“Overview of Adaptive Protection System

for Modern Power Systems”

A
SEMINAR REPORT
submitted
in partial fulfilment
for the award of the Degree of
Bachelor of Technology
in Department of Electrical Engineering

Submitted To :- Submitted By: -

Er. Naveen kumar Tiwari Ritik sharma
Associate Professor RollNo.:17EAREE078
Electrical Engineering B.Tech. Final Year

Department of Electrical Engineering

Arya College of Engineering & Information Technology, Jaipur
Rajasthan technical university, Kota (Raj.)
Session: 2020-21
 CERTIFICATE

This is to certify that the Seminar Report on “Overview of Adaptive Protection System
Modern Power Systems” Submitted by Ritik Sharma Roll No.: - 17EAREE078
fulfilment for the award of degree of Bachelor of Technology in Department of Elect
Engineering, Arya College of Engg. & I.T., Jaipur.

Date: -
Place: - ACEIT, JAIPUR

Er. Surendra Khichar Er. Naveen Tiwari

Assistance Professor Seminar Coordinator
Electrical Engineering Electrical Engineerin

Dr. Prabhat Kumar

Head of Electrical Department

Department of Electrical Engineering

Arya College of Engineering and Information Technology, Jaipur
Rajasthan Technical University, Kota
 ACKNOWLEDGEMENT

It is matter of great pleasure and privilege for me to present this seminar project report on the
basis of practical knowledge gained by me during this seminar project report “Overview of
Adaptive Protection System for Modern Power Systems”.
I would also like to thank Er. Naveen Kumar Tiwari (Seminar Coordinator) & Surendra
Khichar(Assistance Professore) for their contribution in preparing my seminar project report and
to guide me for delivering the seminar project report. And for his pain tasking effort and
enthusiastic cooperation to make possible.
I express sincere thanks to Dr. Prabhat Kumar, Head of Electrical department of A.C.E.I.T. for
giving me an opportunity for conducting the seminar report.
I would like to express my deep gratitude to Management and Principle of Arya College of
Engineering & I.T. for extending me the opportunity for making seminar project report and
providing all the necessary resources for this purpose.

Ritik Sharma
Roll No.-17EAREE078
 Abstract
There exists a critical mass in research related to adaptive protection approaches that address some of the
shortcomings of conventional protection functions. This is in response to concerns in the reliability of
conventional protection which manifested itself in some severe disturbances in more recent years. Despite the
fact that adaptive protection offers a compelling technical solution to some of these performance problems, the
industry has not widely adopted adaptive protection approaches as a de facto policy for future protection scheme
implementations.
This is attributed to the difficulties associated with the testing of such schemes where no significant work has
been reported yet. Furthermore, the benefits vs. the risks associated with such a protection strategy are not well
understood. This is coupled with the conservatism towards radical changes in the way the power system is
operated. As such the work reported in this thesis complements the existing body of research in order to address
some of the major technical and institutional challenges associated with adopting adaptive protection schemes for
future networks, especially those networks that exhibit flexibility in operation to deal with uncertainty in
generation and to maximise asset utilisation. These are network characteristics that adaptive protection
approaches are seen to be an effective enabler of.
This thesis focuses on formal structural and behavioural modelling of adaptive protection schemes as means to
effectively validate their functional operation and verify their performance. Novel contributions have been made
in formalising a user requirements driven architecture for these schemes. Furthermore, significant contributions
have been made to conducting formal algorithm verification that complements inherently limited standard
protection scheme validation techniques. The thesis makes thorough use of a proposed adaptive distance
protection scheme for circuits with quadrature booster transformers to communicate the challenges, lessons
learned and contributions in designing, implementing and testing adaptive protection schemes.

TABLE OF CONTENTS
 S.No Topic Page No.
i Certificate i

ii Acknowledgement ii

iii Abstract iii

iv List of Figures vi
LIST v Glossary of Abbreviation vii
OF
Chapter - 1 Introduction 1-2

Chapter-2 Literature Survey 3

Chapter-3 Overview Protect On System 4

3.1 Zone Protection 4-5

3.2 Overcurrent Protection 5-7

3.2.1 Protection device coordination 6-7

3.3 Directional Overcurrent Protection 7-8

3.4 Differential Protection 9

3.5 Distance Protection 10-11

Emerging Challenges In Protection

Chapter-4 12-13
4.1. Self-Healing 13

Chapter-5 Adaptive Protection Scheme 14-15

Chapter-6 Recommendation For Future Power System 16-18

Protection

A Mordern Perspective On Power System

Chapter-7 Protection

Chapter-8 Achieve Adaptiveprotection Functionality

Chapter-9 The Developed Adaptive Distance Protection

Logic

Design of adaptive protection schemes

Chapter-10
Conclusion

Reference
 FIGURES

Fig. No. Title Page No.

1 Overall structure of general protection system 5

2 Representation of overcurrent relay 5

3 Time Current Characteristics (TCC) Curve 7

4 8
Directional Overcurrent Protection
5 9
Differential Protection
6 10
Distance Protection
7 12
Smart Protection System
8 14
Adaptive protection of power supply and
17
distribution system
9
Block diagram of relay co-ordination software
 GLLOSSARY OF ABBRIVATION

Abbreviation Description

AC Alternating current

DC Direct current

C Capacitor

CKT Circuit

MW MegaWatt
 Chapter-1
INTRODUCTION
Electromechanical relay has been used for more than 60 years to provide protection for the
power. These types of relays have longer lifespan, needs low maintenance and reliable. However
static relay has shown higher usage demand from 1960 onwards. It provides lower burden,
improves the performance characteristic, self-monitoring and many more. For protection, one of
these options can be used. However, it should be consistent for the whole circuit Electromechanical
and static relays are examples of analogue relay. They do not determine the exact location of the
fault. Instead, will establish a zone of protection which the relay will operate. Analogue relays
models are not easy to make precisely. The main reason has been the lack of detailed design data
on which to base the model and the difficulties involved such as fitting a digital model to a non-
linear analogue device. These analogue protective relays are off- line and constant when
operating. This setting method cannot adapt to the changes of topology structure and operation
mode. It is because the relay cannot get the best protection effect.
The modern relays are digital and have an automatic setting. These relays are based on
microprocessors operating in real time. They can adapt to their settings. They do not need to
duplicate the conventual relay characteristic. Instead it can calculate the exact location of a fault
and arc. Therefore, there will be almost zero incorrect operations. These types of relays are used
by adaptive protection. By tallying and changing the information online, this relay can adapt to
the adjustment of the topology architecture, functioning method and the type of the fault
occurring in the system .
Based on the functions and information used, Adaptive Protection scheme can be categorized as
two separate areas; adaptive protection device and adaptive protection system. The device, in
general, uses actual data and is able to adapt to the changes of the processing mode for certain
zones . The integrated communication network shares the information throughout the system and
adjust to the changes of operation mode for large areas.
The key importance of a protection system should be reliable (operate when required), secure (not
to be operated unnecessarily), selective (only a minimum number of fault interrupting devices
required should operate), and fast (isolation of fault). In addition, backup protection should be
cleared when any fault upon failure of the protective equipment happened in the primary
protection system .
This paper presents a literature review of the overall protection system, modern power system and
adaptive protection sequentially. The adaptive protection, its usage and the evolution are also
being described in detail.
1
The rest of the paper is planned in the following sequence; Section II presents the overall
protection scheme in a power system. Section III discusses about the modern power system and
its scheme of protection. Section IV focuses on adaptive protection schemes. Section V gives
recommendation for future power systems. The Conclusions are summarized in section VI.

