600/A^93/167
For presentation at the AWMA/EPA International Symposium, "Measurement
of Toxic and Related Air Pollutants", May 3-7, 1993, Durham NC.
SOURCE APPORTIONMENT OF FINE PARTICLE ORGANICS
AND MUTAGENICITY IN WINTERTIME ROANOKE
Charles W. Lewis and Roy B. Zweidinger
U.S. Environmental Protection Agency
Atmospheric Research and Exposure Assessment Laboratory
Research Triangle Park, NC 27711
Larry D. Claxton
U.S. Environmental Protection Agency
Health Effects Research Laboratory
Research Triangle Park, NC 27711
Donna B. Klinedinst
National Institute of Standards and Technology
Gaithersburg, MD 20899
Sarah H. Warren
Environmental Health Research and Testing
Research Triangle Park, NC 27709
ABSTRACT
During the 1988-1989 winter the U.S. EPA conducted a
comprehensive field study in Roanoke VA as part of its Integrated Air
Cancer Project (IACP). This paper presents results of the source
apportionment of fine particle extractable organic matter (EOM) and
its associated mutagenicity (Salmonella typhimurium TA98 +S9 and
TA98 -S9). The source apportionment methodology is based on multiple
linear regression (MLR) using a variety of tracer species: 14C,
metallic elements and volatile hydrocarbons (VHC) whose ambient
concentrations were measured simultaneously with the EOM and
mutagenicity. The results are compared with those from previous IACP
studies in other locales.
INTRODUCTION
As part of the IACP the U.S. Environmental Protection Agency
has conducted a series of wintertime field studies in U.S. cities to
measure ambient concentrations of fine particle EOM and associated
mutagenicity. Receptor modeling has been employed with these
measurements to determine the quantitative contributions of various
emissions sources to both EOM and mutagenicity. The present work
gives receptor modeling results for the 1988-1989 field study in
Roanoke VA, an airshed whose principal sources of ambient EOM were
anticipated to be woodsmoke, mobile sources and residential distillate
oil combustion (RDOC).
EXPERIMENTAL
Overviews of the Roanoke ambient field sampling program have been
given1'2. The results that follow are from analyses of 12-h fine
particle (0 - 2.5 jim dia) and VHC samples collected simultaneously at
the two primary sites — Morningside Park (residential site) and Civic
-------
Center (roadway site). The analyses followed previous IACP procedures
and generated an ambient data set of the following parameters:
elemental composition of fine aerosol from x-ray fluorescence (XRF)3;
fine EOM from dichloromethane Soxhlet extraction3; TA98 +S9 and TA98 -
S9 mutagenicities of the EOM from plate incorporation bioassays3; 14C
content of the EOM from accelerator mass spectrometry4; and VHCs from
gas chromatography-flame ionization detection5.
MLR RECEPTOR MODELING
The same MLR modeling approach was used as in earlier IACP work.
The measured concentration of the pollutant of interest (i.e, EOM or
mutagenicity) is represented by a sum of individual source
contribution terms, with each term being the product of the measured
concentration of a tracer species for that source and an initially
unknown coefficient that is subsequently determined by an MLR
calculation. A non-zero intercept is allowed for, which can be
regarded as the average contribution of additional sources not
represented in one of the explicitly identified source terms.
In previous IACP work the concentration of fine-particle Pb was
used as a mobile-source tracer. In recent years, however, this has
become less tenable, because of the phaseout of leaded gasoline.
During the preceding IACP study in Boise -- while Pb was still a
satisfactory tracer — the use of any one of several VHC species as
tracers was shown to give estimates for the mobile source contribution
which were virtually identical to those produced with Pb6. For the
Roanoke data however the VHC species were clearly superior to Pb, as
judged from the quality of the MLR fits that could be achieved
(largest r2 value). The unimpressive Pb-Br correlation (r2 = 0.65)
exhibited by the Roanoke data also suggested that Pb was not a
reliable mobile source tracer. Consequently, the VHC species 2-
methylhexane (2MeHx) was used instead of Pb in the present work, as it
produced a slightly better MLR fit than the other VHCs.
