S.J. Watson
info
Please Note
<p>This page displays the records of the person named above and is not linked to a unique person identifier. This record may need to be merged to a profile.</p>
58 records found
1
Wind turbine wake dynamics subjected to atmospheric gravity waves
A measurement-driven large-eddy simulation study
Atmospheric gravity waves (AGWs) are large-scale wave-like flow structures commonly generated when atmospheric flows are vertically perturbed by topographical features or meteorological phenomena. These transient phenomena can significantly affect wind turbine outputs and loads; however, their influence on wake dynamics remains poorly understood, posing challenges for accurate wind farm modeling. In this study, we perform large-eddy simulation of wind turbines operating under an atmospheric condition reconstructed by assimilating lidar measurements of AGWs. Our results show that (i) low-frequency wake meandering becomes more pronounced owing to large-scale AGW flow structures and intensified smaller-scale turbulent structures. This enhanced meandering, combined with stronger turbulent mixing, accelerates mean wake recovery. (ii) The turbulence kinetic energy (TKE) spectrum in the wake region exhibits a peak Strouhal number of approximately 0.3, although the inflow spectrum peaks at significantly lower frequencies. This observation indicates that, under AGW conditions, wake turbulence generation follows a convective instability mechanism. Notably, faster wake recovery reduces wake shear, leading to lower amplification of TKE. Power analysis for two turbines arranged in a streamwise column further highlights the dominant role of convective instabilities. Large-amplitude, low-frequency power fluctuations observed at the most upstream turbine are significantly attenuated for downstream turbines as low-frequency velocity fluctuations are suppressed in the far-wake region. These findings add further insights into wake meandering and turbulence generation, offering guidance for modeling wind turbine and farm flows under non-stationary atmospheric conditions.
...
Atmospheric gravity waves (AGWs) are large-scale wave-like flow structures commonly generated when atmospheric flows are vertically perturbed by topographical features or meteorological phenomena. These transient phenomena can significantly affect wind turbine outputs and loads; however, their influence on wake dynamics remains poorly understood, posing challenges for accurate wind farm modeling. In this study, we perform large-eddy simulation of wind turbines operating under an atmospheric condition reconstructed by assimilating lidar measurements of AGWs. Our results show that (i) low-frequency wake meandering becomes more pronounced owing to large-scale AGW flow structures and intensified smaller-scale turbulent structures. This enhanced meandering, combined with stronger turbulent mixing, accelerates mean wake recovery. (ii) The turbulence kinetic energy (TKE) spectrum in the wake region exhibits a peak Strouhal number of approximately 0.3, although the inflow spectrum peaks at significantly lower frequencies. This observation indicates that, under AGW conditions, wake turbulence generation follows a convective instability mechanism. Notably, faster wake recovery reduces wake shear, leading to lower amplification of TKE. Power analysis for two turbines arranged in a streamwise column further highlights the dominant role of convective instabilities. Large-amplitude, low-frequency power fluctuations observed at the most upstream turbine are significantly attenuated for downstream turbines as low-frequency velocity fluctuations are suppressed in the far-wake region. These findings add further insights into wake meandering and turbulence generation, offering guidance for modeling wind turbine and farm flows under non-stationary atmospheric conditions.
Impact of atmospheric turbulence on performance and loads of wind turbines
Knowledge gaps and research challenges
Review
(2026)
-
Branko Kosović, Sukanta Basu, Jacob Berg, Larry K. Berg, Sue E. Haupt, Xiaoli G. Larsén, Joachim Peinke, Richard J.A.M. Stevens, Paul Veers, Simon Watson
Wind energy harvesting from the atmosphere takes place in the atmospheric boundary layer. The boundary layer shear and buoyancy create three-dimensional turbulent eddies spanning a range of scales that form a continuous forward cascade of kinetic energy to the smallest scales of motion where energy is dissipated. Large-scale atmospheric circulations modulate the boundary layer turbulence, characterized by coherence and intermittency. As wind turbines grow in size and the integrated control of both turbines and wind farms spans greater distances, the relationship between the scales of atmospheric turbulence and the design and operation of wind energy facilities has entered new territory. The boundary layer turbulence impacts both wind turbine power production and turbine loads. Optimizing wind turbine and wind farm performance requires an understanding of how turbulence affects both wind turbine efficiency and reliability. While the characteristics of atmospheric boundary layer turbulence have been observed and studied in detail over the last few decades, there are still significant gaps in our understanding of the impact of turbulence on wind power resources and wind farm operations. This paper outlines the current state of turbulence research relevant to wind energy applications and points to gaps in our knowledge that need to be addressed to effectively utilize wind resources.
...
Wind energy harvesting from the atmosphere takes place in the atmospheric boundary layer. The boundary layer shear and buoyancy create three-dimensional turbulent eddies spanning a range of scales that form a continuous forward cascade of kinetic energy to the smallest scales of motion where energy is dissipated. Large-scale atmospheric circulations modulate the boundary layer turbulence, characterized by coherence and intermittency. As wind turbines grow in size and the integrated control of both turbines and wind farms spans greater distances, the relationship between the scales of atmospheric turbulence and the design and operation of wind energy facilities has entered new territory. The boundary layer turbulence impacts both wind turbine power production and turbine loads. Optimizing wind turbine and wind farm performance requires an understanding of how turbulence affects both wind turbine efficiency and reliability. While the characteristics of atmospheric boundary layer turbulence have been observed and studied in detail over the last few decades, there are still significant gaps in our understanding of the impact of turbulence on wind power resources and wind farm operations. This paper outlines the current state of turbulence research relevant to wind energy applications and points to gaps in our knowledge that need to be addressed to effectively utilize wind resources.
Power ramp events represent an important challenge to grid stability, motivating the enforcement of ramp limits. For wind farm operation, decisions to respect these limits are based on imperfect forecast data, where errors can lead to deviations from the prescribed limit. In this study, we propose two different methods to mitigate the impact of forecast uncertainties on the operation of ramp-constrained wind farms: the use of a pessimistic forecast, where ramp events are worsened artificially, and the use of a storage system. The two methods are assessed by solving an online dispatch optimization problem for one year of operation. Forecast data are generated from numerical weather prediction models of the ECMWF. The dependence of power production on wind speed and direction changes is captured by an engineering wake deficit model. Results for 20 different offshore sites in Europe show that using a pessimistic forecast reduces the number of ramp events exceeding the limit by one third but increases curtailment by 0.2 percentage points on average. Instead, adding a storage system to the wind farm is more effective at reducing curtailment, proportionally to its size. The impact of forecast errors is mitigated most effectively by combining the two methods.
...
