Pareto-based Multi-Objective Optimization of Underground Hydrogen Storage Operational Condition Utilizing Machine Learning-Based Surrogate Models

Hossein  Kheirollahi; Shahab  Ayatollahi; Hassan Mahani

doi:10.69631/m0g16y60

Authors

Hossein Kheirollahi Department of Chemical and Petroleum Engineering, Sharif University of Technology, Tehran, Iran https://orcid.org/0009-0006-7534-992X
Shahab Ayatollahi Department of Chemical and Petroleum Engineering, Sharif University of Technology, Tehran, Iran
Hassan Mahani Department of Chemical and Petroleum Engineering, Sharif University of Technology, Tehran, Iran

DOI:

https://doi.org/10.69631/m0g16y60

Keywords:

Underground Hydrogen Storage, Depleted Gas Reservoirs, Machine Learning, Surrogate models, Genetic algorithms, Energy Transition

Abstract

Underground hydrogen storage (UHS) in depleted hydrocarbon reservoirs, saline aquifers, and salt caverns has emerged as a promising solution for large-scale, long-term energy storage to support the energy transition. However, evaluating the feasibility of UHS requires addressing substantial uncertainties in geological, operational, and economic factors. Physics-based reservoir simulations, though accurate, are computationally expensive and impractical for exploring a wide range of operational scenarios and optimization problems. To overcome these challenges, this study employs an intelligent surrogate model constructed using data collected from Latin hypercube sampling and validated against detailed reservoir simulations to perform sensitivity analysis and formulate a multi-objective optimization framework. The primary objectives are to maximize hydrogen recovery factor and hydrogen purity in the produced gas, while also incorporating net present value (NPV) as an economic criterion. Multi-objective optimization is performed using the Non-dominated Sorting Genetic Algorithm II (NSGA-II), which enables efficient exploration of the trade-offs between conflicting objectives and provides a Pareto front of non-dominated solutions. This Pareto-based decision-making framework offers valuable insights into selecting optimal storage strategies under uncertainty. The results highlight two key findings. First, pure nitrogen (N₂) as a cushion gas yields higher hydrogen mole fraction in production stream compared to CO₂ and CH₄, because of favorable segregation and flow behavior in porous media. Second, the optimal cushion gas mixture was determined to be 45% N₂, 40% CO₂, and 15% CH₄, balancing recovery and purity objectives. Nevertheless, from both economic and operational efficiency perspectives, a CO₂ fraction of 100% emerges as the optimal solution. Additionally, injecting from bottom perforations and producing from top perforations was identified as the most effective configuration due to the gravity override effect of hydrogen. Importantly, re-simulation of the Pareto-optimal solutions confirmed the robustness and reliability of the surrogate-based optimization approach. This study demonstrates the capability of coupling surrogate modeling with evolutionary multi-objective optimization to design cushion gas, optimize injection/withdrawal schemes, and maximize both technical and economic performance of large-scale underground hydrogen storage projects.

Downloads

Download data is not yet available.

Author Biographies

Shahab Ayatollahi, Department of Chemical and Petroleum Engineering, Sharif University of Technology, Tehran, Iran

Department of Chemical and Petroleum Engineering, Sharif University of Technology, Tehran, Iran
Hassan Mahani, Department of Chemical and Petroleum Engineering, Sharif University of Technology, Tehran, Iran

Hassan Mahani is a Professor of Petroleum and Geo-Engineering at Sharif University of Technology. He holds a PhD degree in Petroleum Engineering from Imperial College London and a MSc degree in Chemical Engineering from Sharif University of Technology. Before joining Sharif University of Technology, he worked as a principal investigator and scientist at Shell Technology Centers in The Netherlands. He currently serves as Associate Editor for SPE Journal and Geoenergy Science and Engineering Journal. His research is geared toward fundamental understanding, quantitative evaluation, modeling, and prediction of multi-phase flow in porous media with application to IOR/EOR, geo-engineering, and the environment. His current activities focus on water-based/low-salinity EOR; Underground Hydrogen & Energy Storage, Contaminant Gas Storage; Pore-scale physics & Digital rocks; Geochemistry, and Geomechanics.

References

1. Alavala, C. R. (2008). Fuzzy logic and neural networks: Basic concepts & application (1st ed). New Age International (P) Ltd., Publishers.

