IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 58, NO. 4, APRIL 2011

1081

Overview of Multi-MW Wind Turbines and Wind Parks Marco Liserre, Senior Member, IEEE, Roberto Cárdenas, Senior Member, IEEE, Marta Molinas, Member, IEEE, and José Rodríguez, Fellow, IEEE

Abstract—Multimegawatt wind-turbine systems, often organized in a wind park, are the backbone of the power generation based on renewable-energy systems. This paper reviews the most-adopted wind-turbine systems, the adopted generators, the topologies of the converters, the generator control and grid connection issues, as well as their arrangement in wind parks. Index Terms—Control, generators, power converters, wind parks, wind turbine.

I. I NTRODUCTION

E

LECTRICITY production from wind turbines has been the focus of considerable attention when it comes to the fulfillment of renewable-energy targets set by governments worldwide. Multimegawatt (multi-MW) wind turbines, often organized in wind parks, are the main solution to achieve these goals [1]. In the last years, the focus has been shifted toward offshore resources not only due to the higher wind-energy potential but also because of the limitations and the polemic issues raised around environmental impacts of land-based wind turbines. In addition to efficiency and reliability, which are generally required for all conversion systems onshore, the size and the weight of components will be of extreme importance for offshore installations, considering that expensive platforms must be placed to support the total weight of the structure and all the components of the wind-energy conversion system. In that sense, state-of-the-art conversion systems developed and installed worldwide in land-based wind turbines will not necessarily be the most-suitable ones offshore in terms of the weight, the size, and the reliability. Fig. 1 shows some of the several options available for the implementation of a variable-speed wind-energy conversion system (WECS). The generators conventionally used in large WECSs are the doubly fed induction generator (DFIG) [1]–[3], the cage induction generator (IG), and the synchronous generator (SG) [3]. The power electronics shown in Fig. 1 correspond to a back-to-back converter. Notice that the connection of the Manuscript received June 8, 2010; revised November 9, 2010; accepted December 13, 2010. Date of publication January 6, 2011; date of current version March 11, 2011. The work of R. Cárdenas was supported by Fondecyt Chile under Grant 1085289. M. Liserre is with the Polytechnic University of Bari, 70125 Bari, Italy (e-mail: [email protected]). R. Cárdenas is with the University of Santiago de Chile, 9170124 Santiago, Chile (e-mail: [email protected]). M. Molinas is with the Norwegian University of Science and Technology (NTNU), 7491 Trondheim, Norway (e-mail: [email protected]). J. Rodríguez is with the Federico Santa Maria Technical University, Valparaiso, Chile (e-mail: [email protected]). Digital Object Identifier 10.1109/TIE.2010.2103910

power electronics in a DFIG is different from that required for the SG and the IG. Finally, as shown in Fig. 1, the WECS might be connected to a large utility, a microgrid (weak grid), or a stand-alone load (which is rather unusual for a large wind turbine). This paper aims at giving an update of the most-recent trends regarding generators, power converters, and their control with respect to overviews already published in the past years [2], [3]. On the other hand, this paper highlights the most-recent issues in terms of inertia emulation, energy storage, harmonics, faults/unbalances, and system-oriented approach for a windpark design. In Section II, an overview of wind-turbine systems is presented. In Section III, the power-converter topologies are discussed. The control systems of some of the electrical machines used in WECSs are presented in Section IV. In Section V, some grid-connection issues are analyzed. A discussion of offshorepark collection systems is presented in Section VI. Finally, the conclusions are presented in Section VII. II. W IND -T URBINE S YSTEM OVERVIEW The most-common wind-turbine systems and the control issues (as depicted in Fig. 1) are reviewed here [1]–[4]. Windturbine systems directly coupled to the grid and/or without any power converter directly or indirectly controlling the rotor speed will not be taken into consideration. The electrical generators currently used for the implementation of multi-MW WECSs are the DFIG, the cage IG, and the SG. A short overview of these generators is also presented here. A. DFIG The DFIG is widely used for variable-speed generation and is one of the most important generators for wind-energy applications [3]–[8]. Nowadays, this topology has a fraction of the wind-energy market, which is close to 50%. For a typical DFIG, the power converters are connected to the rotor and, for a restricted speed range, are rated at a fraction of the machine nominal power [5], [6], i.e., typically 30% of this value. The speed range is limited, and slip rings are required in order to connect the machine-side converter to the rotor. For WECSs based on DFIGs, gearboxes are still required because a multipole low-speed DFIG is not technically feasible [3]. The WECS speed is regulated, adjusting the electrical torque via the rotor-side converter. The speed regulation is mostly used to optimize the power extraction from the wind. However,

0278-0046/$26.00 © 2011 IEEE

1082

IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 58, NO. 4, APRIL 2011

Fig. 1. Wind-turbine systems and their control.

the possibility of controlling the active power and the reactive power gives to this system the rolling capacity on the grid [9]–[11] because the active-power injection is controlled not only with the pitch or active stall but also via the machine-side pulsewidth-modulation (PWM) converter. A DFIG-based WECS can contribute to the short-circuit power because the stator is directly coupled to the grid. Therefore, during a grid fault, relatively high currents may be produced in the DFIG stator windings. This could be an advantage from the viewpoint of simplicity to coordinate the overcurrent protection. However, to have the DFIG stator directly connected to the grid may limit the capacity of this generator to stay connected to the system, reducing the power injection and acting as a rolling capacity in the grid, which can be used to restore the system stability after the fault. To improve the fault handling capacity, usually, a crowbar is adopted in order to limit to a safe level the currents and voltages in the rotor circuit where the back-to-back power converter is used [12], [13]. The three-phase rotor winding is thus short-circuited via the closed crowbar switch, transforming the DFIG in a standard IG. Nevertheless, it has to be considered that, during the switching operation, the high currents produced may cause sudden torque loads on the drive train. Most major wind-turbine producers manufacture WECSs based on DFIGs. However, the difficulties associated in complying with grid-fault ride-through requirements may limit its use in the future [14], [15]. B. Squirrel Cage Induction Machine The squirrel-cage IG (SCIG) is a very popular machine due to its mechanical simplicity and robust construction [3]. The rotor is provided by metallic bars, which are resistant to the effects of dirt and vibration. Unlike the DFIG, no brushes are required for the operation of this machine, and little maintenance is necessary [3], mainly bearing lubrication only. The SCIG was widely used in fixed-speed WECS [3] (first Danish wind turbines), and it is still used for variable-speed wind-energy generation [3], [16]–[18]. The IG with a frequency

converter is completely decoupled from the grid, and as a consequence, this system has a complete rolling capacity. The control system of a WECS based on a SCIG and on back-toback converters could be designed to avoid increasing the shortcircuit power because the control loops limit the fault current at the grid-side converter output. The main drawbacks of the SCIG are in the fact that two fullpower converters are required for the operation of this machine and that a multipole direct-drive operation is not technically feasible [3]. Therefore, SCIGs do not have the advantage of variable-speed operation using reduced-size power converters (as in the DFIG); SCIGs can neither be used in direct-driven WECS [as in permanent-magnet generators (PMGs)]. Hence, as shown later in Table III, the number of WECS producers manufacturing topologies based on SCIGs is relatively low. C. SG SGs are considered one of the most-promising technologies [3], [19] for multi-MW wind-energy systems. Excitation is provided either with rotor windings or permanent magnets. Hence, full-scale power converters (FSCs) are needed, and a reducedscale converter for the excitation is required for synchronous machines without permanent magnets. In Table I, some characteristics of three of the most−popular wind generators, in the 3-MW power range, are presented and compared [20], [21]. It is assumed that the DFIG-based WECS uses a three-stage gearbox. For all the generators shown in Table I, it is assumed that back-to-back power converters are used to interface the wind generators to the grid. The cost, the weight, the size, and the losses of the generators are evaluated using a value between 0% and 100%, where the value of 100% corresponds to a standard WECS implemented with a DFIG and a three-stage gearbox. As discussed in Table I, the most-efficient generator is the direct-drive permanent-magnet SG (PMSG) with power losses of about 65% of that of a typical DFIG-based WECS [21]. However, in terms of costs, weight, and size, the DFIG has advantages over the direct-drive generators. Nevertheless, the

LISERRE et al.: OVERVIEW OF MULTI-MW WIND TURBINES AND WIND PARKS

1083

TABLE I T HREE G ENERATORS U SED IN L ARGE W IND T URBINES

reliability problems associated with a three-stage gearbox, the slipping rings, and the brushes make the DFIG generator rather unsuitable for applications where the logistic could be a problem and robustness is of paramount importance, e.g., offshore wind parks. Moreover, as discussed before, there are some difficulties in providing the ride-through capability with this generator. The generator with the largest weight and size is the conventional rotor-excited SG. However, there is at least one windturbine manufacturer, which is Enercon, successfully using this technology [22]. Currently, this company manufactures 3- and 6-MW WECSs equipped with annular SGs. A wind turbine of 7.5 MW is currently under development by Enercon, considering the same type of generator [23]. Multipole PMSGs with full-power back-to-back converters appears to be the configuration to be adopted by most of the large wind-turbine manufactures in the near future, gradually replacing the doubly fed generator as the main generator in the wind-energy market. An additional advantage of direct-drive generators is the noise reduction achieved when the gearbox is eliminated from the WECS [24]. For offshore applications, the increased reliability and elimination of possible oil spills from the gearbox is another advantage.

