Performance testing of engineered corium cooling systems

The coolability of ex-vessel core debris continues to be an issue of concern in the realm of light water reactor safety. Extensive research into corium/concrete interaction phenomena has been unable to establish the certainty of melt quench and stabilization within the containment boundary for all credible cases of cooling restricted to top flooding. As a result, there has been continuing interest in engineered systems that can augment cooling. This paper describes the testing of two passive cooling concepts that inject water into corium from below via nozzles embedded within the basemat: one with porous concrete nozzles and the other with a type of composite nozzle. The latter supplements water injection with noncondensable gas to stabilize flow and suppress vapor explosions. Each test involved a 136 kg melt composed of 56/23/14 wt% UO2/ZrO2/siliceous concrete at an initial depth of 30 cm. The setup with the porous concrete nozzles successfully injected water into the melt at heads as low as 2.3 m. The composite nozzle test was partially successful, with three nozzles delivering coolant while a fourth was damaged by the melt and failed to inject water. The melts cooled twice as fast as similar ones tested in a top flooding configuration. These experiments confirmed earlier work at Forschungszentrum Karlsruhe and elsewhere indicating that cooling via bottom water injection is a particularly effective method for quenching ex-vessel corium melts.

► Experiments tested two engineered corium cooling systems.
► The systems passively inject water into corium from below.
► These systems cool corium much faster than top flooding.