Review of Energy Conversion System for Large Wind ...

Review of Energy Conversion System for Large Wind Turbines Shrestha G., Polinder H., Bang D.J., Ferreira J.A. Electrical Power Processing Group, Delft University of Technology Mekelweg 4, 2628CD, Delft, Netherlands Phone: 0031-15-2782955, Fax: 0031-15-2782968 [email protected], [email protected], [email protected], [email protected]

Abstract Wind turbines in the future will be more placed at offshore locations. This will require the wind turbines to be more reliable and large single units will be favorable for cost reasons. This paper will review the various energy conversion system available in the market based on the upscaling and reliability. The review shows that present state of art favors the single geared version in terms of cost of energy. Direct drive machines have very attractive attributes for future wind turbines but the weight of the machine should be reduced to make them competitive. 1. Introduction There has been a tremendous increase in wind power in the present decades. Wind turbines represent an integral part of the electricity network. This trend is expected to increase in the coming decades. The installed capacity of wind turbines till the end of 2006 was about 74.3GW. The year 2006 had the highest installation of wind power ever with 15GW of power added which is 30% more than the year 2005. An increase of 17.4% per year in new installation is expected till the year 2011 [8]. Figure 1 shows the trend in wind turbine installed worldwide. In Europe wind power produces about 3.3% of the electricity consumption in EU [9]. Similarly the size of the units has increased. Figure 2 shows the growth of size of wind turbine from 1980 when the commercialization of modern wind turbine began till the year 2005.

Figure 1: Wind power development in the world [8]

Figure 2: Increase of wind turbine size from 1980-2005 [Bundesverband WindEnergie e.V]

The technology for wind energy conversion system has also seen various changes in the past two decades. Various trend and development of the energy conversion system has been initiated because of three fundamental goals: 1. to achieve lower cost of energy (high efficiency, low mechanical loads, up-scaling, reliability) 2. to achieve better power quality and grid connection (flicker problem, grid connection requirement) 3. public acceptance (noise, visual effect) The trend now and in the future is towards large offshore wind farms. This trend has led to demand for high reliability of wind turbine and larger single unit of such wind turbine. This is mainly due to the difficulty and cost of maintenance in offshore wind turbine. Some estimate shows the cost of

maintenance in offshore wind turbine to be double that of the onshore ones [39]. Similarly the high cost of foundation and other auxiliaries demand large single units for economic reasons. This paper will review various concepts, configuration and topologies used by modern wind turbine manufacturer. The advantages and disadvantages are outlined for these configurations. Then some relevant comparisons are reviewed to find configurations that are better suited future wind turbines based on upscaling and reliability. Then some conclusions are drawn based on the results from the comparison. 2. Various drive train and control concept There are various drive train concepts for energy production in wind turbines. Based on the control concept used they can be categorized as follows: 1. Fixed speed 2. Semi Variable speed 3. Variable speed Within these concepts there are various configuration used by various manufacturer. All the concepts used by the recent top ten manufacturers [8] in the year 2006 and other existing configuration are presented in table 1 & 2. a. Fixed speed The drive train of a fixed speed wind turbine has a multiple stage gearbox in between the shaft and the generator to convert the huge torque in the shaft to a higher speed at the generator. A squirrel cage induction generator with 4 or 6 poles is used in this concept and is directly connected to the 50 or 60 Hz grid via a transformer. The fixed speed wind turbines use the stall control method to control the speed of the turbine. The shaft rotates at a fixed speed below and at the rated wind speed. Once the wind speed is over the rated wind speed, the power is limited by the stalling effect of the rotor blades. This concept is also referred as a ‘Danish concept’ because this concept was used by Danish wind turbine manufacturer in the 1980-2000. The grid connected SCIG draws reactive power from the grid therefore they incorporate capacitor banks for reactive power compensation. Soft starter are used in these wind turbines for smooth connection to the grid. Some of the manufacturers are listed in table 1. The advantage of such wind turbine is its simplicity, robustness. However there are several disadvantages of such a system. They are • The fluctuation in wind speed and tower shadow is converted directly to electromechanical torque variations, which causes mechanical stress and fatigue to the components of the wind turbine and cause high flicker in the output voltage. • A gearbox is one component that requires the most maintenance. • Grid support during faults is not possible. A pole changeable (4-6 poles) SCIG has been used by some manufacturers, which enables the wind turbine to operate at 2 different speeds. They use either pitch or active stall principle to control the wind turbine. This will increase the efficiency of the wind turbine at low speeds as it can run at different tip speed ratio at low or high winds respectively. Using the active stall control, the wind rotor speed can be controlled to two different fixed speeds for different power generation. Manufacturer of this concept is listed in table 1. This concept is not suitable for large wind farms primarily because its incapability to support the grid. b. Semi Variable speed: This concept uses a multiple stage gearbox with a WRIG. This concept is also a trademark concept of Vestas named Optislip®. The rotor of the generator is connected to an electronically controllable resistance unit which can be varied so that the slip of the induction generator is varied. Variable

