United States
Environmental Protection
Agency
Office of
Toxic Substances
Washington DC 20406
EPA-560/6-79-006
April 1979
Toxic Substances
v>EPA
Environmental Monitoring
Benzene
-------
ENVIRONMENTAL MONITORING
BENZENE
by
A. F. Fentiman, M. B. Neher, 6. W. Kinzer,
P. R. Sticksel, R. W. Coutant, G. A. Jungclaus,
N. A. Edie, J. McNulty, and C. W. Townley
BATTELLE
Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
Contract Nos. 68-01-3420 and 68-01-3858
Project Officers
Vincent J. DeCarlo
Office of Toxic Substances
U.S. Environmental Protection Agency
Washington, D.C. 20460
and
Richard Johnson
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
April 1979
-------
ACKNOWLEDGMENTS
Battalia's Columbus Laboratories and the project team members
wish to acknowledge Dr. Vincent DeCarlo of EPA's Office of Toxic Substances
and Mr. Richard J. Johnson of EPA's Office of Air Quality Planning and
Standards for their guidance and assistance in the planning of this program,
John H. Margeson and Ralph Baumgardner of EPA's Quality Assurance Branch
at RTP for their advice and assistance on quality assurance, Phil Youngblood
of EPA's Office of Air Quality Planning and Standards for his advice on mon-
itoring of meteorological conditions and location of sampling stations,
Mr. James A. Davis and Mr. Dallas Sumner of the Division of Traffic Engin-
eering of the City of Columbus for the traffic count which they conducted,
and numerous homeowners, industry personnel, and others who provided assis-
tance to the BatteHe sampling teams in the field.
-------
TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS i
TABLE OF CONTENTS 1l'
ABSTRACT iv
INTRODUCTION 1
PROCEDURES 2
Development of Sample Collection System 2
Evaluation of Sorbents 2
Charcoal 2
Polymeric Microporous Sorbents 2
Ambient Air Sampling System 7
Analytical Procedure for Benzene 9
Analytical Method 11
Calibration of Gas Chromatograph 13
Quality Assurance Studies 15
Sample Collection 15
Sampling System Design 15
Collection Efficiency of Tenax 17
Benzene Analysis 19
Statistical Analysis 21
Analysis of Environmental Samples for Benzene . . 22
Air 22
Water 22
Soil 24
RESULTS AND DISCUSSION 26
Industrial Locations 26
Selection of Sampling Locations 26
Analysis of the Air Monitoring Data 29
Analysis of Water and Soil Data 33
Water 33
Soil 33
Mobay Chemical Company 33
Gulf Oil Corporation 37
Petro-Tex Corporation 42
Chevron Corporation 47
Union Carbide Corporation 52
U.S. Steel Corporation 56
Ashland Chemical Company 59
ii
-------
TABLE OF CONTENTS
(Continued)
Service Station Locations 64
Selection of Sampling Locations 64
Sampling Protocol 65
Analysis of the Benzene Sampling Data 66
Location 1. The Morse Road and. 1-71 Monitoring 67
Location 2. The Fishinger Road and Mountview Avenue
Monitoring 75
Location 3. The Morse Road and Maize Road Monitoring ... 82
Emission Calculations 90
Urban Locations 92
Selection of Sampling Locations 92
Sampling Protocol 97
Traffic Data 98
Benzene Monitoring Results ..... 98
SUMMARY 105
REFERENCES 110
APPENDIX A
CONDITIONING OF TENAX GC FOR BENZENE ANALYSIS A-l
Stainless Steel Versus Glass A-l
Extracted Versus Unextracted Tenax GC A-3
Effect of Temperature and Time in Thermal Conditioning of
Tenax GC A-5
Tenax Conditioning System A-7
Preconditioning of Tenax A-9
Preconditioned Tenax, Type A A-l2
Preconditioned Tenax, Type B A-l2
Preconditioned Tenax, Type C A-l5
Preconditioned Tenax, Type D A-l5
Alltech Associates Preconditioned Tenax A-15
APPENDIX B
BENZENE PRODUCTION AND CONSUMPTION FACILITIES B-l
-------
ABSTRACT
Procedures were developed for the collection and analysis of
benzene in environmental samples. These procedures were employed in air,
water, and soil sampling for benzene in the vicinity of five industrial
facilities using or producing benzene. In addition, these procedures were
employed in air sampling for benzene in the vicinity of two other industrial
facilities, in the area around three gasoline service station locations, and
at three urban locations in Columbus, Ohio.
Air samples were collected by pulling air through a bed of Tenax GC.
The benzene was thermally desorbed from the Tenax and analyzed by cryogenic
capillary gas chromatography. Benzene present in water and soil samples was
determined by sparging it from the water or soil with nitrogen, adsorbing it
in Tenax absorption tubes, and analyzing the Tenax tubes in the same manner as
air samples.
Average 24-hour benzene concentrations at the air-sampling stations
o
around the industrial facilities ranged from 2 to 51 yg/m (0.5 to 19 ppb).
In general, the highest concentrations in air were found at sampling stations
downwind of the facilities or at stations closest to the facilities. The ben-
zene concentrations found in environmental water samples ranged from <1 to 13 ppb,
with levels higher than these being observed in plant effluents. Downstream of
the plant outfalls the concentrations were 1 to 2 ppb or less. The highest ben-
zene concentrations were found in soil samples, ranging from 2 to 191 ppb.
At air-sampling points around the service stations the average benzene
concentrations were in the range of 1 to 7 yg/m (0.3 to 2.0 ppb). These averages
are for periods between 12 and 30 hours. The levels in residential neighborhoods
upwind of the service stations ranged from 0.6 to 1.4 yg/m (0.2 to 0.4 ppb).
Maximum concentrations of 32.4 yg/m3 (10.3 ppb) and 68.6 yg/m3 (21.5 ppb) were
measured downwind of a service station during refueling of that station's under-
ground tanks. Automobile traffic was found to contribute to the benzene levels
at the service station locations and at the three urban locations near the business
district in Columbus. In a residential neighborhood near the business district,
o
the 25-hour average benzene level was 5 yg/m (1.5 ppb). In the center of the
3
business district it was 12 yg/m (4 ppb). Immediately adjacent to a busy highway
leading into the business district, the average levels for the 25-hour period of
3 3
the study were 9 yg/m (3 ppb) on one side of the highway and 23 yg/m (7 ppb) on
the other side. The higher levels were on the side of the highway carrying the
heavier traffic.
iv
-------
FINAL TASK REPORT
on
ENVIRONMENTAL MONITORING
BENZENE
by
A. F. Fentiman, M. B. Neher, G. W. Kinzer,
P. R. Sticksel, R. W. Coutant, G. A. Jungclaus,
N. A. Edie, J. McNulty, and C. W. Townley
February 1979
INTRODUCTION
In recent years, there has been increasing concern about
environmental levels of many organic compounds, including benzene. This
study was undertaken for the United States Environmental Protection Agency
in order to define environmental levels of benzene. To that end some po-
tential benzene sources were chosen including industrial facilities where
benzene is processed or used, gasoline service stations, and an urban area
where traffic may be a source of benzene in the atmosphere. Accomplishment
of the aims of this program required selection of suitable sampling locations,
development of methodology for benzene sampling and analysis, and the collec-
tion and analysis of environmental samples. The major focus of the program
was air sampling; however, a limited amount of data was obtained on water and
soil samples collected in the vicinity of industrial facilities.
-------
PROCEDURES
Development of Sample Collection System
This part of the program was divided into two parallel studies:
evaluation of sorbents and design of the field sampling systems.
Evaluation of Sorbents
Little data are available on the environmental levels of benzene.
Therefore, it was considered essential to use as sensitive an analytical
method as possible and determine benzene levels down to less than 1 ppb
(3 yg/m ). This made it necessary to use a sampling system which concen-
trates the benzene in a suitable medium. It was decided that an adsorbent
system was the only feasible approach. Cold traps would not only be very
inconvenient for field sampling, they would collect much more water than
benzene, and it is questionable whether a cold trap would work efficiently
for benzene in the parts per billion range.
Charcoal. Charcoal was considered as a sorbent, but discarded on
the basis that desorption with any solvent would give a significant back-
ground of solvent impurities leading to reduced sensitivity. The NIOSH
method for benzene utilizes CS~ as solvent. However, that method was de-
signed for concentrations in the parts per million range, so that relatively
large amounts of benzene are expected. Also, for low concentrations of
benzene, the C$2 presents serious interferences to the quantitative analysis. '
In addition humidity is known to have an adverse effect on the sorption capacity
of charcoal, and no capacity or breakthrough data is available for benzene con-
centration in the parts per billion range.
Thermal desorption of benzene would avoid the interferences from
the eluting solvent and the dilution effect. However, Pellizzari has found
that the recovery of benzene from charcoal by thermal desorption is very low.
Polymeric Microporous Sorbents. Two polymeric microporous adsorbents
were considered. Chromosorb 102 and Tenax GC. In other work at Battelle, both
have been used satisfactorily for atmospheric sampling. Tenax GC has superior
thermal stability, but Chromosorb 102 is significantly cheaper, so it was chosen
for initial evaluation.
-------
Chromosorb 102. To determine the sorbent bed size and air flow
rate for sampling, it is necessary to determine the retention volume for the
sorbate, benzene in this case. The polymeric microporous sorbents function
like a gas chromatographic column rather than by a sorption process by which
charcoal and silica gel function. To determine this retention volume a gas
chromatographic technique was used with a Chromosorb 102 column. A small
amount of the benzene was injected into the gas chromatograph and the time
at which the benzene peak started to elute was recorded. Plotting log of
elution time (log t) versus the reciprocal absolute temperature [l/T(k)],
gives a straight line which can be extrapolated to room remperature (or any
other desired temperature) to determine the elution time at the given flow
rate. The determination was made at two flow rates to demonstrate that the
retention volume was dependent upon the temperature. The data are given in
Figure 1, along with similar data for Tenax GC. Based on these data, a large
trap (4 cm I.D. x 10 cm length containing 50 g of Chromosorb 102) should have
a retention time for benzene of over 24 hours at 30 C with a flow rate of
100 ml/min -- which was considered adequate for the program.
Checkout runs showed that 1-hour desorption of benzene on Chromo-
sorb 102 gave essentially complete desorption of benzene. Benzene (30 pi)
was spiked into a Chromosorb 102 sample collection tube and laboratory air
drawn through at 100 ml/min for 1/2 hour. Desorption was then carried out
for 90 minutes at 185 C with nitrogen at a flow rate of 100 ml/min. The ef-
fluent was collected in three 18 in. x 1/4 in. stainless steel traps containing
1 g each of Tenax GC. Each trap was used for 30 minutes (0-30, 30-60, and
60-90 minutes). The Tenax traps were analyzed for benzene using gas chromato-
graphic analysis. The traps were heated in a Wood's metal bath at 200 C while
nitrogen was passed through the heated traps at 200 ml/min for 30 minutes and
into a 3 m x 2 mm 00 glass column packed with 5 percent DC510 on Gas Chrom Q
and maintained at -50 C. The analyte is frozen on the head of the column under
these conditions. The oven was then programmed from -50 C to 30 C at 20 C/min
and from 30 C to 200 C at 4 C/min. The 0-30 and 30-60 min traps contained
nearly all of the benzene; the 60-90 minute trap contained much less than
1 percent of the total amount of benzene in the three traps. The benzene
peak was saturated for the first two traps so a more accurate estimate is not
possible.
-------
(0
0)
4J
B
600
400
300
200
150
100
80
60
40
30
20
15
8
6
10
I I ' '
.5 g Tenax GC-50 ml/min
.5 g Chromosorb 102-
25 ml/min
.7 g Chromosorb 102-
50 ml/min
20 30 40 50 60
80
100
o
Temperature, C
(Scale Linear with Reciprocal of Absolute Temperature)
FIGURE 1. PLOT OF ELUTION DATA FOR BENZENE
FROM CHROMOSORB 102 AND TENAX GC
-------
5
Although these experiments showed that complete desorption could
be obtained, it also showed that the overnight conditioning of (200 C with
50 ml/mm flow of dry nitrogen) Chromosorb did not clean up the Chromosorb
adequately. A large number of peaks other than benzene also appeared on
the chromatogram which were only attributable to "bleed" from the Chromosorb.
Even 72-hour conditioning of the Chromosorb 102 did not lower the bleed suf-
ficiently to permit its use as the adsorbent for collection of benzene.
Tenax GC. Since it was not possible to condition Chromosorb 102
adequately, it was decided to switch to Tenax GC. Because of the much higher
cost of Tenax, however, it was decided that smaller traps should be used.
Based on the retention time versus temperature plots for benzene on Tenax
shown in Figure 1, a U-tube trap of .95 cm (3/8 in.) O.D. stainless steel tub-
ing 30 cm (12 in.) long was selected. This design was expected to give some-
what over 8-hour retention for benzene at 40 C using a 30 ml/min flow rate,
before breakthrough occurs. A lower flow rate would give longer retention
time, but 30 ml/min is the minimum reasonable attainable with the field
sampler system. The U-tube trap design was chosen because the U-tube traps
fit directly onto the gas chromatographic system for direct desorption
onto the column.
Tenax has been and is being used at Battelle as an adsorbent for
hydrocarbons sparged from water samples with nitrogen gas. After adsorption
of the hydrocarbons, the water vapor is flushed from the Tenax with dry nitro-
gen. The Tenax is then heated to desorb the hydrocarbons onto the chromato-
graphic column. For this work, the Tenax is conditioned by heating overnight
in a slow stream of dry nitrogen at 250 C.
For this program several hundred traps were required; so condition-
ing traps one at a time is impractical. Conditioning was carried out by hook-
ing six traps in series. A Varian 2100 biomedical gas chromatograph with
controls for four columns permitted activation of 24 traps (four groups of
six traps each) at one time. The traps were conditioned overnight at 250 C
with a flow of 100 ml/mm of helium. Selected traps were then run through the
thermal desorption procedure described above and a chromatogram of the de-
sorption products was obtained. Overnight conditioning did not clean up the
Tenax satisfactorily. There was a large amount of background including a
-------
peak exactly where benzene elutes. Conditioning for 48 hours gave a significant
improvement and in some cases the traps were considered to be satisfactory. How-
ever, in subsequent studies, some of the "cleaned" traps were still found to
dive unacceptable background levels.
A search of the literature gave several clues about the source of the
problem. Pellizzari, et ar ' report the use of Tenax GC for GC-MS analysis of
atmospheric pollutants. They prepare Tenax GC for use by extracting 24 hours
with methanol (or acetone^ '), vacuum drying, and thermal conditioning at 275 C
for 20 minutes in a stream of dry helium. Bertsch, et aP ' also report the use
of Tenax GC for collection of atmospheric volatiles for GC-MS analysis. Their
conditioning consists of heating Tenax at 340 C for 1 hour in a stream of dry
nitrogen.
After rather extensive study, it was found that much more drastic con-
ditions than Pellizzari and Bertsch used were required to clean up the Tenax GC
satisfactorily for this program. Details of the experimental studies are given
in Appendix A. The procedure developed is:
(1) Extract with methanol for 48 hours in Soxhlet Extractor
(2) Vacuum dry overnight at 100-125 C
(3) Condition in a large glass tube at 300 C for 16-20 hr,
with a purified nitrogen stream flowing through the tube
(4) Pack in stainless steel traps
(5) Condition at 275 C for 16-20 hr, with a purified
nitrogen stream flowing through the trap at "25 ml/min.
A variation of this procedure was used, at the suggestion of
Pellizarv ', for cleanup of Tenax which had been used previously for benzene
sampling:
(1) Extract successively with methanol for 48 hours and
then with pentane for 24 hours in Soxhlet Extractor
(2) Vacuum dry overnight at 100 to 125 C
(3) Pack in stainless steel traps
(4) Condition traps at 275 C for 24 hours with a purified
nitrogen stream flowing through the trap at -25 ml/min.
The pentane is used to ensure removal of any hydrocarbons that may be present
from the previous use.
-------
Calculation of breakthrough capacity based on the breakthrough data
shown in Figure 1 indicated that a breakthrough time of 8 to 9 hours was ex-
pected for benzene at 40 C (the maximum temperature that should be encountered
during sampling) and a flow of 30 ml/min. However, those data were based on
a small lot of Tenax already on hand. A new batch of Tenax ordered to fill
the first 100 traps had a lower bulk density. It took 3.5 g of the original
material to fill one trap, and only 2.2 g of the new lot. The effect of the
lower bulk density on breakthrough time was not known. Consequently, break-
through was determined for the new lot of Tenax using one of the stainless
steel U-tubes.
Breakthrough was determined by a gas chromatographic procedure using
the trap as a column at 40 C. The analyte was injected through the standard
injection port and swept onto the trap by the carrier gas. Since peak broaden-
ing after several hours retention time becomes severe, direct observation of
breakthrough is difficult. Consequently, a second Tenax trap was attached to
the effluent end of the first trap and changed hourly. These backup traps were
then gas chromatographically analyzed by thermal desorption. The retention time
of benzene was 10-12 hours based on a 30 ml/min flow rate. Precise breakthrough
times are not determined by this method since only hourly values are obtained.
However, the data clearly show that adequate breakthrough time is available for
the needs of this program.
Ambient Air Sampling System
The system shown in Figure 2 was developed to sample for benzene in
ambient air. The system permits concurrent sampling with up to four 0.95 cm
O.D. (3/8 in.) x 30 cm (12 in.) U-tube traps or other trap configurations.
Two traps may also be connected in series to check for breakthrough during
sampling. Flow through each trap is controlled by a calibrated jewel orifice.
A flow meter is used to monitor total system flow. Sample system vacuum is
monitored on a gauge and controlled by a vacuum relief valve. A Gast carbon
vane, continuous duty pump is used to provide air flow through the system.
The system is contained in a weather-proof housing. Air inlet to the system is
through a filter mounted about 5 ft above ground level. Six systems were con-
structed for field use. Necessary spare parts were available to repair or
replace system components in the field, if necessary.
-------
TRAP
MANIFOLDS
47mm
FILTER
r
THERMOMETER
TRAP ASSEMBLY DETAIL
VACUUK1
FLOWrtETER
VACUUM] CONTROL
VALVE
r—Q^ST PUrlP
FIGURE 2. AMBIENT AIR SAMPLING SYSTEM
-------
Two types of situations are encountered in which it is expedient
to operate air sampling devices without dependence on 110 volts alternating
current utility lines.
One of these situations entails traversing an air current downwind
of a suspect source. Such traversals are made by hand-carrying a personal -
type battery operated sampler for as far as a mile or more perpendicular to the
wind; connections to mains is impractical in such situations.
The second situation involves continuous samplings of 24 hours or
longer on sites which are remote from electrical services. The sites may be
chosen to deliberately avoid extraneous emission sources (motor vehicles, in-
cinerators, space heaters, paint, black-top, etc.) usually associated with
117 VAC terminals. To provide extended sampling capability on fixed remote
sites, a battery-powered pumping system was assembled.
The system consists of a small 30 ampere-hour 12-volt lead-acid
battery, a Trico 70 112-1 Electr-Vac vacuum windshield wiper booster pump,
a pair of check-valves, a 30-pound refrigerant can, a Barksdale No. D2T-H18
pressure switch and protective relay, a plywood box (fitted with a hinged lid
and handles) to contain the above, and appropriate plumbing, wiring, and hard-
ware. Sketches of the pumping system appear in Figure 3.
Analytical Procedure for Benzene
A cryogenic capillary GC technique was chosen for analyses of benzene
in environmental substrates on the basis of previous studies performed at BCL
on analysis of hydrocarbons in marine materials. No extensive developmental
work was required other than modification of a Varian 1400 gas chromatograph
to operate under cryogenic conditions and establishment of appropriate gas
chromatographic conditions to achieve satisfactory resolution of benzene from
other volatile components in environmental substrates.
-------
10
TO SAMPLER
VACUUM CAGE
CHECK VALVES
PUMP
EXHAUST TO
OUTSIDE OF BOX
TO VACUUM SWITCH
TO LOW VAC .TO HI VAC
LIMIT #1 LIMIT #2
CONDUIT BOSS RED TERM. N,C, YEL TERM. N.C.
SAWED OFF \. \ 7
r—
VACUUM
- -• - r
+19 vnr i
(COMMON) J
PT-- o
1 *. '
! i
' ! ~
A , , ,i i-
SWITCH ;
i
. i * i . ,/-»
!\
/
/^-^ pt f .ip
o
1
I 12 VDC
-12 VDC
FIGURE 3. PUMPING SYSTEM FOR PORTABLE SAMPLER
-------
11
Analytical Method
Benzene adsorbed on conditioned Tenax GC (s2.7 g) packed in 3/8 in.
x 12 in. stainless steel U-shaped tubes is injected into a capillary GC by
means of a dual valving arrangement utilizing two Carle 8-part valves as shown
in Figure 4. The tubes are flushed with helium at 50 ml/min at room temperature
to remove air and water. The flush period varies depending upon the environ-
mental substrate being analyzed for benzene. For air samples, the flush period
is 10 minutes and for water and soil samples, 60 minutes.
The Tenax tube is then rapidly heated to 200 C with a Woods metal
bath. The flushed gas is vented. The benzene is desorbed from the Tenax with
a 10 ml/min stream of helium for 10 minutes and collected on the head of the
column at -70 C. A 50-meter glass capillary column (.01 in. I.D.) with SF96
as the liquid phase is employed. The oven is allowed to warm for 3 minutes
(which brings the oven temperature to about 0 C). It is then temperature
programmed at a rate of 1 C/min with 0 C as the starting point. Under these
conditions, benzene elution time is approximately 14.5 minutes. Peak areas
are determined by a computer or electronic integration system. A Spectra Physics
SP4000 computer was used in this work.
A schematic diagram of the valving system used for injecting the sam-
ple into the GC is shown in Figure 4. The trap is flushed in the same direction
as it is sampled. By convention the traps are made with one arm 1/2 in. shorter
-- this is the inlet used for sampling. 'During the flush mode (Figure 4a) the
marker gas loop is also flushed with marker gas (150 ppm chloroform in nitrogen).
Figure 4b shows the system in the "inject" mode. In this mode a
10 ml/min stream of helium is flushed through the sample trap for 10 minutes
through the chloroform marker loop and onto the head of the column! After the
10-minute inject cycle, the system is returned to the flush mode and the GC oven
temperature program is started.
The chloroform marker gas can be useful as an aid in identifying the
benzene peak if there are no interfering peaks in the vicinity of the chloro-
form peak. With the particular column, packing, column length, carrier gas flow
rate and temperature program, used in the current work, chloroform elutes at
11.5 minutes and benzene at 14.5 minutes, exactly 3 minutes later. However,
because the Varian 1400 gas chromatograph used is not under temperature control
-------
12
Vent
Flush Gas
50 rol/min.
Tenax
Trap
Marker
'V'S ml/min
Carrier Gaa
2 ml/min
Vent
.
Inject Gas
10 ml/ruin
Markej
Loop
Vent
To Column
a. Flush Mode
Flush Gas
50 ral/min
Vent
Tenax
Trap
Carrier Gas
2 ml/min
Marker
ml/min
Vent
Inject Gas
10 ml/min
Marker
Loop
b. Inject Mode
FIGURE 4. VALVE INLET SYSTEM FOR ANALYSIS OF "TENAX TRAPS
-------
13
for the first three minutes (-70 to OC), the time of chloroform and benzene
can vary ±~0.2 minutes. Actually the chloroform marker was of little value
in this work because of the interferences encountered.
The identity of benzene in a random selection of air, water, and
soil samples was confirmed by GC-MS. In all samples analyzed, the benzene
peak was always the largest peak eluting in the benzene region. In air sam-
ples the closest interferences were methyl chloroform and cyclohexane, for
water the only close peaks were so small that they could not be identified,
and for soil the interferences were methylcyclopentane and, to a small extent,
methyl chloroform.
The minimum detectable level of benzene on column above background
by this analytical method is approximately 2 ng which is equivalent to about
3
0.2 ug/m or >0.1 ppb.
Calibration of Gas Chromatograph
Calibration of the gas chromatograph was performed two or three
times weekly by use of a calibrated gas mixture and a loop injector system.
A variety of valve configurations can be employed, and the loop sizes and
standard gas concentrations can be varied over fairly wide limits, provided
that the total amount of benzene injected for each loop covers the range of
interest. In this work, four loops and two standard gas mixtures of 183 and
25 ppm of benzene in nitrogen obtained from Matheson Coleman and Bell were used.
