Study finds silicon in unirradiated TRISO particles

Researchers at the University of Western Ontario in Canada analyzed TRISO particles and found small regions of silicon in the unirradiated fuel, raising questions about its impact on fuel performance.

TRISO particles consist of a central fuel kernel surrounded by layers of carbon and silicon carbide. Silicon is not generally one of the intended elements in the kernel.

According to a paper published in the Journal of Nuclear Materials, the team obtained as-fabricated uranium oxycarbide (UCO) kernels and fully coated UCO TRISO particles. While characterizing the particles' chemistry, the team found silicon inside the kernels.

"At first, we saw the silicon and thought, 'okay, that's interesting. It's a little weird that it's there, but let's keep going.' We had other things we wanted to understand about this material. But then we started asking where that silicon came from, and that's when the investigation really began," said Gateman, a chemistry professor at Western and one of the authors of the study.

Where the silicon comes from: Silicon in TRISO fuel is not an unprecedented finding. An Oak Ridge National Laboratory study of TRISO particles from the Advanced Gas Reactor (AGR) Fuel Development and Qualification program found silicon-rich regions inside irradiated TRISO particles and posited that the silicon had migrated into the kernel from the outer silicon carbide shell.

But Gateman and her team were using backscattered electron microscopy to look at particles that had not been irradiated.

"We found silicon in the fuel kernel before it had ever been inside a reactor. That means when we see silicon inside irradiated fuel, we can't automatically assume it migrated there from the silicon-carbide coating," said Gateman.

The researchers looked at both bare kernels and fully coated particles and found silicon was present at both stages of fabrication, indicating its presence was not a consequence of the coating process.

Some TRISO fuel kernels are produced using a silicon-based fluid to form the fuel kernels, and the paper suggests that silicon could be introduced into the kernel during that process.

According to Paul Demkowicz, nuclear science and technology directorate fellow at Idaho National Laboratory and the TRISO fuel qualification lead for the Department of Energy's Advanced Fuels Campaign, silicon could be incorporated into the kernel during formation and remain if inadequate kernel washing techniques are used. However, for the AGR kernels, trichloroethylene rather than a silicon-based fluid was used as the forming fluid, so the fabrication pathway would not be expected to introduce silicon.

"The authors of this paper not mentioning the possibilities of different fabrication pathways for different fuel is a major omission, particularly because of the connection they make between the fuel they looked at and the AGR fuel," he said.

He said another possible pathway for the early introduction of silicon is if it's present in the uranium feedstock. The AGR kernels were fabricated at BWX Technologies and therefore cannot necessarily be directly compared with the kernels studied by Gateman, which were fabricated by Standard Nuclear.

Demkowicz said that during preirradiation analysis, the AGR team is not specifically looking for silicon but that they analyze samples using backscattered electron microscopy and have not observed silicon phases.

What's the effect: Gateman's study indicates that at least some unirradiated TRISO kernels contain silicon that was not intended to be there, raising questions about its potential impact.

"We don't know yet what effect these silicon-rich phases have on fuel performance, but there are some interesting possibilities," said Gateman. "The first step is recognizing they're there and making sure we account for them when we study how the fuel behaves."

The paper states that silicon carbide reacting with metallic fission products could "act as preferential degradation sites during reactor operation" which "may lead to the formation of voids or pores within the kernel, potentially compromising its mechanical integrity." It also hypothesizes that silicon impurities could "serve as preferential nucleation sites for fission gas bubble formation."

Demkowicz agreed this is an unanswered question but believes there is not a strong cause for concern. He said the AGR program team has "lots of data on the radiation of UCO fuel and kernel corrosion isn't identified as a performance-limiting phenomenon." He also said that pore formation is "extremely common," and even if silicon is promoting that process, he doesn't see that as a significant concern.

Demkowicz said extensive data from the AGR series of TRISO irradiation tests have shown "really good performance" of the fuel.

"Whether [silicon] was there before or not, or came in afterwards, or during the process of radiation and then heating, it doesn't change the fact that on aggregate, particle failure rates are low," he said.