Leak Detection At High-Volume Retail Sites
Few, if any, solutions to the challenges currently exist.
By Steve Purpora
Most fuel retailers dream of owning a string of million-gallon-a-month facilities. For sure, these facilities come with many operational challenges, such as not running out of product and making sure dispenser meters are accurate. While taking their money to the bank, however, many high-volume facility operators neglect to think about what could happen if a leak occurred. Most prefer to hide their heads in the sand, cross their fingers, and hope that it won't happen. Unfortunately, this strategy is woefully inadequate to the task at hand. A million-gallon-per-month throughput means that an average of 23 gallons per minute are flowing through the system every minute of every day. A million gallons a month means that pumps rarely stop pumping. A million gallons a month means there is not enough quiet time to conduct leak detection in either the tank or the piping system.
The consequence is that high-volume facilities have high-volume releases. Tens of thousands of gallons is a “small” release for a truck stop. The largest release I know of is estimated to be a million gallons. Another release created a lake of diesel fuel two inches deep and two acres in area. Releases like this do not happen overnight. Releases grow to such enormous volumes because standard leak detection methods are ineffective at high-volume facilities.
The list of reasons why leak detection is ineffective is long. The usual culprits include:
- Lack of quiet time to conduct automatic tests
- Satellite dispensers
- Lengthy piping runs
- Deep burial of piping
- Air pockets.
Let's look at each of these factors in turn.
Challenge  Lack of Quiet Time to Conduct Automatic Tests
Although many sophisticated techniques for leak detection are available, the one thread that almost all of them have in common is that the storage system has to be idle while measurements are made. At high-volume facilities, idle times are rare. So rare, in fact, that there is not sufficient time to gather enough data, even with automated systems, to complete a valid test.
About the only commercially available method of leak detection that will handle high throughputs at single-wall facilities (up to 2.7 million gallons a month) is a completely automated inventory system that conducts sophisticated statistical analyses on mountains of inventory data to detect leaks.
Challenge Satellite Dispensers
An issue unique to truck stops is satellite dispensers. Satellite dispensers enable fuel to be delivered to tanks on both sides of a truck in a single sales transaction. To do this, piping is usually run from a master dispenser underground to a satellite dispenser. Depending on the arrangement of solenoid valves in the master dispenser, the satellite piping may be isolated from the main piping run during automated leak tests. This is an issue even for lower volume truck stop type operations such as a fleet fueling facility. Because the satellite piping is downstream from the meter in the master dispenser, leaks in the satellite piping will also be invisible to any leak detection method based on inventory control.
Challenge Lengthy Piping Runs
While it is easy to see how big a high-volume retail site is on the surface, it is harder to envision how much piping is underground at a facility. Many sites have many hundreds of feet of piping underground. The longer the pipe, the greater the volume of product contained in the pipe and the greater the possibility that temperature changes will affect the test. Sometimes the piping material will be a mix of steel, fiberglass and flexible piping, complicating the task of leak detection even more because of the variations in elasticity of the different materials.
Challenge Deep Burial of Piping
Closely associated with lengthy piping is the issue of deep burial. Standard industry practice is to slope piping uniformly from the furthest dispenser back to the tank. If this distance is several hundred feet, the tank top can end up being 6 to 10 feet underground. This in turn can create excessive hydrostatic pressure on some mechanical leak detectors, preventing them from detecting leaks in the piping.
Challenge Air Pockets
The longer and more complex the piping run, the greater the chance that there will be nooks and crannies within the piping that will trap air. In a piping system with no air pockets that is undergoing a tightness test, small changes in liquid volume (i.e., liquid leaking out of the pipe) will create large changes in the pressure in the pipe because liquids are not compressible. When air pockets are present, however, small changes in the liquid volume will produce smaller changes in the pressure because the air pocket will expand to fill the void left by the loss of liquid. Air pockets decrease the sensitivity of pressure-based piping tightness tests. Air pockets also can cause pressure fluctuations during a test as they are affected by changes in ambient temperature. In short, air pockets create significant challenges to effective testing when they are present.
Is There an Answer?
With all of the challenges to automated leak detection posed by high-volume facilities, the most frequently used solutions are inspections and periodic tightness tests conducted by outside personnel. This approach also faces several challenges. High-volume facility operators prefer to have these inspections and tests conducted at night to minimize the disruption of their fueling operations. Conducting an inspection of a complex facility in the dark of night while dodging the ever-present traffic flow is hardly an ideal methodology for conducting a careful inspection. In addition, severe time pressure is often imposed on the tester to complete the test as quickly as possible so that the facility can resume normal operations. These conditions are not conducive to thorough inspections or quality tightness tests. Typically, these inspections and tests occur no more than once a year, leaving the facility vulnerable to undetected leaks for the vast majority of the time it is in operation.
What Now?
In the short term, I see two improvements that need to be implemented:
Inspection and testing needs to be conducted in daylight with adequate time given to conduct a quality job. This may require some regulatory intervention as otherwise there always will be a testing contractor somewhere willing to make a buck no matter what impossible constraints the facility owner places on the people actually doing the work.
Facility operators need to implement the routine daily and monthly operational inspections described in PEI RP500, Recommended Practices for Inspection and Maintenance of Motor Fuel Dispensing Equipment and RP900, Recommended Practices for the Inspection and Maintenance of UST Systems. Because many leaks can be seen if care is taken to look for them, these routine inspections could help minimize the volume of releases from high-volume facilities.
In the long term, we need to take a closer look at how these types of facilities are constructed and develop designs that treat leak detection as a critical necessity and not a bothersome detail. For example, most facilities today consist of a single distribution system where all pumps feed all dispensers It is not possible to shut down a portion of the facility so that it can be tested while another portion of the facility can remain open and service customers. The addition of a few valves and perhaps an extra pump or two could create several independent fueling systems operating side by side that allow a portion of the system to be shut down for testing or even routine service like changing filters without necessitating a complete shutdown of the facility. The amount of fuel each system pumps would also be reduced, simplifying some of the challenges described in this article.
We have come a long way in the last few decades in terms of reducing the volume and frequency of releases from underground storage systems. The dramatic reduction in the number of facilities during this time has also created a substantial number of facilities with tremendous throughputs that pose leak detection challenges that were hardly envisioned 20 years ago when federal tank regulations were established. I believe the time has come to address these glaring holes in our groundwater protection strategy.
|
