Probabilistic Minimum Night Flow Estimation in Water Distribution Networks and Comparison with the Water Balance Approach: Large-Scale Application to the City Center of Patras in Western Greece

Quantification of water losses (WL) in distribution networks (WDNs) is a crucial task towards the development proper strategies to reduce them. Currently, WL estimation methods rely on semi-empirical assumptions and different implementation that increase uncertainty obtained estimates. In this work, we compare effectiveness robustness two widely applied approaches found international literature: (a) balance, or top-down, approach introduced by International Water Association (IWA), (b) bottom-up minimum night flow (MNF) approach, based recently proposed probabilistic MNF method. doing so, use users’ consumption flow-pressure data from 4 largest pressure management areas (PMAs) WDN city Patras (the third Greece), which consist more than 200 km pipeline, cover entire center Patras, serve approximately 58,000 consumers. The results show that: when done rigorous statistical setting high resolution timeseries, there sufficient understanding types patterns during day hours, effectively converge, allowing for reliable individual components. addition, timeseries are available at inlets PMAs, suggested version with should be preferred as less sensitive, while confidence interval components