United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-453/R-93-054
November 1993
Air
National Emission
Standards for Hazardous
Air Pollutants:
Halogenated Solvent
Cleaning- Background
Information Document
ENVIRONMENTAL
v PROTECTION
\
DALLAS, TEXAS
-------
EPA-453/R-93-054
National Emission Standards
for Hazardous Air Pollutants
Halogenated Solvent Cleaning
- Background Information
Document
Emission Standards Division
U.S. Environmental Protection Agency
Office of Air and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
November 1993
-------
TABLE OF CONTENTS
Chapter Page
1.0 OVERVIEW 1-1
1.2 PROJECT HISTORY 1-2
1.2.1 Background 1-2
1.2.2 Data Gathering 1-3
1.2.3 Emissions and Control Data 1-4
1.3 REFERENCES 1-5
2.0 ORGANIC SOLVENT CLEANER CHARACTERISTICS AND EMISSIONS 2-1
2.1 GENERAL 2-1
2.2 ORGANIC SOLVENT CLEANING EQUIPMENT AND PROCESSES 2-2
2.2.1 Batch Vapor Cleaners 2-3
2.2.1.1 OTVC's 2-3
2.2.1.1.1 The cleaning cycle . . 2-4
2.2.1.1.2 Design parameters . . . 2-6
2.2.1.1.3 Design variations . . . 2-8
2.2.1.1.4 Operational variations 2-11
2.2.1.2 Other Batch Cleaners 2-13
2.2.2 In-line Cleaners 2-16
2.2..3 Batch Cold Cleaners 2-22
2.3 EMISSION MECHANISMS AND TYPES 2-23
2.3.1 Air/Solvent Vapor Interface Losses
During Idling (Idling Losses) 2-25
2.3.1.1 OTVC'S 2-25
2.3.1.2 In-line Cleaners and Batch
Non-OTVC's 2-27
2.3.1.3 Cold Cleaners 2-30
2.3.2 Workload Losses 2-30
2.3.2.1 OTVC'S 2-30
2.3.2.2 In-line Cleaners and Batch '
Non-OTVC's 2-34
2.3.2.3 Cold Cleaners 2-36
ii
-------
TABLE OF CONTENTS (CONTINUED)
Chapter Page
2.3.3 Other Losses 2-36
2.3.3.1 Downtime 2-37
2.3.3.2 Leaks 2-37
2.3.3.3 Filling/Draining 2-38
2.3.3.4 Wastevater 2-38
2.3.3.5 Start-up/Shutdown 2-38
2.3.3.6 Distillation Losses/Sludge
Disposal 2-39
2.3.3.7 Solvent Decomposition 2-39
2.4 NATIONAL BASELINE EMISSIONS 2-39
2.5 REFERENCES 2-41
3.0 EMISSION CONTROL TECHNIQUES 3-1
3.1 INTRODUCTION 3-1
3.2 BATCH VAPOR CLEANERS 3-4
3.2.1 Controls for Interface Emissions 3-5
3.2.1.1 Covers 3-5
3.2.1.2 Freeboard Refrigeration Devices . 3-15
3.2.1.3 Refrigerated Primary Condenser . 3-21
3.2.1.4 Increased Freeboard Ratio . . . . 3-25
3.2.1.5 Reduction in Room Draft/Lip
Exhaust Velocities 3-30
3.2.1.6 Carbon Adsorption ... 3-33
3.2.1.7 Enclosed Design 3-37
3.2.1.8 Vacuum Chamber Cleaners 3-38
3.2.2 Controls for Working Emissions 3-40
3.2.2.1 Mechanically-Assisted Parts
Handling/Parts Movement Speed . . 3-40
3.2.2.2 Hot Vapor Recycle/Superheated
Vapor 3-44
3.2.3 Proper Operating and Maintenance Practices 3-45
3.2.3.1 Reducing Drafts 3-45
3.2.3.2 Spray Techniques 3-45
3.2.3.3 start-Up/Shutdown Procedures, . . 3-45
3.2.3.4 Downtime Losses 3-47
3.2.3.5 Workload Introduction/Removal . . 3-47
iii
-------
TABLE OF CONTENTS (CONTINUED)
Chapter Page
3.2.3.6 Parts Drainage 3-48
3.2.3.7 Leak Detection/Repair 3-48
3.2.3.8 Solvent Transfer 3-49
3.2.3.9 Safety Switches 3-49
3.3 IN-LINE CLEANERS 3-51
3.3.1 Controls for interface Emissions 3-53
3.3.1.1 Minimizing Entrance/Exit Openings 3-53
3.3.1.2 Carbon Adsorption ........ 3-53
3.3.1.3 Freeboard Refrigeration Devices . 3-55
3.3.2 Controls for Workload Emissions 3-55
3.3.2.1 Drying Tunnels 3-55
3.3.2.2 Rotating Baskets 3-56
3.3.2.3 Hot Vapor Recycle/Superheated
Vapor 3-56
3.3.3 Proper Operating and Maintenance Practices 3-59
3.3.3.1 Conveyor Speed 3-59
3.3.3.2 Spray Techniques 3-59
3.3.3.3 Start-up/Shutdown Procedures . . 3-59
3.3.3.4 Carbon Adsorber Procedures . . . 3-60
3.3.3.5 Parts Drainage 3-60
3.3.3.6 Leak Detection/Repair 3-60
3.3.3.7 Solvent Transfer 3-61
3.3.3.8 Safety Switches 3-61
3.4 COLD CLEANERS 3-61
3.5 ALTERNATIVE CLEANING TECHNOLOGIES 3-62
3.5.1 Alternative Cleaning Solvents 3-62
3.5.1.1 Hydrochlorofluorocarbons .... 3-63
3.5.1.2 Aqueous 3-65
3.5.1.3 Semi-Aqueous 3-66
3.5.1.4 Organic 3-68
3.5.2 Alternative Cleaning Technologies . ,. . . 3-68
3.5.2.1 Ice Particles 3-68
3.5.2.2 Plasma 3-69
iv
-------
TABLE OF CONTENTS (CONTINUED)
Chapter Page
3.5.2.3 Pressurized Gases 3-69
3.5.2.4 Supercritical Fluids 3-69
3.5.2.5 Ultraviolet/Ozone 3-69
3.5.2.6 Mechanical 3-69
3.5.2.7 Thermal Vacuum Deoiling 3-70
3.5.3 No-Clean Technologies 3-70
3.5.3.1 Low-Solids Flux 3-70
3.5.3.2 Controlled Atmosphere Soldering . 3-70
3.5.3.3 Process Modifications ...... 3-71
3.6 REFERENCES 3-71
4.0 DESCRIPTION OF MODEL CLEANERS AND THE SOLVENT
CLEANER POPULATION 4-1
4.1 MODEL CLEANERS 4-1
4.1.1 Model Batch Cleaners 4-7
4.1.2 Model In-line Cleaners 4-9
4.2 SOLVENT CLEANER CONTROL COMBINATIONS 4-9
4.3 NATIONAL ESTIMATES OF SOLVENT CONSUMPTION AND
NUMBER OF CLEANERS 4-12
4.3.1 National Solvent Use Estimates 4-12
4.3.2 Estimate of the National Number of
Solvent Cleaners 4-22
4.4 REFERENCES 4-25
5.0 REGULATORY APPROACH 5-1
5.1 MACT FLOOR 5-1
5.2 DEVELOPMENT OF ADDITIONAL REGULATORY
ALTERNATIVES 5-2
5.3 NATIONAL IMPACTS 5-15
5.3.1 National Costs 5-15
5.3.2 Air, Water and Solid Waste Impacts .... 5-17
5.4 REFERENCES '. . . 5-17
APPENDIX A: Derivation of the Control Efficiency Formula
-------
LIST OF FIGURES
Tables Page
2-1 Open top vapor cleaner 2-5
2-2 Water separator with cooling coil 2-7
2-3 Two compartment vapor cleaner 2-9
2-4 Two compartment vapor cleaner with offset boiler
chamber 2-10
2-5 Open top vapor cleaner with lip exhaust 2-12
2-6 Cross-rod cleaner 2-14
2-7 Vibra-cleaner 2-15
2-8 Ferris wheel cleaner 2-17
2-9 Monorail in-line cleaner 2-19
2-10 Mesh belt in-line cleaner 2-20
2-11 Schematic diagram of an in-line photoresist
stripping machine . 2-21
2-12 Carburetor cleaner 2-24
2-13 Batch cleaner idling emission sources 2-26
2-14 In-line cleaner emission sources 2-29
2-15 Batch cleaner workload-related emission sources . . 2-31
3-1 Typical open-top vapor cleaner covers 3-12
3-2 Reduction in emission rate from the addition of a
cover 3-13
3-3 Open-top vapor cleaner with freeboard refrigeration
device- 3-16
3-4 Freeboard refrigeration device tests-working
conditions 3-18
/
3-5 Freeboard refrigeration device tests-idling
conditions 3-19
3-6 Effect of primary condenser temperature on
uncontrolled idling and working conditions 3-23
vi
-------
LIST OF FIGURES (CONTINUED)
Tables Page
3-7 Open-top vapor cleaner with increased freeboard . . 3-26
3-8 Solvent loss rate versus freeboard height for
Genesolv* D under idling conditions 3-28
3-9 Effect of freeboard ratio-working conditions:
six open-top vapor cleaner tests 3-29
3-10 Effect of wind speed 3-32
3-11 Lip exhaust effects - idling conditions 3-34
3-12 Lip exhaust effects - working conditions 3-35
3-13 Enclosed batch vapor cleaners 3-39
3-14 Automated parts handling system 3-41
3-15 Baffled monorail in-line cleaner 3-54
3-16 Baffled monorail in-line cleaner with superheat
device 3-58
vii
-------
LIST OF TABLES
Tables Pace
2-1 HALOGENATED SOLVENT EVAPORATION RATES 2-28
2-2 NATIONAL BASELINE EMISSIONS (Mg/yr) . . 2-40
3-1 SUMMARY OF SOLVENT CLEANER CONTROL TECHNIQUES ... 3-2
3-2 SOLVENT VAPOR EMISSION CONTROL EFFICIENCIES FOR
SELECT CONTROL TECHNIQUES 3-3
3-3 SUMMARY OF TESTS ON IDLING OPEN-TOP VAPOR CLEANERS . 3-6
3-4 SUMMARY OF TESTS ON WORKING OPEN-TOP VAPOR CLEANERS 3-8
3-5 SOLVENT LOSS RATE VERSUS SPRAYING PRACTICES . . . . 3-46
3-6 SUMMARY OF AVAILABLE TESTS FOR IN-LINE CLEANERS . . 3-52
3-7 COMMON HYDROCHLOROFLUOROCARBON ALTERNATIVES . . . . 3-64
4-1 MODEL CLEANER PARAMETERS FOR SMALL BATCH VAPOR
CLEANERS OPERATING TWO HOURS PER WORK DAY 4-2
4-2 MODEL CLEANER PARAMETERS FOR MEDIUM BATCH VAPOR
CLEANERS OPERATING TWO HOURS PER WORK DAY 4-3
4-3 MODEL CLEANER PARAMETERS FOR LARGE BATCH VAPOR
CLEANERS OPERATING SIX HOURS PER WORK DAY ..... 4-4
4-4 MODEL CLEANER PARAMETERS FOR VERY LARGE BATCH VAPOR
CLEANERS OPERATING SIX HOURS PER WORK DAY 4-5
4-5 MODEL CLEANER PARAMETERS FOR IN-LINE CLEANERS ... 4-6
4-6 MODEL CLEANER SIZES 4-8
4-7 SOLVENT VAPOR EMISSION CONTROL EFFICIENCIES FOR
VARIOUS CONTROL TECHNIQUES 4-10
4-8 EXISTING BATCH VAPOR CLEANER CONTROL COMBINATIONS . 4-13
4-9 NEW BATCH VAPOR CLEANER CONTROL COMBINATIONS . . . . 4-15
4-10 CONTROL EFFICIENCIES FOR EXISTING IN-LINE SINGLE
AUD COMBINED CONTROLS 4-18
4-11 CONTROL EFFICIENCIES FOR NEW IN-LINE SINGLE AND
COMBINED CONTROLS 4-19
Viii
-------
LIST OF TABLES (CONTINUED)
Tables Page
4-12 CLEANING SOLVENT USAGE SUMMARY FOR 1991 (Mg/yr) . . 4-20
4-13 NUMBER OF HALOGENATEO SOLVENT CLEANERS 4-23
5-1 ANNUALIZED COSTS FOR EXISTING SMALL BATCH CLEANERS
(2 HOUR SCHEDULE) 5-4
5-2 ANNUALIZED COST FOR EXISTING MEDIUM BATCH CLEANERS
(2 HOUR SCHEDULE) 5-5
5-3 ANNUALIZED COST FOR EXISTING LARGE BATCH CLEANERS
(6 HOUR SCHEDULE) 5-6
5-4 ANNUALIZED COST FOR EXISTING VERY LARGE BATCH
CLEANERS (6 HOUR SCHEDULE) 5-7
5-5 SUMMARY OF EMISSION CONTROL COSTS FOR EXISTING
IN-LINE CLEANERS (8 HOUR SCHEDULE) 5-8
5-6 ANNUALIZED COSTS FOR NEW SMALL BATCH CLEANERS
(2 HOUR SCHEDULE) 5-9
5-7 ANNUALIZED COST FOR NEW MEDIUM BATCH CLEANERS
(2 HOUR SCHEDULE) 5-10
5-8 ANNUALIZED COST FOR NEW LARGE BATCH CLEANERS
(6 HOUR SCHEDULE) 5-11
5-9 ANNUALIZED COST FOR NEW VERY LARGE BATCH CLEANERS
(6 HOUR SCHEDULE) 5-12
5-10 SUMMARY OF EMISSION CONTROL COSTS FOR NEW IN-LINE
CLEANERS (8 HOUR SCHEDULE) 5-13
5-11 SOLVENT CLEANER SOURCE SUBCATEGORY REGULATORY
ALTERNATIVES 5-14
5-12 NATIONAL COSTS AND EMISSION REDUCTIONS FOR
SOLVENT CLEANERS 5-16
ix
-------
1.0 INTRODUCTION
1.1 OVERVIEW
Section 112 of the Clean Air Act (Act) requires that the
U. S. Environmental Protection Agency (EPA) establish emission
standards for all categories of sources of hazardous air
pollutants (HAP). These national emission standards for
hazardous air pollutants (NESHAP) must represent the maximum
achievable control technology (MACT) for all major sources and
either MACT or generally achievable control technology (GACT)
for all area sources. The Act defines a major source as:
...any stationary source or group of stationary
sources located within a contiguous area and under
common control that emits or has the potential to
emit, in the aggregate, 10 tons per year or more of
any hazardous air pollutant or 25 tons per year or
more of any combination of hazardous air pollutants.
An area source is any stationary source of HAP that is not a
major source. (A discussion of MACT and GACT is presented in
chapter 5.0, section 5.2 of this document.)
On July 16, 1992 (57 FR 31576), the initial list of
categories of sources that will be regulated under section 112
was published. "Halogenated Solvent Cleaners" was included on
the list both as a category of major sources and as a category
of area sources. The halogenated solvent cleaning/degreasing
NESHAP project will establish standards for both the area and
the major sources of halogenated solvent cleaners.
1-1
-------
The purpose of this document is to summarize the
background information gathered during the development of the
halogenated solvent cleaning NESHAP.
1.2 PROJECT HISTORY
1.2.1 Background
In 1977, a control technique guidelines (CTG) document
was established for the control of volatile organic compounds
(VOC) from solvent cleaners.1 In 1980, new source performance
standards (NSPS) were proposed for this industry; the NSPS has
not been promulgated.2 Also in 1980, five compounds were
proposed as designated pollutants under the authority of
section lll(d): trichloroethylene (TCE); perchloroethylene
(PCE); methylene chloride (MC); 1,1,1-trichloroethane (TCA);
and trichlorotrifluoroethane (CFC-113). in 1989, the
"Alternative Control Technology Document—Halogenated Solvent
Cleaners" (ACT) was published.3 Because the NSPS has not been
promulgated, a lawsuit was brought against the EPA. As a
result of the lawsuit, the EPA entered a consent decree to
propose the NESHAP within 3 years of the passage of the 1990
Amendments (1990 Amendments) to the Act (November 15, 1990)
and to promulgate standards within 1 year of proposal.
In the CTG, guidelines were presented for three'.
categories of solvent cleaners (also known as degreasers):
cold cleaners, batch cleaners (primarily open-top vapor
cleaners [OTVC]), and continuous, or in-line, (cold and vapor)
cleaners.
All of the halogenated solvents, that have been discussed
above, except CFC-113, are Act HAP and will be regulated by
the NESHAP. As a result of the Montreal Protocol and
subsequent 1990 Amendments, CFC-113 consumption will be phased
out by the year 2000. As a result of the Montreal Protocol
and the subsequent Presidential edict by President Bush, TCA
use is to be phased out by the end of 1995.
During recent data gathering efforts, no evidence of cold
batch cleaners using halogenated solvents, with the exception
of carburetor cleaners, has been identified. Carburetor
1-2
-------
cleaners occasionally use a MC blend and are veil controlled
at baseline. Chapters 2.0 and 3.0 discuss this further. As a
result of the limited use of halogenated solvents in batch
cold cleaners, the focus of this document is on batch vapor
cleaners (primarily OTVC) and on in-line cleaners (cold and
vapor).
1.2.2 Data Gathering
During the past several years, two .sets of questionnaires
were sent out under the authority of section 114 of the Act.
In 1987, nine questionnaires were sent to manufacturers and
vendors of halogenated solvent cleaners. The purpose of these
questionnaires was to obtain information (e.g., cleaner costs
and availability) on solvent cleaners and control techniques,
as well as to request any emissions data that were available
for these units.
In 1991, new questionnaires were developed to obtain
updated information.4 In addition, additional information was
needed to respond to the new method for establishing NESHAP
control levels. The 1990 Amendments require that minimum
control levels ("MACT floors11) be established for each
category regulated under section 112. These floors, discussed
in more detail in section 5.2 of this document, are based on
an analysis of the level of control existing in a category
prior to NESHAP regulation (i.e., at baseline). In order to
obtain information .to evaluate the level of control at
baseline, several approaches were considered, including:
• Use of existing State regulations;
• Use of national data bases;
• Use of questionnaires sent to users; and
• Use of questionnaires sent to manufacturers/vendors.
After evaluating each option, a questionnaire sent tp
manufacturers/vendors was determined to be the most effective
method of obtaining the maximum amount of information for the
maximum number of solvent cleaners. Use of existing data
1-3
-------
bases and State regulations were not expected to provide the
level of detail required for a MACT floor analysis.
While a questionnaire sent to the actual owners or
operators of the solvent cleaners would likely provide the
most accurate information, there are estimated to be
approximately 24,500 batch vapor and in-line solvent cleaners
in operation in the United States (see chapter 4.0, section
4.3.2). These cleaners do not belong to a single industry,
but rather are a process or maintenance step in most
industries. Identifying the users, or even an appropriate
cross section, was not feasible.
In addition to information obtained from these
questionnaires, several site visits have been made to solvent
cleaner manufacturers and to solvent cleaner users. Also, the
EPA has met with multiple trade associations and equipment
vendors over the past several years.
1.2.3 Emissions and Control Data
The available emissions and control information for the
solvent cleaning source category is summarized in chapter 3.0.
The EPA has made several attempts to supplement the existing
data base, including requests in both the 1987 and 1991
questionnaires; however, much of the data dates back to the
NSPS regulatory development activities. These data are still
relevant since the controls tested in the early 1980*s are
still in use today*. More recent test data have been obtained
for some relatively new technologies, such as super heated
vapor.
As part of the regulatory development activity, the EPA's
Emission Measurement Branch (EMB) has conducted some tests on
an OTVC.5 While the preliminary focus of the test was to
establish a reference method for determining the solvent
emissions from OTVC's during the idling mode, control
efficiency data were developed for different operating modes
(i.e., idling, working, and downtime) and control techniques.
1-4
-------
1.3 REFERENCES
1. u. S. Environmental Protection Agency. Control of
Volatile Organic Emissions from Solvent Metal Cleaning.
Publication No. EPA-450/2-77-022. Research Triangle
Park, N.C. November 1977. 201 pp.
2. U. S. Environmental Protection Agency. Federal Register.
Standards of Performance for New Stationary Sources
Organic Solvent Cleaning. Proposed Rules and Notice of .
Public Hearing. 45 FR IV-JJ-2 thru IV-JJ-20. June 11,
1980.
3. u. S. Environmental Protection Agency. Alternative
Control Technology Document — Halogenated Solvent
Cleaners. Publication No. EPA-450/3-89-030. Research
Triangle Park, N.C. August 1989. 216 pp.
4. Memorandum and attachments from Gerald, L. and
Falling, A., Radian Corporation, to Almod6var, P.,
U. S. Environmental Protection Agency. October 27, 1992.
Compilation of information obtained from Solvent Vapor
and Cold Cleaner Vendor Questionnaires.
5. Radian Corporation. Measurement of Emissions from Open
Top Vapor cleaners. Prepared for U. S. Environmental
Protection Agency. Research Triangle Park, N.C.
May 1992. 80 pp.
1-5
-------
2.0 ORGANIC SOLVENT CLEANER CHARACTERISTICS AND EMISSIONS
2.1 GENERAL
Organic solvent cleaners use halogenated and
nonhalogenated solvents, solvent blends, or their vapors to
remove water-insoluble soils such as grease, oils, waxes,
carbon deposits, fluxes, and tars from metal, plastic,
fiberglass, printed circuit boards, and other surfaces.
Organic solvent cleaning is performed prior to processes such
as painting, plating, inspection, repair, assembly, heat
treatment, and machining. The same type of machine that is
used in cleaning applications can also be used for drying wet
parts (by displacing surface moisture with solvent and
evaporating the solvent) and for conditioning the surface of
plastic parts.
Organic solvent cleaning does not constitute a distinct
industrial category, but is an integral part of many major
industries. The five 2-digit Standard Industrial
Classification (SIC) codes that use the largest quantities of
halogenated solvents for cleaning are SIC 25 (furniture and
fixtures), SIC 34 (fabricated metal products), SIC 36
(electric and electronic equipment), SIC 37 (transportation
equipment), and SIC 39 (miscellaneous manufacturing
industries). Additional industries that use halogenated
solvents in cleaning include SIC 20 (food and kindred
products), SIC 33 (primary metals), SIC 35 (nonelectric
machinery), and SIC 38 (instruments and clocks).
Nonmanufacturing industries such as railroad, bus, aircraft,
and truck maintenance facilities; automotive and electric tool
repair shops; automobile dealers; and service stations
2-1
-------
(SIC 40, 41, 42, 45, 49, 55, and 75, respectively) also use
organic solvent cleaners.
The five most commonly used halogenated solvents are
methylene chloride (MC), perchloroethylene (PCE),
trichloroethylene (TCE), 1,1,1-trichloroethane (TCA), and
trichlorotrifluoroethane (CFC-113). Chloroform and carbon
tetrachloride may also be used. These solvents can be used
alone or in blends that contain two or more halogenated
solvents, and sometimes alcohols or other solvents.
Use of TCA is expected to decline as a result of the
phaseout mandated by the Clean Air Act as amended (1990
Amendments), and will be completely eliminated from solvent
cleaning operations by the end of 1995 because of a
presidential edict.
Machines with controls that limit emissions of other
solvents will limit CFC-113 emissions because their compound
properties are similar. Therefore, CFC-113 is included in
discussions of solvent cleaning, emissions characteristics,
and control devices in this background information document
(BID). However, CFC-113 is not included in the regulatory
analyses, such as for cost or environmental impacts. Examples
cc nonhalogenated solvents typically used are mineral spirits,
;oddard solvents, and alcohols.
The remainder of this chapter describes typical organic
solvent cleaning processes and emissions from machines using
halogenated solvents. Section 2.2 describes the various types
of cleaners. Section 2.3 identifies emission mechanisms and
presents test data on cleaner emission rates and section 2-4
presents national baseline emissions.
2.2 ORGANIC SOLVENT CLEANING EQUIPMENT AND PROCESSES
There are two basic types of solvent cleaning equipment:
batch cleaners and in-line cleaners. Both cleaner types can
be designed to use either solvent at room temperature (cold
cleaners) or solvent vapor (vapor cleaners). Each category of
solvent cleaners is described in this section.
2-2
-------
The vast majority of halogenated solvent use is in
vapor cleaning, both batch and in-line. The most common type
of batch cleaner that uses halogenated solvents is the
open-top vapor cleaner (OTVC). Cold cleaners typically use
mineral spirits, Stoddard solvents, and alcohols. Very little
halogenated solvent use has been identified in batch cold
cleaning.
In 1991, an estimated 95,000 megagrams (Mg)
[105,000 tons] of halogenated solvents were used by batch
vapor cleaners (mostly OTVC's); 35,000 Mg (39,000 tons) by in-
line vapor cleaners; 59,000 Mg (65,000 tons) by cold cleaning
operations (including carburetor cleaners and excluding
in-line cold cleaners); and 11,000 Mg (12,000 tons) by in-line
cold cleaners (including the estimated use for photoresist
stripping). There were an estimated 8,100 in-line cleaners
and 16,400 batch vapor cleaners using halogenated solvents in
1991.1*2
Batch vapor cleaner descriptions are presented in
section 2.2.1. Section 2.2.2 presents information on in-line
cleaners (batch and vapor), and section 2.3.3 contains
descriptive information on batch cold cleaners.
2.2.1 Batch Vapor Cleaners
Batch vapor cleaners represent the largest source of
halogenated solvent emissions from batch solvent cleaners.
The most common type of batch vapor cleaner is the OTVC;
however, other (non-OTVC) batch cleaners have been developed
to meet the specific needs of a variety of industries.
2.2.1.1 OTVC's. Open-top vapor cleaners are used
primarily in metalworking operations and other manufacturing
facilities. They are seldom used for ordinary maintenance
cleaning because this type of cleaning can usually be
performed at a lower cost by cold cleaners using petroleum
distillate solvents. Exceptions include maintenance cleaning
of electronic components, small equipment parts, and aircraft
parts, where a high degree of cleanliness is needed.
2-3
-------
A basic OTVC, shown in figure 2-1, is a tank designed to
generate and contain solvent vapor. At least one section of
the tank is equipped with a heating system that uses steam,
electricity, hot water, or heat pumps to boil liquid solvent.
As the solvent boils, dense solvent vapors rise and displace
the air inside the tank. The solvent vapors rise to the level
of the primary condensing coils. Coolant (such as water or
refrigerant) is circulated or recirculated through the
condensing coils to provide continuous condensation of rising
solvent vapors, thereby creating a controlled vapor zone that
prevents vapors from escaping the tank. Condensing coils
generally are located around the inside walls of the cleaner,
although in some equipment the primary coils are offset at one
end or side of the cleaner.
2.2.1.1.1 The cleaning cvcle. During the vapor cleaning
operation, solvent vapors condense on the cooler workload
entering the vapor zone. Condensing solvent dissolves some
contaminants and flushes both dissolved and undissolved soils
from the workload. Condensed solvent and dissolved or
entrained contaminants then drain back into the sump below.
When the temperature of the workload reaches that of the
vapor, condensation ceases and the vapor phase cleaning
process is complete. In many instances, the vapor cleaning
cycle is supplemented, or even replaced, by the immersion of
the part into the hot, liquid solvent.
Organic impurities (greases, soils, etc.) cleaned from
parts will accumulate in the solvent sump. However, because
of their higher boiling points, these impurities do not
appreciably contaminate the solvent vapors. Because the
solvent vapor remains relatively pure, solvent can ,be used for
longer periods with vapor cleaning than with cold cleaning,
where the solvent often becomes contaminated with dissolved
and suspended impurities more quickly. Eventually,
accumulated impurities will compromise the performance or
safety of vapor cleaners. To avoid these problems,
2-4
-------
Primary
Condensing Coils
Temperature
Indicator
Cleanout Door
Freeboard
Solvent Level Sight Glass
Condensate Trough
Water
Separator
.Heating Elements
Work Rest and Protective Grate
Figure 2-1. Open top vapor cleaner.
2-5
-------
contaminated solvent is periodically drained from the machine
and replaced with fresh solvent.
Alternatively, a still adjacent to the cleaner can be
used to extract soils building up in the solvent sump and
return clean solvent to the machine. The solvent feed system
to the still can include a filter to remove insolubles such as
metal fines. Using a still can increase the useful life of
solvent and will concentrate the impurities. Solvent cleaning
operation vastestreams are considered hazardous wastes under
the EPA's regulations implementing the Resource Conservation
and Recovery Act (RCRA), and the lover volume, concentrated
vastestream from the still can decrease the expense of proper
disposal.
2.2.1.1.2 Design parameters. Air currents within an
OTVC can disturb the vapor zone and cause excessive solvent
emissions. All machines have covers of varying design to
limit solvent losses and contamination during downtime or
idling. Additional control of the solvent vapor is provided
by the freeboard, which is that part of the tank wall
extending from the top of the solvent vapor level to the tank
lip. The freeboard reduces the effect of room draft.
The freeboard ratio (FBR), or ratio of freeboard height
to machine width (the smaller dimension of the vapor-air
interface area), usually ranges from 0.75 to 1.0, depending on
the manufacturer's design. The FBR can be as low as 0.5 on
some older machines. Increasing the FBR reduces disturbance
of the vapor zone due to workplace air currents and slows
solvent diffusion out of the machine.
Moisture may enter the OTVC on workloads and can also
condense from ambient air on primary cooling coils or
freeboard refrigeration coils (see section 3.0). If allowed
to accumulate, water in an OTVC will lead to higher emissions
and may contribute to solvent decomposition and corrosion in
the cleaner. Therefore, nearly all vapor cleaners are
equipped with a water separator based on the principle
depicted in figure 2-2. The condensed mixture of water and
2-6
-------
GooRnQ Syvfewn
Coolant MM
CootantOuMI
Witw Overflow
Solvent Outtt
Dram
Figure 2-2. Water separator with cooling coil,
2-7
-------
solvent is collected in a trough below the condenser coils and
directed to the water separator. The water separator is a
simple container where the water phase (being essentially
immiscible with, and less dense than, halogenated solvents)
separates from liquid solvent. The water is directed to
disposal while solvent is allowed to return to the cleaner.
