Renewable and Sustainable Energy Reviews 63 (2016) 112

Contents lists available at ScienceDirect

Renewable and Sustainable Energy Reviews

journal homepage: www.elsevier.com/locate/rser

A comprehensive review of the voltage stability indices

Javad Modarresi, Eskandar Gholipour n, Amin Khodabakhshian
Department of Electrical Engineering, Faculty of Engineering, University of Isfahan, Isfahan, Iran

art ic l e i nf o

a b s t r a c t

Article history:
Received 6 May 2015
Received in revised form
8 November 2015
Accepted 3 May 2016

Voltage stability assessment is a major issue in monitoring the power system stability. Different voltage
stability indices (VSIs) have been proposed in the literature for voltage stability assessment. These indices
can be used for distributed generation (DG) placement and sizing, detecting the weak lines and buses
and triggering the countermeasures against voltage instability. This paper reviews the VSIs from different
aspects such as concepts, assumptions, critical values and equations. The review results provide a
comprehensive background to nd out the future works in this eld and select the best VSI for different
applications like DG placement and sizing and voltage stability assessment.
& 2016 Elsevier Ltd. All rights reserved.

Keywords:
Voltage stability index
Voltage stability
DG placement
Weak line
Weak bus

Contents
1.
2.
3.
4.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
VSIs in DG placement and sizing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Characteristics of the voltage collapse point . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Voltage stability indices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
4.1.
Line voltage stability indices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4.1.1.
Fast voltage stability index (FVSI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4.1.2.
Line Stability Index (Lmn) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4.1.3.
Line Stability Factor (LQP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4.1.4.
Line Stability Index (Lp) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4.1.5.
Novel line stability index (NLSI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4.1.6.
Voltage collapse proximity index (VCPI). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4.1.7.
New Voltage Stability Index (NVSI). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4.1.8.
Voltage reactive power index (VQILine) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4.1.9.
Power transfer stability index (PTSI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4.1.10.
Voltage stability index (VSI_1). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4.1.11.
Voltage Stability Load Index (VSLI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4.1.12.
Voltage stability margin (VSMs). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4.1.13.
Voltage Collapse Proximity Index (VCPI_1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4.1.14.
Voltage Stability Indicator (VSI_2). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
4.1.15.
Voltage Stability Margin Index (VSMI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
4.1.16.
Voltage Stability Load Bus Index (VSLBI). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
4.1.17.
Stability Index (SI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
4.1.18.
Line Collapse Proximity Index (LCPI). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
4.2.
Bus voltage stability indices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
4.2.1.
Voltage collapse prediction index (VCPIbus) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
4.2.2.
L-index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.2.3.
S difference criterion (SDC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

Corresponding author. Tel.: 98 3137935605.

E-mail addresses: J.modarresi@eng.ui.ac.ir (J. Modarresi), e.gholipour@eng.ui.ac.ir (E. Gholipour), aminkh@eng.ui.ac.ir (A. Khodabakhshian).

http://dx.doi.org/10.1016/j.rser.2016.05.010
1364-0321/& 2016 Elsevier Ltd. All rights reserved.

J. Modarresi et al. / Renewable and Sustainable Energy Reviews 63 (2016) 112

4.2.4.
Voltage stability index (VSIbus) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.5.
Impedance matching Stability Index (ISI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.6.
ZL=ZS ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.7.
Simplied Voltage Stability Index (SVSI). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.
Overall voltage stability indices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.1.
Network sensitivity approach (SG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5. Classication of VSIs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction
Voltage instability, which is essentially a local phenomenon,
has been the cause of many major blackouts in the world [1]. In
[2], 12 blackouts have been studied from 1965 to 2005 and indicating that voltage instability had been a major incident in 7 cases.
The term voltage collapse is also used instead of voltage instability
and it is the process by which the sequence of events accompanying voltage instability leads to an abnormally low voltages or
blackout in a large part of the power system [3].
Voltage stability has been dened by the IEEE power system
engineering committee as follows: "Voltage stability is the ability
of a system to maintain voltage so that, when load admittance is
increased, load power will increase, and so that both power and
voltage are controllable" [4]. The IEEE/CIGRE Joint Task Force
provides another denition for voltage stability which is: "Voltage
stability refers to the ability of a power system to maintain steady
voltages at all buses in the system after being subjected to a disturbance from a given initial operating condition" [3].
Voltage instability can be caused either by a failure of reactive
power sources in producing enough reactive power or by a failure
of power system lines in transmitting the required reactive power
[5]. The reactive power can be supplied by generators and reactive
power compensators such as shunt capacitors. The main incidents
causing voltage instability are load increase, tripping of power
system equipment (such as transmission lines, power transformers and generators), exceeding some generators reactive power
limits and the malfunction of on-load tap changing transformers.
There are countermeasures to avoid voltage instability. The
most important ones are [6]

.9
.9
10
10
10
10
10
11
11

Another application of VSIs is triggering the countermeasures

against voltage instability [1719]. In this case, the indices must be
used in real-time and the required data are provided by the PMUs
which constitute a part of the wide area measurement system
(WAMS).
Many indices have been proposed in the literature to assess the
voltage stability [2023]. Some indices are functions of the power
system impedance but some others are independent of it and only
need the voltage and current of buses. In practice, determining
power system impedance is not possible with high precision due
to the atmospheric effects and insufcient information about the
power system. So, the performance of indices which are functions
of power system impedance is always associated with error.
A comparison of the VSIs have been made in some papers [24
31]. Nevertheless, the overall characteristics, classications and
differences have not been precisely investigated. So, in this paper,
the VSIs are reviewed by considering different aspects and views
such as assumptions, concept, equations and critical values (CVs).
The review results can be employed to understand the future
works in this eld and select the best VSI for different applications
such as DG placement and sizing, voltage stability assessment,
ranking the buses and lines according to the voltage stability and
activating the countermeasures to control the voltage collapse.
The rest of this paper is organized in the following way:
Section 2 examines the application of VSIs in DG placement and sizing. Section 3 discusses the characteristics of the voltage collapse
point. Section 4 reviews the VSIs. In Section 5 the VSIs are classied
and Section 6 concludes the discussions.

2. VSIs in DG placement and sizing

1. Improvement of weak buses and lines at the planning step of
power system (by distributed generation units or other voltage
supporting equipment)
2. Load shedding, shunt capacitor switching
3. FACTS devices to extend the voltage stability margin (VSM)
4. Tap changer blocking
where load shedding is the last line defense.
One of the important applications of the voltage stability
indices (VSIs) is identifying the weak lines and buses in the power
systems [711]. In this case, the VSIs are used in ofine/online
mode, and the required data are obtained from the static analysis
or phasor measurement units (PMUs). Then, the line/bus with a
VSI closest to the critical value is selected as the weakest line/bus.
This application of VSIs can be used in different cases such as, the
placement and sizing of distributed generation (DG) units, capacitor allocation and the planning of power systems [1216]. In the
DG placement and sizing problems, the VSIs are used in two steps.
In the rst step, the weak buses and lines are selected to determine the candidate location for DG units. In the second one, the
optimal location of DG units are determined to maximize the VSM
where the VSM is calculated by the VSIs.