Power ramp events represent an important challenge to grid stability, motivating the enforcement of ramp limits. For wind farm operation, decisions to respect these limits are based on imperfect forecast data, where errors can lead to deviations from the prescribed limit. In this study, we propose two different methods to mitigate the impact of forecast uncertainties on the operation of ramp-constrained wind farms: the use of a pessimistic forecast, where ramp events are worsened artificially, and the use of a storage system. The two methods are assessed by solving an online dispatch optimization problem for one year of operation. Forecast data are generated from numerical weather prediction models of the ECMWF. The dependence of power production on wind speed and direction changes is captured by an engineering wake deficit model. Results for 20 different offshore sites in Europe show that using a pessimistic forecast reduces the number of ramp events exceeding the limit by one third but increases curtailment by 0.2 percentage points on average. Instead, adding a storage system to the wind farm is more effective at reducing curtailment, proportionally to its size. The impact of forecast errors is mitigated most effectively by combining the two methods.
This large-eddy simulation (LES) study examines how wind-farm-induced atmospheric gravity waves (AGWs) and wind farm performance depend on non-dimensional atmospheric parameters and simulation configuration. A hypothetical aligned wind farm of actuator disks is simulated under neutral surface conditions, with a stable capping inversion and a mildly stable free atmosphere, to assess the effects of stratification beyond the atmospheric boundary layer (ABL) on ABL flow. Simulation set-ups fully resolving AGWs are validated to minimize spurious wave generation and reflection from the domain boundaries. The validated set-up is then used to analyze AGW types and characteristics, as well as stratification impacts under conventionally neutral boundary layer (CNBL) conditions. These conditions are governed by four non-dimensional parameters: the Froude numbers of the free atmosphere and capping inversion (Fr, Fri), and the aspect ratios of the ABL and wind farm (H̃i Sh). Simulation configurations that fully resolve AGWs-capturing at least one wavelength both horizontally and vertically-yield the most realistic stratification effects on ABL flow, whereas partial or unresolved configurations produce non-physical, channel-like behavior. A coherent description of the AGW phenomena is provided, highlighting the central role of capping inversion displacement in linking ABL fluctuations with AGWs. Trapped waves are confined within the capping inversion, while interfacial and internal waves aloft are identified as the AGW types most relevant to wind farm performance. The wavy inversion, analogous to an interfacial wave, forms converging and diverging zones that drive power fluctuations across the farm. The interfacial wavelength, measured over the wind farm, corresponds to one diverging, one converging, and one mildly diverging zone. As the interfacial wavelength decreases with Fri, multiple convergence–divergence zones develop under sub-critical conditions (Fri<1.0), while for super-critical conditions (Fri>1.0), the wavelength approaches the farm length. Wave amplitude increases with decreasing H̃i (i.e., shallower capping inversions). Wind farm performance is most sensitive to H̃i: shallow boundary layers increase blockage and reduce efficiency, while deeper layers enhance efficiency. Increasing Fr and Fri mitigates blockage, and increasing Sh mainly improves wake recovery. Although local power fluctuations arise from AGWs, overall wind farm efficiency remains nearly constant with Fr and Fri, improving primarily with larger H̃i and Sh.
...
This large-eddy simulation (LES) study examines how wind-farm-induced atmospheric gravity waves (AGWs) and wind farm performance depend on non-dimensional atmospheric parameters and simulation configuration. A hypothetical aligned wind farm of actuator disks is simulated under neutral surface conditions, with a stable capping inversion and a mildly stable free atmosphere, to assess the effects of stratification beyond the atmospheric boundary layer (ABL) on ABL flow. Simulation set-ups fully resolving AGWs are validated to minimize spurious wave generation and reflection from the domain boundaries. The validated set-up is then used to analyze AGW types and characteristics, as well as stratification impacts under conventionally neutral boundary layer (CNBL) conditions. These conditions are governed by four non-dimensional parameters: the Froude numbers of the free atmosphere and capping inversion (Fr, Fri), and the aspect ratios of the ABL and wind farm (H̃i Sh). Simulation configurations that fully resolve AGWs-capturing at least one wavelength both horizontally and vertically-yield the most realistic stratification effects on ABL flow, whereas partial or unresolved configurations produce non-physical, channel-like behavior. A coherent description of the AGW phenomena is provided, highlighting the central role of capping inversion displacement in linking ABL fluctuations with AGWs. Trapped waves are confined within the capping inversion, while interfacial and internal waves aloft are identified as the AGW types most relevant to wind farm performance. The wavy inversion, analogous to an interfacial wave, forms converging and diverging zones that drive power fluctuations across the farm. The interfacial wavelength, measured over the wind farm, corresponds to one diverging, one converging, and one mildly diverging zone. As the interfacial wavelength decreases with Fri, multiple convergence–divergence zones develop under sub-critical conditions (Fri<1.0), while for super-critical conditions (Fri>1.0), the wavelength approaches the farm length. Wave amplitude increases with decreasing H̃i (i.e., shallower capping inversions). Wind farm performance is most sensitive to H̃i: shallow boundary layers increase blockage and reduce efficiency, while deeper layers enhance efficiency. Increasing Fr and Fri mitigates blockage, and increasing Sh mainly improves wake recovery. Although local power fluctuations arise from AGWs, overall wind farm efficiency remains nearly constant with Fr and Fri, improving primarily with larger H̃i and Sh.
This work experimentally investigated the feasibility and complementarity of aeroacoustic and infrared thermography (IRT) techniques for detecting damage in rotating wind turbine blades under controlled wind tunnel conditions. Two representative types of damage were considered: trailing edge cracks and internal shear web delamination, created in the scaled blades manufactured in-house. Experiments were conducted in the open jet facility at Delft University of Technology. Acoustic measurements using a two-dimensional microphone array revealed that trailing edge cracks induce distinct tonal noise modifications, which depend on the effective trailing edge thickness and are captured through spectral analysis and acoustic beamforming. The crack-induced tonal noise peaks at a trailing-edge-thickness-based Strouhal number, (Formula presented), between 0.15 and 0.25. IRT, by contrast, are highly sensitive to internal structural features; delaminated regions exhibited localized temperature variations due to changes in thermal properties. Principal component thermography was applied to further enhance the visualization of the internal shear webs and internal delamination. The results demonstrate that the use of aeroacoustic and IRT methods provides a complementary strategy for detecting both edge and internal damage in wind turbine blades.
...
