In English

Modeling of PWR LOCA experiments in RELAP5 based on PREMIUM benchmark

Magnusson Rickard
Göteborg : Chalmers tekniska högskola, 2013. 39 s.
[Examensarbete på avancerad nivå]

Safety in nuclear power plants is very important since release of radioactive ma- terials to the environment could be extremely dangerous. Thus reliable safety analyses are essential in order to design and operate this kind of systems. In the past, nuclear power plants were designed according to a conservative ap- proach so that possible uncertainties in the reactor modeling could be taken in account. However, this turned out to give non-realistic analyses and limited op- erational conditions. In order to obtain more realistic predictions and to improve safely the performance of the reactors, the scienti c community addressed huge e orts in the development of best-estimate methodologies and in the investiga- tion/quanti cation of the uncertainties that can a ect reactor calculations. In this framework the international project PREMIUM is focused on the study of the uncertainties in thermal-hydraulic models that are applied to simulate the nal phase of a Loss-of-Coolant Accident (LOCA), namely the core re ooding. Such a work is based on the core re ooding experiments that were carried out in the German test facility FEBA. In the current thesis, the FEBA experiments are simulated with the thermal- hydraulic program RELAP5. The sensitivity of these calculations is investigated with respect to a group of relevant modeling parameters. To do so, the software DAKOTA was employed. The RELAP5 model developed for this work provided satisfactory predictions of the behavior of the cladding temperature during the core re ooding phase of a Loss-Of-Coolant Accident. In addition, the sensitivity analysis showed which input parameters can have a relevant impact on the calculated results (e.g., mass ow).

Nyckelord: Nuclear Power Plants, Safety Analysis, RELAP5, Thermal-hydraulics,

Publikationen registrerades 2014-03-05.

CPL ID: 194451

Detta är en tjänst från Chalmers bibliotek