DGs are small-scale power generations (typically in the range

from 3 kW to 10,000 kW) are used to improve the traditional
electric power systems [32,33]. The DG technologies include small
gas turbines, micro-turbines, fuel cells, wind and solar energy,
biomass, small hydro-power, etc. [34]. Integrating DG units into
the power systems has many benets like improving voltage
proles and load factors, grid reinforcement, deferring or eliminating for system upgrades, reducing power losses and on-peak
operating costs, and improving system integrity, reliability, and
efciency [13,35,36]. To achieve these benets, an appropriate size
and place must be selected where the DG units are installed.
Installing these units in the inappropriate places may diminish the
reliability of the system and may also lead to an increase in the
system losses and costs [14]. Moreover, if several DG units are
installed, optimal approach for the placement and sizing of DGs
becomes more crucial in order to maintain the reliability and
stability of the power system [35].
The problems concerning the DG placement and sizing can be
divided into two groups in terms of their objective functions and
each group consists of two steps. In the rst step, the candidate
location for DG units are determined and secondly the optimal

J. Modarresi et al. / Renewable and Sustainable Energy Reviews 63 (2016) 112

locations of DG units are obtained by solving an optimization

problem in the next step. The rst group only minimizes the
power losses [3739] and the candidate locations for DG units are
usually selected randomly while in the second group the objective
is to maximize the VSM and simultaneously minimize the power
losses where the candidate locations for DG units are chosen by
the VSIs [12,4042]. Thus, the VSIs are not used in the rst group
of DG placement and sizing problems.
In the rst step of the second group, the candidate location of
DG units is chosen according to VSIs. Thus, nearly all the VSIs that
are able to determine the weak buses and lines can be used in the
rst step. In the second step of the second group, the optimal
location of DG units are determined by solving an optimization
problem that minimizes the power losses and maximizes VSM.
Hence, for the reason that it is not necessary to determine the
weak buses and lines in the second step, almost all kinds of voltage stability indices can be used in this step.
With more accurate VSIs, better places can be obtained for the
DG units. Moreover, complex VSIs increase the running time of the
optimization problem especially in large power systems. So, ideally, an accurate and simple VSI must be used in DG placement and
sizing problems. In order to provide an overview for the selection
of VSIs, the VSIs are classied according to their types (i.e. line, bus
and overall VSIs), assumptions, concept, equations and critical
values. Then, their accuracy is examined by their assumptions in
the next sections.

3. Characteristics of the voltage collapse point

All of VSIs have been derived from the characteristics of the
voltage collapse point. Thus, investigating the characteristics of
voltage collapse point can lead to a better understanding of the
VSIs. Therefore, the characteristics of voltage collapse point are
examined in this section.
To explain what happens when the voltage collapse point is
reached, the two bus representation of a power system is considered as in Fig. 1.
The symbols in Fig. 1 are as below:
Vs, Vr: voltage magnitude at the sending and receiving buses,
respectively.
Ps, Qs: active and reactive power at the sending bus.
Pr, Qr: active and reactive power at the receiving bus.
Ss, Sr: apparent power at the sending and receiving buses,
respectively.
s, r: voltage angle at the sending and receiving buses,
respectively.
Y: line shunt admittance.
R, X, : line resistance, line reactance and line impedance angle.
The line impedance amplitude is Z and the currents at the
sending and receiving of the line are I s and I r , respectively. To
continue, the shunt admittances are neglected for simplicity. The
active and reactive power at the receiving bus can be written as

(1) and (2) [28].

Pr

V s V r cos  V 2r  Q r X
R

P r X  V s V r sin
R

Qr

where
s r

From (1) and (2), the following forth degree equation can be
derived.
V 4r 2P r R 2Q r X  V 2s V 2r P 2r Q 2r Z 2 0

Or
V 4r 2P r R 2Q r X  V 2s V 2r S2r Z 2

At the voltage collapse point, (4) have two pairs of real identical
roots (the Jacobian matrix is singular). If the load is increased
further, then the roots become complex with real and imaginary
parts. On the other hand, for the voltage to be stable, the discriminant of (4) must be greater than or equal to zero [28].
In practice, the receiving end active and reactive powers are a
function of receiving end voltage. In [43], the inuence of load
models on the roots of (4) has been studied and it has been
indicated that if both active and reactive powers are a square


2


2
functions of voltage (P r V r =V n P n and Q r V r =V n Q n where
V n , P n and Q n are nominal values of voltage and powers at the
receiving bus), then there are no limits on the value of the
receiving end voltage. In other words, the voltage collapse does
not occur. At the limit, when the discriminant of (5) is equal to
zero, it gives maximum load ability of line as (6) [2].
V2

s

6
4Z
 =2



where tan  1 Q r =P r and the operating point reach the knee
point of the PV curve [44]. According to the maximum power
transfer theorem, the following conditions are also satised at the
maximum transferable power:
Sr max

cos 2

1. Load absorbs maximum power.

2. Thevenin and load impedance are equal in magnitude.
3. The amplitude of the voltage drop across the Thevenin impedance is equal to the amplitude of load voltage.
Moreover, in the vicinity of the voltage collapse, increasing the
apparent power at the sending end of the line no longer yields an
increase in the receiving end apparent power [5]. The stressed
lines become big reactive power consumers and they limit the
reactive power supply to the loads. As a result, the reactive power
loading of other lines is increased and the lines reach the maximum transferable point. In the nal step, all of the connected
lines cannot supply the reactive power of the loads, which is the
starting point of its voltage instability [5].

4. Voltage stability indices

Many VSIs have been proposed in literature. These indices have
been classied based on the following ways [2430]:
1. Jacobian matrix and system variables based VSIs
2. Bus, line and overall VSIs

Fig. 1. The two bus representation of a power system.

Jacobian matrix based VSIs can calculate the voltage collapse

point and determine the VSM. But the computation time is high

J. Modarresi et al. / Renewable and Sustainable Energy Reviews 63 (2016) 112

and any topological change leads to change the Jacobian matrix

and this matrix must be recalculated; hence, they are not suitable
for real-time voltage stability assessment. Moreover, using the VSIs
which are based on the Jacobian matrix causes an increase in the
running time of DG placement and sizing problems. On the other
hand, the VSIs which are based on system variables require less
computation and are adequate for real-time applications. The
disadvantage of these indices is that they cannot accurately estimate the VSM so they can just present critical lines and buses. In
many applications such as the rst step of DG placement and
sizing problems, VSIs are used to detect the weakest bus and line
of power system or triggering the countermeasures against voltage instability. Therefore, the classication of VSIs according to
the bus, line and overall indices, which are presented in this paper,
can be very useful.

4.1.3. Line Stability Factor (LQP)

Mohamed et al. developed the line stability factor, LQP, based
on the same concept as two previous line VSIs [48]. The LQP is
obtained as follows:
!
!
X
P 2s X
LQ P 4

Q
11
r
V 2s
V 2s
For the transmission line to be stable, it should be LQ P o 1. In
this index, the lines are assumed to be lossless (R=X o o 1) and the
shunt admittance of lines is neglected.
4.1.4. Line Stability Index (Lp)
The Lp has been designated by Moghavvemi et al. based on the
same concept as previous line VSIs. For any value of Lp greater than
one, the system is considered as unstable [49]. The Lp for a
transmission line is dened as follows:

4.1. Line voltage stability indices

Lp
Voltage stability analysis can be evaluated by the voltage stability index referred to a line. All of the line VSIs are formulated
based on the two bus representation of a system as Fig. 1 where
the shunt admittances are neglected. So, the theoretical base of
most of the line VSIs are the same and the difference is in the
assumptions used in each index. In proving most of the line VSIs,
the discriminant of the voltage quadratic equation is set to be
greater than or equal to zero to achieve the stability [28]. The line
VSIs which have been proposed in the literature are briey
described in this section.
4.1.1. Fast voltage stability index (FVSI)
Musirin et al. proposed the FVSI based on the concept in which
the discriminant of the voltage quadratic equation (Eq. (5)) is set to
be greater than or equal to zero [45]. For a typical transmission
line, the FVSI is calculated by
FVSI

4Z 2 Q r

V 2s X

The FVSI must be below 1 for a stable transmission line. If FVSI

goes beyond 1.00, one of the buses that is connected to the line
will experience a sudden voltage drop leading to system collapse.
In this index, the shunt admittance is neglected and the following
assumptions are also considered.

