Fault Analysis in Transmission Line Using MATLAB

Rushikesh Masule
Electrical Engineering Department Gaurav B. Patil Ruhul Desale
SVKM’s Intitute of o, Dhule Electrical Engineering Department Electrical Engineering Department
Dhule, India SVKM’s Intitute of Technology, Dhule SVKM’s Intitute of Technology, Dhule
[email protected] Dhule, India Dhule, India
[email protected] [email protected]
Nikhil More
Electrical Engineering Department Sandeep Ushkewar Gaurav Kele
SVKM’s Intitute of Technology, Dhule Electrical Engineering Department Electrical Engineering Department
Dhule, India SVKM’s Intitute of Technology, Dhule SVKM’s Intitute of Technology, Dhule
[email protected] Dhule, India Dhule, India
[email protected] [email protected]

Abstract - Since the demand for electricity has issues. The dependability and affordability of the electrical
skyrocketed, networks of electric power have become system are adversely affected when transmission line
increasingly important. Consequently, there are now a failures occur. As a result, recognizing and accurately
considerably greater number and length of power diagnosis. The failures of the AC high voltage transmission
transmission lines. A widespread power loss could be lines shown in fall into the following categories: Seventy-
caused by a transmission line tripping or interruption. eight percent of a fault is single line to ground (SLG),
This means that these lines must be effectively seventeen percent is line to line (LL), ten percent is double
safeguarded. An effective protective system is produced line to ground (LLG), and three percent of lines have a fault
through the examination of faults at various loads, which (LLL).
aids in transient identification and, eventually, in the
localization, detection, and classification of power system
defects. This paper describes the faults in discrete
wavelet transmission lines. Among the faults examined
were double-line, three-phase, and line-to-ground faults.
The MATLAB Simulink environment has been used to
conduct extensive research on defect identification.
Keywords: discrete wavelet transform The Phase Measurement Unit (PMU), Support Vector
transmission line, power, mother wavelet, MATLAB Machine (SVM), Artificial Neural Network (ANN), Wavelet
Simulink; system Transformation (WT), WT with fuzzy logic, and WT with
ANN techniques are among the technologies that have been
I. Introduction studied and analysed by researchers for defect diagnosis and
Power Wavelet network fault detection analyses have been detection. (Source:) Voltage fluctuations can be used in two
used to reduce equipment damage from short circuits. This ways to identify a PMU issue: As stated in [2], the first
is accomplished by promptly locating the problematic line phase was matching the matching degree with the degree
and turning it off. M. Tarafdar Hagh et al. [1] used a single index to identify problematic buses because the matching
circuit, two machine power system simulation to find and degree was low. Mathematical methods were utilized in the
isolate transmission line issues. They employed an artificial second step to identify the precise location of the flaw.
neural network to categorize and locate errors. O. Dag et al. Networks with long transmission lines use this technique.
[2] developed a classifier that could identify ten distinct Errors can be categorized and sorted using artificial neural
types of 3-phase system failures using six-channel data of networks. Transmission line networks in ring-connected
current and voltage signals from a power system, hence power systems are susceptible to power interruptions. Using
enhancing fault classification performance. A case study of a neural network (ANN) technique with backpropagation,
a distribution system in Istanbul, Turkey was looked at in which delivers voltage and current as input sources, the
order to establish the recommended approach. Using flaws of this method are discovered. This method does a
Computer Aided Design (Power Systems) Design, simulate respectable job of identifying and categorizing transmission
Electromagnetic Transients integrating DC. A technique for line problems. But this approach can deal with the issue
classifying and classifying different transmission line more simply than with nonlinear loads.
operating conditions was developed by M. Geethanjali et al. Three currents and three voltages from a three-phase source
to ensure the timely and reliable operation of protective made up the six inputs in the ANN approach described in.
relaying networks. The network has four outputs in total. In the power system,
Adopting a modern, precise, and safe power system is MATLAB Simulink is used to model two generators and
imperative due to the intricate transmission line network that transmission lines. Every cycle's sampling rate is set at 1
comprises the energy grid of today. Compared to other kHz in order to train the artificial neural network data.
critical components of the power system, electrical Identifying and categorizing errors are the two steps of the
