United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-450/3-79-018
April 1979
Air
Evaluation of Vapor Leaks
and Development of
Monitoring Procedures
for Gasoline Tank Trucks
and Vapor Piping
-------
EPA-450/3-79-018
Evaluation of Vapor Leaks
and Development of Monitoring
Procedures for Gasoline Tank
Trucks and Vapor Piping
by
Robert L. Norton
Pacific Environmental Services, Inc.
1930 14th Street
Santa Monica, California 90404
Contract No. 68-02-2606
Task No. 11
EPA Project Officers: Nancy Mclaughlin and Stephen A. Shedd
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
April 1979
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations - in limited quantities - from the
Library Services Office (MD-35), U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina 27711; or for a nominal fee,
from the National Technical Information Service, 5285 Port Royal Road,
Springfield,Virginia 22161.
This report was furnished to the Environmental Protection Agency by
Pacific Environmental Services, Inc. , 1930 14th Street, Santa Monica,
California 90404, in fulfillment of Contract No. 68-02-2606, Task No. 11.
The contents of this report are reproduced herein as received from
Pacific Environmental Services, Inc. The opinions, findings, and conclusions
expressed are those of the author and not necessarily those of the Environmental
Protection Agency. Mention of company or product names is not to be
considered as an endorsement by the Environmental Protection Agency.
Publication No. EPA-450/3-79-018
11
-------
TABLE OF CONTENTS
Section Page
1.0 INTRODUCTION 1-1
2.0 HYDROCARBON VAPOR LEAKAGE. 2-1
2.1 Sources of Hydrocarbon Leakage 2-1
2.1.1 Tank Truck 2-1
2.1.2 Bulk Plants and Terminals 2-4
2.1.3 Service Stations 2-6
2.2 Potential and Controlled Emissions ...... 2-7
2.3 Available Equipment to Control Emissions ......... 2-7
2.3.1 Tank Truck Dome Covers and P-V Vents ..,.., 2-7
2.3.2 Tank Truck Vapor Collection Piping and
Internal Vents . 2-11
2.3.3 Vapor Transfer Piping 2-13
2.3.4 Vapor Transfer Couplers 2-13
2.3.5 Storage Tank Pressure-Vacuum Relief Vents . 2-14
2.4 Operating and Maintenance Procedures 2-15
2.4.1 Dome Covers and P-V Vents 2-15
2.4.2 Vapor Collection Piping and Internal Vents 2-18
2.4.3 Vapor Transfer Piping 2-18
2.4.4 Vapor Transfer Couplers 2-19
2.4.5 Storage Tank Pressure-Vacuum Relief Vents . 2-20
2.4.6 Miscellaneous Emission Sources 2-20
2.5 Costs to Maintain Vapor Containing Equipment 2-21
2.5.1 Tank Trucks 2-21
2.5.2 Other Emission Sources 2-23
3.0 DEVELOPMENT OF MONITORING PROCEDURE 3-1
3.1 Test Methods 3-1
3.1.1 Vapor To Liquid Volume Determination (V/L) 3-1
3.1.2 Explosimeter 3-2
3.1.3 Sonic Detector 3-2
3.1.4 San Diego "Bag" Test 3-2
3.1.5 Pressure-Vacuum Test (CARB) 3-3
3.1.6 Bubble Indication Method 3-4
3.1.7 Quick Leak Decay 3-4
3.1.8 Volume Leakage 3-5
111
-------
TABLE OF CONTENTS (Concluded)
Section Page
3.2 Evaluation of Test Procedures 3-5
3.2.1 Pressure-Vacuum Test (CARB) 3-10
3.2.2 Quick Leak Decay Method 3-16
3.2.3 Volume Leakage 3-20
3.2.4. V/L Ratio Method 3-24
3.2.5 Explosimeter Method 3-33
3.2.6 Sonic Detector 3-40
3.2.7 Bubble Indication Method 3-41
3.2.8 San Diego "Bag" Method 3-41
3.3 Pass/Fail Criteria 3-43
4.0 CONCLUSIONS 4-1
4.1 Vapor Containing Equipment and Maintenance 4-1
4.2 Costs of Maintaining Vapor Tight Conditions 4-1
4.3 Monitoring Procedures 4-2
4.4 Pass/Fail Criteria 4-3
APPENDIX A — Suggested Monthly Visual Maintenance
Inspection Checklist A-l
APPENDIX B — Actual Maintenance Performed on Delivery Tanks
During Field Test Phase B-l
APPENDIX C — Suggested Enforcement Inspection Checklist ... C-l
IV
-------
LIST OF ILLUSTRATIONS
Figure Page
2-1 Vapor Recovery System Using Overturn Rail ,... , . 2-3
2-2 C-B Dome Assembly 2-8
2-3 Tiona Dome Cover ., 2-9
2-4 Typical Vent Cover Configuration 2-12
3-1 Frequency of Occurrences of Back Pressure During
Loading Operations , , 3-11
3-2 Typical CARB Pressure Test Results at Bottom Loading
Terminal 3-14
3-3 Typical CARB Pressure Test Results at Top Loading
Terminal , 3-15
3-4 Laboratory Test Results for CARB Pressure Test 3-17
3-5 Quick Leak Decay Test Apparatus 3-19
3-6 Typical Pressure Versus Time Curves for Laboratory
Tests of Quick Leak Decay Method 3-21
3-7 Volume Leakage Versus Pressure Decay for Bottom Loaded
Tanks at Various Pressures 3-22
3-8 Volume Leakage Versus Pressure Decay Rate for Top
Loading Tanks at Various Pressures 3-23
3-9 Frequency Distribution of V/L Ratio for Tanks That Pass
Certification Tests and Tanks That Fail Certification
Tests 3-34
-------
LIST OF TABLES
Table Page
2-1 Costs for Maintaining Trucks in Leak Tight Conditions . 2-22
2-2 Total Annual Maintenance Costs for Product Delivery
Equipment 2-24
3-1 Test Fleet Physical Data 3-7
3-2 Tank Tightness History 3-13
3-3 Correlation Coefficients for Volume Leakage Results
With Respect to CARB Test Results 3-25
3-4 V/L Results for Top Loading 3-26
3-5 V/L Results for Bottom Loading 3-29
3-6 Correlation Factors (r) for V/L Ratios and Tank
Loading Backpressure 3-35
3-7 Percent of Compartments With Leaks at Specific
Locations 3-37
3-8 Occurrence of Hydrocarbon Leakage at Various
Explosimeter Levels 3-38
3-9 Percent of Tanks Identified Correctly at Varying LEL
Levels Using the Explosimeter Method 3-39
3-10 Volume Leakage Rate Required for Various Containment
Requirements 3-45
VI
-------
1.0 INTRODUCTION
In the prevention of hydrocarbon emissions during the transfer
of gasoline, hydrocarbon vapor recovery systems have been installed.
These systems have been developed to contain hydrocarbon emissions
and to transfer the vapors through piping systems either to the
delivery tank during gasoline deliveries, or to the storage tanks or
to the vapor processor during the loading of the delivery vehicle.
The effectiveness of these vapor recovery systems is dependent upon
the absence of leaks in the vapor containing equipment and assorted
piping.
As a task under EPA Contract No. 68-02-2606, Pacific Environmental
Services, Inc. (PES) conducted a study to define the leakage areas,
the equipment necessary to contain the vapors, and the costs of
maintaining the vapor containing equipment in a leak tight condition.
The study also aimed at developing a monitoring procedure which would
be a quick, low cost technique. This procedure would be used as an
enforcement tool to determine if the vapor transfer system was
operating without leaks as defined by some pass/fail criteria.
To determine the availability and the cost of maintaining the
necessary vapor containing equipment, numerous equipment manufac-
turers and equipment operators were contacted. Identification of
leakage areas were performed by observing delivery tank loadings.
Maintenance costs were obtained by observing actual tank maintenance
procedures conducted by several tank truck operators.
To determine an appropriate field monitoring procedure, a field
test program of candidate methods was developed by PES. The field
test program, conducted by another .contractor, was supervised by
both PES and EPA personnel. Tests were conducted in the Los
Angeles, California area because truck fleet operators in California
are required to maintain their trucks in a leak tight condition as
1-1
-------
defined by the California Air Resources Board (GARB) certification
criteria. Maintenance and equipment specifications were also
obtained mostly from California sources because of the leak tight
requirements. Tests were conducted at both bottom and top loading
terminals with a total of over 150 tank loadings monitored.
1-2
-------
2.0 HYDROCARBON VAPOR LEAKAGE
2.1 SOURCES OF HYDROCARBON LEAKAGE
2.1.1 TANK TRUCK
Sources of hydrocarbon leakage from truck delivery tanks
include dome covers, pressure-vacuum vents, and vapor collection
piping and vents. Smaller instances of leakage occur at tank
welds, liquid and vapor transfer hoses, overfill sensors, and
vapor couplers.
2.1.1.1 Dome Covers and P-V Vents
Dome covers consist of a series of openings, clamps and
seals each of which is a potential hydrocarbon vapor leakage
point. The first potential source is the seal where the dome
assembly itself attaches to the truck tank. A gasket material is
placed between the dome base ring and the tank welding ring, and
the dome cover clamped to the tanks. Hydrocarbon leakage can
occur at this seal if dirt or foreign material becomes lodged in
the interface, if the gasket material becomes cracked or worn, or
if the dome base-ring becomes warped or damaged.
Another source of hydrocarbon leakage from the dome cover is
at the seal between the dome lid which covers the hatch opening.
This seal can be easily damaged if foreign material lodges in the
interface, especially if the lid is opened or closed regularly as
in top loading. The dome lid is also spring loaded and acts as a
secondary pressure relief vent normally set to open if the tank
pressure reaches 3 psi. The hatch cover can become warped or
damaged and leakage can occur.
Fugitive hydrocarbon emissions can also occur at the pressure-
vacuum (P-V) vents which are normally installed in the dome lid.
The pressure side of the vent is commonly set to be full open at
1 psi and the vacuum side of the vent is commonly full open at
2-1
-------
6 ounces vacuum. These vents are installed primarily for safety
purposes to allow venting or inbreathing into the tank to offset
pressure changes caused by temperature variations. These vents
also act as a vapor containing device to reduce emissions of
hydrocarbons from the compartment vapor space. Emissions or leaks
may occur if the P-V vent is not installed or is not maintained
properly. The valve seat may become dirty or damaged which may
cause the valve to seal improperly. The valve actuating device,
such as a springloaded valve, may become damaged also allowing
improper sealing and causing hydrocarbon leakage.
2.1.1.2 Vapor Collection Piping and Internal Vents
For those truck delivery tanks that have vapor recovery
installed, hydrocarbons can leak from the vapor collection and
piping systems. Normally, each compartment has a vent valve
which is opened when that compartment is being loaded or unloaded.
This vent allows vapors to be removed from or returned to the
compartment through piping into the vapor recovery system. The
compartment vent valve is covered either with a rubber boot
assembly or metal bolted or welded cover to contain the vapors in
the vapor transfer system. The vapor return line can be either
rubber hoses or metal pipe placed on top of the tank or incorpor-
ated into the overturn rail or any combination of these.
Figure 2-1 illustrates one of these configurations. The vapor
return line, which is manifolded to each compartment, will have
joints or connectors in the piping for each compartment.
Hydrocarbon vapors can leak from the vent valve cover due to
tears in the rubber boot, leaks in gaskets from bolted covers or
faulty welds from welded covers. Leaks can occur in the vapor
line connectors from poor seals or-clamping mechanisms with the
rubber hoses or faulty welds or seals with metal piping. Hydrocarbon
emissions may also be detected at the vapor return coupler. This
would be caused by vapors leaking out through the vapor transfer lines
due to improperly sealing internal valves.
2-2
-------
I
U>
Figure 2-1, Vapor Recovery System Using Overturn Rail
(Courtesy of The Heil Company)
-------
2.1.1.3 Liquid and Vapor Transfer Hoses
Leaks can occur from liquid and vapor transfer hoses and from
their respective couplers. Hoses can become torn, worn, or cracked
to produce hydrocarbon vapor leaks. Fugitive hydrocarbons can
occur from vapor coupler connections if these are not coupled
or closed properly. Coupler gasket material can also be worn or
damaged causing a poor seal. If dry break or vapor tight couplers
are used, the valve seat may become worn or foreign matter may
become lodged in the seal causing hydrocarbon vapors to leak to
the atmosphere.
2.1.1.4 Miscellaneous Emission Sources
Other sources of leakage from truck delivery tanks are
possible but occur considerably less frequently than those already
discussed. Leakage can occur from flaws in the tank shell,
improperly welded seams, or improperly installed or loosened over-
fill protection sensors.
2.1.2 BULK PLANTS AND TERMINALS.
Various leakage of hydrocarbon vapors from bulk plants and
terminals can occur from vapor couplers and hoses corresponding
to the vapor recovery system, top loading connectors (if applicable),
vapor piping to storage tanks and pressure relief vents on fixed
roof storage tanks or sump tanks (if applicable).
2.1.2.1 Vapor Piping to Storage Tanks and Processors
Vapor recovery piping can be installed at bulk plants for both
incoming loads to the storage tank and for vapor control at the
loading racks. Vapor recovery piping installed at terminals will
run from the vapor processors to the loading rack. This piping is
usually above ground and is normally flanged or threaded metal
2-4
-------
pipe. Hydrocarbon vapor leaks can occur at piping joints or
connections due to improper installation, faulty flange gaskets or
accidental damage.
2.1.2.2 Vapor Couplers and Hoses
As discussed in the delivery tank section, losses can occur
from damaged or worn transfer hoses or improperly connected or
damaged vapor transfer couplers. Bottom loading coupler losses can
occur from worn or contaminated vapor tight valve seats or from
worn gaskets.
