United States
Environmental Protection
Agency
Environmental Monitoring Systems
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S4-83-024 Aug. 1983
Project Summary
Field Validation of EPA Reference
Method 23
Byron B. Bradd
The accuracy and precision of EPA
Reference Method 23 for use in the
field was evaluated at a trichloroethylene
degreasing facility and an ethylene di-
chloride manufacturing facility. The
method consists of a procedure for
obtaining an integrated sample followed
by gas chromatographic analysis. This
study identified a number of conditions,
such as sunlight and sample volume,
that may affect the performance of the
method in the field. Results obtained
in the presence and absence of these
conditions were compared. The most
significant variable was the use of
Tedlar (E I. Dupont de Nemours &
Company, Inc.) polyvinyl bags versus
mylar polyvinyl bags as sampling con-
tainers, a choice left open in the method
as originally defined.
A paired sampling technique was
employed to obtain identical chemical
samples from two sampling trains
operating simultaneously. Certain re-
sults fell outside the range of accept-
able precision, due to the use of mylar
bags and to rigid sample containers
that were prone to leakage. The method
was modified to specify that only Tedlar
bags be used for sampling; to incorpo-
rate an on-site technique for early leak
detection; and to discontinue pro-
cedures for moisture removal. With
these modifications, EPA Reference
Method 23 was fully validated as ac-
ceptable for field use.
This Project Summary was developed
by EPA's Environmental Monitoring
Systems Laboratory, Research Triangle
Park NC, to announce key findings of
the research project that is fully docu-
mented in a separate report of the
same title (see Project Report ordering
information at back).
Introduction
On June 11,1980, under Section III of
the Clean Air Act the U.S. Environmental
Protection Agency (EPA) proposed stand-
ards limiting the emission of chlorinated
hydrocarbons from solvent degreasers,
after determining that these substances
presented a potential risk to public health
and welfare. A national emission standard
was proposed which included a method
for measuring chlorinated hydrocarbons,
designated EPA Reference Method 23.
The Quality Assurance Division of the
Environmental Monitoring Systems Labora-
tory, Research Triangle Park, North Carolina
has a program to evaluate and standardize
such EPA source test methods.
Method 23 for the Determination of
Halogenated Organics from Stationary
Sources specifies integrated bag samples
of a stack gas containing one or more
halogenated organic compounds, which
are subjected to gas chromatographic (GQ
analysis using a flame ionization detector.
The method is applicable to the measure-
ment of halogenated organics such as
carbon tetrachloride, ethylene dichloride
(EDQ, perchloroethylene, trichloroethylene
(TCE), methylene chloride, methyl chloro-
form, and trichlorotrifluoroethane in stack
gases from specified processes The method
is not applicable to paniculate matter con-
taining organics.
The goal in this evaluation was to test
the accuracy of the Reference Method 23
apparatus and procedures. In this study
EDC and TCE were measured by varying
the field conditions of sunlight sample
volume, storage temperature and storage
time before analysis. Perhaps the most
significant variable being tested concerned
the use of polyvinyl bags as sampling
containers. The option of using either
-------
Tedlar or aluminized mylar bags was left
open in the Method as originally written.
At the high concentrations encountered
in source sampling, major problems result
from sample contamination due to prior
use of the bags. Although in the past
experience of this laboratory Tedlar bags
have been able to preserve sample con-
centrations better than mylar, for this
evaluation program we wished to establish
definitive results.
A potential difficulty with Tedlar bags is
their capacity to retain " memory traces" of
EDC, which, under conditions of simulated
sunlight, results in contamination peaks.
In order to avoid this problem, new bags
were used for each experiment More
positive results were reported recently, in
which Tedlar bags were successfully em-
ployed under laboratory conditions for a
number of halogenated hydrocarbon gas
mixtures.
Paired sampling trains were used to test
the utility of Tedlar and mylar bags for
preserving field samples and to vary the
conditions mentioned above. Results ob-
tained under strict adherence to Method
23 were compared to those where dif-
ferent parameters were varied, using both
Tedlar and mylar bags. The concentration
figures were then verified with the use of
audit standards. In addition to validation
of Method 23, the most significant out-
come of the investigation regarded the
reliability of Tedlar bags. By contrast
mylar bags produced unacceptable results;
measured concentrations were consistent-
ly lower when compared to Tedlar bag
samples.
Experimental Method
The field evaluations were conducted at
a metal degreasing facility using TCE and
at a plant which manufacturers EDC
Sampling variables were selepted based
on their likelihood of occurring under actual
field conditions. Some variations were
controllable, such as changes in the sam-
pling rate or the type of bag used, whereas
others were subject to chance, such as
ambient temperature, sunlight exposure,
or moisture condensation. The variables
studied in the field were as follows:
1) Sampling and storage temperature,
2) Moisture removal,
3) Tedlar or mylar bags,
4) Sample volume,
5) Exposure to sunlight
6) Storage time before analysis.
All gas samples were collected in pairs,
using identical sampling trains. One sam-
pling train was always operated in strict
adherence to Method 23, i.e., protected
from heat shielded from sunlight during
removal of the bag, sampled to obtain a full
bag, and analyzed within 30 to 60 min.
The second sampling train was varied in
one or more of the above conditions. At
each site both high level and low level
concentrations were sampled and analyzed
by GC with flame ionization detection.
