As Japan struggles with its worst nuclear disaster in decades, I thought it might be useful to provide a few more technical details about the reactors at Fukushima, and the most likely scenario for what is happening there.

The plant

The Fukushima Daiichi power station operates six boiling water reactors all completed during the 1970s. Details of the reactors vary but the concept is the same: the core consists of a pill-shaped pressure vessel filled with several hundred fuel assemblies. Each fuel assembly is in turn filled with about a hundred fuel rods. A fuel rod is a long, narrow tube of zirconium alloy filled with pellets of uranium which has been enriched to around 3-5% of the energy-producing isotope U-235. (In the case of unit 3, Plutonium-239 is also an active part of the fuel).

When enough fuel is brought together at the core, a chain reaction begins that generates heat, and ultimately power. The core of a modern reactor can hum along for a year or more before the fuel needs to be changed.

The accident

The key to the crisis is water. In addition to the uranium fuel rods, the fuel assemblies have channels which carry highly purified water between the fuel. The water acts as both a moderator for the nuclear reactions and a coolant for the reactor core. On top of it all, it makes the electricity: as it is heated by the reactor, it turns into steam that drives the power turbines. Once the water passes through the turbines it is cooled and re-injected into the core to do it all again.

It all goes great unless the water stops flowing, and that’s exactly what it appears has happened in the wake of a massive magnitude 9.0 earthquake that shook the region on 11 March. Diesel generators designed to keep feeding water to Fukushima Unit 1 apparently shutdown about an hour after the quake. Yesterday, the water supply to Unit 3 was interrupted. In both cases, the cores began to heat up.

Meltdown

Immediately after the earthquake, the Fukushima reactors, and many others, went into an automatic shutdown mode. Special rods of neutron-absorbing material, known as control rods, were inserted between the fuel assemblies, halting the power-producing nuclear reactions. But power-producing reactions are not the only ones happening at the core: as nuclear fuel burns it creates new elements that themselves generate a great deal of heat through their radioactive decay. A small but significant amount of the core’s heat is generated by these elements, and there is no way to turn them off.

So, without emergency cooling, the temperature at the core of both reactors began to rise. As it did, what water that remained began to boil off, increasing the pressure inside the pellet-shape pod.

When temperatures reached around a thousand degrees Celsius, the zirconium alloy holding the fuel pellets probably began to melt or split apart. As it did, it reacted with the steam and created hydrogen gas, which is highly volatile.

Operators may or may not have known what was happening when they decided to release some of the pressure from Unit 1 on Saturday. The hydrogen apparently caused a massive explosion which blew the roof off of the fuel hall, though the reactor’s primary containment vessel appears to have remained intact (see diagram, from NEI).

If, as it appears, the zirconium came apart, then some of the uranium and plutonium pellets in the fuel rods may have become loose or melted and sunk to the bottom of the pressure vessel. In that case, the cores of units 1 and 3 are now a volatile test tube filled with radioactive fuel, melted zirconium and water.

The real danger is the fuel. If enough fuel gathers at the bottom of the reactor, it could burn through the concrete containment vessel. In a worse case scenario, the fuel could again gather to form a critical mass outside the fuel assembly. The loose fuel would restart the power-producing reactions, but in a completely uncontrolled way. This, if it happened, would lead to a full-scale nuclear meltdown.

Emergency procedures

To prevent such a catastrophe, plant operators have decided to swamp both units with seawater. The decision is not made easily: the impurities in the seawater will contaminate the reactor cores, effectively ruining them. But it should allow the temperature inside to again drop, preventing further melting of the zirconium rods and the fuel elements. On top of this, the reactors are being filled with boronic boric acid. Boron is an excellent neutron absorber, and it should hamper nuclear reactions, even if fuel pellets are loose inside the core.

The reactors are also venting excess steam, reducing the pressure inside.

What’s next?

It is very difficult to say. In the best case scenario, the fuel will be sufficiently cooled to stabilize the situation. But its important to understand that there’s no way to “shut off” the residual heat inside these reactors. Unless the fuel can be moved, which seems unlikely for now, they will need to be actively cooled for weeks in order to prevent a crisis. (Although the half-life radionuclides in the fuel mean that cooling will become less urgent with time). Even after the immediate crisis is past, decommissioning the reactors could take decades.

The latest update from World Nuclear News indicates progress in the injection of seawater into unit 3.

For full coverage of the Fukushima disaster, go to Nature’s news special.

Updated 14 March at 12:43 PM GMT