Turbine Wake Flow Research Articles

Aerodynamic performance and flow field characteristics of wind turbines under shear incoming flow conditions

Abstract With the large size of the units, wind farms are evaluated in order to improve the accuracy of the incoming flow measurement and to improve the unit control strategy. In this paper, the turbulent wind field generated by autoregressive linear filtering using a large eddy simulation turbulence model is used to study the wind speed-induced zone under shear incoming flow conditions. At the same time, the wake and aerodynamic performances of the wind turbine under different incoming flow conditions are investigated, and conclusions are obtained. 1) The larger the shear index is, the larger the amplitude is at the first frequency component of the power and the thrust, and the larger the fatigue load is for the wind turbine. 2) For the wind turbine wake flow in the case of smaller wind turbine size, the wake velocity profiles at different shear indices cross and overlap in the transition and far wake regions. At position 0.5 D, the turbulence profiles with different shear indices basically overlap; the larger the shear index is, the more intense the wake flow exchange is and the greater the turbulence intensity in the wake region will be. 3) For the wind speed-induced region, at position −0.1 D, the incoming flow velocity distribution is subject to the combined effect of induction and shear, and the induction effect of the wind turbine wheel is a little larger. The larger the shear index, the greater the turbulence intensity.

Reconstruction of dynamic wind turbine wake flow fields from virtual Lidar measurements via physics-informed neural networks

Accurate characterisation of wind turbine wakes is important for the optimal design and operation of wind farms. However, current techniques for full-scale wind measurements are still limited to point characterisation. To address the research challenge in obtaining field characterisation of real-world wind turbine wakes, this work investigates the reconstruction of the dynamic wake flow fields based on a virtual turbine-mounted Lidar and physics-informed neural networks. Specifically, the wake flow field is reconstructed by fusing the sparse measurements with the two-dimensional Navier-Stokes equations without imposing any models for the unsteady wake. Different from supervised machine learning approaches which need the measured values for the quantities of interest in the first place, the proposed method can achieve the prediction of the wind velocity at new locations where there is no measurement available. The reconstruction performance is evaluated via high-fidelity numerical experiments and it is shown that the dynamic wind turbine wake flow fields are predicted accurately, where the main wake features, including the downwind development and crosswind meandering of the wake, are both captured. This work thus paves the way for investigating full-scale in situ wake flow dynamics in real-world wind energy sites.

An evaluation of different measurement strategies to measure wind turbine near wake flow with small multicopter UAS

Wind turbine wake flow, especially in the near wake, that is up to one rotor diameter D downstream, is subject to interaction between tip vortices and ambient turbulence. These interactions are important to understand wake decay, but most difficult to measure with common instrumentation. Small uncrewed aerial systems (UAS) can help to measure at such locations where no masts can be installed. We contrast two measurement strategies, the hover flight with multiple UAS and cross-section flights with single UAS. We show that both strategies have advantages; the cross-section flights provide a full picture of the width and wind speed deficit across the rotor diameter whereas multi-UAS hover flights can provide more reliable turbulence intensity and turbulent flux measurements at specific locations. With both strategies, tip vortices can be detected and qualified to characterize the state of wake decay at different positions. A fit to the vortex models Lamb-Oseen and Burnham-Hallock allows to estimate circulation and core radius of the vortices. For best characterization of the wake, we recommend to combine the hover and cross-section flight strategies in future.

A reduced order modeling-based machine learning approach for wind turbine wake flow estimation from sparse sensor measurements

