Quantifying Background Leakage in Water Distribution Networks Using High-Resolution Demand Data

<p dir="ltr">Leakage in water distribution networks (WDNs) remains a persistent and costly challenge, with background leakage—small, continuous losses—often accounting for a substantial but uncertain portion of total water loss. These subtle leaks are notoriously difficult to detect using conventional techniques. This study, conducted as part of the UK’s Ofwat-funded Managing Background Leakage project, introduces a methodology that leverages high-resolution customer demand data from Stop.Watch loggers deployed across 22 District Metered Areas (DMAs) in the UK. By applying spatial extrapolation and scaling techniques, the approach estimates total household consumption, which is then integrated with DMA inflow measurements to refine the mass balance. The enhanced mass balance allows for precise disaggregation of water use into legitimate customer night consumption, measured customer-side leakage, and unaccounted for water, providing a clearer picture of leakage dynamics. The results demonstrate that high-resolution sensors significantly improve the accuracy of quantifying network leakage components, offering a promising pathway for utilities to reduce water losses and improve system efficiency.</p><p dir="ltr">This paper was presented at the 21st Computing and Control in the Water Industry Conference (CCWI 2025) at the University of Sheffield (1st - 3rd September 2025).</p>