Short Course On

"Power System Voltage Stability"

Stability
Presented by:

Dr. Prabha Kundur

Fellow Member And
Distinguished Lecturer of PES-IEEE
December 4 and 5, 2006
Cuernavaca, Morelos

A Short COURSE ON
power system Voltage STABILITY

Presented by

Dr. P. Kundur

Copyright P. Kundur
This material should not be used without the author's consent

1538pk

Power System Voltage Stability

Course Outline

1. Introduction and Basic Concepts

2. Equipment Characteristics Impacting on
Voltage Stability

3. Control of Reactive Power and Voltage

4. Typical Scenarios of Voltage Instability
5. Methods of Analysis
6. Prevention of Voltage instability
7. Examples of Major Disturbances due to
Voltage Instability

1538pk

1. INTRODUCTION AND BASIC

CONCEPTS

Copyright P. Kundur
This material should not be used without the author's consent

1538pk

Introduction and Basic Concepts

Outline

1. Definition and Classification of Power System

Stability
2. Conceptual Relationship between Power
System Stability, Security and Reliability
3. Review of the Concepts of Active Power and
Reactive Power
4. An Elementary View of Voltage Stability
Phenomenon

1538pk

Power System Stability: Basic Concepts

and Definition
z

Power System Stability denotes the ability of

an electric power system, for a given initial
operating condition, to regain a state of
operating equilibrium after being subjected to
a physical disturbance, with all system
variables bounded so that the system integrity
is preserved
) Integrity of the system is preserved when
practically the entire power system remains
intact with no tripping of generators or loads,
except for those disconnected by isolation of
the faulted elements or intentionally tripped to
preserve the continuity of operation of the rest
of the system

Stability is a condition of equilibrium between

opposing forces:
) instability results when a disturbance leads to
a sustained imbalance between the opposing
forces
) instability is a run-away or run-down situation

1538pk

Basic Concepts (contd)

The power system is a highly nonlinear

system which operates in a constantly
changing environment:
) loads, generator outputs, topology and key
operating parameters change continually

When subjected to a disturbance, the system

stability depends on:
) the nature of the disturbance, as well as
) the initial operating condition

The disturbances may be small or large:

) small disturbances in the form of load changes
occur continually
) large disturbances of a severe nature, such as
a short-circuit on a transmission line or loss of
a large generator

1538pk

Basic Concepts (contd)

Following a transient disturbance, if the power

system is stable it will reach a new equilibrium
state with practically the entire system intact:
) faulted element and any connected load
disconnected
) actions of automatic controls and possibly
operator action will eventually restore system
to normal state

On the other hand, if the system is unstable, it

will result in a run-away or run-down situation;
for example,
) a progressive increase in angular separation
of generator rotors, or
) a progressive decrease in bus voltages

An unstable system condition could lead to

cascading outages, and a shut-down of a
major portion of the power system

1538pk

Need for Classification of Stability

Power system dynamic performance is

influenced by a wide array of devices with
different response rates and characteristics

Instability may be manifested in many different

ways depending on system configuration and
operating conditions
) mode of instability depends on which set of
balancing forces experience a sustained
imbalance

Due to the high dimensionality and complexity

of the system, it is essential to make
simplifying assumptions and to analyze
specific problems using the right degree of
detail

Not very effective to study power system

stability as a single problem

1538pk

Classification of Power System Stability

z

Classification of stability into various categories

greatly facilitates:
) analysis of stability problems
) identification of essential factors which
contribute to instability
) devising methods of improving stable operation

Classification is based on the following

considerations:
) physical nature of the resulting instability
) size of the disturbance considered
) most appropriate method of analysis
) devices, processes, and the time span involved

We should always keep in mind the overall

stability
) solutions to problems of one category should
not be at the expense of another

1538pk

Classification of Power System Stability

Power
PowerSystem
SystemStability
Stability
ability to remain in operating equilibrium
equilibrium between opposing forces

Angle
Angle
Stability
Stability
ability to maintain synchronism
torque balance of synchronous
machines

Small
Small
Signal
Signal
Stability
Stability

Frequency
Frequency
Stability
Stability

Short
Short
Term
Term

Voltage
Voltage
Stability
Stability
ability to maintain steady
voltages
reactive power balance
equilibrium of voltage control

ability to maintain frequency

within nominal range
generation/load balance

Transient
Transient
Stability
Stability

Consideration
for
Classification

Large
Large
Disturbance
Disturbance
Voltage
Voltage
Stability
Stability

Short
Short
Term
Term

Long
Long
Term
Term

Short
Short
Term
Term

Small
Small
Disturbance
Disturbance
Voltage
Voltage
Stability
Stability

Long
Long
Term
Term

Physical Nature/
Main System
Parameter

Size of
Disturbance

Time Span

1538pk

Rotor Angle Stability

z

Ability of interconnected synchronous machines

to remain in synchronism under normal
conditions and after being subjected to a
disturbance

Depends on the ability to maintain/restore

equilibrium between electromagnetic torque and
mechanical torque of each synchronous
machine in the system

If the generators become unstable when

perturbed, it is as a result of
) a run-away situation due to torque imbalance

A fundamental factor is the manner in which

power outputs of synchronous machines vary
as their rotor angles swing

Instability that may result occurs in the form of

increasing angular swings of some generators
leading to loss of synchronism with other
generators
10

1538pk

Voltage Stability
z

Ability of power system to maintain steady

voltages at all buses in the system after being
subjected to a disturbance from a given initial
operating condition

A system experiences voltage instability when a

disturbance, increase in load demand, or
change in system condition causes:
) a progressive and uncontrollable fall or rise in
voltage of some buses

Main factor causing voltage instability is the

inability of power system to maintain a proper
balance of reactive power and voltage control
actions

The driving force for voltage instability is

usually the loads. Following a condition of
reduced transmission system voltages,
) power consumed by the loads tend to be
restored by the action if distribution voltage
regulators, tap changing transformers, and
thermostats
11

1538pk

Voltage Stability (contd)

The possible outcome of voltage instability:

) loss of load in the area where voltages reach
unacceptably low levels, or
) loss of integrity of the power system

Progressive but rapid drop in bus voltage can

also be associated with rotor angles going out
of step:
) voltages at points close to the electrical center
reach very low values
) a much faster phenomenon
) not a voltage instability phenomenon

In contrast, sustained fall of voltage related to

voltage instability occurs where rotor angle
stability is not an issue or the cause:
) usually a slower phenomenon

12

1538pk

Frequency Stability
z

Ability to maintain steady frequency within a

nominal range following a disturbance resulting
in a significant imbalance between generation
and load:
) instability that may result occurs in the form of
sustained frequency swings leading to tripping of
generating units and/or loads

In a small island system, frequency stability

could be of concern for any disturbance causing
a significant loss of load or generation

In a large interconnected system, frequency

stability could be of concern only following a
severe system upset resulting in the system
splitting into one or more islands

Depends on the ability to restore balance

between generation and load of island systems
with minimum loss of load and generation

Generally, frequency stability problems are

associated with inadequacies in equipment
responses, poor coordination of control and
protection systems

13

1538pk

Relationship Between the Concepts of

Reliability, Security and Stability
of a Power System
z

Stability: refers to the continuance of intact

operation following a disturbance
) depends on the operating condition and the
nature of the physical disturbance

Security: the degree of risk in the ability to

survive imminent disturbances (contingencies)
without interruption of customer service.
) depends on the system operating condition as
well as the contingent probability of
disturbances.

Reliability: probability of satisfactory operation

over the long run
) denotes the ability to supply adequate electric
service on a nearly continuous basis, with few
interruptions over an extended time period

14

1538pk

Essential Differences Among the Three

Aspects of Power System Performance

Reliability is the overall objective in power

system design and operation
) To be reliable the power system must be
secure most of the time.

To be secure the system must be stable but

must also be secure against other
contingencies that would not be classified as
stability problems, e.g. damage to equipment
such as an explosive failure of a cable, fall of
transmission towers due to ice loading or
sabotage.

As well, a system may be stable following a

contingency, yet insecure due to post-fault
system conditions resulting in equipment
overloads or voltage violations

15

1538pk

Essential Differences (cont'd)

System security may be further distinguished

from stability in terms of the resulting
consequences.
) For example, two systems may both be stable
with equal stability margins, but one may be
relatively more secure because the
consequences of instability are less severe

Security and stability are time-varying

attributes which can be judged by studying
the performance of the power system under a
particular set of conditions.

Reliability, on the other hand, is a function of

the time-average performance of the power
system; it can only be judged by consideration
of the system's behaviour over an appreciable
period of time.

16

1538pk

Concepts of Active and

Reactive Power

17

1538pk

Power in an A.C. Circuit

Let us first look at simple dc circuits:

Energy is stored in inductance L and capacitance C

z

With A.C., energy is stored and discharged twice

every cycle

Instantaneous power, p = ei

18

1538pk

Concepts of Active and Reactive Powers

Consider the single phase circuit shown in Fig. A.1

with
i

e = Em Sin t
i = Im Sin (t - )

Instantaneous power

FIG A.1

19

1538pk

Figure A.2 shows plots of e, i, p, pp and pq.

Active Power (pp) represents the component of p

utilized for permanent irreversible consumption.
It has an average value of P.

Reactive Power (pq) is utilized in establishing

either a magnetic or electrostatic field; it is
stored in inductance or capacitance and then
returned to the source. It has a zero average
value.

20

1538pk

ACTIVE POWER

pp = P (1 - cos 2 t)

REACTIVE POWER

pq = -Q sin t

Fig. A.2

21

1538pk

3-Phase Power

Under balanced conditions, with

ea = Em sin t
ia = Im sin ( t - )
We have
p = ea ia + eb ib + ec ic
= 3 EI cos
= 3 x average active power per phase
z

Total instantaneous power is constant

We are tempted to assume that reactive power is

not important in a 3-phase network:
) not appropriate to do so;
) power in each phase is of significance

We commonly refer to 3-phase reactive power as

being equal to 3 times single phase Q
) instantaneous values of Q in all three phases add
up to zero; similar to sum of the 3-phase currents
being equal to zero

22

1538pk

Reactive Power - Sign Conventions

z By convention, the Q associated with:

) inductive load is positive

) capacitive load is negative

An inductive load absorbs Q

A capacitive load supplies Q

z A synchronous machine:

) when overexcited, supplies reactive power

) when underexcited, absorbs reactive

23

1538pk

Complex Power

Complex power, S, is defined as:

The reason for using conjugate of I in the complex

multiplication is to arrive at the agreed upon
convention for reactive power
) Q is positive with I lagging E

S = E I = P2 + Q 2

The magnitude of the complex power, S, is normally

used to specify the rating of equipment
24

1538pk

Special Comments on Reactive Power

z

Although the Reactive Power associated with an

element has a zero average value, it represents
"real" power required
) to store and discharge magnetic energy in an
inductance or electrostatic energy in a capacitance,
twice every cycle.

In a network the interchange of energy takes place

between the source, inductive elements and
capacitive elements.
) The net energy associated with reactive power is the
sum of various inductive and capacitive stored
energies.

The oscillatory transfer of reactive power between

points in a power system results in voltage drops
and losses in generation and transmission
equipment.
As efficiency and voltage regulation are very
important in the operation of power systems, the
efficient management and control of reactive power
are of prime importance.

25

1538pk

General Observations on Active and

Reactive Power Flow
z

In a practical transmission system, normally:

) the active power flow is determined primarily by
angular differences between bus voltages; and
) the reactive power flow by magnitude differences of
bus voltages

Active Power is supplied only by generators:

) the desired flow of active power from a generator is
achieved by control of prime mover mechanical
torque.
) Increasing the mechanical torque advances the
generator rotor and hence, the "internal voltage"
with respect to other system voltages.

Sources of Reactive Power:

) synchronous machines (over excited)
) static capacitors
) capacitance of transmission lines

Consumers of Reactive Power:

) synchronous machines (under excited)
) induction motors
) inductive static loads
) inductance of transmission lines, transformers
) AC/DC and DC/AC converters

26

1538pk

At any junction:
P=0
Q=0
They can be added arithmetically

For a system:
) a balance sheet of active and reactive power can be
drawn;
) the total injected P and Q are equal to the total
extracted P and Q, plus any P and Q losses.

Under steady-state conditions;

) P and Q flow over a network are fairly independent
of each other and are influenced by different control
actions.

27

1538pk

An Elementary View of Voltage

Stability Phenomenon

28

1538pk

Simple Radial System

29

1538pk

As load demand increases (ZLD decreases)

z

IR increases, and VR decreases

PR increases rapidly at first, and then slowly

before reaching a maximum, and finally decreases

PR maximum when voltage drop in the line is equal in

magnitude to VR, i.e. ZLD = ZLN
z

Represents limit of satisfactory operation, i.e.

critical operating point

For a load demand higher than the maximum

z

Control of power by varying load is unstable

With constant-admittance load characteristic,

conditions stabilize at a voltage level lower than
normal

If load is supplied by a transformer with ULTC, the

tap changer action leads to a progressive voltage
reduction, i.e. voltage instability
Maximum PR can be increased by increasing ES
and/or decreasing f

30

1538pk

P - V Characteristics
z

A more traditional method of illustrating the

phenomenon

Figure below shows VR - PR characteristic of the

simple radial system for different values of load
power factor

Fig. 14.2 VR-PR characteristics of the radial

system

31

1538pk

Let us next look at the VR - PR characteristics of

an actual line:

Fig. 6.7 Voltage-power characteristics of a 300 km lossless

radial line

VR - PR characteristics of a practical power

system:

Fig. 14.4 V-P curve at bus 530 of system shown in Fig. 14.3

32

1538pk

Fig. 14.3 39 bus, 10 machine test system

33

1538pk

Thus, complex systems with a large number of

voltage sources and load buses have P-V
characteristics similar to that of a radial system
) represent the basic property of networks with
predominantly inductive elements

Q-V Characteristic
z

Another characteristic often used for voltage

stability analysis is the Qi - VR relationship
) shows sensitivity and variation of bus voltages
with respect to injected Q
) readily determined
) suited for examining requirements of VAR
compensation
) voltage instability when dQ/dV negative

The bottom of Q-V curve represents stability limit

) operation on the right side is stable and on the
left side is unstable

34

1538pk

Fig. 14.5 Q-V curves for system shown in Fig. 14.3

35

1538pk

Principal causes of voltage instability are seen

to be:
) load on transmission lines too high
) voltage sources too far from load centres
) source voltages too low, and
) insufficient VAR compensation

The principle driving force for voltage

instability is usually the loads:
) in response to a disturbance, power consumed
by loads tends to be restored by motor slip
adjustment, distribution voltage regulators,
and thermostats
) restored loads increase stress on the HV
network causing further voltage reduction
) voltage instability occurs when load dynamics
attempt to restore power consumption beyond
the capability of the transmission network

The term voltage collapse is also often used.

