United States
Environmental Protection
Agency
National Exposure
Research Laboratory
Research Triangle Park, NC 27711
Research and Development
EPA/600/SR-97/076 August 1997
vxEPA Project Summary
Field Evaluation (First) of VOST
and SemiVOST Methods for
Selected CAAA Organic
Compounds at a Coal-Fired
Power Plant
Joan Bursey, James F. McGaughey, Raymond G. Merrill
Abstract
Laboratory and field studies for vola-
tile organic sampling trains (VOST) and
semivolatile organic sampling trains
(SemiVOST) have been performed to
evaluate the performance of halogenated
volatile and semivolatile organic analytes
from Title III of the Clean Air Act Amend-
ments (CAAA) of 1990. Laboratory tests
were conducted to determine if candi-
date nonhalogenated CAAA analytes and
pesticides could be analyzed by gas
chromatography/mass spectrometry
(GC/MS). Organic compounds that could
not be analyzed successfully were elimi-
nated from further study. Gaseous and
liquid dynamic spiking experiments were
performed in the laboratory using the
candidate sampling trains and the re-
maining analytes. A method evaluation
study was subsequently conducted in
the field. Three spiking schemes were
employed in the field evaluation study:
Volatile organic compounds were spiked
into VOST trains as a gaseous spike,
and semivolatile organic compounds
were spiked into SemiVOST trains as a
mixture of either Acid/Neutral com-
pounds or Base/Neutral compounds.
These two mixtures were evaluated in
separate sampling runs to avoid losses
due to known acid/base chemical reac-
tions. This field evaluation study was
performed at a clean source with low
moisture levels and minimal background
levels of organic compounds. Ten qua-
druple sampling (each) runs were
performed for VOST, Acid/Neutral
SemiVOST, and Base/Neutral SemiVOST.
Each quadruple run used four col located
sampling probes into four similar sam-
pling trains, with two spiked trains and
two unspiked trains. Statistical analysis
of the results was performed according
to the guidelines of EPA Method 301.
This Project Summary was developed
by the National Exposure Research
Laboratory's Air Measurements Re-
search Division, Research 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 informa-
tion at back).
Introduction
After laboratory and field studies were
performed to evaluate the performance of
halogenated volatile and semivolatile or-
ganic analytes from Title III of the Clean Air
Act Amendments (CAAA) of 1990 in
VOST and SemiVOST methodology,
nonhalogenated CAAA organic analytes and
pesticides were similarly evaluated. Labo-
ratory studies were performed to deter-
mine if candidate analytes could be ana-
lyzed by GC/MS because both VOST and
SemiVOST require GC/MS analysis. The
analytical methodology used for the VOST
was SW-846 Method 5041; SW-846 Method
8270 was used for the SemiVOST.
When the 19 nonhalogenated candidate
VOST analytes from the CAAA list were
spiked onto VOST tubes and analyzed by
thermal desorption/purge and trap/GC/MS,
candidate VOST analytes with poor analyti-
cal response were eliminated from further
testing. The five remaining candidate VOST
analytes were supplied in a certified gas
cylinder by a commercial vendor for use in
laboratory and field dynamic spiking ex-
periments.
-------
The semivolatile nonhalogenated organic
compounds and pesticides selected from
the CAAA list were combined into multi-
component solutions and the solutions were
analyzed by GC/MS according to the pro-
cedures of EPA Method 8270. Several dif-
ferent combinations of analytes were ana-
lyzed, and several analytes were eliminated
from further testing because of chemical
incompatibilities in solution and lack of solu-
bility in methylene chloride at the levels
desired for a spiking solution. Semivolatile
organic analytes were ultimately combined
into three individual solutions: Acid, Base,
and Neutral. Methylene chloride solutions
of combined Acid and Neutral compounds
or Base and Neutral compounds were pre-
pared for use in SemiVOST evaluation pro-
cedures.
Laboratory dynamic spiking studies
were performed to evaluate VOST and
SemiVOST method precision and bias
under controlled conditions. Laboratory
sampling conditions included clean labora-
tory air sampled with dual VOST or
SemiVOST trains. The results of the labo-
ratory dynamic spiking studies were used
to predict the behavior of the volatile and
semivolatile organic compounds under field
test conditions, with results from the pre-
liminary laboratory study used to direct the
spiking strategy for the first field test.
