United States
Environmental Protection
Agency
Atmospheric Research and
Exposure Assessment Laboratory
Research Triangle Park, NC 27711
Research and Development
EPA/600/SR-94/130 May 1995
EPA Project Summary
Field Test of a Generic
Method for Halogenated
Hydrocarbons: A VOST Test at a
Chemical Manufacturing Facility
Using a Modified VOST
Sampling Method
James F. McGaughey, Joan T. Bursey, and Raymond G. Merrill
Because of the lack of commercial
availability of SKC Lot 104 petroleum-
based charcoal, the lot of charcoal
specified for use in the Volatile Or-
ganic Sampling Train (VOST, SW-846
Method 0030), a laboratory study was
performed to select an appropriate sub-
stitute. On the basis of performance,
availability, and cost, Anasorb 747®*
was selected as the substitute for SKC
Lot 104, and a draft sampling protocol
for the modified VOST train was writ-
ten. In the modified VOST sampling
train, the second tube containing se-
quential beds of Tenax® and charcoal
was replaced by a tube containing only
Anasorb®. This replacement necessi-
tated the use of two Tenax® front tubes
in the train, for a total of three sorbent
tubes. A field evaluation study, using
dynamic spiking and quadruple sam-
pling trains, was conducted to provide
a statistical comparison of the perfor-
mance of the VOST and the modified
VOST methodologies for the volatile
halogenated organic compounds listed
in Title III of the Clean Air Act Amend-
ments of 1990.
* Mention of trade names or commercial prod-
ucts does not constitute endorsement or recom-
mendation for use.
This Project Summary was developed
by EPA's Atmospheric Research and
Exposure Assessment Laboratory, Re-
search Triangle Park, NC, to announce
key findings of the research project
that is fully documented in a separate
report of the same title (see Project
Report ordering information at back).
Introduction
Validation of a method for a particular
analyte or group of analytes means that
the performance of the methodology for
these analytes has been established and
demonstrated through field tests at the
type of source category of interest: that is,
the precision and bias of the method have
been established experimentally. Field vali-
dation may be performed by performing
spiking operations in the field so that the
precision and bias of the method can be
demonstrated from sample collection
through analysis, using either static or dy-
namic spiking procedures as specified in
EPA Method 301, the Validation Protocol.
Another procedure for validation of a
method is a side-by-side comparison of a
candidate method to a validated method
to establish comparable performance for
the same analytes in the same matrix
(same source category).
-------
Under Title III of the Clean Air Act
Amendments of 1990, identification and/
or validation of sampling and analytical
methods are required for the volatile halo-
genated organic compounds which are
listed (Table 1). The candidate test method
is the Volatile Organic Sampling Train
(VOST), which consists of SW-846 Sam-
pling Method 0030 and SW-846 Analytical
Method 5040 or 5041. Method 0030, the
sampling protocol, requires the use of
Tenax® and a specific petroleum-based
charcoal (SKC Lot 104) or its equivalent.
Since SKC Lot 104 charcoal is no longer
commercially available and Method 0030
contains no performance specifications for
the charcoal, a laboratory study was per-
formed to identify an acceptable equiva-
lent. Several candidate sorbents were stud-
ied in the laboratory, and Anasorb 747®
was selected as a candidate sorbent for a
field test because of its performance, avail-
ability, and cost. In the course of the labo-
ratory studies, the decision was made to
use only Anasorb® in the back tube rather
than the sequential beds of Tenax® and
charcoal which had been used previously.
Use of a back tube consisting of only
Anasorb® necessitated the use of two
Tenax® front tubes in a modified VOST
train. A new sampling protocol was writ-
ten to incorporate the use of the modified
sorbent tube structure. The analytical pro-
cedures of Method 5041 were also modi-
fied to include individual analysis of each
of the three sorbent tubes to examine
compound distribution. In routine applica-
tion, the analysis would be performed with
paired Tenax® tubes.