Fine particle soil-corrected potassium (K1) has proven to be very
useful as a woodsmoke tracer in previous IACP work. In the planning
stage of the Roanoke project, it was conjectured that K1 might serve
as a tracer of overall residential heating. This was because RDOC as
well as wood combustion were anticipated to be important contributors
to ambient EOM concentrations, and the expected similar diurnal
emission patterns of these two source categories would frustrate their
separate estimations by an MLR technique. As shown below however
RDOC was so small in comparison to wood combustion that K1 retained
the same role for Roanoke as it had in previous IACP work. Ondov et
al.7 report a similarly small estimate of the RDOC contribution,
through use of an enriched isotope of samarium as an intentional
tracer.
RESULTS
Modeling
Multiple linear regression of the measured concentrations of EOM
(/ug m"3) and R^ and R^. mutagenicities (revertants m"3) resulted in the
following equations:
[EOM]; = (40 ± 3)[K']i + ( 656 ± 83 )[2MeHx]i - 0.3 ± 0.7 (1)
r2 =0.92, n = 40;
-------
[1*98+h = (32 ± 6)[K']j + (1960 ± 140)[2MeHx]i + 2.6 ± 1.1 (2)
=0.92, n = 37;
= (21 ± 4)[K']i + (1100 ± 130)[2MeHx]j + 4.2 ± 0.9 (3)
r2 =0.86, n=38.
The units of the tracer concentrations [K1] and [2MeHx] are ^g m"3 and
ppmC, respectively, and
[K1] = [K] - (0.22 ± 0.01)[Fe]. (4)
The numerical coefficient in equation 4 is the average potassium-to-
iron ratio measured in the coarse-particle (2.5 - 10 ^im dia) fraction,
as done previously3.
The cases used in each of the three fits were approximately
equally divided between the Civic Center and Morningside Park sampling
sites, and between day and night. The quality of the fits, as judged
by the r2 values, were similar to or better than those achieved in the
three earlier IACP studies, each of which involved approximately the
same number of cases as in the present study.
Table I shows the average source contributions to the measured
concentrations of EOM and mutagenicity, averaged over both sampling
sites as well as day and night. The source contributions were
calculated by inserting the averages of the measured values of [K1]
and [2MeHx] into equations 1-3. Table I also gives the calculated
mutagenic potency of the EOM associated with each source. The source
potency (revertants (/xg EOM)"1) is simply the ratio of the R to EOM
regression coefficient for that source. For comparison Table I also
includes corresponding results from the three earlier studies (Boise,
Albuquerque and Raleigh). The latter results are from Lewis et al.3,
which gives references to the original work. It is important to note
that the source potency values listed in Table I are derived entirely
from ambient measurements.
Validation by 14C
Since 14C is absent from fossil fuels, its presence in fine
particle atmospheric samples is a direct (non-statistical) indication
of the contribution of contemporary carbon sources, assumed to be
essentially only residential woodburning in the wintertime Roanoke
airshed. The 14C-derived estimate of woodsmoke EOM concentration in a
sample i is given by
[Woodsmoke EOM]; = [EOM]; (fRWC)i (5)
with fRWc being the fraction of residential wood combustion carbon in
the sample. On the other hand the term in equation 1 involving [K'];
is an independent statistical estimate of the same quantity. Thus the
right side of equation 5 and [K1], should be linearly related, with a
slope that is the same as the regression coefficient for [K1] in
equation 1, and with no intercept. Figure 1 shows [EOM];* (fRWC) j values
for all available data (n = 20) recommended for use by Klinedinst et
al4, plotted vs [K1]. The straight line in the figure is the product
of [K1] and its regression coefficient from equation 1, with the
dotted band representing the coefficient's uncertainty.
-------
For the most part the points in Figure 1 cluster around the line,
showing that the potassium-traced portion of EOM is wood combustion-
related, and indicating by inference that the RDOC portion must be
small in comparison. This conclusion is also supported by a Roanoke
TSP emissions inventory specific to the 1988-1989 wintertime period of
the IACP field study which indicates an RDOC emissions rate that is
only a few percent of that for wood combustion8.
