(234h) Mitigating Demand Uncertainty By Developing a Capacity Expansion Roadmap for Hydrogen Networks

The global demand for hydrogen is projected to grow fivefold by 2050 [1], driven by the increasing push for sustainable operations and emission reductions. However, the transition to a hydrogen-based economy is challenged in several ways including high production costs, inadequate supply chain infrastructure and uncertain forecasts. Salt caverns offer a promising solution due to their large capacity and cost-effectiveness. Additionally, different hydrogen production pathways vary in their carbon intensity. Gray hydrogen is produced from natural gas-based processes such as autothermal reforming (ATR) or steam methane reforming (SMR) while blue hydrogen is produced from processes using fossil fuels in conjunction with carbon capture. In contrast, green hydrogen is primarily generated via electrolyzers using renewable energy. Given the uncertainties in demand growth and hydrogen “color” uptake [2], long-term forecasts remain the best available tool for planning a flexible and cost-effective hydrogen network that can meet current needs while adapting to future transitions. In this talk, we present a mixed-integer linear programming (MILP) model developed to optimize the long-term network investment plan that accounts for investment decisions in hydrogen production, pipelines, trucking, storage, and market operations.

Pérez-Uresti et al. [3] developed an MINLP model for the strategic investment planning of large-scale hydrogen economy using surrogate models derived from rigorous simulation models. Jamali et al. [4] also modeled the network using an MINLP but placed significant emphasis on modelling the transmission pipelines accurately. Despite their high fidelity, models such as these are intractable due to their non-linearity, especially for long horizons. Jodry et al. [5] build an LP model analyzing the long-term network investment into electricity generation, hydrogen production and storage. Using this model, they analyze the development trajectory and sensitivity of the results found to various parameters. Almansoori and Shah [6] generated an MILP model and used a set of scenarios to optimize the network under various demand uncertainties. While these models address various aspects of hydrogen network optimization, they overlook the recent shift toward a "color"-based hydrogen classification [7] and fail to incorporate uncertainty within this evolving framework.

To develop a robust capacity expansion roadmap, we begin by integrating operational data and formulating an MILP which is used to optimize investment decisions (production, transportation and storage) over the time horizon while maintaining viability over finer time grid to satisfy demands and operating conditions. Given the uncertainty in the transition rate toward clean hydrogen, we incorporate an uncertainty set to capture the evolving demand proportions of brown, blue, and green hydrogen. While binary variables are introduced sparingly to ensure computational tractability, this also allows for nearly complete recourse when reformulating the robust counterpart while preserving linearity. We then assess the tractability of our approach by testing various scenarios and evaluating the price of robustness, providing key insights into the trade-offs between cost, flexibility, and resilience in hydrogen infrastructure planning.

References

[1] International Energy Agency, “Global Hydrogen Review 2024,” Paris, 2024.

[2] J. M. M. Arcos and D. M. F. Santos, “The Hydrogen Color Spectrum: Techno-Economic Analysis of the Available Technologies for Hydrogen Production,” Gases, vol. 3, no. 1, pp. 25–46, Feb. 2023, doi: 10.3390/gases3010002.

[3] S. I. Pérez-Uresti, G. Gallardo, and D. K. Varvarezos, “Strategic investment planning for the hydrogen economy – A mixed integer non-linear framework for the development and capacity expansion of hydrogen supply chain networks,” Comput Chem Eng, vol. 179, p. 108412, Nov. 2023, doi: 10.1016/j.compchemeng.2023.108412.

[4] D. Hamedi Jamali, C. Ganzer, and K. Sundmacher, “Hydrogen Network Topology Optimization by Minlp: Comparing Retrofit with New-Built Design Scenarios,” 2025, doi: 10.2139/ssrn.5118373.

[5] A. Jodry, R. Girard, P. H. A. Nóbrega, R. Molinier, and M.-D. El Alaoui Faris, “Hydrogen production via electrolysis in 2030: comparing four connection schemes through energy system optimization,” in 2023 IEEE PES Innovative Smart Grid Technologies Europe (ISGT EUROPE), IEEE, Oct. 2023, pp. 1–5. doi: 10.1109/ISGTEUROPE56780.2023.10407454.

[6] A. Almansoori and N. Shah, “Design and operation of a stochastic hydrogen supply chain network under demand uncertainty,” Int J Hydrogen Energy, vol. 37, no. 5, pp. 3965–3977, Mar. 2012, doi: 10.1016/j.ijhydene.2011.11.091.

[7] A. Ajanovic, M. Sayer, and R. Haas, “The economics and the environmental benignity of different colors of hydrogen,” Int J Hydrogen Energy, vol. 47, no. 57, pp. 24136–24154, Jul. 2022, doi: 10.1016/j.ijhydene.2022.02.094.