A presurvey conducted at a chemical
manufacturing facility established that no
compounds of interest were present above
a level of 5 times the Method Detection
Limit. Evaluation of the modified VOST
method was accomplished by performing
sampling and analysis following Draft
Method 0031 and Method 5041, using
Method 0030 as the reference method re-
quired by EPA Method 301. The bias of
the method can be evaluated by compar-
ing the results obtained by one candidate
method to the results obtained by a vali-
dated method. To determine the precision
of the test method, multiple or collocated
Table 1. Volatile Halogenated Organic Compounds, Title III of the Clean Air Act Amendments, for Which Laboratory Testing has Determined the
Applicability of the VOST Method
Compound
Boiling
point
Comments
allyl chloride (3-chloropropene)
bis(chloromethyl) ether
carbon tetrachloride
chlorobenzene
chloroform
chloromethyl methyl ether
chloroprene
1,3-dichloropropylene
epichlorohydrin
ethyl chloride (chloroethane)
ethylene dibromide (1,2-dibromoethane)
ethylene dichloride (1,2-dichloroethane)
ethylidene dichloride (1,1-dichloroethylene)
methyl bromide (bromomethane)
methyl chloride (chloromethane)
methyl chloroform (1,1,1-trichloroethane)
methylene chloride
methyl iodide (iodomethane)
propylene dichloride (1,2-dichloropropane)
tetrachloroethylene
1,1,2-trichloroethane
trichloroethylene
vinyl chloride
vinyl bromide (1,1-dichloroethene)
vinylidene chloride
44-46
1061
77
1321
60.5-61.5
55-57
59.4
105-1062
115-1171
123
737-732'
83
57
43
-24.23
74-76
39.8-40
41-43
95-96
727'
770-775'
86.9
-13.43
1&
30-32
Acceptable performance in laboratory
Decomposes in water; cannot be analyzed
Recovery too high in laboratory study
Acceptable performance in laboratory
Acceptable performance in laboratory
Decomposes in water; cannot be analyzed
Acceptable performance in laboratory
Acceptable performance in laboratory
Soluble in water; cannot be analyzed
Acceptable performance in laboratory
Acceptable performance in laboratory
Acceptable performance in laboratory
Acceptable performance in laboratory
Recovery unacceptably high in laboratory
Erratic and unacceptably high recovery in laboratory
Recovery too high in laboratory study
Recovery too high in laboratory study
Acceptable performance in laboratory
Acceptable performance in laboratory
Acceptable performance in laboratory
Acceptable performance in laboratory
Acceptable performance in laboratory
Acceptable performance in laboratory
Acceptable performance in laboratory
Acceptable performance in laboratory
' Above the maximum VOS T boiling point of 100°C; included in the testing because compounds in the range of 100-132°C are frequently tested by the VOS T
method.
2 Boiling temperature at 730 mm Hg.
3 Below the common lower temperature limit of30°C usually used for VOST.
4 Boiling temperature at 750 mm Hg.
-------
simultaneous samples are taken and ana-
lyzed. In this field test, known concentra-
tions of volatile halogenated analytes were
introduced into the VOST trains and into
the modified VOST trains, using the same
certified cylinder as a source of the com-
pounds. Using the same dynamic spiking
procedures, multiple collocated simulta-
neous samples were taken using both
methods at the same source.
Bias is any systematic positive or nega-
tive difference between the measured val-
ues and the true value of a sample. Bias
may result from analytical interferences,
errors in calibration, or inefficiencies in
the collection of the analyte. When the
bias of the method is determined for a
given analyte, a correction for the bias
may be made. The EPA Method 301 al-
lows for this correction within a range of
90-110% when a candidate method is be-
ing compared to a reference method. Bias
values outside this range may be grounds
for rejecting the candidate method.
