1. EWEC06 ( , Athens)
Numerical Site Calibration on a Complex Terrain and its Application for Wind Turbine Performance Measurements
Toshiyuki SANADA
(Mechanical Engineering Science, Kyushu University, Japan)
Masatoshi FUJINO
Daisuke MATSUSHITA
Takanori UCHIDA
Hikaru MATSUMIYA Masao WATANABE
Yoshinori HARA
Minori SHIROTA
Standardization research of wind turbine in Japan
Supported by NEDO (New Energy Development Organization)

2. Outline 1 Numerical Site Calibration A pair of masts
EWEC06 ( , Athens)
Outline 1
A pair of masts
Numerical Site Calibration
Reference meteorological mast
Wind Speed at a wind turbine
Wind speed
CFD (Computational Fluid Dynamics)
Meteorological mast
Fluid Dynamics Lab, Kyushu University

3. Outline 2 Flow Correction Factor due to Complex Terrain LES
EWEC06 ( , Athens)
Outline 2
Flow Correction Factor due to Complex Terrain
LES
Prediction of wind speed
GIS
Meteorological mast
Wind Turbine Performance Measurements

4. ~ Wind Turbine in Japan ~
Introduction
~ Wind Turbine in Japan ~
Renewable energy
1.35% of total electricity supply(2010)
Ratification of Kyoto Protocol
Renewable Portfolio Standard (RPS) Law 　
Wind power
almost 1/10 of Germany’s !!
March, MW
,000 MW (the official government target)
,800 MW (the new target of JWPA)
Challenges for Japanese wind turbine development
Complex terrain, Typhoon, Turbulence, Gust, Thunder storm …

5. ~ Wind Turbine in Japan ~
Introduction
~ Wind Turbine in Japan ~
Wind Turbines in Japan
Complex Terrain
for Japanese wind turbine development
mountain area 73% planes 14%
50% of the people live in plains.
Wind Turbine Performance Measurements at Complex Terrain
for next step…
Prediction of wind power generation

6. Performance Measurements
Introduction
Performance Measurements
IEC
Example
wind speed at reference mast　　 = wind speed at wind turbine
wind speed not influenced by wind turbine
requirement of high accuracy in measuring wind speed

7. Site Calibration Complex Terrain
Introduction
Site Calibration Complex Terrain
Flow correction factor b (=the wind speed at the wind turbine location divided by the wind speed at the meteorology mast)
Test site requirements (topographical variations)
・Maximum slope <5%　（between 2L and 4L）
・Eliminate the direction > 0.02 in b between neighboring sector
・Each wind direction bin, no larger than 10°
V1=bV2
IEC

8. Site Calibration Japan
Introduction
Site Calibration Japan
Wind Turbine Performance Measurements
Imamura et al. (EWEC01)
Complex terrain that fails to satisfy the IEC standard
Two points correlation analysis (before WT is constructed)
Correlation equation using 10miniuts average.
Direction with high correlation
(eliminate the direction with high turbulent intensity)
After measurements….,
If there is no direction with high correlation?
COST
RISK

9. Objective
To propose a numerical site calibration, which employs CFD for wind field simulation over a complex terrain to evaluate flow correction factor.

10. Methods To measure the wind speed at reference mast
Measurements
Numerical Analysis
Methods
To measure the wind speed at reference mast
To calculate the flow correction factor b using CFD
To correct the wind speed using b
To evaluate the wind turbine performance
V1=bV2
V1=bV2

11. Numerical Analysis Nonlinear, Unsteady simulation !
RIAM-COMPACT (Uchida & Ohya, 1999, 2003)
Grid scale vortex : Navier-Stokes Eq.
Sub-grid scale vortex : model
LES (Large Eddy Simulation)
Nonlinear, Unsteady simulation !
time series analysis
correlation
turbulent intensity
etc…

12. finite-difference method
Numerical Analysis
coordinate system
generalized curvilinear coordinate
variable arrangement
collocated arrangement
discretization method
finite-difference method
coupling algorithm
fractional step method
time advancement method
Euler explicit method
poisson equation for pressure
SOR method
convective terms
3rd-order upwind scheme based on an interpolation method (a=0.5)
other spatial derivative term
2nd-order central scheme
SGS model
Smagorinsky model + wall damping function
boundary condition
inflow
1/7 power low
top and side
free-slip
outflow
convective outflow
ground
no-slip

