Predicting corrosion remaining life of underground pipelines with a mechanically-based probabilistic model

A methodology is presented for predicting corrosion remaining life of underground pipelines with a mechanically-based probabilistic model by taking effect of randomness into account in pipeline corrosion. Monte Carlo simulation technique is employed to calculate the remaining life and its cumulative distribution function (CDF). The sensitivity analysis is performed to identify the most important parameters that affect pipeline failure. The results show that the corrosion defect depth and radial corrosion rate are the key factors influencing pipeline failure probability and remaining life. The pipeline remaining life can be prolonged greatly by reducing mean value of corrosion defect depth and radial corrosion rate. CDF is more appropriate to characterize the pipeline failure probability compared to probability density function (PDF) and reliability index. (c) 2009 Elsevier B.V. All rights reserved.