Excitation S.

yste,ms

The basic function of an excitation system is to provide direct current to the

synchronous machine field winding. In addition, the excitation system perforrns
control an.d protective functions essential to the satisfactory performance of the power
system by controlling the field voltage and thereby the field current.
The control functions include the control of voltage and reactive power flow,
and the enhancement of system stability. The protective functions ensure that the
capability Iimits of the synchronous machine, excitation system, and other equipment
are not exceeded.
This chapter describes the characteristics and modelling of different types of
synchronous generator excitation systems. In addition, it discusses dynamic
performance criteria and provides definitions of related terms useful in the
identification and specification of excitation system requ.irements. This subject has
been covered in several IEEE reports []-8]. These serve as useful references to
utilities, manufacturers, and system analysts by establishing a common nomenclature,
by standardizing models, and by providing guides for specifications and testing.
Models and terminologies used in this chapter largely conform to these publications.

8.1 EXCITATION SYSTEM REQUIREMENTS

The performance requirements of the excitation system are determined by

considerations of the synchronous generator as well as the power system [6,9].

315
 316 Excitation Systems Chap. 8

Generator considerations

The basic requirement is that the excitation system supply and automatically
adjust the field current of the synchronous generator to maintain the terminal voltage
as the output varies within the continuous capability of the generator. This
requirement can be visualized from the generator V-curves, such as those shown in
Figure 5.19 of Chapter 5. Margins for temperature variations, component failures )

emergency overrating, etc., must be factored in when the steady-state power rating is
determined. Normally, the exciter rating varies from 2.0 to 3.5 kW/MV A generator
rating.
In addition, the excitation system must be able to respond to transient
disturbances with field forcing consistent with the generator instantaneous and shor],
term capabilities. The generator capabilities in this regard are limited by severa]
factors: rotor insulation failure due to high field voltage, rotor heating due to high
field current, stator heating due to high armature current loading, core end heating
during underexcited operation, and heating dueto excess flux (volts/Hz). The thermal limits
have time-dependent characteristics, and the short-term overload capability of the
generators may extend from 15 to 60 seconds. To ensure the best utilization of the
excitation system, it should be capable of meeting the system needs by taking full
advantage of the generator's short-term capabilities without exceeding their limits.

Power system considerations

From the power system viewpoint, the excitation system should contribute to
effective control of voltage and enhancement of system stability. It should be capable
of responding rapidly to a disturbance so as to enhance transient stability, and of
modulating the generator field so as to enhance small-signal stability.
Historically, the role of the excitation system in enhancing power system
performance has been growing continually. Early excitation systems were controlled
manually to maintain the desired generator términal voltage and reactive power
loading. When the voltage control was first automated, it was very slow, basically
filling the role of an alert operator. In the early l 920s, the potential for enhancing
small-signal and transient stability through use of continuous and fast-acting regulators
was recognized. Greater interest in the design of excitation systems developed, and
exciters and voltage regulators with faster response were soon introduced to the
industry. Excitation systems have since undergone continuous evolution. In the early l
960s, the role of the excitation system was expanded by using auxiliary stabilizing
signals, in addition to the terminal voltage error signal, to control the fíeld voltage to
damp system oscillations. This part of excitation control is referred to as the power
system stabilizer. Modern excitation systems are capable of providing practically
instantaneous response with high ceiling voltages. The combination of high field-
forcing capability and the use of auxiliary stabilizing signals contributes to substantial
enhancement of the overall system dynamic performance. This will be discussed in
detail in Chapters 12, 13, and 17.
. sec. s.z Elements of an Excitation System 317

To fulfill the above roles satisfactorily, the excitation system must satisfy the
following requirements:

• Meet specified response criteria .

• Provide limiting and protective functions as required to prevent damage to

itself, the generator, and other equipment.

• Meet specified requirements for operating flexibility .

• Meet the desired reliability and availability, by incorporating the necessary

Ievel of redundancy and interna} fault detection and isolation capability.

8.2 ELEMENTS OF AN EXCITATION SYSTEM

Figure 8.1 shows the functional block diagram of a typical excitation control
system for a large synchronous generator. The following is a brief description of the
various subsystems identified in the figure.

Limiters and
protective circuits -
®
Terminal voltage
transducer and
load compensator
-
~
@ CD
- To power
Ref.
-
Regulator - Exciter -....- Generator system

r ©
Power system
stabilizer -
Figure 8.1 Functional block diagram of a synchronous
generator excitation control system
 318 Excitation Systems Chap. 8

(1) Exciter. Provides de power to the synchronous machine field winding,

constituting the power stage of the excitation system.

(2) Regulator. Processes and amplifies input control signals to a level and fonn.
appropriate for control of the exciter. This includes both regulating and
excitation system stabilizing functions (rate feedback or lead-lag
compensation).

(3) Terminal voltage transducer and load compensator. Senses generator terminal
voltage, rectifies and filters it to de quantity, and compares it with a reference
which represents the desired terminal voltage. In addition, load (or line-drop,
or reactive) compensation may be provided, if it is desired to hold constant
voltage at sorne point electrically remote from the generator terminal (for
example, partway through the step-up transformer).

(4) Power system stabilizer. Provides an additional input signa! to the regulator to
damp power system oscillations. Sorne commonly used input signals are rotor
speed deviation, accelerating power, and frequency deviation.

( 5) Limiters and protective circuits. These include a wide array of control and
protective functions which ensure that the capability limits of the exciter and
synchronous generator are not exceeded. Sorne of the commonly used
functions are the field-current limiter, maximum excitation limiter, terminal
voltage limiter, volts-per-Hertz regulator and protection, and underexcitation
limiter. These are normally distinct circuits and their output signals may be
applied to the excitation system at various locations as a summing input or a
gated input. For convenience, they have been grouped and shown in Figure 8.1
as a single block.

8.3 TYPES OF EXCITATION SYSTEMS

Excitation systems have taken many forms over the years of their evolution.
They may be classified into the following three broad categories based on the
excitation power source used [4,8]:

• DC excitation systems
• AC excitation systems
• Static excitation systems

This section provides a description of the above classes of excitation systems,

the different forms they take, and their general structure. Details regarding various
regulating and protective functions often included with the excitation systems will be
covered in Section 8.5.
sec. s.s Types of Excitation Systems 319

g,3.1 OC Excitation Systems

The excitation systems of this category utilize de generators as sources of

xcitation power and provide current to the rotor of the synchronous machine through
~lip rings. The exciter may be -driven by a motor or the shaft of the generator. It may
be either self-excited or separately excited. When separately excited, the exciter field
is supplied by a pilot exciter comprising a permanent magnet generator.
DC excitation systems represent early systems, spanning the years from the
l920s to the 1960s. They lost favour in the mid-1960s and were superseded by ac
exciters.
The voltage regulators for such systems range all the way from the early non-
continuously acting rheostatic type to the later systems utilizing many stages of
magnetic amplifiers and rotating amplifiers [ 1O,11].
DC excitation systems are gradually disappearing, as many older systems are
being replaced by ac or static type systems. In sorne cases, the voltage regulators
alone have been replaced by modem solid-state electronic regulators. As many of the
de excítation systems are still in service, they still require modelling in stability
studies.
Figure 8.2 shows a simplified schematic representation of a typical de
excitation system with an amplidyne voltage regulator. It consists of a de commutator
exciter which supplies direct current to the main generator field through slip rings.
The exciter field is controlled by an amplidyne.

DC exciter Main generator

Field Annature Field Annature
Amplidyne
1
r----------------,
1

':
1
Exc. field
! 1 rheostat
L------ ---------1

Voltage
regulator

Figure 8.2 DC excitation system with amplidyne voltage regulator

320 Excitation Systems Chap. 8

An amplidyne is a special type of the general class of rotating amplifiers

known as metadynes [11,12]. It is a de machine of special construction having tw0
sets of brushes 90 electrical degrees apart, one set on its direct (d) axis and the other
set on its quadrature (q) axis. The control-field windings are located on the d-axis. A
compensating winding in series with the d-axis load produces flux equal and opposite
to the d-axis armature current, thereby cancelling negative feedback of the armature
reaction. The brushes on the q-axis are shorted, and very little control-field power is
required to produce a large current in the q-axis armature. The q-axis current produces
the principal magnetic field, and the power required to sustain the q-axis current is
supplied mechanically by the motor driving the amplidyne. The result is a device with
power amplification on the order of 10,000 to 100,000 and a time constant in the
range from 0.02 to 0.25 seconds.
In the excitation system of Figure 8.2, the amplidyne provides incremental
changes to the exciter field in a "buck-boost" scheme. The exciteroutput provides the
rest of its own field by self-excitation. If the amplidyne regulator is out of service, the
exciter field is on "manual control" and is changed through adjustment of the field
rheostat.

8.3.2 AC Excitation Systems

The excitation systems of this category utilize altemators (ac machines) as

sources of the main generator excitation power. Usually, the exciter is on the same
shaft as the turbine generator. The ac output of the exciter is rectified by either
controlled or non-controlled rectifiers to produce the direct current needed for the
generator field. The rectifiers may be stationary or rotating.
The early ac excitation systems used a combination of magnetic and rotating
amplifiers as regulators [11]. Most new systems use electronic amplifier regulators.
AC excitation systems can thus take many forms depending on the rectifier
arrangement, method of exciter output control, and source of excitation for the exciter
[13-17]. The following is a description of different forms of ac excitation systems in
use.

(a) Stationary rectifier systems

With stationary rectifiers, the de output is fed to the field winding of the main
generator through slip rings.
When non-controlled rectifiers are used, the regulator controls the field of the
ac exciter, which in turn controls the exciter output voltage. A simplified one-line
diagram of such a field controlled alternator rectifier excitation system is shown in
Figure 8.3. In the system shown, which is representative of the General Electric
ALTERREX1 excitation system [14], the alternator exciter is driven from the main

1 ALTERREX is a trademark of General Electric Co.

sec. 8.3 Types of Excitation -Svstems 321

enerator rotor. The exciter is self-excited with its field power derived through
~yristor rectifiers. The voltage regulator derives its power from the exciter output
voltage.
An alternative form of field-controlled alternator rectifier system uses a pilot
exciter as the source of exciter field power.
When controlled rectifiers (thyristors) are used, the regulator directly controls
the de output voltage of the exciter. Figure 8.4 shows the schematic diagram of such
an alternator supplied controlled-rectifier system, representative of the General
Electric ALTHYREX 1 System [17]. The voltage regulator controls the firing of the
thyristors. The exciter altemator is self-excited and uses an independent static voltage
regulatorto maintain its output voltage. Since the thyristors directly control the exciter
output, this system inherently provides high initial response (small response time).
As shown in Figures 8.3 and 8.4, two independent modes of regulation are
provided: (1) ac regulator to automatically maintain the main generator stator terminal
voltage at a desired value corresponding to the ac reference; and (2) de regulator to
maintain constant generator field voltage as determined by the de reference. The de
regulator or manual control mode caters to situations where the ac regulator is faulty
or needs to be disabled. The input signals to the ac regulator include auxiliary inputs
which provide additional control and protective functions which will be described in
Section 8.5.