2. Asghari, M., Emami Niri, M., & Sedaee, B. (2025). UHSNet: Deep learning-based smart proxy modeling for underground hydrogen storage. Energy, 329, 136763. https://doi.org/10.1016/j.energy.2025.136763

3. Atrbarmohammadi, S., Kheirollahi, H., Ayatollahi, S., & Pishvaie, S. (2025). Utilizing machine learning classification models for acid–oil emulsion prediction in laboratory acidizing static tests using a hybrid databank. Journal of Petroleum Research, 34(6), 73–91.

4. Bahrami, M., Izadi Amiri, E., Zivar, D., Ayatollahi, S., & Mahani, H. (2023). Challenges in the simulation of underground hydrogen storage: A review of relative permeability and hysteresis in hydrogen-water system. Journal of Energy Storage, 73, 108886. https://doi.org/10.1016/j.est.2023.108886

5. Bahrami, P., Sahari Moghaddam, F., & James, L. A. (2022). A review of proxy modeling highlighting applications for reservoir engineering. Energies, 15(14), 5247. https://doi.org/10.3390/en15145247

6. Eltahan, E., Albadan, D. J., Delshad, M., Sepehrnoori, K., & Alpak, F. O. (2024). Optimizing hydrogen storage in the subsurface using a reservoir-simulation-based and deep-learning-accelerated optimization method. SPE Annual Technical Conference and Exhibition, D021S028R002. https://doi.org/10.2118/220865-MS

7. Esfandi, T., Noruzi, Y., Safavi, M. S., & Sadeghnejad, S. (2025). Analyzing key parameters in underground hydrogen storage using machine learning surrogate models. In J. Yao, Y. Yang, W. Wang, H. Sun, L. Zhang, & K. Zhang (Eds.), Progress and Challenge of Porous Media: Proceedings of the 16th Annual Meeting Conference on Porous Media (pp. 978–986). Springer Nature Singapore. https://doi.org/10.1007/978-981-96-2983-1_83

8. Eslami Nia, M., Maleki Asl, P., Mahani, H., & Ayatollahi, S. (2025). Techno-economic feasibility of underground hydrogen storage in Iran: A path to sustainable energy integration. International Journal of Energy and Water Resources, 9(4), 1767–1778. https://doi.org/10.1007/s42108-025-00350-2

9. Fu, S., Mao, S., Carbonero, A., Srikishan, B., Creasy, N., Chellal, H., & Mehana, M. (2025). Deep learning-based surrogate modeling for underground hydrogen storage. Advances in Water Resources, 203, 105014. https://doi.org/10.1016/j.advwatres.2025.105014

10. Han, Z., Tariq, Z., & Yan, B. (2025). A novel robust optimization framework based on surrogate modeling for underground hydrogen storage in depleted natural gas reservoirs. InterPore Journal, 2(3), IPJ250825-6. https://doi.org/10.69631/ipj.v2i3nr69

11. He, Q., Mohaghegh, S. D., & Liu, Z. (2016). Reservoir simulation using smart proxy in SACROC unit—Case study. SPE Eastern Regional Meeting, SPE-184069-MS. https://doi.org/10.2118/184069-MS

12. Juez-Larré, J., Gonçalves Machado, C., Yousefi, H., Wang, T.-K., Groenenberg, R., & Van Gessel, S. (2023, May 15). (Pre)Feasibility study of underground hydrogen storage potential in depleted gas fields and salt caverns in the Netherlands. https://doi.org/10.5194/egusphere-egu23-10005

13. Kanaani, M., Sedaee, B., Asadian-Pakfar, M., Gilavand, M., & Almahmoudi, Z. (2023). Development of multi-objective co-optimization framework for underground hydrogen storage and carbon dioxide storage using machine learning algorithms. Journal of Cleaner Production, 386, 135785. https://doi.org/10.1016/j.jclepro.2022.135785

14. Kheirollahi, H., Zayedi, M., Sobhani, S., Chahardowli, M., & Simjoo, M. (2023). Using an ensembled artificial intelligence approach for EOR methods screening in oil fields. Petroleum Research, 33(1402–5), 51–62.

