Separate effects tests on hydrogen combustion during direct containment heating events

In the frame of severe accident research for light water reactors Forschungszentrum Karlsruhe (FZK/IKET) operates the facilities DISCO-C and DISCO-H since 1998, conceived to investigate the direct containment heating (DCH) issue. Previous DCH experiments have investigated the corium dispersion and containment pressurization during DCH in different European reactor geometries using an iron-alumina melt and steam as model fluids. The analysis of these experiments showed that the containment was pressurized by the debris-to-gas heat transfer but also to a large part by hydrogen combustion. The need was identified to better characterize the hydrogen combustion during DCH. To address this issue separate effect tests in the DISCO-H facility were conducted. These tests reproduced phenomena occurring during DCH (injection of a hot steam-hydrogen mixture jet into the containment and ignition of the air-steam-hydrogen mixture) with the exception of corium dispersion. The effect of corium particles as igniters was simulated using sparkler systems. The data will be used to validate models in combustion codes and to extrapolate to prototypic scale. Tests have been conducted in the DISCO-H facility in two steps. First a small series of six tests was done in a simplified geometry to study fundamental parameters. Then, two tests were done with a containment geometry subdivided into a subcompartment and the containment dome. The test conditions were as follows: As initial condition in the containment an atmosphere was used either with air or with a homogeneous air-steam mixture containing hydrogen concentrations between 0 and 7 mol%, temperatures around 100 C and pressure at 2 bar (representative of the containment atmosphere conditions at vessel failure). Injection of a hot steam-hydrogen jet mixture into the reactor cavity pit at 20 bar, representative of the primary circuit blow down through the vessel and hydrogen produced during this phase. The most important variables measured were (1) the increase in pressure in the containment vessel, (2) gas temperatures, and (3) the number of moles of hydrogen burnt. These tests characterize the time scale of the hydrogen combustion, its completeness and the combustion mode for different initial conditions in the containment. The fraction of burnt hydrogen was between 55% and 100% of total available hydrogen in the basic geometry, and 46% and 67%, respectively in the more prototypic geometry. The efficiency of the hydrogen combustion in respect to pressure rise in the containment was between 46% and 67%. Dedicated combustion codes must be applied to verify, if these results prove true for reactor scale. (orig.)