U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
PB-295 743
Revisions to Emission
Factors for AP-42 Organic
Source Categories
Pacific Environmental Services, Inc, Santa Monica, CA
Prepared for
Environmental Protection Agency, Research Triangle Park, NC Office of Air
Quality Planning and Standards
Oct 78
-------
vvEPA
United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC
Air
EPA-450/3-78-108
Octooer 1978
PB 295743
Revisions to Emission
Factors for AP-42
Organic Source
Categories
a^^
REPRODUCED BY
NATIONAL TECHNICAL
INFORMATION SERVICE
U.S. DEPARTMENT OF COMMERCE
SPRINGFIELD, VA, 221S1
-------
NOTICE
THIS DOCUMENT HAS BEEN REPRODUCED
FROM THE BEST COPY FURNISHED US BY
THE SPONSORING AGEN.CY. ALTHOUGH IT
IS RECOGNIZED THAT CERTAIN PORTIONS
ARE ILLEGIBLE, IT IS BEING RELEASED
IN THE INTEREST OF MAKING AVAILABLE
AS MUCH INFORMATION AS POSSIBLE.
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing!
1. REPORT NO. 2.
EPA No. 450/3-78-108
4. TITLE AND SUBTITLE
Revisions to Emission Factors for AP-42
Organic Source Categories
7. AUTHOR(S)
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Pacific Environmental Services, Inc.
1930 14th Street
Santa Monica, California 90404
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
3' "WflWf^810?1^- T
PH^MK f h ^
5. RE#ORl"t>A*>e * ^ ~r -^
October 1978
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
2AA635
11. CONTRACT/GRANT NO.
68-02-2583 '
Work Assignment No. 7
13. TYPE OF REPORT AND PERIOD COVERED ij
Final {
14. SPONSORING AGENCY CODE [
IS. SUPPLEMENTARY NOTES i
EPA Project Officer: Audrey McBath ;
16. ABSTRACT
This document contains the text, as revised or originally written, of several
sections prepared for inclusion in AP-42, "Compilation of Air Pollutant Emission
Factors." Each section, or chapter of this document, includes: a description of the
process, a process flow diagram, a characterization of the emissions, a discussion of
applicable control options, and a quantification of the process's emissions, usually
in the form of emission factors.
The source categories included herein are: solvent degreasing (or cleaning);
transportation and marketing of petroleum liquids (gasoline trucks in transit);
industrial surface coating including coil, can, wire, automobile, light-duty truck,
large appliance,, metal furniture, miscellaneous metal parts and products, flat wood
interior panel, paper, and fabric coating; waste solvent reclamation; cleaning of
rail tank cars, tank trucks, and drums; and asphaltic concrete plants (hot-mix
asphalt). For each section, the background document which discusses the derivation
and source of emission information is also included. \
.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
emission factors
industrial processes
air pollution control
evaporation loss sources
mineral products industry
IS. DISTRIBUTION STATEMENT
Unlimited
b.lDENTIFIERS/OPEN ENDED TERMS
19. SECURITY CLASS (This Report/
Unclassified
c. COSATI Field/Group
~i
I
JO. SECURITY CLASS (This page) I22. PRICE / i
Unclassified \fC'A(K//Wfr4bl \
EPA Form 2220-1 (9-73)
-------
EPA-450/3-78-108
REVISIONS TO EMISSION FACTORS
FOR AP-42
ORGANIC SOURCE CATEGORIES
by
PACIFIC ENVIRONMENTAL SERVICES, INC.
1930 14th Street
Santa Monica, California 90404
Contract No. 68-02-2583
Work Assignment No. 7
EPA Project Officer: Audrey McBath
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
October 1978
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations - in limited quantities - from the
Library Services Office (MD-35), U.S.. Environmental Protection Agency,
Research Triangle Park, North Carolina 27711; or, for a fee, from the
National Technical Information Service, 5285 Port Royal Road, Springfield,
Virginia 22161.
This report was furnished to the Environmental Protection Agency by
Pacific Environmental Services, Inc., 1930 14th Street, Santa Monica,
California 90404. The contents of this report are reproduced herein
as received from Pacific Environmental Services, Inc. The opinions,
findings, and conclusions expressed are those of the authors and not
necessarily those of the Environmental Protection Agency. Mention of
company or product names is not to be considered as an endorsement by
the Environmental Protection Agency.
Publi cati on No. EPA-450/3-78- 108
-------
TABLE OF CONTENTS
AP-42
Section Page
INTRODUCTION 1
4.2 INDUSTRIAL SURFACE COATING 3
4.2.1 General 3
4.2.2 Coil and Can Coating 8
4.2.3 Magnet Wire Coating 17
4.2.4 Automobile and Light-Duty Truck Coating 19
4.2.5 Other Metal Coating 23
4.2.6 Flat Wood Interior Panel Coating 33
4.2.7 Paper Coating 38
4.2.8 Fabric Coating 44
References 48
Background Document 50
4.4 TRANSPORTATION AND MARKETING OF PETROLEUM LIQUIDS 54
References 67
Background Document 69
4.5 WASTE SOLVENT RECLAMATION 80
References 90
Background Document 91
4.6 SOLVENT DECREASING 92
References 97
Background Document 101
4.7 TANK AND DRUM CLEANING 108
References 113
Background Document 114
8.1 ASPHALTIC CONCRETE PLANTS 117
References 130
Background Document 133
-------
LIST OF ILLUSTRATIONS
Figure Page
4.2-1 Coil Coating Line Emissions Points 10
4.2-2 Three-Piece Can Sheet Printing Operation Emission
Points 11
4.2-3 Wire Coating Line Emissions Points 18
4.2-4 Automobile and Light-Duty Truck Coating Line
Emission Points 22
4.2-5 Metal Product Coating Line Emission Points 27
4.2-6 Flatwood Interior Panel Coating Line Emission Points . 35
4.2-7 Paper Coating Line Emission Points 41
4.2-8 Fabric Coating Line Emission Points 45
4.4-1 Flowsheet of Petroleum Production, Refining and Distri-
buti on Systems 55
4.4-2 Splash Loading Method 56
4.4-3 Submerged Fill Pipe 56
4.4-4 Bottom Loading 57
4.4-5 Tanktruck Unloading Into an Underground Service
Station Storage Tank 58
4.4-6 Tanktruck Loading with Vapor Recovery 62
4.4-7 Automobile Refueling Vapor-Recovery system 67
4.5-1 General Waste Solvent Reclamation Scheme and
Emission Points 81
4.5-2 Typical Fixed-Bed Activated Carbon Solvent Recovery
System 83
4,5-3 Distillation Process for Solvent Reclaiming 84
4.6-1 Degreaser Emission Points 94
8.1-1 Batch Hot-Mix Asphalt Plant 118
8.1-2 Continuous Hot-Mix Asphalt Plant 119
8.1-3 Dryer Drum Hot-Mix Asphalt Plant 121
11
-------
LIST OF TABLES
Table Page
4.2-1 VOC Emission Factors for Uncontrolled Surface Coating. 7
4.2-2 Typical Densities and Solids Contents of Coatings 7
4.2-3 Emissions Control Efficiencies for Surface Coating
Operations . 8
4.2-4 VOC Emission Factors for Can Coating Processes 15
4.2-5 Emissions Control Efficiencies for Coil and Can
Coating Lines 16
4.2-6 Organic Solvent Emissions From a Typical Wire Coating
Line 20
4.2-7 Emissions Factors for Typical Plant Assembling
Intermediate-Sized Automobiles and Light-Duty Trucks . 24
4.2-8 Control Efficiencies for Automobile and Light-Duty
Truck Coating Lines 25
4.2-9 Emission Factors for Typical Uncontrolled Metal
Coating Lines 30
4.2-10 Estimated Control Technology Efficiencies for Metal
Coating Lines 32
4.2-11 VOC Emission Factors for Interior Printed Panels 37
4.2-12 Solvent Emissions From Uncontrolled Paper Coating
Lines 43
4.2-13 Control Efficiencies for Paper Coating Lines 43
4.2-14 Solvent Emissions From Uncontrolled Fabric Coating
Plants 47
4.4-1 S Factors for Calculating Petroleum Loading Losses ... 59
4.4-2 Hydrocarbon Emission Factors for Gasoline Loading
Operations 60
4.4-3 Hydrocarbon Emission Factors for Petroleum Liquid
Transportation and Marketing Sources .„ 65
4.4-4 Hydrocarbon Emissions From Gasoline Service Station
Operations 66
4.5-1 Emission Factors for Solvent Reclaiming 88
4.6-1 Solvent Loss Emission Factors for Degreasing
Operations 99
4.6-2 Projected Emission Reduction Factors for Solvent
Degreasing 100
4.7-1 Emission Factors for Rail Tank Car Cleaning ~. Ill
4.7-2 Emission Factors for Tank Truck Cleaning 112
4.7-3 Emission Factors for Drum Burning 113
iii
-------
Table
LIST OF TABLES
8.1-1 Emission Factors for Selected Materials From an
Asphaltic Concrete Plant Stack 124
8.1-2 Characertistics of an Asphaltic Concrete Plant
Selected for Samp! ing 126
8.1-3 Participate Emission Factors for Conventional
Hot-Mix Asphaltic Plants 128
8.1-4 Potential Uncontrolled Fugitive Participate Emission
Factors for Conventioanl Asphaltic Concrete Plants.... 129
8.1-5 Participate Emission Factors for Dryer Drum Hot-Mix
Asphalt Plants 129
iv
-------
INTRODUCTION
As the Table of Contents indicates, the chapters of this
report were written for inclusion in EPA Publication No. AP-42,
"Compilation of Air Pollutant Emission Factors." In February
1978, the following sections were assigned and were prepared in
approximately the order listed:
4.4 Transportation and Marketing of Petroleum Liquids.
The 4/77 version of this section was revised by PES
to include "Transit Losses From Gasoline Tank Trucks;"
the background document accompanying the revised text
pertains only to the emission factors for such losses.
4.6 Solvent Degreasing.
4.2 Industrial Surface Coating.
In August 1978, three additional sections were assigned and
written:
4.5 Waste Solvent Reclamation.
4.7 Tank and Drum Cleaning.
8.1 Asphaltic Concrete Plants. In December 1977, PES com-
pleted this section and a background document, for
particulate emissions only, under a separate task
assignment. In August 1978, the additions of organic
emission factors and individual factor ratings were
assigned under the present contract and task order.
This assignment entailed partial revision of the text,
tables, and the background document.
Five AP-42 sections were assigned that are not included
herein:
5.23 Rubber Tire Manufacturing. A preliminary draft of
this section and a partial background document were
submitted to EPA in April 1978. The CTG document on
"Rubber Products Manufacture" has since been post-
poned from July to December 1978, so revision and
review of this section have also been delayed.
-------
4.5 Graphic Arts. A preliminary draft of this section was
started when the draft version of the CTG document on
"Graphic Arts" was published. Since that time, the
final publication date of this document has been post-
poned from July to December 1978, so completion of the
AP-42 section also had to be postponed.
5o22 Adhesives. Publication of a CTG document on "Adhesives"
has been postponed indefinitely, so this part of the
original work assignment was canceled.
5.22 Explosives Detonation. A source document for this
section was received in August 1978, along with the
suggestion that its author be contacted for "other
materials which he has access to and suggests using."
PES agreed to read the source document and make
requests for additional materials if so indicated,
but suggested deferring preparation of the section
due to the 5-week time limit involved. The Project
Officer concurred.
5.22 Vinyl Chloride. An SSEIS document that deals with
emissions of vinyl chloride from all kinds of sources
was received, along with this section assignment, in
August 1978. Because of the time limit involved and
the fact that the source document includes no infor-
mation on the conventional criteria pollutants, PES
requested that preparation of this section be deferred.
The Project Officer granted this request.
-------
4.2 INDUSTRIAL SURFACE COATING
4.2.1 General 1-4
4.2.1M Process Description - Surface coating involves the applica-
tion of decorative or protective materials in liquid or powder form
to any of a number of substrates. These coatings normally include
general solvent-type paints, varnishes, lacquers, and water-thinned
paints. After application by one of a variety of methods, such as
brushing, rolling, spraying, dipping, and flow coating, the surface
is air and/or oven dried to remove the volatile solvents leaving the
coated surface. Powder type coatings can be applied to a hot sur-
face or melted after application and caused to flow together. Other
coatings can be applied normally, then polymerized by curing ther-
mally with infrared or electron beam curing systems.
Coating Operations - There are both "toll" (also called "inde-
pendent") and "captive" surface coating operations. Toll opera-
tions fill orders to various manufaturers' specifications, and thus
change coatings and solvents more frequently than captive compan-
ies, which coat and fabricate products within a single facility and
may operate continuously with the same solvents. Whether a surface
coating operation is toll or captive makes a difference to what
emission control systems are applicable to its coating lines be-
cause not all controls are technically feasible in toll situations.
Coating Formulations - Conventional coatings, which are still
the most widely used, contain at least 30 volume percent solvents
to permit easy handling and application. More typically, they con-
tain 70 to 85 percent solvents by volume. These solvents may be of
one component or a multicomponent mixture of volatile ethers, ace-
tates, aromatics, cellosolves, aliphatic hydrocarbons, and/or wa-
ter. Coatings with 30 volume percent of solvent or less are called
low solvent (or "high solids") coatings.
-------
Water-borne coatings, which have recently gained substantially
in usage, are of several types: water emulsions, water soluble and
colloidal dispersions, and electrocoats. Common ratios of water to
organics in emulsion and dispersion coatings are 80/20 and 70/30.
Two part catalyzed coatings that can be dried, powder coatings,
hot melts, and radiation cured (ultraviolet and electron beam)
coatings contain essentially no volatile organic compounds (VOC),
although some monomers and other lower molecular weight organics may
volatilize.
Depending on the product requirements and material being coated
a surface may have one or more layers of coating applied. The first
coat may be applied to smooth out surface imperfections or to ensure
adhesion of the coating. The intermediate coats usually provide the
required color, texture, or print, and a final clear topcoat is
often added as a protective measure. Although the intended use and
material to be coated determines the composition and resins used in
the coatings, the general coating types do not differ from those
described.
Coating Application Procedures - Conventional spray, which is
air-atomized and normally hand operated, is one of the most versa-
tile coating methods. Colors can be changed easily and a variety of
sizes and shapes can be painted under many operating conditions with
good results. Conventional, catalyzed, and water-borne coatings can
be applied with little modification. Disadvantages are low effi-
ciency due to overspray and high energy requirements for the air
compressor.
In hot, airless spray, the paint is forced through an atomiz-
ing nozzle under pressure. Since volumetric flow is lesss over-
spray is reduced. Less solvent is also required, thus reducing VOC
-------
emissions. Care must be taken to avoid plugging and abrading of
the nozzle orifice to ensure proper flow and atomization of the
coating. Electrostatic spray is most efficient with low
viscosity paints. Charged paint particles are attracted to an
oppositely charged surface. Spray guns, spinning discs, or
bell-shaped atomizers can be used to atomize the paint. Application
efficiencies of 90 to 95 percent are possible with good wraparound
and edge coating. Interiors and recessed surfaces are difficult to
coat, however.
Roller coating is used to apply coatings and inks to flat sur-
faces. If the cylindrical rollers move in the same direction as
the surface to be coated, the system is called a direct roll coat-
er; if they rotate in the opposite direction, the system is a re-
verse roll coater. Coatings can be applied efficiently, uniformly,
and at high speeds to any flat surface. Printing and decorative
graining are applied with direct rollers. Reverse rollers are used
to apply fillers to porous or imperfect substrates including papers,
and fabrics, to give a smooth, uniform surface.
Knife coating is relatively inexpensive, but is not appropri-
ate for coating unstable materials, such as some knitgoods, or when
a high degree of accuracy in the coating thickness is required.
Rotogravure printing is widely used in vinyl coating of imita-
tion leathers and wallpaper, and in the application of a transpar-
ent, protective pattern over the printed pattern. In rotogravure
printing, the image area is recessed or "intaglio" relative to the
copper-plated cylinder on which the image is engraved. The ink is
picked up on the engraved area, and excess ink is scraped off the
nonimage area with a "doctor blade." The image is transferred di-
rectly to the paper or other substrate, which is web fed; the pro-
duct is then dried.
-------
Dip coating requires that the surface to be coated be immersed
into a bath of paint. Dipping is effective for coating irregular,
bulky items and for priming. All surfaces are covered but coating
thickness varies, edge blistering can occur, and a good appearance
is not achieved.
In flow coating, materials to be coated are conveyed through a
flow of paint. Paint flow is directed, without atomization, towards
the surface through multiple nozzles, then caught in a trough and
recycled. For flat surfaces, close control of film thickness can be
maintained by passing the surface through a constantly flowing cur-
tain of paint at a controlled rate.
4.2.1.2 Emissions and Controls - Essentially all of the volatile
organic compounds (VOC) emitted from the surface coating industry
result from the solvents which are either part of the paint formu- .
1 ations, used to thin paints at the coating facility, or used for
cleanupo All unrecovered solvent can be considered as potential
emissions. Monomers and low molecular weight organics can be emitted
from those coatings that do not include solvents, but these emis-
sions are essentially negligible.
Emissions from surface coating can be estimated for an
uncontrolled facility by assuming that all VOC in the coatings are
emitted as pollutants. Usually the coating consumption in gallons
will be known and some information about the coating will be avail-
able. The factors in Table 4.2-1 can be used to calculate
emissions. The choice of the particular factor will .depend on the
kind of coating data that is available. If no specific information
is given for the coating, it may be estimated from the data in Table
4,2-2.
All solvents that are used in surface coating operations and
not subsequently recovered can be considered as potential emissions.
-------
Table 4.2-1. VOC EMISSION FACTORS FOR UNCONTROLLED SURFACE COATING
(MATERIAL BALANCE WHEN VOLUME USAGE OF COATINGS IS KNOWN)
EMISSION FACTOR RATING: B
Available Information on Coating
Conventional or Water-borne Paints
VOC as weight percent (d)
VOC as volume percent (V)
Water-Borne Paint
VOC as wight percent of total
volatile* - Including water (X);
total volatiles as weight
'- percent of coating (d)
VOC as volume percent of total
volatiles - Including water (Y);
total volatlles as volume
percent of coating (V)
Emissions of VOCa
kg/liter of coating
d-coat1nq density
100 '
V-0.88
160
d-X-coat1nq density
100
V-Y-0.88
166
Ib/gal of coating
d-coatino density
To"5
V-7.36
100
d'X'Coatinq density
100
V-Y-7.36
T5o~
'.For special purposes, factors expressed as Ib/gal of coating less water
Bay be desired. These may be computed as follows:
Factor as Ib/gal of coating
1 - volume % water
100
Factor as Ib/gal of coating less water
If the coating density 1s not known, It can be estimated from the Information
1n Table 4.2-2.
Table 4.2-2. TYPICAL DENSITIES AND SOLIDS CONTENTS OF COATINGS
Type of Coating
Enamel, air dry
Enamel , baking
Acryl 1c enamel
Alkyd enamel
Primer surfacer
Primer, epoxy
Varnish, baking
Laquer, spraying
Vinyl, roller coat
Polyurethane
Stain
Sealer
Density
kg/liter Ib/gal
0.91 7.6
1.09 9.1
1.07 8.9
0.96 8.0
1.13 9.4
1.26 10.5
0.79 6.6
0.95 7.9
0.92 7.7
1.10 9.2
0.88 7.3
0.84 7.0
Percent Sol Ids
(by volume)
39.6
42.8
30.3
47.2
49.0
57.2
35.3
26.1
12.0
31.7
21.6
11.7
-------
Such VOC emissions at a facility can result from on-site dilution
of coatings with solvent, from "make-up solvents" required in flow
coating and in some instances in dip coating, and from the solvents
used for cleanup. Make-up solvents are solvents that are added to
coatings to compensate for standing losses in pressure, concentra-
tion, or amount, thus bringing the coating back to working
specifications. Solvent emissions should be added to VOC emissions
from coatings to arrive at total emissions from a coating
facility.
Emission controls normally fall under one of three cate-
gories: modifications in paint formulations, process changes, or
add-on controls. These are discussed more fully in the specific
sections which follow. Typical ranges of control efficiencies are
given in Table 4.2-3.
Table 4.2-3. EMISSIONS CONTROL EFFICIENCIES FOR
SURFACE COATING OPERATIONS1"3
Control Option
Substitute water-borne coatings
Substitute low-solvent coatings
Substitute powder coatings
Add afterburners/incinerators
Percentage .
Reduction3 • '
60-95
40-80
92-98
95
3 Expressed in percentage of total uncontrolled
emission load compared to that of a conventional
solvent-borne coating.
4.2.2 Coil and Can Coating5"8
4.2.2.1 Process Description - Coil coating is the coating of
any flat metal sheet or strip that comes in rolls or coils. Cans
are made from two or three flat pieces of metal, so can coating is
included within this broad category, which also includes the coating
-------
of screens, fencing, metal doors, aluminum siding, and a variety
of other products. Figure 4.2-1 shows a typical coil coating line,
while Figure 4.2-2 depicts a three-piece can sheet printing
operation.
Coil Coating - There are both "toll" and "captive" coil
coating operations. The former fill orders to customer specifica-
tions; while the latter coat the metal and fabricate the products
within one facility. Some coil coaters serve both purposes.
Coil coating lines have one or more coaters, each followed by
an oven (refer to Figure 4.2-1). The metal is cleaned and treated
for corrosion protection and proper coating adhesion (refer to Sec-
tion 4.6, Solvent Degreasing). The prime coat is applied, on one
or both sides, by three or more power-driven rollers; it is typi-
cally reverse-roller coated. This coating is dried or baked, then
cooled in a quench chamber either by a spray of water or by a blast
of air followed by water cooling. Another method of applying a
prime or single coat when a water-borne coating is used is electro-
deposition.
Oven temperatures range from 40 to 380°C (100 to 1,000°F),
depending on the type and desired thickness of the coating and on
the type of metal being coated. A topcoat may be applied and cured
in a similar manner.
Can Coating - As with coil coating, there are both toll and
captive manufacturers. Some plants coat metal sheets, some fabri-
cate three-piece cans, some fabricate and coat two-piece cans, and
some fabricate can ends. Others perform combinations of these
processes.
Cans may be either made from a rectangular sheet (body blank)
and two circular ends ("three-piece" cans) or drawn and wall-ironed
-------
ttf SfllHO ?n?TRt ATWf fsT
OV(N,
ni
our«cH
9vrn
(I)
Figure 4.2-1 Coil Coating Line Emissions Points
-------
INK
• APPLICATORS
•LANKET
CYLINDER
SHEET (PlATEi
FEEDER
LITHOGRAPH
COATER
OVER VARNISH
COATER
WICKET OVEN
^
2
SHEET (PLATE)
STACKER
Figure 4.2-2. Three-Piece Can Sheet Printing Operation
Emission Points7
-------
from a shallow cup to which an end is attached after the can is
filled ("two-piece" cans). There are major differences in coating
practices depending on the can and the type of product packaged in
it,
Three-piece can manufacturing involves sheet coating (refer to
Figure 4.2-2) and can fabricating; sheet coating includes base
coating and printing or lithographing, which is followed by curing
at temperatures of up to 220°C (425°F). When the sheets have been
formed into cylinders, the seam is sprayed - usually with an air-
dry lacquer - to protect the exposed metal. If they are to contain
a product for human consumption, the interiors are spray coated and
the cans baked at up to 220°C (425°F).
Two-piece cans are typically used by beer and other beverage
industries. The exteriors may be reverse-roll coated with a white
basecoat, which is cured at 170 to 200°C (325 to 400°F). Several
colors of ink are then transferred (sometimes by lithographic
printing) to the cans as they rotate on a mandrel. A protective
varnish may be roll coated over the inks; the coating is then cured
in a continuous, single or multipass oven at temperatures of 180 to
200°C (350 to 400°F). The cans are spray coated on the interior and
spray and/or roll coated on the exterior of the bottom end. A final
baking at 110 to 200°C (225 to 400°F) completes the process.
4.2.2o2 Emissions and Controls - Emissions from coil and can
coating operations depend on composition of the coating, area to be
coated, thickness of coat, and efficiency of application. Post-
application chemical changes and nonsolvent contaminants, such as
oven fuel combustion products, may also affect the composition of
emissions. All solvent used and not recovered can be considered as
potential emissions.
12
-------
Coll Coating - Coil coating emissions come from (1) the coat-
ing area, (2) the oven, and (3) the quench area (these numbers
refer to Figure 4.2-1). They consist of volatile organics and
other compounds, such as aldehydes, which result from thermal
degradation of volatile organics. Emissions from combustion of
natural gas, which is typically used to heat the ovens, are dis-
cussed in Section 1.4. Emissions from coil coating can be esti-
mated from the amount of coating applied by using the factors in
Table 4.2-1.
Both incineration and the use of water-borne and low solvent
coatings reduce organic vapor emissions; other technically feasible
control options, such as electrostatically sprayed powder coatings,
are not presently applicable to the industry as a whole. Both cat-
alytic and therma-l incinerators can be used, preferably with pri-
mary and/or secondary heat recovery systems. Water-borne primers,
backers (coatings on the reverse or backside of the coil), and some
low to medium gloss topcoats that equal the performance of organic
solvent-borne coatings have been developed for aluminum, but have
not yet achieved full line speed in all cases. Water-borne coat-
ings for other metals are in the early stages of development.
Can Coating - Sources of can coating VOC emissions include
(1) the coating area and (2) the oven area of the sheet base and
lithographic coating lines (these numbers refer to Figure 4.2-2);
the three-piece can side seam and interior body spray coating
processes; and the two-piece can coating and end sealing compound
lines. Emission rates vary depending on line speed, size of can or
sheet, and type of coating. On sheet coating lines, where the coat-
ing is applied by rollers (refer to Figure 4.2-2), most solvent
evaporates in the oven; for other coating processes, the coating
operation itself is the major source of emissions. Emissions can be
estimated from the amount of coating applied by using the factors in
13
-------
Table 4.2-1, or from Table 4.2-4 if the number and general nature of
the coating lines are known.
Available control technology includes the use of add-on devices
(i.e., incinerators and carbon adsorbers) and the conversion to low
solvent and ultraviolet curable coatings. Both thermal and cata-
lytic incinerators may be used to control emissions from threepiece
can sheet base coating lines, sheet lithographic coating lines, and
interior spray coating; incineration is also applicable to two-piece
can coating lines. Carbon adsorption, on the other hand, is most
acceptable to low temperature processes which use a limited number
of solvents. Such processes include two- and three- piece can
interior spray coating, two-piece can end sealing compound lines,
and three-piece can side seam spray coating.
Low solvent coatings are not yet available to replace all the
organic solvent-borne formulations presently used in the can indus~
try. Water-borne, low solvent, and powder coatings are in various
stages of development. The requirement to gain approval by the Food
and Drug Administration is a lengthy part of the testing program.
Also, they cannot yet compete with conventional coatings in meeting
customer specifications. Water-borne basecoats have been success-
fully applied to two-piece cans, however. Powder coating technology
is being pursued for two-piece can interiors, and has been used for
the side seam coating of noncemented three-piece cans.
