Water & Wastewater Treatment

www.wwtonline.co.uk | WWT | JULY 2015 | 37 In the know Technically speaking: Smart networks S mart is the new buzzword in utilities. There is a lot of discussion about what smart actually is, but whatever you think it is, data is at its heart. However, it is intelligence, rather than data, that enables operators to make decisions - so how does data become actionable intelligence? In this article we will look at how data can assist pressure management, why data needs to be used to trigger alarms for burst detection, how additional information can be captured to monitor asset health, and how basic statistical data can be used to add further intelligence. To describe the principles we will look at the pressure and • ow data that we could collect for a DMA (District Meter Area) that has an inlet, with a meter and PRV (Pressure Regulating Valve) combination. Smart pressure management Firstly let's look at pressure management. Each DMA will have its own unique pressure - demand (• ow) pattern. This can change in a DMA with seasonal demand, sporting events, religious festivals etc. Using basic data, it is possible to investigate whether there is a beneˆ t to smart Actionable intelligence Smarter network management requires moving from the collection of unfocused data to specifi c intelligence that can be acted upon pressure management, which can include time regulating the pressures, • ow modulation or continuously automatically optimising pressures. Data becomes even more powerful when the burst data is collected against a PRV that is controlling the pressure in a region. If there are multiple bursts in an area fed from a PRV then the beneˆ ts to reducing the pressure even slightly can prove substantial in reducing burst frequency. The savings in terms of leaks, burst reduction and operational e‰ ciency can be signiˆ cant. This can then form the heart of targeted, smart pressure management strategies. Leakage and burst management Bursts impact on customers in a rapid and direct way that leakage doesn't. A proactive approach to burst management requires rapid data analysis: if data is being uploaded by the loggers every 24 hours, and it is not possible to view the data till the next day, this is too slow to be of any value when it comes to bursts. By using the location of customers that have complained, or using local knowledge, it may be possible to work out where a burst has occurred. A smarter approach is where pressure loggers are scattered throughout a network to indicate areas of low pressure. This will begin to localise the burst and allow for a targeted response. PRV and meter health The requirement in these circumstances is not the level of detail in the data, but the speed with which it is transmitted. If appropriate thresholds are set on loggers, these can trigger the sending of alarms and/or data when the pressure is low and • ows are high, indicating a burst. Ideally the • ow and pressure ˆ ngerprint for the DMA will be known, so the thresholds are proˆ led around the likely pressures and • ows. Logging the data is only part of the overall strategy, and the method of getting the data back to a central location, or sending alerts to MARK SMITH CEO, WRC Fig 1: All four data sets, indicated by coloured lines, have similar 15-minute average pressures MICHAEL BOLD, SENIOR CONSULTANT, WRC

• Cover