Development of a standard data base for FBR core nuclear design (V); Consistency evaluation of JUPITER experimental analysis

not registered; not registered; not registered; Sato, Wakaei*; not registered; Sanda, Toshio*

In order to improve the design method and accuracy of large fast breeder cores, extensive work has been performed to accumulate and evaluate many kinds of results of fast reactor physics experiments and analyses. As a part of efforts to develop a standard data base for LMFBR core nuclear design, the present report evaluates the physical consistency of JUPITER experimental analysis, especially concentrating on criticality. Here, the judgment of consistency is based on not only the deviation degree of C/E values from unity, but also various viewpoints such as the comparison with other cores or other nuclear characteristics by sensitivity analysis, the effect of changing nuclear data library, the analysis of FCA and JOYO which have completely different source of data from JUPITER, and the use of the Monte Carlo method as an analytical reference. (1)The C/E values of JUPITER criticality are slightly underestimated in the range of 0.993-0.999, using the JFS-3-J2 (1989) group constant set based on JENDL-2 and three-dimensional XYZ transport theory with the most detailed analytical model. There is an obvious dependency of C/Es on reactor core concepts with homogeneous or heterogeneous structure, the main cause of which is considered to be the effect of internal blanket existence and cross-section errors of JFS-3-J2, judged from sensitivity analysis. (2)The latest analytical method and model based on three-dimensional XYZ transport theory has sufficient ability to predict the relative changes of JUPITER criticality caused by the effect of reactor core size, CRP sodium channel, control rod and internal blankets. (3)The analytical error of JUPITER criticality was evaluated as approximately 0.3%dk and this seems reasonable, because the results of Monte Carlo analysis for ZPPR-9 criticality were almost identical with those of our standard analytical method. (4)The analytical results based on the latest JENDL-3.2 library were very close to those of JENDL-2 results, ...