Pump Scheduling for a Large Water Distribution Network. Milan, Italy

Abstract

Large water distribution networks require efficient use of their resources. One of the ways to become more efficient is to reduce the energy consumption due to pumping systems [1] [2]. In the European context the city of Milan has a large water supply system for 1.3 million inhabitants and around 4.0 million commuters, which is supplied entirely by 26 pumping stations. The system currently supplies its ∼50,000 customers with 103 pumps which are actively operated during the day [3]. In previous years a pump scheduling algorithm has been proposed to the utility for a Pressure Management Zone (PMZ) in the south of the city named Abbiategrasso containing only 4 pumps [4] [5]. However, it is of the interest for the utility to extend the analysis to the whole system [6] [7]. For that reason it is necessary to perform a proper pump scheduling. The solution proposed here, is a Multi-Objective Optimization (MOO) for the energy consumption reduction of the whole Water Distribution Network (WDN) using an EPANET model of the whole network. Results show that there is room for improvement of energy and pressure management in the system. The solution presented here can be applied to other utilities with similar challenges.