Hybrid BEM–FEM framework for PSHA-informed local seismic response assessment of small modular reactor sites

Small modular reactors (SMRs) are being developed as flexible nuclear energy systems for deployment across diverse site conditions, including urban, industrial, and remote areas. As safety-critical facilities with stringent seismic requirements, their assessment should account for local site effects that can modify the seismic response of foundations and structural components. While probabilistic seismic hazard analysis (PSHA) quantifies regional ground-motion levels and their probabilities of exceedance, local wave-propagation and soil–structure response modeling is essential for characterizing how terrain, soil, foundations, and structures modify the resulting site response. This study integrates an OpenQuake-based PSHA workflow with hybrid Boundary Element–Finite Element Method (BEM–FEM) modeling for a representative, idealized terrain–soil–SMR configuration. Hazard curves, uniform hazard spectra (UHS), V S 30 sensitivity, conditional spectra (CS), and deaggregation results are used to characterize regional hazard levels and dominant magnitude–distance contributions. The UHS ordinates are used as period-specific scaling amplitudes in a linear frequency-domain hybrid BEM–FEM model, while spatial phase variation and depth attenuation are introduced analytically. The BEM region represents the unbounded exterior rock domain, free-surface terrain relief, and wave radiation, while the FEM region contains the local valley soil, foundations, and SMR structures. Responses are evaluated using normalized transfer functions, UHS-based spectral-period-specific acceleration indicators and CS-selected broadband ground-motion response. The results show that the local terrain–soil–structure system can amplify, attenuate, and redistribute seismic motion across response periods, excitation directions, response components, and locations. The UHS-based and CS-selected broadband analyses show consistent dominant response characteristics while also resolving record-to-record variability and component-dependent redistribution of motion. The representative configuration demonstrates how period-dependent local transfer effects, spatial variability, and directional and cross-component interactions combine to shape the resulting acceleration response. The proposed workflow links regional hazard characterization with period-wise and broadband local response indicators relevant to SMR siting evaluation and feasibility screening, thereby providing a physics-based foundation for more detailed site-specific assessment.