Environmental Monitoring Series
STANDARDIZATION OF STATIONARY SOURCE
METHOD FOR VINYL CHLORIDE
Environmental Monitoring and Support Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL MONITORING series.
This series describes research conducted to develop new or improved methods
and instrumentation for the identification and quantification of environmenta!
pollutants at the lowest conceivably significant concentrations. It also includes
studies to determine the ambient concentrations of pollutants in the environment
and/or the variance of pollutants as a function of time or meteorological factors.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
STANDARDIZATION OF STATIONARY SOURCE
METHOD FOR VINYL CHLORIDE
by
George W. Scheil
Midwest Research Institute
Kansas City, Missouri 64110
EPA Contract No. 68-02-1098
EPA Project Officer
M. Rodney Midgett
Quality Assurance Branch
Environmental Monitoring and Support Laboratory
Research Triangle Park
North Carolina 27711
Prepared for
Environmental Monitoring and Support Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park
North Carolina 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring and Sup-
port Laboratory, U.S. Environmental Protection Agency, and approved for pub-
lication. Approval does not signify that the contents necessarily reflect
the views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or rec-
ommendation for use.
-------
FOREWORD
Midwest Research Institute, under Task 7 of EPA Contract No. 68-02-
1098, conducted in-house work toward the standardization of Method 106 -
Determination of Vinyl Chloride From Stationary Sources. Upon completion of
the method evaluation, field tests at a polyvinyl chloride plant and a vi-
nyl chloride monomer plant were conducted under Task 9 of the contract.
Measurements of retention indices of interfering compounds were done under
Task 5 of EPA Contract No. 68-02-1780. This report contains the results of
the Method 106 laboratory evaluation and of the field tests of the method.
Approved for:
MIDWEST RESEARCH INSTITUTE
.11
c
J. Shannon, Director
Environmental and Materials
Sciences Division
August 22, 1977
iii
-------
CONTENTS
Figures vi
Tables vii
Abstract ix
Acknowledgements x
Section
I Introduction 1
1*1
II Laboratory Method Evaluation 3
Vinyl Chloride Standardization 3
Interferences. .. ......... 6
Sample Stability 10
Other Considerations .......... 13
III Field Testing 15
Test at a Vinyl Chloride Polymer Plant Incinerator . 15
Test at a Vinyl Chloride Monomer Plant 16
Appendix A - Method 106 - Determination of Vinyl Chloride From
Stationary Sources, Federal Register Procedure (Proposed) .... 25
Appendix B - Method 106 - Determination of Vinyl Chloride From
Stationary Sources, MRI Modified Procedure 28
Appendix C - Method 106 - Determination of Vinyl Chloride From
Stationary Sources 38
-------
FIGURES
Number
1 Vinyl chloride permeation assembly 7
2 Retention time determination on 2 m Chromosorb 102 plus
1 m Porapak S at 110°C 11
3 Retention time determination on 2 m Chromosorb 102 plus
1 m Porapak T at 110°C 12
4 Sample Chromatogram Bag No. 3, June 4, 1976, vinyl
chloride = 80 ppb 18
5 Bag No. 4 sample of June 8, 1976 22
6 Analysis of condensate sample from June 8, 1976, on
platform sample .............. 23
7 Temperature programmed chromatogram.showing major compo-
nents of the oxychlorination vent 24
B-l Integrated bag sampling train .............. 31
VI
-------
TABLES
Number Page
1 Permeation Tube Calibrations 4
2 Retention Indices for Possible Vinyl Chloride Interfer-
ences 8
3 Test Results - Vinyl Chloride Incinerator 17
4 Vinyl Chloride Analysis Results of Sampling at Shell
Chemical Company 20
B-l Retention Indices for Possible Vinyl Chloride Interfer-
ences 30
vii
-------
ABSTRACT
Method 106 - Determination of Vinyl Chloride From Stationary Sources -
was evaluated for possible problems with sample stability, interferences and
methods of standardization. The method was then field tested at a polymer
plant incinerator and a monomer plant scrubber. Vinyl chloride field samples
were found to be stable in Tedlar bags for periods of greater than 1 week.
Laboratory tests revealed that aluminized Mylar bags also worked well if the
samples were analyzed within 24 to 48 hr. Several compounds can interfere
with the gas chromatographic analysis. Acetaldehyde and ethylene oxide are
the primary inteferences. Under some conditions methanol and isobutane also
interfere. Retention indices measured on the three most promising columns
indicate that most samples can be successfully resolved. Pressurized gas cyl-
inders of vinyl chloride in nitrogen were found to have excellent long-term
stability, and a gravimetrically calibrated vinyl chloride permeation tube
is an excellent primary standard. The field tests indicated no serious prob-
lems with the procedure and demonstrated that vinyl chloride had excellent
stability in the presence of HCl, C1-, H«0, and other reactive compounds.
This report was submitted in fulfillment of Tasks 7 and 9 of Contract
No. 68-02-1098 and in partial fulfillment of Task 5 of Contract No. 68-02-
1780 by Midwest Research Institute under the sponsorship of the U.S. Envi-
ronmental Protection Agency. This report covers a period from January 1975
to December 1976, and work was completed as of January 1977.
ix
-------
ACKNOWLEDGEMENTS
This work was conducted under the technical management of Mr. Paul C.
Constant, Jr., Head, Environmental Measurements Section of Midwest Research
Institute's Environmental and Materials Sciences Division, who is the pro- .
gram manager. Dr. George Scheil was task leader. He was assisted by Mr. George
Cobb of Midwest Research Institute. The assistance of Mr. William Roberts of
the Shell Chemical Company during the field test is gratefully acknowledged.
-------
SECTION I
INTRODUCTION
On December 24, 1975, under Section 112 of the Clean Air Act as amended,
the Environmental Protection Agency (EPA) added vinyl chloride to the list
of hazardous air pollutants— because it has been implicated as the causal
agent of angiosarcoma and other serious disorders. A national emission stan-
dard has been promulgated— that covers plants that manufacture ethylene di-
chloride, vinyl chloride and polyvinyl chloride (PVC). These regulations in-
clude a method for determining vinyl chloride emissions from stationary sources,
EPA Method 106. The Quality Assurance Branch of the Environmental Monitoring
and Support Laboratory at Research Triangle Park, North Carolina, has as its
task the evaluation and standardization of EPA source test methods. While par-
ticipating in this program, Midwest Research Institute (MRI) has undertaken
a study of EPA Method 106.
This report covers the laboratory evaluation and field testing conducted
by MRI of Method 106 - Determination of Vinyl Chloride from Stationary Sources.
The testing was done for EPA under Contract No. 68-02-1098, "Standardization
of Stationary Source Emission Measurement Methods." This report includes Task
7, Laboratory Evaluation, and Task 9, Field Testing. Additional laboratory
testing was done under Task 5 of EPA Contract No. 68-02-1780.
Method 106 is intended for use in both vinyl chloride monomer plants
and PVC plants. When testing the exit of an incinerator-type control device,
the level of organics is low and generally consists of only a few compounds.
Other control devices or ventilator ducts can contain a very complex mixture
of compounds with many constituents being at concentrations equal to or
greater than the vinyl chloride concentration. Thus, the primary problem in
analysis of vinyl chloride can often be the satisfactory resolution of vinyl
chloride from other possible constituents.
I/ Federal Register, 40, 59477, December 24, 1975.
2/ Federal Register, 41, 46564-46573, October 21, 1976.
-------
The following sections of this report examine the laboratory evaluation
of Method 106, including sample stability, interferences, methods of stan-
dardization, and finally the results of field tests of the method at a poly-
mer plant incinerator and a monomer plant scrubber.
-------
SECTION II
LABORATORY METHOD EVALUATION
The originally proposed Federal Register procedure is given in Appen-
dix A. Three possible problem areas were examined in some detail: a reli-
able standardization procedure, satisfactory separation of interferences,
and the stability of vinyl chloride samples. These and other problems are
discussed in detail in the following sections.
VINYL CHLORIDE STANDARDIZATION
Because of the strict safety regulations necessary for handling pure
vinyl chloride, the primary calibration procedure cannot be used in many
situations. The alternate procedure using gas cylinders containing 15 to
150 mg/m^* (5 to 50 ppm) vinyl chloride mixtures in nitrogen can be used
with less stringent precautions. However, the accuracy and stability of
calibration gases can be poor if sufficient care is not taken during their
preparation. A reliable method of verifying the cylinder concentrations is
then necessary. The dilution of pure vinyl chloride in a calibrated dynamic
dilution system is possible, as well as the injection of vinyl chloride
into a Tedlar bag as given in the method. A vinyl chloride permeation tube
assembly allows a reliable, reasonably safe calibration and was used as the
primary calibration standard in all project work. With a standard water-
heating bath supplying water at 30.0 + 0.1°C to the water jacket of a
straight-tube condenser, a simple permeation tube assembly was constructed.
The permeation rate through the Teflon walls of the permeation tube is de-
pendent almost solely on temperature. Carrier gas flow rate or most other
variables have no effect on permeation rate as long as liquid vinyl chlo-
ride remains in the tube. By periodically weighing the tubes on an analyt-
ical balance the rate of release can be calculated. Table 1 summarizes the
calibration data for three sets of two 100 mm (1.0 mm wall) permeation
tubes. While the apparent weight loss over a few days can vary widely, due
* Specified at conditions of 293°K and 101.3 kPa (760 mm Hg). To convert
from parts per million (v/v) to milligrams per cubic meter at this
temperature and pressure, multiply by 2.59.
