The effects of new cross-section evaluations on criticality of a typical material test research rector were analysed. The reactor selected for the study was Pakistan Research Reactor-I, which is a general purpose, swimming pool MTR type research reactor fueled with low enriched uranium.
Such a study was necessary for selection of the most suitable cross section library for future use in MTR type research reactors similar to PARR-I, as well as the investigation of those cross-sections in new evaluations whose effects on the reactor physics parameters are dominant among all upgraded data.
WIMSD libraries based on ENDFB-VI.8, JENDL-3.2, JEF-2.2 and JEFF-3.1 were used for the analysis. Static core calculations were carried out with reactor simulation codes WIMS-D4 and CITATION. The multiplication factor and neutron energy spectrum obtained using these different cross-section data sets were compared. As WIMS- D4 does not support extended data related to fuel depletion in these libraries, the code was upgraded to WIMS- D4S which can use these libraries for depletion calculations. For fuel cycle analyses, this upgraded code was coupled with CITATION by developing a burnup analyses system SARC (System for Analysis of Reactor Core). The fuel cycle analysis was carried out for the equilibrium core of PARR-I, and the effects of various cross sections were studied for burnup credits as well as production of various other nuclides in the core due to depletion of fuel.
The analysis carried out using simplified modelling of the reactor with WIMS-D4 code alone, concluded that the newly released cross-section libraries show significant and similar differences in neutron energy spectrum from that obtained using WIMS81 (i.e., the WIMSD library released with the code in 1981). It was found that the cross-sections of hydrogen (bound-in-water) are the main reason for these differences. For further analysis, the PARR-l core was modelled in three dimensions with WIMS-D4 and CITATION. The multiplication factor and neutron energy spectrum for first critical core of the reactor were calculated. Up to 14% differences were encountered in neutron energy spectra from new "libraries with that obtained using WIMS81. The keƒƒ values obtained using all newly released libraries were within 0.45% of the experimental value, whereas the WIMS81 library resulted in calculated value 1.05% larger than experimental value. Among all libraries, JENDL-3.2 based WlMSD library results were closest to the experimental values.
The use of only newly evaluated cross sections of hydrogen (bound-in-water) in WIMS81 library resulted in keƒƒ values within 0.43% to the experimental value. Whereas, the neutron energy spectra below 0.1 MeV agreed with new libraries within 0.5% except for JEF-2.2 hydrogen data, which resulted in 1% difference. Moreover the mutual differences among the new libraries in the thermal energy range are also due to these cross sections. The differences in the fast energy range are mainly due to differences in fission spectra. The replacement of fission spectra reduced significantly these differences in the fast energy range.
In order to reduce the simulation time especially in fuel cycle calculations, the effective thickness for water reflector on all sides and graphite thermal column were computed for all libraries. It was found that 13 cm water layer is sufficient, independent of the library used, whereas 32 cm thick graphite layer based on ENDFB- VI may be taken to represent the thermal column.
The full core burnup dependent neutronics calculations using a newly developed code system SARC and WIMSD multigroup library based on JENDL-3.2 are in good agreement with the experimental data. The calculated cycle length of equilibrium core was 40.3 EFPD (effective full power days) which is in good agreement with experimental value. It was also found that reactivity, power densities, thermal flux, depletion of 235U and 239Pu production, show linear relationship with burnup of the core. On these bases, the linear reactivity model can also be used for fuel cycle analyses of Material Test Research Reactors with LEV cores.
The SARC system was then applied to investigate the effects of various cross-sections of the isotopes present in equilibrium core. It was found that 3% life of operating cycle of equilibrium core of PARR-1 is due to the production of plutonium isotopes. The production of actinides in the core during equilibrium cycle was also examined using new data libraries. The maximum difference in concentrations among all produced actinides is for 242mAm from JEFF-3.1library with other libraries. The results from all other libraries are within 3% for this nuclide. From burnup chain analysis, it was concluded that differences in capture cross-sections of nuclides among various libraries are responsible for the differences in calculated amounts of actinides.