Monte Carlo simulation for calculation of kinetic parameters in an Accelerator Driven Subcritical TRIGA reactor

In this study, noise analysis techniques including Feynman-alpha (variance-to-mean) and Rossi-alpha (correlation) and dynamic method such as slope fit method have been used to calculate effective delayed neutron fraction (beta(eff)) and neutron reproduction time (Lambda) in Accelerator Driven Subcritical TRIGA reactor. The obtained results have been compared with MCNPX code results. The relative difference between MCNPX code with Feynman-alpha and Rossi-alpha techniques and slope fit method for beta(eff) are approximately -5.4%, 1.2%, and -10.6%, -14.8%, respectively, and also for A is approximately 2.1%. According to results, the noise methods can been considered ideal for detection with high efficiency and zero dead time and in the slope fit method, the decay of the delayed neutrons has been neglected and only the prompt neutrons have been taken into account. In addition, quantities simulated in the current study are validated against both the reference data and the results of MCNPX code. Therefore, the purpose of this study is to simulate the commonly used experimental methods by MCNPX code and investigate the convergence as well as accuracy of the computational results for different analysis methods in calculation of the kinetic parameters in an Accelerator Driven Subcritical TRIGA reactor. (C) 2015 Elsevier Ltd. All rights reserved.