-------
TABLE 1. PERMEATION TUBE CALIBRATIONS
Ue/min of Vinvl chloride
Date
3-28-75
4-1-75
4-4-75
4-8-75
4-9-75
4-11-75
4-15-75
4-18-75
4-21-75
4-25-75
4-30-75
5-2-75
5-6-75
5-9-75
5-13-75
5-16-75
5-27-75
6-3-75
6-6-75
6-10-75
6-13-75
6-25-75
7-2-75
7-11-75
7-18-75
7-29-75
9-9-75
9-18-75
Time
1645
1600
1625
1645
1635
1635
1615
1615
1505
1030
1635
1615
1635
1640
1500
1530
1510
1635
0825
1440
1515
0950
1020
0900
0830
1000
1100
1150
Weight loss (me)
Tube Tube
No. 1 No. 2
._
12.52
15.28
13.57
4.16
6.87
16.52
11.06
12.23
14.27
24.76
4.85
16.05
11.19
16.18
15.12
40.49
28.93
10.62
17.10
12.18
47.31
28.19
36.46
27.03
44.58
170.44
34.32
16.64
13.15
15.87
2.22
8.51
15.70
12.08
14.45
13.28
22.56
8.08
16.68
12.00
16.26
13.92
46.19
27.45
10.88
17.87
12.71
48.83
29.46
36.95
29.30
45.91
175.49
36.76
Tube No
Since previous
weighing
2.19
3.52
2.35
2.94
2.38
2.88
2.55
2.88
2.60
3.27
1.70
2.77
2.59
2.86
3.46
2.56
2.85
2.77
2.79
2.79
2.79
2.79
2.83
2.69
2.80
2.82
2.65
. 1
Cumulative
2.19
2.76
2.61
2.66
2.58
2.65
2.63
2.67
2.66
2.76
2.70
2.70
2.70
2.71
2.76
2.72
2.73
2.74
2.74
2.74
2.75
2.75
2.76
2.75
2.76
2.77
2.77
Tube No. 2
Since previous
weighing
2.91
3.03
2.75
1.55
2.96
2.74
2.80
3.40
2.42
2.98
2.83
2.89
2.77
2.87
3.20
2.92
2.70
2.84
2.91
2.92
2.88
2.91
2.87
2.92
2.88
2.90
2.84
Cumulative
2.91
2.96
2.88
2.75
2.89
2.86
2.85
2.92
2.85
2.87
2.87
2.87
2.86
2.86
2.88
2.89
2.87
2.87
2.87
2.87
2.87
2.88
2.88
2.88
2.88
2.88
2.88
(continued)
-------
TABLE 1 (continued)
Ue/min of Vinvl chloride
Date
New Set of
10-9-75
1030-75
11-10-75
11-25-75
12-18-75
New Set of
6-18-76
6-28-76
10-8-76
Time
Tubes
1640
0835
1510
1630
1425
Tubes
1025
1010
1450
Weieht
Tube
No. 1
« •
81.53
46.06
59.46
88.81
__
36.45
379.62
loss (me)
Tube No. 1
Tube Since previous
No. 2 weighing
—••
83.55
47.17
61.46
91.37
«••
35.99
370.28
••••
2.74
2.84
2.74
2.69
_„
2.53
2.58
Cumulative
— ^
2.74
2.77
2.76
2.74
_M
2.53
2.58
Tube
Since previous
weiehine
• ••
2.81
2.91
2.83
2.77
«*••
2.50
2.52
No. 2
Cumulative
^ —
2.81
2.84
2.84
2.82
«•
2.50
2.51
-------
to dust on the tubes or the test tubes used to hold them during weighing,
the cumulative permeation rate remains constant for several months within
1 to 2%.
The calibration assembly is shown in Figure 1. The optional dilution
gas was normally kept plugged off and was used only for low levels of vinyl
chloride when the main rotameter could not supply enough nitrogen. A Singer
charcoal test meter was then used to measure the additional dilution nitro-
gen.
The permeation tubes were supplied by two companies: Metronics Asso-
ciates, Inc., Palo Alto, California, and AID, Inc., Avondale, Pennsylvania.
The AID tubes are slightly larger in diameter with a longer life span but
7 of 12 of the tubes (one batch of four were good; a second set of eight
had one good tube) had contamination present and polymerized within a few
days. No problems were experienced with the set of four tubes from Metronics.
A 3.2 mm (1/8 in.) O.D. Teflon tube was used to withdraw gas from the per-
meation assembly and the gas was pulled through the gas chromatograph sam-
pling valve with a small vacuum pump at 20 to 50 ml/min.
Three gas cylinders supplied by Scott Environmental Technology, Inc.,
Plumsteadvilie, Pennsylvania, were checked against the permeation tubes.
On three separate days the 12.9 mg/m^ cylinder gave 12.3, 12.4, and 12.5
mg/nP vinyl chloride. The average, 12.4 rag/'nr, was assigned as the true
value. The 25.6 and 117 mg/m cylinders were checked once and showed 23.5
and 108 mg/m^, respectively. All results are the averages of at least three
measurements. No nonlinear!ty was found in response from 1 to 120 mg/m^.
All readings were by peak height, which was used in the original Method 106
procedure.
A permeation tube comparison is an excellent primary calibration for
gas cylinders. Although the permeation assembly is not portable, and some
tubes polymerize, the gravimetric calibration of the tubes is an absolute
reference and the permeation rate is constant over the life of the tube.
The calibration is necessary only for new cylinders with a check every 6
months or so after the initial calibration.
INTERFERENCES
Table 2 lists several compounds which might be present in vinyl chlo-
ride process streams together with their retention indices on a Chromosorb
102 column, 2 m x 3.2 mm (1/8 in.) O.D.* stainless steel at 100°C and on
two modified columns. The only serious interferences are acetaldehyde and
ethylene oxide which are usually not resolvable from vinyl chloride on a
* All chromatographic columns had a wall thickness of ~ 0.5 mm (0.020 in.).
-------
Purified Nitrogen
Cylinder Gas
Constant
Temperarure
Water Bath
30.0°C +0.1°
Calibrated
Rotometer
Water Jacketed Condenser
Permeation
Tubes
Optional Dilution
Nitrogen
Figure 1. Vinyl chloride permeation assembly
-------
TABLE 2. RETENTION INDICES FOR POSSIBLE VINYL CHLORIDE INTERFERENCES
00
2 m Chromosorb 102
100° C
2 m Chromosorb 102
+ 1 m Porapak T
120°C
2 m Chromosorb 102
+ 2 m SF-96
120°C
2 m Chromosorb 102
+ new 2 m SF-96
120° C
Methane*
Ethylene
Ethane*
Propane*
Methyl chloride
Methanol
100
180
200
300
320
330
100
180
200
300
340
395
100
175
200
300
320
350
100
-
200
300
325
375
Acetaldehyde
E thy le ne ox id e
Vinyl chloride
Isobutane
Isobutylene
1-Butene
n-Butane*
1,3-Butadiene
trans-2 -Butene
Ethanol
Ethyl chloride
cis-2-Butene
1,1-Dichloroethylene
trans-1,2-Dichloroethylene
355
355
360
380
395
395
400
400
400
415
415
415
480t
400
395
375
380
395
395
400
410
410
475
430
415
4901"
4951"
375
375
360
380
385
395
400
395
400
435
415
410
4801"
(low levels retained)
400
(low levels retained)
(low levels retained)
360
380
390
400
395
410
*
t
Reference compounds for indices.
Column at 150°C.
-------
2 m column. Acetaldehyde can be present in significant amounts— in process
streams. (Ethylene oxide appears to be converted to acetaldehyde on the OC
columns.) The use of a precolumn containing sodium bisulfite has been sug-
gested to remove acetaldehyde.— However, in MRI's tests the removal was
not always complete and the precolumn can be inactivated by many other com-
pounds so that the removal efficiency can change suddenly during a series
of analyses.
No single column is likely to resolve all possible interferences. How-
ever, since no known interference with a retention time identical to vinyl
chloride at 100°C on Chromosorb 102 has been found, the presence of unre-
solved interference peaks can be detected by a comparison of the quantita-
tive results by peak height and by peak area. If the results by both meth-
ods are the same, then the peak is probably pure vinyl chloride. If peak
height has a lower result than peak area, an interference is present and a
second column should be added after the Chromosorb 102. The second column
takes advantage of the separation already obtained on the Chromosorb 102
column by separating the particular interfering compound(s) without obtain-
ing a significant shift of the other peaks in the chromatogram. A 2 m x 3.2
mm (1/8 in. ) 0»D. stainless steel column packed with 20% SF-96 on acid-
washed 60/80 mesh Chromosorb P has been found to retard acetaldehyde, which
elutes as a broad peak immediately following vinyl chloride. This combina-
tion column was used during most of MRI's field tests of the method. How-
ever, resolution of methanol and vinyl chloride is not good. Since methanol
tends to elute as a severely tailing peak it will usually cause only a
change in baseline. Three percent columns of other similar— columns (OV-1,
OV-101) do not provide enough separation at the 100°C operating temperature
of Chromosorb 102.
Present evidence indicates that the separation depends partially upon
the high loading of liquid phase on the column and upon unique differences
between SF-96 and other nonpolar liquid phases. Although Chromosorb P is
not a high performance support, it does allow heavier loadings than newer
supports, such as Chromosorb W or Chromosorb 750. The high loading allows
reasonable separation at 100°C, saturates the active sites of the support,
and improves the reproducibility of results from one column to another.
The last column in Table 2 shows selected retention indices measured
using a newly prepared and conditioned SF-96 column. The only compounds
which show significant changes are acetaldehyde, ethylene oxide, and the
alcohols. All of these compounds are strongly adsorbed by the column and
elute as observable peaks only at concentrations above about 0.1%. At lower
levels they will probably not be eluted as peaks but only as gradual base-
line changes, unless the column has become partially deactivated by long
3_/ Krishen, A., and R. Tucker, Anal* Chem., 48_, 455-456 (1976).
4/ Supina, W., The Packed Column in Gas Chromatography, Supelco, Inc.
(1974).
-------
use, as in the case of the older SF-96 column results. If large amounts of
these compounds, especially methanol, are present, the use of a different
column would be preferable. However, aim section of Porapak S followed by
1 m of Porapak T has been shown to achieve good separation^/ although iso-
butane may be shifted enough to interfere.
Further tests by MRI have revealed that Porapak S will not separate
vinyl chloride and acetaldehyde (see Figure 2). A 1 m x 3.2 mm (1/8 in.)
0»D. stainless steel column packed with 80/100 mesh Porapak T added after
the Chromosorb 102 column does obtain a clean separation of vinyl chloride
and acetaldehyde (Figure 3). However, isobutane is now very difficult to
resolve from vinyl chloride. Operating the combined columns at 120°C should
maintain good resolution with an analysis time comparable to the Chromosorb
102 column alone at 100°C.
The suggested analysis procedure is to compare results by height and
area on the Chromosorb 102 column. If the height result is lower, then a
second column (SF-96 or Porapak T) is added after the Chromosorb 102 column
and chromatographic conditions adjusted until the interfering peak is re-
solved, height and area results agree, and all other nearby peaks on the
initial column can be definitely accounted for in the modified chromatogram.
Table 2 should be used to choose the proper secondary column.
The best method of confirming purity of a suspect peak is mass spec-
trometry, a procedure which is not suitable for routine analysis. An alter-
nate possibility is an electron capture detector which has a strong response
to vinyl chloride, very little response to acetaldehyde and hydrocarbons.
This method is very carrier-flow rate sensitive and can be used only under
stable laboratory conditions.