Precision is the variability in the data
from the entire measurement system (both
sampling and analysis) as determined from
multiple or collocated sampling trains. Fol-
lowing the EPA Method 301 procedures,
two paired sampling trains are used to
determine the precision of the entire sys-
tem. Use of quadruple (Quad) trains with
four collocated sampling probes allows
operation of two spiked trains and two
unspiked trains. The EPA Method 301 re-
quires that the precision of the candidate
method not be significantly different from
the reference method in order for the can-
didate method to be considered accept-
able. To determine bias and precision in
the field, a total of at least 16 samples
using quadruple collocated sampling trains
was collected using both VOST and modi-
fied VOST.
Statistical analysis of the results from
the field studies with dynamic spiking show
that the two methods are at least equiva-
lent, with the modified VOST methodol-
ogy frequently being somewhat superior
in compound recovery and precision.
Procedure
Sampling was performed by withdraw-
ing stack gas from a single port in the
stack through a Quad probe, then direct-
ing the sampled gas simultaneously to
four similar VOST or modified VOST sam-
pling trains. The Quad probe contains four
similar heated probes that can be inserted
into the stack as one unit. The front end
of the Quad probe was positioned in the
center of the stack and remained in that
location during each day of testing; no
traverse of the stack was performed with
the Quad probe. The true concentration
of the components of the stack gas was
of no interest to this program as long as
any quantities of the compounds of inter-
est were equal for each train, so travers-
ing was not required. For modified VOST
sampling, two of the trains of each Quad
run were dynamically spiked (Trains A
and B) and two were unspiked (Trains C
and D). The configuration of the modified
VOST sampling train is shown in Figure 1.
Dynamic spiking of candidate VOST
compounds into the modified VOST train
Glass
wool
filter
Three way
glass/teflon
valve
Spiking gas in
To bubble
flow meter
Stack
Heated glass
lined probe
( j 1/4" Teflon
v--} tube
Condensate
trap
Meter box
Sampling module
Figure 1. Modified VOST sampling train.
-------
was performed using a compressed gas
cylinder containing the volatile halogenated
organic compounds. The same cylinder
was used for spiking both the VOST train
and the modified VOST train, with the
composition and concentration of the vola-
tile halogenated organic compounds in the
cylinder established by independent labo-
ratory tests. The boiling point of 100°C
cited in Method 0030 was extended to
approximately 135°C to allow the inclu-
sion of chlorobenzene, a compound fre-
quently determined by the VOST method-
ology. The Modified VOST spiking appa-
ratus shown in Figure 2, was used for
dynamic spiking. In the field, the spiking
gas was allowed to flow through the spik-
ing apparatus for two hours before direct-
ing the flow to the sampling trains to mini-
mize any adsorptive losses during actual
Quad sampling runs. A total of ten (10)
Quad sampling runs for both VOST and
modified VOST was performed in the field
to provide adequate samples for statisti-
cal comparison of the two methods and to
allow backup samples in the event that
any samples became invalid due to break-
age or loss of data during analysis.
Analysis of field samples was performed
according to SW-846 Method 5041
(VOST), analyzing each tube individually
to establish distribution of compounds. The
three tubes associated with each modified
VOST sample set were analyzed in the
following order: Anasorb®, Tenax® #2,
Tenax®#1. No condensate was observed
in the collection of samples.
Results and Discussion
Mean recoveries of the volatile haloge-
nated organic compounds using the modi-
fied VOST sampling method are shown in
Table 2. No background corrections were
made because background levels of the
analytes were less than 5 times the Method
Detection Limit. Methyl chloride exhibited
erratic recoveries, far above 100% in most
analyses. This erratic behavior upon analy-
sis had been observed in laboratory stud-
ies, in the first field test of the VOST
method, and in this field test of both the
VOST and modified VOST methods. The
compound is apparently being formed dur-
ing sampling and analysis.
Surrogate compound recoveries from
the Tenax® tubes were uniformly high, so
correction for surrogate recoveries was
not made. The same surrogate compounds
were used to spike the Anasorb® tubes
prior to analysis, and recovery of the less
volatile surrogate compounds from
Anasorb® is poor; correction of Anasorb®
results for surrogate recovery would dis-
tort the analytical results.
