Burton T. (et. al.) Wind energy Handbook

Подождите немного. Документ загружается.

Moreover, as regards the controller, pitch-regulation introduces the need for fast

response closed loop control, which is not required for the supervisory functions on

a stall-regulated machine. Thus the beneﬁts of pitch control have to be weighed

carefully against all the additional costs involved, including the cost of mainte-

nance.

Another factor that needs to be considered is fatigue loading. This increases

signiﬁcantly on full-span pitch-regulated machines, because the rate of change of

lift coefﬁcient with angle of attack remains at about 2 (see Equation (A3.9) in

Appendix A3.1) instead of reducing to zero as the blade goes into stall, with the

result that rapid changes in wind speed above rated will cause bigger thrust load

changes.

Pitch system controller design is considered in detail in Chapter 8.

6.7.3 Passive pitch control

An attractive alternative to active control of blade pitch to limit power is to design

the blade and/or its hub mounting to twist under the action of loads on the blades

in order to achieve the desired pitch changes at higher wind speeds. Unfortunately,

although the principle is easy to state, it is difﬁcult to achieve it in practice, because

the required variation in blade twist with wind speed generally does not match the

corresponding variation in blade load. In the case of stand- alone wind turbines, the

optimization of energy yield is not the key objective, so passive pitch control is

sometimes adopted, but the concept has not been utilized as yet for many grid-

connected machines.

Corbet and Morgan (1991) give a survey of how different types of blade loads

might be utilized. Harnessing the centrifugal load is obviously promising in the

case of variable speed machines, and this has been demonstrated using a screw

cylinder and preloaded spring to passively control each tip blade, within the Dutch

FLEXHAT programme. When the centrifugal load on the tip exceeds the preload,

the tip blade is driven outwards against the spring and pitches (see Figure 6.13 for

illustration of the concept).

Joose and Kraan (1997) have proposed replacing this mechanism by a mainte-

nance-free ‘Tentortube’, which would twist under tension loading. This tube wo uld

be carbon-ﬁbre reinforced with all the ﬁbres set at an angle to the axis, so that

centrifugal loading induced twist. It would be placed inside a hollow steel tip shaft,

which would carry the aerodynamic loading on the tip blade.

6.7.4 Active stall control

Active stall control achiev es power limitation above rated wind speed by pitching

the blades initially into stall, i.e., in the opposite direction to that employed for

active pitch control, and is thus sometimes known as negati ve pitch control. At

higher wind speeds, however, it is usually necessary to pitch the blades back

towards feather in order to maintain power output at rated.

A signiﬁcant advantage of active stall control is that the blade remains essentially

POWER CONTROL 355

stalled above the rated wind speed, so that gust slicing (see Section 6.7.2) results in

much smaller cyclic ﬂuctuations in blade loads and power output. It is found that

only small changes of pitch angle are required to maintain the power output at

rated, so pitch rates do not need to be as large as for positive pitch control.

Moreover, full aerodynamic braking requires pitc h angles of only about 208, so the

travel of the pitch mechanism is very much reduced compared with positive pitch

control.

Figure 6.11 compares schedules of pitch angle against wind speed for active stall

control and active pitch control for the same blade. The active stall control schedule

is derived from the intersection of the family of power curves for different negative

pitch angles with the 500 kW abscissa in Figure 6.12, while the active pitch control

schedule is derived from Figure 6.7.

The principal disadvantage of active stall control is the difﬁculty in predicting

aerodynamic behaviour accurately in stalled ﬂow conditions. Active stall control is

considered further in Section 8.2.1.

6.7.5 Yaw control

As most horizontal-axis wind turbines employ a yaw drive mechanism to keep the

turbine headed into the wind, the use of the same mechanism to yaw the turbine

out of wind to limit power output is obviously an attractive one. However, there

are two factors which militate against the rapid response of such a system to limit

power: ﬁrst, the large moment of inertia of the nacelle and rotor about the yaw axis,

-5

0 5 10 15 20 25 30

Wind speed (m/s)

Pitch angle (ⴗ)

Active stall control

Active pitch

control

Figure 6.11 Specimen Pitch Angle Schedules for Active Pitch Control and Active Stall

Control

356

CONCEPTUAL DESIGN OF HORIZONTAL-AXIS TURBINES

and second, the cosine relationship between the component of wind speed perpen-

dicular to the rotor disc and the yaw angle. The latter factor means that, at small

initial yaw angles, yaw changes of, say, 108 only bring about reductions in power of

a few percent, whereas blade pitch changes of this magnitude can easily halve the

power output. Thus active yaw control is only practicable for variable speed

machines where the ex tra energy of a wind gust can be stored as rotor kinetic

energy until the yaw drive has mad e the necessary yaw correction. This design

philosophy has been exploited successfully in Italy on the 60 m diameter Gamma 60

prototype, which has an impressive maximum yaw rate of 88/s (Coiante et al.,

1989).

