InterPore2026

Underground Hydrogen Storage: A Data-Driven Approach to Site Selection and Performance Optimisation

21 May 2026, 15:35

1h 30m

Poster Presentation (MS01) Porous Media for a Green World: Energy & Climate Poster

Speaker

Mr Abdolali Mosallanezhad (PhD Student, Research Centre for Carbon Solutions (RCCS), School of Engineering and Physical Sciences, Heriot-Watt University, Edinburgh, UK)

Description

Hydrogen is widely recognised as a cornerstone of global decarbonisation and a critical component of the pathway to net-zero emissions. By enabling the conversion of renewable electricity into chemical energy, a process known as Power-to-X, it offers a robust solution to the temporal and spatial mismatches in renewable generation, effectively tackling the intermittency of wind and solar power. In the UK, for instance, the transition strategy is supported by a strategic "twin-track" roadmap, targeting a production capacity of 10 GW, approximately 4.88 million kg/day. With current natural gas consumption exceeding 444,000 GWh annually, transitioning this massive demand requires infrastructure capable of managing regional imbalances.

Underground hydrogen storage (UHS) provides the essential temporal balancing required to absorb surplus renewable energy, preventing curtailment and preserving value for industry, transport, and heating. However, the success of this infrastructure depends on identifying efficient and reliable geological storage sites. Traditionally, site screening has been dominated by assessing static parameters, which remain constant over time, such as rock properties. While essential, these assessments overlook dynamic factors that evolve over time and in response to operating conditions, including pressure changes and hysteresis in flow functions. These dynamic processes are critical for determining realistic storage capacity and operational efficiency. This study addresses the current gap by integrating static and dynamic screening approaches, enabling more accurate evaluation of potential storage sites and advancing underground hydrogen storage readiness. A significant barrier to dynamic screening has historically been the lack of detailed reservoir input data required for reliable simulations. To address this, machine learning is utilised to develop reservoir-specific relative permeability correlations for hydrogen flow in porous media, derived directly from experimental data. These data-driven correlations supply the missing parameters needed to model complex fluid dynamics, enabling a comprehensive assessment of trapping mechanisms.

To operationalise the findings, we conduct a UK-specific study that advances dynamic screening by simulating various reservoirs under diverse operational conditions driven by UK regional supply and demand. By incorporating specific limiting factors, such as the steady baseload requirements of UK industrial clusters versus the intermittent hydrogen surpluses, the model predicts reservoir behaviour under realistic operating conditions. This framework facilitates the identification of bottleneck scenarios and allows for the selection of top storage options for each major UK cluster, matching geological candidates to local infrastructure needs.

The initial results underscore the risks of relying solely on static models. Numerical simulations show that ignoring hysteresis can lead to an overestimation of hydrogen recovery by up to 20%. Furthermore, in geological models featuring high-permeability layers, flow instabilities reduced recovery rates by an additional 10%. By capturing these key dynamic processes, our research provides a vital tool for enhanced site screening and candidate selection, ensuring that the UK’s storage infrastructure is developed with the efficiency and reliability required for a low-carbon future.