International Journal of Computer Applications (0975 – 8887)

International Conference on Advances in Emerging Technology (ICAET 2016)

Different Methods of Differentiating Inrush Current from

Internal Fault Current in Transformer
Anupam Sinha Sarpreet Kaur
Post Graduate Student Assistant Professor
University Institute of University Institute of
Engineering and Technology, Engineering and Technology
Panjab University, Panjab University,
Chandigarh, India Chandigarh, India

ABSTRACT researchers. Based on the discussion some conclusion has

When a transformer is energized the phenomenon of been drawn.
magnetizing inrush current occurs and causes a pseudo
tripping signal to the differential relay which leads to the
2. INRUSH CURRENT
problem of mal operation or false tripping of the relay. In When transformer is energized, the core flux and the
order to avoid this false tripping of the differential relay and corresponding exciting current undergo a transient before
for safe operation of the transformer the distinction of inrush reaching steady state values. The severity of the switching
current with internal fault current is very important. transient is related to the instant of switching. Under steady
Conventionally, second harmonic restraint relay is used but as state condition if the applied voltage is sinusoidal, the
size of power system network is increasing day by day the instantaneous value of common flux in the core (with no
electrical network is becoming more and more complicated residual flux) changes from –φmaximum to +φmaximum in half
and some disadvantages of conventional system are slowly cycle to balance the applied voltage and lag its voltage by 90
understood. Therefore, some other methods which can also be degree. If the transformer is switched on at the instant of its
used for proper distinction between inrush current and internal positive peak then the flux rises from zero and transformer is
fault current are highlighted in this paper. Different switched on with normal magnetizing current, the same would
techniques used for discriminating inrush current from happen if the applied voltage is at its negative peak at the
internal fault current are discussed and some conclusion has switching instant. However if at the instant of switching, the
been drawn. applied voltage is at zero and say going toward positive then
the flux must change from zero to 2φmaximum in half cycle for a
Keywords flux less core and if the flux contains residual flux then this
Second harmonic; Equivalent Instantaneous Inductance (EII); value will increase because of the effect of residual flux. This
Instantaneous Magnetizing Inductance (IMI); Morphological gives a rise to almost double the flux and is known as
Gradient Algorithm (MGA); Sinusoidal Proximity Factor doubling effect and further causes a huge magnetizing inrush
(SPF); Waveform Singularity Factor (WSF); Finite Element in the primary current. An analogous situation would arise
Method (FEM); Inrush current; Wavelet Transform when applied voltage is going toward negative. The value of
inrush current can reach five times the full load current in the
1. INTRODUCTION transformer and is therefore nearly 100 times the normal no
A Power transformer is an essential component in electrical load current.[15-16] Figure 1 displays the diagram of
power systems and the relays used for its protection must be generation of inrush current in the transformer.
reliable, dependable, and should take less operating time. Techniques are invented which can be used to study inrush
Differential protection is mostly used for the protection of current such as Finite Element Method[17-19], Coupled
transformer. However, false tripping of differential relay Electromagnetic model[20], Operational
occurs when a transformer is energized. This is because of Matrices[21].Investigation of several factor which affect
magnetizing inrush current phenomenon which occurs when inrush current of transformer is done using Finite Element
the transformer is energized. Therefore, distinction of inrush Method[22]. Inrush current also causes forces on the
current from internal fault is very important in order to transformer winding and sometimes these forces can be
improve the reliability and security of differential protection. greater than short circuit forces. It can be minimized by using
Some of the method used for recognition of inrush current superconductor, controlled switching, sequential phase
from internal fault current are voltage restraint [2],second energization technique, Virtual air gap technique, changing
harmonic restraint, dead-angle restraint, flux-based inrush distribution of the coil winding [23-30].
restraint [3],wavelet transform [4-5],inverse
inductance[6],transformer model-based modal analysis [7],
power differential [8], hyperbolic s transform[9], low
frequency component of Discrete wave transform[10],Scheme
of waveform symmetry. Techniques using Fuzzy logic and
artificial intelligence system have also been developed. [11-
14].
In this paper some of the method or ways for differentiating
inrush current from internal fault current has been discussed.
These discussions are a review of the work done by the

