&EPA
United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EMB Report 79-DRY-6
December 1979
Air
Emission Test Report
Leaks from
Perchloroethylene
Dry Cleaners
-------
PERCHLOROETHYLENE EMISSION TESTING
AT
KLEEN KORNOR
CORTLAND, NEW YORK
By
Robert F. Jongleux
TRW
ENVIRONMENTAL ENGINEERING DIVISION
P.O. BOX 13000
Research Triangle Park, North Carolina 27709
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Emission Measurement Branch
Research Triangle Park, NC 27711
November 1979
-------
TABLE OF CONTENTS
age
LIST OF TABLES AND FIGURES ii
I. INTRODUCTION 1
11. SUMMARY OF RESULTS 2
III. PROCESS DESCRIPTION 14
IV. DISCUSSION OF RESULTS 20
V. SAMPLING AND ANALYTICAL PROCEDURES 24
VI. APPENDIX A - SAMPLE CALCULATION A-l
VII. APPENDIX B - FIELD AND LABORATORY DATA B-l
VIII. APPENDIX C - GAS STANDARD CERTIFICATION C-l
IX. APPENDIX D - PROJECT PARTICIPANTS D-l
X. APPENDIX E - TEST LOG (NOTES) E-l
XI. APPENDIX F - CANDIDATE INSTRUMENT SELECTION CRITERIA F-l
(i)
-------
LIST OF TABLES AND FIGURES
Tables Page
TABLE 2.1 - SUMMARY - TESTING RESULTS 3
TABLE 2.2 - MONITOR RESULTS CARBON BREAKTHROUGH (HLD-440) 5
TABLE 2.3 - MONITOR RESULTS CARBON BREAKTHROUGH (TLV-SNIFFER) 6
TABLE 2.4 - MONITOR RESULTS CARBON BREAKTHROUGH (METER-ALL) 7
TABLE 2.5 - LEAK DETECTION - WASH CYCLE 8
TABLE 2.6 - LEAK DETECTION - EXTRACT CYCLE 9
TABLE 2.7 - LEAK DETECTION - DRY CYCLE 10
TABLE 2.8 - LEAK DETECTION - AERATION CYCLE 11
TABLE 2.9 - EVAPORATIVE LOSS TEST 13
TABLE 3.1 - MACHINE SPECIFICATIONS 16
TABLE 3.2 - PLANT THROUGHPUT 18
TABLE 4.1 - CARBON BED EFFICIENCY DATA 21
TABLE 4.2 - ATMOSPHERIC EMISSIONS (OUTLET CARBON ADSORBER) 22
TABLE B.I - CARTRIDGE FILTER WEIGHT LOSS DATA B-2
TABLE B,2 - VOLUMETRIC FLOW RATE DATA B-3
TABLE B.3 - WEATHER DATA B-4
TABLE B.4 - CANDIDATE INSTRUMENTS - LABORATORY TESTING B-7
Figures
FIGURE 3.1 - DETREX DRY-TO-DRY MACHINE 15
FIGURE 3.2 - HOYT CARBON ABSORPTION SYSTEM 17
FIGURE 3.3 - SAMPLING LOCATION SCHEMATIC (FRONT VIEW) 19
FIGURE 5.1 - INSTRUMENT PLUMBING 26
FIGURE 5.2 - S-TYPE PITOT - MANOMETER ASSEMBLY 27
(ii)
-------
Figures (cont'd) Page
FIGURE B-l EXAMPLE - CARBON BED BREAKTHROUGH B-5
FIGURE'B-2 EXAMPLE - INLET CONCENTRATIONS DURING DRY CLEANING CYCLE.. B-6
FIGURE B-3 SENSITIVITY RESPONSE APPARATUS B-9
FIGURE B-4 SENSITIVITY TO PERCHLOROETHYLENE B-10
FIGURE B-5 SENSITIVITY TO PERCHLOROETHYLENE B-ll
FIGURE B-6 SENSITIVITY TO PERCHLOROETHYLENE B-12
FIGURE B-7 SENSITIVITY TO PERCHLOROETHYLENE B-l3
FIGURE B-8 TEMPERATURE RESPONSE APPARATUS B-15
FIGURE B-9 WATER VAPOR TEST APPARATUS B-16
(iii)
-------
I. INTRODUCTION
During the week of March 26, 1979, a two person test crew from TRW per-
formed an emission test at the Kleen Kornor, located at 10012 Homer Avenue,
Cortland, New York. The Kleen Kornor is a small commercial dry cleaning and
laundry establishment, owned and operated by the Ames Linen Service of Cortland,
New York.
The specific process tested at the Kleen Kornor was the dry cleaning system.
This plant utilizes a dry-to-dry perchloroethylene machine, and the rated capacity
of the machine was 40-45 pounds. The dry cleaning machine was a Detrex Commander
Model #11-20-H of approximately 2£ years of age. The machine specifications and
details are elaborated upon in Section III - (Process Description). The dry
cleaning system utilized a Kleen-Rite (Model #3H-1200) cartridge filter system
for purifying the dry cleaning solvent. Emissions from the process were controlled
by a carbon adsorption system manufactured by Hoyt (Model #1-662). The carbon bed
was approximately seventeen (17) years old.
There was a multifold purpose of conducting this emission test. The reasons
include determination of the removal efficiency for a typical carbon adsorption
unit, testing of candidate leak detectors, and establishment of data for a mass
balance around a dry-to-dry perchloroethylene dry cleaning unit.
-1-
-------
II. SUMMARY OF RESULTS
The original work assignment for this project consisted of writing a guideline
document method for determining general and specific leaks at perchloroethylene
(f,2 014) dry cleaning plants. The procedure recommended by the work assignment
was to incorporate an inexpensive (< $250) portable leak detector and parallel the
method used in the EPA guideline series on Control of Volatile Organic Compound
Leaks from petroleum refining equipment. However, such a parallel method was not
practical, because of the simplistic nature of the leak detectors. In addition,
the scope of the work assignment shifted to include additional objectives, all of
which were related to developing a New Source Standard (NSS) for perchloroethylene
dry cleaning plants. In order to clarify testing objectives, the program at the
Kleen Kornor was divided into four segments. These segments include mass balance
(MB), carbon breakthrough monitor evaluation (CBME), leak detector evaluation (LDE)
and mass evaporation losses (MEL). The information collected and determined fosr
these segments overlapped to some degree.
The testing information is summarized in Table 2.1. Data presented
includes the plant throughput, the perchloroethylene loss due to changing cartridge
filters, the average carbon bed removal efficiency, the calculated emission rate
from the carbon adsorption system, and the amount of solvent recovered and used
during the testing period. The total plant throughput was 735 kg (1614.8 Ibs) for
the testing period. This represent a total of forty-four (44) machine loads. A
breakdown of the plant throughput is detailed in Table 3.2. The perchloroethylene
loss due to changing the cartridge filters was calculated as 2.74 kg/100 kg
throughput (2.74 lbs/100 Ibs throughput). The method for calculating this loss
is described in Appendix A. The average removal efficiency of the carbon bed
was 89.4% for the week of testing. The perchloroethylene loss to the atmosphere.
was calculated as approximately 762 g (1.7 Ibs.) for the four day testing period.