Cooling coils may be used inside the separator to cool
condensed solvent and enhance solvent/water separation.
To further reduce water contamination, or to replace the
water separator, some manufacturers produce machines that use
a canister of desiccant, such as a molecular sieve. Use of
desiccants prevents prolonged contact between water and
solvent, where removal of water-soluble stabilizers or
co-solvents (such as alcohols) from certain solvents and
blends can occur. Desiccants also prevent corrosion of the
solvent due to hydrolysis.
2.2.1.1.3 Design variations. Variations in design of
vapor cleaners reflect their many industrial applications.
Workload characteristics and the degree of cleanliness
required by the particular application dictate many additional
features on the basic model. Additional examples of vapor
cleaners are shown in figures 2-3 and 2-4. These figures show
OTVC's with two chambers: one for generating the solvent
vapor, the other for immersion cleaning or spraying
applications.
Some units with multiple sumps include an ultrasonics
chamber. In a machine using ultrasonics, high-frequency sound
waves are used to produce pressure waves in the liquid
solvent. In areas of low pressure within the liquid, minute
vapor pockets are formed. These pockets collapse as the
pressure in the zone cycles to high pressure. The constant
creation and collapse of these vapor pockets (called
cavitation) provides a scrubbing action to aid cleaning.
Ultrasonically agitated liquids often need to be heated to
specific temperatures to achieve optimum cavitation.
2-8
-------
Primary
Condensing
Coils
Solvent
Vapor Zone
Warm Solvent
Overflow Dam
Solvent
Boiling Chamber
Freeboard
Water
Separator
Warm Solvent or Ultrasonics
Immersion Chamber
Figure 2-3. Two compartment vapor cleaner,
2-9
-------
Primary
Condensing Coils
Offset Solvent
Boiling Chamber
Freeboard
Solvent Warm Solvent or Ultrasonics
Overflow Dam Immersion Chamber
Figure 2-4. Two compartment vapor cleaner with offset
boiler chamber.
2-10
-------
Another design modification included in some units is a
lip or slot exhaust, such as those shown in figure 2-5. These
exhausts are designed to capture solvent vapors escaping from
the OTVC and carry them away from the operating personnel in
order to reduce occupational exposures. However, these
exhaust systems also disturb the vapor zone and enhance
diffusion, thereby increasing solvent losses. The increased
losses can be significant. In properly designed lip exhaust
systems, the cover closes below the lip exhaust inlet level.
The effect of lip exhausts is discussed further in
section 3.0.
2.2.1.1.4 Operational variations. The cleaning cycle is
often modified from a basic vapor cleaning cycle in order to
obtain an effective cleaning. One OTVC operational variation
is an immersion-vapor-spray cycle. In this design, the
workload is lowered into a warm or boiling immersion
compartment for precleaning. The immersion compartment may be
equipped with ultrasonics (see section 2.2.1.1.3). After this
first stage of cleaning is completed, the workload is cleaned
in a vapor section and then sprayed.with solvent.
Many other cleaning cycles are possible, including some
that incorporate nonboiling solvent sections with vapor
sections. Spraying may not be necessary or desirable for some
applications.
Another common variation in cleaning cycle design is a
vapor-spray-vapor cycle. In this design, the workload is
lowered into the vapor zone, where the condensing solvent
performs the preliminary cleaning. After condensation ceases,
the workload is sprayed with warm solvent. The pressure of
the spray aids in physical removal of soil. In some cases,
the warm spray may be cooler than the workload and will lower
the workload temperature, promoting further solvent /
condensation on the workload. The spray nozzle must be below
the vapor line to avoid spraying solvent directly to the
atmosphere and directed downward to avoid turbulence at the
air/solvent vapor interface.
2-11
-------
Exhaust Outlet
(to control device
or to atmosphere)
Lip Exhaust
Inlet
Primary
Condensing Coils
Figure 2-5. Open top vapor cleaner with lip exhaust.
2-12
-------
Parts cleaning in an OTVC can be performed either
manually or with the use of an automated parts handling
system. In manual operation, the attendant must lower the
parts basket into the cleaner and remove the basket after the
cleaning has been completed. An electrically operated parts
handling system can be operated by push buttons; some can be
programmed to cycle parts through the cleaning cycle
automatically. With a hoist, the speed of parts entry and
removal can be controlled and will be consistent from cycle to
cycle. The effect of automated parts handling systems on
emissions is discussed in section 3.0.
2.2.1.2 Other Batch Cleaners. As the solvent cleaning
industry has developed, a multitude of cleaner designs have
been developed for specialized cleaning needs. Other
(non-OTVC) batch cleaners are a hybrid of an OTVC and a
continuous cleaner. They tend to be larger and have more of
an enclosed design than typical OTVC's, with conveyorized
systems for moving large parts and parts baskets. Non-OTVC
types of cleaners include the cross-rod, the vibra, the ferris
wheel, and the carousel cleaners.
The cross-rod cleaner (figure 2-6) derives its name from
the rods that suspend the parts baskets as they are conveyed
through the machine by a pair of power-driven chains. The
parts are contained in pendant baskets or perforated or wire
mesh cylinders when tumbling of the parts is desired. These
cylinders may be rotated within the liquid solvent and/or the
vapor zone. This type of equipment works well for small parts
that require immersion in solvent for satisfactory cleaning or
tumbling to remove metal chips and/or drain solvent from
cavities.
In the vibra cleaner (figure 2-7), soiled parts are fed
through a chute into a pan flooded with boiling solvent. The
pan, which is at the bottom of the cleaner, is connected to a
vibrating spiral elevator. Both the pan and spiral elevator
vibrate, causing the parts to move from the pan up the spiral
to the exit chute. The cooler parts condense solvent vapor as
2-13
-------
Conveyor
Path
Chain
Supports
Work /
Basket'
Water
Jacket
Boiling Chamber
Figure 2-6. Cross-rod cleaner.
2-14
-------
Workload Discharger Chute
Ascending
Vibrating
Trough
Distillate Return
for Counter-
flow Wash
Figure 2-7. vibra cleaner.
2-15
-------
they are vibrated up the spiral and dry as soon as they leave
the vapor zone. Vibra cleaners are capable of processing
large quantities of small parts; however, the vibrating action
creates considerable noise; therefore the equipment must be
acoustically insulated or enclosed in a noise-control booth.
The ferris wheel cleaner (figure 2-8) is one of the least
expensive and smallest of the non-OTVC batch cleaners. It
commonly features perforated parts baskets, as does the
cross-rod cleaner. As a large gear wheel rotates, it tumbles
the perforated baskets attached to it, allowing better contact
of the parts with the solvent and draining cavities that could
otherwise retain solvent.
The carousel cleaner (not shown) is a four-chamber
machine similar to the ferris wheel cleaner except that parts
travel on a horizontal plane. The first chamber is the
loading area. The remaining three chambers are cleaning
units. All cleaning chambers can contain halogenated solvent
(typically vapor phase with or without immersion sumps), or
one chamber can be used for steam cleaning. Usually, this
type of machine is used to clean large parts such as airplane
wheels. In operation, a four-arm carousel carries the parts
to be cleaned sequentially through each of the four chambers.
2.2.2 In-Line Cleaners
In-line cleaners (also called continuous cleaners) employ
automated loading on a continuous basis. Although in-line
cleaners can operate in the vapor or nonvapor phase, the
majority of all in-line machines that use halogenated solvents
are vapor cleaners.
A continuous or multiple-batch loading system greatly
reduces the manual parts handling associated with batch
cleaning. The same cleaning techniques are used in in-line
cleaning but usually on a larger scale than with batch units.
To help control solvent losses from the system, in-line
cleaners are nearly always enclosed, except for parts/conveyor
inlet and exit openings. In-line cleaners are used by a broad
spectrum of industries but are most often found in plants
2-16
-------
Work
Basket
Boiling
Chamber
Gear to Tumble
Baskets
Figure 2-8. Ferris wheel cleaner,
2-17
-------
where there is a constant stream of parts to be cleaned, in
such instances, the advantages of continuous cleaning outweigh
the lower capital cost of the batch cleaners. Usually, an
in-line cleaner is individually designed for a specific
workload and production rate situation, rather than being an
"off the shelf item.
There are four main types of in-line cleaners that use
halogenated solvents: monorail, belt, strip, and printed
circuit board processing equipment (photoresist strippers,
flux cleaners, and developers). In addition, the cross-rod
cleaner described in section 2.2.1.2 could be designed as a
continuous cleaner with both an entry and exit post.
A monorail vapor cleaner (figure 2-9) is usually chosen
when the parts to be cleaned are being transported between
manufacturing operations on a monorail conveyor. The monorail
cleaner is well-suited to automatic cleaning with solvent
spray and vapor. It can be of the straight-through design
illustrated or can incorporate a u-turn within the machine so
that parts exit through an opening parallel to the entrance
opening. The u-turn monorail cleaner has lower vapor loss
because the design eliminates the possibility of drafts
flowing through the machine.
Both the belt cleaner (figure 2-10) and the strip cleaner
are designed to allow simple and rapid loading and unloading
of parts. A belt cleaner conveys parts through a long and
narrow boiling chamber where the parts are cleaned either by
the condensing vapor or by immersion in the solvent sump. The
strip cleaner is similar to the belt cleaner except that the
strip itself is the material being cleaned. As with the belt
cleaner, th6 material in a strip cleaner can be cleaned by the
condensing vapor or by immersion in the solvent sump.
The cleaning of printed circuit boards is a common
>
application of a mesh belt cleaner (figure 2-11). In the
production of printed circuit boards, solvent-based
photo-processable resists can be used. The circuit pattern is
contained in an artwork film. This pattern is reproduced by
2-18
-------
Conveyor
Path
Conveyor
Path
Spray
Pump
Boiling
Chamber
Water
Jacket
Figure 2-9. Monorail in-line cleaner.
2-19
-------
0)
c
m
o
0)
c
-p
--I
0)
0)
£
o
iH
I
0)
3
2-20
-------
CO
•H
to
0
h
O
4J
O
a
Q)
c
I
c
c
a
0)
«J .
'S
m
o ST
._! »-
E
0)
O
to
a
a
-H
J-l
4J
U)
S
I
O)
0)
b
3
2-21
-------
projecting ultraviolet rays through the artwork film onto a
copper sheet covered with resist. A developer (typically TCA)
dissolves the unexposed areas of the resist, thereby revealing
the circuit pattern. The resist-covered board is then placed
in plating solutions to add more metal to the circuit pattern
areas. Next, a photoresist stripper dissolves the remaining
resist. The circuit boards are then put in an alkaline
etching solution to remove all the copper in the noncircuitry
areas. The processing is completed by passing the circuit
boards through a wave of molten solder.
Because of the nature of the materials being cleaned,
photoresist strippers use ambient (room temperature) solvents.
Spraying and brushing may be used to enhance cleaning.
Methylene chloride is the solvent most often used in
photoresist stripping; however, the printed circuit board
industry has largely converted to aqueous and semi-aqueous
materials to replace the use of both TCA and MC. The switch
to aqueous systems is discussed further in section 3.0.
Circuit board cleaners are used to dissolve and remove
flux from the circuit board after the molten soldering step.
Unlike photoresist strippers, circuit board cleaners have a
heated or boiling sump. Although a vapor phase may be
present, circuit board cleaning occurs in the liquid solvent
phase. Circuit board cleaners commonly use chlorofluoro-
carbons; however, aqueous fluxes and aqueous flux cleaners are
becoming more widely used in the printed circuit industry.
This switch is also discussed further in section 3.0.
2.2.3 Batch Cold Cleaners
Cold cleaners use room-temperature liquid solvent for
parts cleaning. Most cold cleaners are small maintenance
cleaners or parts washers that use aliphatic petroleum
distillates such as mineral spirits or sometimes alcohol
blends or naphthas. These nonhalogenated solvents are not
covered in this document.
Cold cleaning operations include spraying, flushing,
solvent or parts agitation, wipe cleaning, and immersion. The
2-22
-------
only identified machines currently being manufactured that use
halogenated solvents in a cold cleaning application (except
for nonvapor in-line cleaning) are of a type called carburetor
cleaners. In these cleaners, MC is blended with other
solvents and additives to reduce flammability and increase
dissolving power. A typical carburetor cleaner is shown in
figure 2-12. Emissions from these cleaners are typically well
controlled because the cleaning solution used contain* water,
which forms as a layer above the solvent mixture in the tank.
The water layer significantly reduces evaporation of MC.
2.3 EMISSION MECHANISMS AND TYPES
There are many types of solvent losses to the atmosphere
from an organic solvent cleaner. Two significant types are
air/solvent vapor interface losses and workload-related losses
(hereinafter called workload losses). Air/solvent vapor
interface losses during idling (i.e., when the machine is
turned on and ready to operate) consist of solvent vapor
diffusion (or evaporation from liquid solvent in a cold
cleaner) and solvent vapor convection induced by warm
freeboards.
Workload losses are solvent emissions that are created or
increased by the introduction and extraction of parts during
the cleaning process and by spraying of parts during cleaning
(if sprays are used). Other potentially significant losses
that contribute to total solvent emissions from a solvent
cleaner include filling/draining losses, wastewater losses,
start-up/shutdown losses, downtime losses, and leaks from the
cleaner or associated equipment. Diffusion and convection
losses are described in section 2.3.1. Workload and other
losses are described in sections 2.3.2 and 2.3.3,
respectively.
Because non-OTVC batch vapor units tend to be of a design
that is more enclosed than an OTVC, their emission mechanisms
are more closely related to the mechanisms for in-line units.
Therefore, OTVC emission mechanisms are discussed separately.
2-23
-------
Air Motor and
Drive Assembly
/r
Basket
"On" and "Off"
Valve
Figure 2-12. Carburetor cleaner,
2-24
-------
2.3.1 Air/Solvent Vapor Interface Losses Purina Idling
(Idling Losses)
2.3.1.1 OTVC's. The principal emission sources in
idling OTVC's are shown in figure 2-13. These losses can be
increased dramatically by external factors.
The main source of idling losses from an OTVC is
diffusion, which is the movement of solvent vapors from the
vapor zone to the ambient air above. Diffusion occurs because
molecules of solvent move from higher concentrations in the
vapor zone to lower concentrations in the air. Because
molecular activity increases at higher temperatures, diffusion
rates are temperature-dependent. An idling machine will reach
a point where a steady state diffusion rate is established.
At this point the emission rate will not fluctuate greatly
unless steady state conditions are disturbed.
Additional losses can be caused by convection. The heat
of the boiling solvent is conducted from the boiling solvent
and hot vapor to the walls of the solvent cleaner. This
heating of the walls creates a convective flow up along the
freeboard, carrying solvent vapor out of the cleaner. The
amount of convective loss depends on how warm the freeboard
walls become. If OTVC walls are kept close to ambient
conditions, convective losses will be minimized. Some
machines have a water jacket around the periphery of the
cleaner to help cool the walls of the machine and reduce
convective losses. However, a water jacket is not necessary
on all machines, as, for example, when adequate cooling of the
tank walls is provided by primary coils in contact with the
OTVC walls.
The diffusion rate steady state can also be disturbed if
an air flow is introduced across the air/solvent vapor
interface as the result of room drafts or a lip exhaust. Room
drafts create turbulence in the interface area, which can
cause the air and solvent vapor to mix. This creates a
mixture that is lighter than the pure solvent vapor and,
therefore, more readily lost to the atmosphere. Room drafts
2-25
-------
M
01
o
14
0
M
O
•H
w
w
•H
0)
•H
Q)
C
(0
0)
r-l
O
o
-p
n
1-1
I
CM
0)
2-26
-------
also sweep solvent-laden air from the freeboard area into the
ambient air. This allows more solvent to diffuse into the
"fresh air" in the freeboard area.
Lip (or lateral) exhausts create similar disturbances
across the air/solvent vapor interface. The exhaust system
draws in solvent-laden air from around the top perimeter of
the cleaner in order to lower the solvent concentration in the
area where operators are working. As discussed in
section 3.0, these exhausts do not capture all of the vapors
that escape from the cleaner. Properly operating lip exhausts
can double vapor cleaner diffusion losses. Some lip exhaust
systems include carbon adsorbers to collect the exhausted
solvent for reuse; however, emissions not captured by the lip
exhaust remain uncontrolled.
A summary of the available idling emission data for
OTVC's is included in table 3-3 of section 3.0. All of the
data were obtained on uncovered machines with no refrigerated
freeboard devices or lip exhausts. The emission rates range
from .3 kilogram/square meter/hour (kg/m2*hr)
[0.06 pound/square feet/hour (Ib/ft2*hr)] to 0.85 kg/m2*hr
(0.17 Ib/ft2*hr). The variation in emission rates for the
same solvent can be explained by the varying primary
condensing temperatures during these tests. Emission rates
are lowest in tests where the primary condensing temperature
of the cleaner is lowest. The use of a reduced primary
condenser temperature as a control technology is discussed in
more detail in section 3.0. At the mid-range primary
condensing temperature during the tests (table 2-1, tests 3
and 6), the emissions ranged from 0.45 kg/m2*hr
(0.09 Ib/ft2*hr) to 0.60 kg/m2*hr (0.12 Ib/ft2*hr).
2.3.1.2 In-Line Cleaners and Batch Non-OTVC's. The
primary sources of idling losses from in-line and non-OTVC
batch vapor cleaners are the same as those for OTVC's:
convection and diffusion. These types of losses are
illustrated in figure 2-14 and the mechanisms are described in
detail above. No data were available on idling losses from
2-27
-------
TABLE 2-1. HALOGENATED SOLVENT EVAPORATION RATES
Relative evaporation ratea
Solvent (CClA - 100)
TCE 84
PCE 39
TCA 100
MC 147
CFC-113 170
aReference 3.
2-28
-------
= r
> "a-
(0
0)
U
^
9
O
(0
o
•H
to
(0
•H
0)
0)
C
(0
01
I— I
o
0)
C
I
C
(N
a>
u
3
2-29
-------
in-line cleaners. However, the idling diffusion and
convection losses from these cleaners would likely be less per
unit of air/solvent vapor interface area than from an OTVC
because the units are enclosed and less subject to drafts.
2.3.1.3 Cold Cleaners. The source of solvent losses
from an idle cold cleaner is evaporation from the liquid
surface and subsequent diffusion. The rate of solvent loss is
solvent-dependent and is affected by room drafts. As with
OTVC's, room drafts can remove solvent-laden air from above
the liquid surface, thus increasing steady state evaporation
rates over quiescent conditions. However, the only identified
type of cold cleaner currently being manufactured that uses a
halogenated solvent is a carburetor cleaner. As mentioned
previously, these units typically use MC and have water
covers. Because the solvent is heavier than water and only
slightly soluble in water, little solvent reaches the air
interface and evaporates.
2.3.2 Workload Losses
As stated above, workload losses are defined as all
losses that are created or increased by the cycling of parts
through the solvent cleaner. During the operation of a
solvent cleaner, the losses at the air/solvent vapor interface
continue. However, the rate of these losses will be increased
by disturbances caused by the parts cleaning.
2.3.2.1 OTVC's. Figure 2-15 illustrates the losses that
occur during workload entry, cleaning, and workload removal.
One cause of increased losses during solvent cleaner operation
is the turbulence in the air/solvent vapor interface that is
created when parts and parts baskets enter the cleaner. This
loss result's from the increase in diffusion and convection
losses that occur at the air/solvent vapor interface. The
amount of loss depends on the speed of the basket, as well as
the characteristics of the parts being cleaned. Some loss can
occur when solvent vapor is displaced out of the cleaner by a
piston-type effect as parts are lowered into the cleaner. The
2-30
-------
CO
Q)
O
fc
i
to
0
-H
CO
to
•H
0)
•O
0)
*J
10
iH
0)
u
I
•o
(0
o
r-l
X
t-l
o
OOOfl
N
is
CO
2-31
0)
c
10
0)
I—I
o
o
CQ
•
in
r-4
I
CM
0)
3
-------
amount of loss from parts entry is increased as the speed of
parts introduction increases.
The piston effect is greater when the parts and baskets
take up a larger percentage of the interface area. It is
generally recommended that workloads not take up more than
50 percent of the total interface area, although large
workloads can be used if the lowering speed is very slow.
Also, if a large part is being lowered into a cleaner, the
part can possibly be angled to limit the piston effect.
Vapor line fluctuation also contributes to solvent loss.
Several factors can affect the degree of vapor line
fluctuation. If very cold parts or a large quantity of parts
are introduced into the cleaner, more heat will be required to
bring the parts up to the temperature of the solvent vapor.
When the heat is transferred from the solvent vapor to the
parts, the vapor line lowers. As the vapor line rebuilds and
rises back to its original level, the air/solvent vapor
mixture above the layer is displaced out of the cleaner. One
manufacturer has determined through testing that solvent loss
rates begin to increase substantially when the vapor line is
deflected by more than 6.25 centimeters (cm) [2.5 inches
(in.)].2 These test data also indicated that solvent loss
rates are about twice as high at a deflection of 25 cm
(10 in.) as they are at a deflection of 2.5 inches.
During parts cleaning, additional losses can occur if
sprays are used to aid in cleaning. Spraying from either
fixed nozzles or spray wands is common. The sprayed solvent
can cause turbulence in the air/solvent vapor interface and
vapor line lowering, thereby increasing emissions. If the
spray pressure is too high, splashing of the solvent against
the parts, parts basket, or wall of the cleaner can also
increase emissions. Both of these spray sources should be
/
mounted so that spraying occurs only beneath the vapor zone.
As parts are removed from the cleaner, the air/solvent
vapor interface is again disturbed. As with workload entry,
the speed of workload removal directly affects the amount of
2-32
-------
solvent loss. The effect of parts movement rate on emission
rates is discussed in section 3.0. A large portion of this
loss is vapor entrainment. If parts are extracted rapidly/
solvent vapor will be entrained behind the workload and pulled
out of the cleaner (wake effect).
A final source of loss during workload removal is liquid
carry-out. This includes liquid pooled .in cavities or on flat
horizontal surfaces of the parts, as well as the solvent film
remaining on all surfaces of clean parts as they leave the
cleaner. If the workload is withdrawn slowly and allowed to
dwell in the freeboard area (if needed), then the solvent film
and much of the pooled solvent can evaporate before the
workload is withdrawn. A significant portion of this
evaporated solvent in the freeboard area will sink back into
the vapor layer or be condensed on the.coils and return to the
cleaner. If, however, the workload is withdrawn quickly, most
liquid solvent will not evaporate from the parts until after
they are withdrawn from the cleaner.
It is very difficult to remove parts slowly by manual
operation. Generally, manually operated cleaners will have
high workload losses, and these losses will dominate other
losses from the machine.
A summary of the available data on working emission rates
(i.e., diffusion/convection and workload losses combined) is
presented in table 3-4 of section 3.0. The emission rates
range from 0.3 kg/m2*hr (0.06 Ib/ft2*hr) to 3.9 kg/m2*hr
(0.78 Ib/ft2*hr), with most rates in the range of about 0.5 to
1.5 kg/m2*hr (0.1 to 0.3 Ib/ft2*hr). The large variability in
the data is' due to the wide range of operating parameters
during the tests. Unlike idling emissions, which are more a
factor of machine design, workload emissions are largely a
factor of the operating parameters previously discussed in
this section.
The speed of parts movement in many of the tests is
unknown. All of these tests were performed using electric
hoists for parts entry and removal. Test results with
2-33
-------
manually operated machines would be significantly higher
because it is difficult on impossible for a human operator to
consistently achieve the low workload-related losses exhibited
by hoists. As stated above, the speed of the parts can
directly affect the emissions from a cleaner.
Furthermore, these tests also included a wide range of
room air speeds, which can also affect emission rates. The
tests of idling rates did not include lip exhausts, which
would greatly increase emissions. A more complete discussion
of the effects of operating parameters on emission rates is
presented in section 3.0.
2.3.2.2 In-line Cleaners and Batch Non-P^VC's. The
principal sources of workload emissions from in-line cleaners
are presented in figure 2-14 (above). Many of the losses are
similar to the losses from OTVC's. Because in-line and non-
OTVC batch systems are automated, the workload losses are less
on a per-part-basis than in a manually operated OTVC.
However, because of the large volume of parts cleaned in these
systems, overall losses are typically higher from in-line
cleaners than from OTVC's.
The loss due to turbulence at the air/solvent vapor
interface (air/solvent interface with in-line cold cleaners)
caused by parts entry and exit is generally less for these
cleaners than for manually operated OTVC's because automated
parts handling allows better control of parts entry and exit
speed. However, if the conveyor speed is too high,
considerable turbulence will be generated, and parts may exit
the cleaner wet with solvent. The piston effect is also
lessened because in-line machines have large air/solvent vapor
interfaces (air/solvent interface with in-line cold cleaners)
relative to the size of the parts and baskets. In general,
States that have solvent cleaner regulations limit the
conveyor speed to 3.35 meters per minute (mpm) [11 feet per
minute (fpm)].
Solvent loss due to vapor line fluctuation is not as
significant a problem for in-line and enclosed batch vapor
2-34
-------
cleaners as for OTVC's. Because there is a constant flow of
parts into in-line vapor cleaners, the heat balance of the
machine can be adjusted to compensate for the constant thermal
shock. This practice tends to limit vapor line fluctuation in
these machines.
During parts cleaning, additional losses can occur if
spraying is employed. Spraying is done from either fixed
nozzles, spray wands, or rotating arms. The solvent spray can
cause turbulence within the cleaner and thereby increase
emissions, although the enclosure around in-line machines
would help minimize losses to the atmosphere.
The configuration of entry and exit openings will
influence the amount of loss from turbulence inside the
machine. If the spray pressure is too high, splashing of the
solvent against the parts, parts basket, or wall of the
cleaner can also increase emissions. Fixed or rotating spray
nozzles should be mounted so that spraying occurs only beneath
the vapor zone. For in-line cold cleaners, spraying should
occur only at a downward angle into the machine unless the
spray section is baffled to effectively shield air/solvent
interface from the effects of the spray.
Some manufacturers have developed cleaners that have high
pressure spray zones completely segregated from the
air/solvent vapor interface. These machines are discussed in
section 3.0.
As parts are removed from the cleaner, more disturbances
of the air/solvent vapor or air/solvent interface can occur.
Again, the speed of the parts movement can directly affect the
amount of solvent loss. The effect of parts movement rate on
emission rates is discussed in section 3.0. The majority of
this loss is vapor entrainment. If parts are extracted
t
rapidly, solvent vapor will be entrained behind the workload
and pulled out of the cleaner.
Another source of loss during part removal is liquid
carry-out. This includes liquid solvent pooled in cavities or
on flat horizontal surfaces of parts, as well as the solvent
2-35
-------
film remaining on all surfaces of clean parts as they leave
the cleaner. As discussed in section 2.3.2.1, the speed of
part removal can affect these losses. Some in-line cleaners
include a drying tunnel to allow for evaporation of solvent
before parts exit the cleaner.
Many in-line cleaners also have an exhaust system. This
exhaust system (see figure 2-14) can increase solvent
consumption. If solvent in the exhaust is not controlled by a
carbon adsorber before being vented to the atmosphere, overall
solvent emissions will increase.
2.3.2.3 Cold Cleaners. Workload losses from cold
cleaners are primarily due to agitation and spraying of the
solvent during cleaning and to carry-out (and subsequent
evaporation) of liquid solvent on parts being removed from the
machine. Carry-out losses may be substantially reduced by
allowing longer drainage time and by tipping parts to drain
solvent-filled cavities before removal from the cleaner.
Agitation can increase evaporation from the solvent bath
by increasing the effective air/solvent interface area. The
amount of solvent loss depends on the rate of agitation. In
the case of carburetor cleaners, the water layer over the
solvent bath minimizes the loss from increased turbulence. A
water layer could also reduce similar losses from other cold
cleaners. Spraying can increase the amount of solvent
evaporation by exposing more solvent to the air. The amount
of solvent loss from spraying depends on the spray pressure
(which influences turbulence and splashing).
2.3.3 Other Losses
In addition to idling and working operations, several
other operations or mechanisms can contribute to overall
solvent losses from an organic solvent cleaner, including
/
downtime operations, leaks, filling/draining, wastewater,
start-up/shutdown, distillation/sludge disposal, and solvent
decomposition. The magnitude of these losses relative to
total loss is dependent on machine design, machine integrity,
and operating techniques. For example, poor technique during
2-36
-------
filling and emptying of the cleaner can cause spills that
could amount to a large portion of overall losses from an
otherwise well-operated and maintained machine. Similarly, a
leak that goes undetected and uncorrected can also be a large
source of emissions. A brief discussion of these losses is
presented below.
2.3.3.1 Downtime. Downtime losses are defined as
solvent loss when the heat to the sump is turned off and the
machine is not operating. The losses are due to evaporation .
of solvent from the liquid solvent surface and subsequent
diffusion into the ambient air. These losses can be slowed
through use of a tight-fitting cover during downtime.
However, even with covers in place, the more volatile
halogenated solvents will evaporate at significant rates.
Relative evaporation rates of the halogenated solvents are
presented in table 2-1. Equipment vendor estimates of
downtime losses range from 0.15 kg/m2*hr (0.03 Ib/ft2*hr) to
0.35 kg/m2*hr (0.07 Ib/ft2*hr), comparable to the low end of
idling loss rates.4 Losses will be greatest from machines
using solvents with a higher vapor pressure, such as MC,
CFC-113, or solvent blends made with MC or CFC-113.
2.3.3.2 Leaks. Loss of solvent through leaks can occur
continuously (depending on where the leak is located), whether
the machine is turned on or off. Leaks can result from
manufacturing defects or from machine use. They can occur
from piping connections, cracks in the machine or tank, and
gasketed portholes or viewing windows. Leaks are often
difficult to detect because the solvent will evaporate quickly
when it reaches the atmosphere and may not leave telltale
drips or wet areas. Because solvent has a low surface
tension, it can escape through cracks that may not be easily
visible. These characteristics magnify the chance of leaks
becoming a serious source of solvent loss. Many manufacturers
leak-test their machines before they are sold, but cracks can
occur during shipping. If not detected and repaired, leaks
can become a major source of solvent loss.
