AckermannTh. (ed) Wind Power in Power Systems

Подождите немного. Документ загружается.

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 421 – [411–432/22]

17.12.2004 10:55PM

The third type is called the ‘direct-drive wind turbine’ because it does not need a

gearbox. It corresponds to Type D in Section 4.2.3. A low-speed multipole synchronous

ring generator with the same rotational speed as the wind turbine rotor converts the

mechanical energy into electricity. The generator can have a wound rotor or a rotor with

permanent magnets. The stator is not coupled directly to the grid but to a power electronic

converter. This may consist of a back-to-back voltage source converter or a diode rectifier

with a single voltage source converter. The electronic converter makes it possible to

operate the wind turbine at variable speed. Similar to Type C, pitch control limits the

mechanical power input. Figure 19.2 presents the three main wind turbine types.

Apart from the wind turbine types depicted in Figure 19.2, other types have been

developed and used. They include wind turbines with directly grid coupled synchronous

generators and with conventional synchronous as well as squirrel cage induction gen-

erators combined with a gearbox and coupled to the grid with a full-scale power

electronic converter. Currently, there are hardly any of these wind turbine types on

the market, though, and we will not take them into consideration here.

19.3.2 Wind turbine voltage control capabilities

As mentioned above, node voltages are dependent on branch characteristics and on

branch currents. In transmission networks and distribution grids, node voltage and

Rotor

Direct-drive

synchronous

generator

Grid

Converter

Rotor

Grid

Converter

Doubly fed

(wound rotor)

induction

generator

Gearbox

Squirrel

cage

induction

generator

Compensating

capacitors

(a)

(b)

(c)

Figure 19.2 Widely used wind turbine types: (a) constant-speed wind turbine (Type A); (b)

variable-speed wind turbine with doubly fed induction generator (Type C); and (c) direct-drive

variable-speed wind turbine with multipole synchronous generator (Type D) Note: U ¼ voltage;

I ¼ current; subscripts s, r, c, stand for stator, rotor and converter, respectively

Wind Power in Power Systems 421

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 422 – [411–432/22]

17.12.2004 10:55PM

reactive power are correlated and therefore node voltages can be controlled by changing

the reactive power generation or consumption of generators. An analysis of the voltage

control capabilities of the wind turbine types described in Section 19.3.1 has thus to show

to what extent they can vary their reactive power output. In the following, we will discuss

the voltage control capabilities of each of the wind turbine types described above.

Constant-speed wind turbines (Type A) have squirrel cage induction generators that

always consume reactive power. The amount of the reactive power consumption depends

on the terminal voltage, active power generation and rotor speed. Figure 19.3 illustrates

the relation between terminal voltage, rotor speed, active power generation and reactive

power consumption. This figure shows that a squirrel cage induction generator cannot be

used for voltage control, because it can only consume and not generate reactive power and

because the reactive power exchange with the grid cannot be controlled but is governed by

rotor speed, active power generation and terminal voltage.

–0.08–0.1 –0.06 –0.04 –0.02 0 0.02 0.04 0.06 0.08 0.10

–0.08 –0.06 –0.04 –0.02 0 0.02 0.04 0.06 0.08 0.10

–0.1

Active power generation (p.u.)

Reactive power consumption (p.u.)

Slip

Slip (p.u.)

= 1.1 p.u.

= 1.0 p.u.

= 0.9 p.u.

= 0.8 p.u.

= 1.1 p.u.

= 1.0 p.u.

= 0.9 p.u.

= 0.8 p.u.

–2

–4

(a)

(b)

Figure 19.3 Dependence of (a) active and (b) reactive power of a squirrel cage induction

generator on the rotor speed, with the terminal voltage, U

, as a parameter

422 Voltage Control

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 423 – [411–432/22]

17.12.2004 10:55PM

The fact that a squirrel cage induction generato r consumes reactive power can be a

disadvantage, particularly in the case of large wind turbines or wind farms and/or weak

grids. In such cases, the reactive power consump tion may cause severe node voltage

drops. Therefore, the reactive power consumption of the generator is in most cases

compensated by capacitors, as depicted in Figure 19.2(a). In this way, the reactive power

exchange between the combination of the generator and the capacitors, on the one hand,

and the grid, on the other, can be reduced thus improving the power factor of the system

as a whole.

