Investigation of a Pin-PowerReconstruction Method Used bythe Nodal Code SIMULATE-5
Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
This master thesis aims to investigate a method used by the software package CASMO/SIMULATE to calculate the power distribution in a nuclear reactor core.
The method is a two-part solution where the nodal code SIMULATE is used to calculate the axial homogenous power distribution for all fuel assemblies. The homogeneous solution is then distributed on the rods in the fuel bundles by the multiplication of so-called form factors. Form factors are derived from the lattice code CASMO. The lattice code CASMO is based on a 2D approach, which is an approximation that doesn’t capture any 3D dependencies among the form factors.
However, a 3D dependence is expected among real form factors. This is obvious because the presence of a control rod in an axial segment will affect and capture neutrons, although these are located in an adjacent zone.
Another more exact approach for calculating neutron transport is to use the Monte Carlo principle and the program MCNP. The method is a non-deterministic, stochastic, method which is more accurate if the computation is performed long enough to get a sufficient statistical basis.
In this work, a power distribution on pin-level predicted by the 2D/3D method deployed by CASMO/SIMULATE has been compared to a more realistic power distribution predicted using a three-dimensional MCNP reference model. By doing this, “real” form factors can be derived and compared to the 2D form factors that has been predicted by CASMO.
This study has shown that the use of two-dimensional form factors is a very accurate approximation. However, if pin-powers are to be predicted in a zone located close to an interface between two materials, with strongly different cross-section data, the form factors will be affected by the three dimensional nature of the problem that the 2D CASMO principle cannot completely account for.
The largest error obtained is 12.6% and it can be concluded that it takes almost 100 axial nodes to obtain those errors in interface zones. However, this is not a major issue in reality because the number of nodes used is often a lot fewer.
Place, publisher, year, edition, pages
2014. , 40 p.
UPTEC ES, ISSN 1650-8300 ; 14011
Engineering and Technology
IdentifiersURN: urn:nbn:se:uu:diva-224904OAI: oai:DiVA.org:uu-224904DiVA: diva2:719322
Vattenfall Nuclear Fuel AB
Master Programme in Energy Systems Engineering
Loberg, JohnLindahl, Sten-Örjan
Österlund, MichaelJönsson, Petra