STUDY TO SUPPORT NEW SOURCE PERFORMANCE
STANDARDS FOR
SOLVENT METAL CLEANING OPERATIONS
Contract No. 68-02-1329
Task Order No. 9
Appendix Reports
June 30, 1976
Prepared By:
D. W, Richards
K. S. Surprenant
The Dow Chemical Company
Midland, Michigan
Prepared For:
Emission Standards and Engineering Division
Office of Air Quality Planning
U.S. Environmental Protection Agency
-------
TJ
m
Z
D
-------
APPENDIX - A
Study to Support New Source
Performance Standards for
Solvent Metal Cleaning Operations
EMISSION SURVEY
Prepared By:
D. R. Heinz
H. W. Krimbill
Prepared for:
Emission Standards and Engineering Division
Office of Air Quality Planning
U.S. Environmental Protection Agency
-------
TABLE OF CONTENTS
SECTION PAGE
INDEX OF TABLES AND GRAPHS 2
I, INTRODUCTION 4
II, THE METAL WORKING INDUSTRY ,,, 6
III, PLANTS METAL CLEANING , ,,, 12
IV, TYPES OF SOLVENT CLEANING ,,,,,,, 22
V, VAPOR DEGREASERS ,,,,,,,,,,,,,,, , 32
VI, QUANTITY OF SOLVENTS USED 46
VII, VAPOR RECOVERY AND CONTROL SYSTEMS ,,, - 54
VIII, SOLVENT DISPOSAL 68
IX, SURVEY PROCEDURES r 84
1,
-------
INDEX
OF
TABLES AND GRAPHS
EXHIBIT TITLE PAGE
II-A SIZE OF METAL WORKING 8
I I-B CONCENTRATION OF. METAL WORKING ,,, 9
II-c GEOGRAPHIC ZONES , 11
III-A TOTAL PLANTS METAL CLEANING , ,,,, 13
11I-B NUMBER OF PLANTS METAL CLEANING (By SIC) 15
III-c PERCENT OF PLANTS.METAL CLEANING (BY SIC) ,,,,,,, 17
III-D PLANTS METAL CLEANING (BY PLANT SIZE) 19
III-E PLANTS METAL CLEANING (BY LOCATION) 21
IV-A SOLVENT CLEANING - ROOM TEMPERATURE
AND VAPOR DECREASING ,.,.' ,,,, 23
IV-B PLANTS SOLVENT CLEANING (BY SIC) 25
IV-c PERCENT SOLVENT CLEANING (BY SIC) ,,,, 27
IV-D SOLVENT CLEANING (BY PLANT SIZE) ,,,, 29
IV-E SOLVENT CLEANING (BY PLANT LOCATION) 31
V-A NUMBER OF VAPOR DEGREASERS , ,, 33
V-B NUMBER OF VAPOR DEGREASERS (BY SIC) ,,,, 35
V-c NUMBER OF VAPOR DEGREASERS (BY PLANT SIZE) 37
V-D NUMBER OF VAPOR DEGREASERS (BY PLANT LOCATION) ,, 39
V-E TYPES OF VAPOR DEGREASERS (BY SIC) 41
V-F TYPES OF VAPOR DEGREASERS (BY PLANT SIZE) 43
V-G TYPES OF VAPOR DEGREASERS (BY PLANT LOCATION) ,,, 45
2,
-------
EXHIBIT TITLE PAGE
VI-A SOLVENTS USED FOR VAPOR DECREASING ,,,, 47
VI-B SOLVENT USAGE FOR VAPOR DECREASING 49
VI-c SOLVENT USED FOR ROOM TEMPERATURE CLEANING 51
VI-D SOLVENT USAGE FOR ROOM TEMPERATURE CLEANING 53
VII-A USE OF VAPOR RECOVERY AND CONTROL SYSTEMS ,,, 55
VII-B USE OF VAPOR RECOVERY AND CONTROL SYSTEMS (By SIC),, 57
VII-c PERCENT OF PLANTS USING VAPOR RECOVERY AND
CONTROL SYSTEMS 59
VII-D USE OF VAPOR RECOVERY AND CONTROL SYSTEMS
(Bv PLANT SIZE) 61
VII-E PERCENT OF PLANTS USING VAPOR RECOVERY AND
CONTROL SYSTEMS (BY PLANT SIZE) ,, 63
VII-F USE OF VAPOR RECOVERY AND CONTROL SYSTEMS
(BY PLANT LOCATION) ,, .,,,., 65
VII-G PERCENT OF PLANTS USING VAPOR RECOVERY AND
CONTROL SYSTEMS (BY PLANT LOCATION) 67
VIII-A DISPOSAL ROUTES FOR SOLVENTS ,, 69
VIII-B QUANTITY OF SOLVENTS DISPOSED 71
VIII-c NUMBER OF PLANTS DISPOSING (BY SIC) 73
VIII-D PERCENT OF PLANTS DISPOSING (BY SIC) ,,, ,, 75
VIII-E NUMBER OF PLANTS DISPOSING (BY PLANT SIZE) ,, 77
VIII-F PERCENT OF PLANTS DISPOSING (BY PLANT SIZE) ,,,,,, 79
VIII-G NUMBER OF PLANTS DISPOSING (BY LOCATION) 81
VIII-H PERCENT OF PLANTS DISPOSING (BY LOCATION) ,,,,,,,, 83
3,
-------
I
INTRODUCTION
THIS STUDY HAS IDENTIFIED THE USE OF ORGANIC SOLVENTS IN
THE METAL WORKING INDUSTRY, THE STUDY EMPHASIZES HOW THE
SOLVENTS ARE USED/ THE VOLUMES/ THE USE OF VAPOR RECOVERY
AND CONTROL SYSTEMS/ AND THE METHODS OF SOLVENT DISPOSAL,
EACH ASPECT OF THE STUDY HIGHLIGHTS THE RELEVANCY OF THE
TYPE OF INDUSTRY/ THE PLANT SIZE/ AND THE GEOGRAPHIC
LOCATION,
IT IS IMPORTANT TO NOTE THIS IS A SURVEY OF THE METAL
WORKING INDUSTRY (SEE SECTION II FOR DEFINITION) AND THE
DATA CONTAINED HEREIN PERTAINS ONLY TO THIS SEGMENT OF ALL
INDUSTRY,
THE AREAS INCLUDED IN THIS STUDY ARE REPORTED AS FOLLOWS:
1, A DEFINITION OF THE METAL WORKING INDUSTRY WITH
REGARD TO ITS SIZE/ SCOPE/ AND GEOGRAPHIC CON-
CENTRATION, (SECTION II)
2, THE EXTENT OF METAL CLEANING OUTLINING THE SEGMENT
OF THE METAL WORKING INDUSTRY USING A METAL CLEANING
SYSTEM WITH EMPHASIS ON THE PORTION OF THAT SEGMENT
WHICH USES ORGANIC SOLVENTS, (SECTION III)
3, THE EXTENT OF SOLVENT CLEANING INCLUDING THE
PROPORTION OF VAPOR DECREASING AND ROOM TEMPERATURE
CLEANING, (SECTION IV)
-------
THE EXTENT OF VAPOR DECREASING REPORTING THE
NUMBER OF UNITS IN USE AND TYPES OF UNITS.
(SECTION V)
5, THE QUANTITY OF SOLVENTS USED FOR CLEANING IN
THE METAL WORKING INDUSTRY, (SECTION VI)
6, THE USE OF VAPOR RECOVERY AND CONTROL SYSTEMS
INCLUDING THE TYPES OF SYSTEMS AND THE EXTENT
OF THEIR USE, (SECTION VII)
7, THE METHODS USED FOR SOLVENT DISPOSAL AND THE
QUANTITIES OF SOLVENTS DISPOSED OF BY THESE
VARIOUS ROUTES, (SECTION VIII)
8, THE PROCEDURES USED IN OBTAINING THE DATA
PRESENTED IN THIS STUDY ARE REPORTED IN
SECTION IX,
5,
-------
II
THE METAL WORKING INDUSTRY
ORGANIC SOLVENTS ARE USED IN A WIDE VARIETY OF WAYS, SINCE
OUR CONCERN IS WITH THE USE OF SOLVENTS IN INDUSTRIAL DE-
GREASING AND CLEANING/ A DEFINABLE POPULATION TO SURVEY HAD
TO BE SELECTED, THIS POPULATION NOT ONLY HAD TO DEMONSTRATE
THE END USE/ BUT REPRESENT THE LARGEST SEGMENT OF THE END
USE, WITH THE CRITERIA MENTIONED/ THE METAL WORKING INDUSTRY
WAS SELECTED,
SCOPE OF METAL WORKING
THE UNITED STATES GOVERNMENT CLASSIFIES ALL INDUSTRY BY THE
STANDARD INDUSTRIAL CLASSIFICATION (S.I.C.) CODE SYSTEM,
IN 1972 THE METAL WORKING INDUSTRY WAS CLASSIFIED INTO THESE
EIGHT S.I.C.s:
25 FURNITURE AND FIXTURES
33 PRIMARY METAL INDUSTRIES
34 FABRICATED METAL PRODUCTS
35 MACHINERY EXCEPT ELECTRICAL
36 ELECTRICAL AND ELECTRONIC EQUIPMENT
37 TRANSPORTATION EQUIPMENT
38 INSTRUMENTS AND RELATED PRODUCTS
39 MISCELLANEOUS MANUFACTURING INDUSTRIES
THESE 2-oiGiT S.I.C, CODES ARE FURTHER SUBDIVIDED TO ^-
AND 6-DIGIT CODES GIVING EACH INDUSTRY A FURTHER SUBDIVISION
THAN LISTED IN THE ABOVE 2-DIGIT CODES, FOR THE PURPOSE OF
THIS STUDY, DATA IS PROVIDED TO THE 2-DIGIT CLASSIFICATION,
6.
-------
SIZE OF METAL WORKING *
NEARLY HALF OF THE TOTAL INDUSTRIAL ACTIVITY IN THE U,S, is
CLASSIFIED AS THE METAL WORKING MARKET, IT IS THE LARGEST
SINGLE MARKET ACROSS ALL U,S, INDUSTRY,
THE METAL WORKING MARKET:
SPENDS 45% OF TOTAL DOLLARS EXPENDED BY INDUSTRIAL
PLANTS FOR'MATERIALS,
MAKES 40% OF ALL CAPITAL EXPENDITURES,
EMPLOYS 47% OF ALL INDUSTRIAL WORKERS,
PAYROLL is 53% OF THE TOTAL FOR ALL INDUSTRIES,
As SHOWN IN EXHIBIT H-A, THERE ARE 43/562 PLANTS EMPLOYING
20 OR MORE WITH A TOTAL EMPLOYMENT OF 10/147/834 IN THE
METAL WORKING INDUSTRY, You WILL NOTE IN THE STUDY THE
NUMBER USED FOR TOTAL PLANTS IS 41/670, THIS IS DUE TO THE
ELIMINATION OF SITES NOT MANUFACTURING LIKE R&D SITES OR
HEADQUARTER LOCATIONS,
IT IS ALSO IMPORTANT TO NOTE THERE ARE 83/074 PLANTS EMPLOYING
LESS THAN 20, THE DATA IN THIS SURVEY EXCLUDES THIS SEGMENT
OF THE METAL WORKING INDUSTRY, BECAUSE OF THE LIMITED AVAIL-
ABILITY OF INFORMATION ON THIS SEGMENT OF THE METAL WORKING
INDUSTRY/ SAMPLING WAS NOT POSSIBLE, WHEN CONSIDERING TOTALS
IN THIS REPORT/ KEEP IN MIND IT DOES NOT INCLUDE THE 83/074
PLANTS EMPLOYING LESS THAN 20,
EXHIBIT II-B IS A COLOR CODED MAP BY COUNTY REFLECTING THE
CONCENTRATION OF METAL WORKING PLANTS,
'SOURCE: 1972 CENSUS OF MANUFACTURING
7,
-------
SIZE OF METAL WORKING
STANDARD
INDUSTRIAL
CODE
25
33
34
35
36
37
38
39
TITLES OF INDUSTRY GROUPS
FURNITURE a FIXTURES
PRIMARY ML INDUSTRIES
FABRICATED METAL PRODUCTS
MACHINERY EXCEPT ELECTRICAL
ELKTRIC & ELECTRONIC EQUIP,
TRANSPORTATION EQUIPMENT
INSTRUMENTS & RELATED PRODUCTS
MISC, MANUFACTURING INDUSTRIES
U, S, TOTAL
LESS THAN 20
EMPLOYEES
8,218
3,019
17,355
28,046
4,922
4,836
3,249
13,429
83,074
BY PRIMARY PRODUCT
TOTAL
PU\NTS
EMPWYING
20 OR MORE
893
3,672
11,686
12,263
7,042
3,806
2,617
1,583
43,562
TOTAL
IN PLANTS .
EMPTYING
20 OR MORE
140,033
1,325,592
1,646,258
2,135,282
2,137,686
2,042,694
486,755
233,534
10,147,834
NUMBER OF PLANTS
2O99
472
1,828
7,722
8,180
3,563
1,930
1,532
946
26,173
100-499
352
1,304
3,324
3,049 .
2,374
1,129
781
527
12,840
500 OR MORE
64
494
554
870
925
642
227
98
3,874
CO
-------
PAGE NOT
AVAILABLE
DIGITALLY
-------
Exhibit 11-B
9.
-------
GEOGRAPHICAL BREAKDOWN
GEOGRAPHIC ZONES
IN SEVERAL SECTIONS OF THIS STUDY/ THE DATA IS REPORTED WITH
REFERENCE TO GEOGRAPHIC LOCATIONS, EXHIBIT II~C SHOWS THIS
BREAKDOWN,
THE FOLLOWING LISTS THE STATES IN EACH ZONE:
NORTHEAST
SOUTHEAST
CONNECTICUT
DELAWARE
MAINE
MARYLAND
MASSACHUSETTS
NEW HAMPSHIRE
NEW JERSEY
NEW YORK
PENNSYLVANIA
RHODE ISLAND
VERMONT
ALABAMA
FLORIDA
GEORGIA
No, CAROLINA
So, CAROLINA
TENNESSEE
VIRGINIA
ARKANSAS
WASHINGTON/ D,C,
PUERTO Rico
MID-WEST
ILLINOIS
INDIANA
KENTUCKY
MICHIGAN
OHIO
WEST VIRGINIA
WISCONSIN
FAR WEST
ARIZONA
CALIFORNIA
HAWAII
IDAHO
NEVADA
OREGON
WASHINGTON
UTAH
NORTH CENTRAL
SOUTHWEST
COLORADO
IOWA
KANSAS
MINNESOTA
MISSOURI
MONTANA
NEBRASKA
NORTH DAKOTA
SOUTH DAKOTA
WYOMING
LOUISIANA
MISSISSIPPI
OKLAHOMA
TEXAS
NEW MEXICO
10,
-------
X
1—i
to
o
-------
Ill
PLANTS METAL CLEANING
TO BEST DETERMINE THE NUMBER OF PLANTS USING ORGANIC
SOLVENTS/ A COMPARISON IS MADE WITH THE OTHER OPTION
A PLANT HAS FOR METAL CLEANING/ THE USE OF AN ALKALINE
WASH SYSTEM,
EXHIBIT III-A; IT CAN BE SEEN IN THIS GRAPH THAT SOME
OF THE PLANTS USE BOTH SOLVENT AND ALKALINE SYSTEMS/
WHILE OTHERS USE JUST ONE SYSTEM, 34% OF THE PLANTS
REPORTED THEY DO NOT USE ANY SYSTEM,
THE FOLLOWING IS SEEN IN THIS GRAPH:
• 66% OR 27/315 PLANTS IN THE METAL WORKING INDUSTRY
DO METAL CLEANING,
• 29% OR 12/015 PLANTS USE A SOLVENT SYSTEM ONLY FOR
METAL CLEANING,
• 20% OR 8,305 PLANTS USE A SOLVENT SYSTEM AND ALSO
DO ALKALINE WASHING,
• THEREFORE/ 49% OF THE METAL WORKING INDUSTRY OR
20/320 PLANTS USE SOLVENTS,
12,
-------
EXHIBIT III-A
PLANTS METAL CLEANING
TOTAL PLANTS: 41,670
Do Not Use Cleaning System
Alkaline Wash System Only
Both Solvent and Alkaline
Solvent System Only
14,355
(34%)
66%
49%
$12,015^
' (29%)
p
S> N
-------
EXHIBIT III-B; THIS TABLE is A BREAKDOWN OF THE PLANTS
METAL CLEANING BY S.I.C, GROUP, IMMEDIATELY BELOW EACH
GROUP TITLE IS THE NUMBER OF PLANTS REPRESENTED IN THAT
GROUP. AS EXPECTED/ THE GROUPS WITH THE GREATEST NUMBER
OF PLANTS (34, 35, AND 36) HAVE THE MAJORITY OF PLANTS
METAL CLEANING,
14,
-------
NUMBER OF PLANTS METAL CLEANING
,,, USE THIS TYPE OF METAL
CLEANING PROCESS:
ALKALINE WASH SYSTEMS ONLY
SOLVENT SYSTEMS ONLY
BOTH TYPES OF SYSTEMS
THIS MANY PLANTS IN THESE S,l
TOTAL
41,670
6,995
12,015
8,305
25
METAL
•FURNITUR
878
221
252
147
33
PRIMARY
METALS
3,531
635
733
684
I.C, GROUPS
34
FABRICTD
PRODUCTS
11,474
2,428
2,653
2,151
• t i
35
NON-ELEC
MACHINRY
11,812
1,620
3,923
2,265
36
ELECTRIC
EQUIPMNT
6,532
909
2,208
1,408
37
TRANSPTN
EQUIPMNT
3,462
699
745
957
38
INSTMTS
& CLOCKS
2,430
141
1,028
562
39
MISC
INDUSTRY
1,551
342
473
131
,,, DO NOT USE A CLEANING SYSTEM: 14,355
258
1,479
4,242
4,004
2,007
1,061
699
605
-------
EXHIBIT III-c'. THIS GRAPH GIVES A BETTER PICTURE OF
THE USE OF SOLVENT CLEANING IN EACH S.I.C, GROUP,
IT SHOWS THE PERCENT OF PLANTS IN EACH GROUP USING
SOLVENTS, LOOKING AT "TOTAL SOLVENTS"/ IT CAN BE
SEEN THAT GROUPS 35 / 36/ 37 / AND 38 ARE ABOVE THE
INDUSTRY AVERAGE OF 49%, GROUP 38/ INSTRUMENTS AND
CLOCKS/ HAVE A LARGE PERCENT OF PLANTS USING SOLVENTS
AT 65%,
IF YOU NOTE GROUP 25 YOU CAN SEE THAT ALTHOUGH THIS
GROUP SHOWS 71% OF THE PLANTS METAL CLEANING/ ONLY
46% USE SOLVENTS, THIS IS DUE TO THE EXTENSIVE USE
OF ALKALINE CLEANING IN THIS GROUP,
16,
-------
PERCENT OF PLANTS METAL CLEANING
(By S.I.C. Groups)
23 33 34 35 36 37 g
Metal Primary Fabrictd Non-Elec Electric Transptn
Total Furniture Metals Products Machinry Equipmnt Equipmnt
Percent Of Plants Use This
Type System. . .
Both Solvent & Alkaline
Solvent System Only
•
&20^
29
•
^7 ^
§m
29
•
^19^
^^
21
•
§19$
fc
23
•
^ 1 Q ^
o. iy N
^
33
I21 $
w
34
§P^
§28^
22
38
nstmts
i Clock
bs|
^
42
s 39
Misc
Industry
W
^85
31
Total Solvents: 49 46 40 42 52 55 50 65 39
Alkaline Only 17 25 18 21 J4 JI4 20 _6 22
Total Metal Cleaning: 66 71 58 63 66 69 70 71 61
Do Not Use Metal
Cleaning System:
34
29
42
37
34
31
30
29
39
x
5
o
-------
EXHIBIT III-p; IT is EASILY SEEN IN THIS GRAPH THAT
THE LARGER THE PLANT/ THE GREATER THE CHANCE SOLVENTS
ARE USED, BUT/ THE LARGER PLANTS (EMPLOYING 500+)
REPRESENT A SMALL PORTION OF TOTAL PLANTS, ESPECIALLY
WHEN WE CONSIDER 83/074 PLANTS EMPLOYING LESS THAN 20,
18,
-------
METAL CLEANING - PLANT SIZE
This Many Plants Employing.
Total
41,670
Use This Type Of Metal
Cleaning System:
Alkaline Wash Systems Only
Solvent Systems Only
Both Types Of Systems
Do Not Use A Cleaning System:
This Percent Of Plants. . .
Use A Solvent Cleaning System
20-100
People
25,641
100-500
People
12,408
500+
People
3,621
6,995 3,843 2,430 722
12,015 7,557 3,581 877
8,305 3,762 3,022 1,521
14,355 10,479 3,375 501
49
44
53
66
X
X
^^
DO
<£>
-------
EXHIBIT III-E; THIS SHOWS THE NUMBER AND PERCENT OF
PLANTS IN EACH ZONE USING SOLVENTS, BY TAKING THE
PLANTS IN THREE OF THE SIX ZONES/ THE NORTHEAST/
THE MID-WEST/ AND THE FAR WEST/ IT REPRESENTS 78%
OF THE TOTAL PLANTS, THESE THREE ZONES ALSO SHOW
THE HIGHEST PERCENT OF PLANTS USING SOLVENTS,
20,
-------
METAL CLEANING - LOCATION
This Many Use:
Alkaline Wash Systems
Only
Solvent Systems Only
Both Types Of Systems
Do Not Use A Cleaning
System
This Percent Of
Plants....
Use A Solvent
Cleaning System
Of Plants Located In The
Total Northeast Southeast
Mid-West Southwest North Central Far West
41,670 13,050 3,957 13,492 2,403 2,960 5.808
6,995 1,864 563 2,686 246 613 1,023
12,015 4,173 948 3,920 377 775 1,822
8,305 2,797 930 2,492 368 611 1,107
14,355 4,216 1.516 4,394 1,412 961 1,856
-
49
53
47
48
31
47
50
X
X
w
-------
IV
TYPES OF SOLVENT CLEANING
SOLVENTS ARE USED TO CLEAN OR DECREASE IN BASICALLY TWO
WAYS/ AT ROOM TEMPERATURE OR IN A VAPOR STATE, ROOM
TEMPERATURE OR "COLD CLEANING" (MOST USED TERM) IS DONE
BY SEVERAL METHODS, THESE WOULD INCLUDE WIPING THE AREA
TO BE CLEANED/ SPRAYING OR DIPPING IN A TANK OF SOLVENT,
VAPOR DECREASING INVOLVES THE HEATING OF A SOLVENT TO
ITS BOILING POINT AND CREATING A VAPOR ZONE, THE PARTS
TO BE CLEANED ARE PLACED INTO THIS VAPOR ZONE, THE SOLVENT
VAPOR CONDENSES ON THE PART AND THE SOILS DRIP OFF WITH
THE SOLVENT,
EXHIBIT IV-A; OF THE TOTAL 20/320 PLANTS USING SOLVENTS/
54% OR 11/028 USE ROOM TEMPERATURE CLEANING ONLY/ 25%
OR 5/365 VAPOR DECREASE ONLY/ AND 20% OR 3/927 USE BOTH
SYSTEMS,
FROM THIS/ IT CAN BE DETERMINED THAT OF THE 20/320 PLANTS
74% OR 14/955 USE ROOM TEMPERATURE CLEANING WHILE 46% OR
9/292 VAPOR DECREASE,
22,
-------
EXHIBIT W-A
SOLVENT CLEANING
ROOM TEMPERATURE & VAPOR DECREASING
Plants Vapor Degreasing
Only
•mTOUMMft!
>!•!•
M.M.W.W.
Plants Using Both Systems
Plants Room Temperature
Cleaning Only
$11,0283
3 (54%) <
Total Plants
Vapor Degreasing
9,292-46%
Total Plants
Room Temperature
Cleaning
14,955-74%
23,
-------
EXHIBIT IV-B: THIS TABLE GIVES THE TOTAL NUMBER OF
PLANTS IN EACH S,I,C, GROUP USING SOLVENTS, IT ALSO
GIVES THE TYPE OF SOLVENT CLEANING DONE IN EACH GROUP,
GROUP 35, WITH THE LARGEST NUMBER OF PLANTS USING
SOLVENTS; HAS 4/147 PLANTS ROOM TEMPERATURE CLEANING,
THIS BY FAR EXCEEDS ANY OF THE OTHER GROUPS, THE
MAJORITY OF THE PLANTS VAPOR DECREASING FALL IN GROUPS
34, 35, AND 36,
24,
-------
SOLVENT CLEANING
ROOM TEMPERATURE & VAPOR DECREASING
THIS MANY PLANTS IN THESE S.I.C, GROUPS ,,,
,, USE THIS TYPE OF SOLVENT
SYSTEM FOR METAL CLEANING:
VAPOR DECREASING SYSTEMS ONLY
ROOM TEMPERATURE SYSTEMS ONLY
BOTH TYPES OF SYSTEMS
TOTAL
20,320
25
METAL
FURNITUR
399
33
PRIMARY
METALS
1,417
34
FABRICTD
PRODUCTS
4,804
35
NON-ELEC
MACHINRY
6,188
36
ELECTRIC
EQUIPMNT
3,616
37
TRANSPTN
EQUIPMNT
1,702
38
INSTRMTS
& CLOCKS
1,590
39
MISC
INDUSTR"
604
5,365
11,028
3,927
95
209
95
348.
825
246
1,282
2,826
696
1,086
4,147
955
1,367
1,191
1,058
376
962
362
575
605
410
236
263
105
DO
hO
un
-------
EXHIBIT IV-c: THIS GRAPH SHOWS THE RELATIONSHIP BY
S.I.C, GROUPS OF THE TYPE OF SOLVENT CLEANING SYSTEM
USED, THE PERCENT OF ONE OF THE SYSTEMS ONLY AND THE
PERCENT USING BOTH SYSTEMS TOGETHER GIVE THE TOTALS
AT THE BOTTOM FOR EACH GROUP,
GROUPS 36 (ELECTRIC EQUIPMENT) AND 38 (INSTRUMENTS &
CLOCKS) SHOW LARGEST USE OF VAPOR DECREASING, WHILE
GROUP 35 (NON-ELECTRIC MACHINERY) is LOW WITH ONLY 33%
OF THE PLANTS USING SOLVENTS VAPOR DECREASING-AND THERE-
FORE HAS THE HIGHEST PERCENT OF PLANTS USING ROOM TEMPERA-
TURE CLEANING AT 82%,
IT CAN BE ASSURED THAT THE GROUPS WITH THE HIGHEST
PERCENT USING BOTH SYSTEMS HAVE MORE SOPHISTICATED
CLEANING REQUIREMENTS, EXAMPLES OF THIS WOULD BE
GROUPS 36, ELECTRIC EQUIPMENT/ AND 38/ INSTRUMENTS &
CLOCKS,
26,
-------
SOLVENT CLEANING
ROOM TEMPERATURE & VAPOR DECREASING
Total
25 33 34 35 36 37 38 39
Metal Primary Fabrictd Non-Elec Electric Transptn Instrmts Misc
Furniture Metals Products Machinry Equipmnt Equipmnt % Clocks Industry
ro
Percent Vapor
Degreasing Only
Percent Using Both
Systems
Percent Room
Temperature
Cleaning Only
54
Total: -
Vapor Degreasing 46
Room Temperature
Cleaning 74
52
48
76
58
59
67
(38:
33
57
42
75
41
73
33
82
67
62
43
78
38
62
64
44
56
61
x
X
t—t
03
i
o
-------
EXHIBIT IV-p; PLANT SIZE HAS A DIRECT RELATIONSHIP
TO THE TYPE OF CLEANING DONE, THE GRAPH CLEARLY IN-
DICATES THE FOLLOWING:
o THE LARGER THE PLANT/ THE GREATER THE PERCENT
OF VAPOR DECREASING ONLY
• THE LARGER THE PLANT/ THE SMALLER THE PERCENT
. OF ROOM TEMPERATURE CLEANING,
• THE LARGER THE PLANT/ THE GREATER PERCENT OF
USING BOTH TYPES OF SYSTEMS
28,
-------
SOLVENT CLEANING
ROOM TEMPERATURE & VAPOR DECREASING
Use This Type Of Solvent
System For Metal Degreasing:
Vapor Degreasing Systems Only
Room Temperature Systems Only
Both Types Of Systems
This Percent Of Plants. . .
Vapor Degrease Only
Both Types Of Systems
Room Temperature Only
Total
20.320
5,365
11,028
3,927
54
This Many Plants Employing. .
20-100
People
11,319
2,562
7,494
1,263
66
100-500
People
6,603
1.883
2,868
1,852
m/m
M
&28S
bJ
43
500+
People
2,398
920
666
812
^^
3P
28
x
n:
»—«
DO
-------
EXHIBIT IV-E; THE DIFFERENT REGIONS OF THE COUNTRY
VARY TO A GREAT DEGREE ON THE TYPES OF SOLVENT CLEANING
SYSTEM USED, THE NORTHEAST AND FAR WEST HAVE THE
LARGEST PERCENT OF PLANTS IN THEIR ZONES VAPOR DECREASING,
50% AND 55%, THE SOUTHWEST HAS THE LEAST PERCENT VAPOR
DECREASING AT 29%, OTHER THAN THE SOUTHWEST/ THE BALANCE
OF THE COUNTRY HAS AN ALMOST IDENTICAL PERCENT OF ROOM
CLEANING SOLVENT USE,
30,
-------
SOLVENT CLEANING
ROOM TEMPERATURE & VAPOR DECREASING
Of Plants Located In The. . .
This Many Use:
Total
Northeast Southeast Mid-West Southwest North Central Far West
20,320
6,970 1,878 6,412 745 1,386 2,929
Vapor Oegreasing
Systems Only 5,365 1,910 442 1,738 119 328 828
Room Temperature
Systems Only 11,028 3,511 1,178 3,614 530 872 1,323
Both Types Of Systems 3,927 1,549 258 1,062 96 186 778
This Percent Of Plants. . .
Vapor Degrease Only
Both Types Of Systems
Room Temperature Only
S
54
B
S
50
H
|l4^
62
H
^^
56
H
lh3N
71
B
§13^
63
8
x27 v
45
m
x
-------
VAPOR DEGREASERS
IT HAS BEEN SHOWN THAT 20/320 OF THE 41/670 PLANTS IN
THE METAL WORKING INDUSTRY USE SOLVENTS, ALSO/ THAT
9,292 OF THE PLANTS USING SOLVENTS/ VAPOR DECREASE,
THIS SECTION IDENTIFIES THE NUMBER OF VAPOR DEGREASERS
AND THE TYPE USED IN THE DIFFERENT S.I.C, GROUPS/ BY
PLANT SIZE AND BY GEOGRAPHIC LOCATION,
EXHIBIT V-A: THERE ARE 18/090 VAPOR DEGREASERS IN THE
41/670 PLANTS INCLUDED IN THIS SURVEY, THE 18/090
VAPOR DEGREASERS ARE FOUND IN 9/292 PLANTS GIVING AN
AVERAGE NUMBER OF 1,95 UNITS PER PLANT,
32,
-------
EXHIBIT V-A
NUMBER OF VAPOR DEGREASERS
TOTAL NUMBER OF PLANTS
VAPOR DECREASING: 9,292
TOTAL NUMBER OF VAPOR
DECREASING UNITS: 18,090
AVERAGE NUMBER OF UNITS
PER PLANT: 1,95
33,
-------
EXHIBIT V-BJ GROUPS 34, 35, AND 36 HAVE 69% OF THE
PLANTS VAPOR DECREASING AND 63% OF THE TOTAL VAPOR
DEGREASERS, OTHER GROUPS LIKE 33, 37, AND 38 HAVE
FEWER PLANTS VAPOR DECREASING BUT THE AVERAGE NUMBER
OF UNITS PER PLANT IS MUCH HIGHER,
-------
U1
NUMBER OF VAPOR DEGREASERS
(BY S.I.C, GROUP)
TOTAL
25 33 34 35 36 37 38 39
METAL PRIMARY FABRICTD NON-ELEC ELECTRIC TRANSPTN INSTRMTS MISC
FURNITUR METALS PRODUCTS MACHINRY EQUIPMNT EQUIPMNT & CLOCKS INDUSTRY
NUMBER OF PLANTS
VAPOR DEGREASING:
9,292
190
591 1,978 2,011 2,425
738
985
341
NUMBER OF VAPOR
DEGREASING UNITS:
18,090 376
1,315 3,214 3,285 4,919
2,400 2,233
348
AVERAGE NUMBER OF
UNITS PER PLANT:
1,95 1,98
2,21
1,62
1,61 2,03
3,25 2,27
1,02
-------
EXHIBIT V-c: As THE PLANT SIZE INCREASES/ THE NUMBER
OF PLANTS VAPOR DECREASING FALLS OFF CONSIDERABLY,
THE NUMBER OF VAPOR DEGREASERS REMAINS FAIRLY CONSTANT
IN EACH PLANT SIZE GROUPING, THEREFORE, AS EXPECTED/
THE LARGER PLANTS HAVE MORE UNITS PER PLANT,
36,
-------
NUMBER OF VAPOR DEGREASERS
(BY PLANT SIZE)
EMPLOYMENT
TOTAL
20-100
PEOPLE
100-500
PEOPLE
500+
PEOPLE
NUMBER OF PLANTS VAPOR
DECREASING:
9,292
1,287
1,591
NUMBER OF VAPOR DECREASING
UNITS:
18,090
6,149
5,966
5,975
AVERAGE NUMBER OF UNITS
PER PLANT:
1,95
1,75
3,76
-------
EXHIBIT V-D; BY LOOKING AT THE NUMBER OF PLANTS VAPOR
DECREASING BY GEOGRAPHIC LOCATION/ THE NORTHWEST/
MID-WEST/ AND FAR WEST REGIONS RESPECTIVELY HAVE THE
LARGEST NUMBER, THIS IS ALSO TRUE WITH REGARD TO
NUMBER OF UNITS, BUT/ THE FAR WEST HAS A MUCH HIGHER
AVERAGE NUMBER OF UNITS PER PLANT AT 2,58, THIS BRINGS
THE BALANCE OF THE REGIONS BELOW THE NATIONAL AVERAGE
OF 1,95,
38,
-------
NUMBER OF VAPOR DEGREASERS
(BY PLANT LOCATION)
TOTAL NORTHEAST SOUTHEAST MID-WEST SOUTHWEST NORTH CENTRAL FAR WEST
NUMBER OF PLANTS
VAPOR DECREASING:
9,292 3,452
698
2,798
212
506
1,626
NUMBER OF VAPOR
DECREASING UNITS:
18,090
6,357
1,206
5,090
377
870
4,190
AVERAGE NUMBER OF
UNITS PER PLANT:
1,95
1,84
1,73
1,82
1,78
1,72
2,58
co
LO
-------
EXHIBIT V-E; IN CONSIDERING THE TYPES OF VAPOR
DEGREASERS/ THEY WERE CLASSIFIED IN TWO GENERAL
CATEGORIES: CLOSED CONVEYOR AND OPEN TOP, THERE
ARE A TOTAL OF 18/090 VAPOR DEGREASERS, 2,796 OR
15% ARE CLOSED CONVEYOR AND 15/294 OR 85% ARE OPEN
TOP, IN THE GRAPH/ THE BAR REPRESENTS 100% OF THE
DEGREASERS, THE AMOUNT OF THE BAR ABOVE OR BELOW
THIS LINE REPRESENTS THE PERCENT OF DEGREASERS OF
THE CLOSED CONVEYOR OR OPEN TOP TYPE, THIS DEMON-
STATES THAT GROUPS 25/ 33 / AND 34 HAVE THE LARGEST
PERCENT OF CLOSED CONVEYOR TYPE DEGREASERS WITH
GROUP 38 THE LEAST,
40,
-------
TYPES OF VAPOR DECREASERS
(BY S.I.C. GROUP)
Total Units:
Closed Conveyor:
Open Top:
25 33 34 35 36 37 38 39
Metal Primary Fabrictd Non-Elec Electric Transptn Instrmts Misc
Total Furnitur Metals Products Machinry Equipmnt Equipmnt & Clocks Industry
18,090
2,796
:•:•:•:•:•:
&:*#:
15,29<
I
376
100
m
* • • • •
276
1,315
282
m
ftWS
.V.'.V.
1,064
3,214
675
:i:i:i::!:
m%
2,539
3,285
624
:#*:
>:**
,'"•*• • •
2,661
4,919
618
vX»X
j:jx*:
4,301
2,400
322
Xw
•'•'•'•'•'
2,078
2,233
135
.•.•••;•;•;
2,098
348
40
m
308
rr
X
X
m
-------
EXHIBIT V-F: IT WOULD BE ASSUMED THAT THE LARGER
PLANTS WOULD HAVE THE HIGHER PERCENTAGE OF CLOSED
CONVEYOR DEGREASERS, BUT BY LOOKING AT THIS GRAPH,
IT CAN BE SEEN THAT PLANT SIZE HAS NO EFFECT,
42,
-------
TYPES OF VAPOR DEGREASERS
(BY PLANT SIZE)
Total Units:
Closed Conveyor:
Total
18,090
2,796
::<:*:#
$m%
•• :•!•%• v.
20-100
People
6,149
1,003
::***
;«!
•"•I* •»•-•-•
1 00-500
People
5,966
809
ill
1ft • • • '
500+
People
5,975
984
••: ::••
'-'•: ::•*••
"' i *• •
::• ••::
Open Top:
15,294
5,146
5,157
4,991
x
-------
EXHIBIT V-G; THE TYPE OF VAPOR DEGREASER USED DOES
VARY WITH REGARD TO GEOGRAPHIC LOCATION, BY LOOKING
AT THAT BAR GRAPH/ YOU CAN SEE THE REGIONS WITH A
LOW PERCENT OF OPEN TOP (SOUTHWEST AND FAR WEST),
WHILE THE SOUTHEAST AND MID-WEST HAVE A HIGH PERCENT
OF CLOSED CONVEYOR,
-------
Total Units:
Closed Conveyor:
18,090
2,796
Open Top:
TYPES OF VAPOR DECREASERS
(BY PLANT LOCATION)
Total Northeast
North
Southeast Mid-West Southwest Central
6,357
924
1,206
265
5,090
1,054
377
34
870
145
Far West
4,190
373
5,29
4
5,432
941
4,036
342
/2b
3,817
X
n:
I-*
CD
-------
VI
QUANTITY OF SOLVENT USED
THIS SECTION OUTLINES WHICH SOLVENTS ARE USED IN
VAPOR DECREASING AND ROOM TEMPERATURE CLEANING,
IT IS DETERMINED HOW MANY PLANTS USE A PARTICULAR
SOLVENT AND THE QUANTITY USED,
EXHIBIT VI-A: TRICHLOROETHYLENE is THE MOST WIDELY
USED SOLVENT FOR VAPOR DECREASING WITH 5,W PLANTS
OR 59% OF THE TOTAL 9/292 PLANTS VAPOR DECREASING
USING IT, I/I/I-TRICHLOROETHANE is USED IN 21%
OF THE PLANTS FOR VAPOR DECREASING,
46,
-------
SOLVENTS USED FOR
VAPOR DECREASING
SOLVENT:
EXHIBIT VI-A
NUMBER OF PERCENT OF
PLANTS USING PLANTS USING
TRICHLOROETHYLENE 5,W 59
1,1,1-TRICHLOROETHANE 1,910 21
PERCHLOROETHYLENE 1,486 16
METHYLENE CHLORIDE 142 1,5
FLUOROCARBONS LOW 11
47,
-------
EXHIBIT VI-B; THIS TABLE GIVES THE QUANTITY OF
EACH VAPOR DECREASING SOLVENT USED AND THE AVERAGE
USE PER PLANT IN A YEAR, THIS TABLE SHOWS THAT
ALTHOUGH TRICHLOROETHYLENE AND 1/1/1-TRICHLOROETHANE
USAGE IS SIMILAR/ THE AVERAGE QUANTITY USED PER PLANT
OF 1/1/1 IS THREE TIMES GREATER THAN TRI, THIS
INDICATES THAT PLANTS WITH A LARGE SOLVENT REQUIRE-
MENT HAVE SWITCHED FROM TRICHLOROETHYLENE TO 1/1/1-
TRICHLOROETHANE,
48,
-------
SOLVENT:
SOLVENT USAGE FOR
VAPOR DECREASING
EXHIBIT VI-B
PLANTS GAL/MO
USING
(xlO3)
LBS/YR
(xlO3)
AVERAGE
LBS/YR/PLANT
TRICHLOROETHYLENE
1,1,1-TRICHLOROETHANE
PERCHLOROETHYLENE
METHYLENE CHLORIDE
FLUOROCARBONS
5,117
1,910
1,186
112
1,011
771
899
283
62
235
111,156
118,668
11,118
8,181
36,660
20,161
62,130
29,709
57,631
36,151
19,
-------
EXHIBIT VI-c: THERE ARE 14,955 PLANTS, OR 74%
OF THE PLANTS SOLVENT CLEANING, WHICH ARE ROOM
TEMPERATURE CLEANING, MANY OF THESE PLANTS USE
MORE THAN ONE OF THE SOLVENTS, SO THE NUMBER OF
PLANTS USING WILL NOT TOTAL 14,955, PETROLEUM
SOLVENTS ARE THE MOST WIDELY USED AT 42%, TRI-
CHLOROETHYLENE, 1,1,1-TRICHLOROETHANE AND SAFETY
BLENDS ARE A FAR SECOND CHOICE,
50,
-------
EXHIBIT VI-c
SOLVENTS USED FOR
ROOM TEMPERATURE CLEANING
SOLVENT:
NUMBER OF PERCENT OF
PLANTS USING PLANTS USING
TRICHLOROETHYLENE 2,295 15
1,1,1-TRICHLOROETHANE. 2,106- 11
PERCHLOROETHYLENE 702 5
METHYLENE CHLORIDE 324 2
FLUOROCARBONS 1,026 7
PETROLEUM SOLVENTS 6,344 42
ACETONE 1,215 8
METHYL-ETHYL-KETONE 648 4
TOLUENE 837 6
ALCOHOLS 945 6
ETHERS 27
SAFETY BLENDS 2,079 14
CARBON TETRACHLORIDE 162 1
51
-------
EXHIBIT VI-p; BY TAKING THE POUNDS PER YEAR USED
BY ANY SOLVENT AND DIVIDING IT BY THE PLANTS USING
THAT PARTICULAR SOLVENT/ THE AVERAGE POUNDS USED
PER YEAR PER PLANT IS GIVEN, THE LARGEST SINGLE
USE IS PETROLEUM SOLVENTS AT 73 MILLION POUNDS
PER YEAR, BUT THE AVERAGE PLANT USES ONLY 11/560
POUNDS PER YEAR WHILE THE AVERAGE PLANT USING
1/1/1-TRICHLOROETHANE USES 32/028 POUNDS PER YEAR,
52,
-------
EXHIBIT VI-D
SOLVENT USAGE FOR
ROOM TEMPERATURE CLEANING
SOLVENT:
TRICHLOROETHYLENE
1,1,1-TRICHLOROETHYLENE
PERCHLOROETHYLENE r
METHYLENE CHLORIDE
FLUOROCARBONS
PETROLEUM SOLVENTS
ACETONE
METHYL ETHYL KETONE
TOLUENE
ALCOHOLS
ETHERS
SAFETY BLENDS
CARBON TETRACHLORIDE
PLANTS
USING
2,295
2 JOG
702
32'!
1,026
6,344
1,215
648
837
945
27
2,079
162
GAL/MO
(xlO3)
299
511
61
52
125
926
110
86
138
77
2
180
12
LBS/YR
(xlO3)
43,056
67^52
9)516
6,864
19,500
73,339
8,712
6,811
11,923
6,098
-
16,200
1,584
AVERAGE
LBS/YR/PLANT
18,760
32,028
13,556
21,185
19,006
11,560
7,170
10,511
14,245
6,453
7,792
9,778
53,
-------
VII
VAPOR RECOVERY AND CONTROL SYSTEMS
THIS SECTION GIVES THE NUMBER AND PERCENT OF PLANTS
USING A VAPOR RECOVERY OR CONTROL SYSTEM, SOLVENT
VAPORS CAN BE RECOVERED FROM AIR BY THE USE OF A
CARBON ABSORPTION UNIT, VAPORS ARE CONTROLLED BY
REFRIGERATION/ A WATER BARRIER/ AND VAPOR BURNING,
EXHIBIT VII-A; 76% OF THE PLANTS USING SOLVENTS USE
NO VAPOR RECOVERY OR CONTROL SYSTEM, THE ONLY SYSTEMS
USED TO ANY EXTENT IS REFRIGERATION AT 12%, 8% OF
THE PLANTS ARE LISTED IN THE "OTHER7' CATEGORY/ BUT
CANNOT BE CONSIDERED AS USING GOOD RECOVERY AND
CONTROL SYSTEMS, (A PLANT MIGHT CONSIDER THE LEAVING
OF THE BACK DOOR OPEN/ A VAPOR CONTROL SYSTEM,)
-------
EXHIBIT VII-A
USE OF VAPOR RECOVERY AND CONTROL SYSTEMS
TYPE OF SYSTEM:
CARBON ABSORPTION
REFRIGERATION
VAPOR BURNING
WATER BARRIER
OTHER
NONE
NUMBER
OF PLANTS
305
2,177
351
37
1,632
15.519-
PERCENT
OF PLANTS
1,5
12
2
,2
8
76
PLANT USING SOLVENTS: 20,320
55,
-------
56,
EXHIBIT VII-B! THIS TABLE PROJECTS THE NUMBER OF
PLANTS USING VAPOR RECOVERY AND CONTROL SYSTEMS BY
S,I,C, GROUP, AREAS LEFT BLANK INDICATE TOO SMALL
A SAMPLING TO PROJECT A TOTAL, THE LARGEST USER
OF VAPOR RECOVERY AND CONTROL SYSTEMS IS GROUP 36/
ELECTRIC EQUIPMENT,
-------
USE OF VAPOR RECOVERY & CONTROL SYSTEMS
,, USE THIS TYPE OF SYSTEM:
CARBON ABSORPTION
REFRIGERATION
VAPOR BURNING
WATER BARRIER
SOME OTHER TYPE
,, DO NOT USE RECOVERY OR
CONTROL SYSTEMS:
THIS MANY PLANTS IN THESE S,
TOTAL
20,320
305
2,477
351
37
1,631
15,519
25
METAL
FURNITUR
399
-
18
36
18
18
309
33
PRIMARY
METALS
1,417
57
95
-
-
210
1,055
I,C, GROUPS
34
FABRICTD
PRODUCTS
4,804
40
436
99
-
415
3,814
iii
35
NON-ELEC
MACHINRY
6,188
100
477
59
-
379
5,173
36
ELECTRIC
EQUIPMNT
3,616
83
712
42
-
356
2,423
37
TRANSPTN
EQUIPMNT
1,702
-
310
-
19
39
1,334
•38
INSTRNTS
& CLOCKS
1,590
25
350
75
-
175
965
39
MISC
INDUSTRY
604
-
79
40
-
39
446
-------
EXHIBIT VII-c: THIS GRAPH GIVES A CLEAR PICTURE OF
THE USE OF VAPOR RECOVERY AND CONTROL SYSTEMS BY
S.I.C, GROUP, GROUPS 36 AND 38 HAVE THE LARGEST
PERCENT OF PLANTS USING RECOVERY AND CONTROL SYSTEMS,
58,
-------
PERCENT OF PLANTS USING
VAPOR RECOVERY & CONTROL SYSTEMS
(By S.I.C. GROUP)
Plants Using
Solvents:
Total
20,320
25
Metal
Furnitur
399
33
Primary
Metals
1,417
34
Fabrictd
Products
4,804
35
Non-Elec
Machinry
6,188
36
Electric
Equipmnt
3,616
37
Transptn
Equipmnt
1,702
38
Instrmts
& Clocks
1,590
39
Misc
Industry
604
24
23
26
21
16
33
22
39
26
x
X
o
-------
EXHIBIT VII-p; THE LARGER THE PLANT/ THE GREATER
THE USE OF VAPOR RECOVERY AND CONTROL SYSTEM, IN
PLANTS EMPLOYING FROM 20 TO 100 PEOPLE ONLY ,8% USE
CARBON ABSORPTION/ WHILE PLANTS EMPLOYING OVER 500
PEOPLE HAVE 4,6% OF THE PLANTS USING SOLVENTS USING
CARBON ABSORPTION,
60, .
