Elliot Simon

Research from DTU predicts the wind within 0-15 minutes

Tuesday 06 Mar 18


Elliot Simon
PhD student
DTU Wind Energy
An improved prediction of the wind means a more efficient utilization of it for many players in wind energy.
If a skilled prediction of the wind is possible, costly system imbalances can be significantly reduced as well as the need for spinning reserves, i.e. power plants working as back up if the wind suddenly drops.

Elliot Simon is a PhD student at DTU Wind Energy. In his PhD project he works on improving wind speed and power forecasts within very short time scales - that is 15 minutes or less. He does so by making measurements of the wind with so-called scanning Doppler LiDARs.
LiDARs are remote sensing instruments which can be used for measuring the wind speed and direction across long distances (up to 10km). In this case, pointed far upstream of a wind farm, measured from a LiDAR placed in the wind farm.

DTU has a fleet of 7 long-range scanning lidars which are used in field experiments around the world. The scanners had just come home to Risø after the Perdigão measurement campaign in Portugal and had 6 months until being shipped out to Spain as part of the New European Wind Atlas project.
“We had 3 million Euros worth of scanners sitting around. So let’s use them”, was Elliot’s reaction.

Elliot and his colleague Guillaume Lea received support from Elliot’s Ph.D. supervisor Michael Courtney for deploying the LiDARs at Risø in order to demonstrate and refine the concept.
At the peak of the project they had six lidars – at the moment only two are used. The lidars are measuring both the vertical and horizontal structure of the wind, which are joined with mast and turbine data also at Risø, ending in a large number of independent observations.
Elliot processes the LiDAR data with time series analysis and machine learning techniques in order to build a prediction engine. The predictions are then compared against actual turbine and mast measurements in order to evaluate the performance of the model.
“The greatest benefit from the LiDAR forecast system is that it reduces uncertainties in energy production, supporting the electrical grid and energy markets and opening up new areas in windfarm control”, Elliot concludes.