Microscopic depletion with the correction of microscopic cross sections in nodal diffusion code DYN3D


Microscopic depletion with the correction of microscopic cross sections in nodal diffusion code DYN3D

Bilodid, Y.; Kotlyar, D.; Shwageraus, E.; Fridman, E.; Kliem, S.

Nodal diffusion codes are used routinely for nuclear reactor simulations. The homogenized few-group macroscopic reaction cross sections (XS) for the nodal codes are generated beforehand in single assembly calculations using the lattice neutron transport codes. Usually core- and cycle-averaged operational conditions (coolant density, fuel temperature, boron concentration, etc.) and nominal power are utilized for a single assembly depletion simulation, and the variations of operational conditions are used for branching calculations.
The spectral conditions of single assembly depletion differ from local conditions in a real reactor core. Deviation of local spectral history from core-averaged values leads to deviations in fuel nuclide content and thus influences macroscopic cross sections. Dependence of XS on spectral history is taken into account by various methods: micro-depletion (Bahadir et al., 2005), Pu-correction (Bilodid and Mittag, 2010), spectral indexes (Baturin and Vygovskii, 2001) and exposure-weighted operational parameters (Bahadir et al., 2005).
DYN3D is a 3D nodal reactor dynamic code developed at the Helmholtz-Zentrum Dresden-Rossendorf mainly for transients, but also for steady-state and fuel cycles analysis in LWR cores with hexagonal or square fuel assemblies (Grundmann et al., 2005). Spectral history effects are taken into account in DYN3D by the Pu-correction method (Bilodid and Mittag, 2010). However it is not able to reproduce fuel reactivity changes due to outage periods.
In this work the modified microscopic depletion methodology was proposed and implemented in DYN3D. The most important innovations are: a.) correction of microscopic cross sections, scattering matrix, diffusion coefficients and kinetic parameters using local fissile content and, b.) the depletion solver which utilizes fast and accurate Chebyshev rational approximation method (CRAM).
The use of the CRAM (Gonchar and Rakhmanov, 1989, Pusa, 2011) in depletion solver allows to accurately calculate concentrations of all nuclides, which are present in nuclear fuel in considerable amounts. In the shown test cases about 1100 nuclides were considered in DYN3D, in contrast to about 50 nuclides considered in codes like SIMULATE and ANC.
In DYN3D depletion solver, the number of considered nuclides is chosen by user according to the task of simulation. This research has shown that about 80 nuclides (out of considered 1100) actually are important from neutronics point of view. However, for an accurate tracking of these 80 important nuclides it is necessary to consider all intermediate nuclides in transmutation chains, which results in significantly larger nuclide inventory (>300). On the other hand, knowledge of the full nuclide content can be used for realistic decay heat and radiotoxicity calculation.
In this work Serpent (Leppänen et al., 2014) continuous energy Monte Carlo neutron transport code with JEFF-3.1 isotopic library was used to obtain reference solutions for the examined test cases and to generate homogenized macro- and microscopic cross sections for DYN3D.

  • Contribution to proceedings
    PHYSOR 2016, 01.-05.05.2016, Sun Valley, ID, USA
    Proceedings of PHYSOR 2016
  • Lecture (Conference)
    PHYSOR 2016, 01.-05.05.2016, Sun Valley, ID, USA

Permalink: https://www.hzdr.de/publications/Publ-22424