Overall Equipment Effectiveness (OEE) is calculated by multiplying three component rates: Availability, Performance, and Quality. The OEE formula is: OEE = Availability × Performance × Quality. Each component is expressed as a decimal between 0 and 1, and the result converts to a percentage for reporting.
The formula itself is straightforward. What trips up most operations teams is not the multiplication, it's the inputs.
Two specific errors appear repeatedly in practice: misidentifying ideal cycle time in the performance calculation, and including planned downtime in the availability calculation. Both errors produce an OEE score that looks credible but points improvement effort in the wrong direction.
Key Takeaways
- OEE is only meaningful when planned production time is used as the baseline for all three component calculations; using total calendar time or shift time without excluding scheduled breaks produces a figure that cannot be compared across shifts or sites.
- Misidentifying ideal cycle time causes the performance rate to appear higher than it actually is, hiding speed losses that are often the largest contributor to below-target OEE.
- Planned downtime, including scheduled maintenance and planned changeovers, must be excluded from the availability calculation before it begins; coding these events as unplanned downtime systematically understates true availability.
- Averaging individual machine OEE scores across a production line does not produce a valid line-level figure; the constraint machine's OEE determines total line output, and averaging obscures the actual bottleneck.
What OEE Measures and Why the Formula Has Three Parts
OEE is a compound metric built from three independent rates, each measuring a different category of productive loss. The structure is not arbitrary. The six big losses identified in Total Productive Maintenance map directly onto these three components, so the formula reflects a structured loss classification system, not a convenient arithmetic convention.
Understanding which loss each component captures is essential before applying any numbers.
Availability, what it measures and how it differs from uptime
Availability is the ratio of run time to planned production time, where run time is planned production time minus unplanned downtime only. Planned downtime does not reduce run time in the standard OEE calculation, which is a critical distinction: a naive "uptime" figure that lumps all downtime events together will understate availability and misrepresent equipment downtime losses.
Performance, what it measures and why ideal cycle time matters
Performance is the ratio of actual output to the maximum possible output achievable during run time, and that maximum is derived from ideal cycle time, the fastest rate the equipment is designed or has been demonstrated to sustain. This is where the most common calculation error occurs, addressed directly in the pitfalls section below.
Quality, what it measures and how rework is treated
Quality is the ratio of good parts produced to total parts produced. Total parts includes both scrap and rework, since parts requiring rework consume productive run time even if they're ultimately accepted, and excluding rework from the denominator overstates the quality rate while understating defect-related losses.
The OEE Formula
The master formula and its three sub-formulas are presented below. All components must be calculated before the final multiplication.
- OEE = Availability × Performance × Quality
- Availability = Run Time / Planned Production Time
- Performance = (Ideal Cycle Time × Total Parts Produced) / Run Time
- Quality = Good Parts Produced / Total Parts Produced
All three components are decimal values between 0 and 1, and OEE is calculated the same way before being converted to a percentage for reporting.
A world-class OEE benchmark of 85% gets cited often, but that figure originated in discrete manufacturing contexts, so treat it as a reference point rather than something to chase as a universal target.
Sub-formulas for each component
Each sub-formula captures a specific loss category. Availability divides the time the equipment actually ran by the time it was scheduled to run. Performance divides actual output by the theoretical maximum output possible at ideal speed. Quality divides conforming output by everything the line produced, including scrap and rework.
What OEE of 100% would mean in practice
An OEE of 100% would mean only good parts produced, at maximum speed, with no unplanned stops. This is a theoretical ceiling. No production environment achieves it consistently. Its value is as a reference point that makes clear how far from lossless any real line operates.
Worked Example, An Australian Packaging Line
Formulas make more sense with real numbers behind them, so here's the calculation applied end to end, using figures typical of an Australian production line rather than an abstract example.
The following example applies the OEE formula to a beverage filling line running a standard day shift in an Australian food manufacturing facility. Work through each step in sequence before combining the components.
