Report No. 78-BEZ-4
CD
O !
EXXON COMPANY
PHILADELPHIA7, PENNSYLVANIA
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Emission Measurement Branch
Research Triangle Park. North Carolina
-------
SET 1683 01 0278
FINAL REPORT:
BENZENE CONTROL EFFICIENCY OF
VAPOR PROCESSOR AT THE EXXON
BULK GASOLINE
LOADING TERMINAL
PHILADELPHIA, PENNSYLVANIA
EPA Contract No. 68-02-2813
Work Assignment No. 12
Prepared For:
Emission Measurement Branch
ESED, Mail Drop 13
Environmental Protection Agency
Research Triangle Park, NC 27711
February 1978
SCOTT ENVIRONMENTAL TECHNOLOGY, INC.
Plumsteadville, Pennsylvania 18949
Scott Environmental Techndosy '
-------
TABLE OF CONTENTS
Page
1.0 INTRODUCTION 1
2.0 SUMMARY OF RESULTS 2
3.0 PROCESS DESCRIPTION 4
4.0 TEST PROCEDURE 7
4.1 HYDROCARBON SAMPLING.METHOD ...... 7
4.2 HYDROCARBON AND BENZENE ANALYSIS OF VAPOR SAMPLES .... 8
4.3 LIQUID GASOLINE SAMPLES 10
4.4 FLOW MEASUREMENT 10
4.5 GASOLINE PUMPED DURING TEST 11
5.0 CALCULATIONS 12
5.1 TOTAL HYDROCARBON AND BENZENE RECOVERY EFFICIENCY .... 12
6.0 PRESENTATION OF DATA 15
6.1 GASOLINE VAPOR ANALYSIS . . . 15
6.2 BENZENE CONCENTRATIONS IN LIQUID GASOLINE SAMPLES .... 15
6.3 GASOLINE DISPENSED DURING TEST ....'...' 21
6.4 METER AND SAMPLE LOGS 21
7.0 LABORATORY EVALUATION OF 10 LITER TEDLAR BAG SAMPLING TECHNIQUES 26
8.0 REFERENCES 29
-------
-1-
SET 1683 01 0278
1.0 INTRODUCTION
Under EPA Contract 68-02-2813, Work Assignment No. 12, Scott
Environmental Technology, Inc. has performed hydrocarbon emission measure-
ments on the vapor recovery processor at the Exxon gasoline bulk loading
terminal in Philadelphia, Pennsylvania. The EPA Project Number was 78-BEZ-4.
At this Exxon bulk loading terminal, gasoline is bottom loaded into tank
trucks. The recovered tank truck vapors are processed in a compression-
refrigeration-abSOrption \vapor recovery unit. The primary objective of the
test program was to determine the processor unit efficiency in the removal
of benzene from the collected gasoline vapors. This loading terminal has
been studied previously as a complete system by the EPA and the results are
reported in Reference 1. The current program was done specifically for the
study of the processor benzene efficiency.
Measurements were made of the total hydrocarbon concentration and
the hydrocarbon characteristics at the inlet arid outlet to the processor,
and the volume of exhausted vapors. Total hydrocarbon and benzene concen-
trations were measured by taking integrated Tedlar bag samples and analyzing
them by gas chromatography. Other data collected included 'the amount of
gasoline loaded into the tank trucks during the test period.
Scott Environmental Technology Inc.
-------
-2-
SET 1683 01 0278
2.0 SUMMARY OF RESULTS
The efficiency of a Parker-Hannifin compression-refrigeration-
absorption vapor recovery unit for removing hydrocarbons and benzene was
determined on the unit installed at the Exxon Philadelphia, Pennsylvania
bulk loading terminal. The results of five test runs performed on December
16, 1977 are summarized in Table 2-1. The hydrocarbon removal efficiency
relates the amount of hydrocarbons recovered by the processor to the amount
present at the inlet of the processor. Similarly the benzene removal
efficiency relates the amount of benzene recovered by the processor to
the amount of benzene present at the processor inlet. Emission rates of
benzene in grams per run and grams per gallon of gasoline dispensed are
also presented.
The weighted average efficiency for hydrocarbons was 90.6% and
that for benzene was 95.6%. The run to run variation in efficiency was
very small, and all efficiency data for individual test runs were within
2% of the weighted average. Since the inlet concentrations and processor
operating conditions were much the same in each test, no conclusions can
be drawn on efficiencies at other conditions, e.g. different ambient
temperatures. The efficiency for benzene removal is seen to be 5% better
than for total hydrocarbons. The average benzene emissions were 0.0004
grams per gallon of gasoline dispensed.
*Mention of manufacturer or trade name does not constitute EPA endorsement.
Scott Environmental ^chnoJogy Inc
-------
SET 1683 01 0278
-3-
TABLE 2-1
TEST RESULTS
(1)
Hydrocarbon Benzene
Removal Removal
Run Efficiency Efficiency
No. % as C0H0 %
1
2
3
4(2)
5
6
J u
89.2 94.7
90.5 95.5
90.4 95.3
N 0
91.4 96,3
91.6 96.4
Benzene
Emitted
grams
14,7
17.6
13.3
DATA
9.26
14,6
Benzene
Emitted
Micrograms/gal
of Gasoline
378
453
414
389
375
Benzene
Emitted
ppm V
(as C&H6
61.8
62.8
63.4
54.2
55.7
Gallonage
Weighted
Averages
90.6%
95.6%
14.3
402
59.9
(1) The test results presented do not consider the effects of truck leaks,
These results describe emissions of the processor only. •
(E) The bag samples for run 4 were not valid due to a leak in the inlet
sample line and a defective pump in the outlet sampling system.
