Large Wind Farms Research Articles

LiDAR Measurements to Investigate Farm-to-Farm Interactions at the AWAKEN Experiment

The exponential growth of wind energy and the need to exploit wind resources over areas with higher energy potential have led to the construction of neighboring wind turbines and farms with relatively small separation distances. As a result, for specific wind and atmospheric conditions, the wakes generated by an upstream wind farm may affect wind resources available for a downstream wind farm resulting in detrimental impacts on energy harvesting and structural loads for the downwind wind turbines. Distances between neighboring wind farms are typically larger than those associated with intra-wind-farm wake interactions, generating cumulative wakes whose characteristics might differ from those predicted through classical engineering wake models. These phenomena are referred to as farm-to-farm interactions. A better understanding and characterization of farm-to-farm interactions is one of the science goals tackled by the ongoing American WAKE experimeNt (AWAKEN). The site under investigation for this field campaign comprises two large wind farms in northern Oklahoma, USA, which are spaced roughly 5km apart along the prevailing South-North wind direction. To investigate possible interactions between these two wind farms, the WindFluX mobile LiDAR station has been deployed mainly to perform volumetric scans over their gap region. In this paper, preliminary results from these LiDAR volumetric scans will be discussed, specifically for a case with multiple wind turbine wakes evolving during the occurrence of a low-level jet.

Graph machine learning for predicting wake interaction losses based on SCADA data

Six wind turbine generators (WTGs) within a larger wind farm were chosen to model wake interaction losses by means of graph machine learning. 10-minute averaged measurement data values at each turbine over a period of two years were available. Based on the measured power output and ”free-stream” (undisturbed) wind conditions, determined at an upstream WTG unaffected by wakes, a data-driven free-stream power curve model was built. The cluster of WTGs was then represented by a graph data structure in the form of an adjacency matrix, where the turbines are nodes, which in turn are connected by edges. The edges are represented by weightings, indicating the strength of wake interactions between connected turbines. To learn the weightings, the predicted power outputs of the free-stream power curve model were compared to the measured power outputs and then optimised to account for these differences, which are attributed to wake interaction losses. After that a new data-driven model was trained, with free-stream wind speed, direction and yaw misalignment as features and the optimised weightings as targets. It was shown that the model was able to predict power curves that show clear patterns due to wake effects and reduced the root mean squared error, based on the measured and predicted values, by 16.29% compared to the free-stream power curve model.

Comparison of effects of four subgrid-scale turbulence models in large eddy simulation of a large wind farm

Comparison of effects of four subgrid-scale turbulence models in large eddy simulation of a large wind farm

Cross-code verification of non-neutral ABL and single wind turbine wake modelling in LES

Large Eddy Simulations (LES) of atmospheric boundary layer (ABL) flow with the actuator disc (AD) for turbine modelling is a widely used method of simulating wind farm flows. Hence, it is important to understand the requirements for achieving a good comparison between ABL flow and turbine wakes between different research group setups, despite unavoidable differences which include LES numerical framework, sub-grid scale (SGS) model, and turbine modelling, and at grid resolutions achievable in large wind farm simulations. In this work, conventionally neutral (CNBL), stable (SBL) and convective (CBL) boundary layers, and single turbine wakes under these different conditions, are compared between the EPFL pseudo-spectral code WIRE LES and the DTU finite-volume code EllipSys3D. ABL profiles largely agree well with hub height velocity magnitudes agreeing to 0.7%, 3.8% and 0.5% for the CNBL, SBL and CBL respectively. The scale-dependent SGS model of WIRE LES results in reduced grid dependency, while EllipSys3D required higher grid resolution in the SBL. Wake flows show improved wake recovery and greater added turbulence intensity with increasing grid resolution, and good agreement is achieved with a radius R to cell size ratio of R/(dxdydz)1/3 ≥ 6.5. Trends in wake flow with different stability conditions, such as the influence of inflow turbulence intensity or shear, are well replicated between codes. Likewise, wake deficit and added TI profiles, and distributions of turbine power and thrust also agree well. Mean power output predictions match to 4.3%, 7.2% and 3.8% in the CNBL, SBL and CBL respectively between the two codes. Overall, these results demonstrate that good agreement is possible with aligned turbine data and sufficient grid resolution.