SAMPLE STABILITY
As indicated in Section III of this report, vinyl chloride samples in
Tedlar bags have been found to decay at a rate of less than 10%/month even
in complex samples. Teflon bags have no economic advantage over Tedlar and
are not good vapor barriers, thus Teflon sample bags are not suitable for
vinyl chloride samples. Aluminized Mylar has been suggested as an alterna-
tive.—' In a test of Tedlar and aluminized Mylar bags at the level of 15
mg/m vinyl chloride, both materials showed less than 5% loss of vinyl
chloride over a 2-week period. However, at the end of that period, outgas-
sing from the Mylar resulted in several peaks of very long retention time
5_/ Mr. William Roberts, Shell Chemical Company, Houston, Texas, Private
Communication.
6/ Levine, S. P., K. G. Hebel, J. Bolton, Jr., and R. E. Kugel, Anal. Chem.,
47, 1075A-1080A (1975).
10
-------
Acetaldehyde
Rj =6.4 min.
Vinyl Choride
& Acetaldehyde
R-r = 6.5 min.
6 80
Time from Injection (min.)
Figure 2. Retention time determination on 2 m Chromosorb 102
plus 1 m Porapak S at 110°C
11
-------
to
Attenuation
Change
Acetaldehyde
Rj = 10.2 min
v-
Vinyl Chloride
Rj = 8.1 min.
Acetaldehyde
Ry = 10.2 min.
10
Time from Injection (min.)
8
10
Figure 3. Retention time determination on 2 m Ghromosorb 102 plus 1 m Porapak T at 110° C
-------
(•«' 1 hr at lOCPc). Unless the chromatograph can be backf lushed or programmed
after elution of vinyl chloride, aluminized Mylar bag samples should always
be analyzed quickly. Mylar bags are about half as expensive as Tedlar but
are commercially available from only one known source, Calibrated Instruments,
Inc., Ardsley, New York. The Mylar, aluminum foil, vinyl outer film laminate
used is, however, an excellent vapor barrier and is more flexible and dura-
ble than Tedlar.
While Tedlar should remain as the primary bag material, the use of alu-
minized Mylar is also acceptable if the analyst is aware of the limitations
of aluminized Mylar.
OTHER CONSIDERATIONS
Several other changes have been made in the version of Method 106 which
was developed by MRI based on the results of this study. This version is
given in Appendix B. The column temperature, 155°C, of the proposed method
is too high, from the standpoint of unnecessary column bleed and low resol-
ution due to a short retention time. Temperatures of 100 to 110°C have been
used in MRI's work and provide a reasonable retention time of about 4 to 5
min at 25 ml/min with very low bleed (and noise). The Varian 2440 chromato-
< ^
graph used has a detection limit of « 100 (ig/m (x2 noise) when operating
at 100°C with a 2 ml sample loop volume.
Additional leak checks have also been added to improve the reliability
of the sampling bags and enclosures. An optional glass condensate trap has
been added for the significant fraction of sources with condensibles pres-
ent. The simple vacuum traps have proved reasonably rugged and very efficient
in reducing the sample dew point without causing undue absorption of vinyl
chloride.
A more significant change is the placement of the rotameter (made of
glass or stainless steel for sample compatibility) between the sample bag
and the probe. If the rotameter is made of glass and stainless steel, sam-
ple contamination or reaction is improbable and the condenser protects the
rotameter from condensation. But the change in position of the rotameter
significantly increases the probability of obtaining a representative sam-
ple. Even with the added leak checks, many bags have failed to fill or
filled with ambient air during MRI's tests with this and derivative methods
when the rotameter is after the pump. The pump pulsations cause inaccurate
readings, but more serious is the fact that a slight leak in the bag, en-
closure, pump or connecting tubing and fittings can cause a properly run
test to fail because a very small leak will allow in enough ambient air to
greatly reduce the sample flow rate into the sample bag. With the rotameter
before the sample bag only a leak in the sections ahead of the rotameter
will foul a test. A leak anywhere else will be more obvious, and may not
even stop the test if it is reasonably small and constant and is in the en-
closure or pump, since then the bag will still fill with a slightly higher
13
-------
pumping, rate. Most important, the rotameter can now show that gas is going
into the bag (although the bag can still leak) and not merely that gas is
passing out of the pump. The relocation of the rotameter reduces by at least
50% the number of fatal leak points, a very important factor in a system
which must operate at subambient pressures.
14
-------
SECTION III
FIELD TESTING
Two sources were chosen to field test the method. One test was at a
vinyl chloride polymer plant which used an incinerator/scrubber to destroy
vinyl chloride. The second test was at Shell Chemical Company, Deer Park,
Texas. The stack from their vinyl chloride monomer plant oxychlorination
vent was sampled. This stream was very complex, containing significant
amounts of ethylene, chlorine, hydrogen chloride, acetaldehyde, methyl chlo-
ride, ethyl chloride, dich1oroethane, etc. The presence of large amounts
of high-boiling compounds required the use of a backflush valve. All equip-
ment for these tests was loaded into a van and driven to the sites where
the gas chromatograph was then set up inside the van. Both tests were made
in a single trip. Due to the safety problems and difficult access, stack
flow rate measurements were not made for either test. To remove some of the
acid gases possibly present in both streams, a section of tubing packed
with granular potassium carbonate was used to remove Cl£ and HC1. A labor-
atory test showed that the potassium carbonate had no effect on vinyl chlo-
ride levels.
TEST AT A VINYL CHLORIDE POLYMER PLANT INCINERATOR
The unit tested was an incinerator-scrubber unit which handled process
off-gases from a PVC operation. The gases are stored in holding tanks and
then pumped to the incinerator whenever the tanks are full. The incinerator
is then started and continues to operate until the holding tanks are empty.
The operating time can vary from 1/2 hr to several hours depending on the
level of activity in the polymer operation. The gas-fired incinerator is
followed by an alkali scrubber which vents through a stack. The stack is
mounted on top of the scrubber and has an exit height of about 15 m. Two
150 mm (6 in.) flanges are mounted to the stack at a height of 10.5 m above
ground level.
A bored-through 1/4 in. NPT to 1/4 in. Swagelok male connector was
mounted in one of the flanges and a section of 6.4 mm (1/4 in.) 0,D«,
Teflon-lined stainless steel tubing was slipped through the connector so
that the end of the tubing was at the center of the 300 mm diameter stack.
15
-------
The probe tip was packed with glass wool before being inserted. The sample
line was then connected to the probe tubing with an elbow fitting. The sam-
ple line consisted of a 15 m section of 6.4 mm (1/4 in.) O.D. Teflon-lined
stainless steel tubing inside a sheath of 12.7 mm (1/2 in) O.D. copper tub-
ing. A steam line was connected to the outer tube by a 12.7 mm (1/2 in.)
tee at ground level and the probe end of the sheath was left open so that a
continuous flow of steam passed through the sheath to prevent condensation
in the sample line. The sample probe and line had to be installed with a
"cherry picker" crane, since no ladder or other access was available on the
small diameter stack. The bottom end of the sample line was connected to
the center tube of a 20mmx 2 00 mm glass vacuum trap which was immersed in
an ice bath. A 150 mm section of 6.4 mm (1/4 in.) O.D, stainless steel tub-
ing packed with granular potassium carbonate followed the condenser, and
the sample stream was then split three ways. Two integrated bag samples were
obtained in parallel, and on the 2nd day direct GC sampling was made on the
sample stream. The remaining parts of the trains were as given in the pro-
cedure given in Appendix B.
The equipment was set up on June 3, 1976, and a pair of samples was
obtained that day and analyzed on the following day. The Chromosorb 102
column was used for all tests and operated at 110°C.
One bag on June 4 was found to be contaminated, but the other three
for that day, one set of bag samples the 1st day, and the direct QC sampling
were consistent with a vinyl chloride concentration of 0.25 +0.1 mg/m^.
Table 3 shows the results of the tests. Figure 4 is a sample chromatogram.
Due to the very small quantities of vinyl chloride present; peak height was
used instead of area. Approximately 5 ml-of water condensate was obtained
from a total gas volume of 200 liters. After returning to MRI the condensate
was analyzed by QC and no organic matter was detected in the water. One in-
tegrated gas bag sample was returned to MRI for stability measurements.
The sample contains no acetaldehyde or other known interferences. The
condensate sample was found to contain 0.0037 M chloride.
TEST AT A VINYL CHLORIDE MONOMER PLANT
On June 8 through 10, further field testing was done on a vinyl chlo-
ride monomer oxychlorination vent at the Shell Chemical Company in Deer Park,
Texas. The sampling platform on this stack is approximately 15 m above the
ground. The sample line was connected to the stack through a 2.54 cm (1.0
in.) gate valve. A tee connection at the platform allowed sampling at the
platform as well as at ground level via the heated sample line. The sample
line was connected to the stack in a manner similar to that used on the first
test except that the truck containing the gas chromatograph could not be
parked close enough to the stack to allow direct QC sampling. One pair of
samples was taken at ground level. Three pairs of samples were obtained with
one sample taken at ground level and one at the sampling platform. The analysis
16
-------
TABLE 3. TEST RESULTS - VINYL CHLORIDE INCINERATOR
Sample
6-3-76 analyzed on 6-4-76
23.5 mg/m-* standard
Bag No. 1
Bag No. 3
6-4-76
Direct sampling
Direct sampling
Direct sampling
Direct sampling
Direct sampling
Direct sampling
Direct sampling
Direct sampling
Direct sampling
Bag No . 4
Bag No. 2
23.5 mg/m^ standard
Bag No. 1
Bag No. 3
Reana lysis of Bag No. 2 from
6-4-76 on 6-24-76
12.4 mg/nP standard
Reanalysis of Bag No. 2 from
6-4-76 on 7-27-76
12 .4 mg/nP standard
Time
1410-1500
1410-1500
0955
1000
1007
1014
1020
1027
1034
1041
1048
0954-1054
0954-1054
1225-1310
1225-1310
Attenuation
4
1
1
2
2
2
2
2
2
2
2
2
1
1
4
1
1
1
4
1
4
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
-12
-12
-12
-12
-12
-12
-12
-12
-12
-12
-12
-12
-12
-12
-12
-12
-12
-12
-12
-12
-12
76
4.
3.
1.
1.
1.
1.
1.
1.
1.
0.
0.
5.
2.
66
3,
2.
2.
46
5.
47
Vinyl chloride
Peak heights concentration (mg/m-^)
.5
o,
2,
0
0
0
0
0
0
0
8
8
o,
5,
.4
5,
5,
5,
= 8
0,
.5
, 74.7, 75.3
2.7
3.8
4.5
2.0
, 67.6, 66.8
3.2, 3,2
2.5, 2.5
3.0
, 46.6
4.5
, 47.0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.28
.28
.16
.16
.16
.16
.16
.16
.16
.13
.13
.41*
.18
,28
.21
.17
.28
* Artifact peaks present.
-------
Vinyl Chloride
0
Time from Injection (min.)
