Back to the programme printer.gif Print




Delegates are invited to meet and discuss with the poster presenters in this topic directly after the session 'How does the wind blow behind wind turbines and in wind farms?' taking place on Tuesday, 11 March 2014 at 16:30-18:00. The meet-the-authors will take place in the poster area.

Ndaona Chokani Laboratory for Energy Conversion - ETH Zürich, Switzerland
Co-authors:
Michel Mansour (1) F P Thomas Schmid (1) Ndaona Chokani (1) Abhari Reza (1)
(1) Laboratory for Energy Conversion - ETH Zürich, Zürich, Switzerland

Printer friendly version: printer.gif Print

Presenter's biography

Biographies are supplied directly by presenters at EWEA 2014 and are published here unedited

Michel Mansour received his Master of Mechanical Engineering from EPF Lausanne in 2002 and his Ph.D. in engineering from ETH Zürich in 2008. He is presently project manager at Limmat Scientific AG, a spin-off company of ETH, and senior assistant in charge of the instrumentation group at the Laboratory of Energy Conversion at ETH Zürich.

Abstract

Assessment of power performance of a 2MW wind turbine located in complex terrain

Introduction

In the fast growing wind energy market, an accurate prediction of on-site power performance is of primary importance to minimize the financial risk, and thereby reduce the cost of energy. As reported in [1], the terrain (flat or complex), diurnal and seasonal effects, as well as wind farm induced effects, are among the reasons why on-site power performance often differs substantially from the turbine’s performance under idealized conditions.

This work details the impacts of terrain and atmospheric conditions on the on-site power performance of a 2MW wind turbine that is located in complex terrain.


Approach

Measurements of the atmospheric conditions that are made with a novel, fast response aerodynamic probe are complemented with mesoscale and microscale computations. This combined use of measurements and computations provides a framework to better clarify the factors that impact wind turbine performance.

The measurements are made on a Vestas V90-2MW wind turbine, turbine WTG10, in the Mont Crosin wind farm that is located in complex terrain, Fig. 1a. An examination of the wind rose from mesoscale computations shows that this turbine operates free of wake influences from adjacent wind turbines. (Note: over the period of measurements, the turbines WTG1 and WTG3 were not operational). The SCADA measured power output is correlated to the atmospheric conditions that are measured by a nacelle-mounted, fast-response aerodynamic probe that is in close proximity to the turbine’s ultrasonic anemometer, Fig. 1b.

Fig. 1: (a) Layout of the Mont-Crosin windpark, Switzerland’s largest wind farm that is located in the complex terrain of the Jura. (b) Photograph of the fast-response aerodynamic probe system and ultrasonic anemometer on turbine WTG10.

The measurements of the atmospheric conditions (that is, wind speed, wind direction, and turbulence intensity) are made with a custom-built, 5-sensor fast-response aerodynamic probe. Details of the probe have been presented elsewhere [3]; this measurement technology is based on longstanding experience in the development and application as described in [4, 5]. The probe’s measurements are acquired at 200Hz, which allows the upstream incoming wind conditions to be distinguished from the wake and secondary flows that evolve from the blade and rotor hub. The probe’s measurement accuracies are 0.41m/s in wind speed, and 0.08° and 0.13° in yaw and pitch angles, respectively. The differential pressure transducers used for the aerodynamic probe are automatically re-zeroed every ten minutes in order to ensure the long-term accuracy, and the system is also equipped with a purging and anti-icing system

Main body of abstract

Fig. 2 shows representative time-resolved measurements of wind speed and the associated turbulence intensity; the measurements are at mean wind speeds of 5m/s, Fig. 2a, and 10m/s, Fig. 2b. The rotor and hub-induced wakes and secondary flows are identifiable from the elevated levels of turbulence intensity and the relatively large fluctuations in the crosswise wind speeds. A window filtering method that is based on the turbulence levels is applied to filter out the measurements related to the wakes and secondary flows. This filtering retains only the data that are representative of the upstream incoming wind conditions. Fig. 2c compares the filtered 10-minute averaged wind speeds measured with the aerodynamic probe to the SCADA measurements.

Fig. 2: Time-resolved measurements of absolute wind speed (blue), crosswise wind speeds (black and magenta) and turbulence intensity (red) measured over 1 rotor revolution. Mean wind speeds are (a) V = 5m/s and (b) V= 10m/s. Shaded grey regions show measurements that are filtered out on the basis of high turbulence levels to retain only measurements that are representative of the upstream incoming wind conditions. (c) Correlation of the filtered 10-minute averaged wind speeds measured with the aerodynamic probe (FRAP) and the SCADA measured wind speeds.

The onsite measured power curve is shown in Fig. 3a. The power curve shows large scatter in power output over all wind speeds. The measured power curve differs substantially from the manufacturer’s power curve. At lower wind speeds, 4-6 m/s, the turbine under performs, whereas for wind speeds of 7.5 m/s to 11.5m/s the turbine over performs compared to the manufacturer’s power curve with a 0.18kW increase (13% increase) in power at 9.8m/s.

Fig. 3: (a) Measured on-site power curve of a Vestas V90 at Mont-Crosin wind park. Measured on-site power curve binned with respect to (b) wind directions 0 to 90° (blue) and 135° to 230° (red), (c) pitch angle of incoming flow -13° to -7.5° (blue) and -4 to 0° (green), (c) turbulence intensity 0 to 10% (blue) 15% to 25% (red), (d) time of day nighttime 7h00 to 19h00 (blue) and daytime 19h00 to 7h00 (red).

