You are on page 1of 9

This article has been accepted for inclusion in a future issue of this journal.

Content is final as presented, with the exception of pagination.


IEEE TRANSACTIONS ON POWER SYSTEMS

Estimating the Voltage Stability Margin Using


PMU Measurements
Heng-Yi Su, Member, IEEE, and Chih-Wen Liu, Fellow, IEEE

AbstractThis paper presents a new method based on phasor


measurement units (PMUs) for the estimation of voltage stability
margin in a power system to increase operator's situational awareness. The method assumes a PMU measurement preprocessing
technique as a priori in order to eliminate data inconsistency and
uncertainty caused by random load disturbances. The measurements are then used in the computation of voltage stability margin
based on the coupled single-port Thevenin equivalent model and
the cubic spline extrapolation technique. Moreover, some practical
operating constraints such as the generator reactive power limits
are taken into account for practical assessment of the method's
performance. Extensive case studies conducted on several standard IEEE test systems are used to demonstrate the effectiveness
of the proposed method.
Index TermsPhasor measurement unit (PMU), voltage stability, voltage stability margin, wide-area measurement system
(WAMS).

I. INTRODUCTION

OLTAGE instability has always been a crucial concern


by power system engineers over the past two decades.
Signicant research efforts have been devoted to understanding
voltage instability phenomenon and proposing corresponding
countermeasures [1], [2]. The phenomenon of voltage instability is caused by an uncontrollable drop in system voltage after
being subjected to a disturbance. This deterioration may ultimately result in voltage collapse that has been responsible for
several blackout incidents throughout the world [3]. No matter
how infrequent power system blackout is, this may charge additional cost, and everyone would always be affected.
Voltage stability is an essential dynamic phenomenon. However, a large portion of the research studies has been focused on
the steady-state aspects of voltage stability [4]. In spite of the
fact that the static analysis is simple, it still provides some practical advantages over the dynamic analysis, and an example is
to give results with acceptable accuracy and little computational
effort [5]. In this paper, the main concern is static voltage stability analysis.
Manuscript received April 23, 2015; revised July 29, 2015; accepted
September 03, 2015. This work was supported in part by the Ministry
of Science and Technology of Taiwan, R.O.C., under Contract MOST
103-2221-E-035-101 and Contract MOST 104-3113-E-002-013. Paper no.
TPWRS-00562-2015.
H.-Y. Su is with the Department of Electrical Engineering, Feng Chia University, Taichung 40724, Taiwan, R.O.C. (e-mail: hengyisu@fcu.edu.tw).
C.-W. Liu is with the Department of Electrical Engineering, National Taiwan
University, Taipei 10617, Taiwan, R.O.C. (e-mail: liucw@ntu.edu.tw).
Digital Object Identier 10.1109/TPWRS.2015.2477426

In order to measure steady-state voltage stability of a power


system, the concept of voltage stability margin (VSM) has been
proposed to demonstrate the closeness of the current operating
point to the point of voltage collapse [6]. A review of literature reveals that there are different kinds of technique for VSM
evaluation, such as sensitivity [7], [8], singular value decomposition [9], [10], and continuation power ow (CPFLOW) [11],
[12] method. Among those model-based approaches, CPFLOW
based methods are widely used to trace a solution path from any
given operating point to voltage-collapse point. Since modelbased tools require a lot of computation time for a buck power
system model, they may not be applied in the eld of real-time
applications in practical power systems; therefore, an alternative approach is required.
With increasing deployment of phasor measurement units
(PMUs) on transmission systems, PMU-based wide area
measurement system (WAMS) has already attracted lots of
concerns from the academia and industry [13]. PMU-based
WAMS is able to provide time-synchronized measurements
in both voltage and current phasors over wide areas. Each
measured data from PMU is time-stamped based on a common
time reference with a high precision provided by global positioning system (GPS). According to the time stamps and
sample numbers, the collected phasor data are synchronized
and correlated. The real-time, accurate, and time-aligned phasor
measurements enable to create a precise snapshot of the power
system; therefore, WAMS is introduced to perform a wide
range of applications [14]. Recently, some online tools using
PMU measurements have been successfully applied to voltage
stability analysis, such as regression tree [15], relationships
exploration [16], channel components transform [17], and
modied coupled single-port model [18]. Indeed, most of the
measurement-based methods are the impedance match methods
[18][25]. Such approaches are based on an estimated Thevenin
equivalent network behind a load bus. The works [19][24]
handle such problem by using a single-port model. To improve
the performance of the single-port model under multiple load
change scenario, an innovative method based on the coupled
single-port model was proposed in [25]. An extension of [25]
was proposed in [18] to modify the couple single-port model by
introducing a mitigation factor for measurement-based voltage
stability assessment.
In recent years, many measurement-based approaches have
been proposed to determine voltage stability of a power system.
However, to the author's knowledge, most of them do not
consider generator reactive power limits. Under these circumstances, the performances of these methods would be severely
degraded.