35
 International Journal of Computer Applications (0975 – 8887)
International Conference on Advances in Emerging Technology (ICAET 2016)

Fig 3: The Diagram of virtual third harmonic restraint

Fig 1: Inrush current generation in transformer scheme

3. METHODS During practical application of this method some problems

should be taken care to maintain the reliability of the system.
3.1 Wavelet based method
In this method first the transformer model with sufficient 1. The DC component should be removed in order to avoid
precision are used for inrush current and short circuit current the mal operation of the relay.
measurement. The simulation shows current waveform with 2. The symmetrical axis of the spiry pulse and data window
respect to time (in seconds) for both the cases. Then wavelet should beat same position.
transformation maps the time function of the result obtained
into a two dimensional function of α and t.α is called scaling 3. The threshold value for the operation of relay should be
factor and t is the change of the function along the time axis. carefully selected.
Daubechies family wavelets are selected for the investigation.
Results using wavelet transformation for inrush current and 3.3 Inductance based technique
short circuit current were quite different from each other and 3.3.1 EII based technique
thus this method provides a way to distinguish inrush current The change in Equivalent Instantaneous Inductance (EII) in
from short circuit current [43]. case of inrush current and internal fault is the criteria used for
distinguishing inrush current and internal fault current.
3.2 Virtual third harmonic restraint Experiments were conducted for verifying the performance of
In this method the technology of waveform construction is this technique and it was found that, there is extreme variation
used. Inrush current shown in fig.2 has a dominant even of the Equivalent Instantaneous Inductance (EII) for inrush
harmonic component in it. By using waveform construction current, but the value of EII for faulty phase is almost
technique the spiry pulse is moved for half cycle in backward constant.
direction and then reversed. This will form a waveform as
shown in fig. 3.In fig.3; the practical inrush current shown by During normal operating state of power transformer and for
thin line is the same current waveform which has been shown internal fault condition, the iron core is not saturated and the
in fig.2 while the thick line shows the result after applying the value of magnetizing current is extremely small, which results
process of wave construction. Now this waveform contains a in the approximately constant Instantaneous Magnetizing
dominant third harmonic component in it. However it to be Inductance (IMI) so the operation lies in the linear area of the
noted that the third harmonic component is not really very magnetizing characteristic. But in case of inrush current the
teeming in the whole process. It is only because of the core of transformer moves between saturation and non
waveform construction the third harmonic comes in picture. saturation region of operation due to this a sudden variation of
This method offers many advantages over conventional the IMI, as shown in Fig. 4
second harmonic relay. Firstly, the operating time has
reduced. Secondly, the restrain scheme can be performed by
phase. Thirdly, the problem of symmetrical inrush current can
also be solved [35].

Fig 4: Diagram showing the variation of IMI waveform

Fig 2: Transformer inrush current diagram

36
 International Journal of Computer Applications (0975 – 8887)
International Conference on Advances in Emerging Technology (ICAET 2016)

There are two methods which are used to compute the value
of EII. First method is Indirect method while second method
is Direct method.
Indirect method: When the supply of 50Hz is given to the
transformer the value of IMI varies and it has a fundamental
frequency of 50Hz. The value of fundamental frequency
component of variation of IMI in case of normal operation
and internal fault condition is zero and same is with the value
of EII while this is not the case for inrush current. Thus the
above mentioned criterion is used to differentiate inrush
current from internal fault or normal operation of transformer.
Fig 5: Data window of filter algorithm
If the magnitude of the fundamental frequency component in
the EII is more than the threshold or pick up value, then there When Fourier analysis for full window is done, the signal
is inrush current in the transformer and the relay tripping is magnitudes found for inrush current will not change or very
blocked, if it is more than the threshold value then relay little change will be there.
operating signal is send.
Direct method: The RMS expression which gives the variation
in EII is given as

Δ k= (1)

k= (2)

Where, N is number of samples in a single frequency cycle .

