-------
GASOLINE TANK TRUCKS AND BULK PLANTS:
‘VALUATION OF VAPOR LEAKS AND DEVELOPMENT OF
MONITORING PROCEDURE
Contract No. 68—02—2606
Work Assignment No. 11
EPA Project Officer —- Nancy D. McLaughlin
Project Manager -- Robert Norton
September 1978
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Emission Measurement Branch
Emission Standards and Engineering Division
Research Triangle Park, North Carolina 27711
Pirifir ftrnnrnbr t ,I Cflr ,rrir ,& r
LJRAFT
-------
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—3
2.1.3 Service Stations 2-5
2.2 Potential and Controlled Emissions 2-5
2.3 Available Equipment to Control Emissions . . . . 2—6
2.3.1 Tank Truck Dome Covers 2-6
2.3.2 Tank Truck Vapor Collection Piping
and Internal Vents 2-10
2.3.3 Vapor Transfer Piping 2-12
2.3.4 Vapor Transfer Couplers 2-12
2.3.5 Storage Tank Pressure—Vacuum Relief Vents 2-13
2.4 Operating and Maintenance Procedures 2-14
2.4.1 Dome Covers 2-14
2.4.2 Vapor Collection Piping and Internal
Vents 2-17
2.4.3 Vapor Transfer Piping 2-17
2.4.4 Vapor Transfer Couplers 2-18
2.4.5 Storage Tank Pressure—Vacuum Relief Vents 2-19
2.4.6 Miscellaneous Emission Sources 2-19
2.5 Costs and Man—Hours Necessary to Maintain Vapor
Containing Equipment 2-20
2.5.1 Tank Trucks 2-20
2.5.2 Other Emission Sources 2—24
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
•1
-------
3.1.5 Pressure—Vacuum Test (GARB)
3.1.6 Bubble Indication Method.
3.1.7 Quick Leak Decay
3.1.8 Volume Leakage
3.2 Evaluation 0 f Test Procedures.
3.2.1 V/L Ratio Method
3.2.2 Explosimeter Method .
3.2.3 Sonic Detector
3.2.4 San Diego ‘ 1 Bag Method.
3.2.5 Pressure—Vacuum Test (CARB)
3.2.6 Bubble Indication Method.
3.2.7 Quick Leak Decay Method
3.2.8 Volume Leakage
3.3 Pass/Fail Criteria
Page
3-3
3-4
3-4
3—5
3—5
3-9
3-18
3—23
3—23
3-25
3—31
3—33
3-35
3—39
4-i
4-1
4-1
4-1
4-2
APPENDIX A - Suggested Monthly Visual Maintenance Inspec-
tion Checklist
APPENDIX B - Actual Maintenance Performed on Delivery
Tanks During Field Test Phase
APPENDIX C - Suggested Enforcement Inspection Checklist
A-i
B—i
C-i
Section
4.0 CONCLUSIONS
4.1 Vapor Containing Equipment and Maintenance
4.2 Costs of Maintaining Vapor Tight Conditions.
4.3 Monitorinc Procedures
4.4 Pass/Fail Criteria
11
-------
LIST OF ILLUSTRATIONS
Figure Page
2-1 C-B Dome Assembly 2-7
2-2 Tiona Dome Cover 2-8
2-3 Typical Vent Cover Configuration 2-Il
3-1 Frequency Distribution of V/L Ratio for Tanks That
Pass Certification Tests and Tanks That Fail
Certification Tests 3-17
3-2 Frequency of Occurrences of Back Pressure During
Loading Operations 3-26
3-3 Typical CARB Pressure Test Results at Bottom Loaded
Terminal 3-29
3-4 Typical CARB Pressure Test Results at Top Loaded
Terminal 3-30
3-5 Laboratory Test Results for CARB Pressure Test. . . . 3-32
3-6 Quick Leak Decay Test Apparatus 3-34
3-7 Typical Pressure Versus Time Curves for Laboratory
Tests of Quick Leak Decay Method 3-36
3—8 Volume Leakage vs Pressure Decay for Bottom Loaded
Tanks at Various Pressures 3—37
3-9 Volume Leakage vs Pressure Decay Rate for Top Loading
Tanks at Various Pressures 3-38
LIST OF TABLES
Table Page
2-1 Costs for Maintaining Trucks in Leak Tight Conditions 2-21
2-2 Total Annual Maintenance Costs for Product Delivery
Equipment 2-23
3-1 Test Fleet Physical. Data 3-7
3—2 V/L Results for Top Loading 3—10
3-3 V/L Results for Bottom Loading 3-13
3-4 Correlation Factors (r) for V/L Ratios and Tank Load-
ing Backpressure 3-19
3-5 Percent of Compartments With Leaks at Specific
Locations 3—21
3-6 Occurrence of Hydrocarbon Leakage at Various
Explosimeter Levels 3-22
3-7 Tank Tightness History 3-28
3-8 Correlation Coefficients for Volume Leakage Results
tIith Respect to GARB Test Results 3-40
3-9 Volume Leakage Rate Required for Various Vapor
Containment Requirements 3-43
•11 1
-------
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
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 special 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 necessary to maintain 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. i he field
test program \ conducted by another contractor4 s 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 tracks in a leak tight condition as
1—1
-------
defined by the California Air Resources Board (CARB) certification
criteria. Maintenance and equipment specifications were also
obtained mostly from California sources because of the leak tight
requirements. 1 , 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
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 open 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.
These vents are installed as a vapor control measure to reduce the
emission of hydrocarbons from the vapor space of the compartments.
2—1
-------
during transit. 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 would not allow the valve to seal
properly. The valve actuating devise, such as a spring loaded
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 compart-
ment 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. The vapor *‘ o :,
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.
2.1.1.3 Liquid and V por Transfer Hoses
Leaks can occur from liquid and vapor transfer hoses and from
their respective couplers. Hoses can become torn, worn, cracked,
etc. to produce hydrocarbon vapor leaks. Fugitive hydrocarbons can
2—2
-------
occur from vapor coupler connections if these are not coupled or
closed properlye 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 frequent than those already
discussed. Leakage can occur from flaws in the tank shells,
improperly welded seams, or improperly installed overfill
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 (if applicable).
- S p
2.1.2.1 Vapor Piping to Storage Tanks
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 storage tanks to the loading rack. This piping is
usually above ground and is normally flanged or threaded metal
pipe. Hydrocarbon vapor leaks can occur at piping joints or
connections dur to improper installation, faulty flange gaskets) or
accidental damage.
2—3
-------
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 imporperly 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 loading 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 loading operations. 2
2.1.2.4 Storage Tank Pressure Relief Vents
At either bulk plants or terminals where fixed roof tanks are
employed, pressure-vacuum vents are used to control breathing and
working 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.
The spring or weights system may get out of allignment and not
allow the valve to return to its seat properly, thereby causing
leaks.
2-4
-------
2.1.3 SERVICES 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. 3
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 under-
ground 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.
2.2 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 equilavent orifice size of
the leak opening and the tank pressure. 4 All of the leakage
sources prescribed in Section 2.1 have the potential to be large
2—5
-------
leaks, however, some of the leak sources are normally more
predominate than others.
