First International Conference on Emerging Trends in Engineering and Technology

A Comparison of Voltage Stability Indices for Placing Shunt FACTS

Controllers

A. R. Phadke S. K. Bansal K. R. Niazi

Malviya National Institute of Technology, Jaipur (Rajastan), India
phadkeatul@gmail.com bansal_sanjay30@yahoo.com krniazi@mnit.ac.in

Abstract capacitors and/or FACTS controllers at the appropriate

location. Introducing FACTS devices is the most
Location of shunt compensation devices is effective way for utilities to improve the voltage
important for the enhancement of the voltage stability profile and voltage stability margin of the system [8].
for practical power systems. This paper presents a However, to obtain good performance from these
comparison of several voltage stability indices in controllers, proper placement of these devices in the
electric power system to identify the weakest bus/ area grid is important.
of the system. Shunt FACTS controller is introduced at The best location for reactive power compensation
the weakest bus in IEEE 14-bus test system and its for improving steady state voltage stability margin is
effectiveness is assessed by comparing voltage profile the weakest bus in the system [7, 16, 18]. A review of
and loading margin enhancement. Various line the literature on methods or techniques to study voltage
stability indices are also compared with and without collapse reveals that various analytical tools based on
shunt FACTS controller. It is shown that the best different concepts have been proposed to predict
location for reactive power compensation for voltage collapse. These indices provide reliable
improving static voltage stability margin is the information about the proximity of voltage instability
“weakest bus” of the system. and the weakest bus/area and line in the system.
Several methods have been proposed in the literature
for conducting the voltage stability analysis such as the
Key Words- Static voltage stability, FACTS P-V and Q-V curves [2, 3, 5, 6], modal analysis [17],
controllers, loading margin, placement L-index [9], V/V0 index [3]. The indices derived for
voltage stability analysis are either referred to a bus or
line. Line stability indices [10-12] can be used to
1. Introduction determine the weakest line in the system and to
evaluate the voltage stability condition, whereas bus
Due to ever increasing load demand, power stability indices can be used to identify the weakest bus
utilities are now forced to increase the utilization of or area in the system.
existing transmission facilities. It is quite difficult to In this paper the effectiveness of various voltage
construct new lines due to environmental and stability indices is compared for the identification of
economic considerations. Thus modern power systems the weakest bus in the system. The performance of
are forced to carry increasingly more power over long shunt FACTS controller connected at the weakest bus
distances. One of the major problems that may is assessed by comparing voltage profile and steady
associate with such a stressed system is the voltage state voltage stability margin of the system. The
instability or collapse. Many incidents of system performance is also investigated by comparing various
blackout due to voltage collapse have been reported line stability indices.
worldwide [5, 6]. Voltage stability is the ability of a Rest of the paper is structured as follows: Section
power system to maintain steady voltages at all buses 2 describes various voltage stability indices used to
in the system after being subjected to a disturbance examine the stability of the system. Test systems and
from a given initial operating condition [1, 2, 3]. analysis tools are given in section 3. In section 4
Under stressed condition, the only way to save the results are presented along with discussion. Finally,
system from voltage collapse is to reduce the reactive conclusions are summarized in section 5.
power load or add additional reactive power by
introducing the sources of reactive power, i.e., shunt