-------
USE OF VAPOR RECOVERY AND CONTROL SYSTEMS
,,, USE THIS TYPE OF
SYSTEM:
CARBON ABSORPTION
REFRIGERATION
VAPOR BURNING
WATER BARRIER
OTHER
,,, DO NOT USE RECOVERY
ON CONTROL SYSTEM
(BY PLANT SIZE)
THIS MANY PLANTS EMPLOYING.,,,
TOTAL
20,320
305
2,477
351
37
1,632
20-100 100-500
PEOPLE PEOPLE
. 11,319 6,603
90 104
976 1,030
183 93
37
722 675
500+
PEOPLE
2,398
111
471
75
-
234
15,519
9,310
4,701
1,508
cn
-------
EXHIBIT VII-E: THE PERCENT OF PLANTS USING ANY
TYPE OF RECOVERY OR CONTROL SYSTEM INCREASES
DRAMATICALLY AS PLANT SIZE INCREASES,
62,
-------
PERCENT OF PLANTS USING
VAPOR RECOVERY & CONTROL SYSTEMS
(BY PLANT SIZE)
Plants Using Solvents:
Total
20,320
20-100
People
11,319
100-500
People
6,603
500+
People
2,398
24
18
29
37
x
X
DO
cn
i
m
-------
EXHIBIT VII-F: THIS TABLE PROJECTS THE NUMBER OF
PLANTS USING THE RECOVERY AND CONTROL SYSTEMS IN
THE VARIOUS GEOGRAPHIC ZONES, SOME PROJECTIONS
ARE NOT MADE DUE TO AN INADEQUATE SAMPLE SIZE,
-------
USE OF VAPOR RECOVERY 8 CONTROL SYSTEMS
,,, USE THIS TYPE OF SYSTEM;
CARBON ABSORPTION
REFRIGERATION
VAPOR BURNING
WATER BARRIER
SOME OTHER TYPE
OF PLANTS LOCATED IN
TOTAL
20,320
305
2,477
351
37
1,631
NORTHEAST
6,970
127
861
142
19
612
THE ,,,
SOUTHEAST
1,878
—
173
26
-
73
MID-WEST
6,412
141
717
136
18
439
SOUTHWEST
745
—
36
26
-
36
NORTH CENTRAL FAR WEST
1,386 2,929
37
124 566
21
-
143 328
,,, DO NOT USE RECOVERY
OR CONTROL SYSTEMS
15,519
5,209
1,606
4,961
647
1,119
1,977
cn
-------
EXHIBIT VII-G! THIS GRAPH INDICATES THE PERCENT OF
PLANTS USING VAPOR RECOVERY AND CONTROL SYSTEMS BY
GEOGRAPHIC ZONE, THE FAR WEST HAS THE LARGEST
PERCENT OF PLANTS USING RECOVERY AND CONTROL SYSTEMS
WHILE THE SOUTHEAST AND SOUTHWEST HAS THE LEAST PERCENT,
66,
-------
PERCENT OF PLANTS USING
VAPOR RECOVERY & CONTROL SYSTEMS
(BY PLANT LOCATION)
Plants Using Solvents:
Total
20,320
Northeast Southeast Mid-West Southwest
6,970
1,878
6,412
745
North
Central
1,386
Far West
2,929
24
25
14
23
13
19
33
x
DO
I
en
-------
VIII
SOLVENT DISPOSAL
SOLVENTS CAN BE DISPOSED OF BY SEVERAL ROUTES: BURNING/
FLUSHING, LAND, FILL/ DISPOSAL SERVICE AND RECLAIMER,
THIS SECTION LOOKS AT THESE ROUTES GIVING THE NUMBER
AND PERCENT OF PLANTS USING THE VARIOUS DISPOSAL ROUTES,
THE QUANTITY OF SOLVENTS DISPOSED OF BY THESE ROUTES IS
ALSO REPORTED,
EXHIBIT III-A; THIS TABLE INDICATES THE NUMBER OF PLANTS
USING THE VARIOUS DISPOSAL ROUTES AMD THE PERCENT EACH
IS OF THE TOTAL, (IT DOES NOT TOTAL 100% SINCE SOME
PLANTS MAY USE MORE THAN ONE ROUTE), THE LARGEST METHOD
OF DISPOSAL IS DONE WITH A DISPOSAL SERVICE AT 39% OR
7/989 PLANTS, ONLY 2% OF THE PLANTS BURN THEIR SOLVENTS
AS THEIR METHOD OF DISPOSAL,
68,
-------
DISPOSAL ROUTES FOR SOLVENTS
DISPOSAL ROUTE:
NUMBER
OF PLANTS
EXHIBIT VIII-A
PERCENT
OF PLANTS
BURNING
FLUSHING
LAND FILL
DISPOSAL SERVICE
RECLAIMER
382
2,667
3,674
7,989
4,323
2
13
18
39
21
69,
-------
EXHIBIT VIII-B: ALTHOUGH MORE PLANTS USE A DISPOSAL
SERVICE/ THE LARGEST QUANTITY OF SOLVENTS GOES TO RE-
CLAIMERS, THE AVERAGE GALLONS OF SOLVENTS PER PLANT
GOING TO A RECLAIMER IS 2/509 GALLONS PER YEAR,
BY TOTALING THE GALLONS PER MONTH DISPOSED AND THE
QUANTITY OF SOLVENTS USED/ THE FOLLOWING RELATIONSHIP
CAN BE SEEN:
QUANTITY OF SOLVENTS USED: GAL/|V!ONTH
4,832/000
QUANTITY OF SOLVENTS DISPOSED: 1/706/000
QUANTITY LOST 3/126,000
65% OF THE SOLVENT USED DOES NOT GO TO ONE OF THE DISPOSAL
ROUTES,
70,
-------
QUANTITY OF SOLVENT BY DISPOSAL ROUTES
THIS MANY PLANTS ,,,
,,, DISPOSE OF THIS MUCH SOLVENT
,,, BY THIS ROUTE:
BURNING
FLUSHING
LAND FILL
DISPOSAL SERVICE
RECLAIMER
382
2,667
3,674
7,989
4,323
GALLON/MONTH
(x 103)
GALLON/YEAR
(x 103)
8
112
135
547
904
96
1,344
1,620
6,564
10,848
AVERAGE GALLON/YEAR
PER PLANT
251
504
822
2,509
X
I
I—«
to
-------
EXHIBIT VIII-c; THIS TABLE LISTS THE PROJECTED NUMBER
OF PLANTS IN EACH GROUP USING THE VARIOUS DISPOSAL
ROUTES, SOME AREAS ARE LEFT BLANK DUE TO AN INSUFFIC-
IENT SAMPLING TO MAKE A PROJECTION,
72,
-------
DISPOSAL ROUTES FOR SOLVENTS
THIS MANY PLANTS IN THESE S.I.C, GROUPS ,,,
,,, USE THESE DISPOSAL
ROUTES:
BURNING
FLUSHING
LAND FILL
DISPOSAL SERVICE
RECLAIMER
25
METAL
TOTAL FURNITUR
20,320 '399
332
2,667
3,674
7,989
4,323
18
19
114
171
33
PRIMARY
METALS
1,417
54-
156
276
630
319
34
FABRICTD
PRODUCTS
4,804
56
617
.965
1,959
927
35
NON-ELEC
MACHINRY
6,188
132
840
1,319
2,382
870
36
ELECTRIC
EQUIPMNT
3,616
56
533
481
1,366
1,147
37
TRANSPTN
EOUIPKNT
1,702
66
278
281
534
287
38
INSTRKiTS
S CLOCKS
1,590
224
101
867
410
39
KISC
INDUSTRY
604
137
137
192
x
IE
00
Osl
I
O
-------
EXHIBIT VIII-p; THE PERCENT OF PLANTS USING THE
VARIOUS DISPOSAL ROUTES IN THIS TABLE SHOWS THE USE
OF BURNING TO BE FAIRLY CONSTANT AMONG THE INDIVIDUAL
S.I.C, GROUPS, A SMALLER PERCENTAGE OF PLANTS IN
METAL FURNITURE USE FLUSHING AS A MEANS OF DISPOSAL,
WHILE THE USE OF LAND FILL is WIDELY USED IN GROUP 25
AND LITTLE USED IN GROUP 38 / 55% OF THE PLANTS IN
GROUP 38 USE A DISPOSAL SERVICE AND ONLY 14% USE THIS
SERVICE IN GROUP 33, THE LARGEST PERCENT OF RECLAIMING
IS DONE IN GROUP 25,
74,
-------
DISPOSAL ROUTES FOR SOLVENTS
,,. USE THESE DISPOSAL
ROUTES:
BURNING
FLUSHING
LAND FILL
DISPOSAL SERVICE
RECLAIMER
THIS PERCENT OF PLANTS IN THESE
25 33 34
METAL PRIMARY FABRICTD
TOTAL FURNITUR METALS PRODUCTS
20,320 399 1,417 4,804
2 5 4 1
13 5 11 .13
18 29 19 20
39 29 14 41
21 43 23 . 19
S.I.C, GROUPS,,,
35 36 37 38
NON-ELEC ELECTRIC TRANSPTN INSTRMTS
MACHINRY EQUIPMNT EOUIPMHT & CLOCKS
6,138 3,616 1,702 1,590
224-
14 15 16 14
21 13 17 6
38 38 31 55
14 32 .17 26
s
o
39
MISC
INDUSTRY
604
. _
-
23
23
32
m
X
:c
CO
H
-------
EXHIBIT VIII-E; THE NUMBER OF PLANTS USING THE
VARIOUS DISPOSAL ROUTES ACCORDING TO PLANT SIZE
REFLECTS THE CATEGORIES CONTAINING THE MOST PLANTS/
IN TURN, HAVE THE GREATEST NUMBER OF PLANTS USING
THE VARIOUS DISPOSAL ROUTES,
76,
-------
DISPOSAL ROUTES FOR SOLVENTS
,,, USE THESE DISPOSAL
ROUTES:
BURNING
FLUSHING
LAND FILL
DISPOSAL SERVICE
RECLAIMER
THIS MANY PLANTS EMPLOYING ,,,
TOTAL
20,320
382
2,667
3,674
7,989
4,323
20-100
PEOPLE
11,319
182
1,627
2,487
4,207
2,062
100-500
PEOPLE
6,603
128
889
976
2,798
1,666
50+
PEOPLE
2,398
72
151
211
984
595
-------
EXHIBIT VIII-F: IT CAN BE SEEN THAT THE LARGER THE
PLANT/ THE BETTER THE PRACTICE OF SOLVENT DISPOSAL,'
THE USE OF BURNING/ A DISPOSAL SERVICE AND A RECLAIMER
INCREASES WITH PLANT SIZE, WHILE THE PRACTICE OF
USING LAND FILL AND FLUSHING DECREASE AS PLANT SIZE
INCREASES,
78,
-------
DISPOSAL ROUTES FOR SOLVENTS
,,, USE THESE DISPOSAL
ROUTES:
BURNING
FLUSHING
LAND FILL
DISPOSAL SERVICE
RECLAIMER
THIS PERCENT OF PLANTS EMPLOYING,,,
TOTAL
20,320
2
13
18
39
21
20-100
PEOPLE
11,319
2
m
22
37
18
100-500
PEOPLE
6,603
2
13
15
42
25
50+
PEOPLE
2,398
3
6
9
41
. 25
X
00
UD
-------
EXHIBIT VIII-G; THIS TABLE LISTS THE NUMBER OF
PLANTS IN EACH GEOGRAPHIC REGION USING THE VARIOUS
DISPOSAL ROUTES, THE GREATEST NUMBER OF PLANTS
BURNING APPEAR IN THE NORTHEAST REGION, THE NORTH-
EAST AND MID-WEST WITH THE LARGEST NUMBER OF TOTAL
PLANTS HAVE THE LARGEST NUMBER OF PLANTS USING A
DISPOSAL SERVICE OR A RECLAIMER,
80,
-------
DISPOSAL ROUTES FOR SOLVENTS
OF PLANTS LOCATED IN THE ...
,,, USE THESE DISPOSAL
ROUTES:
BURNING
FLUSHING
LAND FILL
DISPOSAL SERVICE
RECLAIMER
TOTAL
20,322
NORTHEAST
6,970
SOUTHEAST MID-WEST SOUTHWEST NORTH CENTRAL FAR WEST
1,878
6,412
745
1,386
2,929
332
2,667
3,674
7,989
4,323
200
927
880
2,837
1,507
36
359
457
444
139
73
. 700
1,263
2,663
1,388
37
170
249
405
178
18
208
364
403
218
18
303
459
1,177
892
oo
CO
-------
EXHIBIT VIII-H: THE PERCENT OF PLANTS GIVE A BETTER
PICTURE OF THE DISPOSAL ROUTES USED IN THE DIFFERENT
GEOGRAPHIC LOCATIONS, THE SOUTHWEST HAS THE GREATEST
PERCENT OF PLANTS USING DISPOSAL ROUTES, RECLAIMING
IS DONE IN THE LARGEST PERCENTAGE OF PLANTS IN THE
FAR WEST AND THE LEAST IN THE SOUTHEAST, THE SOUTH-
EAST ALSO IS LOW ON THE USE OF A DISPOSAL SERVICE,
AND ABOVE AVERAGE ON THE USE OF LAND FILL AND FLUSHING,
THE NORTHEAST AND FAR WEST APPEAR TO USE THE BEST
DISPOSAL PRACTICES,
82,
-------
DISPOSAL ROUTES FOR SOLVENTS
THIS PERCENT OF PLANTS LOCATED IN THE ,,,
,,, USE THESE DISPOSAL
ROUTES:
BURNING
FLUSHING
LAND FILL
DISPOSAL SERVICE
RECLAIMER
TOTAL NORTHEAST SOUTHEAST MID-WEST SOUTHWEST NORTH CENTRAL FAR WEST
20,322 6,970
1,878
715
1,386
2,929
2
13
13
39
21
3
13
13
41
22
2
19
21
21
7
1
11
20
12
22
5
23
33
51
21
1
15
26
29
16
1
10
16
10
30
oo
CO
-------
IX
SURVEY PROCEDURES
THE RESULTS REPORTED IN THE ACCOMPANYING TABLES ARE DERIVED
FROM PERSONAL INTERVIEWS CONDUCTED WITH A RESPONSIBLE PERSON
AT EACH OF 2/578 PLANT SITES ENGAGED IN A MANUFACTURING
ACTIVITY WITHIN'ONE OF THE METALWORKING INDUSTRIES, ALL
INTERVIEWS WERE CONDUCTED BY TELEPHONE, THE FINDINGS MUST
BE INTERPRETED WITHIN THE LIMITS SUGGESTED BY THE FOLLOWING
DETAILS:
THE STUDY POPULATION; MANUFACTURING LOCATIONS WITHIN THE
EIGHT INDUSTRIAL CLASSES COMPRISING THE METALWORKING INDUSTRY/
AT WHICH 20 OR MORE PEOPLE ARE EMPLOYED/ WERE DESIGNATED AS
THE POPULATION FOR STUDY, (THE EIGHT INDUSTRIAL CLASSES ARE
THOSE NAMED IN THE VARIOUS TABLES,) A PRESUMED COMPLETE
LISTING OF THIS POPULATION WAS OBTAINED FROM THE CHILTON
MARKET/PLANT DATA BANK/ CHILTON COMPANY/ RADNOR/ PENNSYLVANIA,
THE LIST NAMED 41/670 MANUFACTURING PLANT SITES, (ALSO IN-
CLUDED IN THE LIST WERE A NUMBER OF NON-MANUFACTURING LOCATIONS
AND DISTRIBUTORS, THESE WERE IGNORED,)
SAMPLING THE POPULATION; THE OBJECTIVE OF SAMPLING WAS TO
OBTAIN AN ADEQUATE REPRESENTATION OF THAT SEGMENT OF THE
METALWORKING INDUSTRY WHICH MAKES USE OF EITHER CHLORINATED
OR PETROLEUM SOLVENTS IN SOME TYPE OF METAL CLEANING OPERATION,
ON THE BASIS OF PRIOR KNOWLEDGE WE GUESSED THAT SOMEWHAT LESS
THAN 50% OF THE STUDY POPULATION MIGHT BE INCLUDED IN THIS
SEGMENT, FOR THE PURPOSES OF THE ANALYSIS WE SOUGHT TO COMPLETE
APPROXIMATELY 1/000 INTERVIEWS WITHIN THE SEGMENT, WE THEREFORE
DETERMINED TO DRAW A SAMPLE OF 2/300 SITES FOR SCREENING,
84,
-------
THE POPULATION LIST WAS FIRST ORDERED IN TERMS OF THE STANDARD
INDUSTRIAL CLASSIFICATION OF EACH PLANT/ AND FOR NUMBER OF
EMPLOYEES WITHIN INDUSTRIAL CLASS, A NEW LIST WAS MADE OF
EVERY 18TH PLANT NAMED IN THE REORDERED POPULATION LIST,
THIS WAS DESIGNATED AS THE "PRIMARY SAMPLE", AT THE SAME
TIME, PARALLEL LISTINGS OF EVERY 19TH AND EVERY 20TH PLANT
WERE CREATED AND WERE DESIGNATED AS "ALTERNATIVE SAMPLE #1
OR #2", BECAUSE OF THE WAY THE ORIGINAL POPULATION LIST HAD
BEEN ORDERED/ A PLANT IN EITHER OF THE ALTERNATIVE SAMPLES
WAS SIMILAR TO THE CORRESPONDING PLANT IN THE PRIMARY SAMPLE
IN BOTH SIZE AND TYPE OF MANUFACTURING ACTIVITY, IF, FOR ANY
REASON/ AN INTERVIEW COULD NOT BE COMPLETED WITH A SPECIFIC
PLANT IN THE PRIMARY SAMPLE/ AN INTERVIEW WAS ATTEMPTED WITH
THE CORRESPONDING PLANT IN ALTERNATIVE SAMPLE #1, IF THIS
FAILED/ ANOTHER ATTEMPT WAS MADE WITH THE CORRESPONDING PLANT
IN ALTERNATIVE SAMPLE #2,
THE CHOICE OF SPOKESMEN FOR THE PLANT SITES: IN PLANTS EMPLOY-
ING BETWEEN 20 AND 100 PEOPLE/ THE SPOKESMAN SOUGHT WAS THE
OWNER OR GENERAL'MANAGER; IN PLANTS EMPLOYING BETWEEN 100 AND
500 PEOPLE/ THE SPOKESMAN SOUGHT WAS THE PLANT MANAGER; AND IN
PLANTS EMPLOYING MORE THAN 500 PEOPLE/ THE SPOKESMAN SOUGHT
WAS THE PRODUCTION SUPERVISOR OR PRODUCTION SUPERINTENDENT,
IN ANY CASE/ IF THE DESIGNATED PERSON SEEMED NOT TO HAVE THE
INFORMATION WE WERE SEEKING (l,E,/ ANSWERED "DON'T KNOW" TO
CERTAIN KEY QUESTIONS)/ HE WAS ASKED TO NAME SOMEONE ELSE WHO
MIGHT BE KNOWLEDGEABLE IN THE AREA OF THE INQUIRY, THIS REQUEST
USUALLY PRODUCED A REFERRAL TO A SUBORDINATE MORE FAMILIAR
WITH THE DETAILS OF THE MANUFACTURING OPERATIONS (AS DISTINGUISHED
FROM THE BUSINESS OPERATIONS) OF THE PLANT,
85,
-------
THE INTERVIEW; A DETAILED INTERVIEWING GUIDE WAS DRAWN UP
BY MEMBERS OF THE NATIONAL MARKETING SURVEYS STAFF IN CO-
OPERATION WITH THOSE MEMBERS OF THE INORGANIC CHEMICALS
DEPARTMENT WHO WERE KNOWLEDGEABLE ABOUT THE OBJECTIVES OF
THIS STUDY, THE INTERVIEWS WERE CONDUCTED BY SKILLED INTER-
VIEWERS OF NATIONAL MARKETING .SURVEYS, WHO HAD BEEN THOROUGHLY
BRIEFED ABOUT THE, INTERVIEWING REQUIREMENTS AND PROCEDURE
FOR THIS STUDY, ALL INTERVIEWS WERE COMPLETED BETWEEN
JANUARY 9 AND FEBRUARY 3/ 1975,
EDITING OF COMPLETED INTERVIEWS; IN THE COURSE OF THE
INTERVIEW THE SPOKESMEN WERE ASKED TO NAME THE SPECIFIC
SOLVENTS USED IN HIS PLANT/ AND TO STATE THE QUANTITIES
USED AND DISPOSED OF BY VARIOUS ROUTES, RESPONSES CON-
CERNING SOLVENTS USED WERE MOST FREQUENTLY GIVEN AS BRAND
NAMES, THESE WERE EDITED TO ONE OF THE FOURTEEN GENERIC
CLASSES (NAMED IN THE TABLES) BY A MEMBER OF THE INORGANIC
CHEMICALS DEPARTMENT, RESPONSES CONCERNING QUANTITIES COULD
BE GIVEN EITHER IN TERMS OF "DRUMS"/ "POUNDS"/ OR "GALLONS",
ALL RESPONSES WERE EDITED TO THEIR EQUIVALENT IN GALLONS,
ALL QUANTITATIVE RESPONSES WERE EDITED TO REASONABLE INTERVALS/
FOR EASE OF KEYPUNCHING AND LATER TABULATING AND ANALYSIS,
TABULATING AND PROJECTING TO POPULATION; THE RAW DATA WERE
KEYPUNCHED AT NATIONAL MARKETING SURVEYS/ AND CAST INTO
SUMMARY TABLES AT THE COMPUTATIONS RESEARCH LABORATORY ACCORD-
ING TO SPECIFICATIONS DRAWN UP BY THE NATIONAL MARKETING
SURVEYS STAFF,
86,
-------
IN PROJECTING THE SAMPLE TO THE POPULATION/ EACH INDUSTRIAL
CLASS WAS TREATED SEPARATELY, A "MULTIPLIER" FOR EACH
INTERVIEW AS DETERMINED BY DIVIDING THE POPULATION SIZE
BY THE NUMBER OF INTERVIEWS COMPLETED WITHIN THAT POPU-
LATION, NOT SURPRISINGLY/ THE MULTIPLIER IN MOST CASES
WAS APPROXIMATELY 18, HOWEVER, FOR MOST QUESTIONS WITHIN
THE QUESTIONNAIRE THE PLANT'S SPOKESMAN WAS ALLOWED THE
OPTION OF REPLYING "I DON'T KNOW", IN SUCH CASES HE WAS
ASKED TO PROVIDE HIS "BEST ESTIMATE"/ BUT IN SOME CASES
THE SPOKESMAN WAS UNABLE TO ESTIMATE OR HAD NO BASIS FOR
AN ESTIMATE, THESE "DON'T KNOWS" WERE DROPPED FROM THE
NUMBER OF COMPLETED INTERVIEWS IN THE TREATMENT OF THE
QUESTIONS INVOLVED, IN EFFECT/ THIS PROCEDURE IS IDENTICAL
TO MAKING THE ASSUMPTION THAT WHAT IS. NOT KNOWN BY THOSE
WHO DON'T KNOW is THE SAME AS WHAT Ji KNOWN BY THOSE WHO
DO KNOW,
THIS STUDY WAS DESIGNATED AS PROJECT #221 IN NATIONAL
MARKETING SURVEYS/ AND ALL MATERIALS RELEVANT TO ITS
EXECUTION ARE MAINTAINED UNDER THAT FILE NUMBER,
87,
-------
-o
m
Z
D
00
-------
APPENDIX B
STUDY TO SUPPORT NEW SOURCE PERFORMANCE
STANDARDS FOR SOLVENT METAL CLEANING OPERATIONS
Review of Analytical Methods and Techniques
Prepared By:
John Woods
Richard Melcher
The Dow Chemical Company
Prepared For:
Emission Standards and Engineering Division
Office of Air Quality Planning
U.S. Environmental Protection Agency
-------
THEVIEW OF ANALYTICAL METHODS AND TECHNIQUES - SUB TASK 2 OF
PROJECT PLAN FOR EPA CONTRACT NO 68-02-1329
INFORMATIVE SUMMARY WITH CONCLUSIONS
An EPA contract to conduct a "Study to Support New Source
Performance Standards for Industrial Degreasing Operations"
has been awarded The Dow Chemical Company. As part of this
contract a review of analytical methods and techniques for
sampling and analyzing industrial degreasing solvents emissions
from stationary sources has been conducted. Literature sources
include EPA methodology, ASTM, ISC, instrument manufacturers
literature, books and journals.
This report describes how a sampling and measurement task
should be undertaken to obtain the best information possible
from current public technology. These steps for an analytical
chemist are:
1. Obtain preliminary information
2. Choose analytical technique and sampling methodology
3. Prepare and implement choices
4. Conduct flowrate measurements
5. Complete the analysis and suitable report.
-------
Review of Analytical Methods and Techniques - Sub Task 2 of EPA
Contract "Study to Support New Source Performance Standards for
Industrial Degreasing Operations."
I. Preliminary Information
In a pre-sampling visit to a customer the analytical chemist
should gather much information. A general knowledge of the process
or operation, the variety of solvents used, the physical arrangement
of the venting systems and control systems, cost factors, local
legislative requirements, and the nature of the data required by
the customer will guide the choice of sampling and analytical
techniques to be employed. The ultimate choice of techniques may
be a compromise resulting from the above factors plus the expertise
of the analytical chemist. Three approaches are available:
1. Analytical technique for specific compounds
2. Analytical technique for general organics
3. Techniques when the analytical facilities are remote from
the sampling site.
However, flowrate measurement techniques and compatable sampling
methodology must accompany any analytical approach.
II. Flow Rate Measurements
The emission rate of a compound from a stationary source is the
product of the volumetric flow rate and the concentration (by volume)
If stationary sources are to include exhaust fans and hood vents
with blowers, flowrate measurements and data should also be included.
Flow Rate Measurement References
"Determination of Stack Gas Velocity and Volumetric Flow ..-Rate"
Method 2 Federal Register 36_ (247) 24884 Dec. 23, 1971.
Average Velocity in a Duct (Pitot Tube Method) D3154 ASTM Annual
Standards Part 26 pp 642-652 1974.
Average Velocity in a Duct (Thermal Anemometer Method) Method
being drafted in ASTM Committee D-22.
-------
Rotorceter Calibration D3195
ASTM Annual Standards Part 26 pp 690-693, 1974
Air Sampling Instruments
ACGIH 4th Ed. 1972 Section B
III Sampling Methodology
For on-site analysis the analyzer can be piped up on a serai-
permanent basis, samples can be brought to the analyzer, or the
analyzer can be taken to the sampling point (SAFETY). If a
minimum number of sampling points are needed and can be monitored
individually for a period of time, a sample transport line can
be installed between the sampling point and the analyzer. If a
large number of sampling points are to be examined simultaneously
or in a timed sequence, instantaneous or "grab" samples can be
taken to the analyzer. If the sampling point is in a convenient
and non-hazardous location, the analyzer could be used at the
sampling point without long sample lines and many sample "grabbers".
For reiiiGLe i-ampling, see Section VI.
Sampling Methodology References
Stern, A.C., "Air Pollution" Vol. II 2nd Ed. Academic Press,
New York, N.Y., 1968, Chapter 16, p. 3-54.
Sampling Atmospheres for Analysis of Gases and Vapors D1605 ASTM
Annual Standards Part 26 pp 285-306, 1974.
Sampling and Storage of Gases and Vapors, Methods of Air
Sampling and Analysis - Intersociety Committee APHA Part 1, pp 43-55
1972.
Air Sampling Instruments, ACGIH 4th Ed. Section A, 1972.
Ledbetter, J., "Air Pollution Analysis" Part A, Marcel Dekker Inc.
New York, N.Y. 1972 Chapter 6, page 195.
-------
IV» Analytical Technique for Specific Compounds
An instrumental analytical technique for specific compounds
would be an easily movable gas chromatograph with a flame ionization
detector (GC-FID), temperature programming capability and/or
switching valves with a variety of analytical columns for the
various groups of compounds. This analysis could be done on-site
in an approved (safe) atmosphere.
This approach would probably be of "high" initial investment
incorporating a preventive maintainance program and some initial
selection of appropriate analytical columns along with a compilation
of gas chromatographic operating conditions and retention times
for compounds on various, columns.
Gas Chromatography References
Gas Chromatograph Terms and Relationships E 355
ASTM Annual Standards Part 42 1974
General Gas Chromatography Procedure E 260 Ibid
Stern, A.C., "Air Pollution" Vol. II 2nd Ed.
Academic Fzess, New York, N.Y., 196t Chap. 18 pp 120-145
Leithe, W., "The Analysis of Air Pollutants"
Ann Arbor-Humphrey, London, 1970 Chap. 4 pp 78-95
V. Analytical Technique for General Organic
Various detectors can be employed without sophisticated sample
t
preparation to present a "total" measure of organic vapors or gases.
These physical methods of detection include electric conductivity,
coulometry, ionization, thermal conductivity, combustion, UV and
IR spectroscopy, chemiluminescence, and flame photometry. As a
detector alone they are not specific but can be calibrated for a
single vapor or a known mixture. Many of these detectors are
incorporated as "direct reading" instruments.
General Organic References
Total Hydrocarbons by GC-FID ISC 43101-02-71T
Methods of Air Sampling and Analysis - Intersociety Committee
APHA, Washington, D.C., 1972
-------
"Flame lonization Hydrocarbon Analyzer", R.A. Morris and R.L. Chapman;
JAPCA 11 (10) 467-9 Oct. 1961
"Continuous Trace Hydrocarbon Analysis by Flame lonization", A. J.
Andreatch and R. Feinland; Anal. Chem. 32: (8) 1021-4 July, 1960
Stern, A. C., "Air Pollution," Vol II 2nd Ed.
Academic Press, New York, N.Y., 1968 Chap. 18 p 118.
Air Sampling Instruments
ACGIH, 4th Ed. 1972 Section U.
-------
VI Remote Sampling
Because of some of the disadvantages of the portable gas
chromatograph, such as high initial investment and the increased
maintainance and training necessary, a second approach for the
determination of specific compounds may be necessary. The remote
sampling approach (i.e., taking a sample and transporting to
a laboratory for specific quantitative analysis of the target
compounds) may be a more practical approach for a wide range of
solvent mixtures.
A disadvantage of the remote sampling approach would be the
time lag between sampling and obtaining the finished results.
There are, however, a number of advantages which would make this
the preferred approach.
A number of different techniques have been used for taking
remote samples. These are 1) whole air sample, 2) use of absorber
solvents, and 3) solid adsorbents.
Whole air Samples. Whole air samples are taken by pulling the
test air into an evacuated glass bulb, can, tube, syringe, etc.,
or by pumping the air into a container or plastic bag. Although
these techniques may work well for many solvents when Lhe samples
can be analyzed the same day, such samples are more difficult to
transport, may show loss of the compound due to adsorption,
condensation, or diffusion over long periods and do not offer
component concentration and therefore limit sensitivity.
Solvent Absorber. A widely used technique for the collection
of organic vapor from air is the solvent absorber (Kl, >K2). The
difficulty with this system is to find an absorber solvent which is
compatable with all the compounds to be collected, low enough
volatility so that it will not be lost during sampling, and it
will not interfere with gas chromatographic analysis. One of the
main disadvantages of this system is that many of the solvents
used for collection cannot be mailed or transported by common
carriers.
Solid Adsorbents. A number of solid adsorbents have been used
for collection and concentration of organics. These adsorbents
may be coated or non-coated and desorbed in the laboratory by
heat or a suitable solvent. The advantages of solid adsorbents are:
-------
1. Samples can be taken with existing equipment such as a small
laboratory pump or portable battery-operated pump.
2. A minimum amount of time would be spent on-site by using the
rapid sampling technique. Simultaneous samples can be taken
with two or more pumps.
3. The samples could be transported easily or mailed back to the
laboratory while the investigator continues with his sampling
program.
4. A complete analysis can be made on the returned samples using
the more versatile, sophisticated laboratory instrumentation.
5. The tubes containing the adsorbent can be prepared in advance
and a relatively large supply can be carried with the investi-
gator when traveling.
This technique generally involves pulling the test air through
a small tube containing the adsorbent thereby collecting and
concentrating. The tubes are then returned to the laboratory and
the collected components are desorbed using either thermal or
solvent techniques.
THERMAL DESORPTION OF SOLID ADSORBENTS
Thermal desorption necessitates the collection of organic
vapors on a solid adsorbent contained in a rigid tube i.e.
stainless steel. The tube is heated and back-flushed with carrier
gas directly into a gas chromatograph. This technique offers
i
very high sensitivity since the entire collected sample is injected
and is advantagous for repetitive analysis of low concentrations.
Difficulty may arise when a mixture of components especially with
a wide volatility range is sampled. The adsorbent must be able
to collect efficiently all vapors and efficiently desorb with heat.
Special equipment may be necessary to interface the collection
tube to the gas chromatograph. The Bendix HS-10 Flasher is a
commercially available unit that utilizes this technique.
SOLVENT DESORPTION OF SOLID ADSORBENTS
When samples are analyzed by solvent desorption, the adsorbent
is removed from the tube and shaken with the solvent. The extract
can then be analyzed. The advantages of solvent desorption are:
-------
1. Repeated analysis can be made on the sample using different
chromatographic conditions and columns.
2. Other techniques such as infrared can be used on the same
sample.
3. No special additional equipment is necessary to analyze the
samples by gas chromatography.
4. The desorbed samples can be set up with available automatic
injection systems which can run a large number of samples
with a minimum of attention.
5. The desorbed sample can be saved for a period of time in case
additional analysis is required.
TYPES OF SOLID ADSORBENTS
The most widely used adsorbents are silica gel and charcoal,
however, many other adsorbents have been reported for specific
sampling problems.
SILICA GEL (References Gl - G14)
Good collection and recovery for a large number of organic
vapors has been reported. Many variations as to the size of the
collection tube, mesh size and type of silica gel, collection
volumes and desorption solvents have been used. Table II in a
publication by Feldstein et al. (G6) (Copy attached) indicates
the collection and recoveries of a number of organic vapors.
The adsorption of water by silica gel is a serious problem in
that high concentration of water deactivates silica gel and reduces
the collection efficiency. However, with the proper balance of
sampling rate, time and sample tube size, good recoveries may he
obtained.
In addition to the desorption solvents described by Feldstein,
ethanol, methanol, acetone, diethylether, and combinations of these
solvents have been used with success.
The main difficulty arises in selecting a solvent which will
give high desorption for all the organics collected and still not
interfere with the analysis. The most troublesome class of com-
pounds is the volatile, polar compounds. Methanol can be used to
-------
desorb most organic solvents efficiently, but then methanol itself
cannot be determined. Feldstein has suggested dimethylsulfoxide
as a solvent for polar compounds but no data is given for methanol.
CHARCOAL (References Hi— HlO)
A review written by Mueller and Miller (H 10) effectively
summarizes the use of charcoal for sampling organic vapors in air.
A copy of this report is attached. The technique employed is
based on work of Kupel and coworkers at the National Institute
for Occupational Safety and Health (NIOSH) (H 6).
Good collection and recovery has been obtained for many
organic vapors, however, several important facts must be considered.
1. Most of the testing has been done at concentrations relative
to the recommended OSHA limits. The concentrations found in
source emissions may be many factors higher and the use of
larger tubes and smaller sample volumes may be necessary.
For example, methylene chloride was found to break through the
small NIOSH tube at 500 ppm when collected at the recommended
rate of one liter per minute for ten minutes.
2. Carbon disulfide does not readily displace all organic
compounds from carbon. The total recovery for volatile polar
compounds is generally low, for example; methanol 10%, ethanol
67%, butanol 75%, and diacetone alcohol 60%.
ACTIVATED ALUMINA
Although there may be some difficulties, either silica gel or
charcoal could be used to collect the solvents of interest except
perhaps the volatile, polar compounds. Activated alumina has been
suggested as a general adsorbent in some references, but no
literature has been found which describes its practical use for
ambient air testing. Recently it has been shown (M 26) that
activated alumina will efficiently collect volatile polar organics
and that the organics can be desorbed with water with high recovery.
The water extract is then injected directly into a gas chromatograph
for analysis.
-------
VII DISCUSSION FOR REMOTE SAMPLING
From the preceding information, it appears that by using an
adsorption tube packed with silica gel, charcoal, activated
alumina, or other suitable adsorbent, or a combination of the
tubes, all the organic solvents of interest can be collected and
recovered with good efficiency. Gas chromatography can be used
for analyzing mixtures of the solvents. Stoddard-type solvents
which are composed of a mixture of hydrocarbons within a specified
boiling range, present a special problem. This type of solvent
can be determined by gas chromatography by selecting the strongest
peaks of the mixture for calculation or by integrating the total
area for the boiling point range. This type of solvent could
also be determined by infrared by measuring absorption at 3.4
microns. In either case a standard of the actual solvent being
used would be necessary to obtain satisfactory results. Inter-
ference corrections for other solvents present would be necessary.
It is difficult at this point to recommend one specific
collection and analysis method for all compounds of interest. In
practice, a diverse but smaller number of compounds would be
present at any specific degreasing operation. Before actual
sampling is undertaken the possible compounds at that location
should be determined from customers application, and from this
information the proper type of sampling tubes could be selected.
This information should also be supplied to the laboratory so
that the proper analytical conditions can be selected and calibrated,
VIII SUMMARY FOR REMOTE SAMPLING
Recommended steps for designing a system for remote collection
and subsequent analysis.
1. The types of solvents and mixtures used in various degreasing
operations should be determined.
2. The concentration range of interest should be suggested for
each solvent.
3. A recommended sampling tube system could be determined from
the above list of solvent mixtures and concentrations, and
from the recovery data in the literature.
-------
The retention times for the solvents of interest should be
compiled for a number of suggested gas chromatographic columns
so that the proper analytical condition can be determined for
all combinations and mixtures. See References J 1 - J 4.
Collection and recovery data should be determined in the
laboratory for the expected mixtures and concentration
ranges.
This study should also include time storage data.
-------
SILICA GEL REFERENCES
G-l.
A. P. Altshuller, T.A. Bellar, and C.A. demons
Concentration of Hydrocarbon on Silica Gel Prior to .Gas Chromatographic
Analysis. Am. Ind. Hyg. Assoc. J. 23, Apr. 1962. pp. 164-6.
G-2
E.E. Campbell and H.M. Ide
AIR SAMPLING AND ANALYSIS WITH MICROCOLUMNS OF SILICA GEL.
Am. Ind. Hyg. Assoc. J. 21, (4) 323-31, Aug. 1966.
G-3
S. Horiguchi, K. Shinagawa, T. Utsunomiya, K. lyoda, N. Tanaka
ATMOSPHERIC DETERMINATIONS OF CHLORINATED HYDROCARBONS, ESPECIALLY
THE DETERMINATION OF 1,1,1-TRICHLOROETHANE. Text in Japanese.
Japan J. Ind. Health (Tokyo), 7 (5):25-28, May 1965. 5 refs.
G-4
N.E. Whitman and A.E. Johnston
SAMPLING AND ANALYSIS OF AROMATIC HYDROCARBON VAPORS IN AIR: A
GAS-LIQUID CHROMATOGRAPHIC METHOD. Am. Ind. Hyg. Assoc. J. 25 (5)
464-9 Oct. 1964.
G-5
A.Y. Ping, L.R. Clayton, T.E. McEwen, and J.S. Paydo
THE APPLICATION OF SILICA GEL IN SOURCE TESTING. PART I: COLLECTION
OF SAMPLES. Preprint. (Presented at the 59th Annual Meeting, Air
Pollution Control Association, San Francisco, Calif., June 20-25,
1966, Paper No. 66-79.)
G-6
Feldstein, M., S. Balestrieri, and D. A. Levaggi
THE USE OF SILICA GEL IN SOURCE TESTING. Am. Ind. Hyg. Assoc. J.,
28 (4):381-385, July-Aug. 1967. 9 refs.
-------
SILICA GEL REFERENCES
G-7
I. G. Kachmar
DETERMINATION OF BENZENE TOLUENE AND XYLENE SIMULTANEOUSLY PRESENT
IN THE AIR. U.S.S.R. Literature on Air Pollution and Related
Occupational Dieseases, Vol. 7, 143-7 1962. (Digiena i Sanit.)
258 (5) 58-62, 1960. Translated from Russian.
G-8
H. Buchwald
ACTIVATED SILICA GEL AS AN ADSORBENT FOR ATMOSPHERIC CONTAMINANTS.
Occupational Health Rev. (Ottawa) Vol. 17 (4):14-18, 1965.
G-9
J.E. Peterson, H.R. Hoyle and E.J. Schneider, HALOGENATED HYDROCARBON
CONTAMINANTS BY MEANS OF ABSORPTION ON SILICA GEL Amer. Ind. Hyg.
Assoc. Quarterly 17:4, 429-433 (Dec. 1965).
G-10
R.J. Sherwood. "THE MONiri'OwTNG OF BF-N7-ENE EXPOSURE BY a.IR SAMPLING".
Amer. Ind. Hyg. Assoc. J. (Dec. 1971.)
G-ll
J.H.C. Van Mourik, "EXPERIENCES WITH SILICA GEL AS ADSORBENT",
Amer. Ind. Hyg. Assoc. J. Sept. Oct. 1965.
G-12
Elkins, H.B., L.D. Pagnotto, and E.M. Comprone: THE ULTRAVIOLET
SPECTROPHOTOMETRIC DETERMINATION OF BENZENE IN AIR SAMPLES ADSORBED
ON SILICA GEL. Anal. Chem. 13: 1797 (1962).
G-13
Maffet, P.A., Doherty, T.F. and Monkmann: COLLECTION AND DETERMINATION
OF MICRO AMOUNTS OF BENZENE OR TOLUENE IN AIR. Amer. Ind. Hyg.
Quart. 17, 186 (1956).
G-14
Erley, D.S. INFRARED ANALYSIS OF AIR CONTAMINANTS TRAPPED ON SILICA
GEL. Amer. Ind. Hyg. Assoc. J. 23:388 (1962).
-------
CHARCOAL REFERENCES
H-l
C.L. Rraust, E.R. Hermann
CHARCOAL SAMPLING TUBES FOR ORGANIC VAPOR ANALYSIS BY GAS
CHROMATOGRAPHY. Am. Ind. Hyg. Assoc. J., 27 (l):68-74, Feb. 1966.
H-2
THE COLLECTION OF TRICHLOROETHYLENE AND PERCHLORETHYLENE VAPORS.
\
La Captation des Vapeurs de Tri et de per. (Galvano (Paris))
36, (361 1)2, l]5-6, 152, Feb. 1967, Fr.
H-3
S.B. Smith, and R.J. Grant
NON-SELECTIVE COLLECTOR FOR SAMPLING GASEOUS AIR POLLUTANTS FINAL
REPT.) Pittsburgh Coke and Chemical Co., Research and Development
Dept. Dec. 15, 1958. 63 pp
H-4
Y. Natsunvura
THE ADSORPTION PROPERTIES OF ACTIVE CARBON. II, PRELIMINARY STUDY
ON ADSORPTION OF VARIOUS ORGANIC VAPORS ON ACTIVE CARBON BY GAS
CHROMATOGRAPHY. Ind. Health (Japan) 3, 121-5, Dec. 1965.
H-5
Reid, Frank H. and Walter R. Halpin
DETERMINATION OF HALOGENATED AND AROMATIC HYDROCARBONS IN AIR BY
CHARCOAL TUBE AND GAS CHROMATOGRAPHY. Am. Ind. Hyg. Assoc. J.,
29 (4):390-396, July-Aug. 1968.
H-6
L.D. White, D.G. Taylor, P.A. Mauer and R.E. Kupel,
A CONVENIENT OPTIMIZED METHOD FOR THE ANALYSIS OF SELECTED SOLVENT
VAPORS IN THE INDUSTRIAL ATMOSPHERE. Amer. Ind. Hyg. Assoc. J., 31
(2) 225-(.1970)
-------
CHARCOAL REFERENCES
H-7
C. V. Cooper, L.D. White and R. E. Kupel, QUALITATIVE DETECTION
LIMITS FOR SPECIFIC COMPOUNDS UTILIZING GAS CHROMATOGRAPHIC
FRACTIONS ACTIVATED CHARCOAL AND A MASS SPECTROMETER, Amer., Ind.
Hyg. Assoc. 32 383-(1971).
H-8
Otterson, E.J., and C.U. Guy: A METHOD OF ATMOSPHERIC SOLVENT VAPOR
SAMPLING ON ACTIVATED CHARCOAL IN CONNECTION WITH GAS CHROMATOGRAPHY.
TRANCATIONS OF THE TWENTY-SIXTH ANNUAL MEETING OF THE AMERICAN
CONFERENCE OF GOVERNMENTAL INDUSTRIAL HYGIENISTS, Philadelphia, Pa.,
Page 37, American Conference of Governmental Industrial Hygienists,
Cincinnati, Ohio (.1964) .
H-9
NIOSH SAMPLING DATA SHEET 6 (June 16, 1972).
H-10
F.X. Mueller and James A. Miller DETERMINATION OF ORGANIC VAPORS
IN INDUSTRIAL ATMOSPHERES. American Laboratory, May 1974.
H-ll
Turk, A. and C.J. D'Angio, COMPOSITION OF NATURAL FRESH AIR.
J. Air Pollution Control Assoc. 12:29 (.1962).
H-12
Turk, A,, J.I. Morrow and B.E. Kaplan, OLEFIN ISOMERIZATION IN
ADSORPTIVE SAMPLING ON ACTIVATED CARBON. Anal. Chem. 34:561 (1962).
H-13
Health Service & Mental Health Administration contract HSM 99-72-98:
Collaborative Testing of Activated Charcoal Sampling Tubes for
Seven Organic Solvents, Scott Research Laboratories, Inc. SRL 1316
10 0973, 1973.
-------
THERMAL DESORPTION
1-1
Urone, P., and J.E. Smith: ANALYSIS OF CHLORINATED HYDROCARBONS
WITH THE GAS CHROMATOGRAPH. American Ind. Hyg. Assoc. J.,
22:36 (1961).
1-2
West, P.W., B. Sen and N.A. Gibson: GAS-LIQUID CHROMATOGRAPHIC
ANALYSIS APPLIED TO AIR POLLUTION SAMPLING. Anal. Chem. 30:1390
(1958).
1-3
Altshuller, A.P., T.A. Bellar, and C.A. demons: CONCENTRATION OF
HYDROCARBONS ON SILICA GEL PRIOR TO GAS CHROMATOGRAPHIC ANALYSIS
Amer. Ind. Hyg. Assoc. J. 25:646 (1964)
1-4
T.A. Bellar, M.F. Brown, J.E.. Sigsby, Jr.: DETERMINATION OF ATMOS_
PHERIC POLLUTANTS IN THE PART-PER-BILLION RANGE BY GAS CHROMATOGRAPHY.
Anal. Chem. 19.24 (1963) .
1-5
Cropper, F.R. and S. Kaminsky, DETERMINATION OF TOXIC ORGANIC
COMPOUNDS IN ADMIXTURE IN THE ATMOSPHERE BY GAS CHROMATOGRAPHY.
Anal. Chem. 35, 735 U963).
1-6
Novak, J., V. Vasak, and J. Janak: CHROMATOGRAPHIC METHOD FOR THE
CONCENTRATION OF TRACE IMPURITIES IN THE ATMOSPHERE AND OTHER GASES.
Anal. Chem. 37 660 (1965).
Ir.7
Shadoff, L.A., G.J. Kallos, and J.S. Woods: Determination of
Bischloromethyl Ether in Air by Gas Chromatography-Mass Spectro-
metry. Anal. Chem. 45 2341-44 (Dec. 1973).
-------
GAS CHROMATOGRAPHIC ANALYSIS
J-l
Rushing, D.E., GAS CHROMATOGRAPHY IN INDUSTRIAL HYGIENE AND AIR
POLLUTION PROBLEMS. Amer. Ind. Hyg. Assoc. J. 19:238 (1958).
3-2
Deans, D.R. and Scott, I., GAS CHROMATOGRAPHIC COLUMNS WITH ADJUST--)
ABLE SEPARATION CHARACTERISTICS, Anal. Chem. 45 (7) 1137 (1973).
J-3
Sugar, J.W. and Conway, R.A., GAS LIQUID CHROMATOGRAPHIC TECHNIQUES
FOR PETROCHEMICAL WASTE WATER ANALYSIS, Journal of the Water Pollution
Control Federation, JWPFA, Vol, 40, 1968, pp 1622 to 1631.
J-4
ASTMD 2908-70T, TENTATIVE RECOMMENDED PRACTICE FOR MEASURING VOLATILE
ORGANIC MATTER IN WATER BY AQUEOUS-INJECTION GAS CHROMATOGRAPHY.
-------
LIQUID ADSORBERS REFERENCES
K-l
Mansur, R.H., R.F. Pero, and L.A. Krause, VAPOR PHASE CHROMATOGRAPI1Y
IN QUANTITATIVE DETERMINATION OF AIR SAMPLES COLLECTED IN THE FIELD.
Amer. Ind. Hyg. Assoc. J. 20:175 (1959).
K-2
Levadie, B., and J.F. Harwood, AN APPLICATION OF GAS CHROMATOGRAPHY TO
ANALYSIS OF SOLVENT VAPORS IN INDUSTRIAL AIR. Amer. Ind. Hyg. Assoc.
J. 21:20-.(1960),.
MOLECULAR SIEVE REFERENCES
L-l
F. Gustafson and S.H. Smith, Jr. REMOVAL OF ORGANIC CONTAMINANTS
FROM AIR BY TYPE 13X MOLECULAR SIEVE. Naval Research Lab, Washington,
D.C. (NKL Kept. 5560} Dec. 6, 136C 2T pp
ACTIVATED ALUMINA REFERENCES
M-26
Internal Dow Method in Process for Release.
-------
Application Bulletin 801
MODEL 800-VTP AIRFLOW MULTIMETER
A basic tool to help you meet the
"Occupational Safety and Health Standards"
(OSHA)
MEASURES
GAS FLOW
PARAMETERS:
• Velocity
« Ambient
Temperature
e Static
Pressure
a Mass Flow
~\^' ••' ~ " '••• '. •
w»./.'j/:.'TjeA -^ .^-^••.-
This Application Bulletin illustrates simple methods to measure a number of Ventilation Require-
ments in accordance with Subpart G, Section 1910 of the Rules and Regulations of the Occupa-
tional Safety and Health Standards.
(FIRST PAGE ONLY)
-------
©1973, ISA AID T3l»2l» (113-120)
SAMPLING SYSTEM GUIDELINES
John A. Chapman
Director of Marketing
ARCAS
Houston, Texas
Glen D. Payne
Chief Engineer
ARCAS
Houston, Texas
ABSTRACT
This paper covers basic design factors for typical
sampling problems. Guidelines are given to
provide the Instrument engineer with a checklist
of Items to consider when designing sample
handling systems. The guidelines are also useful
to aid communication between Instrument engineer
and process engineer. Data Is presented on
system time lags and flow rates required for
given lag times with various size sample lines.
Several examples are given showing proper and
Improper Installations.
INTRODUCTION
Much has been written on the subject of sampling.
It's very Interesting to read some of the earlier
papers and see the progression of sampling ideas
and methods. Recently, there have been several
excellent papers given that cover specific
sampling problems. This paper Is Intended to
give general guidelines for approaching sampling
problems.
STATEMENTS AND DEFINITIONS
First, let us cover some basic terms and state
the problem a little more clearly. Analyzers
and sampling systems are not In themselves
Important. Analyzers,and sampling systems are
only tools that make the analysis possible. The
analysis cannot be accomplished without both a
good analyzer and a good sampling system. The
purpose of the sampling system is not to deliver
to the analyzer a sample of the exact composition
as that In the process line. The purpose Is to
deliver a representative sample suitable for the
analyzer In which the variables to be measured
vary according to the stream composition.
The sampling system and ther analyzer should be
complimentary of each other to accomplish the
required analysis and not make Impractical demands
on either because of poor design in the other.
All too often Impractical and unnecessary demands
are made of on-stream analyzer systems. Don't
lower your necessary requirements but, "If you
don't need It, don't do it". This Is the basic
philosophy of design to use throughout the analysis
system.
Engineering neglect contributes to the problems
of sampling. The specifications for analyzers
are generally very detailed but the sampling
system frequently receives only a casual mention.
A sampling system doesn't have to be high priced
to be good, but it should be adequate to do the
required job. This point will be shown throughout
the discussion.
The requirements of an on-stream analysis (which
Involves both the analyzer and the sampi'ng
system) Is to analyze the process stream:
I. as completely as required,
2. as fast as required,
3. as accurately as required,
4. and present the analysis in a desired
and usable form.
Since the analyzer is so closely tied in with
the sampling system, the analyzer location is
very important. The proper location of the
analyzer can sometimes lessen or even eliminate
some of the difficult sampling problems and
requirements. Some guidelines for locating the
analyzer:
I. Locate as near as practical to the
samp I ing point.
2. Locate where It will be convenient for
servicing.
3. Locate where the required utilities
are aval(able.
4. Do not block accessibility to other
equipment.
5. Provide adequate shelter from the
elements.
6. Avoid locating In a hazardous working
area or near hazardous equipment.
BASIC FUNCTIONS OF A SAMPLING SYSTEM
In view of the previous comments, now let us
consider the sampling system portion of the
analysis system. There are four basic functions
required of the sampling system.
I. Obtain a representative sample from
the process.
2. Condition the sample to make It suitable
for the analyzer without modifying its
basic characteristics.
FIRST PAGE ONLY
113
-------
SON ANALYZED
© Gun-Type Analyzer
Section
© Flame-Out Alarm.
© Minimum Detectable
of 0.1 ppm.
© Recorder Output.
© Digital Readout
in ppm.
Completely
(self-contained).
Heated Sample Probe and Detector.
Adjustable Concentration Alarm System.
The AID Model 550 is a completely portable HEATED Total Hydrocarbon Analyzer which deter-
mines the concentration of organic vapors in an air sample. An integral digital readout allows
the operator to directly determine the concentration of the organic compounds. A range
switch is used to select the readout to either 0-200 ppm or 0-2000 ppm full scale, with sensitivity
to 0.1 ppm in the 0-200 ppm scale. Condensation of vapors is avoided on the Model 550 by
having both the sampling probe and detector heated to above 70°C. An integral adjustable
alarm system allows the operator to preselect a concentration which will activate an audible
alarm system when the level is reached. In addition, a FLAME-OUT audible alarm is also built
into the instrument.
For operator ease and comfort the Model 550 is designed in two sections. The first or analyzer
section which is hand held consists of the heated probe and detector, supply batteries, ignitor
and a digital readout. This is connected to the second or supply section by means of an
umbilican cord which is used to supply power for heating purposes, vacuum for sampling and
hydrogen for detector operation. The supply section contains the refillable hydrogen cylinder,
pressure gauges and regulators, sample pump and rechargeable Nickel-Cadmium batteries.
In addition, a jack is provided so that the signal may record continuously on a 2 millivolt
recorder. There is sufficient power and hydrogen in the supply section to operate the instru-
ment in the portable mode for at least 8 hours. The instrument may be operated indefinitely
using A.C. line and an external hydrogen supply.
FIRST PAGE ONLY
-------
THE Gas Tech HALIDE DETECTOR
t'"'-", 'j}-^:->'f^»rrr-; v;-^V;;C'" -'j_l"• ' ' ''"."•' :'v? "V^'p?
i. »-1" -•* •jrfrjr^f^^Ti* • '^f*V ;*"' "•""<•*•• ' •'^iL' "X^ '*-*"-'--'••' ' y' •'• Jfc ' * ^'"•'."• ' ' •» "•''.> "'-1." 'i**M ' '--v
• -. v*.^^^,^^^^^^^,^^2-lI^ll!^^t^***M^v-L>;^^ ' 'Tif 'jX.jSf «*?'iii '---'V £' '^ *'•-'«• '*'ift *•'**«
•-Y«3
A new portable instrument for detection and meas-
urement of all airborne halogen compounds. Light
weight and small size are attained by careful design
of detection and flow systems, and by use of solid-
state circuitry throughout.
DETECTION PRINCIPLE
The instrument depends upon the long-recognized
fact that the ultraviolet spectral content of an elec-
trical spark is increased by the presence of halo-
genated compounds, and that the degree of spec-
tral enhancement is related to the concentration of
halogens in the atmosphere surrounding the spark.
DESCRIPTION
A portable metal case with carrying handle has all
the normal operating controls and indicators dis-
played on the front panel, and the 11 SAC power
line and 10'sample tubeextend from the rear. Halide
concentration is read out on a meter, graduated
0 -100, and interpreted in terms of ppm of a specific
compound by use of calibration curves.
A built-in diaphragm pump draws sample through
the sample line and delivers it to the spark chamber,
where the spark intensity is monitored by a photo-
cell/UV filter combination. Cell output is amplified
by a FET-input solid-state amplifier, and spark con-
tinuity is assured by a spark sensing and restarting
circuit.
Recorder terminals are provided on rear of case, so
that a continuous record can be made of variations
in halide concentration. Zero stability is adequate
for long term as well as short term monitoring.
APPLICATION
The instrument will give a useful reading over most
practical ranges of halogenated hydrocarbons, as
used in research and industry. Threshold limit con-
centrations of carbon tetrachloride, trichloroethyl-
ene, chloroform and most other halogen-containing
compounds are readily determinable on the meter,
yet expanded scales also permit readings up to
10,000 ppm. Response is a function of the type and
number of halogen atoms, with greatest sensitivity
obtainable on chlorine and fluorine compounds.
Primary field of application is by industrial hygien-
ists in industrial, laboratory and government facili-
ties. It will also be useful as a process monitor and
as a leak detector in many industries where halo-
genated compounds are used; such as solvent de-
greasing, dry cleaning, painting, refrigeration and
aerosol container plants.
Argus Supply Company
Industrie S«l«ty 4 Fir* Equipment
5340 CASMERE AVENUE
DETROIT, MICHIGAN 48212
(313) 892-8350
-------
PPIICATION
INSTRUMENT SENSITIVITY
MINIMUM DETECTABLE
GCAN 131
AID's Portable Gas Chromatographs have proven to be an effective
tool in the Environmental Health and Air Pollution areas. Be-
cause of the varied types of analysis needed to be performed in
these areas, the following table was prepared. Compounds are
listed according to class. Each class is intended to give re-
presentative examples of what analysis detectability can be achieved
with other compounds in the class. If the compound of interest
is not present, it is reasonable to assume it will have sensitiv-
ity similar to a compound which is similar in structure. For
further clarification AID should be contacted.
Each class is listed under the detector type which is suggested
for that analysis. This does not mean to imply that the compound
would not be detected with a differenct detector type, but that
this detector type is the most widely used for that species.
In addition to minimum detectable (MD) in parts per million
(ppm), the threshold level value (TLV) is stated in ppm. This
,'jives the operator an indication as to analysis confidence. For
•example, in the case of Acetone with a TLV of 1000 ppm, minimum
detectable of 0.2 ppm is unimportant. The table also lists sug-
gested columns and column temperatures for the analysis of interest.
Several points should be made. The table is by no means meant
to be all inclusive. It is representative of.the type of compounds
and the level of sensitivity that can be expected from a gas
chromatographic analysis. The column and column temperature may
be changed depending on the compound present in the sample to
maximize system performance and eliminate interference. This
analysis adjustment may result in a change in system response for
an individual compound. It should be remembered that minimum
detectable (MD) is affected by sample size and retention time.
The following table illustrates compounds and conditions which
give retention times of under 6 minutes and a sample size of 1 ml.
in air.
ANALYTICAL INSTRUMENT DEVELOPMENT, INC
RT. 41 & NEWARK RD., AVONDALE, PA. 19311 PHONE: (215)-268-3181
-------
Mater:al
FLAME IONIZATION DII3CTOR MODEL 511
TLV (ppm) MD (ppm) Column
ethane
Mothy1 Chloro-
form (1,1,1,-Tri-
chloroethane)
Trichloro-
ethylene
350
100
0.1
0.15
(1)
(1)
Oven Temp,
Keton- s
Ace one
Methyl Ethyl Ke- .
ton? ( 2-Butar one )
Methyl Isobu-iyl
Ketone
Methyl n-Butjl Ke-
tonci (2-Hexanone)
2-Pentanone
3-Pentanone
Aromatics
Benzene
Toluene
Xylenes (mixed
isomers)
Kthyl Benzene
Chlorinated
Methyl Chloride
Ethyl Chloride .
