CO CD
^t CO *—
O
^t
O
CD"
ScS i §§§
O 00 O O O
CM CM CD CO CM
r^ co i *— «~ [^
00 CO
00
o o
l
a £
o-cocotocoOOOO
co
it
Q-
CM CO T- W
^ ^ ^ o
OJ +^
-o .22
CN CM
T- CM
«- CM
CO 1^
00 «-
in ^i-
•* ^-
^•" co"
CO
1
CN CM CM
CM CM CM
CM CM CM
r^ r^ r^
T t— T
CM
^ 1 1
i^
*-
1 1
CD
in
o
CO
^~
•a-
r-_
o"
"~
1
i^
CD
CM"
0j
CO ^° 1 00 00 CD CD f^*
in n" cs
p^* t — r-
w
1 1 1
3 CD LO r- CO
- T- r~~ in <-
II II
CM
CM
1 (N..
r~-
v —
1 1
CM
CM
CM
1^.
i —
1
CM
CD
CM_
CM"
CO
CD
O)
o"
CO
CO
1
CO
O)
o"
CO
CO
o
CD
c\
CO
O 00 O O O
00 CM CO CO O
^ m* I *~. *~. °°;.
CM" co" •*" •*" r-."
o o
III-l
y, u 3 -o ^j
— C r- *- CO
O S — I— Q-
CD (1) — —
~ LO CM CM CM CM CM
SQ-' ' o3 E
£ g c £ £ >- -2
Qi •— •— QJ CU CD O
V Jr y —- — o
-£ co cu — — co i:
O > OC LL LL O W
—
CO
-
COCMCMCMCMin T
CO
c o
Q CM LU O D I—
-a
c
CO
l_
a
339
-------
~O
CO
m,
CD"
CO
i--
co"
o
+^
Q.
^
&)
00
LO
(A
O 'TT
O O)
- CD
o" ^
o c
11
Is
LU
2
o"
CO
oS
o
oo"
CM
5.
CM
in
CM
in
CD"
CO
CD
CO
co"
o^
in
•a
CU
ro
^
CO
LU
CU
O)
c
1 1
CO
o
U
CD
O
a
E
LU
cu
6
§
o"
00
CO
CO
CM
cn"
^
p^
C3>
co"
in
in
in
CO
CD"
CD
CO
CO
(3)
CO
o"
00
2
cu
Q.
O
Z^
^^
S
•a
CO
CO
08
1_
(/)
D
^
<
CO
ro
-Q
5?
O
•f
O
s?
o
CO
c
o
CO
O
3
T3
E
a.
[ro
'^_
«
T3
C
_
68
cu
2
H
c
,1 located betwee
<:
CO
^^
5
8
c
CO
c
0)
4-<
c
'ro
^_
CO
3
CU
c
8
(A
i
"D
CO
o
^
o
cu
ro
CO
C
'co
cu
n~
m
ndustrii
•M
C/)
O
CO
"to
_Q
s?
o
CM
0
go
Q
00
c
o
CO
4-*
O
^^
0
tx
ro
^
CO
D
-a
c
—
08
cu
-a
ro
L
c
,1 located betwee
<
0
CO
O)
ro
CJ
ro
Q.
i_
cu
O)
ro
O
4-1
CU
D
T3
O)
handlin
CO
CO
CO
cu
*_
IT
cu
^_
1
b
^
O
CO
t—
accour
4-»
Q.
D
O
."2
tt
T3
C
CO
CO
_Q
6?
in
CM
0
4-<
s?
in
c
0
to
o
D
•D
0
CL
"ro
i-
«
-D
c
—
68
cu
•a
ro
^
c
,1 located betwee
<.
Q
cu
0
Q.
E
cu
g
intenan
CD
0>
C
o
c
o
-Q
c
CD
4-J
M—
v>
£•
•M
C
o
(A
OJ
-^
Q
O)
.E
._
8
cu
^-
O
C
D.
0)
CJ
X
cu
cu
0
J2
ro
m
CO
ro
o>
E
ro
OO
LU
in
-a
c
o
CJ
cu
00
I
5?
a
CM
E
o
L.
4—
00
CO
co'
M—
o
cu
o
cu
D)
ro
§ «'
« c
•a o
cu *-
_o ^:
ro -5
o O
cu <—
4-1
ro i;
II
O
-------
TABLE 88
MODEL PLANT
DEPRECIATION SCHEDULE
(Finished Output 1.9 Million Gallons)
Asset Straight Annual
Capitalized Line Depr.
Value Depr. % Amount
LAND $ 100,000
BUILDING & SITE PREPARATION:
Raw Material Warehouse 219,700 1.25 $ 2,746
Manufacturing Building 912,450 1.25 11,406
Finished Goods Warehouse 526,700 1.25 6,584
Site Preparation 167,100 1.25 2,089
Engineering 217,200 1.25 2,715
Other Building Costs 310,300 1.25 2,879
Total Buildings, Etc. $2,353,450 $ 29,419
EQUIPMENT & UTILITIES:
Paint Plant $ 521,900 10.00 $ 52,190
Resin Plant 222,500 10.00 22,250
Storage 211,600 10.00 21,160
Miscellaneous 142,750 10.00 14,275
Utilities 178,400 10.00 17,840
Total Equipment & Utilities $1,277,150 $ 127,715
GRAND TOTAL $3,730,600 ===== $ 157,134
Note:
1. All freight & sales tax was assumed to be expensed in prior year.
2. Assumption is that plant is operating and on stream: All start-up costs have expired in prior years.
3. Depreciation will be allocated between trade and industrial products on 60% to 40% basis.
Trade 60% $ 94,280
Industrial 40% 62,854
$157,134
341
-------
TABLE 89
MODEL PLANT
INCOME STATEMENT
(Year Ending December 31, 19-2)
Trade Industrial
Sales Sales Total
Products Products All
REVENUE (NET)
A) 1,140,000 gal @$3.78/gal $4,309,200 $4,309,200
A) 760,000 gal @ $3.02/gal $2,295,200 2,295,200
Total Revenue $4,309,200 $2,295,200 $6,604,400
COST OF GOODS SOLD
Raw Materials (See Table 85) $1,550,228 $1,203,045 $2,753,273
Packaging (See Table 86) 447,078 184,828 631,906
Salaries & Wages (See Table 87) 447,051 233,010 680,061
Depreciation (See Table 88) 94,280 62,854 157,134
B) Other Manufacturing Costs
Trade-Estimated @ 500/gal 570,000 570,000
Industrial-Estimated @ 300/gal 228,000 228,000
Total Cost of Goods Sold $3,108,637 $1,911,737 $5,020,374
Gross Profit $1,200,563 $ 383,463 $1,584,026
SELLING, GENERAL & ADMINISTRATIVE EXPENSES
Trade-Estimated @ 15% of Sales $ 646,380 $ 646,380
Industrial-Estimated @ 5% of Sales $ 114,761 114,761
Total S, G & A Expense $ 646,380 $ 114,761 $ 761,141
Operating Income $ 554,183 $ 268,702 $ 822,885
INTEREST EXPENSE
Long-Term Debt $ 151,200 $ 100,800 $ 252,000
Income Before Taxes $ 402,983 $ 167,902 $ 570,885
INCOME TAXES
Federal Income Tax (50%) $ 201,492 $ 83,951 $ 285,443
State & Local Taxes (10%) $ 40,298 $ 16,790 $ 57,088
Total Taxes $ 241,790 $ 100,741 $ 342,531
NET INCOME $ 161,193 $ 67,161 228,354
342
-------
TABLE 89
(Continued)
MODEL PLANT
INCpME STATEMENT
(Year Ending December 31, 19-2)
Notes:
A) Selling price calculated as follows using data from current industrial reports, M28F - - year 1971.
TRADE -$1,563 million divided by 431 million gal = $3.63/gal
$3.63/gal times growth rate of 4% = $3.78/gal
INDUSTRIAL -$1,268 million divided by 443 million gal = $2.86/gal
$2.68/gal times growth rate of 5.5% = $3.02/gal
B) The estimated other manufacturing costs is to cover such items as:
1. Repair parts which are expensed
2. Operating, heating and utility costs
3. Property insurance and taxes
4. Miscellaneous operating expenses
343
-------
o
cn
LLJ
CO
h-
co
L1J r-
-J Z
~^ LLJ
Q i_ S
LU — j Q;
o 1— J/5
CO ~3
Z O <
O co 3
Sol
0 < o
o- z; s?
HO10
^— 1 ^^
^, .— 1 f S
^J cn k
LLJ ,_' >-
Q 1—
2 LL
S <
CO
O O ^f
+2
o
h-
O O ^"
CO_ CD_ CO_
oo" oo" *-"
P*- P^- CO
CM CM
CU
Q
O
~7
in o in
CO O CO
p-._ cn »-__
CM" CM" in"
«- CM 00
in o o
p-_ cn CD
03" CM" in"
CM CM
in o in
ti
O
Q.
CO
cn
D
^
—
—>
00 O «-
o" CM" CM"
<- CM «-
in o
o" CM" CN"
in CM CM
in o in
00 O CO
O 00 00^
oo" CM" o"
CM CM in
£88
00^ OD_ CO^
CM" CN" in
co CM in
in o in
•vf 00 *~
*}•" CM" in"
t- CM CO
O O O
00 O CM
^» w^ ix^
^- CM CO
T- CM in
in o o
oo o in
oo_ oo_ p_
cn CN oo
CM ^r
CU
CO
"^
E
D
c s 1
o 2 a,
» ,S c
c 0 0 S
'co 1 1 ™
°- J, c °°
CD *-* O >•
3) C -r; i-
2 S " 2
.2 E 3 £
I Q. "D 0)
>., •- 0 >
*"» _L^ i_ £
£2 CO Q. —
to
<* o
CM
CO
in
o
in"
r—
in
r—
cn
oo"
"*
s
r-
oo"
CO
in
pv
o
in
in o
cn
in"
CM
o
00
CO
co"
in
CM
CO
<—
o
CO
o
o"
CN
CO
CO
in
CO
^>
^_"
o
in
,.
0
CN
CM"
O O
in
00
co"
00
in
§
S
§
CM
8
in
in
cn
oo"
o
o
in
o"
0
PX,
^J"
CD
CM
03
CM"
in
CM
«~
in
00
*~
o
in
00
o
00
§
m
—
CO
CO
CO
O c
§0
—
J2' I (S
- c 1
.E m "
.E CD
> .^ ^
^~ o
•o 'p
to >- ff.
+_i ^ C/J
.2 1
o in
O CM
O3 00
in" «-'
in o
CM CO
oo_ o_
<-" CM"
in o
CM in
oo_ "*.
^-." ,^*
in in
CM cn
00^ CO
t—"
in
CM
CO 1
* !
r—
in o
CM ^~
, —
in in
CM CO
CO 00
,_"
in o
CM «-
*~ *"
in in
CM O
oo^ cn_
»- CM"
in in
CM O
CO CM
«-" co"
in in
CM in
«-* co"
in o
CM 00
CO CM
T-* CO"
in in
CM (O
»-" CM"
cn
o >
S '«
a -5
1 E
1"
tJ —
n cn
T3
o >
rt~ ^
O O O CM
CM
CO
CM"
o o •sr
•*" •*" o"
CN CM CO
in in
in
CM
CO
CM"
in
CM
CM"
o o o
co o cn
^•" co" p~.~
CM ^ CO
o o o
cn o CM
oo" co" oo"
o o o o
cn
r-T
in
cn
10
00
cn
8
< —
,_r
cn o r-
co i""*1 co
o" co" oo"
CM f CM
O O
^- p^ i
oo" co"
o o o
CD O <-
CM P^ P^
«-" co" •*"
in •*
in o o
^- o P-
°\ ^ PH.
't" oo" CM"
P- •Sf r-
in o o in
o
CM"
8
oo"
o
o
"*
o
in
^j."
in
P-.
o_
o T-
CO P~ CO
co" co" co"
in f 03
o o o
O O T-
t-~ CO" CO"
co ^r o
o o o
*— O t—
r-" co" ^r"
CM ^- CM
o o o
«- O CM
cn^ r*-_ oo
co" co" co"
CM <3" O
o o o
CO O CO
oo" co" CM"
<- •* cn
CO
__ (/3 r- ' — -
ca c k QJ
oi O ^ .>
LO CO "m
p5 jp (j t3 -5
- u- | 1 E
c o j2 § 0
~ 'C C •:-: — '
(U tD 'JJ —
S c E ^ co
^" ^_ O (J
X 'p O >
CO ^ r7\ n~ CI
o ™
CM o o in
CD
00
co"
00
o o cn
co cn oq
oo co in
CM CM CM
CN CN
O
in
co"
O3
o
CN
^-"
P-
o in in
CM in in
*t co r-
o" oo" co"
T- r- CM
in o o
p^ CD CM
o co to
r-T oo" o"
T- CN
o in o in
CO
cn"
o
CM
co"
CM
o
CO
cn
o"
CO
o
cn
^j-
oo"
CO
CM CD 00
cn co^ oq
oo" oo" 03"
•—
o in
CD in
CD CO I
o" oo"
CM t-
in o in
P~ CO O
°°. P. P
t-" oo" CM"
CN «-
O O O
cn CD CM
cn co CM
r-" oo" in"
CO «-
in o in o
CO
cn"
CO
in
CO
CM"
CM
O
00
CM"
00
o
00
,_"
in
o
cS
in"
CO
o
in
CM
oo"
^
CM in in
oo^ co_ in^
»- 00 CO
o o in
t CO <-
cn co CM
CM" oo" i-"
CM t- ^>
o o in
cn CD a>
^" *Q. ^.
co" oo" in"
CM r- ^f
in in in
CM in CM
00 CO CO
«-' co" oo"
T- t- m
in o in
oo co cn
«-" oo" co"
'-'-•*
in o o
cn co CM
O CO 00
oo" oo" 03"
T- CO
—
CO
"to
cn w ~
V) r- 0}
o c 5 >
CM 0 .2 +3
^ "to ra —
(5 "> (3 ^ c
_• o i i i
'•—• '£ c '43
E a> tj "S
r- rn C ^ CQ
o ^D .
— "o S P 2
to > ^ rt c
+_! ^ W7 LJ-
£ 1
o
cn
CD
P-"
CO
o
in
cn
cn
CO
in
P^
^ —
00
o
CO
o
t— **
CM
in
cn
« —
,_"
«—
8
CM
oo"
•—
in
^~
CD"
^~
in
^j*
i"\
CT>
CN
O
5-
CN"
in
§
CO
co"
in
o
CM
in
^r
CO
o
03
ps.
in
m
LO
o"
in
—
CO
co
"co
cn
in
03
* ^
^
•
o
^
CD
+-i
O
344
-------
CD
C
C
B
o
CD
CO
<
CO
O
1-
CD
Q
^"
O
H
o
0
Q.
&
CT)
<
—
—3
C
^
>.
CD
^»
jj
a
<^
^j
CD
^
(U
LL
C
CD
—)
0
CN
CN"
o
o
m
r^
st"
o
CO
co.
co"
o
CN
^
in
CO
CD"
o
CD
cn
CD"
o
CO
•*"
o
CO
cn
co"
in
^-
o"
o
o
CN"
o
CD
m.
in"
in
CO
CM.
in"
in
00
CO
co"
8
CN
CN"
o
o
CN
in
oo"
in
^—
in
oo"
in
in
oo"
o
CN
in
oo"
in
r—
in
oo"
in
in
co"
o
CN
in
co"
in
in
co"
in
in
co"
o
CM
in
oo"
in
in.
co"
in
r—
in
co"
CT
>
5;
00.
*
g
t— .
co"
o
00
CO
CD"
in
CD
•*"
1
in
r—
CD
O
CD
CO
CM"
in
CO
co"
in
00
oo"
*~
in
o
CO
o"
CM
o
CD
CO
•*"
CN
§
CM.
T
CN
O
CO
cn
* —
di
<
in
cn
CD"
t—
§
CN
oo"
O
CD
2
in
o
CD.
CD"
o
in"
in
CM
o
co"
8
in
r-."
in
CD
in
co"
in
in
CD
co"
CM
in
°l
in
CM
O
o
in
CD"
CN
o
r^-
co.
co"
CN
0
CD
£3
co"
CN
co
o o
o o
O O CO
o" o" «-'
*d- *3- CO
m
CD
«- m o
T- CD r-
in
o
CD
§in
CN
O CD
CM m CD
CD CD CM
«- CD CD
O O
r~ o
CD" in"
T- CD
O O
t o
CO
^f- (
03 •
CO CD
T- CN
CM
o o
CD CD
CO CO
to" ;
«- CN
§ §
•*. ^
^ CD
O CD
o o o
oo co in
CM CO O
oo" oo" o"
o in in
in o
in •»
in co i
oo" co"
CD m
o
m
o
in
CD
o o
^r co
CN I
o" m
«- o
in o in m
CO CO CM CN
CM. CO CD CO
CD" cn" CD" *-"
CM in CM CM
CM «-
in
in
o o
^3" CO
co in
co" co" •*"
i^ in CD
CN <-
o o in
o co «a-
O 00
co in
^s
O CO
co" oo"
cn in
o o
CD *3"
in co.
co" oo"
CN in
o o
O CO
00. CO.
<-" co"
CN in
in o
«* CO
in co_
*t~ co"
o in
o
CO
in
o"
CO
§ S
CN CO
«- CO
CO CN
CM CO
in m
in m
«- co"
r~ co
CM co
in
p~
CD
in
co
CO CO
CM CO
in in
Bin
CD
CM CM
CO
(fl
CO
CO
CO
CO
c
LU
(/)
(/)
_o
a
CD
o -V o
I 2
0) o
II
C/5 0-
I
CJ
"ro
CQ
U
c
CO
4-1
o
_co
CO
CO
CO
•u
CO
c 1-
_o 2
"co ra
5 ^
I I
u> CI
p c
fl
in
ol J= ™
= "S <2
CD
vu m 3
^ \ I
$ §£
c
cu
ii
CO CL —
co
CO
O
o
o
co"
CO
"5
c
a
"co
CO
o
CL
CO
•(-•
o
345
-------
TABLE 91
MODEL PLANT
BALANCE SHEET
(Year Ending December 31, 19-2)
ASSETS
Current Assets
Cash
Accounts Receivable (Dec. Sales)
Inventories:
Finished Goods
Trade-203,360 gal @ $2.73/gal
lndustrial-38,000 gal @ 2.52/gal
Raw Materials (Vt2 Of Annual Cost)
Packages, Etc. (1/12 Of Annual Cost)
Total Inventories
Total Current Assets
Property, Plant & Equipment
Land
Buildings
Equipment
Total (On The Basis Of Cost)
Less Allowance For Depreciation
Total Property, Plant & Equipment
($ 14,512)
391,929
$
555,173
95,760
229,439
52,659
$
933,031
$ 100,000
2,353,450
1,277,150
$ 1,310,448
$ 3,730,600
314,268
$ 3,416,332
TOTAL ASSETS
$ 4,726,780
LIABILITIES & SHAREHOLDERS' EQUITY
Current Liabilities
Accounts Payable (Via Of R.M. & Pkg. Cost)
Income Taxes (1/4 Of 342,531)
Total Current Liabilities
Lortg Term Debt
Bank Loan ($4,000,000 - - 10 yr Term - 7%)
Shareholders' Equity
Common Stock
100,000 Shares® $10
Retained Earnings
Total Shareholders' Equity
TOTAL LIABILITIES & SHAREHOLDERS' EQUITY
$ 282,098
85,633
$ 367,731
3,200,000
$ 1,000,000
159,049
$ 1,159,049
$ 4,726,780
346
-------
TABLE 92
MODEL PLANT
CASH FLOW STATEMENT
(1.9 Million Gallons Sold & Manufactured)
(Year 19-2)
REVENUE: NET $6,604,400
EXPENDITURES (CASH)
Raw Materials, Packages, Etc. $3,385,179
Salaries & Wages 680,061
Other Mfg. Expenses 798,000
S, G&A 761,141
Interest Expense 252,000
Federal Income Tax 285,443
State & Local Taxes 57,088
Principal Payment Long-Term Debt 400,000
6,618,912
NET CASH FLOW $ (14,512)
347
-------
TABLE 93
MODEL PLANT
RETURN ON INVESTMENT
(Year 19-2)
Return on investment is calculated using four of the more common methods. The first three methods are
essentially asset efficiency measurements while the last method centers attention on the rate of return
that will be earned by the business owners.
1. Total Gross Assets Available Method
Income Before Interest $329,154
= 6.53%
Total Gross Assets $5,041,048
2. Total Net Assets Available Method
Income Before Interest $329,154
Total Net Assets $4,726,780
3. Stockholders' Equity Plus Long-Term Debt Method
Income Before Interest $329,154
•= 7.55%
Stockholders' Equity Plus Long Term Debt $4,359,049
4. Stockholders' Equity Method
Net Income $228,354 .= 1970o/o
Stockholders'Equity $1,159,049
348
-------
-a _
c ro _
(D \-
0.
O
Q
LU
I
O oo
.— oo
fc O
I 5
CD .-
= P
•E E
x
LU
_ CD
0)
O
O
O
in"
in
en
o
o
o
co"
oo
CM
o
o
CM
r-
o
o
o
CM
o
o
co
o
o
-2 TO
u- a
o
o
o
CM"
co
CM
QJ
C/3
CO
CO
O
i_
i
0
£
+2
c
\-
oooooooooooo o
oooooooooooo o
coi^cDooco*-inoocMr^coin co
'-cocoin-tf^cNinT-m'-oo oo
•* — 25 >*
oooooooooo o o
CO CO CO CO CO CO CO CO CO CO CO CO
o co in *— 000*— ^fco«— | co ^
CO •* 00
o ooo oooo
o ooo oooo
1 oo j *— ^ o co^ co co^ oo in_
«-" <-" CM" «-" «-" o>"
c
Q) Q) ^~
>. c E? = £
o §0_|m 2^.^
-Ci-™ — ^"D^^-^^^' -^
CTOJ^CD^CU^CUCOICDJPI ^
-i-So>-lJS-iSE^SO P
"TO
4-<
0
1-
o
o
CO.
co"
00
o
o
l~^~
«•»
CO
o
o
CM
CD"
00
o
o
en
co"
CM
CM
8
co_
"
o
o
in"
CO
8
CM
in
o
o
in
^_"
^>
CM
00
+-•
—
1 —
o3
V)
0)
TO
CO
c/)~
CD
+-•
'S.
TO
O
H
O
O
in
00
o
o
,_"
1
1
8
CO
o
o
CN
1^
1
8
CM
1^
00
*~ CO
o >
<3 O
^
"o
C/3
0 0
O O
CM r^
00 CN
o o
o o
in CN
in" r-~
o o
o o
CO CD
1 1
o o
o o
CM «-
o o
o o
r^ in
CM in
1 1
o o
o o
r^ in
CM in
i— QJ
C3 DC
O O
O O
00 CO
CM r^
O O
o o
CO CO
^f CN
O 0
^D ^3
in co
CM «-
1 1
O O
o o
CO O
CNJ «-
88
O CO
oo in
1
o o
o o
O CO
oo in
~m
o ^
m 6
O 0
0 0
•- oo
CO
o o
o o
CM"
0
^ 1
CM
1 1
o
o
1 °°
o
o
"* 1
CO
1 1
o
o
^" 1
CO
O CD
ro 13
CD GO
8
co"
in
o
o
CD
^j-"
CM
O
8
co"
8
co"
o
o
CM"
CO
1
8
* —
CM"
CO
oo
1_
_p
O
0
Q
00
"TO
O
t-
O
O
§
-
0
o
0_
CM"
^f
O
o
CN
CM"
O
o
O
co"
CM
CM
O
O
in"
o
o
oo"
O)
8
CN
in
o
o
co"
i^
CM
00
Q)
TO
CO
T3
TO
k_
t
<
"ro
O
H
—
1
TO
-°
>
TO
oo
|S
oo
"D
C
._
•+-
0
^0
b
in
c.
~
o
^
"D
O
Q.
-Q
c
g
O
-a
^
TO
CD
O
C
00
ro
'i-
D
T3
—
O
8
o
CO
4-»
Q.
O
s
.c
'E
LL
C
ro
Q_
"TO
4^
I—
O
O
o
o"
o
en
^
>
z
s?
o
in
/o\
(s)
.D
O
O
o_
CO
CM
^ —
CO
c
CO
0)
oc
"D
v
"o
C;
0
O
D
T3
O
it
"S
C
CD
C
349
-------
ce
z<
< z
V-
t-
LU
ce
O
UJ
to
u
X
CO
a
ce
o
to a
ce =•
LU LU CL.
a. to •—
^ X X
co 3 CO
(M
u.
X
in
VI
< ZLU
=3 C X
O o o
OK
tO LL
ce X
CC
LU
Q. h-
3 tO
to _
STOCKrEN
02)
STOCKMEN
02)
1ST SHIFT
TECHNICIAN
d\
FILLERS
SHADER AND
ADJUSTER(I)
1J f!
3 ™
* C
£ Z
—
sT
5
U
2ND S.HIFT
TECHNICIAN
m
ce
LU
to
0
58 HOURLY PAID PEOPLE
12 SALARIED PAID PEOPLE
70 PEOPLE IN TOTAL
1
LUO
M
P
LU
_J
CL
u z
tO LU
•si z
z:
1 1 1
1
i-
tO LU to
cea=i
LU< C
_jce z
-It- —
•— l
LL. CVJCV
SHADER AND
ABJUSTER(I)
ce
LU
to
o
o."^
o
s
LU
X
1"
LU
o-
i-
Z
—i to
to or
LU CL
or o
UJ
s &
tO LU
z:
1 1 1
MIXERS
z —
— CO UJ
co ce <_>
LU O. LU
ce o D-
CO
z
o
CD
I—
Ct Z
< o
X —
O _l
0>
2:0:0.
C h-
o o
<
u- Q
LU s:
Q
350
-------
TABLE 95
MODEL PAINT PLANT COST, DOLLARS
Total
Item Material Total
Summary
Equipment
Paint Plant 521,900
Resin Plant 222,500
Storage 211,600
Miscellaneous 142,750
Sub-Total 1,098,750 1,098,750
Utilities 178,400 178,400
Sub-Total 1,277,150 1,277,150
Building
Sitework 167,100
Part I - Raw Material Warehouse 219,700
Part II - Manufacturing Building 919,450
Part III - Finished Goods Warehouse 526,700
Sub-Total 1,825,950 1,825,950
Total 3,103,100
Freight and Sales Tax 124,400
Engineering - 7% 217,200
Contingency - 10% 310,300
Project Total* 3,755,000
* Based on Chicago ENR Construction Cost Index of 1964 for December, 1972.
351
-------
TABLE 96
MODEL PAINT PLANT COST, DOLLARS
Item
Total
Material
Total
Equipment and Utilities Summary
Equipment
Paint Plant
Pebble Mills
Ball Mills
Sand Mills
High Speed Dispersers
Mixing and Finishing Tanks
Filling and Packaging
Laboratory
Sub-Total
Resin Plant
Storage
Miscellaneous
Sub-Total
Utilities
Total
128,150
37,500
26,500
44,800
166,450
105,000
13,500
521,900 521,900
222,500
211,600
142,750
576,850
178,400
576,850
178,400
1,277,150
352
-------
TABLE 97
MODEL CONTROLLED PLANT
DEPRECIATION SCHEDULE
(Finished Output 1.9 Million Gallons)
Asset Straight Annual
Capitalized Line Depr.
Value Depr. % Amount
LAND $ 100,000
BUILDING & SITE PREPARATION:
Raw Material Warehouse 219,700 1.25 $ 2,746
Manufacturing Building 912,450 1.25 11,406
Finished Goods Warehouse 526,700 1.25 6,584
Site Preparation 167,100 1.25 2,089
Engineering 217,200 1.25 2,715
Other Building Costs 310,300 1.25 2,879
Total Buildings, Etc. $2,353,450 $ 29,419
EQUIPMENT & UTILITIES:
Paint Plant $ 521,900 10.00 $ 52,190
Resin Plant 222,500 10.00 22,250
Storage 211,600 10.00 21,160
Miscellaneous 142,750 10.00 14,275
Utilities 178,400 10.00 17,840
Resin Plant Afterburner 26,000 10.00 2,600
Paint Plant Baghouse 7,350 10.00 735
Total Equipment & Utilities $1,310,500 $ 160,469
GRAND TOTAL $3,763,950 $ 160,469
Note:
1. All freight & sales tax was assumed to be expensed in prior year.
2. Assumption is that plant is operating and on stream: All start-up costs have expired in prior years.
3. Depreciation will be allocated between trade and industrial products on 60% to 40% basis.
Trade 60% $ 96,281
Industrial 40% 64,188
$160,469
353
-------
TABLE 98
MODEL CONTROLLED PLANT
INCOME STATEMENT
(Year Ending December 31, 19-2)
REVENUE (NET)
A) 1,140,000 gal @ $3.78/gal
A) 760,000 gal @ $3.02/gal
Total Revenue
COST OF GOODS SOLD
Raw Materials (See Table 85)
Packaging (See Table 86)
Salaries & Wages (See Table 87)
Depreciation (See Table 88)
B) Other Manufacturing Costs
Trade-Estimated @ 500/gal
Industrial-Estimated @ 300/gal
C) Operating Cost-Resin Reactor Afterburner
C) Baghouse Operating Cost
Gross Profit
Trade
Sales
Products
$4,309,200
Industrial
Sales
Products
Total
All
$4,309,200
$2,295,200 2,295,200
$4,309,200 $2,295,200 $6,604,400
$1,550,228 $1,203,045 $2,753,273
447,078 184,828 631,906
447,051 233,010 680,061
96,281 64,188 160,469
VI U,UUU
4,476
192
228,000
1,119
106
228,000
5,595
298
$3,115,306 $1,914,296 $5,029,602
$1,193,894 $ 380,904 $1,574,798
SELLING, GENERAL & ADMINISTRATIVE EXPENSES
Trade-Estimated @ 15% of Sales
Industrial-Estimated @ 5% of Sales
Total S, G & A Expense
Operating Income
INTEREST EXPENSE
Long-Term Debt
Income Before Taxes
INCOME TAXES
Federal Income Tax (50%)
State & Local Taxes (10%)
Total Taxes
NET INCOME
$ 646,380
$ 114,761
$ 646,380
114,761
$ 646,380 $ 114,761 $ 761,141
$ 547,514 $ 266,143 $ 813,657
$ 152,461 $ 101,641 $ 254,101
$ 395,053 $ 164,503 $ 559,556
$ 197,527 $ 82,251 $ 279,778
39,505 16,450 55,955
$ 237,032 $ 98,701 $ 335,733
$ 158,021 $ 65,802 $ 223,823
354
-------
TABLE 98
(Continued)
MODEL CONTROLLED PLANT
INCOME STATEMENT
(Year Ending December 31, 19-2)
Notes:
A) Selling price calculated as follows using data from current industrial reports, M28F - - year 1971.
TRADE -$1,563 million divided by 431 million gal = $3.63/gal
$3.63/gal times growth rate of 4% = $3.78/gal
INDUSTRIAL -$1,268 million divided by 443 million gal = $2.86/gal
$2.68/gal times growth rate of 5.5% = $3.02/gal
B) The estimated other manufacturing costs is to cover such items as:
1. Repair parts which are expensed
2. Operating, heating and utility costs
3. Property insurance and taxes
Cx o T bl 10? 4' Miscellaneous operating expenses
355
-------
TABLE 99
MODEL CONTROLLED PLANT
BALANCE SHEET
(Year Ending December 31, 19-2)
ASSETS
Current Assets
Cash
Accounts Receivable (Dec. Sales)
Inventories:
Finished Goods
Trade-203,360 gal @ $2.73/gal
lndustrial-38,000 gal @ 2.52/gal
Raw Materials (Viz Of Annual Cost)
Packages, Etc. (Vi2 Of Annual Cost)
Total Inventories
Total Current Assets
Property, Plant & Equipment
Land
Buildings
Equipment
Air Emission Control Devices
Total (On The Basis Of Cost)
Less Allowance For Depreciation
Total Property, Plant & Equipment
($
$ 100,000
2,353,450
1,277,150
33,350
19,043)
391,929
$
555,173
95,760
229,439
52,659
$
933,031
$ 1,305,917
$ 3,763,950
320,938
$ 3,443,012
TOTAL ASSETS
$ 4,748,929
LIABILITIES & SHAREHOLDERS' EQUITY
Current Liabilities
Accounts Payable (Vi2 Of R.M. & Pkg. Cost)
Income Taxes (1/4 Of 335,733)
Total Current Liabilities
Long-Term Debt
Bank Loan ($4,033,350 - - 10 yr Term - 7%)
Shareholders' Equity
Common Stock
100,000 Shares @ $10
Retained Earnings
Total Shareholders' Equity
$ 282,098
83,933
$ 366,031
3,226,680
$ 1,000,000
156,218
1,156,218
TOTAL LIABILITIES & SHAREHOLDERS' EQUITY
$ 4,748,929
356
-------
TABLE 100
MODEL CONTROLLED PLANT
CASH FLOW STATEMENT
(1.9 Million Gallons Sold & Manufactured)
(Year 19-2)
REVENUE: NET $6,604,400
EXPENDITURES (CASH)
Raw Materials, Packages, Etc. $3,385,179
Salaries & Wages . 680,061
Other Mfg. Expenses 798,000
S, G&A 761,141
Interest Expense 254,101
Federal Income Tax 279,778
State & Local Taxes 55,955
Principal Payment Long-Term Debt 403,335
Operating Cost of Emission Controls 5,893
6,623,443
NET CASH FLOW $ (19,043)
357
-------
TABLE 101
MODEL CONTROLLED PLANT
RETURN ON INVESTMENT
(Year 19-2)
Return on investment is calculated using four of the more common methods. The first three methods are
essentially asset efficiency measurements while the last method centers attention on the rate of return
that will be earned by the business owners.
1) Total Gross Assets Available Method
Income Before Interest $325,463 _
— 6.42 /o
Total Gross Assets $5,069,867
2) Total Net Assets Available Method
Income Before Interest $325,463 _ 0_0.
= 6.85%
Total Net Assets $4,748,929
3) Stockholders' Equity Plus Long-Term Debt Method
Income Before Interest $325,463 _
• — / .T-O /o
Stockholders' Equity Plus Long-Term Debt $4,382,898
4) Stockholders' Equity Method
Net Income $223,823
Stockholders'Equity $1,156,218
= 19.36%
358
-------
TABLE 102
MODEL CONTROLLED PLANT
AIR EMISSION CONTROL DEVICES
RESIN REACTOR AFTERBURNER:
1200 SCFM
Thermal
Type
1200 SCFM
Catalytic
Type
Average
of Two
A) Capital Cost - (Includes Installation) $25,000 $27,000 $26,000
B) Operating Cost Per Year - 16 hrs/day 6,850 4,340 5,595
Effect On Yield None None None
Note (A)
Capital cost is assumed to be financed with long-term debt and as part of initial financing arrangements.
Note (B)
Operating cost is apportioned between trade sales and industrial finishes. Products on same ratio as
manufactured resins are used within the finished products. This will be 80% trade and 20% industrial.
PAINT PLANT BAGHOUSE:
C) Capital Cost - (Includes Installation)
D) Operating Cost Per Year - 8 hr/day
(Includes Maintenance, Utilities, Etc.)
D) Pigment Recovered Per Year
Note (C)
Pulse
Test
Type
$ 6,800
350
Shaker
Type
$ 7,900
245
2,850 Ib 2,850 Ib
Average
of Two
$ 7,350
298
2,850 Ib
Same assumption on capital cost as indicated in Note A above.
Note (D)
Operating cost and credit for pigment recovered will be apportioned between trade sales and industrial
finishes on the same ratio as pigment used between the two. This will be 64.5% for trade and 35.5% for
industrials.
