Martin Robel

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

The problem of identifying the provenance of unknown nuclear material in the environment by multivariate statistical analysis of its uranium and/or plutonium isotopic composition is considered. Such material can be introduced into the environment as a result of nuclear accidents, inadvertent processing losses, illegal dumping of waste, or deliberate trafficking in nuclear materials. Various combinations of reactor type and fuel composition were analyzed using Principal Components Analysis (PCA) and Partial Least Squares Discriminant Analysis (PLSDA) of the concentrations of nine U and Pu isotopes in fuel as a function of burnup. Real-world variation in the concentrations of (234)U and (236)U in the fresh (unirradiated) fuel was incorporated. The U and Pu were also analyzed separately, with results that suggest that, even after reprocessing or environmental fractionation, Pu isotopes can be used to determine both the source reactor type and the initial fuel composition with good discrimination.

Nuclear forensic analysis of uranium oxide powders interdicted in Victoria, Australia

Abstract Nuclear forensic analysis was conducted on two uranium samples confiscated during a police investigation in Victoria, Australia. The first sample, designated NSR-F-270409-1, was a depleted uranium powder of moderate purity (∼ 1000 μg/g total elemental impurities). The chemical form of the uranium was a compound similar to K2(UO2)3O4 · 4H2O. While aliquoting NSR-F-270409-1 for analysis, the body and head of a Tineid moth was discovered in the sample. The second sample, designated NSR-F-270409-2, was also a depleted uranium powder. It was of reasonably high purity (∼ 380 μg/g total elemental impurities). The chemical form of the uranium was primarily UO3 · 2H2O, with minor phases of U3O8 and UO2. While aliquoting NSR-F-270409-2 for analysis, a metal staple of unknown origin was discovered in the sample. The presence of 236U and 232U in both samples indicates that the uranium feed stocks for these samples experienced a neutron flux at some point in their history. The reactor burn-up calculated from the isotopic composition of the uranium is consistent with that of spent fuel from natural uranium (NU) fueled Pu production. These nuclear forensic conclusions allow us to categorically exclude Australia as the origin of the material and greatly reduce the number of candidate sources.

Physical characterization of uranium oxide pellets and powder applied in the Nuclear Forensics International Technical Working Group Collaborative Materials Exercise 4

Physical characterization is one of the most broad and important categories of techniques to apply in a nuclear forensic examination. Physical characterization techniques vary from simple weighing and dimensional measurements to complex sample preparation and scanning electron microscopy-electron backscatter diffraction analysis. This paper reports on the physical characterization conducted by several international laboratories participating in the fourth Collaborative Materials Exercise, organized by the Nuclear Forensics International Technical Working Group. Methods include a range of physical measurements, microscopy-based observations, and profilometry. The value of these results for addressing key investigative questions concerning two uranium dioxide pellets and a uranium dioxide powder is discussed.

A composite position independent monitor of reactor fuel irradiation using Pu, Cs, and Ba isotope ratios

When post-irradiation materials from the nuclear fuel cycle are released to the environment, certain isotopes of actinides and fission products carry signatures of irradiation history that can potentially aid a nuclear forensic investigation into the materials provenance. In this study, combinations of Pu, Cs, and Ba isotope ratios that produce position (in the reactor core) independent monitors of irradiation history in spent light water reactor fuel are identified and explored. These position independent monitors (PIMs) are modeled for various irradiation scenarios using automated depletion codes as well as ordinary differential equation solutions to approximate nuclear physics models. Experimental validation was performed using irradiated low enriched uranium oxide fuel from a light water reactor, which was sampled at 8 axial positions from a single rod. Plutonium, barium and cesium were chemically separated and isotope ratio measurements of the separated solutions were made by quadrupole and multi-collector inductively coupled mass spectrometry (Cs and Pu, respectively) and thermal ionization mass spectrometry (Ba). The effect of axial variations in neutron fluence and energy spectrum are evident in the measured isotope ratios. Two versions of a combined Pu and Cs based PIM are developed. A linear PIM model, which can be used to solve for irradiation time is found to work well for natural U fuel with <10% 240Pu and known or short cooling times. A non-linear PIM model, which cannot be solved explicitly for irradiation time without additional information, can nonetheless still group samples by irradiation history, including high burnup LEU fuel with unknown cooling time. 137Ba/138Ba is also observed to act as a position independent monitor; it is nearly single valued across the sampled fuel rod, indicating that samples sharing an irradiation history (same irradiation time and cooling time) in a reactor despite experiencing different neutron fluxes will have a common 137Ba/138Ba ratio. Modeling of this Ba PIM shows it increases monotonically with irradiation and cooling time, and a confirmatory first order analytical solution is also presented.

Sensitivity and Uncertainty Analysis of Plutonium and Cesium Isotopes in Modeling of BR3 Reactor Spent Fuel

Nuclear forensics has a broad task to characterize recovered nuclear or radiological material and interpret the results of investigation. One approach to isotopic characterization of nuclear material obtained from a reactor is to chemically separate and perform isotopic measurements on the sample and verify the results with modeling of the sample history, for example, operation of a nuclear reactor. The major actinide plutonium and fission product cesium are commonly measured signatures of the fuel history in a reactor core. This study investigates the uncertainty of the plutonium and cesium isotope ratios of a fuel rod discharged from a research pressurized water reactor when the location of the sample is not known a priori. A sensitivity analysis showed overpredicted values for the 240Pu/239Pu ratio toward the axial center of the rod and revealed a lower probability of the rod of interest (ROI) being on the periphery of the assembly. The uncertainty analysis found the relative errors due to only the rod position and boron concentration to be 17% to 36% and 7% to 15% for the 240Pu/239Pu and 137Cs/135Cs ratios, respectively. This study provides a method for uncertainty quantification of isotope concentrations due to the location of the ROI. Similar analyses can be performed to verify future chemical and isotopic analyses.

Leveraging existing information for use in a National Nuclear Forensics Library (NNFL)

A National Nuclear Forensics Library (NNFL) assists a State to assess whether nuclear material encountered out of regulatory control is of domestic or international origin. By leveraging nuclear material registries, nuclear enterprise records, and safeguards accountancy information, as well as existing domestic technical capability and subject-matter domain expertise, states can better assess the effort required for setting up an NNFL. States who are largely recipients of nuclear and radiological materials and have no internal production capabilities may create an NNFL that relies on existing information rather than carry out advanced analyses on domestic materials.