Procedure
A coal-fired power plant that does not
routinely emit high levels of the hazardous
air pollutants (HAPs) was selected as a
test site for a field evaluation of the VOST
and SemiVOST methods applied to
nonhalogenated volatile and semivolatile
organic compounds. The absence of high
levels of volatile and semivolatile organic
compounds in the background matrix of the
stationary source was established during a
pretest survey. Field testing was conducted
to assess the effect of sampling a combus-
tion matrix (stack gas).
Field evaluation of both VOST and
SemiVOST was accomplished by sampling
and analysis following the EPA methods
(VOST: sampling, EPA Method 0030; analy-
sis, EPA Method 5041; SemiVOST: sam-
pling, EPA Method 0010; sample prepara-
tion, EPA Proposed Method 3542; analy-
sis, EPA Method 8270) except for the use
of four collocated sampling probes (Quad
probe). This modification of the sampling
trains was necessary to conform to the
requirements of EPA Method 301 for deter-
mining the bias and precision of test meth-
ods. EPA methods were applied exactly as
written, with no deviation from the written
methodology.
Generation of data for the validation of
test methodology according to EPA
Method 301 allows the bias (systematic er-
ror) and precision (reproducibility of mea-
surement) to be determined in a statisti-
cally valid manner.
To determine bias and precision in the
field, ten quad sampling runs (40 sampling
trains) were performed for each of the
VOST, Acid/Neutral SemiVOST, and Base/
Neutral SemiVOST. The additional qua-
druple runs were performed in the field to
allow for possible loss of samples or invali-
dation of a run for any reason.
Analytical Results
Laboratory studies using GC/MS dem-
onstrated that 5 of 19 volatile organic com-
pounds could be used for field testing. The
volatile organic compounds that were not
tested included acetonitrile, acrylonitrile, 1,1-
dimethylhydrazine, 1,4-dioxane, ethyl aery-
late, ethylene imine, methyl ethyl ketone,
methyl isobutyl ketone, methyl methacry-
late, methyl fe/t-butyl ether, propylene ox-
ide, 1,2-propylene imine, triethylamine, and
vinyl acetate. The semivolatile organic com-
pounds that were not tested in the field
included hydroquinone, 1,4-phenylenedi-
amine, captan, 1,3-propane sultone, eth-
ylene thiourea, hexamethylene-1,6-
diisocyanate, maleic anhydride, quinone,
catechol, and 2,4-toluenediamine.
Analytical field data were evaluated ac-
cording to the statistical procedures of EPA
Method 301, and according to the criteria
of the EPA Handbook for Quality Assur-
ance/Quality Control (QA/QC) Procedures
for Hazardous Waste Incineration (EPA/
625/6-89/023, January, 1990). The com-
pounds tested in the field study are listed in
Table 1. Their recoveries and whether they
meet the criteria of the two statistical pro-
cedures are also shown in Table 1.
Discussion
Based on the work performed in the labo-
ratory and the field evaluation of the VOST
and SemiVOST method for nonhalogenated
analytes and pesticides listed in Title III of
the Clean Air Act Amendments of 1990,
the following observations may be made:
• Where the test atmosphere is clean
dry laboratory air, performance in labo-
ratory testing is a reasonably reliable
indicator of field performance. How-
ever, there are a few examples where
marginal performance in the labora-
tory precedes acceptable field perfor-
mance as well as several instances
where good laboratory performance
precedes unacceptable field perfor-
mance.
• The two modes of statistical calcula-
tion (EPA Method 301 and EPA QA
Handbook) generally produce com-
parable results. An analyte which
meets one set of acceptance criteria
usually meets the other set of accep-
tance criteria as well, or the analyte
will fail to meet both sets of criteria. An
exception to this observation occurs
when the Method 301 correction factor
is in the range of 1.30 to approxi-
mately 2.00: Analytes with a Method
301 correction factor in this range meet
QA Handbook acceptance criteria al-
though the Method 301 correction fac-
tor is not acceptable. An analyte with a
Method 301 correction factor lower than
0.70 can also meet EPA Handbook
acceptance criteria.
Several of the nonhalogenated organic
analytes demonstrated solubility prob-
lems in the course of preparation of a
stock solution in methylene chloride.
