Abstract: Monte-Carlo sampling technique is employed to investigate relationship among random variables such as crack depth, ratio of crack depth to length, fracture toughness, yield strength, transmission pressure and failure probability of oil and gas pipeline with defects. And criticality importance analysis of fault tree analysis is used to evaluate sensitivity of the random variables. The results show that the fracture toughness and crack depth have significant effect on failure probability of pipeline, as well as the ratio of crack depth to length. In addition, the influence of strength-mismatched factor of welded joint is considered. And annual failure probabilities of different failure modes are evaluated.

Key words: Failure mode, Failure probability, Monte-Carlo sampling technique, Pipeline, Sensitivity analysis, Strength match, Cyber-Physical System Models, Online Monitoring and Control, Product Quality