1368
3, JULY 2012
A Novel Approach to Detect Symmetrical
Faults Occurring During Power Swings by
Using Frequency Components of Instantaneous
Three-Phase Active Power
Behnam Mahamedi and Jian Guo Zhu, Senior Member, IEEE
Abstract—Since distance relays are prone to interpret a power
swing as a three-phase fault, they should be blocked during the
power swing to prevent undesired trips. On the other hand, if any
fault occurs during a power swing, they should be fast and reliably
unblocked. Although unblocking the relay is straightforward in the
case of asymmetrical faults by using the zero-sequence and/or neg-
ative-sequence component of current, detecting symmetrical faults
during a power swing is still a challenge. This paper presents a
novel method for detecting symmetrical faults occurring during a
power swing. Based on the damping frequency component of 50 (or
60) Hz created on instantaneous three-phase active power profile
after inception of a symmetrical fault, the proposed method will
be able to detect the fault in less than one power cycle. This de-
tection method can be readily implemented, and is immune to the
power swing slip frequency, fault inception time, and fault loca-
tion. To test the proposed method, several power swings and faults
are numerically simulated in MATLAB/SIMULNK, and the simu-
lation results show that the proposed method is sensitive as well as
reliable.
Index Terms—Distance relay, instantaneous three-phase active
power, power swings, symmetrical faults.
I. INTRODUCTION
D ISTANCE relays operate on the basis of impedance they
measure. Once a fault occurs, the impedance will enter
into the relay protective zones and consequently the relay will
trip after a predefined time delay. On the other hand, during
power swings, the impedance trajectory can also enter into the
protective zones and eventually cause relay to issue an undesired
trip [1]. Power systems may experience a power swing when
they are recovered from a disturbance like a fault and/or heavy
load rejection. To be precise, when the frequency of a gener-
ator (or a group of generators) deviates from the nominal power
system frequency (50 or 60 Hz), a power swing will occur. Since
there is no unbalance during a power swing, this phenomenon
Manuscript received June 27, 2011; revised April 07, 2012; accepted May 12,
2012. Date of publication June 12, 2012; date of current version June 20, 2012.
Paper no. TPWRD-00552-2011.
B. Mahamedi is with the School of Electrical, Mechanical and Mechatronic
Systems, University of Technology Sydney, Ultimo, NSW 2007, Australia, and
also with the Iran Grid Management Company, Tehran IR-15175–648, Iran
(e-mail: behnam.mahamedi@student.uts.edu.au).
J. G. Zhu is with the School of Electrical, Mechanical and Mechatronic Sys-
tems, University of Technology Sydney, Ultimo NSW 2007, Australia (e-mail:
jianguo.zhu@uts.edu.au).
Digital Object Identifier 10.1109/TPWRD.2012.2200265
0885-8977/$31.00 © 2012 IEEE
IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 27, NO.
appears like a three-phase fault. It is crucial then for distance
relays to distinguish a power swing from a three-phase fault.
To overcome the adverse effect of power swings, several power
swing detection methods that block the protective zones during
power swings have been proposed. However, if an asymmetrical
or a symmetrical fault occurs during a power swing, the protec-
tive zones should be released and allowed to trip in a proper
manner.
The most popular method for detecting a power swing is to
measure the rate of change of impedance as it travels into the
protective zones of the relay [2]. Once a distance relay is blocked
by using this method, zero-sequence and/or negative-sequence
components of current are used to unblock the relay when a fault
occurs as it is the most common fault detection method in nu-
merical distance relays [3], [4]. It is clearly known that this fault
detection method fails to unblock a distance relay when a sym-
metrical fault occurs since there is neither a zero-sequence nor
negative-sequence component. This is the main disadvantage of
this fault detection method.
Su et al. [5] utilized the features of swing center voltage
(SCV) to detect symmetrical faults occurring during power
swings. This method is based on the SCV profile during a
power swing and after the fault inception. Li et al. [6] presented
a method based on the measured three-phase active and reac-
tive powers for detecting symmetrical faults. Mechraoui and
Thomas [7], [8] introduced a method by monitoring the voltage
phase angle at the relay location to discriminate faults from
power swings. Only single-phase faults have been simulated,
and the method has not been applied to symmetrical faults.
Lotfifard et al. [9] presented a method using the dc component
of three-phase currents, extracted by the Prony method.
Wavelet-based signal-processing techniques are effective
tools for power system transient analysis and feature extraction
[10]–[12]. Several fault detection methods based on the wavelet
transform (WT) are proposed [13], [14]. Brahma [15] utilized
the WT for detecting faults during power swings, taking ad-