Figure 4. Sample Chromatogram Bag No. 3, June 4, 1976,
vinyl chloride =80 ppb
18
-------
results are given in Table 4. Sample chromatograms are shown in Figure 5,
which compares the different columns and the effect of a sodium bisulfite
scrubber.
Condensate samples were analyzed for vinyl chloride after returning to
MRI. The levels varied from ~ 3 |o,g/g to less than 0.5 (J»g/g. Approximately
six times as much acetaldehyde was present and varying amounts of other or-
ganic compounds were also present, primarily alcohols, judging by the odor
given off by the samples. A sample chromatogram is shown in Figure 6. Chlo-
ride analyses were made on these condensate samples and the results are
shown in Table 4. The volume of condensate was always less than 1 ml. The
sampling was done from an existing tap about 15 m above ground. The connec-
tion was to a gate valve on a 1 in. NPT pipe which allowed the sample probe
to be inserted to the stack center. Stack diameter was about 3 m at the sam-
ple point, midway through a tapered section of the stack.
Sample backflush to vent was used to remove several slow eluting com-
pounds from the samples. The backflush valve was thrown after elution of
the vinyl chloride peak and was flushed for a slightly longer time than the
time from sample injection to when backflushing began. A Carle double sam-
pling loop valve was maintained at 97°C for all analyses. Matched sample
loops were used with volumes of 2 ml.
An attempt was made to split the column effluent for parallel FID and
electron capture detection. However, the electron capture standing current
could not be stabilized due to the backflushing and no usable data were ob-
tained by electron capture.
Figure 7 is a sample chromatogram obtained by Shell Chemical on a log-
arithmic response recorder, programming from 60 to 190°C. The known compo-
nents of the sample are identified. The column used was a 3.2 mm (1/8 in.)
O.D. x 3 m Chromosorb 102. Shell used these conditions for all of their
analyses. They used a midget impinger containing a 5% solution of sodium bi-
sulfite to remove acetaldehyde. They also obtained their samples in small
evacuated propane cylinders with a needle orifice to fill the cylinder at a
constant rate.
The results of the laboratory and field tests have been incorporated
into the suggested procedure given in Appendix B. The proof of a satisfac-
tory separation of vinyl chloride and acetaldehyde removes the last major
problem in the analysis. One of the columns suggested should be able to re-
solve most of the mixtures which occur in vinyl chloride monomer and poly-
mer plants.
Method 106 was published in the Federal Register, Vol. 41, Thursday,
October 21, 1976, pp. 46569-46571, before the final version of this report
was completed. A copy of the procedure is given in Appendix C.
19
-------
TABLE 4. VINYL CHLORIDE ANALYSIS RESULTS OF SAMPLING AT SHELL CHEMICAL COMPANY
(all results reported as vinyl chloride, me/m3)
NJ
O
Date/time
6-8-76
1030-1130
1030-1130
6-9-76
1000-1100
1000-1100
6-10-76
0930-1030
0930-1030
1325-1425
1325-1425
Reana lysis of 6-24-76
Reana lysis of 7-27-76
Sample
108
Bag
Bag
108
Bag
Bag
108
108
Bag
Bag
Bag
Bag
108
Bag
Bag
108
Bag
108
Bag
Bag
Bag
108
mg/m3
No. 3
No. 4
mg/m3
No. 1
No. 3
mg/m3
mg/m3
No. 3
No. 1
No. 3
No. 1
mg/m
No. 1
No. 3
mg/m3
No. 4
mg/m
No. 3
No. 4
No. 1
mg/m3
standard
on ground
on platformt
standard
on platform
on ground
standard
standard
on platform
on ground
on ground t
on groundt
standard
standard
standard
standard
Attenuation
factor
16 x
32 x
32 x
16 x
64 x
64 x
16 x
8 x
8 x
8 x
16 x
16 x
8 x
8 x
8 x
4 x
8 x
4 x
8 x
4 x
8 x
4 x
io-12
10" 12
io-12
io-12
io-12
10- 12
io-12
io-11
lO'11
io-11
io-11
io-11
io-11
10"11
io-11
io-11
io-11
lO'11
lO'11
io-11
io-11
io-11
Peak areas
820, 813
1,050, 1,047
667, 647
-
514
652, 647
867, 845
1,884, 1,849
No organlcs
No organlcs
2,165, 2,245, 2,155
2,190, 2,175
1,895, 1,845
3,315
3,280, 3,340
2,950, 2,970, 2,995
1,325, 1,310
3,015, 3,030
3,115, 3,165
2,470, 2,550
2,750, 2,790
3,065, 3,085
MRI vinyl
chloride result
by area Peak height
.
285
176
-
264
334
_
_
found*
found*
254
256
-
243
243
_
98
_
225
91
199
-
66.9,
60.4,
37.6,
84.5,
38.9,
47.1,
76.9,
47.0,
55.0,
53.0,
45.0,
67
59
36
84
38
46
76
47
_
.
55
52
44
.
_
.
_
_
_
_
_
-
.2
.8
.8
.5, 84.5
.0, 37.5
.8
.4
.2
.9, 53.0
.5
.8
MRI result
by height
.
197
122
.
205
256
_
-
_
.
262
254
-
.
_
_
_
.
_
_
_
-
(continued)
-------
TABLE 4 (continued)
Analyses with sodium bisulfite trap
for acetaldehyde removal
Date /time
6-8-76
1030-1130
1030-1130
6-9-76
1000-1100
1000-1100
6^10-76
0930-1030
0930-1030
1325-1425
1325-1425
Reanalysls of 6-24-76
Reana lysis of 7-27-76
Sample
108 mg/m3 standard
Bag No. 3 on ground
Bag No. 4 on platformt
108 mg/in3 standard
Bag No. 1 on platform
Bag No. 3 on ground
108 mg/m3 standard
108 mg/m3 standard
Bag No. 3 on platform
Bag No. 1 on ground
Bag No. 3 on groundt
Bag No . 1 on groundt
108 mg/m3 standard
Bag No. 1
Bag No. 3
108 mg/m3 standard
Bag No. 4
108 mg/m3 standard
Bag No. 3
Bag No. 4
Bag No. 1
108 mg/m3 standard
MRI result
Peak areas by area
_
976 262
541, 536 145
-
469, 500 251
629, 643 326
-
-
-
-
-
•
-
-
-
-
-
-
-
-
-
-
Peak MRI result
height by height
_ _
50.4 166
30.8, 30.6 101
-
34.5, 36.3 194
45.5, 46.2 251
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
- -
Plant process
analyzer
.
259
259
-
251
251
-
-
-
-
316
316
-
-
-
-
-
-
-
-
-
-
Shell lab
result**
_
264 (189)
-
-
355 (262)
275 (251)
-
-
Ott
ott
329 (280)
-
-
-
-
-
-
-
-
-
-
-
Condensate
chloride
concentration
.
-
0.005 M
-
0.0014 M
-
-
-
0.0011 M
0.0046 M
0.0017 M
-
-
-
-
-
-
-
-
-
-
-
tt
A 2 m x 3.2 mm (1/8 in.) O.D. Chromosorb 102 column was used on 6-8-76, and a 150-mi section of 6.4 tun (1/4 In.) O.D. stainless steel
tubing packed with sodium bisulfite was used for acetaldehyde removal.
A 25 cm x 3.2 mm (1/8 in.) O.D. OV-101 column was added after the 2 m Chromosorb 102 on 6-9-76. Acetaldehyde still Interfered and was
removed with sodium bisulfite.
A 2 m x 3.2 mm (1/8 in.) O.D. Chromosorb 102 column followed by a 2 m x 3.2 mm (1/8 In.) O.D. SF-96 column was used at 110°C on 6-10-76
and all reanalyses. Acetaldehyde eluted approximately 1 min after vinyl chloride on this column as a broad tailing peak.
A test for oxygen content showed that a small leak was present somewhere in the line between the stack sample bag and (probably) the
K2C03 filter. The difference in oxygen showed about 207. air infiltration on the platform samples.
Sample saved for reanalysis.
Leak in the sample line. Oxygen content 217..
Values in parentheses indicate results after acetaldehyde was removed by treatment with sodium bisulfite.
Shell lab analysis of MRI bag sample.
-------
ho
ISi
Vinyl Chloride
& Acetaldehyde
Vinyl Chloride
& Acetaldehyde
Vinyl Chloride
I
I
_L
Acetaldehyde
Attenuation 32 x KT12
0 24
Time from Injection (min.)
Attenuation 32 x 10"12
Attenuation 8 x 10-"
Figure 5. Bag No. 4 sample of June 8, 1976. First run on Chromosorb 102 column; second run,
same column with sodium bisulfite. Last run on Chromosorb 102/SF-96 column June 24, 1976
-------
I
0246
Time from Injection (min.)
Figure 6. Analysis of condensate sample from June 8, 1976, on
platform sample. Chromatogram run on 2 m Chromosorb 102/
2 m SF-96 column at 110°C
23
-------
Figure 7. Temperature programmed chromatogram showing major components
of the oxychlorination vent. Courtesy of Shell Chemical
24
-------
APPENDIX A
METHOD 106 - DETERMINATION OF VINYL CHLORIDE FROM STATIONARY SOURCES
FEDERAL REGISTER PROCEDURE (PROPOSED)
25
-------
PROPOSED RULES ,
METHOD 100—DETERMINATION OP VINYL
CHLORIDE FROM STATIONAKY SOURCES
INTRODUCTION
Performance of this method should not bo
attempted by. persons unfamiliar with the
operation of a gas chromntograph, nor by
those who are unfamiliar with source sam-
pling, as there nre many details that are
beyond the scope of this presentation. Care
must be exercised to prevent exposure of
sampling personnel to vinyl chloride', a car-
cinogen.
1. Principle and Applicability.
1.1 An Integrated bug sample of stack gas
containing vinyl chloride (chloroethylene)
Is subjected to chromatographlc analysis,
using a flame lonlnatlon detector.
1.2 The method Is applicable to the meas-
urement of vinyl chloride In stack gases from
both vinyl chloride and polyvlnyl chloride
manufacturing processes, except where the
vinyl chloride Is contained In paniculate
matter.
2. Range and Sensitivity.
The lower limit of detection will vary ac-
cording to the chromatograph used. Values
reported Include 1 X 10-T mg and 4 X 10-'
mg.
3. Interferences.
In the course of a study to Identify the
Interference potential of several hydrocar-
bons associated with vinyl chloride, none
were found to prevent resolution of the vinyl
chloride peak with the Chromosorb 102'
column. However, if resolution of the vinyl
chloride peak is not satisfactory for a par-
ticular sample, then chromatograph param-
eters may be altered with prior approval of
the Administrator. If there Is reason to be-
lieve that some other hydrocarbon with an
identical retention time Is present In the
sample, then supplemental confirmation of
the vinyl chloride peak through an absoltue
analytical technique, such as mp.ss spec-
troscopy, should be performed.
