DESIGN OF WATER-MODERATED HETEROGENEOUS CORES IN NEW STACY FACILITY THROUGH JAEA/IRSN COLLABORATION

摘要

In the post-Fukushima prospect, Japan Atomic Energy Agency, JAEA, intends to re-open its Static Experiment Critical Facility STACY, in order to contribute to the validation of fuel debris calculations in conditions close to those affected by the Fukushima Daiichi Nuclear Power Station accident. The reactor core of STACY is being modified by JAEA from homogeneous solution fuel core to heterogeneous pin-type fuel core. The first step of critical experiments is to establish reference experiments involving lattices of UO_2 rods without fuel debris. A design of such experiments is in progress for the widest range of moderation ratios in the frame of JAEA/IRSN collaboration. To anticipate the specification values of materials to be manufactured (grids, rods), an evaluation of uncertainties pertaining to the different parameters driving k_(eff) is paramount. The following uncertainties of experimental cores are evaluated to estimate the reactivity effect of engineering tolerance/or measurements of the modified STACY: (1) Density of fuel pellets, (2) Enrichment of uranium, (3) Diameter of fuel pellets, (4) Thickness of fuel claddings, (5) Impurities of fuel pellets, (6) Impurities of fuel claddings and (7) Positioning of fuel rods in the lattice. For these calculations, three types of moderation conditions were considered within the experimental parameters of STACY: the lowest moderation (lattice pitch = 1.15 cm; V_m/V_f=1.2), optimum moderation (lattice pitch = 1.5 cm, V_m/V_f = 2.9) and an over moderated case (lattice pitch = 2.55 cm; V_m/V_f = 11). The calculations to derive the overall uncertainty of modified STACY were performed propagating the lσ uncertainties of parameters of existing STACY in terms of k_(eff). For that purpose, a perturbation calculation (differential operator) technique and a stochastic geometry modeling technique were used with the MCNP5 continuous energy Monte Carlo code and a point-wise cross section library based on the JENDL-4.0 evaluated nuclear data library. Moreover, the results of calculations were compared with those calculated by two continuous energy Monte Carlo codes using the correlated sampling technique: 1. An experimental version of the Japanese MVP2 code combined to the JENDL-4.0 nuclear data library, 2. The MORET 5 code, developed by IRSN, combined to the JEFF-3.1 nuclear data library. The proposed paper discusses all the aforementioned issues, and confirms that the uncertainty pertaining to the critical experiments of the STACY is sufficiently small.