These slides presents an overall discussion on fault location techniques generally used in present power transmission and distribution system. Later of the class we will discuss about the implementation principles and mathematical formulations.

1. Class-17: Fault Location
Techniques in Smart
Distribution System
Prof. (Dr.) Pravat Kumar Rout
Department of EEE
ITER
Siksha ‘O’ Anusandhan (Deemed to be University),
Bhubaneswar, Odisha, India

2. Introduction: Types of fault 1/5
 Balanced faults and unbalanced faults ; also known as symmetrical
faults and asymmetrical faults respectively. Faults can be also
categorised as series and shunt faults.
 Shunt faults: (1) Single line to ground faults (70% occurrence, less
severe); (2) Line to line fault (15% less severe ; (3) Double line to
ground fault (10% less severe ; (4)Three phase to ground fault (5%
more severe);
 Fault occurrence due to power system elements: (1) Transformers
(10%); (2) Overhead lines (50%); (3) Underground cables (9%); (4)
Switch Gears (12%); (5) CT, PT relays, Control equipment etc.(12%);
(6)Generators (7%);
2

3. Introduction 2/5
 Rapid fault location techniques are very much
necessary in order to restore the power supply
quickly in order to restore the power supply
quickly by reducing outage duration and revenue
losses.
 The accuracy of traditional fault lactation
techniques is also being affected significantly by
DG integration in smart distribution system.
 Almost 80% of distribution system interruptions
are caused by faults.
3

4. Introduction 3/5
 Additionally, power quality (PQ) and reliability requirements
from the deregulated electricity markets motivate the
researchers to improve the fault location techniques in order to
speed up the restoration processes.
 Accurate fault location and rapid restoration ensure secure and
stable operation of smart distribution grid .
 The inherent complexities of distribution networks such as
non-homogeneity, unbalanced structures, the presence of
laterals, and so on prevent the straightforward application of
developed fault location methodologies for transmission
networks.
4

5. Introduction 4/5
 In traditional distribution systems, substation voltage and cur-rent are
the only available measurements mostly having a sampling frequency
of 0.4–6.4 kHz. However, implementation of some fault location
methods, such as travelling waves-based methods, requires
measurements with more than 100 kHz sampling rates.
 The required input data is an important criterion which
differentiates fault and outage area location methods and
determines their practicality for a certain distribution network.
5

6. Introduction 5/5
 The nature of fault location schemes designed for distribution system has
traditionally being based on unidirectional power flow assumption, thus,
making them ineffective for a distribution system with a higher level of
DG penetration.
 New grid regulations requires DGs to remain connected to the system in
the vent of a fault to provide support to the grid and improve system
reliability.
 The fault current indicator operation too gets affected with the
presence of DG on a distribution system. A fault current indicator placed
between the fault point and the DG unit can get activated if a DG unit is
present downstream of a fault.
6

7. Fault location Techniques
 Traditional Techniques ( like Impedance-based fault location)
 Travelling-wave-based fault location
 Knowledge-based fault location
Though several attempts have been reported for fault location in
DNs, the integration of distributed generators (DGs) to
distribution grids makes the systems complex and affects
traditional protection schemes as well as the accuracy of
available fault location techniques.
7

8. Fault location problems and challenges
in distribution networks
 Geographic dispersion of distribution networks over a vast area;
 Existence of non-homogeneous lines;
 Presence of laterals, load taps and sometimes single and two phase
loads;
 Limited measurements, typically only available at substations;
 Dynamic topology of distribution networks;
 The effect of fault resistance which is usually non-negligible;
 Multiple fault location in distribution networks due to presence of
several branches.
8

9. FAULT LOCATION TECHNIQUES: Impedance-
Based Techniques 1/4
 The integration of DGs to distribution grids is imposing new
challenges as the nature of the power flow changes from
unidirectional to multi-directional. This affects the accuracy
of traditional impedance based fault location techniques.
 Useful only to the radial distribution system.
 PMU measurements can be incorporated in impedance based
fault location technique to enhance its accuracy.
9

10. FAULT LOCATION TECHNIQUES: Impedance-
Based Techniques 2/4
 The major limitations of impedance based techniques include:
1: Multiple estimations
2: Hectic iterative process
3: Accuracy depends on the dynamic behaviour of the loads,
the presence of DGs, laterals, line parameters, operating
modes, saturation characteristics of transformers, the
presence of noise in data measurements etc.
4: The effectiveness of the approach also be affected by non-
homogeneity of the networks and fault resistance.
10

