engleski

I2S-LWR Pressure Vessel Fast Fluence Calculations

The I2S-LWR (Integral Inherently Safe Light Water Reactor) is a high-power 1000 MWe reactor with improved inherent (passive) safety features. This new reactor concept, led by Georgia Institute of Technology, is based on the integral primary circuit configuration, a new type of the fuel/cladding system, and a novel steam generation system. This paper presents shielding studies of the I2S-LWR reactor model using SCALE6.1 code package to identify the fast neutron fluence rate distribution. The reactor pressure vessel (RPV) fast fluence calculation with uniform fission source showed that the results are conservative and below proposed upper limit. The SCALE6.1/MAVRIC shielding sequence was used to optimize neutron fluence results in the complete RPV for energies E > 0.1 MeV, E > 1.0 MeV and for the complete neutron spectra. The FW-CADIS method, based on multigroup forward-adjoint discrete ordinates solver Denovo, was used to accelerate the final Monte Carlo calculation with Monaco code. Denovo utilizes Koch-Baker-Alcouffe parallel transport sweep algorithm over the XYZ meshes covering the problem domain and Krylov iteration on multigroup equations giving space-energy dependent fluxes. Such hybrid shielding methodology with mesh-based variance reduction parameters is very efficient for complex shielding problems with large computational domains. The same shielding methodology was used in the second part of the paper, where we calculated the radial neutron reflector heating and RPV power-level monitor's feasibility. The FW-CADIS was successfully used to produce well converged neutron fluence rate (in fast and thermal region) over the reduced I2S-LWR model, extending from the core to the biological shield exterior. Obtained results were then used to optimize calculations involving inelastic neutron scattering on 28Si and 12C, since these isotopes comprise the SiC type detector which will be used for power level monitoring.

I2S-LWR, pressure vessel, SCALE6.1, Monte Carlo, shielding, adjoint, variance reduction

nije evidentirano