0885-8950 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.
2

IEEE TRANSACTIONS ON POWER SYSTEMS

In this context, the contribution of this paper is the development of a new approach to estimate the maximum
loading point, considering the generator reactive power limits,
for real-time voltage stability monitoring. To this end, the
PMU-based WAMS technology is employed to improve
voltage stability monitoring through the increased situational
awareness of system operators. In the proposed method, the
coupled single-port Thevenin equivalent model and the cubic
spline extrapolation technique are used to determine the point
of voltage collapse. Furthermore, the developed method is
able to handle limits violation cases by calculating an index
which measures the extreme reactive power of each generator
for a given load increase direction; therefore, it is suitable for
practical use.
The succeeding sections of this paper are organized as
follows: Section II reviews the main concepts of coupled
single-port circuit and impedance match method. The proposed
method for real-time voltage stability monitoring is presented
in Section III. Numerical examples and results are illustrated in
Section IV. Concluding remarks are drawn in Section V.

Fig. 1. Coupled single-port circuit used in [18] and [25]: an extra impedance
is used to represent coupling effects of other loads.

where
(7)
(8)
is called the impedance matrix and can be obtained from the
system admittance matrix . For a specic load bus , we have

II. PRELIMINARIES

(9)

This section gives a brief overview of the coupled singleport circuit and reviews the concept of impedance match for the
estimation of the maximum loading point.
A. Coupled Single-Port Circuit
A general multi-node network for a power system can be
converted into a multi-branch Thevenin equivalent circuit. To
start with, consider the standard node-voltage equation in matrix from
(1)
where is the vector of the injected bus currents, is the vector
of bus voltages, and is known as the system admittance matrix. The system buses are generally classied into three types:
generator bus
, load bus
, and tie bus
which has no
generator or load. Thus, expanding (1) leads to the expression

where
is the diagonal element of
and
is the
element of . Note that the third term in (9) denotes the coupling
effects of other loads on bus .
Several methods were proposed to model such coupling term
while maintaining the single-port structure [18], [25]. The literature [25] employs an extra impedance, which is called the virtual impedance, to represent the coupling effects of other loads,
as shown in Fig. 1. To improve the coupled single-port circuit
under the cases when loads in different areas increase at different rates and keep their power factors constant, a modied
coupled single-port circuit is developed [18]. Specically, a mitigation factor , which is derived from the extended Ward-type
equivalent, is introduced to modify the Thevenin equivalent parameters, i.e.,

(2)
(10)
Since the current vector entering the tie bus is equal to zero,
can be written as
(3)
Rearranging (3) gives
(4)
Substituting for

in (2) yields

Notice that
and can be calculated from two consecutive PMU measured data [18]. Observations from a great
amount of simulations under the cases with proportional load increase,
, which means that the modied coupled singleport model is approximately equal to the coupled single-port
model when the loads of the whole system increase at the same
rate and keep their power factors constant.
B. Impedance Match Concept

(5)
Equation (5) can be rearranged as
(6)

Most of the measurement-based methods rely on impedance


match theorem to identify voltage instability or the proximity to
voltage collapse. A key idea of these methods can be summarized as follows. When maximum loading condition occurs at a
load bus, the Thevenin equivalent circuit satises the following
criterion:

This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.
SU AND LIU: ESTIMATING THE VOLTAGE STABILITY MARGIN USING PMU MEASUREMENTS

Fig. 2. Coupled single-port Thevenin equivalent model used in this paper: an


extra voltage source is used to represent coupling effects of other loads.