AC exciter Main generator

Field Armature Field Armature

DC
C'-o-nt_ro_l_le_d , +<}-
.
f---
,,.....--' 1
1
DC
regulator
ref
1
rectifier 1
1
1
1
¡
AC
~
1
1
1
ref.
AC
regulator
----- Aux. inputs

Figure 8.3 Field-controlled altemator rectifier excitation system

1
ALTHYREX is a trademark of General Electric Co.
322 Excitation Systems Chap.a

AC exciter Main generator

Field J\.rrnature Field Armature

e¡ '-,,-'
TPT
DC
DC ref.
regulator
J\C
Exc. ref.
reg. J\C

regulator Aux. inputs

Figure 8.4 Alternator-supplied controlled-rectifier excitation system

(b) Rotating rectifier systems

With rotating rectifiers, the need for slip rings and brushes is eliminated, and
the de output is directly fed to the main generator field. As shown in Figure 8.5, the
armature of the ac exciter and the diode rectifiers rotate with the main generator field.
A small ac pilot exciter, with a permanent magnet rotor (shown as NS in the figure),
rotates with the exciter armature and the diode rectifiers. The rectified output of the
pilot exciter stator energizes the stationary field of the ac exciter. The voltage
regulator controls the ac exciter field, which in turn controls the field of the main
generator.
Such a system is referred to as a brushless excitatign system. It was developed
to avoid problems with the use ofbrushes that were perceived to exist when supplying
the high field currents of very large generators; for example, the power supplied to
the field of a 600 MW generator is on the order of 1 MW. However, with well-
maintained brushes and slip rings, these perceived problems <lid not actually develop.
AC excitation systems with and without brushes have performed equally well.
High initial-response performance of brushless excitation can be achieved by
special design of the ac exciter and high voltage forcing of the exciter stationary field
winding. An example of such a system is the Westinghouse high initial response
brushless excitation system [13].
Brushless excitation systems do not allow direct measurement of generator
field current or voltage. Manual control of main generator voltage is provided by an
adjustable de input setting to the thyristor gating circuits. For the sake of simplicity, ,
the functions of the control circuitry are not shown in detail in Figure 8.5.
 sec. 8.3 Types of Excitation Systems 323

Pilot exciter Rotating structure Main generator

r-----------------------------------------------1
1

Armature Field : Armature

1
1

~-_, -et
1
1

f---CX ¡
1
=
]1

Three- Field
phase ac

Regulator ------ Manual control

----Aux. inputs
~----~

Figure 8.5 Brushless excitation system

8.3.3 Static Excitation Systems

All components in these systems are static or stationary. Static rectifiers,

controlled or uncontrolled, supply the excitation current directly to the field of the
main synchronous generator through slip rings. The supply of power to the rectifiers is
from the main generator (or the station auxiliary bus) through a transformer to step
down the voltage to an appropriate level, or in sorne cases from auxiliary windings
in the generator.
The following is a description of three fonns of static excitation systems that
have been widely used.

(a) Potential-source controlled-rectifier systems

In this system, the excitation power is supplied through a transfonner from the
generator terminals or the station auxiliary bus, and is regulated by a controlled
rectifier (see Figure 8.6). This type of excitation system is also commonly known as a
bus-fed or transformer-fed static system.
This system has a very small inherent time constant. The maximum exciter
output voltage (ceiling voltage) is, however, dependent on the input ac voltage. Hence,
during system-fault conditions causing depressed generator terminal voltage, the
· available exciter ceiling voltage is reduced, This limitation of the excitation system
is, to a large extent, offset by its virtually instantaneous response and high post-fault
field-forcing capability [18, 19]. In addition, it is inexpensive and easily maintainable.
For generators connected to large power systems such excitation systems perform
· satisfactorily [18]. Compounding ancillaries, such as those described below, are not
normallyjustified; they are likely important for generators feeding power directly into
small industrial networks with slow fault-clearing. -
324 Excitation Systems Chap. 8

Exciter Main generator

transformer Controlled rectifier Field J\rtnature

Three-
phase ac*
Field
DC DC ref.
regulator
* Altematively, from
auxiliary bus AC ref
AC
regulator Aux. inputs

Figure 8.6 Potential-source controlled-rectifier excitation system

Examples of this type of excitation system are: Canadian General Electric

Silicomatic excitation system, Westinghouse type PS excitation system, General
Electric potential source static excitation system, and ABB, Reyrolle-Parsons, GEC-
Eliott, Toshiba, Mitsubishi, and Hitachi static excitation systems.

(b) Compound-source rectifier systems

The power to the excitation system in this case is formed by utilizing the
current as well as the voltage of the main generator. This may be achieved by means
of a power potential transformer (PPT) and a saturable-current transformer (SCT) as
illustrated in Figure 8.7. Altematively, the voltage and current sources may be
combined by utilizing a single excitation transformer, referred to as a saturable-current
potential transformer (SCPT). ·
The regulator controls the exciter output through controlled saturation of the
excitation transformer. When the generator is not supplying a load, the armature
current is zero and the potential source supplies the entire excitation power. Under
loaded conditions, part of the excitation power is derived from the generator current.
During a system-fault condition, with severely depressed generator terminal voltage,
the current input enables the exciter to provide high field-forcing capability.
Examples of this type of excitation system are General Electric SCT-PPT and
SCPT static excitation systems.
sec. s.s Types of Excitation Svsterns 325

Saturable Armature
current transformer

To neutral Power rectifier Field

grounding Slip Power
rmgs potential
transformer
Current
1

source Voltage source

Linear reactor

Regulator
---- Aux. inputs

Figure 8.7 Compound-source rectifier excitation system

(e) Compound-controlled rectifier excitation systems

This system utilizes controlled rectifiers in the exciter output circuits and the
compounding of voltage and current-derived sources within the generator stator to
provide excitation power. The result is a high initial-response static excitation system
with full "fault-on" forcing capability.
An example of this type of system is the compound power-source
GENERREX1 excitation system [15,16]. Figure 8.8 shows an elementary single-line
diagram of the system. The voltage source is formed by a set of three-phase windings
placed in three slots in the generator stator and a series linear reactor. The current
source is obtained from current transformers mounted in the neutral end of the stator
windings. These sources are combined through transformer action and the resultant
ac output is rectified by stationary power semiconductors. The means of control is
provided by a combination of diodes and thyristors connected to form a shunt bridge.
A static ac voltage regulator controls the firing circuits of the thyristors and thus
regulates the excitation to the generator field.
The excitation transformer consists of three single-phase units with three
windings: current (C) and potential (P) primary windings, and a secondary output
winding (F). Under fault conditions, the fault current flowing through the excitation
transformer "C" windings provides the field-forcing capability when the generator
voltage is depressed.

1 GENERREX is a trademark of General Electric Co.

 326 Excitation Systems Chap. a
Excitation
transformer
Generator frame Generator

i1
:"C''
'P"::
11
::
CTT TPT
1 11 '----.,-'
i
---------
"F" ----------
¡¡
1
Field
~------ !

Rectifier
with shunt
thyristor DC A_ux. inputs
control regulator

DC ref. AC ref.

Figure 8.8 GENERREX compound-controlled rectifier

excitation system ©JEEE1976 [16]

The reactor serves two functions: contribution to the desired compounding

characteristic of the excitation system and reduction of fault currents for faults in the
excitation system or the generator.
The excitation transformers and reactors are contained in an excitation dome
that is bolted to the top of the generator frame, forming an integral part of the frame.

Field .flashing for static exciters:

Since the source of power to a static excitation system is the main generator,
it is in effect self-excited. The generator cannot produce any voltage until there ÍS'
sorne field current. It is therefore necessary to have another source of power for a few
seconds to initially provide the field current and energize the generator. This process
of build-up of generator field flux is called field flashing. The usual field-flashing
source is a station battery.

8. 3.4 Recent Developments and Future Trends

The advances in excitation control systems over the last 20 years have been
influenced by developments in solid-state electronics. Developments in analog-
integrated circuitry have made it possible to easily implement complex control
strategies.
 sec. a.4 Dynamic Performance. Measures 327
The latest development in excitation systems has been the introduction of
digital technology. T?yristors. continue to be use? for the power. s~áge. The co~trol,
rotection, and logic functions have been implemented digitally, essentially
~uplicating the functions previously provided by analog circuitry.
The digital controls are likely to be used extensively in the future as they
rovide a cheaper and possibly more reliable alternative to analog circuitry. They have
~e added advantage of being more flexible, allowing easy implementation of more
cornplexcontrol strategies, and interfacing with other generator control and protective
[unctions.

s.4 DVNAMIC PERFORMANCE MEASURES

The effectiveness of an excitation system in enhancing power system stability

is determined by sorne of its key characteristics. In this section, we identify and define
performance measures which determine these characteristics and serve as a basis for
evaluatingand specifying the performance of the excitation control system. Figure 8.9
·shows the representation of the overall excitation control system in the classical form
used for describing feedback control systems.

Power Plant
.
Controller VR amplifier Efd
(generator and Et
~ (regulator)
- (exciter) power system)
ve
Feedback
elements

Figure 8.9 Excitation control system in the classical feedback control form

The performance of the excitation control system depends on the characteristics

of excitation system, the generator, and the power system. Since the system is
nonlinear, it is convenient to classify its dynamic performance into large-signal
performance and small-signal performance. For large-signal performance, the
nonlinearities are significant; for small-signal performance, the response is effectively
linear.

8.4.1 Large-Signal Performance Measures [7]

Large-signal performance measures provide a means of assessing the excitation

system performance for severe transients such as those encountered in the
 328 Excitation Systems Chap, a·
consideration oftransient, mid-term and long-term stability of the power system. Such
measures are based on the quantities defined below. To permit maximum flexibility
in the design, manufacture, and application of excitation equipment, sorne of the
performance measures are defined "under specified conditions"; these conditions rnay
be specified as appropriate for the specific situation.

(a) Excitation system ceiling voltage: The maximum direct voltage that the excitation
system is able to supply from its terminals under specified conditions (7,20].
Ceiling voltage is indicative of the field-forcing capability of the excitation
system; higher ceiling voltages tend to improve transient stability.
For potential source and compound source static excitation systems, whose
supply depends on the generator voltage and current, the ceíling voltage is defined at
specified supply voltage and current. For excitation systems with rotating exciters, the
ceiling voltage is determined at rated speed.

(b) Excitation system ceiling current: The maximum direct current that the excitation
system is able to supply from its terminals for a specified time [7,20].
When prolonged disturbances are a concern, the ceiling current may be based
on the excitation system thermal duty.

(e) Excitation system voltage time response: The excitation system output voltage
expressed as a function of time under specified conditions [7,20].

(d) Excitation system voltage response time: The time in seconds for the excitation
voltage to attain 95% of the difference between the ceiling voltage and rated load-
field voltage under specified conditions [7,20].
The rated load field voltage is the generator field voltage under rated
continuous load conditions with the field winding at (i) 75ºC for windings designed
to operate at rating with a temperature rise of 60ºC or less; or (ii) 1 OOºC for windings
designed to operate at rating with a temperature rise greater than 60ºC.