TABLE II C OMPARISON OF G ENERATORS C ONSIDERING A S INGLE -S TAGE G EARBOX

WECS. However, the efficiency is reduced because of the gearbox power losses, and this has to be considered in the evaluation of WECSs implemented using the Multibrid concept. WECSs manufactured with single-stage gearboxes and PMSGs are commercially available, e.g., the 5-MW M5000 Multibrid and the 3-MW WinWinD [23]. Another solution for very high power operation is currently being investigated and developed [26]. The high-temperature superconductor (HTS) generator can be built with at least twice the power-to-weight ratio of the generators that are available nowadays. This is achieved using superconductors, with a current capacity much higher than that of copper wires. A large 10-MW WECS (the Sea Titan) composed of a direct-drive HTS generator is currently being developed for offshore applications by the Austrian-based company WindTec [23], [26].

D. Multibrid Concept With the increase in the WECS rated power, the direct-drive operation of generators might require electrical machines of very large size, weight, and cost. In this case, the topology introduced by the German company Multibrid is an interesting concept, which may offer some advantages for the manufacturing of large WECSs in the future. Multibrid (now Areva) developed a WECS composed of a medium-speed PMSG and a single-stage gearbox with a gear ratio of about 6–10 [23], [25]. This allows reducing the weight and the size of the generators combined with the advantages of using a gearbox technology, which is lighter, more reliable, and cheaper than that of the standard three-stage gearbox with a typical ratio of 80–100 times. In [25], the characteristics of two 3-MW generators designed to operate with a single-stage gearbox are compared with a direct-drive PMSG for the same power range. One of the generators compared is a DFIG designed for a medium-speed operation, which is not commercially available and has been proposed in [25] as a possible alternative for high-power WECS applications. The other machine is a medium-speed PMSG. A summary is presented in Table II. As shown in this table, the “Multibrid” concept may achieve substantial reduction in the size and the weight of the generators used in high-power

E. Control Issues The two subsystems, i.e., the electrical and mechanical ones, that compose the WECS are characterized by different control goals but interact in view of the main aim, i.e., the control of the power injected into the grid. The electrical control system regulates the supply of the active/reactive power to the grid [9]–[11], [18]. The electrical system also provides overload protection. The mechanical subsystem is responsible of the power limitation (with pitch adjustment), the maximum energy capture, the speed limitation, and the reduction of the acoustical noise [27], [28]. In fact, the power has to react based on a set point given by the power-grid dispatch center or locally with the goal to maximize the production based on the available wind power [28]. The control of the WECS electrical subsystem (see Fig. 1) can be divided into three different stages. The first stage includes the basic functions that guarantee the proper operation of the power converters, hence taking care of voltages and currents on the generator side, in the intermediate direct-current (dc) link if present, and on the grid side [2]. The second stage includes the WECS specific functions, hence the maximization and the limitation of the power. The

1084

IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 58, NO. 4, APRIL 2011

control of the WECS is organized such that below the maximum power production, the wind turbine will typically vary the speed proportional to the wind speed and keep the pitch angle fixed [27]. At very low wind, the speed of the turbine will be fixed at the maximum allowable slip in order not to have overvoltage. A pitch-angle controller will limit the power when the turbine reaches the nominal power. The third stage includes extra functions that will become crucial in the future power system, characterized by a significant inflow of the distributed power generation. In fact, the WECS is expected to contribute and improve the power quality, to offer an energy storage to buffer the energy production, and to contribute to the grid stability with the inertia-emulation functionality. In this sense, the transmission-system operator may also provide a supervisory command to take advantage of these extra functions when required. F. Inertia Emulation and Droop Control In a typical power system, the grid frequency is controlled by the conventional power plants. However, with the increasing penetration of wind energy, it is expected than, in the near future, some grid-frequency support will be provided by the WECSs. There are several publications related to the subject of frequency support using WECSs [29]–[32]. Most of the proposed methods use the kinetic energy stored in the wind-turbine rotating mass to provide additional power to the grid in case of frequency variation. In power system, the inertia constant H is used instead of inertia J. Constant H is defined as [30] H=

Jωr2 Ek = S 2S

(1)

where S is the nominal apparent power of the WECS, J is the rotor inertia, and ωr is the rotating speed. As shown by (1) H is equal to the time that a WECS can supply the nominal power using the kinetic energy stored in the rotor. As reported in [30], the inertia constants for WECSs are in the range of 2–6 s. Meanwhile, H for a typical power-system generator is in the range of 2–9 s. Therefore, WECSs do not reduce the amount of kinetic energy stored in the system. In some publications, it is suggested to control the WECS to maintain a reserve of the active power ΔPw , e.g., regulating the pitch angle to avoid extracting the maximum power from the wind [29]. Using this power reserve, the frequency regulation is accomplished by regulating the wind-generator electrical torque to [29], [30] Te∗ = Trefω + Kd (fgrid − fref ) + Kei

dfgrid dt

(2)

where the first term Trefω represents the torque demand for the normal steady-state operation of the power system when the grid frequency fgrid is equal to the reference frequency. This torque demand might be obtained, for instance, from a look-up table where a relationship between the rotational speed and the demanded output power is stored.

The second term represents the droop torque. In a typical system, when the power is unbalanced (e.g., there is more power consumption than power generation), the grid frequency changes. In this case, the droop torque is increased/decreased in order to support the generation. In [30], it is suggested to activate the droop torque only when the grid frequency exceeds some predefined limits. The last term corresponds to the inertia emulation. In this case, the torque demand is varied according to the rate of the change in the grid frequency. This component emulates the inertia response of a conventional synchronous machine [29]. In [30], Ke is constant. In [29], [31], and [32], it is proposed to change Kei with the grid frequency. The performance of the control law [see (2)] has been evaluated, through simulation, in [29]–[32]. In these publications, it is reported that the frequency variations, produced when one of the power-system conventional generators is tripped, are smaller when the droop torque and the inertia-emulation torque are considered. Moreover, the system total-inertia constant H is increased when the last two terms of (2) are used. G. Energy Storage and Power Smoothing Because the power supplied by a WECS is proportional to the cube of the wind velocity, the wind-speed variability can produce unacceptable variations on the power or the voltages supplied into a stand-alone load or weak grid [33]. Power smoothing is accomplished by supplying a compensating power ΔPc from an energy-storage system (ESS) as Pgrid = Pw + ΔPc .

(3)

The compensating power is regulated in order to reduce or eliminate the variability in power Pw (captured from the wind), maintaining Pgrid almost constant. Several topologies have been proposed in the literature for the ESS and power-electronic interfaces required to interface the energy storage to the system [33]–[36]. For instance, the ESS can be implemented using flywheels [33], [36], supercapacitors [34], [35], lead–acid batteries, and superconducting magnetic storage devices [34]. The energy storage can be connected to the alternatingcurrent (ac) grid or to the dc link of the variable-speed WECS. For instance, in [34], supercapacitors are connected to the dc link of a back-to-back converter. In [36], the compensation is achieved by connecting a flywheel-based energy storage to the ac grid using back-to-back converters. In [35], supercapacitors are connected to the dc link of the full-power back-to-back converter required for the connection of a PMG to the grid. As reported in [34], an energy buffer can be also used to improve the low-voltage ride-through capability of WECSs. The WECS can also store energy in its dc-link and in the inertia of the machine accepting that the speed increases. III. P OWER -C ONVERTER T OPOLOGIES Table III shows the commercial WECSs that are available in the market for power ranges between 1.5 and 3 MW. These

LISERRE et al.: OVERVIEW OF MULTI-MW WIND TURBINES AND WIND PARKS

TABLE III S UMMARY OF THE C OMMERCIALLY AVAILABLE WECS IN THE 1.5- TO 3-MW R ANGE

1085

1086

IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 58, NO. 4, APRIL 2011

Fig. 2. Power-converter topology for the DFIG.