2

speed can be achieved by controlling the energy extracted from the rotor. A typical variable speed range achieved from this concept is 10% above the synchronous speed. A higher value can be achieved at the cost of lower efficiency because more energy extracted from the rotor means more power needs to be dissipated. Reactive power compensation and soft starter should be added to the design to connect it to the grid. These wind turbines can stay connected to the grid during grid faults; however they cannot support the grid during grid faults. Only Suzlon from the top ten manufacturers still has this concept in its product range. This configuration helps to reduce mechanical loads during wind gusts, by letting the rotor accelerate to higher speeds. This concept is not suitable for future wind turbines because of its incapability to support the grid as the primary reason.

Table 1: List of the top ten manufacturers in the year 2006 and their product range [8] Manufacturer

% share

Vestas (DK)

28.2%

Gamesa (ES)

15.6%

GE Wind Energy (US/D)

15.5%

Concept /Control FS/AS VS/PS VS/PS

Generator/ Converter PSCIG/DG DFIG/PC DFIG/PC

Gear

Power range

Voltage level 690V, 1000V

MG MG MG

1.65 MW 0.85-3 MW 850 kW - 2 MW

VS/PS VS/PS VS/PS

DFIG/PC PMSG/FC EESG/DG

MG MG DD

1.5– 3.6 MW 2.5-3.0MW 300 kW – 4,5 MW

FS/CS SV/PS FS/AS VS/PS

SCIG/DG WRIG/DG PSCIG/DG SCIG/FC

MG MG MG MG

350 kW -1.25 MW 2 MW 1.3–2.3 MW 2.3-3.6MW

690V

690V ***

Enercon (D)

15.4%

Suzlon (IND)

7.7%

440V

Siemens/ Bonus (D/ DK)

7.3%

Nordex (D)

3.4%

FS/AS VS/PS

PSCIG/DG DFIG/PC

MG MG

1.3 MW 1.5 – 2.5 MW

660V

Repower Systems (D) Acciona (ES)

3.2% 2.8%

VS/PS VS/PS

DFIG/PC DFIG/PC

MG MG

1.5 – 5 MW 1.5MW

690V 12kV

Goldwind (CN)

2.8%

FS/AS VS/PS VS/PS

PSCIG/DG DFIG/PC PMSG/FC

MG MG 1 DD

600kW 750kW 1.2-1.5MW

690V

690V

1

DD : Outer rotor type

Table 2: Some manufacturer with alternative drivetrain configuration in the market Dewind (UK/D) Clipper

VS/PS VS/PS VS/PS

DFIG/PC EESG/PC PMSG/FC

MG MG/HC MP

1.2-2MW 2MW 2.5MW

4.16, 13.8kV 690V

Harakosan

VS/PS

PMSG/FC

DD

1.5/2 MW

3kV

Multibrid **Jeumont **MADE

VS/PS VS/PS VS/PS

PMSG/FC (A)PMSG/FC EESG/FC

SG DD MG

5MW 0.75-2MW 2MW

3kV *** 1kV

*All machines are radial flux machines unless otherwise stated. ** Discontinued FS/CS: Fixed speed /classic stall DD: Direct drive SV/AS: Semi Variable (Extended slip)/Active stall MG: Multiple stage gear FS/AS: Fixed speed/ Active stall SG: Single stage gear FS/PS: Fixed speed/ pitch control PSCIG: Pole changing squirrel cage induction generator RF: Radial Flux DFIG: Doubly fed induction generator HC: Hydraulic Converter WRIG: Wound rotor induction generator EESG:Electrically excited synchronous generator PC: Partial converter (A)PMSG:(Axial)Permanent magnet synchronous FC: Full converter generator DG: Direct grid connection SCIG: Squirrel cage induction generator MP: Multiple path generators VS/PS: Variable speed/Pitch control

