Key takeaways
Short answer: MTBF measures how often equipment fails between repairs. MTTR measures how long each repair takes. Together they determine Availability: Availability = MTBF / (MTBF + MTTR). Improving MTTR is usually faster (spares, procedures, technician access); improving MTBF takes longer (root cause work, design changes). Plants chasing easy wins focus on MTTR; plants chasing big wins focus on MTBF. See also MTTF vs MTBF vs MTTR.
Mean Time Between Failures is the average operating time an asset runs before failing.
MTBF = Total operating time / Number of failures
If a pump ran for 10,000 hours and failed 5 times, MTBF = 2,000 hours per failure. It is a reliability metric — it measures how trustworthy the asset is between fixes.
Mean Time To Repair is the average time from failure to back-in-service.
MTTR = Total repair time / Number of failures
If those 5 pump failures cost 30 hours of total repair time, MTTR = 6 hours per failure. It is a maintainability metric — it measures how fast your team can recover when something does fail.
The classical Availability formula:
Availability = MTBF / (MTBF + MTTR)
Using the pump example: 2000 / (2000 + 6) = 99.7%. That looks great until you realize the same Availability can hide very different operations:
Same number, very different reality. MTBF and MTTR each carry signal that Availability alone hides.
Improving MTTR usually means:
These are all internal maintenance improvements. Most plants can cut MTTR 20-30% in 6-12 months with focused work.
Improving MTBF means making the asset itself more reliable. That requires:
All of this is slower. Engineering work, OEM coordination, sometimes capital. MTBF doubles are typically a 1-3 year program.
The ratio matters more than the absolute numbers. A rough rule of thumb for industrial equipment:
If your ratio is below 50, attacking MTTR alone will not be enough. The asset is failing too often.
1. Reporting Availability without the underlying MTBF and MTTR. The breakdown is the signal; the aggregate hides it.
2. Setting MTTR targets without spares discipline. Fast repair is impossible without the right part on hand.
3. Treating MTBF as fixed. "This kind of equipment fails every X hours" is a vendor average, not your asset's truth. Your data is what matters.
4. Excluding planned maintenance from MTTR. If a PM event takes the asset down, it counts toward unavailability even if it does not count toward MTTR.
A modern CMMS calculates MTBF and MTTR per asset from the work order history. The data quality depends on every fix having a WO with start/end times and a clear failure-vs-PM tag. If the WO data is sloppy, the reliability data is fiction.
Fabrico's CMMS reports MTBF and MTTR per asset with trend lines, surfaces the worst-MTBF assets for focused improvement, and ties downtime events from the OEE module directly to MTTR — keeping the reliability math honest.
See how Fabrico captures this automatically — explore OEE for manufacturing or book a demo.
No. MTBF applies to repairable assets — it is the time between failures. MTTF (Mean Time To Failure) applies to non-repairable assets that are replaced rather than fixed.
Yes, in the standard definition. MTTR is from failure detected to back-in-service, including diagnosis, parts wait, repair, and verification.
Highly equipment-dependent. For a CNC machine, an MTTR of 1-3 hours is good. For a process pump, 30-60 minutes. The target should be set per asset class, not as a single number.
Sometimes. Operating discipline, lubrication, basic care often improve MTBF without engineering changes. For larger gains, engineering work is usually required.
Both. Per-asset MTBF identifies bad actors. Per-asset-class MTBF identifies systemic problems (e.g., all bearings of type X failing too soon).
Login
Blog