(e) High initial-response excitation system: An excitation system having a voltage

response time of 0.1 seconds or less [7]. It represents a high response and fast-acting
system.

(/) Excitation system nominal response 1: The rate of increase of the excitation
system output voltage determined from the excitation system voltage response curve,
divided by the rated field voltage. This rate, if maintained constant, would develop the
same voltage-time area as obtained from the actual curve over the first half-second
interval (unless a different time interval is specified) [7,20].

1 Historically, the excitation system nominal response has been referred to as the exatation
system response ratio (see 1978 version of [7], and [20]).
sec- 8.4 Dynamic Performance Measures 329

The nominal response is determined by initially operating the excitation system

t the rated load field voltage (and field current) and then suddenly creating the three-
abase terminal voltage input signal conditions necessary to drive the excitation system
poltage to ceiling. It should include any delay time that may be present before the
vxcitation system responds to the initiating disturbance.
e Referring to Figure 8.1 O, the excitation response is illustrated by line ac. This
Iine is determined by establishing area acd equal to area abd:
cd
N ominal response =
(ao)(oe)
where
oe = 0.5 s
ao = rated load field voltage
The basis for considering a nominal time span of 0.5 s in the above definition
is that, following a severe disturbance, the generator rotor angle swing normally peaks
between 0.4 s and O. 75 s. The excitation system must act within this time period to
be effective in enhancing transient stability. Accordingly, 0.5 s was chosen for the
definition time period of nominal response.
In the past, the nominal response has been a well-established and useful
criterion for evaluating the large-signal performance of excitation systems. With older
and slower excitation systems, this was an acceptable performance measure, but it is

Line ac such that 1

area acd = area abd ----- :e
...... 1
... rb
......... »>

...... r
...... 1 Actual response
_,~' 1
...... 1
..,. ... .,,. 1

-~----------------------------~d
,,.,. ...... -- 1
a Rated load field voltage
1
'1
1

1
1
1
1
1
1
1
r
1
1
1
1
1
1
Time in seconds
Ü'-------------------------------------
e

Figure 8.10 Excitation system nominal response

330 Excitation Systems Chap. 8

not adequate to cover many of the modem excitation systems. In particular, it is not
a good figure of merit for excitation systems supplied from the generator or the power
system, due to the reduced capability of such systems during a system fault.
For high initial-response excitation systems, the nominal response merely
establishes the required ceiling voltage. The ceiling voltage and voltage response time
are more meaningful parameters for such systems.

8.4.2 Small-Signal Performance Measures [3, 71

Small-signal performance measures provide a means of evaluating the response

of the closed-Ioop excitation control systems to incremental changes in system
conditions, In addition, small-signal performance characteristics provide a convenient
means for determining or verifying excitation system model parameters f or system
studies.
Small-signal performance may be expressed in terms of performance indices
used in feedback control system theory:

• Indices associated with time response; and

• Indices associated with frequency response

The typical time response of a . feedback control system to a step change in

input is shown in Figure 8.11. The associated indices are rise time, overshoot, and
settling time.
Output

- ~l~ ~ -~ ~- - - - - - - - - -
Specified band
for settling time

Peak
val u e

0.1
Initial value
Delay time Time
Time to reach
peak value
Sett 1.in time

Figure 8.11 Typical time response to step input. ©IEEE 1990 [7]
 sec. 8.4 Dynamic Performance Measures 331

40 Asymptotic o
approximation -40 dB/dec
r/'J
(1)
(1)
of gain 6h
(1)
20 -50 "O

~ ·-
t::
,-..,
-e-
....
i::::
o 1 -100 '-'
o
i::::
·~
1
1 <Pe ~
--- -- -G-m---~----------
o
-20
~
:
1

1 <Pm
-150
~
ª
!l.)

..d
e : 180º ~
Excitation Synchronous
--- - -----------
-40 system machine -200
VR
+ VE Vr
G2

-60 -250

0.01 0.1 1.0 10 100 1000

co in rad/s

Figure 8.12 Typical open-loop frequency response of an excitation

control system with generator open-circuited. © IEEE 1990 [7]

A typical open-loop frequency response characteristic of an excitation control

system with the generator open-circuited is shown in Figure 8.12.
The performance indices associated with the open-loop frequency response are
the low frequency gain G, crossover frequency coc, phase margin $m, and gain margin
Gm. Larger values of G provide better steady-state voltage regulation, and larger
crossover frequency roe indicates faster response. Larger values of phase margin $m
and gain margin Gm provide a more stable excitation control loop. (The reference here
is to excitation control system stability and not power system synchronous stability.)
In tuning the voltage regulator, an improvement to one index will most likely
be to the detriment of other indices. For example, an increase in regulator gain will
shift the gain curve in Figure 8.12 upward. This has the beneficia} effect of increasing
the low-frequency gain and crossover frequency, but has the undesirable effect of
decreasing the gain and phase margins. In general, a phase margin of 40º or more and
a gain margin of 6 dB or more are considered a good design practice for obtaining
a stable, non-oscillatory voltage regulator system.
Figure 8.13 shows the corresponding closed-loop frequency response with the
generator open-circuited.
332 Excitation Systems Chap, 8

M
o -------------~Al 50
V)

Bandwidth t P (l)

(l.)
-3 ~
OS)
<P lú)M
o (1.)
-5 "O

o:¡
·-
i:::::

"O ~
-e-
·-
i:::
-1 o -50 '--'

·ti-
(l)
e:: ~
c.tj
§
(l)

-15 -100 V)
c.tj
.r:::
e,

Excitation Synchronous
-20 machine -150
VT
VR G2

-25 -200

0.1 1.0 10 100

co in rad/s

Figure 8.13 Typical closed-loop frequency response with

generator open-circuited. © IEEE 1990 [7]

The indices of interest associated with the closed-loop frequency response are
the bandwidth ro 8 and peak value MP.
A high value of MP (> 1.6) is indicative of an oscillatory system exhibiting
large overshoot in its transient response. In general, a value of MP between 1.1 and
1. 5 is considered a good design practice.
Bandwidth is an important closed-loop frequency response index. Larger values
indicate faster response. It approximately describes filtering or noise-rejection
characteristics of the system.
Generally accepted values of performance indices characterizing good feedback
control system performance are:

Gain margin ¿ 6 dB
Phase margin ¿ 40°
Overshoot = 5-15%
MP = 1.1-1.6
 Control and Protective Functions 333
sec. 8.5

It is not possible to define such generally acceptable ranges of values for other
all-signal performance indices: rise time, settling time, and bandwidth. These
~rnd. ces are a measure of the relative speed of the control action. They are primarily
dinetterrninedby t he synchr?º~us m~chine dynarm·c e ha~acten·st1·cs.

The performance indices grven above are applicable to any feedback control
stem having a single majar feedback loop, i.e., a single controlled-output variable.
itierefore, they are applicable to an excitation control system with the synchronous
achine on open circuit or feeding an isolated load. Stable operation of the excitation
~ntrol system with the generator off-line is ensured based on these performance
~ndices and associated analytical techniques [3]. On the other hand, synchronous
~achines connected to a power system form a complex multiloop, multivariable, high-
order control system. Por such a system, the performance indices identified above are
not applicable. The state-space approach using eigenvalue techniques is an effective
method of assessing the performance of such complex systems. This is covered in
detail in Chapter 12.

8.5 CONTROL ANO PROTECTIVE FUNCTIONS

A modern excitation control system is much more than a simple voltage

regulator. It includes a number of control, limiting, and protective functions which
assist in fulfilling the performance requirements identified in Section 8.1. The
extensive nature of these functions and the manner in which they interface with each
other are illustrated in Figure 8.14. Any given excitation system may include only
sorne or all of these functions, depending on the requirements of the specific
application and on the type of exciter. The philosophy is to have the control functions
regulate specific quantities at the desired level, and the limiting functions prevent
certain quantities from exceeding set limits. If any of the limiters fail, then the
protective functions remove appropriate components or the unit from service.
The following is a brief description of the various control and protective
functions, and the associated elements identified in Figure 8.14.

8.5.1 AC and DC Regulators

The basic function of the ac regulator is to maintain the generator stator

voltage. In addition, other auxiliary control and protection functions act through the
ac regulator to control the generator field voltage as shown in Figure 8.14.
The de regulator holds constant generator field voltage and is commonly
referred to as manual control. It is used primarily for testing and start-up, and to cater
to situations where the ac regulator is faulty. In this mode of operation, it is the field
voltage that is regulated; only operator intervention by adjusting the setpoint will
modify the field voltage. In sorne excitation systems, facilities for automatic setpoint
tracking are provided. This will cause the manual setpoint to continually track the
generator excitation variation due to the ac regulator and thus minimize the voltage
 334 Excitation Systems Chap. 8
~
Voltage sensing and
load compensation
Power system
stabilizer
Voltage
sensmg
**

~
-- DC
regulator 11
1
1
1

DC voltage 1 Field *
1
1 Exciter ~ Generator
adjust 1
1 shorting
1
AC 11Y
11'1
~ regulator
AC voltage ~ r Exc. sys.
adjust
stab. circuits

Overexc. ***
limiter

Underexc.
limiter

V/Hz limiter
and protection

* Field-shorting circuits are applicable to ac and static exciters only.

** Sorne systems have open-loop de regulator.
*** Overexcitation limiter may also be used with de regulator

Figure 8.14 Excitation system control and protective circuits

and reactive power excursions in the event the ac regulator is removed from service
abruptly. Care must be taken to ensure that a trip of the unit operating on manual
control <loes not leave the generator in an overexcited condition.

8.5.2 Excitation System Stabilizing Circuits

Excitation systems comprised of elements with significant time delays have

poor inherent dynamic performance. This is particularly true of de and ac type
excitation systems. Unless a very low steady-state regulator gain is used, the
excitation control (through feedback of generator stator voltage) is unstable when the
generator is on open circuit. Therefore, excitation control system stabilization,
comprising either series or feedback compensation, is used to improve the dynamic
performance of the control system. The most commonly used form of compensation
is a derivative feedback as shown in Figure 8.15. The effect of the compensation is
 sec. 8.5 Control and Protective Functions 335
minímize the phase shift introduced by the time delays over a selected frequency
~~ge [3]. This ~esults in a stab~e ~ff-line perfon:iance of the ?en~rator, such as that
xisting just pnor to synchronization or followmg a load rejection. The feedback
earameters can also be adjusted to improve the on-line performance of the generating
~it. Depending on the type of excitation system, there may be many levels of
xcitation control system stabilization involving the major outer loop and minor inner
~oops. Static excitation systems have negligible inherent time delays and do not
require excitation control-system stabilization to ensure stable operation with the
generatoroff-line (see closure of reference 21).

