Multi-Mode Imaging for TRISO-fueled Pebble Identification

A common nuclear reactor design uses fuel in the form of billiard ball sized pebbles which flow through the core continuously and can be reinserted into the core until target burnup is reached. Understanding how much radiation and damage a pebble accrues as it passes through the core, as well as the reactivity effects of reinserting an irradiated pebble into the core, are critical to designing safe and economical fuel handling protocols. This requires tracking individual pebbles as they move through the core and adjusting for damage to surfaces. Pebbles can be identified by the unique microstruc- ture of their fuel matrices through neutron multiplicity counting, which is identified upon loading and tracked thereafter. During reactor operation, the pebbles are scanned and identified using a deep learning identification algorithm, so that the radiation and damage accrued by that individual pebble can be tracked. In addition to experimental methods, detailed pebble flow will be simulated using open source software, allowing each pebble’s trajectory to be calculated. Surface interactions can be computed dynamically and stored, providing abrasion characteristics for each pebble. A method for scanning and identifying pebbles based on the unique microstructure of their fuel matrix will be developed using sophisticated active interrogation techniques. An open source physics-based code will be used to simulate the motion of pebbles through the reactor and calculate forces on the pebbles. Data from both approaches will be used to provide detailed information about pebbles such as their burnup state and material integrity. Software will be developed to be licensed to prospective vendors who are developing pebble bed reactor concepts. The experimental method will be marketed to Federal entities, vendors, and regulators interested in better characterizing these fuel types.