© 2014, Bert Blocken, Eindhoven University of Technology.

No part of this publication may be reproduced or transmitted in any form or by any

means without the prior written permission of the author.
MODULE QUESTION
Top view of a converging building arrangement. At which position does the highest
amplification of wind speed occur?

A) Position A: upstream of the passage

B) Position B: in the passage
C) Position C: downstream of the passage

wind
A
B
C
LEARNING OUTCOMES FOR MODULE 10

At the end of this module, you will:

• Understand a particular case study of wind energy harvesting in the built environment
• Understand how the aerodynamic design and wind energy output of the Bahrain
World Trade Center could have been improved
Wind speed in converging and diverging building arrangements

Top view. Two buildings. Wind direction from bottom to top.

Which configuration yields the highest wind speed in the passage?

wind wind

B
A

CONVERGING DIVERGING
Wind speed in converging and diverging building arrangements

 Counter-intuitive result: wind-blocking effect: upstream wind-speed slow-down

“Subsonic upstream disturbance in the wind-flow pattern”
 High-speed
High- High- area
speed area speed area

Low-speed area
This work was published, received very positive reviews and quite a good number of
citations.

Blocken B, Stathopoulos T, Carmeliet J. 2008. A numerical study on the existence of the Venturi-effect
in passages between perpendicular buildings. Journal of Engineering Mechanics - ASCE 134(12):
1021-1028.

Blocken B, Stathopoulos T, Carmeliet J. 2008. Wind environmental conditions in passages between

two long narrow perpendicular buildings. Journal of Aerospace Engineering - ASCE 21(4): 280-287.

Blocken B, Carmeliet J, Stathopoulos T. 2007. CFD evaluation of the wind speed conditions in
passages between buildings – effect of wall-function roughness modifications on the atmospheric
boundary layer flow. Journal of Wind Engineering and Industrial Aerodynamics 95(9-11): 941-962.

Who cares?
Wind energy in the built environment
Wind energy in the built environment
Wind energy in the built environment
Wind energy in the built environment
 Bahrain World Trade Center
- Opened in 2008
- USD $150 million
Wind energy in the built environment
- 240 m height
- Three 29 m wind turbines (225 kW)
- Oriented to “capture the prevailing
on-shore Gulf breeze”…
 Bahrain World Trade Center
- Opened in 2008
- USD $150 million
Wind energy in the built environment
- 240 m height
- Three 29 m wind turbines (225 kW)
- Oriented to “capture the prevailing
on-shore Gulf breeze”…
 Prevailing on-shore
Gulf breeze
Wind energy in the built environment

CONVERGING
CONFIGURATION
?
Hypothesis

The Bahrain WTC design would have yielded higher wind energy output if the
buildings were positioned in diverging rather than converging arrangement.

In other words: from wind energy point of view, the towers should have been
turned 180° around.
Investigation

 Detailed study by wind-tunnel testing and Computational Fluid Dynamics

simulations.
Wind-tunnel testing: results

Not velocity magnitude, but x-velocity component, where the x-axis is parallel to the
passage axis and the wind turbine axes. Only the x-velocity component is relevant for
wind energy output and should be used for a fair comparison.

Kx = x-velocity component (U), divided

by wind speed V240 at tower height
(H = 240 m).
= always taken at position of wind
turbine axis (crossing of blades)

ܷ
‫ܭ‬௫ ൌ
x ܸଶସ଴
 Kx

0.0
0.2
0.4
0.6
0.8
1.0
1.2

θ=
converging

1
0°
diverging

converging

2
15°
diverging

Turbine 1
converging

3
30°
diverging
Wind-tunnel testing: results

converging

40°
diverging
‫ܭ‬௫ ൌ

converging
s
ܷ

5
ܸଶସ଴

15°
diverging
Turbine 2
converging
6

diverging
30°

converging
7
0°

diverging

converging
8
15°

diverging
Turbine 3
1
2
3

converging
9
30°

diverging
Wind-tunnel testing: results
Note: these are all point measurements, so not representative of the velocity to which
the wind turbine is exposed  whole-flow field data  CFD
Computational geometry and grid – based on grid-sensitivity analysis

1.6 million cells

Computational settings and parameters
Computational domain and grid
• Following best practice guidelines
• Grid-sensitivity analysis

Boundary conditions
• Logarithmic wind speed profile (U240 = 15 m/s, y0 = 0.1 m)
• Turbulent kinetic energy from wind-tunnel measurements: k = 0.5(σU2 + σV2 + σW2 )
• Turbulence dissipation rate: ε = (u*ABL)3/κ(y+y0)
• Ground surface roughness: kS = 9.793y0/CS
• Outlet: zero static pressure
• Top of computational domain: slip wall (zero normal velocity and zero normal gradients
of all variables).
Computational settings and parameters
Additional computational settings

• Realizable k-ε turbulence model (Shih et al. 1995)

• Standard wall functions (Launder and Spalding 1974) with sand-grain roughness
modifications (Cebeci and Bradshaw 1977) and roughness parameters according
to (Blocken et al. 2007)
• Pressure-velocity coupling: SIMPLE
• Pressure interpolation: second order
• Second-order discretization schemes (for both convection terms and viscous
terms)
 CFD simulations: validation: comparison with wind-tunnel experiments

