We present an experimental method for aerodynamic characterization of flexible membrane kites by in situ measurement of the relative flow, while performing complex flight maneuvers. We find that the aerodynamics of this type of wing depend not only on the angle of attack, but also on the level of aerodynamic loading and the aeroelastic deformation. We recommend using the relative power setting of the kite as a secondary influencing parameter.
This work summarizes the results of the intelligence-sharing initiative of the Power Curve Working Group. Participants in this share exercise applied a handful of selected power curve modeling correction methods on their power performance test data, and they submitted the results for the coauthors to analyze. In this paper, we describe the share exercise, explain the analysis methodologies, and perform statistical tests to evaluate the correction methods in various inflow conditions.
This research was conducted with the help of computer models to give argumentation on how the reliability of wind turbine rotor blade structures can be increased using subcomponent testing (SCT) as a supplement to full-scale blade testing (FST). It was found that the use of SCT can significantly reduce the testing time compared to FST while replicating more realistic loading conditions for an outboard blade segment as it occurs in the field.
Robust and accurate dynamic stall modeling remains one of the most difficult tasks in wind turbine load calculations despite its long research effort in the past. The present paper describes a new
second-order dynamic stall model for wind turbine airfoils. The new model is robust and improves the prediction for the aerodynamic forces and their higher-harmonic effects due to vortex shedding but also provides improved predictions for pitching moment and drag.
Airborne wind energy systems aim to operate at altitudes above conventional wind turbines where reliable high-resolution wind data are scarce. Wind measurements and computational simulations both have advantages and disadvantages when assessing the wind resource at such heights. This article investigates whether assimilating measurements into the model generates a more accurate wind data set up to 1100 m. These wind data sets are used to estimate optimal AWES operating altitudes and power.
Several methods have been proposed in the past for extracting the blade performance of wind turbines from simulations. In this work, we present a new method that allows obtaining those data easily not only from simulation results but also from flow measurements. We apply the method to both experiments and simulations of a well-known wind turbine model. The results provide insight into the wind turbine aerodynamics and open up new possibilities for the validation of simulation models.
The interaction between wind turbines in a wind farm through their wakes is a widely studied research area. Until recently, research was focused on finding constant turbine inputs that optimize the performance of the wind farm. However, recent studies have shown that time-varying, dynamic inputs might be more beneficial. In this paper, the validity of this approach is further investigated by implementing it in scaled wind tunnel experiments and assessing load effects, showing promising results.
Before constructing wind farms we need to know how much energy they will produce. This requires knowledge of long-term wind conditions from either measurements or models. At the US East Coast there are few wind measurements and little experience with offshore wind farms. Therefore, we created a satellite-based high-resolution wind resource map to quantify spatial variations in the wind conditions over potential sites for wind farms and found larger variation than modelling suggested.
This paper identifies the most sensitive parameters for the load response of a 5 MW wind turbine. Two sets of parameters are examined: one set relating to the wind excitation characteristics and a second related to the physical properties of the wind turbine. The two sensitivity analyses are done separately, and the top most-sensitive parameters are identified for different load outputs throughout the structure. The findings will guide future validation campaigns and measurement needs.
Revised manuscript under review for WES(discussion: final response, 8 comments)
Floating offshore wind technology has high potential, but still faces challenges for gaining economic competitiveness to allow commercial market uptake. Hence, design optimization plays a key role, however, the final optimum floater obtained highly depends on the specified optimization problem. Thus, by considering alternative structural realization approaches, not that stringent limitations on the structure and dimensions are required. This way, more innovative floater designs can be captured.
This paper addresses the topic of far-offshore wind energy exploitation. Far-offshore wind energy exploitation is not feasible with grid-connected floating wind turbines because grid-connection cost, installation cost and O&M cost would be prohibitive. An enabling technology is the energy ship concept, which is described and modeled in the paper. A design of an energy ship is proposed. It is estimated that it could produce 5 GWh per annum of chemical energy (methanol).
In this paper, the influence of optimal wind farm control and optimal wind farm layout is investigated in terms of power production. The capabilities of the developed optimization platform is demonstrated on the Swedish offshore wind farm, Lillgrund. It shows that the expected annual energy production can be increased by 4 % by integrating the wind farm control into the design of the wind farm layout, which is 1.2 % higher than what is achieved by optimizing the layout only.
