Nondestructive gamma-ray spectrometry of nuclear fuel is routinely performed in axial gamma scanning devices and more recently with gamma emission tomography. Following the irradiation of a fresh nuclear fuel with high intensity neutron flux in a nuclear reactor core, a great number of gamma-emitting radionuclides are created. These can be utilized for gamma spectrometric techniques. However, due to the high density and atomic number of the nuclear fuel, self-attenuation of gamma-rays is a challenge, which requires attenuation correction in order to perform accurate analysis of the source activity in the fuel.

In this study, the degradation of the gamma-ray mass attenuation with burnup was investigated and, in addition, a predictive model was created by investigating the attenuation change at various gamma energies caused by the burnup of the nuclear fuel. This model is intended for use by spectrometry practitioners inspecting nuclear fuel. To this aim, the energy-dependent gamma-ray mass-attenuation coefficients were investigated as a function of burnup for UOX, and three MOX fuels having different initial Pu contents. The Serpent 2 reactor physics code was used to simulate the burnup history of the fuel pins. The nuclide inventory of the Serpent 2 output is combined with the NIST XCOM database to calculate the mass attenuation coefficients.

The mass attenuation coefficient of the fuel was found to decrease with the fuel burnup, in the range of a few percent, depending on the burnup and gamma energy. Also, a theoretical burnup dependent swelling model was imposed on fuel density to see how linear attenuation coefficient of fuel material is changed. Furthermore, greater effect may be expected on the transmitted intensity, where a simulation study of a PWR assembly revealed that the contribution from the inner rods in a scanned fuel assembly increased by tens of percent compared to the one with non-irradiated fresh fuels, when shielded by the outer rods of the assembly. A sensitivity analysis was performed in order to test the effect of a number of geometrical and operational reactor parameters that were considered to potentially effect the mass attenuation coefficient. Finally, a simple-to-use predictive model was constructed providing the mass-attenuation coefficient [cm²/g] of fuel as a function of burnup [MWd/kg_HM] and initial Pu content [wt%]. The resulting predictive model was optimized by using the nonlinear regression method.