Different AEP output with diffrent wake models
Hi Mads @mmpe and DTU team,
I was running some simulations with PyWake, trying different wake models, and to my surprise, some of the models are producing very different results, almost a 25% difference in AEP. WF contains around 200 turbines and is set on an offshore site.
I am not sure if I am making some mistakes or if it is an issue.
The models used are:
model_li = [
#Jensen
PropagateDownwind(site, windTurbines,
wake_deficitModel=NOJDeficit(k=0.04),
superpositionModel=SquaredSum(),
turbulenceModel=CrespoHernandez(c=[0.66, 0.83, 0.03, 0.32])),
#TurboJensen
PropagateDownwind(site, windTurbines, rotorAvgModel=None,
wake_deficitModel=TurboNOJDeficit( A=.6, cTI=[1.5, 0.8],
use_effective_ws=True,
use_effective_ti=False),
superpositionModel=LinearSum(),
turbulenceModel=CrespoHernandez(c=[0.66, 0.83, 0.03, 0.32])),
#TurboGaussian
PropagateDownwind(site, windTurbines,
wake_deficitModel=TurboGaussianDeficit(A=.04, cTI=[1.5, 0.8],
ceps=0.25, use_effective_ws=True,
use_effective_ti=False,
groundModel=Mirror(),
rotorAvgModel=GaussianOverlapAvgModel()),
superpositionModel=SquaredSum(),
turbulenceModel=CrespoHernandez(c=[0.66, 0.83, 0.03, 0.32])),
#Niayifar Gaussian
PropagateDownwind(site, windTurbines,
wake_deficitModel=NiayifarGaussianDeficit(a=[0.38, 4e-3], ceps=.2, use_effective_ws=True,
use_effective_ti=True),
superpositionModel=LinearSum(),
turbulenceModel=CrespoHernandez(c=[0.66, 0.83, 0.03, 0.32])),
# Zong
PropagateDownwind(site, windTurbines, wake_deficitModel=ZongGaussianDeficit(use_effective_ws=True,
use_effective_ti=True),
superpositionModel=WeightedSum(),
turbulenceModel=CrespoHernandez(c=[0.66, 0.83, 0.03, 0.32])),
# Blondel20
PropagateDownwind(site, windTurbines, wake_deficitModel=BlondelSuperGaussianDeficit2020(use_effective_ws=True,
use_effective_ti=True),
superpositionModel=LinearSum(),
turbulenceModel=CrespoHernandez(c=[0.66, 0.83, 0.03, 0.32])),
# Blondel23
PropagateDownwind(site, windTurbines, wake_deficitModel=BlondelSuperGaussianDeficit2023(use_effective_ws=True,
use_effective_ti=True),
superpositionModel=LinearSum(),
turbulenceModel=CrespoHernandez(c=[0.66, 0.83, 0.03, 0.32])),
#Blondel23+blockage
All2AllIterative(site, windTurbines, wake_deficitModel=BlondelSuperGaussianDeficit2023(use_effective_ws=True,
use_effective_ti=True),
superpositionModel=LinearSum(),
blockage_deficitModel=SelfSimilarityDeficit(groundModel=Mirror(), superpositionModel=LinearSum()),
turbulenceModel=CrespoHernandez(c=[0.66, 0.83, 0.03, 0.32])),
]
AEP outcome: GWh
Jensen: 198.43 TurboJensen: 210.02 TurbOPark: 188.62 Niayifar Gaussian: 245.38 Zong: 245.38 Blondel2020: 178.36 Blondel2023: 227.04 Blondel2023+blockage(self-similarity): 226.71