Scott Environmental Technology Inc.
-------
-4-
SET 1683 01 0278
3.0 PROCESS DESCRIPTION
At the Exxon Philadelphia Terminal, three out of the eight tanker
loading racks dispense gasoline. Each of the three racks is equipped with
an automatic bottom load dispenser for regular, premium, and unleaded grades
of gasoline. On one of the gasoline racks, two grades of aviation fuel are
also dispensed. A vapor return hose at each rack is manifolded overhead to
a common piping system which routes the collected vapors to the vapor
processor. Figure 3-1 is a schematic of the loading system. The processor
is a Parker-Hannifin 300 cfm compression-refrigeration-absorption unit.
This processor is shown schematically as Figure 3-2.
The gasoline vapors collected from tank truck loading operations
are first sprayed with gasoline in a saturator to raise the concentration
above the explosive range. They are then stored in a vapor holder. When
a sufficient volume of vapors has accumulated in the vapor holder, the
processor is activated. The vapors are drawn from the holding tank, com-
pressed to 50 psig and then passed through a finned tube heat exchanger for
cooling. The water and heavy hydrocarbons that condense are collected in
a separator. The remaining vapors are absorbed by bubbling through gasoline
chilled to 0°F. The liquid gasoline in the absorber is continuously
recirculated, cooled and replenished with fresh gasoline. Air with some
residual hydrocarbons collects in the top of the absorber and is vented
to the atmosphere through a control valve and flame arrester.
Premium gasoline from storage is used both to cool the refrigerant
condenser and as a source of fresh absorbent. The fresh absorbent stream
is first used in the saturator, then it passes through an economizing heat
exchanger as it enters the absorber. The absorbent also passes through
the economizing heat exchanger before being pumped back to storage.
The vapors from A to 5 tanker trucks are required to fill the
storage tank. The processor then operates until the storage tank is drained
to its low limit. The processor running time is dependent on the number of
trucks which continue to load after initial start-up. During these tests, the
minimum running time was about 8 minutes, and the maximum was 14 minutes.
Scott Environmental "technology Inc.
-------
SET 1683 01 0278
-5-
/1ATION
FUEL
AVI 00
AV80
&SOLINE
PLEADED
EXTRA
EXXON
(g
C
)
i
1
J
3
GASOLINE
UNLEADED
EXTRA
_ EXXON
® ®
C
>
\
-J
_J
1
J
j)
1
6ASOL1ME
UNLEADED
EXTRA
EXXON
® ®
r
c
i
"i
_j
~\
^
>
i
^^
~^^ *.
® ©
-7
DISTILLATE
PRODUCTS
(NO VAPOR
^RECOVERY)
»
© RACK NUMBER
VAPOR RETURN LINE
TO SATURATOR
FIGURE 3-1 DISPENSER RACK LAYOUT
EXXON BULK GASOLINE LOADING TERMINAL
PHILADELPHIA, PENNSYLVANIA
Scott Environmental Technology Inc.
-------
-6-
SET 1683 01 0278
A/WVV\/\/\/V\AAAA
ABSORBER
REFRIGERATOR
MODULE
CHILLER
CONDENSER
r
i
i
i
r.-z
i
' T
-j-- ._— -^-— ^
r- FUEL -
STORAGE
LOADING RACK
SATURATOR
FIGURE 3-2 SCHEMATIC OF COMPRESSION-REFRIGERATION-ABSORPTION
VAPOR RECOVERY PROCESSOR AT
EXXON BULK GASOLINE LOADING TERMINAL
PHILADELPHIA, PENNSYLVANIA
Scott Environmental Technology Inc.
-------
-7-
SET 1683 01 0278
4.0 TEST PROCEDURE
The processor collection efficiency for benzene and hydrocarbons
was determined by measurements of the inlet and outlet hydrocarbon concen-
trations and outlet flow rate. At the outlet, a Rockwell T-9 turbine flow
meter was installed using an adapter flange to mate the 4 inch outlet vent
to the 3 inch meter. The outlet sample port was located on the turbine flow
meter. A 10 liter Tedlar bag was filled at a constant rate from this sample
port during the course of a processor run. The temperature and pressure of
the gas in the flow meter was measured using a 0 - 10 inch water manometer
and an iron - constantan thermocouple. The inlet sample port was located
in the 6 inch line from the vapor holder to the compressor near the compressor
inlet. The inlet vapor temperature and pressure were measured using an iron -
constantan thermocouple and a 0 - 10 inch water manometer. A 10 liter Tedlar
bag was filled at a constant rate from the inlet sample port during the
course of a processor run. The sampling pumps, control valves, flow meters
and Tedlar sample bags were located in the processor control house and
connected to the inlet and outlet sample ports by approximately 30 foot
lengths of % inch Teflon tubing.
A mobile laboratory was located adjacent to the processor and
powered from AC mains. This laboratory housed the hydrocarbon analysis
equipment which consisted of a Shimadzu GC-1 gas chromatograph and a
Chromatopac E1A integrator and a Beckman Model 108 total hydrocarbon
analyzer modified for sample injection. These analyzers were calibrated
with precision standards of 51.4% propane in nitrogen, 1.22% propane in
nitrogen, and 49.8 ppm benzene in nitrogen.
