Effects of Fracture Connectivity on Rayleigh Wave Dispersion

Abstract

Passive seismic characterization is an environmentally friendly method to estimate the seismic properties of the subsurface. Among its applications, we find the monitoring of geothermal reservoirs. One key characteristic to ensure a productive management of these reservoirs is the degree of fracture connectivity and its evolution, as it affects the flow of fluids within the formation. In this work, we explore the effects of fracture connectivity on Rayleigh wave velocity dispersion accounting for wave‐induced fluid pressure diffusion (FPD) effects. To this end, we consider a stratified reservoir model with a fractured water‐bearing formation. For the stochastic fracture network prevailing in this formation, we consider varying levels of fracture density and connectivity. A numerical upscaling procedure that accounts for FPD effects is employed to determine the corresponding body wave velocities. We use a Monte‐Carlo‐type approach to obtain these velocities and incorporate them in the considered fractured reservoir model to assess the sensitivity of Rayleigh wave velocity dispersion to fracture connectivity. Our results show that Rayleigh wave phase and group velocities exhibit a significant sensitivity to the degree of fracture connectivity, which is mainly due to a reduction of the stiffening effect of the fluid residing in connected fractures in response to wave‐induced FPD. These effects cannot be accounted for by classical elastic approaches. This suggests that Rayleigh wave velocity changes, which are commonly associated with changes in fracture density, may also be related to changes in interconnectivity of pre‐existing or newly generated fractures.

Dynamic SV‐Wave Signatures of Fluid‐Saturated Porous Rocks Containing Intersecting Fractures

“…To this end, Quiroga et al. (2022) calculated the P‐ and S‐ wave velocities in the layer with different fracture connectivity through numerical simulations, based on which the Rayleigh wave velocity dispersion was computed. They found that the fracture connectivity can have significant influence on Rayleigh wave velocity dispersion.…”

Seismic signatures of partial steam saturation in fractured geothermal reservoirs: Insights from poroelasticity

“…This result shows that employing classic elastic approaches may lead to an overestimation of the sensitivity of S-wave velocities to changes in saturation in fractured media. It is worth noting that, while S-wave velocities appear to be insensitive to changes in saturation in fractured media for the fully connected and randomly connected cases, previous research shows that they are sensitive to changes in fracture density in a geothermal reservoir context (e.g., Quiroga et al, 2022).…”

“…When considering complex fracture networks, generating samples of the medium that are large enough to constitute a REV may not be feasible. To overcome this difficulty, we follow the approach of Rubino et al (2009) and Quiroga et al (2022), who employ the previously outlined upscaling approach in a Monte Carlo fashion on sub-REVsize samples that are within our numerical capabilities. The Monte Carlo procedure consists of obtaining representative mechanical properties by averaging a sufficient number of stochastic realizations of samples with the same statistical properties.…”

“…The latter is widely regarded as a seemingly universal and ubiquitous characteristic of fractures (e.g., de Dreuzy et al, 2001;Bonnet et al, 2001). Following previous works on this topic (e.g., Hunziker et al, 2018;Quiroga et al, 2022), we use…”

“…Given that changes of both fracture density and steam saturation can result from pressure fluctuations in geothermal reservoirs, let us analyze the sensitivity of P-and S-wave velocities to both parameters. For this, we use data from Quiroga et al (2022) where the sensitivity of P-and S-wave velocity to changes in fracture density was Seismic signatures of steam (2022). Both increments in S f s and F d are associated with similar relative decrements in P-wave velocity, therefore, it would not be possible to distinguish steam variations from fracture density changes using this parameter alone.…”