Gasoline and Methanol Exposures from
Automobiles within Residences and Attached Garages
Environmental and Occupational Health Sciences Inst., Piscataway, NJ
Prepared for:
Environmental Protection Agency, Research Triangle Park, NC
1993
L
J
U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
HTIS
-------
TECHNICAL REPORT DATA
1, REPORT NO.
EPA/600/A-93/190
2,
4. TITLE AND SUBTITLE
5.REPORT DATE
Gasoline and Methanol Exposures from Autornobles
Within Residences and Attached Garages
6.PERFORMING ORGANIZATION CODE
7. AUTHOR(5)
Clifford P. Meisel, Nicholas J. Lawryk,
Alan H. Huber, Gennaro H. Crescenti
8,PERFORMING ORGANISATION REPORT
HO,
9 PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental and Occupational Health Sciences Inst.
UMDNJ--Robert Wood Johnson Medical School
Piscataway, NJ
10.PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
C182Q23501
12. SPONSORING AGENCY NAME AND ADDRESS
Atmospheric Research and Exposure Assessment Lab.
Office of Research and Development
U.S. IPA
Research Triangle Park, NC 27711
13.TYPE OF REPORT AND PERIOD
COVERED
14. SPONSORING AQENCY CODE
IS. SUPPLEMENTARY NOTES
Presentation at 6th International Conference on Indoor Air Quality,
July 4-8, 1993, Helsinki, Finland
16. ABSTRACT
A pilot study was conducted to evaluate the characteristics of the air
concentrations within a garage microenvironment. The air exchange rate between the
garage and the house, the windspeed in front of the garage door, the fuel tank
temperature, and the air concentrations of benzene (from gasoline) and methanol
(from MlOO fuel) were measured after an automobile containing US summer grade
gasoline or a fabricated fuel tank containing MlOO fuel entered the garage and its
door was closed. The air concentrations in the garage were greatly elevated after
the car or MlOO fuel tank entered the garage compared to the ambient levels which
were present prior to the caar's entry. A steady state concentration was often
reached within 90 minutes of the automobile or fuel tank entering the garage and
the air concentration remained level until the fuel tank temperature returned to
ambient levels, several hours later. The maximum concentration obtained was a
function of the fuel tank temperature. These studies indicate that parking an
automobile in residential garages results in increased exposures to prisons
spending time within the horns and in the garage microenvironment:.
17,
KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Human Exposures
Air Pollution
Mobile Sources
Methanol
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
b. IDENTIFIERS/ OPEN ENDED
TERMS
19, SECURITY CLASS (This
Report)
UNCLASSIFIED
20. SECURITY CLASS (This
Page)
UNCLASSIFIED
C.COSATI
21. NO. OF PAGES
8
22 . PRICE
-------
EPA/600/A-93/190
GASOLINE AND METHANOL EXPOSURES FROM AUTOMOBILES WITHIN RESIDENCES
AND ATTACHED GARAGES
Clifford P. Weisel1-5, Nicholas J. Lawryk*, Alan H. HuberJ, Gennaro
H. Crescenti3
1 Environmental and Occupation Health Science Institute, Piscataway,
NJ, USA
2 UMDNJ-Robert Wood Johnson Medical School, Piscataway, NJ, USA
3 Atmospheric Sciences Modeling Division, Air Resources Laboratory,
NOAA, Research Triangle Park, NC 27711 USA (On assignment to the
Atmospheric Research and Exposure Assessment Laboratory, U.S.