4.1 HYDROCARBON SAMPLING METHOD
During processor runs, the inlet and outlet sample ports were
continuously sampled. Ten liter Tedlar bags were filled at a constant rate
in order to obtain an integrated sample over the duration of the processor
Scott Environmental Technology Inc
-------
SET 1683 01 0278
run. Figure 4-1 is a schematic of the apparatus'used to fill the Tedlar
bags. The sample was pumped by a stainless steel metal bellows pump through
a Teflon sample line at a rate of 6 - 10 SCFH. Approximately one-tenth of
this flow was directed through a stainless steel flow control valve into the
ten liter Tedlar sample bag. All surfaces in contact with the sample were
either stainless steel or Teflon. Flow rate to the Tedlar bag could be
checked and adjusted by momentarily directing the sample flow through a
selector valve to a 0 - 2 SCFH rotameter. Bypass flow rate was measured in
a 0 - 16 SCFH rotameter. Sample flow into the Tedlar bag was set at 1 SCFH
for the reported test. Identical systems were used for processor inlet and
outlet sampling.
4.2 HYDROCARBON AND BENZENE ANALYSIS OF VAPOR SAMPLES
The Tedlar bag samples were analyzed for individual hydrocarbons
and benzene using a Shimadzu - GC - Mini 1 gas chromatograph equipped with
dual flame ionization detectors. A Chromatopac E1A Shimadzu Data Processor
was used to measure peak areas. The column used was a Supelco 20% SP
2100/0.1% Carbowax 1500 on 100/120 mesh Supelcoport (D-4536) packed in
10 feet of 1/8 inch stainless steel tubing. The chromatograph was programmed
from 40°C to 160°C initially at a rate of 4°C/minute for ten minutes then
the program rate was increased to 20°C/minute. Upon reaching 160°C, it was
held isothermally until no more peaks eluted. The total analysis time was
twenty minutes. The calibration gases were a 1.22% propane in nitrogen
and 49.8 ppm benzene in air.
Samples for injection into the chromatograph were extracted from
the Tedlar bags through a rubber septum into a 100 cc gas sampling syringe.
The inlet samples were diluted 50% with room air before injection into the
chromatograph. The outlet samples were analyzed without dilution. Approxi-
mately 42 hydrocarbon species were identified and measured by chromatographic
separation.
Scott Environmental "fechnoJogy Inc
-------
SET 1683 Ql Q278
-9-
% Teflon
Sample Line
S.S. Surge
Tank
Metal Bellows
MB-21 Pump
0-2 SCFH
Flow Control
Needle Valve
t
Flow
Control
Valve
Vent
Flow Meter
0-16
FIGURE 4-1 INTEGRATED BAG SAMPLER APPARATUS
Scott Environmental Technology Inc.
-------
-10-
SET 1683 01 0278
As a check on the total hydrocarbon levels In the sample bags,
the contents of each sample bag was measured using a modified Beckman
Model 108 analyzer. This analyzer which uses a flame ionization detector
was modified by adding a *£ ml injection loop at the sample inlet to the
analyzer. Hydrocarbon free air was used as a carrier gas. Syringe samples
from the Tedlar bags were admitted into the sample loop. The resulting
peak-shaped response to an injection of a hydrocarbon sample was integrated
by the Chromatopac Integrator producing a response in units of millivolts-
seconds just like a chromatograph. This total hydrocarbon analyzer was
calibrated by injection of the same propane in nitrogen standard as the
chromatograph.
4.3 LIQUID GASOLINE SAMPLES
One sample of each gasoline product was collected during the time
of the processor tests. These samples of Regular, Unleaded, Premium, Avgas 80
and Avgas 100 gasolines were analyzed for benzene using a gas chromatograph.
Since it is difficult to duplicate liquid injections into a
chromatograph and because of the complexity of gasoline, the following
procedure was adopted for the liquid analysis. The density of each sample
was determined by weighing 50 ml at room temperature prior to each analysis.
100 yl of the sample was injected into a glass 6 liter dilution flask and
vaporized. A vapor sample from the dilution flask was injected into the
chromatograph and analyzed in the same manner.as the vapor samples collected
in the field (Section 4.2) using the 49.8 ppm benzene standard for calibration.
By applying standard gas law corrections, the weight percent of the benzene
in the gasoline was then calculated.
4.4 FLOW MEASUREMENT
The flow at the outlet of the processor was measured using a
Rockwell Model T-9 turbine flow meter. This meter was installed in place of
the normal vent pipe using a four to three inch reducer. An eight foot length
of 3" ID rubber hose was added to the outlet of the flow meter to duct the
Scott Environmental Technology Inc.
-------
-11-
SET 1683 01 0278
the exhaust vapors away from the processor and to provide normal back
pressure usually supplied by the vent pipe.
The T-9 meter dial is calibrated in cubic feet with interpolation
to 0.1 cubic foot. The flow meter was read before and after each processor
run and at two minute intervals during processor runs.
4.5 GASOLINE PUMPED DURING TEST
The gasoline loaded into the tank trucks displaces the vapors
which then enter the processor. The quantity of gasoline pumped was
recorded by reading the gasoline totalizers before and after each truck
fill. The time of the truck fill was logged and the processor run number
to which the vapors contributed was noted.
