OEE Calculation: Formula, Steps and a Worked Example
OEE calculation looks simple on paper. In real factory, the hard part is getting numbers you can trust. This guide uses one shift in a 15-machine wood factory to show the full calculation from start to finish. The example mirrors the kind of wood processing reality GlobalReader sees with customers like DaiΔΌrade Koks, where real-time data turns machine output into clearer decisions.
You will see the OEE formula, the three components, an Excel-friendly method, and where manual tracking starts to break down. If you already know the basics of OEE and just need the math, start here.
What is the OEE formula?
OEE = Availability Γ Performance Γ Quality
OEE, or Overall Equipment Effectiveness, shows how much of your planned production time turns into good output at the expected speed.
The full formula is:
OEE = (Run Time Γ· Planned Production Time) Γ ((Ideal Cycle Time Γ Total Count) Γ· Run Time) Γ (Good Count Γ· Total Count)
Use percentages or decimals, but do not mix both in the same calculation.
For example:
Availability: 86.7% = 0.867
Performance: 90.0% = 0.900
Quality: 95.0% = 0.950
OEE: 0.867 Γ 0.900 Γ 0.950 = 74.1%
Treat the final number as a map, not a judgement on your team.
The OEE calculation shows whether output is lost through downtime, speed loss, or scrap.
OEE calculator: check the number quickly
Use this simple OEE calculator layout when you want a fast check before building a full report.
| Input | What to enter | Example |
|---|---|---|
| Planned Production Time | Time the machine or line was scheduled to run | 6,750 machine-minutes |
| Run Time | Planned time minus unplanned stop time | 5,850 machine-minutes |
| Ideal Cycle Time | Fastest expected time per part in normal production | 1.5 minutes |
| Total Count | All produced units, good and rejected | 3,510 parts |
| Good Count | Units that passed quality | 3,335 parts |
Then calculate:
Availability: Run Time Γ· Planned Production Time
Performance: Ideal Cycle Time Γ Total Count Γ· Run Time
Quality: Good Count Γ· Total Count
OEE: Availability Γ Performance Γ Quality
Use the same inputs in the Excel template below if you want to compare machines, shifts, orders, or product groups.
The three OEE components at a glance
OEE is built from three separate questions.
| Component | Question it answers | Basic formula | What it points to |
|---|---|---|---|
| Availability | Did the machine run when it was planned to run? | Run Time Γ· Planned Production Time | Downtime, changeovers, waiting, breakdowns |
| Performance | Did the machine run at the expected speed? | Ideal Cycle Time Γ Total Count Γ· Run Time | Slow cycles, micro-stops, reduced speed |
| Quality | Did the machine produce good parts first time? | Good Count Γ· Total Count | Scrap, rework, defects, rejects |
The combined OEE number matters because each component points to a different shop-floor problem.
Downtime, speed loss, and scrap need different fixes, so do not treat 74.1% OEE as one single problem.
Availability
Availability measures how much planned production time was actually spent running. It turns downtime into a clear percentage that production and maintenance can discuss together.
If your machine was planned for 450 minutes but stopped for 60 minutes, availability is based on the remaining 390 minutes of run time.
In the full 15-machine example, those same minutes become machine-minutes so one long stop on one machine does not hide inside the shift average.
Performance
Performance measures speed loss while the machine is running.
A machine can be available all shift and still lose OEE when real cycle time drifts from the ideal cycle time. For deeper speed-loss math, read the dedicated guide to Performance in OEE.
Quality
Quality measures how many units passed inspection first time compared with total output. First-pass yield matters in high-mix production because rework still uses machine time, operator attention, and delivery capacity.
Rejected parts count against OEE because production time was used, but the factory did not get sellable output from that time.
For the wood factory example, 175 rejected parts reduce the quality score even though the machines already spent time producing them.
How to calculate OEE step by step
Use the same method every shift, with the same definitions for downtime, cycle time, and scrap.
The calculation only becomes useful when operators, planners, and managers agree on what each input means.
Quick formula tree: Availability, Performance, and Quality feed the final OEE result.
