CHAPTER 1

INTRODUCTION

1.1 INTRODUCTION
Underground cables have been widely applied in power distribution networks
due to the benefits of underground connection, involving more secure than
overhead lines in bad weather, less liable to damage by storms or lightning, no
susceptible to trees, less expensive for shorter distance, environment-friendly and
low maintenance. However, the disadvantages of underground cables should also
be mentioned, including 8 to 15 times more expensive than equivalent overhead
lines, less power transfer capability, more liable to permanent damage following a
flash-over, and difficult to locate fault. Faults in underground cables can be
normally classified as two categories: incipient faults and permanent faults.
Usually, incipient faults in power cables are gradually resulted from the aging
process, where the localized deterioration in insulations exists. Electrical
overstress in conjunction with mechanical deficiency, unfavorable environmental
condition and chemical pollution, can cause the irreparable and irreversible
damages in insulations. Eventually, incipient faults would fail into permanent
faults sooner or later. The detection of incipient faults can provide an early
warning for the breakdown of the defective cable, even trip the suspected feeder to
limit the repetitive voltage transients. The location of permanent faults in cables is
essential for electric power distribution networks to improve network reliability,
ensure customer power quality, speed up restoration process, minimize outage
time, reduce repairing cost, dispatch crews more efficiently and maintain network
reliability. The state estimation (SE) is an auxiliary tool to provide the necessary
information for the proposed location algorithms. The related methods published
in journals and proceedings are reviewed, summarized and compared in the next
subsections.

1
1.2 INCIPIENT FAULT DETECTION METHODS
Comparing with the detection methods for arcing faults in overhead lines,
there are relatively fewer literatures and reports discussing the detection of
incipient faults in underground cables. The existing detection methods are
generally based on the analysis of waveforms rather than phasor. Basically, the
process of detection is to examine the characteristics of voltages and currents in
time domain, frequency domain and time frequency domain.
The advantages and disadvantages of four existing techniques developed for
field applications were reviewed and evaluated from the point of a power engineer.
These techniques include detection of partial discharges, time and frequency
domain reflectometry, measurements of dielectric ohmic and polarization, and
acoustic and pressure wave techniques the feasibility of detecting arcing faults in
underground cables. Experiment was carried out in one secondary distribution
network by personnel from the Consolidated Edison Company of New York.
Through analyzing the collected data, three feasible methods are considered, i.e.
analysis of voltages and currents in time domain, in frequency domain and in time-
frequency domain with the aid of the wavelet analysis. Furthermore, it is pointed
out that the potential approaches can process the instantaneous values of currents,
and combine the arc fault features in time, frequency and time-frequency domains
cable splice failure detection scheme, which is integrated into a universal relay
platform as an additional function to enhance the distribution feeder protection.
The basic principle is to monitor instantaneous overcurrent, counter the number of
fault occurrences, record the frequency of fault occurrences, and provide alarming
or tripping capability. a simple, fast and robust method for detecting incipient
faults in cables and implemented it in a commercial relay. The method employs
the superimposed current components and neutral current to monitor the
consistency of currents before and after the event, find the phase where the event
occurs, check the event duration, and set the alarming or tripping signal. Both the

2
magnitude of neutral current and the magnitude of rate of change of neutral current
were used to detect self-clearing cable transient faults and distinguish them from
normal system switching as well as other system faults, such as fast fuse
operations The faulty phase is selected by a phase current rate of change based
detector. The wavelet analysis and neural network were combined to detect on-line
incipient transients in underground distribution cable laterals and predict the
remaining life of the cable lateral. The wavelet packet analysis technique is
applied to decompose the current into separate frequency bands and to extract
features. Then, a type of artificial neural network, self-organizing map, is used for
pattern identification. Therefore, the data sets are clustered and incipient behavior
is identified and categorized. The pattern analysis techniques were applied to
classify load change transients and incipient abnormalities in an underground
distribution cable lateral. A set of features are exacted by the wavelet packet
analysis and output to k-nearest-neighbor classifiers. The methods of
dimensionality reduction are used to reduce the dimensionality for the pattern
recognition and preserve the good classification accuracy as well.
1.3 FAULT LOCATION METHODS FOR CABLES
Basically, the location methods for cables are divided into the offline and
online methods. The offline methods employ the special instruments to test the out
of service cable in field. On the other hand, the online methods utilize and process
the sampled voltages and currents to determine the fault point.
1.3.1 OFFLINE METHODS
There are two offline location approaches, i.e. terminal approaches and
tracer approaches. Terminal methods rely on measurements made from either one
or both terminals of the cable to percolate fault points approximately, but not
accurately enough to allow dig. Tracer methods rely on measurements taken along
the cable to pinpoint the fault location very accurately. Both methods are on-site
technique and performed with low efficiency. Eighteen terminal methods are
introduced in and listed below.

3
 • Halfway Approach Method
• Voltage Drop Ratio Method
• Charging Current Method
• Insulation Resistance Ratio Method
• Murray Loop Method
• Capacitance Ratio Method
• Murray Loop Two-End Method
• Murray-Fisher Loop Method
• Varley Loop Method
• Hilborn Loop Method
• Open-and-closed Loop Method
• Werren Overlap Method
• Impulse Current Method
• Pulse Decay Method
• Standing Wave Differential
• DC Charging Current Method
• Time Domain Reflectrometer (TDR)/Cable Radar Method
• TDR/Cable Radar and Thumper Method
Following the prelocation by the terminal methods, a tracer method is generally
applied to pinpoint the fault point and this method usually requests the repair
crews to walk along the cable route. Nine tracer methods introduced in are listed
below.
• Magnetic Pickup Method
• Tracing current Method
• Earth gradient Method
• Hill-of-Potential Method
• Thumper/Acoustic Method
• Thumper/Electromagnetic Wave Method
• Sheath Coil Method

4
 • Pick Method
• DC Sheath Potential Difference
Method some extension works were proposed as an aid for the offline methods.
An expert system was developed for the Electric Power Research Institute
(EPRI) .The system creates a reference manual to provide the guidance for field
crews to diagnose a cable failure, recommend applicable fault location techniques,
and trouble-shoot resulting difficulties which occur during the process of locating
underground cable faults. For the sake of clarifying the results obtained from the
terminal methods, an expert system approach was proposed to locate fault on high
voltage underground cable systems. The experience and expertise of many
different engineers is accumulated to build a truly expert system. With the data
acquired from diagnostic tests, the system can infer the fault type, advise the
further location techniques, and conclude the probable fault location. The operator
is then advised to carry out the tracer methods to locate the fault precisely.
1.3.2 ONLINE METHODS
The online location methods for underground cables are comparatively
fewer than the ones applied for overhead lines. Two principal techniques have
been proposed for the online location, i.e. signal analysis and knowledge-based.
The former one is further classified into the approaches based on fundamental
frequency phasor quantities and high frequency traveling waves.
1.3.2.1 FUNDAMENTAL PHASOR-BASED METHODS
The fundamental phasor-based methods utilize the voltage and current
phasor at the fundamental frequency. Basically, the impedance is calculated and
used to decide the fault distance, so it is also called the impedance-based method
extended the traditional impedance-based location algorithms to calculate the
apparent impedance of cables in cases of single phase to ground fault (SLG) and
three-phase fault .The single-end voltages and currents are used. An iterative
algorithm is proposed to compensate the capacitive characteristic in typical
underground cables. The fault location scheme can be applied in balanced or

5
unbalanced distribution systems with laterals and tapped loads. Based on the
estimation of the fault-loop impedance, SLG fault with measurements available in
the faulty feeder (voltages and currents), SLG fault with measurements available
at the substation level (total currents are measured at the supplying transformer),
phase to phase (LL) or 3L fault with measurements available in the faulty feeder,
LL or 3L fault with measurements available at the substation. Only positive
sequence impedance calculation is needed for LL or 3L fault, while the zero-
sequence impedance calculation is required for SLG fault. The algorithms can be
applied in radial medium voltage (MV) systems, which include many intermediate
load taps. The non-homogeneity of the feeder sections is also taken into account.
The apparent seen impedance was calculated using local measuring quantities
available at substation. Upon the different fault type, the different apparent
impedance parameters, voltage and current quantities are utilized. Then, a fault
distance is estimated using the conventional apparent impedance computation.
Finally, an iterative compensation mechanism is executed to eliminate the
estimation errors caused by the charging currents in cables. The basic procedure is
similar to the work in except that the symmetrical components are used. The
location algorithm in extended the traditional Takagi’s method into distribution
cable networks. The sequence phase impedance model is used to model laterals
and circuit sections. The line shunt capacitance is taken into account to optimize
the result so that the major source of error in conventional impedance based
methods, particularly for cable networks, is minimized. Differentiating from the
above extended impedance-based methods, an iterative algorithm was proposed
for locating faults in cables. The circuit is modeled by the distributed parameter
approach and the voltage and current equations are formulated based on the
sequence networks. The Newton–Rap son method is applied to calculate the fault
distance. The algorithm is also extended to the radial multi-section cables with
tapped loads. A double-end based location algorithm was presented, particularly
for aged power cables. The aging process in cables would cause the change of the