978-0-7695-3267-7/08 $25.00 © 2008 IEEE 939

DOI 10.1109/ICETET.2008.30
2. Voltage stability indices  g  F ji 
L j = MAX 1 − ∑   (1)
i =1 
 V 
The purpose of voltage stability indices is to  j 
determine the point of voltage instability, the weakest
bus in the system and the critical line referred to a bus. 2.5. Modal analysis
These indices are referred either to a bus or a line.
These indices are briefly discussed in this section. In modal analysis, the smallest eigenvalue and
associated eigenvectors of the reduced Jacobian matrix
2.1. P-V curve are computed based on the steady state system model.
Eigenvalue of the reduced Jacobian matrix close to
As the power transfer increases, the voltage at the zero indicates that the system is on the border of
receiving end decreases. Finally, the critical or nose voltage instability [6, 17]. The closeness of the system
point is reached. It is the point at which the system to voltage collapse situation can be predicted through
reactive power is out of use. The curve between the the evaluation of the minimum positive eigenvalue.
variation of bus voltages with loading factor (λ) is The weakest bus of the system can be determined by
popularly called as P-V curve or ‘Nose’ curve. PV computing the eigenvector for different buses in the
curves are used to determine the loading margin of the system.
power system. The margin between the voltage
collapse point and the current operating point is used as 2.6. Line stability index Lmn
voltage stability criterion [2, 3, 5, 6].
This stability criterion is used to find the stability
2.2. Q-V curve index for each line connected between two busbars in
an interconnected network. As long as the stability
With the help of Q-V curve, it is possible for the index Lmn remains less than 1, the system is stable. This
operators, to know the maximum reactive power that voltage stability criterion is based on a power
can be added to the bus before reaching minimum transmission concept in a single line [10]. Stability
voltage limit. The MVAr distance from the operating criterion is developed considering a single line of a
point to the bottom of the Q-V curve is called as the network as shown in Fig. 1.
reactive power margin. Q-V curve can be used as an j
i
index for voltage stability limit. The point where
dQ Si, Pi, Qi
= 0 is the point of voltage stability limit [2, 3, 5]. Sj, Pj , Qj
dV
Fig. 1. Typical one line diagram of a
2.3. V/ V0 Index transmission line.
Line stability index Lmn is given by –
The ratio V/V0 at each bus shows the voltage 4X Qj
stability map of the system. V is the bus voltage at Lmn = (2)
certain load obtained from load flow study. Voltages [
V i sin θ −δ
2
( )]
V0 are obtained by solving load flow of the system at Where -
an identical state but with all the loads set to zero. This θ = line impedance angle
index allows immediate detection of weakest bus and δ = angle difference between the sending end and the
corrective action can be taken to prevent the voltage receiving end voltage
instability [3, 13]. X = line reactance
Qj = reactive power flow at the receiving end
2.4. L-index Vi = sending end voltage.
To maintain a secure condition, should be
L-index estimates the distance between the present
less than 1.
state of the system and the stability limit. L-index
describes the stability of the complete system and it
varies in a range between 0 (no load) to 1 (voltage 2.7. Line stability index FVSI
collapse). Bus with highest L-index is the most
The line stability index FVSI is also based on the
vulnerable bus in the system [9]. Using load flow
concept of power flow through a single line. For a
results the L-index can be computed as –

940
typical transmission line, the stability index FVSI is 4. Results and Discussion
calculated by equation given in [11] as:
4Z Qi
2
The system described in the previous section is
FVSI ij = 2
(3) simulated and load flow equations are solved to
Vi X determine the weakest bus in the system. The results of
Where - comparison and the performance of the system with
Z = line impedance and without shunt FACTS controller are presented in
X = line reactance this section.
Qj = the reactive power flow at the receiving end
Vi = sending end voltage. 4.1. PV curves
The line that exhibits FVSI close to 1.00 implies
that it is approaching the instability point. If FVSI goes P-V curves show the bus voltage level as the
beyond 1, one of the buses connected to the line loading factor λ increases. The loading factor is 1 at
experiences a sudden voltage drop leading to system the base case and it is gradually increased, until
collapse. maximum loading point is reached. The load at all the
buses is increased maintaining the power factor
2.8. Line stability index LQP constant. Fig. 2 shows the P-V curves for IEEE 14 bus
Line stability index LQP can be obtained from the test system. The maximum permissible loading limit
equation given in [4] as – for a system without shunt FACTS controller is
approximately equal to 1.72. PV curves indicate that
 X  X 
LQP = 4 2  2 P i + Q j 
2
(4) bus no. 14 is the weakest bus in the system.
  
V
 i  iV  Bus 1
1.05
Where – Bus 2
Bus 3
X = line reactance Bus 4
Qj = reactive power flow at the receiving bus 0.95 Bus 5
Vi = voltage at the sending bus and
Volatge (p.u.)

