RELIABILITY OBSERVATIONS AND
EMISSION MEASUREMENTS
AT
GASOLINE TRANSFER
VAPOR RECOVERY SYSTEMS
P.J. Powell
D.E. Hasselmann
TRW, Inc.
Transportation and Environmental Operations
One Space Park, Redondo Beach, California 90278
Contract 68-02-0235
November 1974
Prepared for:
Emission Measurements Branch
Office of A1r Quality Planning and Standards
Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RELIABILITY OBSERVATIONS AND
EMISSION MEASUREMENTS
AT
GASOLINE TRANSFER
VAPOR RECOVERY SYSTEMS
D.J. Powell
D.E. Hasselmann
TRW, Inc.
Transportation and Environmental Operations
One Space Park, Redondo Beach, California 90278
Contract 68-02-0235
November 1974
Prepared for:
Emission Measurements Branch
Office of Air Quality Planning and Standards
Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
1.0 INTRODUCTION
This report describes the results of a five week study performed by
TRW Environmental Services pursuant to Contract Number 68-02-0235 with the
EPA. The purposes of the study were: to assess the reliability of vapor
control systems at gasoline service stations and to measure gasoline vapor
concentrations at various emission points of the systems. The study con-
sisted of random and unannounced inspections of gasoline vapor recovery
devices at filling stations in the San Diego area between 7/11/74 and 8/16/74.
The tests were observed by representatives of the San Diego Air Pollution
Control District and the California Air Resources Board.
Five secondary vapor recovery systems were reviewed at 23 locations:
Process Products, Environics, Intermark, Clean Air Engineering, and Shell Oil.
In addition, two displacement (balanced) vapor recovery systems were visited
during the test period: Standard Oil and Gulf Oil.
This study parallels a series of tests conducted by TRW Environmental
Services for the EPA and County of San Diego between 6/6/74 and 6/26/74. In
that study five different gasoline vapor recovery systems in the San Diego
area were tested for process efficiency.
1-1
-------
2.0 CONCLUSIONS
The 23 vacuum assisted vapor recovery systems were found to be inoperative
5.4% of the time during the test period. The individual average down time for
each make of vapor recovery system are listed in Table 1.
Table 1 Average System Down Time
Vapor Recovery System
Environics
Intermark
Process Products
Clean Air Engineerinc
Shell Oil
# Units
Studied
14
6
1
1
1
# Of Days In
Study Period
36
36
36/u \
25(b)
36
Unit Down
Days
25
17
1
1
0
% Of Test / »
Period Down^a;
4.96
7.87
2.77
3.85
0
(a)
(b)
total unit down days x 100 =% of test
(# of units studied)(# of days in study period) period down
from 7/11/74 to 8/6/74 inclusive; on the evening of 8/6/74 the process
unit was struck by an automobile and was inoperative for the remainder
of the test period.
Aside from the Intermark unit at Carmel Valley Rd., and the Environics
unit at Baltimore Rd., the breakdowns noted did not involve failure of any
major component of the units, and were repaired within three days. Most of the
breakdowns noted were the result of the failure of a small high speed blower
belt which is used in the Environics, Process Products and Intermark units.
Several of the vapor recovery units were found to have vapor leaks which
would significantly lower their efficiency. The Intermark units at Carmel
Valley Rd. and at the Gemco Station in National City had considerable volumes
of vapor leaking from their surge tank vents. The manufacturer subsequently
confirmed that a tear had been found in the surge tank bladder of the unit at
Carmel Valley Rd.
2-1
-------
Vapor leakage from a crack in the concrete next to the Process Products
unit at the Union Station on Waring Rd. became evident on July 27th. It was
subsequently found to be due to leakage in the underground vapor lines. Al-
though the lines were excavated and one leak was repaired, leakage was still
occurring at the end of the test period. This incident indicates the need
for more comprehensive pressure testing of systems, particularly those designed
to operate with pressure on the underground storage tank and vapor piping.
Three of the Environics Vapox 400 units visited had no pressure/vacuum
(P/V) valve on the underground storage tank vent (San Diego Marine Construc-
tion, Frank Motors, San Diego County at San Marcos). Leakage was also verified
by an explosimeter check of the P/V valve on the underground storage tank vent
at the Darr's Liquor Environics unit. Since the Environics system doesn't
pressurize the underground tank, emissions from this source would be due mainly
to breathing losses and decreased efficiency of retention of vapors during
bulk gasoline delivery. A slight leak was also discovered in the Process
Products unit P/V valve which, due to the induced positive pressure in this
system, resulted in continuous vapor leakage.
An intentional leakage was found to have been brought about by the lodging
of a piece of rubber tubing in the P/V valve at the Intermark unit at Carmel
Valley Rd. A few days later the entire interior of the valve was found to have
been removed. As a result large volumes of vapor were vented to the atmosphere
rather than going through the Process unit.
Appreciable leakage was noted at the nozzle-filler neck interface with all
systems. Of the 60 fuel ings monitored at the vacuum assisted systems, 32 showed
some explosimeter reading. Of the 11 vehicle fuel ings monitored at balanced
stations, 8 showed some explosimeter reading. The major causes of this leakage
appeared to be the lack of a tight seal at the nozzle-filler neck interface, or
insufficient applied vacuum to compensate for the pressure buildup in the vehicle
tank caused by the liquid flowing into it.
Several cases of vapor breakthrough in the carbon beds of Environics and
Process Products units were indicated by explosimeter measurements at their
2-2
-------
vents. This indicates a need for either larger carbon beds or longer regenera-
tion time.
Several station operators (both vacuum assist and balanced units) stated
that they had noticed an increase in gasoline spitback since the vapor recovery
systems had been installed. Spitback occurred in 15% of the vehicles fuelings
observed during the test period. An increase in spitback could be attributed
to the fact that some of the nozzles now being used have been converted to
vapor recovery by the addition of attachments. The addition of vapor recovery
apparatus to an ordinary nozzle results in the automatic shutoff sensor not
reaching as far down into the vehicle filler neck as it would without the
modifications. This allows the gasoline to rise farther up the filler neck
as the gasoline tank becomes full, and permits less reaction time for the sensor
to shut off the pump.
Another probably cause of spitback is the fact that in automobiles which
have their filler necks under the license plate or rear bumper it is often
necessary to place the nozzle in sideways or even upside down. Since the
automatic shutoff sensor is designed to perform its function.with the nozzle
in the normal vertical position, such extraordinary positioning can preclude
that function and result in spitback.