359
-------
360
-------
CHAPTER 8
PIGMENT INDUSTRY
I. INTRODUCTION
Pigments represent one of the major building blocks in the formulation of an acceptable
coating. Historically, the use of coloring agents was largely for decorative effect. While this is still
of major importance today, pigments are now recognized to produce a variety of other important
properties in paint films. The selection of the correct pigment for a given application must take
into account factors other than coloring ability alone. Pigments can be used which improve the
physical properties of both the liquid paint and the final paint film. They can act as corrosion inhibitors
and as mildew inhibitors. Some pigments have the ability to absorb ultraviolet radiation and so
protect the paint film as well as the coated material from degradation from this source. Pigments
may react chemically with the binder to produce desirable (and sometimes undesirable) effects.
Finally, and this has become more important in recent years, the biological effect of the pigment
material must be taken into account.
The feature which distinguishes a pigment from other coloring agents is its insolubility in
the medium in which it is employed. A pigment may be defined as a solid material, present in
the form of small particles, which is essentially insoluble in the medium in which it is present. The
ability to color or opacity the medium is not considered essential to the definition. As will be seen
later, a class of pigments called extenders is used which impart little or no color or opacity even
when used in large amounts.
A coloring agent which is soluble in its medium is called a dye. Some of the organic pigments
such as wood stains could more properly be classed as dyes. Two other types of coloring agents
are "lakes" and "toners". A lake is a water insoluble pigment formed from coloring matter in the
presence of a substrate. If the formation takes place without a substrate, the coloring agent is
called a toner. Lakes and toners are usually organic.
A. Classification and Statistics3
Pigments represent one of the most important raw materials consumed by the paint and
361
-------
coating industry. On a weight basis, consumption of pigments exceeded that of any other single
class of raw materials. In 1970, the coatings industry used 1,727 thousand tons of pigments of
all types. Consumption of solvents was a close second at 1,695 thousand tons.
In addition to their use in paints, pigments find extensive use in other industries such as
rubber, paper, plastics and printing. In fact, consumption by the paint industry represents only
about half the total pigment production in this country. Total pigment production in 1968 stood
at 3,640.6 thousand tons. Production by major categories for 1968 is presented in Table 103.
In addition to those listed, an estimated 52,500 tons of metallic pigments, mostly zinc and aluminum,
were consumed by the paint industry in 1970.
The black organic pigments consist almost entirely of carbon blacks produced from petroleum
feed stocks. Inorganic blacks are primarily black iron oxide.
The white pigments fall into two classes:2 hiding (opaque) and extender (non-opaque)
pigments. The extenders are not pigments in the usual sense of the word. That is, they do not
add color or opacity to the paint film. Their lack of these properties is due to their relatively low
index of refraction. Historically, the extenders were introduced as a cheap substitute for the more
expensive true pigment. Extenders, today, are recognized to contribute to stability, texture, durability
and a variety of other paint properties. Production of the important white hiding and extender
pigments is given in Table 104.
The organic color pigments, almost entirely synthetic, include a wide variety of types.2-3
The one feature most of the organic dyes and pigments have in common is the presence of ring
structures in their molecules. The most common starting materials are benzene, toluene, xylenes,
naphthalene and anthracene. Reactions such as nitration, halogenation, and sulfonation are used
to transform these materials into the intermediates from which the final dye or pigment is produced.
Generally speaking, a dye is a coloring material which is soluble in the medium in which it is
employed while a pigment consists of discrete particles which are insoluble in the medium. It is
not uncommon for an organic coloring agent to be soluble in one medium and insoluble in another.
Thus, it may act as either a dye or a pigment.
No single organic pigment, or class of pigments, accounts for more than a relatively small
fraction of the total. Production of the major types is shown in Table 105.
Within any given type, there exists a large variety of pigments. For instance, among the
azo group there are at least 22 different pigment materials having little in common except the
presence of a (— N = N —) bridge between aromatic groups. Most colors, except white, are
362
-------
TABLE 103
PIGMENT PRODUCTION BY MAJOR TYPE3
1968 Production
Pigment (1,000 ton/yr)
Black Organic 66.5
Black Inorganic 3.6
White Hiding 682.3
White Extender 2,670.1
Colored Organic 27.1
Colored Inorganic 191.0
Total 3,640.6
TABLE 104
MAJOR WHITE PIGMENTS3
1968 Production
Pigment (1,000 ton/yr)
Hiding
Titanium Dioxide 623.7
Zinc Oxide (lead free) 36.4
White Lead 11.1
Leaded Zinc Oxide 11.1
Extender
Kaolin (China clay) 2,153.0
CaCOs, precipitated 226.0
Talc 205.2
Bartyes 60.9
Mica 25.0
363
-------
TABLE 105
PRODUCTION OF ORGANIC PIGMENTS3
1968 Production
Organic Pigment (1,000 ton)
Insoluble Azo 7.0
Soluble Azo 6.8
Phthalocyanine 6.0
Condensation Acid 2.5
Basic 1.7
Miscellaneous 2.6
TABLE 106
PRODUCTION OF MAJOR INORGANIC COLOR PIGMENTS3
1968 Production
Pigment (1,000 ton/yr)
Synthetic Iron Oxide 72.4
Natural Iron Oxide 57.6
Chromate 44.2
Ferrocyanide 6.0
Sulfide 3.0
Mixed Chromate and
Ferrocyanide 2.8
364
-------
available with the organics. Their principal disadvantages are high cost and inferior light, heat
and chemical stability.
The inorganic color pigments include several important types. Production of the major types
is given in Table 106. As a group, the inorganics are less expensive and possess higher stability
than the organics. They are considered more suitable for exteriors and for environments where
high temperatures or corrosive conditions are encountered. They do not possess the range of
colors and tones of the organic pigments, but have representatives in most of the major color groups.
B. Purpose and Scope
The purpose of this study is to identify, and study further, those pigment groups which
represent major contributors to air pollution. This decision is based on several factors. First, it is
impossible from a time standpoint to study in detail all of the pigments used by the coating industry.
It becomes necessary, then, to concentrate effort on those areas where the problem is most acute.
This is reasonable on technical grounds, also; since many of the production processes used are
by their nature relatively free of air emissions. Also, some pigments are produced in such small
amounts that diversion of effort to these would be unjustified.
An initial screening was done to eliminate certain groups from consideration. The production
of the carbon black pigments is included in a separate EPA study and so will not be studied here.
The production of white lead and leaded zinc oxide will not be considered, in spite of lead's potential
toxicity, since present and future regulations on lead content in paint should virtually eliminate
these pigments for the coatings industry. Recent trends, as well as industry projections, support
this conclusion. Paint industry projections predict the use of white lead to decrease to 20% of its
1970 value by 1975.31
The white extenders have been eliminated on several grounds. They are produced primarily
by mining operations followed by grinding, classification, drying, etc. Any problems which might
exist would be of a paniculate nature. Technology for particulate control is well developed and
the extender industry does not present any unique problems. Furthermore, the extenders are
considered non-toxic with one possible exception.
Certain types of talc contain asbestos fibers known as "tremolyte". Asbestos fibers including
tremolyte are subject to very strict airborne concentration limitations. Not all varieties of talc contain
tremolyte and those that do contain varying amounts. New York State talc may contain 10% to
30% tremolyte, Montana talc less than 1%, and Georgia steatite talc (soapstone) 10%. The extent
of the toxicity problem associated with tremolyte has not yet been agreed upon. In any event,
365
-------
it is not known to what extent tremolyte is present in the talc used by the paint industry.
The organic pigments have been eliminated from consideration also. No single organic
pigment is produced in sufficient quantity to be a significant factor in the total emissions picture.
Furthermore, the processes used are largely liquid phase reactions under fairly mild conditions.
Air pollution problems on either a total emission basis or on a per unit product basis should be
quite low. Many of the compounds involved in manufacture are considered hazardous to humans,
however, so that there may exist occupational safety problems but such problems are not within
the scope of this study.
The pigments being studied in some detail include, among the whites, titanium dioxide
and zinc oxide. Among the inorganic color pigments the iron oxides, the chrome pigments and the
cadmiums are included. The cadmiums are present by virtue of their potential toxicity rather than
production volume. The following sections will detail the work that has been completed on these
pigments to date.
II. REVIEW OF MAJOR PIGMENTS
A. Cadmium Pigments32>33
A visit was made to the offices of Harshaw Chemical Company on June 29, 1972 to discuss
the manufacturing of cadmium pigments. Harshaw is one of the t'wo major producers of cadmium
pigments and has two plants which are located in Elyria, Ohio and Louisville, Kentucky. They
produce 20 cadmium lithopone yellows, 20 cadmium lithopone reds, 32 full strength cadmium
yellows and 64 full strength cadmium reds. They also market but do not produce 12 mercury
cadmium lithopone reds and 12 full strength mercury cadmium reds. The colors produced range
from primrose yellow to dark maroon. The pure cadmium sulfide is a deep orange color. The
color variations are achieved by the addition of controlled amounts of zinc sulfate for the yellows
and selenium compounds for the reds as shown below:
Zinc Sulfide Addition < CdS * Selenium Addition
Primrose <- Lemon <- Golden « Orange » Lt Reds —» Dark Reds -» Maroon
The cadmium lithopone pigments contain barium sulfate as a diluent and offer better tinting
strength and hiding power on a equal cost basis with the high strength pigments. They are used
primarily where higher pigment loadings can be tolerated.
The full strength pigments have approximately twice the tinctorial strength of the lithopone
366
-------
types and are used primarily in the manufacturing of plastic color concentrate where low pigment
concentration and maximum hiding in thin films are desired.
The mercury cadmium pigments are similar in color to the cadmium sulfoselenide oranges
and reds but are mixtures of mercury sulfide with cadmium sulfide. They are slightly lower in cost
but resistance to heat, moisture and exterior weathering is not quite as good as the cadmium
sulfoselenides.
Use of cadmium pigments in trade sale products is limited almost exclusively to artist
paints. Because of its high cost, it is used primarily for high quality industrial finishes where its
heat resistance and good weather ability are required. Typical applications would be for high
temperature bake enamel finishes and commercial airline and automotive finishes.
A simplified block flow diagram for the production of cadmium sulfide as it pertains to
potential emissions is given as Figure 74.
The cadmium pigments are prepared by reacting an aqueous solution of cadmium sulfate,
CdSO4, (or cadmium chloride) with a solution of sodium sulfide, Na2S, (or HaS) in a stirred reactor.
Zinc sulfate or selenium may be added in the reactor or calciner to obtain the various shades.
The precipitate is filtered, washed, calcined, wet ground in a ball mill, dried in a shelf type steam
heated oven and dry ground in a hammer mill. Cadmium lithopones are produced in the same
manner except the starting materials are cadmium sulfate and barium sulfide and the precipitate
is a cadmium sulfide — barium sulfate mixture.
The cadmium sulfate solution is produced by dissolving cadmium bar stock or sponge in
sulfuric acid containing some nitric acid in a percolating tower at about 30°C. Flower of sulfur is
sometimes added to the calciner and results in an SOa emission at the gas exit end.
This manufacturing process produces the following emissions:32'33
Percolating Tower CdO — About 0.2% of charge
NOz — Amount unknown
SOs — Amount unknown
Calciner CdS — About 1.2% of cadmium charge
Selenium — Amount unknown
ZnS — Amount unknown
SOa — Amount unknown
Calciner Scrubber Liquid water-to-sewage pretreatment. Uncollected
emission — amount unknown but considered insignifi-
cant
367
-------
1L
0
f
i
<
u
C
H3J.-UJ
"«
HSV/VS
FILTER
3TIRRED
REACTOR
i I
1
u
0
y
0
K3
u
3
T>
O
E
•o
o
o
k.
o
"•6
368
-------
Oven Drier Some entrained pigment but not a significant
emission
Dry Hammer Mill Paniculate pigment emission but collected in a fabric
and Product Blender filter for recycle raw material. Cannot be used for
product due to color contamination. Raw material
costs high enough to make recovery economically
desirable.
In as much as the contribution of environmental pollution from manufacturing of cadmium
pigment is miniscule compared to a great number of other pigments, the only justification for
studying it in place of these other pigments would be its potential as a health hazardous chemical.
This problem was discussed with Harshaw. They are quite concerned, as one might expect, that
cadmium could be, in their opinion, erroneously labeled a health hazardous material. They went
to considerable length to point out that practically no cadmium pigments are used in trade sales
coatings, that cadmium pigments are manufactured as insoluble cadmium sulfide (0.000001 gram/
100 ml) and are non-toxic due to their insolubility and that emissions in the manufacturing process
are cadmium oxide and sulfide. Cadmium oxide is also considered an insoluble compound but
no data was given.
Harshaw is very interested in cooperating with the EPA in any way. They would be glad
to let the EPA source test their plants and are quite interested to learn about any information the
EPA may have on toxicity of cadmium, especially chronic poisoning. They can furnish a bibli-
ography on all work published to date on cadmium toxicity. They also mentioned that the NPCA
is sponsoring a study for the Fall of 1972 on the toxicity of 23 heavy metals at the Kettering Institute.
B. Zinc Oxide2.3-34-35
Zinc oxide, either leaded or unleaded, is one of the most important white color pigments
used by the coating industry. While its optical properties are inferior to those of titanium dioxide,
it is able to maintain its position in the paint industry through its ability to impart a variety of
desirable properties to paint films. Also, it is an effective fungistat and mildewcide.
Use of zinc oxide by the paint industry had been declining somewhat in recent years since
it was not compatible with the latex type paints then being formulated. Its situation is beginning
to improve again and paint industry projections predict an increase in zinc oxide use of about
4% per year over the next several years.31 Zinc oxide production in 1969 was about 220,000
369
-------
tons. The U.S. paint industry accounts for an estimated 35,000 tons of this production. Other
major uses include the rubber, pottery and glass industries.
Zinc oxide itself is considered non-toxic. Leaded oxides have a toxicity in proportion to
the amount of lead present. The use of leaded zinc oxide is decreasing rapidly and is expected
to continue to do so.
There are three principle methods used in the manufacture of zinc oxide: the French,
modified American and the Electrothermal.
In the French process, metallic zinc is vaporized in a retort under a reducing atmosphere.
The zinc vapor is then burned in a combustion chamber to give zinc oxide particles which are
blown through cooling pipes and then collected in baghouses and packaged. In this process, the
furnace gases do not come into contact with the zinc oxide production system. A flow sheet is
presented in Figure 75.
The American process starts with zinc ore. The two types used are franklinite (ZnO) and
sphalerite (ZnS). Franklinite has the advantages of being almost free from lead impurities as well
as not requiring a roasting operation to remove sulfur. Sphalerite is more available, however, and
most of the oxide produced by the American process uses this as its starting material. Some
franklinite is still used, however. In addition to lead, other common impurities include iron, cadmium
and manganese. Cadmium and lead are usually produced as by-products in zinc processing
operations.
A typical flow sheet for the American process is shown in Figure 76. A ZnS ore concentrate
is roasted in air to remove the sulfur and a portion of the lead and cadmium. The solid product
from the roaster is mixed with some fluxes and coke and fed to a sintering machine where it is
sintered at temperatures as high as 1600°C but usually 1100 to 1200°C. Most of the remaining
lead and cadmium are released in the form of oxide particles which are collected with electro-
static precipitation or in a fabric filter.
The product from the sintering machine is crushed, mixed with coke and fed to a furnace
where the zinc oxide is reduced and the vaporized zinc subsequently re-oxidized. The ZnO is
collected, calcined again to remove residual impurities and/or control particle size and packaged.
The electrothermal process is similar to the American process except for the furnace. In
this case, the product from the sintering machine is mixed with coke and charged to an electrical
resistance furnace. Electrodes are inserted into the charge which serves as the resistance. The
zinc vapor and CO leaving the furnace are oxidized by air entering at the furnace exit. The product
370
-------
o
z
g
8
T
cr
o
i-
LiJ
cr
cr
u.
O>
5
o
in
(U
UJ
cr
c
N
371
-------
1
H2SO-4-
PLANIT
i
RECOVERY
1
0
c
N
I
I
U
0
<
7
/_
DC
D
L
i
o"
(/)
T)
J3
£L
TJ
U «
^
.0
Q.
n t
A^I
w ^^^
(/)
in LI
SINTERINC
MACHINE
t
DC
U
h
5
8
«-£
LLECTOR
o
u
<
!
3VICO
4
r
QC
y
0
u
(/) i
u
"1 UJ
x
/A
D
«X
U
x:
0
- 0
p
nr
u
b
(I
<
*
[
4
t
J
1
.
,
Qh
U1
U. T U
«-- I ~
< V) Q
z§
1 _ /\
en
o>
o
o
c
o
u
'k_
(U
E
O)
<
QC
Z
0
u
to
5
J3
Ll_
<£>
N
cu
372
-------
is then collected and packaged.
All of the above processes are presently in use. Most producers use both the French
and American processes with a larger share of the production from the American. St. Joseph
uses the electrothermal process exclusively. All of the producers except Eagle-Picher roast their
own ore at the plant site. In the latter case, the ore is roasted at Blackwell, Oklahoma by AMAX
and shipped to the zinc oxide plant. Where roasting is done, the SOa produced is used to manu-
facture sulfuric acid.
The characteristics of the three processes described are similar with respect to particulate
emissions. The average particle size of ZnO produced is 0.2 to 0.25 microns. Some control of
the particle size can be accomplished by a variation in the rate of oxidation, rate of cooling, air
flow rate, etc. In this way, average particle sizes of 0.1 microns or less can be produced. Larger
sizes of 0.4 microns and up can be produced by slow chilling or separate calcination of fine particles.
Product collection is achieved through the use of baghouse type filters. Collection ef-
ficiencies in excess of 99.9% (weight losses) are usually given for this type of collector. Very little
hard data is available on efficiency, however, since designers tend to be more interested in such
parameters as bag life, flow rate per square foot, etc. It must be kept in mind that extremely
fine particles are involved here and this may affect efficiency. Eagle-Picher did some tests a few
years ago and concluded that their filters operated at 99.9% efficiency. High efficiency in a bag-
house is, of course, dependent upon proper mechanical operation and on the integrity of the bag
filters.
Since almost all of the 220,000 ton/yr of zinc oxide is from only five producers, an average
plant production of 44,000 ton/yr can be assumed. Assuming 99.9% collection efficiency this represents
an emission of 44.0 ton/yr per plant, or about 0.12 ton per day per plant.
In addition to effluent through the baghouse filters, losses can occur in the handling and
packaging operations. The amounts lost in this manner are largely a function of plant housekeeping
and control. Product losses from this source should be confined to the plant site and probably do
not significantly affect ambient air quality.
For the American and electrothermal processes additional particulate losses may occur in
raw materials handling. These are estimated to be about 0.25 tons per 100 tons of zinc oxide
produced.
The gas phase effluent from the zinc sulfide roaster contains SOa, and some lead and
cadmium, probably in the form of oxide particles, as well as smaller amounts of zinc. The metals
373
-------
are recovered in a cadmium recovery plant. The sulfur dioxide is fed to a sulfuric acid plant. The
emissions characteristics of these materials, then, are those characteristic of sulfuric acid manu-
facture and cadmium recovery. Roasting of zinc ores is being examined in detail in a separate
EPA study of smelters.
Most of the remaining 862, cadmium and lead are volatilized in the sintering and calcining
steps. The vapors are fed to the acid plant and cadmium recovery units. In addition, small amounts
of chloride compounds from any fluxes used in sintering are released in these steps.
Each of the three processes includes a furnace, or burner, in which the zinc oxide is pro-
duced. In the case of the American and Electrothermal processes a reduction of the crude ZnO
also takes place and is included in the description of this operation. All of the effluent gases from
these operations eventually pass through the baghouse filters discussed previously.
In addition to particulate materials there exist other potential sources of gas phase pollutants
from the furnace. These come from the fuels and reducing agents used. For the American process,
coal is charged to the furnace in the ratio of two tons of roasted ore per ton of coal. In the French
process, vaporization of zinc metal takes place under a reducing atmosphere produced by the
incomplete combustion of about 0.7 ton of coal per ton of ZnO product. The quality of the coal
used in these operations can affect emissions with respect to the oxides of sulfur and nitrogen. The
coke used in the Electrothermal process should be relatively free of these contaminants.
There exists, also, some possibility of CO emissions. Carbon monoxide is part of the zinc
vapor stream leaving the furnace. This is burned in an excess of air which should minimize the
amount of CO leaving the burner. This is confirmed by field experience.
Process heat for the retort in the French process and, perhaps, for air preheating in all
the processes is supplied by burning either coal, gas or oil. The combustion gases are released
directly to the atmosphere and present a potential source of emissions. The extent of emissions
depends on the quality of the fuels used.
There are five major producers of zinc oxide as shown in Table 107. Two of these facilities
(Eagle-Picher and American Smelting & Refining) are devoted solely to zinc oxide production.
The other twa companies also produce zinc metal at the same locations. All of the producers
except Eagle-Picher roast their own sulfide ores. Approximately 50% of New Jersey Zinc's production
by the American process starts with franklinite and so does not require a roasting step.
374
-------
1
o
l_
Q.
o
o
CD
o"
CM
O O o
O O O
in to o
r-" in" o
Mm CM
m
<
CO
cc
LU
o
D
Q
O
CC
CL
X
o
CJ
2.
N
O
i-
D.
CO
Q-
c
CO
o _
'Z CO
c E
J te
0
c
CO
CJ
c
CO
I
L.
CO
<
C
CO
o
cB
13
0)
O
c
g
•*-•
CD
O
O
O
c
o
o
C
C
Q)
a.
8
CO
c
o
_
o
o
C
CO
c
c
0)
CO
03
CO
Ol
C
III
DC
63
O
ct
O
0-
O)
c
a
CO
CO
c
8
->. I
•"-1 ^
CJ
c
N
en
(0
i
c
CD
.C
CO
375
-------
C. Chrome Pigments2'30
The term "chrome pigment" includes a variety of compounds containing the chromate
group, CrO42ror chromium oxide. The chromates of lead and zinc are used as pigments alone or
in combinations with materials such as lead molybdate, PbMoO4. The colors available with chrome
pigments range from green to yellow to red. Consumption of chrome pigments in 1970 is estimated
at 65,000 tons.3 The most recent year for which production by types is available is 1968:3
Production
Type Formula ton/yr
Chrome green PbCrO4 + Prussian Blue 2,830
Chromium oxide green Cr2O3 6,230
Lead chromate yellow PbCrO4 32,790
Molybdate orange PbCrO4 + PbMoO4 11,380
Zinc chromate ZnCrO4 7,400
Chromate compounds are considered toxic. Lead chromes have additional toxicity potential
due to the presence of lead. In both cases they are not considered as hazardous as the white
lead pigments.
Paint industry projections predict a 12% decrease in the use of the lead chrome yellow
by 1975 and slight increases for the other chrome pigments.31 The possible introduction of more
stringent regulations with respect to lead content in paint could affect this picture considerably,
however.
1. Lead Chromes — A variety of reaction schemes, all involving precipitation, are used in the
production of lead chromates. Precipitation time, concentration of solution, pH and temperature
and other reaction parameters influence the properties of the final product. The color depends
primarily on the crystalline form of the PbCrO4. The lemon yellow rhombic form, usually the desired
product, readily converts to the stable, reddish monoclinic form. Various techniques, such as co-
precipitation with lead sulfate, have been found which largely stabilize the rhombic form. Another
stability problem is the photochemical conversion of adsorbed soluble lead salts to metallic lead
and/or the lower oxides of lead, which causes a darkening of the paint film. Methods are available
to partially, though not completely, overcome this problem.
Four commonly used reaction schemes are illustrated below:
Pb(NO3)2 + Na2CrO4 -» PbCrO4 + 2NaNO3 (1)
2Pb(NO3)2 + Na2Cr2O7 + 2NaOH -» 2PbCrO4 + 4NaNO3 + H2O (2)
376
-------
Pb(acetate)2 + Na2CrO4 ->• PbCrO4 + 2Na(acetate) (3)
PbO + 2HNO3 + Na2CrO7 -» 2PbCrO4 + 2NaNO3 + H2O (4)
Lighter shades are produced containing about 40% lead sulfate. Darker shades close to orange
are produced by alkali treatment and contain up to 40% PbO.
2. Zinc Chromes — Pigmentary zinc chrome is most commonly a basic zinc potassium chromate,
K2CrO4 • 4Zn(OH)2. Zinc tetroxychromate, ZnCrO4 • 4Zn(OH)2 is also coming into prominence.
These pigments are lemon yellow in color and have some useful properties, though they are
inferior to lead chromes in opacity and in staining power. Their main use is the manufacture of
anti-corrosive coatings. Both pigments are produced by precipitation from aquous suspensions of
zinc oxide. Two methods are used in the production of basic zinc potassium chromate:
4ZnO + 2CrO3 + K2Cr2O7 + H2O -» product (1)
4ZnO + 2K2Cr2O7 + H2SO4 -> product + K2SO4 (2)
Zinc tetroxychromate is produced from zinc oxide and chromic acid as follows:
5ZnO + CrO3 + 4H2O ->• ZnCrO4 • 4Zn(OH)2
3. Chrome Green — Lead chrome greens are produced by mechanical mixture of lead chromes
and prussian blue, KFe • Fe(CN)e • xH2O, or by precipitation of lead chromes in the presence of
Prussian blue. Manufacture of these pigments represents an extreme fire hazard, since prussian
blue is combustible and lead chrome will act as a source of oxygen. This enables chrome green
to burn independently of an external source of oxygen. Large volumes of ammonia are produced
during combustion.
4. Chromium Oxide Greens — This pigment consists of pure chromic oxide, Cr2Oa. It is prepared
by burning a mixture of sodium dichromate and sulfur. About 30 to 40% excess sulfur is used.
The reaction is indicated by the following equation:
Na2 Cr2O? + S -» Na2SO4 + Cr2O3
The excess sulfur is largely evolved as SO2. Small amounts of SO3 might also be present since
Cr2Os is known to catalyze the oxidation of SO2 somewhat.
Chromium oxide greens have exceptional chemical, light and thermal stability. Its color,
unfortunately, is considered rather unattractive and this limits its use. It is confined largely to
applications where durability is more important than appearance. It finds considerable military use
since its infrared reflectance spectrum approximates that of natural foliage.
5. Molybdate Orange — Molybdate oranges are produced by coprecipitation of lead chromate,
lead sulfate and lead molybdate to produce a mixed crystal PbCrO4 • PbSO4 • PbMoO4. This
377
-------
forms a fairly stable system ranging in color from orange to scarlet. It is often used in combination
with organic reds and maroons to produce a variety of shades.
As a general group, chrome pigment ranks second among the inorganic color pigments
in total tons produced for the paint industry. Of the five major types of chrome pigments produced,
four are made via wet chemistry and offer no significant air pollution emission. The fifth, C^Oa, is
the only pigment that emits a significant air pollutant. The production of this pigment is only 10%
of the total chrome pigment production and its use has been declining. As outlined earlier, it is
produced by reaction with elemental sulfur. Assuming a maximum of 40% excess sulfur is required
in the production, an emission of 1,050 tons/year of 862 can be calculated from the yearly pro-
duction of 6,230 tons/year of chromic oxide. Current technology for control of SOa is well developed
and requires no further discussion in this report, and it is sufficient to say that this emission can
be easily controlled.
It is also anticipated that the use of the leaded chrome pigment will also be declining in
the future by restriction of the lead content of paint. This decline has been reflected by the National
Paint & Coating Association's "Raw Materials Usage Survey" for the year 1970. They project, for
example, a drop of consumption of chrome yellow pigment of 9% for the year 1972. This is not
as severe as the 50% drop projected for lead oxide pigments over the same period; but people
in the industry are not as yet certain how the new ban on lead contents will be applied. It is fair
to assume, however, that the lead ban on paints will reduce the usage of leaded chrome pigment
significantly. This fact coupled with the earlier discussion eliminates these pigments from more
detailed study.
D. Iron Oxides2-30
Iron oxide pigments as a group represent the most widely used color pigments in the
coatings industry. The estimated consumption of these pigments in 1970 stands at 142,500 tons.3
This represents an amount exceeding the consumption of all other inorganic color pigments
combined. The popularity of iron oxide pigments can be attributed to their low cost, good physical
and chemical properties and the range of colors available. Iron oxide is non-toxic.
Iron oxide pigments may be broadly classified as either natural or synthetic. About 53%
of 1968 consumption consisted of synthetic oxide.35 Paint industry projections predict only a small
increase in the use of the natural material. An increase in the use of synthetic iron oxide of
from 25% to 42%, depending on the type, is projected between 1970 and 1975.31
378
-------
The color depends on the type of iron compound. Red pigments are essentially ferric oxide,
Fe22 production).
Yellow iron oxides, ranging in color from light lemon to deep orange, are produced by
the following steps:
1. Precipitate ferrous hydroxide, Fe(OH)2, by the addition of alkali. The precipitate forms
a suspension.
2. Blow air at a controlled temperature through the suspension to form geothite, -FeO • OH,
"seed" crystals.
3. Provide an additional source of ferrous ion and continue to blow air through the
suspension. One way this is accomplished is to introduce scrap iron into the tank. As the
ferrous ion is oxidized, hydrogen ions are formed which simultaneously react with the iron
to form additional ferrous. The reaction scheme may be represented by:
379
-------
4Fe2+ + O2 + 6H2O -» 4FeO • OH + 8H+
8H++ 4Fe^4Fe2++ 4H2
4Fe + O2 + 6H2O -» 4FeO • OH + 4H2
The newly formed oxide is deposited on the "seed" crystals causing them to grow until
the desired size is reached.
4. Wash, filter, and dry.
Paint industry projections predict a 42% increase in the consumption of synthetic red iron
oxide pigments between 1970 and 1975. Production of synthetic reds in 1968 was 34,342 tons.35
This figure does not necessarily reflect production for uses other than as pigments. Particle sizes
often encountered with red iron oxides range from 0.2 to 0.8 microns. There are three principal
methods used in manufacturing these pigments.
-, Copperas reds are produced by calcining ferrous sulfate at a temperature sufficiently high
to cause decomposition to ferric oxide, Fe2O3- The ferrous sulfate is first subjected to a mild calci-
nation to remove hydrated water. This is followed by a severe calcination in air to cause decomposition.
The reaction may be represented by the following equation:
2FeSO4 -* Fe2O3 + SO2 + SO3
This reaction has a dissociation pressure of 546 mm at 654°C. Copperas reds are considered the
most important group. They accounted for 18,910 tons production in 1968.35
Ferrite reds are produced by the dehydration of yellow ferric oxide. If calcined at 400 to
600°C, "Turkey Reds" are produced. If calcined at upwards of 900°C, "Indian Reds" result.
Precipitated reds are produced by precipitation from copperas or ferrous chloride solutions.
The ferrous hydroxide is precipitated with caustic, aerated at 60 to 90°C, and then washed, filtered
and dried. No calcining is involved. This produces a pigment free of aggregates and quite uniform
in size.
A fourth type of red is made in very small amounts. This is Venetian Red made by precipitating
copperas with lime, aeration and calcination to produce a pigment containing 40% Fe2C-3 and
60% CaSO4.
Brown iron oxide pigments can be made by blending the pure reds, yellows and blacks
or by precipitation and calcination. Blacks are produced by complete precipitation from copperas
or ferrous chloride followed by aeration at the boiling point of the solution. 6,177 tons of brown and
3,560 tons of black iron oxides were manufactured in 1 968. 35
With one exception, all of the processes used for manufacturing iron oxide pigments should
380
-------
be relatively free of atmospheric emission problems. Most of the operations are aqueous in nature
and do not produce gas phase pollutants. Since the fine particles of dry pigment eventually produced
represent product, it would be expected that paniculate emissions would be well controlled. Adequate
methods are available for highly efficient collection of fine particulate.
The one process which can produce serious problems is the manufacture of copperas
reds. Here the production of large volumes of SO2 and SCb must be contended with. The production
of 18,910 tons of copperas red causes the release of 7,600 tons of 862 and 9,500 tons of SOs.
This must be recovered as sulfuric acid or otherwise accounted for.
There are two major producers of synthetic iron oxide pigments along with a number of
smaller manufacturers. The largest producer is Cities Service, Inc. with a plant producing precipi-
tated yellows and reds at St. Louis, Missouri and another producing synthetic reds at Monmouth
Junction, New Jersey. The second major producer is Pfizer Minerals with plants at East St. Louis,
Illinois, Easton, Pennsylvania and Emeryville, California. The plants at East St. Louis and Easton
produce calcined copperas.
The smaller producers include Reichard-Coulston, Bethlehem, Pennsylvania. Their production
includes calcined copperas. They report that they have a scrubber of unspecified type on their
copperas calciner.
Mineral Pigments Corporation at the present time does not manufacture red iron oxide and
deals primarily in natural materials. Chemtron and Hilton-Davis both produce a line of transparent
iron oxides for the specialty market, primarily automotive finishes. These are made by precipitation
processes followed by mild calcining to remove water.
E. Titanium Dioxide Pigments3'36
Titanium dioxide is an opaque, white material. When used as a pigment in the paint and
lacquers, it provides excellent hiding power due to the very high refractive index of the two crystalline
forms, rutile and anatase. This excellent hiding power and lack of toxicity combine to make titanium
dioxide the leading hiding pigment in the paint industry, measured by either tonnage of material
consumed or dollar value.
Two processes are used for the production of titanium dioxide pigments. These are the
sulfate process and the chloride process. These processes use several titanium-bearing raw materials.
The principal sources of these are:
1. llmenite, or iron titanate, which is mined in the U.S., as well as several foreign countries.
381
-------
2. Slag from the electric-furnace smelting of ilmenite which yields a salable iron product
and a high — TiOa content slag.
3. Rutile, which occurs naturally in Africa, Australia, and several other countries outside
the U.S.
4. Leucoxene, a highly-weathered ilmenite possibly supplemented with minor amounts of
ilmenite containing 70 to 80% TiC>2.
The sulfate process is the oldest and most versatile of the two processes. It may use
ilmenite or slag, and is capable of producing either rutile or anatase. The chloride process produces
equivalent products. The details of chloride process technology are often considered proprietary
by the industry and it is difficult to define the state of the art. Historically, the chloride process
has produced rutile pigment only and has required rutile ore as raw material. Currently, however,
techniques have been developed which permit use of ilmenite ore (possibly upgraded). Also, it is
now possible to produce anatase TiOa by the chloride process.
This discussion is concerned principally with the manufacturing process for titanium dioxide
as they relate to atmospheric emissions. Much of the information available relating to the economics
of the application of the two principal pigment types and the many variables with regard to particle
size, surface chemistry, etc., will be omitted from the discussion if they do not have a particular
bearing on the emission-related characteristics of the process. On the other hand, those details
of the manufacturing process and properties of the materials which do relate to variations in atmos-
pheric emissions from one plant to another or within a given plant, will be included.
Of particular interest in the consideration of the process in the following paragraphs is the
fact that there is a limited number of producers of titanium dioxide in the U.S. Each of these
practices the processes described in ways that incorporate some characteristics peculiar to that
manufacturer. This discussion will not identify all of the processing variations and relate them to
specific manufacturers. However, those which are important to the atmospheric emission aspects
of the process have been characterized insofar as possible, with respect to the particular manu-
facturer. Those companies known to be engaged in the production of titanium dioxide within the
U.S. at this time are tabulated on the following page.
382
-------
Sulfate Chloride
Manufacturing Company Processing Processing
American Cyanamid Yes Yes
Kerr-McGee No Yes
Cabot Corporation No Yes*
DuPont Yes** Yes
New Jersey Zinc Yes Yes*
NL Industries Yes No
SCM Glidden Yes Yes
Sherwin-Williams No Yes
1 . Sulfate Process — The most common raw material for the sulfate process is ilmenite. The
principal constituent of ilmenite is iron titanate, FeO • TiO2. This is found associated with various
minerals, such as vanadia, alumina, etc. Table 108 gives the chemical composition of several of
the ores with each of the elements reported as its oxide form. In addition to the common ilmenite,
found in the U.S., in the Adirondacks and in Florida, and in Australia and Canada, a co-product
slag produced by the smelting of Canadian deposits is a frequent raw material for titania manu-
facture in the U.S.
