MTBF vs MTTR: Mean Time Between Failures and Mean Time to Repair, With a Worked Example

Mean time between failures (MTBF) measures how long equipment runs, on average, before it fails. Mean time to repair (MTTR) measures how long it takes, on average, to fix it. The first is a reliability number, the second a maintainability number, and together they give you availability. Both can be calculated from work orders most teams already have.

Most explanations use a factory line as the example. Field equipment is different: the pump that fails is on a pad two hours from the shop, so the time to get it running again is mostly travel and parts, not wrench time. This guide works through both formulas on a fleet pump, adds the third number field teams need (mean downtime) and shows which number to attack first.

What Is Mean Time Between Failures (MTBF)?

MTBF is the total operating time of an asset divided by the number of failures in that period.

MTBF = Total operating hours / Number of failures

Two details matter. First, what counts as operating time. For mobile equipment, take it from the hour meter. Calendar time or a default weekly schedule counts the days a unit sat in the yard as running time and inflates MTBF. Second, what counts as a failure: the asset could not perform its function. A scheduled filter change is not a failure. The SMRP Best Practice metrics define both terms if you want a standard to point to.

MTBF is expressed in hours and a higher number is better. For trucks and trailers, the same idea works in miles: mean miles between failures from the odometer.

MTBF vs MTTF

MTBF applies to repairable assets. For a part that is replaced rather than repaired, the equivalent is mean time to failure (MTTF): the average life of the part from install to failure. The pump in this guide has an MTBF; its mechanical seal has an MTTF. Both come from the same work orders, as long as the part replaced is recorded.

What Is Mean Time to Repair (MTTR)?

MTTR is the total time spent repairing an asset divided by the number of repairs.

MTTR = Total repair hours / Number of repairs

In reliability engineering (MIL-HDBK-338B takes this view), MTTR means active repair time: from the start of the repair to the moment the asset is back in service, covering diagnosis, the fix and testing. Travel and waiting for parts are logistics delays and sit outside it. Many CMMS tools and IT teams use the same letters for mean time to recovery, which runs from the failure to back in service and is closer to what this guide calls mean downtime. Before you compare your MTTR with a vendor's or another site's, check which one they mean.

Repair time is elapsed time, not labor hours. Two technicians working four hours is a four-hour repair, not an eight-hour one. For a plant with a maintenance shop next to the line, the difference between repair time and downtime is small. For mobile equipment it is the whole story, so track the third number alongside it:

Mean downtime (MDT) = Total downtime hours / Number of failures

MDT runs from the moment the asset stopped to the moment it was back in service and includes notification, travel, waiting for parts and the repair itself. MTTR tells you how good your technicians are. MDT tells you how good your operation is. Report both; the gap between them shows where the time goes.

A Worked Example From One Pump's Work Orders

Take a transfer pump on a rental fleet over the last 12 months. The work order history shows:

• MTBF = 8,000 / 6 = 1,333 hours. Running around the clock, that is about 56 days between failures.

• MTTR = 48 / 6 = 8 hours. The crew fixes it in a shift once they are on it.

• MDT = 150 / 6 = 25 hours. The pump is down for more than a day each time.

Seventeen of every 25 downtime hours are not repair. Of those 17, about 2 were the call reaching dispatch, 4 were the drive out and 11 were waiting on a seal kit. That gap is 102 hours a year on this one pump, more than four rental days. No amount of technician training fixes it. A seal kit on every service truck and faster dispatch do.

MTBF for a Fleet, Not Just One Unit

The same formula works at fleet level: add up the operating hours of every unit of a model and divide by the total failures across them. Twelve transfer pumps that ran 90,000 hours between them with 40 failures give a fleet MTBF of 2,250 hours.

That fleet number does two jobs. It sets stock levels: if failure codes show 25 of the 40 failures were seals, the fleet uses about two seal kits a month, and the stock level follows from that and the supplier's lead time. And it finds the outlier. Without the example pump, the other eleven ran 82,000 hours with 34 failures, about 2,400 hours MTBF. At 1,333 hours, the example pump fails almost twice as often as its siblings. The problem is that pump, its duty or its operator, not the model.

Keep fleet MTBF by model and by duty. A pump on continuous transfer duty and the same pump on intermittent standby will not fail at the same rate, and mixing them hides both numbers. Across the whole 12-pump fleet, the 17 non-repair hours per failure add up to 680 hours a year, about four weeks of pump time.

MTBF vs MTTR: Which One to Work On First

The three numbers point at different fixes. Use the pattern you see to pick the first move.

For the pump above, the failure rate is the bigger problem (twice its siblings' rate), but the downtime is the faster fix. Getting a seal kit onto every service truck would cut MDT before the next failure. Finding out why the seal fails takes longer and belongs in a predictive or proactive program.

Availability From the Two Numbers

Availability is the share of time the asset is able to run. There are two versions, and they give different answers.

