OEE Calculation: Formula and Worked Example

OEE is the most talked-about and the most miscalculated metric on the production floor. If two teams can arrive at two different OEE figures for the same line, the problem is not the formula but the definition of what goes into it. In this article we work through the calculation step by step, using a real shift.

What is OEE?

OEE (Overall Equipment Effectiveness) is a combined metric showing how much of its planned production time a piece of equipment actually spends creating value. It is the product of three components:

OEE = Availability × Performance × Quality

The three are multiplied because losses compound one another. Three components of 90 per cent each do not give a result of 90 per cent but of 72.9 per cent.

The three components and their formulas

Availability

Measures downtime losses.

Availability = Run Time ÷ Planned Production Time

Planned production time is the shift length minus planned stoppages (breaks, planned maintenance, hours outside the shift). Run time is what remains once unplanned stoppages are also deducted: breakdowns, changeovers, waiting for material, adjustment.

Performance

Measures speed losses: the machine is running, but not at the speed it should be.

Performance = (Total Output × Ideal Cycle Time) ÷ Run Time

The most critical input here is the ideal cycle time. It should not be the machine's catalogue speed but the best speed that is sustainably achievable on that product; and once set, this value must stay fixed.

Quality

Measures defect losses.

Quality = Good Units ÷ Total Output

Reworked units do not count as good. Rework is a loss, and without that distinction the quality component comes out higher than it really is.

A step-by-step worked example

An eight-hour shift on an injection moulding machine:

• Shift length: 480 minutes
• Planned break: 45 minutes → Planned production time: 435 minutes
• Breakdown: 38 minutes · Changeover: 42 minutes · Waiting for material: 15 minutes → Unplanned downtime: 95 minutes
• Run time: 435 − 95 = 340 minutes
• Total output: 5,100 units · Ideal cycle time: 3.6 seconds/unit (0.06 minutes)
• Defective units: 190

The calculation:

• Availability = 340 ÷ 435 = %78,2
• Performance = (5.100 × 0,06) ÷ 340 = 306 ÷ 340 = %90,0
• Quality = (5.100 − 190) ÷ 5.100 = 4.910 ÷ 5.100 = %96,3
• OEE = 0,782 × 0,900 × 0,963 = %67,8

This table says far more than a single percentage: the biggest loss is in availability, and within it the largest single item is changeover (42 minutes). So the first job on this line is a SMED exercise as a lean manufacturing tool — not pushing performance harder.

OEE calculator

Enter the figures from your own shift; the three components, OEE and the minutes lost in each category are calculated instantly. The fields start with the worked example above; because the calculator multiplies unrounded components, the example’s 67.8% shows here as 67.7%.

Time

Shift length min Planned stops (breaks, planned maintenance) min Unplanned stops (breakdowns, changeovers, waiting) min

Output

Ideal cycle time sec/unit Total output units Defects + rework units

Back to example values Clear fields

OEE –

Availability–

Performance–

Quality–

World-class target: 90 · 95 · 99 → 85

Where did the planned production time go?

• Value-adding time –
• Downtime loss –
• Speed loss –
• Quality loss –

Where to look first

Let's set up OEE measurement on your shop floor

The six big losses

The OEE components map directly onto the six big losses in the TPM literature:

• Availability: breakdown downtime · setup and changeover
• Performance: minor stops and idling · running at reduced speed
• Quality: start-up losses (first parts) · in-process defects

This mapping turns OEE from a scoreboard into a tool that generates action. Once the loss category is identified, so is the method to use: TPM and autonomous maintenance for breakdown losses, line balancing for speed losses, root cause analysis for quality losses.

Six common mistakes that distort OEE

1. Stretching the definition of planned downtime. Reclassifying every unplanned stoppage as "planned maintenance" after the fact raises OEE but hides the truth. The definition must be written down and fixed.
2. Changing the ideal cycle time from shift to shift. Raising the ideal time whenever the target is missed makes the metric meaningless.
3. Not recording minor stops. Stoppages under five minutes usually go unrecorded — yet they are often the single largest source of performance loss.
4. Counting rework as good product. This inflates the quality component artificially.
5. Using OEE as a measure of individual performance. If operators are appraised on OEE, the data loses its reliability. OEE measures the equipment, not the person.
6. Averaging to find a line's OEE. On stations connected in series, a line's OEE is not the average but is set by the performance of the constraint station.

What should the target be?

The world-class reference for discrete manufacturing is 85 per cent (availability 90% · performance 95% · quality 99%). But a figure of 50 per cent on a first measurement is not bad news; it is a map of the improvement potential.

A more useful approach than setting an absolute target is to track the trend of a figure measured the same way each time, and to turn the largest loss item each week into a Kaizen topic. To tie the measurement to a daily management rhythm, a KPI board and short shop-floor meetings are the most effective structure.