An Integrated Study on Data-Driven Corrosion Prediction and Maintenance Decision-Making for Buried Pipelines

Tai Zhang; Jianchang Zhang; Ni Zhang; Dan Yu

doi:10.62051/ijnres.v7n2.04

Authors

Tai Zhang
Jianchang Zhang
Ni Zhang
Dan Yu

DOI:

https://doi.org/10.62051/ijnres.v7n2.04

Keywords:

Buried Pipeline; Corrosion Prediction; Neural Network; Corrosion Protection Decision-Making.

Abstract

This paper addresses the corrosion issues of buried pipelines by constructing a corrosion rate prediction model based on neural network algorithms, integrating multi-source data such as in-situ soil physicochemical parameters. Key input parameters of the model were identified through feature analysis, and the model was trained and validated using historical monitoring data. The results demonstrate that the model effectively captures the nonlinear relationships between multiple factors and corrosion rate with high prediction accuracy, particularly in the low corrosion rate range. Furthermore, the paper systematically reviews pipeline anti-corrosion maintenance decision-making methods based on maintenance timing, strategies, and evaluation metrics, comparing the applicability and limitations of reactive, periodic, and predictive maintenance strategies. It emphasizes that an intelligent maintenance system integrating multidisciplinary approaches such as machine learning, risk assessment, and reliability analysis can achieve closed-loop management of pipeline corrosion protection. This approach enhances the economy, safety, and sustainability of maintenance practices, providing theoretical support and decision-making guidance for the intelligent operation and maintenance of oil and gas station pipelines.

References

[1]X. Wu, H.Y. Wu, B. Peng, et al., Soil Corrosion Rate Prediction along Sichuan to Eastern China Gas Transmission Pipeline Based on BP Neural Network, Pipeline Technique and Equipment. (01) (2015) 7-9.

[2]Senouci Ahmed, Elabbasy Mohamed, Elwakil Emad, Abdrabou Bassem, Zayed Tarek. A model for predicting failure of oil pipelines [J]. Structure and Infrastructure Engineering, 2014, 10(3): 375-387.

[3]Ma Haonan, Geng Mengying, Wang Fan, Zheng Wenyue, Ai Yibo, Zhang Weidong. Data augmentation of a corrosion dataset for defect growth prediction of pipelines using conditional tabular generative adversarial networks [J]. Materials, 2024, 17(5): 1142.

[4]Li Yun-Tao, He Xiao-Ning, Shuai Jian. Risk analysis and maintenance decision making of natural gas pipelines with external corrosion based on Bayesian network [J]. Petroleum Science, 2022, 19(3): 1250-1261.

[5]Gong Changqing, Zhou Wenxing. Multi-objective maintenance strategy for in-service corroding pipelines using genetic algorithms [J]. Structure and Infrastructure Engineering, 2018, 14(11): 1561-1571.

[6]Chen Yunxiang, Wang Zezhou, Cai Zhongyi. Optimal maintenance decision based on remaining useful lifetime prediction for the equipment subject to imperfect maintenance [J]. Ieee Access, 2020, 8: 6704-6716.

[7]He Chu, Tan Zijing, Chen Qing, Sha Chaofeng. Repair diversification: A new approach for data repairing [J]. Information Sciences, 2016, 346: 90-105.

[8]Xie Mingjiang, Zhao Jianli, Zuo Ming J, Tian Zhigang, Liu Libin, Wu Jinming. Multi-objective maintenance decision-making of corroded parallel pipeline systems [J]. Applied Energy, 2023, 351: 121822.

[9]Jia Xuefei, Fu Chao, Chang Wenjun. A transfer-based decision-making method based on expert risk attitude and reliability [J]. Applied Intelligence, 2025, 55(7): 1-14.

[10]Tulabandhula Theja, Rudin Cynthia. On combining machine learning with decision making [J]. Machine learning, 2014, 97: 33-64.

[11]Z.B. Yin, S.S. Wang, Z.H. Zhu, et al., Key parameters of soil corrosivity and a model for predicting the corrosion rate of Q235steel in Beijing, Chinese Journal of Engineering. 45 (11) (2023) 1939-1947.

[12]Y. Guang, W.H. Wang, H.W. Song, et al., Prediction of external corrosion rate for buried oil and gas pipelines: A novel deep learning method with DNN and attention mechanism, International Journal of Pressure Vessels and Piping. 209 (2024) 105218, https://doi.org/10.1016/j.ijpvp.2024.105218.

[13]Tee Kong Fah, Pesinis Konstantinos. Reliability prediction for corroding natural gas pipelines [J]. Tunnelling and Underground Space Technology, 2017, 65: 91-105.

[14]Wang Wenbin, Jiang Shibin, Liu Jianguo, Cui Gan. An evaluation method for pipeline corrosion risk index weighting in beach and sea oil fields based on combined weighting with improved hierarchical analysis and Bayesian networks [J]. Applied Ocean Research, 2025, 158: 104522.

[15]Ossai Chinedu I. Corrosion defect modelling of aged pipelines with a feed-forward multi-layer neural network for leak and burst failure estimation [J]. Engineering Failure Analysis, 2020, 110: 104397.

[16]Elshaboury Nehal, Al-Sakkaf Abobakr, Alfalah Ghasan, Abdelkader Eslam Mohammed. Data-driven models for forecasting failure modes in oil and gas pipes [J]. Processes, 2022, 10(2): 400.

[17]Jiang Fengyuan, Dong Sheng, Zhao Enjin. A study on burst failure mechanism analysis and quantitative risk assessment of corroded pipelines with random pitting clusters [J]. Ocean Engineering, 2023, 284: 115258.

[18]Zhou Wenxing. System reliability of corroding pipelines [J]. International Journal of Pressure Vessels and Piping, 2010, 87(10): 587-595.

[19]Gong Changqing, Zhou Wenxing. First-order reliability method-based system reliability analyses of corroding pipelines considering multiple defects and failure modes [J]. Structure and Infrastructure Engineering, 2017, 13(11): 1451-1461.

[20]C.T. Wang, M.F. Hassanein, M.M. Li, Numerical simulation of oil and gas pipeline corrosion based on single- or coupled-factor modeling: A critical review, Natural Gas Industry B. 10 (5) (2023) 445-465, https://doi.org/10.1016/j.ngib.2023.08.001.