Dual-Doppler Radar Meets Big Data: A New Vision for Wake Analytics

Introduction

As the wind industry continues to mature, much focus is shifting towards wind farm level control. Many wind farms experience production losses from reduced downstream turbine power production due to lower wind flow after upstream turbine wind extraction, often referred to as the wake effect. Indeed, a common mantra in the O&M world is “One more Megawatt,” portraying the promise of big data applied to wind turbine and wind plant control. While the pursuit of a global production optimum has been the focus of much analytics research and modeling platforms, industrial validation often results in challenges. Specifically, one such challenge includes the integrity of the data collected, where compounded uncertainties often make consideration of new technology or operational changes impossible to assess in terms of their feasibility/economic return to owners/operators.

Approach

For wake analytics, the most pressing variables to understand are the true wind speed and wind direction at each turbine, including how such parameters vary in all spatial dimensions around it. While remote sensing technologies such as LiDAR (Light Detection and Ranging) and SODAR (Sonic Detection and Ranging) can be used from turbine-to-turbine to make calibration adjustments to facilitate proper analytics, the big picture of interaction of wind flows between turbines is often missing. Envision, with its partners, deployed the use of novel Dual-Doppler Radar to measure wind speed and wind direction across all spatial dimensions for multiple turbine rows at a time, allowing for such site-specific wind flow interactions to be measured in addition to turbine-level calibration for those parameters – all at once.

Main body of abstract

Envision’s use of such breakthrough radar technology allowed for new analytic insights not otherwise possible. Precise wake channeling effects could be measured and replicated such that downstream turbine production was at times higher than freestream inflow turbine production! SCADA data could be precisely calibrated to allow for detection of wind speed and direction gradients between subsequent turbine rows, facilitating improved predictive turbine fatigue analytics and power production levels without the continued use of the radar technology.

Conclusion

In conclusion, mainstream remote sensing solutions and physics-based modeling are often not sufficient to correct for calibration errors at each turbine or assess complex wind flows that hamper the promise of big data analytics to lower the levelized cost of energy (LCoE). Envision’s use of Dual-Doppler Radar technology to reduce the uncertainty inherent in SCADA data allows for novel wake analytics to come to economical fruition in pursuit of a global production optimum.


Learning objectives
1. Measurement uncertainties from turbine-to-turbine inhibit the promise of big data analytics to lower the levelized cost of energy at the wind plant level.
2. Site-specific measurement of wind direction and wind speed across all spatial dimensions within a wind farm can reduce measurement uncertainty and allow for wake analytics tuning to complex wind flows.
3. Brief deployment of Dual-Doppler Radar can resurrect the promise of big data wind farm control analytics to provide an economic return without an ongoing capital expenditure.