Injection-induced aseismic slip in tight fractured rocks

TL;DR

This study models fluid injection in fractured rocks, revealing how fracture orientation and stress conditions influence aseismic slip and its relation to injected volume, with implications for understanding subsurface fluid-induced seismicity.

Contribution

The paper extends single-fracture aseismic slip models to complex 2D fracture networks, deriving closed-form solutions and identifying key parameters governing slip behavior.

Findings

Aseismic slip is governed by a single dimensionless parameter T.

In critically stressed conditions, slip propagates rapidly regardless of network percolation.

Aseismic moment scales quadratically with injected volume in both stress regimes.

Abstract

We investigate the problem of fluid injection at constant pressure in a 2D Discrete Fracture Network (DFN) with randomly oriented and uniformly distributed frictionally-stable fractures. We show that this problem shares similarities with the simpler scenario of injection in a single planar shear fracture, investigated by Bhattacharya and Viesca (2019); Viesca (2021) and whose results are here extended to include closed form solutions for aseismic moment as function of injected volume Vinj. Notably, we demonstrate that the hydro-mechanical response of the fractured rock mass is at first order governed by a single dimensionless parameter T associated with favourably oriented fractures: low values of T (critically stressed conditions) lead to fast migration of aseismic slip from injection point due to elastic stress transfer on critically stressed fractures. In this case, therefore, there is no effect of the DFN percolation number on the spatio-temporal evolution of aseismic slip. On the other hand, in marginally pressurized conditions (T > 1), the slipping patch lags behind the pressurized region and hence the percolation number affects to a first order the response of the medium. Furthermore, we show that the aseismic moment scales Vinj^2 in both limiting conditions, similarly to the case of a single planar fracture subjected to the same injection condition. The factor of proportionality, however, depends on the DFN characteristics in marginally pressurized conditions, while it appears to be only mildly dependent on the DFN properties in critically stressed conditions.