Study various method and use Wavelet Transform for

Fault Detection and Classification of Underground

Transmission Line by using MATLAB

Md.Zamilis Siam Diponkar Kundu

Department of EEE Assistant Professor of Department of EEE
Pabna University of Science and Technology Pabna University of Science and Technology
Pabna,Bangladesh Pabna,Bangladesh
[email protected] [email protected]
[email protected]
fix faults. There are so many techniques in an
Abstract: Underground cables are faces with a overhead system like the Phasor Measurement Unit
wide variety of faults due to underground which is capable of identifying the accurate location
conditions, wear and tear, etc. Diagnosing fault of the fault and its types. But in the case of
source is difficult and the entire cable should be underground systems, such systems cannot be
taken out from the ground to check and fix faults. used.[5] So this paper core objective is detact the
This paper presents the details of faults and the underground fault with 100% accuracy by using
Wavelet transform-based technique for fault Wavelet Transform.
detection, classification in the Underground
transmission line. Due to the fault in the power a. Fault location method can be classified as
system, a high-frequency current and voltage
generate and propagate along with the power. These Online method utilize process the sampled voltage
generated signals contain a lot of information and & current to determine the fault points. Online
can be used for fault detection and location. The method for underground cable is less then overhead
high-frequency components generated are extracted line.[6, 7] Example: Artificial Neural Network,
using the wavelet technique and analysis of the Continuous/complex wavelet transform
extracted signals is carried. The MATLAB
Simulink is used to model the underground cable Offline method special instrument is used to test
network and faults at various locations are out service of cable in the field there are two offline
simulated. The resulting waveforms are subjected methods as following: [6, 7]
through a wavelet transform to extract the required
signals for analysis The presented model is user
friendly and can be used as a common platform for In tracer method fault point is detected by walking
both control and power system engineers. The on the cable line. Fault point is indicated form audio
accuracy of the wavelet transform scheme is 100%. able signal or electromagnetic signal. It is used to
pinpoint fault location very accurately.[7]Example:
Tracing current method ,Sheath coil method
Keyword: Fault,Wavelet Transform, SIMULINK,
MATLAB,Underground,Transmission
Terminal method is technique to detect fault
location of cable from one or both ends without
I. Introduction tracing. This method use to locate general area of
fault expedite tracing on buried cable. Example:
The economic growth and development of a country Murray loop method ,Impulse current
depend greatly on the reliability and quality of the method[7]
electric power supply[1].Overhead transmission
lines are being used in the current power system
which is not much reliable or safe so to overcome b. fault in a cable can be classified into
its problem and developed a smart city underground different types such as
cables are used[2]. Underground cable installations
are costly as compared to overhead cables, but Open circuit faults are better than short circuit
underground cables are more preferable since fault because when this fault occurs current
underground cables are not affected by a climatic flows through cable become zero. Open circuit fault
condition like earthquakes, lightning, heavy rain,
winds, snowfall, etc.[1-3] [4] is a kind of fault that occurs as a result of the
conductor breaking or the conductor being pulled
out of its joint sometimes faults occur when one or
In the underground system, it is quite difficult to more phase conductors break[1-4, 7]
find the faulty area than the overhead system
because of a few obvious reasons the entire cable
should be taken out from the ground to check and
Short Circuit Fault when two or more conductor underground fault accurately with the help of
of the same cable in contact with each other ,then IOT(internet of Things).but our country does not
this is called a short circuit fault. It is impossible to have enough internet speed and if there is any
detect visually without taking the cable apart. A interrupt in providing internet then the model or
short circuit fault occurs when the individual technique is not accepted. On the other side wavelet
insulation of the cables is damaged.[1-4, 7] based underground fault detaction does not have
this type of problem.
Short circuit fault can be categorized in two
types Zhiling Long, Nicolas H. Younan Mohatkar [8]the
paper titles as “Underground Power Cable Fault
Symmetrical fault Three-phase fault is called Detection Using Complex Wavelet Analysis” in this
symmetrical fault. In this all three phases are short paper the author used Complex Wavelet Transform
circuited.[2, 7] for identifying the underground fault not that
accurately then Wavelet Transform in wavelet
Unsymmetrical fault In this fault magnitude of transform the mother signal by discrete filter bank
current is not equal & not displaced by 120 and extract the feature then complex Wavelet
degree.[2, 7] Transform.

