EPA-450/2-78-050
OAQPS No. 1.2-117
Control of Volatile Organic
Emissions from Perchloroethylene
Dry Cleaning Systems
Emission Standards and Engineering Division
Strategies and Air Standards Division
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
December 1978
-------
OAQPS GUIDELINE SERIES , j
-1
The guideline series of reports is being issued by the Office of Air Quality Planning and Standards (OAQPS) to
provide information to state and local air pollution control agencies; for example, to provide guidance on the
acquisition and processing of air quality data and on the'planning and analysis requisite for the maintenance of
air quality. Reports published in this series will be available - as supplies permit -from the Library Services Office
(MD-35). U.S. Environmental Protection Agency, Research Triangle Park, North Carolina 2771 1; or, for a
nominal fee from the National Technical Information Service, 5285 Port Royal Road, Springfield, Virginia ^~
22161. W
Publication No. EPA-450/2-78-050
(OAQPS No. 1.2-117)
-------
TABLE OF CONTENTS
Chapter 1.0
1.1
1.2
, 1.3
Chapter 2.0
2.1
2.2
2.3
Chapter 3.0
3.1
3.2
3.3
3.4
Chapter 4.0
4.1
4.2
4.3
Chapter 5.0
5.1
5.2
5.3
Page
Introduction . 1-1
Need to Regulate 1_1
Sources and Control of VOC , . . . „ i _2
Regulatory Approach . . 1-3
Sources and types of Emissions ... . . . 2-1
i
Industry Description . 2-1
Dry Cleaning Processes and Emissions . . . . 2-2
References ^ 2_s
Emission Control Technology ....... 3-1
Use of Control Techniques . 3-1
Types of Control Techniques .-. . . . •. 1 . . 3.2
Suimiary ,i %_]$
References .* . . 3-12
Cost Analysis ..... 4-1
Introduction . 4_1
Perch!oroethylene Solvent Plant Cost Analysis . 4-3
References > t 4.3
Effects of Applying the Technology . . . . .5-1
Impacts on VOC Emissions . 5-1
.
Other Air Pollution Impacts . 5-4
Water Pollution Impacts 5.4
111
-------
Page
Chapter 5.4 Solid Waste Impact 5-7
5.5 Energy Impact 5-7
5.6 References 5-9
Chapter 6.0 Enforcement Aspects . . . . 6-1
6.1 Affected Facilities 6-1
6.2 Suggested Regulation , . 6-1
6.3 Discussion 6-4
Appendix A Emission Source Test Data A-T
A.I Plant A A-2
A.2 Plant B A-3
A.3 Plant C A-4
A. 4 References A-7
Appendix B Compliance Test Methods and Leak Detection
Equipment B-V
B.I Compliance Test Methods ......... B-l
B.2' Leak Detection Methods . B-7
B.3 Summary . B-8
iv
-------
LIST OF TABLES
Table 2-1
Table 3-1
Table 3-2
Table 3-3
Table 3-4
Table 4-1
Table 4-2
Table 4-3
Table 5-1
Table 5-2
Table 5-3
Table 5-4
Table 5-5
Table A-l
Table A-2
Solvent Losses in Perchloroethylene Plants .
i'
Potential and Applied Control Techniques for
Dry Cleaning Plants .
Carbon Adsorber Test Summary
Effect of Housekeeping Practices . . . i, .
Control Techniques and Solvent Emission Levels
Cost Parameters for Model Perchl oroethylena
Plants
Costs for Carbon Adsorption . .
Cost Effectiveness
Hydrocarbon Emission Factors (Uncontrolled); .
Effect of Applying Available Air Pollution
Control Techniques .
Impact of Control Systems on Water Usage . .
Perchloroethylene Solvent in Effluent Water, .
Energy Imoact . .
Plants Tested by EPA . .
Dry Cleaning Test Data
Page
. . 2-6
. . 3-1
. . 3-3
3-8
. . 3-11
. . 4-4
. . 4-5
. . 4-7
. . 5-2
, . 5-3
. . 5-5
. . 5-6
5-8
A-l
. , A-6
-------
LIST OF FIGURES
Page
Figure 2-1 Perch!oroethylene Dry Cleaning Plant Flow
Diagram .............. 2-4
vi
-------
1.0 INTRODUCTION i.
This document is related to the control of volatile organic compounds
(VOC), specifically perchloroethylene (perc), from all dry cleaning systems
which use this solvent. Other solvents used in the dry cleaning industry,
petroleum distillate (Stoddard Solvent) and trichlorotrifluoroethane
(fluorocarbon), may be discussed in later documents.
Methodology described in this document represents the presumptive
norm or reasonably available control technology (RACT) that can be applied
to existing perch!oroethylene dry cleaning systems. RACT is defined as the
lowest emission limit that a particular source is capable of meeting by the
application of control technology that is reasonably available considering
technological and economic feasibility. It may require technology that has
been applied to similar, but not necessarily identical, source categories.
It is not intended that extensive research and development be conducted before
a given control technology can be applied to the source. This does not,
however, preclude requiring a short term evaluation program to permit the
application of a given technology to a particular source.' The latter effort
is an appropriate technology forcing aspect of RACT.
1.1 NEED TO REGULATE \ .
Control techniques guidelines concerning RACT are Being prepared for
those industries that emit significant quantities of air'pollutants in areas
of the country where National Ambient Air Quality Standards (NAAQS) are not
1-1
-------
being attained. Perch!oroethylene dry cleaning systems are a significant
source of VOC and are predominantly found in urban areas.
Annual nationwide emissions from perch!oroethylene dry cleaning systems
are estimated to be 158,000 metric tons per year or about 0.9 percent of
total stationary source emissions.
The other two solvents used in the industry are not as significant as
perch!oroethylene. Petroleum solvent systems emit 68,000 metric tons per
year and trichlorotrifluoroethane (not considered a photochemically reactive
VOC) emissions are estimated to be only 820 metric tons.
1.2 SOURCES AND CONTROL OF VOLATILE ORGANIC COMPOUNDS FROM PERCHLOROETHYLENE
DRY CLEANING SYSTEMS
Dry cleaning systems have several sources of emissions. The major source
is the dryer (known as the recovery tumbler or reclaimer). While every
perch!oroethylene dryer is equipped with a condenser, significant quantities
(about 50 percent) of emissions occur from this source. The disposal of waste
materials is the second most significant source followed by the losses from
liquid and vapor leaks.
Control techniques are available and have been widely applied in this
industry. It is estimated that about 35 percent of all commercial and
industrial perch!oroethylene system dryers are equipped with carbon adsorbers.
Emissions from waste material disposal can be reduced by several methods, %
among them the proper operation of cookers and cartridge filters. Finally,
V
leaks can be prevented by visual inspection and by periodic monitoring with
a leak detection instrument.
Capital costs of carbon adsorbers are $4500 for a large commercial plant
of 46,000 kg (100,000 pounds) of clothes throughput per year. Cost
41
1-2
-------
effectiveness of controls in this perch!oroethylene system is $90 savings
per metric ton of perchloroethylene removed.
1.3 REGULATORY APPROACH ,;
The application of RACT will reduce dryer outlet concentration
to less than 100 ppm; reduce emissions from filter waste and still
bottoms; and eliminate liquid and vapor leaks. A study is underway to
determine the significance of vapor leaks. A test procedure to define
a "leak tight system" is being developed and will be available in the
near future 1f vaP°r leaks are shown to be a problem.
The following sample regulation, discussed in Chapter 6.0, incorporates
all of the above recommendations: ;
Sec- 1* Solvent emissions from perchloroethylene dry cleaning systems
must be limited in accordance with the provisions of this Rule..
Sec- 2- Compliance with this Rule requires the following:
(a) There shall be no liquid leakage of organic solvent from the
system.
(b) Gaseous leakage shall not exceed ppm. -^
(c) The entire dryer exhaust must be vented through a carbon
adsorber or equally effective control device.
1 if H^mJ 9""™"*,Assessing the significance of vapor leaks.
If deemed significant, a test method for detecting leaks will
be developed and issued to interested parties
1-3
-------
(d) The maximum organic solvent concentration in the vent from
2/
the dryer control device shall not exceed 100 ppm before dilution.—'
(e) Filter and distillation wastes.
(1) The residue from any diatomaceous earth filter shall be
cooked or treated so that wastes shall not contain more than 25 kg
of solvent per 100 kg of wet waste material.
(2) The residue from a solvent still shall not contain more than
60 kg of solvent per 100 kg of wet waste material.
(3) Filtration cartridges must be drained in the filter housing
for at least 24 hours before being discarded. The drained cartridges
should be dried in the dryer tumbler if at all possible.
(4) Any other filtration or distillation system can be used if
equivalency to these guidelines is demonstrated. For purposes of
equivalency demonstration any system reducing waste losses below
1 kg solvent per 100 kg clothes cleaned will be considered equivalent.
Sec. 3. Sections 2(c) and (d) are not applicable to plants where an
an adsorber cannot be accommodated because of inadequate space or to
plants where no or insufficient steam capacity is available to desorb
adsorbers. The District may exclude other plants from the scope of
Sections 2(c) and (d) if it is demonstrated that other hardships justify
such an exclusion.
2/ Enforcement of these provisions is dependent on the development
of a satisfactory detection instrument and test method.
1-4
-------
Sec. 4. Compliance Procedures
(a) Liquid leakage shall be determined by visual inspection of the
following sources:
(1) Hose connections, unions, couplings and valves;
(2) Machine door gasket and seating;
(3) Filter head gasket and seating;
(4) Pumps; ' ;
(5) Base tanks and storage containers;
(6) Water separators;
(7) Filter sludge recovery;
(8) Distillation unit;
(9) Divertor valves;
(10) Saturated lint from lint basket; and
(11) Cartridge filters.
(b) Vaporaleakage shall be determined by .-•
(c) Dryer exhaust concentration shall be determined by .—'
(d) The amount of solvent in earth filter (2.e.'l) and distillation
wastes (2.e.2) shall be determined by utilizing the test method
described by the American National Standards Institute in the paper,
"Standard Method of Test for Dilution of Gasoline-Engine Crankcase
Oils."
3/ See footnote 1, above.
4/ See footnote 2, above.
i
1-5
-------
-------
2.0 SOURCES AND TYPES OF EMISSIONS i
2.1 INDUSTRY DESCRIPTION
Dry cleaning is a service industry, involved in the cleaning of apparel
or renting of apparel. Basically, the industry is segregated into three
areas based on customers and types of services offered. These services
are: (1) coin-operated, (2) commercial, and (3) industrial.
Coin-operated dry cleaning facilities are usually part of a "laundromat"
facility (although there are separate installations), and are operated on
either an independent or a franchise basis. They provide a low cost "self-
service" type of dry cleaning without pressing, spotting, or other services.
Processing is generally about 7200 kilograms (16,000 pounds) of clothes per
year per store (two systems per store). Commercial dry cleaning plants are
the most familiar type of facilities, offering the normal services of cleaning
soiled apparel and other fine goods. They include: small neighborhood dry
cleaning shops operating on an independent basis ("Mom and Pop" dry cleaners),
franchised shops (e.g., "One Hour Martinizing") and specialty cleaners,
handling leather and other fine goods. Neighborhood dry cleaners generally
process about 23,000 kilograms (60,000 pounds) of clothes per year. The
industrial cleaners are the largest dry cleaning plants predominantly supplying
rental services of uniforms or other items to business, industrial, or
institutional consumers. A typical industrial cleaner processes 240,000 to
700,000 kilograms (600,000 to 1,500,000 pounds) of clothes per year. They
are generally associated with large water laundry services. Nationwide perc
emissions are 21,400 metric tons for coin-op, 123,000 metric tons for
commercial and 13,600 metric tons for industrial dry cleaners.