Methylene Chloride
( Dichlorome thane )
Chloroform
Vi.nylidene Chloride
Carbon Tetrachloride
1,2 - Dichloro-
1000
2 JO
100
100
200
-
10
200
100
100
100
1000
500
25
10
10
_
0.2
0.2
0.10
0.10
0.10
0.10
0.05
0.04
0.12
i p-xylene )
. 0.1
0.11
0.08
0.2
0.4
0.2
0.5
0.05
(1)
(1)
(1)
.(1)
(1)
(1)
(1) •
(0
(0
(1)
.(2)
(2)
(2)
(1)
(1)
(1)
(1)
80°
80°
80°
80°
80°
80°
80°
80°
80°
80°
140°
140°
140°
55°
50°
.55°
55°
55'
80°
(131)
-------
Material
FLAME IONIZATIOH DETECTOR MODEL 511
TLV (ppn) MD (ppm) Column
Oven Temp.
Chlorinated (Cont'd)
Tetrachloro-
ethylene
Vinyl Chloride
Fiaon 12 (Difluoro-
dj chlorome thane)
Fraon 114 (1,2 Di-
ch Lorotetraf lu-
oro ethane)
Freon 11 (Fluoro-
trichlorome thane )
Acetates
Etiyl Acetate
i-?ropyl Acetate
n- Propyl Acetate
n-."Butyl Acetate
n- vmyl Acetate
AlCO. LOlS
Me ;hanol
Et .anol
1- 'ropanol
2- ropanol
l-!3utanol
2-3utanol
1-. .'entanol
2- 'entanol
3- 'entanol
2- lexanol
3-Hexanol
100
1.0
1000
1000
1000
400
250
200
150
100
200
1000
200
400
100
100
-
-
-
-
_
0.3
0.05
0.62
0.26
0.85
0.04
0.03
0.05
0.09 .
0.09
0.5
0.06
0.05
0.04
0.06
0.06
0.09
•0.05
0.05
0.05
0.07
(1)
(2)
(1)
(1)
(1)
(10)
(10)
(10)
(10)
(10)
(2)
(2)
(10)
do)
(io)
(10)
do)
(10)
(10)
(10)
(10)
80°
125°
80°
80°
80°
100°
100°
100°
125°
125°
135°
135°
125°
125°
125°
125°
150°
150°
150°
150°
150°
- 2 -
(131)
-------
FLAME IONIZATION DETECTOR MODEL 511
Material
Para ff ina
Methane
Et.iane
Pi opane
i- Butane
n- Butane
n-Pentane
n- Hexane
n-Heptane
•n-Octane
Material
CO
CH,
TLV (ppm)
-
-
1000
-
500
1000
500
500
500
FLAME IONIZATION DETECTOR
MODEL 514 OR
TLV (ppm)
50
MD (ppm)
0.04
0.04
0.02
0.02
0.02
0.01
0.02
0.02
0.05
W/CATALYSIS
511 OPTION 1
MD (ppm)
0.3
0.2
Column
(2)
(2)
: (2)
(2)
(2)
(1)
(1)
(1)
(1)
CHAMBER
4
Column
(8)
(8)
Oven Temp.
80°
80°
100°
130°
130°
60°
60°
80°
80°
Oven Temt>.
140°
140°
Total Hydro
carbon
.05
(9)
- 3 -
(131)
-------
Haterial
FLAME PHOTOMETRIC DETECTOR MODEL 513
TLV (ppm) MD (ppm) Column
so2
H2S
COS
cs2
Methyl Mer cap-
tan
Ethyl Mercap-
tan
Dimethyl Sulfide
Diir ethyl Disulfide
5
10
20
0.5
0.5
THERMAL CONDUCTIVITY DETECTOR
Matei ial
Or sat type analysis
CO
co2
°2
N2
CH4
NO
H,
NGAAA procedure
Propane
Butane
Propane
Butane
Benzene
Toluene
TLV (ppm)
50
5000
-
-
-
25
-
1000
500
1000
500
10
200
0.03
0.03
0.03
0.03
0.03
0.06
0.1
0.1
MODEL 512 OR
MD (ppm)
50
200
50
200
40
70
40
50
50
200
200
75
75
(3)
(3)
(3)
(3)
(3)
(3)
(3)
(3)
511 OPTION
Column
(4)
(4)
(4)
(4)
(4)
(4)
(5)
(6)
(6)
(2)
(2)
(1)
(1)
75°
75°
75°
75°
75°
100°
100°
100°
17
Oven Temp.
40°
40°
40°
40°
40°
40°
40°
65°
65°
100°
100°
80°
80°
_ 4 _
(13D
-------
ELECTRON CAPTURE DETECTOR MODEL 510 OR 511 OPTION 06
Material TLV (ppm) MD (ppm) Column Oven Temp,
c«4
Chloroform
1,1,1, trichloro-
e thane
Chlorine
Sulfur hex-
aflouride
Freon 112
NITRATE ESTERS
TNT
Pesticides
10
25
350
1
-
503
0.001
0.02
0.005
0.5
0.0001
0.001
0.005
0.01
0.005
(1)
(1)
(1)
(3)
(8)
(2)
(7)
(7)
(7)
50°
50°
50°
40°
40°
. 125°
125°
190°
COLUMN REFERENCES
(1) 6' 10$ DC200 on 80-100 mesh Chromosorb WHP
(2) 6' 80-100 mesh Chromosorb 102
(3) 6' 15$ UCON 50 HB280X on 40-60 mesh Chromosorb T
(4) 2' 80-100 mesh Chromosorb 102 + 4' 45-60 mesh Molecular Sieve 13X
(5) Same as (4) except 2' 1/8" delay line on entrance of MS column
(6) 30' 27$ DC 200 on 45-60 mesh Chromosorb P
(7) 6' 3$ GC SE30 on 80-100 mesh Chromosorb W-HP in Glass
(8) 10' 45-60 mesh Molecular Sieve 5A
(9) No column. Sample direct to detector and corrected for "Zero" air Blank.
(10) 6' 5$ Carbowax 1540 on 80-100 mesh Chromosorb VHP
- 5 -
(13D
-------
The Use of Silica Gel in Source Testing
M. FELDSTE1N, S. BALESTRIERI, and U. A. LEVAGGi
Bay Area Air Pollution Contrul District, 1480 Mission Street, San Francitco, California
(gj This is a study of the adsorption of a large group of solvent vapors upon .silica
gel and their subsequent quantitative dcsorption. Esters, kctones. .-iromatic and
aliphatic hydrocarbons, and halo^enated hydrocarbons were ninoni; those studied.
Except for certain low-niolccular-weight hydrocarbons, the silica i:el was extremely
efficient for adsorbing organic solvents. In general, dimcthylsulfo.xidc proved ideal
for clution of adsorbed materials and for subsequent analysis by pji cliromaiography.
It could not be used for hydrocarbon solvent mixtures or for higher boilini; solvents.
In these cases, carbon disulfide, alone or with water, was used for desorption.
Introduction
'T'HE LITERATURE contains many re-
•*• ferences on the use of adsorbing materials
for the collection of organic vapors. Char-
coal,1'3 silica gel/'0 and iiiolccular sieves7 are
only a few of the adsorbing materials which
have been suggested. Desorption has been ac-
complished by heating at elevated tempera-
tures,8 by using polar and nonpolar organic
solvents,1-* and by using water and acid and
alkaline solutions.0 Analysis ranged from sim-
ple weighing0 to complex instrumental meth-
ods including infrared,8 gas chromatograph-
ic,1 and chemical procedures.' Recovery
data have shown values ranging from 0 to
100% for various types of organic materials.
This study was prompted by the need to
collect and analyze organic vapors emitted
by various industrial processes, including
paint spraying and drying, lithographic opur-
ations, dcgrcasing, drying cleaning, and pe-
troleum refining operations. The range of
organic vapors to be collected and analyxed
included alcohols, esters, kctones, aromatic
and aliphatic hydrocarbons, halogenated hy-
drocarbons, and a variety of other types of
organic compounds. The study centered
around two problems: the efficient collection
of organic vapors on an adsorbing medium,
.ind the quantitative removal of the adsorbed
material for subsequent analysis.
Areation System
Prior to field testing of the procedure final-
ly developed, a laboratory system was devised
for the collection of organic vapors. The ap-
paratu* inccl is shown ir. Figure 1. Silica gel
was chosen as the adsorbing material. Three
probes (^4 inch X f> inches) containing 20
gm of Davidson silica gel PA 400 (8 to 16
mesh, dried at 110° to 120°C for at least 2
hours prior to use; were arranged in scries.
The organic material was injected into the
flask in quantities ranging from 25 to 50 .ul,
with the use of a constant-drive syringe
pump. Larger volumes of sample (250 /*!)
were injected into the flask manually over a
5- to 10-minute period. The flask was heat-
ed by means of a healing mantle to 45° to
50°C. In the case of higher boiling materials
the kettle tcniperaimv was raised to 120CC.
Air was drawn through ihc ketch- and the
probes at a rate of 0.25 cfm for a period of
1 hour for a total of 15 cubic feet. The con-
tents of the probes were then poured into
separate 125-ml T/S stoppered Erleiimeycr
Masks for desorption and analysis.
Desorption
A variety of desorbing materials was stud-
ied to cli'U-rmine the efficiency of desorption.
Since infrared (IR) and g;\.s-liquid-chioma-
tographic (GLC) analytical procedures were
3HI
-------
July-August, l'JI,7
FIGURE 1. Apparatus used for adsorption of
known concentrations of solvent vapors on silica
gel.
to be used, it was necessary to select sol-
vents compatible with the final analytical
procedure. For gas chromatographic analyses,
it was necessary to employ desorbing materi-
'als which did not interfere with the material
being eluted. For infrared analysis, the sol-
vent had to be nonabsorbing in the infrared
region of interest (2 to 15 microns).
TABLE I
Gas Chromatographic Analytical Data
(4-foo( Carbowax 20 M Column)
Temp- Approximate Time
erature Flow Ra'e for Peak
Compound f C) (ml/min) (minutes)
Methyl acetate*
Butyl acetate
Ethyl acetate
Vinyl acetate
Butyl Celtosolve
acetate
Ethyl acrylate
Isopropyl alcohol
r-Uutyl olcuhol
Uulyl Cellasolve
Benzene
Toluene
Xylcni;
Trie htoro ethane
Trichlorjcthylcr.e
Tctrachlorocthylene
T el rachluro ethane
Mcthylcthyl ketonc
100
100
100
100
100
100
100
100
100
100
100
100
70
70
70
70
100
Methyl isobutyl ki-tunclOO
Dimcthylsulf oxide
100
78
73
87
75
90
75
75
75
75
61
61
6!
81
81
81
81
87
75
75
1.05
2.8
1.05
1.2
20.0
1.65
1.3
3.6
22.2
0.9
3.1
5.9
2.6S
5.3
6.5
8.1
1.2
2.25
35.0
The choice of analytical procedure was
determined by the nature of the collected
material. Simple: compounds like aromatic
hydrocarbons, esters, and alcohols could l>c
analyzed by GLC procedures, whereas com-
plex materials like Stoddard solvent and V.M.
P naphtha, which are mixtures of innumera-
ble hydrocarbons of varying chain length,
had to be analyzed by infrared procedures.
The desorbing procedure was as follows:
To the flask containing the silica gel, 25 ml
of desorbing solution was added. The flask
was stoppered and allowed to stand for 2
hours with occasional shaking. The solvent
was decanted from the gel and stored in a
T/S stoppered test tube. Aliquots of the sol-
vent were'then taken for GLC or IR analysis.
Where CS2 and H2O were used as the de-
sorbing material, the procedure was as fol-
lows: 25 ml of CS2 was added to the flask
confining the gel, as described above. After
the two-hour period, 15 ml of water was
added to the mixture in the flask. The flask
was permitted to stand further for one hour
with occasional shaking. The solvent was
then filtered into a T/S stoppered test tube
for subsequent analysis. Under these condi-
tions, no water layer was present with the
CS2, since most of the water is retained on
the silica gel.
Analytical Procedure
The GLC procedure consisted in injecting 1
to 5 /il of eluate into the gas chromatograph.
A Beckman Model GC2A gas chromatograph
with flame ionization detector was used. The
column was a 4-foot X j4-inch stainless-steel
tube filled with 20% Carbowax X 20 M on
firebrick 60/80 mesh. Carrier gas was N; at
flow rates that varied depending on the com-
pound being analyzed (Table I). Tempera-
ture was usually 100°C.
For IR analysis, a Beckman IR 4 spcc-
trophotometer was used. The eluate was
transferred to a 0.1 -mm cell, and the absorp-
tion was determined at the wavelength of
interest. For Stoddard-typc solvents ab-
sorption at 3.4 microns was found to be pro-
portional to concentration of solvent. For
other solvents measured by IR spectiopho-
-------
American Intlu.iliial Hygiene A\*uciati(tn Journal
tometry. the wavelength used depended on
tlic absorption characteristics of I lie material.
Thus llic 5.7-inicion Land was used to meas-
ure concentration of ketoncs and esters. All
IR analyses were done in CS- solution.
1 It was found that diincthylsulfoxicle (DM-
SO) was an ideal solvent for clution of ad-
sorbed inaterials from silica gel and for sub-
sequent GLC analysis. Having a high boiling
point, the DM SO peak appears after the ap-
pearance of the peaks of the solvent being
studied. In two general cases, DMSO could
not be used for GLCJ analyses. These includ-
ed hydrocarbon solvent mixtures such as
Stodclard and VMP naphtha, which contain
many kinds of hydrocarbons, and higher boil-
ing solvents which may appear with the
DMSO peak. In these cases CS: and CS-,/
H2O mixtures were used for desorption, and
analysis was accomplished by infrared.
Results of Analyses
The results of the experiments are shown
in Table-II. Analysis for each of the ma-
terials v/as performed separately on each siii-
ca gel probe, la all cases shown in the table,
there were no solvents found on the second
or third probes. Recovery figures shown for
DMSO as an eluant indicate that all the ma-
terial is adsorbed on the first probe. In field
tests, where the concentration of solvent is
.unknown, and generally high, some of the
materials were found on the second and third
probes. Certain generalizations can be drawn
from the table:
1. Carbon disulfide is a poor eluant for
most'of the inaterials studied. The notable
exceptions are solvent mixtures such as Stod-
dard and naphtha and hydrocarbons of more
than eight carbons.
2. Addition of water to the carbon disul-
fide in the eluting flask improves the elution
of such compounds as toluene and xylene and
certain halogenated hydrocarbons and ethyl
acrylatc. Poor recovery is still obtained with
alcohols, ketones, most acetates, and some
halogenated hydrocarbons.
3. DMSO is an efficient eluant for all the
materials studied.
With the exception of certain low-molecu-
lar-weight hydrocarbons, the silica gel was cx-
TAnLF. II
Adsorption and Ri-covi-ry of Solvents
from Silica Gi-l
Microhlcrs
PTO VTY
Solvent injected Microl HIT*
Kluei
Ethyl acelnie
Vinyl acetate
Uutyl acetate
Methyl acct.ite
Ethyl ocrylaie
Isopropyl alcohol
n-But:inol
Butyl cetlosolve
Benzene
Toluene
Xylene
Stoddard?
Naphtha8
Trichloroelhane
Tetrachloroethylene
Tetrachloroethane
Methylethyl ketone
Methylisobutyl ketone
n-Octane
Eluent -
Ethyl acetate
Vinyl acetate
Butyl acetate
Methyl acets'.s
Ethyl ncrylatc
Isopropyl alcohol
n-Butnnol
Buty! cellosolve
Benzene
Toluene
Xylene
Stoddard3
Naphtha3
Trichloroelhane
- Tetrachloroelhylene
Tetrachloroe thane
Methylethyl ketone
Methylisobutyl ketone
if — t'i'j
US
250
250
125
250
125
250
250
125
250
125
125
100
430
100
300
100
250
100
210
100
300
35
100
35
100
35
100
250
125
25,0,
- qs2/H2c
ID
250
250
125
250
125"~
250
250
125
125
125
100
430
100
303
100
250
100
200
100
300
35
100
35
100
35
100
250
125
—
5 .
13
48
105
25
53
12.5
b
—
6
6
5
12
39
124
38
105
98
204
95
288
1.7
5
10.5
35
27.5
77
10
6.5
253
>
74
152
170
121
i-44
41.5
80
240
6
6
6
95
415
98
303
97
238 .
97
204
93
283
34
95
34
98
34
95
•88
94
—
2
S
38
42
20
** 1
5
S
u
S
5
5
.1
.19
41
38
•! 2
98
97
95.
96
5
5
35
35
79
77
4
5
100.9
59
61
63
97
98
33
32
96
S
S
5
95
96
.98
101
97
95
97
97
93
96
97
93
97
98
97
95
35 '
75
Eluent - DMSO
Ethyl acetate
Vinyl acetate
Bulyl acela'.f
Methyl acetale
Ethyl acrylnte
Isopropyl alcohol
n-Dutanol
Butyl Ccllosolvc
Benzene
Toluene
Xylene
Trichloroelhane
Trichlorocthylcnc
Telrachloroelhanc
Methylethyl ketonr
MrthyllBobutyl kelunc
n-Oclnne
7 ?
250
250
250
250
250
125
250
125
125
125
430
100
300
100
250
100
100
35
100
2SO
125
250
70
256
243
244
251
240
121
240
121
r.'i
122. S
425
97
303
104
253
96
97
36
96
242
12.1
2SO
93
102
97
97.6
100.2
96
97
96
97
96
9d
99
97
101
104
103
96
97
102
96
97
93
100
"infrared unalynU.
-------
384
July-Aunu.it. 1'J67
TAULEllI
Efficiency of Adsorption of Hydrocarbons
on Silica Gel
Efficiency of Adsorption
Hytlrucarbon
n-fiMilnnv
n-hcxanc
n-heptunc
n-Octane
n-Nonanu
Cyclopcntanc
Cycluhoxane
Cyclohexene
i
!• irst
Probe
• 5
*N S
IS
90-100
90-100
'-.5
40-50
90-100
!'/, niHnrhpil)
Second
1'rol.c
SS
"* 5
60
^5
to sample until 'the third probe begins to be-
come warm to the touch, indicating that ad-
sorption is beginning to occur on the third
probe.
References
1. FR.M.-ST. C. L.. ami E. K. HUIM\N.N: Clutrco.il &=>•
plinit Tublr* for On;:inic Vapor Aiialvii* by Ci* Chran-
j!osr-|>hr. .liner. Ind. llvg. Auoc. ]. 17: C3 (linjl>)-
'.'. Tunn, A., unit C. J. D'A.vyiu: Comix.iitinn of Natural
r"rr>li Air. / Air 1'utlulinn Control Aiioi. )-'• ?•*
iiye.'i.
3. TUIK. A.. J. 1. Mouow. jucl D. K. KAfuM: Olrlia
-------
American Industrial Hygiene Association Journal
385
Jsomr i i/jlU'.i in \(|MTPII\C S.iiiipliiii* yn Activated Atfi-f. /. _'»: 4^V ll'Mr4j..
Citl.'.n .Inal. (.'/i.»i. .M 11.1 ll'HiJ). 7. (ili.Mi>VM>.\, I'., .mil S II. S>nril. In Hrmi-:al ol
Air l>y Ttlu- ll.\' M.:lr,u.
•n Krut.
4. V*.\ MIIIHIK, J II. C : KxiMrirncr!' with Silu.i < •"
a> Ail-.... l«iu. HIH.T. ;»rf. y/xff. .-lu,... /. 36: !'
3. liucnwMj). II.: Actuated Silica CIH a% .in Aiitjiminaii[s. Otttif. Hnilili H't:
(Oilaa-al 11: 14 (l')(ii).
d. WHITMAN, N. K., and A. K. JOIINNTON: Saiuplinic and
Analvsi-i of Aromalir I lyiii DcaHion V.ipors in Air: A
"r> L\ini{ the Mrih«>d of A'Korption.
.V. r. .S'lni^ /J,p(. /.ui,,,r /nrf. /(,.//. 17: IliO (I9M).
Received November 24, 1966
> frs^rf.*vaS3i. i <•'.-••'
^ P\-^|^8^4fe^
X XV^B^^^V J:5
^% t \ 7
^& \ M /
>vS^-" V //
. - --->-—rf•<•>-, //
f^£% -(/
AIHA Presidents
At the annual banquet of the American Industrial Hygiene Association dar-
ing the annual meeting in Chicago, Dr. Clyde M. Berry (right). President of
AIHA for this year, presented to William T. McCormick, retiring President,
a plaque symbolizing the appreciation of the American Industrial Hygiene
Association for the leadership and devotion "Bill" has given to the Association
over the years and particularly during his term in this office.
-------
Determination of
organic vapors in
industrial atmospheres
SOLVENT VAPORS ARE ubiquitous components of
the atmospheres in many manufacturing plants.
Solvents used in degrcasing, paint applica-
tion, and other industrial processes contribute to
this condition. Prior to 1970, solvent vapor concen-
trations in factory environments were voluntarily
controlled to accepted limits, such as the threshold
limit values of the American Conference of Govern-
mental Industrial Hygienists. This control was ex-
ercised for worker safety as veil as for die con-
comitant material economies. The Williams-Stciger
Occupational Safety and Health Act of 1970 elim-
inated all options. Solvent vapor concentrations in
the workplace came under statutory limits, and
compliance with these OSHA limits became man-
datory.
Semiquantitative solvent vapor concentration
measurements have been made for many years uti-
lizing manual sampling pumps and proprietary color
indicator tubes. While such devices are useful as
range-finding means, they usually lack the specific-
ity, precision, and accuracy necessary to document
worker exposure. In consequence of these limita-
tions, a literature search was undertaken to uncover
a suitably selective, quantitative method applicable
to the needs of a large, multiplant manufacturing
facility utilizing a variety of solvents.
Kupcl ct al.1 had prepared a directly pertinent
literature survey early in 1970. They found that
difficulties in collection and concentration of air
samples had long been an impediment to the success-
ful evaluation of workplace solvent vapor conccntra-
lions. The pas chromatograph (GC) was found to
be most often mentioned as the analytical tool,
whereas sampling techniques ranged from solvent
Dr. MiH'ller is Chemist, Air Quality, unit Mr. Miller i.f
Environmental Scientist, Major Appliance laboratories.
Central likctric Company.
absorption to charcoal adsorption. Kupcl and co-
workers at the National Institute for Occupational
Safety and Health (NIOSH) chose the charcoal
adsorption technique of Ottcrson and Guy,2 and
then systematically evaluated collection efficiency,
carbon disulfide dcsorption efficiency, and quantita-
tive determination of .14 solvents by GC.1 The
sampling and analysis techniques developed by
these workers were accepted by the Occupation::!
Safety and Health Administration (OSHA) as
standard methods for compliance investigations.3
Kupel et al.4 devised a technique for collection and
identification of individual GC fractions applicable
to the complex atmospheres likely to be encountered
in factory situations. They utilized a mass spec-
trometer (MS) to identify the GC fractions which
were individually collected in capillary charcoal
traps.
Two other methods have appeared in the litera-
ture which merit'consideration; both portable in-
frared spcctrophotometcrs* and portable gas chro-
matographs6 have been used to good advantage by
workers in field evaluations. Each of these methods
requires a significant capital equipment outlay,
which, in the case of our laboratory, would have
duplicated more versatile, sophisticated laboratory
instrumentation already available. Therefore, the
decision was made to meld the techniques of the
NIOSH researchers with existing laboratory instru-
mentation.
Experimental
Solvent vapors arc collected on activated car-
bon, dcsorhcd with a suitable agent, and separated
by gas chromatographic procedures. Individual
peaks appearing in the chroinatograms arc identi-
fied by retention times and/or by mass spectrometer
analysis; concentrations arc calculated from cali-
bration curves.
AMERICAN LABORATORY : 49
-------
ORGANIC VAPORS continued
SAMPLING TUBE. COMMERCIAL VERSION
C at 20°C/mln
5minaM40°C
2) 1208C for 10 mln. 120°-200°C at 20°C/mln
50 : MAY 1974
-------
ORGANIC VAPORS continued
Tabled
FFAP column
Organic solvent analyzed
1,1,1-trlchloroethane
l-propyl alcohol
Methyl ethyl ketone
Benzene
Trlchloroethylene
Perchloroethylene
Toluene
n-butyl alcohol
Methylene chloride
m-xylene
Styrene
Retention data and FID .
relative sensitivities for organic compounds
Carrier (low rate, 24 ml/min
FID relative sensitivities Relative retention timcsc
Literature values' Experimental* Literature values Experimental
0.53
0.61
1.12
1.07
0.66
1.04
0.18
0.61
1.21
0.21
0.16
1.07
1.09
0.85
1.00
1.16
1.38
1.51
2.10
0.83
0.86
0.86
1.00
1.10
1.28
1.45
1.90
1.93
2.14
3.41
•Response factors reported by Dietz7 with factors relative to heptane.
^Response factors calculated from a seven-component mixture with the toluene response set equal to 1.07, facilitating com-
pari
-------
ORGANIC VAPORS continued
was inserted in the exit port of the GC as the peak
of interest appeared on the chromatogram. and was
withdrawn on the downside of the peak prior to
reaching the baseline. The carbon particle was trans-
ferred to the mass spectrometer direct inlet probe.
A Bcndix inlet probe (model MA016) was con-
nected through a vacuum lock into the ion chamber
of a Bendix time-of-fiight mass spectrometer mod-
el 12-107). The probe temperature was raised grad-
ually to release the material from the carbon for
recording of the mass spectrum. The limits of detec-
tion for several organic compounds using this pro-
cedure have been determined.4
Quantitative analysis
The NIOSH method1 employs an absolute cali-
bration approach to achieve quantitative analysis of
eluted samples. Standard solutions arc prepared
where
volumetrically in the elution solvent. Concentrations
equivalent to 0.5, 1.0, 2.0, and 5.0 times the OSHA
limn were prepared using Eq. (1):
/*! material in 1.0 ml standard solution =
FXLXVXM
PX 24.45 XD (I)
F = fraction or multiple of OSHA limit
L = OSHA limit, ppm
V = air sample volume, ms .
M = molecular weight, g/mole
p = density, g/ml
24.45 = liters/mole @ 25°C, 760 mm Hg
D = volume desorbing agent, ml
5~Ml portions of the standards and samples are
analyzed by GC. Peak areas of the standards are
related to concentrations to obtain calibration
curves; linear relationships were obtained in every
T..t.f*.
UkMW
Compound
l-amyl acetate
n-aniyl acetate
.Benzene
2-butanone
n-bulyl acetate
n-bulyl alcohol , .
Butyl carbitol Do'>Mt?\
Carbon tetrachlorlde
Chloroform
Diacctono alcohol
p-dioxane
Ethanol
Gthyl ether
Iso-octane
-Perchloroelhylcno
|-propyl alcohol
Pyridino
Toluene
Trichlorocthylcno
SC-100 (aromatic distillate)
Xylene
Desbrption percentages
OSHA
Limits (Ret. 9),
ppm
100
100
10
200
150
100
None established"
. 10
50
50
100
1000
400
None established11
100
400
5
200
100
None established8
: 100
°1 organic solvents
Desorplion percentages
at OSKA limits
Average Range
97 95-99
96 94^-99
96 94-100
96 90-99
94 88-97
91 87-97
98 95-100
101 96-113
100 98-101
97 95-100
68 84-98
67 64-72
90 86-96
95 92-97
96 95-97
80 75-87
48 46-50
98 93-99
97 93-99
97 94-100
98 97-100
Reference"
GE
GE
NIOSH. GE
NIOSH. GE
NIOSH
GE
GE
NIOSH
NIOSH
GE
NIOSH
NIOSH
NIOSH
NIOSH
NIOSH, GE
GE
NIOSH
NIOSH. GE
NIOSH, GE
GE
NIOSH. GE
"Concentrations employed for compounds v.'ith no established limits were as follows'
butyl c:irbilol = 50 ppm. iso-ocl.inc — 400 ppm, SO 100 = I Go ppm_
fcl'iTCciiLipcs represent llie avcincc v;iluc of three to twelve determinations
CNIOS1I Oiii.i t;iken from Kef. 1; Gl; v:\hics determined experimentally jn our laboratory
54 : MAY 1974
-------
ORGANIC VAPORS continued
Table 5
Benzene
Toluene
m-xylene
l-propyl alcohol
n-butyl alcohol
Diacetone alcohol
Butyl carbitol
SC-100 and xylene (14 peaks)
Styrene
Methylene chloride (3 carbon
Efficiency of charcoal technique
Aluminum
odor chamber
concentration,
ppm
18.7
103
87
153
61.6
26.3
11.1
100
90
600
OSHA
• Limit
(Rel. 9),
ppm
10
200
100
400
100
50
a
100»
100
500
Peak areas
Carbon
lubes
0.91 ±0.01
5.65+0.12
5.97+0.13
1.99+0.20
0.90±0.05
0.29±0.02
0.13±0.01
10.8+0.1
9.5+0.4
7.8+0.2
(counts X 10-s)«
Standard
solution'1 .
1.01 ±0.08
5.65+0.33
5.15+0.30
2.29
1.20
0.48
0.15
11.3
11.5±0.1
8.4+01
Analytical
efficiency,
%«
90
100
115
87
75
60
87
96
83
93
tubes in series)
•Average values of peak areas and standard deviation for three separate determinations: X+S, with N=3.
^Standard solutions prepared volumetrically in carbon disulfide equivalent to aluminum odor chamber concentrations.
eAnaSytica! efficiency calculated from Eq. (?.).
o'No OSHA iimii.
•100 ppm OSHA limit for xylene; no OSHA limit for SC-100. a complex m-'xture of aromatics.
case. Sample solution peak areas are converted to
concentrations using the calibration curves.
Some disadvantages of this absolute calibration
approach arc that standards arid samples must be
rcproducibly injected and detector sensitivity must
remain constant.8 After initial calibration, one stand-
ard should be run at the time of analysis to verify
instrument sensitivity.
Method evaluation
Desorplion efficiency
Adsoibed organic compounds may or may not be
readily displaced from carbon by carbon disulfidc,
and it is necessary to know the amount dcsorbcd to
avoid inaccurate interpretation of analytical results.
Kupel ct al.1 list dcsorption percentages for fourteen
compounds, and another seven materials were eval-
uated in this laboratory (Table 4). To obtain
dcsorption efficiency, a known volume of the sub-
stance was injected directly into the carbon sam-
pling lube. The tube was capped and left sitting
overnight to provide ample time for sorption equi-
librium to be attained. The carbon was cluted, and
the dcsorbcd material was analyzed by GC. The
dcsorption efficiency for each material was calcu-
lated as the ratio of the amount dcsorbcd to the
amount added to the carbon tube.
58 : MAY 1974
Analytical efficiency
A test was devised to measure the overall pre-
cision and accuracy of this method, including sam-
pling, adsorption, .dcsorption, and . gas chromato-
graphic analysis. Known concentrations of organic
vapors were generated in a 76.6-ft3 test chamber.'
Thorough mixing was provided by fans located
within the chamber. Hydrocarbon levels 'in the
chamber were monitored with a Beckman total
hydrocarbon analyzer (model 109 A) and a Varian
gas chromatograph to ensure that the proper quanti-
ties were present for sampling. Three 10-liter air
samples were taken from the chamber through car-
bon tubes, dcsorbcd, and analyxcd by GC. Ana-
lytical efficiencies were calculated using the follow-
ing relationship:
Analytical efficiency (% ) =
Organic vaporconc. from carbon lube analysis X 1 00
Known organic vapor cone, in chamber
Analytical efficiencies for several materials arc given
in Table 5. Results obtained for benzene, toluene,
and xylene were in agreement with chamber do-
sorption percentages reported by Kupcl.1
Polar compounds, such as alcohols and amines,
exhibit lower dcsorption percentages with carbon
disulfidc (CS2), and the overall efficiency of the
-------
Precision microsampling . .. made easy!
Drummond MICRODISPENSER'
The MICRODISPENSER is a
precision 1 to 5 microlitcr
dispenser featuring a stainless
steel plunger and disposable
glass bores.
In immunodiffusion and
electrophoresis .... a
proven valuable aid.
Drummond Scientific Company
500 PARKWAY
BROOMALL • PENNSYLVANIA • 19008
Circle Read;,- Ccrvicc Cere! Mcv 1?
ORGANIC VAPORS continued
method is reduced.1 Polycat 8 (N.N'-dimethylcyclo-
hexylamine) was examined as a polar air con-
taminant, using the aluminum test chamber. Only
a fraction of its original concentration was evident
from dcsorbcd carbon samples. The efficiency of
measurement varied with the choice of desorption
solvent, with- chloroform being more effective than
CSj in removing this amine from carbon. The limit
of detection for Polycat 8 with chloroform as the
cluting solvent is between 4.2 and 9.6 ppm, since
the amine was not detected in chromatograms ob-
tained from samples at the lower level.
There was some question about the applicability
of this method to polymerizable monomers or sub-
stances with relatively low adsorptivity on carbon.
Styrcne vapor sorbcd on carbon could polymerize
due to surface catalysis, or polymer formation could
occur on the column packing of the gas chromato-
graph. Neither reaction prohibited valid analysis by
this method, and styrene was effectively determined
using the carbon tube technique (Table 5). A single
cnrbon tube (Figure 1) did not adequately trap
mctliylcnc chloride vapors at the OSITA limit of
500 ppm, and the backup carbon section contained
a considerable amount of this contaminant. Increas-
ing the quantity of carbon available for sorption by
utilizing three carbon tubes in scries resulted in a
high analytical efficiency for methylcne chloride
(Table 5). For substances with relatively highx
OSHA limits and appreciable vapor pressures at \
room temperature, it may .be necessary to increase '
the amount of activated carbon or lower the volume )
of air sampled in order to trap the material effi-
ciently.
Practical applications
Four separate experiences will serve to illustrate
our use of the carbon tube method in industrial
situations.
Cose 1 — Simple mixture of readily identified
components
An employee performing a spraying operation
complained of exposure to odorous vapors. Carbon
tube samples were obtained from the breathing
zone of this operator. Normal and isoamyl acetates
were identified by chromatographic retention times
and wore the only substances found in the dcsorbcd
air samples. The spray applied by this worker con-
tained a mixture of the acetate isomcrs, and MS
analysis was not required. The desorption efficiency
of the carbon disulfklc clution was greater than
95% for both normal and isoamyl acetates, and
58 : MAY 1974
-------
FFAP COLUMN 9 1OT»C
CARRIER FLOW- 24 ML/MIN.
RANGE-10"11 AMPS/MV
FID ATTENUATION-6
STANDARD SOLUTION.
FACTORY XYLENE.
EQUIVALENT TO OSHA
LIMIT
AIR SAMPLE FROM SPRAY BOOTH.
DESORBED FROM CARBON TUBE
Figure 3 Comparison ol desorbed air sample with
standard solution.
GC analysis revealed acceptable workroom vapor
concentrations.
Cose 2 — Complex mixture with single component
above OSHA limit
A spray booth was surveyed for airborne organic
vapors because employees were concerned about
persistent disagreeable odors present in the area. A
chromatogram of the desorbed sample contained
over twenty separate peaks. Comparison with chro-
matograms of bulk liquid samples revealed that
three of the major peaks were due to xylene. The
concentration of xylene alone was above the OSHA
limit (Figure 3), and it was not necessary to
identify the smaller unknown peaks. This informa-
tion was reported to the responsible manager, and
corrective ventilation action was immediately taken.
Case 3 — Mixture of solvents identified by
mass spectrometry
A paint mix area was surveyed for vapor con-
centrations of six organic materials: methyl ethyl
kclone (MEK), isopropyl alcohol (1PA), butyl
alcohol, xylenc, diacctone alcohol, and butyl car-
bitol. CS2 desorption percentages of the four alco-
hols were determined (Table 4) since they were
not available from the literature. Analytical effi-
ciency for these alcohols was in the range of
60-90% (Table 5). These values can be compared
with Kupel's clesorplion percentages of 10%. _ for
methyl alcohol" and 67% for ethyl alcohol.1 GC
analysis of a qualitative air sample indicated thnt
the factory atmosphere contained a number of com-
pounds (more than 20 peaks in chromatogram).
Peaks assigned to dincctone alcohol, butyl alcohol,
and xylene were present, but there was no evidence
of butyl carbitol in the atmosphere. lPA_jmd MEK
were not separated sntisfnctorily, the two giving n
single cinematographic peak under the analytical
conditions used. The single peak was evaluated as
Gas-Dry RLTER TR AP
for drying and filtering
Carri0r gas Remove moisture, oil
and dust from carrier gas before it flows
into your instrument. Get improved base-
line stability; more accurate data.
FEATURES:
• Clear acrylic tube allows visual monitoring
• No wrench needed to change cartridge contents
• "0" ring seal allows pressure to 100 psi
• Two sizes: 120 and 400 cc
New! BASE PLATE
Allows mounting Gas-Dry
Filter Trap on bench top for
continuous observation
Q
.f, ;
!!
j
Gas-Dry Filler Traps Ana hundred:
of other hclplul G. C. tools
»n described in Catalog 600.
Send lor your FREE copy.
Send me a copy of Cataloo 600.
Name/Title
Company/ Institution-
Address
City.
-State & Zip-
^Chemical Research Services, Inc.
£52W«tgateDrive • Addison. Illinois 60101
AL/S
Circle Reader Service Card No. 90
Circle Ro.idor Service Cnrd No. 79
AMERICAN LABORATORY
59
-------
ORGANIC VAPORS continued
MEK, which has a lower OSHA limit than docs
IPA. MS analysis was employed to identify un-
known peaks and to confirm assignment of peaks
according to retention times. Identification of the
minor peaks was not attempted, and only the larger
peaks were considered. MS analysis identified the
presence of IPA, MEK, toluene, Ai-butyl alcohol,
and xylene. It is interesting to note that toluene
contributed to the total worker exposure although
its presence was not considered prior to qualitative
analysis.
Case 4 — Paint diluents containing a variety of
aromatic compounds
A finishing operation was monitored for OSHA
compliance, and as suspected, the solvents were
multicomponcnt mixtures. Air samples contained
a multitude of peaks, and retention time compari-
sons with bulk liquid samples indicated the atmos-
pheres in the various spray booths were mixtures of
benzene, tolupne, xylene, and SC-100. Concentra-
tions varied with the sampling location, and differ-
ent chromatograms were realized for different parts
of the finishing area (Figure 4). These results dem-
onstrate the complexity of industrial atmospheres
where the process materials arc mixtures and where
the composition of the air can vary with locality or
time of sampling. To calculate the concentration of
SC-100, a single clean peak was chosen for com-
parison with standard solutions prepared volu-
mctricaliy. This approach was suggested by Kupel10
as an alternative to attempting identification of a
number of peaks in a solvent which is a distillation
cut. The variety of aromatic compounds potentially
present is aptly demonstrated by the work of Deans
and Scott"; their chromatogram of a heart-cut
naphtha showed the presence of benzene, toluene,
ethyl bciizcne, in- and p-xylene, isopropylbcnzene,
o-xylcne, 3- and 4-ethyl toluene, mcsitylcne, 2-cthyl.
toluene, and pseudo cumcnc. An OSHA limit has
not been established for SC-100, and so a limit of
100 ppm was assumed for this material. Analytical
efficiency for SC-100 v/as 95 %.
Discussion
The carbon tube method cannot be properly
used or interpreted without understanding the fol-
lowing:
1. Some air contaminants are poorly adsorbed
by carbon and would not be efficiently collected by
this method. The relative capacity of carbon for
FFAP COLUMN C1SO°C
CARRIER FLOW - 24 ML/WIN.
SAMPLE 1A
RANGE - 10"11 AMPS/MV
HID ATTENUATION =• 1
SAMPLE «A
Figure 4 Chromatogrnms ot desorbed air samples trom
adjacent factory orcas.
GO
MAY 1974
-------
various vapors is available from the literature" and
can be used to estimate adsorption efficiency.
2. Chemically reactive gases can be converted to
other species on the carbon surface and would not
be properly identified. This would be an important
consideration if the area of interest contained air-
borne species formed by thermal, radiant, or chem-
ical decomposition.
3. Carbon disulfide docs not readily displace all
organic compounds from carbon, and a different
desorption solvent would be necessary for some
compounds to obtain acceptable overall efficiency.
4. Organic compounds can be displaced from
charcoal tubes by the presence of chemical species
that are more strongly adsorbed. If organic ma-
terials are present on the backup carbon section of
the sample tube, the data should be carefully
scrutinized. „-.
5. The 180-mg "A" section of a carbon sampling
tube has a saturation limit of 28-45 mg (approxi-
mately 1000-1500 ppm) of total solvents.1 Each
section of the sampling tube collects solvent vapors
until the saturation limit is reached.
Summary
The adsorption of organic vapors by activated
carbon followed by liquid desorption and GC anal-
ysis is a reliable method for assessing occupational
hazards, provided certain limitations inherent in the
procedure are realized. The method is not univer-
sally applicable to all organic vapors, and each
determination must be considered a separate case.
References
1. KUPEL, R. E. et al., A.I.H.A. J. 31, 225 (1970).
2. OTTERSON, E. J. and GUY, c. U., Trans. Twenty-Sixth
Ann. Meeting A.C.G.I.II.. 37 (1964).
3. NIOSH Sampling Data Sheet 6 (June 16. 1972).
4. KUPI-L, R. E. ct al., A.I.U.A. J. 32. 383 (1971).
5. WILKS, p. A., Amer. Lab. 5 (12), 67 (December 1973).
6. nnmiRF.orr, P. j. ct al., Amer. Lab. A (7), 48 (July
1972).
7. UIKTZ, w. A.. /. Gas Clirom. 5, 68 (1967).
8. MCNAMI, ii. M. and PONELLI, E. J., flusic Gas Chroina-
tography, Varian Aerograph (1969), p. 149.
9. MAPI-S. w. ii. and VANCE, R. F.. ASI1RAE Trans. 71
(II), 52 (1971).
)0. KUPEI., R. t... NIOSH, Cincinnati, Ohio, private com-
munication (May 1973).
11. DEANS, n. R. and SCOTT, I., Anal. Cliern. 45 (7), 1137
(1973).
12. IIAKNI-IIY, ii. L., ASURAE Trans.. 64 (1946); 481
(1958).
•-<
Analytical Injector Valve: One
of the most reproducible and
accurate ways of injection in
high-pressure liquid chroma-
tography. This valve, adjust-
able to 4000 psi, allows plug
injections of 1, 2, 4 or 8 micro-
liter sample volumes by simply
turning a handle. The valve
has a column bypass position
for quick solvent change over.
Wetted parts are stainless
steel or filled Teflon."
Model 710 $425.00
Model 720 $475.00
(includes 250 microliter loop)
Preparative Injector Valve:
Designed for reproducible
sample injection in preparative
liquid chromatography. It is
essentially identical to the An-
alytical Injector Valve in con-
struction and specifications.
The sample volume, 15 to
1,000 microliters, is deter-
mined by loop fitted to the
valve. The column bypass
position is standard for quick
solvent change over.
'•""IS^i^Mi HI* Particle T«cK>nology end Liquid Cnro
CHJQ"; mScpomepSi
,—;:...• :x_i instrument corporation
5680 goshsn springs rd, norcross. ga. 30071 (404) 448-6232
Circle Reader Service Card No. 131
3O25 Sampling Manifold
The basics of sample preparation with common
LSC cocktails using various Millipore (liters.
The propertias of Millipore filters in protein and
nucleic acid binding applications. Also filtration for
microbiological sampling and separation, micro-
chemical analysis, find radio tracer ultracleaning.
In all, 32 pages of useful information. Free. Ask for
Manual AM304. Or call toll free 800-225-1380
(in Mass, call (617) 275-9200).
R/KLLEPGRE Corporation
Bedford, Mass. 01730
csa.
Circle Roador Servico Card No. -19
AMERICAN LABORATORY
G1
-------
CHROMATOGRAPHY / LIP1OS
GC ScparGfion of Solvent
& Determination of Water in Solvents
Introduction
This bulletin deals with columns useful for the complex mixtures of solvents. There are several columns which
may serve as a general purpose column and these may be supplemented by one or more special purpose columns. No
single packed column will be able to separate every possible solvent combination; consequently, more than one col-
umn may be needed to handle a variety of solvent problems.
General Purpose Column A
Carbopack C/0.1% SP-1000 is a good general purpose
column; a major feature is the ability to separate alcohols
without tailing of the peaks. It can be used to separate
complex mixtures of alcohols, ketones, esters, aromatics
and chlorinated hydrocarbons as illustrated in Figure 1 with
the separation of a 19 component mixture. The column
can be used isothermally as shown in Figure 2; however, if
a wide boiling range mixture is to be separated, this is done
better with temperature programming rather than isother-
mally. Table 1 lists the retention data for a variety cf
compounds.
Figure 1 — Solvent Mixture
1. Methyl alcohol
2. Ethyl alcohol
3. Acetone
4. Isopropyl alcohol
5. MEK
6. Isobutyl alcohol
7. Ethyl acetate
8. n-butyl alcohol
9. Isopropyl acetate
10. Cyclohexanone
11
12
11. MIBK
12. Isobutyl acetate
13. n-butyl acetate
14. Toluene
15. Butyl cellosolve
14
13
16. Cellosolve acetate
17. Ethyl berzene
18. m & p-xylene
19. o-xylene
15 16
17
19
l l I l I I I I I I I
9 10 11 12 13 14 15 16 17 18 19
Minutes
Carbopack C/0.1% SP-1000, 6 ft. x 2mm ID, Glass, Col. Temp.: 100-225°C S> 8°C/mln., Flow Rate: 20 ml/min.. Nitrogen 9 50
psl. Sample Size: 0.15 JUI. Detector: FID, Sensitivity: 32xlCT10AFS.
©Copyright 1975, Supnlco, Inc., Bellefonte, Pennsylvania 16823
-------
T»Me 1 - Metemtai Dele tmlmaml
Ateoheh
Methyl
Ethyl
••preeyl
tvpropyt
.iutyl
n-t>utyl
DlMMtM
CvdPMlvM
Methyl
Ethyl
Butyl
KMWM
Acetone
MEK
MIBK
Metrtyl O.de
CVdOttnVflDM
Ixwrwm
AMMM
Ethyl
iprooyl
rvwopyl
tally)
n -butyl
Celknoke
Methyl
Ethyl
MethyleoeC.ilw.de
l.l.l.-WicMo'OMh.ne
Aromilici
Bmrene
Toluene
m at p.Kylene
o-xvlene
AI.pK.te
o-hTtane
tvnctefie
n-nottre
rwdeckne
THF
CMf
2-nttrooroO*ne
C«l Tern,
175°C
0.1% Sf 1000
065
075
088
095
1.35
1.55
DmmpwM
1.3
2.1
Lite
0.9
US
1.71
6.25
3.4
Nol Elund
145
2.10
2.70
505
590
08
0.75
100
0.90
165
2.23
6.35
UN
Lete
2.85
65
168
Lilt
Lite
1.19
1.95
1.55
100»C
70% S»2tOOA
1.12
1J4
142
1.82
2.5
2.98
10
2.87
398
1Z35
138
2.1
4.56
6.35
11.1
43.9
2.33
2.95
386
543
68
11.14
1.18
1.55
1.68
30
3.25
5.79
10.54
12.1
2.4
: 4.0
7.12
13.0
24.3
-2.72
' 6.45
3.6
100°C
20» S».2401 A
1.1
1J5
1.4
1.68
2.15
2.5
1438
2.85
378
97
2.16
312
6.12
7.58
18.26
-
251
3.05
3.95
5.46
6.48
11.95
1.32
1.78
. 105
1 66
2.22
3.47
55
6.65
1.33
1.78
2.55
3.89
6.25
_
7.05
100°C
16VTHEEDA
3.0
3.35
M
489
6.46
7.85
NotEiund
12.12
1229
28.95
1.43
1.75
2.28
4.1
Not Etuted
NotEkited
1.3
1.3
1.69
1.81
2.25
6.48
-
1.15
1 33
1.41
191
2.65
33
0.79
0.8!
0.91
1.09
1.31
1.65
-
3.5
Figure 2 — Solvent Mixture
1. Mtthanol 6. THF
2. Ethanol 7. MEK
3. Mettiylene chloride 8. sec-butyl alcohol
4. Acetone 9. Isobutyl alcohol
5. Isopropyl alcohol 10. Ethyl acetate
10
10
Minutes
Carbopack C/0.1% SP-1000, 6 ft. x 2mm Glass. Col. Temp.: 65°C,
Inlet & Dot. 100°C, Flow Rate: 20 ml/min.. N2 ® 34 psl. Sample
Size: 0.1 JUI, Del. FID, 16 x lO'l 0 AFS.
General Purpose Column B
Our SP-2100, methyl silicone, makes a very good general
purpose column for a variety of solvents, with the compo-
nents eluted generally in order of their boiling points. Fig-
ure 3 shows the separation of an 18 component mixture of
alcohols, esters, ketones, aromatics, etc. In Table 1, we
list the retention time of these and a number of other com-
mon solvents. We have been able to solve 3 large number of
different solvent problems with this column.
Figure 3 — Solvent Mixture
in • 6 ft. * 2f*wn.pit other columnt -10 h. « 1/8" SS
1. Wethjnol
2. Einanol
3. Acetone
4. MEK
5. Lthyl acetjte
7. Methyl celloiotv* 13. n-butyt atcoftol
6. Isop
9. 2-fiitroproMne
10. MIBK
11. Itobutyl acetate
6. liohuly" jiconot 12. Toluene
14. Di.K«ton« alcohol
IS. Ethyl benzene
16. m & p-xyl«n«
1 7. CeHosolve acvlata
18. Butyl ceMosolvi & o-xy*l«n«
The Carbopack C column is best used with small sample
sizes and consequently a flame ionization detector rather
than a thermal conductivity detector should be used. Sam-
ple sizes of 0.1 to 0.5 (i\ are recommended, but with the
smaller volume preferred. If large sample sizes are used,
the column is overloaded and poor peak shapes and poor
separation results. Although with properly packed stainless
steel columns, 700 plates/foot can be obtained, the best col-
umn efficiencies are obtained using a glass column; however,
with stainless steel columns, there is some tailing of peaks.
The Carbopack C/0.1% SP-1000 has the ability to sepa-
rate isomers very readily. As a result, if complex mixtures
of hydrocarbons are to be separated, this is done very well.
However, if alcohols, ketones, etc. are present as well as
complex mixtures of aliphatic hydrocarbons, there will be
considerable overlap of the peaks. Upper column tempera-
ture limit is225°C. Carbopack C/0.1% SP-1000 can be used
interchangeably with Carbopack A/0.2% SP-1000. For more
details on the use of Carbopacks, see Bulletin 738.
12
17
18
12
20% SP-2100/0.1% Carbowax 1500 on 100/120 Supelcoport, 10ft.
x 1/8" SS, Col. Temp.: 100°C, Flow Rate: 20 ml/mln., Ni. Sam-
ple Size: 0.5 yn, Det. FID.
The 20% SP-2100 packing contains 0.1% Carbowax 1500
as a tail reducer to reduce the tailing of alcohols. The col-
umn generally does a good job with most alcohols, but the
methyl and ethyl do tail when present in low concentrations.
There are better columns for trace alcohols and these are
described elsewhere. The SP-2100 column is not well suited
for separation of complex mixtures of aliphatic solvents
-------
along with alcohols, ketones, etc. The hydrocarbons tend
to be eluted along with the other components,obscuring the
results, unless the aliphatics are either lower or higher boil-
ing than the other components. Because of the presence of
Carbowax 1500, the upper temperature limit of this packing
is limited to 175°C.
Hydrocarbon "Fingerprint"
The SP-2100 can be used as a special purpose column to
examine the composition of hydrocarbon mixtures such as
naphthas. One can obtain a "fingerprint" of the hydrocar-
bon mixture and compare it to other batches of the material.
as shown in Figure 4 which is a naphtha sample analyzed
with the 20% SP-2100 column. If better resolution of the
mixture is needed, a 10% SP-2100 is a good choice. For
details see Bulletin 743, Section 7.
Figure 4 — Naptha Sample
Minutes
20% SP-2100/0.1% Carbowax 1500 on 100/120 Supelcoport. 10 ft.
X 1/8" SS, Col. Temp.: 125°C. Flow Rate: 20 ml/mln., Nj, Sam-
ple Slzei 0.5/11. Dot. FID.
Figure 5 — Ketone Retarder Column
1. Isopropyl alcohol
2. Ethyl acetate
3. MEK
4. Toluene
5. n-propyl acetate
6. Isobutyl acetate
7. MIBK
8. Mesltyl oxide
Ketone Retarder
The SP-2401 is a stationary phase which has the ability to
elute ketones relatively later when compared to other sta-
tionary phases. As an example, SP-2401 elutes MEK after
ethyl acetate with good separation of the two, as shown in
Figure 5 along with several other compounds. In contrast
with SP-2100, this pair is barely separated with MEK eluted
first (Figure 3). Isophorone, a ketone, is eluted from the
column, but one must increase the column temperature to
elute it in a reasonable length of time. We recommend
SP-2401 as a special purpose phase rather than a general
purpose one. Our 20% SP-2401 packing also contains 0.1%
Carbowax-1500 as a tail reducer to improve alcohol peak
shapes. Upper temperature limit is 175°C. Retention data
for many common solvents is shown in Table 1.
Alcohol Retarder
Tetrahydroxyethylenediamine (THEED) is a stationary
phase which retards the elution of alcohols relative to other
classes of compounds. It also has the ability to elute hydro-
carbons very rapidly compared to other compounds boiling
in the same temperature range. As an example of this.
Figure 6 shows the rapid elution of heptane and octane,
then esters, aromatics and finally alcohols. The 15% THEED
is coated on 100/120 ChromosorbW AW. We do not recom-
mend the use of a silanized support with THEED because
of the wetting problem. Upper temperature limit for this
column is 135°C. Retention data for common solvents is
listed in Table 1.
Figure 6 — Alcohol Retarder
Heptane
Octane
Ethyl acetate
n-propyl acetate
Toluene
6.
7.
8.
9.
10.
Ethyl benzene
Xylene
Isopropyl alcohol
Ethyl alcohol
n-propyl alcohol
20% SP-2100/0.1% Carbowax 1500 on 100/120 Supelcoport, 10 ft.
