Are power curves affected by wake?
A turbine in the wake of another sees slower wind, so it should make less power. Yet plotted against its own anemometer, one Hill of Towie turbine shows 16.5% more power in the wake at the same measured wind speed.
A turbine in the wake of another turbine sees slower wind and makes less power. That loss is real. The power curve plotted against the turbine’s own nacelle anemometer tells a different story.
Splitting almost three years of data from one Hill of Towie turbine between waked and unwaked periods shows the waked curve sitting higher, not lower.
The waked curve sits higher below rated
At the same measured wind speed, waked power is 16.5% higher between 4 and 11 m/s. Above rated the two curves meet, because the turbine is limiting its own output.
The turbine is not producing more energy in the wake. It is producing more power than its anemometer’s wind speed suggests.
The anemometer sits in the slowest part of the wake
A wake can be thought of as a cone of slow air behind the upstream turbine, slowest at its centre and recovering towards its edges.
When the wind blows along the line between two turbines, the downstream nacelle anemometer sits on the wake centre line at hub height. The outer part of the rotor sees faster air than the anemometer measures, so the anemometer under reads the wind the rotor is using.
In this illustrative wake the anemometer sees 79% of the free stream wind and the rotor as a whole sees 83%, a gap of about 5%. Below rated, power rises with roughly the cube of wind speed, so a 5% gap in wind becomes roughly 17% in power. On this turbine, binning the data by power rather than by wind speed puts the waked anemometer about 6% low.
Turbulence adds to the effect. The nacelle anemometer reads a turbulence intensity of about 0.23 in the wake against 0.16 out of it. Higher turbulence increases power at lower wind speeds and reduces it around the knee of the power curve. Turbulence from a nacelle anemometer is unreliable record by record, as covered in whether a nacelle anemometer can measure turbulence, so this shows the direction of the effect rather than its size.
What the data shows
- On one turbine, waked power is 16.5% higher than unwaked power at the same measured wind speed, between 4 and 11 m/s.
- Across 19 turbine pairs on the site, waked power is higher in every pair, from 2% to 37% with a median of 21%.
- Binned by power, the waked anemometer on this turbine reads about 6% low.
- Roughly 37% of this turbine’s classified records are waked.
Why this matters for performance assessment
This is a measurement effect, separate from the real wake loss, which still reduces the output of the downstream turbine.
IEC 61400-12-1 excludes waked sectors from power curve tests. That makes sense when testing what a turbine model can produce. It does not suit an operational assessment of what a turbine will produce, where waked periods can be a large share of the data. A power curve that mixes the two makes a turbine look better than it is, and a change in the share of waked data between two periods can look like a change in performance.
PowerVeritas keeps all operational data in its Performance Reviews, then categorises and quantifies each major loss, so wake effects are measured rather than filtered away.
Charts and analysis by PowerVeritas. Where open datasets are used, sources are credited on the attributions page.