~ + Exciter E¡d To generator

L and AVR field
-

Compensation
sKF
I+sTF

Figure 8.15 Derivative feedback excitation control system stabilization

8.5.3 Power System Stabilizer (PSS)

The power system stabilizer uses auxiliary stabilizing signals to control the
excitation system so as to improve power system dynamic performance. Commonly
used input signals to the power system stabilizer are shaft speed, terminal frequency
and power. Power system dynamic performance is improved by the damping of
system oscillations. This is a very effective method of enhancing small-signal stability
performance.
The principie of operation of power system stabilizers and their structure and
tuning are discussed in detail in Chapters 12 and 1 7.

8.5.4 Load Compensation

The automatic voltage regulator (AVR) normally controls the generator stator
terminal voltage. Sometimes, load compensation is used to control a voltage which
is representative of the voltage at a point either within or external to the generator.
This is achieved by building additional circuitry into the AVR loop as shown in
Figure 8.16. The compensator has adjustable resistance (Re) and inductive reactance
(Xc) that simulate the impedance between the generator terminals and the point at
which the voltage is being effectively controlled. Using this impedance and the
measured armature current, a voltage drop is computed and added to or subtracted
336 Excitation Systems~ Chap. 8

Generator Step-up
Field
Annaru~t---+-~-t---~--- .......... ----tr-~~sfu~~~er----------~j~H-T_b_u_s ...._

º ~
PT
CT '~

Load compensator
Topo~
system

Voltage
Exciter
regulator

Figure 8.16 Schematic diagram of a load compensator

from the terminal voltage. The magnitude of the resulting compensated voltage (Ve),
which is fed to the AVR, is given by

(8.1)

With Re and Xc positive in Equation 8 .1, the voltage drop across the compensator is
added to the terminal voltage. The compensator regulates the voltage at a point within
the generator and thus provides voltage droop. This is used to ensure proper sharing
of reactive power between generators bussed together at their terminals, sharing a
common step-up transformer. Such an arrangement is commonly used with hydro
electric generating units and cross-compound thermal units. The compensator
functions as a reactive-current compensator by creating an artificial coupling between
the generators. Without this provision, one of the generators would try to control the
terminal voltage slightly higher than the other; hence, one generator would tend to
supply all of the required reactive power while the other would absorb reactive power
to the extent allowed by underexcited limits.
With Re and Xc negative, the compensator regulates the voltage at a point
beyond the machine terminals. This form of compensation is used to compensate for
the voltage drop across the step-up transformer, when two or more units are connected
through individual transformers. Typically, 50% to 80% ofthe transformer impedance
is compensated, ensuring voltage droop at the paralleling point so that generators can
operate in parallel satisfactorily. This device is commonly referred to as a line-drop
compensator although it is practically always used to compensate only for transformer
drop. The nomenclature appears to have been derived from a similar compensator
used on distribution system voltage regulators (see Chapter 11, Section 11.2).
 sec. 8.5 Control and Protective Functions 337

In rnost cases, the resistance cornponent of the impedance to be compensated

. negligible and Re may be set to zero.
15
A}ternative forms of reactive-current and line-drop compensators are described
in references 8 and 22.

s.S.5 Underexcitation Limiter [23-26]

The underexcitation limiter (UEL) is intended to prevent reduction of generator

xcitation to a level where the small-signal (steady-state) stability limit or the stator
~ore end-region heating limit (see Chapter 5, Figure 5.16) is exceeded. This limiter
is also referred to by other names such as underexcitation reactive-ampere limiter
(URAL) and mínimum excitation limiter (MEL ).
The control signal of the UEL is derived from a combination of either voltage
and current or active and reactive power of the generator. The limits are determined
by the signal exceeding a reference level. There are a wide variety of implementations
of the U.E:L function. Sorne UEL applications act on the voltage error signal of the
AVR; when the UEL set limit is reached, a nonlinear element (such as a diode) begins to
conduct and the limiter output signal is combined with other signals controlling the
excitation system. In a more widely used form of UEL application, the limiter output
signal is fed into an auctioneering circuit (high-value gate) which gives control to the
larger of the voltage regulator and UEL signals; when the UEL set limit is reached,
the limiter is given full control of the excitation system until the limiter signal is
below the set limit.
Methods of setting the UEL characteristics are described in references 23 to
25. The settings should be based on the needed protection, i.e., system instability or
stator core heating. In addition, the limiter performance should be coordinated with
the generator loss-of-excitation protection (see Chapter 13). Figure 8.17 indicates the
way in which the UEL characteristic (represented on a P-Q plane) is usually
coordinated with the calculated small-signal stability limit and the loss-of-excitation
(LOE) relay characteristic [28]. If the UEL is used to protect against stator end-region
heating, the coordination is done in a similar manner, except that the stability limit
is replaced by the heating limit.
If the input signals to the UEL are the generator stator voltage and current, the
limiting characteristic appears circular on a P-Q plane as shown in Figure 8.17. With
active and reactive power as input signals, the limiting characteristic would be a
straight line.
Care should be taken to ensure that the UEL performance during a transient
disturbance is not to the detriment of the power system performance [26,27].

8.5.6 Overexcitation Limiter

The purpose of the overexcitation limiter (OXL) is to protect the generator

from overheating due to prolonged field overcurrent. This Iimiter is also commonly
referred to as the maximum excitation limiter (MXL).
 338 Excitation Systems Chap. a
'"O
(l.)
.......
.......
u
X
(l.)

~
~
> o>
~ PinMW
...i..:.:. o
~
'"O
.E
......
u
X
(l.)
¡....;
Q)
'"O
~
:::i
Small-signal stability limit
Loss-of-excitation relay

Figure 8.17 Coordination between UEL, LOE relay and stability limit

The generator field winding is designed to operate continuously at a value

corresponding to rated load conditions. The permissible thermal overload of the field
winding of round rotor generators, as specified by ANSI Standard CS0.13-1977, is
given by the solid curve of Figure 8.18. The curve passes through the following
points:

Time (seconds) 10 30 60 120

Field voltage/current 208 146 125 112
(Percent of rated)

The actual implementation of overexcitation limiting function varies depending

on the manufacturer and vintage of the unit. Limiters supplied by two manufacturers
are described in references 28 and 29.
The overexcitation limiting function typically detects the high field current
condition and, after a time delay, acts through the ac regulator to ramp down the
excitation to a preset value (typically 100% to 110% of rated field current). If this is
unsuccessful, it trips the ac regulator, transfers control to the de regulator, and
repositions the setpoint to a value corresponding to the rated value. If this also <loes
not reduce the excitation to a safe value, the limiter will initiate an exciter field
breaker trip and a unit trip.
Two types of time delays are used: (a) fixed time and (b) inverse time. The
fixed time limiters operate when the field current exceeds the pickup value for a fixed
set time, irrespective of the degree of overexcitation. The inverse time limiters operate
with the time delay matching the field thermal capability, as shown in Figure 8.18.
 sec. 8.5 Control and Protective Functions 339

120
\
90
B\
]
tfl
\
o
<)
(1.)
tfl
\
...i.:.:.
(1.)
60 \
......s .\ Field thermal capability
~ ·~
-.
30
<.
Overexcitation limiting '-.......'-.......

o "--y;--_._~~_._~------~--'-~~_._~---........._

o.o 1.0 1.1 1.2 1.3 1.4 1.5

Generator field voltage per unit of rated value

Figure 8.18 Coordination of overexcitation limiting

with field thermal capability

Exciters with very high ceiling voltages may be provided with an additional
jield current limiter, which acts instantaneously through the ac regulator and limits the
field current to the short time limit (typically 160% of rated value).

8.5. 7 Volts-per-Hertz Limiter and Protection

These are used to protect the generator and step-up transformer from damage
due to excessive magnetic flux resulting from low frequency and/or overvoltage.
Excessive magnetic flux, if sustained, can cause serious overheating and may result
in damage to the unit transformer and to the generator core.
The ratio of per unit voltage to per unit frequency, referred to as volts per
hertz (V/Hz), is a readily measurable quantity that is proportional to magnetic flux.
Typical V/Hz limitations for generators (GEN) and step-up transformers (XFMR) are
shown in the following table.

V/Hz (pu) 1.25 1.2 1.15 1.10 1.05

Damage Time GEN 0.2 1.0 6.0 20.0 00

in Minutes XFMR 1.0 5.0 20.0 00

340 Excitation Systems Chap. 8
The unit step-up transformer low voltage rating is frequently 5% below the
generator voltage rating; therefore, V/Hz limiting and protection requirements are
usually determined by the transformer limitation. If, however, the generator and
transformer voltage ratings are the same, the generator limitation would be more
restrictive.
The V/Hz limiter (or regulator, as it is sometimes called) controls the field
voltage so as to limit the generator voltage when the V/Hz value exceeds a preset
value.
The V/Hz protection trips the generator, when the V!Hz value exceeds a preset
value for a specified time. Usually, a dual-level protection is provided, one with a
higher V!Hz setting and a shorter time setting, and the other with a lower V/Hz
setting and a longer time setting. When used in conjunction with a V/Hz Iimiter, it
serves as a backup.
For many units, the V/Hz protection becomes overvoltage protection above
60 Hz.

8.5.8 Field-Shorting Circuits

Since rectifiers cannot conduct in the reverse direction, the exciter curreni
cannot be negative in the case of ac and static exciters. Under conditions of pole
slipping and system short circuits, the induced current in the generator field winding
may be negative. If a path is not provided for this negative current to flow, very high
voltages may result across the field circuit. Therefore, special circuitry is usually
provided to bypass the exciter to allow negative field current to flow. This takes the
form of either a field-shorting circuit, commonly referred to as crowbar, ora varistor
[8,30].
A crowbar consists of a thyristor and a field discharge resistor (FDR)
connected across the generator field as shown in Figure 8.19. The thyristor is gated
in response to an overvoltage condition that is created by the induced current not
initially having a path in which to flow. The thyristor so gated conducts induced field
current through the field discharge resistor.

Generator

Field
From exciter
FDR

Figure 8.19 Field bypass circuit using a crowbar

 sec. 8.B Modelling of Excitation Systems 341

A varistor is a nonlinear resistor. When connected across the generator field

winding, as shown in Figure 8.20, it provides an effective means of bypassing the
citer under conditions of high induced voltage. With normal exciter voltage across
:e varistor, it has very high resistance and hence carries negligible current. As the
oltage across the varistor increases beyond a threshold value, its resistance decreases
~d current through it increases very rapidly. Thus the varistor provides a low
resistancepath to induced negative field current and limits the voltage across the field
and exciter.

Generator

Field
Varistor Armature

Figure 8.20 Field bypass circuit using a varistor

In sorne cases, no special field shorting circuits are provided. The amortisseurs
associated with the solid rotor iron provide paths for the induced rotor currents. This
is sufficient to limit the induced voltage to a level that is below the withstand
capabilities of the generator field and the exciter. Since the field is not shorted, it
carries no current in the negative direction.