(11)
In other words, voltage collapse takes place when the impedance
match theory holds. That is, the critical point at voltage collapse
is also the point in which the maximum available power can
be delivered to the load. Based on this fact, a new method to
estimate the maximum loading point by using the extrapolation
technique is proposed. The details are given in Section III.B.
III. PROPOSED METHOD
In this section, a PMU-based approach to real-time voltage
stability margin estimation is proposed. It is necessary to lter
out unwanted random load induced dynamics and errors from
PMU measurements before performing the proposed voltage
stability margin computation. To this aim, one can use ltering
techniques such as the published methods [26], [27] to create
clean measurements. These data preprocessing techniques are
not addressed here. Instead, we focus on the method of voltage
stability margin computation using assumed clean PMU measurements. The proposed method is able to deal with the cases
which cause generators reaching their reactive power limits. The
techniques used in the proposed method are described in the following subsections.
A. Coupled Single-Port Thevenin Equivalent Model With an
Extra Voltage Source
The proposed method is based on the coupled single-port
Thevenin equivalent circuit. Fig. 2 illustrates the circuit representation of (9), in which the coupling term is modeled by
an extra voltage source. Thus, the Thevenin equivalent voltage
and Thevenin equivalent impedance at the th load bus can be
expressed as

Fig. 3. Trajectories of
and
is identical to
.

as a function of load parameter . At

where
and
are the load voltage and current phaosrs. In
this study, we assume that the data points of
and
are
available from PMU measurements. In addition, the load model
considered in this research is the constant PQ model.
B. Cubic Spline Extrapolation Technique
Fig. 3 depicts the trajectories of
and
as a function
of load parameter for the Thevenin's network of Fig. 2. When
the maximum power transfer takes place, the impedance match
theory holds. Therefore, the estimation of the maximum loading
point at load bus , denoted by
, can be made by equating
an approximate function that extrapolates the trajectory of
,
to be
.
In this study, the cubic spline extrapolation method is utilized as the function approximation. The reason why cubic
spline function is used as an extrapolation function is that
this function ts the impedance trajectory quite well based on
observed extensive simulation results. The fundamental idea
behind constructing a spline function is to divide the interval
into a collection of subintervals and construct a different approximating low-order polynomial in each subinterval. In cubic
splines, third-order polynomials are used for interpolation in the
subintervals between each successive pair of data points (the
points are connected with curves). Thus, the curves obtained
from each subinterval are smooth.
For given points, there are
subintervals, the mathematical formula of the polynomial in the th subinterval is given
by
(14)
for each
. Overall, there are
equations,
and since each equation has four coefcients, a total of
coefcients have to be determined. The coefcients
are found by applying the following conditions [28]:

(12)
From (12), it is obvious that the Thevenin equivalent impedance
remains the same if there are no changes in system topologies
and bus types. Based on (12), a coupled single-port Thevenin
equivalent model as viewed from the load bus can be acquired.
Referring to Fig. 2, the load impedance
is given by

(13)

(15)
Cubic splines with the boundary conditions given in part (f) of
the denition are called clamped cubic splines. Applying all the
conditions gives a system of

This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.
4

IEEE TRANSACTIONS ON POWER SYSTEMS

linear equations for the


coefcients. The system of linear
equations that has to be solved for the coefcient is given by

Consider the current entering the typical bus of the power


system is given by (1). Expressing (1) in polar form, we have
(20)

(16)
where
and
. Once the values
of are determined, it is a simple matter to nd the remainder
of the constants
, and
by

(17)
Details on how to determine the coefcients in (14) are presented in [28].
In this work, the inputs to the cubic spline extrapolation function are 3 sets of
and and the point of
at which
is to be extrapolated. The output from the function is the
extrapolated value at the given point. Since the impedance
matches when the maximum power transfer occurs, the function is then used for determining the extrapolated value of for
. In other words, the output from the function is
the estimated maximum loading point
at the th load bus.
In general, the estimated maximum loading point
from
each load bus is different. According to [25], the bus with the
smallest maximum loading point among all load buses is the
weakest bus in the power system, i.e., the smallest maximum
loading point, denoted by
, can be utilized to represent the
maximum loading point of the entire power grid:

where

is the bus admittance matrix;


represents the bus voltage; and
stands for the total
number of buses in a power system. In the above equation,
includes . The complex power at bus is
(21)
Substituting from (20) for in (21), the reactive power at generator bus can be expressed as

(22)
is the imaginary part of the diagonal
where
elements of the bus admittance matrix. For the sake of illustration, assume that
, (22) reduces to
(23)
where
is the reactance of the transmission line. In a typical power system,
, and the resistance may be neglected. In this case,
. Replacing
in (23) with
results in

(18)
represents the number of load buses in a power
where
system.
Like many of the existing measurement-based methods that
employ several consecutive PMU measurements for real-time
voltage stability margin estimation, the one proposed in this
paper uses three consecutive PMU measurements (three sets of
calculated
and ) to estimate the maximum loading point.
With the estimated maximum loading point, the voltage stability margin can be determined. A widely accepted measure
of voltage stability monitoring is expressed by the load power
margin which is dened to be the distance between the current
operating point
and system maximum loading point
.
Thus, the system voltage stability margin (VSM) dened by the
percentage of the load power margin is
%