This work experimentally investigated the feasibility and complementarity of aeroacoustic and infrared thermography (IRT) techniques for detecting damage in rotating wind turbine blades under controlled wind tunnel conditions. Two representative types of damage were considered: trailing edge cracks and internal shear web delamination, created in the scaled blades manufactured in-house. Experiments were conducted in the open jet facility at Delft University of Technology. Acoustic measurements using a two-dimensional microphone array revealed that trailing edge cracks induce distinct tonal noise modifications, which depend on the effective trailing edge thickness and are captured through spectral analysis and acoustic beamforming. The crack-induced tonal noise peaks at a trailing-edge-thickness-based Strouhal number, (Formula presented), between 0.15 and 0.25. IRT, by contrast, are highly sensitive to internal structural features; delaminated regions exhibited localized temperature variations due to changes in thermal properties. Principal component thermography was applied to further enhance the visualization of the internal shear webs and internal delamination. The results demonstrate that the use of aeroacoustic and IRT methods provides a complementary strategy for detecting both edge and internal damage in wind turbine blades.
Conventional Deep Learning (DL) methods for bearing health indicator (HI) adopt supervised approaches, requiring expert knowledge of the component degradation trend. Since bearings experience various failure modes, assuming a particular degradation trend for HI is suboptimal. Unsupervised DL methods are scarce in this domain. They generally maximise the HI monotonicity built in the middle layer of an Autoencoder (AE) trained to reconstruct the run-to-failure signals. The backpropagation (BP) training algorithm is unable to perform this maximisation since the monotonicity of HI subsections corresponding to input sample batches does not guarantee the monotonicity of the whole HI. Therefore, existing methods achieve this by searching AE hyperparameters so that its BP training to minimise the reconstruction error also leads to a highly monotonic HI in its middle layer. This is done using expensive search algorithms where the AE is trained numerous times using various hyperparameter settings, rendering them impractical for large datasets. To address this limitation, a small Convolutional Autoencoder (CAE) architecture and a hybrid training algorithm combining Particle Swarm Optimisation and BP are proposed in this work to enable simultaneous maximisation of the HI monotonicity and minimisation of the reconstruction error. As a result, the HI is built by training the CAE only once. The results from three case studies demonstrate this method’s lower computational burden compared to other unsupervised DL methods. Furthermore, the CAE-based HIs outperform the indicators built by equivalent and significantly larger models trained with a BP-based supervised approach, leading to 85% lower remaining useful life prediction errors.
...
Conventional Deep Learning (DL) methods for bearing health indicator (HI) adopt supervised approaches, requiring expert knowledge of the component degradation trend. Since bearings experience various failure modes, assuming a particular degradation trend for HI is suboptimal. Unsupervised DL methods are scarce in this domain. They generally maximise the HI monotonicity built in the middle layer of an Autoencoder (AE) trained to reconstruct the run-to-failure signals. The backpropagation (BP) training algorithm is unable to perform this maximisation since the monotonicity of HI subsections corresponding to input sample batches does not guarantee the monotonicity of the whole HI. Therefore, existing methods achieve this by searching AE hyperparameters so that its BP training to minimise the reconstruction error also leads to a highly monotonic HI in its middle layer. This is done using expensive search algorithms where the AE is trained numerous times using various hyperparameter settings, rendering them impractical for large datasets. To address this limitation, a small Convolutional Autoencoder (CAE) architecture and a hybrid training algorithm combining Particle Swarm Optimisation and BP are proposed in this work to enable simultaneous maximisation of the HI monotonicity and minimisation of the reconstruction error. As a result, the HI is built by training the CAE only once. The results from three case studies demonstrate this method’s lower computational burden compared to other unsupervised DL methods. Furthermore, the CAE-based HIs outperform the indicators built by equivalent and significantly larger models trained with a BP-based supervised approach, leading to 85% lower remaining useful life prediction errors.
An experimental investigation is carried out to characterize the physical mechanisms by which a trailing-edge crack, idealized as a rectangular cavity to represent delamination damage, affects boundary layer development, coherent vortex shedding, and far-field noise of a National Advisory Committee for Aeronautics 0018 airfoil. Both clean and turbulent inflow conditions are considered to isolate the role of inflow disturbance in modifying these mechanisms. The primary objective is to gain insight into how a geometrical discontinuity at the trailing edge alters the coupled aerodynamic and aeroacoustic behavior. Far-field acoustic measurements and near-wake velocity field data are obtained in the anechoic wind tunnel at Delft University of Technology. Acoustic data from a phased microphone array (from prior work) are combined with new velocity field measurements using particle image velocimetry. The results reveal that increasing crack size leads to enhanced near-wall velocity gradients and stronger coherent vortex shedding, resulting in higher tonal noise levels, particularly at higher frequencies. Normalized tonal frequencies agree with the empirical prediction model of Brooks, Pope, and Marcolini for blunt trailing-edge noise, affirming the relevance of this model even in the presence of geometric imperfections. Under turbulent inflow, the coherent structure scale diminishes slightly, and the tonal frequency increases in the trailing-edge noise spectrum, indicating that inflow turbulence modifies the vortex shedding dynamics and should be accounted for in predictive models. This study is a first step toward understanding and modeling trailing-edge noise in the presence of structural damage, under varying flow conditions.
...
An experimental investigation is carried out to characterize the physical mechanisms by which a trailing-edge crack, idealized as a rectangular cavity to represent delamination damage, affects boundary layer development, coherent vortex shedding, and far-field noise of a National Advisory Committee for Aeronautics 0018 airfoil. Both clean and turbulent inflow conditions are considered to isolate the role of inflow disturbance in modifying these mechanisms. The primary objective is to gain insight into how a geometrical discontinuity at the trailing edge alters the coupled aerodynamic and aeroacoustic behavior. Far-field acoustic measurements and near-wake velocity field data are obtained in the anechoic wind tunnel at Delft University of Technology. Acoustic data from a phased microphone array (from prior work) are combined with new velocity field measurements using particle image velocimetry. The results reveal that increasing crack size leads to enhanced near-wall velocity gradients and stronger coherent vortex shedding, resulting in higher tonal noise levels, particularly at higher frequencies. Normalized tonal frequencies agree with the empirical prediction model of Brooks, Pope, and Marcolini for blunt trailing-edge noise, affirming the relevance of this model even in the presence of geometric imperfections. Under turbulent inflow, the coherent structure scale diminishes slightly, and the tonal frequency increases in the trailing-edge noise spectrum, indicating that inflow turbulence modifies the vortex shedding dynamics and should be accounted for in predictive models. This study is a first step toward understanding and modeling trailing-edge noise in the presence of structural damage, under varying flow conditions.