Simulating wake losses of the Danish Energy Island wind farm cluster

The increase in the number of installed offshore wind farms has led to clustering, where wind farm interaction can cause energy losses. The Danish government is planning an Energy Island in the North Sea consisting of ten 1 GW wind farms. The initial wind farm layout design from a consultant company (COWI) has reported a 5% annual energy production (AEP) wake loss based on a low-fidelity engineering wake model that is known to underestimate wind farm interaction. The present work employs higher-fidelity wake models based on Reynolds-averaged Navier-Stokes (RANS) and a mesoscale model, which predict higher AEP losses due to wind farm interaction; namely, between 8.6-10.1%. In addition, we investigate how a microscale model as RANS can be employed to simulate a large wind farm cluster efficiently, and we compare its results with mesoscale model simulations.

Study on boundary conditions of k-ε-fp turbulence model for wind turbine wake simulation

ABSTRACT Wind turbine wake has a great impact on the power output and fatigue damage of the whole wind farm. As a result, wind turbine wake simulation has attracted attention from both researchers and engineers for decades, in which turbulence modeling is the key point. Among all the turbulence modeling methods for wind turbine wake, k − ε − f P , a RANS model proposed by Laan in 2015, can balance well between efficiency and accuracy. In the current research, the boundary conditions of k − ε − f P turbulence model have been studied comprehensively. It was found that the boundary conditions of the model are affected by the wind conditions of the actual wind turbine, resulting in poor simulation accuracy under high wind shear conditions. Five simulation cases have been carried and compared with field measurement results. It was found that the roughness length z 0 is the most important factor for the determination of boundary conditions. Considering the wind shear conditions around wind turbine rotor, a new determination method for roughness length z 0 of the ground has been constructed, which corrects the wind profile prediction under high wind shear conditions. This research will contribute to the micro-sitting of future large wind farms.

Large-eddy simulation of a 15 GW wind farm: Flow effects, energy budgets and comparison with wake models

Planned offshore wind farm clusters have a rated capacity of more than 10 GW. The layout optimization and yield estimation of wind farms is often performed with computationally inexpensive, analytical wake models. As recent research results show, the flow physics in large (multi-gigawatt) offshore wind farms are more complex than in small (sub-gigawatt) wind farms. Since analytical wake models are tuned with data of existing, sub-gigawatt wind farms they might not produce accurate results for large wind farm clusters. In this study the results of a large-eddy simulation of a 15 GW wind farm are compared with two analytical wake models to demonstrate potential discrepancies. The TurbOPark model and the Niayifar and Porté-Agel model are chosen because they use a Gaussian wake profile and a turbulence model. The wind farm has a finite size in the crosswise direction, unlike as in many other large-eddy simulation wind farm studies, in which the wind farm is effectively infinitely wide due to the cyclic boundary conditions. The results show that new effects like crosswise divergence and convergence occur in such a finite-size multi-gigawatt wind farm. The comparison with the wake models shows that there are large discrepancies of up to 40% between the predicted wind farm power output of the wake models and the large-eddy simulation. An energy budget analysis is made to explain the discrepancies. It shows that the wake models neglect relevant kinetic energy sources and sinks like the geostrophic forcing, the energy input by pressure gradients and energy dissipation. Taking some of these sources and sinks into account could improve the accuracy of the wake models.

Addressing deep array effects and impacts to wake steering with the cumulative-curl wake model

Abstract. Wind farm design and analysis heavily rely on computationally efficient engineering models that are evaluated many times to find an optimal solution. A recent article compared the state-of-the-art Gauss-curl hybrid (GCH) model to historical data of three offshore wind farms. Two points of model discrepancy were identified therein: poor wake predictions for turbines experiencing a lot of wakes and wake interactions between two turbines over long distances. The present article addresses those two concerns and presents the cumulative-curl (CC) model. Comparison of the CC model to high-fidelity simulation data and historical data of three offshore wind farms confirms the improved accuracy of the CC model over the GCH model in situations with large wake losses and wake recovery over large inter-turbine distances. Additionally, the CC model performs comparably to the GCH model for single- and fewer-turbine wake interactions, which were already accurately modeled. Lastly, the CC model has been implemented in a vectorized form, greatly reducing the computation time for many wind conditions. The CC model now enables reliable simulation studies for both small and large offshore wind farms at a low computational cost, thereby making it an ideal candidate for wake-steering optimization and layout optimization.

Multilevel emulation for stochastic computer models with application to large offshore wind farms

Abstract Renewable energy projects, such as large offshore wind farms, are critical to achieving low-emission targets set by governments. Stochastic computer models allow us to explore future scenarios to aid decision making while considering the most relevant uncertainties. Complex stochastic computer models can be prohibitively slow, and thus an emulator may be constructed and deployed to allow for efficient computation. We present a novel heteroscedastic Gaussian Process emulator that exploits cheap approximations to a stochastic offshore wind farm simulator. We also conduct a probabilistic sensitivity analysis to understand the influence of key parameters in the wind farm model, which will help us to plan a probability elicitation in the future.

