Why can an IT room need a HIGHER protection level than any single system inside it — even with redundant servers?

Because of the cumulation effect: if mirrored servers stand in the SAME room, one room-level incident kills both — destroying the redundancy that justified their "normal" rating.

For IT rooms, the protection need is derived (inheritance + maximum principle) from the systems and information in the room. But cumulation effects must be checked:

The classic example from the slide: two mirrored (redundant) servers, each individually rated normal availability — precisely because the other one provides redundancy. If both mirrors stand in the same server room, a single fire, water leak, or power failure in that room takes down both at once. The room therefore has a higher availability requirement than each individual server.

Practical consequences: separate fire compartments or sites for redundant components — or accept that the room must be protected to the higher level.

Tip: Redundancy only counts if the redundant parts don't share a fate. Shared room = shared fate ("common cause failure").