) It is the process by which the sequence of
events accompanying voltage instability leads
to a blackout or abnormally low voltages in a
significant part of the power system

36

1538pk

Factors Influencing Voltage Instability

Problem associated with the transfer of active

and reactive power through highly inductive
network

In addition to strength of transmission network

and power transfer levels, the principal
contributing factors are:
) generator reactive power and voltage control
limitations
) load characteristics
) distribution system voltage regulator and
substation transformer tap-changer action
) reactive power compensating device
characteristics

37

1538pk

38

1538pk

39

1538pk

40

1538pk

41

1538pk

42

1538pk

43

1538pk

44

1538pk

45

1538pk

46

1538pk

47

1538pk

48

1538pk

49

1538pk

50

1538pk

51

1538pk

52

1538pk

2. EQUIPMENT CHARACTERISTICS
IMPACTING ON VOLTAGE STABILITY

Copyright P. Kundur
This material should not be used without the author's consent

1538pk

Equipment Characteristics Impacting on

Voltage Stability

Generators (Synchronous Machines)

Excitation Systems

AC Transmission

Power System Loads

Voltage Control and Reactive Compensating

Devices
(discussed in next section)

1538pk

Influence of Generator Characteristics

Action of generator AVRs provide the primary source of

voltage support in power systems

Under normal conditions, terminal voltages are

maintained constant

During conditions of low/high system voltages

) the VAr demand on generators may exceed their reactive
capability limits

With VAr output limited, terminal voltage is no longer

maintained constant

The case illustrated in Fig. 14.6 demonstrates the impact

of loss of voltage control capability in critical situations

1538pk

(a) Schematic diagram

(b) The VR-PR characteristics

Fig. 14.6 Impact of loss of regulation of intermediate bus voltage

An operating condition represented by point A is considerably

more stable on curve 1 than on curve 2

Results demonstrate the importance of maintaining the voltage

control capability of generators

Results show also that the degree of stability cannot be judged

based on how close the bus voltage is to normal level
This situation is similar to that which led to voltage collapse of
Brittany region of the French system in December 1965 and in
November 1975.
3

1538pk

Reactive Capability Limits of Synchronous

Machines

In voltage stability studies, it is important to consider

the reactive capability limits of synchronous machines

Synchronous generators are rated in terms of maximum

MVA output at a specified voltage and power factor
which can be carried continuously without overheating

The active power output is limited by the prime mover

capability

The continuous reactive power output capability is

limited by three considerations
) armature current limit
) field current limit
) end region heating limit

1538pk

Armature Current Limit

Armature current results in I2aRa power loss, and the

resulting heat imposes a limit on the output
The per unit complex output power is
~ *
S = P + jQ = E t ~I t = E t It (cos + j sin )

where is the power factor angle

In a P-Q plane the armature current limit, as shown in

Fig. 5.12, appears as a circle with centre at the origin and
radius equal to the MVA rating

Fig. 5.12 Armature current heating limit

1538pk

Field Current Limit

Because of the heating resulting from RfdI2fd power loss,

the field current imposes the second limit

The phasor diagram relating Et, It and Eq (with Ra

neglected) is shown in Fig. 5.13
Equating the components along and perpendicular to the
phasor Et
X adi fd sin i = X slt cos
X adi fd cos i = E t + X slt sin

Therefore
X ad
E tifd sin i
Xs
X ad
E2t
Q = E tlt sin =
E tifd cos i
Xs
Xs

P = E tlt cos =

The relationship between P and Q for a given field current

is a circle centered at on the Q-axis and with as the
radius. The effect of the maximum field current on the
capability of the machine is shown in Fig. 5.14

In any balanced design, the thermal limits for the field and
armature intersect at a point (A) which represents the
machine name-plate MVA and power factor rating

1538pk

Field Current Limit

Fig. 5.13 Steady state phasor diagram

Fig. 5.14 Field current heating limit

7

1538pk

End Region Heating Limit

The localized heating in the end region of the armature

affects the capability of the machine in the underexcited
condition

The end-turn leakage flux, as shown in Fig. 5.15, enters and

leaves in a direction perpendicular (axial) to the stator
lamination. This causes eddy currents in the laminations
resulting in localized heating in the end region

The high field currents corresponding to the overexcited

condition keep the retaining ring saturated, so that end
leakage flux is small. However, in the underexcited region
the field current is low and the retaining ring is not
saturated; this permits an increase in armature end leakage
flux

Also, in the underexcited condition, the flux produced by the

armature current adds to the flux produced by the field
current. Therefore, the end-turn flux enhances the axial flux
in the end region and the resulting heating effect may
severely limit the generator output, particularly in the case
of a round rotor machine

Fig. 5.16 shows the locus of end region heating limit on a

P-Q plane

1538pk

End Region Heating Limit

Fig. 5.15 Sectional view end region of a generator

Fig. 5.16 End region of heating limit

1538pk

Reactive Capability Limit of a 400 MVA

Hydrogen Cooled Steam Turbine Generator

Fig. 5.18 shows the reactive capability curves of a 400 MVA

hydrogen cooled steam turbine driven generator at rated
armature voltage
) the effectiveness of cooling and hence the allowable
machine loading depends on hydrogen pressure
) for each pressure, the segment AB represents the field
heating limit, the segment BC armature heating limit, and the
segment CD the end region heating limit

Fig. 5.18 Reactive capability curves of a hydrogen cooled

generator at rated voltage

10

1538pk

Fig. 5.17: Effect of reducing the armature voltage on the

generator capability curve

11

1538pk

Excitation Systems

12

1538pk

Functions and Performance Requirements of

Excitation Systems

The functions of an excitation system are

) to provide direct current to the synchronous
generator field winding, and
) to perform control and protective functions essential
to the satisfactory operation of the power system

The performance requirements of the excitation

system are determined by
a) Generator considerations:
) supply and adjust field current as the generator
output varies within its continuous capability
) respond to transient disturbances with field
forcing consistent with the generator short term
capabilities:
) rotor insulation failure due to high field
voltage
) rotor heating due to high field current
) stator heating due to high VAR loading
) heating due to excess flux (volts/Hz)
b) Power system considerations:
) contribute to effective control of system voltage
and improvement of system stability

13

1538pk

Elements of an Excitation System

Exciter: provides dc power to the generator field winding

Regulator: processes and amplifies input control signals to

a level and form appropriate for control of the exciter

Terminal voltage transducer and load compensator: senses

generator terminal voltage, rectifies and filters it to dc
quantity and compares with a reference; load comp may be
provided if desired to hold voltage at a remote point

Power system stabilizer: provides additional input signal to

the regulator to damp power system oscillations

Limiters and protective circuits: ensure that the capability

limits of exciter and generator are not exceeded

14

1538pk

Fig. 8.14: Excitation system control and protective circuits

15

1538pk

Underexcitation Limiter (UEL):

) intended to prevent reduction of generator excitation to
a level where steady-state stability limit or stator core
end-region heating limit is exceeded
) control signal derived from a combination of either
voltage and current or active and reactive power of the
generator
) a wide variety of forms used for implementation
) should be coordinated with the loss-of-excitation
protection (see Figure 8.17)

Overexcitation Limiter (OXL)

) purpose is to protect the generator from overheating
due to prolonged field overcurrent
) Fig. 8.18 shows thermal overload capability of the field
winding
) OXL detects the high field current condition and, after a
time delay, acts through the ac regulator to ramp down
the excitation to about 110% of rated field current; if
unsuccessful, trips the ac regulator, transfers to dc
regulator, and repositions the set point corresponding
to rated value
) two types of time delays used: (a) fixed time, and (b)
inverse time
) with inverse time, the delay matches the thermal
capability as shown in Figure 8.18

16

1538pk

Fig. 8.17: Coordination between UEL, LOE relay and stability

limit

Fig. 8.18: Coordination of over-excitation limiting with field

thermal capability
17

1538pk

(a) Block diagram representation

(b) Limiting characteristics

Fig. 8.47: Field-current limiter model
18

1538pk

AC Transmission

19

1538pk

Performance Equations and Parameters of

Transmission Lines

A transmission line is characterized by four

parameters:
) series resistance (R) due to conductor resistivity
) shunt conductance (G) due to currents along insulator
strings and corona; effect is small and usually
neglected
) series inductance (L) due to magnetic field
surrounding the conductor
) shunt capacitance (C) due to the electric field between
the conductors

These are distributed parameters.

The parameters and hence the characteristics of

cables differ significantly from those of overhead lines
because the conductors in a cable are
) much closer to each other
) surrounded by metallic bodies such as shields, lead or
aluminum sheets, and steel pipes
) separated by insulating material such as impregnated
paper, oil, or inert gas

20

1538pk

For balanced steady-state operation, the performance of

transmission lines may be analyzed in terms of singlephase equivalents.

Fig. 6.1 Voltage and current relationship of a distributed

parameter line

The general solution for voltage and current at a

distance x from the receiving end (see book: page 202)
is:
~
~
~
~
~ VR + ZC I R x VR Z C I R x
V=
e +
e
(6.8)
2

VR

I =

where

ZC

ZC =

+ IR

VR
x

ZC
2

~
IR

(6.9)

= zy = + j
21

1538pk

The constant ZC is called the characteristic

impedance and is called the propagation constant.

The constants and ZC are complex quantities. The

real part of the propagation constant is called the
attenuation constant , and the imaginary part the

phase constant .

If losses are completely neglected,

ZC =

L
= Real Number
C
(pure resistance)

= j = Imaginary number

22

1538pk

For a lossless line, Equations 6.8 and 6.9 simplify to

V =VR cos x + jZ C I R sin x

(6.17)

I = I R cos x + j V R Z sin x

(6.18)

When dealing with lightening/switching surges, HV lines

are assumed to be lossless. Hence, ZC with losses
neglected is commonly referred to as the surge
impedance.
The power delivered by a line when terminated by its
surge impedance is known as the natural load or surge
impedance load.
V02
SIL =
watts
ZC

where V0 is the rated voltage

At SIL, Equations 6.17 and 6.18 further simplify to

~

V = VR ex
~

I = I R e x

23

(6.20)
(6.21)

1538pk

Hence, for a lossless line at SIL,

) V and I have constant amplitude along the line
) V and I are in phase throughout the length of the line
) The line neither generates nor absorbs VARS

As we will see later, the SIL serves as a convenient

reference quantity for evaluating and expressing line
performance

Typical values of SIL for overhead lines:

nominal (kV): 230
SIL (MW):
140

345
420

500
1000

765
2300

Underground cables have higher shunt capacitance;

hence, ZC is much smaller and SIL is much higher than
those for overhead lines.
) for example, the SIL of a 230 kV cable is about
1400 MW
) generate VARs at all loads

24

1538pk

Voltage - Power Characteristics

of a Radial Line

Corresponding to a load of PR+jQR at the receiving end, we

have
~ PR jQR
IR =
~*

VR

Assuming the line to be lossless, from Equation 6.17

with x = l
~

PR jQR

V~*
R

ES = VR cos + jZC sin

Fig. 6.7 shows the relationship between VR and PR for a 300

km line with different loads and power factors.
The load is normalized by dividing PR by P0, the natural load
(SIL), so that the results are applicable to overhead lines of
all voltage ratings.

From Figure 6.7 the following fundamental properties of ac

transmission are evident:
a) There is an inherent maximum limit of power that can be
transmitted at any load power factor. Obviously, there has to
be such a limit, since, with ES constant, the only way to
increase power is by lowering the load impedance. This will
result in increased current, but decreased VR and large line
losses. Up to a certain point the increase of current dominates
the decrease of VR, thereby resulting in an increased PR.
Finally, the decrease in VR is such that the trend reverses.
25

1538pk

Fig. 6.7 Voltage-power characteristics of a 300 km lossless

radial line

26

1538pk

Voltage - Power Characteristics

of a Radial Line (cont'd)

b) Any value of power below the maximum can be

transmitted at two different values of VR. The normal
operation is at the upper value, within narrow limits
around 1.0 pu. At the lower voltage, the current is higher
and may exceed thermal limits. The feasibility of
operation at the lower voltage also depends on load
characteristics, and may lead to voltage instability.
c) The load power factor has a significant influence on VR
and the maximum power that can be transmitted. This
means that the receiving end voltage can be regulated by
the addition of shunt capacitive compensation.
up to a certain limit as we will see in the next
section.

Fig. 6.8 depicts the effect of line length:

) For longer lines, VR is very sensitive to variations in PR.
) For lines longer than 600 km ( > 45), VR at natural load
is the lower of the two values which satisfies Equation
6.46. Such operation is likely to be unstable.

