Blog / OEE
OEE · Guide

OEE for Batch Processes: How to Calculate It

OL By OEE Lab Editorial |Updated 14 November 2026 |6 min read
Short answer

Batch OEE = (good output / ideal rate) / scheduled time, measured on the bottleneck vessel. Use kilograms or liters as the unit and the best accredited batch cycle as the ideal. In the reactor example, 27,440 good kg in 120 scheduled hours gives an OEE of 68.6%.

Batch OEE calculator

OEE, the hours lost to stops, speed and rejects, and the biggest loss, for a vessel with 1 to 10 logged stops.

Open the calculator

Can you use OEE on a batch process?

Yes. OEE works for batch, discrete and continuous processes. The three questions stay the same: how much of the scheduled time did the vessel make product, how fast, and how much was good.

What changes is the "unit". In a batch plant you count kilograms, liters or batches. And the clock is the batch cycle, not the cycle of a single part.

The method here follows Overall Equipment Effectiveness (R. C. Hansen, 2001), which sets the same rules for batch as for discrete work. For the general formula, read how to calculate OEE first.

Set it up for batches

  1. Pick the bottleneck vessel. Apply OEE where the plant is limited: the reactor, fermenter, mixer or oven that sets the pace. See bottleneck analysis.
  2. Choose the unit. Kilograms or liters are best when batch sizes vary. Batches work when they are fixed.
  3. Set the ideal rate. It is the best rate the vessel is accredited for. For a batch it is the nominal batch size over the shortest accredited batch cycle. If you do not know it, Hansen suggests the best 4 hours of the last 400. With cycles of several hours, take the fastest few good batches of the last few hundred hours instead.
  4. Count the schedule. Scheduled time is the time you planned to use the vessel. Weekends, planned shutdowns and "no product scheduled" are excluded time.
  5. Record every stop with a cause. Between and inside batches.

Fixed process steps such as reaction, cure or fermentation are part of the ideal cycle. They are not a loss. Only time beyond the recipe is a loss.

Worked example

Example: a reactor is scheduled for 120 hours in a week. Its ideal cycle is 6.0 hours for a 2,000 kg batch. That is an ideal rate of 2,000 / 6 = 333.3 kg per hour.

The stops in the week:

StopHansen categoryHours
Cleaning between campaignsStop time, operational6
Waiting for raw materialStop time, induced4
Waiting for lab releaseStop time, induced3
Pump failureDowntime, technical5
Total18

Run time = 120 - 18 = 102 hours. The reactor finishes 15 batches in that time, an average of 6.8 hours each.

It makes 29,400 kg in total (an average of 1,960 kg a batch). One batch of 1,960 kg fails the quality check. Good output: 29,400 - 1,960 = 27,440 kg.

FactorCalculationResult
Availability102 / 12085.0%
Performance(29,400 / 333.3) / 102 = 88.2 / 10286.5%
Quality27,440 / 29,40093.3%
OEE0.850 x 0.865 x 0.93368.6%
Check (good units method)(27,440 / 333.3) / 120 = 82.3 / 12068.6%
OEE (good kg at the ideal rate)68.6%
Slow batches and short yield (speed loss)11.5%
Waiting on material and lab (induced stops)5.8%
Cleaning between campaigns (operational stop)5.0%
Rejected batch (waste)4.9%
Pump failure (downtime)4.2%
The week as one bar. OEE plus the five losses add up to 100% of the 120 scheduled hours.

The losses in hours: speed 13.8 h, induced stops 7 h, cleaning 6 h, waste 5.9 h, downtime 5 h. Together 37.7 h, which is 31.4% of 120. And 68.6% + 31.4% = 100%.

Batch OEE calculator

Enter one period on the bottleneck vessel. The defaults are the reactor week above. Updates as you type.

Vessel and schedule
h
Time you planned to use the vessel. Leave out weekends, planned shutdowns and no product scheduled.
h
Shortest accredited cycle, charge to discharge, with fixed steps like reaction or cure.
kg
What one batch should make at the ideal cycle.
Kilograms or liters work best when batch sizes vary.
Stop log (1 to 10)

Every stop in or between batches, with a cause. Planned cleaning counts too, as an operational stop. Time beyond the recipe inside a batch, such as slow charging, heating or discharge, is not a stop: it shows up as speed loss.

Output
kg
Everything made in the period, good and rejected.
kg
Failed the quality check, or released and later found bad.
Splits the speed loss into slow and short batches.
Defaults: the 120-hour reactor week above.
Batch OEE
-%
Availability x performance x quality
0%85% = batch target
Availability
-
Performance
-
Quality
-
Biggest loss
-
Lost to stops
-
Lost to speed
-
Lost to rejects
-
Vs 85% target
-

Where the scheduled hours go

What is in the speed loss

The working

Losses are in ideal-rate hours, so OEE plus all losses is always 100% of scheduled time. If your own numbers do not add up, a loss is missing.