6
relative permittivity and in turn result in the changes in the positive, negative, and
zero sequence capacitance. The fault location scheme is based on phasor
measurements from both ends of the cable, incorporating with the distributed line
model, Clarke transformation theory and discrete Fourier transform (DFT). One
algorithm implemented in the Con Edison of New York. The voltages and currents
are recorded by the power quality monitors and processed for calculations in the
control center. The reactance to fault is calculated based on the fault measurements
and prior knowledge of known fault information. The calculation results combined
with up-to-date distribution feeder models and geographic information system data
are used to generate the estimated fault location tables and viewing maps. The
estimation would typically take ten minutes after the inception of a fault. The
location accuracy is within 10% of the total number of feeder structures, for about
80% of the single phase faults. One more implementation in the Dutch grid
operator. The fault locators only use the calculated reactance since the reactance of
fault impedance is zero and the cable reactance is well known and not current
dependent. Then, the scenarios of short circuits on all nodes in the faulted feeder
are simulated on an actual network model. The calculated impedance is compared
with the simulated impedances to find the exact location. The location algorithm is
known to find the distance within 5 minutes after the occurrence of a fault. The
system is able to locate LL and 3L faults within 100 meters and SLG faults within
500 meters.
1.3.2.2 TRAVELING WAVE-BASED METHODS
Traveling waves are generated by the change of stored energy in
capacitance and inductance in lines or cables after the occurrence of a fault. Both
voltage and current traveling waves propagate along the circuit at the speed as
high as the light speed until meeting any impedance discontinuities, and then the
fault-induced high frequency waves would reflect back to the origin and transmit
through towards other side. Almost all traveling waves-based methods are based
on the principle of the Bewley lattice diagram, and the fault distance is calculated

7
by the multiplication of the propagation velocity and the interval, which is the time
difference between the arrival instant of the initial wave front and the arrival
instant of the reflected wave front. The basic location principle and common
locator visually illustrates and explains an example of the traveling wave, which is
generated by an SLG fault in a transmission line and can be used for the purpose
of the line protection and fault location. Bo et al. designed a special transient
capturing unit to extract the fault-generated high frequency voltage transient
signals in cables .The principle of the fault location method is to identify the
successive arrivals of the traveling high frequency voltage signals arriving at the
bus bar where the locator is installed. Particularly the first and the 8 subsequent
arriving wave fronts with reference to the first wave front are used to locate the
fault position. The above work is enhanced by applying new technique, wavelet
transform, to effectively extract a band of high frequency transient voltage signals.
A cable fault location scheme was proposed based on the principle of the
traditional traveling wave principle, synchronized sampling technique and wavelet
analysis. The current signals at the two terminals are synchronized with the help of
GPS and the arrival time of fault-induced traveling waves is precisely detected by
the wavelet analysis. Then, the location is obtained from the multiplication of the
propagation velocity and the time interval. Similarly, based on the principle of the
traditional Bewley lattice, a double-end traveling wave fault location scheme was
proposed for locating faults in aged cables. The wavelet analysis is applied to
analyze the synchronized voltage singles at the two terminals to capture the
singularity in high frequency transients. The calculations are processed with the
modal quantities rather than the phase quantities. The effect of changes in the
propagation velocity of traveling wave is eliminated. The fault section and
location was determined by the analysis of traveling waves in current signals.
First, the fault section is identified by the comparison between the distance of each
peak in the high frequency current signals and the known reflection points in
distribution feeders. Then, the simulation is processed with the possible location in

8
a transient power system simulator, which is modeled from the actual network.
The simulated currents are cross correlated with the measured currents to find the
match degree in high frequency transients of both current signals. The cross-
correlation coefficients would be a high positive value if the estimated fault
location is correct.
1.3.2.3 KNOWLEDGE-BASED METHODS
Knowledge-based techniques, such neural network, fuzzy logic and expert
system, are applied to fault location for cables. The usage of artificial intelligence
techniques usually requires the specific learning process for each analyzed feeder.
Additionally, the signal processing techniques can also be used to preprocess the
signals and extract the features fed into the analysis of artificial intelligence. Fault
location algorithm for combined overhead transmission line with underground
power cable. First, one adaptive network-based fuzzy inference system (ANFIS) is
used to classify the fault type. Then, another ANFIS is applied to detect the faulty
section, whether the fault is on the overhead line or on the underground cable.
Other eight ANFIS networks are utilized to pinpoint the fault, in which two
networks are used for one fault type. The neuro-fuzzy inference systems are
trained by the data obtained from simulations. a location algorithm combined the
neural network and wavelet analysis. The power distribution system transient
signals are generated by the EMTP software, analyzed using the wavelet analysis
to extract the useful fault features, and applied to the artificial neural networks
(ANNs) for locating ungrounded shunt faults. A three-layer feed-forward ANN
with Levenberg-Marquardt learning algorithm is used for the fault classification
and fault location. One network is designed to classify the fault type and several
ANNs related to each fault type are designed to locate the actual ungrounded fault
position.
1.4 FAULT LOCATION METHODS FOR DISTRIBUTION NETWORKS
The fault location techniques have been well developed and applied in
transmission systems. However, relatively less research work has been conducted

9
in the development of fault location approaches for distribution networks. An
effective and accurate fault location algorithm is essential for electric power
distribution networks to locate the fault point, improve the service reliability,
ensure the customer power quality, and speed up the restoration process.
Particularly, it appears more important for locating faults in underground
distribution cables due to the complexities in electrical characteristics of cables,
underground placement environment and wide applications in high density
commercial districts. Similarly, two principal techniques have been proposed for
such methods, i.e., signal analysis and knowledge-based. The former one is further
classified into the approaches based on fundamental frequency phasor quantities
and high frequency traveling waves. The knowledge-based and traveling wave-
based techniques have been briefly
The utility companies and researchers have been turning more and more
attention to the location methods only using voltages and currents recorded at
substation. The fundamental phasor-based methods utilize and process the
recorded voltages and currents to determine the fault point. Since the proposed
algorithm is to use the fundamental phasor, the existing fundamental phasor-based
methods would be discussed in this subsection. The basic location methods, such
as the reactance method and Takagi method. Ten most cited impedance-based
fault location methods are compared, analyzed and tested, and thereafter the main
problems existing in these methods are concluded. The practical experience and
the fault location systems used in utilities .Most of the previously proposed
location techniques concern the location problem in overhead distribution lines,
and a few of literatures discuss the algorithms for underground distribution cables.
The specifications of the proposed algorithm are listed as well.
1.5 DISTRIBUTION STATE ESTIMATION METHODS
The state estimation for distribution networks is an important application in the
distribution management system (DMS) to provide the essential information for
operation, management, control and planning in distribution networks. It also

10
assists in the fault location algorithms by providing the necessary information of
load flows and bus states (voltage magnitudes and phases). The present
distribution state estimation (DSE) methods are reviewed below since a DSE
algorithm is proposed for underground distribution networks in this work. Usually,
the weighted least squares (WLS) technique is employed. Wan et al. proposed two
WLS approaches to estimate loads in unbalanced power distribution networks.
One is the WLS load parameter method to solve the constrained optimization
problem where loads are treated as variables. The constrained optimization
problem is transformed into an unconstrained problem by the exterior penalty
method. The loads and voltages are estimated simultaneously. Incorporating the
operating and loading constraints, the other one is a constrained WLS distribution
state estimation-based method to estimate voltages by a constrained WLS DSE,
then to estimate loads sequentially based on the estimated voltages. Three-phase
state estimation method based on the WLS method in. A two-stage algorithm is
developed to overcome the observability problems associated with the branch
current magnitude measurements. Rather than using nodal voltages as estimation
variables, the branch currents are used as state variables in the state estimation to
solve the WLS problem, where the Jacobian matrix is well 25 conditioned and can
be decoupled on a phase basis. This method was substantially where a new
algorithm with the constant gain matrix and a decoupled form was developed. The
problem of load estimation was formulated as a weighted least absolute values
estimation problem and solved by WLS. The Newton-Rap son approach is applied
to eliminate the nonlinear effect of power losses. In addition to WLS methods, the
modified conventional algorithms were also proposed. Extending the work in, the
custom-tailored Gauss-Seidel load flow analysis was proposed in. A
computationally efficient solution scheme based on the NewtonRaphson method
was proposed in an algorithm was developed to build a constant Jacobean matrix
and the Newton-Rap son algorithm was also used to solve the load flow problem.
The load flow problem of a radial distribution system was formulated as a convex