The volumes of the loops and weight of benzene in each are:
Benzene Weight, ng
Volume of Loop, y£ 25 ppm Std 183 ppm Std
126 10 72
360 28 205
982 77 559
2,236 175 —(not used)
Figure 5 shows the calibration valve system used. It consists of
two Hamilton Micro-Volume, 8-port valves (Model #2013). There are two loops
on each valve. In Figure 5a, loops 2 and 4 are being filled with calibration
gas and loops 1 and 3 are being swept by carrier gas. In Figure 5b, the valve
on the left has been turned to inject loop 2 and fill loop 1 with calibration
gas. In Figure 5c, the valve on the right has been turned to inject loop 4 and
fill loop 3 with calibration gas.
-------
14
Carrier gas
CC
Loop 1
LOOP y
Calibration Car>>
—' /Loop 2
Loop 4
Vent
a. Fill Loops 2 and A
Carrier gas
Loop
Calibration Gas>i—
b. Inject Loop 2, Fill Loop 1
GC
Carrier gas
y
Loop 1
Loop 3
T
Calibration Gas
^1
' Loop 2 V/ Loop 4
^S
Vent
c. Inject Loop 4, Fill Loop 3
FIGURE 5. LOOP SYSTEM FOR CALIBRATION OF GC
-------
15
The calibration curve is prepared by plotting the weight of benzene
injected versus the integrator output. For the SP4000, the output is in units
of microvolt seconds. At least one calibration point should be checked daily,
and if significant discrepancy (greater than 2-4 percent) is found, the entire
calibration curve is rerun.
Quality Assurance Studies
In addition to routine quality assurance procedures a series of
experiments were performed to confirm the validity of the data obtained in
the sampling and analysis of benzene in ambient air.
Sample Collection
The issues addressed regarding the influence of the sample collection
procedures on data validity were the effects of sampling system design and the
collection efficiency of Tenax under various conditions.
Sampling System Design. During the conduct of this program two
points of concern arose regarding the sampling system design:
(1) The effect of the positioning of the sampling
tube inlet on representative sampling of am-
bient air
(2) The possibility of sampling pump exhaust mixing
with the ambient air and resulting in an erron-
eously low analysis for benzene.
Through the end of 1977, the sampling tube inlet was situated below a
flat protective surface of aluminum foil. When concern about the design was
raised, a new design was adopted for use in subsequent sampling, and an exper-
iment was performed to compare the two. The modified design involved an exten-
sion of the sampling tube (with an inverted funnel at the end) upwards through
the protective aluminum cover. The experiment was conducted at a busy traffic
intersection in Columbus, Ohio.
-------
16
The sampling period was 11 hours (0730 to 1830) and the sampling
rate 31 to 35 ml per minute. Sampling was performed in duplicate with two
traps in tandem (foretrap and backup trap). The two sampling systems were
positioned about 6 feet apart downwind and approximately equidistant (about
50 feet) from the center of the intersection which was controlled by a
traffic light. The systems were oriented so that pump exhausts were also
downwind:
x modified design
x initial design
t
wind direction
The analytical data are presented in Table 1. The samples collected on the
initial design analyzed 6.1 and 6.7 ppb and those collected on the modified
design, 5.7 and 5.9 ppm. Based on these data the current design was just
as effective as the modified one in collecting representative samples of
ambient air.
TABLE 1. EFFECT OF SAMPLING SYSTEM DESIGN IN
COLLECTION OF AMBIENT AIR SAMPLES
Sampling Period: December 30, 1977, 0730-1830
Rate: 31 to 35 ml/min
Temperature: 30 to 36 F
RH: 57 to 65 percent
Wind: Speed, calm to 1 m/sec; Direction, 124-235
Sampling
Design
Initial 1
2
Modified 1
2
Volume
Sampled,
a
22.8
23.1
22.8
20.8
Benzene Analysis
21.1
23.4
19.7
20.4
PPb
6.1
6.7
Avg.6.4
5TT
5.9
Avg.5.8
Benzene Breakthrough
on Backup Trap,
<2
-------
17
During the first two sampling trips (in 1976), 3 in. x 8 ft cardboard
tubes were used to divert pump exhaust away from the sampling systems. When
this practice was abandoned in the 1977 sampling, the question was raised of
the effect of possible mixing of exhaust with sampled ambient air. To allevi-
ate this concern, 25 feet of exhaust hose was used in the 1978 sampling. In
addition, an analysis was made of a sample of the exhaust taken 1 inch from the
exhaust port of the box in which the pump is housed and compared with an analy-
sis of ambient air. The benzene content of the exhaust was 2.5 ppb, which was
essentially the same as the 3.2 ppb found in the ambient air.
The sampling system utilizes a Cast carbon valve pump modified by an
air bleed valve on the low pressure side adjusted to achieve a vacuum of 20 in-
ches of mercury when the air flow through the inlet to the Tenax traps (controlled
by calibrated orifice) is 30 cc/min. When sampling in duplicate the ratio of
exit air from two Tenax traps to exhaust air* is approximately 0.06 H/2B.3 a or
stated another way, the exhaust air is approximately 99.8 percent ambient air.
Also, the sampling systems are always positioned so that the exhaust ports,
which are 4 feet below and to the side of the sample inlets, are downwind. Hence,
on the basis of these facts, it must be concluded, even when wind speed and
direction are variable, that pump exhaust will not constitute a significant
fraction nor have any significant effect on benzene composition of ambient air
samples.
Collection Efficiency of Tenax. Backup traps were used routinely in
field sampling to confirm that breakthrough of benzene was not occurring. Data
from these traps indicate that retention of benzene by Tenax is essentially
quantitative; in two typical instances foretraps analyzed 278 ng (7.5 ppb) and
32 ng (0.8 ppb) of benzene while the corresponding backup traps in each case
analyzed less than 2 ng of benzene (<0.1 ppb). The largest quantity of benzene
found on any backup trap was about 15 percent of the quantity on the foretrap.
* Manufacturer's data cites exhaust rate of 0.95 cfm (approximately 26.9 a) at
20 inches of mercury vacuum.
-------
18
Experiments were conducted to determine the effects of high temper-
ature, high humidity, and flow rate through the sampler.
Atmospheres of 12 ppb of benzene in nitrogen and 100 percent relative
humidity (RH) were prepared by injection of 585 ng of benzene by calibrated
loop into evacuated 20 a Tedlar bags followed by introduction of sufficient
water to provide 100 percent RH at 72 F (•-'320 rug water) and 102 F (-700 mg
water). The bags were charged with 15 to 17 a of nitrogen and the contents
allowed to equilibrate about 1 hour at the temperature of collection. The en-
tire contents of the bags were collected on two Tenax traps in tandem (foretrap
and backup trap) at a sampling rate of 30 to 35 cc/min using the BCL sampling
rig. Traps were analyzed in the usual manner by cryogenic capillary GC. The
experiments were performed in duplicate at each temperature. Control experi-
ments were performed in the same manner using the same bags prior to the actual
benzene runs. The analytical data are shown in Table 2. Recoveries of benzene
were 99 and 109 percent at 72 F and 115 and 98 percent at 102 F. Backup traps
analyzed 1 to 3 ng of benzene background, corresponding to a breakthrough of
0.2 to 0.5 percent, which can be considered negligible.
TABLE 2. BENZENE RECOVERY STUDIES OF TENAX
Atmosphere: 12 ppb benzene (585
ng in 15 to 17s,
nitrogen); 100% RH.
Benzene Collection Volume Break- Benzene
Experi- Temp, Charge, Rate, Collected, through, RecoveredU)
ment
1
2
3
4
F
72
72
102
102
ng
585
585
585
585
cc/min
35
35
32
31
a
15.0
15.2
15.5
16.3
ng
<2
<2
<3
1
ng
580
635
675
575
%
99
109
115
98
(a) The Tedlar bags used in these experiments had a background inter-
ference to 10 to 15 ng of benzene.
-------
19
The collection efficiency of Tenax for benzene at the 1 ppb level
was determined at collection rates of 10, 100, and 1,000 cc/min. The data
in Table 3 indicate that the collection efficiency at the 1 ppb level is in-
dependent of collection rate; average collection efficiency at the three
rates was 104 percent. The benzene atmospheres were prepared in 20 a Tedlar
bags (^72 to 80 ng benzene in 15.5 a air or nitrogen (equivalent to approx-
imately 1.5 ppb) and sampled at the indicated rates on Tenax traps. The traps
were analyzed by cryogenic GC in the usual manner.
TABLE 3. EFFECT OF COLLECTION RATE ON COLLECTION EFFICIENCY
OF TENAX FOR BENZENE AT 1 PPB
Temperature, 72 F; RH, 0; Benzene Concentration,
72 to 80 ng in 15.5 £ of air or nitrogen.
Experiment
No.
1
2
3
4
5
6
Benzene
Concentration,
ng
72
80
72
80
72
80
Collection
Rate, cc/min Time, hr
ll.l(b) 22.5
16.0^ 17.0
yo . u L. • /
100. Q(C) 2.5
800. 0^ 0.33
1,000.0^°^ 0.25
Benzene Analysis(a)
ng
81
85
91
55
81
70
Efficiency, %
113
106
126
69
113
88
N
(a) The Tedlar bags used in these experiments had a background inter-
ference equivalent to 10 to 20 ng of benzene.
(b) In air.
(c) In nitrogen.
Benzene Analysis
Three methods were used routinely to judge the accuracy of the
data: (1) by spiking of the Tenax GC traps with known concentrations
(10 to 1000 ng) of benzene in methanol; (2) calibrated loop injections
(62 to 1100 ng) directly onto the GC column, and (3) spiking of the Tenax
GC traps using the calibrated loops. The three methods agree within 10 to
15 percent of each other over the concentration range of 10 to 1100 ng.
-------
20
On two occasions (in conjunction with the urban location sampling
in Columbus, Ohio, and the U.S. Steel Corp. sampling in Clairton, Pennsylvania)
Tenax traps were sent to EPA where they were spiked with known quantities of
benzene by a gas-permeation technique. The traps were returned and analyzed
periodically during the analysis of the field samples. The results are shown
in Tables 4 and 5. Agreement with the EPA values was good in all cases.
TABLE 4. DETERMINATION OF BENZENE IN EPA-SPIKED SAMPLING TUBES FOR
QUALITY CONTROL — URBAN COLUMBUS SAMPLING
Trap
No.
234
181
290
m
96
177
67
170
Amount
Determined,
ng
131
132
135
127
410
423
444
382
of Benzene^3'
EPA Values,
ng
124
124
124
124
388
388
388
388
Di f f erence ,
%
+5.6
6.5
8.8
2.4
Average +5.8
5.7
8.5
14.6
-1.5
Average +6.8
(a) Tubes supplied by Battelle-Columbus were spiked by the Quality Assurance
Branch of the U.S. EPA at Research Triangle Park, N.C., and returned for analyses
TABLE 5. DETERMINATION OF BENZENE IN EPA-SPIKED TRAPS FOR
QUALITY CONTROL — U.S.STEEL CORP. SAMPLING
Trap
No.
133
76
63
223
65
257
288
259
236
159
Amount
Determined,
ng
96.3
97.0
96.9
99.7
99.9
100.5
350.4
366.9
368.5
340.5
of Benzene^3'
EPA Values,
ng
no
no
no
no
no
no
379
379
379
379
Difference,
%
-12.5
-11.8
-11.9
- 9.4
- 9.2
- 8.6
Average -10.6
- 7.5
- 3.2
- 2.8
-10.2
Average - 5.9
(a) Traps supplied by Battelle-Columbus were spiked by the Quality Assurance
Branch of the U.S. EPA at Research Triangle Park, NC., and returned for analysed
-------
21
At least one or two field blanks were taken on each sampling trip
and analyzed with the field samples. No more than negligible quantities of
benzene were found in any of the field blanks.
Prior to the end of 1977, the 183 ppm benzene standard used for
calibration of the gas chromatograph was checked only twice. An assay of
the standard by mass spectrometry in December, 1977, showed 179 ppm and by
gas chromatography, 187 ppm. The assay in February, 1977, by.mass spec-
trometry was 183 ppm. In early 1978, a new 25 ppm gas standard was obtained.
In order to increase calibration reliability of the cryogenic gas chromatograph,
this standard was assayed daily by an independent GC method which is calibrated
against liquid standards of benzene in carbon disulfide (50 to 500 ng).
To verify that no benzene is lost during the flushing of the Tenax
trap to remove water, an experiment was performed to monitor the effluent from
the flushing operation. A Tenax trap sample (15 n) of a 10 ppb benzene atmos-
phere at 100 percent RH was prepared and backflushed at room temperature onto
the GC capillary column at the highest flow rate possible, 10 ml/min for 50
minutes; this is equivalent to the normal flush of 50 ml/min for 10 minutes.
The Tenax trap was then replaced with an empty trap and the analysis of the
flush effluent performed in the normal manner. No benzene peak was detected in
the region 12.1 to 12.4 minutes where benzene normally elutes. This result was
not unexpected since the previous studies showed a collection efficiency of
>99 percent at 100 F and 100 percent RH for a 15 £ sample.
Statistical Analysis
It was not considered necessary nor would it have been cost-
effective to analyze all samples in duplicate. However, a sufficient number
of duplicate analyses (and two triplicate) were performed to provide data for
a variance analysis.
An analysis of variance for a one-way design was run using the
(e\
Battelle Computer Center's version of the BMD programsv '. A logarithmic
transformation was used. From the variance analysis, a pooled standard
deviation was obtained. This value, s, with the Students "t" value for
95 percent confidence and appropriate number of degrees of freedom was then
-------
22
applied to the individual measurements to determine the 95 percent confidence
limits according to the following formulas:
Upper Confidence Limit = 10(log x + ts/V?f)
Lower Confidence Limit = 10(1og x ' ts/W)
where
£ is each individual value (or average if replicates were run),
n_ is the number of replicates,
^ is the standard derivatives described above (.053 is the
actual value), and
t_ is Students "t" value, 2.069 was used, representing 95 percent
confidence limits for 23 degrees of freedom.
Analysis of Environmental Samples for Benzene
Air
Samples of ambient air, collected in 2 to 12 hours at the rate of
30 ml/min on Tenax GC, were analyzed directly for benzene as described above.
Most of the samples were collected for 8 hours. At some locations, grab sam-
ples were also collected, at flow rates of 200 to 220 ml/min for 10 to 20-
minute periods.
Water
Samples were analyzed by sparging for 45 minutes with nitrogen at
100 ml/min at room temperature onto a Tenax GC tube which was then analyzed
for benzene in the same manner as air samples. The sparging was conducted in
the apparatus shown in Figure 6. The volume of sample used varied from 0.5
to about 20 ml depending upon the concentration of benzene. The volume of
the first sample analyzed was approximately 10 ml and if the benzene content
exceeded 500 ng (50 ppb), the analysis was repeated using a sample volume cal-
culated to give about 75 to 100 ng of benzene on column. The smaller sample
volumes were added to approximately 20 ml of presparged distilled water pre-
viously analyzed for benzene then sparged onto the Tenax tube.
-------
Tenax Trap
Water
CO
Fill Position
Sparge Position
FIGURE 6. SPARGING APPARATUS FOR WATER SAMPLES
-------
24
Soil
Samples of 3 to 6 g of soil were transferred to the sparging apparatus
shown in Figure 7 containing 15 ml of H^O and then sparged one hour at 75-80 C
with a stream of nitrogen at 100 ml/min. The sample was stirred with a magnetic
stirring bar during sparging. The spargate was collected on a Tenax GC tube
and analyzed for benzene in the manner previously described. Prior to injection
of the sample into GC, the tube was flushed 45 minutes to remove water.
-------
25
Tenax Trap
I j
FIGURE 7. SPARGING APPARATUS FOR SOIL SAMPLES
-------
26
RESULTS AND DISCUSSION
Three types of locations where benzene is likely to be present
were chosen for study in this program. Industrial locations where benzene
is either produced or consumed represent one type. Service stations that
dispense gasoline containing benzene represent a second type. An urban
area with a heavy traffic volume was chosen as a third type of location.
Industrial Locations
Selection of Sampling Locations
The first step in selection of sites for collection of samples was
collection of data concerning the production and consumption of benzene by
industrial plants. The original goal was to study storage and transportation
facilities, as well, but in the case of benzene, the major storage facilities
are located at the manufacturing or using facility. These are normally very
close so that there is little "transportation" in the usual sense.
Table 6 lists the U.S. cities with the largest production and/or
consumption of benzene. From this list a preliminary selection of plants for
sampling was made, based on covering as many types of processes as possible,
but sampling at the largest plants where practicable. This list of plants is
given in Table 7. Appendix B gives a listing of benzene production and con-
sumption facilities by various categories (Tables B-l through B-8).
The selection of plants to be sampled was made from this list,
based on discussions with the Sponsor. Later in the program a coke oven plant
in Clairton, Pennsylvania, and a maleic anhydride plant in Neal, West Virginia,
were added at the request of the Sponsor. The plants sampled are listed in
Table 8.
At each location, air sampling sites were chosen on the basis of the
following factors:
(1) Prevailing weather conditions
(2) Topographical details
(3) The availability of electric current to power the samplers
(4) The locations of populated areas that might be exposed
(5) The presence of other potential sources of benzene.
At those locations where soil and water sampling was done, the soil samples were
taken at the same sites as the air samples, and the water samples were taken
upstream and downstream of the plant.
-------
27
TABLE 6. MAJOR U.S. BENZENE CENTERS
State/City
California/Los Angeles
San Francisco
Illinois/Chicago
Kansas/Wichita
Kentucky/Ashland
Louisiana/Baton Rouge
Lake Charles
New Orleans
Shreveport
Maryland/Baltimore
Michigan/Midland
Missouri /St. Louis
New Jersey/Elizabeth
New York/Buffalo
Oklahoma/Tulsa
Pennsyl van ia/Phi 1 adel phi a
Pittsburgh
Texas/Houston
Corpus Christi
Odessa
West Virginia/Parkersburg
Puerto Rico
Virgin Islands
Others
Grand Total
Total Benzene
(in millions
Production
256
176
95
366
476
183
1099
117
no
220
293
—
169
176
931
403
4988
959
374
—
2161
183
80
13815
Capacity
of pounds)
Consumption
107
107
181
108
280
1254
564
104
701
257
130
14
--
568
100
5254
655
248
402
734
--
162
11930
Total
363
107
357
203
646
1730
183
1663
117
214
921
550
130
183
176
1499
503
10242
1614
622
402
2895
183
208
25711
-------
TABLE 7. PRELIMINARY PLAN FOR BENZENE SAMPLING LOCATIONS
Benzene Capacity
Within 100 Miles
of Large City
Process
Benzene
(Cat.-Ref.)
Benzene
(Py. Gas.)
Benzene
(Toluene) .
Ethyl benzene
Cumene
Nitrobenzene
Chlorobenzene
Detergent
Al kyl ate
Cyclohexane
Maleic Anhydride
Location
Texas City, TX
Taft, LA
Corpus Christi , TX
Baton Rouge, LA
Philadelphia, PA
New Marti nsvi lie, WV
Midland, MI
Richmond, CA
Sweeny, TX
St. Louis, MO
Nearest Large
City, miles
Houston (30)
New Orleans (20)
Corpus Christi
Baton Rouge
Philadelphia
Wheeling (20)
Midland
San Francisco
(10)
Houston (50)
St. Louis
Company
Amoco
Union Carbide
Coastal St. Gas
Foster Grant
Gulf Oil
Mo bay
Dow
Chevron
Phillips Pet.
Monsanto
Capacity^
623
513
513
720
360
88
285
107
80
141
Number
of
Plants
34
7
9
3
5
6
4
1
34
4
Capacity^9 '
10242
1663
1614
1730
1499
402
701
107
10242
550
ro
oo
(a) Capacity figures given in millions of pounds.
-------
29
TABLE 8. BENZENE SAMPLING LOCATIONS
Company
Location
Process/Product
Petro-Tex
Union Carbide
Chevron
Mo bay
Gulf Oil
Ashland
U.S.Steel
Houston, TTa'
Taft, LA
Richmond, CA
New Martinsville:
Philadephia, PA
Neal, WV^A
Clairton, PA
(a)
Maleic anhydride from benzene
Benzene from pyrolysis gas
Detergent alky!ate from benzene
WV Nitrobenzene from benzene
Cumene from benzene
Maleic anhydride from benzene
Coke ovens
(a) Selected at recommendation of Sponsor.
Analysis of the Air Monitoring Data
During the benzene sampling exercises, data were gathered for
several meteorological parameters which were expected to help explain the
benzene concentrations measured around the different plants. These param-
eters were wind speed, wind direction, temperature, cloud cover, and weather
type. The last two parameters in combination with wind speed and time of
day provided an estimate of atmospheric stability. Temperature can be re-
lated to evaporation rate. Wind speed can be used to estimate dilution as
well as rate of transport. Calm conditions indicate a potential buildup of
emissions in the vicinity of the source. Wind direction is an indicator of
the direction from which the benzene came to the monitor.
In analyzing the benzene sampling data with relation to these
meteorological conditions, the most striking observation was the number of
benzene measurements which appeared to have little or no relation to the
meteorological parameters. The conclusion was that some of the most notable
benzene measurements were the result of sources other than the one being
monitored or of a marked change in the emission rate of the source under ob-
servation. High benzene concentrations were obtained in grab samples when
the wind was not blowing from the Chevron plant. A local transitory source
was indicated. High concentrations measured to the north of Petro-Tex when
-------
30
the wind was from a northerly direction pointed to a large stationary source
of benzene to the north of the plant site. There was a marked rise in the
benzene measured at all four monitors around the Gulf Oil Company during a
nighttime sampling period. One possible explanation for this phenomenon
would be a sharp increase in emissions from the source company or one of its
neighbors.
Analyses of the monitoring results from each of the plants are
discussed in the following sections. In some of these analyses, estimates
are made of maximum benzene concentrations. For these estimates, considerable
use was made of the nomograms (Figures 8 and 9) in Turner's Workbook of
Atmospheric Dispersion Estimates.' ^ It was assumed that the effective height
of emission of benzene from the plants was 15 to 20 meters which would cor-
respond to a height greater than the height of large storage tanks, one of the
principal sources of escaping benzene. As shown by Turner's Figure 3-9, the
magnitude of the maximum xu/Q* for these heights is relatively uniform for
different atmospheric stabilities, while the distance from the source at
which the maximum concentration will occur varies by a factor of 10. Thus,
if the wind speed does not change, the major result of a change in atmospheric
stability is that the distance from the source to the point of maximum benzene
downwind will change. The concentration will remain approximately constant.
Turner's Figure 3-5 provides an estimate of the ratio of the maximum
benzene concentration and the concentration at some other downwind point.
Thus, from the information in these two nomograms and the field ob-
servations, an estimate can be made of the maximum concentration and the point
at which it occurs. It is assumed that no chemical transformation of the benzene
takes place within the distance from the source to the monitors.
* X = ambient condensation; u = wind velocity; and Q = source strength.
-------
31
L=50
L=750
L=IOOO
L=I500
L=2000
0.1 1.0 10 100
Distance, km
FIGURE 8. xu/Q WITH DISTANCE FOR VARIOUS HEIGHTS OF EMISSION (H) AND
LIMITS TO VERTICAL DISPERSION (L), B STABILITY
Source: Figure 3-5B from Turner, D.B., Workbook of Atmospheric Estimates.
-------
100
10
X
o
E
X
O.I
GO
ro
10
-7
10
-6
10
-5
(Xu/Q)max ,m
-2
10
-4
10
-3
FlUUkt 9.
DISTANCE OF MAXIMUM CONCENTRATION AND MAXIMUM xu/Q AS A FUNCTION OF STABILITY (CURVES) AND
EFFECTIVE HEIGHT OF EMISSION (NUMBERS)
Source: Figure 3-9 from Turner, D.B., Workbook at Atmospheric Estimates.
10
-2
-------
33
Analysis of Hater and Soil Data
Water. Benzene levels in water ranged from <1 ppb to a high of
179 ppb, which was found in a plant effluent. In general, benzene in plant
effluents quickly dispersed in rivers or streams to levels of 1 to 2 ppb or
less.
