Fuzzy logic-based efficiency index for water supply systems in Northeastern Brazil

Abstract

Water supply systems (WSS) face high levels of water losses and energy consumption, leading to substantial operational costs and inefficient use of natural resources. This scenario prompted the enactment of Brazil’s New Legal Framework for Sanitation, which established ambitious targets for service efficiency and universal access, reinforcing the need for robust and reliable performance assessment tools. To provide a comprehensive evaluation of system performance, this study develops a global efficiency index for assessing water supply services in municipalities across the Northeast Region of Brazil. The proposed methodology integrates official datasets, detects and removes outliers using the Isolation Forest algorithm, and combines multiple performance indicators through a fuzzy logicbased model, with membership functions defined according to the statistical distribution of the variables. The resulting index, termed the Water Supply System Assessment Index (IASA), revealed substantial disparities in efficiency among the states of the region, with Bahia achieving the highest average performance and Maranhão exhibiting the lowest efficiency levels. The analysis by population size showed that larger municipalities generally achieved higher energy efficiency but faced greater challenges in controlling water losses. Correlation analysis confirmed the internal consistency of the proposed model, demonstrating an inverse relationship between the IASA and water loss indicators. The findings provide public managers and water utilities with a practical decision-support tool, enabling more efficient resource allocation and supporting strategic planning aimed at improving the operational performance of WSS.