Ultraviolet curing technology is available for rapid drying of
the first two colors of ink on three-piece can sheet lithographic
coating lines. It is still in the development stage for curing
three-piece exterior basecoat, ink, and overvarnish in a single pass
and for curing the exterior coating of two-piece beer and beverage
cans.
Table 4.2-5 shows control efficiencies for typical, average
coil and can coating lines.
14
-------
Table 4.2-4 VOC EMISSION FACTORS FOR CAN COATING PROCESSES'
EMISSION FACTOR RATING: B
Process
Three-piece can sheet base coating line
Three-piece can sheet lithographic
coating line
Three-piece beer and beverage can -
side seam spray coating process
Three-piece beer and beverage can -
Interior body spray coating process
Two-piece can coating line
Two-piece can end sealing compound
line
Typical volatile
organic emissions.
from coating line8
kg/hr
51
30
5
25
39
4
Ib/hr
112
65
12
54
86
B
Estimated
fraction .
of emissions
from cofltcr
area (percent)
9-12
8-11
100
75-85
uncertain .
100
Estimated
fraction of.
emissions from
oven (percent)
88-91
89-92
air dried
15-25
uncertain
air-dried
Typical organic
emissions"
MT/yr
176
55
20
88
287
15
English
ton/yr
160
50
16
80
260
14
* Organic solvent emissions will vary from line to line as a result of line speed.
size of can or sheet being ocated, and type of coating used. A typical line may
coat 500,000-200.000,000 cans per day or 100,000 sheets per day.
Based upon normal operating conditions, which range from 1,500 to 7,000 hr/yr
depending on the process.
-------
Table 4.2-5.
EMISSIONS CONTROL EFFICIENCIES FOR COIL
AND CAN COATING LINES7
Affected .
Facility*
COIL COATINS LIKES
TWO-PIECE CAN LINES
Exterior coating
Interior spray
CO* ting
THREE-MECf CAN LIKIS
Sheet coetlnc lines
Exterior coating'
Interior spray
easting
tef fab-1c«t1nc lines
Side team spray
coating
Interior spray
coating
End coating lines
"Sealing compound
•Sheet coating
Control Option
Themal Incineration
Catalytic InciRmtion
Hater-borne and fiigh cellos
KM tings
Thermal ondreatalJFtie
Incineration '
Hater-borne end high col Ids
coatings
Ultraviolet curing
Thennel and catalytic
Incineration
Hater-borne and high colitis
coatings
Powder coating
Carbon adsorption
Thermal and catalytic
•incineration
Hater-borne and high solids
coatings
Ultraviolet curing
Thermal and catalytic
Incineration
Hater-bom «R£ Mtfn colWs
coatings
Hater-borne and high solids
coatings
Powde" (only for fton-
eewented seams)
Therms 1 and catalytic
incineration
Htter-borne and high solids
coatings
Powder (only for ran-
ctnented seams)
Carbon adsorptien
Hater-borne and high solids
coatings
Carbon adsorption
Therms 1 and catalytic
Incineration
Htter. borne and high sol Ids
coatings
Percentage
Reouctior.0
PO-9E
50
70-95
90
60-90
up to 100
90
60-90
100
90
;
90
60-90
pjp to 100
90
€0-90
60-90
100
90
60-90
100
SO
70-95
92
90
60- 90
* Coll coating lines consist of coaters, ovens, and e-ueneh BPSSS.
Sheet, can and end coatinq lines consist o'f sew tors and ovens.
Coroerei tc conventional solvent-based eecfings
add-On controls.
oittcaut eny
16
-------
4.2.3 Magnet Wire Coating^
4.2.3.1 Process Description - Magnet wire coating is the process
of applying a coating of electrically insulating varnish or enamel
to aluminum or copper wire for use in electrical machinery. Wire is
usually coated in large plants that both draw and insulate it; it is
then sold to manufacturers of electrical equipment. The wire coat-
ing must meet rigid specifications of electrical, thermal, and
abrasion resistance.
Figure 4.2-3 shows a typical wire coating operation. The wire
is unwound from spools and passed through an annealing furnace.
Annealing softens the wire and also cleans it by burning off oil
and dirt. Typically, the wire then passes through a bath of coat-
ing at the coating applicator, and is subsequently drawn through an
orifice or coating die that scrapes off the excess. It is then
dried and cured in a two zone oven at about 200 and 430°C (400 and
806°F), respectively. Wire may pass through the coating applicator
and the oven as many as 12 times in order to acquire the necessary
thickness of coating.
4.2.3.2 Emissions and Controls - Emissions from wire coating
operations depend on composition of the coating, thickness of coat,
and efficiency of application. Postapplication chemical changes
and nonsolvent contaminants, such as oven fuel combustion products,
may also affect the composition of emissions.
The exhaust from the oven (1) is the most important source of
solvent emissions in the wire coating plant. Emissions from the
applicator (2) are low in comparison, primarily because a dip
coating technique is used (these numbers refer to Figure 4.2-3).
All solvent used and not recovered can be considered as potential
emissions.
17
-------
00
DRYING
AND
CURING
Figure 4.2-3. Wire Coating Line Emission Points
-------
VOC emissions may be estimated from the factors in Table 4.2-1
if the coating usage is known and it is known that the coater does
not have any control. Since most wire coaters built since 1960 do
have controls, the information in the following paragraph is prob-
ably applicable. Table 4.2-6 gives estimated emission factors for
wire coating.
Incineration is the only control technique commonly used to
control emissions from wire coating operations. Since about 1960,
all major wire coating designers have incorporated catalytic
incinerators into their oven designs because of the economic
benefits. The internal catalytic incinerator burns solvent fumes
and recirculates the heat back into the wire drying zone. Fuel
otherwise needed to operate the oven is eliminated or greatly
reduced, thus resulting in cost savings. Essentially all solvent
emissions from the oven can be directed to an incinerator with a
combustion efficiency of at least 90 percent.
Ultraviolet-cured coatings are available for specialized sys-
tems. Carbon adsorption is not practical for several reasons. Low
solvent coatings are only a potential control because they have not
yet been developed with the properties that will meet all the
industry's needs.
4.2.4 Automobile and Light-Duty Truck Coating
4.2.4.1 Process Description? _ Automobile (passenger vehicles
seating 12 or fewer people) and light-duty truck (8,500 Ib gross
vehicle weight or less) assembly lines usually produce from 30 to 65
units per hour. The assembly industry receives parts from a variety
of sources and produces finished vehicles for dealers, assembling
one general body style in various models on each line.
19
-------
Table 4.2-6. ORGANIC SOLVENT EMISSIONS FROM A
TYPICAL WIRE COATING LINEy
EMISSION FACTOR RATING: B
Typical volatile organic
emissions from coating
line
kg/hr
12
Ib/hr
26
Typical annual organic
emissions^
MT/yr
102
English ton/yr
93
Organic solvent emissions vary from line to line as a
result of size and speed of the wire, number of wires
per oven, and the number of passes through the oven.
A typical line may coat 1,200 pounds of wire per day.
Based upon normal operating conditions of 7,000 hr/yr.
20
-------
Lacquers and enamels are both used in automobile coating.
Most cars are coated with a two part prime. The first coat is
applied by electrodipping in a water-borne coating that is equiva-
lent to about 75 percent solids. The second coat may be water-
borne or an organic-borne enamel with 25 to 35 volume percent
solids. Acrylic coatings (either lacquers or enamels) are widely
used for topcoats. Approximately half of the total production is
topcoated with lacquer finishes, which average about 88 percent
solvent. The balance is coated with an enamel which averages 68
percent solvent.
r Although no "typical" assembly line exists, features common to
all are shown in Figure 4.2-4. The electrodeposited prime coat re-
quires two steps; the first prime coat is applied by dipping, while
the second prime is spray coated. The unit is baked after the
application of each prime coat. The topcoat follows in from one to
three steps, usually with a bake step after each. The painted body
then goes to the trim shop; coatings damaged during the finishing
operation are repainted in a repair spray booth.
4.2.4.2 Emissions and Controls?**0 - Emission points in auto-
mobile coating include (1) the prime application and flashoff areas,
(2) the prime cure ovens, (3) the topcoat application and flashoff
areas, and (4) the topcoat cure ovens (these numbers refer to Figure
4.2-4). The flashoff area is the evaporation area between the
application area and the oven. When using organic solvent-borne
spray, 85 to 90 percent evaporates in the booth and flashoff areas,
and the remaining 10 to 15 percent evaporates in the ovens.
Uncontrolled emissions vary greatly depending on vehicle size,
line speed, coating thickness, and, most importantly, the solvent
content and composition of the coatings. The amount of organic
solvent emitted per vehicle for several coating categories is shown
21
-------
1 0 C '
FROM BODY SHOP
METAL
PRETREATMENT
DRY-OFF OVEN
FIRST PRIME
APPLICATION
ARIA
*-
FIRST PRIME
CURE OVEN
SECOND PRIME
fc APPLICATION
"1 AREA ^-.
SECOND PRIME
CURE OVEN .-
FIRST TOPCOAT
APPLICATION AREA
1
FIRST TOPCOAT
CURE OVEN
»
THIRD TOPCOAT
APPLICATION AREA
(IF ANY) x-v
^
SECOND TOP
APPLICATION
(IF ANY)
COATED »
OT>
THIRD TOPCOAT
CURE OVEN
(IF ANY) ^
T •
REPAIR TOPCOAT
APPLICATION AREA
»-
RE
(LOW
»AIR TOPCOAT
OVEN
TEMPERATURE)
'ARTS FROM
-------
In Table 4.2-7. If total coating usage is known, uncontrolled
emissions can be estimated using the factors from Table 4.2-1. All
solvent used and not recovered can be considered as potential
emissions.
Emissions from most prime application systems have been re-
duced by converting to an electrodeposited water-borne prime. Other
control options for those plants that still prime with organic-borne
coatings include spraying or dipping with a water-borne surfacer
(also called a guidecoat) and using a higher solids (less than 50
volume percent solvent) prime. Incineration and carbon adsorption
may also reduce emissions from solvent-borne primes. Emissions from
the prime cure oven are reduced by using water-borne electrodepo-
sition, low solvent primes, or incineration.
About 66 percent of uncontrolled emissions from automobile
coating lines come from the topcoat application areas. Powder
coatings generally have not been employed in the U.S. because of
the difficulty of changing colors and because metallic coatings
cannot yet be applied satisfactorily by this method. Incineration
and carbon adsorption of spray booth exhaust are technically feas-
ible, and major auto firms are now evaluating these add-on devices
for spray booths. Thus, control options for this major emission
point are currently limited to water-borne or low solvent topcoats.
Topcoat cure oven emissions can be reduced by conversion to water-
borne or low solvent coatings and by incineration of oven exhaust.
Table 4.2-8 gives control efficiencies for auto coating lines.
4.2.5 Other Metal Coating11'13
4.2.5.1 Process Description - Large appliance, metal furni-
ture, and miscellaneous part and product coating lines have most
23
-------
Table 4.2-7.
EMISSIONS FACTORS FOR A TYPICAL PLANT ASSEMBLING INTERMEDIATE-SIZED
AUTOMOBILES AND LIGHT-DUTY TRUCKS?
EMISSION FACTOR RATING: B
Coating Type
(percent solids
by volume)
121 Solution
lacquer
I8X Dispersion
Lacquer
321 Enamel
501 Enamel
Water-Borhe
Prime Application
Solvent
Exhaust Rate
kg/hr
b
b
163
77
negli-
gible
Ib/hr
b
b
360
170
negli-
gible
Solvent
Emissions
kg/car
b
b
2.7
1.3
small
tb/car
b
b
6.0
2.8
small
Prime Oven
Solvent
Exhaust Rate
kg/hr
b
b
29
13.6
18. ?
Ib/hr
b
b
64
30
40
Solvent
Emissions
kg/car
b
b
0.5
0.2
0.3
Ib/car
b
b
1.1
0.5
0.7
Topcoat Application
Solvent
Exhaust Rate
kg/hr
1.129
701
327
154
38.5
Ib/hr
2.490
1.546
720
340
85
Solvent
Emissions
kg/hr
18.8
11.7
5.4
2.6
0.6
Ib/car
41.5
25.8
12.0
5.7
1.9
Topcoat Oven
Solvent
Exhaust Rate
kg/hr
199
96
58
27.2
3R.5
Ib/hr
438
212
127
60
85
Solvent
Emissions
kg/car
3.3
1.6
0.9
0.5
0.6
Ib/car
7.3
3.5
2.1
1.0
1.4
From Reference 7. Assumptions: Hate of as$en*>1y of 60 vehicles per hour. Solvent corresponds to 50-50 mole percent Hexone-Benipne;
55 percent of emissions are In Application Area(s) for Solvent-Borne Coatings; 50 percent of emissions are In Application Area(sJ for
Bater-Borne Topcoat; 30 gal/hr of Solids are applied for Prime Coat; SO pal/hr of Solids are applied for Topcoat; Application Arpa
£»haust ot 100 ppm was 0.0228 tb/10 scf; Oven Enhaust et 10 percent of LEI was 0.296 Ib/IO3 scf; Organic Solvent Density was
S.665 Ib/gal. For Mater-Borne, percent of LEL was lower.
b riot applicable.
-------
Table 4.2-8. CONTROL EFFICIENCIES FOR AUTOMOBILE AND
LIGHT-DUTY TRUCK COATING LINES7
Affected
Facility*
Prime application13
and flashoff area
Prime cure oven
Topcoat application
and flashoff area"
Topcoat cure oven
1
1
1
i
i
Control Option
Hater-borne (electro-
deposition)
50 volume percent solids
prime
Incineration
Carbon adsorption
Water-borne (electro-
deposition)0
>50 volume percent sol Ids
prime
Incineration
Water-borne topcoat0
>50 volume percent sol Ids
topcoat
Incineration ,
Carbon adsorption
Hater-borne topcoat0
>50 volume percent sol Ids
topcoat
Incineration
Percentage
Reduction
80d -93e
Od-€5e
90+
85+
80d-93e
Od -65e
90+
40d'9 -92f'9
Od -86f
90+
85+
40d.9 _Mf .9
Of -86f
90+
These control options are applicable to all assembly and subassembly
lines 1n the plant,1 Including those for frames, small parts, wneels,
and main body parts.
K I
HThe application areas are the areas where the coating 1s applied
by dip or spray. The flashoff area 1s the space between the appli-
cation area and the oven.
options reduce emissions from application, flashoff and cure.
The percentage reduction given Is the reduction from all of these
sources .
Based on an original coating with 50 volume percent solids (3.7
Ibs/gal less water).
eBased on an original coating with 26 -volume percent solids (5.5
Ibs/gal less water). Water-borne surfacer (guidecoat) 1s Included.
Based on an original coating with 12 volume percent sol Ids (6.5
Ibs/gal less water).
'Based on a water-borne topcoat with 2.76 pounds of organic solvent
per gallon of coating minus water (e.g., 25 volume percent solids,
15 volume percent organic solvent and 60 volume percent water).
25
-------
steps in common; these processes also have similar emissions, emis-
sion points, and available control technology. Figure 4.2-5 shows
a typical metal furniture coating line.
Large appliances include doors, cases, lids, panels, and
interior support parts of washers, dryers, ranges, refrigerators,
freezers, water heaters, air conditioners, and other associated
products. Metal furniture includes both outdoor and indoor pieces
manufactured for household and for business and institutional use.
"Miscellaneous parts and products" herein denotes large and small
farm machinery, small appliances, commercial and industrial machin-
ery, fabricated metal products, and other industrial categories that
coat metal under Standard Industrial Classification (SIC) codes 33
to 41.
Large Appliances - The coatings typically applied to large
appliances are epoxy, epoxy-acrylic, or polyester enamels for the
primer or single coat, and acrylic enamel for the topcoat; coatings
containing alkyd resins are also used in some cases. Prime and
interior single coats are applied at about 25 to 36 volume percent
solids; topcoats and exterior single coats are applied 30 to 40
volume .percent. Lacquers may be used to touch up any scratches that
occur during assembly. Coatings contain 2 to 15 solvents; typical
solvents used are esters, ketones, aliphatics, alcohols, aromatics,
ethers, and terpenes.
Small parts are typically dip coated, while flow or spray
coating is used for larger parts. Dip and flow coating are either
performed in an enclosed room vented by a roof fan or emissions are
vented by an exhaust system adjoining"the drain board or tunnel.
Down- or side-draft booths remove overspray and organic vapors from
prime coat spraying. Spray booths are also equipped with dry fil-
ters or a water wash to trap overspray.
26
-------
MOM
MACHINE SHOP
ro
ELECTROSTATIC. OR
CONVENTIONAL AIR OR
AIRLESS STRAY COATING
FLASHOFF
AREA
(OPTIONAL: USED
IN TWO COAT
OPERATION ONLY)
CLEANSING MNC
PRETRCATMINT
HOW COATING
TOPCOAT OR SINGIE
COAT APPLICATION
Figure 4.2-5. Metal Product Coating Line Emission Points
11
-------
Parts may be manually touched up with conventional or airless
spray equipment; they then go to a flashoff area (either open or
tunneled) for about 7 minutes, and are baked in a multipass oven
for about 20 minutes at 180 to 230°C (350 to 450°F). At this
point, exterior large appliance parts go on to the topcoat applica-
tion area, while the"single coated interior parts are moved to the
assembly area of the plant.
The topcoat, as well as some primes, are applied by automated
electrostatic disc, bell, or other types of spray equipment. Top-
coats usually consist of several colors, which are changed by auto-
matically flushing out the system with solvent. Both the topcoat
and touchup spray booths are designed with side- or down-draft
exhaust control. The parts go through about a 10 minute flashoff
period, followed by baking in a multipass oven for 20 to 30 minutes
at 140 to 180°C (270 to 350°F).
Metal Furniture - Most metal furniture coatings are enamels,
although some lacquers and metallic coatings are also used. The
most common coatings are alkyds, epoxies, and acrylics, which con-
tain the same solvents used in large appliance coatings and are
applied at about 25 to 35 percent solids.
On a typical metal furniture coating line (refer to Figure
4.2-5), the prime coat can be applied with the same methods used
for large appliances, but may be cured at slightly lower tempera-
tures, i.e., 160 to 200°C (300 to 400°F). The topcoat, or usually
just a single coat, is applied electrostatically or with conven-
tional airless or air spray methods, especially if frequent color
changes are required. Most spray coating is done manually, .in con-
trast to large appliance operations. Flow coating or dip coating is
employed if the plant generally uses only one or two colors on its
products.
28
-------
The coated furniture is usually baked, but in some cases it is
air dried. If the furniture is to be baked, it passes through a
flashoff area into a multizone oven at temperatures ranging from 160
to 230°C (300 to 450°F).
/
Miscellaneous Metal Parts and Products - Both enamels (30 to
40 volume percent solids) and lacquers (10 to 20 volume percent
solids) are used to coat miscellaneous metal parts and products,
although enamels are more commonly used. Coatings are often pur-
chased at higher volume percent solids but thinned prior to appli-
cation (frequently with aromatic blends). Alkyds are popular among
industrial and farm machinery manufacturers. Most of the coatings
contain several (up to 10) different solvents, including ketones,
esters, alcohols, aliphatics, ethers, aromatics, and terpenes.
Coatings are applied in both conveyorized and batch, single and
two coat operations. Spraying is usually employed for single coats,
but flow and dip coating may be used when only one or two colors are
applied. For two coat operations, primers are typically applied by
flow or dip coating, while topcoats are almost always applied by
spraying. Electrostatic spraying is common. Spray booths are
maintained at a slight negative pressure to capture overspray.
A manual two coat operation may be used for large items, such
as industrial and farm machinery. The coatings on such products
are air dried rather than oven baked because the machinery, which
is completely assembled, includes heat sensitive materials and may
be too large to be cured in an oven. However, smaller miscellan-
eous parts and products are baked in single or multipass ovens at
150 to 230°C (275 to 450°F).
4.2.5.2 Emissions and Controls - Volatile organic compounds
(VOC) are emitted from (1) the application and flashoff areas and
29
-------
(2) the ovens of metal coating lines (these numbers refer to Figure
4.2-5). The composition of emissions varies among coating lines
according to construction, coating method, and type of coating
applied, but the distribution of emissions among individual opera-
tions is fairly constant regardless of the specific product coated,
as Table 4.2-9 indicates. All solvent used and not recovered can be
considered to be potential emissions.. Emissions can be calculated
from the factors in Table 4.2-1 if coatings usage is known, or from
the factors in Table 4.2-9 if only a general description of the
plant is available. For emissions from the cleansing and pretreat-
ment area, see Section 4.6, Solvent Degreasing.
Table 4.2-9. EMISSION FACTORS FOR TYPICAL UNCONTROLLED METAL COATING PLANTS11"13
EMISSION FACTOR RATING: B
Type of Plant
Urge Appliances -
Prime and topcoats, spray
fetal Furniture -
Single spray coat
Single dip coat
Miscellaneous fetal -
Conveyori zed single flow coat
Conveyor i zed single dip-coat
Conveyori zed single spray-coat
Conveyorl zed two-coat, flow and spray
Conveyorized two-coat, dip and spray
Conveyorl zed two-coat , ' spray
Manual two-coat, spray and air dry
Production Rate
768,000 un1ts/yr
c o
48 x 10? ft:/yr
23 x 10° ft /yr
£ •)
16 x 10° ftf/yr
16 x 10? fti/yr
16 x 10? ft£/yr
16 x 10? ft,/yr
16 x 10? ft£/yr
16 x 10? ftt/yr
8.5 x 10e ftVyr
Cfn« c c i nn c
con 99 tuna
HT/yr
630
355
HO
86
86
103
190
190
206
30
English
ton/yr
571
322
127
78
78
93
172
172
167
27
Estimated Percentage
of Emissions From
Application
and Flasnoff
BO
65 - 80
50 - 60
50 - 60
40 - 50
70 - 80
60 - 70
60 - 70
70 - 80
100
Ovens
20
20 - 35
40 - 50
40 - 50
50 - 60
20 - 30
30 - 40
30 - 40
20 - 30
0
Powder coatings, which contain almost no VOC, can be applied
to some metal products as a modification in coating formulation to
reduce emissions. Powder coatings are applied as single coats on
some interior large appliance parts and as topcoats for ranges;
30
-------
they are also used on metal bed and chair frames, shelving, and
stadium seating, and have been applied as single coats on small
appliances, small farm machinery, fabricated metal product parts,
and industrial machinery component parts. The usual application
method is manual or automatic electrostatic spraying.
An additional control consists of electrostatic coating of
metal products. Many large appliance manufacturers are now using
electrodeposition to apply the prime coat to exterior parts and the
single coat to interior parts because this technique increases
corrosion protection and detergent resistance. Electrodeposition
of water-borne coatings is also being used at several metal furni-
ture coating plants and at some farm and commercial machinery and
fabricated metal product facilities. Water-borne coatings may be
applied with conventional spray, dip, and flow coating equipment as
wel 1.
Low solvent coatings (45 to 50 volume percent solids) are
being applied as topcoats on some refrigerators, thereby reducing
solvent input and thus VOC emissions. Automated electrostatic
spraying is most efficient, but manual and conventional techniques
can also be used. Roll coating is another option on some
miscellaneous parts.
Carbon adsorption is technically feasible for collecting
emissions from the prime, top, and single coat application and
flashoff areas, but there are no known installations to date on
metal coating lines.
Incineration has been used to reduce organic vapor emissions
from large appliance, metal furniture, and miscellaneous product
baking ovens. If sufficient heat recovery can be used to reduce
fuel consumption, incinerators are also an option for controlling
application and flashoff area emissions.
31
-------
Additional control techniques include trapping overspray by
means of dry filters or a water wash in the spray booths and.enclos-
ing flashoff areas. Air inlet velocities required to prevent efflu-
ent from spilling into the work areas from such openings can be
minimized by using air curtains at these openings.
Table 4.2-10 gives estimated control efficiencies for large
appliance, metal furniture, and miscellaneous metal part and pro-
duct coating lines.
Table 4.2-10. ESTIMATED CONTROL TECHNOLOGY
EFFICIENCIES FOR METAL COATING LINES3
Control Technology
fteder
ltiter-bon»
(spray, elp. or
flOCOat)
tater.borne
(elect reesposit ton)
HIcRjr solid!
(tprej)
Carter.
adsorption
lnct£*ratton
Application
Urge
Appliances
Top. Mterlor,
or interior
ttno.lt CMt
*!) applica-
tion!
rV«»« or
interior
tingle coat
loo or
oiterior
•Ingle coat
•no sound
•aadnnir
•riot, tingle.
or topcoat
tppl ication
•no flasnoff
arm
r>tat, toe.
or tingle
coat ovens
Metal
furniture
lop or tingle
coat
Prix, top,
v tingle
coat
»1e» or
tingle
coat
lop or
ttngte
coat
Prle*. too,
or tingle
coat appll-
cation and
flatlwff
areas
Oveni
Klscelleneous
Ooen baked ttngU
coat or toccoat
(Krer. bated tingle
coal, prior, and
topcoat; air dried
priacr and topcoat
Oren baked tingle
coat and prtcer
Oven bokrt tingle
coat and topcoat;
air dried onaer
and toscoal
Oven hated tingle
coat, priier. ond
topcoat applica-
tion and flainoM
areas; air dried
prteer ond top-
coai application
and drying areas
toem
Meductior in OrunU Iwisslons loertent)
Large
Appliances
«5-99b
ro-w1"
CB-9511
aw»"
tf
90s
Heul
Furniture
9S-9S"
to-»c6
TO-95"
Sv-BO11
Wc
Wc
RtsctlUfWout
».9Bd
oO-W*
"3-95S
SO-8C11
Mc
«P-C
• Frco fteforwcts 11-13.
fc Ha base cote tgatntt Bhlrt ttete percent raouctlons »*re calcualtod it a nlo» organic tolvent coating
micr. contains » voluw percent tollos and n volm percent or».n1c tolrants. 1« J""|e'' Jl^t
cles tor llould eoattnjs wm assuw! to be (bout 80 percent for tpre, and CO p?ro«it for dip or flcccont.
for RMOers about 93 percent, and )0' electrooeposUton W furcent.
e ttiis percent reduction In WC tvltilont is on!. »crots Hie amtrel tevlce. ««! toss rat take into account
t«e captur* »ffic1*nc».
* These figures reflect only ttie range In reduction possible. T»* ecuul redsctlon ornlewd dswnas on the
COKSosltlons of the co«»e«it10Ml coating orlqlnjllj used and'tfte reoUccia>M lea organic tolwnl coating,
on transfer efficiency, and on tfte relative fllo thlcknsttes of tM teo ccotlKjs.
32
-------
4.2.6 Flat Wood Interior Panel Coating
4.2.6.1 Process Description1^ - Prefinished flat wood construc-
tion products included in this subcategory of industrial surface
coating are interior panel ings made of hardwood plywood (natural and
lauan), particleboard, and hardboard.
Less than 25 percent of the manufacturers of such flat wood
products coat the products in their plants, and in some of the
plants that do coat, only a small percentage of the total production
capacity is coated. At present, most coating is done by toll coat-
ers who receive panels from manufacturers and prefinish or finish
them according to customer specifications and product requirements.