X 1/8" SS, Col. Temp.: 125°C. Flow Rate: 20 ml/mln.. Nj. Sam-
ple SUe: 0.5 JZI, Det. FID.
15% THEEp on 100/120 Chromosorb W AW, 10 ft. x 1/8" SS, Col.
Temp.: 80°C, Flow Rate: 20 ml/mln., Nj. Sample Size: 0.5 £11,
Det. FID.
Chlorinated Solvents
The 20% SP-2100/0.1% Carbowax 1500 on Supelcoport
makes a good column for separation of complex mixtures
of chlorinated solvents or for determination of impurities in
a single chlorinated solvent. Figure 7 shows the separation
of a mixture of eleven components made with a 10 ft. x
1/8" SS column at 100°C.
SUPELCO, IMC.
Sup*lcof>»k
Phoo. (814) 35*2784
, P«nmyh.»ni» 1M2]
TWX 5ia87»36OO
-------
Figure 7 — Chlorinated Solvents
1. Methylene chloride
2. 1,1-dlchloroethana
3. Chloroform
4. 1,2-dichloroetMane
5. 1,1,1-trichloroethane
6. Carbon tetrachloride
7. Trichloroethylene
8. 1.1,2-trichlOfoethane
9. Perchloroethylene
10. 1,1.1,2-tetrachloroethane
11. 1,1,2,2-tetrachloroethane
20% SP-2100/0.1% Carbowax 1500 on 100/120 Supe.coport 10 ft.
x 1/8" SS, Col. Temp.: 100°C, Flow Rate: 20 ml/mln.. Nitrogen,
Oet. FID.
Packings
1-1820
1-1821
1-1822
1-1823
GP Carbopack C/0.1 % SP-1000, 25g $75
20% SP-2100/0.1 % Carbowax 1500 on
100/120 Supelcoport, 25g 50
20% SP-240170.1% Carbowax 1500 on
100/120 Supelcoport, 25g 50
15%THEEDon 100/120 Chromosorb
W AW, 25g 30
Packed Columns (except Cai bopack & Carbosieve)
1/4", 3/1 6" or 1/8" Stainless Steel $25 + $3.50/ft.
Packed Columns (Carbopack & Carbosieve)
1/4" or 3/16" Stainless Steel $30 + $15/ft.
1 /8" Stainless Steel $30 + $ 8/ft.
Carbopack/Glass Column
&e °Ur SpeC'a' 8'Pa^ 9laSS CO'Umn ^^^ 3nd OUr latest
GC Catalog.
DETERMINATION OF WATER IN SOLVENTS
Water can be rapidly determined in solvents by using a
Chromosorb 101 porous polymer, packed into a short metal
tube; Figure 8 shows 0.08% water in toluene. The work
shown here was carried out with a Fisher Model 2400 GC
equipped with a thermal conductivity detector. The
Chromosorb 101 column can be used specifically for the
water determination or used to separate some solvent mix-
tures.
The determination of water by GC is difficult because
the water peak will tail severely with many columns making
quantitative results very poor. The tailing of the water
peak can be caused by a variety of factors including the
column packing, the column tubing and the inlet of the GC
instrument. The various diatomite supports, even when AW-
DMCS treated, still cause tailing, whereas the porous poly-
mers, such as Chromosorb 101, are much more inert and
water does not tail on them.
Metal column tubing has also been blamed for the tailing
but this depends on both the quality of the tubing and the
concentration of water to be determined. Thus far, we
have tested our stainless steel tubing successfully down to
0.01% water. Some workers prefer either glass or Teflon
column tubing to avoid tailing of the water peak. The
overall column efficiency obtained with a glass column is
higher than obtained with stainless steel, but that obtained
with Teflon is rather poor. The metal inlet of an instrument
can also be highly adsorptive causing tailing. This can be
avoided by using a glass insert in the inlet or by injecting
the sample directly into the column, avoiding the inlet.
A calibration curve for water can be obtained by injecting
the same size sample of each of several standards containing
known amounts of water into the chromatograph. Plot
peak heights versus concentration of water as in Figure 9.
The calibration mixtures should be prepared with a solvent
Figure 8 - 0.08% Water
in Toluene
Toluene
Water
0.08%
Air
1
I)
70
. 60
T~*.
\f S°
I 40
S 30
.
20
10
I I I
024
Mln.
Figure 9 —
Peak heights vs.
concentration of
water
O .01 .02 .03 .04 .05 .06 .07
% water
Chromosorb 101, 60/80 mesh, 3 ft. x 1/8" SS, Col. Temp.: 165°C,
Flow Rate: 20 ml/'min., helium, Sample Size: 1 pi, Det. TC 0>
250MA, Sensitivity 2X, Recorder 1MV.
which is free of water. This can be done by drying the sol-
vent with a good drying agent such as Molecular Sieve 5A.
The syringe used to inject the sample should also be free of
water so as not to contaminate the standards or sample.
For details on the use of an internal standard procedure
for water, see Hogan et al. (3). Other useful references
dealing with the determination of water are 1, 2 and 4.
Packed Columns (except Carbopack or Carbosieve)
1/4", 3/16" or 1/8" Stainless Steel $25 + $3.50/ft.
1/4", 3/16" or 1/8" Teflon $40 + $4.00/ft.
Tubing
2-0526
2-0532
1/8" Stainless Steel, 50 ft. coil
1/8" Teflon, 50 ft. coil
Chromosorb 101
2-0213 60/80,50g
2-0214 80/100,50g
2-0215 100/120,50g
$80
40
$33
33
31
REFERENCES
1. Bennett, O. F., Anal. Cham. 36, 684 (1964).
2. Gvo/dovlch. T. N., Orlnberg, G. S., Zuyeva. L. V. and Yashin,
V. I., Petroleum Chemistry (USSR), I S, No. 2.120 (1972) English.
3. Hogan, J. M., Engel, R. A. and Stevenson, H. F., Anal. Clem. 42,
249 (1970).
4. Hollls.O. l_and Haye.W. W..J. Gat Chromotog,, 4,235 (1966).
-------
-a
m
Z
x
n
-------
APPENDIX - Cl
STUDY TO SUPPORT NEW SOURCE PERFORMANCE
STANDARDS FOR SOLVENT METAL CLEANING OPERATIONS
COMPARISON OF ALKALINE WASHING AND VAPOR DECREASING
Diesel Equipment Division CMC
Grand Rapids, Michigan
PREPARED BY:
K. S. Surprenant
The Dow Chemical Company
PREPARED FOR:
Emission Standards and Engineering Division
Office of Air Quality Planning
U. S. Environmental Protection Agency
-------
Summary
A cost and energy comparative study between alkaline washing
and vapor degreasing was made. The equipment involved a
spiral alkaline washer and a Vibra degreaser. The cost of
cleaning a ton of work was found to be the same as to 17%
less with vapor degreasing than alkaline washing. The
energy demand for the washer was just over twice that of
the degreaser.
Both alkaline washing and vapor degreasing satisfy unique
cleaning needs and are not interchangeable metal cleaning
processes. This fact was repeatedly brought out in seeking
a comparative evaluation testing site and during the
evaluation itself. The substitution of alkaline washing for
solvent metal cleaning to control hydrocarbon emissions would
be possible in relatively few operations.
-------
Objective
Alkaline washing is another means of metal cleaning. No
hydrocarbon emissions result from this cleaning method
because it depends on water detergent solutions rather
than organic solvents. The purpose of this evaluation is
to define the cost and energy relationships between alkaline
washing and solvent metal cleaning, especilly vapor degreasing,
In those instances where either cleaning process might be
used, this background information may be valuable in
evaluating the choice.
-------
Introduction
The qualitative parameters involved in choosing between
alkaline washing and solvent metal cleaning have been
outlined in the body of the main report. In most cases,
the cleaning requirement defines the most practical choice
between alkaline washing and vapor degreasing. Because
these cleaning processes satisfy different cleaning
needs, it is difficult to locate industrial applications
where both processes are employed to clean essentially
similar production.
This evaluation site was located through the assistance
of Detrex Chemical Industries Incorporated who manufacture
both alkaline washing and vapor degreasing equipment.
Potential sites were sought with several other firms
manufacturing equipment for both processes. However,
this location was the only recommendation received.
In addition to the performance characteristics of the
processes, there are other causes for selecting one or
both cleaning methods. Some of these reasons would
include: the availability of a surplus washer or degreaser
within the corporation, changes in the cost of energy,
equipment, or chemicals or a reassessment of the hazard
-------
of alkaline washing compound or chlorinated hydrocarbons.
The reasons for the existence of both cleaning processes
at Diesel Equipment were not explored but could have been
any of the above or other valid causes. Even though the
work processed by the washer and degreaser at this location
is largely similar, there are a large number of specific
cleaning operations which are done exclusively by one or
the other process. It is also valuable to note that this
manufacturing location utilizes a number of vapor degreasers
of several different design types as well as a number of
washers. In spite of this variety of equipment, it is quite
unusual for parts being cleaned in one operation to be
cleaned by the other cleaning process either routinely
or even alternatively.
-------
Equipment
Vapor Degreasing Operation - The vapor degreasing
operation studied was a Detrex Vibra degreaser Model
No. RV918-8-75 Special. This degreaser is 64" x 64" x 110"
high and uses trichloroethylene. The attached speci-
fication sheet'shows its diagramatic construction. The
parts are introduced through the loading chute to the
flooded bottom pan and are conveyed up the spiral by the
vibratory motion of the spiral. With the exception
of the vibratory conveying means, the degreaser operates
similarly to other vapor degreasing operations in
producing solvent vapors and condensing and controlling
them. This degreaser is equipped with two Detrex
Model S185 stills. The still design is shown on the
specification sheet also attached. Both stills were
located on a storage tank 65" x 78" x 33". The
degreaser is identified as Serial No. 54715 and the
stills are identified by Serial No.'s 54515A and 54515B.
Alkaline Washer - The spiral washer evaluated was supplied
by Ransohoff Company, Serial No. 10443 (Figure 1). The
Equipment No. is 36948. This washer has three stages:
wash, rinse and dry. The overall equipment dimensions
are 21' x 6' with a machine height of about 7'10".
-------
Further design information is listed on the attached
Ransohoff specification sheet. Parts are fed into this
machine through a loading chute to the inside of a
rotating drum with a spiral screw. The spiral screw
causes the parts to move through a spray washing station
to a spray rinse station and a drying zone before being
discharged. The drum is perforated to allow the
drainage of wash and rinse water as well as the flow
of hot air for drying. The operating conditions are
spelled out on the washer maintenance specification
sheet supplied by Diesel Equipment and attached to
this report.
Experimental Design - This evaluation differs from other
emission testing in that no emission control devices were
a portion of this study. Rather, an objective cost and
energy comparison between alkaline washing and vapor
degreasing was the purpose of this effort. For this
reason, no attempt was made to improve the operation
of either the alkaline washer or the vapor degreasing
system. However, in a number of ways both processes
studied demonstrated excellent operation.
Briefly, the concept involved in this evaluation was simply
to observe and measure each operation for a sufficient time
-------
to permit reliable estimates of their consumption of
chemicals and utilities and their production workloads.
Equipment cost data was provided by the original equipment
suppliers.
-------
Data Discussion
The method of comparison between these two cleaning processes
is on a cost per ton basis. The following data is organized
to develop the cost of each input expressed in dollars per
ton. Costs per unit of energy, chemical, etc. are selected
to represent national averages, not Diesel Equipment costs.
Work Processed
The work process record for the Vibra degreaser is summarized
below.
Vibra Degreaser Work Record
Date Gondolas Cleaned
9-16-75 9 (1 Shift Only)
9-17-75 9 (1 Shift Only)
9-18-75 34
9-19-75 11 (1 Shift Only)
9-23-75 17 (1 Shift Only)
9-24-75 16 (1 Shift Only)
9-25-75 29
9-26-75 32
9-27-75 36
9-29-75 27
9-30-75 28
10-2-75 34
10-3-75 22
10-4-75 34
A similar record for the spiral washer follows:
-------
Washer Work Record
Date Gondolas Cleaned
8-5-75 21
8-6-75 21
8-7-75 11 (1 Shift Only)
8-8-75 13
8-9-75 16
8-10-75 20 1/2
8-11-75 24 1/2
8-12-75 25
8-13-75 . 25 1/4
8-14-75 10 3/4
In determining the average work processed, only those days
having production on two shifts were used. The plant schedule
during both operations was two shifts per day, six days per
week. The average number of gondolas per day was 19.8 for
the washer or 5940 gondolas per year. The degreaser averaged
30.7 gondolas/day or 9210/years. The weight of a gondola load
was found to be essentially one ton (1980 Ibs.). In the
calculation of costs per ton, the gondolas were assumed to
weigh a ton each.
Capital Investment - The current prices for each
of the process equipment were obtained from the
original suppliers. An installation cost of 15%
of the equipment price was assumed. The building
capital cost was charged on 1 1/2 x the square foot
area occupied by each process x $25.9 per square foot.
This information is summarized in Tables 1 and 2.
-------
The capital linear Annual Payment for capital is
calculated by:
LAP = 1(1 + I)n P
(1 + I)n -1
Where I = Interest Rate
n = Equipment or Building Life
P = Borrowed Principle.
The function 1(1 + I)n ..
— equals:
(1 + I)n -1
0.131476 for 15 yrs.
0.11746 for 20 yrs.
0.110168 for 25 yrs.
when the interest rate is 10%.
Using the equipment and installation cost of $61,502
for the Vibra degreaser and a 20 year life factor,
the annual cost for equipment is $7224. The cost per
ton is obtained by dividing that value by 9210 tons
per year or $0.784 per ton. The total equipment
capital for the alkaline washer ($42,320) x the factor
for a 20 year equipment life, yields a value of $4971
per year. At 5940 tons per year the cost per ton is
$0.837 per ton. The building capital is calculated at
a 25 year life and amounts to $0.030 for the degreaser
and $0.091 for the washer on a tonnage basis.
-------
Insurance - The insurance costs for both processes
was figured at 2% of the total equipment and building
capital, without the installation costs. The capital
costs ($55,951 for the degreaser and $41,695 for the
washer) was multiplied by 0.02 and divided by the tons
processed per year. Thus, the interest cost per ton
was found to be $0.122 per ton for the degreaser and
$0.140 per ton for the washer.
Maintenance - The Maintenance Department estimated
the montly cost for the degreaser at $715. When this
number is multiplied by 12 to obtain the cost per year
and divided by 9210 tons be year cleaned, a value of
$0.932 per ton is obtained for the degreaser. The
maintenance cost for the washer was based on 100 man-
hours per year. Assuming a $7.00 per man-hour, an annual
cost of $700 per year is obtained. This divided by the
5940 tons of work cleaned yields $0.118 per ton.
The maintenance costs as stated above represent 16% and
2% of the equipment prices per year respectively for
vapor degreasing and alkaline washing. Although the
vibratory conveyor means can be expected to cause
more wear than is experienced in other metal cleaning
-------
equipment, 16% of the original equipment price per
year is regarded as a high estimate. Based on conversations
with Detrex and the operational schedule for degreaser
and still clean-outs, the following maintenance estimate
was prepared.
MAINTENANCE SCHEDULE
Still Operation 4 hrs./50 x 50 wks./yr x $7.00/hr.
— $1400
Degreaser Clean-out (2 Per Yr.)
2 men x 3-8 hr. shifts x 2/yr. x $7.00/hr.
— $ 672
Bearing Replacements (2 Per Yr.)
2 men x 2-8 hr. shifts x 2/yr. x $7.00/hr.
— $ 448
Bearing Cost ($200 ea.)
$ 400
Replacement of Spiral and 2 Motors Every 6 Yrs.
Spiral $7500
2 Motors $4000
$11,500
Installation (15%) $ 1,725
$13,225 - 6 yrs.—$2204
$5124
This maintenance estimate would be excessive for most
Vibra degreaser operations. However, the work process
load handled by the Diesel Equipment Vibra degreasers
-------
is extraordinarily high and requires extra maintenance
activity. The $5124 estimated is reasonably close to
10% ($5348) of the original purchase price. If 10%
of the original purchase price is taken as the main-
tenance cost per year, the extra dollars would provide
for the replacement of small components of the degreaser
system. This value divided by the 9210 tons per year
equals $0.581 as a cost of maintenance per ton of
product.
Chemicals - The Vibra degreaser was studied for three
weeks beginning September 15, 1975. No solvent was
added to the system during this interval. The storage
tank (78" x 65") had an initial level of 25 5/8 inches
and a final level of 17 7/8 inches. Each inch within
the storage tank was equivalent to 21.95 gallons. Thus,
the 7 3/4 inches difference in level is equal to 170
gallons. This level could be determined within 1/4
of an inch or +_ 5.5 gallons. The degreaser level at
the start of the trial was 11 1/8 inches and at the
end was 9 7/8 inches. This tank measures 64 inches
square and 1 inch of elevation is equal to 17.73 gallons.
The 1 1/4 inch level change in the degreaser equals 22
gallons of solvent consumed. An error of +_ 1/4 inch is
equal to +_ 4.4 gallons. The total consumption for the
-------
three weeks was 192 gallons + 10 gallons. The degreaser
was operated two shifts per day and scheduled six days
per week. Only five days of operation occurred in the
third week, however, this difference is corrected by
expressing all costs on a tonnage of product cleaned
basis. When the 192 gallons used is divided by 17 work-
days, 11.3 gallons consumption per workday is found.
When this is multiplied by $2.15 per gallon and divided
by 30.7 tons per day, a cost per ton of $0.791 is
obtained.
The Ransohoff washer used Detrex 75LN detergent solution.
This product weighs 10.6 pounds per gallon and has a
value of $.26 per pound according to Detrex. When the
alkaline washer was filled initially on August 5, 20
gallons of 75LN were added to the wash tank and 6
gallons to the rinse tank. A total of 39 gallons were
added during the course of the next week to maintain
the cleaning solution concentrations. The 75LN drum
was removed from the operation between additions to
prevent use for other purposes. Although the clean-out
schedule was weekly, the washer was being cleaned on a
two week basis. The second week consumption should be
the same as the first week with the exception of the
initial fill chemical requirements. Therefore, 104
gallons would be consumed in two weeks for an average
-------
of 52 gallons per week on a six day work week. The
average consumption per day is 8.67 gallons or 91.9
pounds. A cost per ton of $1.206 is obtained by
multiplying by $0.26 per pound and dividing by 19.8
tons per day.
Steam - The steam condensate from both stills and the
degreaser were plumbed to a condensate collection tank
and pumped to the boiler. To determine the steam
condensate use a water meter was installed in the boiler
return line. This meter read 686590 gallons on
September 15, and 700229 on October 6, for a total of
13,639 gallons of steam condensate used. The degreaser
is heated only during operating shifts. Thus, 802
gallons were used per day for the 17 operating days.
A value of $0.493 per ton as a cost for steam is
obtained by using $2.30 per thousand pounds of steam
condensate, 8.2 pounds per gallon and 30.7 tons per
day.
The washer steam condensate was collected four times.
The average of these readings was 554 pounds of steam
per hour. The cost of steam per ton was calculated as
follows:
-------
554 lbs./hr. x 16 hrs./day x $2.3/1000 Ibs. x 1/19.8 tons/day,
By this means, a value of $1.030 per ton is identified
as the cost of steam. This value appears to be quite
low when compared to the equipment engineering steam
consumption specification of 1620 pounds per hour or
a verbal report from Ransohoff that this washer would
be estimated to consume 1000 pounds of steam per hour.
Consequently, the actual steam consumption could be
2-3 times higher than that determined. A measurement
of the steam condensate required for the dryer section
alone was found to be 260 pounds per hour. Another
significant difference between the washer and the
degreaser is the pressure of the steam used. The
degreaser uses 15 psig steam whereas the washer uses
100 psig steam.
Electric Power - The electric power demand for the
Detrex Vibra degreaser was found to be 10 amperes without
work being processed and 11 amperes with work. The empty
power demand for the Ransohoff washer was 17 amperes and
18 amperes were found with work being processed. In
both cases, the current voltage was 460 volts. The
average ampere demand was used along with the formula
below.
-------
Amperes x Volts x 1.73 x Power Factor
1000 = Kilowatt Hours
Amperes x 460 x 1.73 x 0.85
= Kilowatt Hours
Using this formula, the kilowatt hours consumed by
the degreaser were 7.10 KWH while that consumed by
the washer was 11.84 KWH. The cost per ton calculated
using 16 hours per day, $0.025 per kilowatt hour and
the tons per day by each process. The electric power
cost per ton was $0.093 for the degreaser and $0.239
for the washer.
Water - The water consumption for the Vibra degreaser
and two stills was measured by water meters. The
total measured flow for the 21 days of testing was
1,355,540 gallons. This consumption rate is approximately
three times higher than that needed based on the equipment
specifications. A water consumption rate of 2597 gallons
per ton is obtained by dividing by 17 workdays and 30.7
tons per day. Using $0.04 per 1000 gallons, a water
cost of $0.104 per ton is obtained.
A water meter on the Ransohoff washer showed a consumption
of 4769 gallons per week excluding 1190 gallons to fill
-------
the equipment. The water requirement for two weeks of
operation including the initial fill was 10,728 gallons.
Dividing by 12 workdays and 19.8 tons per day, 45.2
gallons per ton was used. This equates to $0.002 per
ton.
Waste Water Treatment - No waste water treatment charge
is assigned to the degreaser system because this water
can be immediately recycled through a cooling tower for
reuse or can be used in other plant water requirements.
The gallonage of waste water from the Ransohoff washer
was determined by a water meter to be 1391 gallons per week,
At the end of two weeks, the washer capacity (1190 gallons)
would be discharged to the Waste Water Treatment Plant.
Thus, the total quantity of waste water requiring treatment
per two weeks is estimated at 3972 gallons. Using 12
workdays per two weeks and 19.8 tons per day, 16.7
gallons per ton of water must be processed by the waste
water treatment plant. A cost of $0.100 per ton is
obtained as the water treatment cost when a value of
$6.00 per thousand gallons is taken.
Solvent Disposal - The quantity of residue obtained
from the stills was measured and sampled. This infor-
mation is summarized in the table following.
-------
Date
9-18
9-22
9-25
9-29
10-1
10-6
Non-Volatile
60.0
51.1-
81.5
87.6
88.2
84.0
89.8
86.3
88.3
80.9
STILL RESIDUES
Weight %
Gals. Res. Wt. of Residue
100 965 Lbs.
(42.5) 414 Lbs.
40 496 Lbs.
55 532 Lbs.
100 726 Lbs.
338 Gals. 3133 Lbs.
-------
The gallons (inserted for September 22, 1975) were
calculated based on the average density of the other
residues. All other results are as reported. The
gallons of waste for disposal per ton of work processed
was 0.648. The cost per ton was $0.130 using a disposal
cost of $0.20 per gallon.
•
Assuming that the volatile portion of each sample is
trichloroethylene, the quantity of trichloroethylene
in the residues from the still amounted to about 60
gallons or about 31% of the total solvent consumed
during the evaluation.
Make-up Air Heat Requirement - The Detrex Vibra degreaser
required no ventilation. Thus/ no charge is assessed to
heating plant make-up air for this process.
The high water content of the exhaust ventilation from
the washer prevented the actual measurement of the
ventilation rate. The Ransohoff specification indicates
a ventilation rate of 2240 cfm. The exhaust temperature
was found to be 182°F. For estimating purposes, the mean
temperature during the four cold months of the year in
Michigan is estimated at 30°F. The expansion of the air
due to heating within the washer decreases the air
make-up to approximately 1700 cfm.
-------
The heat requirement for this air is estimated by
1700 cfm x 0.0808 lb. x 0.25 Btu x 40°F x 60 mins.
FtT7 lb.-°F hr.
= 82416 Btu/hr.
On a 16 hour workday, 1,318,656 Btu's per day would be
needed. A daily cost of $1.648 is found using $1.25 per
million Btu's for natural gas heating. However, this
heating requirement would be needed only about four months
per year. Annualized, the cost would amount to $0.550
per day or $0.028 per ton of work.
-------
Two cases for an operating cost comparison are offered in
this report. The first (Figure ) is a direct summation of
the cost developed in the preceding sub-paragraphs. The
second (Figure ) develops slightly different costs which
present the case least favorable, to vapor degreasing. In the
least favorable case, the equipment life is recalculated on
a 15 year basis. In addition, the maintenance cost is
based on $715 per month or 16% of the equipment purchase
price as discussed in the maintenance sub-paragraph earlier.
The waste water treatment plant and the cost of heating
make-up air is treated as a general overhead in the second
case rather than assigned to the operation requiring these
services and no charge is assessed.
No cost of direct labor was identified for either process.
However, the material handling operations before and after
both the spiral washing operation and the Vibra degreasing
operation are identical and the direct labor should be
exactly the same on a cost per ton basis. The differences
in maintenance labor are charged appropriately to each process
-------
Conclusions
1. In the realistic case (Figure ), the cost comparison
between alkaline washing and vapor degreasing shows
a savings of 17% by vapor degreasing.
2. The energy requirement of alkaline washing was
found to be approximately twice as large as vapor
degreasing. A number of literature sources
indicate the energy demand of alkaline washing
to be six to eight times that of vapor degreasing.
3. Alkaline washing equipment costs have been reported
to be higher than comparable vapor degreasing
equipment in earlier studies. Although the results
may be specific to the case studied, the vapor
degreasing equipment costs were 45% higher than
the alkaline washing equipment.
4. In the cost comparison least favorable to vapor
degreasing (Figure ), alkaline washing and vapor
degreasing are about equal in cost per ton of
work processed.
5. Solvent consumption per ton of work was 0.37 gallons.
This level of consumption is exceptionally low.
-------
High production loads, good operating practices,
free draining parts and the confinement provided
by the degreasing equipment make this possible
in this specific operation. However, this level
could not be achieved in most vapor degreasing
operations.
6. Of the total solvent consumed, 31% could be accounted
for in the still residues in spite of good distillation
practices. The oil content of still residues exceeded
80% by weight. The quantity of solvent remaining
in the residues could not be recovered by any of
the standard emission control equipment such as
carbon adsorption.
-------
IM
r. 11/72
UTHO IN U.VA
3/1-5
SPECIFICATIONS--VIBRA DEGREASERS
STD.
SIZE
1
2
3
4
5
6
7
8
MODEL
NO.
RV118-40-15-0
RV218-40-11-7
RV218-50-21-0
RV312-50-15-7
RV312-60-25-0
RV312-60-21-7
RV918-80-35-0
RV918-80-30-7
OVERALL
DIMENSIONS
5'-0"x4'-9"x5'-8"
5'-0"x4'-9"x5'-8"
51-0"x4'-9"x6'-0"
5'-0" x 4'-9" x 6'-0"
5'-10"x4'-9"x6'-7"
51-10"x4'-9"x6'-7"
6'-10"x5'-7"x9'-0"
6l-10"x5'-7"x9'-0"
LOAD
HGT.
42"
42"
46"
46"
46"
46"
63"
63"
UNLOAD
HEIGHT
34"
34"
39"
39"
41"
41"
54"
54"
PROD. LBS
STEEL/HR
1200
1200
2400
2400
4500
4500
7500
7500
SOLVENT
CAP. GAL
40
40
40
40
60
60
95
95
•STEAM LBS/HR
@ 15 psi
100
100
150
150
210
210
325
325
•WATER GALS/HR
@ 50°F RISE
200
200
305
305
430
430
665
665
•DILUTION
GALS/HR
50
50
60
60
75
75
100
100
APPROX.
WEIGHT
1500 Ib
1600 Ib
3300 Ib
3450 Ib
3800 Ib
3950 Ib
5500 Ib
5050 Ib
STILL
SIZE
S60 _j
S60
S185
S185
S185
S185
S185
S185
"COMBINED
BOILER HP
7
7
15
15
16
16
20
20
NOTE: 'Degreaser Only. "Degreaser and Still.
Standard Vibra Units can be supplied with a tote pan
cleaning or basketed work section.
Submit sample or drawing of tote pan or basket and
advise number per hour to be cleaned. This section
can also be used for occasional work load of work not
feasable to process thru Vibra Unit.
Optional Model S225 recommended where soils that tend
to cause "foaming" upon distillation are present.
IMMER-
SION
DIP
WASHBACK
RINSE
. VA
an
45'-6" of TRAVEL-
VAPOR CLEAN
and DRY-OFF
•17'-8"-
DRY and UNLOAD
_, -
90° 180°
0
\
2 3 4 5 6 7 8 9 10 11 12
REVOLUTIONS (WORK)
Work Cleaning Cycle
Schematic Arrangement of
Vibra Degreaser
RUBBER DRIVE MOUNTS
STEAM TO SPIRAL
STEAM JACKETS
LOAD CHUTE
COPPER FINNED
CONDENSATE COIL
CONDENSATE TROUGH
SOLVENT LEVEL
STEAM COIL
HOOD
DUAL MOTOR VIBRATORY DRIVE
SPIRAL ELEVATOR
UNLOAD CHUTE
STEAM JACKET (Optional)
VAPOR LEVEL
FREEBOARD COOLER
SOLVENT DISTILLATE WASHBACK
ENTERS SPIRAL TRACK
SOLVENT DISTILLATE ENTERS
BOTTOM PAN IMMERSION
WATER SEPARATOR
STILL DISTILLATE
RETURN (Optional)
BOTTOM PAN
VIBRA DEGREASERS
-------
IM 9. 17
l"NO M UAA
9/68
SPECIFICATIONS - DETREX MODEL S185 and 5225 STILLS for HALOGENATED SOLVENTS
SOLVENT
Trichloroethylene
(Perm-A-ClorNAi
Perchloroethylene
(Perk)
1,1,1-Trichloroethane
(Perm-Ethane)
Trichlorotrifluoroethane
(Detron "0")
Methylene Chloride
•DISTILLATION RATE gph
Model S185
185
175
175
70"
35"
Model S225
225
200
185
70"
35"
STEAM REQ'O. LB/HR.
Model S185
275@15psig
265 § 50psig
225 @ 7psig
70@ 3psig
65@3psig
Model S225
3306 15psig
300@50psig
235§7psig
70 § 3psig
65 6 3psig
•WATER q
Model SI85
620
565
525
400
710
ph MAX.
Model S225
750
640
550
400
710
Overall Dimensions
Distillate Discharge Height
Weight - Approx.
Liquid Capacity - Gal.
Pump Capacity
MODEL S 185
Length 4'-10" Width 3'-8"
Std. Height: 7' -3"
Opt. Height: 9' -3"
Std. Height 4' -2"
Opt. Height: 6' -2"
1.500 Ibs.
115
10gpmei5ft.
MODEL S225
Length 4' -10" Width 3'-8"
Std. Height: 7' -3"
Opt. Height: 9' -3"
Std. Height: 4' -2"
Opt. Height: 6' -2"
1,600 Ibs.
90
lOgpme 15ft. Hd.
Current Characteristics S185& S225
110/220/1/60 - 1/3 hp, 1750 rpm TEBB
230/460/3/60 - 1/4 hp, TEBB
•Based on 60°F water inlet 50°F rise - Trichloroethylene, Perchloroethylene and 1,1,1-Trichloroethane:
10°F rise for Methylene Chloride, 20°F rise for Trichlorotrifluoroethane.
•Distillation rate can be increased by supplying chilled (40°F) water.
250
200
150
a.
o>
100
50
m
:prr AVERAGE DISTILLATION RATE :
;p_" OF TRICHLOROETHYLENE £
T^Tj' vs. Boiling Point Solvent-Oil
[-Mixture in Lower Sump® 15 psig :
D • Increase in Distillation
Rate S 225 over S185 -40-60 gphi-
190 200 210 220
Boil Chamber Temp. F°
DISTILLATION RATE CURVES
4'-10"O.A. «• 48" COIL REMOVAL
4'-0" I.S.
STEAM PRESSURE
REDUCER
SAFETY VALVE
3'-6"
APPX.
STEAM IN
15'-6" Opt.)
Model S185 Detrex Solvent Still
STILLS
-------
ENGINEERING SPECIFICATIONS
RANSOHOFF COMPANY
HAMILTON, OHIO
DESIGN , ,
Washer Si/e (l.ulh. * Wiillh) fj»_.^.£—*_.JL
Washer Si/i: (l.ulh. «. Wiil
Washer Typ,-4>TM!"
Description
No. of Slagps—
Continuous
T
Parts/hour
45 CU. ft.
REC1RCULATING SYSTEM .
Pump Manufaclun-i-...y.^.mi!?.9
Pump Type Vortical
Pump Capacity (C..P
Kinso
EXHAUST SYSTEM
Fan M;miif;n:tiir.-|AmerJ_caj!_ F_a_n_
Fiin C;ip.icily (CFM K, Kl») 2?40 P 3/4
INSULATION
Insulation M
Insula'ion Tj
TANK I IKA
Steam (ia
Type 1
TEMPERATURE CONTROLS
Trerice
-..:<;
fr
BLOW-OFF SYSTEM
lll.mrr
Typ.
9100
900
MK.uvr t:.ip.i..ily (CFM K SP^80.JO61
- 'i-H_e.atjed
SPRAY NOZZLES
Ncizzlo MiiniifaiiliiriM- Ransohoff
Nozzle Typo. ' deflector
PRKSSURK RKGULATOR
Manul.icliircr
Type
(iiipiicily.
No. of Spray Nozzles
30-
CONSTRUCTION
W.tsher I lousing pi.mSi r _... --
Washer" Tanks ((laujje) .. _J._/.i" ..H_r
Tank Capacilx1 (tiallons).
W'ash_ '6/0
Rinse
MS
TRAPS
Manufacturer ^a^ay
Type FIT!
Capai-.ily ..... ._ . ... 4860 _.-.
520
SERVICE REQUIREMENTS
WATER
Initial Filling
I'.on sumption
1 190
TZT.TVHT
HORSEPOWER U-'.ACII M()Tf)K|
Wash, I'ump 5 '
Kinso I'ump
AIR
guuntily (CFM)
PSI
STEAM
Consumption
1620
Com'eyor Drive-
Drum Drive
F.xhausl Fan
•
Ulow-off System
Other
7.5
Ilis. /hour
CONDENSATE
1 620
T.ilal III'
Consumption..-. __
.. CFM
-------
Investment Capital
Vibra Degreaser Washer
Equipment $53,480 $36,800
Installation (15%) 8,022 5,520
$61,502 $42,320
Building $ 2,471 $ 4,895
Total $63,973 $47,215
-------
TABLE I
Industrial Building*
Shell (M&L) Cost
Lighting and Electrical
Heating and Ventilating
Plumbing
Fire Prevention
$ 4.09/Ft.'
1.75
1.50
1.70
1.10
Sub-Contract Cost (1.3)
Contingency (15%)
$10.14 (1968 Base)
$17.3/Ft.2 (8%/Annum
1.71 Multiple in 1975)
$22.5/Ft.~
$25.9/Ft/
*Derived from "Modern Cost-Engineering Techniques by
H. Popper. The 8% inflation rate was estimated by
the author.
-------
WASHER MAINTENANCE
TYPE WASHER
LQCATIOM
••
•TYPE CLZAKBR
ft/SHsartort: £sr#e* fie***? t2
7<5* / &
WASHER INFORMATION;
1. OPER. TAMP.
2. TANK CAPACITY
3. CLEANER CONCENTRATION
(A) NEW CHARGE
(B) ADDITIONS
4. CHANGE PERIOD^
NOTES:
WASH TANK
76 &U.
v^«^«
M
PEP NO , !
BAY
RINSE TANK
* X
-------
Equipment
Building
Insurance
Maintenance
Chemicals
Steam
Electric Power
Water
Waste Water
Solvent Waste Disposal
Make-Up Air Heat
Figure
Case
g Cost
st Per
1
Comparison
Ton)
Vibra Degreaser
$0.784
0.030
0.122
0.581
0.791
0.493
0.093
0.104
0.130
Washer
$0.833
0.091
0.140
0.118
1.206
1.030
0.239
0.002
0.100
0.028
$3.128
$3.787
-------
Figure
Case 2
Operating Cost Comparison
(Cost Per Ton)
Equipment
Building
Insurance
Maintenance
Chemicals
Steam
Electric Power
Water
Waste Water
Solvent Waste Disposal
Make-Up Air Heat
Vibra Degreaser
$0.878
0.030
0.122
0.932
0.791
0.493
0.093
0.104
0.130
Washer
$0.837
0.091
0.140
0.118
1.206
1.030
0.239
0.002
$3.573
$3.663
-------
TJ
TJ
m
n
K)
-------
APPENDIX - C2
STUDY TO SUPPORT NEW SOURCE PERFORMANCE
STANDARDS FOR SOLVENT METAL CLEANING OPERATIONS
Evaluation of Vapor Degreaser Covers
Eaton Corporation
Saginaw, Michigan
Prepared By:
K. S. Surprenant
The Dow Chemical Company
Prepared For:
Emission Standards and Engineering Division
Office of Air Quality Planning
U. S. Environmental Protection Agency
-------
-2-
INTHODUCTION
Chlorothene NU (inhibited 1,1,1-trichloroethane) was developed as
an improved grade of Chlorothene, more easily recovered by
distillation and exhibiting much improved corrosion characteristics
Due to the similarity in the conditions of distillation and vapor
degreasing, it was thought that this solvent might find further
markets in this application. In fact, -several companies have
successfully used this solvent in vapor degreasing for some time.
The laboratory vapor degreasing studies made on Chlorothene NU
showed sufficient success to encourage field evaluations. For the
purpose of these field tests, Chlorothene NU is identified as
MC154 to prevent damage to present Chlorothene markets in the
event of failure and to conceal from competition the entry of this
solvent into the degreasing market.
The foremost purpose of this test was to determine the relative
operating costs of trichloroethylene and MC15^+. In an effort to
insure our obtaining this information, the test site conditions
were chosen to reduce the probability of solvent decomposition.
This test was designed to provide data on the stability of the
solvent and stabilizer system, the relative toxicity and cleaning
performance, and to define the equipment changes necessary to use
this solvent in a degreaser designed to use trichloroethylene.
PROCEDURES AND EQUIPMENT
The equipment used in this test consisted of a Detrex open-top
degreaser with only one compartment- a boiling chamber. The
accessory equipment consisted of a water separator, storage tank,
and spray equipment. This unit was heated electrically and held
about two drums of solvent. The degreaser had developed
considerable rust on the walls.
All the parts processed by this equipment were steel. These
parts were given a heat treatment and oil quench Just prior to
-------
-3-
cleaning. Infra-red analysis of the quench oil revealed it to be
mostly aliphatic hydrocarbon with 5-10$ aromatic hydrocarbon.
In general, the procedures normally practiced by this concern
were not changed so that the cost information developed would not
be prejudiced whether these techniques were good or bad. Their
cleaning cycle waa vapor-spray-vapor and the machine was operated
three shifts each day, five days per week. On Monday mornings
the equipment was boiled down, the residues removed, and clean
solvent added. The only change in their operating practice was
the operation of the degreaser with one half of their electrical
coils. Pour coils were used for trichloroethylene while only two
coils were used with MC15^. this was necessary to contain the
vapors. Trichloroethylene and MC15^ were operated in the degreaser
for periods of three to four weeks with the top left open all the
time in each case and with the top closed during the down time to
determine the savings resultant from simply closing the top. It
was their normal practice to leave it open.
All of the solvent added to or taken from the machine was weighed -
including the samples and residues. A record of the parts processed
was also kept. This data was used to prepare the cost comparison
developed in this report. Samples were taken daily during the MC15^
phases of the operation for analysis of the solvent and stabilizer
system. The analytical procedures used are enclosed in Report
AA-1370, "Chlorothene NU Laboratory Vapor Degreasing Evaluations."
DISCUSSION:
Economics
There were two aspects of this cost study - one, the relative costs
of heating, the other a comparison of the consumption of the solvents
All other costs either were the same' or were assumed to be the same.
As mentioned, four electrical heating elements were required for
trichloroethylene while only two were required for MC154. Each of
-------
-4-
these elements operated on a 460 volt circuit and drew about 10
/
amperes. They are wired in parallel. Since the difference between
operating with the two solvents was two elements/ the power these
would have consumed is a measure of the heating savings experienced
by operating with MC154.
2 elements (460 volts)(10 amperes)(24 hrs) _ 220.8 KWH
1000 day
220 KWH ($.0132/KWH) = $2-90 saved
or about $710.00 per year
Tables I through IV are the records of the solvent consumed, hours
of operation, and the pounds of parts processed. Again, it should
be noted that the tables Identified as "degreaser top closed" refers
to closing the top only during down periods. In order to cancel
out the differences in the amounts of solvent remaining in the
residues after bolldown, nonvolatile content determinations were
made on each sample. The volatile matter was assumed to be
solvent and was subtracted from that added to the equipment to
obtain the net solvent consumption. This technique assumes that
both solvents can be recovered equally well from the oil; this
should be approximately correct. In the case of trichloroethylene,
only the solvent in the residues and the amount remaining in the
machine at the completion of the test were subtracted from the
total added because no samples were taken. In calculating the
consumption of MC154, the estimated weight of the samples taken was
also subtracted to get the net consumption.
On Table V the four phases of this test are summarized. It was
interesting to note that closing the degreaser top in down periods
saved between 30-31 pounds of solvnet per ton of work and that
under both operating conditions MC154 used from 18-19 pounds of
solvent per ton of work less than trichloroethylene.
-------
-5-
Toxiclty
Table VI shows the results of the Halide Meter Surveys run on
trichloroethylene and MC15^ under similar conditions. The
rectangles represent the surface of the open top areas of the
degreasers and the numbers are the parts per million (ppm)
concentration of the vapors found at each location. Note: The
American Conference of Governmental Industrial Hygenist suggest
MAC values of 100 ppm for trichloroethylene and 500 ppm for
MC15^. While some of the readings were above the recommended
values, these were all taken Just above the lip of the degreaser
and not in the air which the operators would be inhaling. The
values obtained for the air the operator would be breathing were
safe for both solvents: ^5 ppm for trichloroethylene and 50 ppm
for
Cleaning Performance
The general opinion of the operation personnel on the cleanliness
of the parts was that it was equivalent to that of trichloroethylene
No more detailed study of this attribute was made.
Stability
Boiling Sump Samples:
On Figure I the graphs of Acidity, Specific Gravity, and the
Stabilizers are plotted. The stabilizer concentrations are
expressed as percent by volume of the total sample and not
corrected for the oil content. The encircled points
correspond to the boildown residue samples. It should be
noted that all three stabilizers maintained a satisfactory
concentration throughout the test period except for these
residue samples. However, the condition represented by the
residue samples is prevalent in the degreaser only briefly
once each week. In addition, high concentrations of oils
inhibit some of the solvent decomposition reactions e.g. the
aluminum reaction.
The graph plotting the acid concentrations shows the residue
samples to be somewhat high in acidity. The oil in these
-------
-6-
samples interfered with the titratlon and the silver
nitrate qualitative test showed little or no chloride.
Therefore, the acidity fourd la not due to HC1 and not a
result of solvent decomposition.
Warm Dip Samples:
Acidity, Chloride, and the stabilizer concentrations are
plotted on Figure II. These four graphs demonstrate the
excellent "stability of the solvent under the conditions of
this test.
It should be noted that much of the variation in the
stabilizer concentrations of both warm dip and boiling
sump samples is due to the gradual increase of soils in
the boiling sump. This increase in oil concentration
reduces the specific gravity and increases the boiling
temperature of the boiling sump. The distribution of the
concentrations of the stabilizers between the warm dip
and the boiling sump is a function of the temperature of
the boiling sump
In this instance, the boildown cycle was once each week
and this cycle Is observable in the stabilizer concentrations
of both sets of samples.
CONCLUSIONS:.
1. $2.90 per day ($710 per year based on 2^5 days/year) was
saved In the cost of heating the degreaser by using MC151*.
2. When the equipment was operated with the top open all the
time, MC15^ used 18 pounds of solvent less than trichloro-
ethylene per ton of work, or saved $1.^ per ton of work
processed (using Dow list prices).
-------
-7-
3. When the equipment was operated with the top closed in down
periods, MC152* used 19 pounds of solvent less than trichloro-
ethylene per ton of work, or saved $1.8 per ton.
4. Comparing trlchloroethylene operations with open and closed top,
30 pounds solvent per ton less was consumed when the top was
closed, or a savings of $3.7 dollars per ton.
5. Comparing MC15^ operations with open and closed top, 31 pounds
solvent , or $4.1, was saved by closing the top.
6. Neither solvent presented a toxicity hazard under normal
operating conditions. However, in an accident (e.g. as spill),
MC15^ would provide a greater degree of safety.
7. The cleaning performance of MC154 was considered equal to that
of trichloroethylene.
8. The solvent stability was excellent under the operating
conditions found in this test site.
RECOMMENDATIONS.:
1. Field testing of MC15^ should be continued and the conditions
of operation expanded with regard to metals processed and soils
cleaned.
2. Further field tests should be found in other types of
equipment to further document the cost of operating with MC15^
as compared to that of trichloroethylene.
-------
TABLE I
Consumption of Trichloroethylene
(Degreaser Top Closed in Down Time)
Date
2-6-61
2-7-61
2-8
2-9
2-10
2-13
2-14
2-15
2-16
2-20
2-21
2-22
2-23
2-27
Tri
Added (Ibs)
1152
410
470
Boildown
Residues (Ibs)
127
125
112
% tri in
Residues
27
61
7
Wt. of tri in
Residues (Ibs)
34
76
8
Operating
Time (hrs)
20
24
24
24
20
20
24
24
21
19
24
24
20
Work Load
(pounds)
2720
2681
1459
1537
2033
1022
2044
1287
1905
1644
2922
2830
3089
2032 Ibs used gross
-118 Ibs in residues
T3T4~ Ibs
-576 ibs good tri left
T33S Ibs used net
118 Ibs
288 hrs
27,173 Ibs
13.59 Tons
99 Ibs tri used per ton of work
-------
TABLE II
Consumption of Trichloroethylene
(Degreaser Top Open)
Date
10-31-60
11-1-60
11-2
11-3
11-4
11-7
11-8
11-9
11-10
11-11
11-14
11-15
11-16
11-17
11-18
Tri
Added (Ibs)
1062
258
448
370
Boildown
Residues (Ibs)
166
117
106
7. tri in
Residues
28.3
9.4
11.4
Wt. of tri in
Residues (Ibs)
47
11
12
Operating
Time (hrs}
17
24
24
16
20
21
24
24
24
20
21
24
24
24
20
Work Load
(pounds^
911
1538
1673
1875
1667
1248
2180
1183
1954
1683
980
2059
2433
1225
1323
Totals
2138 Ibs used gross
-70 Ibs in residues
20~6~8" Ibs
-524 Ibs good tri left
1544 Ibs used net
70 Ibs
327 hrs
129 Ibs Cri used per ton of work
23,932 Ibs
11.97 tons
-------
TABLE HI
Consumption of MC-154
(Degreaser Top Closed)
Date
11-21-60
11-22
11-23
12-1
12-2
12-5
12-6
12-7
12-8
12-9
12-12
12-13
12-14
12-15
12-16
12-19
12-20
12-21
12-22
12-27
12-28
12-29
12-30
MC-154
Added (Lbs)
922
353
373
345
329
Boildown % MC-154 Wt . of MC-154
Residue (Lbs) in Residue in Residue (Lbs)
72 20 14
54 95
161 27 ^3
207 19 39
95 72 68
Operating Time
(Hours)
20
24
22
24
20
20
24
24
24
24
19
24
24
24
24
20
24
24
23
10
24
24
20
Work Load
(Pounds)
1229
2385
915
448
336
1089
1593
1784
1132
1993
1112
1005
1383
1848
2646
803
1763
2152
1870
655
1900
2949
2000
2322 Ibs used gross in residues
-169 Ibs
2153 Ibs
-750 Ibs good MC-154 left
1403 Ibs
-12 ibs (30 samples 0.4 Ibs/sample)
1291 Ibs MC-154 used net
169 Lbs
510 Hours
3^,990 Lbs
17.50 Tons
80 Lbs MC-154
used per
ton
-------
TABLE IV
Consumption of MC154
(Degreaser Top Open)
Date
1-3-61
1-4-61
1-5-61
1-6-61
1-9-61
1-10-61
1-11-61
1-12-61
1-13-61
1-16-61
1-17-61
i- 18-61
1-.19-61
1-.23-61
1-24-61
1-25-61
1-26-61
1-30-61
MC154
Added (Ibs)
1020
406
254
362
Boildown
Residues (Ibs)
92
91
85
67
% MC154
in Residue
39
13
42.5
17
Wt. MC154
in Residues (Ibs)
36
12
36
11
Operating Time
(Hours)
19
24
24
20
19
24
24
24
20
20
24
24
21
21-
24
2';
20
— —
Work Load
(Pounds )
1154
1834
2170
2928
574
1773
1060
556
601
465
755
1980
1204
1250
2054
1608
1908
— — — —
2042 Ibs used gross
- 33 Ibs in residues
95 Ibs
-610 Ibs good MC154
1337 Ibs
- 12 Ibs (30 samples at 0.4 Ibs/sample)
T325 Ibs
376 hrs
23,883 Ibs
11.94 Tons
111 Ibs MC154 used per Ton of work
-------
TABLE V
SUMMARY OF CONSUMPTION DATA
Consumption
(Open Top)
Consumption
(Closed Top)
Solvent Lbs sc
Trlchloroethylene
MC15^
Difference
>lvent/ton work
129
111
18
Lbs solvent/ton work
99
80
19
Difference
30
31
COST PER TON OP WORK
Cost of MC15^ (truckload) $0.1325 per Ib.
Cost of Trlchloroethylene (truckload) 0.1250 per Ib.
TRICHLOROETHYLENE - OPEN TOP
129 Ibs per ton ($0.1250) $16.1 per ton of work
IF a
TRICHLOROETHYLENE - CLOSED TOP
99 Ibs per ton ($0.1250) =
^ per ton Qf
MC15U - OPEN TOP
111 Ibs per ton ($.1325)
"ID
$14.7 per ton of work
MC15** - CLOSED TOP
80 Iba per ton ($.1325) $10.5 per ton of work
-------
T3
TJ
m
n
CO
-------
APPENDIX - C3
STUDY TO SUPPORT NEW SOURCE PERFORMANCE
STANDARDS FOR SOLVENT METAL CLEANING OPERATIONS
Evaluation of Two Refrigerated Freeboard Chillers
Hamilton Standard
Windsor Locks, Connecticut
Prepared By:
K, S. Surprenant
The Dow Chemical Company
Prepared For:
Emission Standards and Engineering Div,
Office of Air Quality Planning
U. S. Environmental Protection Agency
-------
Summary
Two open top degreasing operations equipped with refrigerated
freeboard chillers were evaluated to determine the emission
control efficiency achieved by this device. One system
was operated consistently with a cover while the other
was left open.
The freeboard chillers effected 43 percent and 40 percent
lower emission rates of methylene chloride than the same
equipment without the emission control operating. A 50
percent +_ 5 percent reduction in solvent losses to the
atmosphere was derived from the use of a cover.
The added energy required by either emission control
method is negligible compared to the basic degreasing
operation.
-------
Introduction
Autosonics Inc. recommended Hamilton Standard as a suitable
site for the evaluation of their refrigerated freeboard
chiller equipment. This device is referred to as a Cold
Trap when manufactured by Autosonics. Two Cold Traps were
f
in operation at this location. Both were reported to be
recent installations with the most up-to-date design factors.
The Cold Trap installations at Hamilton Standard are a
portion of a program to comply with Connecticut Air
Pollution Control Regulations. Earlier, the degreasers
at this location were operated with trichloroethylene.
Currently, the degreasing operations are being converted
over to exempt solvents, including methylene chloride.
Due to the lower boiling point of methylene chloride
(104°F), vapor degreaser equipment design must be
modified. The equipment modifications include: 1) a
reduced heat input, 2) an assured supply of coolant for
vapor condensation and an extended freeboard to width
ratio (0.75). These equipment modifications were sufficiently
extensive to cause the purchase of new equipment rather thar.
reconstruction of existing degreasers. Cold Traps were
purchased with the new vapor degreasing equipment to assure
minimum solvent emission rates. Methylene chloride was the
solvent used in this equipment.
-------
Objective
The purpose of this test program is to evaluate two
refrigerated freeboard chillers (Cold Traps) as a means
of controlling solvent emissions from two open top degreasers.
The information needed for this evaluation includes determining:
1. the efficiency of this device in reducing
solvent emissions to the atmosphere,
2. the cost/benefit relationship of this
emission control system,
3. the energy requirement of the emission
control system,
4. any alternate emissions created by the
emission control system.
This data base is being developed to forecast -the
magnitude of emission reductions which can be achieved
nationally and the effect on businesses involved. This
information combined with the results of other testing
will be used to design emission control regulations
which effectively limit air pollution and are practical
for industrial application.
-------
Equipment
Department 203
Crest Ultrasonics Corp. supplied the open top degreaser used
in this plant location. This stainless steel degreaser has
an open top area of 41" x 65". The inside working area is
36" x 60". The degreaser is supplied coolant from a five
horsepower Tecumseh refrigeration system, Serial CL522HT
and is electrically heated. It is Hamilton Standard No.
E30407. The distance from the top of the vapor zone to the
lip of the degreaser (freeboard) is 30 inches. The refrig-
erated freeboard chiller or Cold Trap is an Autosonics Model
No. 75, Serial No. 2558. This 3/4 horsepower refrigeration
system supplies coolant to dual heat exchange coils constructed
of copper tubing with fins. These two coils are located
immediately above the primary condenser coils around the
perimeter of the degreaser. This construction is shown
along with the rough degreaser design in Figure 1.
This degreaser works three shifts per day and six days per
week. When shut down, both the heat and the primary
condenser refrigeration system are turned off. However,
the Cold Trap is left on. This degreaser is equipped with
a counterbalanced metal cover. The degreaser is covered
whenever it is not in immediate use.
-------
Department O&R
The stainless steel degreaser is this department is another
Crest Ultrasonics open top design. The open top work area
of this degreaser is 45" x 45" while the internal work area
is 40" x 40". Again, the freeboard was found to be 30 inches,
This degreaser is supported with a small 20-30 gallon per
hour stainless steel still and both pieces of equipment are
steam heated. Both the degreaser and the still are equipped
with separate refrigeration systems to provide coolant for
the primary solvent condensation. E29957 is the Hamilton
Standard identification number for this equipment. The
cover on this O&R Department degreaser is not hinged to
the equipment as in the case of the Department 203 degreaser.