A conventional capacitor is an uncontrollable source of reactive power. By adding

compensating capacit ors, the impact of the wind turbine on the node voltages is

reduced. But this is only a qualitative improvement. The voltage control capab ilities

as such are not enhanced, because there is still a unique relation between rotor speed,

terminal voltage and active and reactive power generation. The voltage control capabil-

ities of a constant-speed wind turbine can be enhanced only with use of more advanced

solutions instead of conventional capacitors. Such advanced solutions include control-

lable sources of reactive power, such as switched capacitors or capacitor banks, a static

condensor (Statcon) or a static VAR compensator (SVC). An example of this is the use

of an American Superconductor D-VAR

system in combination with a wind farm

with constant-speed wind turbi nes at Minot, ND.

(1)

The reactive power generation of a doubly fed induction generator (Type C wind

turbine) can be controlled by the rotor current, as will be discussed in more detail in

Chapter 25. In this case, there is no unique relation between reactive power and other

quantities, such as rotor speed and active power generation. Instead, at a particular

rotor speed and the corresponding active power generation a widely varying amount of

reactive power can be generated or consumed.

Figure 19.4 illustrates the operating range of a doubly fed induction generator at

nominal terminal voltage. It shows that the amount of reactive power is, to a certain

extent, affected by rotor speed and active power generation, as in the case of a squirrel

cage induction generator, even though it does not directly depend on these quantities.

The reason is that both generator torque and reactive power generat ion depend directly

on the current that the power electronic converter feeds into the rotor. The part of the

current that generates torque depends on the torque set point that the rotor speed

controller derives from the actual rotor speed. The current that is needed to generate

the desir ed torque determines, in turn, the converter capacity that is left to circulate

current to generate or consume reactive power.

In the case of a direct-drive variable-speed wind turbine (Type D), the reactive power

exchange with the grid is not determined by the properties of the generator but by the

characteristics of the grid side of the power electronic converter. The generator is fully

decoupled from the grid. Therefore, the reactive power exchange between the generator

itself and the generator side of the converter as well as between the grid side of the

converter and the grid are decoupled. This means that the power factor of the generator

and the power factor of the grid side of the converter can be controlled indep endently.

(1)

D-Var

is a trademark of American Superconductor.

Wind Power in Power Systems 423

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 424 – [411–432/22]

17.12.2004 10:55PM

Figure 19.5 shows the operating range of a variable-speed wind turbine with a direct-

drive synchronous generator, with the terminal voltage as a parameter. The rotor speed

is not taken into accou nt. As the generator and the grid are decoupled, the rotor speed

hardly affects the grid interaction. It is assumed that at nominal voltage and power the

wind turbine can operate with a power factor of between 0.9 leadi ng and 0.9 lagging.

Figure 19.5 illustrates that a variable-speed wind turbine with a direct-drive synchro-

nous generator allows control of react ive power or terminal voltage, because many

values of reactive power correspond to a single value of active power. It should be noted

Generator speed (p.u.)

Active power (p.u.)

1.5

1.0

0.5

–1

–2

Reactive power (p.u.)

Figure 19.4 Operating range of a doubly fed induction generator

Reactive power (p.u.)

–1.5 –1 –0.5 0 0.5 1.51

1.2

0.6

0.8

0.4

0.2

Active power (p.u.)

= 1.1 p.u.

= 1.0 p.u.

= 0.9 p.u.

= 0.8 p.u.

Figure 19.5 Operating range of a wind turbine with a direct-drive synchronous generator with

the terminal voltage, U

, as a parameter; the curves indicate the boundary of the operating range

with U

as a parameter

424 Voltage Control

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 425 – [411–432/22]

17.12.2004 10:55PM

that in Figure 19.5 the curves indicate the boundary of the operating region, whereas in

Figure 19.3, they indicate a unique relationship.

