The Challenge:
The RWA, located in New Haven, Connecticut delivers 43 million gallons of water to 15 South Central Connecticut municipalities daily. Their new ethos is to be proactive in their approach to non-revenue water (NRW)’ and bring about long-term sustainable reductions across their 1700 miles of water distribution mains.
On an operational level, the RWA is challenged with six-foot deep valve chambers and therefore requires a flexible solution that is both easy to deploy and retrieve. In addition, the system experiences elevated system pressures, averaging>100psi. Undetected leaks thrive in high-pressure environments, and spontaneous breaks are common due to added operational stress.
The FIDO AI Solution:
The RWA practised efficiency by deploying of FIDO Bugs to strategically sweep their system for leaks in conjunction with valve maintenance programs. They prioritized rural areas based on AMI data, pipe age, and high break frequencies.
The northern section of their service area was a logical location for the RWA teams to initiate a FIDO sweep for leaks. Before this could begin, the six-foot deep chambers were cleaned out to allow for valve exercising and deployment of FIDO Bugs. Once completed, the Bugs were placed across the defined area and set to record overnight. This setting allows for five recordings to be taken consecutively over a five-hour period.
FIDO AI identified the leak and its size. FIDO Cloud correlation was then used to pinpoint the location of the leak between two POIs flagged by FIDO AI. These locations were then verified independently and confirmed.
The Results
The RWA excavated in the pinpointed location and discovered that FIDO AI had correctly pinpointed a circumferential fracture on a cast iron pipe (CIP). Not only was the leak confirmed, it was also classified as a “large” leak.
This break alone lost an estimated 100gpm and would have gone undetected indefinitely without FIDO AI. On an annual basis, this leak would equate to 52.5 MG of water, a significant loss of revenue for the RWA.
