LiDAR-based turbine performance verification

Key benefits of a LiDAR-based performance verification:

• increase energy production
• avoid premature wear due to excessive stress and loads
• prevent critical component failures and extend turbine lifetime
• assess faster the power performance

A 4-beam LiDAR is temporarily mounted on top of the nacelle, together with a number of calibrated instruments, and a data collection and communication unit in the nacelle. Every second, the LiDAR measures the horizontal wind speed and directionat hub height in front of the turbine at 10 simulatenous measurement ranges, between 50m to 400m.
Compared to met mast-mounted cup anemometers, sufficient data to evaluate the wind turbine power performance can be collected much faster by the nacelle-based LiDAR.

The average relative wind direction and wind speed (at hub height) are computed every 10 minutes from several measurement ranges in front of the turbine. These measurements are validated or discarded based on standard or more advanced criteria, such as cut-in & rated wind speed, low data quality, etc.

Results:

• low wind speeds produce larger variability in the yaw error, as the turbine is still trying to face the wind, but this fluctuation becomes smaller at high wind speeds, as seen in Fig. 1
• the average yaw error is influenced by the wind speed intensity, causing an unusual yawing behavior (see Tab. 1 or Fig. 1)
• large difference between the yaw error computed with and without the split on wind speed ranges

Conclusion:

• LiDAR-based YM campaign can be reduced to less than 7 days, if wind conditions are favorable
• potential production gain of ≈ 0.3% (based on cos² relationship)
• YM for wind speeds > 12 m/s has an impact on the turbine loads

TI results are based on wind reconstructions calculated at a distance closest to 2.5 times the diameter of the rotor (i.e., 240m). The blocked sectors (blue) in Fig. 3 are calculated acc. to the requirements from IEC 61400-12-1 2017, Annex A. All plots are based only on the data during the campaign.

Results:

• the relative wind direction plotted against the 360° nacelle direction (Fig. 5) shows that almost all wind sectors are disturbed by nearby turbines (Fig. 3)
• there is higher TI in the N, S, SW sectors, see Fig. 2
• even with less wind from S, SW sectors (see Fig. 4), there is indication of disturbed wind sector

 
Conclusion:

• increased TI produces a lower wind speed which leads to a decreased energy production
• sectors affected by wakes have to be excluded from the analysis

The power curve was determined using the power output measured by Ventus instruments and LiDAR wind speed measurements (blue) at 240m (distance closest to 2.5 times the rotor diameter), and compared to the power curve based on SCADA wind speeds and power output (orange). Also, Fig. 6 shows the warranted power curve, as provided by the manufacturer

Results:

• the total downtime (turbine shows a code) of ≈3 days ⇒ production loss of 4.7%
• warranted vs. measured prod. during campaign shows a ≈2% underperformance
• wind speed transfer function (Fig. 7) seems to overestimate the upcoming wind speed, and gives the impression of a SCADA PC underestimation (with orange in Fig. 6)

Conclusion:

• the observed large PC underestimation seen at SCADA PC is not verified by LiDAR PC, but the latter suggests the wind transfer function as the root cause
• further investigation of the downtime period is also recommended