ΔL k̅ is criterion which sets the threshold and is used to Fig 6: Data window of filter algorithm
distinguish the inrush current from the internal fault . If ΔL k̅
exceeds a threshold or pick up value, the relay take it is as On the contrary, when the analysis for short window by
inrush current and the operation of relay is blocked. If this applying short window algorithm was done, the value of
value is less than the threshold value the relay takes it as inrush current calculated by short-window filter algorithm
internal fault. The second method of analysis is successful in will be different from full window analysis. The variation in
producing the result in time as well as in frequency domain magnetizing inductance is shown by solid lines in fig 6.
analysis, while first method is effective only in frequency However, this is not the case in the analysis of internal fault
domain [1]. current. The result is shown as the dotted line for full window
and solid line for short window filter algorithm (Fig.6 and
3.3.2 Instantaneous Inductance technique Fig.7)
In this method voltage and current signal are used to find the
differential inductance of the transformer from the primary
side. The differential inductance is calculated from every
phase of the transformer. An algorithm is developed which
compared this value of differential inductance with the
threshold value. If this value is more than the threshold value
then it is inrush current else it is internal fault. The operating
time of this method is also very less(5ms ,less than 1/4th of
the power frequency cycle).This method works even when
there is variation in the tapping of transformer, fault resistance Fig 7: Magnitude of internal fault current
comes into picture, saturation of Current Transformer(CT)
occurs [40]. Based on the above discussion inrush current detection
criterion is set. The detection criteria is given by
3.4 Short Window Filter Algorithm
For calculation of magnitude of transformer differential K= (3)
current the data window in one cycle Fourier filter is
considered. Short data window is used for the detection of Where, Imshortwindow shows the magnitude evaluated by short
inrush detection. Data window for filter algorithm is shown in window filter, ImFourier shows the magnitude evaluated by
fig. 5. W1 shows the data window of Fourier filter which is Fourier filter, N is sampling rate and Cth is the value of pickup
having span of one power frequency cycle. W2 shows the data threshold for detecting inrush current in differential relay [36].
window for short-window filter algorithm and its span is
shorter than span of one power frequency cycle. 3.5 Waveform Singularity Technique
In this technique, waveform singularity factor is used to
differentiate inrush current from internal fault current in
power transformer.
Algorithm:
For sinusoidal wave form,f(t) can be described as

f(t) Asin(ωtθ) (4)

37
 International Journal of Computer Applications (0975 – 8887)
International Conference on Advances in Emerging Technology (ICAET 2016)

Where, magnitude of sinusoidal waveform is denoted by A, sinusoidal proximity factor (SPF) is used to evaluate this
Angular frequency of power signal is denoted by ω, and initial difference.
phase angle is represented by θ.
η(tk) abs(ρ(tk) sin(2ωtk 2θ)) (14)
Time interval for quarter cycle Δt is 0.5
Where, η is the SPF and abs stands for absolute difference
f (t) for instant t + Δt is between ρ(tk) and sin(2ωtk+ 2θ) .When internal fault is there,
the value of SPF is close to zero and when magnetizing inrush
Acos(ωt+θ) (5) current is generated there is a drastic variation in the value of
SPF. Therefore, SPF is a measure to distinguish between
f(t) for instanttΔt/2is inrush current and internal fault current. If the value of SPF of
Acos(ωt+θ+π/4) (6) some phase is less than the threshold or pick up value of 0.5,
the relay take this as an internal fault and gets tripped and for
g(t) is defined as values greater than or equal to 0.5 the relay senses it as inrush
current and dismisses the tripping. This method is more useful
when the value of internal fault current is low [38].
g(t) f (t) f (t Δt)  f (t Δt2) (7)
3.7 Identifying the Difference between the
The value of g(t) is constant with time and is ideally equal to Waveform of Inrush Current and
zero, g(t).But in case of practical current waveform harmonics
comes into picture. So, a waveform singularity factor is used Internal Fault Current
to evaluate this difference The distinction between inrush current and internal fault
current can be done by identifying the difference between the
waveform of inrush current and internal fault current. The
h(t) = (8) different features of inrush current and internal fault current
are shown in fig 8.
e= (9)

