PowerPoint-Präsentation

Report
Wind Farms in Complex Terrain: Numerical Simulation
of Wind and Wakes for Optimized Micrositing
S. Jafari, N. Chokani, R.S. Abhari
ETH Zürich, Switzerland
EWEA February 4-7, 2013
Vienna, Austria
Overview

Motivation

Objectives

Modeling Approach: Immersed Wind Turbine Model

Validation and Results
 Single Wake
 Multiple Wake
 Wind farm in complex terrain

Summary and Conclusion
February 6, 2013
Samira Jafari
2
Motivation

Wind turbines operating in wakes of upwind wind turbines may have
 30-40 % power losses compared to upstream turbine
 up to 80% larger fatigue loads than upstream turbine
 Most models fail in predictions when more complex inflow or ground features
are present
 Interaction of topography with wake and wind flow not addressed satisfactorily
to date
 Microscale wind and wake flow must be simultaneously simulated to be able to:
 Account for change in inflow wind due to terrain effects
 Assess effect of elevated turbulence on wake’s evolution
 Investigate interaction of wake with adverse/favorable pressure gradients caused by
topography
February 6, 2013
Samira Jafari
3
Objectives

Develop computationally efficient wake model to be embedded in RANS solver
used for microscale wind simulations with comparable grid requirements

Perform simultaneous simulations of microscale wind and wind turbine wake

Validate and evaluate predictions of flow field and power performance in wind
farms
February 6, 2013
Samira Jafari
4
Numerical Approach

Turbines modeled using immersed wind turbine model (IWTM), embedded in
LEC’s RANS solver, MULTI3

Turbine represented as streamtube defined based on turbine operating point

Near wake modeled, velocity and turbulent field mapped at the end of inviscid
expansion of wake, but far wake resolved on computational grid

Boundary conditions imposed on Cartesian grid using immersed boundary
method
February 6, 2013
Samira Jafari
5
Validation: Microscale Wind

Prediction of wind speed compared with field measurements over Askervein
(moderate terrain, Jafari et al., 2011) and Bolund Hill (complex terrain, Jafari et al., 2012)

Good agreement observed for both cases, up- and downstream of hill
Bolund Hill, 270o
Askervein Hill
February 6, 2013
Samira Jafari
6
Mean Flow of Single Wake

Predicted evolution of wake in good agreement with wind tunnel experiments,
(Hassan,1992)

Maximum 12% difference between predicted and measured wind speed
February 6, 2013
Samira Jafari
7
Turbulence Intensity in Single Wake

At x=2.5D, two peaks observed in turbulence intensity profile as expected

Evolution of turbulence intensity captured well both qualitatively and qualitatively
February 6, 2013
Samira Jafari
8
Single Wake: Full-scale Measurement

Predictions compared
with full-scale
measurements at
Sexbierum wind farm

5.4 MW farm
consisting of 18
turbines, D=30 m

Maximum deficit
underestimated by
20%

Wake width predicted
well
February 6, 2013
Samira Jafari
9
Validation: Multiple Wakes

Interactions of multiple wakes examined for offshore wind farms
 Horns Rev (offshore, Denmark), 80 Vestas V2.8-80
 Lillgrund (offshore, Sweden), 48 Siemens SWT-2.3-93
30

28
27
Power loss in array and sensitivity to wind direction captured for all cases
February 6, 2013
Samira Jafari
10
Wind Farm in Complex Terrain

23.7 MW Mont Crosin wind farm located in Jura region, Switzerland (complex
terrain) consisting of 16 turbines with hub heights of 45 and 95 m

SCADA data collected and analyzed over one and a half years period
270o
February 6, 2013
Samira Jafari
11
Simulation Set-Up

Upstream conditions of wind speed and turbulence and specified based on:

Long-term mesoscale [Weather Research Forecast Model (WRF)] simulations
performed over Switzerland, Jafari et al., 2012

Measurements using LEC’s nacelle mounted probe, Mansour et al., 2013
Computational grid for wind direction 170o
February 6, 2013
Dominant wind direction from south-west quadrant
Samira Jafari
12
Mont Crosin Wind Farm: Results

Impact of terrain on local wind evident

Performance of turbine 14 decreases
260o
relative to turbine 13 up to 65%
90 m AGL
260o
230o
45 m AGL
February 6, 2013
Samira Jafari
13
Summary & Future Work

Simultaneous simulation of microscale wind and wakes accomplished with
computationally efficient wind turbine model

Model evaluated with broad range of test cases including wind tunnel/field
experiments, onshore/offshore, and flat/complex terrain

IWTM brings grid requirements for wake simulations closer to microscale wind
and facilitates use of Computational Fluid Dynamics for micrositing in complex
terrain
February 6, 2013
Samira Jafari
14
Thank you.
February 6, 2013
Samira Jafari
15

similar documents