transmission lines often have a greater failure rate since their process outlined in [3]. Are identifying and classifying
start. Experts in power protection therefore place a great mistakes. Six ANN system inputs are created, and for fault
importance on the ability to quickly recognize and diagnose
detection, we compare the inputs to the pre-fault value. The 11.7% in the SFCL's limitation when the copper stabiliser
output of an ANN system can malfunction or not; a simple 1 layer was lowered from 40 µm to 20 µm was unaffected by
or 0 communicates this. The same process is used to classify a 3 K temperature increase.
faults, with the four. The following outputs could be D. Guillen, M. Roberto, [3] This shows that the signals'
generated by an ANN system: A, B, C, or G. The simulation detail coefficients are almost zero (straight line) in the
results demonstrated that the defect detection and absence of an error, with the only effect being the
classification capabilities of the concept of ANN system Daubechies wavelet's conclusion, which is likewise
design worked as expected. The learning machines known minuscule and almost zero (1*10-3) and that each signal's
as support vector machines are a powerful technique for energy is present with a negligible rate of change in energy
regression and classification. When employing support following signal compression, close to 0.0. Due to the high-
vector machines as classifiers, there are two primary phases frequency component that fault inception introduces into
involved: training and testing. Support vector machines can signals, which has a substantial impact on the detail
classify data in two ways: linear and nonlinear. Series coefficient, the ratio of energy change from the starting level
compensated transmission line faults (SVM) are the primary was calculated while these signals were compressed.
application for support vector machines. The support vector maintaining the approximation. Finding and shutting off the
machine receives a single cycle of series-compensated circuit breaker at a line required one cycle of fault creation.
transmission lines delivering three-phase, zero-sequence Using MATLAB, we simulated a number of errors. We
currents. Lastly, each support vector machine's output— utilized the wavelet toolbox to decompose the transient
which can have one of two values—must be used to signals after they were recorded in order to obtain the
categorize defects. maximum details coefficient and energy. We then
compressed the signals and calculated the ratio of energy
change from the first level while keeping the approximation
II.Literature Review because fault inception adds a high frequency component to
N. Iqbal, K. R. Abbasi, R. Shinwari, [1] Analyse a the signals that significantly affects the detail coefficient.
photovoltaic farm's transmission line fault behaviours. The The signals were first analyzed after one cycle of fault
MATLAB/Simulink environment has been used to model formation at a line in order to identify the fault and turn off
the solar farm's transmission lines and power system the circuit breaker. We simulated several errors using
architecture. It is estimated that there are two kilometres MATLAB. To ascertain the ratio of energy change from the
separating the DCDC converters and the source of the issue. first level and the ways in which faults impact the energy of
A capacitance of 0.5 mF is anticipated for each DC-DC these signals, we compressed the decomposed transient
converter, which keeps the DC-link voltage constant. This signals using the wavelet toolbox. This was carried out after
covers the inductance and resistance that exist between the the signals were recorded to obtain their maximum detail
failure point and the converters. At position "Fault" on the coefficient, or energy. In our simulation undertook.
transmission line, we applied a pole-to-pole fault for twenty
milliseconds, from t=1.0 sec to t=1.02 sec. The transmission Mohammad Hassan Khooban, and Taher Niknam, [4] Fault
line current with and without SFCL, as well as with SFCL analysis and an evaluation of the SFCL performance under
and a 40 µm copper stabilizer. Two types of copper different fault scenarios have been carried out using a
stabilisers are used: the SFCL 20 µm and the SFCL 40 µm. MATLAB/Simulink model of a community solar farm. The
Without requiring an SFCL, the current surged to 4.1 kA 1 MW solar farm is comprised of four 250 kWp PV arrays.
during the fault due to the converters' instantaneous A DC/DC converter increases the output voltage from 600 V
capacitor discharge at t= 1.0 sec, as shown by the blue line. to 3 kVDC in order to connect each PV array to the main
Utilizing the SFCL with a 40 µm copper stabiliser allowed DC bus. Subsequently, an inverter is used to flip the DC bus
for a 22% reduction in fault current, as indicated by the red so that it can be connected to the utility. For DC system fault
dotted line, to 3.2kA. A 20 µm copper line (yellow dotted analysis, one of the most crucial variables is the line
line) stabiliser reduced the fault current to 2.8 kA, compared impedance. Due to their large amplitude fault currents and
