Wind Turbine Cluster Research Articles

Control of a wind turbine cluster based on squirrel cage induction generators connected to a single VSC power converter

A control scheme for Squirrel Cage Induction Generator (SCIG) based wind turbine cluster (or entire wind farm) connected by means of a single high voltage direct current–Voltage Source Converter (VSC–HVDC) is presented. The proposed scheme extends the SCIG vector control of a single wind turbine with a full power converter to a whole wind turbine cluster with a single power converter calculating and regulating the optimal wind turbine cluster electrical frequency. This concept can be used in wind farms composed of multiple clusters of multiple wind turbines, and for entire offshore wind farms connected to the main grid using a HVDC point-to-point connection or a multiterminal HVDC grid.

The NOWITECH Reference Wind Farm

The NOWITECH Reference Wind Farm is a base case wind farm model created for research and benchmarking. The reference wind farm is using an adapted version of the DTU 10 MW reference turbine, and is similar to the Dogger Bank Creyke Beck A wind farm in terms of size and distance to shore. The focus of this paper has been to research cost effective solutions for the internal grid in an offshore wind farm. As offshore wind farms and wind turbines grow in size, it could be a better solution to install 66kV collector grids within the wind farm rather than 33kV collector grids. In order to save costs it is also researched if it is possible to eliminate the expensive transformer substation within offshore wind farms connected to shore with HVDC technology, as the HVDC converter platform already contains a transformer. The reference wind farm was implemented as a MATLAB Simulink model in order to run load flow simulations to analyze the efficiency of these internal grid configurations. A cost analysis was also carried out for the internal grid in the wind farm. Results show that installing a 66kV collector network saves cost both in terms of lower power losses and cheaper cables. Overall cost might still increase due to more expensive transformers, but uncertainties in the cost estimates makes it difficult to assess whether 33 or 66kV collector grids are more economical. Eliminating the transformer substation also saves a large amount of costs, as long as the distance between the clusters of wind turbines and the HVDC converter platform is moderate. The savings is realized mainly due to the elimination of the expensive platform structure, but is offset by higher cable costs as distance to the converter platform increases. A more detailed study of challenges related to removing the transformer platform in terms of fault handling and redundancy should be conducted in the future.

Development of Optimal Maintenance Strategies for Offshore Wind Turbine by Using Artificial Neural Network

Nowadays offshore wind turbines are an emerging area of research in renewable energy fields. They are penetrating in the energy market with a very rapid scale which entails to address their availiblity issues in novel and cost effective way. The maintenance activities for the array of wind turbines in a wind farm are complex in an offshore environment due to access, weather, and logistic issues. A method has been proposed to address the complexity and intricacies in the planning and formulation of maintenance strategies by using neural network. The approach consists of undertaking the clustering analysis of the array of wind turbines and then to segregate the similar wind turbines based on the analysis results by using Self Organizing Map, (SOM) neural network. Then to predict the expected power output of the wind turbines being part of a certain cluster with the help of another Standard Back Propagation (SBP) neural network. Based on the results, a number of implications have been outlined regarding how the results obtained could provide opportunities in optimizing maintenance strategies. Moreover the weather window issues will be addressed coherently to exploit the information of expected power output in advance by using SBP neural network. So based on the proposed method, the link among the condition of the components, clusters of wind turbines, prediction of power output and the access levels based on the weather conditions could be established which will provide the sound basis for developing optimal and cost effective maintenance strategies. It is expected that the implementation of the proposed approach will enhance the reliability and availability of the offshore wind turbines at the wind farm level in a better and viable way.

Path Creation: Co-creation of Heterogeneous Resources in the Emergence of the Danish Wind Turbine Cluster

This paper employs path creation as a lens to follow the emergence of the Danish wind turbine cluster. Supplier competencies, regulations, user preferences and a market for wind power did not pre-exist; all had to emerge in a tranformative manner involving multiple actors and artefacts. Competencies emerged through processes and mechanisms such as co-creation that implicated multiple learning processes. The process was not an orderly linear one as emergent contingencies influenced the learning processes. An implication is that public policy to catalyse clusters cannot be based on an assumption that linear learning dynamics will unfold.

On the Doing of Local History in New York

New York has a long tradition of attention to, caring for, and publishing local history. Approaches and intensity of interest have varied over the years, with strong local history publication in the late nineteenth century, an ebb in the early twentieth century but also a state requirement in 1919 for local governments to designate official historians, a sharp rise in the period of the bicentennial of the American Revolution, and, since that time, an increase in the publication of local history and other examples of progress such as the annual conference on New York State history. At the same time, though, there are substantial needs for better training, advocacy, and resources. New York's history programs are “like a cluster of wind turbines,” many exuding energy, others quiet and still. A stronger role for the State Historian is essential to strengthening the preservation, management, and use of state and local history.

