SNEAK-12 Experiment – Two Years of Analysis

Significant knowledge gaps exist in the scientific literature regarding the neutron dynamic behavior in a reactor core during severe accident progression. In case of light water reactors, the neutronic aspect is negligible and largely overlooked. In fast reactors, however, recriticality due to material relocation and reflooding is a major problem. The super prompt criticality excursion due to material relocation needs additional examination. The experimental data from experimental programs that are dedicated to material relocation available for neutronic code validation is scarce. The SNEAK-12 is an excellent example for such a program, which aims to study the effect of material relocation in fast lattices.

The SNEAK-12 experimental data restoration project consisted of several major steps. First, both technological and experimental data were gradually recovered from various sources at CEA Cadarache and Karlsruhe Institute of Technology, e.g., SNEDAX data base, technical notes, published papers and personal communication with participants to the project. The data was thoroughly analyzed and compiled into a set of benchmark problems, for different degradation situations. The experiments simulated various sizes of axial and radial fuel slump-in/out, molten pool formation and radial expansion, streaming channels blockage and utilization of different fuels. Second, the data was analyzed with state-of-the-art nuclear codes, such as Serpent 2, MCNP5 and Tripoli MC Codes, as well as the reference ERANOS deterministic code for SFR applications. Static calculations, however, do not provide a full picture of the code performance. Therefore, the third step include a comprehensive analysis of the sensitivity and uncertainty associated with the JEFF-3.1.1 nuclear data library, utilizing the COvariance Matrix Cadarache (COMAC) V01. The sensitivity analysis shows some interesting sensitivity coefficients behavior for the reactivity variations. The uncertainty analysis shows a high level of uncertainty in the estimation of the criticality for all the experimental multiplication factors. However, for the reactivity variation uncertainty the answer was slightly different with respect to the type of experiment. Finally, some questions regarding the performance of the newest JEFF-3.2 data library are raised. The JEFF-3.2 based analysis of the SNEAK-12 experiments show a constant discrepancy. The analysis of this discrepancy results in a feedback (cross section adjustments) on the nuclear data, in particular the capture cross section of ²³⁵U.

Currently, the SNEAK-12 experimental program is under preparation for the OECD IRPhE project, as part of the benchmark book evaluation.