Some component deterioration was noted during the study period. Vapor re-
turn hosing was found to have significant amounts of wear at the Standard Sta-
tion on Carmel Valley Rd., at the Texaco Station at Baltimore and Fletcher, and
at the Crest Beverage Co. This hose wear appeared to be the result of the hoses
rubbing against the pump housing or against the vehicle during fueling. Nozzle
boot wear was noted at all of the Intermark units and the Shell unit. The boot
wear also seemed to be a result of normal use, and would eventually necessitate
replacement to maintain optimum collection efficiency.
A number of potential fire and safety hazards were noted during the .test
period. The primary cause appears to be careless installation and use of inap-
propriate materials. To preclude these hazardous conditions, specifications
should be set as to the type of wiring and other building materials used in
constructing the vapor recovery devices. Requirement of flame and waterproof
circuitry, pressure tested piping, and specific foundation material would
remedy most of the deficiencies.
2-3
-------
The age of the vapor recovery devices visited ranged from two years to
less than a few weeks. Several of the units visited were prototypes which
were the first ones built by their companies. Modifications of nozzles,
hosing, and process unit components were noted during the test period. These
modifications appeared to be normal developmental changes to improve the re-
liability of the systems.
2-4
-------
3.0 DESCRIPTION OF DEVICES AND INSPECTION POINTS
3.1 CLEAN AIR ENGINEERING
The Clean Air Engineering vapor recovery device uses a combination of car-
bon adsorption, air sweep regeneration, and incineration of the vapors in
an internal combustion engine fitted with a catalytic exhaust purifier. The
blower comes on when any gasoline dispenser is turned on. Vacuum at each
nozzle is not started until the nozzle lever is depressed. As gasoline is
dispensed, the vapors are drawn into the vapor return lines, some of the vapors
going to the underground tank to replace the dispensed liquid, and the excess
going into the carbon beds. When the vapor is forced into the carbon beds, the
gasoline is adsorbed and clean air is vented to the atmosphere. A time delay
relay activates the internal combustion engine starter when gasoline dispensing
begins. The internal combustion engine continues to run, drawing the gasoline
off of the carbon bed, until the air stream coming from the carbon beds no
longer contains enough gasoline to keep the engine running.
Explosimeter checks were done at the carbon bed vents, engine exhaust
(see Figure 1), and at the nozzle/filler neck interface during vehicle fueling.
This unit at Frank Dana's Service was being used to operate an air com-
pressor to provide compressed air for the station. The internal combustion
engine ran more often than was necessary to perform this function, however,
so that more compressed air was produced than was needed. The internal combus-
tion engine ran intermittently, its daily running time depending upon the amount
of gasoline dispensed.
3.2 SHELL SYSTEM
This system uses a combination of carbon adsorption, vacuum regeneration,
and absorption of the collected vapors in the gasoline in the underground tank.
As in the Clean Air Engineering system, the blower comes on when a dispenser
is turned on, and the vacuum at the nozzle is activated by the nozzle lever.
The vapors displaced by the dispensed gasoline are drawn into one of the carbon
beds (each carbon bed weighs approximately one ton). The carbon beds alternate;
one is adsorbing vapors while the other is being stripped by the vacuum pump.
3-1
-------
The vapors desorbed from the carbon bed are routed to the underground storage
tank where they are absorbed into the gasoline. Explosimeter checks were
made at the system vent (see Figure 2), and at the nozzle/filler neck inter-
face.
3.3 INTERMARK SYSTEM
The Intermark System uses refrigeration and compression to condense the
vapors back to liquid gasoline, which are diverted back to the underground
tank. As in the two systems described above, the blower comes on when any
pump is turned on, and the vacuum at each nozzle is activated by the nozzle
lever. Vapors displaced during vehicle fueling are drawn into the vapor re-
turn lines, part of it going to the underground tank to replace the volume
of gasoline dispensed, and the excess is bubbled through gasoline in the
bottom of the surge tank. The vapors in the surge tank flow through a two
stage high pressure refrigeration unit. The condensed vapors are returned
to the surge tank reservoir, which overflows into the underground tank. The
dry air is vented to the atmosphere from the refrigeration unit.
Explosimeter checks were made at the refrigeration unit vent, surge tank
vent (see Figure 3), amd at the nozzle/filler neck interface during vehicle
refueling.
3.4 ENVIRONICS SYSTEM
Environics has several different models of its system installed in gasoline
filling stations in San Diego County. All of these models use a combination
of carbon adsorption and incineration of the vapors in a reactor.
The simplest Environics model is the Vapox 400 (see Figure 4). The reactor
in this model is controlled by a timer which is set so that the reactor runs
30 minutes for every 20 seconds of fill time. The displaced vapors are drawn
through the vapor return lines into the carbon bed. When filling of the
vehicle ceases, the gasoline is desorbed from the carbon bed into the reactor.
In some of the Vapox 400 units visited, the reactor ran continuously 24 hours
a day (Frank Motors, San Diego Marine Construction, and Sparkletts Drinking
Water).
The Environics Vapox 1500, 3000, and 4000 models work on the same principle
as the 400 model, but have larger carbon beds. The Vapox 1500 model has two
3-2
-------
carbon beds connected in parallel to process a larger volume of vapors. The
carbon beds are separated by solenoid valves so that one can be isolated to
accept vapors during a bulk delivery of gasoline to the station.
The Vapox 3000 has three carbon beds, two of which adsorb vapors from
vehicle refueling, and one to accept excess vapors during a bulk delivery. The
carbon beds are linked in parallel through solenoid valves which allow
them to be purged of vapors separately after vehicle fueling has ceased.
The Vapox 4000 model has a single large canister (approximately 1000
pounds) des gned to accept the excess vapors generated during the filling of
gasoline tanker trucks. Since the tankers have standarized filler necks, a
tight seal is achieved through a coupling, and intake of excess air is minimized.
Explosimeter checks were made at the reactor exhaust, carbon bed vent
(see Figures 4 and 5), and at the nozzle/filler neck interface during vehicle
fueling. The pressure/vacuum valve on the underground tank was also checked
for leakage.
3.5 PROCESS PRODUCTS
The Process Products system uses a combination of carbon adsorption,
vacuum regeneration, and refrigeration to condense the vapors and divert the
resultant gasoline to the underground tank. The blower comes on when any
gasoline dispenser is turned on, and the vacuum at each nozzle is regulated
by a valve which adjusts air flow into the vapor recovery system in propor-
tion to the dispensed liquid flow. This regulating valve reduces the amount
of excess air brought into the system.