11
optimization problem, particularly a conic quadratic program. The solution of the
distribution load flow problem can be obtained in polynomial time using interior-
point methods.
 Design of the incipient fault detection scheme for distribution cables;
Development of the fault location scheme for a medium voltage cable
with no laterals;
 Design of the state estimation algorithm for underground distribution
networks;
 Extension of the proposed location algorithm to underground
distribution networks with the aid of the proposed state estimation
algorithm.
1.6 CONTRIBUTIONS OF THE THESIS
The contributions of the thesis are summarized as follows:
 A wavelet analysis-based method is developed to detect incipient faults
in cables in time and frequency domains, additionally, identify transient
and fault types, remove effect of noise and supervise almost entire cable
circuit.
 A simple and practical algorithm based on the analysis of superimposed
components and negative sequence is particularly designed to detect
single-line-to-ground incipient faults in cables. The fewer thresholds
and less computation are required.
 A two-layer π circuit is formulated and examined to approximate the
behavior and characteristic of a typical medium voltage cable.
 A set of fault location algorithms are proposed for underground cables.
The characteristics of underground cables in real systems are
comprehensively considered and analyzed in the development of algorithms,
such as the shunt capacitance, metallic sheath, heterogeneity and
untranspositon. The cable configurations and fault scenarios are taken into
account as well, such as five bonding methods and three fault pathways.

12
 Besides, a large number of fault equations are solved effectively and efficiently
and the fault resistance can be calculated. The state estimation for underground
distribution networks is formulated as a nonlinear optimization problem and
solved by the sequential quadratic programming technique.
The characteristics and configurations of underground cables and
distribution networks are considered in the development of the algorithm, such
as the shunt capacitance, metallic sheath, bonding method, unbalance loads
and presence of laterals and tapped loads.
 A section-by-section estimation algorithm combined with the
backward/forward sweep algorithm is presented to estimate the nodal
voltage and branch current for each circuit section in a distribution
network with laterals and tapped loads.
 The combination of the fault location and state estimation algorithms is
proposed to solve the fault location problem in distribution cables.
 The faulty section in distribution networks can be determined and the
problem of multiple estimations is eliminated.
 Only the fundamental voltage and current pastors recorded at the single-
end are utilized in the proposed fault location and state estimation
methods.
 The performance and functionalities of the all proposed algorithms are
examined and verified with the extensive simulations, considering
various fault conditions and system configurations.
1.7 INCIPIENT FAULT DETECTION SCHEMES FOR DISTRIBUTION
CABLES
The incipient faults in underground cables are largely caused by voids in cable
insulations or defects in splices or other accessories. This type of fault would
repeatedly occur and subsequently develop to a permanent fault sooner or later
after its first occurrence. Two algorithms are presented to detect and classify the
incipient faults in underground cables at the distribution voltage levels. Based on

13
the methodology of wavelet analysis, one algorithm is to detect the fault-induced
transients, and therefore identify the incipient faults. Based on the analysis of the
superimposed fault current and negative sequence current in the time domain, the
other algorithm is particularly suitable to detect the single-line-to-ground incipient
faults, which are mostly occurring in underground cables. Both methods are
designed to be applied in real systems. Hence, to verify the effectiveness and
functionalities of the proposed schemes, different fault conditions, various system
configurations and real field cases are examined, and other normal operating
transients caused by permanent fault, capacitor switching, load changing, etc., are
studied as well. The basic concept of incipient faults in cables is first introduced
and the model of arc is formulated. Then the wavelet-based scheme is explained
and the system structure, time sequence diagram, detection rules and classification
rules are also discussed. Subsequently, the details of the superimposed
components-based scheme are presented. Two test distribution systems, extensive
simulation cases, field cases, and simulation results are examined, where, more
specially, the detailed detection process is explicated by analyzing four incipient
faults recorded from real systems.
1.8 INCIPIENT FAULTS IN CABLES UNDERGROUND CABLES
Incipient faults for an unpredicted duration before they fail into permanent
faults. Usually, incipient faults in power cables are 30 gradually resulted from the
aging process, where the localized deterioration in insulations exists. The local
defect or void initiates a process such that the insulation damage spot can
propagate through a section of the insulation, branch into channels, and evolve to a
tree-shape damage area. Two trees are mostly observed, i.e. water tree (WT) and
electrical tree (ET). The water tree in insulation can initiate from a water-filled
micro cavity and would be growing under the influence of moisture and electric
field. The voltage drop on a water tree is quite small compared to the voltage
across the dry insulation surrounding it since the insulation at the water tree area
has a higher conductivity. The progress of water trees is permanent and there is no

14
detectable partial discharge existing in water trees. The electrical tree can initiate
from a point of high stress due to a local defect and/or water tree in dry dielectrics
and propagate relatively quickly through the insulation due to the repetitive partial
discharges. The formation of electrical trees would lead to final cable failure
sooner or later within a relatively short time.

1.8.1APPARENT IMPEDANCE-BASED
It is well known that the apparent impedance can be calculated by the
voltages and Currents of the faulty phase and/or zero sequence current. For
example, the apparent Impedance for an SLG fault in phase A can be expressed as,

(1.1)
Where Zapp is the apparent impedance, Va is the phase A voltage, Ia is the
phase A current, k is the compensating factor, I0 is the zero sequence current, Z0
and Z1 are the zero and Positive impedances of the line.
The KVL equation for Va can be given as,

(1.2)
Where Rf is the fault resistance, Icomp is the compensating current flowing
through the fault Resistance, which can be described as below for an SLG fault,

(1.3)
Therefore,

(1.4)

15
 Where D is the fault distance, z1 is the positive impedance per unit length.
There are two unknown real variables in Equation (1.4), i.e. D and Rf, and other
variables Can be measured at the substation or obtained from the database. The
equation can be re written in terms of real and imaginary components so that the
unknown variables can be solved.
The apparent impedance for other faults can be calculated accordingly.
Basically, the apparent impedance-based technique is used.
The impedance measurement principle is also used, and the real value
nature of the fault resistance is employed to find the fault distance.
1.8.2 DIRECT CIRCUIT KVL EQUATIONS-BASED
Taking an SLG in phase A as an example, the KVL equation describing the
circuit
Between the sending terminal and the fault point can be given as,

(1.5)
Where Va,b,c is the three-phase voltages, Ia,b,c is the three-phase currents,
If is the fault current, Rf is the fault resistance, D is the fault distance, zaa is the
self-impedance of phase A, zab is the mutual impedance between phase A and B,
and so on.
The KVL equation for phase A can be expressed as,

(1.6)
Where Va is the phase A voltage, Ia, Ib, and Ic are the currents of phases A,
B, and C. In the fault current or load current is first assumed or estimated, thus the
fault distance D and the fault resistance Rf can be solved by two real equations ,
which are generated by separating Equation (1.6) into the real part and the
16
imaginary part. Then, an iterative process is carried out to update the fault current
or load current until a small tolerance is satisfied.
In starting with a set of assumed fault distances and using the superimposed
components, the KVL equations describing the circuit between the fault point and
the receiving terminal are also formulated to determine the fault distance on the
condition that there exists a minimal value of the difference between healthy phase
currents around the exact fault point. The equations are simplified on the
assumption that the fault current is equal to the phase current.
1.8.3 FAULT RESISTANCE-BASED
The fault resistance is a non-negative real number, which can be used as a
criterion to find the fault distance. Taking an SLG in phase A as an example, the
imaginary part of fault resistance is given as,