,i
2-1
-------
2.2 DRY CLEANING PROCESSES AND EMISSIONS £K
2.2.1 The Basic Process
Dry cleaning is the cleaning of fabrics in an essentially non-aqueous
solvent. The principal steps in the process are identical to those of
ordinary laundering in water: (1) one or more washes (baths) in solvent;
(2) extraction of excess solvent by spinning; and (3) drying by tumbling
in an air stream. The solvents used are categorized into two broad groups:
(1) petroleum solvents which are mixtures of paraffins and aromatic hydrocarbons
similar—but not identical—to kerosene, and (2) synthetic solvents which are
halogenated hydrocarbons—perchloroethylene and trichlorotrifluoroethane.
Differences between the dry cleaning procedures for these two groups of
solvents are due primarily to three factors:
• Synthetic solvents are much more expensive than petroleum solvents.
• Petroleum solvents are combustible, while synthetic solvents are tf|))
nonflammable.
• The densities of synthetic solvents are about twice that of petroleum
solvents.
This document discusses one synthetic solvent, perchloroethylene, only, as it
is by far the most prevalent solvent type. The other synthetic solvent,
trichlorotrifluoroethane, is not considered to be a photochemically reactive
VOC. Petroleum solvent systems as discussed in Chapter 1 are being examined *
in a separate EPA study at present. By way of illustration, Figure 2-1 is a
>f
schematic of a perchloroethylene plant.
2.2.2 Perchloroethylene Systems and Emissions
Perchloroethylene machines find their major use in commercial dry cleaning
plants (about 74 percent of systems). The typical neighborhood dry cleaner uses
2-2
-------
a perch!oroethylene based process. However, perch!oroethylene-based equipment
is also used in the industrial sector (EPA tested one in their test program -
See Appendix A), making up about 50 percent of the systems and is used in the
coin-op sector where it is the predominant solvent by far (fluorocarbon
machines account for about 3 percent of the market; petroleum, none).
2.2.2.1 Solvent Characteristics - Although other chlorinated hydrocarbon
solvents have been used for dry cleaning in the United States, perch!oro-
ethylene is the only chlorinated solvent seeing significant; use at this time.
An estimated 160 million kilograms (346 million pounds) of "perc" is used
annually for dry cleaning purposes. The solvent may be generally
characterized as follows: ' ; •
. Non-flammable,
. Very high vapor density, , • .• -
. High cost ($.49/kg) - ;
. Aggressive solvent properties.
In spite of the higher cost per gallon of perc, solvent costs for perc plants
are quite competitive with those for petroleum solvent plants, its chief competitor,
because the former are always used with solvent recovery equipment. Stricter
fire codes, increases in petroleum solvent costs, and environmental considerations
have resulted in the use of perc-based equipment for many new plants.
2.2.2.2 Equipment Characteristics - Perc machines may be Cither transfer or
dry-to-dry types. This refers to the method of drying the clothes. In a dry-
to-dry system, the drying is done in the same tumbler as the washing. Clothes
are put in dry and come out dry. For transfer systems, the dryer is separate and
clothes are transferred from washer to dryer. The great majority of perc machines
are transfer units. ', '
2-3 '
-------
r
s_
OJ
(O
•1
to
IS
r
5-
(1)
to
E
CU
•o
E
O
o
P
(13
tO
Q.
O)
to
I
O)
(O
T
cr>
c
•p-
c
(O
cu
o
CU
cu
O
o
U
CM
CD
3
CU.
cu
-a
•o
cu
to
to, to •
O). CU
S-
O)
C -Q
O S-
JQ O
S- to
CO "O
O to
OJ
to
a>
T3
O
O
cu
•p
(O
4 -
S.
cu
•a -P
cu c
s- cu
cu >
•p i—
i— O
•I- IO
s-
o
p
rei
to
a.
cu
to
*(
condenser
L..
S-
o
O
X
CU
s-
cu
^z
10
ns
4-5
CU
CU i— E
S- O <0
3 10 4->
O-
-O E
CU CU
01 >'.
i- r—
IO O
O
to
tt)
en
cu
4-»
CU
-o
O
3
V)
CU
1/5
(C
cu
o
in
E
O
•r—
P
CL
S-
o
to
cu
-a
2-4
-------
A typical commercial perc plant has a 14-27 kg (30-60 Ib) capacity
washer-extractor with a reclaiming tumbler of equivalent size. According
to one survey about half the plants have carbon adsorption units to reduce
? q
solvent consumption.c A more complete survey puts this figure at 35 percent.
Apparently many large perc plants in the industrial sector use adsorption
whereas the majority of commercial plants do not. These control devices
are discussed in Chapter 3.0, Emission Control Technology. ;
2.2.2.3 Emission Characteristics (see Table 2-1 for summary of emissions) -
As stated above, perc plants frequently have vapor adsorbers to reduce solvent
usage. Typical solvent losses for both controlled and uncontrolled perchloroethylene
dry cleaning plants are shown in Table 2-1 as reported by IFI.4 These are
for well-operated plants. Table 2-1 also gives average emissions from three
EPA tests discussed in Appendix A. 5'6'7 ;
I
Table 2-1 shows that the uncontrolled plant can have high emission rates
from filter muck and the dryer exhaust. The figure for evaporation at the
dryer assumes that a condenser is used to recover a certain portion of the
stream. Actually, after wash and extraction, dry cleaned materials contain
about 20-25 kilograms of solvent per 100 kilograms of clothes. All of this
solvent is vented to the condenser. A well-operated condenser reduces this
level to 3-6 kg per 100 kg.
Other sources include evaporation at the washer (from transfer operations
generally), distillation and filter waste disposal, and miscellaneous emission
sources. These miscellaneous emission sources include: losses from pumps, valves,
flanges, and seals; evaporative leak losses from storage vessels; chemical
and water separators; and minor inefficiencies in handling solvent and material.
2-5
-------
Table 2-1. SOLVENT LOSSES FROM WELL OPERATED PERCHLOROETHYLENE PLANTS
(kilograms of solvent per 100 kilograms of clothing)
8,9,10,11
Source
Evaporation @ washer
Evaporation @ dryer
Vapor adsorber exhaust
(properly operated)
Retention in filter muck
• Rigid tube filter-no cooker
• Rigid tube filter-muck cooker
• Regenerative filter-muck cooker
Retention in paper cartridges
• Drained
• Dried in cabinet vented to
adsorber
Retention in still residue
Miscellaneous losses (leaks)
Average Total Loss
IFI data .
(EPA data)b
Plants without
vapor adsorber
0.54 (1)
3 (6)
-
14
1.6
1 (1)
1.8 (0.6)
-
1 .6 (no data)
2 (1)
8-21 c
IFI data ,
(EPA data)0
Plants with
vapor adsorber
0
0
0.3 (0.3)
14
1.6
1. (1)
1.8 (0.6)
1.2
1.6
2 (1)
6-18c (3-5)
a Figures represent well-operated systems. Average emission rates
by industry survey estimated at 12 kg/100 kg.
b EPA data in parenthesis. •
c These ranges are high because plants could not operate economically
without a muck cooker if filter is used.
o
2-6
-------
According to IFI data, the usual plant has a regenerative filter with
a muck cooker, and this results in a total consumption rate of about 8 kg
of solvent per 100 kg of clothing. (According to one vendor, more and more
plants are using cartridge filters-, now in about 55 percent of commercial
12
plants.) For an adsorber-equipped plant, the corresponding solvent usage
is less than 5 kg per 100 kg of clothing, which is equivalent to a 40 percent
loss reduction. It should be emphasized that these usage levels are for
well-operated commercial and industrial plants; average losses--including
controlled and uncontrolled plants—are estimated to be about 10-12 kg of
solvent per 100 kg of clothes cleaned and 20 kg per 100 kg for coin-op.14
Coin-op stores generally have higher emission rates because of underloading
of equipment, lack of carbon adsorption technology, and unattended systems.
2-7
-------
2.3 REFERENCES
1. Data supplied by industry representatives—compiled by Ernst & Ernst,
December 14, 1976. Eight percent added as estimate of imported perc.
2. Watt, Andrew, IV, and William E. Fisher, "Results of Membership Survey
of Dry Cleaning Operations," IFI Special Reporter No.3-1, January-February 1975.
3. Mayberry, J.L., President, R.R. Streets and Company, Inc., letter
to John H. Haines, EPA, March 2, 1977.
4. Fisher, William E., "The ABC's of Solvent Mileage," Part One, IFI
Special Reporter, No.3-4, July-August, 1975.
5. Kleeberg, Charles F., "Material Balance of a Perchloroethylene Dry
Cleaner Unit," test report to James F. Durham, on test in Hershey, Pennsylvania,
March 17, 1976.
6. Kleeberg, Charles F., "Material Balance of an Industrial
Perchloroethylene Dry Cleaner," test report to James F. Durham on test
in San Antonio, Texas, May 14, 1976.
7. Kleeberg, Charles F., "Material Balance of a Small Commercial
Perchloroethylene Dry Cleaner," test report to James F. Durham on test in
Kalamazoo, Michigan, May 17, 1976.
8. Ibid, Reference 4.
9. Ibid, Reference 5.
10. Ibid, Reference 6.
11. Ibid, Reference 7.
12. Cunniff, Joseph L., Vice President of Puritan Division, letter to
Robert T. Walsh, EPA, November 21, 1978.
2-8
-------
13. Ibid, Reference 2.
14. Anonymous Dow Chemical Survey submitted by Joseph Qunniff, Puritan
Filters, to EPA on March 3, 1977.
2-9
-------
-------
3.0 EMISSION CONTROL TECHNOLOGY
The purpose of this chapter is to discuss control techniques for both
existing and new perchloroethylene dry cleaning plants and to define emission
levels that can be achieved with available control technology. Chapter 4.0
is an assessment of the costs of applying the technology.
3.7 USE OF CONTROL TECHNIQUES
For the most part, solvent emission controls for dry cleaning plants
have developed out of economic necessity. In order for a more costly
?
synthetic solvent like perchloroethylene to compete with inexpensive
petroleum solvents, a substantial degree of solvent recovery is necessary
during the drying operation. Solvent is recovered by condensation on all
perc solvent dryers; many are equipped with adsorbers. Table 3-1 shows
the extent of controls on perchloroethylene systems in the three industry
sectors.
Table 3-1. POTENTIAL AND APPLIED CONTROL TECHNIQUES
FOR DRY CLEANING PLANTS ,
Carbon adsorption
Housekeeping
Incineration
Minimize solvent loss
in wastes
Industry Sector
Coin-Op
N/U
Very limited
N/A
To a degree
Commercial
35%
To a degree
N/A
To a degree
Industrial
50% (est.)
To a degree
N/A
To a degree
N/A - Not applicable
N/U - None used
3-1
-------
3.2 TYPES OF CONTROL TECHNIQUES
3.2.1 Carbon Adsorption
Activated carbon is used in many applications for the removal of organic
compounds from carrier gases (usually air) by adsorption. It has been used
extensively to recover perch!oroethylene from dry cleaning systems. Adsorption
is the property of a surface to retain molecules of a fluid which have con-
tacted the surface. Perch!oroethylene can be retained on carbon very easily.
The working bed capacity (weight of solvent per weight of carbon, expressed
as percent) for perch!oroethylene is about 20 percent.