When the pressure in the vapor recovery system reaches three inches of
water, the regrigeration unit is activated. If the pressure decreases to one
inch of water, the refrigeration unit stops and the vacuum pump begins a
thirty minute cycle to strip the carbon canister. When the pressure in the
system rises to five inches of water, the solenoid at the top of the carbon
canister opens to bring its inside pressure up to atmospheric pressure and
the refrigeration unit is turned off. When the system pressure reaches six
and three-eighths inches of water the solenoid at the bottom of the carbon
canister opens and the excess pressure is vented out through the carbon
canister. During the venting, the vapors pass across the refrigeration coils
3-3
-------
which are at a temperature of minus 20°F, which condenses most of the vapor,
the rest being adsorbed on the carbon. When the pressure in the system is
down to one inch of water, the top solenoid valve closes and the vacuum pump
begins a stripping cycle.
Explosimeter checks were made at the carbon canister vent (see Figure 6),
and at the nozzle/filler neck interface during vehicle refueling. The pressure/
vacuum valve on the vent to the underground tank was checked for leakage by
placing a plastic bag over it.
3.6 BALANCED SYSTEMS
3.6.1 Standard Oil of California
The Standard balances system relies on the pressure differential between
the vehicle tank and underground tank to transfer vapors from the vehicle tank
into the underground tank. Since the collection of vapors at the vehicle-
nozzle interface isn't assisted by a vacuum as with the systems described above,
a tight seal is required at the interface to prevent vapors from leaking out.
Explosimeter checks were made at the vents to the underground tanks and at the
vehicle-nozzle interface. OPW Type 7VN nozzles were fitted to all dispenser
hoses.
3.6.2 Gulf Oil
The Gulf balanced system also relies on the pressure difference between
the vehicle tank and underground tank during filling to displace the vapors
to an underground tank, but instead of displacing vapors to the tank from which
the gasoline is being dispensed, all vapors are routed to the premium grade
gasoline tank. Another variation from the Standard system is that whereas each
u/g tank in the Standard system was vented directly to the atmosphere in-
dividually, the Gulf system u/g tank vents were manifolded together with one
2" galvanized pipe vented to the atmosphere. In the last week of the study
the manifolding was taken off and each of the three 2" galvanized u/g tank vents
was fitted with a 2" x 1/2" bell reducer. Explosimeter checks were performed
at the u/g tank vent and at the nozzle-vehicle filler neck interface. OPW
Type 7VN nozzles were fitted to all dispenser hoses.
3-4
-------
Pressure/Vacuum
Valve
Carbon Bed Vent
co
en
O)
CO
C CVJ
O =«*=
•e
to
o
Check
Valve>
Opening
Valve
0)
CO
c
o
-Q r—
I- =*t=
Catalytic
Exhaust
urifier
Internal
Combustion
.Engine
Pressure Openinq
Valve
ravity Drain
Point
Check Valve
Underground Tank
-n-
Figure 1
Clean Air Engineering, Inc. .'f'^-l
Carbon Adsorption, Incineration! ^
i F !
From Dispensers, Each
"With Its Own Solenoid
Valve And Flame Arrestor
Pressure/Vacuum Valve
1^1 Manual Valve
I. j -J
Solenoid Valve
Pressure Regulator
Flame Arrestor
-------
Pressure/Vacuum
Vent Valve
co
i
IFF
System
Vent
HX}
Underground Tank
Carbon Bed #1
Carbon Bed #2
Blower
'^aturator Line
Vapor Return Line
To Dispensers
Product Line
'•%'•£] Sparger
Figure 2 Shell System
Carbon Adsorption, Vacuum Regeneration, Absorption
Vacuum
Pump
-------
PRESSURE/VACUUM
VALVE
PTf
DISPENSED
LIQUID
UNDERGROUND TANK
r
RETURNED
VAPOR
BLOWER
1
SURGE TANK VENT
1
^
S"
/
1 —
SURGE TAfJK
• *c .
* * •
R
TWO
-------
Pressure/Vacuum
/\ Valve
Co
00
Underground Tank
Carbon Bed
n
Carbon Bed
#1
Carbon Bed
Vent
Pump
To Dispenser
Reactor
L
c
Reactor
Exhaust
Flame Arrestor
Solenoid Valve
[X] Manual Valve
Figure 4
. Environics, Inc.
Vapox 400
Adsorption, Air Sweep Regeneration, Incineration
-------
PRESSURE/VACUUM
VALVE
CARBON BED
VENT
OJ
vo
o
co
a:
-------
CARBON CANISTER
VENT
co
o
VALVE T
CO
I—I
-------
4.0 SUMMARY
The following tables summarize the results of the five week test
period.
Table 2 enumerates the gasoline filling stations visited, the type of
vapor recovery unit, the San Diego APCD estimate of monthly throughput, and
the average monthly throughput observed during the test period. Of the
twenty-five vapor recovery units in the test group, fourteen were Environics
units, six were Intermark units, two were balanced units, and the Shell,
Process Products and Clean Air Engineering systems were represented by one
unit each. Throughput is an indication of the vapor load processed by the
units. The table shows that the Intermark units visited had the largest
monthly throughputs, while the average throughputs for the Environics units
were the smallest (except for one Environics unit installed at a Texaco bulk
gasoline facility).
Table 3 summarizes observations of the operating status of the vapor re-
covery systems during each inspection. It also notes those instances where
the station manager reported the vapor recovery unit to be inoperative on days
that the station was not visited. The table indicates that only two of the
units were down for extended periods of time, the Intermark unit at the
Standard Station on Carmel Valley Rd., and the Environics unit at the Texaco
Station on Baltimore Dr. Of the other twenty-one secondary units in the test
group, nine were down for short periods due to relatively minor problems. The
balanced system units were not included in this table because no criteria had
been established by which to judge whether or not they were functioning
properly.
4-1
-------
Table 2 - Gasoline Filling Stations Visited
Station
Address
Vapor Recovery Device
Throughput
APCD
Estimate Observed
Frank Danna Service
Standard Oil
Standard Oil
San Diego Gas & Electric
County of San Diego
County of San Diego
County of San Diego
Texaco Bulk Plant
PhD Truck Rentals
Union Oil
Texaco
Auto Wash
San Diego Marine Construction
San Diego Terminix
Sparklett's Drinking Water
Darr's Liquor
Shell Oil
California State University, S.D
Cajon Valley School District
Sears and Roebuck, Co.