Examination of the observed distribution
of analytes among the three sorbent tubes
of the modified VOST sampling train dem-
onstrates the necessity of a second Tenax®
tube in the sampling train. Methyl chloride,
vinyl chloride, and ethyl chloride are found
primarily on the back (Anasorb®) tube. The
following compounds are found primarily
(>90%) on the front Tenax® tube: ethylidene
dichloride, chlorobenzene, chloroform, pro-
pylene dichloride, methyl chloroform, car-
bon tetrachloride, ethylene dichloride, tri-
chloroethene, cis-1,3-dichloropropene,
trans-1,3-dichloropropene, 1,1,2-trichloro-
Regulator
(temperature controlled)
Certified
standard
(spiking
compounds)
Chassis (temperature controlled)
To Train A •
-^- To Train B
Fine metering
valves
(flow control)
1/4" teflon line
(heat traced and
insulated)
• ••n Cm a tin a
Temperature
display
Variable transformers
(to provide temperature control)
Thermocouple selector
Figure 2. Modified VOS T spiking apparatus.
-------
Table 2. Mean Recoveries of Volatile Halogenated Organic Compounds, Using the Modified VOST Method
Compound
Mean
Recovery
(Percent)1
Percent
Relative
Standard
Deviation
methyl chloride (chloromethane)
167.5
56.4
methyl bromide (bromomethane)
ethyl chloride (chloroethane)
vinyl chloride
vinyl bromide
allyl chloride (3-chloropropene)
45.7
45.3
44.2
38.0
26.0
46.7
30.0
24.2
22.5
21.1
' The shaded area indicates recoveries from 50 to 150 percent with percent relative standard deviation of 50 or less.
ethane, tetrachloroethene, methyl iodide,
allyl chloride, ethylene dibromide, and chlo-
roprene. Other analytes were distributed
among the three tubes.
Comparative results for the VOST and
modified VOST are shown in Table 3. In
most instances, recovery for modified
VOST is equivalent to or better than re-
covery for VOST, and precision for modi-
fied VOST is equivalent to or better than
precision for VOST.
Conclusions and
Recommendations
Based on the field evaluation of the
modified VOST method as compared to
the VOST method, the following conclu-
sions may be drawn:
• On the basis of a compound-by-com-
pound comparison of recovery and
precision, the two methods are com-
parable and, where they differ, the
modified VOST usually shows better
performance with respect to both re-
covery and precision.
The only compound which showed
acceptable performance in the VOST
but not in the modified VOST is me-
thyl bromide. However, the range in
both the recovery and precision is
sufficiently large that the differences
between the results for the VOST and
the modified VOST cannot be consid-
ered statistically significant.
Methyl chloride (chloromethane) has
shown erratic performance, with ex-
tremely high recoveries and poor pre-
cision in several laboratory studies, in
two field tests using VOST, and in
the field test using the modified VOST
methodology. Methyl chloride is not
an appropriate analyte for any form
of the VOST method because quanti-
tative calculations are unreliable for
this analyte. Even qualitative informa-
tion (i.e., presence or absence) is sus-
pect because methyl chloride may be
formed on the sorbent tubes from de-
composition of other halogenated com-
pounds.
The following compounds were not
sampled and analyzed successfully in
this field evaluation using either VOST
or modified VOST: methyl chloride
(chloromethane), vinyl chloride, ethyl
chloride (chloroethane), allyl chloride
(3-chloropropene), and vinyl bromide.