Scott Environmental "fechnotosy Inc
-------
-12-
SET 1683 01 0278
5.0 CALCULATIONS
Using the raw data inputs of outlet flow volume, inlet and outlet
total hydrocarbon and benzene concentrations and gasoline volume dispensed
the following parameters were calculated:
1. Removal efficiency for benzene
2. Removal efficiency for total hydrocarbons
3. Benzene emitted in grams per test
4. Benzene emitted in grams per gallon of gasoline pumped
5. Gallonage weighted averages of efficiency and emissions
In discussing the efficiency of a compression refrigerator -
absorption gasoline vapor processor the device may be considered as having
three ports.
.2 )Vent
I
Vapor
Processor
Vapor
Inlet
FIGURE 5-1
SjLiquid Outlet
Figure 5-1 represents the processor and is characterized by the
gas vapor inlet at uO» the liquid outlet at MM and the "air" vent at \2
Subscript numbers 1 and 2 will be used to reference the engineering param-
eters at the vapor inlet and vent.
5.1 TOTAL HYDROCARBON AND BENZENE RECOVERY EFFICIENCY
The hydrocarbon and benzene removal efficiency was calculated
from the inlet and outlet benzene and total hydrocarbon concentration and
the outlet volume. The inlet and outlet gas is a mixture of hydrocarbons
and air. It is assumed that all the air entering the system exits the
Scott Environmental Technotogy Inc
-------
-13-
SET 1683 01 0278
.system at the vent. Only the liquified hydrocarbons are removed at port 3.
Expressed mathematically:
Air in at (T) = air out at (T) (1)
HC, HC
Where V, and V^.are the inlet and outlet gas volume at standard
conditions respectively. HC.. and HC2 are the inlet and outlet concentrations
of hydrocarbons expressed as volume percent.. These actual hydrocarbon concen- /
trations were calculated from the individual hydrocarbon constituent concentrations
measured (as propane) by the instrument and adjusted for the actual effective
carbon number of each constituent. (The actual hydrocarbon concentration of each
constituent is shown in Tables 6-4 and 6-5.)
Equation (2) is rearranged to solve for V,.
v 0 v 100
1 9
100
Since V-^ and V"2 are now known the hydrocarbon recovery efficiency
as propane can be calculated.
(A)
VHC1P>
(HC. ) and (HC« ) are the inlet and outlet concentrations as propane
Ip 2p • c
expressed in parts per million by volume. Similarly the benzene recovery
efficiency can be calculated.
V (Bz ) - V (Bz )
EBE7 = ^ ' (5)
BEZ V1(Bz1)
Where Bz^ and Bz2 are the inlet and outlet benzene concentrations expressed
in parts per million by volume.
Scott Environmental Technology inc.
-------
-14-
SET 1683 01 0278
The recorded outlet volume was first corrected to standard
conditions before being used in equations 2-5.
v (*r^ - 17 fis v BP + 0.07355 P9
V2(SCF) - 17.65 V2R T + 46Q (6)
Where Von/is the recorded meter volume in cubic feet, P2 is the outlet sample
pressure in inches of water, BP is the barometric pressure in inches of
mercury and T2 is the outlet sample temperature in °F.
The emitted benzene in grams for each processor, run was obtained
from the average benzene concentration and outlet volume,
= 91.96 V2(Bz2) x io"6 (7)
The hydrocarbon mass emitted was similarly obtained.
M^ = 51.6 V2(HC2p) x 10"6 (8)
The benzene emitted per gallon pumped is obtained by dividing by
the amount of gasoline pumped which contributed to' the given processor run.
Where (M/L) m is grams benzene per gallon of gasoline pumped, and L, is
e oz . d
the total number of gallons of gasoline pumped.
Scott Environmental Techndosy Inc
-------
-15-
SET 1683 01 0278
6.0 PRESENTATION OF DATA
The test data used in the calculation of the results is presented
in Table 6-1. Table 6-2 compares the total hydrocarbon analyses (as propane)
obtained with the GC to those data obtained with the modified Beckman 108
analyzer. The results calculated using the considerations of Section 5.0
are presented in Table 6-3.
6.1 GASOLINE VAPOR ANALYSIS
The results of the analysis of the gasoline vapor sample from the
inlet and outlet of the processor are tabulated in Tables 6-4 and 6-5. The
concentrations are reported as per cent by volume. Table 6-4 lists the inlet
samples and Table 6-5 lists the outlet samples.
6.2 BENZENE CONCENTRATIONS IN LIQUID GASOLINE SAMPLES
The gasoline samples collected at the Exxon Philadelphia loading
terminal were analyzed for benzene concentration using the Shimadzu GC-1
gas chromatograph. The procedure used is described in Section 4.3 of this
report. Five samples of product were analyzed. The results are tabulated
in Table 6-6. The concentrations of benzene are reported in per cent by
weight.
TABLE 6-6
BENZENE ANALYSIS OF LIQUID GASpLINE SAMPLES
Benzene Concentration
Product Sample % by Weight
Premium 1.28
Regular 1.81
Unleaded 2.49
Avgas 80 0.-36
Avgas 100 1.06
Scott Environmental Technotogy Inc
-------
©
Scott Environmental "
5
^^
R
Inlet
HC
Run Cone .
No. %V C0
.3
1 30.17
2 35.64
3 34 . 85
4*
5 36.69
6 39.13
* Inlet pump
Outlet
HC
Cone.