Mini Super
Crest Beveridge Co.
Frank Motors
Gemco
Gulf Oil
3746 Main St., San Diego
Baltimore & Lake Murray Rd. S.D.
3063 Carmel Valley Rd., San Diego
4848 Santa Fe Ave., San Diego
325 S. Mel rose, Vista
1600 Descanso Ave., San Marcos
5555 Overland Dr., San Diego
Via Vera Cruz & Grand Ave., San Marcos
2130 Mission Rd., Escondido
5194 Waring Rd., San Diego
5261 Baltimore Dr., San Diego
1300 E. Valley Parkway, Escondido
Samson St., San Diego
8222 Vicker St., San Diego
5930 Mission Gorge, San Diego
350 Main St., Ramona
1-805 and Balboa, San Diego
5402 College, San Diego
777 Park Avenue, El Cajon
575 Fletcher Parkway, El Cajon
9775 Maine St., Lakeside
7745 Carroll Rd., San Diego
2400 National Ave., National City
30th & Highland, National City
Boundary & University Ave., San Diego
Clean Air Engineering 2500
Balanced
Intermark, Mark I
Environics, Vapox 400
Environics, Vapox 400M
Environics, Vapox 400M
Intermark, Mark I
Environics, Vapox 4000
Intermark, Mark I
Process Products, Vapor Saver 200
Environics, Vapox 3000
Intermark, Mark I
Environics, Vapox 400
Environics, Vapox 400
Environics, Vapox 400
Environics, Vapox 1500
Shell
Environics, Vapox 400
Environics, Vapox 400
Intermark, Mark I
Environics, Vapox 1500
Environics, Vapox 1500
Environics, Vapox 400
Intermark, Mark I
Balanced
104,000
16,000
20,000
8,000
33,000
250,000
3,000
70,000
52,000
60,000
3,000
4,000
4,500
30,000
55,000
2,500
5,000
90,000
50,000
7,000
2,500
180,000
47,571
41,000
179,000
(1)
17,000
7,300
35,000
(1)
3,600
33,000
49,000
20,000
3,000
6,000
(2)
26,000
53,000
3,200
1 ,200
70,000
37,000
7,000
2,500
234,000
80,000
ro
(1) A bulk delivery facility with no gallonage meter
(2) No gallonage meter on pump
-------
Table 3 Vapor Recovery Unit Down Time
July August
STATION
Environics
San Diego Gas & Electric Co.
San Diego County, Vista
San Diego County, San Marcos
Texaco Bulk Plant, San Marcos
Texaco, Baltimore Dr., S.D.
San Diego Marine Constr. Co.
San Diego Terminix
Sparkletts, Mission Gorge Rd.
Dorr's Liquor, Ramona
California State Univ., S.D.
Cajon Valley School District
Mini Super, Lakeside
Crest Beveridge Co.
Frank Motors, National City
Intermark
Standard, Carmel Valley Rd.
County of San Diego, S.D.
PhD Truck Rentals
11 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
0 0
0 0 00 °
0 0 00 °
0 00 °
oo o oo oXXXAXXXAAAXXXAAAAA
0 0 ° ^ 0
o o • o ± oo
o o A o o °
0 00
0 0 0 o O
0 0
o XX°X o oo
O 0 o O 0
A o o o
•
o oAA o A OAAXXXA.XXAA
0 00 00
0 0 0 00
Auto Wash, Escondido
Sears, El Cajon o
Gemco, National City o o
Process Products
Union, Waring & Zion, S.D. o
Clean Air Engineering
Frank Danna Service o
Shell Oil
Shell, Balboa, S.D. o o
o
o o o o o
o o
oo oo
X o
o oo
XXX o
o A
000
o A o o
Table 6
o - unit functioning properly during a visit
A- unit not functioning during visit
X - manager said unit was not functioning
-------
Table 4 categorizes the apparent reasons for the vapor recovery unit
breakdowns listed in Table 3, and the total number of days down for each of
those units. The reasons stated in Table 4 were obtained from the manufac-
turers of the units. The most frequent cause of breakdown was the breaking
of belts which drive the blower on the Environics, Process Products, and
Intermark units. The list below enumerates these causes by unit:
Environics
Texaco Station, Baltimore Drive, S.D. - Vapox 3000
7/30 - 8/16 - The motor driving the blower burned out due to metal filings
falling into it from careless servicing
San Diego Marine Construction, Samson St., San Diego - Vapox 400
8/7 - Broken blower belt
San Diego Terminix, Vickers Street, San Diego
8/8 - Broken blower belt
Sparkletts Drinking Water, Mission Gorge, San Diego
7/25 - Broken blower belt
Mini-Super, Lakeside
8/16 - Broken blower belt
Frank Motors, National City
7/17 - Reason for breakdown not known
Intermark
Standard Station, Carmel Valley Rd. - Intermark, Mark 1. Unit had
compressors changed twice during test period. The unit also had problems
with a valve at the outlet of the refrigeration unit which would freeze
up periodically, causing the discharge of concentrated vapor to the
atmosphere.
8/2 - Freeze up of control valve on vapor condensation line
8/7 - 8/16 - Compressor inoperative
Sears, Fletcher Parkway, El Cajon - Intermark, Mark 1
8/5 - 8/8 - Apparently broken blower belt (attendant said there was no
vacuum at nozzles)
Gemco, National City - Intermark, Mark 1
8/5 - 8/8 - Freeze up of.control valve on vapor condensation line
4-4
-------
Table 4 Causes of Vapor Recovery Unit Malfunctions
System/ Station
Environics
Texaco, Baltimore Dr., San Diego
San Diego Marine Construction
San Diego Terminix
Sparklett's Drinking Water
Mini Super, Lakeside
Frank Motors, National City
Intermark
Standard, Carmel Valley Rd.
Sears, El Cajon
Gemco, National City
Clean Air Engineering
Frank Dana Service
Process Products
Union Station, Waring Rd. , S.D.
Blower Belt
1
1 day
1 day
1 day
3 days
1 day
Blower Motor
17 days
Compressor
15 days
Service- Internal
Combustion Engine
1 day
Freeze Up of Con-
trol Valve on Vapor
Condensation Line
1 day
1 day
Not Known
1 day
-------
Process Products
Union Station, Waring Road, San Diego - Vapor Saver 200
8/14 - Broken blower belt
This unit had a leak in the underground vapor lines throughout the test period
Clean Air Engineering
Frank Dana Service, Main Street, San Diego - Model 2500
7/29 - Internal combustion engine would not start. New spark plug and
breaker points were installed,
8/6 - Automobile ran into unit, putting it out of operation for the remainder
of the test period.
