MTTR Explained: Formula, Example & Reduction Guide

Learning outcomes

After completing this guide, you should be able to:

• calculate MTTR from a consistently defined repair-time record;
• separate active repair work from access, permit, diagnosis, logistics and release delays;
• identify the delay category that deserves corrective action;
• recognize changes that make the KPI look better without improving restoration performance; and
• interpret MTTR with MTBF and availability rather than using it alone.

MTTR formula

MTTR = Total included repair or restoration time ÷ Number of completed repairs.

If four qualifying repairs take a combined 20 hours under the same measurement rule, MTTR is 5 hours. State the unit, reporting period, asset population and failure definition beside the result. A number without these boundaries is difficult to compare or audit.

Define when the clock starts and stops

Organizations use the letters in MTTR to mean “repair,” “restore,” “recovery” or “resolution.” These are not automatically equivalent. Before trending the KPI, document exactly what your clock includes.

• Start: equipment trip, operator notification, work-order creation or technician acceptance.
• Stop: repair physically complete, functional test passed, production release granted or stable operation confirmed.
• Included delays: decide whether permits, isolation, access, diagnosis, spare-part waiting and operational release remain inside the clock.
• Excluded events: define planned maintenance, minor stops, duplicate work orders, observation-only calls and incomplete repairs.

Break each restoration into useful time categories

A single duration shows how long restoration took, but not why. Record timestamps or durations for the following stages where practical:

1. Detection and response: failure occurs, alarm is recognized and maintenance is notified.
2. Make safe and gain access: approved PTW/LOTO, risk controls, production clearance, cooling, draining, depressurization or access preparation.
3. Diagnosis: evidence gathering, drawings review, tests and confirmation of the failed function or component.
4. Logistics: locating spares, tools, lifting equipment, contractor support or OEM information.
5. Active repair: removal, correction, installation, adjustment and reassembly.
6. Test and release: inspection, functional testing, safeguards restoration, housekeeping, documentation and operational handback.

This breakdown prevents every long event from being labelled “slow maintenance.” For example, a two-hour repair inside an eight-hour outage may show that six hours were spent waiting for an approved spare or safe access. The corrective action should address the dominant verified delay, not simply pressure technicians to work faster.

Worked maintenance example

A conveyor has three completed corrective repairs in one month. Total restoration times are 2, 4 and 6 hours. Total included time is 12 hours, so MTTR is 12 ÷ 3 = 4 hours.

The delay review shows that the longest event contained 1 hour of diagnosis, 3 hours waiting for a drive belt, 1.5 hours of active replacement and 0.5 hour of testing and release. The evidence points to spare availability as the largest controllable delay. After the correct critical spare is stocked and its storage location is added to the job plan, three later comparable repairs take 1.5, 2 and 2.5 hours. The new MTTR is 6 ÷ 3 = 2 hours.

The lower result is useful evidence, but it does not prove causation by itself. Check whether failure type, equipment duty, technician coverage, access conditions and the measurement boundary remained comparable.

Practical MTTR improvement worksheet

For each qualifying failure, capture enough non-confidential evidence to answer these questions:

• What timestamp started and stopped the agreed clock?
• Which failure mode and functional loss were confirmed?
• How much time fell into each delay category?
• Which delay was largest, repeated or realistically controllable?
• What corrective action, owner and due date were agreed?
• What evidence will show whether the action worked on later comparable events?

Review the distribution as well as the average. One exceptional outage can dominate a small sample. Display the number of repairs and, where useful, the individual durations or median beside MTTR so reviewers can see whether the average represents typical performance.

Actions that commonly reduce MTTR

• Create evidence-based fault guides for repeated symptoms.
• Keep controlled drawings, parameter records and OEM contact routes accessible.
• Define and review critical-spare levels using failure history, lead time and consequence.
• Pre-plan approved isolation, access, lifting and test requirements for critical tasks.
• Improve shift handover with symptoms, measurements, actions already tried and outstanding risks.
• Train competent personnel on frequent high-impact failure modes and safe test methods.
• Use post-job reviews to update job plans, troubleshooting steps and spare-part records.

When an improving MTTR can be misleading

• Clock changes: waiting time was included before but excluded later.
• Premature closure: a work order was closed before functional testing or stable operation.
• Temporary restoration: the machine returned quickly but failed again because the cause remained.
• Failure mix changes: several easy repairs replaced one complex event in the reporting period.
• Missing records: long contractor or night-shift repairs were not captured.
• Unsafe acceleration: required isolation, testing or risk controls were shortened to improve the number.

Use MTTR with MTBF and availability

MTTR describes restoration performance after a qualifying failure. It does not show how often failures occur. Pair it with MTBF, repeat-failure rate, maintenance cost, criticality and operational downtime. Simplified inherent availability uses MTBF ÷ (MTBF + MTTR), but real operational availability can also be affected by planned maintenance and administrative or logistics delays.

Related engineering guides

← Back to Engineering Guides