I was thinking, the main source of problems for reactors is cooling down the reaction in an emergency (I mean, in terms of the neutrons as opposed to heat). Chernobyl was an example of where that didn't happen - the rods overheated, bent out of shape and wouldn't fit back in as I'm aware of things. But I seem to remember the first nuclear pile had guys standing by with pales of ?boron? loaded water that they'd pour onto the reactor if it started to overheat. Could you not position some kind of big water tower filled with such a solution over the core of a reactor and have it empty out into it should things start going wrong? You could have a similarly sized tank underneath the core to dump the normal coolant into. The liquid moderator would obviously benefit from being as absorptive as possible and not breaking down into an explosive hydogen & oxygen mixture - and also acting normally in the normal coolant loop if at all possible. I can think of two reasons why this wouldn't happen. 1.) money 2.) there isn't a liquid form of moderator that will soak up enough of the neutron flux to stop the reaction (but just slowing it would be enough I would have thought, anything is better than letting it runaway). Have I missed some point here or do they just rely on the solid moderators to do their job 100% of the time?