4RP r
V s cos  2

12

In this index, the effect of reactive power on voltage stability as

well as line shunt admittance are neglected and it is assumed that
only the active power affects the line voltage stability.
4.1.5. Novel line stability index (NLSI)
Yazdanpanah-Goharrizi et al. derived a line stability index
based on the same concept as Lp. Any line in the system whose
NLSI is close to unity indicates that the line is approaching its
stability limit [50]. This index is given by
NLSI

Pr R Q r X
0:25V 2s

13

In this index, the angular difference between the sending and

the receiving voltage is assumed to be very small (  0) and also
the line shunt admittance is neglected.
4.1.6. Voltage collapse proximity index (VCPI)
Moghavvemi et al. had proposed four VCPIs for the assessment
of the line voltage stability based on the concept of maximum
power transferable through a line [51]. VCPIs consider the maximum power (P rmax and Q rmax ) transferred through a line and
maximum power loss (P rmax and Q rmax ) in a line. These indices
are dened as follows [51]:

VCPI1

Pr
P rmax

14

In [46], the assumption of (8) is not considered and the index Lij
has been obtained as following:

VCPI2

Qr
Q rmax

15

VCPI3

Pl
P lmax

16

VCPI4

Ql
Q lmax

17

sin  0;

cos  1;

R sin  0;

X cos  X

Lij

4Z Q r X
V 2s R sin  X cos 2

4.1.2. Line Stability Index (Lmn)

Line stability index, Lmn has been proposed by Moghavemmi et
al., and is obtained using the same concept as FVSI in which the
discriminant of the voltage quadratic equation is set to be greater
than or equal to zero [47]. For a typical transmission line, the Lmn
calculated by
Lmn

P rmax

4XQ r
V s sin 

where maximum powers and losses are calculated by (18)(21)

[51].

10

In the Lmn, the effect of the active power on the voltage stability
as well as the line shunt admittance are neglected. As long as the
Lmn remains less than 1, the system is stable and when this index
exceeds the value 1, the system loses its stability and the voltage
collapses.

V 2s
cos



Z 4 cos 2 2

18

V 2s
sin



Z 4 cos 2 2

19

Q rmax

P lmax

V 2s
cos



Z 4 cos 2 2

20

J. Modarresi et al. / Renewable and Sustainable Energy Reviews 63 (2016) 112

Q lmax

V 2s
sin



Z 4 cos 2 2

21

where is the phase angle of the load impedance or load power

factor. For any values of VCPIs greater than one, the system is
exposed to a voltage collapse. In [51], it is indicated that VCPI1
V CPI2 and VCPI3 VCPI4. Therefore, instead of considering
four indicators, either real or reactive terms can be considered. In
[52] the index Lsr is proposed based on the maximum apparent
power transferred through a line as follows:
Lsr

Sr

22

Srmax

It can be easily noted that the index Lsr has the same value as
the index VCPI(1) and VCPI(2) [53]. In the VCPIs, the power factor is
assumed to be constant and the line shunt admittance is
neglected.
4.1.7. New Voltage Stability Index (NVSI)
The NVSI derived by Kanimozhi et al. is obtained using the
same concept as Lp [54]. This index is dened for a transmission
line as follows:
q
2X P 2r Q 2r
23
NVSI
2Q r X  V 2s
For the transmission line to be stable NVSI o 1. In this index,
the line resistance as well as the line shunt admittance are
neglected.
4.1.8. Voltage reactive power index (VQILine)
This index has been derived by Althowibi et al. based on the
same concept as Lp. This index is given by [55]
VQ I Line

4Q r

24

jBjV 2s

where B ImY Im1=R jX. The critical value (CV) of VQILine

is 1 and, beyond this value, the voltage will collapse. In this index,
is assumed to be zero and the line shunt admittance is neglected.
4.1.9. Power transfer stability index (PTSI)
The PTSI has been proposed by Nizam et al. as follows [56]:
PTSI

2Sr Z1 cos 
V 2s

25

This index is based on the same concept as VCPI in which the

maximum transferable power through a line is limited. For the
safe operation of the network, the PTSI must be less than unity.
When PTSI reaches 1, it indicates that voltage collapse has occurred. In this index, the line shunt admittance is neglected the same
as all the previous line VSIs.
4.1.10. Voltage stability index (VSI_1)
This index has been taken from [57], but the notation of VSI has
been changed to VSI_1 to avoid confusing it with other notations.
The VSI_1 is formulated as follows based on the same concept as
VCPI and the PTSI:


P margin Q margin Smargin
VSI_1 min
;
;
26
P max Q max Smax
where
P margin P max  P r

27

Q margin Q max  Q r

28

Smargin Smax  Sr

29

s
V 2s
V2
P max
Qr s
2
X
4X

30

V 2s P 2r X

4X V 2s

31

1  sin V 2s
2X cos 2

32

Q max

Smax

To maintain a secure condition, the value of VSI_1 index should

be maintained greater than 0 [57]. In this index, the line resistance
as well line shunt admittance are neglected.
4.1.11. Voltage Stability Load Index (VSLI)
The VSLI is proposed by Abdul Rahman et al. for line voltage
stability assessment based on the same concept as Lp. The VSLI is
given by [58,59]
h
i
4 V s V r cos  V 2r cos
33
VSLI
V 2s
VSLI must be kept below 1.0 to maintain voltage stability. If VSLI
exceeds 1.0, it indicates a voltage collapse in the system. The VSLI
can be calculated from PMU measurements of voltage magnitudes
and angles [59]. In this index, the line shunt admittances are
neglected as all of previous indices. In [60], the value of cos is
considered equal to 1 and the stability index L is formulated as
(34).


4 V s V r  V 2r
34
L
V 2s
4.1.12. Voltage stability margin (VSMs)
Guiping et al. proposed VSMs based on the same concept as
VCPI and PTSI as follows [61]:
VSM s

Scr  SL
Scr

35

where
Scr

V 21
2Z 1 cos  

36

The CV of VSM index is zero and the system will experience the
voltage collapse in values less than zero. In the VSMs, the power
factor is assumed to be constant and the line shunt admittance is
neglected.
4.1.13. Voltage Collapse Proximity Index (VCPI_1)
This index has been taken from [62] but the notation of VCPI
has been changed to the VCPI_1 to avoid confusing it with other
notations. The main idea of VCPI_1 is that at the collapse point, the
voltage drop across the Thevenin impedance is equal to the load
voltage. Therefore, to assess the risk of voltage collapse, VCPI_1 has
been dened as
VCPI_1 V r cos  0:5V s

37

When VCPI Z 0, the line voltage is stable and otherwise, it is

unstable. In this index, the lines are not simplied. However, it is
assumed that an ideal voltage source is connected to the sending
bus and the network equivalent Thevenin impedance which is
connected to the sending bus is neglected. In [62], another Voltage
stability index for power transmission paths has been proposed
based on VCPI_1. In this case, the path is transformed into a twobus equivalent and the stability index of the power transmission

J. Modarresi et al. / Renewable and Sustainable Energy Reviews 63 (2016) 112

Table 1
Characteristics of VSIs.
Ref. no. VSI

CV
1

Equation

Assumptions


 
sin  0; cos  1
FV SI 4Z 2 Q r = V 2s X

 

Y 0
Lij 4Z 2 Q r X = V 2s R sin  X cos 2


1 [45]

FVSI

2 [46]

Lij

3 [47]

Lmn

4 [48]

LQP

Lmn 4XQ r =V s sin 






LQ P 4 X=V 2s Q r P 2s X =V 2s

5 [49]

Lp

Lp V

6 [50]

NLSI

7 [51]

VCPI

Effect of active power neglected, Y  0

R  0; Y  0
Effect of reactive power neglected, Y  0

4RP r
cos  2

s


NLSI P r R Q r X = 0:25V 2s

 0, Y  0

V CPI1 P r =P rmax ; VCPI2 Q r =Q rmax

Constant power factor, Y  0

V CPI3 P l =P lmax ; V CPI4 Q l =Q lmax

Lsr
NVSI

1
1

10 [55]

VQILine

11 [56]

PTSI

12 [57]

VSI_1

8 [52]
9 [54]