2-37
-------
2.3.3.3 Filling/Draining. Filling and emptying of the
solvent cleaner, if not performed properly, can be a major
contributor to overall emissions. Open handling procedures/
such as manual filling or emptying of machines using open
buckets or drums, will cause significant solvent loss and
operator exposure. This loss will increase if a large amount
of splashing occurs during filling. If solvent is spilled
during filling or draining, the operator may be subject to
Comprehensive Environmental Response, Compensation, and
Liability Act (CERCLA) regulations, which require notification
of all spills above reportable quantities.
2.3.3.4 Wastewater. Water separators are typically
minor sources of solvent loss on vapor cleaners. The solvent
loss occurs when the water is decanted from the separator
containing a slight amount of solvent due to solvents being
slightly soluble in water. Water separators are used to
recover solvent from the solvent/water mixture that condenses
at the primary condenser or at the refrigerated freeboard
device. If not properly designed, freeboard refrigeration
devices may increase wastewater loss because they condense
water vapor as well as solvent from the atmosphere. However,
if a separator is correctly designed, operated, and
maintained, little solvent will be lost. Wastewater impacts
due to the use of a carbon adsorber to recover solvent are
discussed in section 3.0.
2.3.3.5 Start-up/Shutdown. Losses that occur during the
transition time from when a vapor solvent cleaner is turned on
or off to the time when steady state is achieved are called
start-up and shutdown losses.
Start-up losses are due to pump-out of solvent-laden air
within the machine after the sump heat has been activated and
as the solvent vapor layer is being established. One estimate
of start-up losses from a typical vapor cleaner is 11.4 liters
(L) [3 gallons (gal)] of solvent per cycle.5 However, the
amount of loss from a cleaner depends on the cleaner size and
design.
2-38
-------
Shutdown losses are due to evaporation of hot liquid
solvent from the sump (after the heat has been turned off and
the vapor layer has collapsed) and subsequent diffusion of
solvent vapor from the machine. If a cleaner is not
controlled, shutdown losses will be significant because the
solvent is near the boiling point at the beginning of the
shutdown period.
2.3.3.6 Distillation Losses/Sludge Disposal. Losses
occur when spent solvent is regenerated through on-site
distillation. Solvent lost during this process stems from
evaporation during transfer to and from the distillation unit
or, if a piping system is used, from leaks in the equipment.
Solvent may also evaporate from distillation sludge or spent
solvent that is removed for disposal.
2.3.3.7 Solvent Decomposition. Certain solvents and
blends contain stabilizers that prevent the mixture from
turning acidic after reacting with water (where water/solvent
contact occurs). If solvent is not properly monitored and
allowed to become acidic, the solvent will have to be
discarded. Dangerous fumes (chlorine gas, hydrochloric acid)
can be emitted from solvent decomposition. Emissions could
occur during handling and disposal of the solvent. This
solvent would be subject to hazardous waste guidelines under
RCRA.
2.4 NATIONAL BASELINE EMISSIONS
A baseline emissions level represents existing emissions
and reflects the level of emissions control achieved by an
affected industry in the absence of additional EPA standards.
The baseline emissions level is established to facilitate the
comparison of economic, energy, and environmental impacts of
regulatory alternatives. Table 2-2 presents the national
baseline emissions estimates for halogenated solvent batch
vapor and in-line cleaners. These estimates were calculated
using controlled and uncontrolled emission factors determined
for these cleaners in earlier analyses,6 cleaner size (solvent
vapor/air interface),6 solvent consumption estimates for
2-39
-------
CO
2
O
M
CO
CO
H
U
•J
w <-»
CO h
00 -^
J as
«£ %-*
z
o
M
3
2
CM
1
CM
W
J
CO
<
EH
,_!
id
O
EH
H
U
H
3
EH
•J;
0) U
> u
_-J rt
•HO,
CO
o
<•*
Q)
Q
4J
|
V4
leane
0
0
en
«-«
VO
O
CM
O\
H
O
O
in
n
o
in
in
• o
CM
CM
10
fc
%
s
0
H
CM
CM
O
H
rH
CO
H
0
vO
00
n
CM
O
CO
f-
n
0
in
H
^^
CM
in
n
0)
cn
id
eH
0)
o
in
VO
in
CO
0
0
VO
VO
CM
O
CM
CO
<*
O
p*
in
t^
o
VO
o
n
rH
id
I
0
VO
00
H
O
CO
0
H
O
VO
n
CM
o
o
CM
n
o
in
•w
ca
0)
c
id
(U
>-\
u
0)
•H
iH
1
H
0
p^
O
H
1
1
1
1
1
O
O
rt
ners
id
0)
rH
O
•a
•H
O
u
u
id
CO
o
CO
in
oo
CM
H
O
n
o\
VO
n
o
CO
H
CM
r^
O
^^
^^
o
H
O
O
CO
iH
id
0
•P
•a
id
o
Q)
C
(0
0) 0) 4-> Q)
-a c Q) c
-H a) o 0)
O >i O >i
rH JC I— 1 J3
J3 *J £ 4J
U 0) U 0)
O-H O
Q) >H V-i IH
C O EH O
Q) i— 1 1 (— 1
>i O » O
,C U H -H
•P
-------
1991;2 and the number of cleaners estimated to be controlled
versus uncontrolled based on section 114 questionnaire solvent
cleaner vendor responses.7
To estimate emissions, the solvent use allocated to
cleaners estimated to be uncontrolled at baseline was
multiplied by uncontrolled emission factors, expressed in
kilogram emitted per kilogram fresh solvent use. The solvent
use allocated to cleaners estimated to be controlled at
baseline was multiplied by controlled emission factors,
expressed in kilogram emitted per kilogram fresh solvent use.
Because of the numerous factors that affect OTVC and in-line
cleaner emission rates, no attempt was made to estimate
individual emission factors for each solvent. The "National
Baseline Emissions from Halogenated Solvent Cleaners"
memorandum presents the calculations used to derive the
national baseline emissions presented in table 2-2.8
2.5 REFERENCES
1. Memorandum from O'Loughlin, J., Radian Corporation to
Almod6var, P., U. S. Environmental Protection Agency.
May 26, 1993. Estimate of the number of halogenated
solvent cleaners.
2. Memorandum from O'Loughlin, J., Radian Corporation to
Almod&var, P., U. S. Environmental Protection Agency.
June 2, 1992. Estimates of national solvent consumption
by the solvent cleaning industry.
3. American Society for Testing and Materials (ASTM). Cold
Cleaning with Halogenated Solvents. Philadelphia, PA.
July 1966. p. 9.
4. Letter and attachments from Delta Sonics, to Beck, D. A.,
U. S. Environmental Protection Agency. February 1988.
Estimation of freon solvent usage in open top series
Delta sonics degreasers.
/
5. Trip report. Miller, S. J., Radian Corporation,
submitted to D. A. Beck, U. S. Environmental Protection
Agency. October 19, 1988. Summary of visit to Unique
Industries, Sun Valley, CA.
2-41
-------
6. Memorandum from Mead, R., and Pandullo, R., Radian
Corporation and Beck, D., U. s. Environmental Protection
Agency/Chemicals and Petroleum Branch to Degreasing
NESHAP File, U. S. Environmental Protection Agency.
September 8, 1987. Calculation of number of organic
solvent cleaners and solvent emissions and use per model
plant.
7. Memorandum from Gerald, L., and Falling, A., Radian
Corporation, to Almod6var, P., U. S. Environmental
Protection Agency. October 27, 1992. Compilation of
information obtained from solvent vapor and cold cleaner
vendor questionnaires.
8. Memorandum from O'Loughlin, Radian Corporation, to
Almod6var, P., U. S. Environmental Protection Agency.
June 3, 1993. National baseline emissions from
halogenated solvent cleaners.
2-42
-------
3.0 EMISSION CONTROL TECHNIQUES
3.1 INTRODUCTION
As discussed in detail in chapter 2.0, there are several
significant solvent loss sources from cleaners using
halogenated solvents. To achieve lower emissions during
solvent cleaning, owners or operators must consider minimizing
losses from each source. Good control can be achieved through
use of a cleaner with solvent-saving features and through
implementation of sound operating practices.
This section presents solvent control strategies covering
both machine design and operating practices. Table 3-1
presents a section outline and lists the solvent vapor
emission control techniques studied. Table 3-2 shows the
control efficiencies for select control techniques. The
control efficiencies were derived from emission tests
performed on idling and working batch vapor cleaners and
in-line cleaners. For both batch vapor and in-line cleaners,
there are separate sections devoted to diffusion/convection
controls, workload-related controls, and control of other
fugitive emission sources. Following these sections is a
discussion on which design elements and operating practices
should be incorporated to achieve a very-well-controlled
solvent cleaning operation. Finally, the section ends with
remarks about alternatives to solvent cleaning with the four
common halogenated solvents that will be regulated under
section 112 of the 1990 Clean Air Act as amended (1990
Amendments).
3-1
-------
TABLE 3-1. SUMMARY OF SOLVENT CLEANER CONTROL TECHNIQUES
Cleaner control technique Section
BATCH VAPOR CLEANERS
Interface Emission Controls 3.2.1
Covers 3.2.1.1
Freeboard refrigeration devices 3.2.1.2
Refrigerated primary condensers 3.2.1.3
Increased freeboard ratio 3.2.1.4
Reduced room draft/lip exhaust 3.2.1.5
Velocities
Carbon adsorption 3.2.1.6
Enclosed design 3.2.1.7
Workload Emission Controls 3.2.2
Mechanically assisted parts handling 3.2.2.1
Reduced parts movement speed 3.2.2.1
Carbon adsorption 3.2.1.7
Hot vapor recycle/superheated vapor 3.2.2.2
Proper Operating and Maintenance Practices 3.2.3
IN-LINE CLEANERS
Interface Emissions Controls 3.3.1
Minimize entrance/exit openings 3.3.1.1
Carbon adsorption 3.3.1.2
Freeboard refrigeration devices 3.3.1.3
Workload Emissions Controls 3.3.2
Carbon adsorption 3.3.1.2
Drying tunnels 3.3.2.1
Rotating baskets 3.3.2.2
Hot vapor recycle/superheated vapor 3.3.2.3
Proper Operating and Maintenance Practices 3.3.3
3-2
-------
TABLE 3-2.
SOLVENT VAPOR EMISSION CONTROL EFFICIENCIES
FOR SELECT CONTROL TECHNIQUES
Cleaner
Control technique
Control
efficiency (%)
Batch vapor
cleaners - OTVC's
«
In-line cleaners
Cover
Bi-parting cover
AFC
BFC
RPCTa
Lip exhaust
FBR 0.75 — >1.0
FBR 1.0— >1.25
Reduce wind speed
100 — >calm (<50 fpm)
Hoist
Dwell
AFC
BFC
CADS
Idling
40
40
40
40
40
40
20
10
50
NA
NA
60
60
60
Working
NA
40
40
40
40
45
20
10
50
35
30
60
60
60
a30 to 40 percent of the solvent boiling point.
OTVC = open-top vapor cleaners
AFC « above-freezing freeboard refrigeration device
BFC = below-freezing freeboard refrigeration device
RPCT = reduced primary condenser temperature
FBR » freeboard ratio
NA - not applicable
CADS - carbon adsorption system
3-3
-------
3.2 BATCH VAPOR CLEANERS
As discussed in chapter 2.0, batch vapor cleaners use a
heating system to boil liquid solvent, which creates a solvent
vapor zone for cleaning. The primary condenser contains the
vapor zone within the cleaner. Batch vapor cleaners range in
size from 0.2 to over 7.0 square meters (m2) [2.2 to over
75 square feet (ft2)]. The most common type of batch vapor
cleaner is the open-top vapor cleaner (OTVC). Other types of
batch vapor cleaners incorporate a more enclosed design, as
discussed in chapter 2.0.
A typical OTVC has a 0.75 to 1.0 freeboard ratio (FBR), a
water-cooled primary condenser, and a cover that is used
during downtime (see figure 2-1 in chapter 2.0). Some
machines have an external water jacket to cool the cleaner
walls.
Applicable control techniques for batch vapor cleaners
vary according to the size, design, application, and operation
of the cleaner. In general, the emissions reduction
efficiency of the various control options depends on the
fraction of time that the cleaner is idling versus the time it
is processing work because each control has different effects
on idling and working emission mechanisms.
The discussion of control techniques for batch vapor
cleaners is divided into two major areas. Section 3.2.1
presents information on interface emission controls—covers,
freeboard refrigeration devices, refrigerated primary
condensers, raised freeboards, reduced room drafts, enclosed
designs, and carbon adsorbers. Section 3.2.2 discusses
controls that limit workload solvent emissions, including
mechanically-assisted parts handling, reduced parts movement
speed, and hot vapor recycle/superheated vapor.
The data in the following sections are predominantly for
OTVC*s. Because non-OTVC cleaners are basically the same as
OTVC's, except for having an enclosed design (see
section 3.2.1.6), the discussions of OTVC controls would also
apply to other batch vapor cleaners.
3-4
-------
Summaries of emissions test data for OTVC's are presented
in tables 3-3 (idling machines) and 3-4 (working machines).
All idling tests are numbered using an "I" prefix. These
tests were performed by the EPA and by companies that either
manufacture solvent cleaning equipment or sell solvents. No
standard test methods were used. Each company established its
own test procedure.
The data and test procedures included in tables 3-3 and
3-4 were reviewed by the EPA and appear to have produced
valid, repeatable results. The data and test procedures not
included in the tables are documented in the "Summary of
Emission Reductions for Selected Control Techniques for
Organic Solvent Cleaners" memorandum.1 In some cases, the
company test facilities have been visited by personnel of EPA.
The OTVC test data in table 3-4, unless otherwise noted,
are from machines with automated mechanical systems for parts
handling. In most cases, workloads used for these tests can
be described as inherently producing low carry-out losses.
Therefore, emission rates would likely be higher from machines
in regular industrial applications. The following sections
discuss the inferences on control efficiencies that can be
drawn from these data.
3.2.1 Controls for Interface Emissions
3.2.1.1 Covers. Covers are used on OTVC cleaners to
eliminate drafts within the freeboard and to reduce diffusion
losses. Covers can be manually operated or electrically
powered. Some powered models are automated to work with the
cleaning cycle.
Manual covers are normally provided as standard
equipment. These covers are intended to reduce cleaner
emissions during idle time and periods of nonuse
(e.g., downtime). Manual covers should fit well and be
carefully operated to ensure that they do not become bent or
otherwise damaged. If a lip exhaust is used, the cover should
fit between the solvent vapor and the exhaust inlet.
3-5
-------
CO
1
8
K
2
$
0,
g
w
0,
o
o
H
Q
M
§
Srf
8
CO
CM
O
X
CO
.
1
u
i
»_
Control
•fficimcy
(X) Ri
II
• ^
55
&
ttm
~O 9t
2 8
1
k.
^S f
4i
gi
S j[
55
o
8
*£ S
S fi
u
u
41
• S S
S •— 5
u
I
5
- 1
*§
w .
85
»..>...
IQXJSSSSSS
ouuouuu*^
•-•^oooeoo
if iiii i i
o o o o o o
2.2.3SC58B
oeoooooo
oooooooo
f f ii iiii
*» **. **- <*• «*. «•»
O O O O O O
• ••••***
oooooooo
» 1 ^ «"• *•
5 o o g I 5 | ;
|S88333^
e
uuuuuuu*
<••••••&».
Ziiiii"*
S ?S b H
F* «* »n fn
3 3 S S
St <^ * in O>
d •- dodo
- - , i 1 I
C 1C a a n a
r» p*» o o O o
d d «- «- «- «-
gx «•• »^ «»• •*-
v 5 oooo
J S
• 2 2 *» - 2
U Ok K< »- « K(
fM fU « « e •>
K> s. S5 M S5 IM
d d d d d d
2 2 3 S 3 S
~ -D
22 1
A A M
i 1 I. t. U
" tC " C S S 5 "?
S*S« o o o*^
u. o u. o u u u a
e> o >» m >o N.
I 1 i I 1 t
3-6
-------
a
o
3
u
I
w
1
5
j
u
a
2
>
a.
o
i
z
u
a-
o
z
H
J
O
|.^
O
CO
w
H
E*
0
^
cn
1
CO
n
1
TABLE
«*-
w
et
X
!ls
^
s
i
^
i
Usion «t
|/*Z*hr)
2£
§
•t
•
u
*
c
u
„
^ 8.^
* w
v
7
•
If
en 'jj
•— ^5
Se
s
£
^
|
u
k"g?
« 8.^
M 8
j-~
_•;•-*>
U *
1
*£
S •*
•* ,
>- *
in
3
in
in
o
i
e
•" •
^
o
(0
s
o
1
*
o
^
0
o
10
s
o
-
•-
•g
'5
(O
rg
—
^*
o
in
i
«
3
a
1
k
sj
i
§
u
X
k
e
i.
gg 8
v w i
II l
'k *k O
" ^
k k
'5 "s *- c c
2 41 (A o .*»
rrlxf
|J5|o.
^ «t O.**> v
* " -^ i »
" " « M E
u u —
U. Ik UJ O •
< a -i S u
i
M
W
1
|
*_
o
o
g
u
£
U
u
JL .
f|
VI
i s
^ «•
II
a u
> S
•*• C
tl O
il
*- »*-
§'.
(I •
1*
* *^
«
u •
lg C
I/I K O
5e!:|
iii&
« «! i! «
1111
k k k k
Stl «l 41
M W «
ssss
k k k k
fff-f
3-7
-------
CO
fi
ri
u
A
1
04
|
04
0
u
S5
M
|
jg
O
CO
EH
CO
U
o
i
CO
1
i
5
».
m
u «•
£
1*
S*
8
•9
_
1
1
H
**
|^
ii
S
S S
^N
=li
• S
Il>
£
"81
:x
_s
8<^taMO'»m!9fMO>
-»3P>«»«»K»*>rIsi
t
SoooQinwtininin
in-
tC 1C 1C ICICIC S C
» • • • • • *• •
o o o o o o ee
i i i i i i i i ] i
7; o o> ~i e »
3 < >» 5 * -*
emi/taomtnein|n
mw^tom«^«o>^-
o o «- o o o «- oo»-
ICfttC (CtCIC SIC
eoo o o o eo
I I i i I I 1 i 1 I
| 5 5 * 5 S
*
^saaaaaaa^^
««§s» ss ssa
o o o o o e N o o
"-•-"slsill
0 00
o o o o o «- «-*-
iiiiiipiioo
!!!!!!S|3S'
o p in i/\ ip m oiAin
lA ^« ^9 (A ^^ *O O* ^5 C3
55555slsiil
i i ! i I ! i i i s i
i i i i.i 1 j ! ! s :
o e o e e o • e • •
o o o
w w
D A »• Ik
H^TTi'Sfi" "S ^
||i||¥|Si|l
g g 2 ft 3 5
'" * * * s i * s 4 -
3-8
-------
S"
z
H
W
H
$
VAPOR
O
§3
Cu
O
u
H
£
i
o
to
E-t
to
W
H
fe
0
X
(0
t
0
W
•J
H
^.
Control
•ion tfficttncy
?«hr) (X) Rt
ll
s
2 S
e S
|
if
I?
l|
M Ik
S S
• c
u
nl
u
*
1
5
?«
*J W
5 8
«5
S 3 3 R
s e s s
o e e e
5555
o e e e
1111
o o o o
U IU IU IU
O in o in
^ ^ ^* o>
5 S S S
o e o e
1111
5 8 8 |
IU IU IU u
e e o o
^22?
ib
u
o p fi o
S 1 1 e
a. ^ BL o
i st i i
S Jo ft R
S,BSSSSSSS
S*^*^^«^«f^^^ik
ooeoeoo«<
g*lC«t.kb<.ki.
o o o &&CL&&&
S^SS^oo^^g
uj *** ui ui w ^ ^r
SOpmoomtAOO
>»('1*iA*A'«*)*-lA«
o«-ooooo«*oo
<^.^^*»«»-«^«^*^<^'*.
oooooooooo
oooooeoooo
1111111111
•^A^^^SSS^S
0 ° 0 0 m<< tn
m ^ m |^
l52ST33Sa3S
»->->-u*-^»-^-^^
oooo>>>><<
minin^oeeoMw
o o o e > >
^^f???""
M w *> w e e
Sio£S£££|££
SSSSSS8S
$%&&&£§£
i i i i i a S S
SO.,,.00^0,0.
o«-ooo--«-o
oooooooo
00000000
i i i i i i i i
sassssj^
J
CC????oo
-------
a
§
0
2
o
O
i
2
1
OS
2
P*
g
1
/b
0*
o
19
2
M
CK
i
0
M
1
fa
O
>
1
(0
1
n
S
§
2
2
1
8
|
,.
it.
SMB 3
U **
~- **
u **•
I?
1-
Ul *
§
s
""
1
}l
±-
^
f!
g
I
u "S •£
•S
t
ift
u
1
a
_—
|h
•- o
^ *
**
!8
o
o
e
|
S
e
S
e
o
i
e
%
•
u
Ik
o
A
1C
e
s
IM
in
Sin M« in e
•« f\ « K
e e e e e
5*^ *^ *^ <^
O O O O
e o o XJ XJ
iiiii
m m o o u\
•o m •« o> ^>
o o e e e
S S o 5 5
1C 1C R o e
iiiii
o o o o o
"1- * 1
M *" *" S
2 5' 2 » «
A A A ^ •-
JC 1C 1C o i
s s s s s
po ^ in >O **
in tn in in m
31ft 19 ^9 CD
•* ^ K m
e e o o e
a « « « ic
iiiii
tA O W> IA tf\
O O O O *-
o o o o £
o o o e 1C
iiiii
•*
•
0 0 O 0 •-
* * ^c ^ \
i s g g S
» M
« fi n n J
».' ~ ^ ^ 8
» t t * e
o o o e 9
^
1 S S I I
« o> e ^ (v
m m 4
-------
Manual covers can be flat-hinged, sliding, or roll top.
Hinged covers are not recommended because opening and closing
these covers can disturb the vapor layer and unnecessarily
expose the operator to solvent vapors. If a flat-hinged cover
moves too quickly, it can cause turbulence that can disturb
the air/solvent vapor interface and increase emissions. Flat
covers that slide horizontally off the machine reduce
disturbance to the vapor layer. Figure 3-1 presents both a
manual and power-controlled roll-top cover. Roll-top models
are typically made of plastic (mylar). Flat covers (not
shown) may be made of mylar or metal.
Three idling tests were conducted to evaluate the effect
of a manual cover on solvent 1,1,1-trichloroethane (TCA)
emissions (table 3-3, tests 1-24, 1-25, and 1-26). In these
tests, a manually controlled flat cover that slides off
horizontally was used. These tests were conducted at three
different windspeeds (30.48, 15.24, and 7.62 m/min [100, 50
and 25 ft/min]). Figure 3-2 illustrates the test results. At
wind speeds of 30.48 m/min (100 ft/min), idling emission rates
without the cover were 0.95 kilograms per square meter per
hour (kg/m2*hr) [0.19 pounds per square feet per hour
(Ib/ft2*hr)]. With the addition of the manual cover, TCA
emission rates decreased to 0.40 kg/m2*hr (0.08 Ib/ft2*hr), a
58-percent emissions reduction. At windspeeds of 15.24 m/min
(50 ft/min), idling emission rates without a cover in place
were 0.60 kg/m2*hr (0.12 Ib/ft2*hr). Addition of the manual
cover reduced emission rates to 0.40 kg/m2*hr
(0.08 Ib/ft2*hr), a 33-percent emissions reduction. At
windspeeds of 7.62 m/min (25 ft/min), idling emission rates
without a cover in place were 0.55 kg/m2*hr (O.ll Ib/ft2*hr).
Addition of the manual cover reduced emission rates to
0.40 kg/m2*hr (0.08 Ib/ft2*hr), a 27-percent emissions
reduction.
As shown, the emissions reduction effect of covers
decreased when the OTVC was exposed to lower wind velocities.
3-11
-------
A. Boll Top Cover
(manual)
B. Si-Parting RoM-Top Cover
(power)
Figure 3-1. Typical open-top vapor cleaner covers,
3-12
-------
1.2 —
1.1 —
0.9 —
^0.8 -
i"-1
0.6 -
V)
•| 0.5 H
UJ
0.4
0.3 —
0.2 —
0.1 —
20
Windspeed (fpm) Ccalm* = 30 fpm)
40 60 80
I
I
I
100
I
LEGEND
CovwOn
— — m — — CowrOff
33%
27%
58%
— 0.24
— 0.22
— 0.2
— 0.18
— 0.16
c.
«
- 0.14
— 0.12 c
-o,!
m
— 0.08
— 0.06
— 0.04
— 0.02
12 18 24
Windspeed (mpm) Ocalm1 = 9 mpm)
30
Figure 3-2. Reduction in emission rate from the addition
of a cover.
3-13
-------
When the cover was in place, solvent emissions remained
constant despite external wind velocities.
Automated covers, which minimize disturbance of the
air/solvent vapor interface, include roll-top plastic (mylar)
covers (figure 4-1.B in chapter 4.0), canvas curtains, and
biparting (guillotine) covers that close horizontally.
Biparting covers can be made to close around the cables
holding the parts baskets when the basket is inside the
cleaner. This affords complete enclosure during the cleaning
phase. Powered biparting covers are usually operated by a
push-button control with an automatic shut-off and are either
pneumatically or electrically driven. The most advanced
biparting covers are automated to coordinate cover movement
with the movement of an automated parts handling system. In
this design the cover is opened only during parts entry and
exit. Powered biparting covers, which are closed during the
cleaning cycle, reduce both idling and working losses due to
diffusion by minimizing air drafts that disturb the
air/solvent vapor interface. On larger machines, it is
generally desirable to have powered (mechanically assisted) or
automated covers.
Four tests were available for an automatic cover that was
closed during most of the cleaning operation (table 3-4,
tests 38, 39, 40, and 41). In these tests, a biparting
roll-top cover that was closed 79 percent of the time
(275 seconds out of the 350-second OTVC cycle) was evaluated.
Without the automated cover, working emission rates varied
from 0.50 kg/m2*hr (0.10 Ib/ft2*hr) under calm (<9 m/min
[28.63 ft/min]) air conditions to 1.15 kg/m2*hr
(0.23 Ib/ft2*hr) with 48.77-m/min (160-ft/min) room drafts.
With an automated cover in use, working emission rates
decreased to between 0.30 kg/m2*hr (0.06 Ib/ft2*hr) (calm-)
and 0.55 kg/m2*hr (0.11 Ib/ft2*hr) (at 160 ft/min). This
corresponds to working loss reductions of 40 percent (calm) to
52 percent (at 160 ft/min). As expected, covers are more
effective at higher air draft velocity. The effects of
3-14
-------
reducing room drafts on emissions is discussed further in
section 3.2.1.5.
Based on available information,1 a cover is capable of
reducing emissions from a typical OTVC by 40 percent under
both working and idling conditions.
3.2.1.2 Freeboard Refrigeration Devicea. In all vapor
cleaners, solvent vapor created within the machine is
prevented from overflowing by using primary condenser coils.
Freeboard refrigeration devices include a second set of
cooling coils located above the primary condenser coils of the
cleaner. Functionally, the primary condenser coils define the
upper limit of the vapor zone. The freeboard refrigeration
coils chill the air immediately above the vapor zone, forming
a cool air blanket. The cool air blanket slows solvent
diffusion and creates a temperature inversion zone within the
freeboard, thereby reducing the mixing of air and solvent
vapors. Also, the cool air blanket supports lower solvent
concentrations than does warm air. Thus, some solvent at the
interface between the solvent vapor zone and cool air blanket
will condense into the cleaner.
Freeboard refrigeration devices have proven to be an
effective control for diffusional losses from an OTVC.
However, their effect is lessened if a refrigerated primary
condenser is present (see section 3.2.1.3). A drawing of an
OTVC equipped with a freeboard refrigeration device is
presented in figure 3-3.
There are two types of freeboard refrigeration devices:
above-freezing chillers (AFC's) and below-freezing chillers
(BFC's). Above-freezing chillers operate at a temperature
range around 5 °C (41 °F). Below-freezing chillers operate
with refrigerant temperatures usually in the range of -20 °C
to -30 °C (-4 °F to -22 °F).
Because of the low operating temperatures of BFC's,
provisions are made for a timed defrost cycle to melt
solvent/water ice that may form on the coils. If allowed to
accumulate on the refrigerant coils, this ice layer would
3-15
-------
Primary
Condensing Coils
Temperature
Indicator
Cleanout Door
Freeboard
Refrigeration
Device
Solvent Level Sight Glass
Heating Elements
Work Rest and Protective Grate
Freeboard
Condensate Trough
Water
Separator
Figure 3-3. Open-top vapor cleaner with freeboard
refrigeration device.
3-16
-------
compromise heat transfer efficiency. The solvent/water
mixture melted from the freeboard coils during the defrost
cycle drains into a trough located below the freeboard
refrigerator coils. To minimize water contamination of the
solvent, the melted solvent/water mixture should be directed
to a second water separator (distinct from the separator
employed for the condensate from the primary condensing coils)
for removal.
Above-freezing chillers condense water from the air. The
condensed water can strip the stabilizers that are present in
many solvent mixtures. A cleaner equipped with such a device
may also benefit from a second water separator.
Theoretically, a BFC should be more efficient than an AFC
because it can achieve lower freeboard temperatures. Lower
freeboard temperatures establish a cooler, more stable
inversion layer, which lowers diffusion rates. However, the
need to periodically defrost a BFC can somewhat offset its
performance advantage.
Twenty-two tests from four sources were available to
evaluate the effect of freeboard refrigeration devices on
OTVC's under working (16 tests) and idling (7 tests)
conditions. Four tests evaluated AFC's and the remainder
evaluated BFC's. All of the tests under idling conditions
evaluated BFC's. Figures 3-4 and 3-5 summarize the data for
idling and working conditions, respectively. Test numbers
refer to the tests listed in tables 3-3 and 3-4.
