@11: Nice try, but thoroughly and entirely wrong.

In certain disaster scenarios (like a major earthquake, or loss of power from the grid) the reactor automatically SCRAMs with no input from the operators (control rods are inserted into the core to obstruct the fission chain-reaction going on in the core).

In the US, reactors are designed to function primarily from grid power for safety. If external power from the grid is interrupted, that is an emergency and the plant is immediately SCRAMmed. They are not designed to function on their own power. If there were any problems at the plant and it was self-powered, then it would essentially remove your ability to do anything because everything requires electricity to function.

Massive amounts of power are always needed to cool the reactor system. Depending on the type BWR/PWR, there are two or three loops of water respectively. Either way, there are one or two loops of hot water/steam to get to the turbine. The last loop is a massive amount of water (think a river, or an ocean) that is pumped through the equivalent of massive giant water radiators to cool the steam, which then loops back either directly or through another exchanger back to the core to cool the reactor. This takes ENORMOUS water pumps and massive flows of water, which require large amounts of power.

When the reactor SCRAMs, the reaction stops immediately and it will start to cool, and the turbines are no longer going to be giving you power (which isn't designed to self-power the plant, as above). Shutting down a nuclear reactor requires external power (the grid, or several huge diesel generators, essentially small power plants themselves) to drive the massive water pumps to keep the core covered with water, and the water cooled and not boiling away. It takes years for the fuel to cool to room temperature even when removed from the core. It still takes a long time (hours/days) when actively cooled properly for it reach cold shutdown. It may take more than a year for what is left of the cores at Fukushima to reach 'cold shutdown' (the use of the term there may be meaningless, at this point).

Losing power to a nuclear power plant is a very serious problem, and a station blackout is the biggest threat to a nuclear plant, since you are essentially relying 100% on all your backup systems to cool the core. And if for whatever reasons you lose the backup generators (as in, they don't work, they run out of fuel, etc) then you are toast, and you have a meltdown on your hands.

There are some designs, which can self-cool without external power, such as by using huge reservoirs of water propelled by gravity, but we don't build reactors that way in the US. And we don't have battery systems as robust as they have in, say, Japan. But batteries will never be able to power any plant for more than a few hours, so that's still kind of a moot point.