Pressure-Transient Monitoring System Improves Leak Detection and Asset Management for Georgia Utility

With three pump stations operating within a single pressure zone, the system’s flow direction shifts based on demand and input balance around the loop. The network carries water long distances at low velocities to minimize head loss, with pressures ranging from 20 to 230 psi due to elevation differences across the service area. Five elevated storage tanks positioned on ridgelines from southwest to northeast provide additional pressure stability, preventing many pressure-transient events from propagating throughout the network.

In 2016, CCMWA experienced a series of tapping saddle failures on 6- to 12-inch-diameter outlets. Each failure created an orifice flow condition that limited the leak rate to within the capacity of the looped transmission pipe network. Because the resulting head loss was not easily visible in steady-state pressure readings, the utility began exploring new monitoring technologies to better detect and locate these events.

An 8-inch tapping saddle failure occurred on one of CCMWA’s 42-inch mains. By collecting pressure-transient data in a high-sample-rate mode (128 times per second), operators were able to quickly locate, isolate, and identify the leak.

Spotting System Faults

Previously, CCMWA’s method for identifying large faults involved calculating total system demand by comparing high-service pumping rates and tank filling levels, along with rainfall data, to estimate expected demand. For example, during rainy summer periods, demand would typically drop to a minimum and increase again during dry spells.

An analysis of a 12-MGD fault in 2014 showed an additional head loss of only 2 psi in the 36-inch transmission main, which had a design capacity of 140 MGD Further evaluation revealed that even a 10-MGD fault would create only a 5.3 psi (0.37 bar) head loss at the most distant point—making such leaks difficult to spot through steady-state pressure analysis alone.

This insight led CCMWA to consider alternative monitoring technologies. Acoustic monitoring was ruled out as impractical due to the network’s 200-mile length and the density of sensors required. The utility ultimately selected pressure-transient monitoring as the most effective solution for its needs.

In March 2016, CCMWA installed 10 PIPEMINDER-ONE Standard devices spaced 5 to 10 miles apart, focusing on areas with a history of leaks. This advanced pressure and transient monitoring solution from Badger Meter is designed to help water utilities detect, locate, and analyze transient events—such as pressure surges, bursts, or leaks—across their pipeline networks.

Pressure Monitoring in Action

One Friday night, at 7:56 p.m., two abnormal events were detected: a significant and unexplained change in flow direction leaving the treatment plant and an unexpected pressure drop at the plant and another monitoring site. Operators immediately suspected a major leak on the 42-inch line extending east from the plant. Crews were dispatched and located the issue—a failed 8-inch tapping saddle—by 10 p.m.

A comparison of the high-sample-rate pressure data (128 samples per second) with the supervisory control and data acquisition (SCADA) data collected at one-minute intervals demonstrated the advantages of transient monitoring. The high-resolution data clearly showed transient wave travel times and anomalies, while the lower-resolution SCADA data did not provide enough detail to reveal significant events.

Since adopting the pressure monitoring system, CCMWA has developed key analysis methods to interpret transient data more efficiently. This approach enables staff to focus on meaningful events rather than spending time analyzing normal network behavior.

The Power of Triangulation

The term “triangulation” refers to locating the source of major transient events, such as bursts. This automated feature within the RADAR pressure management platform detects pressure waves at multiple monitoring sites and calculates the event’s origin using high-resolution time data.

In June 2019, a major burst triggered an alert. Operators were immediately aware of the incident through the SCADA system but were also able to pinpoint the break’s location using the transient monitoring data before crews arrived on site. This allowed field teams and local authorities to respond directly to the correct location without delay.

Three devices recorded the pressure drop, including one that reached 0 psi. The estimated leak flow was 100 MGD, and pressure drops of up to 188 psi were observed at nearby monitoring points. The triangulation algorithm in RADAR estimated the break’s location within 150 meters of the actual site—an impressive level of precision delivered automatically within minutes.

CCMWA also uses another platform feature known as event clustering, or shape classification. This pattern recognition capability compares transient waveforms against a reference library to classify events, such as pump starts or stops. By identifying and filtering recurring operational transients, the software allows users to focus on anomalies that require investigation, improving overall efficiency.

Green markers indicate the devices that picked up the June 2019 break and were used to automatically triangulate its location. The triangulation algorithm in RADAR estimated the break’s location within 150 meters of the actual site.

A Faster Detection Method

With the monitoring system in place, CCMWA now has a clear, accessible view of pressure-transient data in real time. Before implementation, operators relied solely on SCADA pressure readings, which lacked the resolution needed to identify transient behavior.

Today, the system enables operators to detect fast-developing leaks that once went unnoticed, respond more quickly to network issues, and make informed decisions for asset maintenance and capital planning. By combining accurate monitoring with advanced analytics, CCMWA continues to strengthen its water system reliability and reduce costly main break events.

An earlier version of this article was published in AWWA’s OpFlow magazine.