-2-
-------
oo
I
PLANT THROUGHPUT
CARTRIDGE FILTER LOSSES 2
CARBON BED REMOVAL EFFICIENCY
ATMOSPHERIC EMISSIONS
SOLVENT USED (C2 C14)
MACHINE MILEAGE 4
MACHINE MILEAGE
TOTAL MILEAGE
TOTAL MILEAGE
735.7 kg (1614.8 Ibs.) (44 loads)
2.74 kg/100 kg throughput
(2.74 Ibs./lOO Ibs. throughput
89.4% Average
762 g/week (1.7 Ibs/week)
18.1 q/1oad (.04 Ibs/load)
21.6 liters (5.7 gal)
3.88 kg solvent/100 kg clothes
4.73 kg solvent/100 kg clothes
6.62 kg/solvent/100 kg clothes
7.47 kg/solvent/100 kg clothes
1
6
See Table 3.2
'Calculated - See Appendix A
From Outlet of Carbon Adsorber Only
Based on Machine Capacity
DBased on Machine Throughput
'Machine Mileage and Cartridne Filter Losses
TABLE 2.1 SUMMARY - TESTING RESULTS
-------
Both the removal efficiency and the atmospheric emissions are further discussed
in Section 4. The amount of solvent (perchloroethylene) used was 21.6 liters
(5.7 gal). Likewise, these calculations are detailed in Appendix A.
Tables 2.2, 2.3 and 2.4 tabulate the results of carbon breakthrough monitor
evaluation. The three selected candidate instruments were compared during
various stages of the dry cleaning cycle (Selection criteria - See Appendix F).
Only the Halogen Leak Detector Model 440 instrument responded favorable during
this field evaluation. In order to effectively monitor breakthrough of the
carbon bed, a detector (instrument) must have a pre-determined setting at which
it will alarm. The Halogen Leak Detector Model 440 does not have an effective
means to set the alarm mechanism to a pre-determined level. Therefore, the
usefulness of this instrument to monitor breakthrough of a carbon bed is severely
limited. The Bacharach TLV Sniffer and the Meter-All instruments responded
erratically to lower perchloroethylene concentrations and thus seemed inferior.
The same candidate instruments were evaluated for applicability to detect
perchloroethylene leaks. The results of the leak detector evaluation are
summarized in Tables 2.5, 2.6, 2.7 and 2.8. Three locations of major solvent
leaks were determined by visual inspection and the continuous monitor (Beckman
402). Each candidate leak detector then was tested during the major portions
of the machine cycle. Again, the most favorable instrument was the HLD-440.
Since the response of the HLD-440 is strictly qualitative, the instrument is
well suited in finding relatively large solvent leaks only.
Due to the erratic performance at this test site, the candidate instruments
were returned to the laboratory for further evaluation. Based upon laboratory
and field data none of the candidate's instruments proved satisfactory for the
purpose of detecting various concentrations of leaks for a variety of reasons.
These reasons include: sensitivity to slight increases in water vapor, sensitivity
to increased temperature, erratic response to changing concentrations of perchloro-
-4-
-------
I
tn
r
TIME
0956
0958
1000
1002
1004
1006
1008
1000
1012
j 1014
1016
1018
1
CONCENTRATION
(ppm C2 C14)
37
37
37
36
35
34
42
40
41
50
58
47
TEMPFRATURE
(uc)
28
27
28
27
27
28
32
31
32
32
32
30
(OF)
83
82
83
82
82
83
90
89
90
90
90
86
MONITOR
RESPONSE
NONE
MACHINE
CYCLE
DRY
NONE ^ | DRY
NONE ! DRY
NONE I DRY
NONE ( DRY
NONE I DRY
NONE ) AERATION
NONE ! LOADING
NONE i LOADING i
ALARM
ALARM
NONE
LOADING 1
LOADING
WASH
j
i !
i
i
j
1
CONCENTRATION - measured at outlet duct by Beckman 402 Hydrocarbon Analyzer.
TABLE 2.2 MONITOR RESULTS CARBON BREAKTHROUGH
(HALOGEN LEAK DETECTOR - MODEL 440)
-------
TIME
0956
0958
1000
1002
1004
1006
1008
1010
1012
1014
1016
1018
1
CONCENTRATION
(ppm C2C14)
37
37
37
36
35
34
42
40
41
50
58
47
TEMPERATURE
(°C)
28
27
28
27
27
28
32
31
32
32
32
30
(OF)
83
82
83
82
32
83
90
89
90
90
90
86
MONITOR
RESPONSE
118
125
120
no
122
no
no
118
108
112
112
no
Jf
MACHINE
CYCLE
DRY
DRY
DRY
DRY
DRY
DRY
AERATION
LOADING j
LOADING i
LOADING i
LOADING 1
WASH
i
!
1. CONCENTRATION - measured at outlet duct by. Beckman 402 Hydrocarbon Analyzer.
2. Zero instrument in HC free air.
TABLE 2.3 MONITOR RESULTS CARBON BREAKTHROUGH
(BACHARACK-TLV-SNIFFER)
-------
TIME
0956
0958
1000
1002
1004
1006
1008
1010
1012
1014
1016
1018
1
CONCENTRATION
(ppm C2 C14)
37
37
37
36
35
34
42
40
41
50
58
47
TEMPERATURE
PC)
28
27
28
27
27
28
32
31
32
32
32
30
(°F)
83
82
83
82
82
83
90
89
90
90
90
86
MONITOR
RESPONSE
NONE
ALARM
NONE
NONE
ALARM
NONE
MACHINE
CYCLE
DRY
DRY
DRY
DRY
DRY
DRY
NONE 1 AERATION
ALARM
ALARM
ALARM
ALARM
NONE
LOADING
LOADING 1
LOADING
LOADING
WASH
1
I
i
1
1
CONCENTRATION - measured at outlet duct by Beckman 402 Hydrocarbon Analyzer.
TABLE 2.4 MONITOR RESULTS CARBON BREAKTHROUGH
(METER-ALL^
-------
I
00
MONITOR
METER-ALL
HLD-440
TLV-SNIFFER
METER-ALL
HLD-440
TLV-SNIFFER
METER-ALL
HLD-440
TLV-SNIFFER
SOURCE
FAN
FAN
FAN
SHAFT
SHAFT
SHAFT
DRUM SIDE
DRUM SIDE
DRUM SIDE
RESPONSE
ALARM
ALARM
DOWNS CALE
ALARM
ALARM
5-10
NONE
NONE
NEGATIVE
AMBIENT
TEMPERATURE
TO .
28
28
28
29
29
29
24
24
24
(OF)
83
83
83
84
84
84
75
75
75
AMBIENT
CONCENTRATION
(ppm C2C14) 1
10
10
10
50-150
50-150
50-150
3
3
3
CONCENTRATION - in the proximity of the suspected leak as measured by a Beckman 402
Hydrocarbon Analyzer.
TABLE 2.5 LEAK DETECTION DURING WASH CYCLE
-------
MONITOR
METER-ALL
HLD-440
TLV-SNIFFER
METER-ALL
HLD-440
TLV-SNIFFER
METER-ALL
HLD-440
TLV-SNIFFER
SOURCE
FAN
FAN
FAN
SHAFT
SHAFT
SHAFT
DRUM
DRUM
DRUM
RESPONSE
ALARM
ALARM
8 (SCALE xl)
ALARM
ALARM
38 (xl)
ALARM
ALARM
14 (x 1)
AMBIENT
TEMPERATURE
(°C)
30
30
30
30
30
30
28
28
28
(OF)
86
86
86
86
86
86
83
83
82
AMBIENT
CONCENTRATION
(ppm C2 014) 1
25
25
25
120
120
120
10-40
10-40
10-40
1
CONCENTRATION - in the proximity of the suspected leak as measured by a Beckman 402
Hydrocarbon Analyzer.