Neither degreaser (Department 203 or O&R) is equipped with
a lip exhaust system and both have a spray pump and spray
lance. The degreaser and still operate eight hours per
day and five days per week but the Cold Trap supporting
it operates continuously.
The Cold Trap is installed exclusively on the vapor degreaser
and has three finned heat exchange tubes around the periphery
of the degreaser just above the primary condenser coils.
The refrigeration system for the Cold Trap is one
horsepower and was supplied by Autosonics Inc. It is
identified by Model 100 and Serial Mo. 2487. The
essentials of this equipment design are shown in Figure 2.
-------
Experiment Design
It was planned to operate both degreasers with the Cold
Trap "on" from June 25 through July 16 and with the Cold
Trap "off" from July 16 to August 1. A longer operating
time with the Cold Trap on was deliberately planned due
to the expected lower consumption during this operating
interval. The solvent emission rate was determined by
difference in solvent inventory in the equipment. Since
all chambers within both degreasers overflow to the boiling
chamber, the difference in solvent inventory could be
obtained by measuring the solvent level in the boiling
chamber when the other compartments were full to the
overflow level. The only additional precaution needed
for the degreasing system in Department O&R was to assure
that the still was full to the upper level of the float
control. This was done manually before each measurement.
With the heat turned off, the solvent level in each of
the degreasers could be measured to plus or minus 1/8
of an inch.
Both degreasers were operated in their normal manner
and on the schedules described earlier. However, the
Department O&R degreaser was left open at all times
throughout the test period. Department 203 degreaser
-------
was covered except when work was being processed. This
use and nonuse of covers was deliberate to provide
further information on the emission control which can
be experienced through the consistent use of a cover.
Due to the inconvenience of using most covers, most
open top vapor degreasing equipment is left open most
of the operating time. The better operations cover the
degreasers during down shifts and weekends. The practice
of covering open top degreasers when not in immediate use,
as is done at Hamilton Standard, is a rare practice.
-------
Data Discussion
Department 203
The measurement of solvent emission rates was made by
difference from solvent inventory within the equipment
as discussed earlier. Using the inside dimensions of
the degreaser (36" x 60") a factor of 9.35 gallons per
inch was calculated. This value was confirmed by measuring
the solvent level before (20.5 inches) and after (26.5
inches) a 55 gallon drum of solvent was added to the
degreaser. The six inch difference in level times 9.35
gallons per inch provides an estimated volume of solvent
at 56 gallons. The volume of solvent contained in this
drum was probably higher than normal (54 gallons) because
it was loaded quite full from the Hamilton Standard bulk
tank without precise measurement. The solvent level
could be determined within plus or minus 1/8 on an inch.
This means that errors in measurement from this source
would be plus or minus 1.2 gallons.
Table 1 summarizes the solvent use record for this
department with the Cold Trap "on" and "off". No records
were kept between July 16, 1975 and July 24, 1975 due
to the failure of an electric resistance heater. This
resistance heater burn out caused the entire solvent to
be withdrawn from the equipment so that the heater could
-------
be replaced. When resistance heaters short out, solvent
decomposition often occurs. In many cases, this solvent
decomposition requires the disposal of the solvent. This
solvent loss is not recorded in the data of this report
because it is not related to the performance of the
emission control device. However, it does illustrate
a type of solvent loss which is not recoverable by
any emission control method.
The degreaser was recharged, and the test re-initiated
early on the morning of July 24. The test was aborted
on the late evening shift of July 30, 1975 soon after
the installation of ceiling fans which caused high solvent
vapor odors in the area.
The average solvent consumption for total days of testing
and per operating day are derived on Table 2. Percent
emission control achieved is the difference between the
consumption rate with the Cold Trap "on" versus that with
the Cold Trap "off" on a daily basis.
The number of workloads processed with the Cold Trap "on"
totaled 723 whereas 170 workloads were processed with the
Cold Trap "off". Thus, the workloads processed on a daily
basis with the Cold Trap "on" was somewhat higher than the
operation with the Cold Trap "off". This tends to make
-------
10
the percent emission control determined slightly conservative.
When this information is weighted with the fact that this
equipment is in use three shifts per day and six days per
week the emission control (43 percent) obtained on a total
day basis is more realistic.
The capital and operating costs for the Cold Trap are
outlined in Table 3. Again referring to Table 2, a
volume of 2.76 gallons per day were conserved through
the use of the Cold Trap. Thus, the reduced solvent
consumption can be calculated by:
2.76 Gals./Day x 7 Days/Wk. x 50 Wks./Yr. = 966 Gals./Yr.
The July 7, 1975 issue of Chemical Marketing Reporter
shows the price of methylene chloride at $0.165 per
pound. Using this value and 11 pounds per gallon, the
dollar value of 966 gallons per year equals $1753. A
net profit of $977 is obtained when the operating cost
of $776 is subtracted from the annual savings. This
represents 2.26 times the operating cost per year.
-------
11
Department O&R Degreaser
The method of emission measurements for this degreasing
system was the same as that for Department 203 with the
exception that the still was manually filled to a constant
volume before each reading. The inside working dimensions
of this degreaser are 40" x 40". One inch of level change
in the degreaser therefore equals 6.93 gallons of methylene
chloride. Ah error of plus or minus 1/8 of an" inch in
measuring the solvent level would result in the error
of 0.87 gallons in this degreaser.
The gallons used, total days of testing and operating
days are summarized on Table 3. No data was taken during
a shut-down period between July 16 and July 22 because of
the failure of the refrigeration system for the primary
condenser coils on the degreaser during this interval.
During the testing between July 10 and July 16, a drum
of methylene chloride was added to the degreasing operation.
None of the other test intervals on this degreasing operation
or that of Department 203 included an estimate of solvent
delivered from bulk storage. As mentioned, these drums
are not filled to a specific volume. This combined with
the fact that the primary refrigeration system failed
toward the end of this test interval reduces the
confidence level of this data. Due to these facts,
-------
12
this portion of the data was disregarded in developing
the average consumption rate per day of testing and
per operating day in Table 4. Again, on Table 5 the
emission control efficiency is calculated as before.
This degreaser was left open 100 percent of the time even
though it was used only one shift per day and five days
per week. As the result of this method of operation,
the emission control (40 percent) obtained on a total
test day basis provides a more factual evaluation of the
Cold Trap.
The number of workloads processed with the Cold Trap "on"
was 45 and with it "off" was 23. Thus, the loads processed
per operating day are essentially the same as shown below.
Cold Trap Loads Per Day of Testing Loads Per Operating Day
"on" 2.14 3.21
"off" 2.30 2.88
Table No. 6 outlines the capital investment and operating
costs. A volume of 1.17 gallons per day was prevented
from being lost with the Cold Trap "on". On an annual
basis (350 days), 410 gallons of solvent would have been
prevented from being lost from the system. The value of
this solvent is obtained by:
-------
13
410 Gals./Yr. x 11 Lbs./Gal. x $0.165/Lbs. = $744
Since the operating cost was estimated at $1016 per year,
this installation achieves emission control at a loss
of $272 per year or an annual savings to cost ratio of 0.73.
Comparison of Covered Vs. Open Operation
The solvent emission rate in pounds per square foot-hour
is calculated from the following formula.
2
Lbs./Ft. -Hr. = Gals./Oper. Day x 11 Lbs./Gals.
2
Oper. Hrs./Day x Ft. of Open Top Area
The emission rates determined by this formula are summarized
in Table 7. The final column of this table indicates the
emission control efficiency experienced with the covered
degreaser versus the degreaser operating in the open
condition. The available data permits the calculation
of this emission control efficiency both with and without
the operation of a Cold Trap. Of course, this calculation
is hypothetical to the extent that it presumes that both
operations are similar. In fact, the operations are
similar with regard to their general open top design and
their mutual use of methylene chloride. They are
dissimilar in their dimensions, the parts processed
-------
14
through them and the rooms (air movement) in which they
operated.
In spite of these variances between the two operations,
this data can be interpreted to say that a 50 percent
emission control + five percent could be expected by
covering a methylene chloride open top degreasing system.
It should be noted that the higher evaporation rates of
\
methylene chloride or Fluorocarbon 113 will cause greater
emission rates from open equipment at room temperature than
the other vapor degreasing solvents. In addition/ methylene
chloride has the lowest vapor density. This results in
greater susceptibility to vapor disturbance as the result
of air movement in the work area.
-------
15
Conclusions;
1. Two Cold Traps operating on open top degreasers were
found to be accomplishing a reduced emission rate
of methylene chloride vapors of 43 percent and 40
percent.
t
2. Based on present replacement pricing obtained from
the original supplier (Autosonics Inc.), the Department
203 Cold Trap was returning 266 percent per year on the
total annual operating cost while the O&R Department
unit per year year was controlling emissions at a loss
of $272.
3. By contrasting the two degreasing operations, a
forecast of a 50 percent +_ 5 percent emission
reduction was determined by ths consistent use of
a cover versus leaving the degreaser open.
4. The energy consumption by either the Cold Trap
or covers is nominal when compared to the energy
demand of basic operation.
5. No alternate air, water or solid pollution is
created by use of these emission control techniques.
-------
TABLE 1
DEPT. 203 SOLVENT RECORDS
Cold Trap
"on"
"on"
"off"
Dates Gallons
6/25/75-7/10/75 58
7/10/75-7/16/75 19
7/24/75-7/30/75 45
Total Days
15
6
7
Operating Days
12
5
6
16
-------
TABLE 2
DEPT. 203 - SUMMARY OF DATA
Cold Trap Gallons
"on"
"off"
Emission Control
77
45
Total Days
21
7
Gals./Day
3.67
6.43
43%
Operating Days
17
6
Gals./Oper. Day
4.53
7.50
40%
17
-------
TABLE 3
DEPT. 203
Capital Investment
18
Price
Installation (Included) «
Floor Space (6.2 Ft. x $25.9/Ft. )
$3300
161
TOTAL $3461
Operating Costs
Capital
Equipment (15 Yrs.)
Building (25 Yrs.)
Insurance (2 of Capital)
Equipment
Building
Maintenance (4% of Capital)
Utilities
Electricity (3/4 Hp. Motor)*
Labor
Return on Investment
$ 434
18
66
3
132
123
0
0
TOTAL COST/YEAR
$ 776
*3/4 Hp. X 0.746 KWH/Hp. X 24 Hrs./Day X 365 Days/Yr
X $0.025/KWH
-------
TABLE 4
DEPT. O&R - SOLVENT RECORD
Cold Trap
"on"
"on"
'off
Dates
6/25/75-7/10/75
7/10/75-7/16/75
7/22/75-8/01/75
Gallons
26
7
29
Total Days
15
6
10
Operating Days
10
4
8
19
-------
TABLE 5
DEPT. O&R - SUMMARY OF DATA
Cold Trap
"on"
"off"
Emission Control
Gallons
26
29
Total Days
15
10
Gals./Day
1.73
2.90
40%
Operating Days
10
8
Gals./Oper. Day
2.60
3.63
28%
20
-------
TABLE 6
DEPT. O&R
21
Capital Investment
Price
Installation (Included)
Floor Space
$4460
Operating Costs
Capital
Equipment (15 Yrs.)
Insurance (2% of Capital)
Maintenance (4% of Capital)
Utilities
Electricity (1 Hp. Motor)*
Labor
Return on Investment
$ 586
89
178
163
0
0
TOTAL COST/YEAR $1016
*1 Hp. x 0.746 KWH/Hp. x 24 Hrs./Day x 365 Days/Yr
X $0.025/KWH
-------
TABLE 7
COVERED VERSUS OPEN OPERATION
LOSS RATE PER OPERATING HOUR
Cold Trap Covered (Dept. 203) Open (Dept. O&R) Emission Control
"on" 0.112 lbs./ft?-Hr. 0.254 lbs./ft?-Hr. 56%
"off" 0.186 lbs./ft?-Hr. 0.354 Ibs./ft?-Hr. 47%
22
-------
Figure 1
DEPT. 203 NO E 30407
23
. Cover
Cold Trap
Primary Condenser Coils
Electric Heaters
60"
X
'©
I/R\
O
o
-.-Cover Handle
• Jacket
-------
Figure 2
DEPT. O&R NO E 29957
• Float Control
1:
s)J
^ OOOO 1
fS
1 ©
^v.
o
Primary Cond. o
Coil §
•
o o
^Cold Trap
L hr
Spray
Qtnr
Tank
r
r.
V
/: N
HH
••' '
ket
)
3
Refr. For
Cold Trap
Refr. For
Degr. Condenser
Coils
Refr. For
Still Condenser
Coils
\ Motor
Spray
Pump
24
-------
m
Z
o
X
n
-------
Appendix C-4
STUDY TO SUPPORT NEW SOURCE PERFORMANCE
STANDARDS OF SOLVENT METAL CLEANING OPERATIONS
EVALUATION OF CARBON ADSORPTION RECOVERY
AT HEWLETT PACKARD CORPORATION
IN LOVELAND, COLORADO
PREPARED BY:
T. A. Vivian
The Dow Chemical Company
PREPARED FOR:
Emission Standards and Engineering Division
Office of Air Quality Planning
U.S. Environmental Protection Agency
-------
-2-
Summary
The evaluation of a carbon adsorption system as an emission
control device for 1,1,1-trichloroethane used in a Riston®
system was conducted with the assistance of Mr. Virgil Hebert
at Hewlett Packard Corporation in Loveland, Colorado. It
was found that the installation of the carbon adsorption
unit reduced the overall emissions per unit of production
by 21%. Eighty-five gallons of 1/1,1-trichloroethane were
saved per week at a total recovery cost of $1.14 per gallon.
When all costs are expressed on an annual basis/ the savings
(in solvent) equals 1.89 times the annual cost.
The recovered solvent contained good levels of "acid acceptor"
stabilizer and acceptable water concentrations. However,
the stabilizers designed to protect the solvent from reaction
with aluminum were recovered in totally inadequate quantities.
The water effluent from the water separator of the carbon
adsorber contained from 2,000 ppm to 14,000 ppm of total
hydrocarbons at the point of discharge to the drain.
©Riston is a registered trademark of the E. I. DuPont Company.
-------
-3-
Objective
The objective of this test was to evaluate carbon adsorption
as a means of controlling 1,1,1-trichloroethane vapor
emissions in air. The information required for this
evaluation required the development of specific data
concerning:
1. Efficiency of this system.
2. Financial impact or cost/benefit of this system.
3. Energy requirement of the system.
4. Effect of the system on the present stabilizer levels,
5. The effect of the system as a potential water
pollutant.
-------
-4-
Introduction
Carbon adsorption has been offered to the solvent metal
cleaning users since 1958. Other than specialty processes
for example/ film cleaning or textiles, metal cleaning carbon
adsorption has been limited to perchloroethylene, trichloro-
ethylene, methylene chloride, and Freons®. 1,1,1-trichloro-
ethane was omitted from carbon adsorption due to the corrosion
of the vapor adsorption equipment and the reactivity of the
recovered solvent to aluminum.
Hewlett Packard was selected as an emission control test
site for 1,1,1-trichloroethane for three main reasons.
1. The unit was new having started operation in
January of 1975. This meant a system at optimum
working order.
2. The unit was commercially available.
3. The unit was constructed of materials specifically
designed for the recovery of 1,1,1-trichloroethane
as inhibited for the metal cleaning industry.
^Registered trademark of the E.I. DuPont Company,
-------
-5-
Equipment
The developer (see Figure 1) was built by Hewlett Packard.
It is a long narrow spray chamber with a conveyorized
monorail. Circuit boards are hung from the monorail and
conveyed past various spray tanks of both CHLOROTHENE NU
and water. After the developer, the boards are dried by
blowing them with air and passing them by heat lamps prior
to further treatment. Lip vent exhausts from both the
entrance and exit lead to a Vic Model 536AD carbon adsorber.
The cost of the develop line to Hewlett was $13,361.60 (parts
and labor). The ovens were installed at an additional cost
of $5,847.00.
Production was scheduled so as to maximize idling time by
running only when a sufficient backlog of boards warrants
the developer operation, usually two hours per day, but
recently four hours per day. When idling, the spray tanks
were turned off and approximately two inches of water settled
out on top of the solvent in the develop tank.
The "Riston C100" still was made by DuPont and resembles an open
top vapor degreaser with a cover. Its solvent level is main-
tained by float controls. The rate of distillation is dependent
on contaminant level and varied from 5 gals./hr. to 45 gals./hr.
The average was 9 gals./hr. based on 13 random sample rates.
-------
-6-
The primary function of the carbon adsorption system was
to recovery solvent used in the Riston system. Its
secondary purpose was to capture the evaporation losses
from the recovered solvent holding tank (see Figure 2).
This adsorption system had a three horsepower fan, about
350 pounds of carbon per bed and operates with 85 psig
air pressure and 10 psig steam pressure. Approximately
340-360 gallons per hour of condenser water was required
and the effluent remained fairly steady at 145°F. The cost
of the adsorption system was $16,042.00 and the installation
was $1,940.00 additional. This adsorption system was higher
in price than normal for a model 536AD due to the special
materials of construction used. These materials included
Hastelloy for the shell of the beds, Monel condenser tubes,
Heresite coated milesteel for the water separator and stainless
steel for the holding tank.
Experiment Design
This evaluation was designed to determine the following:
1. Efficiency—overall solvent recovery efficiency
was determined by comparing the solvent use rate
per unit of production for six months before and
six months after the installation of the carbon
adsorption system. The efficiency of the carbon
bed proper was determined by measuring the vapor
concentrations of the air stream before and after
passage through the beds.
-------
-7-
2. Effect of the adsorption system on stabilizer
levels: The stabilizer concentrations in the
recovered solvent and the Riston process equip-
ment were analyzed and compared to the stablizer
levels in fresh solvent.
3. Evaluation of the adsorption system as a source
of water contamination: Steam condensate samples
were taken from the water separator discharge to
the drain and analyzed for total hydrocarbon
content.
4. Energy: An energy balance was determined for both
phases of operation by measuring solvent recovered,
the steam condensate used and the volume of con-
denser water.
5. Recovery cost per gallon of solvent conserved.
-------
-8-
DISCUSSION
Overall Efficiency
Production records were kept for six months prior to the
installation date and six months after startup of the
carbon adsorption equipment. These records included both
the number of circuit boards processed and the total amount
of .1,1,1-trichloroethane used in the development process.
Fortunately, all circuit boards are of uniform size eli-
minating this variable. Consumption was calculated by
dividing the number of boards processed into the number of
gallons used. The overall emission control efficiency
was based upon the comparison of these two six month periods.
This data is summarized below.
Solvent Use Record
Without Carbon Adsorption 0.33 gals./board
With Carbon Adsorption 0.26 gals./board
Recovered Solvent 0.07 gals./board
Recovery Efficiency 21%
Bed Efficiency
Solvent vapor concentrations in air were analyzed by means
of a "Siphon" pump connected to a carbon tube from
-------
-9-
"Organic vapor SKC, Inc.". The "Siphon" pump has a counter
and each count was calibrated for the cubic centimeters of
air sampled. The total number of counts in each air sample
was recorded and converted to an air volume. The solvent
vapors were adsorbed in the carbon tube, extracted, and the
total weight of 1,1,1-trichloroethane in the sample was
determined by a gas chromatograph. Air flow measurements
were taken using an "Alnor" thermo-anemometer. These data
are summarized on Table 1.
TABLE 1
Solvent Air Vapor Concentrations
Develop Line "Off"
PPM AV. %
Exit Bed Efficiency
11
2£
Average 148 18 88%
" Hn "
Develop Line "On
Inlet Exit
73 ppm
4395 ppm 133 ppm
4912 ppm 145 ppm
205 ppm
4927 ppm 496 ppm
789 ppm
Average 4745 307 94%
-------
-10-
Thus, the carbon bed efficiency ranged between 88-94%.
The percent recovery efficiency overall reflects the actual
savings to the customer and the emission control achieved.
The percent bed efficiency is the maximum potential of the
system if 100% of the vapors were captured by the ventilation
system and directed through the carbon adsorption beds. The
difference in the two percentages demonstrates that the
efficiency of the carbon adsorption system as a whole is a
function of its ability to capture the vapors emitted rather
than the ability of the bed to recover solvent delivered to it.
Relying principally on the same data, a solvent material
balance for the total system can be constructed (see
Figure 3). Although the developed line operates only 2-4 hours
per day, the carbon adsorption system functions eight hours
per day. The material balanced describes these two modes of
operation: (A) Development line "Off" and (B) Development
line "On".
Effect of Adsorption on Solvent Stabilizers
Solvent stabilizers are needed for essentially all of the
chlorinated solvents. Each of the stablizer systems are
designed to protect the individual weaknesses of a specific
chlorinated solvent. In the case of 1,1,1-trichloroethane
-------
-11-
three stabilizers will be discussed. The "acid acceptor"
protects the solvent by removing any strong acids present
in it. Two aluminum stabilizers prevent the reaction between
1,1,1-trichloroethane and aluminum. These stablizers will be
referred to as the primary and secondary aluminum stablizers.
Stablizer concentrations were determined from daily samples
of solvent taken from seven points shown on Figure 4.
These daily samples were analyzed over an interval of
approximately six months. Samples of solvent recovered by
the carbon adsorption system were found to contain 80% of
the original concentration of "acid acceptor" but a 87%
loss in the primary aluminum stablizer concentration was
experienced. A serious loss in the secondary aluminum
stablizer was also experienced, but an exact percent loss
could not be defined due to the low levels present. Since
only about one gallon in five of solvent required for the
developed system was recovered solvent, the stablizer lost
effect on the process solvent inventory was much less. The
following observations were made on the effect of the
carbon adsorption system on the stablizer levels in the
develop line after six months.
1. No decomposition products are seen in the develop-
ment line either before or after the installation
of the carbon adsorption system.
-------
-12-
2. The acid acceptor level in the developed system was
down 15% from original solvent levels. However,
this amount of stabilizer loss is not detrimental
to the system.
3. In the developed system, the primary aluminum
inhibitor was reduced only 15% from the fresh
solvent level although 87% of this inhibitor was
not recovered with the solvent from the carbon
adsorber.
4. The secondary aluminum inhibitor concentration
was reduced 53%. Although the solvent could
be used in this specific operation due to the
large dilution with fresh solvent, the recovered
solvent by itself was inadequately stabilized for
use in any aluminum operations or systems containing
aluminum as a material of construction.
Examination of Water Contamination From the Carbon
Adsorption System
Samples of steam condensate from the carbon adsorber
system were collected at the drain during two desorption
cycles. One desorption cycle was representative of the
operation with the developed line on, the other with the
-------
-13-
develop line off. These samples were analyzed for total
hydrocarbon content and the data is summarized on Table 2.
Although chlorinated solvents are essentially insoluble
in water, their stablizer systems are frequently soluble
in both water and solvent. As the steam in solvent
condenses during the adsorption cycle, intimate mixing takes
place. This mixing is continued until separation takes
place in the water separator. This results in solublizing
significant quantities of the stablizers in the water phase
which is discharged to the drain. The volume of steam
condensate generated per hour averaged 22.6 gallons and
ranged in hydrocarbon content from 2,000-14,000 ppm. This
volume of steam condensate was diluted approximately
15 times by the 350 gallon per hour discharge of condenser
water. In most cases, this stream would then be diluted
perhaps a 1,000 or more times from other plant water uses.
It should be possible to remove much of this potential
water contamination by sparging air through the steam
condensate prior to discharging it at the drain.
Energy Balance
The gallons of recovered solvent and steam condensate were
measured from several desorption cycles. These quantities
are reviewed in Table 3. It should be noted that the
volume of steam condensate used is related to the time of
-------
-14-
desorption rather than the volume of solvent retained by
the bed. Other measurements determined that the condenser
water flow rate ranged between 340-360 gallons per hour.
The energy balances on Figures 4 and 5 for the adsorption
and desorption cycles were developed from this information.
The carbon adsorption energy balances are also tabulated on
Table 4. The overall costs of operations are developed on
Tables 5, 6, and 7. The calculations for utility costs are
reviewed on Figure _. Prior to the installation of the
carbon adsorption system, 412 gallons of 1,1,1-trichloro-
ethane were consumed per week. With the carbon adsorption
system in operation/ the solvent consumption was reduced
to 327 gallons per week. Thus, 85 gallons of solvent were
conserved per week by the carbon adsorption system or
4,250 gallons per year (50 weeks per year). The value of
that solvent at $2.12 per gallon is $9,010.00. This is
compared to the $4,838.00 in cost per year developed on
Table 7, a savings to cost ratio of 1.86 is found. This
means that for every dollar invested in cost per year,
$1.86 worth of solvent was recovered. Alternately, a cost
per gallon of solvent recovered can be calculated by
dividing the annual cost of $4,838.00 by the 4,250 gallons
per year. This results in a cost per gallon of recovered
solvent of $1.14.
-------
-15-
Conclusions
1. The overall recovery efficiency of the carbon adsorption
system was 21%. This represents the ability of the
system to capture and recover solvent vapors from the
develop operation.
2. The carbon bed efficiency was found to be 88% during
develop line "off" operation and 94% with the develop
line "on".
3. The cost of recovered solvent was $1.14 per gallon when
all costs were expressed on an annual basis. Similarly,
$1.86 worth of solvent was recovered per $1.00 of
annual cost.
4. Serious losses of solvent stabilization were experienced
in the recovered solvent. However, the recovered solvent
could be re-used in this operation without restabilization.
If the operation was more demanding or the recovery
efficiency higher, the solvent would require restabilization,
5. A small volume (20 gals./hr.) of water is contaminated
with 2,000-14,000 ppm of total hydrocarbon.
-------
TABLE 2
Drain Water Contamination Levels
Desorption
Time Develop Line Develop Line
(Min.) "Off" "On"
Total Hydrocarbons Concentrations (PPM)
In Steam Condensate (At Drain)
4 1,951
12 1,995
20 4,121
23 1,867
31 6,053
33 6,619
40 5,199
42 9,693
44 14,342
48 4,049
55 2,805
-------
TABLE 3
Develop Line "Off"
Av.
Develop Line "On"
Av.
Gal.
Solvent
1.2
2.6
1.8
9.9
7.9
8.7
8.8
Gal
Steam Con
24.2
16.2
24.5
23.9
14.3
-------
TABLE 4
Energy Balance on Carbon Adsorber (1 Bed)
Adsorption (Solvent Recovery)
Solvent Latent Heat (20.8 Lbs. A)
(9617 Lbs. B)
Air Flow @ 700 Cfm
Desorption
Steam 113 Lbs.
Heating Tank (900 Lbs.)
Heating Carbon (300 Lbs.)
Condenser Water (8 GpM)
Vaporizing Solvent*
Adsorption (Drying and Cooling Bed)
Cooling Tank
Cooling Carbon
Air Flow
Input
2,122 Btu's A
Output
(9,863 Btu's B) 2,122 Btu's A
(9,863 Btu's B)
113,000 Btu's
13,860 Btu's
10,500 Btu's
88,640 Btu's
2,122 Btu's A 2,122 Btu's A
9,863 Btu's B 9,863 Btu's B
13,860 Btu's
10,500 Btu's
24,360 Btu's
*Vaporizing solvent from bed requires heat from steam but this same
heat is given up to the condenser water when the vapor is condensed
to liquid solvent.
A) Develop line off
B) Develop line on
-------
TABLE 5
Industrial Building*
Shell (M&L) Cost
Lighting and Electrical
Heating and Ventilating
Plumbing
Fire Prevention
Sub-Contract Cost (1.3)
Contingency (15%)
$ 4.09/Ft.
1.75
1.50
1.70
1.10
$10.14 (1968 Base)
$17.3/Ft.2 (8% Annum
1.71 Multiple in 1975)
$22.5/Ft.2
$25.9/Ft.2
*Derived from "Modern Cost-Engineering Techniques" by
H. Popper. The 8% inflation rate was estimated by
K. S. Surprenant.
-------
TABLE 6
Carbon Adsorber Capital Coats
Building Space
Direct 100 Ft.2
Indirect (50%) 50 Ft.2
Total 150 Ft.2
Value @ $25.9/Ft.2 = $3,885 (Table 1)
Cost/Annum = $3,885 x .11017*1 = $428
Direct Capital
Price $16,042.00
Shipping and Installation $ 1,940.00
Total Capital $17,982.00
Cost/Annum $17,982.00 x .13147*2 = $2,364
* Factor for linear return of capital with 10% interest,
* Factor for linear return of capital with 10% interest,
-------
TABLE 7
MODEL 536 AD CARBON ADSORBER
OPERATING COST PER ANNUM
Capital
Building $ 428
Equipment $2364
Insurance (2%)
Equipment $ 321
Building $ 78
Maintenance (4%) $ 642
Utilities
Steam $ 865
Electricity $ 112
Water $ 28
Compressed Air Nil
Labor Nil
ROI Nil
Total Cost/Annum $4838
Solvent Recovered/Annum 4250/Gal.
Cost/Gal. Recovered $1.14
-------
Figure 1
-------
Figure 2
-------
FIGURE 3
Utility Cost Calculations
Steam
22.6 gals/hr x 2000 hrs/yr x 8.33 Ibs/gal x 1000 Btu/lb
= 377 x 106 Btu/yr
377 x 106 Btu/yr x $2.3/106 Btu T 106 = $866
Electricity (3 Hp fan)
3 Hp x 0.746 KWH/Hp x 2000 hr/yr x $0.025/KWH = $112
Water
350 gal/hr x 2000 hrs/yr x $0.04/103 gal f 1000 gal = $28
-------
-o
-o
m
Z
D
X
n
61
-------
APPENDIX - C5
STUDY TO SUPPORT NEW SOURCE PERFORMANCE
STANDARDS FOR SOLVENT METAL CLEANING OPERATIONS
Evaluation of (1) A Pneumatic Cover (2) Refrigeration
to Control Solvent Emissions at
Pratt Whitney
Hartford, Connecticut
Prepared By:
K. S. Surprenant
The Dow Chemical Company
Prepared For:
Emission Standards and Engineering Div.
Office of Air Quality Planning
U. S. Environmental Protection Agency
-------
Summary
Two emission control methods for solvent metal cleaning
operations were evaluated at Pratt Whitney. A pneumatic
cover was found to effect a 40 percent reduction in
emissions from an open top degreaser using 1,1,1-trichloro-
ethane. This emission control requires little investment
and energy and yields a positive return on investment
(156 percent per year).
A refrigerated freeboard chiller was studied on another
open top degreaser. This degreaser was equipped with
a manually operated cover. A 16 percent lower emission
rate was indicated with the refrigeration system in
service then without it. The capital investment is
about twice that of the pneumatic cover. The solvent
conserved by this device would amortize about 78% of
the investment and operating costs. The energy demand
is less than carbon adsorption but greater than a
pneumatic cover.
-------
Objective
The purpose of this test program is to evaluate:
1. a pneumatic vapor degreaser cover, and
2. a refrigerated freeboard chiller (Cold
Trap) as a means of controlling solvent
emissions from metal cleaning operations.
The information needed for this evaluation includes
determining:
1. the efficiency of this device in reducing
solvent emissions to the atmosphere,
2. the cost/benefit relationship of this
emission control system,
3. the energy requirement of the emission
control system,
4. any alternate emissions created by the
emission control system.
This data base is being developed to forecast the
-------
magnitude of emission reductions which can be achieved
nationally and the effect on businesses involved. This
information combined with the results of other testing
will be used to design emission control regulations
which effectively limit air pollution and are practical
for industrial application.
-------
Introduction
Pratt Whitney Division of The United Technologies, Inc.
was recommended by Dow technical representatives because
of their wide use of solvent metal cleaning and highly
qualified technical staff. These assets made the prospect
of locating one or more techniques of controlling solvent
emissions likely at this location.
A preliminary meeting with Pratt Whitney people revealed
two opportunities to evaluate emission control technology.
The first of these was the use of a pneumatically actuated
cover on an open top degreaser. Although essentially all
open top degreasers are manufactured with covers, most are
equipped with heavy metal covers which are seldom used.
In contrast to industry wide normal procedures, Pratt
Whitney reported a well disciplined use of canvas covers.
Further, they found the installation of a pneumatic cover
had increased the effectiveness and decreased the labor
associated with manually closing the cover. The second
technique employed to control solvent emissions was a
refrigerated freeboard chiller (Cold Trap). This device
reduces solvent diffusion loss from the vapor zone by
establishing a cold air mass immediately above the vapors.
This cold air mass acts as a barrier to the transport of
-------
solvent vapor out of a vapor degreaser. Both systems were
regarded as worthy of examination.
Each of these evaluations will be treated separately through
the balance of this report.
-------
Pneumatic Cover Degreaser
Equipment
The degreaser proper is a moderately large open top Detrex
Model 1 DVS-800-S, Serial No. 52301-A. This degreaser is
constructed of stainless steel, steam heated and equipped
with a spray lance. It is 65 inches wide x 110 inches long
and is operated with a 40 inch freeboard (the distance from
the top of the vapor zone to the top of the degreaser). The
degreaser is divided to provide a solvent immersion chamber
and a boiling chamber. This equipment is diagramed in
Figure 1.
The cover is constructed of canvas and is supported by
rods approximately every 18 inches.
-------
Experiment Design
Degreaser operation was unchanged except for the use of
the cover. Two weeks, as a minimum, were planned with the
cover in normal use. The cover was to be deactivated and
i
the degreaser left open for a minimum of two weeks. The
efficiency of the emission control device could then be
determined directly by comparing the solvent losses
experienced with and without it.
This degreasing operation is used three shifts per day
and five and one-half days per week. The standard Pratt
Whitney procedure for operation of the degreaser involves
closing the cover after every use. Obviously, the equip-
ment would be closed on weekends and holidays. During
the test period without the cover, the degreaser was left
open 100 percent of the time including weekends. This
method of operation provided the maximum control of test
conditions and avoided possible human errors of various
operators using the equipment.
-------
Data Discussion
The standard practice at Pratt Whitney includes pumping
essentially all the dirty solvent out of the boiling
chamber of the degreaser once every two weeks. Clean
solvent is added to the degreaser as needed during the
two week intervals. For convenience in making test
measurements, the degreaser was pumped out on a weekly
basis and no solvent additions were made during the
week. This was done in all cases except for the two
week interval beginning July 17, 1975. During this
test period no solvent was pumped from the degreaser
or added to the degreaser for the full two weeks.
A cart containing two chambers is used to deliver fresh
1,1,1-trichloroethane and remove dirty solvent from the
operation. Each chamber is 34 inches square by approximately
31 inches deep. Consequently, an inch of solvent in
one of these chambers is equal to 5.00 gallons. The
depth of solvent was measured before and after dirty.
solvent was pumped from the boiling chamber as well as
before and after the clean solvent was pumped into the
degreaser. The solvent depth in these chambers could
be measured to within one quarter of an inch easily.
Thus, the error in measurement of solvent volume from
this source would be less than plus or minus 1.25
gallons. Variance due to the completeness of pumping
-------
solvent from the degreaser is estimated at plus or
minus two gallons. The solvent consumed was taken
to be the total solvent added minus the total volume
of solvent removed. This information is reported in
Table 1 along with the number of parts processed on
a biweekly basis.
The degreaser was operated three shifts per day, five
and one-half days per week. Two exceptions to this
schedule occurred. July 4th was celebrated as a holiday
and the degreaser was not operated/ during a two week
interval of testing with the cover off. Less than one
day of operation was lost with the cover operating
during the two week interval beginning July 17, 1975.
This interruption was caused by the need to repair the
solvent pump for the spray lance. No need to adjust
solvent consumption is indicated because one shut down
occurred with the cover in operation while the other
without the cover.
-------
The work parts processed through this equipment are large
sections of turbine motors. A typical piece might be roughly
40 inches in diameter and 10 inches thick. These pieces
contain numerous large cavities. Often these cavities
present difficulty in obtaining good solvent drainage to
avoid drag-out losses. The variance in the quantity of
work processed with the cover on versus that with the cover
off was approximately 4 percent. This variance yields
a slightly higher emission control efficiency (42 percent)
on a work piece basis versus the 40 percent efficiency
obtained on a weekly solvent consumption basis shown on
Table 2. The volumes of solvent consumed per unit piece
are shown below:
0.159 gallons per piece with cover
0.276 gallons per piece without cover
-------
An estimate of $2,300 was provided by Detrex Chemical
Industries, Inc. as a replacement cost for the pneumatically
operated canvas cover. Shipping and installation is
estimated at $345 (15 percent of capital). Since no
floor space is required for this equipment, no building
capital is assessed. With a 15 year write-off and a 10%
time value of money, the annual cost of capital is $348
($2645 x 0.13147). Maintenance (4 percent of capital)
$106 and insurance (2 percent of capital) $53 increase
the annual operating cost to $507. The only power used
on this equipment is plant compressed air. No measurement
of air requirement was made. However, the power requirement
is minimal since it is used only about 10 to 15 seconds per
load and only about 360 loads are processed per week. Thus,
a value of $25 per year is assigned for the utility expense.
The total operating cost would be approximately $532.
The "Chemical Marketing Report," July 7, 1975 price of
1,1,1-trichloroethane is $2.12 per gallon. The total
operating cost of $532 for the pneumatic cover would be
recovered by conserving 251 gallons of solvent. The
actual solvent saved by this device was 39 gallons per
week. On a yearly basis, 1,950 gallons of solvent are
being conserved or $4134. Thus, the value of the
solvent saved is 7.77 times the annual operating cost,
including capital.
-------
Conclusions
1. A 40 percent reduction in solvent emissions was
found to occur through the use of a pneumatic
cover on an open top degreaser.
2. The pneumatic cover yielded a return on investment
of 156 percent per year. The savings to annual
cost ratio including capital was 7.77.
3. The energy requirement for this emission control
is minimal.
4. No adverse effect on the solvent stabilization
is observed as a result of this emission control
means.
-------
Figure 1
DETREX MODEL IDVS-800-S R 511853
Spray
Stor.
o
o
o
Clean Solvent
Immersion
Chamber
O O O O
O O O O
110"
Floor Level
-------
TABLE 1
Pneumatic Cover Degreaser Records
Operational Mode Two Weeks Beginning Solvent Used Parts Processed
With Cover
Without
Without
Cover
Cover
With Cover
May 29 ,
June
June
12
26
July 10
(1 Week
With Cover July
17
1975
, 1975
, 1975
, 1975
Only)
, 1975
116
202
188
67
109
Gals.
Gals.
Gals.
Gals.
Gals.
851
729
681
357
630
-------
TABLE 2
Summary of Pneumatic Cover Degreaser Data
Operational Mode Total Test Time Average Solvent Used Average Parts Processed
With Cover 5 Weeks 58.4 Gals./Week 368/Week
Without Cover 4 Weeks 97.5 Gals./Week 353/Week
% Emission Control = (97.5 - 58.4) 100 = 40%
97.5
-------
Refrigerated Freeboard Chiller
Equipment
The vapor degreaser used in this evaluation is sketched
in Figure 2. The degreaser is a Detrex 1 DVS-800-S model,
Serial No. 52298. The Pratt Whitney equipment number is
R511854. Again, this open top degreaser is equipped with a
canvas cover and a spray lance and is steam heated. The
view shown in Figure 2 is from the end of the degreaser.
This equipment .had an open top surface area of 56 inches
wide by 90 inches long. The overall depth of the equipment
was 139 inches and the freeboard was found to be approximately
33 inches. As was the case with the pneumatic cover degreaser,
this equipment is divided to provide a clean immersion chamber
so that parts can be rinsed in the liquid solvent (1,1,1-
trichloroethane). Unlike the earlier degreaser, the cover
on this equipment is manually operated.
The refrigerated freeboard chiller on this equipment is
a Cold Trap Model 150, Serial No. 2171. This portion
of the equipment was supplied by Autosonics Incorporated
in August of 1968. The present value of this one and
one-half horsepower equipment ($4,900) was provided by
the original supplier. This price includes the equipment
cost, shipping and installation. At the time of installation
Autosonics advised that the expected reduction in solvent
usage would not be achieved unless:
-------
1. A steam condensate return pump was installed
to avoid flooding the steam coils and disturbing
the degreaser heat balance.
2. The steam line supplying the heating coil
should enter the degreaser below the vapor
zone. This steam line was originally piped
through the freeboard, vapor interface and
vapor zone to the heating coils. This method
of supplying steam disturbed both the cold
air barrier and the vapor zone.
Both of these modifications had been made prior to this
evaluation.
It should be noted that current freeboard chiller equipment
supplied by Autosonics employs different design parameters
than exist on this equipment. These more advanced design
parameters reportedly provide more efficient vapor emission
control. Other evaluations are designed to determine the
effectiveness of the improved parameters.
This evaluation provides valuable data on the effective
life of this equipment and its performance several years
after original installation. In addition, other suppliers
of refrigerated freeboard chillers offer equipment with
still different design parameters.
-------
Experiment Design
Two weeks of testing were planned as a minimum with the
refrigerated freeboard chiller on and another two weeks
with it off. No other operational procedures were
changed. As noted earlier, it is standard practice at
Pratt Whitney to cover all degreasers when they are not
in immediate use. Thus, this evaluation determined what
additional emission control could be achieved with a
refrigerated freeboard chiller when an effective emission
control (the degreaser cover) was already in use.
All solvent handling and measurement techniques employed
with the pneumatic cover degreaser were used in this
evaluation as well.
-------
Data Discussion
All of the test records are summarized in Table 3. The
data accumulated during the week of June 20, 1975 was
discarded because of the short six day week with only
four operating days. Again, the week of June 26, 1975
was discarded from analysis due to the three working days
in the week caused by shut downs and the 'low part process
load. Some consideration was given to discarding the
data for the week of July 3 due to the holiday. However,
it was decided to incorporated this data since the part
process load was more than comparable to the succeeding
work weeks.
The averages reported in Table 4 represent the two
weeks of July 3 and July 10, 1975 with the refrigeration
on and the next two weeks with the refrigeration off.
The percent emission control is caluclated on the weekly
solvent consumption. The parts processed with the
refrigeration system on were 10 percent higher than
the workload processed with the refrigeration off.
However, one less day of work was experienced with
the refrigeration on. These factors should tend
to cancel one another. If the percent emission
control is calculated on a workday basis, the efficiency
-------
is reduced. In contrast, if the efficiency were
expressed on a work piece basis the efficiency of
control would be increased.
Table 5 derives a capital cost per square foot of plant
space. Table 6 provides a breakdown of the total capital
and operating cost estimated for the refrigerated free-
board chiller. A value of $2.12 per gallon is taken for
1,1,1-trichloroethane from the July 7, 1975 issue of
"Chemical Marketing Reporter". Using this value of solvent
a total of 574 gallons of solvent would be needed to off-
set the annual cost of $1216 per year. If nine gallons of
solvent are saved per week on an average, 450 gallons would
be conserved in the course of the year. The value of this
solvent would be $954 and would result in a $262 loss using
the operating cost outlined in Table 6. These values can
be expressed as a ratio of savings to cost ($954/$1216 = 0.785) .
Thus, 78.5% of the operating cost (including capital) per
year was returned via solvent savings.
The refrigerated freeboard chiller cannot be expected to
conserve solvent which is already being controlled through
the use of a manual cover. If this degreasing operation
was operated in the open condition as is common practice,
the control achieved by the refrigerated freeboard chiller
could be expected to be notably higher. This and the
improved design parameters of newer equipment will be
evaluated at other locations.
-------
The dirty solvent removed from all of the Pratt Whitney
degreasers is distilled in a central still. Two samples
of the still residues from this system averaged 69 percent
by weight nonvolatile matter. The individual readings
were 75.9 percent and 61.5 percent. Since the oil content
of the solvent withdrawn from the degreasing operations
during test contain only low percentiles of oils, no
significant solvent losses were experienced through
solvent distillation. Solvent stabilization of both
systems was found to be adequate. Water contamination
was observed visually and by solvent analysis throughout
the evaluation of the freeboard chiller. A mixture
of solvent and water freezes on the freeboard chiller
coils. From ice samples taken, the composition appeared
to be approximately one part water to two parts solvent.
This mixture should be removed from the system during
the defrost cycle by the water separator. However,
water removal was not being effected during the evaluation.
-------
Conclusions
1. A refrigerated freeboard chiller can achieve
a low level of emission control (16 percent)
even with a cover in operation.
2. The emission control obtained by adding two
systems to one solvent metal cleaning operation
is not a sum of the expected emission control
of each separately.
3. The refrigerated freeboard chiller studied in
this evaluation was recovering 78.5% of the annual
operating costs per year, including the capital.
4. A refrigerated chiller requires relatively
low energy requirements compared to carbon
adsorption but much greater energy than a
powered cover.
5. A mixture of solvent and water freezes on
the freeboard chiller coils. During the
defrost cycle, the mixture enters the degreaser
condensate trough. If not effectively removed,
water can be a source of serious equipment
corrosion.
-------
Figure 2
DETREX 1 DVS-800-S NO 511854
Refrigerated
Freeboard
Chiller
Water Jacket
Steam Coils
©;
l!
o o o o
"b o o o
56'
O
O
1
•Primary Condenser Coils
. Water Sep
•Spray Storage TK
Spray Pump
Spray I
(fa
I P— Motor
Floor
-------
TABLE 3
Refrigerated Freeboard Chiller Records
Operational Mode Week
Refrigeration On
<
June
June
July
„ July
r July
Refrigeration Off < julv
Beginning
20
26
3,
10
17
24
, 1975
, 1975
1975
, 1975
, 1975
, 1975
Total Days
6
7
7
7
7
7
Work Days
4
3
4
5
5
5
Solvent Used Parts Processed
48
41
40
57
52
63
Gals.
Gals'.
Gals.
Gals.
Gals.
Gals.
-
207
277
238
205
263
-------
TABLE 4
Summary of Refrigerated Freeboard Chiller Data
Weekly Weekly
Operational Mode Average Solvent Used Parts Processed
Refrigeration On 49 Gals. 258 Pieces
Refrigeration Off 58 Gals. 234 Pieces
% Emission Control = (58 - 49) 100 = 16%
58
-------
TABLE 5
Industrial Building*
Shell (M&L) Cost $ 4.09/Ft.2
Lighting and Electrical 1.75
Heating and Ventilating 1.50
Plumbing 1.70
Fire Prevention 1.10
$10.14 (1968 Base)
$17.3/Ft.2 (8%/Annum
1.71 Multiple in 1975)
Sub-Contract Cost (1.3) $22.5/Ft.?
Contingency (15%) $25.9/Ft.
•Derived from "Modern Cost-Engineering Techniques" by
H. Popper. The 8% inflation rate was estimated by
the author.
-------
TABLE 6
Capital Investment
Price $4900
Installation (Included)
Floor Space* 256
Operating Costs
Capital*
Equipment (15 Yr.) 644
Building (25 Yr.) 28
Insurance (2%)
Equipment 98
Building 5
Maintenance (4%) 196
Utilities
Electricity (1 1/2 Hp. Motor)*** 245
Labor Nil
Return on Investment 0
Total Cost/Year $1216
*6.6 Ft. of Direct Floor Space Times 1.5 for Indirect
Space Times $25.9 Per Ft.
**The annual cost of capital was calculated at a 10%
time value of money. The factor for a 15 year life
is 0.13147, for a 25 year life is 0.11017.
***! 1/2 Horsepower x 0.746 KWH/Hp. x 365 Days x 24 Hrs.
x $0.025/KWH
-------
-o
•o
m
Z
n
-------
APPENDIX - C6
STUDY TO SUPPORT NEW SOURCE PERFORMANCE
STANDARDS FOR SOLVENT METAL CLEANING OPERATIONS
Evaluation of Vapor Degreasing Vs. Cold Cleaning
Prestolite Corporation
Bay City, Michigan
Prepared By:
T. A. Vivian
W. C. Douglas
The Dow Chemical Company
Prepared For:
Emission Standards and Engineering Division
Office of Air Quality Planning
U. S. Environmental Protection Agency
-------
Summary
The comparison of cold cleaning vs. vapor degreasing as sources
of emission was conducted at the Prestolite Corporation. The
particular cleaner used in the study had been under close
observation by Mr. Ray Gittens of Prestolite for a period
exceeding two years in order to ascertain the effect of vapor
degreasing with 1,1,1-trichloroethane. Due to the thoroughness
of Mr. Gitten's study, this data was not repeated but used as
the reference on vapor degreasing, saving considerable time and
money. The degreaser was used as a cold cleaner for five weeks
after which the surface of the unit was covered with a layer
of 1 1/2" diameter plastic balls. Cold cleaning with 1,1,1-
trichloroethane as opposed to vapor degreasing with 1,1,1-
trichloroethane reduced emissions 47%. The addition of the
plastic balls to the cold cleaning operation reduced emission
levels an additional 3% beyond the cold cleaning level, however,
due to their hindrance of production, these balls had to be
removed. Two additional points should be made in the 47% cold
cleaning emission reduction. Not all operations can clean without
vapor degreasing, which entails always clean solvent condensing
on the part being cleaned. Second, the consumption figure for
cold cleaning does not reflect distillation of contaminated
solvent which would reduce losses even greater. The data
does show that if cold cleaning can be used and the dirty
solvent distilled or reclaimed, significant emission and
cost savings can be recognized.
-------
Objective
The objective of this test program was to compare energy
and emission data via vapor degreasing vs. cold cleaning.
The information sought for this test included the following:
1. Trichloroethane vapor degreasing
Solvent consumption
Heat requirements
2. Trichloroethane cold cleaning
Solvent consumption
3. Cold cleaning with plastic balls as vapor depressants
-------
Conclusions
At Prestolite:
A. Cold cleaning was 47% more efficient than
vapor degreasing with 1,1.,1-trichloroethane.
B. Plastic balls as vapor depressants in cold
cleaning were only marginally effective and
were too great a hindrance to production.
General Conclusions^
Where feasible from a cleaning standpoint, cold cleaning
offers a unique opportunity to reduce solvent costs,
solvent emissions and energy. The addition of a still
to a cold cleaning operation would further enhance the
process.
-------
Equipment
Prestolite has a Baron Blakeslee Model HD425 vapor degreaser
(see sketch). The dimensions are: width 30.5 inches,
height 59 inches, length 48.3 inches. The degreaser is
equipped with a roll lid to prevent evaporation and a lip
vent exhaust to prevent a concentration of vapors to the
operator's breathing zone.
1. Trichloroethane vapor degreaser:
The consumption for 1,1,1-trichloroethane for
a one year period was 550 gallons. The work
schedule for that year was 250 days, 16 hours
per day. An estimated 211,200 Btu's per hour
or $.42 per hour was the cost of the heat input.
The consumption average was 1.5 Ibs./hr.
2. Trichloroethane cold cleaning:
On July 28, after being down for two weeks, the
degreaser was filled with 1,1,1-trichloroethane to
a height of 30 inches from the top; the overall
height being 59 inches, with 185 gallons of
1,1,1-trichloroethane. This degreaser then was
run as a cold cleaning operation with all heat
turned off. On July 31, Halide measurements were
taken around the top of the vapor degreaser and
within the lip exhaust of the degreaser itself.
-------
The Halide meter used was a Gas Tech Model 1192
calibrated 7-30-75. The recorder was a Rustrak
Model P4002 with a chart speed of one inch per
minute. On September 3, 1975, the solvent level
was down three inches and 18 gallons of make-up
solvent was added. This equates to an average
consumption of 0.49 Ibs./hr. of 1,1,1-trichloro-
ethane. Halide measurements were again taken,
this time from a set position at the lower right
side of the degreaser. On September 18th, the solvent
was too dirty and had to be disposed of.
This equates to an additional .3 Ib./hr. consumption
for a total consumption of .79 Ib./hr. of 1,1,1-
trichloroethane with cold cleaning. While it is not
documented in this report, a common commerical still
would be expected to get about 70% of the .3 Ib./hr.
back to give an estimated consumption of .59 lb./
hr. of 1,1,1-trichloroethane cold cleaning with
distillation.
3. Cold cleaning with plastic balls as vapor depressants:
After the readings, plastic balls were added to
the degreaser to determine their effort as a vapor
depressant and readings were again taken.
-------
On September 10, 1975 we lost our balls at
Prestolite. They interferred too much with
production and had to be removed.
Summary: Consumption - All on same equipment
1,1,1-trichloroethane 1.5 Lb./Hr. Degreasing
1,1,1-trichloroethane .8 Lb./Hr. Cold Cleaning
-------
July 28, 1975
1. Fan Off Lower Right Corner
Meter
PPM
15
25
27
45
85
15
53
85
30
60
70
210
710
30
300
710
Range 30-710 PPM
Average 265 PPM
2. Top Right Corner
17
38
40
28
1
80
23
33
45
35
150
160
80
0
630
55
100
210
Range 0-630 PPM
Average 158 PPM
3. Lower Left Corner
100
75
100
58
70
75
70
1000
560
1000
360
490
560
490
Range 360-1000 PPM
Average 637 PPM
-------
4. Upper Left Corner >1000 PPM
Fan On Air Speed Ft./Min. Lips 2" x 50"
Lower Lip Upper Lip
Left 260 260
Right 320 375
Middle 350 315
Average Air Velocity 313 Ft./Min. Facial Velocity
2/12 x 50/12 x 2 = Sq.1 Face 1.39
= 435 Cfm
5. Readings Taken At Inlet To Lip
PPM
0
1
15
20
0
0
5
5
0
1
30
50
0
0
5
5
-------
September 3, 1975
Prior Addition of After Addition of
Plastic Balls Plastic Balls
Readings PPM Readings PPM
60 380 45 210
100 1000 55 320
65 430 65 430
75 560 40 160
45 210 40 160
60 380 82 660
55 320 100 1000
40 160 100 1000
60 380 60 380
100 1000 80 630
50 260 36 120
100 1000 52 290
55 320 80 630
85 710
492 Av. 479 Av.
3% Less
Evaporation Rate Calculations for Cold Cleaning (Based on
Halide Meter Readings - Does Not Include Drag-out):
350 PPM = 1900 Mg./M3
11.4 PPM = 62 Mg./M3
62 Mg. x 10~3 G./Mg. x 10"3 Kg./G. = 62 x 10~6 Kg. = 1.37 x 10~4 Lb.
1 M3 = 35.3 Ft.3
1.37 x 104 - 35.3 = 3.88 x 106 Lb./Ft.3
11.4 PPM = 3.88 x 10"6 Lb./Ft.3
3.88 x 10"6 Lb./Ft.3 x 435 Ft.3/Min. x 60 Min./Hr. = .101 Lb. Hr.
