Generic numerical investigations of the critical overlap of the suction nozzles of pumps to avoid air entrainment

In this work a new and effective method is introduced to analyse the surface vortex formation and determine its characteristic parameter, the critical submergence, in large and complex water intake structures of nuclear installations. Surface vortex formation induces flow disturbances and could transport gas directly into the pump intakes with negative consequences on the pump operation. Unfavorable operating conditions could furthermore lead to pump damage in the long term. Therefore, a reliable pump operation requires vortex free inflow conditions. The critical submergence marks the minimum water level above the intake, where a surface vortex formation can still be excluded. The safe and reliable long-term operation of the emergency core cooling system is an important requirement to achieve the nuclear safety objectives. The pumps of these systems transport the coolant from large and complex emergency reservoirs and sumps during an accident. In these cases, the prediction of the critical submergence is difficult. In common industrial processes the critical submergence is determined with extensive experiments or with empirical and semi-empirical correlations. In case of sump operation in a Pressurized Water Reactor (PWR) the German Reactor Safety Commission (RSK) recommends to determinate the critical submergence with large scale experiments. If no experimental data is available, the application of the simple correlation of ANSI (American National Standards Institute) is recommended, which is based on experimental and analytical results. The in this work developed CVA-Method (Combined Vortex Analyses) is based on the combination of numerical CFD-simulations and the analytical vortex model of Burgers and Rott. Therefore, it provides an effective alternative solution to calculate the critical submergence and the main parameters of a surface vortex like circulation, gas-core length and tangential velocity distribution. The CFD-simulations calculate well the parameters of the fluid flow in complex intake structures beyond the core region of the surface vortex. However, the simulation of the vortex-core region as well as the gas core evolution is very time-consuming. The analytical Burgers-Rott-model calculates the gas-core length of a surface vortex if two parameters: the circulation and the suction parameter determined beyond the vortex-core region are known. In the CVA-Method these parameters are calculated with the help of the commercial CFD code ANSYS CFX and are used in analytical equations derived from the Burges-Rott-model. With the CVA-Method, based on the results of two appropriate CFD simulations, the critical submergence can be analytically calculated for a wide range of parameters, e.g. intake mass flow. The CVA-Method has been validated on the bases of two experiments with vertical water intakes. In the experiment of Moriya the phases of the surface vortex formation and its local parameters have been measured. The physical models, the modelling parameters and the boundary conditions of the CFD calculations have been defined based on the comparison with the experimental results. Moreover, according to this experiment, the CVA-Method has been successfully validated to calculate the gas-core length of a surface vortex. The determination of the gas-core length is an important step of the CVA-Method and allows the analytical calculation of the critical submergence. The second experiment used for the validation has been performed by Jain et al. In this experiment the critical submergence has been measured by different influence parameter. According to the experimental results the CVA-Method has been successfully validated on the bases of the critical submergence's calculation at diverse intake velocities, circulations, viscosities and intake diameters. Furthermore, based on the results a simple analytical formula has been derived: the CVA-Formula. This describes the correlation between the critical submergence, the Froude number and the intake diameter considering the circulation with a constant parameter. During the post- test calculations of the Jain et al. experiment this constant has been determined for three cases. In following, the validated CVA-Method has been applied to analyse the correlations to predict the critical submergence widely used in the industry and to quantify their conservatism. Next to the ANSI correlation, which also used in nuclear safety procedures, the correlation of Odgaard, Rindells and Gulliver, Jain et al. and Knauss has been selected for the analyses. The application of these correlations is only recommended in a defined, limited parameter range. The comparison of the results obtained with the correlation and results provided by the CVA-Method demonstrated well the conservatism, the limitations and the possible extension potential of the correlations in question. Furthermore, the circulation limit of the ANSI correlation has been determined with the CVA-Method. The results confirm that the CVA-Method can make an important contribution to the nuclear safety procedures in order to prove the applicability of the correlations. Finally, the CVA-Method has been applied to analyse the surface vortex formation in the containment sump of a Vor-Konvoi type PWR. During the postulated accident scenario, the emergency core cooling system (TH-system) is in sump operation. Above the pump intakes of the TH-system a concrete ceiling of the sump can be found. Generally, in safety analyses this concrete ceiling is considered as a vortex breaker, when the water level exceeds its height. Therefore, first an accident scenario with a water level above the concrete ceiling has been analysed with ANSYS CFX simulations. A 400 cm${}^2$ break in the cold leg of the primary circuit has been postulated. Furthermore, only two of the four TH-pumps are available, each delivering a mass flow rate of 300 kg/s. The simulations show a strong vortex formation in the intake chambers of the active pumps. According to the vortex shape the simulations have proved the effectivity of the concrete ceiling as a vortex breaker. To determinate the critical submergence of the TH-pumps at further mass flow and water level another accident scenario has been postulated. Because the application of the CVA-Method requires a free surface above the intakes, in the second scenario, the water level has been decreased under the concrete ceiling. To determinate the circulation and the suction parameter for the analytical part of the CVA-Method, two CFD simulations have been performed with the same water level, but with different mass flow rates. Based on the results the critical submergence has been determined with the help of the CVA-Method for the total operational mass flow range of the TH-pumps. This demonstrated well the application of the CVA-Method for large and complex intake structures. The performed analysis demonstrates the versatile application potential of the CVA-Method, and shows, that extended with further validation calculations this method can provide an effective alternative for the current nuclear safety calculations.