Final Report Summary - WINDRIVE (Industrialization of a 3 MW Medium-Speed Brushless DFIG Drivetrain for Wind Turbine Applications)
Executive Summary:
The Windrive project aims to further develop a Brushless Doubly-Fed Induction Generator (Brushless DFIG) technology for widespread implementation in wind power generation, moving it from being a promising and proven concept, demonstrated on small scales, to an optimized industrial-scale machine for multi-megawatt wind turbine applications. This will have a range of impacts – growth in sales and jobs for the SMEs collaborating in the project, knock-on benefits through the supply chain and substantial environmental and social benefits through a more cost-effective and reliable wind power sector.
The wind industry is growing rapidly. In 2009, the EU adopted a wide-ranging package on climate change with a target for a 20% in emissions of Green House Gases by 2020, compared with 1990 levels, a 20% increase in the share of renewables in the energy mix, and a 20% cut in energy consumption. To achieve this, there will be a significant dependence on the generation of energy using wind turbines and, accordingly, wind turbines are increasingly being deployed in large numbers.
However, the reliability and cost-effectiveness of wind turbines still need major improvement to make wind power competitive with conventional power plants. This is particularly important for the growth of offshore wind power generation.
The proposed medium-speed drivetrain incorporates the following components:
- Medium-speed Brushless DFIG, excluding brush-gear and slip-rings, known to be the highest failure rate components in the generator;
- Partially (typically a third) rated converter;
- Two-stage gearbox, excluding the third high-speed stage, known to be the highest failure rate part of the gearbox.
A recent study performed by an independent wind energy consultant shows that the whole balance of life saving as percentage of drivetrain (i.e. generator, converter and gearbox) capital costs can be more than 83%.
The technology is proven in concept on small scales, including a 20 kW prototype wind turbine and most recently, a 250 kW prototype generator on test bed. However, there is a lack of an industrial size machine to prove the technology on real size. Further, engaging with the market is difficult without a real size prototype.
This project has studied several aspects of the Brushless DFIG including its design, operation, control and grid connection with an aim to optimize and improve its performance for multi-MW wind turbines. Further, its integration into a wind turbine drivetrain and optimization on a system-level has been carried out. A 3 MW medium-speed Brushless DFIG drivetrain system has been designed and its design, modelling and analysis is documented in a comprehensive report which can be used by the SME partners during the exploitation phase which will follow the completion of the project.
Project Context and Objectives:
The generation of electrical power from wind energy is a proven technology. There is already a substantial installed capacity in a number of countries and new capacity is being continually added. Various designs of wind turbines have been proposed but horizontal axis, three bladed machines with a capacity from 600 kW to 3MW or above is the currently preferred option. In these machines a variety of architectures is used, including direct drive, in which the turbine is directly coupled to a low speed generator, and indirect drive, with the turbine coupled through a gearbox to a medium or high speed generator. There is only one large manufacturer of direct drive turbines and they have a significant installed capacity, especially in northern Europe. However, the majority of wind turbines in the world employ the indirect drive architecture. In these machines, generation is normally from a doubly-fed slip-ring induction generator (DFIG). As the gearbox gives an increased shaft speed, a four or six pole machine can be used. The stator of the generator is connected directly to the fixed voltage, fixed frequency electricity grid and the rotor is supplied through the slip rings with a variable voltage and frequency supply generated from a converter.
This form of double feed enables generation to take place over a range of turbine speeds and provided that the range of speeds is moderate, the converter rating need only be a fraction of the total generator output thereby keeping the system cost low yet retaining a reasonable level of reliability. The converter rating is generally about one third of the rating of the generator, allowing speed variations of ±33%. Varying the voltage supplied to the rotor can be used to manage the reactive power flow from the generator. Control schemes have also been developed to enhance the response of the system to changing wind speed and to accommodate varying grid conditions.
However, there are drawbacks to the use of DFIGs, particularly the additional cost and bulk of a machine which incorporates slip rings and the need to maintain brush-gear and to replace brushes on a regular basis. Studies have shown that the reliability of large wind turbines is improving but that faults with generator and converter sub-assemblies contribute significantly to turbine failure rates. Further studies have shown that problems with brush-gear are significant in wind turbine operation and that the problem will be more severe in machines deployed offshore where additional wind resources are available.
Recent reliability studies have shown that the failure rate of wind turbine gearboxes is modest and in fact is considerably less than some other components of the wind turbine such as the power electronics, pitch control and the yaw system. On the other hand, the repair of the gearbox once it fails is significantly costly and in most cases, the gearbox is replaced. Therefore, improvement of gearbox reliability is of significant benefit to the wind industry. The high-speed stage of the 3-stage gearbox, which is commonly used in DFIG drivetrains, accounts for the highest number of failures and therefore eliminating this stage will substantially improve the gearbox reliability. A number of wind turbine manufacturers, such as Multibrid and Winwind have incorporated simplified 1 or 2 stage gearboxes in recent years, but all these drivetrains utilize permanent magnet generators.
Brushless DFIG drivetrain
The proposed novel and disruptive drivetrain technology leverages the advantages offered by the widely used DFIG drivetrains, but also achieves an intrinsically enhanced reliability and lower capital and O&M costs by incorporating a Brushless DFIG in combination with a simplified 2-stage gearbox. The Brushless DFIG requires an identical size converter as that of the DFIG, which is of fractional size, typically a third of the generator rating. Eliminating the slip-ring and brush-gear system in the generator and the high-speed stage of the gearbox, which are reported as the highest cause of failures in these components, enables achieving a significantly improved reliability. Further, the Brushless DFIG has shown to have an intrinsically improved low-voltage ride through (LVRT) performance during grid faults, which removes the need to have complex additional hardware in the converter, such as a crowbar. In summary, the Brushless DFIG drivetrain has the following unique advantages:
• High reliability due to:
- elimination of slip-rings and brush-gear in the generator which is known as the highest failure rate component in the generator;
- elimination of high-speed stage of the gearbox which is known as the highest failure rate part in the gearbox;
• Low capital costs due to:
- elimination of slip-rings, brush-gear and carbon extraction system in the generator;
- utilization of a simplified 2-stage gearbox;
- utilization of a fractionally rated converter;
- elimination/simplification of the crowbar design used for LVRT protection;
• Low O&M costs due to:
- eliminating the need for regular maintenance and replacement of brush-gear in the generator;
- higher reliability of the components, reducing the unplanned maintenance costs;
• Improved grid compatibility due to:
- an intrinsically better LVRT performance of the Brushless DFIG because it has higher rotor inductance by design;
• enhanced reactive power injection capability of the Brushless DFIG during grid faults.
Modern Brushless DFIG
The modern Brushless DFIG concept comprises two electrically separate stator windings, one (i.e. the power winding) connected directly to the mains and the other (i.e. the control winding) supplied from a variable voltage, variable frequency inverter, as shown in Figure 3. The design of the Brushless DFIG is not straightforward as there are more variables to consider than in a conventional induction machine. Nevertheless, attempts have been made to manufacture large Brushless DFIGs, beginning with the 75 kW machine of Runcos et al. and more recently Chinese designs of 200 kW. WTL’s work on a 20 kW Brushless DFIG has confirmed that a Brushless DFIG can be designed to a specification and can work in a wind turbine – an important part of this was implementing effective control of the machine, which is not stable in the open loop over the normally used speed range. Over more than six years’ operation has given confidence in the proposition of a large scale Brushless DFIG; the company has recently designed and tested a 250 kW Brushless DFIG system which was conceived as a stepping-stone towards a multi-megawatt scale machine.
Windrive Project
In order to make the Brushless DFIG attractive for commercial applications, its performance must be optimized to commercially acceptable levels and be demonstrated on real size (i.e. multi-MW size). Further, its mechanical and electrical integration within the wind turbine drivetrain and optimization of the system as a whole must be carried out. The aim of WINDRIVE project is to address these needs and to industrialize the design, construction, integration and operation of the Brushless DFIG in a 3 MW wind turbine drivetrain and achieve an optimized solution attractive to the industry.
The project benefits from the experience and know-how gained in the design and testing of previous smaller size prototypes, including the 20 and 250 kW systems. In the course of the project, several aspects of the new drivetrain technology have been studied and optimized by the RTD performers with inputs from the SME partners. A 3 MW drivetrain system have been designed and modelled, which is the first of its kind. The design will be utilized by the SME partners in the follow-on exploitation phase to build and test a prototype system. The project has nine distinct objectives:
Objective 1: Achieve an overall optimization of the Brushless DFIG design with respect to its manufacturing costs, size, weight and performance.
Objective 2: Achieve optimized control and operation of the Brushless DFIG within a wind turbine.
Objective 3: Establish a grid-friendly system design compatible with recent grid codes of EU countries.
Objective 4: Establish industrial low-cost, reliable mechanical and electrical integration of the Brushless DFIG within a wind turbine with a customized/improved components.
Objective 5: Establish industrialized design, construction and operation of the Brushless DFIG drivetrain.
Objective 6: Achieve a quantified assessment of reliability and economics of the Brushless DFIG drivetrain.
Objective 7: Establish a complete design of a 3 MW medium-speed Brushless DFIG drivetrain, ready to be built and tested during the follow-on exploitation phase.
Objective 8: Establish a portfolio of IP in the form of patents and knowhow related to the design and operation of the Brushless DFIG drivetrain, to be exploited by the SME partners.
Objective 9: Disseminate the project’s finding across the wind generation sector in the EU in order to facilitate the wider societal impacts offered by this technology.
Project Results:
1. Brushless DFIG modelling tools:
Several modelling tools including the Finite Element (FE), coupled-circuit, d-q and equivalent circuit models for the Brushless DFIG (also known as the BDFM) have been developed to study the electric and magnetic circuits in the machine. The predictions from the FE model have been compared with analytical models and the results obtained from experiments, which confirmed the accuracy of the modelling tools.
An iron loss model has been developed for the machine based on conventional techniques used for AC machines. The model, implemented in FE, calculates the iron losses at each time step for each element of the iron circuit. Preliminary core loss measurements have been carried out on a laboratory machine located in UCAM using voltage and current measurements.
A lumped parameter thermal model for the Brushless DFIG has also been developed for determining the machine rating and to study machine design aspects related to the electric loading of the machine. The model is used to determine the thermal coupling between the stator windings and the rotor heating in simulation.
2. Brushless DFIG design practicality:
A design procedure for the Brushless DFIG has been developed based on equations derived from a simplified equivalent circuit. The method allows the many variables in the design of this machine to be handled in straightforward way. Relationships are given for the division of slot area between the two stator windings and for the design of the magnetic circuit. The design method is applied to a frame size 180 machine. In particular, calculated values for flux densities in the machine have been verified by time stepping finite element analysis for actual operating conditions. Practical limitations in the construction of the BDFM are also assessed.
The design procedure for the BDFM starts with the choice of pole pair numbers to set the speed. Normal balanced three-phase windings are used for the stator windings, and these can be short-pitched. The rotor should have a turns ratio close to optimum and the winding must be insulated to give determinate current paths. Equations are derived from the machines equivalent circuit to give an initial design. This can be refined by an iterative procedure which keeps the machine within electric and magnetic loading limits over a specified operating range. The approach is illustrated with results from a frame size 180 BDFM. However, other methods will be needed for a detailed assessment of harmonics and the checking of the magnetic circuit to prevent saturation.
Design optimization to improve the BDFM PW power factor is carried out. It is shown that for smaller machines, leading power factors are difficult to achieve unless the magnetizing burden from the PW is significantly reduced by increasing the converter rating or with capacitor banks. However, as the rating of the BDFM increases to the megawatt scale, achieving a leading power factor becomes easier.
A comparison of different pole pair machines for optimal design of the power factor, torque performance and back iron thickness is carried out. It is observed that for lower pole pair count machines with their poles further apart, better performance (power factor and torque) can be attained, however this comes at a cost of increased weight since more back-iron is required.
3. Design methodology for multi-MW Brushless DFIG:
A design methodology has been developed for the megawatt Brushless DFIG for the wind turbine drivetrain application. The methodology consists of analytical and Finite Element (FE) models, for determining the Brushless DFIG performance and an optimization algorithm for geometry optimization. Based on the presented design methodology, a Brushless DFIG optimization design program has been developed, using MATLAB script and functions. For Finite Element (FE) calculations Comsol software is used. All FE models and calculations are set-up using MATLAB script and are executed through a MATLAB-Comsol interface connection. Therefore, the whole design optimization process is executed through MATLAB script.
The design flowchart incorporates the relations of the used input requirements, models and algorithms in order to come to an optimized Brushless DFIG prototype design that is suitable for industrialization. The optimization program combines the use of several models, such as the geometric model, magnetic-field model and the thermal model, to analyze an input machine design. The optimization algorithm then optimizes the machine design according the provided optimization criterion.
4. Rotor design methodology:
A procedure for obtaining rotor parameters for BDFMs with nested loop rotors has been established. Winding factors are used to give parameters valid for one operating condition but in fact there is little dependence on speed. Parameters determined from the analytical winding factor method are close to those determined experimentally. The ability to calculate rotor parameters accurately enables the exploration of optimal rotor designs and overall machine design. Although this study has looked at the nested-loop type rotor, the methods can be readily extended to other arrangements, including loops of arbitrary pitch and placement.
The ability to accurately calculate rotor parameters analytically enables research into optimal rotor designs. Optimization of the rotor aims to achieve a near optimum turns ratio, while keeping rotor leakage and resistance within a desired range. A low rotor leakage value may be beneficial because it increases the pull-out torque of the machine, however a higher rotor leakage will lead to better Low-Voltage Ride Through (LVRT) performance. A low rotor resistance will lead to lower rotor power losses. It is therefore important to have the ability to tailor rotor parameters in response to design specifications.
5. Rotor design optimisation:
The rotor teeth and core back sections of electrical machines are sized according to the expected peak flux density. Since the peak flux density of the rotor teeth and core back sections in the BDFM are not equal, the size of the teeth and core back sections should similarly not.
Analytical methods have been developed to determine the location of the largest and smallest peak flux density in the rotor teeth and core back sections and therefore the geometry of the rotor can be optimized. However, this field behavior cannot be exploited in the stator because the peak flux density value is different in each teeth and core back section at specific operating speeds. This will be particularly useful for the nested loop rotor because the magnitude of the current in the different loops of the nested loop rotor are not equal - the outer loop carries the greatest current and the inner loop the smallest. It has therefore been recommended to size the rotor bars according to the expected current for each loop.
The peak flux density in the back iron is similarly not equal, with the largest peak flux density found between the innermost rotor loops and the smallest found between the outer loops. It is therefore recommended to optimize the depth of the back iron (non uniform thickness) based on the peak flux density.
6. Design of alternative rotors:
The stator of a BDFM has two standard three-phase windings that are placed one on top of
the other in the same slots. The windings having different pole numbers chosen such that the winding factor of one winding is zero when analyzed for the fundamental harmonic of the other winding. This will ensure that there is no direct coupling. The rotor requires a winding which is capable of coupling with both stator windings.
It was shown by Browadway and Burbridge that a cage rotor of p1 + p2 bars (where p1 and p2 are the pole-pairs of the power and control winding respectively) performs the required linking function. It was further showed that a rotor with p1 + p2 equally spaced loops will also work regardless of the span of the loops, and thus proposed the 'nested loop rotor' made up of sets on concentric loops, with or without an enclosing cage. Subsequent to its discovery many nested-loop rotors have been built by other research groups, some with and some without the enclosing cage, but no particular analysis offered as to which has the best performance.
Two simple BDFM rotors have been studied including a cage rotor with p1 + p2 bars, and a loop rotor with p1 + p2 loops, drawing from analysis of the rotor structure and its equivalent circuit parameters. Five different rotor configurations were built for the prototype 160 frame size BDFM and their parameters were measured and compared.