Table 5 summarizes instances and reasons for vapor recovery units to be
partially inoperative. These include cases where the vapor return mechanisms
and plumbing to one or more of the gasoline pumps were inoperative. In such
cases the vapor recovery efficiency at the nozzles affected were severely
curtailed, so that most of the displaced vapors were not collected. The only
vapor recovery unit in which liquid blockage in the vapor return hoses could
readily be determined was the Shell Oil unit on Balboa Rd. This unit had
clear plastic vapor return hoses. Liquid blockage was also suspected on one
occasion at the Intermark Station at Sears, El Cajon. On that occasion it
was found that there was no vacuum at the nozzle and a gurgling noise could be
heard in the vapor return hose, which indicates that the lack of vacuum may
have been caused by liquid blockage.
Table 6 summarizes the explosimeter checks made at various process unit
vents and suspected emission points. These emission points are described in
Section 3.0 of this report. Some of these readings indicate vapor leakage
and apparent loss of system efficiency.
The TRW test made during June 1974 of vapor recovery system efficiency
indicated that in order for an Environics process unit to achieve 90% efficiency,
the reactor exhaust must contain less than 0.125% (1250 ppm) hydrocarbons -
0.125% corresponds to approximately 0.1 LEL. Explosimeter readings at the re-
actor exhaust which are above 0.1 LEL are an indication that the process unit
is probably operating at less than 90% efficiency. The average explosimeter
readings of the reactor exhausts of the Environics units at San Diego Terminix,
4-6
-------
Table 5 Summary of Units Partially Inoperative
System/Station
Environics
Texaco, Baltimore Dr.
Intermark
Standard, Carmel Valley Rd.
County of San Diego, S.D.
Sear's, El Cajon
Gemco, National City
Shell
Shell , Balboa Ave.
Nozzle Out
of Order
No.
3
1
1
Date
7/22-25
7/29-31
7/17-31
Hoses
Twisted
No.
1
2
2
1
1
6
Date
7/22-25
7/29
8/8
8/12
7/25
7/25-
8/15
No Vacuum
at Nozzle
No.
1
1
1
Date
8/8
8/12
7/16
Torn Nozzle
Boot
No.
2
Date
8/8-13
Severed Vapor
Return Hose
No.
1
Date
8/12-16
Liquid
Blockage in
Vapor Lines
No.
2
Date
7/31
-------
Table 6
EXPLOSIMETER READINGS AT EMISSION POINTS
o:
Environics Units
San Diego Gas & Electric Co.
County of San Diego-Vista
County of San Diego-San Marcos
Texaco Bulk Plant-San ..Marcos
Texaco - Baltimore, S.D.
San Diego Marine Construction
San Diego Terminix
Sparklett's-Hission Gorge Rd.
Darr's Liquor - Ramona
California State Univ., S.D.
Cajon Valley School Dist.
Mini Super - Lakeside
Crest Beveridge Co.
Frank Motors
Intermark Units
Standard - Carmel Valley Rd.
County of San Diego, S.D.
PhD Truck Rentals - Escondido
Auto Wash - Escondido
Sear's - El Cajon
Gemco - National City
Process Products
Union - Waring & Zion, S.D.
Clean Air Engineering
Frank Danna Service
Balanced
Standard - Lake Murray Rd.
Gulf - Boundary & University
Shell
Shell - Balboa Ave, S.D.
0
0
70 ppm
0
600 ppm
600 ppm
0.1 LEL
180 ppm
0.3 LEL
100 ppm
100 ppm
0
0
0
1.0 LEL
0.1 LEL
.06LEL
.08LEL
0
1.0 LEL
0.14LEL
I
0
Reactor Exhaust
80 ppm
800 ppm
0
0.4 LEL
0
0.2 LEL
80 ppm
100 ppm
0.4 LEL
0
250 ppm
20 ppm
.0.2 LEL
0.2 LEL
Process Unit Vent
1.0 LEL
1.0 LEL
400 ppm
50 ppm
0.1 LEL
1.0 LEL
900 ppm
0
Process Unit Vent
0.16LEL
/C Engine
U/G Tank Vent
800 ppm
1.0 LEL
Process
.3 LEL
0.3 LEL
Exhaust
Average
0
40 ppm
70 ppm
400 ppm
600 ppm
425 ppm
0.25LEL
70 ppm
.23 LEL
90 ppm
100 ppm
0.2 LEL
0
0
Average
1.0 LEL
0.6 LEL
.05LEL
.07LEL
0
1.0 LEL
0.2 LEL
Average
Unit Vent
0
Average
.15LEL
0
0
0.2 LEL
0
0
0
0.2 LEL
0
0.2 LEL
0
0
0.3 LEL
0
20 ppm
1.0 LEL
.08LEL
200 ppm
900 ppm
1.0 LEL
1.0 LEL
0.5 LEL
0
Carb
0.86LEL
0.15LEL
0
200 ppm
0.2 LEL
400 ppm
0.9 LEL
0
0
0.7 LEL
0.6 LEL
500 ppm
Surg
300 ppm
0.4 LEL
1.0 LEL
Average
0.5 LEL
on Bed Ve
0.2 LEL
0.2 LEL
0
0
0.35LEL
0
0.3 LEL
0
0.26LEL
0
e Tank Ve
0.2 LEL
1.0 LEL
JTt
0.8LEL
0
0.2LEL
l.OLEL
jrt
l.OLEL
Average
0.5LEL
0
0.2LEL
0
0
70ppm
.3LEL
lOOppm
.5LEL
0
0
.5LEL
.3LEL
0.3LEL
Average
0.6LEL
.08LEL
200ppm
900ppm
0.7LEL
l.OLEL
Carbon Bed Vent
0
0.5 LEL
Average
0.2LEL
(1)
Explosimeter readings were taken with a Bacharach model SS-P Super
Sensitive Combustible Gas Indicator, Mo. 503-001. This model has
a dual sensitivity range: readina from 0 to 1000 opm, and in deci-
mal fraction of lower explosive level (LEL). The unit is calibrated
to give a direct reading for aromatic hydrocarbons and for pentane.
1.0 LEL is equal to 14,000 ppm (1.4 percent) as pentane. The fila-
ments in the explosimeter were replaced twice during the test period
and the unit recalibrated using methane.