6.8 Braking Systems

6.8.1 Independent braking systems—requirements of standards

DS 472 (Section 5.1.4) and the GL rules (Section 5.1.3) both require that a wind

turbine shall have two independent braking systems. On the other hand, IEC 61400-

1 (Section 5.1.2) does not explicitly require the provision of two braking systems

(stating that the protection system shall include one or more systems capable of

bringing the rotor to rest or to an idling state), but it does require the protection

system to remain effective even after the failure of any non-safe-life protection

system component.

IEC 61400-1 and the GL rules require that at least one of the braking systems

100

200

300

400

500

600

700

800

900

0 5 10 15 20 25 30 35

Wind speed (m/s)

Power output (kW)

-5⬚

-4⬚

-3⬚

-2⬚

0⬚

500 kW

Figure 6.12 Power Curves for Different Negative Pitch Angles: 40 m Diameter Rotor Rotat-

ing at 33 r.p.m.

BRAKING SYSTEMS 357

should act on the rotor or low-speed shaft, but DS 472 goes further and requires

that one should be an aerodynamic brake.

Normal practice is to provide both aerodynamic and mechanical braking. How-

ever, if independent aerodynamic braking systems are provided on each blade, and

each has the capacity to decelerate the rotor after the worst-case grid loss, then the

mechanical brake will not necessarily be designed to do this. The function of the

mechanical brake in this case is solely to bring the rotor to rest, i.e., to park it, as

aerodynamic braking is unable do this.

6.8.2 Aerodynamic brake options

Active pitch control

Blade pitching to feather (i.e., to align the blade chord with the wind direction)

provides a highly effective means of aerodynamic braking. Blade pitch rates of 108

per second are generally found adequate, and this is of the same order as the pitch

rate required for power control. The utilization of the blade pitch system for start-

up and power control means that it is regularly exercised with the result that the

existence of a dormant fault is highly unlikely.

In machines relying solely on blade pitc hing for emergency braking, independent

actuation of each blade is required, together with fail-safe operation should power

or hydraulic supplies passing through a hollow low-speed shaft from the nacelle be

interrupted. In the case of hydraulic actuators, oil at pressure is commonly stored in

accumulators in the hub for this purpose.

Pitching blade tips

Blade tips which pitch to feather have become the standard form of aerodynamic

braking for stall-regulated turbines. Typically the tip blade is mounted on a tip

shaft, as illustrated in Figure 6.13, and held in against ce ntrifugal force during

normal operation by a hydraulic cylinder. On release of the hydraulic pressure

(which is triggered by the control system, or directly by an overspeed sensor), the

tip blade ﬂies outwards under the action of centrifugal force, pitching to feather

simultaneously on the shaft screw. The length of the tip blade is commonly some 15

percent of the tip radius.

The ability of the control system to trigger blade tip activation is of crucial

importance. On a number of early machine designs, the blade tips were centrifug-

ally activated only, so there could be long periods without overspeed events when

they did not operate. As a result there was a risk of seizure when operation was

eventually required. With the now commonplace arrangement enabling the control

system to activate the tip as well, the syst em can be rou tinely tested automatically.

The penalty is that the low-speed shaft needs to be hollow to accommodate the feed

to the hydraulic cylinder.

358 CONCEPTUAL DESIGN OF HORIZONTAL-AXIS TURBINES

Spoilers

Spoilers are hinged ﬂaps, which conform to the aerofoil proﬁle when retracted, and

stick out at right angles to it when deployed. However, although such devices have

been used in the past, they have to be of considerable length in order to decelerate

the rotor adequately (Jamieson and Agius, 1990). Moreover, unless the design

allows for their operation to be regularly teste d, there is a risk that they wil l fail to

deploy when actually needed.

