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 – 411 – [411–432/22]

17.12.2004 10:55PM

Part C

Future Concepts

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 412 – [411–432/22]

17.12.2004 10:55PM

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 413 – [411–432/22]

17.12.2004 10:55PM

Wind Power and Voltage Control

J. G. Slootweg, S. W. H. de Haan, H. Polinder and W. L. Kling

19.1 Introduction

The main function of an electrical power system is to trans port electrical power from the

generators to the loads. In order to function properly, it is essential that the voltage is

kept close to the nominal value, in the entire power system. Traditionall y, this is

achieved differently for transmission networks and for distribution grids. In transmis-

sion networks, the large-scale centralised power plants keep the node voltages within the

allowed deviation from their nominal value and the number of dedicated voltage control

devices is limited. These power plants use the energy that is released by burning fossil

fuels or by nuclear fission for generating electrical power. The power they generate

makes up the largest part of the consumed electricity. Distribution grids, in contrast,

incorporate dedicated equipment for voltage control and the generators connected to

the distribution grid are hardly, if at all, involved in controlling the node voltages. The

most frequently used voltage control devices in distribution grids are tap changers –

transformers that can change their turns ratio – but switched capacitors or reactors are

also applied. By using large-scale power plants to regulate voltages in the transmission

network and by using dedicated devices in distribution grids to regulate the voltages at

the distribution level, a well-designed, traditional power system can keep the voltage at

all nodes within the allowed band width.

However, a number of recent developments challenge this traditional approach. One of

these is the increased use of wind turbines for generating electricity. Until now, most wind

turbines have been erected as single plants or in small groups and are connected to

distribution grids. They affect the currents, or power flows, in the distribution grid to which

they are connected, and, because node voltages are strongly related to power flows, they

also change no de voltages. This can lead to problems if the devices installed in the distribution

Wind Power in Power Systems Edited by T. Ackermann

Ó 2005 John Wiley & Sons, Ltd ISBN: 0-470-85508-8 (HB)

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 414 – [411–432/22]

17.12.2004 10:55PM

grid cannot compensate the impact of the wind turbines on the node voltages. In this

case, the voltages at some nodes within the distribution grid cannot be kept within the

allowed deviation from their nominal value, and appropriate measures have to be taken.

A similar problem arises when large-scale wind farms are connected to the transmis-

sion network. Again, the wind farm affects the power flows and hence the node voltages,

but this time in the transmission network. Here, voltages are controlled mainly by large-

scale conventional power plants. If their capability to control voltages within the

transmission network is not sufficient to compensate the impact of the wind farm on

the node voltages, again, the voltage at some nodes can no longer be kept within the

allowable deviation from its nominal value and appropriate measures have to be taken.

This chapter looks at the impact of wind power on voltage con trol. First, we will

review the essential and basic principles of voltage control and comment on the impact

of wind power on voltage control. Where appropriate, a distinction will be made

between transmis sion networks and distribution grids. Then, we will briefly introduce

the three wind turbine types that are most frequently used today and discuss their

voltage control capabilities. We will illustrate voltage control with wind turbines and

use steady-state and dynamic simulation results for this. To conclude, we will discuss the

drawback of voltage c ontrol with wind turbines, namely, the increase in the rating of the

power electronic converter.

19.2 Voltage Control

19.2.1 The need for voltage control

Voltage control is necessa ry because of the capacitance, resistance and inductance of

transformers, lines and cab les, which will hereafter be referred to as branches. Because

branches have a capacitance, resistance and inductance (the first is often referred to as

susceptance, the latter two as impedance), a current flowing through a branch causes a

voltage difference between the ends of the branch (i.e. between the nodes being con-

nected by the branch). However, even though there is a voltage difference between the

two ends of the branch, the node voltage is not allowed to deviate from the nominal

value of the voltage in excess of a certain value (normally 5 % to 10 %). Appropriate

measures must be taken to prevent such deviat ion. Voltage control refers to the task of

keeping the node voltages in the system within the required limits and of preventing any

deviation from the nominal value to become larger than allowed.