Where, N is no. of samples per power frequency cycle. When

internal fault occurs the value of h(t) is zero and when
magnetizing inrush current comes into picture this value is
non zero. If the WSF of every phase gets more than the
threshold or pickup value of 1.0, the relay will take it as
inrush current and blocks the tripping. For all values less than
or equal to 1.0 the relay will take it as an internal fault and
will operate. The theoretical value of threshold is around zero. (a) (b)
Thus h(t), also known as Waveform Singularity Factor (WSF)
provides a measure to differentiate inrush current from Fig 8: (a) fault current, (b) inrush current.
internal fault current [37].
From the figure we can see that the internal fault current
3.6 Sinusoidal Proximity Factor waveform has high slope at the time of switching and after
The difference in the value of SPF in case of inrush current some time duration its slope gets reduced while in case of
and internal fault current is a measure to distinguish internal inrush current the slope at the time of switching is low and it
fault current from inrush current. The algorithm used for the increases till the time current reaches its maximum value. This
calculation of SPF is explained below distinction is used to discriminate the current as inrush current
or internal fault current. The methods are discussed as below.
The sinusoidal waveform can be expressed as
3.7.1 Improved Morphological Gradient
f(t) Asin(ωtθ) (10)
Algorithm
Where, magnitude of sinusoidal waveform is denoted by A, In order to identify the slope feature of the waveform an
Angular frequency of power signal is denoted by ω, and initial improved Morphological Gradient Algorithm (MGA) is used.
phase angle is represented by θ. Fundamental morphological operator dilation and erosion are
used in this scheme. Dilation and erosion for a one
Let a signal f (t) is defined which has n even numbers of dimensional signal is given as
sampling point. For sampling point at the instant tk
(f g) (x) = max {f(x-s)+g(s),x Df,s Dg} (15)
f(tk) Asin(ωtkθ) (11)
Time interval for quarter cycle Δt is 0.5 s
(f ⊖g) (x) = max {f(x+s)-g(s), x Df,s Dg} (16)
f(tk) for the instant tΔt is
Acos(ωtk+θ) (12) s
Normalizing f(t) and multiply equation (11) and (12) Where f is the field, g stands for structuring element (SE),
field of definition of f and g are denoted as Df and Dg
ρ(tk)=sin(2ωtk + 2θ) (13) respectively. For edge detection morphological gradient used
is:
This is the expression for pure sine wave. But for practical
current waveform harmonics comes into picture. So a G(f)=(f ⊕ g) – (f ⊖ g) (18)

38
 International Journal of Computer Applications (0975 – 8887)
International Conference on Advances in Emerging Technology (ICAET 2016)

Structuring element defined to find the rising and falling

edges are
g+ = {01,02,….0l} (19)
g- = {01,02…...0l} (20)
+ -
Where g is used for rising edge and g is used for falling
edge. For the length of flat SE of 3, the improved
morphological gradient for inrush current and internal fault
current is as shown in figure 9(a) and 9(b).

Fig 10: Flow chart of the scheme.

3.7.2 Using mathematical Morphology

The above discussed morphological approach fails when CT
Fig 9: (a) Morphological gradient for inrush current. is saturated so a new method is developed using
morphological fundamentals as described above. It uses two
different algorithms the first algorithm can discriminate
between inrush and internal fault currents when there is no CT
saturation. When CT gets saturated, the first algorithm fails to
provide a decision. In this case, the second algorithm which
uses a symmetrical morphological gradient criterion is
triggered automatically to differentiate between internal fault
and inrush current in the system. This method can also be
used to detect sympathetic inrush current [42].

