Data center redundancy: N, N+1 and 2N without a table

Redundancy means having more capacity or more paths than the minimum required, so the failure of one component does not interrupt operation. It costs space, energy and money, so the right question is how much protection the operation really needs.

N, N+1 and 2N

N is exactly the capacity needed for the load. N+1 adds one extra unit, so one can fail or be taken out for maintenance without impact. 2N duplicates the entire system in two independent paths, each able to carry the load alone.

Redundancy is not only quantity

Two units that depend on the same switchboard, the same control or the same piping fail together. What protects is the independence of the paths. That is why the analysis starts from failure modes, not from a count of equipment.

Maintenance is also a scenario

A system must remain protected while part of it is shut down for maintenance. This is one of the criteria that most separates a good design from one that only looks redundant on the diagram.

About classifications

Classification terms such as the Tier levels describe design requirements, and a facility can only call itself certified if it actually went through the corresponding certification. This text only covers the concepts.

All articles

Want to discuss this topic in your project?

Talk to an engineer