Uncertainty quantification for TRACE simulation of FIX-II No. 5052 test

• The Best Estimate Plus Uncertainty approach requires the knowledge of input uncertainties for the uncertainty propagation with best-estimate codes. Inaccurate judgement of some model parameter uncertainties related to the dominant physical phenomena may result in misestimation of the safety margin. This paper presents a framework of inverse uncertainty quantification (UQ) to quantify model parameter uncertainties in order to address this issue. It is applied to TRACE simulation of a large break loss of coolant accident conducted on the FIX-II facility, and peak cladding temperature (PCT) is the simulation output. Sensitivity analysis identifies the parameters of the critical flow model as the most influential to the PCT. The inverse UQ is performed based on Bayesian framework, which adopts Markov Chain Monte Carlo sampling and surrogate modelling algorithms. The quantified uncertainties of the model parameters are the desired results from the inverse UQ process, which are useful in BEPU studies.

Subject headings

NATURVETENSKAP  -- Fysik (hsv//swe)

NATURAL SCIENCES  -- Physical Sciences (hsv//eng)

Keyword

LOCA

Thermal-hydraulics simulation

TRACE

Uncertainty quantification

Publication and Content Type

ref (subject category)

art (subject category)