Maintenance KPIs: MTBF, MTTR, availability
The three reliability indicators from your operating hours, your stoppages and their duration.
Over a given period, for one item of equipment or one line.
The definition used
The MTBF is the uptime divided by the number of stoppages, and not the length of the period divided by that number. The distinction matters as soon as downtime becomes significant.
The MTTR covers detection, the wait for parts and resources, the intervention and the return to service. Reducing it to the repair time gives a flattering and useless figure: most of the time lost is almost always waiting.
What these indicators do not tell you
They describe the past and assume stoppages are comparable to one another. A single long, exceptional stoppage crushes the average and makes the indicator misleading. Always look at the distribution before drawing conclusions from the average. See reducing unplanned stoppages.
Definition
MTBF (mean time between failures) and MTTR (mean time to repair) are two indicators of an equipment's reliability and maintainability. The first measures the average time the equipment runs without failing, the second the average time needed to restore it. Combined, they give availability, the share of time the equipment is actually fit to produce.
Formula
MTBF = uptime / number of stoppages. MTTR = total repair time / number of repairs. Availability = MTBF / (MTBF + MTTR). MTBF and MTTR are expressed in hours; availability is a dimensionless ratio, often given as a percentage.
How to use this tool
- Choose the observation period and the equipment, then record the total effective operating time over that period.
- Count the number of failure stoppages and divide the uptime by this number to obtain the MTBF.
- Add up the repair durations, then divide by the number of repairs to obtain the MTTR.
- Compute availability with MTBF / (MTBF + MTTR).
- Compare over time rather than in absolute value: it is the trend that reveals declining reliability or improving maintenance.
Worked example
Over one year, an equipment totals 8,000 hours of uptime and undergoes 5 failure stoppages: MTBF = 8,000 / 5 = 1,600 hours. The repairs add up to 40 hours over these 5 interventions: MTTR = 40 / 5 = 8 hours. Availability is 1,600 / (1,600 + 8) = 0.995, that is 99.5 percent.
References
- EN 13306, maintenance terminology: definitions of reliability, maintainability and availability.
Frequently asked questions
Is the MTBF the period divided by the number of stoppages?
No. It is the uptime, that is the observed period minus the downtime, divided by the number of stoppages. The difference stays small as long as the downtime is marginal, and becomes misleading as soon as it becomes significant.
Should planned stoppages be counted?
Not in the MTBF. Otherwise preventive maintenance degrades the very indicator it is meant to improve, and a site that maintains its equipment well shows a lower MTBF than a site that waits for the breakdown.
Should the MTTR include waiting for parts?
That is the choice made here: from detection to return to service. An MTTR limited to hands-on tool time gives a flattering and useless figure: most of the time lost is almost always spent waiting.
Is 98% availability good?
Out of context it means nothing: everything depends on the criticality of the equipment, the operating regime and the cost of an hour of downtime. Comparing a site with itself over time remains sounder than comparing two sites with each other.
The other tools