19.3.3 Factors affecting voltage control

As discussed above, a wind turbine with voltage control capability can control node

voltages by changing the amount of reactive power that is generated or consumed. To

govern the control actions, the voltag e at a certain node is measur ed and fed into the

voltage controller. The controller determines the amount of reactive power to be

generated or consumed, according to the transfer function of the controller. It is easiest

to let the wind turbine control its own terminal voltage but sometimes the voltage at a

node somewhere else in the grid is controlled, although that node must be in the vicinity

of the turbine because of the local nature of the grid voltage. When the measured voltage

is too low, reactive power generation is increased; when it is too high, reactive power

generation is decreased.

The following factors influence the relation between the amount of reactive power

that is generated or consumed and the converter current that controls the reactive power

generation:

generator parameters (only in a wind turbine with a doubly fed induction generator);

the set point of the voltage controller;

values of the resistance, R, and the reactance, X, of the wind turbine grid connection;

the amount of active power flowing through the grid connection.

This chapter will look only at the influence of the last factor (i.e. the amount of active

power flowing through the grid connection). With respect to the other factors, the

following can be said:

Machine parameters are strongly related to machine size; therefore, all doubly fed

induction generat or based wind turbines of a given nominal power will have similar

generator parame ters. Therefore, generator parameters will hardly make a difference

in this context (Heier, 1998).

In practice, the set point of the terminal voltage will always be (nearly) equal to the

nominal value of the voltage. Therefore, the terminal voltage controller set point will

not play a major role.

For a more detailed discussion on the influence of the values of R and X of the grid

connection on the amount of reactive power that needs to be generated to keep the

voltage at the reference value, see Svensson (1996), among others.

19.4 Simulation Results

19.4.1 Test system

We use a simple arti ficial test system in order to quantitatively investigate the voltage

control capabilities of the various wind turbine concepts. The test system consists of a

wind turbine that is connected through an impedance to a strong grid, represented by an

Wind Power in Power Systems 425

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 426 – [411–432/22]

17.12.2004 10:55PM

infinite bus. The nominal power of the wind turbine equals 2 MW, and its nominal

voltage is 700 V. The transformer is a three phase 700 V/10 kV transformer; the grid

voltage is 10 kV. The nominal frequency is 50 Hz. The resistance of the transformer and

the cable is 2:5  and their inductance is 16 mH, which corresponds to a reactance of

5j . Figure 19.6 shows the setup.

19.4.2 Steady-state analysis

First, the constant-speed wind turbine is analysed. For this type, voltage control is not

possible when a compensating capacitor of fixed size is used. There are studies of

systems with a reactive power source that can be controlled, such as capacitor banks

or FACTS (e.g. see Abdin and Xu, 2000). However, the product portfolios of major

wind turbine manufacturers show that this type of equipment is not part of standard

products. Further, when this kind of additional equipment is used it is still not the wind

turbine itself that controls the voltage but the extra device. As this chapter discusses the

voltage control capabilities of wind turbines, rather than dedicated voltage control

equipment, no further attention will be paid to such devices.

Chapter 25 includes the equations that can be used to calculate the amount of reactive

power consumed by a squirrel cage induction generator at a certain terminal voltage and

power output. The generator used here consumes 1.05 MVAR reactive power when

generating nominal active power at nominal terminal vo ltage. It is assumed that this

reactive power demand is fully compensated. To this end, a capacitor of approxim ately

6.8 mF has to be installed at each phase.

Figure 19.7 presents the results of the steady-state analysis of the constant -speed wind

turbine. The horizontal axis shows the active power generated by the wind turbine. The

vertical axis depicts the terminal voltage and the reactive power supplied to the grid. The

reactive power supplied to the grid equals the reactive power generated by the capacitor

minus the reactive power consumed by the squirrel cage inductio n generator.

Figure 19.7 shows that the reactive power decreases an d the terminal voltage increases

when active power generation increases. The decrease in reactive power generation can

be explained by the fact that when the active power generation increases, the reactive

power consumption of the squirrel cage induction generator increases as well. There-

fore, a minor part of the reactive power generated by the capacitor will be supplied to

the grid. The exact quantitative behaviour in a specific situation depends on the para-

meters of the generator and the grid connection (Svensson, 1996).