II. Literature Servay Abhishek Pandey Nicolas H. Younan

[9]“Underground Cable Fault Detection and
Identification via Fourier Analysis”in this paper
A literature survey is necessary to complete a underground fault detact by fourier transform
journal/paper because it describes the previous .Wavelet transform method offers important
works about the working topic and analysis advantages over other methods such as FT due to
thoroughly this analysis supply the researcher with good time and frequency localization
necessary information about the methodologies and characteristics. Analysis results presented clearly
technologies available on that working topic. show that particular wavelet components can be
used as the features to locate the fault in
Mr. M. R. Hans, Ms. Snehal C. Kor, Ms. A. S. Patil underground cables in distribution syste.
[5]the paper titles as “IDENTIFICATION OF
UNDERGROUND CABLE, in this paper the Dhivya Dharani.A, Sowmya.T [10] the paper titles
author” This prototype uses the simple concept of as―Development of a Prototype Underground
OHMs law. The current would vary depending upon Cable Fault Detector‖ —Cable faults are damage to
the length of fault of the cable determined the cables which affects the resistance in the cable. If
underground fault by using a terminal method with allowed to persist, this can lead to a voltage
prototype Murray loop method it identifies the fault breakdown. To locate a fault in the cable, the cable
accurately but the method is a costly and is a slow must first be tested for faults.
method. In this paper, fault detects in a short
transmission line but in my paper, I detect an
underground problems in long-distance III. Theory
transmission lines. And also the proposed model
and fault detection technique is friendly that any When a mathematical function represented by
operator or engineer can be handled easily. wavelets it is a wavelet transform[11]. Wavelet
transform is similar to Fourier transforms, but it has
Dauda A. Folarin, Japhet D. Sakala, Edwin a difference it assumes time localization of various
Matlotse& Mandu A. Gasennelwe-Jeffrey[1] the frequency components of a particular signal. For
paper titles as “Modeling and Simulation of Faults detecting precisely it can analyze a signal into high-
in Distribution Network System Using frequency components and low-frequency
MATLAB/Simulink” in this paper the author components. Sometimes Windowed Fourier
determined the underground fault by using a transforms also showing this type of advantage but
MATLAB SIMULATION and detect the it has a limitation. Because of using a fixed wide
underground fault but the main problem is the windowing function[12]. Wavelet transforms also
model used in this paper is so complex and costlier. show a property called multi-resolution in which
But in this paper provided a model that is simple and higher frequency wavelet narrow and lower
any operator can easily maintain. frequency broader, By using this property transients
fault can be analyzed which contains high-
frequency components in power signals. Because of
Kajal Bagade, Gita Bhardwaj ,Shweta Wasnik, having these advantages by using wavelet transform
Dheeraj Kawade, Payal Mool, Rupali Mohatkar[7] underground transient fault can be identified easily.
the paper titles as”IOT Based Underground Cable
Fault Detector” in this paper the author provide a There are two types of Wavelet transform (1)
Discrete wavelet transform (2) Continuous wavelet
prototype block diagram and detect the transform
Continuous wavelet transform define as a and 11 kV,a two winding transformer, a 250km long
transmission line, V-I measurement, and a three-
+∞ 1 𝑡−𝑏 phase series R-L-C load has been provided in Fig.
CWT (a,b; 𝑓(𝑡) , 𝛹(𝑡))= ∫−∞ 𝑓(𝑡) 𝑎
𝛹 ∗ ( 𝑎 ) 𝑑𝑡 1. The three-phase resistive-inductive-Capacitive
√
(RLC) the load is considered in this study. The
Where a is the scaling parameter,b is the time transmission line has been considered in two parts
parameter, 𝑓(𝑡) is the signal which is transformed first side having a length of 200km and the second
, 𝛹(𝑡)is the mother wavelet [12] side having 50 km. These two parts are designated
as TRL-1 and TRL-2. The fault under investigation
By using Continuous wavelet transform we can has occurred close to the second side of the
transmission line i.e. at a distance of 50 km from the
transform the inner outcome of a signal. By load. The voltages illustrated in the simulation
rewriting the CWT equation into frequency-based results are measured on the V-I measurement on the
signal we can get inverse Fourier transform.[12] load terminal (Bus B-2).

CWT (a,b; 𝑓(𝑡) , 𝛹(𝑡))=

1 +∞
∫−∞ 𝑓̂(𝜔)√𝑎𝛹
̂ (𝜔)(𝑎𝜔)𝑒 𝑗𝑎𝑏 𝑑𝛚 V. Proposed Cable Configaration
2𝜋

Where 𝑓(𝑡) , 𝛹(𝑡) is the signal and the wavelet for OVER SHEATH
Fourier transform. Curving the center frequency for INNER SHEATH
the wavelet shifts towards lower frequencies CWT
equation act as a bandpass filtering for the input INSULATOR
signal. This property can be used for feature CONDUCTOR
extracting from output signal analyze. CORE