2
 Chapter-2
LITERATURE SURVEY

The process of literature review of TCSC provides brief application of power system network.
Many researchers proposed most sensitive method for obtaining optimal allocation of
FACTScontroller under various operating condition with best optimization techniques in multi
objective function, single objective function for system stability. P.Naderi et.al, proposed
sensitive index method is implementedto determining the exact location of TCSC on line
based on positive value of sensitive index reactive power loss. S.Varshney et.al, in proposed
system voltage collapse critical line is determined based on line stability index. Optimization
techniques focus on mitigation in line contingencies on small network.H.I shakeen et.al, proposed
system suggested for multi TCSC with minimum cost, optimal allocation, numbers and
parameter of TCSC has determined by Genetic Algorithm ( GA) and Particle Swarm
Optimization (PSO). Two methods compared for optimal process and simulated in mat
lab.Ahmad khan et.al, this paper suggested transient stability steady state stability dynamic
stability has improved on TCSC, optimizing techniques of Eigen value on considering IEEE 14
bus system. S.Sakthivel et.al,[8]proposed stability of voltage in interconnected system applied
Differential Equation Algorithm ( DEA) to maintain loadability,line contingency outage
condition. The proposed method test under IEEE 14 bus DEA with TCSC controllers maintain
stability margin in power system. M.Hamed et.al,proposed multi objective function,
reduction in power loss system voltage deviation PSO algorithm tends to determine the
parameter of TCSC result in stability of the system. The proposed method analyses two
methods PSO with TCSC, Simple Genetic Algorithm (SGA) with TCSC for IEEE 30 bus
system. Main drawback usingFACTS device In power system issues is planning and control for
multi objective process.

3
 Chapter-3

OVERVIEW PROTECTION SYSTEM

Protection system is one of the key stages to take note when designing a good power system. The
relays are designed where it finds the fault and quickly isolates the fault area without harming the
other parts of the system. Therefore, the information helps the industry to evade of damaging the
equipment with huge losses in the manufacturing process. This is due to the overload or fault of
the system. This further helps to isolate and replace faulty equipment, minimizes the property
losses and safety of the people.
There are some criteria to be followed throughout when choosing a relay protection. The main
key is that, it should isolate any fault very quickly, at the same time neighboring areas should
work smoothly. The relay must be affordable and should able to achieve high efficiency of the
power system safety. It should also be reliable and support high-speed operation. .

3.1. Zone Protection

The most basic type of protection for any type of system is the zone protection. It acts as a
primary and/or backup protection. When a fault happens, the first line of defense would be the
primary protection. The backup protections occur when it fails to isolate the fault primarily, and
by re-tripping the circuit breaker. Another option is tripping the circuit breakers of the
neighboring areas to the fault.
Backup Protections can be either Local or Remote based on their location in the system. Local
backup protections are based on the same zone as the fault happened. The circuit breakers of the
primary or the nearby zones would trip when a fault occurs. The remote backup protection is pre-
defined, and the circuit breakers will only trip when a fault occurs in that zone.
Overlapping neighboring zones is one of the standard conditions to maintain the power system as
shown in Fig 1. This method safeguards and protects the entire system. However, there are
consequences as both overlap zones will be isolated when a fault occurs. The system shall be
designed in such a way to minimize these overlapping zones as a solution .

The following diagram shows how circuit breakers are connected in the power system-

4
 Fig 1: Overall structure of general protection system

3.2 Overcurrent Protection

The Overcurrent Protection, is simple to use and is the mostly used protection scenario in the
radical distribution. Fig 2 is an illustration of how the overcurrent relay works.

Fig 2: Representation of overcurrent relay

Where, I’ can be defined as the generated current from the secondary side of the CT (Current

5
Transformer) and Ipickup is a standard rating of the relay setting, also known as the pick-up
current or the threshold current.
The following the principle for overcurrent protection: Normal operation: I’ < Ipickup
Short Circuit fault: I’ > Ipickup.
There is no trip signal and the circuit breaker is closed as shown in the formula for “Normal
operation”. However, the circuit breaker will be tripped and isolate the fault, when the secondary
current (I’) coming from transformer is greater than pick-up current (Ipickup). Two fundamental
parameters, the pickup current threshold and the time-dial setting of the relay shall be selected
properly to protect the system.
Overcurrent relays are the most commonly-used protective relay type. Overcurrent relay has two
types of protection; instantaneous overcurrent relay (Relay-50) and time-overcurrent relay
(Relay-51). The characteristics of the Relay-50 is that it operates as soon as the current gets
higher than a predetermined value when there is no intentional time delay is set. However, there
could be an inherent time delay in an order of milliseconds based on the manufacturer’s data.
In Relay-51, operating time is inversely changed with the current. There are three commonly used
characteristics which is inverse, very inverse and extremely inverse. It is possible to have
combined Relay-50 and Relay-51 in a System and is referred to as a Relay 50/51.

3.2.1 Protection device coordination

The basic role of the coordination technique is to correctly choose the information and
parameters for the relays on the distribution side of the power system. Ipickup and the time delays
define the thresholds, letting the system to disregard transient overloads and allowing it to operate
within norms up to the closest to a fault or breakdown. Without harming the sensitivity, the
protection device coordination needs to obtain the maximum selectivity.
Time Current Characteristic (TCC) curve tells how fast a protection device triggers when a fault
occurs. There is no characteristics for high voltage circuit breaker under time current curve. There
is a directional overcurrent relay being used with a directional element added to it. Under the
TCC, the low voltage circuit breaker has its own characteristic.
Commercially available electrical overcurrent device protection and coordination
evaluationsoftware provides an intuitive and logical approach to Time-Current Characteristic
curve selectivity analysis.

6
The above curve is the interpretation of a part of a circuit with high voltage circuit breakers and is
illustrated as the picture insert. Based on the latching times of this curve, Relay1-50 which
controls CB1 will trip first, followed by CB2 and CB3. After Relay7-50, Relay1-51 will continue
the process until force termination using the operator console.

3.3 Directional Overcurrent Protection

Relay 67 is used for Directional Overcurrent Protections. The concept is like the overcurrent
protection. However, for directional overcurrent can be flown in either forward or reverse
direction. This kind of protection is mainly used in parallel lines. For this type of protection there
are three types of parameters needed, which are voltage, current and the phase angle between
voltage and current. Relay-67 is often used as the intelligent electronic device to determine the
direction, the fault current is flowing and to trip the appropriate devices accordingly. The residual
voltage is three times the zero-sequence voltage drop on the source impedance during a ground
fault. The phasor voltage is balanced when the residual voltage is zero. The values must be
calculated precisely in order to have the exact directional protection.

Fig 3: Time Current Characteristics (TCC) Curve

7
 Fig.4: Directional Overcurrent Protection

3.4 Differential Protection

The differential transformer protection is denoted by 87T. It is a fast, selective method of

protection against short circuit in transformer. The concept it follows is the Kirchhoff’s Law. This
protection mainly works for internal protection of transformers where differential current is
greater than zero .
8
The differential protection does not support external faults and overloading of large motors,
generators, lines, cables, bus, etc., but only support and protect them from internal faults. An
overcurrent relays shall be used should there be a requirement to protect such systems from
external faults and overloads.

Fig.5: Differential Protection

3.5 Distance Protection

The distance relay is denoted by (Device 21). It is operated by using voltage and current to
determine if a fault is in the protection zone of the relay. These types of relay can detect phase
faults as well as ground faults. The characteristic of these relays can be seen in a R-X diagram.

9
The following are some basic characteristics of the relay.

Fig.6: Distance Protection

3.5.1 Impedance: The impedance characteristic does not take into account the phase angle
between the voltage and the current applied to it. In the R-X plane impedance characteristics
shown as a circle with its center at the origin. The relay operates when the measured impedance is
less than the settings.