Strong

grid

Wind

turbine

P, Q

= (2.5 + 5.0j) Ω

Figure 19.6 Test system for investigating the voltage control capabilities of wind turbine types.

Note: U

¼ terminal voltage; P ¼ active power; Q ¼ reactive power; Z

¼ line impedance;

j ¼

ﬃﬃﬃﬃﬃﬃﬃ

1

426 Voltage Control

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 427 – [411–432/22]

17.12.2004 10:55PM

Although the two variable-speed wind turbine types studied here have a different

generation system, they both allow full active and reactive power control, as described

above. In the steady-state analysis, only the terminal voltage and the exchange of active

and reactive power with the grid are studied. Therefore, it is unnecessary to analyse the

steady-state behaviour of the two varia ble-speed wind turbin e types separately.

Figure 19.8 shows the results of the steady-state analysis of variable-speed wind

turbines. In both cases, the wind turbine voltage controller control s the voltage at the

low-voltage (700 V) side of the transformer. The horizontal axis gives the active power

generated by the wind turbine. The vertical axis depicts the terminal voltage and the

generated reactive power. There is a marked difference between Figures 19.7 and 19.8.

In Figure 19.8 the terminal voltage varies; in Figure 19.8 it is constant.

Active power (MW)

734

732

730

728

726

724

722

720

718

716

714

Terminal voltage (V)

Reactive power (MVAR)

0.45

0.35

0.25

0.15

0.05

0.40

0.30

0.20

0.10

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Figure 19.7 Results of the steady-state analysis of a constant-speed wind turbine using the test

system of Figure 19.6. Note: reactive power generated ¼ solid line; terminal voltage ¼ dotted line

Active power (MW)

Terminal voltage (V)

Reactive power (MVAR)

701.0

700.8

700.6

700.4

700.2

700.0

699.8

699.6

699.4

699.2

–1.0

–0.1

–0.2

–0.3

–0.4

–0.5

–0.6

–0.7

–0.8

–0.9

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

699.0

Figure 19.8 Results of the steady-state analysis of a variable-speed wind turbine with voltage

controller, using the test system of Figure 19.6. Note: reactive power generated ¼ solid line;

terminal voltage ¼ dotted line

Wind Power in Power Systems 427

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 428 – [411–432/22]

17.12.2004 10:55PM

Presently, converters in variable-speed wind turbines are often controlled in such a

way that no reactive power is exchanged with the grid. That means that a current with

the same phase angle as the grid voltage is injected, and the power factor, PF, equals 1

(see Section 19.2.2). This mode of operation will be further referred to as unity power

factor operation. Figure 19.9 presents the results of the steady-state analys is of this

operating mode. Owing to the relatively high impedance of the grid connection in the

test system, the variable-speed wind turbines did not perform very well at unity power

factor operation.

19.4.3 Dynamic analysis

This section analyses the dynamic behaviour of the three wind turbine concepts. We use

the same setup as for the steady-state analysis. When analysing the dynamic behaviour,

the two different variable-speed wind turbine types must be treated separately, because

they have different controllers for controlling the reactive power generation. In the wind

turbine with the doubly fed induction generator, reactive power is generated by con-

trolling the rotor current. In the wind turbine with a direct-drive synchronous generator,

reactive power is changed by controlling the grid side of the converter. Because the

transfer functions are different, the performance of the control systems may also differ

and the two types of variable-speed wind turbines are therefore simulated separately. To

illustrate the difference between wind turbines with and without a voltage controller, the

grid voltage will be dropped after 30 seconds from 10 kV to 9.75 kV. In reality, this can

happen when a large load is switched on or when a nearby synchronous generator trips.

Figure 19.10 shows the results of the simulation for a measured wind speed sequence.

Figure 19.10(a) depicts the wind speed sequence. Then, for each of the wind turbi ne

types, the generated active and reactive power and the terminal voltage are shown. For

the variable speed turbines, both voltage control operating mode and unity power factor

operating mode are studied. In the figure, the individual curves are labelled with the

wind turbine type to which they refer.

