Snowflakes reveal the real near-wake behavior of a utility-scale wind turbine

Present day wind turbines rise over 100m thereby occupying a substantial portion of the atmospheric boundary layer. As such, the turbulent flows around these gigantic structures impose unprecedented challenges for the research community to implement the knowledge from laboratory research at the utility scale. Wake loss, i.e. loss occurring due to turbine wakes, regions of low energy and high turbulence within wind farms, contributes to sub-optimal (power loss of 10%–20%) performance of turbines. Furthermore, wake-generated turbulence is a major source of fatigue loading, which drives premature component failure, results in expensive over-design criteria and limits the turbine and wind farm innovation potential. These wake characteristics and accompanying losses remain poorly understood.

Wind turbine wake is divided into a near wake and a far wake. The former is dominated by turbine-generated coherent structures, such as blade tip and root vortices, trailing vortex sheets, hub vortices and tower vortices and is strongly dependent on rotor properties and turbine operations. The interaction of these structures in the near wake affects the evolution and turbulent flow characteristics of the far wake. Hence near wake is most crucial for the overall wake development and stability. Also, a better understanding of the dependence of near-wake flow characteristics on turbine operation and incoming flow can yield a more accurate prediction of wake growth and provide guidance for developing ‘smart’ control to reduce wake loss and structural impact for future wind farms.

In a recent research work published in Journal of Fluid Mechanics, University of Minnesota scientists Dr. Teja Dasari, Dr. Yue Wu, Dr. Yun Liu, and Professor Jiarong Hong further developed on the novel snowflake based visualization and super-large-scale particle image velocimetry (SLPIV) and studied the near wake. Specifically, the researchers used data collected during multiple snowstorms from 2014 to 2016 around a 2.5MW turbine at the EOLOS field station to examine the turbulent velocity field and coherent structures.

Field campaigns were conducted at EOLOS Wind Energy Research Field Station which has a 3-bladed horizontal axis wind turbine and a meteorological tower, both well instrumented to characterize the atmospheric and turbine operating conditions. The pioneering high-resolution field-scale measurement recorded the average wake deficit to be 0.29 validating the values obtained from the wake models proposed in several prior studies. The near-wake turbulence was also quantified over the entire span of the rotor, a region about the size of a football field, and compared against high-resolution sonic measurements. Interestingly, the instantaneous velocity fields indicated the presence of intermittent wake transitions to a contraction state which were in clear contrast with the expansion states typically associated with wind turbine wakes.

The surprising phenomena occurred for a substantial 25% of the time and the underlying physical mechanisms were revealed. This wake behavior, minimally reported until now, would have a significant impact on the overall wake mixing and growth indicating a need for further exploration and incorporation into wake models. The article further established crucial correlations between the tip vortex behavioral patterns representative of the wake stability and the turbine operation/response characteristics. The analysis established a clear statistical correspondence between turbine parameters and tip vortex behaviors under different turbine operation conditions, which was further substantiated by examining samples of time series of the turbine parameters and tip vortex patterns.

In summary, Professor Jiarong Hong and his research team presented the first systematic experimental investigation of the velocity field and the coherent vortex structures in the near wake of a utility-scale wind turbine at unprecedented spatial and temporal resolution using the super-large-scale particle image velocimetry (SLPIV) and flow visualization with natural snow. These correlations can be used to develop simple wake prediction models based on the readily available SCADA (Supervisory Control and Data Acquisition) parameters and be incorporated into turbine control algorithms to improve efficiencies.

Remarkably, the University of Minnesota study not only offers benchmark grade near-wake datasets for comparison with the-state-of-the-art numerical simulation, laboratory and field measurements, but also sheds light on understanding wake characteristics and the downstream development of the wake, turbine performance and regulation, as well as developing novel turbine or wind farm control strategies.