Environmental Protection Agency)
Prepared for Presentation and Publication in the Proceedings of the
6th International conference on Indoor Air Quality and Climate,
July 5-8, 1993, Helsinki, Finland
-------
GASOLINE AND METHANOL EXPOSURES FROM AUTOMOBILES WITHIN
RESIDENCES AND ATTACHED GARAGES
Clifford P. Weisel12, Nicholas J. Lawryk2, Alan H. Huber*, Gennaro H. Crescent!3
1 Environmental and Occupation Health Science Institute, Piscataway, NJ, USA
3 UMDNJ-Robert Wood Johnson Medical School, Piscataway, NJ, USA
3 Atmospheric Sciences Modeling Division, Air Resources Laboratory, NOAA, Research
Triangle Park, NC 27711 USA (On assignment to the Atmospheric Research and Exposure
Assessment Laboratory, U.S. Environmental Protection Agency)
ABSTRACT
When an automobile is driven into a residential garage, evaporating fuel from the warm fuel
tank can exceed the capacity of the emission control systems resulting in the volatile components
of the fuel being released into the air. This causes an elevation of the concentration of fuel
vapors within the garage and within rooms of the attached home. A pilot study was conducted
to evaluate the characteristics of the air concentrations within a garage microenvironmem. The
a<- exchange rate between the garage and the house, the windspeed in front of the garage door,
the fuel tank temperature, and the air concentrations of benzene (from gasoline) and methanol
(from M100 fuel) were measured after an automobile containing US summer grade gasoline or
a fabricated fuel tank containing M100 fuel entered the garage and its door was closed. The air
concentrations in the garage were greatly elevated after the car or Ml00 fuel tank entered the
garage compared to the ambient levels which were present prior to the car's entry. A steady
state concentration was often reached within 90 minutes of the automobile or fuel tank entering
the garage and the air concentration remained level until the fuel tank temperature returned to
ambient levels, several hours later. The maximum concentration obtained was a function of the
fuel tank temperature. Sulfur hexafluoride (SF6) was used as a tracer to determine the exchange
rate between the garage and the attached house. The SF6 indoor air concentrations in the room
adjacent to the garage was approximately 3% of the air concentration in the garage. These
studies indicate that parking an automobile in residential garages results in increased exposures
to persons spending time within the home and in the garage microenvironment. Estimates of
the daily exposure to automobile evaporative emissions should consider these microenvironments
prior to the substitution of alternate fuels, such as methanol, for gasoline since the methanol
levels within the garage may approach occupational standards which are set for the healthy adult
population not the general public which includes more sensitive individuals,
INTRODUCTION
The daily human exposure to a contaminant is the sum of all of the concentrations within
microenvironments where an individual spends his or her day multiplied by the time spent within
each microenvironment. Automobile fuel is one source of volatile organic compound exposures
to the general population. Currently, gasoline is the predominant fuel used globally, but
alternate fuels are being actively pursued within the US and elsewhere, with methanol being one
-------
candidate. Automotive fuel emissions occur both from tailpipe emission and from evaporation
of fuel in the engine block and fuel tank (1), If emissions occur within an enclosed area, such
as the cabin of an automobile or a garage, the ncentrations will be elevated relative to the
ambient levels or other mieroenvironments. For example, previous studies have shown that the
interior of automobile cabins have concentrations of gasoline derived compounds elevated
compared to ambient air and that the time spent commuting is a major contributor to the daily
exposure for these compounds (2,3)- Mode* estimates of various mieroenvironments impacted
by automobiles have suggested that a residential garage would result in elevated exposure if an
automobile is parked there after it is driven in on a hot sum.ner day (4), A series of
experimental measurements were therefore made to test the hypothesis that elevated exposure
would occur because of large increases in the air concentrations of benzene and methanol within
a residential garage due to evaporative emissions from an automobile fuel tank containing
gasoline and M100 fuel, respectively. These tests were a pilot study to evaluate the
characteristics of the air concentrations in the garage mieroenvironments. The actual emissions
and resulting measured concentrations provide an example to demonstrate important
characteristics for planning future studies. Research is ongoing to determine the potential
exposures to larger populations,
METHODS
The benzene and methanol air concentration within a residential garage in Raleigh, NC, USA
was measured during the summer of 1992 before and after gasoline powered automobiles were
driven into it or a fabricated fuel tank containing Ml00 fuel was pulled in. The fabricated fuel
tank was wrapped with a heating blanket in order to raise the tank's temperature up to 45°C,
stimulating the heating that occurs when an automobile is driven during the summer, The
fabricated fuel tank and one of the gasoline powered vehicles had temperature probes within the
fuel tank and on the surface of the fuel tank to measure the temperature of the fuel and tank
liurface, while a thermocouple was place on the outer surface of the fuel tank of the second
vehicle whenever it was driven into the garage. Each car was driven for approximately 30
minutes prior to entering the garage, during which time the temperature of the fuel tank reached
its maximum, which varied between 30° and 42°C. The maximum fuel tank temperature was
a function of the ambient air temperature, the roadway temperature and the amount of sunshine.
The M100 tank's temperature was adjusted to between 30° and 45°C prior to its being placed
in the garage. The temperature of the automobile's fuel tank remained elevated for several
hours due to the high heat capacity of the automobile. The methanol (M100) tank's temperature
was programmed to mimic the automobile's tank temperature in one set of experiments and
maintained at elevated temperature in a second to obtain a steady state condition. Two
experiments were also done placing either 5 mL gasoline or M100 fuel were on a glass
evaporating dish and allowing the fuel to evaporate to characterized the increase and falloff that
might be associated with a small spill of gasoline or Ml00 fuel.
