2020 IEEE International Conference on Computing, Power and Communication Technologies (GUCON)

Galgotias University, Greater Noida, UP, India. Oct 2-4, 2020

Methodologies for the Detection of Magnetizing Inrush

and Fault Condition in Power Transformer
Maulik Raichura Nilesh Chothani Dharmesh Patel Jigyesh Sharma
Electrical Department, Electrical Department, Adani Electrical Department, Electrical Engineer at Worley,
Gujarat Technological University Institute of Infrastructure Government Engineering Mumbai, Maharashtra, India,
(GTU), Gujarat, India, Engineering (AIIE), Ahmedabad, College, Bharuch, Gujarat, India, [email protected]
[email protected], Gujarat, India, [email protected]
[email protected]

Abstract: Trustworthy protective scheme of a power To design a transformer and its protective schemes
transformer is the key criterion due to its vital part in power generally main parameters considered are voltage, flux and
system. Protective schemes are mainly affected by nonlinear current rating.
characteristics of transformer core and its inrush condition.
Hence, a dedicated transformer protection scheme is the Maximum applied voltage can be derived as,
necessity for smooth operation of power system. This article డ‫׎‬೘
includes effect of inrush, past methodologies for detection of ܸ௠ •‹ሺ߱‫ ݐ‬൅ ߠሻ ൌ ݅௠ ܴ௣ ൅ ܰ௣ (1)
డ௧
inrush and fault condition. Several journal articles, various
books and research thesis are thoroughly analyzed for the Where, Vm, ω, im, Rp, ϕm, Np are maximum voltage,
identification of inrush and fault conditions. Different angular velocity, magnetizing current, primary resistance,
methodologies used for successful discrimination of fault and instantaneous flux and primary turns sequentially.
inrush situations are elaborated by comparative evaluation.
Moreover, various inrush and fault conditions are performed on
hardware developed in laboratory environment. Authors have
also suggested two techniques for transformer inrush and fault
identification purpose in this article.

Index Terms—Transformer Protection, Inrush, CT

Saturation, Fault.

I. INTRODUCTION
As being a heart of power system, protection of power
transformer is a burning issue in HVAC system. Usually for
stability point of views, protective schemes desire accurate
and high speed fault clearing features to avoid anonymous
damages with system disturbance under faulty conditions in
more than 132kV line [1]. Nonlinear characteristics of the core
in power transformer produces magnetizing inrush which
results in erroneous measurement of authentic quantities.
During no load condition, transformer when energized, it
Fig. 1. Flux Linkages, Core Characteristics and Magnetizing Current in
generates 5 to 10 times transient current than rated normal Power Transformer
current, which persist up to several cycle, is called an inrush
current. It is unbearable to avoid inrush in actual practice Instantaneous flux is derived as,
which provoke mal-operation of protective scheme. Residual
ೃ೛
magnetism (flux) share key role in core saturation. As per ି
ಽ೛
௧
constant flux linkage theorem [2], due to core saturation, flux ‫׎‬௠ ൌ ൫‫׎‬௣௠௔௫ ܿ‫ ߠݏ݋‬േ ‫׎‬௥௘௦௜ ൯݁ െ ‫׎‬௣௠௔௫ ‘•ሺ߱‫ ݐ‬൅ ߠሻ(2)
becomes three times as shown in fig.1. Polarity of the core Where, ϕpmax = maximum value of flux associated with
remanence also perform major role in inrush current primary, θ = switching angle, & ϕresi = residual flux.
generation. Tremendous inrush current is provoked in primary
winding of the transformer at instant of primary voltage zero Decaying rate is high in starting time of system due to
crossing and same polarity of core remanence. involvement of steady state AC flux with transient DC
component flux. Due to core saturation, value of L is not
As shown in fig.1 normally core is saturated near “knee” constant so time constant L/R also varies. The nonlinear
region of hysteresis characteristic loop and inrush current is portion of core (B-H curve (knee portion)) is responsible for
initialized in the system. Also, during core saturation generating harmonics in secondary voltage wave.
conditions, current (inrush) is very rich of harmonics due to
sharp reduction in winding inductance. Due to low impedance Transformer inrush is further sub divided in three portion
of differential relay in power transformer, inrush condition (1) Initial inrush (magnetic inrush) (2) sympathetic inrush (3)
affect adversely [3]. In large transformers, winding resistances recovery inrush [4]. Sympathetic inrush is generated when two
are lower and inductances are high, so inrush current remain transformers are placed parallel and one of them is energized
presents for several seconds. without load, at that time inrush effect is observed itself in
transformer but also affects side by transformer which is
known as sympathetic inrush. Recovery inrush is observed
when transformer voltage is restored after clearance of

