Benchmark of evaluated nuclear data libraries using post-irradiation experimental data on fuel composition changes of the fast reactor JOYO

Post irradiation experiment (PIE) data on depleted fuel composition of the experimental fast reactor JOYO MK-II have been accumulated since 1986 by the JNC (the former of JAEA). In the present study, all available PIE data of JOYO MK-II driver fuel were analyzed and integral data concerning 235 U depletion and 236 U generation were prepared. Both the integral data are sensitive to 235 U capture cross section and applicable to nuclear data benchmarks. The recent evaluated nuclear data libraries, JENDL-3.2, -3.3, JEFF-3.1 and ENDF/B-VII, have a tendency to overestimate the generation of 236 U. A cross section adjustment demonstrated that re-evaluation of 235 U capture cross section improved the overestimation. In fast reactor core design, it is important to improve the prediction accuracy of nuclear characteristics. A cross sec- tion adjustment technique using integral experimental data is an effective approach for the reduction of nuclear data induced uncertainties among analytical results. An adjusted cross section set, ADJ2000, based on a fast reactor experiment data base was developed in the past (1). In the adjustments, a few hundred integral data were adopted; however, almost of these came from zero-power experiments. In order to improve reliability and applicability for power reactors, burn- up characteristics should be added to next version of the adjusted cross section library. At the JAEA, measurement of burn-up in post-irradiation experiment (PIE) is performed. In the experimental fast reactor JOYO MK-II, a total of 362 driver fuel and 69 irradiation test subassemblies have been irradiated. PIE data on depleted fuel composition of JOYO have been accumulated since 1986 (2). In the present study, all the available PIE data of JOYO MK-II driver fuel were analyzed from the perspective of prediction accuracy evaluation of the isotopic composition change. The measured isotopic composition was translated to the rate of change of atomic number density (RC) in order to discuss prediction uncertainties induced from cross section data. The measured RCs were utilized for a benchmark calculation of the recent evaluated nuclear data libraries, JENDL-3.2 (3), JENDL-3.3 (4), JEFF-3.1 (5) and ENDF/B-VII (6).