Prediction of maximum pitting corrosion depth in oil and gas pipelines

Avoiding failures of corroded steel structures are critical in offshore oil and gas operations. An accurate prediction of maximum depth of pitting corrosion in oil and gas pipelines has significance importance, not only to prevent potential accidents in future but also to reduce the economic charges to both industry and owners. In the present paper, efficient hybrid intelligent model based on the feasibility of Support Vector Regression (SVR) has been developed to predict the maximum depth of pitting corrosion in oil and gas pipelines, whereas the performance of well-known meta-heuristic optimization techniques, such as Genetic Algorithm (GA), Particle Swarm Optimization (PSO) and Firefly Algorithm (FFA), are considered to select optimal SVR hyper-parameters. These nature-inspired algorithms are capable of presenting precise optimal predictions and therefore, hybrid models are developed to integrate SVR with GA, PSO, and FFA techniques. The performances of the proposed models are compared with the traditional SVR model where its hyper-parameters are attained through trial and error process on the one hand and empirical models on the other. The developed models have been applied to a large database of maximum pitting corrosion depth. Computational results indicate that hybrid SVR models are efficient tools, which are capable of conducting a more precise prediction of maximum pitting corrosion depth. Moreover, the results revealed that the SVR-FFA model outperformed all other models considered in this study. The developed SVR-FFA model could be adopted to support pipeline operators in the maintenance decision-making process of oil and gas facilities.