Analysis of the simplest BDFM rotors suggest that increasing the loop span results in a lower referred resistance, however minimum leakage inductance occurred at a span smaller than fully pitched. A comparison between the cage rotor and the full span simple loop rotor shows that even though the two options filled the same slots, the cage rotor had considerably lower referred resistance and slot leakage inductance. The trends from the simple analysis apply directly to the five experimental rotors. The impedance of the series wound rotor is higher than that of the nested loop because current is forced down narrow loop spans. Design studies were performed to find the optimum number of series coils for a progressive loop rotor and loops per nest for a nested loop rotor. In each case it appeared that 'the more loops the better', up to the point where manufacturing the rotor becomes prohibitively expensive. Switching from a nested loop to a cage+loop offers a dramatic reduction in impedance (especially resistance) because the rotor will have the largest possible loop spans and because of the phasor addition of the currents in the bars. This suggests that for BDFM rotor designs the cage+loop format is likely to offer the best performance, as would be expected from initial comparison of the simple loop and cage rotors. Finally it was shown that some improvement in leakage inductance and resistance may be had by optimizing the span of the rotor loops. Optimizing the loop span further improves impedance by reducing the harmonics in the Magnetomotive Force (MMF) waveform. Though these changes move the ratio further from optimum, it has little impact on machine rating.
7. Grid connection for Brushless DFIG:
A robust controller for the Brushless DFIG has been developed which is essentially a power winding flux orientated vector control scheme without cross-coupling compensation. Theoretical analysis has been used to represent the decoupled control scheme for real and reactive power control. A Brushless DFIG model and controller have been implemented in MATLAB/Simulink and analyzed through simulations. The proposed control scheme successfully controls the static and dynamic demands of real and reactive power. An LVRT fault was also simulated on the model, and the results show that the proposed controller can successfully ride through the fault, without additional electronics, and inject the required reactive power. Future research should examine the effects of cross-coupling on the Brushless DFIG and to include a cross-coupling compensator into the controller.
8. Grid synchronization for Brushless DFIG:
At start-up the Brushless DFIG is required to synchronize to the grid so that the power can be extracted from the turbine. Control winding (CW) excitation is regulated according to the grid voltage and rotor speed to ensure the power winding (PW) voltage, phase and frequency matches the grids. Post synchronization, the control strategy is switched to a power control loop where the real and reactive power of the generator is controlled.
The developed grid synchronization model is implemented on the D180 frame size Cambridge Brushless DFIG. Initially the PW is not connected to the grid (through a contactor) and the rotor speed is ramped up. When the rotor speed reaches a speed of 450 rpm, a voltage is applied to the CW which excites the open-circuit PW. A closed loop controller uses a phase-lock-loop to ensure the amplitude, frequency and phase of the PW voltage matches the grid, this is done using the controller. When the two voltages are well matched (a small error), the contactor is closed and the PW is connected to the grid. Once the rotor speed reaches 450 rpm, the complete synchronization process takes 0.36 s. After grid synchronization, a real and reactive power control loop is switched in to control the output power of the generator.
For maintenance on the wind turbine, it is beneficial if the rotor of the Brushless DFIG can be controlled at a slow speed. For this mode of operating the Brushless DFIG should be connected in simple induction mode where the CW is connected to the grid through the frequency converter and the PW is open-circuit (through a contactor). A simple V/f controller is implemented where the amplitude and frequency of the voltage applied to the CW is adjusted.
It was shown through experiments that the speed can be controlled right done to 5 rpm. In this mode of operation the machine is self-starting however the speed cannot be controlled down to zero-speed. If this mode of operation shows commercial potential a closed-loop vector controller may be implemented, this will allow tighter speed control.
9. Sensorless control for the Brushless DFIG:
A control system for the Brushless DFIG including sensorless position estimator has been developed. In order to verify and show the working principles of the control system, first a time dynamic model of the complete system, including a Brushless DFIG machine model and the sensorless controller model (control algorithm including sensorless position estimator), were developed.
A Brushless DFIG model was developed using the concept of ideal rotating transformers (IRTF). To control the Brushless DFIG output torque in an efficient and stable manner, the developed Brushless DFIG control algorithm uses closed loop vector control of the CW current vector transformed to the CW flux coordinate system (dq‐reference frame). This enables fast and smooth control. In this way torque is proportional to the CW q‐axis current component, while the flow of reactive power in both PW and CW is influenced by the CW d-axis current component. To apply the reference transformation to the CW flux coordinate system the CW flux position angle must be obtained. This position angle is obtained using the sensorless position estimator, which derives the CW flux position angle from the measured CW flux (obtained from voltage and current measurements). By measuring the CW flux, the controller controls the CW voltage (and hence current) to follow the measured flux and hence the Brushless DFIG is kept in its synchronous mode of operation. There is no need to know the mechanical position angle at all. Though the mechanical position angle could easily be derived if required, by using the estimated PW and CW flux position angles.
To show the effectiveness of the developed controller including sensorless position estimator, its performance is simulated on a Brushless DFIG model of an existing D160 frame size machine. First the cascaded torque speed characteristics of the Brushless DFIG were determined, by simulating the Brushless DFIG time‐dynamic behaviour when connected direct online with one of the windings short circuited. The resulting torque response was compared to steady‐state calculations, showing that the Brushless DFIG model behaved as expected. Then the control system including Brushless DFIG controller with sensorless position estimator was simulated. The system response to a step input on the CW d‐axis and on the CW q‐axis current components was simulated, showing that the torque is proportional to the CW q‐axis current component and that the flow of reactive power is proportional to the CW d‐axis current component. Then a speed ramp up from sub‐synchronous to super synchronous speed was simulated, showing that the controller is perfectly able to keep the Brushless DFIG in synchronous operation while the speed varies. It could also be observed that the torque collapses when the rotor frequency reaches 0 Hz. For stable control it is advised to keep away from this operating point. However, it should be noted that this operating point is beyond the speed range required for wind turbine drivetrain operation.
The developed models provided insight in the Brushless DFIG control algorithm and how to obtain sensorless position estimation. The simulations showed the effectiveness of the developed control algorithm including sensorless position estimator. The control algorithm including sensorless position estimator was also implemented in hardware successfully.
10. Optimisation of control system:
A stability model has been developed for the wind turbine drivetrain (WTDT) incorporating a Brushless DFIG. Using the developed WTDT model and control algorithm, the dynamic behaviour of the WTDT to a step change on the input is simulated. The stability of operation is key in electrical generators, on the other hand, improving the dynamics would typically trades off with stability margins. A steady state model is developed which is used to analyse the Brushless DFIG stability.
Two cascaded control loops are used to control active‐ and reactive power independently. Cross compensation has been included to limit cross coupling between control loops. Tuning of PI controllers is based on machine parameters and offer good performance during variation of parameters during operation. The machine controller offers good dynamic behavior to a change in wind speed and is able to maintain optimum power extraction.
11. Condition monitoring for the Brushless DFIG:
To improve WT availability and reduce O&M costs, two condition monitoring techniques have been developed based on analyzing the generator power signal using Continuous wavelet transform (CWT). In comparison with conventional vibration, temperature measurements, and lubrication oil analysis techniques currently been used in industry, the developed techniques have the advantages of reduced capital cost, ability to detect both electrical and mechanical faults, and can be applied to both geared and direct drive WT's with different generator topologies.
1. The new CWT-based condition monitoring algorithm identifies the faults, their associated frequencies and their duration, however conventional approaches also pick up other frequencies associated with normal operation of the wind turbine, and the algorithms take considerable time to execute.
2. The new CWT-based energy tracking method identifies fault perturbations applied to the generator shaft. This method considerably reduces the calculation time, however some fault signals were lost when wind turbulence was included. With more refined signal processing, this technique does have potential and should be further researched.
The two approaches now need to be applied to the power signals obtained from real wind turbine’s during mechanical and/or electrical faults to determine the detectability of the techniques.
12. Definition of drivetrain topologies:
The technical parameters of several wind turbine drivetrain topologies have been defined. These specifications are used in mathematical modelling and simulation work to support the development of a new Brushless Double-Fed Induction Generator turbine. The results of this modelling have been used to influence design decisions in all aspects of the new turbine (structural, mechanical, electrical etc.) so as to verify and optimize the design.
Three turbine specifications were proposed, covering a range of existing generator technologies and transmission types. The designs are derived from existing, commercially available wind turbine products, and are intended to be comparable in terms of operational environment and power output.
13. Stochastic energy analysis:
The aim of this work is to develop mathematical models of the dynamical properties of wind turbine drivetrains, and tools to analyse them so that comparisons may be made between competing designs. Particularly of interest are the vibrational response characteristics of the drivetrains. Unanticipated vibration of any mechanical system has implications for component lifetimes (due to frictional wearing), and radiated audible noise (which is undesirable in a turbine). The desired outcomes of this work are a thorough written analysis of the proposed design options, and a set of software tools to aid in the design process. These results and tools will then aid in the development of a complete turbine design for the new Brushless DFIG drivetrain.
Significant work has been undertaken to develop three wind turbine drivetrain models and analysis tools necessary for the project. These have been implemented as a series of modular MATLAB scripts that produce mass and stiffness matrices representing the masses and stiffnesses of the drivetrain components. Linear Vibration theory is applied to these system matrices, resulting in lists of resonant frequencies and mode shapes. These results can then be used to predict further characteristics such as frequency transfer functions and transient responses. Additionally, software has been written to visualize the modal results from these scripts, allowing fast and intuitive analysis and verification of the results. This information is crucial for predicting problematic vibrations, and will influence the design of other aspects of the turbine.
The stochastic energy analysis enables us to quickly finalize the three drivetrain models with realistic design parameters, and provide analysis of their relative dynamical characteristics. The outcomes of the analysis has facilitated rapid iteration of design process, and optimization of the candidate BDFIG design. In particular, the torque-ripple data from simulation of the generator have been used for comparison. Further work will focus on adding more complexity to the modelling (particularly in the areas of structural coupling and electromagnetic effects within the generator) as necessary to match the frequency ranges of interest.
14. Brushless DFIG drivetrain efficiency and cost modelling:
A wind turbine drivetrain (WTDR) incorporating the Brushless DFIG efficiency and cost model has been developed. The model includes efficiency models for each WTDT system component, as summarized below:
• Wind availability model.
• Wind turbine rotor efficiency model.
• Gearbox efficiency model.
• Brushless DFIG efficiency model.
• PE converter efficiency model.
Additionally a system performance model for the WTDT as a whole has been developed. Further, a simple capital costs (CAPEX) model including all WTDT system components is developed. The resulting models enable the analysis of a complete WTDT performance over its full steady-state operating area. The resulting models are useful for WTDT system comparison and speed-range optimization analysis.
15. Drivetrain speed range optimization:
The generator speed-range and rated speed will vary depending on the WTDT gearbox ratio and generator operating frequencies. This work studied those trends for different generator systems in a WTDT. The results can be used for speed-range optimization, by selecting the most favorable combination of system components.
A 3.2MW wind turbine drive train has been analyzed as a case study. Different generator systems are compared over a wide speed range. For this purpose 10 PMG systems, 8 DFIG systems and 3 Brushless DFIG systems have been analyzed. Each system with different gearbox ratio and rated generator speed. Resulting trends provide the following insight:
• Generator CAPEX getting cheaper with increased ratio.
• Gearbox CAPEX getting expensive with increased ratio.
• For three stage gearbox systems, the total system CAPEX always decreases with increasing gearbox ratios.
A medium-speed 2-stage gearbox solution is favoured for the further development of a multi megawatt Brushless DFIG WTDT. The lowest total capital cost and highest energy yield per euro of capital cost was obtained for a 2-stage gearbox systems with corresponding generator speed in the range of 200 to 300 RPM, but the speed is also related to the wind turbine rotor rated speed. A medium-speed generator solution is also favoured over a high-speed solution, because of historical reliability problems of high-speed 3-stage gearboxes.
In order to provide a more thorough generator system comparison, future research work should focus on:
• System performance evaluation of Brushless DFIG machines.
• The Brushless DFIG cost trend as function of gearbox ratio should be extended over the whole gearbox ratio range.
• Multi-objective design optimization, to provide a better trade-of between generator cost and generator efficiency, which has impact on the system performance and therefore on the energy yield per euro of capital cost.
16. Drivetrain simulation under grid events:
Wind turbines connected to the grid must achieve good grid quality and voltage ride through. The stator windings of a Brushless DFIG and DFIG are connected directly to the grid. In the event of a voltage dip on the grid the turbine must remain connected and not shut down. However, if a voltage dip is seen in the generator this has an effect on the generator torque which reacts the mechanical torque. Effects that are studied included the combined effect of controller algorithms and grid connection on the reaction current and torque response of the different generator systems in a wind turbine drivetrain. A comparative study has been carried out. The generated torque profiles for each generator system enable understanding of grid events and controller dynamics on mechanical component response and reliability. The knowledge gathered will facilitate an understanding of the robustness of the Brushless DFIG in comparison with PMGs and DFIGs. The investigation included the protection of the power electronics during voltage events and the reaction of the generator in these situations.
It has been found that for the PMG based wind turbine, the issue with Low Voltage Ride Through (LVRT) is the rise in the DC link voltage. This is due to the mismatch in power generated by the machine and the power transferred to the grid (which is limited by the reduced voltage at the grid side converter terminals). Methods for controlling this DC voltage rise, such as the use of an energy discharge circuit, have also been investigated. The method Power Balancing was proposed, where the power generated by the machine is controlled to match the power transferred to the grid in the event of a low voltage event. This method has also been used to control the issue of DC link voltage rise, however, it introduces torque disturbances in the machine.
For the DFIG it has been found that voltage dips cause large transient currents in the stator and rotor circuit. The over‐currents in the rotor circuit are a cause for concern as they may lead to adverse effects on the power electronic converter connected to the circuit. A method to overcome this issue i.e. the use of a crowbar circuit, has been studied. The crowbar circuit manages the problem through bypassing the power electronic converter in the event of current rise.
In the case of the Brushless DFIG, it has been found that again transient currents are present in the power and control windings in the event of a voltage dip. However, the magnitude of these currents is lower when compared to that in a DFIG. This is attributed to the higher leakage inductance of the Brushless DFIG. A crowbar‐less method has been studied which is successful in controlling the high currents generated in the control winding. The crowbar‐less control method also reduces the torque oscillations observed in the Brushless DFIG for a low voltage event.
In conclusion, apart from offering better reliability through the exclusion of slip rings and brushes, the Brushless DFIG also has an improved LVRT performance when compared with the DFIG. The protection is simpler and does not require an external circuit, like the crowbar, and can be built into the control algorithm of the machine controller.
Also, in this study the complete reactive power interaction is considered from the power winding of the machine. In reality, there is a possibility to use the grid side converter for the transfer of reactive power to and from the grid as well as the generator windings. This aspect and its effect on the machine response should be looked into in future.
17. Generator torque modelling and response:
The torque and Unbalanced Magnetic Pull (UMP) profiles, due to the electromagnetic fields in the Brushless DFIG, are important input characteristics for the WTDT mechanical structure models. Large torque ripples and UMP could negatively affect component life‐time and contribute to additional noise and losses.
Brushless DFIG Torque and UMP responses have been derived for a 3.2 MW case‐study machine. Based on the specifications of the model 3.2 MW wind turbine, the torque and UMP response profiles were determined under normal (nominal) operating conditions. Using FE modelling and Maxwell’s stress tensor method the torque and UMP response could be calculated.
Then using space and time Fourier analyses, both the torque response and UMP response have been analyzed. It is found that mainly winding space‐harmonics and slotting harmonics are responsible for torque ripple. The amplitudes, frequencies and origin of these space‐harmonics have been determined, in order to recreate the torque profile. The UMP response showed little unbalance, since the pole‐pair combination of the machine was chosen to cancel out any UMP. A space-harmonic analysis has been performed to determine the stator and rotor eccentricity modes. Time-harmonic analysis has shown that there is some fluctuation in the amplitudes of the eccentricity modes over time. This can mainly be attributed to winding and slotting harmonics.
18. Control hardware industrialisation:
The hardware design plays an important role in the performance and reliability of wind turbines in industrial environments. Suitable hardware designs for the control and instrumentation systems of the Brushless DFIG have been developed. The designs have utilized the input and output parameters required to control the Brushless DFIG using the Speedgoat real-time target machine. Input and output voltage signals from Speedgoat are required to be converted to the correct voltage level as well as buffered to protect the device, this is done with a signal conditioning board with custom built printed circuit board.
For a multi-MW Brushless DFIG designed for wind turbine applications it is envisioned that the Educational real-time target machine will be invaluable during the testing process, however an embedded controller would probably be used for commercial applications.