Leakage of hydrocarbons from hatch covers, hatch base ring 5
and pressure vacuum vents are the sources where hydrocarbon vapors
most often occur. These uncontrolled leakage rates have the
potential to exceed 10 percent of the vapor transferred. 5 Under
controlled conditions these sources should not leak in excess of 1
percent of the volume of vapor transferred. This is based on the
CARB 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 va or piping,
cou lers 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. Yapor 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
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
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—1 and 2—2). This added support makes
the C—B dome less apt to succomb to bends or warpage. However, the
2—6
-------
sc* i5
Figure 2-1. C-B Dome Assembly
(Courtesy C-B Equipment, Lynwood, California)
I
2
3
4
I
S
I . ’
*
‘S
7’
7’
I’
I ,
Is
a’
U
53
‘4
‘S
a’
21
‘I
tool.
d3
“I,
“is
4a•
4
14 4%
‘4”
14kg
‘4’,
54 5
,4,t
14 %
4 74
.74 1
. 7’,’
. 74 . 71
‘a . ,’
£ e4
£0’
. 74 4t
.74—.
a
S
S
a
4
4
2
t
S
a
a
2 .
4
4
4
U..l4MG
ISIL SIlK
“It’,
bOM cap
bows — t
5 4?, 5$$S S V
L*C t P 5.111 I ’Il 5
s”v’.
wIw4 515.
ML rru
Dowi. ct. w. it
I..OC II* I4DI. .L
L &tIlc bG T
WIIIGI. SOi.?
**
cAP
‘ gcww V*I. 1 )4IC
‘w ’
W.CIØ.)M VL SItING
‘*.MwsItt.Vt DuG
,5uII7s& lN V1 $41.7.
,t as,ts %11I. $41.1. 51. .
pt7.Isut1. vu. I
PflhllItI
C
Dewl C7. w’ & $&Du .
2—7
4-A
-------
No. Description
Material
Part No.
Fill Cover
Aluminum
8046-AL
2
—
Fill Gasket
Buna ’N
3 1 19-BN
Viton
3119-VT
Bolt -Spring
1 Spring
Steel Cd. Pt.
Van. Steel
3139 -CF
3129 -CF
S
Bolt Clamp
Steel Cd. Pt.
3029-CF
Stls 304
3029-SL
6
Nut -Clamp
Brass
3030-BR
Stls 304
3030 -SL
- Clamp
I
B Clamping Ring
Steel
3031 -MS
StIs 304
Steel
3031 -SL
3036-MS
Steel Cd. Pt.
3036-CF
StIs 304 1 3036-SL
No
9
Description [ Material
Part No.
fr nhoIeGasket
Channel Type
C
Cork-Buna.N
3175 -CS
B
A
Buna-N
3 175-SN
Asbestos
3175-AS
E
Teflon-Asb.
3175-AT
10
ManholeCollar
(Channel Type)
.
.
E
) . .
vi
F
Steel
3176 -MS
¼ Aluminum
3 176-AE
304
3176 -SL
K
StIs 316
3l76-S$
I I
Closure Assem
(Includes Cover
Plate. Strong
Back, Latch,
Hinge Lugs &
Pins)
S
Steel
6083-MS
B
—
Steel Cd. Pt
6083-CP
C
Stls 304
6083-SP
C
StIs 304
6083-SL
DStIs 316
6083-SS
Latch and Strongback Steel Cadmium Plated.
Figure 2-2. Tiona Dome Cover
(Courtesy of Tiona-Betts, Warren, Pennsylvania)
I — 20/4 O.D.
2-8
-------
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 dome can be worked and manipulated to
retain vapor tightness, where as the entire C-B dome, if damaged,
must be replaced. This can be significant when considering the
cost of the dome covers (Tiona approximately $80; C—B dome approxi-
mately $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 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 directon 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 pistàn 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 contains the liquid and will not allow it to
escape from its container. One operator has devised a conversion
kit made so that the C-B vent can be installed on the Tiona
dome.
2—9
-------
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 CARS 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. Leaks can be detected by using a bubble indicating
solution or so nic detector. The internal vent allows vapors to
enter the vapor return system when loading or unloading liquid into
the compartment and are pneumatically coupled with the compartment
loading. The vent opens into a covered area which in turn is piped
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—3). The vapor
collection or return piping can also be made of metal or rubber and
can take severall 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—10
-------
/ fl.. .* • ,4 #*w$ d
4.* /..cA
m.—’ r,./ ‘ , ,‘#.*‘.* j
1 n .h *
øi .ui/. o. — 4
,_ ,,
O . ..c. AL.i
, ,A, 4 ,, — . ...4y.
144• 4 *4’ •
*.,., ‘d i.. .,4 .,..1
I
Cvi i I
Figure 2-3. Typical Vent Cover Configuration
t
U
-J
—I
Vifta? / g(pr’Rr Oi.-r44
I.
I IfrOA’ A’i(OIify lvii i 2 1’4/L
I ’
I
I• I.
W,t#i Our V4i’pvy (‘E v(1Yj
a —,,
A..f/ i(g
‘J /.c 4 •r
,4, f _Ia_v C 4
I ••.•
-------
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 but 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—12
-------
couplers become leak tight through a compression mechanism incor-
porating a gasket material chemically resistant to the liquid being
transfered. 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 work or damaged or the coupler connectors
or body becomes 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 pressusre 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—13
-------
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 on recomended maintenance beyond the fact that damaged
parts should be replaced. However, several of the operators
interviewed had devised their own operating or maintenance
procedures and these are discussed in this section, along with
reconinended maintenance procedures for maintenance not currently
conducted. A visual inspection of the vapor containing equipment
is an integral part 0 f the maintenance program. A suggested
checklist for a periodic visual inspection of eq ioment 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
Dome cover maintenance procedures range from visual
observation to severe manual adjustments. Maintenance practices
vary greatly between operators from nearly nonexistent to monthly
inspections. For pruposes 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. Comon causes for leakage around the
domes, as discussed in Section 2.1, can be caused by 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 shows signs of excessive
2—14
-------
wear or damage, it should be replaced. The dome lid 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 or 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 circumferrence
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 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 should 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—15
-------
a leak occurs between the dome lid and base ring, bending or
reshaping of the dome base can be done by pounding with hamers or
applying leverage to the hatch opening in an attempt to create a
good seal at the dome lid. The dame lid itself can be adusted 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 dissambling 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 th 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 or at least
every two months. This should not require pressurization of the
tank but only replacement of visably damaged or faulty equipment.
Some operators perform visual inspections as often as once every
two to three weeks. 6
/
J 2—16
-------
2.4.2 VAPOR COLLECTION PIPING AND INTERNAL VENTS
Vent valve covers and vapor piping joints should be visually
checked with 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 detector or explosimeters can be used to pinpoint
the hydrocarbon vapor emission 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 occurs at bolted covers, the bolts should be
checked for tightness. If the 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 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 installation,
or loosened flange clamping mechanisms. The piping, where above
ground, should be visibly inspected for damage or obvious leakage
2-17
-------
areas. An explosimeter, sonic detector or bubble indicating
solution can be used to find smaller leaks. For 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 canndPbe 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 can be applied to the
hose 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. ,,
• t
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 work, 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.
2-18
-------
2.4.5 STORAGE TANK PRESSURE-VACUUM RELIEF VENTS
As in other pressure-vacuum reflief vents, the most common
leakage point would be around valve seats. Dirt or other foreign
debris can become lodged on the valve seat face causing the valve
to close incompletely and vapors to escape. Valves can also have a
problem of reseating improperly once they have opened. The closing
disc can get out of allignment 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 potentially 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 too much of a
burnden since most fixed roof tanks are gaged for liquid level from
the toj> ,wheneyer liquid deliyeries^arf trjade.
2.4.i rtisCEllANEduS EMISSION' SOURCES '
This catagory includes leaks in tank shells, poorly welded
seams, damage caused by an accident, or poorly installed 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,
normally when performing a tank pressurization test. Probability
of a leak occuring at other sources is so much higher that these
sources are usually the last checked.