Soil. The highest levels of benzene observed in the three environ-
mental media were found in soil. Levels ranged from <2 ppb to a high of
191 ppb. These benzene levels in soil are to be expected; since soil is a
natural adsorbent and benzene would accumulate through the process of ab-
sorption from air or water (rain). With one or two exceptions the highest
levels of benzene were usually found in samples taken closest to the plant.
Comparison of soil concentrations from different plants or even from different
quadrants around one plant is restricted by the differences in soils.
Mobay Chemical Company -- November, 1976
The benzene monitoring sites for the Mobay Chemical Company nitro-
benzene plant at New Martinsville, West Virginia, are shown in Figure 10.
Results for air, water, and soil samples are given in Tables 9 and 10.
During the air sampling period at this plant, there were no oc-
casions when the wind blew directly toward a monitoring site during an 8-hour
integrated sampling period. During the first 8-hour sample (noon to evening),
the winds were relatively strong (4.4-7.5 m/sec) and from the west. In the
second period (evening to before sunrise), conditions were calm with a tem-
perature inversion indicated.
3
The maximum measured benzene concentration was 18.7 yg/m , observed
during the calm period at the monitoring site closest to the plant fence. The
3
lowest concentration (1.2 yg/m ) was observed during a period of strong winds
at the site farthest from the plant. The second highest measurement
(14.1 yg/m ) was observed when the wind was blowing in the general direction
of one of the monitors but not directly toward it.
-------
34
RIDGE
HIGHWAY
RAILROAD
PLANT SITE
RESIDENTIAL AREA
O AIR SAMPLE
# SITE NUMBER
SOIL SAMPLE
WATER SAMPLE
EMISSION POINT
KILOMETER
FIGURE 10- BENZENE MONITORING SITES AT MOBAY CHEMICAL COMPANY
-------
TABLE 9. ANALYTICAL DATA FOR AIR SAMPLES FROM MOBAY CHEMICAL COMPANY
Weather Conditions
Wind
Sampling
Station
Direction
Speed,
m/sec
Temp.
Range,
F
General
Sampl
ing
Period
Benzene Analysis
ppb .
(v/v) vg/m3'a)
270 4.4-7.5 37-60 Clear-dry front passed through
Calm 28-37 Clear and still
225° 3.5-4.8 28-43 Calm-si haze
270° 4.4-7.5 37-60 Clear-dry front passed through
Calm 28-37 Clear and still
225° 3.5-4.8 28-43 Haze to clear and bright
270° 4.4-7.5 37-60 Dry front passage
Calm 28-37 Clear and still
225° 3.5-4.8 28-43
279 4.4-7.5 37-60 Clear-dry front passed through
Calm 28-37 Clear and still
225 3.5-4.8 28-43 Haze to bright and clear
11/19/76 1110-1925 1.3
11/19/76 1945-0525 3.7
11/20/76 0545-1120 5.2
24 hour average 4
11/19/76 1200-2015 2.8
11/19/76 1230-0555 7.0
11/20/76 0615-1210 2.8
24 hour average 4
11/19/76 1400-2200 0.4
11/19/76 2210-0635 1.2
11/20/76 0647-1410 1.3
24 hour average 1
11/19/76 1525-2245 1.4
11/19/76 2300-0815 3.0
11/20/76 0835-1525 1.2
24 hour average 2
3.4 (2.7-4.4)
9.9 (8.6-11.5) (3)
14.1 (12.0-16.5)(3)
CO
7.6 (6.5-8.9) (3)
18.7 (14.5-24.0)
7.5 (5.8-9.7)
11
1.2 (0-1.6)
3.3 (2.8-4.0) (2)
3.4 (2.7-4.4)
3
3.8 (2.9-4.9)
8.1 (6.3-10.5)
3.2 (2.5-4.1)
5
(a) The numbers in parentheses are the 95 percent confidence limits, determined as discussed under the quality
assurance section. A single number appearing in parentheses is the number of replicates, otherwise the
result is based on a single sample.
-------
36
TABLE 10. ANALYTICAL DATA FOR WATER AND SOIL SAMPLES
FROM MOBAY CHEMICAL COMPANY
Sampling
Station
1
2
3
4
5
6
7
Benzene Analysis, ppb^a'
Water
—
--
Ohio River, downstream:
surface
6-ft depth
--
Ohio River, upstream:
surface
6-ft depth
Plant Influent
Plant Effluent
3.3
4.9
12.0
11.9
4.2
104.
(2.6-4.3)
(3.8-6.3)
(9.3-15.4)
(9.2-15.3)
(3.3-5.4)
(81-134)
Soil
2 (1.6-2.6)
Not analyzed
51 (40-66)
17 (13-22)
•» ••
—
—
(a) The numbers in parentheses are the 95 percent confidence limits,
determined as discussed under the quality assurance section. A
single number appearing in parentheses is the number of replicates,
otherwise the result is based on a single sample.
-------
37
This plant lies in a deep vally; thus the only opportunity for
the benzene to escape from the area is with winds up or down the Ohio River.
The 24-hour average concentrations ranged from 3 yg/m (1 ppb) to 11 yg/m
q
(4 ppb), with 48.7 yg/m (7 ppb) representing the peak concentration observed
during a calm 18-hour period.
Water samples were obtained upstream and downstream from the plant
and in the plant influent and effluent. The higher level in the upstream as
opposed to the downstream sample is very possibly due to a contribution from
the PPG plant upstream of Mobay which manufactures chlorobenzene from benzene.
This same contribution could also account for the Mobay plant influent having
a benzene concentration of 4.2 ppb.
The benzene concentrations found in soil ranged from 2 to 51 ppb.
The highest level found at this location was at the site most distant from
the plant.
Gulf Oil Corporation — December, 1976
The benzene monitoring sites for the Gulf Oil Corporation cumene plant
at Philadelphia, Pennsylvania, are shown in Figure 11. Air, water, and soil
samples were obtained, and the results are given in Tables 11 and 12.
Three integrated air samples varying from 7 to 21 hours were taken at
sites in each of the four quadrants around the plant site. During the first
sampling period, winds were generally from the south-southwest or southwest
directions (202.5° - 225°) away from Site No. 3 and toward Site No. 1. This
correlates with the benzene concentrations being twice as high at Site No. 1
as at Site No. 3. Winds were light during this period (1-3 meters/second) and
thus the plume rise would be expected to be greater than at higher speeds.
These light speeds would also mean little dilution and a small distance traveled
from the source. It is probable that the atmosphere was neutrally stable and
that dispersion was relatively good.
-------
~Ki—r I LJrS""-fvvi> X
-------
TABLE 11. ANALYTICAL DATA FOR AIR SAMPLES FROM GULF OIL CORPORATION
Weather Conditions
Wind
Sampling
Station Direction
1 202.5°
202.5°
450
Speed,
m/sec
1.8
1.8
0.1
Temp.
Range ,
F
34-42
42-40
40-40
Benzene Analysis
General
Overcast
Partly cloudy
Partly cloudy
ppb
Sampling Period (v/v) yg/m3va)
12/15/76
12/15/76
12/16/76
24
0900-1620
1620-1300
1300-2103
hour average
9.1
34.8
10.7
19
24
93
28
51
.4
.5
.8
(18.9-31.
(78.0-112
(22.4-37.
4)
1)
202.5
202.
450
202.5L
Calm
Calm
225U
Calm
450
o
1.0
0.5
0.01
1.01
34-40
40-30 Partly cloudy
30-40 Partly cloudy-overcast
40 Partly sunny
40 Overcast
40 Overcast
3
0.9
42-40
40-40
40-44
Cloudy
12/15/76 0957-1720 1.2
12/15/76 1720-0130 19.4
12/16/76 0130-1023 9.7
24 hour average 10
12/15/76 1053-1920 5.4
12/15/76 1920-0220 20.5
12/16/76 0220-1520 11.2
24 hour average 12 .
12/15/76 1130-2018 7.4
12/15/76 2018-0300 14.4
12/16/76 0390-1138 7.4
24 hour average 9
3.1 (2.4-4.0)
52.2 (40.6-67.4)
26.1 (20.2-33.6)
28
14.5 (11.3-18.7)
55.0 (45.2-65.9)(2)
30.2 (23.4-88.9)
31
20.0 (15.5-25.8)
38.7 (30.1-50.0)
19.8 (15.4-25.5)
CO
25
(a) The numbers in parentheses are the 95 percent confidence limits, determined as discussed under the
quality assurance section. A single number appearing in parentheses is the number of replicates,
otherwise the result is based on a single sample.
-------
40
TABLE 12. ANALYTICAL DATA FOR WATER AND SOIL SAMPLES
FROM GULF OIL CORPORATION
Sampling
Station
1
2
3
4
5
6
7
Benzene Analysis, pptr '
Water
--
--
--
--
Wastewater treatment
unit: outfall 4.3 (3.3-5.4)
from pump 56.6 (43.9-72.9)
River 1 (0-1.3)
Gulf Dock 1.9 (1.5-2.4)
Soil
34 (27-44)
>73
18 (14-23)
>34
-------
41
In the second sampling period which was primarily during night-
time hours, calm conditions or very light winds prevailed. The sky was
partly cloudy to overcast; therefore, if there was a temperature inver-
sion, it was not a strong one. Benzene concentrations rose sharply at
all monitoring sites increasing from 3 to 15 times above their values for
the first sampling period.
In the third sampling period, the wind speeds picked up very
slightly and the direction was generally from the northeast. In agree-
ment with this shift, the benzene concentration at Site No. 3 to the
southwest was slightly greater than the concentrations at the other sites.
Concentrations in all quadrants dropped between the second and third
periods but were rather uniform throughout the area. This may have been
the result of the calm conditions during Period No. 2 when the benzene
probably formed a stationary cloud over the area. When the wind picked up,
this cloud plus the new emissions from the source moved toward the southwest.
Considering that wind speeds were relatively light throughout the
three sampling periods, the low benzene concentrations measured at Site No. 2
during Period 1 is surprising. It is the site closest to the refinery. While
the wind did not blow directly toward this site, the benzene concentration was
markedly lower than at Site No. 3 which was upwind of the refinery. It is
possible that there is enough plume rise so that the benzene plume passes
above Site No. 2. This explanation loses some validity when it is recalled
that the benzene concentrations were high at Site No. 2 during Periods 2 and 3
which had even lighter winds.
The benzene concentrations in water samples obtained from the river
were only 1 to 2 ppb. Higher concentrations were found in water from the
plant wastewater treatment unit.'
The benzene levels in soil were the highest of the three media, and
the highest level was found closest to the plant.
-------
42
Petro-Tex Corporation -- July, 1977
Figure 12 shows the benzene monitoring sites for the Petro-Tex
Corporation maleic anhydride plant at Houston, Texas. The analytical
data for the air, water, and soil samples are given in Tables 13 and 14.
Winds were light (0-1.59 m/s) and variable in direction during
the entire air sampling period at the Petro-Tex plant. Benzene concen-
trations measured by the 8-hour integrating samplers ranged from 6.9 to
44.0 yg/m . The highest of these 8-hour concentrations was observed at
Site No. 3 located about 0.5 km northeast of the plant. It was measured
in the third period during daytime hours when the winds were calm to
0.5 m/sec.
In the sampling Period #2 from 0300 to 1115, the benzene
2
concentration was only slightly smaller (39.1 yg/m ). During this
night-to-midday period the winds were similar to those in the following
period, but the atmosphere was more stable. It is possible that there
was a cloud of benzene from the Petro-Tex plant and the winds dispersed
it or carried it aloft during either Period #2 or #3. The central portion
of the plume downwind of the plant had benzene concentrations of about
45 yg/m .
-------
Plant Site
Highway
""in Railroad
I ndust r ial
Residental
Area
A VVater
Site Number
CO
FIGURE 12. BENZENE MONITORING SITES AT PF/TKO-TEX CORPORATION
-------
TABLE 13. ANALYTICAL DATA FOR AIR SAMPLES FROM PETRO-TEX CORPORATION
Weather Conditions
Wind
Sampling
Station Direction
Speed,
m/sec
Temp.
Range,
F
Benzene
General Sampl
ing Period
Analysis
,ppt\ *(a)
(v/v) Mg/m
Continuous Monitoring Stations
1 180°-calm
Calm-225°
225°-190°
2 Calm
Calm-900
90° -calm
3 Calm
Calm-225°
2250-calm
0.48-calm
Calm-0.48
0.48-1.45
0.48-0
0-1.45
1.45-0
0.97-0
0-0.48
0.48-0
58-37
37-67
67-57
57-39
37-59
59-50
50-39
39-62
62-50
Clear
Partly
Sunny
Clear
Partly
Clear
Clear
Partly
Clear
1/20/77
cloudy 1/21/77
1/21/77
1/20/77
cloudy 1/21/77
1/21/77
1/20/77
cloudy 1/21/77
1/21/77
1650-0210
0210-1040
1040-1532
24 hour average
1605-0100
0100-0845
0805-1615
24 hour average
1810-0300
0300-1115
1115-1830
3
9
3
6
2
9
6
8
4
14
16
.9
.0
.4
10.
24.
9.
6
3
3
(8..2-13.7)
(18.9-31.3)
(7.8-11.1)
(2)
15
.6
.1
.2
7.
24.
16.
0
5
6
(5.4-9.0)
(20.1-29.7)
(13.7-20.2)
(2)
(2)
22
.5
.5
.4
24 hour averagell
4 225°-calm
Calm-315°
315°-135°
1 Calm
2 135°
45°
90°
4 Calm
67.5°
1.45-0
0-0.97
0.97-0.48
0
0.48
0.48
1.59
0
1.45
50-39
39-78
78-51
38
62
62
57
38
50
Sunny
Partly
Clear
Partly
Clear
Rain
Clear
Partly
Rain
1/20/77
cloudy 1/21/77
1/21/77
Grab Samples
cloudy 1/21/77
1/21/77
1/22/77
1/22/77
cloudy 1/21/77
1/22/77
1935-0355
0355-1200
1200-1935
24 hour average
0225-0245
0920-0940
1530-1540
1643-1703
1540-1600
1610-1620
6
3
3
4
5
10
1
0
12
21
.6
.2
.1
.2
.5
.9
.2
.1
.2
12.
39.
44.
31
17.
8.
6.
11
13.
28.
5.
0.
32.
57.
0
1
0
7
6
9
9"
1
0
45
5
0
(9.3-15.4)
(30.4-50.5)
(14.6-21.2)
(6.7-11.1)
(5.3-8.9)
(10.8-17.9)
(21.8-36.2)
(3.9-6.5)
(0-.6)
(25.2-41.8)
(44.3-73.4)
(2)
-------
TABLE 13. (Continued)
Weather
Wind
Sampling
Station
7
9
10
11
12
Speed,
Direction m/sec
30°
45°
Calm
Calm
135°
Calm
0.44
0.48
0
0
0.43
0
Temp.
Range,
F
50
50
38
38
62
78
Benzene Analysis
General
Grab
Rain
Rain
Partly
Partly
Sunny
Partly
Samples
cloudy
cloudy
cloudy
Sampling Period
ppb
(v/v)
yg
/m3(*>
(Continued)
1/22/77
1/22/77
1/21/77
1/21/77
1/21/77
1/21/77
1410-1430
1500-1510
0140-0154
1256-1310
1100-1120
1330-1350
9.2
8.1
3.2
2.2
7.9
2.6
24.6
21.9
8.6
5.8
21.3
6.9
(1
(1
(6
(4
(1
(5
9.1-31.7)
7.0-28.2)
.7-11.1)
.5-7.5)
6.5-27.4)
.3-8.9)
(a) The numbers in parentheses are the 95 percent confidence limits, determined as discussed under the
quality assurance section. A single number appearing in parentheses is the number of replicates,
otherwise the result is based on a single sample.
en
-------
46
TABLE 14. ANALYTICAL DATA FOR WATER AND SOIL SAMPLES
FROM PETRO-TEX CORPORATION
Sampling
Station
1
2
3
4
5
6
7
8
Benzene Analysis, ppb^a'
Water Soi 1
22 (17-28)
Not analyzed
Not analyzed
Not analyzed
Plant outfall 8 (6.2-10.3)
Midway 13.0 (10.1-16.7)
Upstream <1
Downstream of outfall 2 (1.6-2.6)
(a) The numbers in parentheses are the 95 percent confidence limits,
determined as discussed under the quality assurance section. A
single number appearing in parentheses is the number of replicates,
otherwise the result is based on a single sample.
-------
47
Two of the grab samples from Site No. 2 were taken with a light
wind blowing from the source toward this site which is one kilometer away.
Nevertheless, the concentrations that were measured there were the lowest
3 3
observed during the entire sampling period (0.45 yg/m and 5.0 yg/m ).
There are several possible reasons for these low readings including:
(1) light winds and large plume rise caused the plume to pass over this
site, (2) winds taken during these grab samples were too light and did
not persist long enough to bring the benzene to the site, (3) the rain
removed the benzene from the plume, or (4) the vegetation in the park
setting of this site removed the benzene. There is at least one other
reading taken during the Petro-Tex sampling that was made under similar
conditions, but which had a higher benzene concentration.
A reading that might have been selected as a background because
the wind was directed toward the plant occurred at Site No. 4 on January 22,
1977. However, instead of being a low concentration, it was the highest
o
(57.0 yg/m ). A possible explanation of this reading is that it was the
result of benzene being blown in from the refinery complex which is on the
other side of the Houston Ship Channel.
In the water samples, a benzene concentration of 13 ppb was found
in the stream near the plant outfall. The concentration measured in the out-
fall itself was 8 ppb. Downstream of the outfall, the concentration was down
to 2 ppb.
A concentration of 22 ppb of benzene was found in a soil sample
taken near the plant. This is the only soil sample which was analyzed.
Chevron Corporation — January, 1977
The sites for benzene monitoring near the Chevron Corporation
detergent alkylate plant at Richmond, California, are shown in Figure 13.
The results for air, water, and soil samples are given in Tables 15 and 16.
-------
0
HIGHWAY
RAILROAD
INDUSTRAIL
SWAMP AREA
RESIDENTIAL AREA
AIR, 4 & 8 HR,.
AIR, GRAB
SOIL
WATER
PLANT SITE
METER
1/2
.5 1
KILOMETER
co
FIGURE 13. BENZENE MONITORING SITES AT CHEVRON, INC.
-------
TABLE 15. ANALYTICAL DATA FOR AIR SAMPLES FROM CHEVRON CORPORATION
Weather Conditions
Sampling
Station
Wind
Direction
Speed,
m/sec
(a)
Temp.
Range,
F
Sampling Period
ppb
(v/v)
Benzene Analysis
ng/m3(b)
Continuous Monitoring Stations
1
4
6
2
3
4
5
180-360°
60-180
60-1800
60-180°
240°
180
110
90
240°
180
90
90
180°
1800
60
1.1-1.5
0.2-0.7
1.6-4.1
1.6-4.1
1.1
2.1
4.1
3.7
1.2
2.1
3.7
3.7
1.1
0.7
1.6
49-52
41-48
42-45
42-45
54
46
49
47
52
46
47
47
49
48
45
1/4/77 1345-1750
1/4/77 1840-2240
1/5/77 0820-1620
1/5/77 0900-1700
Grab Samples
1/4/77 1515-1535
1/5/77 1012-1032
1/6/77 1337-1357
1/5/77 1400-1420
1/4/77 1630-1650
1/5/77 1043-1103
1/5/77 1431-1451
1/5/77 1454-1514
1/5/77 1745-1805
1/4/77 1835-1855
1/5/77 0925-0945
3
2
1
3
1
1
1
1
18
8
1
1
3
6
12
7.3 (6.1-8.8)(2)
3.9 (3.3-4.7)(2)
3.0 (2.5-3.6)(2)
6.8 (5.7-8.1)(2)
2.7 (2.1-3.5)
3.5 (2.7-4.5)
2.9 (2.2-3.7)
3.0 (2.3-3.9)
55.4 (43.0-71.4)
25.9 (19.7-32.8)
3.1 (2.4-4.0)
3.0 (2.3-3.9)
8.9 (6.9-11.5)
18.7 (14.5-24.0)
36.8 (28.6-47.4)
(a) General: sunny with a few light clouds. No precipitation during sampling period. Winds variable to
highly variable.
(b) The numbers in parentheses are the 95 percent confidence limits, determined as discussed under the
quality assurance section. A single number appearing in parentheses is the number of replicates,
otherwise the result is based on a single sample.
IO
-------
50
TABLE 16. ANALYTICAL DATA FOR WATER AND SOIL SAMPLES
FOR CHEVRON CORPORATION
Sampling
Station
1
2
3
4
5
Benzene Analysis, ppb^a'
Water Soil
Not analyzed
191 (159-229)(2)
Yacht basin <1 70 (54-90)
51 (40-66)
<148(b>
Municipal water
(a) The numbers in parentheses are the 95 percent confidence limits
determined as discussed under the quality assurance section. A
single number appearing in parentheses is the number of repli-
cates, otherwise the result is based on a single sample.
(b) Sample is not heated; so this is a minimum value.
-------
51
The large size of this plant and its location on a peninsula
prevented air samples being taken any closer than 1 kilometer from the
benzene source. One set of grab samples was made as far away as 3 kilom-
eters from the benzene source. Winds were variable in direction during the
sampling at all the sites. They ranged from light to moderate in speed.
Two long periods of calm winds occurred during the monitoring.
There were several instances when benzene measurements were
made at sites upwind of the plant's benzene source. One of these was the
grab sample taken at Site No. 3. This was the highest observed concentration
and suggests the possibility that the observed benzene did not originate at
the Chevron benzene source but was produced by a source near the monitoring
o
site. The other high readings (25.5, 36.8, and 18.7 yg/m ) were also not
observed downwind of the suspected benzene sources. If there are no other
benzene sources in the area, the answer for these high readings would have
to be a complicated trajectory.
One grab sample measurement taken at Site No. 2 with a 4.1 m/sec
wind blowing from 110 is a possible example of a downwind sample. Site No. 2
was 1.9 km away from the suspected benzene source. The wind speed, cloud
cover, and time of day indicated a B-type atmospheric stability. With B sta-
bility and an effective emission height of 15-20 meters, the maximum concen-
tration would occur between 0.1 and 0.14 km downwind according to Turner's
Figure 3-9. His nomogram for B stability (Figure 3-5B) shows that the ratio
between the maximum concentration at 0.14 km and the concentration at 1.9 km
would be:
Xly/Q X] 3.5 x 10"4
X2y/Q x2 5 x 10'6
The benzene concentration observed at Site No. 2 was 2.9 yg/m so the ex-
3
pected maximum would be 243 yg/m under this line of reasoning. The con-
centration 1 kilometer from the source (the property line) would be about
60 ug/m , However, there is some doubt that this reasoning is correct. Four
grab samples were taken at Site No. 2 at different times for different wind
speed and direction conditions. The readings varied from 2.7 to 3.5 yg/m .
These observations could be interpreted as a local source effect, a background
concentration, or a complicated trajectory.
-------
52
Access to suitable water sampling sites was severely restricted by
the condition that we not sample on plant property. Without access via plant
property, it was necessary to obtain water samples about 1 to 1-1/2 miles
from the plant discharge point into San Pablo Bay at the Yacht basin. Benzene
levels in samples of brackish water and municipal water taken at this site
were less than 1 ppb.
In view of the relatively low levels of benzene found in the air
samples from this location, it is difficult to explain levels as high as
191 ppb of benzene being found in the soil samples. It is conceivable that
the benzene may have come from a source other than the Chevron plant (old oil
spills, gasoline spills, etc.).
Union Carbide Corporation -- January, 1977
The benzene monitoring sites for the Union Carbide Corporation
benzene plant at Taft, Louisiana, are shown in Figure 14. Results of the air,
water, and soil sampling are given in Tables 17 and 18.
Winds during the Union Carbide air sampling were light to moderate
(0.2 to 3.8 m/s) and generally with a southerly component. Areas to the north
of the plant were across the Mississippi River and were ruled out as sampling
sites because of the lack of bridges or ferries across the river in this vicin-
ity. Consequently, there were several measurements which can be considered as
representative of background (0.6 to 2.9 yg/m ). The others could give little
information on maximum off-plant observations.
There is one phenomenon — the sharp increase in benzene concentra-
tion during the 0200 to 1130 monitoring period at Sites 1 and 2 — that does not
appear explainable by changes in the weather parameters. This increase in am-
bient concentration may have been the result of an increase in emissions. If
not, then a more detailed monitoring program will be required to determine
causes of similar occurrences.