2.1.2.3 Top Loading Vapor Connectors
Vapor leakage from top 1oading collection and loading arms can
occur from movable joints or swivels. The arms can be either
pneumatically operated or manually swiveled as in smaller bulk plant.
type top loading operations. Even with vapor recovery loading
arms, recent test data has shown that hydrocarbons can escape during
over 95 percent of the loading operations. Liquid spillage and
leaking joints, such as swivels and flange gaskets, account for a
number of hydrocarbon vapor sources. Hydrocarbons can also escape
from the loading-arm-hatch opening interface. Test data show that
this can be the most significant source of leakage from the top
2
loading operations.
2.1.2.4 Storage Tank Pressure Relief Vents
At either bulk plants or terminals where fixed roof tanks or sump
tanks are employed, pressure-vacuum vents are used to control vapor
losses from the storage tanks. These valves are similar in
concept to those discussed in Section 2.1.1.1. The valves can be
either spring loaded or weighted to open at the desired internal
pressure. Dirt or other debris can become lodged in the valve seat
causing it to seat poorly and become a hydrocarbon leak source.
2-5
-------
The spring or weights system may get out of allignment and not
allow the valve to return to its seat properly, thereby causing
leaks.
2.1.3 SERVICE STATIONS
Fugitive hydrocarbon emissions can occur during service
station gasoline deliveries at the delivery and vapor transfer
couplers, at the underground tank vent, and at the underground
vapor piping.
2.1.3.1 Vapor Piping
The vapor piping at service stations is almost exclusively
underground and therefore should not be a significant source of
hydrocarbons. However, if improper installation does occur vapors
can escape from the piping and reach the atmosphere.
2.1.3.2 Underground Tank Vent
Hydrocarbon vapors can be emitted from the underground tank
vent during unloading of the gasoline. This could be caused by
restrictions in the vapor return line, by not connecting the vapor
line during the delivery, or by temperature differences between
the gasoline being unloaded and that which is present in the
underground tank.
2.1.3.3 Vapor and Liquid Transfer Couplers
Leakage can occur due to damaged or improperly attached vapor
couplers as discussed in previous sections. Damaged couplers may
not allow the sealing mechanisms to operate properly and, if not
coupled tightly, hydrocarbon vapors' can be emitted.
2-6
-------
2.2 POTENTIAL AND CONTROLLED EMISSIONS
Quantifying a leakage source is difficult since the size of
the leak and the corresponding leakage rate can vary significantly.
The leakage rate is proportionate to the equivalent orifice size
4
of the leak opening and the tank pressure. All of the leakage
sources prescribed in Section 2.1 have the potential to be large
leaks; however, some of the leak sources are normally more
predominant than others.
Leakage of hydrocarbons from hatch covers, hatch base rings
and pressure vacuum vents are the sources where hydrocarbon vapors
most often occur. These uncontrolled leakage rates have the
potential to exceed 70 percent of the vapor transferred. Under
controlled conditions these sources should not leak in excess of 1
percent of the volume of vapor transferred. This is based on the
CARS tank truck pressure loss criteria which, when calculated, does
not allow the delivery tank to leak greater than 1 percent of the
volume of vapors transferred (99 percent containment).
The other sources discussed, such as the vapor piping,
couplers and storage tank vents also have the potential for large
leaks depending upon the size of the leakage area. Vapor losses
from properly installed and maintained piping and couplers should
be eliminated. Vapor losses from P-V vents, when the tank pressure
is below the venting level, can also be eliminated with properly
installed and maintained equipment.
2.3 AVAILABLE EQUIPMENT TO CONTROL EMISSIONS
2.3.1 TANK TRUCK DOME COVERS AND P-V VENTS
In California where leakage must be contained within the
limits specified in the certification criteria, only two firms
manufacture hatch covers which can meet these requirements. These
domes are supplied by C-B Equipment, Inc., Lynwood, California and
2-7
-------
Tiona-Betts, Inc., Warren, Pennsylvania, From interviewing tank
truck operators, conflicting opinions were obtained on a preference
of either of the hatch covers. The Tiona dome base ring is made
from a flat plate while the C-B dome base ring is cast and has
support ridges (see Figures 2-2 and 2-3). This added support makes
the C-B dome less apt to succumb to bends or warpage. However, the
rigidity of the C-B dome does not allow the type of maintenance
that can be performed on the Tiona dome (see Section 2.4). The
flat plate of the Tiona base ring can be worked and manipulated to
retain vapor tightness, whereas the C-B base ring, if damaged,
must be replaced. This can be significant when considering the
cost of the dome covers (Tiona approximately $80; C-B dome
approximately $120). The dome lids themselves are somewhat similar
(although the C-B cover has reinforced ribs) and both use a spring
loaded closure mechanism for containing vapors.
The pressure-vacuum vents are built into the dome lids in
both cases. The Tiona dome incorporates a separate piston-type
valve for both the pressure and vacuum release vents. These
valves are both spring loaded pistons which will open when the
actuation pressure is reached. The C-B dome uses a valve which can
seat in either the pressure relief direction or the vacuum release
direction. The C-B vent does not use pistons but uses spring
loaded discs. One spring loaded disc constitutes the pressure vent
and another spring loaded disc constitutes the vacuum vent. These
discs will then move as the pressure reaches the critical point.
This vent system is easier to repair and clean than the Tiona
system and is less susceptible to leakage caused by debris because
there are no pistons. The piston system has more tendency to leak
due to tight or sticky piston movement caused by dirt or other
foreign material becoming lodged in the piston sleeves. A ball
is inserted in both the C-B and Tiona vents to act as a shutoff
valve in case the tank is rolled over. This prevents the liquid
from escaping from its container. One operator has devised a
conversion kit so that the C-B- vent can be installed on the Tiona
dome.
2-8
-------
Figure 2-2. C-B Dome Assembly
(Courtesy C-B Equipment, Lynwood, California)
IO01
7444
7417
14St
•MH
1440
14*1
T44I
7*4*
1444
1441
1444
144t
1444.
1410
7411
74»t
741*
141S
1414
14H
mi
lit
MW Mtitmc,
CASKVT
«OMV Ck»
vem. LOC^KM »»»
SUAfT Atit>>>*k-r
LtCLSf HIM IPR1W4
*\.I.LVt.
LOCKIWC *4( C4TCU
DOMi ITVO
DOWh CU4MV kU«
LOCK I
IATCM
SOMi CA» S4MTVT
VACUUM VAkVt 61 DC
VACUUM *•«.*» mi«C
tMtimvai.** o«»c
rtUfUtt MUVt MU.L
Mtb «VT
O- (IMC
OOMb CLAHf- WUC
2-9
k-A
-------
SECONDARY SAFETY LATCH (ft)
4 IN.
STD. WELD
RING
Na
1
2
3
4
5
6
7
8
Description
Fill Cover
Fill Gasket
Bolt - Spring
Spring
Bolt - Clamp
Nut - Clamp
Washer • Clamp
Clamping Ring
Material
Aluminum
Buna-N
Viton
Steel Cd. Pt. j
Van. Steel
Steel Cd. Pt.
Stls 304
Brass
Stls 304
Steel
Stls 304
Part No.
8046 -AL
31 19-BN
3II9-VT
3139-CP
3129-CP
3029-CP
3029 -SL
3030-BR
3030 -SL
3031 -MS
3031 -SL
Steel , 3036-MS
Steel Cd. Pt.
Stls 304
3036-CP
3036-SL
No.
9
10
II
Description
Manhole Gasket
Channel Type
Manhole Collar
(Channel Type)
2
X
i/)
Closure Assem.
(Includes Cover
Plate, Strong
Back, Latch,
Hinge Lugs &
Pins)
C
B
A
E
F
H
J
K
S
B
G
C
D
Material
Cork-Buna-N
Buna-N
Asbestos
Teflon -Asb.
Steel
'/« Aluminum
Stls 304
Stls 316
Steel
Steel Cd. Pt.
Stls 304'
Stls 304
Stls 316
Part No.
3175-CB
3I7S-BN
3175-AS
3 175 -AT
3 176 -MS
3176-AE
3176 -SL
3176-SS
6083 -MS
6083-CP
6083 -SP
6083-SL
6083 -SS
'Latch and Strongback Steel Cadmium Plated.
Figure 2-3. Tiona Dome Cover
(Courtesy of Tiona-Betts, Warren, Pennsylvania)
2-10
-------
Emissions from poorly installed or maintained dome covers or
pressure-vacuum vents can be sizable. However, if proper domes
are installed and maintained correctly, emissions can be reduced
significantly. The CARB certification requirements, which require
leak-tight trucks, illustrates that emissions from domes and
pressure-vacuum vents can be controlled and trucks maintained to
reduce hydrocarbon leakage.
2.3.2 TANK TRUCK VAPOR COLLECTION PIPING AND INTERNAL VENTS
Leakage can occur around the internal vent covers and vapor
piping joints. The internal vent allows vapors to enter the vapor
return system when loading or unloading liquid into the compartment
and is pneumatically or mechanically coupled with the compartment
loading. The vent opens into a covered area which in turn is
plumbed into the vapor return piping system. These vent covers
are made of either metal or rubber and are either welded, bolted,
or clamped into position over the vent valve (see Figure 2-4).
The vapor collection or return piping can also be made of metal or
rubber and can take several configurations. Separate piping may be
used for the vapor return or use may be made of the overturn rail.
If separate piping is used, the piping could be of rubber or metal
pipe, manifolding the exhausts from each compartment into the main
exhaust line. If metal piping is used, joints could be welded or
flanged with gasket material. If the vapor line is rubber, band
clamps are used most often at the joints to maintain tightness. If
the overturn rail is used as the vapor return line, piping is run
from the vent valve to the overturn rail and can again be welded
pipe or rubber hose. Joints could also be welded, flanged, or
clamped. Welded pipe and vent valve covers provide a better vapor
tight transfer system than the rubber boot or rubber hose transfer
2-11
-------
-\_>T~-
J
y 3
1
>
i'-~.
\
t
}
i.
{
/
1
H
,'~\
/
t
V
^^_ 1
•at
"f~\
r
h";— g~
- ' ,¥
-' is
t-
K«k4.>U.
fk^l »>(X.
b uta« _<«..
-\
Figure 2-4. Typical Vent Cover Configuration
-------
systems. Leaks are less frequent when using welded piping and
covers and replacement equipment losses are greatly reduced.
However, installation costs for the welded system would be higher
than that of the rubber system.
With properly installed welded vent covers and piping,
emissions should be eliminated from these areas with the exception
of damage caused by accidents. Proper installation, and a proper
maintenance program should reduce leakage from flanged and rubber
clamped vapor hoses significantly. These last two methods do
require a conscientious maintenance plan to maintain the system in
proper working order.
2.3.3 VAPOR TRANSFER PIPING
Vapor transfer piping includes flexible vapor hoses for the
tank truck loading rack, and all vapor transfer piping at terminals,
bulk plants, and service stations. Flexible vapor transfer hoses
are made from gasoline resistant rubber and normally are attached
to coupling fittings using band clamps. Vapor transfer piping at
terminals and bulk plants are normally rigid metal pipe and can be
found above or below ground in either welded, threaded, or flanged
pipe. Service station piping is almost exclusively underground.
Leakage can occur from poorly installed piping, leaking gaskets,
loosely applied clamps at the couplers, etc. However, welded or
threaded vapor piping, if correctly installed should eliminate
hydrocarbon vapor losses. Flanged pipe must be maintained and
monitored more frequently because of the gaskets involved at the
joints. However, flanged pipe, can all but eliminate hydrocarbon.
leakage if maintained properly,
2.3.4 VAPOR TRANSFER COUPLERS
Vapor transfer couplers for bulk plant, terminal, and service
station vapor transfers can take on numerous configurations. The
2-13
-------
couplers become leak tight through a compression mechanism incor-
porating a gasket material chemically resistant to the liquid being
transferred. These couplers could include: dry-break couplers or
vapor tight spring-loaded couplers for vapor lines at bulk plants,
terminals, and service stations; kam-lock type couplers for tank
truck vapor connections; coaxial fittings at service station drops.
The couplers used are dependent upon the vapor transfer configura-
tion selected. These couplers are readily available from several
manufacturers and are similar to liquid transfer couplers which
have been used for years. The leakage problems from couplers occur
when the gasket becomes worn or damaged or the coupler connectors
or body become damaged, not allowing an adequate seal.
If the vapor transfer couplers are in good working order and
coupled properly, emissions from the coupler joints should be
minimized. Small leaks may be encountered through the vapor hoses
or vapor to coupler joints, but these will be very small if the
system is maintained properly.
2.3.5 STORAGE TANK PRESSURE-VACUUM RELIEF VENTS
For fixed roof storage tanks, pressure-vacuum vents are
installed to relieve positive or negative pressures which exceed
their set point. These P-V vents are similar in approach to those
discussed in the tank compartment dome cover. The valves can have
either separate vent locations or may be designed to have a single
vapor outlet. The valves are held closed by either a series of
weights or force supplied by a spring. The weights or spring force
are designed to be offset by the internal tank pressure and will
be full open at the desired pressure setting. Pressure settings on
storage tanks are normally 6 ounce pressure and one-half ounce
vacuum.
Hydrocarbon vapor emissions cannot be eliminated from these
vents since their purpose is to release pressure. However, if the
2-14
-------
valve is maintained properly, emissions from the valve when the
tank is below the actuation pressure should be minimized.
2.4 OPERATING AND MAINTENANCE PROCEDURES
The equipment manufacturers contacted did not give much
information about recommended maintenance beyond the fact that
damaged parts should be replaced. However, several of the
operators interviewed had devised their own operating or maintenance
procedures which are discussed in this section, along with
recommended maintenance procedures for maintenance not currently
conducted. A visual inspection of the vapor containing equipment
is an integral part of the maintenance program. A suggested
checklist for a periodic visual inspection of equipment is shown in
Appendix A. Actual maintenance performed on truck tanks during the
field test phase are outlined in Appendix B.
2.4.1 DOME COVERS AND P-V VENTS
Dome cover maintenance procedures range from visual
observation to manual adjustments. Maintenance practices vary
greatly between operators from infrequent to monthly inspections.