978-1-7281-5070-3/20/$31.00 ©2020 IEEE 146

Authorized licensed use limited to: Ontario Tech University. Downloaded on March 12,2024 at 22:33:07 UTC from IEEE Xplore. Restrictions apply.
temporary fault outside the transformer zone. It is good to note 2) Wave-form examination based Techniques
that the Severity of sympathetic inrush and recovery inrush are Figure 3 shows basic diagram of different waveform
not as high as that of initial inrush. recognition techniques in protective schemes. The electrical
parameters fetched from the transformer are collected in
II. MAGNETIZING INRUSH DETECTION parameter block. These data may be of voltage or current. The
METHODOLOGIES desired parameter can be selected with the help of parameter
Due to magnetizing inrush current, transformer terminal selection block. These collected parameters are then given to
currents may imbalance on primary and secondary sides and the block that contains Anti Aliasing Filter (AAF), Signal
due to this; possibility of protective scheme mal-operation is Conditioner (SC) and Analog to Digital Converter (ADC).
taking place. Due to these difficulties, researchers have The collected signals will be filtered through AAF and then
presented various inrush current detection methodologies, through SC block, signals will get conditioned and converted
which are described in subsequent section. into digital form with the help of ADC. Waveform analysis
will be done based on the gradient [6], angle difference [7][8]
A. Various Techniques to Detect Inrush Conditions [9], derivation [10] [11], shape of the waveform [12], size of
Among many protective schemes and research work the waveform, rate of change of magnitude or waveform
carried out by researchers, inrush detection schemes are symmetry. The above mentioned analysis extracts important
divided in 6 major categories with its advantage, command information of waveform and then compared with the
and general block diagram. reference threshold [13]. Post comparison, decision of trip or
blocking signal will be made by the processor through a
1) Harmonics and DC Component based Analysis developed scheme.

Fig. 2. Harmonic and DC Component based Analysis

Figure 2 shows the logical block diagram based on

Fig. 3. Waveform based Analysis
harmonic and decaying DC component. In the block diagram,
voltage, current or power data are fetched from both sides of 3) Decomposing and Machine Learning (Classifier)
transformer. These collected data are then given to Analog to
Figure 4 shows elementary block diagram for
Digital Converter (ADC) as shown in the fig. 2, which
decomposing and classifier based techniques. Wavelet
converts collected analog data into digital form. Filters (High
Transform (WT) [14], Wavelet Packet Transform (WPT) [15],
pass filter or low pass filter or band pass filter) are evolved in
Hyperbolic S-Transform based decomposing and Artificial
between to fetch out the required frequency signals and
Neural Network (ANN) [16], Support Vector Machine (SVM)
eliminate the noise signals. After that, these data are given to
[17] [18], Relevance Vector Machine (RVM) [19] and
Harmonic or DC component separation block, which will
Extreme learning machine [20] type classifier techniques are
analyze the harmonics or DC component present in the
implicated in this analysis. The electrical or non-electrical
signals. Another block shown here is used for the purpose to
parameters fetched from the transformer are collected in
block, bypass and provide desired delay in signals and to
parameter block. These data may be voltage, current,
conditioner it. These data are then processed in a processor
frequency, Dissolved Gas, oil and winding temperature etc.
which is shown as a junction point here. After all these
The collected data is conditioned and properly sampled by
process, the calculated values are then compared with the
ADC block.
preset threshold. The processor also works as comparator here
and generate trip or blocking signal based on the result of
comparator. Here, “1” shows the trip signal during faulty
condition and “0” presents blocking signal which ultimately
shows the healthy condition of the transformer. The proposed
block diagram of the protection is generalized and hence we
can select the component as per the requirement. Generally
2nd, 3rd and 5th harmonics are utilized and mostly 2nd harmonic
effect are considered because it is dominant, up to 63% of
fundamental current [5]. Advantages of the schemes are: It is
not considerably exaggerated by the variation of sampling &
method needs to calculate only one harmonic component for
discrimination. Contrarily, disadvantages of the scheme,
Harmonics and DC components are generated during external Fig. 4. Decomposing and Classifier based Analysis based Techniques
fault with CT saturation & due to nonlinear load appearance,
which results in disconnections, so, protective scheme may Analysis includes structural, behavioral and functional or
mal-operate during external fault with CT saturation. combination of these. For above mentioned analysis
Moreover, nonlinear load also generates harmonics. decomposing method or machine learning (classifier) based
technique is used. A comparator then compares the measured