What is in the speed loss?

The performance of 86.5% hides two effects. Separate them, because the fixes differ.

  • Slow batches. The ideal for 15 batches is 90 hours. They took 102. The time factor is 90 / 102 = 88.2%.
  • Short batches. Each batch made 1,960 kg, not 2,000. That is a fill or yield factor of 98.0%.

88.2% x 98.0% = 86.5%. So heating, charging and discharging too slowly cost more than the yield shortfall. Look at the phase times inside the batch to find out which phase is slow.

Ideal: 15 batches x 2,000 kg in 90 h100%
After slow batches (90 h / 102 h)88.2%
After short batches (1,960 / 2,000 kg): performance86.5%
Performance 86.5% = 88.2% (time) x 98.0% (fill). Slow batches cost more than the yield shortfall.

Batch-specific traps

  • One bad batch is a big quality hit. With 15 batches a week, one reject is 6.7% of the count. Count in kg, and report over a longer period to smooth it.
  • Rejected product is waste, not downtime. Count it in units. Note the root cause. If product is released and later found bad, it goes into waste too.
  • Rework. If you blend off-spec product back in on the same vessel during scheduled time, treat that material as new input. Its time, speed and quality count in the current OEE.
  • Cleaning. A planned clean between products is a normal part of the schedule, so Hansen files it as operational stop time. Cleaning that runs over is still a loss to look at.
  • Waiting for lab results. If you wait longer than the planned test time, that is an induced stop. Name it so it can be fixed.
  • Several products on one vessel. Give each product its own ideal rate. Turn each product's good kg into ideal hours (good kg / its ideal rate), add the hours up and divide by scheduled time. Do not average the OEE of single batches.
  • Hold times. Waiting for a downstream tank to free up is an induced stop. It is a flow problem, not a vessel problem.

How good is 68.6%?

Hansen sets world-class targets by process type, building on Nakajima's TPM figures: 85% or more for discrete and batch processes, 90% or more for dedicated discrete and non-dedicated continuous, and 95% or more for dedicated continuous. He calls a result between 65 and 75 passable only when the trend is improving. So 68.6% is not bad, but it is far from the batch target. Read is 85% OEE realistic before you set a target.

Compare with OEE for continuous processes, where there is no batch cycle.

How to measure it on the floor

  • Use the batch record. It has start and end time, quantity and quality result.
  • Add a stop log with a cause for every wait, in or between batches. See how to track downtime.
  • Log phase times (charge, heat, hold, cool, discharge) to find the slow phase.
  • Check the numbers: OEE plus all losses must be 100%. If not, a loss is missing.

For the data options, see OEE data collection methods. To go on after the first numbers, use how to improve OEE.

See it live in Fabrico

Batch plants lose hours to waiting: for material, for the lab, for a free tank. Fabrico tracks stops per line and shift and operators tag the cause at the line, so each wait gets a name. See how Fabrico tracks stops and availability.

What people ask next

What is the ideal cycle time for a batch process?

It is the shortest batch cycle the vessel has been accredited to run for that product, from charge to discharge, including fixed process steps like reaction or cure. If you do not know it, use the best result you have seen. Keep it fixed per product family.

Should I measure OEE in kilograms or in batches?

Use kilograms or liters when batch size varies, as the yield is then part of performance. Use batches only when each batch has a fixed size. Both give the same OEE if the size is fixed.

Is cleaning between batches a loss?

Planned cleaning that is a normal part of the schedule is operational stop time in the Hansen scheme. It still counts against OEE, so shorten it with a better method. See SMED changeover reduction.

How do I count a rejected batch?

As waste, in units, with the root cause noted. Do not log it as downtime. A single batch can be a big share of the quality loss, so use kilograms and a longer period.

Should I measure OEE on every vessel?

Start with the bottleneck vessel, because only its gain raises plant output. Then move to the next critical one. See line OEE vs machine OEE.

See OEE per line and shift on your own machines

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.

Book a Fabrico demo

Related guides: How to Calculate Production Yield by Process · OEE for Bakery Lines: Ovens, Mixers, Proofers · MES vs DCS: Control Loop and Batch Record · OEE for Process Manufacturing: A Plant Guide

← More OEE Lab guides · How to calculate OEE →

OEE essentials: OEE calculator · What is OEE · Six big losses · Availability · TEEP · Cost of downtime · OEE benchmark · Free OEE tracker · Free OEE software · How to track downtime · How to implement OEE · Light MES vs full MES