Scenario parameters
The scenario inputs are as follows:
- Shift length: 480 minutes
- Scheduled breaks (excluded from planned production time): 30 minutes
- Planned production time: 450 minutes
- Unplanned downtime (conveyor fault, mid-shift): 45 minutes
- Run time: 450 - 45 = 405 minutes
- Ideal cycle time: 1.0 minute per unit
- Total parts produced: 360 units
- Good parts produced: 324 units (36 units rejected as scrap or rework)
Step-by-step calculation
Work through each component in order, then multiply.
- Calculate Availability. Run Time / Planned Production Time = 405 / 450 = 0.90 (90%)
- Calculate Performance. (Ideal Cycle Time × Total Parts Produced) / Run Time = (1.0 × 360) / 405 = 360 / 405 = 0.889 (88.9%)
- Calculate Quality. Good Parts Produced / Total Parts Produced = 324 / 360 = 0.90 (90%)
- Calculate OEE. 0.90 × 0.889 × 0.90 = 0.72 (72%)
An OEE of 72% on this line shows no single dominant loss category. Availability and Quality are each losing exactly 10% of potential output, while Performance is losing slightly more, at 11.1%.
The performance rate at 88.9% warrants particular attention: if ideal cycle time has been misidentified, that figure may be flattering a more significant speed problem. The quality rate at 90% represents 36 rejected units per shift, which compounds the speed loss into a meaningful throughput gap.
Two Common Mistakes That Distort Your OEE Score
These two errors appear frequently when organisations begin measuring OEE without a structured measurement framework in place. Each produces a number that passes a basic reasonableness check but actively misleads root-cause analysis.
Using average cycle time instead of ideal cycle time
This one is easy to get wrong because the mistake feels reasonable at the time. Reaching for the number you already have, the current average, seems like the practical shortcut, but it quietly builds today's speed problems into tomorrow's baseline.
- Ideal cycle time is the maximum speed the equipment is designed or has been demonstrated to sustain, taken from the manufacturer's spec or the best sustained production run.
- Using the current average cycle time instead bakes existing speed losses into the baseline, which makes the performance rate look better than it actually is.
- The result: speed-related losses, one of the six big losses in the TPM framework, disappear from the OEE measurement entirely, and improvement effort gets misdirected toward downtime or quality instead.
Including planned downtime in the availability calculation
This second error is less about the wrong number and more about where a number gets filed. Once a planned event is coded as unplanned, the whole downstream analysis points at the wrong problem.
- Planned downtime, including scheduled maintenance, planned changeovers, and regulatory shutdowns, is excluded from planned production time before the availability calculation begins.
- Coding planned events as unplanned downtime instead understates the availability rate.
- The consequence: the organisation chases equipment reliability improvements and directs continuous improvement consulting resources toward the wrong loss category, when the real opportunity is in the planned events themselves, whether that's a SMED programme or better maintenance scheduling.
Refer to your maintenance and production planning records to verify the downtime classification before finalising any availability figure. A consistent, documented rule for what counts as planned versus unplanned is what makes the classification defensible, not a specific external framework.
A Note on Line-Level OEE
A single-machine OEE figure is not the same as a line OEE figure. Averaging individual machine scores across a multi-machine or interlinked production line does not produce a meaningful line-level result, and the error it introduces is not trivial.
Why averaging machine OEE scores misleads
The constraint machine sets the ceiling for the whole line, no matter how well every other asset is running. Take a line where one machine sits at 60% OEE while every other machine runs at 85%: total throughput is still capped by that one machine at 60%, not lifted by the rest.
Average the scores instead, and you get something closer to 80%, a figure that makes the line look far healthier than its actual output confirms. That's the real danger: averaging doesn't just blur the picture; it points improvement effort at the wrong asset entirely and hides how big the true efficiency gap really is.
Value stream mapping is one structured way to find that constraint asset on an interlinked line before you apply OEE measurement to it.
How to approach measurement on an interlinked line
Line-level OEE should be anchored to the constraint asset. Identifying the true constraint often requires a structured line diagnostic rather than a formula applied in isolation to each machine.
How OE Partners Supports Line Performance Measurement and Improvement
Calculating OEE correctly is a starting point, not an improvement strategy. Many operations teams apply the formula accurately but then interpret the result incorrectly, act on the wrong loss category, or have no structured pathway from a measurement to a prioritised improvement project. This is where manufacturing efficiency gains are left on the table.