3

c. Variable speed wind turbines Most wind turbines with a power rating of more than 1.5 MW are variable speed wind turbines. The switch to variable speed has been due to the following factors. • The power that can be extracted from wind is a function of the power coefficient Cp as given in equation 1, which is a dependent on the tip speed ratio which is a ratio of tip speed of the turbine to the wind speed and the pitch angle. So varying the speed of the wind turbine and blade pitch to give an optimal power coefficient will increase the energy yield [2] [10]. This equation is valid for all wind turbine configurations. 1 (1) P = ρ air C p (λ ,θ )πrr2 vw3 2 • The power and torque peaks caused by wind gusts, tower shadow can be dampened by letting the rotor accelerate, which will reduce the mechanical and fatigue loads to the components of the wind turbine such as gears, blades, tower. • The flicker problem seen in constant speed wind turbines is reduced. • The grid connected wind turbine should stay connected during grid faults and should assist in the recovery of the grid, which are prescribed by the grid operators. This voltage-dip ride through capability requires the control of active and reactive power. All variable speed wind turbine can support the grid. All variable speed wind turbines manufactured are pitch controlled type. Between the cut-in wind speed and the rated wind speed, the speed of the rotor changes to maintain an optimal tip speed ratio for maximum energy yield. Over the rated wind speed, the blade pitches to decrease the power coefficient. The voltage and frequency control is possible due to the use of power electronics in variable speed wind turbines (exception Dewind 8.2). Variable speed wind turbine has been the focus of research from the 1990’s onward. Six different configuration of such wind turbines exist in the market.

converter

M.Gear generator

generator

e. Direct drive with full converter a. Fixed speed v. resistor S.Gear

converter

M.Gear generator generator

f. Single stage gear with full converter

b. Semi variable speed converter M.Gear M.Gear

Torque converter generator

g. Multiple stage gear with without converter

generator c. Multiple stage gear with partial converter (DFIG)

Squirrel cage or PMSG

M.Gear

converter converter

Bull gear

converter

converter

converter

generator d. Multiple stage gear with full converter

generator h. Multiple generator with full converter

Figure 3: All the present concepts and configuration of wind turbines.

4

i. Multiple stage gear with partial converter (DFIG) This configuration uses a multiple stage gear, doubly fed induction generator with the rotor of the generator connected to a partial rated converter. This concept is an extension of the Optislip concept. The wound rotor is connected to the power converter instead of a variable resistor. So the energy extracted from the rotor is not dissipated but fed to the grid using the power electronic converter. This configuration can also feed reactive power to the grid due to its power electronics in the rotor circuit. The variable speed range of such wind turbines are about 30% below and above the synchronous speed. The rating of the power converter which is a back to back voltage source converter is 25-35% of the generator capacity [10], [36], which made this an economical choice compared to the full converter design of most other variable speed wind turbines. Some major manufacturers using this technology are listed in table 1. However there are several disadvantages of this type of conversion system. • A multiple stage gearbox is prone to failure. Gearboxes are one of the components that fail the contributing significantly to cost and downtime [35]. Upscaling the wind turbine means that the wind turbine rotor speed decreases as power level increases. So either the gear ratio or the number of stages in the gears must be increased to match up with the generator speed. Thus upscaling can have negative effect to the reliability of gearboxes of larger wind turbines. • Slip rings are used to transfer rotor power to the converter. Slip rings require maintenance and can cause failure of the wind turbine. • A grid fault can induce high voltage and current in the rotor. To stay connect to the grid during grid faults require more complicated control strategies [10]. The wind turbine should be connected to the grid with 85% drop in voltage and when the voltage is above the red line in figure 4.