WECSs are manufactured by 44 companies in 163 models. For instance, as shown in Table III, the Canadian company AAER manufactures six models of WECSs (models A1650–A2000) with power ranges between 1.65 and 2.0 MW. In Table III, the label “NM” stands for “number of models” According to the information depicted in Table III, in the power range of 1.5–3 MW, there are 60 models of WECSs implemented using DFIGs, including a model with a stator winding designed for direct connection to a medium-voltage grid (a WECS manufactured by the Spanish company Acciona). There are 66 models implemented with PMSGs, 18 models implemented with IGs (cage or wound-rotor machines), and 19 models of WECSs implemented using wound-rotor SGs with external excitation. Regarding power converters, most of the WECSs presented in Table III are equipped with back-to-back converters based on low-voltage insulated-gate bipolar transistor (IGBT) switches (typically 1700-V devices for a 690-V rated voltage). A few exceptions are shown in Table III, e.g., WECSs implemented with a variable ratio gearbox (Dewind) and fixed-speed SGs. From Table III, it is concluded that back-to-back converters implemented with low-voltage IGBTs are the most-important power-electronic topology for that power range. However, as discussed in Section III-B, it is not convenient to use lowvoltage converters for high-power WECSs. Aside from the WECS shown in Table III, there is a 2.3-MW prototype for offshore applications, which is currently being tested. This is the floating HyWind system designed by Siemens for deep-sea applications. A. Power Converters for DFIGs As discussed in Section II-A, the variable-speed operation of DFIGs is achieved with a power converter designed for about 30% of the nominal power. Therefore, the commercially available DFIG of up to 6 MW are equipped with back-toback converters designed for a low-voltage operation (690 V). A transformer is typically used to interface the WECSs to the grid (see Fig. 2). There are some converters particularly manufactured for the control of DFIGs, e.g., the water-cooled Prowind topology of the company Converteam [37]. These converters are designed for 690 V, with a modular technology that allows the parallel connection of several back-to-back converters to increase the total power. Another off-the-shelf solution for DFIGs is the ACS800 converter manufactured by ABB with a power range of

Fig. 3. Power-converter topology with full-scale converter organized in parallel units.

1–3.8 MW and 525–690 V, which allows the control of DFIGs up to ≈ 6 MW [39]. The power density of the converter typically used for DFIGs is not appropriate for offshore applications where the weight and the space are of paramount importance. For instance, the AC800 converter has a power density of about 0.43 MW/m3 [39]. The power converter Alstom has a power density of 0.24 MW/m3 [39]. A higher power density can be obtained using multilevel converters as discussed in [40] and [41], where the application of neutral-point-clamped (NPC) converters is proposed for DFIG-based WECSs. The main advantage of this topology is that the converter is directly connected to a medium-voltage grid (with voltages between 1 and ≈5 kV). The medium-voltage operation allows a substantial reduction in the current handled by the devices. Another solution proposed to increase the power density is to use all silicon power converters. In this topology, the bulky dc-link capacitors are not required, and this increases the power density. Recently, the control of DFIGs using matrix converters has been proposed in [42] and [43]. Nevertheless, more research is required for the successful application of direct and indirect matrix converters to high-power WECSs, for instance, to investigate the performance of these converters for the ride-through operation. Difficulties associated with complying with the ride-through requirement may arise from the fact that matrix converters do not have dc-link capacitors, and this produces a more-direct coupling between wind generators and the grid [42]. B. FSCs Variable-speed WECSs implemented with PMSGs, SGs, and SCIGs are connected to the grid using full-power converters. The approach currently used is to implement a back-to-back topology using several converter modules connected in parallel, as shown in Fig. 3. Parallel modules also provide redundancy because, if one of the modules fails, the system can still provide a part of the power. Finally, the commutation of different modules could be synchronized to reduce harmonics using interleaved modulation. The Spanish company Gamesa is using a 4.5-MW converter system composed of six modules, each one of which is

LISERRE et al.: OVERVIEW OF MULTI-MW WIND TURBINES AND WIND PARKS

1087

TABLE IV L ARGE W IND T URBINES FOR O FFSHORE A PPLICATIONS

Fig. 4. Power-converter topology with full-scale converter.

implemented with low-voltage IGBTs [39], [44]. Every converter module has its own circuit breaker, chokes, filters, measuring system, and control unit [39]. According to [39] and [44], the power density achieved is 0.58 MW/m3 , which is about 30% smaller than similar systems manufactured by ABB and Vacon. However, the new generation of WECSs will substantially increase the power range. Table IV shows some of the most powerful wind turbines, which are either available or currently under development mainly for offshore applications [23] (the 10-MW Britannia manufactured by Clipper Wind Power will be used in “Round Three” in U.K., 2011). As depicted in the table, the nominal power is expected to reach 10 MW in the near future, with direct-drive WECSs of 6 MW already available in the market. Considering the future increase in the power size, the connection of the parallel low-voltage module is no longer convenient. For instance, a 7.5-MW power converter operating at 690 V is handling ≈6300 A. Moreover, because of the large voltage drop in the cables, the power-converter modules have to be installed in the nacelle close to the wind generator [45], [46] using a large fraction of the available space and increasing the nacelle total weight. To increase the power density, some alternatives have been discussed in the literature. For instance, in [45] and [47], the research realized in ABB is presented. A full-power backto-back NPC converter implemented using integrated gatecommutated thyristors (IGCTs) is compared with the performance obtained by using 6- or 12-pulse bridge rectifiers in the PMSG side followed by a single NPC power converter in the grid side. It is concluded that bridge rectifiers have many drawbacks, e.g., torque pulsations of frequencies below 200 Hz in the mechanical structure, high harmonic distortion in the stator currents, increased eddy-current losses in the rotor, and even demagnetization of the permanent magnets at a relatively low load [3], [48]–[50]. In [45], it is concluded that fourquadrant three-level back-to-back converters are appropriate for high-power WECSs (see Fig. 4). However, further research in the subject is required because there is still some room to improve the power density using novel converter topologies. For instance, as reported in [55], a two-quadrant rectifier in the

PMSG side, providing sinusoidal stator currents, could reduce the cost and increase the power density of the FSC. There are already some commercial solutions for the control of high-power WECSs. ABB has developed the PCS6000 wind-power converter, which is designed for medium-voltage applications (defined as 1–5 kV in [47]) and implemented with two NPC inverters connected in a back-to-back configuration. PC6000 is water-cooled and designed for a power range of up to 8 MW. The grid-side inverter uses an optimized-pulsepattern PWM to eliminate the distortion created by the loworder harmonics [47]. IGCTs with reduced power losses are used for this implementation. Efficiencies of about 98% are claimed for this converter [45], [47]. Another commercial implementation of an NPC back-toback converter is reported in [51]. The MV7000 converter is a solution offered by Converteam. The converter is implemented using power-pack IGBTs, which is a proprietary technology developed by this company. The power range is similar to that achieved with ABB PC6000. The German company Enercon is using a proprietary solution for the FSC provided in the WECSs manufactured by this company. The converter in the annular SG side is a rectifier. A small-scale power converter is used to regulate the rotor field current to maintain a constant dc voltage. The frontend converter is a four-quadrant inverter providing sinusoidal currents to the grid. However, no information about the power density and the total harmonic distortion in the generator side is available. The Britannia 10-MW WECS, currently under development, is also going to be equipped with a full-power converter similar to that used by Enercon, i.e., a rectifier in the machine side and a four-quadrant inverter providing sinusoidal currents to the grid. C. Future Trends in Power Converters for Wind-Power Generation The back-to-back current-source rectifiers and the currentsource inverter topology must be also studied for high-power application in the range of 5–10 MW. Some recent works have demonstrated that this solution is feasible, as discussed in [52] and [53]. The prevalence of one particular powerconverter topology in the wind-energy industry is related to the compromise between the cost and the performance. Regarding the power-electronic devices, it is likely that new technologies will increase the power density of medium-voltage converters. For instance, the reverse-conducting IGBT (bimode insulated-gate transistor) is reported to increase the power density up to 50% [54]. The high-power technology IGCT may

1088

IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 58, NO. 4, APRIL 2011

also have advantages, for instance, improved fault handling capabilities [54]. Finally, the increase power density and reliability can be also obtained using novel power-converter topologies; for instance, in [55], experimental results obtained with a 6-MW converter prototype are reported. This prototype is a five-level hybridconverter-based active NPC converters. To increase the density, the machine-side converter is a two-quadrant three-level rectifier unit. The authors claim that this topology achieves one of the most-compact designs for wind-energy applications ever reported. IV. C ONTROL OF G ENERATORS Here, the control of generators for wind-turbine applications is reviewed. A. DFIGs As reported in [5], [42], and [43], the DFIG is usually controlled using a vector-control scheme oriented along the stator flux [5]. However, the stator-voltage orientation is also feasible [56]. As shown at the bottom of Fig. 5, an estimation of the stator flux can be obtained as  (4) ψˆs = (vs − Rs is )dt where vs is the grid voltage and is is the stator current. A bandpass filter is usually applied to implement (4), avoiding the drift produced by low-frequency components in vs and is . From the α−β components of the stator flux, the electrical angle is obtained from θe = tan−1 (ψβ /ψα ). Alternatively, θe can be calculated from a system based on a phase-lock loop [57]. This provides better performance when the DFIG is connected to relatively weak grids, with distorted or unbalanced voltages [4], [57]. The slip angle, used to modulate/demodulate the rotor currents and voltages, is calculated using θsl = θe − θr .