15. Le Duigou, A., Bader, A.-G., Lanoix, J.-C., & Nadau, L. (2017). Relevance and costs of large scale underground hydrogen storage in France. International Journal of Hydrogen Energy, 42(36), 22987–23003. https://doi.org/10.1016/j.ijhydene.2017.06.239

16. Li, Y., & Nghiem, L. X. (1986). Phase equilibria of oil, gas and water/brine mixtures from a cubic equation of state and henry’s law. The Canadian Journal of Chemical Engineering, 64(3), 486–496. https://doi.org/10.1002/cjce.5450640319

17. Malki, M. L., Chen, B., Mao, S., Chen, F., & Mehana, M. (2024). OPERATE–H2: A tool for optimizing underground hydrogen storage. Journal of Energy Storage, 90, 111715. https://doi.org/10.1016/j.est.2024.111715

18. Mao, S., Chen, B., Malki, M., Chen, F., Morales, M., Ma, Z., & Mehana, M. (2024). Efficient prediction of hydrogen storage performance in depleted gas reservoirs using machine learning. Applied Energy, 361, 122914. https://doi.org/10.1016/j.apenergy.2024.122914

19. Matthew, D. A. M., Jahanbani Ghahfarokhi, A., Ng, C. S. W., & Nait Amar, M. (2023). Proxy model development for the optimization of water alternating CO2 gas for enhanced oil recovery. Energies, 16(8), 3337. https://doi.org/10.3390/en16083337

20. Nazari, F., Farajzadeh, R., Shokri, J., Vahabzadeh, E., Lopez-Porfiri, P., Perez-Page, M., & Niasar, V. (2025). Implications of non-ideal gas dispersion for underground hydrogen storage. Chemical Engineering Journal, 507, 160143. https://doi.org/10.1016/j.cej.2025.160143

21. Ng, C. S. W., Jahanbani Ghahfarokhi, A., Nait Amar, M., & Torsæter, O. (2021). Smart proxy modeling of a fractured reservoir model for production optimization: Implementation of metaheuristic algorithm and probabilistic application. Natural Resources Research, 30(3), 2431–2462. https://doi.org/10.1007/s11053-021-09844-2

22. Omeke, J., Misra, S., & Retnanto, A. (2024). Rapid forecasting of underground hydrogen storage using Fourier-integrated hybrid neural framework. ADIPEC, D031S088R005. https://doi.org/10.2118/222738-MS

23. Peaceman, D. W. (2010). Fundamentals of numerical reservoir simulation. Elsevier Scientific Pub. Co.

24. Pérez-Vigueras, M., Sotelo-Boyás, R., González-Huerta, R. D. G., & Bañuelos-Ruedas, F. (2023). Feasibility analysis of green hydrogen production from oceanic energy. Heliyon, 9(9), e20046. https://doi.org/10.1016/j.heliyon.2023.e20046

25. Risso, F. V. A., Risso, F. F., & Schiozer, D. J. (2008). Risk assessment of oil fields using proxy models: A case study. Journal of Canadian Petroleum Technology, 47(08). https://doi.org/10.2118/08-08-09-TN

26. Schuetter, J., Mishra, K. S., Ganesh, P. R., & Mooney, D. (2014). Building statistical proxy models for CO2 geologic sequestration. Energy Procedia, 63, 3702–3714. https://doi.org/10.1016/j.egypro.2014.11.399

27. Vapnik, V. N. (1995). The nature of statistical learning theory. Springer New York. https://doi.org/10.1007/978-1-4757-2440-0

28. Vaziri, P., & Sedaee, B. (2024). An application of a genetic algorithm in co-optimization of geological CO2 storage based on artificial neural networks. Clean Energy, 8(1), 111–125. https://doi.org/10.1093/ce/zkad077

29. Yan, H., Zhang, W., Kang, J., & Yuan, T. (2023). The necessity and feasibility of hydrogen storage for large-scale, long-term energy storage in the new power system in China. Energies, 16(13), 4837. https://doi.org/10.3390/en16134837

30. Zhao, Y., Wei, Y., Zhang, S., Guo, Y., & Sun, H. (2024). Multi-objective robust optimization of integrated energy system with hydrogen energy storage. Energies, 17(5), 1132. https://doi.org/10.3390/en17051132

31. Zhou, Z.-H. (2012). Ensemble methods: Foundations and algorithms (1st ed.). Chapman and Hall/CRC. https://doi.org/10.1201/b12207