High frequency, high amplitude and low energy earthquake study of nuclear power plants

Abstract

Nuclear power plants are designed for a seismic input spectrum based on U.S. acceleration time histories. However, data recorded near several recent earthquakes, mostly in the Eastern U.S., are richer in high frequency energy. This paper focuses on the evaluation of one of these events, i.e., the 1986 Ohio earthquake approximately 10 miles from the Perry nuclear power plant. The Perry Seismic Category I structures were reanalyzed using the in-structure recorded earthquake motions. The calculated in-structure response spectra and recorded response spectra have the same general trends, which shows the buildings are capable of responding to high frequency earthquake motion. Dynamic stresses calculated using the Ohio earthquake recorded motions are substantially lower than the design stresses. The seismic qualification of a wide sample of equipment was reassessed using the Ohio earthquake recorded motions and the margins were found to be larger than one. The 1986 Ohio earthquake was also shown to possess much lower energy content and ductility demand than the design spectra. For the Perry case, the seismic design was shown to have adequate safety margins to accommodate the 1986 Ohio earthquake, even though the design spectra were exceeded at about 20 Hz. The NRC is evaluating the need to generically modify design spectra in light of the recent high frequency recordings.