It should be noted that if branches would not have an impedance and susceptance, the

voltage anywhere in the system would be equal to that at the generators, and voltage

control would not be necessary. This is a hypothetical case, though. It is also important

to stress that node voltage is a local qua ntity, as opposed to system frequency, which is a

global or systemwide quantity. It is therefore not possible to control the voltage at a

certain node from any point in the system, as is the case with frequency (provided that

no branch overloads are caused by the associated power flows). Instead, the voltage of a

certain node can be controlled only at that particular node or in its direct vicinity. It is

very important to keep this specific property of the voltage control problem in mind,

because otherwise it is impossible to understand what impact the replacement of con-

ventional generation by wind power has on voltage control.

414 Voltage Control

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 415 – [411–432/22]

17.12.2004 10:55PM

There are various ways to affect node voltages. They differ fundamentally between

transmission networks and distribution grids. This is because of the different char-

acteristics of the branches in transmission networks and distribution grids and the

divergent numbers and characteristics of the generators connected to both. Transmis-

sion networks consist of overhead lines with very low resistance. The voltage differ-

ence between two ends of a line with a high inductive reactance, X, when compared

with its resistance, R (i.e. with a low R/X ratio) is strongly affected by what is called

the reactive power flow through the line. Owing to the characteristics of transmission

networks and the connected generators, node voltages are controlled mainly by

changing the reactive power generation or consumption of large-scale centralised

generators connected to the transmission network. These generators are easily acces-

sible by network operators because of their relatively small number and the fact that

they are continuously monitored and staffed. Further, they are very flexible in

operation and allow a con tinuous control of reactive power generation over a wide

range, as can be seen in Figure 19.1, which depicts a generator loading capability

diagram.

Sometimes, dedicated equipment is used [e.g. capacitor banks or technologies

referred to as flexible AC transmission systems (FACTS)]. These are, in principle,

controllable reactive power sources. There are, however, differences with respect to

how the reactive power is generated and with respect to the speed and accuracy of

the control of the generated amount of reactive power. A detailed explanation of the

Prime mover limit

Armature heating limit

Per unit reactive power (p.u.)

–0.6

–0.4

–0.2

0.0

0.2

0.4

0.6

0.8

Underexcitation limit

Operating region

Per unit active power (p.u.)

Field heating limit

0.40.2 0.8 1.00.6

Figure 19.1 Example of a synchronous generator loading capability diagram

Wind Power in Power Systems 415

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 416 – [411–432/22]

17.12.2004 10:55PM

working principles and pros and cons of the various devices that are capable of

generating controllable reactive power and that therefore can be used for voltage

control purposes is beyond the scope of this chapter (for further reading, see, for

example, Hingorani and Gyugyi, 1999).

In contrast, distribution grids consist of overhead lines or underground cables in

which the resistance is not negligible when compared with the inductance (i.e. that have

a much higher R/X ratio than transmission lines). Therefore, the impact of reactive

power on node voltages is less pronounced than in the case of transmission networks.

Further, the generators connected to distribution grids are not always capable of

varying their reactive power output for contributing to voltage control. Node voltages

in distribution grids are therefore controlled mainly by changing the turns ratio of the

transformer that connects the distribution grid to the higher voltage level and sometimes

also by devices that generate or consume reactive power, such as shunt reactors and

capacitors. In general, distribution grids offer far fewer possibilities for node voltage

control than do transmission networks.

19.2.2 Active and reactive power

In most cases, there is a phase difference between the sinusoidal current supplied to the

grid by an alternating current (AC) generator and the voltage at the generator’s

terminals. The magnitude of this phase difference depends on the resistance, induc-

tance and capacitance of the network to which the generator supplies its power and on

the characteristics and the operating point of the generator itself. In the case of AC

generation, the amou nt of generat ed power is, in general, not e qual to the product of

the root-mean-square (RM S) value of generator voltage and current, as in case of

direct current (DC) generation. The amount of generated power depends not only on

the amplitude of voltage and current but also on the magnitude of the phase angle

between them.