The cost of perch!oroethylene solvent has encouraged and necessitated
recovery of some kind. The earliest units used water cooled or refrigerated
condensers to control 85-90 percent of losses from the dryer. Rising solvent
costs made adsorption of the remaining 10-15 percent attractive. Carbon
adsorption has been used on perch!oroethylene units for years.
EPA collected data during plant tests on three carbon adsorption units
used with perch!oroethylene systems. Appendix A of this report details the
results of those tests. Table 3-2 summarizes the data and shows inlet and
outlet concentrations associated with each of the three tests. Outlet
concentrations ranged from 2 to 100 ppm as perch!oroethylene. Collection
efficiencies ranged from 96 percent to 99.6 percent.
Also seen in Table 3-2 is a list of the sources controlled. In each
case, vapors were drawn off at the dryer and washer, at least. Generally,
a current of fresh air is required for occupational safety at the operator's
face when loading and unloading. This is usually accomplished by an
internal fan (activated by door opening) which draws air through
-------
CO
I—
z
ca:
_1
Q-
UJ
LU
CD
C£
o
o
o;
LU
UJ
£
>-
oo
oo
LU
D;
LlJ
CQ
a:
o
oo
Q
03 03
CU i- ~O
p— *^^
•(-> CO CO
O 03
E
E
0
-l-> 03 ,
cu J-co
,1} ^— ci
3 cu a.
O O Q.
E
o
o
E
o
•r—
-(-5
03
•i-^ t-
CU -M E
i — E Q.
E CU Q.
1—1 O
E
O
O
CO T3 >,
OJ CU 03
-c: E T3
4-> 03
0 CU S-
r— r— CU
O O Q.
^~~^
T3
CU
CO 4->
CO E
CU CU
0 >
0
S- CO
0- •!->
•r—
E
3
4->
E
03
Q_
•
.{_>
M-
•
CT
CO
O
•=*-
r—
CO
CM
0
O
CO
>>
03
•a
CO
o
CO
*=sf~
CO
s-
O)
^_>
cu
E
•
cr
co
(^
CO
—
E
o
•1—
4—*
03
sT£l
CU T-
cu >>+•>
E S- CO
CU -O -r-
f— ^
JE* iT
4-> O CO
i— CU 0 -M-— .
"03 O -O E -f->
S- -r- S- CU E
CU O O S- > CU
4- S- r— CU >
co cu jc JE s~
E E 0 CO 0 -!->
S-
1—
o-
CO r-
CVJ f—
to
pw.
^^J 1
f^
o
r—
n
o
o
CO
CO
>^
03
"P
a>
-X
o
1 —
„, ^
S- CO ,
cu o +->
E O E
CU T3 CU
i— >
>> S-
JE CU S-
« «4-> >> O
>5i— CU S- 0
S- 03 O T3 r—
•0 •>- S_ ^s. C|_ ,
O O S-
O S- i — CU «
+•» CU JE JE i-
E O co CU
S- O CU 3 S_
Q U Q.^ T3
: ^1
o
.
cu
S»- CO
CU T-
ja co
s- cu
o -a
CO S-
•o cu
o3 "O
E
E 3
o
J3 E
03 O)
O f~>
CU CU T3
-E > CU
•4~^ O3 f^
JE
0 CU
4-> O r—
4-> CO
T3 E
CU i- -r-
4-> CU CO
E JO
CU S- 03
> 0
^/) M-mi
OJ "O O
S- 03
CU E
3 CO 0
+•> JC +»
03 +J a.
JC S-
-•-> 2 0
O CO
CO JE CU
cu co "o
O 1
S- 03 E
3 4-> O
O 03 •!—
CO "O +J
Q.
E CO S-
he emissio
icontinuou
le of adso
4J E 0
cu >,
CO CO O
cu
-l-> T3 CU
03 CU E
O +-> O
"5 E S-
E -r- 0
r— i CM! col
3-3
-------
a duct at the machine door lip. The air, laden with solvent vapor, is then
passed to the carbon adsorber.
Dryers usually vent during specific points in the drying process. The
dryer exhaust is generally chilled to remove solvent and then reheated and
recirculated to the dryer. At the end of the drying cycle, the clothes are
hot and must be cooled gradually to avoid wrinkles. Fresh air is drawn in
(in a process called deodorizing) and is vented to the adsorber (since
condensation would not be effective on the low concentration stream). Dryers
also vent to the adsorber whenever the overheat thermostat is actuated causing
cool air to enter an overheated dryer. At least one system design vents
dryer exhaust to the adsorber continuously (Plant C was an example) for
system simplification.
Floor vents are installed to control fugitive vapors around the machines
and to draw vapors from solvent spills. These vents have been located on
the floor next to the front of machines and next to filter systems. There
is some evidence that these vents are more effective if they are located
at the same level as the solvent emission; perchloroethylene vapors do not
2
necessarily drop to the floor because of the solvent's density.
There is no technical reason why all sources in dry cleaners vented
through a stack or duct to atmosphere cannot be directed to a carbon adsorber.
This includes distillation units, washer loading vents, storage tanks, and
chemical separators. None of these vents has an extremely high volume of
vapor to be treated. Emissions from these sources and other pertinent data
are described in Appendix A.
For perch!oroethylene based units, carbon adsorption can be used to
achieve 100 ppni or less outlet concentration on the sources discussed above.
3-4
-------
Space requirements vary with the size of the unit. For the three plants
tested the adsorber floor space is shown in Table 3-2. These area estimates
include piping, canister, and ductwork.
Coin-operated perch!oroethylene systems have special ^problems. There
is generally no steam demand at coin-ops and thus no steam boiler. In most
cases, the steam necessary to desorb a carbon bed does not exist at these
plants and necessary space for an adsorber is not available. Either the
carbon bed must be portable and taken off-site for regeneration or a steam
boiler must be added at each site. EPA examined the feasibility of
regenerating carbon beds off-site and found space requirements and costs
high. (The capacity of the bed must be large to accommodate solvent
recovered over long periods of time or else the carbon must be regenerated
often.) While coin-operated perch!oroethylene dry cleaners have had only limited
use of carbon adsorbers, the technology for perchloroethylene recovery is
certainly demonstrated. Costs are evaluated in Chapter 4.b and include boiler
installation costs. EPA will continue to evaluate methods! of controlling
coin-operated systems.
3.2.2 Housekeeping
The losses associated with poor maintenance of equipment are difficult
to quantify. A few devices, however, control major emission sources in dry
cleaning plants; neglect of these devices can significantly contribute to
high solvent loss. Other sources of emissions—fugitive or miscellaneous--
are not associated with "point losses" or losses from obvious areas such as
venting of dryers or disposal of filter wastes. Fugitive emission points
include leaks from valves, flanges, seals, and covers on storage tanks.
3-5
-------
There are two types of losses from both point and fugitive emission
sources—liquid and vapor. Liquid losses can be detected by sight—the
brown residue associated with a solvent leak is familiar to any operator.
3
One solvent manufacturing company estimates that a leak of one drip per
second equates to as much as four litres of solvent per day. Because of
the volatility of the solvents, these liquid leaks are eventually evaporated
to atmosphere. Vapor leaks can be detected by smell, application of soap and
water to sources, or hydrocarbon detectors. EPA is currently evaluating the
significance of vapor leaks and also a number of methods of detecting vapor
leaks and will advise at a later date on the optimum approach. Our objective
is to develop an inexpensive monitor which can be used to detect major vapor
leak sources. Vapor losses usually occur at evaporative points and tears in
ductwork. The solvent manufacturer has submitted a list of common emission
areas4 which should be checked periodically to control these losses. The
c c -j
following checklist is similar to those used by other vendors » ' to advise
customers on how to maintain equipment.
Liquid leakage areas include:
a) Hose connections, unions, couplings and valves.
b) Machine door gasket and seating.
c) Filter head gasket and seating.
d) Pumps. "
e) Base tanks and storage containers.
*.
f) Water separators (lost in water due to poor separation).
g) Filter sludge recovery (lost in sludge by improper recovery).
h) Distillation unit.
i) Divertor valves. ^
j) Saturated lint from lint basket.
3-6
-------
k) Cartridge filters.
Vapor leakage areas include:
a) Deodorizing and aeration valves on dryers (the seals on these
valves need periodic replacement).
b) Air and exhaust ductwork (solvent lost through tears in duct).
c) Doors left open are problems. Leaks in the system should be
confined to the closed washer and/or dryer if possible.
d) Button traps and lint baskets should be opened only as long
as necessary.
Other areas include: ,
a) Lint screens and bags, fan blades and condensers can adversely
affect capture systems if they are clogged or caked with ,lint.
b) Overloading and underloading can increase losses. Overloading
makes drying difficult. Underloading is self-defeating since most losses
are fixed in the system.
c) Inefficient extraction; due to overloading or loose belts can
cause poor drying.
Rapid detection and repair of leaks is essential to minimize solvent
losses. Table 3-3 shows how neglect of certain pieces of equipment can
increase solvent consumption from the well-controlled plant usage of 3-5 kg
per 100 kg of clothes cleaned to the neglected plant losssof greater than
15 kg per 100 kg. These data were derived from plant tests, vendors,
industry data, and estimates. In one solvent company survey, plants reported
solvent usages from less than 2 kg per 100 kg to above 35 kg per 100 kg.
Average use was around 12 kg per 100 kg.8 Good housekeeping practices
require very little additional effort in existing plants. I No new equipment
is needed and little cost is incurred.
3-7
-------
LU
CO
CO
o
LU
O
CO
Ul
fc
o
o
LU
D-
O
to
LU
O
O
•=c
Di
Q-
CD
OL.
LU
LU
i^:
LU
CO
1
LU
OO
co
cu
(O
to
(0
1
c
n^
QJ
O
S-
*J|
*O CD
CU .M
•p
0 O
CU CD
CU CD
-a
cu
£3 ""^
o«— 1
s-
1 2
0
CU CD
^? J^£
cu
o
o
CO
•p
E
(O
ol
E
s- cu
(O-Q
0 S-
o
CU CO
QJ-O
to (O
1
o
cu
o
cu
E
cu
>,
•^
cu
0
S-
o
r~
.E
O
s-
cu
O-
— . — •• - — 1 ......
1 § =
E -£°
•s s-g.
\^ S-
o o o
8 ^S
E -^ "O
3 CU
t— S- 10
£ .= § S
E Q. i —
OJ f
S ^^ "S
?. CU CU -P
o to J3 - to i- to
•i- to -P * "? ^
u_ <_s co 3 <-> s:
rO
ns
•a
S-
o
•a
cu
T3
E
(O
to
-P
to
cu
•p
(0
r—
Q.
^£
D-
LU
E
O
-a
cu
CO
fO
ca
^^^
^^ i
1
CD
•P
CO
CU
-p
-a
re
to
cu
S-
3
to
^
[ ^
CO
3
-a
-0
•a
cu
PI
o
"a5
>
cu
•a
ni ssion factors
cu
cu
CD
rO
i-
cu
^
(0
E
O
•a
cu
to
(O
CO
•--^
3-8
-------
3.2.3 Incineration
Incineration, while technically feasible for control of perch!oroethylene
is impractical for halocarbons.
