MTBF and MTTR calculator: free formula and worked example

Total operating time

Hours the machine actually ran in the period. Leave out the hours it was down for repair.

Number of failures

Breakdowns in the same period that stopped the machine or needed a repair.

Total repair time

Hours from each breakdown to the machine running again, added up over all failures. Include diagnosis, waiting for parts, repair and the test run, unless you choose a narrower rule and say so.

Number of failures

The same count used for MTBF.

MTBF

From the first formula, in hours.

MTTR

From the second formula, in hours.

With zero failures, MTBF cannot be calculated yet. That is good news for the machine, not a number. Keep counting operating hours until the first failure.

MTBF tells you how often a machine lets you down. MTTR tells you how long it hurts each time. Read them together. A machine with high MTBF and high MTTR fails rarely but stops the line for a long time, so spares and repair instructions matter most. A machine with low MTBF and low MTTR fails often but is quick to fix, so root cause and preventive maintenance matter most.

Inherent availability is a ceiling, not your real availability. It ignores planned maintenance, setups and waiting for an operator, so the real figure on the shop floor is usually lower.

Track the trend for each machine, month on month, with the same rules. One month with two failures is a weak guide on its own.

The arithmetic is simple. The errors come from what you count.

• Using calendar time instead of operating time. If the machine runs one shift a day, a month of calendar hours inflates MTBF three times over.
• Leaving repair hours inside operating time. Take them out. A machine is not operating while it is being repaired.
• Counting planned stops as failures. PMs, setups and changeovers are not failures. Count them separately.
• Changing when the repair clock starts. Decide whether MTTR starts when the machine stops or when the technician arrives, and keep that rule every month.
• Using MTBF for parts you throw away. For parts you replace rather than repair, mean time to failure (MTTF) is the right measure.

You get MTBF and MTTR for every machine without a monthly spreadsheet. When your team logs a breakdown in MaintenanceIQ with a start time and a resolution, MTBF, MTTR and downtime follow for that asset, along with a bad-actor ranking of the machines costing you most. See reliability analytics.

• Repairable systems, non-repairable populations and lifetime distribution models (opens in a new tab), NIST/SEMATECH e-Handbook of Statistical Methods. What makes a system repairable, and why repairable and non-repairable items are measured differently.
• Homogeneous Poisson Process (HPP) (opens in a new tab), NIST/SEMATECH e-Handbook of Statistical Methods. MTBF as the reciprocal of the repair rate for a repairable system.