3.5.2 Reactance: It only measures the reactive component of the impedance. The characteristic of
the impedance relay in the R-X plane is a straight line parallel to the R axis.

3.5.3 Quadrilateral: This characteristic can be achieved by combining directional and reactance
characteristics with two resistive reach blind characteristics.

10
3.5.4 Mho: The characteristic of Mho relay in the R-X plane is a circle whose circumference
passes through the origin. This type of relay operates when measured impedance falls within the
circle.

Chapter-4

EMERGING CHALLENGES IN PROTECTION

In recent years, electric power sources have been expanded and the large scale renewable energy
deployments, which are environmental friendly and avoid any affect to the climate changes, are
key to this energy transition.
11
To find solutions for those problems, Europeans and other countries suggested and came up with
the idea of smart grid. This is now being the key concept in the national power strategy.
These days smart grid has taken the name of the next generation of the power grid. Smart grid
uses digital technology to supply electricity and has two-way communication. It helps to improve
various things. They are reliable, reduces the peak demand and increase the energy efficiency.
This will benefit the environment along with the efficiency .
Analysis and Isolation is the most identified factor when designing a modern power system. It
should isolate fully when a fault occurs in the grid. Also, it should quickly isolate any affected
zones in the main grid or other parts in the grid for the safety of the load.
The smart protection system is the subsystem of the smart grid. It provides advanced grid reliability
analysis, failure protection, security and privacy protection services.

Fig 7: Smart Protection System

main objective of the adaptive protection scheme is to change the relay setting to match with the
main power system conditions.
Under the adaptive protective scheme, there are four main elements. They are hardware,
communication and control, software and human factors. The hardware refers to the digital
relays. Communication and control refers to the computational systems used to monitor the relays
and coordinate the changes needed in the face of changing network conditions. The software
refers to algorithms controlling the systems use to coordinate relays.

4.1. Self-Healing

12
In the principle of self-healing protection, the circuit breaker opens automatically without any
human intervention, when a fault happens and is intelligent enough to self-adjust quickly to
maintain the normal operation conditions. The circuit breakers are closed for the neighboring
areas and will not interrupt the system. It is one of the smart way of providing protection when a
fault occurs.
To achieve self-healing in a power grid, the system should have sensors, automated controls, and
advanced software that uses real-time distribution of data to detect, isolate faults and to
reconfigure the distribution network to minimize the power outage and customer impact. Self-
healing, supported by these control electronics and software, intelligently does reconfiguration,
load shedding, and also controls the output power of connected generators.
System reliability is one of the key objectives in self-healing. This can be done by rearranging the
circuit breakers and relays. To quickly isolate the fault, reclosers are mounted on the distribution
network. They can be integrated with other alternate sources to restore services to as many
customers as possible.
Many constraints due to non-linear optimization can be seen as one of the deadlocks in Self-
healing. The solutions using modern digital algorithms are underway using techniques such as
artificial intelligence.

Chapter-5

ADAPTIVE PROTECTION SCHEME

Adaptive protection has been established since the 1980s. It was enabled by the increase of the
13
computer-based relaying. This enabled to change the characteristic of the relay setting. The
evolving networks create the need for adaptive protection such as existing relay settings and
protection methods may become inappropriate.
The idea behind the adaptive protection defines the ability of the protection system to adapt to the
current operating condition of the power system. This technique is more relevant than modern
digital relays because it has automatic settings.

5.1 Overall structure of the adaptive protection system

The structure of the system offers its overall frame and define its functions and correlations of its
elements. Each module in the system would achieve their parts and interact with each other at the
same time. The following diagram Fig 5. is the representation of the overall structure of adaptive
protection system of power supply and distribution system.

Fig 8: Adaptive protection of power supply and distribution system

The following are the main functions of the modules:

5.1.1 Man-machine Interactive Module: All main operating parameters, information of
operating mode and topology structure, operating parameters of voltage, current and the
respective protection methodologies, verification techniques of selectivity and their sensitivities,
etc., are being fed and queried while it displays and provides the scenarios of the adaptive
protection device of the power supply and the distribution line.
14
5.1.2 Verification module of selectivity and sensitivity: It confirms whether it is reasonable for
the setting of the device and checks for both selectivity and sensitivity. The selectivity function
confirms whether the circuit breaker will trip or not if a fault happens at the end of the protection
scope.
5.1.3 Verification Setting and Calculating Module: It is one of the key systems. It checks
whether things change according to the information given by the SCADA system and operating
parameters of power supply and distributing line.

15
 Chapter-6

RECOMMENDATION FOR FUTURE POWER SYSTEM PROTECTION

We are in the era of Smart Solutions and the prediction is that the number of smart cities will
reach 600 worldwide by 2020. Digital technologies in the Smart Cities could become the engine
of economic progress, and Smart Grids, without a doubt, could be one of them. While opening
new trends and opportunities this may also generate many physical, social, behavioral, economic,
and infrastructural challenges. When addressing these challenges in implementing adaptive
protection solutions in modern power systems, two scenarios are explained in detail in following
paragraphs.

6.1 Software for adaptive protection system

Adaptive relaying is a new viewpoint in protecting electric power systems. It operates the
continuously changing status of the power system as the basis for on-line adjustment. It is
required because fault current and load alter with time. The reason for that is the generation and
topology alter. In order for this to occurs, the information in the relay needs to be altered.
Therefore, time delay can be reduced and maintaining the coordinate margins. The following
paragraphs will discuss about the different types of software for adaptive protection system.

6.1.1 Relay modeling software: Off-line mode is where relay’s features are being chosen from.
The different setting characteristics are stored in Time Multiplier Setting (TMS) of the adaptive
system. The relay’s characteristics at a selected TMS should be stored in the relay memory. If the
relays are fixed at different places in the distribution system, it might need to implement different
sets of parameters. The standard product characteristics are available on-line and can be used
whenever required. A significant role is being played by the pick-up value of the current relay
(Ipick up) and formulates the trip decision logic along with the TSM data.

6.1.2 Relay co-ordination software: The software studies the altering of the system and for the
setting of the relay, it will find the most suitable one. This software has four packages which is
topology detection, state estimator, fault analysis and relay setting program which can also be
seen in Fig 6.

16
 Fig 9: Block diagram of relay co-ordination software

6.1.3 Relay communication software: The main role of adaptive relay system is the
communication. It interchanges data between station computer and main computer There are
three types of packages in communication software. The first package supports various types of
relays processors to interact with each other. The second package support the station computers to
interact with the relay processors. The last package support interaction between the main
computer and station computers.
6.2 Functioning of adaptive protection system
Adaptive protection systems function in two modes; changes of system during normal operation
condition and the operations during faults and transients.
6.2.1 Changes of system operating conditions: In the first mode, the relay would test the
substation bus voltages and the circuit currents. This information will then be translated to
current and voltage phasor before transporting to the substation computer.
The changes that happen subsequently will be found by the system. It would be known by this
data. The main computer will receive any data which has been altered from the original. Then the
system starts computing a state estimation study and a fault analysis study. New relay settings
will be generated and transported back to the station computers to be restored in the individual
relays . It will then be communicated back to the individual relays.

17
6.2.2 Operation during faults and transients: In this second mode of operation, a form of look-
up table based on the relay characteristics is used when deciding the Ipick up and trip decision
logic. It will always monitor the circuit. If an abnormal condition happens, it will issue a
command to the necessary circuit breaker to trip and all the circuit configuration changes will be
transported to the main computer for calculating the new relay setting. It will then be
communicated back to the individual relays.