4. Apparatus.
4.1 Sampling (Figure 1).
4.1.1 Probe—Stainless steel, Pyrex glass.
or Teflon tubing according to stack temper-
ature, each equipped with a glass wool plug
to remove partlculate matter.
4.1.2 Sample line—Teflon, 6.4 mm outside
diameter, of sufficient length to connect
probe to bag. A new unused piece Is employed
for each series of bag samples that constitutes
an emission test.
4.1.3 Male (2) and female (2) stainless
steel quick-connects, -with ball checks (one
pair without) located as shown in Figure 1.
4.1.4 Tedlar bags, 100 liter capacity—To
contain sample.
4.1.5 Rigid leakproof containers for 4.1.4,
with covering to protect contents from sun-
light. ^ -
4.1.6 Needle valve—To adjust sample flow
rate. .
4.1.7 Pump—Leak-Tree. Minimum capac-
ity 2 liters per minute.
4.1.8 Charcoal tube—To prevent admis-
sion of vinyl chloride to atmosphere hi vicin-
ity of samplers.
4.1.9 Flow meter—For observing sample
flow rate; capable of measuring a flow range
from 0.10 to 1.00 liter per minute.
4.1.10 Connecting tubing—Teflon, 6.4 mm
outside- diameter, to assemble sample train
(Figure 1).
4.1.11 Pltot tube—Type E (or equivalent),
attached to the probe so that the sampling
1 Mention of trade names on specific prod-
ucts does not constitute endorsement by the
Environmental Protection Agency.
How rate can be regulated proportional tu
the stack gas, velocity.
4.2 Sample recovery.
4.2.1 Tubing—Teflon, 6.4 mm outside
diameter, to connect bog to gas chromato-
graph sample loop. A new unused piece la
employed for each series of bag samples that
constitutes an emission test, and Is to be dis-
carded upon conclusion of analysis of those
bags.
4.3 Analysis.
4.3.1 Gas chromatograph—With flame
lontzatton detector, potentlometrlc strip
chart recorder and 1.0 to 6.0 ml heated sam-
pling loop In automatic sample valve.
4.3.2 Chromatographlc column—Stainless
steel, 2.5 m X 3.2 mm, containing 80/100
mesh Chromosorb 102.
4.3.3 Flow meters (2)—Rotameter type,
0 to 100 ml/mln capacity, with flow control
valves.
4.3.4 Gas regulators—For required gas
cylinder:;:
4.3 5 Thermometer—Accurate to one de-
gree centigrade, to measure temperature of
'heated sample loop at time of sample Injec-
tion.
4.3.6 Barometer—Accurate to 5 mm Hg, to'
measure atmospheric pressure around gas
chromatograph during sample analysis.
4.3.7 Pump—Leak-Tree. Minimum capac-
ity 100 ml/mln.
4.4 Calibration.
4.4.1 Tubing—Teflon, 6.4 mm outside
diameter, separate pieces marked for each
calibration concentration.
4.4.2 Tedlar bags—Sl'xteen-lnch square
size, separate bag marked for each calibra-
tion concentration.
4.4.3 Syringe—0.5 ml, gas tight.
4.4.4 Syringe—50M, gas tight.
4.4.5 Flow meter—Rotameter type, 0 to.
1000 ml/mln range accurate to ±1%, to
meter nitrogen in. preparation of standard
gas mixtures.
4.4.6 Stop watch—Of known accuracy, to
time gas flow in preparation of standard gas
mixtures.
5. Reagents. It is necessary that all rea-
gents be of chroma-tographlc grade.
5.1 Analysis.
5.1.1 Helium gas or nitrogen gas—Zero
grade, for chromatographlc carrier gas.
5.1.2 Hydrogen gas—Zero grade.
5.1.3 Oxygen gas—Zero grade,
5.2 Calibration.
5.2.1 Vinyl chloride, 99.9+%—For prep-
aration of standard ga-s mixtures.
5.2.2 Calibration cylinders (3'), optional—
One each of 50, 10 and 6 ppm vinyl chloride
In nitrogen with certified analysis.
5.2.3 Nitrogen gas—Zero grade, for prep-
aration of standard gas mixtures.
6. Procedure.
6.1 Sampling. Assemble the sample train
as In Figure 106-1. Perform a bag leak check
according to Section 7.4. Observe that all
connections between the bag and the probo
are tight. Place the end of the probe at the
centrold of the stack and start the pump
with the needle valve adjusted to yield a
flow of 0.5 1pm. After a period of time suffi-
cient to purge the line several times has
elapsed, connect the vacuum line to the
bag and evacuate the bag until the rotam-
etcr indicates no lloCv. Then reposition the
sample and vacuum lines and begin the ac-
tual sampling, keeping the rate proportional
to the stock velocity. Direct the gas existing
the rotameter away from sampling personnel.
At the end of the sample period, shut off the
pump, disconnect the sample line from the
bag, and disconnect the vacuum line from
the bag container. Protect the bag container
from sunlight.
FEDERAL REGISTER, VOl. 40, NO. 248—WEDNESDAY, DECEMBER 24, 1975
' 26
-------
r>!>5r>o
PROPOSED RULES
0.2 Sample storage. Sample bni;s must bo
kept out of direct sunlight. When ut all pos-
sible, analysis Is to be performed within 24
hours of sample collection.
0.3 Sample recovery. With n piece of Tef-
lon tubing Identified for that bug. connect a
bag Inlet valve to the gas chromatogruph
F.;unple valve. Switch the valve to withdraw
gas from the bag through the sample loop.
Plumb the equipment so the sample gas
passes from the sample valve to the Inak-frce
pump, and tlirn to a charcoal tube, followed
by a 0-100 ml/mlti rotamclcr with How con-
trol valve.
6.4 Analysis. Set the column temperature
to 155" C. the detector temperature to 225'
C, and the sample loop temperature to 70° C.
When optimum hydrogen and oxygen flow
rates have been determine, verify and main-
tain these flow rates during all chromato-
graph operations. Using zero helium or
nitrogen as the carrier gas, establish a flow
rate In the range consistent with the manu-
facturer's requirements for satisfactory de-
tector operation. A flow rate of approxi-
mately 15 ml/min should produce adequate
separations. Observe the base line periodi-
cally and determine that the noise level has
stabilized and that base line drift has ceased.
Purge the sample loop for thirty seconds at
the rate of 100 ml/min, then activate the
sample valve. Record the Injection time (the
position of the pen on the chart at the time
of sample Injection), the sample number, the
sample loop temperature, the column tem-
perature, carrier gas flow rate, chart speed
and the attenuator setting. Record the lab-
oratory pressure. From the chart, select the
peak having the retention time correspond-
ing to vinyl chloride, as determined In Sec-
tion 7.2. Measure the peak area, Am, by use
of the automatic Integrator. Record Am and
the retention time. Repeat the Injection at
least two times or until two consecutive vinyl
chloride peaks do not vary In area, more than
6%. The average value for these two areas
1 T.'lll be used to compute the bag concentra-
tion.
7. Calibration and Standards.
7.1 Preparation of vinyl chloride standard
gas mixtures. Evacuate a sixteen-lnch square
Tedlar bag that has passed a leak check
(described In Section 7.4) and meter Is 5.0
liters of nitrogen. While the bag is filling, use
the 0.5 ml syringe to Inject 250/d of 99.9+%
vinyl chloride through the wall of the bag.
Upon withdrawing the syringe needle, Im-
mediately cover the resulting hole with a
piece of adhesive tape. This gives a concen-"
tratlon of 50 ppm of vinyl chloride. In a like
manner use the other syringe to prepare dilu-
tions having 10 and 6 ppm vinyl chloride
concentrations. Place each. bag. on a smooth
eurface and alternately depress opposite
sides of the bag 50 times to further mix the
gases. •
7.2 Determination of vinyl chloride re-
tention time. This section can be performed
simultaneously with Section 7.3. Establish
chromatograph conditions Identical with
those In Section 6.3, above. Set attenuator
to X 1 position. Flush the sampling loop
with zero helium or nitrogen and activate
the sample valve. Record the Injection time,
the sample loop temperature, the column
temperature, the carrier gas flow rate, the
chart speed and the attenuator • setting.
Record peaks and detector' responses that
occur In the absence of vinyl chloride. Main-
tain conditions. With the equipment plumb-
Ing arranged Identically to Section G.3, flush
the sample loop for 30 seconds at the rate of
100 ml/mln with one of the vinyl chloride
calibration mixtures and activate the sample
valve. Record the Injection time. Select the
peak that corresponds to vinyl chloride.
Measure the distance on the chart from the
Injection time to the time at which the peak
maximum occurs. This quantity, divided by
the chart spued, Is defined as the retention
time. Record.
7.3 Preparation of chromatograph cali-
bration curve. Make a gas chromatographlc
nu-nsurcinent of each standard gas mixture
(described In Section T.I) using conditions
Identical with those listed In Section 6.3
above. Flush the sampling loop for 30 seconds
at the rate of 100 ml/mln with each standard
Ka« mixture and activate the sample valve.
Record Cr, the concentrations of vinyl chlo-
rklo Injected, the attenuator setting, chart
speed, peak area, sample loop temperature,
column temperature, carrier gas flow rate,
and retention time. Record the laboratory
pressure. Calculate Ac, the peak area multi-
plied by the attenuator setting. Repeat until
two Injection areas are within G%, then plot
those points vs Cc. When the other concen-
trations have been plotted, draw a smooth
curve through the points. Perform calibra-
tion dally, or before and after each set of
bag samples, whichever Is more frequent.
7.4 Tedlar bag leak checks. Before each
use, make sure a bag Is leak-free by checking
It for leaks. To leak check, connect a water
manometer and pressurize the bag to 5-10
cm HaO (2-4 In. H..O). Allow to stand for
10 minutes. Any displacement In the water
manometer Indicates a leak.
(NOTE: An alternative leak check method
Is to pressurize the bag to 6-10 cm H2O or
2-4 In. H2O and allow to stand overnight.
A deflated" bag Indicates a leak.)
8. Calculations.
8.1 Determine the sample peak area as
follows:
whore:
A ,-Th« sample peak area:
/(."The lucuBurcri peak area;
Ai-Tbe attenuation factor.
8.2 Vinyl chloride concentrations. From
the calibration curve described in Section
7.3, above,, select the value of C. that cor-
responds to A,, the sample peak area. Cal-
culate Cb as follows:
where
C»=
c.=
/>,=
Pi-
T,--
• Equation 100-2
The concentration of vinyl chloride In the bag
sample in ppm.
