Grain boundary percolation modeling of fission gas release in oxide fuels

We present a new approach to fission gas release modeling in oxide fuels based on grain boundary network percolation. The method accounts for variability in the bubble growth and coalescence rates on individual grain boundaries, and the resulting effect on macroscopic fission gas release. Two-dimensional representations of fuel pellet microstructures are considered, and the resulting gas release rates are compared with traditional 2-stage Booth models, which do not account for long-range percolation on grain boundary networks. The results show that accounting for the percolation of saturated grain boundaries can considerably reduce the predicted gas release rates, particularly when gas resolution is considered.