Other devices

Various other devices have been suggested, such as

• ailerons,

• the sliding leading edge device, or SLEDGE, in which a length of leading edge at

the tip slides radially outwards,

• the ﬂying leading edge device, or FLEDGE, in which the whole leading edge

together with an adjacent section of the camber face is pitched towards feather.

Jamieson and Agius (1990) and Armstrong and Hancock (1991) give useful surveys

of these and other aerodynamic braking devices, and note that the SLEDGE device,

which utilizes only 2 percent or 3 percent of the blade area, is highly effective

aerodynamically. Derrick (1992) examines the cap abilities of the SLEDGE and

FLEDGE devices for both braking and power control in more detail. Despite their

promise, these devices have not yet found commercial application.

Rotary bearing

Tip shaft

Screw

Compressed spring

Flow

Blade pitch angle

Figure 6.13 Passive Control of Tip Blade, Using Screw on Tip Shaft and Spring

BRAKING SYSTEMS 359

6.8.3 Mechanical brake options

As noted in Section 6.8.1, the duty of the mechanical brake need only be that of a

parking brake on machines where the aerodynamic brakes can be actuated inde-

pendently. However, on pitch-regulated machines where blade position is con-

trolled by a single actuator, full independent braking capability has to be provided

by the mecha nical brake. It is worth noting that several manufacturers of stal l-

regulated machines ﬁtted with independent tip brakes ensure that the mechanical

brake can stop the rotor unassisted. This may be to satisfy requ irements in certain

countries that two independent braking systems of a different type are provided.

A wind turbine brake typically consists of a steel brake disc acted on by one or

more brake callipers. The disc can be mounted on either the rotor shaft (known as the

low-speed shaft) or on the shaft between the gearbox and the generator (known as

the high-speed shaft). The latter option is much the mo re common because the

braking torque is reduced in inverse proportion to the shaft speeds, but it carries with

it the signiﬁcant disadvantage that the braking torques are experienced by the gear

train. This can increase the gearbox torque rating required by as much as ca 50

percent, depending on the frequency of brake application (see Section 7.4.5). Another

consideration is that the material quality of brake discs mounted on the high-speed

shaft is more critical, because of the magnitude of the centrifugal stresses developed.

The brake callipers are almost always arranged so that the brakes are spring

applied and hydraulically retracted, i.e., fail-safe.

Aerodynamic braking is much more benign than mechanical braking as far as

loading of the blade structure and drive train is concerned, so it is always used in

preference for normal shut-downs.

6.8.4 Parking

versus

idling

Although a mechanical parking brake is essential for bringing the rotor to rest for

maintenance purposes, many manufacturers allow their machines to idle in low

winds and some do so during high wind shut-downs. The idling strategy has two

clear advantages: it reduces the frequency of imposition of braking loads on the

gear train, and gives the impression to members of the public that the turbine is

operating even when it is not generating. On the other hand, gearbox and bearing

lubrication must be maintained throughout.

6.9 Fixed Speed, Two-speed or Variable-speed

Operation

Wind turbine rotors develop their peak efﬁciency at one particular tip speed ratio (see

Figure 3.15), so ﬁxed speed machines operate sub-optimally, except at the wind speed

corresponding to this tip speed ratio. Energy capture can clearly be increased by

varying the rotational speed in proportion to the wind speed so that the turbine

is always running at optimum tip speed ratio, or alternatively a slightly reduced

360 CONCEPTUAL DESIGN OF HORIZONTAL-AXIS TURBINES

improvement can be obtained by runn ing the turbine at one of two ﬁxed speeds so that

the tip speed ratio is closer to the optimum than with a single ﬁxed speed.

Noise considerations are often of greater signiﬁcance than energy capture in the

decision to opt for non-ﬁxed speed operation. As noted in Section 6.4.4, the

aerodynamic noise generated by a wind turbine is approximately proportional to

the ﬁfth power of the tip speed. Both variable speed and two-speed operation allow

the rotational speed to be substantially reduced in low winds, thus reducing turbine

aerodynamic noise dramatically when it could otherwise be objectionable bec ause

of low ambient noise.

6.9.1 Two-speed operation

At one time, two-speed operation was relatively expensive to implement because

separate generators were required for each speed of turbine rotation. Either

generators of differing numbers of poles were connected to gearbox output shafts

rotating at the same speed, or generators with the same number of poles were

connected to output shafts rotating at differing speeds. The rating of the generator

for low-speed operation would normally be much less than the turbine rating.