A comprehensive understanding of wind turbine wake characteristics is vital, particularly in the context of expanding large offshore wind farms. Existing wake measurement techniques provide only spatially sparse wake measurement data, limiting their utility in precise wind turbine design and control. This paper introduces a data-driven approach that combines proper orthogonal decomposition (POD) with machine learning (ML) techniques, designing a Reduced Order Modeling-based Wake Flow Estimation (ROM-WFE) framework. This framework establishes a nonlinear mapping between sensor measurements and low-dimensional POD coefficients. Two distinct sensor placements, wall-mounted and wake-mounted, are investigated for real measurement scenarios. The results highlight the effectiveness of the proposed wake flow estimation method in reconstructing a complete flow field from exceptionally sparse sensor data, with both wall-mounted and wake-mounted strategies, exhibiting promising results with maximum relative errors of 6.37% and 4.51%, respectively. From the reliability assessments considering various configurations of POD modes and sensor numbers, the ROM-WFE framework demonstrates its capability to estimate wake flow effectively, offering a cost-effective tool for practical applications. Furthermore, the framework maintains accuracy even with high-noise and low-frequency data, demonstrating robustness and generalization. This method significantly contributes to wind turbine wake prediction controller design, promising accurate and robust wake flow field estimation, potentially revolutionizing active wake control and enhancing wind farm operational efficiency.

NUMERICAL INVESTIGATION OF THERMAL STRATIFICATION EFFECTS ON WIND TURBINE WAKE FLOW

Numerical investigation of thermal stratification effects on wind turbine wake flow with and without yawed conditions was conducted in this paper. The three-dimensional wind turbine model from National Renewable Energy Laboratory (NREL) Phase VI experiment was adopted in all the simulations. Model validation was done by using three commonly used turbulence models: realizable <i>k-ε</i> model, shear-stress transport (SST) <i>k-ω</i> model, and large eddy simulation (LES) model. Simulation results were compared with experimental data in the literature. The realizable <i>k-ε</i> model was selected for further simulation as it showed a good balance in achieving accurate results and lowering computational cost. Followed by the validation, wind turbine wake flow with and without yawed effects under different thermal stratification conditions were simulated. Wake flow parameters were discussed and analyzed, followed by suggestions of wind turbine layouts based on different operation conditions to efficiently harness wind energy.

The influence of the lobed ejector on the horizontal axial tidal turbine and flow characteristic analysis

The power coefficient of the horizontal axis tidal turbine (HATT) is limited to 0.593 by the Betz limitation, however, the turbine's efficiency can be improved by incorporating a diffuser or lobe ejector. In this study. In this paper, water tunnel experiments were conducted to obtain the power and thrust coefficients of the original turbine and the turbine with a lobe ejector at different tip speed ratios (TSR). The accuracy of the numerical simulation was then verified by comparing it with the experimental results. The results reveal that the power coefficient of the turbine with a lobe ejector is up to 623% higher than that of the original turbine; however, it also leads to a negative thrust coefficient higher than the original turbine by 94.64%. To determine the optimal position of the lobe ejector, a numerical simulation of the turbine with a diffuser, which consists of a single inner lobe channel, was conducted under the same conditions. The findings demonstrate that the power coefficient of the diffuser is 123% higher than that of the original turbine, and the negative thrust is higher at 26.64%, significantly less than that of the lobe ejector. This indicates that the effect of the lobe ejector largely depends on the ejector effect of the outflow channel. Subsequently, to investigate the ejection effect of the lobe ejector, various aspects, including the vortex structure of the flow field, pressure coefficient of the blade section, velocity vector field of the rotor's center plane, and wake velocity curve, were analyzed. Moreover, the changes in the position of flow separation points along different sections of the blade from the blade root to the blade tip, the variation of the spiral structure of the blade root vortex, the evolution of the velocity vector field in the low-velocity region of the turbine's wake flow field and the alteration of the velocity curve of the wake flow field were further examined.

Numerical investigation of blade roughness impact on the aerodynamic performance and wake behavior of horizontal axis wind turbine

