Probabilistic modeling and optimization of leakages in water distribution networks

• Κύριος συγγραφέας: Σεραφείμ, Αθανάσιος
• Άλλοι συγγραφείς: Serafeim, Athanasios V.
• Γλώσσα: English
• Έκδοση: 2022
• Θέματα: Water distribution networks; Minimum night flow; Confidence interval estimation; Water losses; Water balance; Leakage; Parametric modeling; Pressure effect; Statistical clustering; Water networks partitioning; Leakage management; Hydraulic resilience; Pressure management areas; Δίκτυα ύδρευσης; Απώλειες νερού; Ζωνοποίηση δικτύων ύδρευσης; Ζώνες διαχείρισης πίεσης; Παραμετρική μοντελοποίηση; Ελάχιστη νυχτερινή ροή; Εκτίμηση διαστημάτων εμπιστοσύνης; Υδατικό ισοζύγιο; Διαρροές

We develop an integrated framework for resilient reduction of leakages in water distribution networks (WDNs), which combines: a) a set of probabilistic approaches for minimum night flow (MNF) estimation and parametric modeling of water losses in WDNs, and b) a combination of statistical clustering and hydraulic modeling techniques for WDN partitioning into pressure management areas (PMAs; or districted metered areas, DMAs). The strong point of the proposed probabilistic methods is that they allow for confidence interval estimation of both observed and parametrized MNFs, while taking into account the effect of operational pressures on leakages. The statistically rigorous and user unbiased clustering approach produces realistic PMA (or DMA) configurations, allowing water experts and engineers to develop effective strategies for management and reduction of leakages, based on a fast and easy to implement algorithmic procedure. The efficiency of the developed framework is tested in the water distribution network of the City of Patras in western Greece, which serves approximately 213 000 consumers, consists of more than 700 km of pipeline grid (mainly HDPE and PVC pipes), and includes 86 pressure management areas (PMAs) equipped with automated local stations for water flow rates monitoring and pressure regulation.