Perch!oroethylene is virtually inflammable and large quantities of
supplemental fuel must be used to combust it. Incineration can produce
hydrogen chloride (HC1), chlorine (Cl) and phosgene (COCK).9 All of these
compounds can be removed by scrubbing exhaust gases with water. However,
some water treatment would likely be required. ;
3.2.4 Waste Solvent Treatment
Solvent is recovered from filter muck (diatomaceous earth, carbon,
lint, detergents, oils, and solvent) and from distillation bottoms. In many
perch!oroethylene systems solvent is "cooked" out of filter materials. EPA
data ' show that well controlled plants can make this potentially large
emission source an insignificant one by direct and indirect steam distillation.
Other options for disposal include recovery off site by a solvent
disposal vendor and cartridge filtration. Cartridge filters have inherent
design advantages (they are confined and contained) which give them a low
emission factor (1 kg/100 kg) when properly drained and dried and are
applicable to low soil loadings such as commercial operations.12
Solvent losses from distillation bottom disposal can be reduced in
oil cookers (similar to muck cookers) to levels well below 1 kg/100 kg of
clothes cleaned by proper operation of existing equipment according to a
test conducted by EPA.13 Operators should avoid premature shutdown of the
distillation unit.
3-9 '
-------
There would be no additional space requirements for filter units in
perch!oroethylene systems and, of course, no additional space would be
required to improve the "cooking" of existing distillation systems.'
3.3 SUMMARY
This chapter has discussed control techniques for both existing and
new perch!oroethylene dry cleaning plants. Carbon adsorption can be used
to control perc vapor vented from the washer, dryer, storage tanks,
distillation systems, and chemical separators to less than 100 ppm.
Incineration does not appear applicable to synthetic solvent plants
because of associated environmental penalties.
Muck cookers are generally used in perch!oroethylene plants and, if
operated properly, maintain losses at less than 1 kg/100 kg of clothes
cleaned. Drained and dried cartridge filters achieve less than 1 kg per
100 kg of clothes cleaned based on EPA tests and thus are another effective
means of control of this source.
Miscellaneous emissions can be controlled through the use of better
housekeeping—aided by portable, inexpensive monitors (to be developed
by EPA).
In short, the emissions from dryers, washers, distillation units,
*s
holding tanks, filter systems, and fugitive emission sources can all be
controlled by the above named systems. Table 3-4 summarizes sources,
applicable control techniques, and achievable emission levels.
3-10
-------
OO
_l
UJ
LU
OO
oo
LU
LU
O
00 C£3
Q i— i
LU
OO _J
LU CJ
O
LU
CD
eel
•=*•
CO
a>
r—
-a
(O
-a
cu en
i — r—
O CU CD
S- > 0
+J CD , —
E 1 -v^
o cr>
C_5 NX
CU
cr
E
_E
0
cu
\—
o
S-
E
0
o
en
f— vx
O i—
•r- CU 0
to > o
to cu r—
i """ ~o>
LU -\t
CU
0
s_
0
oo
E
re
i—
D_
^~
o
cu
CL
>}
1—
O
1
LO
•
O
s_
a>
JD
S-
o
CO
-a
CO
E
0
JO
S-
fO
O
CM
i-
CU
JE:
co
ro
CU
5
-j ^
CD
O
O
r—
O
s_
cu
o.
0 0
r— i— •
10 10
0 O
i- to
O CU E
O CO T- • •!—
s_ +j a: •<->
S— C7> CU 4-5 (O
CU E 4-> CD to i —
S- E -r- -r- 3: ^
O *r— t^- v/* j ^
to ro o s- to
•a s- cu o cu T-
res T3 en o +J ~o
TD r—
C S- ••- S- -r- S-
O CU i- CO lj_ CU
JQ Q.+J C7) CD
i- 0 S- E 5- E
"3 t- to O O O
CJ> Q. o i — 4- I
LO IO
IO i-^ i-^
O)
0! 3
•(j -a
CO •!—
no to
3: CU
£_
i- CU r—
CU 4-> r—
S- -i— 4J
Q LL 00
00
1
CO
E
Q.
CU
CU
a>
to
0
c-
T3
O
0
ro
1
to
o
a>
E
to
^_
CU
o
to
•r"
s-n
-------
3.4 REFERENCES
1. Barber, J.W., Research Director, Vic Manufacturing Company,
Minneapolis, Minnesota, letter to C.F. Kleeberg, EPA, February 6, 1976.
2. Discussion with William Fisher, IFI, Silver Spring, Maryland,
August 7, 1975.
3. Anonymous, Dow Chemical U.S.A., "Poor Solvent Mileage - Professional
Dry Cleaning Plants," submitted by Bob Lundy, Dow Chemical to Charles F. Kleeberg,
EPA, March 16, 1976.
4. Reference 3, Op. Cit.
5. Anonymous, Hooker Industrial Chemicals, Bulletin Number 185,
"Hooker Handbook for Dry Cleaners," p.10.
6. Vic Manufacturing Company, "Installation and Operation Instruction
for Vic Models 221 and 222," VMC 1195.
7. Reeves, H.E., "Causes of Excessive Loss of Perch!oroethylene,"
IFI Practical Operating Trips Bulletin, p.91, January, 1969.
8. Anonymous, Dow Chemical USA, "Dow Customer Survey," submitted by
Bob Lundy, Dow Chemical to Billy C. McCoy, TRW Services.
9. "Chlorinated Solvents: Toxicity, Handling Precautions, First-Aid,"
Dow Chemical, U.S.A., Form No.lOO-5449-74R.
10. Kleeberg, C.F., "Testing of Commercial Perch!oroethylene Dry
Cleaner," test report on Hershey, Pennsylvania, test, to James F. Durham,
EPA, May 14, 1976.
11. Kleeberg, C.F., "Testing of Industrial Perchloroethylene Dry
Cleaner»" test report on San Antonio, Texas, test, to James F. Durham, EPA,
May 14, 1976.
! 3-12
-------
12. Kleeberg, Charles F., "Testing of Commercial Perch!oroethylene
.1
Dry Cleaner," test report on Kalamazoo, Michigan, test, to James F. Durham
EPA, May 17, 1976. ;
13. Ibid, Reference 11.
3-13
-------
o
-------
4.0 COST ANALYSIS i
4.1 INTRODUCTION ,
4.1.1 Purpose
The purpose of this chapter is to present estimated costs for applying
emission control techniques to perchloroethylene dry cleaning systems.
Cost data will be supplied for hydrocarbon control at perchloroethylene
solvent plants.
4.1.2 Scope , ' ,
Control cost estimates will be presented for three types of facilities
using perchloroethylene solvents: coin-operated plants, industrial plants,
and commercial dry cleaners. These estimates will reflect the retrofit
control cost of carbon adsorption for control of washer and dryer emissions.
No incremental costs fpr housekeeping controls are presented.
4.1.3 Use of Model Plants
Control cost estimates are presented for typical model plants in the
dry cleaning industry. Specific model plant parameters will be presented in
subsequent portions of this chapter. It is admitted that control costs at
actual installations may vary, sometimes appreciably, from the costs
described for the model plants. However, the difficulty of obtaining
actual plant control cost information makes the use of model plants a
necessity. To the extent possible, EPA has incorporated: actual plant cost
information into the cost analysis.
Cost information is presented for typical existing model facilities.
i
In some cases, model plants of varying sizes have been developed. The
purpose of this is to show the relative variation in control equipment
4-1
-------
costs with plant size. Whereas the plant sizes chosen for analysis are , W
believed to be representative of plants in the industry, no attempt has
been made to span the range of existing plant sizes.
4.1.4 Bases for Capital Cost Estimates
Control cost estimates are comprised of installed capital costs and
annualized operating costs. The installed capital cost estimates reflect
the cost of designing, purchasing, and installing a particular control
device. These estimates include costs for both major and auxiliary equip-
ment, removal of any existing equipment, site preparation, equipment
installation, and design engineering. No attempt has been made to include
costs for lost production during equipment installation or start-up. All
capital costs reflect first quarter 1978 costs. In general, information
on capital costs for alternative control systems has been developed through
UUHUW^;. ni»ii «,«,,„,«, v^M.^...-..- ,^,.~~.*. „.. ,
from EPA files has been used along with data from previous contractor
studies of the dry cleaning industry.
4.1.5 Bases for Annualized Cost Estimates
Annualized cost estimates include costs for operating labor, mainte-
nance, utilities, credits for solvent recovery, costs for waste disposal,
and charges for depreciation, interest, administrative overhead, property
taxes, and insurance. A return on the pollution control investment is
not included in the annual cost estimate. All annualized costs reflect
*•.
second quarter 1978 costs. Operating cost estimates have been developed
by EPA from in-house files. Credits for solvent recovery have been calcu-
lated based on emission factors presented in Chapter 3 and the current
market price of $0.49/Kg for perchloroethylene solvent. It is estimated ^
4-2
-------
that this solvent price could vary 2Q% depending on location and
quantities purchased. Estimates of depreciation and interest costs
have been calculated by EPA by using a capital recovery factor based
on the assumptions of an interest rate of 10 percent and a depreciable
equipment life of 10 years. In addition to costs for depreciation and
interest, an additional charge of 4 percent of total capital has been
added for administrative overhead, property taxes, and insurance.
4.2 PERCHLQROETHYLENE SOLVENT PLANT COST ANALYSIS
4.2.1 Model Plant Parameters
Control costs have been developed for three types of perchloroethylene
solvent dry cleaning plants. These are coin-operated plants, commercial "
plants, and industrial plants. The model plant parameters that were
developed for these facilities are displayed in Table 4--1. The model
perchloroethylene plant parameters are based upon industry contacts and
EPA studies of the industry. Typical plant sizes for perchloroethylene
solvent plants are two 3.6 Kg unit in a coin-op store, one 11 Kg unit in
a commercial plant and one 93 Kg unit in an industrial plant.
4.2.2 Control Costs - Perchloroethylene Plants ;,
Costs for control of washer and dryer emissions from coin-operated,
commercial, and industrial perchloroethylene solvent plants have been
calculated.
Table 4-2 presents costs for carbon adsorber controls for five sizes
of model new and existing perchloroethylene plants - 3.6 Kg/load, 11 Kg/load,
23 Kg/load, 91 Kg/load, and 114 Kg/load. Costs are presented in terms of
installed capital costs, annualized costs* and the cost per kilogram of
4-3
-------
fO
CO
0)
J3
CO
1 _1
CO
ol
•r-
S_
to
3
-O
t— 1
CO
si
CO
•r-
o
cu
o
Q.
O
1
'o
0
CM
i^ o «*
5 9. 2 ~
LO
r— • "^ •*
cn 'co g CM
•* ^
co 'co °"> •=*•
CM r- CM CM
-r—
CM
r— O _. ^1"
r— i — O> CM
CM LO
CM
CO LO ^O *5f
CO CM «*• CM
•p Oi LO '
E
3 10
CM CO
^—s ^-. CU
cn t/2 S-
5. §5
C ro
•P (/) CU
, i- -a s ex
o co o E
CO O i — CU
CO
:CU t/1 to
-O -E 3 3
CO to co CO
O tO .E -E
r- 3 X X
CU CO S- J- -O
_E -3 O) CU 0
to .E >, >»,> —
CO E S- S-
3
LO "vf
CO O •
cr»
LO «*•
co o
LO <*•
co o
J3
to
0
T3 -r—
E tO -
CO 0 3 -P tO -fa*
S'^ E-^ wi ~
-O 01 'T3 — - "O CU -P
CU •!- ,CU CD CU E to
i— E ii— i^ r— CO 0
r— CU i— O i— r— O
O O O O r—
i- i- 5-i— S- CU -P
•P CU -P
E >> E
0 S- O
o -o o
E E
Z> ^3
•^ -P O E
cn E
_- u E i—
E 0
ro oo
•-*£
Q.
