Hybrid machine learning and computational fluid dynamics framework for optimizing supercritical CO₂ pipeline networks in large-scale carbon capture and storage
Full Text (PDF)
Abiola Samuel Ajayi 1 * , Ayoola David Bodude 2 , Shokenu Emmanuel Segun 3 , Eniola Isaac Ajayi 4 , Dawudu Ibrahim Shanzhi 5
More Detail
1 Department of Computer Science, Texas Southern University, Houston, TX, USA2 Department of Chemical Engineering, University of Strathclyde, Galsgow, UK3 Department of System Engineering, African University of Science and Technology, Abuja, NIGERIA4 Department of Computer Science, Aptech Vocational Training Institute, Abuja, NIGERIA5 Department of System Engineering, African University of Science and Technology, NIGERIA* Corresponding Author

Abstract

This study presents an integrated framework combining computational simulations and machine learning (ML) to design and optimize supercritical carbon dioxide (CO₂) pipeline networks for large-scale carbon capture and storage (CCS) applications. Using ANSYS fluent, the fluid dynamics of supercritical CO₂ transport were analyzed under various operational conditions to generate synthetic insights and complement real-world datasets collected from CO₂ sequestration activities and industrial sources. Key variables such as pressure, temperature, and pipeline diameter were extracted, engineered into time-lagged and rolling features, and used to train a random forest regressor model. Feature importance analysis revealed that short-term flow trends, volatility, and geographic segmentation are dominant predictors of CO₂ flow behavior. The ML model achieved high performance, with an R² score of 0.9897 and a low RMSE of 0.0335, indicating strong predictive reliability. Hyperparameter tuning further optimized model accuracy, while visual analyses including actual vs. predicted comparisons, residual distributions, and feature interaction heatmaps validated the model’s robustness. The results demonstrate that data-driven approaches can complement simulation-based methods to improve the design, monitoring, and operational efficiency of CO₂ transport systems. This work provides a scalable, intelligent solution for CCS pipeline optimization, supporting the global transition toward low-carbon infrastructure.

Keywords

License

This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Article Type: Research Article

EUR J SUSTAIN DEV RES, Volume 10, Issue 2, 2026, Article No: em0378

https://doi.org/10.29333/ejosdr/17977

Publication date: 01 Apr 2026

Online publication date: 25 Feb 2026

Article Views: 14

Article Downloads: 5

Open Access References How to cite this article