The European project FLOMIX-R: Description of the experimental and numerical studies of flow distribution in the reactor primary circuit(Final report on WP 3)

The flow distribution in the primary circuit of the pressurized water reactor was studied with experiments and Computational Fluid Dynamics (CFD) simulations. The main focus was on the flow field and mixing in the downcomer of the pressure vessel: how the different factors like the orientation of operating loops, the total loop flow rate and the asymmetry of the loop flow rates affect the outcome. In addition to the flow field studies the overall applicability of CFD methods for primary circuit thermal-hydraulic analysis was evaluated based on the CFD simulations of the mixing experiments of the ROCOM (Rossendorf Coolant Mixing Model) test facility and the mixing experiments of the Paks NPP. The experimental part of the work in work package 3 included series of steady state mixing experiments with the ROCOM test facility and the publication of results of Paks VVER-440 NPP thermal mixing experiments. The ROCOM test facility models a 4-loop KONVOI type reactor. In the steady-state mixing experiments the velocity field in the downcomer was measured using laser Doppler anemometry and the concentration of the tracer solution fed from one loop was measured at the downcomer and at the core inlet plane. The varied parameters were the number and orientation of the operating loops, the total flow rate and the (asymmetric) flow rate of individual loops. The Paks NPP thermal mixing experiments took place during commissioning tests of replaced steam generator safety valves in 1987-1989. It was assumed that in the reactor vessels of Paks VVER-440 NPP equipped with six loops the mixing of the coolant is not ideal. For the realistic determination of the active core inlet temperature field for the transients and accidents associated with different level temperature asymmetry a set of mixing factors were determined. Based on data from the online core monitoring system and a separate mathematical model the mixing factors for loop flows at the core inlet were determined. In the numerical simulation part of the work package 3 the detailed measurements of ROCOM tests were used for the validation of CFD methods for primary circuit studies. The selected steady state mixing experiments were simulated with CFD codes CFX-4, CFX-5 and FLUENT. The velocity field in the downcomer and the mixing of the scalar were compared between CFD simulations and experiments. The CFD simulations of full scale PWR included the simulation of Paks VVER-440 mixing experiment and the simulation of Loviisa VVER-440 downcomer flow field. In the simulations of Paks experiments the experimental and simulated concentration field at the core inlet were compared and conclusions made concerning the results overall and the VVER-440 specific geometry modelling aspects like how to model the perforated elliptic bottom plate and what is the effect of the cold leg bends to the flow field entering to the downcomer. With Loviisa simulations the qualitative comparison was made against the original commissioning experiments but the emphasis was on the CFD method validation and testing. The overall conclusion concerning the CFD modelling of the flow field and mixing in the PWR primary circuit could be that the current computation capacity and physical models also in commercial codes is beginning to be sufficient for simulations giving reliable and useful results for many real primary circuit applications. However the misuse of CFD methods is easy, and the general as well as the nuclear power specific modelling guidelines should be followed when the CFD simulations are made