Power Transmission

System Analysis Against

Faults and Attacks
 Power Transmission
System Analysis Against
Faults and Attacks

Tamalika Chowdhury, Abhijit Chakrabarti, and

Chandan Kumar Chanda
First edition published 2021

by CRC Press
6000 Broken Sound Parkway NW, Suite 300, Boca Raton, FL 33487-2742

and by CRC Press

2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN

© 2021 Taylor & Francis Group, LLC

CRC Press is an imprint of Taylor & Francis Group, LLC

The right of Tamalika Chowdhury, Abhijit Chakrabarti, and Chandan Kumar Chanda to be identified
as the authors of the editorial material, and of the authors for their individual chapters, has been
asserted in accordance with Sections 77 and 78 of the Copyright, Designs and Patents Act 1988.

Reasonable efforts have been made to publish reliable data and information, but the author and
publisher cannot assume responsibility for the validity of all materials or the consequences of
their use. The authors and publishers have attempted to trace the copyright holders of all material
reproduced in this publication and apologize to copyright holders if permission to publish in this
form has not been obtained. If any copyright material has not been acknowledged please write and
let us know so we may rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced,
transmitted, or utilized in any form by any electronic, mechanical, or other means, now known
or hereafter invented, including photocopying, microfilming, and recording, or in any information
storage or retrieval system, without written permission from the publishers.

For permission to photocopy or use material electronically from this work, access www.copyright.
com or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA
01923, 978-750-8400. For works that are not available on CCC please contact mpkbookspermissions@
tandf.co.uk.

Trademark notice: Product or corporate names may be trademarks or registered trademarks and are
used only for identification and explanation without intent to infringe.

ISBN: 978-0-367-49777-4 (hbk)

ISBN: 978-0-367-49954-9 (pbk)
ISBN: 978-1-003-04733-9 (ebk)

Typeset in Times
by codeMantra

Visit the CRC Press Web site at https://www.routledge.com/Power-Transmission-System-Analysis-

Against-Faults-and-Attacks/Chowdhury-Chakrabarti-Chanda/p/book/9780367497774
To my husband and parents
–Tamalika Chowdhury
To my spouse
–Abhijit Chakrabarti
To my mother and grandson
–Chandan Kumar Chanda
Contents
Foreword ...................................................................................................................xi
Preface.................................................................................................................... xiii
Acknowledgments..................................................................................................xvii
Authors....................................................................................................................xix

Chapter 1 Introduction ..........................................................................................1

1.1 Background and Motivation ......................................................1
1.2 Growth of Modern Large-Scale Power Systems .......................2
1.3 Complexity in Power Systems ...................................................5
1.3.1 Complexity and Complex Systems...............................5
1.3.1.1 Regular Networks .........................................7
1.3.1.2 Random Networks ........................................7
1.3.1.3 Small-World Networks..................................7
1.3.1.4 Scale-Free Networks.....................................7
1.3.2 Power Grid as a Complex System ................................8
1.4 Transmission System Faults ......................................................9
1.5 Conventional Contingencies in Power Transmission
Network................................................................................ 9
1.6 Threats and their Consequence in Power System Operation.....10
1.7 Modelling of Power Transmission Network as Complex
Network ....................................................................................11
1.8 Structural Property Analysis of Power Transmission
Network...................................................................................18
1.9 Complex Network Approach to Vulnerability Assessment
of Power Transmission Network...............................................19
1.10 Cyber Security and Vulnerability Towards Cascaded
Collapse ................................................................................... 20
References .......................................................................................... 21

Chapter 2 Traditional Approach in Analysis of Faults in Power System............25

2.1 Introduction .............................................................................25
2.2 Formation of [Ybus] Matrix.......................................................25
2.3 Formation of [Ybus] with Line Transformers Present ............... 30
2.4 Concept of Symmetrical Component Analysis ....................... 37
2.5 Operator “a” ............................................................................ 38
2.6 Symmetrical Component Representation................................ 39
2.6.1 Determination of Sequence Components...................40
2.7 Concept of Sequence Impedances........................................... 42
2.8 Sequence Component Models of Principal Power System
Equipment.................................................................................42
2.8.1 Generator Sequence Model .........................................42
vii
viii Contents

2.8.2  equence Network Model of Transformers................. 42

S
2.8.3 Sequence Impedance Circuit for Transmission
Lines������������������������������������������������������������������������ 45
2.8.4 Sequence Impedance Circuit for Motors��������������������45
2.9 Three-Phase Power in Terms of Symmetrical Components�����45
2.10 [Zbus] Building (Step-by-Step Method)������������������������������������ 50
2.10.1 Adding a Branch (or Link) Zb from a New Bus to
the Reference Bus (Type 1 Modification)������������������� 50
2.10.2 Addition of a Branch (or Link) Zb from a
New Bus to an Old Bus (Type 2 Modification)�����������52
2.10.3 Addition of a Branch (or Link) Zb from an Old
Bus to the Reference Bus (Type 3 Modification)��������53
2.10.4 Addition of a Branch (or Link) Zb between Two
Old Buses (Type 4 Modification)������������������������������� 54
2.10.5 Addition of Two Branches Z a and Zb with
Mutual Impedance (Z m) between Four Buses
(Type 5 Modification)������������������������������������������������� 56
2.11 Algorithm for Formation of Bus Impedance Matrix [Zbus]
using Step-by-Step Method������������������������������������������������������58
2.12 Determination of Symmetrical Fault Current Using [Zbus]
Inversion�����������������������������������������������������������������������������������61
2.13 Phase-Sequence Component Network�������������������������������������62
2.13.1 Three-Phase Balanced Fault�������������������������������������� 65
2.13.2 Single Line to Ground Fault��������������������������������������� 66
2.13.3 Line-to-Line Fault������������������������������������������������������ 68
2.13.4 Double Line-to-Ground (L-L-G) Fault�����������������������70
2.13.5 Sequence Currents in the Interconnecting Line
between the Faulted Bus i and Healthy Bus k�������������73
2.14 Severity of Fault Currents and the Effect of Neutral
Grounding Reactance�������������������������������������������������������������� 80
2.15 Open Conductor Fault��������������������������������������������������������������81
2.15.1 Single Conductor Open Fault������������������������������������� 82
2.15.2 Two Conductor Open Fault���������������������������������������� 83
Reference�������������������������������������������������������������������������������������������� 84

Chapter 3 Contingency Analysis������������������������������������������������������������������������� 85

3.1 I ntroduction����������������������������������������������������������������������������� 85
3.2 Relationship between Thevenin’s Theorem and Bus
Impedance Matrix [ Z BUS ]������������������������������������������������������� 85
3.3 Addition and Removal of a Line in a Power System Network��� 90
3.4 Current Distribution Factor (Ψi ) and Line Outage
Factor (λ )������������������������������������������������������������������������������ 95
3.5 Single Line Contingency��������������������������������������������������������� 97
3.6 Contingency Analysis of Interconnectors����������������������������� 100
3.7 Contingency Analysis Employing DC Power Flow Model������103
Contents ix

3.8 Concept of Equivalencing in Power System Network........... 104

Reference.......................................................................................... 107

Chapter 4 Fundamental Concepts of Complex Network Theory ..................... 109

4.1 Introduction ........................................................................... 109
4.2 Pure Topological Approach ................................................... 110
4.2.1 Geodesic Distance.................................................... 110
4.2.2 Average Shortest Path Length .................................. 110
4.2.3 Degree (Connectivity) .............................................. 111
4.2.4 Distance and Effciency............................................ 111
4.2.5 Electrical Distance ................................................... 112
4.2.6 Electrical Centrality and Its Measurement............... 112
4.2.7 Betweenness ............................................................. 114
4.3 Extended Topological Approach ........................................... 114
4.3.1 Electrical Betweenness ............................................ 116
4.3.2 Vulnerability Using Global Effciency..................... 117
4.3.3 Vulnerability Using Netability ................................. 118
4.3.4 Criticality Assessment Using [Zbus] Centrality......... 119
References ........................................................................................ 120

Chapter 5 Vulnerability Assessment of Power Transmission Network ............ 123

5.1 Introduction ........................................................................... 123
5.2 Vulnerability Indices: Tools of Vulnerability Assessment ..... 124
5.2.1 Grid Vulnerability Index (GVI) ............................... 125
5.2.2 Effciency-Based Vulnerability Index ...................... 126
5.3 Aspects of Vulnerability Analysis of Power Transmission
Network ..................................................................................126
5.4 Simulation for Vulnerability Indices in Power
Transmission Network............................................................127
5.4.1 Using Global Effciency and Netability Method.......127
5.4.2 Using Betweenness and Netability Metric................130
References ........................................................................................ 134

Chapter 6 Analysis of Cascading Failure and Islanding in Grid Network ........135

6.1 Introduction ............................................................................135
6.1.1 Fundamental Modelling Approach in Islanding .......137
6.2 Concept of Islanding...............................................................138
6.3 Modelling of Cascading Failure and Islanding ......................139
6.4 Application of Preferential Probability in Modelling of
Cascading Failures .................................................................139
6.5 Simulation...............................................................................141
6.6 Summary ................................................................................146
References ........................................................................................ 146
x Contents

Chapter 7 Assessment of Resilience in Power Transmission Network��������������149

7.1 I ntroduction����������������������������������������������������������������������������149
7.2 Concepts of Resiliency, Robustness, Reliability and Stability��149
7.2.1 Definitions of Resiliency�������������������������������������������153
7.2.2 Framework of Resiliency������������������������������������������154
7.3 Relationship between Resiliency and Percolation Threshold���156
7.4 Power System Behaviour at Percolation Threshold����������������157
7.5 Assessment of Resiliency for Transmission Network������������160
7.6 Simulation�������������������������������������������������������������������������������161
7.7 Summary��������������������������������������������������������������������������������166
References���������������������������������������������������������������������������������������� 166

Chapter 8 Effect of Distributed Energy Sources�����������������������������������������������169

8.1 I ntroduction����������������������������������������������������������������������������169
8.1.1 Solar Photovoltaic Energy�����������������������������������������169
8.1.2 Wind Energy�������������������������������������������������������������172
8.1.3 Small Hydro Power���������������������������������������������������174
8.1.4 Biomass and Biogas Energy��������������������������������������174
8.1.5 Tidal Energy��������������������������������������������������������������175
8.1.6 Ocean Thermal Energy Conversion (OTEC)������������176
8.1.7 Geothermal Energy���������������������������������������������������176
8.1.8 MHD Generation�������������������������������������������������������177
8.1.9 Energy Storage����������������������������������������������������������177
8.2 Role of Distributed Energy Sources in Power
Transmission Networks����������������������������������������������������������178
8.2.1 DGs in Restructured Environment����������������������������178
8.2.2 Applications of DGs��������������������������������������������������179
8.3 Mitigation Strategy Employing Distributed Energy Sources�����180
8.3.1 Network Criticality Assessment Using [Zbus]
Centrality Employing DGs����������������������������������������180
8.3.2 Incorporation of Distributed Energy Sources
in Criticality Assessment Using Betweenness
Metric: Simulation for DG Incorporation in
Criticality Assessment of IEEE 57 Bus System
Using Betweenness Metric����������������������������������������180
8.3.3 Vulnerability Assessment Using GVI Metric
Incorporating DER: Simulation for Vulnerability
Assessment of IEEE 57 Bus System Using GVI
Metric on Incorporation of DG���������������������������������183
References���������������������������������������������������������������������������������������� 186

Appendix A: IEEE 57-BUS TEST SYSTEM (American Electric Power)�������189

Appendix B: West Bengal State Electricity Board’s (WBSEB) 203-Bus
System (A Semi-Government Power Utility in Eastern Part of India)������������195
Index�����������������������������������������������������������������������������������������������������������������������219
Foreword
Dynamic and diverse disruptions are continuously challenging the functionality of
power grid. Modern communication technologies lead to new hazards of attacks and
outages in the past few years, which need to be addressed with the integration of
distributed energy resources. This requires the identifcation of critical buses and
lines in power grid networks that need to be protected and handled for vulnerabil-
ity assessment. The vulnerability assessment process involves mathematical analysis
employing different tools, metrics and indices of complex network theory.
Further, it is important that power grid networks recover after an outage to minimise
the economic, social and political impacts. Resilience evaluates response and recov-
ery processes. An assessment of resilience performance analysis is attempted consid-
ering statistical concepts of percolation threshold that evolved around the response
of the grid against any kind of disruption. A framework for assessment of cascading
failure has been exploited, which involves preferential attachment model to deter-
mine the consequence of grid network islanding.
Modern power system security is prone to unplanned outages due to natural
calamity and cyber threats that may lead to destructive consequences for the power
system network. In traditional concepts of power system security, researches were
conducted in restoring the power network and maintaining the operation of the sys-
tem against electrical contingencies. However, with time and increasing complexity
in power network, the complex power network has become susceptible to unplanned
outages due to cyber-attacks and physical failures of key elements of the power
system that includes grid system.

xi
Preface
Large-scale disruptions of the power grid have occurred several times across the globe
confrming the existence of inherent vulnerabilities. The grid network of Switzerland
suffered a major setback due to electrical failures on 28th September 2003. North
America also faced a major blackout in 2003. Such failures affect millions of people
and cause losses of billions of dollars. Similar incidents were reported in Canada in
1998, in Europe in 2006, in United States due to Katrina, in 2007, etc. On 30th and
31st July 2012, the failure of the Indian grid affected 680 million people and loss of
billions of dollars.
To minimise the economic, social and political impacts of power system outages,
the grid must be suffciently strong to face the vulnerability. The response of the grid
against such diverse disruptions is a matter of concern. Indeed, it is clear that after
decades of human and technological growth, problems are inevitable; however, it
is important is respond to the problems. These challenges motivate research on the
aspects of vulnerability analysis of power system networks.
To assist in developing tools or metrics that could be used for early identifcation
of the zones of possible unplanned failure or attack, this book attempts to develop
methodologies to identify the vulnerable elements, vulnerability of the network and
its resiliency while encountering natural calamities or attacks at its key elements.
The modelling of power system as a complex network offers a framework to anal-
yse the response of the grid system and distribution system while the network is
subjected to the threat of failure and even the threat of cascading collapse. To achieve
this target, different mathematical and computational techniques in line with the
techniques used in complex network theory and some aspects of the concepts of
probability have been utilised in this book.
Chapter 1 introduces the vulnerability analysis of power grids from a complex
network perspective. In this chapter, the background of the work is described as
the missing link between previous studies and how these details make the study of
vulnerability analysis of power grids much more useful. A brief history of complex
networks under this research domain are also presented. Moreover, how power grid
is characterised with complex features is discussed in the chapter.
In Chapter 2, the concepts of bus admittance matrix and bus impedance matrix are
discussed in detail for assessing the vulnerability of the transmission system employ-
ing complex network theory. In this chapter, traditional fault analysis approach in
power system using symmetrical components has been utilised. Formation of Ybus
matrix as well as Zbus matrix are explained in the chapter followed by determination
of symmetrical fault current using [Zbus] inversion. Three phase balanced fault, Line
to ground (L-G) fault, Line to line (L-L) fault, Double line to ground (L-L-G) and
Open Conductor faults are analysed using [Zbus]. Line currents during fault condition
and open conductor faults are also determine.
In Chapter 3, the concept of contingency plays an important role in planning an
islanding condition during emergency. Post-switching condition (transient state) of
any component in power transmission system, any overload or overvoltage conditions

xiii
xiv Preface

are checked to predict the steady bus voltage and line currents. Hence, contingency
analysis is necessary to assess every step after addition or removal of lines along
with the use of Zbus building algorithm. This chapter depicts cases of single as well
as multiple contingencies. The concept of equivalencing of more than one network
area by using tie lines and, consequently, inter-tie contingency analysis is described.
In Chapter 4, a comprehensive description of the parameters in complex net-
work theory is presented employing a pure topological approach. The concept of
incorporating specifc electrical features of power systems is introduced in model-
ling of a power network as a complex network. Applying the mathematical concepts
of complex network theory, an extended topological methodology (betweenness,
vulnerability using global effciency and netability) is discussed, which involves
complex network theory and electrical features of power systems. Criticality assess-
ment employing electrical centrality is also discussed, and subsequently, concepts of
vulnerability indices are introduced.
In Chapter 5, vulnerability and criticality of power networks are assessed in
grid-connected power system. Commonly used power system operating param-
eters have been utilised to rank critical lines based on the magnitudes of electrical
betweenness. In the simulation, physical parameters of the power system that includes
line fow, power limits and transmission capabilities are incorporated to analyse the
vulnerability of the power network after obtaining the concept of netability using
complex network theory. A new index called Grid Vulnerability Index (GVI) has
been introduced in the context of advancing system performance.
Chapter 6 details the analysis of cascading failure and islanding in grid networks.
In this chapter, power systems are modelled for unplanned outages and attacks lead-
ing to cascading failure. The electrical betweenness of transmission lines and/or
nodes has been utilised to ascertain the criticality of transmission line(s) that might
pose a potential threat of outage.
Considering the concept of preferential load distribution among adjacent lines,
the simulation examines whether any line exceeds its ultimate capacity and becomes
an overloaded line prone to cascading failure, subsequent to dynamic redistribution
of load. The evolution of probabilistic tools along with their correlation with com-
plex parameters illustrates their strong infuence in modelling cascading failures.
Cascading failure model is analysed using preferential probability ultimately leading
to system islanding.
In Chapter 7, the immunity of a power network against unplanned or designated
vulnerable attacks is defned. The concepts of complex network theory, integrated
with statistical methods, are employed to determine the criticality of damage to the
nodes and edges of topographical model of an electrical power system encounter-
ing a vulnerable attack on the power system. The capability of the network to resist
fragmentation and mitigate and overcome stresses and failures are characterised by
the resiliency of the network. Betweenness centrality of an electrical power network
being one of the prime parameters of assessment for vulnerability of that network,
computations have been conducted to obtain the moments of degree distribution of
the network. This chapter highlights the quantifcation of critical fraction of dam-
aged nodes as well as the percolation threshold of the power network and the char-
acterisation of resiliency of a typical electrical power system following a routine
Preface xv

electrical failure, outage or designated attack on that electrical power system. This
chapter assesses metrics that resist and mitigate stresses and failures of the power
system to ensure network survival in case of extreme events and natural or man-made
disasters. Probabilistic approach has been applied to determine the loss of connectiv-
ity of the networks, and hence assessment of resiliency metrics in both transmission
and distribution networks.
In Chapter 8, distributed generation is defined and described in detail includ-
ing the most widely used renewable energy sources. Solar, wind, small hydro and
tidal energy sources are discussed. The chapter explains how distributed renewable
energy sources that are a part of the power system are a better alternative. Distributed
generation has been applied to develop an effective mitigation strategy to enhance
the robustness of the distribution system. In this process, simulations are conducted
to obtain an optimal size and placement position for the distributed generation
in the complex power network using distributed energy resources. The impact of
­distributed renewable energy sources on the power loss and voltage profile of sub-
transmission network at different penetration levels are worth investigating in this
chapter. This chapter shows how application of distributed renewable energy sources
can reduce the criticality of the entire network. The benefit of installing distributed
generation at the critical nodes (buses) of the same electrical power network has also
been discussed.
The authors have attempted to present the topics covered in this book in a lucid
language. Any constructive criticism and/or suggestion to further enhance the knowl-
edge related to the topics presented will be highly appreciated.

Tamalika Chowdhury
Abhijit Chakrabarti
Chandan Kumar Chanda

Kolkata, India
Acknowledgments
The authors acknowledge the motivation and support given by the management of
IIEST Shibpur and Jain College of Engineering, Belgaum, Karnataka, India. The
authors also express their sincere thanks to Dr. Debraj Sarkar for his motivation
and support in preparing the book. Dr. Tamalika Chowdhury further acknowl-
edges the blessings and love of her parents Prof. (Rtd.) Tapan Kumar Chowdhury
and Prof. Alpana Chowdhury. The authors fnally mention the criticism and queries
raised by their students and scholars while they teach the courses related to this topic.
These criticisms and queries helped the authors to re-orient the mode of presentation
of the topics covered in this book.

Tamalika Chowdhury
Abhijit Chakrabarti
Chandan Kumar Chanda

xvii
Authors
Tamalika Chowdhury Ph.D. (Engg.) is an assistant professor in Jain College of
Engineering, Belagavi. She has been a research scholar in the Department of Electrical
Engineering, Indian Institute of Engineering Science & Technology (IIEST), Shibpur
(formerly BE College Shibpur) Howrah, West Bengal, India. She pursued her Ph.D.
program under UGC-BSR Fellowship. She received her B.Tech and M.Tech degree
in electrical engineering from West Bengal University of Technology (WBUT). Her
research interests include power grid, power system vulnerability, manifestation of
renewable energy-based distributed generation in power system and assessment of
power grid resilience using complex network theory.

Abhijit Chakrabarti Ph.D. (Tech) is a professor of electrical engineering in

IIEST, Shibpur. He is the former Head of the Department, Electrical Engineering,
IIEST Shibpur and former Vice-Chancellor of Jadavpur University. He is also
the former Vice Chairman of West Bengal State Council of Higher Education,
Government of West Bengal. He has published 132 research papers in peer-
reviewed international and national journals as well as in conferences. He has
authored 17 books in electrical engineering. He is a member of EEAC of National
Board of Accreditation, Government of India, as well as a visiting member of IET,
UK in accreditation process of ADAMS, UK. He has visited foreign countries
like Japan, United States and UK. He is a fellow of the Institution of Engineers,
India and member of IET, UK. Professor Chakrabarti has received many awards
and medals for his research in the feld of power system engineering. He is a
member of different boards and committees in universities in India and an expert
member in numerous technical and standing committees. He has an active inter-
est in research related to power system engineering, electrical machines, power
electronics and electric circuits.

Chandan Kumar Chanda is working as a professor in the Department of Electrical

Engineering, IIEST, Shibpur, India. He earned his Ph.D. degree from the Department
of Electrical Engineering, B.E. College (DU), Shibpur, India with specialisation
in power systems. Dr. Chandan Kumar Chanda has over 31 years of teaching and
research experience in the diverse feld of power systems engineering and almost
5 years of experience in industry. His areas of interest include smart grid, resiliency,
stability and renewable energy. He is a recipient of Tata Rao Gold Medal. He is
actively involved in various research projects funded by centrally funded organisa-
tions like DST and UGC. He has published 150 research articles in reputed national/
international journals and conferences including 45 research papers in SCI/SCOPUS-
indexed journals. He is a member of the editorial board and a guest editor of numer-
ous reputed Journals. He has authored and co-authored four books with reputed
publishing houses like Mc Graw Hill, PHI, etc. He has contributed fve book chap-
ters in International Proceedings. Ten research scholars have received their  Ph.D.

xix
xx Authors

degree under the supervision of Dr. Chanda. He has visited several countries includ-
ing United States, UK, Australia, Japan and China for academic purposes. He is a
senior member of IEEE (United States), member of IET (UK), Fellow of Institution
of Engineers (I), C-Engineering (I) and life-member of ISTE.
 1 Introduction

1.1 BACKGROUND AND MOTIVATION

Electrical energy has become one of the primary sources of energy. Consumption
of electrical energy is increasing day by day. To meet this power demand the power-
generating stations are growing in both size and capacity. However, smooth operation
and uninterrupted power supply has become a big challenge for engineers, and with
the growth of network size and capacity, the problem has increased in both severity
and complexity.
The last decade has been very challenging for the energy sector, especially the
electrical power system. It is not that the previous years’ challenges were less but
the consequences are much severe in today’s world. The reason is very simple: the
electrical system over decades has evolved in such a way that it has become a need
for survival of the modern society as well as economic development. Integration of
extensive and modern development in information and communications technologies
has revolutionised the power system in terms of observability over the operation of
the widespread power networks, and hence, the response has also become faster and
effcient. However, this has created another challenge of cyber security to the power
system. Natural disasters such as hurricanes, ice storms and earthquakes are con-
tinuously challenging power system researchers to fnd ways to reduce the damage
caused by these events.
In the modern world, the civilisation depends upon bulk-scale usage of electric-
ity, for which the infrastructure involves proper operation of the electric grid. The
electric grid, commonly known as power grid, is also equipped with advanced and
intelligent computer-based sensing and automation system as well as information
technology-related software. This network possesses an effcient communication
system and distributed energy resources in addition to conventional equipment and
transmission lines. Integrating these new technologies has resulted in more intercon-
nections and interdependencies between the physical and cyber components of the
grid. Natural calamities and electrical contingencies have been the traditional forces
of disruption of part or whole of the grid. However, complexities in the political
world and threats on society from terrorist outfts have given rise to the apprehen-
sion that there is a risk to electric grid from intentional attacks. The vulnerability of
infrastructure networks to electric grid outages is becoming a major global concern.
Power systems, composed of grids and sub-grids, support the generation, trans-
mission and distribution of electricity. Power grids can be vulnerable to attacks and
failures. There are instances where failure of component(s) in the grid or line(s) have
caused partial or complete grid collapse. It is the need of the day that the utilities
must maintain high quality of customer service. Therefore, it is essential to minimise
such problems and fx them quickly if they occur. Therefore, intensive research is
needed to study the robustness and vulnerability of a power network. Complex net-
work theory has been investigated extensively in recent years due to its potential for
1
2 Power Transmission System Analysis Against Faults and Attacks

solving large-scale practical problems. In modelling of different complex systems

such as biological systems, social systems can be modelled by the use of complex
network theory. While power system researchers are working towards the upgrada-
tion and safety of the grid and sub-grid, the research on power grid and sub-grid is
an interdisciplinary problem, crossing over several disciplines. The structure of the
network of transmission lines and buses stores a huge amount of useful informa-
tion that can be used for the beneft of the society through the science of complex
networks. It is with this goal that researchers have modelled the power grid and sub-
grid as a complex network to analyse the dynamics of the grid and sub-grid from a
network’s perspective, suggesting some strategies that can help strengthen the grid
and sub-grid.
The developments in the area of application of complex network theory in real
power systems have facilitated advanced research in power system. Following the
concepts described in complex network theory, a power system can be modelled as a
graph with nodes and vertices and further analysis can help in identifying the critical
lines and locating faults. It is an attempt to add knowledge to the existing techniques
for structural vulnerability analysis of power networks.
The modelling of complex networks has signifcantly advanced in the last decade
and has provided valuable insights into the properties of real-world systems by evalu-
ating their structure and construction. Several phenomena occurring in real techno-
logical and social systems can be studied, evaluated, quantifed and remedied using
network science. The electric power grid is one such real technological system that
can be studied through the science of complex networks. The electric grid consists
of three basic subsystems: generation, transmission and distribution. Power trans-
mission lines and feeders are the key components of transmission and distribution
systems. The power grid infrastructure transfers electricity from the generation to the
load end through a complex confguration of transmission lines, buses and different
power equipment. Some of these elements of the network are more important than
others either due to their location in the network, due to the load they carry or due
to their inherent characteristics. If these key elements fail due to outage or attack,
then the performance of the entire system can be affected signifcantly. This book
is motivated to identify such critical and vulnerable elements in the power system
using some concepts from the complex network theory. Attempts have been made to
highlight mitigation strategies so that the resiliency of the network is improved, the
power network can successfully encounter unplanned outages and attacks and the
system can survive from cascading collapse.

1.2 GROWTH OF MODERN LARGE-SCALE POWER SYSTEMS

Modern power systems are large-scale power grids belonging to a typical class of
complex systems, whose characteristics can be summarised as follows:

i) Power supply and consumption should be balanced instantaneously:

Electricity is to be generated and transmitted by lines, transformers and
switches to end-users, with the speed of light. So far, it is yet impossible
to achieve industrial-scale and high-capacity storage of electrical energy.
Introduction 3

Hence, one critical feature of power systems is that the production (genera-
tion) and consumption are completed almost simultaneously, and thus, the
power supply and demand need to be balanced instantaneously. If the bal-
ance is broken, the system will lose its stability, which will lead to power
outages of different scales.
ii) Large-scale power: A modern power system consists of three subsystems:
the frst system, where the energy conversion, transmission, distribution and
consumption take place. The second one is the automatic control system,
also called the secondary system, which is responsible for the secure, stable
and economic operation of the power system. And the third is the power
trading system. The secondary system usually has more components than
the primary system as, in addition to the communication equipment, a large
number of sensors are installed to monitor various states, such as voltage,
current and temperature, of the devices in the primary system. Since the
last couple of decades, the interconnection of regional power systems has
attempted to optimise the use of hydroelectric, thermal and nuclear pow-
ers together with other energy resources inviting more constraints into the
picture.
iii) Modern power system components are complex: The power system
components are not only huge and of numerous kinds but the components
have different complexities as well. Generators and motors that are based
on electromagnetic induction law and Newton’s second law are the most
common components that are used to realise the conversion between elec-
trical and mechanical energies. Others components such as boilers, pres-
sure pipes, coal mills and reactors help to convert chemical, hydraulic and
nuclear energies to mechanical energy. Post 1990s, with the rapid devel-
opment in electronic technologies, high-voltage direct current transmis-
sion (HVDC), thyristor-controlled series compensation and static VAR
compensator as well as other FACTS devices have been utilised widely,
which makes the electric components in power systems more complicated.
Different from traditional power equipment, their main functionality is to
fexibly switch the power electronic devices to control power transmission.
Loads are the most complicated class of components in power systems. The
modelling of dynamic loads has not been fully solved yet. Researchers usu-
ally classify loads into constant impedance loads, temperature-controlled
loads, compressor loads and other simplifed models. In addition, various
protections and controls are important classes of complex components in
power systems. For example, for a generator, there are usually excitation
controllers, speed controllers, low-frequency protectors, under- and over-
excitation limiters. There are also centralised control systems, such as
Automatic Generation Control and Automatic Voltage Control, located in
dispatch centres to coordinate and control the overall performance of the
entire system.
iv) Factors affecting secure operation of power grid: Climatic changes,
human activities and natural calamities such as earthquakes, lightning,
snow and storms force power systems to be constantly exposed to various
4 Power Transmission System Analysis Against Faults and Attacks

random perturbations. For example, lightning or untrimmed trees may

cause short-circuit faults on transmission lines, and may destroy electricity
poles and substations. Due to hot weather, more people use air conditioners;
and major sport and political events may change the type of power loads.
Such random factors increase the complexity in operating and managing
power systems. In addition, as electricity markets are deregulated, energy
management systems become more complicated. In combination with the
random effects just described, the planning, operation and management of
power grids are facing demanding challenges.

From the above four aspects, it is clear that modern power systems are large-scale
complex systems. Therefore, naturally, they become one subject in complexity
studies. Although modern power systems have complex dynamic characteristics,
their robustness and reliability are relatively high due to the implementation of
various protections and controls. However, serious losses were caused in the few
power failures reported in the past years. The interconnection between differ-
ent areas/zones brings both great social and economic benefts as well as risk
for failures, especially cascading failures and blackouts. Large-scale disruptions
of the power grid have occurred several times across the globe confrming the
existence of inherent vulnerabilities. The grid network of Switzerland suffered a
major setback due to electrical failures on 28th September 2003. North America
also faced a major blackout in 2003. Such failures affect millions of people and
cause losses of billions of dollars. Similar incidents have occurred in Canada in
1998, in Europe in 2006, in United States due to Katrina, in 2007, etc. On 30th
and 31st July 2012, the failure of Indian grid affected 680 million people and loss
of billions of dollars.
To minimise the economic, social and political impacts of power system outages,
the grid must be suffciently strong to face the vulnerability. The response of the grid
against such diverse disruptions is a matter of concern. Indeed, one thing is clear that
after decades of human and technological growth, problems are inevitable; however,
it is important to respond to these problems. These challenges motivate to work on
the aspects of vulnerability analysis of power system networks.
Modern power system security is prone to unplanned outages due to natural
calamities and cyber threats that may lead to destructive consequences for the power
system network. In traditional concepts of power system security, researches restored
the power network and maintained the operation of the system against electrical con-
tingencies. However, with time and increasing complexity in the power network, the
complex power network is much more susceptible to unplanned outages as well as to
cyber-attacks and physical failures of key elements of the power system that includes
both grid and distribution systems.
To assist in developing tools or metrics that could be used for early identifcation
of the zones of possible unplanned failure or attack, the upcoming chapters state
methodologies to analyse the power system faults and contingencies in a classical
way and to identify the vulnerable elements, vulnerability of the network and its
resiliency while encountering outages due to natural calamities or attacks at its key
elements.
Introduction 5

1.3 COMPLEXITY IN POWER SYSTEMS

In recent years, the difference between complexity and complication has been noticed
by power system researchers. A few studies regarding complexity and applications
of its theories to power systems have been reported. Instead of looking at the details
of particular blackouts, some researchers [1] studied the statistics and dynamics of a
series of blackouts with approximate global models. However, some researchers [2]
employed topology analysis to fgure out the vulnerability of a given transmission
system and concluded that when a network is attacked following a delicate sequence
corresponding to their criticality, the network would illustrate more vulnerability.
Energy infrastructures, such as power systems, are characterised by a large num-
ber of components and many different types of interactions among them. Size itself
does not infer complexity. Continental-scale power grid, for example, is the biggest
dynamic system in the world, but from a physical viewpoint, it can be modelled
by a huge set of differential and algebraic equations. It may conjure complexity in
the computational effciency; however, it is somehow solvable using computationally
powerful facilities and advanced algorithms. In contrast, complexity arises when
the physical substrate interacts with the rest of hierarchical levels governing and
using the infrastructure. The overall expected performance and dynamic evolution
are related to interactions at the “individual” scale. These phenomena can neither
be handled nor studied with a set of equations in any form. Studies and applica-
tions related to the deregulation towards market environment have thrived in both
academia and industries. This change brought a great challenge to power systems
in production and transmission. With the prevalence of distributed generation and
smart grids, the distribution and utilisation are confronting new scenarios in which a
large number of users have transformed from passive recipients to active participants.
The emerging situation and newly introduced players with clear self-interest display
an important role for the future power system, which will further increase the com-
plexity of the power system.
The complexity of power systems also increases with the change in its adminis-
trative mechanism. Initially, in power systems, each utility and/or pool of utilities
has control centres which support today’s hierarchical monitoring and grid control.
Moreover, electrical market is gradually introduced into power systems to transmit
the least expensive power in power grids. An important consequence of this situation
is that utilities require systematic integration of monitoring, computing and control-
ling for improved performance. Therefore, the interaction between power grids and
decision information via a cyber layer is more complex than before.
Furthermore, renewable energy, such as wind power, solar energy and fuel cell,
is drastically emerging and developing in the distribution level of traditional power
industries. This trend increases the complexity of the overall power systems.

1.3.1 COMPLEXITY AND COMPLEX SYSTEMS

There is no general accepted formal defnition of complexity science as Heyligen
stated: “Conceptually, the most diffcult aspect of complexity is still its defnition
and the deeper understanding that goes with it” [3]. There have been many different
6 Power Transmission System Analysis Against Faults and Attacks

endeavours in complexity over almost all disciplines, covering various defnitions

and measures of complexity. In general, they can be categorised into three groups:
Algorithmic complexity [4,5], deterministic complexity [5] and aggravated complex-
ity [5]. The frst group covers the complexity of describing system characteristics,
such as mathematical complexity theory and information theory. The second group
includes the interaction of very few key variables that create largely stable systems
prone to sudden discontinuities, such as chaos theory and catastrophe theory. The
last one mostly concerns how individual elements work in a synergy that generates
complexity in a system.
To have a lucid understanding of complexity, as well as generalising the concept
without losing any useful positive meaning, complexity can be defned as follows:
“Complexity is a property that makes it diffcult to analytically formulate its over-
all behaviour even when knowing the complete information about its elements and
their relationships.” Here “diffcult” can involve several aspects such as size, depth,
computational indication, efforts in a search for the most apt representation, etc.
Accordingly, a general and logical defnition of complex system is one that exhib-
its complexity: “A system, that can be decomposed in a set of different types of ele-
mentary parts with autonomous behaviours, goals and attitudes and an environment,
is complex if its modelling and related simulation tools cannot be done resorting to a
set of whichever type of equations expressing the overall performance of the system,
in terms of quantitative metrics, or a function on the basis of state variables and other
quantitative inputs” [23]. This defnition is an articulated and practical way targeting
at engineering systems.
Over the last several years, complexity science has changed the way scientists
approach all felds of life form biology to medicine, and from economics to engineer-
ing [6–10]. The concepts or techniques such as self-organisation, genetic algorithm,
cellular automata, criticality, artifcial life or chaos theory are now widely accepted
and used as new means of understanding the constantly changing reality. The history
of complex systems research including these concepts begins in the 1950s, emerging
with the advent of von Bertalanffy’s systems theory, the appearance of nonlinear
phenomena in scientifc felds away from physics, such as chemistry and biology, and
the study of feedback concepts in communication and control in living organisms,
machines and organisations. From these early stages, the idea of threshold became
the cornerstone of much of the complexity science developments of the 1980s, espe-
cially in the cellular automata and artifcial life felds, where complex behaviour
seemed to appear suddenly [11–13]. Since then, many books, journals, conferences
and even whole institutes devoted to the feld have fourished everywhere, and even
computer modelling of complex systems has become widely accepted as a valid
scientifc activity. Marti Rosas-Casals [32] cited a conceptual framework in which
complexity pervades both the (i) structure (i.e. formal arrangement of the constitu-
ent parts), (ii) dynamics (i.e. functional behaviours) and (iii) evolution (i.e. the way it
has reached its actual formal and functional state) of any system. It covers the major
aspects of complexity science and different technologies and methods. These meth-
ods and techniques of complex systems are grouped into three categories: (i) those
for analysing data, (ii) those for building and understanding models and (iii) those
for measuring complexity.
Introduction 7

With new advances in complex network theory, most of the complex systems in
the world can be modelled and described in the form of a complex network. Several
network models have been proposed with the intention of studying the topological
characteristics and behaviour of such complex systems. The types of networks can be
broadly classifed as regular, small-world, random and scale-free networks.

1.3.1.1 Regular Networks

Initially, complex networks were assumed to be completely regular. A few examples
of regular networks are chains, grids, lattices, fully connected graphs, etc. The rela-
tively simple architecture allows us to focus on the complexity caused by the nonlin-
ear dynamics of nodes and edges without the additional complexity of the topology
itself [14].
Hence, these network models have been used quite often to study dynamic sys-
tems such as disease spread and ecosystems. For a regular network, the clustering is
high and the average distance between nodes is high [15].

