Service reservoir performance assessment: a data driven approach for smart water infrastructure management

Abstract

Service reservoirs (SRs) are critical assets within drinking water distribution systems, yet water service providers (WSPs) lack methods for in-service performance assessment. Using data from 17 SRs and 2 WSP, a novel framework assessing SR integrity and mixing characteristics is developed and validated. While single-parameter analysis yields limited insights, combining level and flow data enabled mass balance calculations to detect integrity issues such as leaks and ingress up to 3 months before the scheduled inspections. Hydraulic residence time calculations, based on flow and level data, identified SRs operating beyond their recommended range with water age, sedimentation, and disinfectant residual consequences. By comparing hydraulic to cross-correlation residence times, calculated using inlet and outlet conductivity data, SRs exhibiting short-circuiting were identified. These were linked to poor mixing behaviour, creating zones of reduced disinfectant residual and higher bacteriological risk. An operational trial was subsequently conducted that demonstrated that by enhancing the SR level cycling, an approach highlighted by self-organising map analysis of data from 329 SRs, a 33% improvement in mixing was obtained. This new framework, using standard network data, facilitates WSPs to transition from reactive and periodic assessment of SRs to proactive and effective real-time smart water infrastructure management.