Airborne wind energy systems (AWESs) leverage the generally less variable and higher wind speeds at increased altitudes by utilizing kites, with significantly reduced material costs compared to conventional wind turbines. Energy is commonly harnessed by flying crosswind trajectories, which allow the kite to achieve speeds significantly higher than the ambient wind speed. However, the airborne nature of these systems demands active control and makes them highly sensitive to changes in wind conditions, making accurate wind measurements essential for steering the kite along its optimal trajectory. This paper presents an advanced sensor fusion technique based on an iterated extended Kalman filter (EKF) for state and wind estimation for AWESs. By integrating position, velocity, tether force, and reeling speed, this method provides accurate estimations of system dynamics, including kite orientation and tether shape. The estimates of the wind speed and direction are compared to lidar measurements, showing good agreement across various atmospheric conditions, with 10 min averaged root mean square error (RMSE) values below 1 m s−1 and 5°, respectively. The results demonstrate that this approach can effectively capture the transient dynamics of atmospheric wind using sensors typically already present in AWESs, making it suitable for supervisory control strategies and ultimately enhancing energy efficiency and system reliability across diverse atmospheric conditions.
...
Airborne wind energy systems (AWESs) leverage the generally less variable and higher wind speeds at increased altitudes by utilizing kites, with significantly reduced material costs compared to conventional wind turbines. Energy is commonly harnessed by flying crosswind trajectories, which allow the kite to achieve speeds significantly higher than the ambient wind speed. However, the airborne nature of these systems demands active control and makes them highly sensitive to changes in wind conditions, making accurate wind measurements essential for steering the kite along its optimal trajectory. This paper presents an advanced sensor fusion technique based on an iterated extended Kalman filter (EKF) for state and wind estimation for AWESs. By integrating position, velocity, tether force, and reeling speed, this method provides accurate estimations of system dynamics, including kite orientation and tether shape. The estimates of the wind speed and direction are compared to lidar measurements, showing good agreement across various atmospheric conditions, with 10 min averaged root mean square error (RMSE) values below 1 m s−1 and 5°, respectively. The results demonstrate that this approach can effectively capture the transient dynamics of atmospheric wind using sensors typically already present in AWESs, making it suitable for supervisory control strategies and ultimately enhancing energy efficiency and system reliability across diverse atmospheric conditions.
This paper studies the sensitivity of drivetrain condition monitoring system (CMS) signals to blade damage, exploring how these signals, particularly vibration, can serve as a potential tool for detection and tracking damage progression. This is achieved using a decoupled simulation approach, combining an aeroelastic solver with a drivetrain model. First, aeroelastic simulations are performed in OpenFAST, where the low-speed shaft (LSS) forces, moments, and tower top position vector are extracted and transferred to the drivetrain model. The drivetrain is modelled using the multi-body simulation environment SIMPACK. Blade damage is introduced in OpenFAST by reducing stiffness in the flap-wise or edgewise direction. The reference DTU-10MW onshore wind turbine is used as a test case. First, the impact of blade damage on LSS shear forces is analysed. Then the drivetrain response is assessed using virtual velocity sensors placed at the main bearing, rear bearing and gearbox housing. Results indicate that damage occurring in the blade mid-span region shows higher sensitivity compared to tip and root locations. A positive correlation is observed between LSS shear force and bearings side-side velocity, with higher forces leading to increased vibration. Additionally, the trend suggests that higher stiffness reduction results in higher velocity, indicating damage progression.
...
This paper studies the sensitivity of drivetrain condition monitoring system (CMS) signals to blade damage, exploring how these signals, particularly vibration, can serve as a potential tool for detection and tracking damage progression. This is achieved using a decoupled simulation approach, combining an aeroelastic solver with a drivetrain model. First, aeroelastic simulations are performed in OpenFAST, where the low-speed shaft (LSS) forces, moments, and tower top position vector are extracted and transferred to the drivetrain model. The drivetrain is modelled using the multi-body simulation environment SIMPACK. Blade damage is introduced in OpenFAST by reducing stiffness in the flap-wise or edgewise direction. The reference DTU-10MW onshore wind turbine is used as a test case. First, the impact of blade damage on LSS shear forces is analysed. Then the drivetrain response is assessed using virtual velocity sensors placed at the main bearing, rear bearing and gearbox housing. Results indicate that damage occurring in the blade mid-span region shows higher sensitivity compared to tip and root locations. A positive correlation is observed between LSS shear force and bearings side-side velocity, with higher forces leading to increased vibration. Additionally, the trend suggests that higher stiffness reduction results in higher velocity, indicating damage progression.
As the penetration of renewable energy increases in the generation mix, the problem of power dispatchability becomes more critical. The co-location of storage systems with wind energy is a promising solution to shift power delivery from periods of high wind resource availability to periods of high electricity demand. Producing baseload power from wind farms all or most of the time is an example of such dispatchability. In this work, we present an optimization-based dispatch strategy to produce baseload power. At every time step, an optimization problem is solved to decide the storage operation, maximize revenues on the electricity market and reach a given reliability target. In order to reduce the impact of forecast uncertainties on the reliability of the power delivery, a robust formulation of the dispatch optimization is used, based on a pessimistic version of the forecast. The proposed method is evaluated for 18 offshore sites with a 100 MW wind farm and storage system, for one year of operation. By using a robust dispatch strategy, the reliability increases by up to 0.9 points, with a minor impact on revenues (+2% on average), compared to the reference dispatch strategy using a regular forecast. Our study demonstrates the feasibility of providing a reliable baseload power from wind energy in the presence of forecast uncertainty.
...
As the penetration of renewable energy increases in the generation mix, the problem of power dispatchability becomes more critical. The co-location of storage systems with wind energy is a promising solution to shift power delivery from periods of high wind resource availability to periods of high electricity demand. Producing baseload power from wind farms all or most of the time is an example of such dispatchability. In this work, we present an optimization-based dispatch strategy to produce baseload power. At every time step, an optimization problem is solved to decide the storage operation, maximize revenues on the electricity market and reach a given reliability target. In order to reduce the impact of forecast uncertainties on the reliability of the power delivery, a robust formulation of the dispatch optimization is used, based on a pessimistic version of the forecast. The proposed method is evaluated for 18 offshore sites with a 100 MW wind farm and storage system, for one year of operation. By using a robust dispatch strategy, the reliability increases by up to 0.9 points, with a minor impact on revenues (+2% on average), compared to the reference dispatch strategy using a regular forecast. Our study demonstrates the feasibility of providing a reliable baseload power from wind energy in the presence of forecast uncertainty.
Wind turbine supervisory control and data acquisition (SCADA) datasets available for research usually contain a limited number of failure events. This limitation hinders the successful application of deep learning (DL) methods for fault detection and prognosis, as they require large datasets for robust training and generalisation. This work proposes a method using conditional generative adversarial networks (cGANs) to generate synthetic SCADA time series that replicate wind turbine behaviour under controllable operational, environmental, and degradation conditions. Given a set of SCADA time series representing these conditions, the cGAN generates temperature and pressure time series simulating gearbox operation. Results show that augmenting the training set of an artificial neural network (ANN) fault detection model with synthetic time series reduces false positives in the detected gearbox faults by 84 % on average, enabling the model to blindly detect a fault in a test wind turbine without prior knowledge of the event. Furthermore, training a convolutional autoencoder-based unsupervised health indicator (HI) model with both real and synthetic SCADA time series leads to an HI that more accurately captures the expected degradation trend. Using this HI, the gearbox's remaining useful life (RUL) can be predicted within the defined error bounds from around 4.5 months before the detection of the fault, while the HI obtained without the synthetic data fails to produce reliable RUL estimations.