(1.7)
Where Rf is the fault resistance, Vf is the fault voltage, If is the fault
current, Vfp, Vfn, and Vfz are the positive, negative and zero sequence voltages at
the fault point, Ifp, Ifn, and Ifz are the positive, negative and zero sequence fault
currents. Basically, an initial variable, for example, the fault distance, is first
guessed or estimated, Vf and If can be estimated by the application of some skills,
an iterative procedure is used to calculate the mismatch between the new estimated
variable and the old one, then the assumed variable is adjusted until a small
tolerance is satisfied.
1.9 INCIPIENT FAULTS IN CABLES
Underground cables may first experience incipient faults for an unpredicted
duration before they fail into permanent faults. Usually, incipient faults in power
cables are gradually resulted from the aging process, where the localized
deterioration in insulations exists. The local defect or void initiates a process such

17
that the insulation damage spot can propagate through a section of the insulation,
branch into channels, and evolve to a tree-shape damage area. Two trees are
mostly observed, i.e. water tree (WT) and electrical tree (ET). The water tree in
insulation can initiate from a water-filled micro cavity and would be growing
under the influence of moisture and electric field.
The voltage drop on a water tree is quite small compared to the voltage
across the dry insulation surrounding it since the insulation at the water tree area
has a higher conductivity. The progress of water trees is permanent and there is no
detectable partial discharge existing in water trees.
The electrical tree can initiate from a point of high stress due to a local
defect and/or water tree in dry dielectrics and propagate relatively quickly through
the insulation due to the repetitive partial discharges. The formation of electrical
trees would lead to final cable failure sooner or later within a relatively short time.
The example of water tree and electrical tree are shown in Figure.

Figure 1.1Illustrations of water tree (WT) and electrical tree (ET)

Overall, electrical overstress in conjunction with mechanical deficiency,
unfavorable environmental condition, and chemical pollution, can cause the
irreparable and irreversible damages in insulations. The formation of electrical
trees would generate partial discharges, which can be considered as the early stage
of incipient faults before the condition of insulation gets worse. The partial
discharge is characterized by a series of short discharge current pulses with the
width of about one nanosecond and with the time interval of several tens of
nanoseconds between successive discharges.

18
 Therefore, the detection of early cable defects or failures can be classified
into two categories: detection of partial discharges and detection of incipient
faults. Both of them are concerned by the utility companies, and the power
protection engineers would pay more attention on the latter one. The proposed
method is also directly associated with the latter one. Incipient faults are normally
characterized as the faulty phenomena with the relatively low fault currents and
the relatively short duration ranging from one-quarter cycle to multi-cycle.
These short lasting current variations cannot be detected by the traditional
distribution protection schemes because of their short duration and low increment
in magnitude. However, such faults must be detected at the early stage to avoid the
consequent catastrophe induced by the degradation themselves. In underground
cables, the incipient fault is one type of transients in power systems, which is
prone to an intermittent arc fault. The typical incipient faults are composed of two
types: sub-cycle incipient fault and multi-cycle incipient fault. The sub-cycle
incipient fault always occurs near a voltage peak where arc is ignited, lasts around
one quarter cycle, and self-clears when the current crosses zero.

Figure 1.2 Sub-cycle incipient fault

19
 Figure1.3 Multi-cycle incipient fault

Figure 1.2 shows the three phase feeder currents when a sub-cycle incipient
fault occurs between phase A and ground at the 2 km location of a 9 km cable in
the first test system.
1.10 TEST SYSTEMS
Two distribution systems are selected for simulations. The first one is
modified from a 110/10.5 kV distribution network, containing five underground
cables, two overhead lines, and one combination of line and cable as shown in
Figure 1.4.

Figure 1.4 Test system 1

The important system data include

20
 The second test system is simplified from an IEEE 13-node test feeder,
including two underground cables and eight overhead lines shown in Figure 1.5.

Figure 1.5 Test System 2

1.11 STRUCTURE OF A TYPICAL XLPE CABLE
The typical structure of a widely used single-conductor cable is shown in
Figure 3.1 and the each part numbered in the figure is explained as below:
 Aluminum or copper stranded conductor.
 Semi-conducting conductor screen extruded around conductor.
 Insulation, XLPE are used in most modern MV and HV cables.
 Semi-conducting insulation screen. The semi-conducting swelling tapes
wrapped around the insulation screen are considered as part of the
insulation screen since the electrical properties of this layer are similar to
those of the insulation screen.
 Copper wire sheath.
 Outer jacket, usually polyethylene (PE).

21
Figure 1.6 Structure of a typical single-conductor XLPE cable and laid formations
of three-phase cables
The cables could be directly buried or installed in an underground duct laid
in the flat or trefoil formation. The choice depends on several factors like sheath
bonding method, conductor area and available space for installation.
1.12 ADVANTAGES
 Low voltage drop,
 Low maintenance,
 Low chances of fault occurring and is very suitable in urban areas where
overhead transmission lines are not easy to install.
1.13 APPLICATIONS
 Used in industrial application, sub station
1.14 OBJECTIVE
 To Find out Underground lines fault detection
 To determine the fault occur in the underground cable and to detect the
exact location of the fault from the power station.
1.15 THESIS ORGANIZATION
 Chapter 2: Literature Review
 Chapter 3: Existing System
 Chapter 4: Proposed System
 Chapter 5: Result and discussion

22
 Chapter 6: Conclusion
 Chapter 7: Reference

23
 CHAPTER-2
LITERATURE REVIEW
Massimo Marzinotto et al [2015] provides at checking the possibility that
cable sheath faults in unipolar three-phase HV and extra HV cables with sheath
cross-bonding systems can be detected by monitoring sheath-to-ground currents at
the solid bonding ends of major cross-bonding sections. This goal is achieved by
calculating such currents for some reference case studies of faults of the outer
thermoplastic jacket in unipolar three-phase land cables for power transmission in
flat formation. The case studies differ as to cable voltage ratings, resistance and
location of the fault, symmetry of minor sections, and number of major sections.
Giuseppe Parise et al [2019] proposed in electrical power systems, cords
and extension cords are exposed to mechanical damage and other insulation
stresses. Mechanical damage to stranded conductors can reduce locally their cross
section or break them and cause anomalous local conditions of overheating or
arcing. The ordinary protective devices cannot detect the series faults that persist,
so the fault point remains energized and is subject to electric shock and fire
hazards. Effective protection can be accomplished by implementing active and
passive measures: installing arc-fault circuit interrupters or arc-fault detection
devices, able to detect arcing faults, or wiring the circuits with a grounding
protection conductor to involve the ground in every fault. In this way, residual
current protective devices (residual current devices or ground fault protective
devices) quickly protect the series faults not only with arc, but also without it.
Ground-fault-forced cables facilitate by design the conversion of any kind of cable
fault to a ground fault.
Mashad Uddin Saleh et al [2020] we evaluate the detectability and
localizability from both existing literature and our own measurements using
SSTDR of open-circuit faults, connection faults, short-circuit faults, ground faults,
arc faults, shading faults, bypass diode faults, and accelerated degradation faults in
PV cells and mini-modules. Reflection magnitudes for these faults are compared.

24
Preliminary data on buried and grounded PV cable along with arc fault detection
are presented.
Reddipalli Bhargav et al [2020] accurate estimation of fault location is
highly crucial for swift maintenance and early power restoration. Since faults in dc
networks are time critical, this article proposes a new fault detection and
localization scheme for a low-voltage direct current (LVdc) Microgrid network.
Low-resistance faults are detected by observing voltage across the inductor
whereas high-resistance ground faults are identified by measuring ground current
at the relay location. Thereafter, based on iterative method, fault location is
estimated by comparing analytically derived fault current with the measured value
of fault current. Genuineness of the suggested technique has been assessed by
simulating various internal, external, and simultaneous faults on a typical LVdc
Microgrid network modeled in power system computer aided
design/electromagnetic transients including dc (PSCAD/EMTDC) environment.
The proposed method is capable of detecting and locating both low- and high-
resistance dc faults without utilizing remote-end quantities. Subsequently, initial
guess has minimal impact on the convergence and accuracy of the proposed
algorithm. Comparative evaluation of the proposed technique with other
techniques clearly proves its superiority in terms of better discrimination against
external faults, rapid detection during internal faults, independency on the network
topology, and higher accuracy for fault distance estimation against all types of
internal faults.
Zhinong Wei et al [2020] address this issue by applying generalized S-
transform (GST) with a variable factor to conduct denoising of transient zero-
sequence currents based on threshold filtering followed by time-frequency
distribution filtering in sequence. Meanwhile, we proposes a comprehensive multi-
criteria faulty feeder detection method based on the transient zero-sequence
current polarity (criterion 1), the energy relative entropy (criterion 2), and the total
transient current energy (criterion 3). Here, criteria 2 and 3 are based on the time-