2-19
-------
2.5 COSTS AND MAN—HOURS NECESSARY TO MAINTAIN VAPOR CONTAINING
EQU I PMENT
2.5.1 TANK TRUCKS
The costs presented here are for maintaining a tank in a leak
tight condition. These costs include performing the required main-
tenance to reduce leakage to an acceptable level, supplying the
necessary replacement materials ,and performing the required CARB
test to verify the tank leakage integrity.
The costs have been divided into two categories. The first
category deals with the cost of maintenance and equiment to ini-
tially bring an existing tank truck into the limits of the speci-
fied vapor tightness. These costs are generally higher because
there may be many leak sources, which after being maintained ini-
tially, do not require maintenance at every succeeding certifcation
test. The second category deals with the cost of maintaining a
truck within specified vapor tightness limits which has previously
been certified. The costs are also given for two degrees of vapor
tightness. The first, or more stringent case, deals with San Diego
County which allowed a pressure drop of 1 inch in 5 minutes. The
second case was that required in the remainder of the State of
California, which allowed, at the time these costs were generated,
a pressure drop of 3 inches in 5 minutes. The costs are shown in
Table 2-1.
The costs for the more stringent case are three to four times
greater because as the allowable leak rate becomes smaller the
significance of smaller leaks increases. Additional man—hours must
be spent to identify and repair the smaller leaks.
Currently these maintenance procedures and certification tests
are performed on an annual basis. Several operators indicated that
visual observations and minor maintenance is performed on the tanks
between annual certifications. This maintenance is usually
2—20
-------
Table 2-1. COSTS FOR MAINTAINING TRUCKS IN LEAK TIGHT CONDITION?,b
Labor
Hours
Labor
$C
Materials
$
Total Cost
$
San Diegod
Initiale
Retest
Initial
Retest
34
11
8
3.5
748
242
176
77
30
20
20
I 20
778
262
196
97
aCosts obtained from John Snyder, Chevron, USA From A Presenta-
tion to California Air Resources Board, December 2, 1976, and
from Larry Cowie, Shell Oil, from file data on actual main-
tenance performed.
bLeak tight conditions specified as passing certification tests.
CLabor rate = $22/hr.
dSan Diego tests allow leak rate of 1 in H 2 0/5 minutes.
e lnitial = First certification
Retest = Any certification following the first.
9 California tests allow leak rate 3 in H 2 0/5 minutes (1976).
2-21
-------
coordinated with the normal truck power unit maintenance schedule
which is performed approximately every 4 to 6 weeks. This periodic
maintenance normally costs approximately $50 to $100 per occurrence
depending upon the amount of replacement parts required. The aver-
age cost is about $70 per occurrence. 7
Estimates were received on the total tank maintenance required
on the tank truck product handling equipment. This included main-
tenance 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 $158/month. These costs are actual
monthly average costs based upon over 100 months of actual mainten-
ance performed. The costs to maintain the truck in a vapor tight
condition would be about 45 percent (70/158) of the total
maintenance costs required monthjly on the product delivery equip-
ment (this does not include maintenance costs for the power unit).
The total annual maintenance required on the tank equipment would
be $1,896 per year.
Incorporating these figures 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 ofter 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). Also shown in the table are
the annual costs expressed as a percentage of original purchase
price (assumed $30,000 for tank product delivery equipment). As
the maintenance requirements increase so does the percentage of
2-22
-------
Table 2-2. TOTAL ANNUAL MAINTENANCE CQSTS FOR PRODUCT DELIVERY
EQUIpMENTa ,u
Monthly
Costs
Total Annual
Costs
Per:ent of
Purchase Price
Original
per YearC
$
$
%
1. Product Delivery
Maintenance 88 1,056 3.5
2. Product Delivery
and Vapor Tightness
Maintenance 158 1 ,896 6.3
3. Product Delivery
and Vapor Tightness
Maintenance and
Annual Recertifica-
tion maintenance
and testing
San Diego 2,158 7.2
California 1,993 6.6
4. Product Delivery
and Vapor Tightness
Maintenance and
Semi-Annual Re-
certification
• Maintenance and
• Testing
San Diego 2,420 8.1
California 2,090 7.0
5. Product Delivery
and Vapor Tightness
Maintenance and
Quarterly Recerti-
fication Mainte-
nance and Testing
San Diego 2,944 9.8
California 2,284 1 7.6
aproduct delivery equipment includes delivery tank, couplers, internal
valves, vapor recovery requipment, overfill protection, and dome covers.
bAll costs are averages given on a per truck basis.
Coriginal purchase price of product delivery equipment estimated at
$30,000.
2-23
-------
the purchase price, until at the maximum rate the percentage
reached approximately 10 percent of the original purchase price per
year. Even at this upper level, this is not an unreasonable rate.
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
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. Like in the tank truck maintenance, the
ir,ital 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 use for the truck maintenance is used, the inspection should
not cost more than $44 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 mainten-
ance would mostly 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 on a similar scale to that of
tank trucks ($20), the total cost of this maintenance would be $108
per occurrence. If this maintenance were required on a monthly
basis, this cost would not be unreasonbie.
2-24
-------
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. I asselman, Reliability Observations and
Emission Measurements at Gasoline Transfer Vapor Recovery
Systems . TRW, Inc., EPA Contract No. 68—02—0235, Novemer 1974.
4 Letter from R.A. Nichols Engineering to H.B. tihlig, Chevron
U.S.A., June 10, 1977.
5. Ibid.
6. Presentation by John Snyder, Chevron U.S.A. to California Air
Resources Board, December 2, 1976.
7. Ibid.
2—25
-------
3.0 DEVELOPMENT OF MONITORING PROCEDURE
3.1 TEST METHODS
Several test methods were explored as to there acceptability
or usability as a monitoring procedure within the confines of this
task. A test method was sought that would give pass—fail compli-
ance 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 ex-
hausted 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 temperatue can also be monitored during
transfers. The V/L ratio is a simple volume ratio without
corrections. However the additional physical data obtained can be
used to explain some phenomenon which take place. This method has
been used before in conjunction with EPA mass emission
determinations at both bulk plants and terminals. The EPA method
called for obtaining a leak tight truck and determining the V/L
ratio. This V/L ratio was then compared to all the other trucks
checked during the test period. The leak tight trucks were used to
determine a baseline for comparison of the other trucks tested.
3—1
-------
3.1.2 EXPLOSIMETER
The explósimeter 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 rela-
tive 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 soinic 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 would 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 1120 in 5
3—2
-------
minutes). The bags are oversized so that on filling, San Diego
County inspectors are certain that a violation has taken place.
The bag is placed over the compartment which is being loaded and
the number of times the bag fills or the approximate volume of
vapors collected in the bag are estimted.
3.1.5 PRESSURE-VACUUM TEST (GARB)
The California Air Resources Board (CARB) has passed regula-
tions which define the degree of tightness that is required on
gasoline delivery tanks. To ensure that this tightness is main-
tained, all trucks must pass a pressure tightness test each year.
A test procedure was derived by the CARS which would be used to
test the trucks as to their tightness. 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 for 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
3-3
-------
time is then monitored and checked against the allowable leakage
rate. The current allowable leakage rate is 2 inches of water in 5
minutes (from 18 to 16 inches of water). The truck is then placed
under vacuum, most corrinonly using the vacuum supplied by the ex-
haust 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 is currently from 6 inches of water vac-
uum to 4 inches of water vacuum. Many of these other test methods
discussed are based upon estmating the amount of leakage that is
allowed or specified by the GARB certification test procedures.