19. Control software industrialisation:
The industrialization of software is crucial in protecting the IP related to the control algorithms. The Simulink model is compiled into a binary file (.dlm) on the host PC before it is transferred to the Speedgoat target machine. Therefore the model is not visible on the target machine. To protect or hide the Simulink model from the end user on the Host PC, a graphical user interface (GUI) has been built, where parameters (eg. Real and Reactive power reference signals) can be passed from the GUI to the Simulink model without knowing the actual parameter names. The Simulink model with the important intellectual property (IP) for controlling the Brushless DFIG can then be password protected from the end-user.
MATLAB provides the GUIDE (graphical user interface design environment) toolbox for designing user interfaces for custom applications. The GUIDE Layout Editor allows the user to graphically design his/her own interface. GUIDE then automatically generates the MATLAB code and the user can modify this code to program the behaviour of the specific application. The GUI will be composed of an m-file and a .fig file, and will also reference the MATLAB/Simulink binary file with the .dlm suffix.
Currently the Speedgoat xPc real-time target machine connected to a Host PC is used to control the Brushless DFIG. This setup allows the user to easily adjust parameters on a MATLAB/Simulink model running on the Host PC , which can be executed in real-time via the Speedgoat device. This control setup allows rapid prototyping of the controller - the user can quickly and easily modify the Simulink model. This is useful for testing the Brushless DFIG, however it is believed that in the future, an embedded controller will be used.
This embedded system will be based on a microcontroller, programmed to handle the specific task of controlling the Brushless DFIG. Since the embedded system is dedicated to the specific task, the designer will be able to optimize the module to reduce the size and cost of the product, and if mass-produced will benefit from cost savings through economies of scale. This embedded controller will need to have sufficient analog and digital inputs and outputs and sufficient CPU speed to be able to compute the control loop.
20. Quality control management for Brushless DFIG manufacture:
Unbalanced magnetic pull (UMP) increases eccentricity in the Brushless Doubly-Fed Machine (BDFM) and may cause serious damage to the machine or drivetrain. In addition, UMP acts as a major source of vibration and noise. In order to build a BDFM generator successfully, methods to reduce UMP need to be studied.
This work focuses on the UMP in the BDFM, caused by both static and dynamic rotor eccentricities. Several parallel winding designs for the two stator windings are proposed and the practicality of such designs is discussed with respect to direct coupling between the stator windings and with rotor undesirable harmonic fields. Once practical parallel winding designs are established, their effects on reducing deflection as a result of static and dynamic eccentricities are shown and compared with series wound stator.
The study of direct coupling is particularly important when short pitched windings are utilized. It has been shown that with appropriate connection of coil groups in series and parallel, direct coupling of stator windings can be removed. There may be several possibilities of practical parallel winding designs, in which case the design with more parallel paths is shown to have stronger effects in suppressing the UMP.
Larger scale BDFMs are likely to be designed for slower natural speeds, hence with higher stator pole numbers. This will provide more possibilities for parallel connection of stator coils, but the practicality of such designs must be carefully assessed. The study carried out in this work, though is shown for a specific BDFM with 4 and 8 pole stator windings and a nested loop rotor, can be generalized for other BDFM designs.
21. Development of reliability model:
A reliability model for the drivetrain of a geared wind turbine fitted with a doubly fed induction generator (DFIG) and the novel Brushless DFIG has been. The model is derived from considering the failure rates of the main electrical assemblies namely the generator and converter as well as the gearbox.
The analysis of the reliability model suggests that the reliability and reparability of a Brushless DFIG, partially rated converter and 2-stage gearbox is better than the conventional wind turbine drivetrain comprising of a DFIG, partially rated converter and 3-stage gearbox. The main reason for the lower reliability of the system with the DFIG is the presence of the brushes in the configuration and the additional stage in the gearbox (high-speed parallel).
Future work should include field data from actual wind turbines of failure rates of gearboxes. It would also be interesting to compare permanent magnet generators with DFIG's and Brushless DFIGS and to compare direct drive with geared wind turbine systems. It would also be useful to compare the reliability and availability of wind turbines (and farms) across geographical regions as well as on and offshore turbines.
22. 3 MW Brushless DFIG and controller design:
A design for a 3.45 MW Brushless DFIG, its associated converter and controller has been developed. The system has been designed to match the output power, speed and voltage requirements for the Senvion 3.2M114 wind turbine, with the high speed 3-stage gearbox being replaced with a medium speed 2-stage gearbox. Simulated efficiency, power factor and heat tests were conducted on the proposed design, to confirm the generator will produce rated power at an adequate efficiency and power factor. The design was simulated in MATLAB/Simulink to test the dynamic response of the system by applying step changes in real and reactive power as well as simulating LVRT grid faults. The simulations showed that the controller can adequately control the 3.45MW Brushless DFIG.
23. Design of grid connection and condition monitoring:
This work deals with the electrical systems required to connect a wind turbine fitted with a Brushless DFIG to the grid. The different electrical systems required for grid connection and a start-up and shut-down procedure for the wind turbine have been developed. Fault detection and how the system responds to these faults is also discussed as well as condition monitoring alerts. The electrical systems comprised of the Machine-Side Inverter (MSI) and Grid-Side Inverter (GSI), contactors and wind turbine controllers. Fault conditions were defined which will raise a fault flag to the turbine controller and shut the wind turbine down in a safe way. The condition monitoring system previously developed is also incorporated into the system. Finally a start up and shut down procedure for the Brushless DFIG wind turbine was created.
The converter rating for the Brushless DFIG wind turbine with respect to reactive power management and grid LVRT requirements has been studied. It has been shown that at a given speed, taking into account both steady-state and transient operating conditions, there is a minimum sum of the MSI and GSI ratings but this condition may require an undesirably high control winding voltage so the minimum rating may not be practically achievable. Furthermore, it is shown that the MSI rating is typically significantly larger than the GSI rating because the MSI contributes to the supply of machine’s magnetizing currents and must handle large transient currents for crowbar-less LVRT operation.
The effects of a capacitor bank in reducing the rating of the GSI is also discussed. It is shown that with the use of a suitably rated capacitor bank, the minimum sum of the MSI and GSI ratings can be achieved at rated CW excitation voltage.
The effects of the machine’s leakage inductance on MSI rating and efficiency has also been studied. A higher leakage inductance limits the current during LVRT conditions but comes at the price of a restricted power factor range without de-rating the system. The effect of leakage inductance on machine losses was also investigated and these can rise significantly when operating at leading power factors. However, the problem becomes less critical for larger BDFMs, i.e. MW scale machines, since the per unit value of the rotor reactance will drop with size. The use of a capacitor banks is therefore more valuable with smaller machines unless the operating speed range above the natural speed can be limited to around 20%.
24. Brushless DFIG controller implementation:
The hardware and software implementation of a controller for the Brushless DFIG has been studied. For the development of the practical Brushless DFIG controller a test set‐up was built. The test set‐up featured a D160 frame size prototype Brushless DFIG and a specially developed controller to control the Brushless DFIG in an efficient and stable manner. The controller was built using an old converter, equipped with a new Digital Signals Processing (DSP) and measurement circuitry. The control algorithm with sensorless position estimator, as developed earlier in the project, was programmed into the DSP.
The result is a functioning control system for the Brushless DFIG. Measurements were performed demonstrating the control system performance. The measurements showed that the control system for the Brushless DFIG is able to control the machine in a stable manner over its complete speed range. The used D160 frame size Brushless DFIG in the test set‐up provided some difficulties, because it suffers from excessive slotting induced time-harmonics. This caused the sensorless position estimator to lose the position around the natural speed, which resulted in the loss of synchronization. By implementing an additional notch ripple filter the slotting induced time‐harmonic frequency was filtered out. Then, the controller worked fine and was able to keep the Brushless DFIG in stable operation even around the natural speed. The control system for the Brushless DFIG developed in this project, features the world’s first practical and fully functional control system with sensorless position estimator for the Brushless DFIG.
25. Comparison of drivetrain dynamic response:
The purpose of this work is to investigate predicted force and deflection spectra of components within the three candidate drivetrains under investigation. It is known that different wind turbine designs can exhibit very different dynamic characteristics, and so the comparisons made between these three technologies will enable valuable insight into how the different drivetrain options behave.
The vibrational characteristics of three contrasting wind turbine drivetrain designs, containing a range of different generator types (Doubly-Fed Induction Generator, Brushless Doubly-Fed Induction Generator, and Permanent Magnet Generator) have been analyzed. Parameter sensitivity studies have been carried out on all three drivetrains which showed them to be relatively robust to changes in both mounting stiffness and approximated magnetic stiffness.
For the two drivetrain designs containing gearboxes (Drivetrain 1 and Drivetrain 2), excitation sources internal to the same have been identified, and compared with the predicted frequency response functions of their respective generators. This analysis shows that there is a real risk of transmission error-induced vibration affecting the generator, potentially leading to complex interactions with the electromagnetic properties of said generators.
Additionally for the novel BDFIG generator, simulated ripple-torques and unbalanced magnetic pull results were analyzed, and also compared with the frequency response functions. These electromagnetic-induced vibrations were also found to lie close to resonant frequencies of the system, which is a potential concern for gearbox longevity. It is recommended that future work designing a gearbox for the BDFIG generator takes these effects into account.
Analysis of the direct-drive Permanent Magnet machine was confounded by the small number of mechanical components. The resulting analysis results are therefore dominated by the turbine rotor blades and tower stiffnesses.
26. Design documentation:
Finally, all the tools developed for the design and modelling of the Brushless DFIG and its associated converter and control system have been documented so that they can be utilized by the SMEs during the follow-on exploitation phase. The software codes have been copied onto safe and password protected external memories and handed over to the SMEs. Notes have been provided in the codes/programs so that they can be easily understood and potentially modified by operators in future.
Potential Impact:
WINDRIVE supports the exploitation of a more reliable and cost-effective generator system for the wind industry. Improved reliability and cost-effectiveness will inherently lead to wider adoption of wind power generation, especially in offshore, which will boost the low-carbon economy, reduce CO2 emissions and increase energy security for the EU. Wind power development has been a top priority for the EU and Member States, and the targets for 2020 can only be achieved if wind power gains momentum without relying on substantial government subsidies. This can be realized through technological advancement to reduce the capital and O&M costs of wind turbines and enhance their availability through greater reliability. The Brushless DFIG technology offers a significant contribution to achieve this. Further, by eliminating brushes that are carbon-based, the new technology reduces their usage.
The exploitation of the Brushless DFIG drivetrain will support the reduction of CO2 emission, estimated as much as 224 million tonnes in 2022. The followings describe the calculation methodology which is based on the guidelines suggested by the Carbon Trust. The assumptions include:
1. The CO2 emission reduction factor is assumed to be 0.43 kgCO2/kWh for a wind turbine;
2. A wind turbine has a size of 3 MW with a lifespan of 20 years;
3. A capacity factor of 30% is assumed for a wind turbine, which represents the intermittent nature of the wind, the availability of the wind turbines and array losses;
4. A saving of EUR 475,000 can be achieved by utilizing a Brushless DFIG drivetrain (with a simplified 2-stage gearbox) instead of a DFIG drivetrain in an onshore wind turbine over its lifespan, estimated by a study carried out by an independent consultant employed by the Carbon Trust. The saving for an offshore wind turbine with a same size is assumed to be three times greater.
5. The capital cost of onshore and offshore wind farm development is assumed to be 1,150 EUR/kW and 2,500 EUR/kW respectively.
6. From points 4 and 5, with the cost savings achieved by employing the Brushless DFIG technology in 7 onshore 3 MW wind turbines or 5 offshore 3 MW wind turbines, an extra wind turbine can be deployed.
According to the figures published by EWEA, the total installed capacity of onshore and offshore wind power is expected to reach 210 GW and 50 GW by 2022, respectively. It is expected that between 2017 and 2022, during which the exploitation of Brushless DFIG drivetrain will occur, the EU will install 50 GW of new onshore and 30 GW of new offshore wind farms. We anticipate that 1.5 % of the new wind turbines installed between 2017 and 2020 will incorporate the Brushless DFIG drivetrain, which corresponds to 400 3 MW wind turbines, of which 250 turbines will be installed onshore, and hence 150 offshore, each rated at 3 MW. From point 6 above, the utilization of the Brushless DFIG architecture will therefore enable extra 36 onshore and 30 offshore wind turbines to be deployed. This is equivalent to an overall saving of EUR 347 million. The annual ‘extra’ CO2 emission reduction by deploying the extra wind turbines can be calculated as:
66(turbines) x 3000(kW) x 8760(days pa) x 0.3(capacity factor) x 0.43(kgCO2/kWh) = 223,745,038 tonnes.
Improving the strategic competitive position of the participant SMEs
The WINDRIVE project provides its SME participants with specific knowledge, technology and products that will enable them to further differentiate and strengthen their competitive position relative to global competitors. Strategic competitive developments will be achieved in the following key areas:
• Direct economic business growth through exploitation of the WINDRIVE outcomes, enabling further investment, expanding the SMEs’ skill base, technology portfolio and production capacity;
• New knowledge in the application of the SMEs technology, products and services; thereby enabling value-added services and deeper integration within end-user research and development activities;
• Protection of IPR and development of own innovative products; enabling expansion and further development of the SMEs activities towards higher value, sustainable and more profitable business growth (transition further up the supply chain and revenue generation through exploitation of IPR);
• Entrance or improved credibility within the wind turbine market.
Each SME has identified the specific competitive and economic benefits that will result from:
1. Their specific activities within the supply chain which will lead to increased business growth, increased profits and new employee recruitment;
2. Their IPR ownership for a range of new products which will enable the SME to expand its company product portfolio and/or license the technology for other applications;
3. The specific knowledge that they acquire as a result of the WINDRIVE project, thereby enabling deeper integration within supply chain R&D activities.
IPR benefits to the SME partners
Wind Technologies:
- New design features of the Brushless DFIG (patent/know-how);
- New rotor winding design of the Brushless DFIG (patent/know-how);
- New methods in minimizing the unbalance magnetic pull (know-how);
- New designs for cost-effective manufacturing (know-how);
- New methods for magnetic circuit design, analysis and optimization of Brushless DFIG (know-how);
- New optimization methods for core losses in the Brushless DFIG (know-how);
- New optimization methods of rotor performance (know-how);
- New LVRT control algorithms for the Brushless DFIG (patent/know-how);
- New encoderless control algorithms for the Brushless DFIG (patent/know-how).
Romax Technology:
- New design of integrated simplified gearbox for the drivetrain (patent/know-how);
- New optimized drivetrain design for the Brushless DFIG (patent/knowhow);
- New methods for improved dynamics of the drivetrain (patent/know-how);
- New modelling approach for the Brushless DFIG drivetrain (know-how);
- New reliability models for the Brushless DFIG drivetrain (know-how);
Mita:
- New control algorithms optimized for operation of Brushless DFIG drivetrains (know-how);
- New methods of maximizing stability and dynamics of the Brushless DFIG drivetrain (know-how);
Speedgoat:
- New cost-effective and reliable design of control hardware (know-how);
- New method for embedding of control software for protection of IP (know-how);
Improving the competitiveness of the products and services sold by the SMEs
The Brushless DFIG drivetrain is a disruptive technology which, once proven, can potentially address a substantial market opportunity addressed by the conventional DFIG technology. WINDRIVE is a stepping-stone in the commercialization of the new technology and the outcomes of the project give a significant competitive edge to the participating SMEs since the Brushless DFIG drivetrain is a totally innovative concept. WTL will benefit from the exploitation of the core technology, i.e. the generator and its control system, and the other SME partners who are part of the drivetrain supply chain will benefit from gaining the expertise of the other components which are specifically designed and optimized for the Brushless DFIG concept. WINDRIVE will therefore provide them with an immediate competitiveness over the other competitors since they will offer a solution specifically optimized for the Brushless DFIG application which is not available to others.