TABLE 2.6 LEAK DETECTION DURING EXTRACT CYCLE
-------
o
I
MONITOR
METER-ALL
HLD-440
TLV-SNIFFER
METER-ALL
HLD-440
TLV SNIFFER
METER- ALL
HLD-440
TLV-SNIFFER
SOURCE
FAN
FAN
FAN
SHAFT
SHAFT
SHAFT
DRUM
DRUM
DRUM
RESPONSE
ALARM
ALARM
NEGATIVE
ALARM
ALARM
15.5 (ALARM
XI 0)
ALARM
ALARM
25 (xl)
AMBIENT
TEMPERATURE
UC)
3D
3D
3D
32
32
32
31
31
31
(°f)
87
87
87
90
90
90
89
89
89
AMBIENT
CONCENTRATION
(ppmC2Cl4) ]
15
15
15
500-1000
500-1900
500-1000
40-50
40-50
40-50
1
CONCENTRATION - in proximity of the suspected leak as measured by a Beckman 402
Hydrocarbon Analyzer.
TABLE 2.7 LEAK DETECTION DURING DRY CYCLE
-------
MONITOR
METER-ALL
HLD-440
TLV-SNIFFER
METER-ALL
HLD-440
TLV-SNIFFER
SOURCE
FAN
FAN
FAN
DRUM
DRUM
DRUM
RESPONSE
NO ALARM
NO ALARM
NEGATIVE
NONE
NONE
NEGATIVE
AMBIENT
TEMPERATURE
(°C)
33
33
33
33
33
33
(°F)
91
91
91
92
92
92
AMBIENT
CONCENTRATION
(ppm C2 Cl4) 1
5
5
5
5
5
5
1
CONCENTRATION - in proximity of the suspected leak as measured by a Beckman 402
Hydrocarbon Analyzer.
TABLE 2.8 LEAK DETECTION DURING AERATION CYCLE.
-------
ethylene and the lack of an effective zeroing mechanism, Therefore, the method
developed in the EPA guideline document pertaining to petroleum refineries cannot
be applied to the dry cleaning industry if these inexpensive (.< $250) leak
detectors that were tested are to be used. (Laboratory testing - see Appendix B).
During the Evaporation Loss Test a total of 1.04 kg (2.29 Ibs.) was lost
(evaporated from a load of clothes). The test was sixty minutes in duration.
Ambient temperature was 29°C (85°F). Table 2.9 lists the evaporative losses
every five minutes over the course of an hour. The evaporative loss test was
conducted to approximate losses that might take place in a transfer rather than
a dry-to-dry process operation.
-12-
-------
co
:
DATE
TIME
*.
3/30/79 j 1058
3/30/79 I 1103
3/30/79
3/30/79
3/30/79
3/30/79
3/30/79
3/30/79
3/30/79
3/30/79
3/30/79
3/30/79
3/30/79
1108
1113
1118
1123
1128
1133
1138
1143
1148
1153
1158
60
WEIGHT
Kg
20.50
20.41
20.37
20.32
20.14
20.05
19.96
19.87
19.82
19.73
19.55
19.50
19.46
1.04
(Ibs.)
45.2
45.0
44.9
44.8
44.4
44.2
44.0
43.8
43.7
43.5
43.1
43.0
42.9
2.3
TABLE 2.9
EVAPORATIVE LOSS TEST
-------
III. PROCESS DESCRIPTION
The test was performed on Detrex dry-to-dry machine, model 11-20-H (see
Figure 3.1 and Table 3.1 for machine specifications), with Kleen Rite cartridge
filters. A Hoyt Model 1 carbon absorber (with the original carbon) was used to
recover perc from the dry cleaning machine. (See figure 3.2.). Floor vents
also venting to the carbon absorber had been removed due to problems with lint
fouling the dampers. Additional vents had also been connected to two coin-ops
machines (now removed) and to the cartridge filters (also now disconnected).
The net result was that the carbon absorber was connected only to the dry clean-
ing machine with one additional opening, a 3/4 inch pipe, opening to room air
(the pipe was a remnant of the filter venting scheme that was removed).
To establish a reference point with this equipment, one bank of filters
(four cartridges) were changed, the carbon absorber was desorbed with the reclaimed
perc added to the system, and all traps, button and lint, were cleaned. New filter
cartridges were then installed and the dry cleaning machine was turned on to fill
the new filters. A level reading was taken at the wash tank sight tube with a
tape measure. The sight tube for the clean solvent tank showed the tank to be
completely full. For this machine the clean solvent tank will remain full (unless
a clean solvent rinse is used during dry cleaning) since condensed solvent from
the reclaiming cycle is piped to this tank (after water separation). An internal
weir allows overflow from the clean solvent tank to the wash tank, hence, the
clean solvent tank is always full.
The loss attributed to the filters was measured. The Kleen Rite filter at
this installation is composed of two tubes of four cartridges. New cartridges
weigh 16 kg (35.5 Ibs) and are installed in each tube about every 12 weeks after
processing about 5400 kg (12 000 Ibs) of clothes. Replacement of the cartridges
is alternated with only one tube being changed at a time on six-week intervals.
-14-
-------
The cartridge removed Monday morning were weighed on the scales at the front
of the. dry cleaning store. Each of the cartridges in a carton weighed an indi-
cated 52.5 pounds. Later it was discovered that these scales did not register
past 52.5 pounds. Therefore, to obtain an estimate of the perc loss from these
filters, the second tube of cartridges was drained from Thursday afternoon until
Friday morning. One cartridge was removed from this tube and weighed on a floor
scales at a linen supply house. The cartridge removed Monday was dried under a
laboratory hood and reweighed to establish the actual weight of perc which was
present in the drained filter. That weight, along with the weight of the cartridge
removed Friday and the amount of clothes processed on each cartridge are calcu-
lated in the test report (Equation A-l).
Final values for losses were found by adding the loss attributed to the
filter to the loss from the machine itself. The loss from the machine was found
Thursday afternoon (before draining the second filter tube) by desorbing the carbon
bed, adding the reclaimed perc to the machine, and then putting the machine into
the continuous recirculation mode. The perc level in the sight tube was again mea-
sured with tape measure. Note that no perc or additive was added to the machine
during the period of the test.
Losses from the machine are attributable to only two categories of source -
losses during venting on the machine and fugitive losses. Venting occurs during
the aeration cycle and during loading/unloading operations when the door is open.
Fugitive losses are vapor leaks or liquid leaks. The major leaks appeared to be
from valves in the solvent lines to the filters where perc leaked enough during
the night to form a small puddle on the base tank of the machine.
-14a-
-------
FIGURE 3.1 - DETREX DRY—to—DRY MACHINE
-15-
-------
—SO—H
DIMENSIONS (Crated)
Washer-Extractor-6'-11" w x 5'-0" d x 7'-l" h
WEIGHT - 3,000 Lbs. (Crated)
INSTALLATION DIMENSIONS
Washer-Extractor-6'-3" w x 4'-0"d x 6'-4"h
MINIMUM OPENING REQUIRED
Width - 4'-0" Height - 6'-4"
CYLINDER
Diameter-3'-0"
Depth - 19 1/2" - 11.2 Cubic Feet
Material - Stainless Steel
Number of Ribs - 4
Door Opening - 16 1/2" Dia.