"O
c:
CO
an
Q)
•g
S-
O
o
cc
:o
o
o
cO
+J
0
O
3
-o
-a
(U
to
CO
jQ
O.
to
.
CO
•=C
D-
UJ
•a
cu
•p
o
CO
LO
I
co
Q_ CU
O JQ
X *
a>
o
O
co
-------
UJ
o
co
UJ
LU
UJ
o
O
a:
o
u_
Q.
cz.
o
oo
Q
a;
S
a:
S
oo
oo
CM
cu
•—
01
CO
CM
i— •
*~
•
ca> ta
cu
to
•1-3 Q}
E
3 VO
.
CM CO
^^
cn
C5
CU
N
•r-
OO
•i~
E
•r-
s-
•M
3
•a
i— )
r—
fO
•1—
3
•o
E
HH
(O
•r—
£
CU
£^r
£^
O
c_>
res
'o
S-
O)
0
0
Q.
0
1
E
•r—
O
O
> >>
O-o o'
0^ 0
to O' — »O
Oj '— •• *O- i--OO-
'J3 o -^?x~'
•T- 0 -4-> O *-*•
__ -feo- to o -JJ
. .7Z ^^ 0 0 -r-
o O -y)- ^3
frt r^ ^""^ O
"" -2 E* 10 CJ
c^ ^- "^ O>"^^
. .;: « to s- •*->
40 CJ S- n3 c/1
.Jf -a a. o o
5 cu o
1,1 r™ p— • «fJ
r— 4-> CO E
, fO O 4-> CU
^ +J CU -r- >
•^-j tO S- CL^-
O E "I— CO O
gr I-H Q CJ OO
CO CM ^
J£? CO °
*^-^
0? 00 '£o
**~^* CM ii
, *^^V '
CO CM O
• * •
CD OO. Q
^__ • •
LO ^o co
o •—
oo
CM CO oo
1 CVI P^^
O)
IT
-Q? > T3 ~a
to — cu cu*— *
3 O •— S- 0)
E O r— O i^
E O O ' — .
— ' +J +J
•«-> E W)
CU O O
2: ocj
••
CU
>
+»
ro
S-
•1-3
• l_
E
| • 'i
~a
ro
-o
C
cu
o
c
.^_
'3
!E
40 «*
a; i
! X CO
i iB
4-3 CO
1^ pH
!w «^»
O CO
o-S E
-1- E
r/> >
8. -i-
S m
.>> C/)
S-
u
!- CO
M—
c-
5^ O
CD T-
— V)
to
r-™,tG2"i
• -ID CU
a> cu aj E
E E E CO O
CU CU O) i —
s- s- s_ :a-a
cu cu cu u cu
4- q- q_ r- w
CU CU O) CO CO
a; cc c£ o ca
».*««•
"3 -Q CJ "O CU
4-5
-------
hydrocarbon controlled for the different sizes. Note that for carbon
adsorption that not only are total capital costs and total annual ized
costs presented but also presented is information on the cost-effectiveness
of each size plant. For example, for the model 11 Kg/load commercial
plant the total installed capital cost from Table 4-2 for a carbon adsorber
is estimated to be $3,200, the net annual ized cost is estimated to be
$500/year, and the cost per kilogram of hydrocarbon controlled is estimated
to be $0.31 /Kg.
This estimate of $0.31/Kg is determined by dividing the net annual ized
cost of $500/year by the controlled emissions of 1600 Kg/year. The
controlled emissions were determined from Table 3-4. Table 4-3, which
shows that control option #1 for perch! oroethylene solvent plants
combines carbon adsorption, waste solvent disposal, and good housekeeping
practices. The recovered emissions are 6.5 Kg/100 Kg (11.5 Kg-5 Kg). An
11 Kg load system doing 2210 loads per year (Ref. Table 4-1) cleans
24,300 Kg/year of clothes. Multiplying this figure by 6.5 Kg/100 Kg results
in controlled emissions of approximately 1600 Kg/year.
It should be noted that emission reductions attributable to housekeeping
controls have been included in some control options. As stated before,
however, no costs for housekeeping controls have been included since they
are believed to be adequately accounted for by the charge of four percent
of total capital that is allocated to all control systems to cover adminis-
trative overhead taxes, and insurance and the 6 percent of total capital
allocated to cover operating and maintenance. Also note that costs for a
carbon adsorber for 3.6 Kg plants are larger than carbon adsorber costs for
4-6
-------
oo
1 —
1
Q_
•;y
>
|
o •
LU
z
LU
1
>-
~T—
^_
LU
O
Qi
o
|
O
Di
LU
n.
CD
Z
i— i
I >
i^
^*
X
LU
r*^
O
1 1
oo
O T?
o; cu
i_1 _n i
S i —
0 0 .
0 i.
[-TjJ £~
> 0
i — i o
1—
^y xx
LU ^
|_ * s
•=t
U_
0
00
oo
1 1 1
•^r
LU
>
|_}
i t
r~
1,1
u_
1 1
LU
I
1 |
oo
o
o
0
OO
1
CD
fO
I —
^J-
f—
r— -
cri
co
CM
r—
r~~
c£> n3
cu
V)
•P en
E
3 <£>
CM OO
CU
N
00
E
n3
Q.
r— •
ro
il
4-3
to
^3
E
l — l
^^^
^~"
- o
o
V — '
tO
•r—
S-
V)
^5
t^j
C
H-l
CO
ro
e ° '"
r—
O3
•r^
O
<1)
£
o
o
*•— •%
o^
. 0
o
eo
•i—.
O
CU
i
o
,
OO O
*
o'
Q.
o
1
E
•r—
O
CJ
ro *
oo o
to
cu
S- 0
Or- O
5- 03 (O
>i Q- to s- >>
•— O Q. r—
cu Cn «> Q. cu cn
-P ^ E tO CD 4-J •*/
— E ro
E O -i — O T- E O
•r- CD -P Q. -i- O
X •— Q--P CU X t—
O -^^ S- E CU O ^v.
s- cn o cu -i
Q i«i tO > CU Q. ^^
Q "O i — to o.
=^C LO fO O 13 ea; CM
to o i —
E .E
CO •• O CU ••
O- r— •• .Q-PTD CM-'
,>) =tt= o s- to o =tfe o
l — -P to us o -p
E o 3 cn s=
-P O i— o r-
E T- (O . -t- rtS
' — Q ,CT Q. or
a- c5)uj O]LU
s- cu
-O i—
CU "3
to -i—
CU M-
5 -P
E
if- tO
o =s
cr
cu
CJ -P
E O
"3 E
E
CU CU
•P S-
E (O
•i—
(O to
to
T3 O
CU O
in
ea cu
cu vt
S- CU
CJ JE
E -P
•r-
S-
O CU
CU
cu s
rs o
to
10 E
o o
O *I—
CU D.
e o
o
to to
S- -P
13
0 S-
E CU
•r- -O
E
E to
to cu
S- -P
cu to
E >>
S to
0
cn
a. E
O-r-
1 E
E
-------
either an 11 Kg plant or a 23 Kg plant. This is because it was assumed
that the 3.6 Kg plants would not have steam available for regeneration
of the carbon but the larger plants would have such capacity. Therefore,
it was necessary to include the cost of a small steam boiler in with
the cost of the carbon adsorber for the 3.6 Kg plants. In the case of
the larger perch!oroethylene plants it was assumed that steam was available
and no costs were included for purchase of a boiler.
4.2.3 Cost Effectiveness
Summary costs in terms of the cost per kilogram of solvent emissions
controlled is presented in Table 4-3 for different size perch!oroethylene
plants. Control costs decrease rapidly as the size of the unit controlled
increases. For example, carbon adsorber controls cost $7.33/Kg in a 3.6
Kg/load facility but decrease to a net credit of ($.41/Kg) for a 114 Kg/
load facility.
4.3 REFERENCES FOR CHAPTER 4.0
(1) Cost data and equipment brochures furnished by Mrs. Pat King, Executive
Assistant, HOYT Manufacturing .Corporation, and Mr. Peter Zizzi, Sales
and Service Engineer, Fulton Boiler Works, Incorporated.
(2) Information furnished fay Mr. A_ C. Cullins, Laundry and Dry Cleaning
Consultant, Standard Laundry Machine Company, Inc.
(3) Virginia-Carolina Laundry Supply Company, 639 Junction Road,
Durham, North Carolina.
(4) Operating cost based on projections of equipment brochures and
specifications furnished by Vic Manufacturing Company, 1620 Central Ave
N.E., Minneapolis, Minnesota 5541.
O
4-8
-------
5.0 EFFECTS OF APPLYING THE TECHNOLOGY
The air pollution impacts and the other environmental consequences of
applying the control technology presented in Chapter 3.0;are discussed in
this section. A comparison will be made between emissions from a typical
uncontrolled plant and those from plants using alternative control techniques.
Beneficial -and adverse impacts which may be directly or Indirectly attributed
to the operation of these systems will be assessed.
Both direct and indirect impacts are involved in the control
of dry cleaning plants. For example, reduced air emissions, increased water
consumption, and increased energy demand are all impacts directly related to
the use of carbon adsorption recovery systems. Incremental emissions from
a boiler used to supply additional steam to the adsorber are an indirect impact.
5.1 IMPACTS ON VOC EMISSIONS
Pollutant emission factors for the individual uncontrolled plant
are shown in Table 5-1. They are based.upon data from the literature
(including trade associations, '2 equipment vendors,3 and solvent
companies4) and from stack test data5'6'7 obtained during this study.
Table 5-2 shows the individual sources of emission within the plant
and the achievable level with applicable control technology for each source.
The methods include carbon adsorption for washers and dryers;; longer dis-
tillation times for distillation units; longer cooking times or cartridge filter
substitution for filter muck; and leak prevention measures for miscellaneous
losses.
5-1
-------
to
»-*>-
srt-
LU
O
ZTUJ
0—1
CQ—J
Q£0
"fc
§1
oo
in
cu
S
rO
S-
•**^
JQ
r~
•a >>
2 oi
r— -^
O
S-
c -o
o cu
o f- c
c; o ret
13 CU
-^ o
O I/)
o cu
en o
"o
3
O.
0>
3
O
C"
f—
4->
S=
5 >>
D) -Q
**S g—
0 0
0 0
co o
f-T «3
r^
*~^
CU
cu
">^
r~t
•P
CU
o
s-
o
*^
o
cu
Ou
CL
0
1
•t—
o
o
o
§
o
CM
•t
£•
o
CM
CO
CVJ
f—
S- S-
>) >>
oi JD
v* r™*
O 0
O 0
CD O
f>» O
CM IO
CU
c
cu
^
f
+J
cu
o
S-
o
*^
o
cu
D_
-
^-«
(O
•r—
O
s.
cu
o
o
o
o
•t
CM
O
0
ft
CM
CO
CM
^—
s- T?
O) .0"
*^ r~*
O O
o o
0 CD
o cT
r*^> CD
CM <£>
cu
cu
">>
s:
+j
cu
o
S-
o
*7^
o
S-
cu
Q.
^-«
(0
S-
(/)
•a
HH
5-2
-------
a: Q.
LU o
_J O
co s:
CO
UJ
cs cr
n:
O
U_ O
CO O
LLJ CJ>
CM
I
LO
CU
1
•P to
to cu
CO -P
o
CU i—
E O
CO
•P 01
CU -M
0
S-0
O O
r— r—
O O)
CU
Q.
