FROM EDN EUROPE: Wind chimes in: wind turbines

Capturing the sun's energy via wind power is not new, but today's power semiconductors and control systems make this energy source more viable.

By David Marsh, Contributing Technical Editor -- EDN, October 14, 2004

"Look at Mother Nature on the run...in the 1970s." More than 30 years after Canadian rock-star Neil Young's plea to a generation fuelled by cheap gasoline and material abundance, today's ecowarriors argue that progress toward a more sustainable existence has reversed in the intervening period. Dire predictions of rapid climate change due to burning fossil fuels prompted Sir David King, the UK Government's chief scientific advisor, to state in March 2004: "Climate change is the most severe problem we are facing today, more serious even than the threat of terrorism." And if the first oil crisis of 1973 issued a wakeup call to sceptical industrial and public sectors, today's global terrorist threat and record prices of more than $47/barrel for crude oil provide pragmatic impetus for sustainable energy production. Of the available solar-fuelled technologies, wind turbines emerge as the frontrunner in large-scale "green-electricity" generation.

Ironically, history shows that it was only the availability of cheap fossil fuels that led to the decline in wind turbine use—notably following the United States' Rural Electrification Program of the 1930s. But today, both the US government and its European counterparts have substantial commitments to sustainable energy production. Last year, the United States installed about $1.6 billions' worth of wind-power plant, and the country intends to install 100,000 MW by 2020 to satisfy some 6% of its needs. It already boasts the world's largest onshore wind farm at Tehachapi in the Mojave Desert. But 2002 figures show that 90% of new capacity was installed in Europe (Figure 1). Thanks to Germany's Electricity Feed Law, which permits renewable-source electricity connections to the grid and specifies the price for the electricity that's generated, two-thirds of its wind-power plant belongs to domestic and small-business operators. Similar initiatives exist in countries including France, Greece, Spain, and Portugal. Meanwhile, the United Kingdom's offshore developments, such as Scroby Sands, begin to match Denmark's pioneering Horns Rev installation.

Of course, exploiting wind power is anything but new, with the first reliable evidence dating from 644 AD and the Afghan-Persian border. Although these installations used a vertically orientated rotor, European designers subsequently perfected the horizontal-axis design that characterises the classical "Dutch" four-bladed windmill. An alternative design—the US turbine—using a multiple-vane rotor dominated US installations throughout the late 19th and early 20th centuries, with more than 6 million being manufactured by 1930. Many farmers adapted these turbines, originally intended to pump water, to generate electricity. US industrialist Charles F Brush built the first fully automatic wind turbine to charge the lead-acid accumulators that powered his mansion back in 1887 and 1888. His giant device had a rotor diameter of about 17m but produced only about 12 kW at a dynamo speed of 500 rpm. Danish pioneer Paul La Cour recognised that higher generator shaft speeds enable greater efficiency and in 1891 built the first experimental wind turbine. Since then, Denmark, Germany, and the United States have led wind-turbine design and manufacture but now face stiff competition from countries such as Spain.

Variable input challenges designers

It's salutary to appreciate the extent to which the pioneers overcame many of the problems that plague today's designers. The largest of these problems by far is the variable nature of the energy supply. A conventional steam-turbine plant contains four major mechanisms to regulate generator speed and power output: the rate of primary energy consumption to generate steam, the rate of steam release to the turbine, the generator's electrical excitation level, and rotor load-angle variation. Such generators are synchronous machines, in which the rotor spins synchronously with and at an exact multiple of grid frequency. Varying the rotor's angle relative to its zero-phase-difference "idle" position either adds or subtracts power to or from the grid, respectively enabling generator or motor operation. In typical operation, the rotor leads the grid by about 30°. Because the electrical output directly couples to the grid, strong grid conditions provide generator-shaft torque that controls its speed and sustains constant grid frequency.