-------
53
RAILROAD
LEVEE
— HIGHWAY
O AIR, GRAB
SOIL
A WATER
0 AIR, 8 HR.
.5 1
KILOMETER
Emission Point
FIGURE 14. BENZENE MONITORING SITES AT UNION CARBIDE
-------
TABLE 17. ANALYTICAL DATA FOR AIR SAMPLES FROM UNION CARBIDE CORPORATION
Weather Conditions
Sampling
Station
1
2
3
5
2(b)
Wind Temp.
Speed, Range
Direction m/sec F
202.5° 1.5 48-54
1800 0.2 54-56
1800 1.6 56-61
202.5° 1.3 49-58
135° 0.6 58-54
202.5° 1.8 54-59
180° 3.8 54
135° 0.6 58-54
180° 1.4 58
157.5° 1.5 55
292.5° 2.1 69
292.5° 2.1 69
J
General
Continuous Monitoring Stations
Overcast
Overcast and int. rain
Partly cloudy w/mist and It. fog
Overcast
Overcast
Partly cloudy
Overcast
Partly cloudy
Grab Samples
Overcast
Overcast
Partly cloudy
Partly cloudy
ppb
Sampling Period (v/v)
1/26/77 0945-1920 0.7
1/26/77 1920-0225 0.6
1/27/77 0225-1015 4.5
24 hour average 2
1/26/77 1030-1955 2.6
1/27/77 1955-0255 1.1
1/27/77 0255-1125 6.4
24 hour average 4
1/26/77 1110-2025 12.8
1/27/77 2025-0335 13.5
16 hour average 13
1/26/77 0346-0426 0.6
1/26/77 0923-1003 0.2
1/28/77 1111-1157 3.9
1/28/77 2420-2430 2.4
Benzene Analysis
.yg/m3(a)
1.8 (1.5-2.1)(2)
1.6 (1.3-1.9)(2)
12.1 (10.1-14.5)(2)
6
7.0 (5.8-8.4)(2) *"
2.9 (2.4-3.5)(2)
17.3 (14.4-20.7)(2)
9
34.4 (26.7-44.4)
36.2 (28.1-46.5)
35
1.5 (1.2-1.9)
0.6 (0.0-0.8)
10.4 (8.1-13.4)
6.4 (5.0-8.2)
(a) The numbers in parentheses are the 95 percent confidence limits,
assurance section. A single number appearing in parentheses is
is based on a single sample.
(b) Traversed along line shown in Figure 14.
determined as discussed under the quality
the number of replicates, otherwise the result
-------
55
TABLE 18. ANALYTICAL DATA FOR WATER AND SOIL SAMPLES
FROM UNION CARBIDE CORPORATION
Sampling
Station
1
2
Benzene Analysis, ppb
Water
—
[a)
Soil
12 (9.3-15.4)
14 (11-18)
6 Plant Effluent 179 (155-207)(3)
7 Mississippi River:
8
upstream
Mississippi River:
downstream
2 (1.6-2.6)
1 (0-1.3)
(a) The numbers in parentheses are the 95 percent confidence limits,
determined as discussed under the quality assurance section. A
single number appearing in parentheses is the number of replicates,
otherwise the result is based on a single sample.
-------
56
Two grab samples in the vicinity of Site No. 2 were taken when
this site was downwind of the suspected benzene source. The distance be-
tween this site and the source was about 2 km. A combination of the
2.1 m/sec wind speed, partly cloudy sky, and daytime conditions indicate
a B-type atmospheric stability. Under these conditions, the maximum con-
centration is expected to have been within 0.2 km of the source (from
Turner's Figure 3-9). Thus, it would have fallen on the plant property.
3
Using the observed benzene concentration at Site No. 2 (10.4 yg/m ) and
Turner's nomogram (Figure 3-5B), one can estimate that the maximum short-
3
term concentration was on the order of 700 ug/m on plant property and
40 yg/m at the plant's fence line.
Water samples were obtained from the Mississippi River upstream
and downstream from the plant and from the plant effluent. The benzene
levels in the river were only 1 to 2 ppb though the concentration in the
effluent was 179 ppb.
The benzene .concentrations found in soil samples were 12 to 14 ppb.
U.S. Steel Corporation -- July. 1978
The benzene sampling sites for the U.S. Steel coke oven facility
at Clairton, Pennsylvania, are shown in Figure 15. Site No. 1 was at 1035
Monongahela Avenue, at the south end of the Westinghouse Apartment Division.
Site No. IB was at the South Allegheny High School approximately a mile and
one-half from the coke ovens and on a ridge above the coke ovens. Site No.
1A was at the Allegheny County Airport, and was generally downwind from the
plant, but somewhat further removed (about 4 miles) from the coke ovens and
on a ridge.
Site No. 2, at Belle Bridge, Pennsylvania, was generally out of
the air plume crossing the coke ovens and was intended for background pur-
poses. Sites 3 and 4, as shown in Figure 15, were at the Clairton Sewage
Plant and the Clairton Auxiliary Police Stations, respectively. These two
sites were the closest to the plant, but because of the variable wind di-
rection were only periodically in the emission plume from the coke ovens.
Only air sampling was done at this location, and the results are
given in Table 19.
-------
wmfcikd® mw /
? •' -'J:,);1 »'• •'••. ••••N; r'WK'STiH-} • •:/ / • //•// j fh*f...,\^~i 9}' ' ' C*
Wei««yv.\ ,-v • .,'!• ^ale'/v:f**i:^rW / *!'. I !> /'•' r^^ ; ^f~ -;-'^
<-V'-^ -| \ •.. • i l:>v-£^t^-i -?;4 fo t/ /1^F>- K, ^ r..,X/ >:^
t^>7:^L:v^4-. • :-v,T.:^>r»5r-; ••'/// ^^ fi'l^^f. Hrr1'.'^'^fs-^-x
1 \ri/);f'isiM« ? tinT-''';^ i'^*^-^-^^-
^J i. \i: l«;y||j^' J ft: fefevLr- i,V^---:::?•- -, • •,*&>:££$*:$&&&
*/ t^^-'-k ? M^^jm^^ • -?s. ->^-i^ft?.
A^.: V'£r-i.*v' :: N *•-?-#?••-.. \ T\\i- f'-^'i r-^ "FIFlnFrn^f- I/ /' X"yj V" -?"-:.-'-?^S^*^. S.:^^-•* '• •'•:
T*e4^&^^vs^^ '^ :m^&f
>%-X>v. -I vl \>«i?fA^x .; • ->. ji^f ^V>>^--/': ^:.VJ'.;-VT^\?.--/
Ss^^^fe^^^fr"', ':^rr^il&^4KH.'.\'i'::.^,/":x'i^.-'
^fiifl^S^^lir" -:x :^^t^f
-•V-^'--''^--'-.-;''/^ ^~- \ --^ ^ -••a%=2stv.\
4«sawp^,.-- f.K^^f^
mm^^m^^iir^
•^A'^:-^&mm
FIGURE 15. BENZENE MONITORING SITES AT U.S. STEEL
-------
TABLE 19. ANALYTICAL DATA FOR AIR SAMPLES FROM U. S. STEEL CORPORATION
Weather Conditions
Sampling
Station
Wind(a)
Speed,
Direction m/sec
Temp.
Range,
F
Relative
Humidity,
percent
General
Sampling Period
Benzene Analysis
PPb /.x
(v/v) yg/rn^ '
1A 110°-180° 1.6-4.1 77-73 31
110°-180° 1.6-4.1 73-62
1800-2200 3.6-6.2 62.81
IB 110°-1800 2.1-4.1 64-82 43
180°-nO° 1.6-4.1 84-59 55
1 340°-90° 1.8-2.5 70-82
900-1800 1.2-2.2 79-58
180° 0.6-2.4 58-74 51
2 340°-90° 1.7-2.5 69-75
90°-180° 1.3-1.7 75-56
180° 0.6-2.5 58-74
3 340°-90° 1.3-2.5 70-76
1800 0.9-1.6 72-51
180°-70° 0.6-3.2 51-79 59
4 180° 1.2-1.7 71-57
180°-70° 0.6-2.5 57-75 46
Calm
Clear,calm
Partly sunny
Clear, calm
Clear
Clear
Clear, calm
Clear, calm
Clear, calm
7/12/78 1230-2103 2.0
7/17/78 2103-0425 6.0
7/13/78 0425-1323 2.3
24 hour average 3
7/12/78 0957-1809 0.4
7/12/78 1809-0202 0.6
7/12/78 1041-1835 2.3
7/12/78 1835-0228 3.7
7/13/78 0228-1039 12.9
24 hour average 6
7/12/78 1102-1911 0.7
7/12/78 1911-0251 0.7
Partly cloudy, haze 7/13/78 0261-1103 2.0
24 hour average 1
7/12/78 1148-2019 1.5
7/12/78 2019-0352 3.0
7/13/78 0352-1250 2.5
24 hour average 2
Clear, calm
7/12/78 1947-0323 3.5
7/13/78 0323-1226 6.0
6.5 (5.0-8.4)
19.2 (14.9-24.8)
7.3 (5.7-9.4)
TF~
1.2 (0-1.6
2.0 (1.6-2.6)
7.5 (5.8-9.7)
12.0 (9.3-15.4)
41.5 (32.2-53.5)
19
2.4 (1.9-3.1)
2.2 (1.7-2.8)
6.4 (5.0-8.2)
4
4.9 (3.8-6.3)
9.7 (7.5-12.5)
8.0 (6.7-9.6)(2)
11.4 (8.8-14.7)
19.3 (16,2-23.1)(2)
C7J
CO
(a) Wind direction and speed were determined separately for ridge sites (1A and IB) and valley sites (1 to 4).
(b) The numbers in parentheses are the 95 percent confidence limits, determined as discussed under the quality
assurance section. A single number appearing in parentheses is the number of replicates, otherwise the
result is based on a single sample.
-------
59
The highest concentration observed was 12.9 ppb which was at
Site No. 1 in the valley, across the river from the coke ovens. The
readings at Site IB were lower than expected. This site was downwind
from the coke ovens during most of the sampling periods and the sampling
team could smell "coke oven odors" at this site. However, the site is on
a ridge and the winds were slightly stronger on the ridges than they were
in the valley. Background levels-appear to be in the 0.4 to 0.7 ppb range
as judged from Site IB and Site 2 when upwind from the coke ovens. The
levels at the airport are relatively high, 2 to 6 ppb, compared to values
obtained at the closer sites. Other sources of benzene at the airport may
have contributed to these higher values.
The winds were generally out of the South and Southeast, but
the direction was rather variable during any given sampling period.
Ashland Chemical Company — September. 1978
The benzene sampling sites for the Ashland Chemical Company maleic
anhydride plant at Neal, West Virginia, are shown in Figure 16. Site No. 1
is on a dairy farm north and east of the plant. Site No. 2 is in Neal, West
Virginia, at a greenhouse at 2244 Big Sandy Road. Site No. 3 is across the
river on U.S. Route 23, west of the Novamont Chemical Company's polypropylene
plant. Site No. 4 is in Neal on Big Sandy Road. Site No. 5 is north of Neal,
across from the southernmost regions of the refinery. Site No. 6 is farthest
from the maleic anhydride plant, south of Neal on Big Sandy Road; and Site
No. 7 is at the corner of Route 52 and Big Sandy Road, slightly north and east
of the plant.
Directly across the river from Sites 1 and 7 is the Ashland refinery.
This is a possible source of benzene which may complicate the interpretation
of the results. Approximately 2 miles north of the plant, there are coke ovens
in Catlettsburg, Kentucky. In addition, there are coke ovens in South Point,
Ohio, which is approximately 3 miles north of the maleic anhydride plant.
These coke ovens are also possible sources of benzene which may make the results
difficult to interpret.
-------
1000
1000 2000 3000 4000 5000 6000 7000 FEET
1 KILOMETER
FIGURE 16. BENZENE MONITORING SITES AT ASHLAND CHEMICAL
-------
61
The meteorological data were obtained from the Tri-State Airport
which is not shown on the map, but which is just slightly over 1-1/2 miles
east of the maleic anhydride plant.
Only air sampling was done at this location, and the results are
given in Table 20.
In general, the highest determinations were made on samples taken
at Sites 1 and 7 which were closest to the maleic anhydride plant, but also
closest to the refinery. At Site 4 the highest level, 29.7 ppb, was de-
termined; but there is no obvious explanation for this high value when all
other values at Site 4 were 1.1 ppb or below. In general, the benzene
levels within the town of Neal (Sites 2, 4, and 6) directly across the river
from Neal (Site 3) were less than 2 ppb.
The prevailing wind was out of the south and west which makes most
of the sites upwind from the maleic anhydride plant. Only Sites 1 and 7
were downwind of the maleic anhydride plant.
There was good agreement between duplicates, and confidence can
be placed in the analytical results. However, because of the location of the
plant in relationship to the river and the refinery, it was difficult to ob-
tain samples that would accurately reflect the benzene emissions from the
maleic anhydride plant.
-------
TABLE 20. ANALYTICAL DATA FOR AIR SAMPLES FROM ASHLAND CHEMICAL COMPANY
Weather Conditions' '
Wind
Sampling
Station Direction
1 260°-270°
240°-120°
3400-2000
Speed,
m/sec
2.6-5.7
1.6-2.6
1.6-3.1
Temp.
Range,
F
85-75
75-70
70-87
Relative
Humidity,
percent
55-80
80-91
91-59
Benzene Analysis
PPD /h\
General Sampling Period (v/v) yg/m3^ '
9/12/78
9/12/78
Cloudy, calm 9/13/78
24
1:428-2225
2240-0904
0904-1647
hour average
9
4
5
6
.3
.0
.1
30.
12.
16.
19
0
9
5
(23
(10
(12
.2-38.6)
.0-16.6)
.8-21.3)
270°-240° 2.6-5.7 85-73 57-84
2400-1200 1.6-2.6 73-70 84-91
340°-180° 1.6-3.1 73-87 91-59
Cloudy, windy
Cloudy, calm
9/12/78 1517-2337 0.3
9/12/78 2337-0949 0.8
9/13/78 0949-1725 1.1
0.9 (0-1.2)
2.7 (2.1-3.5)
3.4 (2.8-4.1) (2)
3 270°-250°
3400-1800
3 1800-2200
2200-3100
4 260°-230°
230°-28QO
280°-340°
4 180°-220°
220°-250°
250°-310°
5 270°-240°
240°-120°
340°-180°
2.6-5.7
1.6-3.1
2.6-4.6
2.6-3.6
2.1-5.7
1.6-2.6
1.6-3.1
2.6-4.6
2.6-4.1
2.6-4.6
2.6-5.7
1.6-2.6
1.6-3.1
85-77
72-87
80-74
80-70
84-72
70-76
76-87
80-75
75-72
70-84
84.74
74-70
74-87
55-80
91-59
68-85
88-55
57-85
85-91
91-59
80-85
82-85
87-40
55-83
83-91
91-59
Very cloudy
Cloudy
Clear,
Cloudy
-
Cloudy
Cloudy
-
Sunny,
Cloudy
-
Cloudy
sunny
, windy
-
, calm
, windy
-
Calm
, windy
-
, calm
24
9/12/78
9/13/78
9/14/78
9/14/78
24
9/12/78
9/13/78
9/13/78
24
9/14/78
9/14/78
9/15/78
24
9/12/78
9/12/78
9/13/78
24
hour average
1345-2155
0825-1625
1112-2130
2130-1112
hour average
1 530-0003
0003-1017
1017-1757
hour average
1323-2040
2040-0425
0425-1430
hour average
1450-2310
2310-0926
0926-1710
hour average
1
0.8
1.7
1.1
0.3
T
1.1
0.7
0.7
r~
29.7
0.3
0.9
9
6.8
0.7
1.6
3
2
2.5
5.4
3.7
1.1
2
3.6
2.2
2.1
2~
95.8
1.1
2.8
29
21.9
2.3
5.3
10
(1.9-3.2)
(4.2-7.0)
(2.9-4.8)
(0-1.4)
(2.8-4.7)
(1.8-2.6)
(1.6-2.7)
(74.3-123
(0-1.4)
(2.2-36)
(18.2-26.
(1.8-3.0)
(4.1-6.8)
(2)
.5)
2) (2)
ro
-------
TABLE 20. (Continued)
Weather Conditions
(a)
Mind
Sampling
Station
Direction
Speed,
m/sec
Temp.
Range,
F
Relative
Humidity,
percent
Benzene Analysis
General
Sampl
ing
Period
ppb
(v/v)
yg/m
3(b)
260°-230°
1200-300°
3400-180°
190°-210°
210°-250°
180°-200°
200°-320°
2.1-5.7
1.6-2.6
1.6-3.1
84-73
74-70
74-87
2.6-4.6 87-77
2.6-4.1 77-70
2.6-4.6 80-75
2.6-4.1 70-84
57-85
85-91
91-59
80-85
85-82
85-75
87-40
Cloudy, calm
Cloudy, calm
9/12/78 1545-0025 0.3
9/13/78 0025-1037 0.7
9/13/78 1037-1813 1.1
24 hour average 1
9/14/78 1355-2100 0.6
9/14/78 2100-0445 0.5
9/14/78 1240-2000 1.0
9/14/78 2000-1345 13.8
24 hour averagelO
1.0 (0-1.3)
2.1 (1.6-2.7)
3.5 (2.7-4.5)
2
2.0 (1.7-2.4) (2)
1.5 (1.2-1.9)
3.2 (2.5-4.1)
44.5 (34.6-57.5)
32
(a) Barometric pressure throughout the sampling was 29.1 to 29.2 in. Hg.
(b) The numbers in parentheses are the 95 percent confidence limits, determined as discussed under the
quality assurance section. A single number appearing in parentheses is the number of replicates,
otherwise the result is based on a single sample.
cr>
CO
-------
64
Service Station Locations
Benzene is an important ingredient of motor fuels, with the amount
fluctuating with the grade of fuel. Samples of the benzene content of U.S.
gasolines in a 1968 study^ ' showed that the mean benzene content averaged
about 1 percent by volume, with a range between 0.54 and 2.0 percent. As a
consequence of the requirements for reduced levels of lead anti-knock addi-
tives, the average benzene percentage in one oil company's gasolines was
increased from less than 1 percent to less than 2 percent between 1974 and
to\
1976.v ' Samples from this company's refineries in 1976 indicated an over-
all average of 1.25 percent among three grades of gasoline with individual
measurements ranging from 0.54 to 2.39 percent.
Selection of Sampling Locations
Ambient air was sampled around three gasoline service station
locations in Columbus, Ohio. Sampling was performed at each location with
the objective of determining maximum benzene concentrations to which in-
habitants of the area are exposed in their daily activities.
The service station locations selected represented different sets
of source conditions. These were:
(1) Three or four stations at a busy intersection or
beside each other along a roadway with residences
nearby. A location at 1-71 and Morse Road was
selected to satisfy this condition.
(2) A single heavily patronized full-service/self -
service station in a residential area. Fishinger
Road and Mountview Avenue was chosen to meet this
condition.
(3) A fully self-service station with residences nearby.
A location at Morse and Maize Roads was chosen to
satisfy this condition.
-------
65
The following additional criteria were used in selecting the
service station locations:
• Maximum throughput of gallons of gasoline pumped
• High traffic density near the station
• Large population affected
• Grades of gasolines available include those
high in benzene content
t Minimum obstructions to wind flow between the
station(s) and the surrounding neighborhood.
Sampling Protocol
Monitoring for atmospheric benzene levels was performed during
the period the stations Were in operation, 16-20 hours, at seven or eight
sites around each location taking both continuous and grab samples of air.
Samples at the stationary monitors were collected in duplicate at
8 to 10-hour intervals on Tenax GC traps at a rate of 30 ml/min. Grab sam-
ples were collected for 20-minute periods at the rate of 200 to 220 ml/min.
The continuous monitors were placed where the impact on people living or
working in the area could be determined and where the changes in concentra-
tion with distance downwind could be distinguished. Four monitors were
placed in the downwind quadrant (to the southeast or northeast depending on
the expected wind direction) at distances of 50, 100, 150, and 200 m from
the station or station complex. One monitor was placed in each of the other
three quadrants between 50 to 150 m from the station complex. Grab samples
were taken at various sites off the station property with the objectives of
estimating areas of peak benzene concentrations, determining the ambient
concentrations directly downwind of the stations, comparing short-term
averages with long-term averages, and measuring the variations in ambient
benzene concentrations between hours of high and low gasoline sales.
During the sampling periods at Locations 2 and 3 (but not Location 1)
wind direction, wind speed, and temperature were monitored with a battery-
operated weather station. Intermittently, additional checks on the wind mea-
surements were made with a hand-held anemometer. Periodic observations of
cloud cover, current weather type, and humidity were also made. Hourly ob-
servations of all these variables were obtained for the sampling periods from
the National Weather Service station at Port Columbus Airport.
-------
66
Analysis of the Benzene Sampling Data
In the discussions of the ambient data from the three service
station locations, an attempt has been made to develop some hypotheses which
are consistent with the measured ambient concentrations of benzene, the
known or suspected sources, and the meteorological parameters. One can
assume a benzene dispersion model which has service station emissions being
injected into the ambient air near the ground (within 1 m of the surface)
with little or no plume rise and then these emissions being diffused laterally
and vertically as they move downwind. The benzene emissions are generated as
a series of puffs from each vehicle's gasoline tank as gasoline is being
pumped into the tank and vapors are being forced out. Some benzene may also
be emitted by evaporation from the hot engine block as the gas pumping opera-
tion proceeds. Under very high ambient temperatures gasoline vapors, includ-
ing benzene, from normal service station operations (pumping and storage) may
be emitted to the atmosphere at an accelerated rate.
All of the ambient measurements made during the service station
monitoring program cannot be explained by this model. In some cases the
explanation probably lies in the uncertainties of the measurement themselves,
for example, short-term variations of wind speed and wind direction. It is
also likely that the sampler network was unable to pick up the maximum benzene
concentration during each sampling period. In some cases, the discrepancy
between hypothesis and observation was probably a benzene source which was
not included in the original model.
One source of benzene emissions which was unexpectedly significant
in this measurement program was motor vehicles -- in traffic, idling, or
parked. Concentrations at several sampling sites were apparently the result
of benzene evaporated from engine blocks or emitted from exhausts. Automobile
emissions were suspected as a source when: (a) no service station was upwind
of the sampler, (b) the benzene concentration was higher than background, and
(c) the increases of benzene concentration during the day coincided with the
presence of automobiles upwind of the sampler. Some of the highest 8-hour
averages were obtained at sites which appeared to be more auto-traffic related
than service-station related.
-------
67
At one service station location the sampling program was fortunate
to include a period when a tank truck was filling the underground storage
tanks. The highest grab sample concentrations of benzene measured during
the entire service station sampling program were obtained downwind of the
station during this operation. Eight-hour average concentrations at the
samplers in the vicinity of these grab samples were markedly higher during
this period than the 8-hour concentrations downwind of the station during
other periods.
Location 1. The Morse Road and 1-71 Monitoring
(Friday, June Id—Saturday. June 11. 1977)
This location (Figures 17, 18, and 19) fulfilled source condition
(1) and comprised three adjacent stations (T, S, and H) on the south side of
Morse Road and one station (M) north of Morse Road and directly across from
Station T. The location was approximately 200 m east of the I-71-Morse Road
interchange. The point where the 1-71 exit and entrance ramps join Morse
Road is depressed below the area of the service stations. A motel at the
same elevation as the service stations lies on the south side of Morse Road
between the stations and the 1-71 exit ramp. A group of two-story apartment
buildings with several mature trees is located on the south side of Morse Road
to the east of the stations. There are three 50-car and six 10-car parking
lots in this complex. One-story houses with a few trees are located to the
south of the service stations. On the north side of Morse Road there are no
trees. It is generally open with an auto agency, a theater, and a restaurant
being the closest buildings.
While this location was chosen for the high density of service sta-
tions it is also an area of heavy traffic -- commuters into Columbus via
Interstate 71 and vehicles going to or from the shopping complexes which are
on Morse Road east of the sampling location. Morse Road is a six-lane
thoroughfare east of 1-71
Table 21 lists the ambient benzene concentrations measured at each
sampling site, the gasoline pumped at the various service stations, the ben-
zene content of the gasoline, and the range of meteorological parameters
measured at the Port Columbus Airport during the sampling. Hourly values of
the meteorological variables measured at the airport's National Weather
Service station during the sampling are provided in Table 22. No continuous
meteorological measurements were taken at Location 1.
