Title: RECOMMENDED PRACTICES WHEN ANALYZING WIND FLOW NEAR A FOREST EDGE WITH WAsP Benoit Dalp and Christia
1RECOMMENDED PRACTICES WHEN ANALYZING WIND FLOW
NEAR A FOREST EDGE WITH WAsPBenoit Dalpé and
Christian Masson
Canada Research Chair on Nordic Environment
Aerodynamics of Wind Turbines. École de
Technologie Supérieure (ÉTS), Montréal, Canada
2Presentation overview
- Objectives
- Experimental measurements
- Wind flow entering the forest
- Wind flow leaving the forest.
- CFD simulations
- Mathematical model
- Numerical method.
- WAsP
- Results
- Conclusion and recommendations.
31. Objectives
- Evaluate the influence of the meteorological
station (met. station) position on the predicted
wind flow obtained with WAsP - Compare WAsP to experimental measurements at low
altitude and CFD simulations at high altitude for
two wind flow directions.
42. Experimental measurements i) Wind flow
entering the forest
- Measurements obtained by Irvine et al., 1990
- Forest
- Uniform Sitka Spruce plantation
- Average height (h) 7.5m
- LAI2.15
- Assumed a distribution
52. Experimental measurements i) Wind flow
entering the forest
- Wind velocity measurements obtained at three
heights (z/h 0.5, 1, 2) on four masts
62. Experimental measurements ii) Wind flow
leaving the forest
- Measurements obtained by Raynor, 1971
- Forest
- Pine forest
- Average height (h) 10.5m
- Assumed a distribution
72. Experimental measurements ii) Wind flow
leaving the forest
- Wind velocity measurements obtained at four
heights (z/h 0.17, 0.33, 0.67, 1.33) on five
masts
83. CFD simulations i) Mathematical model
- Two dimensionnal (x-z)
- Incompressible flow
- Steady state
- Neutral stratification
- Negligible Coriolis force
- Horizontally homogeneous forest.
93. CFD simulations i) Mathematical model
- Momentum source term
- Cd forest drag coefficient
- a leaf area density.
103. CFD simulations i) Mathematical model
- k-e turbulence model
- Original constants Jones and Launder,
1972 - Modified constants Katul et al., 2004
113. CFD simulations i) Mathematical model
- Source terms in k and e equationsKatul et al.,
2004
123. CFD simulations ii) Numerical method
- Numerical solution with FLUENT 6.2
- Mesh with Gambit 2.2 for two directions
- Wind flow enteringthe forest
- Wind flow leavingthe forest
133. CFD simulations ii) Numerical method (wind
flow entering the forest)
- Boundary conditions
- Inlet and top of the domain
143. CFD simulations ii) Numerical method (wind
flow entering the forest)
- Outlet boundary Patankar, 1980 outflow
condition - On the ground outside the forest shear
boundary condition of Richards and Hoxey,
1993 - On the ground inside the forest transition
from a shear boundary condition to a full slip
wall.
153. CFD simulations ii) Numerical method (wind
flow leaving the forest)
- Boundary conditions
- Same as for wind flow entering the forest except
for the inlet and top boundary - At the inlet and top boundary a fully
developed solution was used.
163. CFD simulations ii) Numerical method (wind
flow leaving the forest)
- Fully developed solution with FLUENT 6.2
- Boundary conditions
- Ground full slip wall
- Top constant friction velocity.
174. WAsP
- Forest representationDellwick et al.,
2004d displacement height 0.65
hzo roughness length 0.1 hWAsP users
guide
184. WAsP
- Zero heat flux was imposed to simulate a neutral
atmosphere - For each wind flow direction (entering and
leaving the forest), nine met. station positions
were considered.
195. Resultsi) Wind flow entering the forest
- Meteorological station positions
205. Results i) Wind flow entering the forest
215. Resultsii) Wind flow leaving the forest
- Meteorological station positions
225. Results ii) Wind flow leaving the forest
236. Conclusion and recommendations
- For each wind flow direction, nine different met.
station positions in WAsP were considered - Predictions obtained from WAsP were compared to
experimental measurements at low altitude and to
CFD simulations at high altitude
246. Conclusion and recommendations
- When in the neighbouring of a forest edge, WAsP
is very sensitive to the met. station position - Compared to CFD simulations, WAsP had differences
up to 20 at typical hub height - To obtain acceptable results, the met. station
should be located outside the forest or above the
forest at z gt 5h.
25Acknowledgments
- NSERC for the funding of this research
- CORUS Center in Murdochville, Canada, for the
WAsP license.