In the sulfate process, the iron titanate is dissolved in concentrated (66° Baume) sulfuric
acid to yield titanyl sulfate, ferric sulfate, ferrous sulfate and other soluble mineral sulfates and
gangue residue. The principal reaction, greatly simplified, for this digestion is:
FeO • TiO2 + 2H2SO4 -» TiOSO4 + FeSO4 + 2H2O
This reaction is carried out batch-wise in multiple digesters which feed a semi-continuous
process. The solid sulfate cake is dissolved in water and recirculated acid to yield "sulfate solution".
The immediate objective is to put as much of the titanium into solution as possible. At this point,
it is objectionable to have metals of high valence present in the solution and the solution is treated
with a reducing agent such as metallic iron to bring about reduction of ferric ions and other metal
impurities to a low valence state. The typical reduction reaction taking place in the digester is:
Fe° + 2Fe+++^ 3Fe++
After digestion, all of the insoluble metallic components are filtered and removed from the solution.
All of the iron remains in solution at this point.
*N.J. Zinc has recently been reported to have leased the chloride production facilities of the Cabot
Corporation.
**Will be discontinued by the end of 1974.
383
-------
TABLE 108
ANALYSIS OF ILMENITE ORES
United States
Canada
Virginia
Chemical
Constituent
TiO2
FeO
Fe2O3
Si02
AI2O3
P205
ZrO2
MgO
MnO
CaO
V205
Cr2O3
Piney
River
44.3
35.9
13.8
2.0
1.21
1.01
0.55
0.07
0.52
0.15
0.16
0.27
Roseland
51.4
37.9
1.6
4.6
0.55
0.17
• 2.35
0.70-
0.59
0.07
New York
44.4
36.7
4.4
3.2
0.19
0.07
0.006
0.80
0.35
1.0
0.24
0.001
Florida
64.1
4.7
25.6
0.3
1.5
0.21
0.35
1.35
0.13
0.13
0.1
California
48.2
39.1
10.4
1.4
0.2
0.05
0.6
0.1
0.1
0.05
0.03
Ivry
42.5
39.1
20.7
0.88
1.05
2.0
0.04
0.1
0.36
0.15
Bourget
22.4
36.9
31.2
1.0
6.01
0.93
1.50
0.55
Allard
37.3
26.3
30.0
0.004
0.10
0.39
384
-------
When ilmenite is used, the next processing step consists of cooling the solution to crystal-
lize out some of the iron as FeSO4—7H2O. The next processing step involves the dilution and
heating of the sulfate liquor, which hydrolyzes the titanyl sulfate and brings about precipitation
of titania. The basic chemical equation for this hydrolysis step is:
TiOSO4 + H2O -> TiO2 + H2SO4
During subsequent processing, the titania is present as a precipitate, and most of the
impurities remain in solution. Several filtering and washing steps follow, in each of which the titania
is retained on the filter, and the filtrate and wash water are recycled or discarded. Although the
basic TiO2 material is formed during the hydrolysis step, the product of this part of the reaction
is a fine, non-pigmentary material.
Subsequent processing steps are aimed at modifying the particle size and crystallinity,
conditioning the surface of the particles and performing other proprietary treatment steps to produce
a final pigment material with the desired properties. The next step in the preparation of the finished
pigment is calcination.
In a rotary calciner, the TiO2 "hydrate" cake is dried and calcined to achieve the final
crystalline product. Here, no basic chemical change takes place, but the recrystallization of the
titania to form either anatase or rutile crystals of the desired size, is accomplished.
Following calcination, the crystallites are milled and packaged for shipment. Alternatively,
the particles may be treated with various surface-coating agents to achieve particular properties of
dispersibility, resistance to weathering, etc.
Figure 77 contains a detailed process flow diagram for the manufacture of either rutile or
anatase from ilmenite by the sulfate process. Some of the processing steps in this process are
carried out batch-wise, while others are continuous. Similarly, some of the emission sources produce
variable rates and compositions of air pollutants while others are relatively steady and continuous.
The ilmenite ore* is a black, powdery material of relatively coarse particle size, which
produces some dusting as it is transferred from rail car, ship or truck to an ore holding yard
to await processing. Again, some dusting may occur as the ore is loaded onto belt or bucket
conveyors for transfer to the ilmenite dryer, which forms the first step in the manufacturing process.
The ilmenite is dried in a counter-current gas or oil-fired dryer. The dryer is generally
equipped with a mechanical dust collector, fabric collector or electrostatic precipitator at the gas
discharge end. Dust captured by the collector is returned directly to the process. Because of the
*lt will be understood that in the following discussion, the term "ilmenite" is meant to include, also,
high TiO2 slag.
385
-------
-------
00
LJ
0
J
0.
9-
5
387
-------
relatively coarse size of the raw ilmenite, there is little net loss of dust from the drier. Sulfur
dioxide may be present as a gaseous emission if fuel oil is used rather than natural gas. The
drier ordinarily runs continuously and feeds dry ilmenite ore to a ball or pebble mill for grinding.
The rotary mill operates without any gas source, and is relatively free of a dusting problem.
Following the mill, the ground ore passes through a classifier which separates oversized particles
for return to the mill and passes particles of sufficient fineness to one of the batch digesters.
Here again, there is no gas emission from the classifier and therefore, little likelihood of emission
of dust or fumes, although electrostatic precipitators may be used on some milling systems.
The digesters are charged sequentially with dry, milled ilmenite ore. During charging, there
may be some dust emission through the ventilating stack on each of the digesters. Charging is
complete in a matter of a few minutes, and the dust emission ceases when the ore has become
wetted with the sulfuric acid added to the digester.
The initial digestion reaction starts relatively slowly, with the reaction mass at a temperature
slightly above ambient temperature. Steam is added to increase the temperature and the reaction
rate, and air sparging for agitation may be used. As the exothermic reaction proceeds, the temper-
ature of the reaction mass increases and the rate of reaction increases. After approximately 5 to
30 minutes, the reaction rate reaches a peak for about 10 to 15 minutes and the maximum rate
of emission of steam, sulfuric acid mist, particulate matter and sulfur dioxide through the stack
occurs. Subsequently, the reaction slows down, but steam emission continues until the reaction
mass has "set" to form a solid.
After the digestion reaction is complete, water and possibly some recirculated acid, is
added to the digester to dissolve the solid. At this point in the cycle, scrap iron in a basket-like
container is immersed in the solution and allowed to remain in contact with the solution until the
reduction of ferric ions to ferrous and some slight reduction of titanyl sulfate has taken place. Other
reduction techniques may be used in some plants. When the desired degree of reduction is achieved,
the sulfate solution is transferred to a clarifier.
The clarifier is basically a large settling tank in which a rotating rake processes settled
material toward the center for removal as a solid "mud". Although the digesters discharge inter-
mittently, the clarification tanks operate on a continuous basis. In the clarification process, pro-
prietary floculating and coagulating agents may be used to improve sedimentation of the solid
impurities. In some cases, hydrogen sulfide or other chemical reagents may be added to aid in
the floculation and removal of solids. It is customary to use completely closed clarifier vessels,
388
-------
even if hydrogen sulfide or other gaseous reagents are not used. Typically, the clarifiers operate
at 60 to 80°C and some evolution of water vapor is normal. The mud collected at the bottom of
the clarifier is washed with recycled sulfuric acid to remove any dissolved titania and the acid
from this washing is returned to the digester. The mud is disposed of by landfill, by sluicing or
may be used as a source of minerals in a recovery process.
The clarified sulfate solution may be passed through a filter process to remove the last
traces of solid material and then passed into an iron sulfate cyrstallizer. This is simply a closed
vessel which is evacuated through a steam jet ejector to a barometric condenser. The pressure
reduction brings about evaporation of water from the sulfate solution which cools the solution and
concentrates it with respect to both titanium sulfate and iron sulfate. The objective of this step
is to crystallize some of the iron sulfate for removal from the process.* The crystallization takes
place in multiple-batch vessels which are mechanically agitated as the cooling, evaporation and
crystallization take place. After sufficient water has been removed, the slurry of iron sulfate-titanium
sulfate solution is passed through a drum filter where the ferrous sulfate solid is removed. This
solid product consists almost completely of FeSO4 • 7H2O, and is normally called "copperas".
Copperas is not produced when slag is the principal raw material.
Copperas is the principal raw material used in the manufacture of iron oxide pigments, and
certain other iron containing pigments. It is fairly common for a plant preparing iron pigments to be
located nearby and to use the wet copperas filter cake, as it comes off the filter press, for raw
material. In some cases, the use of the copperas is restricted to plants located at some distance
from the titanium oxide plant. When this is the case, the filter cake is conveyed to a rotary drier
and copperas is dried for bulk shipment to distant points. Where a drier is used, it is necessary
to provide a mechanical dust collector and possibly a fabric collector on the flue gas discharged
from the drier.
The filtrate from the copperas filter is conducted to evaporation and concentration vessels
in which further concentration of the titanium sulfate solution takes place batch-wise. This vessel
is provided with steam heat coils and operates under vacuum provided by a barometric condenser.
This concentration step increases the titanium concentration to the equivalent of about 200 grams
(measured as TiCk) per liter of solution.
Further heating is accomplished in a separate stirring-heating vessel using steam coils as
the heating source. After this heating step, the strong solution is diluted and additional heating
•Normally, crystallization is not done when processing slag solutions.
389
-------
brings about hydrolysis of the titanyl sulfate to hydrated titanium dioxide. It is in this step that the
predominant color of the titanium compounds changes. In all prior steps, the predominant color
of the solution and slurry materials is the black of the ilmenite and the sulfate solutions. In the
hydrolysis step, the white color associated with titanium dioxide pigment develops. From this point
forward in the processing, the processing equipment is referred to as the "white end" of the plant.
After boiling from 3 to 6 hours in the hydrolysis reactor, the slurry is transferred to filters
in which the precipitated titania is separated as a solid product and the filtrate is circulated to a
closed spent-acid clarifier. The overflow from this clarifier is partially recycled to the digestion
process and to the mud washing step, and the remainder to disposal. The underflow contains some
titania which passed through the filters. This sludge is recycled to the filters for recovery. The
solid titania removed from the leaf filters is reslurried with water and transferred to a bleaching
reactor. In this step, conditioning agents, dilute acid and materials such as aluminum or zinc are
added to bring about a slight reduction in the titania and produce the optimum brightness. The
material leaving the conditioner is filtered and dilute acid filtrate recycled through a slurry settler.
Overflow from the dilute acid settler is recycled or discarded while the solids are recycled to the
bleaching step. The titania hydrate retained by the filter is washed on the filter and transferred
to a second conditioning step. Here, proprietary conditioning agents are added to modify the
crystalline structure and improve the properties of the final pigment. For example, antimony may
be added at this step to modify the chalking properties of the titanium pigment.
The slurry is again filtered, with washing of the filter cake on the drum, and recycle of
the dilute acid filtrate into the dilute acid settler. The washed solid material is discharged into the
calciner.
Catalysts for promotion of rutile formation may be added at the calciner inlet if rutile is
the desired end product.
Calcination comprises a critical step in the pigment processing in that it is in the calciner
that the final crystalline form is established as either anatase or rutile. Also, the gaseous discharge
from the calcination step comprises one of the largest and most difficult single emission sources
in the manufacture of titanium pigments.
The calcination step involves the removal of water of hydration from the titanium slurry,
and, subsequently, crystallation of the titanium dioxide. During the first half of the calcination process,
water alone is removed from the solids being transported through the rotary calciner. As the temper-
ature reaches 200°C or so, water evaporation is complete and sulfuric acid vapor is released from
390
-------
the solid. The temperature continues to increase as the solid passes through the calciner toward
the firing end and temperatures of 900°C or so are reached at the discharge end.
The equilibrium product from the calciner is the rutile form of TiC>2. However, the anatase
form tends to be produced as an intermediate, and severe calcination conditions, that is, high
temperature and long residence time, are required to produce recrystallization to the rutile form.
It has been found that the formation of rutile is promoted by the addition of catalytic materials.
These are used as "rutile promoters" if rutile is the preferred product.
The calciner is generally fired with natural gas or oil. In addition to the products of combustion,
sulfuric acid vapor driven from the titania is released into the gas stream, and this partially decomposes
to produce SOa and small amounts of 862 by decomposition of the sulfuric acid. In addition, there
may be some entrainment of the titanium dioxide at the feed end, although this is nominal because
of the large size of agglomerate particles and the relatively high moisture content at the feed
end of the calciner.
The treatment of the flue gas discharge from the calciner varies from plant to plant. One
of the common combinations of processing steps is shown in Figure 77. This involves water quenching
of the flue gas to reduce the temperature to well below the condensation temperature of sulfuric
acid mist (below 240°F or so) and provides an initial water wash step. Effluent from the cooler-
condenser then passes upflow through a tubular electrostatic precipitator. Lead is the preferred
material of construction for ductwork, precipitator tubes and housing. The discharge from the
electrostatic precipitator then passes through a final scrubber, constructed of redwood. Condensate
from the precipitator as well as the water from the initial quenching step is recycled to various
parts of the process.
In the final washer, large volumes of water are used on a once-through basis and achieve
substantial reductions in sulfur dioxide concentration in the gas. This water is ordinarily discarded
after use in the scrubber.
An alternative processing scheme utilizes a venturi scrubber with approximately 40 in.
water column pressure drop across the throat to accomplish the sulfuric acid mist recovery in
place of the electrostatic precipitator. This use of venturi scrubbers at this stage of the process
must be considered experimental, as of this date, since fully satisfactory operation of such scrubbers
has not yet been achieved.
The hot titanium dioxide leaving the discharge end of the calciner passes through a cooler
and then into the final finishing line. Finishing may be accomplished by a sequence of wet finishing
391
-------
operations, or, alternatively, by dry finishing.
In the wet finishing scheme, the calcine passes in sequence through a Hammermill and
rotary pebble or ball mill. Water is added in the Hammermill and all of the subsequent operations
are carried out with the titania wet. Wet classification with recycle of oversized material to the
pebble mill is followed by thickening of the slurry. The underflow from the thickener is treated in
a conditioning reactor in which surface coating agents and other conditioners may be added to
modify the properties of the titania. For example, aluminum hydroxide or silica may be added
to alter the dispersion properties of the pigment, according to whether it is intended for dispersion
in oil-based vehicles or in water.
After the final conditioning, the pigment is again filtered and washed on a rotary drum filter
from which the filtrate is recycled to the final thickener. The filter cake is dried and discharged
into a final milling step. This mill or "micronizer" deagglomerates the titania to produce the basic
particle size set originally in the crystallization and calcining steps. The micronizer operates by
subjecting the dried solid to extremely high velocity jets of steam. The stream leaving the micronizer
is condensed and the finished dry product is packaged in bags or conveyed to bulk containers
for bulk shipment.
The dry finishing alternate consists of roller milling and Hammermilling prior to packaging.
The roller milling step involves the production of a substantial quantity of entrained titania dust
which is collected in a large cyclone and a final fabric filter. The material collected from both of
these recovery stages is passed on to the Hammermill, which produces the final particle size
desired. Again, the final product may be bagged or prepared for bulk container shipment. Materials
intended for water base paint applications or for use in paper manufacture may be reslurried for
wet transport.
In a complex industrial process involving many stages, there are numerous potential sources
of atmospheric emissions. Many of the potential sources relate to housekeeping and retention of
product within the process rather than to necessary gas emissions into the atmosphere. In the
production of titanium dioxide, many opportunities present themselves for loss of product as dry
transfer operations and handling of the ilmenite ore, calcine or final product are carried out. In
this discussion, only those sources which relate to necessary emissions of gas from the processing
stages will be considered as legitimate potential air pollution emission sources.
The most significant of these legitimate sources involves the drying of the initial ore, digestion
of the ore or slag and calcination of the titania after hydrolysis. In some of these, the gaseous
392
-------
emission consists principally of the products of combustion of the drier or calciner into which
paniculate matter is entrained. In addition, sulfuric acid and SO2 are emitted by the calciner as
the result of vaporization and decomposition of sulfuric acid from the titania hydrate calciner feed.
In the case of the digester, the gaseous emission consists principally of steam produced by the
vaporization of water as the digestion reaction is carried on. Each of these sources will be discussed
in some detail in the following paragraphs. In addition, a number of secondary sources exist which
involve limited discharge of gases into the atmosphere or from which the discharge is routinely
and uniformly controlled in order to preserve the basic product. Examples of these secondary sources
are:
Ore milling
Copperas drying
Final milling of pigment product.
2. Chloride Process — The chloride process for titanium dioxide pigment production is an
alternative route to the manufacture of rutile pigments. The process is continuous to a higher
degree than is the sulfate process, yields a pigment of lower impurity content which may be important
in some applications, and, generally, has fewer sources of emissions requiring control. On the
other side of the ledger, the chloride process has historically been limited in the range of raw
materials which can be used to rutile or high - TiO2 feed stocks. It involves corrosion, heat transfer
and materials handling problems of unusual severity.
The fundamental chemical reactions involved consist of chlorination of titanium dioxide
according to:
C + TiO2 + 2CI2 ->• TiCU + CO2
and 2C + TiO2 + 2CI2 -» TiCI4 + 2CO
After chlorination, the liquid titanium tetrachloride is purified by various solids removal, chemical
treatment and distillation procedures and is then oxidized according to:
TiCU + O2 -» 2CI2 + TiO2
These simple equations do not, of course, express the complexity of, nor sophistication
required for, the manufacture of TiO2 by the chloride process.
The basic flow scheme is illustrated in Figure 78. Ore, or combinations of ore and slag,
are charged into a continuous chlorinator along with coke. The TiO2 bearing ore and the coke
are frequently shipped by rail or ocean-going vessels and unloaded for storage in open areas.
393
-------
394
-------
The moisture content of both coke and ore stored outside is likely to become too high for use
in the chlorinator, unless it is dried in gas-fired rotary driers as shown in the box in the lower
left hand corner of Figure 78. If these driers are used, the dust generated in handling is a potential
source of pollution, and the flue gases must be treated by mechanical collectors, fabric filters
or other particulate abatement device prior to discharge into the atmosphere.
In some cases, it is possible to arrange for unloading and storage of the ore in enclosed
buildings or sheds. This prevents the accumulation of excessive moisture contents in either raw
material, and obviates the need for drying.
In either case, some dusting of the dry ore and coke in the handling process involved in
removing them from storage and elevating them to the reactor level is unavoidable. It is customary
to ventilate the conveyors, elevators, etc., through a fabric collector, or other high-efficiency particu-
late control device. Figure 78 shows a pressurized fabric collector on the ventilating air drawn from
two bucket elevators.
The coke serves as a reducing agent and receptor for the oxygen liberated during chlori-
nation of the titanium ore. Gaseous chlorine is fed into the reactor and carbon dioxide, carbon
monoxide and titanium tetrachloride vapor are the principal products. Substantially all of the chlorine
and all of the coke are used up in the chlorination process. Following condensation to remove
titanium tetrachloride, gases vented from the chlorination reactor consist principally of carbon monoxide
and carbon dioxide and other impurities such as HCI, traces of sulfurous gases and possibly free
chlorine.
Where free chlorine is present at this point, it may produce a serious corrosion problem
in subsequent piping or equipment, and may contribute to pollution. For this reason, the reactor
is carefully designed to minimize the free C\2 content at the reactor exit, and methane may be
introduced at this point to eliminate or minimize the Cla content. The reaction of methane with Cb
produces HCI. Also, there is a possibility of discharging substantial amounts of Cla, oxygen, and
water during an upset in reactor operating conditions, and the use of methane to minimize oxygen
and C\2 discharge under these conditions is helpful.
This gas is potentially valuable due to the heating content of the unburned carbon monoxide.
However, practice varies as to the purification and use of the carbon monoxide values. Some chloride
process operations involve scrubbing the off-gases for chlorine, HCI, and TiCU removal and flaring
or direct discharge to the atmosphere to dispose of the unwanted gas. Others follow these steps
and then remove the carbon monoxide gas for use as an auxiliary fuel in the titanium tetrachloride
395
-------
burner.
Where flaring of the carbon monoxide-rich gas is practiced, the removal of traces of sul-
furous gases, chlorine or chlorides may be required to avoid an emission problem. Scrubbing with
water and an alkaline liquor in a countercurrent packed scrubber should provide satisfactory air
pollution control.
The titanium tetrachloride produced in the chlorinator is purified by distillation and chemical
treatment. Distillation may consist of a number of flash vaporizations, as shown in Figure 78, or
it may be carried out in rectification and stripping towers. The removal of unreacted solids is
accomplished by sedimentation and distillation. Generally, these operations are carried out in a
pressurized system without any gaseous discharge to the atmosphere. In sections of the plant
containing liquid titanium tetrachloride, particular attention must be given to the handling of relief
valve vents, potential leaks, spills, sampling, etc. Titanium tetrachloride hydrolyzes spontaneously
upon discharge into the atmosphere to form extremely dense white fumes of fine particle size,
and even a minor leak may create emissions of high opacity.
The heart of the chloride process is the titanium tetrachloride burner. In this burner, the
titanium tetrachloride is burned with air or pure oxygen, together with, in some processes, an
auxiliary fuel. The auxiliary fuel may consist of carbon monoxide, from the chlorinator, or a hydro-
carbon, or it may be omitted altogether. The use of auxiliary fuel permits an additional degree of
freedom in controlling the temperature and chemical composition of the burner feeds.
In the specially designed burner, the oxidation of titanium tetrachloride must be completed,
and crystallites of proper structure and size must be formed. In order to accomplish the chemical
reaction and proper control of the crystallite structure and size, and to avoid plugging of the burner
by TiOz deposits, the titanium tetrachloride burner design is critical. Mechanical design of the burner
to produce the desired product is one of the key factors in the success of the chloride process.
The reaction taking place in the burner produces free chlorine plus titanium dioxide as reaction
products. The combination of the high chlorine concentration in the effluent gas, coupled with the
high temperature generated in the burner (1800°C) produces a corrosive situation much more severe
than that generally encountered in chemical processes. Exotic materials of construction, inorganic
refractories, etc. are the rule rather than the exception in this part of the process.
The quenching of the hot combustion products in order to prevent undesirable TiO2 sintering,
the separation of the titania from the gaseous chlorine and the ultimate removal of heat from the
process are among the most formidable and challenging problems encountered in a chemical
396
-------
process. In one process variation (NL Industries patent No. 3,560,152) the combustion products
are quenched with cold recycled chlorine. The gas stream then passes to bag filters in which the
TiO2 is removed by means of Inconel fabric filter elements. Process heat removal is then accomplished
by direct contact scrubbing of the hot chlorine by concentrated sulfuric acid, and the heated acid
is indirectly cooled by cascade coolers. The chlorine is then recycled to the quenching step, and
to the chlorination reactors. The titanium dioxide is conveyed from the bag filters, slurried in water,
and further processed by wet-finishing as previously described.
Practice varies widely with regard to the method of quenching the hot burner effluent and
separating the TiO2 pigment from the hot gas. However, regardless of the approach used, there
should be no interconnection of process streams with ambient air, and no potential for air pollution
during normal operation.
The final pigment product is milled and packaged as in the sulfate process. These steps
involve the routine recovery of pigment for product conservation, and are not significant sources
of air pollution. Here again, practice in the finishing steps vary widely from one plant to another.
A manufacturer may be required to process TiC>2 for a wide variety of special applications, and,
therefore, need many trains of finishing equipment, including driers, reslurrying, filtering, washing
and treating tanks. On the other hand, it may be possible to produce only a few products, all
of which are treated in a single train.
The total amount of ventilation air required to avoid unsatisfactory working conditions in
the treating area may vary from as little as 1500 SCFM to many thousands, depending on the
nature and complexity of the treating equipment. However, the potential for air contaminant emissions
from this part of the process is quite limited in either case.
In summary, the principal source of atmospheric emission in the chloride process is the
chlorinator vent gases, consisting mainly of carbon monoxide and carbon dioxide with traces of
sulfurous gases, and possibly hydrochloric acid, chlorine and titanium tetrachloride. Secondary
sources are equipment and tank vent systems, but these are usually easily controlled by caustic
scrubbers. Finally, special attention is required to avoid accidental spills of titanium tetrachloride,
and leakage of gaseous chlorine.
3. Industry Statistics — Questionnaires —
a. Products and Raw Materials — Table 109 presents data obtained from the TiOa industry question-
naires on products and production. Most of the data covers the year ending 12-31-72. Total TiO2
produced by the six plants which did not consider such data confidential was 140,040 tons by the
397
-------
o o
8 8
o
Tfr
Lf>
in
o
o
o
o
CD
CN
CO
CO
CM
o
o
in
CM
C-
O)
CO
o
o
CM
ID"
o
o
CD
CO
O
O
o
r-
CD
o
o
CD
0>
o
T—
LU
*n
^l.
2
^
D
CO
z
o
i-
"ABULA
i—
ill
tr
_>.
"s
o
in
«
J>
CB
5
1
cc
o
CD
.M
o
O
O
o^
CM"
CO
_cu
'•p
D
CC
in
£.
Z
O
S>
LU
3
O
a:
i-
O)
ID
CO
01
+-<
'c
CD
JE
—
o
o
o
co"
CO
CO "~
=)
Q
Z
O
o
o
O
o
o
o
o
CO
o
o
O
CD
o
o
CD
"
c o
CD in
-o
o
CM
8 8
CO O
T-~ •*"
CM O
oo
O)"
CD
o
o
o
o
CD
00
(T>
CM
CM
1*-
a>
CO
5-
CD
CD
Q.
CD
i_
m
nj
Q.
cb
r>-
cn
h_
^>
c
0
c
o
'^
o
IJ
o
o
CL
CD
•JO
o
6
V
^B
tr
00 g °
0°, R
00 °.
rC in |§
0 CM *
2
CD
M—
CO
.
"c
O
8
in
CM"
CM
CO
a.
!c
CO
s
<
73
CD
73
CO
0)
CL
3
CD
2
X
O
(N
O
i-
s?
O
CD
^
73
CD
CQ
398
-------
sulfate process and 249,000 tons by the chloride process. Two of the plants produced pigment by
both techniques.
A summary of production and raw materials for those five plants which reported both is
given in Table 110. Examination of this table reveals some interesting features. First, even though
the chloride process has by far the majority of the production, ilmenite is the dominant titanium
raw material with rutile maintaining a rather minor position. Second, sulfuric acid consumption in
the sulfate process amounts to about 3 tons of H2SO4 per ton of finished TiO2 shipped. A large
part of this is potentially part of an air or water pollution problem, though a significant amount
can be accounted for in the copperas by-product.
Finally, chlorine consumption in the chloride process is considerably higher than might be
expected in light of the fact that chlorine recycle is employed. Making an adjustment for the TiCU
product shipped, chlorine consumption amounts to somewhat more than 1/2 ton per ton of finished
TiO2. It cannot be concluded that all of this chlorine represents potential atmospheric emissions.
Instead, it may be related to the use of ilmenite ore. The manner in which ilmenite is used in
the chloride process is a well kept secret. However, unless some other means are available to
separate out the iron and other impurities, it is logical to assume that they are disposed of as
chlorides. This could account for a large part of the chlorine consumption.
b. Process Equipment — Table 111 summarizes the major pieces of sulfate process equipment as
reported in the questionnaires. The information for chloride process equipment is listed in Table
112. The degree of confidentiality required suggests that these manufacturers are somewhat hyper-
sensitive. The data which was reported in the questionnaires reveals no startling or unusual aspects.
A more detailed description of the mills in use is presented in Table 113. Only one plant reports
sulfuric acid manufacturing facilities while one other has a dilute acid concentrator.
III. EMISSIONS
The emissions characteristics of the various pigment processes have been briefly discussed
in the process narratives reviewed earlier. The discussion in this section will be limited to a more
detailed examination of TiO2 manufacture. This is the pigment manufacturing process whose emission
potential justifies a detailed examination.
A. Description of Emission
Each of the manufacturing processes for TiO2, sulfate and chloride, has associated with it
a particular set of atmospheric pollutants. This is in addition to particulate emissions which are
399
-------
TABLE 110
TiO2 INDUSTRY QUESTIONNAIRE
PRODUCTION - RAW MATERIALS INVENTORY FROM
FIVE TiO2 PLANTS
lon/yr
Production
Sulfate TiO2 90,040
Chloride TiO2 224,000
TiCI4 22,500
Raw Materials
llmenite 258,600
Slag 77,200
Rutile 69,800
Leocoxene 0
Coke 72,000
H2SO4 275,600
CI2 141,500
Other Ore 59,000
400
-------
UJ I-
DC Z
T- CO
00
<
cc
CO
Q
CO
CO
LU
O
O
CC
Q.
LU
Q 3
\- CO
^
.,_,
c
CO
51
5-
^-J
'o . — .
CO >
Q. CO
CO ~O
01 0
<^
CM CN
o q
\— h-
n n
T T
£t
E
-1
Z
LO
4_j
C
CO
CL
-t-i
'o -^
2 >
Q. ru
co ~o
en O
> £.
<£
n 1 co
LL
Z
cu
E
13
Z
CO
•4— '
c
CO
a.
'o —
CO >
CL co
CO ~&
O ~^"
o> O
> H
<^
O
0
"d- 'd- o o
CO ^ CN >*
CN «- v- CM
03
E
—1
Z
•4--
C
CD
CL
+-•
"o
CO ' — '
Q. >
CO
o -5,
II
r- CM CO
CO CO CD
C C C
03 CO Q)
2 2 S
4— M— 4—
C C C
888
cB
"E
Z
«- ro ui
CN
O
LU
o o co
r-^ CD r^
CM co
CO
'^
c
C
0)
c c
o o
O O
a>
6
CD
6
0)
01
a>
a.
a.
o
O
0)
c
c
0)
E
O)
Q.
401
-------
LU
<
LU ^
< 2
i I
P o
o ^
I 2
CO
Q Q
?. CC
N O
o
>
'o . — .
CO >«
Q. co
of
iv. en O
c
'o —
(0 >
Q. CO
CO ~n
CO O -£
o3 0) O
m <; —
«- 1 1 CM 1
LU
Q
= ^ "•
co i_ Z
E _g
E
z
'o —
CO >
D. CO
of
•* 01 O
+j > it
c <
O
u
o in in o
CM if CO I-.
CM CM
to
o.
CO
1
Z
'o —
CO >
Q. CO
CM 01 O
<-• 3 it
c <
O «- CO CD 1
O O O
o o m
Pv ^f <-
CO
Q_
1
E
D
>- O O CO 1
V)
E
CO
>
u, M «
w fe 0) 0
CO 'l- .— CO
S Q » -1 g
« 5 £ ° E
"00^3
O O O O no
1 1 1
1
^
\- ~
o o o
rv co iv.
CM
t- «- CO
in in o
T- it CM
T- «- CO
CO
in C
c 'm
2 H
1- c
c o
O '43
•j_, CO (/)
£ £ =
<5 ~ s:
& „= -
^ % |
0 0 g,
\- \- b.
402
-------
TABLE 113
TiO2 INDUSTRY QUESTIONNAIRE
MILLS, ETC.
Plant Tvoe
1 Ball
Roller
Fluid Energy
Pebble
2 Micronizers
Sand
Belt Driers
Rotary Filters
3 Ball
Hammer
Micronizer
Roller
4 Jet
5
6
7
Number
3
9
6
1
3
3
2
7
2
6
5
1
3
CONFIDENTIAL
CONFIDENTIAL
Size
7ft x 16ft
50 in. dia.
42 in. dia.
6ft x 22ft
200 ton/day
200 ton/day
200 ton/day
200 ton/day
8ft x 6ft, 8ft x 5ft
4 ton/hr
4-42in., 1-36in.
—
36in. dia
403
-------
similar for both processes.
The major sulfate process emissions can be attributed to the use of sulfuric acid. Emissions
from various parts of the plant include SO2, SO3, H2SO4, sulfuric acid mist and possibly various
metal sulfate particulate. The properties of these various pollutants as they pertain to emission control
are sufficiently well understood, due to their widespread occurence, that discussion of their individual
properties will be omitted here. Their generation in the sulfate process is such that they can all
be expected to be present together in any given source. Their relative concentrations will vary
from source to source and, for a given source, can vary with time and operating conditions. Further-
more, the physical properties (temperature, dew point, etc.) of the gas streams range from relatively
mild to quite severe. These matters will be discussed in more detail in the next section.
Chloride process emissions include (besides ore and TiO2 particulate) CO, HCI, C\2, TiCU
and coke particles. Large amounts of energy are often generated on site and this can be the source
of significant amounts of emissions where oil or coal fired boilers are used.
Titanium tetrachloride hydrolyzes readily in air to produce a dense fume of finely divided
particulate so that any leaks in the system or stack emission of this material tends to be very
prominent. Carbon monoxide is often present in large enough quantities that its use as an auxiliary
fuel is sometimes feasible.
B. Source of Emissions
1_. Sulfate Process — The principal digestion reaction, in simplified form, is:
FeO • TiC-2 + 2H2SO4 -> TiOSO4 + FeSO4 + 2H2O
A typical charge to the digester may consist of 1.1 to 1.5 tons of 94% H2SO4 per ton of ore.36
The proportions depend partly on the composition of the ore. Water and steam are added to dilute
the acid slightly and to raise the temperature. The reaction is exothermic and once initiated proceeds
vigorously with considerable evolution of heat and steam. The temperature may exceed 200°C
during the peak reaction period which may last from a few minutes to a half hour.
The severity of the reaction may in some cases be controlled somewhat by gradual addition
of the reactants. In this way, a period of moderately high reaction will be maintained which eliminates
the high peaks. Temperature in this case can be made to remain below 200°C.
The emissions from the digesters will be highly cyclical in nature. During the peak reaction
periods, large amounts of steam containing SOs, SO2 and sulfuric acid mist will be evolved. En-
trainment would be expected to be the principal mechanism by which emissions are released. The
vapor pressure of H2SO4 for a 95% acid solution is 4.8 mm at 200°C. This is small but may
404
-------
still be significant. The quantity of contaminants may be reduced by using gradual addition of
reactants.
Few measurements are available, at the present time, of the emissions from digester stacks.
An estimate of the amount of steam evolved can be made by assuming one mole of steam per
mole of sulfuric acid charged. This calculation gives about 450 pounds of steam per ton of 95%
acid charged.
The calcining operation is the most critical step in the manufacturing operation as well
as the most significant from a gaseous emission standpoint. Gas or oil fired rotary calciners are
usually used. Temperatures as high as 900 to 1000°C are encountered at the discharge end.
Residence time for the TiOa may be as high as 24 hours.
The hydrous titanium dioxide undergoes several washing steps prior to calcination. Even
after thorough washing, however, the TiO2 retains around 5 to 10 percent h^SCX36 This represents
the most important emission from calcining. Processes have been proposed for neutralizing this
residual acid with alkaline material prior to calcining, but this is not normally done.
Most of the water is removed by the time the material reaches 200°C. At this temperature,
the sulfuric acid begins to evolve. Removal of sulfuric acid and SOa is complete at about 650°C.
The atmosphere in the calciner is reported in the literature to contain 2 to 10% oxygen
on a dry basis and 30 to 50% water vapor.36 The oxygen content of the calciner can be important
in that it may influence the ratio of SOa to 862 in the exhaust gas. The reaction equilibrium be-
tween SOa and SOa may be written:
''so
SO2 +1/2O2 ^ SOa , Kp = ^— 1/2
Pso2 Po2
At 650°C, the equilibrium constant, Kp, is about equal to 5 (for pressures measured in atm.) and
Kp = 23 at 550°C. If the calciner is relatively rich in oxygen, virtually all the sulfur oxide should
be in the form of SO3. Where very low oxygen concentrations are present, a significant amount
of SO2 will be formed. If a slightly reducing atmosphere is maintained, even more SOa will dis-
sociate. This assumes that the gas will approach equalibrium at some point in the calciner. This
in turn will depend on the residence time of the gas phase and on the degree of catalytic effect
in the calciner. Furthermore, the kinetics may allow an approach to equilibrium above a certain
temperature but not below.