147
Authorized licensed use limited to: Ontario Tech University. Downloaded on March 12,2024 at 22:33:07 UTC from IEEE Xplore. Restrictions apply.
data with predefined threshold. After that, a processor will by inrush currents [25] based on which analytical, numerical
decide the suitable signal to issue trip signal or not. and experimental approach are used to identify inrush.
Limitations: (1) Easily portray excessive detail so increases However, the computer program is expensive and good
complexity (2) No surety regarding captured details and technical skill is required. Inrush condition may change the
relation with every necessary component (3) Easily generate structural behavior of transformer with extensive axial and
complicity with functional activity (4) Functional radial force [26]. It is also justifiable with the help of axial and
decomposition is generating criticism on algorithm. redial displacement of winding whether it is due to inrush,
internal fault, external fault or mechanical defects [27].
4) Power consumption based techniques
Figure 5 shows the block diagram of the scheme that 6) Methods to minimize inrush current in transformer
works with the help of active and reactive power based Main factors that have effect on inrush condition are
analysis. Initially the voltage, current and power of the instant of energization voltage, magnitude and polarity of
transformer are measured. With the help of a comparator the remnant flux, primary winding resistance, source inductance,
analyzed power is then compared with reference threshold and winding air core inductance between winding and core, core
depends on these comparisons, processor will decide to issue material geometry and magnetic characteristics, ability of
trip signal or not. Active power is almost zero during maximum flux carrying capacity of core etc. Nowadays,
energizing period, while during an internal fault, it consumes renewable converter provide additional flux to reduce
large active power [21]. However, mal-operation takes place magnetic flux due to inrush and reduce stress called Low
during first cycle due to higher active power consumption for Voltage Ride Through (LVRT) resources [28]. In power
few samples only, after that it is reduced and also detection system, to suppress inrush current, fault current or any rapid
time is larger [10]. change in current, Transient Current Limiter (TCL) is used
[29]. Even switching angle also plays a major role in inrush.
Inrush condition identification based on instantaneous So, with controlled switching techniques inrush can be
frequency of differential power signal [22] can accurately minimized [30]. But, switching techniques required proper
identified with less than a quarter cycle in comparison of operation with resolution and preciseness. To avoid inrush in
ANN, WT, 2nd harmonic based technique and active power
transformer, it is required to connect inductive load
consumption based techniques. Time-domain analysis of synchronously near supply. Also inrush blocking scheme [31]
differential power signal [23] employed to detect magnetizing based on amplitude, 2ndharmonic content, saturation state etc.
inrush in power transformers. Initially, intrinsic features of are available.
differential power during inrush conditions are introduced
with time domain wave shape classification. III. HARDWARE SETUP FOR VARIOUS RESULT
ANALYSIS
Figure 6 depicts hardware set up which is taken into
consideration for result analysis. Two, 2-kVA, single phase
220/110 V transformers are connected in parallel for
generating various inrush conditions (in Fig. 6 only one
transformer is shown as a model). A Lamp load having 15
ampere rating is connected via switch S4 as a load as well as
contactors is utilized as circuit breakers in line. Various
Fig. 5. Power Consumption based Analysis external and internal faults are generated with the help of fault
switches (S1 and S2 are used as internal faults and S3 for
5) Electro Magnetic Stress Phenomenon based external faults) via a 12 A, 18 ohm resisters (R1 and R2) as
Technique shown in Fig. 6. CTs of suitable rating are connected on both
Transformer protection, also utilize various uncertainty in sides of one of the transformers to measure and investigate the
parameters, Electro Magnetic (EM) stress is one of them. current signals. IEEE guidelines [32] gives direction to build
Nowadays, logical computational and evidential reasoning proposed hardware set up. Following subsection introduces
based techniques are available to detect uncertainty [24]. the quantitative fault and inrush condition generated in
Radial electromechanical stresses in transformers are caused laboratory setup.