4-8
-------
Darr's Liquor, and the Mini Super at Lakeside were 0.25 LEL, 0.35 LEL, and
0.2 LEL respectively. The fact that all other Environics units checked were
well below 0.1 LEL at the reactor exhaust (see Table 8) underlines the sub-
normal performance of the three units mentioned above.
Explosimeter readings at the carbon bed vents of seven of the fourteen
Environics units visited showed significant hydrocarbon breakthrough. There
was also some breakthrough noted at the Shell, Clean Air Engineering and Process
Products units. THe case of the Clean Air Engineering unit was due to a break-
down of the internal combustion engine which is supposed to incinerate the vapors,
however, and not due to normal operation. Vapor breakthrough was also noted
during a venting of the Process Products unit visited.
The explosimeter checks made at the Intermark unit sites indicated leaks
in the surge tank bladders of three of the six units visited Cat Standard Sta-
tion, Carmel Valley Rd.; Gemco, National City and Sears, El Cajon). Gasoline
vapors above the lower explosive level were also measured coming from the
process unit vent of three of the six units (Standard Station, Carmel Valley
Rd., County of San Diego and Gemco, National City). The vapor leaks at the
surge tank vent and process unit vent of the Gemco and Standard Stations were
of considerable volume on at least one visit to each site.
Table 7 summarizes the vehicle fill checks made at gasoline filling sta-
tions. Each check included monitoring the vehicle filler neck/nozzle inter-
face with an explosimeter during fueling. Table 8 below shows the number of
automobile fuel ings in which an explosimeter check showed some vapor leakage
at the vehicle/nozzle interface. This does not show the magnitude of vapor
loss since no measurement of duration or volume of vapor leakage was attempted
in this test. A positive explosimeter reading during a fueling does indicate
that 100% vapor recovery was not achieved, but cannot give an accurate estimate
of the actual recovery efficiency.
Table 8
Percentage of vehicle Fuel ings Having
Explosimeter Readings
Environics 50%
Intermark 64%
Clean Air Engineering 25%
Balanced 73%
Overal1 56%
Test program time limitations precluded sufficient explosimeter
readings at the Process Products unit to include it in these
averages. 4.9
-------
Table 7 Vehicle Fill Data
1!
Environlcs
S.D. County, San Marcos
S.D. County, Vista
Cal-State Univ., S.D.
Cajon Valley School Dist
Sparklett's
Texaco, Baltimore Dr.
Frank Motors
S.D. Marine Construction
Crest Beveridge Co. .
Mini Super, Lakeside
S.D. County, San Marcos
Sparklett's
S.D. Terminix
Darr's Liquor
Intermark
Gemco, National City
Auto Wash, Escondido
S.D. County Ops. Center
PhD Truck Rental
a
(O
0
7/24
7/24
7/24
7/24
7/24
7/24
7/26
7/26
7/29
7/29
7/29
7/29
7/29
7/29
7/29
7/30
7/30
7/31
7/31
7/31
8/1
8/1
8/2
8/12
8/15
8/16
7/23
7/23
7/23
7/24
7/24
7/25
7/31
7/31
7/31
7/31
7/31
7/31
7/31
8/1
8/6
8/15
D)
• C
O-i-^
r— -O _l
O- id ui
X
:n n:
Yes
Yes
Mo
No
Yes
. Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
No
Yes
Yes
No
Yes
No
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
No
No
Yes
No
No
No
No
Yes
No
Yes
Yes
Yes
Yes
1 Stretched!
Hose |
No
No
No
No
No
Yes
Yes
No
No
No
Yes
No
No
No
No
No
No
No
No
No
Yes
Yes
No
No
No
No
No
No
No
Yes
Yes
No
No
No
No
No
No
No
Yes
No
No
No
t- LU
Dodge
Pay loader
Ford
Dodge
Ford
Chev.
Chev.
Intern.
Chev.
Jaguar
(2nd tank
Audi
Cad.
Rambler
Dodge
Lincoln
Towmotor
Ford
Ford
Ford
Chev.
Plym.
Dodge
CMC
Chev.
Chev.
Plym.
Chev.
Ford
Chrys.
Chevy
Ford
Dodge
Ford
Ford
Ford
Dodge
Chev.
Chev.
Chev
Ford
Ford
01
T3
i
Dump
Van
Sedan
Mav.
Sedan
P.U.
School Bus
P.U.
Sedan
TOOLS
El Dor.
S.W.
G.T.
Cont.
Forklift
Truck
Truck
Truck
Van
Valiant
Four by
5500
P.U.
LT
Rd. Runner
Vega
Pinto
N.Y.
Caprice
Sedan
Mav.
Truck
Truck
P.U.
Nova
S.W.
II
C-700
P.U.
i.
n>
0)
>-
'73
'73
'74
'69
'61
'66
'70
'65
'73
'68
'66
'68
'74
'71
'72
'72
'69
'63
'50
-
'72
'73
'69
'73
'71
'73
'69
'70
'72
'74
'71
'65
'74
'74
'67
'64
'73
'73
4-10
-------
Table 7 Vehicle Fill Data (Continued)
Intermark (continued)
Sears, El Cajon
Process Products
Union, Waring Rd.
Clean Air Engineering
Frank Danna Service
Shell
Shell - Balboa Ave.
Balanced
Standard-Lake Murray Rd.
Gulf - Univ. & Boundary
0>
+J
to
o
7/29
7/29
7/29
7/29
7/29
7/29
8/2
8/2
7/30
7/30
7/30
7.30
7/31
7/31
7/31
7/31
7/31
8/15
7/26
7/26
7/25
7/25
7/25
7/25
7/25
8/15
8/15
8/15
8/15
at
• c i
0 M-^
r— -O _l
Q. IO Ul
X 0) —1
ui o:^
-(?)-
1.0
0.3
0
1.0
0
1.0
0
0.3
1.0
0
0
0
1.0
1.0
1.0
0
0
0
i.a
1.0
0
1.0
1.0
0
0.2
1.0
1.0
1.0
0
n^i
•r- ID ^
n-*^
01
-(3) ~
7.8
9.5
7.8
-
4.7
4.3
-
5.9
7.9
11.4
6.5
6.8
6.5
7.8
10
10.9
7.2
2.9
-
4.9
5.7
2.5
4.6
-
-
-
-
i. •*
° s
i— -Q
i— +-1
a. 5.
00 °°(4)
No
No
No
No
No
No
No
lOc.c.
