Downloaded from orbit.dtu.

dk on: Apr 18, 2019

Prediction of wind energy distribution in complex terrain using CFD

Xu, Chang; Li, Chenqi; Yang, Jianchuan; Shen, Wen Zhong; Zheng, Yuan; Liu, Deyou

Published in:
Proceedings of the 2013 International Conference on aerodynamics of Offshore Wind Energy Systems and
wakes (ICOWES2013)

Publication date:
2013

Document Version
Publisher's PDF, also known as Version of record

Link back to DTU Orbit

Citation (APA):
Xu, C., Li, C., Yang, J., Shen, W. Z., Zheng, Y., & Liu, D. (2013). Prediction of wind energy distribution in
complex terrain using CFD. In W. Shen (Ed.), Proceedings of the 2013 International Conference on
aerodynamics of Offshore Wind Energy Systems and wakes (ICOWES2013) (pp. 549-557). Technical University
of Denmark (DTU).

General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright
owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

 Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
 You may not further distribute the material or use it for any profit-making activity or commercial gain
 You may freely distribute the URL identifying the publication in the public portal

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.
 Prediction of Wind Energy Distribution in Complex Terrain using CFD
Chang Xu 1, Chenqi Li1, Jianchuan Yang1, Wenzhong Shen2, Yuan Zheng1, Deyou Liu1
1
Hohai University, College of Energy and Electrical Engineering, Nanjing, Jiangsu Province, 211100, China,
[email protected] ,[email protected]
2
Technical University of Denmark, Department of Wind Energy, Nils Koppels Alle, Building 403, 2800
Lyngby, Denmark, [email protected]

ABSTRACT
Based on linear models, WAsP software predicts wind energy distribution, with a good accuracy for flat
terrain, but with a large error under complicated topography. In this paper, numerical simulations are carried
out using the FLUENT software on a mesh generated by the GAMBIT and ARGIS software to predict wind
speed distribution in complex terrain. TECPLOT software post-processing is used to get the whole wind
flow field, the wind speed distribution characteristics and distribution of wind energy. The obtained results
are compared with the results of WAsP software and are also more accordance with the actual conditions.
Keywords: wind farm, complex terrain, wind energy resources evaluation, CFD numerical simulation

0 INTRODUCTION
In recent years, wind energy has become the fastest growing renewable energy [1-3]. Wind
energy resource assessment is a basic work for the construction of wind farms. Power calculation of
a wind farm is directly related to economic factor and wind power grid connection. Currently, wind
energy resource and power capacity are usually calculated by WAsP software developed by Danish
RisФ laboratory. The linear model was adopted in the software.Wind energy resource assessment
and wind power output are often accurate for the flat terrain and would be higher error for the
complex terrain. The software should be further improved for increasing the prediction accuracy [4-5].
With the wind energy resource exploitation further development, wind farm site is transferring from
flat terrain of wind speed stability and good construction conditions to complex terrain of high
turbulence and bad construction conditions [6-8]. So it is essential to develop a novel method or code
to estimate the wind energy distribution or the wind farm layout.
Due to the complex topography, wind energy distribution and turbine output calculations are
difficult currently because of the aerodynamic detouring flow and the turbine wake models, and the
former is usually the key point. At present wind mast data are very too limited to estimate the whole
wind farm energy distribution accurately. With rapid development of CFD technology in recent
years, many researchers tried to apply the technology to estimate the wind energy for the wind farm
[9-13]
. This paper tried to develop a numerical calculation method by combining the topography data
process, CFD model and calculation, post-processing. Through the calculation analysis of wind
speed and wind energy distribution of a wind farm, the results obtained are compared to the results
of WAsP software and it could show the more accordance with the actual conditions. At the same
time the method developed by the paper is easy to apply to the actual engineering.