Fig. 6. Proposed Hardware Setup

148
Authorized licensed use limited to: Ontario Tech University. Downloaded on March 12,2024 at 22:33:07 UTC from IEEE Xplore. Restrictions apply.
 1) Initial Inrush 3) Recovery Inrush
Initial Inrush is commonly takes place during no load During various disturbances like sudden change in
operation of transformer and totally associated to transformers voltages followed by recovery of rated voltage, the effect of
core saturation characteristics. The Core saturation rely on that sudden change and recovery in voltage, will indulged the
switching angle and hence, magnetizing inrush current transformer operation. Because of synchronism in the power
magnitude as well as peak of positive or negative are distinct system network these effects may raise in the transformers.
as per the Fault Inception Angle (FIA) as shown in fig.7. The sudden changes in voltages may take place because of
clearance of fault, voltage swing, transitory trip, auto reclosing
etc. Though the impact is not much prominent like initial
inrush but it may affect the transformer operation. The inrush
generated when the system voltages recuperate to rated
voltage level is known as recovery inrush condition. If the
fault cleared is of three phase type and near to transformer, the
effect of recovery inrush is dominant. Fig. 9 shows the
waveform signature of recovery inrush when a fault is cleared
close to the transformer.

Fig. 9. Waveform during Recovery Inrush

Fig. 7. Waveform during Inrush 4) No Load Magnetizing Current Waveform and its
Harmonic
Many research personals had attempted to alleviate the The transformer when energized at no load condition then
amount of inrush current which is takes place at the time of it fetches some magnetizing current in addition to initial inrush
transformer energization by restructuring the material of current, which is continuously required to energize the
transformer core, by controlling inception angle etc. Various windings. The winding behaves as inductors and hence
other inrush conditions are also takes place in the transformer requires reactive power to be injected. The no load
other than initial inrush like recovery inrush, sympathetic magnetizing current is shown in the fig. 10. From the fig 10,
inrush etc. that are described in the subsequent sections. we can see that the transformer magnetizing current is not pure
2) Sympathetic Inrush sinusoidal but it is somewhat distorted this is because of
The condition like Sympathetic inrush, normally take nonlinear characteristic of the transformer core.
place during power transformers parallel operation. As soon as
the second transformer is energized on no load conditions,
then that particular transformer itself is getting inrush current
and also affect in-service parallel transformer. This situation is
known as sympathetic inrush condition. Sympathetic inrush is
captured on hardware based prototype as shown in figure 8.

Fig. 10. No Load Magnetizing Current Waveform

5) External Fault Conditions

When transformer is undergoing through external fault, it
takes more current than the normal condition. Fault is defined
as the undesired path of the current which has low resistance
and due to which the magnitude of current taken from source
by the fault is too high.
This will lead to stress on various parts of the asset as the
asset is not designed for that much amount of current. This can
be more understood from the fig. 11.