%v c3
3.94
4.28
4.21
—
4.03
4.33
failed
Average C Number
Inlet Outlet
4.53 3.92
4.48 3.84
4.48 3.85
—
4.50 3.88
4.46 3.86
during test
TABLE 6-1
TEST DATA
Benzene Cone.
ppmV CgH,-
Volume
Inlet Outlet cu. ft.
968.6 61.8 2227
1091 62.8 2604
1072 63.4 1959
1460
1136 54.2 1622
1165 55.7 2442
Outlet
Temp.
OF
4.3
1.8
5.4
7.0
12.4
4.6
Press.
Inches
H00
5.3
5.9
5.3
5.5
5.6
6.1
Gasoline
gallons
39,001
38,933
32,001
22,786
23,804
38,907
Operating
Time
(minutes)
12.23
14.00
10.88
7.87
8.88
13.33
w
H-
CO
o
I-1
0
to
oo
Ancillary Data
Test Date: December 16, 1977
Barometric Pressure: 30.24 inches Hg
Air Temperature: 30-45°F
-------
Sc
£
$
TABLE 6-3
CALCULATED DATA
Run
No.
1
2
3
4
5
6
Inlet
Volume
SCF
3142.6
3876.3
2870.3
—
2382.9
3733.8
Outlet
Volume
SCF
2593
3053
2275
1691
1858
2847
Actual
Hydrocarbon
Cone. (%)
Inlet
19.98
23.87
23.34
—
24.46
26.32
Outlet
3.02
3.34
3.28
—
3.12
3.37
HC Mass
as CgHg
(grams)
Inlet
48923
71286
51616
—
45113
75390
Outlet
5272
6443
4942
—
3864
6361
Benzene
Mass
(grams)
Inlet
279.9
388.9
283.0
—
248.9
400.0
Outlet
14.74
17.63
13.26
—
9.26
14.58
Outlet
Benzene
y gm/gal
378
453
414
—
389
375
Benzene
Removal
Efficiency
% as CfiH,
94.7
95.5
95.3
—
96.3
96.4
en
. w
H
oo
u>
o
HC
o
Removal N>
Efficiency oo
% as CH.
J O "
89.2
90.5
90.4
—
91.4
91.6 £
-------
-17-
SET 1683 01 0278
TABLE 6-2
COMPARISON OF TOTAL HYDROCARBON ANALYSES USING TWO INSTRUMENTS
Total Hydrocarbons as Propane
Gas Chromatograph
Run No.
1
2
3
5
6
Inlet
30.17
35.64
34.85
36.69
39.13
Outlet
3.94
4.28
4.21
4.03
4.33
Modified Beckman 108
Inlet
31.7
38.2
35.9
38.9
41,0
Outlet
4.20
4.39
4.05
3.99
4.46
Scott Environmental Technolosy Inc
-------
-19-
SET 1683 01 0278 TABLE 6"4
COMPOSITION OF HYDROCARBON VAPOR AT PROCESSOR INLET
Volume %
"~" — -— -^Samj>le Number
Compound • _
TOTAL HC (% C7)
METHANE
ETHANE
PROPANE
ISOBUTANE
N-BUTANE+C°
2-BUTENES
3-ME-l-BOTENE
ISOPENTANE
N-PENTANE+Cs
2-ME-2-BUTENE
2, 2-DiME- BUTANE
Cfi OLEFIN
2-ME-PENTANE+CYCLOPENTANE
3-ME-PENTANE
N-HEXANE+Cg
Cf, OLEFIN
C^ OLEFIN
2 , 4-DiME-PENTANE+ME-CYCLOPENTANE
BENZENE
CYCLOHEXANE
2-ME-HEXANE+2 , 3-DiME-PENTANE
3-ME-HEXANE
ISO-OCTANE
N-HEPTANE
C7 OLEFIN
C7 SATURATE
2,2, 3-TRI-ME-PENTANE+C8 SATURATES
TOLUENE+2 , 3 , 4-TRI-ME-PENTANE
C% SATURATE
Cg SATURATE
N-OCTANE
Cg SATURATE
Cg SATURATE
M+P-XYLENE+ETHYL-BENZENE
0-XYLENE
N-NONANE
ISOPROPYL-BENZENE
l-ET,4-ME-BENZ+l-ET,3-ME-BENZ+MESITYLENt
1-ETHYL , 2-METHYL-BENZENE
1,2, 4-TRI-ME- BENZ+N-PR-BENZ
CIQ AROMATIC
Cio AROMATIC
TOTAL (%)
1
30.17
0.2616
0.0609
0.3634
2.3280
. 8.9194
0.0531
0.0133
: 3.9806
1.9700
0.3151
0.0781
0.0239
0.5871
0.2396
0.2362
~
~
0.1434
0.0969
~
0.1203
—
0.1246
0.0326
~
0.0124
0.0210
0.1566
—
0.0122
0.0065
0.0013
0.0023
0.0365
0.0092
—
0.0003
0.0041
0.0006
0.0016
—
—
19.977
2
35.64
0.3261
0.0754
0.4339
2.7164
10.6175
0.0646
0.0157
4.7153
2.2939
0.3643
0.0927
0.0278
0.6892
0.2809
0.2884
—
