Addendum the second to
The_Custodian's write-up:
Another huge advantage to using pressurized
water concerns water's unique properties as a
moderator. A moderator slows down and reflects
neutrons back into the
fuel, increasing the probability that they will be absorbed by the fuel, thus causing reactor power to increase. This increase in power, in turn, causes temperatures inside the reactor to rise. As temperatures rise, the density of the water decreases and thus reduces its ability to reflect neutrons back into the
fuel. This causes reactor power to go down. This cycling makes a pressurized water reactor inherently more stable. This property is known as a negative coefficient of reactivity.
For an RBMK reactor, such as
Chernobyl, water was used as a coolant, but
graphite was used as the
moderator. The process begins the same as in a pressurized water reactor, but when the temperature goes up, the ability of
graphite to reflect
neutrons does not change significantly, unlike water. This causes power and temperature to go up, which will eventually cause the coolant to boil off, removing the reactor's ability to cool itself. Since
steam isn't a
moderator, more
neutrons reach the graphite, causing power and temperature to continue rising higher and higher in an unending cycle now that there is nothing cooling the reactor.
see a better explanation at http://www.eng-tips.com/gpviewthread.cfm/qid/16663/pid/466/lev2/19/lev3/64