A New Jacobian Matrix Method for Assessing Similarity between Critical Experiments and Real Reactor

For a metal fueled Sodium-cooled Fast Reactor (SFR), innovative reactors such as Prototype Gen-IV Sodium cooled Fast Reactor (PGSFR), unfortunately, experiment data from an operating reactor are unavailable because there are few operating reactors in the world. Hence, a critical experiment is the only way to obtain meaningful experiment data for the target core. However, there is a considerable geometrical difference between the critical assembly for a critical experiment and the target core. The neutron characteristics of a system are influenced by the geometrical difference. A number of researches have been performed to confirm the similarity between a critical experiment and a real reactor using a conventional representativity factor. The conventional representativity factor defined as S^T_EUS_R √S^T_EUS_E, √S^T_RUS_R provides insight of similarity between two sensitivity vectors for cross sections, but it did not provide a quantitative value. Hence, up to now, the influence of geometrical difference to the reactivity is believed to be negligible. In this paper, a new Jacobian matrix method is proposed to provide a quantitative error for geometrical differences between two systems. In this method, reactivity of the critical assembly is decomposed into phenomenon-based reactivity and geometry-based reactivity. The reactivity is then transformed into a target core geometry using a Jacobian matrix. Then, non-linearity of two different systems can be derived by comparing the transformed reactivity with the original reactivity of the target core. The maximum error of the transformed reactivity can be used as an additional uncertainty of the geometrical difference