Some of the layers in any given species of coatings that can be
factory applied to flat woods are filler, sealer, groove coat, pri-
mer, stain, basecoat, inks, and topcoat. The solvents used in
organic-based flat wood coatings are normally multicomponent mix-
tures, including methyl ethyl ketone, methyl isobutyl ketone,
toluene, xylene, butyl acetates, propanol, ethanol, butanol, VM&P
naphtha, methanol, amyl acetate, mineral spirits, SoCal I and II,
glycols, and glycol ethers. Those most often used in water-borne
coatings are glycol, glycol ethers, propanol, and butanol.
Different forms of roll coating are the preferred techniques
for applying coatings to flat woods. Coatings used for surface
coverage can be applied with a direct roller coater, while reverse
roll coaters are generally used to apply fillers in order to force
the filler into panel cracks and voids. Precision coating and
printing (usually with offset gravure grain printers) are also forms
of roll coating, and several types of curtain coating may be em-
ployed as well (usually for application of the topcoat). Various
spray techniques and brush coating may be utilized, too.
Printed interior panel ings are produced from plywoods with
hardwood surfaces (primarily lauan) and from various wood composi-
33
-------
tion panels, including hardboard and particlaboard. Finishing tech-
niques are used primarily to cover the original surface; they also
serve to produce various decorative effects. Figure 4.2-6 is a gen-
eral flow diagram showing some but not all typical production line
variations for printed interior paneling.
Groove coatings, which can be applied in different ways and at
different points in the coating procedure, are usually pigmented,
low resin solids that are reduced with water prior to use and so
yield few if any emissions. Fillers, which are normally applied by
reverse roll coating, may be of various formulations: (1) polyester
(which is ultraviolet cured), (2) water-based, (3) lacquer-based,
(4) polyurethane, and (5) alkyd urea-based. Water-based fillers are
in common use on printed paneling lines.
Sealers may be water- or solvent-based, and are usually applied
by airless spray or direct roll coating, respectively. Basecoats,
which are usually direct roll coated, generally fall into the fol-
lowing categories: lacquer, synthetic, vinyl, modified alkyd urea,
catalyzed vinyl, and water-based (which are now used at some lauan
finishing plants).
Inks are applied by an offset gravure printing operation simi-
lar to direct roll coating. Most lauan printing inks are pigments
dispersed in alkyd resin with some nitrocellulose added for better
wipe and printability. Water-based inks have a good future for
clarity, cost, and ecological reasons. After printing, the board
goes through one or two direct or precision roll coaters for appli-
cation of the clear, protective topcoat. Some topcoats are synthet-
ic, prepared from solvent soluble alkyd or polyester resins, urea
formaldehyde cross linkings, resins, and solvents.
Natural hardwood plywood panels are coated with transparent or
clear finishes to enhance and protect their face ply of hardwood
veneer. Typical production lines are similar to those for printed
34
-------
TEMPERING
FEEKR
I J
BRUSHER
CUT
GROOVE
FIUER
(RRC)
(l)
OVEN
Cv)
SANDER
SEAIER
OR FIRST
BASECOAf
(DRC OR
SPRAT)
OVEN
COOIINQ
OJ
IRKS
(orrsET
GRAWJRt)
TOPCOAT
(ORC)
CO
INKS
(orrstT
GRAVURE)
OVEN
FIRST OR
SECOND
BASFCOAT
(ORC)
(f)
SANDER
PACKAGING
SHIPMENT
RRC • Reverie roll CM ting
ORC • Direct roll coiling
Figure 4.2-6. Flatwood Interior Panel Coating Line Emission Points
14
-------
interior paneling, except that a primer sealer is applied to the
filled panel, normally by direct roll coating; the panel is then em-
bossed and valley printed to give a distressed or antique appearance
(no basecoat is required). A sealer is also applied following
printing but prior to application of the topcoat, which may be cur-
tain coated (although direct roll coating remains the usual
technique).
4.2.6.2 Emissions and Controls8*14 - Emissions of volatile
organic compounds (VOC) at flat wood coating plants occur primarily
from (1) reverse roll coating of filler, (2) direct roll coating of
sealer and basecoat, (3) printing of wood grain patterns, (4) direct
roll or curtain coating of topcoat(s), and (5) oven drying after one
or more of these operations (these numbers refer to Figure 4.2-6).
All solvent used and not recovered can be considered as potential
emissions. Emissions can ba calculated from the factors in Table
4.2-1 if the coating usage is known. Emissions for interior printed
panels can be estimated from the factors in Table 4»2-ll if the
total area of coated panels is known.
Water-borne coatings, as a change in process materials to
reduce emissions, are increasing in usage and can be applied to
almost all flat wood, with the exception of redwood and possibly
cedar. The major use of water-borne flat wood coatings is in the
filler and basecoat applied to printed interior paneling. Limited
use has been made of water-borne materials for inks, groove coats,
and topcoats for printed paneling, and for inks and groove coats for
natural hardwood panels.
-Ultraviolet curing systems are applicable for clear to semi-
transparent fillers, topcoats on particleboard coating lines, and in
specialty coating operations. Polyester, acrylic, urethane, and
alkyd coatings can be cured by this method.,
36
-------
Table 4.2-11. VOC EMISSION FACTORS FOR INTERIOR PRINTED PANELS14
EMISSION FACTOR RATING: B
Paint
Category
Filler
Seal er
Basecoat
Ink
Topcoat
Total
Coverage3
liter/100 m2
Water-
borne
6.5
1.4
2.6
0.4
2.6
13.7
Conven-
tional
Paint
6.9
1.2
3.2
0.4
2.8
14.7
gal /I, 000 ft2
Water-
borne
1.6
0.35
0.65
0.1
0.65
3.4
Conven-
tional
Paint
1.7
0.3
0.8
0.1
0.7
3.6
Uncontrolled VOC Emissions
kg/100 m2 coated
Water-
borne
0.3
0.2
0.2
0.1
0.4
1.2
Conven-
tional
Paint
3.0
0.54
2.4
0.3
1.8
8.0
Ultra-.
v1oletb
nil
0
0.24
0.10
nil
0.4
lb/1,000 ft2 coated
Water-
borne
0.6
0.4
0.5
0.2
0.9
2.6
Conven-
tional
Paint
6.1
1.1
5.0
0.6
3.7
16.5
Ultra-
violet"
nil
0
0.5
0.2
nil
0.8
a Paint coverage based on information by Reference 1 from Abitlbi Corp., Cucamonga,. CA.
Adjustments between water and conventional paints were made using typical nonvolatiles content.
UV line uses no sealer, uses water-borne basecoat and ink. Total is adjusted to cover
potential emissions from the UV coatings.
-------
Afterburners can be used to control VOC emissions from baking
ovens, and there would seem to be ample opportunity for using
recovered heat. Extremely few flat wood coating operations have
afterburners as add-on controls, however, despite the fact that they
are a viable control option for reducing emissions where other con-
trol techniques are not applicable due to product requirements.
Carbon adsorption is technically feasible, especially for
specific applications (e.g., in redwood surface treatment), but
multicomponent solvents and the use of different coating formula-
tions in the several steps along the coating line have thus far
precluded its use to control flat wood coating emissions by reclaim-
ing solvents. The use of low solvent coatings to fill pores and
seal wood has been demonstrated, but they do not appear practicable
for current use in the flat wood coating industry. Costs of both
the installed system and the coating itself limit the applicability
of electron beam (EB) curing as a control technique.
4.2.7 Paper Coating
4.2.7.1 Process Description^*' - Paper is coated for a vari-
ety of decorative and functional purposes using water-borne, organic
sol vent-borne, or solventless extrusion type materials. Paper
coating is not to be confused with printing operations, which use
contrast coatings that must show a difference in brightness with the
paper to be visible and thus convey their message to the observer.
Moreover, inks are always applied with a printing machine.
Water-borne coatings improve printability and gloss, but cannot
compete with organic solvent-borne coatings in terms of weather,
scuff, and chemical resistance. Solvent-borne coatings have the
38
-------
added advantage of permitting a wide range of surface textures.
Most solvent-borne coating is done by paper converting companies
that buy paper from mills and apply coatings to produce a final
product. Among the products that are coated using solvents are
adhesive tapes and labels, decorated paper, book covers, zinc oxide-
coated office copier paper, carbon paper, typewriter ribbons, and
photographic films.
Organic solvent formulations generally used are made up of
film-forming materials, plasticizers, pigments, and solvents. The
main classes of film formers used in paper coating are cellulose
derivatives (usually nitrocellulose) and vinyl resins (usually the
copolymer of vinyl chloride and vinyl acetate). Three common plas-
ticizers are dioctyl phthalate, tricresyl phosphate, and castor oil.
The major solvents used are toluene, xylene, methyl ethyl ketone,
isopropyl alcohol, methanol, acetone, and ethanol. Although a
single solvent is frequently used, a solvent mixture is often neces-
sary to obtain the optimum drying rate, flexibility, toughness, and
abrasion resistance.
A variety of low solvent coatings with negligible emissions
have been developed for some uses that can form organic resin films
equal to those of conventional solvent-borne coatings. They can be
applied up to 1/8-inch thick (usually by reverse roller coating) to
make products such as artificial leather goods, book covers, and
carbon paper. Smooth hot melt finishes can be applied over rough
textured paper by heated gravure or roll coaters at temperatures
from 60 to 230°C (150 to 450°F).
Plastic extrusion coating is a type of hot melt coating in
which a molten thermoplastic sheet (usually low or medium density
polyethylene) is extruded from a from a slotted dye at temperatures
of up to 315°C (600°F). The moving substrate and the molten plas-
tic are combined in a nip between a rubber roll and a chill roll.
39
-------
Hundreds of products are coated with solvent!ess extrusion coatings;
an example is the polyethylene-coated milk carton.
Figure 4.2-7 shows a typical paper coating line that uses
organic solvent-borne formulations. The application device is
usually a reverse roller (as illustrated), a knife (refer to inset),
or a rotogravure printer. A knife coater can apply solutions of
much higher viscosity than roll coaters, thus emitting less solvent
per pound of solids applied. The gravure printer can print patterns
or a solid sheet of color on a paper web.
Ovens may be divided into two to five temperature zones. The
first zone is usually at about 43°C (110°F); other zones have pro-
gressively higher temperatures that cure the coating after most of
the solvent has evaporated. The typical curing temperature is 120°C
(250°F); ovens are generally limited to operating temperatures of
200°C (400°F) to avoid damage to most types of paper. Although
natural gas is the fuel most often used for direct-fired ovens, fuel
oil is sometimes used. Some of the heavier grades of fuel oil
can create problems because S0£ and particulate may contaminate
the paper coating. Distillate fuel oil usually can be used satis-
factorily. Steam produced from burning the solvent that is re-
trieved from an adsorber or vented to an incinerator may also be
used to heat curing ovens.
4.2.7.2 Emissions and Controls7 - The main emission points
from paper coating lines are (1) the coating applicator and (2) the
oven (these numbers refer to Figure 4.2-7). In a typical paper
coating plant, about 70 percent of all solvents used are emitted
from the coating lines; most of these come from the first zone of
the oven. The other 30 percent are emitted from solvent transfer,
storage, and mixing operations and can be reduced through good
40
-------
KNIFE COATING
HEATED AIR FROM BURNER,
BOILER, OR HEAT_
RECOVERY
REVERSE ROLL COATER
MAJOR
EMISSION
POINT
ZONE 1
EXHAUST
ZONE 2
EXHAUST
HOT AIR NOZZLES
TENSION ROLLS
REWIND
Figure 4.2-7. Paper Coating Line Emission Points
-------
housekeeping practices. All solvent used and not recovered can be
considered as potential emissions.
VOC emissions from individual paper coating plants vary with
the size and number of coating lines, line construction, coating
formulation, and substrate composition, and must thus be evaluated
on a case-by-case basis. VOC emissions can be estimated from the
factors in Table 4.2-1 if coating usage is known and sufficient
information on coating composition is available. Since many paper
coatings are proprietary, it may be necessary to have the user
supply information on the total solvent used and employ the factor
from Table 4.2-12 to estimate emissions.
Almost all solvent emissions from the coating lines can be
collected and sent to a control device. Thermal incinerators have
been retrofitted to a large number of oven exhausts, with primary
and even secondary heat recovery systems heating the ovens. Carbon
adsorption is most adaptable to lines using single solvent coatings,
although solvent mixtures collected by adsorbers can be subsequently
distilled for reuse.
Although applicable for certain types of products, low solvent
coatings are not yet available for all paper coating operations.
The nature of some products, such as photographic films, may limit
their adaptability to lower solvent coatings or to efficient add-on
controls.
Table 4.2-13 lists efficiencies of several control devices.
42
-------
Table 4.2-12. SOLVENT EMISSIONS FROM UNCONTROLLED PAPER COATING LINES
EMISSION FACTOR RATING: D
Solvent Emissions
kg/MT solvent used3
- - 1,000
Ib/ton solvent used3
2,000
a composition of coatings is usually proprietary.
Table 4.2-13. CONTROL EFFICIENCIES FOR PAPER COATING LINES7
Affected Facility
Coating line
Control Technique
Incineration
Carbon adsorption
Low solvent coatings
Percentage Reduction
95
90+
80-993
a Based on comparison with a conventional coating containing
35 percent solids and 65 percent organic solvent by volume.
43
-------
4.2.8 Fabric Coating
4.2.8.1 Process Description^'1^ - Fabric coating consists of
coating a textile substrate to impart properties such as strength
stability, water or acid repel!ancy, or appearance. This includes
vinyl coating, which refers to any printing or decorative or pro-
tective topcoat applied over vinyl coated fabric or vinyl sheets,
but does not include the application of the vinyl plastisol to the
fabric, which produces negligible emissions.
Products that involve fabric coating include rainwear, tents,
industrial and electrical tapes, tire cord, seals and gaskets, imita-
tion leathers, shoe material, and upholstery fabrics. The industry
is comprised mostly of small to moderate sized plants, many of which
are toll coaters rather than specialists in a particular product
line.
Figure 4.2-8 is a flow chart of a typical fabric coating opera-
tion. If the fabric is to be coated with rubber, the rubber is
milled with pigments, curing agents, and fillers prior to being dis-
solved (mixed) in a suitable solvent; when nonrubber coatings are
employed, milling is rarely necessary.
Coatings are normally applied by means of a knife, a roller, or
a rotogravure printer. In addition, U.S. plants have recently begun
to apply heat-transfer printing to textiles by the gravure process:
the pattern is applied to dry-transfer paper, then heated to volati-
lize it onto the fabric. Flat and rotary screen printing processes
are also used to print textiles. Screen printing employs a fine
screen for the image area, through which ink or paint is forced; non-
image areas are produced by coating the screen to mask off the ink.
Water-based emulsions and, more frequently, resin-bonded pigments
are used with fabric screen printing.
44
-------
RUBBER
PIGMENTS
I
.CURING AGENTS SOLVENT
MILLING
1
MIXING
(T)
in
DRYING AND
CURING
OVENS
(T)
COATING
APPLICATION
(KNIFE, ROLLER
OR ROTOGRAVURE)
(2)
FABRIC
COATED PRODUCT
Figure 4.2-8. Fabric Coating Line Emission Points7
-------
4.2.8.2 Emissions and Controls7 - The VOC emissions in a fabric
coating plant occur at (1) the mixer, (2) the coating applicator,
and (3) the oven (these numbers refer to Figure 4.2-8). Emissions
from the coating process depend upon the formulation employed, and
all solvent used and not recovered can be considered as potential -
emissions.
VOC emissions can be estimated from the factors in Table 4.2-1
if coating usage is known and sufficient information on coating com-
position is available. Because many fabric coatings are proprie-
tary, it may be necessary to have the user supply information on the
total solvent used and employ the factor from Table 4.2-1 to estimate
emissions.
Sometimes only small emissions occur at the mixer, but some
vinyl coaters estimate that as much as 25 percent of all solvents
used in the plant are emitted in mixing operations. Solvent emis-
sions .from the coating applicator account for 25 to 35 percent of all
solvent emitted from the coating line, with rotogravure being a large
source of such emissions compared to knife or roll coating. The
major emission point, however, is the drying and curing ovens, at 65
to 75 percent of all coating line losses. Some plants report that
over 70 percent of solvents used within the plant are emitted from
the coating applicators and ovens; other plants, especially vinyl
coating ones, report that only 40 to 60 percent of solvents purchased
are emitted from the coating line. The remaining percentage of sol-
vent used is lost as fugitive emissions.
Fugitive losses result from solvent transfer, storage tank
breathing losses, agitation of mixing tanks, waste solvent disposal,
various stages of cleanup, and evaporation from the coated fabric
after it leaves the line. Controls include the use of tightly fit-
ting covers for open tanks, collection hoods for cleanup areas, and
closed containers for solvent wiping cloths.
46
-------
Ovens heated by natural gas, steam, or electricity are used to
dry and cure the coated fabric. Oven evaporation rates are often
controlled to produce desired chemical changes in the coating solids
(curing). In most ovens, almost all solvent emissions are captured
and vented with exhaust gases; coating applicator emissions may also
be ducted to the oven and included with the oven exhaust.
Incineration is probably the most viable control option for
coating application and curing on commission coating lines, which use
a variety of coating formulations to comply with customer specifica-
tions. The primary and secondary heat recovery and higher solvent
concentrations in exhaust gases that are results of this control
option help reduce the fuel requirements of the coating process.
As with other surface coating operations, carbon adsorption is
most applicable to sources which use a single solvent blend that can
be recovered for reuse.
Where high solids or water-borne coatings have been developed
that can replace conventional coatings, their use may preclude the
need for a control device.
Table 4.2-14. SOLVENT EMISSIONS FROM UNCONTROLLED
FABRIC COATING PLANTS
EMISSION FACTOR RATING: D
Solvent Emissions
kg/MT solvent used3
1,000
Ib/ton
solvent used3
2,000
a Composition of coatings is usually proprietary.
47
-------
References for Section 4.2
1. Products Finishing Directory, Volume 41, No. 6A. 600 Main
Street, Cincinnati, Ohio, March 1977, pp. 4-54.
2. Controlling Pollution from the Manufacturing and Coating of
Metal Products: Metal Coating Air Pollution Control-1. EPA-
625/3-77-009. U.S. Environmental Protection Agency, Environ-
mental Research Information Center, Technology Transfer, May
1977, Chapter 2.
3. Powers, H.R., "Economic and Energy Savings Through Coating
Selection." The Sherwin-Williams Company, Chicago, Illinois
60628, February 8, 1978.
4. Danielson, J.A., ed. Air Pollution Engineering Manual, Second
Edition, AP-40. U.S. Environmental Protection Agency,
Research Triangle Park, N.C. 27711, May 1S73, p. 864.
5. Hughes, T.W.; Horn, D.A.; Sandy, C.W.; and Serth, R.W. Source
Assessment: Prioritization of Air Pollution from Industrial
Surface Coating Operations. EPA-650/2-75-019a. Monsanto
Research Corporation, Dayton, Ohio. Prepared for U.S.
Environmental Protection. Agency, Research Triangle Park, N.C.
27711 under Contract No. 68-02-1320, February 1975.
6. Control of Volatile Organic Emissions from Existing Stationary
Sources - Volume I: Control Methods for Surface Coating
Operations. EPA-450/2-76-028 (OAQPS No. 1.2-067). U.S.
Environmental Protection Agency, Research Triangle Park, N.C.
27711, November 1976.
7. Control of Volatile Organic Emissions from Existing Stationary
Sources - Volume II: Surface Coating of Cans, Coils, Paper
Fabrics, Automobiles, and Light-Duty Trucks. EPA-450/2-77-008
(OAQPS No. 1.2-073). U.S. Environmental Protection Agency,
Research Triangle Park, N.C. 27711, May 1977.
8. Air Pollution Control Technology Applicable to 26 Sources of
Volatile Organic Compounds. Emission Standards and Engineering
Division, OAQPS, U.S. Environmental Protection Agency, Research
Triangle Park, N.C. 27711, May 27, 1977.
9. Control of Volatile Organic Emissions from Existing Stationary
Sources - Volume IV: Surface Coating for Insulation of
Magnet Wire. EPA-450/2-77-033 (OAQPS No. 1.2-087). U.S.
Environmental Protection Agency, Research Triangle Park, N.C.
27711, December 1977.
48
-------
10. Controlled and Uncontrolled Emission Rates and Applicable
Limitations for Eighty Processes. TRC of New England,
Wethersfield, Conn. 06109. Prepared for U.S. Environmental
Protection Agency, Research Triangle Park, N.C. 27711, under
Contract No. 68-02-1382', Task Order 12, September 1976, pp.
XI-l-XI-3.
11. Control of Volatile Organic Emissions from Existing Stationary
Sources - Volume III: Surface Coating of Metal Furniture.
EPA-450/2-77-032 (OAQPS No. 1.2-086). U.S. Environmental
Protection Agency, Research Triangle Park, N.C. 27711,
December 1977.
12. Control of Volatile Organic Emissions from Existing Stationary
Sources - Volume V: Surface Coating of Large Appliances.
EPA-450/2-77-034 (OAQPS No. 1.2-088). U.S. Environmental
Protection Agency, Research Triangle Park, N.C. 27711,
December 1977.
13. Control of Volatile Organic Emissions From Existing Stationary
Sources - Volume VI: Surface Coating of Miscellaneous Metal
Parts and Products. EPA-450/2-78-015 (OAQPS No. 1.2-101).
U.S. Environmental Protection Agency, Research Triangle Park,
N.C. 27711, June 1978.
14. Control of Volatile Organic Emissions from Existing Stationary
Sources - Volume VII: Factory Surface Coating of Flat Wood
Interior Paneling. EPA-450/2-78-032 (OAQPS No. 1.2-112).
U.S. Environmental Protection Agency, Research Triangle Park,
N.C. 27711, June 1978.
15. Carpenter, B.H. and Milliard, G.K. "Environmental Aspects of
Chemical Use in Printing Operations." EPA-560/1-75-005
(PB 251406). U.S. Environmental Protection Agency, Washington,
D.C. 20460, January 1976, pp. 5-30.
49
-------
BACKGROUND DOCUMENT
SECTION 4.2 INDUSTRIAL SURFACE COATING
1.0 INTRODUCTION
The section on Industrial Surface Coating is organized as eight
separate sections, as follows:
4.2.1 General
4.2.2 Coil and Can Coating
4.2.3 Magnet Wire Coating
4.2.4 Auto and Light-Duty Truck Coating
4.2.5 Other Metal Coating
4.2.6 Flat Wood Interior Panel Coating
4.2.7 Paper Coating
4.2.8 Fabric Coating
The section titles correspond to several of the subjects in a
series of OAQPS Guideline Documents for the control of volatile
organic emissions from existing stationary sources (References 6,
7, 9 and 11-14). These documents were used as primary information
sources for the preparation of this section.
2.0 GENERAL SURFACE COATING (Tables 4.2-1 to 4.2-3) .
Section 4.2.1 contains no emission factors per se. It does
contain information on the densities and volatile organic contents
of coatings and formulas for using this information to compute VOC
emissions on the assumption that all the VOC "in a coating are
evaporated. No claims are made for the "representativeness" of
these compositional data. They are quoted directly from AP-40,
Second ed. (Reference 4, p. 864), which is assumed to be
authoritative and accurate.
50
-------
3.0 COIL AND CAN COATING (Tables 4.2-4 and 4.2-5)
The emissions from can coating that are summarized in Table
4.2-4 of Section 4.2.2 are taken directly from Table 2-2, page 2-15
of Volume II of the EPA's Control of Volatile Organic Emissions From
Existing Stationary Sources (Reference 7).
The data contained in Table 4.2-5 are taken from pages 2-1 and
2-2 of Reference 7. The author of Reference 7 estimated emissions
and control efficiencies from data contained in trip reports
prepared by EPA personnel V.N. Gallagher and W.L. Johnson. The data
presumably represent typical values based on the best engineering
judgments of the authors of the reports.
Coil and can coating were considered together because they both
involve the industrial surface coating of metal sheets or webs on
coating lines that usually consist of roll coaters and drying
ovens.
4.0 MAGNET WIRE COATING (Table 4.2-6)
The wire coating emissions presented in Table 4.2-6 are taken
directly from Table 3-1, page 3-3 of Reference 9. Tons per year
were computed from the operating factor of 7,000 hr/yr given in
Table 3-1.
5.0 AUTOMOBILE AND LIGHT-DUTY TRUCK COATING (Tables 4.2-7 and
4.2-8)
The emissions data presented in Table 4.2-7 of Section 4.2.4
are taken directly from Reference 7, Figure 6.3, page 6-15, and the
control efficiencies given in Table 4.2-8 are taken from page 6-1 of
Reference 7.
51
-------
No specific sources for individual data items are given in
Reference 7, but nearly all data sources for this document are
listed as comments from industry or as trip reports from EPA per-
sonnel. Presumably the data in Tables 4.2-7 and 4.2-8 are derived
from the industry comments, the EPA trip reports, and the best
engineering judgment of the authors of Reference 7.
Auto refinishing/topcoat repair is not included in this section
because repair production is intermittent and generally limited to
manual spraying with solvent-borne coatings that can be dried at
temperatures low enough for the trim to tolerate.
6.0 OTHER METAL COATING (Tables 4.2-9 and 4.2-10)
The emission factors in Table 4.2-9 were computed from infor-
mation in Reference 11, Table 3-1, Reference 12, Tables 3-3 an 3-4,
and Reference 13, Tables 3-3 to 3-9. In each case, the reference
gave information about the number of tons of VOC removed by con-
trols and also gave the percentage reduction based on uncontrolled
emissions. The tons of uncontrolled VOC were computed from these
two data items. The production rates are those given in References
11, 12, and 13 for model plants. No references or other justifica-
tion are given for the technical parameters listed for these
plants, but presumably they represent the best engineering judgment
of the authors of References 11, 12, and 13.
'Table 4.2-10 is compiled from data taken directly from
Reference 11 (page'2-1), Reference 12 (page 2-1), and Reference 13
(page 2-1). No specific documentation is given for any of these
data, but, in general, they are based on trip reports and comments
from industry.
52
-------
7.0 FLAT WOOD INTERIOR PANEL COATING (Table 4.2-11)
The information in Tables 4.2-11 of Section 4.2.6 is taken
directly from page 2-4 of Volume VII of the series cited above
(Reference 14), which, in turn, is based on information supplied by
industry.
8.0 PAPER COATING (Tables 4.2-12 and 4.2-13)
The information in Table 4.2-12 of Section 4.2.7 is based on
the premise of mass balance, and the information in Table 4.2-13 is
taken directly from page 5-1 of Reference 7. The primary sources
for these data are not given in Reference 7.
9.0 FABRIC COATING (Table 4.2-14)
The information in Table 4.2-14 of Section 4.2.8 is based on
the premise of mass balance. No estimates of emissions from fabric
coating lines were presented in the reference documents on which
this section is based (i.e., References 7 and 15; the latter
consists of one chapter of EPA's "Environmental Aspects of Chemical
Use In Printing Operations").
53
-------
4.4 TRANSPORTATION AND MARKETING
OF PETROLEUM LIQUIDS'
4.4.1 Process Description
As Figure 4.4-1 indie. • the tr-^portation and marketing of petroleum liquids involves many
dittinct operations, each «... which represents a potential source of hydrocarbon evaporation loss.
