An Intelligent Approach for the Condition Assessment of Watermains

Abstract

Frequent occurrences of pipe failure pose a huge threat to potable water security worldwide. The condition assessment of watermains is one of the key strategies that can pinpoint risky pipes and maintain their sustainability. Intelligent systems such as fuzzy inference system (FIS) and adaptive neuro-fuzzy inference system (ANFIS) have proved their efficacy in simulating and predicting intricate water infrastructure problems. This research paper proposes a novel methodology for the development of a risk scale, along with the evaluation and quantification of water network’s condition index. The Arequipa region in Peru that comprises eight provinces is chosen to demonstrate the proposed methodology due to the fast pace of urban sprawl, as well as the economic boom that make sustaining underground pipelines a difficult task. The methodology builds on various algorithms, computational intelligence and interactions between different variables. It involves developing two intelligent models; the first is the ANFIS model that is designed to estimate the watermains condition index of each province through the grid partitioning and hybrid optimization function. Several neuro-fuzzy networks are created and tested through different statistical indicators to select the optimal network that can be used to predict the condition indices of each province. The produced condition indices are then streamlined and entered into the FIS engine to develop the second (FIS) model, which is built on the basis of Mamdani system. The FIS engine runs an iterative simulation process through which the input variables are fuzzified, fuzzy rules are evaluated, outputs are aggregated, and results are de-fuzzified. Finally, the fuzzy consolidator generates one crisp number that represents the water network condition index of the region. The resulted risk scale indicates that the condition of water distribution networks of the Arequipa region is medium, in accordance to the questionnaire of professionals and field experts. This research provides insights for infrastructure managers concerning their maintenance, replacement or rehabilitation plans.