The development of induction generators with two sets of independent windings

has allowed the number of poles within a single generator to be varied by con-

necting them together in different ways. Standard generators of this type are now

available which can be switched between four- and six-pole operation, giving a

speed ratio of 1.5. Given correct selection of the two operating speeds, this ratio

produces close to the optimum increase of energy capture.

It is important to note that, in the case of stall-regulated machines, only limited

energy gains are to be had by converting from ﬁxed-speed to two-speed operation

for a given rated power, because the maximum rotational speed of the two-speed

machine is restricted to the rotational speed of the ﬁxed speed machine in order to

limit the power to the req uired value. Energy gains are only of the order of 2 or 3

percent, but , nevertheless, two-speed opera tion is often deemed worthwhile on stall

regulated machines because of noise considerations.

In the case of pitch- regulated machines, the energy gain obtainable by moving to

two-speed operation depends on the power rating and rotational speed of the

baseline ﬁxed speed machine. Where these parameters have been chosen to be close

to the optimum, in relation to rotor diameter and rotor chord respectively, an

energy gain of only about 3 percent is attainable. However, energy gains of up to 10

percent are possible when the baseline design is sub-optimal.

Some disadvantages of two-speed operation are:

• additional generator cost,

• extra switchgear is required, which is subjected to a demanding duty in terms of

frequency of operation,

• control of turbine speed is required at each speed change,

• energy is lost while the generator is disconnected during each speed change.

FIXED SPEED, TWO-SPEED OR VARIABLE-SPEED OPERATION 361

6.9.2 Variable-speed operation

By interposing a frequency converter between the generator and the network, it is

possible to dec ouple the rotational speed from the network frequency. As well as

allowing the rotor speed to vary, this also allows the generator air-gap torque to be

controlled.

Variable-speed operation has a number of advantages:

• below rated wind speed, the rotor speed can be made to vary with wind speed to

maintain peak aerodynamic efﬁciency;

• the reduced rotor speed in low winds results in a signiﬁcant reduction in

aerodynamically-generated acoustic noise – noise is especially important in low

winds, where ambient wind noise is less effective at masking the turbine noise;

• the rotor can act as a ﬂywheel, smoothing out aerodynamic torque ﬂuctuations

before they enter the drive train – this is particularly important at the blade

passing frequency;

• direct control of the air-gap torque allows gearbox torque variations above the

mean rated level to be kept very small;

• both active and reactive power can be controlled, so that unity power factor can

be maintained – it is even possible to use a variable speed wind farm as a source

of reactive power to compensate for the poor power factor of other consumers on

the network; variable speed turbines will also produce a much lower level of

electrical ﬂicker .

In practice, losses in the frequency converter may amount to several per cent,

counteracting the increased aerodynamic efﬁciency below rated. In terms of energy

capture, there is often little to choose between a two-speed and a variable-speed

machine. The load reduction possibilities, however, mean that most large MW-scale

turbines now use variable speed in some form. Variations in aerodynamic torque at

blade passing frequency are particularly signiﬁcant in larger turbines because of the

size of the rotor compared to the lateral and vertical length scales of turbulence.

Clearly there is a signiﬁcant cost associated with the variable-speed drive or

frequency converter, which must be weighed against the advantages. Other draw-

backs include increased complexity, although there is no particular evidence of

reduced availability due to power converter problems, and the generation of

electrical noise and harmonics by the inverter system. Modern PWM inverters

produce much lower levels of undesirable harmonics than earlier devices because

of the high switching frequencies which can be achieved. Electrical noi se can be a

problem for control signals within the turbine if insufﬁcient care is taken to shield

cables. Fibre optic transmission is increasingly being used, and this is not affected.

There are two principal methods of achieving variable speed operation: ‘broad

range’ varia ble speed, in which the generator stator is connected to the network via

the frequency converter, and ‘narrow range’ variable speed where both the gen-

362 CONCEPTUAL DESIGN OF HORIZONTAL-AXIS TURBINES

erator stator and rotor are connected to the network, the stator direc tly, and the

rotor through slip rings and a frequency converter.

Broad range variable speed allows the speed to vary under load from zero to the

full rated speed, but all the power output has to pass through the frequency

converter. Narrow range variable speed allows a much cheaper frequency conver-

ter, since only a fraction of the power passes through it, but the speed can only vary

by a more limited amount, typically 30–50% either side of synchronous speed. In

practice, this is enough to achieve almost all the advantages of variable-speed

operation. This is the most commonly used approach. A disadvantage is the

maintenance requirements of the slip rings.