Wind turbines rotate clockwise. The rotational direction of the rotor interacts with the nighttime veering wind, resulting in a rotational-direction impact on the wake. In the case of counterclockwise-rotating blades the streamwise velocity in the wake is larger in the Northern Hemisphere whereas it is smaller in the Southern Hemisphere.
To get a better understanding of noise emissions from wind turbines at frequencies far below the audible range, simulations with increasing complexity were conducted. Consistent with the literature, it has been found that acoustic emission is dominated by the noise generated when the rotor blades pass the tower. These specific frequencies are less dominant in the structure-borne emission. Considering aerodynamic forces acting on the tower is important for the correct modeling of emissions.
This work provides a possible solution to closed-loop flow control in a wind farm.
The remote sensing technology, lidar, which is a laser-based measurement system, is used to obtain wind speed information behind a wind turbine. The measurements are processed using a model-based approach to estimate position information of the wake. The information is then used in a controller to redirect the wake to the desired position. Altogether, the concept aims to increase the power output of a wind farm.
In this study we aimed to use the spinner anemometer calibration and nacelle transfer function determined on a reference wind turbine to assess the power performance of a second, identical turbine. An experiment was set up with a met mast and spinner anemometer on each turbine. For each wind turbine, the nacelle power curve agreed with the corresponding power curve within 0.10 % of AEP for the reference wind turbine and within 0.38 % for the second wind turbine, for a mean wind speed of 8 m s−1.
Year-to-year variability of wind speeds limits the certainty of wind-plant preconstruction energy estimates ("resource assessments"). Using 62-year records from 60 stations across Canada we show that resource highs and lows persist for decades, which makes estimates 2–3 times less certain than if annual levels were uncorrelated. Comparing chronological data records with randomly permuted versions of the same data reveals this in an unambiguous and easy-to-understand way.
When assessing wind resources for wind farm development, the first step is to measure the wind from tall meteorological masts. As met masts are expensive, they are not built at every planned wind turbine position but sparsely while trying to minimize the distance. However, this paper shows that it is better to focus on the similarity between the met mast and the wind turbines than the distance. Met masts at similar positions reduce the uncertainty of wind resource assessments significantly.
A method for performing numerical wind resource assessments in the absence of on-site measurements is presented and validated against field measurements. Numerical wind resource assessment is at least 2 orders of magnitude faster and less expensive than using conventional site measurements. This enables analysis of a larger number of projects and thereby increases the chances of discovering the best available sites. The uncertainty in mean wind speed predictions is found to be about 4 %.
Blade bearings of wind turbines experience unusual loads compared to bearings in other industrial applications, which adds some difficulty to the application of otherwise well-established calculation methods, like fatigue lifetime. As a result, different methods for such calculations can be found in the literature. This paper compares three approaches of varying complexity and comes to the conclusion that the simplest of the methods is very inaccurate compared to the more complex methods.
An accurate assessment of the wind resource at hub height is necessary for an efficient and bankable wind farm project. Conventional techniques for wind speed vertical extrapolation include a power law and a logarithmic law. Here, we propose a round-robin validation to assess the benefits that a machine-learning-based approach can provide in vertically extrapolating wind speed at a location different from the training site – the most practically useful application for the wind energy industry.
This research paper proposes a generic structure of electrical test benches and a novel categorization of test options for experimental analysis of wind turbines and wind power plants. The new proposed test structure would concern the increasing challenges in wind power integration and control including reliability, stability, harmonic interactions, and control performance of WPPs in connection to different types of AC and HVDC transmission systems.
Wind speeds can be measured remotely from the ground with lidars. Their estimates are accurate for mean speeds, but turbulence leads to measurement errors. We predict these errors using computer-generated data and compare lidar measurements with data from a meteorological mast. The comparison shows that deviations depend on wind direction, measurement height, and wind conditions. Our method to reduce the measurement error is successful when the wind is aligned with one of the lidar beams.
A computational fluid dynamics (CFD) solver is coupled with a structure solver to predict the dynamic response of a horizontal axis wind turbine structure. CFD provides much more accurate and more realistic aerodynamic loads that cannot be achieved by traditional methods such as blade element momentum theory. As a result, the aeroelastic response of the wind turbine structure, taking into account blade–tower interactions, is described in more detail.