A limited set of opportunities exists to apply this control strategy to a wind turbine. To characterise the energy supply, maps document wind-speed data across continents, including the United States (Reference 1). Similar resources exist for Europe, where the United Kingdom is the windiest region, with a theoretical annual generation potential of some 1000 TW/hours—equivalent to three times the nation's consumption (Reference 2). Although most wind turbines supply a wide-area grid-distribution network, there's increasing interest in relatively small turbines for individual consumers or small consumer groups. In either case, the installation must supply power at a voltage and frequency that suits its connection. From La Cour's time to beyond World War II, many grids in remote parts of Denmark used dc power to simplify integrating wind turbines with small diesel-fuel generators. Today, self-commutating frequency inverters allow stand-alone installations to provide a constant frequency supply. Similarly, turbines that connect to the public-utility grid use frequency inverters to simplify power control, but this is just one approach of various aerodynamic, mechanical, hydraulic, and electrical techniques.

So, how much power is there in the wind? The theory states that, for a given air density, the available watt-per-square-metre energy metric varies to the cube power of the airflow. Accordingly, rotor performance is crucial to every aspect of wind-turbine design. One of the crucial parameters is tip-speed ratio, which is the ratio of the velocity of the blade's tip to the free-flowing air stream. This parameter describes the rotor's power coefficient, which in 1919 German physicist Albert Betz identified as being unable to exceed 0.593. In practice, typical performance scarcely exceeds 0.4 at a tip-speed ratio of 7 (Figure 2). For a fixed rotor speed and neglecting efficiency losses, you can calculate the power output of a wind turbine as:

where C_ρ is the rotor's power coefficient, ρ is air density in kilograms per cubic metre, vw is the wind speed in metres per second, and A is the rotor's swept area in square metres. It's therefore useful to consider wind turbines in terms of the rotor's swept area, as well as kilowatt-per-hour capacity. The designer's task is to maximise overall power coefficient by finding the best combination of rotor configuration and generator concept at an acceptable price for series production.

Spanning power outputs from 20 to 30 kW to the current maximum of 4.5 MW, production wind turbines conventionally use three rotor blades, because experience shows that this configuration offers the best balance of efficiency, dynamic performance, and construction economics. Core elements typically comprise the rotor, a gearbox to increase generator-shaft speed, the generator, the electrical interface, and the control loop (Figure 3). The big problem is always how best to stabilise rotor speed to maximise electricity production. Although a wind turbine is a mechatronic system, in which it's impossible to isolate key elements, the rotor-control concept is the dominant factor. Its control systems must safeguard operation in conditions that range from a dead calm to gusts with directional changes and velocities that may occur only once a century. As an indication of the mass that's involved, the rotor assembly of Vestas' 3-MW, V90-series wind turbine weighs 40 tonnes, despite the extensive use of carbon-fibre composites.

Stall control's simplicity belies issues

One way to limit power capture is to turn the rotor assembly out of the wind. The yaw system that normally functions to keep the rotor head-on into the wind comprises wind-speed and -direction sensors, an electrical or hydraulic motor-drive, interface electronics, and the gears and bearings that allow the nacelle to rotate. The sensor assembly often sits to the rear of the nacelle, traditionally comprising a three-cup anemometer with a vane. Alternative technologies include ultrasonic devices, such as the pair that the Vestas V90-3.0MW uses. In practice, the wind speed behind the rotor is slightly lower than the true wind speed, due to local low-pressure effects from the turning blades. Although this discrepancy is unimportant, characterisation can compensate such errors. But, because experience shows that yaw-based speed control produces poor results, conventional designs either maintain the head-on, maximum-power position or turn the nacelle to minimise wind capture for shutdown.

The simplest aerodynamic approach to stabilising power capture employs passive stall control, in which the rotor has a fixed pitch angle. At a given rotor speed, a wind-velocity increase causes the airflow to separate across the aerofoil's surface, creating the stall effect. This flow separation automatically limits power capture but depends on air density and the quality of the aerofoil's surface finish. The technique also requires strong grid conditions and a powerful generator to maintain stability. It's essential to guard against rotor overspeed if the grid connection fails or an electrical fault occurs, requiring aerodynamic brakes on the rotor as well as the normal mechanical disc brake on the input shaft. Because the rotor has a fixed pitch angle and can't turn to the maximum torque position for starting up, it's sometimes necessary to accelerate the rotor to grid-synchronisation speed by running the generator in motor mode. Finally, the structure must be robust to withstand the high dynamic loads that typify the stall-control concept.

