A Theoretical Model for Fission Gas Release from UCO TRISO Fuel

This paper describes a theoretical model for fission-gas release from TRISO-coated particle fuel with UCO kernels. Model parameters are derived from data obtained from legacy irradiation capsule HFR-B1. The model correlation is expressed as the ratio of release rate to birth rate (R/B), which is numerically equivalent to fractional release under steady-state conditions. Compared to legacy models, the theoretical basis for this model allows for more accurate predictions of fission-gas release over a wider range of temperature, isotope half-life, and fission rate density. The model is derived from an overall mass balance for gaseous fission products, and includes components for low-temperature release, high-temperature release, and transport through fuel compact and/or graphite structures to a reactor coolant or capsule sweep gas. For low-temperature release from small-diameter kernels, the dominate mechanism is direct recoil. The high-temperature release component is derived by considering a diffusion mechanism with trapping into closed porosity. The trapped component is not permanently trapped and can be released by two mechanisms: (1) a direct, rapid release by a thermally activated re-solution mechanism that opens previously closed porosity as temperatures are increased (thermal re-solution) and (2) a slower release mechanism that returns trapped inventory for release by diffusion (fission-induced re-solution). Based on existing theory, the trapping and fission-induced re-solution rates are assumed to be proportional to the fission-rate density. Because the trapped inventory is dominated by longer-lived isotopes, this feature allows the model to more accurately capture the different release behaviors of shorter-lived vs. longer-lived isotopes. Transport to the coolant or sweep gas is modeled as a diffusion process, accounting for the tortuous transport path of the porous media (e.g., fuel-compact matrix and graphite) that surround the exposed fuel kernels. Data obtained from the U.S. Department of Energy (DOE) Advanced Gas Reactor (AGR) TRISO Fuel Development and Qualification program provide an opportunity to update the model parameters and improve the predictive accuracy of the model.