Evaluating the Failure Probability of the Water Distribution Networks Quality in Order to Estimate Risk Using Bayesian Networks

Water distribution networks are one of the vital infrastructures of any country, which one of its main tasks is to deliver good quality water to consumers. These networks are always exposed to various types of threats, and since they are at the end of the water supply chain and are in direct contact with the consumer, the occurrence of quality failure in it can lead to irreparable health, life and social damages. The acceptable performance of water supply networks in terms of quality is dependent on operation and maintenance programs, and risk analysis and management can be mentioned as one of these programs. One of the main components of risk due to the occurrence of quality failure is the probability of failure. In this research, in order to calculate the probability of failure of the quality of water networks, at first, the factors affecting the occurrence of quality failure are identified and with the help of Bayesian networks, the probability of failure of the quality of each pipe is calculated. This work is done in four basic steps: preparing input data, training the Bayesian network, validating it and receiving the output results. In addition, by carrying out sensitivity analysis, four factors consist of pipe type, water pressure, water age and water velocity have been identified as the most effective factors on the occurrence of water quality failure in this area; which controlling them by managers and decision makers of water and sewage companies in the form of risk management programs, the probability of quality failure can be significantly reduced.