PISTON LEAK DL._CTOR stream from . If, two or three dispensers on the same
<br /> product rade with very small, weeping joints can be
<br /> P 9 rY P� 9
<br /> enough to make the leak detector take considerably
<br /> longer than normal to open.
<br /> i IS ENOUGH TIME ALLOWED FOR THE LEAK
<br /> TEST? (I.E. IS THE SYSTEM STAYING CLOSED
<br /> METERING
<br /> PISTON L ONG ENOUGH FOR THE TEST TO BE COMPLETED?)
<br /> SEAL 1) Time delays
<br /> History has taught us that one cannot rely upon the
<br /> nozzles being closed long enough priorto dispensing,for
<br /> the leak detector to complete its test. As a result, some
<br /> �\ dispensing equipment provides a delay in the opening of
<br /> the solenoid valve after activation of the pump or a delay
<br /> FIGURE 4Bp
<br /> \ \� in allowing flow to the nozzle after activation of the um
<br /> and solenoid valve. This characteristic may be an
<br /> simulated leak in place and allow the leak detector to integral part of the design or accomplished by retrofit.
<br /> meter for about ten (10) minutes. The theory in these cases, is that flow is not allowed
<br /> Leaking product through to vent to the hose until completion of the three-step leak test is
<br /> tubing must be caught in a suitable accomplished. In other words, the system remains
<br /> container to avoid accidental dis- sealed until the leak detectorcancomplete the three-step
<br /> charge of product. This can cause leak test. However, if a leak at the rate of approximately
<br /> severe personal injury, death or 3 gph or greater is present in the system,downstream of
<br /> substantial property damage if ig- the leak detector, the line pressure cannot increase
<br /> nored. adequately to allow the poppet to go from the metering
<br /> If the metering seal is bad,fluid will flowthrough the leaky position to the full flow position, regardless of the delay
<br /> seal, through the piston relief valve and out of the vent allowed. Even with a built-in delay, if the test time is
<br /> tube while the leak detector is in the metering position. longer than the delay allows for, the result may be
<br /> Line pressure should be 15 - 22 PSI during this time. restricted flow by the leak detector.
<br /> To check if sufficient time is available for the leak
<br /> 5) Is thermal contraction slowing down the leak detector to perform its test, a pressure gauge should be
<br /> detector? installed as in Figure 3. (Observe warning on Fig. 3).
<br /> Thermal contraction occurring while the leak detec- The gauge is part of the leak detection test apparatus.
<br /> for is performing a leak check will slow down the opening When the dispenser switch is activated afterthe pressure
<br /> time of a leak detector. Product that is lost due to has been bled to zero through the ball valve,the pressure
<br /> contraction will be viewed by the leak detector the same must increase to full pump pressure (indicating that the
<br /> as product lost due to a leak. If the leak detector is LLD has completed the line test)before the sound of the
<br /> metering product into the line at a rate of 3 GPH and solenoid opening is heard. This test must be performed
<br /> thermal contraction is occurring at a rate of 2 GPH, the with the bibcock closed. All solenoid valves in the system
<br /> net pressurization flow into the piping system is only 1/3 must be checked for appropriate timing.
<br /> as much or 1 GPH. Accordingly, the opening time is
<br /> extended. The opening time will vary with the rate of IN SUMMARY, by using the previous ABC steps, it
<br /> thermal contraction present at the time of the leak test. is possible to diagnose the reasons leak detectors are
<br /> going into a restricted flow condition. The problem may
<br /> 6) Are there minor product leaks in the system? be an inadequate solenoid delay time, prolonged test
<br /> Leak detectors are designed to detect a leak of 3 time or both. If the delay time is sufficient for the leak
<br /> GPH downstream of the leak detector. Leaks smaller detector to perform its test,uninterrupted flow will occur.
<br /> than this will not necessarily result in restricted flow. To state it another way, if B and C are synchronized,
<br /> However, a leak smaller than 3 GPH will affect the leak the leak detector will not cause interruption in flow even
<br /> detector by prolonging the opening time. if the leak detector performs a leak test each time the
<br /> If there are leaks greater than approximately 3 gph, pump is activated. If the system stays sealed long
<br /> the leak detector will never open but will function as enough for the leak detector to complete the test, no
<br /> described above. However, if there are leaks in the interruption of flow by the leak detectorwill occur regard-
<br /> system which are less than approximately 3 gph,the leak less of the opening of nozzles.
<br /> detector will eventually open, but, not in the prescribed However, if B and C are not synchornized, it is
<br /> time. If small leaks are present while the leak detector probable that, periodically, the system will be open,
<br /> meters 3 gallons of fuel per hour into the line,a portion of previous to the completion of the line test by the leak
<br /> this fluid will leak out. The net effect will be a longer time detector and resulting in the restriction of flow by the leak
<br /> for the leak detector to open. The nearer the leak is to detector. If the delay time is insufficient,or the test time
<br /> zero, the nearer the time is to the normal test time. istoo long,or whether thermal effects,system resiliency,
<br /> However, the nearer the leak is to 3 gph, the longer the internal leak detection seals, upstream pump seals or
<br /> test time will be.
<br /> From our experience, most of the small leaks occur actual system leaks are causing the line pressure to fall
<br /> in the dispenser usually above the shear valve and below to zero, necessitating a line test by the leak detector,this
<br /> the meter. Since the leak detector senses leaks down report will aid in determining the reason(s).
<br /> -5-
<br />
|