Experiments and analyses on reactivity worth of gas expansion module (GEM) in FCA XX-1. Joint research

Gas Expansion Module (GEM) will be installed in large scale fast reactors in order to insert negative reactivity feedback in a transition of the primary pump trip. A mock-up experiment was performed at the XX-1 core of the Fast Critical Assembly (FCA) to evaluate and improve the prediction accuracy of the reactivity effect inserted by the GEM (GEM reactivity worth). The GEM was categorized into 'External GEM' and 'In-core GEM' according to its position; the experiment was performed for both cases, and the prediction accuracy was discussed. As for the external GEM reactivity worth, it was indicated that the self-shielding effect for iron used as the structural material of the GEM should be taken into account in the preparation process of the effective cross section of the GEM region. The overestimation of 10-20% was observed when the deterministic calculation was adopted for the transport correction in RZ geometry. When the Monte-Carlo method was used instead, the overestimation was reduced below 10%. As for the in-core GEM reactivity worth, it was pointed out that the cell homogenization should be carefully carried out since the neutron streaming affects the prediction accuracy. The prediction accuracy shows a dependence on the sodium level in the GEM. The prediction accuracy was evaluated by separating it into the non-leakage term and the leakage term by a least-squares fitting. It was found that the non-leakage term shows 10-20% overestimation, while the leakage term agrees well with the measurement within the experimental error. (author)