Quantification of TRISO fuel heterogeneity effects in HTGR lattice physics calculations

A large number of LEU-MHR fuel compact models were generated with randomly distributed TRISO particle fuel and were simulated using MCNP5, and it was determined how several neutronic parameters, including k-infinite, the thermal and fast diffusion coefficients, and the four factors, varied across the randomly-generated cases. A sensitivity study was also performed to determine how the four factors depend on the definition of the thermal energy group. Values of k-infinite for the cases had a sample standard deviation of 248 pcm and were found to follow an approximately normal distribution about the mean value of k-infinite. Although all of the four factors were found to have similar sample standard deviations, the resonance escape probability was found to be the most variable parameter with a sample relative standard deviation between 0.07% and 0.08%. HTGR fuel compact homogenization methods typically examine only one reference fuel compact that contains a uniform distribution of TRISO particles, but in reality the TRISO particles are randomly distributed throughout the fuel compact. Thus, the neutronic parameters for actual fuel compacts differ randomly from those in the reference model. To license next-generation High-Temperature Gas Reactors engineers must quantify all uncertainties of the design and this random variation in neutron parameters is a previously unmeasured quantity; this study measures this uncertainty by examining the variation in k-infinite for HTGR fuel compact models with randomly distributed TRISO fuel. (authors)