Wash Speed - 31 RPM
Dry Speed - 45 RPM
Extract Speed - 450 RPM
CAPACITIES of TANKS
Main Filter Tank - 91 Gal.
Clean Solvent Tank - 50 Gal.
ELECTRICAL SPECIFICATIONS
Machine - 60 amp - 230 v - 60 cy - 3 ph
MOTORS
Washer-Extractor Motor - 3/.75hp
Fan Motor - 1 hp
Filter Pump Motor - 1 hp
SERVICE CONNECTIONS
Air - 75-80 osi -1/4 ips
Water
Water Inlet - 3/4 ips
Water Outlet - 3/4 ips,
12 Gal. per min. at 70°F.
Steam
Steam Inlet - 3/4 ips
Condensate Return - 3/4 ips
1-3/4 Boiler hp
75psi
CONTROLS - ELECTRICAL and AIR
Type - Automatic and Manual
...Variable Level Control Stainless Steel Recovery Housing...
Vent - 4"
Water Separator Outlet - 1/2"
TADLE 3.1 - MACHINE SPECIFICATIONS
-16-
-------
Q -SAMPLING LOCATIONS
FIGURES.? CARBON-ABSORBER- (HOYT MODEL # 1;
-17-
-------
MONDAY
TUESDAY
WEDNESDAY
THURSDAY
LOAD # (kg) (Ibs) (kg) (Ibs) (kq) (Ibs) (kg) (Ibs)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
2.3
17.9
16.6
17.2
17.2
23.6
19.5
20.6
13.2
(5.0)
(39.7)
(37.5)
(38.0)
(38.0)
(55.0)
(43.0)
(45.5)
(29.0)
18.1
18.4
17.5
17.5
17.5
17.5
16.1
10.7
10.7
17.5
(40.0)
(40.5)
(38.5)
(38.5)
(38.5)
(38.5)
(35.5)
(23.5)
(23.5)
(38.5)
17.5
17.5
15.2
17.5
17.5
17.5
17.5
17.5
17.5
•
17.5
17.5
17.5
17.5
6.1
(38.5)
(38.5)
(33.5)
(38.5)
(38.5)
(38.5)
(38.5)
(38.5)
(38.5')
(38.5)
(38.5)
(38.5)
(38.5)
(13.5)
17.5
17.5
17.5
17.5
17.5
17.5
17.5
17.5
17.5
17.5
17.5
(38.5)
(38.5)
(38.5)
(38.5)
(38.5)
(38.5)
(38.5)
(38.5)
(38.5)
(38.5)
(38.5)
DAILY TOTAL 148.1 (326.7) 161.5 (355.5) 231.3 (509. l) 192.5 (423.5)
WEEKLY TOTAL 735.7 kq (1614.8 Ibs)
TABLE 3.2 PROCESS DATA - CLOTHES THROUGHPUT
-------
o o
o o
FIGURE 3.3- SAMPLING SCHEMATIC (FRONT VIEW)
-------
IV. DISCUSSION OF TEST RESULTS
The results summarized in Section 2 require elaboration. Table 4.1 shows
the analytical results used to determine the efficiency of the carbon adsorption
system. The inlet and outlet were not sampled simultaneously with the continuous
hydrocarbon analyzer due to equipment limitations. However, comparative data
was generated by utilizing an integrated bag sample at one location, while the
continuous monitor (FID) was located at the other location. Those samples
collected by the integrated bag method are denoted with an asterisk in Table 4.1.
In calculating the removal efficiency, the average inlet concentration was
determined only from those samples drawn from the inlet during the aeration and
loading portions of the dry cleaning cycle. This was due to the fact that only
during this segment was the carbon bed actively receiving perchloroethylene
emissions. During the remaining portions of the cycle (wash, extraction and dry)
the damper from the dry cleaning unit to the carbon absorption system was closed.
This removal efficiency then was used to calculate the average removal efficiency
for each day of the test period. The removal efficiency for the week was in turn
calculated by weighting the daily average removal efficiencies in proportion to
the number of loads cleaned per day.
The emissions emitted to the atmosphere are highlighted in Table 4.2. The
average daily outlet concentration was calculated from the data outlined in Table
4.1. The emission rate was calculated by Equation 4.2 in Appendix A and reported
in units of milligrams per minute. The emissions attributed to a portion of the
cycle, either aeration or loading, were calculated by Equation 4.3 in Appendix
A. The daily total emissions were the sum of the emissions during both the
aeration and loading cycles multiplied by the number of loads on a given day
(Table 3.2).
It should be noted that there was no flow in the outlet duct during a major
portion (80%) of the dry cleaning cycle; consequently, no atmospheric emissions
can be calculated directly. The continuous monitor indicated a C2
-20-
-------
MONDAY
3/26/79
TUESDAY
3/27/79
WEDNESDAY
3/28/79
THURSDAY
3/29/79
FRIDAY
3/30/79
TNI FT
1 INLL. 1
(ppm as 03014)
N.A.
60
—
--
25
275*
—
—
23
11
13
400*
55
--
—
—
275*
--
290*
355*
--'
--
335
nun FT
UU 1 L-U 1
(ppm as 02614)
N.A.
18
15
43*
62
80
72
90
DESORPTION
16
8
25*
11
15
17
40
42
51
56
65
77
90
100
DESORPTION
9
EFFICIENCY
(%)]
70
—
--
--
70.9
--
--
30.4
27.3
--
97.25
72.7
69.1
—
__
82.5
--
77.5
78.3
--
97.3
(%)2
94
95.3
86.6
80.7
75.1
77.6
72.0
95.0
93.5
97.5
96.6
95.3
94.7
87.6
86.9
84.1
82.6
80.7
76.1
72.0
68.9
97.2
(%)3
95. 654
83.04
95.65
82.89
__
5
89.4°
*Measurement Taken From Integrated Bag Sample - FID Analysis Readout
1 Removal Efficiency - Calculated at one Point of Cycle
2Removal Efficiency - Calculated from Average Inlet Concentration During Aeration &
3Average Removal - Calculated (Daily Basis) Loading
^Assumed - Based on Wednesday Efficiency Data (After Bed Desorption)
5Average For Week - Weighted by #Loads/Day (Monday-Thursday Only)
N.A. - Not Ascertained - Instrument Set-up.
TABLE: 4.1 CARBON BED EFFICIENCY DATA
-21-
-------
I
r\5
ro
i
DATT
3/26/79 (A)
(L)
3/27/79 (A)
!
