Uncertainty analysis of water quality in water distribution system

Abstract

Water quality simulation is affected by uncertain parameters such as pipe roughness coefficients, chlorine bulk decay coefficients, and chlorine wall decay coefficients, which are usually considered to be fuzzy variables. The minimum and maximum nodal chlorine concentrations and water ages at each α-cut level were obtained by the genetic algorithm (GA) based on EPANET hydraulic and water quality simulation toolkit. The fuzziness of nodal chlorine concentrations and water ages were measured using the fuzziness measure (FM) proposed in this paper. The method was applied to four networks to analyze the fuzziness of nodal chlorine concentrations and water ages. The results indicated that the distribution of nodal chlorine concentrations does not follow typical trapezoid distribution, while the distribution of nodal water ages follows typical trapezoid distribution. In addition, the chlorine concentration and water ages of nodes farther from the source are affected by uncertain parameters to a greater extent. The greater demands of nodes lead to less effects of uncertain parameters on chlorine concentration, and greater effects of uncertain parameters on water ages. This study would help in analyzing the fuzziness of hydraulic and water quality simulation results in WDS under uncertain conditions.