3.7.3 Wavelet Based Technique by using

Finite Element Method
Discrete wavelet transform (DWT) is applied to the
Fig 9: (b) Morphological gradient for internal fault differential current and distinct characters and features for
current inrush current and internal fault current is extracted based on
the wavelet components. With the help of these character and
Based on these differences between the patterns of waveform, features the current will be recognized as inrush current or
the inrush current can be distinguished with internal fault internal fault current. The ability of the wavelet transform to
current in the sampling window of half a long cycle. For the focus on short time intervals for high frequency components
sake of quantification the criterion used for discrimination is and long-time interval for low frequency components
provides a better way of investigation for signals which have
(17) localized impulses and oscillation. So, wavelet decomposition
is very helpful for studying transient signals. It also helps in
Where, current signal in the data sampling window is finding a much better current characterization with more
represented by Is and morphological gradient is represented reliable distinction. The process of execution of DWT is
by Gs. The value of σ is very high in case of inrush current as shown in fig. 11 in which s denotes the original signal and
compared to its value in case of internal fault current. For the high pass filter and low pass filter are denoted as HPF and
next half cycle, if the gradient result shows a flat waveform, LPF respectively.
then it is inrush current otherwise, there is internal fault [41].
The above process is explained with the help of flow chart as
shown in figure 10.

Fig 11: Implementation procedure of DWT.

By observing the waveform for inrush current and internal
fault current (fig. 8) it can be said that in case of fault current
the magnitude of high frequencies at starting time has falling
trend, while in case of inrush current the trend has a rising.
These trends and features are found in the frequency level D3

39
 International Journal of Computer Applications (0975 – 8887)
International Conference on Advances in Emerging Technology (ICAET 2016)