(19)

C. Consideration of Generator Reactive Power Limits


An index, which measures the extreme reactive power of
each generator for a given load increase direction, is introduced
rst. Afterward, a new algorithm based on PMU measurements
to cope with the cases with generator
limits violation is
presented.

which can be reordered as


(24)
Indeed, the above equation denes a surface in the
space. Cutting this surface with constant power
factor planes, the
curves, which are similar to the
curves (or
curves), except that normal operating points
now lie on the lower part of the curves, can be obtained [2].
Since the approximate quadratic property of the
curve of
the load bus [29], [30], the
curve can be represented by a
quadratic curve. According to this, the
curve can be
approximately modeled by
(25)
where
, and
are parameters to be determined. Using
3 sets of
and , the system of (26) is obtained and the 3
parameters can be computed by

(26)

This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.
SU AND LIU: ESTIMATING THE VOLTAGE STABILITY MARGIN USING PMU MEASUREMENTS

RESULTS

TABLE I
LIMITS VIOLATION
NUMBER)

OF THE IDENTIFICATION OF
(#INDICATES

AT

PV BUS

where the number between parenthesis in (26) denotes the time


index of three consecutive PMU measurements. It should be emphasized that the time span between two consecutive measurements is not the high sampling rate of PMU such as 1 cycle. Instead, the time span, which varies from several seconds to minutes according to the prevailing dynamics of load variation of a
specic power system [31], is acceptable to serve the purpose
of the proposed method.
With the values of
, and , an estimate of the extreme
value of
is derived from setting
, i.e.,
(27)
for each
, and
represents the number of
generator buses in a power system. Let
and
be the
lower and upper bounds of the th generator reactive power
output, respectively. For a given load increase direction, if there
is no generator
limits violation at PV bus , then
. On the other hand, when
or
, generator limits are encountered at PV bus
. Thus, the value of
can be utilized as an indicator for
identifying -limit violation. That is, after the execution of this
method for
, a list of limits violation
at PV buses is created.
In order to verify the accuracy and effectiveness of the proposed method for identifying generator limits violation, we
have studied a lot of test cases. These include different test systems, different load levels, different load patterns, and various
-limit violations. The correctness of the results was conrmed
by running the CPFLOW software package of PSAT [32]. The
selected simulation results among those test cases are presented
in Table I, in which the third column represents the results obtained by the CPFLOW method. An inspection of this table
clearly states that the results given by the proposed method coincide with the results obtained by the CPFLOW method.
In this work, we assume that PMUs are placed optimally
based on one of the methods proposed in [13]; therefore, the
entire system becomes observable. In this case, time-synchronized voltage and current phasors at each bus are available. In
other words, the values of
, and
can be estimated
from PMU measurements.
Fig. 4 shows the owchart of the proposed algorithm with
consideration of generator reactive power limits for voltage stability margin estimation. The main steps are as follows.

Fig. 4. Flowchart of the proposed method for voltage stability margin


estimation.

1) Given the network admittance matrix


and three sets of
, and
calculated from PMU measurements;
2) Compute the parameters of
, and
for each
by (26);
3) Compute
by (27);
4) If there is no generator reactive power limits violation, then
go to step 6. Otherwise, go to step 5;
5) Change the bus type from PV bus to PQ bus for the generator reaching its limits;
6) Compute impedance matrix by (7);
7) Compute
by (12);
8) Estimate
by using the cubic
spline extrapolation technique according to Section III.B;
9) Find
;
10) Determine VSM by (19).
To evaluate the performance of the proposed method, the
experiments have been conducted with respect to various IEEE
test systems, load conditions, and -limit violations. Among
those simulation studies, Fig. 5 shows some selected case
results for the most critical bus 30 on the IEEE 30-bus system.
Fig. 5(a) depicts the case without
limits violation, while
Fig. 5(b) to (d) depict the cases with limits violation, in which
the sharp jumps in
coincide with the generator reaching

This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.
6

IEEE TRANSACTIONS ON POWER SYSTEMS

Fig. 6. Estimation of
.
actual

Fig. 5. Estimation of
estimated
; actual
; actual
.
tual

at bus 30 for the IEEE 30-bus system. (a) Case 1:


; actual
. (b) Case 2: estimated
. (c) Case 3: estimated
. (d) Case 4: estimated
; ac-

for the IEEE 300-bus system:

, the

TABLE III
SELECTED CASES OF SIMULATION SCENARIO I AND SIMULATION SCENARIO II

TABLE II
SIMULATION SCENARIOS FOR THE TEST SYSTEM (PR INDICATES
PERCENTAGE RATE)

its capacity limits. From Fig. 5, it reveals that the estimated


maximum loading points are very close to the actual ones.
IV. SIMULATION RESULTS
The effectiveness of the proposed method is examined on
three commonly used power networks, namely IEEE 30-, 118-,
and 300-bus power systems. The system data including line parameters and bus data are given in [33]. The scenarios, which
are conducted in the simulations, are summarized in Table II.
The experiments are implemented in MATLAB and run on a
Intel Core 2 Duo 2.66 GHz computer with 4 GB of RAM.
The performance measure utilized in this research is dened
as
%

(28)

, is very close to the actual


.
In this study, the authors have performed many simulation
examples with respect to different load levels, different load
patterns, and different -limit violations. Among those investigated cases, several test results for the scenarios given
in Table III are briey summarized. Moreover, the simulation
results obtained are all compared with those found by the
method [18] and method [25]. More details are presented and
discussed in the following examples.
A. IEEE 30-Bus System

is the VSM determined by the proposed method and


where
is the actual VSM obtained by the use of the CPFLOW software package of PSAT [32]. In addition, for each test system in
this study, it is simply assumed that the necessary PMU measurements are available at all generator and load buses, i.e., the
points of
, and
can be calculated from PMU measured data.
A lot of test cases on several IEEE test systems have been
studied. Fig. 6 shows a typical case of the proposed method
for the estimation of
. From the shown gure, it is clearly
seen that
, which is the smallest one among

The rst test system is concerned with the IEEE 30-bus


system which is composed of 41 branches, 6 generators, and
24 loads [33].
In the rst phase of the simulation, loads are increased with
the same percentage rate, i.e., all the loads in the IEEE 30-bus
system are increased simultaneously based on their initial load
levels. In order to examine the effect of generator reactive power
limits on voltage stability monitoring, a lot of -limit violation
cases have been studied. Table IV shows the estimations of the
voltage stability margin VSM via the proposed method, method
[18], and method [25] for the selected cases. In Table IV, it

This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.
SU AND LIU: ESTIMATING THE VOLTAGE STABILITY MARGIN USING PMU MEASUREMENTS

TABLE IV
PERFORMANCE COMPARISON BETWEEN THE PROPOSED METHOD, METHOD
[18], AND METHOD [25] UNDER THE SELECTED CASES OF SIMULATION
SCENARIO I FOR THE IEEE 30-BUS SYSTEM

TABLE VI
PERFORMANCE COMPARISON BETWEEN THE PROPOSED METHOD, METHOD
[18], AND METHOD [25] UNDER THE SELECTED CASES OF SIMULATION
SCENARIO I FOR THE IEEE 118-BUS SYSTEM

TABLE V
PERFORMANCE COMPARISON BETWEEN THE PROPOSED METHOD, METHOD
[18], AND METHOD [25] UNDER THE SELECTED CASES OF SIMULATION
SCENARIO II FOR THE IEEE 30-BUS SYSTEM

TABLE VII
PERFORMANCE COMPARISON BETWEEN THE PROPOSED METHOD, METHOD
[18], AND METHOD [25] UNDER THE SELECTED CASES OF SIMULATION
SCENARIO II FOR THE IEEE 118-BUS SYSTEM

apparently indicates that both the method [18] and method [25]
have approximately the same results under the scenarios with
proportional load increase. Also, due to the impact of limits
violation, the values of VSM estimated by the method [18] and
method [25] are much higher. On the contrary, the points of
VSM given by the proposed method (with considering generator
limits) are very close to the actual ones.
In the second phase of the simulation, loads are increased
with different percentage rate at different load buses. Meanwhile, generator limits are considered during the simulations.
The estimations of VSM under the selected cases by the three
compared methods are shown in Table V. Since loads are not
increased proportionally in these cases, the mismatch of the result provided by the method [25] is much greater than that of
the result calculated by the method [18]. Comparing the estimated VSM for each case in Table V, it is clearly seen that the
proposed method is able to provide the results with acceptable
accuracy.

