Modelling the impact of wakes on power output at Nysted and Horns Rev

1
Modelling the impact of wakes on power output at
Nysted and Horns Rev

• R.J. Barthelmie, Indiana University USA/Risoe DTU
  DK
• K. Hansen, DTU Denmark
• S.T. Frandsen, O. Rathmann, RISOE DTU Denmark
• G. Schepers ECN, Netherlands
• K. Rados, NTUA, Greece
• W. Schlez, A. Neubert, GH, Germany
• L.E. Jensen, DONG Energy, Denmark
• S. Neckelmann, Vattenfall, Denmark

Funding NSF CBET-0828655, EU UPWIND SES6
019945 EU POWWOW SES6 019898 Data DONG Energy
A/S Vattenfall AB (Horns Rev), Vattenfall and
E. On Sweden (Nysted)
2
Modelling wakes in the UpWind project

• Problems
• Preliminary analysis of wake power losses in
  large offshore wind farms larger than expected.
  Amended but high uncertainty
• First v. large wind farms on land
• Resources and wakes difficult to model in complex
  terrain
• Multiple wind farms developed in same area

• Solutions
• New parameters and/or next generation of wake
  models able to account for deep array effect
• (Assess the magnitude of the issue onshore)
• Develop, apply and evaluate CFD
• Assess, develop and evaluate models for whole
  wind farm modelling

3
Data

• In agreement with data owners some wind farm data
  have been made available
• Access is open and free (registration necessary)
• Offshore wake data from
• Vindeby
• Middelgrunden
• Horns Rev
• Nysted (in proc)
• Data processed into case studies for Horns Rev,
  Nysted (performance remains confidential)

Access /registration details http//mypage.iu.edu
/rbarthel/wakeslab.html
4
Wake models used in this project
5
Offshore wind farms
http//earthobservatory.nasa.gov/IOTD/view.php?id
3389
6
Horns Rev and Nysted layouts
Nysted 10.5D x 5.8D
Horns Rev 7D x 7D
7
Data comparison

• Data from 2004-2006 Nysted, 2005 Horns Rev
• Selection on
• wind speed 0.5 ms-1
• direction 2.5º
• all turbines working in row
• all turbines working in neighbouring rows
• two subsequent observations for stationarity
• Gives relatively few observations in each
  category
• Data differences mainly due to spacing?

8
Model comparison at Horns Rev and Nysted
9
Model comparison at Horns Rev and Nysted

• Exact row, narrow directions
• Seems to be a special case
• Agreement on wake behaviour at Horns Rev and
  Nysted
• Model agreement within 10
• Cross row angles
• Asymmetry in obs. and models
• Larger uncertainty

ER
ER10º
10
Model comparison at Horns Rev and Nysted

• Consistency improved in model results
• High degree of uncertainty
• Differences between models
• Data issues
• Lower wind speeds
• Ongoing issues
• Asymmetry around central row
• Developing quantitative methods of evaluation
  e.g. efficiency
• Stability

11
Stability at Nysted
Results from Barthelmie et al. European Offshore
Wind 2007
12
Summary and future work

• Objective
• Reducing uncertainty in predicting power losses
  from wakes
• UpWind project
• Provides platform for undertaking model
  evaluation data sharing
• Progress made
• Data analysis and modelling
• Wakes can be modelled with appropriate parameters
• Future
• Physical understanding of wake processes within
  and downwind wind farms

13
Other UpWind wake presentations and posters