The importance of this is that it may be possible to exercise a degree of control over
the SOa to SC>2 ratio. Economic, or other, considerations may in some cases favor the removal
405
-------
of one form of sulfur oxide over another.
The digesters and the calciners represent the major sources of emission in a sulfate process
plant. Ore and pigment handling and grinding operations can be significant sources of dry particulate
emissions. Finally, the various treatment and finishing steps are potential sources, primarily of the
odor nuisance type. While the nature of such items is usually proprietary, various reduced sulfur
compounds (HsS, mercaptans, etc.) are sometimes cited as treatment agents. These have very
low odor thresholds and, for those plants which use such substances, present a potential local
odor problem.
2: Chloride Process — The chloride process, a more recently developed technology than the
sulfate process, is subject to a greater degree of process variation from plant to plant. These
variations are difficult to define since they are often considered proprietary by the practitioners of
the chloride process. The chlorination and the oxidation steps are basic to the process and are
always practiced in some form.
The principal source of emission from the chloride process is the chlorinator off gas. This
may contain CC>2, CO, HCI, Cb and small quantities of sulfur compounds (from sulfur in the
coke). If air, rather than pure oxygen, is used in the TiCU burner, then significant volumes of nitrogen
will also be present in the exit gas.
The various parts of the plant which contain TiCU are potential emission sources. Vents,
relief valves and rupture disks are found in various parts of the system such as purification trains,
storage tanks, etc. These can be the source, perhaps only intermittently, of the dense, white
hydrolysis product formed when TiCU comes in contact with water vapor. While the quantities
from these sources tend to be small, they can present opacity problems.
As in the sulfate process, the various treatment steps offer potential problems, primarily
odor. It is believed, however, that chloride process pigment requires less subsequent treatment
so that the pollution potential from this source may be less than that from the sulfate process.
C. Measurement of Emissions
Each of the TiO2 manufacturing processes deals with a different set of pollutants. The
sulfate process emissions are similar to those encountered in the manufacture of sulfuric acid.
The methods for measuring such emissions have been extensively developed and procedures can
be found in The Federal Register (Vol. 36, No. 247, p. 24893). This method makes use of an
impinger train, with a filter between the first and second impinger, and a barium perchlorate —
thorin indicator titration. Some investigators have suggested that the filter should be placed before
406
-------
any of the impingers and the assembly heated up to and including that point to prevent the oxi-
dation of SO2. TiO2 and metallic sulfate can interfere with the procedure and it is important that
these be excluded from the sample. Where it is suspected that these interferents have not been
totally eliminated, a sodium hydroxide — phenolphthalein indicator titration can be used to check
the results, though this method will respond to any other acid substance.
The chloride process emissions include CO, Cl2 and HCI. Carbon monoxide in chlorinator
off gas is usually present in such quantities that an orsat type analysis can be performed. A pro-
cedure for measuring HCI and free chlorine is available and is described in a NAPCA document.37
The gas sample is drawn through an impinger train using alkaline sodium arsenite in the absorbing
solution. Total chlorides are determined using the Volhard titration method (back titration with
ammonium thiocyanate — ferric alum indicator) on a suitably prepared sample. Free chlorine is
determined by titrating an aliquot with iodine solution and a starch indicator. The Volhard chloride
titration method will respond to any chloride (e.g. TiCU, FeCb) so care must be exercised in sampling
the gas stream and in interpreting results.
D. Raw Data Tabulated
1. Questionnaires — A considerable amount of emission data has been reported in the question-
naires. This has been gathered together by type of operation and summarized below. Wherever
possible, emissions are summarized at the process outlet before any control devices. In some cases,
this has involved taking emissions at a control device outlet and back calculating using the reported
efficiency for the device. It should be realized, of course, that there exists some danger in this
type of approach in that a slight difference in the efficiency reported can result in a large change
in the number obtained by back calculation. This is particularly true when very high efficiencies
are claimed as in the present case. For this reason, the actual data reported at the control device
outlet (where this is the only information given) will always be reported in the summary below as
well as the calculated inlet loadings. The reader should refer to Tables 109 through 113 for pro-
duction, raw material and process equipment data on these plants.
a. Digestors — Only one questionnaire (Plant 5) presented any numbers which are usable. For
particulate (including sulfuric acid), emissions after control are 0.66 Ib/ton TiO2 produced. For
SO2, emissions are 0.31 Ib/ton TiO2. Since the details of the control device have been requested
confidential, it is not possible to back calculate the emissions from the digestor stack itself. Total
flow rate from the digestor stacks is reported to be 19,200 to 40,000 SCFM.
407
-------
b. Sulfate Calciners — Plants 1, 5 and 7 reported emission data from sulfate process calciners.
Plant 1 reports two calciners with gas flows of 65,000 ACFM and 28,000 ACFM, respectively (both
at 170°F) and atmospheric emissions after a 95% efficient control system of 12 Ib/hr and 5 Ib/hr
H2SO4, respectively. Back calculating, this translates to 30 Ib H2SC>4 per ton of TiO2 at the calciner
exit.
The same plant reports SO2 emissions from the two calciner scrubbers as 110 Ib/hr and
48 Ib/hr, respectively. The scrubber system used has minimal effect on SO2, so these can be
considered to represent emissions at the calciner exits themselves. By calculation, an S02 emission
factor of 17.1 Ib/ton of TiO2 product processed is obtained.
Plant 5 has reported SO2 emissions from the scrubber stacks for its calciners. These
calculated to be 11,830 Ib/day. This results in an emission factor of 80 Ib SO2 per ton of TiO2 pro-
duced. Insufficient information was presented to permit a calculation of an H2SC>4 emission factor
at the calciner exit.
Plant 7 reports a calciner effluent of 44,000 cfm at 950°F having a loading of 0.95 gr/cf
for particulate including sulfuric acid. This resulted in a calculated emission factor of 63 Ib/ton of
TiO2 for total particulate including sulfuric acid.
c. Drying and Milling — Drying and milling operations find extensive use in the TiOa industry whether
the sulfate process or the chloride process is practiced. They can be applied to ore, coke or
finished pigment. A summary of the emission information given for this area is presented in Table
114 for the sulfate process and in Table 115 for the chloride process. The last column in these
tables, the emission factor, is a calculated estimate based on the information given.
The emission factors in pounds per ton of TiO2 product show a good degree of consistency,
with one exception, as far as orders of magnitude are concerned. This is remarkable when com-
pared to similar data obtained for other aspects of the paint and varnish industry.
d. Chlorinators — Five plants reported some information on their chlorinator emissions, though, in
general, the data was not as well defined as in the previous section. Plants reported either off gas
analyses or actual emission rates for various substances. In some cases, emissions are reported
after an emission control system. In other cases, either no emission control devices are present or
emissions are reported upstream from such devices. In the summary that follows, an indication will
be made as to whether the reported emissions are for controlled or uncontrolled processes. Finally,
consistent with other sections of this report, emission factors for uncontrolled chlorinator emissions
will be estimated.
408
-------
CM
o
CM
O
CSI
o
CM
O
CM
O
LLI
c
o
c
o
c
o
o
co
c
o
.£
LD
o
_Q
O)
C
cu
CO
s?
ca
<
DC
a
I
CO
CO
CO
O
O
or
a.
LLJ
(-
D
CO
LLJ
"5
•M
C
O
0
C> 1
O>
LL
cn
C
T3
CO
O
— '
O
=£ ^
& cn
ID «N
«- LO
1
LL
CJ
CO
cn
0
in
cn
CD
.c
CM
o
d
1
LL
O
CO
IB
LO
0
o
o
ci
u--
cn
O)
_
.c
CD
«
0
'4-"
CO
o
o
3
O
S
(0
ii:
'o.
£
it
LL
_0
LL
CO
0
< LL
Is
o
CD (SJ
cn
C
to
t/t
s
o
i
E
c
CD
cn
c
1"
CD
O
4^
C
cu
+-1
LL
^
LL
0
< LL
§°LO
ID_ f*-
ID" (§*
Ol
c
•a
cu
6
_cu
c
CD
LL
2
LL
O
< LL
||
D)
C
'E
cu
CD
•4-J
CD
O
a>
LL
S
LL
CJ
§°s
cn
C
'2-
TD
E
O
cu
c
CD
LL
5
LL
0
<
co"
cn
.E
E
cu
CD
CJ
<§>
O
cu
LL
O
CO
O
8
00
ro
c
cu
cn
b_
•a
0)
CD
_a
o
•t-j
o
c
o
O)
c
cu
E
o
o
c
a>
cu
CD
O
a
CD
c
CO
O
4-J
CD
Q
409
-------
-1?
CD CD —
3. §
LU
CM
O
O
CD
O
CM
O
c
O
in
to
CM
O
c
O
.0
o
CM CM
g g
i- i-
c
O
£
LO
c
O
O
CO
LU
_l
CD
C3
cc
Q
I
C/)
2
O
CO
CO
LU
0
O
QC
D.
LLJ
Q
QC
O
O
LL
O
CO
en
CM
d
LL
CJ
CO
OJ
O
O
q
d
u
CO
_
O)
CO
c
O)
c
cu
c
_o
o
LL
U
CM
CM
LL
O
CO
^_
en
LO
CM
CO
O
CD
C
_o
o
o
LL
CJ
CO
en
ITi
CD
C
^
o
CJ
8
o
c
cu
en
O.
O)
c
T3
CU
6
0)
c
en
c
T3
0)
O
CJ
en
c
CD
CO
410
-------
Plants 2 and 5 present emission information from controlled chlorinator stacks. This infor-
mation is summarized in Table 116. Plant 2 has given the off gas analysis for various years. Of
interest are the trends of improvement for all reported constituents except carbon monoxide. Plant
5 has reported actual emission rates.
Plants 4, 6 and 7 have reported data for either uncontrolled chlorinator stacks or for con-
ditions upstream from control equipment. Plant 4 reports a typical off gas analysis which is tabulated
in Table 117. Emissions of TiCU for this plant were reported to be 22 to 88 Ib/hr.
Plant 6 has listed emissions directly out of the chlorinator tail gas stack. These are also
summarized in Table 117. Emission factors for chlorine and HCI are 4.3 Ib/ton TiC>2 and 39 Ib/ton
TiOa respectively. This plant has presented a complete emission inventory of 48 sources which
includes boilers, chlorinator, dryers, kilns, tank vents, etc. Table 118 lists emissions for the power
house, chlorinator, and the total for all sources including power house and chlorinator. Examination
of the table indicates that the power house and chlorinator account for the major fraction of total
emissions from this chloride process plant.
Finally, plant 7 has listed uncontrolled emissions from its chlorinator off gas as including
1 Ib-mole TiCU/hr and 6 Ib-moles HCI/hr. Correlating these with production, emissions factors for
TiCU and HCI are obtained which calculate to be 67 Ib/ton TiO2 and 77 Ib/ton TiOa respectively.
Plants 2 and 5 present emission information from controlled chlorinator stacks. This infor-
mation is summarized in Table 116. Plant 2 has given the off-gas analysis for various years. Of
interest are the trends of improvement for all reported constituents except carbon monoxide. Plant
5 has reported actual emission rates.
Plants 4, 6, and 7 have reported data for either uncontrolled chlorinator stacks or for con-
ditions upstream from control equipment. Plant 4 reports a typical off-gas analysis which is tabulated
in Table 117. Emissions of TiCU for this plant were reported to be 22 to 88 Ib/hr.
Plant 6 has listed emissions directly out of the chlorinator tail gas stack. These are also
summarized in Table 117. Emission factors for chlorine and HCI are 4.3 Ib/ton TiO2 and 39 Ib/ton
TiO2, respectively. This plant has presented a complete emission inventory of 48 sources which
includes boilers, chlorinator, driers, kilns, tank vents, etc. Table 118 lists emissions for the power
house, chlorinator and the total for all sources including power house and chlorinator. Examination
of the table indicates that the power house and chlorinator account for the major fraction of total
emissions from this chloride process plant.
411
-------
TABLE 116
CHLORINATOR EMISSIONS AFTER CONTROL
Plant 2 - CHLORINATOR OFF GAS
TYPICAL ANALYSIS — PERCENT
Constituent 1972 1971 1970 1968 to 69
CI2
S02
Cl
02
CO2
CO
N2
TiO2
0.00006 0.0001
0.00007 0.0012
0.056 1.5
2.7 4.0
25.5 30.0
27.5 24.0
44.0 40.0
3 Ib/hr 6 Ib/hr
0.0001
—
2.0
4.0
30.0
24.0
40.0
8 Ib/hr
0.0001
—
3.0
4.0
22
26.0
45
12 Ib/hr
Plant 5 - CHLORINATOR OFF GAS
EMISSIONS — LB/HR
Constituent
NOx
CO
Participate
CI2
Rate, Ib/hr
0.34
1570
0.82
10
412
-------
TABLE 117
CHLORINATOR EMISSIONS BEFORE CONTROL
Plant 4 - CHLORINATOR OFF GAS
TYPICAL ANALYSIS, PERCENT
Constituent %
N2
CO2
02
CO
TiCU
HCI
SiCI4
5 to 10
50
0.5
10
0.2
0.1
0.1
to 70
to 2.5
to 20
to 0.5
to 0.5
to 0.4
Plant 6 - CHLORINATOR EMISSIONS, LB/HR
Constituent Emission Rate
SO2 0.00
Particulate 0.00
NOx 1.05
CO 3,195.00
Hydrocarbons 0.00
CI2 53.00
HCI 420.00
413
-------
TABLE 118
Plant 6 - EMISSION INVENTORY
Power House Chlorinator Total All Sources
Constituent Ib/hr Ib/hr Ib/hr
SO2 855.0 0.00 855.13
Particulate 104.0 0.00 126.38
NOx 130.2 1.05 140.57
CO 17.4 3,195.00 3,213.94
Hydrocarbons 8.4 0.00 10.84
C\2 — 53.0 86.40
HCI — 420.0 422.2
414
-------
Finally, plant 7 has listed uncontrolled emissions from its chlorinator off gas as including
1 Ib-mole TiCU/hr and 6 Ib-moles HCI/hr. Correlating these with production, emissions factors for
TiCU and HCI are obtained which calculate to be 67 Ib/ton TiO2 and 77 Ib/ton TiO2, respectively.
e. Summary of Emission Factors — The above information was used to calculate the emission
factors presented in Table 119 for the various steps in TiC>2 pigment manufacture. It should be
emphasized that these represent estimates for uncontrolled processes. Most of the plants reported
emission control equipment on most of these processes. Where required, stated control efficiencies
and outlet loadings were used to back calculate the uncontrolled emission factor. With the exception
of SC>2 and CO, existing controlled plants report a significant reduction in emissions from the levels
suggested by the emission factors. Sulfur dioxide and carbon monoxide are relatively unaffected
by the types of control devices presently in use.
It should be emphasized here that considerable variation exists in the manner in which
these various process are run, particularly the chlorinators. The emissions from chlorinators can be
a function not only of control devices but also of process variations, type of ore used, etc. Since
much of the technology in this area is proprietary in nature, not all processing options are available
to all producers. Consequently, a producer operating a process that is inherently higher in emissions
than a process used by some of his competitors may not be able to achieve as low emissions
by process modification and may require the addition of an economically unfeasible amount of
control equipment. This could put him in a non-competitive position if uniform emission standards
were set for the industry.
One other source of emission is mentioned in one of the questionnaires which seems to
be unique to that particular plant (Plant 4). This plant reports a significant emission of sulfur mono-
chloride, SaCb, from an off gas stream from the TiCU oxidizer. Emission rates of 5.1 to 22 Ib/ton of
TiO2 produced are reported. In the normal operation of the chloride process, the oxidizer does
not vent to the atmosphere directly so that the precise source of this emission is not clear. It is
believed that, as this plant operates the chloride process, S2CI2 is used to temporarily absorb
the chlorine from the oxidizer. Chlorine recycle can then be accomplished by stripping the absorbed
material from the SaCb. The S2CI2 emissions may result from the tail gas from a chlorine absorption
step.
Z Other Sources — Two other sources of information have been obtained on digester emissions.
The St. Louis County Health Department supplied test results for the ML Industries sulfate plant
in St. Louis. They report emissions downstream from a set of scrubbers to be "nil" for sulfuric
415
-------
TABLE 119
ESTIMATE OF EMISSION FACTORS FOR UNCONTROLLED PROCESSES
Process
Sulfate Digester*
Sulfate Calciner
Pollutant
Emission Factor*
(Ib/ton TiO2 Produced)
H2SO4
SO3
SO2
H2S
H2SO4
SO4
14.1
8.8
4.2
0.9
46
48
Ore Drying
Sulfate
Chloride
Participate
Participate
47
44
Coke Drying
Paniculate
75
Ore Milling
Sulfate
Chloride
Particulate
Particulate
42
Pigment Milling
Sulfate
Chloride
Particulate
Particulate
49
Chlorinator
TiCI4
CI2
HCI
CO
32
4.3
55
260
"These are average emission factors. Ranges are given on the preceding tables and discussed in the
text.
"Emission factors for this process based on paper by L.L. Falk discussed in the following section. The
remainder of the factors based on the TiO2 Industry Questionnaires.
416
-------
acid and 7 Ib/hr for SC>2. These scrubbers are claimed to be "100%" efficient for H2SO4 so it is
not possible to determine the inlet loading.
The best defined information obtained thus far is contained in a paper presented by L. L.
Falk.38 He reports digester emissions from the duPont plant in Baltimore (not included in the
questionnaires). Figure 79 gives total gas evolution from the digester as a function of time after
oleum addition. It can be seen that the reaction proceeds at a negligible rate for about the first
45 minutes and then proceeds at a vigorous rate for a very short period of time. A considerable
amount of gas (mostly steam) is evolved during this period.
Typical emissions per batch are given below:
Material Ib/batch
Steam 8,000
H2SO4 120
SO2 36
H2S 8
SO3 76
The quantity of materials charged to the digester in a typical run was not given but it
is possible to make an estimate. Assume a batch charge of 20 tons of ilmenite (Barksdale, p. 220).
Further assuming the ilmenite to contain 50% TiO2, and overall plant recovery to be 85%, the
emission factors can be estimated as follows:
Emission Factors for Sulfate Process Digestors
Quantity
Emission (Ib per ton TiO2 Product)
Steam 940
H2SO4 14.1
SO2 4.2
H2S 0.9
SO3 8.8
Atmospheric emissions tests on the National Lead Company, Sayreville, New Jersey,
sulfate process plant was conducted by the Division of Air Pollution of the Public Health Service
in September, 1966.39 This plant has seven calciners. The sampling team tested the emission control
system for one of the calciners. The system consisted of an initial water scrubber, followed by
electrostatic precipitators, followed by another scrubber. A sketch of the system is shown in Figure
417
-------
E
<*-
u
O
125
100
75
50
25
30
35
40
45
5O
55
TIME FROM OLEUM ADDITION , MINUTES
FIGURE 79
EXHAUST RATE FROM SULFATE PROCESS DIGESTION TANKS
418
-------
I
,JO.
0
a:
u
D
it
o
h
uo
^.
FROM
CAL-CIIMER
w
Z
0
u
0
J
ID
Z
J
QL
2
LJ
cr
O
Li
h
0)
in
0
cr
h-
o
o
o
tn
(/)
y
cr
u
z
o
0
419
-------
80. They were unable to obtain reliable measurements at the calciner outlet (due to interference
with the analytical technique from TiOa) and made no attempt to monitor the exhaust from the
first scrubber. They reported the results of measurements at the outlet from the precipitators and
on the water from the last scrubber. The average results obtained after the precipitators are shown
in the following table:
Pollutant Rate (Ib/day)
SO2 7,900
SO3 310
Paniculate 380
Acid Mist 180
Total gas flow rate at this point was measured (on a dry basis) to be about 3 x 107 SCF/day.
Several comments should be made before making any conclusions from this data. First, the results
of individual runs showed deviations of as much as a factor of two. It is not known whether
these are due to inherent difficulties in obtaining good samples or whether there are, in fact, large
fluctuations in the calciner exhaust.
The quantity of emissions reported raises some questions. It was assumed by the sampling
team that each calciner handles about the same amount of material. If this is correct, then the
calciner in question processes about 62 ton/day of TiO2. If the charge had retained 10% HfeSCU,
then this would result in an emission of about 4 tons of sulfur (as 862) per day. It can be seen
that the 862 reported after the precipitators is about sufficient to account for almost the entire
total. It is not known how much SO2 had been removed by the first scrubber nor how much acid
mist was removed in the precipitators.
If these devices account for considerable amounts of sulfur oxides, then the reported emissions
seem to be too high. On the other hand, most of the calciner exhaust emissions can be in the
form of SO2 only if a somewhat reducing atmosphere is maintained in the calciner. Since no infor-
mation is available on these aspects of the problem, some caution should be exercised in accepting
the data as typical.
NL Industries has reported that, due to an error in the method of calculation, the Public
Health Service results are high by a factor of two. Variable oxidation of sulfite to sulfate in the
last tower was felt to have further increased the error, making the results high by perhaps as
much as a factor of three.
In view of the above comments, several suggestions can be made for source testing. If
420
-------
at all possible, the calciner exit itself should be sampled. The problem here is that TiO2 participate
may interfere with the analysis for sulfate. Techniques should be sought to overcome this.
It should be determined whether the calciner operates in a relatively steady state condition
or whether large fluctuations occur in operating conditions. A sample of the TiC>2 cake prior to
calcination should be obtained and analyzed for sulfuric acid. This, plus a knowledge of the calciner
throughput, should provide a good estimate of the total emissions. Finally, an analysis of the
calciner exhaust for oxygen would provide a means to check whether the SOa to SC>2 ratio measured
is reasonable.
IV. EMISSION CONTROL TECHNOLOGY
A. Description of Currently Used Control Systems
Table 120 summarizes the particulate emission control devices reported in the question-
naires. Included in this table are those devices which collect dry, solid particulate from drying and
milling type operations only. Table 121 discusses those devices whose purpose is to control pro-
cesses applicable only to TiO2 manufacture such as digestion, chlorination, etc. Plant 4 reported no
emission control at all, while plants 5 and 6 considered at least part of the information proprietary.
A wide variety of devices is reported including cyclones, fabric filters, several varieties of
scrubbers and electrostatic precipitators. There are several instances of devices connected in series
in order to obtain good control. Over all control efficiencies, where reported, are said to range
from 90 to 99+% depending to some extent on the process being controlled.
B. Other Methods of Control
Most conventional pollution control devices that would be applicable to the types of emis-
sions found in TiO2 plants are presently in use. The application of more advanced concepts is
discussed in Chapter 9.
C. Performance of Currently Used Control Systems
Performance data on presently used systems, as reported in the industry questionnaires,
is included in Tables 120 and 121 listed earlier. For particulate control devices, efficiencies up to
99%, and higher, are reported. Such efficiencies, on a weight basis at least, can be achieved
under favorable conditions with the types of devices in use. Beyond that, it is not possible to
comment on the accuracy of the reported performance figures. It is not known whether the ef-
ficiencies are based on actual tests.
For other processes (calciners, chlorinators, etc.), data on efficiencies is not consistently
421
-------
o
CM
LJJ
CO
CO
LJJ
rr
z
o
CO
UJ
O
cr
CO
13
Q
d
i-
co
LLJ
o
LJJ
Q
0
cr
i-
z
o
o
LJJ
5
O
1
cr
Q.
•SS
ro (J
"1
o
c
0)
'o
M—
LU
°>2
slfe
C ;P C/3
•"• O i-^C
~" 0)
IE
iT 13
CO
CD
_ O
E >
CD CD
03 Q
C/) "6
O CD
CD •?
a. b
>, ZJ
1- Z
III 1 1 "*
O III 00 CN CM ' ' CN 00 CN r-
CN CN CO CO C) "3- CO CO
4-1
in
°°. III+ + +I — + + +
O)O) O)O)O) O)O) O> O) O)
' ' ' CN O O ' ' CO •* -
Lfi IT) LT) CM CD p~-' -
H
o ^
0 C
°°- u
ooo oooo'~o 0002
| O ID O O CD ^" ID O O O O O C
O CM LD ID * — ^" ^f +~( O 00 CD CO f
«-" »-" ID" CN" «-" § «-" CN" oo" CD"
o "~
C\J
CD
^
D
CJ
CO
3
0 - : »- 'br 'fc 't -p £1 '!= '= '^ '^
Q,^ 0^0 ^^^Cgj-Q ii^lJ3
— — CD ^^ CD •^ CD CD CD UJ ^> Cu CO CD CD
LLJU. OLL> LL U. U. > > LL. LLLLLL
ID
+
^f O> O) O) ID O) CN
O) O> 00 O5 O^ O3 C3^
i CD ID CN
: <- CN co ID CM
- od co r~ d «— 06
CM T-
7
J
3
I CMCDOO OOO
3 CNO5OCN OOO
j O)IDCOO LDOCD
CD «* ID O) CN O «—
ID ^T «-
1
*§
"D
^* C\J
^
>, (n ^4- ^ .Q
2 * i -§ •§
^- O ' Q CD o
CD CD ^« CD t^ CD
C C 0 C "Z. 0 C '^
J2 S ~ J2 2 -c J2 3
< c
' I
£ o,
O Q.
c _cn _co jo
S co II I
ci .2 i i E S S
2 ^ E -D E E * ^ ^
Q (£ O OOOCLCLCL
.
5 "
•n <"
= co « *
coco E fOqJo-- i- ° 2
E S > to o o
OOOCL OCJCLCO" OOO
c
51
CO
CD
422
-------
c\i
UJ
CO
£
~
CO
UJ
rr
<
z
O
H
co
UJ
0
>-
oc
f-
co
D
_
6
f-
CJ
c
a;
'o
4—
UJ
se
a
o
D
CU
k_
^
0
g
tfl (-^
If) c
Q)
1_
Q_
CO
UJ
^s
1 °^
1 LO
O5
1 §
o _
UJ O £
Q LL
_J
£
O
O
^
0
CO
CO
_CJ
1 1
CO
. S
1 E >
*£ CO 0
UJ 1 D
>• H-
CO 0
"o 0
&l
>. 3
1- Z
LO CM
1 1
1 i
CO '=
V- D
CO <-
_o c
JO 0
D >
0 t
CO 1
1 1
5 i
T3 ^
S «
£ Q-
5 co
to
-
JD 0
-0 -9
2 "§
O w
2 us
fl) Q)
U,
c
o
Chlorinator
TiC^ purificati
CM
1
1
O
8
CM"
Venturi scrubber
4->
C
CU
>
Waste chloride
1
1
o
o
o.
o>~
0
«
Q.
_co
CO
JO ,_
0 0
| |
S? §
00 CO
£
CD
+-J
(/>
>s
•M (/>
C -
1 "<"
E x.
O) CD
Oxidizer and pi
Separation, "sn
c
o
o
_cu
"o
CJ
CD
O
O
L_
O
1
1
8
!-•__
^"
Packed column
15% NaOH
i_ C
0 0
**— '~
^J
CD
"- CD
OD 5
_Q ^^"
1 a
2 CM
"Emergency sc
oxidizer & TiO
equipment
s?
1 0)
O)
S N
0
o
CO
CO 0
0 o.
o •*
0 CM
0
cJ
*
|CO
1
1 CO
E I
co l_
CO "5 O
ut
2 &'§
*s^
•c « 6
3 CO 0
S (CM LU
Digestion tank;
Calciners
CO
~1
CO
1
><
CO
Q.
CO
_i,
_l —I _
< < <
1- 1- H
-z.-z.-i
LJJ LJJ U
D D C
LL LL L
Z Z 2
O 0 C
out
1
I
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
^
c
CD
+->
C (/j
° CD
+-j £-
Q) .—
§ &s
Z DO
as
?
5?
CM
CO |
f^-
C
CO
c" +
LD .E
CM
^
^
LL
<£
8 8
o o
J -1 LO" •*
; < CM «s-
- H
: -z.
J LJJ
3 D
L LL Q.
S Z 'o
30 £
JO a
i
o
CU
"CD
t
*_
CD
JD
JO
D
£
^ ^
CD CD
-Q C
J2 O
S "o
0 >•
CO O
!_
0
ra S2 S2
C O
c
f£
CU
en
c
"a.
E
t
3 >
= 1
§ ^
t JS
>
en
CD
O
*J
Wl
C
o
-4-J
CD
O
Chlorinator
TiCI4 Purifi
CO
CU
't-
CD
CO
c
CO
CD
0
>
0
"O
"5
"c
o
CO O
* -c
CO ~
co S
C CO
rol devices ii
-allel system
C &
o a-
0 co
423
-------
reported. For calciners, three plants report efficiencies in excess of 90%. The basis of calculation
is not stated. In each case, the emission control system consists of two or more devices in series.
One plant reported control efficiency for a chlorinator emission control system. Efficiency of
90% or better was reported for TiCU and HCI. An overall pressure drop of 45 inches of water
through three devices in series (one a venturi scrubber) suggests that the level reported is not
unreasonable. As before, however, no further evaluation of the data can be made.
424
-------
CHAPTER 9
RECOMMENDED RESEARCH AND DEVELOPMENT PROGRAMS
Significant among the purposes of this industry study is the recommendation of Research
and Development programs which can lead to improvement in air pollution measurement and control.
In the following pages, the deficiencies in both air pollution technology and emission measurement
are treated in detail. Consideration is also given to that control technology which is more efficient
than "best control" but is considered to be economically unfeasible. Based on the findings of the
Research and Development study presented herein, priorities to improve control technology are
recommended along with suitable programs to achieve these improvements.
I. EMISSION CONTROL TECHNOLOGY
The areas of control technology within the coating industry which would benefit most by
the application of research and development efforts are described. In those segments of industry
where a problem has been selected, an attempt has been made to assess the probable requirements
in time and cost for successful program accomplishment.
A. Technical Developments for Reduced Levels of Emission
The areas where improvements could be made to reduce levels of emission are categorized
as follows:
PROCESS CHEMISTRY AND KINETICS
PROCESS EQUIPMENT AND/OR OPERATIONS
CONTROL EQUIPMENT AND/OR OPERATIONS
Requirements in these categories are outlined below.
1_. Process Chemistry and Kinetics — Odor and solvent emissions from opened and closed
kettles represents one of the major air pollution problems for the industry, due primarily to the
local nuisance problem generated by this type of emission. Cooking in open kettle has dropped
off to the point where it no longer deserves further attention. Most resin and varnishes are now
cooked in closed kettles which have resulted in significant reduction in emission levels. These
emissions could be further reduced if process chemistry or cooking formulas could be developed
for processing in a truly closed or pressurized kettle.
425
-------
2^ Process Equipment and/or Operations — Raw material handling as well as process chemistry
also must be developed for the use of pressurized kettle cooking. These are required to allow
for addition and removal of materials during the cook while still maintaining a closed system.
Development of an in-line, closed thinning, blending and filtering system for handling of the materials
after cooking would also substantially eliminate the emission currently encountered with existing
thin tanks and leaf filters.
3. Control Equipment and/or Operations — Although a wide variety of air pollution control devices
are used in the paint industry, only a few types may be considered applicable. These are:
Thermal and catalytic afterburners for cooking operations
Scrubbers or condenser-scrubbers for cooking operations
Fabric collectors for milling, grinding and other dry solid processing operations
The principal deficiencies associated with the operation of afterburners fall into two categories:
1. Those problems dealing with the inherently large fuel consumption and cost.
2. Potential for safety hazards and damage to the equipment by fire or explosion.
The first of these two problems is becoming increasingly urgent as natural gas becomes more
difficult to obtain. The second problem has caused numerous casualty losses, as well as injury to
employees over many years in the cooking industry.
For those process reactions which are not amenable to closed or pressurized cooks, or in
circumstances where physical or capital limitations do not permit replacement of the present cooking
equipment, a more economical form of oxidation, with regard to operating cost and fuel conservation,
is needed. Three directions suggest themselves for the correction of deficiencies in afterburner systems
for kettle operations. These are:
1. Improved heat exchanger efficiency for thermal afterburner units
2. Improved activity and stability of catalysts for use in catalytic afterburner units
3. Direct flame incinerator
Of these three, a direct flame incineration system of low first cost offers the most potential to provide
optimum cost-benefit relationships for the public.
Thermal afterburners are inherently expensive. Their capital cost increases as the gas flow
rate increases, almost without relationship to the concentration of organics in the fume stream. The
addition of heat exchangers increases thermal efficiency, decreases operating cost and significantly
increases initial capital cost. The best heat exchangers of conventional design (i.e. fixed tube exchangers
and rotary matrix heat exchangers) have thermal efficiencies on the order of 60 to 70%. Some
nonconventional regenerative designs using fixed beds of checkerwork, stone, etc. are being offered
426
-------
with design efficiencies in the range of 80%. Continued competition among the manufacturers of
afterburner equipment is likely to lead to marginal improvements in the relationship between first
cost and heat exchanger efficiency. Therefore, it is not felt that thermal oxidation should be considered
as an area for funded R&D work.
Catalytic afterburners have been used for many years in the abatement of organic fumes
from kettle cooking operations. Operations on a single kettle tend to be widely varied as alternative
formulations are prepared, and, therefore, the potential for inclusion in the cook of one or more
catalyst deactivating or poisoning agents is always present. The problem of maintaining activity in
the presence of contaminating or deactivating agents becomes increasingly severe as the improved
catalyst performance is desired. For example, a catalyst requiring 1,000°F inlet temperature to oxidize
toluene is only marginally better than a thermal system with no catalyst at all. This low level of
activity is likely to be quite tolerant of poisons, misoperation and the like..On the other hand, a
very active catalyst, capable of oxidizing toluene at 450°F is likely to exhibit a marked sensitivity
to overheating, contamination, etc. The "fine edge" of activity is easily damaged by a variety of
chemicals and by non-uniform or extreme operating conditions. These factors suggest that a major
R&D effort should not be made in the area of catalyst development.
Since most emission from kettle cooking is high enough in combustible concentration to
be flammable, the direct flame incinerator discussed earlier in Chapter 5 offers great potential for
incineration with little or no fuel cost. These units are also commercially available. The area of
research and development required is in the transfer system from the kettle to the unit.
B. Economic Deficiencies Preventing Reduced Levels of Emissions
The minimum size of the independent paint manufacturer is quite small. Even seemingly
modest demands for pollution abatement equipment may present a substantial economic burden
to a small independent. Probably the most difficult problem for the small manufacturer is the
prevention of solvent loss during transfer, storage and mixing processes. Of these, those losses
associated with storage of solvents are most likely to be of concern to air pollution agencies,
and most difficult for the small manufacturer to cope with. The small scale of these emission
problems and the high cost-benefit relationship for emission control systems present a difficult
economic problem.
C. R&D Priorities to Improve Control Technology
Four R&D programs have been selected as a part of this industry study to improve control
427
-------
technology. These programs are as follows:
1. Closed Reactor Design Program
2. Investigation of Methods for Handling Resin Raw Materials in the Liquid Form
3. Development of Inexpensive Scavenging Systems
4. Development of Transfer System for Direct Flame Incineration
The details of these programs along with the work and cost requirements are detailed in the next
report section.
D. Recommended Programs for Achieving R&D Requirements
1^ Closed Reactor Design Program — This program involves the analysis of requirements for
general purpose reactors to produce resin and varnish products in a completely closed reactor
system.
The following paragraphs list the program steps envisioned in the overall R&D program:
1. Detailed library investigation of equipment literature and patent literature pertaining to
closed reactor systems already developed.
2. Detailed investigation of the nature and scope of individual reactions carried out in
existing kettles.
3. Reduction of the physical and chemical requirements of the cooking process to simple
concepts regarding those factors which determine the nature of the equipment required.
