Three-dimensional fluid-driven frictional ruptures: theory and applications

Fluid-driven frictional ruptures are important in a broad range of subsurface engineering technologies and natural earthquake-related phenomena. Some examples of subsurface operations where borehole fluid injections can induce frictional slip are deep geothermal energy, CO2 and hydrogen geological storage, and wastewater disposal from oil and gas production, among many others. On the other hand, natural phenomena such as seismic swarms, aftershock sequences, and slow earthquakes, are often associated with transients of pore-fluid pressure and fault slip. Motivated by the aforementioned applications and phenomena, this doctoral thesis aims to mechanistically understand how pre-existing geological structural discontinuities such as fractures and faults slide due to the pressurization of pore fluids, in a three-dimensional configuration that can be used for quantitative comparisons with field observations and preliminary engineering designs.

We first develop a numerical solver for frictional slip and fault opening along pre-existing networks of discontinuities in 3D elastic media. We then examine a model with Coulombâ s friction to reproduce the initiation, propagation, and arrest of fluid-driven stable frictional ruptures. We show that a dimensionless number containing information about the initial stress state of the fault and the intensity of the injection governs the dynamics and shape of the ruptures in all its stages. Next, we extend the model to account for a friction coefficient that weakens with slip. This results in a broader range of fault slip behaviors, from unconditionally stable slip to dynamic instabilities. We quantify the conditions controlling both the propagation of stable slip and the nucleation of earthquakes. It is shown that the Coulombâ s friction model is both an early-time and late-time asymptotic solution of the more general slip-weakening model. After, we generalize some important fault rupture regimes to account for fairly arbitrary fluid injections. In particular, the connection between the expansion rate of the slipping surface and the history of injection volume rate is established.

Using field observations, we then apply and explore the implications of our modeling results to injection-induced seismicity â a critical concern in the geo-energy industry. We show how the history of injection rate may control the migration patterns of micro-seismicity, and how these patterns may contain important information about in-situ conditions such as the fault stress state. Further, we elaborate on how post-injection pulses of stable slip can continue triggering seismicity due to stress-transfer effects, long after fluid injections stop. Finally, we propose a theoretical scaling relation for the maximum magnitude of injection-induced slow slip events, which sheds light on how aseismic motions release potential energy during injection operations.

This doctoral research offers for the first time a quantitative and conceptual