...
Wind turbine supervisory control and data acquisition (SCADA) datasets available for research usually contain a limited number of failure events. This limitation hinders the successful application of deep learning (DL) methods for fault detection and prognosis, as they require large datasets for robust training and generalisation. This work proposes a method using conditional generative adversarial networks (cGANs) to generate synthetic SCADA time series that replicate wind turbine behaviour under controllable operational, environmental, and degradation conditions. Given a set of SCADA time series representing these conditions, the cGAN generates temperature and pressure time series simulating gearbox operation. Results show that augmenting the training set of an artificial neural network (ANN) fault detection model with synthetic time series reduces false positives in the detected gearbox faults by 84 % on average, enabling the model to blindly detect a fault in a test wind turbine without prior knowledge of the event. Furthermore, training a convolutional autoencoder-based unsupervised health indicator (HI) model with both real and synthetic SCADA time series leads to an HI that more accurately captures the expected degradation trend. Using this HI, the gearbox's remaining useful life (RUL) can be predicted within the defined error bounds from around 4.5 months before the detection of the fault, while the HI obtained without the synthetic data fails to produce reliable RUL estimations.
Wind farms, particularly offshore clusters, are becoming larger than ever before. Besides influencing the surface wind flow and the inflow for downstream wind farms, large wind farms can trigger atmospheric gravity waves in the inversion layer and the free atmosphere aloft. Wind farm-induced gravity waves can cause adverse pressure gradients upstream of the wind farm, that contribute to the global blockage effect, and can induce favorable pressure gradients above and downstream of the wind farm that enhance wake recovery. Numerical modeling is a powerful means of studying these wind farm-induced atmospheric gravity waves, but it comes with the challenge of handling spurious reflections of these waves from domain boundaries. Typically, approaches which employ radiation boundary conditions and forcing zones are used to avoid these reflections. However, the simulation setup of these approaches heavily relies on ad-hoc processes. For instance, the widely used Rayleigh damping method requires ad-hoc tuning to produce a setup which may be only produce satisfactory results for a particular case. To provide more systematic guidance on setting up realistic simulations of atmospheric gravity waves, we conduct an LES study of flow over a 2D hill and through a wind farm canopy that explores the optimum domain size and damping layer setup depending on the fundamental parameters which determine the flow characteristics.
In this work, we only consider linearly stratified conditions (i.e., no inversion layer), thereby focusing on internal gravity waves in the free atmosphere and their reflections from the domain boundaries. This type of flow is governed by a single Froude number, which dictates most of the internal wave properties, such as wavelength, amplitude, and direction. This in turn will dictate the optimum domain size and Rayleigh damping layer setup. We find the effective horizontal and vertical wavelengths, (the representative wavelengths of the entire wave spectrum), to be the appropriate length scales to size the domain and damping layer thickness, and the optimal Rayleigh damping coefficient scales with the Brunt–Väisälä frequency.
Considering Froude numbers seen in wind farm applications, we propose recommendations to limit the reflections to less than 10 % of the total upwards propagating wave energy. Typically, damping is done at the top boundary, but given the non-periodic lateral boundary conditions of practical wind farm simulation domains, we find that damping the inflow-outflow boundaries is of equal importance to the top boundary. The Brunt–Väisälä frequency-normalized damping coefficient should be between 1 and 10. The damping layer thickness should be at least one effective vertical wavelength; damping layers exceeding 1.5 times the vertical wavelength are found to be unnecessary. The domain length and height should accommodate at least one effective horizontal and vertical wavelength, respectively. Moreover, Rayleigh damping does not damp the waves completely, and the non-damped energy might accumulate over the simulation time. ...
In this work, we only consider linearly stratified conditions (i.e., no inversion layer), thereby focusing on internal gravity waves in the free atmosphere and their reflections from the domain boundaries. This type of flow is governed by a single Froude number, which dictates most of the internal wave properties, such as wavelength, amplitude, and direction. This in turn will dictate the optimum domain size and Rayleigh damping layer setup. We find the effective horizontal and vertical wavelengths, (the representative wavelengths of the entire wave spectrum), to be the appropriate length scales to size the domain and damping layer thickness, and the optimal Rayleigh damping coefficient scales with the Brunt–Väisälä frequency.
Considering Froude numbers seen in wind farm applications, we propose recommendations to limit the reflections to less than 10 % of the total upwards propagating wave energy. Typically, damping is done at the top boundary, but given the non-periodic lateral boundary conditions of practical wind farm simulation domains, we find that damping the inflow-outflow boundaries is of equal importance to the top boundary. The Brunt–Väisälä frequency-normalized damping coefficient should be between 1 and 10. The damping layer thickness should be at least one effective vertical wavelength; damping layers exceeding 1.5 times the vertical wavelength are found to be unnecessary. The domain length and height should accommodate at least one effective horizontal and vertical wavelength, respectively. Moreover, Rayleigh damping does not damp the waves completely, and the non-damped energy might accumulate over the simulation time. ...
Wind farms, particularly offshore clusters, are becoming larger than ever before. Besides influencing the surface wind flow and the inflow for downstream wind farms, large wind farms can trigger atmospheric gravity waves in the inversion layer and the free atmosphere aloft. Wind farm-induced gravity waves can cause adverse pressure gradients upstream of the wind farm, that contribute to the global blockage effect, and can induce favorable pressure gradients above and downstream of the wind farm that enhance wake recovery. Numerical modeling is a powerful means of studying these wind farm-induced atmospheric gravity waves, but it comes with the challenge of handling spurious reflections of these waves from domain boundaries. Typically, approaches which employ radiation boundary conditions and forcing zones are used to avoid these reflections. However, the simulation setup of these approaches heavily relies on ad-hoc processes. For instance, the widely used Rayleigh damping method requires ad-hoc tuning to produce a setup which may be only produce satisfactory results for a particular case. To provide more systematic guidance on setting up realistic simulations of atmospheric gravity waves, we conduct an LES study of flow over a 2D hill and through a wind farm canopy that explores the optimum domain size and damping layer setup depending on the fundamental parameters which determine the flow characteristics.