1 TERRAIN DIGITAL PROCESSING

Complex terrain of a wind farm is usually depicted by Autocad isohypse contour, But when
calculating aerodynamic field by CFD, it needs to digitize the complex terrain for developing the
physical model. Argis is a powerful software which can analysis geographic information system,
digital map, geographic information acquisition, and is also the most powerful and the most widely
used in geographic information system industry. This paper uses the Argis to get tin file by

549
separating contour figure of Autocad file, and the tin file can be turned into dem file of coordinates
by discretization. In this paper, the wind farm calculated is from north China. The original Autocad
isohypse contour is shown in figure 1, and the terrain discreted is shown in figure 2 by the Argis
software. Figure 3 is the wind farm contour map, which is fitted to the results obtained by Argis at
height. By the comparison and analysis, the topographic map discreted by ARGIS software is in
accordance with the original digital topographic map with very high accuracy.
0

-1000

-2000

-3000
Y (m )

-4000

-5000

-6000

-7000
0 2000 4000 6000 8000
X(m)
Figure 1: Two-dimensional Autocad terrain isohypse contour

Figure 2: Argis map

According to the complex topography and discreted digital files, develop the physical model of
topography in the Gambit software. According to the topographic data, the selection of calculation
area is 7000 m * 7000 m * 500 m, the blank terrain in original map complemented by surrounding
terrain ,as shown in figure 4, The height is 500 m, choose the unstructured grid, ground surface grid
for 30 m by 30 m, In the vertical direction on the surface to mesh with 500 m height is divided into
three layers, 0 to 50 m, 50-200 m, 200-500 m respectively, and the grid interval is 5 m, 10 m and 30
m respectively .

550
 0
1480
-1000
1460
-2000
1440
-3000
Y (m )
-4000
1420
1400
-5000 1380
-6000 1360

-7000 1340
0 2000 4000 6000 8000
X(m)
Figure 3: Wind farms contour map

Figure 4: Wind farm calculation zone

2 MATHEMATICAL MODEL AND CALCULATION

Standard Navier - Stokes equations (cartesian coordinate system) are expressed as
equations (1)-(4) [14]：
 
 div(  u )  0 (1)
t
u  p
  div( uu)  div( grad u)  (2)
t x
v  p
  div(  vu )  div(  grad v)  （3）
t y
w  p
  div(  wu )  div(  grad w)  (4)
t z

551
 This article chooses the steady, constant physical property control equations and
the standard k -  turbulence models are used [15] ：
 k 
 div(  ku )  div( k grad k )  G   （5）
t
 k  C 2
 div(  u )  div( k grad  )  1 G  C 2   （6）
t k k
 is the density of air, and u , v the horizontal wind speeds respectively, w the
vertical wind speed, p the air pressure,  the turbulence viscosity,  k diffusion
coefficient of k , G the turbulent energy generation rate or the kinetic energy
dissipation rate, G   the net source term and the model constant. Equation
parameters are set as: second order discretization form of the upper wind method，
bottom boundary conditions are processed with the wall function method, entrance is
velocity inlet, outlet is outflow, others are all symmetry conditions. Equations are
solved by the simplic algorithm. According to the wind speeds from the wind masts in
the wind farm. Wind rose diagram is divided by 12 sections, the average wind speeds
of each section are calculated. And the boundary condition is used by equation (7):
u  z 
u ( z )  * ln   （7）
  z0 
u* is speed of surface friction coefficient,  is the von Karman constant (0.4), z0 is
surface roughness length . Inlet boundary condition is coupled by Fluent UDF editor.
Wind velocity inlet of 0 to 300 direction is shown as Figure 5, it can be seen the trend
of velocity inlet increases along the height direction. The ground wind speed is close
to zero.

.
Figure 5: speed inlet condition from 0 to 30 degree

3 RESULT ANALYSIS
3.1 Wind calculation results
Wind energy density is defined as wind energy power per area [16], Thus wind
energy density formula, also called wind power density formula is expressed as:
1 3
w v （8）
2

552
 .
Figure 6: Wind energy distribution ( z=1600m)
Aerodynamic field computations are carried out by FLUENT software in 12
directions. Wind energy distributions are summerized according to wind speed
probability distribution of each direction by TECPLOT software. Wind energy
distribution is shown as from Figure 6 to Figure 8 at different heights. Wind energy
distribution will increase with height increasing. But there is very big difference at the
same height level because of different height and detouring flow. Figure 9 and Figure
10 show the local wind energy distribution and velocity distribution at the position of
the topographic map, which the specific scope of coordinates is x = 2500-5000 m and
y = - 3000 m. In Figure 9, the wind energy distribution is obtained from 12 wind
speeds due to probability distribution, Figure 10 is the wind velocity distribution
when the wind blowing hillside at the wind direction of 0 to 30 degree zone, which
can be seen from the figure leeward slope wind velocity is smaller than windward
slope wind velocity, the wind speed decreases caused by slope block as a result of the
energy loss.