Uncontrolled and controlled emission rates vary
considerably under the "working" scenario, much more so than
under idling conditions (see table 3-3; especially notice the
low variability in the "controlled" data). This is to be ,
expected because of the major impact on emissions from the
workload. Tests performed by different companies reflect
differing workload sizes, shapes, and cleaning cycle
frequencies. A higher emission rate does not necessarily mean
that a cleaner was not as well controlled, but is likely to
3-17
-------
1.75 —
1.5 -H
0-
i
cc
o
'55
en
e
0.75 -|
0.5 -J
0.25 H
Uncontrolled
Controlled
P«re«ntag«« r*pr*Mnt
th« percent raduction in
•mitsion* resulting from
control
0.4
— 0.35
- 0.3
— 0.25 ^
,
- 0.2 g
— 0.15
.52
III
— 0.1
— 0.05
#2 #3 *4 #9 #10 #11 #12 #13 #14 #15 #16 #17 #19 #20
Test Number (1 -4 AFC; 9-20 BFC)
Figure 3-4. Freeboard refrigeration device tests
working conditions.
3-18
-------
0.9 —
0.8 —
0.7 —
f 0.6 -\
CM
0>
® 0.5 —
re
DC
O
'55
CO
'§ 0.4 —
HI
0.3 —
0.2 —
0.1 —
Uncontrolled
Controlled
Percentages represent
the percent reduction in
emissions resulting from
control
1-1
1-2
\-7
K3 M 1-5
Test Number
(all tests, except 1-1, are on below-freezing chillers)
Figure 3-5.
Freeboard refrigeration device tests
idling conditions.
3-19
-------
reflect the influence of a more demanding workload schedule or
cleaning of a workload that is more prone to carry-out losses.
For working conditions, control efficiencies ranged from
20 to 48 percent for AFC's (tests 1 through 4). Three of the
four AFC tests showed at least a 33-percent emissions
reduction. Under working conditions (tests 9 through 20,
excluding 18), control efficiencies for BFC's ranged from
27 to 83 percent. The observed 83-percent reduction (test 12)
should be considered atypical.
Freeboard refrigeration devices primarily reduce
diffusional losses. In a working OTVC, losses from solvent
carry-out on parts are significant and are usually greater
than diffusional losses, except where the machine is in a very
drafty location. Therefore, in the more likely situation
where workload losses are significant or dominate, it would be
impossible to achieve an 83-percent emissions reduction from a
device designed to control diffusional losses. Controlled
working emissions rates for AFC's ranged from 0.20 kg/m2*hr
(0.04 Ib/ft2*hr) to 0.6 kg/m2*hr (0.12 Ib/ft2*hr). For BFC's,
controlled emission rates ranged from 0.05 kg/m2*hr
(0.01 Ib/ft2*hr) to 1.25 kg/m2*hr (0.25 Ib/ft2*hr).
Efficiencies for BFC's under idling conditions (tests 1-2
to 1-8, excluding 1-6) ranged from 17 to 59 percent. The
tests were conducted on machines using trichlorotrifluoro-
ethane (CFC-113) and TCA. Most notable to this series of
tests is that the primary condensing temperature affects BFC
effectiveness for CFC-113. As primary condensing temperature
decreases, the additional benefit of a BFC also decreases.
This effect is not nearly as pronounced with TCA. Primary
condensation temperature is discussed further in the next
section. Controlled idling emission rates with the use of a
BFC ranged from 0.20 kg/m2*hr (0.04 Ib/ft2*hr) to
0.35 kg/m2*hr (0.07 Ib/ft2*hr).
It is important that the freeboard refrigeration device
be able to achieve a significant temperature inversion within
the freeboard area (i.e., a temperature less than room
3-20
-------
temperature). Poorly designed freeboard refrigeration devices
may not be able to establish the cooler temperatures at the
center of the freeboard zone.
Based on available information, a freeboard refrigeration
device is capable of reducing emissions from a typical OTVC by
40 percent.10
3.2.1.3 Refrigerated Primary Condenser. Although a
primary condenser is standard equipment on all batch cleaners,
the temperature at which cooling is provided and the design of
the coils and coolant flow have an effect on idling losses.
Heat removal to balance the vapor-generating heat input can be
provided at various temperatures with water, chilled water, or
a direct expansion refrigerant. A lower-temperature primary
condenser, generally using a refrigerant as opposed to water,
will lower diffusion losses. The likely reason for this
effect is that colder primary condenser temperatures, besides
condensing solvent vapor, act to cool the air above the
air/solvent vapor interface, somewhat like a freeboard
refrigeration device. The magnitude of the reduction in
diffusion rate varies by solvent.
The relationship between emission rate and primary
condenser temperature under idling and working conditions for
TCA and CFC-113 are presented in figure 3-6 (table 3-4;
tests 33 through 37). A steeper slope represents a greater
sensitivity to primary condenser temperature. Uncontrolled
working emissions for TCA ranged from 0.80 kg/m2*hr
(0.16 Ib/ft2*hr) at 29 °C (85 °F) to 0.5 kg/m2*hr
(0.10 Ib/ft2*hr) at 10 °C (50 °F).. Thus, a 38-percent
reduction in working emissions of TCA can be obtained by
reducing the primary condenser temperature from 29 °C (85 °F)
to 10 °C (50 OF).
For CFC-113, uncontrolled working emissions ranged from
0.45 kg/m2*hr (0.19 Ib/ft2*hr) at 21 °C (70 °F) to '
0.95 kg/m2*hr (0.09 Ib/ft2*hr) at 4 °C (40 °F). In the case
of CFC-113, lowering the primary condenser temperature from
21 °C (70 °F) to 4 °C (40 °F) yields a 53-percent working
3-21
-------
Primary Condenser Temperature (°F)
30 35 40
.1.1
0.9 —
0.8 —
0.7 —
* 0.6 —
I
2 0.5 —
o
•55
~ 0.4 —|
111
0.3 —
0.2 -
0.1 —
T
2
I
4
45
I
50
I
55
I
60
I
65
I
70
I
75
I
a''
T
I
I
I ' I
I
80
I
85
I
90
CFG (working)
CFC (idto) TCA (working)
7 10 13 16 18 21 24
Primary Condenser Temperature (°C)
I
27
0.2
- 0.18
- 0.16
- 0.14
— 0.12 f
- 0.1 g
o
"35
— 0.08 -|
LU
— 0.08
— 0.04
- 0.02
I
29
32
Figure 3-6. Effect of primary condenser temperature on
uncontrolled idling and working conditions.
3-22
-------
emissions reduction. It should be noted that working
conditions for these tests were simulated by introducing a
water-cooled load using a programmable hoist. The load was
cycled 12 times every hour. This type of setup would be
expected to simulate relatively mild working conditions.
Reducing the primary condenser temperature has a similar
effect on emissions under idling conditions as under working
conditions. Uncontrolled emissions for TCA ranged from
0.70 kg/m2*hr (0.14 Ib/ft2*hr) at 29 °C (85 °F) to
0.45 kg/m2*hr (0.09 Ib/ft2*hr ) at 10 °C (50 °F). This
corresponds to a control efficiency of 36 percent when the
primary condenser temperature is decreased from 29 °C (85 °F)
to 10 °C (50 °F). For CFC-113, uncontrolled idling emissions
were reduced from 0.85 kg/m2*hr (0.17 Ib/ft2*hr) at 21 °C
(70 °F) to 0.30 kg/ro2*hr (0.06 Ib/ft2*hr) at 4 °C (40 °F), or
a control efficiency of 65 percent under idling conditions.
It is unlikely that all solvent vapor cleaners using TCA
and CFC-113 will operate their primary condensers at 29 °C
(85 °F) and 21 °C (70 °F), respectively* In fact, for CFC-113
machines, primary condensation usually is accomplished through
direct expansion refrigeration or chilled water systems
operating at 4 °C to 16 °C (40 °F to 60 °F). However, even if
primary condenser temperatures for TCA and CFC-113 are at
21 °C (70 °F) and 10 °C (50 °F), respectively, additional
diffusion reduction can still be obtained. The tests shown in
figure 3-6 indicate that lowering the primary condenser
temperature for TCA from 21 °C (70 °F) to 10 °C (50 °F)
reduces working emissions from 0.60 kg/m2*hr (0.12 Ib/ft2*hr)
to 0.45 kg/m2*hr (0.09 Ib/ft2*hr); this corresponds to a
29-percent reduction. Similarly, reducing the primary
condenser temperature on a CFC-113 machine from 10 °C (50 °F)
to 4 °C (40 °F) will reduce working emissions from
0.45 kg/m2*hr (0.09 Ib/ft2*hr) to 0.30 kg/m2*hr
(0.06 Ib/ft2*hr), an 18-percent reduction.
These tests also examined the effect of adding a BFC to
an OTVC operating with a refrigerated primary condenser. For
3-23
-------
cleaners using TCA, the addition of a freeboard refrigeration
device to a cleaner with a primary condenser at 10 °C (50 °F)
still had a significant effect on emissions, reducing
emissions by more than 50 percent. Very little reduction was
obtained by adding a freeboard refrigeration device to a
CFC-113 machine operating at a primary condenser temperature
of 4 °C (40 °F).
One drawback to lowering primary condenser temperature is
that it promotes condensation of ambient water vapor,
especially in humid climates. Therefore, it is imperative
that machines employing low-temperature condensation contain
adequately sized water separators or' desiccant dryers to
minimize water contamination.
The test results on primary condenser temperatures
suggest another area of concern for water-cooled batch vapor
cleaners. Machines using tap water, cooling tower water, or
well water will be subject to seasonal temperature variations.
During summer months, condenser water temperatures may rise
significantly and may increase undesirable diffusion loss.
This effect may be exacerbated by increased ambient drafts
from open doors and windows in warm weather. Use of chilling
or refrigerant systems to control condensing temperatures will
minimize seasonal variations.
Based on available information, a refrigerated primary
condenser is capable of reducing emissions from a typical OTVC
by 40 percent, if the temperature of the primary coil is 30 to
40 percent of the solvent boiling point.1
3.2.1.4 Increased Freeboard Ratio. The freeboard height
on a vapor cleaner is the distance from the air/solvent vapor
interface to the top of the tank walls. The freeboard zone
serves to reduce air/solvent vapor interface disturbances
caused by room drafts and provides a column through which
diffusing solvent molecules must migrate before escaping into
the ambient air. Higher freeboards reduce diffusion losses by
diminishing the effects of air currents and lengthening the
3-24
-------
diffusion column. An OTVC with an increased freeboard is
presented in figure 3-7.
In discussing the ability of freeboard height to reduce
solvent loss, it is common to refer to the freeboard ratio
(FBR). The FBR is the freeboard height divided by the
interior width of the solvent cleaner. The freeboard height
should be measured from the established air/solvent vapor
interface to the top of cleaner walls or to the bottom of any
opening in the cleaner walls. Freeboard width is the inside
width of cleaner walls or, if irregular, the smallest width
dimension of the air/solvent vapor interface directly exposed
to the atmosphere. The FBR is used in recognition of the fact
that as cleaner width increases, susceptibility to the adverse
influence of drafts increases unless the freeboard height is
proportionally increased to compensate for increased machine
width. Two cleaners of differing size (width) but with
identical FBR's are equally protected from drafts.
A high freeboard on some machines may make it difficult
for an operator to lower parts into the machine unless an
elevated work platform is installed. However, as discussed in
section 3.2.2.1, a hoist can be used on large machines to
overcome the problem of machine height and reduce
workload-related losses. On very large machines, raised
freeboards may be so tall as to restrict the ability to place
parts in the machine. For these situations, slightly lower
freeboards might be necessary, but special care should be
taken to minimize room drafts.
For small OTVC's, the absolute freeboard height is an
important factor in solvent loss due to diffusion. Despite
having a high FBR, very small machines may not have sufficient
total freeboard height to prevent accelerated diffusion
losses, even in calm environments. Industry tests show that
solvent loss rates can increase substantially with absolute
freeboard heights of less than approximately 30 cm (12 in.).9
An example of how emission rates can vary as a function of
freeboard height is presented in figure 3-8.
3-25
-------
Primary
Condensing Coils
Temperature
Indicator
Cleanout Door
Solvent Level Sight Glass
Heating Elements
Work Rest and Protective Grate
Increased
Freeboard
Condensate Trough
Water
Separator
Figure 3-7. Open-top vapor cleaner with increased freeboard.
3-26
-------
zU/epAo/6 uj emy ssoi
in
in
I— eg
in
N.
d
00
I
n
0)
h
in
o
$
ci
S
in
5
o
S
o
ui ejsy no-)
3-27
-------
Fourteen tests from two sources were available to
evaluate the effect of an increased FBR on solvent emissions.
Twelve of the tests evaluated this effect under working
conditions and two tests evaluated the effect under idling
conditions. Emissions reductions were evaluated for raising
the FBR from 0.75 to 1.0, and for raising the FBR from 1.0 to
1.25. As mentioned previously, a 0.75 FBR is representative
of baseline conditions. Although some older machines may have
0.5 FBR's, most vendors currently sell solvent cleaners with
FBR's of at least 0.75.
The available data on the effect of an increased FBR are
presented in table 3-3 (tests 1-19 and 1-20) and table 3-4
(tests 50 through 61) for idling and working conditions,
respectively. The data for working conditions are presented
graphically in figure 3-9.
Under working conditions, the control efficiencies
associated with raising the FBR from 0.75 to 1.0 ranged from
12 to 24 percent. Controlled emission rates at an FBR of
1.0 ranged from 0.35 kg/m2*hr (0.07 Ib/ft2*hr) to
0.90 kg/m2*hr (0.18 Ib/ft2*hr). The control efficiencies
associated with raising the FBR from 1.0 to 1.25 ranged from
6 to 14 percent. Controlled emission rates at an FBR of 1.25
ranged from 0.30 kg/ro2*hr (0.06 Ib/ft2*hr) to 0.85 kg/m2*hr
(0.17 Ib/ft2*hr). Using the above data, the efficiency
improvement associated with raising the FBR from 0.75 to 1.25
*
were calculated to be approximately 25 percent.
Under idling conditions, data are available to evaluate
the effect of raising the FBR from 0.75 to 1.0, but no data
are available for estimating the efficiency improvement of
increasing the FBR to 1.25. Under idling conditions, the
control efficiencies associated with raising the FBR from 0.75
to 1.0 were 0 and 37 percent, based on two tests. The test
with a 0-percent efficiency was conducted under calnC air
conditions. Therefore, the expected reduction in emissions
would be lower than in tests conducted under higher air speed
conditions. However, the 0-percent efficiency result is
3-28
-------
1.3
1.2 —
1.1 —
1 —
0.9 —
0.8 —
CD
15
OC
0.6
CO
J2
E 0.5
UJ
0.4 —
0.3 -
0.2 —
0.1 —
TM»SO*M-MC
TM»S1&57-MC
TMfc52fc5»CFC.113
TMVS3&50-MC
TwW 54 A eO-TCA
''A.
0.6
I 'I ' I
0.8 1 1.2
Freeboard Ratio
I
1.4
0.26
- 0.24
- 0.22
- 0.2
- 0.18
- 0.16 £
«
- 0.14 £,
e
- 0.12
- 0.1
UJ
- 0.08
- 0.06
- 0.04
- 0.02
Figure 3-9,
Effect of freeboard ratio - working conditions:
six open-top vapor cleaner tests.
n*M«*t.bid:9/24/B3
bjw/039
3-29
-------
likely attributable to measurement inaccuracies. In fact,
measured losses for uncontrolled and controlled scenarios were
very small and could be considered the same, within
experimental precision.
Another strategy related to raising the freeboard for
emission control is to design narrower cleaners. For the same
air/solvent vapor interface area, a square interface
configuration is more susceptible to room drafts than is a
long, narrow rectangular configuration, especially if the
cleaner is oriented in the room so that any drafts blow across
the narrower dimension.
Based on available information, increasing the FBR from
0.75 to 1.0 can reduce emissions from a typical solvent
cleaner by 20 percent. Increasing the FBR from 1.0 to 1.25
can reduce emissions by approximately 10 percent.1
3.2.1.5 Reduction in Room Draft/Lip Exhaust Velocities.
Air movement over a solvent vapor cleaner affects the solvent
emission rate by sweeping away solvent vapors diffused into
the freeboard area. This creates turbulence in the freeboard
area, which will enhance solvent diffusion as well as solvent
air vapor and air mixing.
In industrial manufacturing settings, solvent cleaners
often operate in high-draft areas, with drafts typically in
excess of 39 m/min (130 ft/min).11 Reducing room drafts to
calm conditions (9 m/min [30 ft/min] or less) can greatly
reduce emission rates. The available data for evaluating the
effect of reduced room draft velocity were measured under
working conditions (see figure 3-10). These data are from
tests showing the effects of draft velocity on emissions at a
constant 0.75 FBR (table 3-4, tests 62 and 63). The emission
rates from the tests are 1.15 kg/m2*hr (0.23 Ib/ft2*hr) at
48 m/min (160 ft/min), 0.85 kg/m2*hr (0.17 Ib/ft2*hr) at
39 m/min (130 ft/min), and 0.50 kg/m2*hr (0.1 Ib/ft2*hr) at
calm conditions. Under working conditions, reducing 39-m/min
(130-ft/min) room drafts to 9 m/min (30 ft/min) results in a
41-percent reduction in emissions and reducing 48-m/min
3-30
-------
1.3
Windspeed (fpm) ('calm" = 30 fpm)
20 40 60 80 100 120 140 160 180
I I I I I I I I I I I I I I I I I
1.2 -1
1.1 -
1 —I
0.9 -J
J= 0.8
*
E
'0.7 —
I
tr
0.6 —
(0
to
I 0.5
UJ
0.4 -
0.3 -
0.2 -
0.1 —
1 ' I ' I ' I . ' I ' I ' I
12 18 24 30 36 42 48
Windspeed (mpm) ("calm' = 9 mpm)
0.26
— 0.24
- 0.22
- 0.2
— 0.18
— 0.16
— 0.14
— 0.12
s
- 0.1
UJ
— 0.08
— 0.06
— 0.04
h- 0.02
54
Dogreasar Size: 0.16^(1.8ft2}
FBR: 0.75
1* Condenser T»mpwatur«: &t40*F
Cover: Off
No other controls
A DegreaserSize: 0.28^(3.1ft2)
FBR: 1.0
1 • Condenser Temperature: «40°F
Cover: Off
No other controls
Figure 3-10. Effect of wind speed.
3-31
-------
(160-ft/min) room drafts to 9 m/min (30 ft/min) results in a
51-percent reduction in emissions.
Also illustrated in figure 3-10 are idling test data from
tests conducted by the EPA's Emission Measurement Branch (EMB)
to evaluate the effect of windspeed on solvent (TCA) emission
rates (table 3-3; tests 1-27, 1-28, and 1-29). At wind
velocities of 30 m/min (100 ft/min), the estimated emission
rate was 0.95 kg/m2*hr (0.19 Ib/ft2*hr). When wind velocity
was reduced to 15 m/min (50 ft/min), the TCA emission rate
decreased to 0.60 kg/m2*hr (0.12 Ib/ft2*hr), a 37-percent
emission reduction. Further reduction of the wind velocity
from 30 m/min (100 ft/min) to 8 m/min (25 ft/min) reduced
emissions from 0.95 kg/m2*hr (0.19 Ib/ft2*hr) to 0.55 kg/m2*hr
(0.11 Ib/ft2*hr), a 42-percent emission reduction.
A lip exhaust, described in chapter 2.0, affects
emissions in much the same way as does air speed: it
increases mixing and diffusion in the vapor layer. Tests have
shown that a lip exhaust, even when properly operated, can
double.solvent consumption.12 If the solvent is not recovered
through the use of a carbon adsorber, overall emissions will
increase.
Tests were conducted on the effect of turning off a lip
exhaust under both idling and working conditions (table 3-3,
tests 1-16, 1-17, and 1-18 and table 3-4, tests 30, 31,
and 32, respectively). The lip exhaust was operated at the
rate of 2.5 cubic meters per square meter per minute
(m3/m2*min) [90 cubic feet per square foot per minute
(ft3/ft2*min)] of cleaner area. This corresponds to 25 m3/min
(900 ft3/min) for this particular test. Based on test data
for working conditions, the emission rates associated with a
lip exhaust system in operation ranged from 0.85 kg/m2*hr
(0.17 Ib/ft2*hr) (with primary condenser temperature of 10 °C
[50 °F]) to 1.05 kg/m2*hr (0.21 Ib/ft2*hr) (with primary
condenser temperature of 29 °C [85 °F]). With the lip exhaust
turned off, the emission rates decreased to 0.45 kg/m2*hr
(0.09 Ib/ft2*hr) (at 10 °C [50 °F]) and 0.70 kg/m2*hr
*
3-32
-------
(0.14 Ib/ft2*nr) (at 29 °C [85 °F]). This corresponds to a
reduction in solvent loss ranging from 47 percent (at 10 °c
[50 °F]) to 33 percent (at 29 °C [85 °F]). The data are
presented graphically in figures 3-11 and 3-12 for idling and
working conditions, respectively.
3.2.1.6 Carbon Adsorption. Carbon adsorption can be
employed as a control technique in conjunction with a lip
exhaust system. Lip exhaust/carbon adsorption systems are
most commonly used on large solvent cleaners where the credit
from solvent recovery helps to offset the high capital
equipment cost. With these systems, peripheral exhaust ducts
capture the diffusing solvent vapors, and to some extent
solvent evaporating from clean parts, and directs them through
an activated carbon bed. The solvent vapor molecules are
adsorbed onto the activated carbon, removing the solvent from
the vent stream before the stream is discharged to the
atmosphere.
At intervals, when the carbon becomes saturated with
solvent, the bed is desorbed. Usually,"desorption is
accomplished by passing steam through the bed to remove the
solvent from the carbon. The solvent/steam mixture is then
condensed and passed through a water separator and the
recovered solvent is returned to the cleaner.
The lip exhaust ventilation system should be designed to
maximize solvent capture efficiency and minimize disturbance
of the air/solvent vapor interface. The percentage of vapor
emissions that are captured by the lip exhaust system is
uncertain. Several vendors have indicated a lip exhaust
capture efficiency of 40 to 99 percent,13-15 but no test data
were provided for justification.
Proper operation and maintenance procedures are critical
to maintaining the control efficiency of carbon adsorbers.
Examples of operating procedures that have a negative impact
on control efficiency include: (1) dampers that do not open
and close properly, allowing solvent-laden air to bypass the
3-33
-------
With Up Exhaust
No Up Exhaust
Percentages represent
•auction in
the percent re
•missions resulting from
turning off a Up exhaust
0.2
H6PC@1CTC
- 0.15
M7PC@21*C
Test Number
M8PC@29*C
- 0.1
UJ
- 0.05
Figure 3-11. Lip exhaust effects - idling conditions.
3-34
-------
O29
1.4 —
1.3 —
Up Exhaust On
Up Exhaust Off
PwcantaQos rwysasrit
th* psfcant reduction In
from
30(PC@1CTC) 31(PC©21*Q 32 (PC @ 29^
Test Number (these are working losses)
Figure 3-12. Lip exhaust effects - working conditions.
3-35
-------
carbon beds; (2) use of carbon that does not meet
specifications; and (3) improper timing of the desorption
cycles. Desorption cycles must be frequent enough to prevent
breakthrough of the carbon beds, but not so frequent as to
cause excessive energy consumption. Carbon adsorbers should
not be bypassed during the desorption process. A dual-bed
design can be used so that while one bed is being desorbed,
solvent emissions can be routed to the second bed.
One test was available to evaluate the efficiency of
carbon adsorbers for controlling solvent emissions.4 This
test indicated that a lip exhaust/carbon adsorber system could
control solvent emissions by 65 percent. However, the test
report did not specify whether the baseline emission rate
included lip exhaust. If the baseline OTVC did have a lip
exhaust, the 65-percent emission reduction overstates the
achievable reduction for a carbon adsorber and lip exhaust
installed on a solvent cleaner without a lip exhaust. Thus,
there is some uncertainty about the validity of this data
point.
Another source16 indicated that the overall effect of
installing a lip exhaust/carbon adsorber system on a solvent
cleaner would be a 40-percent reduction in total emissions.
Because of the emissions increase associated with adding a lip
exhaust, the overall control effectiveness of using carbon
adsorption for solvent cleaners is likely closer to 40 percent
than 65 percent.
Depending upon the solvent mixture and the type of
objects being cleaned, adverse effects may be encountered with
carbon adsorption. Where solvent mixtures or stabilizers are
used, the solvent vapor collected by the exhaust system may be
richer in the more volatile' components, and the recovered
solvent mixture will not be identical to fresh solvent. Also,
some stabilizers or co-solvents used in solvent mixtures are
water-soluble. After desorption, the steam used to disturb
the solvent and stabilizers from the carbon bed is condensed.
The water-soluble components remain in the water and are lost
3-36
-------
unless recovered by distillation. Many users are not willing
or able to undertake tasks such as analyzing and reformulating
the solvent or handling toxic or flammable stabilizers.
In addition, by-products of uncontrolled solvent
degradation, such as hydrochloric acid, can be corrosive to
the adsorption equipment and/or hazardous to operators. For
some solvents or cleaning applications, it may be necessary to
use special construction materials for the adsorber (e.g.,
stainless steel or other alloys), or to take other measures to
prevent potential problems that could lead to solvent
degradation and damage to the equipment. One solvent in
particular, TCA, is troublesome when used in carbon
adsorption. It is heavily stabilized and many of the
stabilizers may be removed during carbon adsorption, causing
solvent breakdown and equipment corrosion. Carbon adsorption
probably should not be attempted with this solvent at this
time. However, recent studies indicate that carbon adsorption
systems for use with TCA will be available in the future.17
3.2.1.7 Enclosed Design. The enclosed design as a
control option for OTVC's involves completely enclosing the
cleaner except for a single opening through which parts enter
and leave the enclosure. Use of an enclosure typically
precludes manual parts handling. The non-OTVC batch units
tend to be enclosed design cleaners.
Enclosed design batch cleaners reduce idling and working
losses by creating a still air environment inside the machine,
which limits solvent diffusion. Additionally, automated
loading and unloading of parts at a controlled rate creates
less air turbulence and reduces solvent carry-out oh cleaned
parts.
Schematics of two variations of enclosed designs are
shown in figure 3-13. The enclosed design with a horizontal
entry/exit port (figure 3-13.A) is not affected by room air
drafts and does not require a port cover during machine
operation. The enclosed design with a vertical entry/exit
newest.bid:11/2/93
bjw/039 3-37
-------
3i
M
1 ^
r
i-H
>
CO
T
If*
.H
I
w
LI
0)
1C
0)
r-l
u
o
a
to
A
U
-P
(TJ
•O
01
O
w
n
H
I
n
0)
J-i
3-38
-------
port should have a sliding door that is closed except when
parts are being loaded or unloaded.
3.2.1.8 Vacuum Chamber Cleaners. Vacuum chamber
cleaners are a new design of batch cleaners that clean items
in a sealed decompressed chamber. In this process, the
part(s) to be cleaned is placed in the chamber and the chamber
is then decompressed by a vacuum pump. Warm solvent is pumped
into the chamber and ultrasonic cleaning begins. After a
period of time, the warm solvent is removed from the chamber,
filtered, and placed in the warm solvent tank. Cold solvent
is then pumped into the chamber and ultrasonic cleaning begins
again. The cold solvent is then removed from the chamber,
filtered, and placed in the cold solvent tank. Solvent vapor
is then pumped into the chamber as the final cleaning step.
Before removal from the chamber, the part is dried under
vacuum and residual solvent vapors are recovered by a carbon
adsorption system. The chamber is then returned to normal
pressure and the part can be removed.18
These cleaners reportedly have extremely low emissions.
One manufacturer reports that its enclosed parts cleaner has
demonstrated emissions of less than 50 parts per million (ppm)
per cleaning cycle.
3.2.2 Controls for Working Emissions
3.2.2.1 Mechanically-Assisted Parts Handling/Parts
Movement Speed. The method employed for moving parts through
the batch cleaner cleaning cycle has a direct effect on the
magnitude of working emissions. Rapid movement of parts will
increase solvent loss due to liquid solvent carry-out, solvent
vapor entrainment, and increased disturbance at the
solvent/air interface. As mentioned in chapter 2.0, losses
due to parts handling can be a large portion of total working
losses (see chapter 2.0 for additional discussion of workload
losses). '
Parts can be moved through the cleaning cycle either by a
human operator or by a mechanical system. A human operator is
3-39
-------
generally unable to move parts at or below the maximum speed
of 3 m/min (11 ft/min), as required in many State regulations
and recommended in the EPA guidelines.11/19'20 According to
one vendor, it is difficult to maintain a constant speed if a
full basket weighs around 4 kg (10 Ib) or more (baskets can
weigh in excess of 19 kg [50 lb]).17'20 Operator training may
have limited success in reducing the basket movement rate. In
addition, the speed of the basket is difficult to judge, and
operators will typically return to faster rates, especially if
the load is heavy enough to cause fatigue toward the end of
the workday.^9 In some industries, operators are paid on a
per-piece basis. This may be further incentive to move parts
more quickly.21
Industry estimates of typical parts movement by human
operators are in excess of 18 m/min (60 ft/min).19'20 At
these speeds, working losses would be much higher, perhaps by
several times, than the data presented in section 3.2 (which
reflects use of hoists). Use of a mechanical parts handling
system can reduce emissions by consistently moving parts into
and out of the machine at appropriate rates. A parts handling
system can be operated by push button or it can be automatic
and programmable. Two typical parts handling systems are
shown in figure 3-14. The first is a sin le-axis
hoist that can be operated by a push butten; the second is a
double-axis programmable parts handling system.
Although the emissions reduction benefit of using
mechanically assisted parts handling is generally not
disputed, there are few data available to characterize the
magnitude of the benefit.
Three tests are available that simulate the effects of
switching from a human operator to an automated system
(table 3-4, tests 27, 28, and 29). Test 27 compared a hoist
operated at 6 m/min (20 ft/min) (to simulate a human operator)
to a hoist operated at 3 m/min (10 ft/min). The lower speed
was found to reduce working losses by 28 percent. Because
human operator speeds are generally higher than 6 m/min
3-40
-------
o2
«1
^i:
X
CD
I
o
5
g
o>
S
O)
'35
_<«
'x
<
5
JD
00
L
7.
