Title: Identification of TRISO-Fueled Pebble Based on Neutron Multiplicity Counting and X-Ray CT
Author: Ming Fanga, Sohaib Malika, Jiaqi Chena, Timothy Grunloha, William Dickb, Caleb Brooksa, Angela Di Fulvioa,∗
aDepartment of Nuclear, Plasma, and Radiological Engineering, University of Illinois Urbana-Champaign, Suite 100 Talbot Laboratory, MC-234, 104 South Wright Street, Urbana, 61801, IL, US
bIllinois Rocstar LLC, 108 Hessel Blvd #101, Champaign, 61820, IL, US
Abstract: The Pebble Bed Reactor (PBR) utilizing TRISO-fueled pebbles is a promising Gen-IV reactor design due to its inherent safety and thermal efficiency. Identification of pebbles discharged from the
core may provide IAEA with additional information enhancing material accountancy, help validate
software predicting pebble flow behavior and burnup, and enable optimal fuel utilization. Such identification must rely on non-destructive assay (NDA) techniques and be completed within minutes to
scale to commercial systems with hundreds of thousands of pebbles. To address this, we have integrated two NDA methods, neutron multiplicity counting (NMC) and X-ray CT, to enable fast and reliable TRISO-fueled pebble identification in commercial PBRs.
235U mass decrease during burnup offers a means to estimate the fuel burnup and narrow the search for pebble identity (ID). We have developed an optimized NMC system based on 384 boron-coated straw detectors to measure the 235U mass. The system features high gamma-ray insensitivity, with an intrinsic gamma efficiency of 10-12, surpassing the state-of-the-art 3He-based systems. The NMC can assay the 235U mass in both fresh and spent TRISO-fueled pebbles within 100 s, with an assay uncertainty below 2.5%. This uncertainty allows us to cluster the pebbles based on the burnup they experience as they flow through the core. When searching for a single pebble ID, the 235U mass estimate through NMC reduces the fraction of candidate pebbles to 16%.
The spatial distribution of TRISO fuel kernels is another unique feature associated with each individual fuel pebble and can be extracted with X-ray CT. We have developed a set of advanced algorithms, including 3D image reconstruction and segmentation algorithms to accurately extract the unique 3D distribution of TRISO particles, and a point-cloud registration algorithm capable of retrieving a pebble’s ID despite noise and arbitrary rotations. The imaging approach has been experimentally validated on mock fuel samples with attenuation properties similar to actual TRISO pebbles. Testing of the identification algorithm on a simulated dataset encompassing 100,000 pebbles demonstrated 100% accuracy across 10,000 trials, with the identification process completed in just 50 s per pebble.
By combining neutron multiplicity counting and X-ray CT, our methodology enables the identification of a single discharged TRISO-fueled pebble within an operationally feasible timeframe of 150 s in a PBR containing up to 625,000 pebbles.