27

1538pk

Fig. 6.8 Relationship between receiving end voltage, line

length, and load of a lossless radial line

28

1538pk

Voltage-Power Characteristic of a Line

Connected to Sources at Both Ends

With ES and ER assumed to be equal, the following

conditions exist:
) the midpoint voltage is midway in phase between ES
and ER
) the power factor at midpoint is unity
) with PR>P0, both ends supply reactive power to the line;
with PR<P0, both ends absorb reactive power from the
line.

Fig. 6.9 Voltage and current phase relationships with ES equal

to ER, and PR less than Po

Fig. 6.8 (developed for a radial line) may be used to

analyze how Vm varies with PR.
) with the length equal to half that of the actual line, plots
of VR shown in Figure 6.8 give Vm.

29

1538pk

Power Transfer and Stability Considerations

Assuming a lossless line, from Equation 6.17 with

x = l, we can show that
PR =

ES E R
sin
ZC sin

(6.51)

where = l is the electrical length of line and is the

angle by which ES leads ER, i.e. the load angle.

If ES = ER = rated voltage, then the natural load is

PO =

ES E R
ZC

and Equation 6.51 becomes

PR =

PO
sin
sin

The above is valid for synchronous as well as

asynchronous load at the receiving end.

Fig. 6.10(a) shows the - PR relationship for a 400 km

line.
For comparison, the Vm - PR characteristic of the line is
shown in Fig. 6.10(b).
30

1538pk

Fig. 6.10 PR- and Vm-PR characteristics of 400 km lossless line

transmitting power between two large systems
31

1538pk

Reactive Power Requirements

From Equation 6.17, with x = l and ES = ER = 1.0, we can

show that
QR = QS
ES2 (cos cos )
=
ZC sin

Fig. 6.11 shows the terminal reactive power

requirements of lines of different lengths as a function
of PR.

Adequate VAR sources must be available at the two

ends to operate with varying load and nearly constant
voltage.

32

1538pk

Fig. 6.11 Terminal reactive power as a function of power

transmitted for different line lengths

33

1538pk

Factors Influencing Transfer of Active and

Reactive Power

Consider two sources connected by an inductive

reactance as shown in Figure 6.21.
) representation of two sections of a power system
interconnected by a transmission system
) a purely inductive reactance is considered because
impedances of transmission elements are
predominately inductive
) effects of shunt capacitances do not appear explicitly

(a) Equivalent system diagram

= load angle
= power factor angle
(b) Phasor diagram
Fig. 6.21 Power transfer between two sources

34

1538pk

The complex power at the receiving end is

~

SR

~
~
~ ES E R
= PR + jQ R = ER I = ER

jX
~ ~*

E cos + jES sin ER

= ER S

jX

Hence,
ES E R
sin
X
ES ER cos ER2
=
X

PR =

(6.79)

QR

(6.80)

Similarly,
ES E R
sin
X
ES2 ES ER cos
QS =
X
PS =

(6.81)
(6.82)

Equations 6.79 to 6.82 describe the way in which

active and reactive power are transferred

Let us examine the dependence of P and Q transfer

on the source voltages, by considering separately
the effects of differences in voltage magnitudes and
angles

35

1538pk

(a) Condition with = 0:

From Equations 6.79 to 6.82, we have

PR = PS = 0
QR =

ER (ES ER )
,
X

QS =

ES (ES ER )
X

With ES > ER, QS and QR are positive

With ES < ER, QS and QR are negative

As shown in Fig. 6.22,

) transmission of lagging current through an inductive
reactance causes a drop in receiving end voltage
) transmission of leading current through an inductive
reactance causes a rise in receiving end voltage

Reactive power "consumed" in each case is

(E ER )
= S

QS QR

(a) ES>ER

= XI 2

(b) ER>ES

Fig. 6.22 Phasor diagrams with = 0

36

1538pk

(b) Condition with ES = ER and 0

From Equations 6.79 to 6.82, we now have

E2
sin
PR = PS =
X
E2
(1 cos )
QS = QR =
X
=

1
XI
2

With positive, PS and PR are positive, i.e., active

power flows from sending to receiving end

In each case, there is no reactive power transferred

from one end to the other; instead, each end
supplies half of Q consumed by X.

(a) > 0

(b) < 0

Figure 6.23 Phasor diagram with ES = ER

37

1538pk

(c) General case applicable to any condition:

We now have
ES cos + jES sin ER
jX
ES2 + ER2 2ES ER cos
QS QR =
X
2
(XI ) = XI 2
=
X
I =

(6.83)

(6.84)

If, in addition to X, we consider series resistance R

of the network, then
Qloss

P R2 + Q R2
= XI = X
ER2

(6.85)

Ploss

P R2 + Q R2
= RI = R
E R2

(6.86)

The reactive power "absorbed" by X for all

conditions is X I 2. This leads to the concept of
"reactive power loss", a companion term to active
power loss.

An increase in reactive power transmitted increases

active as well as reactive power losses. This has an
impact on efficiency and voltage regulation.
38

1538pk

Conclusions Regarding Transfer of Active and

Reactive Power

The active power transferred (PR) is a function of

voltage magnitudes and . However, for satisfactory
operation of the power system, the voltage magnitude
at any bus cannot deviate significantly from the
nominal value. Therefore, control of active power
transfer is achieved primarily through variations in
angle .

Reactive power transfer depends mainly on voltage

magnitudes. It is transmitted from the side with higher
voltage magnitude to the side with lower voltage
magnitude.

Reactive power cannot be transmitted over long

distances, since it would require a large voltage
gradient to do so.

An increase in reactive power transfer causes an

increase in active as well as reactive power losses.

Although we have considered a simple system, the general

conclusions are applicable to any practical system, In fact, the
basic characteristics of ac transmission reflected in these
conclusions have a dominant effect on the way in which we
operate and control the power system.

39

1538pk

Power System Loads

40

1538pk

Influence of Load Characteristics

Load characteristics and distribution system voltage

control devices are among key factors influencing
voltage stability

Voltage and power flows settle at values determined by

composite characteristic of the transmission system and
loads

Substation transformer ULTCs and distribution voltage

regulators attempt to maintain voltage at points of
consumption
) within normal control range, load P and Q effectively
constant
) may have destabilizing effect during conditions of voltage
collapse

When the ULTCs reach the end of their tap range,

distribution voltages drop
) residential load P and Q drop, reducing line loading and
reactive losses
) industrial loads with large components of induction
motors change little, however, their capacitors will supply
less VArs, causing a net increase in Q load

41

1538pk

Loads (cont'd)

When distribution voltages remain low for a few minutes,

thermostats and other load regulating devices tend to
restore load
) more such devices will operate at any given time
) loads restored to normal full voltage value in 10 to 15
minutes
) distribution voltages drop further

At voltages below 0.85 to 0.9 pu, some motors may stall

and draw high reactive current
) voltages drop further

Industrial and commercial motors are usually controlled

by magnetically held contactors
) hence, voltage drop causes motors to drop out
) loss of load results in voltage recovery
) motors restored after some time; voltages drop again of
original problem still persists

42

1538pk

Load Modelling

A typical load bus represented in stability studies is

composed of a large number of devices:
) fluorescent and incandescent lamps, refrigerators,
heaters, compressors, furnaces, and so on

The composition changes depending on many

factors, including:
) time
) weather conditions
) state of the economy

The exact composition at any particular time is

difficult to estimate. Even if the load composition
were known, it would be impractical to represent
each individual component.

For the above reasons, load representation is based

on considerable amount of simplification.

43

1538pk

Basic Load Modelling Concepts

The aggregated load is usually represented at a

transmission substation
) includes, in addition to the connected load devices,
the effects of step-down transformers,
subtransmission and distribution feeders, voltage
regulators, and VAr compensation

Fig. 7.1 Power system configuration identifying parts of the

system represented as load at a bulk power delivery point (Bus A)

Load models are traditionally classified into:

) static load models
) dynamic load models

44

1538pk

Static Load Models

Express the load characteristics as algebraic functions

of bus voltage magnitude and frequency.

Traditionally, voltage dependency has been

represented by the exponential model:

( )
Q = Q (V )
P = P0 V

V =

V
V0

P0, Q0, and V0 are the values of the respective variables

at the initial operating condition.

For composite loads,

) exponent "a" ranges between 0.5 and 1.8

) exponent "b" ranges between 1.5 and 6

The exponent "b" is a nonlinear function of voltage.

This is caused by magnetic saturation of distribution
transformers and motors.

45

1538pk

An alternative static model widely used is the

polynomial model:

[
Q = Q [q V

]
+qV +q ]

P = P0 p1V 2 + p2V + p3
0

This model is commonly referred to as the "ZIP" model,

as it is composed of constant impedance (Z), constant
current (I), and constant power (P) components.

The frequency dependency of load characteristics is

usually represented by multiplying the exponential or
polynomial model by a factor:
For example,

[
Q = Q [q V

]
+ q V + q ] (1 + K f )

P = P0 p1V 2 + p2V + p3 (1 + K pf f )
0

qf

where f is the frequency deviation (f-f0). Typically, Kpf

ranges from 0 to 3.0, and Kqf ranges from -2.0 to 0.

Response of most loads is fast and steady state

reached quickly, at least for modest changes in V and f.
) use of static model justified in such cases

46

1538pk

Dynamic Load Models

For voltage stability studies, particularly those dealing

with short-term voltage stability, it is necessary to
account for the dynamics of loads.

Typically, motors consume 60% to 70% of total energy

supplied by a power system
) dynamics attributable to motors are usually the most
significant aspects

Other dynamic aspects of load components include:

) Extinction of discharge (mercury vapour, sodium
vapour, fluorescent) lamps when voltage drops below
0.7 to 0.8 pu and their restart after 1 or 2 seconds delay
when voltage recovers.
) Operation of protective relays. For example, starter
contractors of industrial motors drop open when
voltage drops below 0.55 to 0.75 pu.
) Thermostatic control of loads such as space
heaters/coolers, water heaters and refrigerators operate longer during low voltages and hence, total
number of devices increase in a few minutes.
) Response of ULTCs on distribution transformers and
voltage regulators

47

1538pk

Composite Load Model

Composite model which represents the wide range

of characteristics exhibited by various load
components:

Such representation used to represent each

distribution substation in the area of interest

48

1538pk

Thermostatically Controlled Loads

A realistic model for thermostatically controlled

loads:

Kp
KI
TC
ref
A
Tl
Kl
G0
GMAX

= gain of proportional controller

= gain of integral controller
= time constant of integral controller, s
= reference temperature
= ambient temperature
= load time constant, s
= gain associated with load model
= initial value of G
= maximum value of G

49

1538pk

Discharging Lighting Loads

Fig. 7.5 shows the model used to represent the

characteristics of discharge lighting loads
) at bus voltages less than V1, the lamps extinguish
) for voltages greater than V1, P and Q vary as
nonlinear functions of V

Fig. 7.5 Discharge lighting characteristics

50

1538pk

Modelling of Induction Motors

The general procedure is similar to that of a

synchronous machine
) first write basic equations in terms of phase (a,b,c)
variables
) then, transform equations into 'dq' reference frame

In developing the model of an induction motor it is

worth noting the following of its features which differ
from those of the synchronous machine:
) rotor has a symmetrical structure; hence, d and q axis
equivalent circuits are identical
) rotor speed is not fixed; this has an impact on the
selection of dq reference frame
) there is no excitation source applied to the rotor;
consequently the rotor circuit dynamics are
determined by slip rather than by excitation control.
) currents induced in shorted rotor windings produce a
field with the same number of poles as in the stator;
therefore, rotor windings may be represented by
equivalent 3-phase winding

51

1538pk

The 'dq' transformation:

) the preferred reference frame is one with axes
rotating at synchronous speed, rather than at rotor
speed

The machine equations in dq reference frame:

Stator flux linkages:
ds = Lss ids + Lmidr
qs = Lss iqs + Lmiqr

Rotor flux linkages:

dr = Lrr idr + Lmids
qr = Lrr iqr + Lmiqs

Stator voltages:
Vds = Rs ids s qs + pds
Vqs = Rs iqs + s ds + pqs

Rotor voltages:

Vdr = Rr idr (pr )qr + pdr

Vqr = Rr iqr + (pr )dr + pqr

) The term pr is the slip angular velocity and

represents the relative angular velocity between the
rotor and the reference dq axes.

52

1538pk

Fig. 7.7 Equivalent circuit of a 3-phase induction machine

Fig. 7.13 Deep-bar rotor

construction

Fig. 7.12 Double squirrel-cage

rotor bars

Fig. 7.14 Equivalent circuit of an induction motor with a double-cage

rotor

Fig. 7.15 Equivalent single rotor circuit representation of a motor with

a double-cage rotor or a deep-bar rotor

53

1538pk

Acquisition of Load Model Parameters

Two basic approaches:

) measurement-based approach
) component based approach

Measurement-based approach
) load characteristics measured at representative
substations and feeders at selected times
) parameters of loads throughout the system
extrapolated from the above

Component-based approach
) involves building up the load model from information
on its constituent parts
) load supplied at a bulk power delivery point
categorized into load classes such as residential,
commercial, and industrial
) each load class represented in terms of its
components such as lighting, heating, refrigeration
) individual devices represented by their known
characteristics

54

1538pk

Composite load model derived by aggregating

individual loads
) EPRI LOADSYN program converts data on the load
class mix, components, and their characteristics into
the form required for stability studies

55

1538pk

Component Static Characteristics

Table 7.1 summarizes typical voltage and frequency dependent
characteristics of a number of load components.
Table 7.1
Power
Factor

P/V

Q/V

P/f

Q/f

- 3-phase central

0.90

0.088

2.5

0.98

-1.3

- 1-phase central

0.96

0.202

2.3

0.90

-2.7

- window type

0.82

0.468

2.5

0.56

-2.8

1.0

2.0

Dishwasher

0.99

1.8

3.6

-1.4

Clothes washer

0.65

0.08

1.6

3.0

1.8

Clothes dryer

0.99

2.0

3.2

-2.5

Refrigerator

0.8

0.77

2.5

0.53

-1.5

Television

0.8

2.0

5.1

-4.5

Incandescent lights

1.0

1.55

Fluorescent lights

0.9

0.96

7.4

1.0

-2.8

Industrial motors

0.88

0.07

0.5

2.5

1.2

Fan motors

0.87

0.08

1.6

2.9

1.7

Agricultural pumps

0.85

1.4

1.4

5.0

4.0

Arc furnace

0.70

2.3

1.6

-1.0

-1.0

Transformer (unloaded)

0.64

3.4

11.5

-11.8

Component
Air conditioner

Water heaters,
Range top, oven
Deep fryer

56

1538pk

Load Class Static Characteristics

Table 7.2 summarizes the sample characteristics of different load

classes.