Lsr Sr =Srmax
 q
NV SI 2X P 2r Q 2r =2Q r X  V 2s


V Q I Line 4Q r = jBjV 2s

Constant power factor, Y  0

R  0, Y  0

PTSI 2Sr Z1 cos  =V 2s



Q
P
S
V SI_1 min Pmargin
; Qmargin
; Smargin
max
max
max

Y 0

 0, Y  0
R  0;

h
i
V SLI 4 V s V r cos  V 2r cos 2 =V 2s


L 4 V s V r  V 2r =V 2s

Y 0

Y 0

13 [58,59]

VSLI

14 [60]

15 [61]
16 [62]
17 [64]

VSMs
VCPI_1
VSI_2

0
0
1

V SM s Scr  SL =Scr
V CPI_1 V r cos  0:5V s

  

V SI_2 4Q r R X2 = X V 2s 8RQ r

Constant power factor, Y  0

Z ths  0
Y  0;  0

18 [65]
19 [66]
20 [67]

VSMI
VSLBI
SI

0
1
0

VSMI (max  )/max

V SLBI V r =V

Y  0 and R  0 in lossless case

Z ths  0
Y 0

21 [68]

LCPI

22 [75]
23 [76]

VCPIbus
L-index

1
1

SI 2V 2s V 2r  V 4r  2V 2r P r R Q r X
 Z 2 P 2r Q 2r
"
#
4 A cos P r jBj cos
LCPI Vj jcos 2
s
Q r jBj sin

Lines are modeled as pie model

V CPIbus minfV CPI i g






P


F ji V i =V j 
L max Lj max1 


j A L
j A L
i A G








SDC 1 V r I r = V r I r 


V SI bus 1 I i V i =V i Ii
ISI Z L  Z th =Z L 1  jI r V r j=jV r I r j

24 [5]

SDC

25 [79]

VSIbus

26 [80]
27 [84]

ISI
0
Z L =Z S ratio 1

28 [85]

SVSI

29 [97]

SG

sharp rise SGp P gt , SGq P gt

P dt
Q dt

ZL
ZS

Voltage of a bus is not depend on the other bus voltages

All generator voltages remain unchanged.

V r I r  0
System topology remains unchanged after a disturbance
731 r S r 871

M 1
0:5
 M cos M cos 2  M 2 1

r
SV SI r V
V r

Voltage of the nearest generator to a load bus is equal to the Thevenin voltage of the
network at that bus
Power system efciency is constant

path (PVSI) is dened as follows:

0

PVSI V 2 cos 0:5V 1

0

38

where the V 1 and V 2 are the voltage magnitudes of the sending

and receiving buses in the two-bus equivalent system. The same
VSI called equivalent node voltage collapse index (ENVCI) has been
introduced in [63].
4.1.14. Voltage Stability Indicator (VSI_2)
This index has been taken from [64] but the notation of VSI has
been changed to the VSI_2 to avoid confusing it with other notations. This index is obtained using the same concept as Lp, in which
the discriminant of the voltage quadratic equation is set to be
greater than or equal to zero. The VSI_2 is obtained as follows:

4.1.15. Voltage Stability Margin Index (VSMI)

VSMI has been dened by He et al. based on the relationship
between maximum transferable power through a line and the
angle difference between sending and receiving buses [65].
Operating at secure and stable conditions requires the value of
VSMI to be maintained greater than 0. The formulation of VSMI is
as following:
VSMI max  =max

39

For the safe operation of the network, the VSI_2 must be less
than unity. In this index, the voltage angle is assumed to be very
small and the line shunt admittance is neglected.

40

where max is the maximum angle difference between sending and

receiving buses which can be calculated in lossless and lossy
transmission lines as (41) and (42), respectively.


2

41


2

42

max 2
max

4Q R X

VSI_2  r
X V 2s 8RQ r

Y  0;  0

In this index, the line shunt admittance is neglected as were

many of previous indices.
4.1.16. Voltage Stability Load Bus Index (VSLBI)
Milosevic had proposed the VSLBI for voltage stability assessment using PMU [66]. This index is based on the fact that; under

J. Modarresi et al. / Renewable and Sustainable Energy Reviews 63 (2016) 112

Voltage stability indices

Overall VSIs

Bus VSIs

Line VSIs

SD
Load margin Eigenvalue
EF
RPM
Singular value
SG
VIPI
CMBI
i

TF
SF
TVI
VCPIbus
L-index
SDC
VSI bus
ISI
ZL / Z S ratio
SVSI
V/Vo
P and Q angle

VCPI_1 PTSI
NLSI
VSI_2 VSI_1 VCPI
VSMI VSLI
Lsr
VSLBI
L
NVSI
SI
VSMs VQIline
LCPI

FVSI
Lij
Lmn
LQP
Lp

Fig. 2. Voltage stability indices, classied by their types.

4.1.18. Line Collapse Proximity Index (LCPI)

The LCPI has been proposed by Tiwari et al. based on the same
concept as SI [68]. The formulation of LCPI is as follows:

Suitable VSIs for the second group of DG

placement and sizing problems

LCPI
First step

Second step

Line VSIs

Line VSIs

Bus VSIs except SDC,

VSIbus, ISI and ZL / ZS ratio

Bus VSIs except SDC,

VSIbus, ISI and Z L / Z S ratio

"

Fig. 3. Suitable VSIs for the second group of DG placement and sizing problems.

Simplicity

Bus VSIs

V s cos 2

45

where A and B are the transmission line parameters of the twoport network and and are the phase angles of parameters A and
B respectively. The A and B can be expressed as

Overall VSIs

Line VSIs

 
4A cos P r jBj cos Q r jBj sin 

Vs

Is

"

Vr

Ir

46

where A, B, C and D are calculated by the pie model of a transmission line of a two bus system (Fig. 1) as follows:

Overall VSIs


A 1 Z  Y=2
BZ



C Y  1 Z  Y=4

Accuracy

47

DA

Fig. 4. Comparison between accuracy and simplicity of different types of VSIs.

the maximum power conditions, voltage drop across the transmission line impedance is equal to the load bus voltage. So, the
VSLBI in dened as follows:
VSLBI

Vr
V

43

where V is the dropped voltage across the transmission line. As

long as the VSLBI is greater than 1, the system is stable and when
this index drops to the values less than 1, the system loses its
stability and the voltage collapses. In this index, the same
assumptions as VCPI_1 are used.

4.1.17. Stability Index (SI)

Eminoglu et al. proposed SI based on the voltage quadratic
equation which is mostly used for the calculation of the line VSIs
[67]. The formulation of this index is as follows:
SI 2V 2s V 2r  V 4r  2V 2r P r R Q r X  Z 2 P 2r Q 2r

44

The line, at which the value of the SI is at minimum, is the most

sensitive to the voltage collapse. When the SI value reaches zero,
the voltage collapses. In this index, the line shunt admittance is
neglected as many of previous indices.

To maintain the voltage stability of the system, the LCPI must be

less than unity. In this index, the line shunt admittance is considered and the transmission lines are modeled as model.

4.2. Bus voltage stability indices

Bus VSIs determine the voltage stability of system buses and do
not provide any information about the weak facilities with
potential voltage problems in the system. So, the bus voltage
stability indices cannot be used for the determination of the weak
facilities. In this section, the bus VSIs are reviewed. Some of bus
voltage stability indices such as sensitivity factors (SF) [69,70], test
functions (TF) [71], P and Q angles [70,72], V/Vo Index [70] and
tangent vector index (TVI) [73,74] have been reviewed in [31].
Nevertheless, there are still many of bus VSIs that must be
reviewed.
4.2.1. Voltage collapse prediction index (VCPIbus)
The VCPIbus is derived from the basic power ow equation and
its value varies between 0 and 1. If the value of VCPIbus reaches 1,
the voltage at a bus has collapsed [75]. The formulation of this
index is as follows:
VCPI bus minfVCPI i g

48

J. Modarresi et al. / Renewable and Sustainable Energy Reviews 63 (2016) 112

Concept of VSIs

Existence of
Maximum
transferable power solutions for
voltage
equation
through a line

VCPI VSI_1
Lsr VSMs
PTSI VSMI

VQIline
VSLI
L
VSI_2
SI
L-index
VCPIbus
LCPI

P-V curve

SD
FVSI
CMBI
Lij P and Q angle
Lmn
RPM
Lp
SG
LQP
NVSI
NLSI

TVI
SF
VIPI
TF
V/Vo

Maximum power
transfer theorem

Jacobian matrix
Lyapunov
stability theory

Eigenvalue Singular value SDC

VCPI_1
Load margin
VSIbus VSLBI
i
ISI
SVSI
ZL / ZS ratio
Energy function

Fig. 5. Concepts of the voltage stability indices.

element in j-th row and i-th column of matrix F whose elements

are generated from the admittance matrix as (52) [77].

