MTBF (mean time between failures) is the average time a repairable asset runs between one failure and the next. It is a reliability measure: the higher the MTBF, the longer the asset typically runs before it breaks down. For a repairable asset (or a group of identical ones) over a period:
How to calculate MTBF: a worked example for a group of assets
MTBF is most useful across several identical assets, because one machine rarely fails often enough to give a stable number. Suppose a site runs four identical air compressors and wants their MTBF for the last six months (illustrative figures):
| Compressor | Operating hours | Failures (from work orders) |
|---|---|---|
| C-01 | 2,100 | 1 |
| C-02 | 1,950 | 2 |
| C-03 | 2,050 | 0 |
| C-04 | 1,900 | 1 |
| Total | 8,000 | 4 |
- Add up operating time, not calendar time: 2,100 + 1,950 + 2,050 + 1,900 = 8,000 hours. Exclude time the compressors were switched off or already down for repair.
- Count failures using one consistent definition — here, any unplanned stop that needed a corrective work order. Planned PM stops do not count.
- Divide: MTBF = 8,000 ÷ 4 = 2,000 hours.
- Use it: each compressor runs about 330 hours a month, so on average each one can be expected to run roughly six months between failures — and, across the four, about one failure every six weeks. That is a useful input when setting the PM interval and deciding which spares to stock.
Repeat the calculation each period and watch the trend. If the next six months come in at 8,000 hours and 2 failures, MTBF has doubled to 4,000 hours — evidence that the maintenance changes are working.
Why it matters
A rising MTBF means maintenance is working; a falling one is an early warning. It helps set PM intervals and compare models or suppliers. Combined with MTTR, it gives inherent availability: MTBF ÷ (MTBF + MTTR). With MTBF 300 h and MTTR 3 h, availability = 300 ÷ 303 ≈ 99.0%.
Cautions
- Use operating time, not calendar time, and define "failure" consistently.
- Small numbers of failures make MTBF volatile — look at trends over time.
- For non-repairable items, the related measure is MTTF (mean time to failure).
MTBF vs MTTR
MTBF tells you how often an asset fails; MTTR (mean time to repair) tells you how long it takes to get it working again. A good maintenance programme pushes MTBF up and MTTR down, and the two together give availability.
