Several modelling studies of the wind climate in the North Sea have been done in recent years. As part of the WINS50 project (Winds of the North Sea in 2050), by KNMI, TUDelft and Whiffle, Dutch consultancy Pondera collaborated on a case study for wind farm Strekdammen.
Besides contributing to a better understanding of offshore wind and thus reducing uncertainty in future scenarios, it also provides interesting insights on local wind conditions.
As part of the WINS50 project, the HARMONIE model with wind turbines at 2.5 by 2.5 km resolution has been run for the North Sea (including onshore wind). However, the results of this model need to be validated against wind measurements, which are not often (publicly) available. Pondera has therefore made available wind data acquired in this area. The tricky thing about the Eemshaven site is that the wind measurements (for the purpose of new wind turbines to be built) were made on the coastline. Since large changes in wind (speed, increase with height, etc.) take place on the coastline, HARMONIE’s coarse 2.5 km grid doesn’t provide accurate results. In addition, the measurements themselves were partly influenced by already existing wind turbines.
To determine and correct for the wind and wind speed reduction due to the wind turbines, Pondera conducted an additional 100×100 m modelling study with the GRASP model (Figure 1).
Added value of wind turbines in GRASP
Pondera validated the GRASP results with measurements from the 3 LiDAR sites. Figure 2 shows the results for Strekdammen. The decrease in wind speed due to downstream wind turbines can be clearly seen from the difference between the LiDAR wind measurements and GRASP without wind turbines. Without the wind turbines in GRASP, the wind speed from these angles is overestimated by almost 2 m/s on average. The model results of GRASP with wind turbines are close to the measured values. This allows us to form a more realistic picture of the local wind climate.Figure 2. The left figure shows the wind speed differences, averaged over 15 degrees around hub height (100 m), for the measurement period of Strekdammen. The bottom red line shows the disturbed wind directions (with downstream wind turbines). The blue, yellow and red lines are the LiDAR measurements, GRASP without wind turbines and GRASP with wind turbines. The right figure shows the wind profiles with the 0.3-0.7 quantile intervals. The profiles from the measurements and the GRASP model with wind turbines match well.
What can we learn from the GRASP model results
Pondera modelled the wind speeds in GRASP with and without wind turbines in it, which provides an opportunity to analyse the reduction in wind speed. Figure 3 shows this difference for southwest wind directions. By subtracting the two model fields, the average ‘wake’ effect can be seen, which is clearly visible up to kilometres beyond the coastline. As expected, the wake effect is strongest between the wind turbines (here the wind blows on average 2.5 m/s less strong during the modelled days).
Pondera looked at the wake effect in even more detail and built a model to track the wake downstream. Among other things, this allows them to see how often there are wakes within the model domain and how the wind speed recovers behind the turbines. This will allow more accurate calculation of yield losses from new turbine locations in the future.
Plans for follow-up studies
GRASP allows to provide customised advice. GRASP is already currently being used for many different types of projects to better understand these kinds of more complex wind sites. Every site is unique and will be given a customisable model domain so that, for example, surrounding wind farms are included. There are plenty of new challenges for future projects that Pondera can start working on together with Whiffle. Read the study here. Source: Pondera Image: Strekdammen Wind Turbine (BGF)