—
0.1664
0,1091
—
0.1351
—
0.1352
0.0356
—
0.0136
0.0214
0.1629
—
0.0118
0.0063
0.0015
0.0027
0.0420
0.0146
—
0.0029
0.0125
0.0022
0.0068
0.0017
0.0006
23.89
3
34.58
0.3066
. 0.0729
0.4096
2.6712
10.3458
0.0550
0.0149
4.6268
2.2453
0.3588
0.0912
0.0279
0,6743
0.2764
0.2839
—
—
0.1632
0.1072
--
0.1319
—
0.1345
0.0339
—
0.0133
0.0218
0.1632
—
0.0121
0.0066
0.0014
0.0023
0.0411
0.0131
—
0.0022
0.0119
0.0018
0.0062
—
—
23.33
5
36.69
0.3102
0.0762
0.4186
2.8477
10.9302
0.0592
0.0151
4.8751
2.3718
0.3767
0.0972
0.0294
0.7164
0.2911
0.3035
—
—
0.1722
0.1136
—
0 . 1380
—
0.1386
0.0366
—
0.0140
0.0219
0.1607
—
0.0114
0.0060
0.0012
0.0018
0.0351
0.0100
—
0.0011
6
39.13
0.3036 '
0.0807 3
0.4446
2.8941
12.1476 S
0.0642
0.0158
5.1161
2.4636
0.3920
0.1013
0.0315
0.7381
0.3037
0.3103
—
—
0.1777
0.1164
—
0.1387
—
0.1409
0.0381
—
0.0141
0.0225
0.1607
—
0.0115
0.0064
0.0011
0.0017
0.0352
0.0105
—
0.0016
0.0068 0.0072
, 0.0008 0.0010
0.0033 0.0032
~
—
—
24.44 26.30
-------
SET 1683 01 0278
-20-
TABLE 6-5
COMPOSITION OF HYDROCARBON VAPOR AT PROCESSOR OUTLET
Volume %
" Sample Number
Compound '
TOTAL HC (% C,)
METHANE
•ETHANE
PROPANE
ISOBUTANE .
N-BUTANE+C"
2-BUTENES
3-ME-l-BUTENE
ISOPENTANE
N-PENTANE+C°
2-ME-2-BUTENE
2, 2-DiME- BUTANE
Cf, OLEFIN
2-ME-PENTANE+CYCLOPENTANE
3-ME-PENTANE
N-HEXANE+Cg
Cf. OLEFIN
Cfi OLEFIN
2,4-DiME-PENTANE+ME-CYCLOPENTANE
BENZENE
CTCLOHEXANE
2-ME-HEXANE+2 , 3-DiME-PENTANE
3-ME-HEXANE
ISO-OCTANE
N-HEPTANE
Cy OLEFIN
C7 SATURATE
2,2,3-TRI-ME-PENTANE+Cn SATURATES
TOLUENE-t-2 , 3 , 4-TRI-ME-PENTANE
CR SATURATE
C? SATURATE
N-OCTANE
Cg SATURATE
Cn SATURATE
M+P-XYLENE-f-ETHYL-BENZENE
0-XYLENE
N-NONANE
ISOPROPYL-BENZENE
1-ET , 4-ME-BENZ+L-ET , 3-ME-B.ENZ+MESITYLENI
1-ETHYL , 2-METHYL- BENZENE
1,2, 4-TRI-ME-BENZ+N-PR-BENZ
CIQ AROMATIC
Cm AROMATIC
TOTAL (%)
1
3.94
0.3375
0.0562
0.1280
0.4095
.1.3496
0.0061
0.0013
;. 0.3881
0.1612
0.0228
0.0058
0.0017
0.0401
0.0158
0.0155
—
0.0088
0.0062
—
0.0086
—
0.0095
0.0018
—
0.0006
0.0016
0.0132
—
0.0012
0.0009
—
• —
0.0097
0.0050
—
0.0027
0.0056
—
0.0024
0.0005
0.0013
3.02
2
4.28
0.4050
0.0633
0.1430
0.4591
1.4897
0.0066
0.0013
0.4236
0.1746
0.0244
0.0065
0.0019
0.0428
0.0167
0.0164
—
0.0091
0.0065
—
0.0086
—
0.0108
0.0021
—
0.0006
0.0017
0.0133
—
0.0011
0.0004
<0.0001
0.0003
0.0070
0.0022
—
0.0001
0.0037
0.0006
0.0027
0.0005
0.0013
3.35
3
4.21
0.4107
. 0.0642
0.1402
0.4463
1.4566
0.0063
0.0013
0.4123
0.1696
0.0234
0.0062
0.0018
.0.0457
0.0160
0.0156
—
0.0088
0.0063
—
0.0071
—
0.0079
0.0017
—
0.0006
0.0012
0.0116
—
0.0010
0.0004
—
—
0.0057
0.0021
—
0.0004
0.0035
—
0.0028
0.0009
0.0014
3.28
5
4.03
0.3636
0.0631
0.1326
0.4264
1.4089
0.0063
0.0013
0.3993
0.1630
0.0227
0.0058
0.0017
0.0391
0.0151
0.0146
—
0.0080
0.0054
—
0.0062
—
0.0067
0.0014
—
0.0004
0.0009
> 0.0035
—
0.0007
0.0004
—
—
0.0051
0.0023
—
0.0008
0.0032
—
0.0025
0.0010
0.0029
3.12
6
4.33
0.3864 !
