The project objective was to demonstrate and validate the performance of three innovative technologies for leak detection by assessing their ability to detect and accurately locate leaks in challenging pipe types such as polyvinyl chloride (PVC), asbestos cement (AC), and mixtures of pipe types typically found on Department of Defense (DoD) installations. The fundamental questions addressed by this study include: Is implementation of these technologies technically feasible for use by DoD installations to reduce water loss and to help meet water and energy conservation goals of the EO? Are these technologies cost effective?

The demonstration evaluated two types of cross-correlating leak detection technologies: (1) a continuous monitoring network approach, and (2) an inspection approach that used sensors temporarily deployed to test segments of pipe within a water distribution system. Three different product lines were tested: one for continuous monitoring and two for periodic inspection of pipe segments. Each technology was demonstrated for detecting and pinpointing leaks in metallic and challenging non-metallic pipe types. For each of the technologies, accelerometers or hydrophones were used to detect acoustic signatures of leaks, and time offsets between sensor locations were used to derive leak locations.

Evaluations were conducted under controlled conditions at an underground pipeline test bed (TB) that was configured with simulated leaks followed by testing under operating conditions within the U.S. Army Engineer Research Development Center’s (ERDC’s) water distribution system. The TB included 11 simulated leaks ranging from approximately one gallon per minute (gpm) to 8 gpm that could be controlled from above ground. Projected benefits from water and energy savings and estimated costs for leak detection deployment were also estimated. These projections indicate a savings-to-investment ratio (SIR) greater than one for installations with average rates of water main breaks within their water distribution systems. Actual cost-benefit performance should be monitored as leak detection systems are deployed on a site-specific basis. 

For the TB evaluation, only the technology that used an inspection approach and accelerometers met all the performance criteria. The continuous monitoring technology and the survey technology using both hydrophones and accelerometers did not meet several performance criteria in the TB evaluation. The simulated leak conditions were successfully detected by all the technologies. However, the location accuracy varied between the technologies. Two of the three technologies passed the performance objective of locating 90% of simulated leaks within 4 feet (ft) of the known locations in the TB. The leak location results for PVC pipe ranged from 86% to 100% within 4 ft of the known leak locations. False positives were an issue for two out of the three technologies. There is a potential to mitigate false positives in field applications through focused acoustic surveys that are typically conducted at the correlated location prior to marking the leak location. All three technologies were able to detect small leaks at approximately one gpm. Challenges were encountered with detecting multiple leaks within a bracketed sensor pair (even though the simulated leaks were spaced more than five ft apart) and in spanning mixed pipe materials. Although the capability to detect and locate leaks under these scenarios was claimed, the leak detections were not as accurate compared to the single leak and single pipe material scenarios within the TB.

For the operational water distribution testing, three leaks were detected within the portion of the ERDC water distribution system selected for inspection. The limited number of leaks detected in the field tests did not provide sufficient information for the evaluation of the performance criteria (even though visual indications of one leak were observed during the test). Water, energy, and SIR estimates were developed based upon an industry average water main break frequency and regional water and energy cost data.   

Leak detection systems that rely on an intermittent inspection approach hold the most promise for implementation at military installations at this time. A widely accepted best management practice with this technology is to cover an entire base every 3 to 5 years (American Water Works Association [AWWA], 2009). Both the LeakFinderRT and Correlux leak detection systems process the leak signature data in the field without any requirement for information technology (IT) security, or connection to government IT assets or the Internet. Leak detection using these systems can be procured as a service via a maintenance or job order contract. In addition, if an installation has the manpower, equipment can be procured for in-house use. Since the Correlux technology met all of the performance thresholds for the TB evaluations, this technology could be considered for additional field testing and deployment in its current state.

Further development and investment would be required for widespread adoption of a continuous monitoring system at military installations. Software compatibility issues and the difficulties of securing IT approval would deter implementation at military installations under the current IT security environment. The primary operating concern to be addressed includes network security for information systems that are being deployed in conjunction with advanced metering infrastructure (AMI) systems.