Comparative Analysis of Machine Learning Algorithms for Flow Rate Prediction in Optimizing Pipeline Maintenance Strategies

Abstract

Using machine learning to predict maintenance schedules for crude oil pipelines is crucial for enhancing efficiency and minimizing disruptions in the oil and gas sector. Our research explores the effectiveness of machine learning algorithms in this context, with a specific focus on using oil flow rate as a primary predictor. When trained with a variety of inspection data, machine learning models can accurately predict flow rates, thus improving maintenance planning. Several pipeline scenarios were analyzed, and the Python library was used for dataset augmentation. The study shows a correlation between variations in the buildup deposits and the oil flow rate in the pipeline, indicating that the oil flow rate gives an indication for determining the need for maintenance. The flow rate was categorized into three efficiency levels: High Efficiency (flow rate > 90% of the allowable rate), Moderate Efficiency (flow rate between 70% and 90%), and Low Efficiency (flow rate < 70%). Each efficiency level was linked to specific maintenance intervals: Specifically, a higher flow rate allowed longer intervals between maintenance activities, while a lower flow rate could indicate there is an accumulation of deposits that necessitates urgent intervention. Several machine learning models were trained, and variations in performance were observed. Gradient Boosting and XGBoost Regressor show the best performers with lower values for MSE, RMSE, and MAE and higher R2 scores compared to the Support Vector Regressor. The result shows Gradient Boosting has an MSE of 0.000005, RMSE of 0.002259, MAE of 0.000968, and an R2 of 0.997259, followed by XGBoost Regressor with MSE of 0.000005, an RMSE of 0.002269, an MAE of 0.000922, and an R2 of 0.997234. Support Vector Regressor indicates the least performance, with an MSE of 0.002868, RMSE of 0.053554, MAE of 0.046311, and an R2 of −0.540765. The findings of the study emphasize the necessity of choosing machine learning algorithms that are appropriately suited to the features of the dataset and the task. The findings highlight the importance of selecting machine learning algorithms that are more suitable to the features of the dataset and the task.