Nevertheless, some successful wind turbines use the principle. Citing the elegance of simplicity and low weight, Nordic Windpower's 1-MW model 1000 has a two-bladed, stall-controlled rotor that sweeps 2290m². The turbine is self-starting and has stall strips on the blades to alleviate the peaky power curve that some earlier stall-controlled turbines exhibit, enabling a flat-topped power curve. The rotor uses glass-fibre-reinforced polyester construction, which has high aeroelasticity that contributes toward a "soft," or flexible, structure that can more easily absorb high dynamic loads. Borrowed from helicopter practice, other contributing elements include a "teetering" hub with an elastomeric bearing that damps wind-shear forces between the blades and the input shaft by allowing ±2° of relative motion. Additional damping in the generator- and yaw-control systems further improve structural flexibility.

Made by Weier Electric, the generator is a four-pole, single-speed induction machine, whose rotor turns slightly faster than the rotating electromagnetic field. This "slip" provides a damping action that helps arrest electromechanical oscillations. By switching resistors within the generator's rotor circuit to control excitation current, the slip value can vary from 1 to 10%. Because the torque of an induction generator is directly proportional to slip, this ability provides some speed control that's otherwise difficult to achieve with an asynchronous machine. At 0% slip, the generator is in synchronisation with grid frequency and neither generates nor consumes power (other than the reactive power that its rotor consumes). Similarly, if the generator spins more slowly than grid frequency, it enters motor mode and draws current from the grid. To limit this current drain, an input-shaft disc brake normally arrests rotor motion at wind speeds lower than about 4 to 5m/sec—the turbine's so-called cut-in speed.

Vestas similarly uses slip control for its OptiSlip system, with an optical coupling between the rotor-mounted electronics and the stator-mounted controller. In this case, the control value is about 10%, and the operating time is about 10 msec, smoothing power output and reducing structural loads under turbulent wind conditions. The slip value also affects generation efficiency, with megawatt-class machines typically operating in the 1% region to yield about 95% efficiency. Because the rotor circuit consumes reactive power, the native power factor is typically poor at around 0.87. For this reason, switched-capacitor banks are integral parts of traditional systems, although power electronics increasingly control power factor. In the case of Nordic's model 1000, switched capacitors maintain the output power factor at 1 across the turbine's full operating range.

By factoring some damping into the yaw system's control loop, it's possible to absorb turbulence by allowing a degree of rocking motion around the tower's axis. As a result, the model 1000's structure survives wind speeds of 55m/sec, cutting in to generate power at 4m/sec and out at 25m/sec. At a rotor speed of 25 rpm and rotor-tip speed of 71m/sec, the machine outputs its 1-MW maximum at a wind speed of 17m/sec. At the onset of rotor overspeed, centrifugal force activates hydraulic pressure-relief valves that turn the blade tips to a braking position. The SCADA (supervisory-control-and-data-acquisition) system from wind-power-system specialist Mita-Teknik can also instigate aerodynamic and mechanical braking. The generator outputs three-phase, 690V ac to the tower base over flexible cables that the SCADA system can unwind to prevent accumulative twisting. Communications between the SCADA system and a central facility occur by modem and phone line, and there's provision for a PC to independently monitor and log turbine operation.

Controls simplify power capture

Many wind-turbine designers prefer rotor pitch control, because the technique greatly eases speed variations and system power-capture issues. Two variations appear in contemporary products. The first method incrementally reduces the blade's angle of attack to the air stream from the full-power maximum to a feather position that minimises power capture; the second increases the attack angle to the point at which aerodynamic stall occurs. Danish engineers MB Pedersen and P Nielsen in 1980 tested both concepts in the experimental Nibe-A and Nibe-B turbines (Reference 3). Their results show that full blade-pitch control yields smoother output characteristics and offers greater rotor-thrust reduction potential at high wind speeds (Figure 4). Today, more sophisticated blade aerodynamics and control algorithms help minimise these differences.

