Volume 78, Issue 2

International experience has shown that external hazards can be safety significant contributors to the risk of industrial plants with a high potential of damage to the environment; this has become evident after the Fukushima accident. As a consequence main focus has been set on adequate design measures against external hazards and appropriate assessment methods. Possible methods to analyse existing plants systematically regarding the adequacy of their existing protection equipment against hazards can be deterministic as well as probabilistic. On international level, new recommendations regarding external hazards are recently issued. In that context, in particular earthquakes and flooding scenarios have been re-evaluated to some extent. Also in Germany, a revised guideline on probabilistic safety analyses (PSA) and corresponding technical documents are issued in 2005 addressing external hazards. As a reaction to the accidents in Japan in March 2011, the German Reactor Safety Commission has issued a catalogue of requirements for plant-specific reviews of German nuclear power plants, and the status of the plants have been evaluated with respect to these requirements. The results of these investigations have been the basis for the German report in the frame of the European stress tests; the results with respect to natural external hazards are presented.

As a consequence of the EU stress tests following the Fukushima NPP accident, the Nordic BWR Owners Group (NOG) initiated an analysis related to the capability of Swedish and Finnish BWR plants to withstand potential effects from external flooding and other extreme natural events. A review of existing analyses was performed with the intention to identify and reassess assumptions and simplifications made, and to judge whether all significant aspects of external events had been identified and considered. A previous NOG project dealing with methodologies for analyzing certain external events was used as a reference. Additional analyses were performed within three sub-projects addressing extreme water levels in the Baltic Sea and Skagerrak, methodology for the analysis of multiple external events, and main heat sink accessibility. It was concluded that the outcome from the previous NOG project can still be used as a basis, but some issues requiring further analysis were also identified.

Continuous and reliable sources of electrical energy are required during all modes of operation of a nuclear power plant in order to assure the safety of the plant. The electrical energy to the safety systems of the nuclear power plant is provided from the on-site power system. The objective of this paper is to analyze the nuclear power plant on-site power system reliability during normal operation and after the earthquakes with given intensity. The obtained results show the importance of the on-site power system reliability in general and battery supported section of the Class 1E power system in particular for the safety of the nuclear power plant. High level of reliability is obtained for the Class 1E and Non-1E power system during normal operation and safe shutdown earthquake. Obtained results show high level of reliability of a Class 1E power system after the earthquake with peak ground acceleration equal to the one measured at the Fukushima Daiichi nuclear power plant. Recommendation for revision of the current guideline for assessment of the acceptable station blackout duration capability is recommended considering the results of the analysis and identified deficiencies in the current guideline.

The German PSA Guideline and its technical documents on PSA methods and data require probabilistic safety analyses (PSA) to be carried out in the frame of safety reviews for nuclear power plants. Since 2005 this also includes a seismic PSA (SPSA) for sites with design earthquake intensities exceeding the value VII (MSK-64/EMS-98). It is shown how the plant model of PSA Level 1 for internal events can be extended on the level of fault tree basic events to get a quantifiable seismic plant model. A two-step screening procedure can be applied to derive the seismic equipment list (SEL) and a list of all possibly seismic-induced dependent equipment failures. The screening procedure is supported by a database. The database keeps at hand all the data and information to extend the plant model of PSA Level 1 in a proper manner.

Structures can be exposed to seismic loading. For structures of major importance, extreme seismic loadings have to be considered. The proof of safety for such loadings requires sophisticated analysis. This paper introduces an analysis method which of course still includes simplifications, but yields to a far more realistic estimation of the seismic load bearing capacity of reinforced concrete structures compared to common methods. It is based on the development of pushover curves and the application of time-histories for the dynamic model to a representative harmonic oscillator. Dynamic parameters of the oscillator, such as modal mass and damping are computed using a soil-structure-interaction analysis. Based on the pushover-curve nonlinear force-deformation-capacities are applied to the oscillator including hysteresis behaviour characteristics. The oscillator is then exposed to time-histories of several earthquakes. Based on this computation the ductility is computed. The ductility can be scaled based upon the scaling of the time-histories. Since both, the uncertainty of the earthquake by using different time-histories and the uncertainty of the structure by using characteristic and mean material values, are considered, the uncertainty of the structure under seismic loading can be explicitly represented by a fragility.

The events of Fukushima amplified the scientific interest in the improvement of methods for probabilistic safety assessment (PSA) of extreme external events. The assessment of consequences of external floods belongs to this group of events. The paper presents the key steps of methodology for probabilistic safety assessment of external floods and a recent application for a nuclear power plant in Switzerland. The presented methodology is an extension of earlier activities and provides more focus on the PSA methodology part that may be applicable also for other studies.

External explosions can significantly contribute to risk of damage for industrial plants. External explosions may origin from other plants in the neighbourhood, which store and operate with explosive substances, or from transport of such substances on road, rail, or water. If deflagration is involved, ignition will not necessarily occur at the place of the accident, but a cloud of a combustible gas-air mixture may develop, which will ignite at some distance depending on wind velocity. A probabilistic model has been developed to calculate frequencies of damage based on numerical integration or on Monte Carlo simulation. Geographical information systems provide map material for sites, roads, rail and rivers on a computer. Data has been collected and applied for a nuclear power plant in Germany as an example. The method, however, can be used for any type of plant subject to external explosion hazards.

In recent years international experience has shown that external hazards can be safety significant contributors to the risk for nuclear power plants because such hazards have the potential to reduce simultaneously the level of redundancy by damaging redundant systems. Therefore, effective protections against external hazards have to be accurately assessed applying deterministic, probabilistic or combined methodologies. With respect to the external hazard “explosion” – two different types of risky situations for the plant under consideration have to be investigated: the explosive material is available as a stationary source in the neighbourhood of the plant or the explosive material is mobile (i. e. by ships, rails or trucks). In the latter case, the situation changes with the varying distances. Moreover, the transport way could be a straight line or a bent which also has to be addressed in the calculations. Methods which can be applied to evaluate the probability of occurrence of an external explosion event are, e. g., fault tree analyses, event tree analyses and Monte Carlo simulations. Two case studies applying Monte Carlo simulations for assessing the probability of occurrence of external explosion events, given an accident with creation of an explosive gas cloud, are presented.