8.6 MODELLING OF EXCITATION SYSTEMS

Mathematical models of excitation systems are essential for the assessment of

desired performance requirements, for the design and coordination of supplementary
control and protective circuits, and for system stability studies related to the planning
and operation of power systems. The detail of the model required depends on the
purpose of the study. Referring to Figure 8.14, the control and protective features that
impact on transient and small-signal stability studies are the voltage regulator, power
system stabilizer, and excitation control stabilization. The limiter and protective
circuits identified in the figure normally need to be considered only for mid-term,
long-term, and voltage stability studies. Sorne excitation systems are provided with
fast-acting terminal voltage limiters in conjunction with power system stabilizers;
these have to be modelled in transient stability simulations.
In this section, modelling of excitation systems is described. We begin with
consideration of an appropriate per unit system, then describe models for the various
components, and finally present complete models for selected types of excitation
systems.
 342 Excitation Systems Chap. a
The material presented in this section conforms to the IEEE committee reportg
on excitation system modelling. The IEEE work on standardization of these models
began in the 1960s, and the first set of models was published in 1968 [1]. This work
was extended, and improved models to reflect advances in equipment and better
modelling practices were published in 1981 [4]. These were updated and refined in
1992 [8].

8. 6. 1 Per Unit System

In choosing the per unit system for exciter output voltage and current, there
are severa! options.
First, the per unit system used for the main synchronous machine field circuí¡
would appear to be the obvious choice. While this system was chosen to simplify the
synchronous machine equations (see Chapter 3, Section 3.4), it is not considered
suitable for expressing the exciter output quantities. This is because, for normal
operating conditions, the per unit exciter output voltage would be very small, being
on the arder of 0.001.
Second, for excitation system specification purposes it has become standard
practice to use the rated-load field voltage as 1.0 per unit. However, this is not
convenient for use in the formulation of synchronous machine and excitation system
equations for system studies.
The third choice is to have 1.0 per unit exciter output voltage equal to the field
voltage required to produce rated synchronous machine armature terminal voltage on
the air-gap line; 1.0 per unit exciter output current is the corresponding synchronous
machine field current. This per unit system is universally used in power system
stability studies as it offers considerable simplicity. Here, we refer to this system as
the non-reciproca! per unit system to distinguish it from the reciproca! per unit system
used for modelling synchronous machines.
Excitation system models must interface with the synchronous machine model
at both the field terminals and armature terminals. The input control signals to the
excitation system are the synchronous machine stator quantities and rotor speed. The
per unit systems used for expressing these input variables are the same as those used
for modelling the synchronous machine. Thus, a change of per unit system is required
only for those related to the field circuit.
We now develop the relationship between the per unit values of the exciter
output voltage/current expressed in the non-reciproca! system and the synchronous
machine field voltage/current expressed in the Laábase reciproca! system (see Section
3.4). Forthe synchronous machine, under open-circuit conditions, id=iq=O. Substituting
in Equations 3.139, 3.140, 3.142 and 3.143 yields

(8.2)

(8.3)
 sec. 8.6 Modelling of Excitation Systems 343

Air-gap line, slope = Ladu

occ

l/Ladu

1.0

Figure 8.21 Synchronous machine open circuit characteristics

Referring to Figure 8.21, the field current required to produce 1.0 per unit
stator terminal voltage on the air-gap line (slope = Ladu) is determined by

(8.4)

Therefore, in the reciproca! per unit system, the field current i¡d required to generate
rated stator terminal voltage on the air-gap line is given by

The corresponding field voltage is

=
R~d
J' pu
Ladu

By definition, the corresponding value of exciter output current l¡dis equal to 1.0 per
unit. Therefore,
 344 Excitation Systems Chap. 8

(8.5)

and the corresponding exciter output voltage is

(8.6)

Physically, exciter output voltage/current and generator field voltage/current are the
same; distinction is made only in their per unit values to allow independent selection
of the per unit systems for modelling excitation systems and synchronous machines.
This is illustrated in Figure 8.22.

Exciter model PU conversion Sync. machin

model
N on-reciprocal efd = (Rfd/Laa)Efd
Reciproca!
pu system ifd = l¡af Laau pu system

Figure 8.22 Per unit conversion at the interface between excitation

system and synchronous machine field circuit

Under steady-state conditions, the per unit values of E¡d and l¡d are equal.
During a transient condition, however, E¡d and l¡d differ; E¡d is determined by the
excitation system and l¡d is determined by the dynamics of the field circuits.
A few interesting observations regarding the above per unit conversion are
appropriate at this time:

l. The factor Lad)R¡d in Equation 8.6 is the steady-state value of G(s) (see
Chapter 4, Equations 4.14 and 4.17), in the absence of saturation.

2. Equation 4.24 in terms of M¡d becomes

1 +sTkd
~ lJr d(s) = ~E
(1 +sT~)(l +sT;,) fd (8.7A)
= G(s)~Efd

-
The steady-state value of G(s) is equal to 1.0. There is thus a one to one
relationship between ~d and E¡d·
sec- 8.6 Modelling of Excitation Systems 345
With amortisseurs neglected, Equation 8. 7A becomes
1
A lJr j.s) = áE1d (8.7B)
1 +sT/Jo

3. From the steady equations developed in Section 3.6.3, with the generator on
open circuit, we have
A. 1'il'rd = A.e q = .6.Et

Substitution in Equation 8. 7B yields the following open-circuit transfer

function of the generator:

1
= (8.7C)
l+sT~

Example 8.1

The following are the parameters in per unit on machine rating of the 555 MVA, 0.9
p.f., 24 kV turbine generator considered in Examples 3 .1, 3 .2 and 3 .3 of Chapter 3:
Ladu = 1.66 Laqu = L61 L1=0.15
L¡d = 0.165 R¡d = 0.0006 Rª = 0.003

(a) The field current required to generate rated stator voltage E1 on the air-gap
line is 1300 A and the corresponding field voltage is 92.95 V. Determine the
base values of E¡d and l¡d in the non-reciproca! per unit system and the base
values of e¡d and i¡d in the reciproca! per unit system.

(b) Compute the per unit values of E¡d and I1d,when the generator is delivering
rated MVA at rated power factor and terminal voltage. Assume that the
corresponding values of the saturation factors Ksd and Ksq are equal to 0.835.

Solution

(a) By definítiori, the base values of E¡d and l¡d are respectively equal to the field
vo1tage and field current required to produce rated air-gap line voltage. Hence,

E¡d base = 92.95 V

f¡d base = 1300 A

From Equations 8.5 and 8.6, the base values of e¡d and i¡d are
1¡d base == L cui1.l¡d base
= l.66xl300 == 2158 A
346 Excitation Systems Chap, 8

e¡d base = (LadjR¡d)E¡d base

= (1.66/0.0006)92.95 = 257.2 kV

The above base values of e¡d and i¡d, as expected, agree with the values computed in
Example 3.1.

(b) From the results of Example 3.2, at the rated output conditions,

e¡d = 0.000939 pu
Í¡d = 1.565 pu

The corresponding per unit values of E¡d and l¡d are

E¡d = (LadjR¡d)e¡d
= (1.66/0.0006)0.000939 = 2.598 pu
l¡d =
=
L adu͡d
1.66x 1.565 = 2.598 pu •
Specification o/ temperature

The base exciter output voltage depends on the synchronous machine field
resistance, which in turn depends on the field temperature. The standard temperatures
used for calculating the base exciter output voltage are 1 OOºC for thermal units
(operating temperature rise greater than 60ºC) and 75ºC for hydraulic units (operating
temperature rise 60ºC or less) [5]. However, care should be exercised in using these
temperatures for modelling of excitation systems.
The value of the field resistance used should correspond to the resistance under
the actual operating conditions being simulated, as closely as possible. The value of
TJ0 should be consistent with this value of field resistance.
The field resistance corrected to a specified operating temperature may be
calculated as follows [31]:
t 8-
R=R- +k) (8.8)
s t ( tt +k

where
ts = specified operating temperature, ºC

11 = temperature corresponding to known or measured value of winding

resistance, ºC
R, = winding resistance at temperature ts
R1 = winding resistance at temperature t1

The characteristic constant k depends on winding material. It is equal to 234.5

for pure copper and 225 for aluminium based on a volume conductivity of 62% pure
copper [31].
 sec. 8.6 Modelling of Excitation Systems 347

E:xaJJlple 8.2

The per unit R¡d of a steam-turbine-driven generator is 0.00063 at 75ºC and the
corresponding value of T~ is 7.6125 s. The base exciter output voltage at the standard
temperature of lOOºC is 105.575 V. If the generator is operating ata temperature of
60ºC, find consistent values of R¡d, T~' and base E¡d at this temperature. Assume that
the constant k of the field winding material is 234.5.

Solution

The field resistance at 60ºC is

R = 0.00063 234.5 +60

'fd60 234.5+75
= 0.0006 pu

The time constant T~ is inversely proportional to R¡d· Hence at 60ºC it is

T' = T' R¡a1s

dO dO R¡a60

= 7.6125x 0.00063 = 8.0 s

0.0006

The base E¡d is directly proportional to R¡d· Hence at 60ºC,

234.5+60
Base E¡d60 = base E¡d roo ( 234.5+100
l
= 105.575x0.8804 = 92.95 V
•
8.6.2 Modelling of Excitation System Components

The basic elements which form different types of excitation systems are the
de exciters (self or · separately excited); ac exciters; rectifiers (controlled or non-
controlled); magnetic, rotating, or electronic amplifiers; excitation system stabilizing
feedback circuits; signal sensing and processing circuits. We describe here models for
these individual elements. In the next section we will consider modelling of complete
· excitation systems.

(a) Separately excited de exciter

The circuit model of the exciter is shown in Figure 8.23.

348 Excitation Systems _ Chap, 8

Re¡ Field Armature

+ +

1eJ

Eef i; Ex

Figure 8.23 Separately excited de exciter

For the exciter field circuit, we write

E = R-' + dq¡
e/ ere/ dt (8.9)
with

Neglecting field leakage, the exciter output voltage Ex is given by

(8.1 O)

where Kx depends on the speed and winding configuration of the exciter armature.
The output voltage Ex is a nonlinear function of the exciter field current le¡ due
to magnetic saturation. The voltage Ex is also affected by the load on the exciter. The
common practice [1,4] in de exciter modelling is to account for saturation and load
regulation approximately by combining the two effects and using the constant-
resistance load-saturation curve, as shown in Figure 8.24.
The air-gap line is tangent to the lower linear portian of the open circuit
saturation curve. Let Rg be the slope of the air-gap line and Me¡ denote . the departure
of the load saturation curve from the air-gap line. From Figure 8.24, we write
Ex (8.11)
I ef = -+d.I
R ef
g

where Me¡ is a nonlinear function of Ex and may be expressed as

(8.12)

where SeCEx) is the saturation function dependent on Ex.

 sec. 8.6 Modelling of Excitation Systems 349

Air-gap line
»->: Open circuit curve
,..,... ,..,,../ -- Constant-resistance

/",..... load-saturation curve

Exo ------------¿p:~---- ! <;
: l
f
Operating point
r

- 1 6.1e ¡ i1-
1 1
1 1
1 1
1 1
r 1

Figure 8.24 Exciter load-saturation curve

From Equations 8.9 to 8.12, we have

(8.13)

The above equation gives the relationship between the output Ex and the input voltage
Eef A convenient per unit system for this equation is one with base values of Ex and
le¡ chosen to be equal to those values required to give rated synchronous rnachine
voltage on the air-gap line. Thus,