0.0690 1
0.1454
0.4585
1.5367
0.0061
0.0013
0.4336
0.1765
0.0245
0.0064
0.0019
0.0419
0.0161
0.0155
__
—
0.0085
0.0056
—
0.0061
—
0.0065
0.0013
—
0.0004
0.0008
0.0071
—
0.0003
6.0002
—
0.0035
0.0018
—
0.0004
0.0022
—
0.0013
0.0002
0.0004
3.37
-------
-21-
SET 1683 01 Q278
6.3 GASOLINE DISPENSED DURING TEST
Table 6-7 is a synopsis of the gasoline loading activity on the
processor test day of December 12, 1977. The events are in order of
occurrence and the processor run to which the vapors contributed is marked.
6.4 METER AND SAMPLE LOGS
Table 6-8 is the log of the outlet meter readings and Tedlar bag
sampling times. These are reproductions of the original field logs.
Scott Environmental TechnoSosy Inc
-------
-22-
SET 1683 01 0278
TABLE 6-7
GASOLINE LOADING RECORD AT EXXON PHILADELPHIA TERMINAL
December 12, 1977
Run
No.
1
1
1
1
1
Truck
No.
1
2
3
4
5
Total Run No. 1
2
2
2
2
2
6
7
8
9
10
Total-Run No. 2
3
3
3
3
11
12
13
14
6
6
6
6
21
22
23
24
Rack
No.
8
4
6
4
6
8
6
6
4
8
6
4
6
8
Total Run No. 3
4 15
4 16
4 17
Total Run No. 4
5 18
5 19
5 20
Total Run No. 5
8
6
4
6
6
4
8
6
4
6
Time
EST
0842
0908
0911
0935
0940
1000
1020
1033
1035
1038
1121
1226
1240
1249
1306
1309
1317
-1450
1500
1505
1545
1546
1552
1558
Total Run No. 6
TOTALS
Unleaded
2900
2300
3000
2300
2300
12800
2301
2727
2901
5898
13827
2300
3000
4002
3001
12303
3298
3002
901
7201
5702*
2799
8501
5201
6301
3600
2301
17403
72,035
Premium
1101
1800
1101
1101
1699
6802
2104
1001
1800
2101
7006
2101
1100
1099
2099
6399
701
1100
1799
3600
4603*
1.101
5704
2804
1100
2100
6004
35,515
Regular
4000
2900
3900
4599
4000
19399
3600
4100
3300
5100
16100
3600
3899
2900
2900
13299
4000
3900
4085
11985
5700*
3900
9600
1701
6898
3300
3601
15500
85,883
Avgas 80
2000
2000
2,000
*Trucks 18 and 19 combined.
Scott Environmental Technology Inc.
-------
-23-
TABLE 6-8
METER DATA SHEET
Project 1683
Date / X// 6 /7 >
Site
Barometer
Technician
Meter S/N
Bag
No.
Start/Stop
Clock
Time
Bag
Elapsed
Time
Meter
Reading
\\
Static
P.
Temo.
Notes
. o
. 1
1
O
//*?
ftj>>
lot (
/O
10
.r
/ft;?
( f 1 0
r
2.0? 1
3
2,
T ^« 3,
0
6
IC,oO
0
I&&O
Scott Envircnmenta] T^dinc-Jc^y
-------
-24-
TABLE 6-8
(Continued)
METER DATA SHEET
Project 1683
Date lH/fto/77 Barometer 3> °, 2. *-f
Site ,£""# 7f 0 n Technician / C^/' ±s ^ —
/u./Tr* Meter S/N /&*/Lu^ VM
Sag
fo.
) /
"" >>»
3-S
Start/Stop
cirx
#W (
/0*
/
/i^2_
Clock
Time
o^x?
$r, r
6'nfa
t*$£^*&
\G 2. ^?
10^ 0
10 3 Z.
10 m
fOllf
/o3^
to to
i^^fZ-
(^ V
lu-1 b
IM,?
12. TO
/^rz
/^s~3>
Bag
Elapsed
Time
/v /v
j^f.' a1*
/(}' ' 5\'f
j#
Meter
Reading
ityrity
1 ' O ^f fe? ^
/ te-7
1- 4-'(
r6. X"
^.5^
*^( ^ ''
*-rif
^,1JL4
^ /
^ /
P^. 6
^7 7
y r, b
r^-x
T f»0
fr-^
. rv ~~ -
•f^v *
Temp.
10
JT"
X
0
^
3
^?
0
0
0
o
(3
f (f}
/o
(^
f
f
^o
/)
0
Notes
/7 "fcA^y
•
^ 3^,2.3
A/, 3ar^0
_tr\f\-
{\} Scott Environmental "technology Inc
-------
-25-
TABLE 6-8
(Continued)
METER DATA SHEET
Project 1683
Date /;L
Site
7
Barometer
Technician
Meter S/N
Bag
No.
Start/Stop
Clock
Time
Bag
Elapsed
Time
Meter
Reading
Static
P.
Temp.
Notes
ftf
-0
/c/.1 n-
Qj
-------
-26-
7.0 LABORATORY EVALUATION OF 10 LITER TEDLAR BAG SAMPLING TECHNIQUE
The test procedure specified by EPA for the reported program
was the 9/27/77 draft benzene test method (Reference 1). In this procedure
50 liter Tedlar or aluminized mylar sample bags are filled at 0.5 1pm
using the specified apparatus which is reproduced as Figure 7-1. The
sample bag is evacuated prior to sampling then connected to the sample
probe and filled by drawing air out of the "rigid leak-proof container".
