Author_Institution :
Dept. of Chem., Phys. & Environ., Univ. of Udine, Udine, Italy
Abstract :
In this paper we address the problem of pressure management in water distribution systems. We present a methodology based on single- and multi-objective genetic algorithms (GAs) which, starting from the numerical model of the network, performs calibration and then optimizes the location and control of pressure reducing valves (PRVs), according to some operational constraints that must be satisfied. In particular, the simulation model is calibrated with a real-coded, single-objective GA in order to obtain, on one hand, pipe roughness coefficient values and, on the other, an estimate of the subdivision of the total flow supplied to the network between actual customers demand and water loss. The multi-objective NSGA-II is implemented in order to find the Pareto optimal solutions representing different level of compromise between installation costs and leakage reduction. The application of the methodology to a real network allowed considerable water savings, as evidentiated by the monitoring of the system. Two more advantages are also evident: firstly, the number of interventions for repairing pipes is more than halved, due to the reduced pressure regime (thus enabling the water utility to provide better service more efficiently and reliably); secondly, the surplus water is being diverted to a storage tank of a pumped network, thus allowing a notable reduction of pumping costs.
Keywords :
Pareto optimisation; condition monitoring; cost reduction; genetic algorithms; maintenance engineering; pipelines; pressure control; tanks (containers); valves; water storage; water supply; Pareto optimal solutions; customers demand; energy cost reduction; leakage reduction; multiobjective NSGA-II; multiobjective genetic algorithms; numerical model; optimal pressure management; pipe repairing; pipe roughness coefficient values; pressure reducing valves; pumped network; water distribution systems; water networks; water storage tank; Calibration; Genetic algorithms; Mathematical model; Optimization; Storage tanks; Valves; Water resources; Water supply; leakages; optimization; valves;