Work Order App Builder

Maintenance KPIs: MTTR, MTBF and PM Compliance Explained (2026)

How to calculate MTTR, MTBF, availability, PM compliance and the other maintenance KPIs that matter, with worked numbers and published benchmarks.

Floot ·

Key Takeaways

  • MTTR = total repair time ÷ number of repairs. It measures how quickly you restore failed equipment. Decide in advance whether the clock includes waiting for parts.
  • MTBF = total operating time ÷ number of failures. It measures how long equipment runs between failures. Calculate it per asset or asset class, not across the whole site.
  • PM compliance = PMs completed on time ÷ PMs scheduled. DOE’s benchmark table lists above 90% for preventive maintenance completion.
  • Watch emergency work as a share of hours. DOE’s benchmark is below 10%. A high share means PM is not preventing failures, or work is being mislabeled.
  • Five KPIs reviewed monthly beat twenty nobody reads. Pick the few that drive decisions and publish them on one page.

Quick-Start Prioritization Framework

KPIFormulaPublished benchmark (DOE table)
MTTRTotal repair time ÷ number of repairsNone given; trend it down
MTBFTotal operating time ÷ number of failuresNone given; trend it up
Equipment availabilityHours available ÷ total hours in periodAbove 95%
PM completion (compliance)PMs completed ÷ PMs scheduledAbove 90%
Schedule complianceHours worked on scheduled jobs ÷ hours scheduledAbove 90%
Emergency maintenanceHours on emergency jobs ÷ total hours workedBelow 10%
Maintenance overtimeOvertime hours ÷ regular hoursBelow 5%

Maintenance KPIs are only useful if you calculate them the same way every month and they lead to a decision. MTTR tells you how long repairs take. MTBF tells you how often things break. PM compliance tells you whether the plan is being followed. Together with a few workload measures, they answer the questions your director will ask: is equipment getting more reliable, is the team keeping up, and where is the money going? This guide gives the formulas, worked examples with real arithmetic, the benchmarks published in the U.S. Department of Energy’s O&M Best Practices Guide, and the traps that make maintenance numbers misleading.

Where the benchmarks come from

DOE’s Federal Energy Management Program O&M Best Practices Guide (Release 3.0, 2010) includes a table of industry O&M metrics and benchmarks, citing NASA (2000). It covers equipment availability, schedule compliance, emergency maintenance percentage, overtime percentage and preventive maintenance completion percentage. Those are the benchmarks quoted in this guide.

The guide also lists broader indicators worth tracking: work orders generated and closed out, backlog of corrective maintenance, safety record, energy use, inventory control, overtime, absentee rate and staff turnover. MTTR and MTBF are not in its benchmark table, so for those this guide uses the standard definitions and recommends trending against your own history.

MTTR: mean time to repair

IBM defines MTTR as the average time it takes to repair a system or piece of equipment after it has failed, calculated as total time spent on repairs divided by the number of repairs.

Worked example

A domestic water booster pump failed six times last quarter. The repairs took 1.5, 2, 4, 2.5, 3 and 2 hours. Total repair time is 15 hours. MTTR = 15 ÷ 6 = 2.5 hours.

Decide what the clock includes

Some teams measure only hands-on repair time. Others measure from the moment of failure to the moment the equipment is back in service, which includes detection, travel, waiting for parts and testing. Both are valid; mixing them is not. Write the definition down. If waiting for parts dominates the longer measure, that is a stores problem rather than a skills problem, and you will only see it if you can split the two.

What moves MTTR

Parts on hand for critical assets, asset history and manuals available at the equipment, clear fault-finding procedures, and technicians trained on that equipment. A rising MTTR on one asset class often means a skills gap or an aging fleet that needs harder repairs.

MTBF: mean time between failures

IBM defines MTBF as the average amount of time a system or component operates before it fails, calculated as total operating time divided by number of failures. It estimates failure frequency; it does not predict when any one unit will fail.

Worked example

The same booster pump is meant to run continuously. A 90-day quarter has 2,160 hours. It was down for 15 hours of repairs, so it operated for 2,145 hours. It failed six times. MTBF = 2,145 ÷ 6 = 357.5 hours, or roughly one failure every 15 days.

Use operating time, not calendar time

For equipment that does not run all the time, use run hours from a meter or the building automation system. A chiller that runs 1,200 hours in a season and fails twice has an MTBF of 600 hours, not half a year.

Calculate per asset or asset class

Site-wide MTBF mixes light fixtures with chillers and tells you little. Calculate it for critical assets individually and for groups of identical assets (all rooftop units of one model). The useful signal is a trend: MTBF rising after you change a PM task means the task works.

Availability: putting MTTR and MTBF together

DOE’s table defines equipment availability as hours each unit is available to run at capacity divided by total hours in the reporting period, with a benchmark above 95%.