Figure 4: A voltage dip that wind turbine should handle without disconnection (E-on Netz). ii. Multiple stage gear with full converter: This concept uses squirrel cage induction generator (Siemens), the permanent magnet synchronous generator (GE) or electrically excited synchronous generator (MADE) with a full scale converter. A multiple stage gearbox is used to increase the rotational speed. The gear ratio is divided into 3 stages commonly. Compared to DFIG this configuration has no slip rings but has a full scale converter which is more expensive and has more losses than a partial rated converter. The prices of the converter has fallen over 10 folds in the last decades, therefore the cost of power electronics is not the most significant cost carrier. Grid fault ride through is easy as the generation side is independent of the grid side. But as the converter losses are higher, the overall losses are higher compared to DFIG. iii. Direct drive with full converter This concept has no gears and runs at the same speed as the wind turbine rotor. This low speed means that a high torque is required to produce the same power, which translates into large machine. Direct drive wind turbine has few mechanical parts compared to any other variable speed

5

wind turbine configuration so this also leads to reliable operation. The overall efficiency of the system is improved by the omission of the gears. Various topologies exist within the direct drive machine, depending on the type of excitation and geometry. Electrically excited direct drive machines with radial flux path is manufactured by Enercon and is the most used direct drive machine. Permanent magnet machines are getting the attention of manufacturer owing to various factors such as: • the power control capability of the full converter makes the electrical excitation unnecessary. • The losses are reduced as the losses for coil excitation is not necessary. • Permanent magnet technology is getting cheaper. • It allows making pole pitches small, so are more compact compared to electrical excited machines. • Permanent magnet machines are lighter and more efficient than electrically excited machines. Many new manufacturer of wind turbines are coming to the market with permanent magnet direct drive machines Avantis, Leitwind, Vensys, Goldwind, Scanwind, Harakosan, Darwind and Mitsuibishi. All the permanent magnet machines manufactured till date are radial flux type machines with surface mounted permanent magnets except Jeumont J48 which used an axial flux type permanent magnet machine and Leitwind which uses flux concentrated type machine. Most manufacturers have used the inner rotor type machine whereas Vensys and Goldwind have used the outer rotor type machine. Apart from radial and axial flux type there is a new category called transverse flux machine which has higher force density compared to the other two types of longitudinal flux machines. In principle this should make the machine more compact. But these machines have low power factor and comparison shows that its performance decreases for large diameter machines [4]. The same literature shows the best performance in terms of weight and cost for radial flux machines. A detailed classification on different direct drive machine and comparisons are listed in literature [4], [5], [38]. iv. Single stage gear with full converter This concept uses the single gear to convert the speed to a 1:6-1:7 of the rotational speed of the main shaft [12], and then the permanent magnet machine is used with a full converter. This is also a pitch controlled wind turbine even though the initial concept used a stall control principle to use less moving parts in the wind turbine to increase the reliability [12]. As the power rating of the wind turbines increase, the rotational speed at the shaft decreases. This means that the torque level in the gearbox will increase; making the gearbox larger or an additional stage of gear should be added to the gearbox. It is seen that a conventional design with 3 stage gearbox failures mostly occurs in the high speed components of the gears. This concept is a compromise between the reliability and efficiency of a direct drive wind turbine and the size of the generator i.e. the cost is decreased by increasing the speed of the rotor. This is also known as the Multibrid design. Multibrid and Winwind are the manufacturer of such wind turbines (Table 1). v. Multiple stage gear with hydraulic converter In this configuration a multiple stage gear box is connected to a hydraulic torque converter and this is connected to a high speed generator electrically excited synchronous machine. So basically the speed of the generator is fixed like in other conventional power generators. Due to the electrical excitation the turbine is able to feed reactive power to the grid and control the power. This concept is used by Dewind in its D8.2 model, a 2MW wind turbine. The synchronous machine will run at a fixed speed independent of the rotational speed of the shaft. This is enabled by the combination of hydraulic torque converter and electrically excited synchronous generator. It is similar to the control of conventional power plants, where the guide vanes control the flow of fluid in the hydraulic converter in order to control the torque of the machine. These machines are directly grid connected without a power electronic converter [26]. Any technical advantage of such concept is not foreseen when compared to other variable speed concepts. The losses in the converter are avoided but losses of hydraulic converter and electrical excitation are added. Such concepts can however help the gears from high stresses in case of faults in the grid.

6

vi. Multiple generators with full converter This concept tries to reduce the weight of the wind turbine using a distributed drive train. The torque from the rotor is splitted by a bull gear to parallel secondary pinions connected to parallel generators. This concept is used by Clipper in its Liberty 2.5MW model of wind turbines. 4 permanent magnet generators each of 660kW are used in the design of these wind turbines with full converter. The advantage of such concept is that the torque is divided into smaller units. Similarly the fault in one machine will not stop the operation of the wind turbine.