3*
i
4J
tQ
>i
10
H1
•o
(0
£
n
-P
M
c/5
3-41
-------
(20 ft/min), the reduction attributable to the use of a hoist
is likely larger than 28 percent. Tests 28 and 29 are actual
measurements of the solvent (TCA) emissions reduction effect
of switching from manual operation to an automated hoist. The
automated hoist speeds for these tests were set between 2 to
5 m/min (7.5 to 15 ft/min). The human operator speed was
variable, as in actual working conditions.
Tests 28 and 29 differed in the way in which the hoist
was tested. Test 28 did not incorporate a dwell (a 45-second
pause above the vapor zone); test 29 did incorporate a dwell.
The solvent .(TCA) emission rate for human operator working
conditions was 1.05 kg/m2*hr (0.21 Ib/ft2*hr). The TCA
emission rate using an automated hoist without a dwell
(test 28) was 0.65 kg/m2*hr (0.13 Ib/ft2*hr), a 38-percent
emission reduction. The TCA emission rate using an automated
hoist with a dwell (test 29) decreased the emission rate to
0.55 kg/m2*hr (0.11 Ib/ft2*hr), a 48-percent emission
reduction. The effect of the dwell is discussed further in
the following text.
A reduced hoist speed would allow parts to dry more
thoroughly prior to removal in addition to creating less air
turbulence during parts entry and exit from the cleaner.
Therefore, working losses due to both solvent carry-out and
diffusion would be minimized. One manufacturer has evaluated
the effectiveness of further reducing hoist speed,
particularly as the parts basket moves through the solvent
vapor layer (table 3-4, test 26). During the test, a variable
speed, programmable hoist was used to lower the hoist speed to
1 m/min (3 ft/min) as the parts basket moved through the
solvent vapor zone. Decreasing the hoist speed from 3 to
1 m/min (11 to 3 ft/min), resulted in an 81-percent decrease
in total working losses.
Another advantage of mechanical parts handling 'is the
potential for precise control of dwell time (i.e., the length
of time the part remains in the vapor zone). Proper dwell
time decreases emissions by ensuring that the parts have
3-42
-------
reached the solvent vapor temperature prior to removal from
the machine. If parts.have not reached the solvent vapor
temperature, condensation would still occur as parts are
withdrawn from the machine and solvent carry-out losses would
increase.
A hoist can also be made to pause slightly above the
air/solvent vapor interface within the freeboard area as
cleaned parts are being withdrawn. This reduces carry-out
losses by allowing pooled solvent to drain or evaporate from .
the parts, with much of the evaporated solvent either sinking
back into the vapor zone or being condensed on cooling coils.
One test measuring the effect on emission rates of pausing in
the cold air blanket indicated that adding a 2-minute pause
above the vapor zone reduced working emissions by 46 percent
(table 3-4, test 24). This test was run on parts that
collected substantial amounts of liquid solvent on flat
surfaces. Other types of workloads that do not collect as
much liquid on surfaces would not need as much time to
accomplish adequate drying.
A test with workloads that did not collect as much liquid
on surfaces measured the effect of a 45-second pause above the
vapor zone using a mechanical hoist (table 3-4, test 25). In
this test, there was a 15-percent TCA emission reduction from
a no-dwell (pause) to a 45-second dwell (pause) operation.
Yet another benefit of using a mechanical parts handling
system is reduced worker exposure to solvent vapors. In
manual operations, the person operating the cleaner will
frequently be near the machine and may have to bend over the
top of the cleaner to lower or extract parts. Mechanical
parts handling not only reduces emissions but also allows the
operator to work farther away from the cleaner.
In order to minimize working losses, mechanically
assisted parts handling should be used while parts a're in the
solvent vapor zone, the air/solvent vapor interface, or the
freeboard area. Parts on which liquid solvent has pooled or
otherwise been trapped should remain in the freeboard area
3-43
-------
just above the air/solvent vapor interface until the liquid
solvent has completely.evaporated. Also, parts baskets should
be suspended from metal chain or cables, not from fiber rope
or any porous material that can absorb solvent.
3.2.2.2 Hot Vapor Recycle/Superheated Vapor. Another
means of significantly reducing solvent carry-out on cleaned
parts is by employing hot vapor recycle or superheated vapor
technology. These two technologies aim to create zones of
superheated solvent vapor within the vapor layer. Cleaned
parts are slowly passed through a superheated zone, which
warms the parts so that liquid solvent on parts surfaces
evaporates before the parts are withdrawn from the cleaner.
Solvent vapors heated to approximately 1.5 times the solvent
boiling point are used.22 Hot vapor recycle and superheated
vapor technologies are relatively new and predominantly used
in in-line and enclosed batch vapor cleaners. Development
work is underway for OTVC's. Further discussion of these
control techniques and their effectiveness is contained in
section 3.3.2.3.
3.2.3 Proper Operating and Maintenance Practices
Proper operating and maintenance practices are critical
to keeping solvent emissions at a minimal level; neglect can
result in significant amounts of emissions. The discussion
below recommends practices that will limit solvent loss due to
operating and maintenance activities. No effort was made to
quantify solvent loss reductions associated with these .good
operating and maintenance practices because effectiveness
varies widely, depending on current practices.
3.2.3.1 Reducing Drafts. Emissions due to diffusion and
convection can be reduced by covering the batch cleaner when
parts are not being cleaned. Reducing room drafts through the
use of baffles or by reducing room ventilation flow rate near
the solvent cleaner also reduces emissions.
3.2.3.2 Spray Techniques. For batch vapor cleaners
equipped with spray cleaning 3ysteros, spraying within the
vapor zone and at a downward angle helps control excess
3-44
-------
solvent loss. This practice reduces the amount of liquid
solvent forced out of the batch vapor cleaner and minimizes
turbulence, which can increase diffusion losses. Machines
equipped with permanently mounted spray nozzles eliminate the
possibility of spraying outside the vapor zone. With the
common use of ultrasonics to enhance cleaning, the need for
solvent sprays on many batch cleaners is minimal and could be
eliminated.
Allied Corporation has tested the effects of spraying
location on solvent loss rates. Table 3-5 presents data for
two primary condenser temperatures. The test data show that
it is important to spray parts well below the vapor line.
Solvent losses from spraying 25 cm (10 in.) above the vapor
line are 10 times higher than losses from spraying 10 cm
(4 in.) below the vapor line. These tests were conducted
using a cleaner with a .60 cm (24-in.) freeboard and 10
40-second spraying cycles per hour.
3.2.3.3 Start-Up7Shutdown Procedures. A start-up
practice that reduces solvent emissions involves starting the
primary condenser solvent flow prior to turning on the sump
heater. This practice helps condense solvent from the
saturated zone above the liquid solvent before the air is
forced out of the machine by rising solvent vapors.
Conversely, a good shutdown practice involves allowing the
condenser to stay on after the sump heater has been turned
off, until the vapor layer collapses. Solvent cleaners that
operate on a heat pump design cannot accommodate independent
control of heating and cooling because heat input and
condensation are part of the same thermodynamic cycle.
3.2.3.4 Downtime Losses. Solvent evaporation during
downtime can be significant; especially for CFC-113 and
methylene chloride. Using covers during downtime will reduce
drafts and slow diffusion, but will not stop losses '
completely. Several techniques can be used to reduce downtime
losses—for example, operating a freeboard refrigeration
device, cooling the sump to reduce solvent vapor pressure, and
3-45
-------
id
CO
w
u
I
ft.
O
z
M
i
0-
co
CO
J^
CO
tf
w
u
c^
ȣ
(X
CO
CO
s
z
SM
J>
n3
o
CO
•
in
i
n
W
03
f4?
H
0)
4J
id
0
C
•H
.H
0
o
o
£
o
o
0
rfp
43 -.
* b
1?
Hi
w 43
U) M
o *-
j
IN! "'
H
0)
4J
(C
3
Q
'*j
o
o
u
s
0
o
in
U
O
2
»
f* ^^
* i4
(M 43
B *
'tri'jJ
^
^m^
<*)
H
rH
0
in
VO
CO
in
o
iH
n
x-^
ra
r-
o
•
o
""""
o
r-t
•
O
^*
in
vo
o
1-1
^
o
a
id
0)
>
5
id
*•*
•
c
•H
in
***•
B
o
oj
H
•
O
n
0
o
H
*Bi^
^J"
CO
o\
o
o
eg
•*
VO
CO
^~.
CO
in
•
o
*~*
o
(M
CM
Jj
O
a
id
^
0)
o
id
^.-^
t
c
•H
O
H
«-^
E
^«
in
OJ
•
o
id
•o
m
0)
c
0)
•H
u
Q
J£
^
O
CO
U
u
^
j^
3
O
t^
0)
a
M
0)
iH
U
o
•o
c
o
o
0)
IQ
1
O
0
f-f
• •
10
0)
. Q)
B
id
id
a
4J
Ul
0)
-------
pumping solvent out of the machine to an air-tight storage
drum. Of these techniques, cooling the sump during downtime
is reportedly very effective in reducing solvent losses due to
evaporation. Sump cooling can be accomplished by two methods:
(1) the liquid solvent can be cooled during downtime by
cooling coils, or (2) the air blanket directly above the
liquid solvent can be cooled by an overlay coil. One vendor
indicated that cooling the sump can reduce downtime losses by
90 percent.11 Insufficient data are available for downtime
sump cooling, and cover controls to assign control
efficiencies.
3.2.3.5 Workload Introduction/Removal. Emissions due to
the entry and removal of parts can be reduced with good
operating practices. One such practice is to limit the
introduction rate of the workload in order to minimize the
turbulence created when the load is lowered into the cleaner.
Limiting the introduction rate of the load so that the
air/solvent vapor interface does not fall more than a few
inches will prevent excessive pump-out of mixed solvent vapor
and air as the vapor layer reestablishes. As stated
previously, the use of a mechanical parts mover can
substantially eliminate these emissions.
Emissions can also be reduced by limiting the horizontal
area of the load to be cleaned to 50 percent or less of the
batch vapor cleaner's air/solvent vapor interface area. This
will reduce the displacement and turbulence of solvent vapors
as the load is lowered into the cleaner. Large parts baskets
could be used without increasing emissions if the basket speed
were reduced when the basket moves through the vapor zone.
3.2.3.6 Parts Drainage. An important operating practice
that minimizes solvent carry-out on cleaned parts is proper
racking to avoid solvent puddles. Parts with recesses or
blind holes should be rotated or agitated prior to removal
from the vapor layer to displace trapped solvent. Powered
rotating baskets (discussed in section 3.3.2.2) can also be
used to effectively limit liquid carry-out. Cleaning porous
3-47
-------
or absorbent materials should be avoided because such parts
can carry out excessive quantities of solvent. Also, the part
being cleaned should be allowed to reach the solvent vapor
temperature prior to removal from the vapor layer so that
solvent condensation on the parts no longer occurs.
3.2.3.7 Leak Detection/Repair. Solvent emissions can be
controlled by repairing visible leaks and promptly repairing
or replacing cracked gaskets, malfunctioning pumps, water
separators, and steam traps. Routine equipment inspections
will help locate leaks or problem areas more quickly. Halide
detectors that can be used to identify leaks are available at
a reasonable cost ($150 to $500).
Leaks at welded joints can be avoided if the batch
cleaner vendor tests the joints prior to shipping. The test
must be sensitive enough to detect fine cracks. A simple
water test is not sufficient because the high surface tension
of water prevents penetration of small cracks. Often a dye
penetrant is used. Machines made with 316L stainless steel
walls will be less prone to stress cracks. Pressure fittings,
as opposed to threaded connections, will also reduce leaks.23
Clean-out doors, viewing ports, or other gasketed machine
parts must be carefully designed and manufactured. Gasket
material must be nonporous and resistant to chemical attack
from the solvent. Ill-fitting gaskets or use of improper
gasketing material can result in large solvent losses.
3.2.3.8 Solvent Transfer. Losses during transfer of
solvent into and out of the batch cleaner can be controlled by
correct operating practices. Ideally, solvent filling,
draining, and transfer operations should be by pipe in closed
systems. Some vendors have systems that allow solvent to be
pumped from the solvent drum directly into the solvent
cleaner.22 This could cut down on spill losses and diffusion
associated with solvent filling. If the solvent is pumped
into the cleaner with little or no splashing, such as with
submerged fill piping, less solvent would be lost. Losses
during transfer of contaminated solvent or sump bottoms from
3-48
-------
the batch cleaner sump to stills or waste solvent storage can
be controlled by using leakproof couples. Transfer to a
vented tank or sealed containers will help reduce emissions.
Solvent that has been contaminated with water should
either be purified in a water separator or replaced with fresh
solvent. Water contained in the solvent enhances diffusion
losses (except for CFC-113 solvent).
3.2.3.9 Safety Switches. Control switches are devices
used on vapor cleaners to prevent unsafe conditions such as
vapor overflow, solvent decomposition, and excess solvent
consumption. Common types of control switches include
(1) vapor level control thermostat, (2) condenser water
pressure switch or flow switch and thermostat (for water
cooled machines), (3) sump thermostat, (4) liquid solvent
level control, (5) spray pump control switch, and
(6) secondary heater switch. The first four switches turn off
the sump heat and the fifth turns off the spray when
conditions within the machine exceed proper operating
conditions. The most important switch is the vapor level
control thermostat, which turns off sump heat when the solvent
vapor zone rises above the design operating level. The
secondary heater, found on some machines, is activated when
introduction of a large load causes the vapor level to fall.
Secondary heaters reduce solvent loss from vapor level
fluctuation.
As oils, greases, and other contaminants build up in the
solvent, the boiling point of the mixture increases. Both the
sump thermostat and the liquid solvent level control prevent
the solvent from becoming too hot and decomposing. The sump
thermostat cuts off the heat when the sump temperature rises
significantly above the solvent's boiling point, which will
occur as contamination of solvent increases. The solvent
level control turns off the heat when the liquid leVel of the
boiling sump drops nearly to the height of the sump heater
coils. In the case of electrically heated coils, excessive
3-49
-------
heat could decompose the solvent, emitting toxic and corrosive
decomposition products.
For steam-heated units or units that use a heat pump
system, solvent decomposition is less likely because these
heat sources normally do not reach solvent-decomposing
temperatures. However, solvent level controls can be useful
on machines using these heat sources, especially for the
higher-boiling solvents TCE and PCE, because low liquid levels
permit high concentration of soils that can "bake" onto
heating elements, seriously impairing heat transfer and
possibly contributing to solvent decomposition. Although
these heat sources cannot reach temperatures where solvent
decomposition is rapid, hotter mixtures of solvent and sludges
can cause solvent deterioration more quickly than the cooler
operating temperatures of relatively clean solvent.
Therefore, a solvent level switch can still be useful by
signalling the time for solvent cleanup.
The spray pump control switch is not used as often as the
other safety switches, but it can offer a significant benefit.
If the vapor level drops below a specified level, this control
cuts off the spray pump until the normal vapor level is
resumed. Then the spray can be manually restarted. This
prevents spraying with an inadequate vapor level, which can
cause excessive emissions of sprayed solvent. The spray pump
control switch sometimes also has a feature that cuts off the
spray pump if spraying is outside the vapor zone.
Although the effectiveness of the above controls cannot
be quantified, it is expected that these switches can protect
against potentially significant emissions from upset
conditions.
3.3 IN-LINE CLEANERS
In-line cleaners can be cold cleaners, vapor cleaners, or
4
combination cold/vapor cleaners; however, the majority of
cleaners that use chlorinated/chlorofluorinated solvents are
vapor cleaners.. Those cleaners are nearly always enclosed
except for entrance/exit ports and employ a continuous loading
3-50
-------
system. Unlike batch cleaners, which are often "off-the-
shelf" items, in-line cleaners are normally custom-designed
for a specific workload and production rate. In-line cleaners
are used in a broad spectrum of metal-working industries, but
are most often found in plants where there is a constant
stream of parts to be cleaned and where the advantages of
continuous cleaning outweigh the lower capital cost of a batch
cleaner.
The control techniques applicable for use with in-line
cleaners vary according to machine design and operation. The
following controls are presented in this section: minimizing
entrance/exit openings, carbon adsorption, freeboard
refrigeration devices, drying tunnels, rotating baskets, and
hot vapor recycle/superheated vapor systems.
Test data were not available to evaluate the
effectiveness of all the in-line cleaner control techniques
listed above. Only four tests were available: three that
evaluated the effectiveness of a freeboard refrigeration
device (two BFC's and one AFC) and one that evaluated a carbon
adsorber. These tests are discussed in the appropriate
following subsections and are summarized in table 3-6.
3.3.1 Controls for Interface Emissions
3.3.1.1 Minimizing Entrance/Exit Openings. Although
in-line cleaners are mostly enclosed by design, additional
emissions control can be achieved by minimizing opening areas
and covering the openings during nonoperating hours.
Reducing entrance and exit opening areas reduces idling and
working losses due to diffusion by minimizing air drafts
inside the cleaner. Air drafts increase emissions by sweeping
away solvent-laden air near the air/solvent vapor interface
and promoting mixing and diffusion by increasing turbulence in
the freeboard area.
Among in-line cleaners, monorail cleaners tend to have
the greatest diffusion emissions due to drafts through the
machine caused by openings at opposite ends. In-line machines
with U-bend designs eliminate the problem of air currents
3-51
-------
•
(0
U
|
J
0
u
M
J
1
z
M
OS
g
CO
p.. J
C"1
to
u
E-t
U
J
03
S
i-i
<
<
fa
0
>
f£
CO
•
vo
1
n
M
.J
CQ
E^
X
"o £
L. " ^N
|5*
"t
41 •
|
"o
u
I
S
w
a
£
L.
1!
m S
5|
X
1. b.
ft)
.^
s-g
(/I
.1
•1
VI
(D
00
g
(A
CA
E
u
Carbon
adsorber
X
>
^ JK o> a
« a ^ "*
in a o in
(M O IM (M
1 1 I *
* z z
u u u )t
U. U. U. X
< 00 CO S
^- — ^
f f
i i % "^
^* X. *v *s.
0) 01 0> 01
J* ^< ^ .*
eo eo o o
^ 2 g 5
- r t «•
s s s «
•^ * o .3
CD
X X
ft*. kk.
o o
1 ^. S B
81*"
S S
C/>
u u u o
iiv kek u. *C
< IB CO U
„.!
==l
its
??&
— —IT
M M O
5 • js
SS8
til
•§ * •
* i 0
H H II
-------
flowing through the machine. Also, many in-line cleaners,
such as monorail cleaners, can be constructed so that internal
baffles reduce the effect of air flow through the machine (see
figure 3-15).
Silhouette openings and hanging flaps decrease the area
through which diffusion losses can occur and restrict drafts
inside the cleaner, but they will have a minimal effect on
emissions if the openings are already relatively small. When
the in-line cleaner is not in use, port covers should be used
to reduce downtime emissions.
The extent to which a reduced entrance/exit opening area
affects emissions depends on the total open area. The
relative importance of using port covers in overall emissions
reduction depends on the operating schedule. Port covers are
most essential when the fraction of the daily schedule that
the cleaner spends in downtown mode is substantial.
3.3.1.2 Carbon Adsorption. Venting solvent vapor
emissions to a carbon adsorption system is a major emissions
control technology for both diffusion losses and working loses
from in-line vapor and cold cleaners. Carbon adsorbers are
effective emission control devices and can be cost effective
because captured solvent is recycled. The enclosure around
in-line cleaners makes it easier to capture and duct emissions
to the carbon adsorber, and overall efficiencies are higher on
in-line cleaners than on batch cleaners. The relative degree
of emissions control depends on the cleaner design, workload
characteristics, and the solvent emissions capture efficiency.
See section 3.2.1.7 for further discussion of control by
carbon adsorption.
The available test on carbon adsorbers shows an
approximately 60-percent emissions reduction efficiency when
applied to an in-line cleaner (i.e., circuit board stripper).4
However, with some solvent mixtures, the same operating
problems described for batch cleaners in section 3.2.1.7 could
occur.
3-53
-------
0)
C
-------
3.3.1.3 Freeboard Refrigeration Devices. A freeboard
refrigeration device on an in-line vapor cleaner functions in
the same way as on a batch cleaner. Refrigeration establishes
a cool air layer above the vapor zone, which inhibits
diffusion and solvent/air mixing. (See section 3.2.1.2 for a
more detailed discussion of freeboard refrigeration devices.)
Only three tests evaluating the effect of freeboard
refrigeration devices on in-line vapor cleaner emissions were
available to the EPA. One of these tests evaluated an AFC and
two evaluated BFC's on an in-line circuit board defluxer. The
test data indicated that a BFC can reduce in-line cleaner
emissions by about 60 to 70 percent. An AFC can achieve about
a 10-percent emissions reduction.4
3.3.2 Controls for Workload Emissions
3.3.2.1 Drying Tunnels. A drying tunnel is simply an
add-on enclosure that extends the exit area of in-line
cleaners. The tunnel reduces carry-out losses because solvent
evaporating from cleaned parts exiting the machine may be
contained within the drying tunnel rather than being lost to
the atmosphere. Much of the evaporated solvent in the drying
tunnel will sink back into the vapor zone and be recovered.
If the machine is connected to a carbon adsorber, the
evaporated solvent in the drying tunnel will be drawn into the
adsorber and recovered. A drying tunnel works well in
conjunction with a carbon adsorber.
The effectiveness of a drying tunnel depends on several
factors. Because a drying tunnel primarily reduces carry-out
emissions, the effectiveness of this device depends on the
amount of carry-out before installation of the tunnel. The
amount of control also depends on the length of time the parts
are in the drying tunnel. The length of time necessary will
depend on the solvent type and part configuration. If
sufficient time is allowed, essentially all carry-out
emissions could be eliminated (except from the most intricate
or "solvent trapping" parts).
3-55
-------
A drawback to using a drying tunnel as a control device
is the large amount of floor space they require. Although
floor space may not be available in all plants to add drying
tunnels to existing cleaners, they can be planned for when new
machines are purchased.
3.3.2.2 Rotating Baskets. A rotating basket is a
perforated or wire mesh cylinder that is slowly rotated while
proceeding through the cleaner. The rotation prevents
trapping of liquid solvent on parts and therefore reduces
carry-out emissions. As with drying tunnels, the control
effectiveness of rotating baskets is not easily quantifiable.
The effectiveness depends on the workload shape and the way
the parts are loaded into the basket.
Not all parts can be tumbled in baskets without being
damaged. Therefore, rotating baskets are not applicable to
all operations. Also, rotating baskets are usually designed
into the conveyor and, hence, are not easily retrofitted on
existing cleaners.
3.3.2.3 Hot Vapor Recycle/Superheated Vapor. Hot vapor
recycle and superheated vapor are relatively new but promising
technologies. Vendors are reporting that these technologies
have the potential to significantly reduce carry-out emissions
from both batch cleaners and in-line vapor cleaners. An
in-line cleaner equipped with superheated vapor is shown in
figure 3-16.
Both hot vapor recycle and superheated vapor technologies
operate on the same principle: they aim to create zones of
superheated solvent vapor within the vapor layer. Cleaned
parts are slowly passed through a superheated zone, where they
are warmed so that liquid solvent on part surfaces evaporates
before the parts are withdrawn from the cleaner. Solvent
vapor is heated to approximately 1.5 times the solvent boiling
point.19 (One contact indicated that solvent vapor 'is heated
to the highest temperature possible without decomposing the
solvent to speed drying.24)
3-56
-------
X!
•H
*
0)
(0
0)
rH
o
0)
I
c
id
M
O
§
0)
O
-H
>
0)
4J
A
•O 0)
0) J=
iH ^
-------
The hot vapor recycle process uses continuous
recirculation of the solvent vapor. Solvent vapor is drawn
from the vapor zone, circulated through a heater, and blown
back into the vapor zone through a system of distribution
slots. In the superheated vapor process, heating coils placed
at one end of the vapor zone superheat a sector of solvent
vapor through which cleaned parts are passed.
Hot vapor recycle is generally applicable only to in-line
or non-OTVC batch vapor cleaners because some type of
*
enclosure is necessary for effective recirculation of solvent
vapor. The movement of vapor creates turbulence and tends to
increase solvent loss unless the machine is enclosed or
baffles are present. Superheated vapor technology can
reportedly be applied to both in-line cleaners or batch
cleaners.
Hot vapor recycle and superheated vapor have been
predominantly used with chlorofluorinated solvents. The
technologies are attractive because of the potential savings
of these costly solvents. Hot vapor recycle has been used in
one application to clean condenser coils in a monorail cleaner
using PCE.
One industry contact claims that a 90-percent reduction
in carry-out emissions is possible.4 This would be
approximately a 50 percent reduction of overall working
emissions for a typical cleaning machine.1
The potential for significant emissions reduction is
apparent. Normally, cleaned parts will emerge from the vapor
zone of a cleaner with a thin film of liquid solvent on all
surfaces, and possibly pooled solvent in holes and crevices.
Much of this liquid solvent may not evaporate until parts are
out of the machine. If all solvent film and pooled solvent
are evaporated prior to leaving the vapor zone, large solvent
savings should ensue. The only workload losses remaining
would be associated with air/solvent vapor interface
disturbances and vapor entrainment due to the speed of the
conveyor.
3-58
-------
3.3.3 Proper Operating and Maintenance Practices
3.3.3.1 Conveyor Speed. There are several operating and
maintenance practices that can significantly reduce solvent
emissions from in-line cleaners. By controlling conveyor
rates at or below 3.3 m/min (11 ft/min) absolute speed,
solvent emissions due to vapor zone turbulence and carry-out
can be minimized. The 3.3-m/min (11-ft/min) limit should be
measured as an absolute rate, not a vertical speed (i.e., only
3.3 m [11 ft] of conveyor should pass any spot in 1 minute).
Conveyor rates can be controlled by properly gearing the
electric motor drives.
3.3.3.2 Spray Techniques. Emissions can be minimized by
proper design of fixed spray systems. Nozzles should direct
spray horizontally or downward to keep from piercing the vapor
layer, or the spray area should be separated by baffles from
the rest of the vapor zone.
3.3.3.3 Start-up/Shutdown Procedures. Losses can be
reduced by: (I) starting the primary condenser waste flow
prior to turning on the sump to help condense excess loss as
the vapor layer rises; (2) maintaining the condenser water
flow after shutdown of the sump heater until the vapor layer
has collapsed and the liquid solvent has cooled to room
temperature; (3) cooling the sump during downtime or operating
cooling coils above the sump; and (4) covering the entrance
and exit ports duiing downtime.
. 3.3.3.4 Carbon Adsorber Procedures. For in-line
cleaners with carbon adsorbers (CADS), several operating
practices can be employed that help reduce emissions. The
practices include: (1) not bypassing the carbon adsorber
during the desorption cycle; (2) proper carbon bed
regeneration frequency, so as to prevent solvent breakthrough;
(3) leak checks of the carbon adsorption systems; and (4) good
steam condensate separations. • - .
. 3.3.3.5 Parts Drainage. As with batch cleaners, an
important operating practice that minimizes solvent carry-out
in in-line operations is proper racking to avoid solvent
3-59
-------
puddles. Where pooling of solvent cannot be avoided, rotating
or agitating parts prior to removal from the vapor layer is
necessary to displace trapped solvent. Rotating baskets
(discussed in section 3.3.2.2) can also be used to limit
liquid carry-out. The cleaning of porous or absorbent
materials, which will absorb and carry out excessive
quantities of solvent, must be avoided. Also, conveyor speed
must be adjusted so that parts being cleaned are allowed to
reach the solvent vapor temperature prior to removal from the
vapor layer, and so that solvent is not visible on emerging
parts.
3.3.3.6 Leak Detection/Repair. Solvent emissions can
also be controlled by repairing visible leaks and promptly
repairing or replacing cracked gaskets, malfunctioning pumps,
water separators, and steam traps. Routine equipment
inspections (particularly with a halide detector) will help
locate leaks or problem areas more quickly. Leaks at welded
joints can be avoided if the in-line cleaner vendor tests the
joints prior to shipping. The test must be sensitive enough
to detect fine cracks. A simple water test is not sufficient
because the high surface tension of water prevents penetration
of small cracks. Often a dye penetrant is used. Machines
made with 316L stainless steel walls will be less prone to
stress cracks. Pressure fittings, as opposed to threaded
connections, have also been reported to reduce leaks.
Clean-out doors, viewing ports, or other gasketed machine
parts must be carefully designed and manufactured. Gasket
material must-be nonporous and resistant to chemical attack by
the solvents. Ill-fitting gaskets or use of improper
gasketing material can result in large solvent losses. One
test of an in-line cleaner showed that inadequate sealing
around a viewing door accounted for losses of 0.9 kg/hr
(2.3 lb/hr).9 Sealing the window with duct tape eliminated
these losses.
3.3.3.7 Solvent Transfer. Losses during transfer of
solvent into and out of the in-line cleaner can be controlled
3-60
-------
by correct operating practices. Solvent filling and draining
should be completed in as closed a system as possible. As
stated previously, some vendors have systems that allow
solvent to be pumped from the solvent drum directly into the
solvent cleaner. This could cut down on spill losses and
diffusion associated with solvent filling. If the solvent is
pumped into the cleaner with little or no splashing, such as
with submerged piping, less solvent would be lost. Losses
during transfer of contaminated solvent or sump bottoms from
the in-line cleaner sump can be controlled by using leakproof
couples. Transfer to a vented tank or sealed containers will
help reduce emissions.
3.3.3.8 Safety Switches. In-line cleaners should also
have the appropriate safety switches to ensure proper
operation. A complete discussion of safety switches is
included in section 3.2.3.
3.4 COLD CLEANERS
As discussed in chapter 2.0, carburetor cleaners are the
only type of cold cleaner currently known to be manufactured
for use with a halogenated solvent. These machines are
typically well controlled with a water cover. The water cover
substantially limits evaporation losses because very little
solvent comes into contact with the air. Many such machines
are designed to be closed during the cleaning cycle (as well
as during downtime and idling) and further reduce diffusion
losses due to drafts and splashing of solvents. Based on one
available test, water covers can reduce evaporation losses by
at least 90 percent.25
Cold cleaners other than carburetor cleaners could also
benefit from the use of water covers. Existing cold cleaners
that use halogenated solvents, as well as any new cold
cleaners manufactured for use with halogenated solvents,
should employ water covers to control evaporation. '
Simple work practices can limit working losses. These
practices include allowing adequate drainage of parts and
flushing parts only within the confines of the cleaner. These
3-61
-------
practices would apply to new or existing cold cleaners of any
type.
