With a machine protection system similar to the Large Hadron Collider (LHC), the Future Circular Collider (FCC) would require up to three turns’ to dump the beam synchronously after detection of a failure. The reaction time of the machine protection system can be reduced by several strategies. The time for failure detection can become shorter with faster hardware and beam monitors, e.g. using diamond detectors as fast beam loss monitors. Communication time for the interlock system to the beam dumping system can be reduced by using a straight signal path instead of going through the arc. More than one beam-free abort gap can shorten the time required for synchronization. Different operational and failure scenarios are classified according to beam lifetime, i.e. the speed of the onset and increase of induced beam losses. We put emphasis on so-called ultrafast failures including crab-cavity failures, fast failures such as magnet failures at high beta function positions or with short time constants of field decay, and slow failures. A list is presented, summarizing the critical failure modes and proposing potential mitigation strategies.