This paper presents a review of existing theory and practice relating to main bearings for wind turbines. Topics covered include wind conditions and resulting rotor loads, main-bearing models, damage mechanisms and fault detection procedures.
Multi-element ducts are investigated to further improve the aerodynamic performance of ducted wind turbines. CFD simulations are performed for a multi-element duct geometry consisting of a duct and a flap; the goal is to evaluate the effects on the aerodynamic performance of the radial gap length and the deflection angle of the flap. Increasing the radial gap length results in an augmentation of the total thrust generated by the DWT, whereas a larger deflection angle has an opposite effect.
The US Department of Energy's National Renewable Energy Laboratory (NREL) and industry partners successfully demonstrated a new gearbox design using preloaded tapered roller bearings in the planetary section. The new gearbox design demonstrated improved planetary load-sharing characteristics in the presence of rotor pitch and yaw moments, resulting in a predicted gearbox lifetime that is 3.5 times greater than the previous conventional design with cylindrical roller bearings.
An important design criterion for the electric drive system of a wind turbine is the fulfilment of grid codes given by transmission system operators. The grid codes state how wind turbines/farms must behave when connected to the grid in normal and abnormal conditions. A type of testing equipment that comprises the use of fully-rated voltage source converter in back-to-back configuration for grid code testing is proposed. Test results of a 4 MW wind turbine and an 8 MW test equipment are shown.
The modeling, operating strategy, and controller design for an actual in-field wind turbine with a fixed-displacement hydraulic drivetrain are presented. An analysis is given on a passive torque control strategy for below-rated operation. The turbine lacks the option to influence the system torque by a generator, so the turbine is regulated by a spear valve in the region between below- and above-rated operation. The control design is evaluated on a real-world 500 kW hydraulic wind turbine.
This work presents a new fully automated method to correct for
pitch misalignment imbalances of wind turbine rotors. The method
has minimal requirements, as it only assumes the availability of a
sensor of sufficient accuracy and bandwidth to detect the 1P
harmonic to the desired precision and the ability to command the
pitch setting of each blade independently from the others.
Extensive numerical simulations are used to demonstrate the new
This paper shows how a definitive part of the commonly used Mann (1994) atmospheric turbulence model (its so-called eddy lifetime) implies that the model parameters can be directly related to typical measurements in wind energy projects. Most importantly, the characteristic turbulence length scale is found in terms of commonly measured (10 min mean) quantities (shear and standard deviation of wind speed); this estimator is found to give useful results, over different sites and flow regimes.
Current fast aeroelastic wind turbine codes suitable for certification lack an induction model for standstill conditions. A near-wake model for wind turbines in operation is extended to cover these conditions. The model is validated in aerodynamic simulations of the NREL/NASA Ames Phase VI rotor. Good agreement with the experiments has been obtained in attached flow and beginning separation. Aeroelastic simulations of the DTU 10 MW turbine in standstill indicate a minor impact of the model.
Understanding uncertainties in wind resource assessment associated with the use of the output from numerical weather prediction (NWP) models is important for wind energy applications. A better understanding of the sources of error reduces risk and lowers costs. Here, an intercomparison of the output from 25 NWP models is presented. The study shows that model errors are larger and agreement between models smaller at inland sites and near the surface.
As wind turbines extract energy from the wind, a wind field of reduced wind speed and increased turbulence is left behind for the downstream turbines. For the exact calculation of the annual energy production and lifetime of wind turbines, it is therefore of great importance to be able to accurately calculate this turbulent wake flow for different wind conditions. This paper compares different computational modeling approaches with flow measurements on model turbines in a wind tunnel.
This work characterizes the unsteady aerodynamic response of a scaled version of a 10 MW floating wind turbine subjected to an imposed platform motion. The focus has been put on the simple yet significant motion along the wind's direction (surge). For this purpose, different state-of-the-art aerodynamic codes have been used, validating the outcomes with detailed wind tunnel experiments. This paper sheds light on floating-turbine unsteady aerodynamics for a more conscious controller design.