25
frequency representation of the GST. The performances of the proposed denoising
and faulty feeder detection methods are evaluated under single line to ground
faults based on simulations conducted using a modeled 10 kV RG networks with
overhead and cable mixed lines in addition to reasonably sophisticated permanent
and intermittent arc discharge models to ensure that the simulations faithfully
represent actual complex working conditions.
Laurens Mackay et al [2018] proposed dc distribution grids, residual
ground current protection is more feasible than in ac distribution grids, where stray
capacitance leads to significant ac ground currents. However, when meshing dc
distribution grids, circulating net currents can appear that resemble residual ground
currents and would trip traditional protection schemes. This letter introduces this
phenomenon in both unipolar and bipolar dc distribution grids. Furthermore, worst
case net current magnitudes are derived and examples are shown. Implications for
measurement and protection devices are discussed.
Catherine E. Jones et al [2018] proposed to trend for the use of electrical
power on state-of-the-art aircraft is resulting in significant change to the design of
power system architectures and protection systems for these platforms. There is a
pull from the aerospace industry to integrate the electrical power system with the
aircraft's structural materials to form an embedded system, reducing the need for
bulky cable harnesses. This directly impacts the fault response for ground faults
and ultimately the development of appropriate protection systems. Such structural
materials include composites such as carbon fiber reinforced polymer (CFRP). We
presents the experimental capture and analysis of the response of CFRP to
electrical fault current, which indicates the need for two distinct sets of electrical
ground fault detection criteria for low and high resistance faults and identifies the
threshold resistance for this distinction.
Snehamoy Dhar et al [2018] proposed a multi terminal dc (MTDC)
distribution network is an effective solution for present grid scenario, where local
distribution is incorporated primarily by power electronics based dc loads. PV

26
systems with auxiliary power sources and local loads are used for MTDC
connection, especially when ac utility grid is integrated with it by voltage source
converters. Pole to pole and pole to ground faults are basically considered as dc
distribution network hazards. As PV is connected through dc cable, high resistive
dc arc fault is also studied in present literature. The proposed PV system is
considered with arc-fault circuit interrupters as backup protection and is used to
detect arcing series fault. Fast acting dc switching is considered for proposed
differential current-based unit protection. A discrete frame differential current
solution is considered to classify the fault type by modified cumulative sum
average approach.
Enrico Pons et al [2017] the electrical safety of dc urban traction systems is
analyzed, with particular focus on fault current detection and on dangerous
voltages which could arise in case of fault. For the discussion, the tram network of
Turin, Italy, is used as a case study. First, the structure of the dc traction power
supply is described, analyzing in detail the different components; then, the safety
of the system is analyzed, examining possible types of fault. In particular, ground
faults inside the substation and ground faults along the line are analyzed in detail.
Fault currents and dangerous voltages are calculated, thanks to a simplified steady-
state circuital model of the traction system.
Indra Man Karmacharya et al [2018] proposes a novel fault-location
scheme in which high frequency noise patterns are used to identify the fault
location. The high-frequency noise is generated due to the switching transients of
converters combined with the parasitic capacitance of PV panels and cables.
Discrete wavelet transform is used for the decomposition of the monitored signal
(midpoint voltage of the converters) and features are extracted. Norm values of the
measured waveform at different frequency bands give unique features at different
fault locations and are used as the feature vectors for pattern recognition. Then, a
three-layer feed forward artificial neural networks classifier, which can
automatically classify the fault locations according to the extracted features, is

27
investigated. The proposed fault-location scheme has been primarily developed for
fault location in the PV farm (PV panels and dc cables). The method is tested for
ground faults as well as line-line faults. These faults are simulated with a real-time
digital simulator and the data are then analyzed with wavelets.
Sheng Liang et al [2019] we propose a novel fiber-optic auditory nerve of
ground (FANG) in the suburb based on the fiber-optic distributed vibration sensor
(DVS). The feasibility and effectiveness of the principle prototype FANG for
traffic flow monitoring are proved and investigated by the field experiment. One
of the 31.8 km-long redundant optical fiber of the buried optical-fiber cable for
data transmission is utilized as the sensing fiber. Then, the phase-sensitive optical
time-domain reflecto meter (φ-OTDR) based DVS is realized and regarded as the
FANG. The vibration events at 9 observation points with different ground
conditions along the sensing fiber are detected by a threshold algorithm during 6.5
hours from 8:00 am. Then, the vibration events are analyzed in combination with
the ground conditions to recognize the machine working in the factory, rammer
working and the vehicles passed through near different areas and roads. The traffic
flow is estimated by the vibration-counting with a counting error that is believed
to be in an acceptable range. The distribution and the fluctuation trends of the
estimated traffic flow are useful and enlightening for the traffic monitoring and
pre-warning of special events, such as an accident. 
Wenhai Zhang et al [2017] early and accurate detection of incipient cable
faults is of a great interest to the utility industry. Such faults can be classified into
sub cycle and multi cycle types. Although a lot of research has been conducted on
sub cycle types of faults, little work has been done for multi cycle types of faults.
We proposes a novel method for detecting and locating a multi cycle incipient
fault in a cable. The incipient fault is modeled as a self-clearing arcing fault. The
distortion degree of calculated voltage is used to detect the occurrence of an
incipient fault. The degree of match between the measured and calculated
waveforms is used to guide the search for the fault distance. The accuracy is

28
further improved by taking into account the incipient fault angle as seen in the
voltage waveform and the power loss characteristics. The proposed method has
been verified by PSCAD/EMTDC simulations. Lab experiments are also
conducted to further validate the proposed detection method.
Andrea Cozza [2019] proposed the Time-domain reflectometry (TDR), the
most widely used testing method for fault-detection in cables, is tested against soft
faults of increasing severity. The intensity of TDR echoes is proven to be an
unreliable estimator of fault severity, since both the bandwidth of testing signals
and the length of the fault have a strong impact on the results. Moreover, it is
proven that faults of very different severity may generate virtually identical echoes
making it impossible to assess how critical a fault is. Simple frequency-domain
estimators are introduced, based on models of a fault reflectivity, leading to
minimum requirements for the test bandwidth in order to accurately identify soft
faults. As a practical consequence, it is concluded that only faults above a critical
length can be accurately identified, implying that no clear decision can be taken
about shorter faults.
Chun-Kwon Lee et al [2019] a nuclear power plant (NPP) depends on
instrumentation and control (I&C) systems to ensure its safe and efficient
operation. In particular, I&C cables take on the pivotal role of measuring and
controlling the critical equipment of the NPP. Thus, it is indubitable that the
diagnostic technology of I&C cables for detecting faults and accurately assessing
their health status is required for ensuring the safety and reliability of the NPP
operation. We propose a diagnostic method that combines fault detection and
evaluation algorithm for the I&C cables with stepped-frequency waveform
reflectometry with signal propagation and reflection modeling. The signal
modeling allows the assessment of the fault with an estimated reflection
coefficient by separating the propagation and reflection effects of the measured
signal. In short, cable faults are differentiated and quantified regardless of
distance. The proposed algorithm is verified by characteristic impedance

29
measurement, various fault detection/evaluation experiments, and the fault
evaluation of local accelerated thermal aging cable.
Mohsen Jannati et al [2019] incipient faults in underground cables are often
resulted from electric stress and cable aging. If such faults occur as current spikes
in short periods, permanent faults may appear. Moreover, incipient faults may
disturb electricity transmissions because of detection delays and thus, precise well-
timed protection decisions cannot be made. Therefore, one of the most important
considerations of utilities in the monitoring process is to recognize this type of
faults from other conditions as soon as possible. In this project, a precise approach
based on Cumulative SUM and Adaptive Linear Neuron has been proposed. In
addition to its high-speed detection ability, the practical on-line implementation of
the proposed approach is simple as well.