Impedance dependent and

independent voltage stability indices

1
F  Y LL
Y LG

Impedance
independent

SDC
VSLI
VSIbus
L
TVI
VCPI_1
VSLBI ZL / ZS ratio
SF
ISI
P and Q angle
SG

Impedance
dependent

VSI_2
VSMI
SI
LCPI
VCPIbus
L-index
SD
EF

PTSI
VSI_1
VSMs
i
Load margin
V/Vo
RPM
Eigenvalue

NLSI
VCPI
Lsr
NVSI
VQIline
SVSI
CMBI
Singular value

and
FVSI
Lij
Lmn
LQP
Lp
TF
VIPI

Fig. 6. Impedance dependent and independent voltage stability indices.

where


N
0 

P


V
m


m1




mai


VCPI i 1 



Vi









49

and
Y
V m N im V m
P
Y ij
0

52

50

j1
j ai

where V i and V m are the voltage phasors at bus m and bus i, N is

the number of buses, and Y im is the admittance between the buses
i and m. This index is based on the concept that the voltage
equations must have a solution. In the matrix form, the determinant of a matrix must not be zero.
4.2.2. L-index
Kessel et al. proposed L-index based on the solution of the
power ow equations [76]. This index has been derived as follows:


P

F ji V i 


i A G


L max Lj max1 
51

Vj 
j A L
j A L 


where L is the set of load buses, G is the set of generator buses, V j
and V i are the voltage phasors at bus j and bus i, and F ji is the

"

IL
IG

"

Y LL

Y LG

Y GL

Y GG

#"

VL

VG

53

The values of L-index vary between 0 (no load condition) and 1

(voltage collapse) and the concept of this index is the same as that
of VCPIbus. It can be shown that the stability limit is L 1 when two
conditions are fullled [76]. The rst condition requires that all
generator voltages remain unchanged (amplitude and phase
angle). The second is related to the nodal currents in which the
nodal current at bus j is directly proportional to the current I i and
indirectly proportional to the voltage V i [76]. In general, these two
conditions are satised approximately. So, the exact CV is around
L 1. In [78], L-index has been improved to consider the inuence
of load model.
4.2.3. S difference criterion (SDC)
The SDC proposed by Verbic et al. is based on the fact that in
the vicinity of voltage collapse point an increase in the apparent
power ow at the sending end of the line no longer yields an
increase in the received power [5]. The SDC is dened as (54).



V r I r 
SDC 1

V r I r

54

At the point of the voltage collapse, when Sr 0, the SDC

equals 0. In this index very small values such as VrVr are
neglected in the process of proof. Moreover, this index is calculated when the following conditions are satised.
1. jI r j 4 I min
2. The studied line is a reactive power consumer.
Otherwise, the SDC is set to 1. The SDC cannot determine the
exact time of collapse because it is always a positive value and
never goes below zero.

J. Modarresi et al. / Renewable and Sustainable Energy Reviews 63 (2016) 112

8000

Load margin (MVA)

7000
6000
5000
4000
3000
2000
1000
0
6000

7000

8000

9000
10000
11000
Loading level (MVA)

12000

13000

14000

7000

8000

9000
10000
11000
Loading level (MVA)

12000

13000

14000

13000

14000

second order index (pu)

1
0.8
0.6
0.4
0.2
0
6000

10
9

VSLBI

8
7
6
5
4
3
2

6000

7000

8000

9000
10000 11000
Loading level (MVA)

12000

Fig. 7. Behavior of VSIs versus loading level: (a) load margin, (b) second order index, and (c) VSLBI.

4.2.4. Voltage stability index (VSIbus)

Haque proposed the VSIbus based on the same concept and
assumption of SDC [79]. This index is given by

 

I
V i
VSI i 1 i
55
Vi
I i
where I i and V i are the current and voltage at bus i, Ii and Vi are
current and voltage deviation at bus i, and is a constant number
equal or greater than 1 which is used to give a more or less linear
characteristic to the index. The value of VSIbus varies between unity
(at no load) and zero (at voltage collapse point).
4.2.5. Impedance matching Stability Index (ISI)
According to the circuit theory, when the amplitude of load
impedance is equal to the magnitude of the Thevenin impedance,
the system reaches the maximum transferable power. The ISI

which has been proposed by Smon uses the above law to estimate
the VSM [80]. This index takes values around 1 in normal conditions and at the voltage-collapse point diminishes to 0. The formulation of ISI is as follows:
ISI

jI r V r j
Z L  Z th
1
jV r I r j
ZL

ZL

Vr
;
Ir

Z th

V r
I r

56

57

where Vr and Ir are the voltage and current differences between

two consecutive measurement samples from a PMU at the
receiving bus. The ISI is more sensitive to smaller changes in the
system voltage and current and determining the threshold for Vr
and Ir requires a lot of tests on the system. It must be noted that
all of non-iterative Thevenin equivalent estimation methods such

10

J. Modarresi et al. / Renewable and Sustainable Energy Reviews 63 (2016) 112

as [8183] can be used as real-time voltage stability index [29]. In

this index, it is assumed that after a disturbance, the topology of
the system remains unchanged.
4.2.6. Z L =Z S ratio
The index of Z L =Z S ratio has been dened by Wiszniewski based
on the same concept as ISI [84]. In this index, ratio of the load
impedance at the load bus (Z L ) to the network Thevenin impedance (Z S ) at that bus is calculated as follows:
ZL

ZS

M 1

 M cos M cos 2 M 2 1

i0:5

58

where S  L and, S and L are the phase angles of impedances Z S and Z L , respectively. The factor M is calculated by
measuring the variation of the load apparent power S and the load
admittance Y 1/ZL between the two measurements as (59).
M

S 2  S 1 Y 2 Y 1
S 2 S 1 Y 2  Y 1

59

The Z L =Z S ratio index is very sensitive to the small change of

the load admittance. So, the time difference between the two
measurements ought to be about 500 ms [84]. To maintain a
secure condition, the value of this index should be kept above 1. In
this index, it is assumed that 731 r S r 871.
4.2.7. Simplied Voltage Stability Index (SVSI)
Perez-Londono et al. have proposed the SVSI based on the same
concept as VSLBI [85]. The SVSI is dened as follows:
SVSI r

V r
V r

60

where is the correction factor and is obtained by (61) which is

associated with the highest difference of the voltage magnitude
between two buses (m and l).
  2



61
1  max jV m j  V l 
The Vr is the voltage drop on the Thevenin impedance and it
is estimated by (62).


62
V r V g  V r 
where V g and V r are the voltage phasors of the generator to the
load bus and the analyzed load bus, respectively, which is the
generator is obtained by the relative electrical distance (RED) [86].
When the SVSI value reaches 1 in a bus, the voltage collapse occurs
at that bus. In the SVSI, the voltage of the closest generator to the
load bus is considered to be equal to the Thevenin voltage of the
network at that load bus.
4.3. Overall voltage stability indices
This type of VSIs are not related to the system buses and lines.
So, the overall VSI cannot determine the weakest bus or line and
can only predict the system collapse point. In [31] an almost
complete review on overall VSIs has been performed and several
overall VSIs such as load margin [71,8789], system determinant
(SD) [31], second order index (i) [90], voltage instability proximity
index (VIPI) [91], center manifold based index (CMBI) [92], energy
functions (EF) [93,94], reactive power margins (RPM) [95,96],
singular values and eigenvalues have been examined. So, in this
section only one VSI that not examined in [31] is reviewed.
4.3.1. Network sensitivity approach (SG)
SG has been proposed by Althowibi et al. to calculate voltage
stability and senses how far the system is from its collapse point.