Fig. 8. Waveform during Sympathetic Inrush

149
Authorized licensed use limited to: Ontario Tech University. Downloaded on March 12,2024 at 22:33:07 UTC from IEEE Xplore. Restrictions apply.
 of energization. Contradictory, in the event of fault, due to low
magnetizing current and non-saturation of core, the
instantaneous inductance remain almost constant. Hence, the
fault current has high slope at the inception of fault in
transformer.
B. Time interval between two peaks
The peak of fault current is displaced by half a cycle if the
case is of fault. Whereas for inrush condition, the peak is
Fig. 11. Waveform of Primary and Secondary Current under External Fault displace once in full cycle. Thus, the timing between two
consecutive peaks must be evaluated considering suitable
6) Internal Fault Conditions margin for frequency deviation.
The current waveform during internal fault circumstance is
exemplify in fig. 12. For the period of internal fault conditions V. CONCLUSION
the flow of current increases and which may damage the This paper presents an extract regarding various inrush as
transformer due to electrical and mechanical stresses taking well as fault conditions arising during the transformer
place in the winding and insulation of the transformer. So, to operation. The protection techniques used for detection and
avoid the undesired damages occurred to the transformer and classification of inrush as well as fault in transformer are
consequently to power system, necessary measures should be thoroughly analyzed. Various harmonics, filtration,
taken as early as possible decomposing and classifier based techniques currently in use
for transformer protection are elaborated in this work. Based
on this study authors have pointed out some of the research
gap of different protective schemes. Though, various schemes
such as ANN, fuzzy, GA and decision tree based schemes
provide satisfactory outcome but the required operation time
and accuracy are the main priorities. Moreover, authors have
developed hardware prototype to record, various inrush and
Fig. 12. Waveform of transformer Primary and Secondary side Current under other abnormal conditions. The captured waveforms have
Internal Fault condition been shown in the article which helps readers to understand
different inrush conditions and other abnormalities deeply.
7) Internal Fault alongwith CT Saturation Condition Furthermore, authors have projected transformer protection
For the duration of internal fault condition the current techniques by means of screening of initial slope of the current
fetched from the source is increased which leads to saturate waveform and based on time interval between two peaks of
the CTs. Saturated CTs will deteriorate the waveform of the the current waveform.
current and cannot correctly reflects the magnitude and
waveform of current flowing in the power system. Figure 13 ACKNOWLEDGMENT
elaborate CT saturation condition under internal fault on This work is funded by the Department of Science and
hardware. Technology (DST), SERB, India, under project no.
DST/EMR/2016/006041.
REFERENCES
[1] (Revised by K.G.Mewes) J. Rushton, Power system protection,
Second. London, united kingdom: The institutes of electrical engineers,
1995.
[2] S. V Kulkarni and S. A. Khaparde, Transformer Engineering: Design
and Practice. Taylor & Francis, 2004.
[3] H. Dashti, M. Davarpanah, M. Sanaye-Pasand, and H. Lesani,
“Discriminating transformer large inrush currents from fault currents,”
Fig. 13. Waveform of Primary and Secondary Current under Internal Fault Int. J. Electr. Power Energy Syst., vol. 75, pp. 74–82, 2016.
with CT Saturation Condition [4] W. A. Elmore, Protective Relaying Theory and Applicantions, 2nd ed.
New York, Basel: Marcel Dekker, 2003.
IV. SUGGESTED INRUSH AND FAULT [5] A. R. van C. Warrington, Protective Relays: Their Theory and Practice
DISCRIMINATION SCHEMES FOR TRANSFORMER Volume One. Springer Science & Business Media, 2012.
[6] R. J. N. Alencar, U. H. Bezerra, and A. M. D. Ferreira, “A method to
In the recent era, digital relays provide fast and accurate identify inrush currents in power transformers protection based on the
identification of fault as well as inrush condition of the differential current gradient,” Electr. Power Syst. Res., vol. 111, pp.
78–84, 2014.
transformer. From the above discussion and review of various [7] D. Patel, K. D. Mistry, M. B. Raichura, and N. Chothani, “Three State
papers authors found that, it is easier and faster to identify Kalman Filter based Directional Protection of Power Transformer,” in
these events based on waveform shape analysis. Hence, 20th National Power Systems Conference (NPSC), 2018, pp. 1–6.
current waveform properties can be utilized to detect the [8] Dharmesh D.Patel, K. D. Mistry, N. G. Chothani, “A novel approach to
inrush condition directly compared to internal fault condition. transformer differential protection using sequence component based
algorithm,” J. CPRI, vol. 11, no. 3, pp. 517–528, 2015.
Here, Authors suggesting two methods based on waveform [9] D. D. Patel, N. G. Chothani, and K. D. Mistry, “Sequence component of
recognition techniques. currents based differential protection of power transformer,” in 12th
IEEE International Conference Electronics, Energy, Environment,
A. Initial slope of current waveform Communication, Computer, Control: (E3-C3), INDICON 2015, 2015,
The inrush current setup in transformer has low initial pp. 1–6.
[10] D. D. P. K.D. Mistry, N.G. Chothani, “Transformer inrush/internal
slope due to the change in instantaneous inductance at the time fault identification based on average angle of second order derivative of