3.5 ALTERNATIVE CLEANING TECHNOLOGIES
In some instances, emissions of the five common
halogenated solvents from surface cleaning or degreasing
operations can be eliminated or reduced by using alternative
solvents or technologies. Currently, there are many available
alternative solvents and technologies. Choosing one is
usually case-specific because of the many variables that need
to be considered, such as: (l) the contaminants to be
removed; (2) the material to be cleaned; (3) the size, shape,
and configuration of the part; (4) the level of cleanliness
desired; and (5) current cleaning method(s).
Alternative solvents, alternative cleaning processes, and
no-clean technologies are discussed in sections 3.5.1, 3.5.2,
and 3.5.3, respectively.
3.5.1 Alternative Cleaning Solvents
Alternative cleaning solvents are generally grouped into
one of the following categories: (1) hydrochlorofluorocarbons
(HCFC's), (2) aqueous, (3) semi-aqueous, or (4) organic
solvents. The alternative solvent and the halogenated solvent
that it replaces are described below. Any process or
equipment modifications required by the alternative solvent
are also discussed. Finally, some of the alternative
solvents' benefits and limitations are covered.
3.5.1.1 Hvdrochlorofluorocarbons. The HCFC's are
typically used or considered for use as replacements for
chlorofluorocarbons (CFC's) and TCA, but have the potential to
replace other halogenated solvents as well. Typical HCFC's
used to replace halogenated solvents are presented in
table 3-7. All CFC's will be banned by the year 2000 under
1990 Amendments and the First Renegotiation of the Montreal
Protocol (held in June 1990). However, an accelerated
phase-out of CFC's by 1996 is likely. Because HCFC production
will also be banned in the future, HCFC's are only a temporary
alternative to halogenated solvents.
3-62
-------
TABLE 3-7. COMMON HYDROCHLOROFLUOROCARBON ALTERNATIVES
Common name Chemical formula
HCFC-123 CF3CHC12
HCFO141b CHC12FCH3
HCFC-225ca CF3CF2CHC12
HCFC-225cb CCLF2CF2CHCLF
3-63
-------
By far, the most common HCFC's are HCFC-123 and
HCFC-141b. Lesser quantities of HCFC-225 and blends are also
used. Blends are HCFC mixtures or HCFC's combined with
organic chemicals such as methanol. These three common HCFC's
(HCFC-123, -141b, and -225) are halogenated two- and
three-carbon compounds.
Existing equipment can often be modified for HCFC's.
Equipment modifications for conventional solvent cleaners may
include extending or insulating the freeboard and lowering the
condenser temperature. In addition, superheated vapor drying
or increased dwell time in the freeboard area will reduce
emissions due to carry-out. Most of the HCFC solvent will
volatilize and be condensed in the cleaning step. Therefore,
rinsing and drying stages are not generally required.
The health effects of HCFC's are still being
investigated. Results of preliminary toxicological testing
indicate that rats exposed to HCFC-123 developed cancer.26
Excluding HCFC-225, HCFC's have lower boiling points than
CFC-113 and TCA; therefore, less energy is required to boil
the HCFC's. However, for a given cleaner, HCFC emissions
could be higher than CFC emissions (because of increased
volatility) unless additional controls are applied. Also,
HCFC's have lower ozone-depleting potential (OOP) and global-
warming potential (GWP) than does CFC-113. The HCFC's OOP is
about one-sixth that of the CFC-113 OOP and is similar to the
TCA ODP. Although the HCFC's ODP and GWP are lower than those
for CFC-113, they tend to be higher than those for the other
alternatives mentioned in this section.
3.5.1.2 Aqueous. Aqueous cleaners are being used and
investigated for use as alternatives to most of the common
halogenated solvents. Aqueous cleaners use water as the
primary solvent and can effectively clean inorganic or polar
soils, oils and greases, particles, and films.27'29 In
addition to water, aqueous cleaning solvents typically contain
alkaline salts, surfactants, and additives. Examples of
alkaline salts are phosphates, hydroxides, silicates,
3-64
-------
carbonates, and borates. The surfactants provide detergency,
emulsification, and wetting. Other additives that enhance
cleaning effectiveness include glycol ethers, chelating
compounds, metal salts, and amines.27""29
The major steps in aqueous cleaning are washing, rinsing,
drying, and wastewater treatment and disposal.27'29 During
washing, the aqueous solution loosens and removes the
contaminant in one cleaning tank. This cleaning may be
performed by immersion, spraying, ultrasonics, or some
combination of these. Next, the rinse stage uses a single
immersion tank or multiple-stage tanks, depending on the
residue(s) that need to be removed. The parts must then be
dried. Evaporative drying alone is not usually practical
because of time constraints. In many instances, evaporative
drying will follow mechanical drying. Mechanical drying
methods include compact turbine blowers, compressed air,
slow-pull drying, and hot air recirculation. Some evaporative
drying techniques are infrared heating and vacuum drying.27"29
Drying may not be required if the part proceeds to another wet
process. The choice of rinsing and drying equipment and the
operating parameters of the cleaning system depend on the
requirements of the part to be cleaned.
Finally, the wastewater must be treated. Likely
contaminants are organics, metals, and oils and greases. In
addition, high pK can result when alkaline substances are
present. Organic treatment can be accomplished by using a
carbon adsorber and, sometimes, by using ultrafiltration or
reverse osmosis. Oils and greases are typically treated by
gravity separation, ultrafiltration, coalescence, or with
chemicals. Metal contamination can be controlled by ion
exchange, filters, or lime or caustic soda. For pH
adjustment, sulfuric or hydrochloric acid are regularly
used.27'29
In some cases, existing equipment may be retrofitted for
the washing stage. However, when rinsing, drying, and
wastewater disposal were not part of the original halogenated
3-65
-------
solvent cleaning operations, new equipment is normally
required for these additional steps. These extra rinsing and
drying stages increase water and energy use. In addition,
high-purity water may be required for certain applications
and more floor space may be required for the new equipment.
Nevertheless, chemical costs and consumption are usually lower
for aqueous cleaning than for halogenated solvent cleaning.
Aqueous systems exhibit fewer health and safety problems
than halogenated solvent systems. Most aqueous solutions are
not toxic, flammable, or explosive, although some additives
may present toxicity or volatile organic compound (VOC)
concerns.
3.5.1.3 Semi-Aqueous. Semi-aqueous cleaners are being
used and investigated for use as alternatives to most of the
common halogenated solvents. The semi-aqueous cleaners can
effectively remove heavy grease, tar, waxes, hard-to-remove
soils, and polar as well as nonpolar contaminants.27""29
Typically, semi-aqueous cleaning solutions combine terpenes or
hydrocarbons with surfactants and additional additives such as
corrosion inhibitors. The hydrocarbon or terpene dissolves
the contaminants and the surfactant provides wetting,
emulsifying, and rinsing properties.
As with aqueous systems, the semi-aqueous cleaning
process involves four stages: washing, rinsing, drying, and
wastewater treatment and disposal. In the semi-aqueous
process, the washing stage involves two steps. In the first
step, the contaminants on the part are loosened by a
concentrated hydrocarbon (or terpene)/surfactant mixture. The
cleaning may be performed by either immersion, spray, or a
spray-under-immersion system. Spraying or misting with a
semi-aqueous solution is only used when the system temperature
is below the flash point of the solvent.
In the second step, the part is placed in an emulsion
rinse. The semi-aqueous emulsion can be separated into
hydrocarbon/contaminant and water layers for reclamation. The
remaining solvent may be discharged to a solvent recovery
3-66
-------
system and separated from the contaminants. Aqueous effluent
can be recycled in a closed-loop system or sent to a
wastewater treatment facility. Rinsing, drying, and
wastewater treatment procedures are similar to those in the
aqueous process (section 3.5.1.2). Also, equipment
modifications and purchases are similar to those cited for
aqueous cleaners.
The benefits and limitations of using semi-aqueous
systems are comparable to those listed for aqueous systems
(see section 3.5.1.2). Any benefits or limitations that are
different from those of aqueous cleaners are mentioned below.
In the electronics industry, semi-aqueous cleaners work
well with all common flux types, so the choice of flux type
can be made solely on the basis of soldering performance. The
low surface tension of semi-aqueous cleaners also allows for
penetration into narrow spaces (e.g., surface mount
assemblies).
Semi-aqueous systems may be flammable because of low
flash points, requiring proper equipment design and
temperature control. Also, the toxicity of some of the
additives has not been fully evaluated. Finally, some
cleaners (e.g., some terpenes) have strong odors that may be
objectionable.
3.5.1.4 Organic. Organic cleaners have been used as
alternatives for roost of the common halogenated solvents.
They use no water and include oxygenated hydrocarbon
formulations, aliphatic hydrocarbons, esters, glycol ethers,
alcohols, aromatics, and ketones.
The organic cleaning process typically occurs in a single
step. During the cleaning stage, some type of agitation
(e.g., ultrasonics) is generally used to enhance cleaning
efficiency. Spraying or misting with an organic cleaner is
used only when the system temperature is below the flash point
of the solvent. Each wash stage is followed by a drainage
period to minimize solvent carry-out from the system. A
drying time is required because the organic solvents are not
3-67
-------
as volatile as the halogenated solvents. The drying stage may
involve ambient air or heated forced air.
Generally, new equipment must be purchased for organic
cleaning. This is primarily due to the nature of the organic
chemicals. Some organics are flammable or explosive and,
therefore, proper equipment design and control equipment may
be required for VOC. In addition to flammability and
explosivity concerns, the toxicity of some organic cleaners
has not been fully evaluated.
Organic cleaners are predominantly used because of their
good solvency, but this property causes them to be
incompatible with some materials.
3.5.2 Alternative Cleaning Technologies
Some alternative cleaning technologies can replace the
use of the common halogenated solvents. Most of these
technologies are still in the developmental stages.
3.5.2.1 Ice Particles. Ice particle cleaning is a new
electronic cleaning technology that removes post-solder
residues. Curing this process, ice particles ranging from
O.l to 300 micrometers (0.0000039 in. to 0.0118 in.) in
diameter are sprayed at varying angles and pressures.30
3.5.2.2 Plasma. In plasma cleaning, ions and electrons
in the plasma are energized by radio frequency radiation to
energy levels that are similar to the energy levels of the
bonds typically found in organic contaminants. The high
reactivity of the ions, combined with their kinetic energies,
is sufficient to break the organic bonds. The ions then react
with the freed atomic components to form volatile compounds,
which are then removed by the flow of the process gas. Plasma
cleaning is typically used for precision parts.29
3.5.2.3 Pressurized Gases. This involves cleaning with
high-pressure gases such as air, carbon dioxide,
chlorofluoromethane, or nitrogen. The gases must be'filtered
and 'free of contaminants. Pressurized gases are typically
used in precision cleaning operations for the removal of
nonmetallic dust and particles.30
3-68
-------
3.5.2.4 Supercritical Fluids. Supercritical fluids
(SCF's) are gases that are above their critical temperature
and pressure. At these high temperatures and pressures, the
gases exhibit solvency properties similar to those of fluids.
The most common SCF is carbon dioxide. These SCF's are
generally used for precision parts.29
3.5.2.5 Ultraviolet/Ozone. This cleaning process
involves exposing a contaminated surface to ultraviolet (UV)
light in the presence of ozone. Cleaning occurs as a result
of various photo-sensitized oxidation processes. Contaminant
molecules are excited and/or dissociated by the absorption of
short-length UV light. These molecules and the free radicals
produced by dissociation react with atomic oxygen to form
simpler volatile molecules such as carbon dioxide, water
vapor, and nitrogen. This reaction facilitates the removal of
surface contaminants in precision cleaning operations.28
3.5.2.6 Mechanical. Mechanical technologies include
brushing, wiping with rags and sponges, use of sorbent
materials, wheat starch blasting, and carbon dioxide blasting.
These methods are generally used for less problematic cleaning
requirements.
3.5.2.7 Thermal Vacuum Deoiling. This cleaning system
uses a heated vacuum chamber to remove .oil from parts by
vaporizing the oil. The vapors are then pumped through a
condensing unit and drained from the system.
3.5.3 No-Clean Technologies
Process modifications may eliminate the need for
degreasing or surface cleaning. Several no-clean technologies
are currently available for use and are described below.
3.5.3.1 Low-Solids Flux. For electronic applications,
rosin fluxes typically have a solids content between 15 and
35 percent. A low-solids flux contains 1 to 10 percent
solids. A variety of spray systems are commercially' available
that properly apply the flux to minimize the remaining
residue. Typically, little or no visible flux remains on the
3-69
-------
board after soldering. Therefore, the use of a low-solids
flux eliminates the post-soldering cleaning stage.30
The difficulties with a low-solids flux include: (l) the
low-solids flux may not be compatible with all materials and
equipment designs; (2) the current soldering machine may need
to be adapted because of the change in the flux application
technique; and (3) a residue remains on the board after
soldering, possibly not complying with some specifications.
3.5.3.2 Controlled Atmosphere Soldering. Controlled
atmosphere soldering can be divided into two categories:
inert and reactive. For inert atmosphere soldering, an inert
gas (usually nitrogen) is present in the soldering system
instead of oxygen. Minimizing the oxygen in the system
reduces the formation of oxides, which enhances the wetting of
the molten solder to the surface being soldered. For reactive
gas soldering, an inert gas and another reactive element
(formic acid) are present in the soldering system. Again,
minimizing the oxygen reduces the formation of oxide layers.
The wave soldering equipment used for these two systems
is designed to operate under a vacuum or a nitrogen blanket in
order to reduce the oxygen in the system. Controlled
atmospheric soldering eliminates the need for post-soldering
cleaning. Additionally, a more metallurgically pure solder
may be used. A major drawback for controlled atmosphere
soldering is the capital cost for new equipment.
3.5.3.3 Process Modifications. In the metals industry,
a variety of water-soluble lubricating or cutting oils are
available for metal-working processes. This may allow the
part to avoid being cleaned or to be sufficiently cleaned with
water. "Dry" lubricants or thin polymer sheeting can be used
during some metal-working operations and then peeled from the
surface after processing.30 This decreases the quantity of
contaminants remaining on the part. Using organic peel
coatings instead of oils as preservatives in storage and
shipping reduces oil residuals. These process modifications
3-70
-------
can eliminate the frequency of cleaning in metal-working
operations.
3.6 REFERENCES
1. Memorandum from Miller, S. M., J. O'Lough1in, and
C. Sarsony, Radian Corporation, to Almod6var, P.,
U. S. Environmental Protection Agency. July 2, 1992.
Summary of emission reductions for selected control
techniques for organic solvent cleaners.
2. Letter and attachments from Delta Sonics, to D. A. Beck,
U. S. Environmental Protection Agency. February 1988.
Estimation of freon solvent usage in open top DS-Series
Delta Sonics degreasers.
3. Nylen, G. C. and H. F. Osterman (Allied Corporation).
Cool it to Cut Degreasing Costs. American Machine
November 1982.
4. Suprenant, K. S. and D. W. Richards (Dow Chemical
Company). Study to Support New Source Performance
Standards for Solvent Metal Cleaning Operations.
Prepared for the U. S. Environmental Protection Agency.
June 30, 1976.
5. Emission Measurement Branch Test Data.
U. S. Environmental Protection Agency, Emission
Measurement Branch, Research Triangle Park, NC.
6. Memorandum from Goodrich, J., Detrex Corporation, to
Schlossberg, Detrex Corporation. October 17, 1980.
Degreaser emissions control test report.
7. Irvin, R., GCA/Technology Division. Trip report to
Autosonics, Norristown, Pennsylvania. Prepared for
Jones, L., U. S. Environmental Protection Agency.
June 15, 1979.
8. Letter and attachments from Polhamus, R. L., Branson
Ultrasonics Corporation, to Camroer, P. A., Halogenated
Solvent Industry Alliance. February 10, 1988. Automated
hoist test data.
9. Test data from A. Romig, Allied Signal, to D. A. Beck,
U. S. Environmental Protection Agency. August ,11, 1988.
Effectiveness of the "Cold Trap" in Reducing the Solvent
Loss Rate from a Vapor Degreaser.
3-71
-------
10. Letter and attachments from Halbert, J., Delta Sonics, to
Beck, D. A., U. S. Environmental Protection Agency.
November 24, 1987.
11. Miller, S. J., Radian Corporation. Trip report to Delta
Sonics, Paramount, California. Prepared for Beck, D. A.,
U. S. Environmental Protection Agency. January 1989.
12. Letter from Barber, J. W., Vic Manufacturing Company, to
Shumaker, J. L., U. S. Environmental Protection Agency.
July 8, 1977.
13. Letter and attachments from Arbesmah, P. H., New Jersey
Department of Environmental Protection, to Goodwin,
D. R., U. S. Environmental Protection Agency, August 21,
1978. -Capture efficiencies of lip exhaust systems.
14. Telecon. Miller, S. J., Radian Corporation, with Franz,
O., Phillips Manufacturing, July 27, 1987.
15. Letter and attachments from Stanley, H., Unique
Industries to Farmer, J. R., U. S. Environmental
Protection Agency. April 15, 1987. Section 114 organic
solvent cleaner vendor questionnaire response.
16. Letter from Ward, R. B., DuPont to Rehm, R.,
GCA/Technology Division. May 3, 1979. Controlling OTVC
emissions with carbon adsorption.
17. Letter and attachments from Cammer, P. A., Halogenated
Solvents Industry Alliance, to Farmer, J. R., U. S.
Environmental Protection Agency. June 7, 1989.
Comments at National Air Pollution Control Techniques
Advisory Committee meeting on draft control techniques
document.
18. S&K Products International Inc. Ultrasonic Cleaning and
Drying Systems. Product Brochure, 1990.
19. Miller, S. J., Radian Corporation. Trip report to Unique
Industries, Sun Valley, California. Prepared for Beck,
D. A., U. S. Environmental Protection Agency.
January 1989.
20. Letter and attachments from Sherman, James M., Scanex,
Inc. to S. R. Wyatt, U. S. Environmental Protection
Agency. December 29, 1988. 4 pages. Comments on issue
of solvent release from open top solvent cleaning tanks.
/
21. Telecon. Miller, S. J., Radian Corporation, with
Mulcahey, M., Rhode Island Division of Air and Hazardous
Materials. October 5, 1988.
3-72
-------
22. Telecon. Miller, S. J., Radian Corporation, with
Pennington, C., Detrex Corporation. November 24, 1987.
23. Summary of Meeting between Corpane and U. S.
Environmental Protection Agency. May 20, 1987.
24. Telecon. Miller, S. J., Radian Corporation, with Ramsey,
R., DuPont, November 23, 1987.
25. Memorandum from Brown, J. W. and P. R. West1in, EPA/EMB,
to Beck, D. A., U. S. Environmental Protection Agency.
May 22, 1981. Effect of Water Blanket to Reduce Organic
Evaporation Rates.
26. Wolf, K. HCFC's: Are They Still Viable in Solvent
Applications? Presented at the Second International
Workshop on Solvent Substitution. Phoenix, Arizona,
December 11, 1991.
27. Industry Cooperative for Ozone Layer Protection (ICOLP).
Alternatives for CFC-113 and Methyl Chloroform in Metal
Cleaning. U.S. Environmental Protection Agency, Office of
Air and Radiation. Publication No. EPA/400/1-91/019.
28. Industry Cooperative for Ozone Layer Protection (ICOLP).
Aqueous and Semi-Aqueous Alternatives for CFC-113 and
Methyl Chloroform Cleaning of Printed Circuit Board
Assemblies. U.S. Environmental Protection Agency, Office
of Air and Radiation. Publication No. EPA/400/1-91/016.
29. Industry Cooperative for Ozone Layer Protection (ICOLP).
Eliminating CFC-113 and Methyl Chloroform in Precision
Cleaning Operations. U.S. Environmental Protection
Agency, Office of Air and Radiation. Publication No.
EPA/400/1-91/018.
30. United Nations Environment Program. Electronics,
Degreasing and Dry Cleaning Solvents Technical Options
Report. June 30, 1989.
3-73
-------
4.0 DESCRIPTION OF MODEL CLEANERS AND THE
SOLVENT CLEANER POPULATION
4.1 MODEL CLEANERS
In order to evaluate the impacts of the various control
options on the solvent cleaner source category, model cleaners
were developed. Model cleaners were necessary due to the
large diversity of solvent cleaner sizes and designs. Model
cleaners were chosen instead of model plants because solvent
cleaning is a facet of various industries rather than an
industry with its own plants.
To represent halogenated solvent cleaning operations,
five model cleaners have been selected: four batch vapor
solvent cleaners (all open-top vapor cleaners [OTVC's]) and
one in-line cleaner. For each model cleaner the following
parameters have been specified: size (solvent/air interface),
freeboard ratio (FBR) (freeboard height/smaller interior
dimension of cleaner), and any other controls at baseline.
Size selection was based on the range of sizes reported in
vendor responses to the EPA questionnaires and those used by
the EPA in previous regulatory work under section 111 of the
Clean Air Act (Act).1
Each size model cleaner represents a range of similar
cleaners described by vendors; therefore, the models are not
intended to represent any vendor's actual cleaner. The model
cleaner parameters, as well as the operating schedule,
emissions from uncontrolled cleaners and emission reductions
at various control levels are presented in tables 4.1 through
4.5.
4-1
-------
g
CLEAN!
06
§
*^
I1
2
•J S
4 p.
W 04
S§
«s
p
S U
gg
w u
5 cu
u
w
a
o
x
i
TABLE
,,
i
1
u
**
t
|
M
- §
M
m S <=! 8 «9 2 •* m^mm M
• |S«vjH * * «- IA >»>»-» rj S
^ £mji ••• •••• M
* —•"<« e»-o ooo —
^ S
- 1
* 8
^
K» O
o- I§1 52 5 3 3 5 2 *• s
•
- I
w
«. 8
M- s 1 ? «• 1
1: . 8 § -
w 3 «T 2 S
**t ** ^ It
w | 1^ 1 g .
• xx TJ^ TJ .2 ^
• >«9 n 1 ^ ^ 9 • f ^ B • o —
• C » C — O — B C — OM BC —
1 1| i I I jji i i I || i i I ||
S
•
o
e
d
e
«
«
«J
fM
in
•
o
e
(5
o
g
d
g
»—
I
«V
^
t
1
J
:
1
I
|
fe
d
d
S
d
d
ft
d
*
e
S
d
y
a
d
d
I
W
|
w
1
S
s
z
4-2
-------
1
1
1
X >
HQ
0 &fi
Si
H
S0£
W
aea,
a M
fa o
W S
SK
IE M
04 H
||
2 o
o
.j
u
o
0
CM
1
TABLE 4
m
•»•
£
I
•£
2
i
I
- S
- §
- i
2 1 M * 3 * s * « * ~ §
• * «- ••» o ^eeiM
- §
- i
-1
- i
0 OeiO O**OW w"*»jO«" ^f
i ruS i/\in^ O^w^m N&
e S "> S o«-e ooora M
^ o
"*
2 :
* tf ^ - •
•i -. * s 8 2
^ 5 f : s
tf i «. i s :
' , I f t j 11 1 * i li 1 1 j u
»**— C — *rf^%^C**"*'<-^*C'^*' ^*«
C C ^ *** ^ C ^* "** ^^ C ^ *•* S ^ o 5 Sx ? o 5 S x v e 1 it
$5&~ °^" °ll" ° II
e
?
3
•
5
8
«
e
S
"
o
rg
*«
i
8
8
1
|
^
k
I
3
S
g
Q
0
!R
e
5
(G
S
*
1C
•
e
g
o
|
on reduction. Me/yr i
«••
8
2
4-3
-------
i
i
^^
u
K
>
g<
UQ
<*
°2
w °
1*4 9^
0 *
53 a;
^ w
J£
gw
2g
cog
K *
u v
H5
|u
2 «
2 z
"*< ^^
fl* S
H
w §
3 o<
3 ^
J
0
i-3
U
O
o
X
n
1
^
U
1
<
M
•
*J
•^
a
JS
M
•^
r4
Ot
u
u
~4
b
«
V
i
V
«
0.
-
O O O O O 00
VO 01 VO 00 * * *
t-t in in vo • • •
^ to oi to o
•n O O O Ot i-l 01
"? 04 tO 00 O Ot 04
_1 in in vo . . •
•^ r4 tO «-t O4 O
H
w *
** . 5
M «
• X *J
w ^. «
<-> U U M •
M H X! •
4J g C
•w » 0 O
*• •>! » •** ~*
3 • • •
01 P -5 • •
' « r 1 1 1
« * JS —
18 1^ 1
S « ^^ t *«
-;c"o.?l^3o.c"1^
O> 4J -^ C -^ 4J U " C -^ ** »4
C C 4J -4 ^ C 4J -4 J< C 4J
^STdrHuScu^hie
X>t4'OOOOXi'OOOO
Sl|,M*0^M*0S
» W 0 3^ 9
+
m
n at O 01
vo at to 04 „
• • • • "
O ^ «-i r-
iH
^F
n
r~ o vo vo
in in «v in -,
• • • •
O * «-l vO
_j
!-|
g
^
— JJ/
£ Ot I §<
*J O> C -3
" C -H 4J fH
•^ j< e o
U r-l U S U
x -a o 5 v
I H " ' I
r»
O
vO
s
in
o
*
0
o
t*
o
vO
O
m
j»
O^
qf
«
*
.-i
*4
*H
O
Contn
r«
t-i
•
01
o>
00
•
01
•H
«o
•
m
m
rt
•
+
r»
o>
•
^H
C^
SO
•
OJ
00
01
fl
m
o
•
m
i
41
1<
>,
•^
$
*
•
0
•
•*
S
rH
•-4
O
Contn
in
0
•
in
n
m
•
*
lH
VO
•
r*
at
00
•
01
at
m
•
*
f*>
a>
•
n
00
(N
n
01
to
•
01
X
2
•w
*"'
Vi
>,
1
%
§
M
«
g
!
14
o
•*4
a
a
-H
U
4-4
-------
1
3
u
«
£
£
|j ^
CO O
H K
O 2
2 2
W
>< W
g Ot
> W
is
to x
U (0
u u
5 "*
II
W
« Cw
u o
J
u
u
o
o
&
*
1
•#
TABLE
.
c
D
£
-H
«H
t?
H
U
°£
4J
Q)
14
*
1
01
_ o o o S *
CD 2 vo eo i* *: «H
m S "» * • in •
•" fH vo in J5 tH
u, O O O O> O CM
. CN vo oo in o> in
_: in in vo . . .
m fH VO (M VO O
Ok •
14 9
W "^ *
1 H * e c
»H • •
8 o I 1 1
5 * S ll 3
«~o> o>B»^ja O'S'i
• e o U -O 0 0 0£ *O O 0 0
b >H 0 M X Q U •*•» M » Q O
OOP* c o> • e
» M 0^ D^ 3
,
n
•*s2l "
rH
«
«
m p vo jj:
^2*2 N
fH
•H
5
o
§
i . H *'
«— C -^ 4J -H
**4 Jrf C* 0
«U r*4 14 5 14
x -o o 5 v
-*. M » o e
w* O
X U
o *^
** in
t» «O
0 CO
* VO
o M
in ao
* Ch
0 CM
d
^ f*
o ,
** *
o 71
- S
o ^
in ^
^
o ***
1
M
^J
O*
X
»
•
e
3
•
i 1
1 1
rH
fH fH
80
h
C C
o o
O t.)
§
fH
S
d
in
00
00
o
C*
*
o
•
CO
n
CM'
VO
S
V
jj^
>^
X
•k
o
on reduct
-i
B
1
4-5
-------
CO
fs
w
i
u
u
z
•J
1
z
M
K
o
b.
U
g
M
2
fH
cu
a
u
u
u
o
o
X
,
in
i
*
s
GO
<
£
4J
e
D
«4
kl
0.
OO VO ^ ^" O 9* *"4 S
CO . a CD CO VO f t-l O O CD w
*•> o vo • • • • ••***
CMVO CMCOrH O O> <*> JjJ
in OO «-* f- CM O JC vO H?
. Qcoco Maim o."*^
' ° O vO ... . • . •
CMVD i-tfoo o^n^
H
BU •"
kl «
g*>. «
kl kl + **
J3 • 4
*i e e 4J
«M O O O
— < . -4 ^ (4
CM H ^ • •
• si i i. i
2 • 9 A « —
kl « -H^. f .H >l f
««O> 0» E -^ J3 O» 1 -H Ol O> B
• c 01 e -^ 0^4 o» e -3 o ** o< c -5
O> 4J -^ C •* *< H -» C -H 4J U «- C -< *»
-»e«^wi3cu^k)|cki^kif
j<>kioooo>i'aoo
u ^ a M x a o^ M » o o^ M * o
O O Cu C O* C O*
» (o a a J< ax
* o
(*V O
VO
CM O
in
^
»H 0
^
* 0
n o
vO
CM O
in
t-l O
e
5 1!
it ^
o *^~
iH r-l
0 0
kl kl
4> 4»
e e
8 8
CO
VO
«H
•
in
in
VO
^
r~
•
**
r4
in
n
CO
vO
^
VO
CO
in
o
r*
^
a
ii.
h
0*
a
e
3
•
1
F-l
O
Contr
o
^
*
r-
in
VO
vO
•
in
o
CM
CO
n
0
.
r-
vO
CO
in
CO
VO
*
—
a
*i
Jj,
1
%
ction
I
§
a
a
i
4-6
-------
It is known that many cleaners exist outside the
established range of model cleaners. For example, batch
cleaners significantly larger than the largest model batch
cleaner are known to exist in some industries, such as in the
aerospace industry. Also, some cleaners have recently been
developed that have no solvent vapor/air interface.
At baseline, there is a mixture of controlled and
uncontrolled cleaners, due largely to State implementation
plan (SIP) regulations that require control of solvent
cleaners in certain States and counties (as a result of the
1977 control technique guidelines [CTG]). In general, SIP's
are estimated to control the emissions from a model solvent
cleaner by approximately 40 percent.
The model cleaner emission rates are meant to represent
typical averages and would include emissions through various
pathways, including diffusion and convection during the
idling, working and downtime modes, and carry-out of solvent
on parts during the working mode. It is recognized that
emission rates, particularly during the working mode, can vary
widely. A discussion of each model, and its associated
operating schedule, is presented below.
