Linear Model for Estimating Water Distribution System Reliability

Abstract

The water distribution system (WDS) reliability is the ability of a network to provide the required quantity of water to customers under uncertain system conditions. In this study, eight surrogate measures of WDS reliability were investigated: Two hydraulic reliability measures, two robustness measures, single failure reliability (SFRel) with three valve installation levels, and seismic reliability (SeisRel). The eight reliability measures were calculated for 16 study networks of various sizes (i.e.,in terms of the number of nodes and links), numbers of sources, layouts, demand distributions, and topographies. The network characteristics (e.g.,structural redundancy and overall pipe sizes) were quantified by using system characteristic indicators (SCI), and their correlations with the eight suggested reliability measures were analyzed. The following SCIs were considered: The average node degree, meshedness coefficient, total system demand, total system demand per source, length-weighted average pipe diameter, and total service area. Finally, univariate and multivariate linear reliability models were developed by fitting a linear regression line to the scatter plots of the reliability measures and SCIs.