The OEE formula does not change on a 12-hour shift, but a fair comparison does: compare crews only over a full turn of the rotation, so each has worked the same mix of days, nights and weekends. Then compare loss rates, such as stop minutes per 100 scheduled hours, because a 12-hour shift collects more loss minutes than an 8-hour one simply by being longer.
What a long shift changes
A 12-hour pattern adds three problems that an 8-hour pattern hides. One crew runs longer, crews rotate through days and nights, and one week of data catches only part of the pattern.
The causes of a gap between shifts are covered in why OEE differs between shifts. This page is about making the comparison fair before you look for causes.
Fix the scheduled time first
A 12-hour shift is 720 minutes on the clock, but what you schedule depends on your break rule. Decide it once, for every crew and line.
Example: a 720 minute shift has 60 minutes of breaks. If the line stops for breaks and your rule excludes them, scheduled time is 660 minutes; if a relief crew keeps it running, it is 720.
Practice varies on breaks, and how to count breaks in OEE weighs the options. What cannot vary is the rule between crews, or the comparison means nothing.
Why one week of data misleads
Rotating crews do not work the same hours in the same week. In one week crew A may be on days while crew C is on nights, with different support and different problems.
Example: four crews work a pattern where each spends half its time on days and half on nights. On this line, days run at about 77% and nights about 71%, whichever crew is on.
| Crew A | Crew C | Gap | |
|---|---|---|---|
| One week: A on days, C on nights | 77.0% | 71.0% | 6.0 points |
| Full rotation, day shifts | 77.0% | 76.0% | 1.0 point |
| Full rotation, night shifts | 71.0% | 71.0% | 0.0 points |
| Full rotation, all shifts | 74.0% | 73.5% | 0.5 points |
So find how long your pattern takes to repeat, and compare only whole repeats. Patterns that swap crews between days and nights can take weeks to come round.
A 4-on, 4-off pattern with crews fixed on days or nights repeats every 8 days, but no repeat evens out days and nights there. Compare day crew with day crew, and night crew with night crew.
Several whole repeats beat one, since one bad breakdown can still colour a single turn. For combining them, weight by scheduled time as in how to average OEE.
Compare rates, not minutes
Many plants mix patterns, such as 8-hour weekdays and 12-hour weekends. Then minutes of loss cannot be compared directly.
Express each loss per 100 scheduled hours: stops, short stops, changeovers and scrap. In the example above, that is 10 and 9 stop hours per 100.
The same trap works on one-time events. A 20 minute handover is 4.2% of 480 minutes and 2.8% of 720, so a 12-hour shift scores better on that loss even if its crew hands over no better.
A move from three 8-hour shifts to two 12-hour shifts is a real gain on that one loss, as there are fewer handovers per day: 60 minutes down to 40. To know whether the whole pattern change helped, compare whole days or weeks, not shift scores from before and after.
Look inside the shift
A 12-hour shift is long enough to have a shape. A plain shift total hides it, so split the shift by hour or by block.
Check the total: (120 × 0.62 + 480 × 0.78 + 120 × 0.70) ÷ 720 = 532.8 ÷ 720 = 74.0%. A single shift number of 74% shows none of this.
The shape points to the next step. A low start points to handover and restart, and first-hour losses and the shift handover checklist both help there.
A low end needs care in how you read it. It may be tiredness, but it may also be a rush to clean up early, so check end-of-shift losses and ask the crew before you decide.
Measure the shape by hour on an hour-by-hour board. Do not use a guess or an end-of-shift note, because after 12 hours a memory of what happened at hour 2 is poor.
A fair comparison, step by step
Same break rule, same ideal rate per product and same stop threshold for all crews.
Compare complete repeats of the pattern, and more than one if you can.
Set days against days and nights against nights. Note weekends and support cover.
A crew that drew the hard products needs a rate per product. See the mixed products guide.
Share the loss rates with both crews and ask what differed. Copy what the better shift does.
If one crew always draws the hard products, the mix will drive the gap. OEE with mixed products shows how to keep that out of the score.
Common mistakes on long shifts
- Different break rules by crew. One crew excluding breaks can look several points better.
- Ranking crews weekly. People soon learn the league table depends on the calendar.
- Night support left out. If technicians, material or quality checks are thin at night, a stop runs longer. That is a plant decision, not a crew failing.
- Using OEE to judge the pattern alone. Labour cost and output also change, so look at overall labor effectiveness too.
- Mixed shift lengths in a simple average. A 12-hour shift should carry more weight than an 8-hour one.
A fair crew comparison needs the same clock on every shift, by day and by night. Fabrico takes stops, speed and counts live from PLCs or IoT sensors, operators tag the cause at the line, and OEE shows per line and shift. See how Fabrico tracks stops.
What people ask next
Do I calculate OEE differently on a 12-hour shift?
No. The formula is the same, with 720 minutes or your agreed scheduled time as the base. Try it in the OEE calculator.
Do 12-hour shifts lower OEE?
We found no neutral study that settles this, so measure it on your own lines. Compare OEE by hour of shift and by whole week, before and after any change of pattern.
How long should I compare rotating crews for?
At least one full repeat of the pattern, and several if you can. A single repeat can still be swayed by one breakdown, so look at the downtime Pareto before drawing conclusions.
Should night shift have a lower OEE target?
We would keep one target for the line, and raise support at night if that is the cause. A lower target makes the gap look acceptable. See how to set OEE targets.
Should breaks count on a 12-hour shift?
That depends on whether the line stops, and practice varies. Choose one rule and use it for all crews, as covered in how to count breaks in OEE.
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.
Related guides: Units Per Labor Hour: Formula and Use With OEE