Probabilistic safety assessment of operation

This thesis deals with the method of modelling event trees and fault trees within the probabilistic safety assessment at the Dukovany nuclear power plant. In the first part, probabilistic safety assessment is presented, along with a description of the used analyses, specifically ETA and FTA, and the software used for development of probabilistic models at nuclear power plants in the Czech Republic. Then Large LOCA accident is described, for which a pilot model is developed in the second part of this thesis. The main aim of this thesis is a replacement of the original modelling method with a new modelling. The presented method focuses, in addition to a development of event trees and fault trees, primarily on the substitution of previously used house event for initiating events. In order to prove correctness of the created model, the cut sets from both models are compared. The benefit of the work is the elaboration of a pilot model that utilizes a different modelling method. Thanks to the presented modelling method, the existing number of event trees can be significantly reduced, and manual setting of house events can be removed. This approach simplifies the work with the model and shortens the calculation time