The concentration of vinyl chloride Indicated by
the pas chromalr.eraph, In prim.
The reference pressure, the laboratory pressure
recorded during calibration, mm Ug.
The sample Icop temperature on the absolute
scale at the time of analysis, °K.
•The laboratDry pressure at time of analysis, mm
The reference temperature, the sample loop
temperature recorded during calibration, "K.
Ae=AmAf
Equation 106-1
9. References.
1. Brown, D. W., Loy, E. W. and Stephen-
son, M. H. "Vinyl Chloride Monitoring Near
the B. P. Goodrich Chemical Company In.
Louisville, Kentucky." Region IV, UJS. Envi-
ronmental Protection Agency, Surveillance
and Analysis Division, Athens, Georgia, June
24, 1974.
2. "Evaluation of A Collection and Analy-
tical Procedure for Vinyl Chloride In Air,"
by G. D. Clayton and Associates, December
13, 1974. EPA Contract No. 68-02-1408, Task
Order No. 2. EPA Report oN. 75-VCL-l.
riltet(Cl«» Vool>T|
Sennet V) Type
Meat Tubs
SUcV Mill
s
Yltet KiiKwctcr
106-1. Integrated bag tabling tr«lo.
Kentlnn of trade nanes on specific products 4ooa not constitute
: "by the Ziivicooaantal Protection Agtncj.
FEDERAL REGISTER, VOL 40, NO. 248—WEDNESDAY, DECEMBER 24, 1975
27
-------
APPENDIX B
METHOD 106 - DETERMINATION OF VINYL CHLORIDE FROM STATIONARY SOURCES
MRI MODIFIED PROCEDURE
28
-------
December 3, 1976
METHOD 106 - DETERMINATION OF VINYL CHLORIDE
FROM STATIONARY SOURCES
1. Principle and Applicability
1.1 An integrated bag sample of stack gas containing vinyl chloride
(chloroethene) is subjected to chromatographic analysis using a flame ion-
ization detector.
1.2 The method is applicable to the measurement of vinyl chloride in
stack gases from both vinyl chloride and polyvinyl chloride manufacturing
processes, except where the vinyl chloride is contained in particulate
matter. Care must be exercised to prevent undue exposure of sampling per-
sonnel to vinyl chloride.
2. Range and Sensitivity
The lower limit of detection will vary according to the chromatograph
used. Values reported include 1 x 10"^ g and 4 x 10 g.
3. Interferences
Retention indices for several possible interferences are given in Table
B-l.The only known serious interferences on the normal chromatographic col-
umn are acetaldehyde (vinyl alcohol) and ethylene oxide, which can occur in
monomer streams, ambient air, and some copolymer operations. If resolution
of the vinyl chloride peak is not satisfactory for a particular sample, then
chromatograph parameters may be altered with prior approval of the administrator.
4. Apparatus
4.1 Sampling (Figure 1)
4.1.1 Probe - Stainless steel, Pyrex* glass, or Teflon* tubing
according to stack temperature, equipped with a glass wool plug to remove
particulate matter.
4.1.2 Sample line - Teflon, 6.4 mm diameter, of sufficient
length to connect probe to bag.
4.1.3 Male (2) and Female (2) stainless steel quick-connects,
with ball checks (one pair without) located as shown in Figure 1.
* Mention of trade names on specific products does not constitute
endorsement by the Environmental Protection Agency.
29
-------
TABLE B-l. RETENTION INDICES FOR POSSIBLE VINYL CHLORIDE INTERFERENCES
Methane*
Ethylene
E thane*
Propane*
Methyl chloride
Methanol
Acetaldehyde
Ethylene oxide
Vinyl chloride
Isobutane
Isobutylene
1-Butene
n-Butane*
1,3-Butadiene
trans-2-Butene
Ethanol
Ethyl chloride
cis-2-Butene
1, 1-Dichloroethylene
trans-l» 2-Dichloroethylene
2 m Chromosorb 102
100°C
100
180
200
300
320
330
355
355
360
380
395
395
400
400
400
415
415
415
480t
510f
2 m Chromosorb 102
+ 1 m Porapak T
120°C
100
180
200
300
340
395
400
395
375
380
395
395
400
410
410
475
430
415
490f
495f
2 m Chromosorb 102
+ 2 m SF-96
120°C
100
175
200
300
320
350
375
375
360
380
385
395
400
395
400
435
415
410
480t
505t
2 m Chromosorb 102
+ new 2 m SF-96
120°C
100
-
200
300
325
375
(low levels retained)
400
(low levels retained)
-
(low levels retained)
360
380
390
-
400
395
-
'
410
-
-
™
* Reference compounds for
t Column at 150°C.
indices.
-------
Condenser-
Reverse ("S") Type
Pitot Tube
Stack Wall
Pitot Manometer
Flow Meter
Teflon
Sample Line
Vacuum Line
Male
Quick
Connects
Female"1^,
Needle Valve
Pump
Rigid Leak Proof
Container
Vent
Charcoal Tube
Figure B-l. Integrated bag sampling train
(1) Mention of trade names on specific products does not constitute
endorsement by the Environmental Protection Agency.
-------
4.1.4 Tedlar* bag, 100 liter capacity. Teflon* bags are not
acceptable. Aluminized Mylar* bags may be used provided the samples are
analyzed within 24 hours after collection.
4.1.5 Rigid leakproof container for 4.1.4, with covering to pro-
tect contents from sunlight.
4.1.6 Needle valve - To adjust sample flow rate.
4.1.7 Pump - Minimum capacity 2 liters per minute.
4.1.8 Charcoal tube - To prevent admission of vinyl chloride to
atmosphere in vicinity of samplers.
4.1.9 Flowmeter - For observing sample flow rate; capable of
measuring a flow range from 100 to 1,000 ml/min. The flowmeter and its
fittings must be made of glass or stainless steel.
4.1.10 Connecting tubing - Teflon, 6.4 mm diameter, to assemble
sample train (Figure 1).
4.1.11 Pitot tube - Type S (or equivalent), attached to the probe
so that the sampling flow rate can be regulated proportional to the stack gas
velocity.
4.1.12 Condenser - Glass vacuum trap, 25 x 200 mm with 10 mm
connecting tubes, immersed in an ice bath. Required only if condensation
can occur from the sample stream at ambient temperatures. Teflon* fittings
should be used to connect the condenser.
4.2 Sample Recovery
4.2.1 Tubing - Teflon, 6.4 mm diameter, to connect bag to gas
chromatograph sample loop. Discard the tubing if it shows any contamina-
tion (should be checked periodically by sampling a purified nitrogen stream
with the tubing and checking the chromatogram). The tubing should be re-
placed if condensation occurs or if a very high (> 300 mg/m^) level of vinyl
chloride is sampled, to be followed by samples near 3 mg/m^ (1 ppm).
4.3 Analysis
4.3.1 Gas chromatograph - With flame ionization detector, po-
tentiometric strip chart recorder and 1.0 to 5.0 ml heated sampling loop in
sample valve.
* Mention of trade names on specific products does not constitute
endorsement by the Environmental Protection Agency.
32
-------
4.3.2 Chromatographic column - Stainless steel, 2.0'm x 3.2 mm
O.D. ,* containing 80/100 mesh Chromosorb 102. A secondary column of 20% SF-
96 on AW Chromosorb P, 60/80 mesh, stainless steel column, 2 m x 3.2 mm O.D.,
or Porapak T, 80/100 mesh, stainless steel column, 1.0 m x 3.2 mm O.D., may
be required to separate acetaldehyde from vinyl chloride. If used, the sec-
ondary column is added after the Chromosorb 102 column. Both secondary col-
umns separate the acetaldehyde/vinyl chloride peaks with the acetaldehyde
eluting after vinyl chloride. The combined columns should be operated at
120°C.
4.3.3 Flow meters (2) - Rotometer type, 0 to 100 ml/min capacity.
4.3.4 Gas regulators - For required gas cylinders.
4.3.5 Thermometer - Accurate to one degree centigrade. To measure
temperature of heated sample loop at time of sample injection.
4.3.6 Barometer - Accurate to 500 Pa (5 mm Hg). To measure
atmospheric pressure around gas chromatograph during sample analysis.
4.4 Calibration
4.4.1 Tubing - Teflon, 6.4 mm diameter, for collecting calibra-
tion cylinders to sample loop.
4.4.2 Regulator(s) - For vinyl chloride calibration gases.
4.4.3 Needle valve(s) - For flow control from calibration cylinders.
5. Reagents (It is intended that all reagents be of chromatographic grade.)
5.1 Analysis
5.1.1 Helium gas or nitrogen gas - Zero grade, for chromatographic
carrier gas.
5.1.2 Hydrogen gas - Zero grade
5.1.3 Breathing air
5.2 Calibration
5.2.1 Calibration cylinders (3). One each of 150, 30 and 15
mg/nr- (50, 10 and 5 ppm) vinyl chloride in nitrogen with certified analy-
sis. Analysis should be traceable to NBS or to a gravimetrically cali-
brated vinyl chloride permeation tube.
* All chromatographic columns used have a wall thickness of approximately
0.5 mm (0.020 in.).
t Specified at conditions of 293°K and 101.3 kPa (760 mm Hg). To convert
from parts per million (v/v) to milligrams per cubic meter, multiply
by 2.59.
33
-------
5.2.2 Nitrogen gas - Zero grade
6. Procedure
6.1 Sampling. Assemble the sample train as in Figure 1. Place the
probe in the stack and start the pump with the needle valve adjusted to
yield a flow of 500 ml/min. After a period of time sufficient to purge the
line several times has elapsed, connect the vacuum line to the bag and
evacuate the bag until a rotameter connected to the bag container outlet
indicates no flow. Then reposition the sample and vacuum lines and begin
the actual sampling, keeping the rate proportional to the stack velocity.
Direct the gas exiting the pump away from sampling personnel. At the end
of the sample period, shut off the pump, disconnect the sample line from
the bag, and disconnect the vacuum line from the bag container. Protect
the bag from sunlight.
6.1.1 Tedlar bag leak checks - Before each use, make sure a
bag is leak-free by checking it for leaks. To leak check, connect a water
manometer and pressurize the bag to 500 to 1,000 Pa (5 to 10 cm water col-
umn). Allow to stand for 10 min. Any displacement in the water manometer
indicates a leak. (Note: An alternative leak check method is to pressur-
ize the bag to 500 to 1,000 Pa (5 to 10 cm water column) and allow to stand
overnight. A deflated bag indicates a leak.)
Test the Rigid container for leaks in a similar manner to that
for the bag leak check.
Place a rotameter in-line between the Tedlar bag and pump inlet.
Evacuate the bag. Failure of the rotameter to register zero flow when the
bag is empty indicates a leak.