The air concentration was measured using portable gas chromatographs with selective detectors
set either for benzene or methanol (Microsensor Inc. Model MSI301 and MSI301M). Two
portable MSI gas chromatographs were available for benzene measurements and one for
methanol. They were calibrated prior to the field program and the calibration was checked by
drawing of air containing known quantities of the contaminants from a gas sampling bag or a
-------
Summa canister. The calibration checks for benzene were within 30% of the expected value for
in the Summa canister. The methanol concentration within the Summa canister was only 50 ppb
and therefore not relevant to the air concentrations being measured. A semi-quantitative
methanol standard check was prepared in the field by adding 1 liter of air into a tedlar bag and
spiking the air with /xl quantities of methanol, then warming the bag to evaporate the methanol.
This methanol concentration the MSI301M measured for this type of standard was between 20
and 50% of the expected concentration. A comparison of joint samples collected in summa
canisters and analyzed in the laboratory also produced agreement of better than 50%. A
comparison of the two samplers was made by analyzing the contents of air collected in tedlar
bags throughout the experiments. A scatter plot of the analyses of these samples by the two
instruments are given in figure 1. The result indicate good agreement overall for the 43 pairs
of samples analyzed, though a positive bias for the concentrations measured by GC 2 appears
to be present at low concentrations (<0.2 ppni). Approximately 10% of the readings also
appear to be outliers. The individual instruments had responses of +-10% when a "steady state"
concentration was obtained in the garage over an half hour time period and from repeated
measurements made from a single bag sample.
The MSI gas chroma'egraphs were placed at breathing height within the garage and run
continuously, which resulted in samples being drawn and analyzed automatically every 5 minutes
for nethanol or 6 minutes for benzene. In addition, samples were collected using tedlar
sampling bags and analyzed using the MSI gas chromatographs after an experimental run was
complete to examine the spacial distribution of the contaminant within the garage.
Prior to th** benzene and methanol experiments the ventilation rate from the garage and the air
exchange race between the garage and the attached house was determined by releasing a set
amount sulfur hexafluoride over ten minutes and determining its decay using syringe samplers
which were programmed to sample at specific time intervals. Sulfur hexafluoride releases were
also done during the fuel evaporative emissions studies at a limited number of sites to confirm
that the exchange rates was the same during the garage characterization and the evaporative
emissions experiments.
RESULTS
The gasoline benzene and methanol air concentration within the garage rapidly increased
following the placement of 5 mL of gasoline and M100 fuels in the center of the garage. The
air concentration exponentially declined during the subsequent hour since no additional source
of these fuels were present (figure 2 and 3). Benzene air levels of 0.8 ppm and methanol air
concentrations of 20 ppm were measured during these experiments. The decline observed was
consistent with the air exchange between the garage and the surrounding environment.
Gasoline and methanol vapors would be expected to emitted into the garage from evaporation
from the fuel tank. The gasoline fuel tank was observed to increase to as much as 45CC
following a drive during sunny summer day. On cooler, rainy days the fuel tank temperature
only reached 32CC. The air concentration within the garage was elevated for benzene and
methanol compared to the ambient levels whenever the automobile or fabricated tank containing
M100, respectively, was placed into the garage. The benzene air concentration profiles with
-------
time since the automobile entered a garage is shown in figure 4. The gas tant -Ttperature at
the beginning of these two runs were 37°C and 43°C. The benzene concentration , measured
with two different GC's located within 2 mete s of each other. Differences in absolute
concentrations are noted between the GC's, but the temporal pattern observed were similar,
including a single elevated point that occurred approximately 50 minutes after the experiment
started. The concentrations of benzene increased then declined as the fuel tank temperature
decreased. The concentrations measured were much greater when the fuel tank was hotter, with
the benzene air concentration peaking at approximately 0.5 ppm.
The Ml00 fuel tank was heated to three different temperatures: 32°C, 39°C and 45°C; and
placed in the garage, during three consecutive time periods of a single day (figure 5). The
garage door was opened and near complete air exchange with ambient air occurred (figure 5).
Higher methanol air concentrations were measured within the garage as the fuel tank temperature
v?s increased. The tank was continually heated throughout the experiment until a steady state
appeared to be reached. A vertical profile of methanol concentrations was also determined
during these three experiments, at a position several meters from the monitor by collecting the
air in tedlar bags at 0.6 meter height intervals (figure 6). This sampling site was further from
the Ml00 tank than the GC, thus had lower overall values. They showed a minimum
concentration between 2 and 4 meters with maximum concentration near the floor and between
6 and 8 meters, the height of the ceiling.