3.1.6 BUBBLE INDICATION METHOD
This test method employs the use of a soap solution or other
solution whch will indicate gas leakage by the forming of bubbles
around the leakage area. The solution is applied to hoses, couler
interfaces, hatch covers and pressure vacuum vents and the appear-
ance of bubbles indicates a leakage source.
3.1.7 QUICK LEAK DECAY
The quick leak decay method is similar in concept to the CARB
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 to empty) and gasoline allowed to flow out until the
desired vacuum is reached. This time the increase in pressure (or
3—4
-------
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 GARB test methods.
3.1.8 VOLUME LEAKAGE
The volume leakage method maintains a constant pressure in the
test compartment by continually introducing air into the
compartment. 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 GARB 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 contract. These included the
explosimeter method, sonic detector method, GARB method, V/L
method, bubble indication method, and the volume leakage test. The
quick leak decay method was analyzed under laboratory conditions
and the San Diego “bag” test method was observed in the field, as
performed by San Diego County personnel.
Both a top and bottom loaded 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 loaded 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.
Initial data indicated that the age of the tank and its correspond-
3—5
-------
ing 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 emloyed, 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 other-
hand 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.
The test programs at both the top and bottom loaded 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
monitoring for leaks before and after shop tests and maintenance
were performed giving 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 were necessary. At the bottom loaded
terminal the trucks were scheduled to haul a load of diesel before
being tested. At the top loading terminal diesel was not avail-
able, so compartments were purged with air.
3-6
-------
Table 3—1. TEST FLEET PHYSICAL DATA
Too loading
rank
No. of
10•
iber
.
Age
a
Compart—
ments
Shell
Capacity,
Ga llons
Liquid
Capacity,
Gallons
Hatch
Type
Suspen—
sion
Vapor Recovery Type
Vent Cover
Line
Vapor
13806
53306
13767
3256
13766
i3345
3765
3304
t3804
3307
13803
53297
3805
53305
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
2
3
2
3
2
3
2
3
2
3
2
3
2
3
4408
5159
4408
5380
4408
5315
4408
5159
4408
5159
4408
5159
4408
5159
4000
4800
4000
No
Dat a
4000
4900
4000
4800
4000
4800
4000
4800
4000
4800
No
Data
No
Data
Tiona
Tiona
Tiona
Tiona
C—B
2—C-B/
1—Tiona
Tiona
Ti ona
Ti ona
Tiona
Tiona
Tiona
Ai r
Spring
Air
Spring
Ai r
Spring
Ai r
Spring
Ri r
Spring
Air
Spring
Air
Spring
No
Data
No
Dat a
Welded
Welded
Welded
Rubber
B oat
Bolted
Cover
Bolted
Cover
Bolted
Cover
Bolted
Cover
Bolted
Cover
Bolted
Cover
Bolted
Cover
Bolted
Cover
No
Dat a
No
Dat a
Welded
Welded
Welded
Rubber
Hose
Cl amped
& Welded
Pipe
Cl amped
& Welded
Pipe
Welded
Pipe
We I ded
Pipe
Welded
Pipe
Welded
Pipe
Welded
Pipe
Welded
Pipe
3—7
-------
Table 3-1. TEST FLEET PHYSICAL DATA CONCLUDED)
Bottom
.oadi nq
T
1.0.
Number
a
Age
No. of
Compart
ments
Shell
Capacity
Gallons
Liquid
Capacity
Gallons
Hatch
Type
Suspen—
sion
Vapor Recovery Type
Vent Covei
Vapor Line
67-J75
6/67
5
8650
8150
Tiona
Air
Welded
Welded
67-182
4/71
5
8650
8250
Tiona
Air
Welded
Welded
67-392
4/73
5
8650
8050
C—B
Air
Bolted
Rubber
Hose
67-475
10/74
4
8650
8200
Tiona
Air
Welded
Welded
68-795
4/78
2
5097
4600
C—B
Spring
Welded
Welded
68_795*
4/78
2
5319
4350
C—B
Spring
Welded
Welded
68—597
9/75
2
5087
4300
Tiona
Air
Welded
Welded
68_597*
9/75
2
5327
4550
Tiona
Air
Welded
Welded
68-275
11/62
2
4447
3670
C—B
Spring
Bolted
Rubber
Hose
68_275*
11/62
2
5053
4400
C-B
Spring
Bolted
Rubber
Hose
68—377
8/69
3
5184
4000
Tlona
Spring
Welded
Welded
68-977
8/69
3
5180
4750
Tiona
Spring
Welded
Welded
* Trailer
a Tank age indicates year put into service
3-8
-------
In the shop, equipment was arranged and a volume leakage test
was conducted on the test tank before any mat ntenance 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 estab-
lish a leak rate before any higher pressure might Nblow* 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 CARRB leak rate criteria. A CARS pressure vacuum test and
volume leak rate test were then performed again.
3.2.1 V/I 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 severely effect
the V/I ratio can be expected on a leak tight truck. Table
3—2 and Table 3—3 indicate the results of the V/I tests for both
top and bottom loading and presents the V/L ratio for trucks that
passed the GARB certification tests and those that failed.
The EPA terminal tests using the V/I method prescribe
determining the V/I 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—2 and 3—3 indicate a wide
variability in the V/I ratio for both the tanks that pass and the
tanks that fail the GARB reference test. A frequency distribu-
tion of the V/I ratios for both the tanks that passed and the tanks
that failed is shown in Figure 3—1. As indicated by this figure,
the V/L ratio takes the same frequency of occurrence regardless if
the tank passed or failed the certification tests.
3—9
-------
Table 3—2. V/L RESULTS FOR TOP LOADING
Test
No.
Date
Temperature
Back
Pressure
(inches 1120)
V/L Ratio
Passd Failb
Air Vapor
°F
1 6/19 1.12 80 90 9.25
2 6/19 1.16 80 92 7.5
3 6/19 1.39 80 92 9.8
4 6/19 0.99 80 92 15.0
5 6/19 1.18 80 92 10.7
6 6/19 1.34 80 93 10.8
7 6/19 1.14 80 90 7.8
8 6/20 1.05 66 80 8.0
9 6/20 1.06 70 80 5.6
10 6/20 1.11 70 83 9.8
11 6/20 1.04 76 88 9.7
12 6/20 1.56 80 90 15.6
13 6/20 1.08 80 92 1.1.1
14 6/20 1.05 78 90 6.0
15 6/20 1.20 78 90 12.3
16 6/21 1.02 68 76 10.3
17 6/21 1.43 68 80 9.1
18 6/21 0.99 68 80 8.9
19 6/21 No data 68
20 6/21 No data 70
21 6/21 1.08 74 90 11.0
22 6/22 1.18 74 90 13.0
23 6/21 1.08 80 90 12.2
24 6/21 1.07 80 92 5.1
25 6/21 0.70 80 94 11.5
26 6/21 1.39 80 90 12.6
27 6/22 1.05 70 75 5.9
3-10
-------
Table 3—2. V/L RESULTS FOR TOP LOADING (CONCLUDED)
Test
No.