TABLE VIII
PERFORMANCE COMPARISON BETWEEN THE PROPOSED METHOD, METHOD
[18], AND METHOD [25] UNDER THE SELECTED CASES OF SIMULATION
SCENARIO I FOR THE IEEE 300-BUS SYSTEM

B. IEEE 118-Bus System


In order to illustrate the effectiveness of the proposed method,
the IEEE 118-bus system is used as an example. This sample
system consists of 186 transmission lines, 54 generators, and
64 loads [33].
A series of simulations with respect to different system conditions, loading conditions, and -limit violations on the IEEE
118-bus system has been carried out. The selected simulation
results among those cases are presented as follows.
Table VI shows the comparison with VSM for the selected
cases under the simulation scenario I. From Table VI, it is
clearly seen that the errors of VSM obtained by the proposed
method are within a relatively small range. However, the errors
of VSM computed by the method [18] and method [25] are
much greater. Another selected test results for the simulation
scenario II are shown in Table VII. The results evidently show

that the proposed method performs better than that of the


method [18] and method [25].
C. IEEE 300-Bus System
The IEEE 300-bus system is also used as a test model to
demonstrate the applicability of the proposed method to buck
power systems. This simulated system is composed of 411
branches, 69 generators, and 231 loads [33].
In order to evaluate the performance of the presented methodology to power system voltage stability monitoring, we have
studied a lot of test cases. These include different load levels
and load patterns. Furthermore, generator reactive power limits
are considered in each simulation run.
The comparisons with the selected cases for the simulation
scenario I and simulation scenario II are listed in Tables VIII and
IX, respectively. From these tables, it is once again proven that
the proposed method is capable of offering acceptable accuracy
of results, compared with the results obtained by the method
[18] and method [25].
D. Statistical Evaluation
A series of simulations with respect to different test systems, different load levels, different load patterns, and various

This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.
8

IEEE TRANSACTIONS ON POWER SYSTEMS

TABLE IX
PERFORMANCE COMPARISON BETWEEN THE PROPOSED METHOD, METHOD
[18], AND METHOD [25] UNDER THE SELECTED CASES OF SIMULATION
SCENARIO II FOR THE IEEE 300-BUS SYSTEM

TABLE X
PERFORMANCE EVALUATION OF THE PROPOSED METHOD, METHOD [18], AND
METHOD [25] UNDER 500 TEST CASES FOR THE SIMULATION SCENARIO I

TABLE XII
EFFECTS OF MEASUREMENT ERRORS TO VSM ESTIMATION UNDER 500 TEST
CASES FOR THE SIMULATION SCENARIO I

TABLE XIII
RESULTS OF THE POST-CONTINGENCY CASE FOR THE IEEE 30-BUS TEST
SYSTEM

TABLE XI
PERFORMANCE EVALUATION OF THE PROPOSED METHOD, METHOD [18], AND
METHOD [25] UNDER 500 TEST CASES FOR THE SIMULATION SCENARIO II

generator -limit violations has been carried out. The statistical


results are summarized as below.
The results under 500 simulation runs for each simulation
scenario dened in Table II are listed in Tables X and XI, respectively. From the shown tables, it is clearly observed that
the errors provided by the proposed method are much smaller
than the results obtained by the method [18] and method [25].
In other words, the proposed method is superior over the method
[18] and method [25].
In addition, based on the observations from extensive simulations on the IEEE power systems that range from small systems
to bulk systems, one can see that the values of
estimated
by the proposed method are almost precise but a little less than
the actual values. In other words, the VSM determined by the
proposed method is relatively accurate and conservative. From
system operators' point of view, however, a little conservative
information is better than that of the optimistic one, particularly
in increasing voltage stability monitoring situational awareness.
E. Effects of Measurement Errors
Measured data from PMUs may introduce errors. To investigate the impact of measurement errors on voltage stability
margin estimation, the authors have performed many simulation
examples by simply adding a random noise to original PMU
measurements. Furthermore, an important criteria, namely the
total vector error provided by PMUs, should be less than 1%
based on the IEEE standard [34].
Table XII summarizes the simulation results under 500 test
cases for the simulation scenario I. Extensive simulation studies

