background image design
Customer Story

Yorkshire Water saves water, meets new regulations and adopts predictive maintenance

Water treatment plant at sunrise
  • Challenge
    Better predict and prevent leaks to improve operations, contain costs, and avoid fines from new regulations
  • Solutions
    Revamp the PI System infrastructure via the Visible Network Project to reduce leakage, customer contacts, and water-supply interruptions
  • Benefits
    Identifying leakage trends faster than ever, reducing costs and improving operations

Yorkshire Water has cast-iron pipework that makes the system prone to leakage, affecting water availability to its five million domestic and 136,000 business customers. A recent leak, which was difficult to find, resulted in over 4,000 customer complaints or inquiries from a single failure. Plus, new regulations meant that the company could be fined more than $4,000 for every leak over a targeted threshold. Clearly, something had to change. The UK’s Water Services Regulation Authority set new regulatory targets taking effect in 2017, and Yorkshire was at risk of missing performance targets. Already, in 2012, the company had deployed the PI System as part of its rNet project to gain real-time insights into water pressure and flow in its pipelines. With 4,500 flow and pressure sensors and flat-line notifications, this was a critical first step toward gaining operational insights. However, with the new regulations, it would need something more.

Prediction quality is looking at around a 96% success factor on the data we’ve got. So, a 30-day turnaround, which would normally take to dig and repair a leak, we can now turn that down in less than 10 days.
Andrew Sewell
Telemetry Manager, Yorkshire Water

A server that is a saving

Yorkshire Water produces 1.3 billion liters of clean water and treats 1 billion liters of water in its wastewater networks every day. The company has 686 treatment facilities spread across northern England and 83,000 kilometers of water pipelines, enough to circle the Earth twice. To mitigate leakage risks, Yorkshire turned to the PI System to take a predictive approach in order to significantly shorten the time it takes to identify and fix a system leak, especially considering that one incident could net 4,000 customer complaints.

“Ultimately, that’s pretty much 65% of our target for the year gone in just one event,” said Andrew Sewell, telemetry manager at Yorkshire Water. Not only that, every leak over the target threshold netted a fine of £3,300 (about $4,400). “So . . . you see how we need to get more predictive with our analytics information to support leaks,” he added.

Reducing water leakage

To detect leaks more quickly and meet the new regulations, Yorkshire implemented the Visible Network Project to reduce leakage, customer contacts, and water-supply interruptions. The project encompassed 26 performance commitments that included reducing leakage and water interruptions by 10 million liters, or 4,000 contacts, per day respectively. To meet these goals, Yorkshire partnered with Capula, an automation and real-time integrator in the United Kingdom, to reevaluate internal processes and revamp its PI System infrastructure.

New visibility into water data

With the help of Capula, Yorkshire implemented new PI System technology, including PI Vision, to gain better visibility into its pipeline infrastructure. The team increased the number of pressure and flow sensors to 6,800 to improve its data profiling and developed five integrated modules to deliver full functionality.

The new asset configurator populates the PI System, and the asset-update function syncs data across the different asset structures and automatically updates Asset Framework (AF). The analytics engine manages flow and pressure models based on seasonal variations and creates Event Frames in AF when a leakage event is spotted. All data is now visualized in the leakage viewer. “Using AF really was a key enabler for the project and allowed it to be realized fully and automated in its data structure,” said Alistair Norman, business-sector manager for the operational-intelligence division at Capula.

PI System detecting leakageFaster leakage detection: The PI System allows Yorkshire to detect leakage problems a day and a half before the threshold alarm identifies the issue.

Predicting events four days faster

With the new PI System, Yorkshire is now able to identify developing leakage trends faster than ever before. Now, leakage events are detected 1.5 days before reaching the threshold, and users receive alerts by profile alarms. This is a full 4.5 days faster than previous alerts. “You can see from these . . . examples that significant time would be saved, and dispatching teams to site to investigate the leakage event would have been a lot quicker,” Sewell noted. Thanks to the PI System, Yorkshire is now on track to achieve its targets.

Given the success of the project, Yorkshire is looking to integrate other operations data, such as energy and temperature, into the PI System model as well as integrate this information into its work planning system and to connect it to mobile devices.

For more information about Yorkshire Water and the PI System, watch the full presentation here.

©2021 OSIsoft, LLC