(L)
3/28/79 (A)
(L)
3/29/79 (A)
(L)
fT\ AUCDACC- nin
AVERAGE OUTLET CONCENTRATION
(ppm)U; (mg/m3) (2)
i
20 135.5
54.2 367.2
15 . 101.6
55.3 374.
i
n CT rnMrcMTDATinw ronw TABI c /i i
(3)
(mg/min) (Ibs/min)
1295.5 (.0028)
1212.81 (.0027)
2717.3 (.0060)
3800.5 (.0083)
664.6 (.0015)
916.7 (.0020)
3776.9 (.0083)
4442.91 (.0097)
WEEKLY TOTAL
EMISSION RATE
(4)
(g/cyele)flbs/cycl^
2.8 (.006))
6.2 (.014)
5.8 (.013)
19.6 (.043)
1.4 (.003)
4.7 (.010)
8.1 (.018)
23.9 (.050)
1
(5J
g/day (Ibs/day)
81.2 (.18)
253.9 (.56)
86.0 (.19)
340.6 (.75)
761. 7g (1.68 Ibs
(2) BY EQUATION 4.1
(3) BY EQUATION 4.2
(4) BY EQUATION 4.3
(5) BY EQUATION 4.4
A - AERATION CYCLE
L - LOADING
TABLE 4.2 ATMOSPHERIC EMISSIONS - OUTLET CARBON ABSORBER
-------
concentration present despite no flow in the outlet duct. This concentration was
only significant on Tuesday and Thursday of the test period (i.e. after carbon
bed breakthrough). The magnitude of the emission is assumed to be comparable
to a solvent leak at the machine itself, since ambient conditions were present
in the duct.
In Appendix B several examples of the continuous hydrocarbons monitor chart
are highlighted. Figure B-l graphically illustrates carbon bed breakthrough.
Figure B-2 shows a trace of inlet concentrations for an entire dry cleaning
cycle. There was a measurable concentration of perchloroethylene at all times
at the inlet sampling location, despite the fact that damper to the carbon bed
from the machine was closed. It is judged that this measurable concentration was
the result of the damper leaks and ambient background concentrations drawn in
through the lint trap.
The perchloroethylene concentrations peaked during the aeration and loading
portion of the dry cleaning cycle. The peak was difficult to assess, due to
its magnitude and rapid deterioration. The hydrocarbon analyzer had multiple
scales and a generation of ten thousand parts per million (10,000 ppm) static
gas standard verified the linearity of the hydrocarbon analyzer to + 10 percent.
-23-
-------
V. SAMPLING AND ANALYTICAL PROCEDURES
The perchloroethylene concentrations were monitored with a Beckman 402
Hydrocarbon Analyzer. This instrument is a continuous hydrocarbon monitor which
operates on the principle of flame ionization. A continuous monitor type flame
ionization detector (FID)instrument was selected over a gas chromatograph-flame
ionization detector (GC/FID) because the primary constituent of the flue gas was
assumed to be perchloroethylene (63 014), and no separation of hydrocarbon compounds
was necessary. A continuous monitor FID also offered the advantage of relatively
instaneous reading at one point in time. This was desirable to quantify the
perchloroethylene concentrations during the carbon breakthrough monitor evaluation
(CBME) and leak detector evaluation (IDE). Modification of the calibration system
also provided introduction of an integrated bag sample collected over a specified
period of the machine cycle.
The hydrocarbon monitor was calibrated directly each day with solvent
(perchloroethylene). This procedure is a deviation from the normal procedure.
The normal procedure requires calibration gases of either methane (CH/j) or propane
(CsHs) and determination of a solvent response factor for the solvent under
investigation. By calibrating the instrument directly with solvent (perchlo-
roethylene--C2 Cl4), no further calculations were necessary.
The calibration gases were supplied and certified by Scott Environmental
Technology, Inc. A copy of the standard certification is given in Appendix C.
Three concentrations of calibration gas were utilized. The concentrations were
approximately fifty, one hundred and five hundred parts per million (ppm)
perchloroethylene in air. The standards were analyzed by GC/FID analysis to
be 45.7 ppm, 92.8 ppm and 493 ppm C2 Cl4 prior to shipment into the field. A
post analysis of the gas standards is expected to determine any degradation of
the standards over time. The parameters for analysis were as follows:
-24-
-------
The column was a 6 ft x 1/8" diameter stainless steel
packed with 5% SP-1200 + 5% Bentone 34 on 100/120
mesh Supelcoport. The carrier gas were helium at a
flow rate of 50 cc/min. The column temperature was
75°C. 1 cc injections were made via a gas sampling
valving. The instrument on which these analyses were
performed was a Varian 1800 with flame ionization detector.
The peaks generated were quantified by digital integration.2
The introduction of the calibration gases to the continuous monitor was by
means of a new aluminized gas sampling bag and an auxiliary bellows pump. The
gas sampling bag was outfitted with a new piece of teflon^ tubing and a 1/4 inch
Swage1oc$* quick disconnect. This method of introduction was necessary because
the standards were in low pressure cylinders. Direct interface to the instrument
would not have provided sufficient sample pressure throughout the testing period.
Figure 5.1 is a schematic of the instrument plumbing. The operating conditions
of the instrument: the sample pressure, the air pressure, and fuel pressure
were maintained at two, ten, and twenty pounds per square inch (psig) respectively.
Velocity measurements were conducted daily with a S-type pi tot-tube manometer
assembly (Figure 5.2). Due to the space limitations of sampling location and the
small diameter duct of 30.5 cm (12 inches), an EPA method 2 velocity traverse
could not be conducted. Investigation showed that there was little variation
across the duct during any portion of the dry cleaning cycle. The velocity and
volumetric flowrate calculations are in Appendix A. Field Data is included in
Appendix B. No moisture measurement was undertaken. The molecular weight of
the stack gas (Ms) was assumed to be that of air (28.80). Barometric pressure
was reported as station pressure at the local U.S. Weather Station (Table B-3).
The plant throughput was weighed by means of a basket suspended scale. This
laundry-basket suspended scale was checked against a recently calibrated spring
scale of reliable quality. The difference between the two scales was judged to
Correspondence from R. B. Denyszyn, Scott Environmental Technology, to
R. F. Jongleux (TRW) April 6, 1979.
-25-
-------
DUCT TEFLON SAMPLING LINE
AIR CYCLINDER
r\>
-------
1.90- 2.54 cm*
10.75-1.0 in.)
r €2^533222
, 7.62 cm (3 in.)
TEMPERATURE SENSOR
fEDERAl PECIST6R, VOl. 4J, NO. 160—THURSDAY. AUGUST 18, 1977
LEAK-FREE
CONNECTIONS
•SUGGESTED {INTERFEREPJCE FREE)
PITOT TUBE • THERMOCOUPLE SPACING
' FIGURE 5.2 PITOT--MANOMETER ASSEMBLY
-27-
-------
be insignificant and therefore the basket scale was utilized for the sake of
convenience (See Test Log).
-28-
-------
REFERENCES
Alternative Test Method for Direct Measurement of Total Gaseous Organic
Compounds Using a Flame lonlzation Analyzer. Emission Measurement Branch,
EPA, June 6, 1978.
2
Correspondence from Robert B. Denyszyn, Scott Environmental Technology, to
R. F. Jongleux (TRW), April 6, 1979.
-29-
-------
APPENDIX A
EXAMPLE CALCULATIONS
A-l
-------
EQUATION A.I - CARTRIDGE FILTER LOSSES
!
NOMENCLATURE
X = throughput on 1st Filter (as pounds of clothes)
Y = throughput on 2nd filter (as pounds of clothes)
A = weight of un-dried 2nd filter (in pounds)
B = weight of dry 1st filter (in pounds)
BI= weight of dry 2nd filter (in pounds)
C = weight of new filter (in pounds)
L = loss of perchloroethylene attributed to a cartridge filter
Lj= total perchloroethylene losses attributed to normal filter change
EXPLANATION AND RATIONALE:
In order to determine perchloroethylene losses from cartridge filters at
this plant, the initial plan called for removing one bank (four (4) cartridges)
before the mass balance test and replacing the expended cartridges with new ones.