vantage of the presence of high-frequency components created
on the voltage at inception of fault. Pang and Kezunovic [16]
presented another wavelet-based symmetrical fault detection
method which needs the modal transform in addition to WT.
This paper presents a fast symmetrical fault detector for dis-
tance relays, based on the detection of frequency components
of an instantaneous three-phase active power profile. It will
be demonstrated that during a power swing, the instantaneous
MAHAMEDI AND ZHU: NOVEL APPROACH TO DETECT SYMMETRICAL FAULTS
Fig. 1. Double-frequency power system.
three-phase active power has a frequency component of slip
frequency whereas after the fault inception time, a damping
power-frequency component (50 or 60 Hz) will be created on
its profile. Based on these features and by using the fast Fourier
transform (FFT) analysis, symmetrical faults can be reliably and
quickly detected within one power cycle (i.e., less than 20 ms
in a 50-Hz system and less than 16.67 ms in a 60-Hz one). This
method does not need high sampling frequency and can work
with the common sampling frequency used in digital distance
relays [17]. This detector can be readily implemented and is
immune to the slip frequency, fault location, and fault inception
time. To test the proposed method, several power swings and
faults were numerically simulated in MATLAB/SIMULINK,
and the results prove that this method is sensitive as well as
reliable.
II. THREE-PHASE ACTIVE POWER DURING A POWER SWING
AND AFTER FAULT INCEPTION
Since the proposed method is based on the instantaneous
three-phase active power profile during a power swing and
after symmetrical fault inception, this section is dedicated
to derive and analyze the proper expressions of three-phase
active power. To reduce the length, the word “instantaneous”
in “instantaneous three-phase active power” will be dropped
henceforth.
A. Three-Phase Active Power During a Power Swing
Power swings occur in a power system when two or more
different frequencies exist which are close to the nominal power
system frequency (50 or 60 Hz). The equation illustrating the
three-phase active power variations during a power swing differs
from that held in the normal conditions when there is only one
power system frequency.
To calculate the three-phase active power measured by a dis-
tance relay during a power swing, we consider a power system
with two sources of different frequencies close to the nominal
power system frequency and connected through a transmission
line as shown in Fig. 1.
By the principle of superposition, during a power swing, the
current and voltage of Phase A measured by the relays located
at both ends of the transmission line can be written as [18]
(1)
(2)
where and are the angular frequencies;
and are the current magnitudes at the relay location;
and are the current phase angles; and are the voltage
magnitudes at the relay location; and and are the voltage
1369
phase angles, corresponding to sources 1 and 2, respectively.
All phase angles are presented by assuming an infinite busbar
voltage as a reference. The Phase B and C currents and voltages
can be obtained by introducing a phase-angle delay of 120 and
240 in (1) and (2), respectively.
The three-phase active power can be expressed in term of
current and voltage as [19] and [20]
(3)
Substituting the phase currents and voltages given by (1) and
(2) into (3) and doing some trigonometric derivations, one ob-
tains the following expression of the three-phase active power
during the power swing:
(4)
where , , and are constants.
As discussed previously, during a power swing, the differ-
ence between the two frequencies is very small (otherwise, the
system would suffer from the harmonic condition which is en-
tirely different from power swings), so that the following ap-
proximations are sufficiently accurate:
(5)
(6)
Therefore, the three-phase active power during a power swing
can be approximately expressed as
(7)
where .
To simulate a power swing, a single-machine-infinite busbar
(SMIB) is commonly used by industrial professionals [1]; book
authors [2]; and researchers [5]–[8] , [21], because the simu-
lated signals of this system after a disturbance/fault are consis-
tent with those measured during real power swings. In this status
of an SMIB, one source is at the generator frequency that devi-
ates from the fundamental frequency while the other one has a
frequency equal or very close to the fundamental frequency (50
or 60 Hz).
Fig. 2 shows the per-unit (p.u.) generator rotor speed during a
power swing in an SMIB, simulated by MATLAB/SIMULINK.
Looking at the speed profile and keeping in mind that this profile
is an output of a linear time-invariant system, lead us to the
following equation for the rotor speed:
(8)
where and are constants, is the damping time constant of
the speed profile, and is the slip frequency, which is nothing
other than the rotor speed variations frequency.
The generator frequency, or the frequency of source 1, can
then be obtained by multiplying in per unit and the nom-
inal or base speed as follows:
(9)
1370 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 27, NO. 3, JULY 2012

