Tritium thermal desorption testing of nuclear graphites irradiated at fluoride-salt-cooled high-temperature reactor conditions

Abstract

The Fluoride-Salt-Cooled High-Temperature Reactor (FHR) is a next-generation nuclear plant design that combines successfully demonstrated technologies from other advanced reactor concepts such as tristructural isotropic (TRISO) coated particle fuel, molten flibe salt (LiF-BeF2) coolant, and an Air-Brayton power cycle. A prominent technical challenge for the FHR is maintaining the release of tritium generated from neutron irradiation of flibe below acceptable levels. One proposed method for partitioning tritium from the salt is through adsorption onto graphite. Demonstrating the viability of this type of tritium control system requires further experimental investigation since few studies have examined the combined effect of molten flibe, tritium, and graphite at relevant FHR temperatures. Studying tritium transport experimentally has been recently enabled by three in-core fluoride salt irradiations completed at the Massachusetts Institute of Technology Nuclear Reactor Laboratory.

Tritium release measured from the third irradiation (FS-3) displayed a low initial release rate followed by a roughly 15 day increase towards equilibrium. A difference of 23% was observed between the calculated tritium generation and the total release measured from the experiment. The varying release rate from the capsule and the difference between generation and release was explained by tritium adsorption in the capsule graphite. In this work, a thermal desorption furnace was used to heat graphite samples irradiated in flibe salt in order to release adsorbed tritium. Desorption verses temperature was monitored with a compensated ion chamber while total tritium released was determined from liquid scintillation counting. The primary analysis focused on six subsections of graphite from the second irradiation (FS-2); three from a disc of IG-110U and three from ARB matrix graphite. Most of the desorption occurred in a large release peak centered at 700 °C.

Because the first peak was seen to increase for samples that had a higher salt-facing surface area per unit mass, the main tritium trapping mechanism was assumed to be surface chemisorption. Additionally, the spread in tritium release values from each sample decreased when measured tritium was normalized by surface area instead of sample mass. Therefore, adsorption of tritium was hypothesized to be dominated by surface effects. Measurements of the six samples resulted in 1.62±0.28 [mu]Ci/mm2 of tritium captured by IG-110U and 1.18±0.23 [mu]Ci/mm2 for ARB. The difference between calculated generation of tritium and total tritium release measured from the FS-3 experiment can be explained by adsorption onto graphite.