1.3.1.2 Random Networks

Regular network models were not effcient to describe the phenomenon of real-world
systems; hence, Erdos and Renyi [16] came up with the idea of random networks. The
term random refers to disordered arrangement of links connecting various nodes.
Such networks are created when each pair of nodes is connected by a link with uni-
form probability. These networks were studied by Erdos and Renyi from a purely
mathematical viewpoint. However, networks with complex topology and unknown
properties are often represented as a random graph. Hence, the random graph theory
has a signifcant place in the study of complex networks. In case of random networks,
the average distance between nodes is small and the clustering is low [15]. Such net-
works are usually robust to targeted attacks but very vulnerable to random failures
or attacks [17].

1.3.1.3 Small-World Networks

There are networks whose behaviour falls between a regular network and a ran-
dom network. These types of complex networks are classifed as small-world net-
works, which were frst introduced by Watts and Strogatz [15,18,19]. Even though
there are few long-distance connections present in such networks, the shortest
path length between any two nodes scales logarithmically or at a smaller rate
with increasing network size. This means even in a small-world network with
many nodes, the shortest path length between two nodes is likely to be relatively
small with high clustering [20]. Power system is a good example of a small-world
network.

1.3.1.4 Scale-Free Networks

About the same time when Watts and Strogatz developed the small-world model,
Barabasi and Albert [21] came up with an alternate network model which grew
by preferential attachment and were called scale-free networks. These networks
grow in such a way that nodes with higher degrees receive more new connections
compared to others with low degrees, that is, the probability of making a new
8 Power Transmission System Analysis Against Faults and Attacks

connection to a node is proportional to its degree. Unlike the Gaussian distribution

which has cut-off values where the distribution approaches zero, scale-free distri-
butions have no such cut-offs, and instances of all scales are present; hence, the
term scale-free. These networks are simply those with power law degree distribu-
tion where most nodes have low connectivity but some are highly connected to the
rest of the network [21]. This makes those highly connected nodes or hubs very
vulnerable to attacks. Thus, such network models are robust to random attacks but
can be highly vulnerable to targeted attacks [17]. A classic example of a scale-free
network is the Internet.
Figure 1.1 illustrates simple examples of regular, random, small-world and scale-
free networks [22].

1.3.2 POWER GRID AS A COMPLEX SYSTEM

Power grids have been widely acknowledged as a typical complex network because
of both their huge sizes of components and the complex interactions among them.
Complex network theory has received considerable attention recently which has been
used in many different felds. Numerous studies [23] including basic characteris-
tics, statistical global graph properties, small-world property, scale-free property,
degree distribution, betweenness distribution and vulnerability analysis have been
performed for power grids as they are infrastructures in our society. There is a strong
link between the topological structure and operation performance in power systems
because the structural change could alter operational condition of a power system as
well as its operation performance.
Consequently, there is an increasing interest in analysing structural vulnerabil-
ity of power grids through complex network methodology. With the development of
complex system theory, power grids arise as natural objects of investigation under
the conceptual framework of complex systems, particularly as complex networks.
Therefore, complex network methodology as one of approaches to study complex
systems has been used to analyse and understand power systems from a topological
viewpoint. How complex network theory and methodology are applied in power sys-
tems study, especially in vulnerability analysis, will be addressed in the subsequent
chapters.

(a) (b) (c) (d)

FIGURE 1.1 Different types of complex networks: (a) Regular, (b) random, (c) small word,
and (d) scale free.
Introduction 9

1.4 TRANSMISSION SYSTEM FAULTS

In an electrical power system, a fault is any abnormal fow of unavoidable and unde-
sirable current which can temporarily disturb the stable operating condition of the
system due to insulation failure by any internal or external agent at any point of time.
The agents causing fault may be natural (lightning, wind damage, trees falling across
transmission lines, rodents, squirrels or birds shoring lines) or man-made (outage
due to islanding, vehicles or aircraft colliding accidentally with transmission towers
and poles, any kind of intentional attacks or vandalism). The nature of the faults may
be any abnormal condition which reduces basic insulation strength between phase
conductors, or between phase conductors and earth, or between any earthed screens
surrounding the conductors [24]. In practice, a reduction is not regarded as a fault
until it results either in an excess current or in a reduction of the impedance between
conductors, or between conductors and earth, to a value below that of the lowest load
impedance normal to the circuit.
The faults in transmission system causes overcurrent, undervoltage, unbalance
of phase, reversed power and high-voltage surges. This results in failure of equip-
ment, electrical fres and interruption in normal operation of the network. These
faults occur due to the failure of one or more conductors. The most common faults
that occur in power system are unsymmetrical or asymmetric which lead to unequal
currents with unequal phase shifts in a three-phase system. This kind of fault occurs
in a system due to the presence of an open circuit or short circuit of transmission or
distribution line. The path of the fault current may have either zero impedance (short
circuit) or nonzero impedance [25]. Other type of fault includes one conductor or two
conductors open (open conductor faults). Such faults occur when conductors break or
when one or two phases of a circuit breaker is inadvertently open.
There are mainly three types of faults, namely, line-to-ground (L-G), line-to-line
(L-L) and double line-to-ground (LL-G) faults. Line-to-ground fault (L-G) is the
most common fault, and 65%–70% of faults are of this type. It causes the conductor
to make contact with earth or ground. Line-to-line is a short circuit between lines
caused by ionisation of air, or when lines come into physical contact, for example,
due to a broken insulator. In transmission line faults, roughly 5%–10% are asymmet-
ric line-to-line faults. Line-to-ground is a short circuit between one line and ground,
very often caused by physical contact, for example, due to lightning or storm damage.

1.5 CONVENTIONAL CONTINGENCIES IN POWER

TRANSMISSION NETWORK
Currently, due to large interconnection and stressed operation, power utilities are
facing severe problems of maintaining the required security. Today more emphasis is
paid to the greater utility of generation and transmission capacity, which has caused
the system to operate much closer to their limits. So, it has become indispensable to
do voltage security assessment accurately and instantaneously to prevent the system
from voltage collapse. The concept of security in system operation may be divided
into three components: monitoring, assessment and control. Security monitoring
starts with measurement of real-time system data to provide up-to-date information
10 Power Transmission System Analysis Against Faults and Attacks

of the current condition of power system. Security assessment is the process whereby
any violation of the actual system operating is determined. The conventional meth-
ods for security assessment are based on load fow solution where full AC load fow
is made to run for all contingencies. Load fow constitutes the most important study
in a power system for planning, operation and expansion. The purpose of load fow
study is to compute operating conditions of the power system under steady state.
These operating conditions are normally voltage magnitudes and phase angles at
different buses, line fows (MW and MVAR), real and reactive power supplied by the
generators and power loss. The second and much more demanding function of secu-
rity assessment is contingency analysis. Operations personnel must know which line
or generation outages will cause fows or voltages to fall outside limits. To predict the
effects of outages, contingency analysis techniques are used.
Contingencies are defned as potentially harmful disturbances that occur dur-
ing the steady state operation of a power system. The contingencies are in the form
of network outage such as line or transformer outage or in the form of equipment
outage. The outage considered in transmission network are line outages. Outages
that are important from a limit violation viewpoint are branch fow for line security
or MW security and bus voltage magnitude for voltage security. The conventional
methods for security assessment are based on load fow solution where full AC load
fow is made to run for all contingencies. Contingency procedure includes failure
events one after another in sequence until “all credible outages” have been studied.
For each outage tested, the contingency analysis procedure checks all lines and volt-
ages in the network against their respective limits.

1.6 THREATS AND THEIR CONSEQUENCE IN POWER

SYSTEM OPERATION
The science of power grid technology is progressing worldwide. Most are investing
to transform their traditional power grid to smart grid. They have started realigning
their organisation to support a smart grid vision. At its core, a smart grid utilises
digital communications and control systems to monitor and control power fows, with
the goal of making the power grid more resilient, effcient and cost-effective. Smart
grids increase the connectivity, automation and coordination between suppliers, con-
sumers and network by modernising grid features like demand side management,
generation, real-time pricing and automated metre activation and reading. As smart
grid technology is advancing with time, complexity in managing the power grid is
also increasing.
The grid must be suffciently strong to face the vulnerability. The response of the
grid against such diverse disruptions is a matter of concern. Indeed, one thing is very
clear that after decades of human and technological growth, problems are inevitable,
but it is important to respond to the problems. These challenges motivate to work on
the aspects of vulnerability analysis of power system networks.
Modern power system security is prone to unplanned outages due to natural calam-
ity and cyber threats that may lead to destructive consequences for the power system
network. In traditional concepts of power system security, researches were conducted
in restoring the power network and maintaining the operation of the system against
Introduction 11

electrical contingencies. However, with time and increasing complexity in power net-
work, the complex power network has become susceptible to unplanned outages as
well as cyber-attacks and physical failures of key elements of the power system that
includes grid system and the distribution system.
A cyber attacker is an intelligent agent capable of coordinating attacks that result
in deliberate component failures, which increases the risk of low-probability, high-
impact contingency scenarios. The grid network or the distribution network is also
vulnerable while encountering unplanned contingencies during natural calamity.
The traditional physical contingency analysis tools rely on a version of the power
fow problem, which requires calculating the solution of operating parameters for all
buses in the power system. While personnel within a utility may have knowledge of
specifc bus injections and voltage magnitudes needed to solve the power fow prob-
lem, it is not uncommon that a cyber attacker plans to identify critical elements in the
power network and fulfls their evil desire in making the power system vulnerable
to attack at its critical elements. In addition, the power system faces high degree of
vulnerability when the system is subjected to unplanned electrical failures of a few
of its key elements during natural calamities.
Instances like faults at power stations, damage to electric transmission lines, sub-
stations or other parts of the distribution system, a short circuit, cascading failure,
fuse or circuit breaker operation may cause complete collapse of the power grid. In
addition to these, there has been instances of several targeted attacks on grids in
power system networks in an attempt to collapse the entire system network [33]. To
maintain the reliability of power systems, wide area monitoring systems are exploited
to obtain real-time system status, which is essential for the maintenance and control
of power systems. Most studies on security issues in smart grid have focused on
how to protect data transmission in power grid. Accordingly, the attack on the mea-
surements is called False-data Injection Attack (FIA) [26]. FIA allows attackers to
manipulate the estimated state of the power grid by changing the measurements on
a fraction of monitoring devices. When the attacker accesses the power grid com-
munication network, the attack can bypass the existing bad measurement detection.
The vulnerability assessment of power transmission networks employing com-
plexity science investigates the critical targeted vulnerable areas of such an attack.
Hence, it is a present-day requirement to be equipped with proper knowledge on
vulnerability of grid networks.

1.7 MODELLING OF POWER TRANSMISSION

NETWORK AS COMPLEX NETWORK
A network can be defned as a set of nodes or vertices with connections called links
or edges [27]. A vertex is the fundamental unit of a network, also called a node
(in electrical and computer science). They are connected together by lines called
edges, also known as a link (in electrical and computer science). The nodes or vertex
represent various elements, tangible or otherwise, such as hardware devices, buses,
generators and transformers, and the edges represent the relationship between these
elements or the way they interact with each other. Figure 1.2 shows a network with
nodes and edges in its simplest form.
12 Power Transmission System Analysis Against Faults and Attacks

vertex

edge

FIGURE 1.2 A network with nodes and edges.

Networks can be of different types: For instance, there can be networks with
similar nodes and edges or there can be networks with more than one type of nodes
and different types of edges. Further, these vertices and edges can have various
properties. For example, edges can have weights associated with them which might
represent how strongly or loosely any two nodes are connected. Such networks
are called weighted networks. Any information is transferred within the network
via the nodes using connecting links. Sometimes, the fow of this information can
only be in one direction, in which case the network is termed as directed graphs or
digraphs. These directed networks can either be cyclic containing a closed loop of
edges or acyclic containing no such loops. Then, there are undirected networks in
which fow of information can be in both directions of connections. Figure 1.3 shows

(a) (b)

(c) (d)

FIGURE 1.3 Examples of different types of networks. (a) Network with Identical nodes
and edges, (b) Network with different types of nodes and edges, (c) Network with different
weights put on to nodes and edges, (d) An example of Directed graph.
Introduction 13

simple examples of various types of networks. Figure 1.3a shows a network with
identical nodes and edges, Figure 1.3b shows different types of vertices and con-
nections, Figure 1.3c represents a network in which nodes and edges have differ-
ent weights associated with them and Figure 1.3d shows an example of a directed
network.
To represent the power grid as a complex network, an unweighted and undirected
graph composed of nodes and edges needs to be built frst. Using metrics from graph
theory and modern complex networks analysis, the results provide insight into the
properties of power grids, considering only topological information. In summary, the
goal of this model is to characterise the topological structure of the power transmis-
sion grid.
To study power grids with this model, some simplifcations are necessary. In
the undirected graphs, each node represents a bus. It is important to note that
in  the physical grid, these buses can have different electrical properties; how-
ever, in this chapter, nodes are assumed to be homogeneous. This model ignores
whether generators, loads, transformers or transmission lines connect to the bus.
In the same manner, all transmission lines are modelled as edges with equal
weight. Physical length and electrical impedance are ignored in the undirected
graph representation.
The graph theory is the basic concept from which complex network theory has
been derived. The power grid can be abstracted into the complex network with undi-
rected graph G = {V, E}. It consists of two sets V and E, where the elements of
V = {v1, v2, …, vN} are the nodes (or vertices, or buses) of the graph G, while the ele-
ments E = {e1, e2,…, eL} are its links (or edges or lines). The total number of nodes
and links of the graph are N and L, respectively. The association of nodes with each
other can be shown using adjacency matrix. If an edge eij exits between two nodes i
and j, then the adjacency matrix A having order N × N whose entry aij becomes one
and zero otherwise. Adjacency matrix is a N × N binary matrix in which the value of
[i, j]th cell is 1 if there exists an edge originating from ith vertex and terminating to jth
vertex, otherwise the value is 0.
A power system can be considered a large complex network with nodes and
edges. The generators, bus bars and loads can be identifed as the nodes and the
connecting transmission lines can be modelled as the edges or links. Let us con-
sider a sample network consisting of fve buses and seven lines modelled as an
undirected graph.
Figure 1.4 shows the undirected graph of a system with fve nodes and seven
edges. It can be modelled using the principles defned above and mathematically
represented as a matrix shown below:

° 0 0 1 1 0 ˙
˝ 0 0 1 0 1 ˇ
˝ ˇ
A=˝ 1 1 0 1 1 ˇ
˝ 1 0 1 0 1 ˇ
˝ ˇ
˛˝ 0 1 1 1 0 ˇˆ
14 Power Transmission System Analysis Against Faults and Attacks

1 2

4 5

FIGURE 1.4 Undirected graph of a sample network.

1 5

0 2

FIGURE 1.5 Undirected graph for Example 1.1.

Example 1.1

Problem Statement: Consider a graph given in Figure 1.5 below. Find the adja-
cency matrix for the network.

Solution

In this undirected graph, there are N vertices numbered from 0 to N − 1 and E

number of edges in the form (i, j), where (i, j) represent an edge originating from
ith vertex and terminating on jth vertex. Now, the adjacency matrix will be:
Introduction 15

1 5

0 2

FIGURE 1.6 Directed graph for Example 1.2.

In the directed graph, an edge is represented by an ordered pair of vertices (i, j), in
which edge originates from vertex i and terminates on vertex j.

Example 1.2

Problem Statement: Consider the same sample network in Example 1.1 but as
directed graph, as shown in Figure 1.6. Find the adjacency matrix.
Solution

The adjacency matrix is given by:

Furthermore, we can add weights to the connecting links which could be a mea-
sure of electrical or topological property depending on the application. In that
case, the 1’s will be replaced by the respective weights of the links. This will be
further illustrated through various examples in this section.

WEIGHTED GRAPH
A graph is called a weighted graph when it has weighted edges, which means
there are some “value” associated with each edge in the graph. For example, con-
sider Figure 1.7 where there are weights associated with each edge.
16 Power Transmission System Analysis Against Faults and Attacks

2
1 3

4 7
4
5

0 2 5
4

3 6

2 4

FIGURE 1.7 Directed weighted graph.

Each edge of a graph has an associated numerical value called a weight.

Usually, the edge weights are non-negative integers. Weighted graphs may be
either directed or undirected. The weight of an edge is often referred to as the
“cost” of the edge.

Example 1.3

Consider an undirected weighted graph as shown in the Figure 1.8 and fnd its
adjacency matrix.

4
1 2

5 3
3 4

6 4 3

1
2

5 4
3

FIGURE 1.8 Undirected weighted graph of Example 1.3.

Introduction 17

Solution

The adjacency matrix for this undirected weighted graph is given by

˜ 0 4 0 3 0 5 ˝
˛ 4 0 3 4 4 0 ˆ
˛ 0 3 0 2 0 0 ˆ
A=˛ ˆ
˛ 3 4 2 0 3 0 ˆ
˛ 0 4 0 3 0 1 ˆ
˛° 5 0 0 0 1 0 ˆ˙

Example 1.4

Consider a directed weighted graph as shown in the Figure 1.9 and fnd its adja-
cency matrix.
Solution

The adjacency matrix for this directed weighted graph will be:

˜ 0 4 0 0 0 0 ˝
˛ 0 0 0 4 0 0 ˆ
˛ 0 3 0 0 0 0 ˆ
A=˛ ˆ
˛ 3 0 2 0 3 0 ˆ
˛ 0 4 0 5 0 1 ˆ
˛° 5 0 0 0 0 0 ˆ˙

4
1 2

5 3
3 4

6 4 3

1
2
3
5 4

FIGURE 1.9 Directed weighted graph of Example 1.4.

18 Power Transmission System Analysis Against Faults and Attacks

1.8 STRUCTURAL PROPERTY ANALYSIS OF

POWER TRANSMISSION NETWORK
As mentioned in Section 1.3.1, there are three main models of complex networks:
small-world, scale-free and random networks in addition to regular networks.
To explain the characteristics of these different networks, Figure 1.10 shows
a simple rewiring diagram [], which illustrates the relationship between regular,
small-world and random networks. We start with a ring lattice with n = 20 nodes
with each of them connected to four of their neighbours. Let each edge be rewired
randomly with a probability P, that is, P is the ratio of number of lines rewired
randomly versus total number of lines. Then, for P = 0, the original lattice is
unchanged.
As the value of P is increased, the network becomes increasingly random, and for
P = 1, all the lines are rewired randomly. The small-world phenomenon exists in the
intermediate region 0 < P < 1.
Different network will exhibit different structure and vulnerability property.
Therefore, the frst question to analyse power grid is what type of power grid
is. The frst reference as deduced by Watts and Strogatz [28] is that the western
US power grid seemed to be a small-world network. Later, Barabasi and Albert
[21,29] frst published that degree distribution of a power grid was scale-free fol-
lowing a power law distribution function; however, few subsequent references
supported this fnding. Exponential cumulative degree function was detected in
Californian power grid [30] and the entire US grid [31]. The topological features
of the Union for the Co-ordination of Transport of Electricity (UCTE) power grid
and its individual nation grids were analysed, and the results showed that these
national transmission power grid topologies are similar in terms of mean degree
and degree distribution, suggesting similar topological constraints, mostly associ-
ated with technological considerations and spatial limitations [32]. Furthermore,
the topologies of the North American eastern and western electric grids were anal-
ysed to estimate their reliability based on the Barabasi–Albert network model. The
results were compared to the values of power system reliability indices previously

(a) (b) (c)

p=0 p=1
increased randomness

FIGURE 1.10 Relationship between network models and randomness. (a) Regular, (b) small
word, and (c) random.
Introduction 19

obtained from some standard power engineering methods, which suggested that
scale-free network models are applicable for estimating aggregate electric grid
reliability [33].

1.9 COMPLEX NETWORK APPROACH TO VULNERABILITY

ASSESSMENT OF POWER TRANSMISSION NETWORK
The vulnerability analysis of network is the main motivation for the studies involv-
ing complex network analysis into power grids. When vertices are removed ran-
domly and in decreasing order of their degrees for both generation vertices and
transmission vertices, a connectivity loss is observed, which measures the decrease
of the ability of distribution substations to receive power from the generators. The
loss of generating substations does not signifcantly alter the overall connectivity of
the grid owing to a high level of redundancy at the generating substations. However,
the grid is sensitive to the loss of transmission nodes. Even the removal of a single
transmission node can cause a slight connectivity loss. Especially, the connectivity
loss is substantially higher when intentionally attacking higher-degree or high-load
transmission hubs. The transmission highly connected hubs guarantee the con-
nectivity of the power grid; however, they are also its largest liability in case of
power breakdowns.
In European power grids, the topological properties of the Spanish, Italian and
French power grids are studied and compared, fnding those components whose
removals seriously affected the structure of these graphs [34]. They also treated
power grid as a simple graph with taking physical features into consideration; we
think that the power grid vulnerability results obtained with this approach would be
different from the real situation. Rosato et al. studied the topological properties of
high-voltage power grid in Italy (380 kV), France (400 kV) and Spain (400 kV) [35].
An assessment of the vulnerability of the networks was done by analysing the level
of damage caused by a controlled removal of links. Such topological studies could
be useful for vulnerability assessment and for designing specifc action to reduce
topological weaknesses. Because the grids are the same as used in the former case,
most of the results are consistent. Robustness of the entire European power grid
was studied where the resilience against the failures and attacks of each national
power grid was included. The authors [35,36] found that the European power grid
composed of 33 EU power grids could broadly be classifed into two separate groups:
fragile and robust.
While analysing the frst cascading failures in electrical power grid of the western
United States [37], the degree distribution in this network appeared exponential and
was relatively homogeneous. The distribution of loads, however, was more skewed
than that displayed by semi-random networks with the same distribution of links.
This implied, to a certain extent, that the power grid may have structures that are
not being captured by the existing complex network models. Consequently, global
cascades are probably triggered by load-based intentional attacks than by random
or degree-based removal of vertices. The attack on a single vertex with large load
may make the largest connected component reduce to less than one-half of its initial
size, though the network is highly tolerant. It was observed [38] in North American
20 Power Transmission System Analysis Against Faults and Attacks

power grid that the loss of a single substation can lead to a 25% loss of transmission
effciency caused by an overload cascade in the network. A systematic study of the
damage caused by the loss of vertices suggested that 40% of the disrupted transmis-
sion substations may lead to cascading failures. While the loss of a single vertex
can exacerbate primary substantial damage, the subsequent removals only make the
situation worse.
Large-scale blackouts can be triggered by the failure of vertices with high loads.
Perhaps it is because some highly connected vertices may not be necessarily involved
in a high number of paths. A counterintuitive result was found that the attack on the
vertices with the lowest loads is more harmful than the attack on the ones with the
highest loads.
Within the framework of cascading failures in power grids using a dynamical
fow model based on simple conservation and distribution laws, the role of the tran-
sient dynamics of the redistribution of loads towards the steady state after the failure
of network edges was studies. It was found that considering only load fows in the
steady state gives a best case estimate of the robustness; the worst case of robust-
ness can be determined by the instantaneous dynamic overload failure model. The
results of Norwegian high-voltage power grid showed that the size distribution of
power blackouts in Norwegian power grid seems to follow a power law probability
distribution.

1.10 CYBER SECURITY AND VULNERABILITY

TOWARDS CASCADED COLLAPSE
Cyber security based vulnerability assessment measures tend to come in two forms:
those that assess the vulnerability of cyber assets in the power system for compro-
mise, and those that assess the consequences to the physical power system resulting
from compromised cyber assets. However, because the electric grid is a cyber-
physical system, the distinction between cyber vulnerabilities and the consequence
to the physical system is often diffcult to establish. For example, the existence of
vulnerabilities in a cyber asset such as a control, protection or monitoring device or
system does not necessitate a severe reliability impact to the physical power system
if the asset is compromised. However, if a cyber attacker can utilise the vulnerability
to cause an outage of physical power system components, it is possible for a power
system to be thrust into an emergency state because of the cyber-attack.
Power system cyber asset vulnerabilities are commonly understood through con-
ventional information security principles. Successful cyber-attacks typically use
some vulnerability in the communication protocol, routing or authentication pro-
cesses of a cyber asset to install malware, deny legitimate services or directly intrude
into an information system [39]. Successful cyber-attacks can have various conse-
quences, but the most common are theft or corruption of information, unavailability
of computing resources and physical destruction of equipment [40].
Cyber assets and applications in a power system are uniquely susceptible to
various cyber-attacks. By corrupting the information utilised by a state estimator,
attackers have the potential to alter the state estimation solution to refect a fctitious
power system state while evading bad data detection routines [41]. Furthermore, if
Introduction 21

an attacker can inject false data into a state estimator, they can alter the result of a
security constrained optimal dispatch application, which can lead to the generator
redispatching that results in uneconomic operation of a system and/or an insecure
operating state [42]. Research activities have been pursued to develop more robust
bad data detection algorithms that can detect a state estimator FIA [43].
Moreover, of interest is the vulnerability of Supervisory Control and Data
Acquisition (SCADA) systems to cyber-attacks. While SCADA systems tend to be
isolated from public communication networks, features such as remote vendor and
engineering access into substation networks for legitimate purposes have created an
attack surface from which a skilled attacker can obtain unauthorised entry into a
SCADA system [44].

REFERENCES
1. I. Dobson, B. Carreras, V. Lynch, and D. Newman, “Complex systems analysis of
series of blackouts: Cascading failure, critical points, and selforganization”, Chaos: An
Interdisciplinary Journal of Nonlinear Science, vol. 17, pp. 26–103, 2007.
2. A. Wildberger, “Complex adaptive systems: Concepts and power industry applications”,
IEEE Control Systems Magazine, vol. 17, no. 6, pp. 77–88, 2002.
3. F. Heylighen, “Five questions on complexity”, Arxiv preprint online, 2007.
4. G. Chaitin, Information, Randomness and Incompleteness: Papers on Algorithmic
Information Theory, Singapore: World Scientifc Singapore, 1997.
5. S. Manson, “Simplifying complexity: A review of complexity theory”, Geoforum, vol.
32, no. 3, pp. 405–414, 2001.
6. D.L. Harris, M. Mitchell Waldrop: Complexity-the Emerging Science at the Edge of
Order and Chaos, New York: Simon and Schuster Inc., 1992.
7. Lewin, R., Complejidad. El caos como generador del orden, Barcelona: Tusquets
Editores, 1995.
8. R.V. Solé and S.C. Manrubia, Orden y caos en sistemas complejos, Barcelona: Edicions
UPC, 1996.
9. R. Solé and B.C. Goodwin, Signs of Life: How Complexity Pervades Biology, New
York: Basic Books, 2001.
10. P. Érdi, Complexity Explained, Heidelberg: Springer-Verlag, 2008.
11. J. Singh, Great Ideas in Information Theory, Language and Cybernetics, New York:
Dover Publications Inc., 1966.
12. S. Kauffman, At Home in the Universe, New York: Oxford University Press, 1995.
13. S. Wolfram, A New Kind of Science, Champaign, IL: Wolfram Media, 2002.
14. S.H. Strogatz, “Exploring complex networks”, Nature, vol. 410, pp. 268–276, 2001.
15. V. Latora and M. Marchiori, “Effcient behaviour of small-world networks,” Physical
Review Letters, vol. 87, no. 19, p. 198701, 2001.
16. P. Erdos and A. Renyi, “On the evolution of random graphs,” Publications of the
Mathematical Institute of the Hungarian Academy of Sciences, vol. 5, pp. 17–61, 1960.
17. R. Albert, H. Jeong, and A.L. Barabasi, “Error and attack tolerance of complex net-
works,” Nature, vol. 406, pp. 378–382, 2000.
18. D.J. Watts and S.H. Strogatz, “Collective dynamics of ‘small-world’ networks,” Nature,
vol. 393, no. 6684, pp. 440–442, 1998.
19. D.J. Watts, Small Worlds: The Dynamics of Networks between Order and Randomness,
Princeton, NJ: Princeton University Press, pp. 11–40, 1999.
20. M. Mitchell, “Complex systems: Network thinking,” Artifcial Intelligence, vol. 170, pp.
1194–1212, 2006.
22 Power Transmission System Analysis Against Faults and Attacks

21. A.L. Barabasi and R. Albert, “Emergence of scaling in random networks,” Science, vol.
286, no. 5439, pp. 509–512, 1999.
22. http://cs.wikipedia.org/wiki/Multiagentn%C3%AD_modelov%C3%A1n%C3 %AD.
23. L. Luo and P. Di Torino, Power system vulnerability and performance: Application
from complexity scienze and complex network, Doctoral Thesis, March 2014.
24. T.A. Ernster, Power system vulnerability analysis: A centrality based approach utilizing
limited information, Washington State University, Master of Science Thesis, August
2012.
25. http://web.ecs.baylor.edu/faculty/lee/ELC4340/Lecture%20note/Chapter9_GSO5.pdf.
26. H. Zhong, D. Du, C. Li, and X. Li, “A novel sparse false data injection attack method in
smart grids with incomplete power network information”, Complexity, vol. 2018, no. 2,
pp. 1–16, 2018.
27. A. Dwivedi, Vulnerability analysis and fault location in power systems using complex
network theory, Doctoral Thesis, RMIT University, March 2011.
28. D.J. Watts and S.H. Strogatz, “Collective dynamics of ‘small-world’ networks”, Nature,
vol. 393, pp. 440–442, 1998.
29. R. Albert, et al., “Error and attack tolerance of complex networks”, Nature, vol. 406, pp.
378–382, 2000.
30. L.A.N. Amaral, et al., “Classes of small-world networks”, Proceedings of the National
Academy of Sciences of the United States of America, vol. 97, pp. 11149–11152, 2000.
31. R. Albert, et al., “Structural vulnerability of the North American power grid”, Physical
Review E, vol. 69, pp. 25103–25106, 2004.
32. M. Rosas-Casals, S. Valverde, and R. Sol_e, “Topological vulnerability of the European
power grid under errors and attacks”, International Journal of Bifurcation and Chaos,
vol. 17, no. 7, pp. 2465–2475, 2007.
33. D.P. Chassin and C. Posse, “Evaluating North American electric grid reliability
using the Barabasi-Albert network model”, Physica A-Statistical Mechanics and Its
Applications, vol. 355, pp. 667–677, 2005.
34. P. Crucitti, et al., “Locating critical lines in high voltage electrical power grids”,
Fluctuation and Noise Letters, vol. 5, pp. L210–L208, 2005.
35. V. Rosato, et al., “Topological properties of high-voltage electrical transmission net-
works”, Electric Power Systems Research, vol. 77, pp. 99–105, 2007.
36. R. Solé, et al., “Robustness of the European power grids under intentional attack”,
Physical Review E, vol. 77, p. 26102, 2008.
37. A. Motter and Y. Lai, “Cascade-based attacks on complex networks”, Physical Review
E, vol. 66, p. 65102, 2002.
38. R. Kinney, et al., “Modeling cascading failures in the North American power grid”,
European Physical Journal B, vol. 46, pp. 101–107, 2005.
39. M. Govindarasu, A. Hann, and P. Sauer, “Cyber-physical systems security for smart
grid,” PSERC, Publication 12-02, February 2012.
40. C. Tranchita, N. Hadjsaid, and A. Torres, “Overview of the power systems security with
regard to cyberattacks,” Fourth International Conference on Critical Infrastructures,
ISGT Latin America, Medellin, Colombia, pp. 1–8, 27 March 2009 to 30 April 2009.
41. O. Kosut, L. Jia, R.J. Thomas, and L. Tong, “On malicious data attacks on power sys-
tem state estimation,” 45th International Universities Power Engineering Conference,
Cornell University, USA, pp. 1–6, 31 August to 3 September 2010.
42. Y. Yuan, Z. Li, and K. Ren, “Modeling load redistribution attacks in power systems,”
IEEE Transactions on Smart Grid, vol. 2, no. 2, pp. 382–390, 2011.
43. O. Kosut, L. Jia, R.J. Thomas, and L. Tong, “Malicious data attacks on smart grid
state estimation: Attack strategies and countermeasures,” First IEEE International
Conference on Smart Grid Communications, Gaithersburg, Maryland, pp. 220–225,
4–6 October 2010.
Introduction 23

44. C. Liu, C. Ten, and M. Govindarasu, “Cybersecurity of SCADA Systems: Vulnerability

assessment and mitigation,” IEEE 2009 Power Systems Conference and Exposition,
Seattle, WA, pp. 1–3, 15–18 March 2009.
45. S. Mei, X. Zhang, and M. Cao, “Power Grid Vulnerability”, Berlin Heidelberg: Tsinghua
University Press, Beijing and Springer-Verlag, 2011.
 2 Traditional Approach
in Analysis of Faults
in Power System

2.1 INTRODUCTION
Usually, a power system operates under balanced conditions. However, electrical
fault may appear in the system leading to unbalanced system operation. Out of com-
mon electrical faults, the line fault is the most common. Much less common are the
faults on cables, generators, motors and transformers. The most common faults are
the single line-to-ground fault (L-G) followed by line-to-line (L-L), double line-to-
ground (L-L-G) and three-phase balanced (L-L-L-G) faults.
It is important to study the system under fault condition to provide necessary
system protection. In this chapter, the concept of symmetrical components [1] is
introduced and different types of faults are analysed using symmetrical component
theory so that the fault current and the voltages at different buses can be computed.
Line currents have also been obtained for different fault conditions. Once the fault
currents are determined at any part of the network, it becomes convenient for setting
protective relays for electrical protection of power systems.

2.2 FORMATION OF [YBUS] MATRIX

Let a three-bus simple power network be considered as shown in Figure 2.1a, where
Z A and ZB are reactances associated with EMF sources G1 and G 2 (most commonly
the alternators), and ZC, ZD and ZE are reactances of the interconnecting lines between
the buses 1, 2 and 3.
Let V1, V2 and V3 be the respective bus voltages as shown in Figure 2.1a. Figure 2.1b
is the Norton’s equivalent circuit for the network shown in Figure 2.1a with EMF
sources replaced by the current sources associated with corresponding admittances
YA and YB. The line impedances are also replaced by corresponding admittances YC,
YD and YE.
Application of Kirchhoff’s current law (KCL) at node 1 (bus-1) yields

−I1 + V1YA + (V 1 − V2 ) YD + (V1 − V3 ) YE = 0

(2.1)
i.e. I1 = V1YA + (V1 − V2 ) YD + (V1 − V3 ) YE .

Similarly, application of KCL at node 2 (bus-2) yields

I 2 = V2YB + (V2 − V1 ) YD + (V2 −V3 ) YC (2.2)

25
26 Power Transmission System Analysis Against Faults and Attacks

(a)
G1 G2

ZA ZB

1 2
V1 V2

ZD

ZE ZC

V3
3

(b)
Reference Bus
0

I1 YA I2 YB

1 2
V1 V2

YD

YE YC

V3
3

FIGURE 2.1 (a) A three-bus power network. (b) Norton equivalent circuit of the three-bus
power network.
Analysis of Faults in Power System 27

and application of KCL at node 3 (bus-3) yields,

0 = (V3 − V1 ) YE + (V3 −V2 ) YC (2.3)

Rearranging Equations (2.1)–(2.3),

I1 = V1 (YA +YD +YE ) −V2YD −V3YE (2.4)

I 2 = −V1YD +V2 (YB +YC +YD ) −V3YC (2.5)

0 = −V1YE −V2YC +V3 (YC +YE ) (2.6)

In matrix form, Equations (2.4)–(2.6) can be represented as

˛ I1 ˆ ˛ (YA + YD +YE ) −YD −YE ˆ˛ ˆ

˙ ˘ ˙ ˘ ˙ V1 ˘
˙ I2 ˘=˙ −YD (YB + YC +YD ) −YC ˘ ˙ V2 ˘ (2.7a)
˙ 0 ˘ ˙ −YE −YC
˘˙
(YC +YE ) ˘ˇ ˝ V3 ˘
˝ ˇ ˙˝ ˇ

˜ I1 ˝ ˜ Y11 Y12 Y13 ˝ ˜ V1 ˝

˛ ˆ ˛ ˆ˛ ˆ
i.e., ˛ I 2 ˆ = ˛ Y21 Y22 Y23 ˆ ˛ V2 ˆ (2.7b)
˛ 0 ˆ ˛ Y31 ˆ˛ ˆ
° ˙ ° Y32 Y33 ˙ ° V3 ˙

˛ I1 ˆ ˛ V1 ˆ
˙ ˘ ˙ ˘
i.e. ˙ I 2 ˘ = [YBUS ] ˙ V2 ˘. (2.7c)
˙ 0 ˘ ˙ ˘
˝ ˇ ˝ V3 ˇ

Here Y11 = (YA +YD +YE ), Y12 = −YD , Y13 = −YE , and so on.
[YBUS ] is called bus admittance matrix where Y11, Y22 and Y33 are the diagonal ele-
ments and Y12, Y13, Y21, Y23 , Y31 and Y32 are the off-diagonal elements of the bus admit-
tance matrix. The diagonal elements are also termed as self-admittances, and each
of the elements is the summation of all the admittance connected to the respective
bus, while the off-diagonal elements are termed as transfer admittances, with each
representing the admittances connected between the concerned bus and other buses.
Obviously,
n

Yii = ˜y ij

j=0
j ˛i

and, Yij = −yij

For the power networks, [YBUS ] is a square matrix of order n × n, n being the number
of buses. Moreover, it may be noted that [YBUS ] is symmetrical as yij = y ji . Because
28 Power Transmission System Analysis Against Faults and Attacks

the elements of [YBUS ] matrix are complex numbers in power networks, [YBUS ]
matrix itself is complex. In addition, Yij (i ˛ j ) = 0 if the i th bus is not connected to
bus j through a line. In real-life power system, several interconnections do not exist
between a number of buses, and hence the [YBUS ] matrix becomes highly sparse (con-
taining a number of zero elements in the matrix). This saves a lot of computer storage
and memory requirements.

Example 2.1

A three-bus power network is presented in Figure 2.2. The reactances marked in

the fgure for the interconnecting links are in p.u. Form the [YBUS ] matrix.

Solution

From Figure 2.2, the line admittances can be obtained as

1
y12 = = − j2 p.u. = y21
j0.5

1
y13 = = − j1.25 p.u. = y31
j0.8

1
y23 = = − j2.5 p.u. = y32
j0.4

Y11 = y12 + y13 = − j3.25 p.u.; Y12 = −y12 = j2 p.u. = Y21; Y13 = −y13 = j1.25 p.u. = Y31

Y22 = y21 + y23 = − j4.5 p.u.;  Y23 = −y23 = j2.5 = Y32 ;  Y33 = y31 + y32 = − j3.75 p.u.

1 2

j0.5

j0.8 j0.4

FIGURE 2.2 A three-bus power system.

Analysis of Faults in Power System 29

˙ Y Y12 Y13 ˘˙ − j3.25 j2 j1.25 ˘

ˇ 11 ˇ 
˛ [YBUS ] = ˇ Y21 Y22 Y23 ˇ j − j4.5 j2.5  p.u.
ˇ Y31 Y32 Y33 ˇ j1.25 j2.5 − j3.75 
ˆ ˆ 

Example 2.2

A three-bus power system is shown in Figure 2.3 indicating the p.u. line reactances
of each line. Find the bus admittance matrix.