Hydrocarbon-free nitrogen was injected
into the GC system between analyses to
demonstrate that no contamination re-
mained from previous samples. The stand-
ards shown betow were analyzed frequent-
ly to maintain a constant calibration.
The concentrations of the working stand-
ards were as follows:
EDC (ppm) TCE (ppm)
1024 49.6
51.7 9.18
9.66
The accuracy and linearity of the GC
analysis used in this program were demon-
strated through the use of EPA audit
samples. The concentration of the audit
samples were not disclosed until after
they were analyzed during the field test
One audit cylinder was also sampled with
the Tedlar bag and compared against the
audit value to determine sampling system
loses.
Possible effects of excessive moisture
on source samples were determined by
injecting distilled water into Tedlar bags
that contained 25 ppm EDC and 50 ppm
TCE and measuring the concentrations
after water dispersion. Also, silica gel was
evaluated as a moisture preconditioner in
a 35-1 sample of EDC
Results and Discussion
The results of the Method 23 evaluation
are shown in Tables 1 and 2. Source
concentrations sampled in bags were not
significantly affected by variations in heat
light and sample volume. The difference
in the analytical results fell within the
magnitude of precision for the Method.
When mylar sample bags were used,
consistently lower values were obtained.
These variations far exceeded reasonable
levels of precision (±10%) for Method 23.
The decay of mylar bag samples was
significant as compared to the minimal
decay of Tedlar bag samples.
The values of field analysis of audit
samples were compared to actual audit
values, with close agreement Results
were as follows:
TCE
Audit
value
14.9
566
Field
analysis
15.2
517
%diff.
+2.0
-8.7
EDC
Audit
value
" g
438
Field
analysis
9.75
448
%diff.
+6.0
+2.3
Another factor considered in this evalua-
tion was the possible effect of moisture
content in the sample. The effect of
condensing moisture in the sampling train
could not be demonstrated since the only
Table 1.
TCE Field Evaluation Results
Test no.
1
2
3
4
5
6
7
8
Train A (ppm)
280
345
338
369
32
34
43
51
Train B (ppm)
272
195
192
373
33
20
40
52
Variations in Train B
1/2 Volume/sunlight exposure
'/i Volume/mylar bag
Heat/mylar bag
Heat/sunlight exposure
'/2 Volume
Mylar bag
Heat
Sunlight exposure
Percent
variation
-2.8
-43.5
-43.2
+1.1
+3.1
-41.2
-7.0
+2.0
Table 2. EDC Field Evaluation Results
Test no.
4"
5
6
7
9
10
Train A (ppm)
1816
1764
1796
1764
16.1
17.08
Train B (ppm)
1821
1699
1459
1538
16.2
15.48
Variations in Train B
None
Heat/'/2 volume
Mylar/'/2 volume
Heat/mylar bag
Heat
Mylar bag
Percent
variation
+0.3
-3.7
-18.8
-12.8
+0.6
-9.4
"Test nos. 1-3 contained nondetectable levels of EDC
-------
source containing moisture contained non-
detectable levels of EDC However, the
injection of distilled water into bag samples
did not measurably affect the gas concen-
trations.
Silica gel, as a moisture conditioner,
adsorbed a significant amount of the or-
ganics being sampled (about 9596 of the
EDC from a 35-1 sample). Moisture
preconditioning should be achieved with a
coil condenser, and only where necessary
to prevent blockage of the sampling train.
It was determined that significant errors
could be caused by leaks in the rigid
sample containers, the probe, or bag con-
nection. Depending on its size, a leak in
the rigid container could produce results
ranging from a reduction in the sample
volume to loss of the entire sample. Leaks
in the probe or bag connection, on the
other hand, diluted-the sample. An on-site
procedure for early detection of leaks was
devised to eliminate these problems from
the sampling train.
Sample degradation over time was not
determined to be a major problem for
samples contained in Tedlar bags. Re-
gression analysis of the data collected
indicated that TCE samples held in Tedlar
bags degrade at an average rate of approx-
imately 1 % per day whereas samples in
mylar bags degrade at a rate of 2% per day.
Conclusions and
Recommendations
1. EPA Reference Method23 using Ted-
lar bags produced satisfactory pre-
cision; variations ranged from 0.6% to
7.06%. Mylar bags, however, yielded
reduced concentrations ranging from
9% to 43% lower than the Tedlar bag
counterpart Mylar bags are not re-
commended for use in this method.
2. Field variations of low sample volume,
heat, and sunlight exposure did not
significantly affect the results when
Tedlar bags were used.
3. A rigorous leak check procedure
should be performed immediately be-
fore a sampling run.
4. A container that is reliable, light-
weight, leak-proof, and economical
should be developed for integrated
sampling.
5. It is not necessary to remove moisture
from samples to perform Method 23.
Byron B. Bradd is with Scott Environmental Services, Plumsteadville, PA 18949.
Joseph E. Knoll is the EPA Project Officer (see below).
The complete report, entitled "Field Validation of EPA Reference Method 23."
(Order No. PB 83-214 551; Cost: $10.00, subject to change) will be available only
from:
National Technical Information Service
5285 Port Royal Road
Springfield, V'A 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
•&US GOVERNMENT PRINTING OFFICE- 1983-659-017/7143
-------
United States Center for Environmental Research
Environmental Protection Information
Agency C,nC1nnat, OH 45268
Agency
EPA 335
Official Business
Penalty for Private Use $300
RETURN POSTAGE GUARANTEED f,Sc 2 Ki V I R Pi
------- |