Effective subgroup method employing macro level grid optimization for LWR applications. (July 2019)

• Record Type: Journal Article
• Authors: Park, Hansol; Joo, Han Gyu
• Is Part Of: Annals of nuclear energy. Volume 129(2019)
• Page Start: 461
• Page End: 471
• Publication Date: 2019-07
• Subjects: Resonance self-shielding -- Subgroup method -- Subgroup fixed source problem -- Macro level grid -- Escape cross section -- Direct whole core transport; Nuclear energy -- Periodicals; Nuclear engineering -- Periodicals; 621.4805
• DOI: 10.1016/j.anucene.2019.02.005
• Languages: English
• ISSNs: 0306-4549

Abstract

Highlights: Macroscopic escape XSs are computed at common grid, called macro level grid. Isotopewise escape XSs are interpolated from the macro level grid. Optimum number of subgroup levels for the grid turns out to be 8 for LWR geometry. Total number of SGFSPs are fixed regardless of number of isotopes and their levels. Same accuracy as original isotopewise SGFSP scheme and reduced computing time. Verified at 3 by 3 assembly checkerboard with HFP thermal feedback. Abstract: The Macro Level Grid scheme for the efficient application of the subgroup method is presented, that employs the number density consideration factor and the temperature consideration factor for the treatment of non-uniform number densities and temperature distributions in a core. This scheme provides the efficient resonance treatment in direct whole core calculations of power reactors that involve thermal feedback and isotopic depletion. The new method solves the subgroup fixed source problem only 8 times per energy group with 8 macroscopic subgroup levels, regardless of the number of resonance isotopes in the problem of interest. The escape cross section of each isotope is obtained by interpolation using the pre-calculated ones at the specified macroscopic subgroup levels. This scheme turns out to be superior to the conventional scheme in terms of computing time and accuracy. More than 30% of the computing time for fixed source problems is saved compared to the conventional one with negligible reactivity errors of about a few pcm, whereas the conventional one has consistent reactivity errors of about +60 ∼ +150 pcm for typical pin-cell problems in a light water reactor. Moreover, it turns out that the new method provides high accuracy not only for very heterogeneous uranium dioxide and mixed oxide pin-cell checkerboard problems, but also for the depletion calculation of a multi-assembly problem involving hot full power thermal feedback with significantly shortened times.