[en] Past considerations of hydrogen generated in containment buildings have tended to focus attention on design basis accidents (DBAs) where the extent of the in-core metal-water reaction is limited at low values by the operation of the emergency core cooling systems (ECCS). The radiolysis of water in the core and in the containment sump, together with the possible corrosion of metals and paints in the containment, are all relatively slow processes. Therefore, in DBAs the time scale involved for the generation of hydrogen allows sufficient time for initiation of measures to control the amount of hydrogen in the containment atmosphere and to prevent any burning. Provisions have been made in most plants to keep the local hydrogen concentration below its flammability limit (4% of volume) by means of mixing devices and thermal recombiners. Severe accidents, involving large scale core degradation and possibly even core concrete interactions, raise the possibility of hydrogen release rates greatly exceeding the capacity of conventional DBA hydrogen control measures. The accident at Three Mile Island illustrated the potential of unmitigated hydrogen accumulation to escalate the potential consequences of a severe accident. In a severe accident scenario, local high hydrogen concentrations can be reached in a short time, leading to flammable gas mixtures in containment. Another possibility is that local high steam concentrations will initially create an inert atmosphere and prevent burning for a limited time. While such temporary inerting provides additional time for mixing (dilution) of the hydrogen with containment air, depending on the quantity of hydrogen released, it prevents early intervention by deliberate ignition and sets up conditions for more severe combustion hazards after steam condensation eventually occurs, e.g., by spray initiation or the long term cooling down of the containment atmosphere. As the foregoing example indicates, analysis of the hydrogen threat in post-accident containments is complex and highly plant- and scenario-specific. Many aspects must be considered in analysis of a hydrogen threat: accident sequences, hydrogen production rates, atmosphere thermal hydraulics, mixing processes, combustion phenomena, accident management strategies and mitigation hardware performance. Research activities on these topics have been continuing in several countries for the past two decades. This report summarizes current concepts for hydrogen mitigation in containments, concentrating primarily on measures that are already being implemented or those that show promise in the near future for hydrogen mitigation in severe accidents