Solution

Let the ground bus be numbered as 0.

1 1
˜ y10 = = = − j10 p.u.
Z10 j0.1

1 1
y12 = = = − j5p.u. = y21
Z12 j0.2

1 1
y23 = = = j20 p.u. = y32
Z 23 − j0.05

1 1
y30 = = = − j10 p.u.
Z 30 j0.1

Also, Y11 = y10 + y12 = − j10 − j5 = − j15 p.u.

Y12 = −y12 = j5 p.u. = Y21

Y13 = −y13 = 0 = Y31

1 2 3

j0.2 -j0.05

j0.1 j0.1

FIGURE 2.3 A three-bus power system.

30 Power Transmission System Analysis Against Faults and Attacks

Y22 = y20 + y21 + y23 = 0 + (− j5) + j20 = j15 p.u.

Y23 = −y23 = − j20 p.u. = Y32

Y33 = y30 + y32 = − j10 + j20 = j10 p.u.

˙ Y Y12 Y13 ˘ ˙ − j15 j5 0 ˘

ˇ 11 ˇ 
˛ [YBUS ] = ˇ Y21 Y22 Y23  ˇ j5 j15 − j20  p.u.
ˇ Y31 Y32 Y33 ˇ 0 − j20 j10 
ˆ ˆ 

It may be noted for this problem, in the given data, the reactance between buses 1
and 2 is (j0.2) p.u. while that between buses 2 and 3 is (−j0.05) p.u. It indicates that
the reactance between buses 1 and 2 is inductive, while the reactance between
buses 2 and 3 is capacitive. Moreover, there is no reactance value given between
bus 2 at the reference bus. This makes y20, the admittance between bus 2 and the
reference bus, zero.

2.3 FORMATION OF [YBUS] WITH LINE TRANSFORMERS PRESENT

In real-life power systems, the substation is an integral part of the power network.
The transmission lines are connected with transformers in the substations for ease of
transmission and distribution of the electrical energy. Hence in [YBUS ] formation, the
transformers need to be included as an integral element with the transmission sys-
tem. Similar to lines, the transformers are represented by their complex impedances
along with proper representation of transformation ratio.
Figure 2.4 represents inclusion of a regulating transformer in the π model of a line
connecting bus i and j. The transformer has complex transformation ratio a:1 (a  = 
a ∠ α°). It may be noted that the π model of the line is included at the nonunity side
of the transformer.
Because the transformer is assumed to have the complex off-nominal tap ratio a:1,
assuming the transformer to be loss-less,

Primary voltage
i Vj j
Vi Vpri=aVj(=V'i)

Si yij
I'i Sj

Ii Ij
Ij / a*
yi0 yj0 Secondary voltage
Vsec=Vj
(a:1)

FIGURE 2.4 Representation of a regulating transformer along with the line model.
Analysis of Faults in Power System 31

Vi°
= a = a ˝˜
Vj

or, Vi˛( i.e., Vpri ) = aV j (2.8)

Also, when input power is equal to output power for the transformer, we have,

Vi˜I ˜i*= V j I *j

Vi˜ I*
or, = a = j*
Vj I˜i

°˛I i˜ is the primary current of transformer while I j is the secondary current ˝˙

Ij
I˜i*= (2.9)
a*

Next, we consider the current balance at two buses by the following two equations,

Si*
Ii = = Vi yi0 + (Vi − aV j ) yij = Vi y0 + (Vi − aV j ) yij (2.10)
Vi*

Ij Si*
and = = aV j y j 0 + ( aV j − Vi ) yij = aV j y0 + ( aV j −Vi ) yij
a* ( aV j )*

[assuming yi0 = y j0 = y0 ]
˛ I j = (−a* yij ) Vi + aa* ( y0 + yij ) V j (2.11)

Let us now rewrite Equations (2.10) and (2.11) in pair form as follows,

I i = ( y0 + yij ) Vi + (−ayij ) V j

I j = (−a* yij ) Vi + aa* ( y0 + yij ) V j

In matrix form, these two equations can be represented as

˛ Ii ˆ ˛ y0 + yij −ayij ˆ˛ ˆ
˙ ˘=˙ ˘ ˙ Vi ˘ (2.12)
˙˝ I j ˘ˇ ˙˝ −ayij aa* ( y0 + yij ) ˘ ˙˝ V j ˘ˇ
ˇ

or, [ I ] = [Y ] [V ]
32 Power Transmission System Analysis Against Faults and Attacks

˘ y0 + yij −ayij 
where, [Y ] =   (2.13)
 −ayij aa* ( y0 + yij ) 
 

It may be noted that if (a) is complex, then [Y ] is not symmetric. If (a) be a real quan-
tity, that is, (a) = (kV)base/(kV)tap, then

˘ y0 + yij −ayij 
[Y ] =   (2.14)
 −ayij a 2 ( y0 + yij ) 
 

Thus, the matrix [Y ] becomes symmetric.

Figure 2.5 represents the inclusion of regulating transformer in the line model.
Here the line model is assumed to be in the nonunity side of the complex transforma-
tion ratio (1:a).
With reference to Figure 2.5
V j°
= a = a ˝˜ (2.15)
Vi

i.e., V j˜ = aVi

Also, Vi˛I i* = ( aVi ) I˛j*

(Power is equal at the transformer input and output)

Ij̃ is the secondary current of transformer, while I i is the primary current.

I˜j* 1
=
I i* a

I i = a* I °j (2.16)

Primary voltage
Secondary voltage Vj
Vi Vpri=Vi
i j
Vsec=V'j=aVi
Ij / a*
Si
I'j yij Sj

Ij
Ii
Ij / a*
yi0 yj0

(1:a)

FIGURE 2.5 Inclusion of a regulating transformer with off-nominal tap ratio (1:a) in the
line model.
Analysis of Faults in Power System 33

At bus i

I i = a* I ˝j = a* ˇ˘ aVi y0 + ( aVi − V j ) yij 

or, I i = aa* y0Vi + aa* yijVi − a* yijV j

˛ I i = aa* ( y0 + yij ) Vi − a* yijV j (2.17)

Also, at bus j

I j = y0V j + (V j − aVi ) yij [assuming yi0 = y j0 = y0 ]

˛ I j = y0V j + (V j − aVi ) yij (2.18)

In matrix form, Equations (2.17) and (2.18) can be rearranged as

˛ Ii ˆ ˛ aa* ( y0 + yij ) −a* yij

ˆ˛ ˆ
˙ ˘=˙ ˘ ˙ Vi ˘ (2.19)
˙˝ I j ˘ˇ ˙˝ −ayij y0 + yij ˘ Vj
ˇ ˝˙ ˘ˇ

 
aa* ( y0 + yij ) −a* yij
˙ [Y ] =   (2.20)
 −ayij y0 + yij 
 

It may be noted that (a) being real,

˘ 
a 2 ( y0 + yij ) −ayij
[Y ] =   (2.21)
 −ayij y0 + yij 
 

the [Y ] matrix becomes symmetrical.

[In practical cases, the regulating transformer is designed for either voltage mag-
nitude or phase angle control. In the former case, ˜ = 0 and a can be changed in
discrete steps of ˜ a . In the latter case, a is constant and ˜ is changed in discrete
steps of ˜ a .]
In the next step [YBUS ], can now be modifed with inclusion of the regulating trans-
former. Figure (2.6a) represents the ˜ equivalent model with the regulating trans-
former having off-nominal tap ratio (a:1), while Figure (2.6b) represents the same
with off-nominal tap ratio (1:a); the line model is always placed at the nonunity side
of the transformer. It may be noted that here Yse , Ysh1 and Ysh2 in respective fgures are
used to form Y12 (or Y21) as well as Y11 and Y22 .

In Figure ( 2.6a ) , Yse = ayij ; Ysh1 = ( y0 + yij ) + (−ayij ) = y0 + (1 − a ) yij ;

34 Power Transmission System Analysis Against Faults and Attacks

(a) Yse=ayij

i j

Ysh1=y0+(1-a)yij yj0 Ysh1=a2y0+a(a-1)yij

(b) Yse=ayij

i j

Ysh2=a2y0+a(a-1)yij Ysh2=y0+(1-a)yij
yj0

FIGURE 2.6 (a) Equivalent circuit model with transformer having off-nominal turns ratio
(a:1). (b) Equivalent circuit model with transformer having off-nominal turns ratio (1:a).

Ysh2 = a 2 ( y0 + yij ) + (−ayij ) = a 2 y0 + a ( a −1) yij

In Figure ( 2.6b) , Yse = ayij ; Ysh1 = a 2 ( y0 + yij ) + (−ayij ) = a 2 y0 + a ( a −1) yij

Ysh2 = ( y0 + yij ) + (−ayij ) = y0 + (1− a ) yij

The [YBUS ] matrix can be modifed for inclusion of line with transformer with revised
form of its self (diagonal) and transfer (off-diagonal) elements as given below:

Yii (new) = Yii (old) +Yse +Ysh1

Yij (new) = Y ji (new) = Yij (old) −Yse (2.22)

Y jj (new) = Y jj (old) +Yse +Ysh2

where Yii (old), Y jj (old) and Yij (old) are the elements of [YBUS ] formed without considering
line transformers.

Example 2.3

A three-bus power system is shown in Figure 2.7a. Assume an ideal transformer is

connected between buses 2 and 3 in series with a line reactance j0.5 p.u. Find [YBUS ].
The line data for the given system is shown below (neglect shunt charging effect)
Analysis of Faults in Power System 35

(a)

1 2

z12
V2

V'3
z31 z23

V'3/V2=1.02:1 (or, V2/V'3=1:1.02)

(b)

1 2

z12

z31

(c) Yse

2 3

Ysh1 Ysh2

FIGURE 2.7 (a) A three-bus three-line power system with transformer in line (23). (b) System
before considering the line with transformer. (c) “Pi” model of the line with transformer.

Line no. From Bus To Bus R (in p.u.) X (in p.u.) Off-nominal Tap Ratio of
Transformer
1 1 2 0.05 0.15 -
2 1 3 0.05 0.15 -
3 2 3 0 0.5 1:1.02
36 Power Transmission System Analysis Against Faults and Attacks

Solution

From the circuit of the given power system,

z12 = z13 = ( 0.05 + j0.15) p.u.

1
˛ y12 = y13 = = ( 2 − j6) p.u.
0.05 + j0.15

˘ 
 (4 − j12) (−2 + j6) (−2 + j6) 
[YBUS ] =  (−2 + j6) (2 − j6) (0 + j0)  p.u.
 
 (−2 + j6) (0 + j0) (2 − j6) 

(Y11 = y12 + y31 = (2 − j6) + (2 + j6) = (4 − j12) p.u.)

With regulating transformer placed between buses 2 and 3, the [YBUS ] obtained
needs to be modifed (Figure 2.7b and c).

1
Here, y23 = = − j2
j0.5

˛Y23 = ay23 = 1.02 × (− j2) = − j2.04 p.u.

Y22 = a ( a −1) y23 [ y0 = 0 + j0 ] = 1.02 × (1.02 − 1) × (− j2) = − j0.0408 p.u.

Y33 = (1 − a ) y23 = (1 − 1.02) × (− j2) = j0.04 p.u.

˜ For the transformer in line 2–3,

YBUS22 = YBUS22old + (− j2.04 ) + (− j0.0408) = ( 2 − j6) − j2.0808 = ( 2 − j8.0808) p.u.

YBUS33 = YBUS33old + (− j2.04 ) + ( j0.04 ) = ( 2 − j6) − j2 = (− j8) p.u.

YBUS23 = YBUS32 = − (− j2.04 ) = j2.04 p.u.

 
 (4 − j12) (−2 + j6) (−2 + j6) 
˙ [YBUS ] M =  (−2 + j6) (2 − j8.0808) ( j2.04)  p.u.
 
 (−2 + j6) j2.04 (2 − j8) 

[YBUS ] M represents the modifed [YBUS ] with the transformer included in line
between buses 2 and 3.
Analysis of Faults in Power System 37

2.4 CONCEPT OF SYMMETRICAL COMPONENT ANALYSIS

It is relatively simple to analyse a three-phase circuit in which phase voltages and
currents are balanced (of equal magnitude in three phases and displaced 120° from
each other), and in which the connected circuit elements are symmetrical because
the treatment of a single-phase leads directly to the three-phase solution in such a
case. The analysis of the circuit which is not symmetrical as a result of unbalanced
load, unbalanced faults or short circuits or unbalance in supply system can effectively
be conducted by “symmetrical component” analysis method frst presented by C.L.
Fortescue and developed by Edith Clarke.
The fundamental principle of symmetrical components as applied to the three-
phase circuits is that a set of three phasors forming a three-phase unbalanced system
(Figure 2.8a) can be resolved into:

a. a set of three phasors, equal in magnitude, displaced from each other by

120° in phase and having the same phase sequence as the original phasors.
It forms a three-phase balanced system of positive sequence (Figure 2.8b).
b. a set of three phasors, equal in magnitude, displaced from each other by 120°
in phase and having the phase sequence opposite to that of the original phasors
forming a three-phase balanced system of negative sequence (Figure 2.8c).
c. a set of three phasors, equal in magnitude and with zero-phase displacement
from each other (Figure 2.8d).

(a) Va (b)
Vc1 Va1
120 0

1200 1200
Vc
Vb

Vb1
(c) (d)
Vb2 Va2
120 0

120 0 120 0

Va0= Vb0= Vc0

Vc2

FIGURE 2.8 (a) Unbalanced phasors, (b) positive-sequence components of the phasors,
(c) negative-sequence components of the phasors, and (d) zero-sequence components of the
phasors.
38 Power Transmission System Analysis Against Faults and Attacks

˜ Va1 , Va2 , Va 0 are known as the symmetrical components or, more specifcally, the
positive-sequence component, negative-sequence component and zero-sequence
component, respectively, of Va, and similarly Vb1 , Vb2 , Vb0 for Vb and Vc1 , Vc 2 , Vc 0 for Vc .

2.5 OPERATOR “A”

In Figure 2.9, the three phasors OP, OQ OR, are shown where each one is displaced
from the other by 120°. A phasor OP/ can then be represented as OP  =  OP (cos θ  +
j sin θ) = (OP) e j˜ , with OP being the reference phasor.
Similarly, the other two phasors can be expressed as OQ  =  OP e j120° and OR  =
OP e j240°  = OP e− j120° .
Let “a” be an operator that rotates the phasor it operates upon through 120° in a
counter-clockwise direction.
Applying this defnition,

OQ = OP × a = a ∙ OP
OR = OP × a × a = a2 ∙ OP

Thus, it is possible to list different values of “a” as shown below:

a = e j120° = ( cos120° + jsin120°) = (−0.5+ j0.866)

a 2 = e j240° = ( cos 240° + j sin 240°) = (−0.5 − j0.866) = e− j120°

a 3 = e j360° = ( cos 360° + jsin 360°) = 1

a 4 = a 3 × a = (−0.5+ j0.866)

Q
P'

120 0

P
ref
120 0
120 0

FIGURE 2.9 Rotational representation of three phasors.

Analysis of Faults in Power System 39

a 5 = a 3 × a 2 = (−0.5 − j0.866)

1 + a + a 2 = 0 = 0 + j0

a − a 2 = j1.732 = j 3 = 3˛90°

a + a 2 = −1

1 + a = ( 0.5+ j0.866) = −a 2

1− a = (1.5 − j0.866) = 3˙− 30°

2.6 SYMMETRICAL COMPONENT REPRESENTATION

Applying the concept of “a” in designating the sequence components, the positive-
sequence component system (Figure 2.8b) can be expressed as

Va1 = Va1; Vc1 = aVa1; Vb1 = a 2Va1

Similarly, the symmetrical components of the negative- and zero-sequence compo-

nents system can be expressed as

Va2 = Va2 ; Vb2 = aVa2 ; Vc2 = a 2Va2

and Va0 = Va0 ; Vb0 = Va0 ; Vc0 = Va0

Furthermore, from the concept of symmetrical component theory, the unbalanced

voltages Va ,Vb and Vc can be written as phasor summation of individual symmetrical
component phasors.

i.e. Va = Va1 +Va2 +Va0

Vb = Vb1 +Vb2 +Vb0 (2.23)

Vc = Vc1 +Vc2 +Vc0

Equation (2.23) can further be represented as

Va = Va1 +Va2 +Va0

Vb = a 2Va1 + aVa2 +Va0 (2.24)

Vc = aVa1 + a 2Va2 +Va0

40 Power Transmission System Analysis Against Faults and Attacks

that is, in matrix form,

˛ Va ˆ ˛ 1 1 1 ˆ˛ V ˆ ˛ Va0 ˆ
˙ ˘ ˙ ˘ ˙ a0 ˘ ˙ ˘
˙ Vb ˘=˙ 1 a2 a ˘ ˙ Va1 ˘ = [ A] ˙ Va1 ˘ (2.25a)
˙ Vc ˘ ˙ a2 ˘ ˙ Va2 ˘ ˙ Va2 ˘
˝ ˇ ˝ 1 a ˇ˝ ˇ ˝ ˇ

˝ 1 1 1 ˇ
ˆ 
where [ A] = ˆ 1 a2 a 
ˆ 1 a a2 
˙ ˘

Similar to Equation (2.25a), the unbalanced phase sequence can also be represented
in terms of sequence components (Equation 2.25b)

˜ I ˝ ˜ 1 1 1 ˝˜ I ˝
˛ a ˆ ˛ ˆ ˛ a0 ˆ
˛ Ib ˆ=˛ 1 a2 a ˆ ˛ I a1 ˆ (2.25b)
˛ Ic ˆ ˛ a2 ˆ ˛ Ia2 ˆ
° ˙ ° 1 a ˙° ˙

Also, from Equation (2.25a), we can write

˛ Va0 ˆ ˛ Va ˆ
˙ ˘ −1 ˙ ˘
˙ Va1 ˘ = [ A] ˙ Vb ˘ (2.26a)
˙ Va2 ˘ ˙ Vc ˘
˝ ˇ ˝ ˇ

˝ 1 1 1 ˇ
1ˆ −1 
where, [ A] = ˆ 1 a a2 
3ˆ 
˙ 1 a2 a ˘

and from Equation (2.25b), we have

˛ I a0 ˆ ˛ Ia ˆ
˙ ˘ −1 ˙ ˘
˙ I a1 ˘ = [ A] ˙ I b ˘ (2.26b)
˙ I a2 ˘ ˙ Ic ˘
˝ ˇ ˝ ˇ

2.6.1 DETERMINATION OF SEQUENCE COMPONENTS

Va + Vb + Vc = (1 + a 2 + a ) Va1 + (1 + a + a 2 ) Va2 + 3Va0 = 3Va0 ˙ˆ (1 + a 2 + a ) = 0ˇ˘

1
˛ Va0 = (Va + Vb + Vc ) (2.27)
3
Analysis of Faults in Power System 41

Also,

Va + a 2Vb + aVc = Va1 (1 + a 4 + a 2 ) +Va2 (1+ a 3 + a 3 ) + Va0 (1+ a 2 + a ) = 3Va2

1
˛ Va2 =
3
(Va + a2Vb + aVc ) (2.28)

and, Va + aVb + a 2Vc = Va1 (1+ a 3 + a 3 ) +Va 2 (1+ a 2 + a 4 ) +Va 0 (1+ a + a 2 ) = 3Va1

1
˛ Va1 =
3
(Va + aVb + a2Vc ) (2.29)

Thus, for any unbalanced system of voltages,

1
Va1 =
3
(Va + aVb + a2Vc )
1
Va2 =
3
(Va + a2Vb + aVc ) (2.30a)

1
Va0 = (Va + Vb + Vc )
3

Similarly, for any unbalanced system of currents,

1
I a1 =
3
( I a + aI b + a2 Ic )
1
Ia2 =
3
( I a + a2 I b + aIc ) (2.30b)

1
I a0 = ( I a + I b + I c )
3
From Equation (2.30a), it follows that there will be no zero-sequence component
of voltage when the phasor sum Va + Vb + Vc = 0. Moreover, there will be no zero-
sequence component of current when the phasor sum I a + I b + I c = 0. Because the
phasor sum of the line-to-line voltages in a three-phase balanced system is always
zero, there will be no zero-sequence component of the line voltages for a three-phase
balanced system. Also, because the phasor sum of all the three-line currents in the
three-wire three-phase system is zero, there will also be no zero-sequence compo-
nent of the line current for a three-phase three-wire system.
Zero-sequence components may, however, be present in the phase voltages of a
star-connected circuit or in the phase currents in a delta-connected circuit. In a star-
connected circuit, the zero-sequence components of the phase voltages are all in
phase. The line voltage being the phasor difference of two respective phase voltages,
the zero-sequence components cancel out when line voltages are determined. In a
42 Power Transmission System Analysis Against Faults and Attacks

delta-connected circuit, the zero-sequence components of the phase currents are all
in phase and form a local circulating current inside the delta. Therefore, there is no
zero-sequence component in the delta line currents.

2.7 CONCEPT OF SEQUENCE IMPEDANCES

The sequence impedance is defned as an impedance that is obtained by passing
unit current of that sequence in a three-phase network or equipment and obtaining
the equation for voltage drop. The sequence impedance is expressed normally in
p.u. quantities. The impedance of one sequence component need not be the same to
the other. Linear networks such as transmission lines normally have equal positive-
and negative-sequence impedances. Transformers also have identical positive- and
negative-sequence impedances, but different zero-sequence impedance. On the
other hand, sequence impedances of rotating machines are the transient (X d̃ ) or
sub-transient (X d̃˜) reactances depending on the time of interest. Frequently, only
the sequence reactance is used to calculate the fault currents and the resistance
component of the impedance is neglected as its value is much less than that of
the corresponding reactance. (X d̃˜) is normally used for calculating fault currents,
while (X d̃ ) may be used to calculate currents used to set time-delayed protection
devices.

2.8 SEQUENCE COMPONENT MODELS OF

PRINCIPAL POWER SYSTEM EQUIPMENT
2.8.1 GENERATOR SEQUENCE MODEL
Figure 2.10a represents the equivalent circuit of an alternator for transient conditions.
Let Va = Vb = Vc and Z1 = Z 2 = X d°°. The positive-sequence network contains Va and Z1,
as shown in Figure 2.10b(i). Figure 2.10b(ii) contains only the negative sequence
impedance which is mostly equal to the positive-sequence impedance.
The zero-sequence network (Figure 2.10b(iii)) includes the neutral impedance
and the machine zero-sequence impedance (Z og ). Usually, Z og varies between 0.1 and
0.7 times of (X d̃˜). The zero-sequence current fows in each phase of the generator and
the neutral current is the sum of phase currents. Because zero-sequence current pha-
sors have no angular offset between them (i.e. they are in phase), the neutral current
is three times the current in each phase. The drop (Va0 ) can be expressed as

Va0 = (3I a 0 ) Z n + I a0 Z og = I a 0 (3Z n + Z og ) = I a 0 ˙ Z 0 (2.31)

2.8.2 SEQUENCE NETWORK MODEL OF TRANSFORMERS

Transformers are static electrical machines and their positive- and negative-sequence
impedances are the same to each other. However, the 30° phase shift in a delta star
transformer needs to be considered in the sequence network models of transformers.
For example, the H.V. side of the transformer is delta-connected while the L.V. side
is star-connected (Figure 2.11).
Analysis of Faults in Power System 43

(a)

Ia Xd" Ib Xd" Ic X d"

Va Vb Vc

Zn In

(b)
Reference Bus Reference Bus Reference Bus

- Zog
Va Va1 Z2=Xd" Va2 Z0 Va0
3Zn
+ Ia1
Ia2 Ia0

Z1=Xd" Zog (0.1 to 0.7)Xd"

(i) positive sequence (ii) negative sequence (ii) zero sequence

FIGURE 2.10 (a) Equivalent circuit of a generator; (b) sequence network of generator
impedances: (i) positive sequence, (ii) negative sequence, (iii) zero sequence.

HV LV

FIGURE 2.11 (−30°) Phase shift of delta star transformer.

Figure 2.11 represents the phase shift of 30° for the secondary currents. As per
the ANSI standard, for positive sequence quantities for Y − ° or ˜ −Y transformers,
the positive-sequence voltage to neutral on the H.V. side always leads the positive-
sequence voltage to neutral on the low voltage side by 30°. Thus, when stepping up
from the low-voltage side to the high-voltage side of a ˜ −Y or a Y − ° transformer,
we need to advance the positive-sequence voltages (and currents) by 30° and retard
44 Power Transmission System Analysis Against Faults and Attacks

the negative-sequence voltages (and currents) by 30° as the negative sequence is the
reverse sequence rotation of the positive-sequence quantities.
The zero-sequence impedance of a transformer differs from the positive- and
negative-sequence impedances. For an ungrounded neutral system, there is no fow
of current to ground. Thus, for ungrounded star or delta-connected transformer,
Z 0 ˜ °. Figure 2.12 represents the zero-sequence impedance of common trans-
former connection types.

Connection Zero Sequence Zero Sequence

Supply(S) Lo ad(L) Equivalent Circuit Impedance from
Load Side

( ) =∞

( ) =∞

( )

( ) =∞

( )

( ) =∞

( ) =∞

( ) =∞

FIGURE 2.12 Zero-sequence equivalent circuits for three-phase transformer banks.

Analysis of Faults in Power System 45

2.8.3 SEQUENCE IMPEDANCE CIRCUIT FOR TRANSMISSION LINES

The positive- and negative-sequence impedance of the transmission line (a linear
network) are equal and are the given line impedances. The zero-sequence impedance
is different and sometimes as a thumb rule Z 0 = 3Z1.

2.8.4 SEQUENCE IMPEDANCE CIRCUIT FOR MOTORS

In systems where there are large motors, there is contribution of fault current by the
motor. When a fault occurs, the motor connected to the system continues to rotate for
a short period due to the load inertia. This rotation of the motor due to load inertia
causes the motor to operate as a generator. This effect continues till the motor stops
or disconnects from the system. Thus, during fault period, both the generator and the
motor deliver fault current for the initial period. This effect is predominant for large
motors only.
The equivalent circuit of a synchronous motor is identical to that of a synchronous
generator. The positive- and negative-sequence network for the induction motor are
given by the relation (X1) ≈ (X2) ≈ (Xd) ≈ 0.15 - 0.2 p.u.
The zero-sequence impedance of the induction motor is infnite because the
induction motors are not normally grounded at neutral until the motor phase voltage
exceeds 4160 V. At a phase voltage beyond 4160 V, the winding may be subjected to
corona effect, and hence, the motor neutral needs to be grounded. The zero-sequence
impedance of a grounded neutral induction motor is not infnite, and, in such cases,
the manufacturer states the value of the zero-sequence impedance.

2.9 THREE-PHASE POWER IN TERMS OF

SYMMETRICAL COMPONENTS
Let S be the complex power fowing into a three-phase circuit through the phases
r, y and b, while P and Q are the real and reactive power components of S. Let Vr,
Vy and Vb be the respective phase voltages, while Ir, Iy and Ib are the corresponding
three-phase currents.

Here, S = P + jQ = Vr I r * +Vy I y* +Vb I b*

* T *
° Ir ˝ ° Vr ˝ ° I ˝
˝ ˆˆ I y ˇ ˆ ˇ ˆ r ˇ
i.e. S = ° Vr Vy Vb ˇ = ˆ Vy ˇ ˆ Iy ˇ (2.32)
˛ ˙
ˆ I ˇ ˆ ˇ ˆ I ˇ
˛ b ˙ ˛ Vb ˙ ˛ b ˙

However,

˛ V ˆ ˛ Vr 0 ˆ
˙ r ˘ ˙ ˘
˙ Vy ˘ = [ A] ˙ Vr1 ˘ (2.33)
˙ V ˘ ˙ ˘
˝ b ˇ ˝ Vr 2 ˇ
46 Power Transmission System Analysis Against Faults and Attacks

˝ 1 1 1 ˇ
ˆ 
where, [ A] = ˆ 1 a2 a 
ˆ 1 a a2 
˙ ˘

˛ I ˆ ˛ Ir 0 ˆ
˙ r ˘ ˙ ˘
and, ˙ I y ˘ = [ A] ˙ I r1 ˘ (2.34)
˙ I ˘ ˙ Ir 2 ˘
˝ b ˇ ˝ ˇ

Utilising Equations (2.33) and (2.34) in Equation (2.32)

T *
˝ Vr 0 ˇ ˝ Ir 0 ˇ
Tˆ  *ˆ 
S = [ A] ˆ Vr1  [ A] ˆ I r1  (2.35)
ˆ  ˆ 
˙ Vr 2 ˘ ˙ Ir 2 ˘

However, in Equation (2.35), [ A] [ A] = 3u, u being the unity matrix.

T *

˝ 1 0 0 ˇ
ˆ 
[u ] = ˆ 0 1 0 
ˆ˙ 0 0 1 ˘

*
˛ Ir 0 ˙
˜ S = 3 ˝˛ Vr 0 Vr1 Vr 2 ˙ˇ ˘
ˆ ˇ I r1 ˘
ˇ Ir 2 ˘
˝ ˆ

or S = 3 ˛V
˝ r 0 I r 0 + Vr1 I r1 + Vr 2 I r 2 ˙ˆ
* * *
(2.36)

that is, the sequence power is one third the power in terms of phase quantities.

Example 2.4

Figure 2.13 represents an unbalanced three-phase source having phase voltages

Va , Vb , and, Vc feeding an unbalanced load through a set of three reactances. If the
phase voltages are given by Va = 100°0°; Vb = 100° − 90° and Vc = 100°120°  V , fnd
the sequence currents.

Solution

Let Xa, Xb and Xc represent the total reactances present in each phase of the given
circuit.
Analysis of Faults in Power System 47

FIGURE 2.13 An unbalanced three-phase system.

X a = ( 0 + j1) + ( 0 + j1) = j2˙

X b = ( 0 + j1) + ( 0 + j2) = j3˙

X c = ( 0 + j1) + ( 0 + j3) = j4˙

Va 100˛0°
˜ Ia = = = − j50 A = 50˛ − 90° A
Xa j2

Vb 100° − 90°
Ib = = = 33.33° −180°A
Xb j3

Vc 100°120°
Ic = = = 25°30°A
Xc j4

1
˛ I a1 =
3
( I a + aI b + a 2 Ic )
1
= (50ˆ − 90° + 33.33ˆ (−180° +120°) + 25ˆ(30° + 240°
3
1
= (50ˆ − 90° + 33.33ˆ − 60° + 25ˆ270°
3
= 5.55 − j34.62 = 35.06ˆ − 80.88°A

1
I a2 =
3
( I a + a 2 I b + aIc )
1
= (50˙ − 90° + 33.33˙ (−180° + 240°) + 25˙(30° +120°
3
= −1.66 − j2.88 = 3.32˙ −120°A
48 Power Transmission System Analysis Against Faults and Attacks

1
I a0 = ( I a + I b + I c )
3
1
= (50˙ − 90° + 33.33˙ −180° + 25˙30°)
3
= −3.89 − j12.5 = 13.09˙ −107.299°A

Example 2.5

In the circuit diagram of Figure 2.14, fnd the sequence currents.

Given, I R = 1°0° ; IY = 1°180°; I B = 0

Solution

1
I r1 =
3
( I R + aIY + a 2 I B )
1
=
3
(1˙0° +1˙ (180° +120°) + 0) = 0.5 − j0.2886 = 0.5773˙ − 30°

1
Ir 2 =
3
( I R + a 2 IY + aI B )
1
=
3
(1˙0° +1˙ (240° +180°) + 0) = 0.5 + j0.2886 = 0.5773˙30°

1 1
I r 0 = ( I R + IY + I B ) = (1˙0° + 1˙180° + 0 ) = 0
3 3

Example 2.6

In any three-phase circuit the power is denoted by the following relation

FIGURE 2.14 Circuit diagram of Example 2.5.

Analysis of Faults in Power System 49

˝ IA ˆ
P = ˝˙ VAN VBN VCN ˆ [K ] ˘ I B 
ˇ ˘ 
˘ 
˙ IC ˇ

where, VAN , VBN and VCN are the phase voltages, and I A , I B and I C are the line cur-
rents for a pure star-connected balanced load.

˙ 0 1 −1 ˘
1 ˇ 
If [ K ] = −1 0 1  and I A = I B = I C ,
3 ˇˇ 1 −1 0 
ˆ

Find the magnitude of P.

Solution

ˆ IA ˘
 
P = ˆˇ VAN VBN VCN ˘ [ K ]  I B 
 IC 
ˇ 

ˆ 0 1 −1 ˘ˆ IA ˘
˘ 1  −1  
= ˆˇ VAN VBN VCN  3 0 1  IB 
ˇ 1 −1 0   I C 
ˇ 

ˆV V V ˘
=  AN ( I C − I B ) + BN ( I A − I C ) + CN ( I B − I A )
ˇ 3 3 3 

˛ I A = I B = I C = I (say )

ˆV V V 
P = ˘ AN ( I − I ) + BN ( I − I ) + CN ( I − I ) = 0
ˇ 3 3 3 

Example 2.7

If symmetrical components of voltages and currents are given by

Va 0 = 10° − 30° V ; Va1 = 45°0°V ; Va 2 = 25°40°V

and I a0 = 5°190°A; I a1 = 10°25°A; I a2 = 1°50°A,

Find the complex power.

50 Power Transmission System Analysis Against Faults and Attacks

Solution

S3 = 3  [Va0 I a0 + Va1 I a1 + Va2 I a2 ]

= 3 [10ˇ − 30° × 5ˇ190° + 45ˇ0° × 10ˇ25° + 25ˇ40° × 1ˇ50°]

10(cos30° − j sin 30°) × 5 ( cos190° + j sin190°) + 45(cos 0° + j sin 0°) 

= 3 
×10 ( cos 25° + j sin 25°) + 25 ( cos 40° + j sin 40°) ×1(cos50° + j sin 50°)]
 

= 3 [360.85 + j232.28]

= 1082.56 + j696.84 = 1287.448ˇ32.77°

2.10 [ZBUS] BUILDING (STEP-BY-STEP METHOD)

During the formulation of [YBUS ], [ Z BUS ] can be obtained by inversion of [YBUS ].
However, this technique may consume more computer time and memory once the
size of the system is large and the [YBUS ] matrix is extremely sparse. Moreover, any
modifcation in the network needs reformation of [YBUS ], and for large systems this is
clearly not convenient.
On the other hand, the process of [ Z BUS ] building with step-by-step method is
more convenient and the technique starts working with branch impedance values.
The algorithm starts from scratch and any modifcation in the network does not
require a complete rebuilding of [ Z BUS ].
Once the reference node is defned and the network bus voltage is specifed with
respect to this node, a branch (Z b say) can be easily added to the existing impedance
matrix [ Z BUS ]old to produce the new impedance matrix. There are fve cases where
the addition of the branch can infuence the [ Z BUS ], depending on the type of Z b   with
system buses (i.e. whether with reference bus, new bus or old buses).

2.10.1 ADDING A BRANCH (OR LINK) ZB FROM A NEW BUS TO

THE REFERENCE BUS (TYPE 1 MODIFICATION)

The addition of new bus (k) to the reference bus (or node) through a link/branch of
impedance Z b (Figure 2.15) without any connection to any of the other buses of the
original network will not alter the original bus voltages, even if a new current Ik is
injected in the new bus (k). The voltage Vk at the new bus is given by

Vk = Z b I k

Obviously, Zik = Z ki = 0;  i = 1, 2,3,…, n (2.37)

Z kk = Z b
Analysis of Faults in Power System 51

n port
passive linear
network

FIGURE 2.15 Addition of link Z b from new bus (k) to an old bus (r).

Then, the nodal equation becomes

˝ V1 ˇ ˝ | 0 ˇ˝ I1 ˇ
ˆ  ˆ ˆ 
ˆ V2  ˆ [ Z BUS ]old | ˜ ˆ I2 
ˆ ˜  ˆ |

˜ ˆ ˜ 
ˆ Vn =ˆ ˆ In  (2.38a)
ˆ  ˆ | 0 ˆ 
ˆ …  ˆ       ˆ
 … 
ˆ Vk  ˆ ˆ Ik 
˙ ˘ ˙ˆ 0 ° ° 0 | Z b ˘ ˙ ˘

or [V ] = [ Z BUS ]new [ I ]

ˆ | 0 
˘ 
˘ [ Z BUS ]old | ˜ 
˘ 
|
=˘
˜ 
Here, [ Z BUS ]new
˘ 
˘ | 0  (2.38b)
˘ ˙ ˙ ˙ ˙ ˙ ˙ 
˘ 
ˇ 0 ° ° 0 | Zb 

It may be noted here that the column vector of the current multiplied by [ Z BUS ]new
does not alter the voltage of the original network.
52 Power Transmission System Analysis Against Faults and Attacks

2.10.2 ADDITION OF A BRANCH (OR LINK) ZB FROM A NEW

BUS TO AN OLD BUS (TYPE 2 MODIFICATION)
Let a branch of impedance Z b be added from a new bus (k) to an old bus (j). With
reference to Figure 2.16 the current I k entering into bus j would increase the original
voltage (V j ) of the bus j by a voltage I k Z b, that is,

Vk = Vj + Z b I k

= Z j1I1 + Z j 2 I 2 +  + Z jj ( I j +  I k ) ++ Z jn I n + Z b I k (2.39)

or Vk = Z j1I1 + Z j2 I 2 +  + Z jj I j +  + Z jn I n + ( Z jj + Z b ) I k

Thus, the new row must be combined with [ Z BUS ]old to fnd Vk ; this is
{ Z j1    Z j2  …Z jn   .( Z jj + Zb )}
Because [ Z BUS ] is to be a square matrix around the
principal diagonal, a new column is to be combined that is a transpose of the new
row. The new column accounts for the increase of all bus voltages due to I k. The
generalised matrix equation then becomes

ˆ | Z1 j 
ˆ V1  ˘ ˆ I1 
˘
˘ V2
 ˘
 ˘
[ Z BUS ]old | Z2 j ˘ I2

˘ 
˘ ˜  ˘ | ˜ ˘ ˜ 
˘ Vn =˘ | Z nj ˘ In  (2.40a)
˘  ˘ ˘ 
˘ …  ˘       ˘ … 
˘  ˘  ˘ˇ 
ˇ Vk  ˘ Z j1
ˇ
Z j2 ° Z jn | ( Z jj + Zb ) 
Ik 

n port
passive linear
network

FIGURE 2.16 Addition of a link Z b from new bus (k) to an old bus (j).
Analysis of Faults in Power System 53

or, [V ] = [ Z BUS ]new [ I ]

 | Z1 j 
 
 [ Z BUS ]old | Z2 j 
 
 | ˜ 
Here, [ Z BUS ]new =  (2.40b)
 | Z nj 
 ˇ ˇ ˇ ˇ ˇ ˇ 
 
 Z j1 Z j2 ° Z jn | ( jj + Z b )
Z 

[It may be observed here that the frst elements of the new row of [ Z BUS ]new are the
elements of row j of [ Z BUS ]old , and the frst n elements of the new column are the ele-
ments of column j of [ Z BUS ]old ]

2.10.3 ADDITION OF A BRANCH (OR LINK) ZB FROM AN OLD BUS

TO THE REFERENCE BUS (TYPE 3 MODIFICATION)

Let there be an addition of a branch (link) Z b from an old bus j to the reference bus
(Figure 2.17). This is done by short circuiting bus (k) to reference node (r) by mak-
ing Vk = 0. This will yield the same matrix as shown in Equation (2.40a) except for
Vk = 0.

n port
passive linear
network

FIGURE 2.17 Addition of a link (Z b ) from an old bus (j) to the reference bus (r).
54 Power Transmission System Analysis Against Faults and Attacks

This yields

ˆ | Z1 j 
ˆ V1  ˘ ˆ I1 
˘
V2
 ˘ [ Z BUS ]old | Z2 j ˘ I2

˘  ˘ ˘ 
˘ ˜  ˘ | ˜ ˘ ˜ 
˘ Vn =˘ | Z nj ˘ In  (2.41)
˘  ˘ ˘ 
˘ …  ˘       ˘ … 
˘
ˇ 0  ˘
 ˘ Z j1 Z j2 ° Z jn | ( Z jj + Zb )  ˇ˘ Ik 

ˇ 

A new (n  +  1)th row and a new (n  +  1)th column is obtained. It is now possible to
eliminate (n + 1)th row and (n + 1)th column by Kron reduction. Each element in the
new matrix will then become

Z m (n+1) Z(n+1)i
Z m (new) = Z mi − (2.42)
Z jj + Z b

or, [ Z BUS ]new = [ Z BUS ]old

ˆ Z1 j 
1 ˘  (2.43)
− ˘   ˇˆ Z j1 … Z jn 

Z jj + Z b ˘ Z 
nj
ˇ 

2.10.4 ADDITION OF A BRANCH (OR LINK) ZB BETWEEN

TWO OLD BUSES (TYPE 4 MODIFICATION)
Let a branch (link) Z b be added from an old bus i to another old bus j (Figure 2.18)
For bus 1,

V1 = Z11I1 + Z12 I 2 +  + Z1i ( I i + I k ) + Z1 j (I j − I k ) +  + Z1n I n (2.44)

Rearrangement of Equation (2.44) yields

V1 = Z11I1 + Z12 I 2 +  + ( Z1i − Z1 j ) I k +  + Z1n I n (2.45)

Similar equation can be written for all other buses.