4. The specification of a design basis for a non-ventilated reactor system.
5. The collection of potentially feasible designs for review and screening.
6. The selection of alternative equipment types which appear generally suitable for the
requirements in the prototype specification.
7. The fabrication and testing of a prototype reactor.
The proposed program would be conducted in three phases consisting of:
1. Research and selection of approach
2. Design and construction of prototype
3. Field testing of prototype
Phase I should involve approximately 1,000 man-hours of professional R&D work, and about
600 man-hours of non-professional effort. The total estimated cost for this phase is $35,000.
Phase II should consist of a design portion requiring approximately 750 man-hours of professional
and non-professional time. The total estimated cost of the labor is approximately $28,000. In addition,
428
-------
a one-time fabrication cost of the prototype reactor is estimated at $40,000. The total for Phase II
is estimated at $68,000.
Phase III, the installation and field operation of the prototype, will vary in cost depending
upon the ability to arrange with a commercial paint manufacturing company to contribute to the
installation and operating costs for the kettle required. The estimated cost of this phase of this
program is in the range of $100,000 to $150,000 depending on final commercial arrangements.
Of this total, $25,000 is included to cover an estimated 900 man-hours of field test and observation
work.
2. Investigation of Methods for Handling Liquid Resin Compounds — This investigation involves
an analysis of methods for handling resin components in liquid or molten form. The inherent problems
with solid resin powders are two-fold:
1. The ventilating air requirement is large in order to prevent fuming through the kettle
hatch.
2. The added powders tend to become airborne and are swept out of the hood.
Solid raw materials can potentially be handled in liquid or molten form so that batch additions
could be made without handling dry powders. The handling of molten salts at elevated temperatures
is feasible and has been demonstrated in other industries and should not involve any radically
new techniques. The potential problems lie in the development of equipment which can be operated
economically and safely, particularly in connection with small, unsophisticated cooking operations.
The R&D program consists of three phases which are outlined below:
Phase I consists of a literature search to accomplish the following:
1. Review existing technology and define the types and extent of usage of solid raw materials
in resin cooking.
2. Establish the properties of the identified components, including: fusion point, viscosity,
vapor pressure, corrosiveness, stability and other properties.
In addition, literature research and a field investigation should be made on conventional equipment
for handling molten materials in other manufacturing activities.
Phase II should include a definition of alternative process schemes and equipment components.
Based on this definition, an economic evaluation of alternative process schemes and equipment
should be made.
Phase III should consist of the selection and definition of a prototype process scheme for
handling molten salts in the paint and varnish industry.
429
-------
In this program, field testing of the proposed prototype system is not considered an essential
step. The development of a system design which is reasonable from an engineering standpoint
combined with publication of the bases, calculating methods, etc., used should be sufficient to make
proposed system available to designers of new plants in the coating industry.
The cost and timing for the three phases of the proposed program are as follows:
Phase I should require approximately 320 man-hours of engineering and scientist work and
cost about $10,000.
Phase II will require about 320 man-hours of engineering work as well as 160 man-hours
of drafting and should cost approximately $13,000.
Phase III consists of about 320 man-hours of engineering work and 160 man-hours of
drafting and should cost approximately $13,000.
3; Development of Inexpensive Solvent Scavenging Systems — This development program is
proposed with particular emphasis on the needs of the smaller paint manufacturers. The proposed
program considers the development of a simple and inexpensive system for scavenging solvent
vapors from storage tanks and possibly from transfer points as well.
Three basic mechanisms deserve consideration in limiting solvent emissions: prevention of
evaporation, incineration of the tank effluent and adsorption of organics from the vented gases.
The steps in this proposed program are as follows:
1. Literature and plant surveys to establish the size and type of solvent storage tanks
most widely used in the industry.
2. Library and plant surveys to establish the most frequently used solvents as well as
the physical properties of these solvents.
3. Preparation of calculated emissions, and comparison of these with test results prepared
by the EPA.
4. Formulation of performance criteria for abatement systems.
5. Screening of available mechanisms for abatement, and selection of the most feasible
approach to handle each of the following:
a. Prevention of Vaporization
b. Condensation
c. Thermal Incineration
d. Activated Carbon Adsorption
430
-------
6. Preliminary design of a system utilizing each mechanism.
7. Analysis of the alternative designs for selection of the most economical design approach.
8. Preparation of a prototype design.
9. Construction and testing of the prototype design.
This program can be conducted on a relatively modest scale. The research and design
phases should not entail more than 700 hours of professional work and 400 hours of drafting.
The cost for this part of the work is estimated at $20,000. Fabrication of the prototype and subsequent
testing is estimated at an additional $30,000.
4. Development of Transfer System for Direct Flame Incineration — Direct flame incineration
would offer a very inexpensive system to build and operate. It would also reduce the danger of
fires and explosions currently being experienced by this industry. The main problems preventing
use of this system are:
1. Excessive ventilation of reactor kettles when loading materials. This system will work
better with the closed pressurized reactor discussed earlier.
2. The need for a gas tight low flow process blower capable of pumping the reactor effluent
which will contain resinous material and phthalic anhydride particulate. Corrosion
and heat resistance must also be considered.
3. The need for automatic safety valves to meet the requirements outlined in Item 2.
4. The need for a nozzle mix burner to meet the requirements outlined in Item 2.
The program proposed would be carried out in three phases and would follow the same
approach outlined for the closed reactor design program. The three phases would consist of:
1. Research and selection of approach
2. Design and construction of prototype
3. Field testing of prototype
Costs and man-hour requirements are estimated to be about 65% of those presented
for the closed reactor design program.
II. MEASUREMENT OF EMISSIONS
Significant attention has been devoted to the development of techniques and instruments
for testing particulate and organic emission in general. In the paint and varnish industry,
certain deficiencies in measurement exist due to the nature of the cooking process which suggest
the need for additional development work. These deficiencies and required development work are
431
-------
outlined below.
A. Deficiencies in Manual Methods of Source Sample Collection and Analysis
The cooking process is a cyclical batch operation with respect to gas flow and hydrocarbon
concentration. In order to produce an accurate assessment of the total emissions per batch
(or average emission), it is essential to undertake simultaneous measurements of both gas
flow and of organic concentration over a period of several hours. Using manual source testing
methods, this is both an imperfect and cumbersome procedure at best.
The resinous materials vaporized from the cook have high dew-points, and, therefore,
readily condense on any surface which is at a temperature lower than that of the air contacting
the resinous mass. This leads to deposition of organic material in any sort of sampling
system which operates below the temperatures involved in the cooking operations (as high as
700°F). These deposits are troublesome for mechanical reasons and also since they cause
emissions into the gas stream at periods substantially later than the time they enter the
sampling system.
B. Deficiencies in Techniques and/or Equipment for Continuous Monitoring of Source
Emissions
Continuous monitoring of cooking operations is subject to the same deficiencies as described
previously for manual testing methods. In the case of condensation of resinous materials, the
continuous methods are significantly more sensitive than the manual methods. This is primarily
due to the necessity for somewhat remote location on continuous monitoring equipment and the
attendent long length of sample lines.
Such continuous equipment, as is presently available, is very sensitive and is designed
more for laboratory use than for the production plant. Long sample lines are usually required as well
as a host of auxiliary materials such as calibration gases, etc. A further deficiency exists in that
present continuous monitoring equipment is incapable of continuous operation over long periods
without regular attention by highly skilled instrument specialists.
C. R&D Priorities to Improve Measurement Techniques
Based on the deficiencies which exist in measurement of emissions in the paint and
varnish industry, only one R&D program is suggested. This program is to develop a simple
and inexpensive instrument to detect hydrocarbon emission and measure flow from kettle cooking
operations. The proposed program is outlined in the next report section.
432
-------
D. Recommended Programs for Achieving R&D Requirements
An ideal monitoring instrument for kettle cooking operations would simultaneously
measure the gas flow and the concentration of total organic materials in the gas stream.
In this manner, a weight rate of emission of organics could be deduced, or even generated
within the instrument and recorded. This instrument would be required to run continuously
over long periods of time without the necessity for highly skilled instrument specialists. It
would either operate with the detection element deployed directly in the gas stream and without
inter-connecting sampling tube, etc., or with the entire sampling train operated at a temperature
high enough to prevent condensation of organics. It is, of course, imperative that the instrument
pose no safety hazard.
The characteristics of this instrument are as follows:
1. Organic concentration and gas flow rate must be measured simultaneously.
2. Organic concentration must be measured in terms of either total hydrocarbon or
equivalent methane.
3. Continuous operation must be possible on heavy organic fumes from kettle operations.
4. Maintenance or operation requirements must be simple.
5. First cost must be inexpensive (less than $5,000).
In addition to handling organic emissions, the proposed instrument would need to
incorporate total flow measurement capability with low energy loss or pressure drop. Primary
flow elements and associated connection lines will need to be designed to operate at high
temperature and without interference from deposits of heavy resinous materials.
With a low pressure drop limitation, the selection of the primary flow measurement
device should be limited to the following:
1. Pilot tube (S type)
2. Venturi flowmeter
3. Fluidic sensor
The final selection of device and means for eliminating interference from organic
deposition would be a part of the proposed development program. In addition, the program
would include selection of a read-out approach which would be most compatible, least
expensive and require the lowest maintenance.
433
-------
The proposed development would be carried in three phases as follows:
1. Library and field research.
2. "Bread board" and flow measurement experimentation.
3. Design and fabrication of a working prototype.
The suggested content of each phase is as follows:
Phase I
1. Literature survey to define the present state of the art for flame ionization detection
as well as pitot or venturi flow measurement.
2. Discussions with vendors to establish the range of commercially available pitot and
venturi measuring devices.
3. Definition of alternatives with respect to:
a. Detector configuration
b. Electronic circuitry
c. Hydrogen generation
d. Over-all packaging
e. Primary flow measurement
f. Sample-loop consideration
Phase II
1. Assembly of test set-up.
2. Preparation of alternative "bread board" detectors.
3. Selection of primary flow devices.
4. Testing modification of bread board units and alternative flow devices.
5. Selection of mechanisms applicable to prototype design.
Phase III
1. Prototype design
2. Component acquisition
3. Fabrication of components
4. Assembly of components
5. Field trials
6. Revision of prototype as required
7. Final testing
434
-------
The timing and costs for each of the three phases are indicated below:
Phase I should require approximately 400 man-hours of engineering or scientist effort
and cost approximately $10,000.
Phase II is expected to require 1500 man-hours of engineer or scientist effort and 1000
man-hours of technician effort. This phase of the program should cost approximately $45,000.
In addition, equipment costs of $5,000 would be anticipated.
Phase III should require approximately 1500 man-hours of engineer or scientist effort
and 2000 man-hours of drafting and technician effort which will cost about $65,000. In addition,
about $8,000 will be required to purchase components. It is anticipated that installation and
operation of the prototype unit can be arranged with a paint manufacturer and that it will be an
advantage for him to bear the installation and operation expense.
III. PIGMENT INDUSTRY
The manufacture of pigments is nearly as complex as the manufacture of the coatings.
However, it is easy to divide this industry into three components, which are:
1. The manufacture of titanium dioxide by the sulfate process.
2. Manufacture of titanium dioxide by the chloride process.
3. Manufacture of other pigments by various processes.
While TiO2 production by the chloride process probably exceeds that by the sulfate process
today, the sulfate process will probably remain a significant factor for some time. Also, the
problems inherent in the preparation of titanium dioxide pigments by the sulfate process appear
to be more significant and difficult to control than those in the chloride process and development
efforts should be concentrated in this area.
Unlike the sulfate process, there are a number of different chloride processes used in
the manufacture of TiO2. The large differences in emissions from the various type plant chlorinators
should also be a subject of further study.
The R&D program in this industry should consist of the following steps:
1. Assemble all of the available information on the design basis used for the selection
and sizing of existing electrostatic precipitators and scrubbers by contact with the
manufacturers of titanium dioxide by the sulfate process, and by contact with the
435
-------
manufacturers of air pollution control equipment.
2. Independent definition of the problem in terms of the physics and chemistry of the
fume product and source test information established by the EPA.
3. Preparation of prototype specifications for a satisfactory control system by each means
judged as a feasible approach, without regard to commercial duration of the process.
4. Economic evaluation of the alternatives judged feasible.
5. Selection of one (or possibly two) technically feasible alternatives for prototype
development.
6. Small scale pilot testing of a prototype design.
7. Recommendations for full-scale testing.
The cost of the development program is likely to be large relative to the economic
advantage achieved by any single pigment manufacturer by product recovery or air pollution
control equipment cost reduction.
Estimates of the timing and costs can be divided into three phases. The first involves
the investigatory and evaluation phase. This should require approximately 2000 man-hours of
professional time and 1000 man-hours of non-professional time, and should cost on the order
of $60,000. At the end of this phase, a feasible approach should have been selected and a
pilot design prepared. The second phase involves the construction, field testing and evaluation of
a prototype pilot unit. This is expected to involve an additional 1500 man-hours of professional
time, 1500 man-hours of non-professional time and a lump sum cost for fabrication of the pilot
unit of $25,000. The manpower costs are estimated to be approximately $50,000, for a total
cost of $68,000.
The final phase, which is a preparation for the testing of a full-scale prototype, involves
the design of a full-scale prototype and a good estimate of the installation and operating cost
of the prototype. This phase is estimated to require approximately 1200 professional man-hours,
100 non-professional man-hours and to cost approximately $30,000.
The installation and operation of a full-scale prototype unit may be emitted from the
research and development program proposed, and some cost-sharing arrangement worked out with
a titanium dioxide manufacturer so that the principal costs of installation and operation are borne
by the commercial process. The development costs could reasonably be funded by the public
through the EPA.
436
-------
The second area identified for research and development work is that of treatment of
the flue gas discharged from the pigment digester. An alternative course of action involves the
modification of the digestion process such that little or no gas is vented to the atmosphere.
This has been judged to be less likely of success because of the large investment in existing
sulfate process digesters in the industry. However, should a good approach to process modifications
be proposed within the industry, direct funding of all or a part of the development work would be
appropriate.
437
-------
REFERENCES
1. Marketing Guide to the Paint Industry. Fairfield, N.J., Charles H. Kline & Co., Inc., Patricia
Noble (ed), 1969.
2. Federation Series on Coating Technology.
Unit 1 — "Introduction to Coatings Technology", Oct., 1964.
Unit 2 — "Formation and Structure of Paint Films", June, 1965.
Unit 3 — "Oils for Organic Coatings", Sept., 1965.
Unit 4 — "Modern Varnish Technology", May, 1966.
Unit 5 — "Alkyd Resins", Mar., 1967.
Unit 6 — "Solvents", May, 1967.
Unit 7 — "White Hiding and Extender Pigments", Oct., 1967.
Unit 8 — "Inorganic Color Pigments", Mar., 1968.
Unit 9 — "Organic Color Pigments", July, 1968.
Unit 10 — "Black and Metallic Pigments", Jan., 1969.
Unit 11 — "Paint Driers and Additives", June, 1969.
Unit 12 — "Principles of Formulation and Paint Calculations", June, 1969.
Unit 13 — "Amino Resins in Coatings", Dec., 1969.
Unit 14 — "Silicone Resins for Organic Coatings", Jan., 1970.
Unit 15 — "Urethane Coatings", July, 1970.
Unit 16 — "Dispersion and Grinding", Sept., 1970.
Unit 17 — "Acrylic Resins", Mar., 1971.
Unit 18 — "Phenolic Resins", Mar., 1971.
Unit 19 — "Vinyl Resins", Apr., 1972.
Federation Societies for Paint Technology. Philadelphia, Pa.
3. Chemical Economics Handbook. Stanford Research Institute, Menlo Park, Ca.
4. The Technology of Paints, Varnishes, and Lacquers. New York, Reinhold Book Corp., Charles
R. Martens (ed), 1968.
5. Parker, Dean H. Principles of Surface Coating Technology. New York, John Wiley & Sons,
Interscience Publishers Division, 1965.
6. Raw Materials Index — Resin Section. Washington, D.C., National Paint, Varnish and
Lacquer Association, 1971-72 edition.
7. Paint Red Book. New York, Palmerton Publishing Company, 1972, fifth edition.
8. Current Industrial Reports — Series M28F. Washington, D.C., U.S. Department of Commerce,
Bureau of the Census.
9. Brewster, R. F. Process Layout and Equipment Selection for One Million Gallons a Year
Paint Plant. Presented at Manufacturing Management Forum of 1972 convention of the
National Paint and Coating Association, October 31, 1972.
10. The Ideal Paint Plant. Manufacturing Committee, Toronto Paint Society.
11. The Story of Alkyd Resins-Processing and Equipment. Cincinnati, the Brighton Corporation, 1959.
12. Unpublished, First Interim Report.
439
-------
13. County Business Patterns. Washington, D.C., Department of Commerce, Bureau of the
Census, 1965-71.
14. 1967 Census of Manufacturers. Washington, D.C., U.S. Bureau of the Census, Industry and
Area Statistics.
15. Chemical Engineering. February 19, 1973.
16. Chemical and Engineering News. February 23, 1973.
17. Air Pollution Engineering Manual. Los Angeles County, Air Pollution Control District.
18. Unpublished source tests of 10 different kettles. PPG Industries.
19. Skelly, W. K. Unpublished report on source test of Stresen-Reuter International's plant in
Bensenville, Illinois, conducted by the R&D group of International Minerals & Chemicals
Corporation.
20. Unpublished source tests of 11 different kettles. Montebello, Ca., Hirt Combustion Engineers,
August, 1971.
21. White, H. J. Industrial Electrostatic Precipitation. Reading, Mass., Addison-Wesley Publishing
Company, 1963.
22. Chemical Week. June 20, 1973.
23. Chemical Engineering. February 19, 1973.
24. Modern Plastics. April, 1973.
25. Chemical and Engineering News. February 23, 1973.
26. Chemical Week. April 4, 1973.
27. Rolke, et al. Afterburner Systems Study, Final Report under Contract EHS-D-71-3 for the
U.S. Environmental Protection Agency, Office of Air Programs.
28. Hardison, L. C. A Summary of the Use of Catalyst for Stationary Emission Source Control.
Presented at Franklin Institute, Philadelphia, Pa., Nov., 1968.
29. Hardison, L. C. Controlling Combustible Emissions. Paint and Varnish Production, July, 1967.
30. Paint Technology Manual, Part Six, Pigments, Dyestuffs, and Lakes. Oil and Colour Chemists
Association, New York, Reinhold Publishing Corp., 1966.
31. Raw Materials Usage Survey. National Paint, Varnish and Lacquer Association, June, 1971.
32. Unpublished Data. Harshaw Chemical Company.
33. Nelson, K. W. Cadmium in the Environment. San Francisco, presented at AIME Meeting,
February 21, 1972.
440
-------
34. Shreve, R. N. The Chemical Process Industries. New York, McGraw-Hill Book Company,
1956.
35. Minerals Yearbook 1968. Washington, D.C., Bureau of Mines.
36. Barksdale, J. Titanium — Its Occurrence, Chemistry and Technology. New York, The Ronald
Press Company, 1966.
37. Atmospheric Emissions from Hydrochloric Acid Manufacturing Processes. Durham, N.C.,
U.S. Department of Health, Education and Welfare, Public Health Service, National Air
Pollution Control Administration, 1969. Publication No. AP-54.
38. Falk, L. L. Industrial Hygiene Quarterly. December, 1951. Vol. 12, No. 4.
39. Report of Atmospheric Emission Tests Conducted at Titanium Division, National Lead
Company, Sayreville, New Jersey. Division of Air Pollution, Public Health Service, Sept., 1966.
IND-1-66.
441
-------
LIST OF STANDARD ABBREVIATIONS
actual cubic feet per minute ACFM
atmospheres gage (pressure) atmg
British thermal units Btu
centimeters cm
change of pressure (delta pressure) AP
change of temperature (delta temperature) AT
cubic feet ft3
degrees Centigrade °C
degrees Fahrenheit °F
degrees Rankine °R
diameter dia
dollars $
feet or foot ft
gallon or gallons gal
gallons per minute gpm
grain or grains gr
hour or hours hr
hydrocarbon Hcbn
inch or inches in.
lower explosive limit L.E.L.
meter or meters m
milliliters ml
millimeters mm
millimeters of Mercury (pressure) mmHg
millions (106) MM
minute or minutes min
mole mol
non-volatile material NVM
parts per million ppm
per cent %
pint or pints pt
pound or pounds Ib
pounds per square inch gauge psig
quart or quarts q{
seconds sec
square feet ft2
standard cubic feet SCF
standard cubic feet per minute SCFM
temperature Temp
ton or tons ton
water column (pressure) w-c-
weight percent wt-0/°
year or years Yr
442
-------
APPENDIX A
MODEL PLANT COST DETAIL
-------
APPENDIX A
MODEL PLANT COST DETAIL
Item
Equipment - Paint Plant
101 - Pebble Mill
Abbe - Model 8% - 21" O.D. x 28" LG,
Jacketed Mill with Burrstone lining
complete with 3/4 HP motor §
magnetic brake
Ni-Hard Balls
Electrical Wiring
Foundation § Supports
Installation
Total
102 - Pebble Mill
Abbe - Model 8B - 24" O.D. x 36" LG.
Jacketed Mill with Burrstone lining
complete with 1% HP motor § magnetic
brake
Ni-Hard Balls
Electrical Wiring
Foundation § Supports
Installation
Total
103 - Pebble Mill
Abbe - Model 6 - 32" O.D. x 36" LG.
Jacketed Mill with Burrstone lining
complete with 3 HP motor §
magnetic brake
Ni-Hard Balls
Electrical Wiring
Foundation § Supports
Installation
Total
Quan.
Unit
Cost
3,600
100
1,700
500
400
6,300
4,250
150
1,700
800
500
7,400
5,600
450
1,700
1,100
600
9,450
Total
6,300
7,400
9,450
A-l
-------
Item
Equipment - Paint Plant
104 - Pebble Mill
Abbe - Model 5B - 42" O.D. x 48" LG.
Jacketed Mill with Burrstone lining
complete with 5 HP motor § magnetic
brake
Ni-Hard Balls
Electrical Wiring
Foundation § Supports
Installation
Quan
Total
105 - Pebble Mill
Abbe - Model 2 - 60" O.D. x 48" LG.
Jacketed Mill with Burrstone lining
complete with 15 HP motor § magnetic
brake
Ni-Hard Balls
Electrical Wiring
Foundation § Supports
Installation
Total
106 - Pebble Mill
Abbe - Model 2B - 60" O.D. x 72" LG.
Jacketed Mill with Burrstone lining
complete with 20 HP motor § magnetic
brake
Ni-Hard Balls
Electrical Wiring
Foundation § Supports
Installation
Total
Unit
Cost
7,350
550
1,700
2,200
800
12,600
11,200
1,100
1,800
4,700
1,500
20,100
12,900
1,650
1,800
5,700
1,500
23,550
Total
12,600
20,100
23,550
A-2
-------
Unit
Item Quan. Cost Total
Equipment - Paint Plant
107 - Pebble Mill
Abbe - Model 0 - 72" O.D. x 96" LG. 1 19,700
Jacketed Mill with Burrstone lining
complete with 30 HP motor § magnetic
brake
Ni-Hard Balls 3,200
Electrical Wiring 2,150
Foundation § Supports 10,800
Installation 2,300
Total 38,150 38,150
Total unadjusted cost for Pebble Mills 117,550
Chicago area adjustment 41
Adjustment from Jan. to Dec. '72 5%
Total 9% 10,600
Total installed cost for Pebble Mills 128,150
A-3
-------
Item
Quan.
Unit
Cost
Total
Equipment - Paint Plant
108 - Ball Mill
Abbe - Model 6 - 32" O.D. x 36" LG.
Jacketed Mill with Burrstone lining
complete with 5 HP motor and magnetic
brake
Ni-Hard Balls
Electrical Wiring
Foundation § Supports
Installation
Total
109 - Ball Mill
Epworth - 4'x5' Ball Mill manufac-
tured in accordance with Sherwin-
Williams specifications complete
with 15 HP motor reducer drive
Ni-Hard Balls
Electrical Wiring
Foundation § Supports
Installation
Total
6,000
650
1,700
1,500
700
10,550
13,150
2,500
1,800
4,700
1,700
23,850
10,550
23,850
Total unadjusted cost for Ball Mills
Adjustment - 9%
Total installed cost for Ball Mills
34,400
3,100
37,500
A-4
-------
Unit
Item Quan. Cost Total
Equipment - Paint Plant
110 - Sand Mill
Standard "Red Head" Mill Model 3P 1 3,350
complete with 10 HP motor, feed
pump and two charges of sand
Electrical Wiring 1,800
Foundations § Supports 300
Installation 500
Total 5,950 5,950
111 - Sand Mill
Standard "Red Head" Mill Model 8P 1 5,100
complete with 20 HP motor, feed
pump, discharge pump and two charges
of sand
Electrical Wiring 1,800
Foundations § Supports 600
Installation 700
Total 8,200 8,200
112 - Sand Mill
Standard "Red Head" Mill Model 16P 1 6,300
complete with 30 HP motor, feed pump,
discharge pump and two charges of
sand
Electrical Wiring 2,150
Foundations § Supports 900
Installation 800
Total 10,150 10,150
Total unadjusted cost for Sand Mills 24,300
Total adjusted cost for Sand Mills 26,500
A-5
-------
Unit
Item Quan. Cost Total
Equipment - Paint Plant
113 - High Speed Disperser
Morehouse Cowles - Model 515 VHV - 2 3,300
Disperser complete with 10 HP
explosion-proof motor
Electrical Wiring 1,800
Installation 400
Total 5,500 11,000
114 - High Speed Disperser
Morehouse Cowles - Model 520 VHV - 1 4,550
Disperser complete with 25 HP
explosion-proof motor
Electrical Wiring 2,000
Installation 500
Total 7,050 7,050
115 - High Speed Disperser
Morehouse Cowles - Model 720 VHV - 1 7,050
Disperser complete with 50 HP
explosion-proof motor
Electrical Wiring 2,500
Installation 800
Total 10,350 10,350
116 - High Speed Disperser
Morehouse Cowles - Model 1030 VHV - 1 8,400
Two Speed Disperser complete with
100 HP explosion-proof motor
Electrical Wiring 3,300
Installation 1,000
Total 12,700 12,700
Total unadjusted costs for High Speed Dispersers 41,100
Adjustment 3,700
Total installed costs for High Speed Dispersers 44,800
A-6
-------
Unit
Item Quan. Cost Total
Equipment - Paint Plant
117 - Mixing Tanks
Portable tank, 220 gallon capacity, 20 350 7,000
42" I.D. x 34" straight side with
flat bottom mounted on 6" heavy duty
casters; lift hooks with 2" half
coupling outlet
Carbon steel construction
118 - Finishing Tanks
Floor mounted 550 gallon tank, 14 2,900
4' I.D. x 6' straight side with
dished bottom and flat top complete
with 3 HP top mounted agitator
carbon steel
Electrical Wiring 1,700
Installation 300
Total 4,900 68,600
119 - Finishing Tanks
Floor mounted 1100 gallon tank, 8 4,500
5' I.D. x 8' straight side with
dished bottom and flat top complete
with 7% HP top mounted agitator
Carbon steel construction
Electrical Wiring 1,700
Installation 400
Total 6,600 52,800
120 - Finishing Tanks
Floor mounted 2200 gallon tank, 3 5,700
7' I.D. x 8' straight side with
dished bottom and flat top complete
with 15 HP top mounted agitator
Carbon steel construction
Electrical Wiring 1,800
Installation 600
Total 8,100 24,300
Total unadjusted cost for Mixing and Finishing Tanks 152,700
Adjustment 13,750
Total installed costs for Mixing and Finishing Tanks 166,450
A-7
-------
Unit
Item Quan. Cost Total
Equipment - Paint Plant
121 - Filter
Cuno - Model - FH-530-CW with 1 200
support assembly. Carbon
steel construction
Installation 50
Total 250 250
122 - Filter
Cuno Micro Klean Filter with 6 3 655
cartridges. Carbon steel con-
struction
Installation 75
Total 730 2,190
123 - Screens
Lehman Vorti Sieve 2 3,000
Support Cart 500
Electrical Wiring 1,700
Installation 500
Total 5,500 11,000
124 - Can Filler
Ambrose Model PF-9 Filling and 2 10,000
sealing machine
Lid Dropper § Can Closer 1,500
Installation 1,000
Total 12,500 25,000
125 - Bail-0-Matic
Bail-0-Matic - Model "A" 1 16,000
Electrical Wiring 1,700
Installation 1,000
Total 18,700 18,700
126 - Barrel Filler
Mantes - Model 16 MTFF 1 4,500
Barrel Filler
Installation 400
Total 4,900 4,900
A-8
-------
Item
Quan.
Unit
Cost
Total
Equipment - Paint Plant
127 - Labeler
New .Way Labeler - Model
Electrical Wiring
Installation
128 - Packing Station
New Way Model EC Caser
Electrical Wiring
Installation
129 - Case Sealer
New Way Model F-20 Kiwi
Kiwi Carton Coder
Electrical Wiring
Installation
iEpn
Total
Total
Total
130 - Carton Stencil
131 - Portable Pumps and Carts
Total. Filling and Packaging Equipment
Adjusted Cost
Adjustment
Total Installed Cost for Filling and
Packaging Equipment
6,000
1,700
500
8,200
9,550
100
1,000
10,600
8,200
10,600
1
1
2
7,000
300
1,700
1,000
10,000
500
2,500
10,000
500
5,000
96,340
8,660
105,000
A-9
-------
Unit
Item Quan. Cost Total
Equipment - Paint Plant
Laboratory Equipment
Viscometers 4 250 1,000
Spray Booth - Binks Model PBFL-4
Electrical Wiring 2 475
Installation 4 1,700
Installation 125
Total 2,300 4,600
Drying Oven - "Blue M" Model 2 550
OV-472A-2 with floor stand
Electrical Wiring 1,700
Installation 150
Total 2,400 4,800
Tables, Benches § Disks 2,000
Total Unadjusted Cost for
Laboratory Equipment 12,400
Adjustment 1,100
Total Installed Cost for Laboratory Equipment 13,500
A-10
-------
Item
Equipment - Resin Plant
201 - 500 Gallon Resin Reactor Unit
Complete with:
Agitator $ Baffles
Cooling Coil
Baffles
Fume Scrubber
Spray Tower
Ej ector
Circulating Pump
1000 Gallon Thinning Tank
Thinning Tank Agitator
Thinning Tank Scale
Thinning Tank Condenser
Control Panel
Material of Construction
Stainless Steel
Quotation from
Brighton Corporation
Cincinnati, Ohio
Electrical Wiring
Piping - included in the above
quotation
Installation
Structural Steel
Insulation
Unit
Quan. Cost
Total
55,300
5,000
11,000
3,000
1,800
Total
76,100
76,100
A-ll
-------
Item
Equipment - Resin Plant
202 - 1500 Gallon Resin Reactor Unit
Complete with:
Agitator
Inconel Dimpled Jacket
Sampler
Baffles
Cooling Coil
Partial Condenser
Total Condenser
Decanter-Receiver
Vacuum System
Fume Scrubber
Spray Tower
Ej ector
Circulating Pump
3000 Gallon Thinning Tank
Thinning Tank Agitator
Thinning Tank Scale
Thinning Tank Condenser
Control Panel
Unit
Quan. Cost Total
85,400
Material of Construction
Stainless steel unless specified
otherwise
Quotation from Brighton Corp.
Cincinnati, Ohio
Electrical Wiring
Piping - included in the above
quotation
Installation
Total
8,500
17,000
110,900 110,900
A-12
-------
Item
Equipment - Resin Plant
203 - Filter Feed Pumps
Viking Model K-124 Cast Iron
Electrical Wiring
Foundations § Supports
Installation
Total
204 - Resin Transfer Pump
Viking Model K-124 Cast Iron
Electrical Wiring
Foundations § Supports
Installation
Total
205 - Filter Presses
Shriver - 18" x 18" Filter Press
with 70 £t2 of filtering area
Cast iron with "Hydrd Kloser"
Installation
Total
206 - Relief Tank
1500 Gallon
Foundations
Installation
Carbon Steel
Supports
Total
Quan,
Total Unadjusted Cost for Resin Plant
Total Adjusted Cost for Resin Plant
Unit
Cost
600
1,700
200
100
2,600
600
1,700
200
100
2,600
2,000
200
2,200
1,400
600
500
2,300
Total
5,200
5,200
4,400
2,300
204,100
222,500
A-13
-------
Unit
Item Qua a. Cost Total
Equipment - Storage Area
301 - Odorless Mineral Spirits Stg. Tank
25,000 Gallons - Carbon Steel 1 11,200
Installation 1,500
Total 12,500 12,500
302 - Odorless Mineral Spirits Transfer
Pump
Goulds - 1 x 2 x 10 - Carbon Steel 1 700
Pump complete with 3 HP motor
Electrical Wiring 1,700
Installation 100
Piping 3,250
Tokheim - IV Line Meter 750
Total 6,500 6,500
303 - Industrial Alkyd Solvent Stg. Tank
25,000 Gallons - Carbon Steel 1 12,500
304 - Industrial Alkyd Solvent Transfer Pump
Goulds - 1 x 2 x 10 - Carbon Steel 1 700
Pump complete with 3 HP motor
Electrical Wiring 1,700
Installation 100
Piping 3,250
Tokheim - IV Line Meter 750
Total 6,500 6,500
305 - Waste Solvent Tank
1,000 Gallons - Carbon Steel 2 1,200
Installation 500
Total 1,500 3,000
Pump - Goulds - 1 x 2 x 10 - Carbon 1 700
Steel with 3 HP motor
Installation 100
Electrical Wiring 1,700
Piping 2,000
Total 4,500 4,500
A-14
-------
Unit
Item Quan. Cost Total
Equipment - Storage Area
306 - Aqueous Waste Tank
1,000 Gallons - Carbon Steel 1 1,200
Installation 300
Pump - Goulds -1x2x10 1 800
Electrical Wiring 1,700
Piping 600
Total 4,500 4,500
307 - Industrial Alkyd Tank
1,000 Gallons - Carbon Steel 4 5,250 21,000
Tank with Pump £ Meter
308 - Interior Latex Tank
22,500 Gallons - Stainless Steel 1 15,000
Installation 1,000
Pump - Viking Model K-124 1 1,700
Stainless Steel with Motor
Installation 100
Tokheim Meter - IV Stainless Steel 2,500
Electrical Wiring 1,700
Piping - 304 Stainless Steel 1,100
Total 23,100 23,100
309 - Exterior Latex Tank
17,500 Gallons - Stainless Steel 1 13,500
Installation 800
Pump - Viking Model K-124 1 1,700
Stainless Steel complete with
3 HP Motor
Installation 100
Tokheim Meter - IV Stainless Steel 2,500
Electrical Wiring 1,700
Piping - 304 Stainless Steel 1,100
Total 21,400 21,400
A-15
-------
Unit
Item Quan. Cost Total
Equipment - Storage Area
310 - Oil Storage Tank
12,500 Gallons - Carbon Steel 2 6,300
Installation 700
Pump - Viking Model K-124 550
Carbon Steel complete
with 3 HP motor
Installation 100
Tokheim Meter - IV Carbon Steel 750
Electrical Wiring 1,700
Piping - Carbon Steel 600
Total 10,700 21,400
311 - Industrial Alkyd Storage Tank
10,000 Gallons - Carbon Steel 1 5,500
Installation 600
Pump - Viking Model K-124 550
Installation 100
Tokheim Meter - IV Carbon Steel 750
Electrical Wiring 1,700
Piping - Carbon Steel 600
Total 9,800 9,800
312 - Industrial Alkyd Storage Tank
7,500 Gallons - Carbon Steel 1 4,600
Installation 500
Pump - Viking Model K-124 650
Tokheim Meter - iy Carbon Steel 750
Electrical Wiring 1,700
Piping - Carbon Steel 600
Total 8,800 8,800
A-16
-------
Unit
Item Quan. Cost Total
Equipment - Storage Area
313 - Trade Alkyd Tank
10,000 Gallons - Carbon Steel 1 9,800 9,800
314 - Trade Alkyd Tank
5,000 Gallons - Carbon Steel 1 3,400
Installation 400
Pump - Viking Model K-124 650
Tokheim Meter - IV Carbon Steel 750
Electrical Wiring 1,700
Piping 600
Total 7,500 7,500
315 - Glycerine Storage Tank
7,500 Gallons - Carbon Steel 1 8,800 8,800
Installation, Pump, Piping,
Wiring and Meter
316 - Portable Pumps with Carts 3 2,500 7,500
Tank Gauges 20 250 5,000
Total Storage Area Unadjusted Cost 194,100
Adjustment 17,500
Total Storage Area Cost 211,600
A-17
-------
Item
Equipment - Miscellaneous
Scales - 0-100 Ibs.