Bonus Energy's products are leading examples of active-stall designs, which the company brands CombiStalls. Its "Danish-concept" turbines comprise a three-bladed rotor with constant revolution speed, a generator that supplies electricity directly to the grid, and failsafe safety systems. The company's largest product, the 2.3-MW type B40, has a swept rotor area of 5330m². It's possible to turn the blades, which are built from fibreglass-reinforced epoxy, 80° to a shutdown position. In normal operation, the microprocessor-controlled servo loop continuously adjusts blade position to stall. A dual-generator design improves efficiency at partial loads by facilitating dual-speed operation at either 11 or 17 rpm. By switching in a small, six-pole generator winding at low wind speeds, the generator produces power at two-thirds of its nominal speed. In higher wind speeds, the generator switches to its four-pole main winding and operates at nominal speed.

The turbine self-starts at an average wind speed of around 5 to 6m/sec. The rotor accelerates up to grid-synchronous speed, when a thyristor soft-start circuit connects the generator to the grid. After a few seconds online, the main contactor bypasses the thyristors to eliminate semiconductor losses. The wind turbine's output then increases roughly linearly with maximum wind speeds of about 14 to 15m/sec, when the control loop cuts in to maintain constant power output and prevent the generator from overloading. If the average wind speed exceeds the turbine's operating limit, the control system feathers the blades and applies a brake to shut down the turbine. When the wind speed drops below a restarting limit, the safety systems automatically reset, and the turbine starts up again—unless a fault occurs, in which case, the turbine stays offline. Failsafe operation comes from a redundant system that employs centrifugal devices to override the turbine's control system in case of gross malfunctions.

Inverters simplify operation

The most flexible power-capture and -control ability comes with variable-speed operation, because the turbine's rotor can ideally run at its maximum tip-speed ratio. Early attempts to substitute an automatic gearbox for the fixed-speed, step-up planetary gearbox that most turbines employ failed with cost and reliability issues. Because slip control offers limited speed control in an induction machine, an alternative method that serves as a model for many of today's turbines employs the DFIG (doubly fed induction-generator) that the 3-MW Growian wind-turbine experiment pioneered during the 1980s. The Growian topology comprises a synchronous generator with a three-phase, slip-ring-fed rotor that creates a wound-rotor induction machine. This arrangement enables a cycloconverter to inject ac current into the rotor (Figure 5a). The cycloconverter is an ac/ac frequency converter built from a thyristor array, in which sampling the three-phase line frequency creates a low-frequency control waveform (Figure 5b). Superimposing this control waveform onto the rotor's field helps stabilise the generator's output frequency; controlling its magnitude and phase controls the generator's power factor, thus emulating a synchronous generator's ability to supply true or reactive power. Issues with this topology include relatively greater susceptibility to grid-fault events than alternative schemes (see sidebar, "Power quality comes to wind turbines").

A relatively simple variable-speed technique employs an ac-dc-ac link as a frequency inverter, rectifying the generator's "wild-ac" output before commutating it at line frequency. This technique decouples the generator from the load and allows the use of the more efficient synchronous machine and retains generator torque control by varying the dc-link conditions. In one production example, the Vestas V90-3 MW wind turbine uses full blade-pitch control and the company's OptiSpeed technology to control the rotor's 6362m² swept area. The OptiSpeed system permits the rotational speed of the rotor and the generator to vary by as much as 60%, minimising output variations to the grid, as well as reducing structural strain. The heart of this system is the company's VMP-Top controller and inverter, which make up the power electronics that control the generator and its output to the grid transformer. The machine is otherwise conventional, retaining a gearbox to increase generator speed from the turbine's 9- to 19-rpm operating range.

But in what's conceptually the simplest approach, Enercon pioneered a range of gearless direct-drive wind turbines that now achieve a 4.5-MW rating. In this design, attaching the rotor directly to the generator reduces the number of drive-train bearings to just two slowly revolving components. The problem is then how to generate sufficient power at low rpm and how best to convert it to grid frequency. Enercon solves the generator problem by using an electrically excited synchronous machine with a large number of poles, such as the 84-pole, 4.8m-diameter example that appears in its 600- kW E-40 machine. Here, the rotor speed varies from 18 to 34 rpm and sweeps 1521m². With the company's roots in industrial-inverter-drive design, Enercon uses its own electronics. By contrast, the recent 2-MW model Z72 from Zephyros, which similarly features a direct-drive generator, employs a modified ACS 1000 variable-speed motor-drive controller from ABB. A single drive-shaft bearing supports the permanent-magnet generator, also made by ABB. Citing reduced generator losses, superior partial-load efficiency, and lower potential for failure, Zephyros highlights the permanent-magnet design's advantages. The downside is its greater expense, due to the use of high-permeability magnetic materials, such as neodymium iron and samarium cobalt. Permanent-magnet generators also have poor power-factor characteristics, which the inverter electronics must compensate.