Crude oil is transported from production operations to the refinery via tankers, barges, tank cars, tank
trucks, and pipelines. In the same manner, refined petroleum products are conveyed to fuel market-
ing terminals and petrochemical industries by tankers, barges, tank cars, tank trucks, and pipelines.
From the fuel marketing terminals, the f ueU are delivered via tank trucks to service stations, commer-
cial accounts, and local bulk storage plants. The final destination for gasoline is usually a motor vehicle
gasoline tank. A similar distribution path may also be developed for fuel oils and other petroleum
products.
4.4.2 Emissions and Controls
Evaporative hydrocarbon emissions from the transportation and marketing of petroleum liquids
may be separated into four categories, depending on the storage equipment and mode of transporta-
tion used:
1. Large storage tanks: Breathing, working, and standing storage losses.
2. Marine vessels, tank cars., and tank trucks: Loading, transit, and ballasting losses.
3. Service stations: Bulk fuel drop losses and underground tank breathing losses.
4. Motor vehicle tanks: Refueling losses.
(In addition, evaporative and exhaust emissions are also associoted with motor vehicle operation.
These topics are discussed in Chapter 3.)
4.4.2.1 Large Storage Tanks - Losses from storage tanks are thoroughly discussed in Section 4.3. "
4.4.2.2 Marine Vessels, Tank Cars, and Tank Trucks • Losses from marine vessels, tank cars, and tank
trucks can be categorized into loading losses, transit losses, and ballasting losses.
Loading losses are the primary source of evaporative hydrocarbon emissions from marine vessel.
tank car, and tank truck operations. Loading losses occur as hydrocarbon vapors residing in empty
cargo tanks are displaced to the atmosphere by the liquid being loaded into the cargo tanks. The
hydrocarbon vapors displaced from the cargo tanks are a composite of (1) hydrocarbon vapors formed
in the empty tank by evaporation of residual product from previous hauls and (2) hydrocarbon vapors
generated in the tank as the new product is being loaded. The quantity of hydrocarbon losses from
loading operations is, therefore, a function of the following parameters:
• Physical and chemical characteristics of the previous cargo.
• Method of unloading the previous cargo.
54
-------
PRODUCT
STORAGE
TANKS
CRUDE OIL PRODUCTION
en
01
MARKETING
TERMINAL
STORAGE
TANKS
TANK CAR
BULK
PLANT
STORAGE
TANKS
TANK TRUCK
PETROCHEMICALS
SERVICE
STATIONS
i
COMMfRCIAL
ACCOUNTS*
STORAGE
TANKS
AUTOMOBILES
ANO
OTIIfR MOTOR
VEHICLES
Figure 4.4-1. Flowsheet of petroleum production, refining, and distribution systems.
(Sources of organic evaporative emissions are indicated by vertical arrows.)
-------
• Operations during the transport of the empty carrier to the loading terminal.
• Method of loading the new cargo.
• Physical and chemical characteristics of the new cargo.
The principal methods of loading cargo carriers are presented in Figures 4.4-2,4.4-3. and 4.4-4. In
the splash loading method, the fill pipe dispensing the cargo is only partially lowered into the cargo
tank. Significant turbulence and vapor-liquid contacting occurs during the splash loading operation.
resulting in high levels of vapor generation and loss.. If the turbulence is high enough, liquid droplets
will be entrained in the vented vapors.
VAPOR EMISSIONS
HATCH COVER
?F CARGO TANK
Figure 4.4-2. Splash loading method.
"-FILLPIPE
HATCH COVER
CARGO TANK
Figure 4.4-3. Submerged fill pipe.
A second method of loading it submerged loading. The two type* of submerged loading are the
aubmerged fill pipe method and the bottom loading method. In the submerged fill pipe method, the
fill pipe descends almost to the bottom of the cargo tank. In the bottom lo«ding method, the fill pipe
enters the cargo tank from the bottom. During the major portion of both forms of submerged loading
56
-------
VAPOR VENT
TO RECOVERY
OR ATMOSPHERE
HATCH CLOSED
VAPORS
== PRODUCT ^S=^f=S"f
CARGO TANK
FILL PIPE
Figure 4.4-4. Bottom loading.
method*, the fill pipe opening i* positioned below the liquid level. The lubmerged loading method
•ipnificantly reduces liquid turbulence and vapor-liquid contacting, thereby retailing in much lower
hydrocarbon Io*»e* than encountered during cplath loading methods.
The history of a cargo carrier 1s just as Important a factor
in loading losses as the method of loading. If the carrier has
just been cleaned or has carried a nonvolatile liquid such as fuel
oil, 1t will be full of clean air immediately prior to loading; if
it has just carried gasoline and has not been vented, the carrier
will be full of air saturated with hydrocarbon vapor. In the latter
case, the residual vapors are expelled along with newly generated
vapors during the subsequent loading operation.
Some cargo carriers are dedicated to the transport of only one
product. In this situation, tanks are not cleaned between each
trip and so return for loading containing air fully or partially
saturated with vapor. The degree of dedication differs for marine
vessels, tank cars, and large and small tank trucks. It also varies
with ownership of the carrier, petroleum liquid being transported,
geographic location, season of the year, and control measure employed.
57
-------
Gasoline tank trucks may be in "dedicated balance service,"
where the truck picks up the vapors displaced during unloading opera-
tions and transports them in the empty tank back to the loading
terminal. Figure 4.4-5 shows a tank truck in dedicated vapor balance
service unloading gasoline to an underground service station tank and
filling up with displaced gasoline vapors to be returned to the truck
loading terminal. The vapors in an empty gasoline tank truck in
dedicated balance service are normally saturated with hydrocarbons.
Dedicated balance service is not usually practiced with marine vessels.
MANIFOLD FOR RETURNING VAPORS
VAPOR VENT LINE
\ . \1
^ r »T
PRESSURE RELIEF VALVES-*-
f\\\\\\\\\\\\\rrrrr
111111itt
==; SUBMERGED FILL PIPE
Figure 4.4-5. Tanktruck unloading into an underground service station storage tank.
Tanktruck is practicing "vapor balance" form of vapor control.
58
-------
Emissions from loading hydrocarbon liquid can be estimated (within 30 percent) using the follow-
ing expression:
LL « 12.46
(1)
where: LL * Loading loss, lb/103 gal of liquid loaded.
M • Molecular weight of vapors, Ib/lb-mole (»ee Table 4.3-D-
P « True vapor pressure of liquid loading, psia (see Figure* 4.3-8 and
4.3-9. and Table 4.3-1).
T * Bulk temperature of liquid loaded, °R.
S « A saturation factor (see Table 4.4-1).
The saturation farlor (S) represents the expelled vapor'» fractional approach to saturation and
a«-<>iini- fur the variations observed in emission rates from the different unloading and loading
method.*. Table 4.4-1 lists nuggealed naturation factors (S).
Table 4.4-1. S FACTORS FOR CALCULATING PETROLEUM
LOADING LOSSES
Cargo carrier
Tank trucks and tank cars
Marine vessels8
Mode of operation
Submerged loading of a clean
cargo tank
Splash loading of a clean
cargo tank
Submerged loading: normal
dedicated service
Splash loading: normal
dedicated service
Submerged loading: dedicated.
vapor balance service
Splash loading: dedicated,
vapor balance service
Submerged loading: ships
Submerged loading: barges
S factor
0.50
1.45
0.60
1.45
1.00
1.00
0.2
0.5
*To be used (or oroduett other than gMOdne: ine factor* from Table 4.4-2
for marine loading of gatoline.
59
-------
Ballasting operation! are • major source of hydrocarbon emiuionr tssociated with unloading
petroleum liquids at marine terminal*. It U common practice for large tanken to fill several cargo
tankt with water after unloading their cargo. Thii water, termed ballad, improve* the stability of the
empty tanker on rough teat during the subsequent return voyage. Ballaiting emission* occur ai hydro-
carbon-laden air in the empty cargo lank it displaced to the atmosphere by ballast water being pumped
into the empty cargo tank. Although ballasting practice* vary quite a bit, individual cargo tank* are
ballasted about 80 percent, and the total vessel i* ballasted approximately 40 percent of capacity.
Ballasting emission! from gasoline and crude oil tanker* are approximately 0.8 and 0.6 lb/10s gal.
respectively, based on total Unker capacity. These estimate* »re for motor fasolinei and medium
volatility crude* (RVP. 5 psia).' Upon flrri val
in port, this ballast water is pumped from the cargo tanks before
loading the new cargo. The ballasting of cargo tanks reduces the
quantity of vapor returning in the empty tanker, thereby reducing
the quantity of vapors emitted during subsequent tanker loading
operations.
Recent studiei on gasoline loading lossei from thipi and barge* have led to the development of
more accurate emission factors for these specific loading operation*. These factor* are presented in
Table 4.4-2 and should be used instead of Equation (1) for gasoline loading operation* at marine
terminal*.1
Table 4.4-2. HYDROCARBON EMISSION FACTORS FOR GASOLINE LOADING OPERATIONS
f
Vesse! tank condition
Cleaned «nd vapor free
lb'1()3 oil transferred
kg'ICP liter transferred
Ballasted
Ib'IO^ gai transferred
kgM03 liter transferred
Unctsaned • dedicated service
lb'103 e»' transferred
kg'103 hte' transferred
•Average cargo tank condition
lbM03g»i transferred
ks'10^ l«er transferred
Hydrocarbon emission factors
Ships
Range
0 to 2.3
0 to 0.26
06 to 3
0.05 to 0.36
04 to 4
0.05 to 0«8
a
Average
1.0
0.12
1.6
0.19
24
0.29
1.4
0.17
Ocean barges
Range
Oto3
0 to 0.36
0.5 to 3
0.06 to 0.36
0.5 to 5
0.06 to 0.60
-
a
Average
1.3
0.16
2.1
0.25
3.3
0.40
a
Barges
Range
a
b
1.4 to 9
0.17 to 1.08
a
Average
1.2
0.14
b
4.0
0.4B •
-
4.0
0.46
*Tt»«t v»iuti •'« not amiable.
bfi>'B«t »'» nei normciiy
60
-------
Simple Calculation • Loading louei from a gasoline tank truck in dedicated balance aervice
•nd practicing vapor recovery would be calculated at follow* uiing Equation (1).
Design basis:
Tank truck volume is 8000 gallons
Gasoline RVP i; 9 psia
Dispensing temperature is 80CF
Vapor recover) efficiency is 95^f
Loading loss equation:
where: S « Saturation factor (see Table 4.4-1) = 1.0
P « True vapor pressure of gasoline (see Figure 4.3-81 = 6.6 psia
M * Molecular weight of gasoline vapors (see Table 4.3-1) £66
T « Temperature of gasoline s 540"R
eff • The control efficiency « 95^
« 0.50 lb/103 gal
Total loading loisei are
(O.SO lb/10> gal) (8.0 x 10' gal) » 4.0 Ib of hydrocarbon
Control measures for reducing loading emissions include the application of alternate loading
methods producing lower emission? and the application of vapor recovery equipment. Vapor recovery
equipment captures hydrocarbon vapors displaced during loading and ballasting operations and re-
covers the hydrocarbon vapors by the use of refrigeration, absorption, adsorption, and/or compres-
sion. Figure 4.4-6 demonstrate* the recovery of gasoline vapors from tank trucks during loading oper-
ation at bulk terminals. Control efficiencies range from 90 to 98 percent depending on the nature of
the vapors and the type of recovery equipment employed.4
Emissions from controlled loading operations can be calculated by multiplying the uncontrolled
emission rate calculated in Equations (1) and (2) by the control efficiency term:
f, efficiencv |
L TOO J
61
-------
VAPOR RETURN UNE
i
I TRUCK
\STORAGE V
\ COMPARTMENTS
cr»
ro
VAPOR FREE
AIR VENTED TO
ATMOSPHERE
VAPOR
RECOVERY
UNIT
PRODUCT FROM
LOADING TERMINAL
STORAGE TANK
Figure 4.4-6. Tanktruck loading with vapor recovery.
-------
In addition to loading and ballasting losses, losses occur while the
cargo is in transit. Transit losses are similar in many ways to breathing
losses associated with petroleum storage (refer to section 4.3). Experimental
tests on tankers and barges have indicated that transit losses can be calculated
using Equation (2):
LT • o.i PW . (2)
where: L— • Transit low, lb/week-101 §al transported.
P * True vapor prenure of the transported liquid, psia
<»ee Figures 4.3-8 and 4.3-9, and Table 4.3-1).
XT • Density of the condensed vapors, Ib/gal (aee Table 4.3-1).
In the absence of specific input! for Equation! (1) and (2), typical evaporative hydrocarbon emit-
•ioni from loading operations are presented in Table 4.4-3. It should be noted that, although the crude
oil used to calculate the emission values presented in Table 4.4*3 has an R VP of S. the R VF of crude oili
can range over two orders of magnitude. In areas where loading and transportation sources are major
factors affecting the air qualit) it is advisable to obtain the accessary parameters and to calculate
emission estimates from Equations (I) and (2).
Emissions from gasoline trucks have been studied by a combine-
7 8
tion of theoretical and experimental techniques, * and typical
emission values are presented 1n Table 4.4-3. Emissions depend
upon the extent of venting from the tank truck during transit,
which 1n turn depends on the leak-tightness of the truck, the
pressure relief valve settings, the pressure in the tank at the
start of the trip, the vapor pressure of the fuel being transpor-
ted, and the degree of saturation (with fuel vapor) of the vapor
space 1n the tank. The emissions are not directly proportional
to the time spent 1n transit: as the leakage rate of the truck In-
creases, emissions Increase up to a point and then level off as
other factors take over 1n determining the rate. Tank trucks in
dedicated vapor balance service typically contain saturated
vapors; this leads to lower emissions during transit because no
additional fuel evaporates to raise the pressure In the tank
63
-------
and cause venting. Table 4.4-3 lists "typical" values for emissions
and also "extreme" values which could occur in the unlikely event
that all factors that determine emissions had precisely the proper
values to give maximum emissions.
4.4.2.3 Service Stations - Another major source of evaporative hydrocarbon emission.' if the filling
of underground gasoline storage tanks at service stations. Normally, gasoline is delivered to service
stations in large (8000 gallon) tank trucks. Emissions are generated when hydrocarbon >apoi> in the
underground storage tank are displaced to the atmosphere by the gasoline being loaded into the tank.
As with other loading losses, the quantity of the service station tank loading lo« depend*on several
variables including the aize and length of the fill pipe, the method of filling, the tank configuration.
and the gasoline temperature, vapor pressure, and composition. An average hxdrocarhon emission
rate for submerged filling is 7.3 lb/10s gallons of transferred gasoline, and the rate for splash filling
is 11.5 lb/10' gallons of transferred gasoline (Table 4.4-4).«
Emissions from underground tank fill ing operations at *erv ire station* can he red ured b\ the use of
the vapor balance system (Figure 4.4-5). The vapor balance system employs a vapor return hose which
returns gasoline vapors displaced from the underground tank to the tank truck storage compartment
being emptied. The control efficiency of the balance c)*tem ranee* from 93 to 100 percent. Hydrocar-
bon emissions from underground tank filling operation' at a service station employing the vapor
balance system and submerged filling are not expected to eu-ecri 0.3 Ih 10' gallons of transferred
gasoline.
64
-------
Table 4.4-3.
HYDROCARBON EMISSION FACTORS FOR PETROLEUM LIQUID
TRANSPORTATION AND MARKETING SOURCES
Emission source
Tank cars/trucks
Submerged load Ing- normal service
lb/103 gal transferred
kg/103 liters transferrtd
Splash loading-normal service
lb/103 gal transferred
kg/103 liters transferred
Submerged loading-balance service
lb/103 gal transferred
3
kg/10 liters transferred
Splash loading-balance service
lb/10 gal transferred
kg/103 liters transferred
Transit- loaded with fuel
lb/10 gas transferred
Product emission factors3
6asol1neb
5
0.6
12
1.4
8
1.0
8
1.0
0-0.01
typical
: 0-0.08
kg/103 liters transferred
extreme
0-0.001
typical
0-0.009
extreme
Transit-return with vapor
lb/103 gal transferred • 1 0-0.11
i typical
l
kg/103 liters transferred
i
; Marine vessels
Loading tankers
lb/103 ga! transferred
kg/ 10 liters transferred
Loading Barges
: lb/10 gal transferred
; kjj/103 liters transferred
i Tanker ballasting
lb/10 gal cargo capacity
kg/10 liters cargo capacity
; Transit
Ib/week - 10 gal transported
kg/week - 10 liters transported
0-0.37
extreme
0-0.013
typical
0-0.044
extreme
f
f
f
0.8
0.10
3
0.4
Jet
Crude naphtha
o1lc (JP-4)
3
0.4
7
0.8
5
0.6
5
0.6
e
e
e
e
e
e
0.07
0.08
1.7
0.20
0.6
0.07
1
0.1
1.5
0.18
4
0.5
2.5
0.3
2.5
0.3
e
e
Q
e
e
e
e
0.05
0.06
1.2
0.14
e
0.7
0.08
Distillate Residual
Jet oil
kerosene No. 2
oil
No. 6
0.02
0.002
0.04
0.005
d
d
e
e
£
Q
e
e
e.
e
0.005
0.0006
-
0.0013
0.0016
e
O.OC5
0.0006
0.01 0.0001
0.001
0.03
0.004
d
0.00001
0.0003
0.00004
d
d
d
e i e
e
e
e
e
e
e
e
0.005
0.0006
0.012
0.0014
e
0.005
0.0006
e
e
e
e
e
e
e
0.00004
5 x 10'6
0.00009
1.1 x 10'5
e
3 x 10*5
4 x 10'6
a Emission factors are calculated for dispensed fuel temperature of 60° F.
b The example gasoline has an RVP of 10 psia.
c The example crude oil has an RVP of 5 psia.
d Not normally used. cc
e Not 'available. ' 3
f See Table 4.4-2 for these emission factors.
-------
Table 4.4-4. HYDROCARBON EMISSIONS FROM GASOLINE
SERVICE STATION OPERATIONS
Emission source
Filling underground tank
Submerged filling
Splash filling
Balanced submerged filling
Underground tank breathing
and emptying3
Vehicle refueling operations
Displacement losses
(uncontrolled)
Displacement losses
(controlled)
Spillage
Emission rate
Ib/ID^ ga! throughput
7.3
11.5
0.3
1
9
0.9
0.7
kg.'IO^ liters throughput
O.BB
1.38
0.04
0.12
1.08
0.11
0.084
a Emissions include any vapor loss from the underground tank
to the gas pump.
A second source of hydrocarbon emissions from sen-ice nations it underground tank breathing
Breathing losses occur daily and are attributed to temperature changes barometric pressure changes.
and gasoline evaporation. The type of aervice nation operation also has a large impart on breathing
losses. AD average breathing emission rate is 1 lb/101 gallons throughput.4
4.4.2.4 Motor Vehicle Refueling - An additional aource of evaporative hydrocarbon emission; at
aervice atations is vehicle refueling operations. Vehicle refueling emissions are attributable to vapor;
displaced from the automobile tank by dispensed gasoline and to spillage. The quantity of displaced
vapors is dependent on gasoline temperature, auto tank temperature, gasoline FVP, and dispensing
rates. Althouf h several correlations have been developed to estimate losses due to displaced vapors.
significant controversy exists concerning these correlations. It is estimated that the hydrocarbon
emissions due to vapors displaced during vehicle refueling average 9 lb/101 gallon; of dispensed
gasoline.**1
The quantity of spillage loss is a function of the type of service station, vehicle tank configuration.
operator technique, and operation discomfort indices. An overall average spillage loss i; 0.7 lb '10'
gallons of dispensed gasoline.*. '
Control methods for vehicle refueling emissions are based on conveying the vapor; displaced from
the vehicle fuel tank to the underground storage tank vapor spare through the u»e of a special note and
nozzle (Figure 4.4-7). In the "balance" vapor control system, the vipor* are con*eyed by natural pres-
sure differentials established during refueling. In "vacuum assist" vapor control systems, the eomev-
ance of vapors from the auto fuel lank to the underground fuel tank i* assisted b> a \ aruum pump. The
overall control efficiency of vapor control systems for vehicle refueling emissions it estimated to be 86
to 92 percent.'
66
-------
SERVICE
STATION
PUMP
Figure 4.4-7. Automobile refueling vapor-recovery system.
References for Section 4.4
1. Burklin, C.E. and R.L. Honerkamp. Reviiion of Evaporative Hydrocarbon Emission Factors.
Research Triangle Park, N.C EPA Report No. 450/3-76-039. August 15, 1976.
2. Burklin, Clinton E. et al. Background Information on Hydrocarbon Emissions From Marine
Terminal Operations, 2 Volt., EPA Report No. 450/3-76-038a and b. Research Triangle Park, N.C
November 1976.
3. American Petroleum Inst., Evaporation Loss Committee. Evaporation Loss From Tank Cars.
Tank Trucks, and Marine Vessels. Washington, D.C Bull. 2514. 1959.
4. Burklin, Clinton E. et al. Study of Vapor Control Methods For Gasoline Marketing Operations.
2 Vols. Radian Corporation. Austin, Texas. May 1975.
5. Scott Research Laboratories, Inc. Investigation Of Passenger Car Refueling Losses. Final Report.
2nd year program. EPA Report No. APTD-1453. Research Triangle Park, N.C September 1972.
6. Scott Research Laboratories, Inc. Mathematical Expressions Relating Evaporative Emissions
From Motor Vehicles To Gasoline Volatility, summary report. Plumsteadville, Pennsylvania.
API Publication 4077. March 1971.
67
-------
7. Nichols, R.A. Analytical Calculation of Fuel Transit Breathing
Loss. Prepared by R.A. Nichols Engineering, Corona del Mar, CA,
for Chevron USA, Inc., San Francisco, CA, March 21, 1977.
8. Nichols, R.A. Tank Truck Leakage Measurements. Prepared by
R.A. Nichols Engineering, Corona del Mar, CA, for Chevron USA,
Inc., San Francisco, CA, June 7, 1977.
9. California Air Resources Board (CARB), "Delivery Tank Field
Results." Attachment 2 to Staff Report 77-5-1, March 15, 1977.
68
-------
BACKGROUND DOCUMENT
TRANSIT LOSSES FROM GASOLINE TANK TRUCKS
1.0 INTRODUCTION
This document pertains only, to the emission factors for transit
losses from gasoline tank trucks presented in Table 4.4-3 of AP-42
and to the text that deals with these factors. The remainder of
Section 4.4 was prepared by others, and the background for their
calculations is found elsewhere.
2.0 THEORETICAL CALCULATIONS OF EMISSIONS
A theoretical analysis of truck transit breathing loss was made
by Nichols using an isothermal stirred tank as a model. Calcula-
tions were made for two situations: (1) where the tank has an open
vent, and (2) where the tank has a P/V valve which prevents all venting
until the valve opening pressure is reached and allows free venting
thereafter.
2.1 TRANSIT WITH FULL FUEL LOADS
1, 2
For truck transit with a full fuel load from the terminal, venting
was assumed to occur until the fuel vapor space was saturated to the
fuel vapor pressure. When this state is reached, no further pressure
increases occur. The following equation was used to estimate losses
from tanks with open vents and full fuel loads:
V
gm/gal transferred = 0.7057 G p
Wp - VH \ + PH
-------
P = atmospheric pressure, psia
S1 = vapor saturation on leaving terminal, fraction of 1.00
S2 c vapor saturation upon arrival at unloading station, fraction of 1.00
P£ = fuel vapor pressure, psia
For losses from tanks with P/V valves and full fuel loads, the following
equation was used:
gm/gal transferred = 0.7057 G
D
~
H
(s* -
I
where symbols are as above and
Py s vent opening pressure, psia
S* = vapor saturation at the vent opening pressure on leaving
terminal, fraction of 1.00
S| = vapor saturation at the vent opening pressure upon arrival
at unloading station, fraction of 1.00
The constant of 0.7057 in the equations contains the assumption
that the temperature is 74.1°F (534. 1°R) and the mole weight
of the fuel vapor is 66.7 (Ib m/lb mole).
Calculations were made for both situations using the follow-
ing values. Since no experimental data base was available, a
series of values was used for some of the parameters in an
effort to cover the entire range of reasonable values.
VG = 0.05, 0.10, 0.15
VL
P- = 14.7 psia '
S1 = 0, 0.2, 0.5, 0.85, 0.95
S2 = 1.00
70
-------
P° = 5.87 psla
Py - 15.675 psia (27 inches of HgO)
Sf = 0.1661, 0.3661, 0.6661, 1.00
S* = 1.00
Results are shown on the following page in Table B-l taken
from Reference 1. Calculated losses ranged from 0.0066 to 0.1717
gm/gal (0.015 - 0.379 lb/103 gal; 0.002 - 0.045 kg/103 liters) for
the open vent situation, and from 0 to 0.1538 gm/gal (0 - 0.339
lb/103 gal; 0 - 0.041 kg/103 liters) for the vent valve situation.
The above calculations are based on the assumption that the
tank leaks at a rate sufficient to dissipate all tank pressure
during the course of a trip. Some calculations were done using
2
information on typical leak rates available at the time, which
indicated that pressure within exceptionally tight tanks may not
be dissipated in a 60 minute trip, although it usually is.
2.2 TRANSIST WITH VAPOR LOADS (RETURN TRIP)1
Theoretical values for maximum emissions were calculated by
assuming that the residual fuel present in the tank after the
load is delivered vaporizes immediately and so causes vapor
venting from the tank at the highest initial tank pressure. The
amount of vapor lost depends on the rate of leakage of the tank,
which in turn determines the residual pressure in the tank when
it reaches the refueling terminal. For trips of sufficient
duration to permit maximum dissipation of pressure, venting
losses can be calculated using the same equations that were applied
to the full load case described in Section 2.1. Since no experi-
mental data base was available, a series of values was used for
each parameter in an effort to cover the entire range of reason-
able values.
71
-------
K. A. Nichols
Engineering
TABLE B. 1
VENT LOSS AFTER REFUELING
P * 14.7
. OPEN
S]B
VV/VG
J D
< ui
0 H
•^ 2
2 u
u >
VG/VL *= o. 05
VG/VL* o.io
VG/V, e o. 15
\-i j^
0.
.5097
.0572
.1145
. 1717
IDEAL 27 IN.
P s 15.675 S* *
vv/vG
J O
<; u
U H
•— 2
2 U
o >
VC/VL = 0-05
vc/vL = 0.10
VG/VLS 0.15
.1661
.4050
.0513
. 1025
. 1538
VENT
.2
.4265
.0555
.1110
.1665
CALCULATION
.5 .85
.2870 .0951
.0453 .0182
.0906 .0365
.1359 .0547
95
0327
0066
0132
0198
H2O VENT CALCULATION
.3661
.3217
.0466
.0933
.1399
.6661
.1822
.0316
.0633
.0949
72
-------
VG - 1.05, 1.10, 1.15
\
P « 14.7 psla
S1 - 0, 0.3261, 0.7065, 0.8152
S2 « 0.200, 0.500, 0.850, 0.950
P° = 5.87 psla
Py = 15.675 psla (27 Inches of H20)
S* « 0.1661, 0.4922, 0.8726, 0.9813
S* » 0.2055, 0.5265, 0.9099, 1.000
Results are shown on the following page in Table B-2 taken from
Reference 1. Venting losses ranged from 0.037 to 0.350 gm/gal
(0.082 - 0.772 lb/103 gal; 0.010 - 0.093 kg/103 liters) for the
open vent situation, and from 0.013 to 0.067 gm/gal (0.029 -
0.148 lb/103 gal; 0.003 - 0.018 kg/103 liters) for the vent valve
situation.