Table 7.2
Power
Factor

P/V

Q/V

P/f

Q/f

- summer

0.9

1.2

2.9

0.8

-2.2

- winter

0.99

1.5

3.2

1.0

-1.5

- summer

0.85

0.99

3.5

1.2

-1.6

- winter

0.9

1.3

3.1

1.5

-1.1

Industrial

0.85

0.18

6.0

2.6

1.6

Power plant auxiliaries

0.8

0.1

1.6

2.9

1.8

Load Class
Residential

Commercial

57

1538pk

Dynamic Characteristics

The following are sample data for induction motor equivalents

representing three different types of load (see Fig. 7.7 for
definition of parameters).

(i) The composite dynamic characteristics of a feeder

supplying predominantly a commercial load:

Rs = 0.001

Xs = 0.23

Xr = 0.23

Xm = 5.77

Rr = 0.012

H = 0.663

m = 5.0

(ii) A large industrial motor:

Rs = 0.012

Xs = 0.07

Xr = 0.165

Xm = 3.6

Rr = 0.01

H = 1.6

m = 2.0

(iii) A small industrial motor:

Rx = 0.025

Xs = 0.10

Xr = 0.17

Xm = 3.1

Rr = 0.02

H = 0.9

58

m = 2.0

1538pk

3. CONTROL OF REACTIVE POWER

AND VOLTAGE

Copyright P. Kundur
This material should not be used without the author's consent

1538pk

Reactive Power and Voltage Control

Control objectives contributing to efficient and

reliable operation of power system:
z

Voltage at terminals of all equipment are

within acceptable limits
) both utility and customer equipment designed
to operate at certain voltage rating
) prolonged operation outside allowable range
could cause them damage

System stability is satisfactory

) voltage levels and reactive power control have
significant impact on stability

The reactive power flow is minimized so as to

reduce I 2R and I 2X losses to a practical
minimum
) ensures transmission system operates
efficiently

1538pk

Production and Absorption of Reactive

Power (Q)
z

Synchronous Generators
) can generate or absorb Q depending on excitation
) capability limited by field current, armature
current, and end-region heating limits
) automatic voltage regulator continuously adjusts
excitation to control armature voltage
) primary source of voltage support!

Overhead lines
) at loads below natural or surge impedance load
(SIL), produce Q
) at loads above SIL, absorb Q

Underground cables
) have high SIL due to high capacitance
) always loaded below SIL, and hence generate Q

1538pk

Production and Absorption of Q (cont'd)

Transformers
) absorb Q due to shunt magnetizing reactance
and series leakage inductance

Loads
) a typical "load bus" is composed of a large
number of devices
) composite characteristics are normally such
that a load bus absorbs Q
) industrial loads usually have shunt capacitors
to improve power factor

As power flow conditions vary, reactive power

requirements of transmission network vary

Since Q cannot be transmitted over long

distances, voltage control has to be effected
using special devices dispersed throughout
the system

1538pk

Methods of Voltage Control

Control of voltage levels is accomplished by

controlling the production, absorption, and
flow of reactive power at all levels in the
system

Generating units provide the basic means of

voltage control

Additional means are usually required to

control voltage throughout the system:
) sources or sinks of reactive power, such as
shunt capacitors, shunt reactors, synchronous
condensers, and static var compensators
(SVCs)
) line reactance compensators, such as series
capacitors
) regulating transformers, such as tap-changing
transformers and boosters
cont'd
4

1538pk

Reactive Power Compensation

z

Shunt capacitors and reactors, and series

capacitors provide passive compensation
) are either permanently connected to the
transmission and distribution system, or switched
) contribute to voltage control by modifying the
network characteristics

Synchronous condensers and SVCs provide

active compensation
) the reactive power absorbed/supplied by them are
automatically adjusted so as to maintain voltages
of the buses to which they are connected
) together with the generating units, they establish
voltages at specific points in the system
) voltages at other locations in the system are
determined by active and reactive power flows
through various circuit elements, including the
passive compensating devices

1538pk

Objectives of Reactive Power

Compensation

To control voltage and/or improve maximum

power transfer capability

Achieved by modifying effective line

parameters:
) characteristic impedance, ZC =

L
C

) electrical length, = l
z

The voltage profile is determined by ZC

The maximum power that can be transmitted

depends on ZC as well as

1538pk

Shunt Reactors
z

Used to compensate the undesirable voltage

effects associated with line capacitance
) limit voltage rise on open circuit or light load

Shunt compensation with reactors:

) increases effective ZC
) reduces the effective natural load , i.e., voltage
at which flat voltage profile is achieved

They are connected either:

) directly to the lines at the ends, or
) to transformer tertiary windings; conveniently
switched as var requirements vary

Line reactors assist in limiting switching

surges

In very long lines, at least some reactors are

required to be connected to lines

1538pk

Shunt Capacitors
z

Used in transmission systems to compensate for

I2X losses

Connected either directly to H.V. bus or to tertiary

winding of transformers

Normally distributed throughout the system so as

to minimize losses and voltage drops

Usually switched: a convenient means of

controlling voltage

Shunt capacitor compensation of transmission

lines in effect
) decreases ZC
) increases , i.e., electrical length

Advantages: low cost and flexibility of installation

and operating

Disadvantages: Q output is proportional to square

of the voltage; hence Q output reduced at low
voltages

Shunt capacitors are used extensively in

distribution systems for power factor correction
and feeder voltage control
8

1538pk

Shunt capacitors can be effectively used up to a

certain point to extend VS limits by
) correcting receiving end power factor
) freeing up spinning reactive reserve on generators

However, shunt caps have a number of limitations

) in heavily shunt compensated system, voltage
regulation tends to be poor
) beyond a certain level of compensation, stable
operation is unattainable.

Example: 322 km 500 kV line, limiting transfer is

1700 MW requiring shunt cap compensation of 675
MVAr

1538pk

(a) Schematic diagram

(b) Equivalent p circuit representation of line

322 km, 500 kV line supplying a radial load

System and shunt capacitor Q-V characteristics

(capacitor MVAr shown at rated voltage)
10

1538pk

Series Capacitors
z

Connected in series with the line

Used to reduce effective inductive reactance of

line
) increases maximum power
) reduces I 2X loss

Series capacitive compensation in effect

reduces both:
) characteristic impedance ZC, and
) electrical length

Reactive power produced increases with

increasing power transfer
) Self regulating !

Typical applications
) improve power transfer compatibility
) alter load division among parallel lines
) voltage regulation

11

1538pk

Relative Performance of Shunt and Series Caps

(a) Power transfer as a function of transmission angle

Compensation chosen to keep
Vm at 1.0 pu when P = 1.4 Po
Po = natural load

(b) Midpoint voltage as a function of power transfer

Fig. 11.56 Performance of 600 km line with and without passive

compensation
12

1538pk

Conclusions from Results Presented in

Fig. 11.56

With shunt capacitor compensation (chosen to

keep midpoint voltage at 1.0 pu when P = 1.4 Po)
) maximum power transfer capability increased to
1.58 pu of natural power (SIL); represents an
increase of 0.16 pu over the uncompensated case
) voltage regulation is poor, i.e., the voltage
magnitude is very sensitive to variations in power
transfer

With series capacitor compensation (chosen to

keep mid point voltage at 1.0 pu when P = 1.4 Po)
) maximum power transfer capability increased to
2.65 pu
) voltage regulation significantly improved

13

1538pk

Compensation Requirements

In all cases it is not required to satisfy both the

objectives of:
) increasing the power level at which the voltage
profile is flat; and
) decreasing electrical length in order to
improve power transfer level

Short lines may require voltage support, i.e.,

increase natural load
) This may be achieved by shunt capacitors,
provided does not become excessive as a
result

Lines longer than 500 km cannot be loaded up

to natural load because of excessive
) In such cases, reduction of is the first priority

14

1538pk

Synchronous Condenser

A synchronous machine running without a prime

mover or a mechanical load

Depending on field excitation, it can either

absorb or generate vars

With a voltage regulator, it can automatically

adjust vars to maintain constant voltage

Started as an induction motor and then

synchronized

Normally connected to tertiary windings of

transformers

Unlike a SVC, a synchronous condenser has an

internal voltage

Speed of response not as fast as that of an SVC

15

1538pk

Static VAR Compensators (SVC)

z

Shunt connected static var generators and/or

absorbers whose outputs are varied so as to
control specific power system quantities

The term static is used to denote that there are

no moving or rotating components

Basic types of SVCs:

) thyristor-controlled reactor
) thyristor-switched capacitor
) saturated reactor

A static var system (SVS) is an aggregation of

SVCs and mechanically switched capacitors
or reactors whose outputs are coordinated

When operating at its capacitive limit, an SVC

behaves like a simple capacitor !

16

1538pk

(a) Controllable
reactor

(b) Fixed capacitor

(c) SVS

Fig. 11.41 Composite characteristics of an SVS

Fig. 11.44 Use of switched capacitors to extend

continuous control range

17

1538pk

Fig. 11.52 A typical static var system

(a) Voltage-current
characteristic

(b) Voltage-reactive power

characteristic

Fig. 11.53 SVS steady-state characteristics

18

1538pk

Regulated or Dynamic Shunt Compensation

Provides continuous control

In the steady state, the line performs as if it is

made up of independent sections

For example, with mid-point regulated

compensation, the line performs as if it is made
up of two independent sections
Po - natural load

Fig. 11.59 Power-angle relationship with and without mid-point

regulated compensation of a 600 km line ( = 44.7)

19

1538pk

Fig. 11.60 Performance of a 600 km line with an SVS regulating midpoint voltage

Fig. 11.61 Reactive power supplied by SVS as a function of transmitted

power
20

1538pk

Static Synchronous Compensator

(STATCOM)
z

Can be based on a voltage-sourced or current-sourced

converter

Figure below shows one with voltage-sourced converter

) driven by a dc voltage source: capacitor

Effectively an alternating voltage source behind a

coupling reactance
) controllable in magnitude

Can be operated over its full output current range even

at very low (typically 0.2 pu) system voltage levels

Requires fewer harmonic filters and capacitors than an

SVC, and no reactors
) significantly more compact

21

1538pk

Comparison of
STATCOM and SVC Characteristics
(a) V-I characteristics:

(b) P- characteristic with mid-point compensation:

Source: N.G. Hingorani and L. Gyugi, "Understanding FACTS", IEEE Press, 1999
22

1538pk

Comparative Summary of Alternative

Forms of Compensation
z

Switched shunt capacitor compensation

generally provides the most economical reactive
power source for voltage control
) ideally suited for compensation transmission
lines if reduction of ZC, rather than reduction of
line length is the primary consideration
) however, heavy use of shunt capacitor
compensation could result in poor voltage
regulation and may have an adverse effect on
system stability

Series capacitor is self-regulating, i.e., its

reactive power output increases with line loading
) ideally suited for applications where reduction of
line length () is the primary consideration
) improves voltage regulation and system stability

A combination of series and shunt capacitors

may provide the ideal form of compensation in
some cases
23

1538pk

Comparative Summary (cont'd)

z

A static var compensator (SVC) is ideally suited

for applications requiring direct and rapid control
of voltage
) has advantage over series capacitors where
compensation is required to prevent voltage sag at
a bus involving multiple lines; total cost may be
less than that for series compensation of each of
the lines

When an SVC is used to permit a high power

transfer over a long distance, the possibility of
instability when the SVC is pushed to its reactive
limit must be recognized
) when operating at its capacitive limit, the SVC
becomes a simple capacitor

An SVC has limited overload capability and has

higher losses than series capacitor compensation

STATCOM overcomes some of the limitations of

an SVC

24

1538pk

Tap-Changing Transformers
z

Transformer with tap-changing facilities

constitute an important means of controlling
voltages throughout the power system

Control of a single transformer will cause

changes in voltages at its terminals
) in turn this influences reactive power flow
) resulting effect on the voltages at other buses
will depend on network configuration and
load/generation distribution

Coordinated control of the tap changers of all

transformers interconnecting the subsystems
required to achieve overall desired effect

During high system load conditions, network

voltages are kept at highest practical level to
) minimize reactive power requirements
) increase effectiveness of shunt capacitors and
line charging
cont'd
25

1538pk

The highest allowable operating voltage of the

transmission network is governed by
) requirement that insulation levels of equipment
not be exceeded
) need to take into consideration possible
switching operations and outage conditions

During light load conditions, it is usually

required to lower network voltages
) reduce line charging
) avoid underexcited operation of generators

