Equipment, Fleet and Supplier Operations
Reading breakdowns with MTBF and MTTR
Reading how often equipment fails and how long it stops each time with two averages, and choosing priorities from them.
- Level:
- Practitioner
- Who it is for:
- Maintenance and calibration lead, Plant manager
Definition
MTBF shows how long a repairable piece of equipment runs, on average, between two failures. NIST's reliability handbook estimates this average, under a constant repair rate (HPP/exponential) model, as total operating time divided by the number of failures. MTTR is the average time it takes to get equipment running again after a failure. The two are read together: one tells you how often you stop, the other how long.
When does it help?
- You are choosing which equipment gets priority in the maintenance plan.
- You want to see whether failures have become less frequent after an improvement.
- In TPM work you need to separate whether the downtime loss comes from how often failures happen or how long repairs take.
When does it not help?
- When there were only one or two failures in the period; the average misleads, so read the failure list one by one.
- For a part that is replaced rather than repaired when it fails; there the mean time to failure is used.
Step-by-step method
Define a failure
Write which stops count as failures; planned maintenance, setup and waiting for material are kept separate.
Add up the operating time
Add up the time the equipment actually ran in the period; leave out planned stops.
Calculate MTBF
Divide the total operating time by the number of failures.
Calculate MTTR
Add up the time from stop to running again for each failure and divide by the number of repairs.
Break down the repair time
If MTTR is long, split the time into waiting, finding the fault, getting the part and repairing.
Follow the trend
Recalculate every period with the same definition; decide from the trend, not from one period.
Example scenario
The monthly record of two presses
In a plastic injection workshop the monthly records of two presses are compared. The first press runs 400 hours, fails four times and spends 10 hours in repair in total. For this press MTBF is 100 hours and MTTR is two and a half hours. The second press fails rarely, but every repair waits for a part; the team first sorts out the spare parts list.
A fictional example; the people, company and numbers are not real.
Common mistakes
- Counting planned maintenance as failures; MTBF looks shorter than it is.
- Using calendar time instead of operating time; hours when the equipment was stopped count as running.
- Taking MTBF for the life of the equipment; it is an average interval, not a date of failure.
- Looking only at MTBF; rare but long failures only show up in MTTR.
Terms
Sources
- NIST/SEMATECH e-Handbook of Statistical Methods, 8.4.5.1 Constant repair rate (HPP/exponential) model, National Institute of Standards and Technology (NIST) (accessed 16 November 2026)
- Total Productive Maintenance, Lean Enterprise Institute (accessed 16 November 2026)