Societal Impacts
Given the crucial timeframe up to 2020 during which global emission must start to decline, the speed of deployment of wind farms is of key importance in combating climate change. Building a conventional power plant can take 10 or 12 years or more, and until it is completed, no power is being generated. Wind power deployment is measured in months, and a half completed wind farm is just a smaller power plant, starting to generate power and income as soon as the first turbines are connected to the grid. There are no fuel costs, no geo-political risk and no supply dependence on imported fuels from politically unstable regions. Every kilowatt/hour generated by wind power has the potential to displace fossil fuel imports, improving both security of supply and the national balance of payments. It seems evident that wind farm spreading have many legislation and policy impacts as well. It may strongly impact (and will impacted by) security policies, energy policies, renewable energy, climate, sustainable development, regional, environmental, aviation, economic and research policies and strategies.
By the end of the year 2010, about 670,000 persons were employed worldwide directly and indirectly in the various branches of the wind sector. Within five years, the number of jobs almost tripled, from 235,000 in 2005. There is an increasing demand for a very broad range of jobs, from engineers, skilled workers to mangers, financial, environmental and legal experts. WINDRIVE will lower the cost of wind power and hence will contribute to the growth of the wind industry and creating new jobs.
Environmental Impacts
In March 2004 Greenpeace published “SeaWind Europe”, which examines the ramifications –environmental, technical, social and economic – of offshore wind providing 30% of the EU’s electricity. While ambitious, it shows what could be possible with the vision and ambition to develop renewable energy on such a scale. Greenpeace says that “success on this scale would not only deliver enormous environmental benefits from this clean, safe energy source, but would also generate an economic boom in Europe worth hundreds of billions of Euros and create up to 3 million jobs.”
This statement was as encouraging then as it is now and the real potential of wind technology on the future stabilisation of power generation is only now really being shown to have wide benefits and implications in “keeping the lights on”. The balance, and trade off, of this technology has been, is being, and will continue to be argued by the “fors” and “againsts”, who will state the added benefits and those that voice their concerns as the environmental impact and possible damage that may or not be done. WINDRIVE does not want to be seen entering into this argument or taking a position. It is felt by the parties involved in this project that a more balanced and even approach to the entire issue is in the duty of care that any EU funded project needs to address issues relating to environmental impact and the regulatory standards that need to be adopted. In this regard, WINDRIVE will look to assist regulators in developing those standards and approvals and enhancing and mitigating any environmental potential impact that might arise. In the case of wind technology, both on-shore and off shore, the discussion on regulation and also standards is on-going.
Since 2001, EU Directive 77/2001/EC, which promotes electricity from renewable sources, which included an indicative target for 21% of the EU's overall electricity to come from renewable energy sources by 2010, gave an important boost to the sector, although just over 18% were actually reached by that deadline. At the end of 2010, the EU’s new Renewable Energy Directive (2009/28/EC) entered into force, setting an EU renewable energy target of at least 20% of final energy consumption by 2020. Each EU Member State has a national legally binding target for the share of renewable energy it must achieve by that date. The member states were required to submit National Renewable Energy Action Plans (NREAPs) detailing sectoral targets and measures necessary for them to reach their overall binding RES target. In their NREAPs, 15 member states say they will exceed their targets, ten consider they will meet their targets, and only two expect to not meet their target domestically. Consequently, aggregating the NREAPs, the EU is expected to exceed its 2020 20% target by 0.7%.
Electricity from renewables will play the largest role in meeting the 2020 target, covering over a third of total electricity consumption. With a capacity forecast of 213 GW, of which 43 GW will be offshore, producing 495 TWh of electricity covering about 14% of total consumption, wind comes out as the technology of choice in the NREAPs. The European Wind Energy Association (EWEA)’s baseline scenario assumes a total installed capacity of wind power in the EU by 2020 of 230 GW, producing 582 TWh of electricity, meeting 14.2% of EU electricity demand (depending on demand growth over that period). EWEA’s high scenario assumes that total installed wind power capacity will reach 265 GW by 2020, producing 681 TWh of electricity and meeting 16.7% of the EU’s electricity demand by 2020.
Transnational approach
The WINDRIVE consortium has been possible through a trans-national approach. The need for a wider strategy is increasing due to the increasing demand for innovative green technologies within Europe and globally. This trans-national approach combines the knowledge and experience from a committed consortium enabling delivery of this challenging project which will produce distinct advantages and environmental benefits when applied to wind farm generators. Only through the trans-national combination of these skills, can we increase competitiveness and establish a critical mass of experience within the participants who would otherwise remain a fragmented resource.
Within the supply chain, SMEs make up the large majority of companies offering innovative products. However, innovation within these companies is limited by the small R&D budget or lack of resources, both of which are common to SMEs. Indeed this is also true for the engineering SMEs who supply into this market. Routine customer specific projects leave little room for innovation and ability to develop partnerships on a pan- European level. WINDRIVE facilitates this process by establishing new technology and paving the way for a broader integration between partners. The WINDRIVE consortium brings together an expert core group of SMEs with research organisations who provide word class research and product differentiators.
Plan for the use and dissemination of foreground
Work package 8 was included within the project to concentrate specifically on the dissemination and exploitation of the project results. All project partners recognise and are strongly committed to maximising the dissemination and exploitation of the WINDRIVE results.
The impact of the WINDRIVE project on the state of the art in the field of wind power generation will rely on wide and effective dissemination of the results gathered. WINDRIVE have applied efforts to convey its knowledge and technologies to relevant industrial and academic communities as well as the general public. The WINDRIVE consortium concentrates the expertise of several adjacent and complementary research fields and thus is very well placed to stimulate new methodological paradigms within the partners.
Effective communication of the results of the cross-disciplinary investigations and their promotion to specialists from diverse backgrounds will maximise the project’s impact. The dissemination activities lay the basis for successful uptake and commercial exploitation of the project’s technological output. Dissemination strategies which have been carried out during the WINDRIVE project included:
1. Development of an interactive project website;
2. Development of a Wikipedia page;
3. Development of a video clip and its publication on YouTube;
4. Publications of non-confidential research outcomes in international journals and conferences;
5. Presentation at international conferences and exhibitions and engagement with industrial experts at such events.
6. One-day workshop for invited attendees from a wide range of audience including industrial and academic experts and participants from government and NGOs.
The above will be described in detail in the next section.
Main Dissemination Activities
1. Project website
The project website has been developed to engage with public about the outcomes of WINDRIVE and can be viewed at www.bdfig.com. The name of the website is the acronym for the main technology that WINDRIVE researches, the Brushless Doubly-Fed Induction Generator (BDFIG). The website includes the brief summary of the project, detailed description of the project and its objectives, publication list, information about the consortium partners, latest news and contact details. Any contact made via the website will be forward to the Coordinator who will respond or refer the message to the relevant partner.
2. Project Wikipedia page
A Wikipedia page has been created for WINDRIVE which is open to public and can be accessed at https://en.wikipedia.org/wiki/Windrive. The page contains a brief summary of WINDRIVE, main objectives of the project, list of project consortium partners and a link to the project website.
3. Video clip
A video clip has been designed, with length 4:49 and can be viewed on YouTube: https://www.youtube.com/watch?v=OvodJYWNZMk. The video clip gives information on the project and its aim and shows the prototype wind turbine located in Cambridge, UK. It also shows the laboratory at Cambridge University where a small scale generator is located. The consortium partners are mentioned in the video clip.
4. WINDRIVE publications
There are several publications from the outcomes of WINDRIVE which have been published in leading international conferences and journals. The list of publications is provided below.
Title: Effects of rotor winding structure on the BDFM equivalent circuit parameters
Authors: A. Oraee, E. Abdi, S. Abdi, R. McMahon, P.J. Tavner
Published: IEEE Transactions on Energy Conversion, DOI: 10.1109/TEC.2015.2432272
Title: A study of unbalanced magnetic pull in brushless doubly fed machines
Authors: S. Abdi, E. Abdi, R. McMahon
Published: IEEE Transactions on Energy Conversion, DOI: 10.1109/TEC.2015.2394912
Title: Brushless Doubly-Fed Induction Machines: Torque Ripple
Authors: T. D. Strous, Xuezhou Wang, H. Polinder, and J. A. Ferreira
Published: in International Electrical Machines & Drives Conference (IEMDC), May. 2015
Title: Finite Element Modelling of Brushless Doubly-Fed Induction Machine based on Magneto-Static Simulation
Authors: Xuezhou Wang, T. D. Strous, D. Lahaye, H. Polinder, and J. A. Ferreira
Published: in International Electrical Machines & Drives Conference (IEMDC), May. 2015
Title: Finite Element based Multi-Objective Optimization of a Brushless Doubly-Fed Induction Machine
Authors: T. D. Strous, Xuezhou Wang, H. Polinder, and J. A. Ferreira
Published: in International Electrical Machines & Drives Conference (IEMDC), May. 2015
Title: Effect of Rotor Skew on the Performance of Brushless Doubly-Fed Induction Machine
Authors: Xuezhou Wang, T. D. Strous, D. Lahaye, H. Polinder, and J. A. Ferreira
Published: in International Electrical Machines & Drives Conference (IEMDC), May. 2015
Title: LVRT Performance of Brushless Doubly-Fed Induction Machines – A Comparison
Authors: U. Shipurkar, T. D. Strous, H. Polinder, and J. A. Ferreira
Published: in International Electrical Machines & Drives Conference (IEMDC), May. 2015
Title: Converter rating optimisation for a brushless doubly-fed induction generator
Authors: A. Oraee, E. Abdi, R McMahon
Published: IET Renewable Power Generation, vol. 9, No. 4, 2015
Title: Equivalent circuit parameters for large brushless doubly-fed machines (BDFMs)
Authors: S. Abdi, E. Abdi, A. Oraee, R. McMahon
Published: IEEE Transactions on Energy Conversion, vol. 29, No. 3, 2014
Title: Vector-controlled grid synchronization of the brushless doubly-fed induction generator
Authors: A. Broekhof, M. Tatlow, R. McMahon
Published: 7th IET Int. Conference on Power Electronics, Machines and Drives (PEMD2014), pp. 1-6, 2014, Manchester, UK
Title: Design of the Brushless DFIG for Optimal Inverter Rating
Authors: E. Abdi, A. Oraee, S. Abdi, R.A. McMahon
Published: 7th IET Int. Conference on Power Electronics, Machines and Drives (PEMD2014), pp. 1-6, 2014, Manchester, UK
Title: Brushless Doubly-Fed Induction Machines: Magnetic field modelling
Authors: T. D. Strous, N. H. van der Blij, H. Polinder, and J. A. Ferreira
Published: in Conference on Electrical Machines (ICEM), Sep. 2014, pp. 2702–2708
Title: A novel analytical approach and finite element modelling of a BDFIM
Authors: N. H. van der Blij, T. D. Strous, Xuezhou Wang and H. Polinder
Published: in International Conference on Electrical Machines (ICEM), Sep. 2014, pp. 346–352
Title: Crowbarless Fault Ride-Through of the Brushless Doubly Fed Induction Generator in a Wind Turbine Under Symmetrical Voltage Dips
Authors: T. Long, S. Shao, P. Malliband, E. Abdi, R. McMahon
Published: IEEE Transactions on Industrial Electronics, vol. 60, No. 7, pp. 2833-2841, 2013.
Title: Dynamic Modelling of the Brushless Doubly-Fed Machine
Authors: P.C. Roberts, T. Long, R. McMahon, S. Shao, E. Abdi, M. Maciejowski
Published: IET Electric Power Applications, vol. 7, No. 7, pp. 544-556, 2013
Title: A Study of Converter Rating for Brushless DFIG Wind Turbines
Authors: A. Oraee, E. Abdi, S. Abdi, R.A. McMahon
Published: In Proceeding of the 2nd Renewable Power Generation Conference, September 2013, Beijing, China
Title: Design and Performance Analysis of a 6 MW Medium-Speed Brushless DFIG
Authors: E. Abdi, M.R Tatlow, R.A. McMahon, PJ. Tavner
Published: In Proceeding of the 2nd Renewable Power Generation Conference, September 2013, Beijing, China
Title: Investigation of Magnetic Wedge Effects in Large-Scale BDFMs
Authors: S. Abdi, E. Abdi, A. Oraee, R.A. McMahon
Published: In Proceeding of the 2nd Renewable Power Generation Conference, September 2013, Beijing, China
5. Presentations at International Conferences and Exhibitions
The consortium partners presented at several international conferences, exhibitions and workshops, including the followings:
Name of event: European Wind Energy Association’s Annual Event
WINDRIVE Partners attended/presented: WTL, Romax, Mita, Senvion, UCAM, DUT
Location: Vienna, Austria
Date: 4-7 February 2013
Name of event: International Conference on Electrical Systems for Wind Turbines
WINDRIVE Partners attended/presented: Romax
Location: Bremen, Germany
Date: 13-15 May 2013
Name of event: Renewable Power Generation Conference
WINDRIVE Partners attended/presented: WTL, UCAM
Location: Beijing, China
Date: 9-11 September 2013
Name of event: European Wind Energy Association’s Annual Event
WINDRIVE Partners attended/presented: WTL, Romax, Mita, Speedgoat, Senvion, UCAM
Location: Barcelona, Spain
Date: 10-13 March 2014
Name of event: International Conference on Power Electronics, Machines and Drives
WINDRIVE Partners attended/presented: WTL, UCAM
Location: Manchester, UK
Date: 8-10 April 2014
Name of event: American Wind Energy Association WINDPOWER Conference & Exhibition
WINDRIVE Partners attended/presented: Mita
Location: Las Vegas, United States
Date: 5-7 May 2014
Name of event: International Conference on Electrical Machines
WINDRIVE Partners attended/presented: DUT
Location: Berlin, Germany
Date: 2-5 September 2014
Name of event: Wind Energy O&M Forum Europe
WINDRIVE Partners attended/presented: Romax
Location: Hamburg, Germany
Date: 20-21 January 2015
Name of event: European Wind Energy Association’s Offshore Wind Conference
WINDRIVE Partners attended/presented: WTL, Romax, Mita, Senvion, UCAM, DUT
Location: Copenhagen, Denmark
Date: 10-12 March 2015
Name of event: International Electrical Machines and Drives Conference
WINDRIVE Partners attended/presented: DUT
Location: Coeur d’Alene, United States
Date: 10-13 May 2015
Name of event: International Symposium on Industrial Electronics
WINDRIVE Partners attended/presented: WTL, UCAM
Location: Rio de Janeiro, Brazil
Date: 3-5 June 2015
6. One-day Workshop
A one-day workshop was held in London on 5th March 2015. The attendees included invited experts from the wind power industries (supply chain, wind turbine OEMs, developers, operators and energy companies), academic institutes (university professors and researchers), governmental organisations (UK Department of Energy and Climate Change, Carbon Trust, Energy Institute, etc) and non-governmental organisations (Institution of Engineering and Technology, Cleantech Forum, etc).
The event included several presentations and Q&A sessions covering several aspects of WINDRIVE outcomes, including the followings:
1- Advances in electrical generator design for wind power industry;
2- Drivetrain topologies and understanding the trade-offs;
3- Wind turbine operations and maintenance for different drivetrain topologies;
4- Potentials of WINDRIVE drivetrain for offshore wind;
5- Wind turbine reliability and WINDRIVE contribution;
6- Economics of WINDRIVE technology and reduction on levelised cost of energy.
Dissemination Activities beyond the Life of WINDRIVE
The dissemination activities will continue beyond the life of the project to further maximise the impact of the project. These activities will include:
• Identification of target audiences and dissemination plans to ensure visibility and awareness of the achievements of WINDRIVE and enable effective knowledge transfer.
• Continued update of the project web portal to: (i) organise collaborative activities between project partners and disseminate information and announcements in the consortium, and (ii) distribute WINDRIVE non-confidential outputs among outside visitors, and to collect feedback from potential users.
• Scientific dissemination by the academic partners through international publication/ presentation at relevant conferences, workshops and seminars.
• Development of lectures and information packages for delivery within higher education facilities.
Publication and dissemination of the project knowledge and results will be conducted in agreement with the consortium partners. Careful consideration to timing will be maintained to ensure that the benefits of swift dissemination are achieved whilst safeguarding the intellectual property rights, confidentiality and legitimate commercial interests of the project SME partners.
The consortium have engaged, and will continue to do so, with relevant European Activities, including:
• European Wind Energy Technology Platform (TPWind);
• The European Wind Energy Association (EWEA) and various national associated organizations.