(5)

For the control system shown in Fig. 5, the rotor-position angle is measured by a position encoder. However, a sensorless operation is feasible, for instance, using some of the modelreference adaptive system (MRAS) observer proposed in [58]– [60], to estimate the rotor position and speed. For stator-flux orientation, the q-axis rotor current regulates the machine electrical torque, and the d-axis rotor current regulates the magnetizing power supplied to the rotor. In Fig. 5, conventional proportional–integral controllers are used to regulate the rotor currents. Nevertheless, other controllers are suitable for this application [60]. In a grid-connected DFIG, the torque current irq is typically controlled to drive the WECS to the point of maximum aerodynamic efficiency [17]. A well-known strategy to achieve this goal (in steady state) is to regulate the machine electrical torque proportional to the square of the rotational speed [17] (see Fig. 5). However, other control strategies for regulating the

Fig. 5. Variable-speed generation system based on doubly fed induction machines and back-to-back converters.

power capture are also feasible [30]–[32], e.g., the frequencycontrol strategies discussed in Section II-F. The reactive power supplied by the DFIG to the grid is regulated using the direct rotor current ird [9]. For instance, if ird ≈ 0, the machine is entirely magnetized from the stator, and the reactive power is drawn from the utility. Another approach to regulate the direct rotor current is to maximize the WECS efficiency, reducing the iron, switching, conducting, and copper losses in the generation system [10], [11]. A stand-alone load can be also fed from a vector-controlled DFIG. For this application the control system depicted in Fig. 5 has to be slightly modified. For instance, for a grid-connected generation, a stator-flux control loop is not required [5] because this flux is a function of the grid voltage. On the other hand, for a stand-alone application, a stator-flux control loop is necessary [57], [58] to regulate the output-voltage magnitude. Moreover, for an islanded operation, the electrical angle θe is provided by a fictitious rotating vector (e.g., by using θe = ωe dt), and the stator flux (or stator voltage if the control system is oriented along this vector) is forced to align with this vector [57]. More information about the stand-alone and/or grid-connected operation of DFIGs is presented in [4], [5], and [57]–[60]. B. SCIG Fig. 6 shows a typical sensorless vector-control scheme for the operation of a SCIG in a variable-speed WECS. The sensorless vector-control system, shown in Fig. 6, is oriented along the rotor flux vector, with the rotor-position angle and the rotational speed estimated using an MRAS observer [17], [62], [63]. As discussed in [62], the observer is based on a current model and a voltage model implemented using the voltages and the currents measured in the machine stator. Additionally, a spectral-estimation algorithm (see Fig. 6)

LISERRE et al.: OVERVIEW OF MULTI-MW WIND TURBINES AND WIND PARKS

1089

As in cage machines, the PMG may be controlled using the standard vector control [48]. A sensorless operation is also feasible in this machine [67], [68], increasing the robustness and reliability of the WECS. The main advantages of PMGs, compared with induction machines, is the feasibility of building PMGs with a large number of poles and a reasonably small size for a given power [3], [19]. Therefore, the direct-drive operation is possible with this machine, eliminating the gearbox from WECSs [19]. V. G RID C ONNECTION

Fig. 6.

Variable-speed WECS based on a cage IG.

can be used to provide an additional speed measurement from the rotor slot harmonics [63]. The speed obtained from the spectral-estimation method is used to tune the MRAS observer, eliminating the steady-state speed error from its output [63]. In the vector-controlled SCIG, the stator torque current isq also regulates the electrical torque, driving the WECS to the point of maximum aerodynamic efficiency [17]. Regarding the direct-axis rotor current, the SCIG is usually operated at full flux with the nominal magnetizing current supplied from the machine-side converter. However, in order to improve the efficiency of the WECS by reducing the iron losses, the SCIG may be operated using some sort of reduced flux-control strategy during low wind velocities [64]. With minor modifications, the control system in Fig. 6 can be also used to supply energy to a stand-alone load (as will be the scenario in the North Sea with wind power providing energy to oil platforms), controlling the load-side converter to supply energy of constant voltage and frequency to isolated loads. Induction machines are also used for controlling a rotating energy storage. For instance, in [33], [65], and [66], the applications of cage machines driving flywheels (for winddiesel applications) are discussed. Power smoothing systems with cage machines and flywheels are discussed in [33]. The application of doubly fed induction machines for the control of flywheel energy storage has been also discussed in [56], using simulation studies. C. PMG PMGs have a good efficiency and a good power-to-size ratio, and the maintenance is mainly restricted to lubrication of the bearing [3], [19], [48]–[50], [67], [68]. PMGs can be divided into several categories, e.g., radial-flux permanent-magnet machines, machines built with surface-mounted magnets, machines with damper windings, etc. In all these machines, the excitation is provided by internal magnets in the rotor. Because of the magnets, a major operational consideration in PMGs is to maintain the rotor temperature below the maximum value allowed by the magnetic material [3].

Here, two main issues related to the connection to the grid of the wind generator are briefly addressed for the previously reported topologies, i.e., harmonics and faults/unbalances, whose responsibility is of the grid converter [93]. These topics can be addressed at the component level, as discussed in Section VI, but in the case of wind parks, the system-level approach is more advisable. A. Harmonics Current and voltage harmonics are strictly limited by grid standards and codes for multi-MW wind turbines. The issue is treated both at the hardware level (e.g., use of filters) and at the control level (modulation techniques and harmonic controllers). The introduction of filters changes the system plant. As a consequence, stability issues are faced, and extra damping, either passive or active, is needed. The increased power of the WECS and the consequent limited switching frequency pose stringent constraints on the filter, modulation, and controller/damping designs [90]–[92]. B. Faults/Unbalances Faults are strictly related to unbalance issues; in fact, faults generally lead to the creation of negative-sequence components in the voltages/currents. Until recently, there were no particular requirements for WECSs to remain connected during voltage sags, and hence, to avoid large overcurrents, the generator protection tripped during faults. With the increase in the wind-energy penetration, the disconnection of wind turbines is no longer feasible without compromising the stability of the entire power system [2], [13]–[15], [70]. Therefore, some grid operators, e.g., in Spain and Germany [15], have issued stringent grid codes requiring WECSs to remain connected during a typical voltage sag. As discussed before, one important disadvantage of DFIGs in variable-speed WECSs is that the machine stator is directly connected to the grid. Therefore, grid perturbations, faults, and grid unbalances may severely affect the machine performance [12], [57]. For instance, in weak grids, the negative-sequence voltage produces torque pulsations [12] and localized heating [57] in the machine. This may reduce the life span of a typical generation system. Control systems to compensate the effects of unbalances have been presented in [12] for grid-connected DFIGs and in [57] for the stand-alone generation. In most of the proposed control schemes for the grid-connected operation, the DFIG rotor currents are regulated to compensate

1090

IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 58, NO. 4, APRIL 2011

for some of the problems produced by the negative-sequence components, e.g., to eliminate the pulsations in the machine electrical torque. For stand-alone applications, the grid-side converter may be used to compensate for the machine-stator negative-sequence current [57]. If a four-leg converter is used in the load side, then the stand-alone zero-sequence components could be also supplied by the power converter, reducing the current in the machine windings. When the generator stator is connected to the grid using fullpower converters (see Figs. 3–5), the generator is “isolated” from the grid, and the effects of grid-voltage unbalances in the generator are considerable reduced [2], [14], [70]. Moreover, because the grid-side converter is designed for full-power operation, more margin is available to control the output current during voltage sags, allowing the WECS to ride through during a typical grid fault [13], [71]. VI. PARK C OLLECTION S YSTEM : S ERIES V ERSUS PARALLEL A RRAYS When moving from land to offshore locations, wind technology as established today will be posed with new technological challenges with regard to conversion, collection, and transmission. When global efficiency is on focus, the collection systems should not be separately analyzed from the conversion system as this will strongly affect the offshore-park efficiency. The choice of the electrical-conversion system will decide the overall efficiency offshore. In the current state of the art, wind-park collection grids are built in a medium-voltage ac (10–33 kV). Today, out of the 54 operative offshore wind parks with a total installed capacity of 3007 MW, the largest wind park has a 300-MW capacity installed with a collection grid at 33 kV [72]. According to [73], the series connection of turbines leads to less transmission losses and increases the power density without heavy high-power transformers. However, this grid architecture is reported with high losses in the power-electronic converters. Therefore, the converter topology and its modulation become the key factors to make the series connection viable by reducing the losses in the conversion system inside the nacelle of each wind turbine, as reported in [77]. Due to the long distances, conventional ac transmission is neither technically nor economically attractive, and highvoltage dc (HVDC) appears to be the most-suitable option [74], [79]. The design of the conversion system offshore requires taking into account not only the efficiency and the reliability but also the size and the weight, as expensive platforms must be placed to support the total weight of the structure and the components of the conversion system. Although several grid arrays have been proposed for offshore wind parks, seriesor parallel-collection arrays are the two main well-defined alternatives. A. Collection Grid Topologies Recent studies have shown different topologies for the series connection using line-commutated converters. In [80], a

Fig. 7.

