Probabilistic models for condition assessment of oil and gas pipelines

The paper is primarily concerned with the interpretation of in-line inspection data collected using magnetic flux leakage tools to characterize the actual condition of pipelines vulnerable to metal loss corrosion. The paper presents a probabilistic analysis framework to estimate the pipeline reliability incorporating the impact of inspection and repair activities planned over the service life. The framework is applied to determine the optimal inspection interval and the repair strategy that would satisfy a target reliability requirement. To update the pipeline failure probability after maintenance, a practical approximation is developed and validated using Monte Carlo simulation results.