-O
CU
"o
S-
E
O
O
"S
_
0
•P
c
O
O
s- .•• • '
cu
-Q
s-
o .—.'
to ro '
^y J^
VO Lf> Q r— CM
E • • . I
O O i — r— O4-> O *—
* CO
o tz*
1
*o oo
• I
cu
• •- =3 ' . .
o -0 "*
3 • 4-> J-^ C°
r— r— -i— tO
U. Lu CO O
O)
T3 CU T3 E
CU CD CU rtj
•0 T3 -0 r-
S- -i- • S- J- r—
>- > to S- co to to
a 5 ' o S 3 s
LO
V
CM
1 — • -
t
1
to
cu cu
JC -l->
•»-> (O
E CU
«4- -i- 4->
O 4-> (O
«/> E
CU CU f-
to x •
3 OJ O TD
to t/> S- CU
O CO Q.+J
CO J= Q. (O
JQ I— (O £
10
cu
to
s-
+J +J
E co
, CU CU
to
(O
3
-a
a. n3
cu
s- w
cu
T3 O
E S-
-Q , O
O S- CO
CO -r-
co
•r- M
E CU
CU O
-a co
o s_
i— CU
r— t*-
o »«-
S- T-
•(j -a
£Z
o a>
o -o
cu
4->
o
to 4->
3 tO
T3 O
E E
•r—
S-
E O
O «t-
T3 CU
CU 01
10 E
»O to
-a .*-
cu -a
re 'E
5-3
-------
5.2 OTHER AIR POLLUTION IMPACTS
There are no other air pollution impacts associated with any of the
control techniques.
5.3 WATER POLLUTION IMPACTS
]
Dry cleaning processes usually discharge some water to sewage facilities.
Perch!oroethylene plants use water cooled condensers. Some plants have water
washes to remove soluble soils. Many perchloroethylene plants use steam for
heating dryer air, presses and finishing equipment, and for distillation or
muck cooking purposes. The air pollution control systems envisioned for
dry cleaning facilities will add to the amount of water used as indicated
8 9 10 11
in Table 5-3. Data are based on plant tests *y*tv and vendor submittals.
It should be noted that increased water usage is estimated only for those
sources where water may come in contact with solvent. This does not include
condenser water which will total about 750 liters per day for a commercial
system.
The primary addition would be the steam required to regenerate the
carbon. Typically about 45 kg of steam is required per 100 kg of clothes
cleaned. Condensate is generally disposed of by sewer (about 55 liters per day).
Also shown in Table 5-3 is the steam (and thus water) required for a
muck cooker or distillation unit. These units are generally present in
perchloroethylene plants.
EPA has not promulgated or proposed effluent guidelines for dry cleaning
solvent content in waste water streams. During plant tests for this project,
EPA took water samples of streams from carbon adsorbers and found them to
contain less than 100 ppm perchloroethylene by weight. The effluent was
disposed of in sanitary sewers.
5-4
-------
LLJ
S!
oo
UJ
g
o
oo
00
oo
o
I—
o
o
o
Q.
o:
co
LO
CU
r—
ja
re
(/)
CU
o
CU
co
re
to
S-
>«• ^^
CU ^^x
-u co
re -^
3
c/>
-o
0
CU
s:
Control
CU
Q-
^s.
1—
^_>
C
re
r—
O.
CO
•^
O
o
^^
o
CM
CO
~^
to
^J"
<*-
o
CU
-M
re
S- W
CU
S- J=
CU -M
•4-> O
rer—
3E 0
0 00
o
vo
A
r—
V)
CU
o
o
re
i. i-
CU Q-
-Q CO
S- C CO
O i- C
tO >) •!-
-0 S- G.
re 73 cu
CU
£ S- -^
O CU CU
•e £ s
re T- o
o u_ n:
cu
cu
J?
4-^
cu
o
S-
O i—
i— re
^: «t-
CJ. O O
OS- S-
1 CU CU
CQ. s
O CJ
COS-
JM: cu
o o
0 0
r— O
0.*: T3
CMO CU
3: 3 E
E 3
CO (/)
•^ S- t/1
o re
LO M-
«* CO
o cu o
•4->0
' CU W r—
^_) ^^t
re »*»o
S- tO CM
cu 3:
S-.C
cu 4J co
•M O -^
re r—
3 0 LO
000
LO LO
i— CO
*» <\
CM i—
r—
c_
cu
S. CO
0 S- C
to CU •!-
•O -^ Q.
re o cu
o cu
C O -i£
o cu
-Q -^ ui
1-03
re 3 o
0 S Z
cu
cu
I
•M
cu
o
S-
o
-C
0
cu
Q.
co i-
-^ CU
o o
0 0
r— O
O ^ "O
CMU CU
2C 3 E
E 3
NX ^ lyj
o re
LO <4-
«* CO
«t- 1^
o cu o
+J O
CU W r—
re »*»o
S- to CM
CU T*
S- t~
CU •«-> CO
•»-> O -M
re r—
3 O LO
S °
LO
oo
i1
CU •!—
jii Q.
o cu
o cu
o ^
cu
^c to
O 3
i 5
cu
C
cu
S
cu
o
s_
i— O
re i—
•r- ^T
^ "
to cu
3 Q-
T3
c:
1— 1
o
o ;
LO
r—
t
CO
JD
o
T3
O ;
J_
re
5-5
-------
Table 5-4. PERCHLOROETHYLENE SOLVENT IN EFFLUENT WATER
AS A RESULT OF CARBON ADSORPTION (MODEL PLANTS)
Increased water usage Solvent disposed of
.(kg/year) (kg/year)
Coin-Op 1,600 0.2
Commercial 13,500 1.4
Industrial 135,000 13.4
Assumes 100 ppm in effluent
4ft
-------
Using the figure of TOO ppm in water for perch!oroethylene plants,
Table 5-5 shows that 13.5 kilograms of solvent per year will be added to
effluent from typical industrial plants; less from coin-op arid commercial
establishments. ;
5.4 SOLID WASTE IMPACT
There is little solid waste impact associated with afr pollution control
techniques. Carbon in adsorbers eventually must be replaced because of
"blinding" of the bed by small pieces of lint and other particulate. Vendors
and users have estimated the life of carbon at up to 30 years. The carbon can
be regenerated, but may be discarded every 15 years. Each commercial perch!oro-
ethylene plant uses around 100 kilograms of carbon. Large industrial perc
plants use up to 450 kilograms. The solid waste impact from the entire industry
is estimated to be insignificant—even if all plants used carbon adsorbers.
The techniques used to reduce emissions from solvent .filters do not
increase solid waste at all; they do reduce the amount of solvent in discarded
muck .and filters. The emission reduction from control of filter disposal is
part of the total emission reduction shown in Table 5-2.
5.5 ENERGY IMPACT
Certain control techniques require additional energy.. Carbon adsorbers
require steam for desorption. Muck cookers and distillation oil cookers both
require steam, but in many plants already equipped with boilers the energy
increment is small.
5.5.1 Impact on Model Plant
Table 5-5 shows the energy impact of the above alternatives on model
plants. There is also the possibility of an energy credit from the decreased
use of solvent which would be a result of these alternatives if implemented
in the plants. It is estimated that at least one kilogram of fuel would be
5-7
-------
Table 5-5. ENERGY IMPACT OF ALTERNATIVE CONTROL
LEVELS ON TYPICAL PLANT
Plant
Control
10 BTU/yr
Usage
10 BTU/yr
Savings
Net Energy
Usage (savings)
106,BTU/yr
Coin-Op
Perch!oroethylene
Commercial
Perch!oroethylene
Industrial
Perch!oroethylene
Carbon adsorber
Muck cooker
6.6
Carbon adsorber 27
Carbon adsorber 270
(25)
(45)
(430)
(18).
(18)-
(160)
5-8
-------
required to produce one kilogram of solvent. Table 5-5 shows this solvent
savings as an energy credit for each plant. (Actually, the energy savings
would be creditable to the solvent producer.) Net energy consumption is shown
as a savings.
5.5.2 Impact on Indirect Air'Pollution" ~
Increases or decreases in steam demand as a result of applying the
i
control techniques will influence emissions from the boiler plant. These
emissions are considered insignificant. \
5.6 REFERENCES
1. Watt, Andrew, IV, and William E. Fisher, "Results of Membership
Survey of Dry Cleaning Operation," IFI Special Reporter #3-1, January-
February, 1975.
2, Fisher, William E., "The ABC's of Solvent Mileage," Part One,
IFI Special Reporter #3-4, July-August, 1975.
3. Barber, J.W., Research Director, Vic Manufacturing Company,
Minneapolis, Minnesota, letter to C.F. Kleeberg, U.S. EPA, February 6, 1976.
4. Anonymous, "Dry Cleaning Industry Statistics," submitted by
Robert Lundy of Dow Chemical from Dow Survey. <
5. Kleeberg, Charles F., "Material Balance of a Perch!oroethylene
Dry Cleaner Unit," test report to James F. Durham on test in Hershey,
Pennsylvania, March 17, 1976.
6. Kleeberg, Charles F., "Material Balance of an Industrial,
Perch!oroethylene Dry Cleaner," test report to James F. Durham, on test
in San Antonio, Texas, May 14, 1976.
7. Kleeberg, Charles F., "Material Balance of Small Commercial
Perchloroethylene Dry Cleaner," test report to James F. Durham on test
in Kalamazoo, Michigan, May 17, 1976.
5-9
-------
8. Ibid, Reference 5.
9. Ibid, Reference 6.
10. Ibid, Reference 7.
11. Vic Manufacturing Company, "Model 221 Strato Dry to Dry Series
Specifications," Form Number 221-1083, June, 1971.
5-10
-------
6.0 ENFORCEMENT ASPECTS
6.1 AFFECTED FACILITY
l
In formulating regulations it is suggested that the affected facility
be defined as the dry cleaning system which includes: washer, dryer, filter
and purification systems, waste disposal systems, holding tanks, pumps, and
attendant piping and valves. This definition would cover all significant
VOC sources of emissions in perch!oroethylene plants.
6.2 SUGGESTED REGULATION i
The ease of determining compliance is the most important consideration
in development of regulations for such a prevalent source as perchloro-
i ~
ethylene dry cleaning. The following example regulation ^outlines the
method deemed to be optimum for reducing emissions. •
Rule of ^Air Pollution Control District
Sec- 1» Solvent emissions from perchloroethylene dry cleaning systems
must be limited in accordance with the provisions of this Rule.
Sec. 2. Compliance with this Rule requires the following:
. I1
(a) There shall be no liquid leakage of organic solvent from
the system. . .. ; .
(b) Gaseous leakage shall not exceed ppm.-/
V The EPA is currently assessing the significance of vapor leaks. If
deemed significant, a test method for detecting leaks will be developed
and issued to interested parties. '
6-1
-------
I !
5
(c) The entire dryer exhaust must be vented through a carbon
adsorber or equally effective control device.
(d) The maximum organic solvent concentration in the vent from
the dryer control device shall not exceed 100 ppm before dilution.-
(e) Filter and distillation wastes.
(1) The residue from any diatomaceous earth filter shall be
cooked or treated so that wastes shall not contain more than 25 kg
of solvent per 100 kg of wet waste material.
(2) The residue from a solvent still shall not contain more than
60 kg of solvent per 100 kg of wet waste material.
(3) Filtration cartridges must be drained in the filter housing
for at least 24 hours before being discarded. The drained cartridges
should be dried in the dryer tumbler after draining if at all possible.
