Hybrid microscopic depletion model in nodal code DYN3D. (June 2016)

• Record Type: Journal Article
• Authors: Bilodid, Y.; Kotlyar, D.; Shwageraus, E.; Fridman, E.; Kliem, S.
• Is Part Of: Annals of nuclear energy. Volume 92(2016:Jun.)
• Page Start: 397
• Page End: 406
• Publication Date: 2016-06
• Subjects: History effects -- Spectral history -- Microscopic depletion -- DYN3D; Nuclear energy -- Periodicals; Nuclear engineering -- Periodicals; 621.4805
• DOI: 10.1016/j.anucene.2016.02.012
• Languages: English
• ISSNs: 0306-4549

Abstract

Highlights: A new hybrid method of accounting for spectral history effects is proposed. Local concentrations of over 1000 nuclides are calculated using micro depletion. The new method is implemented in nodal code DYN3D and verified. Abstract: The paper presents a general hybrid method that combines the micro-depletion technique with correction of micro- and macro-diffusion parameters to account for the spectral history effects. The fuel in a core is subjected to time- and space-dependent operational conditions (e.g. coolant density), which cannot be predicted in advance. However, lattice codes assume some average conditions to generate cross sections (XS) for nodal diffusion codes such as DYN3D. Deviation of local operational history from average conditions leads to accumulation of errors in XS, which is referred as spectral history effects. Various methods to account for the spectral history effects, such as spectral index, burnup-averaged operational parameters and micro-depletion, were implemented in some nodal codes. Recently, an alternative method, which characterizes fuel depletion state by burnup and 239 Pu concentration (denoted as Pu-correction) was proposed, implemented in nodal code DYN3D and verified for a wide range of history effects. The method is computationally efficient, however, it has applicability limitations. The current study seeks to improve the accuracy and applicability range of Pu-correction method. The proposed hybrid method combines the micro-depletion method with a XS characterization technique similar to the Pu-correction method. The method was implemented in DYN3D and verified on multiple test cases. The results obtained with DYN3D were compared to those obtained with Monte Carlo code Serpent, which was also used to generate the XS. The observed differences are within the statistical uncertainties.