11. FAULT LOCATION TECHNIQUES: Impedance-
Based Techniques 3/4
 As these techniques depend on fundamental frequency
components, the harmonics and transient nature of current can
create difficulty in accurately extracting the fundamentals.
 Fault resistance and system loading may create serious errors in the
measured impedance.
 In multilateral distribution network these schemes suffers from the
problem of multiple fault position in the network.
 Fault events having small duration presents a challenge, due to
short data window more analysis is required to get accurate results.
For long duration fault, the fault location estimate gives better result.11

12. Flowchart 4/4
12

13.  Generally applicable for the longer lines (transmission
networks)
 Generally, less influenced by operating modes, fault types of
the system, saturation characteristics of current
transformers etc.
 The schemes measure relative arrival times of travelling
waves produced by faults to estimate fault locations that
require a high-speed communication channel with wide
bandwidth for accurate measurement.
FAULT LOCATION TECHNIQUES:
Travelling-Wave-Based Techniques 1/4
13

14. FAULT LOCATION TECHNIQUES: Travelling-
Wave-Based Techniques Limitations 2/4
 Though travelling-wave-based fault location techniques offer satisfactory
solutions for most of the limitations of impedance-based techniques,
these fault location techniques are having few limitations.
➢ Usually applicable for longer lines only and difficult to apply for the
distribution networks with several branches and short lines.
➢ Requires high-speed communication channels with wide
bandwidth.
➢ Additionally, these methods are highly affected by the presence of
noise in data measurement procedures.
➢ Besides, the behaviour of travelling waves originated by faults is very
complex, which may lead to wrong decisions on fault locations.
➢ Requiring measurements with very high sampling rate. 14

15. FAULT LOCATION TECHNIQUES: Travelling-
Wave-Based Techniques Limitations 3/4
 Small fault inception angles and faults close to the place of locator
installation affect the accuracy of fault location estimation.
 There is a need of accurate synchronization of devices in case of
double ended schemes.
 In case of multiple discontinuities (reflection points) in distribution
networks, errors associated with detecting travelling wave.
 Difficulties in the configuration and location of fault transient
detectors due to complexity in distribution network.
15

16. Flowchart 4/4
16

17.  Sometimes, it becomes very difficult to locate faults employing
impedance-based and travelling-wave-based techniques because
of the presence of various complexities and uncertainties like non-
homogeneity, unbalanced structure, shorter length of the
distribution lines and/or cables compared to transmission lines
and/or cables, unknown fault resistance, and the presence of
laterals.
 Consequently, knowledge-based techniques were explored with
reasonable accuracy, which is comparatively less costly. It has
short execution time and generalization capability.
FAULT LOCATION TECHNIQUES: Knowledge-
Based Techniques 1/7
17

18. FAULT LOCATION TECHNIQUES: Types of
Knowledge-Based Techniques 2/7
 Artificial Neural Network (ANN) and Support Vector Machines
(SVM) based approach
 Signal processing based hybrid approaches
 Data mining based approaches
 Hybrid Methods
18

19. 1. Requiring a large amount of data for training;
2. Training should be repeated by any changes in distribution
network topology
3. A limited or inaccurate training data can affect the accuracy
of the fault detection algorithm.
4. Noise in the signal affects the performance of signal
processing based approaches
Limitations 3/7
19

20. Flow chart for knowledge
based fault location
technique 4/7
20

21. Flowchart for data
matching fault
location technique 5/7
21

22. FAULT LOCATION TECHNIQUES: Limitations of
Knowledge-Based Techniques 6/7
 Though knowledge-based approaches are straightforward and do not
require complex mathematical representation, till there are some
prominent limitations are associated with these methods.
 The performances of these approaches are highly dependent on the
quality and amount of training data.
 Additionally, limited or inaccurate information collected from an
insufficient amount of monitoring or measuring devices highly affects
the accuracy of these techniques.
22

23. FAULT LOCATION TECHNIQUES: Limitations of
Knowledge-Based Techniques 7/7
 On the other hand, placing high-quality monitoring devices in each node is very
expensive.
 Consequently, the industry needs to look for an appropriate trade off between
expenses and quality of the measured data.
 Furthermore, selections of optimal/useful features from signal processing
techniques and proper parameters for ANN/SVM/ANFIS/FLS are always a headache.
 Last but not the least, ANN/SVM/ANFIS/FLS are needed to be retrained if there is
any minor or major change in the network topology.
23

24. Challenges for fault location techniques 1/4
 The operation rules of smart DNs with DGs are not yet fully
grasped because of lack of relevant experiences.
 As the conventional fault diagnosis techniques simplify the
targeted networks, the reliability and accuracy of diagnosis
schemes are hampered.
 Though the transmission error and/or loss are very common in
data collection, most of the fault diagnosis schemes are based
on the correct and credible information.
24