In this work, we only consider linearly stratified conditions (i.e., no inversion layer), thereby focusing on internal gravity waves in the free atmosphere and their reflections from the domain boundaries. This type of flow is governed by a single Froude number, which dictates most of the internal wave properties, such as wavelength, amplitude, and direction. This in turn will dictate the optimum domain size and Rayleigh damping layer setup. We find the effective horizontal and vertical wavelengths, (the representative wavelengths of the entire wave spectrum), to be the appropriate length scales to size the domain and damping layer thickness, and the optimal Rayleigh damping coefficient scales with the Brunt–Väisälä frequency.
Considering Froude numbers seen in wind farm applications, we propose recommendations to limit the reflections to less than 10 % of the total upwards propagating wave energy. Typically, damping is done at the top boundary, but given the non-periodic lateral boundary conditions of practical wind farm simulation domains, we find that damping the inflow-outflow boundaries is of equal importance to the top boundary. The Brunt–Väisälä frequency-normalized damping coefficient should be between 1 and 10. The damping layer thickness should be at least one effective vertical wavelength; damping layers exceeding 1.5 times the vertical wavelength are found to be unnecessary. The domain length and height should accommodate at least one effective horizontal and vertical wavelength, respectively. Moreover, Rayleigh damping does not damp the waves completely, and the non-damped energy might accumulate over the simulation time.
In this work, we only consider linearly stratified conditions (i.e., no inversion layer), thereby focusing on internal gravity waves in the free atmosphere and their reflections from the domain boundaries. This type of flow is governed by a single Froude number, which dictates most of the internal wave properties, such as wavelength, amplitude, and direction. This in turn will dictate the optimum domain size and Rayleigh damping layer setup. We find the effective horizontal and vertical wavelengths, (the representative wavelengths of the entire wave spectrum), to be the appropriate length scales to size the domain and damping layer thickness, and the optimal Rayleigh damping coefficient scales with the Brunt–Väisälä frequency.
Considering Froude numbers seen in wind farm applications, we propose recommendations to limit the reflections to less than 10 % of the total upwards propagating wave energy. Typically, damping is done at the top boundary, but given the non-periodic lateral boundary conditions of practical wind farm simulation domains, we find that damping the inflow-outflow boundaries is of equal importance to the top boundary. The Brunt–Väisälä frequency-normalized damping coefficient should be between 1 and 10. The damping layer thickness should be at least one effective vertical wavelength; damping layers exceeding 1.5 times the vertical wavelength are found to be unnecessary. The domain length and height should accommodate at least one effective horizontal and vertical wavelength, respectively. Moreover, Rayleigh damping does not damp the waves completely, and the non-damped energy might accumulate over the simulation time.
Wind farm-induced atmospheric gravity waves have been the subject of recent research as they can impact wind farm performance. Pressure variations associated with gravity waves can contribute to the global blockage effect and wind farm wake recovery. Therefore, accurate numerical simulation of flow fields, where wind-farm-induced gravity waves may be produced, is important. Three main considerations in such simulations are the overall domain size, the use of Rayleigh damping near domain boundaries to dampen gravity waves, and advection damping at the inlet to prevent spurious oscillations. Often these considerations are treated ad hoc rather than systematically. This work aims to test and extend the systematic modelling of internal gravity waves proposed in a preliminary investigation to modelling of both internal and trapped gravity waves. The preliminary study identifies the length scales to set the domain and damping layer sizes and the time scale to configure the Rayleigh damping coefficient but under linearly stratified conditions. Large eddy simulations of flow through a wind farm canopy are performed under conventionally neutral boundary layer (CNBL) conditions to test the validity of proposed setups for CNBL conditions. Background atmospheric parameters, such as Froude number (Fr), inversion height (Hi), and inversion layer Froude number (Fri) control most of the atmospheric gravity wave characteristics. We validated for CBNL conditions that the effective wavelengths of the internal gravity waves are the correct length scale to configure the domain size and damping layer thickness. Likewise, the optimum damping coefficient to dampen the internal gravity waves relates to the free atmosphere's buoyancy frequency or buoyant perturbations' time scale. We infer that the damping coefficient in the inversion layer may relate to the inversion buoyancy frequency to effectively dampen the trapped gravity waves. Moreover, the advection damping length is linked to the horizontal wavelength of the trapped gravity waves in the inversion layer to prevent spurious waves at the inlet by retaining wave energy accumulation.
...
Wind farm-induced atmospheric gravity waves have been the subject of recent research as they can impact wind farm performance. Pressure variations associated with gravity waves can contribute to the global blockage effect and wind farm wake recovery. Therefore, accurate numerical simulation of flow fields, where wind-farm-induced gravity waves may be produced, is important. Three main considerations in such simulations are the overall domain size, the use of Rayleigh damping near domain boundaries to dampen gravity waves, and advection damping at the inlet to prevent spurious oscillations. Often these considerations are treated ad hoc rather than systematically. This work aims to test and extend the systematic modelling of internal gravity waves proposed in a preliminary investigation to modelling of both internal and trapped gravity waves. The preliminary study identifies the length scales to set the domain and damping layer sizes and the time scale to configure the Rayleigh damping coefficient but under linearly stratified conditions. Large eddy simulations of flow through a wind farm canopy are performed under conventionally neutral boundary layer (CNBL) conditions to test the validity of proposed setups for CNBL conditions. Background atmospheric parameters, such as Froude number (Fr), inversion height (Hi), and inversion layer Froude number (Fri) control most of the atmospheric gravity wave characteristics. We validated for CBNL conditions that the effective wavelengths of the internal gravity waves are the correct length scale to configure the domain size and damping layer thickness. Likewise, the optimum damping coefficient to dampen the internal gravity waves relates to the free atmosphere's buoyancy frequency or buoyant perturbations' time scale. We infer that the damping coefficient in the inversion layer may relate to the inversion buoyancy frequency to effectively dampen the trapped gravity waves. Moreover, the advection damping length is linked to the horizontal wavelength of the trapped gravity waves in the inversion layer to prevent spurious waves at the inlet by retaining wave energy accumulation.