Long-Term Wind Speed and Power Forecasting Using Local Recurrent Neural Network Models

This paper deals with the problem of long-term wind speed and power forecasting based on meteorological information. Hourly forecasts up to 72-h ahead are produced for a wind park on the Greek island of Crete. As inputs our models use the numerical forecasts of wind speed and direction provided by atmospheric modeling system SKIRON for four nearby positions up to 30 km away from the wind turbine cluster. Three types of local recurrent neural networks are employed as forecasting models, namely, the infinite impulse response multilayer perceptron (IIR-MLP), the local activation feedback multilayer network (LAF-MLN), and the diagonal recurrent neural network (RNN). These networks contain internal feedback paths, with the neuron connections implemented by means of IIR synaptic filters. Two novel and optimal on-line learning schemes are suggested for the update of the recurrent network's weights based on the recursive prediction error algorithm. The methods assure continuous stability of the network during the learning phase and exhibit improved performance compared to the conventional dynamic back propagation. Extensive experimentation is carried out where the three recurrent networks are additionally compared to two static models, a finite-impulse response NN (FIR-NN) and a conventional static-MLP network. Simulation results demonstrate that the recurrent models, trained by the suggested methods, outperform the static ones while they exhibit significant improvement over the persistent method.

A Fuzzy Model for Wind Speed Prediction and Power Generation in Wind Parks Using Spatial Correlation

In this paper, a fuzzy model is suggested for the prediction of wind speed and the produced electrical power at a wind park. The model is trained using a genetic algorithm-based learning scheme. The training set includes wind speed and direction data, measured at neighboring sites up to 30 km away from the wind turbine clusters. Extensive simulation results are shown for two application cases, providing wind speed forecasts from 30 min to 2 h ahead. It is demonstrated that the suggested model achieves an adequate understanding of the problem while it exhibits significant improvement compared to the persistent method.

A Viscous Three-Dimensional Differential/Actuator-Disk Method for the Aerodynamic Analysis of Wind Farms

Computational Fluid Dynamics (CFD) is a promising tool for the analysis and optimization of wind turbine positioning inside wind parks (also known as wind farms) in order to maximize power production. In this paper, 3-D, time-averaged, steady-state, incompressible Navier-Stokes equations, in which wind turbines are represented by surficial forces, are solved using a Control-Volume Finite Element Method (CVFEM). The fundamentals of developing a practical 3-D method are discussed in this paper, with an emphasis on some of the challenges that arose during their implementation. For isolated turbines, results have indicated that the proposed 3-D method attains the same level of accuracy, in terms of performance predictions, as the previously developed 2-D axisymmetric method and the well-known momentum-strip theory. Furthermore, the capability of the proposed method to predict wind turbine wake characteristics is also illustrated. Satisfactory agreement with experimental measurements has been achieved. The analysis of a two-row periodic wind farm in neutral atmospheric boundary layers demonstrate the existence of positive interference effects (venturi effects) as well as the dominant influence of mutual interference on the performance of dense wind turbine clusters.

Local Wind Climate within and Downwind of Large Offshore Wind Turbine Clusters

This Note reviews presentations made at a meeting for experts, held at the Risø Laboratory, concerning possible approaches to improving understanding of local wind climates within large offshore wind farms and particularly their impact on downwind wind farms. Such effects are detectable both within and beyond the boundaries of each windfarm, which typically has 10 to 100 large turbines of rotor diameter up to 80 m and hub-heights between 60 and 80 m.

Large-scale wind power in Denmark

The goal of the official energy plan for Denmark is to produce 10% of the country's electricity from wind power. A total of 3% has so far been achieved, but it now looks as though there will be a moratorium on further wind power development. If the 10% target is to be reached stricter planning control will be required so as not to destroy the landscape through the building of windfarms and clusters of wind turbines, and in the long run offshore windfarms will be the way forward. However, there are currently obstacles arising from conflicting interests, and there is also a ban on private ‘windmill guilds’ as operators of offshore windfarms. The private guilds have during the last 20 years been the primary driving force behind the dissemination of wind energy in the national energy system.

Reduction of horizontal wind speed in a boundary layer with obstacles

The reduction of horizontal wind speed at hub height in an infinite cluster of wind turbines is computed from a balance between a loss of horizontal momentum due to the drag and replenishment from above by turbulent fluxes. This reduction is derived without assumptions concerning the vertical wind profile above or below hub height, only some basic assumptions on turbulent exchange have been made. Two applications of the result are presented, one considering wind turbines and one pressure drag on orographic obstacles in the atmospheric boundary layer. Both applications are basically governed by the same kind of momentum balance.

