Predicting the Failure Probability and Reliability Based Design, Optimization for Pipelines

Abstract

A probabilistic model is developed in this study for calculating the failure probability and finding the associated reliability for a given pipeline in a plant affected by the time-dependent failure by corrosion. This is accomplished by applying a probabilistic approach to the Shell-92 pipeline failure prediction model. The Monte Carlo simulation scheme is used to predict the probability of failure. The life cycle cost is also estimated based on predicted probability density function, which shows that failure cost is dominating the maintenance cost in overall life cycle cost. A sensitivity analysis is also carried out to investigate the influence of corrosion parameters on the probability of failure which reflects the effect of axial-radial corrosion rates and depth-length of corrosion defect on failure probability. At the end, optimized corrosion rates are also predicted using the reliability based design and optimization (RBDO) technique, in which the probabilistic constraint has been utilized based on the Shell-92 pipeline failure prediction model.