Figure 7: Wind energy distribution ( z=1700m)

553
Figure 8: Wind energy distribution ( z=1800m)

Figure 9: Local wind energy distribution

Figure 10: Local wind velocity distribution

554
 3.2 Comparison between numerical simulation and WAsP software
In the CFD calculation result of the wind farm, along the prevailing wind direction,
20 evenly distribution points are checked by the results of the CFD numerical
simulation and the WAsP software respectively. Results show that there are some
differences between these two methods. In the Tab.1, points from 6 to 20 , CFD
results are smaller around 100-200 w/m2 than the results of the WAsP, and the first
five points is close to the edge of the terrain, the result of range is bigger. From
Figure 11,we can intuitively see, just the second calculation point in the distribution
of wind energy outweighs the WAsP. The other CFD calculation results are less than
the WAsP calculation results. The calculation results are consistent with the
engineering results, which wind distribution calculated in complex terrain using
WAsP is often bigger than the measurement from the wind farm at present. This is
because the WAsP is just taking consideration of the height influence on the wind
speed by Lissaman model, while ignoring the detouring flow of the terrain. Otherwise,
CFD numerical simulation method takes the terrain height and detouring flow effect
on wind speed distribution, thus more accord with the actual wind power distribution
results obtained.
Tab.1 distribution of wind energy in 20 calculated point by CFD and WAsP
CFD/ WAsP/ Difference/
Points X/m Ym/ Z/m
（w/m2） （w/m2） （w/m2）
1 150.8621 -6849.14 1428.775 352.0798 378.3076 26.2278
2 452.5862 -6547.41 1487.5 568.931 432.9725 -135.9585
3 754.3103 -6245.69 1459.047 394.0908 412.5912 18.5004
4 1056.034 -5943.97 1415.502 214.5136 354.0609 139.5473
5 1357.759 -5642.24 1428.367 319.9514 377.6848 57.7334
6 1659.483 -5340.52 1419.449 215.2356 362.272 147.0364
7 1961.207 -5038.79 1447.859 270.9516 402.0661 131.1145
8 2262.931 -4737.07 1434.054 195.2646 385.8512 190.5866
9 2564.655 -4435.34 1438.466 246.8089 391.5182 144.7093
10 2866.379 -4133.62 1419.913 149.0633 363.1724 214.1091
11 3168.103 -3831.9 1435.779 163.1859 388.1296 224.9437
12 3469.828 -3530.17 1480.485 271.2416 428.591 157.3494
13 3771.552 -3228.45 1455.358 220.3219 409.3249 189.003
14 4073.276 -2926.72 1452.328 235.1058 406.4986 171.3928
15 4375 -2625 1441.018 205.4146 394.5711 189.1565
16 4676.724 -2323.28 1459.547 265.2732 413.0201 147.7469
17 4978.448 -2021.55 1482.582 293.8114 429.9375 136.1261
18 5280.172 -1719.83 1477.5 316.9961 426.6165 109.6204
19 5581.897 -1418.1 1433.579 194.8793 385.2088 190.3295
20 5883.621 -1116.38 1415.281 254.5544 353.5692 99.0148

555
 Figure 11: Wind energy diagram in calculation 20 points by CFD and WAsP

4 CONCLUSIONS
(1) In the complex terrain, it was compared that the wind power distribution
calculation results of the CFD and the WAsP software, and CFD calculation method
calculates more accurately the wind flow over the complicated topography and the
distribution of wind energy.
(2) WAsP is a traditional wind resource evaluation software based on the linear wake
model and Lissaman model. Calculation error is often a little big for complex terrain.
CFD numerical simulation method for complex terrain can take the effects of wind
speed change with height by detouring flow accurately, the wind energy calculation
results are often smaller than the results of WAsP. This matches the fact that WAsP
calculation results are often over-estimated in complex terrain.

