Simulating Earthquake Rupture and Near-Fault Fracture Response

Sammanfattning: Sweden is presently a low seismicity area where most earthquakes are small and pose no serious threat to constructions. For the long-term perspectives of safety assessments of geological repositories for spent nuclear fuel, however, the effects of large earthquakes have to be considered. For the Swedish nuclear waste storage concept, seismically induced secondary fracture shear displacements across waste canister positions could pose a long-term seismic risk to the repository.In this thesis, I present earthquake simulations with which I study the potential for near-fault secondary fracture shear displacements. As a measure I use the Coulomb Failure Stress (CFS), but also calculate explicit fracture displacements. I account for both the dynamic and quasi-static stress perturbations generated during the earthquake. As numerical tool I use the 3DEC code, whose performance I validate using Stokes closed-form solution and the Compsyn code as benchmarks. In a model of a Mw 6.4 earthquake, I investigate how fault roughness, the fault rupture propagation model and rupture velocity may impact the near-fault CFS evolution. I find that fault roughness can reduce the amount of fault slip by tens of percent, but also increase the near-fault CFS with similar amounts locally. Furthermore, different fault rupture models generate similar CFS levels. I also find that the secondary stresses scale with rupture velocity.In a model based on data from the Forsmark nuclear waste repository site, and assuming stress conditions prevailing at the end of a glaciation, I simulate several high stress drop ~Mw 5.6 earthquake scenarios on the gently dipping fault zone ZFMA2 and calculate secondary fracture displacements on 300 m diameter planar fractures. Less than 1% of the fractures at the shortest distance from ZFMA2 generate displacements exceeding the 50 mm criterion established by the Swedish Nuclear Fuel and Waste Management Co. Given the high stress drops and the assumption of fracture planarity, I consider the calculated displacements to represent upper bound estimates of possible secondary displacements at Forsmark. Hence, the results should strengthen the confidence in the safety assessment of the nuclear waste repository at the Forsmark site.