Figure 5: A drivetrain configuration of Dewind 8.2 wind turbine.[26]

Figure 6: Model of a Clipper energy conversion system. [32]

4. Comparison of various wind turbine energy conversion system Various quantitative comparisons have been done by researchers to find the best configuration [3], [6], [7]&[12]. These surveys have evolved with new concepts, upscaling of the wind turbine size. Here some relevant comparisons are presented for megawatt ranged wind turbines. A rather extensive comparison has been carried out by Polinder et. al. [6] for a 3MW wind turbine. An extract from the results are presented in table 3. A similar comparison is carried out by NPS(Northern power system) [7] for the multi generator system, single stage gear box design and the direct drive generator with a base line design of a DFIG. Some results are presented in table 4& 5. Table 3: Comparision of various configuration of variable speed wind turbine [6] Generator concept Weight of active material (ktons) Cost of generator system (kEuro) Cost of total system (kEuro) Loss of generator system (MWh) Annual energy yield (GWh) Annual energy yield/ Cost of (kWh/euro)

total

system

DFIG 3G 5.25 320 1870 763 7.73 4.13

EESG DD 45.10 567 2117 739 7.88 3.72

PMSG DD 24.1 432 1982 513 8.04 4.05

PMSG 1G 6.11 333 1883 674 7.84 4.16

DFIG 1G 11.37 287 1837 701 7.80 4.25

Table 4: Comparison of various configuration of variable speed wind turbine. [7] Generator concept (1.5MW) DFIG 3G DD 4m DD 5.3m Projected sales price (k$) 1214 1272 1257 Annual energy yield (GWh) 4.77 4.87 4.87 Annual energy yield/ total cost (kWh/$) 3.92 3.83 3.88 Cost of energy (cent/kWh) 3.77 3.71 3.68 Generator concept (3MW) DFIG 3G DD 4m DD 5.3m Projected sales price (k$) 2222 2333 Annual energy yield (GWh) 9.77 9.95 Annual energy yield/ total cost (kWh/$) 4.4 4.26 Cost of energy (cent/kWh) 3.42 3.42 MS-1: Single stage geared (Multibrid design) , MS-6: Distributed drivetrain with 6 generators

MS-1 1204 4.81 3.99 3.68 MS-1 2227 9.84 4.41 3.39

MS-6 1327 4.78 3.37 4.17 MS-6 -

7

Table 5: Component cost in percentage of total component cost [7] DFIG 3G Component

1.5 MW

DD 5.3m

MS-1

3MW

1.5MW

3 MW

1.5MW

3MW 0%

Main Shaft

2.35%

2.4%

0%

0%

0%

Gearbox

11.7%

11.98%

0%

0%

8.3%

9.3%

Generator

6.7%

5.8%

18.3%

22.8 %

6.5%

8.3%

Converter

6.4%

6.6%

12 %

9.8 %

12.5%

10.2%

Rotor

30.3%

26.8%

29.3 %

25.5 %

30.5%

26.7% 23.5%

Tower

23.6%

23.6%

22.8 %

22.5 %

23.7%

Controller

4.4%

3.2%

4.25 %

3.1 %

4.4%

3.2%

Bearing

1.6%

1.2%

3.6 %

2.4 %

2.8%

2.6%

Yaw

2.8%

3.5%

2.7 %

3.3 %

2.8%

3.4%

The results from NPS [7] show a similar trend to the results presented by Polinder [6]. Some inferences from the results are: • The DFIG- 3G are cheapest and lightest solution. (DFIG 1G seems to be even cheaper but this solution has not been implemented by any manufacturer till date) • The energy yield is highest for the direct drive machine. • It also shows that the energy yield per cost is higher for machine built with larger diameter (4m & 5.3m) direct drive machine for the same power. • The COE (cost of energy) calculation for 20 years life time of the wind turbine shows the lowest cost of energy for single stage geared machine with permanent magnet and direct drive machine for 1.5MW and lower for single stage geared machine for 3MW machine. • The COE decreases for larger machine for all machines. • When scaling up the direct drive machine, the cost of energy is not decreasing as steeply as the DFIG or single gearbox system. • When upscaling from 1.5MW to 3MW the cost of generator in terms of percent of total component cost increases for a direct drive machine, whereas cost of other components stays almost the same or decreases. This explains one reason why a direct drive wind turbine COE is not competitive when upscaling the wind turbine even though the energy yield is high and the O&M cost is lower. a. Factors affecting cost of direct drive generator The main reasons for the significant increase in cost of direct drive generators are • From active material point of view, it is normally economical to make generators with large diameter and shorter axial length as shown by equation 2. Radius has a quadratic affect on power output. Constraints of transportation limits the outer radius therefore the axial length should be increased. But the cost of active material is in the first order approximation proportional to the surface area 2πrl . Therefore a longer axial length increases the cost of active material.

