We address the coupled hydro-mechanical problem of fluid injection in a fractured rock mass in the limit of negligible matrix hydraulic diffusivity. The fact that fuid-induced aseismic slip may outpace pore-fluid diffusion along a planar fault is also valid in the case of fluid injection into a Discrete Fracture Network. Upon revisiting the problem of self-similar fluid driven shear crack propagation on a planar fault under critically stressed and marginally pressurized regimes, we show via numerical simulations that a critically stressed DFN exhibits fast aseismic slipping path that migrates away from both injection point and fluid front location. This scenario persists regardless the geometrical connectivity of the DFN, since elastic stress interactions between active fractures represents the main driving force for such fast slip propagation. On the other hand, the opposite scenario occurs on a marginally pressurized DFN, where the aseismic slip front is localized near the injection point and well within the pressurized region.
Anthropogenic fluid injection in the sub-surface lowers the local effective normal stresses potentially leading to inelastic deformations and slippage of pre-existing planes of discontinuities that necessarily exist at multiple scales in rocks. As a result, micro-seismicity (typically manifested as a cloud of events with moment magnitude lower than 2) migrates away from injection point. The enlarging cloud of micro-seismicity has been observed to correlate well with a diffusion process, suggesting the pore-pressure front evolution as main driving force (Shapiro et al. 1997, 2002, Parotidis et al. 2003, Hainzl & Ogata 2005, Albaric et al. 2014, Hainzl et al. 2012). Many observations reveal that the growth of the microseismic cloud in time is indeed bounded by a power-law which grows similarly to the diffusion length scale (inline-equation), where α is the hydraulic diffusivity (see a 2D sketch in Figure 1). As a result, some authors have directly used the migration of micro-seismic swarms to estimate a global hydraulic diffusivity, obtaining - in a number of cases - relatively large value compared to the ones estimated from core samples in the laboratory (Townend & Zoback 2000, Doan et al. 2006, Xue et al. 2013). Townend & Zoback (2000), for instance, have shown that the diffusivities inferred from micro-seismicity over-estimate the measured values from core samples by one to three order of magnitude. Although this mismatch can be attributed to the presence of highly hydraulically conductive fractures at a large scale, this may not be the only explanation. In particular, the hypothesis that the micro-seismic cloud is directly coinciding with the location of the fluid front does not necessary hold. Indeed, the micro-seismic front is a direct measure of the locus of deformation which may be located either ahead or within the pore-pressure disturbance.
