Data-Driven Approach to Predict Future Oil Production of an Oil Field Using Machine Learning Techniques
DOI:
https://doi.org/10.38032/scse.2025.3.16Keywords:
Future production, GBR, XGBoost, LightGBR, Volve oil fieldAbstract
Determining accurate future production in the oil and gas industry is increasingly challenging with traditional methods. Decline Curve Analysis often fails to provide precise results, while Reservoir Simulation Models require detailed parameters and are time-consuming due to the constantly changing reservoir conditions during the production period. Today, industries are leveraging machine learning and extensive data from oil wells for predictive purposes, can significantly reduce operational costs and minimize negative environmental impacts. This study aims to predict future production using machine learning algorithms. Specifically, Gradient Boosting Regression, Light Gradient Boosting Regression, and Extreme Gradient Boosting Regression models were developed using the production dataset of the Norwegian Volve oil field (well NO159F-11H) within 12 parameters. These models were trained on 80% of the dataset, while the remaining 20% was reserved for testing purposes. The accuracy of the models was assessed using the coefficient of determination (R²), which was found to be 99% for both training and testing data across all models. GBR demonstrated the lowest mean absolute error (MAE = 12.810) and root mean square error (RMSE = 17.802) compared to the other two models based on testing value. On the other hand, based on training dataset, XGBoost showed the lowest MAE (1.192) and RMSE (1.671) values. However, the results for well NO159 F-11H show that GBR outperformed the other two methods but this doesn't imply that GBR is always better than XGBoost or LightGBR in all cases. An extensive study was conducted to evaluate the predictive performance of these models, with systematic assessments and hyperparameter adjustments to reliably anticipate the well's performance.
Downloads
Downloads
Downloads
References
[1] P. S. Omrani et al., “Deep Learning and Hybrid Approaches Applied to Production Forecasting,” Abu Dhabi International Petroleum Exhibition & Conference, Nov. 2019.
[2] M. Bagheri et al., “Data Conditioning and Forecasting Methodology using Machine Learning on Production Data for a Well Pad,” Offshore Technology Conference, May 2020.
[3] D. Han, S. Kwon, H. Son, and J. Lee, “Production forecasting for shale gas well in transient flow using machine learning and decline curve analysis,” Proceedings of the SPE/AAPG/SEG Asia Pacific Unconventional Resources Technology Conference, Jan. 2019.
[4] Y. Li and Y. Han, “Decline Curve Analysis for Production Forecasting Based on Machine Learning,” SPE Symposium: Production Enhancement and Cost Optimisation, Nov. 2017.
[5] Q. . Cao, R. . Banerjee, S. . Gupta, J. . Li, W. . Zhou, and B. . Jeyachandra, “Data Driven Production Forecasting Using Machine Learning,” SPE Argentina Exploration and Production of Unconventional Resources Symposium, May 2016.
[6] E. A. M. Vélez, F. R. Consuegra, and C. A. B. Arias, “EOR Screening and Early Production Forecasting in Heavy Oil Fields: A Machine Learning Approach,” SPE Latin American and Caribbean Petroleum Engineering Conference, Jul. 2020.
[7] C. Zhan, S. Sankaran, V. LeMoine, J. Graybill, and D.-O. S. Mey, “Application of machine learning for production forecasting for unconventional resources,” Proceedings of the 7th Unconventional Resources Technology Conference, Jan. 2019.
[8] E. A. M. Vélez, F. R. Consuegra, and C. A. B. Arias, “EOR Screening and Early Production Forecasting in Heavy Oil Fields: A Machine Learning Approach,” SPE Latin American and Caribbean Petroleum Engineering Conference, Jul. 2020.
[9] E. Naqa and M. J. Murphy, “What is machine learning?,” in Springer eBooks, 2015, pp. 3–11.
[10] B. A. Juliussen, J. P. Rui, and D. Johansen, “Algorithms that forget: Machine unlearning and the right to erasure,” Computer Law & Security Review, vol. 51, p. 105885, Sep. 2023.
[11] Data Preprocessing in Data Mining, 1st ed., vol. 72. springer, 2015. [Online]. Available: https://link.springer.com/book/10.1007/978-3-319-10247-4
[12] J. Cai, K. Xu, Y. Zhu, F. Hu, and L. Li, “Prediction and analysis of net ecosystem carbon exchange based on gradient boosting regression and random forest,” Applied Energy, vol. 262, p. 114566, Feb. 2020.
[13] Shehadeh, O. Alshboul, R. E. A. Mamlook, and O. Hamedat, “Machine learning models for predicting the residual value of heavy construction equipment: An evaluation of modified decision tree, LightGBM, and XGBoost regression,” Automation in Construction, vol. 129, p. 103827, Jul. 2021.
[14] Balamwar, R. Mitra, M. K. Tiwari, and P. Verma, “Prediction and Analysis of Seasonal Dynamic Metal Consumption using Aggregated LightGBM - A Case Study,” IFAC-PapersOnLine, vol. 55, no. 10, pp. 725–730, Jan. 2022.
[15] L. Zhang, W. Bian, W. Qu, L. Tuo, and Y. Wang, “Time series forecast of sales volume based on XGBoost,” Journal of Physics Conference Series, vol. 1873, no. 1, p. 012067, Apr. 2021.
[16] X. Dairu and Z. Shilong, “Machine Learning Model for Sales Forecasting by Using XGBoost,” IEEE, Jan. 2021.
Published
Conference Proceedings Volume
Section
License
Copyright (c) 2025 Ataharuse Samad, Istiaque Muhammad Khan, Md. Shakil Rahaman, Ahmed Sakib, Md. Ashraful Islam (Author)
This work is licensed under a Creative Commons Attribution 4.0 International License.
All the articles published by this journal are licensed under a Creative Commons Attribution 4.0 International License