EX base = Efdbase

/e/base = Efdbasef Rg
(8.14)

Rgbase =R g

Dividing Equation 8 .13 by EXbase» we have

--'-E
ef_ = _R!f_ EX +R S (E ) EX +--1 d( EX l
EXbase Rg EXbase ef e x E Xbase K X dt EXbase
350 Excitation Systems Chap.8
In per unit f orm, we have

-
- Re¡- - - 1 dEx
E =-E [l+S (E)]+--- (8.15)
ef R X e X K dt
g X

In the above equation, Se(Ex) is the per unit saturation function defined as follow~:

(8.16)

From Figure 8.25, with Ex and le¡ expressed in per unit, the per unit saturation
function is given by [4]
A-B (8.17)
B

The parameter Kx defined by Equation 8.10 may be written as

-
Ex R Ex
K = - = ---
g
X 1jT L J-
e/ e/

Constant-resistance
load-saturation curve

Figure 8.25 Exciter saturation characteristic

sec. 8.6 Modelling of Excitation Systems 351

corresponding to any given operating point UefO' Exo), let

-
I
L = L efD (8.18)
'fu ef -
Exo
Therefore, Kx=Rg/Lfu. Substituting in Equation 8.15, we get

(8.19)

where

(8.20)

Equation 8.19 represents the input-output relationship of the exciter. For a separately
excited exciter, the input voltage Eef is the regulator output VR. The output voltage Ex
of a de exciter is directly applied to the field of the synchronous machine. Therefore,
the exciter may be represented in block diagram form as shown in Figure 8.26. In the
diagram, ali variables are in per unit; however, the superbar notation denoting this
has been dropped.

--------- Ex= EFD

VX

Figure 8.26 Block diagram of a de exciter

352 Excitation Systems Chap. 8
The adjustment of fíeld-circuit resistanee Re¡ affeets KE as well as the
saturation function SE(Ex), but not the integration time TE of the forward loop.
There are severa! convenient mathematical expressions that may be used to
approximate the effect of exciter saturation. A commonly used expression is the
exponential function

(8.21)

The block diagram of Figure 8.26 provides a convenient means of representing the
de exeiter in stability studies. However, the effective gain and time eonstant of the
exciter are not readily apparent from it. These are more evident when the block
diagram is reduced to the standard form by considering small-signal response:

For any operating point with Ex=EFD;;;:,EFDO' the effective gain K and time
constant of the exciter for small perturbations are
1
K=----- (8.22)
BEx SE(Epoo) +KE

T =
TE (8.23)

where SE (EXO ) =A EX e B ExEroo

(h) Self-excited de exciter

Figure 8.27 shows a eircuit model of a self-excited de exeiter.

Figure 8.27 Self-excited de exciter

 sec. 8.6 Modelling of Excitation Systems 353

For this exciter, the regulator output VR is in series with the exciter shunt field.
Tberefore, the per unit voltage across the exciter field (without the explicit per unit
notation) is

(8.24)

jhe relationship between the per unit values of Ee¡ and Ex developed for the
separately exci.ted exci:er also applies in this case. Substituting for Ee¡ given by
Equation 8.24 m Equation 8.15, we have
R 1 dE
= _!f_E [l +S(E )] +---x
R X e X K dt
g X
This reduces to

(8.25)

where

TE = (8.26)

R
SE= se (E)~
X R
g

The block diagram of Figure 8.26 also applies to the self-excited de exciter. The value
of KE, however, is now equal to Re/Rg-1 as compared to Re/Rg for the separately
excited case.
The station operators usually track the voltage regulator by periodically
adjusting the rheostat setpoint so as to make the voltage regulator output zero. This
is accounted for by selecting the value of KE so that the initial value of VR is equal
to zero. The parameter KE is therefore not fixed, but varies with the operating
condition.

(e) AC exciters and rectifiers

The ac exciter representation (excluding rectification) recommended m

reference 8 for use in large-scale stability studies is shown in Figure 8.28.
354 Excitation Svstems Chap. 8

+ 1
[ - .

~ sTE

vx = VE·SE(VE)

+
vx
E + KE
+

KD IFD

Figure 8.28 Block diagram of an ac exciter

The general structure of the model is similar to that of the de exciter. However, in
this case the load regulation due to the armature reaction effect is accounted for
distinctly, and the no-load saturation curve is used to define the saturation function
SE. The exciter internal voltage VE is the no-load voltage as determined by the
saturation function. The main generator field current I FD represents the exciter load
current, and the negative feedback of KDIFD accounts for the armature reaction
demagnetizing effect. The constant KD depends on the ac exciter synchronous and
transient reactances [32]. Figure 8.29 illustrates the calculation of the saturation
function SE for a specified value of VE.
The per unit saturation function is
S (V) = A-B (8.27)
E E B

Any convenient mathematical expression can be used to represent the saturation

function. As in the case of de exciters, a commonly used expression for Vx=VESe(VE)
is the exponential function given by Equation 8.21.
Three-phase full-wave bridge rectifier circuits are commonly used to rectify
the ac exciter output voltage. The effective ac source impedance seen by the rectifier
is predominantly an inductive reactance. As described in Chapter 1 O, the effect of this
inductive reactance (referred to as the commutating reactancei is to delay the process of
commutation, i.e., transfer of current from one valve to another. This produces a
 sec. 8.B Modelling of Excitation Systems 355

No-load saturation curve

pu exciter field current

Figure 8.29 AC exciter saturation characteristic

decrease in the average output voltage of the rectifier as its load current increases.
Reference 33 shows that a three-phase full-wave bridge rectifier circuit operates in
one of three distinct modes as the rectifier load current varies from no load to the
short circuit level. The mode of operation depends on the commutating voltage drop
(equal to the product of commutating reactance and load current).
The equations defining the rectifier regulation as a function of commutation
voltage drop may be expressed as follows [8,34]:

EFD = FEXVE (8.28)

where
FEX = f(IN) (8.29)
and

IN =
KCIFD (8.30)
VE

The constant Kc depends on the commutating reactance. The expressions for the
functionf(JN) characterizing the three modes of rectifier circuit operation are
356 Excitation Systems Chap. 8

Mode 1: {(IN) = 1.0 -0.577JN, if INs0.433

Mode 2: {(IN) = ¡ 0.75 -!}¡, if 0.433<[N<0.75 (8.31)

Mode 3: {(IN) = 1.732(1.0-/N), if 0.75s!Nsl.O

Now IN should not be greater than 1.0, but if for sorne reason it is, FEx should be set
to zero [8].
The rectifier regulation effects identified above may be depicted in the block
diagram form as shown in Figure 8.30.

Figure 8.30 Rectifier regulation model

Referring to Figures 8.28 and 8.29, the exciter output voltage Epn is simulated as
the ac exciter interna! voltage VE reduced by the armature reaction (Ipnf(n) and
rectifier regulation (FEx).

(d) Amplifiers

Amplifiers may be the magnetic, rotating, or electronic type. Magnetic and

electronic amplifiers are characterized by a gain and may also include a time constant.
As such they may be represented by the block diagram of Figure 8 .31.

VRMAX

~---
Figure 8.31 Amplifier model

The amplifier output is limited by saturation or power-supply limitations; this

is represented by "non-windup" limits VRMAX and VRMJN in Figure 8.31. A description
of such limits is provided later in this section.
sec. 8.6 Modelling of Excitation Systems 357

The output limits of sorne amplifiers having power supplies from generator or
uxiliarY bus voltage vary with generator terminal voltage. In such cases, VRMAX and
~ vary directly with generator terminal voltage s,
RMIN The transfer function of an amplidyne is derived in reference 12 and has the
general form shown in Figure 8.32.

~ ---· I (1 +sT,~1 +sT2)

Figure 8.32 Amplidyne model

(e) Excitation system stabilizing circuit

There are severa! ways of physically realizing the stabilizing function identified
in Figure 8.15. Sorne excitation systems use series transformers as shown in Figure
S.33.
The transformer equations in Laplace notation are

(8.32)

where subscripts 1 and 2 denote primary and secondary quantities; R, L, and M denote
resistance, leakage inductance and mutual inductance, respectively.

Exciter
Field Armature
~-~-----~+

)Ex

Amplifier _ Stabilizing
---¡- transformer
+ 2

+
~--,o+

verr
~--~-----------0-

Figure 8.33 Excitation system stabilizing transformer

358 Excitation Systems Chap, 8_

The secondary of the transformer is connected to a high impedance c1rcuit

Therefore, neglecting i2, we have

(8.33)

Thus

-----

(8.34)

(/) Windup and non-windup limits

In the modelling of excitation systems, it is necessary to distinguish between

windup and non-windup limits. Such limits are encountered with integrator blocks,
single time constant blocks, and lead-lag blocks.
Figures 8.34(a) and (b) show the differences between the two types of limits
when applied to an integrator block.

1
Representation: U---i
s

dv
S ystem equation: = u
dt
Limiting action:
If LN<v<Lx, then y = v
If v ~ Lx, then y = Lx
If v:::;,LN, then y = LN

Figure 8.34 (a) Integrator with windup limits

sec. 8.6 Modellinq of Excitation Systems 359

1
Representation: u y
s
LN

system equation: dy
= u
dt
Limítingaction:

If LN<y<Lx, then dy = u
dt
If y z Lx and dy >O, then set dy =o y= Lx
dt dt '
dy dy
If y~LN and -<O, then set = o, y= LN
dt dt

Figure 8.34 (b) Integrator with non-windup limits

With windup limits the variable v is not limited. Therefore, the output variable y
cannot come off a limit until v comes within the limit. With non-windup limits, the
output variable y is Iimited; it comes off the limit as soon as the input u changes sign.
Figures 8.35(a) and (b) show the difference between the two types of lirnits
when applied to a single time constant block. The significance of the two types of
limits is similar to that for an integrator. With a .windup limit, the output y cannot
come off a limit until v comes within the limit. With a non-windup limit, however,
the output y comes off the limit as soon as the input u re-enters the range within
limits.

1
Representation: U---
l+sT

dv u-v
System equation: =
dt T

Limitingaction:
If LN<v<Lx, then y = v
If v '2:. Lx, then y = L:x
If v~LN, then y = LN

Figure 8.35 (a) Single time constant block with windup limits
360 Excitation Systems Chap. 8

1
Representation: U---i i---- y
l+sT

u-y
System equation: f=
T
Limiting action:

If LN<y<Lx, then dy = f
dt
If y c.Lx and f>O, then set dy = O, y = Lx
di.
If y -:;,LN and f <O, then set d; = O, y = LN

Figure 8.35 (b) Single time constant block with non-windup limits

LX
l+sTA
Representation: u y
l+sT8

LN TA<TB

Physical realization:
Lx

+ TA
u y
TB

Figure 8.36 Lead-lag function with non-windup limits

(Continued on next page)
 sec. 8.6 Modelling of Excitation Systems 361

Lirniting action:
If LNs,.v ::;.Lx, then y = v

If v>Lx, then y = i;
If v<LN , then y = LN

Figure 8.36 (Continued) Lead-lag function with non-windup limits

With a lead-lag block, the interpretation of the action of a windup limit is

straightforward and is similar to that of a single time constant block. However, the
way in which a non-windup limit can be realized is not unique; the interpretation of
the limiting action should therefore be based on the physical device represented by
the block. Figure 8.36 illustrates such limiting action associated with electronic
iroplementation of lead-lag functions.

