I always wonder why nuclear power plants aren't designed to handle a loss of the heat sink. At the point at which hydrogen is being made from the water, can't you just remove the water and catch the molten corium into a concrete container until it cools off?
This is a really good question and your question is answered directly here:
Basically the corium is so hot that it will react with concrete producing water vapor and carbon dioxide which may further react into just hydrogen gas and carbon monoxide and that the corium can melt the concrete to about a depth of one meter in one hour.
They are. The heat sink is also the neutron moderator. If you take some of the water away, reactors are designed to slow down by more than the loss of coolant. This is what negative void coefficient means. The only reactors that didn't do this were the Chernobyl type designs (RBMKs). They were known to be a danger by 1955. A similar design was built near Arco, Idaho and cycled to destruction to prove the point.