Lesson 2 of 10
OEE — the one number
Lesson one gave you six losses. This lesson folds them into a single figure that a plant manager can read in a second — and then shows you the three ways that figure gets quietly inflated before it reaches him.
Step one
Read the material
Three ideas. What the number is built from, why it multiplies instead of averaging, and where it gets rigged.
Concept one
Three fractions, and each one owns two losses
Overall Equipment Effectiveness is not a new measurement. It is the six big losses from lesson one, sorted into three fractions and multiplied together.
Availability is run time divided by planned production time. It carries the two TIME losses — breakdowns and setup. If you planned to run for 420 minutes and the machine actually ran for 357, availability is 85%.
Performance is what the machine should have taken to make what it made, divided by the time it actually had. Ideal cycle time × total count, over run time. It carries the two SPEED losses — small stops and reduced speed. This is the fraction that catches the losses nobody logged, because it does not ask what happened; it asks only what came out.
Quality is good count divided by total count. It carries the two QUALITY losses — startup scrap and production defects. Rework counts against you here, as it did in lesson one.
Every one of the six losses lands in exactly one fraction. Nothing is counted twice and nothing escapes. That is the whole elegance of the thing.
Concept two
It multiplies, and that is the point
Ninety per cent available. Ninety per cent performance. Ninety per cent quality. Three numbers any supervisor would sign off without a second look.
Multiply them: 0.90 × 0.90 × 0.90 = 0.729. Twenty-seven per cent of your capacity is gone, and not one of the three fractions looked like a problem.
People want to average them. Averaging gives 90% and is simply wrong, because the losses are not parallel — they are in series. Performance only acts on the time that survived availability. Quality only acts on the parts that survived performance. Each loss takes its cut of what the previous one left, so they compound the way interest compounds, not the way marks on a report card average.
This is why OEE is worth calculating at all. Three respectable numbers hide a poor one, and only the product exposes it. The commonly quoted world-class figure is 85% — roughly 90.0 × 95.0 × 99.9. Note what that demands: near-perfect quality is not optional, it is the price of admission.
A first honest measurement of 45% to 60% is completely normal. If your first attempt comes out at 92%, you have not found a good machine. You have found a definition problem, which is concept three.
Concept three
The denominator is where OEE gets rigged
Availability is run time over planned production time, and that second phrase is doing an enormous amount of unexamined work. Planned production time is the shift with the excused absences taken out: breaks, scheduled maintenance, shift handover, meetings, and — the elastic one — time when there was no demand.
Take an eight-hour shift where the machine ran 300 minutes. Excuse 120 minutes as planned downtime and availability is 300 ÷ 360 = 83%. Excuse nothing and it is 300 ÷ 480 = 63%. Same shift. Same machine. Same parts. Twenty points of difference, produced entirely by an accounting choice.
Neither number is a lie. But two plants comparing OEE across a fence are usually not measuring the same thing at all, and the plant with the more generous definition wins a competition nobody agreed to enter. When a supplier quotes you an OEE, the first question is not “how do you calculate it” — it is “what is in your denominator?”
Two defences. First, write the denominator down and do not change it, because the value of OEE is the trend, and a trend across a moving definition means nothing. Second, if you want a figure that cannot be gamed, use all the calendar time you paid for — that is TEEP, and it is brutal, and it is honest.
Step two
Work it with your hands on it
One eight-hour shift. Set the losses, then watch the cascade — each bar is what survived the fraction above it. The last row grades the identical shift both ways, so you can see exactly what the denominator is worth.
Hands on it, off the screen. Find whoever reports OEE at your plant and ask one question: what is in the denominator? Then ask to see it written down. Not explained — written down, dated, and unchanged since.
If nobody can produce that document, the number has been drifting, and every improvement it has ever shown is unverifiable. Getting one page written and signed is a bigger contribution than any project you could run on the machine itself.
Step three
Prove you understood it
One question per idea, new numbers every time.
All three solid.
You can build OEE from the six losses, you know why it multiplies, and you know the first question to ask anyone who quotes you one. Lesson three opens the availability fraction and asks what really stops a machine.
Modern Total Productive Maintenance (TPM)
Lesson two of ten. This class is taught from the Total Productive Maintenance series — the eight pillars, the six big losses, and the roadmap into AI and predictive manufacturing. Written by our founder, Dr. Gene A Constant, and donated to the Foundation.
Read on Kindle The whole class AI + TPMGSU is a 501(c)(3) nonprofit and book sales are its only income. A purchase is not a tax-deductible donation. As an Amazon Associate, Global Sovereign University earns from qualifying purchases. This lesson is free and always will be. Global Sovereign University: Different by Design. Better by Mission.