Development of a standard data base for FBR core nuclear design, VIII; Compilation of JUPITER analytical results

Ishikawa, Makoto; Sato, Wakaei*; Sugino, Kazuteru; Yokoyama, Kenji   ; Numata, Kazuyuki*; Iwai, Takehiko*

A Standard data base for LMFBR core nuclear design has been developed to improve analytical methods and prediction accuracy of nuclear design for large fast breeder cores such as demonstration or commercial FBRs. To develop the data base, extensive work has been prerformed to accumulate and evaluate many kinds of results from fast reactor physics experiments and their analyses. The present report summarizes the analytical results of the JUPITER experiments, using the most recent nuclear data library (JENDEL-3.2) and the lates analytical methods in a consistent manner. In the present work, a great emphasis was placed on guaranteeing the essential requirements for this kind of general data base, that is, "Accountability", "Traceability" and "Consistency". In other words, consistent strategies and analytical methods were applied to all calculations, including detialed corrections; the enormous analytical input data generated were all saved in the form of computer files, so that reanalysis of any experiment could be easily performed for verification or in response to future improvement in nuclear data or analytical methods. The main results of the present JUPITER analysis are as follows: (1) The C/E(calculation/experiment) values of criticality were slightly underestimated by -0.7-0.3%k. (2) The reaction rate ratio of C28/F49 was overestimated by +4+6% with the standard analytical method. However it was found to improve about 2% after the cell factors were revised using the Monte Carlo method. (3) The radial space-dependency of the reaction rate distribution and control rod worth almost disappeared in the homogeneous cores. (4) The previous overestimation of sodium void reactivity was greatly improved in the homogeneous cores.