25. Challenges for fault location techniques 2/4
 A little effort has been offered to diagnose faults of distribution
grid with smart measuring devices like phasor measurement
units/synchro-phasors.
 Though currently, the research of integrating DGs in the smart
distribution network is very popular, fault diagnosis by
combining multiple methods still needs further exploration.
25

26. Challenges for fault location techniques 3/4
 Must of the techniques are dependent on network topology (affected by
change in system configuration)
 The fault location methods yield multiple fault positions when applied to a
multilateral distribution network. Under this conditions the algorithms are
not able to differentiate between a faulty lateral and healthy lateral.
 High impedance faults have current magnitudes similar to those of normal
loads, such that it will not be causing any noticeable voltage drop. So it is
difficult to distinguish these faults.
 The dynamic response and stability of distribution system is changing due
to high penetration of distributed generation. 26

27. Challenges for fault location techniques 4/4
 Most of the proposed protection schemes are either centralised or
distributed in nature. But looking at the limitations of each approach ,
hierarchical approach may be desired.
 For improving distribution system supply reliability, the protection method
used should have the capability to provide transition into the islanded
mode of operation. This ability is not evident in the solutions proposed at
present.
 The integration of Electric Vehicle (EV) in distribution system for both
charging and discharging (for providing good support) is increasing at first
rate. Impact of this in distribution system especially in case of fault is
needed to be explored in depth. 27

28. Discussions, Scope and Conclusions 1/4
 With the recent trends of adopting DGs to distribution grids, it is more likely that
smart distribution grids will have an arbitrary and significant penetration of DGs in
the near future.
 Besides, locating faults in smart distribution grids is one of the major challenges in
order to restore power supply rapidly by shortening the outage duration and to
improve the reliability and the quality of power supply.
 Additionally, the sophisticated fault location techniques available for transmission
networks cannot be applied directly to distribution networks because of the
inherent properties of distribution grid such as non-homogeneity, unsymmetrical
and unbalanced structures, the presence of single, double, and three-phase
loads, and so on.
28

29. Discussions, Scope and Conclusions 2/4
It can be concluded that locating faults in smart distribution grid is still a promising field as most of the
available techniques are facing a different kind of challenge ranging from multiple estimations,
hectic iterative process, need for sophisticated and expensive communications devices and
channels with wide bandwidth, effects of noise in measured data or signal processing techniques,
lack of practical fault data, inaccurate and insufficient training and testing data, inherent
properties of distribution grids, the presence of DGs, frequent changes of the network
topologies, and so on.
Different hybrid methods based on decision trees/ensembles of decision trees, efficient
training algorithms for ANN/SVM/ANFIS/FLS, signal-processing technique-based extreme
learning machine, genetic programming as well as other signal processing techniques like S-
transform can be explored by incorporating the mentioned challenges.
29

30. Discussions, Scope and Conclusions 3/4
 The renewable energy resources are integrated with distribution
grids in order to reduce the emission of greenhouse gasses
and transmission power losses, to improve voltage profile and
phase imbalances, and to supply reactive power.
 This integration changes many technical parameters of the
grids that affects both protection schemes and fault location
techniques.
 Consequently, the researchers should attempt to explore the
improvement of avalable faut location techniques or develop
novel techniques to address the newly evolved challenges. 30

31.  Most of the methods are either good at fast detection capability or at accuracy
level. There is a need for combination of both to have reliable and high speed
protection scheme for distribution system with DG.
 To a great extent the available fault location schemes are formulated for specific
networks with specific DG. There is a need of fault detection techniques to be
developed in such a manner that they should be capable of application in a general
distribution network with multiple types of DGs.
 The choice of load model used in simulations affects the fault location accuracy.
Generally, constant current model and constant impedance model are considered
far deviate than the real case. So, it is important to formulate appropriate load
model as the distribution network behaviour after a fault, depends greatly on how
the load is presented.
Discussions, Scope and Conclusions 4/4
31

32. Questions?
 What are the major challenges for fault location methods for
smart distribution systems.
 Discuss about the major limitations of various fault location
techniques?
32

33. References
 Shafiullah, Md, and Mohammad Ali Abido. "A review on distribution grid
fault location techniques." Electric Power Components and Systems 45.8
(2017): 807-824.
 Bahmanyar, A., et al. "A comparison framework for distribution system
outage and fault location methods." Electric Power Systems
Research 145 (2017): 19-34.
 Kumar, Ranjeet, and Dipti Saxena. "A Literature Review on
Methodologies of Fault Location in Distribution System with
Distributed Generation." Energy Technology.
 Gururajapathy, S. S., H. Mokhlis, and H. A. Illias. "Fault location and
detection techniques in power distribution systems with distributed
generation: A review." Renewable and sustainable energy reviews 74
(2017): 949-958.
33