The prolonged operation of wind turbines in harsh offshore environments leads to deterioration and roughness accumulation on the blade surface. This roughness, particularly on the leading edge and other surfaces, can affect the laminar-to-turbulent transition, alter the flow characteristics in the turbine wake and turbulent boundary layer, and become critical for the accurate design and performance analysis of offshore horizontal axis wind turbines (HAWT). This study investigates the effects of blade surface roughness on the aerodynamic performance and wake evolution of the NREL Phase VI wind turbine rotor using the Reynolds-Averaged Navier-Stokes (RANS) technique. First, 2D simulations are validated against experimental data of the S809 airfoil. Then, full-scale 3D simulations of the complete turbine model are conducted with roughness effects to simulate natural conditions. The results show that surface roughness reduces the blade’s aerodynamic performance. The rough surface increases the boundary layer thickness, causing flow separation and turbulence, which decrease the lift generated by the blade and increase its drag, resulting in decreased overall blade performance. At higher wind speeds, surface roughness has a negligible effect on turbine performance due to flow separation at the leading edge. The analysis of surface roughness effects on the turbine wake flow indicates that blade roughness positively correlates with wake recovery, where the wake velocity recovers faster with an increase in roughness height.

ART-LSTANet: An adaptive intelligent method for wind turbine wake analysis

The analysis of wake effects within wind farms is paramount to elevating power generation efficiency, especially when considering the losses incurred by wake effects. In the present investigation, we introduce an innovative neural network-based model – adaptive reduction three-way long short term attention network (ART-LSTANet) – designed to augment the precision of wind turbine wake flow field predictions. Unlike conventional methodologies that often segregate the reduced-order model from the prediction procedure, our proposed model exploits adaptive order reduction to swiftly procure the necessary input for the predictive model, thus curtailing the time expenditure throughout the entire process. The predictive model subsequently incorporates carefully designed feature extraction components tailored to multiple temporal scales, with parameters being updated via a data-driven approach. A comparative analysis with six established intelligent algorithms underscores the superiority of the ART-LSTANet. Comprehensive results indicate that ART-LSTANet delivers superior performance in the reconstruction of the wake flow field, demonstrating a reduction in the mean squared error by up to 9.0% and in the root mean squared error by up to 3.3% compared to traditional methodologies. Numerical errors calculated under a spectrum of additional evaluation metrics consistently yield the lowest values.

Effectiveness of wake control optimization for multiple in-line wind turbines by combinatorial machine learning wake model

In the wind turbine wake control research, the analytical wake model does not consider the influence of direct control parameters on the wind turbine wake flow field, and its prediction accuracy is by no means satisfying. The numerical simulation method can accurately predict the wake effect, but its low computational efficiency renders it unable to be used in the iterative control optimization study. In this paper, a new wake model coupled with the direct control parameters (tip speed ratio and pitch angle) of a single wind turbine is built based on the machine learning algorithm for which the widely applied Artificial Neural Network (ANN) is applied for instance. By combining the ANN wake model with different wake superposition models, the best combinatorial model for quantitative evaluation of multiple wake interferences is achieved, which not only has great accuracy but also ensures high calculation efficiency. On top of the established combinatorial wake model, three in-line wind turbines are investigated for assessing the effectiveness of wake control optimization with respect to the traditional greedy control. The results show that by optimizing the control parameters (tip speed ratio and pitch angle) of wind turbines, the total power of the three wind turbine units is increased by 1.15~3.48% depending on the incoming wind speeds and distances apart between wind turbines. The power variation of each wind turbine shows that the increase in total power is achieved by the decreased power of the first wind turbine and the increased power of the second and third wind turbines. The research verifies the effectiveness of adopting optimized wind turbine wake control on the total wind farm power improvement numerically.

Data‐driven modelling of turbine wake interactions and flow resistance in large wind farms

AbstractTurbine wake and local blockage effects are known to alter wind farm power production in two different ways: (1) by changing the wind speed locally in front of each turbine and (2) by changing the overall flow resistance in the farm and thus the so‐called farm blockage effect. To better predict these effects with low computational costs, we develop data‐driven emulators of the ‘local’ or ‘internal’ turbine thrust coefficient as a function of turbine layout. We train the model using a multi‐fidelity Gaussian process (GP) regression with a combination of low (engineering wake model) and high‐fidelity (large eddy simulations) simulations of farms with different layouts and wind directions. A large set of low‐fidelity data speeds up the learning process and the high‐fidelity data ensures a high accuracy. The trained multi‐fidelity GP model is shown to give more accurate predictions of compared to a standard (single‐fidelity) GP regression applied only to a limited set of high‐fidelity data. We also use the multi‐fidelity GP model of with the two‐scale momentum theory (Nishino & Dunstan 2020, J. Fluid Mech. 894, A2) to demonstrate that the model can be used to give fast and accurate predictions of large wind farm performance under various mesoscale atmospheric conditions. This new approach could be beneficial for improving annual energy production (AEP) calculations and farm optimization in the future.

