Free OEE calculator and the OEE formula, step by step

The OEE formula, factor by factor

OEE starts from time, not from output. Planned production time is the time the asset was scheduled to run: the shift length minus planned breaks, meals and periods with no production scheduled. Everything that stops the asset inside that window, whether a breakdown or a changeover, is stop time. Run time is what remains.

The three factors then isolate three different kinds of loss. Availability measures time lost to stops. Performance measures speed lost while the asset was running, including the short stops nobody logs. Quality measures output that was not good the first time. Multiply them and you get the share of planned time that produced good parts at ideal speed.

• Run time = planned production time − stop time
• Availability = run time ÷ planned production time
• Performance = (ideal cycle time × total count) ÷ run time
• Quality = good count ÷ total count
• OEE = Availability × Performance × Quality
• Shortcut for checking: OEE = (good count × ideal cycle time) ÷ planned production time

The Six Big Losses and where they land

The Six Big Losses, from the Total Productive Maintenance (TPM) tradition, give every lost minute a home. Classifying losses this way is what turns an OEE percentage into a short list of things to fix.

Rework counts as a quality loss even when the part is saved later: it was not good the first time, and the machine time spent producing it is gone.

• Availability, equipment failure: breakdowns and unplanned repairs.
• Availability, setup and adjustments: changeovers, tool changes, warm-up and material shortages that stop the asset.
• Performance, idling and minor stops: jams, blocked sensors and short stops, usually under five minutes, cleared by the operator.
• Performance, reduced speed: running below the ideal rate because of wear, poor material or a cautious setting.
• Quality, process defects: scrap and rework during steady production.
• Quality, reduced yield: rejects made during startup or after a changeover, until the process stabilises.

Common OEE calculation mistakes

Most OEE disagreements between two sites come from definitions, not from arithmetic. Agree on these points before comparing any numbers, and write them down next to the spreadsheet.

• Counting planned breaks as downtime. Meals and scheduled breaks come out of planned production time; logging them as stops punishes Availability for time nobody intended to produce.
• Taking changeovers out of planned time. A changeover is a loss you can reduce; hiding it in the schedule makes Availability look better than the line really is.
• Using a demonstrated or “standard” speed instead of the ideal one. The ideal cycle time is the fastest the asset is designed or has proven to run, from the nameplate or the best demonstrated rate. A comfortable standard hides the speed loss.
• Accepting a Performance above 100%. It means the ideal cycle time is too slow, the count covers a different period than the time, or a stop was logged twice. Fix the input rather than capping the result.
• Using one cycle time for a product mix. When the asset runs several products, add up ideal cycle time × count for each product.
• Counting reworked parts as good. Good count means right first time.

Is 85% OEE world class?

The 85% figure goes back to Seiichi Nakajima’s work on Total Productive Maintenance, built from about 90% Availability, 95% Performance and 99.9% Quality. It is widely quoted as “world class” and it is a useful reference point.

It is a reference, not a target to copy. A batch process with frequent changeovers, a line running a wide product mix and a dedicated high-volume machine will never land in the same place. The meaningful comparison is the same asset against itself, week after week, with the same definitions. A first honest measurement is often lower than people expect; that is the starting point, not a failure.

When a spreadsheet is no longer enough

One calculation per shift in a spreadsheet works while one person owns it. It breaks down when stop times are typed from memory, when every supervisor uses a slightly different ideal cycle time, and when nobody links the biggest loss to a maintenance action.

In FreeMaint, downtime is logged on the asset itself on every plan, including the free Core tier, with start time, end time and reason. The Continuous Improvement module, part of the Enterprise plan at $299 per company per month (flat, never per user), adds the OEE worksheet per shift or day, the dashboard and the Six Big Losses Pareto. Counts are entered by hand for now; they are not read from machines.

Breakdowns are usually the first loss to attack. FreeMaint Research, 2026, found that 27.1% of sites log no preventive maintenance at all, a gap that tends to surface as unplanned stops and lost Availability.