You can calculate OEE for a generating unit, but it is rarely the right main measure. Generators already report equivalent availability factor (EAF), forced outage rates (EFOR, EFORd) and net capacity factor (NCF), and they use heat rate for fuel efficiency. OEE adds little over these, though it can help on fuel-handling and other production-style lines in a plant.
Why a generating unit is not a production line
A power unit has no scrap bin, and it does not decide how much to make. The grid or the market often decides that, so a perfect unit can sit idle for days.
The industry solved this long ago with its own definitions. In North America, NERC's Generating Availability Data System (GADS) publishes the equations, and its training material ties them to IEEE Std 762.
The measures plants already use
NERC's training module on factors and rates draws a useful line. Factors are generally based on period hours, and rates are generally based on service hours, the hours the unit ran.
| Measure | Equation (simplified) | What it tells you |
|---|---|---|
| Availability factor, AF | Available hours / period hours | Could it run, whether or not it was needed? |
| Equivalent availability factor, EAF | (Available hours - equivalent derated hours) / period hours | The same, with derates counted |
| Service factor, SF | Service hours / period hours | How much of the period it ran |
| Net output factor, NOF | Net generation / (service hours x net max capacity) | How hard it ran while running |
| Net capacity factor, NCF | Net generation / (period hours x net max capacity) | Share of maximum net energy produced |
| Forced outage rate, FOR | Forced outage hours / (forced outage + service hours) | Share of its would-be running time lost to forced outage |
| EFOR and EFORd | FOR with derates added; EFORd adjusts for demand | Chance of forced outage or derate, with EFORd asking "when demanded to run" |
Fuel efficiency sits outside all of these. The US Energy Information Administration defines heat rate as a measure of thermal efficiency, commonly stated as Btu per kilowatthour.
How these map to OEE
Here is the neat part. Net capacity factor already has the shape of OEE with quality set at 100%, because NCF = SF x NOF.
The service factor plays the part of availability, and the net output factor plays the part of performance. Only the time base is different, since NCF uses all period hours.
Worked example: one unit, one month
Example: a 500 MW unit in a 720-hour month ran 600 hours, sat on reserve shutdown for 72 because nobody needed the power, and was on forced outage for 48. It generated 270,000 MWh net, and derates add up to 14 equivalent hours.
To get an OEE-style number, take out the 72 reserve hours as "no demand". That leaves 648 planned hours, so availability is 600 / 648 = 92.6%, performance is 90.0%, quality is 100% and OEE is 83.3%.
Part of that 90.0% may be dispatch, not a problem at the unit, because the same factor holds derates and part load the grid asked for.
That is the same treatment as OEE with no orders. It tells the plant nothing that EAF, EFOR and NCF did not already say, and they have shared definitions behind them.
Why a generating unit strains the formula
- Dispatch. OEE needs a rule for reserve shutdown, part load and ramping. The industry's factors and rates already separate these.
- Quality. Energy delivered at the meter has no reject count that behaves like factory scrap, so the factor adds nothing.
- Fuel. Same output at a heat rate of 10,500 instead of 10,000 Btu per kWh burns 5% more fuel. In this example that is 135,000 MMBtu more, and OEE would not move.
- Comparisons. NERC notes that with pooled statistics, unweighted figures treat a 50 MW gas turbine and a 1,000 MW nuclear unit alike, so it provides capacity-weighted versions. A plain OEE average repeats that mistake.
- Wind and solar. Output follows the resource, so low capacity factor can simply mean a calm month. NERC keeps separate GADS instructions for wind.
OEE does get discussed for thermal plants. The abstract of Souza's introduction to Thermal Power Plant Performance Analysis (Springer, 2012) describes it as a key performance indicator, linked to reliability and maintainability.
That is a reasonable way to explain plant losses to non-specialists. Still, we found no neutral published OEE benchmark for generating units, so treat any percentage you see for one as unverified.
Where OEE can help
A plant holds several smaller systems that behave like production lines. OEE suits those because they have a rate, a run state and a usable output.
| Asset | Use OEE? | Why |
|---|---|---|
| Biomass pellet or fuel preparation line | Yes | Tons per hour and on-spec fuel, a normal continuous line |
| Coal mills, conveyors, ash handling | Sometimes | Useful if they limit output, in tons per hour. Their load follows the unit, so exclude no-demand time. |
| The generating unit | No | Use EAF, EFOR or EFORd, NCF and heat rate |
| A factory's own boiler or generator set | No | It follows site demand. Log its failures as stops on the lines they hit. |
If you are a factory with your own power, the last row matters most. A boiler trip is an upstream cause of line stops, so code it that way in your line's log.
If management insists on one number
Show NCF split into SF and NOF, and put EAF and EFOR next to it. People who know OEE will read the first two as availability and performance within a minute.
If it helps, explain the idea through OEE vs TEEP, since NCF's calendar base is close to how TEEP treats time. For failure patterns, MTBF and MTTR still apply, and the calculator sizes them.
Keep the reliability work on the systems behind the numbers. Reliability-centered maintenance gives a starting framework for that.
Fabrico is a light MES for OEE on production lines, not a generation reporting tool. For a fuel preparation line or a factory hit by utility failures, it shows stops, speed and counts live from PLCs or IoT sensors, with OEE per line and shift. See how Fabrico measures OEE on production lines.
What people ask next
What is OEE for a power plant?
It is availability x performance x quality applied to a generating unit, usually with quality at 100%. In practice it looks much like net capacity factor with no-demand hours taken out, as the example above shows. The general formula is in what is OEE.
What is the difference between EAF and net capacity factor?
EAF says how much of the period the unit could run, after derates, whether or not it was needed. NCF says how much of the maximum energy it actually made, so low demand lowers NCF but does not lower EAF. Planned vs unplanned downtime covers the related factory idea.
What does EFORd mean?
NERC describes it as the probability of being on forced outage or derate when demanded to run. It suits peaking and cycling units, where plain time averages say less, and the full equation is in NERC's Appendix F. For factory failure measures, see the MTBF and MTTR calculator.
Is there a world-class OEE for power plants?
We found no neutral published figure. The usual 85% traces to Nakajima and was not built for generation, so read is 85% OEE realistic before borrowing it.
Can OEE include heat rate?
Not in the standard formula, which has no factor for fuel. Some teams show it beside the unit's other numbers, which is the honest way. OEE vs TEEP shows how far the time-based family goes.
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.