American Nuclear Society
Home

Home / Publications / Journals / Nuclear Technology / Volume 205 / Number 1-2

Uncertainty Analysis of the Conditional Exceedance Probability Calculation for a Probabilistically Significant SBO Sequence

Y. Du, H. X. Li, T. H. Liang, K. S. Liang

Nuclear Technology / Volume 205 / Number 1-2 / January-February 2019 / Pages 128-139

Technical Paper / dx.doi.org/10.1080/00295450.2018.1494998

Received:March 29, 2018
Accepted:June 26, 2018
Published:December 12, 2018

The Risk Informed Safety Margin Characterization methodology combines traditional probabilistic safety assessment (PSA) and the best-estimate plus uncertainty approach. Consequently, both stochastic uncertainty and epistemic uncertainty can be taken into overall consideration to evaluate the risk-informed safety margin. Generally, in calculation of the event sequence success criteria in traditional PSA, the result can only be either success (zero) or failure (unity), which is because uncertainties are not properly taken into consideration. In this paper, the conditional exceedance probability (CEP) of a probabilistically significant station blackout sequence of a typical three-loop pressurized water reactor was calculated with the consideration of both stochastic and epistemic uncertainties by using RELAP5. To get the probability density function of the peak cladding temperature (PCT) of a particular sequence and corresponding CEP, random sampling analysis of major plant status parameters and stochastic parameters was performed. It is assumed that the core is damaged when the PCT reaches 1477.6 K. Through the calculation of CEP of this specific sequence, it can be found that core damage will take place in a certain possibility between zero and unity when taking plant status uncertainties and stochastic uncertainties into consideration. Therefore, the core damage frequency (CDF) of any probabilistically significant sequence can be recalculated to get a more precise CEP.

With the application of the computational risk assessment method, not only can the conditional CDF be reasonably reduced, but also the revised model can be made sensitive to a system design change of limited scope. Compared to the traditional PSA evaluation without uncertainty analysis, the CDF of the loss–of–heat sink dominant group can be reduced by a factor of 8.75 (/).

 
Questions or comments about the site? Contact the ANS Webmaster.
advertisement