An Economical Decision Model for Water Pipe Replacement Using the Bayesian Theory

According to the limited annual budget for pipeline replacement, the most effective and economical plan is needed to available fund. A variety of factors contribute to pipe breaks, including pipe material, diameter, outer coating, operation pressure, inner coating, load and bedding condition. To consolidate information of those factors to decide pipe replacement plan, this study developed a model using Bayesian theory to predict pipe life considering economical cost. An example of this Bayesian model with pipe operation data was presented. Then the influence of each factor on pipe life was estimated. Results suggested pipe life was the more sensitive to the factor of pipe material, inner coating and outer coating. Other factors with smaller weight values and relative stable posterior means, would have less influence on pipe life. This Bayesian model predict pipe in an accuracy way by combined experience estimation and practical information.