SG contains tow indices as follows [97]:

SGp

P gt
P dt

63

SGq

P gt
Q dt

64

where P gt , P dt and Q dt are total active power generation, total

active power demand and total reactive power demand, respectively. The system approaches to its collapse point when SGp and
SGq increment gradually, causing a sharp rise to innite values
[97]. The SG is based on the P-V curve and it assumed that the
power system efciency is constant.

5. Classication of VSIs
In order to establish a comparison, Table 1 shows the main
characteristics of the reviewed VSIs such as equation, CV and
assumptions. Furthermore, the references that proposed each VSI
are included in the table. According to this table, the CV of the
most VSIs is one and in the most of them, the line shunt admittance is neglected.
A summary and classication of the VSIs are presented in Fig. 2.
In this gure, the VSIs are categorized as line VSIs, overall VSIs and
bus VSIs.
In the rst step of the second group of the DG placement and
sizing problems, the weak lines and buses need to be identied.
Thus, the VSIs that cannot determine the weak buses and lines are
not suitable for this step but in the second step, all kinds of VSIs
except for a few ones can be used. DG placement and sizing problems are solved by ofine methods which use load ow to calculate the objective functions and check the constraints. Thus, the
VSIs such as SDC, VSIbus, ISI and Z L =Z S ratio that are based on two
consecutive measurement samples from PMUs are not suitable for
DG placement and sizing problems. According to these explanations, the suitable VSIs for the second group of the DG placement
and sizing problems can be depicted as Fig. 3.
In general, the accuracy of the overall VSIs is better than the
line and bus VSIs but they are complex and need more computational efforts. On the other hand, the line VSIs and some of bus
VSIs are simple but their accuracy is not the same as that of the
overall VSIs and can determine the weak buses and lines. A comparison between accuracy and simplicity of different types of VSIs
is presented in Fig. 4. To increase the precision and decrease the
computation time of DG placement and sizing problems, it is
recommended that the simple bus and line VSIs be used in the rst
step, and in the second step, the overall VSIs and the bus VSIs with
high accuracy must be employed to compute the VSM with high
precision. The accuracy of VSIs can be compared together by taking into view their assumptions which are presented in Table 1.
Another classication of VSIs can be performed based on the
main concepts of the VSIs. The VSIs have main concepts as follows:
1.
2.
3.
4.
5.
6.

Maximum transferable power through a line

Existence of solutions for voltage equation
PV curve
Lyapunov stability theory
Jacobian matrix
Maximum power transfer theorem

In Fig. 5, the VSIs are classied according to their concepts. This

gure indicates that most of VSIs are based on the concept in
which the voltage equation must have a solution to satisfy the
voltage stability.

J. Modarresi et al. / Renewable and Sustainable Energy Reviews 63 (2016) 112

In the WAMS, the magnitude and angle of the voltages and

currents are provided by the PMUs. So, the indices that only use
the magnitude and angle of the voltages and currents are very
compatible with the WAMS. In Fig. 6, the classication of the VSI
based on the dependence on the power system impedance is
provided. According to this gure, VSLI, L, VCPI_1, VSLBI, SDC, ISI,
SG, VSIbus, TVI, Z L =Z S ratio, SF and P and Q angle only need the
provided data by PMUs and thus, these indices are very compatible with WAMS.
The smooth and linear behaviors of the VSI with changes in system load allow for good predictions of collapse proximity. The
behavior of some VSIs with changes in the system load are linear and
quasi linear but some others are non-linear. Examples for each type of
these VSIs are shown in Fig. 7 where the results are provided for New
England test system [98]. As it is clear from this gure, the load
margin is linear and the second order index in quasi-linear while
VSLBI is non-linear. Among the reviewed VSIs, the load margin and
the energy function are linear and the second order index, VIPI and SG
are quasi-linear and nally the other VSIs are non-linear.

6. Conclusion
In this paper, VSIs were reviewed by considering different
aspects and views and it was shown that the reviewed VSIs are
based on six concepts: maximum transferable power through a
line, existence of solutions for voltage equation, Jacobian matrix,
PV curve, Lyapunov stability theory and maximum power transfer
theorem. According to Fig. 5, most of VSIs are based on the concept
that the voltage equation must have a solution and only one of the
reviewed VSIs is based on the Lyapunov stability theory. Another
classication of VSIs was made based on line, bus and overall VSIs.
To increase the accuracy and decrease the running time of the DG
placement and sizing problems, it was recommended that the
simple line and bus VSIs be used in the rst step of the DG placement and sizing problems, and for the second step, the bus and
overall VSIs which have better accuracy must be employed.
In some VSIs, the Thevenin network impedance at the sending
bus is neglected (such as VCPI_1 and VSLBI) and, in some others,
the Thevenin network impedance at a bus is considered but these
indices are very sensitive to the small change of the data for the
two consecutive measurements (such as SDC, VSIbus, ISI and ZL/ZS
ratio). So, introducing a new VSI which considers the Thevenin
network impedance and is insensitive to the small change of the
two consecutive measurement data can be a future work. In the
newly published article [85], an attempt has been made to introduce a VSI that considers the Thevenin impedance at a bus. The
proposed VSI is insensitive to the small change of the two consecutive measurement data. But the proposed VSI (SVSI) is a bus
VSI and does not provide any information about the weak facilities
with potential voltage problems in the system. Moreover, it
assumes that the voltage of the nearest generator to a load bus is
equal to the Thevenin voltage of the network at that bus. This
assumption is not exactly true.

References
[1] Mousavi OA, Bozorg M, Cherkaoui R. Preventive reactive power management
for improving voltage stability margin. Electr Power Syst Res 2013;96:3646.
[2] Eremia M, Shahidehpour M. Handbook of electrical power system dynamics:
modeling, stability, and control. Wiley; 2013.
[3] Kundur P, Paserba J, Ajjarapu V, Andersson G, Bose A, Canizares C, et al. Denition
and classication of power system stability IEEE/CIGRE joint task force on stability
terms and denitions. IEEE Trans Power Syst 2004;19:1387401.
[4] Begovic M, Fulton D, Gonzalez MR, Goossens J, Guro EA, Haas RW, et al.
Summary of system protection and voltage stability. IEEE Trans Power Deliv
1995;10:6318.

