Analysis of time correlation measurements with the Active Well Coincidence Counter

Active well coincidence counters are widely used for nondestructive assay applications in nuclear safeguards and nuclear waste characterization. The method is based on the detection of correlated neutrons from fission by He-3 detectors embedded in a polyethylene moderator. In the assay of uranium, an active measurement must be performed to induce fission in the material, and typically Am/Li neutron sources are used as the active source. Monte Carlo studies of the measurement setup are useful in the design, optimization, and analysis of the entire measurement system. The simulation must take into account many factors, for example the Am/Li neutron spectrum, the multiplicity of neutron emission in induced fission events, and the detection of thermalized neutrons by the He-3 counters. In this study, we address these issues and present a detailed analysis of the measurement system that includes parameters such as the length of fission chains generated in the fissile material by the source neutrons, the time of neutron detection in the He-3 counters, and the generation number of the detected neutrons. The simulations are performed with the MCNP-PoliMi code. The simulation results are compared with measurements performed on uranium oxide standards with an active well coincidence counter that is in use at the Y-12 National Security Complex. The geometry of the MCNP-PoliMi simulation for the active well coincidence counter is shown. In addition to the simulation of traditional multiplicity parameters given by the shift register (singles, doubles and triples), MCNP-PoliMi allows the user to simulate the entire distribution of time correlations between detectors and detector autocorrelations. We show that this approach, also known as time interval analysis and first proposed by Bruggeman and colleagues in 1996, has the potential to lead to a more robust and complete analysis compared to the measurement of multiplicity alone. The results of this study serve as a preliminary validation of the MCNP-PoliMi code for the simulation of measurements performed with active well coincidence counters