In situations where residual pressure still remains in the
tank, losses can be calculated from the following equation:
gm/gal transferred = 4.142 (S|) (Py - PR) (VG)
~V~ \
where symbols are as previously defined and
PH = residual tank pressure, psia
The constant of 4.142 contains the" assumption that P£ = 5.87 psia,
MH = 66.7 lb m/lb mole, and T = 74.TF.
Since no experimental measurements of PR were available,
calculations were made for a wide range of theoretically
possible values. Using the following values, losses were calcu-
lated from the above equation:
73
-------
JR. A. NichoL
Engineering
TABLE B.2
VENT LOSS AFTER FUEL DROP
OPEN VENT LOSS
P = 14.7 Sj =
VV/VG B
GM/GAL
VENTED
VG/VL = i.os
VG/VL « 1.10
.0
.2
.0832
.0367
.0385
.0402
IDEAL 27 IN.
"?•
S| =
VV/VG =
GM/GAL
VENTED
VG/VL = i.os
VG/VL = 1.10
. 0000
.1661
.2055
.0159
.0128
.0134
. 0140
.3261
.5000
.0832
.1499 .
.1571
.1642 .
H2O VENT
.3261
.4922
.5265 .
.0159
.0351
.0368
.0385
7065
8500
0832
2820
2954
3088
LOSS
7065
8726
9009
0159
0612
0641
0670
.8152
.9500
.0832
.3197
.3350
.3502
.8152
.9813
1.000
.0112
. 0481
.0504
.0527
74
-------
SJ •'.LOO, 0.9009, 0.5265, 0.2055
Py » 15.675 psla (27 Inches of Hg
PP = 20 different values ranging from 14.7 to 15.58 psla (0.0
K to 21.6 Inches of
VG « 1.05, 1.10, 1.15
Results are shown on the following page in Table B.7 taken from
Reference 1. Losses ranged from 0.011 to 0.296 gm/gal (0.024 -
0.653 lb/103 gal; 0.0031 - 0.078 kg/103 liters).
3.0 MODIFIED CALCULATIONS BASED ON EXPERIMENTAL DATA
1, 3, 4
The theoretical calculations above covered the entire range
of values for most of the parameters for which no field measure-
ments had been made (S^, S2, Sf, S|, Py, and PR). Subsequently,
experiments were conducted to determine where typical values
lay within the range of values considered theoretically.
Pressure measurements were made on tank trucks while they were
filled with fuel and in transit. Pressures varied widely and
frequently were negative because air, originally present in the
vapor space, dissolved in the freshly charged fuel. This
situation is apparently typical of fuel that has been stored in
floating roof tanks and is not saturated with air.
In addition, vent valves were shown to open partially rather
than fully at the valve opening pressure. Vapors in tanks were
found to be somewhat less than saturated (71 to 96 percent) in-
stead of 100 percent saturated, as originally assumed, when a
value of 1.00 was chosen for S| in the computations shown in Table •
B-l from Reference 1. Moreover, truck leakage rates were shown to be
75
-------
R. A. Nichols
Engineering
TABLE B.7 SLOWDOWN LOSS FOR VARIOUS FUEL DROP
VENT SPACE AND TANK LEAKAGE SITUATIONS
PR IN. H20
J C
<; -H ..
0 .H
.s'S
Vr/VT s 1.05
VG/VL - i.io
VC/VLS 1.15
. 0 >
APP IN. H70
r\ £
J O
<; a
o H
5 1
\'/V, e 1.05
G' L-
VG/VL = 1.10
VG/VL =1.15
U > ~ ~
APR IN. H2O
J D
<; H
Oi ._.
H
~ 2
2 U
Vr/VT = 1. 05
V_l JU
V /V-r =1.10
Cj X^
VG/VL = 1.15
O > . -
APR IN H2O
• J D
< W
0 H
2 §
VC/VL = i.os
VC/VL = 1.10
VC/VL B i.i5
0 >
S| = 1.0
0.0 1.54 6.69 15.23 20.72
.2706 .2552 .2035 .1180 .0629
.2835 .2673 .2132 .1236 .0659
.2964 .2795 .2229 .1292 .0689
Sj = .9009
0.0 1.8 " 7.12 15.35 20.75
.2438 .2275 .1795 .1052 .0564
.2554 .2384 .1880 .1102 .0591
.2670 .2492 .1966 .1152 .0618
A
S* = .5265
0.0 2.2 7.2 15.7 21.0
.1425 .1309 .1045 .0596 .0317
.1492 .1371 .1094 .0625 .0332
.1560 .1433 .1144 .0653 .0347
^t • '
S2 = . 2055
0.45 3.45 7.46 16.7 21.6
.0547 .0485 .0402 .0212 .0111
.0573 .0508 .0422 .0222 .0117
.0599 .0531 .0441" .0232 .0122
0.0
. 0556
.0583
.0609
76
-------
much lower than previously supposed—possibly as little as 5
percent of the rates used in the theoretical calculations.
Based on all these findings—specifically, a tank pressure
(Pv) of 14.81 psia (3 inches of H20), leakage that persisted
for 5 minutes before the tank pressure became negative, and a
leak rate of 5 percent of that used previously—the authors of
Reference 3 estimated that losses from transit with full loads
are 0 - 0.035 gm/gal (0 - 0.077 lb/103 gal; 0 - 0.009 kg/103
liters) rather than the 0 - 0.0172 gm/gal that was computed
theoretically. These calculations were performed using a
computer program that employs the same fundamental equations given
in Sections 2.1 and 2.2 and also considers leakage rates expressed
as equivalent orifice diameters. A complete explanation is
given in References 1 and 3. A summary of the results is shown
in Table.!, taken from a June 10, 1977 letter from R.A. Nichols
to H.B.Uhlig of Chevron USA, Inc., San Francisco, California.
Experimental tests showed that the degree of saturation of
vapors in empty tankers returning to be refilled was lower than
previously estimated. A value for Si of 0.10 was selected by
the authors of Reference 3 as more representative than the values
used in the theoretical calculations. The lower tank leakage rates
also reduced the losses as compared with the original estimates.
A range of values from 0 to 0.166 gm/gal (0 - 0.366 lb/10 gal;
0 - 0.044 kg/10 liters) was selected rather than the 0.011 -
0.350 gm/gal estimated from the theoretical analysis.
No experiments have been attempted for the purpose of moni-
toring the actual hydrocarbon emissions from tank trucks as they
are in transit. The losses are so small that they could not be
detected by weighing (or otherwise measuring) the load at the
start and end of each run. Experiments have been designed to
provide values for the various parameters used to compute losses
from well established theoretical principles, however.
77
-------
Table 1. TRUCK TRANSIT AND TRANSFER LEAKAGE LOSS
CD
&p
Loss
In.H2O
0
0.5
1
•t
3
4
•
Leak Din
5000 Gal
In. O
0
.0528
.0749
. 107
.131
.153
. 200
.250
.300
.400
.500
Terminal
Loss
Gm/Gal
0
.0002
.0003
.0006
.0010
.0013
.0022
.0035
.0050
.0089
.0139
Terminal
Vapor
Vol%Loss
0
.0037
.0076
.0155
.0232
.0316
.0537
.0840
. 1210
.2150
.3358
AP = 6"
QL =600ppm
* Leakage rate 1s
Liquid
Vapor
Trans. Loss Trans. Loss
Gm/Gal Gm/Gal
0
.0001
.0002
.0004
. 0006
.0008
.0013
.0021
.0030
.0053
.0083
t=5min
nA a V " O
J » » *i A. ~~ • \J .
such that
0
.0016
.0032
. 0065
.0097
.0132
.0226
.0354
.0509
r.OS28.
;0905?
..0828
\1415'
20 AP=6"H20
t=60 min
_ _• * -•. "tf • v' O t O
1 S 1 nA *1 V •"
Loss S. S. S. S,
Stage 1
Gm/Gal
.0911
.0915
.0924
.0961
. 0994
.1019
.1110
.1267
. 1400
. 1802
. 2278
Stape I
Vol%Loss
2.20
2.21
2.23
2.32
2.40
2.46
2.68
3.06
3.38
4.35
5.50
TotLoss
w/oStp 1
Gm/Gal
0
.0019
.0037
.0075
.0113
.0153
.0261
.0410
.05H9
(.0970
\ 104?'
r 1051.
\1637'
J51. M^->165?
maximum losses occur.
TotLoss
w/stp I
Gm/Gal
.091 1
.0934
.0961
. 1036
. 1 107
. 1 172
.1371
. 1677
. 1989
2772
\ 2H49'
..3328
\3915'
h
3
oq
V*
3
re
re
-i
-------
REFERENCES
1. Nichols, R.A., "Analytical Calculation of Fuel Transit
Breathing Loss." Prepared by R.A. Nichols Engineering,
Corona del Mar, Ca., for Chevron USA, Inc., San Francisco,
Ca., March 21, 1977.
2. California Air Resources Board (CARB), "Delivery Tank Field
Results." Attachment 2 to Staff Report 77-5-1, March 15, 1977.
3. Nichols, R.A., "Tank Truck Leakage Measurements." Prepared
by R.A. Nichols Engineering, Corona del Mar, Ca., for Chevron
USA, Inc., San Francisco, Ca., June 7, 1977.
4. Private correspondence from R.A. Nichols of R.A. Nichols
Engineering, Corona del Mar, Ca., to H.B. Uhlig of Chevron
USA, Inc., San Francisco, Ca., June 10,:1977.
79
-------
4.5 WASTE SOLVENT RECLAMATION
4.5.1 Process Description*'4
Waste solvents are organic dissolving agerits that are
contaminated with suspended and dissolved solids, organics, water,
other solvents, and/or any substance not added to the solvent
during its manufacture. Reclaiming is the process of restoring a
waste solvent to a condition that permits its reuse, either for its
original purpose or for different industrial needs. The limiting
factor that determines whether a solvent is reclaimed is economic
because the cost of reclamation may exceed the value of the
recovered solvent.
Industries that produce waste solvents include solvent
refining, polymerization processes, vegetable oil extraction,
metallurgical operations, pharmaceutical manufacture, surface
coating, and cleaning operations (dry cleaning and solvent
degreasing). The amount of solvent recovered from the waste
solvent varies from about 40 to 99 percent, depending on the extent
and characterization of the contamination and on the recovery.
process employed.
Design parameters and economic factors determine whether
solvent reclamation is accomplished as a main process by a private
contractor, as an integral part of a main process (such, as solvent
refining), or as an add-on process (as in the surface coating and
cleaning industries). Most contract solvent reprocessing
operations recover halogenated hydrocarbons (e.g., methylene
chloride, trichlorotrifluoroethane, and trichloroethylene) from
degreasing, and/or-aliphatic, aromatic, and naphthenic solvents
such as those used in the paint and coatings industry. They may
also reclaim small quantities of numerous specialty solvents such
as phenols, nitriles, and oils.
80
-------
The general reclamation scheme for solvent reuse is
illustrated in Figure 4.5-1. Industrial operations are, however,
capable of reclaiming their waste solvent without incorporating all
of these steps. For instance, initial treatment is only necessary
when liquid waste solvents contain dissolved contaminants.
STORAGE FUGITIVE FUGITIVE
TANK VENT EMISSIONS EMISSIONS
CONDENSER FUGITIVE FUGITIVE
VCNT EMISSIONS EMISSIONS
STORAGE FUGITIVE
TANK VEST EVISSIONS
©!
V.ASTE | ST01ACEAND
SOLVCMS | HANDLING
_| INITIAL
1 TREATMENT
®
~ 1
< 1
DISTILLATION
WASTE
DISPOSAL
STORAGt AND
PURIFICATION -— H4K0.JNG
—• INCINERATOR STACK
— FUGITIVE EMISSIONS
SOLVENT
Figure 4.5-1. General Waste Solvent Reclamation Scheme
and Emission Points1
Solvent Storage and Handling - Solvents are stored before and
after reclamation in containers ranging in size from 0.2 m3 (55
gal) drums to tanks with capacities of 75 m3 (20,000 gal) or
more. Storage tanks are of fixed or floating roof design. Venting
systems prevent solvent vapors from creating excessive pressure or
vacuum inside fixed roof tanks.
Handling includes loading waste solvent into process
equipment and filling drums and tanks prior to transport and
storage. The filling is most often done through submerged or
bottom loading.
81
-------
Initial Treatment - Waste solvents are initially treated by
vapor recovery or mechanical separation. Vapor recovery entails
removal of solvent vapors from a gas stream in preparation for
further reclaiming operations; in mechanical separation, undis-
solved solid contaminants are removed from liquid solvents.
Vapor recovery or collection methods employed include conden-
sation, adsorption, and absorption. Technical feasibility of the
method chosen depends on solvent miscibility, vapor composition,
and vapor concentration; solvent boiling point, reactivity, and
solubility; and several other factors.
Condensation of solvent vapors is accomplished by water-cooled
condensers and refrigeration units. For adequate recovery, a sol-
vent vapor concentration well above 20 mg/nP is required. To
avoid explosive mixtures of a flammable solvent and air in the
process gas stream, air is replaced with an inert gas, such as
nitrogen. Solvent vapors that escape condensation are recycled
through the main process stream or recovered by adsorption or
absorption.
Activated carbon adsorption is the most common method of cap-
turing solvent emissions. Adsorption systems are capable of
recovering solvent vapors in concentrations below 4 mg/m^ of air.
Solvents with boiling points of 200°C (392°F) or more do not desorb
effectively with the low pressure steam commonly used to regenerate
the carbon beds. Figure 4.5-2 shows a flow diagram of a typical
fixed-bed activated carbon solvent rcovery system. The mixture of
steam .and solvent vapor passes-to a water-cooled condenser. Water
immiscible solvents are simply decanted to separate the solvent,
whereas water-miscible solvents must be distilled and solvent mix-
tures must be both decanted and distilled. Fluidized-bed opera-
tions are also in use.
82
-------
00
CO
CLEAN AIR
EXHAUST
PROCESS BLOWER
DRYING AIR
BLOWER
(OPTIONAL)
. w.«. .•:;.'.%i.- ; i. »TT;.. --..-••:
j .•:.:.,• J ACT!yATEO..CARBON.T.
.LOW PRESSURE STEAM
•COOUNO WATER IN
p—WATER OUT
CONDENSER
WASTE
WATER
DECANTER
RECOVERED
SOLVENT
Figure 4.5-2. Typical Fixed-Bed Activated Carbon Solvent Recovery System
-------
Absorption of solvent vapors is accomplished by passing the
waste gas stream through a liquid in scrubbing towers or spray
chambers. Recovery by condensation and adsorption results in a
mixture of water and liquid solvent, while absorption results in an
oil and solvent mixture. Further reclaiming procedures are requir-
ed after solvent vapors are collected by any of these three
methods.
Initial treatment of liquid waste solvents is accomplished by
mechanical separation methods. This includes both removing water
by means of decanting and removing undissolved solids by means of
filtering, draining, settling, and/or centrifuging. A combination
of initial treatment methods may be necessary to prepare waste
solvents for further processing.
Distillation - After initial treatment, waste solvents are
distilled to remove dissolved impurities and to separate solvent
mixtures. Separation of dissolved impurities is accomplished by
simple batch, simple continuous, or steam distillation. Mixed
solvents are separated by multiple, simple distillation methods
such as batch or continuous rectification. These processes are
shown in Figure 4.5-3.
I
,
RATION
SOLVENT VAPOR
_ !
REFLUX
SniVFNT ( _. I 1 ,
VAPOR 1 !
n ~ i
SATION
SLUDGE
DISTILLED SOLVENT
Figure 4.5-3. Distillation Process for Solvent Reclaiming*
84
-------
In simple distillation, waste solvent is charged to an evapor-
ator; vapors are then continuously removed and condensed, and the
resulting sludge or still bottoms are drawn off. In steam distil-
lation, solvents are vaporized by direct contact with steam which
is injected into the evaporator. Simple batch, continuous, and
steam distillations follow path I in Figure 4.5-3.
The separation of mixed solvents requires multiple simple
distillations or rectification. Batch and continuous rectification
are represented by path II in Figure 4.5.3. In batch rectifica-
tion, solvent vapors pass through a fractionating column where they
contact condensed solvent (reflux) entering at the top of the
column; solvent not returned as reflux is drawn off as overhead
product. In continuous rectification, the waste solvent feed
enters continuously, at an intermediate point in the column; the
more volatile solvents are drawn off at the top, while those with
higher boiling points collect at the bottom.
Design criteria for evaporating vessels depend on waste
solvent composition; scraped-surface stills or agitated thin-film
evaporators are the most suitable for heat sensitive or. viscous
materials. Condensation is accomplished by shell and tube or
barometric condensers. Azeotropic solvent mixtures are separated
by the addition of a third solvent component, while solvents with
higher boiling points, e.g., in the range of high flash naphthas
(155°C, 311°F), are most effectively distilled under vacuum.
Purity requirements for the reclaimed solvent determine the number
of distillations, reflux ratios, and processing time needed.
Purification - After distillation, water is removed from
solvent be decanting or salting. Decanting is accomplished with
immiscible solvent and water which, when condensed, form separate
liquid layers, one or the other of which can be drawn off mechan-
ically. Additional cooling of the solvent-water mix before
85
-------
decanting increases the separation of the two components by reduc-
ing their solubility. In salting, solvent is passed through a
calcium chloride bed where water is removed by absorption.
During purification, reclaimed solvents are stabilized if
necessary. Buffers are added to virgin solvents to ensure that pH
is kept constant during use. Reclaiming the solvent may cause a
loss of buffering capacity. To renew it, special additives are
used during purification. The composition of these additives is
considered proprietary.
Waste Disposal - Waste materials separated from solvents
during initial treatment and distillation are disposed of by
incineration, landfill ing, or deep well injection. The composition
of such waste varies depending on the original use of the solvent,
but up to 50 percent is unreclaimed solvent, which keeps the waste
viscous yet liquid, thus facilitating pumping and handling proce-
dures. The remainder consists of components such as oils, greases,
waxes, detergents, pigments, metal fines, dissolved metals, organ-
ics, vegetable fibers, and resins.
About 80 percent of the waste from solvent reclaiming by pri-
vate contractors is disposed of in incinerators capable of burning
liquid wastes. About 14 percent is deposited in sanitary land-
fills, usually in 55-gallon drums. Deep well injection consists of
injecting wastes between impermeable geologic strata; viscous
wastes may have to be diluted prior to pumping into the desired
stratum level.
4.5.2 Emissions and Controls !»3-5
Volatile organic and particulate emissions result from waste
solvent reclamation. Emission points include (1) storage tank
vents, (2) condenser vents, (3) incinerator stacks, and fugitive
losses (these numbers refer to Figures 4.5-1 and -3). Emission
86
-------
factors for these sources are given in Table 4.5-1 in kilograms of
pollutant per metric ton and pounds per ton of reclaimed solvent.
Solvent storage results in volatile organic emissions from
solvent evaporation (Figure 4.5-1, emission point 1). The conden-
sation of solvent vapors during distillation (Figure 4.5-3) also
involves VOC emisssions and, if steam ejectors are used, emission
of steam and noncondensables as well (Figures 4.5-1 and -3, point
2). Incinerator stack emissions consist of solid contaminants that
are oxidized and released as particulates, unburned organics, and
combustion stack gases (Figure 4.5-1, point 3).
Volatile organ.i.c emissions from equipment leaks, open solvent
sources (sludge draw-off and storage from distillation and initial
treatment operations), solvent loading, and solvent spills are
classified as fugitive; the former two sources are continuously
.released while the latter two are intermittently emitted.
Solvent reclamation is viewed by industry as a form of control
in itself. It is estimated that less than 50 percent of reclama-
tion plants run by private contractors utilize any control technol-
ogy. Carbon adsorption systems can remove up to 95 percent of the
solvent vapors from the air stream.
Volatile organic emissions from the storage of solvents can be
reduced by as much as 98 percent by converting from fixed to float-
ing roof tanks, although the percent reduction also depends on sol-
vent evaporation rate, ambient temperature, loading rate, and tank
capacity. Tanks may also be refrigerated or equipped with conser-
vation vents, which prevent air inflow and vapor escape until some
preset vacuum or pressure develops.
Solvent vapors vented during distillation are controlled by
scrubbers and condensers. Direct flame and catalytic afterburners
can also be used to control noncondensables and solvent vapors not
condensed during distillation; the time required for complete com-
87
-------
Table 4.5-1.
EMISSION FACTORS FOR SOLVENT RECLAIMING*
EMISSION FACTOR RATING: D
Source
Storage tank
ventB
Condenser
vent ...'••
Incinerator
stack
Incinerator
stack
Fugitive
emissions
Spillage
Loading
Leaks
Open
sources
TOTAL
TOTAL
Criteria
Pollutant
Volatile
organics
Volatile
organics
Vdl ati 1 e
organics
Particul ates
Volatile
organics0
Volatile
organics
Volatile
organics
Volatile
organics
Volatile
organics
Particul ates
Emission Factor Emission Factor
Range Average
kg/MT
.002-. 04
.26-4.17
.01
.55-1.0
.095
.00012-
.71
d
d
.38-5.0
.55-1.0
Ib/ton
.004-. 09
.63-10.13
.02
1.34-2.43
.23
.00029-
1.72
d
d
.92-12.15
1.34-2.43
kg/MT
.0072
1.65
.01
.72
.095
.36
d
d
2.1
.72
Ib/ton
.0175
4.01
.02
1.75
.23
.875
d
d
5.10
1.75
aData obtained by Reference 1 from state air pollution control
agencies and presurvey sampling. All emission factors are for
uncontrolled process equipment except those for the incinerator
stack. (Reference 1 does not, however, specify what the control
is on this stack.)
^Storage tank is of fixed roof design.
C0nly one value available.
dNot available.
88
-------
bustIon depends on the flammability of the solvent. Carbon or oil
adsorption may be employed as well, as in the case of vent gases
from the manufacture of vegetable oils.
Wet scrubbers are used to remove particulates from sludge
incinerator exhaust gases, although they do not effectively control
submicron-sized particles.
Submerged rather than splash filling of storage tanks and tank
cars can reduce solvent emissions from this source by more than 50
percent. Proper plant maintenance and loading procedures reduce
emissions from leaks and spills. Open solvent sources can be
covered to reduce these fugitive emissions.
89
-------
References for Section 4.5
1. Tierney, D.R. and T.W. Hughes, Source Assessment: Reclaiming
of Waste Solvents - State of the Art. EPA-600/2-78/004f.
Prepared by Monsanto Research Corporation, Dayton, Ohio
45407, for U.S. Environmental Protection Agency, Industrial
Environmental Research Laboratory, Office of Research and
Development, Cincinnati, Ohio 45268, April 1978
2. Levin, J.E. and F. Scof.ield, "An Assessment of the Solvent
Reclaiming Industry." Preprints of Papers Presented at the
170th Meeting of the American Chemical Society, Division of
Organic Coatings and Plastics Chemistry, Chicago, Illinois-,
35(2): 416-418, August 25-29, 1975
3. Rowson, H.M., "Design Considerations in Solvent Recovery."
Courtaulds Engineering Ltd., Conventry, England. Proceedings
of the Metropolitan Engineers' Council on Air Resources
(MECAR) Symposium on New Developments in Air Pollution Control
New York, October 23, 1961, pp. 110-128
4. Cooper, J.C. and F.T. Cuniff, "Control of Solvent Emissions."
Pittsburgh Activated Carbon Co., Pittsburgh, PA. Proceedings
of the Metropolitan Engineers' Council on Air Resources
(MECAR) Symposium on New Developments in Air Pollution
Control, New York, October 23, 1961, pp. 30-41
5. Meyer, W.R., "Solvent Broke." Vulcan-Cincinnati, Inc.,
Cincinnati, Ohio. Proceedings of TAPPI Testing Paper
Synthetics Conference, Boston, Mass., October 7-9, 1974,
pp. 109-11.5
6. Shaw, Nathan R., "Vapor Adsorption Technology for Recovery
of Chlorinated Hydrocarbons and Other Solvents." Preprint
of Paper Presented at the 80th Annual Meeting of the Air
Pollution Control Association, Boston, Massachusetts,
. Paper 75-123: June 15-20, 1975
90
-------
:: .... BACKGROUND, DOCUMENT '.. • ;, ,' :
SECTION; 4.5-WASTE .SOLVENT.RECLAMATION . ' •
The only available document that lists emission factors for
waste solvent reclamation is Monsanto Research Corporation's
Source Assessment: Reclaiming of Waste Solvents - State of.the
Art (Reference 1).. Accordingly, Table 4.5-1 of Section 4..5 .is .
. •• ' • o • •*
'adapted from Table 6 of this .reference. *, ,-. . ^;::..'vf.- .',
.Footnote a of Table 4i5-1-states that the data in this table
were obtained "from state air pollution control agencies and
presuryey ..sampling."'These agencies are not - documented .elsewhere
in Reference 1.. Appendix B of'Reference 1 is entitled "Results
and Sample Calculations for Pr.esurvey Sampling of Private '
Contractor Solvent Reclaiming Plant," but neither the storage tank
vent nor the condenser vent emission factors calculated therein
correspond to the factors published in Table 6 of the reportcand
reproduced in Table 4.5-1 of AP-42 Section 4.5.
It is possible that the information given in Appendix B of
Reference 1 does pertain to the emission factors that appear in
Table 6 of the same reference (i.e., Table 4.5-1 of AP-42 Section
4.5). However, if this is the case, the authors of Reference 1
have not made the connection clear.
The control efficiencies mentioned in Section 4.5.2 are given
in the references cited, which in no case provide background data
for these percentages.
The amount of source test data is not specified in Reference
1. The process is highly variable and depends on the particular
solvent and recovery scheme used. The data on which material bal-
• ance calculations could be based is also not specified in Re.fer-
.ence 1. For these reasons the rating of p was assigned to all
«• j r
Demission factors in Table 4.5-1.
91
-------
4.-. 6 " SOLVENT DEGREASING ,
12
4.6.1 Process.Description '
Solvent degreasing (or solvent cleaning) is the physical
process of removing grease, fats, oils, wax, or soil from
various metal, glass, and plastic items with organic solvents.
The types of equipment used in this method are categorized as
cold cleaners, open top vapor degreasers, or conveyorized
•begreasers. . Nonaqueous solvents such as petrol eum;distill;ates,
chlorinated'hydrocarbons, ketones, and alcohol are used. ., .
Solvent selection; is based on the solubility of the soil and
the toxicityvfflammability, flash point, evaporation rate;; .-•'
boiling point, cost, and several other properties of the :' '•'
solvent. "
The metal working industry is the major user of solvent
degreasing; examples are the automotive, electronics, plumbing,
aircraft, refrigeration, and:business machine industries.
Solvent cleaning is also used in industries such as printing,
chemicals, plastics, rubber, textiles, glass, paper, and
electric power. Most transportation vehicle and electric
tool repair stations utilize solvent cleaning at least part
of the time. Many industries also use water-based alkaline
wash systems for degreasing. These systems emit no solvent
vapors to the atmosphere and are therefore not included in this
discussion.