Parallel connection of turbines in an offshore wind park.

thyristor-based current-source converter is proposed for the series connection of wind turbines based on PMSGs. Two and three turbines were simulated, showing a good performance of this topology. Current-source converters based on self-commutated switches have been also proposed in [81], concluding that insulation and optimal management of transmission-cable losses are important issues to be taken into account. A comparison between ac and series-dc transmission was also discussed. A series array based on cycloconverters using fast thyristors and a medium-frequency transformer was also proposed in [82]; however, an offshore platform is required for the ac/dc conversion step. In [83], a medium-frequency link was proposed, and a control strategy was presented. In [73], a comparison between different offshore-grid configurations, considering operation and investment cost, is reported. A series connection is presented as an attractive option if the dc–dc converter technology can maximize the efficiency. The series configuration presents less cable losses and requires less conversion stages and investment cost; however, the efficiency and the reliability in the conversion system must be increased to make this configuration economically and technically feasible. In addition, a coordinated control must be used to operate the system as a series collection requires constant and equal current in each turbine. As previously stated, the series collection leads to high losses in the power-electronic converters [73]. On the other hand, the parallel connection is a well-known topology, which permits a more-reliable operation. By connecting in parallel clusters of series-connected turbines, a good compromise between the reliability and the efficiency could be achieved on the condition that losses are reduced in the power-electronic converters [75]. The reduction of losses in the power electronics will then rely on identifying converter topologies and modulation techniques that have the potential for reducing the losses in the conversion system to make series arrays an alternative for offshore collection grids. Fig. 7 shows an offshore park in which wind turbines are connected in a conventional parallel array. Turbines are connected to a low-voltage grid. A high-power transformer is required for increasing the voltage to a transmission level [79]. Losses in the HVDC transmission are low because of the high voltage and the dc. However, losses in the offshore grid are high due to the low voltage and high current. Only one stage of conversion is required. To improve the efficiency in the offshore grid, a medium-voltage power-conversion system can be introduced in

LISERRE et al.: OVERVIEW OF MULTI-MW WIND TURBINES AND WIND PARKS

Fig. 8.

Series connection of wind turbines in the offshore park.

the wind turbine. This will improve the basic power-electronic conversion efficiency and will allow the total elimination of the wind-turbine transformer. With today’s available mediumvoltage power-conversion systems (at a 3.3- to 4-kV ac or a 5- to 6-kV dc), the wind-turbine transformer can be eliminated, and an efficiency improvement up to 1.4% is achievable. With even higher medium-voltage levels in the wind-turbine clusters, further efficiency improvements are possible. Such higher medium-voltage solutions are an interesting option, but it remains a technical challenge to build such power-electronic converters at an attractive cost level [84]. Improvements in the efficiency by introducing higher voltage levels in the offshore grid have also been proposed by feasibility studies. An increase from 33 kV (highest voltage currently used in offshore arrays) to 48 and 66 kV for several grid arrays was investigated in [85]. Fig. 8 shows an offshore park in which the wind turbines are connected in series. In this case, losses in the offshore grid are as low as the losses in the transmission stage. Moreover, the length of cable is shorter than in the parallel connection. The offshore platform might not be required, and the investment in the cable is reduced. With this grid architecture, there will be an improvement in the efficiency and the weight. However, more conversion stages are required, and the global efficiency is still less due to the converter and transformer losses [73]. In order to better exploit the lower losses’ dc-series grid architecture, a conversion system in the nacelle, with a higher efficiency than that reported today, is necessary. The matrix-converter technology represents such a possible alternative if the topology is customized so as to eliminate the intermediate dc-conversion stage of the conventional solutions presented in [73]–[76]. With a matrix-converter-based solution, as presented in [77], [86], and [88], an improved efficiency in the conversion system can be attained, and by that, the series array will be superior in the global efficiency compared with the parallel one, making it feasible as the grid architecture for offshore wind parks. The cluster connection shown in Fig. 9 could increase the reliability of the system, holding the advantages of both series and parallel arrays. The parallel connection can be used for both ac and dc offshore grids. In general, the ac grid leads to more losses than dc grids for the same voltage level and cable resistance. Moreover,

1091

the ac grid could require the ac/ac converter for controlling the machines, which means even more losses. Fig. 10 shows these two possibilities for the parallel connection of offshore grids. A multigenerator turbine, as proposed in [87], fits with the offshore dc grid. Furthermore, the multigenerator approach can be used in the series connection as the one shown in Fig. 8. This approach can contribute to make relatively high voltage and power levels from each turbine, even with limited voltage and current ratings of the available semiconductors. On the park level, this approach can contribute in achieving the transmission voltage level with a lower number of turbines in each series-connected cluster. Thus, the high voltage can be achieved without any dedicated transformer platform. B. Some Challenges for the Series Connection The most-clear advantage of the series connection is the reduction of the offshore-grid losses. Furthermore, in the abovediscussed concept, no platform will be required offshore, and less cable is used. However, the series connection presents technological challenges that need to be further investigated. The variation in the wind velocity will cause variations in the output power and, therefore, in the output voltage. Consequently, wide voltage-variation capability and coordinated control are required in the output voltage. On the other hand, overvoltages are present after a short circuit in one turbine. As a result of that, a wide variation capability of the output voltage is required in each turbine. The insulation in the nacelle of the turbine is a practical challenge. Some authors have presented some advances related to this [81], [89], but it remains an open problem for the research and development community. VII. C ONCLUSION This paper has summarized the most-recent research trends and industrial solutions in the field of the multi-MW windturbine systems and wind parks. Regarding generators and converters, it seems that the most-adopted system is still the doubly fed generator equipped with a back-to-back converter, due to the lower weight and cost. However, for large wind-energy systems mainly designed for offshore applications, where the robustness, the efficiency, and the reliability are of paramount importance, the preferred solution has been the direct-drive SGs, considering PMSGs or machines of the wound-rotor type. Nevertheless, if the nominal power of the WECSs is substantially increased in the future, the large (increased) size and weight of the multipole PMSGs required for direct-drive operation is likely to be too large for commercial applications. In this case, other solutions, e.g., the HTS direct-drive generator or systems based on the “Multibrid” concept, will become more-attractive alternatives. As an example of this trend, the 10-MW Britannia WECS is designed using a topology similar to that of the “Multibrid” concept, i.e., considering a smaller and more-reliable gearbox and a state-of-the-art PMSG. Regarding control issues, while low-voltage ride through is still the most-important topic (challenging particularly

1092

IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 58, NO. 4, APRIL 2011

Fig. 9. Cluster connection of an offshore wind park with a matrix-converter-based conversion system.

Fig. 10. Possible parallel connection of the offshore grid. (a) Offshore dc grid. (b) Offshore ac grid.

DFIG-based WECSs), ancillary services as grid-frequency support (e.g., inertia emulation and droop control) and ESSs are emerging issues. The most-adopted power-converter topology is still based on two-level PWM inverters. However, the most-recent projects of the main wind-turbine manufacturers show a clear preference toward multilevel and multicell structures to manage the increasing power of WECSs, with the aim of increasing the power density particularly for wind parks. Currently, most of the commercial multilevel power converters offered for windenergy applications are based in NPC topologies designed for nominal voltages of up to 5 kV. However, more research in the field of multilevel-converter topologies seems to be required in order to select the best alternatives for high-power wind-energy systems. Beyond the current state of the art, direct ac–ac conversion systems based on the matrix-converter concept are coming into the picture, calling for new developments in bidirectional switches and high-frequency transformers. Moreover, here, the preferable topologies include multigeneration solutions with a

modular conversion system that can make use of the current state-of-the-art components. Regarding the offshore-park topology, a system-oriented approach is suggested in order to increase the power density of the components in the nacelle and, at the same time, to minimize the overall losses in the conversion-and-collection system. One important aspect to remark in the conclusion is that the conversion–collection system topology, and the offshoregrid voltage level will critically influence the choice of the park array when the system is optimized under given targets and constraints. It is however important to highlight that, in this respect, there is still a long way to go until the mostsuitable (optimized) solutions for the given cases are identified and established as solid. However, with the massive ongoing and imminent developments of offshore wind parks, ad hoc, rather than generalizable, solutions will be emerging from the current installations. From these, experience will be learned, and future developments will benefit from the merging of the new knowledge emerging from research, expected technological developments, and the lessons learned from the field.