Power and Flow Analysis of Axial Induction Control in an Array of Model-Scale Wind Turbines

As research on wind energy has progressed, it has broadened from a focus on the wind turbine to include the entire wind farm. In particular, methods to mitigate the negative effects of upstream wakes on downstream turbines have received significant attention. One such mitigation method is axial induction control (AIC) in which upstream turbines are derated to reduce the momentum deficits in their wakes, leaving higher speed flow for downstream turbines. If performed correctly, it is theorized that the power production gains in downstream turbines can compensate for the power sacrificed by derating upstream turbines. Previous work has indicated that the “excess” energy left in the wake of the derated turbine is along the edges of the wake such that a turbine placed directly downstream will see little to no increase in power. To address this hypothesis, we performed a control and treatment experiment with model-scale turbines in a wide flume. Five turbines were arranged in three successive streamwise rows, with the first two rows consisting of two aligned turbines, while three turbines with small transverse spacing were placed in the third row, the central of which was also streamwise-aligned with the upstream two turbines. This arrangement was used to evaluate the difference in power production primarily among the turbines in the third row when the upstream turbines were derated. Particle image velocimetry (PIV) was used to measure the wake in the streamwise-vertical planes along the centerline of the array and along the rotor tips of the centerline turbines between all rows, and high accuracy power measurements were recorded from each turbine. The results show that the total power of the array was decreased while implementing AIC but that individual turbine performance differed from predictions. PIV results show that mean kinetic energy (MKE) is redistributed to the edges of the wakes as has been previously hypothesized. We provide an analysis of the results that connects both the power and flow measurements and that highlights several of the aspects of wind turbine wake flows that make them so complex and challenging to study.

Development of a dynamic wake model accounting for wake advection delays and mesoscale wind transients

With tip height often above 200 meters, currently installed wind turbines are increasingly interacting with the atmospheric boundary layer. Therefore, the research intends to develop a new dynamic wake model that considers weather transient inputs. The Flow Redirection and Induction in Steady State (FLORIS) framework has been proven to be a powerful wake modelling tool even though lacking dynamical effects. An extension of the FLORIS framework is presented in order to include wake advection delays between wind turbines, time-varying and spatially heterogeneous wind conditions. The so-called Observation Points (OPs) are generated at the rotor centre location. Following a Lagrangian approach, the OPs are convected downstream, along the wake of the wind turbine, defining the wake centreline. It is during this process that the dynamics of wind turbine wake and background flow field unsteadiness are included. Finally, using the same approach currently implemented for yaw-misalignment wake deflection, the wake is shifted to match the unsteady wake centreline. The developed model is validated against Large Eddy Simulation for unsteady inflow with variable wind turbine control and wake interactions. The new model is applied to a simulation of the Horns Rev wind farm with a sinusoidally time-varying wind inflow direction to show an hysteresis in the wind farm power extraction curve.

Design of a gust generator and comparison of model wind turbine and porous disc wake flows in a transverse gust

This paper outlines the similarities and differences between the wake characteristics of a porous disc and model wind turbine under transverse gust loading which is generated via a two-vane wind tunnel gust generator. Phase-locked two-dimensional two-component (2D2C) particle image velocimetry (PIV) measurements show that the capability of a porous disc in representing a wind turbine increases under gust loading. Compared to the uniform inflow case, normalized streamwise velocity plots of the porous disc and model turbine are in a better agreement under gusty inflow condition particularly in the near wake region. Comparison of the wake centerlines reveals a similar wavy wake pattern with no significant phase lag suggesting that the gust dynamics is the dominant factor in the determination of the wake dynamics.