Vertical-axis wind turbines (VAWTs) are considered promising solutions for urban wind energy generation due to their design, low maintenance costs, and reduced noise and visual impact compared to horizontal-axis wind turbines (HAWTs). However, deploying these turbines close to densely populated urban areas often triggers considerable local opposition to wind energy projects. Among the primary concerns raised by communities is the issue of noise emissions. Noise annoyance should be considered in the design and decision-making process to foster the social acceptance of VAWTs in urban environments. At the same time, maximising the operational efficiency of VAWTs in terms of power generation and actuation effort is equally important. This paper balances noise and aero-servo-elastic performance by formulating and solving a multi-objective optimisation problem from a controller calibration perspective. Psychoacoustic annoyance is taken as a novel indicator for the noise objective by providing a more reliable estimate of the human perception of wind turbine noise than conventional sound metrics. The computation of the psychoacoustic annoyance metric is made feasible by integrating it with an accurate and computationally efficient low-fidelity noise prediction model. For optimisation, an advanced partial-load control scheme – often used in industrial turbines – is considered, with the Kω2 controller as a baseline for comparison. Optimal solutions balancing the defined objectives are identified using a multi-criteria decision-making method (MCDM) and are subsequently assessed using a frequency-domain controller analysis framework and mid-fidelity time-domain aero-servo-elastic simulations. The MCDM results indicate the potential application of this controller in small-scale urban VAWTs to attain power gains of up to 39 % on one side and to trade off a reduction in actuation effort of up to 25 % at the cost of only a 2 % power decrease and a 6 % increase in psychoacoustic annoyance on the other side compared to the baseline. These findings confirm the flexible structure of the optimally calibrated wind speed estimator and tip-speed ratio (WSE–TSR) tracking controller, effectively balancing aero-servo-elastic performance with noise emissions and marking the first instance of integrating residential concerns into the decision-making process.
...
Vertical-axis wind turbines (VAWTs) are considered promising solutions for urban wind energy generation due to their design, low maintenance costs, and reduced noise and visual impact compared to horizontal-axis wind turbines (HAWTs). However, deploying these turbines close to densely populated urban areas often triggers considerable local opposition to wind energy projects. Among the primary concerns raised by communities is the issue of noise emissions. Noise annoyance should be considered in the design and decision-making process to foster the social acceptance of VAWTs in urban environments. At the same time, maximising the operational efficiency of VAWTs in terms of power generation and actuation effort is equally important. This paper balances noise and aero-servo-elastic performance by formulating and solving a multi-objective optimisation problem from a controller calibration perspective. Psychoacoustic annoyance is taken as a novel indicator for the noise objective by providing a more reliable estimate of the human perception of wind turbine noise than conventional sound metrics. The computation of the psychoacoustic annoyance metric is made feasible by integrating it with an accurate and computationally efficient low-fidelity noise prediction model. For optimisation, an advanced partial-load control scheme – often used in industrial turbines – is considered, with the Kω2 controller as a baseline for comparison. Optimal solutions balancing the defined objectives are identified using a multi-criteria decision-making method (MCDM) and are subsequently assessed using a frequency-domain controller analysis framework and mid-fidelity time-domain aero-servo-elastic simulations. The MCDM results indicate the potential application of this controller in small-scale urban VAWTs to attain power gains of up to 39 % on one side and to trade off a reduction in actuation effort of up to 25 % at the cost of only a 2 % power decrease and a 6 % increase in psychoacoustic annoyance on the other side compared to the baseline. These findings confirm the flexible structure of the optimally calibrated wind speed estimator and tip-speed ratio (WSE–TSR) tracking controller, effectively balancing aero-servo-elastic performance with noise emissions and marking the first instance of integrating residential concerns into the decision-making process.
Journal article
(2024)
-
Ali Eftekhari Milani, Donatella Zappalá, Francesco Castellani, Simon Watson
State-of-the-art Deep Learning (DL) methods based on Supervisory Control and Data Acquisition (SCADA) system data for the detection and prognosis of wind turbine faults require large amounts of failure data for successful training and generalisation, which are generally not available. This limitation prevents benefiting from the superior performance of these methods, especially in SCADA-based failure prognosis. Data augmentation approaches have been proposed in the literature for generating failure data instances within a SCADA sequence to reduce the imbalance between healthy and faulty state data points, which is relevant to fault detection tasks. However, the successful implementation of DL-based failure prognosis methods requires the availability of multiple run-to-failure SCADA sequences. This paper proposes a data-driven method for generating synthetic run-to-failure SCADA sequences with custom operational and environmental conditions and progression of degradation. An Artificial Neural Network (ANN) is trained with signals that represent these factors to reconstruct the SCADA signals. Then, it is used to generate synthetic SCADA datasets based on data available from a wind turbine that experienced a gearbox failure. Synthetic data sets generated are evaluated on the basis of the similarity of their signal distributions, the temporal dynamics within each signal, and the temporal dynamics among different SCADA signals with those in similar field datasets. The results show that the generated synthetic datasets are consistent with their field counterparts, with a comparatively lower diversity in their dynamic behaviour in time.
...
State-of-the-art Deep Learning (DL) methods based on Supervisory Control and Data Acquisition (SCADA) system data for the detection and prognosis of wind turbine faults require large amounts of failure data for successful training and generalisation, which are generally not available. This limitation prevents benefiting from the superior performance of these methods, especially in SCADA-based failure prognosis. Data augmentation approaches have been proposed in the literature for generating failure data instances within a SCADA sequence to reduce the imbalance between healthy and faulty state data points, which is relevant to fault detection tasks. However, the successful implementation of DL-based failure prognosis methods requires the availability of multiple run-to-failure SCADA sequences. This paper proposes a data-driven method for generating synthetic run-to-failure SCADA sequences with custom operational and environmental conditions and progression of degradation. An Artificial Neural Network (ANN) is trained with signals that represent these factors to reconstruct the SCADA signals. Then, it is used to generate synthetic SCADA datasets based on data available from a wind turbine that experienced a gearbox failure. Synthetic data sets generated are evaluated on the basis of the similarity of their signal distributions, the temporal dynamics within each signal, and the temporal dynamics among different SCADA signals with those in similar field datasets. The results show that the generated synthetic datasets are consistent with their field counterparts, with a comparatively lower diversity in their dynamic behaviour in time.
We perform large-eddy simulation of wind turbine wake flow subjected to atmospheric gravity waves. The mesoscale forcing from atmospheric gravity waves is coupled to microscale turbine simulation using an indirect profile assimilation method. Our analysis will center on examining the effects of atmospheric gravity waves on wake meandering and wake-added turbulence, aiming to enhance understanding of the wake dynamics in the presence of atmospheric gravity waves. This study holds promise for guiding the development of wind farm models under realistic atmospheric conditions.
...
We perform large-eddy simulation of wind turbine wake flow subjected to atmospheric gravity waves. The mesoscale forcing from atmospheric gravity waves is coupled to microscale turbine simulation using an indirect profile assimilation method. Our analysis will center on examining the effects of atmospheric gravity waves on wake meandering and wake-added turbulence, aiming to enhance understanding of the wake dynamics in the presence of atmospheric gravity waves. This study holds promise for guiding the development of wind farm models under realistic atmospheric conditions.