Revised manuscript under review for WES(discussion: final response, 6 comments)
This paper describes the design and field testing of a controller for reducing wake interactions on a wind farm. Reducing the power of some turbines weakens their wakes, allowing other turbines to produce more power, so that the total wind farm power may increase. There have been doubts that this is feasible, but these field tests on a full-scale wind farm have demonstrated that this goal has been achieved, also providing convincing validation of the model used for designing the controller.
Revised manuscript accepted for WES(discussion: closed, 6 comments)
This research paper investigates for the first time the potential of thrust set-point optimization in large wind farms for mitigating gravity-wave induced blockage effects, with the aim of increasing the wind-farm energy extraction. The optimization tool is applied to almost two thousand different atmospheric states. Overall, energy gains above 4 % are observed for 77 % of the cases.
Preprint under review for WES(discussion: final response, 6 comments)
One particular problem with structure operating in seas is the so-called fatigue phenomena. Cyclic loads imposed by waves and winds can cause structural failure after a number of cycles. The tradition method have some limitations.
This paper presents a developed design framework based on Fracture Mechanics for offshore wind turbine support structures which enables the design engineer to make most of available inspection capabilities and optimise the design and inspection, simultaneously.
Wind turbine blade prototypes undergo structural tests before they are used in the field so that any design failure can be detected prior to their operation. In this study, strength characteristics of a small-scale existing 5 m composite wind turbine blade is carried out utilizing the finite-element-method software package Ansys. The results show that the blade exhibits sufficient resistance against buckling. Yet, laminate failure is found to play a major role in the ultimate blade failure.
This paper introduces the Tocha wind farm, presents the different layouts adopted in the instrumentation of the wind turbines and shows initial results. At this preliminary stage, the capabilities of the very extensive monitoring layout are demonstrated. The results presented demonstrate the ability of the different monitoring components to track the modal parameters of the system, composed of tower and rotor, and to characterize the internal loads at the tower base and blade roots.
Wake losses significantly reduce the power production of utility-scale wind farms since all wind turbines are operated in a greedy, individual power maximization fashion. In order to mitigate wake losses, collective wind farm operation strategies use wake steering, in which certain turbines are intentionally misaligned with respect to the incoming wind direction. The control strategy developed is dynamic and closed-loop to adapt to changing atmospheric conditions.
With the increase in installed wind capacity, the rotor diameter of wind turbines is becoming larger and larger, and therefore it is necessary to take aeroelasticity into consideration. At the same time, wind turbines are in reality subjected to atmospheric inflow leading to high wind instabilities and fluctuations. Within this work, a high-fidelity chain is used to analyze the effects of both by the use of models of the same turbine with increasing complexity and technical details.
Revised manuscript accepted for WES(discussion: closed, 6 comments)
As windfarms are moving further offshore, logistical concepts increasingly include service operation vessels (SOVs) as the prime means of service delivery. Complex SOV operations make their safety management difficult. Existing risk assessments are done piecemeal and potentially lacking completeness when integrated. The paper performs systemic hazard analysis to (1) bring awareness of hazards that may have been overlooked and (2) allow for a preliminary comparison of various operational stages.
This work presents an improvement to a popular algorithm used in the prediction of the power of vertical axis wind turbines. While, until now, the algorithm was considered valid only for low rotor speeds and a small number of blades, our improvement makes it valid for any configuration. The predictions of the improved algorithm were found to be of considerable accuracy, when validated with measurements of a small model turbine acquired in a compressed air wind tunnel.
Meteorological and oceanic datasets are fundamental to the modeling of offshore wind farms. Data quality issues in one such dataset led us to conduct a study to establish whether such issues are more generally present in these datasets. The answer is yes and users should be aware of this. We therefore also investigated how such issues can be avoided. The result is a set of techniques and recommendations for dataset producers, leading to substantial quality improvements with limited extra effort.
Bat carcass surveys guided by likely fall zone distributions require accurate descriptions of carcass aerodynamics. This research introduces a new methodology resulting in the first direct measurements of bat carcass drag coefficients. The drag coefficient for three carcasses of three different species was found to be within a range of 0.70–1.23, with a terminal velocity between 6.63 and 17.57 m s−1. This information is useful for assessing the impact of wind farm projects on wildlife.