5c.c.
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Yes
No
No
No
Interface 1
Gap |
No
No
No
No
No
Yes
No
No
Yes
No
No
No
No
Yes
Yes
No
No
No
No
No
No
No
No
No
No
Yes
Yes
Yes
No
•a -a
C i—
03 01
3: 3:
Yes
Yes
No
No
No
No
No
Yes
Yes
No
Yes
No
No
No
No
Yes
Yes
Yes
No
No
No
No
No
No
No
Yes
No
Yes
No
jstretched 1
1 Hose 1
NO
No
No
No
No
No
No
No
No
No
Yes
No
No
No
Yes
No
No
No
No
No
No
No
No
Yes
No
Yes
No
No
No
i-"^
<^
.(M-
2
3
5
1
-3
2
-
-7
-8
1
12
-2
0
5
4
12
-
-
-
7
5
2
-
-
-
-
0)
_*:
to
hord
Chrys.
Chev.
Datsun
Honda
Chev.
Pont.
Toyota
Chev.
Ford
Plym.
Chev.
Chev.
Merc.
Chev.
Plym.
Olds.
Dodge
Ford
Olds.
Lincoln
Chev.
Ford
Chev.
Dodge
Chev.
Ford
V.W.
Opal
"oj
•a
i
H.U.
Sedan
Vega
P.U.
Civic
Impala
Tempest
Corolla
Chevelle
LTD
Valiant
II
Vega
Mont.
S.W.
Fury
S.W.
Dart
S.W.
Toronado
Cont.
Chevelle
P.U.
Nova
Dart
P.U.
Sedan
S.W.
S.W.
i.
-
'M
'71
'74
'73
'73
'69
\
'68
'73
'68
'68
'63
'68
'73
'73
'67
'68
'60
'64
'73
'73
'72
.73
'63
'64
'65
'74
'67
'69
(2)
(3)
(4)
(5)
Explosimeter checks were done during vehicle fueling by moving
the explosimeter probe around the vehicle-nozzle interface at
a distance from the interface of approximately 1/4" to 1/2"
throughout the fueling period.
Fill rate was determined by timing fueling with a stop watch
and recording the number of gallons dispensed from the dis-
penser register.
Spitback amounts are visual estimates
'AT = vehicle gasoline temperature - underground tank gasoline
temperature. Temperatures were measured with a Yellow Springs
Instrument model 43 - Tele-Thermometer. . n
-------
Automobile fuel ings were observed at the two balanced system stations in
the test group to determine how often the attendants' hand held the nozzle.
The results of these observations are shown in Table 9 below. Most of the
hand held fuel ings appeared to be due to the customer asking for a specific
small amount of gasoline rather than to assure a tight nozzle fit.
Table 9 Percentage of Hand-Held Fuel ings at Balanced
System Stations
STATION
Standard - Lake Murray
Road
Gulf - Boundary &
University
# FUELINGS OBSERVED
16
26
# HAND HELD
2
13
% HAND HELD
12.5
50
Table 10 shows the average fill rates observed at the different systems
during the test period. The balanced systems had considerably lower fill
rates than the vacuum assisted systems
Table 10
Average FiIT Rates Observed (Gal/Mini
Environics
Intermark
Clean Air Engineering
Shell
Balanced
Overall
7.4
6.8
7.9
8.2
4.1
7.0
4-12
-------
5.0 ANALYSIS OF PROCESS UNIT VENTING
AT UNION STATION, WARING RD., S.D.
From July 19th through August 16th a dry gas meter was connected to
the carbon canister vent of the Process Products unit at the Union Station
on Waring Rd. The purpose of this test was to monitor the amount of proc-
essed air vented by the unit to assure that all excess air being taken into
the unit was being processed. Table 11 shows the volumes of air vented,
gallons of gasoline dispensed, and gallons of gasoline delivered to the sta-
tion from the Union bulk terminal.
In June 1974 a TRW test team tested the Process Products unit and found
that the system drew in 1.3 cubic feet of air for every cubic foot of gaso-
line dispensed. The test team was unable to account for much of this excess
air and so hypothesized a leak somewhere in the system.
The calculations below compare the amount of air/vapor mixture which can
be assumed to have been taken into the vapor recovery unit over the period
7/22/74 to 8/6/74, with the amount of processed air which was vented by the
unit over the same period. The calculations show that a significant amount
of the excess vapors collected by the unit cannot be accounted for and must
be assumed to have leaked out without being processed. This conclusion is
supported by the fact that a leak was discovered in the underground vapor re-
turn lines on 8/8/74 and that vapor leakage was still being detected at the
end of the test period.
. gallons of gasoline dispensed from 7/22 to 8/6 = 19,716 gallons
19,716 gallons x °-1337 cubic feet = 2^36 cu.ftt Of gasoline dispensed
gaiion
f.3
. 2,636 cu.ft. of gasoline dispensed x .3™ excess vaP°rs
ft dispensed
o
= 790.8 ft excess vapors taken into system (air + gasoline vapor)
2
. 790.8 ft excess air/vapor mixture x 0.6 (assuming 60% air)
3
= 474.5 ft excess air
. cubic feet of air vented through carbon canister vent over period
7/22 to 8/6 = 8.74 ft3
5-1
-------
474.5 ft3 of excess air - 8.74 ft3 air vented = 465.8 ft3 excess
air unaccounted for
o
465.8 ft of excess air = (0.6) (total air/vapor mixture)
(assuming 60% air)
total air/vapor mixture unaccounted for = -Q- >• = 776.3 ft
3
776 ft of air/vapor would represent approximately 29% of the un-
controlled (baseline) emissions
5-2
-------
Table 11
Process Unit Venting
UNION STATION - 5194 Waring Rd., S.D.
DATE
July 18
19
20
21
22
23
24
25
26
27
28
29
30
31
August
1
2]
3
4
5
6
7
8
9
10
11
12
u1
15
Total
GALLONS
DISPENSED
1276.1
1296.0
1662.1
3865.9
1203.9
1183.6
685.3
1062.8
6218.8
1261.8
1292.9
4072.7
729.3
2206.3
28017.5
BULK DROP
(GALLONS)
8630
8600
8800
8800
34830
OUTBREAKING
(ft3)
20.41
0
9.34
0
0
0
0
0
27.04 (])
175.25 (2)
43.00 (3)
96.77
371.81
INBREATHING
(ft3)
0.588
0.600
0
0.31
0
0
0
0
0
2.38
6.48
10.36
Lj Unit force vented by Process Products technician
v ' TRW gas meter was removed by Union Oil Co. test
I A leak in vapor
return lines was
repaired on 8/8/74
team for in-house
Dry gas meter was replaced by
gas meter was removed by
testing of Process Products unit.
their own dry gas meter.