4.6.1.1 Cold Cleaners - Cold cleaners are batch loaded, non-
boiling solvent degreasers, and are usually the simplest, least
expensive method of metal cleaning, the two basic types are
maintenance and manufacturing. Maintenance cold cleaners are
92
-------
more numerous, smaller, and generally use petroleum solvents
such as mineral spirits (petroleum distillates and Stoddard
solvents). Manufacturing cold cleaners use a wide variety of
solvents; they perform higher quality cleaning, are more
specialized, and.have about twice the average emission rate
of maintenance cold cleaners. Some cold cleaners can serve both
purposes.
Cold cleaner operations include spraying, brushing, flushing,
and immersion. .In a typical maintenance cleaner (Figure 4.6-1),
dirty parts are cleaned manually by spraying and then soaking in
the tank. After cleaning, the parts are either suspended over
the tank, to drain or placed on an external rack that routes the
drained solvent back into the cleaner. The cover is intended to
be closed whenever parts are not being handled in the cleaner.
Manufacturing cold cleaners vary widely in design, but there are
two basic tank designs: the simple spray sink and the dip tank.
The dip tank provides more thorough cleaning through soaking .and
often Is made to improve cleaning efficiency through agitation.
4.6.1.2 Open Top Vapor Systems - Open top vapor degreasers are
batch loaded, boiling degreasers that clean through the
condensation of hot solvent vapor on colder metal parts. Vapor
degreasing uses halogenated solvents (usually perch!oroethylene,
trichloroethylene, or 1,1,1-trichloroethane) because they are
not flammable and their vapors are much heavier than air.
A typical vapor degreaser (Figure 4.6-1) is a sump
containing a heater that boil's the solvent to generate vapors.
The upper level of these pure vapors is controlled by condenser
coils and/or by a water jacket encircling the device. A
"freeboard" extends above the top of the vapor zone to
93
-------
VD
'AGITATIONS — •£***=«——
• i
rx*^'
( 2 JCABRYflOT
©
) WASTE
SOLVENT
Cold cleaner
Open top vapor degreaser
Figure 4.6-1. Degreaser Emission Points
-------
minimize vapor escape. Parts to be cleaned are immersed in the
vapor zone; condensation continues until they are heated to the
vapor temperature. Residual liquid solvent on -the parts
rapidly evaporates as they are slowly removed from the vapor
zone. Cleaning action is often Increased by spraying the parts
with solvent (below the vapor level) or by immersing them into
the liquid solvent bath. Nearly all vapor degreasers are
equipped with a water separator which, allows the solvent to
flow back into the degreaser; many also have lip-mounted
exhaust systems which capture solvent vapors and carry them
away from operating personnel.
4.6.1.3 Conveyorized Degreasers - Conveyorized degreasers may
operate with either cold or vaporized solvent, but merit
separate consideration because they are continuously loaded
and almost always hooded or enclosed. About 85 percent are
vapor types and 15 percent are nonboiling.
4.6.2 Emissions and Controls1'2'3
Emissions from cold cleaners occur through (1) waste
solvent evaporation, (2) solvent carry-out (evaporation from
wet parts), (3) solvent bath evaporation, (4) spray evaporation,
and (5) agitation (Figure 4.6-1). Waste solvent loss, cold
cleaning's greatest emission source, can be minimized through
distillation and special incineration plants; draining cleaned
parts for at least 15 seconds reduces carry-out emissions. Bath
evaporation can be controlled by regularly using a cover, allowing
an adequate freeboard height, and avoiding excessive drafts in the
workshop. If the solvent used is insoluble in and heavier than
water, a layer of water about two to four inches thick covering
the halogenated solvent can also reduce bath evaporation. This
95
-------
is known as a "water cover." Spraying at low pressure helps reduce
solvent loss from this part of the process. Agitation emissions can be
controlled through use of a cover, agitating no longer than necessary,
and avoiding the use of agitation with low volatility solvents. Emissions
of low volatility solvents increase significantly with agitation; however,
contrary to what one might expect, a.gitation causes only a small increase
in emissions of high volatility solvents. Solvent type, particularly its
volatility at the operating temperature, is the variable which most
affects cold cleaner emission rates.
As with cold cleaning, open top vapor degreasing emissions
depend heavily on proper operating methods. Most emissions are due
to (6) diffusion and convection, which can be minimized by using an
automated cover, regularly using a manual cover, spraying below
the vapor level, optimizing work loads, or using a refrigerated
freeboard chiller (for which a carbon adsorption unit would be sub-
stituted on larger units). Safety switches and thermostats that
prevent emissions during malfunctions and abnormal operation also
reduce diffusion and convection from the vaporized solvent. Addi-
tional sources are (7) solvent carry-out, (8) exhaust systems, and
(9) waste solvent evaporation (Figure 4.6-1). Carry-out is directly
affected by the size and shape of the workload, racking of parts,
and cleaning and drying time. Exhaust emissions can be nearly
eliminated by a carbon adsorber that collects the waste solvent
for reuse. Waste solvent evaporation is not as much of a problem
with vapor degreasers as it is for cold cleaners because the halo-
genated solvents used are often distilled and recycled by solvent
recovery systems.
Because of their large workload capacity and the fact that
they are usually enclosed, conveyorized degreasers emit less solvent
per part cleaned than either of the other two types of degreaser.
Compared to operating practices, design and adjustment are major
factors affecting emissions, the main source of which is carry-
out of vapor and liquid solvent.
96
-------
Emission rates are usually estimated from solvent consumption
data for the particular degreasing operation under consideration.
Solvents are often purchased specifically for use in degreasing
and are not used in any other plant operations; in these cases,
purchase records provide the necessary information and an
emission factor of 1,000 kg of volatile organic emissions per
metric ton of solvent can be applied (Table 4.6-1). This
factor is based on the assumption that solvent consumption
equals emissions. When information on solvent consumption
is not available, emission rates can be estimated if the
number and type of degreasing units are known. The factors in
Table 4.6-1 are based on the number of degreasers and emissions
produced nationwide and therefore may be considerably in error
when applied to one particular unit.
The expected effectiveness of various control devices and
procedures are listed in Table 4.6-2. As a first approximation,
these efficiencies can be applied without regard for the
specific solvent being used; however, efficiencies are generally
higher for more volatile solvents. These solvents also result in
higher emission rates than those computed from the "average"
factors listed in Table 4.6-1.
References for Section 4.6
1. Marn, P.O.; Hoogheem, T.J.; Horn, D.A.; and Hughes, T.W.
Source Assessment: Solvent Evaporation-Degreasing. Draft
document. Monsanto Research Corporation, Dayton, Ohio.
Prepared for U.S. Environmental Protection Agency, Research
Triangle Park, N.C. 27711, under Contract No. 68-02-1874,
January, 1977.
2. Control of Volatile Organic Emissions from Solvent Metal
Cleaning. EPA-450/2-77-022 COAPQS No. 1.2-079). U.S.
Environmental Protection Agency, Research Triangle Park,
N.C. 27711, November 1977.
97
-------
3. Suprenant, K.S. and Richards, D.W. Study to Support New
Source Performance Standards for Solvent Metal Cleaning
Operations, Final and Appendix Reports. Dow Chemical
Company, Midland, Mich. Prepared for U.S. Environmental
Protection Agency, Research Triangle Park, N.C. 27711, .
under Contract No. 68-02-1329, Task No. 9, June 30, 1976.
98
-------
Table 4.6-1. SOLVENT LOSS EMISSION FACTORS FOR DECREASING OPERATIONS
EMISSION FACTOR RATING: C
Type of degreaslng
Activity measure
Uncontrolled organic
emission factor3
CO
All0
Cold cleaner
Entire unit0
Waste solvent loss
Solvent carryout
Bath and spray evaporation
Entire unit
Open top vapor
Entire unit
Entire unit
ConveyoHzed, vapor
Entire unit
Conveyorized, nonbolling
Entire unit
Solvent consumed
Units In operation
Surface area and duty
cycle
Units 1n operation
Surface area and duty
cycle
Units in operation
Units In operation
2,000 Ib/ton
0.33 tons/yr-unlt
0.18 tons/yr-unit
0.08 tons/yr-unlt
0.07 tons/yr.aj)1t
a. 08 Ib/hr. ft2
10.5 tons/yr. unit
0.15 Ib/hr- ft*.
26 tons/yr. unit
52 tons/yr.unit
1,000 kg/MT
0.30 MT/yr-unlt
0.165 MT/yr.unit
0.075 MT/yr-unit
0.060 MT/yr-unlt
Q.4 kg/hr. m2- 9
9.5 MT/yr-unit
0.7 kg/hr-m2e
24 MT/yr-unit
47 MT/yr.unit
b
100 percent nonmethane hydrocarbons or volatile organic compounds.
Solvent consumptionn data will provide much more accurate emission
estimates than any of the other factors presented.
Emissions would generally be higher for manufacturing units and
lower for maintenance units.
For trlchloroethane degreaser. From Reference 3, Appendix C-6.
Tor trlchloroethane degreaser. Does not include waste solvent losses.
-------
Table 4.6-2. PROJECTED EMISSION REDUCTION FACTORS FOR SOLVENT DECREASING3
System Scenario
o
o
System
Control Devices
cover or enclosed design
drainage facility
water cover, refrigerated chiller, carbon adsorption
or high freeboard"
solid, fluid spray streanr
safety switches and thermostats
emission reduction from control devices
Operating Procedures
proper use of equipment
use of high volatility solvent
waste solvent reclamation
reduced exhaust ventilation
reduced conveyer or entry speed
emission reduction from operating procedures
Total Emission Reduction (percentage)
Cold
Cleaner
A
X
X
13-38
X
X
15-15
28-83d
8
X
X
X
NA*
X
X
X
HA*
55-«9f
Vapor
Oegreaser
C
X
X
20-40
X
X
X
X
15-35
30-60
D
X
X
X
30-60
X
X
X
X
20-40
45-75
Conveyor 1 zed
Oegreaser
E
X
X
X
X
X
20-30
20-30
F
X
X
X
X
40-60
X
X
X
X
20-30
50-70
'Reference 2. Ranges of emission reduction represent poor to excellent compliance.
bOnly one of these major control devices would be used 1n any degress1ng system. System 8 could employ any of
them; system D could employ any except water cover; system F could employ any except water cover and high
freeboard.
clf agitation by spraying Is used, the spray should not be a shower type spray.
dA manual or mechanically assisted cover would contribute 6-18* reduction;draining parts 15 seconds within the
degr'easer, 7-201; and storing waste solvent In containers, an additional 15-451.
Breakdown between control equipment and operating procedures Is not available.
Percentages represent average compliance.
-------
BACKGROUND DOCUMENT
SECTION 4.6 SOLVENT DECREASING
1.0 INTRODUCTION
The Information on emissions from solvent degreasing was
developed from (1) telephone surveys of responsible persons in
industries engaged in degreasing, and (2) a series of studies
.carried out under EPA sponsorship to evaluate various emission
control devices. Information from both sources was used to
compute emission factors. A detailed discussion of these compu-
tations follows.
2.0 FACTORS BASED ON SURVEY DATA
A nationwide survey of metal working industries was con-
ducted by the Dow Chemical Company to obtain information on
emissions from solvent degreasing CReference 3, Appendix Report).
Metal working Industries employing 20 or more people were surveyed
within the following industrial classifications:
• Furniture and fixtures
• Primary metal industries
• Fabricated metal products
• Machinery (except electrical)
• Electrical and electronic equipment
• Transportation equipment
• Instruments and related products
• Miscellaneous manufacturing industries
A representative sample of 2,578 out of a possible 41,670
plants was surveyed by telephone to ascertain the types of de-
greasers employed, the specific solvents used, and the quantities
consumed and disposed of by various routes. The results of this
survey were as follows:
101
-------
• Number of maintenance cold cleaners in the
U.S. in 1974 880,000
• Number of manufacturing cold cleaners 340,000
• Number of open top vapor degreasers 21,000
• Number of conveyorized vapor degreasers 3,170
• Number of conveyorized nonboiling degreasers 530
The estimates of solvent usage from the survey were compared with
similar estimates from other surveys (Table 1), and a weighted
average solvent consumption for each type of degreaser was computed.
2.1 COLD CLEANING
3
Of the 450 x 10 MT/yr of solvents consumed in cold cleaning,
3
25 x 10 MT/yr are used in wiping operations, which are not con-
o
sidered cold cleaner emissions, and another 25 x 10 MT/yr are
used in conveyorized nonboiling degreasers, which are considered
separately. Another 20 x 10 MT/yr are incinerated or landfilled
after use in such a manner that no emissions occur. Thus,
380 x 10 MT/yr of solvents find their way into the atmosphere
as a result of cold cleaning operations. The average emissions
per unit can be computed as
380 x IP3 MT/yr
880,000 + 340,000 =0.31 MT/yr per
maintenance units manufacturing units cold cleaning unit
If it is assumed that manufacturing cold cleaners have twice
the emission rate of maintenance cold cleaners, then by simple
.algebra manufacturing units average 0.48 MT/yr and maintenance
units average 0.24 MT/yr. By using their best engineering
judgement, the authors of Reference 2 estimated that 55 percent
of cold cleaning emissions are due to waste solvent evaporation,
20 percent to bath and spray evaporation, and 25 percent to
carryout losses. The emission factors for each part of the cold
102
-------
Table 1. U.S. CONSUMPTION OF DECREASING SOLVENTS1
1974 (10 metric tons/year)
o
to
Solvent Type
Halogenated
Trichloroethylene
1,1.1 trichloroethane
Perchloroethylene
Methylene chloride
Trichlorotrif luoroethane
Total
Aliphatic
Aromatic
Benzene
Toluene
Xylene
Cyclohexane
Heavy aromatics
Total
Oxygenated
Ke tones
Acetone
Methyl ethyl ketone
Alcohols
Butyl
Ethers
Total
TOTAL
Breakdown:
Vapor deg. solvents
Cold clean, solvents
Weighted
Average
yp+CC=Tota1
128 + 25 153
80 + 82 162
41 + 13 54
7+23 30
20 + 10 30
276 +153 429
222
7
14
12
1
12
4~6
10
8
5
6
29
726
- VO = 276*275
= CC = 153 + 222
Honsanto-S.A.D.
Tom Hoogheem-1974
VD + CC = Total
157
90 + 78 - 168
43 + 11 - 54
10 + 46 * 56
17
-
225
7
14
12
1
-
-
10
7.5
3.3
6
ISTff
Expected
+10
+ 46 + 29 +153
U.S. Tariff COM.
Report for 1974
VO + CC = Total
142 + 8 - 150
73 +106 - 179
40 4 19 * 59
7 + 18 - 25
_
-
No
data
No
data
Accuracy :
percent -*• VO
1 t 301 + 50X -— CC
Dow Final 1
Survey for
VD + CC -
103 + 39
110+63
41 * 9
7.5 +6.3
34 + 18
296 +135 =
No
data
10
8
7
H
-
= 275 t 25 »
= (155 t 25)
(eport
1974
Total
142
173
50
13.8
52
431
Ranges
250 to
+ (220
Dow
Chart
for
1974
VD only
143
73
40
9
20
285
No
data
No
data
300 (x 103
+ 65) + (75
Detrex
1975
y_p
114
63
45
8
20
250
No
data
1
No
data
netrlc
+ 35)
Estimates
Projected
19/4
yp
124
53
40
6
18
247
No
data
ton/yr)
J.S. Gunnln
Shell Chemical
Solvent Bus. Ctr
No data
218
12
No data
« 153 + 222 + 75
= 153 + 297
= 450
"From Reference 2 (EPA 450/2-77-022)
(130 + 155 + 38) to (180 + 285 + 112)
323 to 557
450 + 127
-------
cleaning process were derived by applying these percentages
to the overall factor for the entire unit. Emission factors
were not computed for individual solvent types, but represent
composite factors for all solvents.
2.2 OPEN TOP VAPOR DECREASING
3
According to Table 1, 275 x 10 MT/yr of solvents are used
in vapor degreasing. Based on information from the survey and
best engineering judgement, the authors of Reference 2 estimated
that 75 x 10 MT/yr are from conveyorized vapor degreasing, and
the remaining 200 x 10 MT/yr from open top vapor degreasing.
The emissions per unit were computed as
= 9.5 MT/yr per open top"
vapor degreasing unit
No attempt was made to separate this emission factor into sub-
factors that represent various parts of the vapor degreasing
operation (as was done for cold cleaning).
2.3 CONVEYORIZED DEGREASING
Solvent usage in conveyorized nonboiling and vapor de-
greasing was estimated to be 25 x 10 MT/yr and 75 x 10 MT/yr,
respectively (refer to Sections 2.1 and 2.2). Emission factors
for each type of unit were computed as follows:
units •
23-7
As in the previous cases, these factors apply to composite
average units and solvent types.
104
-------
3.0 FACTORS BASED ON TEST DATA
One study (Reference 3, Appendix C-6) was carried out on a
Baron Blakeslee Model HD 425 vapor degreaser, which was operated
both as a cold cleaner and as a vapor degreaser by the Presto!ite
Corporation at Bay City, Michigan, using 1,1,1 trichloroethane
as the solvent. When used as a vapor degreaser for one year, the
unit operated for 16 hours per day for 250 days and consumed
550 gallons (2,737 kg) of solvent. When used as a cold cleaner
for 25 days at 16 hours per day, the unit consumer 18 gallons
(90 kg). No waste solvent disposal was carried out during the
25 day test, so these losses are not included as they are in the
250 day test of vapor degreasing. These figures equate to 1.5
Ib/hr for the vapor degreaser and 0.49 Ib/hr for the cold cleaner.
Waste solvent disposal was estimated to add 0.3 Ib/hr for the-cold
cleaner.
The cleaner used for this test had a surface area of 10.2
? 2
ft (.95 m ), so the emissions were calculated to be 0.048 Ib/hr-
ft (.23 kg/hr-m2) for the cold cleaner without waste solvent
disposal, and 0.077 lb/hr«ft2 (.37 kg/hr-m2) for the cold cleaner
Including the estimated emissions from waste solvent disposal.
Emissions from the vapor degreasing test are calculated as 0.15
Ib/hr.ft2 (.71 kg/hr-m2) for the entire operation. The factors
listed in AP-42, Table 4.6-1 are based entirely on the Presto!ite
Corporation tests for the applications where activity measures
were given in terms of the unit's surface area and duty cycle
(Reference 3, Appendix C-6).
Evaporation rate tests were also conducted for two cold cleaners
under a variety of operating conditions for periods of 16 to
164 hours, and loss rates of 0.006 to 0.59 Ib/hr.ft2 (.003 to
2.8 kg/hr-m2) were measured (Reference 1, Appendix A-2). These
tests illustrate the extreme range of emission rates that may be
encountered; they did not define typical, real-world operation and
therefore were not used in the development of emission factors.
105
-------
4.0 EMISSION REDUCTION FACTORS
The authors of Reference 2 made the emission reduction
factor estimates given in AP-42, Table 4.6-2 by combining
their best engineering judgment and the results of the tests
/
summarized in Table 2.
5.0 EMISSION FACTOR RATING
The conventional rating system gives the following
weighting to the various information categories: measured
emission data, 20 points maximum; process data, 10 points
maximum; engineering analysis, 10 points maximum. In the case
of solvent degreasing, where the most accurate estimate of
emissions is based on a material balance, less importance was
assigned to measured emission data because they do not
appreciably increase the accuracy of an emissions estimate based
on solvent consumption. The following rating system was used:
Maximum Solvent
Points Degreasing
Measured emission data 10 4
Process data 10 6
Engineering analysis 20_ 1_5_
TOTAL: 40 25
< ,
The emission factor rating for 25 points is C.
106
-------
Table 2. TEST RESULTS FROM DOW REPORT*
PM Plppr!*
C-l
C-l
C-ll
C-l
C-M
C-l
C-ll
UPHll.rictp.
C-l
l-t
C-4
C-l
•M.
Pratt 1 IPtllMy
Ut»
ta>CP>
•Mlm
SIM4BH
Ik
kfelift IKk
tt llpctrfc
.te!aUn!«s
Pratt 1 knit**;
tapir MlltPT
It!:,.*
Jl.
IMVIPP
PW.Mr*
tin
«•>
•IPJ
iw. nil
otn
(out
««
CH-
UB
nn
•m
cm
c»
crpiira*
fpppr ATM
U • lit IP.
W.« III
,
M.I • n IP.
1.1 ni
M.I • II IP.
1.1 «ii
n ii1
u. i ii*
11 • i.i n.
M f|l
«i.i it*
.
H • » U.
t • l| rt.
it n*
tPlM
I.I.I
Irl.
I.I.I
I.I.I
I.I.I
•siea
Mtbta
cklirMt
Irl.
NR.
Irl.
I.I.I
P-.
I.I.I
Irl.
IMpplrplM 1
K.I/.II)
II. 1 VIA*
in IWIp>
III IWUP
•!n
clrCMll kMr«
!!.• Itl/Mpiy
MlklM
WP.I/.W
••»•"»*—
,,„,„,
VJ S f.lM
ft.
Itk/ll'.krl
•.U
,
.
•.in
•.in
•.in
• m
•.m
».m
•.m
•.it
_
•-!•
I.U
__ (MnUpUi
(filtetl)
».« ptlM
H IWMP
M PV(P»
;:«S!SS"
l H winpfe
i.ii ni/*i»
• tilM
I.I pl/PPU
•.•n pi/it'
H.I til/kPii
«»I/M
»..«.,
* •* """^
I.Pt |kl/V
«.l HIM
u.
ll/r|!.pr)
•••
•.in
•.in
*.H>
I.BI
•.111
•.0
•.!»
•.M
_
•.in
I.H
CotrpllPt
iiikta*cr
«•
IB
g
IPI
.
«"
2
m
ta
M>
•m
in
m
MS
ClPlrpI Jr>'—
C^^—i II...I, p^W
CPwr (MW«I)
n ••.>
• •.n
n- «.i
• rt
l •
MM trif
C*H irpj
I! • •.!» (WMT ikniw
C.~~-~P.I.
CMItar
UU tnr
C«rtai pfterpltap
UTW-^rM..
Ur*w *4i«fVII«l
Urtop wt«rpll~
CnM.
IWP«f«< Ipr H "»/»T «rf 1 »>M
•• «ck prfprpilltP M tta tnt
UK IIP ««t IpMll. •**•!• pt mill pp»M IP pppr
p»ti pi • It « IP tU
ft - Ml. CPU Irapfcilo ttill* kprp «PI r«»«rlp< II .Pi
;p!rL"7..^1''yrt,r\;!j.i>lt'™*1 *-fn"-1
l» ep.trpl pliliKor «' "• Hnr»tlpp trU" '• IkP^rt U W
Hltirlil kplpKI rpMltl
•tttltt frpp pirckpw ratpr^l
: nn • *~ i«r IM
I«f rivw DrfrMtwr. Cff - tMv.wrtra.) f-vv tefrMitr. CM • Cw
» • 0»| *yt«itr ft U «i>«rtlt«*. wttt • Mr4l.>« .Uj. CI • C«U Ii
it . :r! ***'• lh/l°* * P0*"*1 •' »'«•» ••UUt Hr IM of (Mrl* tf4«9rt»ti«. WO -
l.i.l " •tk.rl ckl«r«fani. l»t. • trUk>oro«Urrlt>..«, ftft. • ( • •
• tf tto aM Hffari «twrtto* Mck U«l l« .tmUII.
-------
4.7 TANK AND DRUM CLEANING
4.7.1 Process Description
Rail tank cars, tank trucks, and drums are used to transport
about 700 different commodities. The interiors of rail cars, and
most tank trucks and drums, are in dedicated service (carrying one
commodity only) and, unless contaminated, are cleaned only prior
to repair or testing. Nondedicated tank trucks (about 20,000 or
22 percent of the total in service) and drums (approximately 5.6
million or 12.5 percent of the total) are cleaned after every trip.
4.7.1.1 Rail Tank Cars - Most rail tank cars are privately owned.
Some of these, like rail road-owned cars, are operated on a
for-hire basis. The commodities hauled (and cleaned) are 35
percent petroleum products (excluding gasoline, fuel oils, and
lubricating oils), 20 percent organic chemicals, 25 percent
inorganic chemicals, 15 percent compressed gases, and 5 percent
food products.
Much tank car cleaning is conducted at shipping and receiving
terminals, where the wastes go to the manufacturers' treatment
systems. However, 30 to 40 percent is done at service stations
operated by tank car owner-lessors. These installations clean out
wastes derived from a wide variety of commodities, many of which
require special cleaning methods.
A typical tank car cleaning facility cleans 4 to 10 cars per
day. Capacity per car varies from 38 to 129 m3 (10,000 to
34,000 gal). Cleaning agents include steam, water, detergents,
and solvents, which are applied using steam hoses, pressure wands,
or rotating spray heads placed through the opening in the top of
the car. Scraping of hardened or crystallized products is often
necessary. Cars carrying gases and volatile materials and those
that are to be pressure tested must be filled or flushed with
108
-------
water. The average amount of residual material cleaned from each
car is estimated to be 250 kg (550 Ib). Vapors from cleaning cars
that are not flared or dissolved in water are dissipated to the
atmosphere.
4.7.1.2 Tank Trucks - Two thirds of the tank trucks in service in
the United States are operated on a for-hire basis; of these, 80
percent are used to haul bulk liquids. Most companies operate
fleets of five trucks or less; wherever possible, these trucks are
consigned to dedicated service. Commodities hauled and cleaned
are 15 percent petroleum products (excluding gasoline, fuel oils,
and lubricating oils), 35 percent organic chemicals, 5 percent
food products, and 10 percent other (paint, inks, naval stores,
and so son).
Interior washing is carried out at many tank truck dispatch
terminals. Cleaning agents include water, steam, detergents,
caustic, acid, and solvents, which are applied with hand-held
pressure wands or by Turco or Butterworth rotating spray nozzles.
Detergent, caustic, and acid solutions are usually recycled until
spent and then sent to treatment facilities; solvents are recycled
in a closed system, with sludges either incinerated or landfilled.
The average amount of material cleaned from each trailer is 100 kg
(220 Ib). Vapors from volatile materials are flared at a few
terminals, but are most commonly dissipated to the atmosphere.
Approximately 0.23 m3 (60 gal) of liquid are used per tank truck
steam cleaning and 20.9 nr* (5,500 gal) for full flushing; this
represents an average of 2 m3 (500 gal) for tank truck cleaning.
4.7.1.3 Drums - Both 0.2 and 0.11 nr* (55 and 30 gal) drums are
used to ship a vast variety of commodities, with organic chemicals
(including solvents) accounting for 50 percent. The remaining 50
109
-------
percent include inorganic chemicals, asphaltic materials, elasto-
meric materials, printing inks, prints, food additives, fuel oils,
and so on.
Drums made entirely of 18-gauge steel have an average life,
with total cleaning, of eight trips; those with 20-gauge bodies and
18-gauge heads have an average life of three trips. Not all drums
are cleaned, especially those of thinner construction.
