OEE cannot pass 100% with correct data, because it is the time the good units needed at the ideal rate, divided by planned time. A reading above 100% means one input is wrong: the ideal rate is too low, the count is too high, or the planned time is too short.
Why 100% is a hard ceiling
OEE is a share of planned time, and a share cannot pass the whole.
The ideal rate is the fastest the line is accredited to run, so 100% already means full speed, no stops and no bad units for the whole planned time. To go higher, the good units would need more minutes than the shift had.
That gives you a test you can run in one minute, with no factors at all.
The same shortcut is proven step by step in time-based vs count-based OEE.
One shift that reads 104%
Example: a bottling line has 480 planned minutes and 30 minutes of stops, so it runs 450. It makes 50,400 bottles and 49,920 are good.
The system holds an ideal rate of 100 bottles per minute. Availability is 93.75%, performance is 112% and quality is 99.05%, which multiply to 104.0%.
Now say the line's real accredited rate is 120 per minute. Performance becomes 93.33% and OEE becomes 86.7%, so the line was never above 100%.
The chart shows how three different data errors on this one shift all land above the ceiling.
Cause 1: the ideal rate is too low
This is the first place to look. The rate on file may belong to an older machine, a slower format or last year's average.
A soft rate gives a steady reading a little over 100%, usually on one product or one line. Our guide to ideal cycle time shows how to set the rate properly.
Watch for a very large reading too. If OEE shows 300%, the count is probably in parts while the rate is in cycles, a classic on multi-cavity tools and multi-lane packers.
Cause 2: the count is too high
In the example, a photo eye that reads one bottle in five twice turns 50,400 into 60,480. The reported good count becomes 60,000, and 60,000 ÷ 120 = 500 minutes, which is 104.2% of the shift.
- Double pulses. A sensor fires twice on one unit, often after a guide or a speed was changed.
- A late reset. The counter resets after the shift starts, so units from the shift before are added to this one.
- Buffer output. On a line, the end counter sends out units that were made earlier and held in an accumulator.
- Good units from elsewhere. Reworked or returned units pass the counter a second time.
A clear test is to compare the counter with what was packed or shipped. Our page on what counts as a good part sets the rules for the good number.
Cause 3: planned time is too short
This cause is the easiest to miss, because the rate and the count both look right. The units are real, but the minutes that made them were taken out of the denominator.
Example: 90 minutes of breaks and a meeting come off the schedule, but a relief crew keeps the line running. The good units need 416 minutes, and 416 ÷ 390 = 106.7%.
If the line ran, those minutes are scheduled time. See how to count breaks in OEE for the staffed and stopped cases.
Which factor is over 100%?
If your system shows the three factors, the one that breaks the limit tells you where to start.
| Factor above 100% | What it means | Check first |
|---|---|---|
| Availability | Run time is longer than planned time | Excluded time while the line ran, or a wrong shift pattern |
| Performance | More units than the ideal rate allows | Ideal rate, unit of measure, counter |
| Quality | More good units than total units | Good and total counted at different points or in different periods |
| None, but OEE is high | Each factor looks fine and the result still looks too good | Run the minute test, then the losses to 100% |
When performance is the one, the deeper guide is why performance goes over 100%.
How to fix the data, in order
Good units ÷ ideal rate, against planned minutes. Note which shifts and products fail.
If the line beats the rate hour after hour, the rate is wrong. If one hour spikes, suspect the counter.
Take one shift. Count what went into boxes or onto pallets and match it to the machine count.
List every excluded minute and ask whether the line really stood still in it.
Then split the loss: the OEE loss tree puts each minute in one box, and all the boxes must add to the scheduled time. A sum that goes over tells you the same thing as the minute test.
The OEE calculator is handy for testing a corrected rate or count against the old one.
Should you cap OEE at 100%?
Overall Equipment Effectiveness (R. C. Hansen, 2001) caps the speed factor at 1.0 and notes any overspeed. So capping performance is standard practice.
A cap guards the report and repairs nothing. The wrong rate or count is still there, and it still flatters every shift that reads below the cap.
Our advice: cap the factor in the report, keep the raw reading visible, and treat any raw reading above 100% as an open data task.
Fixing the history
A corrected rate will drop the reported number. In the example, the line goes from 104.0% to 86.7% with no change on the floor.
Write down the date and the reason for the change, and tell the crews before it happens. Recalculate old shifts if you still have the counts, so the trend does not show a false fall.
Mistakes that keep it coming back
- Setting the ideal rate to whatever makes the number look sensible, which only hides the speed loss.
- Changing a format or a tool without updating the rate in the system.
- Moving a sensor without checking its count against a hand count.
- Reporting averages of shifts, which can bury one bad reading. See how to average OEE.
The rest of the common errors are in OEE calculation mistakes.
A hand-entered count or a stale shift record can push a reading over 100%. Fabrico takes stops, speed and counts live from PLCs or IoT sensors, with add-on AI cameras counting output where a machine has no usable signal, and shows OEE per line and shift, so you check the data against the machine. See how Fabrico connects to machines.
What people ask next
Can OEE ever really be over 100%?
No, not with correct inputs and a true ideal rate. If a line truly beats the rate for long periods, the rate is wrong and should be accredited again. See ideal cycle time.
Why is my OEE over 100% on only one product?
That usually means the ideal rate for that product or format is too low. Check it against the best sustained hour for that product, and compare with similar formats.
Can availability be over 100%?
Only if run time is longer than planned time, which means minutes were excluded while the line was running. Put those minutes back into the schedule. The availability calculator helps you recheck the sum.
Can quality be over 100%?
Not if good and total units come from the same point and the same period. It usually means good units were counted at the end of the line and total units at the machine, or that held or reworked units were added later. See how to calculate OEE quality.
Can I just raise the ideal rate until OEE looks normal?
No. If the real fault is a double-counting sensor, a higher rate hides it and makes every later reading wrong. Change the rate only after you have measured what the line can really hold, with the count checked by hand.
Fabrico is a light MES for OEE. It takes stops, speed and counts from PLCs or IoT sensors, with add-on AI cameras counting output where a machine has no usable signal, and operators tag the cause at the line.