Computationally-generated nuclear forensic characteristics of early production reactors with an emphasis on sensitivity and uncertainty

Abstract

With nuclear technology and analysis advancements, site access restrictions, and ban on nuclear testing, computationally-generated nuclear forensic signatures are becoming more important in gaining knowledge to a reclusive country’s weapon material production capabilities. In particular, graphite-moderated reactors provide an appropriate case study for isotopics relevant in Pu production in a clandestine nuclear program due to the ease of design and low thermal output. We study the production characteristics of the X-10 reactor with a goal to develop statistically-relevant nuclear forensic signatures from early Pu production. In X-10 reactor, a flat flux gradient and low burnup produce exceptionally pure Pu as evident by the 240Pu/239Pu ratio. However, these design aspects also make determining reactor zone attribution, done with the 242Pu/240Pu ratio, uncertain. Alternatively, the same ratios produce statistically differentiable results between Manhattan Project and post-Manhattan Project reactor configurations, allowing for attribution conclusions.