Advanced Studies and Statistical Treatment for Sodium-Cooled Fast Reactor Pin Failures During Unprotected Transient Overpower Accident

Abstract

Usually, simulation tools are validated based on experimental data considering a best estimate simulation case; however, there is no quantification of this validation, which remains based on rough expert judgment. This technical note presents advanced validation treatment of the simulation tool OCARINa devoted to unprotected transient overpower (UTOP) accidents on two CABRI tests considering this time a Best Estimate Plus Uncertainty (BEPU) approach. The output results of interest are both scalar physical data such as the time and location of the pin failure and associated molten mass and vector data such as temperature axial distribution or temperature evolution versus time. This approach is a first step in quantifying the degree of agreement between the calculation results and the experimental results. It is of great interest for the verification, validation, and uncertainty quantification approach, which leads to the qualification of scientific calculation tools.

Within the framework of the Generation IV Sodium-cooled Fast Reactor (SFR) research and development project in which the CEA is involved, OCARINa is a physical tool, relevant for performing preconceptual design studies and devoted to simulation of UTOP accidents on heterogeneous cores. Such accidents could not be simulated with mechanistic calculation tools such as SAS4A or SIMMER with their current capabilities; the thermomechanical models are not finalized in the SIMMER tool, and the SAS4A tool is validated only for homogeneous cores. The final objective aims at deriving the variability of the main results of interest to quantify the safety margins.

The final use of the OCARINa tool being to perform sensitivity studies on the various possible sodium fast nuclear preconceptual core designs, the validation of this tool is first discussed at the pin scale (where separate-effects test measurements are available) based on statistical treatment. This enables one to determine the lacks and uncertainties of this tool. The modeling is then extended from local pin behavior to global core behavior adding a point-kinetics neutronic model. Final simulations of UTOP accidents caused by a uniform space reactivity ramp on an SFR core are realized taking into account the specificities of the pins of the various assemblies. The orders of magnitude of mechanical energy released are derived.

Keywords:

• Validation
• CABRI
• unprotected transient overpower accident
• OCARINa tool, SFR

Notes

^a Kernel density estimation is a nonparametric way to estimate the pdf of a random variable. Kernel density estimation is a fundamental datum smoothing problem where inferences about the population are made, based on a finite data sample.

^b 1$ = 370 pcm in the core.