Simulation studies on a stilbene-based fast-neutron uranium collar for non-destructive assay of fresh fuel assemblies

Year
2019
Author(s)
Shaun D. Clarke - University of Michigan, Ann Arbor
Sara A. Pozzi - University of Michigan
Tony H. Shin - University of Michigan
Angela Di Fulvio - University of Illinois at Urbana-Champaign
Abstract
The University of Michigan (UM) has investigated a stilbene-based fast-neutron uranium collar (FNCL) system for non-destructive assay (NDA) of fresh fuel assemblies as part of the Neutron Rodeo Phase II (NR II) collaborative project. The NR II project aimed at comparing different types of organic scintillators to investigate potential improvements in the speed, accuracy, and reliability of quantifying the 235U linear density of fresh fuel assemblies. MCNPX-PoliMi in conjunction with the MPPost detector response code was used to model a FNCL system consisting of 30 - 5.08 Ø cm x 5.08 cm cylindrical trans-</i>stilbene detectors, arranged in two banks with 15 detectors each. The system uses two AmLi interrogative sources (5 x 104n/s) to induce fissions within each fresh fuel assembly. The simulated fresh fuel assemblies include a set of 17 x 17 calibration assemblies that was used to produce a calibration curve. This calibration curve was then used to quantify the 235U linear density of assemblies that involved burnable poisons, partial removal of fuel pins, and different geometries. Various low-energy detection thresholds were applied for the simulated NDA to investigate any advantages in observing neutrons of different energy groups. UM has conducted a series of dedicated detector characterization experiments in order to provide measured parameters for implementation in the MPPost detector response algorithm. A prototype stilbene-based FNCL has been procured at UM consisting of two banks of 12 detectors each. Simulated detector response for a single and an array of stilbene detectors was validated with measured data, showing an agreement of approximately 3% integral count rate for a single detector, and less than 8% difference in the singles and doubles count rate for an array of 24 detectors. This paper reports the validation of the simulated results by experimental data. This direct comparison shows the accuracy of the measured stilbene physical parameters and the high fidelity of the detector response code. We will then detail the full simulated results of the FNCL for NDA of fresh fuel assemblies. This work was sponsored by the U.S. Department of Energy, National Nuclear Security Administration, Office of Nonproliferation and Arms Control.