Unexpected fault activation due to underground gas storage in produced reservoirs. Part II: Definition of safe operational bandwidths

TL;DR

This study tackles the problem of unexpected fault activation during underground gas storage in faulted Rotliegend reservoirs by deploying a physics-based, Coulomb-friction–driven FE framework (per Part I) with one-way hydro-poro-mechanical coupling to simulate CH$_4$, CO$_2$, $H_2$, and $N_2$ storage cycles. A two-stage sensitivity analysis identifies geometry (compartment offsets, fault dips), initial stress state, fault strength (cohesion, static/dynamic friction), and reservoir stiffness contrasts as the main drivers of fault stability, revealing that activation can occur within pressure ranges previously experienced during production. The results support qualitative safe operational bandwidth guidelines that tie upper/lower pressure bounds to observed seismic pressures during primary production and emphasize site-specific calibration; chemo-mechanical effects of stored gases appear to have limited impact on fault stability under the Rotliegend settings. These insights inform risk assessment and mitigation strategies for current and future UGS, hydrogen storage, and CCS operations in depleted fields, guiding more conservative, buffer-rich operation plans. The framework provides a tractable basis for evaluating seismic risk across different fluids and reservoir geometries, while acknowledging modeling simplifications that warrant future enhancements such as two-way coupling and time-dependent inelastic behavior.

Abstract

Underground gas storage is a versatile tool for managing energy resources and addressing pressing environmental concerns. While natural gas is stored in geological formations since the early 20th century, hydrogen has recently been considered as a potential candidate toward a more flexible and sustainable energy infrastructure. Furthermore, these formations can additionally capture gases that contribute to climate change, such as CO2. When such operations are implemented in faulted basins, however, safety concerns may arise due to the potential reactivation of pre-existing faults, which could trigger (micro)-seismicity events. In the Netherlands, it has been recently noted that fault reactivation can occur "unexpectedly" during the life of an underground gas storage (UGS) site, even when stress conditions are not expected to cause a failure. The present two-part work aims to develop a modeling framework to investigate the physical mechanisms causing such occurrences in previously produced gas reservoirs and define a safe operational bandwidth for pore pressure variation for UGS operations in the faulted reservoirs of the Upper Rotliegend Group, the Netherlands. This paper investigates in detail the mechanisms and crucial factors that result in fault reactivation at various stages of a UGS. The mathematical and numerical model described in Part I is used, also considering how the presence of stored gases may influence the mechanical properties of the reservoir and caprock, in particular the Young modulus. The study investigates the hazard of fault activation caused by the storage of different fluids for various purposes, such as long-term CO2 sequestration, CH4 and H2 injection and extraction cycles, and N2 injection as cushion gas. The results show how geological configuration, geomechanical properties, and reservoir operating conditions may increase the hazard of fault reactivation.

Figures (25)

• Figure 1: Sketch of the main stages during the reservoir lifespan in terms of pressure change over time for CH$_4$, CO$_2$, H$_2$ and N$_2$: Primary Production (PP) for CH$_4$ production, Storage (ST) for CO$_2$ and N$_2$ storage, Cushion Gas Injection (CGI) for H$_2$ and CH$_4$ injection, and Underground H$_2$ Storage (UHS) and Underground Gas Storage (UGS) cycles for seasonal pressure fluctuations during CH$_4$/H$_2$ injection and withdrawal. Notice that, according to the Netherlands regulation, $P$ must remain below $P_i$ regardless of the stage under consideration minEZK_grijpskerk2022tno_ondergrondse2018
• Figure 2: Key geological and seismological features of the Rotliegend formation. a): Tectonic provinces of the Carboniferous–Rotliegend Total Petroleum System with the hydrocarbon field centerpoints distinguished between oil and gas fields (modified after Gau03). b): Top view of the Norg UGS site (in green), with bounding fault traces (red lines) and recorded seismic events (blue circles). The dashed purple line surrounding Blocks 2a, 2b, and 2c illustrates an example of reservoir compartmentalization (modified after TNO15). c): Interpreted seismic profile across the northeast Netherlands, showing the Norg gas field located in fault-bounded blocks at Rotliegend level (modified after JAG07). d)t: Average pressure evolution at the Norg UGS reservoir over time, with annotated seismic events from 1993 and 1999 (modified after NAM16).
• Figure 3: (a) horizontal cross-section of the conceptual model at 2100 m depth. Faults F1 and F2 are inclined and aligned along the y-axis; F4 and F5 are vertical and aligned along the x-axis. Fault F3 separates Block 1 and Block 2, which may experience different pressure histories ($\Delta p_1$, $\Delta p_2$). (b) vertical sections of the conceptual model along the trace A-A and B-B. $\Delta \delta$ represents the possible variation in the dip angle of fault F3. This figure is not to scale to highlight the local features within the reservoir.
• Figure 4: (a): axonometric view of the 3D computational grid used for the geomechanical simulation and the embedded 2D grid used to represent the fault system. Right: cut of the model along a vertical plane of symmetry with the FE mesh grid on the back part and the discretization of fault planes with IE in the front part. The various colors represent the different portions of the domain in agreement with Figure \ref{['fig:conceptual-3D']}.
• Figure 5: (a) spatial location of the injection and production wells (blue and red for injection/production for CH$_4$, red for CO$_2$ and N$_2$ injection, and green for UHS cycles). (b) 3D perspective view of block 1, showing the surrounding fault planes and the subdivision into two sub-blocks, left (red) and right (blue), illustrated in the right-hand panels Right: normalized pressure distribution from the flow-dynamic simulation of CH$_4$ (c) and H$_2$ (d) within the right reservoir block. The grid splits vertically along the dashed line shown on the left panel to provide evidence of the pressure distribution along the vertical direction.