From measurement to structured line improvement
OE Partners works with manufacturing operations to assess line performance, identify the dominant loss categories driving below-target OEE, and structure an improvement approach that addresses root causes rather than symptoms.
For teams that already have a measurement foundation in place and just need to act on it, the lean manufacturing consulting service provides that structured diagnostic and improvement pathway. For organisations thinking further ahead, about how improvement capability gets built and sustained internally, the operational excellence consulting service addresses the broader framework line metrics sit within.
What a line diagnostic typically examines
A structured line performance diagnostic moves beyond the OEE number to examine what is driving it. A typical diagnostic covers:
- Loss classification validation, confirming downtime events are coded correctly as planned or unplanned
- Constraint identification, determining which asset limits total line throughput
- Cycle time analysis, verifying ideal cycle time inputs against actual equipment capability
- Planned versus unplanned downtime categorisation, separating managed losses from reliability failures
- Improvement priority sequencing, ranking loss categories by recoverable output and implementation feasibility
These are the practical next steps once a team has confirmed its OEE calculation approach is sound. In practice, organisations that get the most out of OEE treat the figure as a diagnostic input to guide improvement work, not as the finished output of a measurement programme.
Let's Recap
- OEE is calculated by multiplying Availability, Performance, and Quality, with planned production time as the baseline for all three components.
- Ideal cycle time must reflect the equipment's designed or demonstrated best rate, not the current average; using the average inflates the performance rate and hides speed losses.
- Planned downtime must be excluded from the availability calculation before it begins; including it as unplanned downtime understates true availability and misdirects improvement effort.
- Averaging individual machine OEE scores across a production line obscures the constraint asset and overstates line performance.
- A correctly calculated OEE figure is the starting point for structured improvement, not the conclusion; the value is in what the three component rates reveal about which loss category to address first.
Find Out Where Your Line Is Losing Productive Time
If your OEE figures aren't translating into clear improvement priorities, speak with an OE Partners consultant about a structured line performance diagnostic. We'll focus the conversation on loss classification, constraint identification, and improvement priority, not a generic consulting pitch.
Frequently Asked Questions
What is a good OEE score for an Australian manufacturing operation?
A frequently cited reference point is 85%, drawn from discrete manufacturing benchmarking in Total Productive Maintenance literature. Your own target should reflect your industry, equipment type, and current loss profile, not a universal figure. A food processing line with significant planned changeover time will structurally score differently from a high-volume automated packaging cell.
How does planned downtime affect the availability calculation in OEE?
It doesn't, in the standard OEE formula. Planned downtime is excluded from planned production time before the calculation even begins, so it never shows up in the availability denominator or numerator. If planned events get coded as unplanned downtime instead, your availability rate comes out understated and your improvement analysis follows it off course.
What is the difference between OEE and TEEP?
OEE measures performance against planned production time, while TEEP, Total Effective Equipment Performance, measures it against total calendar time, including all planned downtime and non-scheduled periods. TEEP is the more demanding of the two, reflecting total asset utilisation potential rather than in-shift operational effectiveness.
Should changeover time be treated as planned or unplanned downtime in an OEE calculation?
Planned changeovers count as planned downtime and are excluded from planned production time before the availability calculation begins. Only changeovers that run over their planned duration, or ones that weren't scheduled at all, generate an unplanned downtime event counted against availability.
How do you calculate OEE across multiple machines on the same production line?
You don't average the individual machine scores. The constraint machine's OEE governs total line output, so anchor your line-level measurement to that asset, and use a structured diagnostic to confirm which one it actually is before applying the formula.
What is ideal cycle time and how do you determine it for a piece of equipment?
Ideal cycle time is the fastest rate the equipment can sustain under normal operating conditions, taken from the manufacturer's design specification or from the best sustained production rate in your own operating data. It's not the current average cycle time, which already has existing speed losses baked into it.
When is OEE not the right metric for measuring line performance?
OEE is built for equipment-constrained environments, where one defined asset drives throughput. It's less meaningful for highly variable, low-volume, or project-based production, where planned production time is irregular. In those environments, performance frameworks tied to flow or lead time are usually a better fit.