Analysis of Faults in Power System 55

n port
passive linear
network

FIGURE 2.18 Addition of a link (Z b ) between two old buses (i and j).

Finally,

V j = Vi + Z b I k

= Z j1I1 + Z j2 I 2 +  + Z ji ( I i +  I k ) + Z jj ( I j −  I k ) + + Z jn I n (2.46)

= Zi1I1 + Zi 2 I 2 +  + Zii ( I i +  I k ) + Zij ( I j −  I k ) ++ Zin I n + Z b I k

Rearranging Equation (2.45),

0 = ( Zi1 − Z j1 ) I1 + + ( Zii − Z ji ) I i + ( Zij − Z jj ) I j + 

+  ( Zin − Z jn ) I n + ( Z b + Zii + Z jj − Z ji − Zij ) I k

In a matrix form,

ˆ V1  ˘
ˆ | ( Z1i − Z1 j ) 
ˆ I1 
˘  ˘ [ Z BUS ]old | ˜ ˘ 
˘ ˜  ˘ ˘ ˜ 
˘ Vn =˘ | (Z ni − Z nj ) ˘ In 
˘ …  ˘ ˘ … 
˘  ˘      ˘ 
˘ˇ 0  ˘ ( Zi1 − Z j1 ) ° ° | ( Zb + Zii + Z jj − 2Zij )  ˘ˇ Ik 
ˇ˘ 

(2.47)

Elimination of I k yields Equation (2.48)

56 Power Transmission System Analysis Against Faults and Attacks

˘ Z1i − Z1 j 
1  
[ Z BUS ]new = [ Z BUS ]old −   
( Zb + Zii + Z jj − 2Zij )  Zni − Znj 
 
(2.48)
× ˘ (Zi1 − Z j1 ) … (Zin − Z jn ) 
 

2.10.5 ADDITION OF TWO BRANCHES ZA AND ZB WITH MUTUAL

IMPEDANCE (ZM) BETWEEN FOUR BUSES (TYPE 5 MODIFICATION)
Let there be two branches Z a and Z b added with mutual impedance Z m   between four
buses j, k, l, and m, as shown in Figure 2.19. Here,

Vi = Zi1I1 + Zi2 I 2 +  + Zij ( I j − I a ) + Zik ( I k + I a )

+Zil ( I l − I b ) + Zim ( I m + I b ) +  + Zin I n (2.49)

n port
passive linear
network

FIGURE 2.19 Addition of two branches (Z a and Zb) with mutual impedance (Z m) between
four buses.
Analysis of Faults in Power System 57

On rearrangement, Equation (2.49) becomes

Vi = Zi1 I1 +  + Zij I j + Zik I k + Zil I l + Zim I m + 

+ Zin I n + ( Zik − Zij ) I a + ( Zim − Zil ) I b (2.50)

Again, at the j-th and k-th buses, with mutual impedance Z m between them V j is
given by

V j = Z a I a + Z m I b +Vk (2.51)

Equation (2.50) can be used in Equation (2.51) with proper notations (for j-th and k-th
buses) to yield

Z j1I1 + Z j2 I 2 +  + Z jn I n + ( Z jk − Z jj ) I a + ( Z jm − Z jl ) I b
(2.52)
= Z a I a + Z m I b + Z k 1I1 +  + Z kn I n + ( Z kk − Z kj ) I a + ( Z km − Z kl ) I b

On rearrangement, Equation (2.52) becomes

0 = ( Z k1 − Z j1 ) I1 +  + ( Z kn − Z jn ) I n + ( Z a + Z jj + Z kk − 2Z jk ) I a

+ ( Z m + Z jl + Z km − Z jm − Z kl ) I b (2.53)

Again, for buses l and m,

Vl = Z b I b + Z m I a +Vm

or, Zl1I1 +  + Zln I n + ( Zlk − Zlj ) I a + ( Zlm − Zll ) I b

(2.54)
= Z b I b + Z m I a + Z m1I1 +  + Z mn I n + ( Z mk − Z mj ) I a + ( Z mm − Z ml ) I b

On rearrangement, Equation (2.54) becomes

0 = ( Z m1 − Zl1 ) I1 +  + ( Z mn − Z ln ) I n + ( Z m + Zlj + Z mk − Zlk − Z mj ) I a

+ ( Z b + Zll + Z mn − 2Zlm ) I b (2.55)

Thus, in matrix representation, Equation (2.55) can be represented as

58 Power Transmission System Analysis Against Faults and Attacks

ˆ V1  ˆ | ˜ | ˜ 
˘  ˘ 
˘ ˜  ˘ [ Z BUS ]old | ( Zlk − Zij ) | ( Zim − Zil ) 
˘ Vn  ˘ 
˘  ˘ | ˜ | ˜ 
˘ −− =˘ 
− | − − −
˘ 0  ˘ 
˘ ˜  ˘ ° ( ki − Z ji )°
Z | ( Za + Zij + Z kk − 2Z jk ) | ( Zm + Z jl + Z km − Z kl − Z jm ) 
˘  ˘ 
ˇ˘
0  ˘ °( Z mi − Z li )° | ( Z m + Z lj + Z mk − Z lk − Z mj ) | ( Z b + Zll + Z mm − 2 Zlm ) 
ˇ 

ˆ I1 
˘ 
˘ ˜ 
˘ In 
× ˘˘ −− 

˘ Ia 
˘ −− 
˘ 
˘ˇ Ib 

˝ [ Z BUS ]old [ Z ab ] ˇ
i.e., [ Z BUS ]new = ˆ 
ˆ
˙
[ Z ba ] [ Z bb ] 
˘ (2.56)

= [ Z BUS ]old − [ Z ab ][ Z bb ] [ Z ba ]
−1

The algorithm to obtain bus impedance matrix [ Z BUS ] is given in the next article.

2.11 ALGORITHM FOR FORMATION OF BUS IMPEDANCE

MATRIX [ZBUS] USING STEP-BY-STEP METHOD
1. Read existing [ Z BUS ] matrix, that is, [ Z BUS ]old .
2. Read modifcation type i.
3. Read the added impedance value Z b .
4. If i = 1 then obtain [ Z BUS ]new using Equation (2.38b) and go to step 9, else go
to the next step.
5. If i = 2 then obtain [ Z BUS ]new using Equation (2.40b) and go to step 9, else go
to the next step.
6. If i = 3  obtain [ Z BUS ]new using Equation (2.40a) and go to step 9, else go to
the next step.
7. If i = 4  then obtain [ Z BUS ]new using Equation (2.48) and go to step 9, else go
to the next step.
8. If i = 5  obtain [ Z BUS ]new using Equation (2.56a), else go to the next step.
9. Go to step 2 for further modifcation, else go to the next step.
10. Display and/or store fnal [ Z BUS ].
11. Stop.
Analysis of Faults in Power System 59

Example 2.8

Figure 2.20 shows a four-bus power network. Assuming bus 1 to be the reference
bus, fnd [ Z BUS ] .

Solution

Bus 1 is the reference bus and in the algorithm of [ Z BUS ] building, the bus number-
ing is done excluding the reference bus. Hence, new bus numbering is required
as given below.

Old Bus no. New Bus no.

1 Reference (r)
2 1
3 2
4 3

Figure 2.21 represents the redesignated four-bus network.

FIGURE 2.20 A four-bus network (R = 5 Ω).

FIGURE 2.21 Redesignated four-bus network.

60 Power Transmission System Analysis Against Faults and Attacks

Initially, order of old [ Z BUS ] is assumed to be zero.

Step 1: Consider the element between bus 1 and reference [ Z b = 5˝] (type 1
modifcation) where added impedance is 5 Ω.
∵ Added impedance value = 5Ω,

    1
[ Z Bus ] = 1  5 ˝
[ ]
Step 2: In the next modifcation consider the element between new bus 2 and
reference bus [ Z b = 5˝]. This is also type 1 modifcation.

1 2
[ Z Bus ] = 1 ˝ 5 0 ˇ 
ˆ 
2 ˙ 0 5 ˘

Step 3: Consider the element [ Z b = 5˝] between bus 3 and reference bus. This is
again a type 1 modifcation between the new bus and reference bus.

1 2 3

1 ˙ 5 0 0 ˘
˛ [ Z Bus ] = ˇ  
2 ˇ 0 5 0 
3 ˇˆ 0 0 5 

Step 4: Consider the element [ Z b = 5˝] between buses 1 and 2 in the next modif-
cation. This is type 4 modifcation between the two old buses 1 and 2.

˘ (Z − Z ) 
1  11 12

[ Z Bus ] = [ Z BUS ]old −  ( Z 21 − Z 22 )  × ( Z11 − Z 21 )( Z12 − Z 22 )( Z13 − Z 23 )
5 + Z11 + Z 22 − 2Z12  
 ( Z 31 − Z 32 ) 

ˆ 5 
1 ˘ 
or, [ Z Bus ] = [ Z BUS ]old −   ˘ −5  ˆˇ 5 −5 0 
5+5+5
˘ˇ 0 

˝ 5 0 0 ˇ ˝ 5 ˇ ˝ 25 −25 0 ˇ
ˆ  1 ˆ ˆ 
or, [ Z Bus ] = ˆ 0 5 0 −
 15 ˆ −5  ˆ −25
  25 0 
ˆ˙ 0 0 5 ˘ ˆ˙ 0 ˘ ˆ˙ 0 0 0 ˘

Simplifcation yields

1 2 3

or, [ Z Bus ] = 1 ˝ 3.33 1.67 0 ˇ 

ˆ 
2 ˆ 1.67 3.33 0 
3 ˆ˙ 0 0 5 ˘
Analysis of Faults in Power System 61

Step 5: In the next modifcation, consider the element [ Z b = 5˝] between buses 2
and 3. This is also type 4 modifcation.
Added impedance value is 5 Ω between two old buses 2 and 3.

 (Z − Z ) 
1  12 13

[ Z Bus ] = [ Z BUS ]old −   ×  ( Z 22 − Z 23 )  ( Z 21 − Z 31 )( Z 22 − Z 32 )( Z 23 − Z 33 )
5 + Z 22 + Z 33 − 2Z 23  
 ( Z 32 − Z 33 ) 

ˆ 1.67 
1 ˘ 
= [ Z BUS ]old −   ˘ 3.33  ˆˇ 1.67 3.33 −5 
5 + 3.33 + 5
˘ˇ −5 

˙ 1.67 ˘
1 ˇ 
= [ Z BUS ]old −   3.33  ˙ˆ 1.67 3.33 −5 ˘
13.33 ˇ
ˇˆ −5 

° 3.33 1.67 0 ˙ ° 0.21 0.42 −0.62 ˙

˝ ˇ ˝ ˇ
= ˝ 1.67 3.33 0 ˇ − ˝ 0.42 0.83 −1.25 ˇ
˛˝ 0 0 5 ˇˆ ˝˛ −0.62 −1.25 1.87 ˇˆ

    1              2            3 

1 ° 3.12 1.25 0.62 ˙
=  ˝ ˇ ˘
2 ˝ 1.25 2.5 1.25 ˇ
3 ˝˛ 0.62 1.25 3.12 ˇˆ

Because there is no further modifcation and no bus interchange, the above matrix
is the fnal [ Z Bus ]   matrix.
Therefore,

    1              2            3 
1 ˝ 3.12 1.25 0.62 ˇ
[ Z Bus ] = 2
ˆ   = [ Z Bus ]new
ˆ 1.25 2.5 1.25 
3 ˆ˙ 0.62 1.25 3.12 ˘

2.12 DETERMINATION OF SYMMETRICAL FAULT

CURRENT USING [ZBUS] INVERSION
A simple way to fnd symmetrical fault current at any bus of multi-bus power system
is to use [ Z Bus ] obtained directly by inverting [YBus ] as
62 Power Transmission System Analysis Against Faults and Attacks

−1
˝ Y11 Y12 ˜ Y1i ˜ Y1n ˇ
ˆ 
ˆ Y21 Y22 ˜ Y2i ˜ Y2n 
−1 ˆ ° ° ° ° 
[ Z Bus ] = [YBus ] = ˆ 
ˆ Yi1 Yi2 ˜ Yii ˜ Yin 
ˆ ° ° ° ° 
ˆ Yn1 Yn2 ˜ Yni ˜ Ynn 
˙ ˘
(2.57)
˝ Z11 Z12 ˜ Z1i ˜ Z1n ˇ
ˆ 
ˆ Z 21 Z 22 ˜ Z 2i ˜ Z 2n 
ˆ ° ° ° ° 
=ˆ 
ˆ Zi1 Zi2 ˜ Zii ˜ Zin 
ˆ ° ° ° ° 
ˆ Z n1 Z n2 ˜ Z ni ˜ Z nn 
˙ ˘

This symmetrical fault current at bus 1 (say) is given by

I f1 = 1 Z , assuming the pre-fault bus voltage of bus 1 as 1.00 p.u. Similarly, the
11
symmetrical fault current of any bus i is given by I f1 = 1 Z , assuming the pre-fault i
ii
th bus voltage to be 1.00. However, for large multi-bus power network, this approach
of fnding the symmetrical bus fault current is unacceptable because of the need of
inverting a very large and sparse [YBus ] matrix. Furthermore, for any change in the
network, the full [YBus ] is to be rebuilt, and again it needs to be inverted using eff-
cient [YBus ] inversion technique. On the other hand, fnding fault current from [ Z Bus ]
where the direct building of [ Z Bus ] algorithm is used appears to be more reasonable
and saves computer time and memory.

2.13 PHASE-SEQUENCE COMPONENT NETWORK

It has already been established that using the symmetrical component theory, it is
possible to replace a three-phase set of unbalanced voltages (or currents) defned
as positive-phase sequence (PPS), negative-phase sequence (NPS) and zero-phase
sequence (ZPS) sets. Because of an external unbalanced condition, say a single-phase
short circuit, PPS, NPS and ZPS voltages and currents appear on the network at a
fault point. A phase-sequence component is one that carries current and voltages
of one particular phase sequence such as PPS, NPS or ZPS sequence. It may be
noted that because the actual three-phase network is assumed to be balanced, PPS,
NPS and ZPS networks are separate and there is no intersequence mutual coupling
between them. These sequence networks are only connected at the point of unbal-
ance in the system. Moreover, the assumption of a perfectly balanced three-phase
network means that only positive sequence voltage and currents exist in the PPS
network. Although negative and ZPS network can still be artifcially constructed,
they are totally redundant because they carry no negative or ZPS voltages or cur-
rents. The negative- and zero-sequence voltages in the corresponding negative and
ZPS networks appear as a result of the unbalanced condition imposed on the actual
Analysis of Faults in Power System 63

three-phase network; they should not be confused with the voltage sources that exist
in the PPS network.
Let the unbalanced condition be considered for a three-phase network at point
F, as shown in Figure 2.22a. The three-phase networks can be then constructed
from the actual three-phase network components and network theory and are shown
in Figures 2.22b and c. The entire PPS, NPS and ZPS networks are derived using
Thevenin’s theorem. This reduction results in a single equivalent voltage source at
point F and a single equivalent impedance seen looking back into the relevant net-
work from point F, as illustrated in Figure 2.22c.
From Figure 2.22c, sequence voltage and current relations at the point of fault (F)
can be written for the networks. For the active positive-sequence network,

Vr1 = VF − Z1I r1 (2.58)

where VF is the positive-sequence phase voltage at the point of fault immediately

before the event of unbalance appearing at F. Vr1 is the resultant positive sequence
voltage at the point of fault, and I r1 is the fow of positive sequence current in the
positive sequence network.

(a)
Point of Fault

Balanced 3-phase power

network

(b)

Active single phase Passive single phase Passive single phase

network network network

Positive Sequence Network Negative Sequence Network Zero Sequence Network

(c)

Thevenin's positive Thevenin's negative Thevenin's zero

sequence network sequence network sequence network

FIGURE 2.22 Equivalent Thevenin’s positive-, negative- and zero-sequence networks.

64 Power Transmission System Analysis Against Faults and Attacks

For the passive negative and zero-sequence networks,

Vr2 = −Z 2 I r2 (2.59)

Vr0 = −Z 0 I r0 (2.60)

where Vr2 and Vr0 are the negative-sequence and zero-sequence voltages at F for the
negative- and zero-sequence networks, respectively. I r2 and I r0 are the corresponding
negative- and zero-sequence currents fowing out of the NPS and zero-sequence net-
works into the fault, respectively. Z1, Z 2 and Z 0 represent the positive-, negative- and
zero-sequence impedances of the sequence networks.
In the multi-bus power system, the governing sequence network equation can be
represented, generalising Equations (2.58)–(2.60) as follows:
For the positive-sequence generalised network,
n

Vi1 = E − ˜Z I
ik1 k1 (2.61a)
k=1

where Vi1 is the generalised-positive sequence voltage at ith bus, the bus subjected to
be under fault condition, and E is the pre-fault bus phase voltage. I k1 is the positive-
sequence current at any bus k, while Zik1 is the equivalent positive-sequence imped-
ance looking into the positive-sequence network of bus k. For the convenience of the
modelling, it is the logical assumption that all the currents at bus k ˜ i are zero, while
I f is the fault current at bus i at k = i . Hence, I k1 , I k2 and I k0 are all zero quantities for
k ˜ i . At bus, the positive-, negative- and zero-sequence currents are represented as
I i1 , I i2 and I i0 (k = i).
Then, for the faulted power network, the governing equation of the sequence net-
works are given by

Vi1 = E − [ Z1 ] I i1 (2.61b)
n

Vi2 = − ˜Z I
ik2 k2 (2.62)
k=1

Vi0 = − ˜Z I
ik0 k0
(2.63)
k=1

= − [ Z 0 ] I i0

The following common types of electrical faults are analysed in the following articles:

a. Three-phase balanced fault

b. Single line-to-ground fault
c. Line-to-line fault
d. Line-to-line-to-ground fault
Analysis of Faults in Power System 65

For any of the above type of fault at the i-th bus, I k = 0  ( k = 1, 2,3,…, n; k ˙ i ).

2.13.1 THREE-PHASE BALANCED FAULT

A balanced three-phase system remains symmetrical after the occurrence of a three-
phase fault having the same impedance between each line and a common point.
Only positive-sequence current fows through such a network. The positive-sequence
voltages at the i-th bus, where the three-phase balanced (symmetrical) fault appears
is rewritten from Equation (2.61).
n

Vi(1) = E − ˜Z I
ik(1) k(1)
k=1

= E − ˆ˙ Zi1(1) I1(1) +  + Zii(1) I i(1) +  + Zin(1) I n(1) ˘ˇ

[1 at suffx represents the positive sequence quantity]

Because currents at all buses except the   i-th bus are zero,

I i(1) ˜ 0; I1(1) = I 2(1) = = I k(1) =  = I n(1) = 0

˜Vi(1) = E − ˙ˆ 0 + + Zii(1) I i(1) +  + 0 ˇ˘

E
i.e., I i(1) = (2.64)
Zii(1)

Assuming the series fault impedance to be Z f ,

E
I i(1) = (2.65)
Z f + Zii(1)

The bus voltage at i-th bus would be zero due to direct fault at i-th bus. The bus
voltage at other bus are given by,

Vk(1) = E − Z ki(1) I i(1)

E
= E − Z ki(1)
Z f + Zii(1)

˝ Z ki(1) ˇ
= E ˆ1− 
ˆ˙ Z f + Zii(1) ˘

˙ Z f + Zii(1) − Z ki(1) ˘
˜Vk(1) = E ˇ  (2.66)
ˇˆ Z f + Zii(1) 
66 Power Transmission System Analysis Against Faults and Attacks

2.13.2 SINGLE LINE TO GROUND FAULT

Let single phase to ground fault (phase “a” is shorted with ground through Z f , the
fault impedance) be occurring at bus i (Figure 2.23).

Here, I ib = I ic = 0 (2.67)
Via = I ia Z f (2.68)

Because I ib = I ic , following  Equation ( 2.25a ) ,

I i( 0 ) + a 2 I i(1) + aI i( 2) = I i( 0 ) + aI i(1) + a 2 I i( 2)
or, (a 2
) (
− a I i(1) = a 2 − a I i( 2) ) (2.69)

ˆ I i(1) = I i( 2)

Because I ib = 0 ,

I i( 0 ) + a 2 I i(1) + aI i( 2) = 0
(
i.e. , I i( 0 ) = − a 2 + a I i(1) ) ˆˇ I i(1) = I i( 2) ˘ (2.70)

or, I i( 0 ) = I i(1) ˇ (
ˆ 1 + a + a 2 = 0 ˘
 )
With reference to Figure 2.23,

Via = I ia Z f
(2.71)
Vi( 0 ) + Vi(1) + Vi( 2) = Z f ˛˝ I i( 0 ) + I i(1) + I i( 2) ˙ˆ
i.e., Vi( 0 ) + Vi(1) + Vi( 2) = 3I i(1) Z f

˝ I i( 0 ) = I i(1) = I i( 2) , just shown above in Equtions ( 2.69 and 2.70 ) ˆ

˙ ˇ

point of fault

i-th bus

ground

FIGURE 2.23 Line-to-ground fault at i -th bus.

Analysis of Faults in Power System 67

Zero sequence Positive sequence Negative sequence

network network network

FIGURE 2.24 Interconnection of sequence networks in single line to ground fault.

The equivalent circuit of the Equation (2.71) can then be the network having the
series-connected sequence networks across the fault impedance Z f (Figure 2.24).
In the generalised n – bus network, for a single line-to-ground fault at the i th bus,
Equation (2.71) can be rewritten as

Vi (0) +Vi (1) +Vi (2) = 3Z f I i (1) (2.72)

Also, I k( 0 ) = I k(1) = I k( 2) = 0 ˝˙ for k = 1, 2,… , n; k ˛ iˆˇ

From Equation (2.72) in terms of sequence network quantities we have

− Zii (0) I i (0) + E − Zii (1) I i (1) − Zii (2) I i (2) = 3 Z f I i (1)

or, E − Zii (0) I i (0) − Zii (1) I i (1) − Zii (2) I i (2) = 3 Z f I i (1) (2.73)

Since I i (0) = I i (1) = I i (2) , Equation  ( 2.73) reduces to

E = I i(1) ˛Z
˝ ii( 0 ) + Zii(1) + Zii( 2) + 3Z f ˙ˆ
E
˜ I i (1) = (2.74)
Zii (0) + Zii (1) + Zii (2) + 3Z f

The kt h bus voltage sequence components are given by

Vk (0) = −Z ki (0) I i (0) = −Z ki (0) I i (1)

−Z ki (0) ˝  E
˜Vk (0) = (2.75)
Zii (0) + Zii (1) + Zii (2) + 3Z f

˝ −Z ki(1) ˇ
Also, Vk (1) = E ˆ1 − 
˙ Zii( 0 ) + Zii(1) + Zii( 2) + 3Z f ˘
68 Power Transmission System Analysis Against Faults and Attacks

˛V
˝ k (1) = E − Z ki(1)I i(1) ˙ˆ

Zii (0) + Zii (1) + Zii (2) + 3Z f − Z ki (1)

or, Vk (1) = ˝E (2.76)
Zii (0) + Zii (1) + Zii (2) + 3Z f

and Vk (2) = −Z ki (2) I i (2) = −Z ki (2) I i (1)

−Z ki (2) ˝ E
˜Vk (2) = (2.77)
Zii (0) + Zii (1) + Zii (2) + 3Z f

Z ki (1) , Z ki (0) , Z ki (2) , Zii (1) , Zii (0) ,and, Zii (2) are the respective entries of the [ Z Bus ] matrix
of the n-bus power network in sequence terms. Thus, it is possible to determine the
sequence components of fault currents and the sequence components of voltages at
any bus of the n-bus power network subjected to a single line-to-ground fault at bus i.

2.13.3 LINE-TO-LINE FAULT

Let there be a line to line fault in a three-phase balanced system (Figure 2.25).
Obviously,   I ib = −I ic

i.e., I i (0) + a 2 I i (1) + aI i (2) = − ( I i (0) + aI i (1) + a 2 I i (2) )

or 2I i (0) + ( a 2 + a ) ( I i (1) + I i (2) ) = 0 (2.78)

Also, Vib = Vic + Z f I ib (2.79)

and I ia = 0

i.e. I i (0) + I i (1) + I i (2) = 0 (2.80)

3 phase
i-th bus

ground

FIGURE 2.25 Double-line fault at i -th bus.

Analysis of Faults in Power System 69

Using Equation (2.80) in (2.78), we have

( ) (
2I a( 0 ) + ( −1) −I a( 0 ) = 0 ˆˇ a 2 + a = −1˘ )
i.e., 3I a(0) = 0

or, I a(0) = 0 (2.81)

Using Equation (2.81) in (2.80),

I i (1) = −I i (2) (2.82)

Because Vib = Vic + Z f I ib, we can write,

Vi (0) +Vi (1) +Vi (2) = (Vi (0) + aVi (1) + a 2Vi (2) ) + Z f ( I i (0) + a 2 I i (1) + aI i (2) )

Simplifcation yields,

(a 2
− a ) Vi (1) = ( a 2 − a ) I i (1) Z f + ( a 2 − a ) Vi (2)

i.e., Vi (1) = I i (1) Z f +Vi (2) (2.83)

The equivalent sequence network circuit is shown in Figure 2.26.

Next, the fault (L-L) is assumed to occur at the i-th bus of the generalised n-bus
network. Here Vi (0) = 0 as zero-sequence network is a passive network.
Using Equation (2.83) for the multi-bus network, we have

Ei − Zii (1) I i (1) = I i (1) Z f − Zii (2) I i (2) (2.84)

Since Vi (1) = Ei − Zii (1) I i (1)

and Vi (2) = −Zii (2) I i (2)

Zero sequence Positive sequence Negative sequence

network network network

FIGURE 2.26 Interconnection of sequence network for double line (L-L) fault.
70 Power Transmission System Analysis Against Faults and Attacks

Using the relation I i (1) = −I i (2) from Equation (2.84), we obtain,

Ei
I i (1) = (2.85)
Zii (1) + Zii (2) + Z f

Also,

Vk (1) = Ek − Z ki (1) I i (1)

Ei
= Ek − Z ki (1)
Zii (1) + Zii (2) + Z f

ˆ Z ki (1) 
= E ˘1−
˘ˇ Zii (1) + Zii (2) + Z f


(assuming Ei = Ek = E )

˙ Zii(1) + Zii( 2) + Z f − Z ki(1) ˘

˜Vk(1) = E ˇ  (2.86)
ˆ Zii(1) + Zii( 2) + Z f 

( )
and, Vk( 2) = −Z ki( 2) I i( 2) = −Z ki( 2) +I i( 2) = −Z ki( 2) −I i(1) ( ) ˆˇ I i(1) = −I i( 2) ˘

Ek
or, Vk (2) = Z ki (2) I i (1) = Z ki (2)
Zii (1) + Zii (2) + Z f

Z ki (2) ˝ E
i.e., Vk (2) = ; [ Ek = E ] (2.87)
Zii (1) + Zii (2) + Z f

Equation (2.85) represents the i-th bus positive sequence current while the k-th bus
sequence voltages are represented in Equations (2.86) and (2.87) in the multi-bus
power network subjected to a line-to-line fault at i-th bus.

2.13.4 DOUBLE LINE-TO-GROUND (L-L-G) FAULT

Let a double line-to-ground fault be occurring at bus i as represented in Figure 2.27.

Obviously, I ia = 0 (2.88)

Vib = Vic (2.89)

and Vib = ( I ib + I ic ) Z f (2.90)

As I ia = 0 (as per Equation (2.88)),

I i (0) + I i (1) + I i (2) = 0 (2.91)

Analysis of Faults in Power System 71

ground

FIGURE 2.27 Double line to ground fault at i -th bus.

From Equation (2.89),

Vi (0) + a 2Vi (1) + aVi (2) = Vi (0) + aVi (1) + a 2Vi (2)

Simplifcation yields,

Vi (1) = Vi (2) (2.92)

Also from Equation (2.90)

Vi (0) + a 2Vi (1) + aVi (2) = ( I i (0) + a 2 I i (1) + aI i (2) + I i (0) + aI i (1) + a 2 I i (2) ) Z f

Simplifcation yields

( )(
Vi( 0 ) − Vi(1) = ˆˇ 2I i( 0 ) + a 2 + a I i(1) + I i( 2) ˘ Z f )
or, Vi (0) −Vi (1) = 3Z f I i (0)

Vi (0) = Vi (1) + 3Z f I i (0) (2.93)

The sequence network interconnection is shown in Figure 2.28.

For the multi-bus network, in case of double line-to-ground fault, the positive
sequence network gives

Vi (1) = E − Zii (1) I i (1)

−Vi (1) + E
i.e., I i (1) = (2.94)
Zii (1)

Vi (2) = −Zii (2) I i (2)

72 Power Transmission System Analysis Against Faults and Attacks

Zero sequence Positive sequence Negative sequence

network network network

FIGURE 2.28 Sequence network interconnection for double line to ground fault.

−Vi( 2) −Vi(1)
i.e., I i( 2) = = ˙using
ˆ Equation  ( 2.92 )˘ˇ (2.95)
Zii( 2) Zii( 2)

From Equation (2.93), we can write

Vi(1) = −Zii( 0 ) I i( 0 ) − 3Z f I i( 0 ) ˛V

˝ i( 0 ) = −Zii I i( 0 ) ˙ˆ

Vi (1)
i.e., I i (0) = − (2.96)
Zii (0) + 3Z f

Substituting the values of I i (1) ,  I i (2) and I i (0) from Equations (2.94)–(2.96), respectively,
in Equation (2.91), we get

Vi (1) E − Vi (1) Vi (1)

− + − =0
Zii (0) + 3Z f Zii (1) Zii (2)

Vi (1) Vi 1 − E Vi (1)
or, + () + =0
Zii (0) + 3Z f Zii (1) Zii (2)

Zii (2) ( Zii (0) + 3Z f ) E

˛Vi (1) =
Zii (1) Zii (2) + Zii (1) ( Zii (0) + 3Z f ) + Zii (2) ( Zii (0) + 3Z f )

Zii( 2) ˛Z
˝ ii( 0 ) + 3Z f ˙ˆ
i.e., Vi(1) = (2.97)
ˇZ  

where, ˛Z = Zii (1) Zii (2) + Zii (1) ( Zii (0) + 3Z f ) + Zii (2) ( Zii (0) + 3Z f )
Analysis of Faults in Power System 73

Substituting the value of  Vi (1) from Equation (2.97) in Equations (2.94), (2.95) and
(2.96), the expressions of I i (0) , I i (1) and I i (2) are modifed as follows:

Vi (1) Zii 2 ˝ E
I i (0) = − =− ( ) (2.98)
Zii (0) + 3Z f ˙Z

(
ˇ Zii( 2) Zii( 0 ) + 3Z f E 
E− 
)
I i(1) =
E − Vi(1)
=
˘ ˆZ  
=
(
 E Zii( 2) + Zii( 0 ) + 3Z f ) (2.99)
Zii(1) Zii(1) ˆZ  

−Vi(1) − ˙˝ Zii( 0 ) + 3Z f ˆˇ
and, I i( 2) = = (2.100)
Zii( 2) ˘Z  

The sequence voltages at bus k are obtained as

Z ki (0)   Zii (2) E

Vk (0) = −Z ki (0) I i (0) = (2.101)
˛Z
Here, E = Ei = Ek

∆Z − Z ki (1)   ( Zii (2) + 3 Z f + Zii (0) ) E

Vk (1) = E − Z ki (1) I i (1) = (2.102)
∆Z
and Vk (2) = − Z ki (2) I i (2)

Z ki (2)   (3 Z f + Zii (0) ) E

or, Vk (2) = (2.103)
∆Z

2.13.5 SEQUENCE CURRENTS IN THE INTERCONNECTING LINE

BETWEEN THE FAULTED BUS I AND HEALTHY BUS K

The sequence currents in the interconnecting line between faulted bus i and healthy
bus k are given as

Vi( 0 ) − Vk( 0 ) Vi(1) − Vk(1) Vi( 2) − Vk( 2)

I ik( 0 ) = ; I ik(1) = ; and I ik( 2) = (2.104)
Zik( 0 ) Zik(1) Zik( 2)

Also, I ik (0) = −I ik (0)

I ik (1) = −I ik (1)

and I ik (2) = −I ik (2)

−I ik (0) ,   −I ik (1) and −I ik (2) being the sequence currents fowing from the k-th bus to
the i-th bus and are usually negative of I ik (0) , I ik (1) and, I ik (2), respectively, except the
condition when there is a star-delta transformer in the line.
74 Power Transmission System Analysis Against Faults and Attacks

Example 2.9

For a three-bus power system, the sequence impedances are given as follows:

˝ j0.18 j0.12 j0.15 ˇ

ˆ 
[ Z Bus ](1) = [ Z Bus ](2) = ˆ j0.12 j0.18 j0.15  p.u.
ˆ j0.15 j0.15 j0.25 
˙ ˘
˝ j0.09 j0.06 j0.08 ˇ
ˆ 
[ Z Bus ](0) = ˆ j0.06 j0.09 j0.08  p.u.
ˆ j0.08 j0.08 j0.14 
˙ ˘
Assuming the pre-fault bus voltage to be 1.00 p.u./phase, compute the fault cur-
rent, bus voltages and line currents in the post-fault condition for (i) single line-to-
ground solid fault, (ii) line-to-line short circuit fault, (iii) solid three-phase-to-ground
fault and (iv) solid double line-to-ground fault, assuming the faults occurring sepa-
rately at bus number 3.

Zero-Sequence Positive-Sequence Negative-Sequence

Line no. From Bus To Bus Line Impedance Line Impedance Line Impedance
1 1 2 j0.15 j0.3 j0.3
2 2 3 j0.15 j0.3 j0.3
3 1 3 j0.15 j0.3 j0.3

Solution

I) FOR SINGLE LINE-TO-GROUND FAULT

Assuming the fault occurring in bus number 3,

E ˛˝ Z f = 0, for solid fault˙ˆ

I 3(0) = I 3(1) = I 3(2) =
Z 33(0) + Z33(1) + Z 33(2)

Here, Z 33(1) = Z 33(2) , as given in the statement of this problem. As E = 1 p.u. we have,

1 1
I 3(0) = I 3(1) = I 3(2) = = = − j1.56 p.u.
Z 33(0) + 2Z 33(1) j0.14 + 2 × j0.25

˜ I 3(fault) = 3I a1 = − j 3 ×1.56 = − j4.68 p.u.

Next, we determine the bus voltages after the fault occurs.

V1(0) = −Z13(0) I 3(0) = − j0.08 (− j1.56) = −0.125 p.u.

V2(0) = −Z 23(0) I 3(0) = − j0.08 (− j1.56) = −0.125 p.u.

V3(0) = −Z 33(0) I 3(0) = − j0.14 (− j1.56) = −0.218 p.u. ˙ −0.22 p.u.

Analysis of Faults in Power System 75

V1(1) = E − Z13(1) I 3(1) = 1 − j0.15 (− j1.56) = 0.766 p.u.

V1(2) = −Z13(2) I 3(2) = − j0.15 (− j1.56) = −0.234 p.u.

V2(1) = E − Z 23(1) I 3(1) = 1 − j0.15 (− j1.56) = 0.766 p.u.

V3(1) = E − Z 33(1) I 3(1) = 1 − j0.25 (− j1.56) = 0.61 p.u.

V1(2) = −Z13(2) I 3(2) = − j0.15 (− j1.56) = −0.234 p.u.

V2(2) = −Z 23(2) I 3(2) = − j0.15 (− j1.56) = −0.234 p.u.

V3( 2) = −Z 33( 2) I 3( 2) = − j0.25 ( − j1.56 ) = −0.39 p.u.

