Year
2009
Abstract
International Atomic Energy Agency (IAEA) inspectors currently perform periodic inspections at uranium enrichment plants to verify UF6 cylinder enrichment declarations. Measurements are typically performed with handheld high-resolution sensors on a sampling of cylinders taken to be representative of the facility’s entire product-cylinder inventory. As additional enrichment plants come online to support the expansion of nuclear power, reducing person-days of inspection time will take on greater importance. Pacific Northwest National Laboratory (PNNL) is developing a concept to automate the verification of enrichment plant cylinders to enable 100% product-cylinder verification and potentially, mass-balance calculations on the facility as a whole (by also measuring feed and tails cylinders). The Automated Cylinder Enrichment Verification System (ACEVS) could be located at key measurement points to positively identify each cylinder, measure its mass and enrichment, store the collected data in a secure database, and maintain continuity of knowledge on measured cylinders until IAEA inspector arrival. Given the potential for reduced inspector presence, the operational and manpower-reduction benefits of the portal concept are significant. The key technical question is whether the cylinder portal concept can meet, or potentially improve upon, the enrichment verification performance of today’s technologies. PNNL’s ACEVS concept utilizes sensors that can be operated in an unattended mode: moderated He-3 neutron detectors and large NaI(Tl) scintillators for gamma-ray spectroscopy. While the medium-resolution NaI(Tl) scintillators are a sacrifice in energy resolution, they do provide improved collection efficiency for signatures above 1 MeV. When compared to today’s enrichment meter technologies that utilize only the weakly penetrating 185-keV line, the He-3/NaI sensor combination allows the exploitation of additional, more-penetrating signatures: neutrons produced from F-19(a,n) reactions (spawned primarily from U-234 alpha emission) and high-energy gamma rays (extending up to 8 MeV) induced by neutrons interacting in the steel cylinder. These indirect measures of U-235 require a relatively stable U-234/U-235 ratio in the product material in order to be useful for enrichment verification purposes. The hypothesis in ACEVS development is that the U-234/U-235 ratio is sufficiently constant at the specific facility where the automated system is installed to utilize neutron-related signatures as an indirect measure of U-235 enrichment. Further, these highly penetrating signatures can be combined with a modified form of “traditional” 185-keV enrichment measurements to meet target uncertainties for the verification of product cylinders. If successful, this new enrichment assay approach would provide two important benefits over today’s handheld enrichment-meter method: 1) automation to enable 100% product-cylinder verification while reducing manpower requirements, and 2) assay of the entire cylinder volume rather than a small fraction of that volume. This paper focuses on the “non-traditional” enrichment assay aspects of the ACEVS concept: neutron and high-energy gamma-ray signatures, the radiation sensors designed to collect those signatures, and proof-of- principle cylinder measurements and analysis. Preliminary analysis indicates that an automated cylinder verification approach has the potential to meet target uncertainty values for 30B product cylinders (5%), assuming ore-based enrichment feed and a facility-specific calibration. Also described is the additional work needed to more definitively assess the concept’s viability, most notably an improved understanding of the U- 234/U-235 ratio variability in modern enrichment plants.