13. Digital Map Wind turbine performance measurements
Numerical simulation of wind turbine scale
Digital map of wind turbine scale
c.f. GSI map (50m, most commonly used)
GIS (Geographic Information System) technique
aero-photograph
Printed atlas
aero-photograph
CAD data

14. Computational Grid
160 x 160 x 60

15. Test Site ~takashima~ N
topographical variations
Point
Distance
Height difference
Mean slope
O-A
50m
25m
50%
O-B
75m
33%
O-C
62.5m
40%
Distance
Maximum slope
<2L
<3%
≥2L and <4L
<5%
≥4L and <8L
<10%
fails to satisfy the IEC standard : test site requirement site calibration

16. Measurements Spatial configuration of WT & mast N L=5.13D
mast1: reference mast
mast2: for validity check

17. Wind Characteristics NW, N, S mast1 mast2
1st, April, 05 ~ 30th, November, 2005
Wind Rose Average Wind Speed Average Turbulent Intensity
[%] [m/s] [%] N
NW, N, S

18. How to define the flow correction factor b ?M1
Example, North wind
Condition: 5m/s, 10m/s, 15m/s, 10minuts
time series of wind speed of each points
z=M1
mast 1 WT

19. How to define the flow correction factor b ?
10 minutes averaged value (corresponding to 10 minute data set)
regression line (minimum mean square method)　
correlation coefficient : high
However …
Uncertainly analysis

20. Uncertainly analysis Correction Factor
mast 1 WT
(a)
(b)
Assumption: bivariate normal distribution
u=N(0, s2): component of variation from average wind speed

21. How to define the flow correction factor b ?
confidence interval 95% (±1.97s)
evaluation of correction factor b
error of UWT
In this case,
a=21%
High accuracy in measuring wind speed is required.
The power is approximately proportional to wind speed to the third power.

22. Flow Field Results ~ NORTH WEST ~ N NW y=WT y zM2 M1 WT
Mast1 & WT: flow smooth
mast2: large Uz components
y=M2

23. Flow Correction Factor
Results ~ NORTH WEST ~ N NW
Flow Correction Factor
regression analysis
b=1.059
a=0.6%
Good direction for WT performance measurements

24. Validity Check Results ~ NORTH WEST ~ N NW
validity check : mast 1 & mast 2
Measurements (10min data-set)
Calculation
cup vs. cup
over estimation?
y=M2

25. Results ~ NORTH ~ N
Flow Field
Flow Field M1 M1 WT WT
z=mast1
z=mast1

26. Flow Correction Factor
Results ~ NORTH ~ N
Flow Correction Factor
regression analysis
b=1.2241
a=21%
Poor direction for WT performance measurements

27. Results ~ SOUTH ~ N S
Flow Field
Flow Field M1 WT
y=WT
z=WT WT

28. Flow Correction Factor
Results ~ SOUTH ~ N S
Flow Correction Factor
regression analysis
b=1.0844
a=1.0%
Good direction for WT performance measurements

29. Validity Check Results ~ SOUTH ~ What happen at the Mast 2 ?
Turbulent Intensity WT M2
Siting of Meteorological mast
a=53.0% !!

30. Power Performance ~ NORTH WEST ~NW

31. Power Performance ~ NORTH ~

32. Power Performance ~ SOUTH ~

33. Conclusion
A site calibration by employing CFD (LES and detail digital map obtained from GIS) is proposed.
A numerical site calibration can evaluate directional flow correction factor.
A numerical site calibration can be applied to the wind turbine performance measurements if terrain condition and flow direction are carefully chosen.
Using numerical simulation, the appropriate direction and position of meteorological mast for site calibration where fluctuation of wind speed is small, can be chosen.

34. For Future Numerical site calibration at other sites. (validity check)
Measurements of wind turbine performance at complex terrain
(effects of flow distortion, turbulent intensity)
(For Japanese wind turbine development, a performance testing of WT at complex terrain is required for prediction of electric power generation)

35. Validity Check Results ~ NORTH ~ N validity check : mast 1 & mast 2
Measurements (10min data-set)
Calculation
Measurements: north wind at mast1 = Calculation : north wind into island

36. Validity Check Results ~ SOUTH ~ N S validity check : mast 1 & mast 2
Measurements (10min data-set)
Calculation
Qualitatively good agreement