4.1.1 Model Batch Cleaners
Four model OTVC's that differ only by size have been
selected to represent batch vapor cleaners. At baseline, all
four have an FBR of 0.75, a manual cover used during the
downtime mode, and all the appropriate safety switches
described in chapter 3.0. The model cleaner sizes (solvent
vapor/air interface) are summarized in table 4-6.
The operating schedule selected for small and medium
model batch cleaners is 2 .hours per day (hr/d) working
(i.e., machine actively cleaning parts), 6 hr/d idling
(i.e., machine turned on, but not cleaning), and 16 ftr/d down
(i.e., machine turned off) for 260 days per year (d/yr); the
machine is down 24 hr/d for the remaining 105 days of the
year. The operating schedule for large and very large batch
cleaners is 6 hr/d working, 2 hr/d idling and 16 hr/d down for
4-7
-------
TABLE 4-6. MODEL CLEANER SIZES
Model cleaner Size range (m2) Selected size (a2)
Small OTVC <0.6 0.4
Medium OTVC 0.6 to 1.21 0.8
Large OTVC 1.21 to 2.51 1.5
Very large OTVC >2.51 3.5
In-line All sizes 3.5
4-8
-------
260 d/yr; the machine is down 24 hr/d for the remaining
105 days of the year. Although it is recognized that
operating schedules may vary considerably throughout the
industry, the operating schedules selected are representative
of the average operating schedule for each size cleaner.
4.1.2 Model In-Line Cleaners
The solvent vapor/air interface of the model in-line
cleaner is 3.5 square meters (m2) [38 square feet (ft2)] The
model in-line cleaner has an automated parts handling system,
an enclosed design, and all major safety switches. Although a
freeboard area is present in all in-line machines, no specific
FBR is assumed in the baseline of these model in-line
cleaners. The model in-line cleaner is assumed to have no
covers on the entrance or exit for use during downtime.
The operating schedule for in-line cleaners is 8 hr/d
working and 16 hr/d downtime for 260 d/yr; the machine is down
24 hr/d for the remaining 105 days of the year. No idling
time is included for in-line cleaners since, if the machine
were not to be used constantly during a working shift, a less
costly batch cleaner would likely be used.
4.2 SOLVENT CLEANER CONTROL COMBINATIONS
Descriptions of the various control techniques and their
individual control efficiencies are presented in chapter 3.0.
Each control device was assigned a control efficiency for the
three solvent cleaner operating modes (idling, working, and
downtime), as appropriate.2 The control efficiencies of the
various control techniques are summarized in table 4-7.
In many instances, multiple control techniques are used
on a single cleaner. In order to evaluate the range of
possible emission reductions, control options representing
reasonable groupings of up to four control techniques were
developed. Control combinations containing redundant controls
(i.e., control the same component of emissions) were not
included. An example of redundant controls would be a
super-cooled primary condenser and a freeboard refrigeration
device.
4-9
-------
TABLE 4-7.
SOLVENT VAPOR EMISSION CONTROL EFFICIENCIES
FOR VARIOUS CONTROL TECHNIQUES
Cleaner
In-line
Cleaners
Control technique
Control efficiency*
Batch Vapor
Cleaners -
Cover
Biparting Cover
FBR 0.75— >1.0
FBR 1.0 — >1.25
Freeboard
Idling
40
40
20
10
40
Working
0
40
20
10
40
refrigeration
device
Primary condenser 40
temperature (3 0
to 40 percent of
the solvent
boiling point)
Reduce wind speed 50
30.3 a/min
(100 ft/min) —
>calm (15.2 m/min
[<50 ft/min])
Hoist 0
Dwell 0
Freeboard 60
refrigeration
device
Carbon adsorption 60
system
40
50
35
30
60
60
aControl-efficiency over a typical cleaner.
Typical batch vapor cleaner: OTVC with a 0.75 FBR,
circumferential water-cooled primary condensing coils, a
manual cover, and located in a room with windspeeds in
excess of 30.3 m/min (100 ft/min).
Typical in-line cleaner (vapor and cold): A typical
in-line cleaner has water-cooled condenser coils.
4-10
-------
Using the combined control efficiency formulas (see
appendix A for the derivation), mode-specific control
efficiencies were developed for each option. The overall
emission reduction for each control combination was determined
by combining mode-specific control efficiencies with the
percent of total emissions attributed to each mode.
Under the operating schedule for a large or very large
batch vapor cleaner, 69 percent of the emissions from the
cleaner occur during working, 22 percent during downtime, and
9 percent during idling. If a control combination controlled
working emissions by 60 percent, idling emissions by 40
percent, and downtime emissions by 30 percent, the overall
control efficiency would be:
0.69*0.6 + 0.22*0.40 + 0.09*0.30 - 50 percent control
The mode-specific control efficiency for a control
combination is independent of both the cleaner size and the
operating schedule. However, while the percent of the
emissions attributable to the various operating modes is
independent of the size of cleaner, it can be highly dependent
on the operating schedule. For example, if the same mode-
specific control efficiencies presented in the previous
example remained constant, but the schedule were changed from
6 hr/d working to 2 hr/d working, the overall control
efficiency would be only 40 percent. •
The magnitude of the impact of operating schedule on
overall control efficiency is dependent on the relative
differences between the mode-specific control efficiencies;
the greater the difference, the higher the impact. For
example, if the mode-specific control efficiencies had been
70 percent, 30 percent, and 20 percent for working, idling,
and downtime, respectively, the overall control efficiency
4-11
-------
would be 60 percent for a 6-hr working schedule and 40 percent
for a 2-hr working schedule.
The control combinations evaluated for existing and new
batch vapor cleaners are presented in tables 4-8 and 4-9,
respectively; tables 4-10 and 4-11 present the combinations
for existing and new in-line cleaners, respectively. The
combinations are grouped by overall control efficiency.
4.3 NATIONAL ESTIMATES OF SOLVENT CONSUMPTION
AND NUMBER OF CLEANERS
As part of the regulatory development activities,
estimates have been made of the national annual solvent
consumption for this category and the national number of
solvent cleaners.3/4 This section describes the methods and
information used to develop these estimates. Section 4.3.1
presents baseline estimates of national hazardous air
pollutant (HAP) solvent consumption for the solvent cleaning
source category for methylene chloride (MC), perchloroethylene
(PCE), trichloroethylene (TCE), and trichloroethane (TCA).
Section 4.3.2 presents the estimates of the national number of
solvent cleaners. It should be noted that TCA is included in
both analyses.
4.3.1 National Solvent Use Estimates
The estimates of national HAP solvent consumption by the
solvent cleaning industry were derived from estimates of the
total HAP solvent consumed by all industries in 1991.3
Table 4-12 presents the estimates of HAP solvent consumption
by the solvent cleaning industry, and a breakdown of HAP
solvent consumption within cold cleaning and vapor degreasing.
The HAP solvent consumption data from which the estimates
were derived were obtained from the Chemical Marketing
Reporter's (CMR) Chemical Profiles for MC, PCE, TCA,, and TCE.
Chemical Marketing Reporter was chosen as the source for
solvent consumption information because their estimates of HAP
solvents used in the solvent cleaning industry were the most
recent and included all four HAP solvents.3
4-12
-------
TABLE 4-8. EXISTING BATCH VAPOR CLEANER CONTROL COMBINATIONS
Control •ffici«ncv <\\
R«f«r«nc* . 6-hr working
number Control combination* cchedul*
26
25
S3
56
23
24
22
18
9
20
49
52
6
48
54
17
50
51
14
7
2
10
11
55
19
15
16
1
8
21
57
Hill,
Hill,
Hill,
Hill,
Hill,
Hen,
neii,
Hiii,
H811,
neii,
Hill,
Hill,
nen,
H@ll,
H811,
H@ii,
Hen,
H611,
neii,
Hill,
H611,
Hen,
Hen,
Hen,
- H811,
H811,
H811,
H811,
H811,
Hill,
Hen,
FRD, RRD, DHL
FRD, RRD, AC
AC, RRD, DHL
RRD, AC
AC, FRD, l.OFBR
FRD, RRD, l.OFBR
AC, DHL, l.OFBR
AC, FRD, DHL
FRD, RRD
MC, FRD, DHL
AC, RRD, l.OFBR
FRD, RRD, MC
AC, FRD
RRD, DHL, l.OFBR
RRD, MC, DHL
RRD, DHL
MC, RRD, l.OFBR
DHL, FRD, l.OFBR
l.OFBR, RRD
AC, l.OFBR
FRD
FRD, DHL
MC, FRD
RRD
MC, FRD, l.OFBR
FRD, l.OFBR
l.OFBR, DHL
AC
AC, DHL
MC, l.OFBR, DHL
RRD MC
70
70
70
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
50
50
50
50
50
50
50
50
50
50
50
50
2-hr working
•ch«dul«
50
60
50
50
50
60
40
50
50
50
60
60
50
50
50
40
50
40
50
40
30
40
40
40
50
40
30
' 30
40
40
50
4-13
-------
TABLE 4-8.
EXISTING BATCH CLEANER CLEANER CONTROL COMBINATIONS
(CONCLUDED)
Control efficiencv (\\
Reference
number
13
12
4
5
3
58
Control combination*
H811,
Hill,
H611,
H811,
H611,
H@ll
MC, DHL
MC, 1.0F8R
DWL
MC
l.OFBR
6-hr working
schedule
40
40
40
30
30
-—
.2-hr working
schedule
30
30
20
30
20
10
KEY:
FRO » Freeboard Refrigeration Device
RRD * Reduced Room Draft
H811 - Manual Hoi»t Operating at 11 fpm
DWL » Dwell Time
1.0 FBR • Freeboard Ratio of 1.0
AC * Automated Cover
MC • Manual Cover (Manual cover in on during idling and down time)
4-14
-------
TABLE 4-9. NEW BATCH VAPOR CLEANER CONTROL COMBINATIONS
Control •ffieienev f%)
R*f«r«nc« 6-hr working
nunb*r Control combination* sch*dul«
34
43
25
53
44
26
41
40
51
14
6
45
23
46
38
56
35
54
42
32
24
49
48
20
28
31
22
39
17
18
He 11,
neii,
ne 11,
Hen,
Hen,
Hen,
Hen,
Hen,
Hen,
Hen,
Hen,
Hen,
H611,
H@ll,
Hen,
H611,
neii,
Hill,
H611,
K811,
Hill,
Hill,
Bin,
Hiii,
Hiii,
Hiii,
Hen,
Hill,
Hill,
Hill,
AC, FRD, SHV
RRD, OWL, SHV
AC, RRD, FRD
AC, RRD, OWL
RRD, AC, SHV
FRD, RRD, DWL
FRD, RRD, SHV
RRD, l.OFBR, SHV
DWL, FRD, l.OFBR
l.OFBR, RRD
AC, FRD
RRD, MC, SHV
AC, FRD, l.OFBR
AC, DWL, SHV
FRD, l.OFBR, SHV
RRD, AC
MC, FRD, SHV
RRD, MC, DWL
DWL, FRD, SHV
AC, SHV
FRD, RRD, l.OFBR
AC, RRD, 1.0 FBR
RRD, DWL, l.OFBR
MC, FRD, DWL
FRD, SHV
RRD, SHV
AC, DWL, l.OFBR
DWL, l.OFBR, SHV
RRD, DWL
AC, FRD, DWL
70
70
70
70
70
70
70
70
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
2-hr working
•ch«dul«
SO
SO
60
50
SO
50
50
50
40
50
SO
50
50
40
SO
SO
50
50
40
40
60
60
SO
50
40
SO
40
, 30
40
50
4-15
-------
TABLE 4-9.
NEW BATCH VAPOR CLEANER CONTROL COMBINATIONS
(CONTINUED)
Control *ffici«ncv /%»
R«f«r«nc« 6-hr working
numb«r Control combination* ichadul*
47
52
37
36
9
50
19
30
2
15
16
10
1
55
33
21
29
7
8
27
11
57
12
4
13
5
3
58
Hen,
Hen,
neii.
neii.
Hen,
H611,
neii,
Hen,
H@ll,
Hen,
Hen,
Hen,
neii,
Hen,
Hen,
H@ll,
Hen,
Hen,
Hen,
H611,
Hen,
Hen,
Hen,
Hen,
* Hen,
Hen,
Hen,
Hen
MC, OWL, SHV
FRD, RRD, MC
MC, l.OFBR, SHV
AC, l.OFBR, SHV
FRO, RRD
RRD, MC, l.OFBR
MC, FRD, l.OFBR
DWL, SHV
FRD
FRD, l.OFBR
l.OFBR, DWL
FRD, DWL
AC
RRD
MC, SHV
MC, l.OFBR, DWL
l.OFBR, SHV
AC, l.OFBR
AC, DWL
SHV
MC, FRD
RRD MC
MC, l.OFBR
DWL
MC, DWL
MC
l.OFBR
60
60
60
60
60
60
50
50
50
50
SO
50
50
50
50
50
50
50
50
50
50
50
40
40
40
30
30
— —
2-hr working
•ch*dul«
40
60
40
50
50
50
50
30
30
40
30
40
30
40
40
40
30
40
40
20
40
50
30
20
30
30
y20
10
4-16
-------
TABLE 4-9. NEW BATCH VAPOR CLEANER CONTROL COMBINATIONS
(CONCLUDED)
KEY t
FRO - Freeboard Refrigeration Device
RRD • Reduced Room Draft
SHV * Super Heated Vapor
OWL - Dwell Tin*
1.0 FBR - Freeboard Ratio of 1.0
AC - Automated Cover
MC • Manual Cover (Manual cover in on during idling and down tine)
H011 * Manual Hoist Operating at 11 fpn
4-17
-------
TABLE 4-10. CONTROL EFFICIENCIES FOR EXISTING IN-LINE
SINGLE AND COMBINED CONTROLS
Control
efficiency (%)
Reference
number Control combinations
14
6
9
1
11
8
2
l.OFBR, FRD, DWL
l.OFBR, FRD
DWL, FRD
FRD
DWL, l.OFBR
DWL
l.OFBR
8-hr working
schedule
50
50
50
40
30
20
10
KEY:
1.0 FBR = Freeboard Ratio of 1.0
FRD * Freeboard Refrigeration Device
DWL - Dwell Time
4-18
-------
TABLE 4-11.
CONTROL EFFICIENCIES FOR NEW IN-LINE
SINGLE AND COMBINED CONTROLS
Reference
number
15
7
12
4
13
5
10
14
6
9
3
1
11
8
2
Control combinations
SHV, l.OFBR, FRD, DWL
SHV, l.OFBR, FRD
DWL, FRD, SHV
FRD, SHV
DWL, l.OFBR, SHV
l.OFBR, SHV
DWL, SHV
l.OFBR, FRD, DWL
l.OFBR, FRD
DWL, FRD
SHV
FRD
DWL, l.OFBR
DWL
1 . OFBR
Control
efficiencey
(%)
8-hr working
schedule
70
60
60
60
60
50
50
50
50
50
40
40
30
20
10
KEY:
SHV - Super Heated Vapor
1.0 FBR • Freeboard Ratio of 1.0
FRD - Freeboard Refrigeration Device
DWL - Dwell Time
4-19
-------
M h
D *
§
-H
C
0
0
a
u
•gg
ss
(0 ^
u
-1 C
0 V
JB a
a.
0 «
-u ^"
TJ 0 «
-I U C
« u
°
5 *
V4
j= 0
u e
4J 0
0 0
a ^
u
s
1
J5 0
o
o
P4
0
o
O
o
ON
§
0
o
O
o
O
o
in
O O
O O
in
o
0
in
e* in
O4 9>
§
0
o
O
o
•9
o
o
o
o
o
0
00
o
0
r-
o
o
VO
O
o
o o
o o
S
u
o
o
vO
1
o
«
n
(M
o o
o o
r» in
in
i g
m t-
M
0 0
0 « B
-H-^ O>
0 s «w e
4J O 0 -4
eo
** 4J a
u e a a
000-^
I W-
Vi U
o»u c
X O 0 0
«M U 4J
« U
m M **
2583
0 c o oo
i' «0
M • 0 O
h TJ
o c u .p
-------
2J§£
isS5
0 4J
0 «
W '^l'^ «J M
•-4 O>JB U HJ
-4X1 0
O O B "0 0
a o a 4J
>H a 0 a
e «. JG I
0 VO TJ 4J-4
u 0
h Lt « 2
c .c
8*J U
« T3 0
e 4t
TJ s o e
c CTJ
-4M «, -a h
a an cut o
S « o — 5co
CM u a >,« ~*
»-*.«-! TJ • a •
M .C O O>9 0 * O <~
W U >H X O *» CO I 14
a, *j^ -* «eo o x
a o o a ~4 oa>*. »
p. g • ~—.^ | . 1*1 u
vo c « w «i * o 0
-4 4JKUUOO O C
BCO.IHOV X«
0 0 O _ ,
-4 LI u — ao
4J 0 JB 3 L) « O
« o 4J Lf«>*i » e
rfhw *V f^ ^*» i
aa uxw^i
52
4-21
-------
The data estimates presented in table 4-12 include HAP
solvent consumption for use in electronics (i.e., photoresist
stripping, photoresist development, and printed circuit board
assembly/defluxing), metal cleaning/degreasing, vapor degreasing,
vapor degreasing of fabricated parts, and cold cleaning. As
mentioned previously, no evidence can be found of batch cold
cleaners (except carburetor cleaners) using any of the four HAP.
It has been concluded, therefore, that the batch cold cleaning
HAP consumption estimate presented in table 4-12 includes other
cold cleaning operations, such as wipe cleaning.
The 1991 solvent consumption estimate presented here is
33 percent less than the 1987 estimate (296,000 megagrams [Mg])
[326,000 tons] that was extrapolated from 1983 data. The largest
consumption reduction within the solvent cleaning industry was
for PCE (53 percent), followed by MC (47 percent), TCA (28
percent), and TCE (24 percent). Qualitative explanations for
these HAP solvent consumption reductions include recycling,
escalating excise taxes, future mandated production phase-outs
for TCA, and environmental and occupational health concerns
surrounding the use of PCE and MC.
4.3.2 Estimate of the National Number of Solvent Cleaners
Information on the location and number of solvent batch
solvent cleaners (primarily OTVC's) and in-line vapor cleaners
(ILVC's) is difficult to obtain because of the large number of
solvent cleaning operations existing within many different
industries. Therefore, a method to estimate the number of
cleaners nationwide was developed using the nationwide solvent
consumption estimates described in section 4.3.1, assumptions on
the breakdown of the national cleaner population (e.g., sizes of
cleaners, percent of controlled cleaners), control efficiency
estimates, and typical solvent consumption rates.4 The estimated
national number of cleaners using each solvent are presented in
table 4-13.
4-22
-------
Table 4-13. KUMBER OF HALOGENATED SOLVENT CLEANERS
Model cleaner
size
BATCH CLEANERS
Small
Medium
Large
Very Large
IN-LINE CLEANERS
Solvent/
vapor/air
interface
m2 (ft2)
0.4 (4.5)
0.8 (8.6)
1.5 (16)
3.5 (38)
3.5 (38)
Solvent
MC
TCE
PCE
TCA
Total
MC
TCE
PCE
TCA
Total
MC
TCE
PCE
TCA
Total
MC
TCE
PCE
TCA
Total
Total Batch:
MC
TCE
PCE
TCA
Total ILC:
Grand Total:
Estimated
number of
cleaners
150
1,270
360
2,320
4,100
180
1,530
440
2,780
4,930
150
1,270
360
2,320
4,100
120
1,020
290
1,860
3,290
16,420
880
1990
620
4,590
8,080
24,500
MC - methylene chloride
TCE * trichloroethylene
PCE « perchloroethylene
TCA « 1,1,1-trichloroethane
ILC * In-line cleaner
4-23
-------
The model cleaners and operating schedules presented in
section 4.1 of this report, and the typical cleaner emission
rates presented in chapter 3.0, were used for this analysis. It
was assumed that the percentage of machines with controls sold in
the last decade is representative of the current percent of
solvent cleaners controlled nationally. These data were obtained
from the 1991 section 114 survey of equipment vendors. The
percentage of controlled versus uncontrolled estimates for all
representative OTVC sizes and ILVC's were obtained from the sales
figures reported in the same questionnaire responses. An average
of 40 percent emission reduction over uncontrolled cleaners was
assigned to all controlled cleaners.
Because of the substantial nationwide reduction in solvent
use, the sales data from the 1991 survey do not necessarily
represent the actual number of cleaners still in use. Therefore,
the percentage of batch cleaners in each size range is from
estimates reported in the 1987 vendor surveys, where each vendor
estimated national distribution for all vendors. The emission
factor estimates used in this analysis were developed during
previous regulatory efforts for this industry.
For the ILVC, the annual emissions per surface area were
divided by the emission factor (with.recycle) to determine the
annual amount of solvent consumed per surface area. This was
then multiplied by the model surface area to determine the annual
amount of solvent consumed. The annual amount of solvent
consumed was calculated for both controlled and uncontrolled
cleaners. The fraction of cleaners controlled and uncontrolled
were multiplied by their respective annual solvent consumptions
and summed to obtain a weighted average of the annual amount of
solvent consumed by an ILVC. Finally, the average ILVC solvent
consumption was divided into the estimated amount of solvent
consumed nationwide by all ILVC's (table 4-12) to d. armine -.he
number of ILVC's in operation nationwide.
4-24
-------
The sane basic approach was used for OTVC's. The average
annual amount of solvent consumed by each of the four model
OTVC's was calculated by the same method described above. The
number of cleaners in each range of OTVC sizes was then
determined using the relative percentages determined from the
1987 questionnaires. The percentage of batch vapor cleaners in
each range size is: 25 percent small, 30 percent medium, 25
percent large, and 20 percent very large. The number of small
OTVC's ("I") was then determined using the following relation:
Total national OTVC - I*A + 1.2*I*B + I* C + 0.8*I*D
solvent consumption
or I « Total national OTVC solvent consumption
A + 1.2*B + C + 0.8*D
where A, B, C, and D are the respective average annual solvent
consumptions for the four model size cleaners. From the number
of small OTVC's, the number of cleaners in the remaining ranges
were calculated.
4.4 REFERENCES
1. Memorandum and attachments from Gerald, L. and Falling, A.,
Radian Corporation, to Almod6var, P., EPA/CPB. October 27,
1992. Compilation of information obtained from Solvent
Vapor and Cold Cleaner Vendor Questionnaires.
2. Memorandum and attachments from O'Loughlin, J., Sarsony, C.,
and Miller, S., Radian Corporation to Almod6var, P.,
EPA/CPB. May 26, 1993. Summary of emissions reductions for
selected control techniques for halogenated solvent
cleaners.
3. Memorandum and attachments from O'Loughlin, J., Radian
Corporation, to Almod6var, P., EPA/CPB. July 2f 1992.
Estimates of national solvent consumption by the solvent
cleaning industry.
4. Memorandum and attachments from O'Loughlin, J., Radian
Corporation to Almod6var, P., EPA/CPB. May 26, 1993.
Estimate of the number of halogenated solvent cleaners.
4-25
-------
5.0 REGULATORY APPROACH
5.1 MAXIMUM ACHIEVABLE CONTROL TECHNOLOGY (MACT) FLOOR
Emission standards for new and existing sources promulgated
under section 112(d) of the Act must represent the maximum degree
of reduction achievable; this is typically referred to as the
MACT. The Act establishes minimum levels, or "floors," for MACT
standards. These floors are to be determined as follows:
(1) for new sources, the MACT floor cannot be "less
stringent than the emission control that is achieved in
practice by the best controlled similar source...."
(2) for existing sources with more than 30 sources, the
MACT floor cannot be less stringent than "the average
emission limitation achieved by the best performing 12
percent of the existing sources...."
The MACT floors for the halogenated solvent cleaning machine
source batch vapor and in-line subcategories are based on control
efficiency and sales data obtained from section 114
questionnaires sent to solvent cleaning machine vendors.1 In
these section 114 questionnaires data were gathered for cleaning
machines in each subcategory. These data are assumed to be
representative of the control levels achieved by the industry as
a whole.
Using the control efficiencies for the individual control
devices, the combined control efficiency formula, and the model
cleaning machine operating schedules, a control efficiency was
calculated for each reported cleaning machine. The control
combination efficiencies were rounded to the nearest 10 percent
5-1
-------
increment (i.e., 10 percent control, 20 percent control, etc.) to
reflect the precision of the data. The combinations were then
grouped by their combined control efficiencies. All the cleaning
machines were grouped based on control efficiencies and ranked
from the highest control efficiency combination to the lowest
control efficiency combination. The floor for existing sources
in each subcategory was then determined by calculating the
weighted-average level of control for the top performing
12 percent. The MACT floors, for batch vapor cleaning machine
existing sources are as follows: 10 percent for small,
40 percent for medium, 40 percent for large, 60 percent for very
large. The MACT floor is 30 percent for existing source in-line
cleaners. The MACT floor for new sources in each subcategory was
determined by determining the maximum control level achieved
(control level of the best-controlled existing source) for each
subcategory. The MACT floors for batch cleaning machine new
sources are as follows: 40 percent for small, 50 percent for
medium, 60 percent for large, 60 percent for very large. The
MACT fioui. is 40 percent for new source in-line cleaning
machines.
5.2 DEVELOPMENT OF ADDITIONAL REGULATORY ALTERNATIVES
Regulatory alternatives were developed for each of the five
solvent cleaning machine subcategories. The least stringent
regulatory alternative that was considered is the MACT floor;
therefore, the MACT floor is always presented as the first
alternative. To develop the regulatory alternatives, potential
control levels for each subcategory were developed and analyzed.
First, all reasonable control combinations were evaluated for
their emission reduction potential for each subcategory, based on
a typical operating schedule.. Then the control combinations were
grouped into control levels, rounded to the nearest 10 percent
increment. Next, the capital costs were evaluated and'the median
cost combination was determined for each control level.2 Median
costs were selected instead of average costs because the median
5-2
-------
costs have a particular control combination associated with each,
whereas an average cost may not correspond to any of the
available control combinations. Median costs were selected as
opposed to the lowest cost combination because some combinations
may not be feasible for all cleaners. For each subcategory, the
control combination cost that represented the median capital cost
was chosen to represent the control level.
Once the median cost control combinations were selected, the
costs of each control level were evaluated for each subcategory.
Tables 5-1 through 5-5 show the annualized costs for existing
cleaners using median costs, tables 5-6 through 5-10 show the
annualized costs for new cleaners using median costs. The total
annual cost for each control level was estimated from the
annualized capital costs, the annual operation costs; the
monitoring, reporting and recordkeeping costs, and the solvent
cost savings realized from reduced emissions.2 The cost per unit
of emissions reduction, or cost effectiveness, was then
determined for each control level and inferior control options
were identified. Inferior control options are defined as those
control levels with higher costs than other levels that achieve
the same or greater emission reductions. The remaining
non-inferior control options made up the set of regulatory
alternatives for each subcategory. Both the existing and the new
small batch cleaning machine control options at the MACT floor
levels were identified as inferior options; however, because they
are the MACT floor they are included as the first regulatory
alternatives for this subcategory. Table 5-11 presents the
regulatory alternatives for each subcategory.
The control cost analysis for the regulatory alternatives
indicates that a large number of solvent cleaning facilities will
have negative costs for moving from the estimated current level
of control to meeting each regulatory alternative. This implies
that many facilities have had the possibility of reducing their
5-3
-------
TABLE 5-1. ANNUALJZED COSTS FOR EXISTING SMALL BATCH CLEANERS (2 HOUR SCHEDULE)
Emission Reduction Levels
Description
1. Capital Costs
-hoist
- increased dwell
- 1.0 freeboard ratio
- automated cover
— reduced room draft
-manual cover
- freeboard refrigeration device
- additional floor space required
Total Capital Cost* (TCC)
Includes taxes, freight, & installation)
Annualized Total Capital Costs (15yr @ 10%)
2. Annual Operation Costs
-Labor
- Utilities (electricity)
- Miscellaneous Operating Expenses
- Reporting/Recordkeeping Expenses*
- Monitoring Expenses
Total Amualixed Cost
Level 1
10%
$2.000
...
— __
...
...
...
...
$421
$2.421
$318
$0
$75
$60
$279
$265
$1.017
Level 2
30%
$2,000
$0
$1,200
...
...
...
...
$421
$3,821
$476
$0
$75
$80
$279
$265
$1.175
Levels
40%
$2.000
$0
...
...
...
...
$4.700
$631
$7£81
$964
$206
$150
$268
$279
$315
$2.181
Level 4
50%
$2,000
$0
...
$4.500
$1.400
...
...
$1.709
19,009
$1.264
$0
$99
$260
$279
$315
$2*17
Levels
60%
$2,000
___
$1.200
...
$1.400
...