Large Scale, Mid Term Wind Farms Power Generation Prediction

Renewable energy sources, such as wind turbines, have become much more prevalent in recent years, and thus a popular form of energy generation. This is in part due to the ‘Fit for 55’ EU initiative, and in part, to rising fossil fuel prices, as well as the perceived requirement for nations to have power independence, and due to the influence of renewable energy sources we can see a marked increase in large wind farms in particular. However, wind farms by their very nature are highly inconsistent regarding power generation and are weather-dependent, thus presenting several challenges for transmission system operators. One of the options to overcome these issues is a system being able to forecast the generated power in a wide-ranging period—ranging from 15 min up to 36 h, and with an adequate resolution. Such a system would better help manage the power grid and allow for greater utilization of the green energy produced. In this document, we present a process of development for such a system, along with a comparison of the various steps of the process, including data preparation, feature importance analysis, and the impact of various data sources on the forecast horizon. Lastly, we also compare multiple machine learning models and their influence on the system quality and execution time. Additionally, we propose an ensemble that concatenates predictions over the forecast horizon. The conducted experiments have been evaluated on seven wind farms located in Central Europe. Out of the experiments conducted, the most efficient solution with the lowest error rate and required computational resources has been obtained for random forest regression, and two independent models; one for the short-term horizon, and the other, for the mid- to long-term horizon, which was combined into one forecasting system.

Understanding wind farm power densities

Kirby et al. (J. Fluid Mech., vol. 953, 2022, A39) adapted the two-scale momentum theory (Nishino & Dunstan, J. Fluid Mech., vol. 894, A2) to large finite-sized farms. They demonstrated that analytical estimates agree excellently with large eddy simulations, and that the model provides a good upper limit of the power production for a given array density. Crucially, they introduced the concepts of farm-scale losses, caused by the atmospheric response to the whole farm, and turbine-scale losses, owing to internal flow interactions in the wind farm. These two new theoretical concepts offer a novel way to analyse the performance of extended wind farms. For large offshore wind farms, losses at the wind-farm scale are typically twice as high as at the turbine scale. This demonstrates that there is limited potential for layout optimizations of extended arrays. Instead, optimization strategies should focus on developing methods to increase the energy entrainment into the wind farm. This work provides an exciting roadmap for analysing the effective efficiency of large wind farms.

Study on Atmospheric Stability and Wake Attenuation Constant of Large Offshore Wind Farm in Yellow Sea

Power generation estimation is one of the key steps in wind farm micro-sitting, and its accuracy is related to the wake decay constant in the wake model. Considering the influence of wind resource distribution in different regions, this study carried out interval optimization for the wake decay constant of offshore wind farms in the Yellow Sea region of China. Given the very small length of the sea surface roughness, atmospheric stability is a critical factor influencing the wake extent and recovery speed of offshore wind farms. WAsP 10 (Wind Atlas Analysis and Application software) simulates the wake of various scenarios, and the selection range of wake attenuation constant is investigated by combining the cases of two offshore wind farms in the Yellow Sea region. The study found that the higher the atmospheric stability, the larger the wake and the lower the wake attenuation coefficient. The Yellow Sea wind farm’s wake error system is between 0.03 and 0.04, and the forecast error can be controlled to within 3%. When simulating the wind farm at Yellow Sea offshore for improving power generation and economic evaluation, it is critical to select the correct value range of wake decay constant.

A Bayesian hierarchical assessment of night shift working for offshore wind farms

AbstractThis article presents a Bayesian data‐modelling approach to assessing operational efficiency at offshore wind farms. Input data are provided by an operational database provided by a large offshore wind farm which employs an advanced data management system. We explore the combination of datasets making up the database, using them to train a Bayesian hierarchical model which predicts weekly lost production from corrective maintenance and time‐based availability. The approach is used to investigate the effect of technician work shift patterns, specifically addressing a strategy involving night shifts for corrective maintenance which was employed at the site throughout the winter. It was found that, for this particular site, there is an approximate annual increase in time‐based technical availability of 0.64%. We explore the effect of modelling assumptions on cost savings; specifically, we explore variations in failure rate, price of electricity, number of technicians working night shift, extra staff wages, months of the year employing 24/7 working and extra vessel provision. Results vary quite significantly among the scenarios investigated, exemplifying the need to consider the question on a farm‐by‐farm basis.