T = Fd A.r = 2πr 2lFd •

(2)

From inactive material point of view, for smaller generator, the structural material doesnot play an important role for the cost of generator. As the size of the generator increases, this becomes more significant. As shown in figure 7, the weight of a 10MW direct drive machine is 13 times more than the weight of a 2 MW machine. Most of the weight is structural material. The specific cost of the structural material is much lower than the cost of permanent magnet and copper. But the amount of such material required makes it significant. In a 10MW machine the cost of structural material is about 63% of the total cost of generator.

8

Torque Vs Weight 350 300

Weight Tons

250 Inactive

200

Total 150

Active

100 50 0 0

2

4

6

8

10

12

Torque MNm

Figure 7: Comparison of weight of 2, 3, 5 & 10 MW (in terms of torque rating) direct drive wind turbine. [41][42] b. Factors affecting the cost of energy The COE given are for the onshore wind turbines. But for offshore wind turbine the COE can be different because • The operational and maintenance cost of offshore wind farms are about 2 times the O&M cost on onshore location. It contributes to about 25-30% of the COE [39]. • The cost of foundation and tower are more significant in offshore wind turbine, therefore the affecting the COE. • Transportation and installation cost of heavy generators will affect the cost of energy significantly. Cranes for installation to handle such heavy weight, logistics are major cost carrier. 6. Conclusion The paper reviewed various concepts and configurations of energy conversion systems in modern wind turbines that are in the market. Some of the conclusions are • Variable speed wind turbines are more suitable for future wind turbines. This is also the present trend. • DFIG with multiple stage gear are the lightest and cost effective solution, permanent magnet direct drive configuration are the heaviest and most expensive (except for its electrically excited counterpart) but has highest energy yield. • The cost of energy is lowest for permanent magnet direct drive and single stage geared configuration for range of 1.5MW; however the single stage geared configuration has lower cost of energy for larger power ratings (upscaling). COE of large direct drive machine is not competitive. • Upscaling the wind turbine shows decreasing or similar cost of major components in terms of percentage of total component cost, except for the direct drive generator which sees a significant increase in cost per total component cost. • Upscaling of wind turbines seems to favor single geared machine (Multibrid type) at present state of art in terms of COE. However the COE might change for offshore wind turbines due to higher O&M costs compared to direct drive machine. • Permanent magnet direct drive machine can be more favorable if the weight is reduced (therefore the cost), transportability of machine is improved. Reference 1. 2.

J.G. Slootweg, E. de Vries, “Inside wind turbines - Fixed vs. variable speed”, Renewable Energy World, 2003, pp. 3040. H. Polinder, S.W.H. de Haan, M.R. Dubois, J.G. Slootweg, ‘Basic operation principles and electrical conversion systems of wind turbines’. Accepted for publication in EPE Journal, 2005.

9

3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40.