Transformers with under-load tap-changers

(ULTC) are used to take care of daily, hourly,
and minute-by-minute variations in system
conditions

Off-load tap-changing transformers used to

take care of long-term variations due to system
expansion, load growth, or seasonal changes

26

1538pk

Modelling of Transformer ULTC Control

Systems
z

Functional block diagram of ULTC control

system shown in Fig. 11.79 and block diagram
suitable for system studies

Line drop compensator regulates voltage at a

remote point along the line or feeder

Measuring element consists of adjustable dead

band relay with hysteresis. The output of the
measuring element is Vm; which takes a value of
0, 1, or -1, depending on input Verr

Time delay element prevents unnecessary tap

changes

Fig. 11.79 Functional block diagram of control system for automatic

changing of transformer taps

27

1538pk

Fig. 11.80 ULTC control system model

28

1538pk

Distribution System Voltage Regulation

z

Substation bus regulation

) substation transformer equipped with ULTC
facilities to control secondary voltage
) alternatively, substation may have a separate
voltage regulator

Feeder regulation
) feeder regulators control the voltage of each
feeder
) older units are the induction type - provide
accurate and continuous control; however, they
are costly and have been superseded by step
type regulator
) step voltage regulator (SVR) is basically an
autotransformer with taps or steps in the series
winding; however, it is purely a voltage control
device and not used for voltage transformation
cont'd

29

1538pk

Fig. 11.75 Schematic of an induction regulator

Fig. 11.76 Schematic of a step voltage regulator

Fig. 11.77 SVR control mechanism

30

1538pk

Distribution System Voltage Regulation

(cont'd)
z

Application of voltage regulators and

capacitors for control of voltage profile along
a feeder is illustrated in Fig. 11.78
) curve 1 shows voltage with distributed loads
along the line, without any regulation
) the addition of voltage regulator R1, capacitor
C and voltage regulator R2, brings the voltage
profile along the entire feeder (from the first
consumer to the last) to within max and min
limits

Fig. 11.78 Voltage profile of a feeder with a station regulation (R1),

supplementary regulator (R2) and a shunt capacitor bank (C)

31

1538pk

Implementation of Overall Reactive Power

Control
z

Effect of reactive power control is felt mostly

locally:
) equipment for supplying Q at appropriate points
throughout the system necessary

Coordination of the overall scheme a complex

task:
) approach is still largely based on operator
experience and off-line load flow studies
) implementation of automated schemes with
optimum dispatch is feasible and practical
methods are being pursued

EDF and ENEL have used secondary and

tertiary voltage control to provide coordinated
voltage control in HV networks
) CIGRE TF 38.02.23 set up to assess the potential
and provide guidelines

32

1538pk

4. TYPICAL SCENARIO OF
VOLTAGE INSTABILITY

Copyright P. Kundur
This material should not be used without the author's consent

1538pk

Practical Aspects of Voltage Collapse

z

When a power system is subjected to a sudden

increase of reactive power demand following a
system contingency
) the additional demand is met by the reactive
power reserves carried by generators and
compensators
) generally there are sufficient reserves and the
system settles to a stable voltage level

It is possible, because of a combination of events

and system conditions, that the additional
reactive power demand may lead to voltage
instability, causing a major breakdown of part or
all of the system.
) result of an accumulative process involving
actions and interactions of many devices,
controls and protective systems

The time frame by which voltage instability

occurs could be in the range of a few seconds or
tens of minutes
) may be a short-term or a long-term phenomenon
1

1538pk

Long-Term and Short-Term Voltage

Stability
z

Long-term voltage stability involves slower acting

equipment such as tap-changing transformers,
thermostatically controlled loads, and generator
field current limiters
) study period may extend to several minutes
) may be effectively studied using static analysis
techniques with complementary use of dynamic
analysis

Short-term voltage stability involves dynamics of

fast acting load components such as induction
motors, electronically controlled loads and HVDC
converters
) study period of interest is in the order of several
seconds
) dynamic modeling of loads often essential;
analysis requires solution of differential equations
using time-domain simulations
) particularly vulnerable to low inertia compressor
motors used for air conditioning heat pumps and
refrigerators
) faults/short circuits near loads could be important
2

1538pk

A Typical Scenario of
Long-Term Voltage Instability
z

The power system is experiencing abnormal

operating conditions with large generating units
near the load centres out of service
) some EHV lines are heavily loaded and reactive
power reserves are at a minimum

The triggering event is the loss of a heavily

loaded line
) causes additional loading on the remaining
adjacent lines
) reactive losses in the lines increase
) heavy reactive power demand on the system

Immediately, there would be a considerable

reduction of voltages at adjacent EHV buses
which would be reflected into the distribution
system
) causes load reduction, and the resulting reduction
in power flow through the EHV lines would have a
stabilizing effect
) generator AVRs quickly restore terminal voltages
by increasing excitation
) resulting additional reactive power flow through
the inductances associated with generator
transformers and lines would cause increased I2x
losses and voltage drops
3

1538pk

At this stage, generators are likely to be within

the reactive power output capabilities
) speed governors would regulate frequency by
reducing MW output.

ULTCs of substation transformers would restore

distribution voltages and loads to pre-fault levels
in about 2 to 4 minutes
) with each tap change operation, the resulting
increment in load on EHV lines increases the line
I2x losses, which in turn cause greater drop in
EHV voltages

If the EHV line is loaded considerably above the

SIL, each MVA increase in line flow causes
several MVArs of line losses.

1538pk

With each tap-changing operation, the reactive

output of generators throughout the system
would increase
) gradually, the generators would hit their reactive
power capability limits one by one
) when the first generator reaches its field current
limit, its terminal voltage will drop
) at reduced terminal voltage for a fixed MW output,
the armature current would increase
) this may further limit reactive output to keep the
armature current within allowable limits
) its share of reactive loading will be transferred to
other generators, leading to cascading field
current limiting and armature current overloads of
more and more generators
) with fewer generators on automatic excitation
control, the system is much more prone to voltage
instability. This is likely to be compounded by the
reduced effectiveness of shunt compensators at
low voltages.

The process will eventually lead to voltage collapse or

avalanche, possibly leading to a major blackout

1538pk

Overvoltage Instability
z

While the most common form of voltage

instability is the progressive drop of bus voltages,
the risk of overvoltage instability also exists

Caused under light load conditions

) EHV lines operating well below SIL
) underexcitation limiters limit Q absorbed by
generators
) instability caused by the inability of combined
generation and transmission system to operate
below certain load level
) transformer tap changers, attempting to restore
load voltage and power, cause long-term voltage
instability

One such situation reported in the paper:

T. Vancutsem and R. Mailhot, "Validation of Fast Voltage
Stability Analysis Method on the Hydro Quebec
System", IEEE Trans. on Power Systems vol. 12, pp.
282-292, February 1997

1538pk

A Typical Scenario of
Short-Term Voltage Instability
z

The power system is operating in a stressed

condition during hot weather with a high level of
air conditioning load

The triggering event is a multi-phase fault near a

load center
) causes voltage dips at distribution buses
) air conditioner compressor motors decelerate,
drawing high current

Following fault clearing with

transmission/distribution line tripping, motors
draw very high current while attempting to
reaccelerate
) motors stall if power system is weak

Under-voltage load rejection may not be fast

enough to be effective

Loss of much of the area load and voltage

collapse

1538pk

Short-Term Voltage Instability

z

Has not received much attention in the past

Growing industry problem

) many incidents have occurred in recent years:
Southern California on 5 August 1997; Atlanta,
Georgia on 30 July 1999

Factors contributing to this trend

) increasing use of low inertia compressor motors
for air conditioning, heat pumps and refrigeration
) growth in the use of voltage-insensitive loads with
electronic power supplies
) transmission network being pushed harder

Effective countermeasures:
) STATCOM's, particularly smaller units connected
to distribution network
) fast load shedding

1538pk

Voltage Stability Associated with HVDC

Links
z

Usually associated with HVDC links connected to

weak AC systems

May occur at rectifier or invester stations

) associated with the unfavourable reactive power
"load" characteristics of converters

HVDC link control strategies have a very

significant influence on such problems

Recent developments in HVDC technology, such

as voltage source converters and capacitor
commutated converters, have significantly
increased the limits of stable operation as
compared to those with line commutated
converters

Such a phenomenon is relatively fast

) time frame of interest in the order of a second

1538pk

5. METHODS OF VOLTAGE
STABILITY ANALYSIS

Copyright P. Kundur
This material should not be used without the author's consent

1538pk

Methods of Voltage Stability Analysis

Objectives of Analysis:

State: is the current system state stable or

unstable?

Proximity: if stable, how close to instability?

) operating and planning stability margins
) numerous measures possible: load level, power
flow, VAr reserve
) physical units most appropriate are system
dependent

Mechanism: in case of instability,

) what areas are involved?
) why does it happen?
) what factors contribute?
) what measures can prevent or control instability

1538pk

Methods of Analysis cont'd

Voltage instability is a dynamic phenomenon

which may involve the interaction of many
devices

Power system behaviour may be very

nonlinear

Instability may occur in different time frames

and involve different parts of the system

It is generally necessary to examine a wide

range of conditions and contingencies

Analysis must provide information on system

state, proximity to, and mechanism of
instability

The practical approach is to use a variety of

complementary methods to address different aspects of
the phenomenon and analysis requirements

1538pk

Methods of Analysis (cont'd)

Alternative Approaches:

Dynamic Analysis: considers differential

equations (time is explicitly handled)
) enhanced time-domain simulations

Steady-State Analysis: considers only algebraic

equations and time is handled implicitly
) powerflow based techniques

Quasi-Dynamic Analysis: considers algebraic

equations and time is handled explicitly
) fast time-domain simulations

1538pk

Dynamic Analysis

1538pk

Dynamic Analysis

Nonlinear time domain simulations

Uses numerical integration of differential
equations
dx/dt = f (x,V)
I (x,V) = YNV

Advantages:

Captures the events and chronology leading to

voltage instability

Accurately replicates the actual dynamics of

voltage instability

Performance of system and individual devices is

provided

1538pk

Dynamic Analysis (cont'd)

Disadvantages:

Substantial data requirements (beyond those for

transient stability simulations)

Long simulation times required

) even with state-of-the-art techniques, simulations
are time consuming
) multiple contingency/condition analysis generally
impractical

Detailed analysis and interpretation of results

required

Simulations do not readily provide sensitivity

information or the degree of stability
Applications:

Essential for studies involving the coordination of

controls and protections

Short-term voltage stability analysis

Post-mortem studies
Bench marking of simplified (steady-state)
analyses
6

1538pk

Example 14.2
In addition to illustrating the time-domain approach to dynamic
analysis of voltage stability, this example shows the effects of
generator overexcitation limiter (OXL), transformer tap changer
on voltage stability

Fig. E14.4 Test system

We will consider three system load levels:

Load level 1: 6655 MW, 1986 MVAr
Load level 2: 6755 MW, 2016 MVAr
Load level 3: 6805 MW, 2031 MVAr

ULTC for transformer T6 between buses 10 and 11:

Time delay for the first tap movement: 30 seconds
Time delay for subsequent tap movement: 5 seconds
Dead band:
1% pu bus voltage
Tap range:
16 steps
Step size:
0.00625 pu

1538pk

Over-excitation limiter for generator G3:

Ifdmax1 = 3.02 pu
ILIM = 3.85 pu

Ifdmax2 = 4.60 pu
K1 = 0.248

K2 = 12.6

Fig. E14.5 Block diagram of OXL

Fig. E14.6 OXL characteristic

1538pk

The disturbance considered is the loss of one of the

lines between buses 6 and 7 (without a fault).
Investigate the voltage stability of the system with the
following representations of loads at buses 8 and 11.

Case (a):
The load at bus 11 is modelled as 50% constant
impedance and 50% constant current for both active and
reactive components; the action of the ULTC transformer
(T6) supplying this load is modelled in detail, as shown in
Fig. 11.80.
The load at bus 8 is modelled as constant MVA for both
active and reactive components. The transformer T4
supplying this load is assumed to have a fixed tap.
Case (b):
The load at bus 8 is modelled with its active component
as an equivalent induction motor. All other components
modelled as in Case (a).

