Assessment of High-Frequency Earthquake Excitation for Nonstructural Components of Nuclear Power Plant

Abstract

Structures perform as a system by integrating various nonstructural components during the earthquake. System failures occur because of nonstructural damage even though the structural components perform well. However, less attention has been paid to the design of nonstructural components in comparison with structural ones. Therefore, assessment of the seismic demand is important for nonstructural components such as critical infrastructures. This study reviews design standards (ASCE 7-22, NERHP-2020, FEMA P-58 2018, Eurocode 8-2004 and Japanese Regulatory Guide-2013) and prediction of the peak floor acceleration (PFA) which is a major input in determining the seismic force. These standards are applied to predict PFA for the nuclear power plant during the 2011 Tohoku Earthquake, where its foundation is deeply embedded under the ground surface. Moreover, the study applies random vibration theory (RVT) to improve the prediction of PFA by incorporating theoretical (TTF) and empirical transfer functions (ETF) of soil-structure interaction (SSI) and structural responses. Different combinations of TTF and ETF are investigated to understand the applicability and limitations of these TFs. The results show that the RVT works well to predict PFA even though SSI is modelled by TTF. It also shows that the application of ETF requires specific care for high-frequency coherency for which high-cut filter is required to remove the low-coherent signals. The prediction of PFA is substantially improved using RVT with TTF and ETF. Moreover, this study also investigated the influence of site attenuation parameters (𝜅𝑜), SSI and frequency content of PGA and should be considered for the correct prediction of floor acceleration, especially for important infrastructure such as Nuclear Power Plants.

Date of Award	Apr 2023
Original language	American English
Supervisor	Tadahiro Kishida (Supervisor)