Sign up Log in
rjees logo



Prediction of Pipeline Failure using Machine Learning Algorithms

Authors: *Odekanle, E.L. And Abdulsalam, F.M.

DOI Info: http://doi.org/10.5281/zenodo.15778256

ABSTRACT

The integrity of oil and gas pipelines is critical for ensuring environmental safety, operational efficiency, and uninterrupted energy supply. Traditional monitoring methods often fail to detect early-stage pipeline degradation, which can lead to catastrophic failures. This research presents a machine learning-based predictive system for identifying pipelines at risk of failure, utilizing operational and structural parameters. Four supervised learning models - Logistic Regression, Random Forest, Support Vector Machine, and Decision Tree - were trained and evaluated on a 10,000-entry dataset comprising pressure, temperature, corrosion, deformation, age, and material data to predict the failure status of pipeline components. The dataset was preprocessed using MinMax scaling and encoded for consistency. Logistic Regression achieved the highest accuracy of 98.8%, followed closely by Support Vector Machine and Random Forest. Corrosion and deformation levels emerged as the most significant indicators of failure. The study demonstrates that machine learning offers a reliable, interpretable, and scalable solution for predictive maintenance in pipeline systems, reducing risk and supporting cost-effective operation strategies.


Affiliations: Department of Chemical and Petroleum Engineering, Faculty of Engineering and Technology, Abiola Ajimobi Technical University (First Tech-U), Ibadan, Nigeria.
Keywords: Pipeline Failure Machine Learning Logistic Regression Predictive Maintenance Corrosion
Published date: 2025/06/30

Download Full Text

SUBMIT A MANUSCRIPT

ISSN: 2635-3342 (Print)

ISSN: 2635-3350 (Online)

DOI: In progress

ISI Impact Factor: In progress

Indexing & Abstracting
AR Index google scholar Directory of research journal indexing JIFactor Info base index scientific journal impact factor

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License



(+234) 806 927 5563

Chemical Engineering Department, Faculty of Engineering, University of Benin, PMB 1154, Ugbowo, Benin City, Edo State, Nigeria.