ܷ
‫ܭ‬௫ ൌ
ܸଶସ଴
Kx - CFD

Kx - CFD

Kx - CFD
Kx - WT Kx - WT Kx - WT
CFD simulations: amplification factor K in horizontal plane ܸ
‫ܭ‬ൌ
ܸଶସ଴
CFD simulations: amplification factor K in horizontal plane ܸ
‫ܭ‬ൌ
ܸଶସ଴
CFD simulations: amplification factor K in horizontal plane ܸ
‫ܭ‬ൌ
ܸଶସ଴
CFD simulations: amplification factor K in horizontal plane ܸ
‫ܭ‬ൌ
ܸଶସ଴

Note: excellent design of tower

shape, re-directing the flow
parallel to the wind turbine axes
CFD simulations: amplification factor K in horizontal plane ܸ
‫ܭ‬ൌ
ܸଶସ଴
CFD simulations: amplification factor K in horizontal plane ܸ
‫ܭ‬ൌ
ܸଶସ଴
CFD simulations: amplification factor K in horizontal plane ܸ
‫ܭ‬ൌ
ܸଶସ଴

Note: excellent design of tower

shape, re-directing the flow
parallel to the wind turbine axes
 Kx

0.0
0.2
0.4
0.6
0.8
1.0
1.2

θ=
converging

1
T3
diverging

converging

2
T2
0°
diverging

converging

3
T1
diverging

converging

4
T3
diverging

5 converging
15°
T2
diverging

converging
6
T1

diverging

converging
7
T3

diverging

converging
8
T2

diverging
30°

converging
9
T1

diverging
CFD simulations: amplification factor Kx integrated over turbine swept area
CFD simulations: amplification factor K in horizontal plane ܸ
‫ܭ‬ൌ
ܸଶସ଴
CFD simulations: amplification factor K in horizontal plane ܸ
‫ܭ‬ൌ
ܸଶସ଴
 Kx

0
0.2
0.4
0.6
0.8
1
1.2

θ=
converging

1
Converging ‐ optilmal

T3
diverging

converging

2
Converging ‐ optilmal

T2
0°
diverging

converging

3
Converging ‐ optilmal

T1
diverging

converging

4
Converging ‐ optilmal

T3
diverging

converging
5
T2
Converging ‐ optilmal
diverging
15° converging
6

Converging ‐ optilmal
T1

diverging

converging
7

Converging ‐ optilmal
T3

diverging

converging
8

Converging ‐ optilmal
T2
30°

diverging

converging
9

Converging ‐ optilmal
T1

diverging
CFD simulations: amplification factor Kx integrated over turbine swept area
Intermediate conclusions:

• Diverging arrangement gives higher amplification of mean wind speed than

converging arrangement for wind direction parallel to the wind turbine axes (0°) and 15°.

• Diverging arrangement does not give higher amplification of mean wind speed than
converging arrangement for wind direction 30°.

• Improved position of wind turbines in converging arrangement gives higher

amplification of mean wind speed for all wind directions.
Transformation of wind velocity data to wind energy output
Wind statistics

N
 Wind turbine power curve
250
N Reference: Shaun K, Smith RF, 2008.
200 Harnessing energy in tall buildings:
Bahrain World Trace Center and
Beyond. CTBUH Technical paper.
Power (kW)

150

100

50

0
0 5 10 15 20 25
Wind speed (m/s)
 E (MWh)

- 0
50
100
150
200
250

θ=
conv

1
Conv ‐ opt

T3
div

conv

2
T2
Conv ‐ opt

0°
div

conv

3
Conv ‐ opt

T1
div

conv

4
T3
Conv ‐ opt
div

5 conv
Conv ‐ opt
T2
15° div
CFD simulations: yearly wind energy output

conv
6

Conv ‐ opt
T1

div

conv
7

Conv ‐ opt
T3

div

conv
8
T2

Conv ‐ opt
30°

div

conv
9

Conv ‐ opt
T1

div
 CFD simulations: yearly wind energy output
600
Killa & Smith (2008):
500
T3: 400 – 470 MWh/year
400 T2 : 360 – 430 MWh/year
T1: 340 – 400 MWh/year
E (MWh)

300

Conv ‐ opt

Conv ‐ opt
Conv ‐ opt

200
conv

conv

conv
div

div

div
100

-0

Turbine 3 Turbine 2 Turbine 1

 CFD simulations: yearly wind energy output
1 600

1 400 +31%
1 200
+14%

conv ‐ optimal
1 000
E (MWh)

800
converging

diverging
600

400

200

-0
 CFD simulations: yearly wind energy output
1 600

1 400 +31%
1 200
+14%

conv ‐ optimal
1 000
E (MWh)

800
converging

diverging
600

400

200

-0

Conclusion: Bahrain WTC has a good design, but it can be improved significantly.
MODULE QUESTION
Top view of a converging building arrangement. At which position does the highest
amplification of wind speed occur?

A) Position A: upstream of the passage

B) Position B: in the passage
C) Position C: downstream of the passage

wind
A
B
C
In this module, we have learned about:

• A particular case study of wind energy harvesting in the built environment

• How the aerodynamic design and wind energy output of the Bahrain World Trade
Center can be improved
In this week on “Building Aerodynamics”, we have learned about:

• Wind-flow patterns around buildings

• Why these patterns are so complex and why they are often misinterpreted
• How wind flow can create wind nuisance or even wind danger for pedestrians
• Important aerodynamic processes for buildings such as natural ventilation and
wind-driven rain
• Why successful integration of wind energy in the built environment requires the
detailed study of building aerodynamics
In the next week,
we will focus on:

100 m sprint
aerodynamics