Tight-head drums which have carried matarials that are easy to
clean are steamed or washed with caustic. Steam cleaning is done
by inserting a nozzle into the drum, with vapors going .to the
atmosphere. Caustic washing is done by tumbling the drum with a
charge of hot caustic solution and some pieces of chain.
Drums used to carry materials that are difficult to clean are
burned out either in a furnace or in the open; those with tight-
heads have the tops cut out and are reconditioned as open-head
drums. Drum burning furnaces may be batch or continuous. Several
gas burners completely bathe the drum in flame, burning away the
contents, lining, and outside paint within a nominal 4-minute
period at a temperature of at least 480° but not more than 540°C
(900°F-1000°F) in order to prevent warping of the drum. Emissions
are vented to an afterburner or secondary combustion chamber, where
the gases are raised to at least 760°C (1,500°F) for a minimum of
0.5 second. The average amount of material removed from each drum
is 2 kg (4.4 Ib).
4.7.2 Emissions and Controls
4.7.2.1 Rail Tank Cars and Tank Trucks - Atmospheric emissions
from tank car and truck cleaning are predominantly volatile organic
chemical vapors. In order to achieve a practical but representa-
tive picture of these emissions, the organic chemicals hauled by
110
-------
the carriers must be broken down into classes characterized by
high, medium, and low viscosities and by high, medium, and low
vapor pressures. This is because high viscosity materials do not
drain readily, thus affecting the quantity of material remaining in
the tank, while high vapor pressure materials volatilize more
readily during cleaning and tend to lead to higher emissions.
Practical and economically feasible control of atmospheric
emissions from tank car and truck cleaning does not exist except
for containers transporting commodities that produce combustible
gases and water-soluble vapors (such as ammonia and chlorine).
Gases which are displaced as tanks are filled are sent to a flare
and burned. Water soluble vapors are absorbed in water and sent to
the wastewater system. All other emissions are vented to the
atmosphere.
Tables 4.7-1 and 4.7-2 give emission factors for representa-
tive organic chemicals hauled by tank cars and trucks, respectively.
Table 4.7-1.
EMISSION FACTORS FOR RAIL TANK CAR CLEANING1
EMISSION FACTOR RATING: D
Compound
Ethyl ene glycol
Chlorobenzene
o-Dichlorobenzene
Creosote
Chemical Class
Vapor
Pressure
low
medium
low
low
Viscosity
high
medium
medium
high
Total
Emissions3
(g/car)
0.32 -
15.7
75.4
2,350
a Total emissions = (emission rate) x (emission volume)
111
-------
Table 4.7-2.
EMISSION FACTORS FOR TANK TRUCK CLEANING1
EMISSION FACTOR RATING: D
Compound
Acetone
Perch! ordethyl ene
Methyl methacryl ate
Phenol
Propylene glycol
Chemical Class
Vapor
Pressure
high
high
medium
low
low
Viscosity
low
low
med i urn
low
high
Total
Emissions3
(g/truck)
311
215
32.4
5.5
1.07
a Total emissions = (emission rate) x (emission volume)
4.7.2.2 Drums - There is no control for emissions from steaming of
drums. Solution or caustic Washing yields negligible air emissions
because the drum is closed during the wash cycle. Atmospheric
emissions from steaming or washing drums are predominantly organic
chemical vapors.
Air emissions from drum turning furnaces are controlled by
proper operation of the afterburner or secondary combustion cham-
ber, where gases are raised to at least 760°C (1,500°F) for a
minimum of 0.5 second. This normally ensures complete combustion
of organic materials and prevents the formation, and subsequent
release, of large quantities of NOX, CO, and particulates. In
open burning, however, there is no feasible way of controlling the
release of incompletely burned combustion products to the atmos-
phere. Converting open cleaning operations to closed-cycle
cleaning and eliminating open air drum burning seem to be the only
control alternatives for the immediate future.
Table 4.7-3 gives emission factors for representative criteria
pollutants emitted from drum cleaning.
112
-------
Table 4.7-3. EMISSION FACTORS FOR DRUM BURNING1
EMISSION FACTOR RATING: E
Criteria
Pollutant
Participates
N0x
Volatile organics
Total Emissions3
Uncontrolled
(g/drum)
16
0.89
negl igible
Controlled
(g/drum)
12b
0.018
negl igible
a Total emissions = (emission rate) x (emission volume)
° Derived from Reference 1, Table 17 and Appendix A
Reference for Section 4.7
1. Earley, D.R., K.M. Tackett, and T.R. Blackwood, Source
Assessment: Rail Tank Car, Tank Truck, and Drum Cleaning
- State of the Art. EPA-600/2-78-004g. Prepared by
Monsanto Research Corporation, Dayton, Ohio 45407, for
USEPA, Industrial Environmental Research Laboratory,
Office of Research and Development, Cincinnati, Ohio
45268, under Contract No. 68-02-1874, April 1978
113
-------
BACKGROUND DOCUMENT
SECTION 4.7 TANK AND DRUM CLEANING
1.0 INTRODUCTION
The only document readily available for use in writing AP-42
Section 4.7, "Tank and Drum Cleaning," was Monsanto Research
Corporation's Source Assessment: Rail Tank Car. Tank Truck, and
Drum Cleaning - State of the Art (Reference 1).
2.0 TANK CAR AND TRUCK CLEANING EMISSIONS
Tables 4.7-1 and -2 of Section 4.7 are adapted from Table 12
of Reference 1. The only explanation of the origin of the emission
measurements from tank car and truck cleaning that appear in this
table consists of the following text (Reference 1, p. 21):
In order to achieve a practical, but representative, picture
of these emissions, the organic chemicals hauled by the
carriers were broken down into classes by high, medium, and
low vapor pressures. Viscosity affects the quantity of
material remaining in the tank; low viscosity materials drain
readily while high viscosity materials do not. Vapor pressure
affects the air emissions since high vapor-pressure materials
volatile more readily during cleaning and tend to lead to
higher emission rates.
After the classes of chemicals had been established, the
selection of the particular chemical to be sampled for was
dictated by the specific materials which were being cleaned
during the sampling visits.
There is no data supporting the emission factors for rail tank
car cleaning (Table 4.7-1) or tank truck cleaning (Table 4.7-2)
reported in Reference 1. Since the solvents carried in either rail
cars or trucks are extremely variable, the emissions vary by orders
114
-------
of magnitude. The few individual solvents reported are only
roughly representative for broad classes; moreover, the trends are
in opposite directions in the two tables. Therefore the rating of
D was applied to all emission factors in the two tables.
3.0 DRUM BURNING EMISSIONS
3.1 P ARTICULATES
In Table 4.7-3, the controlled emission factor for particu-
lates from drum burning was derived by dividing the total emissions
for each state listed in Table 17 of Reference 1 (in metric tons
per year) by the number of barrels burned annually in that state.
In each case, a factor of approximately 12 grams per drum was
obtained. This agrees with the factor obtained by dividing total
emissions nationwide (119.6 metric tons per year) by total barrels
burned (10.1 million).
The emission factor of .12 grams of particulate per drum burned
is also calculated in Appendix A of Reference 1. This appendix
states that the particulate emission factor for auto incineration
with an afterburner is 0.68 kilograms per car, based on 113 kilo-
grams (250 Ib) of combustible material (CM) on a stripped car body
(refer to AP-42, p. 2.2-1). Assuming 2 kilograms of combustible
material per drum (refer to AP-40, p. 508) gives:
0.68 kg car 2 kg CM
P car 113 kg/CM drum
= 0.012 kg/drum
= 12 g/drum
Performing this calculation using the particulate emission
factor for uncontrolled auto incineration, namely 0.9 kilograms per
car (refer to AP-42, p. 2.2-1), yields an uncontrolled emission
factor of 16 grams of particulate per drum burned.
115
-------
3.2 NITROGEN OXIDES
Employing the nitrogen oxides emission factors for auto
incineration with an afterburner of 0.01 kilograms per car (refer
to AP-42, p. 2.2-1), Appendix A of Reference 1 gives the following
factor for NOX emissions from drum burning:
E ^ 0.01 kg car 2 kg CM
car 113 kg CM drum
= 1.8 x 10~4 kg/drum
= 0.018 g/drum
Use of the NOX emission factor for uncontrolled auto incin-
eration of 0.05 kilograms per car (refer to AP-42, p. 2.2-1) yields
a factor of 0.89 grams NOX emitted per drum burned.
3.3 EMISSION FACTOR RATING
The emission factors for particulates and NOX for drum
burning in Table 4.7-3 are based solely on an analogy to auto
incineration (refer to Appendix A of Reference 1). The state
particulate burning emissions data in Table 17 of that reference
are apparently computed from the factor derived from auto incin-
eration, since they are all the same, within round-off errors. The
auto incineration data has a rating of B. The analogy is poor,
however, because auto combustibles are largely fabric and plastics,
i
whereas drum residues are liquid or semi-sol id, foi[ the most part.
Since there are no direct data whatever for drum burning emissions,
a rating of E was assigned to the emission factors in Table 4.7-3.
116
-------
8.1 ASPHALT1C CONCRETE PLANTS
8.1.1 Process Description
Asphaltic concrete (asphaltic hot-mix) is a paving material which
consists of a combination of graded aggregate that is dried, heated,
and evenly coated with hot asphalt cement.
Asphalt hot-mix is produced by mixing hot, dry aggregate with
hot liquid asphalt cement in batch or continuous processes. In the
dryer drum process, another method of hot-mix asphalt production,
wet -aggregate is dried and mixed with hot liquid asphalt cement
simultaneously in the dryer. Since different applications require
different aggregate size distributions, the aggregate is segregated
by size and proportioned into the mix as required. In 1975, about
90 percent of the total U.S. production was conventional batch
process and most the the remainder was conventional continuous.
The dryer drum process comprised less than 3 percent of the total,
but most new construction favors this design. Plants may be either
permanent or portable.
8.1.1.1 Conventional Plants - Conventional plants produce finished
asphaltic concrete through either batch (Figure 8.1-1) or continuous
(Figure 8.2-2) aggregate mixing operations. The raw aggregate is
normally stock-piled near the plant at a location where the moisture
content will stabilize between 3 and 5 percent by weight.
As processing for either type of operation begins, the aggre-
gate is hauled from the storage piles and placed in the appropriate
hoppers of the cold-feed unit. The material is metered from the
hoppers onto a conveyor belt and is transported into a gas- or oil-
fired rotary dryer. Because a substantial portion of the heat is
117
-------
CO
EXHAUST TO I
ATMOSPHERE '
SECONDARY
COLLECTION
PRIMARY DUST
COLLECTOR
, HOT
\ SCREENS
FINE
AGGREGATE
STORAGE
PILE
COARSE
AGGREGATE
STORAGE
PILE
FINES
RETURN
LINEx,
WEIGH HOPPER
MIXER ,
ELEVATOR^
ROTARY
DRYER
ASPHALT
STORAGE
FEEDERS
CONVEYOR^
Figure 8.1-1. Batch Hot-Mix Asphalt Plant3
a Numbered locations are points of emission for substances
listed in Table 8.1-1.
-------
SECONDARY
COLLECTION
EXHAUST TO
ATMOSPHERE
PRIMARY DUST
COLLECTOR
DRAFT FAN (LOCATION
- DEPENDENT UPON
TYPE OF SECONDARY)
SJ )u— 'DEPENDENT UPON
't \ T
STORAGE
TANK
(OPTIONAL)
FEEDERS
CONVEYOR
'ELEVATORSS
Figure 8.1-2. Continuous Hot-Mix Asphalt Plant9
TRUCK
Numbered locations are points of emission for substances
listed in Table 8.1-1.
-------
transferred by radiation, dryers are equipped with flights that are
designed to tumble the aggregate and promote drying.
As it leaves the dryer, the hot material drops into a bucket
elevator and is transferred to a set of vibrating screens where it
is classified by size into as many as four different grades. The
classified hot materials then enter the mixing operation.
In a batch plant, the classified aggregate drops into one of
four large bins. The operator controls the aggregate size distribu-
tion by opening individual bins and allowing the classified aggregate
to drop into a weigh hopper until the desired weight is obtained.
After all the material is weighted out, the sized aggregates are
dropped into a mixer and mixed dry for about 30 seconds. The asphalt,
which is a solid at ambient temperatures, is pumped from heated
storage tanks, weighted, and then injected into the mixer. The hot-
mixed batch is then dropped into a truck and hauled to the job site.
In a continuous plant, the classified aggregate drops into a set
of small bins which collect and meter the classified aggregate to.the
mixer. From the hot bins, the aggregate is metered through a set of
feeder, conveyors to another bucket elevator and into the mixer.
Asphalt is metered into the inlet end of the mixer and retention
time is controlled by an adjustable dam at the end of the mixer.
The mix flows out of the mixer into a hopper from which the trucks
are loaded.
8.1.1.2 Dryer Drum Plants - The dryer drum process simplifies the
conventional process by replacing hot aggregate storage bins,
vibrating screens, and the mixer with proportioning feed controls.
Figure 8.1-3 is a diagram of the dryer drum process. Both
aggregate and asphalt are introduced near the flame end of the
120
-------
Bins
r\j
Burner and
Turbo-Cowpres sor
Variable Speed
Convenor Belts
Figure 8.1-3. Dryer Drum Hot-Mix Asphalt Plant3
From Reference 20. Numbered locations are points of emission for
substances listed 1n Table 8.1-1.
-------
revolving drum. A variable-flow asphalt pump is electronically
linked to the aggregate belt scales to control mix specifications.
Dryer drum plants generally use parallel flow design for hot
burner gases and aggregate flow. Parallel flow has the advantage
of giving the mixture a longer time to coat and collect dust in the
mix, thereby reducing particulate emissions to the atmosphere. The
amount of particulates generated within the dryer in this process is
lower than that generated within conventional dryers, but because
asphalt is heated to high temperatures for a long period of time,
organic emissions are greater.
The mix is discharged from the revolving dryer drum into
surge bins or storage silos.
8.1.2 Emissions and Controls
8.1.2.1 Emission Locations - Emission points at batch, continuous,
and drum dryer hot-mix asphalt plants are numbered in Figures 8.1-1,
-2, and -3, respectively.
Emissions from the various sources in an asphaltic concrete
plant are vented either through the dryer vent or the scavenger vent.
The dryer vent stream goes to the primary collector. The outputs
of the primary collector and the scavenger vent go to the secondary
collector, then to the stack (1) for release to the atmosphere.
The scavenger vent carries releases from the hot aggregate elevator
(5), vibrating screens (5), hot aggregate storage bins (5), weigh
hopper, and mixer (2). The dryer vent carries emissions only from.
the dryer. In the dryer drum process the screens, weigh hopper,.
and mixer are not in a separate tower. Dryer emissions in conven-
tional plants contain mineral fines and fuel combustion products,
122
-------
while the mixer assembly (2) also emits materials from the hot
asphalt. In dryer drum plants, both types of emissions arise in
the drum. These numbers refer to Figures 8.1-1 to 8.1-3.
Potential fugitive particulate emission sources from asphaltic
concrete plants include unloading of aggregate to aggregate storage
bins (5), conveying aggregate by elevators (5), and aggregate
screening operations (5). Another source of particulate emissions
is the mixer (2) which, although it is generally vented into the
secondary collector, is open to the atmosphere when a batch is
loaded onto a truck. This is an intermittent operation and the con-
ditions of air movement are quite variable, so these emissions are
best regarded as fugitive. The open truck (4) can also be a source
of fugitive VOC emissions, as can the asphalt storage tanks (3),
which may emit small amounts of polycyclics as well. These numbers
refer to Figures 8.1-1 to 8.1-3.
Thus fugitive particulate emissions from hot-mix asphalt plants
consist basically of dust from aggregate storage, handling, and
transfer. Stone dust may range from 0.1 /im to more than 300 Mm in
diameter. On the average, 5 percent of cold aggregate feed is
<4/im (minus 200 mesh). Dust that may escape before reaching pri-
mary dust collection generally is 50 to 70 percent <4 urn (minus 200
mesh). Materials emitted are given in Tables 8.1-1 and 8.1-4.
8.1.2.2 Emission Factors - Emission factors for various materials
emitted from the stack are given in Table 8.1-1. With the exception
of aldehydes, the materials listed in this table are also emitted
from the mixer, but mixer concentrations are 5- to 100-fold smaller
than stack concentrations and last only during the discharge of the
mixer. Reference 16 reports mixer concentrations of SOX, NOX, VOC,
and ozone as less than certain values, so they may not be present at
123
-------
Table 8.1-1. EMISSION FACTORS FOR SELECTED MATERIALS FROM
AN ASPHALTIC CONCRETE PLANT STACK16
Material Emitted3
Participate0
Sulfur oxides (as S02)c>d
Nitrogen oxides (as Nt^)6
Volatile organic compounds
(expressed as methane
equivalents)
Carbon monoxide6
Polycyclic organic
material
Aldehydes6
Formal dehyde
2-Methylpropanal
(isobutyral dehyde)
1-Butanal
(n-butyral dehyde)
3-Methylbutanal
(isovaleraldehyde)
Emission
Factor
Rating
B
C
D
D
D
D
D
D
D
D
D
Emission Factor
g/MT
137
146S
18
14
19
0.013
10
0.077
0.63
1.2
8.3
Ib/ton
.274
.292S
.036
.028
.038
.000026
.020
.00015
.0013
.0024 -
.016
Particulates, carbon monoxide, polycyclics, trace metals, and
hydrogen sulfide were observed in the mixer emissions at con-
centrations that were small relative to stack concentrations.
Refer to Section 8.1.2.2.
Expressed as grams per metric ton and Ib/ton of asphaltic
concrete produced.
c Mean of 400 plant survey source test results.
d S = percent sulfur in fuel. S02 may be attenuated more than
50 percent by adsorption on alkaline aggregate (Reference 21).
e Based on limited test data from the single asphaltic concrete
plant described in Table 8.1-2.
124
-------
all, while participates, carbon monoxide, polycyclics, track metals,
and hydrogen sulfide were observed at concentrations that were small
relative to stack concentrations. Emissions from the mixer are thus
best treated as fugitive.
The materials listed in Table 8.1-1 are discussed below.
Factor ratings are listed for each material in the table. All emis-
sion factors are for controlled operation, either based on average
industry practice shown by survey, or on actual results of testing
in a selected typical plant. The characteristics of this represen-
tative plant are given in Table 8.1-2.
The particulate emission factor was derived from a 400 plant
sample of the U.S. hot-mix asphalt industry. The extremes for this
sample range over three orders of magnitude, but 76 percent of the
plants have emission rates below the mean.
The industrial survey showed that over 66 percent of operating
hot-mix asphalt plants use fuel oil for combustion. Possible
sulfur oxides emissions from the stack were calculated assuming
that all sulfur in the fuel oil is oxidized to SO . (No. 2 fuel
rt
oil has an average sulfur content of 0.22 percent.) The amount of
sulfur oxides actually released through the stack may be attenuated
by water scrubbers or even by the aggregate itself, where limestone
is being dried.
Emission factors of nitrogen oxides, nonmethane volatile
organics, carbon monoxide, polycyclic organic material, and aldehydes
were determined by sampling the stack gas at the representative
asphalt hot-mix plant.
8.1.2.3 Emission Controls - The choice of applicable control equip-
ment ranges from dry mechanical collectors to scrubbers and fabric
125
-------
Table 8.1-2. CHARACTERISTICS OF AN ASPHALTIC CONCRETE
PLANT SELECTED FOR SAMPLING3
Parameter
Plant Sampled
Plant type
Production rate,
Metric tons/hr
(tons/hr)
Mixer capacity,
Metric tons (tons)
Primary collector
Secondary collector
Fuel
Release agent
Stack height, m (ft)
Particulate emission rate,
kg/hr
(Ib/hr)
Conventional permanent
batch plant
160.3 + 16 percent
(177 + 16 percent)
3.6 (4.0)
Cyclone
Wet scrubber (venturi)
Oil
Fuel oil
15.85 (52)
7.7 + 48 percent
(17.0 + 48 percent)
a
From Reference 16, Table 16
126
-------
collectors. Attempts to apply electrostatic precipitators have met
with little success. Practically all plants use primary dust collec-
tion equipment, such as large diameter cyclones, skimmers, or
settling chambers. These chambers are often used as classifiers
where the collected material is returned to the hot aggregate
elevator and combined with the dryer aggregate load. The air dis-
charge from the primary collector is seldom vented to the atmosphere
because high emission levels would result. The primary collector
effluent is therefore ducted to a secondary collection device.
Particulate emission factors for conventional asphaltic con-
crete plants are presented in Table 8.1-3. Particle size distribu-
tion information has not been included because the particle size
distribution varies with the aggregate being used, the mix being
made, and the type of plant operation. Potential fugitive particu-
late emission factors for conventional asphaltic concrete plants are
shown in Table 8.1-4.
Particulate emission factors for dryer drum plants are presented
in Table 8.1-5. (There are no data for other pollutants released
from the dryer drum hot-mix process.) Particle size distribution
has not been included because it varies with the aggregate used,
the mix made, and the type of plant operation. Emission factors
for particulates in an uncontrolled plant can vary by a factor of
10, depending upon the percent of fine particles in the aggregate.
127
-------
Table 8.1-3. PARTICULATE EMISSION FACTORS FOR CONVENTIONAL
HOT-MIX ASPHALTIC PLANTS2
EMISSION FACTOR RATING: B
Type of Control
Uncontrolled0 »d
Prec leaner
High-efficiency cyclone
Spray tower
Baffle spray tower
Multiple centrifugal scrubber
Orifice- type scrubber
Venturi scrubber
Baghouse^
Emission Factor
kg/MT
22.5
7.5
0.85
0.20
0.15
0.035
0.02
0.02
0.01
Ib/ton
45.0
15.0
1.7
0.4
0.3
0.07
(.007-. 138)
0.04 .
0.04
(.025-. 053)
0.02
(0. 07-. 036)
a References 1, 2, 5-10, and 14-16.
Factors experssed in terms of emissions per unit weight of
asphalt concrete produced.
c Almost all plants have at least a precleaner following the
rotary dryer.
These factors differ from those given in Table 8.1-1 be-
cause they are for uncontrolled emissions and are from an
earlier survey. Refer to Reference 16.
e The average emission from a properly designed,, installed,
operated, and maintained scrubber, based on a study to
develop new source performance standards. Refer to
Reference 15.
References 14 and 15.
^ Emissions from a properly designed, installed, operated, and
maintained baghouse, based on a study to develop new source
performance standards. Refer to References 14 and 15.
128
-------
Table 8.1-4. POTENTIAL UNCONTROLLED FUGITIVE PARTICULATE
EMISSION FACTORS FOR CONVENTIONAL ASPHALTIC CONCRETE PLANTS
EMISSION FACTOR RATING: E
Type of Operation
Unloading coarse and fine
aggregate to storage binsb
Cold and dried (and hot)
aggregate elevator'5
Screening hot aggregate0
Participates8
kg/MT
0.05
0.10
0.013
Ib/ton
0.10
0.20
0.026
* Factors expressed as units per unit might of aggregate.
Reference 18. assumed equal to similar sources.
c Reference 19, assumed equal to similar crushed granfte
processes.
Table 8.1-5 PARTICULATE EMISSION FACTORS FOR
DRYER DRUM HOT-MIX ASPHALT PLANTS11
EMISSION FACTOR RATING: B
Type of Control
Uncontrolled
Cyclone or nultlcyclone
Low energy wet scrubber
Venturi scrubber
Emission Factor9
kg/MT
2.45
0.34
0.04
0.02
Ib/ton
4.9
0.67
0.07
0.04
* Factors expressed In terms of emissions per unit weight
of asphalt concrete produced. These factors vary from
those for conventional asphaltic concrete plants
because the aggregate contacts and is coated with
asphalt early in the dryer drum process.
Either stack sprays where water droplets are injected
into the exit stack or a dynamic scrubber that
Incorporates a wet fan.
129
-------
References for Section 8.1
1. Asphaltic Concrete Plants Atmospheric Emissions Study.
Valentine, Fisher, and Tomlinson, Consulting Engineers,
Seattle, Washington. Prepared for U.S. Environmental
Protection Agency, Research Triangle Park, N.C., under
Contract No. 68-02-0076, November 1971.
2. Guide for Air Pollution Control of Hot-Mix Asphalt Plants.
National Asphalt Pavement Association, Riverdale, Md.,
Information Series 17.
3. Daniel son, O.A. Control of Asphaltic Concrete Batching
Plants in Los Angeles County. J. Air Pol. Control Assoc.
10(2):29-33. 1960.
4. Friedrich, H.E. Air Pollution Control Practices and Criteria
for Hot-Mix Asphalt Paving Batch Plants. APCA Paper Number
69.160. American Precision Industries, Inc., Buffalo, N.Y.
Presented at the 62nd Annual Meeting of the Air Pollution
Control Association.
5. Air Pollution Engineering Manual. PHS Publication No. 999-
AP-40. Los Angeles County Air Pollution Control District.
Prepared for U.S. DHEW, Public Health .Service, 1973.
6. Allen, 6.L., F.H. Vicks, and I.e. McCabe. Control of
Metallurgical and Mineral Dust and Fumes in Los Angeles
County, California. U.S. Department of Interior, Bureau
of Mines, Washington, D.C., Information Circular 7627,
April. 1952.
7. Kenllne P.A. Unpublished report on control of air pollu-
tants from chemical process industries. Robert A. Taft
Engineering Center, Cincinnati, Ohio, May 1959.
8. Sal lee, G. Private communication on particulate pollutant
study between Midwest Research Institute and National Air
Pollution Control Administration, Durham, N.C. Prepared
under Contract No. 22-69-104, June 1970.
9. Daniel son, O.A. Unpublished test data from asphalt batching
plants, Los Angeles County Air Pollution Control District.
Presented at Air Pollution Control Institute, University of
Southern California, Los Angeles, November 1966.
130
-------
10. Fogel, M.E. et al. Comprehensive Economic Study of Air
Pollution Control Costs for Selected Industries and Selected
Regions. Prepared for U.S. Environmental Protection Agency,
Research Triangle Park, N.C., under Final Report No. R-OU-455,
February 1970.
11. Preliminary Evaluation of Air Pollution Aspects of the Drum-
Mix Process. JACA Corp., Fort Washington, Pennsylvania.
Prepared for U.S. Environmental Protection Agency, Research
Triangle Park, N.C., under Final Report No. EPA-340/1-77-004,
March 1976.
12. Beaty, R.W. and B.M. Bunnell. The Manufacture of Asphalt
Concrete Mixtures 1n the Dryer-Drum. JACA Corp., Fort
Washington, Pennsylvania. Presented at the CTAA annual
meeting, November 19-21, 1973.
13. Kinsey, O.S. An Evaluation of Control Systems and Mass
Emission Rates From Dryer-Drum Hpt.Ashpalt Plants. Colorado
Air Pollution Control Division, December 1976.
14. Background Information for Proposed New Source Performance
Standards. Publication Nos. APTD 1363A and 1352B. USEPA,
Office of Air Quality Planning and Standards, Research
Triangle Park, N.C., 1973.
15. Background Information for New Source Performance Standards.
EPA 450/2-74-003 (APTD 1352C). USEPA, Office of Air Quality
Planning and Standards, Research Triangle Park, N.C., February
1974.
