Minimizing residence times by rerouting flows to improve water quality in distribution networks

Abstract

Management of water quality is a major issue for water companies, especially as many systems are old and have excess capacity. A methodology which uses an evolutionary algorithm to minimize water age, and hence improve water quality, is presented in this article. A steady-state model is used to find the water age at various nodes of a network. Three parameters are derived from these nodal age values to represent quality for the entire network. The evolutionary algorithm reconfigures the network by selecting a set of pipes for closure. The optimal network configuration is achieved when the chosen water age parameter is minimized subject to maintaining connectivity and hydraulic feasibility in the network. The methodology is applied to an example network to identify the age parameter that best represents quality over the entire network. The evolutionary model is then applied to re-route flows in a real water distribution network and the results are compared with those from the unmodified network. The validity of the use of steady-state hydraulics is tested by conducting an extended period simulation (EPS) on these results.

Keywords:

• Water distribution networks
• Water quality
• Optimization
• Evolutionary computation

Acknowledgements

The work reported in this article was supported by UK Engineering and Physical Sciences Research Council Grant no. GR/R67989.