Availability: Run Time Γ· Planned Production Time
Performance: Ideal Cycle Time Γ Total Count Γ· Run Time
Quality: Good Count Γ· Total Count
OEE: Availability Γ Performance Γ Quality
Step 1: Define planned production time
Start with the time the machine, line, or work center was expected to produce. Then remove planned downtime before calculating OEE.
Before the shift starts, remove these agreed blocks from total shift time:
Planned breaks: For example, a 30-minute lunch break.
Scheduled maintenance: Remove maintenance windows agreed before the shift.
Team meetings: Remove briefings that were not production time.
Agreed shutdowns: Remove any shutdown already known before the shift.
For a single machine on a 480-minute shift with a 30-minute break, planned production time is 450 minutes.
For multiple machines, use machine-minutes.
In a 15-machine factory: 450 planned minutes Γ 15 machines = 6,750 planned machine-minutes
A common mistake is inconsistent planned downtime definitions. If one shift removes lunch breaks and another does not, Monday OEE and Tuesday OEE measure different things. Agree on the definition once, write it down, and use the same rule every shift. Once you have that baseline, the next job is to track every stop against it. This keeps the calculation fair when you look across a full shift.
Step 2: Capture run time and every stop
Run time is planned production time minus unplanned stop time.
Track unplanned stops by reason while the shift is running. For this wood factory shift, the stop log groups loss into four practical buckets:
Tool and changeover overruns
Waiting for material
Changeover adjustment
Micro-stops and short stops
For the wood factory shift, we will use this sample stop log:
| Stop reason | Lost machine-minutes |
|---|---|
| Tool change overrun | 300 |
| Waiting for material | 240 |
| Changeover adjustment | 180 |
| Micro-stops and short stops | 180 |
| Total unplanned stop time | 900 |
Now calculate run time:
6,750 planned machine-minutes - 900 stop minutes = 5,850 run minutes
Then calculate availability:
Availability = 5,850 Γ· 6,750 = 0.867 = 86.7%
Step 3: Count total output vs good output
Next, count everything produced.
Total output includes good and rejected units. Good output only includes units that passed quality and can move to the next process or customer.
For the same shift:
Total output: 3,510 parts
Rejected output: 175 parts
Good output: 3,335 parts
Ideal cycle time: 1.5 minutes per part
Now calculate performance:
Performance = (1.5 Γ 3,510) Γ· 5,850 = 5,265 Γ· 5,850 = 0.900 = 90.0%
Then calculate quality:
Quality = 3,335 Γ· 3,510 = 0.950 = 95.0%
Step 4: Put it together: full worked example
Now multiply the three components.
| Component | Result |
|---|---|
| Availability | 86.7% |
| Performance | 90.0% |
| Quality | 95.0% |
| OEE | 74.1% |
The calculation is:
OEE = 0.867 Γ 0.900 Γ 0.950 = 0.741 = 74.1%
That missing 25.9% is not just a percentage gap on a report.
In a 15-machine factory running two shifts, the gap can turn into hundreds of machine-hours per month that become overtime, late orders, or unused capacity.
This shift lost most output through availability first, then performance, then quality.
Each loss type points to a different fix:
Availability loss: Review stop reasons first: changeover overruns, material waits, and breakdown response time are the usual culprits.
Performance loss: Check speed settings, micro-stop frequency, and whether the ideal cycle time still reflects how the process actually runs.
Quality loss: Connect rejected parts to a specific product, material batch, machine, or shift.
Together, the three components show how much output was lost and which loss type to fix first.
If you want to run these four steps in a spreadsheet, the next section includes a template layout and the most common places manual Excel OEE breaks down.
OEE calculation in Excel: free template and where it breaks down
Excel is a not the best first step for OEE calculation. But it's an option many still use.
Most production managers start there because Excel is already available, easy to adjust, and good enough for a first baseline.
A simple OEE Excel template should include:
Planned production time: Shift time minus planned downtime.
Stop time: Every unplanned stop, grouped by reason.
Run time: Planned production time minus stop time.
Output: Total count, rejected count, and good count.
Ideal cycle time: The standard cycle time for the product or process.
Availability, Performance, Quality, and OEE: Automatic formulas for each component.
The important rule is simple.