30
 CHAPTER-3
EXISTING SYSTEM
3.1 INTRODUCTION
A bundle of electrical conductors used for carrying electricity is called as a
cable. An underground cable generally has one or more conductors covered with
suitable insulation and a protective cover. Commonly used materials for insulation
are varnished cambric or impregnated paper. Fault in a cable can be any defect or
no homogeneity that diverts the path of current or affects the performance of the
cable. So it is necessary to correct the fault. Power Transmission can be done in
both overhead as well as in underground cables. But unlike underground cables
the overhead cables have the drawback of being easily prone to the effects of
rainfall, snow, thunder, lightningetc.This requires cables with reliability, increased
safety, ruggedness and greater service. So underground cables are preferred in
many areas especially in urban places. When it is easy to detect and correct the
faults in overhead line by mere observation, it is not possible to do so in an
underground cable. As they are buried deep in the soil it is not easy to detect the
abnormalities in them. Even when a fault is found to be present it is very difficult
to detect the exact location of the fault. This leads to dogging of the entire area to
detect and correct the fault which in turn causes wastage of money and manpower.
So it is necessary to know the exact location of faults in the underground cables.
Whatever the fault is, the voltage of the cable has the tendency to change
abruptly whenever a fault occurs. We make use of this voltage change across the
series resistors to detect the fault.
3.2. FAULTS IN UNDERGROUND CABLES
3.2.1 OPEN CIRCUIT FAULTS
These faults occur due to the failure of one or more conductors. The most
common causes of these faults include joint failures of cables and overhead lines,
and failure of one or more phase of circuit breaker and also due to melting of a
fuse or conductor in one or more phases. Open circuit faults are also called as

31
series faults. These are unsymmetrical or unbalanced type of faults except three
phase open fault.
3.2.2 SHORT CIRCUIT FAULTS
A short circuit can be defined as an abnormal connection of very low
impedance between two points of different potential, whether made intentionally
or accidentally. These are the most common and severe kind of faults, resulting in
the flow of abnormal high currents through the equipment or transmission lines. If
these faults are allowed to persist even for a short period, it leads to the extensive
damage to the equipment. Short circuit faults are also called as shunt faults. These
faults are caused due to the insulation failure between phase conductors or
between earth and phase conductors or both.
The various possible short circuit fault conditions include three phase to
earth, phase to phase, single phase to earth, two phase to earth and phase to phase.
In single line to ground fault, fault occurs between any one of the three lines and
the ground. In double line to ground fault, fault occurs between any two of the
three lines and the ground. In line to line fault, fault occurs between any two lines.
When fault occurs there is an abrupt change in voltage. This change in voltage
may cause serious damages to the system if not corrected in time. So immediate
step of fault correction is isolation of the faulty part from the rest of the system
3.3 FAULT DETECTION METHODS
3.3.1 ONLINE METHOD
This method utilizes and processes the sampled voltages and current to
determine the fault points. Online method for underground cable are less common
than overhead lines.
3.3.2 OFFLINE METHOD
In this method special instrument is used to test out service of cable in the
field. This offline method can be divided into two methods. They are tracer
method and terminal method.

32
3.3.3 TRACER METHOD
In this method fault point is detected by walking on the cable lines. Fault
point is indicated from audible signal or electromagnetic signal. It is used to
pinpoint fault location very accurately.
3.3.4 TERMINAL METHOD
It is a technique used to detect fault location of cable from one or both ends
without tracing. This method use to locate general area of fault, to expedite tracing
on buried cable.
3.4 RELATED WORK
In underground cable fault detection and identification via Fourier analysis.
The methods of impedance calculation via sending end voltage and differential
voltage can be used for differentiating between the different types of cable defects
from phase information. It needs study to be conducted to find the best way of
visualizing the results, especially the magnitude response. Behaviour of
simultaneous fault signals in distribution underground cable using DWT. The
simulations were performed using ATP/EMTP, and the analysis behaviour of
characteristics signals was Performed using DWT. Various case studies have been
carried out including the single fault and simultaneous fault.
Fault location algorithms without utilizing line parameters. By utilizing
unsynchronized voltage and current measurements from both ends of line without
requiring line parameters based on the distributed parameter line model. The fault
location estimate is not sensitive to measurement errors while line parameter
estimates are sensitive to measurement errors. Thus relatively precise
measurements are required to obtain accurate line parameter estimates. an
automatic fault location method using TDR. This method uses acquired data from
an existing TDR instrument. It enables user of TDR equipment to locate ULVDN
cable faults without user interpretation. An impedance based fault location method
for phase to phase and three phase fault. This method utilized the measured
impedance by distance relay and the super imposed current factor to discriminate

33
the fault location. This method is sensitive to the measured impedance accuracy
and super imposed current factor a resilient incipient fault location algorithm in
the time-domain, which utilizes data collected by PQ monitors to estimate the fault
location in terms of the line impedance by taking into account the arc voltage
associated with the incipient cable faults. So the algorithm predicts cable fault
location between two adjacent manholes. The ANNs are a family of statistical
learning algorithm inspired by biological neural networks and are used to
appropriate functions that depend on the large number of inputs. The existing
algorithm exactly pin-points the exact fault in the underground cable.
3.5. EXISTING METHOOD
The circuit consists of a power supply, 4 line display, and Arduino and
resistance measurement circuit. To induce faults manually in the kit, fault switches
are used. About 12 fault switches are used which are arranged in three rows with
each row having 4 switches. The 3 rows represent the 3 phases namely R,Y and B.
The fault switches: have 2 positions-No fault position(NF) and fault
position(F).Main component of the underground cable fault detection circuit is
low value resistance measurement. It is constructed using a constant current source
of 100mAmps. It can measure very low value resistance as the cables have around
0.01 Ohm/meter resistance. For 10meter cable resistance becomes 0.1 Ohm. This
circuit can measure resistance up 50 Ohm, Maximum cable length it can check up
to 4 kilometers. So starting from the reference point 4 sets of resistances are placed
in series. These 4 sets of resistances represent the three places and the neutral.
Short circuit faults, Symmetrical and unsymmetrical faults can be determined by
this method. This project uses three set of resistances in series (i.e.) R10- R11-
R12-R12,R17-R16-R14R21,R20-R19-R18-R25 one for each phase. Each series
resistor represents the resistance of the underground cable for a particular distance
and so here four resistances in series represent 1-4kms.Value of each resistance is
10kΩ.

34
 CHAPTER-4
PROPOSED SYSTEM
4.1 INTRODUCTION
The electrical energy is sent through the electrical transmission and
distribution system to the customers from the generation unit. When a line hit with
a fault, it is necessary to detect fault for the power system and to clear it before the
damage of the power system is increased. The underground cable system has an
edge over the overhead line system but to find the fault location is hard. The
development of locating fault techniques are on demand for reliable service. The
resistance in the cables gets affected by the cable faults. If allowed to continue, a
voltage breakdown can happen. The fault location method is a technique that helps
in finding the fault that occurs periodically in the line, an Arduino microcontroller
which is programmed shows the accurate and specific value of the resistance
voltage drop digitally, which are connected in series. The ADC converts the
distance in units from the base station. This paper comprises with LCD, GSM,
Arduino and IoT. The time of the system is reduced and constructively operates on
the system.
4.2 OBJECTIVE
This project deals with the location of fault occurs in the cable by using the
Arduino board that gives us the fault distance from the station. In common, urban
region practices the digging method to find the underground cable fault which
consume huge amount of time to find exact location of fault. The method used by
underground lines fault detection is a technique for locating the fault. This project
shows us the way to find the fault location which helps in avoiding the digging
method for the whole line. This will save time and reduce human efforts.
4.3 SYSTEM ANALYSIS
The electrical energy is distributed to the different loads which are
generated in the stations and consumed by villages, towns and cities then. The

35
voltage is stepped up in this process to minimize the heat energy loss in the lines.
The grid station sends the stepped up voltage which is stepped down by the local
transformers and send to the consumers for the use. The physically cutting and
splicing the cable is the basic method of locating a cable fault. For the fault
location, we start diving the cable into smaller parts so that the search is narrow
down and fault get detected. For example, let’s take a 1000-ft length wire which
has to be divided and cut into further smaller section i.e.500- ft length. Again this
500-ft length wire is cut into 250- ft length sections.
This section is then measured by ohmmeter or by an insulation resister (IR)
tester at both ends. The lower IR is shown by the defective section than at the
correct section of the wire. The method of divide and conquer is used until we can
find the smaller section of the cable at which fault occurs. This process becomes
very hectic and involves repetition in cable evacuation.
There are many disadvantages to the existing system. For the detection of
the fault in a cable whole cable is checked whenever a fault occurs in a cable. It is
time- consuming and requires a lot of human efforts. This method is used by the
short distance case only. During the repair work the probability of miss happening
increases. The interruption is caused due to striking of light on overhead cables as
they are vulnerable to it. The bare conductors are used which causes damage if
they break. The voltage drop is high and the maintenance cost is also high.
4.4 PROPOSED SYSTEM
This document mainly focuses to detect location range of the underground
wire fault. This is going to use the idea of Ohm's law. As per this principle, when
small DC voltage is feed at one end of cable using network of resistors, there is
change in current based on position of fault in the wire. If there occur any Short
Circuit, then there is change in voltage in the wire, now this is given to the Analog
to digital Converter of Arduino Mega development board that provide accurate
digital data and will be displayed on LCD. In this way, we are able to get the
specific range of defected wire (fault) from the base station. This System alarms

