Hydrogen Explosions

The fuel rods are tubes of zirconium, which contain uranium pellets. An odd feature of zirconium is that when exposed to boiling water, it disassociate the boiling water and subsequently creates lots of hydrogen.

As to why zirconium is preferred in a wet, hot environment is anyone's guess.

Water starts to decompose to hydrogen and oxygen at around 2,000C. At 3,000C about 50% has decomposed. As long as the surrounding temperature remains high, then the hydrogen and oxygen will not recombine. Once they are vented to a lower temperature, then recombination takes with the release of much energy - an explosion. All that energy that was used to decompose the water now comes back with a vengeance. Hydrogen is the smallest molecule and can pass right through metals, particularly when hot. Perhaps in nuclear reactors, the hydrogen escapes from the overheated reactor core, causing it to explode within the containment vessel. Any nuclear engineers out there? Please advise.

As regards magnesium, you are correct, when hot from burning in oxygen, it will also react with water, reducing it to hydrogen, which then also burns. So adding water to extinguish burning magnesium can only be accomplished by adding a large excess.

At around 1200 c zirconium takes oxigen from water, forming zirconium oxide. When wented out, the steam with hidrogen in it, meets the oxigen from air. A litle spark or static charges or a catalist and it explodes.

The hydrogen explosions are dwarfed in size by the "hot air" explosion in the popular press.

Australian papers are full of nuclear horror stories, unscaled full color graphs, lists of symptoms etc. One report showed hundreds of people lining up for a radiation screening, pictures of the scary looking test room, space-suited workers and quotes from scared mothers. Truly horrific (think of the children). It would have been even worse if any radiation had actually been detected.

The only things missing, in the coverage, are actual measurements and information that would allow people to keep this disaster in proportion.

The replies explaining the zirc-oxide reaction are correct. Hydrogen release after a loss of coolant accident (or LOCA) is a well-studied scenario in nuclear safety. Unfortunately, hydrogen and oxygen (or air) are flammable over a large range of concentrations, and even a relatively slow burn can cause a pressure buildup in the containment building that is too much for the building to contain unless there is venting or condensation. Large energetic explosions or detonations can be very destructive, as we have seen in Japan. Some nuclear plants have ignitors as a safety feature to ignite the hydrogen before it builds up to dangerous levels. (Small pressure increases spread out over a long period of time are less dangerous than one large pressure increase in a short time, the steam will have time to condense or vent safely and the pressure returns to normal before the next mini-burn). Some reactors are using catalytic recombiners, which do the same thing but without the flame. These features are only useful if the release of hydrogen is slow, however. Catalytic recombiners have their limits, they can only process so much H2-air-steam at a time (imagine a steel duct about 2' x 2' and 4-10' long. The gases enter at the bottom of the duct and the heat of recombination creates a chimney effect in the duct which pulls in new gases at the bottom).

PWR reactor containment buildings are extra strong to withstand the pressure buildup from steam release and potential hydrogen deflagrations. BWRs vent the steam/hydrogen to a pool of water to condense the gases and reduce the pressure increase. Many CANDU reactors are connected to a vacuum building which would rapidly draw in any steam from an accident and keep the containment building sub-atmospheric.