Replacing a 252Cf Source with a Neutron Generator in a Shuffler - A Conceptual Design Performed with MCNPX

摘要

The 252Cf shuffler has been widely used in nuclear safeguards and radioactive waste management to assay fissile isotopes, such as ~(235)U or ~(239)Pu, present in a variety of samples, ranging from small cans of uranium waste to metal samples weighing several kilograms. Like other non-destructive assay instruments, the shuffler uses an interrogating neutron source to induce fissions in the sample. Although shufflers with ~(252)Cf sources have been reliably used for several decades, replacing this isotopic source with a neutron generator presents some distinct advantages. Neutron generators can be run in a continuous or pulsed mode, and may be turned off, eliminating the need for shielding and a shuffling mechanism in the shuffler. There is also essentially no dose to personnel during installation, and no reliance on the availability of ~(252)Cf. Despite these advantages, the more energetic neutrons emitted from the neutron generator (14.1 MeV for D-T generators) present some challenges for certain material types. For example when the enrichment of a uranium sample is unknown, the fission of ~(238)U is generally undesirable. Since measuring uranium is one of the main uses of a shuffler, reducing the delayed neutron contribution from 238U is desirable. Hence, the shuffler hardware must be modified to accommodate a moderator configuration near the source to tailor the interrogating spectrum in a manner which promotes sub-threshold fissions (below 1 MeV) but avoids the over-moderation of the interrogating neutrons so as to avoid self-shielding.. In this study, where there are many material and geometry combinations, we use the Monte Carlo N-Particle eXtended (MCNPX)1 transport code to model, design, and optimize the moderator configuration within the shuffler geometry. We then use the code to evaluate and compare the assay performances of both the modified shuffler and the current ~(252)Cf shuffler designs for different test samples. We investigate and quantify the matrix effect and the non-uniformity of the interrogating flux in each case. The modified geometry we propose in this study can serve as a guide in retrofitting shufflers that are already in use.