-------
co
M
(0
O.
M
0)
Norse Road.
13.7
Motel
0.7 1515
G^T,5,6
A
0.9
10
Auto Sales
June 11
Theater Prevailing Winds
::::«S' T VA*:::
Wft 1 SAV:
Homes
A
G-2
1.8 1545
3.9 1620
G-3
A
• 1.2
5
Homes
A c-4
25 m
Apartments
•6.7
Traffic Light
• Continuous Monitors
A Grab Samples
Service Stations
FIGURE 17. LOCATION 1 — BEf. .,c CONCENTRATIONS (yg/m ) AT MORSE ROAD
AND 1-71 DURING PERIOD 1 (1320-2100, JUNE 10, 1977)
-------
:v
\
co
CO
rt
tx
0)
S
.H
f-.
1
M
v \
V
\
\
\
\ -
\\
\ \
\ \
G-1,5
A
,6
Auto Sales
••.v.'rv»*.vX'»»Xi:
HM.!!
1 9 une 10 /f vJl
i.y /r vd»
\ — =^H
\ ^y
v June 11
Theater Prevailing Winds
Morse Road . ^
•/
/
/
I/
V
/
/
1
/£! Motel
/ 7
/
/
/
r
•31
Homes
, i S:
A
G-2
Oc
. J
•
1
•
;l"u"«;
m^,M
2f\ i
,2'
• 0 4
5
Homes
G-3
A
fe
9
2
A c-4
3.4
• ^
Apartments
• 5,5
3
N
4 Traffic Light
• Continuous Monitors
A Grab Samples
Service Stations
25 m
FIGURE 18. LOCATION 1 -- BENZENE CONCENTKAiiuiw vHy/m ) AT MORSE ROAD A.ND
1-71 DURING PERIOD 2 (2100, JUNE 10 — 1000, JUNE 11, 1977) •
-------
(0
w
m
a.
M
a)
Morse Road
Motel
G-1,5
Ai
5
.9
.0
,6
1200
1600
Auto Sales
•:.£v:;.*;viv>»x*»:
11.M..II
||:"T"||
A
G-2
0.8
Homes 0
1
luCi
«
2.1 June *° ^ v
\ -^
\ *
^' TN ,
^ June 11
Theater Prevailing Winds
G-3
A
H.8
2
•1.4
5
Homes
0.9
Apartments
• 1.7
3
^Traffic Light
• Continuous Monitors
A G-4
2.7 1115
AGrab Samples
Service Stations
- 25 ra
FIGURE 19. LOCATION 1 -- BENZENE CONCENTRATIONS (yg/nf) AT MORSE ROAD
AND 1-71 DURING PERIOD 3 (1000-1800, JUNE 11, 1977)
-------
TABLE 21. BENZENE ANALYSES OF AIR SAMPLES FROM SERVICE STATION LOCATION AT 1-71 AND MORSE ROAD
Weather Conditions
Wind
Temp.
Atmos Relative
Benzene
Sampling Speed, Range, Pressure, Humidity, ppb
Station Direction m/sec F mm Hg percent General Sampling Period (v/v)
1 320-360 1.5-4.1 71-61 742-745 43-54 Clear 6/10,
030-calm 2.1-calm 61-48 741-742 52-86 Partly cloudy 6/10,
Calm-250 1.5-4.1 56-62 739-742 60-75 Overcast inter- 6/11,
mittent rain
2 Ditto Ditto Ditto Ditto Ditto Ditto 6/10,
6/10,
6/11,
3 Ditto Ditto Ditto Ditto Ditto Ditto 6/10,
6/10,
6/11,
1240-2145
2145-0825
0825-1800
29 hr avg
1320-2300
2300-0950
0950-1800
29 hr avg
1330-2235
2235-1010
1010-1800
29 hr avg
0
0
0
0
1
0
0
0
2
1
0
1
.8
.2
.3
.5
.0
.7
.6
.8
.1
.7
.5
.5
2
0
0
1
3
2
1
2
6
5
1
5
.5
.5
.8
.2
.2
.8
.7
.5
.7
yg/m
(2.1-3.0)*
(0-0.7)
(0-1.0)
(2.7-8.8)*
(1.8-2.6)*
(1.5-2.1)*
(5.1-8.7)
(4.3-7.1)
(1.3-2.2)
Ditto Ditto Ditto Ditto
Ditto Ditto Ditto Ditto
Ditto Ditto Ditto Ditto
Ditto Ditto
Ditto Ditto
Ditto Ditto
6/10,
6/10,
6/11,
6/10,
6/10,
6/11,
6/10,
6/10,
6/11,
1355-2320 2.0
2320-1030 1.1
1030-1800 0.3
28 hr avg 1.1
1300-2210 0.4
2210-0845 0.1
0845-1800 0.4
29 hr avg 0.3 1
6.3 (4.9-8.1)
3.4 (2.7-4.4)
0.9 (0-1.2)
4
1.2 (0-1.6)
0.4 (0-0.6)
1.4 (1.1-1.8)
1410-2340 0.3
2340-1045 0.6
1045-1800 0.7
28 hr avg 0.5
0.9 (0-1.1)*
1.9 (1.6-2.3)*
2.1 (1.6-2.7)
2
7 Ditto Ditto Ditto Ditto Ditto Ditto 6/10,
6/10,
6/11,
1220-2120
2120-0745
0745-1800
30 hr avg
4
1
0
2
.3
.3
.6
.0
13
4
1
7
.7
.3
.9
(1
(3
(1
0
f
,
.6-1
3-5.
6-2.
7.
5)
3)
6)
*
-------
TABLE 21. (Continued)
Wind
Sampling
Station Direction
G-l
6-2
G-3
6-4
6-5
G-6
360
360
360
250
130
250
Speed,
m/sec
4.1
4.1
4.1
1.5
2.6
3.6
Weather
Temp.
Range,
F
71 .
71
71
62
62
62
Conditions
Atmos
Pressure
mm Hg
743
743
743
741
740
742
Relative
, Humidity,
percent
Grab Samples
47
46
43
60
75
60
General Sampling Period
Clear 6/10, 1510-1530
Clear 6/10, 1539-1559
Clear 6/10, 1608-1633
Overcast 6/11, 1106-1129
Overcast 6/11, 1545-1605
Overcast 6/11, 1140-1200
Benzene
ppb
(v/v)
0.2 0.
0.6 1.
1.2 3.
0.8 2.
1.6 5.
0.9 2.
Analysi
yg/m
7 (0-0.9
8 (1.4-2
9 (3.0-5
7 (2.1-3
0 (3.9-6
9 (2.2-3
s(a)(b)
.3)
.0)
•5)
-5)
.7)
(a)
(b)
The numbers in parentheses are the
assurance section. Values marked
a single sample.
Gasoline Data:
Station Date
T
S
H
M
6/10,
6/11,
LV %
6/10,
6/11,
LV %
6/10,
6/11,
LV %
6/10,
LV %
0600-2200
0600-2200
Benzene
0600-2200
0600-2200
Benzene
0600-2400
0600-2400
Benzene
1200-1800
Benzene
95 percent confidence 1
* are based on duplicate
Gallons Pumped
Regular
1.1
1.4
1487
1650
0.6
1.6
imits, determined as discussed
determinations, otherwise the
under the
result is
quality
based on
Unleaded Premium Totals
0.9 1.
1.3 1.
5182 1095
5700 1200
0.5 0.
1.3
750
800
0 1550
3400
3400
3 6800
7764
8500
5 16264
1184
ro
-------
73
TABLE 22. HOURLY WEATHER OBSERVATIONS FOR 1-71 AND MORSE ROAD LOCATION
(a)
Time
1200
1300
1400
1500
1600
1700
1800
1900
2000
2100
2200
2300
2400
0100
0200
0300
0400
0500
0600
0700
0800
0900
1000
1100
1200
1300
1400
1500
1600
1700
1800
Pressure,
mm Hg
745
744
744
743
743
743
743
743
743
742
742
742
743
743
743
742
742
742
741
742
742
741
741
742
742
741
740
740
739
739
739
Temp,
F
65
66
69
71
71
71
71
70
68
65
61
60
58
55
55
55
54
52
48
49
56
61
61
62
62
62
62
61
62
62
61
Wind
Speed,
Direction m/sec
June 10
360
350
320
360
360
360
360
350
350
030
030
030
040
June 11
040
Calm
ii
H
ii
ii
ii
it
170
130
170
250
160
150
140
130
120
120
1.5
3.6
2.6
4.1
4.1
3.6
3.6
3.1
3.1
2.1
2.1
2.1
2.1
2.1
—
•--
—
-_
—
--
—
3.6
2.6
1.5
3.6
3.1
4.1
2.1
2.6
3.1
4.1
Relative
Humidity
percent
54
52
47
47
46
43
44
44
49
52
62
64
72
72
69
66
69
74
86
83
72
60
60
60
60
60
60
67
75
75
72
5
Character
Clear
Ditto
H
ii
M
n
H
n
M
Partly Cl
Ditto
II
II
Partly Cl
Ditto
n
n
n
n
n
n
Overcast
Di tto
Overcast,
mittent
Ditto
n
M
n
n
n
n
oudy
oudy
inter-
rain
(a) National Weather Service data.
-------
74
During the first two sampling periods (Figures 17 and 18) at
Location 1 (approximately 1300 to 2100 local daylight time on Friday,
June 10, and 2100 on June 10 to 1000 on June 11) the most significant
observations were the high benzene concentrations to the east and west
of the service station complex in comparison with concentrations down-
wind (south) of the complex. Eight-hour concentrations at the downwind
sites (Numbers 1, 2, and 3) were from 2 to 10 times smaller than concen-
trations on the east (Sites 3 and 4) and west (Site 7). A possible ex-
planation of these higher concentrations crosswind from the service stations
is motor vehicle emissions. Sites 3 and 4 are in the apartment complex. It
can be hypothesized that over 150 automobiles returned to th.is area during the
evening of the 10th and benzene from the engines evaporated while the engines
cooled off. Low wind or calm conditions during the night (Period 2) in ad-
dition to a nocturnal temperature inversion caused the benzene to remain in
the area. Benzene concentrations in this apartment area during Period 3
(Figure 19) were among the lowest measured by the monitoring network. During
this last period (approximately 1000 to 1800 LOT on June 11) the evaporated
benzene was dispersed by light winds and unstable conditions while little new
benzene was evaporated in the apartment parking lots.
Site 7 was on the east side of the exit ramp from northbound 1-71.
This exit ramp handles a large volume of evening commuter traffic from down-
town Columbus as well as cars going to the shopping centers on the eastern
portion of Morse Road. At 6:00 p.m. cars are lined up in this exit ramp
and along one lane of 1-71 waiting to enter Morse Road going both east and
west. Morse Road passes under 1-71 and the exit ramp is depressed beneath
the level of the service station complex. The high benzene concentrations
at Site 7 during sampling Period 1 are possibly the result of "hot soak"
emissions from automobiles waiting to enter Morse Road plus emissions blown
from 1-71 toward the sampler by the north wind. Another important source
for this site is an oil company's small tank field about 1 km to the north.
Because Morse Road and the exit-entrance ramps are depressed in this area
benzene probably built up in this topographical depression. The high concen-
tration remained there during the light-wind and temperature-inversion condi-
tions of the nighttime portion of Period 2.
-------
75
Ambient benzene concentrations which are probably related to
gasoline station emissions did show a decrease with increasing distance.
In Period 1 the highest long-term measurement was 3.7 yg/m at Site 2 dropr.'..
ping off to 1.2 yg/m at Site 5 further downwind. As a consequence of the
southeast wind direction during Period 3 (Figure 19) the highest ambient
concentrations attributable to the service stations probably were to the
northwest of the stations in the area of grab sample Sites 5 and 6. This
supposition is borne out by the 5.0 yg/nP observation at G-5 at 1600 and
the fact that Site 6 measured the highest concentration in the network
during the period.
Background benzene concentrations upwind of service stations, high-
o
ways, and parking lots appear to be about 1.0 yg/m . Support for this con-
clusion comes from the measurements at Site 6 during Period 1 and Sites 1
and 4 during Period 3. However, it should be noted that the minimum long-
term observations were about 0.5 yg/m3 at Sites 1 and 5 during Period 2.
Grab samples at Sites G-2 and G-3 during Period 1 are in the range
of values that would be expected from the long-term measurements. However, it
is difficult to explain the grab-sample measurement of 2.7 yg/m3 taken at G-4
during mid-morning of Period 3. No known large source is near and upwind of
this site.
Location 2. The Fishinger Road
and Mountview Avenue Monitoring
(Thursday. July 7, 1977J
This location (Figures 20 and 21) fulfilled source condition (2)
and comprised a single full-service/self-service station (Station D) at the
southwest corner of Fishinger Road and Mountview Avenue. The area around
the church north of the service station is open while the homes to the west,
south, and east of the station have mature trees. The home on the southeast
corner of Fishinger Road and Mountview Avenue is shielded from the service
station by a hedge six feet high. Shoreham Road which is parallel to
Fishinger on the north is depressed below the level of Fishinger. Fishinger
Road is four lanes wide and serves as a feeder to two commuter arteries at
its east and west ends each about 1.5 km from Mountview.
-------
Shoreham Road
G-l,
A0.7 0800
1.7 0930
Open Field
Fishinger Road
0.6
Homes
Church
Station D
House-^
\
House— >| |
Prevailing —.
Homes
1.5
2.5 1000
G-6
a)
0)
•H
AG-5
2.3
Residential
Traffic Light
0 Continuous Monitors
A Grab Samples
FIGURE 20. LOCATION 2 -- BENZENE CONCENTRATIONS (ug/ni ) AT FISHINGER ROAD
AND MOUNTVIEW AVENUE DURING PERIOD 1 (0700-1500, JULY 7, 1977)
-------
Shoreham Road
A
G-1,2
Open Field
Church ^
A • 6 2-°
^
2s i 7 ^f»
• -> -L / JU
Flshinger Road
Station D
• 2 . *~ — t
~ 1 axcKii
2 4 House^ ) ;
A |
i ft 11
-L • O O " ' L'-1"-'*-"-1'-'1*-"-"!-' i —
Homes . ' 5 ^
House— ^[ |
3^
Prevailing -^s "^=jj«^'
^ Homes
3.2
'
/ *,
/ / • !
/ / A 2'7 AO /
/ / ^ G-3 AG-A
•
fi-fi An-s
b b AG 5 ^
m2.; 1700 "„
5.0
§1 • 8 Residential
< 3.7
3
•H
> A
^ f Traffic Light
3
O
s 0 Continuous Monitors
A Grab Samples
^ Weather Station
t—=J
20 M
FIGURE 21. LOCATION 2 -- BENZENE CONCENTRATIONS (yg/mJ) AT FISHINGEk
AND MOUNTVIEW AVENUE DURING PERIOD 2 (1500-2000, JULY 7, 1977)
-------
78
This sampling took place in a residential neighborhood along a
busy road, but one that does not carry nearly the amount of traffic that
travels over Morse Road. There was only a single station at the Fishinger
Road location and the volume of gasoline pumped was considerably less than
at the Morse Road station complex. The benzene content of the gasolines at
this location were similar to those at Location 1.
Despite the differences in amounts of gasoline pumped, the benzene
concentrations downwind of the Fishinger Road service station were of a mag-
nitude comparable to those measured downwind of the stations on Morse Road.
One explanation for the Fishinger Road concentrations being larger than ex-
pected may lie in the meteorological conditions -- specifically, the ex-
tremely high temperatures (over 35 C). These temperatures could have caused
an increase in the benzene vaporization rate which resulted in higher ambient
concentrations.
Sampling at this location was conducted simultaneously by Battelle
and by personnel from the Sohio Research Center, Cleveland, Ohio. Sohio col-
lected NIOSH charcoal tube samples (600 mg) at 8-hour intervals with a sam-
pling flow rate of 1 &/min.. The benzene analyses of these tubes by Sohio
personnel were in good agreement with Battelle data (see values in brackets
in Table 23).
An MRI Mechanical Weather Station was placed on the roof of a
church across the street from the service station. Continuous measurements
of wind speed and wind direction were made with this instrument. Hourly
meteorological readings taken by the National Weather Service at the Port
Columbus Airport are presented in Table 24. Variables between the National
Weather Service hourly average wind direction measurements and those made on
location ranged from -40° to +40° with an average deviation of 17°.
During the first sampling period, extending from approximately 0700
to 1500 local daylight time (Figure 20), benzene measurements at the con-
tinuous sampling sites indicated a pattern of benzene transport and diffusion
along the downwind direction. The maximum observation in the neighborhood was
3 3
2.9 yg/m made at Site 4. Benzene concentrations dropped off to 1.2 yg/m
further downwind at Site 1. Based on the pattern of 8-hour measurements, the
highest concentrations in the neighborhood probably occurred in the area just
o
north of Sites .G-5 and G-6. The grab sample measurement of 5.9 yg/m taken
at G-5 about 9:30 a.m. fits this pattern.
-------
TABLE 23. BENZENE ANALYSES OF AIR SAMPLES FROM SERVICE STATION LOCATION
AT FISHINGER ROAD AND MOUNTVIEW AVENUE
Weather Conditions
Sampling
Station
1
2
3
4
5
6
7
8
Wind
Direction
270-360
225-270
270-360
225-270
225-270
270-360
225-270
270-360
225-270
270-360
225-270
270-360
225-270
270-360
225-270
Speed,
m/sec
4.6-7.2
7.2-9.3
4.6-7.3
7.2-9.3
7.2-9.3
4.6-7.2
7.2-9.3
4.6-7.2
7.2-9.3
4.6-7.2
7.2-9.3
4.6-7.2
7.2-9.3
4.6-7.2
7.2-9.3
Temp,
F
86
94
87
94
87
89
99
86
85
86
86
Atmos Relative
Pressure, Humidity,
mm Hg percent
760 48-74
760 45-48
Ditto 48-74
45-48
45-48
48-74
45-48
48-74
45-48
48-74
45-48
48-74
45-48
48-74
45-48
General
Clear-cloudy
Cloudy
Clear-cloudy
Cloudy
Cloudy
Clear-cloudy
Cloudy
Clear-cloudy
Cloudy
Clear-cloudy
Cloudy
Clear-cloudy
Cloudy
Cloudy
Rain
Sampling Period
0600-1435
1435-2030
14 hr avg
0715-1525
1525-2000
13 hr avg
1155-2005
0640-1550
1550-2035
13 hr avg
0700-1555
1555-2015
0540-1410
1410-2100
15 hr avg
0730-1505
1505-2030
13 hr avg
0815-1615
1615-2050
13 hr avg
Benzene Analysis^3'' '
ppb
(v/v)
0.4
0.8
0.7
0.2
0.8
0.5
0.6
0.9
1.6
1.3
0.7
Lost
0.4
0.6
0.5
0.5
1.0
O"
0.7
1.2
wg/m
1.2 (0-1.4)*[<11
2.7 (2.1-3.5)[<11
2
0.6 (0-0. 8) [<1]
2.4 (1.9-3. !)[
-------
TABLE 23. (Continued)
Weather Conditions
Sampling
Station
Wind
Speed,
Direction m/sec
Temp,
F
Atmos
Pressure,
mm hg General
Benzene )
ppb
Sampling Period (v/v)
\nalysis(a)(b)
yg/m
Grab Samples
6-1
G-2
6-3
6-4
6-5
6-6
6-7
6-8
250
240
260
260
280
210
210
4.1
2.6
5.7
5.7
7.2
8.2
9.3
Incomplete sample
83 760
81 Di tto
86
86
90
95
95
- rain
Clear-hazy
Clear
Hazy
Hazy
Hazy
Cloudy
Cloudy
0900-0905
0740-0800
0929-0949
0955-1005
1200-1220
1645-1715
1725-1745
0.5
0.2
1.9
0.8
Lost
0.8
0.8
1.7 (1.3-2.2)
0.7 (0-0.9)
5.9 (4.6-7.6
2.5 (1.9-3.2)
2.7 (2.1-3.5)
2.5 (1.9-3.2)
(a) The numbers in parentheses are the 95 percent confidence limits, determined as discussed under
the quality assurance section. Values marked * are based on duplicate determinations, other-
wise the result is based on a single sample. Values in brackets are results of analyses re-
ported by personnel of Sohio Research Center of charcoal tube samples taken simultaneously
with the Battelle samples. The unit is ppb.
(b) Gasoline data:
Station
D
Date
7/7. 0600-2030
LV % Benzene
Regular
0.3
Gallons Pumped
Unleaded Premium Totals
1.0
0.6
2000 (estimate)
CO
o
-------
81
TABLE 24. HOURLY WEATHER OBSERVATIONS.FOR FISHINGER ROAD AND
MOUNTVIEW AVENUE LOCATION!9)
Time
0600
0700
0800
0900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
Pressure,
mm Hg
760
760
700
760
760
760
760
760
760
759
760
760
760
760
759
Temp,
F
79
75
81
83
86
88
90
91
94
94
95
95
94
93
91
Wind
Direction
July 7
240
240
240
250
260
291
280
270
230
240
240
210
220
230
250
Speed,
m/sec
4.6
4.6
2.6
4.1
5.7
5.2
7.2
6.2
712
8.8
8.2
8.2
9.3
9.3
8.8
Relative
Humidity,
percent
74
77
69
67
63
61
58
54
48
48
45
45
46
46
47
Character
Clear
Ditto
ii
Hazy
Ditto
ii
ii
Cloudy
Ditto
M
n
H
n
n
Cloudy, rain
(a) National Weather Service data.
-------
82
All grab samples taken in the 9:30 to 10:00 a.m. period are higher
than the longer term averages which may indicate a benzene buildup over the
entire area during this time.
Site 2 was selected as an upwind background measurement for this
3
location. A comparison of the Site 2 measurement of 0.6 yg/m and the Site 6
measurement of 1.3 yg/m offers some indication of the magnitudes of benzene
emissions from Fishinger Road Traffic.
During the second sampling period extending from 1500 to 2000 local
daylight time (Figure 21) the same general pattern of benzene distribution
prevailed downwind of the service station as existed during Period 1. However,
the concentrations were 1.5 to 2.0 times as great. The increase resulted from
an areawide benzene increase as well as an increase in emissions from the
service station.
The reading at Site 2's background sampler was 2.5 yg/m . This was
3 3
larger than the measurements at Sites 6 (2.0 yg/m ) and 3 (1.8 yg/m ) which
were downwind of Site 2. This suggests that benzene from some other source
may have affected Site 2 during this period.
From the general pattern of benzene distribution downwind of the
service station, it is suggested that the maximum 5-hour concentration in the
3
residential neighborhood was probably in excess of 6 yg/m and occurred in
the vicinity of Sites G-5 and G-6. However, the grab sample taken at 1700 at
3
G-6 was only 2.7 yg/m . An examination of the on-location wind direction
trace for this period indicated that the winds varied from southwesterly to
westerly to northerly. The low reading for the G-6 samples was possibly made
when the wind was not blowing directly from the station to the sampler.
