The neutron-gamma Feynman variance to mean approach: gamma detection and total neutron-gamma detection (theory and practice)

TL;DR

This paper derives and compares neutron-gamma variance to mean formulas for gamma-only and total neutron-gamma detection, with theoretical, simulation, and experimental validation for applications in reactor monitoring and safeguards.

Contribution

It introduces new formulas for neutron-gamma variance to mean analysis applicable to gamma-only and combined detection scenarios, validated through simulations and experiments.

Findings

Derived new variance to mean formulas for gamma and neutron-gamma detection.

Validated formulas with MCNPX simulations for different detectors.

Experimentally confirmed the applicability of formulas with various sources.

Abstract

Two versions of the neutron-gamma variance to mean (Feynman-alpha method or Feynman-Y function) formula for either gamma detection only or total neutron-gamma detection, respectively, are derived and compared in this paper. The new formulas have a particular importance for detectors of either gamma photons or detectors sensitive to both neutron and gamma radiation. If applied to a plastic or liquid scintillation detector, the total neutron-gamma detection Feynman-Y expression corresponds to a situation where no discrimination is made between neutrons and gamma particles. The gamma variance to mean formulas are useful when a detector of only gamma radiation is used or when working with a combined neutron-gamma detector at high count rates. The theoretical derivation is based on the Chapman-Kolmogorov equation with inclusion of general reactions and passage intensities for neutrons and gammas, but with the inclusion of prompt reactions only. A one energy group approximation is considered. The comparison of the two different theories is made by using reaction intensities obtained in MCNPX simulations with a simplified geometry for two scintillation detectors and a 252Cf-source enclosed in a steel container. In addition, the variance to mean ratios, neutron, gamma and total neutron-gamma, are evaluated experimentally for a weak 252Cf neutron-gamma source in a steel container, a 137Cs random gamma source and a 22Na correlated gamma source. Due to the focus being on the possibility of using neutron-gamma variance to mean theories for both reactor and safeguards applications, we limited the present study to the general analytical expressions for Feynman-Y formulas.