A Flag-Based Algorithm for Explosives Detection in Sea-Land Cargo Containers using Active Neutron Interrogation.

Abstract

The high volume and minimal screening of sea land cargo containers presents a vulnerability in which explosive devices may be smuggled across national borders. Fast neutrons are a strong candidate for use in container screening due to their high target penetration and ability to discriminate between materials of low atomic mass, such as explosives and non metallic container contents. An algorithm has been developed that uses flags, calculated from specific measurements of the reflected neutrons and photons produced during active neutron interrogation, to discern explosives hidden in cargo containers. Steps in algorithm development included Monte Carlo simulations for scatter characterization, identification of flags in idealized scenarios, refinement of flags in realistic scenarios, combining the flags into a detection algorithm, and evaluation of the algorithm and associated detection system. Simulations compared favorably with small scale neutron scatter measurements using the explosives surrogate, melamine. The detection algorithm included corrections for different types of cargo contents and cargo inhomogeneity, surrounding environment, and realistic neutron sources and radiation detectors. The proposed algorithm has two variations, one of which can be easily implemented with today’s technology. The proposed scanning system utilizes a shielded 14.1 MeV neutron generator, eleven large liquid scintillators neutron detectors, and several inorganic scintillators for photon spectroscopy. This system should cost less than $1M to install and dose estimates fall well within acceptable levels for both operators and smuggled persons. Algorithm performance has been quantified with various explosive sizes and positions, as well as heterogeneous cargo configurations, with typical minimum detectable amounts not exceeding 200 kg.