Novel Methods for Ranking District Metered Areas for Water Distribution Network Maintenance Scheduling

Abstract

To prevent the accumulation of material in pipes which leads to the potential for discolouration events to occur, UK water companies often operate five year cleaning schedules. To organise the schedule District Metered Areas (DMAs), the case study water company assigns a score based on several key performance indicators and water quality levels, which are used to place each DMA into one of three categories: good, poor and urgent. This paper investigates alternative methods of ranking DMAs in order to generate better maintenance schedules. We demonstrate how DMAs can be both partially and totally ordered with methods from multi-objective optimisation, and show how it is possible to prioritise and progressively apply Discolouration Propensity Modelling (DPM) to help guide interventions in the most effective and efficient way. Results obtained from sample DMAs show a good correlation between the DPM scores and the rankings produced by the multiobjective methods. We apply both methods to water networks from a UK water company and demonstrate that used in combination the power index and DPM, have advantages over the current ranking method.