Events at Fukushima have raised significant concerns, both practical and political, for the safety of any nuclear power plant. As advocates of Thorium and ADSRs we have to address the question of how such a system would have fared under similar circumstances - major earthquake, unprecedentedly large tsunami, prolonged loss of external power, and extremely difficult operating conditions - in an honest way. It is too early for a full analysis, but we have to start thinking about the issue. We may believe that ADSRs are intrinsically safer than conventional reactors, but this has to be demonstrated in this particular situation.
Problems at Fukushina had nothing to do with criticality but were due to fuel melt-down. The reactors were easily switched off, and indeed some were already shut down before the incident. However they continued to generate heat through their residual radioactivity. The rate for this is usually quoted as 7% of the full power. An ADSR would be slightly different but not greatly so: there are fewer actinides in the fuel but the fission product mix is pretty much the same. It would be useful to do a full study of this, and the figure may be a bit lower due to the smaller contribution from spontaneous fission, but it cannot be vastly different.
So the fuel will still generate heat. At Fukushima this caused a build up in temperature as the primary cooling circuit had lost power. However in an ADSR the coolant is molten lead, or lead-bismuth, rather than pressurised water. This carries away the heat by convection. In some designs convection is assisted by electrical pumps, however these are only needed when the reactor is at full power. Even in such designs the unasssited power of the convective heating is sufficient to take away the 7% residual power.
In this scenario the water/steam secondary cooling will not be operational, so the lead is going to get hot. However there is quite a safety margin: Lead does not boil until 1740 degrees C, and as the temperature rises the cooling effects of conduction and convection will increase. A large vat of very hot lead is, of course, dangerous, and the design will have to ensure that it is robust and stable against leakages, etc. But it is not as dangerous as a pressure vessel full of superheated steam, which escapes though leaks and dissociates into an explosive mixture.
Further problems occured due to the spent fuel rods kept in cooling ponds, in which the water evaporated so that their temperature increased and they fractured. The rods were apparently kept in the reactor for no reason other than it was a convenient place.This problem would (almost certainly) not have arisen in an ADSR system, as the fuel rods stay in the reactor for 5-10 years rather than 5-10 months. In a thorium fuelled system there will be far fewer fuel rods to handle, and such expediants should not become necessary.
Fukushima was affected by an extroadinary disaster, and it is a tribute to the workers on site that the damage has been, despite the press coverage and global public panic, so limited.It is fair to claim that ADSR systems (assuming the design was sensible) would not have undergone such leakages, but this is due to the secondary features of the design and not to their primary 'accelerator driven' nature.