Power fluctuations in spatially dispersed wind turbine systems

We report on a second-order statistical analysis of power output time series from spatially dispersed wind turbine clusters (typical area dimension 50 × 100 km). Power fluctuations in the time scale of hours are investigated. It is shown that, for the analyzed climatic region (northwest Germany), a main effect of geographical dispersion is the reduction of large hour to hour changes in the overall power output.

On the wind speed reduction in the center of large clusters of wind turbines

The paper intends to probe the feasibility of a method to determine the wind speed reduction in the center of a large wind farm by means of simple boundary layer theory. Therefore, the simplest possible assumptions have been chosen. The paper presents an approach for determination of the reduction of the wind speed assuming the wind turbines effectively act as roughness elements. The model makes use of similarities to so-called canopy flows, where the surface drag and the drag on individual obstacles are added to form the total drag. Results are compared with existing models for reduction of efficiency within wind turbine clusters and good aggeement is found.

Aerodynamic interference of vertical axis wind turbines

The laminar flowfield and performance of clusters of two-dimensional vertical axis wind turbines are analyzed by idealizing the rotors as momentum sources. The flowfield dominated by the pressure field of the operating turbines is determined by solving the incompressible Navier-Stokes equations and mutual interference is observed to be elliptic in nature. Physical positioning of the turbines with respect to each other significantly affects the aerodynamic performance of the turbines. Several cases are examined starting with one pair of turbines to a cluster as big as 21 turbines. From the two-dimensional cases presented here, it is concluded that downwind turbine performance can be improved by judiciously choosing the angular relative orientation of the turbines. Qualitative analysis of the laminar flowfield is conducted with the help of streamlines and pressure contours. Performance comparisons of the turbines are made with a stand-alone turbine operating under similar flow conditions. This investigation lays the groundwork for the optimization of a wind farm.

Application of a Monte Carlo Optimization Technique to a Cluster of Wind Turbines

An optimization process based on Monte Carlo sampling of the variables of the objective function is developed. The technique is applied to a grouping of wind turbine generators. The permissible dimensions of the wind turbines and the range of spacing between adjacent machines allows a search for a maximum profit over a wide mix of system configurations. The computer program for the optimization algorithm finds cluster configurations for maximum profit with a high degree of certainty for the relatively simple cost and design constraints used in the model.

Wind Array Power Prediction for Improved Operating Economics and Reliability

A methodology for predicting wind power variations one or more hours ahead is developed. This prediction method is required for unit commitment and generation control strategies that have been developed to provide economic and reliable operation for utilities with large wind penetrations. The methodology utilizes a set of meteorological towers that encircle the wind turbine clusters at a radius of 100 miles. The prediction methodology must determine (1) the direction of propagation of the meteorological event, (2) the subset of meteorological towers that are in the direction of motion of the meteorological event and thus are to be used in the prediction, (3) the delays between the prediction sites in the wind turbine cluster and the selected subset of reference sites in the ring of meteorological towers to be used for prediction, and (4) the parameters of the predictive model that predicts wind speed based on weighted delayed wind speed measurements at the selected reference sites.

Method for Evaluating Wind Turbine Wake Effects on Wind Farm Performance

A method of testing the performance of a cluster of wind turbine units and data analysis equations are presented which together form a simple and direct procedure for determining the reduction in energy output caused by the wake of an upwind turbine. This method appears to solve the problems presented by data scatter and wind variability. Test data from the three-unit Mod-2 wind turbine cluster at Goldendale, Washington, are analyzed to illustrate the application of the proposed method. In this sample case the reduction in energy was found to be about 10 percent when the Mod-2 units were separated a distance equal to 7 dia and winds were below rated.

Operations Requirements of Utilites With Wind Power Generation

A computational method and the necessary wind speed data are presented in this paper for quantifying in a probabilistic framework the load- following, operating-reserve and unloadable-generation requirements for a utility with one or more spatially dispersed wind turbine clusters. Results of this method are valid for random atmospheric conditions, excluding significant atmospheric events such as storm fronts. An example illustrating the proposed method is included.

Operations Requirements of Utilities with Wind Power Generation

A computational method and the necessary wind-speed data are presented in this paper for quantifying in a probabilistic framework the load-following, operating-reserve and unloadable-generation requirements for a utility with one or more spatially dispersed wind-turbine clusters. Results of this method are valid for random atmospheric conditions, excluding significant atmospheric events such as storm fronts. An example illustrating the proposed method is included.