$4.700
$1.919
$11*19
$1.475
$205
$150
$268
$279
$364
$2.741
3. Cost Effectiveness
Emission Reduction (Mg/yr)
0.13
0.39
0.52
0.66
0.79
Recovered Solvent Credit ($A
- Methylene Chloride ($1.02
- Perchloroethylene ($1.11/1
- Trichloroethylene ($1 .43/k
-1,1,1 -Trichtoroethane ($2
Net Annualized Cost of Control (S/yr)
-Methylene Chloride
- Perchloroethylene
- Trichloroethylene
- 1,1,1 -Trichtoroethane
Cost of Control ($/Mg)
-Methylene Chloride
- Perchloroethylene
-Trichloroethylene
-1.1.1 -Trichtoroethane
Inremental Costs ($/Mg)b
-Methylene Chloride
- Perchloroethylene
- Trichtoroethytene
-1,1.1 -Trichtoroethane
($134) ($402) ($536)
($144) ($433) ($577)
($186) ($558) ($744)
($263) ($788) ($1.050)
$883 $773 $1.645
$873 $742 $1,604
$831 $617 $1.437
$755 $387 $1,131
$6.798 $1.863 $3,164
$6.716 $1.903 $3.064
$6.396 $1,583 $2,764
$5,806 $993 $2,174
INFERIOR $1,963 INFERIOR
INFERIOR $1.903 INFEROR
INFEROR $1.563 INFERIOR
INFEROR $993 INFEROR
facility were divided by 2.6 to get $279 per cleaner.
b Incremental cost = Net Annualized Control Cost -
Cmieeinn Q^Hi M»Tirtn * 1 t
1 •*• ft iMMiMeWffrw ^^tinri kt^ ArMiiieilii^
' L3SI NOmnTBnOf UPuOn NCR AnnuwiZBI
Mt KL-minfArvv Dntinn Pmiounn Retritjr^
($680)
($733)
($»**)
($1,353)
$1.537
$1,484
$1.273
$683
$2.248
$1.928
$1,338
$2.827
$2.747
$2.427
,$1,857
(per
d Control Cost
inn
($814)
($877)
($1.130)
($1.596)
$1,928
$1,864
$1.612
$1,145
$2.440
$2.360
$2.040
$1.450
$3.006
$2.926
$2.606
$2,016
5-4
-------
TAOE5-2. ANNUAUZED COOT FOR BQSTMGMBXUM BATCH CLEANBB (2 HOUR SCHEDULE)
Uvrtf Uwlt UMlt Uwf4
80% 40ft
1. Capital Cos*
-hoot $2.000 t2.000 $2,000 $2400
-LOtMtevditifo $2,100 -— $2.100
-reduced room Ml $1,500 $1.500
-increBMddwel $0 $0 $0
-teeboe/drettgembon device 1(000 $5.000
- automated cover — $4,500
-acfctttoi»JiocTK»ee required $526 $784 $1.888 $2.111
Total Cap* Costs (TCC) $4.628 97.794 $9.8U $12.711
Include* taxee, teight, & ratateton
AnnuafaedTctel Capital Costs (15 yr@ 10%) $808 $1,025 $1,800 $1,671
2. Annual Operation Cacti
— tflvf $0 $205 $0 $30ff
- Utilities (etocttdty) $119 $194 $209 $194
-Miscellaneous Operating Exp«ne«8 $80 $280 $280 $280
- Reportng/Recorefteeping Expcmas* $279 $279 $279 $279
- Moratof ing Expense $285 $818 $815 $364
Total Annuals** Cost $1,351 tS&8 $2.363 $2.903
3. Cost Effectiveness
Emission Reduction (Mg/yr) 0.75 W \SS 1.50
Recovered Solvent Credit ($/yr)
- Metnylene Chloride ($1 .03/kg)
- PerchJofoethytene ($1 .1 iykg)
- TrichJoroethylene ($1.43Ag)
- 1 .1 ,1 -Trichloroelhjne ($2.02/Vg)
Net Amualized Cost of Control (t/yr)
- Methytene Chloride
- Perchtoroethylene
- TrJchloroethytene
- I.l.t-Trichtoroethane
CostofControl{$/Mg)
- Methytene CNoride
- PefchJoroethytene
- Trichlaoethyiene
-1.1.1-Trichloroethane
hcremental Costs («Mg)b
- Metnylene Chloride
— Parchloroethyiene
- Trichlaoethyiene
-1.1,1-Tricnloroethane
($778)
(S833)
($1.073)
($1.515)
$579
$519
$279
<*1«)
$772
$892
$972
(S218)
$772
$892
$372
<»8)
($1,080)
($1.110)
(11,430)
($2.020)
$1^88
$1,168
$869
$278
$1.268
$1,188
$868
$278
NFERDR
WFEWOR
NFEHOR
NFEMOR
($1^88)
($1,888)
($1,788)
92^25)
$1,078
$976
$576
0162)
$861
$781
$461
(1129)
$994
$914
$594
$4
($1.545)
($1.665)
($2,145)
($3.080)
$1.448
$1.328
$648
($37)
$966
$866
$566
024)
$1,491
$1,411
$1,091
$501
facility were divided by 2.6 to get $279 per dun*.
Incremental cost- NetAnnusJMCprtroLCot^^
Emission Reduction - Latt Norirfariot Option Emiukan Reduction
5-5
-------
TABLE 5-1 ANNUAUZED COST FOR EXISTING LARGE BATCH CLEANERS (6 HOUR SCHEDULE)
Description
LeveM
40%
Lev* 2
50%
Bon
Level3
Level 4
70%
1.
Capital Cm*
-hoiat
-LOteeboardnrtto
-manual cam
-reduced room draft
-increased dwell
- freeboard ratigaration davica
-acaftonalloor space required
$2,000
$2.800
$0
11,630
$2.600
$2,800
$5.700
It .630
$2,600
$2.800
$5.700
$1.700
$8.155
Total Annuitized Cost
3. Cost Effectiveness
Emission Reduction (Mg/yr)
Recovered Solvent Credit ($/yr)
$1.741
2.82
$2.794
328
3.93
$2.600
$1.700
$0
$0.000
$3.307
Toti Capital Cot* (ICC) $7.030 $12.730 $15.965 $18,097
Includes taxes, freight & rotabtion)
Annuatized Total Capital Costs (15 yr@ 10%) $824 $1.674 $2.006 $2.198
Annual Operation Coats
- Labor $0 $0 $0 $205
- Utilities (elecfrterry) $119 $194 $194 $310
-Miscellaneous Operating Expenses $104 $332 $332 $464
- ReporHng/Racardkaeping Expanses* $279 $279 $279 $279
- Monitoring Expenses $315 $315 $364 $364
$3.818
4.59
- Methylene Chloride ($1 .03/kg)
- Perchkxoethylene ($1.1 1/kg)
- Trichloroethylene ($1 .43/kg)
- i.i,i-incnioroeinane(K.u
-------
TABLE 5-4. ANNUALIZED COST FOR EXBTWQ VERY IARQE BATCH CLEANERS (8 HOUR SCHEDULE)
LtVtJI
40%
LffMl 2
80%
Level 3
00%
Level 4
70%
1. Capital Costs
-hoist
-1.0 freeboard rate
- automated cower
-manual cover
— reduced room draft
12.600
$3,200
$2.600
$&aoo
refrigeration device
- aajftonal floor space required
$1,630
$1.680
$2400
$3400
$1.900
•8.860
82.600
$1.900
$0
$14.000
88.986
Total dpi* Costs (TCC)
Includes taxes, freight & installation
Annuafized Total Capital Costs (15 yr @ 10%)
Annual Operation Costs
- Labor
- Utilities (electricity)
- Miscellaneous Operating Expenses
- Reportng/Recordkeeping Expenses'
- Monitoring Expenses
17,430
$977
SO
$119
$104
$270
$315
$13,890
$1,619
$0
$209
$360
$279
$815
$17.480
$2496
$0
$209
$360
$279
$364
$2.951
$205
$465
$664
$279
$364
Total Annuitized Cost
3. Cost Effectiveness
Emission Reduction (Mg/yr)
$1,794
6.23
$2.962
7.78
$1506
9.34
10.9
Recovered Solvent Credit ($/yr)
- Methyterw Chloride ($1 .03/kg)
- Perchloroethylene ($L11/kg)
- Trichloroethylene ($1 .43/kg)
-1,1.1 -Trichloroetrtane($2.02/kg)
Net Amualized Cost of Control ($/yr)
- Metrtylene Chloride
- Perchloroethylene
- Trichloroethylene
-1.1,1-Trichloroethane
Cost of Control ($/Mg)
-Methylene Chloride
Hail liln.naita^
— refcniofoeuiyiene
- Tricnloroelnylene
- 1.1.1-Trichloroetnane
Incremental Costs ($/Mg)b
-Methylene Chloride
- Perchloroethylene
- Trichloroethylene
-1.1,1-Trichloroethane
($6.417)
($6.915)
($8.909)
($12.585)
($4.623)
($5.121)
($7.115)
($10.791)
($742)
/feA »»et imart fc* henta 4 at nl^eiiiara tbtmimjJntm immrf1i(m*r£nrt SM^rl ia«>m
-------
TABLE 8-8. SUMMARY Of EMISSION CONTROL COSTS FOR BOSTMQ M-UNE CLEANERS (S HOUR SCHEDULE)
Earieete* RcdycMoei Level
Level 1 Levels Levels Level 4 Levels
10% 20%
I.OFroefcoerd
1. CepftalCods
Flocf Speee Coet»
$1.WO
. to
so
so
S1.SOO
so
S10.100
S210
S11.SOO
mo
Tote/ Cap** Cacti (TCC), Mud* Tare*.
Fnigtt, und IrtittlfHon
Amu*fecdTc
Optio
I 1^^ AlUUMVvArif' f*«
EmlMlon Reduction - Uet NonMwIor Opion EmlMlon Reduction
5-8
-------
TABLE 5-6. ANNUAUZED COSTS FOR NEW SMALL BATCH CLEANERS (2 HOUR SCHEDULE)
Description
level 1
10%
Gmteion ftodudtan UMris
Level 2
90%
Level*
40%
Level 4
Levels
1. Capital Costs
— hoist
$2,000
-1.0 freeboard ratio
— reduced room draft
- automated cover
- superheated vapor
-freeboard refrigeration device
- additional Boor space required
1421
(2*000
$0
$1.200
$421
$4.700
$881
$1.400
$4,500
tl.TOP
12.000
$1.200
$1.400
$4.700
$1.919
Tot* Opt* Cot* (TCQ $2.421 $3,821 $7*31
Includes taxes, freight, & instalation)
AnnualizBd Total Capital Costs (15yr @ 10%) $318 $470 $884
2. Annual Operation Costs
-Labor $0 $0 $205
- Utilities (electricity) $75 $75 $150
- Miscellaneous Operating Expenses $80 $80 $268
- Reportng/Hecordkeeping Expenses* $279 $279 $279
- Monitoring Expenses $149 $149 $198
$1,264
.$280
$279
$198
$11*19
$1.475
$205
$150
$268
$279
$248
Total AnnualiZKl Cost
3. Cost Effectiveness
Emission Reduction (Mg/yr)
$901
0.13
$1.059
0.39
$2,064
0.52
$2.100
0.06
$2,625
0.79
Recovered Solvent Credit ($/yr)
- Methytone Chloride ($1 .03/kg)
- Perchloroethylene ($1.11/kg)
- Trichiof oethytene ((1 .43/kg)
- 1.1.1 -Trichloroethane ($2.02/kg)
Net Annualized Cost of Control ($/yr)
-Methylene Chloride
- Perchloroethylene
- Trichloroethyiene
- 1.1.1 -Trichloroethane
Cost of Control (S/Mg)
-Methylene Chloride
- Perchloroethylene
- Trichloroethyiene
-1.1.1 -Trichloroethane
Incremental Costs ($/Mg)b
-Methylene Chloride
- Perchloroethylene
-Trichloroethyiene
-1.1.1 -Trichloroethane
($134)
($144)
($186)
($263)
$767
$757
$715
$639
$5.904
$5324
$5.504
$4.914
INFEHOR
INFERIOR
INFEFVOR
INFERIOR
($402)
($433)
($556)
($788)
1657
$626
$501
$271
$1.686
$1.606
$1.286
$696
$1,686
$1.608
$1.286
$696
&&^*j
($577)
($744)
($1.050)
$1,526
$1.487
$1,320
$1.014
$2,939
$2459
$2,539
$1X9
INFERKDR
INFEHOR
INFERKJR
INFERKM
»680)
($733)
9*4)
($1.333)
$1.420
$1,3«7
$1.156
$766
$2.151
$2,071
$1.751
$1.161
$1,231
$1.151
$831
$241
($814)
($877)
($1.130)
($1.596)
$1,812
$1.748
$1.496
$1.029
$2.293
$2,213
$1.893
$1,303
$3.014
$£934
$2,614
$592
• C»/»h fe»ai*u U •«•! im«rl tet k*t« 9 II -" «K»>»W» rannrrlln ' mnit n»Mi4l>>M MM^ «< *7
-------
TABLE 5-7. ANNUALJZH) COST FOR NEW MEDIUM BATCH CLEANERS (2 HOUR SCHEDULE)
Description
Emission Reduction Levels
Level 1 Level 2 Levels Level 4
80% 40% 50% 60%
1. Capital Costs
-hoist
-1.0 freeboard ratio
~ rsducsd room drsfl
-increased dwell
- freeboard refrigeration device
- automated cow
- additional ioor space required
$2.000
$2,100
10
SS26
$2.000
10
$5.000
$740
$2.000
$1,500
$0
$4,500
$1.688
$2.000
$2.100
$1.500
$5.000
$2.111
Indudst hues, freight. & imitation
Annualized Total Capital Costs (15 yr @ 10%)
2. Annual Operation Costs
-Labor
- Utilities (etecMtity)
- Miscellaneous Operating Expenses
- Reporting/Recordkeaping Expenses*
- Monitoring Expenses
$4.ese
$606
$0
$119
$80
$270
$148
17,749
$1.019
$205
$194
$260
$279
$198
$0,888
$1.300
$0
$200
$260
$279
$196
$12.711
$1.671
$205
$194
$280
$279
$248
Total AnnaMztd Cost
3. Cost Effectiveness
Emission Reduction (Mg/yr)
$1,235
0.75
(2.175
1.00
$2340
1.25
$2.877
1.50
Recovered Solvent Credit ($/yr)
- Methylene Chloride ($1.03/kg)
- Perchloroethytene ($l.ll/kg)
- Trichloroethylene ($1.43/kg)
-1.1.1 -Trichloroethane ($2.02Ag)
Net Annualized Cost of Control ($/yr)
-Methylene Chloride
- Perchloroethylene
- Trichloroethylene
- 1.1,1 -Trichloroethane
CostofConfrol($/Mg)
- Methylene Chtoride
— PefchJoroethylene
— incrvoroetnyiens
-1.1,1 -Trichloroethane
Incfsmsntal Costt($7Mg)k
- Methylene Chtoride
- Perchloroethylsne
- Trichloroethylene
-1.1,1 -Trichloroethane
($773)
($833)
($1.073)
($1.515)
$463
$403
$163
($280)
$617
$537
$217
($373)
$617
$597
$217
^378)
($1,030)
($1.110)
($1.430)
($2.020)
$1.145
$1.065
$745
$155
$1.145
$1.066
$745
$155
WFERJOR
MFEROR
MFEROR
MFEROR
($1,268)
($1.368)
($1.788)
($2.525)
$959
$859
$459
($279)
$767
$667
$367
($223)
$992
$912
$592
$2
($1,545)
($1.665)
($2,145)
($3.030)
$1,332
$1,212
$732
($153)
$888
$808
$488
($102)
$1.495
$1.415
. $1,095
$505
t e*«h «^.X«.. \~ .MI ,m^4 tr, h»>_ O • i l.«n«i« II|«| ^ u " '- ww4 r«*tn<4bu. JXM* fJ *7
-------
TABLE 5-8. ANNUALJZB) COST FDR NEW UWOE BATCH O£ANERS (6 HOUR SCHEDULE)
Description
40%
Lewi!
60%
Lewi*
LsvsU
70%
1.
Capital Carts
— hoist
-I.Oteeboerd ratio
- automated cover
— Increased dwel
— reduced room draft
- superheated vapor
- freeboard legation device
$2.600
$2,800
10
$1.630
$2,000
$2,800
$5,700
$1.830
$2,600
$Z800
$5,700
$1.700
$3.158
$2.800
$0
$1.700
$0,000
$3,367
TottlCtpMCotti(7CQ $7,(OO $11790 $13*55 $10,087
vxludes taxes, teight, & irmtaastton)
Anruuzed Total Capital Costa (15 yr@ 10%) $924 $1.674 $2.096 $2,196
Annual Operation Costa
-labor $0 $0 $0 $205
- Utilities (etoctWty) $119 $194 $194 $310
-Miscellaneous Operating Expenses $104 $332 $332 $464
- Repcrtng/Hecordkeeping Expenses* $279 $279 $279 $279
- Monitoring Expensei $196 $196 $248 $246
Total AnnuaMzrt Cost
3. Cost Effactivi
Emission Reduction (Mg/yf)
2.62
3JB
3.93
13,702
4.56
Recovered Solvent Credit ($/yr)
- Methylene Chloride (11 .03/Vg)
- Perchtoroetnytene ($1.1 lAg)
- Trichtoroethytene ($1.43/kg)
- 1,1.1-Trichkxoe1h«r»($2.02/)cg)
Net Amualized Cost of Contol ($/yr)
- Methylene Cttortde
- PerchtoroethytenB
- Tricniaroethylene
-1,1,1-Trichloroettww
Cost of Contol ($/Mg)
— Metnytene Chtaride
- Perchkroefiytorw
- TrichtorotthytorM
-1.1,1-Tftchtaroethane
Incremental Costa ($/Mg)k
-Melhytene Chloride
- PercNoroethytone
- Tr fchtoroetrytone
- 1,1,1-TrichtaroethBne
($2.899)
($2.908)
($3.747)
($5.292}
($1.074)
($1^84)
($2.122)
($3,688)
0410)
($490)
($810)
($1.400)
($410)
NFEHOR
NFEROR
NFEROR
($3,378)
($3.641)
($4,890)
($6,626)
($701)
($864)
($2,013)
($3.949)
e»4)
d294)
(»4)
(11^04)
NFBVOR
WFEROR
MFBVOR
MFBtOR
($4,048)
($4,362)
($5.620)
($7,939)
($897)
($1.211)
($2,469)
($4.788)
C$226)
fOQ8\
*v w«*/
($626)
($1^18)
WFEROR
MFEROR
NFERK>fl
WFERJOR
($4.728)
($5.085)
($6.564)
($9.272)
($1.026)
($1.393)
($2.862)
($5.570)
($224)
($304)
($624)
($1.214)
$24
($304)
($824)
($1.214)
* Eachlaeitybaasurnedtohm&6cleiflers,toefcrerecadke^
tacity were divided by 2.8 to get $279 per deansr.
b Incremental cost <
Emission Reduction - LaatNorinlBrior Option Emission Reduction
5-11
-------
TABLE 5-9. ANNUAU2B) COOT FOR NEW VERY LARGE BATCH OJEANERS (6 HOUR 8CHffiU£)
Description
Laweil
40%
Lev* 2
50%
UmiS
60%
LaweU
70%
Capital Cods
-hoist
-1.0 freeboard ratio
- automated cover
-manual cover
-fcxraaMddwal
— reduced room draft
- superheated vepor
fc i h n mt4 *^MMA«M4lMA -J— J - -
— •eeooara rea igamon oevice
- addrttonal tea apace teqdred
$2.000
ttioo
S2.flOO
$3^00
$6.400
$i.eao
$1.630
$2.600
$3200
$6.400
$1^00
$3.860
$2.600
$0
$1.900
$14.000
$3.986
Indudaa tana, taight & IrvMtaten
Anraabad Total dpiWCoaliO Syr® 10%)
Annual Operation Costa
-Labor
- Utilities (atactidty)
-MiscflllanaousOparating&aMnMa
- Raporting/Ftoordkaapina Bpansaa*
- Monitoring Expanaea
$7.490
$977
$0
$119
$104
$279
$196
$13.830
$1,819
$0
$209
$360
$279
$196
$17.400
$2296
$0
$209
$360
$279
$248
$32,438
$2.961
$206
$465
$664
$279
$248
3.
Total Amutind Cost
Cost Effectiveness
Emission Reduction (Mg/yr)
Recovered Solvent Credt ($/yr)
- Melnytene Chloride ($1.Otykg)
- TrichJoroetnylene ($1.43Afl)
- 1,1,1-TrchJaoettwe<$2JaVQ)
Net Annualized Cost of Control ($/yr)
-Methytene Chloride
- PercNoroetnytene
— Tnohtoroethytene
- 1,1,1-Trichlaroelhane
Co6tofContot($/Mg)
— Matt lylai le Chlonde
— Perchloroetfiylene
- Tricntaroethytene
- 1.1.1-TMchtaroethane
IrersmentalCoffJt/Mg)*
— Metfiylene Chtende
- Parchloroetfiylana
- Trtchtaroethyiana
- 1.1.1-Tfichloroethane
($6,417)
($6.915)
($8.909)
($12.585)
($4.740)
($5238)
($7232)
($10.908)
($761)
(1841)
($1.161)
($1.751)
NFOTOfl
NFEnOn
MFEHOR
NFEHOR
7.78
($8,013)
($8,636)
($11,125)
($15,716)
($5.140)
($5.771)
($8261)
($12*61)
($662)
($742)
($1.082)
($1.652)
NFERJOR
NFEROR
MFEROR
MFEROR
0.34
($9.620)
($10.367)
($13.356)
($18.867)
($6226)
($6.975)
($9,964)
($15.475)
9867)
($747)
($1.067)
91.657)
MFEROR
MFEHOR
MFEROR
NFEROR
$4,812
10.9
($11227)
($12.099)
915.587)
922.018)
9«.415)
97287)
910.775)
917.206)
9589)
9669)
9989)
91.579)
9589)
9869)
9989)
91.579)
• Ea^tect^li assumed to heve 2^ deanery tfwetaereca*eaping and
tecaity were dMded by 2.6 to get $279 par deanar.
I cost of $726 per
b Incremental coat = Net AnruJ'gadCon>olCoat-
Dniaei on Reduction -JjathjoryrtykyOpltonEiniaaionReduclion
5-12
-------
8
$
EANI
i
2
J
I
COSTS
o
u
0
IU
MMARY
?*
1
•«•£
V *•
II
h:
ill
c
«
!i!
K
C
•
o
ll
r
r
8S
IS
82
O O
Is
81
ll
ssxs:
s i 2 s ;
88di
28829
ss$
i! « s fi
II I i
II
E c c , i i i
£ < < ' ' ' '
:
2
ill!
fill
a^lft
fill
»ABN
saea
5522
III!
eaaa
nil
i s if?? mi
III!
aftfta
fill
aaaJ
IKI
?ff
c a
E&
ecce
II2-8-
attfttt
f|!§
e
:- o
N N - -
• • • *
i!
ii
!!!§
e a e £
a
oooo
Ml I
lll
«l
e c c E
" O 0 0
KK K
ui yj IU
_!!!
K • • S
iiii
IU Ml II W
fc IL IL IL
z z z z
§000
ESS
IU II HI HI
Illl
:- 3
i £ S - * I £ S - «i£E- £i££-
i i i i | i i i i g i i i i 5 i i i i
e
U
U e
S
I
c
j
4
5-13
-------
TABLE 5-11,
SOLVENT CLEANER SOURCE SUBCATEGORY
REGULATORY ALTERNATIVES
Existing
New
Batch
Small
Medium
Large
Very Large
In-Line
I - 10% (floor)
II - 30%
III - 50%
IV - 60%
I - 40% (floor)
II - 50%
III - 60%
I - 40% (floor)
II - 70%
I - 60% (floor)
II - 70%
I - 30% (floor)
II - 50%
I - 40% (floor)
II - 50%
III - 60%
I - 50% (floor)
II - 60%
I - 60% (floor)
II - 70%
I - 60% (floor)
II - 70%
I - 40% (floor)
II - 50%
III - 70%
5-14
-------
costs but have not yet taken the opportunity to do so in the
absence of the standard.
5.3 NATIONAL IMPACTS
5.3.1 National Costs
After the regulatory alternatives were determined, the
national costs were estimated based on an estimated population of
24,500 halogenated solvent cleaning machines.3 Assuming an
average 15 year lifespan for a solvent cleaning machine and that
the number of cleaning machines will remain constant from 1993 to
1996, it was determined that in 1996 (first year of compliance),
the cleaning machine population will be 20 percent new cleaning
machines (4,900) and 80 percent existing cleaning machines
(19,600).
Based on the responses to section 114 surveys, the
distribution of cleaning machines among the different control
levels (i.e., 10 percent, 20 percent, etc.) was determined for
all cleaning machines sold in the last 10-year period. This
distribution was used as the control efficiency distribution for
existing cleaning machines. The control efficiency distribution
of those cleaning machines sold in 1990 was used as the control
efficiency distribution for new cleaning machines.
The national cost for cleaning machines to achieve a
particular regulatory alternative comprised the sum of the
capital; and monitoring, reporting and recordkeeping costs for
all cleaning machines' to move from their current levels of
control to the regulatory alternative level.3 If a cleaning
machine was at or above a regulatory alternative level, that
cleaning machine incurred no capital costs; but did incur
monitoring, reporting and recordkeeping costs. The corresponding
total emission reductions were also estimated and, with the total
national costs, used to calculate the average cost effectiveness
for each regulatory alternative. Then the incremental cost of
moving from one regulatory alternative to the next regulatory
alternative was determined. Table 5-12 presents the national
5-15
-------
^
y
c
(
iS
!«:
• °
i|i
a *>
9 a
e o
c o v>
a
h
• •
o
2
O 0 0
«n •-• c*
r- ^ oo
N 01
o o o o
* o* * oo
»n «n o» o
CO >H f-IN
o o o o
o o o o
m CM i-i in
•H M PH
o o o o
o o o o
o o o o
«*> * ro M
va o r» o
in o» o (N
CM rH * in
*o *
« O m o
on I *
1 MM>
Small 1
1
1
]
o"o
O^ tN
O <-l
fH fH
O O O
•-I V «O
r-l in VO
iH
O O O
000
o^ o\ o^
wn *
0 O O
§§§
•w m co
•H N CO
CM «-l CN
«n M «n
•
« O
*0 «0
o tn i
1 H M
Medium ]
]
]
0
m
n
O O
so o
in «
00
0 0
«n r»
« 0
r*
00
0 0
0 0
00 O
CM VO
* m
CM *
*s
^7
1 M
&
U
5
o
00
*
0 0
vo in
oo r»
§o
0
i-l CM
O «
IH fH
0*O
§§
* «n
VO CM
«O VO
o o
— t-1
* o
S7
1 H
&
3
•
>
0
00
»
oo
•SO
•H 00
rS
§§
VO rl
V O»
»H CM
§§
O O
O 00
iH fH
CM CM
vo m
•H CM
SO
in
1 M
8
J
e
M
o~o
0V fO
CM 00
fH CM
000
f VO 00
CM fH fH
§§§
« in vo
O o o
000
o o o
* » *
in vo ON
vo n fH
ON 00 fH
«
fH
«
*O vo
o m i
0 0
o o
r- co
CM f>
§o
0
o o
» «
0 0
fH O
* ON
•ft *
Ol W
O f
v0 1
^
w
>3
>
oo
in v
fH f»
fH fH
O OO
00 00 fH
r» oo
§0 O
oo
in v CM
in i"*fH
fH
o oo
o o o
o oo
% •» %
O 00 CM
«n -9 fH
% %
* >0
* OP-
o tn i
« 1 M
8
i4
1
M
§
£
•^
S
M
S,
e
it
M
a
9
9
a
~4
I
*
|
VM
a
rH
•^
*J
C
W
a
Jr
0
4J
9
I
I
f|
5-16
-------
costs, emission reductions, and the average and incremental cost
effectiveness values for the regulatory alternatives for each
subcategory.
5.3.2 Air. Water and Solid Waste Impacts
There are no adverse air water, or solid waste impacts
anticipated from the promulgation of these standards. However,
the use of some halogenated solvents is expected to decline as a
result of the phaseout mandated by the 1990 Amendments, the
Montreal Protocol, and presidential edict; and aqueous cleaners
have been chosen by many owners or operators as a cleaning
alternative. These standards may have the potential to cause
owners or operators to switch to aqueous cleaners or other
alternative cleaning technologies; although, this effect is
expected to be minor compared to the other factors listed above.
There is a potential impact on water quality from the use of
carbon adsorption from the regeneration of carbon beds. The use
of carbon adsorbers, however, are not recommended under these
standards. Carbon adsorbers are not expected to be selected by
owners or operators to control halogenated HAP solvent emissions
because of their expense relative to other control options. The
quantity of waste solvent or carbon from existing carbon
adsorption units disposed as hazardous waste would not be
affected by these standards.
5.4 REFERENCES
1. Memorandum and attachment from Sarsony, C. and Kane, C.,
Radian Corporation to Almod6var, P., EPA/CPB. July 6, 1993.
Summary of MACT Floor Determination.
•
2. Memorandum and attachments from Contos, L., Falling, A. and
Sarsony, C., Radian Corporation to Almod6var, P., EPA/CPB.
July 16, 1993. Summary of Costs and Cost Effectiveness
Associated with Emission Reductions for Selected Control
Techniques for Halogenated Solvent Cleaners.
5-17
-------
3. Memorandum and attachments from Sarsony, C., and Kane, C.,
Radian Corporation to Almod6var, P., EPA/CPB. July 19,
1993. National Impacts for Solvent Cleaning NESHAP.
5-18
-------
APPENDIX A
DERIVATION OF COMBINED
EFFICIENCY FORMULA
A-l
-------
Combined Efficiency Formula for Two Control Techniques
1. Let,
B • baseline emissions (no control)
Ej • removal tfficitncy for control 1 only
E2 • removal efficiency for control 2 only
2. Then,
BE, • mount of baseline Missions reduction using control 1
3. B - BE, • amount of baseline emissions remaining after application
of control 1
4. E^CB - BE,) • emission reduction control 2 acting on remaining
emissions
5. Total Emission Reduction
- BEj + E2(B - BEj)
• BEj + Bt, - BEIE2
6. The effective removal efficiency (Epcc) of control 1 and control 2
acting simultaneously on the baseline emissions (B) is, therefore,:
E£FF - E
Combined Efficiency formula for Three Control Techniques
1. Lit, .
B • baseline Missions (no control)
Ej • removal percent for control 1 only
E2 • removal percent for control 2 only
E3 • removal percent for control 3 only
A-2
-------
2. Then,
BEj • amount of baseline emissions reduction using control 1
3. B - BEj • amount of emission remaining after application of control 1
4. (B - BEj)E2 » amount of emission reduction using control 2
5. B - BEj - (BE* - BE,E2) • amount of emission remaining after
ap Hcatlon of control 2
6. [B - BEj -(BE2 - BEj£2)]E3 • amount of emission reduction using
control 3
7. Total Emissions Reduction
- BEj + (B - BEj)E2 + [B - BEj -(BE2 - BEjE2)]E3
- BEj + BE2 - BEjEj + BE3 • Btfa - BE2E3 * BEjE2E3
• BEj + BE2 + BE3 - BEjE2 - ttfa - BE2E3 + BE,E2E3
8. The effective removal efficiency (Egpp) of controls 1, 2, and 3
acting on baseline emissions 1s:
Ej * E2 * E3 - (EjE2) - (EjE3) - (E2E3) * (tfoty
A-3
------- |