List of Websites:
www.bdfig.com
The Windrive project aims to further develop a Brushless Doubly-Fed Induction Generator (Brushless DFIG) technology for widespread implementation in wind power generation, moving it from being a promising and proven concept, demonstrated on small scales, to an optimized industrial-scale machine for multi-megawatt wind turbine applications. This will have a range of impacts – growth in sales and jobs for the SMEs collaborating in the project, knock-on benefits through the supply chain and substantial environmental and social benefits through a more cost-effective and reliable wind power sector.
The wind industry is growing rapidly. In 2009, the EU adopted a wide-ranging package on climate change with a target for a 20% in emissions of Green House Gases by 2020, compared with 1990 levels, a 20% increase in the share of renewables in the energy mix, and a 20% cut in energy consumption. To achieve this, there will be a significant dependence on the generation of energy using wind turbines and, accordingly, wind turbines are increasingly being deployed in large numbers.
However, the reliability and cost-effectiveness of wind turbines still need major improvement to make wind power competitive with conventional power plants. This is particularly important for the growth of offshore wind power generation.
The proposed medium-speed drivetrain incorporates the following components:
- Medium-speed Brushless DFIG, excluding brush-gear and slip-rings, known to be the highest failure rate components in the generator;
- Partially (typically a third) rated converter;
- Two-stage gearbox, excluding the third high-speed stage, known to be the highest failure rate part of the gearbox.
A recent study performed by an independent wind energy consultant shows that the whole balance of life saving as percentage of drivetrain (i.e. generator, converter and gearbox) capital costs can be more than 83%.
The technology is proven in concept on small scales, including a 20 kW prototype wind turbine and most recently, a 250 kW prototype generator on test bed. However, there is a lack of an industrial size machine to prove the technology on real size. Further, engaging with the market is difficult without a real size prototype.
This project has studied several aspects of the Brushless DFIG including its design, operation, control and grid connection with an aim to optimize and improve its performance for multi-MW wind turbines. Further, its integration into a wind turbine drivetrain and optimization on a system-level has been carried out. A 3 MW medium-speed Brushless DFIG drivetrain system has been designed and its design, modelling and analysis is documented in a comprehensive report which can be used by the SME partners during the exploitation phase which will follow the completion of the project.
Project Context and Objectives:
The generation of electrical power from wind energy is a proven technology. There is already a substantial installed capacity in a number of countries and new capacity is being continually added. Various designs of wind turbines have been proposed but horizontal axis, three bladed machines with a capacity from 600 kW to 3MW or above is the currently preferred option. In these machines a variety of architectures is used, including direct drive, in which the turbine is directly coupled to a low speed generator, and indirect drive, with the turbine coupled through a gearbox to a medium or high speed generator. There is only one large manufacturer of direct drive turbines and they have a significant installed capacity, especially in northern Europe. However, the majority of wind turbines in the world employ the indirect drive architecture. In these machines, generation is normally from a doubly-fed slip-ring induction generator (DFIG). As the gearbox gives an increased shaft speed, a four or six pole machine can be used. The stator of the generator is connected directly to the fixed voltage, fixed frequency electricity grid and the rotor is supplied through the slip rings with a variable voltage and frequency supply generated from a converter.
This form of double feed enables generation to take place over a range of turbine speeds and provided that the range of speeds is moderate, the converter rating need only be a fraction of the total generator output thereby keeping the system cost low yet retaining a reasonable level of reliability. The converter rating is generally about one third of the rating of the generator, allowing speed variations of ±33%. Varying the voltage supplied to the rotor can be used to manage the reactive power flow from the generator. Control schemes have also been developed to enhance the response of the system to changing wind speed and to accommodate varying grid conditions.
However, there are drawbacks to the use of DFIGs, particularly the additional cost and bulk of a machine which incorporates slip rings and the need to maintain brush-gear and to replace brushes on a regular basis. Studies have shown that the reliability of large wind turbines is improving but that faults with generator and converter sub-assemblies contribute significantly to turbine failure rates. Further studies have shown that problems with brush-gear are significant in wind turbine operation and that the problem will be more severe in machines deployed offshore where additional wind resources are available.
Recent reliability studies have shown that the failure rate of wind turbine gearboxes is modest and in fact is considerably less than some other components of the wind turbine such as the power electronics, pitch control and the yaw system. On the other hand, the repair of the gearbox once it fails is significantly costly and in most cases, the gearbox is replaced. Therefore, improvement of gearbox reliability is of significant benefit to the wind industry. The high-speed stage of the 3-stage gearbox, which is commonly used in DFIG drivetrains, accounts for the highest number of failures and therefore eliminating this stage will substantially improve the gearbox reliability. A number of wind turbine manufacturers, such as Multibrid and Winwind have incorporated simplified 1 or 2 stage gearboxes in recent years, but all these drivetrains utilize permanent magnet generators.
Brushless DFIG drivetrain
The proposed novel and disruptive drivetrain technology leverages the advantages offered by the widely used DFIG drivetrains, but also achieves an intrinsically enhanced reliability and lower capital and O&M costs by incorporating a Brushless DFIG in combination with a simplified 2-stage gearbox. The Brushless DFIG requires an identical size converter as that of the DFIG, which is of fractional size, typically a third of the generator rating. Eliminating the slip-ring and brush-gear system in the generator and the high-speed stage of the gearbox, which are reported as the highest cause of failures in these components, enables achieving a significantly improved reliability. Further, the Brushless DFIG has shown to have an intrinsically improved low-voltage ride through (LVRT) performance during grid faults, which removes the need to have complex additional hardware in the converter, such as a crowbar. In summary, the Brushless DFIG drivetrain has the following unique advantages:
• High reliability due to:
- elimination of slip-rings and brush-gear in the generator which is known as the highest failure rate component in the generator;
- elimination of high-speed stage of the gearbox which is known as the highest failure rate part in the gearbox;
• Low capital costs due to:
- elimination of slip-rings, brush-gear and carbon extraction system in the generator;
- utilization of a simplified 2-stage gearbox;
- utilization of a fractionally rated converter;
- elimination/simplification of the crowbar design used for LVRT protection;
• Low O&M costs due to:
- eliminating the need for regular maintenance and replacement of brush-gear in the generator;
- higher reliability of the components, reducing the unplanned maintenance costs;
• Improved grid compatibility due to:
- an intrinsically better LVRT performance of the Brushless DFIG because it has higher rotor inductance by design;
• enhanced reactive power injection capability of the Brushless DFIG during grid faults.
Modern Brushless DFIG
The modern Brushless DFIG concept comprises two electrically separate stator windings, one (i.e. the power winding) connected directly to the mains and the other (i.e. the control winding) supplied from a variable voltage, variable frequency inverter, as shown in Figure 3. The design of the Brushless DFIG is not straightforward as there are more variables to consider than in a conventional induction machine. Nevertheless, attempts have been made to manufacture large Brushless DFIGs, beginning with the 75 kW machine of Runcos et al. and more recently Chinese designs of 200 kW. WTL’s work on a 20 kW Brushless DFIG has confirmed that a Brushless DFIG can be designed to a specification and can work in a wind turbine – an important part of this was implementing effective control of the machine, which is not stable in the open loop over the normally used speed range. Over more than six years’ operation has given confidence in the proposition of a large scale Brushless DFIG; the company has recently designed and tested a 250 kW Brushless DFIG system which was conceived as a stepping-stone towards a multi-megawatt scale machine.
Windrive Project
In order to make the Brushless DFIG attractive for commercial applications, its performance must be optimized to commercially acceptable levels and be demonstrated on real size (i.e. multi-MW size). Further, its mechanical and electrical integration within the wind turbine drivetrain and optimization of the system as a whole must be carried out. The aim of WINDRIVE project is to address these needs and to industrialize the design, construction, integration and operation of the Brushless DFIG in a 3 MW wind turbine drivetrain and achieve an optimized solution attractive to the industry.
The project benefits from the experience and know-how gained in the design and testing of previous smaller size prototypes, including the 20 and 250 kW systems. In the course of the project, several aspects of the new drivetrain technology have been studied and optimized by the RTD performers with inputs from the SME partners. A 3 MW drivetrain system have been designed and modelled, which is the first of its kind. The design will be utilized by the SME partners in the follow-on exploitation phase to build and test a prototype system. The project has nine distinct objectives:
Objective 1: Achieve an overall optimization of the Brushless DFIG design with respect to its manufacturing costs, size, weight and performance.
Objective 2: Achieve optimized control and operation of the Brushless DFIG within a wind turbine.
Objective 3: Establish a grid-friendly system design compatible with recent grid codes of EU countries.
Objective 4: Establish industrial low-cost, reliable mechanical and electrical integration of the Brushless DFIG within a wind turbine with a customized/improved components.
Objective 5: Establish industrialized design, construction and operation of the Brushless DFIG drivetrain.
Objective 6: Achieve a quantified assessment of reliability and economics of the Brushless DFIG drivetrain.
Objective 7: Establish a complete design of a 3 MW medium-speed Brushless DFIG drivetrain, ready to be built and tested during the follow-on exploitation phase.
Objective 8: Establish a portfolio of IP in the form of patents and knowhow related to the design and operation of the Brushless DFIG drivetrain, to be exploited by the SME partners.
Objective 9: Disseminate the project’s finding across the wind generation sector in the EU in order to facilitate the wider societal impacts offered by this technology.
Project Results:
1. Brushless DFIG modelling tools:
Several modelling tools including the Finite Element (FE), coupled-circuit, d-q and equivalent circuit models for the Brushless DFIG (also known as the BDFM) have been developed to study the electric and magnetic circuits in the machine. The predictions from the FE model have been compared with analytical models and the results obtained from experiments, which confirmed the accuracy of the modelling tools.
An iron loss model has been developed for the machine based on conventional techniques used for AC machines. The model, implemented in FE, calculates the iron losses at each time step for each element of the iron circuit. Preliminary core loss measurements have been carried out on a laboratory machine located in UCAM using voltage and current measurements.
A lumped parameter thermal model for the Brushless DFIG has also been developed for determining the machine rating and to study machine design aspects related to the electric loading of the machine. The model is used to determine the thermal coupling between the stator windings and the rotor heating in simulation.
2. Brushless DFIG design practicality:
A design procedure for the Brushless DFIG has been developed based on equations derived from a simplified equivalent circuit. The method allows the many variables in the design of this machine to be handled in straightforward way. Relationships are given for the division of slot area between the two stator windings and for the design of the magnetic circuit. The design method is applied to a frame size 180 machine. In particular, calculated values for flux densities in the machine have been verified by time stepping finite element analysis for actual operating conditions. Practical limitations in the construction of the BDFM are also assessed.
The design procedure for the BDFM starts with the choice of pole pair numbers to set the speed. Normal balanced three-phase windings are used for the stator windings, and these can be short-pitched. The rotor should have a turns ratio close to optimum and the winding must be insulated to give determinate current paths. Equations are derived from the machines equivalent circuit to give an initial design. This can be refined by an iterative procedure which keeps the machine within electric and magnetic loading limits over a specified operating range. The approach is illustrated with results from a frame size 180 BDFM. However, other methods will be needed for a detailed assessment of harmonics and the checking of the magnetic circuit to prevent saturation.
Design optimization to improve the BDFM PW power factor is carried out. It is shown that for smaller machines, leading power factors are difficult to achieve unless the magnetizing burden from the PW is significantly reduced by increasing the converter rating or with capacitor banks. However, as the rating of the BDFM increases to the megawatt scale, achieving a leading power factor becomes easier.
A comparison of different pole pair machines for optimal design of the power factor, torque performance and back iron thickness is carried out. It is observed that for lower pole pair count machines with their poles further apart, better performance (power factor and torque) can be attained, however this comes at a cost of increased weight since more back-iron is required.
3. Design methodology for multi-MW Brushless DFIG:
A design methodology has been developed for the megawatt Brushless DFIG for the wind turbine drivetrain application. The methodology consists of analytical and Finite Element (FE) models, for determining the Brushless DFIG performance and an optimization algorithm for geometry optimization. Based on the presented design methodology, a Brushless DFIG optimization design program has been developed, using MATLAB script and functions. For Finite Element (FE) calculations Comsol software is used. All FE models and calculations are set-up using MATLAB script and are executed through a MATLAB-Comsol interface connection. Therefore, the whole design optimization process is executed through MATLAB script.
The design flowchart incorporates the relations of the used input requirements, models and algorithms in order to come to an optimized Brushless DFIG prototype design that is suitable for industrialization. The optimization program combines the use of several models, such as the geometric model, magnetic-field model and the thermal model, to analyze an input machine design. The optimization algorithm then optimizes the machine design according the provided optimization criterion.
4. Rotor design methodology:
A procedure for obtaining rotor parameters for BDFMs with nested loop rotors has been established. Winding factors are used to give parameters valid for one operating condition but in fact there is little dependence on speed. Parameters determined from the analytical winding factor method are close to those determined experimentally. The ability to calculate rotor parameters accurately enables the exploration of optimal rotor designs and overall machine design. Although this study has looked at the nested-loop type rotor, the methods can be readily extended to other arrangements, including loops of arbitrary pitch and placement.
The ability to accurately calculate rotor parameters analytically enables research into optimal rotor designs. Optimization of the rotor aims to achieve a near optimum turns ratio, while keeping rotor leakage and resistance within a desired range. A low rotor leakage value may be beneficial because it increases the pull-out torque of the machine, however a higher rotor leakage will lead to better Low-Voltage Ride Through (LVRT) performance. A low rotor resistance will lead to lower rotor power losses. It is therefore important to have the ability to tailor rotor parameters in response to design specifications.
5. Rotor design optimisation:
The rotor teeth and core back sections of electrical machines are sized according to the expected peak flux density. Since the peak flux density of the rotor teeth and core back sections in the BDFM are not equal, the size of the teeth and core back sections should similarly not.
Analytical methods have been developed to determine the location of the largest and smallest peak flux density in the rotor teeth and core back sections and therefore the geometry of the rotor can be optimized. However, this field behavior cannot be exploited in the stator because the peak flux density value is different in each teeth and core back section at specific operating speeds. This will be particularly useful for the nested loop rotor because the magnitude of the current in the different loops of the nested loop rotor are not equal - the outer loop carries the greatest current and the inner loop the smallest. It has therefore been recommended to size the rotor bars according to the expected current for each loop.
The peak flux density in the back iron is similarly not equal, with the largest peak flux density found between the innermost rotor loops and the smallest found between the outer loops. It is therefore recommended to optimize the depth of the back iron (non uniform thickness) based on the peak flux density.
6. Design of alternative rotors:
The stator of a BDFM has two standard three-phase windings that are placed one on top of
the other in the same slots. The windings having different pole numbers chosen such that the winding factor of one winding is zero when analyzed for the fundamental harmonic of the other winding. This will ensure that there is no direct coupling. The rotor requires a winding which is capable of coupling with both stator windings.
It was shown by Browadway and Burbridge that a cage rotor of p1 + p2 bars (where p1 and p2 are the pole-pairs of the power and control winding respectively) performs the required linking function. It was further showed that a rotor with p1 + p2 equally spaced loops will also work regardless of the span of the loops, and thus proposed the 'nested loop rotor' made up of sets on concentric loops, with or without an enclosing cage. Subsequent to its discovery many nested-loop rotors have been built by other research groups, some with and some without the enclosing cage, but no particular analysis offered as to which has the best performance.
Two simple BDFM rotors have been studied including a cage rotor with p1 + p2 bars, and a loop rotor with p1 + p2 loops, drawing from analysis of the rotor structure and its equivalent circuit parameters. Five different rotor configurations were built for the prototype 160 frame size BDFM and their parameters were measured and compared.
