MTBF, MTTR and availability calculator
Enter total run time, number of failures and repair time, or paste your failure log. See MTBF, MTTR, failure rate, availability and expected yearly downtime.
Projection assumptions
The calculation runs in your browser; the log you paste is not sent anywhere.
- Inherent availability A
- 99.291%
- MTBF (mean operating time between failures)
- 700 hours
- MTTR (mean time to repair)
- 5 hours
- Failure rate λ
- 0.00143 1/hour
- Failures used
- 6
- Total run time
- 4,200 hours
- Total repair time
- 30 hours
- Expected failures per year
- 12.43 per year
- Expected downtime per year
- 62.13 hours
Reliability R(t)
- Probability of running failure-free for t, R(t)
- 78.7%
- Probability of at least one failure within t
- 21.3%
At t = MTBF, R = 36.8% (e⁻¹): MTBF does not mean more than half of the units will run failure-free for that long.
Assumption: constant failure rate (exponential distribution). It does not hold for equipment that is wearing out or newly commissioned.
System availability (n identical units)
| Series (all must work)System downtime per year: 185.06 hours | 97.8874% |
|---|---|
| Parallel, redundant (at least one must work)System downtime per year: 0 hours | 100% |
Units are assumed to fail independently, switching is perfect and each unit has its own repair crew. In reality common-cause failures and shared repair resources lower the result.
MTBF = run time / failures · MTTR = repair time / failures · A = MTBF / (MTBF + MTTR) · λ = 1 / MTBF · R(t) = e^(−t/MTBF)
Results are only as reliable as the record you enter: an MTBF based on few failures carries wide uncertainty, and the constant failure rate assumption does not suit every piece of equipment. This tool gives a first estimate.
Let's look at which failures in your records could be spotted early, and what that would do for availability.
Request a conversation01
How to use it
A
Enter the summary values (run time, number of failures, repair time) or paste your failure log line by line.
B
Read MTBF, MTTR, failure rate, availability and the expected failures and downtime per year.
C
Try mission time t for reliability, and unit count n for series and redundant parallel system availability.
02
MTBF, MTTR and availability
MTBF (mean time between failures) shows how long a repairable item runs on average between failures: total run time divided by the number of failures. MTTR (mean time to repair) is the average time from the start of a failure until the item is back in service. The failure rate λ is the inverse of MTBF.
Inherent availability A = MTBF / (MTBF + MTTR) counts only downtime from failures and repairs; planned maintenance, waiting for material and time outside shifts are not included. This tool defines MTBF on run time excluding repair; some sources use the time between two failure starts, which makes MTBF larger by the repair time.
MTBF and MTTF are often confused. MTTF (mean time to failure) is the average time to first failure for non-repairable items that are replaced (a lamp, a bearing). MTBF is used for repairable equipment, MTTF for non-repairable parts.
03
The constant failure rate assumption and its limit
The formula R(t) = e^(−t/MTBF) assumes the failure rate does not change over time (exponential distribution). That is a reasonable approximation in the useful-life part of the bathtub curve, for equipment such as electronics where random failures dominate.
It does not hold at the two ends of the bathtub curve: early failures after commissioning (infant mortality) and a failure rate that rises with wear. Wear often dominates in rotating machines and mechanical parts; for those, distributions such as the Weibull, which let the failure rate change over time, are used. Where wear dominates, condition monitoring tells you more than MTBF.
An MTBF computed from few failures is not exact either: two or three failures give only a rough idea. Where possible work with records spanning several years and many failures.
04
Availability of series and parallel systems
In a series arrangement every unit must work for the system to work; availability is the product of the units (A^n) and falls quickly as units are added. Three units at 99% in series give about 97% for the system.
In a redundant parallel arrangement one working unit is enough: 1 − (1 − A)^n. A second unit roughly squares the probability of being down. The calculation assumes independent failures, perfect switching and enough repair resources; common-cause failures reduce the real gain.
FAQ
- What is MTBF and how is it calculated?
- The average operating time between failures of a repairable item: total run time / number of failures. Repair time is not part of the run time; MTTR covers repair.
- What is the difference between MTBF and MTTF?
- MTBF is the mean time between failures of repairable equipment. MTTF is the mean time to first failure of non-repairable parts that are replaced.
- If MTBF is 10,000 hours, will the equipment last 10,000 hours?
- No. MTBF is an average failure interval, not a lifetime. At a constant failure rate the probability that a unit runs failure-free for t = MTBF is only about 36.8%.
- How many hours of downtime a year is 99% availability?
- For continuously running equipment a year has 8,760 hours; 99% availability means about 87.6 hours a year of failure-related downtime. Only failures and repairs are counted.
- What format should the failure log have?
- Each line is either a "start;end" date-time pair or a "run hours;repair hours" pair. Use ; , or tab as the separator. If you use a decimal comma, make the separator ;. A header line is skipped.
- Why does availability fall in a series system?
- In a series system one unit failing stops the whole system; probabilities multiply, so availability falls as units are added. A redundant parallel arrangement reverses the effect.
Get reliability out of your failure data
Let's look at which failures your maintenance records and machine data could have flagged early, and what that means for availability.