Year
2007
Abstract
Verification of fresh fuel pellets during Physical Inventory Verification (PIV) requires partial defect detection capability to be achieved. Non-Destructive Assay (NDA) measurements performed to reach that goal traditionally combine 235U enrichment verification with fuel pellet weighing. The 235U enrichment determination is usually performed by use of low-resolution ambient temperature gamma detectors and calibrations based on infinitely thick reference standards. Although this methodology is proven, it does require the use of small opening collimators to meet the infinite thickness criteria when measuring objects as small as fuel pellets. This significantly reduces the achievable count rate and may hamper the verification. Additionally, calibration accuracy by this methodology when combined with a 2 Region of Interest (ROI) determination of the uranium count rate is sensitive to ambient background levels. Consequently, the possibility to rely on an alternative approach based on the absolute calibration of the detector by numerical techniques was investigated. A General Numerical Object Modelling for Enrichment (GNOME) software was developed to generate automatically Monte Carlo N-Particle based calibrations for a specific detection system and fuel pellet characteristics. The code is also intended to support for each measurement systems the benchmark of the numerical modelling against empirical data acquired with reference material and the comparison of the measured and declared enrichments. This paper covers in detail the methodology used to derive the fresh fuel pellet enrichment level as well as the software architecture used to provide a deployable solution to support IAEA inspectors’ work in the field.