On the failure modes of alternative containment designs following postulated core meltdown

The containment response to a postulated core meltdown accident in a PWR ice condenser containment, a BWR Mark III containment and a BWR non-inerted Mark I containment has been examined to see if the WASH-1400 containment failure mode judgement for the Surry large, dry containment and the Peach Bottom Mark I inerted-containment are likely to be appropriate for these alternative containment plant designs. For the PWR, the representative accident chosen for the analysis is a large cold leg break accompanied by a loss of all electric power while the BWR representative event chosen is a recirculation line break without adequate core cooling function. Two containment event paths are studied for each of these two cases, depending on whether or not containment vapor suppression function is assumed to be available. Both the core and the containment pressure and temperature response to the accident events are computed for the four time intervals which characterize (a) blowdown of the pipe break, (b) core melt, (c) vessel melt-through, and (d) containment foundation penetration. The calculations are based on a best estimate of the most probable sequence, but certain phenomena and events were followed down multiple tracks. These include the temperature of the non-condensibles escaping the ice condenser into the upper compartment, the performance of the pressure suppression system, the distribution of non-condensibles between compartments, and the degree and rate of combustion of hydrogen generated from metal-water reactions. For the PWR ice condenser case, results indicate that the containment would be breached by (i) steam overpressurization during the blowdown period (time less than 20 sec) if the ice condenser fails to perform its function, (ii) by overpressurization and thermal stress during the core melt period if 25% or more of the core zirconium reacts with water followed by hydrogen burning and, and (iii) by the overpressurization due to non-condensibles before containment floor penetration is completed. For the BWR Mark III case, similar conclusions can be drawn for the loss of vapor suppression, and for the hydrogen burning if the extent of zirconium-water reaction is more than 35% of the core inventory. If the hydrogen burning fails to materialize, the containment can retain its integrity until containment meltthrough provided the melting is confined to the reactor pedestal area. It appears that the non-inerted Mark I containment is not so vulnerable to overpressurization from hydrogen burning as the Mark III; however, acceptable temperatures may be exceeded.