Fig. 2. Rotor speed during a power swing. Fig. 3. Lumped model for Phase A after a symmetrical fault inception.

As explained before, the angular frequency of the infinite where and are the positive-sequence resistance and induc-
busbar is equal to and, thus, - can be written as tance, respectively, from the relay location to the fault point,
is the magnitude of voltage at the relay location ( is
the power frequency, 50 or 60 Hz), and is the phase angle of
the voltage at the fault inception time.
(10)
Solving this equation, we obtain
Substituting (10) into (7), we obtain an expression relating the
(14)
three-phase active power to the slip frequency as follows:

where can be found by

(11)
(15)
After a trigonometric expansion, the three-phase active power
can eventually be expressed as and is a constant obtained from the initial condition by

(16)

(12)
The Phase B and C currents can be written accordingly as
where , , , and are constants.
Equation (12) shows that during a power swing, the (17)
three-phase active power has two frequency components
of and . Generally, the component is negligible
compared with the component. This is mainly because (18)
the heavier damping effect of than , and is
inherently smaller than . This is confirmed in the simulated
cases to be presented in the next section and, in fact, as reported where
in [6], the three-phase active power oscillates at a single fre-
quency. Therefore, we can conclude that during a power swing, (19)
the three-phase active power oscillates at the frequency of .
Typically, it is in the range of 0.3–7 Hz [7], [22]. (20)

B. Three-Phase Active Power After Symmetrical Fault (21)

Inception
In this section, we are going to find the variation of three- and
phase active power when a symmetrical fault occurs. First, we
will extract the phase currents when a symmetrical fault occurs. (22)
Since our detection method is not based on the transients caused
by the faults, we only need the currents calculated by using the
lumped model of the faulty system. Fig. 3 shows this model for Substituting (14)–(22) into (3), we can finally calculate the
Phase A when a symmetrical fault occurs. Based on the model, three-phase active power measured by the relay after fault in-
after symmetrical fault inception, the Phase A current is gov- ception by
erned by

(13) (23)
MAHAMEDI AND ZHU: NOVEL APPROACH TO DETECT SYMMETRICAL FAULTS 1371

where is a constant and

(24)

(25)

Equation (23) clearly shows that after a symmetrical fault in-

ception, the three-phase active power has damping oscillations
at the nominal frequency. It should be mentioned that since three
phases are symmetric, constants , , and cannot be
zero simultaneously as shown in (16), (21), and (22). There-
fore, it does not matter when the symmetrical fault occurs and
for any fault inception time, the three-phase active power will
have a damping frequency component of power frequency.
Here, it is useful to comment about the equivalent circuit
(Fig. 3) that we used to calculate the current after fault incep-
tion. In the circuit, it has been assumed that the fault resis-
tance is negligible or the fault is a solid one. If the fault re-
sistance cannot be omitted, the current will still have two dif-
ferent frequencies as the two sources cannot be separated by
the fault point. However, a damping dc term will also be cre-
ated in the currents at the fault inception time. In this case, the
three-phase active power will contain both low-frequency (the
slip frequency) and damping power-frequency (50 or 60 Hz)
oscillations. Therefore, the results that have recently been ob-
tained are still valid for normal impedance faults. Here, we use Fig. 4. Flowchart of the proposed method.
the word “normal-impedance” to indicate that we do not con-
sider high-impedance faults (HIFs) since they mostly occur in
phase-to-earth faults rather than three-phase ones [23], [24].

