Yet Another Confused MTBF Definition
Just when I thought we had experienced every possible MTBF definition confusion, here’s another.
This one is courtesy the thread concerning the impact to reliability when adding redundancy to a system.
MTBF is used to indicate the cycle time between failures. This value exceeds the MTTF by a margin which is attributable to the time associated with repair. In the case of most components and systems, the repair time is usually very small compared with the operating time and the numerical values of MTTF and MTBF are therefore very close.
I’m giving the writer a pass on the ‘cycle time between failure’ as that is a common misunderstanding. It is the notion the MTBF value exceed the MTTF value with the association of repair time. That one is new to me.
It is generally understood that we use MTBF for repairable systems and MTTF for non-repairable systems. That is fine.
Have any of you run across an MTBF definition that includes the repair time? The most common include total operating time divided by the number of failures.
It is possible, I suppose, to just use total calendar time the system should be operating divided by the number of failures as an estimate. This would assume the repair times are small, yet really? We could simply delay repairs to improve the MTBF metric, as the chance of failure while already down is often smaller than when fully functional and operating.
Ok, once again, what is your take on the writer’s understanding of MTBF? How have you experienced MTBF confusion? Leave a comment and send me a note. Next week an example of MTBF confusion described by a reader.