Investigation of Failed TRISO Fuel Assay Using Gamma-Ray Spectrometry

Abstract

TRISO microsphere fuel is the fundamental fuel unit for Very High Temperature Reactors (VHTR). A single TRISO particle consists of an inner kernel of Uranium Oxycarbide surrounded by layers of pyrolytic carbon and silicon carbide. The silicon carbide serves as the primary barrier to the release of fission products into the core. If the silicon carbide layer fails, fission gas, especially Kr and Xe, will begin to escape the failed particle.

In order to understand the behavior of TRISO fuel under in-core conditions, a series of experiments is being conducted by Idaho National Lab at the Advanced Test Reactor. AGR-1 is the first of these experiments. It will measure fission product release due to failed TRISO particles. Simulations of this experiment have been conducted at North Carolina State University to develop a method for the analysis of the results of the experiment.

The ATR core was simulated using the Monte Carlo code MCNP to calculate the expected neutron energy spectrum for the AGR-1 experimental test train. This spectrum was used to create one-group cross sections for implementation in ORIGEN calculations of the amount of activity produced in the experiment.

Several theoretical models have been developed to describe the phenomenon of gas release. While each model is based on similar physics, different models contain unique features that distinguish them from one another. These Release to Birth (R⁄B) models are developed and applied to the activity found in the ORIGEN calculations to create expected release activities. The release activity is used to create gamma-ray spectra that are representative of the different R⁄B models.

Expected R⁄B due to a model can be calculated for comparison to the experiment with knowledge of the number of failed particles in the spectra. The comparison of measured to predicted R⁄B ratios gives insight into the physics of release and also helps validate specific models. Direct comparison is possible, but many of the uncertainties associated with direct comparison are nullified through the use of relative indicators.

Each R⁄B model has a unique set of indicators that reflect the physical processes simulated in the model. Trends in the model indicators can be matched up with trends in indicators derived from the release spectra to validate either an entire model or validate the need to consider certain parameters in the creation of a complete and successful release to birth model.

Gamma spectrometry is a useful tool for the understanding of fission gas release from failed TRISO particles. A better understanding of the processes that influence fission gas release will influence the fuel manufacturing and quality assurance protocols during the continued development of the VHTR. Future work in this area includes experiment in which the conditions can be better controlled to document the effects of temperature and fission rate in the fuel.