Benchmarking and Simulation Efforts in Nuclear Safeguards

It is almost impossible to imagine that a new neutron detector for safeguards would be designed without the use of Monte Carlo simulation. The simulation allows the performance of an instrument to be predicted without ever picking up a piece of hardware. In addition it is possible to determine the response of the instrument to extreme items, which cannot be produced in the laboratory. This was not always the case: a large fraction of safeguards detectors that are currently in use were designed using a combination of experiments and physics understanding and intuition, before Monte Carlo methods were applied to safeguards instrumentation. By the time the safeguards NDA ‘bible”, PANDA, was first published in 1991 there were a number of examples of Monte Carlo methods being used for understanding neutron behavior in moderators and shielding. From the 1990s onwards there are numerous examples of the design of safeguards instrumentation using Monte Carlo techniques, one driving force being the increased computing and storage capabilities of desktop machines. Los Alamos National Laboratory has implemented many neutron measurement instruments for safeguards in facilities around the world. The typical lifecycle of such an instrument begins with a statement of the problem and potential approaches are simulated and refined with Monte Carlo calculations until an acceptable solution is found. After the instrument is built an experimental measurement is made (very often with 252Cf) and compared with the simulation results to benchmark the model. Agreement at this stage gives confidence in the model and allows the model to be used to predict the performance with nuclear material items. Nuclear data quality can be an important factor in the quality of the results. Simulation is also used in other applications, such as converting neutron counting rates from large area monitoring into nuclear material masses. This paper describes the use of Monte Carlo for safeguards as well as some recent examples of several kinds of applications.