Aerodynamic upgrades of a Darrieus vertical axis small wind turbine

Wind energy conversion is contributing significantly to the new clean energy transition but, up to now, such contribution is mainly driven by large multi-MW windfarms. On the other hand, the recent needs in terms of distributed energy production are revealing a new interest in small-scale wind energy conversion technologies for residential and urban applications. In this context, vertical axis wind turbines (VAWT) are often the most valid choice from the point of view of building integrating possibilities, but it is well known that they suffer a considerable gap in performance when compared to horizontal axis technologies. Based on these premises, in the present work some innovative solutions for improving the performance of a Darrieus vertical axis wind turbine are investigated through numerical and engineering approaches. The investigated interventions include rotor hybridization by the use of an inner Savonius section for improving the machine’s start-up and the possible application of dimples for improving the NACA 0021 airfoil performances on the outer rotor section. Results demonstrate the effectiveness of the interventions in obtaining a rotor with stable performances in a very wide range of wind regimes.

Turbulence structures and entrainment length scales in large offshore wind farms

Abstract. The flow inside and around large offshore wind farms can range from smaller structures associated with the mechanical turbulence generated by wind turbines to larger structures indicative of the mesoscale flow. In this study, we explore the variation in turbulence structures and dominant scales of vertical entrainment above large offshore wind farms located in the North Sea, using data obtained from a research aircraft. The aircraft was flown upstream, downstream, and above wind farm clusters. Under neutrally stratified conditions, there is high ambient turbulence in the atmosphere and an elevated energy dissipation rate compared to stable conditions. The intensity of small-scale turbulence structures is increased above and downstream of the wind farm, and it prevails over mesoscale fluctuations. But in stable stratification, mesoscale flow structures are not only dominant upstream of the wind farm but also downstream. We observed that the vertical flux of horizontal momentum is the main source of energy recovery in large offshore wind farms, and it strongly depends on the magnitude of the length scales of the vertical wind velocity component. The dominant length scales of entrainment range from 20 to ∼60 m above the wind farm in all stratification strengths, and in the wake flow these scales range from 10 to ∼100 m only under near-neutral stratification. For strongly stable conditions, negligible vertical entrainment of momentum was observed even just 2 km downstream of large wind farms. We also observed that there is a significant lateral momentum flux above the offshore wind farms, especially under strongly stable conditions, which suggests that these wind farms do not satisfy the conditions of an “infinite wind farm”.

Optimizing wind farm cable layout considering ditch sharing

AbstractWind power is becoming an important source of electrical energy production. In an onshore wind farm (WF), the electrical energy is collected at a substation from different wind turbines through electrical cables deployed over ground ditches. This work considers the WF layout design assuming that the substation location and all wind turbine locations are given, and a set of electrical cable types is available. The WF layout problem, taking into account its lifetime and technical constraints, involves selecting the cables to interconnect all wind turbines to the substation and the supporting ditches to minimize the initial investment cost plus the cost of the electrical energy that is lost on the cables over the lifetime of the WF. It is assumed that each ditch can deploy multiple cables, turning this problem into a more complex variant of previously addressed WF layout problems. This variant turns the problem best fitting to the real case and leads to substantial gains in the total cost of the solutions. The problem is defined as an integer linear programming model, which is then strengthened with different sets of valid inequalities. The models are tested with four WFs with up to 115 wind turbines. The computational experiments show that the optimal solutions can be computed with the proposed models for almost all cases. The largest WF was not solved to optimality, but the final relative gaps are small.

Optimization-Based Mitigation Techniques of the Temporary Overvoltage in Large Offshore Wind Farm

Overvoltage transient is the major danger facing large offshore wind farms, as it may lead to a complete failure of the system components. A 150 MW offshore wind farm was modeled using the ATP-EMTP package to analyze the effect of the temporary overvoltage resulting from de-energization conditions, and the ferro resonance phenomenon resulting from the asymmetric opening of circuit breaker pole(s). The results showed that the best scenario to shut down such a system from the grid with minimum values of temporary overvoltage is to disconnect it, feeder, by a feeder. The stuck of one or more poles during the opening operation of the circuit breaker is a dangerous phenomenon, as it induces a high overvoltage value, which lasts for a long time causing more electric stress on the sensitive components of the system. The use of the Pre-Insertion Resistor PIR was recommended as a mitigation technique for the temporary overvoltage in large offshore wind farms, as it proves a significant reduction in the overvoltage values reaches to 79.44 % compared to surge arrester and shunt reactor which reduce the over voltage by 68.33%, 66.29% respectively. The genetic algorithm was used to find the optimum value of the PIR and shunt reactor that leads to the minimum TOV. The optimum value of PIR which was recorded is equal to 57.352 Ohm, which leads to more reduction in the overvoltage value by 9.67%. Also, the optimum values of the resistor and inductor of the shunt reactor were found to be 11.238 Ohms and 37.246 mH which leads to more reduction in the overvoltage value by 21.97% compared to the reduction without optimization.