11

[5] Verbic G, Gubina F. A new concept of voltage-collapse protection based on

local phasors. IEEE Trans Power Deliv 2004;19:57681.
[6] Pereira RMM, Ferreira CMM, Barbosa FPM. Comparative study of STATCOM
and SVC performance on dynamic voltage collapse of an electric power system
with wind generation. Latin Am Trans 2014;12:13845.
[7] Musirin I, Rahman TKA. Estimating maximum loadability for weak bus identication using FVSI. IEEE Power Eng Rev 2002;22:502.
[8] Jalboub MK, Rajamani HS, Abd-Alhameed RA, Ihbal AM. Weakest bus identication based on modal analysis and singular value decomposition techniques. In: Proceedings of the international conference on energy power and
control; 2010. p. 3516.
[9] Hashim H, Omar YR, Abidin IZ, Zahidi RA, Ahmad N, Ali AM. Weak area analysis based on the apparent impedance and voltage indices. In: Proceedings of
the international conference on energy and environment; 2009. p. 2515.
[10] Paramasivam M, Salloum A, Ajjarapu V, Vittal V, Bhatt B, Liu N. Dynamic
optimization based reactive power planning to mitigate slow voltage recovery
and short term voltage instability. IEEE Trans Power Syst 2013;28:386573.
[11] Amrane Y, Boudour M, Belazzoug M. A new optimal reactive power planning
based on differential search algorithm. Int J Electr Power Energy Syst
2015;64:55161.
[12] Ettehadi M, Ghasemi H, Vaez-Zadeh S. Voltage stability-based DG placement
in distribution networks. IEEE Trans Power Deliv 2013;28:1718.
[13] Al Abri RS, El-Saadany EF, Atwa YM. Optimal placement and sizing method to
improve the voltage stability margin in a distribution system using distributed
generation. IEEE Trans Power Syst 2013;28:32634.
[14] Esmaili M. Placement of minimum distributed generation units observing
power losses and voltage stability with network constraints. IET Gener Transm
Distrib 2013;7:81321.
[15] Sheng W, Liu KY, Liu Y, Meng X, Li Y. Optimal placement and sizing of distributed generation via an improved nondominated sorting genetic algorithm
II. IEEE Trans Power Deliv 2015;30:56978.
[16] Zeinalzadeh A, Mohammadi Y, Morad MH. Optimal multi objective placement
and sizing of multiple DGs and shunt capacitor banks simultaneously considering load uncertainty via MOPSO approach. Int J Electr Power Energy Syst
2015;67:33649.
[17] Junjie T, Junqi L, Ponci F, Monti A. Adaptive load shedding based on combined
frequency and voltage stability assessment using synchrophasor measurements. IEEE Trans Power Syst 2013;28:203547.
[18] Sasikala J, Ramaswamy M. Fuzzy based load shedding strategies for avoiding
voltage collapse. Appl Soft Comput 2011;11:317985.
[19] Zahidi RA, Abidin IZ, Hashim H, Omar YR, Ahmad N, Ali AM. Study of static
voltage stability index as an indicator for under voltage load shedding
schemes. In: Proceedings of the international conference on energy and
environment; 2009. p. 25661.
[20] Huang GM, Nair NKC. Detection of dynamic voltage collapse. In: Proceedings
of the IEEE Power Engineering Society summer meeting; 2002. p. 12849.
[21] Nisam M, Mohamed A, Hussain A. Performance evaluation of voltage stability
indices for dynamic voltage collapse prediction. J Appl Sci 2006;6:110413.
[22] Abhyankar SG, Flueck AJ. A new conrmation of voltage collapse via instantaneous time domain simulation. In: Proceedings of the North American
power symposium; 2009. p. 19.
[23] Cupelli M, Cardet CD, Monti A. Comparison of line voltage stability indices
using dynamic real time simulation. In: IEEE PES innovative smart grid technologies Europe; 2013. p. 18.
[24] Karbalaei F, Soleymani H, Afsharnia S. A comparison of voltage collapse
proximity indicators. IPEC 2010:42932.
[25] Reis C, Barbosa FPM. Line indices for voltage stability assessment. In: Proceedings of the IEEE Bucharest power tech conference; 2009. p. 16.
[26] Reis C, Barbosa FPM. A comparison of voltage stability indices. In: Proceedings
of the IEEE Mediterranean electrotechnical conference; 2006. p. 100710.
[27] Ismail NAM, Zin AAM, Khairuddin A, Khokhar S. A comparison of voltage
stability indices. In: Proceedings of the international power engineering and
optimization conference; 2014. p. 304.
[28] Lim ZJ, Mustafa MW, Muda ZB. Evaluation of the effectiveness of voltage
stability indices on different loadings. In: Proceedings of the international
power engineering and optimization conference; 2012. p. 5437.
[29] Zhao J, Yang Y, Gao Z. A review on on-line voltage stability monitoring indices
and methods based on local phasor measurements. In: Proceedings of the
power systems computation conference; 2011. p. 17.
[30] Cardet CED. Analysis on voltage stability indices. RWTH Aachen University;
2012.
[31] Canizares C, Dobson I, Miller N, Ajjarapu V, Caizares C, Hamadanizadeh H,
et al. Voltage stability assessment: concepts, practices and tools. In: IEEE PES
Power System Stability Subcommittee; 2002.
[32] Viral R, Khatod DK. Optimal planning of distributed generation systems in
distribution system: a review. Renew Sustain Energy Rev 2012;16:514665.
[33] Viral R, Khatod DK. An analytical approach for sizing and siting of DGs in
balanced radial distribution networks for loss minimization. Electr
Power Energy Syst 2015;67:191201.
[34] Anaya LK, Michael GP. The role of distribution network operators in promoting
cost-effective distributed generation: lessons from the United States of
America for Europe. Renew Sustain Energy Rev 2015;51:48496.
[35] Kayal P, Chanda CK. Placement of wind and solar based DGs in distribution
system for power loss minimization and voltage stability improvement. Electr
Power Energy Syst 2013;53:795809.

12

J. Modarresi et al. / Renewable and Sustainable Energy Reviews 63 (2016) 112

[36] Murty VVSN, Kumar A. Optimal placement of DG in radial distribution systems

based on new voltage stability index under load growth. Electr Power Energy
Syst 2015;69:24656.
[37] Acharya N, Mahat P, Mithulananthan N. An analytical approach for DG allocation in primary distribution network. Int J Electr Power Energy Syst
2006;28:66978.
[38] Singh AK, Parida SK. Combined optimal placement of solar, wind and fuel cell
based DGs using AHP. In: Proceedings of world renewable energy congress;
2011. p. 311320.
[39] Lee SH, Park JW. Selection of optimal location and size of multiple distributed
generations by using Kalman lter algorithm. IEEE Trans Power Syst
2009;24:1393400.
[40] Hedayati H, Nabaviniaki SA, Akbarimajd A. A method for placement of DG
units in distribution networks. IEEE Trans Power Deliv 2008;23:16208.
[41] Ameli A, Bahrami S, Khazaeli F, Haghifam MR. A Multiobjective particle swarm
optimization for sizing and placement of DGs from DG owner's and distribution company's viewpoints. IEEE Trans Power Deliv 2014;29:183140.
[42] Gampa SR, Das D. Optimum placement and sizing of DGs considering average
hourly variations of load. Int J Electr Power Energy Syst 2015;66:2540.
[43] Machowski J, Bialek J, Bumby J. Power system dynamics: stability and control.
2nd ed. Wiley; 2008. p. 299306.
[44] Pama A, Radman G. A new approach for estimating voltage collapse point
based on quadratic approximation of PV-curves. Electr Power Syst Res
2009;79:6539.
[45] Musirin I, Khawa T, Rahman A. Novel fast voltage stability index (FVSI) for voltage
stability analysis in power transmission system. In: Proceedings of the student
conference on research and development proceedings; 2002. p. 2658.
[46] Subramani C, Dash SS, Bhaskar MA, Jagadeeshkumar M, Sureshkumar K, Parthipan R. Line outage contingency screening and ranking for voltage stability
assessment. In: Proceedings of the international conference on power systems; 2009. p. 15.
[47] Moghavvemi M, Omar FM. Technique for contingency monitoring and voltage
collapse prediction. IEE Proc Gener Transm Distrib 1998;145:63440.
[48] Mohamed A, Jasmon GB, Yusoff S. A static voltage collapse indicator using line
stability factors. J Ind Technol 1989;7:7385.
[49] Moghavvemi M, Faruque MO. Technique for assessment of voltage stability in illconditioned radial distribution network. IEEE Power Eng Rev 2001;21:5860.
[50] Yazdanpanah-Goharrizi A, Asghari R. A novel line stability index (NLSI) for
voltage stability assessment of power systems. In: Proceedings of the international conference on power systems; 2007. p. 1657.
[51] Moghavvemi M, Faruque O. Real-time contingency evaluation and ranking
techniquei. IEE Proc Gener Transm Distrib 1998;145:51724.
[52] Albuquerque MA, Castro CA. A contingency ranking method for voltage stability in real time operation of power systems. In: Proceedings of the Bologna
power tech conference; 2003. p. 15.
[53] Okon T, Wilkosz K. Properties of indices for voltage stability assessment. In:
Proceedings of the international symposium on advanced topics in electrical
engineering; 2013. p. 16.
[54] Kanimozhi R, Selvi K. A novel line stability index for voltage stability analysis
and contingency ranking in power system using fuzzy based load ow. J Electr
Eng Technol 2013;8:66776.
[55] Althowibi FA, Mustafa MW. Line voltage stability calculations in power systems. In: Proceedings of the international conference on power and energy;
2010. p. 396401.
[56] Nizam M, Mohamed A, Hussain A. Dynamic voltage collapse prediction in
power systems using power transfer stability index. In: Proceedings of the
international power and energy coference; 2006. p. 24650.
[57] Gong Y, Schulz N, Guzman A. Synchrophasor-based real-time voltage stability
index. In: Proceedings of the PSCE conference; 2006. p. 102936.
[58] Rahman TKA, Jasmon GB. A new technique for voltage stability analysis in a
power system and improved loadow algorithm for distribution network. In:
Proceedings of the international conference on energy management and
power delivery; 1995. p. 7149.
[59] Makasa KJ, Venayagamoorthy GK. On-line voltage stability load index estimation based on PMU measurements. In: Proceedings of the IEEE power and
energy society general meeting; 2011. p. 16.
[60] Sahari S, Abidin AF, Rahman TKA. Development of articial neural network for
voltage stability monitoring. In: Proceedings of the power and energy conference; 2003. p. 3742.
[61] Guiping D, Yuanzhang S, Jian X. A new index of voltage stability considering
distribution network. In: Proceedings of the power and energy engineering
conference; 2009. p. 14.
[62] Wang L, Liu Y, Luan Z. Power transmission paths based voltage stability
assessment. In: Proceedings of the transmission and distribution conference
and exhibition: Asia and Pacic; 2005. p. 15.
[63] Wang Y, Li W, Lua J. A new node voltage stability index based on local voltage
phasors. Electr Power Syst Res 2009;79:26571.
[64] Chattopadhyay TK, Banerjee S, Chanda CK. Impact of distributed generator on
voltage stability analysis of distribution networks under critical loading conditions for composite loads. In: Proceedings of the international conference on
electronics, communication and instrumentation; 2014. p. 14.
[65] He T, Kolluri S, Mandal S, Galvan F, Rastgoufard P. Identication of weak locations
in bulk transmission systems using voltage stability margin index. In: Proceedings
of the IEEE Power Engineering Society General Meeting; 2004. p. 18149.