Date
Temperature
Back
Pressure
(inches 1120)
V/L Ratio
Passd Fail 1 ’ —
Air Vapor
°F °F
28 6/22 0.95 70 80 9.3
29 6/22 No data 70
30 6/22 1.07 70 82 10.8
31 6/22 1.01 72 80 10.6
32 6/22 1.49 78 83 9.0
33 6/22 1.05 80 89 12.8
34 6/22 No data 80
35 6/22 1.15 80 88 8.9
36 6/22 No data 81
37 6/22 1.42 82 94 12.7
38 6/22 1.47 82 92 10.9
39 6/22 No data 82
40 6/22 No data 72
41 6/22 1.06 72 90 10.5
42 6/22 1.06 77 86 8.6
43 6/22 0.80 77 86 10.5
44 6/22 1.22 77 86 8.5
45 6/22 0.80 78 86 6.7
46 6/23 1.06 80 88 10.8
47 6/23 No data 80
48 6/23 0.71 84 92 6.6
49 6/23 1.92 88 96 6.3
50 6/23 1.09 88 96 8.2
51 6/23 1.01 84 87 9.2
52 6/23 1.04 84 85 8.9
53 6/23 1.12 78 88 5.2
3-11
-------
Notes for Tabi e 3-2
a Pass indicates a tank that will meet the CARS leak tight
criteria.
b Fail indicates a tank that leaks greater than the allowable
rate defined by the CARB leak tight criteria.
C All trucks tested were truck and trailer units but a V/I
•ratio could not be obtained for each tank. Tanks were loaded
simultaneously and all loading arms manifolded together
before the vapor meter.
3-12
-------
Table 3—3. V/L RESULTS FOR BOTTOM LOADING
Test
No.
Date
Temperature
Back
Pressure
(inches 1120)
V/L Ratio
Passd Failli
Air Vapor
°F
1 6/12 1.23 98 11.5
2 6/12 0.86 98 4.4
3 6/12 0.97 101 5.5
4 6/12 0.55/ 100/ 4.1/
110 4.8
5 6/12 0.85/ 108 1.7/
0.75 2.4
6 6/12 0.84/ 108 9.5/
0.95 7.1
7 6/12 1.02 110 7.3
8 6/12 0.96/ 106/ 7.5/
0.63 103 4.9
9 6/13 1.0]. 78 2.7
10 6/13 0.94 79 3.0
11 6/13 1.06 98 5.1
12 6/13 1.02 100 5.5
13 6/13 0.89 90 6.9
14 6/13 0.99 96 15.8
15 6/13 1.26/ 110/ 6.1/
1.02 102 6.3
16 6/13 0.79 95 108 3.4
17 6/13 0.96 95 106 8.9
18 6/13 0.98 95 110 10.1
19 6/13 1.11/ 93 102/ 10.2/
0.78 103 4.1
20 6/13 0.61 91. 100 4.3
21 6/13 0.37/ 91 100/ 3.8/
1e06 103 2.6
22 6/13 0.88 90 100 9.1
23 6/13 0.84/ 90 100 3.8/
0.80 6.0
3—13
-------
Table 3—3. V/L RESULTS FOR BOTTOM LOADING (CONTINUED)
Temperature
Back
Test
No.
Date
V/I Ratio
Passd FailU
of_____
Air Vapor
Pressure
(inches 1120
°F
°F
24 6/14 0.71/ 68 78 17.5/
0.82 16.2
25 6/14 0.84 71 83 3.0
26 6/14 1.15 73 80 6.6
27 6/14 0.69 76 84 3.0
28 6/14 0.87/ 78 89/ 2.7/
1.00 92 2.3
29 6/14 0.83 79 85 4.9
30 6/14 0.93/ 85 90 7.7/
0.67 8.1
31 6/14 0.97 86 90 10.7
32 6/14 0.67 87 100 10.0
33 6/14 0.80/ 88 92/ 4.0/
0.91 102 3.0
34 6/14 0.88/ 87 98/ 9.4/
1.12 104 7.1
35 6/14 0.99 88 96 4.0
36 6/14 0.91 88 100 8.9
37 6/14 0.92/ 87 100/ 4.3/
0.72 90 5.6
38 6/15 1.07 67 75 6.4
39 6/15 0.80 67 70 7.4
40 6/15 0.97 69 76 6.2
41 6/15 0.80/ 71 82/ 8.0/
0.75 80 7.5
42 6/15 0.78 73 86 11.0
43 6/15 0.99 74 83 7.1
44 6/15 1.08 77 90 4.1
45 6/15 0.83/ 79 90/ 6.5
0.97 92 4.3
46 6/15 0.92 83 98 5.4
3—14
-------
Table 3—3. V/L RESULTS FOR BOTTOM LOADING (CONCLUDED)
Test
No.
Date
Temperature
Back
Pressure
(inches 1f 2 0)
V/L Ratio
Passe Failb
of
Air Vapor
°F °F
47 6/15 0.96 83 98 5.4
48 6/15 0.83/ 86 103/ 4.1/
0.70 90 8.3
49 6/15 0.84/ 84 90/ 3.5/
0.88 104 4.0
50 6/15 0.96/ 86 90/ 3.4/
0.93 107 3.4
51 6/15 0.89 86 90 4.7
52 6/15 1.03 87 92 4.7
53 6/15 0.81 86 103 4.1
54 6/15 0.90/ 87 102/ 6.2/
1.04 87 7.3
55 6/16 1.00 67 80 10.5
56 6/16 1.06 70 6.9
57 6/16 0.91/ 72. 72/ 6.9/
0.74 84 4.6
58 6/16 0.99 78 76 12.0
59 6/16 0.88 79 78 3.9
60 6/16 1.02 80 88 3.7
61 6/16 1.52 84 90 4.7
62 6/16 1.07 85 97 6.2
63 6/16 1.04/ 85 100/ 5.5/
1.17 93 6.9
64 6/16 1.12/ 87 79/ 7.5/
1.14 81 9.4
65 6/16 1.03/ 85 101/ 8.0f
0.98 103 6.8
66 6/16 0.95 85 95 3.5.
67 6/16 0.58/ 84 100/ 3.0/
0.88 92 2.3
3—15
-------
Notes for Table 3-3
a Pass indicates a tank that will meet the CARS leak tight
criteria.
b Fail indicates a tank that leaks greater than the allowable
rate defined by the GARB leak tight criteria.
C Indicates the truck tested was a truck and trailer unit.
Data is presented for each tank (truck/trailer).
3-16
-------
______ Pass
Fall
.1 .2 .3 .4 .5 .6 .7 .8 .9 1.0 1.1 1.2
V/I Ratio
FIgure 3-1. Frequency Distribution of V/I Ratio for Tanks That Pass
Certification Tests and Tanks That Fail Certification Tests
1.3 1.4 1.5 1.6
U
I’
1
I
I
25
20
15
10
‘I . .
0 ,
I - )
0,
L
U
U
0
F-
(4 ) 10
4 )
-4 0
—J I-
9-
0
4 - I
0,
0)
0
/p\
5
/
/
1
0
/
/
/
-------
A relationship was attempted 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 presssure. The correlation factors are shown
in Table 3—4, and as indicated there is no correlation between V/L
ratios and back pressure.
In suninary, the V/L ratio showed no relationship between vapor
tight tanks or tanks that leaked. The VIL ratio also showed no
relationship when compared to the back pressure experienced during
the loading operation.
3.2.2 EXPLOSIMETER METHOD
The explosimeter method was extensively tested in the field
test program. The CARS 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. 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, which was calibrated against propane, were then
compared to the CARB results to determine the types 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 loaded facilities.