LISERRE et al.: OVERVIEW OF MULTI-MW WIND TURBINES AND WIND PARKS

ACKNOWLEDGMENT M. Liserre would like to thank M. Dellapigna, B.Sc. student in electrical engineering, Politecnico di Bari, Italy, for collecting the information depicted in Table III. R EFERENCES [1] M. Liserre, T. Sauter, and J. Y. Hung, “Integrating renewable energy sources into the smart power grid through industrial electronics,” IEEE Ind. Electron. Mag., vol. 4, no. 1, pp. 18–37, Mar. 2010. [2] F. Blaabjerg, R. Teodorescu, M. Liserre, and A. V. Timbus, “Overview of control and grid synchronization for distributed power generation systems,” IEEE Trans. Ind. Electron., vol. 53, no. 5, pp. 1398–1409, Oct. 2006. [3] L. H. Hansen, L. Helle, F. Blaabjerg, E. Ritchie, S. Munk-Nielsen, H. Bindner, P. Srensen, and B. Bak-Jensen, “Conceptual survey of generators and power electronics for wind turbines,” Risø Nat. Lab., Roskilde, Denmark, Tech. Rep. Riso-R1205 (EN), Dec. 2001. [4] G. Abad, M. A. Rodriguez, G. Iwanski, and J. Poza, “Direct power control of doubly-fed-induction-generator-based wind turbines under unbalanced grid voltage,” IEEE Trans. Power Electron., vol. 25, no. 2, pp. 442–452, Feb. 2010. [5] R. Peña, J. Clare, and G. Asher, “Doubly-fed induction generators using back-to-back PWM converters and its applications to variable-speed wind-energy generation,” Proc. Inst. Elect. Eng., vol. 143, no. 3, pt. B, pp. 231–241, May 1996. [6] W. Leonhard, Control of Electrical Drives. Berlin, Germany: SpringerVerlag, 1997. [7] S. Muller, M. Deicke, and R. W. De Doncker, “Doubly fed induction generator systems for wind turbines,” IEEE Ind. Appl. Mag., vol. 8, no. 3, pp. 26–33, May 2002. [8] K. Rothenhagen and F. W. Fuchs, “Doubly fed induction generator model-based sensor fault detection and control loop reconfiguration,” IEEE Trans. Ind. Electron., vol. 56, no. 10, pp. 4229–4238, Oct. 2009. [9] M. Kayikci and J. V. Milanovic, “Reactive power control strategies for DFIG-based plants,” IEEE Trans. Energy Convers., vol. 22, no. 2, pp. 389–396, Jun. 2007. [10] B. Rabelo and W. Hofman, “Control of an optimized power flow in wind power plants with doubly-fed induction generators,” in Proc. IEEE Power Electron. Spec. Conf., Acapulco, México, Jun. 2003, pp. 1563–1568. [11] B. Rabelo and W. Hofmann, “Power flow optimisation and grid integration of wind turbines with the doubly-fed induction generator,” in Proc. IEEE PESC, Recife, Brazil, 2005, pp. 2930–2936. [12] M. Kiani and W.-J. Lee, “Effects of voltage unbalance and system harmonics on the performance of doubly fed induction wind generators,” IEEE Trans. Ind. Appl., vol. 46, no. 2, pp. 562–568, Mar./Apr. 2010. [13] J. Lopez, E. Gubia, P. Sanchis, X. Roboam, and L. Marroyo, “Wind turbines based on doubly fed induction generator under asymmetrical voltage dips,” IEEE Trans. Energy Convers., vol. 23, no. 1, pp. 321–330, Mar. 2008. [14] F. Iov, R. Teodorescu, F. Blaabjerg, B. Andersen, J. Birk, and J. Miranda, “Grid code compliance of grid-side converter in wind turbine systems,” in Proc. IEEE PESC, Jun. 18–22, 2006, pp. 1–7. [15] Grid Code. High and Extra High Voltage, E.ON Netz, Bayreuth, Germany. [Online]. Available: www.eon-netz.com [16] E. J. Bueno, S. Cóbreces, F. J. Rodríguez, A. Hernández, and F. Espinosa, “Design of a back-to-back NPC converter interface for wind turbines with squirrel-cage induction generator,” IEEE Trans. Energy Convers., vol. 23, no. 3, pp. 932–945, Sep. 2008. [17] R. Cardenas and R. Pena, “Sensorless vector control of induction machines for variable-speed wind energy applications,” IEEE Trans. Energy Convers., vol. 19, no. 1, pp. 196–205, Mar. 2004. [18] R. Cardenas, R. Pena, P. Wheeler, J. Clare, and G. Asher, “Control of the reactive power supplied by a WECS based on an induction generator fed by a matrix converter,” IEEE Trans. Ind. Electron., vol. 56, no. 2, pp. 429–438, Feb. 2009. [19] J. Zhang, Z. Chen, and M. Cheng, “Design and comparison of a novel stator interior permanent magnet generator for direct-drive wind turbines,” IET Renew. Power Gener., vol. 1, no. 4, pp. 203–210, Dec. 2007. [20] H. Polinder, F. F. A. van der Pijl, G.-J. de Vilder, and P. J. Tavner, “Comparison of direct-drive and geared generator concepts for wind turbines,” IEEE Trans. Energy Convers., vol. 21, no. 3, pp. 725–733, Sep. 2006.

1093

[21] H. Polinder, F. F. A. van der Pijl, G.-J. de Vilder, and P. Tavner, “Comparison of direct-drive and geared generator concepts for wind turbines,” in Proc. IEEE Int. Conf. Elect. Mach. Drives, May 15, 2005, pp. 543–550. [22] ENERCON Wind Energy Converters Technology & Service. accessed Oct. 2010. [Online]. Available: http://www.enercon.de/p/downloads/, file EN_Eng_TandS_0710.pdf [23] Wind Turbines and Wind Farms Database. accessed on Oct. 2010. [Online]. Available: http://www.thewindpower.net/turbines_list.php [24] A. Ragheb and M. Ragheb, “Wind turbine gearbox technologies,” in Proc. 1st INREC, Mar. 21–24, 2010, pp. 1–10. [25] H. Li, Z. Chen, and H. Polinder, “Optimization of Multibrid permanentmagnet wind generator systems,” IEEE Trans. Energy Convers., vol. 24, no. 1, pp. 82–92, Mar. 2009. [26] Sea Titan Data Sheet. [Online]. Available: http://www.amsc-windtec.com/ pdf/SEATITAN_DS_0510_ForWeb.pdf [27] E. Muljadi and C. P. Butterfield, “Pitch-controlled variable-speed wind turbine generation,” IEEE Trans. Ind. Appl., vol. 37, no. 1, pp. 240–246, Jan./Feb. 2001. [28] H. Geng and G. Yang, “Output power control for variable-speed variablepitch wind generation systems,” IEEE Trans. Energy Convers., vol. 25, no. 2, pp. 494–503, Jun. 2010. [29] G. Tarnowski, P. Kjaer, P. Sorensen, and J. Stergaard, “Study of variable speed wind turbines capability for frequency response,” in Proc. EWEC, Marseille, France, 2009, pp. 190–193. [30] J. Morren, S. W. Haan, W. L. Kling, and J. Ferreira, “Wind turbines emulating inertia and supporting primary frequency control,” IEEE Trans. Power Syst., vol. 12, no. 1, pp. 433–434, Feb. 2006. [31] A. Mullane and M. O’Malley, “The inertial response of induction machine based wind turbines,” IEEE Trans. Power Syst., vol. 20, no. 3, pp. 1496–1503, Aug. 2005. [32] N. R. Ullah, T. Thiringer, and D. Karlsson, “Support by variable speed wind turbines—Potential and applications,” IEEE Trans. Power Syst., vol. 23, no. 2, pp. 601–612, May 2008. [33] R. Cardenas, R. Pena, G. M. Asher, J. Clare, and R. Blasco-Gimenez, “Control strategies for power smoothing using a flywheel driven by a sensorless vector-controlled induction machine operating in a wide speed range,” IEEE Trans. Ind. Electron., vol. 51, no. 3, pp. 603–614, Jun. 2004. [34] C. Abbey and G. Joos, “Supercapacitor energy storage for wind energy applications,” IEEE Trans. Ind. Appl., vol. 43, no. 3, pp. 769–776, May/Jun. 2007. [35] S. M. Muyeen, R. Takahashi, T. Murata, and J. Tamura, “Integration of an energy capacitor system with a variable-speed wind generator,” IEEE Trans. Energy Convers., vol. 24, no. 3, pp. 740–749, Sep. 2009. [36] F. Hardan, J. A. M. Bleij, R. Jones, and P. Bromley, “Bi-directional power control for flywheel energy storage system with vector-controlled induction machine drive,” in Proc. 7th Int. Conf. Power Electron. Variable Speed Drives, 1998, pp. 477–482. [37] Prowind Converter Data Sheet. accessed Oct. 2010. [Online]. Available: www.converteam.com/majic/dl/4/doc/.../5520_prowind_en_ansicht.pdf [38] ABB Low Voltage Wind Turbine Converters. accessed Oct. 2010. [Online]. Available: www.abb.com [39] B. Andresen and J. Birk, “A high power density converter system for the Gamesa G10x 4,5 MW Wind turbine,” in Proc. EPE, 2007, pp. 1–8. [40] R. C. Portillo, M. M. Prats, J. I. Leon, J. A. Sanchez, J. M. Carrasco, E. Galvan, and L. G. Franquelo, “Modelling strategy for back-to-back three-level converters applied to high-power wind turbines,” IEEE Trans. Ind. Electron., vol. 53, no. 5, pp. 1483–1491, Oct. 2006. [41] G. Abad, M. A. Rodriguez, and J. Poza, “Three-level NPC converterbased predictive direct power control of the doubly fed induction machine at low constant switching frequency,” IEEE Trans. Ind. Electron., vol. 55, no. 12, pp. 4417–4429, Dec. 2008. [42] R. Cardenas, R. Pena, G. Tobar, J. Clare, P. Wheeler, and G. Asher, “Stability analysis of a wind energy conversion system based on a doubly fed induction generator fed by a matrix converter,” IEEE Trans. Ind. Electron., vol. 56, no. 10, pp. 4194–4206, Oct. 2009. [43] R. Pena, R. Cardenas, E. Reyes, J. Clare, and P. Wheeler, “A topology for multiple generation system with doubly fed induction machines and indirect matrix converter,” IEEE Trans. Ind. Electron., vol. 56, no. 10, pp. 4181–4193, Oct. 2009. [44] J. Birk and B. Andresen, “Parallel-connected converters for optimizing efficiency, reliability and grid harmonics in a wind turbine,” in Proc. Eur. Conf. Power Electron. Appl., Sep. 2–5, 2007, pp. 1–7. [45] P. Maibach, A. Faulstich, M. Eichler, and S. Dewar, Full-Scale MediumVoltage Converters for Wind Power Generators Up to 7 MVA, accessed Oct. 2010. [Online]. Available: www.abb.com

