Advancements in prediction models for corrosion in oil and gas pipelines

Abstract Oil and gas pipelines play an important role in the energy transportation industry, but metal corrosion can affect the safe operation of pipeline equipment. This study uses CiteSpace software to synthesize and analyze corrosion models and keywords from research institutions, countries, and methods related to pipeline corrosion prediction. The investigation into the mechanisms of pipeline metal corrosion, with a specific emphasis on CO 2 and H 2 S corrosion, has revealed that several factors influence the process, including temperature, partial pressure, medium composition and the corrosion product film. In addition, the study provides a comprehensive review of pipeline corrosion prediction methods and models. These include traditional empirical, semi-empirical, and mechanism-based prediction models, as well as advanced machine learning techniques such as random forest, artificial neural network model, support vector machine, and dose-response function. Although there are many ways to improve model performance, no universally accepted methods have been established. Therefore, further in-depth research is needed to improve the accuracy of these models and provide guidance for improving the operational safety of pipelines.