Due to Evaporation
Average Hourly
Consumption
-------
,1" = 1"
Dimensions: |
Width 30.5"
Height 59" I
Length 48.3" ,
To Solvent From Top 29.5*
Vents 2" x 50"
\
1 Plastic Roll Lid
2 Lip Vent
-------
Meter
90 -
80 -
70 -
60 -
50
40 -
30
20
10 -
HALIDE METER CALIBRATION CURVE
x
200
400
600
PPM
i i i r—
800 1000
-------
-o
-o
rn
Z
O
n
XI
-------
APPENDIX - C7
STUDY TO SUPPORT NEW SOURCE PERFORMANCE
STANDARDS FOR SOLVENT METAL CLEANING OPERATIONS
Evaluation of Refrigerated Freeboard Chillers
Schlage Lock
Rocky Mount, North Carolina
Prepared By:
J. C. Bellinger
U. S. Environmental Protection Agency
Prepared For:
Emission Standards and Engineering Division
Office of Air Quality Planning
U. S. Environmental Protection Agency
-------
John'Bollinger
May 1976
Test Report:
Refrigerated Freeboard Chiller
on a Monorail Vapor Degreaser
Test Site:
Schlage Lock Company
Rocky Mount, N.C.
Contents
I. Summary 1
II. Experiment Design 2
III. Data Results 3
IV. Equipment Description _._ .... 5
V. Control Efficiency from Solvent Purchase Records 7
VI. Conclusions 8
VII Appendices
A. Energy Consumption 9
B. Capital Cost Data : 10
C. Savings/Cost Ratios 11
-------
I. Summary
This test measures the control efficiency of a refrigerated freeboard
chiller on a conveyorized vapor degreaser using perchloroethylene. The
refrigerated freeboard chiller uses finned coils with subfreezing coolant
to establish-a cold air "blanket" over the hot solvent vapor zone. (The
cold air retards convection of hot solvent laden air and the diffusion of
the vapors.) The vapor'degreaser is a U-bent monorail. The emission rate
was measured with and without the freeboard chiller.* Records of the
solvent added indicate the solvent emitted. Operating the chiller and
turning off the exhaust yielded a control efficiency of about 60 percent.
Additional test runs could not be performed to confirm the results. Solvent
purchase records showed a control efficiency of roughly 45 percent.
*The degreaser exhaust was used only when the chiller was off. Employees
complained of solvent smell when the exhaust was turned off without the
chiller on. The exhaust increases the emission rate significantly.
-------
II. Experiment Design
The method of measuring solvent emission is material balance. Basically,
the solvent added to the degreaser during the test period, equals the
solvent emitted. A solvent flow meter measures the gallons of solvent
added into the degreaser. It is necessary that the solvent content of the
degreaser at the beginning of the test period equals the content at the end.
Plant personnel record the amount of workload each day in terms of number
of trays. Thus the solvent emission rates can be adjusted to reflect
equal workloads.
Complications arouse when equalizing the solvent content. When equalizinq
the solvent content, two procedures should be followed to avoid inaccuracies.
First, overflow the two higher sumps and level the lowest sump to a constant
depth. No'tYthat sumps if 1 and # 2 overflow into # 3, which.is the boiling
sump. Second, fill the external still to the maximum level. This second
step is an involved maneuver. Specifically, the operator shuts off the
external still immediately after the still pump activates. He waits for
the still condensate stream to cease. Then, he turns on the still so that
the pump automatically fills up the still to the full level and shuts off.
Next, he turns off the still again and proceeds to level the 3 sumps as
previously described.
Previously given information led the measurement personnel to believe
two incorrect premises: (1) sump # 2 was the lowest one and (2) the
variation in solvent content of the still was negligible. Consequently, test
periods # 1 to # 4 contain excess inaccuracy. Inconsistencies resulting
from the previous information necessitated consulting the manufacturer of
the degreaser, Detrex Inc. Thus, only test periods # 5 and # 6 are performed
-------
with the accurate sump leveling procedure, as described in the previous
paragraph.
An upset which probably caused the biggest error in data is a malfunction
of the primary condensing system. The thermostatically operated valve
which automatically adjusts the flow of cooling v/ater was stuck. During
the hot summer months the cooling water was too warm because it was cooled
by a towep. This augmented the cooling problem. Consequently, the vapor
level rose up to the "Cold Trap" coils, or beyond it if it were turned off.
The condensing system was not repaired until the end of test # 4. Thus the
emission rates for tests # 1 to # 4 are all abnormally high.
III. Data Results
The results of the material balance measurements are present on the
following table. Note that the workload is measured in number of trays.
Thus the column to the farthest right can express an emission rate adjusted
for a constant workload (500 trays per workday*).
Only test periods # 5 and 6 are reasonably accurate. The two periods
included a proper sump leveling procedure and a repaired primary condensing
system. Test periods # 1 to # 4 have less percision because the sumps
were not leveled properly. More important though, the primary condensing
system was deficient. This deficiency caused the emissions to be roughly
twice the normal rate.
There are two approaches to calculating the emission control efficiency
for the "Cold Trap." First, all six test periods could be considered.
Second, tests # 1 to # 4- could be excluded, because they did not use
the proper sump leveling procedure and had a defective primary condensing system
*0nly workdays are counted. The evaporation during shutdown days is comparatively
Insignificant. A plant engineer measured less than 1/32 inch loss in sump
level during a 3 day weekend.
-------
"Cold Trap" on Monorail Vapor Degreaser
Data Summary
If 500 tray/
wrkdy
Test
Period
1*
2*
3*
4*
5
5a
5b
6
6a
6b
6b
Dates
Spanned
7/10 to 7/21
7/22 to 8/4
9/3 to 9/15
9/16 to 9/29
10/14 to 10/26
10/14 to 10/20
10/21 to 10/26
10/28 to 11/10
10/28 to 11/3
11/3 to 11/10
1.1/3 to 11/7
Workday
6
8
7
8
10
5
5
11
•5
6
5
No. of Trays
(trays per wrkdy)
2590
(432)
3886
(486)
unknown
unknown
6415
(640)
3615
(720)
2800
(560)
-
-
4175
£ (700)
3479
(697)
% Knobs Cold Trap
68% ON
61% OFF
ON
ON
72% OFF
73% OFF
71% OFF
ON
ON
ON
Solvent .
Consumed
(gal )
95
307
110
140
250
137
113
108
46
62
Emission
Rate
gal /wrkdy
15.8
38
15.7
17
25
27.4
22.6
9.8
9.2
10.3
Adjusted
Em. Rate
gal/500 trays
£18.3.
£39
£16
a$i7
19.5
19.0
20.2
* 7.5
£ 7.7
7.4
*Judge to be defective tests
-------
Thus the second calculation should be the most accurate.
Calculations of Control Efficiencies:
1) Considering tests # 1 to i 6
Uncontrolled emission = 39 + 19.5 = 29.3 gal/day*
Controlled emission = 18.3 + 16 + 17 + 7.5 = 14.7 gal/day
Qpntrol efficiency = Emisison reduction = 29.3 - 14.7 = 50%
Uncontrolled emission 29.3
2) Considering only tests # 5 and # 6
Uncontrolled =19.5 gal/day
Controlled =7.5 gal/day
Control Efficiency = 62% * 60%
Let 1t be noted that omitting just test # 2 would not yield a reasonable
result. At first it appears that only # 2 is far too high to be valid data.
But actually tests #1 to # 4 are all too high because of a defective primary
condensing system.
IV. Equipment Description
Refrigerated Freeboard Chiller:
The refrigerated freeboard chiller uses subfreezing coolant. The
subfreezlng chiller design is patented by Auto Sonics Inc. (Chillers that use
above freezing coolant are not patented.) Two levels of copper finned cooling
coils run the inside perimeter of degreaser a few inches above the primary
condensing coils. The cooling coils chill the air immediately above the
air/vapor interface in order to retard convection of solvent laden air,
diffusion and air/vapor mixing.
* All rates are adjusted to reflect a constant workload of 500 trays per
workday.
-------
Specifications:
- Manufacturer: Auto Sonics Co., Norriston, Pa.
- Model: Cold Trap I 302
- Compressor horsepower: 3 Hp
- Coils: total perimeter: 41 ft.
number of levels: 3
type coil: copper, finned
- Coolant temperature: 0 to "20 °F
Monorail Vapor Degreaser:
The monorail carries the parts in a U-bend pattern through the vapor
zone and solvent sprays. The parts to be cleaned are door knobs and their
base plates, which are loaded on tray racks. The parts experience a prewet
spray, hot spray, cool rinse and then a final spray. The degreaser has three
sumps: (#1) hot spray sump, (#2) cool rinse spray sump and (#3) boiling sump.
The solvent overflows from sumps # 2 to. # 1 to # 3. The solvent is perch!oro-
ethylene. There are two sets of primary condensing coils: one around the
entrance and one around the exit. Both follow the perimeter of a rectangle
90" by 33". The degreaser is manufactured by Detrex Chemical Inc. and is
a Chainveyor series 200 and Model VM 325-1S.
Still:
The still is an integral part of the degreasing system. It consistently
draws dirty solvent from the degreaser sumps and returns the
solvent distillate to the spray sumps. Because it is very important to
totally clean particles off the brass parts before laquering, the rinses must
use the pure distillate. Detrex Inc. manufactured the still and the model
is S 350 S.
-------
V. Control Efficiency from Solvent Purchase Records
A. Emission rate without the refrigerated freeboard chiller (& with exhaust
on)
1. Time Period: 2/15/73 to 11/18/74 = 610 days + 7 = 87 weeks
2. Solvent bought: 14,130 gal : 87 - 3 vacation wks = 84 weeks
3. Emission rate = J|J15jl = 168 gal/wk
B. Emission with Cold Trap
1. Time Periods
#1 3/25 to 7/9/75 = 14.2 wk; 1470 gal. (approximately consumed)*
92 7/9 to 10/10 = 13.4 wk; 1500 gal.
#3 10/10 to 1/7/76 = 10.8 wk; 2040 gal.*
#4 1/7 to 3/25 = 12.0 wk; 1540 gal.
2. Workloads - (Very roughly estimated)
#1 Assume 40 hr/wk
#2 Assume 47 hr/wk: (45 hr/wk x 83 days + 60 hr/wk x 10 days)-?- 93 days
46.6
#3 Assume 72 hr/wk: (60 hr x 21 + 80 hr x 31)-r- 52 = 72-hr/wk
#4 Assume 80 hr/wk
3. Gallons per week
#1 1470 gal/14.2 (40 hr.wk.) = 104 gal/40 hr.wk.
n 150 gal/13.4 (40 hr.wk.) = 111 x 40/47 = 94 gal/40 hr.wk.
#3 2040 gal/10.8 (75 hr.wk.) = 189 x 40/72 = 105 gal/40 hr.wk.
#4 1540 gal/12.0 (80 hr.wk.) = 128 x 40/80 = 64 gal/40 hr.wk.
Average = 104 + 94^ 105 + €4 = Q2 gal/4Q hp>wk
*1 week vacation
-------
C. Results
1. Calculated Control Efficiency
168-92 _ Ac*
168 45%
2. Accuracy: The main cause of inaccuracy in the results is the
differences in workloads before and after the Cold Trap installation.
The number of hours per day and the number of workdays per week both
varied greatly. The average workhours per week can only be roughly
approximated.
3. Assumptions:
a. Bill McPhillips estimated 40 hrs/wk up to June 1975, 45 hrs/wk
for July to September, 60 hrs/wk for October and 80 hrs/wk for
Nov. (1975) and on.
b. One week shutdown in July and one week in December, each year.
c. The solvent tank is filled to an equal level each filling.
VI. Conclusions
The control efficiency resulting from using refrigerated freeboard
chiller on a monorail vapor degreaser and turning off the degreaser exhaust,
was measured at approximately 45% to 603. emission reduction. This value is
derived from solvent purchase data and from the test periods which used an
accurate method of sump leveling and had satisfactory primary condensation.
The uncontrolled emission rate averaged roughly 20 gallons of solvent for an
8 hour work workday,* or 35 Ib/hr.
*This applies after the primary condensing system was repaired.
8
-------
Appendix A - Energy Consumption
Refrigerated Freeboard Chiller
Averages about 2.8 kw £9,600 Btu/hr
(6.4 amp x 440 volt = 2820 watts)
Monorail Degreaser:
Steam for sump heat = 485,000 Btu/hr
Elect, for pumps = 11.8 kw = 40,000 Btu/hr
Still:
Steam = 450,000 Btu/hr
Elect, for pumps = 6 kw = 20,000 Btu/hr
Percent Energy Required for Chiller:
2% of Degreaser energy
1% of Degreaser and Still energy
Note: The energy requirement for the still is unusually high,
because a high rate of distillate solvent is needed for this
cleaning application.
-------
Appendix B
EXECUTIVE OFFICES
AND FACTORY
BAYSHORE BOULEVARD
P.O. BOX 332-1
SAN FRANCISCO 94119
- Capita] Cost Data
ROCKY MOUNT
U.S. HIGHWAY 301 NORTH
P. 0. BOX 552
ROCKY MOUNT, N. C. 27801
July 22, 1975
COST OF DEGREASER
TELEPHONE
(919} 446-'.905
$11-61450
Degreaser, Detrex Model No. VM-325-1S-
With S-350 Still
Install Detrex Degreaser —
Electrical Installation Detrex Degreaser-
Power Controls for Degreaser
Cost of Cold Trap, Model 302-
Cost of Installation
Total
Total
$40,884.37
591.00
1,227.00
1,585.00
$44,287.37
7,944.00
350.00
$8,294.00
Degreaser began operation in April , 1972, Cold Trap installed
January, 1975.
Richard T. Wilcox
Plant Engineer
10
-------
Appendix C - Savings/Cost Ratios
High Control Efficiency = 60%
Savings: 60% x 20 2lL x 250 day*x 2.16 $ = 6480 $/yr.
day year gaT
Annualized Cost:
Capital cost of "Cold Trap" = $8,294 (Includes shipping & installation)
Assume 15 year Hfe (x 0.1315)
Annualized capital cost = 1090 $/yr
Electrical cost 2.8 kw x 3 i x 8 hr x 250 day = 170
lew day
Maintenance cost = 4% x $8,300 = 330
Insurance = 2% x 8,300 = 165
T775~$/yr
Savings/Cost ratio = 6480/1755 = 3.7
Moderate Control Efficiency = 40%
Savings: 6480 x 4J) = 4320 $/yr
60
Savings/Cost ratio = 4320/1755 = 2.5
Breakeven Control Efficiency
= 1 -»• E = 16% control efficiency
*Based on a 5 day workweek. Actually the plant tested, Schlage Lock, has been
running only 4 days per week, but 5 days per week is more typical of decreasing
plants. » a
11
-------
TJ
m
Z
X
n
00
-------
APPENDIX - C8
STUDY TO SUPPORT NEW SOURCE PERFORMANCE
STANDARDS OF SOLVENT METAL CLEANING OPERATIONS
EVALUATION OF CARBON ADSORPTION RECOVERY
Super Radiator Corporation
St. Louis Park, Minnesota
PREPARED BY:
D. W. Richards
The Dow Chemical Company
PREPARED FOR:
Emission Standards and Engineering Division
Office of Air Quality Planning
U. S. Environmental Protection Agency
-------
Summary
An evaluation of a carbon adsorption unit in support of
an open top degreaser was conducted at Super Radiator
Corporation. Due to some unique circumstances, the use of
the adsorber was found to cause a slight increase (~8%)
in solvent emissions. Even though this situation was
explained and corrected using the data obtained during
the test period, it does point out what can occur in an
actual industrial operation. The need to periodically
monitor the performance of this control device is quite
evident. Due to the manner in which the adsorber was
operating, it was increasing plant operating costs by an
estimated $8500 per year. Also, the solvent recovered
by the carbon adsorber did not contain any stabilizers
which means that it must be mixed with new solvent or
restabilized before reuse.
-------
Objective
The purpose of this test program is to evaluate carbon
adsorption as a means of controlling solvent emissions
from metal cleaning operations. The information needed
for this evaluation includes determining:
1. the efficiency of this device in reducing
solvent emissions to the atmosphere,
2. the cost/benefit relationship of this
emission control system/
3. the energy requirement of the emission
control system,
4. any alternate emissions created by the
emission control system.
This data base is being developed to forecast the
magnitude of emission reductions which can be achieved
nationally and the effect on businesses involved. This
information combined with the results of other testing
will be used to design emission control regulations
which effectively limit air pollution and are practical
for industrial application.
-------
Introduction
Carbon adsorption has been offered to solvent metal
cleaning users since 1958. The efficiency of carbon beds
to capture solvent vapor from low concentrations in air
has been estimated to be greater than 95 percent. The
typical efficiency of carbon adsorption in reducing solvent
consumption (controlling solvent emissions) is commonly
reported to be in the range of 50-60 percent. Inspite of
this, only a small fraction (1.5 percent) of the industrial
solvent metal cleaning users surveyed*, reported that they
were using carbon adsorption.
Super Radiator was considered as an emission control test
site on the recommendation of Vic Manufacturing Co. (carbon
adsorption manufacturer). This operation was selected
for evaluation because it was felt that it would demonstrate
typical industrial use of carbon adsorption in support of
an open top degreaser. The unit was installed in December,
1971.
Before the evaluation began, service personnel from Vic
Manufacturing inspected the carbon adsorption equipment to
be sure that it was functioning properly. Severe solvent
"breakthrough" was discovered and two more desorption cycles
were added to alleviate this problem.
*See Appendix A
-------
Equipment
The basic metal cleaning process employed is a Detrex VS
800 (61 x 12') open top degreaser and a carbon adsorption
unit operated with perchloroethylene. The degreaser is
a gas-fired unit (See Figure 1 below).
The carbon adsorption unit is a model 554 AD. Solvent-
laden air is drawn into the unit through a 14" duct.
After passing through the carbon bed it is discharged outside
through another 14" duct. When the carbon adsorption unit
was installed, it was also necessary to install a 10 h.p.
boiler to provide steam for the desorption cycles since
no plant steam was available.
Behind the degreaser are two solvent storage tanks (21
diameter x 4' long) which are used during degreaser clean-
outs. The degreaser contents are boiled down to the heating
tube; pure solvent is condensed and transferred to the
storage tanks; and any remaining "sludge" is removed from
the degreaser. The solvent from the storage tanks is then
returned to the degreaser and new solvent is added to bring
the degreaser back to its operating level. A schematic of
the metal cleaning system is presented in Figure 2.
-------
Figure 1
OPEN TOP DEGREASER
SUPER RADIATOR CORPORATION
Exhaust Duct To Carbon Adsorber,
6'
\
,
\«^,\/\/\/\'\'\/\'\'\/\/\
t+~i
i
r
i
\
Exhaust
Stack
Gas Inlet
^.^^•^V^WSj^^^^S^S^*^* -*>i
Solvent Level
li
Gas-Fired "U" Tube
-------
Figure 2
METAL CLEANING SYSTEM
SUPER RADIATOR CORPORATION
New Solvent From Bulk Storage
1
Degreaser
Solvent-Laden Air
Carbon
Adsorber
To
Atmosphere
Damper
Recovered Solvent
Boiler
Steam ( )
-+-\^
To Atmosphere
Solvent Storage Tanks
-------
Experiment Design
This test was designed to obtain two weeks of data
with the carbon adsorber operating and one or more weeks
with it off. For reasons which will be discussed later,
the carbon adsorber was operated for two weeks and turned
/
off for three weeks during the actual test period. The
test began on May 19, 1975 and concluded in August. Since
all of the new solvent added to the degreaser comes from
a bulk storage tank, a solvent meter was installed in the
delivery line so that solvent additions could be measured
accurately. A gas meter was also installed in the gas line
to the degreaser so that the energy input could be determined,
The difference in solvent inventory was determined by
measuring the amount of solvent necessary to bring the
level in the degreaser to a given mark on a "dip stick".
The degreaser was leveled each morning before any work was
processed. The amount of work processed was also recorded
daily.
-------
Data Discussion
During the initial three weeks of testing, the carbon adsorber
was operated for two weeks and turned off for the final week.
Since the solvent consumption decreased the week the carbon
adsorber was turned off, two additional weeks of data were
obtained with the adsorber off. Once again, a decrease in
solvent consumption was observed. Several parameters were
found to be responsible for this situation, some of which
are unique to this metal cleaning system. Each will be
discussed separately as will the method used to determine
solvent consumption.
Solvent Consumption - As discussed previously, all new solvent
added to the system was added through a solvent meter to the
degreaser. Solvent was added each morning (before any heat
was applied) to the same mark on a wooden "dip stick". The
ability to fill to this mark was estimated to be + 1/8".
Since the mark was 11 1/2" from the bottom of the stick, this
is equivalent to a possible error of + 1.0%. The average new
solvent additions to the degreaser when the carbon adsorber
was operating was 49.0* gals./day and 36.6* gals./day when the
adsorber was turned off. The average work processed by the
degreaser when the adsorber was on was 8995 Ibs./day and
7283 Ibs./day when it was off. This is equivalent to
*New solvent additions to the degreaser were adjusted to
account for the fact that varying amounts of kerosene were
present in the degreaser during the test.
-------
0.00545 gals./lb. with the adsorber "on" vs. 0.00503 gals./lb,
with it "off". The emission control efficiency is therefore:
.00503 - .00545
x 100% = -8.35%
.00503
The negative efficiency calculated by this equation is
obviously not representative of the degree of emission
control possible with a carbon adsorber. For this system,
however, this number is believed to be valid. Some of the
key factors which are felt to be responsible for this
negative percentage are as follows:
1. Minimal Freeboard Height - During the test period
it was observed that the vapor level in the degreaser
was allowed to continually rise to within 3" to
8" of the lip exhaust. Therefore, the effective
freeboard height during the test period was 3" to
8". Standard degreasing practice calls for a
desirable operating freeboard height of 0.5 to
0.6 times the inside width (72") of the degreaser,
which in this case would be 36" - 43". If
this cannot be practically achieved, a minimum
freeboard height of 12" is recommended. Maintaining
an adequate freeboard height in any open top degreaser
-------
is important if solvent losses due to drafts in the
plant are to be minimized. Another important reason
for maintaining a greater freeboard height in this
particular degreaser relates to the exhaust fan
velocity which is discussed in paragraph No. 2.
2. Exhaust Fan Velocity - When the carbon adsorber
is operated, both the degreaser lip exhaust fan
and the adsorber fan are used. The inlet velocity
to the adsorber which is developed with both
fans operating (and both beds adsorbing) was
found to be 2496 ft./min. (2671 cfm). The
velocity dropped to 1460 ft./min. (1562 cfm)
when one bed was desorbing. When the degreaser
is exhausted to the atmosphere, only the degreaser
fan is used. This fan was found to operate at
only 557 ft./min. (741 cfm).
Since the vapor zone was allowed to rise to within
3" to 8" of the degreaser lip exhaust, large
quantities of solvent were available to be drawn
to the carbon adsorber at 1562-2671 cfm. Normally,
this solvent would simply be recycled to the degreaser,
but in this test much of it "broke through" (see para. 3)
the beds and was lost to the atmosphere. This resulted
in an abnormally high solvent consumption during the
time the adsorber was operated. Also, the amount
-------
of solvent lost to the atmosphere when the adsorber
was not operated was actually less because the
exhaust velocity was only 741 cfm.
3. Vapor Concentration Measurements - The solvent vapor
concentrations in the inlet and outlet ducts of the
carbon adsorber were measured from 9:45 a.m. to
3:15 p.m. on August 19, 1975 with a Gas Tech halide
meter. While much was learned from this data, no
attempt will be made to plot concentrations in the
outlet duct. The reason for this is that during the
desorption of both beds, readings were obtained that
were so high that they were off of the curve which
is used to convert meter readings to parts per
million of solvent. For the same reason, the average
efficiency of the carbon beds could not be determined.
Concentration measurements in the outlet duct did in-
dicate, however, that "breakthrough" was occurring
continuously. With both beds adsorbing, measurements'
varied from 160-770 ppm throughout the day. Also,
when Bed "A" was desorbing, measurements in the outlet
duct ranged from 35 ppm to greater than 2000* ppm.
When Bed "B" was desorbing, readings ranged from
12 ppm to greater than 1500* ppm. All of these
measurements were made at the end of the test period.
*Estimated by extrapolating the curve for converting meter
readings to ppm perchloroethylene.
-------
As discussed earlier, the equipment manufacturer
inspected the carbon adsorber before the test
began, observed breakthrough, and added more
desorption cycles to alleviate the problem. There-
fore, no additional checks for breakthrough were
deemed necessary until the solvent consumption
data indicated negative adsorber efficiency.
-------
Concentration measurements in the inlet duct were much higher
than those in the outlet duct and therefore even further
off of the conversion curve. Extrapolation estimates for
inlet concentrations while the degreaser was idling range
from 2,000 ppm to 5,000 ppm. Once again, it would be
unrealistic to attempt to develop an average bed efficiency
based on these estimates.
Solvent recovered by the carbon adsorber averaged 476.75
Ibs./day (35.40 gals./day). This data was obtained by
collecting solvent from all six of the one hour desorption
cycles over a 24 hour period. The average solvent collected
per desorption cycle is, therefore, 79.46 Ibs. or 5.90 gals.
These recovery rates are not indicative of the maximum
capabilities of the carbon adsorber as predicted by the
manufacturer. Manufacturer estimates for solvent recovery
with this size adsorber are 200 Ibs./hr. (14.85 gals./hr.)
of perchloroethylene.
The utilities required to operate the carbon adsorber are
steam, water, compressed air, and electricity. Steam conden-
sate was collected and weighted for four separate one hour
desorption cycles. The weights were: 262.75 Ibs., 272.25
Ibs., 244.25 Ibs., and 288,50 Ibs. (average = 266.94 Ibs.).
Using 1,000 Btu's per pound of steam, 267,000 Btu's are
required per desorption cycle. Thus, 267,000 Btu's/desorption
x 6 desorptions/day x 240 days/yr. = 384 x 10 Btu's/yr.
-------
The carbon adsorption unit's condenser is estimated to use
16 G.P.M. by the manufacturer. Since the water only flows
during the desorptiori cycles/ the total water requirement
is 16 gals./min. x 60 min./hr. x 6 hr./day x 240 days/yr.
= 1,380,000 gals. The carbon adsorber fan is rated at 10
horsepower. This is equivalent to: 10 Hp x .746 KWH/Hp
hour x 17 hrs./day'x 240 days/yr. = 30,440 KWH/yr. An
energy balance for one bed of the carbon adsorber (adsorption
and desorption cycles) is summarized in Table No. 1. Radiation
losses through the shell of the adsorber bed are not included
in order to simplify the overall balance.
-------
The replacement cost for the 554AD carbon adsorption
unit studied during this test is estimated to be $13,990
by Vic Manufacturing. Since plant steam was not available
at this test site, a boiler was also purchased with the
carbon adsorber. The estimated replacement cost for
this is $4000. A special platform was also constructed
above ground level to house the adsorber and boiler and
to save on floor space. The estimated cost of this is
$3300, bringing the total replacement cost for this
equipment to $21,290. This information was used to
develop Tables 2 and 3. The assumptions involved in the
development of these tables include a zero return on
investment; charging the construction of the elevated
platform as direct capital and therefore not charging
any floor space against the carbon adsorber; the omission
of costs for other facilities which are required, but
already exist; and the omission of other minor costs such
as heating, lighting, janitorial, etc. Based on these
assumptions, the total operating cost per year for this
model 554AD carbon adsorber is calculated to be $6293
(see Table 3). If a price of $0.16/lb.* ($2.16/gal.) is
used for perchloroethylene, then 2913 gallons must be
recovered per year by the carbon adsorber to offset the
total annual operating cost. Based on plant purchasing
records through July, 1975, a total of approximately
*Chemical Marketing Reporter, July 7, 1975, Schnell
Publishing Company
-------
12,500 gallons of solvent will be used during 1975.
Using the emission control efficiency of -8.35% observed
during the test, this would mean that approximately 1000
gallons of solvent would be saved per year by not using
the adsorber. This would also mean that, under these
operating conditions, the adsorber would actually add ~$8500/
yr. to this company's operating costs [$6293 (from Table 3)
+ 1000 gals, x $2.16/gal7| .
While this is obviously not a desirable situation, it can
be remedied by making relatively minor changes in the
operation of the degreaser and carbon adsorber.* This does
demonstrate, however, the need to periodically check the
performance of this type of emission control device using
more sophisticated techniques than just measuring the amount
of solvent recovered. One such technique would be to check
the outlet duct of the adsorber to be sure that no significant
solvent "breakthrough" is.occurring.
Samples of solvent recovered from the carbon adsorber were
obtained at various times during the test and analyzed for
stabilizer content. Perchloroethylene is not as highly
stabilized as most of the chlorinated solvents, but depletion
*Changes were made in the operation of the adsorber after the
"breakthrough" problem was discovered. Unfortunately there
was not time to collect data for use in this study, but reports
from Super Radiator indicate that the adsorber is now operating
on a profitable basis.
-------
of the stabilizers can result in considerable problems. No
stabilizers were found in any of the samples of solvent
recovered from the carbon adsorber. Before this solvent
can be reused, therefore, it must be blended with an adequate
amount of new solvent or replenished with a stabilizer concentrate,
The gas flow rate to the degreaser was checked several times
during the test and averaged 51.87 ft. /min. Since the
degreaser is operated approximately 14 hrs./day, the daily
usage rate is: 51.87 ft. /min. x 1000 Btu's/ft. x 60 min./
hr. x 14 hr./day = 43.6 x 10 Btu's day. The steam requirement
for the carbon adsorber is 267,000 Btu's/desorption cycle
x 6 cycles/day = 1.6 x 10 Btu's/day. Expressed in the same
units, the energy required by the carbon adsorber fan is
10 hp. x 2545 Btu's/hp. x 6 cycles/day = 0.15 x 106 Btu/day.
The total energy demand of the carbon adsorber is, therefore,
1.75 x 10 Btu's/day. This represents an energy increase of
only about 4% over that required for the degreaser*.
*The energy required for the small degreaser lip exhaust
fan was considered to be negligible and not included in
the calculation.
-------
Conclusions
1. For this particular metal cleaning system,
solvent emissions were actually reduced when
the carbon adsorber was not operated, based
on new solvent additions to the degreaser. It
is important to recognize that this reflects the
effects of several different parameters, many of
which are unique to this cleaning system. This
demonstrates, however, that this type of emission
control device requires at least occasional
checks to assure that it is functioning properly.
2. Since the solvent vapor level in the degreaser
was maintained so near the lip exhaust, large
quantities of solvent were drawn into the adsorber,
This compounded the existing "breakthrough"
problem and magnified the solvent losses consid-
erably. As a result, operation of the adsorber
in this manner will add ~$8500/yr. to the plant
operating costs.
3. Since plant steam was not available at this test
site, it was necessary to purchase a boiler to
be used in conjunction with the carbon adsorber,
-------
which demonstrates that additional costs may be
incurred if this method of controlling emissions
is selected.
4. Perchloroethylene recovered by the carbon adsorber
contained no measurable amount of stabilzer and
must be blended with new solvent or restabilized
before it is used again.
-------
TABLE 1
Energy Balance On Carbon Adsorber (1 Bed)
Adsorption (Solvent Recovery)
Input Output
Solvent Latent Heat (79.46 Lbs.) 7,200 BTU's
Air Flow At 2671 CFM 7,200 BTU's
Desorption
Steam (267 Lbs.) 267,000 BTU's
Heating Tank ('2000 Lbs.) . 31,240 BTU's
Heating Carbon (1000 Lbs.) :28,400 BTU's
Condenser Water (16 GPM) 207,360 BTU's
Vaporizing Solvent*
7,200 BTU's 7,200 BTU's
Adsorption (Drying and Cooling Bed)
Cooling Tank 31,240 BTU's
Cooling Carbon 28,400 BTU's
Air Flow At 267.1 CFM 59,640 BTU's
*Vaporizing solvent from the bed requires heat from steam, but
this same heat is given up to the condenser water when the
vapor is condensed to liquid solvent.
-------
TABLE 2
Carbon Adsorber
Building Space Not Applicable
(None is assigned since the carbon adsorber
and boiler are housed on an elevated plat-
form and do not occupy any previously existing
floor space.)
Direct Capital
Carbon Adsorber (13,990)^ ' _ QQn .„. M _ . . .
Boiler (4,000) ' $17,990 (Vic Manufacturing)
Platform $ 3,300
Shipping and Installation $ 2,700 (At 15% of Selling Price)
Total Capital $23,990
Cost/Annum = $3,154 (At 15 Yrs. Depreciation
Rate)*
*10% Interest Rate On Investment
-------
TABLE 3
MODEL 554 AD CARBON ADSORBER
OPERATING COST PER ANNUM
Capital
Equipment $3154
Building N.A.
Insurance (2%) •
Equipment $ 480
Building N.A.
Maintenance (4%) $ 960
Utilities
Steam $ 883 (384 M BTU's at $2.30/M BTU's)
Electricity $ 761 (30,440 KWH at $0.025/KWH)
Water $ 55 (1380 M Gals, at $0.04/M Gals.)
Compressed Air Nil
Labor Nil
Return on Investment 0
Total Cost/Annum $6293
-------
TJ
T)
m
r»
-------
APPENDIX - C9
STUDY TO SUPPORT NEW SOURCE PERFORMANCE
STANDARDS FOR SOLVENT METAL CLEANING OPERATIONS
Evaluation of Carbon Adsorption Recovery
at J. L. Thompson Co.
Waltham, Mass.
Prepared by:
K. S. Surprenant
The Dow Chemical Company
Prepared for:
Emission Standards and Engineering Div,
Office of Air Quality Planning
U. S. Environmental Protection Agency
-------
Summary
The evaluation of a carbon adsorption system as an emission
control for a cross-rod degreaser was conducted at J. L.
Thompson Co. This solvent recovery system had been
installed several years ago. After minor repairs, it
was found to be recovering 23 percent of the solvent
supplied to the degreaser. The major costs of operation
and the capital invested exceeded the value of the reclaimed
solvent annually by $817.
Solvent recovery data from earlier operation (when the
process volume was higher) indicated a greater volume of
solvent recovered. This solvent recovery rate would
provide a return of 2.58 times the annual costs of the
recovery system. However, all samples of recovered
solvent were unstabilized and unsuitable for reuse without
restabilization or dilution with fresh solvent.
-------
Objective
The purpose of this test program is to evaluate carbon
adsorption as a means of controlling solvent emissions
from metal cleaning operations. The information needed
for this evaluation includes determining:
1. the efficiency of this device in reducing
solvent emissions to the atmosphere,
2. the cost/benefit relationship of this
emission control system,
3. the energy requirement of the emission
control system,
4. any alternate emissions created by the
emission control system.
This data base is being developed to forecast the
magnitude of emission reductions which can be achieved
nationally and the effect on businesses involved. This
information combined with the results of other testing
will be used to design emission control regulations
which effectively limit air pollution and are practical
for industrial application.
-------
Introduction
Carbon adsorption has been offered to the solvent metal
cleaning users since 1958. The efficiency of carbon
beds to capture solvent vapor from low concentrations
in air has been estimated to be greater than 95 percent.
The typical efficiency of carbon adsorption in reducing
solvent consumption (controlling solvent emissions) is
commonly reported to be in the range of 50-60 percent.
Inspite of this, only a small fraction (1.5 percent)
of the industrial solvent metal cleaning users surveyed*,
reported that they were using carbon adsorption.
J. L. Thompson was selected as an emission control test
site on the recommendation of the carbon adsorption
supplier, Hoyt Manufacturing. During the preliminary
inspection of the equipment, a weekly solvent recovery
rate of 50 gallons was reported along with a solvent
consumption rate of about 50 gallons per week. This
level of solvent recovery or emission control was
judged to be worthy of evaluation.
The initial study time revealed an equipment failure.
This failure caused the inlet damper to carbon Bed B
-------
to remain closed. Bed B was recovering no solvent.
Although some data was lost, the efficient efforts
of J. L. Thompson Co. and Hoyt Manufacturing restored
the carbon adsorber to operating condition in time
to permit brief testing.
In April 1968, this equipment was installed. Carbon
adsorption equipment has been reported to have an
operating life of 15 or more years. Thus, data on
seven year old equipment is both valid and needed.
This combined with the excellent records and coop-
eration of J. L. Thompson, the metal cleaning process
equipment (a cross-rod degreaser) and the process
solvent (trichloroethylene) continued to make this
test site a valuable source of emission control
testing.
*See Appendix A
-------
Equipment
The basic metal cleaning process employed is a three
sump cross-rod degreaser and still operated with
trichloroethylene. The equipment is steam heated.
Every other fixture is a rotating basket designed
to handle small metal parts, fasteners in this case.
The alternate fixtures process the tote pans for
the parts. Rotation is obtained by rack and pinion
design. See Figure 1.
The carbon adsorption system is a Model 536AD. A
three horsepower fan draws air through a 12 inch
duct from the left side of the operator loading
station at the degreaser. After adsorbing the
solvent vapor contained in the air in its passage
down through the carbon bed, the air is discharged
by another 12 inch diameter duct.
-------
Experiment Design
The plan for testing this emission control system called
c
for running the degreaser system for two weeks with the
carbon adsorption system operating and two weeks without
the carbon adsorber. The workload/ solvent consumption
and heat input data would be recorded weekly and compared.
Additional weeks of testing would be used if the data
under similar operating conditions did not compare
favorably.
This plan had to be abandoned due to two factors. First,
the carbon adsorber was found to be mal-functioning.
Second, the firm had arranged to convert the system
to methylene chloride during plant shut-down July 12
thru July 21, 1975. This change was being, made to
comply with local air pollution codes and required
considerable equipment changes. These equipment
changes could not be done when the plant was operating.
Some valuable data was obtained after repair of the
carbon adsorber. However, time did not permit data
collection without the carbon adsorber. For purposes
of evaluation, it will be assumed that the solvent
losses from the degreaser would be the same with or
without the carbon adsorption ventilation system. This
-------
assumption should be essentially valid. The air collection
plenum to the carbon adsorber is sufficiently removed from
the degreaser vapor zone that it would be unlikely to
cause increased losses of solvent vapor. This assumption
is not valid when the adsorber vent system is close
enough to disturb the vapor zone which is often the case
with open top degreasing systems.
-------
Data Discussion
Solvent Consumption
Production records were kept from May 19, 1975 to
July 11, 1975. During this time, an average of
3.50 tons of work was processed per day. Work-
load variations ranged from 3,338 pounds to 10,081
pounds per day. The parts are transferred in tote
boxes which are cleaned in the same degreaser. An
average of 81.1 boxes were processed per day. Their
estimated weight is about 22 pounds, thus, they
contribute to another 0.89 tons per day to the
work processed by the vapor degreaser. The total
metal cleaned (including the tote boxes) per day
was 4.39 tons per day.
A solvent meter was located between the main
storage tank and the degreasing operation. Meter
readings between May 19 and July 11 reported
374 gallons of solvent consumed. Thirty-seven
days of operation occurred during this interval.
Based on this information, the average solvent
consumption rate was 10.1 gallons per day or
50.5 gallons per week.
Solvent purchase records for the months of January,
February, March and April 1975 totaled 842 gallons.
-------
Thus, the solvent consumption rate based on that
17 week period was 49.5 gallons per week. Note
that both figures for solvent consumption make
no adjustment for solvent lost from the system as
still residues or solvent recovered by the carbon
adsorption and used again.
Solvent Recovery
The total solvent available for recovery is the
sum of all solvent additions needed to maintain
a constant inventory in the system. Although
only 50.5 gallons of new solvent was required
per week in this degreasing operation, an additional
5 gallons per week, on an average basis, was added
to the degreaser from the carbon adsorption system.
This includes both that portion of the operating
time when poor solvent recovery was being obtained
and after the carbon adsorption system had been
repaired.
The quantity of solvent removed from the system
with the still residues would not be available for
solvent recovery. A total of 1,987 pounds of
still residues were removed from the system during
the same test period. The average solvent content
of this material was 23 percent by volume. Conse-
-------
quently, 46 gallons of trichloroethylene were
removed from the system with these still residues
and were unavailable to be recovered by the carbon
adsorption system. On a daily basis, this amounts
to 1.2 gallons which were unavailable to be recovered
by the carbon adsorption system. Therefore, only
9.9 gallons of solvent were available for recovery
per day. This was determined as follows: 10.1 gallons
of fresh solvent per day plus 1.0 gallons/day from the
carbon adsorber minus 1.2 gallons/day removed from oil
residues. It is also valuable to note that the 46
gallons of solvent lost in the still residues represent
12 percent of the new solvent added to the system.
This percentile of solvent could not have been recovered
even if the emission control had performed at 100 percent
efficiency.
After repairing the carbon adsorber, the solvent
recovery was determined. From July 1, 1975 to
July 11, 1975, the quantity of solvent recovered
ranged from 2.01 to 3.52 gallons per day. The
solvent was measured by the level change which
occurred in the solvent collection tank (14 inches in
diameter). A factor of 0.667 gallons per inch
was used to convert to gallons. Two calibration
-------
checks were made by weighing the solvent and comparing
the level change in the tank. Both were within the +
0.008 gallons expected from measurements in error by +
1/8 inch. The average solvent collected per day was 2.53
gallons (30.4 pounds/day). An overall recovery efficiency
of 25.5 percent is obtained when this is compared to
the 9.9 gallons available for recovery.
Solvent Use Per Ton of Work
The total quantity of metal cleaned including both tote
boxes and the work pieces amounted to 4.39 tons per day.
When this is divided into the 374 gallons consumed during
the test (10.1 gallons per day), 2.30 gallons per ton of
metal cleaned is found to be the consumption rate on a
tonnage basis. Although not demonstrated in this testing,
the solvent consumption rate per ton is expected to
diminish as the work processed increases. In terms
of the capacity of the vapor degreasing machine, only
about 25 percent of the total work capacity of this
degreaser was being used during the test interval.
It is not uncommon to find metal cleaning equipment
being used at much lower than machine capcity. Over-
design of this equipment is common in order to avoid
having it become the limiting factor in plant production.
-------
Bed Efficiency
The solvent vapor concentrations in the exhaust duct
from the carbon adsorption system for a single day of
the test, July 11, 1975, are plotted on Figure 2. The
average vapor concentration in the exhaust was 10.6 ppm,
The elevated plateaus between 11:00 - 12:00 and
1:30 - 2:30 correspond to the desorption cycle for
Bed B and Bed A respectively. During the time that
one bed is desorbing, the other bed is trying to
do the work of both. Consequently, the higher
concentrations during the desorption cycles are
expected.
The concentrations of vapor being carried from the
vapor degreaser to the carbon adsorption beds were
measured several times during the course of the day
and varied only with whether the machine was processing
work or being maintained at idle. During the work
process periods the vapor concentrations varied from
37 to 145 ppm and had an average of 74 ppm. The
solvent vapor concentrations when no work was being
processed (idling) ranged between 37 and 50 ppm with
an average of 41 ppm. It is valuable to know that no
variations in the exhaust vapor concentration from the
-------
carbon adsorber were observed to correspond to variations
in the inlet composition or the work activity in the vapor
degreasing operation.
Although the vapor degreaser was processing work on and
off throughout the day, the hours of actual vapor degreaser
operation were estimated at two hours per day. When vapor
concentrations in the inlet duct to the carbon adsorber are
time-weighted for the hours of operation and idling, the
average daily concentration was found to be 49 ppm. The
bed efficiency was determined as follows:
Aver. Inlet Cone. (49 ppm) - Aver. Exhaust Cone. (10.6 ppm)
= 78%
Aver. Inlet Cone. (49 ppm)
The bed efficiency can be expect to increase as the vapor
concentrations in the inlet to the carbon adsorption
system increase. This is supported by examining the vapor
concentrations obtained earlier in the workday when the
inlet vapor concentration was 102 ppm and the average
exhaust concentration was 5 ppm. The bed recovery efficiency
during this time interval was 95 percent. Of course, the
recovery efficiency of the bed diminishes as the bed approaches
its saturation capacity for the solvent. The bed
-------
recovery efficiency can also be expected to decline
with time over years of use due to the adsorption
of materials on the bed which cannot be desorbed.
This diminishes the capacity of the bed for solvent.
This carbon adsorption system is seven years old,
as noted earlier.
Vapor concentrations in the inlet duct of the carbon
adsorber can be expected to vary greatly with the type
of work processed. Some work parts carry decidely
more solvent out of the degreaser in spite of the
rotary action of the fixtures. For instance, on
May 28, 1975, vapor concentrations going to the carbon
adsorber varied between 50-86 ppm without work being
processed (average 65 ppm). The concentrations with
work processed from 80 ppm to ~1500 ppm (average 480 ppm)
The average inlet concentration on a daily basis was
estimated to be 169 ppm. The exhaust concentration
was ~5 ppm. In this case a bed recovery efficiency
of ~97 percent was observed.
On the same day, the vapor concentrations in the
operator breathing zone were determined. This was
done to establish whether or not safe conditions
could be maintained without the carbon adsorption
-------
system operating. These concentrations were found
to average 15 ppm trichloroethylene (ranged 0-40 ppm
over 16 minutes) without the carbon adsorber vent fan
or the standing operator fan. Safe levels for
trichloroethylene are 100 ppm for an average daily
exposure.
Solvent samples representing five separate desorption
cycles were analyzed for stabilizer content. In
every case, the stabilizers were almost entirely
lost. If this solvent were used by itself or
in larger proportion to the volume of fresh solvent,
the solvent would be unprotected and would be likely
to decompose. Samples of solvent taken from all
portions of the equipment were found to contain
only marginal stabilizer levels.
Steam Requirement
The carbon adsorption system uses steam and water
for desorption, compressed air to activate the valves
and dampers and a 3 horsepower electric motor for the
fan. The steam condensate from two desorption cycles
was collected and measured (120 pounds, 106 pounds) for
an average of 113 pounds per desorption cycle. At
1,000 btu's per pound of steam, 113,000 btu's were
used per desorption. With two desorptions per day
and 240 days per year, the annual steam requirement
-------
is estimated at 54.2 million btu's.
Water Requirement
Water used by the carbon adsorption system is estimated
by the manufacturer at 8 gallons per minute. The water
flows only during the desorption cycles which last
60 minutes each. Thus, the water per year amounts
to 230,000 gallons. The electric 3 horsepower motor
demand can be estimated as follows: 3 horsepower x .746
kilowatt hours per horsepower hour x 8 hours per day
x 240 days per year = 4,300 kilowatt hours per year.
Cost of Operation
Hoyt Manufacturing estimated a replacement purchase price
of $8,000 for an equivalent model to that sold to J. L.
Thompson in 1968. This information along with the utility
consumption data allows the estimation of the cost of
this operation summarized on Tables 1, 2 and 3. The
total operating cost per year was estimated to be
$2,256 (see Table 3). The 50 percent extra floor space
assigned in Table 2 provides for aisleways, restrooms, etc.,
in proportion to the equipment direct space consumption.
No costs have been included to account for numerous other
facilities which are needed but already present. Some of
-------
these would include a portion of the investment capital
for electrical transformer equipment, a steam boiler, an
air compressor, roads and grounds. Also, minor routine
costs have been omitted such as roads, grounds and
building maintenance, heating and lighting operating
costs for the work area and janitorial services.
If the operating cost per year of $2,256 and a solvent
price of $2.37/gallon ($0.195/pound) are taken 952 gallons
of solvent would be needed to recover these costs. Based
on the 2.53 gallons/day recovery rate, only 607 gallons/
year (2.53 gallons x 240 days/year) would actually be
recovered. A net loss of about $817 per year was being
experienced during the evaluation period.
In contrast, customer records for seven months beginning
December 1970 show a recovery average of 205 gallons/month.
This time period was chosen at random and was roughly two
years after initial installation. At this rate of recovery,
an estimated 2,460 gallons would be saved per year. The
value of this solvent (at today's prices) would be $5,830
per year. At this recovery rate a profit of $3,574 per
year was being created by the carbon adsorber. During 1971,
2453 gallons of trichloroethylene were purchased. Thus,
50 percent of the total solvent supplied to the degreaser
was being recycled by the adsorption system.
-------
The ventilation rate through the carbon adsorption system
was determined to be 1,190 feet per minute. The duct
diameter is 12 inches. Thus, the average ventilation rate
was found to be 935 cubic feet per minute (cfm) when both
beds were adsorbing. When one bed was in the desorption
phase of its cycle, the back pressure created by only one
bed reduced the ventilation velocity to about 850 feet per
minute and the volume of air to about 670 cfm . The solvent
recovery rate reported earlier was 2.53 gallons per day or
30.4 pounds per day. This quanity of solvent is recovered
from both beds desorbing once per, day. The average solvent
yield per desorption cycle was 15.2 pounds.
Since this process has two distinctly different phases of
operation, a material balance for each phase is represented
separately. Figure No. 3 describes the adsorption phase
while Figure No. 4 summarizes the desorption phase. In
both cases, the data represented is the average of results
obtained during testing. The carbon adsorption system, like
the degreasing operation, is operated only eight hours per
day. Each bed is desorbed once per day for one hour. Thus,
each bed operates on the adsorption phase of the cycle seven
hours per day. Table No. 4 summarizes the energy exchanges
which take place in the various portions of the adsorption-
desorption cycle. For simplicity, heat loss through the
wall of the carbon adsorption bed is not considered.
Essentially all of the heat energy transferred to a
-------
vapor degreaser or still results in the generation of
solvent vapors. If it were not for the need to replace
heat lost through the walls of the equipment and to
heat condensed solvent back up to the boiling temperature,
the heat input could be measured directly by measuring
the volume of solvent condensate when no work was being
processed through the equipment. Three separate measurements
of the solvent condensate rate in the degreaser were found
to have an average of 95.4 gallons per hour. Three
determinations of the distillation rate from the still
averaged 16.8 gallons per hour. As in the case of the degreaser,
the still experiences a heat loss through the walls of the
equipment but it does not require a heat input to bring the
solvent to boiling temperature. The solvent pumped to the
still is drawn from the boiling sump of the vapor degreaser.
This information permits the following estimates of heat
requirements for the degreaser and still.
Degreaser Heat Requirement
Heat for Solvent Vapor Generation
95.4 Gal./Hr. x 12 Lb./Gal. x 102 Btu*/Lb. = 117,000 Btu/Hr.
Heat Loss Through Walls
173 Ft.2 x 266 Btu**Ft.2 = 46,000 Btu/Hr.
-------
Heat Required for Solvent Condensate
94.4 Gph. from Degreaser
16.8 Gph. from Still
112. 2 Gph. Total
112 Gph. x 12 Lb./Gal. x 0.22***Btu/Lb.-Hr. x
(189-149°F) = 11.800 Btu/Hr
Total = 175,000 Btu/Hr
Still Heat Requirement
Heat for Solvent Vapor Generation
16.8 Gal./Hr. x 12 Lb. x 102 Btu*/Lb. = 20,600 Btu/Hr,
Heat Loss Through Walls
25 Ft.2 x 266 Btu**/Ft.2 = 6,700 Btu/Hr,
Total = 27,300 Btu/Hr.
-------
The combined heat requirement for the metal cleaning
equipment is approximately 202,000 Btu's per hour. By
comparison, the steam required for two desorption cycles
per day (113,000 Btu's each) is 226,000 Btu's per day
or 28,300 Btu's per hour. The energy demand by the 3
horsepower electric motor expressed in Btu's per hour is
7,600 (3 horsepower x 2,545 Btu's per horsepower). The
increase energy demanded by the carbon adsorption system
represents about 18 percent of the energy required for the
metal cleaning equipment alone. If more desorption cycles
were needed, this energy demand would increase rapidly
as a percentile of the direct metal cleaning heat requirement,
An overall flow diagram for solvent, showing the various
process operations, is shown in Figure 5. This information
is simplified on Figure 6 and reduced to percentiles in
Figure 7. Most of this information has been developed
earlier in the report. No method was available during
this testing to measure the amount of solvent loss from
this system during down-time versus that lost during
operating hours. In order to account for this an estimate
of between 15-20 percent was made and an arbitrary 18
percent was chosen. The solvent loss during non-operating
hours was expressed on a per hour basis for the operating
time to be consistant with the system presented. All of
the solvent not accounted for by losses through the still
-------
or estimated as losses taking place during non-operative
time were assumed to be lost to the atmosphere through
the operator loading station. The quantity of solvent
captured by the ventilation system was calculated using
a 95 percent efficiency of the carbon bed to recover
solvent and the known solvent (2.53 gallons per day)
recovered. Again, the quantity of solvent loss in the
exhaust air was calculated based on a 95 percent carbon
bed efficiency. Of the total solvent feed to the
system only 23% is captured and returned by the
carbon adsorption system. As is generally the case,
the overall recovery efficiency is largely controlled
by the low percentage capture of solvent vapors by
the ventilation system to the carbon adsorber beds.
Conclusions
1. The efficiency of emission control by carbon
adsorption of solvents from this metal cleaning
process is largely a function of the effectiveness
of the vent system in capturing the air containing
the solvent.
2. The control of solvent emissions by carbon
adsorption can be profitable to the manufacturer.
-------
3. The solvent recovered (neutrally stabilized
trichloroethylene) is not suitably stabilized
for reuse as is. If blended with sufficient
fresh unused solvent or if restabilized, it
can be reused.
4. The capital investment for carbon adsorption
is sufficiently large that solvent metal
cleaning process must be relatively large
to justify such an expenditure. A recovery
of 952 gallons (at $2.37/gal.) per year would
be needed to support the investment for a
Model 536AD carbon adsorber or equivalent on
a break-even basis.
5. A solvent emission control of 25 percent was
measured based on the quantity of solvent
required by the metal cleaning system.
-------
TABLE I
Industrial Building*
Shell (M&L) Cost
Lighting and Electrical
Heating and Ventilating
Plumbing
Fire Prevention
$ 4
1,
1
1,
09/Ft,
75
50
70
1.10
Sub-Contract Cost (1.3)
Contingency (15%)
$10.14
(1968 Base)
2
$17.3/Ft.'t (8%/Annum
1.71 Multiple in 1975)
$22.5/Ft.~
$25.9/Ft.
*Derived from "Modern Cost-Engineering Techniques by
H. Popper. The 8% inflation rate was estimated by
the author.
