Dr. Dirk Lucas

Head Computational Fluid Dynamics
Phone: +49 351 260 2047

Reflux-Condenser-Operation with counter-current flow limitation in a hot leg of a Pressurized Water Reactor

Scientific background

If accidents occur on Nuclear Power Plants (NPP), the safe reactor core cooling always has to be guaranteed, even in case of additional failure of components. One of the considered accident scenarios is a little leakage at the primary loop of a Pressurized Water Reactor (PWR) with simultaneous failure of the high-pressure emergency core cooling system after the shut-down of the main coolant pumps. In the course of this scenario a natural circulation starts between the “hot” reactor pressure vessel and the “cold” higher-positioned steam generator.

This regime cools the reactor core by heat removal from the primary to the secondary loop through the steam generator needle pipes. Further, the pressure in the primary loop decreases due to coolant loss through the leakage and steam appears in the reactor that flows together with the liquid coolant by two-phase natural circulation to the steam generator and condensates there. Also in this case the heat removal is continued safely. If the coolant level in the reactor pressure vessel decreases below the position of the main coolant pipes, only steam flows to the steam generator and the natural circulation collapses. The steam condensation in the steam generator continues but approximately one half of the condensate returns to the reactor pressure vessel in counter-current direction to the steam. This phenomenon takes place in the “hot leg” of the main coolant pipe and is called “Reflux-Condenser-Mode”. The process is illustrated in the picture above.

As known, the counter-current flow of steam and condensate in horizontal pipes works in stable state only at defined conditions. At high steam flows the condensate is retained in the steam generator inlet chamber and in the hot leg and a counter-current flow limitation occurs. Thereby the core cooling and the level in the reactor pressure vessel are decreased simultaneous, that may result in an overheating of the reactor core. For this reason this phenomenon has to be considered during the safety-related design phase of nuclear power plants. One important part of safety analysis is fluid dynamic calculation with three-dimensional computer codes. In order that these codes may simulate the flow phenomena with the required accuracy in a geometry-independent way and at a wide parameter range, a code validation with temporal and spatial high-resolved measurement data is necessary.

Experimental setup

To get CFD-grade measurement data of counter-current flow limitation in horizontal channels, a model of the PWR hot leg and steam generator inlet chamber of a German Konvoi NPP was designed and erected in the TOPFLOW pressure tank basically on a scale of 1:3. In doing so, the experiences of previous test series (2007) and of experiments on a horizontal flow channel were considered, so that as the renewed test section as the current operational conditions were improved significantly. In general the test rigs inside the pressure tank are operated at pressure equilibrium. Hence the side walls of the test section may consist of glass windows. Additional the horizontal channel was designed with a rectangular cross-section of 250 x 50 mm² (height x width), that allow optical observation without distortions. Furthermore the steady-state operation of the test rig as a part of a closed loop yield to significantly improved data for the investigation of the partial and complete counter-current flow limitation, because the basis of the data evaluation is a constant water inventory.

Hot leg test rig installed in the TOPFLOW pressure tank

Scheme of the TOPFLOW pressure tank with installed hot leg test rig and auxiliary systems

The above figure shows a scheme of the TOPFLOW pressure tank with two circulation loops, a pressurized gas supply system and a gas cooler. At the right hand side in the tank the high-pressure condenser is arranged. This device condenses residual steam after the test rig and allows the pressure equilibrium between the inner volume of the test section and the tank atmosphere. The horizontal test channel is fixed between two separation tanks, which model the steam generator (right hand side) and the reactor pressure vessel (at the left side).


The current experiments deal with counter-current flows of gas and water both with air-water and steam-water mixtures. In the first case the test rig was operated at pressures of 0.1 and 0.2 MPa and temperatures of app. 20 °C. The non-adiabatic tests was carried out at pressures of 1, 2.5 and 5 MPa and almost saturation conditions. The water mass flow, injected into the steam generator separator, was set to 0.3, 1 and 2 kg/s. The gas volume flow, fed into the reactor pressure vessel simulator, was varied in such a way that the required flow regimes developed, started with undisturbed counter-current flow via increased counter-current flow limitation (CCFL) up to the complete CCFL. Finally the gas volume flow was decreased stepwise to analyze hysteresis effects.

Beside app. 100 operational data, optical picture sequences of the inclined steam generator inlet chamber and of the horizontal channel are available. Additional pressure data were recorded at some position along the horizontal channel. On the basis of the operational data flooding lines were calculated for the liquid and gas phase in terms of the dimensionless Wallis parameter. These charts reveal interesting dependency of the CCFL on both the flow mixtures and the operational pressure. Additional the quality of the measurements may be assessed by analyzing the operational data.

The picture sequences were used to define the flow regime, e.g. for the verification of the complete CCFL. As an example the following figures show the evolution of a slug in a steam water flow at 5 MPa with complete CCFL.

Visualization of slug development in the hot leg test rig

Visualization of slug development in the hot leg model with complete counter current flow limitation; test parameter: steam water flow, pressure: 5 MPa, saturation condition, injected water mass flow: 2 kg/s, steam mass flow: 1.05 kg/s

Otherwise at present time picture analyzing algorithms are under development. This task should result in the estimation of equivalent interfacial areas even on churn turbulent flows.

On the basis of the pressure data along the channel slug frequencies were calculated, which show clear dependency on the pressure and also on the injected water mass flow.

The measured and evaluated experimental data were used for code development and validation. Current results of comparative computer flow simulations calculated by CFD code are published on this page.



This work is based on a research project funded by the German Federal Ministry of Economics and Energy, support codes: 150 1329 and 150 1411. The authors assume the responsibility for the content.



Dr. Dirk Lucas

Head Computational Fluid Dynamics
Phone: +49 351 260 2047