For purposes of this report, California maintenance procedures will
be discussed because of the tank tightness requirements and the
corresponding maintenance necessary to obtain the required tightness.
Common causes for leakage around the domes, as discussed in Section
2.1, can be damaged or warped dome cover-base rings or dome lids,
dirty gaskets, or faulty pressure-vacuum vents. Before pressurizing
the tank, the dome lids should be visually inspected. The gasket
between the dome lid and base ring should be inspected for damage
such as tears or cracks. Dirt, or other foreign material, should also
be removed from the gasket sealing'Surface. If the gasket show signs
of excessive wear or damage, it should be replaced. The dome lid
2-15
-------
itself may be damaged or warped and should be checked for the quality
of the seal between the lid and base ring. Several methods can be
used, two of which are discussed here. The gasket or seal of the
dome lid could be coated with a type of grease or other easily visible
material and the dome lid closed, sealed, and then reopened. The
inability of the dome lid to close or seal around the entire cir-
cumference can then be clearly visible by showing gaps in the
indicating material on the mating surface.
The other method suggested by a tank truck operator would be
to use a piece of thin paper placed between the dome lid and the
base ring with the dome Tide closed securely. If the paper can
then be moved, the seal is not tight enough and a leak will most
likely occur.
The P-V vents should be visually inspected to determine if
foreign material is lodged in the valve seats not allowing the
valve to seal properly. The vent should also be tested to
determine if the spring loaded valve closures are working smoothly
without sticking or rubbing. The bolts and/or clamps used to
attach the base ring to the tank should also be tested for
tightness. If any of these visual techniques indicate the
necessity of repair, the maintenance should be performed before
proceeding.
The tank should then be pressurized to determine the ability
of the tank to maintain pressure. A bubble solution or sonic
detector could be used to indicate the presence of leakage points.
When these leaks have been found, the maintenance necessary to
reduce these leaks to acceptable limits should be performed.
The maintenance required to minimize leakage points identified
around the dome cover would include bending or reshaping the dome
base ring if possible or replacement of the entire dome cover. If
2-16
-------
a leak occurs between the dome lid and base ring, bending or
reshaping of the dome base can be done by pounding with hammers
or applying leverage to the hatch opening in an attempt to create
a good seal at the dome lid. The dome lid itself can be adjusted
by tightening the hold down mechanism. However, since the dome lid
is a secondary pressure relief vent (normally at 3 psi) there is a
limit on this adjustment. If the leak cannot be repaired
satisfactorily, the dome cover must be replaced. For leakage
around the tank/base ring interface, the bolts or attachment clamps
should be tightened. If the leak persists the dome cover should be
removed and the gasket inspected and replaced if necessary. The
dome cover may have to be reshaped or replaced entirely if
damaged to the extent that a good seal cannot be maintained.
Leakage at the P-V vents will require removing the vent from
the dome lid and disassembling and cleaning, the components:. The
valve seats should be cleaned and all foreign material removed to
ensure a good seal. If the components are damaged they should be
replaced. The springs holding the vents closed may need
replacing or stretching to return them to their designed holding
force. The piston housing, if applicable, should also be cleaned
to ensure the piston can move freely without rubbing or sticking.
The valve should be then reassembled and installed. If leakage
occurs which is still not acceptable, the vent valve should be
replaced.
Visual inspection of all dome covers should be performed on a
regularly scheduled basis and equipment which needs repair or
replacement should be fixed accordingly. This should be performed
when the truck is in the shop for normal maintenance. This should
not require pressurization of the tank but only replacement of
visibly damaged or faulty equipment. Some operators perform
visual inspections as often as once every two to three weeks.
2-17
-------
2.4.2 VAPOR COLLECTION PIPING AND INTERNAL VENTS
Vent valve covers and vapor piping joints should be visually
checked when observing the dome covers. During the visual
inspections, bolts in flanged covers should be checked for tightness
and rubber boots and hoses should be inspected for tears or cracks.
Bolts should be tightened and rubber equipment replaced as required.
These leakage points may be less obvious and pressurization of
the delivery tank may be necessary to locate the leak. Bubble
solution, sonic detectors or explosimeters can be used to pinpoint
the leakage sources. Leaks at the vent valve covers can occur
at welded joints, bolted covers, or from rubber covers. The
leaks found in welds should be marked and the weld repaired. If
a leak persists, the vent cover gasket should be inspected
and replaced if excessively cracked or damaged. Rubber covers
should be checked for tears or cracks and replaced as needed.
Vapor piping joints should be checked in a similar fashion.
Welded joints should be inspected for weld integrity and
repaired as needed. Flanged joints should have the bolts
tightened and the gasket material replaced as needed. All
bolted or clamped vapor piping joints should be checked for
tightness. Rubber vapor hoses should be checked for leaks and
replaced if worn or cracked. Gasket materials for flanged
piping should be replaced if leaks persist after tightening.
2.4.3 VAPOR TRANSFER PIPING
Leakage from vapor transfer piping can occur at piping
joints due to worn or deteriorated gasket material, improper instal-
lation, or loosened flange clamping mechanisms. The piping, where
above ground should be visibly inspected for damage or obvious
leakage areas. An explosimeter, sonic detector or bubble
indicating solution can be used to find smaller leaks. For
2-18
-------
welded pipe the leakage point should be marked and the point
repaired or plugged. For flanged pipe, all flange bolts should be
checked for tightness and adjusted. If leaks persist at these
flanged joints, the gasket material should be replaced. For
threaded pipe, if leaks are found at joints the fittings should be
disconnected and reassembled using some type of thread sealing
compound to ensure a tight fit. If the piping cannot be practically
dismantled, the leakage area should be marked and the leak
minimized.
Flexible vapor hoses should be checked visually for obvious
cracks and tears, and the hose-to-coupler clamp should be checked
for tightness. A bubble indicating solution or sonic detector can
be used to indicate the location of leaks. However, before
replacing this equipment, the hose should be tested with the entire
tank truck system since small leaks in the hose can occur and the
system may still pass the test (see Section 3.2). If during
inspection the hose has excessive wear or damage, it should be
replaced.
2.4.4 VAPOR TRANSFER COUPLERS
Vapor transfer couplers should be inspected periodically to
ensure their vapor tightness is maintained. The gasket material
should be visually inspected and replaced if worn, cracked, or
damaged excessively. Vapor tight couplers such as dry breaks or
spring loaded connectors should have the valve seat inspected and
cleaned to maintain a good tight vapor seal. Coupler clamping
mechanisms should be inspected and adjusted as necessary. The
coupler interface can be checked for leaks using an explosimeter
or bubble indication solution. If-leaks persist after maintenance
has been performed the coupler unit should be replaced.
2-19
-------
2.4.5 STORAGE TANK PRESSURE-VACUUM RELIEF VENTS
As in other pressure-vacuum relief vents, the most common
leakage point is around valve seats. Dirt or other foreign
debris can become lodged on the valve seat face not allowing the
valve to close completely and causing vapors to escape. Valves can
also have a problem of reseating improperly once they have opened.
The closing disc can get out of alignment resulting in the valve
face resting at an angle and not firmly on the valve seat. To
assure good valve closure, the valve seats must be periodically
checked to remove dirt and debris. The valves should also be
inspected to ensure they have reseated properly and that the valve
guides are clean and free of obstructions. Because of the poten-
tially large emission source from an open P-V valve, these vents
should be checked at least once per week. This should not prove to
be an excessive burden since many fixed roof tanks are gaged for
liquid level from the top whenever liquid deliveries are made.
2.4.6 MISCELLANEOUS EMISSION SOURCES
This category includes leaks in tank shells, poorly welded
seams, damage caused by an accident, and poorly installed or loose
overfill protection on tank trucks. These leaks are usually small and
therefore hard to detect. However once they are repaired, usually
when the tank is first pressure tested, the occurrence of these
leak sources decrease significantly. These sources are commonly
found using a bubble indication solution while the tank is
pressurized. The sources are marked and repaired as necessary.
Probability of a leak occuring at other sources is so much
higher that these sources are usually the last checked.
2-20
-------
2.5 COSTS TO MAINTAIN VAPOR CONTAINING EQUIPMENT
2.5.1 TANK TRUCKS
Costs are presented for maintaining a tank in leak tight
condition. These costs include labor and materials for performing
the required maintenance to reduce leakage to an acceptable level
and labor for performing the CARB certification test.
Costs are given for two degrees of vapor tightness. The
first, and more stringent case, comes from San Diego County which
allowed a pressure drop of 1 inch of water in 5 minutes. The second
case was that required in the remainder of the State of California
which allowed, at the time the costs were generated, a pressure drop
of 3 inches of water in 5 minutes. The costs are shown in Table 2-1.
The costs have been divided into two categories. The first includes
the cost of labor and equipment to initially bring an existing tank
truck within the limits of the specified vapor tightness. The
second category deals with the cost of maintaining a truck within
specified vapor tightness limits after initial certification.
These latter costs are generally.lower after the tank has been
maintained initially because many leak sources do not reappear at
every succeeding-vapor tightness test.
The costs to comply with San Diego requirements appear to be
about two times greater than the State requirements. It is
believed that as the allowable leak rate becomes smaller the
significance of smaller leaks becomes greater. Hence, additional
man-hours must be spent to identify and repair the smaller leaks.
Currently maintenance procedures and certification tests
are performed on an annual basis. Several operators indicated
that visual observations and minor maintenance are performed on the
tanks between annual certifications. This maintenance is usually
coordinated with the normal truck power unit maintenance schedule
2-21
-------
Table 2-1. COSTS FOR MAINTAINING TRUCKS IN LEAK
TIGHT3 CONDITIONS7'8'9'10
Initial
Certi fi cation c
Recertifi cation
San Diego6
California
Labor
Hours
30
11
4
Labor
$b
630
231
84
Materials
70
50
50
Total
Cost
700
281
134
Leak tight defined as passing certification tests.
bLabor rate = $21/hr.
°Initial certification indicates truck has not been previously
certified as leak tight.
Recertification = Any retest following the first
eSan Diego tests allow leak rate of 1 in.H20 in 5 minutes.
Costs are representative of a California allowable leak rate
of 3 in.H20 in 5 minutes.
2-22
-------
which is performed approximately every 4 to 6 weeks. This periodic
visual observation and minor maintenance normally costs
approximately $50 to $100 per occurrence depending upon the amount
of replacement parts required. The average cost is about $70 per
occurrence.
Estimates were received on the total tank maintenance required
on the tank truck product handling equipment. This included
maintenance for the vapor recovery hoses, couplers and adapters,
internal valves, overfill protection and vapor recovery equipment.
These costs ranged from $65 to $272 per month with an average
monthly maintenance cost of $160 per month. These costs are
average monthly costs based upon over 100 months of actual
12
maintenance performed at two tank truck shops in California. The
total estimated annual maintenance required on the tank equipment
would then be about $1,920 per year.
Incorporating these monthly maintenance figures and vapor
tightness costs into Table 2-2, the total costs including
certification testing and maintenance are shown. Costs are also
shown indicating total costs if certification tests were required
more often than once during the year. This is included because
some operators felt the certification may be more meaningful if
performed more often. However, with this they would like to see a
relaxation of the vapor tightness requirements. The California
certification program includes a gradual tightening of the leak
rate limits until, in 1979, the requirements are the same as San
Diego (1 inch of water in 5 minutes).
2.5.2 OTHER EMISSION SOURCES
Maintenance costs for other fugitive hydrocarbon emission
sources would involve mostly labor requirements for visual
inspections and cleaning of equipment. This would include visual
inspections of vapor return lines, storage tank pressure-vacuum
2-23
-------
Table 2-2. TOTAL ANNUAL MAINTENANCE COSTS FOR PRODUCT DELIVERY
EQUIPMENT3'6
Monthly
Costs
$
Total Annual
Costs
$
Product Delivery and
Vapor Tightness
Maintenance
Product Delivery and
Vapor Tightness
Maintenance and Annual
Recertification
Maintenance and Testing
San Diego
California
Product Delivery and
Vapor Tightness
Maintenance and
Semi-Annual
Recertification
Maintenance and Testing
San Diego
California
Product Delivery and
Vapor Tightness
Maintenance and Quarterly
Recertification
Maintenance and Testing
San Diego
California
160
1920
2200
2050
2480
2190
3040
2460
Product delivery equipment includes delivery tank, couplers, internal
valves, vapor recovery requipment, overfill protection, and dome
covers.
All costs are averages given on a per truck basis
2-24
-------
vents, couplers, and adapters. These activities would be performed
at bulk plants and terminals and by the tank truck operators.
However, this type of maintenance program was not performed by the
operators interviewed so no data on actual hours spent were obtained.
The costs discussed here can therefore only be estimated. A
visual inspection of the equipment including leak indication should
take no more than 2 hours. As in the tank truck maintenance, the
initial equipment inspection will turn up many more leaks than
subsequent inspections assuming the leaks are repaired after the
first inspection. The time required to perform the maintenance
following the initial inspection cannot be estimated because of
the numerous possibilities of leak sources. If a labor rate
similar to that used for the truck maintenance is used, the
inspection should not cost more than $42 per occurrence. The
most likely place for leaks, once piping losses have been
repaired, would be at pressure vacuum-vents or at vapor tight
dry break couplers. This maintenance would generally require
cleaning of valve seats or replacement of gasket material. If
it is assumed this takes an additional 2 hours and that average
replacement parts were similar in cost to that of tank trucks
($20), the total, cost of this maintenance would be $104 per
occurrence.
2-25
-------
References for Section 2.0
1. Leak Testing of Gasoline Tank Trucks, Scott Environmental
Technology, EPA Contract No. 68-02-2813, Work Assignment No.
19, August 1978 (Draft).
2. Ibid.
3. Powell, D.J. and D.E. Hasselman, Reliability Observations and
Emission Measurements at Gasoline Transfer Vapor Recovery
Systems. TRW, Inc., EPA Contract No. 68-02-0235, November 1974.
