Wind, like solar, can be run standalone, co-located with a solar site, or tri-located alongside solar and a battery.
Wind generation is always taken from a custom profile that is uploaded for the site. As with solar, the model reshuffles the uploaded profile so that it lines up with the weather year behind the power prices.
Wind Representation
Wind assets are represented with a small set of parameters:
- Nameplate Capacity (MW): The rated power of the wind site.
- Custom Generation Profile: A user-uploaded, site-specific generation time series.
- Site Location: The latitude and longitude of the site, used to look up local wind speeds for the reshuffle.
- Grid Export Limit (MW): The maximum power the site can export.
Generation Modelling
Wind sites do not use a modelled generation profile. Instead, each site requires a custom wind profile - a full year of the site’s actual generation. See Custom Generation Profile for the required file format.
Why a custom profile is required: Wind output is far more site-specific than solar. It depends on turbine sizing, turbine spacing, site and land layout, and local geography, all of which are difficult to model accurately from a central set of assumptions. The site owner understands these factors better than a general model can, so the model relies on an accurate profile provided for the site rather than generating one.
Capturing Cannibalisation
As wind penetration increases, wholesale prices tend to be lower when the wind is blowing. This effect, known as cannibalisation, reduces the capture price for wind generators relative to the baseload price.
To capture cannibalisation accurately, the wind profile must line up with the weather year used in the Fundamentals Model. To ensure this, the model reshuffles the uploaded profile onto that weather year. This is done in 4 steps:
- Get the hourly wind speed at 100 m at the site for the weather year, from ERA5 reanalysis data.
- Rank every period of the weather year from calmest to windiest.
- Rank every period of the uploaded profile from lowest to highest output.
- Give the lowest output to the calmest period, the next-lowest output to the next-calmest period, and so on.
Wind is reordered period by period across the full year, as solar is in the USA and the NEM. The profile keeps every output value, so its annual energy and duration curve do not change.
The upload therefore does not need to come from the weather year, but it should reflect the expected long-term yield of the site. The weather year for each market is documented below:
| Market | Weather year documentation |
|---|---|
| ERCOT | Generation |
| WECC | Generation |
| SPP | Generation |
| Eastern Interconnection | Generation |
| UK & Europe | Renewables |
| The NEM | Reference Years |
Co-located and Tri-located Sites
Wind can be combined with other assets at the same grid connection:
- Wind co-located with solar: Both assets export into the shared grid connection.
- Wind tri-located with solar and a battery: Wind and solar export into the connection, and the battery is optimised alongside them.
Wind is co-located in the same way as solar, and can be AC- or DC-coupled with the battery. On a DC-coupled site, wind can charge the battery when the export limit would otherwise curtail it. See Co-location for the full coupling logic, which applies identically to wind.
The battery is optimised around the wind and solar generation, with the combined export of wind, solar, and battery kept within the grid export limit.
As with co-located solar and storage, the forecast reports combined revenues for the whole site rather than splitting them between the wind, solar, and battery assets.
Model Outputs
The wind model provides:
- Revenue Projections: Wholesale revenue from exported generation.
- Generation Metrics: Total generation (MWh) and load factor (%).
- Market Value Metrics: Capture price, capture rate, and profile delta, showing how wind generation correlates with market prices.