The UPS industry is all about the reliable devices that can save clients from power interruptions. This fact highlights the importance of the reliability calculations for a UPS application.
The common indicators of reliability are MTBF (mean time between failures) MTTR (mean time to tepair). But such figures are not helpful in building confidence as you can see in the figure below. It shows the different MTBF values for the same availability: A high availability inconjunction with short repair times can be associated with a failure rate that is unsuitable for the application and is not acceptable to the operator of the UPS system.
As pointed out in a recently published article by Newave (“Separating The Wheat From The Chaff”) the MTBF formula has to be used with caution. Applying simple reliability analysis to modular UPS systems (“more parts equals greater risk of failure”) is misleading.
Modularity does not only deliver its more obvious and easily understood benefits like scalability and rapid adaptation to new IT equipment. It also has a less discussed benefit: fault tolerance.
Fault tolerance recognizes that careful control of component quality is the first step toward system reliability, and that continued system operation in the face of component failure is the ultimate reliability tactic.
“The design transition from monolithic to modular is a natural evolution for complex systems because of the advantages it provides in efficiency, flexibility, and reliability. […] As modularity and fault tolerance become the new models for data center design, data center infrastructure must also move in the same direction […] to gain modularity’s benefits for its own efficiency, flexibility, and reliability.”