verify that measurement inaccuracies indeed degrade the performance of the voltage stability monitoring method utilizing PMU
raw data. The problem of data pre-processing can be solved by
use of some ltering techniques [26], [27].
F. Effects of Post-Contingency Cases
The topology changes, which are simulated in this research,
are the post-contingency cases. The considered N-1 contingency
situations are outages of branches, where all possible branch
outages are simulated except the only branches connected to a
generator bus.
Due to limited space, the selected simulation results for the
IEEE 30-bus system are presented in Table XIII. An inspection
of this table clearly observes that the proposed method still provides acceptable results even under post-contingency cases.
V. CONCLUSION
A method, which employs PMU measurements to estimate
the voltage stability margin in real-time, is proposed. This
method takes into account the generator limits for practical
operation. It also aims to advance wide-area situational awareness enhanced with voltage stability monitoring. That is, with
this new method, power system operators are able to rapidly
recognize how far the system is away from a possible instability event. Numerical examples for various IEEE benchmark
systems have been carried out. The test results are promising
and the effectiveness of the proposed method is conrmed.
REFERENCES
[1] P. Kundur, Power System Stability and Control.
USA: McGraw-Hill, 1994.

New York, NY,

This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.
SU AND LIU: ESTIMATING THE VOLTAGE STABILITY MARGIN USING PMU MEASUREMENTS

[2] T. Van Cutsem and C. Vournas, Voltage Stability of Electric Power


Systems. Norwell, MA, USA: Kluwer, 1998.
[3] P. Kundur et al., Denition and classication of power system stability IEEE/CIGRE joint task force on stability terms and denitions,
IEEE Trans. Power Syst., vol. 19, no. 3, pp. 13871401, Aug. 2004.
[4] M. Begovic et al., Summary of system protection and voltage stability, IEEE Trans. Power Del., vol. 10, no. 2, pp. 631638, Apr. 1995.
[5] G. K. Morison, B. Gao, and P. Kundur, Voltage stability analysis using
static and dynamic approaches, IEEE Trans. Power Syst., vol. 8, no.
3, pp. 11591171, Aug. 1993.
[6] R. A. Schlueter, A voltage stability security assessment method,
IEEE Trans. Power Syst., vol. 13, no. 4, pp. 14231438, Nov. 1998.
[7] N. Flatabo, R. Ognedal, and T. Carlsen, Voltage stability condition in
a power transmission system calculated by sensitivity methods, IEEE
Trans. Power Syst., vol. 5, no. 4, pp. 12861293, Nov. 1990.
[8] N. Flatabo, O. Fosso, R. Ognedal, and T. Carlsen, A method for calculation of margins to voltage instability applied on the Norwegian
system for maintaining required security level, IEEE Trans. Power
Syst., vol. 8, no. 3, pp. 920928, Aug. 1993.
[9] A. Tiranuchit and R. J. Thomas, A posturing strategy against voltage
instabilities in electric power systems, IEEE Trans. Power Syst., vol.
3, no. 1, pp. 8793, Feb. 1988.
[10] B. Gao, G. K. Morison, and P. Kundur, Voltage stability evaluation
using modal analysis, IEEE Trans. Power Syst., vol. 7, no. 4, pp.
15291542, Nov. 1992.
[11] V. Ajjarapu and C. Christy, The continuation power ow: A tool for
steady state voltage stability analysis, IEEE Trans. Power Syst., vol.
7, no. 1, pp. 416423, Feb. 1992.
[12] H. D. Chiang, A. J. Flueck, K. S. Shah, and N. Balu, CPFLOW: A
practical tool for tracing power system steady-state stationary behavior
due to load and generation variations, IEEE Trans. Power Syst., vol.
10, no. 2, pp. 623634, May 1995.
[13] N. M. Manousakis, G. N. Korres, and P. S. Georgilakis, Taxonomy
of PMU placement methodologies, IEEE Trans. Power Syst., vol. 27,
no. 2, pp. 10701077, May 2012.
[14] J. De La Ree, V. Centeno, J. S. Thorp, and A. G. Phadke, Synchronized phasor measurement applications in power systems, IEEE
Trans. Smart Grid, vol. 1, no. 1, pp. 2027, Jun. 2010.
[15] C. Zheng, V. Malbasa, and M. Kezunovic, Regression tree for stability
margin prediction using synchrophasor measurements, IEEE Trans.
Power Syst., vol. 28, no. 2, pp. 19781987, May 2013.
[16] Y. Fan, S. Liu, L. Qin, H. Li, and H. Qiu, A novel online estimation
scheme for static voltage stability margin based on relationships exploration in a large data set, IEEE Trans. Power Syst., vol. 30, no. 3, pp.
13801393, May 2015.
[17] I. R. Pordanjani, Y. Wang, and W. Xu, Identication of critical components for voltage stability assessment using channel components
transform, IEEE Trans. Smart Grid, vol. 4, no. 2, Jun. 2013, Art. ID
11221132.
[18] J. H. Liu and C. C. Chu, Wide-area measurement-based voltage stability indicators by modied coupled single-port models, IEEE Trans.
Power Syst., vol. 29, no. 2, pp. 756764, Mar. 2014.
[19] K. Vu, M. M. Begovic, D. Novosel, and M. M. Saha, Use of local measurements to estimate voltage-stability margin, IEEE Trans. Power
Syst., vol. 14, no. 3, pp. 10291035, Aug. 1999.
[20] B. Milosevic and M. Begovic, Voltage-stability protection and control using a wide-area network of phasor measurements, IEEE Trans.
Power Syst., vol. 18, no. 1, pp. 121127, Feb. 2003.
[21] I. Smon, G. Verbic, and F. Gubina, Local voltage-stability index
using Tllegen's theorem, IEEE Trans. Power Syst., vol. 21, no. 3, pp.
12671275, Aug. 2006.
[22] G. Verbic and F. Gubina, A new concept of voltage-collapse protection based on local phasors, IEEE Trans. Power Del., vol. 19, no. 2,
pp. 576581, Apr. 2004.

