|
Issue
Date: November 22, 1995
Revision Date: September 22, 2009
Tanknology
- NDE
VacuTect
NON-VOLUMETRIC
TANK TIGHTNESS TEST METHOD (VACUUM)
|
Certification
|
Leak
rate of 0.1 gph with PD = 100% and PFA = 0%.
|
|
Leak
Threshold
|
A
tank system should not be declared tight when:
sonic emission of air ingress is detected in ullage area
and/or; sonic emission of bubbles formed by air ingress
is detected in product-filled portion of the tank and/or;
water ingress is detected by the water sensor.
|
|
Applicability
|
Gasoline,
diesel, aviation fuel, fuel oil #4, waste oil.
Other liquids may be tested after consultation with the
manufacturer. Water miscible products limit the effectiveness
of water ingress detection.
|
|
Tank
Capacity
|
Maximum of 75,000 gallons.
Test may be performed at any product level, including empty, if total ullage volume does not exceed 20,000 gallons.
Maximum of 30,000 gallons per tank and 60,000 gallons cumulative capacity for manifolded tank systems with microphone (hydrophone) and water sensor in each tank.
|
|
Waiting
Time
|
None
between delivery and testing.
|
|
Test
Period
|
Minimum
not specified to declare a tank "non-tight".
There must be no dispensing or delivery during test.
When water level in tank excavation backfill is
below bottom of tank:
Minimum of 1 hour to declare tank tight.
When water level in tank excavation backfill is
above bottom of tank:
Minimum to declare a tank tight may need to be extended
because a water sensor must be used.
When water level in tank excavation backfill is
not determined:
Minimum to declare a tank tight may need to be extended
because a water sensor must be used.
To determine test period, water level in tank excavation
backfill must be assumed to be just above bottom of tank.
When using a water sensor, the test period
is determined based on tank size, water level inside tank
prior to test, tank tilt, type of water sensor, water
sensor location in the tank, and water level in tank excavation
backfill relative to bottom of tank. Tank tilt is determined
by an inclinometer, or by measuring tank bottom elevations
at two points within tank. Water sensor is placed in low
end of tank or in pre-existing cross section of water
inside tank.
For example:
For a printed circuit board water sensor, the test period
to detect a 0.1 gph leak rate in a 10,000 gallon (96"dia
x 324"lg) level tank without a striker plate and
without water in the tank is 51 minutes for water ingress
to contact the water sensor plus 38 minutes to allow the
sensor to detect the "minimum detectable change in
water level" (see "Water Sensor" section
below).
For example:
For a magnetostrictive water sensor, the test period to
detect a 0.1 gph leak rate in a 10,000 gallon (96"dia
x 324"lg) level tank without a striker plate and
without water in the tank is 37 minutes for water ingress
to contact the water sensor plus 2 minutes to allow the
sensor to detect the "minimum detectable change in
water level" (see "Water Sensor" section
below).
|
|
Test
Pressure
|
Required
test pressure is a function of tank construction, burial
depth, product level in the tank, and water level in tank
excavation backfill.
Pressure differential across tank wall at the bottom of
the tank must be at least 0.5 psi.
Pressure differential across the tank wall is equal to
the absolute value of vacuum applied to tank, plus the
pressure of the tank excavation backfill on tank, plus
groundwater pressure on tank, minus pressure of liquid
in tank.
If water level in tank excavation backfill is not determined
by wells, probes or pump boxes in the tank excavation
backfill, test pressure calculation must account for both:
1) water level just above bottom of tank to achieve minimum
0.5 psi at worst case condition, and
2) tank completely submerged to assure tank is not damaged
from over pressurization.
|
|
Temperature
|
Sonic
emission and water ingress are independent of product
temperature.
|
|
Water
Sensor
|
Must
be used if water level in tank excavation backfill is
above tank bottom or if water level in tank excavation
backfill was not determined.
Printed circuit board sensor minimum detectable water
level is 0.022 inch, and minimum detectable change in
water level is 0.016 inch.
Magnetostrictive sensor minimum detectable water level
is 0.017 inch, and minimum detectable change in water
level 0.001 inch.
|
|
Groundwater
|
Per
California regulations, depth to groundwater in tank excavation
backfill must be determined. If
groundwater level in tank excavation backfill is above
bottom of tank, or if groundwater level in tank excavation
backfill is not determined, test time must be sufficient
to detect water ingress using one of the above water sensors. |
|
Comments
|
Evaluated
using gasoline, diesel, Jet-A, and JP-4.
Microphone (hydrophone) should be located within 60 feet
of any possible leak source.
Vacuum test method may not be effective in some tank excavation
backfill (such as clay) because it may plug holes in tank.
If free product is present in tank excavation backfill,
a leak in the free product zone may not be detected by
a vacuum test method.
A well point or observation well in the tank excavation
backfill can help identify the presence of free product,
tank excavation backfill material, and water elevation
in the tank excavation backfill.
More that 4 psi pressure differential across the tank
wall at any location in the tank could damage tank.
|
|
Tanknology
- NDE
|
Evaluator: Ken Wilcox Associates
|
|
8900 Shoal Creek Blvd.
,
Building 200
|
Tel: (816) 443-2494
|
|
Austin
,
TX
78757
|
Dates
of Evaluation:
10/28/91
,
09/18/92
,
|
|
Tel: (800) 800-4633
|
02/20/92
,
01/18/94
,
02/23/96
|
|