III. PROPOSED SYMMETRICAL FAULT DETECTION METHOD

As just explained, during a power swing, the three-phase ac-
tive power oscillates at the slip frequency, whereas after the
fault inception, it oscillates at the nominal frequency (50 or 60
Hz). On the other hand, since the slip frequency (0.3–7 Hz) is
much lower than the power frequency (50 or 60 Hz), if the relay
performs a fast Fourier transform (FFT) on three-phase active
power, the fundamental (power) frequency coefficient will be
negligible during a power swing, whereas it suddenly increases
to a considerable value after the fault inception. Eventually, we
can propose a method for the relay to detect the symmetrical
faults by judging whether the FFT coefficient of the fundamental
component is more than a predefined threshold. In the next sec-
tion, we will calculate the threshold by numerical simulation.
The proposed method requires that the FFT be carried out
every 1 ms (i.e., 20 times per cycle for a 50-Hz power system),
which should be a tiny extra computing burden. The fault can
then be detected in less than one power cycle. In the following
section, the exact amount of time required for fault detection by
this method will be obtained.
Fig. 4 illustrates a flowchart of the proposed method in a log-
ical pattern. The instantaneous three-phase active power ( ) viewpoint of fault location and fault inception time according
is calculated from the sampled currents and voltages. Then, the
FFT is performed to find the fundamental frequency component
from the samples of a previous power cycle and this value is up-
dated at every sampling. For a 50-Hz power system, the data
window is 20 ms or 20 samples. Once the FFT coefficient is
Fig. 5. Simulated 400-kV power system.
more than the predetermined threshold (k), a three-phase fault
is detected, and the power-swing blocking (PSB) flag will be
reset.
IV. SIMULATION AND EVALUATION
To demonstrate the proposed symmetrical fault detection
method, a series of tests by using MATLAB/SIMULIMK to a
400-kV transmission system shown in Fig. 5 was conducted.
The length of the transmission line is 150 km. The system
details are given in the Appendix. To simulate the power swings
with different slip frequencies, the frequency of source 1 was
modulated by a frequency in the range of 1–5 Hz so that the
profile of generator frequency discussed in Section II can be
available. The slip frequencies simulated in this study are 1, 2,
3, 4, and 5 Hz. The frequency of source 2 is held to 50 Hz as it
is assumed to be an infinite busbar.
During the power swings, different types of faults, from the
to cases shown in Table I, were simulated. Since a fault may
occur at any instantaneous three-phase active power level, in
the simulation, different fault inception times at the maximum,
intermediate, and minimum levels of instantaneous three-phase
active power were considered to investigate the effect of fault
1372
TABLE I
SIMULATED CASES FOR EVALUATING THE PROPOSED METHOD
Fig. 6. Three-phase active power and its FFT coefficient measured in Case 1.
Fig. 7. Three-phase active power and its FFT coefficient measured in Case 2.
inception time on the proposed method. The sampling frequency
was chosen as 1 kHz, which is common in digital relays. The
variation of three-phase active power measured by relay R1 was
studied by examining its FFT coefficients of the fundamental
IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 27, NO. 3, JULY 2012
Fig. 8. Three-phase active power and its FFT coefficient measured in Case 3.
Fig. 9. Three-phase active power and its FFT coefficient measured in Case 4.
Fig. 10. Three-phase active power and its FFT coefficient measured in Case 5.
frequency (50 Hz) component during the power swings and after
the symmetrical fault inception. For the sake of simplicity, the
three-phase active power will be noted as henceforth, and to