4. Letter from R.A. Nichols Engineering to H.B. Uhlig, Chevron
U.S.A., June 10, 1977.
5. Control of Hydrocarbons From Tank Truck Gasoline Loading
Terminals.EPA - 450/2-77-026, October 1977.
6. Presentation by John Snyder, Chevron U.S.A. to California Air
Resources Board, December 2, 1976.
7. Larry Cowie, Shell Oil, File data on Actual Maintenance
Performed.
8. Telephone conversation with Jack Ritterbush, J & L Tank,
Lynwood, California, March 1979.
9. Telephone conversation with Ray Schaffer, Weld-It, Los
Angeles, March 1979.
10. Telephone conversation with Ken Gay, Paramount Tank,
Paramount, California, March 1979.
11. John Snyder, 1976.
12. Larry Cowie, Shell Oil and Frank Canning, Chevron, U.S.A.,
File Data on Actual Maintenance Performed.
2-26
-------
3.0 DEVELOPMENT OF MONITORING PROCEDURE
3.1 TEST METHODS
Several test methods were explored as to their acceptability
or usability as a monitoring procedure within the confines of this
task. A test method was sought that would give pass-fail compliance
information for hydrocarbon leakage at various sources. The test
method desired was to be low cost, quick, and would not require
taking the truck out of service for any great length of time. The
methods researched for their usefulness are described in the
following sections.
3.1.1 VAPOR TO LIQUID VOLUME DETERMINATION (V/L)
This method determines a ratio of the volume of vapor
exhausted versus the volume of liquid loaded (V/L ratio). The
liquid volume is determined by monitoring the gallons of liquid
loaded and converting this to cubic feet. The volume of vapors
displaced are monitored by installing a low pressure drop, positive
displacement meter in the vapor return line. Pressure, vapor
temperature and liquid temperature can also be monitored during
transfers. The V/L ratio is a simple volume ratio without correc-
tions. However the additional physical data obtained can be used
to explain some phenomena which take place. This method has been
used before in conjunction with EPA mass emission tests at
both bulk plants and terminals. The EPA method calls for obtaining
a leak tight truck and determining the V/L ratio. An average V/L
ratio is then calculated for leak tight trucks. The leak tight
trucks are used to determine a baseline for comparison of the
other trucks tested. The V/L for other trucks checked during the
test period is compared to the baseline value to determine the
amount of vapor leakage which occured.
3-1
-------
3.1.2 EXPLOSIMETER
The explosimeter method calls for the use of an explosimeter
or combustible gas analyzer to monitor the potential leakage sources
described in Section 2.1 for evidence of hydrocarbon emissions. The
probe of the portable instrument is positioned around the potential
leak source and the meter reading recorded in percent of the lower
explosive limit (LEL). Explosimeters have long been used to
pinpoint leakage points when handling gasoline or hydrocarbon vapors.
The method calls for monitoring of truck hatches, P-V vents, couplers,
hoses, etc. during loadings and unloadings of gasoline from the truck
tanks and recording the relative leakage observed.
3.1.3 SONIC DETECTOR
The sonic detector is used in a similar fashion-as the explo-
simeter. Instead of measuring hydrocarbons, the sonic detectors
monitor the noise made by the gas escaping through the leak area.
The sonic detector can be used to measure leakage caused by any
gas and can be used if the system is either under pressure (leakage
out) or vacuum (leakage in). The sonic detector could monitor at
all the same emission sources as the explosimeter or combustible
gas analyzer.
3.1.4 SAN DIEGO "BAG" TEST
In this test method, a bag is placed over the dome cover to
capture and quantify the otherwise fugitive vapors. The bag is
attached to a modified bicycle tire which has been filled with sand.
The weight of the sand in the tire forces the assembly against the
truck tank and creates the vapor seal. The bag is sized based upon
calculations of the amount of vapors that would be lost given the
allowable pressure decline rate (1 inch hUO in 5 minutes for San
Diego County). The bag.is placed over the compartment which is being
loaded and the number of times the bag fills or the approximate
3-2
-------
volume of vapors collected in the bag is estimated. The bags are
oversized so that on filling, San Diego County inspectors are
certain that a violation has taken place.
3.1.5 PRESSURE-VACUUM TEST (CARB)
The California Air Resources Board (CARB) has passed regulations
which define the degree of tightness that is required on gasoline
delivery tanks. To ensure that this tightness is maintained, all
trucks must pass a pressure tightness test each year. A test
procedure was derived by the CARB which would be used to test the
trucks as to their tightness. This CARB procedure was then used as
the reference method, for this project, for comparison with other
methods. The truck, if its last load was gasoline, is purged of
volatile hydrocarbon gases by blowing air into the compartments
with the dome lids open. This purging has normally been'done for
about 10 minutes per compartment. This will remove the volatile
vapors and allow a better pressure determination within the
test tank. Some truck owners will either purge the compartment
with diesel or make the last load before testing diesel. This
will eliminate .the volatile vapors in the truck compartments and
eliminate the necessity of purging. The trucks are then brought
into a covered shop area where the effects of temperature variation,
and therefore pressure variation, caused by the sun and wind would
be minimized. The truck hatches are closed and the delivery and
vapor transfer hoses are attached and capped on the ends. The
internal valves are opened and the compartments are all manifolded
together. The compartments can be tested separately, but this is
considerably more time consuming.
The truck is then pressurized, most commonly with shop supplied
compressed air. A manometer is attached to the truck and the truck
pressure brought to 18 inches of water. The pressure loss versus
time is monitored and checked against the allowable leakage rate.
3-3
-------
The allowable leakage rate at the time of this report is 2 inches of
water in 5 minutes (from 18 to 16 inches of water) for all the State
of California except San Diego County. In San Diego, the allowable
leakage rate is 1 inch of water in 5 minutes. As of June 1, 1979,
the entire State will be required to meet the 1 inch in 5 minutes
criteria. The truck is then placed under vacuum, most commonly
using the vacuum supplied by the exhaust manifold of an automobile
engine. The tank is evacuated to 6 inches of water and the pressure
monitored again for 5 minutes. The allowable in-breathing at the
time of this report is 2 inches of water in California and one inch
of water in San Diego. Many of these other test methods discussed
are based upon estimating the amount of leakage that is allowed or
specified by the CARB certification test procedures.
3.1.6 BUBBLE INDICATION METHOD
This test method employs the use of a soap solution or other
solution which, will indicate gas leakage by the forming of bubbles
around the leakage area. The solution is applied to hoses, coupler
interfaces, hatch covers and pressure vacuum vents. The appearance
of bubbles indicates a leakage source.
3.1.7 QUICK LEAK DECAY
The quick leak decay method is similar in concept to the CARB
pressure-vacuum test method except that liquid is used to supply the
pressure or vacuum needed to determine the amount of pressure or vacuum
loss. Liquid, such as gasoline or diesel, would be desirable to use since
the truck would not have to be removed from service. During loading, the
vapor return line would be capped off, and liquid pumped into the
vehicle until the desired pressure is reached. The truck would be
allowed to stabilize and then the pressure decay would be noted.
During unloading at a bulk plant or service station the vapor return
line would be capped off toward the end of the unloading (tank close
3-4
-------
to empty) and gasoline allowed to flow out until the desired vacuum
is reached. This time the increase in pressure (or decrease in
vacuum) is monitored with respect to time. The leak rate during
the vacuum test is performed with a tank as close as possible to
empty to best correlate with the CARB test methods.
No one has used this quick leak decay method as described in
the field. This method was suggested because of its advantages.
The method is a quick, in-service test and is roughly equivalent to
the CARB procedures. Some California companies have used variations
of this test method. One operator pressurizes diesel vapors with
diesel fuel to obtain the desired pressures. Another operator
places the truck inside a garage and pressurizes gasoline vapors
with water. Time is needed however for vapors to stabilize.
3.1.8 VOLUME LEAKAGE
The volume leakage method maintains a constant pressure in the
test compartment by continually introducing air into the compart-
ment. It is assumed that the amount of air introduced into the
compartment to maintain the desired pressure is equal to the leak
rate at that pressure. The tank is pressurized to the desired
pressure in a similar manner as described in the CARB test method
and a rotameter is used to measure the amount of air necessary to
maintain the pressure. When the introduction rate has been
stabilized, the rate of air introduced into the tank is assumed to
be equal to the leak rate of gases out of the tank.
3.2 EVALUATION OF TEST PROCEDURES
The test procedures were included in a field test program
performed under a separate EPA contract (No. 68-02-2813). These
included the explosimeter method, sonic detector method, CARB
method, V/L method, bubble indication method, and the volume
leakage test. The quick leak decay method was analyzed under
3-5
-------
laboratory conditions and the San Diego "bag" test method was
observed in the field, as performed by San Diego County personnel,
Tank trucks were monitored during loading with an explosimeter
and sonic detector to determine leakage and an in-line volume meter
to determine the volume of vapor transferred. Shop tests such as
the CARB test and volume leakage test were also performed on each
test tank both prior to and after maintenance being performed.
Shop tests were performed on 26 tanks and leaks were monitored
during 207 tank loadings.
Both a top and bottom loading terminal were selected for
inclusion in the test program, The top loading terminal, operated
by Shell Oil Company, was located in Los Angeles, California. The
bottom loading terminal, operated by Chevron, U.S.A., was located
in Montebello, California. Selection of the truck fleet to be
tested was important to obtain a representative cross-section of
trucks. At the beginning of the project, data indicated that the
age of the tank and its corresponding vapor containing equipment may
be significant. Tanks were selected for the test program, therefore,
in an attempt to maximize the number of tanks that could be tested
and to obtain a reasonable cross-section of tanks of varying ages
and tanks with varying types of vapor containing equipment. Data
on the trucks selected for the test program are shown in Table 3-1.
Included in the table is information on capacity of the tank shell,
type of dome cover used, type of vapor piping employed, number of
compartments, type of suspension, and the year the tank was put into
service (tank age). The type of suspension was included in the data
because the spring type suspension is sized upon a fully loaded tank.
This results in a very stiff ride when empty and subjects the vapor
containing equipment to additional vibrations. Air suspension on the
otherhand can vary as the load changes and should yield a smoother
ride. All tanks tested used the overturn rail for the manifold line
on the vapor piping system. However, several types of vent valve
covers and piping to the overturn rail were observed.
3-6
-------
Table 3-1. TEST FLEET PHYSICAL DATA
Top Loading
Tank
I.D.
Number
63806
53306
63767
53256
63766
53345
63765
53304
63804
53307
63803
53297
63805
53305
Age3
8/74
12/74
7/73
9/68
7/73
9/77
7/73
11/73
4/74
12/74
4/74
7/74
4/74
12/74
No. of
Compart-
ments
2
3
2
3
2
3
2
3
2
3
2
3
2
3
Tank
Shell
Capacity,
Gallons
4408
5159
4408
5380
4408
5315
4408
5159
4408
5159
4408
5159
4408
5159
Liquid
Capacity,
Gallons
4000
4800
4000
No
Data
4000
4900
4000
4800
4000
4800
4000
4800
4000
4800
Hatch
Type
No
Data
No
Data
Tiona
Ti ona
Tiona
Tiona
C-B
2-C-B/
1-Tiona
Tiona
Tiona
Ti ona
Tiona
Tiona
Tiona
Suspen-
sion
Air
Spring
Air
Spring
Air
Spring
Air
Spring
Air
Spring
Air
Spring
Air
Spring
Vapor Recovery Type
Vent Cover
No
Data
No
Data
Welded
Welded
Welded
Rubber
Boot
Bolted
Cover
Bolted
Cover
Bolted
Cover
Bolted
Cover
Bolted
Cover
Bolted
Cover
Bolted
Cover
Bolted
Cover
Vapor Line
No
Data
No
Data
Welded
Welded
Welded
Rubber
Hose
Cl amped
& Welded
Pipe
Cl amped
& Welded
Pipe
Wei ded
Pipe
Wei ded
Pipe
Wei ded
Pipe
Welded
Pipe
Welded
Pipe
Wei ded
Pipe
3-7
-------
Table 3-1. TEST FLEET PHYSICAL DATA (CONCLUDED)
Bottom Loading
Tank
I.D.
Number
67-775
67-182
67-392
67-475
68-795
68-795*
68-597
68-597*
68-275
68-275*
68-977
68-977*
Age
6/67
4/71
4/73
10/74
4/78
4/78
9/75
9/75
11/62
11/62
8/69
8/69
No. of
Compart-
ments
5
5
5
4
2
2
2
2
2
2
3
3
Tank
Shell
Capacity
Gallons
8650
8650
8650
8650
5097
5319
5087
5327
4447
5053
5184
5180
Liquid
Capacity
Gallons
8150
8250
8050
8200
4600
4350
4300
4550
3670
4400
4000
4750
Hatch
Type
Tiona
Tiona
C-B
Tiona
C-B
C-B
Tiona
Tiona
C-B
C-B
Tiona
Ti ona
Suspen-
sion
Air
Air
Air
Air
Spring
Spring
Air
Air
Spring
Spring
Spring
Spring
Vapor Recovery Type
Vent Cover
Welded
Wei ded
Bolted
Welded
Welded
Welded
Welded
Welded
Bolted
Bolted
Welded
Wei ded
Vapor Line
Welded
Welded
Rubber
Hose
Welded
Welded
Welded
Welded
Welded
Rubber
Hose
Rubber
Hose
Welded
Wei ded
* Trailer
a
Tank age indicates year put into service
3-8
-------
The test programs at both the top and bottom loading terminals
were nearly identical. The trucks to be tested were scheduled into
the shop at varying times during the week for various tests to be
performed. Meanwhile, monitoring of leaks was performed on all
loadings of trucks included in the test plan. This included moni-
toring for leaks before and after shop tests and maintenance to
yield information on tanks that leaked and tanks that were vapor
tight. Loadings were monitored using the explosimeter method, sonic
detector method and the V/L method.
