On AEP prediction and wake modelling at Anholt

Research output: Contribution to conferenceConference abstract for conferenceResearchpeer-review

89 Downloads (Pure)

Abstract

The Anholt wind farm is not only one of the largest parks of the world but also has one of the highest capacity factors (CFs); in 2014 it was 45.85%1. This is mainly due to the low wake effects within the wind farm. Using hub-height hourly simulated winds from the WRF model for the year 2014 at a position in the middle of the wind farm, without accounting for wake effects and assuming flow homogeneity within the wind farm, the CF is 45.07%. The difference between the model-estimated and the reported CFs are partly due to errors in the WRF model but it is also due to the gradients of wind speed and direction. We show that the WRF model is able to reproduce such gradients relatively well by comparison to the wind farm’s SCADA. About 1.5 yr of such SCADA, further quality controlled and filtered, reveals an average wake loss of 3.87% only, whereas results from three wake models, Park, Larsen and FUGA, show average wake losses of 3.46%, 3.69%, and 3.38%, respectively. We employ a bootstrap method to estimate the uncertainty of the wake models. As this is performed with reference to the SCADA, the results provide an idea of the uncertainty of the AEP prediction2. We find all wake models to underpredict the wake loss. The simpler models are as uncertain as the more sophisticated ones.
Original languageEnglish
Publication date2017
Publication statusPublished - 2017
EventWind Energy Science Conference 2017 - Lyngby, Denmark
Duration: 26 Jun 201729 Jun 2017
http://www.wesc2017.org/
http://www.wesc2017.org/

Conference

ConferenceWind Energy Science Conference 2017
CountryDenmark
CityLyngby
Period26/06/201729/06/2017
Internet address

Cite this

Pena Diaz, A., Hansen, K. S., Volker, P., Ott, S., & Hasager, C. B. (2017). On AEP prediction and wake modelling at Anholt. Abstract from Wind Energy Science Conference 2017, Lyngby, Denmark.
Pena Diaz, Alfredo ; Hansen, Kurt Schaldemose ; Volker, Patrick ; Ott, Søren ; Hasager, Charlotte Bay. / On AEP prediction and wake modelling at Anholt. Abstract from Wind Energy Science Conference 2017, Lyngby, Denmark.
@conference{c9bb15da2a09470fb89652596cdcbd23,
title = "On AEP prediction and wake modelling at Anholt",
abstract = "The Anholt wind farm is not only one of the largest parks of the world but also has one of the highest capacity factors (CFs); in 2014 it was 45.85{\%}1. This is mainly due to the low wake effects within the wind farm. Using hub-height hourly simulated winds from the WRF model for the year 2014 at a position in the middle of the wind farm, without accounting for wake effects and assuming flow homogeneity within the wind farm, the CF is 45.07{\%}. The difference between the model-estimated and the reported CFs are partly due to errors in the WRF model but it is also due to the gradients of wind speed and direction. We show that the WRF model is able to reproduce such gradients relatively well by comparison to the wind farm’s SCADA. About 1.5 yr of such SCADA, further quality controlled and filtered, reveals an average wake loss of 3.87{\%} only, whereas results from three wake models, Park, Larsen and FUGA, show average wake losses of 3.46{\%}, 3.69{\%}, and 3.38{\%}, respectively. We employ a bootstrap method to estimate the uncertainty of the wake models. As this is performed with reference to the SCADA, the results provide an idea of the uncertainty of the AEP prediction2. We find all wake models to underpredict the wake loss. The simpler models are as uncertain as the more sophisticated ones.",
author = "{Pena Diaz}, Alfredo and Hansen, {Kurt Schaldemose} and Patrick Volker and S{\o}ren Ott and Hasager, {Charlotte Bay}",
year = "2017",
language = "English",
note = "Wind Energy Science Conference 2017, WESC-2017 ; Conference date: 26-06-2017 Through 29-06-2017",
url = "http://www.wesc2017.org/, http://www.wesc2017.org/",

}

Pena Diaz, A, Hansen, KS, Volker, P, Ott, S & Hasager, CB 2017, 'On AEP prediction and wake modelling at Anholt', Wind Energy Science Conference 2017, Lyngby, Denmark, 26/06/2017 - 29/06/2017.

On AEP prediction and wake modelling at Anholt. / Pena Diaz, Alfredo; Hansen, Kurt Schaldemose; Volker, Patrick; Ott, Søren; Hasager, Charlotte Bay.

2017. Abstract from Wind Energy Science Conference 2017, Lyngby, Denmark.

Research output: Contribution to conferenceConference abstract for conferenceResearchpeer-review

TY - ABST

T1 - On AEP prediction and wake modelling at Anholt

AU - Pena Diaz, Alfredo

AU - Hansen, Kurt Schaldemose

AU - Volker, Patrick

AU - Ott, Søren

AU - Hasager, Charlotte Bay

PY - 2017

Y1 - 2017

N2 - The Anholt wind farm is not only one of the largest parks of the world but also has one of the highest capacity factors (CFs); in 2014 it was 45.85%1. This is mainly due to the low wake effects within the wind farm. Using hub-height hourly simulated winds from the WRF model for the year 2014 at a position in the middle of the wind farm, without accounting for wake effects and assuming flow homogeneity within the wind farm, the CF is 45.07%. The difference between the model-estimated and the reported CFs are partly due to errors in the WRF model but it is also due to the gradients of wind speed and direction. We show that the WRF model is able to reproduce such gradients relatively well by comparison to the wind farm’s SCADA. About 1.5 yr of such SCADA, further quality controlled and filtered, reveals an average wake loss of 3.87% only, whereas results from three wake models, Park, Larsen and FUGA, show average wake losses of 3.46%, 3.69%, and 3.38%, respectively. We employ a bootstrap method to estimate the uncertainty of the wake models. As this is performed with reference to the SCADA, the results provide an idea of the uncertainty of the AEP prediction2. We find all wake models to underpredict the wake loss. The simpler models are as uncertain as the more sophisticated ones.

AB - The Anholt wind farm is not only one of the largest parks of the world but also has one of the highest capacity factors (CFs); in 2014 it was 45.85%1. This is mainly due to the low wake effects within the wind farm. Using hub-height hourly simulated winds from the WRF model for the year 2014 at a position in the middle of the wind farm, without accounting for wake effects and assuming flow homogeneity within the wind farm, the CF is 45.07%. The difference between the model-estimated and the reported CFs are partly due to errors in the WRF model but it is also due to the gradients of wind speed and direction. We show that the WRF model is able to reproduce such gradients relatively well by comparison to the wind farm’s SCADA. About 1.5 yr of such SCADA, further quality controlled and filtered, reveals an average wake loss of 3.87% only, whereas results from three wake models, Park, Larsen and FUGA, show average wake losses of 3.46%, 3.69%, and 3.38%, respectively. We employ a bootstrap method to estimate the uncertainty of the wake models. As this is performed with reference to the SCADA, the results provide an idea of the uncertainty of the AEP prediction2. We find all wake models to underpredict the wake loss. The simpler models are as uncertain as the more sophisticated ones.

M3 - Conference abstract for conference

ER -

Pena Diaz A, Hansen KS, Volker P, Ott S, Hasager CB. On AEP prediction and wake modelling at Anholt. 2017. Abstract from Wind Energy Science Conference 2017, Lyngby, Denmark.