1538pk

Fig. 11.80 ULTC control system model

10

1538pk

Results of Case (a)

Load at as bus 11 modelled as 50% constant Z

and 50% constant I
) transformer T6 with ULTC

Load at bus 8 modelled as constant MVA

11

1538pk

Fig. E14.7 Voltage at buses 11, 10, and 7

12

1538pk

Fig. E14.8 Responses of generator G3 variables

13

1538pk

Results of Case (b):

Load at bus 11 modelled as in Case (a)

Load at bus 8 modelled as an equivalent in
induction motor of 3600 MVA rating

Only Load Level 2 considered

14

1538pk

Fig. E14.9 Induction Motor Response, with Load Level 2

15

1538pk

Fig. E14.10 Voltage magnitude at bus 7 and bus 8 with

induction motor load at bus 8; system at load level 2

Fig. E14.11 Response of voltage magnitude at bus 11 with (a)

constant MVA load and (b) induction motor load at bus 8;
system load at level 2
16

1538pk

Steady-State Analysis

Note: To follow the material in this sub-section, it is

useful to review power-flow analysis techniques
and procedures. Appendix at the end of this
section provides copies of related material from the
book "Power System Stability & Control"

17

1538pk

Steady-State (Static) Analysis

System dynamics influencing voltage stability are
often slow (there are exceptions)

Time frames can be identified in which snapshots

of system conditions can be captured and
analyzed by assuming

dx/dt = 0
with the state variables x set at values appropriate
to the specific time frame

Practical approach is based on powerflow

methods
Advantages:

Can provide insight into state, proximity to, and

mechanism of instability

Low CPU requirements

Well suited for studies involving a large number
of system conditions and multiple contingencies

cont'd
18

1538pk

Static Analysis (cont'd)

Disadvantages:

Time trajectory is not computed and therefore no

indication if snapshot condition can be reached

Modelling assumptions/simplifications required

) care should be exercised in retaining essential
mechanics

May be difficult to predict certain

characterizations of instability (fast dynamic
effects)
Applications:

Bulk of planning and operating studies in which

many system conditions/contingencies must be
examined

On-line system security analysis

Identifying causes of instability and selecting
remedial measures

19

1538pk

Static Analysis Methods

Methods discussed in this presentation:

Q-V Curves
P-V Curves using full powerflows
Modal analysis
Continuation powerflow
Methods not discussed here, but considered
in the literature:

V-Q sensitivities
Singular value decomposition
Shortest distance to instability
MVAR margin using optimization
Direct solution of the point of instability

20

1538pk

Q-V Curves
An approach used in early studies
Variable reactive source is placed at a specified bus
and the bus voltage is controlled through a range of
values to obtain a plot of bus voltage versus reactive
power injection

Curve shows:
) the reactive margin at a bus
) the voltage at which instability occurs (critical
voltage)
) the sensitivity of bus voltage to changes in reactive
load

Popular due to ease of implementing in powerflows

Full powerflow solutions are used. Nonlinearities
associated with generator reactive limits, tap changer
actions, and voltage dependency of loads can be
included

Many buses may have to be examined, and many

powerflow solutions are required for each bus
) choice of appropriate buses critical

The process must be repeated for each load level of

interest to find system MW or MVAR margin

Main disadvantage is that it stresses the system in an

unrealistic manner by injecting Q at one bus
21

1538pk

Example of Criteria Based on Q-V Curves

Used by Some Utilities
a) Reactive power margin "A" between the
voltage stability limit (bottom of curve) and
operating point
) 4.0 MVAr per 100 MW power transfer to the
region

b) Voltage margin "C" between the voltage

stability limit and the operating point
) greater than 5%

c) The bus voltage corresponding to the stability

limit must be less than 0.95 pu

Note: This is not necessarily the best approach with

currently available tools

22

1538pk

Fig 3.30 Q-V curve analysis criteria

23

1538pk

P-V Curves
Load and generation in selected areas are
increased/changed in a predetermined manner to
find the distance to voltage instability

Full powerflow solution is performed at each load

level to obtain bus voltages

PV plots show
) variation (sensitivity) of bus voltages (or other
variables) with load
) distance to instability (VS margin)
) voltage at which instability occurs

24

1538pk

P-V Curves (cont'd)

Full powerflow solution ensures all system
nonlinearities are represented as the system is
stressed

Stressing the system by load/generation increase is

the most relevant measure for assessing the voltage
stability of the system

Computed VS margins are in physical terms (e.g. MW

load increase)

Implementation requires various generation dispatch

options to meet the increasing load

Concern is that the failure of powerflow to converge

may be the result of numerical and algorithmic
problems rather than the actual instability being
reached

Experience has shown that the Fast Decoupled

method (when properly implemented) can solve the
powerflow very close to the instability point (by
gradually decreasing the step size)

Continuation Powerflow method can be used when

convergence problems experienced or if it is desired
to compute full P-V curve, including the lower
portion
25

1538pk

Q-V Modal Analysis

Powerflow Jacobian matrix, including the enhanced
models for generators, static loads, induction motors,
SVCs, HV dc links and other devices
) the device models are based on linearized dynamic
equations with dx/dt = 0

Steady State Jacobian Matrix

P = J P
Q JQ

J PV
JQV V

where
P = incremental change in bus real power injection
Q = incremental change in bus reactive power injection
= incremental change in bus voltage angle
V =incremental change in bus voltage magnitude

Let P = 0,

Q = JQV JQ1 J PV V = J R V
and

V = J R1 Q

JR is called the reduced steady state Jacobian matrix

of the system, and represents the linearized

relationship between the incremental changes in bus
voltage magnitude (V) and bus reactive power
injection (Q).
26

1538pk

 At each operating point we assume that P does not

vary and consider the incremental relationship
between Q and V

JR is an important concept and it provides a

convenient platform for system voltage stability

evaluation

Eliminating the real power and angle part from the

linearized steady state system equations allows us
to focus on the study of the reactive demand and
supply problem of the system as well as minimize
computational effort

If the full Jacobian is used, the results represent the

relationship between (, V) and (P, Q). Since
is included in the formulation , it is difficult to
discern the relationship between V and (P, Q)
which is of primary importance for voltage stability
analysis

Although the incremental changes in P are neglected

in the formation, the effects of changes in system
load or power transfers are considered by studying
the V-Q relationship at different operating
conditions
) similar in concept to computing Q-V curves for
different operating conditions

27

1538pk

Modal Analysis

Voltage stability characteristic analyzed by

computing eigenvalues and eigenvectors of JR:

JR =
Where
= right eigenvector matrix of JR
= left eigenvector matrix
=diagonal eigenvalue matrix
Q = JR V
= V

Rearranging,

1 Q = V

28

1538pk

 For a 3 bus example,

1 1
11 12 13 Q1 11 12 13 V1

21 22 23 Q2 = 21 22 23 V2
1 2

1 3 31 32 33 Q3 31 32 33 V3

For mode 1:
1
( 11 Q 1 + 12 Q 2 + 13 Q 3 ) = ( 11 V 1 + 12 V 2 + 13 V 3 )
1

Or

1
q1 = 1
1

where
q1= 1Q = mode 1 reactive power variation
1 = 1V = mode 1 voltage variation

If i > 0, the ith modal voltage change is positive for a

positive change in ith modal reactive power

Stability:
> 0 stable mode
< 0 unstable mode

The magnitude of eigenvalue can provide a relative

measure of the proximity to instability

29

1538pk

Modal Analysis
Coordinate transformation so that:
) reactive power changes in a given direction
result in proportional bus voltage changes in the
same direction

q- variations are decoupled

When the system is voltage unstable
) individual modes are unstable

30

1538pk

Participation Factors

To gain an understanding of the mechanism of

voltage instability, the participation of the
following elements in the critical mode(s) are
computed:
) buses, branches and generators

For each of these elements, a participation factor

in a given mode is computed using the right and
left eigenvectors of JR

Participation is an indication of the degree to

which each element is associated with a mode

31

1538pk

Bus Participation Factors

The relative participation of bus k in mode i is given
by the bus participation factor:

Pki = kiik
Pki determines the contribution of i to the V-Q
sensitivity at bus k

Bus participation factors determine the areas

associated with each mode

The sum of all the bus participations for each mode

is equal to unity because of the way right and left
eigenvectors are normalized

The size of bus participation in a given mode

indicates the effectiveness of remedial actions
applied at that bus in stabilizing the mode

There are generally two types of modes

) first type has very few buses with large
participations and all the other buses with close to
zero participations, indicating that the mode is very
localized
) second type has many buses with small but similar
degree of participations, and the rest of the buses
with close to zero participations; this indicates that
the mode is not localized
32

1538pk

Branch Participation Factors

Let us compute the branch participation factor
associated with mode i by assuming that the vector of
modal reactive power variations q has all elements
equal to zero except for the ith, which equals 1.
The corresponding vector of bus reactive power
variations is

Q (i ) = 1q = q = i
where i is the ith right eigenvector of JR. We further
assume that all the right eigenvectors are normalized so
that

2
ji

=1

With the vector of bus reactive power variations equal

to Q(i), the vector of bus voltage variations, V(i), is
V (i ) =

1
Q (i )
i

and, the corresponding vector of bus angle variation is

(i ) = JP1 JPV V (i )

33

1538pk

 With the angle and voltage variations for both the

sending end and receiving end known, the
linearized change in branch reactive loss can be
calculated.

The relative participation of branch j in mode i is

given by the participation factor:
Pji =

Qloss for branch j

maximum Qloss for all branches

Indicate, for each mode, which branches consume

the most reactive power in response to an
incremental change in reactive load

Branches with participations are either weak links

or heavily loaded branches

Useful for identifying:

) remedial measures to alleviate voltage stability
problems, and
) critical contingencies

34

1538pk

Generator participation factors

For a given reactive power variation, voltage and

angle variations are determined at each machine
terminal. These in turn are used to compute the
change in reactive power output for each
machine.

The relative participation of machine m in mode i

is given by the generator participation factor:
Pmi =

Q m for machine m
maximum Q for all machines

Indicate, for each mode, which generators supply

the most reactive power in response to an
incremental change in system reactive loading

Provide important information regarding proper

distribution of reactive reserves among all the
machines in order to maintain adequate voltage
stability margin

35

1538pk

Illustration of Modal Analysis

Consider the 39 bus, 10 machine system of Fig. 14.7.
Now perform modal analysis at the three operating
conditions represented by points A, B, and C on the V-P
curve of Fig. 14.4
Table 14.1 Five smallest eigenvalues

Operating point

0.3867

0.1446

0.0083

1.0271

0.5550

0.3209

2.4049

1.5133

0.9334

4.1031

2.6280

1.8757

4.2699

3.0209

2.3373

Fig. 14.4 V-P curve at bus 530 of system shown in Fig. 14.3
36

1538pk

Fig. 14.7 Buses and branches with high participation in

the least stable mode

37

1538pk

Table 14.2 Bus, branch, and generator participations in

the least stable mode for operating point C

Bus participation

Branch participation

Generator participation

Bus no.

Participation

Branch

Participation

Bus no.

Participation

530

0.2638

500-520

1.0000

1311

1.0000

520

0.2091

300-360

0.8414

2412

0.2786

510

0.1025

100-350

0.8175

1011

0.2103

500

0.0941

320-500

0.8093

1014

0.2036

320

0.0482

330-350

0.6534

1013

0.2036

310

0.0319

1012

0.2036

300

0.0296

340

0.0279

38

1538pk

The Continuation Power Flow

The powerflow Jacobian matrix becomes singular

at the voltage stability limit

Conventional powerflow algorithms are prone to

convergence problems at operating conditions
near the stability limit

The continuation powerflow overcomes this

problem by reformulating the powerflow equations
so that they remain well-conditioned at all possible
loading conditions

Allows the solution of the powerflow problem for

stable as well as unstable equilibrium points

Method described is based on work by Ajjarapu

and Christy
) uses a locally-parameterized continuation method
) belongs to a general class of methods for solving
nonlinear algebraic equations known as path-

following methods

39

1538pk

Basic Principle

Uses an iterative process involving predictor and

corrector steps

40

1538pk

 From a known initial solution (A), a tangent

predictor is used to estimate the solution (B) for a
specified pattern of load increase

The corrector step then determines the exact

solution (C) using a conventional powerflow
analysis with the system load assumed to be fixed

The voltages for a further increase in load are

then predicted based on a new tangent predictor

If the new estimated load (D) is now beyond the

maximum load on the exact solution, a corrector
step with loads fixed would not converge;
therefore, a corrector step with a fixed voltage at
the monitored bus is applied to find the exact
solution (E)

As the voltage stability limit is reached, to

determine the exact maximum load, the size of
load increase has to be reduced gradually during
the successive predictor steps

41

1538pk

Sensitivity information

The elements of the tangent vector represent

differential changes in the state variables in
response to a differential change in system load

Therefore, the dV elements in a given tangent

vector are useful in identifying "weak buses", that
is, buses which experience large voltage
variations in response to a change in load
Complementary use of conventional and continuation
methods
Continuation method of powerflow analysis is
robust and flexible

Ideally suited for solving powerflow problems

with convergence difficulties
) however, the method is very slow and timeconsuming

The best overall approach for computing

powerflow solutions up to and beyond the critical
point is to use the conventional and continuation
methods of powerflow analysis in a
complementary manner

42

1538pk

Quasi-Dynamic Analysis

43

1538pk

Quasi-Dynamic Analysis

Based on full time-domain simulations

Fast dynamics are ignored - equations are
algebraic

Simulation steps through time and accounts for

chronology of events triggered by devices such
as switch shunts, OXLs, ULTCs, etc.

Advantages

Compromise between dynamic and steady state

analysis

Fast but also includes "dynamic performance

through time"

44

1538pk

Fast Simulation
Basic Principles and Assumptions
Focus on the evolution of system operating
conditions driven by the following slow dynamics
) change in load
) ULTC operations
) generator field current limiters
) switching of capacitors and reactors
) automatic generation control

Dynamics of all the other devices are assumed to

be so fast that their effects are adequately
captured by solving the post-contingency steady
state equations for each device. The devices
whose dynamics are neglected include:
) electro-mechanical transients of synchronous
machines
) exciters, stabilizers, and governors
) induction motors
) HVDC links
) SVCs and other FACTS devices

45

1538pk

Treatment of Governor Responses

Governor responses are fast dynamics

) only the change in mechanical power at the steady
state is considered

The adjustment of mechanical power is done

during an outer loop as described below
1) the nonlinear algebraic equations are first solved
without changing the mechanical power. The
change in active load and losses are picked up by
the system slack bus(es)
2) the mechanical power of each synchronous
machine is adjusted based on the change in active
load and losses, and the governor parameters
3) the non-linear algebraic equations are resolved
after adjusting the generator mechanical power
4) steps 2 and 3 are repeated until the change in the
mechanical power of the slack bus(es) becomes
less than a specified tolerance

46

1538pk

Mathematical Formulation

The complete set of differential/algebraic

equations of a power system has the following
general form:
X = f (X ,V , Z )

YV = I (X ,V , Z )

Where:
X
V
I
Y
Z

=
=
=
=
=

state vector
bus voltage vector
current injector vector
network admittance matrix
variables associated with the slow
control devices including ULTCs,
loads, switchable reactors and
capacitors, and field current limiters

47

1538pk

 At each equilibrium point, Z=Zi and the system

operating condition is obtained by solving the

following set of nonlinear algebraic equations:

0 = f (X,V, Z i )
YV = I (X,V, Z i )
As time progresses, the slow control devices
operate and the values of Z change. The above
set of nonlinear algebraic equations is solved
every time the values of Z change.

