IAEA-CN-220-279 Alternatives to Truncated Multiplicity Analysis to Improve Verification of Uranium Fuel Cycle Materials

More accurate verification of U enrichment and mass in UF6 storage cylinders and of UO2F2 holdup contained in process equipment is needed to improve international safeguards and nuclear material accountancy at uranium enrichment plants. Small UF6 cylinders (1.5” and 5” diameter) are used to store the full range of enrichments from depleted to highly-enriched UF6. For independent verification of these materials, it is essential that the U mass and enrichment measurements do not rely on facility operator declarations. Furthermore, in order to be deployed by IAEA inspectors to detect undeclared activities (e.g. during complementary access), it is also imperative that the measurement technique is quick, portable, and sensitive to a broad range of U masses. Truncated multiplicity analysis is a technique that reduces both the value and the variance of the measured neutron count rates by only considering values 0, 1, and 2 of the multiplicity distribution. This is especially important for reducing the uncertainty in the measured doubles and triples rates in environments with a high cosmic ray background relative to the uranium signal strength. However, we believe that truncated multiplicity analysis may throw away too much useful data by truncating the distribution at 2. This paper investigates two alternative methods to truncated multiplicity analysis: 1) uses a 1σ outlier rejection limit to truncate the multiplicity distribution based on the measured data and 2) uses a second identical detector next to the detector measuring the item to directly track the cosmic ray variation. Experimental measurements of small UF6 cylinders and UO2F2 working reference materials were performed at Los Alamos National Laboratory (LANL). The data were analyzed using the alternative methods and traditional truncated multiplicity analysis to compare the impact of the methods on the measured count rates and uncertainties. The results from this analysis directly support nuclear safeguards at enrichment plants and provide a more accurate verification method for UF6 cylinders and uranium holdup in high