to the fault current without the SFCL, which resulted in an lack of zero crossing spots, DC power systems are thought
almost 31.7% drop in the SFCL's drop percentage. to require protection. an essential. Factors such as voltage
E. Kato and A. Kurosawa, [2] Electricity passing through level, grounding impedance, line impedance, and DC-link
the SFCL and over the 40 µm copper stabilizer. Overall capacitance influence the fault current behaviour. Both pole-
current through the fault is shown by the blue line. The red, to-ground and pole-to-pole DC faults can occur in bipolar
yellow, purple, and green lines represent, respectively, the systems. A fault that connects the positive and negative
current flowing through the copper, silver, and YBCO poles has a high voltage and low line impedance, making it
layers. The red dashed line indicates that the YBCO layer the most dangerous type. By contrast, a pole-to-ground fault
was currently conducting electricity prior to the fault. The occurs when the ground is connected to the positive or
YBCO resistivity spiked as a result of the current being negative pole directly. Short circuits of the pole-to-ground
diverted to other layers as the temperature increased after type are more frequent than those of the pole-to-pole type,
the fault. Because the copper stabiliser had the lowest but they are less dangerous. Due to its significant challenge,
resistance, the copper flowed through it the most during the this paper focuses on the pole-to-pole fault. Because it has a
failure. The current flow of the SFCL 20 µm copper higher fault current and presents a significant threat to
stabiliser is shown. The total current flowing through the circuit breakers, the pole-to-pole fault is the subject of this
fault is represented by the blue line. Throughout the fault paper.
period, he was fully current. The red, yellow, purple, and X. Li, Q. Song, W. Liu, H. Rao, S. Xu, and L. Li [5]
green lines represent the current flowing through the copper, "Shielding of MMC-based HVDC systems from no
silver, and YBCO layers in that order. The improvement of permanent faults on DC overhead lines, "The maximum,
minimum, standard deviation, and energy levels of the that is almost always zero (a straight line) and only shows
wavelet detailed coefficients are all higher in the presence of up in the absence of a defect. Every signal has energy, and
a fault than in the absence of one. This indicates that the that energy changes at a very small rate (close to 0.001)
error originated in phase A. Wavelet analysis can be used to
following signal compression. Daubechies wavelet also has
distinguish and identify different types of faults. For
example, shaded data can be used to compare the deviation an extremely small final effect that is close to zero (1*10-3)
of the statistical values for the Bus-1 fault from the healthy and very small.
condition and the energy values for the various faulty phase
voltage and current wave breakdown stages with the
corresponding energy levels of the healthy condition. This
applies to symmetrical faults in different phases that occur at
different bus locations or locations, such as three-phase
symmetrical faults, line-to-line, double-line-to-ground, and
single-line-to-ground.[6]-[10] Every time a phase fault has
occurred, the wavelet detailed coefficient's maximum,
minimum, and standard values as well as the usual
behaviour of the energy level data have deviated from
values under healthy conditions. It might be sufficient to
compare and analyse the wavelet data in great detail in order
to distinguish between the various system states (defective
and healthy). The same methodology was used to gather Figure .2: Three phase current signals at normal
statistics about faults and energy level data at the other five
locations.[11]-[14]
B. Single phase to ground fault
Signals for current in three phases. The ground fault is
associated with phase A in Figure 3, which shows three-
III.CIRCUIT AND DESCRIPTION phase current signals. Although there was little to no change
Two three-phase sources, two three-phase transformers, and in the other two phases compared to the massive amount of
a three-phase load make up the circuit. Selecting the ratings change during the A phase, the arrow pointed directly
of different components, including two transmission line toward the beginning of the fault inception (around sample
lengths, two three-phase transformers, and three-phase 1800, which is half the number of samples from the initial
sources, is crucial for the simulation study's objectives [3]. signal fault inception time). Along with the energy change
The MATLAB software model created using the Sim Power ratio being higher than normal and the detail coefficient
System for the implementation of the simulation of various being greater than 0.001, the data also clearly showed that
failures is shown. The simulation study used a discrete the gearbox line was malfunctioning at that particular point.
PowerGrid with a sample time of 3e-05s under various fault
scenarios.

