Influence of fracture network connectivity and size distribution on injection-induced seismicity in two-dimensional fractured rocks

Le résumé fourni par la source

Fluid injection into fractured rocks can trigger seismicity through hydro-mechanical interactions. We perform 2D fully coupled hydro-mechanical simulations, based on the finite element method and the discrete fracture network approach, to quantify how fracture network geometry controls pressure diffusion, damage growth, fracture slip, and induced seismicity. Two parameters are used to constrain the generation of fracture networks: the power law length exponent a and the fracture intensity γ. Together they set the connectivity, measured by the percolation parameter χ. The generated networks span the disconnected (χ < χc) to well-connected (χ > χc) regimes, where the percolation threshold χc ≈ 6 for 2D fracture networks. Our results reveal that well-connected systems accommodate coherent pressure fronts, broad damage halos, and a radially migrating seismic band; disconnected systems produce intermittent pressure pockets, localized damage, and clustered seismic events. Fracture length-magnitude scaling is steeper for sparse networks, where slip localizes along individual long fractures, whereas dense networks partition slip across multiple fractures. Radial pressure profiles reveal a rapid decay of mean pressure elevation near the injection well, but persistent peak perturbations at fracture cluster boundaries. These peaks coincide with Coulomb failure stress hotspots and delayed seismic events. Fracture network connectivity governs how far induced seismicity extends from the injection point and how it is partitioned between matrix damage and fracture slip. These findings link fracture network structures to the spatiotemporal evolution of induced seismicity, with implications for enhanced geothermal systems and other injection related subsurface applications.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.