The results were much more consistant and meaningful at the
bottom loaded terminal than at the top loaded terminal. At the top
loading terminal, since the hatch cover is not closed, the data
does not reflect the relative vapor tightness of the tank which is
traceable back to the GARB pressure test. Instead, the loading
3-18
-------
Table 3-4. CflRRELATION FACTORS (r) FOR V/L RATIOS
AND TANK LOADING BACKPRESSURE
Loading
Pass/Fail
r
Top
Pass
0.49
Top
Fail
—0.02
Bottom
Pass
0.04
Bottom
Fail
—0.02
Combined
Pass
0.03
Combined
Fail
—0.07
3-19
-------
arm/hatch interface, loading arm connections and swivels become the
major sources of hydrocarbon leakage.
Tables 3-5 and 3—6 show the results of the explosimeter tests
performed at the loading terminals. When discussing the usefulness
of this test method as a means of performing tank vapor tightness
the discussion will be limited to the bottom loaded data. Leaks
occurred at almost every top loading occurrence regardless if the
tank had passed or failed the CARB certification test. The data
from the bottom loading terminal indicates that the tanks that fail
the CARB certification test show considerably more leaks 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 hatches indicating this relative size of leak were
those that exceeded the allowable leakage criteria. At the 1.0 LEL
level, over 95 percent of the hatches indicating this relative size
of leak failed the certification tests.
The explosimeter method at the 0.5 LEL level or the 1.0 LEL
level is an acceptable method to determine the compliance of the
tank vehicle based on the on the CARB certification. The data was
more meanigful 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, It was just 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
useable, are based on 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.
3-20
-------
Table 3-5.PERCENT OF COMPARTMENTS WITH LEAKS AT SPECIFIC LOCATIONS
Top
Hatcha
Base
Ring
P-V
Vent
Loading
Arm Vapor
Connectors
Vapor
Collector
Cover
Top Loading
passb
993
85.9
N.A.
43.0
40.8
FaliC
94.9
82.2
N.A.
33.1
34.7
97.3
84.2
N.A.
38.5
38.].
Bottom Loading
passb
5.2
8.6
7.8
N.A.
N.L 9
Fa 1 1C
42.3
42.3
48.7
N.A.
N.L.
Totald e
22.7
22.2
?4.2
N.A.
N.L.
a Top Hatch top loading indicates the loading arm/hatch interface
b Pass indicates a tank that is leak tight as defined by the CARS
criteria
C Fail indicates a tank that has failed the CARS test and
therefore leaks in excess of the allowable criteria
d Total Percent of all compartments where leak occurred at
specific location
e 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 118 in the fail
mode
NA. = Not applicable since compartment loaded with hatch open
g N.i. = No leaks detected
3—21
-------
Table 3—6.
OCCURRENCE OF HYDROCARBON LEAKAGE AT VARIOUS
EXPLOSIMETER LEVELS a
>0
? •
>1.0
> 0
.5
1.O
>0
.5
>1.0
LEL
LEL
LEL
LEL
LEL
LEL
L
LEL
LEL
d For bottom laoding 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 118 in the fail mode
b
Indicates the percent of total leaks at that level found at a
hatch that either passed or failed the CARB test
C Pass indicates a tank that is leak tight as defined by the CARB
criteria
d Fail indicates a tank that has failed the CARS test and therefore
leaks in excess of the allowable criteria
Percent
of Hatches Where
Leak Occurs
Avg. No. of
Occurrences Per
Hatch
Top
Loading
Percent
of All Leaks
Pass c
100
Fail
89.4
85.9
96.6
1.70
88.1
1.05
84.7
0.96
1.80
Bottom
Loading
Pass
53.2
1.21
1.02
53.1
51.0
49.0
19.6
46.8
Fail
5.2
46 • 9
2.6
63.5
0.36
53.2
39.2
1.58
0.07
1.09
25.1
0.03
0.80
8.5
4.5
74.9
91.5
95.5
3—22
-------
3.2.3 SONIC DETECTOR
The sonic detector was used much in 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.
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.
3.2.4 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,
the tire—bag appratus 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.
3-23
-------
As the truck is loaded, the bag is closely watched to observe
any leakage. The bags are sized so that the volume escaping is
about twice the allowable leak rate 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 mose
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 appratus itself may
not be able to fit all tank hatch configurations. Other equip-
ment, such as overfill protectors, tank compartment vapor vent
housing, etc., may get in the way and not allow a good seal between
the tank 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. However, 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.
3-24
-------
3.2.5 PRESSURE—VACUUM TEST (GARB)
The CARB pressure—vacuum tests were performed on each tank at
least once during the test period. The tanks were tested both
prior to mai ntenance and after mai ntenance 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
all ows a decrease from 6 inches øf water vacuum to 4 inches of
water vacuum in 5 minutes. By July 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 0 f water)
vacuum as specified by DOT regulations. The limits selected by
GARB are the maximum pressure or vacuum that can be applied to the
tank before the vent starts to open. 2
Back pressures observed during loading operations at both the
top and bottom loading terminals ranged from as l i 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—2 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.
3—25
-------
20
8
I
/
/
A
I
10
p.
/
L½$
12
14
—- — — Top Loading
/
_________ Bottom Loading
/
\
4 6 16
Back Pressure, inch 1120
Figure 3-2. Frequency of Occurrences of Back Pressure Ourinq Loading Operations
A
18
18
16
14
12
8
I
\,I
6
4
I
I
\
2
I
/
I
0
I
2
-------
Historical data was obtained from the terminal operators on
each of the test tanks which Included the dates of the previous
certification. This data was used to determine how long the trucks
are able to remain in certifiable condition under normal use. The
shop test data was used to indicate the condition of the tanks
which was compared to the historical data (see Table 3—7). As
indicated, only four tanks remained in certifiable condition. The
time frame ranged fr a n one year to one month. Of the tanks that
passed, the time since the last certification date ranged from four
months to four days.
The GARB test 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 3-3 and
3-4 illustrate examples of the pressure decay rate. In about every
case, the pressure continued to decline at an approximate 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 phenomenom 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 phenomenom was duplicated in the laboratory using a
smal 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. When the
degassing was completed for 20 minutes the tank was inmiediately
pressurized and sealed off. The pressure again increaseed with
3-27
-------
Table 3-7. TANK TIGHTNESS HISTORY
Tank
Last
Identification
Certification
Field
Pass!
Type
of
Number
Datea
Test
Date
Fail
Loading
63806 2/24178 6/23/78 F Top
53306 2/24/78 6/23/78 F Top
63767 2/09/78 6/20/78 F Top
53256 2/09/78 6/20/78 p Top
63766 2/08/78 6/20/78 F Top
53345 2/08/78 6/20/78 F Top
63765 2/07/78 6/19/78 F Top
63765 6/19/78 6/23/78 p Top
53304 2/07/78 6/23/78 p Top
63804 2/16/78 6/21/78 F Top
53307 2/16/78 6/22/78 F Top
63803 2/10/78 6/21/78 F Top
53297 2/10/78 6/21/78 F Top
63805 2/28/78 6/22/78 F Top
53305 2/28/78 6/22/78 F Top
67-182 5/18/78 6/14/78 F Bottom
67-392 5/23/78 6/13/78 F Bottom
67-475 5/24/ 78 6/15/78 p Bottom
68-795 5/03/78 6/14/78 F Bottom
68-597 6/13/77 6/13/78 F Bottom
68-275 6/28/77 6/15/78 F Bottom
68-977 5/03/78 6/16/78 F Bottom
67-775 No data 6/12/78 F Bottom
a Passed CARB Certification Test
3-28
-------
_____ ___________ ____ ____ Chevron Terminal
- - — 6-13-78
_____ _____ Truck 68—597 (trailer)
— —— CARB Pressure Test
• Premaintenance • Postmainten ce
— — .— — 1. _ . _____ _ . _ —
— — -. . — — - — _________________________________
——— ——--4 _ —- — ——.
