WWEM - Abstract

Abstract Information :

A sewerage rising main (or pumping main) is a pressurised pipe used to convey wastewater from a pumping station up to higher ground in situations where onward flow by gravity is not feasible. As with other pressurised water pipes, rising mains can be subject to bursts (or breaks) either through poor installation or age-related deterioration.
Although many pumping stations have monitoring in place for pump status, pressure and/or flow, automated monitoring systems are often unable to identify the occurrence of a burst. In many cases the water company is only made aware of the issue many hours later when it is reported by a member of the public, by which time a significant pollution event may have occurred. As well as the obvious environmental and reputational damage, such pollution events can lead to the imposition of large legal and regulatory penalties, as well as the need for expensive clean-up operations.
The presentation will show how cloud-based machine learning tools can be used to identify rising main bursts from existing monitoring data within 30 minutes of occurrence, allowing water company asset managers to be immediately alerted when a burst is detected. Following a rapid review of the relevant data to confirm the event, pumping can be stopped remotely, and a response team deployed to prevent any further environmental damage.
A water company case study will be presented to show the performance of the analytical algorithm, and the value that it is able to provide to the water company by avoiding or minimising significant pollution events.
The presentation will show how the analytics deployed are able to detect bursts on pumping stations with only the bare minimum of monitoring in place, providing immediate benefits across the whole rising main estate without the need for capital investment.