The inertisation of PWR containments by injection of liquid carbondioxide

Several methods are under discussion to mitigate the consequences of the release of large amounts of hydrogen which inevitably accompanies the course of any severe accident in light water cooled power reactors. The installation of ignitors is a method which is in particular difficult to defend because one is requested to support this method with difficult analytical predictions of the resulting combustion process, starting with the expected spectrum of release rates, the associated distribution process ending with the analysis of the variety of possible combustion processes. Hence, less questionable mitigation procedures should be preferably developed. Inertisation of the containment atmosphere aims to prevent combustion either partially (exclusion of highly energetic combustion modes) or totally (exclusion of any combustion mode). Injection of liquefied carbondioxide into a building is a well established method which is widely used, e.g. in inerting crude oil transporters to prevent the possibilities of explosions during the transport. Numerous other applications have been developed by fire fighting institutions and were promoted by insurance companies. The containment code RALOC has been modified and supplemented to simulate the processes characterizing the injection of liquefied carbondioxide through nozzles into a LWR containment after a preceding loss of coolant accident. Main aim of the liquid C02 injection is to arrive rapidly at a uniform distribution of the inert gas within the free containment atmosphere. Large driving temperature differences are formed inside the containment which promote a natural convection driven distribution process. Sufficient heat capacity is available from the post LOCA atmosphere and associated to the large masses of containment structures, which facilitate the rapid evaporation of generated CO2 particles. This process is also modelled by the code. Several analyses have been performed with the modified code showing the transient distribution until uniform concentration of the atmospheric components steam, air and carbondioxide is reached. Time intervals necessary to achieve the required inerting status of approx. 60 vol% may easily be below 30 minutes. The combination of inertisation with the installation of a number of catalytic recombiners remains as a supplementary option. A limited number of experimental results has been made available by fire fighting industry to facilitate the application of the modified code for validation purposes. Comparisons between analyses and experimental results are satisfactory. However, some tests, more typical for PWR containments would be desireable to confirm the post-accident inertisation simulations. (author)