Search

Loss of water due to leakage is a common phenomenon observed in all water distribution systems. However, the water loss can be reduced to an acceptable limit by improved understanding of leakage mechanisms and pathways, and the influencing factors that lead to leakage. This study proposes a methodology to estimate leakage potential (LP) in a water distribution system (WDS). The study has identified various factors which are directly and/or indirectly associated with LP in a WDS. The research has defined a selected set of influencing factors as fuzzy variables and has developed a fuzzy rule-based (FRB) model using Mamdani inferencing algorithm. The impact of influencing factors varies under different operating pressure, therefore a generalized methodology has been proposed. The proposed methodology has been implemented for a real WDS in Bangkok (Thailand). To evaluate the impacts of uncertainties of influencing factors on LP, Monte Carlo simulation-based sensitivity analysis has been carried out. It has been concluded that the operating system pressure and age of the network are the two most important factors that contribute to LP of a WDS. The proposed model will help water utility managers to prioritize their network rehabilitation strategies and to establishing an efficient active leakage control program. [ABSTRACT FROM AUTHOR]

Selecting and prioritizing monitoring locations (zones) in a water distribution network (DN) involves complex decision-making that warrants the use of risk-based decision-making. This paper couples the concept of risk with fuzzy sets and provides a simple approach to prioritizing monitoring locations in a DN. An index-based approach is proposed to estimate vulnerability, sensitivity and risk indices at a given DN location. The approach is based on fuzzy-based techniques that include fuzzy synthetic evaluation and fuzzy rule-base. Two types of analyses (deterministic and probabilistic) are used to demonstrate the proposed approach for a DN in Quebec City (Canada). Fifteen 'input factors' related to structural integrity, hydraulics, water quality and consumer sensitivity are used in the analyses. Based on the risk index results, the DN is divided into two parts. Sensitivity analysis results reveal that four factors, including two sensitivity factors (i.e. % population < 5 and > 75 yr) and two vulnerability factors (free residual chlorine and pipe breaks), explain 85% of the risk index variability. The usefulness of the proposed approach is demonstrated and strategies to improve the model are discussed. [ABSTRACT FROM AUTHOR]