Other than methyl chloride, all of the
analytes which show poor perfor-
mance contain a double bond. Appli-
cation of VOST or modified VOST
methodology for accurate quantitative
determination of analytes containing
-------
Table 3. Results of VOSTand Modified VOST: Recovery and Precision
Modified VOST
VOST
Compound
methyl chloride (chloromethane)
ethylidene dichloride (1,1-dichloroethane)
chlorobenzene
vinyl chloride
vinylidene chloride (1,1-dichloroethene)
chloroform
propylene dichloride (1,2-dichloropropane)
methyl bromide (bromomethane)
ethyl chloride (chloroethane)
methylene chloride
methyl chloroform (1,1,1-trichloroethane)
carbon tetrachloride
ethylene dichloride (1,2-dichloroethane)
trichloroethene
cis- 1 , 3-dichloropropene
trans-1 ,3-dichloropropene
1,1,2-trichloroethane
tetrachloroethene
methyl iodide (iodomethane)
allyl chloride (3-chloropropene)
ethylene dibromide (1,2-dibromoethane)
chloroprene
vinyl bromide
Mean
Percent
Recovery1
167.5
96.2 2
91.62
44.2
96.8 2
98.4 2
149.42
45.7
45.3
120.72
87.1 2
89.3 2
83.2 2
148.72
118.42
75.22
117.32
61. 8 2
89. 02
26.0
108.52
85.8 2
38.0
Percent
Relative
Standard
Deviation
56.4
12.6
13.0
24.2
17.2
20.4
14.0
46.7
30.0
10.9
12.1
12.5
25.1
3.4
21.0
32.6
20.5
8.0
11.9
21.1
23.2
15.3
22.5
Mean
Percent
Recovery1
255.3
86.0 2
84.8 2
37.3
77.8 2
95.3 2
117.72
52. S2
31.4
90.8 2
96.8 2
85. 7 2
78.6 2
124.0 2
83.5 2
47.9
81.42
57.5 2
77.8 2
36.4
81.62
76.4 2
28.4
Percent
Relative
Standard
Deviation
58.1
13.2
27.9
39.5
25.1
14.3
30.0
27.8
37.6
11.7
19.4
13.8
27.7
16.8
16.1
35.0
14.4
12.5
20.4
29.6
31.0
12.3
30.9
' Mean of 6 runs (12 sets of dynamically spiked tubes).
2 Indicates acceptable performance, using the criteria of recovery from 50 to 150 percent, with percent relative standard deviation of 50 or less.
a double bond is questionable. Be-
cause recoveries of these analytes
tend to be low, there tends to be a
low bias in the quantitative determi-
nation.
On the basis of the laboratory and field
efforts conducted for the volatile haloge-
nated organic compounds from Title III of
the Clean Air Act Amendments, the fol-
lowing recommendations are made:
• Careful consideration of the chemical
and physical properties of candidate
VOST or modified VOST analytes can
predict when successful performance
of the sampling or analytical method-
ology is unlikely. For example, analy-
sis of water-soluble compounds will
either be completely unsuccessful
or show a low bias, because
water-soluble compounds cannot be
quantitatively purged from water in
the course of the analysis.
Neither the VOST nor the Modified
VOST is recommended as a sam-
pling or analytical method for methyl
chloride (chloromethane). Even the
accuracy of VOST or modified VOST
as a screening method to establish
the presence of methyl chloride in an
emissions matrix is questionable.
Modification or replacement of the
VOST methodology should be inves-
tigated to provide a sampling and ana-
lytical method feasible for application
to volatile, polar, water-soluble ana-
lytes.
If surrogate compound recoveries are
90% or above, no corrections for sur-
rogate recoveries should be made in
either VOST or modified VOST.
-------
James F. McGaughey, Joan T. Bursey, and Raymond G. Merrill are with Radian
Corporation, Research Triangle Park, NC 27709.
Merrill D. Jackson is the EPA Project Officer (see below).
The complete report, entitled "Field Test of a Generic Method for Halogenated
Hydrocarbons: A VOST Test at a Chemical Manufacturing Facility Using a
Modified VOST Sampling Method," (Order No. PB95-142055; Cost: $27.00;
subject to change) will be available only from
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at
Atmospheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
United States
Environmental Protection Agency
Center for Environmental Research Information
Cincinnati, OH 45268
Official Business
Penalty for Private Use $300
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
EPA/600/SR-94/130
------- |