(g) Gating functions

Gating or auctioneering circuits are used when it is required to give control to

one of two input signals, depending on their relative size with respect to each other.
Figure 8.37 illustrates the functions of a low value (LV) gate and a high value (HV)
gate, and the syrnbols used to represent them in block diagrams.

u LV y u HV y
V gate V gate

If u s v, y=u If u z v, y=u
If u>v, y=v If u<v, y=v

Figure 8.37 Low- and high-value gating functions

(h) Terminal voltage transducer and load compensator {8]

The block diagram representation of these elements is shown in Figure 8.38.

The time constant TR represents rectification and filtering of the synchronous machine
terminal voltage. The parameters of the load compensator (described in Section 8.5.4)
are Re and Xc. The input variables Et and 11 are in phasor forro. \Vhen load
compensation is not used, Re and Xc are set to zero.
The voltage transducer output Ve forms the principal control signal to the
excitation system. If a load compensator is not used and TR is negligible, Vc=E1•
362 Excitation Systems Chap, 8

Et--•
It ,
......_
_ °V"Cl 1

.
1
l+TR
1

Load compensator Voltage transducer

Figure 8.38 Terminal voltage transducer and load compensator model

8.6.3 Modelling of Complete Excitation Systems

Figure 8.39 depicts the general structure of a detailed excitation system model
having a one-to-one correspondence with the physical equipment. While this niodel
structure has the advantage of retaining a direct relationship between model
parameters and physical parameters, such detail is considered too great for general
system studies. Therefore, model reduction techniques are used to simplify and obtain a
practica} model appropriate for the type of study for which it is intended.
The parameters of the reduced model are selected such that the gain and phase
characteristics of the reduced model match those of the detailed model over the
frequency range of O to 3 Hz. In addition, all significant nonlinearities that impact on
system stability are accounted for. With a reduced model, however, direct
correspondence between the model parameters and the actual system parameters is
generally lost.
The appropriate structure for the reduced model depends on the type of
excitation system. The IEEE has standardized 12 model structures in block diagram
form for representing the wide variety of excitation systems currently in use [8].,

Amplifying stages Exciter

Other
signals

' Minor loo/p

stabilization

Major loop stabilization

Figure 8.39 Structure of a detailed excitation system model

 sec. 8.6 Modelling of Excitation Systems 363

These are intended for use in transient stability and small-signal stability studies. For
urposes of illustration, we will consider five of these models. These are shown in
~·gures 8.40 to 8.44, with slight modification of the block diagram conventions to
~nform to the recommendations made in reference 35. The figures include a brief
º!
~escription the key features an~ samp~e.data. The suffix "A" acco~panying the
designationsis for the purpose of differentiating these models developed m 1992 from
5
irnilar models developed previously in 1981 [4].
The principal input signal to each of the excitation systems is the output Ve
ofthe voltage transducer shown in Figure 8.38. At the first summing point, the signal
V. is subtracted from the voltage regulator reference Vref and the output V8 of the
p¿wer system stabilizer, if used, is added to produce the actuating signal which
controls the excitation system. Additional signals, such as the underexcitation Iimiter
output (VUEL), come into play only during extreme or unusual conditions. Under
steady state, V8=0 and Vref takes on a value unique to the synchronous machine
loading condition so that the error signal results in the required field voltage E¡d· This
is illustrated in Example 8.3.

J. Type DCJA exciter model

Vs VUEL v;EL VRMAX

l+sTc HV
Ve l+sTB gate

vref
+

* Altemate input point

Figure 8.40 IEEE type DClA excitation system model. ©IEEE 1992 [8]

The type DClA exciter model represents field-controlled de commutator

exciters, with continuously acting voltage regulators. The exciter may be separately
excited or self-excited, the latter type being more common. When self-excíted, KE is
selected so that initially VR =0, representing operator action of tracking the voltage
regulator by periodically trimming the shunt field rheostat setpoint.
364 Excitation Systems Chap, 8
Sample data

Self-excited de exciter:

KA=181 TA=0.89 TE=l.15 AEx=0.014 BEx=l.55

KF=0.058 TF=0.62 TB=0.06 Tc=0.173 TR=0.05
VRMAX=l.7 VRMJN=-1.7
KE is computed so that initially VR=O, and the load compensator is not usect.

2. Type A CJA exciter model

I+sTc
l+sT8
o
~MIN

sKp
l+sTp

Figure 8.41 IEEE type AClA excitation system model. © IEEE 1992 [8]

The type ACIA exciter model represents a field-controlled altemator excitation

system with non-controlled rectifiers, and is applicable to brushless excitation systems.
The diode rectifier characteristic imposes a lower limit of zero on the exciter output
voltage. The exciter field is supplied by a pilot exciter, and the voltage regulator
power supply is not affected by externa! transients.

Sample data

Exciter and regulator:

KA=400.0 Te =0
TF=l.0 Kn=0.38

VRMAx=7.3 VRMIN=-6.6 VAMAx=lS.0 VAMIN=-15.0 AEx=0.1

BEX=0.03
The load compensator is not used, and TR is negligible.
 sec. 8.6 Modelling of Excitation Systems 365

J. Type A C4A exciter model

+
l+sTc HV
l+sTs gate

vref

Figure 8.42 IEEE type AC4A excitation system model. © IEEE 1992 [8]

The type AC4A exciter model represents an alternator-supplied controlled-

rectifier excitation system - a high initial-response excitation system utilizing a full-
wave thyristor bridge circuit. Excitation system stabilization is usually provided in the
form of a series lag-lead network (transient gain reduction). The time constant
associatedwith the regulator and firing of thyristors is represented by TA. The overall
gain is represented by KA. The rectifier operation is confined to mode 1 region.
Rectifierregulation effects on exciter output limits are accounted for by constant Kc.

Sample data

Exciter and regulator:

Tc=l.0 TB=l2.0 VRMAx=5.64
VIMAX=l.O VIMIN=-l.O

4. Type STJA exciter model

The type STIA exciter model represents a potential-source controlled-rectifier

system. The excitatíon power is supplied through a transformer from generator
terminals;therefore, the exciter ceiling voltage is directly proportional to the generator
terminalvoltage. The effect of rectifier regulation on ceiling voltage is represented by
Kc. The model provides flexibility to represent series lag-lead or rate feedback
stabilization. Because of the very high field-forcing capability of the system, a field-
. currentlimiter is sometimes employed; the limit is defined by ILR and the gain by KLR·
366 Excitation Systems Chap, 8

Vs Vu*EL Vs* Vu*EL

~MAX
VuEL V¡MAX (Et VRMAX- KcfFn)

HV (1 +sTJ(l +sTc1) KA HV LV
gate (1 +sT~(l +sTBI) l+sTA gate gate

V,.e¡ V¡MIN EtVRMIN

~MTN

VaEL
V¡,.
sKF
l+sTF

* Altemate input points

o

Figure 8.43 IEEE type STlA excitation system model. © IEEE 1992 [8]

Sample data

Exciter and regulator:

KA=200.0 TA =O VRMAX=7.0

KLR=4.54 ILR=4.4 TB, Te, TB1, Tc1, KF, TF not used

VIMAx, VIMIN, VAMAX, VAMIN are not represented.

Voltage transducer and load compensator:

R e =O X e =O

5. Type ST2A exciter model

The type ST2A exciter model represents a compound-source rectifier excitation

system. The exciter power source is formed by phasor combination of main generator
armature voltage and current. The regulator controls the exciter output through
controlled saturation of the power transfonner. The parameter TE represents the
integration rate associated with the control windings; EFDMAX represents the limit on
exciter output due to magnetic saturation.

Sample data
KA=120.0 TA =0.15 KE=l.0 TE=0.5 Kc=0.65
KF=0.02 TF=0.56 VRMAx=l.2 VRMIN=-1.2 EFDMAX=3.55
Kp=l.19 K1=1.62
sec. 8.6 Modelling of Excitation Systems 367

Vs *
V uEL
VRMAN
VuEL

HV KA
Ve gate l+s7A
Vs
V,.ej VF o
VRMIN

sKp
l+sTF

-
Vr---
+
~ "E =I x,Vr+i K1lr1
Ir---~---~-----'

* Alternate input point

Figure 8.44 IEEE type ST2A excitation system model. e IEEE 1992 [8]

Example 8.3

A generator is operating under steady-state conditions with an E¡d of 2.598 pu and

E,=1.0 pu.

(a) If it is equipped with a type AC4A excítation system represented by the block
diagram of Figure 8.42, determine the value of vref"

(b) If it is equipped with a self-excíted de excitation system represented by the

block diagram of Figure 8.40, determine the values of KE and vref

The parameters of the excitation systems are the same as for the sample data provided
with the figures.

Solution

(a) Type AC4A excitation system of Figure 8.42

When E¡d is 2.598 pu, we have

 368 Excitation Systems Chap. 8

VR=E¡iKA

2·598
= =0.013 pu
200.0

Since a load compensator is not used,

Under steady-state operation, V1= VR and V8=0. Since the generator is operating under
normal conditions, VUEL =O. Therefore, from Figure 8.42 we see that

Vre/ = Ve+ V1

= 1.0 +0.013 = 1.013 pu

(b) Self-excited de exciter of Figure 8. 40

In this case, KE takes a value such that VR =0. With E¡dat a steady-state value and
VR=O, VFE=O. Hence,

KEEfd = -Vx

= -AEXeBEXE/d = -0.0l4e l.55x2.598

= -0.7852

Therefore,

K = _ 0.7852
E 2.598
= -0.3022

Under steady-state normal operation, VF=O and V5=0. With VR=O,

•
Modelling of limiters

The standard models shown in Figures 8.40 to 8.44 do not include

representation of limiting circuits, namely, the underexcitation limiter, V/Hz limiter,
and maximum excitation limiter. These circuits do not come into play under normal
conditions and are not usually modelled in transient and small-signal stability studies.
They may, however, be important for long-term stability and voltage stability studies.
sec. 8.6 Modelling of Excitati.on Systems 369

The actual implementation of these limiting functions varies widely depending

n the manufacturer, the vintage of the equipment, and the requirements specified by
~e utility. Therefore, models for these circuits have to be established on a case by
case basis. Here we will illustrate how such devices are modelled by considering
specific examples.

(a} Underexcitation limiter:

Figure 8.45 shows the model of an underexcitation limiter used in conjunction

with a type STlA (static) exciter. The parameters Kc and KR determine the
characteristics of the limiter on the P-Q plane. The output VUEL of the limiter is
appliedto the HV gate of the exciter model of Figure 8.43.