36
when defect in wire occurs. It may also be used for the transmission of wire under
the ground. Wires have their own resistance; our prior concern is that, the
resistance of cable may change in accordance with the wire length. When there is
increase in cable length, the resistance value increases accordingly. The Fault
point occur whenever there is fluctuation in the resistance and we identify this
point using GSM module. This location shown on LCD display measuring from
the base station.
4.4.1 OUTCOMES AND BENEFITS
 We are able to find exact range of fault.
 Saves Man Power.
 Saves cost and Convenient.
 This may utilize for overhead and underground cables.
4.5 HARDWARE REQUIREMENTS
Hardware i.e. all the physical components, are essential to govern the flow
of process by sensing the values and send it to Arduino. Following hardware, we
are going to use:
4.5.1 LCD Display
LCD refers to Liquid Crystal Display, used in many devices (to display
output). Similar to gas-plasma technologies and light-emitting diode (LED),
Unlike Cathode Ray Tube (CRT), display of LCD is much thinner. It uses very
low energy as compared to LED(s), gas displays as LCD is based on the principle
of intercepting beam of light in spite of emitting the light. LCD is based on
alphanumeric pattern. Meaning, this is able to show alphabets, numbers and
special symbols as well. This is very convenient device that is used to display
various information.

37
 Figure 4.1 LCD Display

4.5.2 TRANSFORMER
A transformer is an important part of circuit which is used to convert
electrical energy from one coil to another coil. This is based on the principle of
mutual induction among the coils. It can convert power from one circuitry to other
having no alteration of frequency, but there can be difference in the amplitude of
voltage. It can either increase or decrease amplitude of V or I in circuit. In today’s
times, most of these devices increases amplitude of voltage in order to minimize
any kind of trouble along the path. This is mainly employed to regulate the
voltage. Depending upon the behavior, there may be two types of transformers
available: Step up and Step-down.
4.5.3 DIODES
It’s a device that used as one-way switch for flow of current. The main
functionality is to pass the current in only one direction while restricting in the
other direction. In this way, this may be seen as a replica of electronic check valve.
The one-way switch behavior is known as rectification, usually operated within
rated level. It basically restricts the flow in reverse direction. Also, a reverse
voltage lies within a limited range so that to retain reverse barrier from
breakdown.
4.5.4 VOLTAGE REGULATOR

38
 Voltage regulator is a device that gives the voltage of constant level itself.
This produce a voltage having constant magnitude. This constant voltage is fixed
irrespective of changes in load value or its input voltage.

Figure 4.2 Voltage Regulator 7805

4.5.5 ARDUINO BOARD

Arduino is a publicly-available gadget that stages work on the simple to-
utilize system and programming. Microcontroller boards easily understand inputs -
light fall on a sensor, easily sense touching of a button or can read a Twitter
information – and provide an effective output -starting a motor, turning ON a
LED, circulate information on the web. The Arduino Mega is a board that
dependent on the ATmega2560. It consists of 56 I/P and O/P pins (in which 15
pins are allocated for the output of PWM). A sixteen MHz fired resonator, a USB
contact, a power jack, an ICSP header, and a reset button, it has all the features
which are required for the microcontroller basically, interface PC with Arduino
through a USB connection or give supply with a battery or a connector to get
started.

39
 Figure 4.3 Arduino UNO Circuit Board
4.5.6 GSM MODULE (SIM900)
This module is a GSM/GPRS four-band module developed for the
worldwide market. It supports four frequencies with the different range of GSM
850MHz, 900MHz, 1800MHz and 1900MHz. Its unique quality of GPRS multi-
slot classes support coding schemes of GPRS with multi-slot class having
10/8GPRS mobile station of class B.

Figure 4.4 SIM 900 GPRS/GSM Module

With a small configuration, SIM 900 can provide different basic space
requirements in users’ applications, like M2M and smart phone. SIM 900 consist
of sixty-eight SMT pads, and they provide all hardware interfaces. SIM 900
combine TCP/IP protocol and spread TCP/IP the AT commands of TCP/IP which
are important for the transfer of data.
4.5.7 WI-FI (ESP 8266)
In this project we are using ESP8266, a Wi-Fi module which is used for
sending the data to the cloud. For sending the data first we create an account on

40
Ask Sensors IoT platform. Ask Sensors is an IoT platform providing
communication between the cloud and different internet connected devices. This
platform provides a free account so you not need to pay to get started.

Figure 4.5 Wi-Fi Module

4.6 SOFTWARE REQUIREMENT
Software Requirements explain the need of different software program and
precondition that are required to install in the computer system to give best
functioning of the software. These different requirements and preconditioning are
basically, not covered in the software so, we need to set up a package which must
be install independently before the installation of software. The software which is
required for the fault detection system is Arduino programming language.
4.6.1 ARDUINO SOFTWARE (IDE)
The publicly available Arduino Software (IDE) that provide a platform to
write code in simple way and send to the microcontroller. This works on Linux,
Mac OS and also on Windows. The domain of this software is composed in Java
which work on Processing and different publicly-available software. HTML is
ideal mark-up language used for generating Web pages. HTML stands for Hyper
Text Mark-up Language. HTML define the basic structure of webpages. HTML
element is basic block of HTML page by using different tags we can easily
represent the HTML elements.
4.6.2 BLYNK APPLICATION
This application was created for the IoT and used to manage the hardware.
This app shows the data of sensor and visualize it. This application creates an

41
interface between the project and various widgets also responsible for the remote
communication between device and the user.
4.7 GENERAL DESCRIPTION
Today in many metro cities underground cable system is largely familiar
where repairing of cable it very problematic because finding the location of fault is
really a tough task. With this device, we can easily find exact position of fault. An
LCD display and a web net page are two important part of this system. This
system work on the principle of Ohms law i.e., whenever a small voltage (DC) is
given through the network of resistor at the end of the cable, then there is change
in the current value which depend on the exact position of the fault within the
cable. If any short circuit occur, there is a voltage change across the resistors. This
change in voltage is given to A/D convertor which generate the accurate digital
data and displayed on LCD. This project consist of the set of resistors which are
connected in series showing the length of the cable in km and switches are used to
generate the fault at each km to verify the perfection of this system. The location
of the fault can be shown on the display of LCD display which is connected to the
Arduino Mega. There is change in the value of resistance occur when the switches
are open, and the fault distance is easily calculated. The distance can be shown on
the display of the LCD in Km. the Blink App will show the location of faults.
4.7.1 PRODUCT FUNCTIONALITY ARDUINO
Arduino Mega is a microcontroller development board that cause the
different application maximum reachable that includes very interactive objects
around its surroundings. Arduino work as a server in the underground wire fault
detector. This project consists the resistors which are connected in series. These
series resistors are used to show that cables are conductors having current carrying
capability. Whenever we choose any specific cable the particular LED which is
analogous to the cable bloom showing the line is active. The set of switches is
used for fault generation. If switch is open it indicate the fault. Arduino read the

42
resistance value and calculated the distance. The distance which is calculated is
shown on the display of LCD and the BLYNK App.
4.7.2 WI-FI MODULE
ESP8266 is a Wi-Fi microchip. It has microcontroller Capability having full
TCP/IP stack. ESP8266 is 32-bit microcontroller. It has 16 input GPIO pins. It
consists of 32kiB instruction and 80 Ki user data, a power jack, a RJ45
connection, an ICSP header, and button to reset is used. This tiny module provides
the feature of connecting to the Wi-Fi network and generate simple connections of
TCP by using different types of commands. The ESP8266 with 1MiB of built-in
Flash memory help in making the devices having single-chip which having
capability of connecting to Wi-Fi network. It has authentication of WEP or WPA,
or open network.
It is having a UART with the dedicated pins, and consists of a transmit-only
UART that can be enabled on 10-bit ADC and GPIO2. ESP8266 module can
introduce an application. It has feature of unloading networking functions of Wi-Fi
with the help of other application processor. ESP8266 is a pre-programmed
module having set of an AT command. ESP8266 can be easily connect with the
microcontroller. It has powerful storage capability due to which it can be easily
integrated with application devices and the sensors through GPIO.
4.8 BLOCK DIAGRAM