Prediction and Mitigation of Wind Farm Blockage Losses Considering Mesoscale Atmospheric Response

The engineering wind farm models currently used in industry can assess power losses due to turbine wake effects, but the prediction of power losses due to farm blockage is still a challenge. In this study we demonstrate a new prediction method of farm blockage losses and a possible strategy to mitigate them for a large offshore wind farm in the North Sea, by combining a common engineering wind farm model ’FLORIS’ with the ’two-scale momentum theory’ of Nishino and Dunstan (2020). Results show that the farm blockage losses depend significantly on the ’wind extractability’ factor, which reflects the strength of mesoscale atmospheric response. For a typical range of the extractability factor (assessed using a numerical weather prediction model) the farm blockage losses are shown to vary between about 5% and 15% of the annual energy production (AEP). However, these losses may be mitigated by adjusting turbine operating points taking into account the wind extractability. It is shown that a simple adjustment of the blade pitch angle and tip-speed ratio used below the rated wind speed may increase the AEP by up to about 2%.

Two-scale interaction of wake and blockage effects in large wind farms

Turbine wake and farm blockage effects may significantly impact the power produced by large wind farms. In this study, we perform large-eddy simulations (LES) of 50 infinitely large offshore wind farms with different turbine layouts and wind directions. The LES results are combined with the two-scale momentum theory (Nishino & Dunstan, J. Fluid Mech., vol. 894, 2020, p. A2) to investigate the aerodynamic performance of large but finite-sized farms as well. The power of infinitely large farms is found to be a strong function of the array density, whereas the power of large finite-sized farms depends on both the array density and turbine layout. An analytical model derived from the two-scale momentum theory predicts the impact of array density very well for all 50 farms investigated and can therefore be used as an upper limit to farm performance. We also propose a new method to quantify turbine-scale losses (due to turbine–wake interactions) and farm-scale losses (due to the reduction of farm-average wind speed). They both depend on the strength of atmospheric response to the farm, and our results suggest that, for large offshore wind farms, the farm-scale losses are typically more than twice as large as the turbine-scale losses. This is found to be due to a two-scale interaction between turbine wake and farm induction effects, explaining why the impact of turbine layout on farm power varies with the strength of atmospheric response.

Comparison of large eddy simulations against measurements from the Lillgrund offshore wind farm

Abstract. Numerical simulation tools such as large eddy simulations (LESs) have been extensively used in recent years to simulate and analyze turbine–wake interactions within large wind farms. However, to ensure the reliability of the performance and accuracy of such numerical solvers, validation against field measurements is essential. To this end, a measurement campaign is carried out at the Lillgrund offshore wind farm to gather data for the validation of an in-house LES solver. Flow field data are collected from the farm using three long-range WindScanners, along with turbine performance and load measurements from individual turbines. Turbulent inflow conditions are reconstructed from an existing precursor database using a scaling-and-shifting approach in an optimization framework, proposed so that the generated inflow statistics match the measurements. Thus, five different simulation cases are setup, corresponding to five different inflow conditions at the Lillgrund wind farm. Operation of the 48 Siemens 2.3 MW turbines from the Lillgrund wind farm is parameterized in the flow domain using an aeroelastic actuator sector model (AASM). Time-series turbine performance metrics from the simulated cases are compared against field measurements to evaluate the accuracy of the optimization framework, turbine model, and flow solver. In general, results from the numerical solver exhibited a good comparison in terms of the trends in power production, turbine loading, and wake recovery. For four out of the five simulated cases, the total wind farm power error was found to be below 5 %. However, when comparing individual turbine power production, statistical significant errors were observed for 16 % to 84 % of the turbines across the simulated cases, with larger errors being associated with wind directions resulting in configurations with aligned turbines. While the compared flapwise loads in general show a reasonable agreement, errors greater than 100 % were also present in some cases. Larger errors in the wake recovery in the far wake region behind the lidar installed turbines were also observed. An analysis of the observed errors reveals the need for an improved controller implementation, improvement in representing meso-scale effects, and possibly a finer simulation grid for capturing the smaller scales of wake turbulence.