Location 3. The Morse Road and Maize Road Monitoring
(Friday, August 5--Saturday, August 6, 1977T
This location (Figures 22 and 23) fulfilled source condition (3)
and comprised a full self-service station (Station P) located on the north-
east corner of the Morse Road and Maize Road intersection. This corner is
approximately 1000 m east of the first location. A station (Station B)
which was closed during the sampling at Location 3, is on the southeast
corner of the Morse and Maize Roads intersection. A full-service station
(Station L) is located on the soutwest corner of the intersection. With
-------
7.6-1415
Station P
I *
I Office,'
I Bldg
I
Prevailing Winds
Morse Road (8 lanes)
i A G~2 2.1 1115
X
1.0 1140 A I
G-3
Station
Restaurant
••a
fl
N
•rl
• 7 1
Station 0.9 ft FtolH
n nil upen riexu
B Bank r
fc ,AG-I, 10
1.2 1000
4
| I
1
* Restaurant |
L J
t Traffic Light
^ Continuous Monitors
00
CO
25 m
Weather Station
FniURE 22. LOCATION 3 — BENZENE CONCEi\...,-,, 10110 vpg/md) AT MORSE AND
MAIZE ROADS DURING PERIOD 1 (0800-1800, AUGUST 5, 1977)
-------
Station P»
w
*
2.5
w 5.6 1945
| , G-4,5,6,7
Prevailing Winds
G-8,9 32.4 2230
A
Morse Road (8 lanes)
A
Station
G-3 L
1 Restaurant
•o
n>
&
0)
N
•H
M
_ ' • 7 1.4 i I
SLUL1°" _ , 1 Open Field 1 I
B ?ank | 1 Restaurant |
_ !AG-I, 10 1 j
0.92315
^ ^ Traffic Light
0 Continuous Monitors
A fJrnh Samoles
25 m
Weather Station
FIGURE 23. LOCATION 3 -- BENZENE CONCENTRATIONS (ug/rn ) AT MORSE AND MAIZE
ROADS DURING PERIOD 2 (1800, AUGUST 5 -- 0200, AUGUST 6, 1977)
-------
85
the exception of a small three-story office building one block west of
Station P the area west, north, and east of the station is composed of
homes and commercial establishments which are one-story buildings. On
the south side on this stretch of Morse Road the area is generally flat
and open with only a small one-story bank located to the east of Station B.
There are no mature trees in this vicinity.
The sampling was conducted at a location approximately 1/4 mile
to the east of the first sampling area (Location 1) on Morse Road. This
location was deemed to be beyond the influence of the other service sta-
tion complex. Even if this assumption were not true, the wind direction
during the third sampling program would have transported the benzene emis-
sions from 1-71 and Morse Road along a line generally to the west of
Maize Road.
On-location wind measurements were made with the Mechanical
Weather Station in an open field south of Morse Road. Hourly weather data
taken by the National Weather Service at Port Columbus are presented in
Table 25. Variations between hourly average measurements made at the sampling
location and at the airport ranged from 0° to 40° with an average 16°
deviation for the entire sampling time. Hourly average wind directions at
Location 3 varied from 200° to 250° while at the airport the range was be-
tween 180° and 240°.
At this third location, Site 7 served as a background measurement
upwind of Morse Road and Site 5 measured benzene concentrations downwind of
this thoroughfare at a point well removed from service stations. The mea-
3 3
surements of 0.9 yg/m at Site 7 and 2.4 yg/m at Site 5 represent the
background values upwind and downwind of Morse Road against which the other
measurements can be compared. Measurements at the first sampling location
also indicated that 1 yg/m might be a reasonable benzene background concen-
tration for a suburban-residential area such as Location 2.
In the first sampling period (Figure 22, Tables 25 and 26) the
winds were generally from the southwest. To the south of Morse Road (upwind)
long-term and grab-sample measurements indicated benzene concentrations of
o
about 1 yg/m . On the downwind side of the road long-term concentrations
ranged from 2.0 to 2.8 yg/m . Decreasing concentrations from Site 1 to
-------
86
TABLE 25. HOURLY WEATHER OBSERVATIONS FOR MORSE AND
MAIZE ROADS LOCATION^)
Time
0600
0700
0800
0900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2100
2200
2300
2400
0100
0200
Pressure,
mm Hg
765
765
766
766
766
766
766
766
766
765
765
765
765
765
765 .
765
765
765
765
766
765
Temp,
F
73
73
74
75
80
82
85
87
87
87
90
88
88
88
86
83
81
79
76
75
75
Wind
Direction
August
186
220
190
190
230
220
210
240
230
230
240
220
200
210
220
200
190
210
190
180
180
Speed,
m/sec
5
4.1
3.6
4.1
4.6
5.2
5.7
6.1
8.2
9.3
7.2
8.8
4.7
8.8
5.2
7.2
3.6
3.6
3.1
3.1
2.6
2.6
Relative
Humidity,
percent
82
82
79
76
69
65
59
57
55
55
49
50
54
50
53
59
63
67
71
74
74
Character
Cloudy
Ditto
ii
n
Partly cloudy
Cloudy
Ditto
n
n
n
Partly cloudy
Di tto
n
n
n
n
n
n
n
ii
n
(a) National Weather Service data.
-------
TABLE 26. BENZENE ANALYSES OF AIR SAMPLES FROM SERVICE STATION LOCATION AT MORSE AND MAIZE ROADS
Weather Conditions
Wind
Sampling
Station Direction
1 200-240
200-220
2 190-240
180-210
3 190-240
180-210
4 190-240
180-210
5 199-240
180-210
6 190-240
180-210
7 190-240
180-210
Speed,
m/sec
4.7-9.3
3.6-8.8
3.6-8.8
2.6-7.2
Ditto
Ditto
Ditto
Ditto
Ditto
Temp.
Range ,
F
82-88
88-83
73-88
88-75
Ditto
Di tto
Ditto
Ditto
Ditto
Atmos
Pressure,
nun Hg
765
765
Ditto
Ditto
Ditto
Ditto
Ditto
Ditto
Relative
Humidity,
percent
54-59
54-59
54-82
54-74
Di tto
Ditto
Ditto
Ditto
Ditto
Benzene Analys
General
Cloudy
Cloudy
Ditto
Di tto
Ditto
Ditto
Ditto
Ditto
ppb
Sampling Period (v/v)
1138-1820
1820-2100
9 hr avg
0700-1745
1745-0153
19 hr avg
0735-1840
1840-0142
18 hr avg
0825-1855
1855-0130
18 hr avg
0806-1910
1910-0118
17 hr avg
0705-1810
1810-0200
19 hr avg
0850-1935
1935-0210
17 hr avg
0.9
1.4
1.2
0.7
0.7
0.7
0.6
1.3
1.0
0.7
1.1
0.9
0.8
0.7
0.8
0.8
0.6
0.7
0.3
0.4
0.3
2
4
4
2
2
2
2
4
2
3
2
2
2
2
2
1
2
0
1
1
.8
.5
.4
.5
.0
.0
.2
.3
.4
.3
.6
.8
.9
.4
is(a)(b)
yg/m
(2
(3
(1
(1
(1
(3
(1
(2
.2-3.6)
.5-5.8)
.9-3.1)
.9-3.2)
.6-2.6)
.3-4.8)*
.7-2.8)
.8-4.0)*
(1.9-3.1)
(1.9-2.8)*
(2.0-3.3)
(1.5-2.1)*
(0-1.1)*
(1.1-1.8)
00
-------
TABLE 26. (Continued)
Sampling
Station
Wind
Speed,
Direction m/sec
Weather
Temp.
Range ,
F
Atmos
Pressure,
mm Hg
Relative
Humidity,
percent General
Benzene
ppb
Sampling Period (v/v)
Analysis(a)(b)
yg/m
Grab Samples
G-l
G-2
G-3
G-4
G-5
G-6
G-7
G-8
G-9
G-10
210
210
210
230
240
220
210
220
220
210
4
4
6
9
4
4
4
3
3
3
.6
.7
.2
.3
.7
.7
.7
.6
.1
.1
75
85
85
87
89
89
88
81
81
79
766
766
766
766
765
765
765
765
765
765
69
65
59
55
50
54
53
67
67
71
Cloudy 0943-1003
Ditto 1102-1122
1139-1158
1414-1434
1644-1704
1710-1730
1930-1950
2217-2227
2227-2237
2303-2323
0.4
0.7
0.3
2.4
1.2
2.4
1.8
21.5
10.3
0.3
1
2
1
7
3
7
5
68
32
0
.2
.1
.0
.6
.9
.6
.6
.6
.4
.9
(0-1.6)
(1.6-2.7)
(0-1.3)
(5.9-9.8)
(3.0-5.0)
(5.9-9.8)
(4.4-7.2)
(54.5-90.5)
(25.2-41.7)
(0-1.2)
oo
00
(a) The numbers in parentheses are the 95 percent confidence limits, determined as discussed under the quality
assurance section (page B-5). Values marked * are based on duplicate determinations, otherwise the result
is based on a single sample.
(b) Gasoline data:
Station
Date
8/5,
0600-1700
1700-1900
1900-0230
LV % Benzene
8/5. 0600-0600
LV % Benzene
Gallons Pumped
Regular
2960
1025
1745
1.2
2900
0.8
Unleaded
1191
379
662
0.7
4350
1.1
Premium
1550
1.2
Totals
4151
1404
2407
7692
8800
-------
89
Site 6 to Site 2 could be attributed to downwind dispersion of the traffic
emissions. The position of Site 5 would recommend it as a good control site
to compare concentrations downwind of the road at a place without a gasoline
station. Thus, it was believed that the benzene concentration at Site 5
could be deducted from the concentrations at Sites 3 and 4 to isolate the
effect of the Station P emissions at Morse and Maize Roads.
Unexpectedly, the long-term benzene concentrations at Sites 3 and 4
during this period were less than the concentration at Sites 5, 1, 6, and 2.
Unless there were some unidentified benzene sources affecting Sites 5, 1, 6,
and 2, or some chemical reaction destroying benzene, or some microscale dis-
persion fluctuations causing the benzene from Station P to bypass Sites 3 and
4, the measurements at these two sites are suspect.
During the second sampling period at Location 3 (Figure 23) the
north side of Morse Road was again downwind in relation to traffic emissions.
The magnitude of traffic emissions of benzene during this time is best par-
's
trayed by the difference between the measurements at Site 5 (2.3 yg/m ) and
Site 7 (1.4 yg/m3).
There was considerable variation among the benzene measurements at
the other monitoring sites. Some of these variations can be explained, at
least partially, by differences in sampling times or sources. However, there
are still some unexplained measurements.
The largest concentration during this period was observed at Site 1
(4.5 yg/m ). During Period 1 the gradient from Site 1 to Site 2 was attrib-
uted to downwind dispersion. In Period 2 there was no uniform gradient.
Sampling at Site 1 during this period lasted only 2.5 hours while at the
other sites the sampling covered 5 to 8 hours. The 2.5 hours at Site 1 was
done between 6:30 p.m. and 9:00 p.m. when auto traffic and activities at the
bank and at Station L were at a maximum for the period. If the benzene con-
centration at Site 1 had averaged 2.5 yg/m for an additional 4.5 hours, the
overall concentration average would have been 3.2 yg/m which is more in line.
with the concentrations at Sites 2 and 5.
3
The 1.8 yg/m concentration at Site 6 is lower than would be
3 3
predicted considering the 2.5 ug/m at Site 2 and the potential 3.2 yg/m
at Site 1.
-------
90
For 20 minutes at about 10:30 p.m. a tank truck refilled the
underground gasoline storage tanks at Station P. During this operation,
two 10-minute grab samples (G-8 and G-9) were taken downwind of the station
near continuous sampling Site 4. Average concentrations for these 10-minute
3 3
periods were 68.6 yg/m and 32.4 yg/m , which were the highest levels ob-
served during the service station monitoring program.
It would seem that the large benzene levels during this 20-
minute period would result in concentrations at Sites 3 and 4 which were
higher than other sites (e.g., 5 and 2) not downwind of the service station.
What is revealed by a computation, however, is that the effect of the tank
truck emissions should have resulted in an even higher concentration at
Site 4 than was actually measured. Even if the prevailing benzene concen-
iu1
3
3
tration for the remainder of Period 2 was only 2.2 yg/m , then the 20 minutes
of tank-truck emissions should have caused an overall average of 4.7 yg/m"
instead of the 3.3 yg/m that was observed. The measurement at Site 3 should
also have been higher under these calculations. The lower than expected
readings at these two samplers during Period 2 adds to the questionability
of the low readings at these sites during Period 1.
The series of grab samples taken at the downwind edge of Station P
property (G-4 through G-7) at different times indicate the range of benzene
concentration variations that might be experienced at a site during a day's
operations. A dispersion calculation to check on long-term values was made
3
based on the G-4 and G-5 grab samples. These had concentrations of 7.6 yg/m
and 3.9 yg/m , respectively, minus 2.0 yg/m of background benzene downwind of
the road. It was found that the benzene contribution to Site 3 might vary be-
3 3
tween 0.7 yg/m and 2.1 yg/m . When these concentrations are added to the
3 3
2.0 yg/m of assumed background the resulting 2.7 to 4.1 yg/m is greater than
the 2.0 yg/m observed during Period 1.
Emission Calculations
Table 27 presents several estimates of benzene emission rates based
on grab samples taken during the three programs. The samples selected for
calculating emissions were those for which the wind blew from the service sta-
tion to the monitor. In each case the source was considered to be a point
source, which is actually incorrect since at several locations there was more
-------
91
than one service station. In one case the monitor was so close to the station
that the calculation of emissions was made as though the source was an area
source projected backward to a virtual point source. Thus, the values listed
in the table should be considered only as rough estimates.
TABLE 27. ESTIMATED BENZENE EMISSION RATES
Location
1
2
2
3
3
Site
No.
6-5
6-3
6-4
6-4
6-8
Background
Concentration
Subtracted,
yg/m3
2.1
1.2
1.2
2.0
1.8
Atmospheric
Stability
Type
D
C
C
D
D
Distance
From
Source,
m
200
100
200
35^)
155
Time
Day
Day
Day
Day
Night
Emission
- Rate,
g/sec
0.0145
0.0069
0.0083
0.0139
( M
0.061lb)
(a) Virtual distance is 160 m when source is treated as an area source.
(b) During refilling of storage tanks.
No attempt has been made to carry the calculations on to the emission-
factor stage. As mentioned in the last paragraph there is only limited
confidence in the emission-rate calculations. The large amount of benzene
produced by highway traffic adds to the uncertainty. The complications of
different volumes of gasoline of various grades pumped by various gasolines
are too complex to allow derivation of emission factors at this time. Fur-
thermore, there are indications from the Fishinger Road sampling that the
benzene emissions may be related to ambient temperature. If this is true,
then climate can be a significant factor in predicting benzene emissions
and ambient concentrations. The concentrations measured during this sampling
program may be low compared to those found near similar sources in a very hot
climate.
-------
92
Urban Locations
The purpose of this part of the program was to determine
atmospheric benzene levels at urban locations having different traffic
densities.
Selection of Sampling Locations
Three types of sites were chosen in the vicinity of the main
business district of Columbus, Ohio. Air at these sites was sampled and
the benzene concentrations determined. Types of sites included are:
(1) A midtown intersection having a high traffic load
throughout normal business hours,
(2) A highway leading into the business district.
Traffic volume was determined for the sampling
period at this site, and
(3) A residential area situated near the business
district, but relatively isolated from normal
business traffic and having a low service
station density.
The locations of the three sites are indicated on Figure 24. Site
No. 1 was on the Ohio Statehouse lawn at the intersection of Broad (Route 40)
and High (Route 23) Streets, the center of the city. During the sampling
period of March 15 and March 16, 1978, this position was essentially directly
downwind of the traffic passing through that intersection. At Site No. 2,
two samplers, indicated as 2a and 2b in Figure 24, were operated on opposite
sides of the highway between 520 and 550 East Broad Street. Automatic traf-
fic counters were placed by the City on each side of the street in the same
block and approximately 75 feet west of Jefferson Avenue. Site No. 3, Fig-
ure 24, was located in the German Village section of Columbus, south of the
business district. The exact locations of the sites are listed in Table 28.
Sketched maps of each of the sampling locations showing approximate building
placements in the vicinity of the samplers are given in Figures 25, 26, and 27.
-------
93
Fifth Ave.
SCALE IN MILES
1/2
0 1/2
SCALE IN KILOMETER
FIGURE 24. BENZENE MONITORING SITES AT URBAN LOCATIONS IN COLUMBUS, OHIO
-------
94
W. Broad
E. Broad
CP
3
QQ
a>
c
33
a>
c
jg
"5
GO
o>
c
3
0>
00
o>
X
State Street
£$c xo
a g $ *°
CO
• 3
O
OT §
O
£
O
0)
o
c
^ co c Q
° o 1 °
li
State Street
FIGURE 25. BENZENE SAMPLING SITE NO. 1 IN COLUMBUS, OHIO
-------
95
c
"]D
CG
CP
c
GO
Jefferson Ave.
.
LU
Alley
T3
O
O
GO
O
LjJ
O
CJ
o>
O
m
GO
Alley
o>
'5
CD
o»
CD
o o
c c
0) o>
-OX
Washington Ave.
FIGURE 26. BENZENE SAMPLING SITES 2a AND 2b IN COLUMBUS, OHIO
-------
96
C
CD
T3
"
&
C
CD
."2
'to
£
E.Frankfort St.
E. Frankfort St.
c
CD
T3
to
nj '
or
"c
CD
tfy
or
"c
CD
TJ
to
&
•^
QJ
-Q
•to
"»A
CD
or
"c
fli
•g
to
or
c
CD
°O
to
XCD
^v^
0)
Q.^
C ,
o
V)
•£
CD
TJ
&
"o
c
X
E. Columbus St.
E.Columbus St.
"to
&
CD
or
FIGURE 27. BENZENE SAMPLING SITE NO. 3 IN COLUMBUS, OHIO
-------
97
TABLE 28. URBAN SAMPLING LOCATIONS IN COLUMBUS, OHIO
Site
No.
Sampling Protocol
Location
1 Statehouse lawn -- about 100 feet east of
High Street (Route 23) and 250 feet
south of Broad Street (Route 40)
2a 518 East Broad Street (Route 40) — about
6 feet north of the north sidewalk
2b 547 East Broad Street (Route 40) — about
10 feet south of the south sidewalk
3 766 South Fifth Street (iin German Village)
about 3 feet east of the east sidewalk
Air was sampled at all three sites. At Site No. 1, three 8-hour
samples, one duplicate 8-hour sample, and one backup sample were collected
during the sampling period. At the second site, samples consisted of
3 - 8-hour samples at each of the two stations
2 - duplicate samples (one at each station)
2 - backup samples (one at each station)
2 - 2-hour (rush hour) samples at each station
2 - backup samples for the rush-hour samples.
At the residential site, Site No. 3, samples were collected in the same manner
as those at Site No. 1.
In addition, meteorological data were collected at each site. At
Site No. 1, continuous measurements were made of the temperature, wind speed,
and wind direction using a Model 1071 Meteorological Research Inc. instru-
ment. Spot checks of relative humidity, wind speed, wind direction, and tem-
perature were made at each of the sites using portable instruments. These
data were supplemented by hourly readings made by the U.S. Weather Bureau at
Port Columbus.
-------
98
Traffic Data
With the cooperation of the City of Columbus, Division of Traffic
Engineering, it was possible to obtain an exact count of the vehicles pass-
ing the samplers at Site No. 2. This was done with automatic traffic
counters which had hoses extending across all lanes of East Broad Street
and counted the vehicles in both directions. Tables 29 and 30 show the
data that were supplied to us by the Division of Traffic Engineering. Fig-
ure 28 is a plot of the data showing the rush hours and the total movement
of traffic during the sampling period.
Benzene Monitoring Results
Table 31 summarizes the analytical data for the valid determinations
of benzene in the collected samples. Some of the samples at Sites 2a and 2b
were lost due to mechanical failures. This prevented determination of a 24-
hour average benzene concentration at Site 2a and the morning rush-hour ben-
zene concentration at Site 2b.
In general, the observed benzene levels correlate very well with
traffic density. The lowest levels were found in the residential area
(Site No. 3) which is somewhat removed (1/4 to 1/2 mile) from major thorough-
fares. The levels at the midtown location (Site No. 1) were within the range
of concentrations observed along the highway leading into the business district
(Site No. 2). At all three locations, the lowest levels were measured during
the night when the traffic was the lightest.
A comparison of the benzene levels with the traffic volume is given
in Table 32 for Site No. 2, the only location where traffic counts were made.
The highest benzene levels were observed on the same side of the highway as
the heavy traffic flow, the north side. The only time the levels measured
on the south side approached the north-side levels was during the evening rush
hour when the eastbound traffic was heavy. This was true even though the wind
during most of the day was blowing toward the sampler on the south side and
away from the sampler on the north side. However, the winds were light
throughout the day.
-------
99
TABLE 29. TRAFFIC COUNT SUPPLIED BY CITY OF.COLUMBUS
DIVISION OF TRAFFIC ENGINEERING^9)
(East Broad Street at Jefferson,
Westbound Traffic)
One-Hour
Period
Starting
9:00 a.m.
10:00 a.m.
11 :00 a.m.
12:00 p.m.
1 :00 p.m.
2:00 p.m.
3:00 p.m.
4:00 p.m.
5:00 p.m.
6:00 p.m.
7:00 p.m.
8:00 p.m.
9:00 p.m.
10:00 p.m.
11 :00 p.m.
12:00 a.m.
1 :00 a.m.
2:00 a.m.
3:00 a.m.
4:00 a.m.
5:00 a.m.
6:00 a.m.
7:00 a.m.
8:00 a.m.
15-Minute Period Starting
:00
392
305
285
280
291
287
272
274
273
212
189
117
no
105
51
58
24
8
7
7
33
119
479
573
:15
March
364
285
280
313
291
324
301
336
767
180
239
117
107
107
73
March
30
17
13
4
21
52
200
580
590
:30
15
391
310
332
345
321
324
300
308
725
181
211
97
98
88
40
16
31
17
15
15
23
78
244
685
494
:45
305
264
335
356
289
269
295
317
255
180
169
108
90
68
47
24
16
11
10
22
76
358
617
447
One-Hour
Period
Totals
1452
1164
1232
1294
1192
1204
1168
1235
2040
753
808
439
405
368
211
143
74
47
36
73
239
921
2361
2104
Percent
of
Total
6.93
5.55
5.88
6.17
5.69
5.74
5.57
5.89
9.73
3.59
3.85
2.09
1.93
1.76
1.01
0.68
0.35
0.22
0.17
0.35
1.14
4.39
11.26
10.04
(a) Total over 24-hr period: 20963.
-------
TOO
TABLE 30. TRAFFIC COUNT SUPPLIED BY CITY OF COLUMBUS
DIVISION OF TRAFFIC ENGINEERINGS)
(East Broad Street at Jefferson,
Eastbound Traffic)
One-Hour
Period
Starting
9:00 a.m.
10:00 a.m.
11 :00 a.m.
12:00 p.m.
1 :00 p.m.
2:00 p.m.
3:00 p.m.
4:00 p.m.
5:00 p.m.
6:00 p.m.
7:00 p.m.
8:00 p.m.
9:00 p.m.
10:00 p.m.
11 :00 p.m.
12:00 a.m.
1 :00 a.m.
2:00 a.m.
3:00 a.m.
4:00 a.m.
5:00 a.m.
6:00 a.m.
7:00 a.m.
8:00 a.m.
15 -Minute Period Starting
:00
170
172
229
279
196
259
297
443
609
258
147
147
157
134
117
82
22
15
14
6
10
29
101
168
:15
130
179
233
257
229
264
316
390
503
197
175
139
164
85
79
41
22
16
13
9
23
48
118
155
:30
March 15
153
184
235
227
213
281
450
554
409
169
147
171
155
109
74
March 16
45
21
25
7
15
27
70
150
153
:45
184
197
257
223
228
302
231
638
305
180
153
141
121
83
61
43
25
20
14
17
31
85
188
147
One-Hour
Period
Totals
637
732
954
986
866
1106
1294
2025
1826
804
622
598
597
411
331
211
95
76
48
47
91
232
557
623
Percent
of
Total
4.04
4.64
6.05
6.25
5.49
7.01
8.21
12.84
11.58
5.10
3.94
3.79
3.79
2.61
2.10
1.34
0.60
0.48
0.30
0.30
0.58
1.47
3.53
3.95
(a) Total over 24-hr period: 15769.
