In the event of failure of the active residual heat removal systems, four containment cooling condensers (CCC) are designed to remove residual heat from the containment to the dryer — separator storage pool located above the containment. The CCCs are actuated by rising temperatures in the containment. They use natural circulation both on the primary and on the secondary sides. The nominal heat transfer capacity of each condenser is 4 MW based on a containment pressure of 3 bar (absolute) and a cooling water temperature of 100°C. In a hypothetical core melt accident the thermal capacity could be 2 or 3 times higher, depending on the higher containment pressure and temperature. The containment cooling condenser has been experimentally tested at nearly original scale at the PANDA facility of the Paul Scherrer Institute in Switzerland.

The working principle of the CCC is shown in Figure 3. It comprises a simple heat exchanger mounted about 1 m above the water level of the core flooding pool. If the temperature in the drywell atmosphere increases over that in the dryer-separator storage pool, the water inside the heat exchanger tubes heats up. It flows to the outlet line due to the slope of the exchanger tubes. The outlet line ends at a higher elevational level than the inlet line, so the lifting forces are increased for the whole system. Depending on the heat transfer rate and cooling water temperature, secondary-side flow can be either single-phase, intermittent or two-phase.

Depending on the type of loss-of-coolant accident (LOCA) and on the time after onset of accident conditions, the medium on the heat exchanger primary side is either nitrogen, a
nitrogen-steam mixture or pure steam. In the hypothetical case of a core melt accident, a hydrogen-steam mixture would also be possible. Given nitrogen, steam and mixtures thereof, primary flow is downwards because the densities of pure gases and a nitrogen-steam mixture increase with decreasing temperature. This results in the expected downward flow. Condensed steam drops into the core flooding pool. However, the opposite is true for a hydrogen-steam mixture, as the density of this mixture decreases with decreasing temperature, resulting in an upward flow through the heat exchanger tube bundle. This does not pose any problem for the SWR 1000 because both directions of flow on the primary side are equivalent.

Finned-tube

cooler

Nevertheless, problems will arise if more hydrogen is generated than can be accommodated in the containment above the CCCs. The atmosphere in the containment is stratified with a high hydrogen content above the CCCs and a high steam content below. With increasing hydrogen mass the boundary between both stratified regions would drop lower and lower. After a certain time the CCCs would become ineffective because they would be surrounded by cooled hydrogen. To prevent this condition from occurring Siemens designed a hydrogen overflow line. The upper end of this line is located at a higher elevation than the CCCs and its lower end is higher than the lower end of the vent pipes. When the heat transfer capacity of a CCC deteriorates, the pressure in the containment increases until the hydrogen overflow pipe is empty of water (but the vent pipes still contain some water). This produces a forced-flow condition of cooled hydrogen-steam mixture from the drywell to the wetwell. The steam condenses in the wetwell pool and the hydrogen rises into the wetwell pool atmosphere. The hydrogen mass flow is self-controlled. The boundary of the stratified regions stabilizes at a position at which as much steam is condensed as is being generated.

In the case of a hypothetical core melt accident, there first occurs normal natural circulation in the drywell. Later, there is natural circulation with opposed flow directions on the primary side and with stratification between the cold hydrogen above and hot steam below. In the final phases there is a self-induced forced-flow of hydrogen from the drywell to the wetwell. To
simulate all these effects by way of computer code modeling would pose a considerable challenge. The experimental tests performed at the PANDA test facility were much easier, and demonstrated that the entire system is effective and stable, and the pressure differential of several kPa between the drywell and wetwell was generated without any problems.