This paper presents a comparison study of the simplified model QuLAF (Quick Load Analysis of Floating wind turbines) and FAST for the planar version of various design load cases, in order to investigate how accurate results can be obtained from this simplified model.
The overall analysis shows that QuLAF is generally very good at estimating the bending moment at the tower base and the floater motions, whereas the nacelle acceleration is generally underpredicted.
UAS systems provide in situ measurements of turbulence and wind conditions. In the presented paper, the tip vortex generated by wind energy converters (WECs) is measured by a fixed-wing UAS and compared to an analytical model as well as a literature value. The results show good agreement. The presented method is a basis for future measurement campaigns to compare UAS measurements with numerical simulations of WEC wakes.
A new active power control (APC) approach is investigated to simultaneously reduce the wake-induced power tracking errors and structural fatigue loads of individual turbines within a wind farm. The non-unique solution of the APC problem with respect to the distribution of the individual powers is exploited. The simple control architecture and practical measurement system make the proposed approach prominent for real-time control of large wind farms with turbulent flows and wakes.
In the frame of a multi-body simulation of a wind turbine, the lowest rotor blade eigenmodes are used to describe their elastic deformations. In this paper, a finite Timoshenko beam element is proposed based on the transfer matrix method. The element stiffness and mass matrices are derived by numerical integration of the differential equations of motion. A numerical example with generic rotor blade data demonstrates the performance of the method in comparison with FAST/ADAMS software results.
Active wake deflection (AWD) aims to increase the power output of a wind farm by misaligning the yaw of upstream turbines. We analysed the effect of dynamic wind direction changes on AWD. The results show that AWD is very sensitive towards these dynamics. Therefore, we present a robust active wake control, which considers uncertainties and wind direction changes, increasing the overall power output of a wind farm. A side effect is a significant reduction of the yaw actuation of the turbines.
Franz Mühle, Jannik Schottler, Jan Bartl, Romain Futrzynski, Steve Evans, Luca Bernini, Paolo Schito, Martín Draper, Andrés Guggeri, Elektra Kleusberg, Dan S. Henningson, Michael Hölling, Joachim Peinke, Muyiwa S. Adaramola, and Lars Sætran
In this paper, detailed inflow information extracted from measurements is used to improve the accuracy of simulated wind turbine fatigue loads. Inflow information from nearby met masts is utilised as well as information from a blade-mounted flow sensor in combination with a method to compensate for the disturbance to the flow caused by the presence of the wind turbine.
Wind turbines are exposed to harsh weather, leading to surface defects on rotor blades emerging from the first day of operation. Defects
grow quickly and affect the performance of wind turbines. Thus, there is demand for an easily applicable remote-inspection method that is sensitive to small
surface defects. In this work we show that infrared thermography can meet these requirements by visualizing differences in the surface temperature
of the rotor blades downstream of surface defects.
Our experimental wind tunnel study on a pair of model wind turbines demonstrates a significant potential of turbine yaw angle control for the combined optimization of turbine power and rotor loads. Depending on the turbines' relative positions to the incoming wind, a combined power increase and individual rotor load reduction can be achieved by operating the turbine rotors slightly misaligned with the main wind direction (i.e., at a certain yaw angle).
In this work, the wake flows behind two different model wind turbines were investigated in wind tunnel experiments user laser Doppler anemometry. It was found that the width of the wake flow is significantly dependent on the quantities examined, becoming much wider when taking higher-order statistics into account. This effect is stable against yaw misalignment and thus affects not only wind farm layout optimizations but also the applicability of active wake steering methods.
The wind speed measured by a flow sensor mounted on the blade of a wind turbine is disturbed by the turbine. This paper presents a method to obtain the free turbulence inflow by compensating for this disturbance.
The method is tested using numerical simulations and can be used to extract inflow information for accurate aeroelastic load simulations.
This paper presents a well-defined procedure for measuring the displacement on a HAWT via stereo photometry. The method is demonstrated by measuring displacement during operation of a scaled down turbine model. The method is developed in (1) camera calibration and (2) tracking algorithm. We introduce an efficient, easy and practical calibration method for measurement in the large field of views that is always a challenge. Tracking algorithm also tracks the markers during rotation successfully.