These analytes are not appropriate for
the SemiVOST methodology because
quantitative extraction by methylene
chloride cannot be ensured. The utility
of SemiVOST even as a screening
method for the presence of these
analytes is questionable.
Reactive nonhalogenated semivolatile
organic analytes which demonstrated
reaction with other components of the
stock or spiking solution are question-
able as appropriate for the SemiVOST
method because the extent of reaction
cannot be determined; therefore, the
accuracy of the quantitative values is
suspect. The utility of SemiVOST even
as a screening method for the pres-
ence of these analytes is question-
able.
The VOST is appropriate as a sam-
pling/analytical method only for those
CAAA volatile organic analytes which
can be analyzed by GC/MS. The VOST
is not appropriate even as a screening
method for those volatile organic com-
pounds which cannot be analyzed at
all by the methodology.
Quantitative analysis of very polar
semivolatile compounds is difficult with
the SemiVOST method.
The presence of particulate matter in
the SemiVOST method causes reten-
tion of significant amounts of polar
semivolatile organic analytes on the
filter. Even relatively volatile polar
analytes such as aniline are retained
on the heated filter of the sampling
train. Significant retention of relatively
volatile polar analytes on the filter ap-
pears to occur for nitrogen-containing
analytes such as aniline or quinoline;
relatively volatile oxygenated analytes
such as phenol do not appear to be
retained on the filter.
-------
Table 1. Experimental Results Summary
Analyte Percent Recovery ± RSD
Meets Method 301
Criteria in Field?1
Meets QA Handbook
Criteria in Field?2
VOST Analytes
benzene
carbon disulfide
n-hexane
toluene
2,2,4-trimethylpentane
106 ±25.6
63.1 ±18.3
79.2 ±22.6
77.9 ±17.5
63.1 ±18.3
Yes
No; correction factor 1 .47
Yes
Yes
No; correction factor 1 .48
Yes
Yes
Yes
Yes
Yes
SemiVOST Analytes
acetophenone
4-aminobiphenyl
aniline
o-anisidine
benzidine
biphenyl
bis(2-ethylhexyl) phthalate
cumene
dibenzofuran
di-n-butyl phthalate
N,N-diethylaniline
N,N-dimethylaniline
dimethylaminoazobenzene
3, 3'-dimethoxy benzidine
dimethyl phthalate
4,6-dinitro-o-cresol
2,4-dinitrophenol
2,4-dinitrotoluene
3,3'-dimethylbenzidine
isophorone
4,4'-methylene
4,4'-methylene bis(chloroaniline)
4,4'-methylenedianiline
o-cresol
m-/jj-cresol3
96±12
50 ±41
70 ±24
39 ±39
65 ± 1 1 9
103±12
48 ±23
88±11
100±12
46 ±54
95 ±19
67 ±24
31 ±51
37 ±38
82±17
122±14
111 ±31
109±12
92 ±44
93 ±12
89 ±36
89 ±36
6 ±73
90±15
69±14
Yes
Yes
No; correction factor 1 .37
No; correction factor 2.53
No
Yes
No; correction factor 2.16
Yes
Yes
No; correction factor 2.44
Yes
No; correction factor 1 .42
No; correction factor 3.26
No
No; correction factor 1 .31
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
No; correction factor 1 .51
Yes
Yes
Yes
No; recovery 39.4%
No
Yes
No; recovery 48%
Yes
Yes
No; recovery 46%
Yes
Yes
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
'Acceptance criteria: correction factor between 0.70 and 1.30, precision <50% relative standard deviation.
Acceptance criteria: recovery between 50% and 150%, precision <50% relative standard deviation.
3These compounds are listed together in the table because the coelute and are not separated in analysis.
