I have seen descriptions of the Tier-3 datacenters, as having power systems which are N+1, and sized to support "90% N" (versus 100% of "N" for Tier-2).
Does anyone know how to translate that for use with "failover" designs? For example, where one "catcher" UPS provides power to the bypass input of the primary UPS; or where there are cascading transfer switches, and a "swing" genset backs-up the primary genset?
I would guess that you still wouldn't exceed 90% of one UPS. But for the generator, would you want to stay below 90% of Standby load, Prime load, or Continuous load rating? Also, it seems like generators should be sized to handle the maximum load with allowable voltage drop, rather than some percentage of a rating; but I haven't seen this mentioned in datacenter tiers, or in TIA-942.
"Engine generators for Tier III and IV sites" (my situation) "shall not have a limitation on consecutive hours of operation when loaded to 'N' demand."
I note that it doesn't say "loaded to 90% of 'N'". So I'd interpret this to mean that the Continuous, or de-rated Prime rating of the genset, must be greater than total load.
I'll still say that THD or motor load could make this recommendation a moot point.