0-500 Ibs.
Ford Trucks
Electric - 4,000 Ib. Capacity
Charger
Total
I.C. Engine - 4,000 Ib. Capacity
Unit
Quan. Cost
3 200
2 250
1 9,350
250
9,600
2 7,300
Total
600
500
9,600
14,600
Racks
4' wide - 2' deep 6,250 ft2 9,400
Fire Extinguishers 2,200
Emergency Lighting 22,500
Furniture
Office 5,000
Lunch Room 2,500
Locker Room 1,500
Total 9,000 9,000
Drum and Hand Pallet Trucks 1,500 1,500
A-18
-------
Unit
Item Quan. Cost Total
Equipment - Storage Area
313 - Trade Alkyd Tank
10,000 Gallons - Carbon Steel 1 9,800 9,800
314 - Trade Alkyd Tank
5,000 Gallons - Carbon Steel l 3,400
Installation 400
Pump - Viking Model K-124 650
Tokheim Meter - i%" Carbon Steel 750
Electrical Wiring 1,700
Piping 600
Total 7,500 7,500
315 - Glycerine Storage Tank
7,500 Gallons - Carbon Steel 1 8,800 8,800
Installation, Pump, Piping,
Wiring and Meter
316 - Portable Pumps with Carts 3 2,500 7,500
Tank Gauges 20 250 5,000
Total Storage Area Unadjusted Cost 194,100
Adjustment 17,500
Total Storage Area Cost 211,600
A-17
-------
Item
Equipment - Miscellaneous
Scales - 0-100 Ibs.
0-500 Ibs.
Ford Trucks
Electric - 4,000 Ib. Capacity
Charger
Total
I.C. Engine - 4,000 Ib. Capacity
Unit
Quan. Cost
3 200
2 250
1 9,350
250
9,600
2 7,300
Total
600
500
9,600
14,600
Racks
4' wide - 2' deep 6,250 ft2 9,400
Fire Extinguishers 2,200
Emergency Lighting 22,500
Furniture
Office 5,000
Lunch Room 2,500
Locker Room 1,500
Total 9,000 9,000
Drum and Hand Pallet Trucks 1,500 1,500
A-li
-------
Unit
Item Quan. Cost Total
Equipment - Miscellaneous
Shop Equipment
Cabinets § Shelves 3,000
Grinders 1,900
Press 3,000
Radial Drill 12,000
Wells Saw 1,800
Milling Machine 3,600
Lathe 25,000
Benches 850
Vises 700
Welding Machines 900
Band Saw 2,100
Hand Brake 1,800
Roller 2,200
Shearer 1,200
Hand Tools 1,000
Total 61,050 61,050
Total Unadjusted Cost for
Miscellaneous Equipment 130,950
Adjustment 11,800
Total Cost for Miscellaneous Equipment 142,750
A-19
-------
Unit
Item Quan. Cost Total
Utilities
Dowtherm System
Two Million Btu/hr Direct Fired 1 36,000
Heater - Carbon Steel
Piping 5,500
Insulation 2,800
Electrical Wiring 2,500
Total 46,800 46,800
Inert Gas Generator
Kemp - Model 6 - PH - 5000 CFH 1 23,100
Inert Gas System
Storage Tank 1,200
Installation 2,500
Electrical Wiring 2,150
Piping 4,600
Total 33,550 33,550
Compressed Air
150 CFM Packaged Air Compressor 1 3,500
complete with 50 HP Motor
Electrical 2,500
Installation 1,000
Piping 5,200
Total 12,200 12,200
Cooling Water
11,000
7,000
2,520
1,950
1,000
975
600
275
Sub-Total 25,320
Insulation 500
6"
4"
3"
2"
iy
1"
3/4"
k"
Sch.
Sch.
Sch.
Sch.
Sch.
Sch.
Sch.
Sch.
40
40
40
40
40
40
40
40
Carbon
Carbon
Carbon
Carbon
Carbon
Carbon
Carbon
Carbon
Steel
Steel
Steel
Steel
Steel
Steel
Steel
Steel
Total 25,820 25,820
A-20
-------
Unit
Item Quan. Cost Total
Utilities
Portable Water
Piping 950
Hot Water Heaters 800
Electrical Wiring 3,400
Total 5,150 5,150
Sewers § Drains
Process - Piping 7,150
Catch Basins 2,000
Intercepter 1,500
Sanitary - Piping 6,650
Catch Basins 4,000
Total 21,300 21,300
Natural Gas
Piping 7,350
Grounding
Allowance 10,000
Telephones 1,500
Total Unadjusted Cost for Utilities 163,670
Adjustment 14,730
Total Cost for Utilities 178,400
Building Summary
Sitework 167,100
Part I - Raw Material Warehouse 219,700
Part II - Manufacturing Arch 912,450
Part III - Finished Goods Warehouse 526,700
Total 1,825,950 1,825,950
A-21
-------
SITE WORK
Unit Total
Item Quan. Cost Material
Strip Top Soil (8") 2,080 C.Y. .65 1,350
Excavation 3,560 C.Y. 1.36 4,850
Compacted Fill 5,640 C.Y. 8.90 50,100
Asphalt Paving, Base Course, Grading 6,200 S.Y. 4.20 26,000
Fence (8' Chain Link, w/barb wire) 620 L.F. 5.00 3,100
Utility Building 2,000 S.F. 8.00 16,000
Drum Storage Pad (6" cone.) 9,000 S.F. .80 7,200
Tank Farm Pad (6" cone.) 2,400 S.F. .80 1,900
Tank Farm Dike 200 L.F. 37.00 7,400
Seeding 48,000 S.F. .05 2,400
Landscaping L.S. 2,000 2,000
Exterior Lighting 6 400 2,400
124,700
Contractors - Overhead § Profit - 25%
Chicago Area Adjustment - 4%
Adjustment from Jan. to Dec. '72 5%
34% 42,400
Total 167,100
A-22
-------
PART I - RAW MATERIAL WAREHOUSE
Total
Material
850
8,100
2,050
13,600
13,800
32,000
21,200
1,950
750
1,400
5,300
1,300
5,850
6,600
3,600
3,000
1,000
900
450
2,800
400
450
12,800
15,300
8,500
163,950
Contractors - Overhead $ Profit - 25%
Chicago Area Adjustment - 4%
Adjustment from Jan. to Dec.'72 51
341 55,750
Total 219,700
Item
Excavation
Foundations
Compacted Fill
First Floor (6" cone.)
Roof Deck, Insulation § Roofing
Structural Steel
Insulated Sandwich Panel
Partition Walls (8" Block)
Gravel Stop
Roof Drains § Piping
Smoke Vents
Doors (31 x 7')
Doors (81 x 8') w/operator 510 + 465
Dock Levelers
Dock Seals
Dock Canopy
Dock Stairway
Ceilings
V.A. Tile
Toilet Fixtures
Toilet Partitions § Screens
Paint
Heat § Vent
Lighting
Sprinklers ford, hazard)
Quan.
630 C.Y.
90 C.Y.
230 C.Y.
17,000 S.F.
17,000 S.F.
145,000 #
9,000 S.F.
1,100 S.F.
570 L.F.
2 ea.
6 ea.
5 ea.
6 ea.
6 ea.
6 ea.
1,000 S.F.
1 ea.
900 S.F.
900 S.F.
7 ea.
4 ea.
2,200 S.F.
17,000 S.F.
17,000 S.F.
17,000 S.F.
Unit
Cost
1.36
90
8.90
.80
.81
.22
2.35
1.75
1.30
700
880
265
975
1,100
600
3.00
1,000
1.00
.50
400
100
.20
.75
.90
.50
A-23
-------
PART II - MANUFACTURING AREA
Total
Material
1,500
25,200
7,300
17,600
44,000
36,300
10,100
131,000
11,700
10,700
2,000
500
6,200
1,400
12,300
2,650
4,900
109,500
10,600
92,500
70,500
33,000
39,500
680,950
Contractors - Overhead § Profit -25%
Chicago Area Adjustment - 4%
Adjustment from Jan. to Dec. '72 5%
341 231,500
Total 912,450
Item
Excavation
Foundations
Compacted Fill
First Floor (6" cone.)
Second Floor (6" cone.)
Roof (8" Flexicore)
Roof Insulation § Roofing
Structural Steel
Walls (8" Block § 4" Brick)
Walls (8" Cone. Block)
Coping
Flashing
Windows
Roof Drains § Piping
Smoke Vents (4 ' x 8')
Doors (31 x 7')
Doors (8f x 8') w/operator
Fire Proof Structural Steel
Paint Walls
Heat and Vent
Lighting
Sprinklers (call system)
Walls (12" Cone. Block)
Quan.
1,100 C.Y.
280 C.Y.
820 C.Y.
22,000 S.F.
22,000 S.F.
22,000 S.F.
22,000 S.F.
570,000 #
3,600 S.F.
6,100 S.F.
620 L.F.
620 L.F.
880 S.F.
2 ea.
14 ea.
10 ea.
5 ea.
52,000 S.F.
53,000 S.F.
44,000 S.F.
44,000 S.F.
44,000 S.F.
16,800 S.F.
Unit
Cost
1.36
90
8.90
.80
2.00
1.65
.46
.23
3.25
1.75
3.25
.80
7.00
700
880
265
975
2.10
.20
2.10
1.60
.75
2.35
A-24
-------
PART III - FINISHED GOODS WAREHOUSE
Item
Excavation
Foundations
Compacted Fill
First Floor (6" cone.)
Roof Deck, Insulation § Roofing
Structural Steel
Insulated Sandwich Panel
Partition Walls
Gravel Stop § Fascia
Roof Drains
Smoke Vents
Doors (3! x 7')
Doors (8* x 8') w/operator
Dock Levelers
Dock Seals
Dock Canopy
Dock Stairway
Ceilings
V.A. Tile
Toilet Fixtures
Toilet Partitions § Screens
Lockers
Heat § Vent
Lighting
Sprinklers (Calc. Syst.)
Quan.
580 C.Y.
80 C.Y.
500 C.Y.
45,000 S.F.
45,000 S.F.
382,000 Ib.
15,600 S.F.
4,800 S.F.
850 L.F.
4 ea.
14 ea.
11 ea.
9 ea.
9 ea.
9 ea.
1,500 S.F.
1 ea.
6,800 S.F.
6,800 S.F.
20 ea.
9 ea.
70 ea.
45,000 S.F.
45,000 S.F.
45,000 S.F.
Unit
Cost
1.36
90
8.90
.80
.81
.22
2.35
1.75
1.30
700
880
265
975
1,100
600
3.00
1,000
1.00
.50
400
100
40
.75
.90
.75
Total
Material
800
7,200
4,450
36,000
36,400
84,000
36,700
8,400
1,100
2,800
12,300
2,900
8,800
9,900
5,400
4,500
1,000
6,800
3,400
8,000
900
2,800
34,000
40,500
34,000
393,050
Contractors - Overhead § Profit - 25%
Chicago Area Adjustment - 4%
Adjustment from Jan. to Dec. '72 - 5%
34% 153,650
Total 526,700
A-25
-------
A-26
-------
APPENDIX B
PROCESS NARRATIVE OF PAINT AND VARNISH INDUSTRY
-------
APPENDIX B
PROCESS NARRATIVE OF PAINT AND VARNISH INDUSTRY
The Paint and Varnish industry is one of the oldest manufacturing industries in the United
States. The industry is made up of about 1,600 companies operating 1,875 plants.'1' It is well
distributed geographically throughout the country and the number of plants or production
volume is definitely related to density of population. Even though about 27 companies account
for about 57% of the total sales, the industry is one of the few remaining which contains
numerous small companies that specialize in a limited product line to be marketed within a
geographical region. There are fewer than 20 companies that sell paint nationwide.
The industry is now emerging as a scientific business from its beginning as an art 50 years
ago. Even with rapid growth in technology, the industry processing techniques still are not well
defined and vary from one producer to another. To add further complication, the industry is
technically one of the most complex of the chemical industry. A plant that produces a broad
line of products might utilize over 600 different raw materials and purchased intermediates.
These materials can be generally classified in the following categories: oils, metallic driers,
resins, pigment extenders, plasticizers, solvents, dyes, bleaching agents, organic monomers for
resins and additives of many kinds.
The industry produces an equally large number of finished products which are generally
classified as trade sale finishes, maintenance finishes and industrial finishes.
Trade sale products are stock-type paints generally distributed through wholesale-retail
channels and packaged in sizes ranging from 1/2 pint to 1 gallon. A subdivision of trade sale
products are maintenance finishes which are used for the protection and upkeep of factories,
buildings and structures such as bridges and storage tanks. Since they are usually stock type,
they come under the Department of Commerce definition of trade sales.
The other major type of paint products are industrial finishes which are generally defined
as those applied to manufactured products. These finishes such as automotive, aircraft, furniture
and electrical are usually specifically formulated for the using industry. Within these major
product lines there are literally thousands of different products for many different applications
and types of customers. Trade sales finishes and industrial finishes are produced in almost equal
volume with the production for this year estimated at 475 million gallons for each type. Trade
sales, however, are expected to account for 55% of the dollar sales or about $1,685 million
dollars.
B-1
-------
PROCESS DESCRIPTION
Starting with all purchased raw material, the manufacturing process for pigmented
products is deceptively simple from a process viewpoint. Basically, it consists of mixing or
dispersing pigment and vehicle to give the final product. This is schematically illustrated in
Figure 1.
The paint vehicle is defined as the liquid portion of the paint and consists of volatile
solvent and non-volatile binder such as oils and resins. The non-volatile portion is also called the
vehicle solid or film former. The pigment portion of the paint consists of hiding pigments such
as titanium dioxide (Ti02), extenders or fillers such as talc or barium sulfate, and any mineral
matter used for flatting or other purposes.
The incorporation of the pigment in the paint vehicle is accomplished by a combination of
grinding and dispersion or dispersion alone. When it is necessary to further grind the raw
pigment, the pebble or steel ball mills are normally used. With the advent of fine particle grades
of pigment and extenders, as well as the wide spread use of wetting agents, the trend is toward
milling methods that are based on dispersion without grinding. This dispersion consists of
breakup of the pigment clusters and agglomerates, followed by wetting of the individual
particles with the binder or vehicle. Some of the more popular methods currently being used are
high speed disc impellers, high speed impingement mills and the sand mill.
Aside from this dispersion step, pigment paint manufacturing involves handling of raw
material as well as handling and packaging of finished product. Operations of a typical plant
may be summarized as a raw material and finished product handling problem with a variety of
interdispersed batch operations. The inter-relationship of all these operations is schematically
illustrated in Figure 2. The operations depicted are those of a plant that makes its own resins
and produces both trade sale and industrial finishes.
Many of the larger and some of the medium size manufacturers produce a significant
amount of their formulation ingredients, including pigments, resins, modified oils and basic
chemicals. Certain manufacturers produce these ingredients in an amount exceeding their
requirements and sell the excess to other manufacturers. A significant number also produce only
a portion of their resins and purchase the remainder from their competitors or suppliers who
specialize in resin manufacturing.
The manufacturing of resins and varnishes is by far the most complex process in a paint
plant, primarily as the result of the large variety of different raw materials, products and
cooking formulas utilized. The complexity begins with the nomenclature used in classification
of the final product. Originally, varnishes were all made from naturally occurring material and
they were easily defined as a homogeneous solution of drying oils and resins in organic solvents.
As new synthetic resins were developed, the resulting varnishes were classified as resins rather
than varnishes. Examples of this are alkyd, epoxy and polyurethane resin varnishes.
There are two basic types of varnishes, spirit varnishes and oleoresinous varnishes. Spirit
B-2
-------
LJ
Z
J
LJ
or
I
e)
z
Q
N
id
K
1
X
2
X
I *
UJ<
tt *-
a.
0
Ztf
(Old
DO.
0°
zo
&I
Ili
5 ce
(5 ^
^
-------
O
*
lli
tt
O
z
QC
D
I-
U
£
Z
K
Z
c^
(T
Q
LJ
CO
O
CO
<
-------
varnishes are formed by dissolving a resin in a solvent. They dry by solvent evaporation. Shellac
is a good example of a spirit varnish. Another material that might fall in this category is lacquer.
Technically, lacquers are defined as a colloidal dispersion or solution of nitro-cellulose, or of
similar film-forming compounds, with resins and plasticizers, in solvent and diluents which dry
primarily by solvent evaporation.
Oleoresinous varnishes, as the name implies, are solutions of both oils and resins. These
varnishes dry by solvent evaporation and by reaction of the non-volatile liquid portion with
oxygen in the air to form a solid film. They are classified as oxygen convertible varnishes and
the film formed on drying is insoluble in the original solvent. A summary of the various types of
material used in the production of classical varnishes is given in Table 1.
Varnish is cooked in both portable kettles and large reactors. Kettles are used only to a
limited extent and primarily by the smaller manufacturers. The very old, coke fired, 30 gallon
capacity copper kettles are no longer used. The varnish kettles which are used, have capacities of
150 to 375 gallons. These are fabricated of stainless steel, have straight sides and are equipped
with three or four-wheel trucks. Heating is done with natural gas or fuel oil for better
temperature control. The kettles are fitted with retractable hoods and exhaust pipes, some of
which may incorporate solvent condensers. Cooling and thinning are normally done in special
rooms. A typical varnish production operation is illustrated in Figure 3.
The manufacturing of oleoresinous varnishes is somewhat more complex than for spirit
varnishes. This manufacture consists of the heating or cooking of oil and resins together for the
purpose of obtaining compatability of resin and oil and solubility of the mixture in solvent, as
well as for development of higher molecular weight molecules or polymers.
The time and temperature of the cook are the operating variables used to develop the
desired end product polymerization or "body". The chemical reactions which occur are not well
defined. The resin is a polymer before cooking and may or may not increase in molecular size
during the cook. This resin may react with the oil to produce copolymers of oil and resin or it
may exist as a homogeneous mixture or solution of oil homopolymers and resin homopolymers.
It is possible to blend resins and heat-bodied oil and obtain the same varnish that can be
produced by cooking the resin and the unbodied oils. This indicates that copolymerization is
not the fundamental reaction in varnish cooking.
Heat bodying or polymerization of an oil is done to increase its viscosity and is carried out
in a kettle in a fashion similar to varnish cooking. The fundamental reaction that occurs is
polymerization of the oil monomers to form dinners with a small portion of trimers.
There is a large variety of synthetic resins produced for use in the manufacture of surface
coatings. A listing of the more popular resins is given below. They are listed by order of
consumption by the coatings industry:(6>
B-5
-------
CO
LJU
>
H
Q
Q
<
en
c
'+-• CU
C CO
< <
.2 fe
1* -Q
S 8
5 S
CD
^o en
CO
QC
UJ
QC
Q
Si
CD
CO if
-Q CO CO
co 5 -Q
cu J2 o
I'i
o 8
"co .!=
SL* _. ^X 4U • —
-I S O O N
CO
Z
LJJ
>
O
co
cc
111
*- ON 00
5
UJ
_J
CO
<
I
cc
I
CO
DC
<
o
1? -* c
o g -i
0 co »
CO
CO
z
CO
LU
DC
CO
CD
o 8 -Q
^ .^ I •§ ?
•i==5
•i I
03 tn JJ
"- cu cu
Q. S -
O o .«
^- I- +-<
O o «n
-P. O co
^ co p
CO CO
"co 8
o
S E
C C 2 °
C E co co
.E-i
8 £ -
£ «
in
° in
CU
e
T3 O
rrc^v^ui-.— t-'Wi—
54 c-D^'35 E O — '«
ytccuo^O-^o
COW3Cm^uCcaJo-^O
5i£.Soo'ES£-s!!D?c3!
CO
c
in
co O
DC
^ «
'Jn .E J
0) c/i "-
QC Qj o
.y y "o
+-• 'cu c
CD — O)
^ co jr
c ^ Q-
CO co
0)
•» .E to
Q. « c
CO O • —
o?^ e
I E S
5 -I N
cu
^ CO ^J
CM
00
CO
O
o
tt
CO
o
T3
cu
4-J
CO
co
X
.Si -o
:11
O Q
ca -D o
in
T3
'o
CO
cu £
1 -2 co "5
9) H- J° ^
C *- •*" —
.h co o •- to
CO CO U- I-
o
CO CO
CN
CJ T3 CU
ie
odi
C
CD
+->
CO
cu
"5 ~
o QJ
11
cu ^
O cd
oo
B-6
-------
"
Ul
IT
0
0
£
(D
z
X
(/)
D
5? Z
IL It
0
OL
\\\\\\\\\\\
\\\\\X\\\N
\\\\\\\\\\
oz
ZO
OP
00)
OH
a:
uj
h
a:
Id
(D
(D
D
a:
u
a)
0
12
zo
?9
1°^
-------
Alkyd Styrene Butadiene
Vinyl Phenolic
Amino Polyester
Epoxy Urethane
Acrylic Silicone
By far the most widely used of these resins are the alkyds and the vinyls. Alkyd
consumption is approximately five times that of the vinyl, which is approximately twice that of
the amino resins. Further discussion will concentrate on alkyd resins.
Alkyd resins comprise a group of synthetic resins which can be described as oil-modified
polyester resins. They are produced from the reaction of polyols or polyhydric alcohol,
polybasic acid and oil or fatty monobasic acid. A listing and discussion of commonly used raw
materials will follow.
1. Oils or fatty acid (2)
Linseed Castor
Soybean Coconut
Safflower Cottonseed
Tall Oil fatty acid Laurie Acid
Tall Oil Pelargonic Acid
Fish Isodecanoic Acid
Tung (minor)
Oiticica (minor)
Dehydrated Castor (minor)
The materials in the first column are oxidizing or drying types. The materials in the second
column are non-oxidizing and yield soft non-drying alkyds which are used primarily as
plasticizers for hard film resins. The acids shown in this column are the only materials that are
strictly synthetic in origin.
2. Polyols
Name Formula Form
H
Ethylene glycol HC - OH Liquid
HC - OH
H
H H H H
Diethylene HO- C- C-0-C-C-OH Liquid
glycol H H H H
B-8
-------
Propylene glycol
Glycerine
CP-95% glycerine
Super-98%
H
HC -
H
H
c -
I
0
H
H
c -
1
H
OH
H
HC
HC
!
HC
H
Pentaerythritol HOCH2
HOCH2
OH
OH
OH
H2COH
Liquid
Liquid
White Solid
H2COH
Glycerol or glycerine was the first polyol used for alkyds. It is also the most widely used
polyol for alkyds.
The second polyol, based on usage, is pentaerythritol (PE), which came into common use
in the 1940's. PE is supplied as "technical grade" material and contains mono, di, tri and
polypentaerythritol. The material consists primarily of the mono form which was illustrated
previously in the list of polyols.
The important distinguishing feature of the various polyols is the number of potentially
reactive hydroxyl groups in the molecule, known as functionality. The glycols with a
functionality of two produce only straight chain polymers and their resins are soft and flexible.
The resultant products are used primarily as plasticizers for hard resins. Glycerine has a
functionality of three and is used primarily in short and medium oil alkyds. Pentaerythritol,
with a functionality of four, cross-links to a greater extent, forming harder polymers. It is ideal
for use in long oil alkyds.
3. Acids and Anhydrides
Name
Phthalic
anhydride
(ortho)
Formula
T
I
HC^
H
/C<*
^
H
s.
X
,c
&^
Form
White solid
B-9
-------
°Y°H
Isophthalic /C*^ White needles
acid (meta) HC'
I
>0
HVC~C"°»
Terephthalic | white crystals
acid (para)
HC'
HC
OH
Maleic White solid
anhydride
HC C
\
0
HC C
The acidic material can be used as an acid or anhydride. The anhydride is formed from two
molecules of acid minus a molecule of water or removal of one molecule of water from a di
acid. It is preferred, since it reacts faster and yields less water for removal from the cook.
For many years, phthalic anhydride (ortho) (PA) was the only polybasic acid used in
substantial proportions in alkyds. It still remains the predominant dibasic acid. PA is produced
from the catalytic oxidation of naphthalene or ortho-xylene.
The chemistry of alkyd resin systems is very complex. So much so that theoretical
considerations offer only a good starting point. Final formulae and variations are developed by
trial and error changes, based on performance requirements and shortcomings of previous
batches.
B-10
-------
Condensation is the reaction basic to all polyester resins, including alkyds. This reaction
follows the elementary equation for esterification as shown below:
,° o
RC + R.iOH^ * RC + HoO
\ \
OH OR!
Acid + Alcohol (Ester -i- Water) >
For Alkyd Resins
PA + Glycerine $=* Ester + H20
The ester monomer formed is very complex and further reacts to form large polymers
called resins. The polymers formed are low in molecular weight by comparison to other resins.
For example alkyd resins have molecular weights ranging from 1,000 to 7,000 while some vinyl
and acrylic resins have average molecular weights in excess of 100,000 and in some cases as high
as 500,000.
The alkyd polymers also react with oil or fatty acid and are generally classified by the
amount of oil or PA used in the formulation, as described below:
% Oil % PA
Short Oil
Medium Oil
Long Oil
Very Long Oil
33 to 45
46 to 55
56 to 70
71 up
35
30 to 35
20 to 30
20
The resulting reactants of the PA, polyol and oil may be represented in part as shown
below.
Phthalic Anhydride (PA) + Glycerine G(OH)3 )
G-PA-G-PA-G-PA-G-PA
0000
H | H
HOOC-PA PA - COOH
B-11
-------
This will then react with the long chain oil monoglyceride or fatty acid (FA) to yield:
HOOC - PA - G - PA - G - PA - G - PA - G - PA - G - PA - G - OH
0
H
O
H
0
H
0
H
F
A
F
A
Short Oil Alkyd
Alkyds can be manufactured directly from a fatty acid, polyol and acid or from the fatty
acid oil, polyol and acid. The second combination (oil, glycerine and PA) produces glyceryl
phthalate which is insoluble in the oil and precipitates. This problem can be overcome by first
converting the oil to a monoglyceride by heating with a polyol in the presence of a catalyst.
This process is called alcoholysis of the oil. The basic reaction is shown below:
C H2 OOCR
CHOOCR +
CH2OOCR
Triglyceride
CH2OH
2CHOH
CH2OH
Glycerine
CH2OH
3CHOH
CH2OOCR
Monoglyceride
This is an ester interchange reaction with no loss of water.
When fatty acid rather than oil is used as the starting material, this is called the "one-stage"
process. In this process, the fatty acid and glycerine are added to the kettle, the agitator is
started and heat is introduced. When the batch reaches 440° F, the PA is slowly added and
cooking continued for another 3 to 4 hours until the desired body and acid number are reached.
If the fusion process is being used, a continuous purge of inert gas is maintained to remove
the water formed in the reaction. This water may also be removed by what is known as the
solvent process. It is similar to the fusion process except that about 10% aromatic solvent
(usually xylene) is added at the start. The vaporized solvent is passed into a condenser. The
condensate then flows to a decant receiver for separation of reaction water. Recovered solvent is
returned to the reactor.
As discussed earlier, when oil is used rather than fatty acid, the alkyds are produced in a
two stage process. In the first stage the monoglyceride is first produced from the linseed oil and
glycerol. Catalyst and oil are added and the alcoholysis of the polyol and oil is carried out
between 450 and 500° F until the desired end point is reached. When the alcoholysis is
completed, any additional polyol needed is added.
B-12
-------
Following this, the required amount of PA and esterification catalyst are slowly added. If
solvent cooking is to be used, the solvent is also added at this time. Cooking then proceeds as
before.
A typical manufacturing formula for a 50% oil-modified glyceryl phthalate alkyd using the
two stage process is given below.(4)
Ib
First stage
Linseed oil 51.3
Glycerol (95%) 12.8
Catalyst, Ca(OH)2 0.026
Second stage
Glycerol (95%) 6.2
Phthalic Anhydride 39.7
Catalyst
Methyl p-Toluene SuIfonate 0.2
110.2
Approx. Loss 10.2
Solids Yield 100.0
Alkyd and other resins are cooked in closed set kettles, more properly called reactors.
They vary in size in commercial production from 500 to 10,000 gallons. A typical reactor
system is shown in Figure 4. They are generally fabricated of Type 304 or 316 stainless steel
with well polished surfaces to assure easy cleaning. Design pressure is usually 50 psig. These
reactors may be heated electrically, direct fired with gas or oil or indirectly heated using a heat
transfer media such as Dowtherm(R). They are also equipped with a manway, sight-glass,
charging and sampling line, condenser system, weigh tanks, temperature measuring devices and
agitator. The manway is used both for charging solid material and for access to the kettle for
cleaning and repair.
The reactor may be equipped with a variety of different condenser systems. The system
shown in Figure 4 includes a packed fractionating column, a reflux condenser and a main
condenser. The condensers are water cooled shell and tube type and may be either horizontally
or vertically inclined. Vapors are processed and condensed on the tube side and drain to a
decant receiver for separation and possible return of solvent to the reactor. A dual function
aspirator Venturi scrubber is often added to the system. It is used to ventilate the kettle during
addition of solid materials and may also remove entrained unreacted or vaporized solids and
liquids from the venting gases.
B-13
-------
-SPRAY TOWER
REFLUX
COMDEIXISER
vEfxi T
F RAC TICXSIAT IMG
DISTILLATIOM
PORTHOLE
FOR SOLIDS
DIRECT FIRED OR
FOR MIQM
TEMPERATURE VAPOR
R LIQUID
TO RESIM
STORAC3E
FIC3URE1
MODEIRM RESIN PRODUCTION SYSTEIfVI
OF
SOLVENT AND FUSION COOKS
B-14
-------
Thinning tanks are always included as part of the reactor system. They are normally water
cooled and equipped with a condenser and agitator. The partially cooled finished alkyd is
transferred from the reactor to the partially filled thinning tank. Since most alkyd resins are
thinned to 50% solids, the capacities of these tanks are normally twice the capacity of the
reactors. These tanks are also frequently mounted on scales so that thinning solvents may be
accurately added.
The final step in a reactor system is filtering of the thinned resin prior to final storage. This
is normally done while it is still hot. Filter presses are the most commonly used filtering device.
The manufacturing procedures and equipment used for the production of other resins
listed at the beginning of this discussion are quite similar. The major differences are the raw
materials and the process steps utilized. A detailed discussion of these other resins is beyond the
scope of this narrative.
NATURE OF GASEOUS DISCHARGE
There are two major types of emissions from a paint plant. These are non-fugitive and
fugitive. Non-fugitive emissions are those that are collected by and confined within an exhaust
system. Fugitive emissions are those that escape into the plant atmosphere from various
operations and exit the plant building through the doors and windows in an unregulated
fashion.
In today's typical paint plant there are two types of fugitive emissions. These are pigment
particulate and paint solvents. In a small percentage of the plants an attempt is made to collect
these emissions. The incentive for doing so is based on insurance requirements as well as
occupational health and safety rather than air pollution considerations or regulations. The
newly passed OSHA regulations will have a dramatic effect on the paint industry practice and
necessitate the regulation of fugitive emissions.
PARTICULATE CONTAMINANTS
Fugitive particulate emissions consist primarily of the various pigments used. As a general
rule, the pigments are received and stored in 25 to 50 pound paper sacks or fiber drums. Modern
pigment manufacturing has developed fine sized pigment, 0.05 to 0.25 microns, for ease of
dispersion into the paint vehicle. Loading of these fine pigments into grinding equipment results
in fugitive particulate dust emissions into the surrounding plant areas. This dust is either
collected by a ventilation and exhaust system or allowed to settle and later collected as part of
the general housekeeping requirements.
A variety of resins are received as granular or flaked solids which are of large size and do
not result in a fugitive dust emission. The manufacturer of these solid resins, however, does
encounter fugitive emission problems in his flaking or grinding operations.
B-15
-------
GASEOUS CONTAMINANTS
Solvent emissions occur in almost every phase of paint and varnish manufacturing and in
numerous locations throughout individual plants. A listing of emission points is given below.
Location Operation Temp., °F Pressure
1. Resin Plant Thinning 200 to 300° F Atmospheric
2. Resin Plant Filtering 200 to 300° F Atmospheric
3. Resin Plant Storage Tanks 100° F Atmospheric
4. Paint Plant Blending Tanks Ambient Atmospheric
5. Paint Plant Milling Ambient Atmospheric
6. Paint Plant Dispersion Ambient Atmospheric
7. Paint Plant Holding Tank Ambient Atmospheric
8. Paint Plant Filtering Ambient Atmospheric
The extent of these emissions vary with the type of operation and the effort extended to
control atmospheric losses. The high temperature thinning and filtering results in the largest
emissions while packaging in drums and cans contribute the smallest emission. Other operations
contribute intermediate emissions which vary depending on the degree of control exercised and
the vapor pressure of the solvent used.
In some cases, efforts are made to collect fugitive emissions by use of ventilation hoods
and a closed exhaust system. More frequently, however, they are exhausted from the building
by general building exhaust fans which ventilate areas having the highest contaminant
concentration.
Resin plants or paint plants producing resins and varnishes are likely to have a number of
regulated emissions. These emissions consist primarily of gaseous hydrocarbons in air or inert
gas streams. The three major sources of these regulated emissions are:
1. Varnish cooking
2. Resin cooking
3. Thinning
Other less concentrated streams that may or may not be regulated are:
1. Storage and rundown tank vent systems
2. Filter press vent systems
3. Sandmill vent systems
Considerable effort has been expended to identify the various types of chemical
compounds emitted during a varnish cook. The majority of this work was done in the 1950's
and is well summarized by R. L. Stenburg131 in the H.E.W. Technical Report A58-4. A copy of
B-16
-------
his summary is included here as Table 2. In general, one or more of the following compounds
are emitted, depending upon the ingredients in the cook and the cooking temperature; water
vapor, fatty acids, glycerine, acrolein, phenols, aldehydes, ketones, terpene oils and terpene.
These materials are mainly decomposition products of the varnish ingredients.
Varnishes and oils are cooked or bodied at temperatures from 200 to 650° F. At about
350° F decomposition begins and continues throughout the cooking cycle which normally runs
between 8 and 12 hours. The quantity, composition and rate of emissions depends upon the
ingredients in the cook as well as the maximum temperature, the length, the method of
introducing additive, the degree of stirring and the use of inert gas blowing. In general, the
emissions will average between one to three percent of the charge in oil bodying and three to six
percent in varnish cooking. Aside from academic interest, the exact chemical structure of these
emissions is not too important. Of more importance are the characteristics of the emissions
related to ease of removal by the applicable pollution control devices.
Modern resin reactors and varnish cookers account for the majority of clear coatings
production in the paint and varnish industry. As described earlier, these products are cooked in
larger more carefully controlled reactors equipped with product recovery devices which also
help reduce atmospheric emission.
As with the old varnish kettles, the amount of emissions vary with the type of cook, the
cooking time, the maximum temperature, the initial ingredients as well as the type and method
of introducing ingredients.
For solvent cooking the quantity of emission does not vary significantly with the size of
the reactor but is rather more a function of the volatility of the solvent being used and the size
and/or efficiency of the condenser. Since there is no sparge gas used in solvent cooking, exhaust
volumes are small and consist primarily of noncondensed solvent fumes. Emissions will run
from 0.1 to 0.5 pounds per hour and will be less cyclic in nature than for fusion cooks.