150
Authorized licensed use limited to: Ontario Tech University. Downloaded on March 12,2024 at 22:33:07 UTC from IEEE Xplore. Restrictions apply.
 current,” in Asia-Pacific Power and Energy Engineering Conference, [23] A. Hooshyar, M. Sanaye-Pasand, S. Afsharnia, M. Davarpanah, and B.
APPEEC, 2017, pp. 1–6. M. Ebrahimi, “Time-Domain Analysis of Differential Power Signal to
[11] D. Patel, N. Chothani, and K. Mistry, “Discrimination of Inrush, Detect Magnetizing Inrush in Power Transformers,” IEEE Trans.
Internal, and External Fault in Power Transformer Using Phasor Angle Power Deliv., vol. 27, no. 3, pp. 1394–1404, Jul. 2012.
Comparison and Biased Differential Principle,” Electr. Power [24] W. H. Tang and Q. H. Wu, Condition Monitoring and Assessment of
Components Syst., vol. 46, no. 7, pp. 788–801, 2018. Power Transformers Using Computational Intelligence, 1st ed., vol. 58.
[12] A. Sahebi and H. Samet, “Efficient method for discrimination between London, New york: Springer-Verlag London Limited, 2011.
inrush current and internal faults in power transformers based on the [25] R. Guimarães, A. C. Delaiba, J. C. Oliveira, E. Saraiva, and A. J. J. P.
non-saturation zone,” IET Gener. Transm. Distrib., vol. 11, no. 6, pp. Rosentino, “Electromechanical Forces in Transformers Caused by
1486–1493, 2017. Inrush Currents: An Analytical, Numerical and Experimental
[13] E. Ali, A. Helal, H. Desouki, K. Shebl, S. Abdelkader, and O. P. Malik, Approach,” J. Control. Autom. Electr. Syst., vol. 24, no. 6, pp.
“Power transformer differential protection using current and voltage 863–872, Dec. 2013.
ratios,” Electr. Power Syst. Res., vol. 154, pp. 140–150, 2018. [26] W. S. Fonseca, D. S. Lima, A. K. F. Lima, M. V. A. Nunes, U. H.
[14] K. P. Vittal, D. N. Gaonakar, and D. B. Fakruddin, “Development of Bezerra, and N. S. Soeiro, “Analysis of Structural Behavior of
wavelet transform based numeric relay for differential protection of Transformer’s Winding Under Inrush Current Conditions,” IEEE
power transformer,” in TENCON 2003. Conference on Convergent Trans. Ind. Appl., vol. 54, no. 3, pp. 2285–2294, May 2018.
Technologies for Asia-Pacific Region, Oct. 2003, vol. 4, pp. [27] S. Bagheri, Z. Moravej, and G. B. Gharehpetian, “Classification and
1580-1584. discrimination among winding mechanical defects, internal and
[15] S. A. Saleh, A. Aktaibi, R. Ahshan, and M. A. Rahman, “The external electrical faults, and inrush current of transformer,” IEEE
Development of a $d$ #x2013; $q$ Axis WPT-Based Digital Transactions on Industrial Informatics. 2018.
Protection for Power Transformers,” IEEE Trans. Power Deliv., vol. [28] H. C. Chen, P. H. Wu, and P. T. Cheng, “A Transformer Inrush
27, no. 4, pp. 1–8, Oct. 2012. Reduction Technique for Low-Voltage Ride-Through Operation of
[16] N. Kumar, H. Gupta, and S. P. Singh, “Prototype scheme for Renewable Converters,” IEEE Trans. Ind. Appl., vol. 52, no. 3, pp.