Next, we compute line currents (here, suffx L stands for the p.u. impedances of
respective lines)

V1(0) − V2(0)
I12(0) = =0 ˛˝V1(0) = V2(0) ˙ˆ
Z12(0) L

V2(0) − V3(0) −0.125 − (−0.218)

I 23(0) = = = − j0.62 p.u.
Z 23(0) L j0.15

V3(0) − V1(0) −0.218 − (−0.125)

I 31(0) = = = j0.62 p.u.
Z 31(0) L j0.15

V1(1) − V2(1)
I12(1) = =0 ˛˝V1(1) = V2(1) ˙ˆ
Z12(1) L

V2(1) − V3(1) 0.766 − 0.61

I 23(1) = = = − j0.52 p.u.
Z 23(1) L j0.3

V3(1) − V1(1) 0.61− 0.766

I 31(1) = =   = j0.52 p.u.
Z13(1) L j0.3

V1(2) − V2(2)
I12(2) = =0 ˛˝V1(2) = V2(2) ˙ˆ
Z12(2) L

V2(2) − V3(2) 0.234 − (−0.39)

I 23(2) = =   = − j0.52 p.u.
Z 23(2) L j0.3

V3(2) − V1(2) −0.39 − (−0.234 )

I 31(2) = =   = j0.52 p.u.
Z 31(2) L j0.3

˜ Finally, for L-G fault at bus 3, we get the following sequence quantities.
76 Power Transmission System Analysis Against Faults and Attacks

V1(0) = −0.125 p.u.; V1(1) = 0.766 p.u. ; V1(2) = −0.234 p.u.

V2(0) = −0.125 p.u.; V2(1) = 0.766 p.u.; V2(2) = −0.234 p.u.

V3(0) = −0.218 p.u.; V3(1) = 0.61 p.u.; V3(2) = −0.39 p.u.

I 3( 0 ) = I 3(1) = I 3( 2) = − j1.56 p.u.; ˛˝ I 3( fault ) = − j4.68 p.u. ˙ˆ

I12(0) = 0 p.u.; I12(1) = 0 p.u. ;  I12(2) = 0 p.u.

I 23(0) = − j0.62 p.u.; I 23(1) = − j0.52 p.u.; I 23(2) = − j0.52 p.u.

I 31(0) = − j0.62 p.u.; I 31(1) = j0.52 p.u.; I 31(2) = j0.52 p.u.

II) FOR LINE-TO-LINE FAULT

For the line-to-line fault at bus 3,

E
I 3(1) =
Z 33(1) + Z 33(2) + Z f

Here, Z 33(1) = Z 33(2) = j0.25 p.u. ; Z f = 0 and E = 1 p.u.

1 1
˜ I 3(1) = = = − j2 p.u.
j0.25 + j0.25 j0.5
I 3(2) = I 3(1) = j2 p.u.

The zero-sequence component is absent in line-to-line fault.

˜ I 3( 0 ) = 0

Post-fault bus voltages are given as:

V1(1) = E − Z13(1) I 3(1) = 1 − j0.15 (− j 2) = 0.7 p.u.

V1(2) = −Z13(2) I 3(2) = − j0.15 × j2 = 0.3 p.u.

V1(0) = 0 p.u.

V2(1) = E − Z 23(1) I 3(1) = 1 − j0.15 (− j 2) = 0.7 p.u.

V2(2) = −Z 23(2) I 3(2) = − j0.15 × j2 = 0.3 p.u.

V2(0) = 0 p.u.

V3(1) = E − Z 33(1) I 3(1) = 1 − ( j0.25)(− j2) = 0.5 p.u.

V3(2) = −Z 33(2) I 3(2) = − j0.25 × j2 = 0.5 p.u.

V3(0) = 0
Analysis of Faults in Power System 77

Line currents are obtained as (suffx L stands for respective line impedances)

V2(1) − V3(1) 0.7 − 0.5

I 23(1) = = = − j0.67 p.u.
Z 23(1) L j0.3

V2(2) − V3(2) 0.3 − 0.5

I 23(2) = =   = − j0.67 p.u.
Z 23(2) L j0.3

I 23(0) = 0 p.u.

V1(1) − V3(1) 0.7 − 0.5

I13(1) = = = − j0.67 p.u. = −I 31(1)
Z13(1) L j0.3

V1(2) − V3(2) 0.3 − 0.5

I13(2) = =   = j0.67 p.u. = −I 31(2)
Z13(2) L j0.3

I13(0) = 0 p.u. = −I 31(0)

V1(1) − V2(1) 0.7 − 0.7

I12(1) = = = 0 p.u.
Z12(1) L j0.3

V1(2) − V2(2) 0.3 − 0.3

I12(2) = = = 0 p.u.
Z12(2) L j0.3

I12(0) = 0

˜ For line-to-line fault we have

V1(0) = 0 p.u.; V1(1) = 0.7 p.u.;  V1(2) = 0.3 p.u.

V2(0) = 0 p.u.; V2(1) = 0.7 p.u.;V2(2) = 0.3 p.u.

V3(0) = 0 p.u.; V3(1) = 0.5 p.u.;  V3(2) = 0.5 p.u.

I 3(1) = − j2 p.u.; I 3(2) = j2 p.u.; I 3(0) = 0

I12(0) = 0 p.u.; I12(1) = 0 p.u.; I12(2) = 0 p.u.

I 23(0) = 0 p.u.; I 23(1) = − j0.67 p.u.; I 23(2) = j0.67 p.u.

I 31(0) = 0; I 31(1) = j0.67 p.u.; I 31(2) = − j0.67 p.u.

III) FOR SOLID THREE-PHASE-TO-GROUND FAULT

The fault current is given by

E °˛ Z f = 0˝˙
I 3(1) =
Z 33(1)
78 Power Transmission System Analysis Against Faults and Attacks

1
or, I 3(1) = = − j4 p.u.
j0.25

Post-fault bus voltages are given by

V1(1) = E − Z13(1) I 3(1) = 1 − j0.15 (− j 4 ) = 0.4 p.u.

V2(1) = E − Z 23(1) I 3(1) = 1 − j0.15 (− j 4 ) = 0.4 p.u.

V3(1) = E − Z 33(1) I 3(1) = 1 − j0.25 (− j 4 ) = 0 p.u.

Line currents during the fault period are given by

V1(1) − V2(1)
I12(1) = =0 ˛˝V1(1) = V2(1) ˙ˆ
Z12(1) L

V2(1) − V3(1) 0.4 − 0

I 23(1) = = = − j1.33 p.u.
Z 23(1) L j0.3

V1(1) − V3(1) 0.4 − 0

I13(1) = = = − j1.33 p.u. = −I 31(1)
Z13(1) L j0.3

Finally, we have for a three-phase-to-ground fault, I 3(1) = − j4 p.u. [only positive

sequence current fows in balanced three-phase faulted network]

V1(0) = 0 p.u.; V1(1) = 0.4 p.u.;  V1(2) = 0 p.u.

V2(0) = 0 p.u.; V2(1) = 0.4 p.u.; V2(2) = 0 p.u.

V3(0) = 0 p.u.; V3(1) = 0 p.u.;  V3(2) = 0 p.u.

Also, I12(1) = 0 p.u.; I 23(1) = − j1.33 p.u. ; I 31(1) = j1.33 p.u.

IV) FOR DOUBLE-LINE-TO-GROUND FAULT

The fault current is given by

E ( Z 33(2) − Z 33(0) )
I 31(1) = , using equation ( 2.99) with i = 3 and Z f = 0
˙Z
1( j0.25 + j0.14 )
=
j0.25 × j0.25 + j0.25 × j0.14 + j0.25 × j0.14

[ ˜Z = Z 33(1) Z 33(2) + Z 33(1) Z 33(0) + Z 33(2) Z 33(0) for Z f = 0 and i = 3 from Equation
(2.97)]

j0.39
˜ I 31(1) = = − j2.94 p.u.
−0.1325
Analysis of Faults in Power System 79

Also, I 3( 2 ) = −
( Z ( ) ) , with i = 3 and
33 0
Z f = 0 in Equation  ( 2.100 )
˙Z
j0.14
i.e., I 3(2) = − = j1.06 p.u.
−0.1325

and, I 3(0) = −
( Z ( ) ) E , with i = 3 and
33 2
Z f = 0 in Equation ( 2.98)
˙Z
j0.25 ×1
=− = j1.89 p.u.
−0.1325

Bus voltages after the occurrence of the fault can be calculated as follows.

V1(1) = E − Z13(1) I 3(1) = 1 − j0.15 (− j 2.94 ) = 0.559 p.u.

V1(2) = −Z13(2) I 3(2) = −0.15 ( j1.06) = 0.159 p.u.

V1(0) = −Z13(0) I 3(0) = − j0.08 ( j1.89) = 0.1512 p.u.

V2(1) = E − Z 23(1) I 3(1) = 1 − j0.15 (− j 2.94 ) = 0.559 p.u.

V2(2) = −Z 23(2) I 3(2) = − j0.15 ( j1.06) = 0.159 p.u.

V2(0) = −Z 23(0) I 3(0) = − j0.08 ( j1.89) = 0.1512 p.u.

V3(1) = E − Z 33(1) I 3(1) = 1 − j0.25 (− j 2.94 ) = 0.265 p.u.

V3(2) = −Z 33(2) I 3(2) = − j0.25 ( j1.06) = 0.265 p.u.

V3(0) = −Z 33(0) I 3(0) = − j0.14 ( j1.89) = 0.2646 p.u.

Line currents can be calculated as follows:

V1(1) − V2(1) 0.559 − 0.559

I12(1) = = = 0 p.u.
Z12(1) L j0.3

V1(2) − V2(2) 0.159 − 0.159

I12(2) = = =0
Z12(2) L j0.3

V1(0) − V2(0) 0.1512 − 0.1512

I12(0) = = =0
Z12(0) L j0.15

V1(1) − V3(1) 0.559 − 0.265

I13(1) = = = − j0.98 p.u. = −I 31(1)
Z13(1) L j0.3

V1(2) − V3(2) 0.159 − 0.265

I13(2) = = = j0.353 p.u. = −I 31(2)
Z13(2) L j0.3
80 Power Transmission System Analysis Against Faults and Attacks

V1(0) − V3(0) 0.1512 − 0.2646

I13(0) = = = j0.756 p.u. = −I 31(0)
Z13(0) L j0.15

V2(1) − V3(1) 0.559 − 0.265

I 23(1) = = = − j0.98 p.u.
Z 23(1) L j0.3

V2(2) − V3(2) 0.159 − 0.265

I 23(2) = = = j0.353 p.u.
Z 23(2) L j0.3

V2(0) − V3(0) 0.1512 − 0.2646

I 23(0) = = = j0.756 p.u.
Z 23(0) L j0.15

˜ For L-L-G fault at bus 3 we get the following sequence quantities,

V1(0) = 0.1512 p.u.; V1(1) = 0.559 p.u. ; V1(2) = 0.159 p.u.

V2(0) = 0.1512 p.u.; V2(1) = 0.559 p.u.; V2(2) = 0.159 p.u.

V3(0) = 0.2646 p.u.; V3(1) = 0.265 p.u.;  V3(2) = 0.265 p.u.

I 3(1) = − j2.94 p.u.; I 3( 2) = j1.06 p.u.; I 3( 0 ) = j1.89 p.u.

I12(0) = 0 p.u.; I12(1) = 0 p.u.; I12(2) = 0 p.u.

I 23(0) = j0.756 p.u.; I 23(1) = − j0.98 p.u.; I 23(2) = j0.353 p.u.

I 31(0) = − j0.756; I 31(1) = j0.98 p.u.; I 31(2) = − j0.353 p.u.

2.14 SEVERITY OF FAULT CURRENTS AND THE EFFECT

OF NEUTRAL GROUNDING REACTANCE
In the three-phase system, let the fault occurs at phase ‘a’. The type of fault being
L-G fault and   Z f , the fault impedance being zero, the fault current is the given by

˙ E ˘
I af  ( the fault current in phase a  ) = 3 ˇ ,
ˆ Zii( 0 ) + Zii(1) + Zii( 2) 
1
As I (0) = I (1) = I (2) =
I af , for the single line-to-ground fault, I af = 3 × sequence current.
3
Assuming the resistance component of impedances to be negligible and the posi-
tive sequence reactance being equal to the negative-sequence reactance, we get the
modifed fault current expression. Here

˛ E ˆ
  I af   = 3 ˙ ˘ (2.105)
˝ 2X ii(1) + X ii( 0 ) ˇ
Analysis of Faults in Power System 81

On the other hand, for a solid three-phase fault (where Z f ˜ 0), we have

E
I af = (2.106)
X ii (1)

Comparing Equations (2.105) and (2.106), we fnd that for a three-phase system
where the neutral is solidly grounded and the positive reactance is equal to the nega-
tive reactance and the zero-sequence reactance is of low value, we have the single
line-to-ground fault more severe than the three-phase fault. However, if the zero-
sequence reactance ( X ii (0) ) is of much higher value than the positive-sequence reac-
tance X ii (1), there is a possibility that the three-phase fault is more severe than the
single line-to-ground fault.
For an alternator, usually the positive- and negative-sequence reactances are equal
while the zero-sequence reactance is of lower magnitude. To make the fault current
due to the L-G fault lesser than even the fault current due to three-phase faults, we
need to include a grounded reactance ( X n )  at the neutral so that

3E E
<
2X ii (1) + X ii (0) + 3X n X ii (1)

i.e., 2X ii (1) + X ii (0) + 3X n > 3X ii (1) (2.107)

1
or, Xn >
3
( Xii(1) − Xii(0) )
Thus, it is possible to select the grounding reactance (X n) to limit the L-G fault cur-
rent lower than the L-L-L fault current using a suitable grounding reactance at the
neutral of the generator.

2.15 OPEN CONDUCTOR FAULT

Let the three-phase three-wire line be assumed with open conductor fault occurring
at F (Figure 2.29).
Let the voltage between broken conductors in each phase be Vaa˜ , Vbb˜ and Vcc˜ while
the line currents are I a, I b and I c (for generalisation of treatments, the currents are
assumed even in the broken phases and their respective values will be set to zero
successively during the analysis of broken conductor faults).

° Ia ˙ ° Vaa˘ ˙
˝ ˇ ˝ ˇ
Here,  I = ˝ I b ˇ; V = ˝ Vbb˘ ˇ
˝ Ic ˇ ˝ Vcc˘ ˇ
˛ ˆ ˛ ˆ

˜ Ia ˝ ˜ 1 1 1 ˝ ˜ Ia0 ˝
˛ ˆ ˛ ˆ˛ ˆ
and ˛ I b ˆ=˛ 1 a2 a ˆ ˛ I a1 ˆ
˛ Ic ˆ ˛ 1 a a2 ˆ ˛ Ia2 ˆ
° ˙ ° ˙° ˙
82 Power Transmission System Analysis Against Faults and Attacks

FIGURE 2.29 A three-phase three-wire line with broken conductors.

Positive sequence Negative sequence Zero sequence

network network network

FIGURE 2.30 Sequence network of open conductor fault.

FIGURE 2.31 Single conductor fault at a three-phase three-wire system.

˜ Va ˝ ˜ 1 1 1 ˝ ˜ Va0 ˝
˛ ˆ ˛ ˆ˛ ˆ
˛ Vb ˆ=˛ 1 a2 a ˆ ˛ Va1 ˆ
˛ Vc ˆ ˛ 1 a a2 ˆ ˛ Va2 ˆ
° ˙ ° ˙° ˙
The sequence component of currents and voltages are shown in Figure 2.30.

2.15.1 SINGLE CONDUCTOR OPEN FAULT

Let us now assume that phase “a” is only open (Figure 2.31).
Here,

Vbb° = Vcc° = 0 (2.108)

Ia = 0 (2.109)

Using Equations (2.109) and (2.108) in the concept of symmetrical theory related to
Equations (2.30) and (2.25b), we get

1
Vaa˝(1) = Vaa˝(2) = Vaa˝(0) = Vaa˝ (2.110)
3
Analysis of Faults in Power System 83

and I a(1) + I a(2) + I a(0) = 0 (2.111)

Thus, the sequence network is a partially connected network of individual sequence

networks (Figure 2.32).

2.15.2 TWO CONDUCTOR OPEN FAULT

Let us assume conductor “b” and “c” are open (Figure 2.33).

Obviously, Vaa° = 0; I b = I c = 0

From the concept of symmetrical theory, we have,

Vaa˙(1) +Vaa˙(2) +Vaa˙(0) = 0 (2.112)

1
I a(1) = I a(2) = I a(0) =   I a (2.113)
3
Thus, the sequence network connections would be a series connected network, as
shown in Figure 2.34.

Positive sequence Negative sequence Zero sequence

network network network

FIGURE 2.32 Connection of sequence network for one conductor open circuit fault.

FIGURE 2.33 Open conductor fault at two phases of a three-phase system.

84 Power Transmission System Analysis Against Faults and Attacks

Positive sequence Negative sequence Zero sequence

network network network

FIGURE 2.34 Interconnection diagram of sequence networks for two conductor open fault.

REFERENCE
1. A. Chakrabarti and S. Halder, “Power System Analysis: Operation and Control”, Third
Edition, New Delhi: PHI Learning Pvt. Ltd.
 3 Contingency Analysis

3.1 INTRODUCTION
Contingency analysis is a method to predict steady state bus voltages and line cur-
rents in a power system network following switching (addition or removal) of a line
in that network. It helps in checking the components of the system (e.g. bus or line)
being subjected to overloading or overvoltage/undervoltage conditions following
switching of the prescribed line. Contingency analysis is mostly an offline method
and usually in the analysis, line resistance, presence of off-nominal tap ratios of
transformers and line charging effects are neglected. The linear model of the sys-
tem is usually assumed where the principle of superposition can be applied. The
results, thus, have little approximation. Contingency analysis [1] frequently uses bus
impedance matrix [ Z BUS ] where loads are to be treated as constant current injec-
tions. Addition of a line in the system is simply the addition of an impedance, while
removal of the line is accomplished by adding a negative impedance in the pre-
scribed part of the network. The diagonal elements of [ Z BUS ] provide important
characteristics of the power network and facilitate the use of Thevenin’s equivalent
impedance at designated buses. Application of [ Z BUS ] is evident in power system
contingency analysis.

3.2 RELATIONSHIP BETWEEN THEVENIN’S THEOREM

AND BUS IMPEDANCE MATRIX [ Z BUS ]
Figure 3.1 shows a power network with buses numbered from 0 to n, 0th bus being the
reference bus. Let us assume that bus q is fed by a current source of strength ∆I q p.u.
Let V10, V20, …. Vp0, Vq0, … Vn0 be the initial circuit voltage of the respective buses
1, 2, …, p, q, … n (the voltage between the respective buses and the reference). In this
power network, in general we can write

[V0 ] = [ Z BUS ][ I 0 ] (3.1)

where [V0 ] is the column vector of initial bus voltages and [ I 0 ] is the column vector of
initial respective bus currents. However, due to current injection at bus q, there will
be changes in the bus voltages, and the modified governing equation of the system
is expressed as

[V ] = [ Z BUS ][ I 0 + ∆I ] = [ Z BUS ][ I 0 ] + [ Z BUS ]  [ ∆I ]

(3.2)
= [V0 ] + [ ∆V ]

where, [ ∆V ] = [ Z BUS ][ ∆I ] (3.3)

85
86 Power Transmission System Analysis Against Faults and Attacks

n bus
power
network

FIGURE 3.1 Current injection at q-th bus of a multi-bus power network.

Because ˜I q is the only injected current at bus q with no other current injections at
any other bus, the column vector [ ˛I ] can be expressed as

˙ 0 ˘ ˙ ˛V1 ˘
ˇ  ˇ 
ˇ   ˇ ˛V2 
ˇ 0  ˇ  
[ ˛I ] = ˇˇ ˛I q  while [ ˛V ] = ˇ ˛Vp

 ˇ 
ˇ 0  ˇ ˛Vq 
ˇ   ˇ 
ˇ  ˇ  
ˇˆ 0  ˇ ˛Vn 
ˆ 

Equation (3.3) can be written as

° ˜V1 ˙ ° Z11 Z12 ° Z1q ° Z1n ˙° ˙

˝ ˇ ˝ ˇ˝ 0 ˇ
˝ ˜V2 ˇ ˝ Z 21 Z 22 ° Z2q ° Z2n ˇ˝ 0 ˇ
˝ ˜ ˇ ˝ ˜ ˜ ° ˜ ° ˜ ˇ˝ ˜ ˇ
˝ ˇ=˝ ˇ˝ ˇ (3.4)
˝ ˜Vq ˇ ˝ Z q1 Z q2 ° Z qq ° Z qn ˇ ˝ ˜I q ˇ
˝ ˜ ˇ ˝ ˜ ˜ ° ˜ ° ˜ ˇ˝ ˜ ˇ
˝ ˇ ˝ ˇ˝ ˇ
˝˛ ˜Vn ˇˆ ˝˛ Z n1 Z n2 ° Z nq ° Z nn ˇˆ ˛ 0 ˆ
Contingency Analysis 87

° ˜V1 ˙ ° Z1q ˙
˝ ˇ ˝ ˇ
˝ ˜V2 ˇ ˝ Z 2q ˇ
˝  ˇ ˝  ˇ
i.e. ˝ ˇ=˝ ˇ ˜I q (3.5)
˝ ˜Vq ˇ ˝ Z qq ˇ
˝  ˇ ˝  ˇ
˝ ˇ ˝ ˇ
˝˛ ˜Vn ˇˆ ˝˛ Z nq ˇˆ

(as the nonzero quantity in the current matrix exists in the q-th row only).
At the q-th bus, with initial bus voltage Vq0 , we can express the new q-th bus volt-
age as

Vq = Vq0 + Z qq ˛I q (3.6)

Hence, we can interpret Z qq as the Thevenin’s impedance at bus q where Zth ˜ Z qq.
Obviously, Z qq is the diagonal entry in the q-th row in [ Z BUS ] and is the driving point
impedance of bus q (Figure 3.2).
The next task is to obtain the Thevenin’s impedance between bus p and q so that
the bus voltages are V1 , V2 ,… , Vp , Vq ,… , Vn . As [ ˛V ] = [ Z BUS ] [ ˛I ], we can write

° ˜V1 ˙ ° Z11 Z12 ° Z1p Z1q ° Z1n ˙° ˙

˝ ˇ ˝ ˇ˝ 0 ˇ
˝ ˜V2 ˇ ˝ Z 21 Z 22 ° Z2 p Z 2q ° Z 2n ˇ˝ 0 ˇ
˝ ˇ ˝ ˇ˝ ˇ
˝ ˜ ˇ ˝ ˜ ˜ ˜ ˜ ˜ ˇ˝ ˜ ˇ
˝ ˜Vp ˇ=˝ Z p1 Z p2 ° Z pp Z pq ° Z pn ˇ ˝ ˜I p ˇ
˝ ˜Vq ˇ ˝ Z q1 Z q2 ° Z qp Z qq ° Z qn ˇ ˝ ˜I ˇ
˝ ˇ ˝ ˇ˝ q
ˇ
˝ ˜ ˇ ˝ ˜ ˜ ˜ ˜ ˜ ˇ˝ ˜ ˇ
˝ ˜Vn ˇ ˝ Z n1 Z n2 ° Z np Z nq ° Z nn ˇ˝ 0 ˇ
˛ ˆ ˛˝ ˆˇ ˛ ˆ

° Z1p ˜I p + Z1q ˜I q ˙
˝ ˇ
˝ Z 2 p ˜I p + Z 2q ˜I q ˇ
˝ ˇ
˝ ˜ ˇ
=˝ Z pp ˜I p + Z pq ˜I q ˇ (3.7)
˝ ˇ
˝ Z qp ˜I p + Z qq ˜I q ˇ
˝ ˜ ˇ
˝ ˇ
˝˛ Z np ˜I p + Z nq ˜I q ˇˆ

(for assumed current injections at bus p and q).

Thus, for row p and q, we have

˜Vp = Z pp ˜I p + Z pq ˜I q (3.8a)
88 Power Transmission System Analysis Against Faults and Attacks

FIGURE 3.2 Thevenin’s equivalent circuit at bus q.

˜Vq = Z qp ˜I p + Z qq ˜I q (3.8b)

As ˜Vp = Vp −Vp0 and ˜Vq = Vq −Vq0 , we can further write

Vp = Vp0 + Z pp ˛I p + Z pq ˛I q (3.9a)

Vq = Vq0 + Z qp ˛I p + Z qq ˛I q (3.9b)

Adding and subtracting Z pq ˜I p in Equation (3.9a) and Z qp ˜I q in Equation (3.9b), we

get

Vp = Vp0 + ( Z pp − Z pq ) ˆI p + Z pq ( ˆI p + ˆI q ) (3.10a)

Vq = Vq0 + ( Z qq − Z qp ) ˆI q + Z qp ( ˆI p + ˆI q ) (3.10b)

As [ Z BUS ] is symmetrical in power system networks, Z pq = Z qp . Thus, Equations

(3.10a) and (3.10b) can be rewritten as

Vp = Vp0 + ( Z pp − Z pq ) ˆI p + Z pq ( ˆI p + ˆI q )

and Vq = Vq0 + ( Z qq − Z pq ) ˆI q + Z pq ( ˆI p + ˆI q )

The equivalent circuit is shown in Figure 3.3.

To fnd equivalent impedance (Zth) between bus p and q, the current and voltage
sources are deactivated. The circuit of Figure 3.3 reduces to that shown in Figure 3.4.
Obviously in Figure 3.4, the Thevenin’s impedance between bus p and q is
given by

Zth( p−q) = ( Z pp + Z qq − 2Z pq ) (3.11)

Contingency Analysis 89

FIGURE 3.3 Thevenin’s equivalent circuit of the p-th, q-th bus in the multi-bus power
network.

FIGURE 3.4 Determination of Thevenin’s equivalent impedance between bus p and q.

Assuming a load impedance Zl between bus p and q, the Thevenin’s current through
the load impedance Zl is given by

Vp0 −Vq0
Il =   (3.12)
Zth( p−q) + Zl

where I l is the current through Zl and (Vp0 −Vq0 ) is the open circuit voltage between
bus p and q.
90 Power Transmission System Analysis Against Faults and Attacks

3.3 ADDITION AND REMOVAL OF A LINE IN

A POWER SYSTEM NETWORK
Let V1, V2, …, Vn = Bus voltages (in p.u) in the power system network
I1, I2, …, In = Known bus current injection (in p.u.) at respective buses.
Z x and Zy = p.u. impedances of lines to be added in the system between buses p-q
and r-s, respectively.
Ix and Iy = Currents (in p.u.) in branches Z x and Zy, respectively.
V1˜, V2˜,…,Vn˜  =  bus voltages (in p.u.) in the same power network after addition of
Z x and Zy in the network.
Here,

˝ V1 ˇ ˝ I1 ˇ ˝ V1 ˇ
ˆ  ˆ  ˆ 
ˆ V2  ˆ I2  ˆ V2 
ˆ   ˆ   ˆ  
ˆ  ˆ  ˆ 
Vp Ip ˆ Vp 
ˆ  ˆ 
[V ] = ˆ Vq  ;[I ] = ˆ Iq  ; [V  ] = ˆ Vq 
ˆ  ˆ  ˆ 
ˆ Vr  ˆ Ir  ˆ Vr 
ˆ Vs  ˆ Is  ˆ Vs 
ˆ  ˆ  ˆ 
ˆ   ˆ   ˆ  
ˆ Vn  ˆ In  ˆ Vn 
˙ ˘ ˙ ˘ ˙ ˘

˝ Z11 Z12 ˜ Z1p Z1q Z1r Z1s ˜ Z1n ˇ

ˆ 
ˆ Z 21 Z 22 ˜ Z2 p Z 2q Z 2r Z 2s ˜ Z 2n 
ˆ 
ˆ ° ° ° ° ° ° ° 
ˆ Z p1 Z p2 ˜ Z pp Z pq Z pr Z ps ˜ Z pn 
ˆ 
[ Z Bus ] = ˆ Z q1 Z q2 ˜ Z qp Z qq Z qr Z qs ˜ Z qn 
ˆ Zr1 Zr1 ˜ Zrp Z rq Z rr Zrs ˜ Z rn 
ˆ 
ˆ Z s1 Z s2 ˜ Z sp Z sq Z sr Z ss ˜ Z sn 
ˆ ° ° ˜ ° ° ° ° ˜ ° 
ˆ 
ˆ˙ Z n1 Z n2 ˜ Z np Z nq Z nr Z ns ˜ Z nn ˘

Also, [V ] = [ Z BUS ]   [ I ]

Figure 3.5a represents the system in the initial state, while Figure 3.5b represents the
system after Z x and Zy are added at the designated buses.
From Figure 3.5b, we have

Z x I x = Vp° − Vq° and Z y I y = Vr° − Vs° (3.13)

Equation (3.13) can be represented in matrix form as shown in Equation (3.13a)

Contingency Analysis 91

n bus n bus
power power
network network

FIGURE 3.5 n-bus power system with addition of lines in designated buses.

˜ V1 ˝
˛ ˆ
˛ V2 ˆ
˛ ° ˆ
˛ ˆ
˛ Vp ˆ
˜ Zx 0 ˝ ˜ Ix ˝ x ˜ 0 ˜ 1 −1 0 0 ˜ 0 ˛˝
˛ ˆ˛ ˆ= ˛ Vq ˆ
˛° 0 Zy ˆ ˛ Iy ˆ˙ y ° 0 ˜ 0 0 1 −1 ˜ 0 ˆ˙ ˛ ˆ
˙° ˛ Vr ˆ
˛ Vs ˆ
˛ ˆ
˛ ° ˆ
˛ Vn ˆ˙
°
(3.13 a)
˛ Zx 0 ˆ ˛ Ix ˆ
i.e., ˙ ˘  ˙ ˘ =  [ AC ] [V  ] (AC being the connection matrix) (3.13 b)
0 Zy ˘ ˙ Iy ˘
˝˙ ˇ˝ ˇ
As (−I x ) enters bus p while (+I x) enters bus q and (−I y ) enters bus r while (+I y) enters
bus s in Figure 3.5b, we can express the current vector [ I ˛ ] as

˙ 0 ˘ 1 ˙ 0 ˘
ˇ  0
ˇ 
ˇ   
ˇ   
ˇ −I x  p ˇ −1 0 
ˇ Ix  ˇ 1  ˙ Ix ˘ ˙ Ix ˘
q 0
[ I  ] = ˇ −I = ˇ ˇ I  = − ˙ˆ ACT ˘ ˇ  (3.14)
ˇ y  r ˇ 0 −1  ˇˆ y  ˇˆ I y 
ˇ I  s ˇ 0 1 
ˇ y
 ˇ  
ˇ    
ˇ 
ˇ 0  n ˇˆ 0 0 
ˆ 
92 Power Transmission System Analysis Against Faults and Attacks

Change in bus voltages can be obtained using Equation (3.14)

ˇ Ix 
[ ˛V ] = [V ˙ ] − [V ] = [ Z BUS ][ I ˛ ] = − [ Z BUS ]  ˇ˘ ACT    (3.15)
˘ I y 

Also, [V ˛] = [V ] + [ ˆV ] = [V ] + [ Z BUS ] [ I ˆ ]

ˆ Ix ˘
[V ˛ ] = [V ] − [ Z BUS ]  ˆˇ ACT ˘   (3.16)
ˇ I y 

Substituting Equation (3.16) in (3.13) yields

˛ Zx 0 ˆ ˛ Ix ˆ  ˛ Ix ˆ 
˙ ˘  ˙ ˘ =  [ AC ] [V  ] = [ AC ] [V ] − [ Z BUS ]  ˛˝ ACT ˆˇ ˙ ˘
˙˝ 0 Zy ˘ ˙ Iy ˘ˇ  ˙˝ I y ˘ˇ 
ˇ˝

˛ Ix ˆ
= [ AC ] [V ] − [ AC ] [ Z BUS ]  ˛˝ ACT ˆˇ ˙ ˘
˙˝ I y ˇ˘

˛˛ Zx 0 ˆ ˆ ˛ Ix ˆ ˛ Vp − Vq ˆ
or, ˙˙ ˘ + [ AC ] [ Z BUS ]  ˝˛ ACT ˆˇ ˘ ˙ ˘ = [ AC ] [V ] = ˙ ˘
˙˝ ˙˝ 0 Zy ˘
ˇ ˇ˘ ˙˝
Iy ˘ˇ ˙˝ Vr − Vs ˘ˇ

(3.17)

−1
˛ Ix ˆ ˛˛ Zx 0 ˆ ˆ
or, ˙ ˘ = ˙˙ ˘ + [ AC ] [ Z BUS ] ˛˝ ACT ˆˇ ˘ [ AC ][V ]
I 0 Zy ˘
˝˙ y ˇ˘ ˙˝ ˙˝ ˇ ˇ˘

˛ Ix ˆ ˛ Vp − Vq ˆ
˘ = [ Z ] [ AC ] [V ] = [ Z ] [ AC ] ˙
−1 −1
or, ˙ ˘ (3.18)
˙˝ I y ˘ˇ ˙˝ Vr − Vs ˘ˇ

˝˝ Zx 0 ˇ ˇ
where [Z ] = ˆˆ 0  + [ AC ] [ Z BUS ]  ˝˙ ACT ˘ˇ 
ˆ ˆ˙ Zy  ˘
˙ ˘

(Vp − Vq )) and (Vr − Vs ) are the open circuit voltages between buses (p-q) and
(r-s), respectively In the original power network (Figure 3.5a) and as evident from
Equation (3.18),
Contingency Analysis 93

x ˙ 1 −1 0 0 ˇ
[ AC ][ Z BUS ] ˙ˆ ACT ˇ˘ = y ˆ 0 0 1 −1 ˘

˙ Z pp Z pq Z pr Z ps ˇ ˙ 1 ˇ
  0
 Z pq Z qq Z qr Z qs   −1 0 
×  
Z pr Zrq Zrr Zrs   0 1 
 
 Z ps Z sq Z sr Z ss  ˆ 0 −1 
˘ (3.19)
ˆ ˘

˙ ( Z pp − Z pq ) − ( Zqp − Zqq ) ( Z pr − Z ps ) − ( Zqr − Zqs ) ˇ


=

ˆ ( Zrp − Zrq ) − ( Zsp − Zsq ) ( Zrr − Zrs ) − ( Zsr − Zss ) 
˘

However, from Equation (3.11) the Thevenin’s impedance can be represented as

Zth( p−q) = ( Z pp + Z qq − 2Z pq ) (3.20a)

and, Zth(r −s) = ( Zrr + Z ss − 2Zrs ) (3.20 b)

Finally, substituting Equation (3.19) in (3.17), we can have

˙
ˇ ( Z pp − Z pq ) − ( Zqp − Zqq ) + Z x ( Z pr − Z ps ) − ( Zqr − Zqs ) ˘˙
 ˇ Ix
˘

ˇ
ˇˆ ( Zrp − Zrq ) − ( Zsp − Zsq ) ( Zrr − Zrs ) − ( Z sr − Z ss ) + Z y  ˇ Iy
 ˆ 

˙ Vp − Vq ˘
=ˇ  (3.21)
ˇˆ Vr − Vs 

I x and I y can be obtained from Equation (3.21) to observe the addition of impedances
Z x and Z y between respective buses in the system. The removal of the line imped-
ances Z x and Z y from the original network in Figure (3.5a) can be accomplished in a
similar manner by treating the removals as additions of negative impedances (−Z x )
and (−Z y ) in Figure 3.5b and Equation (3.21).

Example 3.1

In a four-bus power system, the bus voltages are given by

V1 = 1.00∠0° p.u.; V2 = 0.99∠0° p.u., V3 = 0.98∠0° p.u. and V4 = 0.97∠0° p.u.
The [ Z BUS ] matrix is given by

˝ j0.28 j0.25 j0.20 j0.15 ˇ

ˆ 
j0.25 j0.22 j0.15 j0.15
[ Z BUS ] = ˆˆ j0.20 j0.15 j0.20 j0.12

   p.u.
ˆ 
ˆ˙ j0.15 j0.15 j0.12 j0.20 ˘
94 Power Transmission System Analysis Against Faults and Attacks

Two lines Z x and Z y of p.u. reactances j0.08 and j1 p.u. are connected between
buses 2–3 an 3–4, respectively. Find currents I x and I y fowing through line imped-
ances Z x and Z y . Also, fnd the changes in bus voltages.