DISCUSSION
Potential exposure to volatile components of automobile fuels will result from placing an
automobile within a residential garage after being driven on a hot, sunny day. Volatile
components from the fuel within the heated tank can breakthrough the charcoal canister that
provides the evaporative emissions control. One of automobiles used in this study was an older
vehicle for which the canister had been subject to extensive use and testing. In addition, this
vehicle had a carburetor engine which retains fuel in the engine block that is released via
evaporation when it is stopped. A second vehicle with a fuel injection system was also used and
lower air concentrations were measured with this vehicle. However, it was not operated on the
two hottest days during the experiments. The canister on the Ml00 fuel tank was not
regenerated during the 10 days of testing and may have saturated by the end of the experimental
period. However, saturation of the canister might occur in normal automotive operation when
numerous short trips are taken, so this test represents a valid, though possibly worse case,
emission test. The ventilation rate of the garage when the door was closed is within the range
expected for residential garages and was found to vary with the windspeed and direction relative
to the door. When the garage door was opened the air exchange rate from the garage and the
air concentrations returned to ambient levels within minutes.
The data generated is also being used to evaluate theoretical models of the concentration
distribution of gasoline constituents and methanol from Ml00 within garage based on source
emission data collected in controlled laboratory conditions, diffusion rates and ventilation rate
assumptions. The air exchange rate into the home was a conservative estimate, since the
doorway was kept closed during the rate exchange measurements. Increased infiltration of fuel
components into the home would be expected to be greater than the 3% measured in this study
-------
if there was frequent movement between the home and garage. Measurements in all appropriate
microenvironments should be done to validate the human exposure models prior to implementing
any large alteration to the automobile fuel supp'y.
ACKNOWLEDGEMENTS
The research described in this paper has been funded by the US EPA cooperative agreement
number CR82023501. The authors greatly appreciate the cooperation of Robert Burton for the
use of his residential garage.
This paper is being reviewed in accordance with the U.S. Environmental Protection Agency's
peer and administrative review policies and approved for presentation and publication. Mention
of trade names or commerical products does not constitute endorsement or recommendation for
use.
REFERENCES
1. Stump FD, Knapp KT and Ray, WD. Seasonal impact of blending oxygenated organics
with gasoline on motor vehicle tailpipe and evaporative emissions, J. Air Waste Manage.
Assoc. 1990;40:872-880.
2. Chan CC, Ozkaynak H, Spengler J and Sheidon L. Driver exposure to volatile organic
compounds, CO, Ozone and NO2 under different driving conditions, Environ. Sci.
Tcchnol. 1991:25:964-972.
3. Weisel CP, Lawryk NJ and Lioy PJ. Exposure to emissions from gasoline within
automobile cabins, J. of Expos, Analy. and Environ. Epide. 1992a,2:79-95.
4. Kavet R, and Niuss KM, The toxicity of inhaled methanol
vapors, Critical Reviews in Toxicology 1990,21:21-50,
Comparison of Portable CC
Concentration in ppm
Benzene
0,1!
to
B.I
..
e
o
6.
cc
;
^ a
fi 2 p Cfe
tL
c.e f
e.c
E
Q.
Q. BM
o.w
O »•«
C
O
" » f
*C
»c
6* ••• «••
PortobH CC 1
figur* 1
Co»p-»ri*on ef Rttpeait of th*
two Pori»bl« CCi
Time Since Spike (min)
Fijurt 2
Ti»e Prpfilt AftiT 5 »L
Caselint Ev«per*ttd in
Ccragt
-------
Benzene
Methanol E '
Q 0.fi
f^~ ~-% *S
CL
CL 20
-.
r^ IS
VB/
O
U. 10
"c
Q>
U *
C
o
O o.
i i i i
Q
*•/•••
A
*•
»•
' * *%•• '
- .
•
CL
^"^ i.S
O 0.4
"rn
2 0.3
"= 0.2
U
C 0.1
O i
O oo«
»
* «•
v
»™»
/ * *%
»
%
V
V** VV CC ^ *»»,, « 1 43'C
WV ***
•
,a>iiiiiifittt cci 37-c
0 SO 100 150 20
Time Since Auto Entered fm
20 40 60 80 100
Time Since Spike (mm)
Figure 3
Tine Profile After 5mL M100
Fuel Evaporated in Garage
Figure 4
Time Profile After Automobile
Entered Garage
Methano
Methano
T
CL
CL ,
^ , 30
C
O
'->-> 20
o
I—
C
—- 2.1
xS- ^i
^ 1.5
.? 1.2
0)
X O.B
O.f
0.3
Q 20 40 §0 iO tOO
O.B
// "
32*C >/ .
/ \
L^
Si
\ I
\ /
V
f\
39TC\45'C .
\ \
Since Auto Entered (mil
Figure 5
Tisae Profile After M100 Fuel
Tank Entered Garage
0.0 0.5 1.0 1.5 2.0 10.0 12.5 15.0 17.5 2D.O
Concentration (ppm)
Figure 6
Height Profile of Methanol
from Emission froa a Ml00
Fuel Tank at Three Temperatures
------- |