CONCLUSIONS
The Roanoke results presented in this paper together with
corresponding results from the previous IACP studies (Table I) show
some very clear consistencies across four geographically scattered
U.S. airsheds: (1) ambient EOM is dominated by woodsmoke, rather than
mobile sources emissions, for these wintertime studies; (2) the
mutagenic potency (Salmonella typhimurium TA98 +S9) of ambient
woodsmoke is approximately 1 revertant per microgram of EOM; (3) the
mutagenic potency of ambient mobile source emissions is about three
times that of woodsmoke. These consistencies span a period of four
years, and depend neither on the type of wood used nor on the choice
of a mobile source tracer (particulate Pb or volatile hydrocarbons).
The use of I4C measurements have served to confirm the MLR-based
apportionment of EOM.
The woodsmoke domination of EOM may not be surprising for
Albuquerque and Raleigh, since the sampling site for both was in a
residential neighborhood. For Boise and Roanoke however the results
come from the combining of measurements at both residential and
roadway sites, and woodsmoke still dominates EOM overall. The choice
of sampling site locations within the cities presumably had little
effect on the potency values that were obtained.
The Roanoke results for TA -98 mutagenicity and potency are the
first known ambient-derived values for these parameters.
ACKNOWLEDGEMENTS
We thank Alan Hoffman for his supervision of the Roanoke field
study, Bob Kellogg for XRF analyses and Gwen Belk for EOM analyses.
DISCLAIMER
The information in this document has been funded wholly or in
part by the United States Environmental Protection Agency. It has
been subjected to Agency review and approved for publication. Mention
of trade names or commercial products does not constitute endorsement
or recommendation for use.
REFERENCES
1. V.R. Highsmith, A.J. Hoffman, R.B. Zweidinger et al., "The IACP:
Overview of the Boise, Idaho, and the Roanoke, Virginia, field
studies," in Proceedings of the 84th Annual Meeting of A&WMA, paper
91-131.1, 1991.
2. R.K. Stevens, A.J. Hoffman, J.D. Baugh et al., "Air quality
measurement in Roanoke, VA in support of the 1988-1989 Integrated Air
Cancer Project," these proceedings, 1993.
3. C.W. Lewis, R.K. Stevens, R.B. Zweidinger et al., "Source
apportionment of mutagenic activity of fine particle organics in
Boise, Idaho," in Proceedings of the 84th Annual Meeting of A&WMA,
paper 91-131.3, 1991.
-------
4. D.B. Klinedinst, G.A. Klouda, L.A. Currie, "Radiocarbon
measurements of extractable organic matter from the Integrated Air
Cancer Project Study in Roanoke, VA," these proceedings, 1993.
5. R.B. Zweidinger, T.E. Kleindienst and E. Hudgens, "Apportionment of
residential indoor VOCs and aldehydes to indoor and outdoor sources in
Roanoke," in Proceedings of the 84th Annual Meeting of &&WMA, paper
91-131.10, 1991.
6. R.B. Zweidinger, R.K. Stevens, C.W. Lewis and H. Westburg,
"Identification of volatile hydrocarbons as mobile source tracers for
fine-particle organics," Environ. Sci, Technol. 24(4):538 1990.
7. J.M. Ondov, Z.C. Lin, W.R. Kelly et al., "Enriched stable isotopes
of Sm as intentional tracers of diesel and residential oil furnace
emissions in Roanoke, VA," these proceedings, 1993.
8. TRC Environmental Corporation, "IACP Source Inventory for Roanoke,1'
Contract No. 68-D9-0173, Work Assignment No. 2/236, 1993.
0.2 0.3 0.4
Soil-Corrected K, ug/m3
0.5
0.6
Figure 1. 14C-determined extractable organic matter vs soil-corrected
potassium. The straight line and its uncertainty band is given by
(40 ±3) [K1], from equation 1.