Do not only track the final OEE percentage. Track the reason behind the loss.
Download the template
Use the free OEE Excel template as a clean starting point for manual calculation.
The template should help you calculate:
Availability from planned time and stop time.
Performance from ideal cycle time, total count, and run time.
Quality from good count and total count.
OEE from all three components.
You can use the template for one machine, one line, or a full shift.
If you compare several machines, weight the result by planned production time. Do not average machine percentages without checking the time behind each number.
Why manually logged stops are fiction
Manual OEE breaks down when stop reasons are logged after the fact.
A common example from sales calls is simple: an operator logs 15 minutes for a tool change that actually took 3 hours.
That does not always happen because someone is trying to hide the truth.
It happens because the operator is dealing with several jobs at once:
Fixing the machine
Finding tools
Checking quality
Answering questions
Getting production moving again
By the time the stop is written down, the number is a guess.
The most common logging errors in manual OEE are easy to recognise:
Short stops under 5 minutes are missed.
Changeover time gets split between outgoing and incoming orders.
Breakdown time is logged from fix completion, not from the actual stop.
Reason codes are picked at end of shift instead of when the stop happens.
Each error moves the OEE calculation in a different direction and makes improvement work harder.
That guess feeds into Excel, the daily report, and the improvement meeting.
The final OEE percentage looks exact. The stop data underneath it is not.
If you want to stop guessing at stop reasons, GlobalReader Operator captures downtime reasons at the machine while the event is happening.
The official OEE formula: what ISO 22400 actually defines
Many teams cite ISO 22400, but the useful part comes before the math: agreeing on what counts as planned time, unplanned downtime, speed loss, and scrap.
In plain language, ISO 22400 treats OEE as a KPI built from availability, performance, and quality.
ISO-style OEE formula:
OEE = Availability Γ Performance Γ Quality
The standard matters because it keeps the calculation consistent.
It helps teams avoid mixing OEE with related metrics. TEEP, or Total Effective Equipment Performance, extends OEE by including all calendar time, not just planned production time.
A machine with 74% OEE but only two scheduled shifts per week will show a much lower TEEP. Utilization and schedule adherence answer separate questions.
For practical factory use, the most important point is this:
Availability covers planned time that was lost to stops.
Performance covers speed loss while running.
Quality covers output that was produced but not good.
Use ISO 22400 as the reference for the structure of the formula.
Use your own agreed shop floor definitions for stop reasons, product cycle times, and scrap categories.
Simple OEE vs true OEE: the two-sensor method
Simple OEE usually uses basic production inputs.
You record planned time, stop time, output, rejects, and ideal cycle time. That gives you a useful first calculation, especially in Excel.
True OEE needs better source data.
A two-sensor method makes the number more reliable because it separates machine state from product flow.
| Method | What it captures | What can go wrong |
|---|---|---|
| Simple OEE | Manual stop time, output, rejects, and cycle time | Stops are guessed later, short stops get missed, reasons are inconsistent |
| True OEE with two sensors | Machine running state plus real production count or movement | Requires setup, but gives cleaner stop and speed data |
For most machine types, the first sensor monitors the motor, drive signal, or running state.
The second sensor monitors output through a proximity sensor, press counter, conveyor signal, or machine controller.
If the machine has a PLC, the controller output is usually the cleanest source for both signals without adding new hardware.
In practice, one signal shows whether the machine is running.
The second signal confirms whether product is actually moving, counting, or being processed.
That difference matters.
A machine can be powered and running, but not making good parts. A machine can also stop for short periods that nobody logs manually.
The two-sensor method helps catch both problems.
Whether you start with a spreadsheet or a two-sensor setup, the OEE number itself is only the starting point.
What to do with your OEE number
Printing 74.1% OEE on a report and calling it improvement wastes a useful calculation.
Start with three checks:
Find the biggest component loss: Availability, Performance, or Quality.
Sort losses by reason: Breakdown, changeover, waiting, speed loss, rejects, or rework.
Pick one improvement target: Choose the loss that costs the most time or blocks the most orders.
These three checks map to the Six Big Losses framework in ISO 22400:
Availability: Breakdowns, setup, and adjustment losses.
