In 1979 the world experienced its first severe nuclear plant accident at Three Miles Island, when TMI-2 underwent a LOCA with devastating consequences for the reactor core. About 50 % of it melted and some 20 tonnes of core material ended up on the reactor vessel's floor.
However, the reactor had effectively been shut down early on in the crisis, by SCRAMming the control rods, so no more fission took place and 'only' fission product decay heat had to be dealt with.
Imagine now a LOCA where part or all of the core is no longer water cooled and the SCRAM fails (the control rod lowering mechanism fails, gets blocked or breaks).
How bad could this get, considering nuclear fission would now continue but bearing in mind also that in these PWR reactors the primary coolant is also the neutron moderator and absence of the moderator must also affect fission yield?