Grauers, “Design of direct-driven permanent-magnet generators for wind turbines”, PhD-thesis, Göteburg, Sweden, 1996. M. Dubois, “Optimized permanent magnet generator topologies for direct drive wind turbines”, PhD-thesis Delft University of Technology, Delft, The Netherlands, 2004. M.R. Dubois, H. Polinder, J.A. Ferreira, “Comparison of generator topologies for direct-drive wind turbines”, Proc. Nordic Countries Power and Industrial Electronics Conference (NORPIE), Aalborg, 2000, pp. 22-26. H. Polinder, F.F.A. van der Pijl, G.J. de Vilder, P. Tavner, “Comparison of direct-drive and geared generator concepts for wind turbines”, accepted for publication in Proc. of the international conference on electrical machines and drives, St. Antonio, 2005. G. Bywaters, V. John, J. Lynch, P. Mattila, G. Norton, J. Stowell, M. Salata, O. Labath, A. Chertok, D. Hablanian, Northern Power Systems WindPACT drive train alternative design study report, report NREL/SR-500-35524, 2004, http://www.nrel.gov/wind/windpact/drivetrain.html BTM Consult, World market update (Press release 26-03-2007), http://www.btm.dk/world-index.htm, last accessed Sept. 2007-09-19 Global wind 2006 report, GWEC, http://www.gwec.net/index.php?id=78 H. Polinder and J. Morren, “Developments in wind turbine generator systems”, Electrimacs 2005, Hammamet, Tunisia S. Siegfriedsen and G. Böhmeke, “Multibrid technology-a significant step to multi-megawatt wind turbines”, Wind Energy, Vol. 1, Issue 2, pp. 89-100, December 1998. G. Böhmeke, “Development and operational experience of the wind energy converter WWD-1”, in Proc. 2003 Europ. Wind Energy Conf. T. Hartkopf, M. Hofmann and S. Jöckel, “Direct-drive generators for megawatt wind turbines”, in Proc. 1997 Europ. Wind Energy Conf., pp. 668-671. L. H. Hansen, L. Helle, et al. “Conceptual survey of generators and power electronics for wind turbines”, Risø-R1205(EN) J. Manwell, J. McGowan, A. Rogers, “Wind energy explained”, Wiley ENERCON GmbH, http://www.enercon.de/en/_home.htm, last accessed September 2007. Winwind Oy, http://www.winwind.fi/english/tuotteet.html, last accessed September 2007. Harakosan Europe BV, http://www.harakosan.nl/products/, last accessed September 2007. Vestas Wind Systems, http://www.vestas.com/vestas/global/en/Downloads/Downloads/Download_brochurer.htm, last accessed September 2007. Siemens AG, http://www.powergeneration.siemens.com/en/windpower/products/index.cfm, last accessed September 2006. Repower Systems AG, http://www.repower.de/index.php?id=12&L=1, last accessed September 2007. Nordex AG, http://www.nordex-online.com/en/products-services/wind-turbines.html, last accessed September 2007. Multibrid , http://www.multibrid.com/m5000/data.html, last accessed September 2007. GE Energy, http://www.gepower.com/prod_serv/products/wind_turbines/en/index.htm, last accessed September 2007. Gamesa Elórica, http://www.gamesa.es/gamesa/index.html, last accessed September 2007. DeWind, http://www.dewind.de, last accessed September 2007. Suzlon Energy, http://www.suzlon.com/product_overview.htm, last accessed September 2007. Avantis, http://www.avantis-energy.com/, last accessed September 2007. Goldwind, http://cn.goldwind.cn/en/cp_index.asp, last accessed September 2007. Leitwind, http://www.leitwind.com/content.asp?L=3&IdMen=245, last accessed September 2007. Vensys, http://www.vensys.de/index.php?sprache=en, last accessed September 2007. Clipper, http://www.clipperwind.com/, last accessed September 2007. Scanwind, http://www.scanwind.com/index.php?option=com_content&task=view&id=27&Itemid=54, last accessed September 2007. Acciona, http://www.acciona-energia.com/default.asp, last accessed September 2007. Ribrant J., Bertling L.: “Survey of failures in wind power systems with focus on Swedish wind power plants during 19972005”, IEEE 2007. Muller S, Deicke M., Rik W., Doncker D.: “Doubly fed induction generator systems for wind turbines, IEEE trans inds. Appl.”, 2002. J.G. Slootweg, S.W.H. de Haan, H. Polinder, W.L. Kling, “Voltage Control Methods with Grid Connected Wind Turbines: a tutorial review”, Wind Engineering, 2001, pp. 353-365. Chen Y., Pillay P., Khan A.: “ PM wind Generator comparison of different topologies” IEEE 2004 Rademakers L., Braam H., Zaaijer M., Bussel G.: “ Assessment and optimization of operation and maintenance of offshore wind turbines” Jockel S., Herrmann A., Rinck J.: „High energy production plus built in reliability- the new VENSYS 70/77 gearless wind turbines in the 1.5MW class”,

41. Polinder H. Bang D., Rooij R.P.J.O.M.V, McDonald A.S, Mueller M.A.: “10MW wind turbine direct drive generator design with pitch and active stall control”, IEMDC 2007, Vol 2, pp 1390-1395 42. McDonald A.S., Mueller M.A., Polinder H.: “Comparison of generator topologies for direct drive wind turbines including structural mass”, ICEM06, Chania 2006

10