Data-driven wind turbine wake modeling via probabilistic machine learning

Wind farm design primarily depends on the variability of the wind turbine wake flows to the atmospheric wind conditions and the interaction between wakes. Physics-based models that capture the wake flow field with high-fidelity are computationally very expensive to perform layout optimization of wind farms, and, thus, data-driven reduced-order models can represent an efficient alternative for simulating wind farms. In this work, we use real-world light detection and ranging (LiDAR) measurements of wind-turbine wakes to construct predictive surrogate models using machine learning. Specifically, we first demonstrate the use of deep autoencoders to find a low-dimensional latent space that gives a computationally tractable approximation of the wake LiDAR measurements. Then, we learn the mapping between the parameter space and the (latent space) wake flow fields using a deep neural network. Additionally, we also demonstrate the use of a probabilistic machine learning technique, namely, Gaussian process modeling, to learn the parameter-space-latent-space mapping in addition to the epistemic and aleatoric uncertainty in the data. Finally, to cope with training large datasets, we demonstrate the use of variational Gaussian process models that provide a tractable alternative to the conventional Gaussian process models for large datasets. Furthermore, we introduce the use of active learning to adaptively build and improve a conventional Gaussian process model predictive capability. Overall, we find that our approach provides accurate approximations of the wind-turbine wake flow field that can be queried at an orders-of-magnitude cheaper cost than those generated with high-fidelity physics-based simulations.

Wind farm wake modeling based on deep convolutional conditional generative adversarial network

Modeling of wind farm wakes is of great importance for the optimal design and operation of wind farms. In this work a surrogate modeling method for parametrized fluid flows is proposed for wind farm wake modeling, based on the state-of-the-art deep learning framework i.e. deep convolutional conditional generative adversarial network. Based on the proposed method and the data generated by high-fidelity large eddy simulations, a novel wind farm wake model is developed. The developed model is first validated against high-fidelity data and the results show that it achieves accurate, efficient, and robust prediction of wind turbine wake flow, at all the streamwise locations including both near wake and far wake, for both streamwise and spanwise velocity components, and at the cases with different inflow wind profiles. Then an extensive parametric study is carried out and the results show that the model generalizes well to unknown flow scenarios. Furthermore, a case study for a wind farm is investigated by the developed model. The prediction results are then compared with high-fidelity simulations, showing that the model can predict the wind farm wake flow (including both the streamwise and spanwise velocity fields) very well.

Lagrangian coherent structures and material transport in unsteady flow of vertical-axis turbine wakes

In this study, the unsteady phenomena of vertical-axis turbine wake flow were analyzed in two rotational speed conditions from a Lagrangian perspective. The wake flow field was obtained using particle image velocimetry. The spectral proper orthogonal decomposition method was used to decompose the wake modes with different frequencies. The Lagrangian coherent structure theory based on the finite-time Lyapunov exponent was used to study the material transport and mixing in the wake evolution process. The results show that the Lagrangian coherent structures can capture the boundary of the wake vortex. At a low rotation speed, a vorticity–flux window is found, through which the fluid is transported to the rotation-induced structures. At a high rotation speed, the increasing material transport between the separation bubbles and the incoming flow eventually leads to shedding of the large-scale vortex structures, accompanied by a decrease in the length of the separation zone. The material transport and mixing revealed in the unsteady flow of turbine wakes have significance in guiding the study of the flow control mechanism.

Doppler Lidar Investigations of Wind Turbine Near-Wakes and LES Modeling with New Porous Disc Approach

