Analysis of the influence of subchannel effects on the PWR average channel calculation

Full text of publication follows: Whole core analysis of nuclear reactors is presently made by the coupling of neutronic and thermalhydraulic codes. Nodes size is commonly one node per fuel assembly or higher what makes to lose the detailed representation of neutronic and thermal-hydraulic variables. In the first case, neutronic variables, the detailed distribution within different types of assemblies is used to generate the nodal distribution in the whole core but for thermal-hydraulic variables it is only calculated in an off-line manner for the hot channel in order to obtain the safety limits after the whole core calculation. The main problem of this whole core calculation is that some thermal-hydraulic variables are under or over-predicted. One of them is the enthalpy exchange between channels that is usually utilized as boundary condition for the hot channel subchannel analysis. This fact makes the subchannel analysis not to be very realistic. The important implication of these conclusions is that the refinement of the subchannel analysis models will not be useful if subchannel effects are not taken into account when the whole core analysis is carried out. Systematic analysis of these effects has been developed using the SEANAP system, a neutronic and thermal-hydraulic integrated system that includes three main calculation levels: fuel rod clusters (WIMS-D code, 24 groups, S6 approximation, 2-D); fuel assemblies and core planes (COBAYA code, 2 groups, 2-D); and whole core (SIMULA/SIMTRAN code, 3-D). SIMTRAN code is a whole reactor core dynamic model consisting in the neutronic code SIMULA, with six delayed neutrons precursor groups, and the thermalhydraulic code COBRA IIIC/ MIT2, that is a HEM code (4 equations) with explicit cross flows. A quarter of fuel assembly nodalization is used for both codes. Relations between whole core and subchannel calculations for critical variables like fuel and clad temperatures, DNBR, void fraction and cross flows have been analyzed for the 68 hottest channels (one eighth of the core) at different levels of power, pressure and flow but steady state scenario and some correlations can be extracted from them. The effect of geometry within the assembly and the over-prediction of cross flows in the whole core analysis are clearly observed in the results. Effect of the enthalpy distribution averaging in the enthalpy exchange term and therefore in the solution of the HEM equations has been analyzed. As second step in the study, a two fluid model subchannel thermal-hydraulic code, COBRA-TF, developed by the Pacific Northwest Laboratory for the United States Nuclear Regulatory Commission (US-NRC), is being integrated in the system to carry out the analysis over the 68 hottest channels in order to compare the results obtained previously with the ones generated by a more refined model code. (authors)