EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Multi-output machine learning prediction of CO₂ enhanced oil recovery and geological storage performance

Abdullah A. Hashim, Farqad A. Hadi, Dheiaa Alfarge

Iraqi Journal of Chemical and Petroleum Engineering · 30.03.2026 · 27 (1) · 85-92

DOI: 10.31699/ijcpe.2026.1.7 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Iraqi Journal of Chemical and Petroleum Engineering, online erschienen am 30.03.2026. Verfasst von Abdullah A. Hashim und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Carbon dioxide-enhanced oil recovery (CO₂-EOR) is one of the best commercial carbon capture, utilization, and storage (CCUS) technologies, enhancing oil recovery and achieving safe geological storage. However, data density and computational costs hinder conventional reservoir simulations. This research presents a multi-output machine learning model to simultaneously predict both incremental oil production and CO₂ storage volume. A selected global dataset of 173 projects was compiled, preprocessed, and optimized. Among the evaluated algorithms (random forest, gradient boosting, and random forest chain), gradient boosting achieved the best predictive performance, with a weighted coefficient of determination (R²) of 0.722. Model analysis showed that permeability, depth, and miscibility strongly influence oil recovery (R² for gradient boosting = 0.87), while the predictability of storage volume is low (R² for gradient e boosting = 0.5), which is significantly affected by reservoir depth and porosity. This study highlights the potential of machine learning as a rapid screening tool for evaluating carbon capture, utilization, and storage projects, identifying the performance of key factors, and emphasizing the need for standardized storage reporting to improve prediction accuracy.

Zitieren

APA

Hashim, A. A., Hadi, F. A., & Alfarge, D. (2026). Multi-output machine learning prediction of CO₂ enhanced oil recovery and geological storage performance. Iraqi Journal of Chemical and Petroleum Engineering, 27(1), 85-92. https://doi.org/10.31699/ijcpe.2026.1.7

Vancouver

Hashim AA, Hadi FA, Alfarge D. Multi-output machine learning prediction of CO₂ enhanced oil recovery and geological storage performance. Iraqi Journal of Chemical and Petroleum Engineering. 2026;27(1):85-92. doi:10.31699/ijcpe.2026.1.7

BibTeX
@article{Hashim2026,
  author = {Hashim, Abdullah A. and Hadi, Farqad A. and Alfarge, Dheiaa},
  title = {Multi-output machine learning prediction of CO₂ enhanced oil recovery and geological storage performance},
  journal = {Iraqi Journal of Chemical and Petroleum Engineering},
  year = {2026},
  volume = {27},
  number = {1},
  pages = {85--92},
  doi = {10.31699/ijcpe.2026.1.7}
}
Kostenlos starten →