• Journal of the Earthquake Engineering Society of Korea
• /
• v.18 no.6
• /
• pp.291-299
• /
• 2014

There has been an increasing demand for introducing a base isolation system to secure the seismic safety of a nuclear power plant. However, the design criteria and the safety assessment methodology of a base isolated nuclear facility are still being developed. A performance based design concept for the base isolation system needs to be added to the general seismic design procedures. For the base isolation system, the displacement responses of isolators excited by the extended design basis earthquake are important as well as the design displacement. The possible displacement response by the extended design basis earthquake should be limited less than the failure displacement of the isolator. The failure of isolators were investigated by an experimental test to define the ultimate strain level of rubber bearings. The uncertainty analysis, considering the variations of the mechanical properties of isolators and input ground motions, was performed to estimate the probabilistic distribution of the isolator displacement. The relationship of the displacement response by each ground motion level was compared in view of a period elongation and a reduction of damping. Finally, several examples of isolator parameters are calculated and the considerations for an acceptable isolation design is discussed.

• Nuclear Engineering and Technology
• /
• v.44 no.3
• /
• pp.225-236
• /
• 2012

The Fukushima-Daiichi accident shook the world, as a well-known plant design, the General Electric BWR Mark I, was heavily damaged in the tsunami, which followed the Great Japanese Earthquake of 11 March 2011. Plant safety functions were lost and, as both AC and DC failed, manoeuvrability of the plants at the site virtually came to a full stop. The traditional system of Emergency Operating Procedures (EOPs) and Severe Accident Management Guidelines (SAMG) failed to protect core and containment, and severe core damage resulted, followed by devastating hydrogen explosions and, finally, considerable radioactive releases. The root cause may not only have been that the design against tsunamis was incorrect, but that the defence against accidents in most power plants is based on traditional assumptions, such as Large Break LOCA as the limiting event, whereas there is no engineered design against severe accidents in most plants. Accidents beyond the licensed design basis have hardly been considered in the various designs, and if they were included, they often were not classified for their safety role, as most system safety classifications considered only design basis accidents. It is, hence, time to again consider the Design Basis Accident, and ask ourselves whether the time has not come to consider engineered safety functions to mitigate core damage accidents. Associated is a proper classification of those systems that do the job. Also associated are safety criteria, which so far are only related to 'public health and safety'; in reality, nuclear accidents cause few casualties, but create immense economical and societal effects-for which there are no criteria to be met. Severe accidents create an environment far surpassing the imagination of those who developed EOPs and SAMG, most of which was developed after Three Mile Island - an accident where all was still in place, except the insight in the event was lost. It requires fundamental changes in our present safety approach and safety thinking and, hence, also in our EOPs and SAMG, in order to prevent future 'Fukushimas'.

• Proceedings of the Korea Concrete Institute Conference
• /
• 1993.10a
• /
• pp.267-272
• /
• 1993

Currently about 60 contries in the world require earthquake resistant structural design in case of multistory building constructions. In these cases the equivalent lateral force procedure is commonly adopted because of its simplicity and convenience. This procedure, however, is developed based mainly on the first vibration mode response of building structure. The dynamic analysis of tall building shows that the effect of higher modes of vibration on the response of the building can not be neglected. In this paper, the effect of higher modes of vibration on seismic response is evaluated through modal analysis of tall building structures. On the basis of evaluation results, an improved procedure is to be proposed for the extended application of the equivalent lateral force procedure.

• Structural Engineering and Mechanics
• /
• v.42 no.1
• /
• pp.121-129
• /
• 2012

All contemporary seismic Codes have adopted smooth design acceleration response spectra, which have derived by statistical analysis of many elastic response spectra of natural accelerograms. The above smooth design spectra are characterized by two main branches, an horizontal branch that is 2.5 times higher than the peak ground acceleration, and a declining parabolic branch. According to Eurocode EN/1998, the period range of the horizontal, flat branch is extended from 0.1 s, for rock soils, up to 0.8 s for softer ones. However, from many natural recorded accelerograms of important earthquakes, the real spectral amplification factor appears to be much higher than 2.5 and this means that the spectrum leads to an unsafe seismic design of the structures. This point is an issue open to question and it is the object of the present study. In the present paper, the spectral amplification factor of the smooth design acceleration spectra is re-calculated on the grounds of a known "reliability index" for a desired probability of exceedance. As a pilot scheme, the seismic area of Greece is chosen, as it is the most seismically hazardous area in Europe. The accelerograms of the 82 most important earthquakes, which have occurred in Greece during the last 38 years, are used. The soil categories are taken into account according to EN/1998. The results that have been concluded from these data are compared with the results obtained from other strong earthquakes reported in the World literature.

• Structural Engineering and Mechanics
• /
• v.20 no.6
• /
• pp.655-672
• /
• 2005

Seismic random responses due to the wave passage effect are extensively investigated by using the pseudo excitation method (PEM). Two examples are used. The first is very simple but also very informative, while the second is a realistic suspension bridge. Numerical results show that the seismic responses vary significantly with wave speed, especially for low velocity or large span. Such variations are not monotonic, especially for flexible structures. The contributions of the dynamic and quasi-static components depend heavily on the seismic wave velocity and the natural frequencies of structures. For the lower natural frequency cases, the dynamic component has significant effects on the dynamic responses of the structure, whereas the quasi-static component dominates for higher natural frequencies unless the wave speed is also high. It is concluded that if insufficient data on local seismic wave velocity is available, it is advisable to select several possible velocity values in the seismic analysis and to choose the most conservative of the results thus obtained as the basis for design.

• Earthquakes and Structures
• /
• v.12 no.6
• /
• pp.653-663
• /
• 2017

A seismic margin assessment evaluates how much margin exists for the system under beyond design basis earthquake events. Specifically, the seismic margin for the entire system is evaluated by utilizing a systems analysis based on the sub-system and component seismic fragility data. Each seismic fragility curve is obtained by using empirical, experimental, and/or numerical simulation data. The systems analysis is generally performed by employing a fault tree analysis. However, the current practice has clear limitations in that it cannot deal with the uncertainties of basic components and accommodate the newly observed data. Therefore, in this paper, we present a Bayesian-based seismic margin assessment that is conducted using seismic fragility data and fault tree analysis including Bayesian inference. This proposed approach is first applied to the pooltype nuclear research reactor system for the quantitative evaluation of the seismic margin. The results show that the applied approach can allow updating by considering the newly available data/information at any level of the fault tree, and can identify critical scenarios modified due to new information. Also, given the seismic hazard information, this approach is further extended to the real-time risk evaluation. Thus, the proposed approach can finally be expected to solve the fundamental restrictions of the current method.