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.

Samuel Davoust Avent Lidar Technology, France
Samuel Davoust (1) F P Andrew Scholbrock (2) Paul Fleming (2) Alban Jéhu (1) Mathieu Bardon (1) Michael Bouillet (1) Benoist Vercherin (1) Alan Wright (2)
(1) Avent Lidar Technology, Orsay, France (2) National Renewable Energy Laboratory, Boulder, United States

Printer friendly version: printer.gif Print


Assessment and optimization of lidar measurement availability for wind turbine control


Integrating Lidar to improve turbine controls is a potential breakthrough for reducing the cost of wind energy [1]. By providing undisturbed wind measurements up to 200m in front of the rotor, Lidars provide an update of the turbine inflow with a preview time of several seconds. Several studies have evaluated potential benefits using integrated Lidar, either by simulation [2, 3] or full scale field testing [4]. Leading turbine manufacturers are also evaluating the technology [5]. One of the key aspects to be validated is the availability of Lidar measurements: if the Lidar is not delivering measurements, benefits cannot be obtained.


The present work provides a first overview of the measurement availability which can be achieved with a turbine mounted Lidar under realistic operational conditions. The objective is to describe future methods and solutions to optimize Lidar integration for turbine control. The two main parameters which are considered are the atmospheric conditions and the Lidar configuration on the wind turbine.

First, the effects of atmospheric conditions are investigated using operational feed-back from a Lidar installed at the National Renewable Energy Laboratory test site in Boulder Colorado. This has allowed to characterize the effects of several meteorological events such as rain, snow or fog on Lidar availability. These events are likely to occur during the operational life of a future wind turbine controlled using inputs from a Lidar, with different probabilities which will depend on the localization of the turbine. In parallel, a model of the atmosphere transmission and backscattering coefficients is employed to generate simulations matching the observations.

Second, data collected from over 20 turbine mounted Lidars deployed in several configurations is used to investigate operational measurement availability. The mains parameters which are considered are the Lidar mounting position, the Lidar sampling rate and the rotor rotation speed. Recorded measurement availabilities are then compared to what can be expected from simple geometrical models, taking into account the available Lidar field of view, the integration time for a Lidar measurement and the duration of a blade’s passage in front of a Lidar beam. A study is performed to determine optimal positions for a Lidar on the nacelle, and guidelines are provided to configure the best Lidar sampling rate as a function of measurement objectives.

Main body of abstract

1. Effect of atmospheric conditions

Under specific conditions, a Lidar measurement can become non-available at one range if the returned signal is not strong enough or does not match quality filters. An analysis of Lidar availability under various atmospheric conditions has been performed. Over the course of a 6 month campaign, a Wind Iris Lidar was installed on the NREL 2-Bladed Controls Advanced Research Turbine located in Boulder. The Lidar is a 2 beam pulsed Doppler Lidar measuring from 50 to 200m (10 ranges acquired every 2Hz). Using instrumentation including a webcam facing the turbine, weather categories were created an shown as selected examples .

Under nominal conditions (clear sky), availability is constant and near 100% up to 160m. During the heavy rain event (7 mm in 10 minutes), availability is observed to be maintained, suggesting that rain is not degrading it. Degradation of availability is however observed during snow event. In a thick fog (visibility below 50m), CNR strongly increases at close range, but availability is seen strangely to be lower at 50m than at 200m.

The signal to noise ratio “CNR” is the one of the main parameters to flag one Lidar wind measurement valid [6]. A theoretical computation of Lidar signal was performed in order to understand the behavior of average CNR. The main parameters are the atmospheric transmission (α) and backscattering (β) coefficients. The results obtained {f2] show that the simulations approximate well the Lidar signal in clear sky and heavy fog situations, allowing to extrapolate and further results in harsher conditions.

2. Effect of Lidar configuration

Different mounting positions on the turbine have an effect on the Lidar availability based on the interference between the Lidar laser signal and the turbine’s blades. The Lidar can be mounted with various positions on the turbine nacelle . Since the turbine’s blades can interfere with the Lidar beams, special consideration is needed to ensure this doesn’t affect the wind measurement. Another possibility is to install the Lidar inside the turbine’s spinner [7].

Geometrical aspects:
When the Lidar is behind the blades, a percentage of average available field of view Δ is allowed by the rotor. The evolution of Δ as a function of the Lidar position behind the rotor can be predicted , considering different Lidar beam opening angle and heights above the rotor axis. Bringing the Lidar at the back of the nacelle and raising it above the hub both improve Δ. Similarly wider beam opening angles improve availability. These results are in line with field observations from over 20 documented Lidar installations on various nacelle types from 1.5MW up to 7.5MW turbines.

