Hydrogen Recombiners have a long murky history. Maybe I should be a blogger -- but this is a Primer:
- When the Design Bases LOCA was established and the ECCS rule was finally put in place it was recognized that there were several sources of hydrogen to be dealt with: (a) high temperature Zircalloy-steam reaction which oxidizes the zircalloy and liberates hydrogen, (b) longer term radiolytic decomposition of water (2 parts H2O --> 2 parts H2 and one part O2), (c) really long term corrosion on surfaces.
- Because the ECCS rule -- after postulating various single failures still leaves some remaining functioning ECCS -- the metal water reaction is limited to a very small amount of hydrogen being generated (<1% of the clad). This would be a short term release in containment of hydrogen and would be proportional to the total cladding volume.
- Over weeks and months the ECCS recirculation of water through a reactor will result in separation of small amounts of hydrogen and oxygen -- that post-accident would be into containment. (Under normal reactor operation, a slight overpressure of hydrogen gas - dissolved in the water - is used to keep the coolant from separating too much water into hydrogen and oxygen). The rates of radiolysis are fairly well known and measured by experiment -- so one can predict just how much comes from radiolysis.
- Because the long term buildup of hydrogen in containment original regulations required recombiners be installed. These are units that take in containment atmosphere and with heated elements such as platinum force the hydrogen and oxygen to recombine in a slow controlled fashion. The problem with such units is that they can only function in a situation where the amount of hydrogen is limited (I seem to recall - but my memory is rusty -- that hydrogen had to be <5-6% hydrogen). Their operating procedures would require them to be shut down if hydrogen levels were higher -- for two reasons: they'd overheat permanently damaging the catalytic elements -- but also in a high concentration hydrogen-oxygen environment they would become an ignition source for a possible explosion.
So the problem with recombiners is:
(1) Recombiners were originally designed for a narrowly postulated accident scenario (e.g. <1% metal water reaction) that experience such as TMI, but also all other severe accident studies shows is too narrow to be useful for much.
(2) BWRs with Mark I containments are inerted by replacing the containment atmosphere with enough nitrogen gas that the oxygen is typically <1% -- vs. the normal atmospheric oxygen level of ~20.9%. A recombiner would not be effective in removing hydrogen because there is insufficient oxygen to run the recombiner. The industry demonstrated back in the 1980's that: (a) the DBA could not credibly cause a flammable situation even when radiolysis and corrosion were added in, (b) a severe accident with greater than 1% clad reacted would add more hydrogen but would also proportionally decrease the amount of oxygen thus yielding a less flammable situation in containment.
(3) Other designs with relatively smaller containments (later BWR designs and the Westinghouse Ice-Condensers) intalled igniters (basically like automotive "glow plugs" used in diesel cars). These would be turned on to burn the hydrogen as it was generated and released.
The specific problem at Fukushima was that there was no electrcial power (AC or DC) to run recombiners or igniters. The containment atmosphere was inerted by excess hydrogen while it remained in containment. Recombiners would not have helped there because there was no power and not enough oxygen in containment. But, outside contaiment where it could mix with atmospheric air the hydrogen was highly flammable. The pressures experienced required pressure relief via venting. Supposedly the Fukushima units had vents -- but I do not know their design details and don't wish to comment where the hydrogen went.