(continued)
-------
Table 1. Continued
Analyte
naphthalene
nitrobenzene
4-nitrobiphenyl
4-nitrophenol
N-nitrosodimethylamine
N-nitrosomorpholine
phenol
g-toluidine
trifluralin
ethylbenzene
styrene
g-xylene
m-/j>xylene3
2-acetylaminofluorene
caprolactam
carbaryl
chlordane
DDE
dichlorvos
heptachlor
hexamethylphosphoramide
lindane
methoxychlor
parathion
propoxur
quinoline
styrene oxide
1 ,4-dioxane
ethyl carbamate
methyl isobutyl ketone
phthalic anhydride
Percent Recovery ± RSD
96±11
109±12
102±14
114 ±31
117± 13
116±12
89 ±8
56 ±30
27 ±41
89±12
84 ±10
85 ±11
79±12
1 47 ± 23
114±12
99 ±14
142±16
102±15
101 ±18
103±12
14±118
104±12
73 ±19
89 ±28
123±12
80 ±19
0.5 ±1481
87 ±11
103±14
112±11
5.3 ±144
Meets Method 301
Criteria in Field?1
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No; correction factor 1 .77
No; correction factor 3.81
Yes
Yes
Yes
Yes
No; correction factor 0.68
Yes
Yes
Yes
Yes
Yes
Yes
No; correction factor 6.93
Yes
No; correction factor 1 .40
No; correction factor 1 .31
Yes
Yes
No
Yes
Yes
Yes
No
Meets QA Handbook
Criteria in Field?2
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
No
'Acceptance criteria: correction factor between 0.70 and 1.30, precision <50% relative standard deviation.
Acceptance criteria: recovery between 50% and 150%, precision <50% relative standard deviation.
3These compounds are listed together in the table because they coelute and are not separated in analysis.
-------
• The least volatile semivolatile organic
analytes are distributed mostly on the
heated filter of the sampling train. For
some types of particulate matter, it may
not be possible to recover organic
analytes quantitatively using methyl-
ene chloride extraction techniques.
• Analytes tested using both VOST and
SemiVOST include 1,4-dioxane and
methyl isobutyl ketone. For these two
polar volatile organic analytes, VOST
was unsuccessful; SemiVOST was
successful.
Conclusions and
Recommendations
Successful operation of the SemiVOST
methodology, including validation, depends
on several considerations which can be
evaluated only by testing the operation of
the entire methodology:
• The chemical nature of the compounds
of interest must be considered. EPA
Method 0010 states that the range of
applicability is all organic compounds
with a boiling point above 100°C. This
statement provides only a general
guideline that excludes organic com-
pounds with a boiling point below
100°C. The statement does not guar-
antee that any organic compound in
the appropriate boiling point range will
function successfully in the SemiVOST
methodology. Polar compounds are
more reactive, more difficult to extract,
and more difficult to analyze than non-
polar compounds. Any of these factors
may make it impossible to sample or
analyze the compound using the
SemiVOST methodology.
• The nature of the background station-
ary source matrix is an important con-
sideration in determining the success-
ful function of the SemiVOST method.
Highly acidic stationary source matri-
ces affect compound recoveries. Ex-
tremely wet source matrices introduce
difficulties in sample extraction. High
levels of particulate material can retain
organic compounds. All of these fac-
tors of the stationary source demon-
strate why successful laboratory per-
formance of a given compound cannot
guarantee that the compound will per-
form successfully in the field. How-
ever, inability to analyze the compound
by GC/MS in the laboratory does guar-
antee that performance of the sam-
pling and analytical methodology can-
not even be evaluated.
• Neutral, non-polar compounds as a
class tend to demonstrate the best per-
formance in the SemiVOST methodol-
ogy. Acidic or basic compounds are
reactive. These compounds tend to re-
act with each other or with constitu-
ents of the source matrix, thus limiting
or eliminating their recovery using the
SemiVOST method.
The successful performance of the VOST
method also demands evaluation of the
entire sampling and analytical method:
• The chemical nature of the compounds
of interest must be considered. EPA
Method 0030 states that the range of
applicability is all organic compounds
with a boiling point below 100°C. This
statement provides only a general
guideline that excludes organic com-
pounds with a boiling point above
100°C, although in practical applica-
tion of the method this boiling point
limitation is frequently extended to
132°C. The statement does not guar-
antee that any organic compound in
the appropriate boiling point range will
function successfully in the VOST
methodology. Polar compounds are
more reactive, more difficult to desorb
and purge, and more difficult to ana-
lyze than non-polar compounds. Any
of these factors may make it impos-
sible to sample or analyze the com-
pound using the VOST methodology.