Bus 6
Pi = active power flow at the sending bus. Bus 7
0.85 Bus 8
To maintain a secure condition, LQP index should
Bus 9
be maintained below 1. Bus 10
0.75 Bus 11
3. Test system and analysis tools Bus 12
Bus 13
0.65 Bus 14
The system studied in this paper is IEEE 14-bus 1.3 1.4 1.5 1.6 1.7
test system. System data is obtained from [20]. All the Loading Factor
results are produced with the help of a program Fig. 2. P-V curves for IEEE 14-bus test system.
developed in MATLAB. In this paper, unified
approach suggested in [14, 15] is used for solving load
4.2. QV curves
flow equations of the system consisting a shunt
FACTS controller. The unified or simultaneous
Q-V curve gives the maximum reactive power
approach combines the state variables of FACTS
demand at a particular bus that can be increased before
devices with the state variables of network in a single
the system suffers a voltage collapse. Maximum
frame of reference. This set of non linear algebraic
reactive power margin at different buses in IEEE 14-
equations is then solved by the Newton-Raphson
bus test system is shown in Fig. 3. Busbars 14 and 12
algorithm. The unified method retains Newton’s
have the lowest margin of reactive power. This
quadratic convergence characteristics.
indicates that these are the most critical buses in the
The best location for the shunt reactive power
system.
compensation, as far as the improvement of steady
state voltage margin is concerned, is the weakest bus of
the system [7, 16, 18]. From simulation results the
weakest bus is identified and the shunt FACTS
controller is connected at the weakest bus. The voltage
magnitude at this bus is maintained at 1 p.u.

941
 300 factor λ = 0.4. This figure shows that the bus no. 14 is
the weakest bus of the system.
Reactive power margin (MVAr)

250

200 0.5

150 0.4

0.3

L index
100

50 0.2

0.1
0
4 5 9 10 11 12 13 14
0
Bus numbe r
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Fig. 3. Reactive power margin in IEEE 14-bus Bus No.
test system. Fig. 5. L indices for IEEE 14-bus test system at
loading factor λ = 0.4.
4.3. V/ V0 index
0.4
This index is very simple to compute. Fig. 4 shows
the V/V0 index without shunt FACTS controller plotted 0.3
at a loading factor λ = 0.4. At this loading as the
Eigenvector
system is away from the voltage instability point, as 0.2
expected, the voltages at the busbars are nearly equal
to the voltage V0. It can be noticed from the figure that 0.1
without FACTS controller, the weakest bus in the
system is bus no. 14. 0
1 2 3 4 5 6 7 8 9 10 11 12 13 14
1 Bus No.

0.8 Fig. 6. Eigenvector for IEEE 14-bus test

V / V0 Index

system at loading factor λ = 0.4.

0.6

0.4 Simulation results shown above indicate that bus

no. 14 of the IEEE 14-bus test system is the weakest
0.2
bus in the system. These stability indices provide
0 accurate information regarding the stability condition.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 Table 1 gives the ranking of various buses according to
Bus No.
the different voltage stability indices.
Fig. 4. V/V0 index for IEEE 14-bus test system
at loading factor λ = 0.4. 4.6. Placement and sizing of shunt FACTS
controller
4.4. L-index
From Table 1, bus no. 14 could be considered as
In Fig. 5, L-indices for various buses are plotted at the best location for providing a shunt reactive power
loading factor λ = 0.4. At this loading, the value of L- support. The capacity of shunt FACTS controller can
index at various buses is less than 1 because at this be determined by plotting the relationship between the
load the voltage stability of the system is guaranteed. maximum loading factor and the corresponding
Again bus no. 14 is the weakest bus. capacity of the shunt controller. This relationship is
given in Fig. 7. From Fig. 7 it is clear that the optimum
4.5. Modal analysis capacity required is about 100 MVAr. Thus a shunt
FACTS controller of 100 MVAr is connected at bus
At different loading factors, eigenvalues of the no. 14 and its voltage setting is kept at 1 p.u.
reduced Jacobian matrix are computed. Fig. 6 shows
the eigenvector for different buses for IEEE 14-bus test
system without shunt FACTS controller at a loading

942
 Table 1. Bus ranking according to various 1.1 Without FACTS controller
voltage stability indices With FACTS controller
1.05