But many experts believe that permanent-magnet machines are the way forward, especially for large direct-drive designs. Adrian Wilson, electrical technology specialist at the United Kingdom's NaREC (New and Renewable Energy Centre), reports that this approach forms the heart of a current research project whose key objective is weight reduction. Because a wind turbine's theoretical power output increases with the cube of the volume of air that it captures, structures tend to gain weight in proportion. Wilson says that, because current design approaches simply won't scale to the 10-MW level—let alone to the 20 or 30 MW that future demands require—his agency is investigating a direct-drive design to dispense with gearbox mass. This approach in turn requires a large-diameter generator. At the size that the project concerns, one possible approach departs from convention and substitutes a bicycle-wheel-like structure, with spokes supporting the generator's pole-pairs. The grid output connection demands a full-power ac-dc-ac frequency-inverter link, which needs multiple parallel inverters.

IGBTs oust thyristors

The power semiconductors that wind turbines require are unfamiliar to anyone who works with microelectronics. Rather than submicron line widths, consider single-device modules that occupy European-standard footprints from 34×94 to 140×190 mm. Such devices handle kiloamp-level currents at several kilovolts, and progress within this technology has made the biggest single contribution to wind-turbine advances over the past several decades. In the Growian's day, thyristor technology handled high-power applications, but conduction losses were large, and switching-time performance was poor—often in the 100-µsec region. Accordingly, frequency-converter stages typically used six- or 12-step waveforms to approximate a sine wave's energy distribution, resulting in especially strong odd-order harmonics, such as fifth and seventh. These limitations led to the need for harmonic-frequency filters. In the 1.2-MW WKA-60, or "Growian-2," wind turbine that served Helgoland, Germany, in the 1980s, the electrical equipment also included a rotating phase-shifter built from a second synchronous generator driven by a motor. This arrangement performed reactive power compensation across the turbine's output range to maintain stability within the island's weak 3-MW grid. Today's power-transmission engineers expect the electronics to cater to such requirements.

Substituting IGBTs (insulated-gate-bipolar transistors) for Growian's first-generation thyristors allows PWM (pulse-width-modulation) control that overcomes poor harmonic performance. The technique also makes it easier to control real and reactive power generation. Although they're rugged and although today's devices, such as Mitsubishi's FT1500AU-240, switch as much as 1.5 kA at 12 kV in 15 µsec, conventional thyristors are impossible to turn off while they're conducting above their holding-current value. GTO (gate-turn-off) thyristors, such as the company's FG6000AU-120D, offer continuous-voltage and -current capabilities of 6 kV at 1.5 kA and enable turn-off control in around 30 µsec, but they are difficult to drive. Worse, all thyristors are difficult to operate in parallel, which is often essential at the power levels that wind-turbine applications require.

High-power IGBTs combine the easy drive and current-sharing characteristics of MOSFETs with 1-µsec switching times. Although the PWM frequencies necessary to commutate line frequency are low at only a few kilohertz, such rapid switching minimises conduction losses as the devices traverse their linear operating region. Devices such as Eupec's FZ600R65KF1 turn on in less than 1 µsec and off in about 6 µsec to control 1.2 kA at 6 kV; lower voltage devices such as the company's FZ3600R12KE3 switch 3.6 kA at 1.2 kV. As a result, IGBTs are the device of choice for high-power inverters and soft-start controllers. Other high-power semiconductor specialists include ABB, Dynex, Fuji Electric, Powerex, and Semikron.