43
 Figure 4.6 Block Diagram of Proposed System
In this project the concept of OHMs law is used. When we applied low
voltage (DC) at the end in the resistors which are connected in series then the
value of current would depend on the length of fault of the cable. The changed
voltage value in the circuit is fed to an ADC which send the value to the
microcontroller. We made this project with the set of resisters which are connected
in series that represent the length of cable in KMs. This project is made in five
major parts –Power supply, controlling, cable part, GSM and Wi-Fi module and
display. Power supply which converts ac signal to dc signal with desired voltage
with the help of bridge rectifier. The set of switches which are connected in series
represent cable length and that represents faults in the cable. Microcontroller
Arduino Mega is the controller which send signals to the connected module and
also make calculation regarding fault.
The LCD screen is used to show the fault location which is connected to the
microcontroller. Simultaneously microcontroller send signal to the GSM module
which send fault location message to the connected device and also this message
signal is send to the server by using Wi-Fi module to the BLYNK application
which maintain record and real time analysis for the faults.

44
 Figure 4.7 Flow Chart
4.9 ALGORITHM
Step1: Initialize LCD, GSM and Wi-Fi module.
Step2: Start an infinite loop.
Step3: Check weather fault is created or not.
Step4: If fault is not created then continuously monitor.
Step5: Fault is created then check analog data for distance.
Step6: Display fault on screen and call send sms and send cloud function.
Step7: Data send to Blynk app.
Step8: Repeat step3 to step7 for other phases.

45
 CHAPTER-5
RESULT AND DISCUSSION

46
 CHAPTER-6
CONCLUSION
Electrical Cables help in distribution of electrical energy. These cables face
so many failures. Which is very complicate task to detect the faults in these cables.
This system with the help of Arduino find the exact position of fault in cable from
the base station in Km. Nowadays in many non-rural areas, the wires in
underground are frequently used rather than atop lines. Whenever there is a fault
in underground cable it become very tough to locate the correct position of the
fault for the repairing of the cable. This system will work effectively for
underground as well as atop cables. An Arduino Mega board used in this system.
Here the Arduino is interfaced by the current sensing circuits which is made of
combination of many resistors. The fault is generated by the set of switches.
To increase the remote controlling capability of this industrial system we
have proposed the low-cost solution. This project on detection of fault with
working Arduino was made and fault distance in km from the ground station will
be shown on the screen of LCD and webpage. Whenever there is the fault occurs
then the switch which is analogous to the phase is recognized as the faulty phase
to which fault switches are operated. In this way the sector having fault can be
easily positioned.
It is durable, safe and low consuming power device. This device can run on
various channels to escape the interference with equipment or another wireless
device. With the help of microcontroller, we can accurately detect the fault
position. As faults occur in the cable, the fault location is displayed on the LCD
display.

47
 REFERENCES
[1] Massimo Marzinotto; Giovanni Mazzanti, 2015, “The Feasibility of Cable
Sheath Fault Detection by Monitoring Sheath-to-Ground Currents at the Ends of
Cross-Bonding Sections”, IEEE Transactions on Industry Applications, vol: 51,
no: 6, pp: 5376 – 5384.
[2] Giuseppe Parise; Erling Hesla; Claudio S. Mardegan; Luigi Parise; Elisabetta
Bragagni Capaccini, 2019, “Measures to Minimize Series Faults in Electrical
Cords and Extension Cords”, IEEE Transactions on Industry Applications, vol: 55,
no: 5, pp: 4551 – 4556.
[3] Mashad Uddin Saleh; Chris Deline; Evan Benoit; Samuel Kingston; Ayobami
S. Edun; Naveen Kumar Tumkur Jayakumar; Joel B. Harley; Cynthia Furse;
Michael Scarpulla, 2020, “An Overview of Spread Spectrum Time Domain
Reflectometry Responses to Photovoltaic Faults”, IEEE Journal of Photovoltaics,
vol: 10, no: 3, pp: 844 – 851.
[4] Reddipalli Bhargav; Bhavesh R. Bhalja; Chandra P. Gupta, 2020, “Novel Fault
Detection and Localization Algorithm for Low-Voltage DC Microgrid”, IEEE
Transactions on Industrial Informatics, vol: 16, no: 7, pp: 4498 – 4511.
[5] Zhinong Wei; Yuexuan Mao; Zhihua Yin; Guoqiang Sun; Haixiang Zang,
2020, “Fault Detection Based on the Generalized S-Transform With a Variable
Factor for Resonant Grounding Distribution Networks”, IEEE Access, vol: 8, pp:
91351 – 91367.
[6] Laurens Mackay; Elisabeth Vandeventer; Laura Ramirez-Elizondo, 2018,
“Circulating Net Currents in Meshed DC Distribution Grids: A Challenge for
Residual Ground Fault Protection”, IEEE Transactions on Power Delivery, vol:
33, no: 2, pp: 1018 – 1019.
[7] Catherine E. Jones; Patrick J. Norman; Michal Sztykiel; Rafael Peña Alzola;
Graeme M. Burt; Stuart J. Galloway; Luiz F. Kawashita; Stephen R. Hallett, 2018,
“Electrical and Thermal Effects of Fault Currents in Aircraft Electrical Power

48
Systems with Composite Aero structures”, IEEE Transactions on Transportation
Electrification, vol: 4, no: 3, pp: 660 – 670.
[8] Snehamoy Dhar; R. K. Patnaik; P. K. Dash, 2018, “Fault Detection and
Location of Photovoltaic Based DC Microgrid Using Differential Protection
Strategy”, IEEE Transactions on Smart Grid, vol: 9, no: 5, pp: 4303 – 4312.
[9] Enrico Pons; Riccardo Tommasini; Pietro Colella, 2017, “Fault Current
Detection and Dangerous Voltages in DC Urban Rail Traction Systems”, IEEE
Transactions on Industry Applications, vol: 53, no: 4, pp: 4109 – 4115.
[10] Indra Man Karmacharya; Ramakrishna Gokaraju, 2018, “Fault Location in
Ungrounded Photovoltaic System Using Wavelets and ANN”, IEEE Transactions
on Power Delivery, vol: 33, no: 2, pp: 549 – 559.
[11] Sheng Liang; Xiaoting Zhao; Ruijun Liu; Xiaodong Zhang; Liang Wang;
Xinghua Zhang; Yisheng Lv; Fanchao Meng, 2019, “Fiber-Optic Auditory Nerve
of Ground in the Suburb: For Traffic Flow Monitoring”, IEEE Access, vol: 7, pp:
166704 – 166710.
[12] Wenhai Zhang; Xianyong Xiao; Kai Zhou; Wilsun Xu; Yindi Jing, 2017,
“Multi cycle Incipient Fault Detection and Location for Medium Voltage
Underground Cable”, IEEE Transactions on Power Delivery, vol: 32, no: 3, pp:
1450 – 1459.
[13] Andrea Cozza, 2019, “Never Trust a Cable Bearing Echoes: Understanding
Ambiguities in Time-Domain Reflectometry Applied to Soft Faults in Cables”,
IEEE Transactions on Electromagnetic Compatibility, vol: 61, no: 2, pp: 586 –
589.
[14] Chun-Kwon Lee; Gu-Young Kwon; Yong-June Shin, 2019, “Condition
Assessment of I&C Cables in Nuclear Power Plants via Stepped-Frequency
Waveform Reflectometry”, IEEE Transactions on Instrumentation and
Measurement, vol: 68, no: 1, pp: 215 – 224.
[15] Mohsen Jannati; Behrooz Vahidi; Seyed Hossein Hosseinian, 2019,
“Incipient Faults Monitoring in Underground Medium Voltage Cables of

49
Distribution Systems Based on a Two-Step Strategy”, IEEE Transactions on
Power Delivery, vol: 34, no: 4, pp: 1647 – 1655.

50