Voltage (p.u.)
R Index 1
a PV React. V / V0 L Eig-
n curve power Index index vect. 0.95
k margin 0.9
1 14 14 14 14 14
2 13 12 13 13 12 0.85
3 10 6 10 10 13 1 2 3 4 5 6 7 8 9
Bus No.
10 11 12 13 14

4 12 11 12 12 11
5 11 13 9 9 10 Fig. 8. Voltage profile of IEEE 14-bus test
6 9 10 11 11 9 system with and without shunt FACTS
7 6 9 6 6 6 controller at loading factor λ = 0.5.
8 7 4 7 3 8
9 8 8 4 7 7 The purpose of line stability indices is to quantify
10 3 5 5 8 4 the closeness of particular operating point to the
11 4 3 3 4 5 voltage collapse, i.e. to assess the system security. To
investigate the effect of shunt controller on the line
12 5 7 8 5 3
indices discussed in section 2, the system is simulated
13 2 2 2 2 2
with large active and reactive power load at all the
14 1 1 1 1 1
buses. Line indices for three highly stressed lines with
and without shunt controller are presented in Table 2.
2.1 Table 2 indicates the stressed condition of the lines
without shunt controller. From Table 2, it is observed
Maximum loading

2 that line that connects bus 7 to 8 is the most critical

line and the system is near to the point of voltage
1.9 collapse. Shunt controller connected at the weakest bus
shows good improvement in the line stability indices.
1.8

Table 2. Line stability indices for highly

1.7
0 20 40 60 80 100 120 140 160 180 200
stressed lines
Capacity of Shunt FACTS controller
Line Lmn FVSI LQP
Fig. 7. Loading factor vs controller capacity of Base case 0.954 0.796 0.954
a shunt controller. 7–8
With controller 0.494 0.445 0.494
Base case 0.373 0.569 0.497
4.7. Performance of the system with shunt 1–5
With controller 0.186 0.255 0.301
FACTS controller
Base case 0.367 0.423 0.378
7–9
With controller 0.182 0.179 0.171
The system described in the previous section is
simulated for assessing the effectiveness of the shunt
Table 3 shows the line stability indices for lines 9
FACTS controller for voltage profile improvement and
– 14 and 13 – 14. From the table it may be observed
increase in the maximum permissible loading. The
that with shunt controller there is increase in line
voltage profile at various buses in the system at a
indices. This is due to the fact that bus no. 14 is a
loading factor λ = 0.5 as compared to the base case is controller bus which is now supplying the reactive
shown in Fig. 8. It clearly shows that the use of shunt support to all other buses through these lines.
controller at the weakest bus results in an improved
network voltage profile. The reactive power support at
the weakest bus results in an improved network voltage
profile. Maximum permissible loading is also
increased from 1.72 to 2.01.