Figure.3: Three phase current signals at Single phase to

ground Fault

Figure. 1: Transmission line model in MATLAB C. . Double phase to ground fault

Just two of the faulty phases at the fault inception time show
a high degree of detail coefficient and significant change,
while the healthy phase shows almost no change (Figure 4).
A. Normal Condition The figure shows three-phase current signals with phases A-
B to the ground fault. The information in Table 4 further
A yellow, B blue, and C red phase current signals at no fault supports this, demonstrating that even though the defective
condition, along with their detail coefficient, are typical phases were very different from each other and had
setups. It follows that these signals have a detail coefficient
maximum detail coefficients higher than 0.001, they were each DC-DC converter should contain a capacitor of about
still in a defective state. Furthermore, it suggests that the 0.5 mF. The transmission line in Fig. 1 was subjected to a
defective phases were connected to the transmission line's pole-to-pole fault at the "Fault" position for 20 milliseconds,
ground because the amounts of energy that changed for the starting at t=1.0 and terminating at t=1.02 seconds. Figure 4
two faulty phases when compression was applied varied. (a) depicts the transmission line current with and without
SFCL, as well as a 40 µm and 20 µm copper stabilizer.
Figure 4 (a) illustrates that without SFCL, the current rose to
4.1 kA when the fault occurred at t= 1.0 sec because the
converters' capacitors emptied immediately. Using the SFCL
with a 40 µm copper stabiliser resulted in a 22% decrease in
fault current, as seen by the red dashed line in Figure 4 (a).
The SFCL obtained a fault current limitation of 2.8 kA and a
decrease percentage of approximately 31.7% when utilizing
a 20 µm copper stabiliser (yellow dotted line). Figure 4(b)
shows the current flowing through the SFCL 40 µm copper
stabiliser. The blue line depicts the whole current flowing
through the fault. Each layer's current is shown by a
separate-colored line: red for YBCO, yellow for copper,
purple for silver, and green for Hastelloy. In Fig. 4 (b), the
Figure. 4: Three phase current signals at Double phase to red dashed line indicates that the current went through the
ground Fault YBCO layer before the fault. The current was directed to the
other layers due to a quick increase in YBCO resistivity
D. Double phase fault produced by the fault's formation and subsequent
Three-phase current signals with faulty phases A–C are temperature rise. Because of its low resistance, the copper
shown in the image below. Out of these, the healthy phase stabiliser received almost all of electricity after the incident.
exhibited almost no change, while only two faulty phases at Figure 4(c) depicts the current passing through the SFCL 20
fault time show significant change. This is corroborated by µm copper stabiliser. The blue line depicts the whole current
Table 4's data, which demonstrates that in terms of flowing through the fault. Each layer's current is shown by a
coefficient, energy, and energy change, there was essentially separate-colored line: red for YBCO, yellow for copper,
no change during the healthy phase and a close resemblance purple for silver, and green for Hastelloy. Figure 4 (d) shows
to the normal condition. the ratio Although the maximum that lowering the copper stabiliser layer from 40 µm to 20
detail coefficient of the faulty phases was greater than 0.001, µm reduced the SFCL restriction by 11.7%, despite a rise in
when compression was applied, the energy change ratio of temperature of 3 K. Figure 4(d) displays both the 40 µm and
the two faulty phases was typically the same or 20 µm SFCL heats. The failure occurred when the two
insignificantly different, indicating that the faulty phases SFCLs were at 77 K, the temperature of LN2. Following the
were not connected to the transmission line's ground. It fault, the two SFCLs' temperatures increased to 91 and 94 K,
means that there were challenges with these stages of respectively. The SFCL with a 40 µm copper stabilizer and
development. the SFCL with a 20 µm stabilizer were impacted differently.
The efficiency of the SFCL restriction ability and the
increase in temperature are clearly linked. The SFCL with a
little copper stabiliser improves fault current limitation, but
it raises the temperature.

Figure 5 Three phase current signals at Double phase fault

IV. SIMULATION RESULTS AND DISCUSSION

As shown in Figure 1, the solar farm's power system
architecture and transmission lines have been modelled in
the MATLAB/Simulink environment to investigate
transmission line failure behaviour. The DC-DC converters
are around 2 kilometres from the point of failure. Figure 1
shows that the DC-DC converters have a resistance of 172
mΩ and an inductance of 0.76 mH, as determined by (1) and
(2), respectively. To maintain a constant DC-link voltage,
 The main conclusions, findings, and implications of the
study should all be summarized in the conclusion of a
transmission line fault detection project. The significance of
the project for improving the efficiency and dependability of
electrical power transmission ought to be discussed as well.
This sample project conclusion demonstrates how modern
technologies, such as machine learning algorithms and
phaser measurement units (PMUs), can be used to identify
transmission line faults early on. Faster response times and
the possibility of reducing downtime and damage to the
electrical infrastructure are brought about by this enhanced
detection capability. Fault analysis of three phase
transmission lines is made much easier with the use of
MATLAB and ETAP tools. Our effort is focused on
modelling these problems to obtain a better understanding of
transmission line properties. Single line-to-ground, double
line, and three-phase faults are among the many fault types
that we model. This simulation technology enhances
accuracy and user experience while improving efficiency in
the identification of problems in transmission lines. It can
also be used with more complex power systems because of
its versatility.