Temporal aspects:
Temporal parameters such as the Lidar measurement integration time, and the rotor rotation speed also have an effect on availability. Indeed, the availability for one Lidar beam is expected to depend on rotor or generator speed . For low rpm, one can see that availability is near 65%, which is the valid field of view Δ, but reaches 100% above 80% of nominal rpm.

Dividing the Lidar integration time by 2 removes 3dB of CNR. This leads to a loss of availability that depends on CNR level with respect to the threshold. Thus, the ratio Σ between Lidar integration time and the time taken for a blade to pass in front of a beam is an important parameter. If Σ>1, then the Lidar will tend to win back the lost availability, which is 1-Δ. If Σ<1, as a first model, one can assume that lost availability is proportional to Σ. Thus, the parameter which should set availability writes Α=Δ + Σ(1-Δ).

In order to evaluate this model, availability was investigated using Lidar data from the 20 documented configurations described above. A wide of parameters Σ and Δ has been covered and a correlation is observed between availability and model A . This confirms the existence of two regimes evidenced above. When A<1 (if Σ is small), the Lidar availability is governed by Δ, the available field of view. When A>1 (if Σ is large) the Lidar measurement availability should be 100%.


Measurement availability has been investigated for a pulsed Doppler turbine mounted Lidar for a wide range of parameters including Lidar characteristics, mounting configurations and atmospheric conditions. It has been demonstrated that the main effects of these parameters can be modeled and predicted. A critical analysis of availability performance with respect to turbine control application is provided, and optimization strategies have been proposed to remedy cases of low performance.

For atmospheric effects, heavy rain did not appear to reduce directly Lidar availability. However, it was evidenced that during a snow event availability was reduced at close and long range, yet acceptable availability could be conserved around the Lidar 100m range. Fog caused a very strong signal a close range, validating the principle that such close range measurements would still be made, providing some modification of current Lidar filtering rules. Nevertheless, availability can be expected to decrease at longer range.

For a Lidar on a wind turbine nacelle, it has been shown that a position at the back and as high as possible results in the highest availability. Considering one Lidar beam, the ratio of blade passage to Lidar integration time was used to show that the effects of blades can be reduced or suppressed.
As a future topic of research, the convection time of wind disturbances with a Lidar acquiring different ranges at the same time can be used as a form of spatio-temporal redundancy. Similarly, a Lidar with multiple beams will need to be considered and may bring further redundancy. Together with further specifications from turbine manufacturer, this opens the perspective of integrating Lidar to improve wind turbine control.

Learning objectives
Benefit from a feasibility study on the use of Lidar for wind turbine control under difficult atmospheric conditions.

Anticipate integration issues to provide optimal availability for future Lidar sensors on tomorrow’s turbine.

[1]: Expected Impacts on Cost of Energy through Lidar Based Wind Turbine Control. A Byrne, T McCoy, K Briggs, T Rogers conference. EWEA, 2012.
[2]: Wind turbine control applications of turbine-mounted LIDAR. E A Bossanyi, A Kumar, O Hugues-Salas. Torque From Wind conference, Oldenburg, 2012.
[3]: Lidar-enabled real-time control of wind turbines. C L Bottasso, F P Pietro Pizzinelli, C E D Riboldi. EWEA conference, 2013.
[4]: Field Testing of Feed forward Collective Pitch Control on the CART2 Using a Nacelle-Based Lidar Scanner. D Schlipf, P Fleming, F Haizmann, A Scholbrock, M Hofäß, A Wright, P W Cheng. Torque From Wind conference, Oldenburg, 2012.
[5]: Lidar Assisted Turbine Control ... An Industrial Perspective. A Koerber, C Mehendale. AWEA conference, 2013.
[6]: Leosphere Pulsed Lidar Principles. Contribution to UpWind WP6 on Remote Sensing Devices. J P Cariou, M Boquet. 2010.
[7]: A spinner-integrated wind Lidar for enhanced wind turbine control. T Mikkelsen, N Angelou, K H Hansen, M Sjöholm, M Harris, C Slinger, P Hadley, R Scullion, G Ellis, G Vives. Wind Energy, Vol. 16, 2013, p. 625–643.
[8]: Nacelle lidar for power curve measurement _ Avedøre campaign. R Wagner, S Davoust. DTU Wind Energy E-0016, 2013.