1094

IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 58, NO. 4, APRIL 2011

[46] A. Faulstich, J. K. Stinke, and F. Wittwer, “Medium voltage converter for permanent magnet wind power generators up to 5 MW,” in Proc. Eur. Conf. Power Electron. Appl., 2005, p. 9. [47] M. Eichler, Off-Shore But Online. accessed Oct. 2010. [Online]. Available: www.abb.com [48] M. Chinchilla, S. Arnaltes, and J. C. Burgos, “Control of permanentmagnet generators applied to variable-speed wind-energy systems connected to the grid,” IEEE Trans. Energy Convers., vol. 21, no. 1, pp. 130–135, Mar. 2006. [49] O. Carranza, E. Figueres, G. Garcerá, and L. G. González, “Peak current mode control of a boost rectifier with low distortion of the input current for wind power systems based on permanent magnet synchronous generators,” in Proc. EPE, 2009, pp. 1–10. [50] S. M. Dehghan, M. Mohamadian, and A. Y. Varjani, “A new variable-speed wind energy conversion system using permanent-magnet synchronous generator and Z-source inverter,” IEEE Trans. Energy Convers., vol. 24, no. 3, pp. 714–724, Sep. 2009. [51] MV7000 Data Sheet. accessed Oct. 2010. [Online]. Available: www. converteam.com [52] X. Zeng, Z. Chen, and F. Blaabjerg, “Design and comparison of full-size converters for large variable-speed wind turbines,” presented at the European Conf. Power Electronics Applications (EPE), Aalborg, Denmark, 2007, Paper 0538. [53] J. Dai, D. D. Xu, and B. Wu, “A novel control scheme for current-sourceconverter-based PMSG wind energy conversion systems,” IEEE Trans. Power Electron., vol. 24, no. 4, pp. 963–972, Apr. 2009. [54] B. Backlund, M. Rahimo, S. Klaka, and J. Siefken, “Topologies, voltage ratings and state of the art high power semiconductor devices for medium voltage wind energy conversion,” in Proc. IEEE PEMWA, Jun. 24–26, 2009, pp. 1–6. [55] M. Winkelnkemper, F. Wildner, and P. K. Steimer, “6 MVA five-level hybrid converter for windpower,” in Proc. IEEE PESC, Jun. 15–19, 2008, pp. 4532–4538. [56] H. Akagi and H. Sato, “Control and performance of a doubly-fed induction machine intended for a flywheel energy storage system,” IEEE Trans. Power Electron., vol. 17, no. 1, pp. 109–116, Jan. 2002. [57] R. Peña, R. Cardenas, E. Escobar, J. Clare, and P. Wheeler, “Control system for unbalanced operation of stand-alone doubly fed induction generators,” IEEE Trans. Energy Convers., vol. 22, no. 2, pp. 544–545, Jun. 2007. [58] R. Cárdenas, R. Peña, J. Clare, G. Asher, and J. Proboste, “MRAS observers for sensorless control of doubly-fed induction generators,” IEEE Trans. Power Electron., vol. 23, no. 3, pp. 1075–1084, May 2008. [59] R. Peña, R. Cardenas, J. Proboste, G. Asher, and J. Clare, “Sensorless control of doubly-fed induction generators using a rotor current based MRAS observer,” IEEE Trans. Ind. Electron., vol. 55, no. 1, pp. 330–339, Jan. 2008. [60] L. Morel, H. Godfroid, A. Mirzaian, and J. M. Kauffmann, “Doublefed induction machine: Converter optimisation and field oriented control without position sensor,” Proc. Inst. Elect. Eng.—Elect. Power Appl., vol. 145, no. 4, pp. 360–368, Jul. 1998. [61] F. Bonnet and M. Pietrzak-David, “Control optimization of a doubly fed induction machine,” in Proc. IEEE PESC, 2008, pp. 2579–2585. [62] C. Schauder, “Adaptive speed identification for vector control of induction motors without rotational transducers,” IEEE Trans. Ind. Appl., vol. 28, no. 5, pp. 1054–1061, Oct. 1992. [63] A. Ferrah, K. J. Bradley, P. J. Hogben-Laing, M. S. Woolfson, G. M. Asher, M. Sumner, J. Cilia, and J. Shuli, “A speed identifier for induction motor drives using real-time adaptive digital filtering,” IEEE Trans. Ind. Appl., vol. 34, no. 1, pp. 156–162, Jan./Feb. 1998. [64] R. D. Lorenz and S.-M. Yang, “Efficiency-optimized flux trajectories for closed-cycle operation of field-orientation induction machine drives,” IEEE Trans. Ind. Appl., vol. 28, no. 3, pp. 574–580, May/Jun. 1992. [65] X.-D. Sun, K.-H. Koh, B.-G. Yu, and M. Matsui, “Fuzzy-logic-based V/f control of an induction motor for a dc grid power-leveling system using flywheel energy storage equipment,” IEEE Trans. Ind. Appl., vol. 56, no. 8, pp. 3161–3168, Aug. 2009. [66] G. O. Cimuca, C. Saudemont, B. Robyns, and M. M. Radulescu, “Control and performance evaluation of a flywheel energy-storage system associated to a variable-speed wind generator,” IEEE Trans. Ind. Electron., vol. 53, no. 4, pp. 1074–1085, Jun. 2006. [67] M. Pahlevaninezhad, S. Eren, A. Bakhshai, and P. Jain, “A model reference adaptive controller for a wind energy conversion system based on a permanent magnet synchronous generator fed by a matrix converter,” in Proc. IECON, 2009, pp. 65–70.

