Convergence of Monte Carlo methods for neutron noise

The neutron noise δφ describes the small variations of the neutron flux around the stationary state φ0, and is typically due to vibrations or oscillations of the core components, induced by fluid-structure interactions and other generally unwanted phenomena. Knowledge of δφ is useful for core monitoring: for this purpose, in recent years several new computational methods have been proposed in order to solve the neutron noise equations with state-of-theart deterministic or Monte Carlo solvers. In this paper we present a preliminary investigation of the convergence properties of Monte Carlo methods for neutron noise analysis, in view of improving their reliability. We carry out our investigation on an infinite-medium benchmark configuration: despite the introduced simplifications, our findings suggest some universal features that represent a stepping stone towards a broader theoretical framework.