avoid large values, the FFT coefficients presented in per unit by
assuming 100 MW as the base power.
Figs. 6–20 illustrate the simulated profiles of three-phase ac-
tive power and FFT coefficients of fundamental components for
Cases 1 to 15, respectively. Table II tabulates the maximum
values of FFT coefficients and the inception times measured by
relay R1 in Cases 1 to 15.
To determine a threshold for detecting a symmetrical fault
by the FFT coefficient of the fundamental component, the
MAHAMEDI AND ZHU: NOVEL APPROACH TO DETECT SYMMETRICAL FAULTS
Fig. 11. Three-phase active power and its FFT coefficient measured in Case 6.
Fig. 12. Three-phase active power and its FFT coefficient measured in Case 7.
Fig. 13. Three-phase active power and its FFT coefficient measured in Case 8.
maximum FFT coefficients during the power swings with dif-
ferent slip frequencies and the minimum FFT coefficients after
faults in the 15 cases tabulated in Table I should be determined.
Table III lists the maximum FFT coefficients during the power
swings with different slip frequencies, and Table IV shows the
maximum value of the FFT coefficient after fault inception
with its corresponding measurement time and time delay for
fault detection.
From Table III, one can find that the maximum FFT coef-
ficient during the power swings before fault inception is 0.39,
whereas the minimum FFT coefficient after fault inception is
1.98. Therefore, a threshold equal to 1 would be appropriate.
1373
Fig. 14. Three-phase active power and its FFT coefficient measured in Case 9.
Fig. 15. Three-phase active power and its FFT coefficient measured in Case
10.
Fig. 16. Three-phase active power and its FFT coefficient measured in Case
11.
From Table IV, one can find that the maximum delay for fault
detection is 14 ms or 0.014 s, which indicates the fast detecting
speed of the proposed method. Regarding the threshold, it is
necessary to mention that for other power systems, it can also
be readily obtained by simulation studies.
Here, it is worthwhile to comment about the three-phase ac-
tive power profile after fault inception. As can be seen from the
waveforms of three-phase active power shown in Figs. 5–19,
apart from the frequency component of 50 Hz, other high-fre-
quency components can be created on the three-phase active
power profile after fault inception. However, as mentioned in
1374
Fig. 17. Three-phase active power and its FFT coefficient measured in Case
12.
Fig. 18. Three-phase active power and its FFT coefficient measured in Case
13.
Fig. 19. Three-phase active power and its FFT coefficient measured in Case
14.
Section II, they are not of concern in the proposed method,
which is the beauty of the proposed concept of three-phase ac-
tive power. Indeed, the proposed method will not be affected by
high-frequency transients created on instantaneous three-phase
active power.
V. CONCLUSION
A novel scheme for detecting symmetrical faults occurring
during power swings was presented based on the variation of the
FFT coefficient of fundamental component of the three-phase
IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 27, NO. 3, JULY 2012
Fig. 20. Three-phase active power and its FFT coefficient measured in Case
15.
TABLE II
MAXIMUM FFT COEFFICIENTS AND INCEPTION TIMES
MEASURED BY RELAY R1 IN CASES 1 TO 15
TABLE III
MAXIMUM FFT COEFFICIENTS DURING POWER
SWINGS OF DIFFERENT SLIP FREQUENCIES
active power profiles before and after the fault inception. By
comparing the FFT coefficient of three-phase active power fun-
damental component with a predetermined threshold, the new
method can detect a symmetric fault occurring during a power
swing within a power cycle. For a 50-Hz, 100-MW base power,
400-kV power system of 150-km transmission line, through a
systematic numerical study of 15 cases, it was found that 1.00
would be an appropriate threshold value. For other power sys-