-------
4000
3800
3600
3400
3200
3000
2800
2600
to
.3? 2400
o
"£ 2200
>
•5 2000
1800
1600
1400
1200
1000
800
600
400
200
o
9AM 10 II I2NIPM 23456789 10 II I2M 1AM 234 56789
March 15, 1978 For the hour ending March 16,1978
FIGURE 28. TRAFFIC FLOW ON ROUTE 40 INTO AND OUT OF THE BUSINESS DISTRICT IN COLUMBUS, OHIO
-------
TABLE 31. ANALYTICAL DATA FOR AIR SAMPLES FROM URBAN SITES IN COLUMBUS, OHIO
Weather Conditions^ '
Wind Temp. Relative
Sampling Speed, Range, Humidity,
Station Direction m/sec F percent
1
2a
2b
3
150°-350° 1.7-2.2 33-39
3500-3600 1.0-2.0 37-40
360°-45° 0.9-2.2 32-40
150° 1.7-1.9 33-36
150°-350° 1.7-2.2 33-40
3500-3600 1.4-1.8 37-40
3300-3600 1.4-2.2 32-33
1500-3500 1.7-2.2 33-40
3500-3600 1.0-2.0 37-40
360°-45° 0.9-2.2 32-37
350°-360° 1.4-1.8 37-40
150°-350° 1.7-2.2 34-39
350°-360° 1.2-2.0 35-40
360°-45° 0.9-2.2 33-40
87-61
57-67
59-75
78
78-61
57-62
77
78-61
57-67
59-75
57-62
78-61
67-57
60-75
Benzene Analysis
PPb /k\
General Sampling Period (v/v) yg/m3^ '
Drizzle 3/15/780712-1455 5.1 16.5(12.8-21.3)
Overcast 3/15/78 1455-2230 4.0 12.8 (10.7-15.3)(2)
Overcast 3/15/78 2237-0815 2.4 7.6 (5.9-9.8)
25 hr average 3.8 12
Drizzle 3/15/780743-0945 8.6 27.8(21.6-35.9)
Overcast 3/15/78 0743-1528 7.5 24.2(18.8-31.2)
Overcast 3/15/78 1528-1754 6.9 22.2 (17.2-28.6)
Overcast 3/16/78 0658-0901 5^5 17.8 (13.8-23.9)
Drizzle 3/15/780812-1542 3,\ 11.0(8.5-14.2)
Overcast 3/15/78 1542-2320 3.3 10.5 (8.1-13.6)
Overcast 3/15/78 2320-0907 2.2 7.2 (5.6-9.3)
25 hr average 3.0 9
Overcast 3/15/78 1542-1759 5.3 17.1(13.3-22.1)
Drizzle 3/15/78 0645-1416 2.0 6.3 (4.9-8.1)
Overcast 3/15/78 1416-2157 1.5 4.8 (4.0-5.8)(2)
Overcast 3/15/78 2157-0804 1.3 4.1 (3.2-5.3)
25 hr average 1 .5 5
(a)
(b)
(c)
Barometric pressure throughout the
The numbers in parentheses are the
quality assurance section. A singl
otherwise the result is based on a
A continuous 25-hour period was not
sampling was 741 to 744 mm Hg.
95 percent confidence limits, determined as discussed under the
e number appearing in parentheses is the number of replicates,
single sample.
covered due to loss of samples during analysis.
o
ro
-------
TABLE 32. CORRELATION OF BENZENE LEVELS WITH TRAFFIC VOLUME AT SITE NO. 2 IN COLUMBUS, OHIO
Without Background Subtraction
Traffic Volume (Vehicles/Hour)
Sampling Period
One -Way
Traffic
Benzene
Benzene Concentra-
tion per Traffic
Unit (ppb/1,000
Vehicles/Hour)
With Background Subtraction^3'
Benzene
Benzene Concentra-
tion per Traffic
Unit (ppb/1,000
Vehicles/Hour)
Total Concentration, One-Way Total Concentration One -Way
Traffic ppb Traffic Traffic ppb Traffic
Sampling Site No.
3/15/78 0743-0945
3/15/78 0743-1528
3/15/78 1528-1754
3/16/78 0658-0901
Westbound Traffic
(Same Side of High-
way as Sampler)
l,903(b)
l!398(b)
1,591
2,126
2,263(b)
3,453
2,688
Sampling Site No.
3/15/78 0812-1542
3/15/78 1542-2320
3/15/78 2320-0907
Eastbound Traffic
(Same Side of High-
way as Sampler)
907
947
224
25 hr average 652
3/15/78 1542-1759
1,788
2)232(b)
1,785
854
1,555
3,352
2a (North
8.6
7.5
6.9
5.5
2b (South
3.4
3.3
2.2
3.0
5.3
Side of Highway)
Westbound
Traffic
4.5
5.4
4.3
2.6
Side of Highway)
Eastbound
Traffic
3.7
3.5
9.8
4.6
3.0
3.4
3.3
2.0
2.0
1.5
1.8
2.6
1.9
1.6
7.3
6.2
5.6
4.2
2.1
2.0
0.9
1.7
4.0
Westbound
Traffic
3.8
4.4
3.5
2.0
Eastbound
Traffic
2.3
2.1
4.0
2.6
2.2
Total
Traffic
2.9
2.7
1.6
1.6
0.9
1.1
1.1
1.1
1.2
o
CO
(a) The background employed was 1.3 ppb, the lowest value observed in the residential neighborhood (Site No. 3).
(b) Portions of this count were estimated from the following morning's count.
-------
104
Since the results indicated a "correlation between benzene concen-
tration and traffic volume, the data were further compared by calculation of
a benzene concentration per unit of traffic. This was done by dividing the
benzene levels (ppb) by the traffic volume (1000 vehicles per hour). The
calculation was performed for both the traffic volume on the same side of
the highway as the sampler and the total traffic volume. It was also done
with and without the subtraction of a background benzene concentration that
might be expected if there were no traffic. The background concentration
used was 1.3 ppb, the night-time level observed in the residential neighbor-
hood (Site No. 3). The results of these calculations are also shown in
Table 32. The best correlations with traffic volume were achieved with the
subtraction of a benzene background. This is most apparent in the data ob-
tained on the south side of the highway, where an excellent correlation was
obtained with the use of the total traffic volume and with background sub-
traction. The wind was blowing from the highway toward the south side sampler
during all of the sampling periods except for a portion of the first. The
reverse situation was true on the north side. There the wind blew toward
the sampler only during the first sampling period and part of the second,
which may have contributed to a poorer correlation of the north side data
with traffic volume. Comparison of the data for the two sides of the highway
is better when only one-way traffic is considered.
-------
105
SUMMARY
Procedures were developed for the collection and analysis of
benzene in air, water, and soil samples. The procedures were applied to
air monitoring for benzene in the vicinity of seven industrial facilities,
in the area around three gasoline service station locations, and at three
urban locations in Columbus, Ohio, and to water and soil sampling for ben-
zene near five of the industrial facilities.
The procedure for collection of benzene in air samples involved
pulling the air through a bed of Tenax GC. It was found that preparation
of the Tenax prior to its use required much more drastic cleanup conditions
than indicated in the literature. The benzene was thermally desorbed from
the Tenax and analyzed by cryogenic capillary gas chromatography. Benzene
present in water and soil samples was determined by sparging it from the
water or soil with nitrogen, adsorbing it in Tenax adsorption tubes, and
analyzing the Tenax tubes in the same manner as air samples.
Results of the air monitoring at the various locations are summar-
ized in Table 33. The 24-hour average benzene concentrations at the air sam-
pling stations around the seven industrial facilities ranged from 2 to 51
3
yg/m (0.5 to 19 ppb) as compared with background concentrations of 0.5 to 3
yg/m (0.2 to 1.2 ppb). At the Mobay Chemical Company nitrobenzene plant, the
3
average levels were in the range of 3 to 11 ug/m (1 to 4 ppb), with the
lowest concentration being observed at the station most distant from the
plant. The average concentrations found near the Gulf Oil Corporation
cumene plant ranged from 25 to 51 yg/m (9 to 19 ppb). Near the Petro-Tex
Corporation maleic anhydride plant, the average concentrations were in the
3
range of 11 to 31 yg/m (4 to 11 ppb). The highest levels were observed at
a station near the plant when the wind, though relatively calm, blew toward
this station. At the Ashland Chemical Company maleic anhydride plant, the
average concentrations ranged from 2 to 32 yg/m (1 to 10 ppb). The highest
levels were found downwind of the plant at stations close to the plant. The
presence of a refinery near this plant complicates the interpretation of the
results. Near the Chevron Corporation detergent alkylate plant the average
benzene levels were in the range of 3 to 7 yg/m (1 to 3 ppb). The location
of this plant prevented samples being taken any closer than 1 km from the
suspected benzene source. The average concentrations near the Union Carbide
Corporation benzene plant ranged from 6 to 35 yg/m (2 to 13 ppb). The
-------
TABLE 33. SUMMARY OF BENZENE AIR-MONITORING DATA.
Location of Source
Relative to Sampler
Facility or Location Monitored
Mobay Nitrobenzene Plant' '
Gulf Oil Cumene Plant^
Petro-Tex Malelc Anhydride Planjb'
Ashland Malelc Anhydride Plant^
Chevron Detergent Alkylate Plant* '
Union Carbide Benzene Plant* '
Sampling
Station
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
5
6
7
1
4
6
1
2
3
Distance, Direction,
km degrees
1.0
1.0
2.5
0.9
2.5
0.8
2.0
2.0
0.2-0.5
1.0
0.5
2.0
0.8
1.6
1.4
1.5
0.5
1.9
0.3
1.0
2.7
1.6
2.0
2.0
1.0
160
305
340
100
230
275
45
145
270-320
45-90
220-270
150
210
350
20
355
330
360
230
15
240
10
250
280
290
Mind
Direction,
degrees
225/270
225/270
225/270
225/279
45/202.5
45/202.5
202.5
45/225
180-225
90
225
135-315
120-340
120-340
180-340
180-340
120-340
120-340
180-320
60-360
60-180
60-180
180-202.5
135/202.5
135/180
Speed ,
m/sec
Calm-7.5
Calm-7.5
Calm-7.5
Calm-7.5
0.1-1.8
0.01-1.0
Calm-1.01
Calm-3
Calm-1.45
Calm-1.45
Calm-0.97
Calm-0.97
1.6-5.7
1.6-5.7
1.6-5.7
1.6-5.7
1.6-5.7
1.6-5.7
2.6-4.6
0.2-1.5
1.6-4.1
1.6-4.1
0.2-1.6
0.6-1.8
0.6-3.8
Benzene Concentrations*8'
ppb
Avg.
4
4
1
2
19
10
12
9
6
8
11
4
6
1
1
5
3
1
10
2
1
3
2
4
13
(v/v)
Range Avg.
1.3-5.2
2.8-7.0
0.4-1.3
1.2-3.0
9.1-34.8
1.2-19.4
5.4-20.5
7.4-14.4
3.4-9.0
2.6-9.1
4.5-16.4
3.1-6.6
4.0-9.3
0.3-1.1
0.3-1.7
0.3-29.7
0.7-6.8
0.3-1.1
1.0-13.8
2-3
1
3
0.6-4.5
1.1-6.4
12.8-13.5
11
11
3
5
51
28
31
25
15
22
31
11
19
2
3
16
10
2
32
6
3
7
6
9
35
ug/m
Range
3.4-14.1
7.5-18.7
1.2-3.4
3.2-8.1
24.4-93.5
3.1-52.2
14.5-55.0
19.8-38.7
9.3-24.3
7.0-24.5
12.0-39.1
6.9-17.7
12.9-30.0
0.9-3.4
1.1-5.4
1.1-95.8
2.3-21.9
1.0-3.5
3.2-44.5
3.9-7.3
3
6.8
1.6-12.1
2.9-17.3
34.4-36.2
-------
TABLE 33. (Continued)
Location of
Relative to
Source
Sampler
Wind
Sampling Distance, Direction, Direction,
Facility or Location Monitored
U.S. Steel Coke Oven Facility(b)
Location 1 : Four-Station
Intersect1on(c)
Location 2: Single Full -Service/
Self-Service Statlon(c)
Location 3:. Single Self-Service
Station(c)
Location 1: Center of the City^ '
Location 2: Both Sides of a
Busy Highway(d)
Station
1A
IB
1
2
3
4
1
2
3
4
5
6
7
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
1
2a
2b
km
7.
2.
2.
0.
1.
1.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
—
_ _
--
0
5
2
5
0
4
1-0.2
05-0.25
1-0.4
15-0. f.
15-0.3
2-0.3
2-0.4
2
2
05
1
03
07
1
05
1
15
05
1
25
1
1
degrees
145
170
135
280
100
100
300-45
280-360
285-310
270-290
300-10
185-245
40-90
250
90
75
270
360
160
210
290
110
170
190
220
270
135
315
—
180
260
Location 3: Residential Neighborhood' '3
degrees
110-220
110-180
90-340
90-340
70-340
70-180
30-360
30-360
30-360
30-360
30-360
30-360
30-360
225-360
225-360
225-270
225-360
225-360
225-360
225-360
225-360
200-240
180-240
180-240
180-240
180-240
180-240
180-240
150-45
150-360
150-45
150-45
Speed,
m/sec
1.6-6.2
1.6-4.1
0.6-2.5
0.6-2.5
0.6-3.2
0.6-2.5
Calm-4.1
Calm-4.1
Calm-4.1
Calm-4.1
Calm-4.1
Calm-4.1
Calm-4.1
4.6-9.3
4.6-9.3
7.2-9.3
4.6-9.3
4.6-9.3
4.6-9.3
4.6-9.3
4.6-9.3
3.6-9.3
2.6-8.8
2.6-8.8
2.6-8.8
2.6-8.8
2.6-8.8
2.6-8.8
0.9-2.2
1.4-2.2
0.9-2.2
0.9-2.2
Benzene Concentrations*3'
ppb
Avg
3
0
6
1
2
5
0
0
1
1
0
0
2
0
0
0
1
0
0
0
1
1
0
1
0
0
0
0
4
7
3
1
•
.5
.5
.8
.5
.1
.3
.5
.0
.7
.5
.6
.3
.7
.5
.8
.0
.2
.7
.0
.9
.8
.7
.3
.5
(v/v)
Range
2.0-6.0
0.4-0.6
2.3-12.3
0.7-2.0
1.5-3.0
3.5-6.0
0.2-0.8
0.6-1.0
0.5-2.1
0.3-2.0
0.1-0.4
0.3-0.7
0.6-4.3
0.4-0.8
0.2-0.8
0.6
0.9-1.6
0.7
0.4-0.6
0.5-1.0
0.7-1.2
0.9-1.4
0.7
0.6-1.3
0.7-1.1
0.7-0.8
0.6-0.8
0.3-0.4
2.4-5.1
5.5-8.6
2.2-5.3
1.3-2.0
Avg
10
2
19
4
7
15
1
2
5
4
1
2
7
2
2
2
4
2
2
2
3
4
2
3
2
2
2
1
12
23
9
5
3
pg/m
. Range
6.5-19.2
1.2-2.0
7.5-41.5
2.2-6.4
4.9-9.7
11.4-19.3
0.5-2.5
1.8-3.2
1.7-6.7
0..9-6.3
0.4-1.4
0.9-2.1
1.9-13.7
1.2-2.7
0.6-2.4
1.8
2.9-5.0
2.3
1.3-2.0
1.5-3.2
2.3-3.7
2.8-4.5
2.4-2.5
2.0-4.0
2.2-3.3 '
2.3-2.4
1.8-2.6
0.9-1.4
7.6-16.5
17.8-27.8
7.2-17.1
4.1-6.3
a) These data are for the continuous monitoring
b) Industrial locations.
c) Service Station locations in
d) Urban locations in Columbus,
Columbus, Ohio.
Ohio.
only
(grab
samples are
not included).
-------
108
highest concentration was observed at the sampling station closest to the
emission point. At the U.S. Steel Corporation coke oven facility, the
o
average concentrations ranged from 2 to 19 yg/m (0.5 to 6 ppb). The
highest level was found in the valley across the river from the coke ovens.
The lowest levels were observed at a station on a ridge above the ovens,
though it was downwind.
A limited .number of soil and water samples were obtained near
five of the industrial facilities and analyzed for benzene. The results
are summarized in Table 34. Exclusive of plant effluent samples, the
benzene concentrations found in water samples ranged from <1 to 13 ppb.
In general, the downstream levels were 1 to 2 ppb or less. The higher levels
were found near the plant outfall in one case and upstream of the plant in
another case where a chlorobenzene plant was located upstream. The highest
levels of benzene found in any of the three environmental media were found
in the soil samples. The levels ranged from 2 to 191 ppb.
Also summarized in Table 33 are the results of air sampling in
Columbus, Ohio, near three service station locations and at three urban
locations near the center of the city. At sampling points near the service
stations, the 12-to-30-hour average benzene concentrations were in the range
of 1 to 7 yg/m (0.3 to 2.0 ppb), with individual measurements ranging from
0.4 to 13.7 yg/m (0.1 to 4.3 ppb). Analysis of the data revealed that emis-
sions from automobiles (idling, parked after a trip, or in traffic) contribute
to observed benzene concentrations. Background benzene concentrations in
neighborhoods upwind of the service stations and automobile traffic ranged from
0.6 to 1.4 yg/m3 (0.2 to 0.4 ppb). Levels of 32 to 69 yg/m3 (10 to 22 ppb)
were observed in the residential neighborhood near one station in grab samples
taken during the filling of underground storage tanks.
Generally higher benzene concentrations were found at the three
urban Columbus locations, all of which were nearer the center of the city
then the service station locations. In a residential neighborhood near the
q
business district, the 25-hour average benzene level was 5 yg/m (1.5 ppb).
In the center of the business district, it was 12 yg/m (4 ppb). Immediately
adjacent to a busy highway leading into the business district, the average
3 3
levels were 9 yg/m (3 ppb) on one side of the highway and 23 yg/m (7 ppb)
on the other side. The higher levels were observed on the side of the high-
way carrying the heavier traffic. The individual measurements made at this
location correlate well with the traffic counts.
-------
TABLE 34. SUMMARY OF BENZENE SOIL- AND WATER-MONITORING DATA
Facility Monitored
Mobay Nitrobenzene Plant
Gulf Oil Cumene Plant
Petro-Tex Maleic Anhydride Plant
Chevron Detergent Alkylate Plant
Union Carbide Benzene Plant
Sampling
Station
1
3
4
5
6
7
1
2
3
4
5
6
7
1
5
6
7
8
2
3
4
5
7
1
2
6
7
8
Water
Sample Description
—
Downstream: Surface
6 ft depth
--
Upstream: Surface
6 ft depth
Plant Influent
Plant Effluent
_ _
--
_ _
__
Wastewater Treatment
Unit: Outfall
From pump
River
Gulf Dock
--
Plant Outfall
Midway, Near Outfall
Upstream
Downstream
--
Yacht Basin
__
—
Municipal Water
__
--
Plant Effluent
Upstream
Downstream
Benzene
Concentration,
ppb
__
3.3
4.9
--
12.0
11.9
4.2
104.
— .
._
_ —
-_
4.3
56.6
1
1.9
—
8
13
<1
2
--
<1
_ _
__
<1
__
--
179
2
1
Distance
from the
Plant, km
1.0
2.5
0.9
--
--
--
—
2.5
0.8
2.0
2.0
--
--
—
—
0.2-0.5
--
—
—
—
1.7
1.3
2.7
3.2
--
1-2
A>2
—
—
—
Soil
Benzene
Concentration,
ppb
2
51
17
--
--
--
—
34
>73U)
1 O
1 O
>34(fl)
--
--
—
—
22
--
—
—
—
191
70
51
>148(a)
--
12
14
--
—
— -
o
UD
"[a] Sample not heated, so this is a minimum value.
-------
no
REFERENCES
(1) Richard Hartle, personal communication.
(2) E. D. Pellizzari, personal communication.
(3) E. D. Pellizzari, J. F. Bund, R. E. Berkley, and J. McRae, "Determination
of Trace Hazardous Organic Vapor Pollutants in Ambient Atmospheres by Gas
Chromatography-Mass Spectrometry Computer", Anal. Chem., 48_, 803-7 (1976).
(4) W. Bertsch, A. Zlatkis, H. M. Leibich, and H. J. Schneider, "Concentration
and Analysis of Organic Volatiles in Sklab 4", J. Chromatog. 99, 673-87
(1974).
(5) W. J. Dixon, "Biomedical Computer Programs: BMD", University of California
Press, 2, 486 (1967).
(6) D. B. Turner, "Workbook of Atmospheric Estimates, AP-26", U.S. Department
of Health, Education, and Welfare, Cincinnati, Ohio (1970).
(7) W. :N. Sanders and J. B. Maynard. 1968. Capillary Gas Chromatographic
Method for Determining the C3-C12 Hydrocarbons in Full-Range Motor Gaso-
lines. Analytical Chemistry 40_:527-535.
(8) H. E. Runion. 1977. Benzene in Gasoline II. American Industrial Hygiene
Association Journal. 38:391-393.
-------
APPENDIX A
CONDITIONING OF TENAX GC FOR BENZENE ANALYSES
-------
A-l
APPENDIX A
CONDITIONING OF TENAX GC FOR BENZENE ANALYSIS
Preparation of satisfactory traps is dependent upon several
important variables; solvent extraction and time and temperature of con-
ditioning. To check out these variables several experiments were made.
Exploration of all possible combinations was precluded by limitations of
time, but the primary variables were studied as discussed below.
Stainless Steel Versus Glass
Stainless steel and glass traps (U tube, 3/8 in. O.D. x 12 in. long)*
were packed with Tenax GC that was solvent extracted in a Soxhlet extractor
with distilled-in-glass methanol for 48 hours, then vacuum dried at 135-150 C
for 16 hours. The traps were then conditioned for 45 min at 200 C with a
helium flow of 40 ml/min on the GC analytical system (see later section on
analysis). The traps were then evaluated by being subjected to essentially
the standard analytical procedure for benzene. Each trap was heated for 5 min-
utes at 200 C while helium at 10 ml/min was passed through the trap and swept
onto the GC column at -70 C. For trap evaluation, the column was programmed
from 0 C at 8 C/min. This is the standard procedure for evaluation of Tenax
traps. Figure A-la and A-lb are chromatograms for evaluation of Tenax con-
ditioned in the glass and stainless steel traps, respectively. To simulate
the effect of aging (storage), the traps were capped and heated overnight
(16 hr) at 100 C. Figure A-lc shows a check of the stainless steel trap. The
seal** on the glass trap apparently leaked air, as the chromatogram of the
glass trap had peaks completely off scale.
These data indicate that stainless steel does not appear to affect
the stability of Tenax.
* All stainless steel traps are this size unless otherwise indicated.
** Swagelok caps with graphite ferrules were used.
-------
A-2
a. Glass Trap
b. Stainless Steel Trap
c. Stainless Steel Trap After Heating Overnight at 100 C
FIGURE A-l. CHROMATOGRAMS FOR COMPARISON OF GLASS AND STAINLESS STEEL TRAPS
-------
A-3
Extracted Versus Unextracted Tenax GC
A stainless steel trap was packed with unextracted Tenax GC and
conditioned at 200 C for 1 hour. It was evaluated by the standard procedure.
The chromatogram is shown in Figure A-2a. This trap was then capped and
heated overnight at 60 C. Figure A-2b is the chromatogram obtained upon eval-
uation of the trap. Comparison* of Figure A-2a with Figure A-lb indicates the
extraction makes little difference in the cleanliness of traps immediately
after conditioning. However, comparison of Figure A-2b with Figure A-lc in-
dicates that extracted Tenax is superior to unextracted Tenax for stability.
Except for one series to be discussed later, all traps have been packed with
extracted Tenax GC.
* A single column GC is used, and thus baseline drift should be ignored.
-------
A-4
a. Initial Evaluation
b. After Aging Overnight at 60 C
FIGURE A-2. CHROMATOGRAMS OF STAINLESS STEEL TRAP FILLED WITH UNEXTRACTED TENAX
-------
A-5
Effect of Temperature and Time
in Thermal Conditioning of Tenax GC
Two times and temperatures were picked for a brief preliminary study
of these two variables, 200 C (used in earlier work at BCL on another project)
and 275 C, and 20 minutes and 2 hours chromatograms for evaluation of these
traps are given in Figure A-3. All were run 24 hours after conditioning. The
chromatogram for the trap conditioned 20 minutes at 200 C had numerous peaks
off-scale. Chromatograms for evaluation of the other three traps are shown in
Figure A-3 as follows:
200 C, 2 hr - Figure A-3a
275 C, 20 min - Figure A-3b
275 C, 2 hr - Figure A-3c.
Based on these chromatograms, 20 minutes is obviously too short a time for
conditioning and 2 hours at 200 C seems better than 2 hours at 275 C.