-------
TABLE I Average ambient concentrations (EOM, TA98 +S9, and TA98 -S9)
and mutagenic potencies attributed to woodsmoke and mobile sources in
four U.S. cities during wintertime. The apparent relative importance
of the sources depends on the the location of sampling sites within
each city and season, and are not necessarily representative of annual
city-wide ratios. Non-Roanoke results are from Ref. 3.
Woodsmoke
Roanoke VA (1988-89)
n = 37 - 40
EOM Conc'n1
RJJ+ Conc'nb
R,8+ Potency0
Rjg. Cone ' nb
Rgg. Potency0
Boise ID (1986-87)d
n = 40
EOM Conc'n'
R,8+ Cone ' nb
R,8+ Potency0
d
8.1
5.9
0.80
4.5
0.53
14
12
0.84
± 0.6
±1.1
± .16
± 0.9
± .11
± 2
± 3
± .25
Mobile Intercept
Sources
4.0
10
3.0
5.8
1.7
6
18
3.0
± 0.5 -.3 ± .7
± 1 2.6+1.1
± 0.4
± 0.7 4.2 ± 0.9
±0.3
•
±2 2 + 2
±3 3 ± 4
±1.1
Meas.
Total
12
19
15
22
32
Albuquerque NM ( 1984-85) e
n = 44
EOM Conc'n"
R98+ Conc'nb
R,8+ Potency0
Raleigh NC (1984-85)
n = 40
EOM Conc'n"
R98+ Conc'nb
R98+ Potency0
15
19
1.3
e
16
12
0.78
± 1
± 2
± .2
± .5
± 1
± .07
3
11
3.7
1
4
3.7
±1 1 ± 1
±3 3 + 3
±1.5
± .3 0.1 + 1
± 1 1 + 1
+ 1.5
19
32
17
18
' Mg / m3
b revertants / m3
c revertants / jzg EOM
d residential and roadway site
e residential site only
-------
TECHNICAL REPORT DATA
1. REPORT NO.
EPA/600/A-93/167
4- TITLE AND SUBTITLE
SOURCE APPORTIONMENT OF FINE PARTICLE ORGANICS AND
MOTAGENICITY IN WINTERTIME ROANOKE
5.REPORT DATE
6.PERFORMING ORGANIZATION CODE
7. AUTHORCS)
Charles W. Lewis, Roy B. Zweidinger, Larry D.
Claxton, Donna B. Klinedinst and Sarah H. Warren
8.PERFORMING ORGANIZATION REPORT
NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Atmos. Res. and Exp. Assess. Lab and Health Effects
Res. Lab, U.S. Environmental Protection Agency,
Research Triangle Park, NC 27711; National Institute
of Standards and Technology, Gaithersburg MD 20899;
Environmental Health Research and Testing, Research
Triangle Park, NC 27709
iO.PROGRAM ELEMENT NO.
A101/C/90/01
11. CONTRACT/GRANT NO.
In-house
12. SPONSORING AGENCY NAME AND ADDRESS
Atmospheric Research and Exposure Assessment
Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13 .TYPE OF REPORT AND PERIOD COVERED
Symposium paper
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
To be published in the Proceedings of the AWMA/EPA Symposium, "Measurement of Air
Toxic and Related Air Pollutants," May 3-7,1993, Durham, NC
16. ABSTRACT
During the 1988-1989 winter the U.S. EPA conducted a comprehensive field study in
Roanoke VA as part of its Integrated Air Cancer Project (IACP). This paper
presents results of the source, apportionment of'fine particle extractable organic
matter (EOM) and its associated mutagenicity (Salmonella typhimurium TA 98 +S9 and
TA98 -S9). The source apportionment methodology is based on multiple linear
regression using a variety of tracer species: HC, metallic elements and volatile
hydrocarbons whose ambient concentrations were measured simultaneously with the EOM
and mutagenicity. The results are compared with those from previous IACP studies
in other locales.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED
TERMS
c.COSATI
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (Tftw Report)
UNCLASSIFIED
21.NO, OF PAGES
f
20. SECURITY CLASS (This Page)
UNCLASSIFIED
22. PR1PF
------- |