6.2 Sample recovery. With an uncontaminated piece of Teflon tubing
connect a bag to the gas chromatograph sample valve. Connect a 0 to 100 ml/min
rotameter between the sample valve outlet and the inlet to a small pump with
a needle valve flow control.
6.3 Analysis. Set the column temperature to 100°C, the detector temper-
ature to 160°C, and the sample loop temperature to 70°C. When optimum hydrogen
and air flow rates have been determined, verify and maintain these flow rates
during all chromatograph operations. Using zero helium or nitrogen as the
carrier gas, establish a flow rate in the range consistent with the manufac-
turer's requirements for satisfactory detector operation. A flow rate of
40 ml/min has been shown to produce adequate separations. Observe the base-
line periodically and determine that the noise level has stabilized and that
baseline drift has ceased. Purge the sample loop for at least one minute at
34
-------
a rate of 50 ml/min, stop the pump, and then activate the sampling valve.
Record the injection time (the position of the pen on the chart at the time
of sample injection), the sample number, the sample loop temperature, the
column temperature, carrier gas flow rate, chart speed and the attenuator
setting. Record the laboratory pressure. From the chart, select the peak
having the retention time corresponding to vinyl chloride. Measure the
peak area, A^, and the peak height, R^. Record A^ Hjjp and the retention
time. Repeat the injection at least two times or until two consecutive
vinyl chloride peaks do not vary in height or area more than 5%. The
average values will be used to compute the bag concentration.
Compare the ratio of 1^:^ for the vinyl chloride sample with the
ratio for the standard which is closest in concentration. If these ratios
differ by more than 107o, the vinyl chloride peak is not pure (acetaldehyde
may be present) and the peaks must be resolved (see Section 4.3.2). If
the height-area test shows impure peaks, the height measurements will be
more accurate than the area measurements. However, the chromatographic con-
ditions should be altered until the height-area test is passed.
7. Calibration
7.1 Determination of vinyl chloride retention time. Perform before
and after analyzing bag samples. Establish chromatograph conditions identi-
cal with those in 6.3 above. Set attenuator to X 1 position. Flush the
sampling loop with zero nitrogen and activate the sample valve. Record the
injection time, the sample loop temperature, the column temperature, the
carrier gas flow rate, the chart speed and the attenuator setting. Record
peaks and detector responses that occur in the absence of vinyl chloride.
Maintain conditions. Flush the sample loop for at least 1 min at the rate
of 50 ml/min with the vinyl chloride calibration mixture and activate the
sample valve. Record the injection time. Select the peak that corresponds
to vinyl chloride. Measure the distance on the chart from the injection
time to the peak maximum. This quantity, divided by the chart speed, is
defined as the retention time. Record.
7.2 Preparation of chromatograph calibration curve. Make a gas
chromotographic measurement of each standard gas mixture using conditions
identical with those listed in Section 6.3 above. Flush the sampling loop
for at least 1 min at the rate of 50 ml/min with each standard gas mixture
and activate the sample valve. Record Cc, the concentrations of vinyl chlo-
ride injected, the attenuator setting, chart speed, peak area, peak height,
sample loop temperature, column temperature, carrier gas flow rate, and re-
tention time. Record the laboratory pressure. Calculate Ac, the peak area
multiplied by the attenuator setting. Repeat until two injection areas are
within 5%, then plot those points versus GC. When the other concentrations
have been plotted, draw a smooth curve through the points. Perform calibra-
tion daily or before and after each set of bag samples, whichever is more
frequent.
35
-------
8. Calculations
8.1 Determine the sample peak area as,follows:
A = A A,. Eq. 106-1
c m f ^
where:
AC = The sample peak area
AJJJ = The measured peak area
Af = The attenuation factor
8.2 Vinyl chloride concentrations. From the calibration curve described
in Section 7.3 above, select the value of C that corresponds to A , the sam-
ple peak area. Calculate C, as follows:
C P T
_ _ c r x Eq. 106-2
b " P-T
iiir
where:
O JU
C, = The concentration of vinyl chloride in the bag sample in mg/m .
C = The concentration of vinyl chloride indicated by the gas
C o -fr
chromatograph, in mg/m .
P = The reference pressure, the laboratory pressure recorded
during calibration, kPa (mm Hg).
T. = The sample loop temperature on the absolute scale at the
time of analysis, °K.
P. = The laboratory pressure at time of analysis, kPa (mm Hg).
T = The reference temperature, the sample loop temperature
recorded during calibration, °K.
9. Sample Storage
Sample bags must be kept out of direct sunlight. Analysis is to be
performed within 24 hours of sample collection.
* Specified at standard conditions of 293°K and 101.3 kPa (760 mm Hg),
36
-------
10. References
1. Brown, D. W., Loy, E. W., and Stephenson, M. H. "Vinyl Chloride
Monitoring Near the B. F. Goodrich Chemical Company in Louisville,
Kentucky." Region IV, U.S. Environmental Protection Agency, Sur-
veillance and Analysis Division, Athens, Georgia, June 24, 1974.
2. "Evaluation of a Collection and Analytical Procedure for Vinyl
Chloride in Air," by G. D. Clayton and Associates, December 13,
1974. EPA Contract No. 68-02-1408, Task Order No. 2. EPA Report
No. 75-VCL-l.
Note 1: (Section 4.1.4) The use of Teflon or aluminized Mylar bags has
been proposed. Teflon apparently allows excessive permeation through the
bag. It has no cost advantage over Tedlar and would not be acceptable.
Aluminized Mylar does provide an effective vapor barrier, is much less ex-
pensive than Tedlar, nor does it alter vinyl chloride concentrations in
the bags up to at least 2 weeks. However, it does tend to outgas various
heavy compounds over a period of several days which can contaminate the
GC column.
Note 2: (Section 4.1.9) With the rotameter position changed, the pump
does not have to be leakfree, nor does the bag container need to be com-
pletely leakfree. The change in the rotameter position does require a
high quality rotameter but it eliminates most leak problems in the system
since it provides a positive check that gas is going into the bag. In
this position, the rotameter is also not prone to false high readings
from pump pulsations.
Note 3; (Section 5.2) The bag calibration method was deleted because of
the safety problems of handling pure vinyl chloride in the field. It is
also more difficult to prepare accurate dilutions by this method. Compressed
cylinder gases are stable, safer to handle at such concentrations, and more
accurate. If special facilities already exist for handling vinyl chloride
in the laboratory, either procedure is suitable.
37
-------
APPENDIX C
METHOD 106 - DETERMINATION OF VINYL CHLORIDE FROM STATIONARY SOURCES
38
-------
METHOD 106—DETERMINATION or VINYI.
CHLORIDE FROM STATIONARY SOURCES
INTRODUCTION
Performance of this method should not be
attempted by persons unfamiliar with the
operation of a gas chromatograph, nor by
those who are unfamiliar with source sam-
pling, as there are many details that are
beyond the scope of this presentation. Care
must be exercised to prevent, exposure of
sampling personnel to vinyl chloride, a car-
cinogen.
1. Principle and Applicability.
1.1 An integrated bag sample of stack gas
containing vinyl chloride (chloroethylenc)
Is subjected to chromatographlc analysis,
using a flame ionlzation detector.
1.2 The method is applicable to the meas-
urement of vinyl chloride In stack gases from
ethylene dichloride, vinyl chloride and poly-
vinyl chloride manufacturing processes, ex-
cept where the vinyl chloride Is contained In
participate matter.
2. Range and Sensitivity.
The lower limit of detection will vary ac-
cording to the chromatograph used. Values
reported Include 1 x 10-' mg and 4 x 10-'
mg.
3. Interferences.
Acetaldehyde, which can occur In some
vinyl chloride sources, will Interfere with the
vinyl chloride peak from the Chromosorb 102
column. See sections 4.3.2 and 6.4. If resolu-
tion of the vinyl chloride peak is still not
satisfactory for a particular sample, then
chromatograph parameters can be further
altered with prior approval of the Admin-
istrator. If alteration of the chromatograph
parameters falls to resolve the vinyl chloride
peak, then supplemental confirmation of the
vinyl chloride peak through an absolute
analytical technique, such as mass spectro-
scopy, must be performed.
4. Apparatus.
4.1 Sampling (Figure 1).
4.1.1 Probe—Stainless steel, Pyrex glass,
or Teflon tubing according to stack temper-
ature, each equipped with a glass wool plug
to remove participate matter.
4.1.2 Sample line—Teflon, 0.4 mm outside
diameter, of sufficient length to connect
probe to bag. A new unused piece Is employed
for each series of bag samples that constitutes
an emission test.
4.1.3 Male (2) and female (2) stainless
steel quick-connects, with ball checks (one
pair without) located as shown in Figure 1.
4.1.4 Tecllar bags, 100 liter capacity—To
contain sample. Teflon bags are not accept-
able. Alumlnlzcd Mylar bags may be used,
provided that the samples are analyzed
within 24 hours of collection.
4.1.5 Rigid leakproof containers for 4.1.4,
with covering to protect contents from sun-
light.
4.1.8 Needle valve—To adjust sample flow
rate.
4.1.7 Pump—Leak-free. Minimum capac-
ity 2 liters per minute.
4.1.8 Charcoal tube—To prevent admis-
sion of vinyl chloride to atmosphere In vicin-
ity of samplers.
4.1.9 Flow meter—For observing sample
flow rate; capable of measuring a flow range
from 0.10 to 1.00 liter per minute.
4.1.10 Connecting tubing—Teflon, 6.4 mm
outside diameter, to assemble sample train
(Figure 1).
4.1.11 Pilot tube—Type S (or equivalent),
attached to the probe so that the sampling
flow rate can be regulated proportional to
the stack gas velocity.
4.2 Sample recovery.,
4.2.1 Tubing—Teflon, 6.4 mm outside
diameter, to connect bag to gas chromato-
graph sample loop. A new unused piece Is
employed for each series of bag samples that
constitutes an emission test, and Is to be dis-
carded upon conclusion of analysis of those
bags.
4.3 Analysis.
4.3.1 Gas chromatograph—With flame
lonization detector, potentiometric strip
chart recorder and 1.0 to 5.0 ml heated sam-
pling loop in automatic sample valve.
4.3.2 Chromatographlc column—Stainless
steel, 2.0 X 3.2 mm, containing 80/100 mesh
Chromosorb 102. A secondary colum of OE
SF-96, 20% on 60/80 mesh AW Chromosorb
P, stainless steel, 2.0 m X 3.2 mm, will be
required If acetaldehyde Is present. If used,
the SF-96 column is placed after the Chromo-
sorb 102 column. The combined columns
should then be operated at 110'C.
4.3.3 Flow meters (2)—Rotameter type,
0 to 100 ml/min capacity, with flow control
valves.