Simulation is controlled by the dynamics of

discrete-time control devices such as ULTCs,
field current limiters, switchable shunts,
protective relays.

48

1538pk

49

1538pk

Practical Approach to Computation of

Voltage Stability Margins

Effective use of P-V curves

Limitations of Q-V curves

50

1538pk

Practical Approach for Computation of

Voltage Stability Margin
The most direct and relevant measure of VS
margin:
) MW load, generation or power transfer margin as
shown by P-V curves

Q-V curves have been used in the past by many

investigators to identify weak points on the
system and to measure relative margin to
instability. This has a number of limitations:
) Q-V curves do not generally reveal the voltage
stability of the system and true weak spots, unless
they are computed not only for every bus in the
study area and for every contingency but also for
different operating points along the P-V curves

On the other hand, P-V curves directly reveal the

margin to instability in terms of the most relevant
and measurable quantities

In addition, Modal Analysis at or close to the nose

of the P-V curves readily identify the "weak
buses" and mode of instability.

51

1538pk

'd

 The weak buses are the ones that are prone

to voltage instability due to lack of reactive
support
) Weak buses are not necessarily the ones
with the lowest voltages

In some cases, the weak region is confined to

a small number of buses
) Controlling this region (isolating it, shedding
some of its load, or adding appropriate VAR
compensation) can prevent instability

In other cases, the weak region could be

spread over a wide area
) Widespread control actions may be
necessary to prevent instability

cont'd

52

1538pk

 To compute VS margin, the generation and load in

some parts of the system are appropriately
changed in discrete steps.
) At each step, after solving pre-contingency
powerflow, contingencies are applied one by one
) For each contingency, the highest load/power
transfer level that results in a stable postcontingency powerflow solution is the stability
limit for that contingency
) The distance between this limit and the initial
load/transfer level is the VS margin for that
contingency as shown in Fig. A.

Fig. A: P-V curves and VS margin

53

1538pk

 It is evident that the stability margin is

independent of which bus voltage is monitored
and plotted in the P-V curve
) the points on the curves for different buses will be
different only in their V values, but their P values
will be the same

On the contrary, in Q-V curves the injected

reactive power and its MVAr stability margin is
strictly applicable to the bus whose voltage is
monitored and plotted

Different P-V margins may be computed choosing

different patterns of load, generation or power
transfer increases
) each of these is a meaningful measure of VS
stability, relevant for a particular situation

The MW margin computed may be supplemented

by MVAr reserve on generating units within the
appropriate voltage control area applicable to the
mode of voltage instability.

To identify situations resulting in stalling and

tripping of motors, dynamic simulations need to
be carried out.

54

1538pk

Comprehensive Assessment
of Voltage Stability for
Power System Planning and
Operation

55

1538pk

Comprehensive Assessment of Voltage

Stability
In addition to providing for good modelling capability
and use of analysis techniques which compute the
necessary stability related information, the following
capability is essential for comprehensive
assessment:

Reliable and efficient procedures for contingency

screening and ranking

Automated procedures for changing system

operating conditions and power flow dispatching

Capability for identifying and ranking appropriate

remedial measures for ensuring sufficient stability
margin consistent with the system design and
operating criteria

56

1538pk

Key Elements of VSA

Model Initialization;
Interface with EMS for on-line application

Contingency screening and selection

Determination of secure operating region
) using static analysis

Determination of remedial actions

Fast time-domain simulation
) validation and checking

57

1538pk

Contingency Selection Module

Impractical to consider every conceivable

contingency

A limited number (typically 20) critical

contingencies determined for detailed studies

Performance Indices based on a few power flow

solutions and reactive reserve not reliable

A fast screening method used:

) based on exact margin to voltage collapse and full
power flow solutions
) number of power flow solutions 1.2 to 2.0 times
number of contingencies

Supplemented with user-specified contingencies

58

1538pk

Fig. 4 Automatic Critical Contingency Selection

59

1538pk

Security Computation Module

Engine for voltage stability analysis

) static analysis with detailed models

Secure region is defined by a number of

Coordinates (SRCs)
) key system parameters: MW generation, area
load, interface transfers, etc.

Voltage stability assessment:

) stability limit/secure operating region
) MVAr reserves of key reactive sources
) post-contingency voltage decline

Specialized powerflow dispatcher and solver to

quickly search for stability limit

60

1538pk

Modelling:
) generator capability curves
) governor response, economic dispatch, AGC
) nonlinear loads
) control of ULTCs, switched shunts, etc.

Inputs and Outputs:

Inputs
) list of contingencies produced by screening and
ranking (+user defined)
) base case powerflow from state estimator
) definition of SCRs
) voltage security criteria and definition of
parameter of stress

Output
) secure region in secure region space

61

1538pk

Secure Operating Region

62

1538pk

Secure Operating Region

63

1538pk

Remedial Measures Module

Determines necessary remedial measures to

) ensure sufficient stability margins
) expand the secure region

Preventative control actions:

) taken prior to a contingency
) caps/reactor switching, generation redispatch,
voltage rescheduling

Corrective (emergency) control actions:

) applied following a contingency
) load shedding, generator runback, transformer
tap changer blocking

Ranking of each remedial measure using

sensitivity analysis

64

1538pk

Ranking and Applying Remedial

Measures
Objective is to identify the most effective
remedial measures to give the desired stability
margin

Obtain solved power flow case for the most

severe contingency
) gradually introduce the effect of the contingency
) bus injection compensation technique

Compute the sensitivities of reactive power (or

bus voltage) to different control measures
) rank the remedial measures

Apply controls one at a time in order of ranking

until power flow solves for the most severe
contingency

65

1538pk

Expanding the Secure Region: Remedial

Measures

66

1538pk

Fig. 3 VSAT Structure

67

1538pk

VSAT main window after completed run

Plot and table of security limits of two-dimensional transfer

68

1538pk

Example of Computational Performance

of VSAT
Computation times for analyzing a power system
with 4655 buses, 156 generators using a 1.7 GHz
Pentium 4 PC with 256 MB memory:

Screening 300 contingencies to select 20 critical

contingencies: 20 secs

Detailed security analysis of base case with 20

critical contingencies: 1.2 secs

One transaction limit search with 20 critical

contingencies: 12 secs

69

1538pk

Appendix to Section on Methods of

Voltage Stability

Copies of:
1. Sections 6.4 and 11.3 from the book "Power
System Stability & control"
provide background information related to power
flow analysis techniques and procedure

2. Paper "Voltage Stability Analysis using Static and

Dynamic Approaches"
3. Paper "Towards the Development of a Systematic
Approach for Voltage Stability Analysis"

70

1538pk

71

1538pk

72

1538pk

73

1538pk

74

1538pk

75

1538pk

76

1538pk

77

1538pk

78

1538pk

79

1538pk

80

1538pk

81

1538pk

82

1538pk

83

1538pk

84

1538pk

85

1538pk

86

1538pk

87

1538pk

88

1538pk

89

1538pk

90

1538pk

91

1538pk

92

1538pk

93

1538pk

94

1538pk

95

1538pk

96

1538pk

97

1538pk

98

1538pk

99

1538pk

100

1538pk

101

1538pk

102

1538pk

103

1538pk

104

1538pk

105

1538pk

106

1538pk

107

1538pk

108

1538pk

109

1538pk

110

1538pk

111

1538pk

6. PREVENTION OF
VOLTAGE INSTABILITY

Copyright P. Kundur
This material should not be used without the author's consent

1538pk

Voltage Instability Prevention

1. System Design Measures

2. System Operating Measures

3. Methods of Identifying Causes of

Instability and Selecting Remedial
Measures

4. Coordinated Voltage control Schemes

1538pk

System Design Measures for Prevention

of Voltage Instability

a) Application of VAR compensating devices:

) ensure adequate stability margins by proper
selection of schemes
) selection of size, rating and location based on
detailed study
) design criteria based on maximum allowable
voltage drop following a contingency not
satisfactory
) stability margins should be based on MW and
MVAr distances to instability
) important to recognize voltage control areas and
weak boundaries

1538pk

Voltage Instability Prevention (cont'd)

b) Control of network voltage and generator reactive

output:
) Load (or line drop) compensation of AVRs has, for
many situations, a beneficial effect
) Alternatively, secondary outer loop control of
generator excitation may be used to regulate
network side voltage. This should be much slower
than normal AVR action
) Several utilities have developed and used special
schemes for coordinated voltage control.
For example, EdF and ENEL use secondary and
tertiary voltage control schemes for controlling
network voltages and generator VAR output
Tokyo Electric Power has an adaptive control of
reactive power

1538pk

Voltage Instability Prevention (cont'd)

c) Control of transformer tap changers:

) Can be controlled, either locally or centrally, so as
to reduce the risk of voltage collapse
) Where tap changing is detrimental, a simple
method is to block tap changing when the source
side sags, and unblock when voltage recovers.
Such schemes have been used by several utilities.
A more sophisticated approach is to reduce the
voltage set point, resulting in a larger load relief

) There is potential for improved control strategies

based on a knowledge of load characteristics:
residential loads provide, at least temporarily,
load relief when secondary voltage depressed
industrial loads are not affected; however, caps
provide more VARs if voltage increased

) Microprocessor-based ULTC controls offer

virtually unlimited flexibility for implementing
control strategies

1538pk

Voltage Instability Prevention (cont'd)

d) Under-voltage load shedding

) used to cater for unplanned or extreme situations
) analogous to under-frequency load shedding,
which is widely used to cater for generation
deficiency
) provides a low-cost means of preventing
widespread system collapse
) particularly attractive if conditions leading to
voltage instability are of low probability but the
consequences are high
) characteristics and locations of loads to be shed
are more important for voltage problems than for
frequency problems
) should be designed to distinguish between faults,
transient voltage dips, and low voltage leading to
voltage collapse

1538pk

Voltage Instability Prevention (cont'd)

e) Coordination of protections/controls
) lack of coordination has been the cause of many
voltage collapse incidents
) adequate coordination should be ensured based
on dynamic simulation studies
) tripping of equipment to protect from overloaded
condition should be the last resort
Where possible, adequate control measures (auto
or manual) should be provided for relieving
overload conditions

1538pk

System Operating Measures for Prevention

of Voltage Instability
a) Stability margin:
) system should be operated with adequate stability
margin by appropriate scheduling of VAR sources
and voltage profile
) the required size of margin and system parameters
used as indices are system dependent
) if required margin cannot be met by VAR
scheduling alone, it may be necessary to limit
power transfers and start up additional generation
near critical areas

b) Spinning reserve:
) adequate spinning VAR reserve must be ensured
by operating generators, if necessary, at moderate
or low excitation, and switching in shunt caps
) required reserve must be identified and
maintained for each "voltage control area"

1538pk

Voltage Instability Prevention cont'd

c) Operators' action:
) operators must be able to recognize voltage
instability related symptoms and take appropriate
actions: voltage and power transfer control, and
load curtailment
) operating strategies that prevent voltage collapse
need to be established
) on-line monitoring and analysis to identify
potential problems and remedial measures would
be invaluable

d) Starting of Gas Turbines:

) may be important means of preventing voltage
collapse when available at the right locations
) can be started quickly (minutes)
) provide active and reactive power to relieve stress
on the system
) more importantly, provide voltage support/control
at additional buses

1538pk

Methods of Identifying Causes of Instability

and Selecting Remedial Measures
Modal analysis identifies:
) individual voltage instability modes
) critical voltage control areas
) bus, branch, and generator participations

Helpful in determining causes of voltage

instability and identifying most effective remedial
actions

Branch participation factors indicate which

branches consume the most reactive power in
response to an incremental change in load.
Branches with high participations are
) weak links, or
) heavily loaded branches

Useful in identifying system enhancements:

upgrading or adding lines, series capacitor
compensation, etc.

1538pk

Methods of Identifying Causes (cont'd)

Generator participation factors indicate which

generators supply the most reactive power in
response to an incremental change in system
reactive loading
) provide information regarding proper distribution
of reactive reserves among all machines to
maintain adequate VS margin

Bus participation factors identify areas associated

with each mode.
In addition, the magnitude of bus participation
indicates the effectiveness of remedial actions
applied at that bus
) helpful in identifying best locations for SVCs,
STATCOMs, switched shunt caps, etc.

