Two ways to measure availability
| Time-based availability | Energy-based availability | |
|---|---|---|
| Question | For how much of the time could the turbine operate? | How much of the possible energy could the turbine produce? |
| Formula | Available time ÷ total time (with agreed exclusions) | Actual energy ÷ (actual energy + lost energy) |
| Strength | Simple, easy to audit | Reflects the commercial impact of downtime |
| Weakness | Ignores when downtime happens | Needs an estimate of lost production |
The IEC 61400-26 series defines both approaches and the categories that sit behind them. Many contracts use time-based availability for warranties and energy-based availability for performance reviews.
Who is responsible for the downtime?
Contracts usually split unavailable time into categories, because only some of it counts against the OEM or O&M provider:
- Turbine faults and repairs (usually the OEM's responsibility)
- Scheduled maintenance, sometimes within an allowance
- Grid outages and curtailment instructions
- Weather outside the design envelope: extreme wind, icing, lightning
- Owner-requested stops
Estimating lost production
- Take the wind speed during the downtime from the turbine's own anemometer, or from neighbouring turbines if its sensors were down.
- Look up the expected output on the turbine's reference power curve, adjusted for air density.
- Sum the expected output over the downtime period: that is the estimated lost energy.
A 3 MW turbine is down for 4 hours while neighbouring turbines record an average 9 m/s. The reference curve gives about 1.6 MW at that speed.
Lost energy ≈ 1.6 MW × 4 h = 6.4 MWh. The same 4 hours on a calm night might have cost almost nothing.
Underperformance that availability misses
A turbine can be 100% available and still lose energy if it runs below its power curve. Comparing measured output with the reference curve by wind-speed band, after filtering for curtailment, icing and data quality, highlights pitch, yaw or sensor problems that never trigger an availability alarm.
How nxtGrid helps
- Turbine-level generation, operating states, alarms and availability across farms, where the data is accessible
- Power-curve deviation analysis by wind-speed band
- Trends of temperatures and repeated fault patterns to support investigations
- Inspections and repairs assigned and tracked through maintenance workflows