An efficient and accurate Monte Carlo approach is presented to assess the lifetime fatigue loading on a floating offshore wind turbine accurately. This is typically challenging in simulation effort due to the many different combinations of relevant environmental conditions which need to be considered. The applied method uses quasi-random Sobol sequences and shows promising performance with respect to convergence and accuracy.
A reliable load history is crucial for a fatigue assessment of wind turbines. However, installing strain sensors to measure the load history on every wind turbine is not economically feasible. In this paper, a technique is proposed to reconstruct the thrust load history of a wind turbine based on high-frequency SCADA data and a trained neural network. Both simulated and real-world results show the potential of high-frequency SCADA for thrust load reconstruction.
We analyze the wake of the Anholt offshore wind farm in Denmark by intercomparing models and measurements. We also look at the effect of the land on the wind farm by intercomparing mesoscale winds and measurements. Annual energy production and capacity factor estimates are performed using different approaches. Lastly, the uncertainty of the wake models is determined by bootstrapping the data; we find that the wake models generally underestimate the wake losses.
This study shows that using probabilistic forecasting can improve next-day production forecasts for wind energy in cold climates. Wind turbines can suffer from severe production losses due to icing on the turbine blades. Short-range forecasts including the icing-related production losses are therefore valuable when planning for next-day energy production. Probabilistic forecasting can also provide a likelihood for icing and icing-related production losses.
The paper discusses load effects on wind turbines operating under misaligned-flow operations, which is part of a strategy to optimize wind-power-plant power production, where upwind turbines can be rotated off the wind axis to redirect their wakes. Analytical simplification, aeroelastic simulations, and field data from an instrumented turbine are compared and interpreted to provide an informed picture on the loads for various components.
A model for quick load analysis is presented. This is a fast model for the calculation of dynamic loads of an offshore wind turbine tower and foundation. The model is compared to the state-of-the-art aeroelastic code. In general, there is good similarity between the two models. This indicates that in the early stage of the design phase a simple dynamic model can be used to make a preliminary design of the foundation and wind turbine tower.
We conducted measurements in a wind tunnel with the remote sensing technique lidar to map the flow around a row of three model wind turbines. Two lidars were positioned near the wind tunnel walls to measure the two-dimensional wind vector over a defined scanning line or area without influencing the flow itself. A comparison of the lidar measurements with a hot-wire probe and a thorough uncertainty analysis confirmed the usefulness of lidar technology for such flow measurements in a wind tunnel.
This paper demonstrates optimization of wind turbine locations within a utility-scale wind plant using a nonlinear flow model and gradient-based optimization techniques made possible through the use of adjoints. This represents a groundbreaking improvement in model fidelity and optimization efficiency for wind energy applications. The optimized wind farms demonstrate significant improvements in annual energy production with turbine layouts that take advantage of nonlinear flow curvature effects.
The series of GABLS model intercomparison benchmarks is revisited in the context of wind energy atmospheric boundary layer (ABL) models. GABLS 1 and 2 are used for verification purposes. Then GABLS 3 is used to develop a methodology for using realistic mesoscale forcing for microscale ABL models. The method also uses profile nudging to dynamically reduce the bias. Different data assimilation strategies are discussed based on typical instrumentation setups of wind energy campaigns.
The paper discusses different concepts for reducing loads on wind turbines using movable blade tips. Passive and semi-passive tip solutions move freely in response to air load fluctuations, while in the active case an actuator drives the tip motion in response to load measurements. The various solutions are compared with a standard blade and with each other in terms of their ability to reduce both fatigue and extreme loads.
Efficient detection of wind turbines operating below their expected power output and immediate corrections help maximize asset value. The method presented estimates the environmental conditions from turbine states and uses pre-calculated power lookup tables from a numeric wake model to predict the expected power output. Deviations between the expected and the measured power output are an indication of underperformance. A demonstration of the method's ability to detect underperformance is given.
Fluid–structure interaction analysis of a membrane blade concept has been performed for a non-rotating blade under steady inflow conditions. The membrane blade consists of a rigid mast at the leading edge, ribs along the blade, tensioned edge cables at the trailing edge, and membranes forming the upper and lower surface of the blade. The studied membrane blade shows a higher lift curve slope and higher lift-to-drag ratio compared with its rigid-blade counterpart.