Discrete Wavelet Transform (DWT), where the

scale and translation are discredited, but not are
independent variables of the original signal. D=Diameter

1 0 𝑛−𝑘2𝑚 1stD=11.3mm
DWT (m,n)= ∫ 𝑓(𝑘) 𝛹 (
√2𝑚 𝑘 2𝑚
) 2ndD=52.2mm
3rdD=98.6mm
𝑛−𝑘2𝑚
Where( ) is the mother wavelet[11]. 4thD=107.7mm
2𝑚
5thD=119.5mm

IV. Proposed Model

Figure 2: Cable Configaration

VI. Proposed Paramet

Table 1: Paramet

Object parameter

Three phase VL=11kV,f=50Hz,

Source Xs:Rs=10,Xs=2Ω,Rs=20Ω
Cables XLPE, Three-phase cabl

Distribution Length =100m, f=50Hz

parameter
cable1,2
Transformer: S=1 MVA VL=11kV,
Winding 1 Rp=1 Ω, Lp=28.6mH
Transformer: S=1 MVA VL =380V,
Winding 2 Rs=0.00044 Ω,
Figure 1: Proposed system for Detecting fault analysis Ls=0.0114mH
Three-phase VL=380 Vrms, f=50 Hz,
The proposed test system consisting of a three- static Load PL=92.62kW, QL=69.252
phase source which generates the power at 50 Hz kVAR
VII. Methodology b. Faullt classifing by current coefficient

fault classification and location identifies means For identification of fault, phase currents Ias, Ibs,
which type of fault occurs in the transmission line Ics are gained from the transmission end of the
and where it occurs. At this point for classifying the transmission line .Then mother signals are
occurring fault, we used a wavelet-based decomposed by DWT filter bank .The ground signal
multiresolution analysis which is basically a special Ig is calculated. Fault is identified on a individual
property of discrete wavelet transform. It is a transmission line if and only if the count of utmost
feature extracting technique in which the value of the detail coefficients of the transmission
multiresolution model is constructed by a filter line currents is higher then threshold limit. After the
which is called a quadrature mirror filter, which fault has been identified, it is easy whether it is
means it contains two filters a low pass filter and a occurred.
high pass filter, and both filters relater each other.
The resultant output signal from the HP filter is The ground current is calculated
known as detailed coefficients, and from the LP
filter are known as approximated coefficients. First, Ig=Ias+Ibs +Ics
the input signal passing through a low pass filter and
high puls filter with impulse response and
completes the first level. For classifying the
identified fault we need detail coefficients because Start
of this decomposition is repeated to further increase
the frequency resolution and the approximation
coefficients decomposed with high and low pass Get Value of Ias ,Ibs,Ics
filters and then down-sampled. This is represented
as a binary tree with nodes representing a sub-space
with a different time-frequency localization. The Calculate Ig
tree is known as a filter bank.

if
Ig>Th1 No Fault
Yes

No
h 2
if Yes
Ig>Th2 Ground Fault

h 2 g 2
No

Underground Fault
g 2 1st Level

Stop
Figure 3:Filter Bannk
Figure 4: Block diagram of fault classification
technique by current
a. Feature Selecting

Proper fault detection and classification technique

is depend on selecting the correct feature selection. c. Faullt classifing by voltage coefficient
After decomposing the mother signal by discrete
filter bank taking the required feature.In filter bank After identifying the grounded and ungrounded
the mother signal decomposed in 3rd level. In this faults. The other faults need to be classified. For
technique those proper of a signal are selected these reasons, the voltage of the first detailed
which have high frequency. coefficients are calculated in standard deviation
corresponding to the energy in each phase I.e. Va,
Vb, and Vc.
 Start a. Single Phase to Ground Fault
Detact Fault in the line

Calculate DWT and Coefficient standard

deviation and change energy

is
Va>Th1 Display fault
phase A

is
Va>Th2 Display fault
phase B

is Display fault
Va>Th3 phase C

Stop
Figure 6: All phase to ground fault with R-L-C load
Figure 5: Block diagram of fault classification
technique by voltage

After completing the fault classification, the next Table 2: Pic point current and distance
step is to determine the location of the fault. The
location of fault means where the fault occurs in the Type of Pic point
power system. Because of a fault occurring in the phase to current(A) Distance
system, a high-frequency voltage and current ground (km)
suddenly propagate along with the power. The fault
traveling wave theory method uses the information AG 7 16.8
got from the faulty place of voltage and current BG 6 13.2
traveling waves for locating the fault. The first peak
time obtained from the faulty bus is employed for CG 6 14.4
calculating the distance of fault from the sending
end. This technique is mostly used for extra high b. Double Phase to Ground Fault
voltage transmission lines than distribution lines
because the distribution system contains many
subsections such as laterals and feeders. It includes
two methods – the single-ended method and the
double-ended

