A Multi-Scale Approach to Fault-Valve Systems and Their Evolution

Cyclic fluid injection can promote repeated fault reactivation and transient permeability changes, a behavior often discussed within the fault-valve concept where faults alternate between acting as hydraulic barriers and conduits. Such cycles are relevant to both natural hydrothermal systems and industrial activities that modify pore pressure in the subsurface.

The complex evolution of fault permeability and strength during and after fault reactivation calls for a more complete description of the underlying physical processes. Current state-of-the-art fault reactivation models generally represent these weakening and strengthening mechanisms at the macroscopic scale using phenomenological laws, such as the widely used rate-and-state framework. Although these formulations have proven successful in reproducing some observed fault behaviors, they rely on empirically determined parameters and still leave part of the relevant physics insufficiently described.

Here we investigate these processes using a discrete element method (DEM) approach coupled with a pore-scale finite volume (PFV) scheme. Similarly to the framework proposed by Nguyen et al. (2021), the DEM models the granular gouge, while PFV simulates pore-pressure evolution and fluid flow through the evolving pore geometry, with full two-way coupling between solid deformation and fluid pressure, thus relating the macroscopic response of the system to the micromechanical phenomena at work.

Using this coupled approach, we simulate both monotonic and cyclic injection protocols designed to represent fault-valve cycles. We quantify how permeability evolves before, during, and after reactivation, and we explore the influence of key controlling factors: (i) initial permeability, (ii) initial stress state prior to injection, and (iii) confining stress. We also estimate seismic moments associated with individual reactivation events where the recovered moments remain bounded by the injected-volume constraint M_0,max =GΔV  (McGarr, 2014). Overall, by adding grain-scale observations to trends reported in laboratory and in situ studies, this work helps interpret permeability transients and their implications for triggered seismicity, in order to provide more realistic models of fluid-induced fault reactivation.

References
Nguyen, H. N. G., Scholtès, L., Guglielmi, Y., Donzé, F. V., Ouraga, Z., & Souley, M. (2021). Micromechanics of sheared granular layers activated by fluid pressurization. Geophysical Research Letters. https://doi.org/10.1029/2021GL093222
McGarr, A. (2014). Maximum magnitude earthquakes induced by fluid injection. Journal of Geophysical Research: Solid Earth, 119, 1008–1019. https://doi.org/10.1002/2013JB010597