Table of Contents
Fetching ...

Pore-scale imaging of hydrogen and methane storage in fractured aquifer rock: The impact of gas type on relative permeability

Sojwal Manoorkar, Gülce Kalyoncu, Hamdi Omar, Soetkin Barbaix, Dominique Ceursters, Maxime Latinis, Stefanie Van Offenwert, Tom Bultreys

TL;DR

This study directly measures drainage relative permeabilities for $H_2$-brine, $CH_4$-brine, and $N_2$-brine in a fractured limestone core from Loenhout at $10\ \mathrm{MPa}$, using X-ray tomography to resolve pore-scale distributions. Hydrogen and methane exhibit similar, but globally low, gas permeabilities $K_{rg}$ due to strong phase interference and intermittent flow, while nitrogen shows substantially higher $K_{rg}$, revealing that nitrogen is an unreliable proxy for hydrogen under reservoir conditions. The results imply that hydrogen storage in fractured reservoirs may have reduced injectivity relative to methane, and that fracture morphology and intermittency must be captured in upscaled models; they also underscore the need for hydrogen-specific fracture data and caution against proxy-based upscaling using nitrogen. Limitations include a single artificially fractured sample and non-matching reservoir temperature, suggesting future work to explore imbibition relative permeabilities, varied fracture geometries, and upscaling to discrete fracture networks (DFN).

Abstract

Underground hydrogen storage in saline aquifers is a potential solution for seasonal renewable energy storage. Among potential storage sites, facilities used for underground natural gas storage have advantages, including well-characterized cyclical injection-withdrawal behavior and partially reusable infrastructure. However, the differences between hydrogen-brine and natural gas-brine flow, particularly through fractures in the reservoir and the sealing caprock, remain unclear due to the complexity of two-phase flow. Therefore, we investigate fracture relative permeability for hydrogen versus methane (natural gas) and nitrogen (commonly used in laboratories). Steady-state relative permeability experiments were conducted at 10 MPa on fractured carbonate rock from the Loenhout natural gas storage in Belgium, where gas flows through {\textmu}m-to-mm scale fractures. Our results reveal that the hydrogen exhibits similar relative permeability curves to methane, but both are significantly lower than those measured for nitrogen. This implies that nitrogen cannot reliably serve as a proxy for hydrogen at typical reservoir pressures. The low relative permeabilities for hydrogen and methane indicate strong fluid phase interference, which traditional relative permeability models fail to capture. This is supported by our observation of periodic pressure fluctuations associated with intermittent fluid connectivity for hydrogen and methane. In conclusion, our findings suggest that the fundamental flow properties of fractured rocks are complex but relatively similar for hydrogen and natural gas. This is an important insight for predictive modeling of the conversion of Loenhout and similar natural gas storage facilities, which is crucial to evaluate their hydrogen storage efficiency and integrity.

Pore-scale imaging of hydrogen and methane storage in fractured aquifer rock: The impact of gas type on relative permeability

TL;DR

This study directly measures drainage relative permeabilities for -brine, -brine, and -brine in a fractured limestone core from Loenhout at , using X-ray tomography to resolve pore-scale distributions. Hydrogen and methane exhibit similar, but globally low, gas permeabilities due to strong phase interference and intermittent flow, while nitrogen shows substantially higher , revealing that nitrogen is an unreliable proxy for hydrogen under reservoir conditions. The results imply that hydrogen storage in fractured reservoirs may have reduced injectivity relative to methane, and that fracture morphology and intermittency must be captured in upscaled models; they also underscore the need for hydrogen-specific fracture data and caution against proxy-based upscaling using nitrogen. Limitations include a single artificially fractured sample and non-matching reservoir temperature, suggesting future work to explore imbibition relative permeabilities, varied fracture geometries, and upscaling to discrete fracture networks (DFN).

Abstract

Underground hydrogen storage in saline aquifers is a potential solution for seasonal renewable energy storage. Among potential storage sites, facilities used for underground natural gas storage have advantages, including well-characterized cyclical injection-withdrawal behavior and partially reusable infrastructure. However, the differences between hydrogen-brine and natural gas-brine flow, particularly through fractures in the reservoir and the sealing caprock, remain unclear due to the complexity of two-phase flow. Therefore, we investigate fracture relative permeability for hydrogen versus methane (natural gas) and nitrogen (commonly used in laboratories). Steady-state relative permeability experiments were conducted at 10 MPa on fractured carbonate rock from the Loenhout natural gas storage in Belgium, where gas flows through {\textmu}m-to-mm scale fractures. Our results reveal that the hydrogen exhibits similar relative permeability curves to methane, but both are significantly lower than those measured for nitrogen. This implies that nitrogen cannot reliably serve as a proxy for hydrogen at typical reservoir pressures. The low relative permeabilities for hydrogen and methane indicate strong fluid phase interference, which traditional relative permeability models fail to capture. This is supported by our observation of periodic pressure fluctuations associated with intermittent fluid connectivity for hydrogen and methane. In conclusion, our findings suggest that the fundamental flow properties of fractured rocks are complex but relatively similar for hydrogen and natural gas. This is an important insight for predictive modeling of the conversion of Loenhout and similar natural gas storage facilities, which is crucial to evaluate their hydrogen storage efficiency and integrity.

Paper Structure

This paper contains 12 sections, 12 figures, 1 table.

Figures (12)

  • Figure 1: (a) 3D rendering of fracture network of the core. (b) Porosity along the length of the core $<\Phi> = 0.011$. (c) 2D local thickness map of a horizontal slice of dry scan. (d) Aperture (local thickness) distribution with $<b>$ = 350.µm.
  • Figure 2: Experimental set-up for two-phase hydrogen/methane/nitrogen-brine core floods performed with X-ray micro computed tomography scanner.
  • Figure 3: Fluid occupancy at $f_g = 0.6$ for different gases: (a) H$_2$ (b) CH$_4$ (c) N$_2$. The distance is derived from the distance map of the segmented dry fracture image, calculated as twice the distance between the center and the nearest wall of the fracture.
  • Figure 4: Drainage steady-state relative permeability comparison between H$_2$-brine, CH$_4$-brine and N$_2$-brine system at 10 MPa and 22$^{\circ}$C.
  • Figure 5: 1-D slice-averaged saturation profile of brine along the length of the core for different gases. (a) H$_2$ (b) CH$_4$ (c) N$_2$.
  • ...and 7 more figures