when simulation based on Finite Element Method was done. [9] Q. Zhang, S. Jiao, S. Wang, ICIEA IEEE, Identification
Finite element method gives a more realistic method of Inrush Current and Internal Faults of Transformer based
simulation. Thus the above mentioned criteria can be used for on Hyperbolic S-transform,258-263,(2009).
differentiating inrush current from fault current. This method
is relatively faster method and can differentiate inrush current [10] C. Jettanasen, C. Pothisarn , J. Klomjit and
from fault current in less than quarter a cycle if the supply is A.Ngaopitakkul, ICEMS IEEE, Discriminating among
of 50Hz [39]. Inrush Current, External Fault and Internal Fault in
Power Transformer using Low Frequency Components
4. CONCLUSIONS AND FUTURE Comparison of DWT,1-6, (2012).
SCOPE [11] L. G. Perez, A. J. Flechsig, J. L. Meador, and Z.
Among the above mentioned techniques all the techniques can Obradoviic, IEEE Trans. Power Del., Training an
be used in real life situation but instantaneous inductance and artificial neural network to discriminate between
wavelet based technique are relatively faster as compared to magnetizing inrush and internal faults, 9, no. 1, 434–
other techniques. Also, Instantaneous inductance technique 441, (1994).
and morphological technique are better than other techniques
[12] Á. L. Orille-Fernández, N. K. I. Ghonaim, and J.A.
because they work even when CT is saturated. For future
Valencia, IEEE Trans. Power Del.,A FIRANN as a
scope, efforts are needed to remove the DC decaying
differential relay for three phase power transformer
component. Better mechanism for finding optimum position
protection, 16, no. 2, 215–218,(2001).
of data window needs to be developed. The sensitivity for the
method of sinusoidal proximity factor should not be limited to [13] M. R. Zaman and M. A. Rahman, IEEE Trans. Power
low level internal faults only. The application of EII can be Del., Experimental testing of the artificial neural network
extended to transformers with more number of turns. Efforts based protection of power transformers,13, no. 2, 510–
should be made to reduce the operating time of relay. More 517, (1998).
study is needed in setting up the threshold criterion of the
relay. [14] A. Wiszniewski and B. Kasztenny, IEEE Trans. Power
Del.,A multi-criteria differential transformer relay based
5. REFERENCES on fuzzy logic, 10, no. 4, 1786–1792, (1995).
[1] Ge Baoming, A T. de Almeida, ZQionglin, and [15] D.P.Kothari, I.J.Nagrath, Electrical Machines;
W.Xiangheng,IEEE Trans. on Power Delivery, An 6thedition.; New Delhi : McGraw Hill
equivalent instantaneous inductance based technique for Education(India)(2013).
discrimination between inrush current and internal faults
in power transformers, 20, 2473-2482, (2005). [16] S.V. Kulkarni, S.A. Khaparde, Transformer Engineering
Design and Practice, New york : Marcel Dekker, 2004.
[2] J. S. Thorp and A. G. Phadke,,IEEE Trans. Power Appl.
Syst.,A microprocessor based three phase transformer [17] Z. Liu, S. Liu, and O. A. Mohammed, IEEE Transactions
differential relay,PAS-101, 426–432, (1982). on Magnetics, A Practical Method for Building the FE-
Based Phase Variable Model of Single Phase
[3] A. G. Phadke and J. S. Thorp, IEEE Trans. on Power Transformers for Dynamic Simulations,43,no. 4, 1761-
Apparatus and Systems, A new computer-based flux 1764, (2007).
restrained current-differential relay for power
transformer protection,PAS-102, 3624-3629, (1983). [18] J.Takehara, M.Kitagawa, T.Nakata and N.Takahashi,
IEEE Trans. Magn., Finite element analysis of inrush
[4] M. Jamali, M. Mirzaie, S. A. Gholamian and S. currents in three phase transformer,23, 2647-2649, no.
MahmodiCherati, IAPEC IEEE, A Wavelet-Based 5,(1987).
Technique for Discrimination of Inrush Currents from
Faults in Transformers Coupled with Finite Element [19] D. Phaengkieoi, W. Somlak2 and S.
Method, 138-142, (2011). Ruangsinchaiwanich, ICEMS IEEE, Transformer Design
by Finite Element Method with DOE Algorithm.,2219-
[5] X.L.Feng, J.C.Tan, Z.Q.Bo, UPEC IEEE, A new wavelet 2224 (2013).
transform approach to discriminate magnetizing inrush
current and fault current,3,876-880, (2006). [20] R. Yacamini,H. Bronzeado, IEE Proceedings science,
measurement and technology, Transformer inrush
[6] K. Inagaki, M. Higaki, Y. Matsui, K. Kurita, M. Suzuki, calculations using a coupled electromagnetic model,
K. Yoshida, and T. Maeda, IEEE Trans Power Del. 141,491-498 (1994).
Digital protection method for power transformers based
on an equivalent circuit composed of inverse [21] J. Jesus Rico, Enrique Acha and Manuel Madrigal, IEEE
inductance3, 1501–1510, (1988) Transactions on Power Delivery, The Study of Inrush
Current Phenomenon Using Operational Matrice, 16
[7] T. S. Sidhu and M. S. Sachdev, IEEE Trans. Power Del., ,231-237, no. 2, (2001).
On line identification of magnetizing inrush and internal
faults in three phase transformer, 7 , no. 4, 1885–1891, [22] M.R.Feyzi and M.B.B.Sharifian, IPEMC IEEE,
(1992). Investigation on the factors affecting inrush Current of
Transformers Based on Finite Element Modelling,1, 1-5,
[8] K.Yabe, IEEE Transactions on power delivery, Power (2006).
differential method for discrimination between fault and
magnetizing inrush current in transformers,12, no. 3, [23] JawafFaiz, Bashir Mahdi Ebrahimi and TahereNoori,
1109-1118, (1997) IEEE Trans Magn.,Three and two dimensional finite
element computation of inrush current and short circuit
electromagnetic forces on windings of a three phase core
type power transformer,44, no. 5, 590-597,(2008).

40
 International Journal of Computer Applications (0975 – 8887)
International Conference on Advances in Emerging Technology (ICAET 2016)

