Detection of failure potential in the clean water network using multi-criteria decision-making (MCDM) methods: the case of Yenişehir, Mersin

Turkey, which was considered a water-rich country before the 2000s, is now among the countries facing water scarcity. The primary causes of water issues include poor water management, deficiencies in water policies, increasing water demand across all sectors due to population growth, and the impacts of global climate change. Water losses in distribution networks are a common problem for many cities and countries worldwide. In Turkey, as in many other nations, the average annual water losses exceed 50%. Due to the rapid increase in water demand and changing climatic conditions, water supply problems are becoming more frequent globally. Therefore, reducing and controlling water losses is of critical importance. To use our water resources more efficiently, it is essential to identify water loss rates and develop methods and technologies to reduce them. Predicting water pipe failures before they occur is particularly crucial. In this thesis, the study area was selected as Yenişehir district in Mersin province, located in the Mediterranean Region. To predict water pipe failures, the environmental criteria affecting the pipes were first identified. These criteria include soil type, geology, slope, elevation, aspect, building density, population density, lineament, and distance to roads. The nine criteria were weighted according to their importance using the Frequency Ratio (FR) and Analytic Hierarchy Process (AHP) methods. This approach introduced an innovative method to the literature, and the resulting model was validated against actual failure data, achieving an accuracy of 92.40%.