Statistical uncertainties of nondestructive assay for spent nuclear fuel by using nuclear resonance fluorescence

We estimated statistical uncertainties of a nondestructive assay system using nuclear resonance fluorescence (NRF) for spent nuclear fuel including low-concentrations of actinide nuclei with an intense, mono-energetic photon beam. Background counts from radioactive materials inside the spent fuel were calculated with the ORIGEN2.2-UPJ burn-up computer code. Coherent scattering contribution associated with Rayleigh, nuclear Thomson, and Delbrück scattering was also considered. The energy of the coherent scattering overlaps with that of NRF transitions to the ground state. Here, we propose to measure NRF transitions to the first excited state to avoid the coherent scattering contribution. Assuming that the total NRF cross-sections are in the range of 3-100 eV b at excitation energies of 2.25, 3.5, and 5 MeV, statistical uncertainties of the NRF measurement were estimated. We concluded that it is possible to assay 1% actinide content in the spent fuel with 2.2-3.2% statistical precision during 4000 s measurement time for the total integrated cross-section of 30 eV b at excitation energies of 3.5-5 MeV by using a photon beam with an intensity of 106 photons/s/eV. We also examined both the experimental and theoretical NRF cross-sections for actinide nuclei. The calculation based on the quasi-particle random phase approximation suggests the existence of strong magnetic dipole resonances at excitation energies ranging from 2 to 6 MeV with the scattering cross-sections of tens eV b around 5 MeV in 238U.