A boiling water reactor has an Achilles heel -- a fatal flaw -- that is invisible under normal operating conditions and most failure scenarios. The flaw has to do with the cooling system.
A boiling water reactor boils water: That's obvious and simple enough. It's a technology that goes back more than a century to the earliest steam engines. As the water boils, it creates a huge amount of pressure -- the pressure that will be used to spin the steam turbine. The boiling water also keeps the reactor core at a safe temperature. When it exits the steam turbine, the steam is cooled and condensed to be reused over and over again in a closed loop. The water is recirculated through the system with electric pumps.
The design's vulnerability comes into play if the electric pumps lose power. Without a fresh supply of water in the boiler, the water continues boiling off, and the water level starts falling. If enough water boils off, the fuel rods are exposed and they overheat. At some point, even with the control rods fully inserted, there is enough heat to melt the nuclear fuel. This is where the term meltdown comes from. Tons of melting uranium flows to the bottom of the pressure vessel. At that point, it's catastrophic. In the worst case, the molten fuel penetrates the pressure vessel gets released into the environment.
Because of this known vulnerability, there is huge redundancy around the pumps and their supply of electricity. There are several sets of redundant pumps, and there are redundant power supplies. Power can come from the power grid. If that fails, there are several layers of backup diesel generators. If they fail, there is a backup battery system. With all of this redundancy, it seems like the vulnerability is completely covered. There is no way for the fatal flaw to ever be exposed.
Unfortunately, shortly after the earthquake, the worst-case scenario unfolded.