Analysis of the simplest BDFM rotors suggest that increasing the loop span results in a lower referred resistance, however minimum leakage inductance occurred at a span smaller than fully pitched. A comparison between the cage rotor and the full span simple loop rotor shows that even though the two options filled the same slots, the cage rotor had considerably lower referred resistance and slot leakage inductance. The trends from the simple analysis apply directly to the five experimental rotors. The impedance of the series wound rotor is higher than that of the nested loop because current is forced down narrow loop spans. Design studies were performed to find the optimum number of series coils for a progressive loop rotor and loops per nest for a nested loop rotor. In each case it appeared that 'the more loops the better', up to the point where manufacturing the rotor becomes prohibitively expensive. Switching from a nested loop to a cage+loop offers a dramatic reduction in impedance (especially resistance) because the rotor will have the largest possible loop spans and because of the phasor addition of the currents in the bars. This suggests that for BDFM rotor designs the cage+loop format is likely to offer the best performance, as would be expected from initial comparison of the simple loop and cage rotors. Finally it was shown that some improvement in leakage inductance and resistance may be had by optimizing the span of the rotor loops. Optimizing the loop span further improves impedance by reducing the harmonics in the Magnetomotive Force (MMF) waveform. Though these changes move the ratio further from optimum, it has little impact on machine rating.
7. Grid connection for Brushless DFIG:
A robust controller for the Brushless DFIG has been developed which is essentially a power winding flux orientated vector control scheme without cross-coupling compensation. Theoretical analysis has been used to represent the decoupled control scheme for real and reactive power control. A Brushless DFIG model and controller have been implemented in MATLAB/Simulink and analyzed through simulations. The proposed control scheme successfully controls the static and dynamic demands of real and reactive power. An LVRT fault was also simulated on the model, and the results show that the proposed controller can successfully ride through the fault, without additional electronics, and inject the required reactive power. Future research should examine the effects of cross-coupling on the Brushless DFIG and to include a cross-coupling compensator into the controller.
8. Grid synchronization for Brushless DFIG:
At start-up the Brushless DFIG is required to synchronize to the grid so that the power can be extracted from the turbine. Control winding (CW) excitation is regulated according to the grid voltage and rotor speed to ensure the power winding (PW) voltage, phase and frequency matches the grids. Post synchronization, the control strategy is switched to a power control loop where the real and reactive power of the generator is controlled.
The developed grid synchronization model is implemented on the D180 frame size Cambridge Brushless DFIG. Initially the PW is not connected to the grid (through a contactor) and the rotor speed is ramped up. When the rotor speed reaches a speed of 450 rpm, a voltage is applied to the CW which excites the open-circuit PW. A closed loop controller uses a phase-lock-loop to ensure the amplitude, frequency and phase of the PW voltage matches the grid, this is done using the controller. When the two voltages are well matched (a small error), the contactor is closed and the PW is connected to the grid. Once the rotor speed reaches 450 rpm, the complete synchronization process takes 0.36 s. After grid synchronization, a real and reactive power control loop is switched in to control the output power of the generator.
For maintenance on the wind turbine, it is beneficial if the rotor of the Brushless DFIG can be controlled at a slow speed. For this mode of operating the Brushless DFIG should be connected in simple induction mode where the CW is connected to the grid through the frequency converter and the PW is open-circuit (through a contactor). A simple V/f controller is implemented where the amplitude and frequency of the voltage applied to the CW is adjusted.
It was shown through experiments that the speed can be controlled right done to 5 rpm. In this mode of operation the machine is self-starting however the speed cannot be controlled down to zero-speed. If this mode of operation shows commercial potential a closed-loop vector controller may be implemented, this will allow tighter speed control.
9. Sensorless control for the Brushless DFIG:
A control system for the Brushless DFIG including sensorless position estimator has been developed. In order to verify and show the working principles of the control system, first a time dynamic model of the complete system, including a Brushless DFIG machine model and the sensorless controller model (control algorithm including sensorless position estimator), were developed.
A Brushless DFIG model was developed using the concept of ideal rotating transformers (IRTF). To control the Brushless DFIG output torque in an efficient and stable manner, the developed Brushless DFIG control algorithm uses closed loop vector control of the CW current vector transformed to the CW flux coordinate system (dq‐reference frame). This enables fast and smooth control. In this way torque is proportional to the CW q‐axis current component, while the flow of reactive power in both PW and CW is influenced by the CW d-axis current component. To apply the reference transformation to the CW flux coordinate system the CW flux position angle must be obtained. This position angle is obtained using the sensorless position estimator, which derives the CW flux position angle from the measured CW flux (obtained from voltage and current measurements). By measuring the CW flux, the controller controls the CW voltage (and hence current) to follow the measured flux and hence the Brushless DFIG is kept in its synchronous mode of operation. There is no need to know the mechanical position angle at all. Though the mechanical position angle could easily be derived if required, by using the estimated PW and CW flux position angles.
To show the effectiveness of the developed controller including sensorless position estimator, its performance is simulated on a Brushless DFIG model of an existing D160 frame size machine. First the cascaded torque speed characteristics of the Brushless DFIG were determined, by simulating the Brushless DFIG time‐dynamic behaviour when connected direct online with one of the windings short circuited. The resulting torque response was compared to steady‐state calculations, showing that the Brushless DFIG model behaved as expected. Then the control system including Brushless DFIG controller with sensorless position estimator was simulated. The system response to a step input on the CW d‐axis and on the CW q‐axis current components was simulated, showing that the torque is proportional to the CW q‐axis current component and that the flow of reactive power is proportional to the CW d‐axis current component. Then a speed ramp up from sub‐synchronous to super synchronous speed was simulated, showing that the controller is perfectly able to keep the Brushless DFIG in synchronous operation while the speed varies. It could also be observed that the torque collapses when the rotor frequency reaches 0 Hz. For stable control it is advised to keep away from this operating point. However, it should be noted that this operating point is beyond the speed range required for wind turbine drivetrain operation.
The developed models provided insight in the Brushless DFIG control algorithm and how to obtain sensorless position estimation. The simulations showed the effectiveness of the developed control algorithm including sensorless position estimator. The control algorithm including sensorless position estimator was also implemented in hardware successfully.
10. Optimisation of control system:
A stability model has been developed for the wind turbine drivetrain (WTDT) incorporating a Brushless DFIG. Using the developed WTDT model and control algorithm, the dynamic behaviour of the WTDT to a step change on the input is simulated. The stability of operation is key in electrical generators, on the other hand, improving the dynamics would typically trades off with stability margins. A steady state model is developed which is used to analyse the Brushless DFIG stability.
Two cascaded control loops are used to control active‐ and reactive power independently. Cross compensation has been included to limit cross coupling between control loops. Tuning of PI controllers is based on machine parameters and offer good performance during variation of parameters during operation. The machine controller offers good dynamic behavior to a change in wind speed and is able to maintain optimum power extraction.
11. Condition monitoring for the Brushless DFIG:
To improve WT availability and reduce O&M costs, two condition monitoring techniques have been developed based on analyzing the generator power signal using Continuous wavelet transform (CWT). In comparison with conventional vibration, temperature measurements, and lubrication oil analysis techniques currently been used in industry, the developed techniques have the advantages of reduced capital cost, ability to detect both electrical and mechanical faults, and can be applied to both geared and direct drive WT's with different generator topologies.
1. The new CWT-based condition monitoring algorithm identifies the faults, their associated frequencies and their duration, however conventional approaches also pick up other frequencies associated with normal operation of the wind turbine, and the algorithms take considerable time to execute.
2. The new CWT-based energy tracking method identifies fault perturbations applied to the generator shaft. This method considerably reduces the calculation time, however some fault signals were lost when wind turbulence was included. With more refined signal processing, this technique does have potential and should be further researched.
The two approaches now need to be applied to the power signals obtained from real wind turbine’s during mechanical and/or electrical faults to determine the detectability of the techniques.
12. Definition of drivetrain topologies:
The technical parameters of several wind turbine drivetrain topologies have been defined. These specifications are used in mathematical modelling and simulation work to support the development of a new Brushless Double-Fed Induction Generator turbine. The results of this modelling have been used to influence design decisions in all aspects of the new turbine (structural, mechanical, electrical etc.) so as to verify and optimize the design.
Three turbine specifications were proposed, covering a range of existing generator technologies and transmission types. The designs are derived from existing, commercially available wind turbine products, and are intended to be comparable in terms of operational environment and power output.
13. Stochastic energy analysis:
The aim of this work is to develop mathematical models of the dynamical properties of wind turbine drivetrains, and tools to analyse them so that comparisons may be made between competing designs. Particularly of interest are the vibrational response characteristics of the drivetrains. Unanticipated vibration of any mechanical system has implications for component lifetimes (due to frictional wearing), and radiated audible noise (which is undesirable in a turbine). The desired outcomes of this work are a thorough written analysis of the proposed design options, and a set of software tools to aid in the design process. These results and tools will then aid in the development of a complete turbine design for the new Brushless DFIG drivetrain.
Significant work has been undertaken to develop three wind turbine drivetrain models and analysis tools necessary for the project. These have been implemented as a series of modular MATLAB scripts that produce mass and stiffness matrices representing the masses and stiffnesses of the drivetrain components. Linear Vibration theory is applied to these system matrices, resulting in lists of resonant frequencies and mode shapes. These results can then be used to predict further characteristics such as frequency transfer functions and transient responses. Additionally, software has been written to visualize the modal results from these scripts, allowing fast and intuitive analysis and verification of the results. This information is crucial for predicting problematic vibrations, and will influence the design of other aspects of the turbine.
The stochastic energy analysis enables us to quickly finalize the three drivetrain models with realistic design parameters, and provide analysis of their relative dynamical characteristics. The outcomes of the analysis has facilitated rapid iteration of design process, and optimization of the candidate BDFIG design. In particular, the torque-ripple data from simulation of the generator have been used for comparison. Further work will focus on adding more complexity to the modelling (particularly in the areas of structural coupling and electromagnetic effects within the generator) as necessary to match the frequency ranges of interest.
14. Brushless DFIG drivetrain efficiency and cost modelling:
A wind turbine drivetrain (WTDR) incorporating the Brushless DFIG efficiency and cost model has been developed. The model includes efficiency models for each WTDT system component, as summarized below:
• Wind availability model.
• Wind turbine rotor efficiency model.
• Gearbox efficiency model.
• Brushless DFIG efficiency model.
• PE converter efficiency model.
Additionally a system performance model for the WTDT as a whole has been developed. Further, a simple capital costs (CAPEX) model including all WTDT system components is developed. The resulting models enable the analysis of a complete WTDT performance over its full steady-state operating area. The resulting models are useful for WTDT system comparison and speed-range optimization analysis.
15. Drivetrain speed range optimization:
The generator speed-range and rated speed will vary depending on the WTDT gearbox ratio and generator operating frequencies. This work studied those trends for different generator systems in a WTDT. The results can be used for speed-range optimization, by selecting the most favorable combination of system components.
A 3.2MW wind turbine drive train has been analyzed as a case study. Different generator systems are compared over a wide speed range. For this purpose 10 PMG systems, 8 DFIG systems and 3 Brushless DFIG systems have been analyzed. Each system with different gearbox ratio and rated generator speed. Resulting trends provide the following insight:
• Generator CAPEX getting cheaper with increased ratio.
• Gearbox CAPEX getting expensive with increased ratio.
• For three stage gearbox systems, the total system CAPEX always decreases with increasing gearbox ratios.
A medium-speed 2-stage gearbox solution is favoured for the further development of a multi megawatt Brushless DFIG WTDT. The lowest total capital cost and highest energy yield per euro of capital cost was obtained for a 2-stage gearbox systems with corresponding generator speed in the range of 200 to 300 RPM, but the speed is also related to the wind turbine rotor rated speed. A medium-speed generator solution is also favoured over a high-speed solution, because of historical reliability problems of high-speed 3-stage gearboxes.
In order to provide a more thorough generator system comparison, future research work should focus on:
• System performance evaluation of Brushless DFIG machines.
• The Brushless DFIG cost trend as function of gearbox ratio should be extended over the whole gearbox ratio range.
• Multi-objective design optimization, to provide a better trade-of between generator cost and generator efficiency, which has impact on the system performance and therefore on the energy yield per euro of capital cost.
16. Drivetrain simulation under grid events:
Wind turbines connected to the grid must achieve good grid quality and voltage ride through. The stator windings of a Brushless DFIG and DFIG are connected directly to the grid. In the event of a voltage dip on the grid the turbine must remain connected and not shut down. However, if a voltage dip is seen in the generator this has an effect on the generator torque which reacts the mechanical torque. Effects that are studied included the combined effect of controller algorithms and grid connection on the reaction current and torque response of the different generator systems in a wind turbine drivetrain. A comparative study has been carried out. The generated torque profiles for each generator system enable understanding of grid events and controller dynamics on mechanical component response and reliability. The knowledge gathered will facilitate an understanding of the robustness of the Brushless DFIG in comparison with PMGs and DFIGs. The investigation included the protection of the power electronics during voltage events and the reaction of the generator in these situations.
It has been found that for the PMG based wind turbine, the issue with Low Voltage Ride Through (LVRT) is the rise in the DC link voltage. This is due to the mismatch in power generated by the machine and the power transferred to the grid (which is limited by the reduced voltage at the grid side converter terminals). Methods for controlling this DC voltage rise, such as the use of an energy discharge circuit, have also been investigated. The method Power Balancing was proposed, where the power generated by the machine is controlled to match the power transferred to the grid in the event of a low voltage event. This method has also been used to control the issue of DC link voltage rise, however, it introduces torque disturbances in the machine.
For the DFIG it has been found that voltage dips cause large transient currents in the stator and rotor circuit. The over‐currents in the rotor circuit are a cause for concern as they may lead to adverse effects on the power electronic converter connected to the circuit. A method to overcome this issue i.e. the use of a crowbar circuit, has been studied. The crowbar circuit manages the problem through bypassing the power electronic converter in the event of current rise.
In the case of the Brushless DFIG, it has been found that again transient currents are present in the power and control windings in the event of a voltage dip. However, the magnitude of these currents is lower when compared to that in a DFIG. This is attributed to the higher leakage inductance of the Brushless DFIG. A crowbar‐less method has been studied which is successful in controlling the high currents generated in the control winding. The crowbar‐less control method also reduces the torque oscillations observed in the Brushless DFIG for a low voltage event.
In conclusion, apart from offering better reliability through the exclusion of slip rings and brushes, the Brushless DFIG also has an improved LVRT performance when compared with the DFIG. The protection is simpler and does not require an external circuit, like the crowbar, and can be built into the control algorithm of the machine controller.
Also, in this study the complete reactive power interaction is considered from the power winding of the machine. In reality, there is a possibility to use the grid side converter for the transfer of reactive power to and from the grid as well as the generator windings. This aspect and its effect on the machine response should be looked into in future.
17. Generator torque modelling and response:
The torque and Unbalanced Magnetic Pull (UMP) profiles, due to the electromagnetic fields in the Brushless DFIG, are important input characteristics for the WTDT mechanical structure models. Large torque ripples and UMP could negatively affect component life‐time and contribute to additional noise and losses.
Brushless DFIG Torque and UMP responses have been derived for a 3.2 MW case‐study machine. Based on the specifications of the model 3.2 MW wind turbine, the torque and UMP response profiles were determined under normal (nominal) operating conditions. Using FE modelling and Maxwell’s stress tensor method the torque and UMP response could be calculated.
Then using space and time Fourier analyses, both the torque response and UMP response have been analyzed. It is found that mainly winding space‐harmonics and slotting harmonics are responsible for torque ripple. The amplitudes, frequencies and origin of these space‐harmonics have been determined, in order to recreate the torque profile. The UMP response showed little unbalance, since the pole‐pair combination of the machine was chosen to cancel out any UMP. A space-harmonic analysis has been performed to determine the stator and rotor eccentricity modes. Time-harmonic analysis has shown that there is some fluctuation in the amplitudes of the eccentricity modes over time. This can mainly be attributed to winding and slotting harmonics.
18. Control hardware industrialisation:
The hardware design plays an important role in the performance and reliability of wind turbines in industrial environments. Suitable hardware designs for the control and instrumentation systems of the Brushless DFIG have been developed. The designs have utilized the input and output parameters required to control the Brushless DFIG using the Speedgoat real-time target machine. Input and output voltage signals from Speedgoat are required to be converted to the correct voltage level as well as buffered to protect the device, this is done with a signal conditioning board with custom built printed circuit board.
For a multi-MW Brushless DFIG designed for wind turbine applications it is envisioned that the Educational real-time target machine will be invaluable during the testing process, however an embedded controller would probably be used for commercial applications.