Emissions during fusion cooking run much higher and vary with the size of the reactor.
The total exhaust volume is dependent primarily on the sparge rate of inert gas. Dean H.
Parker'41 indicated typical sparge rates of 0.04 ft3/min/gal of charge during the first hour, 0.02
ft3 during the second, and 0.01 ft3 during the remainder of the cook. The exhaust rate will
average from 2 ft3/min/100 gallons of capacity on small reactors to 1 ft3/min/100 gallons of
capacity on large reactors. A summary of source test results from a variety of resin reactors is
presented in Table 3.(11'
Since fusion cooking is a cyclic batch process, the concentration of emission will vary from
the start to finish of the cook. Hydrocarbon concentration will vary from 15,000 to 80,000
ppm as methane equivalent, depending on the time of the cycle and the type of cook. There are
at least 100 different emission curves that could be encountered if one tried to cover all of the
different cooking formulas. Particulate phthalic anhydride (PA) is also emitted from the kettle
and concentration levels vary depending on cycle time, types of cook, method of charging and
type of PA used. Charging of liquid PA rather than dry solid PA significantly reduces the
B-17
-------
TABLE 2
COMPOSITION OF OIL AND VARNISH FUMES13'
Bodying Oils
Water vapor
Fatty acids
Glycerine
Acrolein
Aldehydes
Ketones
Carbon dioxide
Running Natural
Gums
Water vapor
Fatty acids
Terpenes
Terpene Oils
Tar
Manufacturing
Oleoresinous
Varnish
Water vapor
Fatty acids
Glycerine
Acrolein
Phenols
Aldehydes
Ketones
Manufacturing
Alkyd Varnish
Water vapor
Fatty acids
Glycerine
Phthalic anhydride
Carbon dioxide
Terpene Oils
Terpenes
Carbon dioxide
B-18
-------
PO
LU
_J
CO
>•
cr
<
Q
Z
g
I
LU
LL
C O
8,0 ~
2 S g-
CD 03 C
> d) QJ
< CC I-
c
CD
o I
•5- E
^ o
o
•; LU
g < -e
CO CC ^
CO LU ^
g§^
0)
•<•
1
T3
O>
CO
co
O
8
0)
"o
CO
o
CO
CD
_Q3
X
Solvent/fusion cook, based upon
7 resins.
CD
03
esins
Solvent cook, based upon
3
O
.C
~tn
-a g)
S'2>
Q. CO
oo •§
"a Q>
-S o>
T3 CO
gj ts
« T5
^ C
OJ §
en o
C 0)
o co
« ?
IS
CD
C
03
X
es
Fusion cook, based upon
400-4
CD
C
03
> O
a. o
o >
3
o
en -n
1 §
3 E
a a>
oo ^
•o o
03 X
~o o)
03
8 9>
X
e
- ff
en CO
§^
missi
ing
c
'en
03
00
c
o
Q.
O
0
°
c
O
'
5 °- i?
T3 e/> LL.
LU "O to
CO
CD
a>
CN
CD
_QJ _
> o
a o
o >
it 5
00
«— CN
O
Q.
O
a.
03
.Q
O
O
O
•M
a>
O
Lf)
CN
CD
CN
CN
O
C
ca
QQ
O
O
"O
CO
O
O
O
•M
CD
"o
CO
o
(C
•M
3
CD
in
CD
O w ~~
— (T L-
w ^ =E
w < >
LU
in
q
d
LO
(^
o
5
d
co
CN
CN
CO
CO
CN
CN
LO
CO
O
CN
CD
CO
CD
un
O CD ^T
CO N CO
» CO^?
0
o
o
o
o
CD
o
o
0
o
o
O
o
O
o
o
o
o
o
-E S.
IH
T3
CD
"5
Q-
en
03
to C
II
O 03
< ^
B-19
-------
emission rate. However, if the linear velocity of the sparge gas is maintained below 150ft/min,
the carryover of PA is also significantly reduced. Entrained and sublimed PA will run between 1
to 3 pounds per hour over a period of 50 to 70 minutes during and following the charging
period. Plots of hydrocarbon emission level vs. time for three of many possible cooks are given
as Figures 5, 6 and 7.(10) These emission concentrations are those measured directly out of a
closed kettle or reactor.
Figure 7 shows typical variations in emissions from one batch to another when cooking the
same product in the same kettle. Variations twice as great as this are not uncommon. Emissions
increase dramatically and rapidly as indicated on Figures 5 and 6 whenever the loading hatches
are opened. This is a result of forced exhaust of the kettle to prevent spillage of fumes into the
room from the open hatch.
The storage of liquid PA will result in significant vaporization losses from the storage tank
and an effort must be made to control these losses. The most widely used method consists of an
inert gas blanketing in conjunction with a pressure controlled unit. The tank is also equipped
with a water cooled condenser used to vent the tank during filling. After filling the condenser is
then heated with steam to remove collected PA by melting.
POLLUTION CONTROL CONSIDERATIONS
Collection of particulate pigment or resin emission is a simple straightforward job. The
only practical control device is a fabric filter, and it is ideally suited for this application.
Collection efficiency for the submicron pigment dust (0.05 to 0.25 microns) is in the range of
99.9%. There are no temperature problems since the exhaust system runs at ambient
temperatures. The grain loading is very low and baglife is extensive. Approximately 0.01% of
the loaded pigments are lost and collected. Grain loadings to the fabric filter run around 0.19
grain/SCF. A typical collection system is shown in Figure 8. The collection system can be a
fixed hood which can handle both dust and pigment bags or a flexible hose positioned above the
loading hatch or attached to the top of the tank. The tank attachment provides the most
positive control of fugitive dust emission but also increases pigment and solvent losses slightly.
The application of control equipment to this problem is quite simple and can be solved
with standard off-the-shelf equipment from a host of suppliers. For this reason, detailed
equipment cost and installation bids will not be required.
The control of hydrocarbon and odors from the various emission sources listed earlier is
not quite as straightforward as the dust emission. There are three types of control equipment
that have been applied to this problem. They are catalytic and thermal combustion devices, and
wet scrubbers.
As a general rule, wet scrubbing does not provide a satisfactory solution for the following
reasons:
B-20
-------
in
u
ir
D
0
lL
I/)
o:
0
i
i
u
- JdOcdVA Nl
B-21
-------
L
0
h
y
h
0
L
*Z
y 0
£ in
3 ">
2 !
"•- y
z
0
ID
0
0
o:
o
Lo
ou.
ii
Lit
fs
$
y it
it y
Dl
y
o:
D
0
>»
y
I
h
(VI
0"
>0
0
o
0
0
(_LN3~IVAinO3
Ml
B-22
-------
0
y
a:
0
z
0
I
y
o
*
QQ
=IOcdVA Nl
B-23
-------
Id
I/)
d
o:
D Z
0 0
h! ^
!5
5
id
h
z
y
2
B-24
-------
1. Removal efficiency of fine hydrocarbon aerosol is not good at economically practical
pressure drops.
2. Noncondensible hydrocarbon solvent vapors will not be removed.
3. Odor removal without the addition of an oxidizing agent such as potassium
permanganate or sodium hypochorite is unsatisfactory. If an oxidizing agent is used,
operating cost will be quite high due to the high concentration of other oxidizable
material such as phthalic anydride, resins and oil.
4. Mobile packing and high make-up water rates are required to prevent plugging of the
scrubber beds and spray nozzles.
5. Correction of the.air pollution problem with wet scrubbing causes an equivalent water
pollution problem which in many areas is more costly to correct than the original air
pollution problem.
The only control technique currently being used that has proven effective for all cases is
combustion. Three general methods are employed to combust waste gases, as shown below.
1. Flame Combustion
2. Thermal Combustion
3. Catalytic Combustion
All of the above methods are oxidation processes. Ordinarily, each requires that the
gaseous effluents be heated to the point where oxidation of the combustible will take place. The
three methods differ basically in the temperature to which the gas stream must be heated.
Flame combustion is the easiest of the three to understand, as it comes the closest to
everyday experience. When a gas stream is contaminated with combustibles at a concentration
approaching the lower flammable limit, it is frequently practical to add a small amount of
natural gas as an auxiliary fuel and sufficient air for combustion when necessary, and then pass
the resulting mixture through a burner. The contaminants in the mixture serve as a part of the
fuel. Flame incinerators of this type are most often used for closed chemical reactors. They are
not used on resin reactors at present. They may be an ideal solution some day, however, when
methods of operating a closed, pressurized resin reactor are developed.
It is far more likely that the concentration of combustible contaminants in an air stream
will be well below the lower limit of flammability. When this is the case, direct thermal
combustion is considerably more economical than flame combustion. Direct thermal
combustion is carried out by equipment such as that illustrated in Figure 9. In this equipment, a
gas burner is used to raise the temperature of the flowing stream sufficiently to cause a slow
thermal reaction to occur in a residence chamber.
B-25
-------
UJ
F r
(/) UJ
o:
u
O> Q;
Q
UJ
UJ
D O
t
<
QC
O
I-
o
<
UJ
tr
1 2
I (/>
I- UJ
tr
B-26
-------
Whereas flame temperatures bring about oxidation by free radical mechanisms at
temperatures of 2500° F and higher, thermal combustion of ordinary hydrocarbon compounds
begins to take place at temperatures as low as 900 to 1000°F. Good conversion efficiencies are
produced at temperatures in the order of 1400° F with a residence time of 0.3 to 0.6 seconds.
Catalytic combustion is carried out by bringing the gas stream into intimate contact with a
bed of catalyst. In this system, the reaction takes place directly upon the surface of the catalyst,
which is usually composed of precious metals, such as platinum and palladium. While thermal
combustion equipment brings about oxidation at concentrations below the limits of flame
combustion, catalytic combustion operated below the limits of flammability and below the
normal oxidation temperatures of the contaminants. The reaction is instantaneous by
comparison to thermal combustion and no residence chamber is required. Catalytic combustion
is carried out by equipment such as that illustrated in Figure 10.
In general, catalytic afterburners are less expensive to operate, however, they depend
directly on the performance of the catalyst for their effectiveness. It will not function properly
if the catalyst become deactivated. Because of this, catalytic units are not inherently functional
when operated at design conditions. In many areas, means for ensuring adequate performance of
the catalyst on a long-term basis will be required by environmental control offices.
The basis for design of either catalytic or thermal combustion is the hydrocarbon
concentration of the exhaust gases handled by the incinerator. The maximum hydrocarbon level
is set by most insurance companies at one-quarter of the lower explosive limit (L.E.L.) which is
equivalent to 13 BTU/SCF of exhaust gas. As outlined earlier, the quantity of emission may
vary significantly with cooking time and the type of cook. There is also likely to be a large
variety of different hydrocarbons emitted. For this reason, theoretical calculations ofemissions
for design purposes are not satisfactory. On site emission measurements, as shown earlier on
Figures 5 and 6, are required. Once the rate of emission is determined, it is then necessary to
calculate the dilution air required to meet 1/4 LEL and set up the duct work system to provide
for this dilution. When possible, dilution air should be utilized to help capture as many fugitive
fume emissions as possible. For example this can be accomplished by taking the dilution air
from a hood positioned over the resin filter press and venting the thinning tanks and product
rundown tanks into the same system.
A concentration of 1/4 LEL or 13 BTU/SCF will give a temperature rise of about 600° F in
the afterburner. This is too high if a heat exchanger is to be used, and in these cases, we will
dilute to a maximum concentration of 12 BTU/SCF. In all cases, the heat exchangers will be the
parallel flow type having a thermal efficiency of 42%. This is required due to the high emission
concentration to assure temperature balance and control.
The major problem with catalytic or thermal afterburners as applied to open or closed
resin and varnish kettles is the danger of fires and/or explosions. This has happened in numerous
occasions in the past due primarily to excessive hydrocarbon emission from kettles. These
problems have been all but eliminated on newer units by assuring that the design was based on
actual emission measurements of the highest emitting cook and the addition of some of the
B-27
-------
Ill
0
y
£
D
0
t
tfl
y
j
0*
0y
<^
o*
u"
y <
^
Oy
10 ?
0
z
B-28
-------
following system safety features:
1. High limit temperature alarm to shut off burner and activate a diversion system.
2. High velocity duct section to assure gas flow to afterburner substantially exceeds
flame propogation velocity of hydrocarbons being burned.
3. Double manifolding or hot gas recycle to prevent condensation of heavy
hydrocarbons or phthalic anhydride.
4. Diversion system to block off hydrocarbon emissions to unit, by-passing them
directly out of separate exhaust and introduce fresh air to purge the unit.
5. Pneumatic operation of the diversion system to assure fast positive action and provide
a fail-safe system in the event of either air or electrical failure.
6. Purging with inert gas in the event of power failure.
The above general requirements are applicable to all types of afterburner control. Specific
details for each type of system will be given in the equipment specifications.
B-29
-------
REFERENCES
1. Marketing Guide to the Paint Industry, Patricia Noble, Ed., Charles H. Kline & Co., Inc.,
Fairfield, New Jersey, 1969.
2. Federation Series on Coating Technology, Units
1. Introduction to Coating Technology, Oct., 1964
3. Oils for Organic Coatings, Sept., 1965
4. Modern Varnish Technology, May, 1966
5. Alkyd Resins, March, 1967
12. Principles of Formulation and Paint Calculation, June, 1969
17. Acrylic Resins, March, 1971
19. Vinyl Resins, April, 1972, Federation of Societies for Paint Technology, Philadelphia,
Pennsylvania
3. Stenburg, R. L, Control of Atmospheric Emissions from Paint and Varnish Manufacturing
Operations, U.S. Department of Health, Education and Welfare, Robert A. Taft Sanitary
Engineering Center, Technical Report A58-4, Cincinnati, Ohio (June, 1958).
4. Parker, Dean H., Principles of Surface Coating Technology, John Wiley & Sons,
Interscience Publishers Division, New York, N.Y., 1965.
5. The Technology of Paints, Varnishes and Lacquers, Charles R. Martens, Ed., Reinhold
Book Corp., New York, N.Y., 1968.
6. Spence, J. W. and Haynie, F. H., Paint Technology and Air Pollution: A Survey and
Economic Assessment, Environmental Protection Agency, National Environmental
Research Center, Office of Air Programs Publication No. AP-103, Research Triangle Park,
North Carolina, February, 1972.
7. Hardison, L. C., A Summary of the Use of Catalyst for Stationary Emission Source
Control, presented at Franklin Institute, Philadelphia, Pennsylvania (Nov., 1968).
8. Hardison, L. C., "Disposal of Gaseous Wastes", presented at East Ohio Gas Company
Seminar on Waste Disposal, Cleveland, Ohio (May, 1967).
9. Hardison, L. C., "Controlling Combustible Emissions", Paint and Varnish Production, July,
1967.
10. Unpublished source tests of 11 different kettles, Hirt Combustion Engineers, Montebello,
California, August, 1971.
11. Unpublished source tests of 10 different kettles, PPG Industries.
B-30
-------
12. Rolke, R. W., et al, Afterburner Systems Study, Shell Development Co., Emeryville,
California.
B-31
-------
APPENDIX C
CAPITAL AND OPERATING COST FORMS
-------
APPENDIX C
ESTIMATED CAPITAL COST DATA
(COSTS IN DOLLARS)
Effluent Gas Flow
ACFM
°F
SCFM
Moisture Content, Vol. %
Effluent Contaminant Loading
gr/ACF (ppm)
Ib/hr
Cleaned Gas Flow
ACFM
°F
SCFM
Moisture Content, Vol. %
Cleaned Gas Contaminant Loading
gr/ACF (ppm)
Ib/hr
Cleaning Efficiency, %
( 1 ) Gas Cleaning Device Cost
(2) Auxiliaries Cost
(a) Fan(s)
(b) Pump(s)
(c) Damper(s)
(d) Conditioning,
Equipment
(e) Dust Disposal
Equipment
(3) Installation Cost
(a) Engineering
(b) Foundations
& Support
Ductwork
Stack
Electrical
Piping
Insulation
Painting
Supervision
Startup
Performance Test
Other
(4) Total Cost
LA Process Wt.
Small
Large
High Efficiency
Small
Large
- C-l -
-------
X
o
z
u
a.
a.
u
a
<
DC
UJ
a.
01
CO
u
o
•H
+J
X
t— I
Ri
R!
u
r— 1
Ri
e
fn
0)
H
i
2
u
CO
LL
O>
_c
V"
CO
0)
a
o
O 0
0 0
VD 00
•to- -se-
ating Labor (if any)
Operator
Supervisor
1 Operating Labor
S -
Q- o
O I-
LO
IT)
i— I
LO
o
LO
i— 1
i— 1
O
o
vO
/ — %
tenance
Labor
Materials (Average
1 Maintenance
i 1
0
o
o
•vf
o
o
(N
1 — 1
0
"*
o
1^.
0)
60
RJ
H
acement Parts (Ave
1 Replacement Parts
"a +S
0) O
oc H
rH
r-H O LO in
O OO CN O
0000
4—
1 g = to
Q O «^
o ty s s
c -5 1 s i I
~ "to
5 2
i/>
03
O
S 1 °
o ci o
o
0 T3 5
CU CU C
•— .s c
— 3 —
-------
APPENDIX D
EPA SOURCE TEST DATA
-------
CO
w
U
o
to
CO
CO
CO
I
o
w
^J
PQ
to
to to
cd e
S 0
•H
• If)
cj to
• -H
0- 6
• UJ
H
M
t/)
o cd
MS
^S
0 ^
bO O
> rH
< PH
^
• O
bo nj
> -P
< to
0
3 O
(ft i-H
Cd PH
CD
S
tO rH
cd CD
U 4-*
• 0)
bo 4-1 g
> rH 1
< O O
C rH
H PH
•
M
>
C
•rH
(ft
0-
CD
S
•H
H
TJ
CD
O
3
13
0
rH
0,
-7
rH
~v^
(/)
=tt=
^
ST
PH
U
* — '
PH
O
S
PH
<^
^ ^
S
PH
U
'^-s
o"
c\
X
H_j
O
o
C
•H
V '
O
to
J- —
o
o
to
«rj-
^H-
o
r*~
o
o
rj-
o
O
£
O LO
OO Tj-
i-H
O
CM
*
z
Oi
1
*H-
4t
LU
. 1
r— i
H
H
,. — N
U
e
fi.
p.
PJ
1
^
o
0
r -\
•^-S
z
CD
^O
0
S
CO
to
CD
O
O
O,
O
hH
C/D
PH
a
^/|
, ._"]
^
,
to
r-H
cd
rH
0)
-p
C
hH
i — .
~>»^
o
i-H
1
^
O
^J-
CM
0
O
O O
•* 0
oo oo
PH
0
u!J i-H O
•H O
cd O to
X
O 13 O *
tO CD 4-J UJ
CD O-
rrj (/) ^J
cd C cd 13
O -H (D 13
rJ rJ X <
vD O
CM tO
LO LO
i-H i-H
1 1
LO \O
tO CM
^ LO
i-H i — 1
,_,
PH
0
o
LO
to
o
4-»
4.)
cd
CD
X
^
CM
O
to
o
i-H
^O
00
Oi
CM
O
01
CM
o
i— 1
OO
i-H
n-
oo
LO
CM
to
o
0
o
1 — 1
00
00
/"""> /•~"'\
PH
4-1 O
to o
r*-» "^"
r— 1 Tj-
cd
4_) o
cd 4->
U
4->
13 CS
13 CD
CSS
LO
O
00
i-H
1
O
to
LO
rH
^
en
Oi
CM
o
O
\D
00
vD
O
o
O
Oi
LO
0
OO
to
CM
00
LO
CM
O
o
o
o
CM
LO
Oi
rH
cd
CD
i — 1
U
rH
O
13
O
LO
to
00
i-H
LO
O
00
I— t
o
00
LO
^j-
1 — 1
[^
PH
o
LO
oo
to
o
p
i-H
O
o
U
LO
1 — I
Oi
1 — 1
LO
to
oo
, — 1
CTl
**o
to
*^~
o
o
rH
O
o
i-H
o
O
o
•^f
0
CM
^
Oi
f-^
LO
[ —
o
0
o
o
T*"
[^
CJ
•H
rH
cd <1>
rC 13
•P -H
r~ rH
OH 13
£x.
rrj f-^
13 C
to
CM
Oi
rH
LO
rH
Ol
rH
CM
r~~
LO
t~-~
o
O
^o
to
t~~
i — i
0
o
,_,.
•l^'
•
to
LO
i-H
VO
OO
LO
to
CM
O
o
o
LO
o
\o
PH
O
o
*«o
•rj-
O
•p
•p
cd
rH
CD 2
C PH
rH CJ
O
CM
O
CM
to
CM
O
rH
Q
bO
C
•H
rH CL,
O O
O rH
O
LO
i-H
O
I
O
CM
o
4-1
X
CD
cd
CD
OH
*
D-l
-------
H T3
• -H -^ O
X S m $H
• U4 - O,
CM
to
to
o
OO
LO
00
OO
o
LO
en
LO
o
to
in
c
in o
rt -H
U
in in
B «=
PL] v—'
• 5 (D PL,
00 O P U
> rH c4 CO
< IX, OS ^-"
LO LO
\O rH
C2 *O
o o
OO vO
LO to
LO r--
LO
o
o
o
en
en
oo
o
CM
CM
LO
CM
LO
CM
OO
CM
H
<
Q
CO
CO
w
u
o
OS
PL,
n a s
3 O OOPL,
in rH rt t_3
rt PL, OQ <
0 >—i
0
o
LO
to
en
to
en
en
CO
o
(—1
co
CO
in i-i
rt o> oo
CD .P c --?
u -H S
i TJ U,
H S rt U
(D O (D ^—'
C n OS
rH tL,
OO -P
00
en
en
en
Ol
to
en
oo
en
en
oo
00
CM
LO
CO
CO
CXI
I
Q
W
PQ
H
CM to
a: *
t-i JH:
00 OH O 5
s> ^. as
£ <3 ""^
C i
•H
v — ' rt
pq
CJ /—" N I-J
S CJ H
S W
• p. S
00 P
> ^ 1
^
8
0) CJ
*"CJ
o tis
S 0
in c/i
in O
0> PL,
U
O Q
h >-
p l ^^
I
rH
rt to
CD r* ^^
S f-i —
•H ~ \ o o o
IOXT*'-|O-H LOPOCM to O
tOrHrtCJvOrH LO3vO P
rt tortcotoortin in
opofn ^T3 rt rt oo
* .pcdP O OP-H Oh OCD C3 C
m cj MHPXVH-PO-P -H
O.P4-1 CUT3 rH .p 4-> i— 1 P,
rt-OrttJi-H XI-H-HPM h OO '-|
T3
to to \o vo ^^ i^ en rn CM rt cu
O O OOO O OrH r-{ rH *
D-2
-------
Q
C/3
u
o
CO
o
I—I
CO
r—I
W
CO
co
to
PQ
<
H
CO
05
V)
in v> <4H
cd c O *^
So a)
•H C 0
• in o S
CJ i/) t""1 'O
~ •£ "^ °
3* S m fn
• u - cu
H *
W r~*
WC*
tn O fn
• cd .H ,e
oo S in in
<: cj -H -
S B ^
sT
•SOU-,
00 O P CJ
> i-H Cd CO
P M S cd cj^
£ rH 05
0
CM
a:
MH
o
boa,
•H
V— J
o
0
0
i-H
o
0
•
0
i-H
•
i-H
o to
I-- vO
vO
00
CM
•
i-H
i-H
*»=
g
pS
&
1
CM
Bl
^?
* Ot
bo pi
^•4
UJ
52
,
O
O
*"CJ
5
M
O
CO
Q
I
t-H
cd
0) >
6 H
E™* P
c
n
to
f^.
v^^
i-H
i-H
*x»^
^
CM
O
i-H
00
CM
•-H /— \ P
13 O f-i cd
•H P t3 O <1> 1
X <£
cd o cd cd co OH *r3 p us* p5
T3P4JCO +JtO X fn 0)
cdCOP -i-l T3X cdMrH
OOfHCO bOO T3C POX
i-J o. cj *— ' <;P •<
-------
•H
t/)
0
OS
E- fn
. .H ^
H
o
LO
CM
o
vO
o%
LO
to
LO
vO
o
to
LO
to
CM
CM
LO
LO
O
to
LO
vO
o
o
w
• cd
rH
< PU
• o
oo cd
> 4->
< co
ssion:
•H
e
0
•p
nS
OS
'
e
0)
H
4
»-
=**=
^^
/— \
S
fc
U
CO
v '
OH
O
rf
O
0
O
O
\O
O
<^-
(N
f-^
O
LO
CTl
LO
t^-* CD
vO it
• •
\o o
00
OO 00
CM vO
,— {
LO
,—4
•
O
I"-
vO
CM
00
LO
O
rH
CM
CM
to
O
00
00
00
CM
CTl
Ol
LO
O
o
LO
a>
to
o
o
o
LO
o
LO
00
CO
w
U
o
Pi
OH
Q
"^
CO
O
1 — (
CO
CO
1 — 1
^*
tlj
CO
CO
sj
^j-
1
Q
S
PQ
^^
EH
T3
0) g%
3 O bOtL,
i/) i— I cd CJ
Cd PL, CQ bO
U 4J C ^
(U -H S
bO 4-» S T3 U,
> H S rt c_)
<; CD o 4> ' — '
C rH QS
rH [i,
, — ,
0
c\
X
OO&i 4-1
r* <3 O
<
C
•H
>. — /
rH
*
2
DS
1
OO
=«=
PJ
f_
f-H
OJ
i^e<
0
H G"
00 ^
> PH
<: ^
1
0
o
CJ
2
rH
W
OS
0
*"C3
O
w
V)
0)
o
o
£H;
CX
QH
W
H
cn
^H
i-J
o
0,
1
tO
t>^
rH
CTJ
C^ ^
S fH
•H CD
H 4->
C
hH
^^
CM
r_f
1
rH
rH
— ^
t^.
\0
LO
f*^
to
r-j
O
o
X
rH
bo
^» rC
rH V)
O Ctf
0, S
LO
Tf
CM
CM
1
\o
o
CM
0
^H-
rH
0
O
to
LO
e
oj
o
^_)
CO
V 1
, — ^
bO f-i
C 0
•H +J
TJ 0
O T-l
J P4
O
^f
o
o
1
V*O
^Hl-
CM
CM
0
o
£_4
O
•H
^
(X X
3
e?£
•H <1>
•M iS
cd
o> o •
X -P
0
to
rH
o
1
0
o
o
o
rH
o
oo
C
•H 0
•P CM
cd x
0 > — '
X
PL, f~ -,
X 0 0
3 0 >
rH •* O
^H PM
0 O 0
OS -P OS
T^-
to
T^-
o
1
o
to
rH
o
00 LO
o oo
rH CM
0 0
Tt tO
i— I O>
LO rH
CM 00
oo to
U)
0
rO r— ^
00 o
Z > X
O U
rH 3
4-1 O O
0 CM rS
OS X >
LO rH
o n-
O rH
1 1
•* LO
to o
^" t —
0 0
CM
=8=
s
OS
1
00
*
tq
H
H
•*"
i
O
o
U
^
rH
CO
U
OS
Di
yj
H
W
nJ
o
PH
1
to
f^
*• — .
to
rH
1
CM
rH
*-*^
r~->
0
to
0
00
^
o
CM
CM
*o
rH
r~|
(/)
rt
^
rH
O
O
X
rH
00
X
rH
o
(X
o
LO
t^
rH
1
t*^
to
LO
rH
0
t^.
to
o
r^
oo
LO
0
rt
0
4^
co
'G
00 O
C -H
•H 4->
T3 O
cd 0
O '!-»
, "] LIJ
o
o
CJ>
rH
1
O
LO
t^
rH
'*
CM
0
f^
LO
rH
00
rH
r-l
O
•H
fH
O- X
^
00 rH
C 4-1
•H 0
•P OS
cd
0 O
X +->
CM
CM
rH
CM
1
O
0
en
rH
D-4A
-------
' **
^
h~H
H
»-^
<£-*
O
U
<
H
<
Q
CO
CO
W
U
§
l-M
Oc
Q
^~3
*^
CO
Z
O
1— 1
CO
CO
HH
•5"
tc-*
w
CO
CO
*3-
1
Q
W
nj
PQ
H
in TJ
t/) 10 4-1 0>
cd G O u
SO 3
1 r- < -p*
•r-1 >-i tj
•woo
U t/> H 5n
. -H ^ 0,
X B
in G •
w o f-i
• cd -H ,c
bO S t/i \
> I/> OT
< U -H -
X 6 *
H W v_.
/ — \
• S to S
00 O 4J UH
> ^n td u
<; OH oJ co
\ — /
OO
o
o
0
0
CM
0
o
0
•
o
o
1— 1
•
o
^ •
• O PH/ — >
DO Ctf B B,
> +J <1>O
< CO H -^
TJ
a) /— v
^2 S
3 O bOtL,
t/l i-H Cd CJ
cd tL, eQ <
<1) v^
-
00
00
1— 1
r— 1
•
o
V) fn
rt CD OO
UJ +-> C /-N
0) -H S
bO +-> S TJ U-,
> H s ca u
<; a) o a> v — '
G r-l 0!
M B-
/^™^
o
o
•
bOfr. 4-1
> ft
< ^
to
LO
LO
\D
*t
bO
CD
TJ
,0
2
W
«
(U
0
o
^
a,
•H O
•P CM
td X
0 v — '
X
B« r-N
"O TJ
X 0 CD
3 0 >
•-I •* O
4-1 S
os
LO
r-t
cd
e n
•H a>
H 4->
G
i— i
o
r-H
0
1
CM
CM
r— 1
CM
OO
O
r--x
LO
to
LO
CM
0
00
VO
•
o»
00
\0
w
LO
0
00
o
r-l
0
OO
(N
,
O
to
LO
r-^.
\o
00
CMT3
X
O O
2 e
0
•> h
X
3 bo
r-t C
4H -H
0) 0)
OS Xi
to
o
to
o
LO
o
!-(
o
LO
0)
1— 1
o
X
u
g
2
3
0
cd
>
0
^1-
o>
o
i
to
o
to
o
D-4B
-------
H
Q
CO
u
o
CX
r^
Q
<£
CO
o
h- 1
co
CO
1— 1
w
CO
CO
^
!>'
LO
1
s
PQ
^^
H
C
•rH
to
to
tO (O <4-(
cd C O Tf
SO 0)
•H C 0
. to O 2
U to H X)
• -H -^ 0
|T{ £3 (/) M
• uj - a.
H =*!=
to / — *
to C •
to O !H
• cd -H Si
b02 W •».
> to to
<; u -H -
X B *,
/-~^
2
• 5 0 O P U
> rH Cd CO
< PU erf v_/
f>^
to
^^
to
•
o
o^
uO
t —
o
•
0
VO
•
en
f^.
• O Q , *--""J t^-
bO cd B U-, °°
> P 0 Ol
*O
0)
£ § b02
Ifl rH Cd PU
Cd PL, CQ CJ
(D H 13
2 cd
,-J O
1
to
C^ LO
rH
cd
E £
•H d)
H .p
p)
rH
^^ rH
!"*• vO
rH rH
1 1
vO vO
i— 1 LO
•"V. LO
r^* t*H
vD
cn
en
CN
rH
O
r^
00
CM
•
O
^.
rH
•
vO
oo
to
00
oo
o
o
•
o
CN
OO
f^
rH
tu
O
o
o
0
^_)
cd
0)
LO
rH
f-^
rH
1
LO
rH
vO
rH
LO
f^
LO
•51-
CM
vO
CN
^>
LO
•
0
^
r-
•
en
vO
to
o
00
CN
rH
0
•
o
LO
00
p^
1
p
X3C
H O
o cd
<2Z
C °^ ' — *
D- rH §
O 0)
T3 -H tO
CM
CN
rH
LO
rH
|*^
rH
O
f^
LO
0*
00
vO
CN
i— 1
•
0
en
to
•
VO
OO
o
00
en
i*^
o
o
•
0
LO
00
f-~
-[•T |
O
o
LO
o
•p
^_)
cd .
Z
O
o
en
i
CN
CN
r —
rH
0
to
f^.
LO
o
LO
CN
rH
.
0
to
CN
•
VD
00
CN
00
at
t"-.
0
o
•
o
LO
oo
[*•-
PU
o
0
O"!
10
o
i— 4
o
o
u
LO
rH
Ol
1— 1
1
o
o
CT>
rH
t —
CN
en
LO
o
en
0
to
rH
•
0
CN
en
•
en
rH
00
00
F-*
vO
O
0
•
O
VO
00
1
rH O
^ cd p
£^
O O bO
•H CO C
rH -H
cd US' P
rC cd
P Q> Q>
a, -H v— 'p.,
rH O
T3 TJ rH 0
T3 x-i-l oo
< X O •*
vO
O
0
CN
1
LO
rH
cn
rH
CN
to
t--
rH
•
O
C-J
00
to
o
•
0
••*
00
•
vO
00
LO
at
t--
o
0
•
o
vO
to
C1
Pi
o
o
oo
•*
o
p
•p
cd
-------
H
Q
CO
c^
TTl
U
0
OH
pi !
Q
l
vO
1
o
r-^T
_
*™^
PQ
^
P"H
£
•H
W)
t/)
w w *w
|g°-S
•H C U
• in o 3
U in H TJ
• -H --. 0
K g w rH
• r M •- o i
in
in (3 /-N
OT O rH
bo 2 m "^
> in in
•< U -H -
Eg*
H W v-/
CM
00
^f
O
rH
CTl
vO
00
•
O
/-N VO
2
• 5
60 O P U
> rH rt CO
• O ft,— N
oo rt fftL.
> 4-> 0) O
< to H ^
TJ
0
3 0 002^
(/) rH rt BH
rt fc co cj
0) <
^
,—4
•
O
00
o
oo
I/) rH
rt ao
CJ 4-> C .->
0) -H 2
OO.P 6 T» E
> rH S rt CJ^
C rH* <£""'
rH U,
CM
3t=
z
3
t — \ Q*
O
N
X
W) &* O
^ ,^-1 *
- CM
O rH
to
v — /
^J
0)
o
(/)
t/)
o
o
o
O,
8 3
u o
Z rt
O X
rH O
to to
[I* T)
rt
1 2
^*
^-*H
rH
... rt
1 >
•H QJ
H 4J
rH
CM
to TJ-
t-- o
^^ 1
r^- r^
rH rH
•v^ ^>
t-^ o
^.
r*^,
t^
00
o
^J-
^o
l-O
r-H
•
o
Ol
rH
•
vO
TJ-
CM
00
CM
rH
CM
O
o
•
o
CM
i-H
OO
i-H
U,
O
o
0
o
•p
•p
rt
CM
•
LO
^f
to
to
f-^
CM
o
•
o
LO
vO
CTi
1
f^
\\
JH CD
P
< fH ^^
vO
rH
vO
O
1
O
vD
0
^.
CM
i-H
vO
O
(SI
O
rH
•
O
vD
•
vD
LO
LO
I--
CM
rH
to
O
O
•
O
vD
CM
O
rH
UH
O
0
LO
O
p
p
rt
a>
K
o
o
o
rH
|
vO
vD
0
00
i-H
CTi
LO
O
LO
LO
O
1— 1
•
o
vO
r--
•
LO
vO
LO
[^^
CM
i-H
O
o
•
0
o^
LO
00
LLc
o
o
Q^
to
o
•p
rH
o
o
CJ
0
to
o
rH
t
o
o
0
rH
OJ
CM
00
00
O
^J-
r*^
LO
rH
•
O
oo
rH
•
to
pv*
o
oo
LO
LO
0
o
•
o
1—t
LO
rH
rH
|
£J
^£ rt
0 O
•H CO
rH
rt "0-
e*
•p a>
rC TJ
(X -H
rH
TJ TJ rH
TJ X.H
< rC O
to
i-H
rH
rH
1
0
to
0
rH
to
LO
LO
CM
O
LO
LO
^~
O
•
o
vO
01
•
a>
r —
to
oo
oo
vD
O
O
•
0
LO
o
to
o
o
00
o
.p
•p
rt
o
LO
*^
to
rH
1
to
rH
i-H
rH
rH
01
t--
<^-
C3
^~
LO
oo
o
•
o
vO
rH
•
CM
Ol
LO
oo
Ol
00
o
o
•
0
VD
01
2
PL,
CJ
0
rH
in
rt
O
4-*
£_(
Qi
c
i— i
to
rj-
rH
|
LO
Tfr
to
rH
VO
to
LO
CM
O
CM
LO
^>
O
•
o
oo
LO
•
CM
to
LO
00
r^
•
00
rH
rH
0
O
•
O
^f
*rt^
|"^.
