Development of an efficient self-optimized Monte Carlo code for neutron porosity logs

Abstract

A novel Monte Carlo code has been developed to stimulate the response of neutron porosity tools. The code features a single variance reduction scheme based on importance sampling. Explicit importance functions are used to continuously determine the particle importance as a function of position and energy. The parameters required for these functions are estimated in a negligible amount of computer time from an analog Monte Carlo simulation prior to the actual simulation. This technique eliminates the difficult task of dividing the medium into cells and does not require setting up an importance map. The code takes into consideration anisotropic elastic and inelastic scattering and thermal neutron scattering with bound nuclei. The counting rate ratio is obtained within a 5% statistical uncertainty in 3 hours of FPS-64 computer time. Benchmark results and comparisons with test-pit data are presented to illustrate the validity and computer efficiency of the proposed code. >