Many bottom-mounted offshore wind farms are currently planned for the coastal areas of Japan, in which wind speeds of 6.0–10.0 m/s are extremely common. The impact of such wind speeds is very relevant for the realization of bottom-mounted offshore wind farms. In evaluating the feasibility of these wind farms, therefore, strict evaluation at wind speeds of 6.0–10.0 m/s is important. In the present study, the airflow characteristics of 2 MW-class downwind wind turbine wake flows were first investigated using a vertically profiling remote sensing wind measurement device (lidar). The wind turbines used in this study are installed at the point where the sea is just in front of the wind turbines. A ground-based continuous-wave (CW) conically scanning wind lidar system (“ZephIR ZX300”) was used. Focusing on the wind turbine near-wakes, the detailed behaviors were considered. We found that the influence of the wind turbine wake, that is, the wake loss (wind velocity deficit), is extremely large in the wind speed range of 6.0–10.0 m/s, and that the wake loss was almost constant at such wind speeds (6.0–10.0 m/s). It was additionally shown that these results correspond to the distribution of the thrust coefficient of the wind turbine. We proposed a computational fluid dynamics (CFD) porous disk (PD) wake model as an intermediate method between engineering wake models and CFD wake models. Based on the above observations, the wind speed range for reproducing the behavior of the wind turbine wakes with the CFD PD wake model we developed was set to 6.0–10.0 m/s. Targeting the vertical wind speed distribution in the near-wake region acquired in the “ZephIR ZX300”, we tuned the parameters of the CFD PD wake model (CRC = 2.5). We found that in practice, when evaluating the mean wind velocity deficit due to wind turbine wakes, applying the CFD PD wake model in the wind turbine swept area was very effective. That is, the CFD PD wake model can reproduce the mean average wind speed distribution in the wind turbine swept area.

Large eddy simulations of wind turbine wakes in typical complex topographies

AbstractThere is a need to clarify the coupling characteristics of terrain‐induced wind fields and wind turbine‐induced wake in wind farm micrositing. However, research investigating the effect of hill shape on this interaction is lacking. In addition, during the optimization of the layout of the turbines over topographies, the flow behind the turbines should be predicted in a fast way. After obtaining the wake flow of the turbine mounted on flat terrain and the flow over the topographies, there are mainly two superposition methods to predict the turbine wake flow over the topographies. One is to add the wind deficit following the location with a vertical distance of the hub height (D‐line). The other is to add the wind deficit following the streamline starting from the turbine center (B‐line). It remains unknown to what extent these two superposition methods can be adopted. Therefore, the effects of hill shape, wind turbine size, and turbine location are investigated. When the wind deficit, obtained from modeling the wake flow of the turbine mounted on flat terrain, is superposed following B‐line, the results are overall better than those when following D‐line.

Development of a streamline wake model for wind farm performance predictions

In the present paper a new streamline model for wake development based on a streamline topology is applied and compared with different approaches for modeling the wind turbine wakes. The contribution of the present work is the comparison of the streamline-and straight-wake models as well as the different wake models, e.g., Jensen and Gaussian. The models have been applied to two wind farm cases. The results from the first case are compared against SCADA measurements and computational fluid dynamics simulations (CFD). The CFD simulations are performed using Reynolds-Averaged Navier Stokes (RANS) together with the actuator disc (AD) approach. The mean percent difference in power using the different wake models ranged between 19-20%. Mean percent difference in power for the AD-RANS was approximately 20%. For the second wind farm case only wake models are compared and approximately ±1% difference exists between them. The present work shows that the streamline topology of the wind turbine wake flow as well as the wake models have an effect on the performance prediction of wind farms.

Predictions of Wake and Central Mixing Region of Double Horizontal Axis Tidal Turbine

Predicting the velocity distribution of double horizontal axis tidal turbines (DHATTs) is significant for the effective development of tidal streams. This current research gives an account on double turbine wake theory and flow structure of DHATT connected to single support by using the joint axial momentum theory and computational fluid dynamics (CFD) method. Characteristics of single turbine wake were previously studied with two theoretical equations predicting the initial upstream velocity closer to the turbine, and it's lateral distributions along the downstream of the turbine. This current works agreed with the previous wake equations, which was used for predicting the velocity region along the downstream of the turbines. Flow field separating the two turbines is complicated in nature due to the indirect disturbance of turbines and no report was found on this central region. The Central region in the downstream flow is initially suppressed due to the blockage effects with a high velocity close to the free stream. Lateral expansion of two turbine wakes penetrated the central region with velocity reduction and followed by the flow recovery further downstream. This work provides more understandings of the wake and its central mixing region for double turbines with a proposed theoretical model.