How to Calculate OEE (Overall Equipment Effectiveness): Formula + Worked Examples

To calculate OEE you multiply three loss factors - Availability x Performance x Quality - into a single percentage showing how much of your planned production time is genuinely productive. This guide walks through the OEE formula factor by factor with worked numbers, the Six Big Losses behind each one, the 85% world-class benchmark, and the mistakes (like a performance score above 100%) that quietly inflate the number.

What OEE Actually Measures

To know how to calculate OEE, start with what it measures. Overall Equipment Effectiveness (OEE) is the percentage of planned production time that an asset spends making good parts at its rated speed. It was formalised within Total Productive Maintenance (TPM) by the Japan Institute of Plant Maintenance (JIPM) and is now the standard yardstick for manufacturing productivity.

Its power is folding three different kinds of loss into one comparable number:

• Availability - did the machine run when it was supposed to? (downtime losses)
• Performance - when running, did it run at full speed? (speed losses)
• Quality - of what it made, how much was right first time? (defect losses)

A 65% OEE does not mean the equipment is broken 35% of the time. It means that after stripping out stoppages, slow running and scrap, only 65% of planned time produced sellable output.

The other 35% is hidden capacity - capacity you already pay for in labour, energy and overhead. That reframing makes invisible loss visible, which is why OEE is the entry point to operational excellence. Our free OEE Calculator applies the exact formula below if you want to skip the arithmetic.

The OEE Formula

The headline formula is the product of the three factors:

OEE = Availability x Performance x Quality

Because it is a product, OEE is unforgiving. Three respectable factors of 90% each multiply to 0.90 x 0.90 x 0.90 = 72.9%, not 90%. A single weak factor drags the whole score down - which is the point: it points you straight at the biggest opportunity.

Here is each factor with its sub-formula.

Availability

Availability = Run Time / Planned Production Time, where Run Time = Planned Production Time minus all unplanned stop time. Planned production time is the time you intended to run after subtracting planned shutdowns such as no demand, unstaffed breaks, or scheduled maintenance.

Performance

Performance = (Ideal Cycle Time x Total Count) / Run Time. Equivalently, it is your actual run rate divided by the ideal (nameplate) rate. It captures small stops and reduced speed - losses that don't register as a downtime event but quietly bleed throughput.

Quality

Quality = Good Count / Total Count. Only parts that pass first time count as good. Reworked parts count against you, because rework consumed capacity even if the part is eventually sold.

Worked Example: Calculating OEE Step by Step

Let's run a complete shift through the formula. These are the same default inputs pre-loaded in the OEE Calculator, so you can reproduce every number.

Step 1 - Run Time and Availability

Run Time = 480 - 60 = 420 min.
Availability = 420 / 480 = 0.875 (87.5%).

Step 2 - Performance

Ideal Cycle Time x Total Count = 0.5 min x 700 = 350 min of work theoretically required.
Performance = 350 / 420 = 0.833 (83.3%).

In plain terms: in 420 minutes of run time you made what should have taken 350 minutes at rated speed, so the line ran at about 83% of its potential pace.

Step 3 - Quality

Quality = 670 / 700 = 0.957 (95.7%).

Step 4 - OEE

OEE = 0.875 x 0.833 x 0.957 = 0.698, or 69.8%.

Sanity check: the line could theoretically have produced 480 min / 0.5 min = 960 units; it actually made 670 good ones; 670 / 960 = 69.8%. Both routes agree. At 69.8% this asset sits in the typical band - with meaningful capacity to recover, most of it hiding in the Performance factor.

The Six Big Losses Behind Each Factor

TPM maps OEE onto Six Big Losses, two per factor. Naming the loss is what turns a score into an action - you can't fix "83% performance," but you can fix "the labeller micro-stops every two minutes."

The Performance losses - small stops and reduced speed - are the most under-measured, because each event is too brief to log, yet collectively they are often the single largest loss in a plant. A continuous downtime record, ideally from a CMMS, is what surfaces them.

The 85% World-Class Benchmark

The widely cited world-class OEE benchmark is 85%. It is not arbitrary - it is the product of three world-class factor targets:

• Availability ~ 90%
• Performance ~ 95%
• Quality ~ 99.9%

0.90 x 0.95 x 0.999 = 0.854, hence 85%. Use these factor-level targets, not just the headline number, to see which factor is holding you back.