Before the trucks were tested in the shop, removal of the
volatile gasoline vapors was necessary. At the bottom loading
terminal the trucks were scheduled to haul a load of diesel before
being tested. The returning truck therefore had less volatile
diesel vapors in the compartment prior to pressurization. At the
top loading terminal diesel was not available so compartments were
purged with air. The purging took place by blowing air through the
compartments for 10-15 minutes each. This method did not prove to
be as effective as the diesel method for removing volatile vapors
because longer stabilization times were experienced.
In the shop, equipment was arranged and a volume leakage test
was conducted on the test tank before any maintenance was performed.
A volume leakage rate was determined for varying pressures starting
with 9 inches of water and increases to 18 inches of water in 3 inch
increments. This was performed to establish a leak rate before any
higher pressure might "blow" a leak and also to determine if the
leak rate increased rapidly with pressure. A CARB pressure and
vacuum test followed the volume leakage tests.
This established the condition of the truck with respect to
leak tightness prior to maintenance. Maintenance was then performed
on the truck tanks to make them leak tight as defined by the CARB
leak rate criteria (See Appendix B). A CARB pressure vacuum test
and volume leak rate test were then performed again to establish the
post maintenance condition of the truck.
3-9
-------
3.2.1 PRESSURE-VACUUM TEST (CARB)
The CARB pressure-vacuum tests were performed on each tank at
least once during the test period. The tanks were tested both prior
to maintenance and after maintenance procedures were performed.
This defined the leak rate and the subsequent compliance status of
the tank with regard to the CARB annual certification compliance
criteria. This criteria is currently set as allowing a leak in
the tank such that when the tank is pressurized to 18 inches of
water, the pressure will not decline more than 2 inches of water in
5 minutes. The vacuum criteria allows a decrease from 6 inches of
water vacuum to 4 inches of water vacuum in 5 minutes. By June
1979, the criteria for both pressure and vacuum variations will be
set at 1 inch of water in 5 minutes. The Air Resources Board was
contacted to determine the reason for the particular pressure and
vacuum limits used in the certification test procedure. The
pressure and vacuum vent valves are spring loaded and designed to
slowly open and be full open at 1 psi (27 inches of water) pressure
and 6 ounce (10 inches of water) vacuum as specified by DOT regula-
tions. The limits selected by CARB are the maximum pressure or
vacuum that can be applied to the tank before the vent starts to
open.
Back pressures observed during loading operations at both the
top and bottom loading terminals ranged from as low as 1.7 inches of
water to as high as 17.5 inches of water which approaches the value
used by CARB in their certification testing. The average back
pressure over 144 tank loadings was 7.3 inches of water. Figure 3-1
shows the frequency distribution of the back pressures experienced.
These indicate, with the types of vapor recovery systems employed,
that the loadings observed at both the top and bottom loading
terminals are consistently lower than the 18 inch criteria used in
the CARB certification test. Thus, if a tank is leak tight as
defined by the CARB criteria, the terminal back pressure at these
test sites would not be expected to cause a leak.
3-10
-------
22_j
20-
18-
16-
14 _
12 _
CO
I
t/l
OJ
OJ
S-
3
0 Q
O O
O
(O
c
O)
O
6 _
< 0
D- 2
Top Loading
Bottom Loading
0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 10-11 11-12 12-13 13-14 14-15 15-16 16-17 17-18
Back Pressure, Inch H0
Figure 3-1. Frequency of Occurrences of Back Pressure During Loading Operations
-------
Historical data were obtained from the terminal operators on
each of the test tanks which included the dates of the previous
certification. These data were used to determine how long the trucks
were able to remain in certifiable condition under normal use. The
shop test data were used to indicate the condition of the tanks
which was compared to the historical data (see Table 3-2). As
indicated, only four tanks remained in certifiable condition. The
time since the last certification date ranged from four months to
four days. Of the tanks that failed, the time frame ranged from
one year to two weeks.
The CARB tests performed in the shops were also monitored for
pressure drop past the 5 minute time requirement. Tests were run
for as long as 20 minutes to determine if the pressure would
continue to drop with time or would level off at a lower pressure.
This was performed predominantly on the tanks before maintenance
was performed to determine the extent of the leak. Figures 3-2 and
3-3 illustrate examples of the pressure decay rate. In almost every
case, the pressure continued to decline at an approximately linear
rate. This indicates the pressure decay rate is more dependent on
the size of the leak than on the pressure in the test tank.
A phenomenon was also noted in the shop tests at the top loading
facility. The tank was pressurized to 18 inches of water and the
pressure increased with time. This occurred at the top loading
terminal where the compartments were degassed by blowing air through
them and not at the bottom loading terminal where the gasoline vapors
were removed by hauling a load of diesel prior to the test. This
phenomenon was duplicated in the laboratory using a small leak tight
tank which had held gasoline. The tank was loaded and unloaded with
gasoline several times. A pump was then used to push air through the
tank to simulate the degassing. After 20 minutes of degassing the tank
was immediately pressurized and sealed. The pressure again increased with
3-12
-------
Table 3-2. TANK TIGHTNESS HISTORY
Tank
Identification
Number
63806
53306
63767
53256
63766
53345
63765
63765d
53304
63804
53307
63803
53297
63805
53305
67-282
67-392
67-475
68-795
68-795b
68-597
68-597b
68-275
68-275b
68-977
68-977b
67-775
Last
Certification
Date3
2/24/78
2/24/78
2/09/78
2/09/78
2/08/78
2/08/78
2/07/78
6/19/78
2/07/78
2/16/78
2/16/78
2/10/78
2/10/78
2/28/78
2/28/78
5/18/78
5/23/78
5/24/78
5/03/78
5/03/78
6/13/77
6/13/77
6/28/77
6/28/77
5/03/78
5/03/78
No data
Field
Test Date
6/23/78
6/23/78
6/20/78
6/20/78
6/20/78
6/20/78
6/19/78
6/23/78
6/23/78
6/21/78
6/22/78
6/21/78
6/21/78
6/22/78
6/22/78
6/14/78
6/13/78
6/15/73
6/14/78
6/14/78
6/13/78
6/13/78
6/15/78
6/15/78
6/16/78
6/16/78
6/12/78
Pass/
Fail
F
F
F
P
Fc
Fc
F
P
P
F
F
F
F
F
FC
F
F
P
F
F
F
F
FC
Fc
FC
Fc
F
Type of
Loading
Top
Top
Top
Top
Top
Top
Top
Top
Top
Top
Top
Top
Top
Top
Top
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
aPassed CARB Certification Test
Trailer of Truck/Trailer Unit
GTanks not maintained to pass conditions after field tests, all
other tanks maintained to pass conditions.
Tested twice during the field tests
3-13
-------
I I-
I •
Chevron Terminal
6-13-78
Truck 68-597 (trailer)
CARD Pressure Test
£ • Premaintenance
Postmaintenance
'^"f r~rT''
' \ i ^^
14__,
__r£
^=^n~
=±=^=4r=:
\—4-
-i—
l • !
—-(=
=^Tf
—I •(-
=t
__ r\
—'TT-
—T"
i 1 1 e fr)
Figure 3-2. Typical CARS Pressure Test Results at Bottom Loading Terminal
(Ref. 1)
3-14
-------
Shell Terminal
6-19-78
Truck 63765 (Trailer)
CARB Pressure Test
• Premaintenance » Postmaintenance
Figure 3-3. Typical CARB Pressure Test Results at Top Loading Terminal
(Ref. 1)
3-15
-------
time. When a vacuum test was attempted, the results appeared normal
with the vacuum decreasing with time. This was repeated several
times with the same results. The tank was then degassed and
allowed to sit for 20 to 30 minutes. The tank was again pressurized
and the results appeared normal with an initial equilization and
then a constant pressure (see Figure 3-4). The data indicates that
the pressurization of highly volatile vapors are greatly affected
by slight changes in temperature. If proper degassing and stabili-
zation is not performed, two possible results could occur. The
pressure may increase with time, as described above, and if it
increased faster than or equal to the actual leak rate the tank may
be erroneously considered certified. The other possibility is that
the pressure may decrease in the tank due to condensation of the
pressurized vapors. The tank may then be wrongly defined as a
leaking tank. In any case, the degassing and stabilization
procedures appear to be important and a period of time after
degassing is necessary before testing to allow the tank to stabilize.
For the purposes of a quick monitoring procedure as outlined
by this project, the CARB method is not very reasonable. The truck
must be taken out of service, degassed, and tested. Regardless if
the maintenance is performed or not, the truck is out of service
for approximately 3 hours. This required using either a back-up
vehicle or rescheduling of deliveries. However, since this method
quantitatively defines what a leak tight truck is, the method is
very useful as an enforcement tool.
3.2.2 QUICK LEAK DECAY METHOD
The purpose of the quick leak decay method was to develop a
quick procedure that was directly equivalent to the CARB procedure.
The quick leak decay method was not included in the field test
program because it was felt that additional information based upon
laboratory tests was needed along with information regarding the
3-16
-------
22 _
20 _
i
L
18 _
1 C
16 _
14 _
s_
01
5 12 _
UJ **-
1 O
- | 10 _,
o
c
-------
safety limitations. Safety was a factor that had to be dealt with
because the method called for the pressurization of the delivery
tank with liquid gasoline. Several fire marshal!s were contacted
and all gave similar responses. Since the tank pressure sought
(18 inches of water) is below the pressure at which the tank is
designed to maintain around the liquid gasoline (27 inches of
water), there should be no safety problems. This was provided that
the normal safety precautions regarding static discharge and
proximity to sources of flame are upheld.
The laboratory tests were conducted by first constructing a
leak tight delivery tank simulation model. The tank had to be leak
tight to ensure that pressure changes taking place inside the tank
could be attributed to liquid-vapor equilibrium changes and not to
leaks in the tank. A pressure transducer was attached to the tank
and all other ports sealed. The tank was pressurized and the
pressure recorded on a strip chart recorder. A leak was found at a
tank weld and repaired. Thermocouples were installed and attached
to a multipoint recorder to monitor ambient temperature, tank
vapor temperature, liquid temperature in the tank, and liquid
loading temperature. The apparatus was assembled as shown in
Figure 3-5. The. tank had a capacity of approximately 10 gallons.
To eliminate the hazards of pumping the gasoline, gravity was used as
a driving force for loading or unloading the liquid gasoline.
After initially loading the tank with gasoline, it took some
time to reach a stabilized condition (the pressure no longer
increased with time). Once the system had stabilized, liquid was
forced into the sealed tank until the desired pressure was reached.
Temperature and pressure were recorded and the time required to
reach a stable pressure noted. The pressure was released and the
tank again sealed with the liquid still inside. The liquid was then
allowed to drain out until the desired vacuum in the tank was
achieved. Again temperature, pressure, and time were recorded until
3-18
-------
1. Multipoint temperature recorder 7.
2. Ambient temp 8.
3. Vapor temp 9.
4. Tank liquid temp 10.
5. Entering liquid temp 11.
6. Liquid gasoline supply 12.
Pressure relief vent
Pressure transducer
Signal conditioner
Strip chart recorder
Test tank
Needle valve
CO
.1.
Figure 3-5. Quick Leak Decay Test Apparatus
-------
a stable pressure was reached. Typical pressure versus time curves
are shown in Figure 3-6. The time for stabilization of the pressure
tests was normally about 10 minutes.
It is difficult to estimate what time period would be necessary
to stabilize pressure in a full scale tank vehicle. It is safe to
say,however, that this time period should be significantly longer.
This time requirement will probably eliminate the usefulness of this
method as a quick detection technique. Diesel fuel has been used
successfully as a pressurizing liquid after the tank has been rinsed
4
or flushed with diesel. Diesel fuel is not available at all loading
facilities so the method was assessed for acceptability using
gasoline. Even though the test demonstrates that this method may
not be usable as a quick monitoring technique, it did illustrate the
need for removing as much gasoline vapor as possible from the test
tank and allowing the test tank to stabilize before testing to
obtain reliable results.
3.2.3 VOLUME LEAKAGE
The volume leakage test was performed in the shop prior to the
CARB test and the results were compared with the results of the CARS
test. The volume leakage test is simply a variation of the CARB test
that measures the volume lost at a specified pressure instead of
pressure drop. The method was selected to determine if it was
simpler and more reliable than the CARB test. The pressure was
first increased and held at 9 inches of water, then held at 12
inches of water, then 15 inches of water and finally, at 18 inches
of water. This approach was used before the CARB test so that the
lower pressure data could be obtained before reaching the higher
pressures and eliminating the possibility of "blowing" a leak at the
18 inch pressure. The results of the volume leakage data as they
compare to the CARB data is shown in Figures 3-7 and 3-8 for bottom
and top loading instances. The top loading data are not as complete.
3-20
-------
CO
I
rsi
i.
0)
l/l
0)
u
c
Ol
s_
3
LO
CO
OJ
22
20
18
16
14
12
-1]
-t
10-A
8 _
6 -
4 _
A
a
a a
A
A
A A
n = Pressure tests
A = Vacuum tests (pressures
noted are inches of
water vacuum).
A A A A
A
2 -
I
2
"T"
3
"T T"
4 5
7
Time, Minutes
I
10
n
12
Figure 3.6. Typical Pressure Versus Time Curves for Laboratory
Tests of Quick Leak Decay Method
-------
6.0
5.0
4.0
9 inch H20
3.0
2.0
1.0 _
15 inch H20
18 inch H20
12 inch H20
0,
A,
a ,
o ,
18
15
12
9
inch
inch
inch
inch
H20
H20
H20
H20
oo
1 1
50 100
1
150
1
200
1
250
1
300
1
350
400
1
450
Volume Leakage, SCFH
Figure 3-7. Volume Leakage vs Pressure Decay for Bottom
Loaded Tanks at Various Pressures
-------
0.4
O)
O
C\J
o
c
0.3
OJ
I
1X0 •
C£
o
i
Ol
to
O)
s_
Q-
0.2
0.1
O A a
D
A
O
_ 18 inch H20
_ 15 inch H20
_ 12 inch H20
I
5
I
10
15
I
20
I
25
30
I
35
40
45
Volume Leakage, SCFH
Figure 3-8. Volume Leakage vs Pressure Decay Rate for Top
Loading Tanks at Various Pressures
-------
as the bottom loading data because time constraints when performing
the field tests did not allow testing of all of the tanks using the
volume leakage method. Linear regression analyses were performed
on the data and the best fit curves are illustrated. The correlation
coefficients were calculated and the bottom loading data are signif-
icant at the 0.1 percent probability level. The top loading data
were significant at the 10 percent level. Correlation coeffi-
cients for the volume leakage versus CARB test are listed in
Table 3-3.
