No abstract

No abstract

Nuclear nonproliferation efforts rely on a variety of safeguards to protect sensitive materials in nuclear facilities. The enrichment of fresh light-water-reactor fuel assemblies is verified by several inspectorates using the uranium neutron coincidence collar (UNCL), which uses neutrons from an americium lithium ( 241 AmLi) source to interrogate the assemblies from one side. Eighteen 3 He tubes on the other three sides are used to count the coincidence neutrons from the induced fission reactions. Experiments have shown that 252 Cf could also be used to complete these measurements, providing several benefits over the use of the standard 241 AmLi source. The UNCL is one of the many instruments that will be available for training purposes in the China Center of Excellence for Nuclear Security (COE), which is located in Beijing, China. This thesis contains a detailed characterization of the response of this detector with 252 Cf as compared with 241 AmLi and an analysis of the technical basis for the use of 252 Cf in place of 241 AmLi in the Antech N2071 Neutron Coincidence Collar. This thesis (1) discusses the development a benchmarked, high-fidelity model of the UNCL using Monte Carlo N-Particle Extended (MCNPX), version 2.7.4.a; (2) fully characterizes the detection parameters, including the efficiency profile, die-away time, and deadtime parameters; and (3) demonstrates the technical basis for the replacement of 241 AmLi sources with 252 Cf sources by assessing the penetrability of neutrons from each source, evaluating the statistical uncertainty in the measurements incurred by each source, and investigating the possibility of a higher effective average number of neutrons produced per fission using 252 Cf rather than 241 AmLi. This work demonstrates the suitability of 252 Cf as a substitute for 241 AmLi and in fact shows approximately a 7.5% improvement in counting statistics over the traditional interrogation source at 4% enrichment.

Experimental results of the assay of mock-up (fresh) fuel with the differential die-away (DDA) instrument were compared to the Monte Carlo N-Particle eXtended (MCNPX) simulation results. Most principal experimental observables, the die-away time and the integral of the DDA signal in several time domains, have been found in good agreement with the MCNPX simulation results. The remaining discrepancies between the simulation and experimental results are likely due to small differences between the actual experimental setup and the simulated geometry, including uncertainty in the DT neutron generator yield. Within this report we also present a sensitivity study of the DDA instrument which is a complex and sensitive system and demonstrate to what degree it can be impacted by geometry, material composition, and electronics performance.

No abstract