Doppler Radar Analysis of Wind Turbine Wakes During the Early Evening Transition

A universal issue in the modeling of wind plant complex flows is the dearth of observational data to validate model output.  Nacelle-mounted anemometry and meteorological masts currently constitute the industry standard for in-situ observations within the wind plant.  While these instruments provide high temporal information about the embedded flow field, the distribution of these measurement platforms lacks the density required to adequately characterize the spatial complexity of wind plant flows.  Further, the research applicability of nacelle-based measurements is often a source of contention due to potential bias imposed by rotor circulation and the physical structures of the turbine (Allik et al. 2014).  Without a comprehensive understanding of the multi-scale physics and aerodynamic interactions governing wind plant flows, turbine control and plant optimization strategies will continue to yield plant underperformance.  Including operational data from 60 North American wind power facilities encompassing 317 years of operation, DNV (2011) found wind plants were operating at an aggregate 9% underperformance level between 2000 and 2010. 

Recent advancements in remote sensing technologies, however, have afforded researchers the opportunity to better quantify the structure and variability of flows within and surrounding wind plants.  In particular, the use of Light Detection and Ranging (LIDAR) systems, which are frequently mounted atop the nacelle of the wind turbine, has become common practice within the industry to document the local flows surrounding an individual turbine.  Current LIDAR technology is spatially limited due to the inverse relationship between maximum range and along-beam range resolution (Trombe et al. 2013).  Thus, analyses of wind plant complex flows through LIDAR remote sensing remain heavily turbine-centric.  Research-grade Doppler radar systems provide the opportunity to expand this spatial footprint.  Equipped with a pulse-dependent constant range resolution, research-grade Doppler radars enable a paradigm shift in research from turbine- to more plant-centric analyses. 

Utilizing Texas Tech University’s Ka-band Doppler radar systems and invoking dual-Doppler scanning strategies, researchers have demonstrated the ability to map the complex flows and wake structures surrounding an individual turbine (Hirth and Schroeder 2013) as well as within wind plants (Hirth et al. 2015).  Isolating a single utility-scale turbine, researchers have been able to spatially characterize the dynamic nature of a wind turbine wake.  In contrast to the perceived length of an average wind turbine wake (~7 rotor diameters (D)), the wake observed exhibited lengths greater than 20D downstream of the turbine.  Transitioning to a wind plant setting, researchers have been able to document the prevalence of turbine-to-turbine interaction and other plant-specific complex flows, including channels of higher momentum.  Until recently, the collection of coherent velocity fields required the presence of light precipitation.  Recent upgrades to the radar systems, however, have enhanced the ability to collect coherent data during periods of non-precipitation, or ‘clear-air’ modes. 

This increase in data availability has allowed for the examination of plant-scale complex flows under a variety of atmospheric stabilities.  Of particular interest is the behavior of the boundary layer between the fully developed convective conditions of mid-day and the more stable nocturnal boundary layer, which forms in the hours after sunset.  Isolating this short period allows researchers to directly examine how the varying structure and physics of the ABL can modulate plant-relevant flows and in particular, wind turbine wakes.

To examine this question, a single Doppler radar was deployed on 10 October 2016 isolating a utility scale turbine.  Data was collected over a period of 3 hours and 36 minutes, effectively capturing the ABL early evening transition (EET).  Sector scans were performed at a 1.0° elevation tilt, which allowed for the scan to adequately capture the downstream wake profile.  During the scanning period, 2,281 individual scans were collected with an average revisit time of 4.7 seconds.  The high frequency of these revisits enables the characterization of the fine-scale structure and variability of the wind turbine wake.  Often, the quantification of these features is lost due to the temporal averaging commonplace within the wind energy community.  In addition, a 200m meteorological tower is contained within the scanned sector.  The tower is instrumented across 10 different levels providing high-frequency thermal and wind data, which allow for the development of both stability and flow characterization parameters. Coupling these tower with radar-derived measurements, we are able to quantify how the stability of the ABL governs the structure and variability of the wind turbine wake. 

As turbine control strategies trend towards becoming more plant-centric, it is increasingly important to know how the ABL modulates plant-specific complex flows.  To expand upon the work above and explore this question, a dual-Doppler deployment was executed in September 2015 surrounding a wind plant.  Coordinated measurements between two TTU Ka-band radars and subsequent dual-Doppler synthesis allow for the 3-dimensional analysis of plant-scale complex flows surrounding 14 multi-MW turbines.  Through the analysis and characterization of complex flows during the EET, specifically wind turbine wakes, researchers will know when and in which atmospheric conditions coordinated control strategies between turbines may be most appropriate. 

REFERENCES:

Allik, A., J. Uiga, and A. Annuk, 2014:  Deviations between wind speed data measured with nacelle-mounted anemometers on small wind turbines and anemometers mounted on measuring masts.  Agron. Res., 12, 433-444.

DNV Renewables Inc., 2011:  Wind power project underperformance white paper.  Report, 11 pp.  [Available online at https://www.dnvgl.com/energy/publications/index.html]

Hirth, B. D. and J. L. Schroeder, 2013:  Documenting wind speed and power deficits behind a utility-scale wind turbine.  J. Appl. Meteor. Climatol., 52, 39-46.

Hirth, B. D., J. L. Schroeder, W. S. Gunter, and J. G. Guynes, 2015:  Coupling Doppler radar-derived wind maps with operational turbine data to document wind farm complex flows. Wind Energ., 18, 529-540. 

Trombe, PJ, and Coauthors, 2013:  Weather radars – the new eyes for offshore wind farms?  Wind Energ., 17, 1767-1787.