The current of an AC generator can be divided into a component that is in phase with

the terminal voltage and a component that is shifted by 90



. It can be shown that only

the component of the current that is in phase with the voltage feeds net power into the

grid (Grainger and Stevenson, 1994). Generated power is hence equal to the product of

the RMS value of the in-phase component of the generator current and the terminal

voltage. This quantity is named the active power or real power. It is measured in watts

(W) or megawatts (MW) and represented by the symbol P.

The product of the RMS values of the source voltage and the out-of-phase (90



shifted) current component is called the reactive power. The unit of reactive power is

voltamperes reactive (VAR). It is represented by the symbol Q and is often referred to as

‘VARs’. The origin of reactive power is the electromagne tic energy stored in the network

inductances and capacitances. The product of the RMS magnitudes of voltage and

current is called the apparent power, which is represented by the symbol S.Itis

measured in volt amperes (VA). The apparent power S does not take into account the

phase difference between voltage and current. It is therefore, in general, not equal to

active power P unless there is no phase difference between the generator current and

terminal voltage.

416 Voltage Control

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 417 – [411–432/22]

17.12.2004 10:55PM

Another important quantity in this context is the power factor (PF). The PF is defined

as follows:

PF ¼ cos ’ ¼

ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ

þ Q

; ð19:1Þ

where ’ is the phase angle between the terminal voltage and terminal current, in degrees

or radians. If ’ equals 0, its cosine equals 1, reactive power Q equals 0 and apparent

power S equals real power P. In that case, only active power is exchanged with the grid

and there is no react ive power. This mode of operation is referred to as ‘cos ’ equal to

one’ or as ‘unity power factor’.

19.2.3 Impact of wind power on voltage control

19.2.3.1 Impact of wind power on voltage control in transmission systems

When discussing the impact of wind power on voltage control, we have to distinguish

between the impact of wind power on voltage control in transmission networks and on

voltage control in distribution grids. As discussed above, in transmission networks, the

task of voltage control was traditionally assigned to large-scale conventional power

plants using synchronous generators, although some dedicated equipment , such as

capacitor banks or FACTS, has been used as well. These traditional, verti cally inte-

grated, utilities have operated power generating units, on the one hand, and power

transmission and distribution systems, on the other hand. They also have handled the

voltage control issue, both short-term (day-to-day dispatch of units) and long-term

(system planning).

Owing to recent developments, this is, however, changing. First, the liberalisation and

restructuring of the electricity sector that is currently being implemented in many

countries has resulted in the unbundling of power generation and grid operation. These

activities are no longer combined in vertically integrated utilities as they used to be. As a

consequence, voltage control is no longer a ‘natural part’ of the planning and dispatch

of power plants. Now, independent generation companies ca rry out the planning and

dispatch, and, in the long term, conventional power stations that are considered unpro-

fitable will be closed down without considering their importance for grid volta ge

control. In addition, grid companies today often have to pay generation companies

for generating reactive power. The grid companies also have to solve any voltage control

problem that may result from the decisions taken by generation companies, either

themselves or supported by the generation company. In the short term, this can be done

by requiring the generation companies to redispatch. In the long term, additional

equipment for controlling the voltage can be installed. Another recent development is

that generation is shifted from the transmission network to the distribution grid.

As a result of these two developments (unbundling and decentralisation), the con-

tribution of conventional power plants to voltage control in transmission networks is

diminishing. It is becoming more difficult to control the voltage in the entire transmis-

sion network from conventional power stat ions only. Grid companies respond by

installing dedicated voltage control equipment and by requiring generation equipment

Wind Power in Power Systems 417

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 418 – [411–432/22]

17.12.2004 10:55PM

to have reactive power capabilities independent of the applied technology. This means

that no exception is made for wind power or other renewables any longer, as has often

been the case until now.