References
V. CONCLUSION [1] N. Iqbal, K. R. Abbasi, R. Shinwari, W. Guangcai, M. Ahmad, and K.
Tang, “Does exports diversification and environmental innovation
 achieve carbon neutrality target of OECD economies?,” J. Environ. [8] PriyaYadav “Simulation and Performance Analysis of a lkWp
Manage., vol. 291, no. April, p. 112648, 2021, doi: Photovoltaic System Using PVsyst” 2015 International Conference on
10.1016/j.jenvman.2021.112648. Computation of Power, Energy, Information and Communication.
[2] Qin Jian, Peng Li-ping and Wang He-chun, “Single Terminal [9] A. Wani, P. Mistari and G. B. Patil, "Comparative Analysis of Power
Methods of Traveling Wave Fault Location in Transmission Line Quality Issues in Grid-Integrated Wind Energy System," 2022
Using Wavelet Transform,” Automation of Electric Power Systems, International Interdisciplinary Humanitarian Conference for
vol. 29, pp. 62–65, 2005. Sustainability (IIHC), Bengaluru, India, 2022, pp. 1485-1489
[3] M. Irwanto, Y. Irwan, I. Safwati, W. Leow, and N. Gomesh, [10] G. B. Patil and L. D. Arya, "Voltage Stability Assessment of Grid
“Analysis simulation of the fault dtection output performance, ” and Connected Solar PV System," 2021 Third International Conference on
Optimization Conference (PEOCO'2014), IEEE 8th International, pp. Intelligent Communication Technologies and Virtual Mobile
477–481, 2014. Networks (ICICV), Tirunelveli, India, 2021, pp. 563-568
[4] Mohammad Hassan Khooban, and Taher Niknam, “Fast fault [11] M. Thakur, A. Patil and G. Patil, "The Current State of the Art for PV
detection and classification based on a combination of wavelet Grid Connected System Issues with Power Quality," 2022
singular entropy theory and fuzzy logic in distribution lines in the International Interdisciplinary Humanitarian Conference for
presence of distributed generations,” International Journal of Sustainability (IIHC), Bengaluru, India, 2022, pp. 1515-1520.
Electrical Power and Energy Systems, vol. 78, pp. 455–462, 2016. [12] Patil, G.B., Raghuwanshi, S.S., Arya, L.D. (2023). Effect of PV
[5] M. Monadi, M. A. Zamani, J. I. Candela, A. Luna, and P. Rodriguez, Penetration on the Steady-State Voltage on Grid-Integrated PV
“Protection of AC and DC distribution systems Embedding System. In: Rajakumar, G., Du, KL., Rocha, Á. (eds) Intelligent
distributed energy resources: A comparative review and analysis,” Communication Technologies and Virtual Mobile Networks. ICICV
Renew. Sustain. Energy Rev., vol. 51, pp. 1578– 1593, 2015, doi: 2023. Lecture Notes on Data Engineering and Communications
10.1016/j.rser.2015.07.013. Superpower Inc., “2G HTS Wire Technologies, vol 171. Springer, Singapore.
Specification.” https://www.superpower-inc.com/specification.aspx. [13] S. Baviskar, G. B. Patil and R. Beldar, "Predictive Analysis of 50 KW
[6] Bindu U Kansara and B.R. Parekh, “Modelling and simulation of Solar Photovoltaic System using PVsyst," 2024
distributed generation system using HOMER software”, IEEE 2011. InternationalConference on Intelligent and Innovative Technologies in
[7] Maria A. Mazuera, Johnny Posada and Paul Manrique, “Analysis and Computing, Electrical and Electronics (IITCEE), Bangalore, India,
Emulation of a fault detection System Using a MATLAB/Simulink 2024, pp. 1-4.
Model and the DS1104 Unit”, IEEE October 2016.