Abstract
Pitting corrosion is a primary type of external corrosion that poses a critical challenge in the oil and gas industry, potentially leading to severe environmental, health, and economic consequences if left unaddressed. Researchers have attempted to comprehend and forecast it through diverse methods, including machine learning and artificial intelligence. However, achieving a precise estimation of pitting growth while considering the influence of various environmental and operational factors remains a concern. This study aims to develop a model by employing tree-based machine learning algorithms, including Decision Tree, Random Forest, LightGBM, CatBoost, and XGBoost, to enhance the prediction accuracy of the maximum depth of pitting corrosion in buried transmission pipelines while also facilitating interpretability. The Random Forest model is found to be the best-performing model with a mean absolute error (MAE) of 0.588. Beyond prediction accuracy, the present analysis contributes practical insights for industry application. First, the model identifies pipes at high risk of severe pitting corrosion under challenging environmental conditions, enabling targeted preventive measures. Secondly, the model's interpretation, utilising game-theoretic Shapely values and permutation importance, informs strategic decisions on the burial of new pipes, ultimately extending their operational lifespan. These applications emphasise the potential of our findings to significantly improve the safety, reliability, and risk mitigation strategies in oil and gas pipeline operations.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.