There is a third development that is specific to wind power. Wind farms that are large

enough to be connected to the transmission system tend to be erected in remote areas or

offshore because of their dimension and impact on the scenery. Given that the node

voltage is a local quantity, it can be difficult to control the voltage at these distant places

by use of conventional power stations elsewhere in the grid. Therefore, wind turbines

have to have voltage control capabilities. The voltage control capabilities of various

wind turbine types are expected to become an increasingly important consideration

regarding grid connection and the turbines’ market potential (Wind, 1999).

A final impact of wind power on voltage control that we would like to mention is that

large-scale wind farms may make it necessary to install voltage control devices in the

transmission network, irrespective of the voltage control capabilities of the wind turbines

themselves. In other words: even if the wind turbines have exactly the same voltage control

capabilities as the conventional synchronous generators whose output they replace, there

will be no guarantee that they can fulfil the voltage control task of these generators.

Therefore, it may be inevitable to take additional measures to control the grid voltage.

The local nature of grid voltage is the reason for this. In most cases, wind turbines will

not be erected at or near the locat ion of the conventional power plant whose output they

replace. It is likely that the power generation of a conventional power plant in or near a

load centre, such as a city or a large industrial site, will be replaced by a distant

(offshore) wind farm. However, because voltage is a local quantity, this wind farm will

not be able to control the grid voltage in the vicinity of the power plant that is replaced.

Therefore, additional equipment must be installed near the location of this power plant

in order to control the grid voltage. The reason for the necessary additional measures for

voltage control is therefore not that conventional power generation is replaced by wind

power generation as such, but that generation moves away from the vicinity of the load

to a more distant location.

There are also, however, other developments that lead to a geographical displacement

of gene ration, such as power transit between various countries as a result of market

liberalisation. They can lead to voltage control problems, too, and additional equipment

for voltage control in transmission network s may have to be installed. The problem is

therefore not specific to the replacement of conventional generation by wind power. It

can be induced by any development that leads to significant changes in generator

locations.

19.2.3.2 Impact of wind power on voltage control in distribution systems

Traditionally, the two most important approaches towards voltage control in distribu-

tion grids are the use of tap-changing transformers (i.e. transformers in which the turns

ratio can be changed) and devices that can generate or consume reactive power (i.e.

shunt capacitors or reactors). Use of tap-changing transformers is a rather cumbersome

way of controlling node voltages. Rather than affecting the voltage at one node and in

its direct vicinity, the whole voltage profile of the distribution grid is shifted up or down,

depending on whether the transformer turns ratio is decreas ed or increased. Capacitors

418 Voltage Control

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 419 – [411–432/22]

17.12.2004 10:55PM

and reactors perform better in this respect, because they affect mainly the voltage of the

node to which they are connected. However, the sensitivity of the node voltage to

changes in reactive power is rather limited and therefore relatively large capacitors

and reactors are necessary. This disadvantage is due to the high R/X ratio of the

branches in distribution grids when compared with that in transmission networks.

As was the case for transmission networks, recent developments are complicating the

task of maintaining node voltages throughout distribution grids. More and more distrib-

uted generation (also referred to as embedded, decentralised or dispersed generation; see

Ackermann, Andersson and So

der, 2001), such as wind turbines, solar photovoltaic (PV)

systems and small combined heat and power (CHP) generation is being connected to

distribution grids. These generators affect the power flows in distribution grids. In

particular, if their output power does not correlate with the load, as is the case with

generators using an uncontrollable prime mover, such as wind or sunlight, the variations

in the current through the branches and therefore in the node voltages increase. The

maximum and minimum value of the current through a certain branch used to depend on

the load only, but, with the connection of distributed generation, the current limits have

become dependent on the load as well as on the output of the distributed generator. The

limits are now determined by a situation with minimum generation and maximum load,

on the one hand, and maximum generation and minimum load on the other, rather than

only by the difference between minimum and maximum load, as used to be the case.