[23] S. Corsi and G. N. Taranto, A real-time voltage instability identication algorithm based on local phasor measurements, IEEE Trans.
Power Syst., vol. 23, no. 3, pp. 12711279, Aug. 2008.
[24] M. Parniani and M. Vanouni, A fast local index for online estimation
of closeness to loadability limit, IEEE Trans. Power Syst., vol. 25, no.
1, pp. 584585, Feb. 2010.
[25] Y. Wang, I. R. Pordanjani, W. Li, W. Xu, T. Chen, E. Vaahedi, and
J. Gurney, Voltage stability monitoring based on the concept of coupled single-port circuit, IEEE Trans. Power Syst., vol. 26, no. 4, pp.
21542163, Nov. 2011.
[26] J. E. Tate and T. Van Cutsem, Extracting steady state values from
phasor measurement unit data using FIR and median lters, in Proc.
IEEE Power Syst. Conf. Expo., 2009, pp. 18.
[27] M. Glavic and T. Van Cutsem, Wide-area detection of voltage instability from synchronized phasor measurements, Part I: Principle,
IEEE Trans. Power Syst., vol. 24, no. 2, pp. 14081416, May 2009.
[28] C. de Boor, A Practical Guide to Splines. New York, NY, USA:
Springer-Verlag, 1978.
[29] G. C. Ejebe, G. D. Irisarri, S. Mokhtari, O. Obadina, P. Ristanovic, and
J. Tong, Methods for contingency screening and ranking for voltage
stability analysis of power systems, IEEE Trans. Power Syst., vol. 11,
pp. 350356, Feb. 1996.
[30] H.-D. Chiang, C.-S. Wang, and A. Flueck, Look-ahead voltage and
load margin contingency selection functions for large-scale power systems, IEEE Trans. Power Syst., vol. 12, no. 1, pp. 173180, Feb. 1997.
[31] P. Kansal and A. Bose, Bandwidth and latency requirements for smart
transmission grid applications, IEEE Trans. Smart Grid, vol. 3, no. 3,
pp. 13441352, Sep. 2012.
[32] F. Milano, Power System Analysis Toolbox (PSAT). ver. 2.0.0, Feb.
2008.
[33] University of Washington College of Engineering, Power systems test
case archive, [Online]. Available: http://www.ee.washington.edu/reserach/pstcal/
[34] IEEE Standard for Synchrophasors for Power Systems, IEEE Std. C37.
118-2005, 2005.

Heng-Yi Su (S'12M'15) was born in Taipei, Taiwan,


ROC, in 1980. He received the M.S. and Ph.D. degrees in electrical engineering from National Taiwan
University, Taipei, Taiwan, in 2005 and 2014, respectively.
Currently, he is an Assistant Professor of electrical
engineering with Feng Chia University, Taichung,
Taiwan. His research interests include applications of
PMUs to power system voltage stability monitoring
and control.

Chih-Wen Liu (S'93M'96SM'02F'13) received


the B.S. degree from National Taiwan University,
Taipei, Taiwan, in 1987, and the M.S. and Ph.D.
degrees from Cornell University, Ithaca, NY, USA,
in 1992 and 1994, respectively, all in electrical
engineering.
Since 1994, he has been with the Department of
Electrical Engineering, National Taiwan University,
Taipei, Taiwan, where he is currently a University
Distinguished Professor. His main research interests
include applications of information/communication
technology (ICT) to power system monitoring, protection, and control.

You might also like