18
 Chapter-7

A Modern Perspective on Power System Protection

7.1 Chapter methodology and contributions

Modern substation technologies provide the building blocks for the realisation of new and
improved protection techniques and systems. This chapter examines the fundamental concepts
and recent developments in power system protection practices. A brief explanation of distance
protection principles is included as it will be revisited in later chapters of the thesis. Other
functions will not be discussed in detail as they have been treated exhaustively in previous theses
and related textbooks. Focus will also be placed on the emerging concepts of the digital
substation and wide area protection systems. Finally, the testing of protection schemes will be
discussed while identifying potential shortfalls of existing testing practices. This will be used as a
springboard for the development of improved functional testing methodologies in later chapters.
All protection functions discussed in this chapter are based on numerical methods.
The main contributions of this chapter are:
 Review of emerging approaches to power system protection including those utilising synchro
phasor technologies and digital substation functions.
 Discussion of limitations in protection system testing practices in coping with scheme
developments and new functional requirements.

7.2 Power system protection principles

Although protection systems represent a 5% capital investment of the overall power system they
are considered a fundamental operational component. Without protection schemes, power
systems cannot be operated in a stable, secure or reliable manner. Different protection functions
are deployed in transmission and distribution networks. This is mainly due to the more stringent
stability requirements placed on transmission networks. Consequently, more complex scheme
configurations are found on transmission networks in addition to redundant schemes. Distribution
networks on the other hand require more cost effective protection solutions due to the sheer
volume of feeders and network assets that need to be protected.
Recent developments in distribution network automation in addition to the increased penetration
of DG are stimulating more interest in distribution network protection. This particular field has
seen growth in research activity which led to the development of many improved protection
19
functions including those dealing with protection performance issues arising from islanded
network operation, changes in network topology, increased use of power electronics in generator
interfaces, etc.
7.2.1 Unit-based protection
Zones of protection are used to define the areas of the primary system which are protected by a
specific protection function. In a unit-based protection scheme (current differential for example),
the zone of protection boundary is defined by the instrument transformers used to measure the
current flow through the protected feeder as shown in Figure 2-1. Such schemes are mostly
applied to transmission networks where the cost of required communications is justified. These
schemes are also highly selective in their operation. However, they may suffer from instabilities if
current transformers are saturated due to high through fault currents. This can be mitigated by the
use of new sensing technologies such as Rogowski coils, hall effect sensors or optical current
sensors. These are referred to as non-conventional instrument transformers (NCITs). More recent
developments make use of optical fibre Bragg gratings to directly measure the line quantities
without intermediate electrical/optical transformations. This allows for faster acquisition of line
current from different points on the line using the same optical fibre

20
Fig. A typical current differential protection scheme showing zone of protection

differential relay circuit breaker current transformer (CT)zone of protection primary current
secondary current Digital current differential protection relies on communications to exchange
measurements made across the protection zone boundary. Dedicated point to point
communications links are commonly used for this purpose. Such links’ latencies can be
characterised so that corresponding measurements made at different physical points can be
compared at the same time regardless of communications channel delay. Alternatively, the delay
compensation algorithms (e.g. ping pong method) can be used to dynamically calculate this delay
and compensate for it. With the advent of non-deterministic packet switched communications
networks, compensating for channel delays becomes more problematic. To tackle this problem,
GPS synchronisation can be used were each measurement can be tagged with a GPS time stamp.
Therefore, only corresponding measurements are compared. When backhaul communications
infrastructure is used for exchanging measurements, routing technologies such as IP/MPLS
(Internet Protocol/Multi-Protocol Label Switching) can guarantee the communications quality of
service by prioritising protection traffic.
7.2.2 Non-unit based protection
Non-unit protection schemes can be found in both transmission and distribution networks. These
rely on local measurements made by instrument transformers to inform the protection functions.
Such functions include overcurrent and distance protection. The latter will be discussed in detail
in the following section due to its relevance to the remainder of the thesis.
Overcurrent protection is mostly applied in distribution networks due to its simplicity. It is
21
however also used in transmission networks as a backup protection function. Phase or earth faults
are detected by simply measuring the current at the relaying point and comparing that with a
predetermined pick-up setting. To achieve selectivity in operation, time delays or fault levels or a
combination of both are used. The latter method is most commonly used and is achieved using an
inverse definite minimum time (IDMT) characteristic as shown in Figure 2-2. The IDMT
characteristic ensures that faster operation is achieved with higher fault currents. A time setting
multiplier is used to coordinate the operation of relays in series (R1-R3) which creates a grading
margin. The grading margin is selected based on breaker operating times, errors in the protection
system and the overall acceptable operating time for the specific network.
7.3 Distance protection
Distance protection is mainly used in transmission systems. It is applied, to a lesser extent, in
meshed distribution systems to improve selectivity with a faster operating time [14]. Distance
protection relies on the simple principle that the protected line impedance is proportional to its
length. Therefore, by measuring (or more practically calculating) the protected line impedance, a
fault can be identified by monitoring changes in the impedance. These changes can then be
compared with impedance characteristics to determine the need to operate or restrain [14].
Multiple distance relays can be made part of a communications based scheme. Such schemes are
used to overcome reach issues or accelerate tripping of time delayed distance relays.

7.3.1 Elements, characteristics and polarisation

Distance protection relies on both current and voltage measurements in order to obtain the
apparent impedance of the circuit at the relaying point. This is then compared with the relay
settings which represent the distance protection reach (or protection zone boundary). Several
operating characteristics exist for distance protection. Modern numerical relays provide the ability
to create a custom characteristic. However, the most common ones used are the Mho and
quadrilateral characteristics. The MHO characteristic, as shown in Figure 2-3, is self-polarised.
The voltage measurement is used to restrict fault detection to those faults that occur downstream
of the relay.
7.3.2 Ground fault detection
Distance protection settings are expressed using positive sequence quantities. Apparent
impedance is also calculated in the same manner. Therefore, to accommodate ground faults, a
compensation factor is used in the calculation. This takes into account the ground loop impedance
during a ground fault situation. Different relay manufacturers implement this compensation factor
in different ways. Therefore, it is important to refer to the relay documentation to ensure the
correct settings of the factor. for instance, is normally calculated using the positive and zero

22
sequence line impedances and respectively as in Alternatively, a residual compensation factor can
be used where ⁄. In this case, the measured ground fault impedance will depend on the residual
current measurement instead of the zero sequence current used in .
7.3.3 Communications based distance schemes
The performance of distance protection schemes can be enhanced using communications
channels. This is particularly useful in interconnected transmission circuits where faults at certain
positions are not immediately cleared by zone 1 elements on both ends. Faster fault clearance
times can be achieved through remote signaling.
7.4 DER interface protection
Engineering recommendations such as G59/2 [26] in the UK or IEEE 1547 guidelines stipulate
the functions necessary to protect the DER. These differ according to the type of DER, the
voltage level it is connected to and the country. provides useful information on international
practices related to the protection of DER. This section focuses on loss of mains (LOM)
protection functions as it will be revisited.
7.4.1 Loss of mains protection
Loss of mains is the condition where a section of the distribution network is disconnected from
the main grid and remains energised by installed DER. This islanded mode of operation is not
currently permitted due to the following reasons:
 The islanded distribution network frequency may drift in relation to the main grid. Therefore,
out of synch re-closures at the point of common coupling are a possibility unless check
synchronism functionality is fitted.

 Power quality usually cannot be maintained by DER.

 Operational procedures normally assume that an islanded network is not energised which if it
were not true would pose a safety risk to personnel working on this network.