differential protection of power transformer using ANN,” in 2008 40th 2467–2474, May 2016.
North American Power Symposium, Sep. 2008, pp. 1–6. [29] P. Amiri and M. Akhbari, “Transient current limiter for suppressing
[17] A. M. Shah and B. R. Bhalja, “Discrimination Between Internal Faults transformer inrush, motor starting and fault currents in power system,”
and Other Disturbances in Transformer Using the Support Vector IET Electr. Power Appl., vol. 11, no. 3, pp. 423–433, 2017.
Machine-Based Protection Scheme,” IEEE Trans. Power Deliv., vol. [30] R. Cano-González, A. Bachiller-Soler, J. A. Rosendo-Macías, and G.
28, no. 3, pp. 1508–1515, Jul. 2013. Álvarez-Cordero, “Controlled switching strategies for transformer
[18] N. G. Chothani, D. D. Patel, and K. D. Mistry, “SupportVector inrush current reduction: A comparative study,” Electr. Power Syst.
Machine Based Classification of Current Transformer Saturation Res., vol. 145, pp. 12–18, 2017.
Phenomenon,” J. Green Eng. River Publ., vol. 7, pp. 25–42, 2017. [31] B. Saravanan and A. Rathinam, “Inrush Blocking Scheme in
[19] D. Patel, N. G. Chothani, K. D. Mistry, and M. Raichura, “Design and Transformer Differential Protection,” Energy Procedia, vol. 117, pp.
development of fault classification algorithm based on relevance vector 1165–1171, 2017.
machine for power transformer,” IET Electr. Power Appl., vol. 12, no. [32] IEEE Guide for Protecting Power Transformers (Revision of IEEE Std
4, pp. 557–565, 2018. C37.91-2000). New York, USA: IEEE Power Engineering Society
[20] M. B. Raichura, N. G. Chothani, and D. D. Patel, “Identification of Sponsored by the Power System Relaying Committee, 2008.
internal fault against external abnormalities in power transformer using [33] R. Nath Shaw, P. Walde, and A. Ghosh, “Review And Analysis Of
hierarchical ensemble extreme learning machine technique,” IET Sci. Photovoltaic Arrays With Different Configuration System In Partial
Meas. Technol., vol. 14, no. 1, pp. 111–121, 2020. Shadowing Condition,” Int. J. Adv. Sci. Technol., vol. 29, no. 9s, pp.
[21] K. Yabe, “Power differential method for discrimination between fault 2945–2956, 2020.
and magnetizing inrush current in transformers,” IEEE Trans. Power [34] R. N. Shaw, P. Walde, and A. Ghosh, “Enhancement of Power and
Deliv., vol. 12, no. 3, pp. 1109–1118, Jul. 1997. Performance of 9x4 PV Arrays by a novel arrangement with shade
[22] A. Hooshyar, S. Afsharnia, M. Sanaye-Pasand, and B. M. Ebrahimi, “A dispersion,” Test Eng. Manag., vol. 82, 2020.
New Algorithm to Identify Magnetizing Inrush Conditions Based on
Instantaneous Frequency of Differential Power Signal,” IEEE Trans.
Power Deliv., vol. 25, no. 4, pp. 2223–2233, Oct. 2010.

151
Authorized licensed use limited to: Ontario Tech University. Downloaded on March 12,2024 at 22:33:07 UTC from IEEE Xplore. Restrictions apply.