[24] W.Neves, D. Fernandes , F. J. A. Baltar , A. J. P. Transform Based Discrimination Between Inrush and
Rosentino , E, Saraiva , A. C. Delaiba , R. Guimaraes , Internal Fault of Indirect Symmetrical Phase Shift
M. Lynce , Jose Carlos de Oliveira, EPQU IEEE, A Transformer,1-5, (2014).
comparative investigation of electromechanical stresses
on transformers caused by inrush and short-circuit [34] V. Barhate, International Journal of Electrical and
currents,1-6,(2011). Electronics Engineering Research, A review of
distinguishing schemes for power transformer’s
[25] M. Steurer and K. Fröhlich, IEEE Trans.Power Del.,The magnetizing inrush and fault currents, 3, 277-284,
impact of inrush currents on the mechanical stress of (2013).
high voltage power transformer coils, 17, no. 1, 155–160,
(2002). [35] Y. Hu, D. Chen, X. Yin and Z. Zhang, Transmission and
Distribution Conference and Exposition,IEEE PES, A
[26] A.A. Adly, IEEE Trans Magn. ,Computation Of Inrush Novel Theory for Identifying Transformer Magnetizing
current forces on transformer windings, 37,no. 4,2855- Inrush Current, 1, 274-278,(2003).
2857,(2001)
[36] Z. Han, S. Liu, S.Gao, Z. Bo, UPEC IEEE,A Novel
[27] C.K Cheng, T.J.Liang, J.F. Chen,S.D. Chen and Detection Criterion for Transformer Inrush Based on
W.H.Yang , IEE Proc. Electr. Power Appl., Novel Short-Window Filter Algorithm, 1-5, (2008).
approach to reducing the inrush current of a power
transformer, 151, No.3, 289-295 (2004). [37] J. Ma, Z. Wang, J. Wu, IPEC IEEE, A novel method for
discrimination of internal faults and inrush currents by
[28] Y. Cui, S. G. Abdulsalam, S. Chen and W. Xu,IEEE using waveform singularity factor, 1035-1039, (2010).
Trans. Power Del., A sequential phase energization
technique for transformer inrush current reduction part- [38] J. Ma, D Ye, Z. Wang, J, Wu IPEC IEEE, Identifying
1,20, no. 2, 943–949, (2005). Inrush Current Using Sinusoidal Proximity Factor, 215-
219, (2010).
[29] W. Xu, S. G. Abdulsalam, Y. Cui, and X. Liu,IEEE
Trans. Power Del., A sequential phase energization [39] M. Jamali, M. Mirzaie, S. A.Gholamian, S. M.Cherati,
technique for transformer inrush current reduction part- IAPEC IEEE,A Wavelet-Based Technique for
2,20, no. 2, 950–957,(2005). Discrimination of Inrush Currents from Faults in
Transformers Coupled with Finite Element Method, 138-
[30] H. S. Nankani1, R. B. Kelkar, International Journal of 142,(2011).
Science and Research, Review on Reduction of
Magnetizing Inrush Current in Transformer, 4, 234- [40] H. Abniki, H. Monsef, P. Khajavi, and H. Dashti, MEPS
242,(2015). IEEE, A Novel Inductance-Based Technique for
Discrimination of Internal Faults from Magnetizing
[31] M.Jing, W.Zengping, Power Engineering Society Inrush Currents in Power Transformer, 1-6, (2010).
General Meeting IEEE, A Novel Algorithm for
Discrimination Between Inrush Currents and Internal [41] W.C. Wu, T. Y. Ji, M. S. Li; L. L. Zhang; Q. H.
Faults Based on Equivalent Instantaneous Leakage Wu,Power and energy society general meeting IEEE,
Inductance., 1-5, (2007). Inrush Identification by Applying Improved
Morphological Gradient Algorithm, 1-5, (2013).
[32] J.Ma, Z.Wang, S.Zheng, T.Wang, Q. Yang, Canadian
Journal of Electrical and Computer Engineering, A new [42] W. Wu, T. Ji, M. Li, Q. Wu, IET generation,
algorithm to discriminate internal fault current and inrush transmission and distribution, Using mathematical
current utilizing features of fundamental current, 36, 26- morphology to discriminate between internal fault and
31,(2013). inrush current of transformers, 10 , 73-80, (2016).

[33] S.K Bhasker, M. Tripathy,V Kumar, PES general [43] S.Y Hong and W. Qin, ICPST IEEE, A wavelet-based
meeting conference and exposition IEEE, Wavelet method to discriminate between inrush current and
internal fault, 2, 927-931, (2000).

IJCATM : www.ijcaonline.org
41