Discrimination of source reactor type by multivariate statistical analysis of uranium and plutonium isotopic concentrations in unknown irradiated nuclear fuel material

Nuclear Forensic Science: Correlating Measurable Material Parameters to the History of Nuclear Material

“…Robel and Kristo also applied multivariate statistics for identifying the source of reactor produced plutonium using a combination of PCA and partial least-squares discriminant analysis (PLSDA). They conclude that all reactor types and fuel combinations are well resolved starting very early in the core life.…”

“…In combination with other parameters, such as the pellet geometry or the presence of neutron poisons (e.g., Er, Gd), a useful nuclear forensic signature is obtained. Because of the large variability of uranium isotopic compositions in fresh fuel and the spectrum of different reactor types, the isotopic composition of uranium after irradiation is of limited use in nuclear forensics as determined by Robel and Kristo …”

Fingerprinting commercial nuclear reactor forensics using the ORIGEN-S simulation code

“…In the most recent revision of the Origen-S code, a substantial amount of work went into benchmarking this code, specifically the crosssection libraries, which improved the computational tool's accuracy. Robel 7 used the updated version of ORIGEN-S, which enabled realistic variation in fresh-fuel composition, as well as classification of actual unknown spent fuels using Partial Least Square Discriminant Analysis (PLSDA). The goal of this study was to illustrate the distinguishing factors among nuclear reactors of the same type by investigating the differences between their unique uranium and plutonium isotopic ratios.…”

Plutonium fingerprinting in nuclear forensics of spent nuclear fuel

“…If the smuggled material is spent fuel, its origin depends on its charge composition, the type of the reactor where it was charged and its final burnup. Fingerprinting methods have been demonstrated to resolve spent fuels of different origin in a 3D space (Nicolaou, 2006(Nicolaou, & 2008Robel and Kristo, 2008;Jones et al, 2014). They have combined the U, Pu and fission product isotopes present in simulated or actual spent nuclear fuels and multivariate statistical/pattern recognition techniques.…”