Worked example

The booster pump was available 2,145 of 2,160 hours. Availability = 2,145 ÷ 2,160 = 99.3%. The same answer comes from MTBF ÷ (MTBF + MTTR) = 357.5 ÷ 360 = 99.3%, because the inputs are the same.

Why availability can hide a problem

99.3% looks excellent and comfortably beats 95%. But six failures in a quarter is a lot, and each one may have meant no water pressure on the upper floors. High availability with low MTBF means you are fast at fixing something that should not be breaking. That is the case for root cause analysis and better PM on that pump, not a reason to celebrate.

PM compliance and schedule compliance

These two show whether the plan is being followed. DOE benchmarks both above 90%.

PM completion percentage

DOE’s formula is preventive maintenance actions completed divided by preventive maintenance actions scheduled. Worked example: 200 PMs were due in October and 184 were completed by their due date (or within your defined grace window). PM compliance = 184 ÷ 200 = 92%.

Define the grace window in writing. A common approach is that a monthly PM counts as on time if completed within a set number of days of its due date. Without a rule, people count PMs completed three weeks late as compliant, and the number stops meaning anything.

Schedule compliance

DOE’s formula is total hours worked on scheduled jobs divided by total hours scheduled. If you scheduled 600 hours of work in a month and 540 of those hours were worked on the scheduled jobs, schedule compliance is 90%. Low schedule compliance usually means reactive work keeps displacing the plan.

Break compliance down

Report PM compliance by category or by critical asset. 92% overall can hide 60% on fire safety inspections, which is the number that matters at an audit.

Workload KPIs: emergency share, overtime and backlog

These tell you whether the team is keeping up and whether staffing matches the work.

Emergency maintenance percentage

Hours on emergency jobs divided by total hours worked, benchmarked by DOE below 10%. If technicians logged 1,100 hours last month and 165 were emergencies, that is 15%: a sign that failures are not being prevented, or that the emergency label is being overused.

Overtime percentage

Overtime hours divided by regular hours, benchmarked by DOE below 5%. Sustained high overtime is a cost, a burnout risk and a signal that you are under-staffed for the reactive load.

Backlog in crew-weeks

Add the estimated hours of all open, approved work and divide by the weekly hours your team can put toward it. 420 hours of backlog with 240 available hours a week is 1.75 weeks of work. DOE lists backlog of corrective maintenance as an indicator of workload and of PM effectiveness; track the trend. A steadily growing backlog means demand exceeds capacity.

Building a monthly KPI report

A one-page report reviewed at the same meeting every month does more than a dashboard with 30 tiles.

  • PM compliance, overall and for life-safety and critical assets.
  • Emergency share of hours and overtime percentage.
  • Backlog in weeks, with the five oldest open work orders listed.
  • MTBF and MTTR for your 10 most critical assets, with last quarter’s figure beside them.
  • Top five assets by reactive cost this month.
  • Work orders opened versus closed.

Tie each number to an action

For every KPI that moved the wrong way, write one sentence on why and one on what you will do. The number is the start of the conversation.

Common mistakes

These make maintenance KPIs look precise while saying very little.

  • Changing the MTTR definition from month to month, or across sites.
  • Calculating MTBF with calendar time for equipment that runs part time.
  • Counting late PMs as compliant because there is no written grace window.
  • Closing work orders without labor time, so MTTR and cost per asset are fiction.
  • Reporting site-wide averages that hide the critical assets.
  • Celebrating high availability while the same asset keeps failing.
  • Collecting KPIs nobody reviews or acts on.

Getting clean data

Every KPI here depends on work orders with a failure date and time, a return-to-service time, labor hours, the asset, and whether the job was planned or reactive. A CMMS records these if technicians fill them in; make the fields required at close.

If you’d rather not do this in an off-the-shelf tool, Floot, an AI app builder, can build a work order app from a plain-English description with the exact fields and KPI calculations described here. It is free to start.

Frequently Asked Questions

How do you calculate MTTR?

Divide total repair time by the number of repairs in the period. Six repairs totaling 15 hours give an MTTR of 2.5 hours.

How do you calculate MTBF?

Divide total operating time by the number of failures. A pump that operated 2,145 hours and failed six times has an MTBF of 357.5 hours.

What is a good PM compliance rate?

DOE’s O&M Best Practices Guide lists a benchmark above 90% for preventive maintenance actions completed versus scheduled.

What is the difference between MTBF and MTTF?

MTBF is used for repairable equipment that returns to service after a failure. MTTF, mean time to failure, is usually used for items that are replaced rather than repaired. IBM notes MTBF is sometimes written as MTTF.

What maintenance KPIs should a small team track?

Start with PM compliance, emergency share of hours, backlog in weeks and work orders opened versus closed. Add MTBF and MTTR for your most critical assets once work orders reliably record times.

How is equipment availability calculated?

Hours the equipment was available to run divided by total hours in the period. DOE’s benchmark table lists above 95%.

Sources