-------
TABLE 2
Carbon Adsorber
Building Space
Direct
Indirect (50%)
TOTAL
Value
Cost/Annum
Direct Capital
Price
Shipping and
Installation
Total Capital
Cost/Annum
50 Ft.!
25 Ft.'
75 Ft.
$1,943 (At $25.9/Ft.2 - Table I)
$214 (At 25 Yrs. Life @ 10% Time
Value of Money)
$8,000 (Hoyt Manufacturing)
$1,200 (At 15% of Selling Price)
$9,200
$1,200 (At 15 Yrs. Life @ 10% Time
Value of Money)
-------
TABLE 3
Model 536AD Carbon Adsorber
Operating Cost Per Annum
Capital
Equipment
Building
Insurance (2%)
Equipment
Building
Maintenance (4%)
Utilities
Steam
Electricity
Water
Compressed Air
Labor
Return on Investment
$1,210
214
184
39
368
125 (54.2 x 106 Btu @ $2.30/106 Btu)
107 (4,300 KWH @ $0.025/KWH) ,
9 (230 x 10J Gal. @ $0.04/10J Gal.)
Nil
Nil
0
Total Cost/Annum
$2,256
-------
TABLE 4
Energy Balance on Carbon Adsorber (1 Bed)
Adsorption (Solvent Recovery)
Input Output
Solvent Latent Heat 15.2 Lbs. 1,550 Btu's
Air Flow @ 935 Cfm 1,550 Btu's
Desorption
Steam 113 Lbs. 113,000 Btu's
Heating Tank (~900 Lbs.) 13,860 Btu's
Heating Carbon (~300 Lbs.) 10,500 Btu's
Condenser Water (8 Gph) 88,640 Btu's
Vaporizing Solvent* 1,550 Btu's 1,550 Btu's
Adsorption (Drying and Cooling Bed)
Cooling Tank 13,860 Btu's
Cooling Carbon 10,500 Btu's
Air Flow 24,360 Btu's
*Vaporizing solvent from bed requires heat from steam but
this same heat is given up to the condenser water when
the vapor is condensed to liquid solvent.
-------
Figure 1
Float Control
Steam
Coil.
•Cond. Coils
-Water Sep
Motor
Gear Rod
r™ o^-
Loj
Cond. Coil
Steam Coil
146"
Water Sep
"U—Water Jacket
-------
Figure 2
SOLVENT VAPOR CONCENTRATION IN CARBON
ADSORPTION EXHAUST DUCT
ppm
-------
Figure 3
ADSORPTION CYCLE PHASE (Seven Hours Per Day)
16 Lbs Solvent
29,300 Lbs. Air*
\
15.2 Lbs
Solvent
Carbon Adsorption
Bed
0.8 Lbs
^ Solvent
29,300 Lbs
Air*
*935 Cfm X 60 Min/Hr. X 7 Mrs/Day X 0.0745 Lb/Cu Ft j
-------
Figure 4
DESORPTION CYCLE PHASE (ONE HOUR PER DAY)
113 Lbs Steam
(@ 5 psig)
15.2 Lbs
Solvent
4000 Lbs Water
(8 Gpm)
4000 Lbs
Water
1 Carbon Adsorption Bed
2 Shell And Tube Condenser
3 SolventAVater Separator
113 Lbs
Steam Condensate
-------
Figure 5
SOLVENT MATERIAL BALANCE FOR TOTAL SYSTEM
1
0.02 Gph In
Exhaust Air
Adsorber
C. Adsorber
t
0.32 Gph Recovered
By C.A.
Storage
Tank
1.26Gph_
Fresh ~
Solvent
Still
16.8 Gph
0.34 Gph
Captured
By Vent
Distilled Solvent
17.5 Gph
Dirty Solvent
Degreaser
95.4 Gph Distillation Rate
Within Degreaser
Total Loss
From
Degreaser
0.80
Gph -
Loss
To Atm
0.68 Gph
Total Residue
\
0.52 Gph
Oil
0.28 Gph Loss
During Non-Operational
Time (18% of Total)
0.16 Gph
Solvent In
Residue
-------
Figure 6
TOTAL METAL CLEANING AND RECOVERY SYSTEM SOLVENT BALANCE
IN
OUT
Fresh Solvent 1.06 Gph
0.32 Gph Returned
By C.A.*
1.38 Gph
0.63 Gph Not
Captured By
Vent
0.34 Gph
Captured By
Vent To C.A.*
0.97 Gph Loss
To Atm
0.16 Gph
Loss In Still Residues
Est. 0.25_Gph _
Loss In Down-time
*Carbon Adsorption
-------
Figure 7
TOTAL METAL CLEANING AND RECOVERY SYSTEM SOLVENT BALANCE
IN
Fresh Solvent 77%
23% Returned
ByC.A._»
100%
OUT
24% Captured
By Vent To C.A.
46% Not
Captured
By Vent
70% Loss
To Atmos-
phere
Loss In Still Residues
Est. 18%
Loss In Down-Time
*Carbon Adsorption
-------
TJ
m
Z
D
X
O
-------
APPENDIX - CIO
STUDY TO SUPPORT NEW SOURCE PERFORMANCE
STANDARDS OF SOLVENT METAL CLEANING OPERATIONS
EVALUATION OF CARBON ADSORPTION RECOVERY
Vic Manufacturing Company
Minneapolis, Minnesota
PREPARED BY:
D. W. Richards
The Dow Chemical Company
PREPARED FOR:
Emission Standards and Engineering Division
Office of Air Quality Planning
U.S. Environmental Protection Agency
-------
Summary
A test was conducted at Vic Manufacturing Company to evaluate
the use of a carbon adsorption unit to control solvent emissions
from an open top degreaser. Since recorded additions of new
solvent to the degreaser fluctuated erratically during the
time the adsorber was operated, the new solvent additions
required with the adsorber "on" had to be estimated from
plant purchasing records. New solvent additions with the
adsorber "off" were then measured. A comparison of the new
solvent additions with the adsorber "on" to the new solvent
additions with it "off" show an emission control efficiency
of 65 percent. It was also determined that the adsorber had
significantly more capacity than was required based on the
amount of solvent available for adsorption. In spite of this,
the adsorber was still shown to be recovering 1.08 times the
amount of solvent required to offset its annual operating cost.
The stabilizers in the recovered solvent, however, were almost
entirely depleted. The recovered solvent must be blended with
sufficient quantities of new solvent or restabilized before it
is used again.
Emission testing at this location was conducted by George
W. Scheil of the Midwest Research Institute. This work is
reported in EPA Project Report No. 76-DEG-l.
-------
Objective
The purpose of this test program is to evaluate carbon
adsorption as a means of controlling solvent emissions
from metal cleaning operations. The information needed
for this evaluation includes determining:
1. the efficiency of this device in reducing
solvent emissions to the atmosphere,
2. the cost/benefit relationship of this
emission control system,
3. the energy requirement of the emission
control system,
4. any alternate emissions created by the
emission control system.
This data base is being developed to forecast the
magnitude of emission reductions which can be achieved
nationally and the effect on businesses involved. This
information combined with the results of other testing
will be used to design emission control regulations
which effectively limit air pollution and are practical
for industrial application.
-------
Introduction
Carbon adsorption has been offered to solvent metal cleaning
users since 1958. The efficiency of carbon beds to capture
solvent vapor from low concentrations in air has been estimated
to be greater than 95 percent. The typical efficiency of
carbon adsorption in reducing solvent consumption (controlling
solvent emissions) is commonly reported to be in the range of
50-60 percent. Inspite of this, only a small fraction (1.5
percent) of the industrial solvent metal cleaning users surveyed*,
reported that they were using carbon adsorption.
Vic Manufacturing was selected as an emission control test
site primarily because their open top degreaser supported
by a carbon adsorption unit was reported to be extremely
efficient. Since open top degreasers account for approxi-
mately 85 percent of the degreasers now in use*, the potential
applicability of the data from this test was judged to be
quite high.
The carbon adsorber was installed in 1968.
*See Appendix A
-------
Equipment
The basic metal cleaning process employed is a G. S.
Blakeslee D-95-P (12' x 4 1/2') open top degreaser and
carbon adsorption unit operated with trichloroethylene.
The degreaser i,s steam heated. See Figure 1 below.
The carbon adsorption unit is a model 572 AD. Solvent-
laden air is drawn into the unit through an 18" duct. After
passing through the carbon bed, it is discharged outside
through another 18" duct.
-------
OPEN TOP DEGREASER
(Vic Manufacturing Company)
12'4'
r^,
\ '
> y
— ^.'
o^. Cooling Coils
•" Water Jacket
c
Cf
Steam Inlet
5'
Solvent Cooler
Solvent
Storage Tank
Water Separator
J
Exhaust Duct to
Carbon Absorber
£ _ j^*- Vapor Line
I
-------
Experiment Design
This test was designed to obtain several weeks of data with
the carbon adsorber operating and one to two weeks of data
with it off. The work processed and solvent consumption
were to be recorded daily and compared. The actual test
period ran from May 19, 1975 to August 15, 1975. Due to
problems encountered during the test (which will be reviewed
in the Data Discussion section), solvent consumption data
obtained while the adsorber was operating could not be used.
Also, solvent consumption data could be obtained for only
four days with the adsorber off.
-------
Data Discussion
Several weeks of data were obtained with the carbon adsorber
operating. The recorded additions of new solvent to the
degreaser, however, fluctuated widely. At different times
during this period, average new solvent additions varied from
~0.5 gal./day to ~11 gals./day. Also, there were periods of
several consecutive days when no new solvent was added.
Discussions with operating personnel produced several possible
explanations for these discrepancies, but the accuracy of the
data was still in doubt. Therefore, the emission control
efficiency of the carbon adsorber could not be determined
directly. An approximation can be made, however, by using
plant purchasing records to determine the average solvent
consumption of the degreaser while the carbon adsorber is
operating. This number can then be compared to the measured
degreaser solvent consumption while the adsorber was turned
off. Purchasing records were available from January 1 through
August 22, 1975 and indicated that ~38 gals./week of new
solvent were added to the degreaser while the adsorber was
operating. During the four days the adsorber was off, 1047 Ibs,
(86.5 gals.) of new solvent was added. This is equivalent to:
86.5 gals, x 5/4 = 108 gals./week. On this basis, the adsorber
emission control efficiency is:
108 - 38 x 100% = 65%
108
-------
As mentioned, this number is only an approximation, but it
does indicate that the adsorber is very efficient.
The average bed efficiency also indicates the effectiveness
of the adsorber as an emission control device. This value
is determined by measuring and comparing the time-weighted
average solvent vapor concentrations in the inlet and outlet
ducts of the adsorber. On August 20, 1975, these measurements
were made from 10:00 a.m. to 1:30 p.m. using a Gas Tech
Halide meter. The average solvent concentration in the
inlet duct while the degreaser was idling was 21 ppm and
29 ppm while work was being processed. Since the amount of
idling time was about equal to t!he amount of work time for
this particular day, the time-weighted average solvent concen-
tration in the inlet duct was 25 ppm.
The solvent concentration in the outlet duct of the adsorber
was extremely low at all times and averaged only 1.6 ppm.
Therefore, the average bed efficiency of the carbon adsorber
for this day was:
25 - 1.6 x 100% = 93.6%
25
Obviously, the solvent concentration in the inlet has a
great effect on the efficiency of this unit. If the solvent
-------
concentration increases significantly above the 25 ppm level
that was observed, the efficiency should approach 100%.
It is reasonable to expect this since the adsorber is
larger than is required for this system and is capable of
handling much greater inlet concentrations. This was verified
during the test period since the average amount of solvent
collected during the two daily desorption cycles (of ~1 hr.
each) was only 167.25 Ibs. (13.81 gals.). Vic Manufacturing
estimates that the model 572 AD adsorber is capable of
recovering at least 450 Ibs. (37.16 gals.) of trichloroethylene
during the same time (two desorption cycles). Separate material
balances for the adsorption and desorption phases of the carbon
adsorption unit are represented in Figures 2 and 3 respectively.
The numbers shown in each figure are average values obtained
during testing except for water values which are manufacturer's
estimates. The carbon adsorption unit is operated eight
hours per day. Each bed is desorbed once per day for one
hour. This means that each bed must adsorb for seven hours
per day.
The inlet velocity to the carbon adsorber was also measured
on August 22, 1975 and found to be 3133 ft./min. when both
beds were adsorbing. Since the duct diameter is 18", this
is equivalent to 5545 cfm. The. back pressure created when
one bed was desorbing dropped the velocity to an average
of 2225 ft./min. (3938 cfm).
-------
The utilities required for the carbon adsorption system are
steam, water, compressed air, and electricity. Steam condensate
was collected and weighted for four separate desorption cycles.
The weights were: 669.5 Ibs., 600.0 Ibs., 606.0 Ibs., and
586.0 Ibs. (average = 615 Ibs.). Using 1000 Btu's per pound
of steam, 615,000 Btu's are required per desorption cycle.
Thus, 615,000 Btu's/desorption x 2 desorptions/day x 250 days/yr,
= 308 x 106 Btu's/yr.
The carbon adsorption unit's condenser is estimated to use
21 G.P.M. by Vic Manufacturing. Since the water only flows
during the desorption cycles, the total water requirement
is 21 gals./min. x 60 min./hr. x 2 hr./day x 250 days/yr.
= 630,000 gals. The carbon adsorber fan is rated at 20 horse-
power. This is equivalent to: 20 hp. x .746 Kwh/hp. x 8 hrs./
day x 250 days/yr. = 29,840 Kwh/yr. An energy balance for one
bed of the carbon adsorber (adsorption and desorption cycles)
is summarized in Table 1. Radiation losses through the shell
of the adsorber bed are not included in order to simplify
the overall balance.
The replacement cost for the 572 AD carbon adsorption unit
studied during this test is estimated to be $22,085 by Vic
Manufacturing. This information and Table 2 were used to
develop Tables 3 and 4. The assumptions involved in the
-------
development of these Tables include a zero return on investment;
the assignment of 50 percent extra floor space for aisleways,
etc.; the omission of costs for other facilities which are
required, but already exist; and the omission of other minor
costs such as heating, lighting, janitorial services, etc.
Based on these assumptions, the total operating cost per
year for the model 572 AD carbon adsorber is calculated to
be $6903 (see Table 4). If a price of $2.15/gal.* is used
for trichloroethylene, then 3211 gallons must be recovered
by the carbon adsorber per year to offset the total annual
operating cost. Based on the measured 13.81 gals./day
recovery rate, a total of 3453 gallons will be recovered
yearly, or 1.08 times more than is required to offset the
adsorber's annual operating cost.
Due to the relatively small quantity of work processed by
the degreaser, the 572AD adsorber has quite a bit more
capacity than is required for this system. Substitution
of a smaller 554AD adsorber, for example, would still allow
the same amount of work to be processed and would also increase
the profitability of the system. The total yearly operating
cost for a 554AD (calculated in the same manner as the
572AD) would be ~$4410. Using the same price of $2.15/gal.
for trichloroethylene, only 2051 gallons must be recovered
to offset the annual operating cost. Comparing this to the
-------
3453 gallon yearly recovery developed earlier, it is found
that it would be possible to recover 1.68 times more than
the amount of solvent required to offset the annual operating
cost instead of only 1.08 times more.
The affect of carbon adsorption on solvent stabilizer levels
was somewhat difficult to ascertain at this test site. Several
impurities (such as methyl chloroform, perchloroethylene, and
Fluorocarbon 113) were identified in the adsorber samples.
Most of these impurities are introduced into the degreaser
on the various parts which are cleaned; however, some are also
present in the new solvent that is added to the degreaser.
Analyses showed that the stabilizers were severely depleted
by the carbon adsorber. This means that any solvent recovered
by the adsorber must be restabilized (by blending with adequate
amounts of new solvent or adding a stabilizer concentrate)
before it is used again.
The solvent condensation rate in the degreaser was found to
be 198.2 gals./hr. while no work was being processed. Using
this information and the knowledge that some heat is lost
via radiation through the degreaser walls, the degreaser heat
requirement can be calculated.
Degreaser Heat Requirement
Heat for Solvent Vapor Generation
-------
Figure 6
IN
Fresh Solvent
37.6%
62.4% Returned by C.A.*
100% To Circuit Board Cleaner
OUT
4.1% Lost To Atmosphere
66.5% Captured
By Vent To C.A.*
33.5% Lost on Boards
*Carbon Adsorption
-------
Figure 5
IN
Fresh Solvent
10.43 Gpd
17.28 Gpd Returned By C.A."
27.71 Gpd To Circuit Board Cleaner
"Carbon Adsorption
OUT
1.16 Gpd Lost To Atmosphere
18.44 Gpd Captured
By Vent To C.A.*
9.27 Gpd Lost On Boards
-------
198.2 Gals./Hr. x 12.11 Lbs./Gal. x (189°-70°F) x .22* Btu/Lb./°F
= 62,387 Btu/Hr.
198.2 Gals./Hr. x 12.11 Lbs./Gal. x 101.6** Btu/Lb.
= 243,860 Btu/Hr.
Heat Loss Through Walls
208 Ft.2 x 266*** Btu/Hr./Ft.2 = 55,328 Btu/Hr.
361,575 Btu/Hr.
The energy required for two desorption cycles per day is
615,000 Btu's x 2 = 1.23 x 10 Btu's/Day. The energy required
by the carbon adsorber's 20 horsepower fan (expressed in
equivalent units) is 20 Hp. x 2545 Btu/Hp./Hr. x 8 Hrs./Day =
407,200 Btu's/Day. Therefore, the total energy demand of
the carbon adsorption unit is 1.64 x 10 Btu's/Day. Since
the degreaser represents the total heat input for the metal
cleaning system, the total heat requirement for the system is
361,575 Btu's/Hr. x 8 Hrs./Day = 2.89 x 106 Btu's/Day. Thus,
the energy required for the carbon adsorber represents an
increase of 57 percent of that required for the degreaser alone,
*Specific Heat For Trichloroethylene "^ Obtained From "Modern
**Heat of Vaporization For Trichloroethylene> Vapor Degreasing",
***Radiation Heat Loss For Trichloroethylene J The Dow Chemical Co.
-------
Conclusions
1. Carbon adsorption can provide significant emission
control for an open top degreaser. In this case,
the reduction in emissions was estimated to be "65
percent by comparing solvent usage with the adsorber
turned off to solvent usage with the adsorber operating
(calculated from plant purchasing records).
2. Controlling emissions with carbon adsorption can be
profitable. At the solvent recovery rate measured
during this test, 1.08 times the amount of solvent
required to offset the model 572AD adsorber's annual
operating cost can be recovered. This can be
theoretically increased to 1.68 times by substituting
a model 554AD adsorber for the model 572AD.
3. Trichloroethylene recovered by the adsorber is not
suitable for reuse unless it is either restabilized
with a concentrate, or blended with a sufficient
amount of new solvent.
-------
TABLE 1
Energy Balance on Carbon Adsorber (1 Bed)
Adsorption (Solvent Recovery)
Output
Solvent Latent Heat (83.63 Lbs.)
Air Flow At 5545 CFM
8500 BTU'S
8500 BTU'S
Desorption
Steam (615 Lbs.)
Heating Tank (2800 Lbs.)
Heating Carbon (1500 Lbs.)
Condenser Water (21 GPH)
Vaporizing Solvent*
615,000 BTU'S
43,740 BTU'S
42,600 BTU'S
528,660 BTU'S
8500 BTU'S 8500 BTU'S
Adsorption (Drying and Cooling Bed)
Cooling Tank
Cooling Carbon
Air Flow At 5545 CFM
43,740 BTU'S
42,600 BTU'S
86,340 BTU'S
*Vaporizing solvent from the bed requires heat from steam, but
this same heat is given up to the condenser water when the
vapor is condensed to liquid solvent.
-------
TABLE 2
Industrial Building*
Shell (M&L) Cost
Lighting and Electrical
Heating and Ventilating,
Plumbing
Fire Prevention
$ 4.09/Ft
1.75
,50
,70
1,
1
1.10
Sub-Contract Cost (1.3)
Contingency (15%)
$10.14/Ft. (1968 Base)
$17.3/Ft.2 (8%/Annum
1.71 Multiple in 1975)
$22.5/Ft.i?
$25.9/Ft.
*Derived from "Modern Cost-Engineering Techniques by
H. Popper (pg. 103) . The 8% inflation rate was estimated
by the author.
-------
TABLE 3
Carbon Adsorber
Building Space
Direct
Indirect (50%)
TOTAL
Value
Cost/Annum
Direct Capital
Price
Shipping and
Installation
TOTAL CAPITAL
Cost/Annum
111 Ft.!
55.5 Ft.'
166.5 Ft.
$4312
$ 475
(At $25.9/Ft. - Table 2)
(At 25 Years Depreciation Rate)*
$22,085 (Vic Manufacturing)
$ 3,313 (At 15% of Selling Price)
$25,398
$ 3,339 (At 15 Years Depreciation Rate)*
*10% Interest Rate On Investment
-------
TABLE 4
Model 572 AD Carbon Adsorber
Operating Cost Per Annum
Capital
Equipment $3339
Building $ 475
Insurance (2%)
Equipment $ 508
Building $ 86
Maintenance (4%) $1016
Utilities
Steam $ 708 (308 M BTU's At $2.30/M BTU's)
Electricity $ 746 (29,840 KWH At $0.025/KWH
Water $ 25 (630 M Gals. At $0.04/M Gals.)
Compressed Air Nil
Labor Nil
Return on Investment 0
Total Cost/Annum $6903
-------
Figure 2
ADSORPTION CYCLE PHASE
(7 HOURS PER DAY)
89.3 Ibs Solvent
173,500 Ibs Air*
1
83.6 Ibs
Solvent
5.7 Ibs Solvent
173,500 Ibs Air*
'5545 CPM x 60 min/hr x 7 hrs/day x 0.0745 Ibs/cu ft
-------
Figure 3
DESORPTION CYCLE PHASE
(ONE HOUR PER DAY)
L'L
1
\^
II
t
615lbsSteam
(at~10psig)
1 o__u«
>
^
1(
w
-*-\ 10,500* Ibs Water
•*
(21 G.P.M.)
2
T
3
IT
1 I
83.6 Ibs 615lbsSteam
Solvent Condensate
r
),500* Ibs
ater
Carbon Adsorption Bed
2. Shell And Tube Condenser
3. Solvent/Water Separator
GPM x 60 min/hr x 8.33 Ibs/gal x 1 hr/day
-------
cc: K. S. Surprenant, 2020
CRC/PF 2192008
Central Files
December 15, 1975
Mr. J. W. Barber
Research Director
Vic Manufacturing Company
1620 Central Ave. N.E.
Minneapolis, Minnesota 55413
REPORT PREPARED FOR ENVIRONMENTAL PROTECTION AGENCY COVERING
VIC CARBON ADSORBER STUDY
Dear Joe:
This is to confirm our telephone conversation of this morning
during which we discussed my report and your December 31, 1975
letter commenting on the report. As I mentioned, I feel that
many of the comments you made concerning Vic's costs, etc.
are quite valid. By using your actual cost experience, I
would have arrived at an annual operating cost for the Model
572AD adsorber of $5052 (as you stated) instead of the $6903
that appears in ray report. The problem that this creates,
however, is one of comparing "apples to oranges". The
assumptions that were used to develop the adsorber operating
cost for the Vic report were also used for all of our other
studies. This was done deliberate so that the EPA can compare
the operating costs of the various emission control devices
tested on an equal basis. It is important to note that
the cost information we have developed applies to the
use of adsorbers and other control devices in new facilities.
Using the actual costs for floor space, maintenance, insurance,
etc. for an existing facility, therefore, would not be
consistent with this approach.
Because of this methodology, therefore, I am going to
submit the Vic report to the EPA without changing the
development of the adsorber annual operating cost. I
-------
Mr. Barber 2 December 12, 1975
also submit a copy of your December 3, 1975 letter
for their inspection.
Thanks again for all of your assistance during this study.
Best regards,
David W. Richards
Inorganic Chemicals Department
Chlorinated Solvents Section
Phone: (517) '636-6640
cc: Mr. John C. Bellinger
Industrial Studies Branch
U. S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
kw
-------
VIC MANUFACTURING COMPANY
1620 CENTRAL AVENUE N.E., MINNEAPOLIS, MINNESOTA 55413/PHONE (612) 781-6601
A»R POLLUTION CONTROL SYSTEMS
December 3, 1975
Dave Richards
Inorganic Chemicals Department
2020 Dow Center
Midland, Michigan 48640
RE: EPA Report
Dear Dave:
We have received your draft copy and find it well-organized and
quite complete. We do, however, have some comments regarding
the treatment of various costs.
1. We believe that the area costs assigned to the equipment to
be excessive. The costs detailed in the report refer to
new building costs. Since the report is concerned with the
VIC Manufacturing Company unit, we believe that the cost
should be related to the present market value of $6.00/S.F.
We suggest that in new buildings the cost/S.F. may dictate
that an adsorption system be mounted on an elevated platform
or on the roof and thus reduce or eliminate the S.F. charges
against the equipment.
While we support the application of all reasonable costs to
provide a true costing picture, we do not agree that new
building costs should be so applied.
2. If we apply $6.00/S.F. to the VIC installation and use the
166.5 S.F. space allocation, then the value would be $999.00
and cost/annum would be $110.00. Then in Table 4 the
following changes are suggested to reflect the actual costs
based on our rates and charges.
AWARDED THE PRESIDENT'S "E" FOR EXCELLENCE IN EXPORT
-------
Inorganic Chemicals Department - 2 - December 3, 1975
Dave Richards
Capitol: Equipment 3340
Building UO
Insurance (our FIA Premium is 0.7% of dollar value)
Equipment . 178
Building 7
Maintenance 200 ^ ^ -, . .
ffrS/ti ''*• <*•
Utilities (our gas costs 1.16/Million BTU's)
Using your 308M" input to the adsorber we would
need 385FT at the boiler (w/80% efficiency) and
that would cost 446
Electricity 746
Water 25_
5052/Annum
3. 'If we apply this annual cost to the valve of the solvent
reclaimed, their 2350 gallons would need to be recovered
and the measured 3453 gallons reclaimed become 1.47 times
the breakeven quantity.
Applying these same values to the 554 assumptions would
probably raise the factor to 2X or more.
4. We also note that there are available two tax credits that
may be applied to Air Pollution Control equipment. One is
the Investment Tax Credit of 10% on new equipment applied to
the Federal Income Tax, and the second is the Air Pollution
Control Equipment Tax Credit that is applicable to many
State Income Tax assessments.
We have appreciated this opportunity to review your excellent
report and hope that you will find our comments helpful.
;ards,
Research Director
JBimek
cc/IVictor
CGorman
-------
-o
•o
m
O
n
-------
APPENDIX - Cll
STUDY TO SUPPORT NEW SOURCE PERFORMANCE
STANDARDS OF SOLVENT METAL CLEANING OPERATIONS
EVALUATION OF CARBON ADSORPTION RECOVERY
Western Electric Company
Hawthorne Station
Chicago, Illinois
PREPARED BY:
D. w. Richards
The Dow Chemical Company
PREPARED FOR:
Emission Standards and Engineering Division
Office of Air Quality Planning
U. S. Environmental Protection Agency
-------
Summary
The use of a carbon adsorption unit to control solvent emissions
from an enclosed "cold cleaning" system (circuit board cleaner)
was studied at Western Electric Company. It was determined that
the carbon adsorber recovered 60% of the new solvent added to
the system and that the solvent recovery rate was 2.5 times
that required to offset the annual adsorber operating costs.
The stabilizers in the recovered solvent, however, were
totally depleted. Before the recovered solvent can be
reused it must be mixed with adequate quantities of new
solvent or restabilized.
-------
Objective
The purpose of this test program is to evaluate carbon
adsorption as a means of controlling solvent emissions
from metal cleaning operations. The information needed
for this evaluation includes determining:
1. the efficiency of this device in reducing
solvent emissions to the atmosphere,
2. the cost/benefit relationship of this
emission control system,
3. the energy requirement of the emission
control system,
4. any alternate emissions created by the
emission control system.
This data base is being developed to forecast the
magnitude of emission reductions which can be achieved
nationally and the effect on businesses involved. This
information combined with the results of other testing
will be used to design emission control regulations
which effectively limit air pollution and are practical
for industrial application.
-------
Introduction
Carbon adsorption has been offered to solvent metal
cleaning users since 1958. The efficiency of carbon
beds to capture solvent vapor, from low concentrations
in air has been estimated to be greater than 95 percent.
The typical efficiency of carbon adsorption in reducing
solvent consumption (controlling solvent emissions) is
commonly reported to be in the range of 50-60 percent.
Inspite of this, only a small fraction (1.5 percent)
of the industrial solvent metal cleaning users surveyed*,
reported that they were using carbon adsorption.
Western Electric was considered as an emission control
test site on the recommendation of Baron-Blakeslee (equip-
ment manufacturer). Western Electric has extensive experience
with solvent metal cleaning and their personnel have excellent
reputations in this area. Also, the carbon adsorption system
was reported to reduce solvent consumption from 76 pounds
per hour to 1.5 pounds per hour (98% recovery). This operation
was selected for evaluation because it was felt that: it could
demonstrate the highest level of efficiency possible through
the use of carbon adsorption in support of an enclosed cold
cleaning system.
Just prior to the first week of the study it was noticed that
the steam-solvent vapor valve on bed "A" was cracked. Fine
*See Appendix A
-------
efforts by both Vic Manufacturing and Western Electric resulted
in the replacement of this valve before the first week of
the study was completed.
The carbon adsorption unit was installed in July, 1973. Since
these units are reported to have operating lives of 15 years
or more, this unit must be considered to be relatively new.
-------
Equipment
The metal cleaning process employed consists of two circuit
board flux cleaners, a still, and a carbon adsorption unit
operated with trichloroethylene. One of the flux cleaners
is oval in shape while the other one is a straight line
(in-line) unit. Both are room temperature cleaners and both
are exhausted to the carbon adsorption unit. During the
test period, however, approximately 75-90% of the work was
processed through the in-line cleaner. Therefore, the oval
cleaner is not shown below in Figure 1 (which is a schematic
of the entire system).
The carbon adsorption unit is a Model 536AD. Solvent-laden
air from both flux cleaners is drawn into the carbon adsorber
through a 12 inch duct. After passing through the carbon bed,
the essentially solvent-free air is then discharged outside
through a 10 inch duct.
The clean solvent storage tank receives solvent from three
different sources: 1) reclaimed solvent from the still, 2)
recovered solvent from the carbon adsorber, and 3) new
solvent pumped from 55 gallon drums. All of the new solvent
is pumped through a meter.
The dirty solvent storage tank receives contaminated solvent
from both flux cleaners. When enough dirty solvent is collected,
it is then transferred automatically to the still to be
reclaimed.
-------
Fresh Solvent
To Clean Solvent
Chamber
Solvent
Meter
Reclaimed Solvent
Clean
Solvent
Storage
Still
Figure 1
CIRCUIT BOARD CLEANING SYSTEM
Western Electric Company
1
1
t
1
1
1
Dirty
Solvent
Storaae
i
t
t
k ^-
Recovered Solvent
Oval Cleaner
1
Carbon
Adsorber
\
Solvent Section Exhaust
r
t
». f
>
In-Line Solvent
Section Exhaust
In-Line Cleaner
Boards Enter
Belt
fc
' — A
I
Preheater
Flux
Solder
k
Solvent Cleaning Section
(Brushes)
OOOiOOOOOO
«-i
i ».
i
Solvent— Fr
Air To Atnr
phere
^ Boards Exit
^^••4
J Dirty
( Solvent Chambers)
Clean
-------
Experiment Design
The original plan for testing was designed to obtain two
weeks of data with the carbon adsorption system operating
and one to two weeks with it off. The test was scheduled
to begin on May 22, 1975, but due to unforeseeable circum-
stances did not begin until June 16, 1975. As a result
of the early delays, the testing ran into Western Electric"s
scheduled shutdown (July 13 - July 26) and had to be extended
into August. The total workload and solvent consumption
for both cleaners were to be recorded weekly and compared.
Since the cleaners' solvent chambers are kept at a constant
level by a float control system, the difference in solvent
inventory could-be measured by manually filling the still
and measuring the solvent levels in the clean and dirty
solvent storage tanks. These quantities could then be added
to or subtracted from the total new solvent added to the
system (which is measured by the solvent meter). Small
quantities of solvent are withdrawn from the clean storage
tank by the operator for miscellaneous cleaning operations.
These withdrawals were recorded during the entire period
of the test and were subtracted from solvent consumption
figures.
-------
Data Discussion
Two weeks of data were obtained with the carbon adsorption
unit "on" and two weeks with it "off". An additional two weeks
of data had to be discarded. During one of these weeks
solvent was lost through a crack in the housing of the
steam-vapor valve on bed "A" of the carbon adsorber. During
the other week a leak developed in the carbon adsorber
condenser coil and the system was contaminated with water.
Solvent emission rates were determined by differences in
solvent inventory within the clean and dirty storage tanks
as discussed previously. The clean and dirty storage tanks
are each 2' high x 4' wide x 5' long. Therefore, an inch
of solvent in either tank is equivalent to 12.5 gallons.
Each tank is equipped with a sight glass, .and the solvent
depth can be accurately measured to within 1/8 inch. Thus,
the error in measuring solvent in either tank is + 1.5 gallons.
The average solvent consumption for the two weeks the carbon
adsorber was operating was 10.4 gal./day. The average work
2
processed during this two week period was 417.7 ft./day.
The daily solvent consumption per square foot of work processed
is therefore, 10.4/417.7 or 0.025 gal./ft . Similarly, the
daily solvent consumption per square foot of work processed
-------
for the two weeks the carbon adsorber was off is 23.8/375.8
or 0.063 gal./ft . On this basis, the emission control
efficiency is determined to be:
°-063 - °-025 x 100% = 60.3%
0.063
The solvent removed from the system as still residue was
insignificant during the test period and was not considered
in this calculation.
The solvent vapor concentrations in the outlet duct of the
carbon adsorber were measured from 10:00 A.M. to 1:00 P.M. on
August 21, 1975 and are plotted below in Figure 2. The
average vapor concentration in the outlet duct was 10.4 ppm.
-------
Figure 2
SOLVENT VAPOR CONCENTRATION IN CARBON
ADSORPTION OUTLET DUCT (August 21, 1975)
20
18
16
14
12
E 10
a
a 8
6
4
2
0
I Desorb I
I Bed I
I "R" I
1 Desorb
1 Bed
"A"
10 10:30 11 11730
am Time
12 12:30
pm
Average = 10.4 ppm
As shown in Figure 2, the portion of the curve between 10:10-
11:10 corresponds to the desorption of bed "B" and the portion
of the curve between 11:45-12:45 corresponds to the desorption
of bed "A". At the beginning of both desorption cycles the
vapor concentration decreases rapidly indicating that the bed
being desorbed was saturated and losing solvent to the atmos-
phere. As the desorption cycles continue, the vapor concen-
tration gradually increases since only one bed is adsorbing.
At the end of the desorption cycle, the concentration decreases
to an equilibrium level since both beds are adsorbing again.
-------
Vapor concentration measurements were also made in the
inlet duct to the carbon adsorber. The readings ranged
from 132 to 177 ppm with an average of 165 ppm. On this
particular day, only the in-line circuit board cleaner
was being operated. This was typical throughout the
test period. No significant differences in vapor concen-
trations were bbserved when work was being processed
compared to the time the machine was idling. This was
anticipated since the brushes in the solvent cleaning section
continue to rotate whether or not work is being processed,
and the entire section is covered by metal "pans" to reduce
splashing and solvent loss to the exhaust duct. The only
additional loss of solvent that occurs while work is being
processed is that which is retained on the circuit boards
and carried outside of the machine. The average inlet
vapor concentration (165 ppm), therefore, was simply
determined by averaging the readings and was not time-
weighted for hours of working operation and idling. The
average bed efficiency for this particular day was:
165 - 10.4 x 100% = 93.7%
165
Samples of solvent were obtained during three different
desorption cycles and analyzed for stabilizer content.
-------
Although the depletion of stabilizers in a room temperature
cleaning operation is not as critical as it is in a vapor
degreasing operation, carbon adsorbers are used in conjunc-
tion with both types of systems. Therefore, the affect of
carbon adsorption on solvent stabilizer levels was studied
during this test. No stabilizers were found in any of the
samples. This indicates that solvent recovered from the
carbon adsorber is not suitable for reuse unless it is
mixed with a greater quantity of fresh solvent, or
replenished with a stabilizer concentrate.
-------
The utilities required for the carbon adsorption system are
steam, water, compressed air, and electricity. Steam
condensate was collected and weighed for three separate
one hour desorption cycles. The weights were: 395 Ibs.,
406 Ibs., and 402 Ibs. (average = 401 Ibs.). While the
weights were consistent, they were extremely high for this
carbon adsorber (manufacturer's estimate for the unit is
180 lbs./hr.). Near the end of the test period a leak
was detected in the condenser coil of the carbon adsorber.
Since there is no way to determine how much additional
water this added to the steam condensate being measured,
the manufacturer's estimate of 180 lbs./hr. will be used
for calculations. Using 1000 Btu's per pound of steam,
180,000 Btu's are required per desorption cycle. Thus,
180,000 Btu's per desorption x 12 desorptions per day
x 240 days per year = 518 x 10 Btu's/year.
The carbon adsorption unit's condenser is estimated to
use eight gallons per minute by the manufacturer. Since
the water only flows during the desorption cycles, the
total water requirement per year is 1,380,000 gallons.
The carbon adsorber fan motor is rated at 3 horsepower.
This is equivalent to: 3 hp. x .746 KWH/hp. hour
x 24 hrs./day x 240 days/yr. = 12,890 KWH/yr.
The replacement cost for the 536AD carbon adsorber unit
studied during this test was estimated to be $9320 by
Vic Manufacturing.
-------
This information and Table 2 were used to develop Tables
3 and 4. The assumption involved in the development of
these Tables include a zero return on investment; the
assignment of 50 percent extra floor space for aisleways,
etc.; the omission of costs for other facilities which
are required, but already exist; and the omission of
other minor costs such as heating, lighting, janitorial
services, etc. Based on these assumptions, the total
operating cost per year for the model 536AD carbon adsorber
is calculated to be $3874 (see Table 4). If a price of
$2.15/gal. ($0.1775/lb.)* is used for trichloroethylene,
then 1802 gallons must be recovered by the carbon adsorber
per year to offset the total annual operating cost. Based
on the measured 17.3 gals./day recovery rate, a total of
4152 gallons will be recovered yearly. Thus, a net savings
of about $5050/yr. will be realized at this recovery rate.
*Chemical Marketing Reporter, July 7, 1975, Schnell Publishing
Company
-------
The inlet velocity through the carbon adsorption unit was
found to be 1458 ft./min. when both beds were adsorbing.
Since the inlet duct diameter is 12 inches, this is
equivalent to an air volume of 1145 cfm. When one bed
was desorbing, the back pressure reduced the velocity to
an average of 804 ft./min. and the air volume to 631 cfm.
The average solvent recovery rate was determined to be
209 Ibs./day or 17.28 gals./day. This data was obtained
by collecting solvent from all twelve of the desorption
cycles during a 24 hour period. Therefore, the average
solvent collected per desorption cycle was 17.4 Ibs.
(1.4 gals.) .
Separate material balances for the adsorption and desorption
phases of the carbon adsorption unit are represented in
Figures 3 and 4 respectively. The numbers shown in each
figure are average values obtained during testing except
for steam and water values which are manufacturer's
estimates. The carbon adsorption unit is operated 24
hours per day. Each bed is desorbed six times per day
for one hour each time. This means that each bed must
adsorb for 18 hours per day. An energy balance for one
bed of the carbon adsorber (adsorption and desorption
cycles) is summarized in Table No. 1. Radiation losses
through the shell of the adsorber bed are not included
in order to simplify the overall balance.
-------
The still distillation rate was checked several times
during the test period and averaged 95.6 gallons per hour.
Dirty solvent from the circuit board cleaner is collected
in the dirty solvent storage tank and is automatically
pumped to the still when enough is collected for distillation.
Therefore, heat must be added to bring the dirty solvent from
room temperature, to the boiling point as well as additional
heat to maintain the distillation. Also, some of the heat
can be lost by radiation through the still walls. Using
this information, the still heat requirements can be
calculated.
Still Heat Requirement
Heat for Solvent Vapor Generation
95.6 Gal./Hr. x 12.11 Lbs./Gal. x (189° - 70°F) x .22* Btu/Lb./°F
= 30,300 Btu/Hr.
95.6 Gal./Hr. x 12.11 Lbs./Gal. x 101.6** Btu/Lb.
= 117,600 Btu/Hr.
Heat Loss Through Walls
62 Ft.2 x 266*** Btu/Hr./Ft.2 = 16,500 Btu/Hr.
164,400 Btu/Hr.
*Specific Heat For Trichloroethylene ~"| Obtained From "Modern
**Heat of Vaporization For Trichloroethylene> Vapor Degreasing",
***Radiation Heat Loss For Trichloroethylene The Dow Chemical Co.
-------
The energy required for 12 desorption cycles per day is
180,000 Btu's x 12 = 2.16 x 106 Btu's/Day. The total energy
required by the carbon adsorber 3 horsepower fan (expressed
in equivalent units) is 3 hp. x 2545 Btu/hp./Hr. x 24 Hrs./Day
= 183,240 Btu's/Day. Thus, the total energy demand of the
carbon adsorption unit is 2.3 x 10 Btu's/Day. Since the
still represents the total heat input for the metal cleaning
system and only operates "2 Hrs./Day, the total heat requirement
for the cleaning system is 164,400 Btu's/Hr. x 2 Hrs./Day =
328,800 Btu's/Day. Therefore, the energy required for the
carbon adsorber represents an increase of .7 times that required
for the metal cleaning operation alone. This is not nearly as
significant as it appears at first glance, since (as discussed
earlier) the carbon adsorber recovers more than enough solvent
to offset its annual operating cost.
A simplified overall flow diagram which traces the flow
of solvent through the cleaning system is presented in
Figure 5. The same information is also presented in
Figure 6 in percentile form. As discussed earlier, solvent
losses in the still bottoms are omitted. During the test
period it was estimated that less than 0.2 Gpd of solvent
added to the system was lost in the still bottoms. Therefore,
all of the losses of liquid solvent were assumed to occur
as carryout on the circuit boards. The losses of solvent
vapor to the atmosphere were determined using a 93.7% carbon
adsorber bed efficiency and the measured amount of solvent
recovered (17.28 Gpd).
-------
Conclusions
1. For an enclosed cold cleaning system, carbon adsorption
can provide significant emission control. In this case
a 60% reduction in emissions was measured based on the
amount of new solvent added to the system.
2. Controlling emissions with carbon adsorption can be
profitable (at the solvent recovery rate measured during
this test, 2.5 times the amount of solvent required
to offset the adsorber yearly operating cost can be
recovered).
3. Trichloroethylene recovered by the the adsorber is not
suitable for reuse unless it is: a) blended with an
adequate amount of new solvent or b) restabilized.
-------
TABLE 1
Energy Balance on Carbon Adsorber (1 Bed)
Adsorption (Solvent Recovery)
Solvent Latent Heat (17.4 Lbs.)
Air Flow At 1145 cfm
Desorption
Steam (180 Lbs.)
Heating Tank (~900 Lbs.)
Heating Carbon (~350 Lbs.)
Condenser Water (8 gpm)
Vaporizing Solvent*
Adsorption (Drying and Cooling Bed)
Cooling Tank
Cooling Carbon
Air Flow At 1145 cfm
Input
1770 Btu's
180,000 Btu's
1770 Btu's
14,050 Btu's
9940 Btu's
Output
1770 Btu's
14,050 Btu's
9940 Btu's
156,010 Btu's
1770 Btu's
23,990 Btu's
*Vaporizing solvent from the bed requires heat from steam, but
this same heat is given up to the condenser water when the
vapor is condensed to liquid solvent.
-------
TABLE 2
Industrial Building*
Shell (M&L) Cost
Lighting and Electrical
Heating and Ventilating
Plumbing
Fire Prevention
Sub-Contract Cost (1.3)
Contingency (15%)
$ 4
1
1
1
1
.09/Ft.
.75
.50
.70
.10
$10.14/Ft." (1968 Base)
$17.3/Ft.2 (8%/Annum
1.71 Multiple in 1975)
$22.5/Ft*
$25.9/Ft/
*Derived from "Modern Cost-Engineering Techniques by
H. Popper (pg. 103)• The 8% inflation rate was estimated
by the author.
-------
TABLE 3
Carbon Adsorber
Building Space
Direct
Indirect (50%)
Total
Value
Cost/Annum
Direct Capital
Price
Shipping and
Installation
Total Capital
Cost/Annum
50 Ft.:
25 Ft.'
75 Ft.
$1940 (At $25.9/Ft.2 - Table I)
$ 214 (At 25 Years Depreciation Rate)*
$9320 (Vic Manufacturing)
$1400 (At 15% of Selling Price)
$10,720
$1410 (At 15 Yrs. Depreciation Rate)*
*10% Interest Rate on Investment
-------
TABLE 4
MODEL 536AD CARBON ADSORBER
OPERATING COST PER ANNUM
Capital
Equipment
Building
Insurance (2%)
Equipment
Building
Maintenance (4%)
Utilities
Steam
Electricity
Water
Compressed Air
Labor
Return on Investment
$1410
$ 214
$ 214
$ 39
$ 429
$1^.91 (518 M Btu's At $2.30/M Btu)
$ 322 (12,890 KWH At $0.025/KWH
$ 55 (1380 M Gal. At $0.04/M Gal.)
Nil
Nil
0
Total Cost/Annum
$3874
-------
Figure 3
ADSORPTION CYCLE PHASE (18 Hours Per Day)
111.5 Lbs Solvent
92,127 Lbs Air*
104.5 Lbs
Solvent
7.0 Lbs Solvent
92,127 Lbs Air*
*1145 CFM X 60 min/hr X 18 hrs/day X 0.0745 Ibs/cubic foot
-------
Figure 4
DESORPTION CYCLE PHASE (6 Hrs. Per Day)
s
1
V.
1 1
t
1080 Lbs Steam
(at 10 psig)
1 Parh
nn A (He
J
lrr»tirin R
\
2
orl
r*-i ^ 23,990» Lbs Water
4
2
3
1 1
104.5 Lbs 1080 Lbs Steam
Solvent Condensate
r
3,990* Lbs
Water
2 Shell And Tube Condenser
3 Solvent/Water Separator
*8 GPM X 60 min/hr X 8.33 Ibs/gal X 6 hrs/day
-------
m
Z
o
X
n
Jo
-------
APPENDIX - C12
STUDY TO SUPPORT NEW SOURCE PERFORMANCE
STANDARDS OF SOLVENT METAL CLEANING OPERATIONS
STUDY OF THE EMISSION CONTROL EFFECTIVENESS OF
INCREASED FREEBOARD ON OPEN TOP VAPOR DEGREASERS
PREPARED BY:
K. S. Surprenant
The Dow Chemical Company
PREPARED FOR:
Emission Standards and Engineering Division
Office of Air Quality Planning
U. S. Environmental Protection Agency
-------
Summary
A series of laboratory tests were made to determine 1) the
emission control effect of increasing freeboard on open top
degreasers, and 2) the relative solvent loss rates with and
without lip exhaust ventilation. In quiet air, increasing
the freeboard to width ratio from 0.50 to 0.75 causes a 46%
reduced emission rate. A 55% emission reduction (in turbulent
air) was achieved by increasing the freeboard to equal to
the width of the degreaser. Solvent losses to the environment
are likely to be doubled by the addition of lip exhaust
ventilation.
-------
Objective
This study was conducted to determine how effectively
solvent emissions from vapor degreasing operations could
be controlled by increased freeboard height. Portions of
this testing were also designed to determine the effect
of lip exhaust ventilation on the solvent emission rate,
the idling degreaser loss rate while heated and the shut
down degreaser loss rate.
The aim of this work is to provide basic background data
to be used in conjunction with emission control evaluations
of actual metal cleaning operations. This background
information will be used to develop New Source Performance
Standards which effectively limit air pollution and are
practical for industrial application.
-------
Introduction
Industrial vapor degreasing operations do not lend them-
selves to the determination of some basic information.
Specifically, working vapor degreasers can not be modified
conveniently to examine solvent loss rates vs. freeboard
height without work interruption. Again, it is of value
to know the solvent emission rates of vapor degreasers
during shut down time and during idling (heated but not
cleaning metal parts). In each of these cases, the work
capacity of the degreaser would have to be lost for
periods of time to determine this data. According to the
survey reported in Appendix A, 85% of the vapor degreasers
are open top degreasers. Both experience and in-plant
emission control testing has shown that open top degreasers
are used for actual metal cleaning only a relatively small
percent (25%) of the time. Typically, open top degreasers
are left heated and open during the balance of the work
shifts. This idling operation represents a significant
portion of the total solvent emissions. Yet, it is
difficult to define this emission rate at in-plant
locations. For this reason, the evaluations made in
this report were conducted under laboratory conditions.
Compliance with the Occupational Safety and Health Act
requirements and concern for employee safety are causing
-------
manufacturing concerns to consider ventilation of vapor
degreasing operations. Although ventilation systems have
been shown to provide limited reductions in exposures of
workers to solvent vapors/ ventilation has also been
known to cause higher solvent emission rates. This also
was studied on a laboratory basis to avoid variations
caused by workloads.
-------
Equipment
The equipment used for this testing was a Model 2D500 E
size D20. This equipment was supplied as a standard model
of the Detrex Chemical Industries Co. and was operated with
1,1,1-trichloroethane throughout all the tests described.
A copy of the Detrex specification sheet for this degreaser
is attached. The test degreaser was heated electrically
with 15 kilowatt heaters. This degreaser was located in
a room with exceptionally quiet air having dimensions of
approximately 20' x 35'.
-------
Experiment Design
The valve from the immersion chamber of the degreaser to
the boiling chamber was left open throughout testing so
that this chamber was maintained in an empty state. The
water separator and solvent spray reservoir were maintained
at an overflow level. Thus, all solvent losses from the
equipment were reflected in a change in the level of the
boiling chamber. The solvent level in the degreaser was
adjusted to a specific elevation before and after each
test with the heat turned "off". The dimensions of the
boiling chamber are roughly 24 1/4" x 22". Every inch
of elevation is equal to approximately 2.31 gallons. Since
the level could be measured within + 1/8 inch, the error in
estimating solvent consumption was controlled to within
+_ 0.298 gallons or 1.09 liters. The solvent loss was
determined in liters of solvent necessary to return the
solvent level to the starting level.
-------
Data Discussion
Solvent Emission Rate Vs. Freeboard Height (Quiet Air)
The roll top cover provided with this degreaser was removed
for this testing, exposing the full open top area of the
degreaser (2' x 4'). A sleeve was constructed to fit
just inside the walls of the degreaser and designed so
that the existing freeboard height of 12 inches could
be increased by 6 inches or by another 12 inches. The
original freeboard height to degreaser width ratio was
0.50, that is, 12 inches of freeboard over 24 inches of
degreaser width. The increased freeboard heights increased
this ratio to 0.75 and 1.0. The degreaser was operated for
one week each at the 12 inch freeboard and 24 inch freeboard
and two weeks at the 18 inch freeboard height. During each
seven day test, the degreaser was held at a boil continuously
and no work was processed. The results of this testing are
summarized in Table 1.
TABLE 1
Emission Rate Versus Freeboard Height (Quiet Air)
2
Freeboard Solvent Used Lbs./Ft. -Hr. Emission Control
12" 44 L./7 Days 0.095
18" f 23 L./7 Days) ... ...
[ 24 L./7 Days ) 0'051 46% ± 3%
24" 25 L./7 Days 0.054 43% + 3%
-------
Due to the workloads processed in industrial degreasers
and the lack of totally quiet air, emission rates from
operating vapor degreasers can not be expected to achieve
these levels. However, most of the emissions which occur
while degreasers are idling can be prevented through the
use of a cover. Thus, a degreasing operation with an
automatic cover could be expected to conserve almost
one-tenth of a pound per square foot-hour of operation
if it has a normal 0.50 freeboard design. Increasing
the freeboard height to a ratio of 0.75 reduces the
solvent emissions in a quiet area by approximately 46%
without a cover. Again, in a quiet area, no further decrease
in solvent emission rate occurs by increasing the freeboard
height to equal to the width of the degreaser.
Solvent Emission Rate Vs. Freeboard Height (With Air Movement)
To determine solvent emission rates on a more realistic
basis, a fan was located 16 feet from the degreaser and
turned on low. The air movement created by this fan was
barely noticeable when standing at the end of the degreaser
closest to the fan. Velometer measurements of the air
velocity near the lip of the vapor degreaser were found
to fluctuate from a low of about 30 feet per minute to
slightly less than 100 feet per minute. Most of the
-------
air velocity measurements were found in the range of
50 to 100 feet per minute or less than one mile per hour.
With the normal freeboard on the equipment, the vapor
zone was not seriously disturbed but wave motions having
an amplitude of one to two inches were observed. When
the six inch freeboard addition was added to provide 18
inches of freeboard the wave motion of the vapors was
almost eliminated. The vapor zone was essentially quiet
with the 24 inch freeboard.
Due to the larger volumes of solvent being consumed measureable
losses would occur in 24 hours and data was taken on that basis.
This data is summarized in Table 2. It should be noted that
the solvent emission rate varied considerably with the
original 12 inch freeboard as would be expected of a
system not being entirely controlled. However, the
solvent loss rate experienced under these conditions
(0.373 pounds per square foot hour) is not as great
as the commonly used industry value of 0.50 pound per
square foot hour for an operating industrial degreaser.
The emission rate was reduced by approximately 27% with
an 18 inch or 0.75 freeboard ratio. The consistency of
the results also demonstrate the better control experienced
by the system. The 24 inch or 1.0 ratio freeboard reduced
solvent emissions by 55%. The last two results reported
-------
for the 12 inch freeboard were measured after the testing
with the 18 inch and the 24 inch freeboard testing, whereas
the first three measurements with the 12 inch freeboard were
made before. This was done to assure that uncontrolled
variables had not influenced the test results. The average
of the three results before the increased freeboard testing
was 25.7 liters per 24 hours and that after the increased
freeboard testing was 23.0 liters per 24 hours.
TABLE 2
Emission Rate Versus Freeboard Height
(With Air Movement)
2
Freeboard Solvent Used Lbs./Ft. -Hr. Emission Control
20 L./24 Hrs.
30 L./24 Hrs.
12" 27 L./24 Hrs. \ 0.373
21 L./24 Hrs.
25 L./24 Hrs.
1R» 18 L-/24 Hrs. \ n ,_. .