Solution

Letting bus nomenclatures p = 2, q = 3, r = 3 and s = 4 from Equation (3.21), we
have

˙ ( Z 22 − Z 23 ) − ( Z32 − Z33 ) + Z x ( Z 23 − Z 24 ) − ( Z33 − Z34 ) ˘˙ I ˘ ˙ V −V ˘

ˇ ˇ x  = ˇ 2 3

ˇ
ˆ
( Z32 − Z33 ) − ( Z 42 − Z 43 ) ( Z33 − Z34 ) − ( Z 43 − Z 44 ) + Z y  ˆˇ I y  ˆˇ V3 − V4 


˙ ˘
ˇ ( j0.22 − j0.15) − ( j0.15 − j0.20) + j0.08 ( j0.15 − j0.15) − ( j0.20 − j0.12) 
ˇ ( j0.15 − j0.2) − ( j0.15 − j0.12) ( j0.2 − j0.12) − ( j0.12 − j0.2) + j1 
ˆ 

˙ Ix ˘ ˙ ˘
ס  = ˇ 0.990° − 0.980° 
ˇˆ I y  ˆ 0.980° − 0.970° 

˙ ˘˙ ˘ ˙
( j0.07 + j0.05 + j0.08) (0 − j0.08) ˇ Ix
˘
or, ˇ  = ˇ 0.01 
ˇ (− j0.05 − j0.03) ( j0.08 + j0.08 + j1)  ˇˆ I y  ˆ 0.01 
ˆ 

−1
˜ Ix ˝ ˜ j 0.2 − j0.08 ˝ ˜ ˝ ˜ − j5.14 − j 0.35 ˝˜ ˝
˛ ˆ=˛ ˆ ˛ 0.01 ˆ = ˛ ˆ ˛ 0.01 ˆ
˛° I y ˆ˙ ˛° − j 0.08 j1.16 ˆ˙ ° 0.01 ˙ ˛° − j 0.35 − j 0.88 ˆ˙ ° 0.01 ˙

˜ − j0.055 ˝
=˛ ˆ  p.u. (3.21a)
˛° − j0.012 ˆ˙

The change in bus voltages can be calculated as

[ ˛V ] = −I x ˇˆ Z BUS
p− q (r − s ) ˘
˘ − I y ˆˇ Z BUS 

˜ Z1p − Z1q ˛
ˇ ˘
ˇ Z 2 p − Z 2q ˘ ˜ ˛
ˇ ˘ ˇ Z12 − Z13
 ˘
ˇ ˘ Z 22 − Z 23
˜ ( p− q) ˛ ˜ (2−3) ˛
where, ° Z BUS ˝ = ° Z BUS ˝ = ˇ Z pp − Z pq ˘ = ˇˇ ˘
˘
ˇ Z32 − Z33
ˇ Z qp − Z qq ˘˘ ˇ ˘
ˇ Z 42 − Z 43 ˘˝
ˇ  ˘ °
ˇ Z np − Z nq ˘˝
°
Contingency Analysis 95

 j 0.25 − j 0.20   j 0.05 

   
( 2 − 3)   j 0.22 − j 0.15   j 0.07 
∴  Z BUS =
   =  − j 0.05 
j 0.15 − j 0.20
   
 j 0.15 − j 0.12   j 0.03 

 Z1r − Z1s 
 
 Z 2r − Z 2 s 
  
   Z13 − Z14 
 Z pr − Z ps   
Z 23 − Z 24
(r − s )   ( )  
Also,  Z BUS 3 − 4 = 
 =  Z BUS  =  Z qr − Z qs
  Z 33 − Z34 
 Zrr − Zrs   Z 43 − Z 44

 Z sr − Z ss   
 
  
 Z nr − Z ns 

 j 0.20 − j 0.15   j 0.05 

   
 j 0.15 − j 0.15   j0 
=  =  j 0.08 
j 0.20 − j 0.12
   
 j 0.12 − j 0.20   − j 0.08 

 ∆V1   j 0.05   j 0.05 

     
∆V2  = − I  j 0.07   j0 
∴ x   − I y  j 0.08 
 ∆V3  − j 0.05
     
 ∆V4   j 0.03   − j 0.08 


Substituting the values of I x and I y from Equation (3.21a), the bus voltage changes
are obtained as

 ∆V1   −0.02756 
   
 ∆V2  =  −0.0038

 ∆V3   +0.0266 
   −0.0007 
 ∆V4   

3.4 CURRENT DISTRIBUTION FACTOR (Ψi ) AND

LINE OUTAGE FACTOR (λ )
Let us assume the current injection at bus r is altered in a multi-bus power network.
This will cause change in bus voltages in the system as

i.e., [ ∆V ] = [V ′] − [V ] = [ Z BUS ][ I ∆ ]
96 Power Transmission System Analysis Against Faults and Attacks

˝ ˛V1 ˇ ˝ V1− V1 ˇ
˝ 0 ˇ
ˆ  ˆ 
ˆ 0 
ˆ ˛V2  ˆ V2 − V2 
ˆ  ˆ  ˆ 
ˆ   ˆ
  ˆ  
or, ˆ ˛Vp =ˆ Vp − Vp  = [ Z BUS ] ˆ   [column r of Z BUS ] ˛I r (3.22)
ˆ  ˆ  ˆ 
ˆ
˛Vq
 ˆ
Vq − Vq  ˆ ˛I r 
ˆ  ˆ  ˆ  
  ˆ 0 
ˆ ˛Vn  ˆ Vn − Vn

˙ ˘
˙ ˘ ˆ˙ ˘

Here, we assumed nonprimed voltages to be the bus voltages before any current injec-
tion and is available from load fow study. The primed voltage quantities are respec-
tive bus voltages following current injection (  ˜I r ) at bus r only. The current injection
is constant, and for each line the line changing capacitance and any shunt connection
to the line are omitted. All bus voltages are with respect to the reference bus.
Moreover, we have

˜Vp = Z pr ˜I r
(3.23)
and, ˜Vq = Z qr ˜I r

Let us now assume that Z pq   is the p.u. impedance of line between buses p and q. The
change of line current in Z pq can then be computed as

˜Vp − ˜Vq Z pr ˜I r − Z qr ˜I r ˝ Z − Z qr ˇ
˜I pq = = = ˜I r ˆ pr 
Z pq Z pq ˙ Z pq ˘
(3.24)
˜I Z − Z qr
 pq = pr
˜I r Z pq

° ˜I ˙
Conventionally, the ratio ˝ pq ˇ is termed as current injection distribution factor
( ˛ i ). ˛ ˜I r ˆ

˜I pq
i.e., = ˛i
˜I r (3.25)
or, ˜I pq = ˛ i ( pq)r ˜I r

In a power system network, often the load increases in a bus leading to line loading
beyond permissible value, and then it is required to reschedule the line loadings to
keep the line loading within safe values. Mathematically, this is equivalent to change
of current injection by ˜I m and ˜I n at buses m and n for reducing the line loading of
line between buses m-n. Equation (3.25) can be rewritten as

˜I pq = ˛ i ( pq)m ˜I m + ˛ i ( pq)n ˜I n (3.26)

Contingency Analysis 97

The next concept we would like to introduce is the line outage distribution factor (˜ ).
˛I pq
˜ pq(m−n) =
I mn
(3.27)
˛I
and, ˜rs(m−n) = rs
I mn
Clearly, the post contingency line currents in lines p-q and r-s are

I°pq = I pq + ˙I pq = I pq + ˜ pq(m−n) I mn (3.28a)

and I rs° = I rs + ˙I rs = I rs + ˜rs(m−n) I mn (3.28b)

Thus, it is now evident that if the pre-contingency line currents are known from usual
load fow as well as the line outage distribution factors ˜ pq   and ˜rs for tripping a line
m-n are known, we can obtain the values of post-contingency line currents I pq ' and
I rs ' in the healthy lines p-q and r-s.

3.5 SINGLE LINE CONTINGENCY

Let us assume that a line Lmn is tripped between buses m and n. Following the con-
cepts developed in the preceding chapters, the removal of the line Lmn my be con-
sidered to be adding an impedance (−Z x ) between buses m and n. Obviously it is
assumed that the magnitude of the imaginary impedance (−Z x ) is identical to the
magnitude of the impedance of the line Lmn removed.
Figure 3.6 represents the Thevenin’s equivalent circuit where (I x) is the current
fowing through (−Z x ). Here,

Vm −Vn
Ix = (3.29)
( Z mm + Z nn − 2  Z mn ) + (− Z x )

FIGURE 3.6 Removal of a line between buses m-n is equivalent to addition of (−Z x ) imped-
ance between buses m-n in Thevenin’s equivalent circuit.
98 Power Transmission System Analysis Against Faults and Attacks

Here (Vm ) and (Vn ) are the pre-contingency voltages at buses m and n, respectively.
On the basis of the analytical reasoning in the preceding articles, we can say that
the addition of impedance (−Z x ) between buses m-n would cause current ˜I m (= −I x) in
bus m and ˜I n (= I x ) in bus n. Thus, the resulting current variation in the line between
buses p-q can be expressed by the Equation (3.30).

˛I pq = ˙ i ( pq)m ˛I m + ˙ i ( pq)n ˛I n

Z pm − Z qm Z − Z qn (3.30)
=   (−I x ) + pn  ( I x )
Z pq Z pq

assuming Z pq to be the impedance for line p-q

ˆ ( Z pn − Z pm ) − ( Z qn − Z qm ) 
or, ˛I pq = I x ˘  (3.31a)
ˇ Z pq 

ˇ Vm −Vn   ( Z pn − Z pm ) − ( Z qn − Z qm ) 
˛I pq =    (3.31b)
˘ Z mm + Z nn − 2  Z mn − Z x   Z pq 

However, during the state of pre-contingency, when the line Lmn was not removed,
the current through this line (i.e. the current through the impedance (−Z x )) was
expressed as

Vm −Vn
I x(0) = (3.32)
Zx

I x (0)   is the current through the line Lmn (= Z x ) in the pre-contingency state, while I x
is the current through (−Z x ) following line removal.
It may be noted here that I x (0) ˜ I x
Substituting Equation (3.32) in (3.31), we get

ˆ ( Z pn − Z pm ) − ( Z qn − Z qm )  I x(0) Z x
˛I pq = ˘  
ˇ Z pq  Z mm + Z nn − 2  Z mn − Z x

 Z  ˆ ( Z pn − Z pm ) − ( Z qn − Z qm )  (0)
=  x  ˘ I x
 Z pq  ˇ Z mm + Z nn − 2  Z mn − Z x 

˛I pq ˆ Z   ( Z pm − Z pn ) − ( Z qm − Z qn ) 
i.e., (0)
= −˘ x    (3.33)
Ix ˇ Z pq   Z mm + Z nn − 2  Z mn − Z x 

˛ °I pq ˆ
Here, ˙ ( 0 ) ˘ is the line utage distribution factor ˜ pq( m− n )
˝ Ix ˇ
Contingency Analysis 99

Then we can say from Equation (3.33)

ˆ Z   ( Z pm − Z pn ) − ( Z qm − Z qn ) 
˜ pq(m−n) = − ˘ x    (3.34)
ˇ Z pq   Z mm + Z nn − 2  Z mn − Z x 

In a similar manner, for removal of line Lmn, the effect on line between buses r and
s can be represented as

ˆ Z   ( Z − Zrn ) − ( Z sm − Z sn ) 
˜rs( m− n ) = − ˘ x     rm (3.35)
ˇ Zrs   Z mm + Z nn − 2  Z mn − Z x 

Thus, from Equations (3.34) and (3.35), we can fnd new line currents Ip̃q and I rs˜ fol-
lowing a line contingency in the power system between buses m-n.

 Z x   ( Z pm − Z pn ) − ( Z qm − Z qn )  (0)
I′pq − I pq = ∆I pq = −
     I x
 Z pq   Z mm + Z nn − 2 Z mn − Z x 

 Z   ( Z pm − Z pn ) − ( Z qm − Z qn )  (0)
i.e., I′pq = I pq   −  x    I x (3.36a)
 Z pq   Z mm + Z nn − 2  Z mn − Z x 

i.e., I°pq = I pq + ˜ pq(m−n) I x(0) (3.36b)

ˆ Z   ( Z − Zrn ) − ( Z sm − Z sn )  (0)
i.e., I rs˛ = I rs   − ˘ x   rm Ix (3.37a)
ˇ Zrs   Z mm + Z nn − 2Z mn − Z x 

i.e., I rs° = I rs + ˜rs(m−n) I x(0) (3.37b)

Example 3.2

Assume fve bus power system. The [ Z BUS ] (in p.u.) is given by

˝ j20.52 j20 j20.09 j20.25 j20.15 ˇ

ˆ 
ˆ j20 j20.85 j20.05 j20 j20.58 
[ Z BUS ] = ˆˆ j20.09 j 20.05 j 20 j 20.61 j 20.56 

ˆ j20.25 j20 j20.61 j20.89 j20.03 
ˆ j20.15 j20.58 j20.56 j20.63 j20.09 
˙ ˘

Pre-contingency bus voltages are given as

V1 = (1 + j0 ) p.u.; V2 = ( 0.99 − j0.08) p.u.; V3 = ( 0.98 − j0.085) p.u.;

V4 = ( 0.975 − j0.06) p.u.; V5 = ( 0.98 − j0.05) p.u.

100 Power Transmission System Analysis Against Faults and Attacks

Assume removal of the line between buses 2 and 4. The series impedance of line
between buses 2 and 4, that is, L34 are j0.05 and j0.1 p.u., respectively. Compute
the new current in the line between buses 3 and 4. In addition, obtain the value of
the line outage distribution factor of L34 for removal of line L24.

Solution
Let p = 3; q = 4; m = 2; n = 4
From Equation (3.34), we can write

ˆ Z 24 L   ( Z − Z 34 ) − ( Z 42 − Z 44 ) 
˜34 (2−4) = − ˘˘ ( )   32 
ˇ Z 34 ( L )   Z 22 + Z 44 − 2  Z 24 − Z 24 

j0.05  ( j20.05 − j20.61) − ( j20 − j20.89) 

=−  
j0.1  j20.85 + j20.89 − 2 ( j 20 ) − j0.05 

 − j0.56 + j0.89 
= −0.5  = −0.0976
 j1.69 

˜ Line outage distribution factor of line L34 for outage of line L24 has been obtained
as (−0.0976).

V3 − V4 ( 0.98 − j0.085) − ( 0.975 − j0.06)

Also, I 34 ( pre contingency ) = =
Z 34 ( L ) j0.1

0.005 − j0.025
= = − ( 0.25 + j0.05) p.u.
j0.1

V2 − V4 ( 0.99 − j0.08) − ( 0.975 − j0.06)

I 24 ( pre contingency ) = =
Z 24 ( L ) j0.05

0.015 − j0.02
= = − ( 0.4 + j0.3) p.u.
j0.05

Following Equation (3.36) we now have, for new line current in L34,

ˆ = I 34 − ˜34 (2−4) I 24 = − ( 0.25 + j0.05) − {(−0.0976) − ( 0.4 + j0.3)}

I 34

= (−0.21− j0.079) p.u. = − ( 0.21+ j0.079) p.u.

3.6 CONTINGENCY ANALYSIS OF INTERCONNECTORS

Sub-grids in a large power system are interconnected by interconnectors (tie-lines),
and the contingency analysis of such lines is important for the power system plan-
ning and operation. [ Z BUS ] technique is an effective tool in contingency analysis of
interconnectors.
Figure 3.7 represents the schematic of two interconnectors being connected
between two sub-grids A and B. Let the interconnector connecting bus p of area A to
Contingency Analysis 101

FIGURE 3.7 Two sub-grids A and B are interconnected through interconnections.

bus q of area B has impedance Z x , while that connecting bus m of area A and bus n
of area B has impedance Z y .
First, let us assume that the sub-grids are not interconnected and V1A , V2 A ,… , VmA
are the bus voltages for the buses 1, 2, …, m in area A, while V1B , V2B ,… , VnB are
the bus voltages for the buses 1, 2, …, n in area B. The respective bus impedance
metrices of area A and area B are [ Z BUS ] A and [ Z BUS ] B during isolated operation
of area A and area B. Moreover, the current injections in the respective buses are
I1A , I 2 A ,… ,  I mA for area A and I1B , I 2B ,… ,  I nB for area B.
Using the relation [V ] = [ Z BUS ]   [ I ] , we have for this system,

° V1 A ˙ ° I1 A ˙
˝ ˇ ˝ ˇ
˝ V2 A ˇ ˝ I2 A ˇ
˝  ˇ ˝  ˇ
˝ ˇ ˝ ˇ
˝ VpA ˇ ˝ I pA ˇ
˝ VmA ˇ ° ˙ ˝˝ I mA ˇ
˝ ˇ ZA 0 ˇ
˝ … ˇ=˝ ˇ˝ … ˇ (3.38)
V1B ˇ ˝˛ 0
ZB ˇˆ I1 B ˇ
˝ ˝
˝ V2 B ˇ ˝ I2B ˇ
˝ ˇ ˝ ˇ
˝  ˇ ˝  ˇ
˝ VqB ˇ ˝ I qB ˇ
˝ ˇ ˝ ˇ
˝˛ VnB ˇ ˝˛ I nB ˇ
ˆ ˆ

where, Z A ˛ [ Z BUS ] A and Z B ˛ [ Z BUS ] B

Let us now assume that the tie-lines are interconnected and the currents fowing
through these lines are I x   and I y, respectively. Let the current fows alter the respec-
tive bus voltages from V1A to V1˜A , V2 A to V2˜A … VpA to VpA
˜ , and so on in area A and V1B
to V1˜B , V2B to V2˜B … VpB to VpB
˜ and so on for area B. The connection matrix can then
be expressed as
102 Power Transmission System Analysis Against Faults and Attacks

[ AC ] = ˝˙ ACA ACB ˆˇ  

x ˝ 0 0 ˜ 1 0 ° 0 0 … 1 0 ˆ
=  (3.39)
y  0 0 ˜ 1 0 ° 0 0 … 0 −1 ˇ
˙
Following Equation (3.16), we have

˛ ˜
V1A ˆ ˛ ˆ
˙ ˘ ˙ V1 A
˘
˙ V2˜A ˘ ˙ V2 A˘
˙  ˘ ˙  ˘
˙ ˘ ˙ ˘
˙ Vp̃A ˘ ˙ VpA ˘
˙ ˜ ˘ ˙
VmA VmA ˘ ˛
˙ ˘ ˙ ˘ ZA 0 ˆ ˛ T ˆ ˛ Ix ˆ
˙ … ˘=˙ … ˘−˙ ˘ AC ˙ ˘ (3.40)
Z B ˘ˇ ˝ ˇ ˙˝ I y
˙ V1B˜ ˘ ˙ V1B ˘ ˙˝ 0 ˘ˇ
˙ ˘
˙ V2˜B ˘ ˙ V2B ˘
˙ ˘
˙  ˘ ˙  ˘
˙ ˜ ˘ ˙
VqB VqB ˘
˙ ˘ ˙ ˘
˙ ˜ ˘ ˙˝
VnB VnB ˘
˝ ˇ ˇ

In Equation (3.21) letting bus r as m and bus s as n we can write

˝ ˇ
ˆ ( Z pp + Z qq + Z x ) ˜ ( Z pn + Z qn ) ˝ I ˇ ˝ Vp − Vq ˇ
ˆ ° ° ° ˆ x =ˆ  (3.41)
ˆ  ˆ˙ I y ˘ ˆ˙ Vm −Vn ˘
ˆ˙ ( Z mp + Z nq ) ˜ ( Z mm + Z nn + Z y ) ˘

Because there are no interconnections between buses p-n and m-q hence   Z pn = Z np
and Z mq = Z qm are nonexistent quantities.
−1
˝ ˇ
˝ Ix ˇ ˆ ( Z pp + Z qq + Z x ) ˜ ( Z pn + Z qn )  ˝ V −V ˇ
˛ˆ =ˆ ° ° °  ˆ p q  (3.41a)
ˆ˙ I y ˘ ˆ  ˆ˙ Vm −Vn ˘
ˆ˙ ( Z mp + Z nq ) ˜ ( Z mm + Z nn + Z y ) ˘

ˇ 
 ( Z pp + Z qq + Z x ) ˜ ( Z pn + Z qn ) 
Here, [ Z ] =  ° ° ° 
 
˘ ( Z mp + Z nq ) ˜ ( Z mm + Z nn + Z y ) 
(3.42)
ˇ Z pp Z pm  ˇ Z x 0  ˇ Z qq Z qn 
= +  +  
˘ Z mp Z mm  ˘ 0 Zy  ˘ Z nq Z nn 
˛˙˙˝˙˙ˆ ˛˙˙˝˙˙ˆ
Sub matrix for system-A Sub matrix for system-B
Contingency Analysis 103

[ Z ] is the [ Z BUS ] matrix for the system with interconnections connected. Moreover,
the change in bus voltages can be obtained using Equation (3.15).

˝ ˛V1 A ˇ ˝ V1A ˇ ˝ V ˇ
ˆ  ˆ  ˆ 1A  ˝ Ix ˇ
ˆ  =ˆ   ˆ  = − Z ˝ T ˇ
 − ˆ VpA [ BUS ] A ˙ ACA ˘ ˆ  (3.34a)
ˆ ˛VpA  ˆ VpA  ˆ˙ I y ˘
ˆ  ˆ  ˆ 
ˆ˙ ˛VmA 
˘ ˆ˙ VmA ˘ ˆ˙ VmA ˘

˝ ˛V1B ˇ ˝ V1B ˇ ˝ V1B ˇ

ˆ  ˆ  ˆ  ˝ ˇ
and ˆ
 =ˆ  −ˆ   = − [ Z ] ˝ AT ˇ ˆ I x 
BUS B ˙ CA ˘
ˆ ˛VpB  ˆ VpB   ˆ VpB  ˆ˙ I y ˘
ˆ  ˆ  ˆ 
ˆ˙ ˛VnB ˘ ˆ˙ VnB ˘ ˆ˙ VnB ˘

Thus, the line currents can be obtained from Equation (3.41a), while the change in
bus voltages are obtained from Equations (3.43a) and (3.43b).

3.7 CONTINGENCY ANALYSIS EMPLOYING

DC POWER FLOW MODEL
In the DC power fow method, we assume that the power system is loss-less and
each transmission line is represented by series reactance only; the sunt charging is
neglected. Angular difference between voltages at adjacent buses i − j are assumed
to be small so that sin (˜i − ˜ j ) ˙ (˜i − ˜ j ) rad;  and cos ˜i ° cos ˜ j . It has been assumed
that the system voltages at the buses are 1.00 p.u. The p.u. current fow from the i-th
bus to the   k -th bus through a series line reactance of ( jxl ) p.u. is expressed as

Vi − V j Vi ( cos ˜i + j sin ˜i ) − V j ( cos ˜ j + j sin ˜ j )

I i− j = =
jxl jxl
(3.44)
or, I i − j =
(cos ˜i − cos ˜ j ) + j (sin ˜i − sin ˜ j ) = ˜i − ˜ j p.u.
jxl xl

( V i = V j = 1.00 p.u.)

˜ I i − j becomes a real quantity. The p.u. power fow in the connecting line between
bus i − j is expressed as

Vi V j ˜i − ˜ j
Pi − j = sin (˜i − ˜ j ) ˆ p.u. (3.45)
xl xl

Comparing Equations (3.44) and (3.45) we see that once the voltages at buses of the
power network are assumed to be 1.00 p.u., the line current and power fow lead to
104 Power Transmission System Analysis Against Faults and Attacks

the same expression; hence, both are synonymous in the DC power flow method, that
δ −δ
is,  I i − j ≡ Pi − j = i j
xl
∆I i − j, the change in line current is given by

∆ (δi − δ j ) ∆δi − ∆δ j
∆I i − j = =  p.u (3.46)
xl xl

Next, we look at Equation (3.27) to apply the concept of line outage factor λi − j (m−n) for
assessing the line outage factor λ for line i − j due to assumed removal of line m − n in
the power system network. Following Equation (3.34) we can write

 Z   ( Zim − Zin ) − ( Z jm − Z jn ) 
λi − j (m−n) = − x   
 Zij   Z mm + Z nn − 2 Z mn − Z x 

∴ I i′− j ( new flow of current in line i − j following contingency at line m − n )

 Z   ( Zim − Zin ) − ( Z jm − Z jn )  (0)

= I ij  ( i.e., pre contingency line current in i − j ) −  x    Ix
 Zij   Z mm + Z nn − 2  Z mn − Z x 
(3.47)

Where, for removal of line m − n, Z x ≡ Z mn and

Vm − Vn
I x(0) = ≡ I x( 0)
  Z mn  

∆I ij
Since λi − j (m−n) =
I mn
(see equation 3.27) , in dc pwer flow model

∆Pij  Z   ( Zim − Zin ) − ( Z jm − Z jn ) 

λi − j( m − n ) = = − x    (3.48)
Pmn  Zij   Z mm + Z nn − 2 Z mn − Z x 

3.8 CONCEPT OF EQUIVALENCING IN POWER

SYSTEM NETWORK
In a complex power system, different areas in the power system network are intercon-
nected through tie-lines. Once the contingency analysis is done for one system, the
other system (i.e. the external system network) needs to be represented by a reduced
equivalent network so that the modified bus impedance matrix is available.
Let us assume that tie-line T − 1 interconnects bus i of area 1 to bus j of area 2, as
shown in Figure 3.8. Let [YBUS ]1 and [YBUS ]2 be the respective bus admittance matri-
ces. The overall [YBUS ] of the entire system is expressed as
Contingency Analysis 105

FIGURE 3.8 Two areas of power system interconnected by tie-lines.

˝ ˜ Ykl 0 ° 0 ˇ
kˆ 
iˆ ˜ 0 Yij ° 0 
ˆ 
˜ˆ ˜ ˜ ˜ ° ˜ 
˜ ˆ ˜ 0 0 ° 0 
[ BUS ]
Y = ˜ˆ ° ° ° ° ˜ ° ° ° °  (3.49)
ˆ 
lˆ Ylk 0 ° 0 ˜ 
jˆ 0 Y ji ° 0 ˜ [YBUS ]2 
˜ ˆˆ ˜ ˜ ° ˜ ˜


1 ˆ 0 0 ° 0 ˜ 
˙ ˘

Once the contingency analysis of area 2 is attempted, proper equivalencing of net-

work at area 1 is required. We can use the popular Kron’s network reduction tech-
nique and use the following expression

Yin ×Ynj
Yij (new) = Yij (old) − ,
Ynn

where n is the node number that is to be eliminated. Kron’s network reduction tech-
nique is applied to [YBUS ]1. Let the new bus admittance matrix of area 1 be [YBUS ]˛1
where all the nodes except nodes k and i are eliminated;

˙ Y˛ Yki˛ ˘
i.e.,   [YBUS ]1˛ = ˇ 
kk

ˆˇ ik˛
Y Yii˛ 

Equation (3.49) can now be modifed as

106 Power Transmission System Analysis Against Faults and Attacks

˙ Ykk˝ Yki˝ ° Ykl 0 ˜ 0 ˘

k ˇ 
i ˇ Yik̋ Yii˝ ° 0 Yij ˜ 0 
ˇ ˜ ˜ ˜ ˜ ˜ ˜ ˜ 
˜ ˇ
[YBUS ]new = l ˇ Ylk˝ 0 °  (3.50)
j ˇ 0 Y ji˝ ° [YBUS ]2 
ˇ 
° ˇ ° ° ° 
1 ˇ 0 0 ° 
ˆ 
The voltage current equation can then be written as

˝ I k˛ˇ ˝ Ykk˛ Yki˛ ° Ykl 0 ˜ 0 ˇ ˝ Vk˛ ˇ

ˆ  ˆ ˆ 
ˆ I i˛ ˆ Yik˛ Yii˛ ° 0 Yij ˜ 0  ˆ Vi˛ 
ˆ ˜  ˆ ˜ ˜ ˜ ˜ ˜ ˜ ˜  ˆ ˜ 
ˆ  ˆ ˆ 
ˆ I l˛=ˆ Ylk 0 °  ˆ Vl˛  (3.51)
ˆ I ˛j ˆ 0 Y ji ° [YBUS ]2 ˆ V˛ 
ˆ  ˆ ˆ j 
ˆ °  ˆ ° ° ° ˆ ° 
ˆ
˙ I1˛ ˘ ˙ˆ 0 0 ° ˆ 
˘ ˙ V1˛ ˘
In the next phase, bus k and i are eliminated. Kron’s Node elimination technique is
used to eliminate bus k and i of area 1 and to retain buses of area 2 only. This gives us

° Y˜ Ylj̃˜ ˜ Yl˜1̃ ˙° ˙ ° ˙
˝ ll̃ ˇ ˝ Vl˜ ˇ ˝ I˜l ˇ
˝ Y jl˜˜ Yljj˜˜ ˜  Y j˜1̃ ˇ ˝ V j˜ ˇ ˝ I ˜j ˇ
˝ ˇ˝ ˇ=˝ ˇ (3.52)
˝ ° ° ° ˇ˝ ° ˇ ˝ ° ˇ
˝˛ Y1˜˜l Y1˜j̃ ˜ Y11˜˜ ˇˆ ˛˝ V1˜ ˆˇ ˛˝ I1˜ ˆˇ

The new [ Z BUS ] with respect to area 2 having equivalent representation of buses of
area 1 is as shown below
−1
˙ Y ˛˛ Ylj˛˛ ˜ Yl1˛˛ ˘
ˇ ll 
ˇ Y jl˛˛ Yljj˛˛ ˜  Y j˛˛1 
[ Z BUS ]˛˛ = ˇ  (3.53)
ˇ ° ° ° 
ˆˇ Y1˛˛l Y1˛˛j ˜ Y11˛˛ 

while the new voltages for the buses in area 2 are given by

° Z ˜˜ Zlj˜˜ ˜ Zl˜1̃ ˙ ° I l˜ ˙ ° Vl˜ ˙

˝ ll ˇ˝ ˇ ˝ ˇ
˝ Z˜˜jl Z˜˜jj ˜ Zl˜1̃ ˇ ˝ I j̃ ˇ ˝ V j˜ ˇ
˝
° ˇ ˝˝
ˇ ˇ=˝ ˇ (3.54)
˝ ° ° ° ˇ ˝ ° ˇ
˛˝ Z1̃l̃ Zl˜˜j ˜ Z1̃˜1 ˇˆ ˛˝ I1˜ ˇˆ ˛˝ V1˜ ˆˇ
Contingency Analysis 107

i.e., [V ˛] = [ Z BUS ]˛˛ [ I˛]

The elements of [ Z BUS ]˛˛ has external network equivalents of area 1, and hence, it is
possible to conduct the study of contingency analysis of the system in area 2 with
equivalent network of area 1 embedded in the new [ Z BUS ].   I l˜ and   Ij̃ are equivalent
current injections at buses l and j, the boundary buses, while other bus current injec-
tions remain unchanged.

REFERENCE
1. A. Chakrabarti and S. Halder, “Power System Analysis: Operation and Control”, Third
Edition, New Delhi: PHI Learning Pvt. Ltd.
 4 Fundamental
Concepts of Complex
Network Theory

4.1 INTRODUCTION
Complex network theory has gained extensive approval and has been successfully
applied in the assessment of complex systems such as computer networks, social
interacting species, internet and many more. With the evolution of complex network
theory, most complex systems in the universe can be abstracted as networks consist-
ing of a set of edges connecting a set of vertices. The inherent structural features of
these networks are assessed statistically using metrics. Earlier, the concept of com-
plex network theory was applied in abstracted networks such as random network
[1], small-world networks [2] and scale-free networks [3,4]. Scale-free networks are
fragile to intentional attacks but robust against random failures of nodes [4]. They are
more prone to cascading failure triggered by intentional attacks rather than random
networks [5,6].
The complexity of power systems arises not just from the instant power balance
of generators and consumers in large-scale transmission network across multitude of
countries but also from the decision making of system operators to keep the system
secure and reliable. Furthermore, there is a strong link between topological structure
and operational performance in power systems. For instance, a large-scale blackout
is more possible to be triggered by removing some critical buses or lines, which are
essential elements of the topological structure of power systems. Consequently, power
systems are naturally analysed under the framework of complex networks [7–9].
Despite the application of complex network theory in simulation of an electrical
power system, it would not be appropriate until it is employed with power system-
related constraints and features. Hence, it is more appropriate to utilise the concepts of
complex network theory in assessing the structural vulnerability of electrical power
grid associated with appropriate constraints and operational aspects related to electri-
cal power system. Literature survey reveals that attempts have already been made to
analyse the structural vulnerability in the North American [10] and European power
grids [11–13]. In this chapter, the complex network theory has been employed to
analyse its features related to vulnerability taking into account the associated param-
eters of power system such as electrical distance, line fow limit and power transmis-
sion distribution. These parameters are inserted through appropriate mathematical
framework into the traditional complex network metrics such as degree, geodesic
distance, effciency and betweenness. The concept of extended topological metrics
of electrical betweenness and netability have been simulated to assess the vulner-
ability of the power network. Furthermore, this chapter explains the theoretical and
109
110 Power Transmission System Analysis Against Faults and Attacks

mathematical concepts of resilience originating from percolation threshold, incor-

porating the concept of preferential probability in modelling of cascading failures.

4.2 PURE TOPOLOGICAL APPROACH

In conventional complex network theory approach, there is a primary and essential
set of centrality indices which are utilised for measuring the importance of a vertex
or an edge in a network. These indices have been classifed into three categories: the
frst being that the centrality of a vertex in a network determines how it is related
to other vertices (degree centrality), the second one is based on the concept that
central vertices stand between others (betweenness centrality) and the third, namely
effciency, quantifes the contribution of a network performance when subjected to
removal of vertices or edges. To determine the criticality of the component and (or)
networks, these metrics form the base of the pure topological method.
In the complex network theory, each bus of the power system, which may be a
power source or a power sink, can be modelled as a vertex (or node), and each trans-
mission line and transformer can be modelled as an edge (or line), in which power
fow may be transmitted between its terminals. The actual ability of a power trans-
mission system to perform properly depends on its topological structure, impedance
and fow limits of its lines [14].
The vulnerability is the ability of a network continuing to provide key services dur-
ing random failures or intentional attacks. Complex network theory provides a feasible
way to study the vulnerability of power grids, which has drawn a link between the
topological structure and the vulnerability of networks. There are some salient con-
cepts that are required for the vulnerability analysis of power grid complex network.

4.2.1 GEODESIC DISTANCE

The number of lines in a path connecting nodes i and j is called the length of the path.
A geodesic path [15] (or the shortest path) between i and j is the path connecting
these nodes with the minimum length. The length of the geodesic path is the geode-
sic distance dij between i and j. If one is dealing with a weighted graph, the length of
a path is the sum of the weights of the lines constituting that path.

4.2.2 AVERAGE SHORTEST PATH LENGTH

The distance dij between the node i and j is defned as the number of the edges which
form the shortest path connecting node i and j. The average shortest path length L is
defned [2] as the mean of geodesic lengths over all node couples:

L=
1
N   ( N −1) ∑d ij (4.1)
i , j =1

In a purely topological model, a power network is considered to be a network com-

posed of vertices (buses) connected by edges (transmission lines). In most cases, an
Complex Network Theory 111

unweighted and undirected network model is utilised. All the vertices and edges are
considered to be identical, without differences in their quantitative features or direc-
tions. In an unweighted and undirected graph, the length of a path is the number of
edges in a path connecting vertices i and j.

4.2.3 DEGREE (CONNECTIVITY)

The degree or connectivity [15] of a node is traditionally measured by its degree in an
unweighted topological model or its strength in a weighted model. In an unweighted
and undirected network model (according to traditional graph theory), the degree of
a vertex i is the number of edges connected to it (or the number of vertices adjacent
to it);

ci = ˜l ij (4.2)
j

where lij represents the number of lines connecting i and j.

In a weighted network model, connectivity can also be expressed by the strength
measured as the sum of the weights of the corresponding edges:

si = ˜w ij (4.3)
j

where wij represents the weight of the line connecting i and j.

Cumulative degree distribution being an essential feature of the topological
confguration of a network, it can be suitably employed in assessing the vulner-
ability of a network when the cumulative degree distribution of a network follows a
Poisson’s distribution. It becomes a homogenous network where each node has the
same degree. However, if the distribution is a power law or exponential, the cor-
responding network is heterogeneous, in which there are some vertices with higher
degree than others.
The degree indicates the connectivity of a node in a network; if a node has higher
connectivity, the node has more connections between other nodes and is more rel-
evant. Therefore, degree could be treated as a metric to measure the criticality of the
nodes in networks.

4.2.4 DISTANCE AND EFFICIENCY

The power grid can be abstracted into the complex network with a graph composed
of numbers of buses as a set of nodes and numbers of lines as a set of links [15]. The
total number of nodes and links of the graph are N and L, respectively. The associa-
tion of nodes with each other can be shown using an adjacency matrix. The walk of
minimal length between two nodes is known as the shortest path or geodesic path,
where a walk from node i to node j is an alternating sequence of nodes and edges (a
sequence of adjacent nodes) that begins with i and ends with j. The length of the walk
is defned as the number of edges in the sequence.
112 Power Transmission System Analysis Against Faults and Attacks

Effciency of power grid can be expressed as the following equation [15]:

E=
1
N ( N −1) ˜ d1 ij
(4.4)
i˝ j

where N is the number of nodes and dij is the geodesic distance between nodes i and j.
The concept of distance dij may be explained as the diffculty to transfer the rel-
evant quantity between the nodes (i, j) of a network. Distance generally depends on
the path that one follows and should be defned as a function of the characteristics of
the lines along the path. The economic and technical diffculties for transfer of elec-
trical power through a path depend on both the power fow through the lines and on
their impedance. With the same impedance, higher power fow increases costs; with
the same power fow, higher impedance increases costs. Consequently, the distance
from node i to node j along path k is related not only to the impedance of each line of
the path but also to the power fows through the lines of the path.

4.2.5 ELECTRICAL DISTANCE

The electrical distance [16] considers characteristics of the lines along the path join-
ing buses. So, it can be defned as the equivalent impedance Zij, that is, the Thevenin
impedance between the two nodes i and j [17].

Zij = zii + z jj − 2zij (4.5)

where Z is the element of the bus impedance matrix. The electrical distance is used
instead of geodesic distance when considering an extended topological approach.

dij = Zij (4.6)

4.2.6 ELECTRICAL CENTRALITY AND ITS MEASUREMENT

The calculation of electrical centrality is entirely based on Zbus. Electrical centrality [18]
is computed using Zbus impedance matrix, which, in turn, can be either obtained from
Ybus matrix of the distribution system or by building up formulation of Zbus algorithm.
The latter is harder to implement but much more practical and faster for larger systems.
However, Kirchhoff’s laws are followed for the calculation of this measure. As the
physical topology of a distribution network may be either radial or ring type, the Ybus
matrix is sparse in nature. The impedance matrix Zbus gives the electrical topology
of the system by representing the strength of connection of every node to every other
node in the network electrically. In this chapter, the approach is to directly build
formulation of Zbus algorithm without using Ybus.
Considering only topological properties, geodesic distance is used to fnd path
impedance in complex networks. The weights of lines may be considered as in a
weighted graph or may not be considered as in an unweighted graph. It is obvious
that the smaller the geodesic distance between a pair of nodes is, the more eff-
ciently the energy transmits. Under normal operations, geodesic distance is replaced
Complex Network Theory 113

by electrical distance. Electrical distance considers characteristics of the lines along

the path joining buses.
Let us call Zij the equivalent impedance of the circuit whose ends are node i and node
j; Vij is the voltage between i and j and Ii is the current injected at node i and extracted at
node j (Ii = −Ij). As shown in Figure 4.1, the equivalent impedance is defned as

Vij
Zij = (4.7)
Ii

Furthermore, let Ii = 1, Ij = −1 and Ih = 0 ∀ h ≠ i, j (meaning that a unit current is

injected at node i and extracted at j, while no current is extracted nor injected in other
nodes), then the computation of equivalent impedance is as shown in Figure 4.1 and
amounts to
Vij
Zij = = Vij ˝ Zij = Vi − V j
Ii

= ( Zii − Zij ) − ( Zij − Z jj )

= Zii + Zij − 2Zij

where Zij is the i-th, j-th element of the impedance matrix.

The magnitude of the corresponding entry in the Zbus matrix represents the electri-
cal distance or geodesic distance between the nodes i and j. Hence, the smaller the
value of Zij, the shorter is the electrical distance or geodesic distance. From electrical
topology, some nodes are heavily connected to the rest of the network while others
are sparsely connected. The size of the node determines their relative importance in
the network with respect to their connectivity with the rest of the network. Larger the
nodes, higher are their connectivity with the remaining nodes. Electrical topology
N ( N −1)
consists of connections, where N is the number of buses or nodes; however,
2
N − 1 connections are actually shown to correspond with the physical topology,
which consists of exactly N − 1 links.

FIGURE 4.1 Equivalent impedance method.