tems, the threshold values can be determined similarly.
As outlined before, the threshold value was determined by
a systematic numerical study of a series of cases for a partic-
ular power system. It could be well argued that a more accurate
threshold value could be obtained by using practical data of a
MAHAMEDI AND ZHU: NOVEL APPROACH TO DETECT SYMMETRICAL FAULTS
TABLE IV
MAXIMUM FFT COEFFICIENTS AFTER SYMMETRICAL FAULT INCEPTION
WITH CORRESPONDING MEASUREMENT TIME AND TIME DELAY
real power system. Under this circumstance, the data-processing
procedure is exactly the same as the one presented with the sim-
ulation data. However, since the threshold value is system de-
pendent and should account for all possible cases of symmet-
rical faults, for a real power system, it is impractical to obtain
the practical data of all possible cases of faults, especially in the
stage of system design, when the threshold value must be deter-
mined theoretically. A more sensible approach would be to code
the theoretical threshold value in the fault detecting algorithm in
the digital relay as an initial value and refine it by applying an
algorithm of artificial intelligence (e.g., neural network), with
the practical data once the relay is installed and operated in the
real system.
The proposed method has several advantages that are superior
to the existing methods. First, since it requires only three-phase
active power, which can be simply calculated by production of
instantaneous voltage and current, and FFT, which is coded in
all digital relays, the new method can be readily implemented.
Second, the new method is very fast. While using a sampling
frequency of only 1 kHz, it can detect the fault within a power
cycle. From the literature, such speed can only be possible with
a high sampling frequency. Finally, in the proposed method,
FFT is performed only on one signal (i.e., three-phase active
power), whereas other existing methods need to perform signal
processing on all three-phase voltages and/or currents. It is clear
that the proposed new method would cause much less extra com-
puting burden to the relays.
APPENDIX
The parameters of the system used for simulation (Fig. 4) are
given.
Source impedance
1375
Transmission line: Distributed model with parameters
REFERENCES
[1] Power Syst. Relay Committee, “Power swing and out-of-step consid-
erations on transmission lines,” IEEE PSRC Working Group D6, New
York, 2005.
[2] J. L. Blackburn, Protective Relaying Principles and Applications, 2nd
ed. New York: Marcel Dekker, 1998.
[3] “REL 670 Technical Reference Manual” ABB. [Online]. Available:
www.pes-psrc.org/h/H5/Manual_ABB.pdf
[4] “MiCOM P444 Application Notes,” AREVA. [Online]. Available: ftp://
ftp.areva-td.com/P44x_EN_T_F65_Global.pdf
[5] B. Su et al., “Fast detector of symmetrical fault during power swing for
distance relay,” in Proc. IEEE Power Eng. Soc. Gen. Meeting, 2005, pp.
1836–1841.
[6] X. Lin, Y. Gao, and P. Liu, “A novel scheme to identify symmetrical
faults occurring during power swings,” IEEE Trans. Power Del., vol.
23, no. 1, pp. 73–78, Jan. 2008.
[7] A. Mechraoui and D. W. P. Thomas, “A new blocking principle with
phase and earth fault detection during fast power swings for distance
protection,” IEEE Trans. Power Del., vol. 10, no. 3, pp. 1242–1248,
Jul. 1995.
[8] A. Mechraoui and D. W. P. Thomas, “A new principle for high resis-
tance earth fault detection during fast power swings for distance pro-
tection,” IEEE Trans. Power Del., vol. 12, no. 4, pp. 1452–1457, Oct.
1997.
[9] S. Lotfifard, J. Faiz, and M. Kezunovic, “Detection of symmetrical
faults by distance relays during power swings,” IEEE Trans. Power
Del., vol. 25, no. 1, pp. 81–87, Jan. 2010.
[10] S. Santoso, E. J. Powers, W. M. Grady, and P. Hofmann, “Power
quality assessment via wavelet transform analysis,” IEEE Trans.
Power Del., vol. 11, no. 2, pp. 924–930, Apr. 1996.