___ ___ - - — -
E - - EEE I ___________
-= -1 - ___ ___
Iii____ - - - -____
L -’ -n - - — -- — - 1 —. kt — -
- I 2 ’- -i ==‘ ---- -- : = — z. =
; ..T4——---- \ ——--- - =z= ____ =— __ —T ’ -—______ ________
-4 ___: — . _______
- —. --r- ’= — -— -‘_-i =_t __ _-
— _ ±r
_______________ -
- T T -= ______ ______ ___________
- ; =F —--—-------—1----- -- - —
____ ____ : =—L==
L —— — =-2 : = = t _ . __ i
-1I E Z i T : iI _
: . JT:T = - ________
- L. -
Trn-’— - _ 4 - - -8 - - -10 - - - -i2 --—i’- _ :.i _ I8= .-2C ; - :: :
i: -- = = .:= -= _-.
— ______ - -— -- ____ _____
—-- — 1 - —, 4.- - ____
F— - ——————----- - -
__- - - - -—, -— --. — - - - ---4-.— , —
- ____
_______ ____________ ___________ _______ ___________
-_J .__ ___________ -i-—— I— ---- - -—-- 4—-- -
Figure 3-3. Typical CARB Pressure Test Results at Bottom Loaded Terminal
(Ref. 1)
3—29
-------
- -
Shell Terrni nal
6-1 9-78
— —----H -‘-
___ — -
__-
n 2 v— -
— -—-—_____ - -
- — . —--- --- - -i
= •-- -- LV - -
V
-
L__ - —
-
ii :
L
- =
— -
-—--—“V
- - -
— I - — — -
—a---- -___________ — —
r E- -- —
L- T — —:
‘
— - -4 -. V - V -__________________
I --—-4— -- — • - [ J___ V - V
4VV_ — — - V
—- •— V__ V - V V
V V
- -- !__ — -_ -2== —— V V E. . ..-
V V_________
- - V —
T
= 1 - - — — —
= - -- _________ -- -i - V
- - - — ---
Eii - - - - — V — — -----= —:_
V _ _. :_ .-‘: __--- . -- - —D ——,.- —-.-- ,_ --it -—- -—-- . ——ic --—-- - r . ______
_________ V c -
— _ ._—— - - -
_____ j _ . L — —l - —----a —
Figure 3-4. Typical CARB Pressure Test Results at Top Loaded Terminal
(Ref. 1)
Truck 63765 (Trailer)
CARB Pressure Test
• Premaintenance I Post ain:ena e
pp
I-
____________ 1 — ——
___ : _ ______
____________________ V
______________________________ I _____________ ___________ — -
TT:— — =— p — - 4---- i -1 i ñ T øc)
-------
time. When a vacuum test was attempted, the results appeared
normal with the vacuum decreasing with time. This was operated
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 ininediate
equilization and then a constant pressure (see Figure 3—5). This
data indicates that after degassing, a period of time is necessary
before testing to allow the tank to stabilize. If not, a leak may
in fact be present but the pressure may increase faster than or
equal to the leak rate and the tank may be considered certified.
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 requires using either a back—up
vehicle or rescheduling of deliveries. However, since this method
defines what a leak tight truck is, the method is very useful as
an enforcement tool.
3.2.6 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 øf 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
method is however, similar to the sonic detector method in that
both can be used as an indicator or locator of leaks.
3—31
-------
22 0
0 Q = Pressure test without
0
O stabilization period
20
0 A = Pressure test with
0 stabilizatIon period
— 0 0 [ ] = Vacuum test (values Indicate
vacuum in Inches of water)
A A A A A A A A A A A A A
16 —
14 —
12 —
10
0
o
o 0 0 c i
S
0 )
0 c i
In
U’
w 6_
4—
2
I I I 1 1 I I I 1 r
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Time, minutes
Fiqure 3-5. Laboratory Test Results for CI RB Pressure Test
-------
3.2.7 QUICK LEAK DECAY METHOD
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
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 gasolinee Several fire marshalls 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 equilibri&zn 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
tempertature. The appratus was assembled as shown in Figure 3—6.
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
3—33
-------
Multipoint temperature recorder
Ambient temp
Vapor temp
Tank liquid temp
Entering liquid temp
Liquid gasoline supply
7.
8.
9.
10.
11.
12.
Pressure relief vent
Pressure transducer
Signal conditioner
Strip chart recorder
Test tank
Needle valve
1.
2.
3.
4.
5.
6.
c. .j
Figure 3-6. Quick Leak Decay Test Apparatus
-------
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 desited vacuum in the tank was
achieved. Again temperature, pressure, and time were recorded
until a stable pressure was reached. Typical pressure versus time
curves are shown in Figure 3—7. 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 successflly as a pressurizing
liquid after the tank has been rinsed or flushed with diesel. 2
Diesel fuel is not available at all loading facilities so the
method was assessed for acceptability using gasoline. It should be
noted that even though this method may not be useable as a quick
monitoring technique, it does illusrate the need of removing as
much gasoline vapors as possible from the test tank and allowing
the test tank to stabilize before testing to obtain reliable
results.
3.2.8 VOLUME LEAKAGE
The volume leakage test was compared in the shop to the CARB
data results. The pressure was 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
CARS 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—8 and 3—9 for bottom and top loading instances. The top
loading data is not as complete as the bottom loading data because
3—35
-------
22
20
18 0
-1 0 Li 0
El (3 0
E] U 0
14 Pressure tests
A= Vacuum tests (pressures
12 noted are inches of
0
water vacuum
1 / )
U
Q 6J A
A
4 A A A A A
2 1
..-i—_.• r — t-•-—— •i•-•• -. .I. I
1 2 3 4 5 6 7 8 9 10 11 12
Time, Minutes
FIqure 3-7. TypIcal Pressure Versus Time Curves for Laboratory
Tests of Quick Leak Decay Method
-------
6.
5.0
4.
0
j 3.
U
C
U)
2
In
In
0
-J
U )
I-
In
In
U)
0
0
0
L
,-. —
.-.- . .
4
5” ..—
4 -
0
0
0’
S
1 —. • —
A.,—. —
0 ,
a
0
18 inch H 2 0
15 inch 20
12 Inch 1120
9 inch 1120
100 150 200 250 300 350
Volume Leakaqe, SCFI1
Figure 3-8. Volume leakage vs Pressure Decay for Bottom
loaded Tanks at Various Pressures
0
50
400
450
-------
0 . L
.
0.4_
o A ci-
0.3
U
.-.-...-.
---. .____.
— .
0.2 C) A o
o — 18 Inch 1120
15 inch 11 0
0.1 2
12 inch 1120
U
I — 1 1 f F F —
5 10 15 20 25 30 15 40 45
Volume Leakage, SCFH
FIgure 3-9. Volume Leakage vs Pressure Decay Rate for Top
Loading Tanks at Various Pressures
-------
of time constraints when performing the field tests. Linear
regression analyses were performed on the data and the best fit
curves are illustrarted. The correlation coefficients were
calculated and the bottom loaded data are significant at the 0.1.
percent probability level. The top loading data was significant
only at the 10 percent level. Correlation coefficients for the
volume leakage versus CARB test are listed in Table 3—B.
The volume leakage test compared very favorably to the CARB
method. However, for this project 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 addition to all the necessary GARB
equipment, rotameters and corresponding valves and tubing for
measuring volume rate are required. This method is however, like
the CARB method in that it can be used to define an aceptable leak
rate.
