Solution for water distribution systems under pressure-deficient conditions

Wah Khim Ang, Paul W. Jowitt

    Research output: Contribution to journalArticlepeer-review

    116 Citations (Scopus)

    Abstract

    In recent years, many researchers have tried to predict the behavior of the water distribution systems under pressure-deficient conditions. The root of the problem is that under these conditions the traditional demand-driven analysis will compute heads below the minimum required for outflow to occur physically at some or all of the nodes. The purpose of this paper is to present a novel algorithm for the solution of a water distribution network under pressure-deficient conditions, and is termed the pressure-deficient network algorithm (PDNA). The proposed model progressively introduces a set of artificial reservoirs into the network to initiate nodal flows, with the ultimate replacement of such reservoirs by full demand loads once it has become clear that the nodal flow can be satisfied. The foundation for the solution methodology is established using a series network for ease of discussion. The PDNA is presented in the form for coding into a computer program. For solving the flows in a looped network, the PDNA has to be used with a hydraulic network solver, as manual computation is too time consuming. Using the EPANET 2 hydraulic network solver, the PDNA is applied to both a single-source and a multiple-source network. The results show that the behavior of a water distribution system under pressure-deficient conditions is complex and nonintuitive. © ASCE.

    Original languageEnglish
    Pages (from-to)175-182
    Number of pages8
    JournalJournal of Water Resources Planning and Management
    Volume132
    Issue number3
    DOIs
    Publication statusPublished - May 2006

    Keywords

    • Networks
    • System reliability
    • Water distribution systems

    Fingerprint

    Dive into the research topics of 'Solution for water distribution systems under pressure-deficient conditions'. Together they form a unique fingerprint.

    Cite this