REFERENCES
[1] HAN Chun-fu. Wind energy resources evaluation method analysis and application [J].
Energy Conservation, Vol.322,No.5 (2009)pp.22-24.
[2] HU Yi, ZHANG Jian. Research on wind speed of wind resources evaluation [J]. Inner
Mongolia Science Technology & Economy, No.21 (2011) pp.76-78.
[3] Liang Si-chao, ZHANG Xiao-dong, KANG Shun, KANG Yalan, ZHAO Yongfeng. A
method of wind farm wind resources evaluation in complex terrain Based on numerical
simulation [J]. ACTA AERODYNAMICA SINICA , Vol.21,No.3 (2012) pp. 415-421.
[4] LAI Yong-lun, WU Qing. Analysis on application of WAsP to wind energy resource
assessment for Sige Wind Farm in Guizhou [J]. HongShui River, Vol.28,No.4 (2009) pp.
106-108
[5] YANG Zhen-bin, XUE Heng, SANG Jian-guo. A trial research on complex terrain wind
energy resources evaluation[J]. Acta Energiae Solaris Sinica, Vol.25,No.6 (2004)
pp.744-749.
[6] Yang X-Y, Xiao Y, Chen S-Y. Wind Speed and Generated Power Forecasting in Wind
Farm[J]. Proceedings- Chinese Socity of Electrical Engineering. Vol.25,No.11 (2005)
pp.1.

556
[7] Berge Erik, Nyhammer Finn,Tallhaug Lars, Jacobsen Øystein. An evaluation of the
WAsP model at a coastal mountainous site in Norway[J]. Wind Energy. Vol.9,No.1
(2006) pp.131-140.
[8] Kusiak Andrew,Zheng Haiyang. Optimization of wind turbine energy and power factor
with an evolutionary computation algorithm[J]. Renewable Energy. Vol.35,No.3 (2010)
pp. 1324-1332
[9] FENG Chang-qing, DU Yan-jun, BAO Zi-guang, XUAN Ji-xin. The applicability
between Wind energy resources evaluation software WAsP and WT [J]. Electric Power,
Vol.43,No.1 (2010) pp. 61-65.
[10] H. A. Madsen, G. C. Larsen and T. J. Larsen. Calibration. validation of the
dynamic wake meandering model for implementation in an aeroelastic code[J]. Journal
of Solar Energy Engineering, Vol.132,No.4 (2010) pp. 41-71.
[11] J. Berg,J. Mann, A. Bechmann,H. E. Jørgensen. The Bolund Experiment, Part I：
Flow Over a Steep, Three-Dimensional Hill[J].Boundary-Layer Meteorology.
Vol.142,No.2 (2011) pp.219
[12] Bingöl Ferhat,Mann Jakob,Foussekis Dimitri. Conically scanning lidar error in
complex terrain [J]. Meteorologische Zeitschrift. Vol.18,No.2 (2009) pp. 189-196.
[13] Kindler Detlef,Oldroyd Andrew,MacAskill Allan,Finch Danny. An eight month test
campaign of the Qinetiq ZephIR system ： Preliminary results[J]. Meteorologische
Zeitschrift. Vol.16,No.5 (2007) pp.479
[14] WANG Mei-ling, LUO Yong, ZHOU Rong-wei. Application of WindSim to Wind
Energy Resources Assessment of Complex Terrain in china [J]. Meteorological
Monthly, Vol.36,No.2 (2010) pp. 113-119.
[15] WANG Fu-Jun. Computational fluid dynamics analysis[M]. Bei Jing: Tsinghua
university press,(2004)pp.113-119
[16] TIAN Zi-Chan, YANG Yong-Ping, LIU Yong-Qian. Complex terrain of wind
resources evaluation research[D]. Bei Jing: North China electric power university,2009.

557