OptimalOEE™ Wiki / Calculating OEE
Formulae
Calculating OEE
OEE is three ratios multiplied together. Here are the equations, a worked example, the world-class benchmark each factor has to hit, and the places an OEE number is quietly wrong.
The formulae
Availability = Run Time ÷ Planned Production Time
Performance = (Ideal Cycle Time × Total Count) ÷ Run Time
Quality = Good Count ÷ Total Count
OEE = Availability × Performance × Quality
Equivalently, OEE is fully productive time over planned time: OEE = (Good Count × Ideal Cycle Time) ÷ Planned Production Time. The two forms always agree, which is a useful self-check. Planned Production Time is shift time minus the agreed non-production (breaks and, by convention, no-demand and planned shutdowns); define it once and hold it, or OEE stops being comparable.
A worked example
One eight-hour shift on a single station:
| Planned production time | 480 min |
|---|---|
| Downtime (stops + changeover) | 60 min, so Run Time = 420 min |
| Ideal cycle time | 1.0 s per unit (3,600 per hour) |
| Total count | 21,000 units |
| Good count | 20,000 units |
Availability = 420 ÷ 480 = 87.5%
Performance = (1.0s × 21,000) ÷ (420×60)s = 21,000 ÷ 25,200 = 83.3%
Quality = 20,000 ÷ 21,000 = 95.2%
OEE = 0.875 × 0.833 × 0.952 = 69.4%
Cross-check: (20,000 × 1.0s) ÷ (480×60)s = 20,000 ÷ 28,800 = 69.4%. The station lost most to availability and speed, not quality, which is where to look first.
World-class, factor by factor
The 85% world-class benchmark is not one target; it is the product of a demanding number on each factor. Miss one and the product falls fast:
| Availability | about 90% – little unplanned downtime, fast changeovers |
|---|---|
| Performance | about 95% – running at rated speed, minimal minor stops |
| Quality | about 99.9% – almost no defects or startup rejects |
| OEE | 0.90 × 0.95 × 0.999 = about 85% |
OEE, OOE and TEEP
OEE, OOE and TEEP are the same three factors, availability × performance × quality, measured against three different clocks. Only the availability denominator changes:
| OEE | Run Time ÷ Planned Production Time. How well you ran when you were meant to be running. The shop-floor lever. |
|---|---|
| OOE | Run Time ÷ Shift Time. Overall operations effectiveness, counting the planned stops and breaks inside a staffed shift. |
| TEEP | Run Time ÷ All Available Time, the full 24/7 calendar. Total effective equipment performance, your true capacity. |
Because TEEP measures against every hour on the calendar it exposes the hidden factory, the capacity idle on nights and weekends. It is also why TEEP reads low: one eight-hour shift five days a week caps TEEP at about 24% (40 of 168 hours) even at a world-class OEE. Use OEE to run the shift well, and TEEP to decide whether to add one.
Where OEE calculations go wrong
Most disputed OEE numbers fail on the same few points, and a from-source calculation is what removes them:
- Ideal cycle time set soft. If it is the actual or a conservative rate, performance is capped and speed loss disappears. A performance over 100% is the tell that the ideal cycle time is wrong. The honest basis is the Maximum Demonstrated Rate (MDR), the fastest speed the asset has reliably sustained.
- Planned production time defined inconsistently. Whether breaks, planned maintenance and no-demand sit inside or outside changes the number; two lines only compare on one definition.
- Minor stops hiding in performance. Short, unlogged stops are absorbed into the speed loss and never named, so the biggest fixable loss stays invisible.
- Quality counting scrap but not rework or startup rejects. First-time-good is the honest quality number.
OptimalOEE™ recomputes each factor from source on every read, against a fixed ideal cycle time and a fixed planned-time definition, so the OEE it reports is consistent, explainable and hard to game.
Honest numbers
OEE is also one of the most manipulated metrics on the shop floor, usually by moving one of three levers: shrinking the time base (dropping maintenance, breaks or changeover out of planned time), softening the speed baseline (an easy MDR, so performance reads over 100%), or flattering the count (rework or startup scrap counted as good). Each can lift the number by 5–15% with no real gain. Two fast red flags: any OEE above 90% deserves scrutiny, and a performance above 100% is impossible by definition.
When a line is first measured from source and end to end, the honest OEE is frequently well below the figure it had been reporting. That is not a step back, it is the starting line, because you can only improve from a number you can defend. It is why OptimalOEE™ computes from the raw signal and keeps the audit trail, so the score is the real one.
Try it
Move the three factors to see the OEE they produce. Because it is a product, watch how one low factor dominates.
World-class is around 85%. Under 60% usually means no structured reliability programme.
Overall OEE
OEE = A × P × Q. Green at or above 85%, amber 60–84%, red below 60%.