The volume leakage test compared favorably to the CARB
method. However, the volume leakage method is not shorter or quicker
than the CARB test but in fact longer and requires more equipment.
The truck must still be taken out of service and degassed. In addi-
tion to all of the necessary CARB equipment, rotameters and correspond-
ing valves and tubing for measuring volume rate are required. This
method is, however, similar to the CARB method in that it can be
used to define an acceptable leak rate.
3.2.4 V/L RATIO METHOD
Vapor to liquid volume ratios were determined for 120 loadings
over the 2 week test period. The results were separated on a daily
basis since the ambient conditions can strongly affect the V/L ratio
which can be expected on a leak tight truck. Table 3-4 and Table
3-5 indicate the results of the V/L tests for both top and bottom
loading. These tables present the V/L ratio for tanks that passed
the CARB certification tests and those that failed.
The EPA terminal tests using the V/L method prescribe
determining the V/L from a leak tight truck for a particular day or
set of conditions and then comparing this value to the other trucks
tested. The data presented in Tables 3-4 and 3-5 indicate a wide
variability in the V/L ratio for both the tanks that pass and the
3-24
-------
Table 3-3. CORRELATION COEFFICIENTS FOR VOLUME
LEAKAGE RESULTS WITH RESPECT TO CARB TEST RESULTS
Constant Pressure Held
Correlation Coefficient
Bottom Loading
9 inches H20
12 inches H20
15 inches H20
18 inches H20
Top Loading
12 inches H20
15 inches H20
18 inches H20
0.96
0.95
0.94
1.00
0.90
0.90
0.65
3-25
-------
Table 3-4. V/L RESULTS FOR TOP LOADING4
Run
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Date
6/19
6/19
6/19
6/19
6/19
6/19
6/19
6/20
6/20
6/20
6/20
6/20
6/20
6/20
6/20
6/21
6/21
6/21
6/21
6/21
6/21
6/21
6/21
6/21
6/21
6/21
6/22
V/L Ratio
Passb
1.-05
1.04
1.05
1.02
0.99
No data
No data
1.08
1.07
1.39
Fail1-
1.12
1.16
1.39
0.99
1.18
1.34
1.14
1.06
1.11
1.56
1.08
1.20
1.43
1.18
1.08
0.70
1.05
Temperature
Air
°F
80
80
80
80
80
80
80
66
70
70
76
80
80
78
78
68
68
68
68
70
74
74
80
80
80
80
70
Vapor
°F
90
92
92
92
92
93
90
80
80
83
88
90
92
90
90
76
80
80
90
90
90
92
94
90
75
Back
Pressure
(inches H20)
9.25
7.5
9.8
15.0
10.7
10.8
7.8
8.0
5.6
9.8
9.7
15.6
11.1
6.0
12.3
10.3
9.1
8.9
11.0
13.0
12.2
5.1
11.5
12.6
5.9
3-26
-------
Table 3-4. V/L RESULTS FOR TOP LOADING (CONCLUDED)
Run
No.
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
Date
6/22
6/22
6/22
6/22
6/22
6/22
6/22
6/22
6/22
6/22
6/22
6/22
6/22
6/22
6/22
6/22
6/22
6/22
6/23
6/23
6/23
6/23
6/23
6/23
6/23
6/23
V/L Ratio
Passb
0.95
No data
1.07
1.01
1.05
No data
1.15
No data
1.42
No data
No data
1.06
0.80"
1.22
0.80
No data
0.71
No data
1.09
1.04
Fai1u
1.49
1.47
1.06
1.06
1.01
1.12
Temperature
Air
°F
70
70
70
72
78
80
80
80
81
82
82
82
72
72
77
77
77
78
80
80
84
88
88
84
84
78
Vapor
°F
80
82
80
83
89
88
94
92
90
86
86
86
86
88
92
96
96
87
85
88
Back
Pressure
(inches H20)
9.3
10.8
10.6
9.0
12.8
8.9
12.7
10.9
10.5
8.6
10.5
8.5
6.7
10.8
6.6
6.3
8.2
9.2
8.9
5.2
3-27
-------
Notes for Table 3-4
All trucks tested were truck and trailer units but a V/L
ratio could not be obtained for each tank. Tanks were loaded
simultaneously and all loading arms manifolded together
before the vapor meter.
Pass indicates a tank that will meet the CARB leak tight
criteria.
c Fail indicates a tank that leaks greater than the allowable
rate defined by the CARB leak tight criteria.
3-28
-------
Table 3-5. V/L RESULTS FOR BOTTOM LOADING
Run
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Date
6/12
6/12
6/12
6/12
6/12
6/12
6/12
6/12
6/13
6/13
6/13
6/13
6/13
6/13
6/13
6/13
6/13
6/13
6/13
6/13
6/13
6/13
6/13
V/L Ratio .
Pass0
1.02
1.02
1.26/
1.02
0.96
l.ll/
0.78
0.88
0.84/
0.80
Fai1u
1.23
0.86
0.97
0.55/
0.88C
0.85/
0.75
0.84/
0.95
0.96/
0.63
1.01
0.94
1.06
0.89
0.99
0.79
0.98
0.61
0.37/
1.06
Temperature
Air
°F
95
95
95
93
91
•91
90
90
Vapor
°F
98
98
101
100/
110
108
108
110
106/
103
78
79
98
100
90
96
no/
102
108
106
110
102/
103
100
100/
103
100
100
Back
Pressure
(inches H20)
11.5
4.4
5.5
4.1/
4.8
1.7/
2.4
9.5/
7.1
7.3
7.5/
4.9
2.7
3.0
5.1
5.5
6.9
15.8
6. 1/
6.3
3.4
8.9
10.1
10. 2/
4.1
4.3
3.8/
2.6
9.1
3.8/
6.0
3-29
-------
Table 3-5. V/L RESULTS FOR BOTTOM LOADING (CONTINUED)
Run
No.
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
Date
6/14
6/14
6/14
6/14
6/14
6/14
6/14
6/14
6/14
6/14
6/14
6/14
6/14
6/14
6/15
6/15
6/15
6/15
6/15
6/15
6/15
6/15
6/15
V/L Ratio .
Passa
0.71/
0.82
0.84
1.15
0.83
0.97
0.67
0.88/
1.12
0.99
0.91
0.92/
0.72
1.07
0.80
0.97
0.78
0.99
0.83/
0.97
Failu
0.69
0.87/
1.00
0.93/
0.67
0.80/
0.91
0.80/
0.75
1.08
0.92
Temperature
Air
°F
68
71
73
76
78
79
85
86
87
88
. 87
88
88
87
67
67
69
71
73
74
77
79
83
Vapor
°F
78
83
80
84
89/
92
85
90
90
100
92 /
102
98/
104
96
100
100/
90
75
70
76
82/
80
86
83
90
90/
92
98
Back
Pressure
(inches H20)
17. 5/
16.2
3.0
6.6
3.0
2.7/
2.3
4.9
1.11
8.1
10.7
10.0
4.0/
3.0
9.4/
7.1
4.0
8.9
4.3/
5.6
6.4
7.4
6.2
8.0/
7.5
11.0
7.1
4.1
6.5
4.3
5.4
3-30
-------
Table 3-5. V/L RESULTS FOR BOTTOM LOADING (CONCLUDED)
Run
No.
47
48
49
50
51
52
53
54
55
56 .
57
58
59
60
61
62
63
64
65
66
67
Date
6/15
6/15
6/15
6/15
6/15
6/15
6/15
6/15
6/16
6/16
6/16
6/16
6/16
6/16
6/16
6/16
6/16
6/16
6/16
6/16
6/16
V/L Ratio .
Passa
0.96
0.83/
0.70
0.84/
0.88
0.96/
0.93
0.89
1.03
0.90/
1.04
1.00
1.06
0.91/
0.74
0.99.
1.02
1.52
1.07
1.04/
1.17
1.03/
0.98
0.95
Fail"
0.81
0.88
1.12/
1.14
0.58/
0.88
Temperature
Air
°F
83
86
84
86
86
87
86
- 87
67
70
72
78
79
80
84
85
85
87
85
85
84
Vapor
Op:
98
103/
90
90/
104
90/
107
90
92
103
102/
87
80
72/
84
76
78
88
90
97
100/
93
79/
81
101/
103
95
100/
92
Back
Pressure
(inches H20)
5.4
4.1/
8.3
3.5/
4.0
3.4/
3.4
4.7
4.7
4.1
6.2/
7.3
10.5
6.9
6.9/
4.6
12.0
3.9
3.7
4.7
6.2
5.5/
6.9
7.5/
9.4
8.0/
6.8
3.5
3.0/
2.3
3-31
-------
Notes for Table 3-5
a Pass indicates a tank that will meet the CARB leak tight
criteria.
Fail indicates a tank that leaks greater than the allowable
rate defined by the CARB leak tight criteria.
c Indicates the truck tested was a truck and trailer unit.
Data is presented for each tank (truck/trailer).
3-32
-------
tanks that fail the CARB reference test. A frequency distribution
of the V/L ratios for both the tanks that passed and the tanks that
failed is shown in Figure 3-9. As indicated by this figure, the
V/L ratio takes the same frequency of occurrence regardless of
whether the tank passed or failed the certification tests.
An attempt was made to determine if there was any correlation
between V/L ratio and the back pressure observed in the tank during
loading. Correlation factors were calculated for the V/L ratio
versus back pressure. The correlation factors are shown in Table
3-6, and as indicated by the low factors, there is no correlation
between V/L ratios and back pressure.
In summary, the V/L ratio showed no relationship between vapor
tight tanks or tanks that leaked. The V/L ratio also showed no
relationship when compared to the back pressure experienced during
the loading operation.
3.2.5 EXPLOSIMETER METHOD
The explosimeter method was extensively tested in the field
test program. The CARB pressure test was used as the reference to
determine the acceptability of the explosimeter method. During
each tank loading, the explosimeter was used to monitor for leaks
at the hatch cover, P-V vent, vapor piping, couplers, etc. The
explosimeter used was calibrated against propane and used a 1/4-inch
diameter probe. . When monitoring for leaks, the explosimeter probe
was held at a distance of 1 inch from the source and the highest
level was recorded. Loadings were monitored before and after the
CARB certification test and maintenance to determine whether the
tank was under a leak tight condition as defined by the CARB criteria.
The readings taken by the explosimeter. were then compared to the
CARB results to determine the magnitude of readings obtained on a
leak tight tank and on a tank which is known to leak. This was
performed at both top and bottom loading facilities.
3-33
-------
30 _
CO
25
20_
o
u
OJ
Q.
5-
Pass
Fail
.1
V/L Ratio
Figure 3-9. Frequency Distribution of V/L Ratio for Tanks That Pass
Certification Tests and Tanks That Fail Certification Tests
-------
Table 3-6. CORRELATION FACTORS (r) FOR V/L RATIOS
AND TANK LOADING BACKPRESSURE
Loading
Top
Top
Bottom
Bottom
Combined
Combined
Pass/Fail
Pass
Fail
Pass
Fail
Pass
Fail
r
0.49
-0.02
0.04
-0.02
0.03
-0.07
3-35
-------
The results were much more consistent and meaningful at the
bottom loading terminal than at the top loading terminal. At the
top loading terminal, since the hatch cover is not closed, the data
do not reflect the relative vapor tightness of the tank which is
traceable back to the CARB pressure test. Instead, the loading arm/
hatch interface, loading arm connections and swivels become the major
sources of hydrocarbon leakage.
Tables 3-7, 3-8, and 3-9 show the results of the explosimeter
tests performed at the loading terminals. Data were not included
in the tables if the previous load was diesel fuel or if liquid
interference would cause an erroneously high reading. If data were
not complete for all compartments of a tank, the tank was not
included in the data base. When discussing the usefulness of this
test method as a means of performing tank vapor tightness, the
discussion will be limited to the bottom loading data. Leaks were
detected at almost every top loading occurrence regardless of whether
the tank had passed or failed the CARB certification test. The data
from the bottom loading terminal indicate that the tanks that fail the
CARB certification test show considerably more leaks at the compart-
ments than the tanks that pass the certification tests. This becomes
more significant as the explosimeter readings become larger. For
example, at the 0.5 lower explosive limit (LEL) level, over 90 percent
of the compartments indicating this relative size of leak were those
that failed the certification tests. At the 1.0 LEL level, over 95
percent of the hatches indicating this relative size of leak, failed
the certification tests.
When discussing the data on a per tank basis rather than on a
per compartment basis, the method appears even better. As shown in
Table 3-9, every tank that failed the field CARB tests was judged
a fail tank by the explosimeter method. In other words, not one
leaking truck was missed using the explosimeter method. At the 1.0
LEL level, 9.5 percent of the tanks which passed the CARB test were
incorrectly identified as a leaking tank.
3-36
-------
Table 3-7. PERCENT OF COMPARTMENTS WITH LEAKS3 AT SPECIFIC LOCATIONS
Top Loading
Pass0
Faild
Total e'f
Bottom Loading
Pass0
Faild
Total e'f
Top
Hatch
99. 3b
94.9
97.3
5.2
42.3
22.7
Base
Ring
85.9
82.2
84.2
8.6
42.3
22.2
P-V n
Vent9
N.A.