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
explosirneter method, San Diego “Bag” method, the volume leakage
method, and the CARB pressure loss rate mehtod.
The explosimeter tests reveal, that an LEL limit of 0.5 LEL or
1.0 LEL can be used as a pass/fail value using this method. The
limit selected will depend upon the reliability of the method that
can be accepted. Using the criteria of allowing no leaks greater
than 1.0 LEL, if a reading of 1.0 LEL was found the chances would
be very high (96/100) that the tank would be one that would fail
the GARB criteria and not one that would pass. This criteria would
yield a high success rate (meaning the tank identified as a
violator is in fact leaking in excess of the criteria) for select-
ing leaking tanks, but may not indicate some tanks that would
3-39
-------
Table 3-8. CORRELATION COEFFICIENTS FOR VOLUME
LEAKAGE RESULTS WITH RESPECT TO CARB TEST RESULTS
Constant Pressure Held Correlation Coefficient
Bottom Loading
9 inches H 2 0 0.96
12 inches H 2 0 0.95
15 inches H 2 0 0.94
18 inches H 2 0 1.00
lop Loading
12 inches H 2 0 0.90
15 inches H 2 0 0.90
18 inches H 2 0 0.65
3-40
-------
fail (e.g., leaks of 0.8 LEL). Using the criteria of no leaks
greater than 0.5 LEL, the chances of identifying a leak tight truck
as one that leaks increases from 4.5/100 to 8.5/100. This lower
allowable level will yield a lower success rate but will exclude
fewer of the tanks actually in violation.
San Diego APCD personnel have seleted pass/fail criteria based
upon the allowable pressure loss defined by the GARB 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
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 twice that allowed by the GARB
criteria. SDAPCD has experienced a 100 percent success rate using
this enforcement technique. 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 will define a leak
tight truck. The leak tightness of the tank can be defined as a
percentage of the total vapors transfered and an acceptable leakage
rate can be selected. For example, if a tank vapor volume of 5,000
gal (670 ft 3 ) 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 ft 3 /rnin)(60 min/hr)(1—0.99) = 40.2 ft 3 /hr
3-41
-------
By referring to Figure 3-5, this rate can be compared with the
CARB leak rate criteria, and for those trucks tested the corres-
ponding pressure leak rate would range from 1.25 inches of water
to 3.25 inches of water in 5 minutes. Table 3-9 shows several
corresponding volume leakage rates for various containment
requirements. The table can also be used to determine the pres-
sure loss requirements for a GARB type test given the vapor
containment required during the loading process.
3-42
-------
RATE REQUIRED FOR VARIOUS VAPOR CONTAINMENT REQUIREMINISa
Table 3.9. VOLUME LEAKAGE
( )
( )
Required
Vapor
Containment,
Percent
Volume
Leakage
Rate,
SCFI I
Corresponding CARB Pressure Loss Rate, In 1120/5 Mlnutesb
9 H 2 0
12 In 1120
15 in 1120
18 in
90
95
98.1
99
402.0
201.0
75.0
40.2
28.3
14.3
5.5
3.3
23.0
11.5
4.3
2.5
20.3
10.3
3.8
2.3
23.0
10.8
3.0
1.3
a Based upon a tank volume of 5,000 gallons and a loading rate of 500 GPM
b See Figure 3.5, data based upon tanks tested during field test phase.
-------
References for Section 3.0
1. Leak Testing of Gasoline Tank Trucks , Scott Environmental
Technology, EPA Contract No. 68—02—2813, Work Assignment No. 19,
August 1978 (Draft)
2. Personal comunication with Dean Simeroth, California Resources
Board, Sacramento, California
3. Ibid
3-44
-------
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 fran 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 maintenance program
some tanks have shown leak tightness maintained for over 6 months.
The CARB certification program specifies testing the trucks
annualy. 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 costs, as outlined in Section 2.5, of maintaining the
equipment in a leak tight condition are not unreasonable. Even if
certification tests were required more often, the costs still remain
reasonable. Requiring certification more often than 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 explosi—
meter method, San Diego Bag method, the CARB pressure loss rate
4-1
-------
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 repeatable
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. 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 selected yielding acceptable results, either of
the two shop tests, the CARB test or the volume leakage test, should
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 should be used as an interim enforcement
of screening procedure. These short methods could be used to
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
4-2
-------
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 0.5 LEL or 1.0
LEL using the explosimeter method, depending upon the success rate
required by the 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 should occur as indicated with either
a sonic detector or a bubble indication solution. No 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
-------
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.
Clean dome lid gasket as required.
2. Inspect P—V valve seals for debris or foreign material on seat
Check valve operation to ensure valve will move smoothly and
will reseat properly.
3. Inspect condition of hatch base ring for severe damage or
warpage.
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
f or cracks or tears.
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 worn or damaged parts as
required.
BULK PLANT AND TERMINAL OPERATORS
1. Inspect all gasoling 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.
A-i
-------
3. Inspect above ground vapor piping using a bubble indicating
solution. Check all piping connections and joints with the
solution and Took 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
mechnaisms for signs of damage or excessive wear.
Repair or replace damaged components as required.
SERVICE STATION OPERATORS
1. Inspect vapor couplers used for vapor transfer to underground
storage tanks. Check yalve seats to make sure they are clean
and free of debris and foreign material.
2. Inspect vapor hoses (if applicable) as stated above.
A-2
-------
APPENDIX B
ACTUAL MAINTENANCE PERFORMED ON DELIVERY TANKS
DURING FIELD TEST PHASE
-------
APPENDIX B
ACTUAL MAINTENANCE PERFORMED ON 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)
Trail er/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, Trail er/53304
Truck degassed, start 0615 — end 0720
Trailer degassed, start 0730 — end 0930
Truck — No maintenance
Trailer — No maintenance
Truck/63804
No. 2 dome lid leak — adjust tension
No. 3 dome lid leak — adjust tension and straightened hatch
(1 man —— 5 minutes)
B—i
-------
Trail er/53307
No. 3 tightened base ring, replaced base ring and gasket
No. 1 dome lid leak - adjust tension
(1 man —— 10 mm)
Trai ler/53305
30 minutes — leak compartment
Start 0630 - end 0850
Truck/6 3805
Degass, start 0810 — end 1030
Trailer
No. 3 dome lid leaking, cannot get over 1 inch H 2 0
(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
Truck
Nos. 1 and 2 dome lid leaking, straightened dome lids and adjusted
spring tension
(1 man —— 8 minutes)
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, vacuum test —— 6.12 to 6.28
inches H 2 0 in 5 minutes —— invalid since not stabilized
8—2
-------
Trailer/53297 — Trailer 63803
Degas, start 0630 — end 0805 (on trailer)
Degas, start 0815 — end 1050 (on truck)
Maintenance/trailer —— all three covers leaked, adjust tension 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
Truck/63767
No. 1 cover leaking, upon adjusting spring tension a bolt broke,
must replace dome lid
(1 man —— 19 minutes)
Trail er/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 H 2 0
(3 minutes 1 man)
No testing done pre—maintenance
Maintenance for GARB Certification 9:10 to 10:30
Truck/68—275, Trailer
Previous load diesel - old CB equipment
Bad leak compartment No. 3 vent, detectable by smell, ear, feel,
leak dur 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 and 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—3
-------
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.
Truck/67— 182
Maintenance 1250 — 1305
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, neq hose with a new material cover, many leaks but
total volume not significant
B-4
-------
APPENDIX C
SUGGESTED ENFORCEMENT INSPECTION CHECKLIST
-------
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 greater than 1.0 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 rood 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.
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—i
-------