June 16, 2025

AI helps prevent major water main breaks

AI helps prevent major water main breaks

Preventing disruptive water main breaks

A large city in the South East used new AI leak detection technology to reveal 50 previously undetected leaks within its drinking water network, enabling repair crews to save over 167M gallons of non-revenue water and prevent costly and disruptive major water main breaks. 

The challenge

A major city in the South East was grappling with aging water supply infrastructure, with many of its thousands of miles of underground pipes more than 80 years old. These aging pipes are prone to leaks, which can result in major water main breaks that cause localized flooding that disrupts homes, businesses and essential services such as hospitals and law enforcement. 

Following two catastrophic and high-profile breaks, the city urgently needed a way to quickly and reliably identify and locate other potential developing leaks before they could erupt and cause further damage and disruption. 

The solution: CivilSense™ AI-driven leak detection 

City officials approached Oldcastle Infrastructure to use CivilSense™ AI-driven leak detection to support its burst mains response program, asking the team to deploy it rapidly across the 70 miles of its water distribution network that it considered to be of elevated break risk. 

The Oldcastle team deployed CivilSense™ in two phases, analyzing network GIS data to identify optimal locations for leak detection data sampling. Sensors were deployed across the network at a total of 223 accessible main line control valves and service curb stops.

In response to the city’s need for urgency, the Oldcastle team expedited data collection by opting for remote sensor data upload. 

The received data was analysed by the CivilSense™ AI, powered by FIDO Tech, to identify probable leak locations and rank them by volume of water lost. 

Each location was further investigated by our expert field team, who used correlation and top sounding techniques to pinpoint, locate and mark up all the significant leaks for repair by the city.

The outcome: reduced water main break risk

CivilSense™ successfully detected and located a total of 47 utility network leaks, along with a further three customer side leaks. These included two large main line leaks, one of which was close to a major road intersection in a busy area of the city. Each of these large leaks was estimated as losing 10 gallons of water a minute.

The city was able to use this data to carry out emergency repairs and to prioritize scheduled water infrastructure upgrade and repair work to address the higher-risk pipe sections first. 

Using AWWA nominal volumes, the total volume of non-revenue water being lost was calculated at 320 gallons per minute. This equates to around 460,000 gallons per day, or a total of almost 168,000,000 gallons every year. 

Each of these leaks represents a persistent risk of a water main break risk, so by addressing these leaks, the city helped to prevent the kinds of major main bursts that cause cost, damage and disruption to local people, businesses and essential services.

Download This Case Study

Learn more