19. Control software industrialisation:
The industrialization of software is crucial in protecting the IP related to the control algorithms. The Simulink model is compiled into a binary file (.dlm) on the host PC before it is transferred to the Speedgoat target machine. Therefore the model is not visible on the target machine. To protect or hide the Simulink model from the end user on the Host PC, a graphical user interface (GUI) has been built, where parameters (eg. Real and Reactive power reference signals) can be passed from the GUI to the Simulink model without knowing the actual parameter names. The Simulink model with the important intellectual property (IP) for controlling the Brushless DFIG can then be password protected from the end-user.
MATLAB provides the GUIDE (graphical user interface design environment) toolbox for designing user interfaces for custom applications. The GUIDE Layout Editor allows the user to graphically design his/her own interface. GUIDE then automatically generates the MATLAB code and the user can modify this code to program the behaviour of the specific application. The GUI will be composed of an m-file and a .fig file, and will also reference the MATLAB/Simulink binary file with the .dlm suffix.
Currently the Speedgoat xPc real-time target machine connected to a Host PC is used to control the Brushless DFIG. This setup allows the user to easily adjust parameters on a MATLAB/Simulink model running on the Host PC , which can be executed in real-time via the Speedgoat device. This control setup allows rapid prototyping of the controller - the user can quickly and easily modify the Simulink model. This is useful for testing the Brushless DFIG, however it is believed that in the future, an embedded controller will be used.
This embedded system will be based on a microcontroller, programmed to handle the specific task of controlling the Brushless DFIG. Since the embedded system is dedicated to the specific task, the designer will be able to optimize the module to reduce the size and cost of the product, and if mass-produced will benefit from cost savings through economies of scale. This embedded controller will need to have sufficient analog and digital inputs and outputs and sufficient CPU speed to be able to compute the control loop.
20. Quality control management for Brushless DFIG manufacture:
Unbalanced magnetic pull (UMP) increases eccentricity in the Brushless Doubly-Fed Machine (BDFM) and may cause serious damage to the machine or drivetrain. In addition, UMP acts as a major source of vibration and noise. In order to build a BDFM generator successfully, methods to reduce UMP need to be studied.
This work focuses on the UMP in the BDFM, caused by both static and dynamic rotor eccentricities. Several parallel winding designs for the two stator windings are proposed and the practicality of such designs is discussed with respect to direct coupling between the stator windings and with rotor undesirable harmonic fields. Once practical parallel winding designs are established, their effects on reducing deflection as a result of static and dynamic eccentricities are shown and compared with series wound stator.
The study of direct coupling is particularly important when short pitched windings are utilized. It has been shown that with appropriate connection of coil groups in series and parallel, direct coupling of stator windings can be removed. There may be several possibilities of practical parallel winding designs, in which case the design with more parallel paths is shown to have stronger effects in suppressing the UMP.
Larger scale BDFMs are likely to be designed for slower natural speeds, hence with higher stator pole numbers. This will provide more possibilities for parallel connection of stator coils, but the practicality of such designs must be carefully assessed. The study carried out in this work, though is shown for a specific BDFM with 4 and 8 pole stator windings and a nested loop rotor, can be generalized for other BDFM designs.
21. Development of reliability model:
A reliability model for the drivetrain of a geared wind turbine fitted with a doubly fed induction generator (DFIG) and the novel Brushless DFIG has been. The model is derived from considering the failure rates of the main electrical assemblies namely the generator and converter as well as the gearbox.
The analysis of the reliability model suggests that the reliability and reparability of a Brushless DFIG, partially rated converter and 2-stage gearbox is better than the conventional wind turbine drivetrain comprising of a DFIG, partially rated converter and 3-stage gearbox. The main reason for the lower reliability of the system with the DFIG is the presence of the brushes in the configuration and the additional stage in the gearbox (high-speed parallel).
Future work should include field data from actual wind turbines of failure rates of gearboxes. It would also be interesting to compare permanent magnet generators with DFIG's and Brushless DFIGS and to compare direct drive with geared wind turbine systems. It would also be useful to compare the reliability and availability of wind turbines (and farms) across geographical regions as well as on and offshore turbines.
22. 3 MW Brushless DFIG and controller design:
A design for a 3.45 MW Brushless DFIG, its associated converter and controller has been developed. The system has been designed to match the output power, speed and voltage requirements for the Senvion 3.2M114 wind turbine, with the high speed 3-stage gearbox being replaced with a medium speed 2-stage gearbox. Simulated efficiency, power factor and heat tests were conducted on the proposed design, to confirm the generator will produce rated power at an adequate efficiency and power factor. The design was simulated in MATLAB/Simulink to test the dynamic response of the system by applying step changes in real and reactive power as well as simulating LVRT grid faults. The simulations showed that the controller can adequately control the 3.45MW Brushless DFIG.
23. Design of grid connection and condition monitoring:
This work deals with the electrical systems required to connect a wind turbine fitted with a Brushless DFIG to the grid. The different electrical systems required for grid connection and a start-up and shut-down procedure for the wind turbine have been developed. Fault detection and how the system responds to these faults is also discussed as well as condition monitoring alerts. The electrical systems comprised of the Machine-Side Inverter (MSI) and Grid-Side Inverter (GSI), contactors and wind turbine controllers. Fault conditions were defined which will raise a fault flag to the turbine controller and shut the wind turbine down in a safe way. The condition monitoring system previously developed is also incorporated into the system. Finally a start up and shut down procedure for the Brushless DFIG wind turbine was created.
The converter rating for the Brushless DFIG wind turbine with respect to reactive power management and grid LVRT requirements has been studied. It has been shown that at a given speed, taking into account both steady-state and transient operating conditions, there is a minimum sum of the MSI and GSI ratings but this condition may require an undesirably high control winding voltage so the minimum rating may not be practically achievable. Furthermore, it is shown that the MSI rating is typically significantly larger than the GSI rating because the MSI contributes to the supply of machine’s magnetizing currents and must handle large transient currents for crowbar-less LVRT operation.
The effects of a capacitor bank in reducing the rating of the GSI is also discussed. It is shown that with the use of a suitably rated capacitor bank, the minimum sum of the MSI and GSI ratings can be achieved at rated CW excitation voltage.
The effects of the machine’s leakage inductance on MSI rating and efficiency has also been studied. A higher leakage inductance limits the current during LVRT conditions but comes at the price of a restricted power factor range without de-rating the system. The effect of leakage inductance on machine losses was also investigated and these can rise significantly when operating at leading power factors. However, the problem becomes less critical for larger BDFMs, i.e. MW scale machines, since the per unit value of the rotor reactance will drop with size. The use of a capacitor banks is therefore more valuable with smaller machines unless the operating speed range above the natural speed can be limited to around 20%.
24. Brushless DFIG controller implementation:
The hardware and software implementation of a controller for the Brushless DFIG has been studied. For the development of the practical Brushless DFIG controller a test set‐up was built. The test set‐up featured a D160 frame size prototype Brushless DFIG and a specially developed controller to control the Brushless DFIG in an efficient and stable manner. The controller was built using an old converter, equipped with a new Digital Signals Processing (DSP) and measurement circuitry. The control algorithm with sensorless position estimator, as developed earlier in the project, was programmed into the DSP.
The result is a functioning control system for the Brushless DFIG. Measurements were performed demonstrating the control system performance. The measurements showed that the control system for the Brushless DFIG is able to control the machine in a stable manner over its complete speed range. The used D160 frame size Brushless DFIG in the test set‐up provided some difficulties, because it suffers from excessive slotting induced time-harmonics. This caused the sensorless position estimator to lose the position around the natural speed, which resulted in the loss of synchronization. By implementing an additional notch ripple filter the slotting induced time‐harmonic frequency was filtered out. Then, the controller worked fine and was able to keep the Brushless DFIG in stable operation even around the natural speed. The control system for the Brushless DFIG developed in this project, features the world’s first practical and fully functional control system with sensorless position estimator for the Brushless DFIG.
25. Comparison of drivetrain dynamic response:
The purpose of this work is to investigate predicted force and deflection spectra of components within the three candidate drivetrains under investigation. It is known that different wind turbine designs can exhibit very different dynamic characteristics, and so the comparisons made between these three technologies will enable valuable insight into how the different drivetrain options behave.
The vibrational characteristics of three contrasting wind turbine drivetrain designs, containing a range of different generator types (Doubly-Fed Induction Generator, Brushless Doubly-Fed Induction Generator, and Permanent Magnet Generator) have been analyzed. Parameter sensitivity studies have been carried out on all three drivetrains which showed them to be relatively robust to changes in both mounting stiffness and approximated magnetic stiffness.
For the two drivetrain designs containing gearboxes (Drivetrain 1 and Drivetrain 2), excitation sources internal to the same have been identified, and compared with the predicted frequency response functions of their respective generators. This analysis shows that there is a real risk of transmission error-induced vibration affecting the generator, potentially leading to complex interactions with the electromagnetic properties of said generators.
Additionally for the novel BDFIG generator, simulated ripple-torques and unbalanced magnetic pull results were analyzed, and also compared with the frequency response functions. These electromagnetic-induced vibrations were also found to lie close to resonant frequencies of the system, which is a potential concern for gearbox longevity. It is recommended that future work designing a gearbox for the BDFIG generator takes these effects into account.
Analysis of the direct-drive Permanent Magnet machine was confounded by the small number of mechanical components. The resulting analysis results are therefore dominated by the turbine rotor blades and tower stiffnesses.
26. Design documentation:
Finally, all the tools developed for the design and modelling of the Brushless DFIG and its associated converter and control system have been documented so that they can be utilized by the SMEs during the follow-on exploitation phase. The software codes have been copied onto safe and password protected external memories and handed over to the SMEs. Notes have been provided in the codes/programs so that they can be easily understood and potentially modified by operators in future.
Potential Impact:
WINDRIVE supports the exploitation of a more reliable and cost-effective generator system for the wind industry. Improved reliability and cost-effectiveness will inherently lead to wider adoption of wind power generation, especially in offshore, which will boost the low-carbon economy, reduce CO2 emissions and increase energy security for the EU. Wind power development has been a top priority for the EU and Member States, and the targets for 2020 can only be achieved if wind power gains momentum without relying on substantial government subsidies. This can be realized through technological advancement to reduce the capital and O&M costs of wind turbines and enhance their availability through greater reliability. The Brushless DFIG technology offers a significant contribution to achieve this. Further, by eliminating brushes that are carbon-based, the new technology reduces their usage.
The exploitation of the Brushless DFIG drivetrain will support the reduction of CO2 emission, estimated as much as 224 million tonnes in 2022. The followings describe the calculation methodology which is based on the guidelines suggested by the Carbon Trust. The assumptions include:
1. The CO2 emission reduction factor is assumed to be 0.43 kgCO2/kWh for a wind turbine;
2. A wind turbine has a size of 3 MW with a lifespan of 20 years;
3. A capacity factor of 30% is assumed for a wind turbine, which represents the intermittent nature of the wind, the availability of the wind turbines and array losses;
4. A saving of EUR 475,000 can be achieved by utilizing a Brushless DFIG drivetrain (with a simplified 2-stage gearbox) instead of a DFIG drivetrain in an onshore wind turbine over its lifespan, estimated by a study carried out by an independent consultant employed by the Carbon Trust. The saving for an offshore wind turbine with a same size is assumed to be three times greater.
5. The capital cost of onshore and offshore wind farm development is assumed to be 1,150 EUR/kW and 2,500 EUR/kW respectively.
6. From points 4 and 5, with the cost savings achieved by employing the Brushless DFIG technology in 7 onshore 3 MW wind turbines or 5 offshore 3 MW wind turbines, an extra wind turbine can be deployed.
According to the figures published by EWEA, the total installed capacity of onshore and offshore wind power is expected to reach 210 GW and 50 GW by 2022, respectively. It is expected that between 2017 and 2022, during which the exploitation of Brushless DFIG drivetrain will occur, the EU will install 50 GW of new onshore and 30 GW of new offshore wind farms. We anticipate that 1.5 % of the new wind turbines installed between 2017 and 2020 will incorporate the Brushless DFIG drivetrain, which corresponds to 400 3 MW wind turbines, of which 250 turbines will be installed onshore, and hence 150 offshore, each rated at 3 MW. From point 6 above, the utilization of the Brushless DFIG architecture will therefore enable extra 36 onshore and 30 offshore wind turbines to be deployed. This is equivalent to an overall saving of EUR 347 million. The annual ‘extra’ CO2 emission reduction by deploying the extra wind turbines can be calculated as:
66(turbines) x 3000(kW) x 8760(days pa) x 0.3(capacity factor) x 0.43(kgCO2/kWh) = 223,745,038 tonnes.
Improving the strategic competitive position of the participant SMEs
The WINDRIVE project provides its SME participants with specific knowledge, technology and products that will enable them to further differentiate and strengthen their competitive position relative to global competitors. Strategic competitive developments will be achieved in the following key areas:
• Direct economic business growth through exploitation of the WINDRIVE outcomes, enabling further investment, expanding the SMEs’ skill base, technology portfolio and production capacity;
• New knowledge in the application of the SMEs technology, products and services; thereby enabling value-added services and deeper integration within end-user research and development activities;
• Protection of IPR and development of own innovative products; enabling expansion and further development of the SMEs activities towards higher value, sustainable and more profitable business growth (transition further up the supply chain and revenue generation through exploitation of IPR);
• Entrance or improved credibility within the wind turbine market.
Each SME has identified the specific competitive and economic benefits that will result from:
1. Their specific activities within the supply chain which will lead to increased business growth, increased profits and new employee recruitment;
2. Their IPR ownership for a range of new products which will enable the SME to expand its company product portfolio and/or license the technology for other applications;
3. The specific knowledge that they acquire as a result of the WINDRIVE project, thereby enabling deeper integration within supply chain R&D activities.
IPR benefits to the SME partners
Wind Technologies:
- New design features of the Brushless DFIG (patent/know-how);
- New rotor winding design of the Brushless DFIG (patent/know-how);
- New methods in minimizing the unbalance magnetic pull (know-how);
- New designs for cost-effective manufacturing (know-how);
- New methods for magnetic circuit design, analysis and optimization of Brushless DFIG (know-how);
- New optimization methods for core losses in the Brushless DFIG (know-how);
- New optimization methods of rotor performance (know-how);
- New LVRT control algorithms for the Brushless DFIG (patent/know-how);
- New encoderless control algorithms for the Brushless DFIG (patent/know-how).
Romax Technology:
- New design of integrated simplified gearbox for the drivetrain (patent/know-how);
- New optimized drivetrain design for the Brushless DFIG (patent/knowhow);
- New methods for improved dynamics of the drivetrain (patent/know-how);
- New modelling approach for the Brushless DFIG drivetrain (know-how);
- New reliability models for the Brushless DFIG drivetrain (know-how);
Mita:
- New control algorithms optimized for operation of Brushless DFIG drivetrains (know-how);
- New methods of maximizing stability and dynamics of the Brushless DFIG drivetrain (know-how);
Speedgoat:
- New cost-effective and reliable design of control hardware (know-how);
- New method for embedding of control software for protection of IP (know-how);
Improving the competitiveness of the products and services sold by the SMEs
The Brushless DFIG drivetrain is a disruptive technology which, once proven, can potentially address a substantial market opportunity addressed by the conventional DFIG technology. WINDRIVE is a stepping-stone in the commercialization of the new technology and the outcomes of the project give a significant competitive edge to the participating SMEs since the Brushless DFIG drivetrain is a totally innovative concept. WTL will benefit from the exploitation of the core technology, i.e. the generator and its control system, and the other SME partners who are part of the drivetrain supply chain will benefit from gaining the expertise of the other components which are specifically designed and optimized for the Brushless DFIG concept. WINDRIVE will therefore provide them with an immediate competitiveness over the other competitors since they will offer a solution specifically optimized for the Brushless DFIG application which is not available to others.