A surplus or deficit in generation capacity provided by the DER compared to the local load in the
islanded network determines the ease of detecting a LOM condition. When these are not matched
then voltage and frequency protection can be effectively used to detect LOM [8]. However, when
generation and local loading are closely matched then it is more difficult to detect the islanding
event. Therefore, more specialised protection functions are included. The most commonly used
functions are rate of change of frequency (ROCOF) and voltage vector shift (VS).
7.5 System integrity protection schemes
In addition to the protection against short circuits, there are schemes that are used to protect the
overall integrity of the power system against certain events that usually lead to unstable

23
transients, overloads or, in extreme cases, blackouts. These are called system integrity protection
schemes (SIPS) [36]. The actions performed by system integrity schemes are designed based on
extensive system studies. For example, frequency excursions lasting longer than a predefined
amount of time usually trigger generation or load disconnection as appropriate. Failing to do so
can result in loss of system synchronism. Similarly, excursions in voltage limits (usually voltage
depression) should be treated to avoid a system voltage collapse. This can be remedied by
managing power flows or switching of FACTS [37]. The advent of wide area measurements
promises more flexibility in available protection actions through the implementation of more
advanced SIPS functions.

7.6 Wide area measurement, protection, automation and control

7.6.1 Synchro phasor measurement technology
Collecting synchronised voltage measurements from remote busses was first discussed in. The
technology has since then developed significantly and currently relies on GPS (global positioning
system) as a universal source of synchronising signals. These signals are used by phasor
measurement units (PMU) to time stamp each measurement made for comparison at a later stage.
The operation of PMUs is described in standard IEEE C37.118. Synchro phasor measurement
technology (SMT) consists of a number of building blocks which provide data measurement,
collection, archiving and visualisation systems. SMT can be used in a range of applications,
mainly in system monitoring where it is usually referred to as a wide area measurement system
(WAMS).
A number of real-time protection and control applications based on PMU measurements have
been proposed. These, however, require further development and the appropriate infrastructure
put in place including suitable communications networks and algorithms. These are usually
referred to as wide area measurement protection and control systems (WAMPAC). Figure 2-9
depicts a typical WAMPAC architecture. An extended version of this architecture can be found in
, where WAMS can be utilised to perform adaptive protection functions to cope with variable
power system operational states.
GPSPMUPMUPMULAN/WANPDCStability Analysis State Estimation Energy Management
SystemPMU-assisted Short Circuit ProtectionSubstationPrimary Plant (e.g. breakers, FACTS)
Synchro phasors & status indications Protection & Control Signalling Trip
commandsMeasurementsPMU-assisted System Integrity Protection.
7.6.2 Protection applications of SMT
SMT is seen as an enabler for more advanced system integrity protection functions. The ability to
compare measurement from the wider network can enable greater flexibility and potentially more

24
selective protection operation. Below is a list of some SMT based protection functions proposed
in the literature:
 Predictive angular and voltage stability protection.

 Fault localisation and classification.

 Precise islanding detection.

 Adaptive load shedding.

 Real-time state measurement or estimation to enable further protection and control functions.

It is envisaged that SMT technology can enhance dependable and secure performance of SIPS. By
shifting the balance between these performance criteria when the system is normally loaded or
under stress respectively, undesirable operation can be avoided.

7.7 The digital substation

The introduction of microprocessor based protection and control devices has enabled the delivery
of more powerful and flexible functions. The term digital substation refers to the integration of
these devices over communications channels. Intelligent electronic devices (IEDs) are considered
the building block of digital substations.
7.7.1 Intelligent electronic devices
Relaying platforms have evolved from electromechanical based protection relays to
multifunctional numerical functions implemented on IEDs. The latter offers a wide range of
protection functionality within a single physical device along with more integration of monitoring
and control functions.

7.7.2 IED advantages over legacy relaying platforms

Greater flexibility in protection scheme deployment is achieved due to a potentially large number
of protection and automation functions that can be activated on any given IED. IEDs are based on
embedded platforms that constitute modular hardware components. This means that upgrading a
scheme’s I/O or hardware capabilities is a relatively straightforward task since complete hardware
replacement is not necessary. Upgrades to the functionality can also be achieved through
firmware upgrades.

25
 Chapter-8
Review of techniques to achieve adaptive protection functionality

As discussed in the introductory section of this chapter, the examination of adaptive protection
schemes proposed in the literature will be limited to those functions which aim to improve
scheme sensitivity or coordination. It should be emphasised that coordination in this case
encompasses ensuring the selective operation of the protection scheme. Other examples exist to
26
address protection stability and speed of operation, but the scope of the first set of examples best
matches the scope of the thesis and indeed the developed adaptive distance protection scheme.
The review will not delve into the problems that these adaptive schemes address as the need for
adaptive protection has already been established in the previous chapter. Nor will it discuss the
intricacies of the techniques used as this is out with the scope of the thesis. The schemes reviewed
fall under one of the following categories in terms of techniques used to achieve adaptive
functionality:
 Artificial neural networks (ANN): neural networks are used to recognise patterns in measured
power system quantities that reflect its prevailing operational state [3]. Based on this pattern
recognition, a decision can be made to alter the behaviour of the protection accordingly. An ANN
requires offline training to tune its functionality [4].

 Optimisation techniques: linear or heuristic optimisation techniques [5, 6] are used to achieve
an objective function which usually aims to minimise operating time or maximise selectivity
based on network conditions.

 Fuzzy logic: a fuzzy membership function is used to determine the primary system state [7]
which affects the setting of the adaptive protection.

 Data mining: offline analysis of power system data is used to classify its operation into states
(e.g. islanded and grid connected network). This analysis is then used to create a ‘decision-tree’
logic for online use to infer the state of the system and feed it into the adaptive protection
algorithm for decision making.
 Multi agent systems (MAS): Distributed ‘intelligent’ functions operate cooperatively over a
communications network to achieve a performance objective which is usually related to the
coordination of a multi-relay scheme.

 Expert systems: the expertise of a protection engineer or power system operator is captured in
an algorithm to provide decision making functionality. This aims to alter the protection behaviour
to deal with varied power system operating conditions, the same way an expert would do.

 Numerical techniques: these techniques usually implement an online form of protection setting
calculations that are normally performed offline. Other approaches involve the use of multiple
prioritised objectives to inform the operation of the protection functions.

 Hybrid techniques: these use a combination of two of the above techniques to adaptive the
protection behaviour.

8.1 Adaptive protection to improve scheme sensitivity

27
As discussed in the previous chapter there are situations when the pick-up threshold of the
protection is not sufficiently tuned to deal with varied power system conditions such as varied
fault levels, distinguishing between islanding and grid connected states or detecting resistive
faults. To this end, a number of adaptive protection schemes have been proposed to deal with
these situations.
The use of ANN has been proposed to detect resistive faults and alter the tripping logic of
distance protection schemes. Self-organising ANNs are used to alter the operating characteristic
of the distance protection such that a temporary extension that encompasses resistive fault
impedances is achieved. A common problem with ANN based adaptive schemes is that the
response is always specific to the training data set for the ANN. Therefore, the impact of
transducer errors or different fault impedances is not fully understood or catered for. Adaptive
scheme examples based on numerical techniques are plentiful. The underlying principle of
operation however remains largely similar – that is the use of power system measurements as
inputs to a short circuit calculation or state estimation which is then used to tune the protection
accordingly. In, generator infeed in a multi terminal transmission circuit are measured to alter the
reach of zone 2 distance protection. Short circuit simulations are performed offline calculate the
seen impedance by zone 2 under different infeed levels. Based on the results, the reach of zone 2
is minimised to avoid overreaching into adjacent lines. In, another numerical algorithm is
proposed to deal with the reach error effect mutual coupling has on distance protection. The
measured impedance is compensated based on the current flowing in the parallel circuit. A
number of states are defined according to the loading of the parallel circuit and fault impedance
locus. These determine the appropriate action for the distance protection scheme. An adaptive
load encroachment algorithm has been proposed in. Based on system wide measurements, the
relay at most risk of load encroachment is identified. This is used to apply an anti-encroachment
zone (AEZ). Then a simple binary logical operation is performed to combine the distance
characteristic and AEZ responses to a fault locus to produce the trip command.
Data mining is increasingly finding new applications in power systems. In, data mining is used as
a means to bias the operation of existing protection schemes towards dependability or security
according to power system conditions (i.e. normal or stressed operations). Data mining is used to
classify the power system measurements that reflect normal or stressed operation. A voting logic
based on a decision tree derived from the classification process is introduced between feeder
protection relays and the final trip command. Whenever significant changes are made to the
network the logic must be revised. So it is unclear how the logic would perform if the system
topology changed often or its dynamics change due to changes in connected generation.
8.2 Adaptive protection to improve scheme coordination

