Neutron Noise Analysis Techniques for Improved Verification of Critical and Subcritical
Cores

Safeguards verifications at research reactors and (sub)critical assemblies are often challenged by the limited access to or complete inaccessibility of in-core material. Additional difficulties are related to the significant impact of reactor design and its operation history on the emitted radiation. In recent years, the IAEA has expanded a standard toolkit of its in-core material verification techniques, which were traditionally based on total neutron counting, towards application of advanced approaches involving detection of correlated neutrons and Monte Carlo modelling. These new verification techniques improved the confidence level for verification of direct use nuclear materials, while reducing the burden on the facility operator. In this paper we present further advancements of the verification techniques, with particular emphasis on the Feynman- and power spectral density (PSD) neutron noise analyses and their application in quantitative verification of the fast highly enriched uranium core of the TAPIRO research reactor (ENEA, Casaccia Research Centre, Italy). In the course of this research and development effort, multiple neutron noise measurements were performed at different reactor criticalities (ranging from deep subcritical to supercritical states) and reactor power levels. The data were acquired using a pulse digitizer capable of recording long waveforms from a shielded 1 in.  1 in. stilbene detector positioned near the reactor core. To accurately process and interpret the data thus obtained, a new Monte Carlo-based approach involving a multi-group representation of the prompt neutron generation time distribution was developed and utilized. The multi-group approach overcomes shortcomings of the classical single-group point-kinetics model observed when applied to fast reflected cores, thereby providing more accurate estimates of the correlation magnitude (Y) and temporal scale () of the neutron multiplication process. A selected combination of measurable characteristics, including the  value and a newly introduced -parameter, provided a robust approach for the verification of both reactor design and in-core fissile mass. The -parameter is independent of the reactor criticality and power levels, yet highly sensitive to removal of small quantities of nuclear material from the core and modifications of the reactor and core designs.