These filters were to be drained for a minimum of 24 hours. Weight (wet) was
i
to be determined for one of the expended filters immediately and again after the
filter was dried to a constant weight.
All four filters showed a wet weight of 52.5 pounds on Monday, March 26th.
On Tuesday, March 27 it was discovered that capacity of scale used, had been
exceeded. Therefore an alternate procedure was necessary. The reformulated
plan called for removing and weighing another used cartridge from the second
filter bank after the mass balance test. The weight (wet) was determined to be
92.0 pounds. This filter was then returned to the filter bank because it had not
been fully expended. Therefore a method to correlate the two filters had to be
derived. This can be accomplish by assuming that each pound of clothes cleaned
deposits the same weight of dirt on a cartridge and the weight of each new cartridge
is equilavent. The following derivation utilizes these assumptions. Note that
the throughput between each change of four (4) cartridges is actually the throughput
A-2
-------
for two (2) tubes of four (4) cartridges each. Plant records showed approximate
throughput between changes of either tube to be 12,000 pounds. Therefore 6,000
pounds throughput was used for four (4) cartridges changed for the test. (Note:
6,000 pounds throughput is also the manufacturer's suggested filter life).
A-3
-------
DERIVATION:
B-C = dirt in 1st filter (in pounds)
B-C = dirt (1bsj_
X throughput 1st filter (Ibs)
~- = dirt Qbs)
throughput 2nd filter (Ibs)
B-C = B1 - C
Solving For
= (B-C) + C
"PERC"
2nd Filter = A - B
1
Substituting
Y_
x (B-C) + C = (2nd Filter)
A _ Y_ "PERC"
L -
EXPLANATION:
The total loss attributed to cartridge filters change is equal to X 4, since
there were four (4) filters in a bank. The loss is commonly expressed in units
of pounds loss per hundred (100) pounds (Ibs) throughput. Therefore:
l - A - Bl
LT ' -~ (4) (100)
LT = 2.74 lbs/100 Ibs throughput (TABLE 2.1)
A-4
-------
EQUATION A 2 - AVERAGE STACK GAS VELOCITY
- KpCp
EQUATION A3 - AVERAGE STACK GAS VOLUMETRIC FLOWRATE
Qsd = 3,600' (Vs) ./
NOMENCLATURE
2 2
A = Cross-sectional area of stack, m (ft )
C = Pitot tube coefficient, dimensionless
K = Pitot tube constant,
34 97 ni [ (g/g-mole)(mm Hgfl
sec I tQu\t ^ Q\ I
f_ WM!
for the metric system and
ft
85'49 sec
tQb/1b-mo1e)(in. Hg)"[ h
(°R)(in. H20) J
for the English system.
M. = Molecular weight of stack gas, dry basis,
g/g-mole (Ib/lb-mole)
M = Molecular weight of stack has, wet basis,
g/g-mole (Ib/lb-mole)
= Md(1 - Bws}
ws
P. = Barometric pressure at measurement site,
Dar mm Hg (in. Hg)
A-5
-------
NOMENCLATURE (cont'd)
P = Stack static pressure, mm Hg (in.. Hg)
P = Absolute stack gas pressure, mm Hg (in. Hg)
= P. + P
bar g
P . . = Standard absolute pressure 760 mm Hg
510 (29.92 in. Hg)
Q . . = Dry volumetric stack gas flowrate corrected to
standard conditions, dscm/hr (dscf/hr)
t = Stack temperature, C (°F)
T = Absolute stack temperature, K ( R)
= 273 + ts for metric
= 460 + ts for English
Tstd = standard absolute temperature, 293 °K (528 °R)
V = Average stack gas velocity, m/sec (ft/sec)
p = Velocity head of stack gas, mm H20 (in. H20)
3,600 = Conversion factor, sec/hr
A-6
-------
EQUATION 4.1 - VOLUME TO MASS CONVERSION
(1000) (MW) (PPM)
(24.5)
WHERE:
EQUATION 4.2
MW = Molecular weight of perchloroethylene
PPM = Parts per million by volume
24.5 = Conversion Factor (Ideal gas law @ STP)
1,000 = Conversion Factor (grams to milligrams)
QSTD
= E
EQUATION 4.3
(ER ) X D = EA or £L
1,000
EQUATION 4.4
WHERE:
(EA + Ep) LD = ET
CM = Concentration mass in milligrams per cubic
meter (mg/m3) (Equation 4.1)
D = duration of cycle in minutes
for aeration cycle (2.14) minutes
for loading cycle (5.15) minutes
E/\ = average emissions during aeration cycle
ER = emission rate (Equation 4.2) in mg/min
E^ = average emission during loading cycle
A-7
-------
EQUATION 4.4 (cont'd)
Ej = total atmospheric emissions (daily)
in grams
LD = loads per day (from Table 3.2)
A-8
-------
APPENDIX B
FIELD & LABORATORY DATA
B-l
-------
CO
I
PO
DATE
April 6, 1979
April 9, 1979
April 10, 1979
April 12, 1979
April 13, 1979
April 18, 1979
April 19, 1979
April 20, 1979
TIME
1000
0845
1545
0945
1335
0830
1645
1000
CARTRIDGE
FILTER WEIGHT
(Kg) (IBS)
30.8 68.0
29.4 64.8
24.2 53.2
23.6 52.0
23.4 51.5
23.0 50.7
22.8 50.2
22.8 50.2
23.2* (51.2 lb*)
(NET FINAL)
*-SCALE CORRECTION FACTOR APPLIED
TABLE B-l- CARTRIDGE FILTER LOSS (RAW DATA)
-------
;
DATE
3/26/79
3/27/79
3/28/79
3/29/79
3/26/79
3/27/79
3/28/79
3/29/79
LOCATION
INLET
INLET
(AERATION)
INLET
(PURGE)
INLET (A)
INLET (P)
INLET (A)
INLET (P)
INLET (A)
INLET (P)
OUTLET (A)
OUTLET (P)
OUTLET (A)
OUTLET (P)
OUTLET (A)
OUTLET (P)
OUTLET (A)
OUTLET (P)
Ap
AVG.
.038
.053
.049
.042
.051
.022
.042
.029
.036
.017
.015
.010
.020
.008
.015
.019
.026
• |
77
120
107
116
118
117
100
116
116
90
90
88
95
97
91
95
94
Ts
(°R)
AVG.
537
580
567
576
578
577
560
576
576
550
550
548
555
557
551
555
554
Pg
("H20)
AVG.
-.92
-.92
-.74
-.92
-.74
-.92
-.74
-.92
-.74
+.06
+ .07
+ .06
+.07
+.06
+ .07
+.06
+ .07
Ps
("H20)
AVG.