18 18 L./24 Hrs. j °'273 27%
24" 10 L'/24 Hrs' 1 n ifi7 SRa
24 12 L./24 Hrs. / °'167 55%
j
As noted before, the quiet air testing indicated no
improvement in solvent emission rate when the freeboard
was increased from 18 inches to 24 inches. The function
of the freeboard is to prevent the vapor zone from being
disturbed by air movements in the room. When the free-
board height becomes high enough to isolate the vapor
zone from the air movements in the room, further increases
-------
in the freeboard would not be expected to produce any
improved control of the vapor zone. However, when the
air in the room is more greatly disturbed as in the testing
summarized in Table 2, the isolation of the vapor zone would
not occur until higher freeboard ratios were employed. This
is well illustrated by the results obtained.
Using this information the graph on Figure 1 can be prepared.
The emission rates for the 12, 18 and 24 inch freeboard
heights construct a straight line. The dotted line
extension of this data is based on the quiet room testing
which indicates this system can be expected to loose
approximately 0.05 pound per square foot hour when it is
totally isolated from the room air movements.
Lip Exhaust Ventilation Effect
In this testing, the degreaser was operated with the
roll top cover in-place but open, seven days per week and
24 hours per day. One week of operation without the lip
exhaust ventilation consumed 33 1/4 liters. When the lip
exhaust was operated for similar period, 71 liters were
consumed. Allowing for the slightly decreased open top
surface area of the degreaser due to the roll cover
fixture, the solvent loss rate without ventilation was
0.083 pounds per square foot hour and with ventilation
-------
was 0.177 pounds per square foot hour. Solvent emissions
increased slightly over 210% with the lip exhaust vent
operating. The ventilation rate is described on the
equipment specification sheet.
-------
Shut Down Loss Rate
The solvent loss rate with a standard 12 inch freeboard
was found to be 0.023 pounds per square foot hour. Solvent
emission rates with the 18 inch freeboard of 0.020 and
0.023 during two 64 hour periods. Therefore, the added
freeboard has little effect on the solvent loss rate
experienced during shut down time when the degreaser is
left open. This data also suggests that it is much less
important to cover a degreaser when it is completely
shut down than when it is idling.
-------
Conclusions
1. Solvent emission rates from open top degreasers
can be reduced by at least 27% by increasing the
freeboard to degreaser width ratio to 0.75.
2. Emission reductions of between 43% and 55% can
be expected when degreaser freeboard ratios are
increased to equal to the width of the degreaser.
3. Automatic covers for open top vapor degreasers
which would maintain the degreaser covered when
idling or during shut down can be expected to
reduce emissions to the atmosphere between 0.1
and 0.4 pounds per square foot per hour of idling
vapor degreaser operation.
4. Lip exhaust ventilation can increase solvent
emission rates by over 200%.
-------
2 D 500 - 0 20
DETREX MODEL 2 D500 S,G and E, TWO-DIP, IMMERSION SPRAY, SIZE D 20 DEGREASERS
Applications and Advantages
1. The Model 2D500 Detrex degreasers
provide maximum flexibility of cleaning
cycles. Work can be cleaned:
(a) By immersion in boiling solvent, then
in a cool solvent rinse that is con-
tinuously replenished with clean sol-
vent distillate.
(b) By immersion in solvent vapors, then
in a cool solvent rinse and finally
in pure solvent vapors. Optionally a
spray pump and lance can be pur-
chased to provide a distillate spray
prior to the final vapor rinse and
drying cycle.
2. The Model 20 500 degreasers rapidly
remove greases, oils, wax and other
soils from all kinds of parts.
3. Available from stock.
Standard Features
4. Water jacket and auxiliary condenser
coil.
5. Spring loaded retractable roll-up cover.
6. Interior sidewall unobstructed.
7. All welded construction - Detrex FF-1
coated interior.
8. All units of material of construction
to allow choice of solvents -trichloro-
ethylene, perchloroethylene, or 1,1,1-
trichloroethane.
9. Lower safety temperature and liquid
level control for gas and electric units.
Optional Accessories Available
10. Solvent spray pump and spray lance
for distillate rinse.
11. Water separator.
12. Lip exhaust system
13. Vapor level control.
14. Water temperature controls.
Wet.r Out
3/4-
Ref. Dwg. SD 30.5020-1 81 -2
Steam-Heated
Optional Accessories (Cont'd)
15. Detrex Freeboard Chiller.
16. Degreaser body fabricated of type 304
stainless steel.
EXHAUST SYSTEM SPECIFICATIONS:
0. S.H. 1445 cfm @ 1" S.P. - 1 hp, 1750 rpm motor.
(Increases "A" dimension approx. 22 inches).
Ref. Dwg. SD 30.5021 -I
Gas-Heated
r
> i
1 1
i
, M
"1
/
V
j)
Work
Ref. Dwg. SD30.S022
Electric-Heated
SPECIFICATIONS
SIZE
020
OVERALL
DIMENSIONS
A
60"
B
33"
Q..
48"
MAX. WORK
CLEARANCE
D
20"
E
24"
F
12"
F1
10"
RATED
PRODUCTION
Lb Steel/Hr
1500
HEAT INPUT
STEAM
Lb/Hr
55
GAS
cf/hr@1000btu
80
ELECTRIC
kw
15
SOLVENT
CAPACITY
Gallons
70
WATER @
50°F RISE
gph
100
••Add 3" for Cover or Exhaust.
IM 1.4
i. 1/72
Litho in U.S.A.
-------
t *~r~ LoO
0,2-Q .
(5.10 -
O.OO-
\
\
\
1 \ \ I
* * _ 4 . n **
.foot — U
QU,
-------
m
Z
g
x
-------
APPENDIX D
STUDY TO SUPPORT NEW SOURCE PERFORMANCE
STANDARDS FOR SOLVENT METAL CLEANING OPERATIONS
Review of Existing Hydrocarbon Emission Regulations
Applicable to Stationary Sources
Prepared By:
Angela J. Williams
The Dow Chemical Company
Prepared For:
Emission Standards and Engineering Division
Office of Air Quality Planning
U. S. Environmental Protection Agency
-------
OF EXISTING HYDROCARBON EMISSION REGULATIONS APPLICABLE TO
STATIONARY SOURCES - SUB-TASK 6 OF PROJECT PLAN FOR EPA CONTRACT
As part of the EPA contract "Study to Support New Source
Performance Standards for Organic Metal Cleaning Operations",
a review of existing state anr1 local emission regulations
applicable to metal cleaning has been conducted.
This report categorizes and compares the emission control
5
-------
REVIEW OF EXISTING HYDROCARBON EMISSION REGULATIONS
APPLICABLE TO STATIONARY SOURCES
INTRODUCTION
Since 1966, when Los Angeles County developed regulations
to control the emission of hydrocarbons into the atmosphere,
many State and local areas have followed by developing and
implementing control regulations applicable to their own areas.
The Clean Air Act was created in 1963 with the goal of
establishing uniform State and local laws relative to the
prevention and control of air pollution, while recognizing
the need for practicable controls in the light of varying
conditions and requirements.
In a major revision of the Act in 1970, many amendments
occurred and the Environmental Protection Agency was formed
and made responsible for air pollution Control.
In 1971, the EPA published their criteria for " National
Primary and Secondary Ambient Air Quality Standards for
Photochemical Oxidants and Hydrocarbons" (Federal Register
Vol. 36, No. 84, April 30, 1971).
In Appendix B of this document, emission guidelines and
controls were promulgated in accordance with EPA's estimation
of the degree of emission control that could be attained using
available control technology.
Los Angeles County - Air Pollution Control District developed
and implemented "Rule 66" in mid-1966 and subsequently amended
it in 1971. It was developed as a result of close consultation
of the industries most affected by the regulations and ultimately
it was accepted as a workable approach to reduce the formation
of photochemical smog.
Most overall emission control plans which have been implemented
are based on either the Rule 66 or Appendix E approach.
DEFINITIONS OF CONTROL
RULE 66
The original rule defines a photochemir^lly reactive solvent
as any solvent with more tjian 20% of its total volume composed
of the following compounds, or which exceeds any of the
individual total composition limitations stated:
-------
1. Combination of hydrocarbons, alcohols, aldehydes
esters or ketones having an olefinic or cyclo-olefinic
type of unsaturation: 5 percent.
2. Combination of aromatic compounds with eight or
more carbon atoms per molecule, except ethylbenzene:
8 percent.
3. Combination of ethylbenzene, ketones having branched
hydrocarbon structures, toluene or trichloroethylene:
20 percent.
Whenever solvents may be described by the above classifications,
they are regarded as reactive and subject to the following
restrictions:
No more than 40 Ibs./day or 8 Ibs. in any one hour may be
discnarged, unless that discharge has been reduced 85%.
The 1971 amendment to Rule 66 exerts the following controls
of total solvent emissions after August 31, 1974:
No more than 3,000 Ibs./day or 450 Ibs. in any one hour of
organic material may be discharged from a single source unless
that discharge has been reduced by 85%.
APPENDIX B
This regulation exerts more stringent and controversial
limitations on solvent emission controls. It states that:
No more than 15 Ibs. in any one day or 3 Ibs. in any one
hour of organic compounds may be emitted unless the emissions
have been reduced at least 85% by: a) incineration - provided
that 90% or more of the carbon is oxid.i .red to carbon d^xide;
b) carbon adsorption.
The EPA regulations have identified hydrocarbon solvents as
pollutants as they are pre-cursors of photochemical oxidants.
However, some organic solvents have been shown to be virtually
photochemically unreactive and may be considered exempt from
the regulations. These include saturated halogenated hydrocarbons,
perchloroethylene, benzene, acetone and C, - C5 paraffins.
Solvents in this group may be considered! for exemption
only if sources are not major contributors to hydrocarbon
emissions.
-------
The following table summarizes the state and local district regulations
currently in operation.
STATE AND LOCAL AREA REGULATIONS AFFECTING
METAL CLEANING OPERATIONS
State or Local
Area
Alabama
Arizona—
Maricopa County
Puma County
California—
Los Angeles APCD
San Francisco Bay Area
Sacramento
San Diegc
San Bernadino
Orange County \
Riverside
Santa Barbara
Ventura
Colorado—
Denver AQCR
Connecticut—
D.C.
Illinois—
Chicago
Indiana
Kentucky
Louisiana
Maryland
Massachusetts--
Boston AQCR
New York >>
New York City \
Metropolitan Area j
New Jersey
North Carolina
Ohio
Oklahoma
Philadelphia
Rhode Island
Texar.--
Houston Galveston
San Antonio AQCR
Virginia
J
Wisconsin
Type of Regulation
Rule 66
Rule 66
*Rule 34A (Rule 66)
Rule 5 (App. B)
*Rule 66
Reg. 3 (Revised Rule 66)
^ EPA
} Imposed
J Regulation
Rule 66
Rule 66
Rule 66 - EPA Imposed
Rule 66 and App. B
App. B
*Rule 66
Rule 66
APC-15 (App. B)
AP-5 (Rule 66)
App. B
Rule 66
Rule 66 - EPA Imposed
Rule 66
Parts 204 and 205
*(Rule 66 Type With
Exception)
Original (Rule 66 Type)
Rule 66
Rule 66
Rule 66
Rule 65
AP-15 (Rule 66 Type)
Reg V (Rule 66 Type)
*Rule 66 and
App. B
Rule 66
Rule 66
Date of
Implementation
January 1973
August 1972
August 1972
August 1972
August 1971
November 1974
March 1374
August 1967
September 1973
May 1975
March 1973 - Jan. 1975
June 1974
December 1973
June 1974
October 1974
April 1972
February 1974
January 1973
February 1974
February 1974
August 1972
January 1974
Open
January 1972
February 1972
January 1972
July 1071
Open
January 1974
May 1975
June 1974
April 1972
-------
The Clean Air Act has scheduled State implementation .plans
for attaining ambient air standards be adopted no later than
May 31, 1975. However, in those cases where implementation
in 1975 is not economically or technically feasible, extensions
up to two years are being granted by EPA.
Regulations affecting metal cleaning operations have so far
been classified under "Organic Solvent Control" or "Control
of Emissions from Stationary Sources". There are only a few
promulgations actually stated for degreasing operations.
These include:
*Maricopa County—Rule 34 A
1. No person shall use or conduct any vapor phase
degreasing operation without minimizing organic
solvent vapor diffusion emissions by good modern
practices such as, but not limited to, the use of
a freeboard chiller cr other Affective device
operated and maintained in accordance with solvent
and equipment manufacturers specifications.
2. Spray degreasing shall be conducted in an enclosure
equipped with controls which will minimize the
emission of organic solvents.
*Houston, Galveston, San Antonio A.Q.C.R.
Under sub-heading "Control of Degreasing Operations", the basic
requirements are modeled upon Rule 66. Some exemptions to
to the Rules' restrictions are defined as follows:
1. Degreasing operations which emit less than 3 Ibs./hr.
or 15 Ibs./day of controlled organic materials.
2. Degreasing operations using pcrchloroethylene 1,1,1-
trichloroethane or saturated halogenated hydrocarbons
as an organic solvent.
Under sub-heading "Regulations for Limitations of Nev; Sources",
Appendix B has been adopted and becomes effective May 31, 1975.
*Los Angeles, S?n Diego, Sacramento Valley, San Joaquin Galley
and San Francisco Bay Area Intrastate Regions - January 1975
"Control of Deqreasing Operations"
(a) Degreasing means any operaton using an organic
solvent as a surface cleaning agent prior to
fabricating, surface coating, electroplating, or
any other process.
-------
(c) Any organic emissions must be reduced by 85% o£
solvent must be classfied non-photochemically"~reactive-
Degreasing operations using perchloroethylene or
saturated halogenated hydrocarbons shall be exempt.
EFFECT OF REGULATIONS ON THE METAL CLEANING INDUSTRY
The largest change since legislation was first introduced
is the decrease in use of trichloroethylene in vapor
degreasers. With two exceptions (*New York City Metropolitan
Area - Part 205, January 28, 1974 and Illinois, December 1973)
trichloroethylene is classified as a photochemical reactant and
therefore subject to severe emission limitations. Since,
for most operations, 85% reduction of emissions is
impossible to meet, the alternative is to change solvents
to using a non-photochemically-reactive classified solvent,
which would then be exempt from severe restrictions, xne
solvents most used from this category are 1,1,1-trichloroethane
and perchloroethylene.
These solvents are used in degreasing operations within
the workable, less stringent emission controls by moderate
equipment modifications and recommended improved work habits.
SUMMARY
Future legislation specifically for metal cleaning should
consider the large quantity of miscellaneous hydrocarbons
used in cold cleaning applications. Generally, small amounts
of solvent are involved for each individual operation, but
collectively the solvent loss from these add up to a large
volume of emissions without any form of control.
Overall regulations, whether modeled on Rule 66 or Appendix
B type of control, as applied to stationary sources are not
wholly viable when applied to vapor degreasing.
In this extremely varied industry, the design and type of
degreasing equipment influence the volume of solvent vapors
emitted.
WorkabJ.e controls could be postulated taking into account
the following:
. Areas of solvent vapor/air interface
. Surface area of metal cleaned
. Installation of equipment to limit vapor loss (covers,
free board chillers, etc.)
-------
. Consideration of solvent vapor recovery systems
(carbon adsorption)
. Improvement of degreaser operating habits (manual
spraying, work racking, hoist speeds, drainage time, etc.)
. Disposal of solvent residues
The solvent industry is cognizant of the necessity for controls
to limit solvent vapor emissions from both health and economic
viewpoints.
Regulations applicable to this industry should take into
consideration the energy and economic impact of emission
controls, as recognized by the Environmental Protection Agency.
-------
Appendix E-l
DOW CHEMICAL U.S.A.
DEPARTMENT:
Inorganic Chemicals
UJ
CL
o
u
F'IRM NAME (In lull, do not abbreviate)
Safety Clean Corporation
S FRKE T
655 Big Timber Road
CITY
Elgin
STATE
Illinois
ZIP CODE
60120
UJ
E. J. Fort, Chicago
S. J. Nolan, Chicago
R. R. Lapine, 2020
G. E. Forrest, 2020
C. R. Crabb, 2020
R. T. Gerard, 2020
KOG/PF 2192008
PF.RSONS INTERVIEWED AND TITLES
Mr. Harry Logue, Sales Manager
Mr. Allen Manteuffel
/VRITTENSY
K. S. Surprenant
FIELD DOW PHONE
X
DEVELOP. SERV. OTHER
x
DATE CALLED
3/25/75
D » TE Vi SI T T EN
3/26/75
OTHER DOW PERSONNEL PRESENT
None
SUBJECT
EPA CONTRACT ON SOLVENT METAL CLEANING
DC.
<
ID
I/O
Oi
Discussions with Graymills Corporation and Kleer-Flo both indicated
that Safety Clean Corporation makes parts washers and provides
them to customers on a lease basis. This leasing arrangement
involves both the supply of solvent for metal cleaning as
well as the equipment.
Mr. Manteuffel indicated that Safety Clean Corporation has
about 140,000 units under lease at the present time. These
metal washers come in two sizes, 16 gallon capacity and
30 gallon capacity. The average fill is approximately
12 gallons. Customers are set up on a solvent replacement
cycle from two to six weeks. The annual replacement need
for solvent to service these customers amounts to 16 - 17
million gallons. Mr. Manteuffel estimated that the solvent
use of these customers would be .in the range of 40 million
gallons per year if they were disposing of this solvent in
landfills rather than using Safety Clean Corporation's service
which recovers the solvent for reuse.
-------
Appendix E-2
DOW CHEMICAL U.S.A.
DEPARTMENT:
Inorganic Chemicals Department
J. C. Car lav;, Sarnia
D. R. Heinz, 2020
H. R. Krimbill, 2020
R. R. Lapine, 2020
S. J. Nolan, Chicago
D. W. Richards, 2020
A. J. Williams, 2020
KOG/PF 2192008
h'ifc'.; HkMf (In full, do not obbrcviote)
Graymills Corporation
s T -< E ?: T
3759 N. Lincoln St.
Chicago
STATE
Illinois
ZID CODE
60613
Mr. Ed Roels
K. S. Surprenant
FIELD OOf
X
DEVELOP. SF.RV
x
. OTHER JO ATE C A L L E C
{L-15-75
1-29-75
OTHE» DO// ret. so>. -.5.1. PRESENT
A. J. Williams
SLBJECT
EPA CONTRACT ON SOLVENT METAL CLEANING
The purpose of this visit was to obtain additional background
on solvent room temperature cleaning or "cold cleaning". Graymills
Corporation is a leading manufacturer of tanks for this cleaning
method which they refer to as parts washers. They also offer
a line of solvents under the trademark Agitene Solvents which
are both straight petroleum solvents and solvent blends made
with chlorinated hydrocarbons and other chemicals. Our discussion
brought out the facts that cold cleaning requires the least capital
investment, is completely portable and requires no energy consumption,
Graymills have been making parts wahser's since 1945 and have marketed
as many as 15,000 units per year. They estimate that between 750,000
to 1,000,000 units are in operation in the U. S. today. The parts
washers require five gallons upwards to 100 or 200 gallons for an
initial fill. The most common model holds about 30 gallons. In
the automotive after-market, it was estimated that this volume of
solvent is changed twice per year whereas the industrial users
change the solvent from every two weeks to four times per year.
The waste solvent was reported to be returned to drums and sold
as scrap. Make-up solvent between solvent changes probably equals
the volume of solvent contained in the parts washer. In addition,
it was estimated that as many as an equal number of units are
"homemade" from solvent drums or pails. Kleer-Flo was reported
to be their major competitor and located in Eden Prairie, Minnesota.
Safety Clean Corporation of New Berlin, Wisconsin leases similar
solvent cleaning equipment but does not sell it. Lesser competition
is received from Phillips Manufacturing and Gunk Incorporated. The
life of this equipment is very long since it experiences little or
no corrosion. Ed Roels has no doubt that some of their equipment
is still in operation from their first year of manufacture. Safety
Clean Corporation reportedly has 120,000 units under lease in.the
field right now.
-------
Appendix E-3
DOW CHEMICAL U.S.A.
DEPARTMENT: Inorganic Chemicals
u
Ftr*M NAME (In full, do not obbreviate)
Klppr-Flo Corp.,
STHEET
6600 Washington Ave., S.
CITY
Eden Prairie
STATE
Minnesota
ZIP CODE
55346
UJ
WRITTEN BY |F
K. S. Surprenant & /-> |
IELD DOW PHONE
X
DEVELOP. SERV. OTHER
X
DATE CALLED DATE /* <=! ; 7 - *_ • .
1/23/75 2/17/-7c
OTHER DO« PERSONNELPRESENT
None
SUBjEC T
EPA CONTRACT ON SOLVENT
METAL CLEANING
UJ
tx.
j. C. Carlaw, Sarnia
D. R. Heinz, 2020
H. R. Krimbill, 2020
R. R. Lapine, 2020
S. J. Nolan, Chicago
D. W. Richards, 2020
A. J. Williams, 2020
KOG/PF
INTERVIEWED Lr;O TITLES
Mr. Graham Pendelton
Mr. Ed Roels of Graymills Corp. indicated that Kleer-Flo
Corp. is their main competitor and would be able to
offer additional information on the cold cleaning market.
After reviewing both companies literature and having
discussions with both, they obviously compete head-to-head.
Kleer-Flo has been marketing parts washers (cold cleaning
tanks), petroleum solvents and safety blends since 1936.
Graham Pendelton estimated the total number of parts washers
manufactured by Kleer-Flo or other competitors to be between
the one half to one million units. Truck and car dealers
and independent garages as well as industrial accounts are
major consumers. The automotive dealers probably average
two to three units each whereas garages average one to two
units. The volume of an average parts washer would be in
the twenty-five to thirty gallon range. Operational instructions
as provided by Kleer-Flo suggest that the solvent be changed
every two to three months. However, in practice many firms
drain the system only once every six months. Graham indicated
the evaporation loss to be 5 to 10 gallons of solvent between
changes. Again as in the case of Graymills Corp., their prime
profit center is in the equipment manufacturer. Graham noted
that Safety Clean Corp. of Wisconsin has greater than 100,000
units on lease, primarily in auromotive after market applications,
sm
-------
Appendix E-4
DOW CHEMICAL U.S.A.
DEPARTMENT: Inorganic Chemicals Dept.
t-iRM NAME (In full, do not obbreviote)
Horton Co.
3TRSE I
2070 Brooklyn Rd.
CITY
Jackson,
STATE
Michigan
ZIP CODE
49203
W^l TTEN BY
K. S. Surprenant ;f55
0THER DOW PERSONNEL PRESENT
None
FIELD DOW PHONE DEVELOP. SERV. OTHER
X X
DATE CALLED DATE WRITTEN
7/1/75 7/30/75
S_ bjEC T
EPA CONTRACT ON SOLVENT METAL CLEANING
L. W. Stump, Detroit
J. W. Hennington, 2020
C. R. Crabb, 2020
D. W. Richards, 2020
KOG/PF 2192008
PERSONS INTERVIEWED AND TITLES
Mr. Robert Ziegler, President
Summary
The Horton Company replaced a small open top degreaser with
a Phillips Manufacturing Rotomatic degreaser. Solvent records
show that the larger more enclosed degreaser system is consuming
at least 40% less solvent on a work load basis and less than
a third as much solvent loss on a sq. ft. of exposed vapor zone
basis. The newer equipment demonstrates two generally recognized
facts:
1. Enclosed conveyorized equipment is much more
efficient in controlling solvent losses than a
number of open top degreasers having equal work
capacity.
2. The sale of enclosed conveyorized vapor degreasers
often reduces the volume of solvent used in the
market rather than increasing it.
Discussion
In earlier conversations with vapor degreasing equipment
manufacturers, reports were received indicating that the
sale of some large cross-rod or monorail type vapor degreasing
systems can result in less solvent usage by the customer
rather than more. The reason offered for this was that large
conveyorized vapor degreasers often replace several open top
vapor degreasers. Because of the greater control over solvent
losses in the conveyorized equipment, the solvent requirement
for the firm involved is actually reduced and sometimes sub-
stantially (see Baron Blakeslee report on January 16, 1975 trip).
j 7 HI' P P f. ! »
-------
Horton Co.
Prior to January 1974, Horton Manufacturing Co. had a gas heated
open top degreaser. The open top dimensions of this degreaser
were 36"x24". This degreaser was replaced by a Rotomatic vapor
degreaser manufactured by Phillips Manufacturing Co., Model
60TRMD. This degreaser is electrically heated with a 24kw rating
the Rotomatic is supported with a still (Model RS-30) having
a 15kw heating capacity. These pieces of equipment are sketched
on the attached sheet with no scale intended. The left figure
represents the open top degreaser while that on the right
represents the Rotomatic and still. The Rotomatic degreaser is
hooded over a substantial proportion of the degreaser system and
has a ferris wheel type of material handling system. This design
is similar to a cross-rod vapor degreaser in that the handling
system causes the parts to rotate providing improved solvent
drainage from the parts. However, the Rotomatic design is less
enclosed then a cross-rod degreaser.
Both the open top and the Rotomatic vapor degreasers have been
operated on a one shift per day five days per week basis.
The open top degreaser was closed during non-operating shifts
on weekends and was equipped with a manually rotatable parts
fixture. This degreaser was not equipped with a still. The
Rotomatic degreaser is equipped with a still, as mentioned,
which provides much greater capability of recovering solvent
from the oil removed from parts. The open top surface area
of this degreaser including that which is located under the
hood is 100"x33". These dimensions in addition to the ferris
wheel design make the use of a cover during non-operating
hours difficult. Horton Co. solved this problem by pumping
all of the solvent in the degreaser into storage tanks during
down shifts and weekends. Of course, this is a highly efficient
wasy to prevent solvent evaporative losses during non-working
hours.
Solvent records were available for over a year before the
new equipment was installed and over half a year after the
conversion to the new equipment. These records indicated 37
drums of 1,1,1-trichloroethane were consumed during a 58
week period prior to the new equipment installation or 0.64
drums per week. The solvent consumption for a 29 week interval
with the Rotomatic degreaser consumed 22 drums of 1,1,1-
trichloroethane. This amounts to 0.76 drums per week. Although
this indicates a solvent consumption of 18% greater then that
experienced with the small open top degreaser, the work load
being processed was estimated to be two to three times greater
after the installation of the Rotomatic degreaser. Thus, on
a work load basis, the solvent consumption had been reduced by
at least 40%. Using 600 pounds/drum, 384 pounds of 1,1,1-
trichloroethane were consumed per week in the open top degreaser
and 456 pounds of solvent were consumed per week in the Rotomatic
-------
Horton Co.
degreaser. The solvent consumption rate can be determined on
a pound per hour per sq. ft. basis. Expressed in this way, values
of 0.50 and 1.60 Ibs/hour/sq. ft. are found for the Rotomatic
and the open top respectively. This calculation assumes that
no solvent is lost during non-operating time. On this basis
the Rotomatic degreaser is three times more efficient then the
open top.
sm
-------
Open Top Degreaser
®
rly-=?r
*"Hr—-|»
i
Rotomatic Degreaser
Still
1. Condenser Coils
2. Freeboard Chiller
3. Hood
4. Fan
5. Rotating Baskets and
Rotating Fixture
-------
Appendix E-5
DOW CHEMICAL U.S.A.
DEPARTMENT- Inorganic Chemicals
111
J. C. Carlaw, Sarnia
D. R. Heinz, 2020
H. R. Krimbill, 2020
R. R. Lapine, 2020
S. J. Nolan, Chicago
D. W. Richards, 2020
A. J. Williams, 2020
KOG/PF 2192008
Central Files
F'SM NtME (In full, do not obb'eviote.i
BARON BLAKESLEE
1620 South Laramie Ave. Chicago
STATE
Illinois
ZIP CODE
Of.
<
Mr. R. A. Kullerstrand, Vice-President
Mr. Parker Johnson, National Sales Manager
WRITTEN BY
K. S. Surprenant /(^
FIELD DOV. PHONE DEVELOP. SERV. OTHER D»TE CALLED O 4 r E /. Si r - E-.
x x 1/16/75 1/29/75
GT^E« DOW P£^SONNEt. PRESENT
A. J. Williams (Dow) , W. Johnson (EPA)
s'jejtc i
EPA CONTRACT ON SOLVENT METAL CLEANING
Baron Blakeslee is the leading manufacturer of vapor
degreasing equipment. This discussion was arranged to
provide further background on solvent metal cleaning
operations, to determine any new emission control
technology for controlling solvent emissions and
to provide Baron Blakeslee with the opportunity to suggest
sites to evaluate emission control methods.
As a major supplier of vapor degreasing equipment and
solvents used in this process, the Baron people felt
that they could be of help in discussing vapor degreasing
but could offer little information on solvent room temperature
or cold cleaning. However, they agreed that the volume
of solvent used in room temperature or cold cleaning is
very large. Mr. Kullerstrand and Mr. Johnson agreed that
there are approximately 25,000 vapor degreasing systems
operating in the U.S. It was estimated that 70% of these
systems are of the open top design. Approximately 2,500
vapor degreasers are manufactured per year with a life
expectance of 15 years or more. Less than one half of these
units were estimated to be replacement of existing systems.
In some cases the sale of a new vapor degreaser can result
in less solvent use rather than more. This occurs where a
conveyorized degreaser replaces one or more open top degreasers,
-------
-2-
Such was the case at Tillotson, Toledo, Ohio where one cross-rod
vapor degreaser reduced the solvent consumption from 25 drums
per month to about 5 drums per month by replacing two open top
vapor degreasers. Mr. Johnson emphasized the point that vapor
degreasing equipment is sold to all kinds of industries not only
metal fabricating plants. These would include maintenance
operations at paper mills or textile mills, automotive repair
and rebuilding shops, heavy duty construction equipment
maintenance and agricultural equipment maintenance locations
as well. This diverse use of vapor degreasing was illustrated
by the vapor degreasing of candelabra at St. Patrick's Cathedral
and by degreasing skeletons for laboratory and educational
use.
Parker Johnson observed that vapor degreasing is different
from cold cleaning by its ability to distill and recover
pure and reusable solvent. When a still is attached to
a vapor degreaser he pointed out that the oils removed from
metal parts can be concentrated up to 90% or better in the
case of perchloroethylene and trichloroethylene^by steam
distillation. With 1,1,1-trichlorotthane, steam distillation
is not recommended and he noted as much as 30% by volume
of solvent may be residual in the concentrated still residues.
In the Baron Blakeslee solvent distillation recovery service
portion of their business, the dirty solvent received for
reclaiming usually contains between 50 and 90% solvent most
of this solvent can be recovered by distillation and is
returned to the customer as a service or sold to new customers.
Mr. Johnson reported that Baron Blakeslee also sells carbon
adsorption recovery equipment and the Econ-0-coil. The latter
is a refrigerated water system to reduce solvent vapor emissions
from vapor degreasing operations. It supplements the existing
condenser coils used to control the vapor zone within the
vapor degreaser, these two techniques were the only two
techniques which had been observed in the U.S. market place.
No solvent recovery or emission control devices were reported
to be used in connection with cold cleaning. The Baron
people reported that 40% solvent recovery can be expected on
open top vapor degreasing equipment but on specially designed
and conveyorized equipment recoveries up to 85% are possible.
However, no known company installations are performing this
well.
Mr. Kullerstrand and Mr. Johnson agreed that vapor degreasing
is regarded as a conservation device vs. alkaline washing
or a cold cleaning. In addition to the water pollution
-------
-3-
created by alkaline washing, an alkaline washing system
reportedly consumes 8 times as much energy as a comparable
vapor degreasing system. As many as 30 to 40% of the vapor
degreasing operations do not require ventilation whereas all
alkaline washing equipment does which requires additional
heat to condition the replenishment air for the factory. They
agreed that locating a single plant where both vapor degreasing
and alkaline washing could be compared on a direct basis
would be very difficult if not impossible. It was suggested
that a real vapor degreasing operation could be studied and
compared to a papei; analysis of a comparable alkaline washing
operation to develop comparable data on the two systems.
Western Electrics Hawthorne Plant was suggested as an excellent
location to evaluate various vapor emission control systems
due to their large number of installations and the knowledgability
of Mr. James Weston of these systems. (Mr. Weston will be
contacted as an evaluation cite).
sm
-------
Appendix E-6
DOW CHEMICAL U.S.A.
DEPARTMENT:
Inorganic Chemicals Dept.
v>
Ul
a.
o
u
S3WVN 1
\j t w-iiianjvwiiiiu ••
F-tRM NAME (In full, do not abbreviate)
Auto Sonics Co. <215> 828-9090
S r F.EE T
P. 0
PE RSON
CITY STATE ZIP CODE
. Box 300 ' Conshohocken , Pennsylvania 19428
S INTERVIEW ED AND TITLES
WRITTEN BV FIELD DOW PHONE DEVELOP. SERV. OTHER DATE CALLED DATE WRITTi:::
K. S. Surprenant .^ x x 5/9/75 5/13/75
OTHER
None
DOW PERSONNEL PRESENT
SOLVENT METAL CLEANING CONTRACT FOR EPA
Ji-
ll:
<
^.
D. W. Richards, 2020
R. R. Lapine, 2020
R. C. Ormiston, Philadelphia
Carlaw, Sarnia
Metal Cleaning Group
Non-Metal Cleaning Group
KOG/PF 2192008
This phone conversation was made to determine'current pricing
of the cold trap. This pricing data will be used to determine
the cost benefit relationship of the cold trap installations to
be tested under the Dow- EPA contract on solvent metal cleaning.
The compressor size is determined by the peripheral footage of
an open top degreaser. The peripheral footage is calculated as
follows:
- Less than 42" width -- 2 (width plus length)
- 42" to 72" width -- 2 (width plus length) times 1.5
- Greater than 72" width -- 2 (width plus length) times 2.0
The footage corrections for degreasers having widths of 42" to 72"
or greater than 72" is done because an extra refrigeration coil
is added for the 42" to 72" width and two extra refrigeration
coils for degreasers with greater than 6' width.
The size of compressor needed and Its cost can be obtained from
the following table:
Horsepower
Peripheral Footage
Dollars
J
£
1/2
3/4
1
1-1/2
2
3
10'
16'
21 '
35'
47'
70'
$2,510
2,595
3,662
4,795
5,575
6,265
-------
-2-
To obtain the total purchase price a value for the peripheral
footage should be added to the compressor cost. Peripheral footage
cost can be obtained by adding the degreaser length and width"and
multiplying by two. The cost 1s $32.00 per foot for a degreaser
having a width of less than 42". $41.00 per foot is the cost for
degreasers 42" to 72" in width. Degreasers 6' wide and larger
cost $50.00 per foot.
-------
Appendix E-7
DOW CHEMICAL U.S.A.
DEPARTMENT:
D. N. DeMott - 2020
R. R. Lapine - 2020
D. W. Richards - 2020
A. J. Williams - 2020
KOG/PF 2192008
H. A. Pastor - 2020
J. C. Copus - 2020
T. A. Vivian - 2020
R. T. Gerard - 2020
FIRM NAME (In lull, do not obbreviote)
Vic Manufacturing Company
STREET
1620 Central Ave.
CITY
Minneapolis
STATE
Minnesota
ZIP CODE
Ul
PERSONS INTERVIEWED AND TITLES
Mr. Irving Victor, Executive Vice-President
Mr. Oscar Victor, President
Mr. Joe Barber, Chief Chemist
Charles Gorman, Sales Manager
WRl T T EN 6V
K. S. Surprenant A'X
FIELD DOW PHONE
DEVELOP. SERV. OTHER
DATE CALLED
12/13/74
OTHER DOW PERSONNEL PRESENT
K. 0. Groves
DATE f RI T 7 E s.
12/20/74
SUBJEC T
CARBON ADSORPTION
Non-Dow Personnel
Mr. David Patrick - EPA
Mr. William Johnson - EPA
Mr. Billy McCoy - TRW Corp.
This call was made to provide an indepth background and understanding
of the process of carbon adsorption as it can be applied to both chlorinated
and flammable solvents used in metal cleaning and drycleaning.
Historically, Vic Manufacturing began marketing carbon adsorption recovery
equipment to the drycleaning market in 1957 and to the metal cleaning
(vapor degreasing) market place in 1958. They have maintained a dominate
position in both markets ever since. The basic Vic carbon adsorption
patents will expire in 1976, however, they should maintain their position
in the market even without patent protection. The typical efficiency
of carbon adsorption on open-top or cross-rod degreasers is estimated to
be in the range of 50 to 60%. The recovery efficiency on a monorail
degreaser is estimated to be about 70%. With special design considerations,
maximum efficiency are estimated to bo in the range of 90% unrter expertly
controlled conditions. With best design techniques, the solvent recovery
efficiency for typical vapor degreasing operations could be improved «-o
about 70%. In drycleaning, the recovery efficiency is estimated at 50»
savings as a general rule. Vic Manufacturing ;:i 11 guarantee solvent
recovery efficiencies up to 50% in both market places. A crude estimate
of the cost of carbon adsorption equipment can be made from the volume
of air which needs to be processed. The equipment can cost as low as
$2.00 per cfm. Common equipment used to recover solvent from vapor
degreasing operations cost between $4.00 and $4.5 per cfm based on the
maximum bed gas flow rates.
-------
Carbon Adsorption - Metal Cleaning Processes
Vic Manufacturing estimates that they have sold about 1,000
carbon adsorption systems to metal cleaning operations and
that they have supplied 90 to 95% of the carbon adsorption
systems to this use. All sales to date have been motivated
by the dollar savings the customer could obtain. Some
interest is being noted by Vic Manufacturing of customers
wanting adsorption to comply with air pollution regulations.
The typical open-top degreaser ventilated at a rate of 60 cfm
per square foot will experience a recovery efficiency of 50 to 60%.
They believe that an open-top degreaser with an optimum design
and top notch maintenance and operation of carbon adsorption can
recover 90 to 95% of the solvent which would escape the system
without recovery equipment. The ventilation rate used in this
best design system is 100 cfm per square foot. Using 100 cfm
per square foot in a typical open-top degreasing operation, they
expect that a 70% recovery would be common. Cross-rod degreasers
are ventilated on tho basis of the vertical section open are:, by
the loading station. Ventilation rates of 35 to 50 cfm per square
foot are used and recovery efficiencies of 50 to 60% are typical.
Monorail degreasers are ventilated at the rate of 35 cfm or greater
per square foot of an open area at both inlet and outlet and obtain
70% recovery. The carbon beds are designed to retain the processed
air for a minimum of 1/100 of a minute. Bed depths run from a
minimum of 9 inches to a maximum of 32 inches and the volume flow
is calculated based on a 100 ft. per minute per square foot of bed,
the working bed capacities are shown below:
Trichloroethylene — 15% by weight of the carbon bed
Perchloroethylene — 20% " " "
Methylene Chloride — 10% " " "
1,1,1-trichloroethane — 12%
Toluene/Xylene — 6% " " "
Stoddard Solvent — 6-7% " "
In general carbon adsorption beds are designed to operate on the
adsorption cycle for at least 1 hour. For vapor degreasing operations
the typical bed would have a four hour capacity. Detection systems
are available to determine breakthrough or saturation of the bed
for combustable solvents, but a similar system is not available for
halogenated solvents.
Typical equipment is mild steel with baked phenolic as a protective
coating. Equipment life is expected to exceed 10 years and some
units have been operating for 15 years. Again, in the case of
bed life the expected life is 10 years or more. 1,1,1-trichloroethane
equipment is designed of Hastaloy and is accompanied with sodium
carbonate neutralization and calcium chloride drying beds. Carbon
adsorption equipment prices change annually or semi-annually and have
been inflating at a rate of 5 to 11% per annum. The equipment is
sold f.o.b. plant and freight and installation costs are estimated to
be between 15 and 25%. Vic Manufacturing estimated the cost of
maintenance and operation to be 3 to 5% of the equipment cost.
-------
Solvent recovered from adsorbers has reduced levels of stabilizers
but can be reused when recovery efficiencies are 50 - 60% or less.
Higher recovery efficiencies require restabilization of the solvent
for reuse.
-------
Carbon Adsorption - Drycleaning
Vic Manufacturing estimates the number of perchloroethylene
drycleaning plants to be about 16,000. Of this number
Vic has sold 7,000 carbon adsorption units and estimates
that Hoyt Manufacturing (a licensee of Vic Manufacturing)
has sold another 3,500. Other carbon adsorption manufacturers
account for about 10% more bringing the total to about
11.6 thousand carbon adsorbers. Of this total, the Vic
Manufacturing share of market is about 70%. This agrees
with their own estimate of share of market. Model 128
is used for drycleaners loosing one drum of solvent per
month or less and is a most commonly sold machine (about
3,000 machines). The next most common model is Model 118
for one to two drum customers (about 2,600 machines have
been sold). Model 108 is designed for customers using
more than two drums of solvent per month (about 900 machines
sold). Again, a 50% effiency and saving of solvent is
estimated for these units. Model 415 is designed for
coin operated drycleaners and about 100 machines have
been sold.
Vic estimates that there are about 12,000 Stoddard
drycleaning plants and that they process approximately
60% of the total drycleaning market. Perchloroethylene
is estimated to process the 40% balance. 3.5% of the
the drycleaning was estimated to be done by Valclene
(trichloro-trifluoroethane). Almost none of the
Stoddard plants have carbon adsorption. Carbon
adsorption equipment for Stoddard plants cost more
because of the lower capacity (6 to 7% by weight of
bed) than that of perchloroethylene (20% by weight)
and because vapor concentrations must be held to less
than 25% of the lower explosive limit. It is estimated
that the equipment pay off for carbon adsorption at
Stoddard plants will require 2 1/2 to 8 years.
Mo mileage rates were offered for Stoddard solvent.
Perchloroethylene mileage on dry to dry or transfer
machines without carbon adsorption was estimated to
be 5,000 pounds per drum. With carbon adsorption the
mileage was estimated to be about 10,000 pounds per
drum and the best systems with carbon adsorption were
reported to achieve 20,000 pounds per drum. The average
income per drycleaning plant was estimated to be about
$800. per week. Income can be converted to pounds of
clothes cleaned using a rate of $1.00 to $1.10 of income
per pound of clothes.
New drycleaning equipment sales have been slow for the
past year or so. Current equipment sales in drycleaning
are about 60% for Valclene and 40% for perchloroethylene.
-------
Large numbers of used perchloroethylene machine are
on the market reducing the demand for new equipment
of this type. Refrigeration is being used instead of
carbon adsorption for the new Valclene equipment. Two-ton
refrigeration capacity units are used for this system and
the heat exchange temperature is set for about 0°F. Vic
estimates the cost per pound of clothes for Valclene at 2C
and for perchloroethylene at 1C. The Valclene equipment is
not vented to the world but uses a plastic lung or bag
to collect the final drying air. The cost of Valclene
was estimated to be about $8.00 per gallon in 5 - 20 gallon
quantities.
-------
l>
SPECIFICATIONS
Vic Air Pollution Control System
(Contact factory for solvents not listed. )
MODEL Single
534 700
1
536 800
f, i
\SSO xvjk*^»— J
548
554
M"r 3 #\
572
&2i m
; 584
I
•*
r
596
1400
1700
3000
3800
5000
VI
Double
-
1300
-
LBS OF
CARBON
Per Tank
150
350
400
3000 | 1000
5500
7500
10.000
j
1500
3000
4500
SOLVENT
Frichloroethylene
Perchloreothylene
Toluene
Freon TF
Methylene Chloride
Trichloroethylene
Perchloroethylene
Toluene
Freon TF
Methylene Chloride
Trichloroethylene
Perchloroethylene
Toluene
Freon TF
Methylene Chloride
Trichloroethylene
Perchloroethylene
1 Toluene
Freon TF
Methylene Chloride
Trichloroethylene
Perchloroethylene
Toluene
Freon TF
Methylene Chloride
Trichloroethylene
Perchloroethylene
Toluene
Freon TF
Methylene Chloride
Trichloroethylene
Perchloroethylene
Toluene.
Freon TF
Methylene Cliloride
LBS SOLVENT
PER TANK/HR
23
30
9
12
15
53
70
21
28
35
60
80
24
32
40
150
180
60
80
100
225
300
90
120
150
450
600
180
240
300
675
900
270
300
450
LBS STEAM
PER HOUR
b7
90
32
42
45
158
210
84 *
98
105
180
240
96
112
120
450
540
240
280
300
675
900
360
420
450
1350
1800
720
840
900
2025
2700
1080
1260
1350
CONDENSING WATER - GPM ;
fr65«
2. 7
3. 5
f
6. 5
8. 5
7
9
18. 5
22
27. 5
37
55
73.5
82
110
(585°
4
5
2
10
13. 5
5. 5
11. 5
15. 5
6. 5
29
35
15. 5
43. 5
58
23. 5
87
115
47
130
173
70
Primary
85°-120°
2.5
2.8
6
6.5
7
8
17
19. 5
25. 5
28
51
56
76
84
Secondary
50°- 60°
— I
1 '
1
1. 5
1.7
2
2. 5
4. 5
5
7
8
14
16
20
24
-------
Appendix E-8
DOW CHEMICAL U.S.A.
DEPARTMENT Inorganic Chemicals Department
UJ
Q.
O
U
C. R. Crabb, 2020
R. R. Lapine, 2020
K. P. Schultz, Stamford
M. Meglio, Boston
D. W. Richards, 2020
KOG/PF 2192008
FIRM NAME (In full, do not obbreviate)
HOYT MANUFACTURING
CITY
Westport
STATE ZIP CODE
Massachusetts 02790
LLJ
•s.
PERSONS INTERVIEWED AND TITLES
Mr. Derek Oakes, Vice President
WRITTEN BY
K. S. Surprenant
X
DEVELOP. SERV. OTHER
04TE CALLED
5-28-75
DATE V'RI T TEN
6-11-75
OTHER DOW PERSONNEL PRESENT
Mr. John Bellinger, EPA
SUBJEC T
EPA CONTRACT ON SOLVENT METAL CLEANING
>-
Oi
<
Hoyt Manufacturing was visited to advise them of the low recovery
efficiency being obtained from the carbon adsorption system at
J. L. Thompson, Waltham, Massachusetts. The J. L. Thompson
location as a test site for carbon adsorption recovery was suggested
by Hoyt Manufacturing and the equipment was supplied by them.
Mr. Oakes agreed to have one of his engineers reinspect the
operation as soon as possible after hearing that our recovery
efficiency had not improved over the earlier report. Since
then, Mr. Oakes has reported that his engineer found the air
inlet damper to bed "B" broken and in the closed position so
that bed "B" could not effectively adsorb solvent vapors. This
agrees well with the observation that no solvent was recovered
from the desorption cycle of bed "B" during the test evaluation.
The model located at J. L. Thompson was Model 536AD. This model
is now redefined as Model ABRS15F and is constructed of mild
steel and Heresite coated. The system contains 300 pounds of
carbon per bed and has the capability of recovering 45 to 50
pounds of trichloroethylene per desorption cycle. Mr. Oakes
indicated that the normal ratio of steam condensate to pounds
of solvent recovered is three to four pounds of steam to one
pound of solvent. This model now costs approximately $8,000.
The installation costs can vary considerably depending upon the
amount of vent ducting and the availability of utilities
(compressed air, steam, electricity) in the immediate area of
installation. However, 15% of the purchase price was thought
to be a reasonable estimate for installation.
F OHM e-'-W PRINT i ;j n-3-7.?
-------
-2-
Mr. Oakes indicated that about 200 Hoyt units have been sold
for solvent metal cleaning operations. Bed efficiency is
over 95%. Hoyt Manufacturing use to guarantee 50% overall
recovery on a total solvent consumption basis and never had
a unit returned to him on this basis. Normal maintenance
costs were estimated to be about $200 per year. Titanium clad
or Hastaloy C are recommended for solvent recovery of 1,1,1-
trichloroethane although they have limited experience in
recovery of this solvent.
-------
Appendix E-9
DOW CHEMICAL U.S.A.
DEPARTMENT: Inorganic Chemicals
u
FIRM. NAME (In full, do not obbreviote)
Detrex Chemical Industries Inc.
STREET CITY
14331 Woodrow Wilson Detroit
STATE
Michigan
ZIP CODE
48232
WRITTEN BY
K. S. Surprenant £-.5
FIELD DOW PHONE
X
DEVELOP. SERV. OTHER
X
DATE CALLED
3/7/75
DATE WRITTEN
3/21/75
OTHER OOW PERSONNEL PRESENT
None
•SUBJECT
EPA CONTRACT ON SOLVENT METAL CLEANING
vO
C. R. Crabb, 2020
R. T. Gerard, 2020
J. W. Hennington, 2020
R. R. Lapine, 2020
D. W. Richards, 2020
L. W. Stump, Detroit
KOG/PF 2192008
Central Files
PERSONS INTERVIEWED AND TITLES
Mr. Tom Kearney, Engineer
Mr. Robert Clark, General Sales Mgr.
Mr. Robert White, Manager Gold Shield Sales
The primary purpose of this call was to discuss the EPA Contract
with Mr. Tom Kearney, one of the most experienced men in solvent
metal cleaning equipment. An earlier letter as a result of
a phone conversation is included with this call report discussing
solvent losses from various kinds of metal cleaning operations
and comparing alkaline cleaning costs vs. vapor degreasing costs.
Mr. Kearney reported that no new technology was available to
control solvent emissions from metal cleaning operations. Free-
board chillers including the "cold trap" and carbon adsorption
represent essentially all the commercial technology available
for solvent recovery or loss control. He emphasized the
importance of good metal cleaning practices and the use of covers
as a means of conserving solvent, and noted that these techniques
are nearly always economically practical.
The ratio of freeboard height to degreaser width has been known to
effect solvent losses from solvent cleaning. Two graphs showing this
relationship we are provided. Copies are attached.
As discussed in his letter, he felt that vapor degreasing results
in less solvent emissions on a general basis than cold cleaning.
He will arrange for the use of a small open top degreaser on a
loan or consignment basis to develop information to confirm or
deny this conclusion. Mr. Kearney agreed to help us locate a
manufacturing plant where the comparative cost of alkaline cleaning
and vapor degreasing could be developed.
-------
Appendix E-10
DOW CHEMICAL U.S.A.
DEPARTMENT:
Inorganic Chemicals Department
0.
o
o
S. J. Nolan, Chicago
FIRM NAME (In full, do not obbreviote)
Phillips Manufacturing Company
STREET CITY
7334 N. Clark Street Chicago
STATE ZIP CODE
Illinois 60626
LLJ
•s.
•z.
K. Surprenant
*;*>!
OIHER COH PERSONNEI. PRESENT
A. J. Williams (Dow) ,
SUBJEC T
ORGANIC SOLVENT
METAL
FIELD DOW PHONE
X
DEVELOP. SERV. OTHER
X
DATE C A •_ L E 0
1-16-75
DATE /^RiT7EN
1-29-75
Mr. William Johnson (EPA)
CLEANING
2
QC.
tx.
J. C. Carlaw, Sarnia
D. R. Heinz, 2020
H. R. Krimbill, 2020
D. W. Richards, 2020
A. J. Williams, 2020
KOG/PF 2192008
PERSONS I M 1 E R. VI E W E D AND TITLES
Mr. Donald L. Racquet
Mr. Oskar Franz
This discussion was held to develop a further general
background in solvent metal cleaning for the Environmental
Protection Agency Contract. Further, it was our purpose
to solicit Phillips Manufacturing cooperation in locating
industrial sites which could be used to evaluate various
emission control technology. Much of the background
information on solvent metal cleaning technology and
emission control technology supported that provided
by various Dow* personnel in the initial orientation
program provided in the Dow laboratories. Phillips
agreed to help select emission control evaluation
sites.
Don Racquet of Phillips Manufacturing discussed the
wide use of "cold cleaning" with Petroleum Solvents
particularly with Stoddard solvent. No estimate
could be made of the volume of cold cleaning solvents
in use today in the U. S., however, the volume of
petroleum solvent used in this manner was estimated
to be very large. Vapor degreasing was observed to
be a more sophisticated solvent cleaning method
providing much greater quality in cleaning and some
capability of separating the solvent from the accumulated
soils. It was noted that well-designed and well-operated
-------
-2-
open top degreasers could function with losses of only
0.25-0.50 pounds per hour per square foot of open top
area. Enclosed vapor degreasers were reported to be
more economical in the use of solvent on a tonnage of
parts basis, although they often consume more solvent
on an individual equipment basis. The enclosed and
conveyorized vapor degreaser designs often solve the
problem of handling parts which would ordinarily trap
solvent and cause large solvent drag-out losses.
Recent concern by customers for preventing pollution has
resulted in an increased ratio of stills to vapor
degreasers so that the ratio is about 1:1 at present.
Phillips reported that they manufacture 200 medium or
larger vapor degreasers per year. Approximately 60%
of these are of the open top design. Four different
vapor control methods are used on their vapor degreasing
equipment. They are: 1) thermostat control; 2) direct
refrigeration; 3) a freeboard chiller (water over 32°F),
and, 4) standard condenser water coils with refrigerated
coils above them ("the cold trap"). Of these methods they
reported that "the cold trap" method is the most effective
in controlling solvent losses but causes problems with
excessive water contamination. Don Racquet also observed
that the OSHA guidelines for ventilation require excessive
ventilation rates (75 to 250 cfm) resulting in higher
solvent losses than necessary.
Both carbon adsorbtion and refrigerated coils are
employed on vapor degreasing equipment to control
solvent emissions. Phillips knew of no practical
applications of liquid absorption or incineration being
used on either petroleum solvents or halogenated solvents.
In fact, no known emission control techniques are being
employed in connection with cold cleaning operations.
Phillips Manufacturing sells six to ten carbon adsorption
units per year. They forecast that a 50% solvent recovery
can be expected and that 60% solvent recovery can be obtained
at some locations. Stabilizers would be expected to be
lost when the solvent recovery efficiency reached 80% or
greater, even in the case of trichloroethylene. It was
noted that 1,1,1-trichloroethane has stabilizer difficulty
at much lower solvent recovery rates by carbon adsorption.
Phillips indicated that their customers lean toward vapor
degreasing as opposed to alkaline washing. The cause for
this tendency was reported to be economics as well as
pollution considerations. Sources for additional
information were reported to be available in: 1) Electroplating
Engineering, published by Van Nostrum; 2) The Handbook of
Industrial Loss Prevention, published by McGraw Hill, and,
3) The Welding Handbook - Sixth Addition, Section Four,
published by The American Welding Society.
-------
TJ
TJ
m
Z
D
m
------- |