16. Kahn, Z.S., and T.W. Hughes. Source Assessment: Asphalt
Paving Hot-Mix. IERL-C1-260. Monsanto Research Corporation,
Dayton, Ohio. Prepared for U.S. Environmental Protection
Agency, Research Triangle Park, N.C., under Contract No. 68-
02-1874, Program Element No. !AB604,July 1977.
17. Puzinauskas, V.P. and L.W. Corbett. Report on Emissions from
Asphalt Hot Mixes. RR-75-1A. The Asphalt Institute, Asphalt
Institute Building, College Park, Md., May 1975.
131
-------
18. Evaluation of Fugitive Dust from Mining, Task 1 Report. PEDCo
Environmental Specialists, Inc. Cincinnati, Ohio. Prepared for
Industrial Environmental Research Laboratory/REHD, U.S. Environ-
mental Protection Agency, Cincinnati, Ohio. Contract No. 68-02-
1321, Task No. 36. June 1976.
19. Chalekode, P.K., and O.A. Peters. Assessment of Open Sources.
Monsanto Research Corporation, Dayton, Ohio. Presented at
Third National Conference on Energy and the Environment, College
Corner, Ohio. October 1, 1975. 9 p.
20. Pacific Environmental Services, Inc., Santa Monica, California.
Illustration of Dryer Drum Hot-Mix Asphalt Plant, 1978.
21. Forsten, Herman H. Applications of Fabric Filters to Asphalt
Plants. APCA 78-56-2. E.I. Du Pont de Nemours and Company,
Chestnut Run, Wilmington, DE 19893.
132
-------
BACKGROUND DOCUMENT
SECTION 8.1 ASPHALTIC CONCRETE PLANTS
1.0 INTRODUCTION . • .
• In December 1977, AP-42 Section 8.1, "Asphaltic Concrete
Plants," was revised, but still included particulate emission
factors only. The explanatory information submitted at that time
corresponds to Sections 4.1 and 4.3 of this background document.
In August 1978, the additions of organic emission factors and
factor ratings were assigned under a separate contract and task
order. The background information for these factors and factor
ratings appears in Sections 2.0, 3.0 and 4.2 of this document.
The stack emission factors presented in Table 8.1-1 are based
either on the mean of a 400 plant survey or on sampling data from a
representative hot-mix asphalt plant with the characteristics
listed in Table 8.1-2. The data, calculations, and emission
factors appear in Monsanto Research Corporation's (MRC) Source
Assessment: Asphalt Hot-Mix (Reference 16).
The particulate emission factors for conventional hot-mix
asphalt plants given in Table 8.1-3 are derived from several
sources by the methods described in Section 4.1 of this document.
The potential uncontrolled fugitive particulate emission
factors for conventional plants given in Table 8.1-4 were deter-
mined from References 18 and 19, as Section 4.2 describes.
133
-------
Section 4.3 presents the data from JACA Corporation's
Preliminary Evaluation of Air Pollution Aspects of the Drum-Mix
Process (Reference 11) that were used to calculate the particulate
emission factors for drum dryer asphaltic concrete plants shown in
Table 8.1-5.
2.0 EMISSION FACTORS FOR SELECTED MATERIALS ROM AN ASPHALTIC
CONCRETE PLANT STACK (Tables 8.1-1 and 8.1-2)
The emission factors presented in Table 8.1-1 appear in Tables
11 and 13 of Reference 16, and are based either on the mean of a
400 plant survey or on the plant characteristics listed in Table
8.1-2, as indicated in the notes to Table 8.1-1. Table 8.1-2
corresponds to Table 16 of Reference 16 (p. 53). The emission
factors were calculated as follows:
2.1 PARTICULATE
As indicated in footnote a of Table 11 (Reference 16, p. 50),
the particulate emission factor of 137 grams of particulate per
metric ton of asphaltic concrete produced was calculated from the
mean value of 6.09 grams per second (983.20 Ib/hr) in Figure 12,
"Asphaltic Hot-Mix Emission Rate" (Reference 16, p. 48), and the
mean average production rate of 160 metric tons of asphalt per hour
(176.4 ton/hr) from Question No. 62 of Table A-l, "Summary of
Asphalt Hot-Mix Industry Survey Data" (Reference 16, p. 96). Table
A-l is from MCR's Asphalt Industry Survey, conducted through the
National Asphalt Pavement Association (NAPA, November 23, 1975, 24
pages):
ED = 983.20 Ib/hr
176.4 ton/hr
= 5.57 Ib/ton « 137 g/MT
134
-------
2.2 SULFUR OXIDES
The following explanation of the calculation of the sulfur
oxides emission factor of 32 grams per metric ton is from Appendix
F of -Reference 16, "Calculation of Emission Factors Based on MRC
Sampling Data" (pp. 150-151):
• 40 percent of the asphalt industry uses No. 2 fuel oil
(according to Asphalt Industry Survey, MRC, op. cit.)
• Average sulfur contents of No. 2 fuel oil = 0.22 (Chemical
Engineers' Handbook. Perry and Chi 1 ton, eds., McGraw-Hill,
An average of 2 gallons of No. 2 fuel oil is consumed to
dry and heat aggregate for one ton of hot-mix ("Fuel
Conservation," Foster and Kloiber, NAPA, Riverdale)
Density of No. 2 fuel oil = 7.31 Ib/gal (Chemical Engineers'
Handbook, op. cit.) Ib of fuel oil consumed to dry 1 ton
hot-mix = 2 x 7.31 = 14.62 Ib
100 Ib of fuel oil contain S Ib sulfur
•M4.62 Ib of fuel oil contain 0.146(S) Ib sulfur
S * Op^SOo
(32) (32) 164)
• 32 Ib sulfur burn to form 64 Ib SOo
.-.0.146 Ib sulfur burns to form 64/32 x 0.146(S) Ib S02
* 0.292(5) Ib S02
Therefore the emission factor is 0.292(S) Ib S02/ton or
0.146(S) kg S02 Per metric ton of asphalt produced.
The calculated S02 was based on knowledge of type of fuel
used and the typical sulfur content, not on fuel analysis for
sulfur. The only characterization of the aggregate was the pH,
although the sources of aggregate (the site locations in Table I of
Reference 21) were widespread: Iowa, Pennsylvania, Georgia, and so
135
-------
on. Tables I and III from Reference 21 are reproduced here as
Tables 1 and 2 respectively.
2.3 NITROGEN OXIDES
Calculation of the nitrogen oxides emission factor of 18 grams
per metric ton is also given in Appendix F of Reference 16 (pp.
149-150):
• Concentration of nitrogen dioxide detected in stack gas is
<29 ppm by volume
• Gas flow rate through stack = 27,487 ft3/min
• Density of nitrogen dioxide = 0.1287 lb/ft3 (Chemical
Engineers' Handbook, op. cit.)
• Production rate during sampling = 176.4 ton/hr
• 106 ft3 of.stack gas contain 29 ft3 of.NOx
.'.27,487 ft3 of stack gas contain 29/106 x 27,487 ft3 NOX
* 0.80 ft3 NOX
• 1 ft3 NOM weighs 0.1287 Ib
.'.0.80 ft3 NOX weighs 0.1287 lb/ft3 x 0.80 ft3 = 0.10 Ib
• In 1 minute, 0.10 Ib NOX flows through the stack
/.In 60 minutes, 6.16 Ib NOX flow through the stack
• 176.4 tons of asphalt produced per hr emit 6.16 Ib NOX
.M ton of asphalt produced emits 0.035 Ib NOX
Therefore the emission factor is 0.035 Ib N0x/ton or 0.0176kg
NOX per metric ton of asphalt produced.
2.4 VOLATILE ORGANIC COMPOUNDS
The emission factor for volatile organic compounds (VOC) of 14
grams per metric ton of asphaltic concrete produced was calculated as
follows (Reference 16, Appendix F, p. 151):
136
-------
Table 1. ASPHALT INDUSTRY MONITORING RESULTS
Site
1
2
3
4
5
6
7
8
Fuel
#2 Oil &
Natural Gas
#2 Oil &
Natural Gas
#2 Oil
#2 Oil
#2 Oil
#2 Oil
#5 Oil
#5 Oil
Dryer
Capacity
T/hr
200
-
100
150
250
150
150
150
Exiting
Temperature
°F
160-210
150-230
325-395
195
260
305
150-185
290-360
Gas
Composition (percent)
C02
1.4
2.2
2.1
2.2
1.6
4.5
2.1
2.4
02
17. b
16.7
17.9
17.1
17.0
15.0
18.5
17.4
H20
-
-
-
8.1
4.2
11.7
10.5
2.8
S02 Concentration
(ppm)
Measured
5
5
20
10
0
10-25
10
65-230
Calculated
12
18
49
52.
100
38
171
214
Dust
' PH
7.0
7.0
11.2
9.3
1 -
9.7
10.8
6.9
Excess Air
Percent
475
388
548
416
404
212
672
459
CO
a From Reference 21, Table I.
-------
Table 2. MONITORING OF TURBULENT MASS MIXER ASPHALT PLANT3
Operating Conditions
Capacity
Fuel
Asphalt Temperature
Injection Rate
Baghouse - Type
- Temperature
- Bags
- Pressure Drop
Gas Streamb
S02
C02
°2
Dust - pH
Hydrocarbons
Dust
Condensate
Excess Air
100 T/D
#2 Oil
315°F
5%
Plenum Pulse
305°F
Nomex
4-5 in
18 ppm (calc
4-4.5%
14-15%
12%
9.7
Trace
Trace
212%
38 ppm)
a From Reference 21,Table III
b Measurements made by Du Pont
138
-------
• Concentration of VOC in stack gas = 42.3 ppm
• Gas flow rate through the stack » 32,350*5,133 ft3/min
• Density of hydrocarbons emitted = 0.0448 lb/ft3 (Chemical
Engineers' Handbook, op. cit«)
• Production rate during sampling = 130 ton/hr
• 106 ft3 of.stack gas contain 42.3 ft3 yOC ,
.-.32,350 ft3 of stack gas contain 42.3/10° x 32,350 ftj HC
» 1.37 ft3 VOC
• 1 ft3 of VOC weighs 0.0448 Ib
.U.37 ft3 VOC weighs 0.0448 lb/ft3 x 1.37 ft3 « 0.061 Ib
• In 1 minute, 0.061 Ib VOC flows through the stack
/.In 60 minutes, 3.68 Ib VOC flow through the stack
• 130 tons of asphalt produced per hr emit 3.68 Ib VOC
.M ton of asphalt produced emits 0.028 Ib VOC
The emission factor is thus 0.028 Ib VOC/ton or 0.014 kg VOC
per metric ton asphalt produced.
2.5 CARBON MONOXIDE
The carbon monoxide emission factor given in Table 8.1-1 is 19
grams per metric ton. Reference 16, Appendix F, p. 149 gives the
following account of how this factor was calculated:
• Concentration of carbon monoxide in stack gas = 32.2 ppm
17.6 percent by volume
• Gas flow rate through stack « 32,350 ft3/min
• Density of carbon monoxide s 0.0781 lb/ft3 (Chemical
. Engineers' Handbook, op. cit.)
• Production rate during sampling » 130 ton/hr
• 106 ft3 of stack gas contain 32.2 ft3 of CO
.'.32,350 ft3 of stack gas contain 32.2/106 x 32,350 ft3 CO
= 1.04 ft3 CO
139
-------
• 1 ft3 CO,weighs 0.0781 Ib
.U.04 ft3 CO weighs 0.0781 lb/ft3 x 1.04 ft3 = 0.081 Ib
• In 1 minute, the CO flowing through the stack is 0.081 Ib
.'.In 60 minutes, the CO flowing through the stack is 4.87 Ib
• 130 tons of asphalt produced per hr emit 4.807 Ib CO
.'.1 ton of asphalt produced per hr emits 0.0375 Ib CO
The emission factor is thus 0.0375 Ib CO/ton or 0.0188 kg CO
per metric ton asphalt produced.
2.6 POLYCYCLIC ORGANIC MATERIAL
The emission factor for polycyclic organic material (POM)
shown in Table 8.1-1 is 0.013 grams per metric ton. Footnote c of
Table 11 of Reference 16 indicates that the calculation of this
emission factor is shown-in Table B-20 of Appendix B, "Results from
Representative Plant Sampled by MRC" (p. 112). Table B-20 is
reproduced here as Table 3:
Table 3. OUTLET POM EMISSION RATE
Run
No.
!*>
2
3
Emission rate,
kg/hr (Ib/hr)
0.0027 (0.006)
0.0014 (0.003)
0.0018 (0.004)
Production rate,
metric tons/hr
(tons/hr)
57.0 (62.8)
119.3 (131.5)
133.1 (146.7) .
Emission factor0
kg/metric ton (Ib/ton)
4.8 x ID'5 (9.6 x 1CT5)
1.2 x 10-5 (2.3 x lO'5)
1.4 x ID'5 (2.8 x 10'5)
a Mean emission factor = 1.25 x 10~f ± 38.2% kg/metric ton
(2.5 x 1.0-'5 ± 38.2% Ib/ton)
b Not included during averaging because of difficulties during
sampling.
140
-------
2.7 ALDEHYDES
2.7.1 GENERAL EMISSION FACTOR
The overall emission factor for aldehydes is 10 grams per
metric ton. Appendix F of Reference 16 (p. 152) gives the follow-
ing explanation of how this factor was calculated:
• Concentration of aldedehydes in stack gas = 14.8 ppm 33
percent
• Gas flow rate through stack = 32,350 ft3/min
• Density of acetaldehyde = 0.1235 lb/ft3 (Chemical
Engineers' Handbook, op. cit.)
• Production rate during sampling = 130 tons/hr
• 106 ft3 of,stack gas contain 14.8 ft3 aldehydes
.'.32,350 ft3 stack gas contain 14.8/106 x 32,350 ft3
aldehydes = 0.48 ft3 aldehydes
• 1 ft3 aldehyde weighs 0.1235 Ib
/.0.48 ft3 aldehyde weighs 0.059 Ib
• In 1 minute, 0.059 Ib aldehyde flow through the stack
/.In 60 minutes, 3.55 Ib aldehyde flow through the stack
• 176.4 tons of asphalt produced per hr emit 3.55 Ib aldehyde
/.I ton of asphalt produced emits 0.020 Ib aldehyde
Thus the emission factor is 0.020 Ib aldehydes/ton or 0.0101 kg
aldehydes/metric ton asphalt produced.
2.7.2 SPECIES EMISSION FACTORS
The emission factors for the four species of aldehydes given
in Table 8.1-1 are shown in Table 13 of Reference 16 (p. 51).
Footnote a of Table 13 indicates that these factors were calculated
from the production rate and emission rate shown in Table B-33 of
141
-------
Reference 16 (p. 113), which is reproduced below as Table 4.
(However, the terms "isobutanol" and "isopentanol" are neither
systematic nor common names, but hybrids not customarily used.)
Table 4. ALDEHYDES DETECTED IN SAMPLES COLLECTED AT OUTLET
Run
No.
1
2
3
1
2
3
1
2
3
1
2
3
Aldehyde
Formal dehyde
Formaldehyde
Formaldehyde
Isobutanal
Isobutanal
Isobutanal
Butanal
But anal
Butanal
Isopentanal
Isopentanal
Isopentanal
Concentration
g/ml
0.2
0.11
0.2
3
<0.2
30
3
2
7
20
«0.1
20
g/m3
215
187
247
3,213
<339
37,076
3,213
<3,390
8,828
21,540
«177
24,718
Emission Rate
g/hr
12.3
10.7
14.2
184
19.4
2,130a
184.2 •
194.3
506 a
1,230
10. la
1,420
Mean mg/s
3.4 26%
28 350%
53 12%
370 30%
a Not averaged
No further explanation of the method used to calculate these
emission factors appears in Appendix B or elsewhere in Reference
16. The values for replicate analyses given in Table 4 (Table B-23
of Reference 16). show poor repeatability and, as footnote a indi-.
cates, some were not averaged with the rest (although no reason for
these omissions is given).
142
-------
3.0 EMISSION FACTOR RATINGS
The factor ratings assigned to each emission factor listed in
Table 8.1-1 were arrived at by applying RES' best engineering
judgment to the source test information upon which each calculation
was based.
The particulate emission factor was rated "B" because it is
based on data on emissions and operating parameters for a large
sample of the asphalt hot-mix industry, covering a broad spectrum
of plant types and conditions.
The sulfur oxides emission factor is also based on emissions
and operating data for a large sample of the asphaltic concrete
industry, but was assigned a rating of "C" because of the variable
reduction in the S02» dependent on the nature of the aggregate
(refer to Section 2.2 and Reference 21).
The nitrogen oxides, VOC, carbon monoxide, POM, and aldehydes
emission factors were rated "D" because they are based on limited
source test data from the single asphaltic concrete plant
described 1n Table 8.1-2 (Table 16, p. 53, of Reference 16).
4.0 PARTICULATE EMISSION FACTORS
4.1 CONVENTIONAL ASPHALTIC CONCRETE PLANTS (Table 8.1-3)
Source test information from EPA's Background Information for
New Source Performance Standards (APTD-1352B and APTD-1352C,
References 14 and 15) was used to update existing AP-42 par-
ticulate emission factors for conventional asphaltic concrete
143
-------
plants shown in Table 8.1-3. Sections 4.1.1 to 4.1.3 give back-
ground information for the factors which were modified or added.
4.1.1 PARTICULATE EMISSION FACTORS FOR PLANTS EQUIPPED WITH
"MULTIPLE CENTRIFUGAL SCRUBBER"
Source test results, which are presented in Table 5, were
taken from Appendix A of Reference 15. In this report, the term
"multiple centrifugal scrubber" applies to various combinations of
cyclones, low energy wet scrubbers, and wet fans in series. In the
study, a total, of 17 source tests were conducted on nine plants
i
equipped with properly designed, operated, and maintained
scrubbers. Data for plants equipped with multiple centrifugal
scrubbers were collected using
tion District's test procedure.
the San Bernardino County Air Pollu-
The average emission factor from
these plants was derived by simple averaging of the source test
results, as follows:
0.0743 w 0.07 Ib/ton
17
(2)
where
average emission factor
emission factor derived from individual source test
source test number (from 1 to 17)
144
-------
The range shown 1n parentheses in Table 8.1-3 is taken from
Table 5.
TableB. PARTICIPATE EMISSION FACTORS FROM
CONVENTIONAL ASPHALTIC CONCRETE PLANTS EQUIPPED WITH
MULTIPLE CENTRIFUGAL SCRUBBER AS CONTROL DEVICE
Code of
Facility
X
Y
Z
AA
BB
CC
DO
EE
FF
Emission Factor
(Ib/ton)
Test 1
0.081
0.007
0.116
0.014
0.0293
0.059
0.057
0.103
0.124
Test 2
0.0975
0.004
0.0408
0.056
0.123
0.120
Test 3
0.094
0.138
Collector
TC*
TC
TC
TC
RFC
TC
PF
PF
PF
Reference
Page3
67
68
69
70
71
72
73
74
75
• page nunber for report APTD-1352C, Vol. 3 (Reference 15)
b Total catch
c Probe and filter catch
4.1.2 PARTICULATE EMISSION FACTORS FOR PLANTS EQUIPPED WITH
VENTURI SCRUBBER
Source test results, which are presented in Table 6, were
taken from the New Source Performance Standards publications
APTD-1352B and APTD-1352C (References 14 and 15). In these
studies, a total of four source tests were conducted on four plants
equipped with venturi scrubbers. The data for plants equipped with
venturi scrubbers were gathered using test procedures in general
conformance with EPA Method 5. The emission factor from a plant
equipped with a venturi scrubber was estimated by averaging the
results from all of the above source tests, as follows:
145
-------
4
E
1=1
= 0.039^ 0.04 Ib/ton
(3)
where
average emission factor
emission factor derived from individual source test
source test number from 1 to 4)
The range shown in Table 8.1-3 is taken from the results given
in Table 6.
Table 6. PARTICIPATE EMISSION FACTORS ROM CONVENTIONAL
ASPHALTIC CONCRETE PLANTS EQUIPPED WITH VENTURI SCRUBBER
AS CONTROL DEVICE
Code of Facility
C
H1
U2
Emission Factor
(Ib/ton)
0.037
0.044
0.025
0.053
Collector
PFa
PFh
TCb
PF
Reference Page
IOC
15C
16S
64d
a probe and filter catch
b Total catch
c page number for report APTD-1352B, Vol. 2 (Reference 14)
d Page number for report APTD-1352C, Vol. 3 (Reference 1,5)
4.1.3 PARTICULATE EMISSION FACTOR FOR PLANTS EQUIPPED WITH
BAGHOUSES
Source test results, which are presented in Table 7, were
taken from the New Source Performance Standards publications
146
-------
APTD-1352B and APTD-1352C (References 14 and 15). In these
studies, a total of six source tests were conducted on six plants
equipped with baghouses. The data for plants equipped with
baghoases were collected using test procedure in general
conformance with EPA Method 5. The emission factor from a plant
equipped with a baghouse was derived by averaging the results from
the above source tests, as follows:
Table 7. PARTICIPATE EMISSION FACTORS FROM CONVENTIONAL
ASPHALTIC CONCRETE PLANTS EQUIPPED WITH BAGHOUSE
AS CONTROL DEVICE
Code of Facility
AI
A?
B
D
L
0
Emission Factor
(Ib/ton)
0.016
0.0243
0.007
0.016
0.0087
0.036
Collector
PFa
PF
PF
PF
PF
PF
Reference Page
7K
8b
gb
llb
20b
58C
* probe and filter catch
b Page number for report APTD-1352B, Vol. 2 (Reference 14)
c Page number for report APTD-1352C, Vol. 3 (Reference 15)
i=l
0.018 « 0.02 Ib/ton
(4)
147
-------
where
average emission factor
emission factor derived from individual source test
i = source test number (from 1 to 6)
The range given in Table 8.1-3 is taken from Table 7.
4.1.4 EMISSION FACTOR RATING
The emission factors presented in Table 8.1-3 were rated "B"
because of the quality of the emission test data, the number of
separate plants tested, and the adequacy of process and engineering
information.
4.2 FUGITIVE PARTICIPATE FROM CONVENTIONAL ASPHALTIC CONCRETE
PLANTS (Table 8.1-4)
References 18 and 19 were the primary sources of the potential
fugitive particulate emision factors given in Table 8.1-4. The
following EPA document was written using these two references and
may be substituted as the source of this information:
Zoller, J., T. Bertke, and T. Janszen. Assessment of Fugi-
tive Particulate Emission Factors for Industrial Processes.
EPA-450/3-78-107. PEDCo Environmental, Cincinnati, Ohio.
Prepared for U.S. Environmental Protection Agency, OAQPS,
Research Triangle Park, N.C., September 1978.
An uncontrolled fugitive particulate factor of negligible to
.05 kg/Mg (0.1 Ib/ton) of aggregate handled has been determined
(Monsanto Research and Environmental Research Technology) for
losses from unloading coarse/fine aggregate to storage bins, and a
factor of negligible to 0.1 kg/Mg (0.2 Ib/ton) has been determined
for elevator conveying of cold and dried (hot) aggregate. These
148
-------
determinations are based on similar processing at granite quarries
and coal mining operations (Reference 18). An uncontrolled fugi-
tive emission rate ranging from negligible to 0.013 kg/Mg (0.026
Ib/ton) of aggregate handled has been determined for the screening
of hot aggregate, based on Monsanto Research Corporation test
values for crushed granite processing (Reference 19). Extensive
particulate fugitive emission test sampling data from asphaltic
concrete batching plants are needed to obtain representative data
for documentation in AP-42. For this reason, these factors have
been assigned a reliability rating of E.
4.3 DRYER DRUM HOT ASPHALT PLANTS (Table 8.1-5)
4.3.1 EMISSION FACTORS
Source test information from Preliminary Evaluation of Air
Pollution Aspects of the Drum-Mix Process (Reference 11) was the
only data used to derive the emission factors presented in Table
8.1-5. In this study, particulate emission concentrations were
obtained from field test data gathered at different plants with
different control devices. The results are summarized in Table 8,
which presents the average particulate concentrations in the stacks
of dryer drum plants with different control devices.
Particulate concentrations in the stacks were obtained from 70
different plant tests. Thirty-one additional test runs that were
from plants where only the aggregate data were made available by
the manufacturer were not included. In deriving the emission fac-
tors, seven test results from the 70 reported were excluded for the
following reasons: two tests were run outside the isokinetic range;
149
-------
TableS. PARTICULATE EMISSION CONCENTRATIONS FROM A
PARALLEL-FLOW DRYER DRUM HOT-MIX ASPHALT PLANT
Number of
Plants Tested
9b
7
24
18
Control
Device
Uncontrolled
Dry mechanical
collector
Wet scrubber
Venturi
scrubber
Particulate
Concentration
(grain/dscf)
6.19
0.853
0.094
0.0557
Reference
Page3
32
34
35
37
a Page number in EPA-340/1-77-004 (Reference 1)
b An additional 5 plants had much lower average reading and were
not included
three tests were performed on nonrepresentative drum dryer plants;
and two tests were duplicates.
Of the 63 tests that were acceptable for inclusion in the
analysis, 14 were on uncontrolled plants; seven were on plants with
dry mechanical controls, such as cyclones and multicyclones; 24
were from plants with scrubbers of the spray impingement or wet fan
type; and 18 were on plants with venturi scrubbers of varying pres-
sure drops.
The 63 tests gave a total of 158 independent runs for analy-
sis, of which 108 reported only "front-half" results, seven
reported only the "total" particulate matter, and 43 reported
both front-half as well as total particulate concentration.
The average emission concentrations obtained from the above
source tests for each degree of control are reported in Table 8.
Emission concentrations multiplied by the average stack gas flow
r -
150
-------
per ton of product yield the emission factors shown in Table 9.
The typical range of stack flow gas from a dryer drum plant is
between 4,000 and 7,000 dscf/ton, depending on burning conditions.
The average stack gas flow used was 5,500 dscf/ton.
Table 9. PARTICIPATE EMISSION FACTORS FROM A PARALLEL-FLOW
DRYER DRUM HOT-MIX ASPHALT PLANT
Degree of Control
Uncontrolled
Cyclone or multi-
eye! one
Wet scrubber5
Venturi scrubber
Emission Factor3
Ib/ton
4.9
0.67
0.07
0.04
kg/MT
2.45
0.34
0.04
0.02
a Emission factors expressed as units per unit
weight of asphalt concrete produced
D Wet scrubbers Include stack spray, wet fan, and
dynamic scrubber
151
-------
Participate emission factors were obtained by using the following
relation:
Particulate emission Participate emission
factor B concentration x °»DUU
Uncontrolled emission * 6.19 grain'x 1 x Ib x 5.500 dscf
emission dscf 7,000 grain 1 ton
« 4.86 % 4.9 Ib/ton
Cyclone or multi- » 0.853 x 1 x 5,500 = 0.67 Ib/ton
cyclone 7,000
Wet scrubber = 0.094 x 1 x 5,500 = 0.074 ^0^07 Ib/ton
77550"
Venturi scrubber -0.0557 x 1 x 5,500 = 0.044 » 0.04 Ib/ton
4.3.2 EMISSION FACTOR RATING
The rating for the particulate emission factors presented in
Table 8.1-5 was designated "B" due to the amount and reliability of
the available source test data.
152
------- |