\
A modified procedure was used during the test program because
the intermittent operation of the process limited the use of the draft
procedure. The length of the vapor processor sampling runs varied from
7 to 14 minutes and use of the 50 liter sample bags would have required
either sampling at an increased sampling rate or integrating the single
sample over several processor runs. This would have been necessary to
collect a sufficient volume of sample to minimize the potential error
introduced by the residual volume of air in the evacuated bag. There is
also a significant period of time required between sampling runs to
properly follow the draft procedure whereas the truck loading operation
could not be controlled that closely without disrupting plant operation.
As a result an alternative procedure was employed with smaller 10 liter
bags using the apparatus shown in Figure 4-1. A laboratory test was
conducted prior to field use to evaluate the sampling method and materials..
A gasoline vapor sample drawn from an automobile fuel tank was used to
fill a 100 liter Tedlar bag in a barrel container. Some of the contents
of the 100 liter bag were transfilled to a 10 liter bag using the
apparatus of Figure 4-1.
-------
SET 1683 01 0278
-27-
STACKWALL
/
FILTER (GLASS WOOL) /
/X PROBE
TEFLON
SAMPLE LINE
NO
CHECKS r—! CHECKS
QUICK
CONNECTS
FEMALE
TEDLAR-
BAG
OR
ALUMINIZED
MYLAR
PUMP
RIGID LEAK-PROOF
CONTAINER '.
FIGURE 7-1 BENZENE METHOD INTEGRATED-BAG SAMPLING TRAIN
Scott Environmental Technoiogy Inc
-------
-28-
Syringe samples from both the 100 liter bag and the ten liter bag were
analyzed for total hydrocarbons by a modified Beckman 109 total hydrocarbon
analyzer and a Shimadzu gas chromatograph. Samples from the ten liter bag
were also analyzed after 24 hour bag storage. The results of this test are
presented in Table 7-1.
The agreement between the two bag samples was within 5% by both
analysis techniques. The agreement between the two techniques was within 9%,
These data were taken during the development of the total hydrocarbon injec-
tion method and better agreement between the chromatograph and total hydro-
carbon analyzer were experienced during the field sampling. The benzene
concentrations in the two bags agreed within 6% for the low concentration
bags and within 4% for the high concentration bags. For the high concentration
case, the GC shows one bag higher than the other with the .THC vice-versa. For
the low concentration bag, the differences between bags and instruments is
reversed. The conclusion was that e?;perimental errors were larger than the
differences between the two sampling systems and that the ten liter bag and
its associated sampling apparatus would give equivalent results to using 50
liter bag of the draft benzene procedure.
TABLE 7-1
10 LITER TEDLAR BAG TEST RESULTS
10 liter bag
100 liter bag
10 liter bag
100 liter bag
Cone, as
by THC
37
36
5.01
5.10
Cone, as
by GC
33.8
35.1
4.53
4.38
Cone,
Cr11,
—6-6—
1180
1235
165
155
Remarks
10 liter bag sample after
storage by THC - 37%
Scott Environmental Technology
-------
-29-
SET 1683 01 0278
8.0 REFERENCES
1. Emissions from Gasoline Transfer Operations at Exxon Company, U.S.A.,
Philadelphia Terminal, Philadelphia, Pennsylvania.
US Environmental Protection Agency EMB Report No. 75-GAS-10, December 1974.
2. EPA Draft Benzene Test Method. September 27, 1977.
Scott Environmental Technology Inc.
-------
Ar-1
SET 1683 0.1 0278
APPENDIX A
ANALYSIS OF AUDIT SAMPLES
An audit sample consisting of a benzene in nitrogen mixture con-
tained in a B size aluminum high pressure cylinder was delivered to Scott
by EPA. The analysis of this audit sample was used to verify the validity
of Scott's benzene analysis technique.
The audit sample was analyzed in the field by gas chromatogr.aph
against Scott's laboratory standards using exactly the same GC, column,
instrumentation and procedure as was used for the reported gasoline vapor
analysis. The cylinder was then returned to Scott's Plumsteadville labora-
tory and reanalyzed using the same equipment and procedure. In the
reanalysis, the concentration results could not be duplicated. It was felt
that the concentration in the cylinder changed while gradually warming to
room temperature (~70°F). It had been outside (average temperature 30°F)
for two days prior to the field analysis. To confirm this temperature
instability, the cylinder was again allowed to stand outside overnight and
then a series of duplicate analyses was performed using two analysts and
two instruments. All samples were taken directly from the cylinder. The
following results were obtained:
Cone, (ppm)
Date
12/16
12/16
12/19
12/20
12/20
12/20
12/20
12/20
12/20
281
280
281-286
281
286
295
297
285
285
Instrument
Shimadzu
Shimadzu
Shimadzu
Perkin Elmer
Shimadzu
Perkin Elmer
Shimadzu
Perkin Elmer
Shimadzu
Notes
Field analysis
Field analysis
Several lab analyses
During warm up from
outside ambient (~30°F)
During warm up from
outside ambient (~30°F)
After warming on hot
plate to >100°F
After warming on hot
plate to >100°F
After equilibrating to
room temperature
After equilibrating to
room temperature
Scott Environmental Technology Inc.
-------
A-2
SET 1683 01 0273
Out conclusion from these results is that wall adsorption occurred
in the audit cylinder. This resulted in gradual desorption and concentration
increase upon warming and vice versa upon cooling. Wall adsorption can also
cause a gradual overall increase in concentration as cylinder pressure drops.
At the time of analysis, the pressure was in the neighborhood of 500 psi.
Scott Environmental Technology Inc.
------- |