N.A.
N.A.
7.8
48.7
24.2
Loading
Arm Vapor
Connectors
43.0
33.1
38.5
N.A.
N.A.
N.A.
Terminal
Vent
Cover
40.8
34.7
38.1
N.L.h .
N.L.
N.L.
For purposes of this table, a leak was defined as any deflection of
the explosimeter scale.
Top Hatch top loading indicates the loading arm/hatch interface
°Pass indicates a tank that is leak tight as defined by the CARB
criteria
Fail indicates a tank that has failed the CARB test and therefore
leaks in excess of the allowable criteria
eTotal = Percent of all compartments where leak occured at specific
location
For bottom loading a total of 195 hatches were tested, 116 in the
pass mode and 79 in the fail mode. For top loading, 260 hatches
were tested, with 142 in the pass mode and 188 in the fail mode
9N.A. = Not applicable for top loading since compartment loaded with
hatch open
hN.L. = No leaks detected
3-37
-------
Table 3-8. OCCURRENCE OF HYDROCARBON LEAKAGE
AT VARIOUS EXPLOSIMETER LEVELS3
Fop Loading
Passc
Faild
Jottom Loading
Pass
Fail
Percent
of Compartments Where
Leak Occurs
>0
LEL
100
96.6
19.6
63.5
>0.5
LEL
89.4
88.1
5.2
53.2
>1.0
LEL
85.9
84.7
2.6
39.2
Percent of
Tanks Where
Leak Occurs
>0
LEL
100
100
42.9
100
>0.5
LEL
96.0
97.0
19.0
100
>1.0
LEL
94.0
97.0
9.5
100
Avg. No. of
Leaks per
Compartment
>0
LEL
1.70
1.80
0.36
1.58
>0.5
LEL
1.05
1.21
0.07
1.09
>1.0
LEL
0.96
1.02
0.03
0.80
Percent
of all
Leaksb
>0
LEL
53.2
46.8
25.1
74.9
>0.5
LEL
51.0
49.0
8.5
91.5
>1.0
LEL
53.1
46.9
4.5
95.5
CO
CO
For bottom loading, a total of 41 tanks (.195 compartments) were tested, 21 tanks (116 compartments)
in the pass mode and 20 tanks (79 compartments) in the fail mode. For top loading, 84 tanks (260
compartments) were tested, with 50 tanks (.142 compartments) in the pass mode and 34 tanks (118
compartments) in the fail mode.
Indicates the percent of all leaks at that explosimeter level found at a compartment on a tank in
either the pass or fail mode.
cPass indicates a tank that was judged leak tight based upon the field CARB tests.
Fail indicates a tank that failed the field CARB tests and therefore leaked in excess of the allow-
able criteria.
-------
Table 3-9. PERCENT OF TANKS IDENTIFIED CORRECTLY AT VARYING
LEL LEVELS USING THE EXPLOSIMETER METHOD9
Passb
Fail0
All Tanksd
Percent of Tanks
Judged Correctly
At
>0
LEL
57.1
100
78.0
>0.5
LEL
81.0
100
90.2
>1.0
LEL
90.5
100
95.1
Percent of Tanks
Judged Incorrectly
At
>0
LEL
42.9
0
22.0
>0.5
"LEL
19.0
0
9.8
XL.O
LEL
9.5
0
4.9
Data for bottom loaded tanks only
Pass = leak tight as defined by the CARB field tests
cFail = tank leaked in excess of the criteria during field tests
Total of all bottom loaded tanks
3-39
-------
The success rate of the explosimeter method can be explained in
two ways. If a leak were found above a certain LEL level, the
success rate could be defined as the percentage of tanks that were
actually identified correctly as a leaking tank. For example, as
shown in Table 3-8, if a leak registering > 1.0 LEL were discovered
on a bottom loaded tank, the tank would be correctly identified as
a leaking tank 95.5 percent of the time.
The other success rate definition would be based upon the
ability of the method to correctly identify the tank as a leak
tight or leaking tank. At the >1.0 LEL level, as shown in Table 3-9,
the explosimeter method correctly identified the tanks 95.1 percent
of the time. The method did not overlook any tank that would
have violated the leak tight criteria.
The explosimeter method, as used in this study, had a success
rate at the > 1.0 LEL level of 95 percent using either definition.
The data was more meaningful at bottom loading operations but this
does not mean that the method is not applicable to top loading
systems. The presence of a leak was still detected, but it was
difficult to determine if the violation was caused by the tank
truck or the loading apparatus. It should also be noted that the
adequacy of the method, and the apparent LEL levels where the
method is usable, are based upon the CARB. leak criteria. If other
criteria are used to define a leak tight truck, the method will have
to be re-evaluated to determine the applicable LEL levels to be used
as pass/fail criteria.
3.2.6 SONIC DETECTOR
The sonic detector was used in much the same manner as the
explosimeter. The hatch covers, P-V vents, vapor collectors and
vapor piping were checked for leakage during the loading operation.
This unit was used almost exclusively when the previous load was
diesel because no volatile gasoline vapors were in the gases being
emitted.
3-40
-------
The sonic detector test method did not prove to be very useful
in the field. The instrument itself worked well in detecting the
presence of gas leakage, however, the instrument could not give
repeatable results on the relative size of the leakage. The
instrument sensitivity and indicator scale would vary with the
volume setting of the instrument. On a given constant leak, two
people could get two readings based on the volume setting of the
instrument and hearing ability of the operator.
However, this instrument and method could become usable if
either a calibration method is devised based upon a standard sound
level or if the instrument were modified to incorporate several set
ranges which in turn could be calibrated.
3.2.7 BUBBLE INDICATION METHOD
The bubble indication method is used by many mechanics during
tests to indicate the presence of leaks. This method was incorpor-
ated into the shop test procedures and used to indicate the locations
of the leak sources. The bubble method proved to be too sensitive
in indicating leaks. For example, a vapor hose connected to a tank
during a pressure test indicated a leak with a series of extremely
small bubbles along the entire length of the hose. The pressure
test indicated no leakage at all. The bubble test would also be
able to relate relative sizes of leaks only to a limited extent and
this would vary with the indicating solution used. This test is also
limited to tanks that are subjected to positive pressures only. This
method is however, similar to the sonic detector method in that both
can be used as an indicator or locator of leaks.
3.2.8 SAN DIEGO "BAG" METHOD
Compliance tests of tank trucks were observed as performed by
San Diego Air Pollution Control District (SDAPCD) personnel at a
bottom loading terminal in San Diego. As described in Section 3.1,
3-41
-------
the tire-bag apparatus was placed over all hatch covers of the tank
being loaded. This included putting bags over compartments that
were not being loaded but were interconnected through the vapor
piping. No leak should occur at the compartment that was not being
loaded because the internal valve should be closed, however, if a
leak is detected this would indicate a faulty internal valve and the
presense of a leak source.
As the truck is loaded, the bag is closely watched to observe
any leakage around the hatch. The bags are sized so that the
volume escaping is about twice the allowable leak volume based upon a
leak decay of 18 inches of water to 17 inches of water in 5 minutes.
A violation can then be easily noted by estimating the volume of
leakage collected in the bag (or several bags). The bags are
placed only over the hatch covers, which include the P-V vents,
because this is the most predominant leak source and other leakage
areas could be determined during the annual State Certification
Test (see CARB pressure test).
The test method has several advantages. The equipment is
inexpensive to buy and easy to use. A visual, easily detected
violation can be determined. Any leakage around the bag or
innertube - truck hatch interface are errors always in favor of the
truck. According to SDAPCD personnel, they have a 100 percent
success rate on trucks they note as violators. In other words,
every truck they have cited for violation has in fact, after a shop
test, been found to leak in excess of the allowed rate. The truck
must then be maintained to a level to pass the annual requirements
again.
This test does have its drawbacks. This method can obviously
be used only on bottom loaded trucks. The bag apparatus itself may
not be able to fit all tank hatch configurations. Other equipment,
such as overfill protectors, tank compartment vapor vent housing,
etc., may get in the way and not allow a good seal between the tank
3-42
-------
and the bag. However, the bag equipment is somewhat flexible and
very inexpensive and bags could be made to fit several configurations.
The other drawback is that there are many errors that may be involved
and that these are all in the favor of the. truck., This means some
trucks get by that may actually fail the criteria. The test is now
set up so that they do catch the larger violators and have a 100
percent success record for their field procedure.
3.3 PASS/FAIL CRITERIA
The pass/fail criteria selected will obviously be determined
by the monitoring technique chosen. Based on the test method
evaluation, the methods to be included in this discussion are the
explosimeter method, San Diego "Bag" method, the volume leakage
method, and the CARB pressure loss rate method.
The explosimeter tests reveal that an LEL limit of 1.0 LEL can
be used as a pass/fail value. Using the criteria of allowing no
leaks > 1.0 LEL, if a reading of 1.0 LEL was found then chances
would be very high (95.5%) that the tank would fail the CARB
criteria and not one that would pass. This method also would correctly
identify all tanks 95.1 percent of the time at the 1.0 LEL level.
Also, the method would exclude only a very small percentage (0
percent for the test fleet) of the tanks that were actually in
violation.
San Diego APCD personnel have selected pass/fail criteria based
upon the allowable pressure loss defined by the CARB certification
tests. The volume of vapor that would be lost due to a pressure
decrease from 18 inches of water to 17 inches of water is calculated
and a loss rate (over the five minute test) can be established. The
pass/fail criteria can then be established by determining the margin
of error that can be allowed in identifying all the violators. In San
Diego, because of the possible errors involved in the method, personnel
have selected a volume of vapor lost during loading operations of
3-43
-------
twice that allowed by the CARB criteria. SDAPCD has experienced a
100 percent success rate, using this enforcement technique
when a leak is found. There .is, however, no information on how many
violators are not detected by this method. The pass/fail criteria
for this screening method is therefore dependent upon the definition
of a leak tight truck.
The pass/fail criteria selected for the volume leakage test
will depend upon the allowable leakage that defines a leak tight
truck. The leak tightness of the tank can be defined as a percentage
of the total vapors transferred and an acceptable leakage rate can
be selected. For example, if a tank vapor volume of 5,000 gal
(670 ft ) and a loading rate of 500 gallons per minute (67 CFM)
is assumed, a pass/fail limit can be determined dependent upon the
vapor containment necessary. If a containment of 99 percent is required,
the allowable leakage can be calculated as follows:
(vapor volume flow rate) (1-containment required)
= Allowable volume leakage rate
For this example, the allowable volume leakage would be:
(67 ft3/min) (60 min/hr) (1-0.99) = 40.2 ft3/hr
By referring to Figure 3-7, this rate can be compared with the CARB
leak rate criteria.. For those trucks tested, the corresponding
pressure leak rate would range from 0.25 inches of water to 0.65
inches of water in 1 minute (or 1.25 inches of water to 3.25 inches
of water in 5 minutes). Table 3-10 shows several corresponding
volume leakage rates for various containment requirements, with
the corresponding pressure loss rate, based upon the field test data.
3-44
-------
Table 3-10. VOLUME LEAKAGE RATE REQUIRED FOR VARIOUS VAPOR CONTAINMENT REQUIREMENTS'
GO
en
Required
Vapor
Containment,
Percent
90
95
98
99
Volume
Leakage
Rate,
SCFH
402.0
201.0
80,4
40.2
Corresponding CARB Pressure Loss Rate, in H20/5 Minutes
9 in H20
28.3
14.3
5.5
3.3
12 in H20
23.0
11.5
4.3
2.5
15 in H20
20.3
10.3
3.8
2.3
18 in H20
23.0
10.8
3.0
1.3
Based upon a tank volume of 5,000 gallons and a loading rate of 500 GPM
See Figure 3-7, data based upon tanks tested during field test phase.
-------
References for Section 3.0
1. Control of Hydrocarbons From Tank Truck Gasoline Loading
Terminals, EPA Guideline Series, EPA-H50/2-77-026. October 1977.
2. Leak Testing of Gasoline Tank Trucks, Scott Environmental
Technology, EPA Contract No. 68-02-2813, Work Assignment No. 19,
August 1978 (Draft).
3. Personal communication with Dean Simeroth, California Resources
Board, Sacramento, California.
4. Ibid
3-46
-------
4.0 CONCLUSIONS
4.1 VAPOR CONTAINING EQUIPMENT AND MAINTENANCE
Vapor containing equipment is available to either eliminate or
minimize hydrocarbon vapor leakage from delivery tanks and from
bulk plant and terminal vapor piping. Test data indicates that
the type of equipment selected or age of equipment was not as much
a factor as the maintenance required to keep the equipment in good
working order. How long the equipment will remain controlling
vapors in a leak tight manner is unknown, but with a proper mainte-
nance program some tanks have shown leak tightness maintained for
over 4 months.
The CARB certification program specifies testing the trucks
annually. However, the program may be more meaningful if the
trucks were tested more often; e.g., either semi-annually or
quarterly. Some truck fleet operators stated that they would be
willing to test the tanks more often than the current annual test
but they felt also that the degree of leak tightness required by
the CARB program should be reduced.
4.2 COSTS OF MAINTAINING VAPOR TIGHT CONDITIONS
The normal annual cost for product delivery maintenance is
$1920 per truck. Annual recertification of trucks to a leak tight
condition would require an additional cost per truck of $130-$280
annually. The range of costs are due to the level of vapor tight-
ness desired (i.e. 3 inches of water in 5 minutes or 1 inch of
water in 5 minutes). If recertification were required semi-annually,
the additional costs would range from $270-$560 per year per truck.
The additional costs would range from $540-$1120 per year per truck
if recertification were required quarterly. The truck fleet
operators interviewed felt requiring certification more often than
4-1
-------
quarterly would not be realistic because of the time required to
certify a truck fleet and the scheduling and delivery adjustments
that would be required.
