A technique to retrieve wind speed at constant height above ground level in complex terrain with a 4-beam nacelle mounted Lidar

Introduction

Turbine performance testing has become a key challenge of wind farm development either dealing with the cost for offshore mast measurement or the measurement of wind speed in moderately complex to complex terrain. Although nacelle mounted Lidar was observed to be a real alternative to mast measurement, wind speed measured is located at the laser beams height inducing a constant height of measurement above sea level (CHASL) which might not fit the requirement for power curve computing or site calibration in terrain with important variation in altitude. Indeed, Lidar permits to achieve quicker power curve measurement campaign by being always aligned with the turbine thus by using wider wind sectors than those available when measuring with a mast. Nevertheless current wind speed reconstruction algorithm provide the wind speed at the same height regardless to the wind sectors. For power curve measurement or site calibration, it is necessary to measure the wind speed always at the same height for each wind sectors. This paper presents a wind reconstruction method adapted to a 4-beam nacelle mounted Lidar which ensure constant height above ground level measurement in any wind sectors.

Approach

4-beam nacelle mounted Lidar (Wind Iris from Leosphere) has two set of beams : a couple of beams measuring downward and a couple measuring upward with a aperture angle of 10°. Assuming local homogeneity, the wind speed can be retrieved for each couple of beams, ie. at two heights for each measurement distance (10). Using these two measurements the wind profile can be retrieved assuming it follows a linear or logarithmic profile and therefore the wind speed can retrieved at any height. Having a wind speed measurement at a constant height above ground level at 10 distances from the wind turbine will permits to further assess the turbine performances in complex terrain and also possibly computing the site calibration.

Main body of abstract

The study will show the computational method of wind speed, wind direction and turbulence intensity using a 4-beam nacelle mounted Lidar. The assumption on the wind profile for simple and complex terrain is discussed. The expected deviations linked to the log-profile assumptions are computed from mast data equipped with 4 anemometer at different height. The wind profile is computed for each possible pairs of anemometer and the extrapolation of wind speed is compared for the two others anemometer. Using data from a campaign performed in 2015 with 4-beam nacelle mounted Lidar, the reconstruction is challenged against a mast showing non log-profile measured by mast and the accuracy of Lidar measurement in these cases. After validating Lidar wind reconstruction, basic wind speed reconstruction (CHASL) and wind speed reconstruction with constant height above ground level CHAGL wind reconstruction are compared and impact on performance estimation is given through wind sector power curve computing. The method requires the terrain elevation data to obtain the CHAGL wind speed in whatever distances ahead the turbine and in whatever direction the turbine is facing. Finally, a method to compute site calibration using CHAGL wind reconstruction is presented and tested.

Conclusion

The results from this study will give confidence in the Lidar measurement in moderately complex to complex terrain giving the associated uncertainty to each step of wind speed reconstruction. Application for power performance testing is presented and the impact of using CHAGL in moderately complex terrain on uncertainty will permit to assess the opportunity of using nacelle mounted Lidar for performance testing in moderately complex to complex terrain. The ability of Lidar to possibly perform the site calibration will be discussed.


Learning objectives
1. The wind reconstruction algorithm for 4-beam Lidar and its associated uncertainties.
2. The effect of different type of reconstruction CHAS and CHAL on power testing uncertainty.