6.9.3 Variable-slip operation

Variable slip represents a compromise between ﬁxed- and variable-speed operation

(Bossanyi and Gamble 1991, Pedersen 1995). The variable-slip generator is essen-

tially an induction generator with a variable resistor in series with the rotor circuit,

controlled by a high-frequency semiconductor switch. Below rated, this acts just

like a conventional ﬁxe d-speed induction generator. Above rated, however, control

of the resistance effective ly allows the air-gap torque to be controlled and the slip

speed to vary, so the behav iour is then similar to a variable-speed system. A speed

range of about 10 percent is typical.

This is cheaper than a variable-speed system, and gives some of the advantages,

in particular the control of torque in the drive train and the smoothing of

aerodynamic torque variations above rated. It does not offer increased aerodynamic

efﬁciency below rated (although it does not suffer from frequency converter losses),

and it does not allow any control of the power factor. Electrical ﬂicker will,

however, be reduced above rated.

Slip rings can be avoided by mounting the variable resistors and control circuitry

on the generator rotor. An advantage of mounting these externally via slip rings is

that it is then easier to dissipate the extra heat which is generated above rated, and

which may otherwise be a limiting factor at large sizes.

6.9.4 Other approaches to variable-speed operation

There are other possible approaches to variable-speed operation, although none of

these has found signiﬁcant commercial application. They include:

• use of a differential gearbox, with the third shaft controlled by a variable speed

electric motor/generator (Law, Doubt and Cooper, 1984, Burton, Mill and

Simpson, 1990) or by a hydraulic pump/motor (Henderson et al., 1990);

• mechanical continuously-variable transmission systems such as have been devel-

oped for automotive applications.

FIXED SPEED, TWO-SPEED OR VARIABLE-SPEED OPERATION 363

6.10 Type of Generator

Fixed-speed wind turbines differ from almost all conventional generating plant by

using induction rather than synchronous generators. This choice is driven by the

requirement for signiﬁcant damping in the drive train due to the cyclic variations in

the torque developed by the aerodynamic rotor.

Both synchronous and induction generators have similar winding arrangements

on the stator which, when connected to the three-phase network voltage, produce a

ﬁxed-speed, rotating magnetic ﬁeld. However, the rotors of the two machines are

quite different (Hindmarsh, 1984, McPherson, 1990). A synchronous machine has

magnets mounted on its rotor and the rotor magnetic ﬁeld then locks into that

produced by the stator leading to operation at synchronous speed. For power

generation applications, electromagnets are used on the roto r excited by an

externally applied direct current. Although the rotor opera tes at the same speed as

the stator magnetic ﬁeld it leads the stator ﬁeld by an angle depending on the

applied torque . In contrast, the rotor of a conventional induction machine has a

‘squirrel cage ’ winding into which currents are induced as the rotor bars cut the

magnetic ﬁeld produced by the stator. Hence, an induction generator can only

develop torque at a rotational speed slightly greater than that of the stator ﬁeld.

This ‘slip speed’ is proportional to the applied torque.

Therefore, to a ﬁrst approximation, the behaviour of a synchronous machine may

be considered to be analogous to a torsional spring. The torque is proportional to

the angle between the rotor and the stator ﬁeld. This angle is known as the load or

power angle. In contrast, an inductio n generator can be thought of as a torsional

damper where the torque is proportional to the difference in speed between the

rotor and the stator ﬁeld (the slip speed). This is illustrated in simple schematic

form in Figure 6.14. It may be seen that if the simple model of a ﬁxed-speed wind

turbine, equipped with a synchronous generator, is excited by the cyclic torque

from the wind-turbine rotor then there is no damping in the drive train to control

the torsional oscillations. It is a simple two-spring, two-mass system. In contr ast,

with an induction generator, the connection of the generator to the network is

represented by a torsional damper. The main cyclic torque of the wind turbine rotor

Blades

Generator

Transmission Connection

Network

Synchronous

generator

Blades

Generator

Transmission Connection

Network

Induction

generator

Spring Damper Mass

Key

Figure 6.14 Mechanical Analogue of Fixed-speed Wind Turbines

364

CONCEPTUAL DESIGN OF HORIZONTAL-AXIS TURBINES