VIII. Simulation Result and Discission

The simulation is carried out for 14-16 cycles. In

each case study, the fault is created in the 4th cycle.
The current is measured at the load terminal in each
case study. The rated voltage is 11 kV and the same
voltage level is considered at all points of the
system. The first peak time obtained from the faulty
bus is employed for calculating the distance of fault
from the sending end. Four different types of
transmission line faults such as a line to ground Figure 7: All Double phase to ground fault with R-
fault, double line fault, double line to ground fault, L Table
and three-phase fault have been investigated in the
presence of linear load (RLC load).
Table 3: Pic point current and distance Table 5: Pic point current and distance

Type of Pic point Type of three Pic point Distance

Double current(A) Distance phase to phase current(A) (km)
phase to (km) fault
ground fault ABC 13,12.5,11 26.456
ABG 10,4 24.8
ACG 10,4.5 24.2
BCG 8,7 21.4
IX. CONCLUSION
c. Double phase to phase fault Detection and location of faults are very important
in the power system as they are the basic of
protection of the power system. If detection and
location of faults are not accurate the protective
devices cannot trip properly and required operation
cannot be done, which may leads to damage of the
large equipments. In this paper at first, a new
method to analyze power distribution system
transient signals based MATLAB version 17a is
proposed by using WT technique. This method
offers important advantages over other methods
such as FFT and STFT due to good time and
frequency localization characteristics. Analysis
results presented clearly show that particular
Figure 8: all double phase to phase fault with R-L- wavelet components can be used as the features to
C load locate the fault in underground cables in distribution
syste
Table 4: Pic point current and distance
References
Type of Pic point
phase to current(A) Distance [1] Dauda A. Folarin et al., "Modeling and
phase fault (km) Simulation of Faults in Distribution Network
AB 8.3 19.8
System Using MATLAB/Simulink," IOSR Journal
AC 5.8 15.2 of Electrical and Electronics Engineering, vol.
BC 7 141 Volume 13, no. Issue 3, 2018.

[2] S. Vhora et al., "Underground Cable Fault

Detection Using
d. Three phase to phase fault with R-L-C
load Matlab," International Journal of Innovative
Research in Science,

Engineering and Technology, vol. Vol. 7, no. Issue

12, December 2018.

[3] H. Sam.E.J et al., "Underground Cable Fault

Detection Using IOT," Asian Journal of Applied
Science and Technology, vol. Volume 2, no. Issue
2, pp. Pages 737-746, 28 April 2018.

[4] N. Sampathraja, L. A. Kumar, V.

Kirubalakshmi, C. Muthumaniyarasi, and K. V.
Murthy, "IOT based underground cable fault
detector," International Journal of Mechanical
Engineering and Technology, vol. 8, no. 8, pp.
Figure 9: abc three phase to phase fault with R-L- 1299-1309, 2018.
C load
[5] M. R. Hans, S. C. Kor, and A. S. Patil,
"Identification of underground cable fault location
and development," in 2017 International
Conference on Data Management, Analytics and
Innovation (ICDMAI), 2017, pp. 5-8.

[6] M. Gilany, D. k. Ibrahim, and E. S. T. Eldin,

"Traveling-Wave-Based Fault-Location Scheme
for Multiend-Aged Underground Cable System,"
IEEE Transactions on Power Delivery, vol. 22, no.
1, pp. 82-89, 2007.

[7] K. Bagade et al., "IOT Based Underground

Cable Fault Detector," IJESC vol. Volume 8 no.
Issue No.3, 2018.

[8] Z. Long, N. H. Younan, and T. O. Bialek,

"Underground power cable fault detection using
complex wavelet analysis," in 2012 International
Conference on High Voltage Engineering and
Application, 2012: IEEE, pp. 59-62.

[9] A. Pandey and N. H. Younan,

"Underground cable fault detection and
identification via fourier analysis," in 2010
International Conference on High Voltage
Engineering and Application, 2010, pp. 618-621.

[10] S. T. Dhivya Dharani.A, "Development of a

Prototype Underground Cable Fault Detector,"
IJREE – International Journal for Research on
Extended, vol. 01, 2014.

[11] A. C and A. Ngaopitakkul, "Identification of

Fault Types for Underground Cable using Discrete
Wavelet Transform," Lecture Notes in Engineering
and Computer Science, vol. 2181, 03/01 2010.

[12] I. Journals, D. Prabhavathi, K. Prakasam,

and B. J. Reddy, "Detection and Location of Faults
in 11KV Underground Cable by using Continuous
Wavelet Transform (CWT)," 2015.