29.102
29.102
29.116
29.392
29.406
29.592
29.606
29.522
29.536
29.174
29.175
29.464
29.465
29.664
29.665
29.594
29.595
Vs
(ft/sec)
11.342
13.897
13.206
12.288
13.567
16.009
12.072
10.168
11.361
7.646
7.176
5.840
8.286
5.222
7.120
8.033
9.432
QSTD
(scf/hr)
30,656.9
34,811.6
33,806.9
31,456.9
34,409.7
40,940.6
31,849.2
26,007.2
29,058.6
20,263.6
18,961.7
15,671.1
21,934.8
13,864.1
19,102.5
21,351.4
25,122.6
(scf/min)
510.9
580.2
563.4
524.3
573.5
682.3
530.8
433.5
484.3
337.7
216.0
261.2
365.6
231.1
318.4
335.9
418.7
( s cm/mi n)
14.47
16.43
15.95
14.85
16.24
19.32
15.03
12.28
13.72
9.56
8.95
7.40
10.35
6.54
9.02
10.08
11.86
TABLE:B-2 VOLUMETRIC FLOWRATE DATA
-------
CO
I
BAROMETRIC
PRESSURE
("Hg.) ]
RELATIVE
HUMIDITY
MONDAY
3/26/79
29.44
65
TUESDAY
3/27/79
29.17
77
WEDNESDAY
3/28/79
29.46
74
THURSDAY
3/28/79
29.665
71
FRIDAY
3/30/79
29.59
51
^Station Pressure - U.S. Weather Station, Syracuse, New York.
TABLE: B-3 WEATHER DATA
-------
TUESEAY MARCH 27th,1979
KLEEN-KORNOR
:ORTLAND , NEW YORK
TIME: APPROXIMATELY 11:00 a.m.
Scale - 0-100 ppn
Chart Speed - 1 cn/nin
Instrument - Becknan 402
FIGURE B-l - EXAMPLE CARBON BED BREAKTHROUGH
B-5
-------
SCALE;- ;q-io0 ppm
"CHART ~SPJEED~-T' cm/mi n.
INSTRUMENT -'BECKMAN 402
FIGURE B-2 - EXAMPLE INLET CONCENTF./.TIG.N DURING ORYCLEARING CYCLE
-------
INTRODUCTION: ADDITIONAL LABORATORY TESTING
The purpose of the laboratory testing was to investigate the applicability
of the candidate instruments to detect perchloroethylene (62^4) within the
range of acceptable concentrations. Initially the parameters sought were that
of minimum detectability and response time. After field observations, it was
decided to investigate further the parameters of temperature and moisture
response. The candidate instruments chosen under the initial phase of this
program are listed in Table B-4. A cost prohibitive instrument (44,000),
an OVA-128 portable hydrocarbon analyzer, manufactured by Century Systems, was
also utilized during the laboratory testing for informational purposes only.
MANUFACTOR COST (APPROXIMATE) MODEL
• TIF-HALOGEN LEAK DETECTOR $ 75-°° #HLD-440
• COOK MFG. - (METER-ALL) . 125.00 #423-100
• BACHARACH INSTRUMENT - (TLV - SNIFFER) 500.00 #0023-7350
TABLE B-4 - CANDIDATE INSTRUMENTS - LABORATORY TESTING
B-7
-------
TEST PROCEDURE AND DISCUSSION
A.) Sensitivity Response to Perchloroethylene
The candidate instruments were tested for their response to perchloroethylene
(02014). The apparatus used is illustrated in Figure B-3. The basic apparatus
consists of two bellows pumps connected in line to a manifold. The candidate
instruments and a continuous hydrocarbon analyzer withdrew samples from the
manifold. In all cases, the concentrations introduced to the instruments were
assumed equivalent. The continuous hydrocarbon analyzer utilized to measure the
perchloroethylene concentrations was a Beckman 402, which operates on the prin-
ciple of hydrogen flame ionization. A matched voltage output strip chart recorder
provided a hard-copy record of the flame ionization detector. Field testing
demonstrated the general applicability of the Beckman instrument to measure
perchloroethylene in air concentrations. The Beckman 402, therefore, was utilized
to quantify all perchloroethylene concentrations during the laboratory testing.
The candidate instruments were zeroed in the manifold with ambient air
(introduced with Pump B). Valve "A" was opened slightly, allowing a small amount
of perchloroethylene into the manifold. The candidate instruments and the con-
tinuous monitor responses were duly noted. Successive increases in concentrations
were accomplished by further opening of Valve "A". The response of the instru-
ments were manually recorded on the strip chart illustrations in Figures B-4
through B-7.
The laboratory results in Figure B-4 shows that both the Meter-All and TIF
instruments will respond to low levels (^20 ppm) of perchloroethylene. The
TLV-Sniffer responded downscale in three out of four trials. Therefore, it can
be said that the TLV-Sniffer is inadequate for detecting perchloroethylene
vapors. The Meter-All instrument, while sensitive to initial introduction of
perchloroethylene (Figures B-4 and B-6), appears to have an inconsistent
B-8
-------
CO
I
ID
ALUMINIZED GAS
SAMPLING BAG
AMBIENT A
f
VALVE A
, V
-/"^
PUMP "A" I
IR INLET V
THEROMETER
o
4 A A A
_.,. £^ £_^ ^_^
PORTS
-j^ C1 A MANIFOLD
1 DIIMD -a" \ "VALVE B
^
f
ooo
OQ ~
ou ^
L_ . —
IS
y
HYDROCARBON ANALYZER
FIGURE B-3 SENSITIVITY RESPONSE APPARATUS
-------
. j _ ; i o_ '
b .. __; 1 _i 4H.-~ "J HIT
I i • , T i t"
," ~ ." ~~ ~t I :"•
PERCHLOROETHYLENE (From 45 ppm
source) ^
'-T.~ii"i:r."T. :.."-izii~.
v-
-*
":• ../ ~ ,i .»
TFMPERATURE - 80°F • -- i ; - - -- :-
i ' ! i ' '
i '''It—
- r
~":.:^iiii — _-
i i
... •-— j-_
r ;- — - — |
.. — .. _- j
P_-«-
— r
FIGURE B-4 - SENSITIVITY TO PERCHLOROETHYLENE
B-10
-------
ITITTTI
i i; ; TIP (Increased response) ^
"
METER-ALL (Alarms
TIP - (Slight increase in response
ill! J : ;j i
i ; i i i ;" ; n
LU
o
AMBIENT AIR
-------
—,o
i ! j
! !_ I
"I T" ~
"~"r~ ."lin
t HT r
.1 I J ;
~~. !izzT— i! "! .". ~---—.-.'. ".t" ri
PERCHLOROETHYLENE (From +100 ppm source)
i. _ ___;__ i
'.._ ' i
. . —^ -|-. ...
FIGURE B-6 - SENSITIVITY TCL R,
OgJ^CHLOROETHYLENE
-------
I I
.._ I r .
— 1 .
TEMPERATURE
FIGURE B-X.- SENSITIVITY TO PERCHLUKUETHYLENE
B-13
-------
response to increased perchloroethylene concentrations (Figure B-4). Figure
B-7 demonstrates that both the TIF and Meter-All instruments will respond to
increased concentrations after zeroing. However, neither instrument has a very
effective means for zeroing due to the crude nature of the potentiometer circuit.
B.) Temperature Response Test
Based upon preliminary field observations, the candidate instruments were
further evaluated for their response to temperature changes. Figure B-8 is
generalized schematic of the testing apparatus. The sensor of the candidate
instruments were inserted into the manifold and zeroed. The air in the manifold
was drawn pass the sensors at 23°C (74°F). The TIF-HLD440 and the Bacharach TLV
Sniffer ahd no response to a 8-10°F temperature rise. The Meter-All #423-100
instrument alarmed with a half a degree (.5°F) increase in temperature. In order
to qualify further the temperature response of the instruments, a second method
was used. The instruments were zeroed in calm ambient air 23°C (74°F) and
inserted into a beaker (Figure B-8). Again the Meter-All #423-100 alarmed.
