Cluster Analysis to Identify Priority Areas for Water Quality Management in Water Supply Systems

Abstract

Zoning methodologies, such as district metered areas (DMA), are commonly employed to robustly maintain water pipe network systems in both normal and abnormal situations. However, the management and evaluation typically associated with existing DMAs primarily focus on ensuring stable water volume and pressure. Consequently, these methods do not adequately address the maintenance of water quality elements within the water supply system, such as adequately managing residual chlorine and reducing water quality complaints. This study introduced a zoning tailored explicitly for managing water quality-oriented elements, facilitating stable water quality management, and enhancing responses to water quality incidents in large-scale domestic water supply networks. A method was proposed to establish priorities for each zone, using various geographic information system (GIS)-based water quality-related structured data (such as water quality measurement data and pipe data) and unstructured data (such as water quality complaints). Comprehensive water quality management was achieved by applying machine learning techniques based on clustering analysis to derive evaluation factors. The proposed methodology was implemented in Metropolitan City A in Korea, leading to the derivation and analysis of evaluation results. This data-centric water supply network priority management area designation methodology, as presented in this study, is anticipated to serve as a valuable decision-making tool for enhancing the accuracy and reliability of water supply network operation and the overall management of water supply operators.