For scenarios of high penetration of renewable energy, it becomes increasingly relevant to improve the dispatchability of supply for wind and solar power plants. Baseload power plants, required to produce a minimum power production at all times, are discussed in this context. The baseload constraint can be satisfied with renewable sources when combined with a storage system but at a high cost. This work studies the design drivers of such a storage system when consisting of short and long-term storage. The capacities of the short-term and long-term storage components are calculated as part of a linear optimization problem with the objective of minimizing the cost of baseload, using a metric based on a net present value formulation. Our analysis, based on 10 locations in Northern Europe, highlights a high sensitivity of optimal storage sizing to storage cost assumptions. In addition, the cost of baseload is found to be correlated to the share of renewable power produced above baseload, but not to the correlation between price and wind power, suggesting arbitrage plays a minor role in the business case.
...
For scenarios of high penetration of renewable energy, it becomes increasingly relevant to improve the dispatchability of supply for wind and solar power plants. Baseload power plants, required to produce a minimum power production at all times, are discussed in this context. The baseload constraint can be satisfied with renewable sources when combined with a storage system but at a high cost. This work studies the design drivers of such a storage system when consisting of short and long-term storage. The capacities of the short-term and long-term storage components are calculated as part of a linear optimization problem with the objective of minimizing the cost of baseload, using a metric based on a net present value formulation. Our analysis, based on 10 locations in Northern Europe, highlights a high sensitivity of optimal storage sizing to storage cost assumptions. In addition, the cost of baseload is found to be correlated to the share of renewable power produced above baseload, but not to the correlation between price and wind power, suggesting arbitrage plays a minor role in the business case.
Wind turbine blades carry the risk of impact damage during transportation, installation, and operation. Such impacts can cause levels of damage that can propagate throughout the structure compromising performance and safety. In this study, the effect of impact damage on fatigue damage propagation in test specimens representative of a spar cap-shear web adhesively-bonded connection of a wind turbine blade was investigated. In addition, the effectiveness of using acoustic emissions to detect early impact-induced fatigue damage was studied. Three impact tests with increasing levels of energy were investigated. The results showed that for an impact test with an average energy of 16.32 J, the fatigue damage accumulation process was not influenced by the size and location of the impact damage. But for impact tests with an average energy of 23.68 J and 32.13 J, greater crack density and accelerated de-lamination and de-bonding of the adhesive from the laminate could be seen in the impact zone. Acoustic emission was shown to identify the position of the damage zone for the higher energy impact tests. It was also effective in showing the progressive accumulation of fatigue damage in this zone during the fatigue test.
...
Wind turbine blades carry the risk of impact damage during transportation, installation, and operation. Such impacts can cause levels of damage that can propagate throughout the structure compromising performance and safety. In this study, the effect of impact damage on fatigue damage propagation in test specimens representative of a spar cap-shear web adhesively-bonded connection of a wind turbine blade was investigated. In addition, the effectiveness of using acoustic emissions to detect early impact-induced fatigue damage was studied. Three impact tests with increasing levels of energy were investigated. The results showed that for an impact test with an average energy of 16.32 J, the fatigue damage accumulation process was not influenced by the size and location of the impact damage. But for impact tests with an average energy of 23.68 J and 32.13 J, greater crack density and accelerated de-lamination and de-bonding of the adhesive from the laminate could be seen in the impact zone. Acoustic emission was shown to identify the position of the damage zone for the higher energy impact tests. It was also effective in showing the progressive accumulation of fatigue damage in this zone during the fatigue test.
Journal article
(2024)
-
Majid Bastankhah, M. Becker, Matthew Churchfield, Caroline Draxl, Jay Prakash Goit, M.A. Khan, Luis A Martínez-Tossas, J.W. van Wingerden, S.J. Watson, More Authors...
In this paper, the feasibility of using far-field acoustic measurements as a non-contact monitoring technique for wind turbine blade leading edge erosion is assessed. For this purpose, a DU96 W180 airfoil with several eroded leading edge configurations of different severities is experimentally investigated. The eroded leading edges are designed with pits, gouges and coating delamination scaled from a real eroded blade. To assess the feasibility of the technique in quasi-realistic configurations, experiments are carried out under clean and turbulent inflow conditions. Acoustic measurements are performed with a phased microphone array. In the absence of inflow turbulence, because of the low Reynolds number at which the experiments are carried out, the case with minor erosion severity shows similar far-field noise spectra as the clean leading-edge cases, i.e., the presence of tonal peaks caused by laminar boundary layer instability noise through a self-sustained feedback loop but with higher tonal amplitudes. Increasing the damage level (considered as moderate erosion), the spectra of the noise scattered from the suction side show that the tonal peaks shift to higher frequencies and have lower amplitudes, thus suggesting that the damage alters the flow features responsible for the acoustic feedback loop; whereas, the spectra from the pressure side show a broadband noise distribution. For heavy erosion, the far-field noise spectra show broadband features from both airfoil sides, thus suggesting that the damage has fully forced the transition to turbulent flow; in which case, an increase in the low-frequency content is observed. Conversely, in the presence of turbulent inflow, when comparing the noise scattered at the trailing edge, no difference is found. However, leading edge impingement noise decreases at medium–high frequency compared with the baseline case at a chord-length-based Strouhal number St_C~10. The experimental results also suggest that the delamination feature is the one which is the most easily detectable and the approach is valid for a wide range of angles of attack and inflow velocity.
...
In this paper, the feasibility of using far-field acoustic measurements as a non-contact monitoring technique for wind turbine blade leading edge erosion is assessed. For this purpose, a DU96 W180 airfoil with several eroded leading edge configurations of different severities is experimentally investigated. The eroded leading edges are designed with pits, gouges and coating delamination scaled from a real eroded blade. To assess the feasibility of the technique in quasi-realistic configurations, experiments are carried out under clean and turbulent inflow conditions. Acoustic measurements are performed with a phased microphone array. In the absence of inflow turbulence, because of the low Reynolds number at which the experiments are carried out, the case with minor erosion severity shows similar far-field noise spectra as the clean leading-edge cases, i.e., the presence of tonal peaks caused by laminar boundary layer instability noise through a self-sustained feedback loop but with higher tonal amplitudes. Increasing the damage level (considered as moderate erosion), the spectra of the noise scattered from the suction side show that the tonal peaks shift to higher frequencies and have lower amplitudes, thus suggesting that the damage alters the flow features responsible for the acoustic feedback loop; whereas, the spectra from the pressure side show a broadband noise distribution. For heavy erosion, the far-field noise spectra show broadband features from both airfoil sides, thus suggesting that the damage has fully forced the transition to turbulent flow; in which case, an increase in the low-frequency content is observed. Conversely, in the presence of turbulent inflow, when comparing the noise scattered at the trailing edge, no difference is found. However, leading edge impingement noise decreases at medium–high frequency compared with the baseline case at a chord-length-based Strouhal number St_C~10. The experimental results also suggest that the delamination feature is the one which is the most easily detectable and the approach is valid for a wide range of angles of attack and inflow velocity.