Societal Impacts
Given the crucial timeframe up to 2020 during which global emission must start to decline, the speed of deployment of wind farms is of key importance in combating climate change. Building a conventional power plant can take 10 or 12 years or more, and until it is completed, no power is being generated. Wind power deployment is measured in months, and a half completed wind farm is just a smaller power plant, starting to generate power and income as soon as the first turbines are connected to the grid. There are no fuel costs, no geo-political risk and no supply dependence on imported fuels from politically unstable regions. Every kilowatt/hour generated by wind power has the potential to displace fossil fuel imports, improving both security of supply and the national balance of payments. It seems evident that wind farm spreading have many legislation and policy impacts as well. It may strongly impact (and will impacted by) security policies, energy policies, renewable energy, climate, sustainable development, regional, environmental, aviation, economic and research policies and strategies.
By the end of the year 2010, about 670,000 persons were employed worldwide directly and indirectly in the various branches of the wind sector. Within five years, the number of jobs almost tripled, from 235,000 in 2005. There is an increasing demand for a very broad range of jobs, from engineers, skilled workers to mangers, financial, environmental and legal experts. WINDRIVE will lower the cost of wind power and hence will contribute to the growth of the wind industry and creating new jobs.
Environmental Impacts
In March 2004 Greenpeace published “SeaWind Europe”, which examines the ramifications –environmental, technical, social and economic – of offshore wind providing 30% of the EU’s electricity. While ambitious, it shows what could be possible with the vision and ambition to develop renewable energy on such a scale. Greenpeace says that “success on this scale would not only deliver enormous environmental benefits from this clean, safe energy source, but would also generate an economic boom in Europe worth hundreds of billions of Euros and create up to 3 million jobs.”
This statement was as encouraging then as it is now and the real potential of wind technology on the future stabilisation of power generation is only now really being shown to have wide benefits and implications in “keeping the lights on”. The balance, and trade off, of this technology has been, is being, and will continue to be argued by the “fors” and “againsts”, who will state the added benefits and those that voice their concerns as the environmental impact and possible damage that may or not be done. WINDRIVE does not want to be seen entering into this argument or taking a position. It is felt by the parties involved in this project that a more balanced and even approach to the entire issue is in the duty of care that any EU funded project needs to address issues relating to environmental impact and the regulatory standards that need to be adopted. In this regard, WINDRIVE will look to assist regulators in developing those standards and approvals and enhancing and mitigating any environmental potential impact that might arise. In the case of wind technology, both on-shore and off shore, the discussion on regulation and also standards is on-going.
Since 2001, EU Directive 77/2001/EC, which promotes electricity from renewable sources, which included an indicative target for 21% of the EU's overall electricity to come from renewable energy sources by 2010, gave an important boost to the sector, although just over 18% were actually reached by that deadline. At the end of 2010, the EU’s new Renewable Energy Directive (2009/28/EC) entered into force, setting an EU renewable energy target of at least 20% of final energy consumption by 2020. Each EU Member State has a national legally binding target for the share of renewable energy it must achieve by that date. The member states were required to submit National Renewable Energy Action Plans (NREAPs) detailing sectoral targets and measures necessary for them to reach their overall binding RES target. In their NREAPs, 15 member states say they will exceed their targets, ten consider they will meet their targets, and only two expect to not meet their target domestically. Consequently, aggregating the NREAPs, the EU is expected to exceed its 2020 20% target by 0.7%.
Electricity from renewables will play the largest role in meeting the 2020 target, covering over a third of total electricity consumption. With a capacity forecast of 213 GW, of which 43 GW will be offshore, producing 495 TWh of electricity covering about 14% of total consumption, wind comes out as the technology of choice in the NREAPs. The European Wind Energy Association (EWEA)’s baseline scenario assumes a total installed capacity of wind power in the EU by 2020 of 230 GW, producing 582 TWh of electricity, meeting 14.2% of EU electricity demand (depending on demand growth over that period). EWEA’s high scenario assumes that total installed wind power capacity will reach 265 GW by 2020, producing 681 TWh of electricity and meeting 16.7% of the EU’s electricity demand by 2020.
Transnational approach
The WINDRIVE consortium has been possible through a trans-national approach. The need for a wider strategy is increasing due to the increasing demand for innovative green technologies within Europe and globally. This trans-national approach combines the knowledge and experience from a committed consortium enabling delivery of this challenging project which will produce distinct advantages and environmental benefits when applied to wind farm generators. Only through the trans-national combination of these skills, can we increase competitiveness and establish a critical mass of experience within the participants who would otherwise remain a fragmented resource.
Within the supply chain, SMEs make up the large majority of companies offering innovative products. However, innovation within these companies is limited by the small R&D budget or lack of resources, both of which are common to SMEs. Indeed this is also true for the engineering SMEs who supply into this market. Routine customer specific projects leave little room for innovation and ability to develop partnerships on a pan- European level. WINDRIVE facilitates this process by establishing new technology and paving the way for a broader integration between partners. The WINDRIVE consortium brings together an expert core group of SMEs with research organisations who provide word class research and product differentiators.
Plan for the use and dissemination of foreground
Work package 8 was included within the project to concentrate specifically on the dissemination and exploitation of the project results. All project partners recognise and are strongly committed to maximising the dissemination and exploitation of the WINDRIVE results.
The impact of the WINDRIVE project on the state of the art in the field of wind power generation will rely on wide and effective dissemination of the results gathered. WINDRIVE have applied efforts to convey its knowledge and technologies to relevant industrial and academic communities as well as the general public. The WINDRIVE consortium concentrates the expertise of several adjacent and complementary research fields and thus is very well placed to stimulate new methodological paradigms within the partners.
Effective communication of the results of the cross-disciplinary investigations and their promotion to specialists from diverse backgrounds will maximise the project’s impact. The dissemination activities lay the basis for successful uptake and commercial exploitation of the project’s technological output. Dissemination strategies which have been carried out during the WINDRIVE project included:
1. Development of an interactive project website;
2. Development of a Wikipedia page;
3. Development of a video clip and its publication on YouTube;
4. Publications of non-confidential research outcomes in international journals and conferences;
5. Presentation at international conferences and exhibitions and engagement with industrial experts at such events.
6. One-day workshop for invited attendees from a wide range of audience including industrial and academic experts and participants from government and NGOs.
The above will be described in detail in the next section.
Main Dissemination Activities
1. Project website
The project website has been developed to engage with public about the outcomes of WINDRIVE and can be viewed at www.bdfig.com. The name of the website is the acronym for the main technology that WINDRIVE researches, the Brushless Doubly-Fed Induction Generator (BDFIG). The website includes the brief summary of the project, detailed description of the project and its objectives, publication list, information about the consortium partners, latest news and contact details. Any contact made via the website will be forward to the Coordinator who will respond or refer the message to the relevant partner.
2. Project Wikipedia page
A Wikipedia page has been created for WINDRIVE which is open to public and can be accessed at https://en.wikipedia.org/wiki/Windrive. The page contains a brief summary of WINDRIVE, main objectives of the project, list of project consortium partners and a link to the project website.
3. Video clip
A video clip has been designed, with length 4:49 and can be viewed on YouTube: https://www.youtube.com/watch?v=OvodJYWNZMk. The video clip gives information on the project and its aim and shows the prototype wind turbine located in Cambridge, UK. It also shows the laboratory at Cambridge University where a small scale generator is located. The consortium partners are mentioned in the video clip.
4. WINDRIVE publications
There are several publications from the outcomes of WINDRIVE which have been published in leading international conferences and journals. The list of publications is provided below.
Title: Effects of rotor winding structure on the BDFM equivalent circuit parameters
Authors: A. Oraee, E. Abdi, S. Abdi, R. McMahon, P.J. Tavner
Published: IEEE Transactions on Energy Conversion, DOI: 10.1109/TEC.2015.2432272
Title: A study of unbalanced magnetic pull in brushless doubly fed machines
Authors: S. Abdi, E. Abdi, R. McMahon
Published: IEEE Transactions on Energy Conversion, DOI: 10.1109/TEC.2015.2394912
Title: Brushless Doubly-Fed Induction Machines: Torque Ripple
Authors: T. D. Strous, Xuezhou Wang, H. Polinder, and J. A. Ferreira
Published: in International Electrical Machines & Drives Conference (IEMDC), May. 2015
Title: Finite Element Modelling of Brushless Doubly-Fed Induction Machine based on Magneto-Static Simulation
Authors: Xuezhou Wang, T. D. Strous, D. Lahaye, H. Polinder, and J. A. Ferreira
Published: in International Electrical Machines & Drives Conference (IEMDC), May. 2015
Title: Finite Element based Multi-Objective Optimization of a Brushless Doubly-Fed Induction Machine
Authors: T. D. Strous, Xuezhou Wang, H. Polinder, and J. A. Ferreira
Published: in International Electrical Machines & Drives Conference (IEMDC), May. 2015
Title: Effect of Rotor Skew on the Performance of Brushless Doubly-Fed Induction Machine
Authors: Xuezhou Wang, T. D. Strous, D. Lahaye, H. Polinder, and J. A. Ferreira
Published: in International Electrical Machines & Drives Conference (IEMDC), May. 2015
Title: LVRT Performance of Brushless Doubly-Fed Induction Machines – A Comparison
Authors: U. Shipurkar, T. D. Strous, H. Polinder, and J. A. Ferreira
Published: in International Electrical Machines & Drives Conference (IEMDC), May. 2015
Title: Converter rating optimisation for a brushless doubly-fed induction generator
Authors: A. Oraee, E. Abdi, R McMahon
Published: IET Renewable Power Generation, vol. 9, No. 4, 2015
Title: Equivalent circuit parameters for large brushless doubly-fed machines (BDFMs)
Authors: S. Abdi, E. Abdi, A. Oraee, R. McMahon
Published: IEEE Transactions on Energy Conversion, vol. 29, No. 3, 2014
Title: Vector-controlled grid synchronization of the brushless doubly-fed induction generator
Authors: A. Broekhof, M. Tatlow, R. McMahon
Published: 7th IET Int. Conference on Power Electronics, Machines and Drives (PEMD2014), pp. 1-6, 2014, Manchester, UK
Title: Design of the Brushless DFIG for Optimal Inverter Rating
Authors: E. Abdi, A. Oraee, S. Abdi, R.A. McMahon
Published: 7th IET Int. Conference on Power Electronics, Machines and Drives (PEMD2014), pp. 1-6, 2014, Manchester, UK
Title: Brushless Doubly-Fed Induction Machines: Magnetic field modelling
Authors: T. D. Strous, N. H. van der Blij, H. Polinder, and J. A. Ferreira
Published: in Conference on Electrical Machines (ICEM), Sep. 2014, pp. 2702–2708
Title: A novel analytical approach and finite element modelling of a BDFIM
Authors: N. H. van der Blij, T. D. Strous, Xuezhou Wang and H. Polinder
Published: in International Conference on Electrical Machines (ICEM), Sep. 2014, pp. 346–352
Title: Crowbarless Fault Ride-Through of the Brushless Doubly Fed Induction Generator in a Wind Turbine Under Symmetrical Voltage Dips
Authors: T. Long, S. Shao, P. Malliband, E. Abdi, R. McMahon
Published: IEEE Transactions on Industrial Electronics, vol. 60, No. 7, pp. 2833-2841, 2013.
Title: Dynamic Modelling of the Brushless Doubly-Fed Machine
Authors: P.C. Roberts, T. Long, R. McMahon, S. Shao, E. Abdi, M. Maciejowski
Published: IET Electric Power Applications, vol. 7, No. 7, pp. 544-556, 2013
Title: A Study of Converter Rating for Brushless DFIG Wind Turbines
Authors: A. Oraee, E. Abdi, S. Abdi, R.A. McMahon
Published: In Proceeding of the 2nd Renewable Power Generation Conference, September 2013, Beijing, China
Title: Design and Performance Analysis of a 6 MW Medium-Speed Brushless DFIG
Authors: E. Abdi, M.R Tatlow, R.A. McMahon, PJ. Tavner
Published: In Proceeding of the 2nd Renewable Power Generation Conference, September 2013, Beijing, China
Title: Investigation of Magnetic Wedge Effects in Large-Scale BDFMs
Authors: S. Abdi, E. Abdi, A. Oraee, R.A. McMahon
Published: In Proceeding of the 2nd Renewable Power Generation Conference, September 2013, Beijing, China
5. Presentations at International Conferences and Exhibitions
The consortium partners presented at several international conferences, exhibitions and workshops, including the followings:
Name of event: European Wind Energy Association’s Annual Event
WINDRIVE Partners attended/presented: WTL, Romax, Mita, Senvion, UCAM, DUT
Location: Vienna, Austria
Date: 4-7 February 2013
Name of event: International Conference on Electrical Systems for Wind Turbines
WINDRIVE Partners attended/presented: Romax
Location: Bremen, Germany
Date: 13-15 May 2013
Name of event: Renewable Power Generation Conference
WINDRIVE Partners attended/presented: WTL, UCAM
Location: Beijing, China
Date: 9-11 September 2013
Name of event: European Wind Energy Association’s Annual Event
WINDRIVE Partners attended/presented: WTL, Romax, Mita, Speedgoat, Senvion, UCAM
Location: Barcelona, Spain
Date: 10-13 March 2014
Name of event: International Conference on Power Electronics, Machines and Drives
WINDRIVE Partners attended/presented: WTL, UCAM
Location: Manchester, UK
Date: 8-10 April 2014
Name of event: American Wind Energy Association WINDPOWER Conference & Exhibition
WINDRIVE Partners attended/presented: Mita
Location: Las Vegas, United States
Date: 5-7 May 2014
Name of event: International Conference on Electrical Machines
WINDRIVE Partners attended/presented: DUT
Location: Berlin, Germany
Date: 2-5 September 2014
Name of event: Wind Energy O&M Forum Europe
WINDRIVE Partners attended/presented: Romax
Location: Hamburg, Germany
Date: 20-21 January 2015
Name of event: European Wind Energy Association’s Offshore Wind Conference
WINDRIVE Partners attended/presented: WTL, Romax, Mita, Senvion, UCAM, DUT
Location: Copenhagen, Denmark
Date: 10-12 March 2015
Name of event: International Electrical Machines and Drives Conference
WINDRIVE Partners attended/presented: DUT
Location: Coeur d’Alene, United States
Date: 10-13 May 2015
Name of event: International Symposium on Industrial Electronics
WINDRIVE Partners attended/presented: WTL, UCAM
Location: Rio de Janeiro, Brazil
Date: 3-5 June 2015
6. One-day Workshop
A one-day workshop was held in London on 5th March 2015. The attendees included invited experts from the wind power industries (supply chain, wind turbine OEMs, developers, operators and energy companies), academic institutes (university professors and researchers), governmental organisations (UK Department of Energy and Climate Change, Carbon Trust, Energy Institute, etc) and non-governmental organisations (Institution of Engineering and Technology, Cleantech Forum, etc).
The event included several presentations and Q&A sessions covering several aspects of WINDRIVE outcomes, including the followings:
1- Advances in electrical generator design for wind power industry;
2- Drivetrain topologies and understanding the trade-offs;
3- Wind turbine operations and maintenance for different drivetrain topologies;
4- Potentials of WINDRIVE drivetrain for offshore wind;
5- Wind turbine reliability and WINDRIVE contribution;
6- Economics of WINDRIVE technology and reduction on levelised cost of energy.
Dissemination Activities beyond the Life of WINDRIVE
The dissemination activities will continue beyond the life of the project to further maximise the impact of the project. These activities will include:
• Identification of target audiences and dissemination plans to ensure visibility and awareness of the achievements of WINDRIVE and enable effective knowledge transfer.
• Continued update of the project web portal to: (i) organise collaborative activities between project partners and disseminate information and announcements in the consortium, and (ii) distribute WINDRIVE non-confidential outputs among outside visitors, and to collect feedback from potential users.
• Scientific dissemination by the academic partners through international publication/ presentation at relevant conferences, workshops and seminars.
• Development of lectures and information packages for delivery within higher education facilities.
Publication and dissemination of the project knowledge and results will be conducted in agreement with the consortium partners. Careful consideration to timing will be maintained to ensure that the benefits of swift dissemination are achieved whilst safeguarding the intellectual property rights, confidentiality and legitimate commercial interests of the project SME partners.
The consortium have engaged, and will continue to do so, with relevant European Activities, including:
• European Wind Energy Technology Platform (TPWind);
• The European Wind Energy Association (EWEA) and various national associated organizations.
List of Websites:
www.bdfig.com