28
Coordination in a multi-relay scheme is an important issue when their behaviour is being altered
dynamically. There are a number of adaptive protection techniques that ensure coordination is
maintained. For instance, optimisation techniques have been used to this effect. In, particle swarm
optimisation (PSO) has been used to coordinate overcurrent protection relays using the grading
margins as a constraint. Several hundreds of iterations are usually required to reach convergence.
It is not clear whether the computing requirements and time to achieve convergence are suitable
for an online application of the method to adapt the overcurrent settings. Another optimisation
method based on genetic algorithm (GA) is proposed in to alter the operation of load shedding
schemes. The amount of load shed is reduced with each stage to minimise customer interruptions.
Numerical approaches have also been used to achieve adaptive protection. Both and, propose an
approach to load shedding using multiple criteria. A combination of bus voltage calculations and
rate of change of frequency measurements determine the load priority and speed of disconnection
to best serve the system stability. Although high-speed communications are a prerequisite for the
effective operation of the scheme, no specific requirements have been given. At the distribution
level, numerical approaches rely on calculating short circuit levels for a given network
configuration and applying the appropriate settings to the overcurrent relays as in.
The nature of MAS lends itself to address the coordination problem since it relies on peer to peer
communication to achieve an overall objective. A MAS based load shedding scheme has been
proposed in. Loads are continuously monitored and when the need for disconnection arises, the
scheme prioritises attempts to minimise the amount of load disconnected. Different agents take
the responsibility of system monitoring. In, overcurrent relays form part of a MAS to coordinate
the settings of overcurrent relays in the presence of DG. The literature, however, tends to focus on
the MAS architecture as opposed to the intricacies of the coordination algorithms. Finally, hybrid
techniques such as that proposed in rely on expert system and fuzzy logic to determine the faulted
section of a network. Initial operation of protection elements surrounding the fault is used to
define a search region. The hybrid approach is then used to zoom in on the faulty feeder.
8.3 Shaping the research direction for adaptive protection
It is clear that there is no shortage of advanced adaptive protection schemes proposed in the
literature that rely on intelligent systems techniques or otherwise to achieve improved
performance levels. But why has none of these or many others previously proposed have never
been deployed in the real power system despite a clear need for such improved performance and a
clear advantage offered by these proposals? In fact therein lies the problem with adaptive
protection. The problem of its applicability and not the problem of gaining an ‘X’ performance
enhancement for a ‘Y’ network that is operated under ‘Z’ conditions. Furthermore, with the
exception of MAS based schemes, there is a lack of consideration to the scalability and future

29
proofing of these schemes. What should be done when the scope of the adaptive scheme needs to
be expanded to encompass more relays? What if more DG has been added to the network? How
does the utility manage and maintain the settings of these schemes? More importantly, what if
these schemes do not rely on crisp settings in the conventional sense?
So it seems that the performance problems facing static protection (that is uncertainty and
variability in the power system) remain so with some of the advanced techniques being proposed.
Some of the performance problems are being effectively addressed. But the problem of being able
to cope with system variability keeps recurring even with the use of adaptive techniques. Perhaps
problem of producing a truly universal adaptive scheme is an unwieldy one. To this end, the
remainder of this thesis will not focus on adding to the pool of existing adaptive algorithms, but it
will attempt to understand the bigger picture issues related to adaptive protection, thus
establishing the scope, requirements and approaches that make adaptive protection a credible
solution.
8.4 Challenges to adopting adaptive protection
Adaptive protection can offer performance advantages compared with conventional schemes.
However, there are technical and institutional challenges that must be overcome beforehand. This
section discusses the main challenges associated with the adoption of an adaptive protection
philosophy. Three main challenges are discussed in this section.
8.4.1 Integration with existing protection arrangements
An overhaul of existing protection arrangement to accommodate adaptive functionality is not an
attractive option. This is especially the case with older substation installations. New substations
may be more suited for introducing adaptive functionality. That said, there must be an assessment
of the impact such introduction can have on the wider system protection. This includes the
requirement for coordination with adjacent circuits and/or other protection functions as well as
any requirement to exchange information with existing substation systems.
Communication standards such as IEC 61850 have created a vehicle to facilitate the
interoperability between protection functions from different vendors and potentially the
interchangeability of those functions. Similarly, adaptive protection functions to be integrated into
existing protection arrangements must adhere to interfaces offered by existing functions. As
discussed in section 4.2.1, inferring the system state by an adaptive scheme may require access to
remote signalling to determine primary plant status. If this information is to be made available
from the existing substation communication network (LAN), then the integration exercise
becomes easier. More issues lie in furnishing the adaptive protection functionality with signals
from remote substations. The absence of suitable communication infrastructure will then need to
be remedied. Factors affecting the choice of a suitable communications infrastructure include the

30
required reliability and timeliness of the information a communications link can offer. The
change of settings will take a finite time to achieve which will be added to the time delays caused
by communications overhead. This must be considered in the context of the application. the case
that critical circuits (e.g. major transmission corridors) will have more stringent information
timeliness requirements.
8.4.2 Adaptive scheme testing
Testing of protection schemes has always been an important aspect in a scheme’s lifecycle. The
testing of adaptive protection schemes is not different and in fact is of greater importance. This is
mainly due to the increased complexity in scheme behaviour due to the introduction of adaptive
functions. The behaviour of the scheme is perceived to be less deterministic especially with the
use of some of the advanced techniques described in section 4.3. Furthermore, adaptive protection
functions respond to an extended range of events. In other words, the testing of the scheme
functionality should not be limited to verifying the response of the protection elements to faults.
Tests must also verify the correct change of settings (or scheme configuration) in response to
changes in prevailing network conditions. Generally speaking, tests performed on adaptive
protection schemes must complement those defined in international standards (e.g. IEC 60255)
but not replace them. More emphasis will be on stimulating the adaptive protection logic rather
than verifying the performance of conventional protection elements. The following two chapters
discuss the requirements for testing and different testing methods for adaptive protection schemes
and demonstrated developed testing methodologies using the adaptive distance protection
algorithm.