Active power (MW)

Terminal voltage (V)

Reactive power (MVAR)

735

730

725

720

715

710

705

700

0.0 0.2 0.4

0.6

0.8 1.0 1.2 1.4 1.6 1.8 2.0

0.10

0.08

0.06

0.04

0.02

–0.02

–0.04

–0.06

–0.08

–0.10

0.00

Figure 19.9 Results of the steady-state analysis of a variable-speed wind turbine with unity

power factor using the test system of Figure 19.6. Note: reactive power generated ¼ solid line;

terminal voltage ¼ dotted line

428 Voltage Control

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 429 – [411–432/22]

17.12.2004 10:55PM

Time (s)

010

Constant speed

Doubly fed

Direct drive

2.0

1.8

1.6

1.4

1.2

0.8

Active power (MVV)

Time (s)

010

Constant speed

Variable speed at unity power factor

Direct drive with voltage controller

Doubly fed with voltage controller

1.0

0.8

0.6

0.4

0.2

–0.2

–0.4

–0.6

–0.8

–1.0

Reactive power (MVAR)

Time (s)

010

Constant speed

Variable speed with voltage controller

Direct drive at unity power factor

Doubly fed at unity power factor

740

730

720

710

700

690

680

Terminal voltage (V)

Time (s)

010

Wind speed (m/s)

(a)

(b)

(c)

(d)

Figure 19.10 Results of dynamic simulation of a constant-speed wind turbine and a variable-

speed wind turbine in either voltage control operating mode or unity power factor operating

mode: (a) wind speed, (b) active power, (c) reactive power and (d) terminal voltage. The individual

curves are labelled with the wind turbine type to which they refer

Wind Power in Power Systems 429

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 430 – [411–432/22]

17.12.2004 10:55PM

The following conclusions can be drawn from Figure 19.10:

The voltage fluctuates most rapidly in the case of constant-speed wind turbines. The

reason is that the rotor has a constant rotational speed, because the generator is directly

coupled to the grid. Therefore, the rotating mass cannot be used as an energy buffer, as

is the case in variable-speed wind turbines. Changes in wind speed, as well as the effect

of the tower shadow, are directly reflected in changes in active power. The reactive

power consumed by the generator depends on the active power, which in turn influences

both the reactive power generated and the terminal voltage. This effect would be even

more pronounced if the tower shadow effect were included in the models.

The amplitude of the terminal voltage variations is high in the case of constant-speed

wind turbines and variable-speed wind turbines in unity power factor mode. This is

caused by the high impedance of the grid connection.

The terminal voltage variation is smoothest in the case of variable-speed wind

turbines with a vo ltage controller. As expected, a wind turbine with a terminal voltage

controller performs better than one without a terminal voltage controller.

Although the voltage controller works in a different way, the performance of both

variable-speed wind turbines is similar. If the controllers are well designed, both

variable speed-wind turbines perform equally well with respect to terminal voltage

control.

Finally, it is clear that wind turbines equipped with a voltage controller compensate

the grid voltage drop and keep the voltage at its reference value. Only variable-speed

wind turbine s with a voltage controller are capable of controlling terminal voltage

independently of the grid voltage, unless their operating limits are exceeded.

19.5 Voltage Control Capability and Converter Rating

The conclusion is that a variable-speed wind turbine with a voltage controller performs

best with respect to voltage control. Unfortunately, this does not come for free. For

both variable-speed wind turbines, voltage control requires power electronic converters

with a rating that is higher than the rating for operation at unity power factor.

In order to illustrate this, the converter current was calculated when the wind turbines

studied above generated nominal real power and:

did not generate reactive power;

generated nominal reactive power ;

consumed nominal reactive power.

We assumed that grid voltage and frequency equal their nominal values. The nominal

reactive power exchange with the grid was assumed to equal 1 MVAR. This c orresponds

to a power factor of approximately 0.9. For reactive power consumption and genera-

tion, we also calculated the relative change in converter current when compared with the

converter current in the unity power factor ope ration mode. The results are given in

Table 19.1. Even though the quantitative values are specific to the wind turbines studied

here, the qualitative conclusions can be applied more generally.

430 Voltage Control