4.3.4 Gas regulators—For required gas
cylinders.
4.3 5 Thermometer—Accurate to one de-
gree centigrade, to measure temperature of
heated sample loop at time of sample Injec-
tion.
4.3.6 Barometer—Accurate to 5 mm Hg, to
measure atmospheric pressure around gas
chromatograph during sample analysis.
4.3.7 Pump—Leak-free. Minimum capac-
ity 100 ml/min.
'4.4 Calibration.
4.4.1 Tubing—Teflon, 6.4 mm outside
diameter, separate pieces marked for each
calibration concentration.
4.4.2 Tedlar bags—Slxteen-lnch square
size, separate bag marked for each calibra-
tion concentration.
4.4.3 Syringe—0.6 ml, gas tight.
4.4.4 Syringe—50/il, gas tight.
1 Mention of trade names on specific prod-
ucts does not constitute endorsement by the
Environmental Protection Agency.
4.4.6 Flow meter—Rotameter type, 0 to
1000 ml/mln range accurate to ±1%, to
meter nitrogen In preparation of standard
gas mixtures.
4.4.6 Stop watch—Of known accuracy, to
tlmo gas flow In preparation of standard gas
mixtures.
5. Reagents. It Is necessary that all rea-
gents be of chromatographlc grade.
5.1 Analysis.
5.1.1 Helium gas or nitrogen gas-*-Zero
grade, for chromatographlc carrier gas.
5.1.2 Hydrogen gas—Zero grade.
5.1.3 Oxygen gas, or Air, ns required by
the detector—Zero grade.
5.2 Calibration.
5.2.1 Vinyl chloride, 90.0+%—For prep-
aration of standard gas mixtures.
5.2.2 Calibration cylinders (3), optional—
One each of 50, 10 and 5 ppm vinyl chloride
In nitrogen with certified analysis. Analysis
must be traceable to NDS (National Bureau
of Standards) or to a grflvimetrlcally cali-
brated vinyl chloride permeation tube.
5.2.3 Nitrogen gas—Zero grade, for prep-
aration of standard gas mixtures.
6. Procedure.
6.1 Sampling. Assemble the sample train
as in Figure 106-1. Perform a bag leak check
according to Section 7.4. Observe that all
connections between the bag and the probe
are tight. Place the end of the probe at the
centroid of the stack and start the pump
with the needle valve adjusted to yield a
flow of 0.5 1pm. After a period of time suffi-
cient to purge the line several times has
elapsed, connect the vacuum line to the
bag and evacuate the bag until the rotam-
eter Indicates no flow. Then reposition the
sample and vacuum lines and begin the ac-
tual sampling, keeping the rate proportional
to the stack velocity. Direct the gas exiting
the rotameter away from sampling personnel.
At the end of the sample period, shut off the
pump, disconnect the sample line from the
bag, and disconnect the vacuum line from
the bag container. Protect the bag container
from sunlight.
6.2 Sample storage. Sample bags must be
kept out of direct sunlight. When at all pos-
sible, analysis Is to be performed within 24
hours of sample collection.
6.3 Sample recovery. With a piece of Tef-
lon tubing identified for that bag, connect a
bag inlet valve to the gas chromatograph
sample valve. Switch the valve to withdraw
gas .from the bag through the sample loop.
Plumb the equipment so the sample gas
passes from the sample valve to the leak-free
pump, and then to a charcoal tube, followed
by a 0-100 ml/mln rotameter with flow con-
trol valve.
6.4 Analysis. Set the column temperature
to 100° C the detector temperature to 150*
C. and the sample loop temperature to 70' C.
When optimum hydrogen and oxygen flow
rates have been determined verify and main-
tain these flow rates during all chromato-
graph operations. Using zero helium or
nitrogen as the carrier gas, establish a flow
rate in the range consistent with the manu-
facturer's requirements for satisfactory de-
tector operation. A flow rate of approxi-
mately 40 ml/mln should produce adequate
separations. Observe the base line periodi-
cally and determine that the noise level has
stabilized and that base line drift has ceased.
Purge the sample loop for thirty seconds at
the rate of 100 ml/min. then activate the
sample valve. Record the injection time (the
position of the pen on the chart at the time
of sample injection), the sample number, the
sample loop temperature, the column tem-
perature, carrier gas flow rate, chart speed
39
-------
and the attenuator setting. Record the lab-
oratory pressure. From the chart, select th&
peak Having the retention time correspond-
ing to vinyl chloride, us determined In Sec-
tion 7.2. Measure the peak area, Am, by use
of Hin, and a disc integrator or a pl.inlmetcr.
Measure the peak height, II.n. Record Am and
the retention time. Repeat the Injection at
least two times or until two consecutive vinyl
chloride peaks do not vary In area more than
5'.;.. The average value for these two areas
v.ill be u.-i'U to compute the bay concentra-
tion.
Compare the ratio of II..i to Am for the vinyl
chloride tomplo with the same ratio for the
standard peak which is closest in height. As
a Kiiidcimo, if these ratios differ by more
tliiin 10', . the vinyl chloride peak may not
be pure (possibly acetnldehyde is present)
and tho secondary column should be em-
plowd (see Section 4.3.2).
O.f) M"iwiiro the ambient temperature and
b.i.'onu'tri'.' pressure near the bag. (Assume
the- relative humidity to be 100 percent.)
From a water saturation vapor pressure table,
determine the record and water vapor con-
tent of the bag.
7. Calibration and Standards.
7.1 Preparation of vinyl chloride standard
Sjns mixtures. Evacuate a sixteen-inch square
Tedlar b^g that lias passed a leak check
(described In Section 7.4) and meter in 5.0
liters of nitrogen. While the bag is filling, xise
the 0.5 ml syringe to inject 250fi\ of 99.9+%
vinyl chloride through the wall of the bag.
Upon withdrawing the syringe needle. Im-
mediately cover the resulting hole with a
piece of adhesive tape. This gives a concen-
tration of 50 ppm of vinyl chloride. In a like
manner use the other syringe to prepare dilu-
tions having 10 and 5 ppm vinyl chloride
concentrations. Place each bug on a smooth
surface and alternately depress opposite
sides of the bag SO times to further mix the
gases.
7.2 Determination of vinyl chloride re-
tention time. This section can be performed
simultaneously with Section 7.3. Establish
chromatopraph conditions identical with
those in Section 6.3, above. Set attenuator
to X 1 position. Flush the sampling loop
with zero helium or nitrogen and activate
the sample valve. Record the injection time,
the sample loop temperature, the column
temperature, the carrier gas flow rate, the
chart speed and the attenuator setting.
Record peaks and detector responses that
occur In the absence of vinyl chloride. Main-
tain conditions. With the equipment plumb-
ing arranged identically to Section 6.3, flush
the sample loop for 30 seconds at the rate of I
100 ml/ min with one of the vinyl chloride
calibration mixtures and activate the sample
valve. Record the injection time. Select the
peak that corresponds to vinyl chloride.
Measure the distance on the chart from the
injection time to the time at which the peak
maximum occurs. This quantity, divided by
the chart speed, is deliiied as the retention
time. Record.
7.3 Preparation of chromatograph cali-
bration curve. Make ft g!\s chromatographic
measurement of each standard gas mixture
(described in Section 7.1) using conditions
Identical with those listed In Section 6.3
above. Plush the sampling loop for 30 seconds
at the rate of 100 ml/min with each standard
gas mixture and activate the sample valve.
Record C,., the concentrations of vinyl chlo-
ride injected, the attenuator setting, chart
speed, peak area, sample loop temperature,
column temperature, carrier gas flow rate,
and retention time. Record the laboratory
pressure. Calculate At, the peak area multi-
i 106-1. Integrated 1>B
train,
'l' Xent'lon of tr«4« ««««• on fpiclfle pioJucH doci not eoaitltuU
•odoiocupnt by (U {avlronMiiUl Protection Agoncjr.
piled by the attenuator setting. Repeat until
two Injection areas are within 6%, then plot
those points vs C,. When tlie other concen-
trations have been plotted, draw a smooth
curve through the points. Perform calibra-
tion daily, or before and after each set of
bag samples, whichever is more frequent.
7.4 Bag leak checks. While performance
of this section is required subsequent to bag
use. it Is also advised that it be performed
prior to bag use. After each use, make sure
a bag did not develop leaks as follows. To leak
check, connect a water manometer and pres-
surize the bag to 5-10 cm H.O (2-4 in HSO).
Allow to stand for 10 minutes. Any displace-
ment In the water manometer indicates a
leak. Also check the rigid container for leaks
In this manner.
(NOTE: An alternative leak check method
is to pressurize the bag to 5-10 cm H..O. or
2-4 in. H.O and allow to stand overnight.
A deflated bag indicates a leak.) For each
sample bag In Its rigid container, place a
rotameter In-line between the bag and the
pump Inlet. Evacuate the bag. Failure of the
rotameter to register zero flow when the bag
appears to be empty indicates a leak.
8. Calculations.
8.1 Determine the sample peak area as
follows:
8.2 Vinyl chloride concentrations. Prom
the calibration curve described In Section
7.3, above, select the value of C, that cor-
responds to A,, the sample peak area. Cal-
culate Cb as follows:
„ C,P,Ti
Equation 106-2
'Where:
J3»i=Tlie water vapor conltml of the bag saiuMu, as
analyzed.
C*=The concentration of vinyl chloride in the bag
sample In ppm.
C,=TIie concentration ol vinyl chloride indicated by
the gas chromntograph, la ppm.
Pr=»The reference pressure, the laboratory pressure
recorded during calibration, nun Hfr
J\=The sample loop temperature on the absolute
scale at the time of analysis, °K.
Pi=The laboratory pressure at time of analysis, mm
Hg.
T,=Tho reference temperature, the sample loop
temperature recorded during calibration, °K'
whore:
X.=The sample peak area.
v4»=The measured peak area.
X/=The attenuation factor.
9. References.
1. Brown, D. W., Loy, E. W. and Stephen-
son, M. H. "Vinyl Chloride Monitoring Near
the B. F. Goodrich Chemical Company In
Louisville, Kentucky." Region IV, US. Envi-
ronmental Protection Agency, Surveillance
and Analysis Division, Athens, Georgia, June
24, 1974.
Equation 106-1 3. "Evaluation of A Collection and Analy-
tical Procedure for Vinyl Chloride in Air."
_ by G. D. Clayton and Associates, December
• 13. 1974. EPA Contract No. 68-02-1408, Tusk
Order No. 2. EPA Report ON. 75-VCL-l.
8. "Standardization of Stationary Source
Emission Method for Vinyl Chloride," by Mid-
west Research Institute, 1976. EPA Contract
No. 68-02-1098. Task Order No. 7.
40
------- |