Fission gas disposition in decay heat environment

Abstract

A computer code is developed to model fission gas disposition in UO 2 fuel during nearly isothermal heating, as would result from decay heat. The intragranular analysis, random diffusion model, uses a spatial solution for random migration and bubble coalescence. Nonequilibrium bubble growth and interactions between bubbles as well as nonequilibrium bubblegrain boundary interactions are considered. In the intergranular analysis, grain growth is allowed until tunnels have formed; this is set at 6% grain edge swelling. Grain face bubbles are assumed uniform in size and distribution. Grain edge tunnels are approximated in toroidal geometry. The model (a system of two grains, one shrinking and one growing but with total volume conserved; each grain originally contained 50% of the total fission gas) is applied to a “postulated” LMFBR accident condition involving a slow “nearly isothermal” heating of the fuel. The intragranular release is computed at 3.4% without grain growth, but at 14% with grain growth. Intragranular release is found to be dominated by grain growth. The analysis was applied also to the FGR-34 transient of HEDL. It is pointed out, however, that in the FGR-34 experiment thermal gradients were present whereas in the present code, only isothermal heating is considered. In spite of this significant difference between the modeled and the observed thermal state of the fuel, the comparison was carried out with a purpose to examine the existence of nonequilibrium attractive forces, between bubbles and grain boundaries, which were suggested by HEDL as perhaps responsible for the bubble denuding observed on both sides of the grain boundary. The computations did demonstrate the existence of nonequilibrium conditions, but the computed intragranular bubble radii, with only random diffusion as the operative mechanism, were well below the reported values. It is likely that this descrepancy between computed and observed bubble radii is due to (1) the presence in FGR-34 tests of thermal gradients, which would make bubble biased migration operative, and/or (2) the possibility of very strong enhancement, significantly more than two orders of magnitude, of the diffusion coefficient due to the prevailing nonequilibrium bubble conditions. The present code treats nonequilibrium conditions, but contains no physical mechanism for diffusion enhancement.