Third Party Testing Results on Interstitial, Vapor and Liquid Monitoring

Compare the test results submitted by various manufacturers to the National Work Group on Leak Detection Evaluations (NWGLDE), which published its results in April 1997. PE&T adds 15 pages of tables, an update and an introduction by Jairus D. Flora, Jr., Ph.D., of the Midwest Research Institute

Behind the Test Scene
Curt Johnson, Chair of the National Work Group on Leak Detection Evaluations (NWGLDE)

How the List was Developed. The US EPA began evaluating leak detection methods back in 1986. Subsequent evaluations became the responsibility of vendors. In 1990, several states began to independently examine the test data from vendors, and discovered that some of the tests were not done according to the EPA’s standard test procedures. The National Work Group on Leak Detection Evaluations (NWGLDE) was formed in 1993. The Group consists of volunteers from seven states and two EPA offices who perform a comprehensive review of third-party tests of UST and piping leak detection methods to ensure that: (1) these tests are run in accordance with EPA or other acceptable test protocols; and (2) the test data does, in fact, support the results of the evaluation.

The Function of the List. The primary function of the work group is to determine whether a leak detection method has been evaluated according to an acceptable evaluation procedure.

The Limitations of the List. Although the List may be a useful tool for locating po-tential vendors, it is not comprehensive, and being included on the List does not guarantee that the leak detection method works. (See OUST Engineer David Wiley’s column.)

Interstitial Monitoring
What is it? Interstitial monitoring applies to double-walled tanks and piping; such systems monitor the space between the two containing walls. The regula-tory requirements for this method (as well as for vapor and out-of-tank liquid monitoring) are found in 40 CFR Subpart D 280.43(e)(f)(g). The regulations require that interstitial monitoring be capable of detecting a release through the inner wall from any portion of the tank that regularly contains product. Most interstitial moni-tors also detect any breach in the outer wall of the double containment system, and sound an alarm when-ever the system is violated.

Evaluation process There is no EPA standard test protocol for evaluating interstitial monitors. In fact, interstitial monitoring can be done through volumetric testing, non-volumetric testing, vapor monitoring or liquid monitoring of the interstitial space. The physical principles of each particular interstitial monitor indi-cate which of the EPA protocols should be adapted and used to test the equipment.

In the accompanying tables, the interstitial monitors use three approaches to test the interstitial space: hydro-static monitoring, pressure/vacuum monitoring and liquid level monitoring. There are 15 vendors with inter-stitial monitors that monitor the product liquid-level in the interstitial space (see Table 1). These various moni-tors use one of several different physical principles to detect a liquid (either product or water) that enters an interstitial space.

The evaluations generally follow the EPA test proce-dure for liquid-phase out-of-tank test methods, which reports on accuracy, response time and lower detection limits. However, since the interstitial space is quite small, an interstitial monitor can be successful even if its lower detection limit is greater than the J inch (0.32 cm) required for out-of-tank liquid monitors. The mon-itor only needs to detect liquid entering the interstitial space at a rate of 0.2 gallon per hour or greater. This rate generally produces enough liquid to be easily detectable in the interstitial space within a few days.

There are five hydrostatic monitors that work like a volumetric tank ti ghtness test, and monitor the change in level in the interstitial space that is filled with a liquid (see Table 3). The evaluations reported detection times and product activation heights for both upward and downward changes in the intersti-tial liquid monitoring levels.

One vendor lists an interstitial monitor that uses the pressure/vacuum approach (see Table 2). This monitor was evaluated using the EPA test method for non-volumet-ric tank tests. The evaluation of this system reported a probability of detection and false alarm for detecting a 0.1 gallon per hour leak.

Comments The evaluations of interstitial monitors do not always provide a probability of detection and a probability of false alarm, as is the case with other leak detection methods. However, when an interstitial monitor provides an alarm, generally no product has been released to the environment. The alarm merely identifies that the double-containment no longer exists and only single containment for the product remains. Thus, while action to restore the double-contain-ment is needed, generally no remediation is required.

Vapor Monitoring
What is it? Vapor monitoring for leak detection is an external monitoring method. One or more sensors monitor the hydrocarbon concentration in the soil gas in the backfill around the tank. A site assessment is required to ensure that:

• The porosity of the backfill material is sufficient to allow the product vapor to migrate through the vadose zone in the backfill to be intercepted by monitoring wells.
• The product in the tank must be sufficiently volatile to produce a vapor level that is high enough to be detected at the monitoring wells.
• The site must not be subject to interferences, such as a high water table, that would prevent the vapor from migrating through the soil and being intercepted.
• The site must not have background contamination to the extent that it would interfere with the detection of a new release.
• The size and nature of the site must be used in deter-mining the required number, spacing, and location of moni toring wel ls, and these must be adequately installed, protected, and identified.
• Assuming that the site is suitable, the vapor monitoring equipment must be capable of detecting any significant increase in the concentration above background of the product vapor level or a tracer compound.

