A Comparative Study of Nuclear Forensic Methods and Lessons Learned: The UK’s Contribution to the Galaxy Serpent Tabletop Exercise

National Nuclear Forensic Libraries (NNFLs) represent an initiative for the international community to maintain data pertaining to the inventory of nuclear substances that are handled within particular countries. NNFLs have been proposed as a means to enhance nuclear security by exploiting these databases to determine the provenance of materials that have been found outside of regulatory control. The Galaxy Serpent Tabletop exercise was organised as a way to engage the international community with the potential applications of using NNFLs as part of a forensic investigation. The exercise required teams to investigate the provenance of a nuclear fuel pin based on its isotopic composition, along with reference data from a set of nuclear reactors. The aim of this work is to highlight the different approaches that can be taken to investigate nuclear materials, as well as the potential drawbacks of each approach. The UK’s contribution to the Galaxy Serpent Virtual Tabletop Exercise is summarised and discussed. The UK team includes representatives from Atomic Weapons Establishment (AWE), National Nuclear Laboratory (NNL), and the University of Liverpool. Each institution in the UK team approached the Galaxy Serpent exercise independently with a number of different methodologies: Radiochemistry and Isotopic Correlations (AWE), Reactor Type Multivariate Analysis (NNL), and Laplacian Eigenmaps (University of Liverpool). All three approaches supported the same conclusion despite the variety of methodologies employed. A full exercise data set was not provided to the UK and the seized fuel pin was not in the data provided. It was not possible for the UK teams to conclude that the pin was inconsistent with our reference data and therefore conclusions were misleading. The conclusions uncovered during this investigation highlight the shortcomings of a nuclear forensic investigation that relies on an incomplete reference database. Particularly when the reference data that is lacking is key to a particular investigation. Therefore we will expand upon this discussion by pursuing pattern recognition methods to determine data that is complete and relevant for forensic investigation. In doing so we provide tools that will allow for the utilization of data that is appropriate and well resolved.