The most cogent explanation I have seen of the nuclear incident at the Fukushima Daiichi power plant is by Ken Haapala. You can download the whole thing here. An excerpt:
All six of the reactors at the Fukushima Daiichi power plant are Generation II boiling water reactors of GE design. The oldest is about 40 years old. Unlike the Chernobyl reactor, these reactors have two strong features designed to contain any accident or mishap, in addition to the thick steel pressure vessel. The primary containment is a concrete and steel structure around the pressure vessel and the secondary containment thick poured concrete around the primary containment. A third structure is a thin shell designed to keep out weather.
In a nuclear plant, the heat from the nuclear reaction is used to create steam that drives turbines generating electricity. Rods made of neutron absorbing material, boron carbide at Fukushima, are placed among the fuel rods to control the rate of the nuclear reaction and the amount of heat produced. The nuclear reaction is shut down by full insertion of these rods. However, the continuing radioactive decay of the products of the nuclear reaction in fuel rods still gives off heat even after the reactor is shut down. For that reason the fuel assemblies must be immersed in water or some other coolant. If the coolant is lost, the fuel rods may reach temperatures so high they melt.
Adding to the difficulty, if the fuel rods are exposed to air / steam, the zirconium that coats the rods combines so strongly with oxygen, that it can strip oxygen from steam, forming free hydrogen. Hydrogen is so chemically reactive that it is not found in the atmosphere. It burns rapidly in the atmosphere, giving the impression of an explosion.
Three reactors were closed for maintenance and refueling, but contained fuel rods or fuel rods that were kept in storage pools. Retrospectively, a design flaw emerges: these cooling pools were built above the containment structures.
The US Geological Survey reports that the earthquake that struck Japan last Friday is the strongest ever recorded with modern instrumentation to hit Japan and the fourth strongest since 1900. The Richter scale used to measure earthquakes is a base ten logarithmic scale. Thus a 9 point earthquake is 10 times stronger than an 8 point earthquake. The strongest earthquake recorded, 9.5, hit Chile in 1960 (which also caused tsunami waves to hit Japan). …
Apparently, the earthquake destroyed the electric grid to which the nuclear plant is connected, but did not damage the plant. With the loss of the grid, the control rods were fully inserted into the fuel assemblies stopping the nuclear reactions in the active reactors. The backup diesel engines to generate emergency electrical power turned on providing the necessary circulation of water needed to keep the fuel rods from overheating.
About an hour later, the tsunami overwhelmed the sea wall damaging or destroying the diesel backup for the electricity to the pumps providing cooling water. A second backup of batteries may or may not have worked, but if they did the batteries were quickly drained. The cooling water overheated into steam and some of the fuel rods likely melted. The reaction between the hot fuel rods and the steam probably created hydrogen within the containment chambers.
When plant workers opened valves to release pressure from the chambers, the hydrogen escaped which burned so rapidly in the atmosphere that it virtually exploded. This resulted in damage to the outer shells, so displayed in photographs, but immaterial to the structural strength of the two containment structures. How the burning of the hydrogen gas affected the cooling pools on top of the containment structures is not clear. With the release of pressure, some radioactive gas escaped into the atmosphere.
The earthquake and tsunami effectively destroyed regular communications between plant operators and the corporate and government leaders. This probably delayed the decision to flood the overheating reactors with sea water which would effectively destroy them for further use.
What appears to be the most prolonged problem were the pools used for cooling the fuel assemblies. While the conditions in the reactors were being brought under control it appears that water in some of the pools overheated and evaporated or boiled off. The fuel rod assemblies may have been exposed to air / steam resulting in the release of hydrogen and radioactive gases. This may have caused the spike in radioactive readings later in the week.
As of this writing, the latest reports are that the reactors are cautiously under control and that the temperatures of the cooling pools are being stabilized. Much of the instrumentation failed, but, based on theory, it appears that a few percent of each active reactor cores had damage.
Although not confirmed, if the above fairly reflects what took place, then it can be a basis to evaluate some of the more excited claims in the press and by anti-nuclear groups and politicians.
1) The reactors withstood the earthquake significantly above designed strength.
2) Although the primary source of electricity failed, initially the backup systems worked as required.
3) The reactors withstood the tsunami which was above planned height.
4) The tsunami disabled the diesel generated backup of electricity and the battery backup was either disabled or inadequate.
5) The destruction of the standard communications between the plant operators and corporate and national leaders lead to a slow decision to flood the active reactors with sea water, which destroyed them.
6) The observed “explosions” were chemical, probably burning of free hydrogen.
7) Some meltdown of active reactors probably occurred.
8) Some of the cooling pools overheated, probably exposing the fuel rods and giving off hydrogen and radioactive gases.
9) Thus far, except for the immediate area around the reactors, the radioactivity released has been insignificant.
From this natural disaster, we can learn that properly built nuclear plants can withstand powerful earthquakes and tsunamis. But backup cooling systems and any on-site cooling pools must be protected from any after effects of an earthquake. Cooling pools should be separated from the reactors. Also, solid communications must be available in spite of the scope of the natural disaster.
UPDATE: From the comments:
There are several errors and omissions, including the problems with the SFP’s in Fukushima Dai’ichi #’s 5 & 6 and also Fukushima Daini #1, 2 & 4.
Another omission is that the containment structure in Dai’ichi #2 is indeed cracked in the area of the torus suppression pool underneath the reactor.
Latest update from the MIT Nuclear Engineering department is two hours ago at.