The TIF #HLD440 emitted a slight increase in signal, while the Bacharach TLV-
Sniffer had no response. The temperature during this experiment was recorded
as 57QC (135°F).
C.) Calm Air Test
In this case, instruments were zeroed in calm air at a temperature of 23°C
(74°F), and inserted into the manifold. Unheated air is then pumped through the
manifold. The Meter-All #423-100 alarmed, while both the TIF and Bacharach TLV-
Sniffer had no response.
D.) Water Vapor Test
The apparatus for this test is highlighted in Figure B-9. The basic idea
of the apparatus was to saturate air and introduce the air to the manifold at
room temperatures. For this experiment the instruments were zeroed in the
manifold. Conditions during zeroing were 23°C (74°F) and relative humidity of
B-14
-------
AMBIENT AIR INLET
GLASS JAR
HOT PLATE
THERMOMETER
VENT
A
A A
PORTS
MANIFOLD
TEMPERATURE RESPONSE APPARATUS (FLOWING)
PORT
THERMOCOUPLE
FIGURE B-8. TEMPERATURE RESPONSE APPARATUS (CALM AIR)
.. B-15
-------
IMPINGERS
AMBIENT AIR INLET
DO
cn
THERMOMETER
SATURATED
AIR
t
VENT
O O
PORTS
MANIFOLD
HOT PLATE
FIGURE B-9. WATER VAPOR TEST APPARATUS
-------
42%. The moisture content in the manifold was increased to a relative humidity
of 100%. This is a rise of 1.06% in absolute water content.
All three instruments responded to the increase in moisture. The Meter-All
#423-100 and TIP HLD-440 alarmed, while the Bacharach TLV-Sniffer showed a slow
upscale response.
B-17
-------
APPENDIX C
GAS STANDARD CERTIFICATION
SCOTT ENVIRONMENTAL TECHNOLOGY, INC.
C-l
-------
Scott Environmental Technology Inc.
Plumsteadville, PA 18949
(215) 766-8861
Madison Heights, Ml 48071
(313) 544-0625
SPECIALTY GAS DIVISION
San Bernardino, CA 92411
(714) 887-2571
TRW
Attn: Bob Jangleau
800 Follin Lane
Vienna, VA 22180
. April 10. 1979
Our Project No.: 3°66Q1
YourP.O.No.: H °8503
Gentlemen:
Thank you for choosing Scott for your Specialty Gas needs. The analyses for the gases ordered, as
reported by our laboratory, are listed below. Results are in volume percent, unless otherwise indicated.
ANALYTICAL REPORT
Cvl Nn C-1414
Component
Analytical
Accuracy ±2%—
Concentration
TETRACHLORO ETHYLENE
45.7 ppm
AIR
BALANCE
Cyl. Nn C-1560
Component
Analytical
Accuracy ±2%
Concentration
TETRACHLORO ETHYLENE
473 nnm\
AIR
BALANCE
Analyst
Tyl Nn C-1682
Analytical
Component
TETRACHLORO ETHYLENE
AIR
Concentration
92.8 com
BALANCE
\ ?
~ *• • . ^
''. *•**'"
Pyl- No V
Component
Analytical
Accuracy,
Concentration
^
Approved By
?o-^>s^^A CjjNte
ROBERT DENYSZYN
The only liability of this Company for gai which falls to comply with thl» analyoU dull be replacement thereof by the Company without extra co»t.
ACUBLEND®« CALIBRATION & SPECIALTY GAS MIXTURES P PURE GASES
ACCESSORY PRODUCTS » CUSTOM ANALYTICAL SERVICES
C-2
-------
APPENDIX D
PROJECT PARTICIPANTS
D-l
-------
MR. KEN CONSTANTINE (TRW) TEST TEAM MEMBER
MR.. DEXTER YOUNG (TRW) PROCESS MONITOR
MR. STEVEN LUTZ (TRW) STANDARD TASK MANAGER
MR. CHUCK KLEEBERG (EPA) CPB PROJECT OFFICER
MR. FRANK CLAY (EPA) EMB TECHNICAL MANAGER
MR. BOB JONGLEUX (TRW) FIELD TEST COORDINATOR
MR. BUD AMES (KLEEN KORNOR) PLANT OWNER & CONTACT
D-2
-------
APPENDIX E
FIELD LOG (NOTES)
E-l
-------
45 'I
-. 5-
L*i
'.J. _l :.,.....„]..• . '
! I . i . • |
! ' i i '
' ' • !
E-2
WITNESS:.
WITNESS:.
.DATE:.
. DATE:.
SIGNED.
DATE
-------
DATE
f
PROJECT
. 5"
MJO NO..
V*V>
2 /• 5" . : "X
_L2i^- J
46
~.
J
^
WITNESS:.
WITNESS:.
DATE:
DATE: E-3
SIGNED.
DATE
-------
''^\:c^r'^ c^-~c->.
WITNESS:
WITNESS:
.DATE:
DATE: E-4
SIGNED.
DATE
-------
MJO NO..
3
4-
7
c\.
3:07
/o
/.'07
WITNESS:.
WITNESS:.
.DATE:.
E-5
SIGNED.
'DATE
-------
DATE
49
.PROJECT.
.MJO NO.
^A^A^AA 5^ ^ -j^, L.*Vr
9 HI
I I'll
3 !
»\n/. <
WITNESS: .
.DATE:.
E-6
SIGNED.
DATE
-------
. ... ^
y^t _, .. ........ . . .... o-
/-j
\
WITNESS: DATE: , SIGNED.
WITNESS: DATF; E-7 DATE '.
-------
DATE
53
MJU NVJ.
CU1U
I l;
1 1:|3
WITNESS:
WITNESS:
DATE:
DATE:
E-8
SIGNED.
DATE _
-------
y
B-c • •?,- c
A - -
E-9
-------
APPENDIX F
CANDIDATE INSTRUMENT
SELECTION CRITERIA
F-l
-------
The criteria used to select the candidate instruments involved the factors
of instrument cost, and instrument availability. The goal was to find an
inexpensive instrument ($100 - $250 range) that would provide a rapid, yet
accurate check of perchloroethylene concentrations around perchloroethylene dry
cleaning operations.
A review of currently available instruments was undertaken (12/78). Manu-
facturers, suppliers and industry spokesmen were contacted for suggestions and
comments. Based upon that information three (3) candidate instruments were
selected. Table F-l tabulates the list from which the candidate instruments
were chosen.
F-2
-------
SELECTION LIST
Manufactor
TIP
TIP
TIP
COOK
SIPIN
JOHNSON
CPO
BACHARACH
GOW-MAC
Model
440
5000
HM 290
423-1000
SP 1 to 7
1565
DM 498
725
0023-7350
21-100
Cost
$125
$130
$250
>$100
^$350
^$500
$250
$370
$500
$300
Disposition
Selected
Not Listed by Manufactor
Not Listed by Manufactor
Selected
Not Commercially Available
(Development Stage)
Too expensive
(Shipment Delay-2 months)
Not Commercially Available
Selected (EPA loaned)
Inappropriate for
Perchloroethylene
F-3
------- |