Performance: Minor stoppages and reduced speed.
Quality: Process defects and reduced yield.
Loss priority chart: Use the lowest component to pick the next job.
Availability is lowest: Fix downtime response and changeover control.
Performance is lowest: Check cycle standards, micro-stops, and speed settings.
Quality is lowest: Trace defects to product, material, machine, and shift.
Avoid comparing your number to generic benchmark tables too early.
Benchmarks can be useful later, but your first goal is to improve your own baseline with clean data.
Use the three component losses to decide where to act first: availability, performance, or quality. See the extra 10 strategies to improve your manufacturing process.
Automate the calculation with GlobalReader
Manual OEE calculation is useful when you are learning the method.
It becomes weak when you need daily accuracy across machines, shifts, orders, and products.
The worked example above is a starting point, not the end state. The 15-minute vs 3-hour tool-change story shows the gap: manual OEE gives you a baseline, while real-time data gives you trust.
GlobalReader helps manufacturers move the same OEE inputs out of paper and Excel and into live machine data.
Each module connects to a failure mode from the calculation:
Analytics: Tracks availability, performance, quality, OEE trends, and root causes in real time.
Operator: Captures downtime and scrap reasons while the event is happening, so a 3-hour changeover is not logged as 15 minutes.
Planner: Compares plan vs actual so production managers see order progress before the shift ends.
Maintenance: Connects downtime to machine history and tasks, so repeat stops become action items.
Smart Factory: Syncs production data with ERP and wider factory systems, reducing manual reporting gaps.
Based on reported customer results, OEE gains in the first year have ranged from 10% to 25%, with several customers reaching payback within 2 to 4 months.
Your subscription can start with the Starter Bundle, which includes hardware and Analytics from β¬109 per month per machine. You can then add Operator, Planner, Maintenance, and Smart Factory as your needs grow.
If you want to see the pricing structure and pick a starting point, browse GlobalReader pricing.
If you want to see the system first, start your free demo. No trial, no financial commitment. Sign in with Google or create a free account and you will be guided through with a quick tour.
FAQ
-
The OEE calculation formula is Availability Γ Performance Γ Quality. It measures how much planned production time becomes good output at the expected speed.
Use decimals in the final multiplication: 0.867 Γ 0.900 Γ 0.950 = 0.741, or 74.1% OEE, because percentage signs can confuse the math.
-
You calculate OEE step by step by calculating Availability, Performance, and Quality separately, then multiplying the three decimal results.
For the wood factory example in this guide, the shift has 86.7% Availability, 90.0% Performance, and 95.0% Quality. That gives 74.1% OEE.
-
A good OEE percentage depends on your process, product mix, machine age, and data quality. A make-to-order wood factory should not chase the same baseline as a fully automated high-volume line.
Use your first trusted OEE number as the baseline. Chasing the 85% world-class benchmark before your stop data is clean is one of the fastest ways to demoralize a production team.
Improve the biggest loss category first, then compare progress shift by shift.
-
Yes, you can calculate OEE in Excel when your sheet captures 5 inputs:
Planned production time
Run time
Ideal cycle time
Total output
Good output
Excel works for a first baseline, but stop logging becomes weak when operators record downtime from memory.
-
Simple OEE often relies on manual inputs: planned time, stop time, output, rejects, and cycle time. The spreadsheet formula can be correct while the source data is weak.
True OEE uses real machine and production signals, such as a two-sensor setup, to capture running state, stops, speed loss, and output more accurately.
-
Planned downtime is normally removed from planned production time before calculating OEE. That includes planned breaks, scheduled maintenance, team meetings, and shutdowns agreed before the shift.
Unplanned downtime stays in the calculation. Stops such as breakdowns, missing material, and changeover overruns show lost production time that the team can reduce.
The argument about whether a 10-minute team huddle counts as planned downtime is real. Settle the rule in writing before your first OEE review.
-
Your OEE number can change between Excel and software because the source data changes. OEE software captures short stops, real run time, and production counts more accurately than manual logs.
Excel depends on what operators remember and record. If a 3-hour tool change is logged as 15 minutes, the final percentage looks precise but points your team in the wrong direction.