One might argue that with an increasing number of generators connected to the

distribution grid, the voltage control possibilities might increase as well. However, in

many cases these small-scale generators are more difficult to use for voltage control than

the generators in large-scale power plants connected to transmission systems because:

they are not always able to vary reactive power generation (depending on the applied

generator type and rating as well as on the rating of the power electronic converter, if

existing);

it may be (very) costly to equip them with voltage control capabilities;

equipping them with voltage control capabilities could increase the risk of ‘islanding’

[i.e. a situation in which (part of) a distribution network remains energised after being

disconnected from the rest of the system];

there are many of them, which makes it very cumbersome to change controller

parameters, such as the voltage set point or time constants, which may be necessary

after a change in the network topology, for example.

There are many reasons for increasing the amount of distributed generation, such as

increased environmental awareness or the desire to reduce investment risk. However, the

impact of an increased amount of distributed generation on power flows and node

voltages in distribut ion grids may lead to problems with maintaini ng node voltages in

distribution grids. In that case, appropriate measures have to be taken. One approach

would be, for example, to install additional tap changers (sometimes with a nominal

turns ratio of 1 further down into the distribution grid; the device is then called a voltage

regulator), reactors and/or capacitors. Another one would be to oblige distributed

generators to contribute to voltage control, despite the above complications and dis-

advantages.

Wind Power in Power Systems 419

//INTEGRAS/KCG/P AGIN ATION/ WILEY /WPS /FINALS_1 4-12- 04/0470855088_ 20_CHA19 .3D – 420 – [411–432/22]

17.12.2004 10:55PM

19.2.3.3 The importance of wind turbine voltage control capabilities

As argued above, the voltage control capabilities of wind turbines erected in large-scale

wind farms and connected to the transmission network become increasingly important.

The contribution of conventional synchronous generators to voltage control is decreas-

ing as a result of unbundling and decentralisation. Further, wind farms are connected at

distant locations at which it is difficult to control the voltage out of conventional power

plants. It has also been shown that as a result of the impact of distributed generation,

such as wind turbines, the range over which currents in distribution grids vary becomes

larger. Because branch currents and node voltages are strongly correlated, this also

applies to the range of variation in node voltages. Wind turbines with voltage control

capabilities could counteract this effect.

It is becoming increasingly important, for wind turbines connected to the transmis-

sion system and for those connected to the distribution system, to be able to contribute

to voltage control. Therefore, the remainder of this chapter will look at the voltage

control capabilities of the various types of wind turbines.

19.3 Voltage Control Capabilities of Wind Turbines

19.3.1 Current wind turbine types

The vast majority of wind turbines that are currently being installed use one of the three

main types of electromechanical conversion system. The first type is known as the

Danish concept. In Section 4.2.3, the Danish concept is introduced as Type A. An

(asynchronous) squirrel cage induction generator is used to convert the mechanical

energy into electricity. Owing to the different operating speeds of the wind turbine rotor

and the generator, a gearbox is necessary to match these speeds. The generator slip

slightly varies with the amount of generated power and is therefore not entirely con-

stant. However, because these speed variations are in the order of 1 %, this wind turbine

type is normally referred to as constant-speed or fixed -speed. The Danish, or constant-

speed, design is nowadays nearly always combined with stall control of the aerodynamic

power, although pitch-controlled constant-speed wind turbin e types have been built too

(type A1; see Section 4.2.3).

The second type uses a doubly fed induction generator instead of a squirrel cage

induction generator, and was introduced as Type C in Section 4.2.3. Similar to the

previous type, it needs a gearbox. The stator winding of the generator is coupled to the

grid, and the rotor winding to a power electronic converter, nowadays usually a back-

to-back voltage source converter with current control loops. In this way, the electrical

and mechanical rotor frequencies are decoupled, because the power electronic converter

compensates the difference between mechanical and electrical frequency by injecting a

rotor current with variable frequency. Variable-speed operation thus becomes possible.

This means that the mechanical rotor speed can be controlled according to a certain goal

function, such as energy yield maximisation or noise minimisation. The rotor speed is

controlled by changing the generator power in such a way that it equals the value

derived from the goal function. In this type of conversion system, the control of

aerodynamic power is usually performed by pitch control.

420 Voltage Control