OEE vs Availability: Why High Availability Can Still Mean Low OEE

Key takeaways

• Availability is ONE of the three OEE factors (Availability x Performance x Quality), not OEE itself.
• A line can be 95% Available and still 60% OEE if it runs slow or scraps parts.
• Plants that report only Availability are blind to Performance loss (slow cycles, micro-stops) and Quality loss (scrap, rework).
• Availability typically drives 40-60% of total OEE loss in discrete plants. Performance and Quality together drive the rest — and are usually under-instrumented.
• Always report all three. Reporting only Availability creates a false sense of how well the line is doing.

Short answer: Availability measures whether the machine was running during planned production time. OEE multiplies Availability by Performance and Quality. A line can have excellent Availability and still mediocre OEE because slow cycles and scrap are invisible to Availability alone. Reporting only Availability is one of the most common ways manufacturers undercount their actual production losses. See also OEE vs Utilization vs Availability.

What Availability measures

Availability is the ratio of actual run time to planned production time.

Availability = Run time / Planned production time

If a shift had 8 hours of planned production time and the machine ran for 7.2 hours, Availability = 90%. The lost 0.8 hour is unplanned downtime — breakdowns, material starvation, changeover overrun.

Why Availability alone is not enough

A machine can be running at 100% Availability and still lose a third of its potential output. Two examples:

Slow cycles. The machine is on, but each part takes 45 seconds instead of the design 30. Availability says 100%; Performance says 67%. OEE drops to 0.67 from a perfect 1.

Scrap. The machine is on, running at full speed, but 8% of parts fail QC. Availability and Performance say 100% and 100%; Quality says 92%. OEE drops to 0.92.

Real-world OEE is the product of all three. Optimizing Availability without seeing Performance and Quality just shifts the loss to where you cannot see it.

The Availability trap

Availability is the easiest of the three factors to instrument. Most CMMS and SCADA systems can give you accurate downtime data. Performance requires ideal cycle time and actual cycle time. Quality requires part counts and scrap data. So plants instrument Availability first, often report only Availability, and call it OEE.

The result is a confident-looking number that hides the bulk of the loss. A plant reporting 90% Availability and calling it 90% OEE may actually be running 60% true OEE — losing 30 points of capacity to Performance and Quality issues nobody is measuring.

What good measurement looks like

True OEE needs:

• Availability — actual run time vs planned production time, with documented stop reasons.
• Performance — actual cycle time vs ideal cycle time, ideally per SKU or per recipe.
• Quality — good parts vs total parts produced, with scrap and rework distinguished.

If you cannot calculate all three, you do not have OEE. You have a partial number.

Where Performance loss usually hides

• Micro-stops under the threshold most systems log as downtime — pauses of a few seconds that accumulate.
• Idle time while operators clear jams or wait for material.
• Reduced speed on aging tooling or older recipes that nobody updated.

These losses usually show up as a Performance number well below 100% on lines that report 90%+ Availability.

Where Quality loss usually hides

• First-pass yield losses — parts that need rework but eventually pass and are not counted as scrap.
• Start-up scrap at the beginning of a run that gets bucketed as planned waste instead of Quality loss.
• Customer returns not fed back to the line as Quality data.

How an OEE platform measures all three

A real OEE platform connects to PLC or machine signals for Availability (run/stop state), Performance (cycle count and time), and Quality (good/bad part counts). It computes all three in real time, shows the breakdown, and points at the dominant loss so the right action gets taken.

Fabrico's OEE module measures all three factors at line cadence and surfaces which one is the dominant loss for each asset — so the team works on the right problem, not the easiest one.

See how Fabrico captures this automatically — explore OEE for manufacturing or book a demo.

Related reading

• OEE vs Utilization vs Availability
• OEE vs Utilization
• OEE and ISO 22400
• OEE for Batch Manufacturing

Frequently asked questions

Is Availability the same as utilization?

They are close. Utilization is usually measured against scheduled time; Availability inside OEE is measured against planned production time, excluding planned breaks.

Why is my Availability high but my OEE low?

Because the line is running slowly (Performance loss) or scrapping parts (Quality loss). Look at the breakdown — the dominant loss is whichever factor is furthest from 100%.

Can I report Availability as my OEE?

Only if you confirm Performance and Quality are both at 100%. That is rare. Otherwise the Availability number overstates the OEE.

What is a good Availability target?

World-class is around 90%. The OEE world-class benchmark of 85% assumes Availability ~90%, Performance ~95%, Quality ~99%.

How do I improve Availability specifically?

Cut unplanned downtime: preventive maintenance, faster changeover (SMED), better material flow, faster mean-time-to-repair. Available time is mostly a maintenance and material problem.