114 Power Transmission System Analysis Against Faults and Attacks

4.2.7 BETWEENNESS
From the concept of complex network theory, it is evident that, if a vertex or edge
participates in more number of paths, the corresponding component becomes more
important for transmission in the entire network. Once it is assumed that the interac-
tions or transmission is always through the shortest paths between two vertices, it can
be used as an indication to justify the importance of a vertex or an edge in terms of
its betweenness.
Energy transfer between two nonadjacent buses depends on the buses and lines
of the geodesic paths connecting those buses. Thus, the vulnerability measure of a
network element can be determined by counting the number of geodesic paths going
through it, and is defned as betweenness centrality of that element.
Node betweenness of a node v can be expressed by the following formula [19]:

N N

B ( v ) =  ˜˜ °°(v)   i ˆ j ˆ v ˇ V
ij

ij
(4.8)
i j

where σij(v) is the number of geodesics from node i and node j through node v and
σij is the total number of geodesics between i and j. Similarly, the importance of
lines in the power grid can be determined by edge betweenness. Mathematically,
line betweenness centrality of a line can be obtained by the following formula [19]:

N N

B (l ) = ˜˜ °° (l ) i ˆ j ˇV , l ˇ E
ij

ij
(4.9)
i j

where σij(l) is the number of geodesics from node i to node j through line l. The
betweenness centrality is based on the topological structure of a complex network.
A component in any network has a specifc value of betweenness depending on
the confguration of the network. When these values are higher, it indicates that a
greater number of shortest paths pass through the component, highlighting the higher
criticality of the component. Thus, critical components of a network can be identifed
by ranking the betweenness value of the network components.

4.3 EXTENDED TOPOLOGICAL APPROACH

The research on application of complex network theory during its early stage dem-
onstrated some important concepts and measures that are suitable for various net-
work types. Nonetheless, the features and physical aspects of different networks are
naturally different from each other. Moreover, some particular features cannot be
included with the general methodologies. If the complex network theory is mechani-
cally applied to some felds without incorporating those particular features of these
networks, results obtained would diverge from reality. For application of the method-
ologies of complex network theory in power systems, it is important to consider the
electrical properties.
Complex Network Theory 115

To apply centrality indices in power system calculations, it is important to rede-

fne the corresponding metrics, particularly effciency. It is convenient to express
effciency for the electrical network in terms of netability. Effciency is redefned as
netability in which line fow limit on each line and electrical distance are introduced
into the effciency index. Similarly, betweenness centrality is redefned as electrical
betweenness by incorporating line fow limit on each line. Power transfer distribu-
tion factor (PTDF) is a matrix which is crucial for netability analysis as it refects
the line power fow sensitivity towards the change in the injected power of buses and
withdrawn power at a reference bus.
In this chapter, four basic features of a power network have been considered that
need to be extended using the topological approach. These features are incorporated
to defne the topological metrics into extended topological metrics.

• Bus Classifcation
To avoid diffculties involved in differentiation and dynamic behaviour
features of network components, all components have to be treated identi-
cally. Here, power fow is the characterising parameter which is consid-
ered to be transmitted from any vertex to any other. Generally, the buses
in power transmission networks are considered as generation, transmis-
sion and load buses. Transmission is normally considered from generation
to load buses.
• Line Flow Limits
In a pure topological approach, edges are generally described in an
unweighted manner to defne related metrics such as distance, degree and
betweenness. However, in power systems, line fow limits restrict the abil-
ity of power fow in a line due to congestion and economic load dispatch.
These factors are important to assess the network security of the power
system. These constraints need to be ascertained and applied for each line
specifcally as different lines may have distinct fow limits. To assess the
vulnerability of the power network, electrical performance parameters are
important to assess the network security.
• Flow-Based Network
Transmission of power between two buses (vertices) is always supposed
to be through the shortest path. This is the most unrealistic assumption from
the viewpoint of electrical engineering. Power transmission from a genera-
tor bus to a load bus involves numerous lines which can be designated as
paths having different extent of contribution. In simulation of power net-
works for a linear model of power fow, the various contributions of lines in
the power network can be described by the PTDF.
The conventional topological model in network theory describes the
graph as unweighted and undirected. The identifcation of possible paths
connecting two vertices is based on graph theory where transmission lines
are assumed to be bidirectional. Some paths in undirected model may be
not valid in the directed power transmission networks. PTDF character-
ises the behaviour of the system when the simulation between two vertices
completely depends on physical rules. Because each element in PTDF is
116 Power Transmission System Analysis Against Faults and Attacks

associated with a sign, the lines connecting to one vertex should be classi-
fed as input or output lines.
• Transmission capacity
To maintain stable and secure operation, the capability of transferring
power for each transmission line can be designated by its own transmission
l
limit   Pmax . In fact, for power transmission, not all lines reach their line fow
limit at the same time. Once the line attains its power transmission limit, the
power transmitted between corresponding pair of buses reaches its upper limit.

4.3.1 ELECTRICAL BETWEENNESS

In conventional topological approach, the concept of betweenness is defned as the
sum of the probability for a vertex or an edge to belong to a randomly selected geo-
desic path linking any other pair of vertices. Betweenness is a more useful tool for
measuring the magnitude of critical load on the given node/edge in the network. The
importance of betweenness also includes network connectivity. It can be considered
as a local metric to describe the degree centrality and corresponding criticality of
components (vertices and edges) in complex networks.
Including the parameters of power grid, the extended betweenness of a bus v can
be defned as [19]
NG N D

T (v) =
1
2 ˜˜C ˜ | f
g ˙G d ˙D
d
g
v
l ˙L
l
gd
| (4.10)

where v ˜ g ˜ d ° V

˜| f
l °Lv
l
gd
| is the sum of PTDF of all the lines connecting bus v when power is injected
at bus g and withdrawn at bus d.
1 d
2
Cg
l °L
˜
v
| fl gd | is the transmission power taken by bus v when power is transmitted
from generator bus g to load bus d; G is the set of generation buses, D is the set of
load buses, NG is the number of generation buses, ND is the number of load buses and
Lv is the set of lines connecting bus v.
Similarly, the extended betweenness of a line l can be defned as follows [66]:

T ( l ) = max ˝˙T p ( l ) ,  T n ( l ) ˆˇ   l   ˘ L (4.11)

where Tp and Tn are the positive extended line betweenness and negative extended
line betweenness of the line l.

T p (l ) = ˜˜C f g ˙ d , fl gd > 0
d gd
g l (4.12)
g ˇG d ˇ D

T n (l ) = ˜˜C f g ˙ d , fl gd < 0
d gd
g l (4.13)
g ˇG d ˇ D
Complex Network Theory 117

where Cgd fl gd represents the power fowing on the line l when power is transmitted
from g to d.
The concept of pure betweenness has been extended by introducing some elec-
trical properties. The set of extended betweenness qualifes the contribution of a
component to power transmission in the entire power grid, and in this respect, the
elements of the power grid can be ranked according to their criticality.

4.3.2 VULNERABILITY USING GLOBAL EFFICIENCY

A general measure for vulnerability assessment of power grid is the so-called eff-
ciency index, which can be used to describe the transmitting effciency of the power
grid. The transmitting effciency of the link connecting nodes i and j is denoted by
E, and the initial transmitting effciency is 1 for every link in the network. The eff-
ciency of a path between nodes i and j is the harmonic mean of effciencies of all the
links the path passed by. In all the paths connecting nodes i and j, the one with the
maximal effciency is called the most effcient path, and its effciency is denoted as
E. If there is no path between two nodes i and j, we have dij = 0.
As discussed in Equation (4.4), if there are N number of nodes and dij is the geo-
desic distance between nodes i and j, then the global effciency of the network is
expressed as [15]

E =
1
N ( N −1) ˜ d1 ij
(4.14)
iˆ j

Here, the sum is taken over all pairs of nodes of the unweighted and undirected
network.
Vulnerability of a line in the network is a measure of drop in network functioning
due to removal of that line in the system [14]. So, the damage of the line can be mea-
sured by the relative decrement in network performance. If E is the global effciency
of the network without any damage to the lines and E(l) is the global effciency of
the network when line l is removed, then vulnerability of that line l can be calculated
as [14]

E − E (l )
Ve (l ) = (4.15)
E
The maximum of the vulnerabilities is vulnerability of the network;

V = max [Ve ( l )] m (4.16)

Though the above method can be applied to power grid, some problems arise due to
distinguished characteristics of power grid network as follows:

a) Geodesic distance considers only the shortest path between the nodes, but
power can fow through any path between a pair of nodes.
118 Power Transmission System Analysis Against Faults and Attacks

b) There is no need to consider all pairs of nodes as power fows only from
generation nodes to load nodes.
c) The transfer capabilities in transmitting power of a network are different
which has not been taken into consideration in global effciency calculation.

4.3.3 VULNERABILITY USING NETABILITY

Including general operating conditions such as power fow limit, transmission
capacity and PTDF, the performance of a power grid can be evaluated using
netability.
Power transmission capability Cij is defned as the power injection at bus i when
the frst line of all the paths connecting the generation node i and the load node j
reaches its limit. It is expressed as [19]

° l ˙
P ˇ
Cij = min l ˘ L ˝ max (4.17)
˝ fijl ˇ
˛ ˆ

l
where Pmax is the transmission limit of transmission line l and fijl is the PTDF of line
l of the path joining generation node i to load node j. This is the change of power on
line l for injection at generation bus i and withdrawal at load bus j. fijl is obtained as
the difference between the entries fli and flj of the PTDF matrix is calculated as [16]

A = H / B / −1 (4.18)

where B = N × N admission matrix

1
Bij = ,i ° j (4.19)
xij

Bii = ˜ x1 ij
(4.20)
j ˛i

To avoid singularity, slack bus column and row are eliminated.

B = Sub-matrix of B where slack bus column and row are eliminated from B.
H = Transmission matrix of order L × N

1
Hli = −Hlj = (4.21)
xij

Hlk = 0 °k ˛ i, j (4.22)

L = Number of lines in the network joining the nodes.

N = Number of nodes in the network
H = Sub-matrix of H where slack bus column is eliminated from H.
Complex Network Theory 119

Considering normal operation criteria, geodesic distance is replaced by electri-

cal distance in netability model. Electrical distance considers characteristics of the
lines along the path joining buses. So electrical distance is defined as the equivalent
impedance Zij is the Thevenin impedance between the two nodes i and j as defined
earlier Zij = Zii + Zij − 2Zij where Zij is the i, jth element of the bus impedance matrix.
Hence, netability of a power grid network is expressed as

A=
1
NG N D ∑∑ CZ (4.23)ij

ij
i ∈G j ∈ D

where G = Set of generation buses

D = Set of load buses
NG = Number of generation buses
ND = Number of load buses
Line resistances are ignored.
Similarly, as in previous model, the vulnerability of line l is defined as the drop in
netability of the network due to removal of the line l, that is,

A − A (l )
Va (l ) = (4.24)
A
where A(l) is the netability of the network when line l is removed.

4.3.4 Criticality Assessment Using [Zbus] Centrality

A metric of electrical centrality that can be useful for installation of DG in sub-
transmission system from the complex network theory viewpoint. It was proposed to
differentiate the electrical topology and physical topology of power grid. This metric
not only enables complex network analysis of power systems but is also more appro-
priate for the power grid than general topological analysis. This measure is based
on Zbus impedance matrix of a power system which finds more electrically central
nodes in the system which is supposed to be highly connected to most other nodes
for the placement of distributed generators in the sub-transmission system. Correct
placement of DG in the transmission system is of strategic importance to improve
the robustness of the grid. Incorrect placement may make the grid more vulnerable
to failures or attacks.
The measure of electrical centrality (electrical node significance) is utilised to
locate nodes where DGs can be placed. The size of the DGs depends on the relative
importance of the nodes. The first step of the procedure is to calculate electrical cen-
trality and significance of every node in the network. This calculation gives a clear
indication of which nodes are more important in the network and offers a candidate set
of locations where DGs can be placed. For all the test networks, most nodes indicated
by electrical centrality differed from those indicated by electrical node significance.
This shows that different methods can indicate multiple possibilities of locations for
DG placement. Individually, they can produce slightly satisfactory results, but with a
combination of these methods, the quality of results can be enhanced.
120 Power Transmission System Analysis Against Faults and Attacks

Electrical centrality is a measure used in topological analysis of power grid

networks, which differentiates electrical structure of the grid from its topological
structure. Electrical centrality uses the impedance matrix, or the Z bus matrix of
the transmission system, to determine which nodes are more electrically central to
the system and indicates them as candidate locations for the placement of DGs. The
electrical topology of the grid indicates that power grids possess “electrical hubs,”
indicating that some nodes in the power grid have strong electrical connections with
other parts of the network. This phenomenon is very different from the physical
topology of the power grids as their average degree is usually between 2 and 5,
indicating the absence of hubs. As mentioned in Section 4.2.6, electrical centrality
is calculated using the Z bus matrix which is computed as the inverse of the Y bus
matrix or the admittance matrix of the system. The Y bus matrix is usually sparse,
and, hence, the Z bus matrix is obtained as a dense matrix. Every element in the
matrix represents an equivalent electrical distance between two nodes. Since the Z
N ( N −1)
bus matrix is a non-sparse, dense matrix, there are different electrical con-
2
nections possible.

REFERENCES
1. P. Erdos and A. Renyi, “On the evolution of random graphs”, Publications of the
Mathematical Institute of the Hungarian Academy of Sciences, vol. 5, pp. 17–61, 1960.
2. D.J. Watts and S.H. Strogatz, “Collective dynamics of ‘small world’ networks”, Nature,
vol. 393, pp. 440–442, 1998.
3. A.L. Barabasi and R. Albert, “Emergence of scaling in random networks”, Science, vol.
286, pp. 509–512, 1999.
4. R. Albert, H. Jeong, and A.L. Barabási, “Error and attack tolerance of complex net-
works”, Nature, vol. 406, pp. 378–382, 2000.
5. A.E. Motter and Y.C. Lai, “Cascade-based attacks on complex networks”, Physical
Review E, vol. 66, p. 065102, 2002.
6. P. Crucitti, V. Latora, and M. Marchiori, “Model for cascading failures in complexnet-
works”, Physical Review E, vol. 69, p. 045104, 2004.
7. R. Albert, I. Albert, and G.L. Nakarado, “Structural vulnerability of the North
American power grid”, Physical Review E, vol. 69, p. 025103, 2004.
8. V. Rosato, S. Bologna, and F. Tiriticco, “Topological properties of high-voltage elec-
trical transmission networks”, Electric Power Systems Research, vol. 77, pp. 99–105,
2007.
9. M. Rosas-Casals, S. Valverde, and R. Sol_e, “Topological vulnerability of the European
power grid under errors and attacks”, International Journal of Bifurcation and Chaos,
vol. 17, no. 7, pp. 2465–2475, 2007.
10. R. Kinney, P. Crucitti, R. Albert, and V. Latora, “Modelling cascading failures in North
American power grid”, European Physical Journal B, vol. 46, pp. 101–107, 2005.
11. P. Crucitti, V. Latora, and M. Marchiori. “A topological analysis of the Italian electric
power grid”, Physica A, vol. 338, pp. 92–97, 2004.
12. P. Crucitti, V. Latora, and M. Marchiori, “Locating critical lines in high-voltage electri-
cal power grids”, Fluctuation and Noise Letters, vol. 5, pp. L201–L208, 2005.
13. M.E.J. Newman, “The structure and function of complex network”, SIAM Review, vol.
45, pp. 167–256, 2003.
14. V. Latora and M. Marchiori, “Vulnerability and protection of infrastructure networks”,
Physical Review E, vol. 71, p. 15103, 2005.
Complex Network Theory 121

15. V. Latora and M. Marchiori, “Effcient behaviour of small-world networks”, Physical

Review Letter, vol. 87, p. 198701, 2001.
16. S. Arianos, E. Bompard, A. Carbone, and F. Xue, “Power grid vulnerability: A complex
network approach”, Chaos, vol. 19, pp. 1–6, 2009.
17. J.J. Grainger and W.D. Stevenson, Jr., Power System Analysis, India: Tata McGraw-
Hill, pp. 287–292, 1994.
18. S. Pahwa, D. Weerasinghe, C. Scoglio, and R. Miller, “A complex networks approach
for sizing and siting of distributed generators in the distribution system”, 45th North
American Power Symposium, Manhattan KA, September 2013.
19. E. Bompard, D. Wu, and F. Xue, Structural vulnerability of power systems: A topologi-
cal approach. Electric Power Systems Research, vol. 81, no. 7, pp. 1334–1340, 2011.
 5 Vulnerability
Assessment of Power
Transmission Network

5.1 INTRODUCTION
Power generation, transmission and distribution, as an integrated system, is one of
the critical infrastructures as it is widely distributed in a vast geographical region and
indispensable to the modern world. Accidental failures and intentional attacks can
lead to disastrous social and economic consequences. Therefore, to ensure reliable
performance, electrical operation engineers need to assess the vulnerability of power
grid networks and identify the critical elements that need essential back-up protec-
tion to ensure a more robust electrical system against natural or malicious threats.
The vastness, size of components and the complex dynamics of various compo-
nents of power grids make them typical complex networks. Numerous researches
including basic characteristics and steady state and transient state performance have
been analysed since long, and in the present scenario of planned and unplanned
attacks on civilisation and society have raised the issue of vulnerability of a power
grid. Interestingly, the close relationship between topological structure and opera-
tion performance in power systems as a change in topological structure can alter the
operational condition of a power system leading to change in its operational perfor-
mance. Hence, there is an increasing interest in topological vulnerability assessment
of power grid networks using complex network methodology.
In this chapter attempts have been made to explore the applicability of complex
network theory in power distribution and transmission systems in assessment of vul-
nerability of any electrical power network.
In general, a vulnerable system operates with a “reduced level of security that
renders it vulnerable to the cumulative effects of a series of moderate disturbances”
[1]. From the viewpoint of power system operation, vulnerability is an estimate of the
fragility of the system following a sequence of occurrences which might involve line
or generator failure, disruptions, interruptions, blackout, breakdown or undesirable
operations of protective relays, information or communication system malfunction-
ing and human errors.
Vulnerability assessment includes identifcation and quantifcation following the
prioritising (or ranking) sequence while the system is under the threat of planned or
unplanned contingencies or attacks [2]. In conventional methods of security assess-
ment and contingency analysis, power system engineers use conventional mathemati-
cal tools with operational data and physical models of the power system components.
Such an analysis is valid under a given contingency and operating condition. It may
not be computationally feasible to simulate all possible combinations of contingencies
123
124 Power Transmission System Analysis Against Faults and Attacks

that could result in partial or complete collapse of the power network as well as events
of various switching actions and continuous load variations. Hence, it is diffcult to
assess the outage of a portion or the entire grid following unforeseen operating con-
ditions. Moreover, the expansion of the bulk power systems, their physical behaviour
and interrelationship intensifes the problem of comprehensive vulnerability study.
Hence, to reduce and simplify these problems, it is important to deepen and contrib-
ute to traditional analysis with novel tools of vulnerability assessment.
The motivation to assess the vulnerability of grid networks employing complex
network analysis is because of cascading failures of power grids globally following
contingencies and other reasons. The North American power grid was the frst ana-
lysed power grid [3] where the vertices were randomly removed, and in decreasing
order of their degrees (of those vertices) the connectivity loss was monitored. The
connectivity loss among various areas of the grid resulted in loss of potential of
substations to draw power from the generators. The depletion of generation nodes
(substations) was not effective in changing the average degree (connectivity) of the
grid due to high redundancy at the generation nodes (substations). As the power grid
is delicate towards the depletion of transmission nodes, the elimination of a single
transmission line can lead to connectivity loss. The analytical methods in complex
network theory revealed that the connectivity loss is substantially higher following
planned attacks, particularly if the degree of attack is higher for high betweenness
transmission hubs. Crucitti et al. [4] made the frst reference to European power grids
and compared the structural properties of Italian, French and Spanish power grids
by identifying the elements whose removal genuinely affected the structure of those
graphs. Rosato et al. [5] implemented high-voltage vulnerability assessment of power
grid networks in Italy, France and Spain to analyse the damage caused by controlled
removal of links. Thus, vulnerability assessments using topological methods are use-
ful in assessing the weakness of an electrical network as well as to design distinct
actions to minimise topological weaknesses.

5.2 VULNERABILITY INDICES: TOOLS OF

VULNERABILITY ASSESSMENT
Several indices like global effciency, electrical betweenness, and netability, defned
in Chapter 4 can be used to identify and rank critical nodes (buses) and links (lines)
in a power network. Indices of vulnerability can include electrical centrality, elec-
trical betweenness, vulnerability assessment using netability and global effciency.
While global effciency and netability are both indices representing vulnerability of
the system when any network component (bus or line) loss occurs, betweenness is
a measure of criticality of each of these components in a network. The reactance of
the lines act as the weights of the edges of the graph along the shortest electric path,
which is the path whose sum of weights is the smallest among all possible paths
between the nodes. The complex power fow through any line or tie in a complex
grid is governed by the line parameters as well as by bus voltages and power angles.
Evidently, the magnitude of power transfer via any transmission line from the send-
ing bus to the receiving bus is inversely proportional to the reactance of the line when
the bus voltages are at nominal p.u. values. In this chapter, the line reactance has
Vulnerability Assessment of Power Transmission Network 125

been assumed to be the weight of the transmission lines and it is presumed that the
magnitude of power fow increases when the line reactance is low.
Some researchers [6] have employed metrics, such as degree and degree distribu-
tion, characteristic path length, clustering coeffcient and betweenness, to determine
how the relative connection of distributed generators (DGs) infuences the topological
structure of the grid network. Some researchers utilised weighted graph indices and
proposed advanced metrics based on the topology and condition of practical working
operation of the power networks for the measure of topological characteristics of the
power networks employing DGs. Authors of [7] performed similar assessment on the
basis of specifc distinct metrics where they introduced three vulnerability indices:
Structural Vulnerability Index (SVI), Contingency Vulnerability Index (CVI) and
Operational Vulnerability Index (OVI). These indices were exploited to measure
structural vulnerability to identify the vulnerable elements in the power network
and assess the state of operation of the network. In this chapter, a new metric called
Grid Vulnerability Index (GVI), which is a combination of the SVI, CVI and OVI, is
calculated to assess the structural change in a transmission system due to incorpora-
tion of DG.

5.2.1 GRID VULNERABILITY INDEX (GVI)

GVI is a new index that assesses the overall functioning capability of a grid when the
grid is equipped with various DGs fulflling the load demand. Because line reactance
is governed by the length of the line, the concept of electrical distance is introduced
as the performance parameters of the power grid. The GVI can be expressed as [8]

GVI =
1
NG N D ˜˜ L Pe j
i
Zij (5.1)
i ˛G j ˛ D

where NG and ND denote number of generation buses and number of load buses,
respectively. Pi and Lj are the active power generation and maximum load of buses,
respectively. Here, the term Zij would be zero if the DG is directly connected to the
load bus. On the other hand, for conventional generators supplying power through
transmission lines, the value of Zij would be higher. In this case, the attribution of
generation bus to load bus reduces with the increase in the value of Zij.
To simulate GVI for a power network, the following assumptions are considered:

a) For every alteration in the allocation of DG in the power system, NG varies.

It rises with the addition of the DGs to the load buses.
b) The quantity of load buses is assumed to be constant despite being included
with DGs.
c) DGs are allocated to load buses only as they are troublesome to synchronise
with traditional generation buses.
d) The capacity of DG has been assumed to be half the capacity of peak load at
load bus. The highest DG capacity is 50 MW, although few load nodes have
more half of peak load than this.
126 Power Transmission System Analysis Against Faults and Attacks

5.2.2 EFFICIENCY-BASED VULNERABILITY INDEX

In general, the rating of a distributed energy resource (DER)-based generator
(i.e.  DGs) is typically small. Such types of resources are installed at load nodes in
remote or isolated geographical area. The load nodes that are distant from these gen-
erators are minimally facilitated. Moreover, the capability of a DG drastically reduces
as the distance between it and a load node increases. Thus, the index effciency does
not apply in analysis of the network vulnerability when the network is equipped with
DG connected at remote load nodes. With respect to this, a novel index (metric) with
the assumption of the local supplying characteristic of DG is proposed:

ei =
1 1
PD j N g ˜ 2P Gi
dij −1 (5.2)
i ˛VG

where ei represents the effciency of supplying power to a load node j. PD j is the active
load demand of node j, PGi is the power capacity of generator node i and dij is the
shortest path between generation node i and load node j. Ng is the number of DGs
in the network. The bigger the value of ei is, the more will be the power supplying
effciency of node i.
Consequently, the Global Average Power Supplying Effciency Index can be
expressed as [8]:

ei =
1
Ng ˜e = N1 i
g
1
ND ˜ PGi
2dij−1
(5.3)
i ˛VG j ˛VD i ˛VG

The exponential function in Equation (5.3) describes the characteristic of power sup-
plying effciency of the DG which decreases sharply with the increase in the trans-
mission distance.
Following any decrement in the value of e, the vulnerability of the network can be
assessed using the following expression:

˝ eˇ
e˜ = ˆ1 −  ×100% (5.4)
˙ e0 ˘

Here, e0 denotes the power supplying effciency before the failure and e denotes the
power supplying effciency after the failure. From Equation (5.4), it is evident that the
bigger e’, more will be the drop in the functioning of the network, and the affected
node (bus) or link (line) will be more prone towards vulnerability.

5.3 ASPECTS OF VULNERABILITY ANALYSIS OF

POWER TRANSMISSION NETWORK
Identifying the vulnerable components in a power grid [9] is vital to the design and
operation of a secure and stable system. One aspect of vulnerability analysis is to
identify those transmission lines for which minor perturbations in their conductive
properties lead to major disruptions to the grid, such as voltage drops, or the need
Vulnerability Assessment of Power Transmission Network 127

for load shedding at demand nodes to restore feasible operation. Most studies focus
on minimising the costs of load shedding and additional generation in the DC model
or in lossless AC models. In case of identifcation of critical components of a power
system, the grid has to be modelled with full AC power fow equations, which are the
most accurate mathematical models of power fow.
A second aspect to measure severity of an attack is to consider the minimum
adjustments to power that must be made to restore the grid to feasible operation.
Power adjustments take the form of shedding load at demand nodes and adjusting
generation at generator nodes. This is explained in the Chapter 6 in the islanding pro-
cedure for modelling cascading failure in power transmission grids where load redis-
tribution model is followed using preferential probability. The next section discusses
the application of vulnerability indices in power transmission network by simulation.

5.4 SIMULATION FOR VULNERABILITY INDICES

IN POWER TRANSMISSION NETWORK
5.4.1 USING GLOBAL EFFICIENCY AND NETABILITY METHOD
Figure 5.1 shows IEEE 14 bus system [10,11] whose vulnerability is being assessed
by obtaining the global effciency and netability values of the lines of this network.

FIGURE 5.1 IEEE 14 bus system.

128 Power Transmission System Analysis Against Faults and Attacks

TABLE 5.1
Line Parameters of IEEE 14 Bus System
Line No. Connecting Nodes Line Reactance Line Flow Limits
01 1–2 0.05917 130
02 1–5 0.22304 130
03 2–3 0.19797 65
04 2–4 0.17632 130
05 2–5 0.17388 130
06 3–4 0.17103 65
07 4–5 0.04211 90
08 4–7 0.20912 70
09 4–9 0.55618 130
10 5–6 0.25202 32
12 6–11 0.19890 65
12 6–12 0.25581 32
13 6–13 0.13027 65
14 7–8 0.17615 65
15 7–9 0.11001 65
16 9–10 0.08450 65
17 9–14 0.27038 32
18 10–11 0.19207 32
19 12–13 0.19988 32
20 13–14 0.34802 16

The line reactance is considered as weights of the edges of the graph. The line
parameters that are mainly considered are shown in Table 5.1:

Sample Calculation:
Global effciency (Ve): From Equations (4.14) and (4.15), the global effciency can
be calculated as follows:
For the given network

˜ d1 = 46.31667, E = 0.254487
ij
i˛ j

For line no. 1, ˜ d1 = 45.65, E = 0.250842

ij
1
i˛ j

Global effciency for removal of line 1 is then given by

E − E1
Ve1 = = 1.43%
E
Netability (Va): From Equations (4.23) and (4.24), the netability can be calculated.
Vulnerability Assessment of Power Transmission Network 129

In the same network,

  ˜˜ CZ ij

ij
= 18218.0365, A = 506.0565
i =G j = D

For line no. 1, A1 = 459.7372

Netability for removal of line 1 is, thus, given by

A − A1
Va1 = = 9.15%
A
Following the above method of calculation, the values of global effciency and
netability for the given network are obtained and shown in Table 5.2.
The line vulnerabilities are calculated using Equations (4.15) and (4.24) and the
comparable graphs are resulted as shown in Figure 5.2:
It is observed that netability metric that includes the electrical properties of
extended topological approach gives enhanced results of vulnerability analysis. The
critical lines can be identifed as listed below in Table 5.3.

TABLE 5.2
Calculated Values of Global Effciency
and Netability
Line No. Connecting Nodes Ve (%) Va (%)
01 1–2 1.43 09.15
02 1–5 2.73 08.44
03 2–3 2.15 17.57
04 2–4 3.6 14.14
05 2–5 2.08 05.47
06 3–4 3.63 19.41
07 4–5 8.44 18.96
08 4–7 8.5 17.17
09 4–9 1.76 11.58
10 5–6 8.2 01.73
12 6–11 9.31 18.86
12 6–12 4.06 05.24
13 6–13 7.12 09.14
14 7–8 7.19 29.6
15 7–9 8.25 17.8
16 9–10 0.6 16.45
17 9–14 4.1 08.93
18 10–11 4.46 14.03
19 12–13 1.43 05.02
20 13–14 5.75 12.04

Bold values are having the highest value, it is the most

vulnerable line
130 Power Transmission System Analysis Against Faults and Attacks

FIGURE 5.2 Graph of vulnerability analysis of lines using global effciency versus netabil-
ity of IEEE 14 bus power system.

TABLE 5.3
Critical Lines Identifed from Graph
Line No. Connecting Nodes Vulnerability in (%)
3 2–3 17.57
6 3–4 19.41
7 4–5 18.96
8 4–7 17.17
11 6–11 18.86
14 7–8 29.60

From the above, it is noticed that line number 11 is highly vulnerable in the given
network which connects a generation bus to a transformer node. One of critical lines,
that is, line no. 3 connects two generation buses. Other critical lines connect trans-
formers and buses. For obvious reasons, these lines are vulnerable as they are associ-
ated with critical components such as transformers and generation buses of the power
grid network.
The grid network has been analysed separately using only topological model
(global effciency) and by adding some power grid characteristics to the topological
model (netability). The IEEE 14 bus system is studied and the critical lines are iden-
tifed which justifes the applicability of the developed model, that is, the netability
model. The analysis identifes critical lines in the power grid which is very important
in maintaining the network security.

5.4.2 USING BETWEENNESS AND NETABILITY METRIC

The given network has been altered to obtain bus impedance matrix directly. Bus num-
ber 1 is considered as a slack bus in case of electrical betweenness and is numbered as
Vulnerability Assessment of Power Transmission Network 131

the 0th node. According to [12], the line parameters and fgure are shown in Appendix.
On the basis of pure topological betweenness or betweenness centrality (Equations 4.8
and 4.9) and extended electrical betweenness (Equations 4.10 and 4.11), the simulation
has been conducted on the test system to fnd line betweenness and node betweenness
with respect to respective line and node numbers (Figures 5.3 and 5.4a–d).
From the above analysis, it has been observed that the lines exhibit different
betweenness among themselves. Figure 5.4a reveals that line 70 has the largest value
of betweenness. Other lines like line number 42, 71, etc. follow line 70 in descending
order of magnitude of betweenness. Because betweenness interprets the criticality of
an element connected between two nodes, it is obvious that ranking lines in the order
of criticality is possible using this method of analysis.

FIGURE 5.3 IEEE 57 bus system.

132 Power Transmission System Analysis Against Faults and Attacks

(a) (b)

(c) (d)

FIGURE 5.4 (a) Line betweenness of IEEE57 bus system, (b) extended line betweenness
of IEEE57 bus system, (c) node betweenness of IEEE57 bus system and (d) extended node
betweenness of IEEE57 bus system.

In the next step, the concept of extended betweenness has been considered. This
extended betweenness indicates the criticality taking into account constraints like
line fow limits, maximum transmission capability, and power transmission distri-
bution factor. This concept is more realistic, and applying this concept in the test
system, the simulation reveals that line number 79 showed the highest magnitude
of extended betweenness followed by line number 25, 14, 27 and 21 (Figure 5.4b).
Hence, the extended betweenness exhibits more specifc criticality of the lines. The
same exercise has been conducted for nodes (buses) revealing that bus number 12
exhibits the highest degree of extended betweenness, that is, criticality.
In the next phase, analysis and simulation has been conducted on the same test
system to determine the netability of the system. The netability exhibits the degree
of vulnerability of the power network following contingencies. While observing the
results (shown in Figure 5.5), it is clear that despite line number 79 displaying high
betweenness under normal operating conditions, failure of line number 14 indicates
the drop of netability of 14.37% compared to other lines (case of single line outage).
In this context, line number 14 has been taken into account as a specifc case of
Vulnerability Assessment of Power Transmission Network 133

(a)

(b)

(c)

FIGURE 5.5 (a) Single-line outages of IEEE 57 bus system, (b) double-line outages of IEEE
57 bus system and (c) triple-line outages of IEEE 57 bus system.
134 Power Transmission System Analysis Against Faults and Attacks

interest. Thus, in multiple contingency studies, line number 14 is considered with

other lines; as in case of two outages (line number 79 and 14) and in case of three
outages (line number 79, 14 and 25). It reveals that the netability magnitudes remain
high indicating higher vulnerability of the system line 14 is associated in contingency
cases. Therefore, with undivided attention of operating personnel, the network can be
arrested from catastrophic failure if line number 14 is checked on priority.
Hence, complex network theory can be successfully applied in vulnerability analy-
sis of a sub-transmission system in conjunction with netability analysis. It is observed
that computation of netability of the network facilitates ranking of critical lines and can
be utilised in real power grid system. IEEE 57 bus system as a complex network is ana-
lysed in both topological and electrical context separately on the standard. The number
of vulnerable lines has been obtained which reveals the practicality of the preferred
model (i.e. the network modelled with extended betweenness). It may be suggested that
topological analysis including electrical properties of the power network could serve
as a complementary tool that can be used in decision-making by the operating person-
nel. Thus, application of complex network theory in power network can suitably be
employed for identifying critical lines and buses in a sub-transmission system.

REFERENCES
1. L.H. Fink and K. Carlsen, “Operating under stress and strain”, IEEE Spectrum, vol. 20,
pp. 1357–1365, 2005.
2. http://en.wikipedia.org/wiki/Vulnerability_assessment.
3. R. Albert, I. Albert, and G.L. Nakarado, “Structural vulnerability of the North
American power grid”, Physical Review E, vol. 69, 2004.
4. P. Crucitti, V. Latora, and M. Marchiori, “Locating critical lines in high-voltage electri-
cal power grids”, Fluctuation and Noise Letters, vol. 5, pp. L201–L208, 2005.
5. V. Rosato, S. Bologna, and F. Tiriticco, “Topological properties of high-voltage elec-
trical transmission networks”, Electric Power Systems Research, vol. 77, pp. 99–105,
2007.
6. L. Feng, M. Yanbin, and M. Shengwei, “On the topological characteristics of power
grids with distributed generation”, In the Proceedings of the 29th Chinese Control
Conference, Bejing, China, pp. 4714–4720, July 2010.
7. Z. Chen, C. Liu, Q. Xu, and C.L. Bak, “Vulnerability evaluation of power system
integrated with large scale distributed generation based on complex network the-
ory”, In Proceedings of the 47th IEEE International Universities Power Engineering
Conference, Middlesex, U.K, pp. 1–5, September 2012.
8. T. Chowdhury, A. Chakrabarti, and C.K. Chanda, “Analysis of vulnerability indices
of power grid integrated DG units based on complex network theory”, In Proceedings
of the 2015 Annual IEEE India Conference (INDICON), New Delhi, pp. 6540–6546,
December 2015.
9. T. Kim, et al., “Vulnerability analysis of power systems”, 2015.
10. http://www.ee.washington.edu/research/pstca/pf14/pg_tca14bus.htm.
11. H.B. Plittgen, “Computational cycle time evalution for steady state power fow calcula-
tions”, December 1985.
12. http://www.ee.washington.edu/research/pstca/.
 6 Analysis of Cascading
Failure and Islanding
in Grid Network

6.1 INTRODUCTION
In most countries power grid is a typical complex network that is distributed
throughout a vast geographical region. Seldom, grid operation poses a threat of fail-
ure because of localised faults, consequently, affecting a part or the entire network
leading to vulnerable collapse of the grid network. Such a failure can affect a large
number of customers resulting in significant economic disruptions. Although most
electrical failures of line(s) and or buses emerge and dissolve locally, some trigger
avalanche mechanisms that can lead to failure of lines and buses as a chain event and
eventually cripple the entire network [1–3]. Such vulnerable failure of an electric grid
that leads to islanding of the grid network is termed as cascading failure. Literature
review reveals that a number of power grids across the world have suffered cascading
failures [1–5] causing blackouts. Cascading failure can achieve massive scale, with
millions of customers affected by loss of billions of dollars [6,7].
In mathematics a popular model, called the “sand pile” [8] model, states that when
sand is continuously piled in a heap, a point arises when a portion of the pile suddenly
begins to subside. If an attempt is made to prevent the breakdown by adding more
sand, the pile just collapses. Similarly, we can say that as a power grid approaches
a critical point, the possibility of a collapse increases. With the addition of new ele-
ments to the grid, increasing demand, interconnections and increase in power fows,
the grid may be quickly approaching its critical point. At such a point, it becomes
necessary to study the complex dynamics arising in the grid and to fnd possible solu-
tions to problems that may arise because of these dynamics.
Cascading failure [2] is common in power transmission grids when one of the ele-
ments fails (completely or partially) and shifts its load to nearby elements in the system.
Those nearby elements are then pushed beyond their capacity and they become over-
loaded and shift their load onto other elements.

The probable causes of cascading failure mainly involve:

• Severe overload
• Resource exhaustion
• Defective load management
• Inaccurate capacity planning
• Delay in communication response
• Sudden load burdening

135
136 Power Transmission System Analysis Against Faults and Attacks

The concerning issues associated with cascading failures are:

• They can take down the entire power system, tumbling the service of each
component of the system one by one, until the entire load-balanced service
is unhealthy.
• They are an exceptional type of failure from which it is hard to recover.
They normally start with some small perturbation – like a transient net-
work issue, a small spike in load or the failure of a few instances. Instead
of recovering to a normal state over time, the system enters into a worse
state. A system in cascading failure won’t self-heal; it can only be restored
through human intervention.
• If the right conditions exist in the system, cascading failures can strike with
no warning. Unfortunately, the basic pre-conditions for cascading failures
are diffcult to avoid: it’s simply failover. If failure of a component can cause
load to shift to other parts of the system, then the basic conditions for cas-
cading failure are present. However, there are patterns we can apply that
help us defend our systems against cascading failures.