10

1538pk

Example 1
Because of delays in obtaining approval to build
500 kV transmission lines, in the mid-1980s the
Ottawa area in Ontario, Canada was supplied
largely by 230 kV transmission

Loss of a critical 230 kV circuit under heavy load

periods could result in voltage collapse of the
area

A coordinated scheme consisting of the following

used as a stop-gap measure to prevent voltage
collapse:
) fast auto reclosure of major 230 kV circuits
supplying the area
) automatic load rejection
) automatic switching of shunt capacitors
) automatic blocking of transformer load tap
changers (LTC)

Fully automated scheme

) triggered by level and duration of voltage drop at
buses in the area

Area peak load 1400 MW

11

1538pk

12

1538pk

Coordinated Scheme:

Fast (0.8 s) auto reclosure of 230 kV circuits:

) used as first measure to prevent collapse
) if reclosure successful, other measures to prevent
voltage collapse not triggered

Automatic load rejection:

) 9 blocks of loads can be selected for rejection of
up to 600 MW of load
) voltage and time dependent
) armed by operator action depending on flow into
Ottawa area

Automatic capacitor switching:

) total of 36 capacitor banks in 17 TSs can be
switched in staggered blocks ranging from 1.8 s to
8.0 s

Automatic transformer LTC Blocking:

) implemented at 14 TSs
) blocked and then unblocked depending on voltage
and time

13

1538pk

14

1538pk

15

1538pk

Ottawa Area Load Characteristics:

Voltage stability is sensitive to load

characteristics

Load models used for LF and time-domain studies

were based on measurements on two transformer
stations in the Ottawa area, one supplying
predominantly commercial loads and the other
predominantly residential loads

For time-domain simulations representation of the

dynamic characteristics of the loads was found to
have a significant effect

16

1538pk

17

1538pk

Example 2

Example of Selection of System Enhancements for

Increasing VS Margins

System with about 2300 buses, 450 generators

representing Ontario Hydro and the
interconnected system

Fig. 1 shows one line diagram of a portion of the

system representing the region of interest

Objective is to identify remedial measures to

increase post-contingency VS margins,
) taking into consideration future load growth

The contingency consists of:

) loss of a double circuit 500 kV line (shown as line
A in Fig. 1) carrying 2500 mW, and
) rejecting two generators (one at Plant G and one at
Plant H) with a total output of about 1500 MW

VS margin is defined as total load growth in the

"area of load increase" (see Fig. 1)
Base load in the area is 8850 MW
18

1538pk

Fig. 1 One Line Diagram of the Internal System

Fig. 2 Voltage Stability Margins

19

1538pk

 From Fig. 2, VS margin:

) for pre-contingency case is 12100-8850 = 3250 MW
) for post-contingency case is 10430-8850 = 1580 MW

For system conditions corresponding to point B on

Fig. 2, system is unstable as indicated by
eigenvalue 1 = -0.0105

Buses with high participation are identified as

"critical mode" in Fig. 1

Alternative remedial measures considered:

) shunt compensation: an SVC
) series caps compensation

20

1538pk

a) Remedial Measure - an SVC of 500 MVAr:

) based on bus participations, bus 5581 is selected
as best location
has highest participation among 230 kV buses

) effect of SVC at different locations

SVC Location

Bus Participation

VS Margin (MW)

No SVC

--

1580

Bus 5581

0.01305

2180

Bus 6580

0.00307

1820

Bus 5660

0.00026

1800

Bus 1583

0.00001

1600

Bus 5995

0.00001

1600

Bus 6180

0.00001

1600

Bus 6391

0.00001

1600

) VS margin increased by 600 MW when SVC is

installed at bus 5581

21

1538pk

b) Remedial Measure - Series Caps:

) based on branch participations, the two circuits of
the 500 kV line D are selected for addition of series
caps
with 30% compensation, VS margin increased by
680 MW

) For comparison, compensation of other 500 kV

circuits considered:
Line D:

5525 - 5537, circuit 1, 187 km, 30%

5524 - 5537, circuit 1, 187 km, 30%

Line B

5506 - 5519, circuit 1, 179 km, 15%

5506 - 5519, circuit 2, 179 km, 15%
5506 - 5519, circuit 3, 179 km, 15%
5506 - 5519, circuit 4, 179 km, 15%

Line C

5531 - 5533, circuit 1, 209 km, 60%

) Voltage stability margins with above:

Line
none
Line D
Line B
Line C

Branch Participation
-1.0
0.567
0.039

22

VS Margin
1580 MW
2260 MW
1600 MW
1580 MW

1538pk

Example 3

Objectives
) Demonstrate VS assessment procedures using
static analysis (VSTAB) and time domain dynamic
analysis (ETMSP)

System
) Based on 94 winter peak condition (5348 buses
and 634 generators) of WSCC system
) BC Hydro system (1058 buses, 74 generators)

Procedure
) Use VSTAB to calculate VS margins for the base
case and all the contingency cases. The critical
contingencies which do not meet VS margin
criterion are identified
) Bench mark VSTAB results using ETMSP
) Design remedial measures using VSTAB and
ETMSP for the critical contingencies

Criterion assumed:
) System must be voltage stable for all
contingencies at BCH load level of 7% above the
base case
23

1538pk

Identification of Critical Contingencies

(VSTAB)
All 230 kV and above single contingencies
considered (74)

Load increased uniformly in BC Hydro

Active generation within BCH scaled up (could be
dispatched)

Device Modelling
Device
Loads
ULTCs

Generators

VSTAB Model

ETMSP Model

Study Area: P V0.5, Q V

External Area: constant MVA
Deadband: 1%
Tap size: 0.625%

All the generators in the study

area and some in BPA are
represented with capability
curves.
All the other generators are
represented by PV buses with
infinite reactive capabilities.

24

Study Area: same as VSTAB

External Area: P V0.1, Q V0.1
Deadband: 1%
Tap size: 0.625%
Initial delay: 30 seconds
Subsequent delay: 5 seconds
All the generators with
capability curves in VSTAB are
represented with OELs.
All the other generators are
represented as either classical
machines with small X'd or
detailed machines with AVRs
but no OELs

1538pk

BC Hydro Case Study

Identification of Critical Contingencies
(VSTAB)

No.

Branch Lost

VS Margin

TIR500-N to DMR500

250 MW (3.3%)

TIR500-S to DMR500

250 MW (3.3%)

CKY500 to KLY500

450 MW (5.9%)

MDN500 to NIC500

450 MW (5.9%)

All the other contingencies have VS margin> 600 MW (7.9%)

Four contingencies do not meet VS criterion

25

1538pk

PV Curves for the Pre and Post Contingency Systems

(Contingency: CKY500 to KLY500)

26

1538pk

VS Margin Calculation (ETMSP)

The contingency case of losing the 500 kV line

from CKY500 to KLY500 simulated using ETMSP

VS margin calculated using ETMSP = 350 MW

(margin calculated using VSTAB = 450 MW)

Comparison of ETMSP and VSTAB

) 100 MW difference on 7573 MW total load
) contributing factors
AVR droop (included in future VSTAB)
chronology
governor action
dynamics

27

1538pk

Time Response of Bus Voltage at ING500

(Load Increase of 300 MW Stable, 400 MW Unstable)

28

1538pk

Time Response of Field Current and Terminal Voltage for

Machine PCN G1 (300 MW Load Increase, Stable)

29

1538pk

Time Response of Field Current and Terminal Voltage for

Machine PCN G1 (400 MW Load Increase, Unstable)

30

1538pk

Design of Remedial Measures

Modal analysis conducted at the nose point of the

PV curve for the contingency case of losing the
line CKY500 to KLY500

100 MVAR of SVC added at Bus 151 (POW132) to

increase post-contingency VS margin to 7%
) This bus has the highest voltage level among the
top 10 buses with the highest participation to the
critical mode

Post contingency VS margins with and without

SVC
) without: 450 MW (5.9%)
) with: 530 MW (7%)

Adding 100 MVAR SVC at the high voltage bus

with high participation to the critical mode
extends the post-contingency VS margin to meet
the specified criterion

31

1538pk

Post-Contingency System PV Curves

(with and without SVC added)

32

1538pk

Coordinated Voltage Control Schemes:

Secondary and Tertiary Voltage Control

An effective way to control voltage and manage
reactive power resources in a transmission
network is to use a hierarchical control scheme
) results in improved efficiency, security and
flexibility of power system operation
) particularly attractive for highly meshed and
interconnected networks
) enhance Voltage Stability by making best use of
VAr resources

Network voltage control function divided into

three levels of hierarchy: primary, secondary and
tertiary
) the three levels of control are "de-coupled" in time
and space
) time scale separation ensures that the three levels
of control mechanisms do not adversely interact
with each other

Such schemes have been developed and used by

several European countries: France, Italy,
Belgium and Spain

33

1538pk

Coordinated Voltage Control (cont'd)

a) Primary Voltage Control:

Relates to automatic control actions on individual
equipment based on local measurements

Equipment utilizing primary control:

) Generators
) Dynamic VAr compensators: SVC, STATCOM,
synchronous condensers
) Switched capacitors and reactors
) Transformer with under load tap changers (ULTC)

Such controls normally used on all power

systems

34

1538pk

Coordinated Voltage Control (cont'd)

b) Secondary Voltage Control (SVC):

Transmission network conceptually separated into
"non-interacting" areas
) low sensitivity of bus voltages in one area to
generator Q in other areas
) voltages within each area independently controlled

Q-V resources within each "coherent" area

controlled in a coordinated manner in a time scale
ranging from one to a few minutes
) control aims to maintain system security and
economy/efficiency operation: enhance voltage
stability and minimize losses

Secondary Voltage control automatically adjusts the

reactive power resources, particularly Q output of
generators, to control voltage at a specific point in
the area referred to as the "pilot node"
) voltage at pilot node considered representative of
the voltages at all points in the area
cont'd
35

1538pk

 Processing all the needed information requires

monitoring of the appropriate system variables in
the area and performing the necessary analysis to
achieve the desired objectives.
) For example, EDF application on the French
system uses a sensitivity matrix model
implemented on computers at Regional Control
Centers

System Security requirements take into

consideration criteria contingencies and reserve
requirements

In effect, Secondary Voltage control performs

real-time adjustment of primary control reference
points and manages control resources to best
satisfy the current system operating condition
and needs

Additional control functions that may be included

as part of Secondary Voltage Control to enhance
system security:
) generator start-up/shut-down, distribution voltage
reductions, transformer ULTC blocking, and load
shedding

36

1538pk

Coordinated Voltage Control (cont'd)

c) Tertiary Voltage Control:

Objective is to achieve voltage control
coordination on a more global scale: at the utility,
pool or country level
) based on real-time measurements

Operates in a time scale of 10 minutes or higher

Determines the optimal voltage profile of the
network and operates on the set points of the
Secondary Voltage Control

A good source of reference on the topic is the report

prepared by CIGRE TF C4.602 (formerly TF38.02.23)
on "Coordinated Voltage Control in Transmission
Network

37

1538pk

Coordinated Voltage Control (cont'd)

General Comments

Coordinated voltage control receiving much

attention with the "de-regulation" of the electricity
supply industry and emergence of "energy
markets"
) better able to cope with power interchange over
longer distances and the need for higher utilization
of transmission assets
) facilitates the identification and recognition of the
contributions of different participants to the
"voltage ancillary service"

Currently available wide-area monitoring,

processing and control technologies make the
application of coordinated voltage control more
attractive and feasible

38

1538pk

7. EXAMPLES OF MAJOR
SYSTEM DISTURBANCES DUE
TO VOLTAGE INSTABILITY

Copyright P. Kundur
This material should not be used without the author's consent

1538pk

Incidents of Short-Term Voltage

Instability

South Florida, USA

17 May 1985

Western Tennessee, USA

22 August 1987

Phoenix area, Arizona, USA

29 July 1995

Southern California, USA

5 August 1997

Atlanta area, Georgia, USA

30 July 1999

1538pk

Incidents of Long-Term Voltage

Instability

Brittany, France

10 November 1976

France

19 December 1978

Western France

12 January 1987

Sweden

27 December 1983

Tokyo, Japan

23 July 1987

Florida, USA

30 December 1982

WSCC

2 July 1996

1538pk

July 2, 1996 WSCC

Disturbance

1538pk

WSCC July 2, 1996 Disturbance

Started in Wyoming and Idaho area at 14:24:37

Loads were high; high temperature around 38C
Heavy power transfers from Pacific NW to
California

LG fault on 345 kV line from Jim Bridger 2000

MW plant in Wyoming to Idaho due to flashover
to a tree
) tripping of parallel line due to relay
misoperation

Tripping of two (of four) Jim Bridger units as

stability control; this should have stabilized the
system

Faulty relay tripped 230 kV line in Eastern

Oregon

Voltage decay in southern Idaho and slow

decay in central Oregon

1538pk

WSCC July 2, 1996 Disturbance (cont'd)

1538pk

WSCC July 2, 1996 Disturbance (cont'd)

About 24 seconds later, generators at a small

hydro plant in Southern Idaho tripped because
of high field current

About the same time, a long 230 kV line (Amps

line) from western Montana to Southern Idaho
tripped
) zone 3 relay operation
) parallel 161 kV line subsequently tripped

Rapid voltage decay in Idaho and Oregon

Pacific intertie lines separation less than two
seconds after the Hells Canyon lines tripped

Cascading to five islands 35 seconds after

initial fault

2.2 million customers experienced outages;

total load lost 11,900 MW

1538pk

WSCC July 2, 1996 Disturbance (cont'd)

4
5
2
1

Event 5
Event 4
Event 3 :
Event 2 :
Event 1 :

14:5:01
14:24:38
14:24:37
14:24:37
14:24:37

Anaconda-Amps 230 kV line tripped (+Dillon-Big Grassy)

LaGrand-Roundup 230kV line tripped
Jim Bridger units 2&4 tripped by remedial action scheme
Jim Bridger-Goshen 345 kV line tripped
Kinport-Jim Bridger 345 kV line on LG fault

1538pk

WSCC July 2, 1996 Disturbance (cont'd)

1538pk

TSAT was used to Replicate Disturbance

in Time Domain

MEASURED RESPONSE

SIMULATED
RESPONSE

1538pk

WSCC July 2, 1996 Disturbance

Modelling for TSAT - (Time-Domain)

Original base case did not capture collapse

Field current limiters added to units in Northwest,
Idaho, Montana, PG&E and BC Hydro

Loads in Northwest represented as 100% constant

P&Q

Loads in Idaho and Montana represented as

exponential loads with exponents of 0.4 and 0.3
respectively - all other loads remained unchanged

Simulation of events up to and including Amps

line outage

Simulation captures event quite closely

10

1538pk