What is it?
Wind power forecasting is estimating the expected production of one or several wind turbines in the near future. Grid operator, energy trader, or plant owner and operators need to assess risks and impacts of wind generation on the electricity grid.
In the electricity grid, balance must be maintained between electricity consumption and generation – to avoid disturbances in power quality or supply. At any time, electricity production has to match consumption. Predicting a renewable energy project’s output days, hours or minutes in advance provides accuracy in your business strategies, cost savings for utility operations, facilitate plant dispatch scheduling, and optimize plant maintenance.
Meteodyn software solutions: meso-to-micro CFD downscaling forecast
Prediction of meteorological variables is done by using numerical weather prediction (NWP) models.
Meteorological forecasts are then extrapolated at the desired location and at turbine hub height taking into account terrain’s roughness, orography and obstacles, and by modeling the local wind profile. The solution consists in using a CFD (Computational Fluid Dynamics) code which ensures more accuracy.
Meteodyn Forecast software combines physical modelling and statistical learning to predict the wind power production. Thanks to the downscaling of Numerical Weather Predictions at the turbine level, get the forecasted production for both the whole park and each turbine. Power production is subject to seasonal variations and daily cycles too. Of course, these fluctuations request very short term forecast to be computed.
- Consideration of air density, atmospheric stability, wakes effects
- Computation of production forecast uncertainties (unsteadiness and turbulence conditions)
- Automatic Neural Network correction based on advanced learning algorithms
- Forecasts for any location in the world
- Display of all the information the user needs in a clear and user-friendly interface
Meteodyn can also provide reports with your forecast performance.
Meteodyn consulting services
- Short (0-5 days) and very short (0-4h) term wind power forecasting
Forecasts can be provided from several minutes to several days to help you trading in real time or day-ahead.This combined use of statistical and physical modelling methods integrates on-site real-time measurements as well as mesoscale forecasts to optimize very short-term forecasting.
- Automatic Neural Network by using SCADA data
- Operations and maintenance (on-site real wind turbine power curves)
Meteodyn R&D Department is continuously conducting forecasting research programs. Currently, they are focusing on the Automatic Neural Network and Scada data. The aim is to improve forecast data accuracy and reduce forecast error.
On the same basis, meteodyn provides solar power forecasting.