Inherent availability = MTBF / (MTBF + MTTR)

Operational availability (close to it) = MTBF / (MTBF + MDT)

For the pump: inherent availability is 1,333 / (1,333 + 8) = 99.4 percent, which looks fine. Operational availability is 1,333 / (1,333 + 25) = 98.2 percent, and that is the number the customer feels, because it includes the day of lost work per failure. This is the most common way availability gets overstated for field equipment: the travel and parts time is left out because nobody recorded it.

Availability is also not utilization. The pump could be available 98.2 percent of the time and still only ran 8,000 of the year's 8,760 hours, about 91 percent. Availability says whether the unit could work; utilization says whether it did.

How to Raise MTBF and Cut Downtime on Field Equipment

Improvement lists written for plants talk about technician training and spare parts rooms. On field equipment the hours go elsewhere, so organise the fixes by where the time is lost.

Getting the Data Out of Your CMMS or Field App

Both metrics are only as good as the work orders. The fields you need on every corrective work order, captured on the digital work order in the field:

1. Asset ID (one per physical unit; a model number is not an asset).

2. A failure flag or failure code, so PMs and inspections can be filtered out.

3. Timestamps for dispatch, arrival on site, repair start and back in service. Labor hours still matter for cost, but they are not repair time.

4. Hour meter reading (or odometer for trucks) at failure, to build operating time.

5. Part replaced, so MTTF per part and seal-kit stock can be worked out later.

A CMMS or field service system that captures the reading and the timestamps on the mobile work order removes most of the data problems in the next section.

Common MTBF and MTTR Mistakes

• Operating hours estimated from the calendar instead of the meter, which inflates MTBF for anything that sits in the yard.

• Failures logged against a location or a model rather than the unit, so no single asset ever looks bad.

• MTTR calculated from labor hours (two technicians, four hours each, logged as eight) instead of elapsed repair time.

• Downtime fields left blank because the technician closed the job from the truck, so MDT cannot be calculated at all.

• Comparing your MTTR with a vendor's without checking whether theirs means repair time or recovery time.

• Reading the average without the spread. Five outages of 12 hours and one of 90 also average 25 hours, but that pattern needs a different fix: find out what made the one outage so long.

• Treating MTBF as a prediction. Under a constant failure rate only about 37 percent of runs last as long as the MTBF; real failures often come well before or after it. Use it to compare and trend, not to schedule. The cost of each failure is what turns these numbers into a business case.

What Is a Good MTBF and MTTR?

There is no universal target. A frac pump, a light tower and a service truck fail at different rates by design, and a number that is good for one is terrible for another. Three comparisons are worth making instead:

• The same asset against its own history: is MTBF rising quarter over quarter?

• Units of the same model against each other: the pump above fails twice as often as its siblings, and that has a cause you can find.

• MDT against MTTR: if MDT is more than twice MTTR, logistics, not repair skill, is costing you the hours.

If you want an external sanity check for oil and gas equipment, the OREDA handbooks publish failure rates and repair times collected from operators, by equipment class. Treat them as a range, not a target.

Track all three numbers on every revenue asset, review them monthly and tie the review to the spare parts decision, because the parts on the truck are usually the quickest way to move the downtime number. Equipt's field service management software tracks every work order from creation to closure with time-stamped evidence, and works offline at remote sites, so the numbers come out of the system rather than a spreadsheet.

Frequently Asked Questions

What is the difference between MTBF and MTTR?

MTBF (mean time between failures) is the average operating time between failures and measures reliability. MTTR (mean time to repair) is the average time to fix a failure and measures maintainability. Higher MTBF and lower MTTR are both better.

How do you calculate MTBF?

Divide total operating hours by the number of failures in the same period. For example, 8,000 operating hours and 6 failures give an MTBF of 1,333 hours. Count only true failures, not scheduled maintenance, and take operating hours from the meter.

How do you calculate MTTR?

Divide total repair time by the number of repairs, using elapsed time from repair start to back in service, not labor hours. Six repairs totalling 48 elapsed hours give an MTTR of 8 hours.

Does MTTR include travel time?

In reliability engineering, no: MTTR covers diagnosis, repair and testing. Travel and parts waiting belong in mean downtime (MDT). Some CMMS tools define MTTR as mean time to recovery, which does include them, so check the definition before comparing numbers.

What is the difference between MTBF and MTTF?

MTBF is for repairable assets and measures time between failures. MTTF (mean time to failure) is for parts that are replaced rather than repaired and measures their average life. A pump has an MTBF; its mechanical seal has an MTTF.

How do you calculate availability from MTBF and MTTR?

Inherent availability is MTBF divided by (MTBF + MTTR). For field equipment, also calculate MTBF divided by (MTBF + MDT), which includes travel and parts delays and is the number customers experience.

What is a good MTBF or MTTR?

It depends on the asset. Compare a unit against its own trend and against other units of the same model rather than a generic benchmark. A rising MTBF and a falling MDT over several quarters show the program is working.

Does MTBF tell you when the next failure will happen?

No. MTBF is an average for comparing and trending, not a prediction for one unit. Under a constant failure rate only about 37 percent of runs last as long as the MTBF, so individual failures often come well before or after it.

Want MTBF, MTTR and downtime per unit without building the spreadsheet? Book a demo