Assessment of fission gas release mechanisms in LWR fuel based on OECD/NEA-P2M benchmark

Fission Gas Release (FGR) is one of the most critical phenomena in nuclear fuel performance, directly influencing fuel rod swelling, internal pressure, thermal conductivity, and ultimately reactor safety margins. An accurate understanding of the mechanisms that govern FGR, as well as the ability to model its evolution under both steady-state and transient conditions, is essential for predicting the safe operational limits of nuclear fuel. Within this context, the OECD/NEA “Power to Melt and Maneuverability” (P2M) benchmark provides a unique framework to evaluate the predictive capability of state-of-the-art fuel performance codes against historical bump test data and to prepare for future irradiation campaigns. This work, carried out as part of a Master’s Thesis internship at SCK CEN, contributes to the P2M benchmark by focusing on the modeling of FGR in four representative experiments: AN3, AN10, and AN4 (added as extra work), as well as P2M-Q1 (not included here due to confidentiality). The ANx experiments were performed in the DR3 reactor at Risø and they are included into the OECD/NEA IFPE database. The activity is carried out with the TRANSURANUS v.2023 fuel performance code, detailed input decks were developed for both base irradiation histories and subsequent bump tests using the data retrieved from benchmark specifications. A systematic modeling strategy was applied, including careful treatment of axial nodalization to be consistent with the nodalization developed by CEA, plenum conditions, and the activation of different mechanistic options for FGR. The simulations demonstrated that TRANSURANUS is capable of reproducing the general trends observed experimentally, with a satisfactory description of base irradiation conditions and a qualitatively correct representation of FGR during bump tests. However, discrepancies remain, particularly regarding the magnitude of burst releases and final release fractions under low-power conditions. The underestimation of these effects suggests that additional refinement of mechanistic models and calibration of input parameters are necessary to enhance predictive accuracy. The sensitivity analysis highlighted the central role of burnup and maximum linear heat rate, which influence both diffusional release and burst phenomena as well. Overall, this work illustrates the value of the P2M benchmark in identifying model strengths and limitations, fostering a feedback loop between experimental data and code development. The results not only reinforce the importance of modeling for safety assessments but also provide input to ongoing international efforts to validate and improve fuel performance simulations. In this way, the internship contributes both to the scientific understanding of FGR and to the collective preparation of the nuclear research community for the upcoming P2M-Q1 irradiation in BR2.