Nuclear Forensics Methodology for Reactor-Type Attribution of Chemically Separated Plutonium

Abstract

A nuclear forensics methodology has been developed that is capable of source attribution of separated weapons-grade plutonium in case of an interdiction. The methodology utilizes plutonium and contaminant fission product isotopes within the separated plutonium sample to determine the characteristics (reactor parameters) of the interdicted material. The reactor parameters of interest include source reactor type, fuel irradiation burnup, and time since irradiation. The MCNPX-2.7 radiation transport code was used to model reactor cores and perform neutronics simulations to estimate the resulting isotopes of irradiated UO2 fuel. The simulation results were used to create a reactor-dependent library of irradiated fuel isotope ratio values as a function of fuel burnup and time since irradiation. Ratios of intra-element isotopes (fission product or actinide) are used as characteristics to determine a combination of reactor parameters of interest that could have produced the interdicted sample. The isotopes selected for the attribution methodology development were based upon the initial criteria of isotope production yield in fuel and half-life. Subsequently, intra-element isotope ratios were formed with the criterion that the ratio must have a functional dependence on at least one of the reactor parameters of interest. The developed methodology compares the values of reactor-dependent intra-element isotope ratios in the library developed to the same ratios of the interdicted sample. A maximum likelihood calculation methodology was utilized to perform the aforementioned multiple intra-element isotope ratio comparison to produce a single metric to depict the result of the comparison. The methodology can predict the reactor type, fuel burnup, and time since irradiation of the sample by selecting the array of reactor-dependent intra-element isotope ratios that provides the maximum likelihood value. The methodology was tested with intra-element ratios of pseudo interdicted sample data and found to be viable for source attribution.