Year
2018
Abstract
The 235U mass assay of bulk uranium items, such as oxide canisters, fuel pellets, and fresh fuel assemblies, is not achievable by traditional passive gamma-ray assay techniques because of the limited penetration of the item by the characteristic 235U gamma rays. Instead, fast neutron interrogation methods such as active neutron coincidence counting must be used. For international safeguards applications, the most commonly applied active neutron systems, the Active Well Coincidence Counter (AWCC) and the Uranium Neutron Coincidence Collar (UNCL), rely on fast neutron interrogation using an Am(Li) isotopic neutron source to achieve better measurement accuracies than are possible using gamma-ray techniques for high-mass, high-density items. However, the Am(Li) sources required for the AWCC and UNCL systems are no longer manufactured, and newly produced systems rely on limited supplies of old sources salvaged from disused instruments. Lack of a suitable alternative neutron interrogation source would leave a potentially significant gap in the safeguarding of uranium processing facilities. This paper examines the potential of the D-D neutron generator as an alternative to the Am(Li) in coincidence counting through a combination of numerical simulations and measurements for both the AWCC and the UNCL. Two commercial-off-the-shelf assay systems, an AWCC and UNCL were modified to accept a commercial available neutron generator. Measurement performance was examined with the neutron generator in both steady state and pulsed neutron modes. The steady state mode allows the modified systems to be operated with the same acquisition electronics and software as the current Am(Li) measurements, operating as a near “drop in” replacement for the Am(Li) source measurement. Operation of the neutron generator in pulsed output mode provides the capability for delayed neutron counting with the potential for improved measurement precision relative to active coincidence counting. Measurement performance of the systems operated in both steady state and pulsed neutron modes is presented along with the performance from the traditional Am(Li) measurements. The relative performance of the three interrogation modes is discussed along with a discussion of what steps are needed to make D-D a robust and practical alternative.