Experimental Analysis of Containment Atmosphere Mixing and Stratification Induced by Vertical Fluid Release

Continuing safety assessments are important for the present and future generations of nuclear reactors to ensure safe and efficient use of nuclear energy. For safety analysis purpose, lumped parameter codes as well as codes with 3D capabilities are used to predict and simulate the containment conditions following a hypothetical accident. The assessment of code capability to predict consequences of hypothetical accidents can be done by comparing code simulations with experimental data obtained in large scale facilities under prototypical thermal hydraulic conditions. The large scale PANDA test facility built at the Paul Scherrer Institute (PSI) in Switzerland is well suited to perform thermal hydraulic experiments for investigating integral containment system response during accidents and also to study multicompartmental 3D effects related to a Light Water Reactor (LWR). This paper presents the experimental results and the analysis of two tests performed in a containment that is represented by two large vessels (volume of each vessel is 90 m³) interconnected by a bended pipe of 1 m diameter. The tests address the possibility of the permanence of hydrogen stratification just beneath the containment dome under conditions determined by low velocity and a plume or jet of higher density which could be, for instance, due to steam produced by evaporation of the sump. Since practically no data are available on these specific phenomena in large scale test facilities, the capabilities of the codes could not be yet verified. During the experiments, a stratified helium-rich layer is created in upper part of one of the two vessels and thereafter, interaction of an ascending buoyant jet with the stratified helium layer is investigated. Helium is used to simulate hydrogen released by cladding metal water reaction. Two tests with initial densimetric Froude number equal to 22.6 (Test_Air/He) and 104 (Test_Steam/He) are investigated. The distance between the pipe exit for vertical injection of the fluid and the bottom front of the stratified helium rich layer is about 2m for both the investigated test cases. Results depicting the mixing of lower and upper vessel regions are presented in terms of helium concentration profiles and temperature contour maps. The results show that the maximum penetration height is controlled by the initial exit conditions (momentum flux and the buoyancy flux) and density stratification inside the test vessel. Additionally, for the given test conditions and geometry, most of the injected fluid remains trapped in upper part of one of the two test vessels in which helium layer is created.