Sweden is generally considered to be a low seismicity area, but for structures within nuclear power facilities, the safety level demand with respect to seismic events are high and thus, these structures are required to be earthquake-resistant. The seismic hazard is here primarily considered to be associated with near-field earthquakes. The nuclear power plants are further founded on hard rock and the expected ground motions are dominated by high frequencies. The design earthquake considered for the nuclear facilities has an annual probability of 10-5 events, that is, the probability of occurrence is once per 100 000 years. The focus of the study is the seismic response of large concrete structures for the nuclear power industry, with regard not only to the structure itself but also to non-structural components attached to the primary structure, and with emphasis on Swedish conditions. The aim of this licentiate thesis is to summarize and demonstrate some important aspects when the seismic load is dominated by high frequencies. Additionally, an overview of laws, regulations, codes, standards, and guidelines important for seismic analysis and design of nuclear power structures is provided.

The thesis includes two case studies investigating the effect of seismic high-frequency content loads. The first study investigates the influence of gaps in the piping supports on the response of a steel piping system subjected to a seismic load dominated by high amplitudes at high frequencies. The gaps are found in the joints of the strut supports or are gaps between the rigid box supports and the pipe. The piping system is assessed to be susceptible to high-frequency loads and is located within the reactor containment building of a nuclear power plant. The stress response of the pipe and the acceleration response of the valves are evaluated. The second study investigates the effect of fluid-structure interaction (FSI) on the response of an elevated rectangular water-containing concrete pool subjected to a seismic load with dominating low and high frequencies, respectively. The pool is located within the reactor containment building of a boiling water reactor at a nuclear power plant. The hydrodynamic pressure distribution is evaluated together with the stress distribution in the walls of the tank.

From the two case studies, it is evident that the response due to a seismic load dominated by high frequencies and low frequencies, respectively, is different. Although the seismic high-frequency load may be considered non-damaging for the structure, the effect may not be negligible for non-structural components attached to the primary structure. Including geometrical non-linear effects such as gaps may however reduce the response. It was shown that the stress response for most of the pipe elements in the first case study was reduced due to the gaps. It may also be that the inclusion of fluid-structure interaction effects changes the dynamic properties of a structural system so that it responds significantly in the high frequency range, thus making it more vulnerable to seismic loads dominated by high frequencies. In the second case study, it was shown that even for a seismic load with small amplitudes and short duration, but with dominating high-frequency content, as the Swedish 10-5 design earthquake, the increase of the dynamic response as fluid-structure interaction is accounted for is significant.