A reliable assessment of the final inventory in the fertile blanket of a Fast Neutron Reactor is an important issue for fuel cycle physics, as it impacts safety, reprocessing and design studies. The performances of fertile blankets neutronics calculations is a long standing issue, as there as specificities in these regions that are quite challenging, especially for depletion codes.The new CEA fuel depletion calculation package is DARWIN3-SFR, which incorporates the deterministic neutronics code APOLLO3® and the depletion module MENDEL. This paper details the validation of DARWIN3-SFR for fertile blanket calculations through the re-interpretation of the DOUBLON pin-irradiation in the Phenix reactor.During this irradiation, nine pins are studied through isotopic ratios, which give us information about the depletion and, indirectly, the neutron flux calculations inside the blanket. This environment is especially challenging for neutronics codes, since there is a strong variation of the neutron energy and population over a short distance. Our recent analysis of TRAPU - which is a similar experiment, but in the core center - has proven DARWIN3-SFR to be reliable for the fuel depletion calculations of fissile subassemblies; nevertheless, its performances in the fertile blankets still require validation.We observe that, once the calculated neutron flux level has been adjusted to the experiment through the 148Nd/238U ratio, DARWIN3-SFR provides results similar to the reference stochastic code TRIPOLI-4®. However, both codes have difficulties to reproduce some of the measured isotopic ratios inside the fertile blanket. Whilst DARWIN3-SFR produces identical results for most of the isotopes analysed (234U, 235U, 236U, 238Pu, 239Pu, 240Pu), variations of the neutron spectrum lead to some disparities for the production of 241Pu and 242Pu. Indeed, the low-energy flux estimation is higher with TRIPOLI-4® than with DARWIN3-SFR in the energy range where the 240Pu has a high capture cross section, which increases the calculated production of 241Pu and 242Pu.DARWIN3-SFR and its predecessor DARWIN-2 are efficient at calculating the average neutron flux level over the entire fertile blanket. However, the results of the two codes show a strong pin-to-pin dispersion, resulting in a different shape of the neutron flux inside the blanket. With DARWIN3-SFR, the estimated neutron spectrum is softer than in DARWIN-2, which impacts the 238U capture and fission reaction rates and hence the production of plutonium.
Read full abstract