For context, many discrete manufacturers run an OEE of 40-60%, and 60% is a common starting point. So the gap between a typical plant and world-class is enormous - frequently a doubling of effective capacity from the same assets with no capital spend.

Treat 85% as a north star, but your most useful benchmark is your own trend line: a steadily rising OEE on a stable measurement basis matters more than any single absolute number.

OEE vs TEEP, and What Each One Excludes

OEE deliberately measures effectiveness only during planned production time. That is the right scope for improving a running asset - but it hides one thing: time you never scheduled the machine at all. Three related metrics zoom out progressively:

• OEE = Availability x Performance x Quality, based on planned production time.
• TEEP (Total Effective Equipment Performance) = OEE x Utilisation, where Utilisation = scheduled time / all calendar time (24/7/365). TEEP exposes capacity lost to running fewer shifts, idle weekends, or no demand.
• OOE (Overall Operations Effectiveness) sits between the two, using operating time as its base.

An asset can post a strong 85% OEE on a single weekday shift yet have a TEEP near 28%, because it sits idle two-thirds of every day and all weekend.

OEE answers "how good are we when we run?"; TEEP answers "how much of our total potential are we capturing?" Use OEE to drive shop-floor improvement and TEEP for capital and capacity decisions - before buying a second machine, check whether TEEP says you already own the capacity.

Common OEE Mistakes (Including Performance Over 100%)

OEE is easy to calculate and easy to game. These are the errors that most often make a number untrustworthy.

Performance above 100%

If your Performance factor exceeds 100%, the machine appears to run faster than physically possible. It isn't - your ideal cycle time is set too slow. Re-baseline the ideal cycle time to the fastest demonstrated sustainable rate (the true nameplate or best stable run), not an outdated or padded standard. An honest ideal cycle time is the foundation of the whole calculation.

Excluding losses to inflate the score

Reclassifying breakdowns as "planned downtime," excluding minor stops, or counting reworked parts as good all inflate OEE. The discipline of TPM is to count all the loss, then attack it.

Inconsistent time basis

If one shift subtracts breaks from planned time and another doesn't, the numbers aren't comparable. Define planned production time once, in writing, and apply it everywhere.

Chasing OEE on a non-bottleneck

Improving OEE on a machine that isn't the constraint just builds inventory. Measure broadly, but improve at the bottleneck first.

Treating OEE as the goal

OEE is a diagnostic, not a target to hit by any means. The aim is more good output at lower cost. Translate recovered availability into money with the Downtime Cost Calculator to keep the focus on value.

How to Improve Each OEE Factor

Because OEE is a product, the fastest gain is almost always in your lowest factor. Improve in this order: find the weakest factor, find its dominant Big Loss, then apply the matching lever.

Improving Availability

• Reduce breakdowns: shift from reactive to preventive and predictive maintenance. Rank assets with the Asset Criticality Calculator, then track reliability with MTBF and MTTR - raising MTBF cuts the number of stops, cutting MTTR shortens each one.
• Slash changeover time: apply SMED (Single-Minute Exchange of Die) to convert internal setup steps into external ones done while the machine still runs.
• Standardise PM: a clear, measurable procedure makes maintenance faster and more repeatable - the PM Procedure Generator builds one with acceptance criteria.

Improving Performance

• Eliminate small stops: autonomous maintenance, better guarding and sensors, and clean-and-inspect routines remove the micro-jams that erode speed.
• Restore rated speed: address worn tooling, lubrication, and operator technique to close the gap between your actual run rate and the nameplate rate used in the Performance factor.

Improving Quality

• Reduce variation: use SPC and process capability - the Cp/Cpk Calculator tells you whether the process can hold spec, and the DPMO & Sigma Calculator converts defect rates into a sigma level.
• Cut startup scrap: standardise warm-up and first-article checks; mistake-proof (poka-yoke) the steps that generate production rejects.

Underpinning all three is workplace discipline - a tidy, organised line surfaces abnormalities fast, which is why many programs start with a 5S audit. When you're ready, the full AMAADOR tool library covers reliability, quality and cost end to end.

Related free calculators

Frequently asked questions