Research On Water Pipe Risk Assessment Using Bayesian Theory

An accurate risk assessment of a pipe is important for maintaining acceptable levels of service and providing appropriate strategies for maintenance and rehabilitation in water supply systems. A risk assessment system was developed in this paper according to the threat during pipe operation.Firstly, understanding of the risk assessment was expressed; the risk was equaled to the product of the threat possibility, vulnerability and importance. The threats studied in this paper were load and corrosion. According to the character of the threats, this system combined the assessment of pipe vulnerability and threat possibility. It meant to examine the pipe condition. Based on the result of pipe condition, we added the element-the importance of pipe to generate pipe risk assessment system.Secondly, deterioration point assignment was used to assess pipe condition. Many factors contribute to pipe deterioration. To consolidate information on these factors to assess the condition of water pipes, this study employed a new approach based on Bayesian configuration against pipe condition to generate factor weights. Eleven pipe factors were used in this study. Informative prior factor weight distributions were based on the literature and previous knowledge. The influence of each factor on pipe assessment was estimated. A traditional pipe assessment method was compared to this Bayesian model. The pipe importance was calculated by analyzing the flow and position of pipe. Therefore the system of pipe risk assessment was obtained.Thirdly, a risk assessment system for cast iron pipes was developed. The condition of cast iron pipes was divided into 3 ratings: good condition, regular monitoring and priority rehab. The relative weights of five condition indicators and three ratings of the indicators were identified based on literature and previous study. Thus the priori probability of each rating for the assessing pipe was obtained. Then the probability of each rating for the pipe was calculated by Bayesian method, with the combination of difference between experience judgment and pipe operation condition; consequently the condition of the cast iron pipe was obtained. The severity of the risk was also divided into two rates. By examine the pipe condition and severity rates; the risk of cast iron pipe was obtained.Finally, the system of cast iron pipe risk assessment was applied to a district in Huabei area, and showed an excellent effect. The proposed Bayesian inference approach provided a reliable decision suggestion for water pipe management.