Statistical analyses of granular flows in pebble bed reactors

Abstract

Fuel costs are a major component of pebble bed reactor (PBR) operating expenses, and uncertainty in pebble trajectories can lead to under- or over-irradiation, reducing fuel utilization and potentially degrading performance. More accurate pebble-path prediction can therefore improve fuel-management economics, particularly during load-following operation. Simulating granular flow in PBRs is a challenging task because it is affected by material properties and contact mechanics, which exhibit a large degree of time variability because of operation and fuel position. These uncertainties propagate directly into predictions of pebble trajectories and residence times, which are central quantities for fuel management. This paper reports the development and validation of Pebble TRACK : a discrete element method (DEM)-informed statistical framework for rapid pebble trajectory assessment. The Pebble TRACK approach uses mesh-free DEM simulations to generate high-resolution trajectory and contact-interaction data, including dynamic surface interactions. Pebble TRACK -predicted trajectories and residence times agree well with experimental data, requiring approximately 5.52% of the computational time as compared to a full DEM simulation. While developed for PBR applications, Pebble TRACK can be extended to produce granular flow predictions for an arbitrary geometry. These components and performances provide a computationally efficient path to predicting real-time pebble trajectories suitable for advanced fuel management strategies that will be combined with the neutron flux history to help determine a pebble’s burnup in the future.