Satellite SAR measurements for offshore wind farm development

Tobias Ahsbahs, Merete Badger, Charlotte B. Hasager,

Kurt S. Hansen, and Patrick Volker

DTU Wind Energy, Technical University of Denmark

Add Presentation Title in Footer via ”Insert”; ”Header & Footer”

Overview

• Data and Location

• Coastal wind speed gradients

• Wind farm wakes

– Long fetch

– Limited fetch

• Conclusions

2 24 October, 2017

Number of turbines: 111

Wind turbine capacity: 3.6 MW

Rotor diameter: 120 meters

Courtesy: DONG Energy (soon Ørsted)

Construction period: 2012-2013

SCADA data for analysis is from January 1st, 2013 to June 30, 2015 (2.5 years)

DTU Wind Energy, Technical University of Denmark

Add Presentation Title in Footer via ”Insert”; ”Header & Footer”

Data

3

Envisat

2002-2012

(without wind farm)

Sentinel-1a/b

2014/2016-present

(with wind farm)

SCADA data from 1 January 2013 to 30 June 2015 (2.5 yr)

Courtesy: DONG Energy (soon Ørsted)

SAR archives are more open and easier available than ever!

Full archive available at: https://satwinds.windenergy.dtu.dk/

DTU Wind Energy, Technical University of Denmark

Add Presentation Title in Footer via ”Insert”; ”Header & Footer”

SAR wind retrieval

4

Radar backscatter High resolution ocean wind

speed at10 m

DTU Wind Energy, Technical University of Denmark

Add Presentation Title in Footer via ”Insert”; ”Header & Footer”

5 24 October, 2017

From radar backscatter to wind

Empirical geophysical model functions (GMF):

2cos),(cos),(1)()( UCUBAUNRCS

NRCS = radar backscatter [dB] θ = incidence angle [degrees]

U = wind speed at 10 m [m/s] Φ = relative wind direction [degrees]

DTU Wind Energy, Technical University of Denmark

Add Presentation Title in Footer via ”Insert”; ”Header & Footer”

6

Wind speed map from SAR

DTU Wind Energy, Technical University of Denmark

Add Presentation Title in Footer via ”Insert”; ”Header & Footer”

Horizontal wind speed gradients in row A for wind directions of 245°to 275°

• Wind speed non-dimensionalized with turbine postion A15

– SCADA: up to 8% difference

– Strong gradient

• SAR retrievals from 72 Envisat images (2002-2012) and WRF (2002-2012) similar

– SAR predicts the wind speed gradient

7

DTU Wind Energy, Technical University of Denmark

Add Presentation Title in Footer via ”Insert”; ”Header & Footer”

Wind farm wakes – long fetch

• a) Envisat before wind farm

• b) Sentinel-1 after wind farm

• Wind direction 75° to 105° from WRF model

8

DTU Wind Energy, Technical University of Denmark

Add Presentation Title in Footer via ”Insert”; ”Header & Footer”

Wind farm wakes – long fetch difference between transects

• Wind farm wake visible as a difference

• Maximum wake deficit in the Southern part with highest wind turbine density

• Few images, but indication for wake is clear

• This analysis is possible for (almost) any wind farm

9

DTU Wind Energy, Technical University of Denmark

Add Presentation Title in Footer via ”Insert”; ”Header & Footer”

Wind farm wakes and coastal gradients

• Wind directions between 214.5° 244.5°

• Three transects (b,c,d) through the wind farm

• Comparison between without WF (ENVISAT) and with (Sentinel-1)

• SCADA data from turbines within the transects

• One transect for reference (1)

10 24 October, 2017

DTU Wind Energy, Technical University of Denmark

Add Presentation Title in Footer via ”Insert”; ”Header & Footer”

Wind farm wakes and coastal gradients

• Transects nondimensionalized with the inflow wind speed

a)

• Without wind farm similar wind speeds before and after WF construction

b)

• No reduction wake visible with wind farm

• SCADA data shows wake

11 24 October, 2017

DTU Wind Energy, Technical University of Denmark

Add Presentation Title in Footer via ”Insert”; ”Header & Footer”

Wind farm wakes and coastal gradients

• Wind speed gradient upstream in good agreement

c)

• Slight reduction after WF

d)

• SAR wind speed consistently reduced

• No reduction close to what SCADA data suggests

12 24 October, 2017

DTU Wind Energy, Technical University of Denmark

Add Presentation Title in Footer via ”Insert”; ”Header & Footer”

Conclusions

SAR archives are more open and easier available than ever!

• Full archive available at: https://satwinds.windenergy.dtu.dk/

Coastal gradient from SAR

• SAR predicts coastal gradients similar to SCADA and WRF.

– Benefit for planning wind farms close to a complex shore line

Influence of the wind farm on the flow:

• Long archive – cases with and without wind farm

– More data is coming from Sentinel-1A and B.

• For long fetch

– Wake can be visible by comparison images before and after WF

• For fetch limited cases:

– No reduction of wind speed is found

– Wake can be visible by comparison with pre WF results

13

DTU Wind Energy, Technical University of Denmark

Add Presentation Title in Footer via ”Insert”; ”Header & Footer”

Acknowledgements

14

We thank DONG Energy (soon Ørsted) and partners for the SCADA data.

Satellite SAR data are from ESA and Copernicus.