[66] Milosevic B, Begovic M. Voltage-stability protection and control using a widearea network of phasor measurements. IEEE Trans Power Syst 2003;18:1217.
[67] Eminoglu U, Hocaoglu MH. A voltage stability index for radial distribution
networks. In: Proceedings of the Universities' power engineering conference;
2007. p. 40813.
[68] Tiwari R, Niazi KR, Gupta V. Line collapse proximity index for prediction of
voltage collapse in power systems. Int J Electr Power Energy Syst
2012;41:10511.
[69] Suggested techniques for voltage stability analysis. Technical report
93TH0620-5PWR, IEEEjPES; 1993.
[70] Indices predicting voltage collapse including dynamic phenomena. Technical
report TF 38-02-11, CIGRE; 1994.
[71] Seydel R. Practical bifurcation and stability analysis-from equilibrium to
Chaos. 2nd ed.. New York: Springer-Verlag; 1994.
[72] Wang L, Girgis AA. Online detection of power system small disturbance voltage instability. IEEE Trans Power Syst 1996;11:130413.
[73] ACZd Souza, Caiiizares CA, Quintana VH. New techniques to speed up voltage
collapse computations using tangent vectors. IEEE Trans Power Syst
1997;12:13807.
[74] Acharya NV, Rao PSN. A new voltage stability index based on the tangent
vector of the power ow Jacobian. Asia: Innovative Smart Grid Technologies;
2013. p. 16.
[75] Balamourougan V, Sidhu TS, Sachdev MS. Technique for online prediction of
voltage collapse. IEE Proc Gener Transm Distrib 2004;151:45360.
[76] Kessel P, Glavitsch H. Estimating the voltage stability of a power system. IEEE
Trans Power Deliv 1986:34654.
[77] Zabaiou T, Dessaint LA, Kamwa I. Preventive control approach for voltage
stability improvement using voltage stability constrained optimal power ow
based on static line voltage stability indices. IET Gener Transm Distrib
2014;8:92434.
[78] Hongjie J, Xiaodan Y, Yixin Y. An improved voltage stability index and its
application. Int J Electr Power Energy Syst 2005;27:56774.
[79] Haque MH. Use of local information to determine the distance to voltage
collapse. In: Proceedings of the international power engineering conference;
2007. p. 40712.
[80] Smon I, Verbic G, Gubina F. Local voltage-stability index using Tellegens
theorem. IEEE Trans Power Syst 2006;21:126775.
[81] Li W, Wang Y, Chen T. Investigation on the Thevenin equivalent parameters
for online estimation of maximum power transfer limits. IET Gener Transm
Distrib 2010;4:11807.
[82] STsai Sj, Wong KH. On-line estimation of Thevenin equivalent with varying
system states. In: Proceedings of the Power and Energy Society general
meeting conversion and delivery of electrical energy; 2008. p. 17.
[83] Mou X, Weixing L, Li W. A preliminary study on the Thevenin equivalent
impedance for power systems monitoring. In: Proceedings of the international
conference on electric utility deregulation and restructuring and power
technologies; 2011. p. 7303.
[84] Wiszniewski A. New criteria of voltage stability margin for the purpose of load
shedding. IEEE Trans Power Deliv 2007;22:136771.
[85] Perez-Londonoa S, Rodrigueza LF, Olivar G. A simplied voltage stability index
(SVSI). Int J Electr Power Energy Syst 2014;63:80613.
[86] Yesuratnam G, Thukaram D. Congestion management in open access based on
relative electrical distances using voltage stability criteria. Electr Power Syst
Res 2007;77:160818.
[87] Alvarado FL, Jung TH. Direct detection of voltage collapse conditions. In: Bulk
power system voltage phenomena voltage stability and security; 1989. p. 5.2338.
[88] Yorino N, Harada S, Cheng H. A method to approximate a closest loadability limit
using multiple load ow solutions. IEEE Trans Power Syst 1997;12:4249.
[89] Nagao T, Tanaka K, Takenaka K. Development of static and simulation programs for voltage stability studies of bulk power system. IEEE Trans Power
Syst 1997;12:27381.
[90] Berizzi A, Finazzi P, Dosi D, Marannino P, Corsi S. First and second order
methods for voltage collapse assessment and security enhancement. IEEE
Trans Power Syst 1998;13:5439.
[91] Tamura Y, Sakamoto K, Tayama Y. Current issues in the analysis of voltage
instability phenomena. In: Bulk power system voltage phenomena voltage s.
tability and security; 1989. p. 5-3953.
[92] Barquin J, Gomez T, Pagola FL. Estimating the loading limit margin taking into
account voltage collapse areas. IEEE Trans Power Syst 1995;10:195262.
[93] DeMarco CL. A new method of constructing Lyapunov functions for power
systems. In: Proceedings of the IEEE international symposium on circuits and
systems; 1988. p. 9058.
[94] Overbye TJ, DeMarco CL. Voltage security enhancement using energy based
sensitivities. IEEE Trans Power Syst 1991;6:1196202.
[95] CIGRE WG 39.02-(Control Center Performance). Possibilities and expectations for
improved man machine interface in power system control. CIGRE; 1992. p. 35/9-03.
[96] Sandberg L, Rouden K, Ekstam L. Security assessment against voltage collapse
based on real-time data including generator reactive power capacity. GIGRE1994. p. 39/11-03.
[97] Althowibi FA, Mustafa MW. Power system network sensitivity to voltage
collapse. In: Proceedings of the international power engineering and optimization conference; 2012. p. 37983.
[98] http://www.ee.washington.edu/research/pstca/dyn30/pg_tcadyn30.htm;
2015 [last accessed 04.11.15].