Chapter-9
The Developed adaptive distance protection logic
31
The analysis focuses on the erroneous adaptive protection behaviour caused by an unsatisfactory
response of the dynamic setting selection function in response to a primary system stimulus. To
this end, reachability analysis is proposed to examine the possibility of reaching an undesired
response. This property is hereafter referred to as the safety of the adaptive protection logic. One
of the key requirements of conducting reachability analysis is defining the unsafe states that the
system must not reach or dwell in. Given that the adaptive logic may take a finite time to
determine the appropriate setting, dwelling in an unsafe state may be acceptable. This is the case
given that a maximum time delay for the unsafe state exit transition is specified. Note that it was
shown experimentally in chapter 5 that the implementation of the scheme has an impact on this
time delay which can be variable. Nevertheless, it is important to determine the acceptable
boundaries for this delay. Therefore, verifying this condition in the adaptive protection scheme
context requires the examination of unsafe states entry and exit during the adaptive logic
operation.
The performance invariant sets previously defined are used to identify these unsafe states . Where
denotes an unsafe state. In Figure 6-7, the performance state combining the ‘default setting’ and
‘QB boost’ states is considered unsafe since this particular combination results in distance
protection under reach. As mentioned previously, invariant sets are mutually exclusive. Thus, the
boundaries of the performance states can be clearly defined in the hybrid state space. Ultimately,
this will result in a clear (binary) indication of whether a particular state can be considered safe or
not.
Where is the ‘eventually’ logical operator. This temporal aspect reflects the finite amount of time
required to exit an unsafe state through adaptive protection setting changes. To formally examine
the temporal dimension from a hybrid system perspective, timed hybrid automata can be
considered. However, this is out of the scope of the thesis.
The backwards trajectory obtained from the unsafe transition can be used to identify faults in the
adaptive logic, by observing the scheme inputs and the resulting adaptive logic state transitions
leading to the unsafe state entry. In light of this, a safety performance verification procedure
based on reachability analysis is proposed and is shown in Figure 6-8. This will be used on the
adaptive distance protection scheme previously developed. While conducting the reachability
analysis as outlined in Figure 6-8, it is important to stimulate the system with inputs for each set
of initial conditions. It is worth noting in this case, that merely residing in an unsafe state does not
necessarily reflect that the system under test is unsafe. Since the adaptive protection logic takes a
finite amount of time to respond to changes in the network, this should always be taken into
account in the analysis.
The dotted region within Figure 6-8 indicates the use of , and defined earlier. The unsafe states

32
are mapped directly to these invariant sets as will be reflected in the implementation below.

Figure: Reachability analysis procedure

Chapter-10
33
 Design of adaptive protection schemes

It was shown that one of the most important design decisions made in developing an adaptive
protection scheme (from the point of view of scheme verification), is the amount of flexibility
that should be introduced by the scheme functionality. This manifested itself in the approach to
choosing the active protection setting. Choosing active protection settings from a limited pool of
pre-calculated settings groups is more desirable from a scheme validation point as opposed to
using online settings calculations. The number of settings groups necessary is highly dependent
on the protection scheme and application. Generally speaking, the trade-off between the
simplicity and flexibility provided by a limited or large number of settings groups respectively is
determined by the increase in risk of continuous settings changes due to a higher resolution
provided by a larger number of settings groups and the additional risk of scheme failure because
of this. Moreover, the more settings groups used, the less distinct the approach becomes
compared to online calculations, which diminishes its validation advantage.
The adaptive distance protection scheme developed in this thesis was limited by the maximum
number of settings groups provided by the relaying platform – four in this case. Theoretically
however, this can be expanded to provide a dedicated settings group for each mode of the QB and
associated tap positions – that is up to 41 settings groups for a typical QB. The effort in validating
this amount of settings groups becomes magnified, where in reality adjusting distance zone reach
is always governed by error margins of about +/-5% of reach setting based on engineering
practices and testing standards. Thus, a theoretical upper limit of settings groups can be applied
based on this error margin. In this particular case, 10 settings groups have been shown to be
sufficient.
The operational scope of adaptive protection functionality was defined, for the first time in this
thesis, to ensure valid performance at the design stage. This was achieved by designating the roles
of adaptive and conventional functions. It was shown that adaptive protection functions are more
suited to performing tasks triggered by non-fault events such as system reconfiguration.
Conversely, tasks performed during a fault transient should be preserved for conventional
protection functions that have been configured by adaptive functionality beforehand – again
appropriately when triggered.

Structural and behavioural modelling of adaptive protection schemes

The concept of an adaptive protection architecture (APA) was further developed through the
definition of minimum functional elements and interfaces for each of its layers. This has been
shown to achieve, for the first time, the following:
34
 Ensured the architecture’s applicability to transmission level adaptive protection applications
and consequently a wider application domain than the original conceived conceptual applications.

 Enabled the implementation and experimental testing of an adaptive protection scheme

(adaptive distance protection developed in the thesis) based on the architecture.

 Facilitated adaptive scheme validation by providing a reference functional and performance

specification that is independent of the scheme implementation. Furthermore, this development of
the architecture concept enabled creating a clear distinction between what constitutes coordination
and management layer functions. This distinction has been shown to be based on three criteria – the
nature of information used to infer system state, protection operation time frames and the breadth of
protection actions each layer exerts on another.
To facilitate the performance verification of adaptive protection algorithms, a novel approach based
on hybrid systems was used to describe its behaviour – that is a characterisation of its response to
measured events in the protected system. The application of hybrid systems modelling to adaptive
protection schemes is the first of its kind.
It has been shown that standard approaches to abstracting the behaviour of hybrid systems (discrete
event system abstractions) had limitations in encompassing the hierarchical ‘control’ nature of
adaptive protection based on the APA. The work reported in this thesis has been shown to overcome
this limitation by extending the definition of these behaviour abstractions. This necessarily required
the definition of additional interfaces and interactions between constituent elements of the behaviour
model.

Validation and verification of adaptive protection schemes

The adaptive distance protection scheme developed and implemented in this thesis was shown, using
hardware in the loop validation, to provide an improvement in selective reach of up to 20% of
protected line impedance for zone 2 – an improvement limited by coordination with adjacent line
protection zones. Thus, performance of backup protection is restored dynamically based on the state
of the QB transformer and circuit being protected.
The functional abstraction of the APA enabled performing unit testing of constituent components of
the developed adaptive distance scheme as well as the validation of the overall scheme more
effectively. This is due to the ability to define more clearly the functional and non-functional
requirements for the adaptive schemes components and the expected overall scheme performance
under varied operating conditions of the QB transformer. In other words the scheme requirements
were more traceable and as such more confidence can be obtained from the adaptive scheme
validation process.
This thesis reported the first application of reachability analysis (based on the hybrid behavioural
model) as a means of verifying the performance of adaptive protection schemes. The reachability
35
analysis methodology presented in this thesis was demonstrated through verifying the safety property
of the adaptive settings selection logic for the developed distance protection scheme. The reachability
analysis methodology reported in this thesis is also novel in the approach to defining the boundaries
of the reachable state space representing unsafe adaptive performance.
It has been demonstrated that the approach to defining the unsafe state space eliminated the need for
explicitly computing the continuous state space of the underlying hybrid behavioural model provided
that:
 The continuous evolution of the primary system states are not directly influenced by the outcome
of the adaptive setting selection logic.

 The adaptive logic forms part of the coordination layer functionality.

As such efficiencies in the verification process are gained. The approach to defining the state space is
based on splitting the hybrid system state space into ‘operational’ and ‘performance’ invariant sets
through which a direct mapping between power system conditions and scheme performance can be
made. This process necessitated defining the state transitions (and guard conditions) associated with
these invariant sets.
The adaptive scheme verification through reachability analysis has been shown to incorporate a
temporal aspect that reflects the adaptive scheme’s finite response time to changes in the power
system. Design and implementation measures can be put in place to minimise this time delay. The
required improvements in the time response are dictated by the acceptable duration of power system
vulnerability caused by the temporary degradation of protection performance levels. To better
characterise this temporal aspect, it is then necessary to use timed automata to model the hybrid state
space.
Testing adaptive protection schemes must generally incorporate the full complement of simulation
and formal testing methods. This is necessary to address inherent limitations of simulation based
testing stemming from the design of the testing scenarios. Therefore, testing approaches stipulated by
testing standards still hold but should be extended with approaches developed in the thesis to
incorporate the characteristics of interactions between conventional and adaptive functions. More
emphasis should be placed on formal testing as it becomes more difficult to test the full set of
potential operating scenarios in the field during scheme commissioning – a task that is even more
difficult when the schemes perform wide area protection functions.