[68] L. Yongdong and Z. Hao, “Sensorless control of permanent magnet synchronous motor—A survey,” in Proc. IEEE VPPC, 2008, pp. 1–8. [69] J. G. Oliveira, A. Larsson, and H. Bernhoff, “Controlling a permanentmagnet motor using PWM converter in flywheel energy storage systems,” in Proc. IECON, 2008, pp. 3364–3369. [70] S. Alepuz, S. Busquets, J. Bordonau, J. Martínez, C. Silva, J. Pontt, and J. Rodríguez, “Control strategies based on symmetrical components for grid-connected converters under voltage dips,” IEEE Trans. Ind. Electron., vol. 56, no. 6, pp. 2162–2173, Jun. 2009. [71] G. Saccomando and J. Svensson, “Transient operation of gridconnected voltage source converter under unbalanced voltage conditions,” in Conf. Rec. IEEE IAS Annu. Meeting, Sep. 30–Oct. 4, 2001, pp. 2419–2424. [72] List of Off-Shore Farms. accessed Oct. 2010. [Online]. Available: http:// www.thewindpower.net/offshore_wind_farms_list.php [73] C. Meyer, “Key components for future off_shore dc grids,” Ph.D. dissertation, Rheinisch-Westfallischen Technischen Hochschule, Aachen, Germany, 2007. [74] N. Negra, J. Todorovic, and T. Ackermann, “Loss evaluation of HVAC and HVDC transmission solutions for large offshore wind farms,” Elect. Power Syst. Res., vol. 76, no. 11, pp. 916–927, Jul. 2006. [75] S. Lundberg, “Evaluation of wind farm layouts,” in Proc. Nordic Workshop Power Ind. Electron., 2004, 7 pp., CD-ROM. [76] P. Wheeler, J. Rodriguez, J. Clare, L. Empringham, and A. Weinstein, “Matrix converters: A technology review,” IEEE Trans. Ind. Electron., vol. 49, no. 2, pp. 276–288, Apr. 2002. [77] A. Garces and M. Molinas, “Electrical conversion system for offshore wind turbines based on high frequency ac link,” in Proc. IX Int. Conf. Exhib. Renewal Energy Ecological Vehicles EVER, Mar. 2009, 8 pp., CD-ROM. [78] M. Stiebler, Wind Energy Systems for Electric Power Generation, Book Series Green Energy and Technology. Berlin, Germany: SpringerVerlag, 2008. [79] P. Bresesti, W. Kling, R. Hendriks, and R. Vailati, “HVDC connection of offshore wind farms to the transmission system,” IEEE Trans. Energy Convers., vol. 22, no. 1, pp. 37–43, Mar. 2007. [80] S. Nishikata and F. Tatsuta, “A new interconnecting method for wind turbine/generators in a wind farm and basic characteristics of the integrated system,” in Proc. 13th EPE-PEMC, 2008, pp. 2343–2348. [81] D. Jovcic, “Offshore wind farm with a series multiterminal CSI HVDC,” Elect. Power Syst. Res., vol. 78, no. 4, pp. 747–755, Apr. 2008. [82] S. Meier, S. Norrga, and H. Nee, “New voltage source converter topology for HVDC grid connection of offshore wind farms,” in Proc. EPE-PEMC, Riga, Latvia, Sep. 2–4, 2004. [83] S. Lundberg, “Wind farm configuration and energy efficiency studies— Series dc versus ac layouts,” Lic. of Eng. thesis, Chalmers Univ. Technol., Goteborg, Sweden, 2006. [84] P. K. Steimer, “Medium voltage power conversion technology for efficient wind park power collection grids,” in Proc. 2nd IEEE Int. Symp. Power Electron. Distrib. Gener. Syst., Hefei, China, Jun. 2010, pp. 12–18. [85] R. Mc Dermott, “Investigation of use of higher ac voltage on offshore wind farms,” in Proc. Eur. Wind Energy Conf., Marseille, France, Mar. 2009. [86] A. Mogstad and M. Molinas, “Power collection and integration on the electric grid from offshore wind parks,” in Proc. Nordic Workshop Power Ind. Electron., Espoo, Finland, Jun. 2008. [87] J. Cotrell, “Preliminary evaluation of a multiple generator drive train configuration for wind turbines,” in Proc. 21st ASME Wind Energy Symp., 2002, pp. 345–352. [88] A. Mogstad, M. Molinas, P. Olsen, and R. Nilsen, “A power conversion system for off shore wind parks,” in Proc. 34th IEEE IECON, Nov. 2008, pp. 2106–2112. [89] P. L. E. Veilleux, “Interconnection of direct-drive wind turbines using dc grid,” in Proc. 8th Int. Workshop Large-Scale Integr. Wind Power Into Power Syst. Transm. Netw. Offshore Wind Farms, Bremen, Germany, Oct. 2009, 8 pp., CD-ROM. [90] J. Phipps, “A transfer function approach to harmonic filter design,” IEEE Ind. Appl. Mag., vol. 3, no. 2, pp. 68–82, Mar./Apr. 1997. [91] T. Wang, Z. Ye, G. Sinha, and X. Yuan, “Output filter design for a gridinterconnected three-phase inverter,” in Proc. 34th IEEE Annu. PESC, Jun. 2003, vol. 2, pp. 779–784. [92] A. A. Rockhill, M. Liserre, R. Teodorescu, and P. Rodriguez, “Grid filter design for a multi-megawatt medium-voltage voltage source inverter,” IEEE Trans. Ind. Electron, vol. 58, no. 4, pp. 1205–1217, Apr. 2011. [93] R. Teodorescu, M. Liserre, and P. Rodriguez, Grid Converters for Photovoltaic and Wind Power Systems: IEEE-Wiley, 2011.

LISERRE et al.: OVERVIEW OF MULTI-MW WIND TURBINES AND WIND PARKS

Marco Liserre (S’00–M’02–SM’07) received the M.Sc. and Ph.D. degrees in electrical engineering from the Polytechnic University of Bari, Bari, Italy, in 1998 and 2002, respectively. He has been a Visiting Professor at Aalborg University, Aalborg, Denmark; Alcala de Henares University, Alcala de Henares, Spain; and ChristianAlbrechts University of Kiel, Kiel, Germany. He is currently an Associate Professor at the Polytechnic University of Bari, teaching courses on power electronics, industrial electronics, and electrical machines. He has published 130 technical papers, 31 of them in international peer-reviewed journals, three book chapters, and the book Grid Converters for Photovoltaic and Wind Power Systems (Wiley, 2011). These works have received more than 1500 citations. Dr. Liserre is an Associate Editor of the IEEE T RANSACTIONS ON I N DUSTRIAL E LECTRONICS and the IEEE T RANSACTIONS ON S USTAINABLE E NERGY. He is the founder and was the Editor-in-Chief of the IEEE Industrial Electronics Magazine in 2007–2009. He is the founder and the Chairman of the Technical Committee on Renewable Energy Systems of the IEEE Industrial Electronics Society (IES). He is the Vice President responsible for publications of the IES. He was the recipient of an IES 2009 Early Career Award. He was the Cochairman of the 2010 International Symposium on Industrial Electronics.

Roberto Cárdenas (S’95–M’97–SM’07) was born in Punta Arenas, Chile. He received the B.S. degree from the University of Magallanes, Punta Arenas, in 1988, and the M.Sc. and Ph.D. degrees from the University of Nottingham, Nottingham, U.K., in 1992 and 1996, respectively. From 1989 to 1991 and 1996 to 2008, he was a Lecturer at the University of Magallanes. From 1991 to 1996, he was with the Power Electronics, Machines, and Control Group, University of Nottingham. He is currently a Professor of power electronics and drives in the Department of Electrical Engineering, University of Santiago de Chile, Santiago, Chile. His main interests are in the control of electrical machines, variable-speed drives, and renewable-energy systems. Prof. Cárdenas received the Best Paper Award from the IEEE T RANS ACTIONS ON I NDUSTRIAL E LECTRONICS in 2004 and the “Ramon Salas Edward” Award for research excellence from the Chilean Institute of Engineers in 2009.

1095

Marta Molinas (M’94) received the B.S. degree in electromechanical engineering from the National University of Asuncion, San Lorenzo, Paraguay, in 1992, the M.Sc. degree from Ryukyu University, Okinawa, Japan, in 1997, and the Dr.-Eng. degree from Tokyo Institute of Technology, Tokyo, Japan, in 2000. In 1998, she was a Guest Researcher at the University of Padova, Padova, Italy. From 2004 to 2007, she was a Postdoctoral Researcher at the Norwegian University of Science and Technology (NTNU), Trondheim, Norway, where she has been a Professor since 2008. From 2008 to 2009, she was a Japan Society for the Promotion of Science Research Fellow at the Energy Technology Research Institute, National Institute of Advanced Industrial Science and Technology, Tsukuba, Japan. Her research interests include wind-/wave-energy conversion systems and power electronics and electrical machines in distributed energy systems. Dr. Molinas is an Associate Editor of the IEEE T RANSACTIONS ON P OWER E LECTRONICS, an active reviewer for the IEEE T RANSACTIONS ON I NDUSTRIAL E LECTRONICS and the IEEE T RANSACTIONS ON P OWER E LECTRONICS , and she is an Administrative Committee member of the IEEE Power Electronics Society.

José Rodríguez (M’81–SM’94–F’11) received the B.Eng. degree in electrical engineering from the Universidad Tecnica Federico Santa Maria (UTFSM), Valparaiso, Chile, in 1977, and the Dr.-Ing. degree in electrical engineering from the University of Erlangen, Erlangen, Germany, in 1985. Since 1977, he has been with the Department of Electronics Engineering, UTFSM, where he was the Director of the department from 2001 to 2004, was the Vice Rector of academic affairs from 2004 to 2005, has been the Rector of the university since 2005, and is also currently a Professor. During his sabbatical leave in 1996, he was with the Mining Division, Siemens Corporation, Santiago, Chile. He has extensive consulting experience in the mining industry, particularly in the application of large drives such as cycloconverter-fed synchronous motors for semiautogenous grinding mills, high-power conveyors, and controlled alternating-current drives for shovels and power-quality issues. He has directed more than 40 research and development projects in the field of industrial electronics. He has coauthored more than 250 journal and conference papers and contributed one book chapter. His research group was recognized as one of the two Centers of Excellence in Engineering in Chile from 2005 to 2008. His main research interests include multilevel inverters, new converter topologies, control of power converters, and adjustable-speed drives. Prof. Rodríguez has been an active Associate Editor of the IEEE T RANS ACTIONS ON P OWER E LECTRONICS and the IEEE T RANSACTIONS ON I NDUSTRIAL E LECTRONICS since 2002. He has served as a Guest Editor for the IEEE T RANSACTIONS ON I NDUSTRIAL E LECTRONICS in six instances [Special Sections on Matrix Converters (2002), Multilevel Inverters (2002), Modern Rectifiers (2005), High-Power Drives (2007), Predictive Control of Power Converters and Drives (2008), and Multilevel Inverters (2009)]. He was the recipient of the Best Paper Award from the IEEE T RANSACTIONS ON I NDUSTRIAL E LECTRONICS in 2007 and the Best Paper Award from the IEEE Industrial Electronics Magazine in 2008.