Identification of the cause of variability of probability of failure for burst models recommended by codes/standards

Failure probability of oil and gas pipelines due to external corrosion defects can be estimated using corrosion growth model and the evaluation of remaining strength. Codes/standards have been developed for the assessment of the remaining strength of corroded pipeline. The remaining strength and the operating pressure were considered to develop the limit state equation and consequently the failure probability of the burst models recommended by codes/standards. In the present paper, comparative analyses of the failure probability estimated by the codes/standards were conducted, using Monte Carlo simulation and first order second moment methods. The analysis revealed that the failure probability of the burst models recommended by codes/standards varies significantly for the same defects size. The study further explored the cause of variability in failure probabilities. The study observed that different defect shape specifications (rectangular, parabolic, etc.) and different stress concentration factor derivations (different contributions of l) for burst pressure estimation are responsible for high variability in the probability of failure. It is important to reduce variability to ensure unified risk-based design approach considering any codes/standards. © 2011 American Society of Mechanical Engineers.