The measured power curve shown in Fig. 3a, is binned with respect to wind direction, Fig. 3b, pitch angle of incoming flow, Fig. 3c, turbulence intensity, Fig. 3d, and time of the day, Fig. 3e. The largest overall difference between the binned and manufacturer’s power curves are seen on account of the wind directions over wind speeds from 7 m/s to 13m/s. The more negative pitch angles of the inflow (-7.5° to -13°) (negative indicating an upward flow direction) show that there is a relatively large under performance over wind speeds of 4-7m/s, Fig. 3c. On the otherhand, it can be seen that at lower turbulence levels (less than 10%) the actual power performance is above that of the manufacturer’s power curve for wind speeds above 9.5m/s, Fig. 3d.

The measured turbulence intensity is shown in Fig. 4a for different wind directions. The largest levels of turbulence intensity are for wind directions between 135 and 220°. The complex topography ahead of the turbine WTG10, Fig. 1a, results in the elevated levels of turbulence in this sector. Microscale simulations of the atmospheric flow using the CFD tool described in [2] are used to detail the impacts of the terrain on the wind shear, Fig. 4. Mesoscale simulations [6] provide the boundary conditions for the microscale simulations. Fig. 4b shows the predicted profiles of wind speeds for the two different wind direction sectors. It can be seen that these profiles differ from the commonly used logarithmic profile.

Fig. 4: (a) Turbulence intensity as a function of wind direction for the flow incoming to turbine WTG10, (b) Predictions from the microscale CFD simulation described in [2]. Vertical profiles of wind speeds across rotor swept area. Profiles are averaged from profiles over wind directions of 0° to 90° (green) and 135° to 220° (blue). The logarithmic wind profile is determined with roughness length of zo= 0.03m). All wind speed profiles have a hub height wind speed of vhub = 10m/s.


Conclusion

The impacts of terrain and atmospheric conditions on the on-site power performance of a 2MW wind turbine that is located in complex terrain has been investigated. The current investigation was conducted using measurements of the atmospheric conditions that are made with a novel fast response aerodynamic probe. The probe provides the simulataneous measurement of wind speed as well as yaw and pitch angles, from which turbulence is inferred. These measurements are completed with mesoscale and microscale computations of the atmospheric boundary layer for various asbolute wind directions .

It has been shown that the current onsite power curve deviates substantially from the manufacturer curve and tends to overperform for wind speeds ranging from 7.5 up to 11m/s due to the complexe topography present around the wind park. The topography induces large variations in wind shear and turbulence resulting in a strong coupling of the power curve to the wind absolute direction. The elevated turbulence intensity levels present between 135° and 220°direction results in a reduction in power performance for wind speeds above 9m/s. However the steep characteristic wind shear profile at this site results in an higher equivalent wind speed over the rotor swept area which is responsible for the onsite power curve overperformance. On the other end large pitch angles of the inflow (-7.5° to -13°) measured around sunset were responsible for large under performance at moderate wind speeds. This combined use of measurements and computations provides a framework to better clarify the factors that impact wind turbine performance.

In the full-length paper, the impacts on the power curve and turbine’s annual energy yield due to turbulence, wind shear and inflow pitch angle will be further discussed and quantified.



Learning objectives
Clarify and quantify the factors that impact onsite wind turbine power performance in order to improve wind park micositting approaches and therefore mitigate the investment risks


References
[1] E. Rareshide, T. Andrew, C. Johnson, A.-M. Graves, E. Simpson, J. Bleeg, T. Harris, and D. Schoborg, "Effects of Complex Wind Regimes on Turbine Performance," in AWEA WINDPOWER, Chicago, USA, 2009.
[2] S. Jafari, N. Chokani, and R. S. Abhari, "Wind Farms in Complex Terrain: Numerical Simulations of Wind and Wakes for Optimised Micrositing," in Proceedings of EWEA Annual Event 2013, Vienna, Austria, 2013.
[3] M. Mansour, C. Atalayer, N. Chokani, and R. S. Abhari,, "Time-Resolved Near-Wake Measurements of a 2MW Wind Turbine," presented at ASME Turbo Expo 2013, San Antonio, USA, 2013.
[4] M. Mansour, G. Kocer, C. Lenherr, N. Chokani, and R. S. Abhari, "Seven-Sensor Fast-Response Probe for Full-Scale Wind Turbine Flowfield Measurements," Journal of Engineering for Gas Turbines and Power-Transactions of the ASME, vol. 133, Aug 2011.
[5] K. Ettlin, M. Mansour, D. Costa, N. Chokani, and R. S. Abhari, "Near-Wake Measurements in an Offshore Reference Field Using a Kite-Borne Aerodynamic Probe System " presented at EWEA Annual Event 2013, Vienna, Austria, 2013.
[6] S. Jafari, T. Sommer, N. Chokani, and R. S. Abhari, " Wind Resource Assessment Using a Mesoscale Model: The Effect of Horizontal Resolution," “Wind Resource Assessment Using a Mesoscale Model: The Effect of Horizontal Resolution,” presented at ASME Turbo Expo 2012, Copenhagen, Denmark, 2012.