Modelling of Safety Functions of Nuclear Power Plants Powered by Fault Tree Analysis

Controlling and safety systems of nuclear power plant are complex, complicated and many times backed up systems which function is to reliable control technology and to prevent occurrence and spread of unwanted dangerous event. Such events could be compared based on potential risk they constitute. Risk has on principle two factors - consequence of dangerous event and frequency of its occurrence. By the term unwanted event it is understood failure or unexpected state of technology. The frequency of failures occurrence could be managed e. g. by high-quality maintenance. Consequences are declined by predefined actions of control systems and if this system is unable to manage the process back into standard boundaries, the function of safety system starts to be performed. In case of safety systems it is not possible (or it is very difficult) to test its function and reaction to the non-standard process option, but it is really essential to know the probability of proper performing of an action of the safety system, when its intervention is demanded. Due to the complexity of each safety functions the dependability (to be more concrete the unavailability) of safety functions is simulated by the Fault Tree Analysis method, powered by the software called RiskSpectrum. Submitted paper deals with the dilemma of choosing of relevant safety functions and their analysis with use of the method mentioned above. There will be introduced model of safety function, principle of the fault tree creation and its quantification based on the two kinds of data - predicted ones and real field data. At the end of the contribution there will be shown how to compare both results of availability computing and how to monitor development of the safety function year by year.