Spanning output-power ranges of 660 kW to 2 MW, Gamesa Eólica's wind-turbine range makes extensive use of IGBT technology to facilitate variable speed and frequency control. Variable-pitch rotor-blade control allows continuous adjustment for optimum power capture and couples to a DFIG system whose generator's speed ranges from 900 to 1900 rpm. The control technology minimises peak values, flicker, and harmonics, which eases connection-licensing issues. The vector-control system allows precise power-factor regulation and contributes to grid-voltage stability by generating or consuming reactive energy. Their power electronics also enable Gamesa Eólica's turbines to stay online during power outages elsewhere in the grid. These issues are economically critical in Spain, where there's a tariff bonus for high-quality connections.

Ivan Novikoff, head of wind energy at French supplier Cegelec, notes that the choice of wind turbine and its technology depends heavily on the location and the characteristics of the local infrastructure. Novikoff says that issues such as cabling, inrush current at start-up, and short-circuit current behaviour depend on the system topology. In specifying a wind turbine for any given application, the company considers a multitude of subordinate but essential issues that range from allowable rotor height and acoustic-noise emissions to the quality of a turbine manufacturer's onsite service. Novikoff explains that, from an investor's viewpoint, economic considerations include the reliability of the wind supply, machine reliability and maintenance costs, and electricity-production tariff variances. In a country that's dominated by nuclear power, Cegelec is about to commission France's largest wind farm, Haute-Lys, in the Pas de Calais (Figure 6). The installation comprises 25 1.5-MW turbines from GE Energy, all of which are variable-speed machines with adaptive VAR (volt-amps-reactive) power generation and low-voltage ride-through capabilities.

You can reach Contributing Technical Editor David Marsh at forncett@btinternet.com.

Talkback

» Submit Feedback

• 
  

  
  

  As a power converter engineer, I got experience in using igbts voltage source drives (with active front end solution) for wind turbine application. I can say that this solution (either for induction or synchronous generators) can solve some issues related to the power quality (very low harmonics, high power factor). It can also provide reactive power to the network, if needed. More, it can provide an efficient solution for riding though the so-called line undervoltage condition, without loosing the generator. It can also be applied to the solution of fig. 5 of this article.

  
  

  
  

  
  Giordano Torri - 2004-25-10 01:24:00 PDT

• 
  

  
  

  Having been experimenting with small scale wind turbines for some years now, I was pleased to read this article. It is the first time I have come across an article that has explained (almost) all the pros & cons of WT design.
  The only thing missing ( maybe because no-one is researching this one) is a design that uses a bevel gearbox behind the WT, which turns the drive through 90 deg. driving a vertical shaft. This allows all the heavy gear to be mounted on the ground. Although I have made a 2.5 kW version of this, the idea is not mine. The first time I came across it was on the Faroes . This machine used a car back axle as the bevel gearbox and about a 5kVA alternator. This approach for making a megawatt size machine has got to be lighter than a direct drive multipole machine.

  
  

  
  

  
  Mr John E. Beech - 2004-18-10 10:44:00 PDT
» Submit Feedback

• Resource Center
• Browse Categories
• Browse Companies

Featured Company

• Mill-Max Mfg. Corp.

  

  The leading US manufacturer of machined interconnect components. Product line includes spring-loaded contacts, SIP, DIP, PGA and BGA sockets, board-to-board interconnects and pin headers, surface mount and application specific products, PCB pins... more

• Rohde & Schwarz

  

  Rohde & Schwarz is an independent group of companies specializing in electronics. It is a leading supplier of solutions in the fields of test and measurement, broadcasting, radiomonitoring and radiolocation, as well as secure communications... more

• ON Semiconductor

  

  ON Semiconductor provides you with the products, solutions and design resources to meet all your power management design needs... more

• Agilent Technologies

  Agilent Technologies Inc. (NYSE: A) is the world's premier measurement company and a technology leader in communications, electronics, life sciences and chemical analysis... more

• Aeroflex

  

  Aeroflex is a leading worldwide provider of highly specialized test and measurement equipment and microelectronic solutions... more

---

Most Recent Resources

• Atmel's QTouch Library 2.0
• Mill-Max Pin Receptacles
• Mill-Max .050" Pitch SMT & Through-Hole Spring-loaded Connectors
• The Impact of Digital Oscilloscope Blind Time on Your Measurements
• LTE-Advanced Technology Introduction