Optimal design of a new water distribution network using a water supply risk assessment

Taeho Choi, Cheolho Bae, Hwisu Shin, Jayong Koo

Research output: Contribution to journalArticlepeer-review

2 Scopus citations

Abstract

This study considered water supply risk (WSR) as well as the cost of a water distribution network (WDN) design, and developed an optimized model of pipe diameters and gate valves within the WDN. Thus, a multipurpose algorithm was used to implement an optimal WDN design that minimizes both WDN establishment costs and the Block WSR. The algorithm adopted was non-dominated sorting genetic algorithm II, which has most frequently been used. The optimal model was applied to Block A2, and the results showed that the Block WSR of each pipe after ProbPB was changed to 0.2 (case/km/year) from that of the existing design method. As a result, the Block WSR of the existing network design was 0.306 m3/year, but that of the optimal network design was reduced by 10.1% to 0.275 m3/year. The construction cost of the existing network design was 139,600,000 won, but that of the optimal network design was reduced by 59.7% to 56,300,000 won. While the construction cost of the network with the optimal design is much lower than that of the network with the existing design, the optimal design method can also reduce the Block WSR; thus, the optimal design model was proven effective. In addition, apart from construction costs, the Block WSR reduction effect due to the optimal design showed that the daily water cut-off of 8.8 persons/year can be prevented by the conversion of 296 L/Lpcd.

Original languageEnglish
Pages (from-to)153-162
Number of pages10
JournalDesalination and Water Treatment
Volume65
DOIs
StatePublished - Feb 2017

Keywords

  • Interpretive structural modeling
  • NSGA II
  • Optimal design
  • Water distribution network
  • Water supply risk

Fingerprint

Dive into the research topics of 'Optimal design of a new water distribution network using a water supply risk assessment'. Together they form a unique fingerprint.

Cite this