Logically, however, higher temperatures should give better condition-
ing, provided that the decomposition temperature of Tenax is avoided. Previous
attempts to condition traps in groups were reported earlier and had not been
successful. Several more attempts were made to condition traps in groups of
from four to twelve at 275 C for 16 hours. The traps were fastened together
in series with 3/8 in. Swagelok unions and heated in a GC oven. Results were
variable; some traps were good -- others bad. A manifold was constructed to
permit conditioning of four groups of traps in .parallel, three traps in series
in each group. Again, results were variable; some traps were good, others were
marginal, and others poor. The difficulty was finally traced to leaks in the
Swagelok connections. Since the Swagelok connection was heated, air oxidation
of the surfaces occurred. The oxidation made it more difficult to tighten the
connections with each successive experiment. Connections that were tight before
heating would develop leaks after being heated and cooled to room temperature.
Leaks permit diffusion of air into the traps, and Tenax GC oxidizes at the
elevated temperatures.
At this point there was a critical need for traps and no really good
method of conditioning Tenax traps more than one at a time had been found. The
only reliable method was to condition traps individually on the GC analytical
system; 200-225.C for 2 hours gave good traps. Long-term stability tests had
-------
A-6
a. 200 C - 2 hr
b. 275 C - 20 min
c. 275 C - 2 hr
FIGURE A-3. CHROMATOGRAMS FOR COMPARISON OF TIME AND TEMPERATURE
IN TENAX CONDITIONING
-------
A-7
not been performed at room temperature but it had been found that traps were
stable when stored at -70 C. Therefore, 40 traps were packed with solvent-
extracted Tenax, conditioned at 225 C for 2 hours on the GC system, and
(2}
stored at dry-ice temperature. It had been learned from Pellizzariv '
that cold storage of traps after sampling was advisable if the traps could
not be analyzed quickly, so it was planned to keep them cold after sampling.
These traps were-used to sample at Mobay Chemical Company, New Martinsvilie,
West Virginia.
Figure A-4a is the chromatogram for evaluation of one of these traps.
Figure A-4b is the chromatogram of a control trap, taken on the sampling trip
and returned with the samples. No significant deterioration of the traps oc-
curred. However, storage at room temperature for 1 week led to significant
worsening of the traps, Figure A-4c; therefore, the short conditioning time is
not satisfactory.
Tenax Conditioning System
To overcome this difficulty, a new system for conditioning Tenax
traps was designed. This system consists .of a stainless steel manifold onto
which 12 Tenax traps can be fastened individually so that the traps are con-
nected in parallel. Each trap is closed with a stainless steel cap into which
a 1-inch piece of .01 in.I.D. stainless steel capillary is silver soldered.
A 300 ml/min flow of nitrogen through the manifold provides a flow of 25 ml/
min through each trap. High purity nitrogen is used. It is further purified
by passing it through an Oxy-Trap and a Drierite-molecular sieve trap. The
traps mounted on the manifold are heated by lowering the assembly into the
Wood's metal bath. This bath is made from an aluminum block, machined to pro-
vide a slot for each tube.* Since the connections are not in the bath they
are not subjected to the maximum temperature, thus reducing oxidation problems.
The top is covered with fiberglass and aluminum foil during the conditioning
period, and the enclosed space is constantly flushed with nitrogen, further re-
ducing oxidation.
* The slots are not necessary, but substantially reduce the amount of Wood's
metal required.
-------
A-8
a. 24 Hours After Conditioning
b. Control Trap Taken on Mobay Sampling
Trip (Stored 3 Weeks at - 70 C)
Ml
c. Stored 1 Week at Room Temperature
FIGURE A-4. CHROMATOGRAMS FROM EVALUATION OF TRAPS CONDITIONED FOR
2 HR AT 225 C
-------
A-9
The first experiments with this apparatus employed Tenax GC which
had been extracted for at least 48 hours with methanol and vacuum dried be-
fore packing the traps. Three sets of traps (12 in each set) were condi-
tioned overnight (16-20 hours) and evaluated as described previously. Of
the 36 traps, slightly over one-half were considered good, the rest were
only fair, and a few poor. Figure A-5 shows chromatograms of the evaluation
of four typical traps. These traps were reasonably good. However, since so
many traps were marginal or unsatisfactory, this method of preparation of
traps was not considered satisfactory.
Preconditioning of Tenax
More thorough conditioning of Tenax would have to involve higher
(4^
temperature, longer time, or both. Bertsch et al,v ' recommended 340 C.
Because of the possible catalytic effects of stainless steel surfaces,
conditioning in the stainless steel tubes did not seem safe. Moreover,
the problem of oxidation of joints would be more severe at higher temper-
atures. Longer time also was undesirable from the standpoint of reasonable
production rate.
It was decided to precondition Tenax at a higher temperature in large
glass tubes. Accordingly, several experiments were run in which the solvent
extracted Tenax GC was heated at 300 C in a large glass tube, with purified
nitrogen flow. Four different conditions were evaluated:
(A) Re-extract with methanol for 72 hours, vacuum dry.
Heat with nitrogen flow for 8 hours at 300 C.
(B) No re-extraction -- heat with nitrogen flow for
8 hours at 300 C.
(C) Similar to (B), but 16 hours at 300 C.
(D) Similar to (B) and (C), but 40 hours at 300 C.
At this point all of the Tenax on hand had been extracted with methanol at
least once and packed into traps. All traps that were marginal or poor were
unpacked, screened to remove any debris, and blended into one large batch of
recovered Tenax. Experiments A and B above were run concurrently to determine
-------
A-10
a. Typical Good Traps
b. Typical Good Traps
FIGURE A-5. CHROMATOGRAMS FOR EVALUATION OF SOLVENT EXTRACTED TENAX
-------
A-n
c. A Fair Trap
d. A Poor Trap
FIGURE A-5. (CONTINUED)
-------
A-12
whether a re-extraction of the recycled Tenax was required. Experiments (C)
and (D) were done later. Several sets of traps were packed with each of the
four kinds of preconditioned Tenax and conditioned in the 12-tube conditioning
system described previously.
The following sections briefly describe the results of the evalua-
tion of the conditioned traps. Overall, Tenax (A) is slightly better than (B),
(C), or (D). Therefore, in all further reconditioning of Tenax a combination
of solvent extraction and thermal conditioning was used.
In evaluation of traps, however, it should be noted that the gas
chromatograph with flame-ionization detector is much more sensitive than
GC/MS. Peaks considered unacceptable interferences for this program are not
detectable by GC/MS, unless the latter is operating in the SIM mode and look-
ing for a specific ion (or ions). The worst traps in the whole series have
peaks corresponding to only about 2 to 4 ng of benzene. Thus, using even the
bad traps on a 14-liter sample would provide an uncertainty of less than
0.1 ppb (volume).
Preconditioned Tenax, Type A
A total of 26 traps were conditioned and checked. Most of these traps
were quite good. The chromatogram given in Figure A-6a is typical.* Two had
larger than usual peaks in the early part of the chromatogram as shown in
Figure A-6c.
Preconditioned Tenax, Type B
Most of these traps were also quite good, as illustrated by Figure A-7a.**
However, two of the traps had significantly larger peaks, in and following the ben-
zene regions, as shown in Figure A-7c.
* Storage stability of these traps was good. Figure A-6b shows the same trap
after 21 days aging.
** Storage stability of these traps was good. Figure A-7b shows the same trap
after 14 days aging.
-------
A-13
a. Typical Good Trap
A
b. Trap A, Aged 14 Days
A
c. Trap With Early Peaks
FIGURE A-6. CHROMATOGRAMS FOR EVALUATION OF PRECONDITIONED TENAX TYPE A
-------
A-14
a. Typical Good Trap
b. Trap A, Aged 14 Days
c. Trap With Larger Peaks
FIGURE A-7. CHROMATOGRAMS FOR EVALUATION OF PRECONDITIONED TENAX TYPE B
-------
A-15
Preconditioned Tenax, Type C
There were 25 traps in this series; all but three were good.
Figure A-8a is typical of the good traps.* The other three were rated
fair and are exemplified by Figure A-8b. These traps were also fairly
stable as shown in Figure A-8c.
Preconditioned Tenax, Type D
There were 59 traps in this series. Most of these were found
to be good. Figure A-9a is representative of the good traps. Two were
found to be fair (Figure A-9b) and two were poor (Figure A-9c).
Alltech Associates Preconditioned Tenax
All tech Associates markets a preconditioned Tenax, advertised as
being ready for use. It is made from Tenax GC by thermal conditioning at
300 C for 3-4 hours in a tube furnace with a stream of purified nitrogen.
A 15-g lot of this material in stainless steel traps was conditioned over-
night at 275 C in the conditioning apparatus previously described. These
traps were all found to be good. Figure A-lOa is representative. In ad-
dition, traps stored for up to 3 weeks at room temperature remained good
(Figures A-lOb and A-lOc).
Several more lots of the Alltech preconditioned Tenax were ordered
for additional tests. Unfortunately, traps packed from these later lots did
not respond as favorably to conditioning. Many traps from these later were
rated as only fair. However, these traps were not checked immediately after
conditioning, whereas most traps have been checked initially within a few
days of conditioning. Also as noted previously, their peaks are not large
enough to cause an uncertainty of more than about 0.1 ppb.
It has been observed that any trap will be better after the first
evaluation. Several of the traps packed with the later lots of Alltech
preconditioned Tenax were re-checked after the initial evaluation. Fig-
ure A-ll** shows one of these traps. On re-checking it is much better.
* Storage stability of these traps was good. Figure A-8b shows the same trap
after 10 days aging.
** These chromatograms were run with a 50-meter column, whereas all the other
chromatograms were run with a 25-meter column. Consequently, the chloroform
marker peak is at about 9 minutes and benzene elutes at about 10 minutes.
-------
A-16
a. Typical Good Trap
b. A Fair Trap
c. Trap C Aged 10 Days
FIGURE A-8. CHROMATOGRAMS FOR EVALUATION OF PRECONDITIONED TENAX TYPE C
-------
A-17
a. Typical Good Trap
b. A Fair Trap
c. A Poor Trap
FIGURE A-9. CHROMATOGRAMS FOR EVALUATION OF PRECONDITIONED TENAX TYPE D
-------
A-18
a. Typical Good Trap
b. a, 7 Days Later
c. Another Good Trap, 21 Days Aging
FIGURE A-10. CHROMATOGRAMS FOR EVALUATION OF SAMPLE LOT OF
ALLTECH PRECONDITIONED TENAX
-------
A-19
a. Trap Rated as Fair
b. Same Trap Rechecked 2 Days Later
FIGURE A-ll.
CHROMATOGRAMS FROM EVALUATION OF LATER LOTS OF
ALLTECH PRECONDITIONED TENAX
-------
APPENDIX B
BENZENE PRODUCTION AND CONSUMPTION FACILITIES
-------
B-l
APPENDIX B
BENZENE PRODUCTION AND CONSUMPTION FACILITIES
To permit a logical choice of sampling locations, data were
assembled on the production and consumption of benzene. Table B-l lists
all of the benzene manufacturing plants for which data were found, and
includes the location, company, capacity, and process. Tables B-2 and B-3
give the benzene requirements for chemical manufacturing processes using
benzene. Tables B-4 through B-10 contain data on plants using benzene in
production of various organic industrial chemicals. Table B-ll lists the
cities in which benzene production and consumption is largest.
-------
B-2
TABLE B-l. BENZENE MANUFACTURING PLANTS
(LOCATIONS, ANNUAL CAPACITIES, AND PROCESSES)
Capacity
State/City
California
El Segundo
Palo Alto
Wilmington
Colorado
Pueblo
Illinois
Lemont
Wood River
Kansas
ElDorado
Kentucky
Ashland
Louisiana
Alliance
Baton Rouge
Chalmette
Lake Charles
Shreveport
Taft
Maryland
Sparrows Point .
Michigan
Bay City
New Jersey
Westville
New York
La cka wanna
North Tonawanda
„ (in
Company ,
Standard Oil (Calif.).
Beckman Instr.
Atlantic Richfield
CF&I Steel
Union Oil
Shell Chemical
Skelly Oil
Ashland Oil
Gulf Oil
Exxon
Tenneco
Cities Service
Pennzoil
Union Carbide
Bethlehem Steel
Dow
Texaco
Bethlehem Steel
Ashland Oil
millions
gallons)
23
NA
12
3
24
40
13
50
70
65
10
25
15
70
15
30
60
8
15
(in millions
of pounds)
168
88
22
176
293
95
366
513
476
73
183
110
513
110
220
440
59
110
Process*
(in millions
of gallons)
23-R
NA
12-R
3-C
18-C
6-R
40-R
13-R
40-C
10-H
45-D
25-R
43-R
22-E
10-R
25-R
15-E
70-E
15-C
30-R
35-R
25-D
8-C
12-C
3-R
-------
B-3
TABLE B-l. (Continued)
Capacity
State/City
Ohio
Middletown
Toledo
Oklahoma
Tulsa
Pennsylvania
Aliquippa
Bethlehem
Clairton
Marcus Hook
Marcus Hook
Philadelphia
Puerto Rico
Guayama
Penuelas
Texas
Bay town
Beaumont
Beaumont
Big Spring
Chocolate Bayou
Corpus Christi
Corpus Christi
Corpus Christi
Corpus Christi
Deer Park
Freeport
Houston
Houston
Company (in
of
Armco Steel
Interlake
Sun Oil
Jones & Laughlin
Bethlehem Steel
United States Steel
Sun Oil
Standard Oil (Ohio)
Gulf Oil
Phillips Petroleum
Commonwealth Oil
Exxon
Mobil Oil
Union Oil
American Petrofina
Monsanto
Union Pacific (Champlin)
Kerr-McGee
Sun Oil
Coastal States Gas
Shell
Crown Central Petrol.
Charter Oil
Atlantic Richfield
millions
gallons)
3
1
24
10
4
45
15
8
40
110
185
65
60
70
45
75
10
16
35
70
75
50
5
44
(in millions
of pounds)
22
7
176
73
29
330
110
59
293
806
1355
476
440
513
330
549
73
117
256
513
549
366
37
322
Process*
(in millions
of gallons)
3-C
1-NA
12-D
12-R
10- C
4-C
45-C
15-R .
8-NA
19-D
21-R
65-D
45-R
110-D
70-R
43-R
22-E
40-R
20-E
70-D
30-D
15-R
35-D
40-E
10-R
16-NA
20-D
15-R
60-D
10-R
40-R
35-E
50-D
5-R
32-iRc
12-D
-------
B-4
TABLE B-l. (Continued)
Capacity
State/City
Texas (Continued)
Lone Star
Odessa
Pasadena
Port Arthur
Port Arthur
Port Arthur
Sweeny
Texas City
Texas City
Winnie
Utah
Geneva
Company (in
of
Northwest Ind.
Shell
Crown Central Petrol.
American Petrofina
Gulf Oil
Texaco
Phillips Petroleum
Standard Oil (Ind.) (Amoco)
Marathon Oil
Allied Chemical
U.S. Steel
millions
gallons)
1
6
23
15
38
45
22
85
6
3
4
(in millions
of pounds)
7
44
168
110
278
330
161
623
44
22
29
Process*
(in millions
of gallons)
1-C
6-R
18-D
5-R
15-D
30-R
8-E
45-R
22-R
80-R
5-E
6-R
3-R
4-C
Virgin Islands
St. Crois
Amerada Hess
25
183
25-R
TOTALS
1886
13815
(see below)
*c -
D -
E -
R -
NA
Coal derived benzene from coke oven light oil:
Dealkylation of toluene
Petroleum product cracking operations for ethylene
production
Catalytic reformate of petroleum products
- Not available
Totals
163
526
307
865
25
Percent
8.6
27.9
16.3
45.9
1.3
1886
100.0
-------
B-5
TABLE B-2. BENZENE REQUIREMENTS FOR MAJOR CHEMICAL PRODUCTS
Product
Benzene required at
100 percent capacity
(Million pounds)
Percent of
Total
Benzene Capacity
Ethylbenzene
Curaene
Nitrobenzene
Chlorobenzene
Detergent Alkylate
Cyclohexane
Maleic Anhydride
Total benzene capacity
6,354
3,028
694
656
420
389
389
11,930
13,815
46
22
5
5
3
3
3
87
(See Table )
TABLE B-3. ESTIMATED BENZENE REQUIREMENTS FOR MINOR CHEMICAL
PRODUCTS
Product
Benzene Required at
100 percent capacity
(Million pounds)
Locations
Anniston/Al.,Sparrows Pt./Md,
Bay City/Mich, Lyndhurst/NJ,
Greensboro/NC, Woonsocket/RI,
Freeport/Texas, Houston/Texas
Petrolia/Pa
LaSalle/Ill., Kingsport/Tenn,
Baysport/Texas
Toms River/NJ
Niagara Falls/NY, Newark/NJ
Midland/Mich, Stamford/Conn,
Jacksonville/Ark.
Midland/Mich
Widely scattered
Biphenyl (diphenyl)
Resorcinol
Hydroquinone
Anthraquinone
Benzene hexachloride
(Lindane)
Trichlorophenol(245T)
Pentachlorophenol
Solvent(paints, rubber
solvent extraction, etc
35
25
7
small
small
small
small
cements,
.) small
-------
B-6
TABLE B-4. MALEIC ANHYDRIDE MANUFACTURING PLANTS
(PLANT LOCATIONS, COMPANIES, AND ANNUAL
CAPACITY FOR BENZENE CONSUMPTION)
State/City
Company
Benzene Requirement at .
100 Percent Capacity
(million pounds)
Illinois
Cicero
. Morris
Missouri
St. Louis
New Jersey
Elizabeth
Fords
Pennsylvania
Bridgeville
Neville Island
Texas
Houston
West Virginia
Moundsville
TOTAL
Koppers
Reichold
Monsanto
Reichold
Tenneco
Koppers
U.S. Steel
Petro-Tex
Allied Chemical
13
80
141
40
35
46
54
67
80
389
(a)
Based on the assumption that 1.34 pounds of benzene are required for
each pound of maleic anhydride produced.
-------
B-7
TABLE B-5. ETHYLBENZENE MANUFACTURING PLANTS (PLANT LOCATIONS,
COMPANIES, AND ANNUAL CAPACITY FOR BENZENE
CONSUMPTION)
Benzene Requirement at
State/City Company 100 Percent Capacity
(million pounds)
ta>
Louisiana
Baton Rouge
Carville
Chalmette
Welcome
Michigan
Midland
Puerto Rico
Penuelas
Texas
Big Spring
Corpus Christi
Freeport
Houston
Houston
Houston
Odessa
Phillips
Port Arthur
Seadrift
Texas City
Texas City
Foster Grant
Cos-Mar
Tenneco
Gulf Oil
Dow
Commonwealth Oil
American Petrofina
Sun Oil
Dow
Charter Oil
ARCO
Joe Oil
ElPaso
Phillips Petroleum
ARCO
Union Carbide
Monsanto
Standard Oil (ind.) (Amoco)
720
534
19
408
408
119
33
70
1384
26
74
NA
204
NA
326
252
1076
701
TOTAL 6354
(ay
Based on the assumption that 0.742 pounds of benzene are required for
each pound of ethylbenzene produced.
-------
B-8
TABLE B-6. CUMENE MANUFACTURING PLANTS (PLANT LOCATIONS,
COMPANIES, AND ANNUAL CAPACITY FOR BENZENE
CONSUMPTION)
State/City
Company
Benzene Requirement at .
100 Percent Capacity
(million pounds)
California
El Segundo
Illinois
Blue Island
Kansas
El Dorado
Kentucky
Ashland
Michigan
Midland
New Jersey
Westville
Pennsylvania
Philadelphia
Puerto Rico
Penuelas
Texas
Chocolate Bayou
Corpus Christi
Corpus Christi
Port Arthur
Texas City
Texas City
TOTAL
Standard Oil (Calif.)(Chevron)
Clark Oil
Skelly Oil
Ashland Oil
Dow
Texaco
Gulf Oil
Union Carbide
Monsanto
Coastal States Gas
Sun Oil
Gulf Oil
Marathon Oil
Standard Oil (Ind.)(Amoco)
80
88
108
280
8
208
360
512
520
112
200
360
152
40
3028
(a)
Based on the estimate that 0.80 pounds of benzene are required for each
pound of cumene produced.
-------
B-9
TABLE B-7. NITROBENZENE MANUFACTURING -PLANTS (PLANT
LOCATIONS, COMPANIES, AND ANNUAL CAPACITY
FOR BENZENE CONSUMPTION)
State/City
Company
Benzene Requirement at .
100 Percent Capacity
(million pounds)
Illinois
Sauget
Louisiana
Geismar
Mississippi
Pascagoula
New Jersey
Bound Brook
Gibbstown
Texas
Beaumont
West Virginia
Moundsville
New Martinsville
Willow Island
TOTAL
Monsanto
Rubicon Chemicals
First Mississippi
American Cyanamid
DuPont
DuPont
Allied Chemical
Mobay Chemicals
American Cyanamid
7
49
88
55
130
202
36
88
39
694
(a)
Based on the assumption that 0.65 pounds of benzene are required for
each pound of nitrobenzene produced.
-------
B-10
TABLE B-8. CHLOROBENZENE MANUFACTURING PLANTS
(PLANT LOCATIONS, COMPANIES, AND ANNUAL
CAPACITY FOR BENZENE CONSUMPTION)
Benzene Requirement at >
State/City Company 100 Percent Capacity
(million pounds)
Delaware
Delaware City Standard Chlorine 71
Illinois
Sauget Monsanto 109
Michigan
Midland Dow 285
Nevada
Henderson Montrose Chemicals 67
New York
Niagara Falls Hooker 14
Syracuse Allied Chemical 24
West Virginia
Natrium PPG 86
TOTAL 656
(a)
Based on the assumption that 0.950 pounds of benzene are required for
each pound of chlorobenzene produced.
-------
B-n
TABLE B-9. DETERGENT ALKYLATE MANUFACTURING PLANTS
(PLANT LOCATIONS, COMPANIES, AND ANNUAL
CAPACITY FOR BENZENE CONSUMPTION)
Benzene Requirement at .
State/City Company 100 Percent Capacity
(million pounds)
California
Carson
• Richmond
Maryland
Baltimore
Baltimore
Texas
Chocolate Bayou
West Virginia
South Charleston
TOTAL
i
Witco Chemicals
Standard Oil (Calif .) (Chevron)
Continental Oil
Continental Oil
Monsanto
Union Carbide
27
107
92
12
109
73
420
SC
BC
SC
BC
SC
SC
(a)
v ' Based on the assumption that 0.485 pounds of benzene are required for
each pound of detergent alkylate produced.
SC = straight chain (biodegradable)
BC = branched chain.
-------
B-12
TABLE B-10. CYCLOHEXANE MANUFACTURING PLANTS (PLANT
LOCATIONS, COMPANIES, AND ANNUAL CAPACITY
FOR BENZENE CONSUMPTION)
State/City
TOTAL
Company
Benzene Requirement at ,
100 Percent Capacity
(million pounds)
Puerto Rico
Guayama
Penuelas
Texas
Bay town
Beaumont
Big Spring
Borger
Corpus Chris ti
Port Arthur
Port Arthur
Sweeny
Phillips Petroleum
Commonwealth Oil
Exxon
Union Oil
American Petrofina
Phillips Petroleum
Union Pacific (Champlin)
Gulf Oil
Texaco
Phillips Petroleum i
66
37
37
32
11
37
21
31
37
80
389
(a)
Based on the assumption that 0.935 pounds of benzene are required for
each pound of cyclohexane produced.
-------
B-13
TABLE B-ll.
MAJOR U.S. BENZENE CENTERS
State/City
California/Los Angeles
San Francisco
Illinois/Chicago
Kansas /Wichita
Kentucky /Ashland
Louisiana/Baton Rouge
Lake Charles
New Orleans
Shreveport
Maryland/Baltimore
Michigan /Midland
Missouri /St. Louis
New Jersey /Elizabeth
New York/Buffalo
Oklahoma /Tulsa
Pennsylvania /Philadelphia
Pittsburgh
Texas /Ho us t on
Corpus Christi
Odessa
West Virginia/Parkersburg
Puerto Rico
Virgin Islands
Others
GRAND TOTAL
Total Benzene
Production
(in millions of
256
176
95
366
476
183
1099
117
110
220
293
—
169
176
931
403
4988
959
374
—
2161
183
80-
13815
Capacity
Consumption
pounds)
107
107
181
108
280
1254
564
104
701
257
130
14
—
568
100
5254
655
248
402
734
—
162
11930
TOTAL
363
107
357
203
646
1730
183
1663
117
214
921
550
130
183
176
1499
503
10242
1614
622
402
2895
183
208
25711
------- |