943
 Table 3. Line stability indices for the lines [8] N. G. Hingorani and L. Gyugyi, Understanding FACTS:
connected to bus no. 14 (Controller bus) Concepts and Technology of Flexible AC Transmission
Systems, New York: IEEE Press, 2000.
[9] P. Kessel and H. Glavitsch, “Estimating the Voltage
Line Lmn FVSI LQP Stability of a Power System”, IEEE Transactions on Power
Base case 0.243 0.266 0.206 Delivery, vol.PWRD-1, July 1986.
9-14 With 0.494 0.410 0.381 [10] M. Moghavvemi, and F. M. Omar, “Technique for
controller Contingency Monitoring and Voltage Collapse Prediction”,
Base case 0.081 0.088 0.070 IEE Proceeding Generation, Transmission and Distribution,
13- 14 With 0.438 0.375 0.351 vol. 145, pp. 634-640 November 1998.
[11] I. Musirin, and T. K. A. Rahman, “Novel Fast Voltage
controller Stability Index (FVSI) for Voltage Stability Analysis in
Power Transmission System”, Student Conference on
5. Conclusion Research and Development Proceedings, Shah Alam,
Malaysia, July 2002.
In this paper, strategy for placement and sizing of [12] M. Moghavvemi, and O. Faruque, “Real-Time
shunt controller is presented to obtain maximum Contingency Evaluation and Ranking Technique”, IEE
Proceeding Generation, Transmission and Distribution, vol.
improvement in the voltage profile and enhancement in
145, September 1998.
steady state voltage stability margin. In the system [13] Claudia Reis and F.P. Maciel Barbosa, “A Comparison
studied, it has been found that various bus and line of Voltage Stability Indices,” IEEE MELECON (Spain) pp.
stability indices are able to identify the critical lines 1007-1010, 2006
and buses correctly. The performance of the indices [14] C. R. Feurte Esquivel and E. Acha, “A Newton-type
shows high degree of accuracy and reliability. Most of algorithm for the control of power flow in electrical power
the indices are very simple in nature and easy to networks,” IEEE Trans. Power Syst., vol. 12, pp. 1474–1481,
calculate. Suitable shunt FACTS controller can be Nov. 1997.
connected at the weakest bus/area in order to improve [15] E. Acha, C. R. Fuerte-Esquivel, H. Ambriz-Pérez and C.
Angeles-Camacho, FACTS: Modelling and simulation in
the performance of the system.
power networks, Wiley, 2004.
However, it is possible that the best location of [16] C. A. Canizares, A. C. Z. de Souza, and V. H. Quintana,
controller from voltage stability viewpoint may not be “Comparison of performance indices for detection of
the best location from transient stability viewpoint. proximity of voltage collapse”, IEEE Trans. Power Syst., vol.
11, pp. 1441-1450, Aug. 1996.
6. References [17] B.Gao, G.K.Morison and P. Kundur “Voltage Stability
Evaluation Using Modal Analysis”, IEEE, Transactions on
Power Systems, Vol.7, November 1992.
[1] P. Kundur, J. Paserba, V. Ajjarapu, G. Andersson, A.
[18] A. Sode-Yome and N. Mithulananthan, “Comparison of
Bose, C. Canizares, N. Hatziargyriou, D. Hill, A. Stankovic,
shunt capacitor, SVC and STATCOM in static voltage
C Taylor, T. Van Cutsem, and V. Vittal, “Definition and
stability margin enhancement”, International Journal of
classification of power system stability IEEE/CIGRE joint
Electrical Engineering Education, UMIST, Vol. 41, No. 3,
task force on stability terms and definitions”, IEEE
pp. 158-171, July 2004.
Transactions on Power Systems, vol. 19, Issue 3, pp. 1387-
[19] B. H. Lee and K. Y. Lee, “Dynamic and Static voltage
1401, Aug. 2004.
stability enhancement of power systems”, IEEE Trans.
[2] Western Electricity Coordinating Council, Guide to
Power Syst., vol. 8, pp. 231-238, Feb. 1993.
WECC/NERC Planning Standards I.D: Voltage Support and
[20] Power System Test Archive-UWEE (University of
Reactive Power, March 30, 2006
Washington) available Online,
[3] IEEE/PES Power System Stability Subcommittee,
http://www.ee.washington.edu/research/pstca
Voltage Stability Assessment: Concepts, Practices and Tools,
[21] J. Zhang, J. Y. Wen, S. J. Cheng and , J. Ma, “A Novel
special publication, final draft, Aug. 2003.
SVC Allocation Method for Power System Voltage Stability
[4] Wiszniewski A., “New Criteria of Voltage Stability
Enhancement by Normal Forms of Diffeomorphism”, IEEE
Margin for the Purpose of Load Shedding”, IEEE
Transactions on Power Systems, vol. 22, Issue 4, pp. 1819 –
Transactions on Power Systems, vol. 22, pp. 1367 – 1371,
1825, Nov. 2007
July 2007.
[5] C. W. Taylor, Power System Voltage Stability, McGraw-
Hill, Inc. 1994.
[6] P. Kundur, Power System Stability and Control,
McGraw-Hill, 1994.
[7] A. Sode-Yome, N. Mithulananthan and Y. Lee Kwang,
“A Comprehensive Comparison of FACTS Devices for
Enhancing Static Voltage Stability”, IEEE Power
Engineering Society General Meeting, 2007.

944