One main cause is leaks. Twenty to 30 per cent of treated water is lost in systems because of this simple and fixable problem.
Repairs need to be as precise as possible because excavation and resurfacing is a costly undertaking. Digging up more than one location, or more area than is needed for the repair, can lead to a problematic domino effect including traffic disruption, commuter frustration and loss of business.
Meanwhile, there are major public health risks associated with contaminants entering the water system through holes in pipes.
Luckily, researchers from Concordia have an innovative solution. In an article recently published by the American Society of Civil Engineers, Tarek Zayed, professor in the Department of Building, Civil and Environmental Engineering, shows how a special tool called a noise logger can detect leaks accurately and efficiently, before major roadwork is required.
“This approach can reduce the duration of a leak, as well as the cost and time involved in locating the site in need of repair,” says Zayed, who co-wrote the article with post-doctoral fellow Mohammed S. El-Abbassy, recent graduate Fadi Mosleh and Ahmed Senouci from the University of Houston and Qatar University.
For the study, the researchers went all the way to Doha, Qatar to test their theories. The small nation has one of the lowest precipitation rates in the world, as well as one of the highest evaporation rates — meaning the little rain that falls is quickly reabsorbed by the atmosphere as water vapour.
“Qatar is currently facing significant challenges regarding its water supply,” explains Zayed. “Its water distribution network currently suffers from 30 to 35 per cent water loss due to leakage.”
Working on-site at Qatar University, the team installed the noise loggers along the institution’s main water network and used them to record the constant noise generated by a leak over a two-hour time period. They then analyzed the readings, comparing sound level and sound spread. A consistent anomaly meant a leak investigation was required.
Over several weeks of monitoring they collected data from across 140 different points. They then ran simulations using mathematical models to determine the location of the leaks. The facilities management team at Qatar University reported back on the actual locations and found that the team had estimated with 99.5 per cent accuracy.
For Zayed and his team, the next step is to collect leak-data surveys of real-life pipelines from municipalities that use noise loggers and develop customized leak location prediction models.