Back to the programme printer.gif Print



Tuesday, 11 March 2014
16:30 - 18:00 How does the wind blow behind wind turbines and in wind farms?
Science & Research  


Room: Llevant
Session description

Wind turbines operate under highly fluctuating wind conditions. It is thus important to achieve a profound understanding of the characteristic features of the micro scale meteorological conditions. Current research activities focus not only on the inflow conditions and their impact on wind turbines, but also on the wake structures and the wind conditions within a wind farm.

Learning objectives

  • Get a better understanding micro scale wind conditions
  • Learn about new advanced measuring techniques
  • See the possibilities of numerical methods to simulate complex wind conditions
  • Learn about the impact of wind on turbine components
Lead Session Chair:
Joachim Peinke, Uni Oldenburg, Germany

Co-chair(s):
Jakob Mann, DTU Wind Energy
Ndaona Chokani ETH Zürich, Switzerland
Co-authors:
Balaji Subramanian (1) F P Ndaona Chokani (1) Reza Abhari (1)
(1) ETH Zürich, Zürich, Switzerland

Printer friendly version: printer.gif Print

Abstract

Instrumented drone measurements of 3D flow structure of multi-MW wind turbines

Introduction

Wakes are responsible for 10-20% loss in farm performance [1]. Moreover, fatigue loads on turbines in wakes are increased by up to 80% [2]. To reduce turbine lifecycle costs, there is substantial effort towards the development of fast and accurate computational tools [3, 4, 5] that may provide optimised turbine layouts within wind farms. In this work, the wind flow around multi-megawatt turbines are measured with a fleet of fully autonomous instrumented drones that have evolved in recent years to provide the required spatial and temporal resolutions in measurements within utility-scale wind farms to support the development of computational tools.

Approach

The fleet of drones is comprised of a series of first- and second-generation drones, Fig. 1 . The first-generation drones, Fig. 1a, were on the one hand proof-of-concept demonstrators used to develop the required software and hardware and on the other hand the platforms used to make the first instrumented drone measurements in utility-scale wind farms [6, 7]. The electric-powered, pusher-propeller drone has a wingspan of 795mm, overall length of 750mm, and take-off mass of 900g. The second-generation drones, windFlyer II, Fig. 1b, evolved from the experiences garnered with the first-generation drones. windFlyer II has a wingspan of 2280mm, overall length of 1150mm, and take-off mass of 2700g. With a potential flight endurance of 3hours (a fivefold improvement compared to the first-generation drone) windFlyer II provides substantially improved capabilities for measurements with superior spatial and temporal resolutions in utility-scale wind farms. The drones are instrumented with a nose-mounted, seven-sensor fast-response aerodynamic probes for wind measurements [8]; the accuracy of the probe measurement system has been verified under a wide range of controlled conditions in extensive wind tunnel tests, Figs. 1c and 1d. The drone is controlled and steered with an on-board, open-source autopilot system based on The hardware suite includes an IMU, a GPS, a magnetometer, a 14-channel 24 bit ADC that samples up to 500 Hz, and an on-board data storage. The sensors’ data are transmitted, in real-time, through an on-board modem, and are logged on to a ground computer. The Paparazzi autopilot system [9] has been adapted for the present system and is used for the fully autonomous flight of the drones. Furthermore, the wind tunnel measured aerodynamic characteristics are used to software to develop flight plans with optimized trajectories [10].

Main body of abstract

The wind velocity in the Earth frame of reference is obtained by a vector sum of the drone’s velocity in the Earth frame of reference and the wind velocity, measured by probe, in drone’s frame of reference. Thus, to determine the total uncertainty in measurement, the uncertainties of all components in the overall measurement chain are systematically combined using the guide to extended uncertainty in measurements (GUM). In the GUM method, all uncertainty sources are modelled as probability distributions, and are combined with Gaussian uncertainty propagation to yield the final uncertainty of the measurement system. The GUM analysis yields a standard uncertainty of 0.7 m/s in wind speed with a confidence level of 67%. In order to assess the in-field measurements of the instrumented drone simultaneous measurements were made with a 3D scanning LIDAR in Switzerland’s largest wind farm that is located in complex terrain. The 3D scanning LIDAR is installed in a customised van, windRover II [11]. For this series of measurements, the drone was programmed to fly along a straight horizontal line passing above windRover II; the LIDAR was programmed to track the drone’s trajectory. Since the LIDAR measures the line-of-sight (LOS) wind speed component, whereas the drone-mounted aerodynamic probe measures the three Cartesian wind speed components, the LOS wind speed was therefore derived from the drone-based measurements. In the conditions of a 8m/s horizontal wind speed, the LOS wind speed varies monotonically from 4m/s to -4m/s, Fig. 2 . Overall, there is very good agreement between the point measurements of the drone and the LIDAR measurements that are averaged within the cylindrical volume of the laser pulse.