2
tu
CJ
o
CM
m
rt
•P
rH
C
I-H
O
o
vD
rH
|
LO
^•
rH
0
[^^
O
to
o
00
*H/
LO
0
•
o
CM
0
•
oo
Ol
CM
00
to
•
^~
i-H
O
0
•
0
CTi
Oi
f-t
o
• H
^1
a,
rH PL
0 0
0 JH
cj a
LO
0
Ol
rH
|
0
o
vO
rH
D-6
-------
c/>
rt
^g
•
CJ
as*
H
t/i
C
O
• H
to
•H
s
PU
c
• H
0)
OS
tw
o
c
0
H
in
V
*
T3
0
O
3
O
OH
vO
n-
CM
oo to
m o
CM o
Tfr OO
vO
o
00
LO
o
CM
LO
o
o
o
o
o
o
o
o
o
CTl
o
o
o
MS:
c
o
• H
t/>
CJ -H -
as 6 *
H W '—
to
Ol
o
O
f-
o
o
i-l
O
o
vO
to
I—I
00
•
o
o
o
o
to
o
o
o
o
o
o
o
o
o
bO O
0) S
-P PL,
cti CJ
oS oo
to
to
o>
1-^
,3.
vO
00
CM
to
vO
00
CM
CM
to
to
vO
O
CO
rH
to
o
CO
vO
i— 1
LO
r— 1
O
to
CNl
O
CM
•
O
O
to
H
<
Q
• o
bO oj
> p
< co
S PL.
(DO
CO
CM
LO
VO
r-
o
OO
O
00
oo
CO
CO
W
U
o
TJ
0)
3 O
to 1-1 cd CJ
cd PL, cQ <
0 v—i
s
CM
to
CM
•
o
to
o
LO
to
CO
1— H
CO
CO
I-H
s
PPl
CO
CO
g
1
Q
w
^^
'•*"j
^
03 CD W)
C3 +J C ,— v
0) -H S
bO -p 6 T3 PL,
> f-t 3 ci CJ
<3 O
<
C
•H
V — /
CO
to
*
Z
^
OS
1
CM
*
tu
PI
CJ
X '-v
H CJ
1
• B
bO PH
i-J
^D
CO
Q
J
i
rH
cd
(15 ^
S3 ?-l
• H »
H ^
^ C
i— (
to
t~^
^-^
CO
^H
^^
t^
T^-
o
0
C3
i— 1
to
vO
•i~>
u u
•H
O to
-P 0)
O -P
?-i co
CJ v_/
00
0
LO
0
1
THy-
T^-
^J-
o
^H-
CM
LO
o
0
r-l
01
CM
i— 1
+J
rt
a>
X
•\
C
O PH
•HO
P 0
rt O
•P to
•H
bO o
< -P
O
r— f
f^
O
1
OO
0
LO
O
LO
•
^-
CM
to
i— 1
O
o
CO
t^
vO
CM
O
•H
r—t
aj D
X T)
-P -H
X ?H
OH T3
^^
5^
CM
CM
p-^
O
1
O
1— I
^^
o
00
o
Ol
1^.
CO
fn
O
W
C 41
0 C
-a a)
C •-!
o x
CJ X
+j -r)
h -d
rt <<
4->
co as1
to
^«
CO
o
1
CM
CM
t\
o
o
CM
O
•p
cS
1)
X
off PU
O
X 0
3 LO
.H tO
m
CD O
OS 4->
0
to
o
1
to
TH/
00
o
vO
CM
to
LO
LO
-p
rt
0)
X
Uff UH
o
X 0
3 oo
iH '*
t4-(
a> o
OS -P
o
o
LO
.-1
1
o
to
Ol
o
o
01
vO
PL,
o
o
00
^4^
4->
rt
tJ PH
i— i B
O 3
X Q
LO
LO
T-l
1
O ^?r
O LO
LO CO
rH rH
D-7
-------
I
C/}
C/3
t— 1
S
w
H
§
oo
Q
m
t>0
cd
fa
<^
e Mass Emission:
4_l
cd
Oi
^J
p^
E~ '
•
g
0
H
O,
M
S
H
O
g
•H
EL*
/^^
•
M
*
* — *
S"
UH
CO
v — '
i
i
f >
u
e
&
v— '
rj^
0
/ — v
o
E
MH
o
*
c
•H
> — >
^j
c
(I)
>
^H
o
CO
B
o
CN
\^
to
•t
r-
to
in
f^
i
M
•H
H
•H
PH
CO
"cd
0)
1
t^-
vO
i-H O
cd
^1
4)
1 1
rj
HH
0>
+J
cd
Q
CN
CN
1
O
LO
i-H
CN
to
r-
\o
iH
•x^
t~~
to
00
in
oo
fx.
•
o
T^-
^
^~
oo
•»
1— *
^
o
o^
m
0
o
0
0)
C
0)
r— t
^
NO
Ol
CN
O
1
to
1— 1
r--
o
to
f^
r^"
i-H
•^x.
^~
00
vO
Tl-
00
i— i
•
^f
to
0
0
CN
*t
to
r--
in
00
rj-
i-H
O
O
tn
•H
FH
•H
O-
CO
cd
CD
g
•H
to
i-H
i-H
Q)
i-H
|
in
o
rH
to
f^
ti"
i-H
^x.
t~-
'
i
\o
i— i
^f
•»
oo
"*
in
r-~
i
CD
C
CD
r-H
X
X
CN
rl-
CN
O
1
o
to
CN
0
to
l~^
00
1— 1
X^
r~-
CN
CN
CN
CN
,-H
0
CTl
•
Tf
\O
O
in
vO
•s
o
o
l-t
o
C71
r-l
m
o
o
tn
•H
f-l
•H
pi,
co
,_4
cd
fa
O
(2
3
in
0
to
^-
i-i
i
00
CN
^<
r— 1
to
r^
00
r— 1
x^
r~-
in
1 C7)
i— 1
CN
00
1 •
in
t-^
r— 1 xO
r^ \o
to \o
•» <\
<-l vO
to to
0 0
0 0
i— 1 i-H
^3-
1— 1
1 0
0
tn
•H
^
•H
PU
CO
rH
cd
^
0)
p^
•H
s
t — vO
CN CN
CN rl"
vO ^O
i-H i-H
1 1
in o^
r-H tO
^O ^O
1— 1 I— 1
to to
r*- t-^
OO OO
i-H rH
•x^. ^x»
r~~ t~-
i— i
to
i-H
r— I
•
^J-
^>
to
1O
^)-
to
•s
CT>
Ol
O
O
i— I
O
^o
Tl-
•-H
o
q
i-H
*
CN
^
in
00
i-H
1
m
^^
00
i-H
to
^
00
1— 1
"*x.
t^
D-8
-------
Oi
i
w
.-J
T3
0)
X
(J
C
cd
h
pa
0\°
t/) UjJ
O I/)
•H / — v t3
•P CM C C
Cd rH S ^
JS U O O
PH 1 C P-
•H OO.X 6
rH U fi O
< ^3 U
T3 in
0) 0)
•H 4->
<*H cd
•H C
o\° 4-> U
o\° cd l
6 vc
0 U
fH >-^
<
to
o
•H / — v
+-> c
td r-
^5 U
o\° P, 1
•H vC
i-H U
< ^
1
c
H
^3
u -^.
X w
E-, -
=t)=
r — \
2
3 0) IX
O 4-> U
rH CS W
B. OS *-^
LO
vD
CM
LO
to
o
.
00
vO
CM
•
O) o>
o^
vO
o
0
0
to
to
00
^
•et
o
LO
00
[*^
.
o
"*
w
+J
•H
^H
•H
+-)
c
(U
>
1— 1
o
t/1
ft
C/}
rH
cd 0)
rH C
o 0
C rH
•H X
S X
OO
t-^
t~^
vO
.
•«*
00
1— 1
.
•*
to
in
•p
•H
rH
• H
PH
co
1 — 1
cd
f-t
u
c
•H
s
CM
CM
rH
CM
.
CM
rH
O
o>
.
•*
vO
in
•(->
•H
!H
• H
PH
to
i— 1
a) cd
C rH
(D (L>
rH C
X -H
X S
CM
vO
00
o
to
CM
ai
at
LO
00
to
LO
r--
CTi
LO
CM
00
LO
On
cn
r-H
rt
fn
0)
C
•H
VO
to
X
X
vO
to
LO
vO
Time
Interval
^D
O
CM
CM
1
0
LO
1— I
CM
CTl
CM
0
1
to
1— 1
o
r— 1
i— 1
l— 1
LO
O
•—I
CM
O
0
to
CM
0
o
to
1— 1
00
CM
r-H
CM
CM
vO
t— 1
LO
rH
vO
i— 1
CM
VO
i— t
1
o>
to
vD
i — i
LO
00
rH
1
1/5
00
s
a
i
vD
to
rH
I
to
I
I
l-~
to
t^~
I
00
to
I
00
to
00
rH
t
to
I
00
I
00'
D-9
-------
t--
o
CM
O
to
CTl
Ol
CM
01
vO
to
CM
f-
LO
CM
Ol
VO
00
Ol
CM
CM
CM
O
w
PQ
•
tofi
•£
,
M
^
'V
Ji
U
s
rH
PH
t/)
t/)
rrt
*>*
1
>
o
r-H
PH
10
0
•rH
4-)
rS
.8-
3
^O
0\°
0\°
PI •
o b
•H rC
(/) t/)
•H -
p3 v — '
0) S
p E
cti CJ
K CO
US'
'Tsipj
i— 1?
U 0
iorS
CJC
0 (D
•H P
<4H rt
•H pj
P CO
fi tut
"
to
u f-^
•H OO
P CJ
i vo
fH V— '
TJ-
rf
O
0
0
Ol
•
o
"*
CN1
:4fc
Z
K
1
=*t=
to
0
p 'c
ftv
•rH '•C
rH CJ
1
Pi
w
^
i
"
i
f\^
8
\^S
z
CD
Q
r§
to
0)
u
o
o
rH
g
^-j
C^
9
1
to
rH
rt
£ rH
•H m
H -P
P:
r— 1
r--
0
i— i
i
Ol
^^
r^
rH
O
•H
^
PL, P
O X
oD" rH O fn
•HOD
rt O to nj
X P
O 13 O C
CO 0) P (D
Q) OH
T3 I/I P
03 P! cS T3
O -H D T)
nJ -J X <
vO 0
CM tO
LO LO
rH rH
1 1
LO VO
tO CM
Tj- LO
rH rH
t N
P.,
0
o
LO
to
o
p
P
rt
C PL,
to O O
CM tO CM
Ol CM tO
rH CM CM
1 1 1
LO to O
rH CM tO
Ol CTl CM
rH rH CM
Ol
to
o
LO
o
O
CM
o
OO
o
CM
S
PL,
CJ
O
to
C?
cd
U
p
^_l
0)
C
rH
O
CM
rH
O
1
O
CM
to
CM
O
CM
LO
O
•
O
vO
•
f^.
to
'_•'•"•'
o
fH
Q
0
P
P?
•H
rH ft
0 0
O fH
CJ Q
\^_j \-_J
O
LO
rH
O
1
0
CM
rH
O
D-10
-------
I
Q
PQ
<
H
to
IO CD
OT to OS
cd C T)
S O M-l CD
•H O O
• to 3
U tft C T)
• -H O O
as 6 E- n
. PJ \ 0.
H to
to C •
to O fn
• cd -H ,C
W) S t/) **^
> to t/j
< U -H -
33 S *
f_ [jj « — /
vO OO to
••* LO O
CM tO O
i— 1 rt OO
• • •
o .-< o
vO \O O
to O i-H
CJi ^^ CD
O O \O
• • •
O t-H O
to *t \o
t — N
S
• S CD PL,
M O P U
> i-l Cd CO
CD O /— •. to
A -H
4-> Cd
S §
"O M
M X
U co
•H U
P 1
cd 6 U
<
to
0
•rH
cd o
(O.U
•rH
rH 1
1 U
o\°
to \o oo
• • •
O> 00 CM
t~^ CM CM
to to
.-1 1— 1
to
^
g
OS
1
CM
=«=
PJ
E:
2
1
j^
o
8
H-3
o
o
g
^**
^*i
i
i— i
cd
C3 ^|
•H CD
H •!->
C
i— i
to
{*-*
••^^
oo
rH
'•x,^
^
^H /— \
FO O £H
•H +J T3 O
O •!-! -H +->
< ^ O O
cd +J T-,
X X 0 pj
O ?H -H
CO CD C 6
cd O cd
T3 4_> 4-> (D
cd C 0 +J
O CD f-i CO
•J O, u v-'
rj- Tf 00
«3- •<* O
to •* LO
O 0 0
1 1 1
O •«* Tt-
O Tf - •!-!
^ C •)->
o\° CD CD -H
Oi +-> i-H TJ
CJ> «4H XT3
? < X <
o
+J Jn +-'
cd o
CD tft +->
33 o c CD cd
•H CD C CD
« I-H 13 CD 33
gcd CD C »— i
PL, X T3 OX "S1
•HO 4-> -H u X
4-) 0 JS fH «
cdo CUT) +-> "d 3 tt.
4J tO X ^1 T) --HO
•H "T3 j£ cd «< Mn o
bO O TJ fi -4-* CD LO
<; OO
O 0 O 0
D-ll
\O C7l
0 0
0 0
0 0
d d
^f r^
0 0
0 0
0 O
• •
o o
t ^5
CM to
o d
o
•P UH
cd o
CD O
33 oo
P
X cd
3 Pu
i— 1 O T3
-------
tsi
rH
I
Q
PQ
<
H
t/1 CD
C 05 T3
WO CD
to -H M-l O
cd if) O 3
2 W T3
•H C 0
• E O rH
U LU E-i (X
• \^
X w
H *=
I/J / v
to C •
W O rH
o Cd -H X
&Q 2 t/) ^^,
> to to
< U -H -
IE E *
£_ tq >— '
^
• S OJ 2
M O -P BH
> i-H Cd C_)
< B. Oi CO
*"O t/) U^"
CD O
O -P CN C
C Cd rH g
cd X u o
rH OH 1 1=1
pa -H oo,x
rH U C
o\° < v— ' "3
T3
CD CD
•H -P
<4-l Cd
•H C
o\° 4-> CD
(M &0
CD X
rH O
CN ^f 1*^
oo r*- rj-
Tt t~- 00
O OO LO
o o •
rH O bO
O) •* CN
vD vO O>
00 LO IO
i — 1 i — ( vD
o • •
O O O
vO O^ O)
rH rH CN
0*0
O VD Tt
OO Tf LO
CN
*:
10
u /— v
•H 00
4-> U
o\° cd i .
B vC
O U
< ^
Z
§
|
to
=tt
W
. T
H
to
o
•H
cd c
X r-
•H 1
rH VC
**• ^£/
C
C^
W
«
1
>H
O
co
L^S
MM
O
O
CD
TJ
O
2
t/i
to
CD
O
O
J_|
(X
QJ
Z
0
rH
CO
D
PL,
Q
^**
— s
1
rH
cd
CD >
S H
•H CD
H 4->
C
rH
to
f*^
^^
^
rH
\^
t*-
X
rH PL. O
•HO Cd
O 0 Z
O
cd Tf air
X
o o * / — .
CO +J MS
cx cd
13 4J CD
cd cd -a -P
O CD p^ t/)
>"H f_L, ^i
vo r^ vo
CN O rH
•^- vD vO
O O O
1 1 1
r^~ vo r^
iH CN O
^t" *H/ \Q
000
00 C71
rH CN
Ov 00
LO 00
tO rH
LO O>
LO t^-
CN Tt
VO
to
LO
CN
O
to
CN LO
Ol LO
O O
LO
LO '*
LO LO
Tt 00
o o
vo
01
LO
vO 00
t^ rH
LO tO
VO
Ol
vO
CN
Ol
CN
LO
O
o
00
vO
CN
to
O
t~-
00
vO
OO
rl-
LO
O
CN
O
00
01
0\°
o
01
o\°
to
o\=
to
O
to
p_
o
o
LO
cd
CD
o
o
o
PL,
o
o
Ol
to
o
o
u
I
o
o
o
o
•H ay
rH
cd CD
rC T3 rH
4-> -H -H
rC rH O
CX T3
X Cd
T) rC X
T) C O
< < CO
B.
O
O
oo
cd
CD
LO
to
i
to
to
cd
•p PU
rH U
CD
C O
LO
LO
to
to
cd
° 2
4-> UH
rH U
CD
C O
h-1 CN
O
O
vD
LO
rH
O
• H
rH
a,
rH CX
O O
O rH
U Q
LO
O
I
o
o
O
4-)
•H
0)
cd
4->
0)
CL,
*
D-12
-------
G
•H
to
to CD
10 G OS T3
CO O CD
ctf -H <+-i O
S w O 3
to T3
« -H G O
U 6 O h
• WHO,
X ^*^.
• 10
H
to
•
o
CTl
CN
LO
r~~
LO
LO
o
O
o
to
LO
CN
CTl
LO
CN
IO
O
O
1^
CTl
LO
r-H
CN
•
O
I
Q
W
J
PQ
<
H
W C •
w o fn
ca -H x!
• S to 'v-.
w> to 10
> C_J -H -
< x e *
f_ gj >_/
«-4
^CH
» 3 CD [i,
W> 0 -P U
> i-H ca co
e &
•H CD
H£
1— f
^.^
t~^
i— 1
|
vO
iH
r^
X
r-H tU O tU
•HO ca o
O 0 Z 0
O LO
ca T* ay •*)•
X
o o * /— v o
CO -PUS P
O. CO
T3 P CD -P
ca ca TJ -P ca
O CD T3 to CD
nJ DC < ^ X
LO LO CN O
i-H i-H CN O
v£> t^- (~- CTl
i— 1 i— 1 r— 1 i— 1
III 1
\O LO LO CN
LO i-H i— 1 CN
LO vO l~» h~
i-H r-H i— 1 i— 1
LO
\D
CN
i— 1
«
O
to
CN
•
vD
00
U.
O
0
CT^
to
O
P
i— I
O
o
u
LO
i-H
CTi
i— (
1
O
o
CTl
T-H
CTi
O
to
i— 1
o
o
CN
CTl
•
CTl
t^
W>
G
•H
P
ca
O CD
•H utr X
rH <— '
cd CD i—\
X! T3
ca
CJ
p
fn
0) 2
C U.
M U
n-
LO
O
o
1
LO
o
to
CN
LO to
to to
Tf LO
0 0
o •
0 0
LO to
00 CTl
• 0
CM CTl
to LO
o
0 .p
CN
t-t
S 00
P> tt, O fH
<: u u Q
i-H O
<-t ft
i-H tO
O O
1 1
TT i-H
LO i-H
O f-H
0 0
o
•p
•H
M
P
X
f-i
0)
ca
•p
CD
0,
*
D-13
-------
C
•H
in
in CD
in in oS
•H o o
• in 3
cj in c *o
• -H o
SC S H
in
=tt=
2
. O*
o
LO
(N
o
O>
IO
r--
to
LO
VO
O
to
to
OO
CN
in
£
^^
§
S3
CO
<
H
<
Q
Z
O
I-H
CO
CO
HH
S
w
CO
CO
> w
< CJ -H
OC 6
H «
• 5 0)
00 O 4->
> I-H rt
<£ Pu OS
O
^
.§
'O
(U
•H
•H
g
*O
0>P
(j$a
0)
•H
.
^
r*t
*v^
in
=
cd
C
wi
^>
0
rH
^
S
OS
i-H
o
c
rt
_r*t
4.)
S
1
00
*
w
^
w
^
tf>
i-H
O
u
i-H
C3
1
J^J
o
o
CJ
z
1— 1
CO
0
*o
o
2
in
m
CD
U
O
o,
OS
os
w
H
CO
w
t-J
o
OH
to
t^^
i-H
rt
>
4-*
c
i— i
^^^
CN
1— 1
1
i-H
i-H
*-*^
t^
^f-^
in
rt
3:
i— t
o
^^
i-H
bO
i— I
O
OH
LO
CN
CN
\o
o
CN
0
vO
LO
Ol
•
LO
r--
\o
00
LO
C
rt
0
^_>
CO
^— '
/ — \
00 FH
C 0
•H +J
TJ 0
rt CD
O •!-»
O
H^-
0
o
rt
(N
CN
0
*^
o
00 CD
o\° (3 fH
O -H -H
o rri
f-i
o
•H
OH X
3
00 r-i
Cfl J
*T1
4J OS
rt
0) O
o
to
r— 1
o
1
0
o
o
LO
i-H
0
in
in
CD
o
o
a,
o
^j
00
c
•H
+-)
rt
a>
ss
•k
^
3
i-H
MH
0)
a:
n-
to
^-
O
o
to
t— 1
o
/~\
"^
o
^
CD
OS
o
CM
SC
\ — /
[T ,
o
o
o
rt
to O
00 i-H
t*» CN
• •
LO CN
LO O
rt Ol
• •
r-~ to
to to
0\°
rt
O>
•
0>
0\°
0
CN
00
e\° o\°
0 0
dP
00 O Ol
i-H Ol O»
fl 1 III
o o
o
CM in r :
3C 0) O O
i-H i-H CN
0 CJ
Z X P P
U rt rt
** t3
X O E S €
3 i> 3 3 3
^-1 0 3 3 3
MH g O O O O O
> DC > SS
LO i-t
O rt *
t~~ to
O i-H
1 1
rt LO
to o
rt t-~
O O
rO
0
^
o
4->
•H
C
1
c
o
rt
IP
O
in
•S
i
f-i
i-H
•H
B
•H
C
•H
Jj
3:
n3
P
f>
g
m
•H
f5
OO
c
•H
(H
3
C
(U
^
rt
P
•P
o
C
0)
^
CD
in
(U
r-t
ft
in
(U
in
CD
£
*
D-14A
-------
**
rH
Q
m
H
CO
os
t/>
tn i/> <+H
cd C O TJ
SO c/i t/i
< CJ -H -
p~| fg =tt~
H tq v— '
/ ,
^ S
M O -U U
> rH Cd W
< tt, OS ^
M
P C
o o
c §
C 0
S u
•13
o> v>
•H CO
•rH CO
c §
V DO
13 >.
rH X
o
o\°
i-H
CM
CM
•
LO
i— (
00
OO
f^
,
CM
vO
vD
i— 1
CM
4(:
2J
*5
OS
pH
o
»Jv.O c~*
OV* fl^
CO
s
o
o
o\«> X
I— 1
CJ
1
00
tu
^J
i
.
»^tf
8
u
5 x:
01 H M
^J S*.
0 -H
a. o
a.
1
to
r — c^
r-i
cd
o >
S ^
•H CO
H 4->
I—I
^v. LO
to r^
?— i i— i
i i
CM r*>
r-l tO
•^^ LO
t*^ »-H
LO vO
0 rH
C7) O
C7> O
* •
to o
m
rH tO
\Q ^^
LO O
r~~ o
C7> O
O ^H
O LO
•-H O
LO
W
CQ
^
1— I
«^
^>
^
W
H
4f
O
O
^
CO f -H CO
T3 O 4J OS
cd co cd
O •!-> CO O
i-J PJ E 4->
O CM
O CM
C71 rH
i-H CM
1 1
o o
in o
rH i-H
o
1—I
o
oo a»
o o
LO rH
00
•H
4-)
cd
cU
K
S
rH
CO
OS
tn
o
rH
O
1
CM
CM
j
CM
O
x}
UH /— X
O CO
0 >
O CO
4-> OS
D-l
0
CMTJ
DC
O
S
rH
CO
a:
to
0
to
0
1
LO
o
o
4B
ro
Q
g
0
00
•H
co
co
rH
O
3
O
cd
o
fi
to
o
to
o
-------
c
• H
(/)
C OS -0
o cu
• H IH CJ
W O 3
u 6 o
« m H
H
o
o
t^- oo
o o
o o
o o
£
^^
^>*
*£
o
CO
<
H
<
Q
2
0
i— i
CO
CO
H- 1
S
w
CO
CO
^}
l^t
rH
Q
fT"]
T
CI"^
f""J
^^
H
•
bO
>
•
bC
>
5
13
0)
fj
O
c
1
0\°
t/1
w
rt
s
u
ac
H
J5
o
i— i
PL,
O
• H
•P
crj
CX
• H
i— 1
W
C
o
•H
(/)
(/)
•H
g
m
0)
P
crj
OS
utr
/ -V
CM
rH
/ — \
•
rH
^C
*^^
w
••
*:
s"
PL,
r^_j
co
v — /
£
|2
r_3 o
l C
OO.X
U
T3
0)
•H
•H
C
13
i— i
tfl
O
•H
P
rt
e
0
^
V)
o
•H
^J
rt
r*i
•H
r— 1
<
C
0
^J
1
vO
CJ
*
w
£_H
E""1
w
1
js^
8
u
I_J
o
co
Q
^H
J
_**»
^-t.
i— 1
crj
fn
O
4_>
c
r— 1
to
f^
*^^
r— (
r-t
*v^
r^
«^
1 1— i / — > P
'O O r^ CfJ
•H p i3 o m
o -H -H P x
< (-t O U
^ < a> •>
rt ^ T-, c
X X o w o
O fH -H -H
CO 0} £- E •P
crj O cS crj
crj C O -P -H
O QiJ ^H t^ WI
, "] p\ t QJ v>_^ ^^
LO
i— 1
T^J-
O
1
0
o
CN
o
o
rH
O
0
•
o
I— (
•
rH
f_l
0\o 0)
C?l 4-*
0^ 4n
o
C -H
0) -P
r-l -H
^» 'O
X 13
<
BH
O
o
LO
to
o
•P
P
crj
(D
PM
off
^
j3
iH
MH
0)
ctf
o
CN
00
0
1
rH
CM
LO
O
^~
o
o
o
•
to
•
o
PU
O
O
00
rj-
o
P
P
crj
0)
X
o5"
X
3
rH
cd
LO
ft
LO
rH
1
O
CM
00
O
^>
o
0
o
•
00
rH
0
O
PL,
O
0
00
^>
4.)
as
13
rH
O
O
0
O
CN
|
LO
ft
LO
rH
ft
D-15
-------
I
Q
W
c
in
in cC
tn tn
cd c ^-i 'O
S O O <1>
•H O
• in c 3
C_? to O "O
• -H H 0
K S -^ rH
• UJ tn CL,
H
in / — >
in c
in O rH
• cd -H X
bo S m *^
> in in
< U -H -
ac e *
f_ tU < — '
/_x
2
• S 0 PL,
bO O 4-> U
> rH Cd CO
< PL, OS V—
(U O /-^
,C -H
M-l 4->
•H Cd
4-* £
G
1 U
G
0\°
(
£
m
H-l
r _,
UJ
|sX
1
^
O
•H ^
_ Q)
4^
c
hH
to
^^
o
>.
^
^^
H
Q
O
z
r— |
•H
o
*^
CD
0)
CO- I PL, PL, cd PL,
• HO X 04->0 0)0
_jO rH OWOrHO
O fl> LO X ^" C_3 vO
Cd tO Cd IO rH Tt tO
X +-• cd fn
O O CrH O4->O O O
rn 4-> 0>O 4-ICd4-> 4-H 4->
(X 4-> U
rrt 4-* *H 4-* 4-* T) *~H
rt cd 'OM ccJ^rt ?H o
3 ac to o ^ in
rC *X3 0) 4-1
Ou T3 rH 4.)
X rH -H 4-J fH
n3 Si O O cd 0)
T3 G O OQ CO C
00 O O O
CM O LO LO
Is- O> rH CM
O O rH rH
1 1 II
LO oo o o
CM CM O LO
t*** t — > O^ rH
O O O rH
LO
0
rH
to
rH
CM
OO
CM
o'
rH
rH
,
rH
p*.
rH
•5?
PL,
U
O
to
c?
cd
CJ
4->
rH
CD
c
rH
LO
O
rH
1
O
LO
CM
rH
/ — v /-~v
p> t
O
J_J
o
4->
bO
'rH ft
O O
0 fH
U Q
v — '< — '
LO
^-
rH
|
LO
0
rH
-------
IO
O i— I tO "3- O i— ltO^J-Or-HtO'3-Or-HtO'd-O!— ( tO •* O t— I tO
IO
CT)
CM
CM
CM
o
tO
oo
00
oo
tO
CMt—li—lOOr-I^1-OI^OOCMIOCM"3-CM
cr>cr>ooioito\OoocMOLOi-ot—IT-H
CMCMtOCMCMtOtOCM^'S-^t-tOtOtO
rH CTi OO O \O *3~ CTi
I I LO LO LO O1 i—I CM K)
I I CM CM CM CM K) K) hO
I O O O O
oo
I I OO
I I CM
tOtOIOlOtOLOr-lt^ OO tO CM CM
OOOOCTlO>CTlLOCMLOtOCMtOtOCT>r—I I
^•«=d-rt^--=t'3-i-OLOcMOr-^LO'*to i
OOOOOOOOrHi—IOOOO
CM OO OO i—I CM i—I CT> CM rH OO OO Tt
I I CTi rH r—I r- "* OO OO I I ^t OO LO vO I I i-H
ooooooo oooo o
O CM CM rH t^ tO
CMOOOOOCTltO'd-LOCnCriOOi—IOCMrHOrH
rHrHrHi-Hi—IrHOOOOOOrHi—IrHrHr-
OO \O tO
CM LO rj- LO O^ rH tO O"l LO tO
rHCDOOCTlOICDOOrHI lO
rHrHtHl—(OLOt^\OCT>CT)vO\OLO[^tO^tvOOO
v£>\ot-~-t--r-^i^voLOLo^j-\OLOto>ot^.!^.o>r-.f-
ooooooooooooooooooo
ooooooooooooooooooo
OOOOOOOO
oooooooo
OO OO tO CM
OO OO F- O
I \O ^t to ^f
IOOOO
oooo
10
OOOOOOOOOOOOO
ooooooooooooo
CMtOCMrHtOCMlOtO
oooooooo
oooooooo
CTl OO tO CM
«* i—I •* vo
rf tO CM CM
O O O O
0000
CMOOCMt^OCTli—iaiOOCMvOCM
CM
tO O
i— I OO
i—l i—I i—IrHrHi—I i—(r-ti—li—ICMCMCMCMCMCMCMCMCMCMtOtOtOtO
D-17A
-------
K5
t^
CTl
CD
s
•(-> -H
bO
ac
u
H
S
o
u
LO O LO O LO O LO
^t" O rH tO ^t- O rH
CTl O O O O rH rH
rH CXI CXI CXI CX] CXI CX]
OO CXI
OO LO
to to
^•LOLOt^-rJ-CTlt-^ CXI
CTi tO to *a- «* tO 00 tM
CXIiotOtOtOtOtO IO
-3" CX] \O LO f-~ VQ CTl rH
rH vO LO O LO f^ r-~ to
LO LO •* oo t^~ r-- \o LO
0000000 O
LO LO CTi O rH «* -=t
CTlt^-CTltOt^-LOrH CTl
CTirHOOOOCTiCTiCT) CTi
OrHOOOOO O
rH
ac
U 43
fn
O w
~ 43
OO
rH
LO
O
o
CTi rH I~~ rH rH VO
LO CX] CXI to \£> f^
^ LO tO I ^ "3- Tf
O O O I O O O
O O O O O O
CTi O ^t
CTi •<*• CTl
CX] tO rH
O O O
o o o
OO O CTl
CXI tO CX]
o o o
o o o
£?
H 43
cxi to ^O
• • •
o o o
rH LO
tO LO CTl rH
o r-- oo
o
\o
rH
o
LO
o
o
to
ex]
to
o
o
o
LO
OO
•H •
^
&
*3- LO \O t-~ OO CTi O
to to to to to to -3-
D-17B
-------
SRL 1372 01 1173
TO EMERGENCY .
EPISODE TANK / \
SPRAY
NOZZLES
SAMPLI NG
POINT
ROOF
"4TH F=1_OOR
STEAM
EJECTOR
3RD FLOOR
» TO SEWER
2ND FLOOR
THINNING
TANK
THINNING CONDENSER
(SURFACE AREA-SOFT2)
1ST FLOOR
REACTOR # 3 &.
FIGURE D- I
D-18
-------
TO EMERGENCY ^
EPISODE TANK / \
SAMPLI NK3
POINT'S
ROOF
RECEIVER TANK
SPRAY
NOZZLES
STEAM
EJECTOR
CONDETSISER
(SURFACE AREA-
-4OO FT2)
FLOOR
3RD FLOOR
P^ »
TO SEWER
2ND FLOOR
THINNING
TANK
THINNING CONDENSER
(SURFACE AREA -3OOFT2)
1ST FLOOR
REACTOR # 2
FIGURE: 0-2.
D-19
-------
SRL I3Y2 01 IIT3
TO EMERGENCY >.
EPISODE TANK /\ SAMPLING
POINTS
GLYCOL
RECOVERY
COND.
CONDEJSJSER
(SURFACE AREA
333 FT*)
STEAM VACUUM
&. COND.
RECEIVER TANK
STEAM
EJECTOR
3RD FLOOR
TO SEWER
THINNINO
TANK
2ND FLOOR
THINNING CONDENSER
(SURFACE AREA - I SOFT2)
1ST FLOOR
REACTOR # 0
FIGURE D-3
D-20
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-450/3-74-031
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Air Pollution Control Engineering and Cost Study
of the Paint and Varnish Industry
5. REPORT DATE
June 1974 (date of issue)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Edward J. Dowd, Project Director
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
Air Resources, Inc.
800 E. Northwest Highway
Palatine, Illinois 60067
11. CONTRACT/GRANT NO.
Contract No. 68-02-0259
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
EPA, Office of Air and Water Programs
Office of Air Quality Planning & Standards
Industrial Studies Branch
Research Triangle Park, N.C. 27711
Final Report
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Under this contract, a comprehensive study was conducted of the Paint
and Varnish Industry and its relationship to the pollution of our
environment. The report presents a description of the industry, its
method of operation and the chemical processes utilized. Also pre-
sented are comprehensive industry statistics including type, size and
location of present day plants and past, present and projected industry
trends. In addition, the following environmental and economic informa-
tion is presented: types and quantities of air pollution emissions and
their geographical distribution; the effect of operations on air
pollution emissions; the impact of emissions on air quality; the type
and effectiveness of existing control technology; performance and costs
of best control technology; the economic impact of the use of best con-
trol by the industry; emission measurement techniques and problems; in-
spection procedures to determine compliance with air pollution
regulations; and areas of needed research and development. The
manufacture of various pigments is included as part of the Paint and
Varnish Industry study. A three page bibliography is included in the
report. The manufacture of Ti02 was studied in detail.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Air Pollution Control Equipment
Scrubbers
Condensation Resins
Zinc Oxide
Iron Oxides
Titanium Dioxide
Cost Analysis
Hydrocarbons
Resin Production
Paint Production
Varnish Production
Particulates
Solvent Emissions
Cadmium Pigment
Chrome Pigment
Thprmal Aftprhurnpr<;. nathlvMr
7/A, 7/B
n/i
13/B, 13/1
8. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21T NO. OF PAGES
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
-------
|