Evaluation process The EPA has developed standard test procedures for evaluating vapor-phase out-of-tank monitoring devices (“Standard Test Procedures for Evalu-ating Leak Detection Methods: Vapor-Phase Out-of-Tank Product Detectors,” EPA/530/UST 90-008, March 1990).

This protocol differs from other EPA protocols in that it does not specify a detectable leak rate together with required false alarm and detection rates. Instead, the proto-col requires the evaluation of the vapor sensors for accu-racy, precision, detection time, fall time, specificity and lower detection limits.

The tests are done with the detector in a test chamber. The vapor detectors are challenged with a variety of com-pounds at different concentration levels to determine the level at which they will respond. The compounds specified are volatile hydrocarbons, such as benzene, as well as com-mercial unleaded gasoline. The evaluation does not test the systems under field conditions where effects such as soil porosity, humidity, and temperature might affect the transport of the vapor or liquid from the tank to the device in a monitoring well.

The NWGLDE List of April 18, 1997 shows 13 different vendors of vapor-phase out-of-tank product detectors, with six different operating principles. Most of the detectors are based on a metal oxide semiconductor or an adsistor for product detection. However, fiber optics, product per-meability, adsorption sampling, and gas chromatography are also used.

Comments The vapor monitoring may be implemented either automatically or manually. In an automatic mode, a console receives a signal from the sensors in the monitoring wells and displays an alarm or prints a warning if there is an increase in the vapor concentration. The manual mode requires a person to inspect the monitoring wells with a vapor sensor periodically (at least once a month) to deter-mine whether a detection occurred. Both automatic and manual systems are available.

With this type of monitoring, increases in vapor concen-tration are not always clear cut. Particularly with manual vapor monitoring systems, the operator must understand the system and be able to interpret the results. A transient increase in vapor concentration may result from a small spill or splash while fueling, or from a larger spill during a delivery. A one-time event would show a spike in the vapor concentration, decreasing toward baseline over time, while a leak would result in a continuing release with the vapor concentration increasing and maintaining a high level.

Liquid-Phase Out-of-Tank Monitoring
What is it? Groundwater monitoring for leak detection is an external monitoring method done in the backfill around the tank. Groundwater monitoring requires that the ground water level must never be more than 20 feet below the surface. Other requirements include:

• The product must be immiscible with water and have a specific gravity less than 1, so that it will float on the water surface.
• The soil characteristics must be such that the liquid can be intercepted by appropriately spaced and designed monitoring wells. A hydraulic conductivity of at least 0.01 cm/sec is required.
• The site must be assessed for suitability and to deter-mine the appropriate number, spacing and location of the monitoring wells. These wells must be correctly designed, installed and identified.
• If the site is suitable, the monitoring device must be capable of detecting a layer of free product on the water table of J inch or less.

Evaluation process The EPA has developed standard test procedures for evaluating groundwater monitoring devices. (“Standard Test Procedures for Evaluating Leak Detection Methods: Liquid-Phase Out-of-Tank Product Detectors,” EPA/530/UST 90-009, March 1990). The pro-tocol is similar to that for vapor monitoring in that it does not specify probabilities of detection and false alarm for specified leak rates. Instead, it tests the liquid-level sen-sors for accuracy, precision, detection time, fall time, specificity and lower detection limits.

The tests are conducted with the detector in a test chamber. The evaluation protocol does not test the sys-tems under field conditions where effects such as soil porosity, humidity and temperature might affect the trans-port of the liquid from the tank to the device in a moni-toring well.

The liquid detectors are challenged with a variety of compounds at different thicknesses on a water column. The compounds specified are volatile hydrocarbons, such as benzene, as well as commercial unleaded gasoline.

The liquid sensors may be implemented either automat-ically or manually. In an automatic mode, a console receives a signal from the sensor and displays an alarm or prints a warning if a hydrocarbon layer is detected on the water surface in a monitoring well. The manual mode would require a person to inspect the monitoring wells periodically (at least once a month) to determine whether a release occurred. Both automatic and manual systems are available.

The detectors are required to detect a layer of 1/8 inch (0.32 cm) of free product. If free product is found on the water table, the result is clear: the tank system has a prob-lem. This may result from either a leak from the tank or its associated piping.

Comments There are 16 vendors of liquid-phase out-of-tank product detectors listed on the most recent list of eval-uations. These detectors use a variety of operating principles, including float switches, electrical conductivity or imped-ance, product soluble cables, and fiber optic sensors.

Conclusion
The external monitoring methods require specific site con-ditions for use. When site conditions permit, these meth-ods have this advantage. One system of monitoring wells can provide monthly monitoring for the site, including both tanks and piping. However, if a release is detected, additional testing may be needed to identify the source.

Discuss