- AND gate versus OR gate
- Logic elements in a fault tree. An AND gate passes a fault only when all its inputs fail together, which is how genuine redundancy appears; an OR gate passes a fault when any single input fails. Fragile systems hide OR gates where the operator believed AND gates existed.
- Automatic transfer switch
- The device that moves load from a failed source to an alternate one without human intervention. Its transfer speed and whether it verifies the alternate source is healthy before transferring appear repeatedly in outage narratives; a slow transfer strands stored energy, and a transfer to an already-faulted source turns one failure into two.
- Common-cause failure
- A single underlying condition that disables multiple supposedly independent paths at once, such as two power paths sharing a conduit, a cooling plant, a control network, or a fuel system. Common-cause failures are the usual reason a nominally doubled topology collapses under a single fault.
- Detection, diagnosis, recovery
- The three functional phases of an outage timeline. Detection measures whether instruments or customers found the fault first; diagnosis measures operator competence under a running clock, with reversed actions signaling confusion; recovery measures the physical architecture's ability to be brought back online. Their ratios diagnose different weaknesses.
- Fault tree
- A top-down logical map of an undesired outcome, such as loss of critical load, that traces downward through AND and OR gates to the combinations of component failures and missing barriers that could produce it. Reading an outage as a fault tree separates the trigger from the barriers that failed to contain it.
- Redundancy notation (N, N+1, 2N)
- Shorthand for installed capacity relative to need. N carries the full load with no spare; N+1 adds one extra unit so any single unit can fail or be serviced; 2N provides two complete independent systems. The notation describes capacity and unit count but cannot express whether the paths are truly independent.
- Thermal runaway
- A self-reinforcing loss of cooling in which a dense hall's inlet temperatures climb through safe limits within minutes, forcing equipment to throttle or trip on over-temperature. It is the classic cause of a recovery phase far longer than the electrical event, because cooling must be restored and heat purged before load can be staged back.
- Trigger, propagation, and barrier
- The three parts of any outage. The trigger is the first component failure; propagation is the path by which it spread, often through hidden shared couplings; the missing barrier is the redundancy that should have halted propagation and was absent, depleted, defeated, mis-operated, or out for maintenance.