VKMX

K¡ VuMX
1 +sT1 Vu
KP VUEL
(1 +sT2) (l +sT3)
1
s VuMN

ILIM

Sample data:

Kp = 0.015 K1 = 10.0 T1 = 6.4 s T2 = 0.8 s

T3 = 0.64 s VKMX = 4.0 VcuMX = 4.0
VUMX = 0.2 V UMN = -0.2 ]LIM = -0.012
KR = Radius of UEL characteristic
Kc = Centre of UEL characteristic

Figure 8.45 An example of UEL model [26]

 370 Excitation Systems Chap. 8
(b) V/Hz limiter:

An example of the V/Hz limiter model is shown in Figure 8.46. The operation
of the limiter is quite straightforward. When the per unit V/Hz value exceeds the
limiting value of VzLM' a strong negative signal drives the excitation down. The VZL
limit is set typically at 1.07 to 1.09 pu. M

AVR summing
Et (pu) VzLM function
1.0 Ve
f (pu)---4 E/f
+ L ~~ 1
s
Kz2 Vz ~ L r----

+
o

Sample data:

VzLM:;: 1.07 KZJ :;: 1000 KZ2 :;: 0.007

Figure 8.46 V/Hz limiter model

(e) Field-current or overexcitation limiter:

Figure 8.47 shows the model of a field-current limiter circuit. It is designed

to have a limiting action as shown in Figure 8.47(b). A high setting provides almost
instantaneous limiting at 1.6 times full-load current (FLC). A low setting of l .05xFLC
in conjunction with a ramp timing function provides a limiting action with time delay
dependent on the level of field current. For example, a field current level of
l.325xFLC will be allowed for 15 s, followed by a reduction in current level to
l.05xFLC over the next 15 s.
Referring to the block diagram of Figure 8.47(a), when I¡d exceeds the high
setting IFLMJ, signal VFJ of control loop CD acts to reduce excitation instantaneously.
When the field current is below I FLMJ, the limiting action is through the control loop
(í). The magnitude of the control signal VF2 and the value of gain K2 determine the
time delay and ramping action. Once the field current reaches the low setting IFLM2'
the select switch changes to the low select position; this ensures that the field current,
in the event of a second disturbance, <loes not exceed the low setting for a mínimum
period to allow cooling of the machine. When the field current is below IFLMl' the
signal VF4 helps to reset rapidly the integrator output to zero.
sec. 8.6 Modelling of Excitation Systems 371

o o AVR summing
1 function
l¡d s +
VFJ High-low
LM2 select switch

K4 @ ---'1'"~~~--=--=...J
1--~~...;::_~~~~~~

CD Normally at
high setting

(a) Block diagram representation

Sample data:
]FLMl= l.6xfull load l¡d IFLM2 = l.05xfull load l¡d
K1 = 150 K2 = 0.248 K3 = 12.6
K4 = 140 LMI = -0.085 LM2 = -3.85

FLC = full load current

IFLM2 = 1.05 FLC

º T 30 s Time

(b) Limiting characteristics

Figure 8.47 Field-current limiter model

372 Excitation Systems Chap. 8
8.6.4 Field Testing for Model Development and Verification

Although most of the data related to excitation system models cari be obtainect
from factory tests, such data can only be considered as typical. The actual settings are
usually determined on site during installation and commissioning of the equipment.
It is therefore desirable to determine the model parameters by performing tests on the
actual equipment on site.
The nature of the tests required will depend on the type of excitation systern
being tested. A general procedure for model verification and development is as
follows:

1. Obtain circuit diagrams, block diagrams, nominal settings and setting ranges ..
Construct a detailed block diagram of the complete excitation system }

identifying gains, time constants, and non-linearities,

2. With the generator (synchronous machine) shut down, perform frequency

response tests and/or transient response tests on the individual elements
comprising the excitation systern. Identify their transfer functions, non-
linearities, saturation characteristics, and ceiling limits. Using these data,
validate as much of the detailed block diagram of the excitation system as
possible and modify the diagram as necessary.

3. Perform frequency response and time-response tests with the generator running at
rated speed and producing rated voltage on open-circuit, Measure the overall
linear response and responses at various points of the system to a step change
in terminal voltage. Perform additional tests with the generator operating near
rated load. Validate the detailed rnodel of the complete system by comparison
with the measured responses.

4. Reduce the detailed model to fit the standard model applicable to the specific
type of excitation systern using techniques of classical control theory. Validate
the responses of the reduced rnodel against the measured responses.

The above procedure is quite involved and time-consuming for the older slow-
response exciter. For a high-response excitation system, frequency response tests are
usually not required as the nurnber of time constants within the major feedback loop
is small. For static excitation systems, model parameters can usually be obtained from
design data; field tests are required only for verification.
Techniques for field testing, performance verification and model development
of excitation systems are described in references 3, 7, and 36.
 References 373

REFERENCES

[U IEEE Committee Report, "Computer Representation of Excitation Systems,"

IEEE Trans., Vol. PAS-87, pp. 1460-1464, June 1968.

[2] IEEE Committee Report, "Proposed Excitation System Definitions for

Synchronous Machines," IEEE Trans., Vol. PAS-88, pp. 1248-1258, August
1969.
[3] IEEE Committee Report, "Excitation System Dynamic Characteristics," IEEE
Trans., Vol. PAS-92, pp. 64-75, January/February 1973.

[4] IEEE Committee Report, "Excitation System Models for Power System
Stability Studies," IEEE Trans., Vol. PAS-100, pp. 494-509, February 1981.

[5] IEEE Standard Definitions for Excitation Systems for Synchronous Machines,
IEEE Standard 421.1-1986.

[6] IEEE Guide for the Preparation of Excitation System Specifications, IEEE
Standard 421.4-1987.

[7] IEEE Guide for Identifzcation, Testing and Evaluation of the Dynamic
Performance of Excitation Control Systems, IEEE Standard 421.2-1990
(revision to IEEE Standard 421A-l978).

[8] IEEE Recommended Practice for Excitation System Models for Power System
Stability Studies, IEEE Standard 421.5-1992.

[9] Guide to the Characteristics, Performance, and Hardware Requirements in the

Specification of Excitation Systems, Report of the Governor and Excitation
Control System Committee of the Canadian Electrical Association, 1982.

(10] Westinghouse Electric Corporation, Electric Transmission and Distribution

Reference Book, East Pittsburgh, Pa., 1964.

[11] E.W. Kimbark, Power System Stability, Vol. JI!: Synchronous Machines, John
Wiley & Sons, 1956.

[12] A.E. Fitzgerald and C. Kingsley, Electric Machinery, Second Edition,

McGraw- Hill, 1961.

[13] T.L. Dillman, J.W. Skooglund, F.W. Keay, W.H. South, and C. Raczkowski,
"A High Initial Response Brushless Excitation System," IEEE Trans., Vol.
374 Excitation Systems Chap. 8
PAS-90, pp. 2089-2094, September/October 1971.

[14] J.S. Bishop, D.H. Miller, and A.C. Shartrand, "Experience with ALTERREx
Excitation for Large Turbine-Generators," Paper presented at the Joint
IEEE/ASME Power Conference, Miami Beach, September 15-19, 1974.

[15] R.K. Gerlitz, R.E. Gorman, and M. Temoshok, "The GENERREX Excitation
System for Large Steam Turbine-Generators," Paper GE3-3003, Pacific Coast
Electric Association Engineering and Operation Conference, Culver City~
Calif., March 20-21, 1975.

[16] P.H. Beagles, K. Carlsen, M.L. Crenshaw, and M. Temoshok, "Generator and
Power System Performance with the GENERREX Excitation System," IEEE
Trans., Vol. PAS-95, pp. 489-493, March/April 1976.

[17] H.M. Rustebakke (editor), Electric Utility Systems and Practices, John Wiley
& Sons, 1983.

[18] F. Peneder and R. Bertschi, "Static Excitation Systems with and without a
Compounding Ancíllary," Brown Boveri Review 7-85, pp. 343-348.

[19] D.C. Lee and P. Kundur, "Advanced Excitation Controls for Power System
Stability Enhancement," CIGRE Paper 38-01, Paris, France, 1986.

[20] IEEE Standard Dictionary of Electrical and Electronics Terms (ANSI), IEEE
Standard 100-1988.

[21] P. Kundur, M. Klein, G.J. Rogers, and M.S. Zywno, "Application of Power
System Stabilizers for Enhancement of Overall System Stability," IEEE Trans.,
Vol. PWRS-4, No. 2, pp. 614-626, May 1989.

[22] A.S. Rubenstein and W.W. Walkley, "Control of Reactive kVA with Modern
Amplidyne Voltage Regulators," AJEE Trans., Part III, pp. 961-970, December
1957.

[23] A.S. Rubenstein and M. Temoshok, "Underexcited Reactive Ampere Limit for
Modern Amplidyne Voltage Regulator," AJEE Trans., Vol. PAS-73, pp.
1433-1438, December 1954.

[24] J.T. Carleton, P.O. Bobo, and D.A. Burt, "Minimum Excitation Limit for
Magnetic Amplifier Regulating System," AJEE Trans., Vol. PAS-73, pp.
869-874, August 1954.

[25] l. Nagy, "Analysis of Mínimum Excitation Limits of Synchronous Machines,"

 References 375

IEEE Trans., Vol. PAS-89, No. 6, pp. 1001-1008, July/August 1970.

[26] J.R. Ribeiro, "Mínimum Excitation Limiter Effects on Generator Response to

System Disturbances," IEEE Trans. on Energy Conversion, Vol. 6, No. 1, pp.
29-38, March 1991.

[27] Discussion of reference 26 by D.C. Lee, R.E. Beaulieu, P. Kundur, and G.J.
Rogers.

[28] M.S. Baldwin and D.P. McFadden, "Power Systems Performance as Affected
by Turbine-Generator Controls Response during Frequency Disturbances,"
IEEE Trans., Vol. PAS-100, pp. 2486-2494, May 1981.

[29] General Electric Instructions GEK-15014C, Inverse Time Maximum Excitation

Limit.

[30] p. Kundur and P.L. Dandeno, "Implementation of Synchronous Machine

Models into Power System Stability Programs," IEEE Trans., Vol. PAS-102,
pp. 2047-2054, July 1983.

[31] IEEE Guide: Test Procedures for Synchronous Machines, IEEE Standard 115-
1983.

[32] R.W. Ferguson, H. Herbst, and R.W. Miller, "Analytical Studies of the
Brushless Excitation System," AJEE Trans., Part III, pp. 961-970, 1957.

[33] R.L. Witzke, J.V. Kresser, and J.K. Dillard, "Influence of AC Reactance on
Voltage Regulation of 6-Phase Rectifiers," AJEE Trans., Vol. 72, pp. 244-253,
July 1953.

[34] L.L. Freris, "Analysis of a Hybrid Bridge Rectifier," Direct Current, pp.
22-23, February 1966.

[35] IEEE Task Force, "Conventions for Block Diagram Representations,'' IEEE
Trans., Vol. PWRS-1, No. 3, pp. 95-100, August 1986.

[36] IEEE Tutorial Course Text, "Power System Stabilization via Excitation
Control - Chapter IV: Field Testing Techniques," Publication 81 EHO 175-0
PWR.