Future work
Interfaces and functions within the adaptive protection architecture have been defined. However,
this can be taken a step further by using standard ways of describing and implementing the
architecture. For instance, IEC 61499 event driven function blocks can be used to represent the

36
architecture’s constituent functions as standard executable elements. The use of standard
interfaces simplifies the process of porting these functions into different platforms. Further
research is necessary to determine methods of distributing adaptive functionality. And the use of
IEC 61499 enables such an approach as it supports describing distributed control functionality.
A full suite of formal performance verification techniques can be applied with the aid of the
behavioural model. These include determinism and observability. Knowledge of the current state
of the adaptive logic and stimulating inputs, results in knowledge of its output if it is
deterministic. Also, the state of the adaptive scheme can be identified by observing its inputs and
outputs in relation to the behavioural model. The significance of determining these properties lies
in offering complementary methods that can be used to verify the adaptive protection functions’
performance.
Making full use of management layer functions necessitates identifying system integrity
protection schemes (SIPS) that would benefit from adaptation in their performance. Moreover,
techniques of establishing the system state and its impact on system protection performance must
be developed. Changing the configuration of system protection, in this case, requires greater
levels of coordination to avoid conflict in performance objectives between coexisting system
protection schemes. The developed hybrid system model will prove to be a powerful approach to
understanding this complex problem and reachability analysis is one of the tools that should be
used to determine the safety of these interactions.

37
 CONCLUSION

This paper presents the protection solutions for overall and modern power systems. Zone
protection can be used for any type of electrical grid and it’s the basic protection of all. Some of
the other protection which are commonly used also mentioned in this paper. When designing an
electrical system reliability, selectivity, speed and sensitivity are the key points to have a good
performance. It is imperative to use the correct protection type to ensure that selected protection
can detect and isolate all faulted portions of the electrical system. To ensure the safe operation
of the power system, the protection equipment should be updated accordingly.
Adaptive protection scheme is taken as the subject matter and explained the overall structure of it.
Software and functioning of adaptive protection system are explained in detail as the
recommendations for the future power system.

38
 REFERENCES

"IEEE Recommended Practice for Protection and Coordination of Industrial and Commercial Power
Systems - IEEE Explore Document", Ieeexplore.ieee.org, 2017.
Energetikaelektro.files.wordpress.com. [Online]. Available:
https://energetikaelektro.files.wordpress.com/2015/01/adaptive- relaying.pdf.
Zhongwei Li, Weiming Tong and Fengge Li, "Study on Adaptive Protection System of Power Supply and
Distribution Line - IEEE Conference Publication", Ieeexplore.ieee.org, 2006.
A.Y.Abedlaziz, H.E.A.Talaat, A.I.Nosseit, and et al, "An adaptive protection scheme for optimal
coordination of overcurrent relays," Electric Power Research, vol. 61, pp. 1-9, 2002.
"C37.95-2014 - IEEE Guide for Protective Relaying of Utility- Consumer Interconnections - IEEE
Standard", Ieeexplore.ieee.org, 2014.
C37.95-2002 - IEEE Guide for Protective Relaying of Utility- Consumer Interconnections - IEEE
Standard", Ieeexplore.ieee.org, 2003.
C37.95-1989 - IEEE Guide for Protective Relaying of Utility- Consumer Interconnections - IEEE
Standard", Ieeexplore.ieee.org, 1999.
C. Chandraratne, T. Logenthiran and R.T. Naayagi, "Smart Grid Protection through Self-Healing", ISGT-
ASIA 2017.
"Distribution System Feeder Overcurrent Protection", Apps.geindustrial.com, 2017.
[Online]. Available: http://apps.geindustrial.com/publibrary/checkout/GET- 6450?
TNR=White%20Papers%7CGET-6450%7Cgeneric.
J. C. Das, "Overcurrent Coordination", Wiley-IEEE Press 2012.
Distribution System Feeder Overcurrent Protection, GE Power management report-GET-6450, 1-21.
Michael Stanbury, Zarko Djekic, “The Impact of Current- Transformer Saturation on Transformer
Differential Protection”, IEEE Transactions on Power Delivery, vol.30, no.3, pp. 1278- 1287, 2015.
"C37.113-2015 - IEEE Guide for Protective Relay Applications to Transmission Lines - IEEE Standard",
Ieeexplore.ieee.org, 2016.
Jinju Zhou, Lina He, Yijia Cao, Xubin Liu, and Yinghui Geng, "IEEE Xplore document - what’s the
difference between traditional power grid and smart grid? — from dispatching perspective,"
M. Khosrow-Pour, Renewable and alternative energy. IGI Global, 2016.
Ayaz Ahmad and Naveed Ul Hassan, “Smart Grid as a Solution for Renewable and Efficient
Energy”, April 2016.
S. Shandilya, Handbook of research on emerging technologies for electrical power planning, analysis,
and optimization. IGI Global, 2016.
J. Zhang and Y. Dong. “Preventing false trips of zone 3 protection relays in smart grid”. Tsinghua Science
and Technology, 20(2):142–154, 2015.
E. K. Ismail and T. K. Sindhu, "Self-healing practice for an LT
11 kV overhead radial distribution network feeder - IEEE Conference Publication", Ieeexplore.ieee.org,
2013.
J. Gao, Y. Xiao, J. Liu, W. Liang, and C. L. P. Chen, “A survey of communication/networking in smart
grids”, Future Generation Computer Systems, vol. 28, no. 2, pp. 391–404, 20

Q. Wu, Z. Lu and T. Ji, Protective relaying of power systems using mathematical morphology. Inglaterra:
Springer - Verlag London Limited, 2010.
Jingjing Zhang and Ming Ding, "Application of adaptive theory in distribution network protective relaying
- IEEE Conference Publication", Ieeexplore.ieee.org, 2009.
S. Horowitz and A. Phadke, Power system relaying, 4th ed. Chichester: Wiley, 2015.
S. Voima, H. Laaksonen and K. Kauhaniemi, "Adaptive protection scheme for smart grids - IET
Conference Publication", Ieeexplore.ieee.org, 2017.
D. Somasundaram, S. Kumar and M. Chandran, "An Overview on Adaptive Protection in Power System",
International Journal of Advanced Research in Electrical, Electronics and Instrumentation Engineering,
2014.
H. Song, S. Bregni, R. Srinivasan, T. Sookoor and S. Jeschke,
Smart Cities. Somerset: John Wiley & Sons, Incorporated, 2017.
J.D. Codling, S.A. House and J.H. Joice, "Adaptive relaying. A new direction in power system protection -
IEEE Journals & Magazine", Ieeexplore.ieee.org, 1996.
39
M. Geidl, "Protection of power systems with distributed generation", Research-collection.ethz.ch, 2017.
E. Orduna, F. Garces, and E. Handschin, ―Algorithim- Knowledge- Based Adaptive Coordination in
Transmission Protection‖, IEEE Transaction 2003.
Li Yinhang, Shi Dongyuan, and Duan Xianzhong, ―An Integrated Power System Relay Coordination
Software‖,IEEE Transaction 2001.
B. Chattopadhyay, M.S. Sachdev, and T.S. Sidhu. Adaptive relaying for protecting a distribution system -
a feasibility study. In WESCANEX '91 IEEE Western Canada Conference on Computer, Power and
Communications Systems in a Rural Environment, pages 20-25, 1991.
C. W. So and K. K. Li, ― Protection Relay Coordination on Ring- fed Distribution Network with
Distributed Generations‖, IEEE Transaction 2002.

40