Experience of Modelling Forested Complex Terrain Peter Stuart, Ian Hunter & Nicola Atkinson 30 th October 2009
Jan 29, 2016
Experience of Modelling Forested Complex TerrainPeter Stuart, Ian Hunter & Nicola Atkinson
30th October 2009
Overview
• The Challenges of Forested Complex Terrain. • Predicting the Breakdown of Linear Flow Models in Forested Complex Terrain.
• Tuning Canopy Model Parameters
• Energy Yield Prediction Verification in Forested Complex Terrain
• Future challenges: Modelling Non-Neutral Canopy Flow
The Challenges of Forested The Challenges of Forested Complex TerrainComplex Terrain
The Challenges of Forested Complex
The Challenges of Forested Complex Terrain – Some Observations
Masts A & B are in simple non
forested terrain
Masts C & D are in complex
forested terrain
However masts A,B,C & D are all on the same wind
farm site!
8%
Predicting the Breakdown of Predicting the Breakdown of Linear Flow Models in Linear Flow Models in Forested Complex TerrainForested Complex Terrain
Predicting the Breakdown of Linear Models in Forested Terrain: Method & Models
• Calculate flow over idealised hills using both CFD and linear models for incrementally increasing slopes and tree heights.
MS3DJH / RES Roughness
• Establish guidelines for where linear models fail by comparing to CFD.
Linear CFD
• Use simple geometrical considerations to assess likely impact on real sites.
• Confirm predicted effects using CFD.
8
Variation of Critical Terrain Slope with Tree Height (2D Symmetric Hill)
Critical angle for recirculation reduced by ~ ¼° per metre of tree height.
c.f. Kaimal and Finnigan (1994): 2D Critical slope ~10° for a very rough hill.
Variation of Critical Terrain Slope with Tree Height (2D Symmetric Hill)
Critical angle for linear model break down reduced by ~½° per metre of tree height.
Predicting the Breakdown of Linear Models in Forested Terrain: Example Site
Predicting the breakdown of linear flow models in forested terrain: Example Site
11
• Establish critical angle considering tree height (20m).
114
16= TR
H 62
16= TN
HPredominant wind direction
Tuning Canopy Tuning Canopy ParametersParameters
The drag due to the canopy is taken into account via an additional term entering the momentum equation :
iDi UUCF 2
1 α (in m2m-3) is the leaf foliage area per unit of volume
CD is the canopy drag coefficient.
The effects of the canopy on turbulence are accounted for by additional source terms Sk and Sε in the transport equations of k and ε
Lopes da Costa, J. C. P., “Atmospheric Flow Over Forested and Non-Forested Complex Terrain”,PhD Thesis University of Porto, July 2007.
kUUCS dpDk
3
UCU
kCCS dpD 5
3
4
Ventos Canopy Model
• European site with complex orography and extensive forest cover (H ~ 15m).
• 6 meteorological masts used for validation.
Tuning Canopy Parameters: Example site
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
M273 M272 M223 M1 M187 M186
Sh
ear
Exp
on
ent
Predicted and measured shear exponents for 330° direction.
Measured ShearCFD Predicted Shear H = 15m, CD = 0.25 and α = 0.2
Tuning Canopy Parameters: Example site
Reducing the canopy density improves agreement, but even with α = 0.05 the predicted shear exponents are still too high.
2nd Iteration: α → 0.13
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
M273 M272 M223 M1 M187 M186
Sh
ear
Exp
on
ent
Measured ShearCFD Predicted Shear
3rd Iteration: α → 0.05
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
M273 M272 M223 M1 M187 M186
Sh
ear
Exp
on
ent
Measured ShearCFD Predicted Shear
Tuning Canopy Parameters: Example site
Further improvement gained by using an effective tree height of ¾ the actual height.
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
M273 M272 M223 M1 M187 M186
Sh
ear
Exp
on
ent
Measured ShearCFD Predicted Shear
Predicted and measured shear exponents for 330° direction.
Final parameters: H = 11.25m, CD = 0.25, α = 0.05
Tuning Canopy Parameters: Example site
Optimized parameters derived from 330° direction applied to 300° direction.
Sh
ear
Exp
on
ent
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
M273 M272 M223 M1 M187 M186
Measured Shear
CFD Predicted Shear
Predicted and measured shear exponents for 300° direction.
Tuning Canopy Parameters: Example site
Energy Yield Prediction Energy Yield Prediction Verification in Forested Verification in Forested Complex TerrainComplex Terrain
20
Prediction Verification in Forested Complex Terrain - Site Overview
• Moderately complex terrain. 11 multi-megawatt class turbines.
• Inhomogeneous forest cover 5-20m in height.
• Two 40m Masts (turbine hub height is 65m, rotor diameter is 82m)
Mast A Mast B
Power Performance Mast T3
21
How is the Wind Farm Performing?
Power performance indicates that turbine T3 is operating as expected.
However, the majority of turbines are found to be under producing.
22
Prediction Verification in Forested Complex Terrain – Terrain Effect
• Plot error in yield vs. predicted terrain effect
• Strong correlation between predicted terrain effect and error in energy yield.
• Using WAsP provides very similar results.
23
Prediction Verification in Forested Complex Terrain – Increased Roughness
• Increasing roughness in orography model improves agreement.
• Roughness increased from default 0.04m to 2m, potentially more representative of the site (tree height up to 20m).
• Site also modelled using CFD, including a canopy model.• CFD and increased roughness orography model in very close
agreement.
24
Prediction Verification in Forested Complex Terrain – Shear Predictions
• Still an error in the predictions; other causes – high shear?
• Strong correlation between high shear and error in energy yield predictions.
Power Performance Turbine
Increasing Shear
Future challenges: Modelling Future challenges: Modelling Non-Neutral Canopy FlowNon-Neutral Canopy Flow
Future challenges: Modelling Non-Neutral Canopy Flow
• How does non-neutral atmospheric stability change canopy flows e.g. Internal Boundary Layer:
Bz
h
x
h
1ln
02
Conclusions
• Trees can cause very different site conditions to occur within the same wind farm e.g. 8% change in turbulence intensity.
• Tree height has a strong impact on the critical slope at which linear models breakdown. • Tuning of canopy parameters help improve agreement between observations and CFD model predictions.
• Operational data suggests that tuning the roughness length in the linear orography model helps improve agreement with operational data and CFD predictions.
• Operational data suggests a link between turbine under performance and wind shear.