(4) Any other filtration or distillation system can be used if
equivalency to these guidelines is demonstrated. For purposes of
equivalency demonstration, any system reducing waste losses below
1 kg solvent per 100 kg clothes cleaned will be considered equivalent.
Sec. 3. Sections 2(c) and (d) are not applicable to plants where an
adsorber cannot be accommodated because of inadequate space or to
plants where no or insufficient steam capacity is available to desorb
adsorbers. The District may exclude other plants from the scope of
Sections 2(c) and (d) if it appears that other hardships justify such ^
an exclusion.
2/ Enforcement of these provisions is dependent on the development of a
satisfactory detector and of test methods.
1
1
i
6-2
-------
Sec. 4. Compliance Procedures -
(a) Liquid leakage shall be determined by visual inspection of
the following sources:
(l) Hose connections, unions, couplings and valves;
(2) Machine door gasket and seating;
(3) Filter head gasket and seating;
(4) Pumps;
(5) Base tanks and storage containers;
(6) Water separators;
(7) Filter sludge recovery;
(8) Distillation unit; .
(9) Divertor valves; '.
(10) Saturated lint from lint basket; and
(11) Cartridge filters.
(b) Vapor leakage shall be determined by __ .-/
(c) Dryer exhaust concentration shall be determined by j/
(d) The amount of solvent in filter and distillation wastes shall
be determined by utilizing the test method described by the
American National Standards Institute in the paper, "Standard
Method of Test for Dilution of Gasoline-Engine Crankcase Oils."
3_/ See footnote 1, above.
4/ See footnote 2, above.
6-3
-------
6.3 DISCUSSION
(Sec. 2.b) As noted, the EPA is now assessing the significance of
vapor leaks in perc dry cleaning systems. If deemed significant, then an
inexpensive monitor will be developed which can be used by operators and
enforcement personnel to locate major vapor leaks. If deemed insignificant,
Section 2.b can be deleted. The study of vapor leaks should be completed
by May 1979.
(Sec. 2.d) Carbon adsorbers tested by the EPA have achieved much
better control than 100 ppm outlet concentration. This figure was chosen
because it is high enough to indicate "breakthrough" of the carbon bed.
Breakthrough is a good indicator to enforcement officials of improper
maintenance or operation of the adsorber.
(Sec. 2.e) Figures given for filter and distillation waste disposal
are based on limited data and thus include margins of safety. A more
stringent standard may be achievable. For purposes of equivalency, waste
losses should be less than 1 kg of solvent per 100 kg of clothes cleaned.
(Sec. 3) Most coin-op cleaners are expected to fall under this exemption
clause since space and steam capacity are not usually available. While some
small commercial plants may fall under this exemption clause, most commercial
and industrial perchloroethylene cleaners should be able to comply with
Section 2 (c) and (d).
It is expected that because of the limited number of carbon adsorption
equipment vendors, there may be problems in obtaining delivery of control
equipment in the time frame outlined by State regulations. Regulatory agencies
should be sensitive to this problem and provide extensions to compliance
schedules where deemed necessary.
6-4
-------
APPENDIX A. EMISSION SOURCE TEST DATA
EPA planned to test only as many plants as necessary to represent
best available control in the dry cleaning industry. A number of
parameters which affect emissions presented themselves for consideration.
Dry cleaning plants differ in size, control techniques, design, capacity,
types of clothes cleaned, climate of locality, soil composition, age of
equipment, and maintenance history. The effect on emissions that some
parameters have is small. EPA tested typical plants in two of the three
industry sectors (commercial and industrial) as shown in Table A-l.
;|
TABLE A-l. PLANTS TESTED BY EPA
(kg capacity of washers given in parentheses)
Perchloroethylene
Coin-Op None j
Commercial X (50,18)
Industrial X (140)
A small and a large commercial perchToroethylene unit were tested.
A description of these tests can be found in Sections A.I;and A.2. The
difference between dry-to-dry and transfer units was explored in these
tests.
A large industrial perchloroethylene unit was tested. The test is
discussed in Section A.3. The unit was a relatively new design of transfer
machine in which the washer and dryer nearly touch during transfer, thus
reducing exposure time of the damp clothes to atmosphere.
A-l
-------
No coin-operated perchloroethylene machines were tested. No
adsorption systems in use on perc coin-ops in the United States were located.
All systems were tested by the methods discussed in Appendix B of this
document.
All systems were tested by the methods discussed in Appendix B of this
document.
A.I PLANT A
Plant A's commercial operation, which uses perchloroethylene solvent in
a 50 kg capacity machine, was tested by material balance (November 3-Npvember 20,
1975). The machine is a Washex SM-11 and was installed in 1967. The system
consists of a washer/extractor, muck cooker, two dryers, a regenerative filter
and a Vic dual canister carbon adsorber. The carbon adsorber collects emissions
from the washer door vent, the dryers, floor vents, and the distillation (muck
cooker) unit. EPA not only performed a material balance of the unit, but also
stack tested the carbon adsorber for perchloroethylene (by test methods also
described in Appendix B).
The plant used two operations which are not normally used in dry cleaning
services—fire-proofing and water repelling applications. The addition of these
materials was accounted for in the material balance.
Table A-2 summarizes data from each test in the dry cleaning test program.
It can be seen that emissions from this unit were about 4.1 kg of solvent per
100 kg of clothes cleaned. Outlet concentrations of perchloroethylene averaged
about 25 ppm. This means that solvent consumption in the whole process was
about 19 kg per day of which 1 kg was from the adsorber. Without an adsorber,
total emissions would have more than doubled.
A-2
-------
The adsorber was installed at least 15 years ago and the unit has had
one major "overhaul" since that date (due tovcojqrosion). It requires
approximately 9 cubic meters or about 4 square meters of floor space. The
original washer dryer system which the unit serviced was replaced in 1967.
This system demonstrated the performance of carbon adsorption as a
control technique. The carbon in this carbon bed is over 15 years old
and outlet concentrations are only 25 ppnfwhen tested", The^stenTsuffeFet from"
inadequate housekeeping, however. Liquid leaks were sighted and buckets
of perchloroethylene draining from water separators were left uncovered.
The scent of perchloroethylene was prevalent. EPA feels ;that operation
of this plant could have been improved by better housekeeping.
A.2 PLANT B
During the period April 7-20, 1976, a material balance was conducted
on a small, commercial dry cleaning operation using percKloroethylene
solvent (Plant B). A stack test of a carbon adsorber on the plant was
conducted during one day of testing by Midwest Research Institute. The
stack test involved integrated samples analyzed for total non-methane hydro-
carbons.
Plant B consists of a dry-to-dry Vic Model 221 Strato System of 18 kg
(40 pounds) capacity. During the course of the test, approximately 170 kg
(370 pounds) of material were cleaned per day. Table A-2 summarizes the
emission data taken from the material balance and stack test.
The system vents to a dual canister carbon adsorber from the dryer
(during the entirety of the drying cycle), from floor vents, and from the
washer door. Each carbon bed operates for one cycle of the washer/dryer
and then is desorbed during the next cycle. There were some indications
that the carbon beds were undersized. Limited data taken;from a semi-
continuous monitor indicate that breakthrough occurred on each bed during
A-3 '
-------
its cycle. Average non-methane hydrocarbon concentration in the exhaust
stream was 100 ppm. .
* \
A 14 cartridge (paper) filter was used to" purrfy solvent in the systems.
It was the only such purification device used in the system.
According to material balance and accounting for cartridge filter loss
prorated to the course of this test, the dry cleaning system at Plant B had
an emission factor of 2.1 kilograms of solvent used per 100 kilograms of
clothes cleaned (based on machine capacity). Approximately 3.6"klT6grams(7,"9
pounds) of solvent were lost from the system per day. Of this 3.6 kilograms,
the carbon adsorber lost 1.2 kilograms (2.6 pounds) at an average outlet
concentration of 100 ppm. The cartridge filter accounted for an estimated
0.7 kilograms (1.5 pounds) loss per day.
The adsorber was built in as an integral part of the unit. It requires
about 1.4 cubic meters of space or about 1 square meter of floor space.
A.3 PLANT C
Plant C is an industrial dry cleaning plant using perchloroethylene
solvent. It is an American Laundry Machinery system which includes washer/
extractor, a "kissing" dryer, distillation unit, chemical separator, oil
cooker and single bed carbon adsorber. The adsorber collects emissions from
the washer and dryer. The capacity of the washer is about 140 kg per load
but shirts are loaded at about 90 kg per load because of the number of articles
per kilogram. Pants are loaded at capacity.
The "kissing" washer/dryer is a relatively new innovation in the industry.
At the conclusion of washing, the dryer is pneumatically rolled to within 0.3
meters of the washer, both doors are opened and the clothes are transferred
by tumbling. This design greatly reduces the time that solvent laden clothes
are exposed to the atmosphere.
A-4
-------
EPA performed a material balance on the system and also tested the
carbon adsorber. Results of the test are shown along with other tests
in Table A-2. The table shows that solvent usage was a very low 2.5 kg
of solvent per 100 kg of clothes cleaned. The entire systeim lost about
40 kilograms of solvent per day of which about 0.1 kilograms were emitted
from the adsorber. Most of the losses were accounted for in a special
washer loading exhaust and in a distillation unit vent. Both were vented
to atmosphere and emitted approximately 24 kilograms of solvent per day.
System changes were being initiated to vent these two sources to the adsorber.
The outlet to the carbon adsorber averaged around 3 ppm.
Both the material balance and the adsorber test demonstrated the
efficiency of this system. Exemplary housekeeping practices were followed at
[
the plant and attention was paid to methods of improving performance. The
equipment was installed from 1970 (washer, distillation unit, and oil cooker)
to 1975 (kissing dryer in early 1974 and the carbon adsorber in May, 1975).
No solvent leaks were detected by sight or smell.
The adsorber required about 20 cubic meters of space and about 6 square
meters of floor space. It was retrofitted in 1975.
A-5
-------
LT>
ro
*\
CM
CO
LU
CO CO r—
O >>•! >»
(— co E cn
III \S
CO =5
CO Q.
JC JC
4-> cn
O 3
r— O
0 S-
1—
E
d)
^_3
(/)
>j
t/>
4J
fO
Q.
*£>
•
^t"
f^.
CD
O
CO
0
CM
O
r—
,-j!
to
•a
^*^
cn
vx
LO
LO
O)
CO
JC
0)
0 S-
i- CO
o q-
i — CO
JC C
o to
s- s-
a.
ef.
CO
CM
i— •
O
o
•
o
^
CD
"^
O
_
r—
CM
10
-a
""•^s,.
cn
"*S
O
r—
CO
C
CO
r—
JC
•p >>
CU S-
O T3
i- 1
O O
_.pT |
CJ >,
s- s-
CU TD
Q"
CQ
P—
•
CO
ID
CM
0
•
O
C
CD CU
CM > LO
O CO
• |"« •
Or—
•r- II
CO
cn
' £=
CM T-
OT3
8fO
o
o
i- I!
JC
to
3;
LO
CM
to
TD
^^^
cn
NX
O
LO
cu
E
CU
JC
cu
0 S-
s- cu
O 4-
i — CO
JC C
O lO
&. i-
CU -P
a.
CO
CO
cu
JC
CJ
to
O
tO
o.
IO
o
-a
cu
CO
to
CQ
*
A-6
-------
A.4 REFERENCES
i
1. Kleeberg, C.F., "Testing of Commercial Perch!oroethylene Dry
Cleaner," test report on Hershey, Pennsylvania test to James F. Durham,
EPA, May 14, 1976.
