Either they can't reduce the pressure in the reactors enough to pump water in, or there's a blockage, or there's a big leak. Or all 3.
What I picture happening is a combination of things:
1) There's hot fuel in the reactor because of a drop in water level. Some of the fuel rods have melted.
2) They vent the excess pressure, leading to even more water boiling away (lower pressure means easier boiling).
3) They pump in water, which instantly turns to steam when it contacts the hot fuel rods, increasing the pressure and making it more difficult to pump more water in.
4) Goto 2.
What I haven't heard anyone talk about is: How effective are the control rods when they're embedded in molten fuel rods? Obviously, it's difficult for water to get to the center of a hot blob of metal, but isn't there also danger of some increased level of fission reactions starting up again since the control rods can't do their job?
IOW, under normal circumstances the reactor cools down in a few days once it's been "scrammed". Does that still hold if the fuel rods have melted?
My understanding is that the control rods "suck up" neutrons and stop the fission chain reaction by reducing the number below a self-sustaining level. But the control rods are designed to be a certain distance from the fuel, or in other words, the fuel rods are supposed to be a certain distance from their neighbors, to control the fission reaction. If everything's melted together, there's nothing to control the reaction (with the important caveat being lack of a moderator to slow the neutrons down), so even if the reaction won't be efficient, it can still heat up the fuel even more (more U atoms are closer together than designed), causing more melting, etc., etc.
I haven't seen anyone talk about that, and it bothers me.
In other accidents (TMI, the Fermi breeder outside Detroit), the core partially melted, so there must be some way (in principle) to prevent it from running away or staying hot for decades. But in those cases, 1) they weren't BWRs, and 2) they didn't suffer loss of coolant anywhere near as severe as some of these TEPCO reactors apparently have. How much of that history transfers here?
I don't envy the folks working there. :-(
Keep an eye on the prevailing winds, all of you West-Coasters.... :-(
Cheers,
Scott.