In order to accomplish time localisation of the frequency components in flowfield measurements on the non-stationary drone, a short-time Fourier Transform algorithm has been developed. This approach allows the characteristics of the flow, such as wind speed and flow angles, to be simultaneously measured with the turbulence spectra. In Fig. 3 , vertical profiles of the wind speed, horizontal flow angle and turbulence intensity one diameter upstream and downstream of a 2MW wind turbine in complex terrain are shown; these profiles are derived from measurements over spanwise distances between Y/D= ±0.5. The jet-like profile of the upstream wind speed is washed out one diameter downstream due to mixing, Fig. 3a. The swept rotor area averaged flow angles are respectively 99deg. and 75deg. upstream and downstream of the turbine, Fig. 3b. The SCADA measured wind direction indicates that the turbine has a yaw misalignment of 9deg.; thus the wake is skewed by 24deg. relative to the incoming flow. The profiles of the mean turbulence intensity show the overall increase in turbulence levels in the wake compared to upstream of the turbine, Fig. 3c. Furthermore, at the locations of the rotor’s vertical extent, downstream of the rotor peaks in the mean turbulence intensity are observed and coincide with the largest deficits in the wind speed.

In the final version of the paper, the detailed measurements in the upstream, near-wake, intermediate-wake and far-wake shall be further discussed and detailed. The implications of these measurements for the development of reduced-order models for simulations of the wind field in wind farms shall be highlighted. Furthermore, one-to-one comparisons of the measurements and predictions from the computational tool that is described in [5] shall be detailed.


Conclusion

A fleet of first- and second-generation instrumented drones have been developed to make high spatial and temporal measurements of the wind field in utility-scale wind farms. The second-generation drones have evolved from the authors' experiences garnered with the first-generation drones. Detailed wind tunnel tests of the complete drone are used to assess the measurement system in well-controlled conditions. Furthermore, the aerodynamic data allow for the development of trajectory optimisation tools that are used to improve the productivity of the measurement system. Profiles of the upstream wind speed, wind veer, and turbulence, and their subsequent downstream evolution are detailed. A comparison of the upstream wind direction with the turbine's SCADA data shows that the turbine is yawed. Thus there is skew in the wake. A Short Time Fourier Transform based approach is used to determine the turbulence intensity. (In the final paper, detailed measurements in the upstream, near-wake, intermediate-wake and far-wake shall be further discussed.) These in-field measurements demonstrate that instrumented drones are well suited to improve our knowledge of flows in wind farms and to advance the development of CFD tools. (In companion work at ETH Zürich, novel immersed turbine models, which simulate the turbines’ three-dimensional flowfield, are integrated in our in-house preconditioned multistage Reynolds-Averaged Navier Stokes solver in connection with a k-ω turbulence model. This approach enables the wind flow and turbine wake to be simulated simultaneously, and allows atmospheric conditions and/or topography to be accounted for. The predicted evolution of wakes, and their associated wind speeds, directions, and turbulence intensities will be compared with the drone-based measurements.)


Learning objectives
• Grasp idea of operation of instrumented drones for wind field measurements
• Gain knowledge of validating measurement systems
• Be familiar with characteristics of wakes in complex and flat terrains
• Understand novel approaches to wake and atmospheric flow modelling in microscale CFD simulations



References
[1] Schepers, J. G. Obdam, T. S. Prospathopoulos, J., “Analysis of Wake Measurements from the ECN Wind Turbine Test Site Wieringermeer, EWTW”, Wind Energy, 15, 575-591, 2012.
[2] Sanderse, B., “Aerodynamics of Wind Turbine Wakes”, Technical Report, ECN-E-09-016, 2009
[3] Jafari, S., Chokani, N., Abhari, R. S., “Wind Farms in Complex Terrain: Numerical Simulation of Wind and Wakes for Optimised Micrositing”, Proceedings of EWEC2013, Vienna, Austria, 2013.
[4] Jafari, S., Chokani, N., and Abhari, R.S., “Simulations of Wake Interactions in Wind Farms Using an Immersed Wind Turbine Model,” presented at ASME Turbo Expo, San Antonio, USA, 2013.
[5] Jafari, S., Basol, A. M., Chokani, N., and Abhari, R. S., “Simulations of Atmospheric Flow and Wind Turbine Wakes Using a Preconditioned Multigrid,” to be presented at AIAA SciTech Meeting, Maryland, USA, 2014.
[6] Kocer G., Mansour M., Chokani N., Abhari R.S., Müller M., “Full-Scale Wind Turbine Near-Wake Measurements Using an Instrumented Uninhabited Aerial Vehicle,” Journal of Solar Energy Engineering-Transactions of the ASME, 133, 041011, 2011.
[7] Kocer G., Chokani N., Abhari R.S., “Wake Structure of a 2MW Wind Turbine Measured Using an Instrumented UAV,” AIAA Paper 2012-0231 presented at 50th AIAA Aerospace Sciences Meeting, Nashville, USA, 2012.
[8] Mansour, M., Kocer, G., Lenherr, C., Chokani, N., and Abhari, R. S., "Seven-Sensor Fast-Response Probe for Full-Scale Wind Turbine Flowfield Measurements," Journal of Engineering for Gas Turbines and Power-Transactions of the ASME, 133, 081601, 2011.
[9] Gati, B., "Open Source Autopilot for Academic Research - The Paparazzi System," presented at 2013 American Control Conference Washington DC, USA, 2013.
[10] Vogel, K.V., “Integrated Optimized Flight Trajectory and Autopilot for Wind Measurements using a UAV,” Internal Report, Laboratory for Energy Conversion, ETH Zürich, 2011.
[11] Zehndehbad, M., Chokani, N., and Abhari, R.S., “Experimental Study of Aero-Mechanical Damping on Full-Scale Wind Turbines,” to be presented at ASME Turbo Expo, Dusseldorf, Germany, 2014.