• The nature of the background station-
ary source matrix is an important con-
sideration in determining the success-
ful function of the VOST method. Highly
acidic stationary source matrices af-
fect compound recoveries. Extremely
wet source matrices introduce difficul-
ties in sample desorption. Reactive con-
stituents of the source matrix can react
with the sorbent or the sorbed com-
pounds. All of these factors of the sta-
tionary source demonstrate why suc-
cessful laboratory performance of a
given compound cannot guarantee that
the compound will perform success-
fully in the field. However, inability to
analyze the compound by GC/MS in
the laboratory does guarantee that per-
formance of the sampling and analyti-
cal methodology cannot even be evalu-
ated.
• Neutral, non-polar compounds as a
class tend to demonstrate the best per-
formance in the VOST methodology.
Acidic or basic compounds are reac-
tive. These compounds tend to react
with each other or with constituents of
the source matrix, thus limiting or elimi-
nating their recovery using the VOST
method.
On the basis of the laboratory and field
effort conducted thus far for the nonha-
logenated volatile and semivolatile organic
analytes listed in Title III of the Clean Air
Act Amendments of 1990, the following
recommendations are made:
• Faced with a candidate analyte for ei-
ther VOST or SemiVOST for which no
method validation information is avail-
able, laboratory experimentation can
aid in establishing the validity of as-
signing the analyte to a particular meth-
odology.
• Careful consideration of the chemical
properties (acid, base, reactivity, etc.)
of candidate analytes is essential in
predicting success or failure in the ap-
plication of VOST or SemiVOST sam-
pling and analytical methodology.
• A complete and accurate formulation
of the complete SemiVOST methodol-
ogy as it is presently used should be
written. Method 0010 is applicable for
sampling procedures but sample prepa-
ration procedures contained in
Method 0010 have been superseded
by Proposed Method 3542. The com-
plete Method 8270 is not applicable as
the analytical methodology; only cer-
tain sections are applicable. Also,
Method 8270 surrogate compounds are
usually used with SemiVOST samples.
• Further study of the VOST analytical
methodology (Method 5041) should be
performed. Internal standards and
surrogate compounds used in
Method 8240 have been used in
Method 5041 for the sake of con-
sistency between the methods. How-
ever, £-bromofluorobenzene is not
an appropriate surrogate compound
for the VOST methodology since the
boiling point for this compound is
150°C, well above the upper limit of
100°C. A better choice of surrogate
compound should be made. Also, no
criteria are available in the method for
acceptability of surrogate recoveries in
the VOST beyond the general guide-
lines of the QA Handbook of recovery
between 50 and 150%, with precision
of 50% relative standard deviation.
Disclaimer
The U.S. Environmental Protection
Agency through its Office of Research and
Development funded and managed the re-
search described here under 68-Do-0022
to Eastern Research Group, Inc. It has
been subjected to the Agency's peer and
administrative review and has been ap-
proved for publication as an EPA docu-
ment. Mention of trade names or com-
mercial products does not constitute en-
dorsement or recommendation for use.
-------
Acknowledgments
Under EPA Contract No. 68-Do-0022,
Eastern Research Group, Inc. prepared
this report. The Project Manager was
Raymond G. Merrill; the Principal Investi-
gator was Joan T. Bursey. We wish to
acknowledge the contributions of the fol-
lowing individuals to the success of this
program: Amy Bederka, Mike Bryant, Tom
Buedel, Kelly Dowler, Sam Foster, Danny
Harrison, Judi McCartney, Linh Nguyen,
Phyllis O'Hara, Mark Owens, Tiffany
Sorrell, Jim Southerland, Julie Swift, and
Donna Tedder. Helpful discussions on sta-
tistical calculations and interpretation of
statistics with Dr. Joseph Knoll and Mr.
Merrill Jackson are gratefully acknowl-
edged.
-------
Joan Bursey, James F. McGaughey, and Raymond G. Merrill are with Eastern
Research Group, Inc., Morrisville, NC 27560
Merrill D. Jackson is the EPA Project Officer (see below).
The complete report, entitled "Field Evaluation (First) of VOST and SemiVOST
Methods for Selected CAAA Organic Compounds at a Coal-Fired Power Plant,"
(Order No. PB97-196117; Cost: $67.00, 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:
Air Measurements Research Division
National Exposure Research Laboratory
U. S. Environmental Protection Agency
Research Triangle Park, NC 27711
United States
Environmental Protection Agency
Center for Environmental Research Information (G-72)
Cincinnati, OH 45268
Official Business
Penalty for Private Use $300
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
EPA/600/SR-97/076
------- |