A hybrid population-based degradation model for pipeline pitting corrosion

This paper presents a novel algorithm to develop a population-based pitting corrosion degradation model for piggable oil and gas pipelines. The algorithm is designed to estimate and predict the distribution of actual depth of existing pits on a pipeline segment, given two or more sets of in-line inspection data that have uncertainty in size and number of the detected pits. This algorithm eliminates the need for a defect-matching procedure for those pits that are not critical, that is required in developing defect-based pitting corrosion degradation models. A hierarchical Bayesian model based on a non-homogeneous gamma process is developed to fuse the uncertain in-line inspection data and physics of failure knowledge of pitting corrosion process. Measurement error (ME), probability of detection (POD), and probability of false call (POFC) are addressed in the developed algorithm. The application of the developed algorithm is demonstrated by implementing it on a simulated case study and the results are compared with the simulated data from a generic degradation model that is available in the literature. Results indicate that this algorithm can predict the degradation level of the pipeline with a high accuracy.

Automated summary

This paper presents a novel algorithm for developing a population-based corrosion degradation model for oil and gas pipelines. It eliminates the need for defect-matching procedures for non-critical pits and uses a hierarchical Bayesian model to combine uncertain in-line inspection data and physics of failure knowledge. The algorithm successfully predicts pipeline degradation levels with high accuracy.