4.3 MONITORING PROCEDURES
Of the eight monitoring procedures studied, only four appear
to show promise as an acceptable procedure. These include the
explosimeter method, San Diego Bag method, the CARB pressure loss
rate method, and the volume leakage method. The other methods
studied include the quick leak decay method, sonic detector method,
bubble indication method, and the V/L ratio method. The quick leak
decay was evaluated using gasoline as the pressurizing liquid
because diesel is not available at all facilities. The quick leak
decay method would take too long to reach stabilizing pressures to
act as a quick monitoring method. The sonic detector indicated
the presence of leaks but because of the way the instrument was
used no relative sizing of the leaks could be obtained on a repeat-
able basis. The bubble indication method also worked very well at
identifying leakage areas but also could not be used to determine
the relative size of the hydrocarbon leaks. This method also often
detected insignificant leaks. The data for the V/L method indicates
that there was no correlation between V/L test results and the leak
tight condition of the truck,thereby eliminating it, as used, as a
possible compliance method.
Of the tests yielding acceptable results, either of the two
shop tests, the CARB test or the volume leakage test, could be
used as the compliance test. The volume leakage test is based
upon actual vapor volumes emitted to the atmosphere but this
determination requires slightly more time and equipment than the
CARB test. Either of the short monitoring methods, the explosimeter
test or San Diego Bag test could be used as an interim enforcement
or screening procedure. These short methods could be used to
4-2
-------
monitor loadings and determine violators, and in turn would
require the leaks to be minimized as specified by the shop test
selected.
4.4 PASS/FAIL CRITERIA
Pass/fail criteria for leaks emanating from the tank trucks
will depend upon the degree of leak tightness defined in the shop
test. Pass/fail criteria have been developed based upon the
existing CARB leak tight definition of allowing a pressure decrease
of 2 inches of water in 5 minutes (from 18 inches of water to 16
inches of water). San Diego has developed pass/fail criteria using
the bag method. A criteria could be developed using 1.0 LEL with
the explosimeter method. If another definition of leak tight were
developed using volume leakage rate criteria or pressure loss rate
criteria, pass/fail criteria for the screening methods would have to
be modified.
For other leakage areas not involving the delivery tank, pass/
fail criteria can be established. From vapor piping to the storage
tanks, no hydrocarbon leakage would be allowed to occur as indicated
with either an explosimeter, a sonic detector or a bubble indication
solution. Mo hydrocarbons as indicated by an explosimeter should leak
from a fixed roof storage tank pressure-vacuum vent if the storage
tank pressure is below the vent open setting. No leakage should
occur from vapor couplers as indicated by an explosimeter or bubble
indication solution.
4-3
-------
APPENDIX A
SUGGESTED MONTHLY VISUAL MAINTENANCE INSPECTION CHECKLIST
A-l
-------
APPENDIX A
SUGGESTED MONTHLY VISUAL MAINTENANCE INSPECTION CHECKLIST
TANK TRUCK OPERATOR — (To be performed on each truck)
1. Inspect hatch cover for integrity of gasket and seal surfaces.
The dome lid may be damaged or warped and should be checked
for the quality of the seal. Two methods are suggested. The
gasket or seal on the dome lid can be coated with a grease or
other easily visible material and the dome lid closed, sealed,
and reopened. The inability of the dome lid to close.or seal
around the entire circumference can then be clearly visible
by showing gaps in the indicating material on the mating
surface.
The other method, suggested by a tank truck operator, would be
to use a piece of thin paper placed between the dome lid and
the base ring with the dome lid closed securely. If the paper
can then be moved, the seal is not tight enough and a leak will
most likely occur.
The dome lid gasket should be visually checked for excessive
cracking or for debris or foreign material on the mating
surface. The gasket should be cleaned or replaced as required.
2. Inspect P-V.valve seals for debris or foreign material on seat.
Check valve operation to ensure valve will move smoothly with-
out sticking or rubbing and will reseat properly.
3. Inspect condition of the hatch base ring for severe damage or
warpage. The bolts and/or clamps used to attach the base ring
to the tank should also be tested for tightness.
4. Inspect condition of compartment vapor vent covers especially
if covers are the flanged bolted type or the rubber boot type.
Inspect flange bolts for tightness and/or inspect rubber boots
for cracks or tears.
A-2
-------
5. Inspect vapor recovery piping. If rubber hoses are used,
check the hoses for tears or cracks and check the tightness
of any connector clamps.
6. Inspect couplers for damage or wear which will not allow the
coupler to close properly. Check the coupler gaskets also for
excessive wear or damage.
7. Inspect vapor and liquid transfer hoses for cracks, tears or
excessive wear or damage. Check hose to coupler clamps for
tightness.
For all items above, repair or replace excessively worn or damaged
parts as required.
BULK PLANT AND TERMINAL OPERATORS
1. Inspect all gasoline delivery tanks as described above.
2. Inspect loading rack vapor and liquid couplers for signs of
wear, damage or liquid leakage. Check flexible vapor hoses
(if applicable) for cracks, tears, or damage. Check all hose
clamps for tightness.
3. Inspect above ground vapor piping using a bubble indicating
solution. Check all piping connections and joints with the
solution and look for signs of damage to the rigid piping.
4. For fixed roof storage tanks, inspect the condition of the
pressure-vacuum vents. Be sure that the valves are seated or
can be seated properly and that the valves can move freely in
the valve guides. Check to make sure all valve seats are
clean and free of debris.
5. For top loading arms, use an explosimeter or bubble indication
solutions to identify leaks at the loading arm swivels and
joints. Check any tapered rubber loading arm hatch sealing
mechanisms for signs of damage or excessive wear.
A-3
-------
Repair or replace damaged components as required.
SERVICE STATION OPERATORS
1. Inspect vapor couplers used for vapor transfer to underground
storage tanks. Check valve seats to make sure they are clean
and free of debris and foreign material.
2. Inspect vapor hoses (if applicable) as stated above.
A-4
-------
APPENDIX B
ACTUAL MAINTENANCE PERFORMED ON DELIVERY TANKS
DURING FIELD TEST PHASE
B-l
-------
APPENDIX B
ACTUAL MAINTENANCE PERFORMED ON SELECTED DELIVERY TANKS
DURING FIELD TEST PHASE
Truck/63806
Degassed, start 1030 - end 1145
Dome No. 1 - bad leak at lid seal
Dome No. 2 - smaller leak at lid seal
Dome No. 2 - high level shutoff also loose
(1 man -- 9 minutes)
Nos. 1 and 2 — straightened the hatch bases and adjusted the lid
spring tension. Tightened the high level shutoff (for bottom
loading)
Trailer/53306
Degassed, start 1200 - end 1415
No. 3 has bad leak at dome lid
(1 man -- 4 minutes)
Straightened base, adjusted spring tension and tightened base ring.
Truck/63765. Trailer/53304 6/23/78
Truck degassed, start 0615 - end 0720
Trailer degassed, start 0730 - end 0930
Truck - No maintenance
Trailer - No maintenance
Passed CARB test
Truck/63804
No. 2 dome lid leak - adjust tension
No. 3 dome lid leak - adjust tension and straightened hatch
(1 man -- 5 minutes)
Passed CARB test after maintenance
B-2
-------
Trailer/53307
No. 3 tightened base ring, replaced base ring and gasket
No. 1 dome lid leak - adjust tension
(1 man — 10 minutes)
Passed CARB test after maintenance
Trailer/53304 6/19/78
No. 3 dome lid leaking, cannot get over 1 inch FUO
(1 man -- 17 minutes)
Nos. 1 and 2 also leaking, adjust spring tension of lid
No. 3 straightened out hatch and adjust spring tension, replace
gasket in vapor recovery outlet
Passed CARB test after maintenance
Truck/63765 6/19/78
Nos. 1 and 2 dome lid leaking, straightened dome lids and adjusted
spring tension
(1 man -- 8 minutes)
Passed CARB test after maintenance
Truck/63804
Degassed 10 minutes for each compartment
No. 1 vent missing, replace
(1 man -- 5 minutes)
Only truck tank tested because of time limitations, truck only
degassed for 10 minute/compartment, complete stabilization not
achieved because of time limitation, vaccum test — 6.12 to 6.28
inches hLO fn 5 minutes -- invalid since not stabilized
Trailer/53297, Truck/63803
Degas, start 0630 - end 0805 (on trailer)
Degas, start 0815 - end 1050 (on truck)
Maintenance/trailer -- all three covers leaked, adjust tension
B-3
-------
and spring clamps, No. 3 has bent dome lid, also tightened high
level shutoff, start 0824 - end 0846
Maintenance/truck -- leak at hose clamp on vapor return hose,
replace clamp (1 man — 5 minutes), dome covers adjusted after EPA
Volume Leakage test
Passed CARB test after maintenance
Truck/63767
No. 1 cover leaking, upon adjusting spring tension a bolt broke,
must replace dome lid
(1 man -- 19 minutes)
Trailer/53345
No. 1 cover leaks, tighten base ring and adjust spring tension on
cover, (5 minutes — 1 man)
Blew the rubber boot off of No. 1 collector at 23 inches H20
(3 minutes -- 1 man)
No testing done pre-maintenance
Truck/68-275, Trailer/68-275
Maintenance for CARB Certification 9:10 to 10:30
Previous load diesel - old CB equipment
Bad leak compartment No. 3 vent, detectable by smell, ear, feel,
leak due to small 1/8 inch diameter rock in seal. Repaired by
cleaning vent/seal 9:10 to 9:30
Repressurize to find more leaks 9:30-9:40
18 inch to 15 inch in 5 minutes (1 inch too much drop), 6
medium/small leaks found, pressure and vacuum valves in each
compartment, repaired—started 9:40 to 10:00; pressure valves
seal, compartment No. 1, small 1/8 inch rock in seal, pressure
valves seal, compartment No. 3; replace both, all other 4 valves
cleaned
B.-4
-------
Repressurize to find more leaks, 10:00 to 10:10
18 inch to 15.1 inch in 5 minutes
Three small leaks found, compartment No. 1, pressure valve;
compartment No. 2, screw threads on vacuum valve; compartment
No. 3, vacuum valve still leaks
Repaired, start 10:10 to 10:20, replaced domes in compartments
Nos. 1 and 3 with brand new domes, leaked approximately 3 inches
in 3 minutes 10:20-10:30
Decision: Will fail CARB certification until new domes are
purchased and installed.
Note: Did not pass CARB test after maintenance completed;
Truck/67-182
Maintenance 1250 - 1350
Replaced vent dome cover Nos. 1,3, and 5 leak very badly and
Nos. 2 and 4 not as bad
1250 - maintenance started (2 minutes)
1350 - end maintenance
One of the vapor hoses has minute seeping leaks of the full length
of the base, new hose with a new material cover, many leaks but
total volume not significant
Passed CARB test after maintenance
B-5
-------
APPENDIX C
SUGGESTED ENFORCEMENT INSPECTION CHECKLIST
C-l
-------
APPENDIX C
SUGGESTED ENFORCEMENT INSPECTION CHECKLIST
GASOLINE DELIVERY TANKS — (Based on explosimeter method and 1.0 LEL
leak criteria)
While the delivery tank is being loaded, check the following
with an explosimeter:
• Dome lid/base ring interface
• Base ring/tank interface
• P-V vent
• Compartment vapor vent cover
• Vapor piping
• Vapor couplers
• Vapor transfer hoses
Any reading .>_l.O LEL constitutes a violation.
BULK PLANT AND TERMINALS
Check the following with an explosimeter:
• Vapor piping
• Vapor tight couplers
Any explosimeter reading constitutes a violation.
P-V vents on fixed roof storage tanks should be inspected to
determine if they are seated properly. Vent valves should be
tested to determine if they open and close smoothly.
SERVICE STATIONS
During gasoline transfers, vapor couplers should be checked
with an explosimeter. Any indication of a leak shall constitute
a violation.
C-2
-------
In all instances, the inspector should note the condition of
the vapor containing equipment and give suggestions on required
maintenance necessary to bring the violation into compliance.
C-3
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-450/3-79-018
2.
4. TITLE AND SUBTITLE
Evaluation of Vapor Leaks and Development c
Monitoring Procedures for Gasoline Tank Tn
Vapor Pioinci
7. AUTHOH(S)
Robert L. Norton
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
f April 1979
6. PERFORMING ORGANIZATION COOS
jcks and
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Pacific Environmental Services, Inc.
1930 14th Street
Santa Monica, California 90404
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standarc
Emission Standards and Engineering Divisior
Research Triangle Park, N.C. 27711
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-2606
Task No. 11
13. TYPE OF REPORT AND PERIOD COVERED
[S 14. SPONSORING AGENCY CODE
1
1S. SUPPLEMENTARY NOTES
EPA Project Officers: Nancy Mclaughlin, Emission Measurement Branch
Stephen A. Shedd, Chemical and Petroleum Branch
16. ABSTRACT
This technical document provides information on control techniques,
monitoring procedures, and costs for maintaining gasoline tank trucks and
vapor oiping in "vapor tight" conditions. The leak sources and evaluation
of different low cost and quick monitoring and test procedures are also
discussed. This document provides the support information for the Office
of Air Quality Planning and Standards guideline series document entitled
"Control of Volatile Organic Compound Leaks from Gasoline Tank Trucks and
Vapor Collection Systems," EPA-450/2-78-051 , December 1978.
17.
KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Air Pollution
Gasoline Vapor Control
Gasoline Tank Trucks
Bulk Plants
Service Stations
Bulk Terminals
Leak Monitoring
Systems
18. DISTRIBUTION STATEMENT
Unlimited
b.lOENTtFIERS/OPSN ENDED TERMS
Air Pollution Control
Stationary Sources
Mobile Sources
Hydrocarbon Emissions
19.' SECURITY CLASS (This Report!
Unclassified
20. SECURITY CLASS /This page>
Unclassified
c. COSATI Reid/Group
21. NO. OF ?AGSS
94
22. PRICE
SPA Form 2220-1 (R«». i-77)
"3SVIOUS iOITION U OBSOLETE
------- |