Evaluation of Prediction Techniques in Corrosion Management of Pipelines External Corrosion

Abstract The most dependable and widely used techniques to inspect and manage pipelines external corrosion are In Line Inspection (ILI) for piggable pipelines and external corrosion direct assessment (ECDA) for non-piggable pipelines. But ILI is a lagging indicator, i.e., it provides information on corrosion after metal loss has happened and some aspects of ECDA may be considered as current indictors but overall, it is a lagging indicator. Therefore, it is essential to integrate lagging indicators with prediction techniques so that pipeline operators can take the required pre-emptive steps and implement corrective actions in a timely manner. This paper reviews concepts and capabilities of various prediction techniques including 5M-Methodology, Machine Learning (ML), Artificial Intelligence (AI) and Deep Learning (DL) in managing external corrosion of oil and gas pipelines and describes the experience of implementing the 5-M Methodology in four (4) oil and gas pipelines.