Wave Propagation in Fluid-Filled Horizontal Fracture Networks: Bridging Effective Medium Theories and Full-Waveform Simulations

Abstract:

Geothermal systems extract energy from fluids circulating through fractured rocks. Their productivity is influenced by the subsurface’s capacity to accommodate fluid, which also alters the elastic properties of the rock. Fracture geometry therefore controls both the flow paths for injected fluids and the propagation of elastic waves. Reservoir properties are often constrained using geophysical observations such as seismic tomography or well-scale measurements. Improving predictions of reservoir performance requires relating these techniques to microscale fracture characteristics, such as fracture length, density, aspect ratio, or orientation.

Effective Medium Theories (EMTs) estimate the bulk elastic properties of fractured media by representing complex fracture networks as an equivalent homogeneous medium. These approaches enable linking seismic velocities to the properties of the rock matrix and fracture parameters such as density, orientation, and fluid filling of fractures.

In this study, we compare EMT predictions with numerical wave propagation simulations in media containing horizontal, water-filled elliptical fractures. First, stochastic Discrete Fracture Networks (DFN) are built for different fracture lengths and densities and meshed with Gmsh. Wave propagation in these media is then simulated using SPECFEM. P-wave arrivals are retrieved from the synthetics for different configurations of fluid pockets to assess how the fracture distribution, and more specifically the length and density of fluid pockets, affects the effective properties of the host rock.

Results show that horizontal fractures induce strong seismic anisotropy, with velocities parallel to fractures remaining close to the host rock value but significantly reduced in the perpendicular direction as the density and length of fluid pockets increase. EMTs reproduce velocity trends at low fracture porosity but show increasing discrepancies as fracture density or length increases.

To link microscale properties of fluid pockets with macroscale observations, we propose expressing anisotropy as a function of fracture-induced porosity, as this parameter remains more accessible than the detailed characteristics of the fracture network.

Short bio:

Diplômée de l’École Normale Supérieure de Paris et Docteur en géophysique de l’Institut de Physique du Globe de Paris (IPGP) et de UC Berkeley, j’ai travaillé sur la structure de la graine à partir de l’analyse de phases sismiques durant ma thèse. J’ai ensuite rejoint l’Université de Lorraine pour me focaliser sur l’évaluation des propriétés effectives de milieux fracturés à partir de l’homogénéisation non périodique. J’ai ensuite fondé une entreprise pour travailler sur l’évaluation des risques climatiques et sismiques grâce à la génération de modèles de subsurface couplant hydrologie et propriétés géologiques.

Je suis actuellement chercheuse postdoctorale au Helmholtz Centre for Environmental Research (UFZ) en Allemagne. Mes travaux portent sur les échanges entre nappes et rivières et la manière dont ces dynamiques contrôlent la réponse des aquifères aux sécheresses. En parallèle, je m’intéresse à l’effet des poches de fluide sur les signaux sismiques pour fournir une description plus fine de la subsurface à partir de données macro-échelle et contribuer à la compréhension des systèmes géothermiques.