Volume readings taken from Union Oil Co. test teams dry gas meter from
from this point on. 5_ ~
-------
6.0 VENTING OF UNDERGROUND TANK
AT STANDARD BALANCED SYSTEM
Two dry gas meters were placed on the vent to the low lead gasoline
tank of the Standard Station on Lake Murray Rd. from July 19th to August 16th.
The meters were connected to the underground tank through solenoid valves.
One dry gas meter was connected through a solenoid valve set to open at 0.15
inches of water of pressure while the second was connected to a solenoid valve
which opened at 0.15 inches of water of vacuum. This allowed separate readings
for inbreathing into the underground tank and outbreathing from it. Records
were kept of the total number of gallons of low lead gasoline dispensed each
day as well. This data and the number of gallons of gasoline delivered to the
station from the Standard Oil bulk terminal are shown in Table 12.
The data in Table 12 shows that there was 2.65 times more inbreathing
than outbreathing during the study period. Most outbreathing appears to have
occurred in connection with bulk deliveries of gasoline to the station. Vapors
displaced from the underground tank while it was being filled were released
through the vent rather than returning to the tank truck. 143.9 cubic feet of
vapor were vented from the underground tank during the test period. This re-
presents 2.6% of uncontrolled emissions (includes vapor displaced during
vehicle fuel ings and during bulk deliveries).
Inbreathing was more evenly distributed throughout the test period than
was outbreathing. Inbreathing was observed during vehicle fuel ings on several
occasions. There are several possible explanations for this including "vapor
contraction," liquid blockage, or restrictions in vapor return hoses.
6-1
-------
Table 12
Underground Tank Venting
STANDARD STATION - Lake Murray Rd., S.D.
DATE
July 16
17
' 18
19
20
21
22
23
24
25
26
27
28
29
30
31
August
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Total
GALLONS
DISPENSED
634
805
425
937
i
1861
t
702 '
782
439
725
t
2939
t
562
629
29 2
549
501
950
847
16,779
(2,243 ft3
BULK DROP
(1)
4600
4000
4600
4000
4600
3500
25,300
(3,379 ft3)
OUTBREAKING
(ft3)
t
7.81
1
3.45
2.11
0.42
0.02
4
44.00
¥
0.44
10.17
3.84
0.33
I
18.11
t
0.45
1.53
t
36! 91
t
1.32
0.41
0.11
12.48
143.91
INBREATHING
(ft3)
t
63.56
.1
12.00
15.22
8.26
21.44
1
29.14
t
22.24
32.95
11.73
13.30
1
0.94
t
13.24
59?06
t
12.60
11.68
23.32
29.44
. 380.12
All figures are for Low Lead Gasoline
(1) Data from bulk facility
6-2
-------
7.0 POTENTIAL HAZARDS
The greatest hazard connected with vapor recovery systems appears to be
that of fire. Several instances of this type of hazard were noted during the
study period. Such hazardous situations were ususlly the result of careless
maintenance or installation rather than inherent faults in the apparatus.
An example of careless installation is that of the Environics Vapox 400
unit installed on the roof of the Mini Super Market in Lakeside. The unit
was placed on a wooden platform, the outlet of the exhaust from the reactor
being directed at the wooden platform and main electrical conduit. The result
of this placement of the exhaust outlet was a charring of both the wooden
platform and the electrical conduit, which could easily have resulted in a fire.
Two potential hazards noted during the study period resulted from care-
less maintenance. These two situations occurred due to an incorrect adjust-
ment of the reactor temperature in Environics units. The first case occurred
at the Union Station on 6th and Robinson Streets in San Diego and resulted in
what the attendants on duty described as "backfiring" noises coming from the
process unit. The entire processing unit has subsequently been replaced by
Environics.
A similar incident occurred to the Vapox 400 unit at Frank Motors in
National City shortly after the unit was installed. Apparently the reactor
overheated and burned out the wiring in the process unit. The heat was suffi-
cient to char the paint on the outside of the unit.
On one day during the test period a fire did occur at one of the gas
stations. This occurred at the Union Station on Waring Road in San Diego on
July 27, when an attendant was burning cobwebs with a lighter close to a curb
next to the Process Products vapor recovery device. A fire sprang up along a
crack which ran along the base of the curb. The attendant stamped the flames
out only to have them spring up again. After putting out the flames, a second
and final time the manager of the station called an engineer of the San Diego
County APCD and asked him to investigate the situation. The engineer attempted
7-1
-------
to seal the crack with cement. The station manager's son then checked for
further leakage by lighting his lighter next to the crack. Since no vapors
were ignited by this test, it was assumed that the leakage problem had been
temporarily solved.
Explosimeter checks along the base of the curb on August 1 and 2 showed
the vapors to be in the explosive range all along the curb base. On August 2
technicians from Union Oil Company drilled core holes at four points along
the base of the curb and took samples of the vapors coming from underground.
Laboratory analysis confirmed them to be petroleum vapors.
On August 7 the underground vapor lines were excavated and a leak was
found where the underground 2-1/2" fiberglass lines connect to the 2" galvanized
lines going into the process unit. Subsequent tests with an explosimeter have
revealed that as recently as August 16 vapors in the explosive range were still
issuing from the crack at the base of the curb.
On August 14, with a negative pressure of 1/2 inch of water on the vapor
recovery system (the blower which normally provides a vacuum at the nozzles
was down) explosimeter readings along the crack at the base of the curb were
zero. On August 15 with a positive pressure of 3 inches of water on the vapor
recovery system explosimeter checks showed the vapors issuing from the cracks
to be above the lower explosive limit. The pressurization of the vapor system,
which is an intrinsic part of the operation of the Process Products unit, seems
to have intensified the vapor leakage.
Another source of possible hazards is the lack of explosion-proof and water-
proof junction boxes in the wiring of some units. An example of this was a
junction box at the Vapox 1500 unit at Darr's Liquor. The junction box had no
lid, and was covered only by a 5" by 4" piece of 3/8" plywood layed on it to
keep the rain out. The fact that the unit (like several others visited) sits on
a roughly constructed wooden platform which was beginning to sag in spots adds
to the possible fire hazards.
7-2
------- |