[11] D. C. Robertson, O. I. Camps, J. S. Mayer, and W. B. Gish, “Wavelets
and electromagnetic power system transient,” IEEE Trans. Power Del.,
vol. 11, no. 2, pp. 1050–1058, Apr. 1996.
[12] M. Michalik, W. Rebizant, M. Lukowicz, S.-J. Lee, and S.-H. Kang,
“High-impedance fault detection in distribution networks with use of
wavelet-based algorithm,” IEEE Trans. Power Del., vol. 21, no. 4, pp.
1793–1802, Oct. 2006.
[13] K. L. Butler-Purry and M. Bagriyanik, “Characterization of transients
in transformers using discrete wavelet transforms,” IEEE Trans. Power
Syst., vol. 18, no. 2, pp. 648–656, May 2003.
[14] J. Faiz and S. Lotfi-Fard, “A novel wavelet-based algorithm for
discrimination of internal faults from magnetizing inrush currents
in power transformers,” IEEE Trans. Power Del., vol. 21, no. 4, pp.
1989–1996, Oct. 2006.
[15] S. M. Brahma, “Distance relay with out-of-step blocking function
using wavelet transform,” IEEE Trans. Power Del., vol. 22, no. 3, pp.
1360–1366, Jul. 2007.
[16] C. Pang and M. Kezunovic, “Fast distance relay scheme for detecting
symmetrical fault during power swing,” IEEE Trans. Power Del., vol.
25, no. 4, pp. 2205–2212, Oct. 2010.
[17] I. Zamora, A. J. Mazón, V. Valverde, E. Torres, and A. Dysko, “Power
quality and digital protection relays,” presented at the Int. Conf. Renew.
Energies Power Qual., Barcelona, Spain, 2004.
[18] C. Wei, P. Zhencun, Y. Zhihao, and Z. Rong, “The method of detecting
asynchronous power swing based on the variation of active power,” in
Proc. Int. Conf. Power Syst. Manage. Control, 2002, pp. 297–302.
[19] W. H. Hayt and J. E. Kemmerly, Engineering Circuit Analysis. New
York: McGraw-Hill, 1978.
[20] P. Kundur, Power System Stability and Control. New York: McGraw-
Hill, 1993.
[21] H. Khorashadi-Zadeh, “Evaluation and performance comparison of
power swing detection algorithms,” in Proc. IEEE Power Eng. Soc.
Gen. Meeting, 2005, pp. 976–982.
1376
[22] G. Benmouyal, D. Hou, and D. Tziouvaras, Zero-setting power-swing
blocking protection. [Online]. Available: http://www.selinc.com/tech-
pprs/6172_ZeroSetting_20050302.pdf
[23] E. O. Schweitzer, III, N. Fischer, and B. Kasztenny, “A fresh look at
limits to the sensitivity of line protection,” in Proc. 64th Annu. Conf.
Protective Relay Eng., College Station, TX, Apr. 2011.
[24] Power Syst. Relay Committee, “IEEE guide for protective relay appli-
cations to transmission lines,” New York, IEEE C37.113-1999, 1999.
Behnam Mahamedi was born in Tehran, Iran, in
1982. He received the B.Sc. degree in electrical
engineering from Iran University of Science and
Technology (IUST), Tehran, in 2005, the M.Sc. de-
gree in electrical engineering from Shahid Beheshti
University (SBU), Tehran, in 2010, and is currently
pursuing the Ph.D. degree in electrical engineering
from the University of Technology Sydney (UTS),
Sydney, Australia, focusing on power system fault
diagnosis systems.
Since 2008, he has been a Protection Engineer with
the Iran Grid Management Company, Tehran. His research interests are power
system protection, power system transient studies, and wavelet-transform appli-
cation in power systems.
IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 27, NO. 3, JULY 2012
Jian Guo Zhu (S’93–M’96–SM’03) received the
B.E. degree in electrical engineering from Jiangsu
Institute of Technology, Zhenjiang, China, in 1982,
the M.E. degree in electrical engineering from
Shanghai University of Technology, Shanghai,
China, in 1987, and the Ph.D. degree in electrical
engineering from the University of Technology
Sydney (UTS), Sydney, Australia, in 1995.
Currently, he is a Professor of Electrical Engi-
neering and the Head of the School of Electrical,
Mechanical, and Mechatronic Systems, UTS. His
research interests include electromagnetics, magnetic properties of materials,
electrical machines and drives, power electronics, and renewable energy
systems.