2. Kleeberg, C.F., "Testing of Industrial Perch!oroethylene Dry
Cleaner," test report on San Antonio, Texas, test to James F. Durham,
EPA, May 14, 1976.
3. Kleeberg, C.F., "Testing of Commercial Perch!oroethylene Dry
Cleaner," test report on Kalamazoo, Michigan, test to James F. Durham,
EPA, May 17, 1976. :
4. Scott Environmental Technology, Inc., "Air Pollution Emission
Test - Hershey Dry Cleaners and Laundry, Hershey, Pennsylvania," Report
No.76-Dry-l to EPA, March, 1976.
5. Midwest Research Institute, "Test of Industry Dry Cleaning
Operations at Texas Industrial Services, San Antonio, Texas," Report
No.76-Dry-2 for EPA, April, 1976.
6. Midwest Research Institute, "Air Pollution Emission Test,
Westwood Cleaners, Kalamazoo, Michigan," Report No.76-Dry-3, for EPA,
June, 1976.
A-7
-------
-------
APPENDIX B ;
COMPLIANCE TEST METHODS AND LEAK DETECTION EQUIPMENT
FOR PERCHLOROETHYLENE DRY CLEANERS
B.I COMPLIANCE TEST METHODS !
An emission measurement can be made by several methods, all of which
were analyzed as possible compliance test methods before choosing the
equipment performance criteria discussed in Chapter 6.0.
a) Material balance i
b) VOC concentration limit on dryer exhausts
c) Total mass limit for all emission points i
d) Equipment performance specification
While the material balance was determined to be the best method of
truly measuring solvent losses, equipment performance specifications are
preferred for enforcement of a standard. Still, the material balance test
method was used to develop background data for this document and is
therefore discussed. The method has the following advantages:
a) Total system emissions can be checked. This is not the case
for a dryer exhaust limit where only one emission point wpuld be monitored.
b) A material balance is more direct and simple than the test
equipment and procedures associated with a stack test.
c) Many existing plants keep records of clothes and solvent
throughput. These records could be used to assist and check the material
balance.
B-l
-------
•I
d) The material balance method, which determines emissions on a
mass per mass basis, does not distinguish between large and small plants as
a mass per day limit does. This means that best control technology is applied
across the board to all plants. v
The primary disadvantage of the material balance is that it is very time
w
consuming. While the material balance is optimum for determining exact emissions
it is suggested that other methods, specifically equipment performance require-
ments, should be used for enforcement.
The following sections of this chapter detail the material balance,
stack test, and-solvent sampling techniques. In addition, leak detection devices
are discussed in Section B.2 in terms of availability and cost.
B.I.I Material Balance Methods
A material balance requires measurement of clothes and solvent over a ™**
number of loads in addition to solvent levels in the system before and after
testing. All significant sources of solvent must be accounted for. The
following method was developed by EPA with the assistance of an EPA contractor
and the International Fabricare Institute. The method outlined here should
be considered flexible for the different processes in the industry.
A. Before the test begins, solvent in the system should be accounted .for
by the following methods:
1. Drain entire filter contents (powder, soil, and solvent) to muck
cooker or to holding tank (if cooker is not used).
2. Begin distillation/cooking or other treatment of muck. Dry cartridge
filters, if applicable.
B-2 III
-------
3. Begin desorbing vapor absorber.
4. On completion of cooking or drying, remove and discard dry
residue. Replace cartridge filters with new filters. \
5. Dry out adsorber bed. Put desorbed solvent into cleaning machine
base tank.
6. Start up wash pump to fill filter housing (ideally, machine should
be on continuous recirculation—solvent circulating between base tank and
filter and returning).
7. Add any detergent needed. (Take solvent sample, if needed—see
below for description of analysis methods.)
8. Measure solvent level by dip stick or gauge in base tank.
(Account for residue volume in bottom of tank.)
9. Put in filter and carbon. (Samples and total weights of this
material can be taken upon each removal from the cooker to determine losses
associated with the filter system.)
B. During the test:
Record weight of all loads.
C. After the test period, recreate conditions of first! sol vent measurement by
repeating Steps A.I through A.7. Another sample is taken to determine
detergent concentration in the "charged" solvent, if needed (see below).
The solvent loss in cartridge filters is a fixed loss for the number
of loads recommended for use. In other words, if a filter vendor recommends
200 loads of solvent as the filter life, the loss from filter change is the
same as the 200 load whether there are 50 loads or 300 loads. The loss from
filters for a test of less than the recommended filter life should be prorated
to the life of the filter. A loss of 1 kilogram after 50 loads on a filter of
B-3
-------
200 load life should be considered in the calculation as a loss of 0.25
kilograms.
Fixed losses are a significant factor in small machines. A 5 kilogram
load in a 12 kilogram capacity machine will have nearly the same loss as a
12 kilogram load in the same machine. In calculating kilograms of clothes
throughput in machines, the vendor capacity times the number of loads should *
be used instead of the actual load. The IFI and other organizations can
relate cubic feet of washer volume to capacity by available factors too
extensive to list here.
To determine solvent consumption, the solvent level (minus detergent,
sizing, etc.) of the initial measurement (Step A.8) is compared to the
solvent level (minus detergent, sizing, etc.) of the final measurement
(Step C.2). All solvent added during the test period should be accounted {|]|}
for.
To determine the system emission factor for the test period (which
should.be for at least one work week), the solvent consumption is divided
by'the clothes TihrougnoUT-the system. Stnce the test site need orrfy ber
prepared by an enforcement official and not attended, total manhours
required per test is_less than _10.
The following discusses sample analyses for solvent taken from the ^
system. A 0.5 liter sample is sufficient for analysis.
B-4
-------
According to the IFI, samples should be analyzed for detergent
concentration, moisture, non-volatiles, dry sizing, and insoluble materials.
A Hyamine 1622 or Aerosol OT Titration should be used for detergent concen-
tration reported on a volume/volume percent basis. The moisture content
is determined by a Karl-Fischer titration procedure and reported as grains
of water/100 millilitres of solution. Non-volatile residue is determined
gravimetrically by a steam bath evaporation of a measured volume of solvent
and weighing the residue. Dry sizing content is determined by extracting
the non-volatile residue with boiling ethyl alcohol. Insoluble material
content is to be determined gravimetrically after filtration of a volume of
solvent through a 0.20 micrometer membrane.
For determining the amount of solvent in filter materials (muck and
distillation waste) the test method described by the American National
Standards Institute in the paper "Standard Method of Test for Dilution of
Gasoline-Engine Crankcase Oils," should be used. To be derived are the
kilograms of VOC per kilogram of discarded filter muck. This method can
be used for the enforcement of the performance requirements of RACT.
EPA found that results were consistently 8-10 percent different when
these accounts for material other than solvent were not made. It is felt
that after determining total system solvent volume consumed during the course
of the test 9 percent can be subtracted out as other materials. The
remaining 91 percent can be considered pure solvent emitted to the atmosphere.
The test methods are described here for reference only.
B-5
-------
B.I.2 Emission Measurement Method for Perch!oroethylene From Adsorber Vent W
The primary method used to gather emission data has been the integrated
bag sampling procedure followed by gas chromatographic/flame ionization
detector analysis. Appendix B, Draft EPA Method 23: '"Determination of Total
Halogenated Organics from Stationary Sources," describes this approach.
For this method, the integrated bag sampling technique was_chosen
over charcoal adsorption tubes for Two "reasons:""Tf)~Tess~uhcertafnty""
about sample recovery efficiency, and (2) only one sample portion to analyze per
sample run. A column identified by a major manufacturer of chromatographic
equipment as useful for the separation of chlorinated solvents is employed.
The method was written after an initial EPA funded study of halogenated
hydrocarbon testing revealed areas where improvements in the bag sampling
technique were needed. In particular, leaking bags and bag containers were
cited as a probable cause of poor correlation between integrated and grab
samples taken at an emission site by that contractor. In light of these
findings, more rigorous leak check procedures were incorporated. The first
test conducted by EPA with the improved method to gather emission data
utilized both integrated bag and grab sampling techniques as a form of
quality control. For the three days during which tests were made, very good
correlation between the two techniques was obtained. Subsequent to these
R-
tests, a final draft of this method was prepared that incorporates further
leak checks as an additional precaution against erroneous data. These *
additions were suggested by an EPA contractor that was studying the vinyl
chloride test method. This contractor coincidentally performed the second
and third dry cleaning emission data tests, and was previously aware of the
need for exercising particular caution with respect to leak detection. ||||
B-6
-------
The costs for conducting a Method 23 emission test in triplicate will
depend on the length of the cleaning cycle and are accordingly estimated at
$5000 to $10,000 per unit. A simplified version of this test may run as
low as $200. The testing costs per unit would be lower if several units
at a single site were serially tested. The high cost of this test
precludes its use on a day-to-day enforcement of RACT. It is expected
that compliance with the 100 ppm RACT definition will be demonstrated
with inexpensive portable analyzers.
B.2 Leak Detection Methods
There are several types of portable, self-contained instruments currently
available for leak monitoring in dry cleaning facilities.! The principles of
operation are catalytic-oxidation., flame ionization, and infrared energy
absorption. All three types of detection will respond to practically all types
of organic materials although the relative responses to the different types
will vary.
For halogenated solvent operations where a single compound is predominant,
the instruments can be calibrated with that compound and the results will be on
that basis. Examples of some manufacturer's reported ranges for perchloroethylene
are: (1) catalytic-oxidation, 27-13,000 ppmv; (2) flame ionization, 2-20,000 ppmv;
and (3) infrared, 0.5-200 ppm +, depending on configuration.
The cost of a monitoring instrument ranges from about $900 to $4000,
depending on the detection principle, operating features, and required
accessories associated with the different instrument types and vendors.
EPA has contracted to examine several of these alternatives,
including less expensive systems than discussed above, the
object of the study being to develop an easy to use,
inexpensive monitor for vapor leak detection.
B-7
-------
B.3 SUMMARY
This chapter has detailed the methods used to develop background
information for this study. For the most part these methods are too
expensive and cumbersome to be used as effective enforcement tools.
It is suggested that portable detectors, to be analyzed and developed
by EPA in the near future, be used to determine the extent of vapor
leaks in a system and also be used to determine compliance with dryer
control requirements. Solvent in filter and distillation system wastes
can be determined'by methods discussed in this chapter.
B-8
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-450/2-78-050
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Control of Volatile Organic Emissions From
Perchloroethylene Dry Cleaning Systems
5. REPORT DATE
December, 1978
6. PE-RFORMING ORGANIZATION CODE
7. AUTHOR(S)
Charles F. Kleeberg, ESED
Jack G. Wright, SASD
8. PE-RFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Emission Standards and Engineering Division
Research Triangle Park, N.C. 27711
OAQPS No. 1.2-117
1O. F'ROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report provides the necessary guidance for development of regulations
limiting emissions of Volatile Organic Compounds (VOC) from perchloroethylene
dry cleaning systems. Reasonably Available Control Technology (RACT) is
defined and a costjanalysis of RACT is included in order that cost effectiveness
may be evaluated fjbr these systems.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Air Pollution
Regulatory Guidance
Dry Cleaning
8. DISTRIBUTION STATEMENT
Unlimited
b. IDENTIFIERS/OPEN ENDED TERMS
Air Pollution Control
Stationary Sources
Organic Vapors
19. SECURITY CLASS (ThisReport)'
Unclassified
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
20. SECURITY CLASS (Thispage)
Unclassified
c. COSATI Field/Group
21. NO. OF PAGES
68
22. PRICE
-------
------- |