Among the above-stated probable causes of cascading failures, two scenarios are
commonly observed [8] to be frequent. These are load growth and random fuctua-
tions or perturbations.
Load growth is the increase in load in a specifc ratio either intentionally or due to
sudden high current operation.
Random perturbations can occur in the grid for several reasons. If random fuc-
tuations are small, their effect may not be felt. However, if the fuctuations are con-
siderable in magnitude, they can cause several undesirable effects on the functioning
of the grid, including the initiation of a cascading failure. For example, random per-
turbations can occur due to the incorporation of renewable distributed generators,
such as wind turbines and solar panels, in the grid.
The complexity of the entire power grid makes it diffcult to model each and every
individual component and study the stability of the entire system. Therefore, it is
often the case that abstract models of the working of the power grid are constructed
and analysed. The lack of a complex network-based model for the analysis of power
grid was a major drawback of the studies conducted earlier. However, Overload
Cascade model [9] was introduced to the complex network approach-based model.
This model assigns capacity to the links in the network proportional to the initial
power owing through them in the absence of any disturbance. When a disturbance
occurs, these power fows are redistributed and the fow dynamics of the network are
observed. The scenarios on which grid dynamics are observed include load growth
and random fuctuations or perturbations.
In this model, the initial load and generation on the nodes represent the initial
state of the system. Power fows are calculated using the DC Power Flow model.
Every link has a capacity which determines the maximum amount of power that can
be carried by that link. During normal operation, the system is stable and all power
fows are within limit. If a disturbance, such as failure of an element, sudden increase
in demand or load fuctuations, occurs in the system, all power fows are recalculated
Cascading Failure and Islanding in Grid Network 137

using equations and utilisation of all links is checked to see whether they are within
their capacity. If any link exceeds its capacity, it is removed from the system, and
its power is redistributed among other links, depending on impedances of the other
links. Even though the model is simple, the presence of the fow equations guarantees
that Kirchhoff’s and Ohm’s laws are properly considered, thus making the model
more realistic for use with the power grid. The reduced complexity of this method
allows its use even with large systems and provides a reasonable balance between
complexity of the method and accuracy of the results.
In load growth scenario, the load growth is modelled by increasing all loads simul-
taneously step-by-step by a factor. Each time the loads are increased, the Overload
Cascade model is run to calculate the power in the network. The increase in the
demand causes increased power fows through the links. Because each link is bound
by a fnite capacity, some of the links exceed their capacity and get overloaded to sup-
ply the increased demand. These overloaded links fail and the power that was being
carried through them is redistributed among the other links in the network. Because
of this redistribution, more links may reach their capacity and fail. This may lead
to a cascade of overload failures. This simulation can be performed with different
values of the factor, and the fraction of links that failed at the end of each simulation
can be recorded.
If a transmission line within the power grid trips, its power fow is automati-
cally shifted to the neighbouring line(s), which in most of the cases are capable
of operating with the extra load. However, in some cases, one or more of these
neighbouring lines may get overloaded and exceed the allowable limit of power
fow leading to overloading of line(s). This may lead to outage of the neighbouring
overloaded line(s) which phenomenon may spread over other line(s) suffering from
overloading. This process of failure of multiple lines as a chain event is a blow to
the contingent power grid which is likely to collapse while encountering multiple
line tripping. To better understand such events, the power grid needs to be analysed
from a network perspective taking the advantage of concepts in complex network
theory [10].

6.1.1 FUNDAMENTAL MODELLING APPROACH IN ISLANDING

A grid network can be modelled as a weighted graph having nodes or vertices (the
buses) and links or edges (the transmission lines) [11]. When there is no direct trans-
mission line from the node i to node j, the off-nominal entry (aij) of such a matrix is
zero, otherwise aij is 1 provided a line exists between node i and node j. In a realistic
approach of a power grid network model, each generator is assumed to transfer power
to all buses (nodes) through transmission lines (links).
Following established concepts on complex network theory [12,13], the electrical
load on a load bus can be defned as the electrical betweenness of that bus (node)
associated with transmitting capabilities with number of most effcient paths passing
from generators to load buses through the particular node. The same concept of elec-
trical betweenness is valid for a transmission line where the line can be associated
with an ultimate power transfer capacity Cij that is proportional to its loading in the
unperturbed state of the network.
138 Power Transmission System Analysis Against Faults and Attacks

Due to external causes when tripping of a line or bus connected to transmission

line(s) occurs, the most effcient paths through which electrical power from each
generator can reach each bus changes and the load gets redistributed among the net-
work. Thus, a new value of electrical betweenness emerges for the surviving adjacent
lines. The increment in load of the adjacent line(s) would raise the value of electrical
betweenness for the line, and because the capacity Cij of the line is characterised by
the maximum load that the line can handle, the surviving adjacent line(s) would face
outage provided the maximum value of the betweenness exceeds the ultimate capac-
ity Cij of the concerned line.

6.2 CONCEPT OF ISLANDING

Due to massive expansion of power system, large blocks of power are transmitted
through long-distance EHV transmission lines from source to load centre in most
power grids. Because of complex network structure, the grid suffers from possi-
bility of lesser stability and is prone to encounter different types of contingencies
and attacks that may lead to failure of a portion or the complete grid. This affects
the capability of the power system to operate in a stable mode with high degree
of reliability. In case the grid suffers from major contingencies and attacks, it may
become vulnerable and there have been many occurrences of cascading failures in
the recent past due to planned and unplanned outages of critical components of the
grid. The cascading failure has a devastating effect on power system operation and
the economy of the state. In western countries, researchers are assessing the possi-
bility of cascading failure of the grid when the grid is subjected to electrical contin-
gencies and unplanned outages. Defned by the North American Electric Reliability
Corporation (NERC), a cascading failure is “the uncontrolled loss of any system
facilities or load, whether because of thermal overload, voltage collapse, or loss of
synchronism, except those occurring as a result of fault isolation” [14].
There have been suggestions such as moderating the dynamic equilibrium of
the grid system to a point of self-organised criticality and strategic load shedding
schemes so that it is possible to reduce the effects of cascading failure. Practical
experiences reveal that intentional islanding of the power system is one such strat-
egy. In intentional islanding, the utility through a pre-program software goes for
the intentional splitting of the grid into separate controllable parts or islands. Each
island must have its own independent generation, and the presence of distributed
generation is preferred. In addition to intentional islanding, some strategic load shed-
ding may be done to balance the generation and load in the sub-systems. Intentional
islanding can be very helpful in isolating failures or localising them within the
region where they occurred and preventing them from spreading throughout the
system.
To ensure that the power grid survives following a major contingency or an attack
on its critical parts employing intentional islanding is one of the solutions; though it
is not always wise to opt for such a remedy until there is an utmost need for it. The
utmost need is generated when the system suffers from cascading failure. The criti-
cal elements in the grid network consists of vulnerable buses and lines, and in case
there is an outage of such component, the system may suffer tremendous damage and
Cascading Failure and Islanding in Grid Network 139

is prone to have cascading failure. If any of these vulnerable links happen to be the
initial failure, islanding must be initiated. Using the vulnerability index, proposed in
[15], it is possible to obtain the order in which the links fail.

6.3 MODELLING OF CASCADING FAILURE AND ISLANDING

Cascading failure in power transmission network is a potential threat which is usu-
ally triggered by a small and benign fault in a part of the transmission network and
subsequently causes islanding of the power transmission system. In the transmission
network model, a bus is treated as a node (or vertex) while a transmission line is
treated as a link (or edge). An interconnecting path within a pair of nodes is defned
as a sequence where any node is not repeated more than once. The geodesic distance
is defned as the path of minimal length between a pair of nodes, that is, the shortest
electrical distance between the pair of nodes. Obviously, the transmission of electri-
cal power between a pair of nonadjacent buses depends on the transmission lines
in the geodesic paths incorporating respective nodes (buses). The vulnerability of
a bus (node) in the power transmission network can be assessed by the quantum of
geodesic paths passing through it and is conventionally denoted by the betweenness
centrality of that bus [16].
The pure topological approach of analysing a network can be employed to fnd
critical elements of a transmission power network. However, this may not accurately
address identifying vulnerable elements in a power grid as such a transmission grid
possesses typical characteristics which needs to be taken into account while analys-
ing the vulnerability of the elements in the grid. This leads to the concept of electri-
cal betweenness [13] where power transmission network buses are categorised with
respect to their characteristics, and each power line performs in power transmission
system subjected to operating equality and inequality constraints.

6.4 APPLICATION OF PREFERENTIAL PROBABILITY

IN MODELLING OF CASCADING FAILURES
Cascading failure in power transmission network is a potential threat which is usu-
ally triggered by a small and apparently benign fault in a part of the transmission
network and subsequently causes islanding of the power transmission system An
interconnecting path within a pair of nodes is defned as a sequence where any node
is not repeated more than once. Obviously, the transmission of electrical power
between a pair of nonadjacent buses depends on the transmission lines in the geode-
sic paths incorporating respective nodes (buses). The vulnerability of a bus (node) in
the power transmission network can be assessed by the quantum of geodesic paths
passing through it and is conventionally denoted by the betweenness centrality of
that bus [16].
The pure topological approach of analysing a network can be employed to fnd
critical elements of a transmission power network. However, this may not accurately
address the identifcation of vulnerable elements in a power grid as such a transmis-
sion grid possesses typical characteristics which needs to be taken into account while
analysing the vulnerability of the elements in the grid. This leads to the concept of
140 Power Transmission System Analysis Against Faults and Attacks

electrical betweenness [13] where power transmission network buses are categorised
with respect to their characteristics and each power line performs in power transmis-
sion system subjected to operating equality and inequality constraints.
As stated earlier, betweenness of a bus (or node m) in a power network can be
defned as [13].

NG N D

T (m) =
1
2 ˜˜C ˜ f d
g
m
l
gd
(6.1)
g ˙G d ˙ D l˙L

where m ˜ g ˜ d ° Lm

˜fm
l
gd
represents the summation of PTDF of all the lines connecting bus m, power
l°L
being injected at bus g with power withdrawal at bus d.
1 d
2
Cg
m
˜
fl gd represents the quantum of transmitted power available at bus m dur-
l°L
ing transmission of electrical power from bus g to bus d. G represents the set of
generation buses; D denotes the set of load buses; NG and ND represent the number
of generation buses and the number of load buses, respectively; Lm is the set of lines
connected with bus m.
Electrical betweenness of a transmission line (link l) in a power network model
can then be represented as [13]

T ( l ) = max ˝˙T p ( l ) ,  T n ( l ) ˆˇ   l ˘ L (6.2)

where T p (l ) and T n (l ) are positive and negative electrical line betweenness of the
transmission line l.

T p (l ) = ˜˜C d gd
f
g l g ˙ d , fl gd > 0 (6.3)
g ˇG d ˇ D

T n (l ) = ˜˜C d gd
g l f g ˙ d , fl gd < 0 (6.4)
g ˇG d ˇ D

Here, Cgd fl gd represents the power fow in the line l. The set of electrical betweenness
qualifes the contribution of the grid network component, for example, a transmis-
sion line, to the power grid network, and thus the criticality of elements of the grid
network can be assessed. With higher electrical betweenness, the criticality of the
line is enhanced. Each line can then be associated with a defnite value of electrical
betweenness Tgd, max where

Tgd ,max = ˜ * T (l ) (6.5)

Cascading Failure and Islanding in Grid Network 141

Here, ε is the tolerance factor that represents the ability of a line to handle increased
electrical betweenness to retain the line operation within specifed limits. It may be
noted that ˜ > 1. A transmission line may be termed as overloaded, making it vulner-
able to failure, if Tgd, max > Cgd.
When considering power transmission [17] from a generator bus g to a load bus
d , as PTDF has a sign, if we specify a reference direction for line l the PTDF of
fl gd should be positive, negative or zero. Then, we determine positive betweenness
( fl gd > 0) and negative betweenness ( fl gd < 0) by Equations (6.3) and (6.4), respec-
tively. In Equation (6.3), if there is no fl gd > 0, then T p (l ) = 0 and in Equation (6.4),
if there is no fl gd < 0, then T n (l ) = 0.
The outage of a line or a bus connected to a line alters the shortest paths between
the buses (nodes), and consequently, the power fow of the tripped line gets redis-
tributed following the concept of preferential probability Πd. Such a process may
create overloads and subsequent overloading in other adjacent lines. The electrical
betweenness Tgd, max of the adjacent line(s) may exceed the corresponding limit(s) Cgd
for the overloaded line(s) following load redistribution and may trigger an avalanche
chain mechanism collapsing the entire system. During load redistribution, the pref-
erential probability can be expressed as [18]
cd˝
°d = (6.6)
˜c ˝
m
m˙ N

where α is a tunable parameter that can be varied from 0.2 to 2.2 in small-world
networks [9,19–23] like power grid, c is the degree (or connectivity) of d and N is the
set of neighbours of d.
Obviously, following collapse of a overloaded line, the additional electrical
betweenness ΔTgd transferred to an adjacent line is then given by

˛Tgd = T (l ) ˙ d (6.7)

This adjacent line may face outage and would induce further redistribution of power
fow making other surviving lines potentially vulnerable for collapse provided for
any surviving line, Tgd, max > Cgd.

Also, Tgd ,max = T (l ) + ˆTgd = ˜ * T (l ) (6.8)

This iterative process for removal of an overloaded line followed by redistribution of loads
among adjacent lines with subsequent identifcation and tripping of adjacent line(s) with
high degree of betweenness leads to cascading failure resulting in islanding condition.

6.5 SIMULATION
To illustrate the dynamic behaviour of cascading failures, the developed concept
has been validated frst in IEEE 57 bus test system [24] and then in the part of the
Eastern grid system of India (203 bus, 267 lines and 24 generator electrical grid
network). The single line diagram of both the test systems is shown in Appendix A
and B respectively.
142 Power Transmission System Analysis Against Faults and Attacks

The following algorithm illustrates the process of testing a grid system against
possible cascading collapse following tripping of a line with highest electrical
betweenness.

Step 1: A generation dispatch process is performed with generation-load bal-

ance using Optimal Power Flow program employing Newton–Raphson
method in grid systems under consideration. The output exhibits parameters
of grid operation including power fow in the lines.
Step 2: Loading at heavily loaded buses are increased and generation-load
balance is re-established. The network power fow model is executed to cal-
culate subsequent power fow in each line.
Step 3: Electrical betweenness of the lines (links)/buses (nodes) for the model
network are calculated.
Step 4: The line (link) having the highest value of electrical betweenness is
considered to be the overloaded line. The bus (node) with the highest electri-
cal betweenness is assumed to be a potential threat facing bus failure.
Step 5: The overloaded line is tripped and/or the bus with highest electrical
betweenness is removed following a simulated failure. The loading of adja-
cent lines is calculated using the developed concept of power fow redis-
tribution associated with the concept of preferential probability for load
redistribution.
Step 6: Criterion of overload (i.e. Tgd, max (the sum of T (l ) and ΔTgd) > Cgd) for
each of the adjacent lines is checked.
Step 7: Any line which satisfes the criterion Tgd ,max > Cgd is overloaded and
tripped with subsequent redistribution of power fow to adjacent lines.
Step 8: Computations in Steps 6 and 7 are executed for these adjacent lines. If
this process ends with surviving lines having Tgd ,max ˜ Cgd , the system sur-
vives. However, if this process ends with an ultimate line where Tgd ,max > Cgd ,
the system is subjected to cascading failure and islanding may occur.

In the simulation, the IEEE 57 bus test system has frst been tested for failure of high-
est betweenness line as well as for failure of the bus with highest electrical between-
ness. In this process, the transmission line with the highest electrical betweenness in
the IEEE 57 bus test system has been identifed (line 79) and removed in the simula-
tion. This induces load redistribution among adjacent lines (line 10, 11, 13, 15 and 25).
Line 11 has now the highest electrical betweenness and is removed (Table 6.1). The
simulation reveals that continuation of the iterative process for removal of an over-
loaded line followed by redistribution of loads among adjacent lines with subsequent
identifcation and tripping of adjacent line(s) with high degree of betweenness leads
to islanding condition. It is clear from Table 6.1 that at the end of the process only
line 21 survives, and tripping of the this line leads to termination of the process and
the system is islanded. The mapping diagram of this process is shown in Figure 6.1
(circles indicate the islanded nodes while the rectangles are the line numbers). This
diagram depicts the mapping pattern of cascading failure, and it may be noted that
the tripping or cascading failure occurs only in the direction shown by arrows and
not otherwise or in any reverse process.
Cascading Failure and Islanding in Grid Network 143

TABLE 6.1
Results on Tripping of Line 79
Line No. Adjacent Lines πd Tgd,max Cgd
79 10 0.4836 22.8165 70
11 0.3631 25.8475 20
13 0.2362 31.8237 30
15 0.5352 45.4841 50
25 0.4849 74.8006 80
11 10 0.4351 22.0706 70
12 0.2672 43.0848 30
13 0.3962 35.9426 30
15 0.2436 36.8447 50
74 0.5964 41.2983 45
12 21 0.5632 64.6761 50
23 0.2975 24.6177 25
13 0.1326 29.1567 30
74 0.8973 49.0825 45
21 22 0.3261 1.6372 50
23 0.8964 35.9826 25

FIGURE 6.1 Load redistribution resulting islanding initiated by outage of line 79.

In the next step, standard load fow program is executed applying contingency
with tripping of lines, as indicated in the preceding steps. It has been found that load
fow convergence is possible till line 23 is tripped. Following tripping of line 23, the
load fow does not converge, indicating that the system is subjected to failure.
In the next step, the bus with the highest electrical betweenness (node 12) is
tripped in IEEE 57 bus test system. This induces load redistribution among adjacent
144 Power Transmission System Analysis Against Faults and Attacks

FIGURE 6.2 Load redistribution resulting islanding initiated by outage of node 12.

lines (line numbers 12, 13, 14, 21 and 23). The simulation reveals that continuation
of the iterative process for removal of an overloaded line followed by redistribution
of loads and subsequent identifcation and tripping of line(s) with high degree of
betweenness leads to cascading failure. The mapping pattern of this process is shown
in Figure 6.2.
In the 203-bus test system, the same process is repeated, and it has been observed
that cascading failure process is initiated with removal of a single line with the high-
est electrical betweenness (line number 206). Outage of line 206 results in redistribu-
tion of power fow associated with preferential probability in the neighbouring lines,
as shown in Table 6.2.
It reveals that for line number 14 and 13, the condition Tgd, max > Cgd is satisfed.
From outage of line number13, it can be observed from Table 6.2 that the criterion of
overload appears in line number 165 and 17 (as for each of these two lines Tgd, max > Cij
is satisfed). Following outage of line 165, the process of further redistribution of
loads is terminated because it is one of the terminal lines creating the state of island-
ing for the 203 bus network. The results obtained in Table 6.2 have been pictorially
depicted in Figure 6.3 where the rectangles represent the transmission lines (links or
edges of complex network) and the circles represent the buses (nodes or vertices of
complex network).
The process of cascading failure for the 203 bus network has been shown in
Figure  6.4 following tripping of the node (node number 101) that has the highest
electrical betweenness in the model power network. During islanding operation, the
system parts into two or more sections and each section may tend to have overvoltage

TABLE 6.2
Results on Tripping of Line 206
Line No. Adjacent Lines πd Tgd,max Cgd
206 207 0.4982 54.6062 85
14 0.5277 48.9331 40
13 0.7782 88.5479 80
241 0.6962 65.5667 70
13 237 0.4932 48.8378 90
165 0.6742 57.8635 45
17 0.4642 62.9876 50
165 157 0.6329 96.5728 90
Cascading Failure and Islanding in Grid Network 145

FIGURE 6.3 Load redistribution resulting islanding initiated with outage of line 206.

FIGURE 6.4 Load redistribution resulting islanding initiated by outage of node 101.

and undervoltage problems depending upon the amount of generation available in

each section. The loading level, nominal rating, short-term emergency rating and
long-term emergency rating for some of the network components (generators, trans-
formers and lines) might have been violated under such a condition.
To check the results obtained, load fow program is again executed, and it has
been observed that with sequence of tripping of lines as considered in the preceding
portion leads to islanding of the 203 bus grid network, validating the use of electrical
betweenness for assessing the cascading failure of a grid network.
The electrical betweenness of the transmission lines in the IEEE 57 bus test grid
as well as the part of Eastern grid of Indian power system have been calculated fol-
lowing the concepts of complex network theory. The lines are arranged in descend-
ing order from the highest value of line electrical betweenness down to the lowest
value. Tripping of a line, with highest betweenness value, has been considered in this
simulation. Its load gets redistributed to the adjacent line(s) following preferential
probability of load redistribution subsequent to outage of the line. The new values
of line betweenness of the surviving adjacent lines are checked to identify the next
overloaded line(s) prone to failure. Once such an overloaded line collapses, its load
is again redistributed to other lines following preferential load redistribution, giving
rise to possibility of further overloading in other surviving line(s) and subsequent
tripping of such line(s). In this investigation, the roadmap of cascading failure with
islanding condition of a grid network followed by failure of overloaded transmission
lines in steps has been mapped to assess the vulnerability of the grid network under
consideration.
146 Power Transmission System Analysis Against Faults and Attacks

6.6 SUMMARY
In this chapter, a topological cascading failure assessment approach is described for
complex power grids. The property of electrical betweenness has been utilised to
identify critical lines and tripping of such a line with the highest electrical between-
ness resulting in load redistribution associated with preferential probability of load
attachment. To validate the developed concepts, IEEE 57 bus test system and a typi-
cal 203 bus grid network of eastern India have been utilised. Outage of a line with the
highest electrical betweenness within such a grid network has been simulated, and
it has been observed that the redistribution of loads to the adjacent lines associated
with preferential attachment leads to overload of an adjacent line(s). The iteration of
simulated outage of such an overload line and determination of subsequent loadings
on surviving adjacent lines is continued in this investigation so that the process is
converged to a single overloaded line. It has been observed that tripping of this ter-
minating line would result in islanding of the power grid. Hence, it is concluded that
this method is an effcient technique for analysing and assessing cascading failure
following outage of overloaded line(s).

REFERENCES
1. R. Albert, I. Albert, and G.L. Nakarado, “Structural vulnerability of the North
American power grid”, Physical Review E, vol. 69, p. 025103, 2004.
2. P. Hines, K. Balasubramaniam, and E.C. Sanchez, “Cascading failures in Power grids”,
IEEE Potentials, vol. 28, no. 5, pp. 24–30, 2009.
3. Yang Nan, Liu Wenying and Guo Wei, “Study on scale-free characteristic on propa-
gation of cascading failures in power grid”, IEEE 2011 EnergyTech, Cleveland, OH,
pp. 1–5, 2011, doi: 10.1109/EnergyTech.2011.5948519.
4. L.L. Lai, H.T. Zhang, C.S. Lai, F.Y. Xu, and S. Mishra, “Investigation on July 2012 Indian
blackout”, 2013 International Conference on Machine Learning and Cybernetics,
14–17 July 2013. doi: 10.1109/ICMLC.2013.6890450.
5. P. Crucitti, V. Latora, and M. Marchiori, “A topological analysis of the Italian electric
power grid”, Physica A, vol. 338, pp. 92–97, 2004.
6. R. Kinney, P. Crucitti, R. Albert, and V. Latora, “Modelling cascading failures in North
American power grid”, European Physical Journal B, vol. 46, pp. 101–107, 2005.
7. IEEE Working Group on Understanding, Prediction, Mitigation and Restoration of
Cascading Failures, “Benchmarking and validation of cascading failure analysis tools”,
IEEE Transactions on Power Systems, vol. 31, no. 6, pp. 4887–4900, 2016.
8. S. Pahwa, “Dynamics on complex networks with application to power grids”,
Dissertation Thesis, Kansas State University, 2013.
9. A.E. Motter and Y.C. Lai., “Cascade-based attacks on complex networks”, Physical
Review E, vol. 66, p. 065102, 2002.
10. G.A. Pagani and M. Aiello, “The power grid as a complex network: A survey”, Physica
A, vol. 392, pp. 2688–2700, 2013.
11. K. Sun, “Complex networks theory: A new method of research in power grid”, In
Proceedings of IEEE PES Transmission and Distribution Conference and Exhibition,
Asia and Pacifc, Dalian, China, pp. 1–6, 2005.
12. D. Wu, E. Bompard, and F. Xue, “The Concept of Betweenness in the Analysis of Power
Grid Vulnerability,” In Engineering. Complexity in, Rome, Italy, pp. 52–54, 2010, doi:
10.1109/COMPENG.2010.10
Cascading Failure and Islanding in Grid Network 147

13. E. Bompard, D. Wu, and F. Xue, “Structural vulnerability of power systems: A topolog-
ical approach”, Electric Power Systems Research, vol. 81, no. 7, pp. 1334–1340, 2011.
14. North American Electric Reliability Corporation, “Evaluation of criteria, methods, and
practices used for system design, planning, and analysis response to NERC blackout
recommendation 13c, 2005. http://www.nerc.com/docs/pc/tis/.
15. M. Youssef, C. Scoglio, and S. Pahwa, “Robustness measure for power grids with respect
to casacading failures, In Proceedings of the International Workshop on Modeling,
Analysis and Control of Complex Networks, San Francisco, CA, pp. 45–49, September
2011.
16. G. Rout, T. Chowdhury, and C.K. Chanda, “Betweenness as a tool of vulnerability anal-
ysis of power system”, Journal of Institution of Engineers (India): Series B, Springer,
vol. 26, no. 8, pp 2443–2451, 2016.
17. A.V. Gheorghe, M. Masera, and P.F. Katina, “Infranomics: Sustainability, Engineering
Design and Governance”, New York: Springer Ltd.
18. J. Wang, L. Rong, and L. Zhang, “A new cascading model on scale free network
with tunable parameter”, First International Conference on Intelligent Networks and
Intelligent Systems, IEEE Computer Society, Washington, DC, 2008.
19. X.F. Wang and G.R. Chen, “Complex networks: Small-world, scale-free and beyond”,
IEEE Circuits and Systems Magazine, vol. 3, no. 1, pp. 6–20, 2003.
20. P. Crucitti, V. Latora, and M. Marchiori, “Model for cascading failures in complexnet-
works”, Physical Review E, vol. 69, p. 045104, 2004.
21. K.-I. Goh, B. Kahng, and D. Kim, “Universal behaviour of load distribution in scale
free networks”, Physical Review Letters, vol. 87, no. 27, p. 278701, 2001.
22. R. Cohen, “Breakdown of the Internet under international attack”, Physical Review
Letters, vol. 86, no. 16, p. 3682, 2001.
23. Y. Zhang, S. Cai, C. Chen, and J. Shi, “Robustness of deterministic hierarchical net-
works against cascading failures”, IEEE International Conference on Electrical and
Control Engineering, USA pp. 4063–4066, 2011.
24. http://www.ee.washington.edu/research/pstca/.
 7 Assessment of
Resilience in Power
Transmission Network

7.1 INTRODUCTION
The word resiliency is derived from the Latin word “resilio,” which literally means
the ability of an entity to rapidly get back to its original state, shape, health or posi-
tion, after enduring and opposing stress applied on it. Based on these, resilience of
networks, systems, infrastructure, machines, human health, community response to
epidemics or natural disasters have been defned. Power grid resiliency is gaining
importance as climate change increases the threats to modern infrastructure.

7.2 CONCEPTS OF RESILIENCY, ROBUSTNESS,

RELIABILITY AND STABILITY
The concept of resilience is still in the early stages of research, and though there
are multiple defnitions, there is no universally accepted one. Overlying concepts
of risk assessment, reliability, recovery and robustness contribute to defnitions of
resiliency. Some sources [1] reveal resilience of a system depends on how capable it
is to decrease the intensity and extent of impact caused by the disturbances, whereas
some researchers [2,3] defne resilience as the ability of a system to identify risks.
Analysis of system performance focuses on system failure defned as any out-
put value in violation of a performance threshold (such as a performance standard).
System performance can be described from three different viewpoints: (i) how often
the system fails (reliability), (ii) how quickly the system returns to a satisfactory state
once a failure has occurred (resiliency) and (iii) how signifcantly strong the system
remains to face the consequence of failure (robustness).
On the basis of power system operating state:
Robustness to a class of disturbances is defned as the ability of a system to main-
tain its function (normal state) when it is subjected to disturbances of this class.
Resilience to a class of unexpected extreme disturbances is defned as the ability
of a system to gracefully degrade and to quickly self-recover to a normal state.
Reliability is the ability of the power system to deliver electricity to customers
with acceptable quality and in the amount desired while maintaining grid functional-
ity even when failures occur.
Or reliability is the probability that no failure occurs within a fxed period of
time, often taken to be the planning period.
Resiliency will describe how quickly a system is likely to recover or bounce back
from failure once failure has occurred. The resilience of a system presented with an
149
150 Power Transmission System Analysis Against Faults and Attacks

unexpected set of disturbances is the system’s ability to reduce the magnitude and
duration of the disruption. A resilient system downgrades its functionality and alters
its structure in an agile manner. If failures are prolonged events and system recovery
is slow, this may have serious implications for the transmission system.
Robustness is the ability of a system to cope with a given set of disturbances and
maintain its functionality. Robustness is concerned with strength, whereas resilience
is concerned with fexibility.
Robustness and resilience belong to two different design philosophies. When a
robust grid is attacked, it may break like an oak tree in a storm. When a resilient grid
is attacked, it can bend and survive like a reed in a storm. From a system engineering
point of view, absolute robustness can actually lead to fragility. The interrelationship
between these three parameters can be explained by Figure 7.1.
In the wake of unprecedented disasters and attacks, robustness and resilience have
become buzz words in many disciplines including biology, ecology, sociology, sys-
tems engineering and infrastructure engineering. The traditional defnition of resil-
ience in systems engineering is the capacity for fast recovery after stress and for
enduring greater stress [4,5]. In systems engineering, resilience includes maintaining
system functionality following disturbances. Robustness, on the other hand, explains
the ability of a system to resist change without losing stability [6]. In some disci-
plines like social systems and organisational systems, the term resilience is similar
to the term robustness. In infrastructural systems, and especially in power systems,
however, the term robustness and resilience are more distinct; this is due to power
systems structure and function centring on conductor lines delivering electric power
to a certain area within specifc voltage and frequency ranges.
Extreme robustness actually leads to fragility. Power systems are usually robust
enough to withstand one contingency [N − 1] or two contingency [N − 2] events,
where N stands for the number of system buses. However, beyond that, they are gen-
erally vulnerable. Moreover, the robustness is usually used with specifc assumptions
for protection system operation under pre-defned operational ranges for voltage and
loading.

FIGURE 7.1 Block diagram of reliability, robustness and resilience in control operation.
Resilience in Power Transmission Network 151

The concept of reliability is also used in industrial and systems engineering and
is accompanied by statistical and probabilistic approaches that characterise system
performance after predicted and unpredicted failures. Reliability and stability are
two more explicit power system concepts that pre-date the terms “robustness” and
“resilience” [6]. Reliability and stability are well studied concepts in power systems.
Similarities and dissimilarities between them and the terms robustness and resilience
can inform future cyber physical resilience studies.
Discussions of resilience often centre around a system survivability that lever-
ages load shedding, generation outages and other actions. Reliability is a measure
of the system’s ability to serve all loads. The system’s ability to serve loads is tra-
ditionally referred to as service availability, which falls under the power systems
defnition of reliability. Reliability indices are usually expressed in terms of the
probability of load loss [6]. The loss of load probability is expressed in days per
year. Reliability is primarily concerned with the risk of service interruption of
device failure.
“Stability is the ability of a system to remain intact after being subjected to small
perturbations” [7]. In power systems, stability for a given initial operating condition
means the system will regain operation equilibrium state after small perturbations.
Stability is focused on the system equilibrium point. However, the concept of robust-
ness in power systems goes beyond stability. To be robust, the electric grid has to be
stable in the face of small perturbations as well as major equipment failures, man-
made attacks and natural disasters [7].
Power system resiliency generally refers to the ability of the system to deal with
unpredictability of catastrophic weather events and robustness of the infrastructure.
Resiliency of transmission lines are [8–10] different from that of distribution lines.
Transmission lines undergo extensive planning, maintenance, making them strong
enough to withstand forces of nature. Distribution systems, on the other hand, have
fnancial constraints for robust planning and 90% of the power outages are due to the
inability of the distribution system network to withstand strong disturbances leading
to power outages downstream of sub-transmission networks.
Metrics of resiliency are necessary to assess about preparation for ensuring proper
operation and safety of existing electrical network when the network is subjected to
contingency or attack. Resiliency implies the immunity of the electrical network and
it is a subjective concept. It can be assessed through its metric but there is practically
no way to quantify the resiliency. The resiliency related to the planning and opera-
tion of electrical power system deals with the effect of sudden disturbances to the
network [11]. For long-term resiliency, it is important to determine how the power
system infrastructure of a country would sustain itself in case imported raw materi-
als suddenly become unavailable.
There are considerable amount of research work on system restoration and
optimal recovery strategies in power system networks following a vulnerable dis-
turbance. Usually, the system operators heavily depend on switching of different
portions of the network to ensure proper restoration of the electrical network sub-
jected to unplanned contingency or attack but in case the switches face catastrophic
damages due to the vulnerability of the contingency or attack, then the restoration
policy fails.
152 Power Transmission System Analysis Against Faults and Attacks

To assess the resiliency, due consideration may be given in the following aspects:

• A way of increasing the resiliency of the system is using microgrids.

Microgrids are normally having self-sustaining resources. They are effec-
tive in several smart grid projects to prevent wide spread outages. When
the multiple microgrid system islands from the grid, it may prevent cascad-
ing outages in the wide area network by lowering load demand. However,
it does not guarantee that the critical loads within the micro grid would
continue to operate properly. Metrics of resilience of the power network
ensures the level of preparedness against unforeseen attacks with integra-
tion of multiple microgrids.
• Resiliency indices of microgrids can adjust with microgrid controllers or
changes in design if it is possible to demonstrate remarkable enhancement in
resiliency after implementing new algorithm or installation of new devices.
• In any electrical system it is very important that there is remarkable resil-
iency to ensure the robustness of critical electrical loads associated with
military bases, hospitals and police stations.
• There is a possibility that resiliency have an impact in the economic aspects
of power exchange between two microgrids in long term, or between a
transmission company and a microgrid.
• Proper resiliency helps to improve the performance and design of the
electrical protection scheme of the system, machinery and algorithms of
operation.

Several isolated yet similar occurrences over the past emphasise the range of chal-
lenges power systems need to overcome. In addition, an eye must be kept on how
these challenges are evolving due to climate change and emergence of new tech-
nologies. Strong winds, especially when combined with rain, hail and snow from
seasonal storms, can damage electricity utility systems, resulting in service inter-
pretations to large numbers of electricity customers, which is likely to include criti-
cal loads like airports, hospitals, city halls and other buildings deemed important
to the community [12]. While most such power outages are caused by faults as a
result of trees (or their branches) falling on local electricity lines and poles, major
power outages tend to be caused by extreme weather events. Here are some of the
examples:

• North-eastern states of the United States were hit by Hurricane Sandy in

2012, which destroyed over 1,00,000 primary electrical wires along with
several substation transformers exploded and fooding numerous substa-
tions. These lead to the interruption of power supply, affecting ~7,000,000
consumers [13].
• During the summer of 2010–2011, Queensland, Australia was affected by
widespread fooding that resulted in signifcant damage to six zonal sub-
stations and a large number of nodes, transformers and overhead wires.
Approximately 150,000 consumers were affected by regular, long power
outages over the entire two summers [14].
Resilience in Power Transmission Network 153

• In 2008, China had to bear the consequences of a severe ice storm, which
led to failure of about 2000 substations, resulting in the collapse of about
8,500 poles, resulting long power disruptions in 13 provinces and more than
170 cities and towns across the world’s most populated country [15].

Commonly used practices for resiliency are making lines underground, and/or
removing the trees which caused faults, and/or taking mitigating action against
rodents causing short circuit fault; however, these approaches are neither environ-
ment friendly nor a sustainable, economic and universal solution.
There is a huge drive to improve the resilience of the power infrastructure to
such increasing occurrences of power outages due to abrupt weather disturbances.
However, without having a formal procedure to measure the existing resiliency of
the system, it may be an uncertain path to follow while reconfguring the existing
network to be more robust. Now, before proceeding to the details about assessment
of resiliency, it is important to review the existing defnitions and interpretations of
resiliency.

7.2.1 DEFINITIONS OF RESILIENCY

The frst formal defnition of resiliency was published in 1973 by Holling [16], who
said that resiliency is a form of persistence, the ability to absorb shock and changes
and “still maintain the same relationships between populations or state variables.”
Gunderson et al. [17] modifed the defnition by adding buffer capacity for absorbing
perturbations in a timely manner. Walker et al. [18] extended the defnition to include
the ability to self heal during disturbances. Kendra and Wachtendorf [19] described
“bouncing back from a disturbance” as a crucial aspect of resilience. The breadth
and number of defnitions for “resilience” has increased signifcantly over the last
decade, making it diffcult to fnd a universal understanding of the term “resilience.”
Merriam Webster defnes resilience as “the ability to recover from or adjust easily
to misfortune or change.” There are numerous defnitions of “resiliency” that have
been introduced for any physical network infrastructure and the socio-economic
system in the past decades [20–24]. In general, it can be referred as the ability of
a system or an organisation to react and recover from unpredictable disturbances
and events [25]. Zio [26] modifed the concept of resiliency by stating “…systems
should not only be made reliable, i.e. with acceptably low failure probability, but also
resilient, i.e., with the ability to recover from disruptions of the nominal operating
conditions.” According to Haimes, “resilience is the ability of the system to with-
stand a major disruption within acceptable degradation parameters and to recover
within an acceptable time and composite costs and risks.” In 2013, White House
declared “resilience is the ability to prepare for and adapt to changing conditions and
withstand and recover rapidly from disruptions… [It] includes the ability to with-
stand and recover from deliberate attacks, accidents or naturally occurring threats or
incidents.” Committee on Increasing National Resilience to Hazards and Disasters,
Committee on Science, Engineering, and Public Policy and the National Academies
on their collaborative effort defned in 2012 that “resilience is the ability to prepare
and plan for, absorb, recover from, and more successfully adapt to adverse events.”
154 Power Transmission System Analysis Against Faults and Attacks

Though there are some distinctions existing among various defnitions, it is wor-
thy to compare them to reveal some relevant aspects of system operations.
Basically, resiliency addresses the concern that assesses the level of preparedness
of a system encountering disruptions. It is important to clarify in resiliency assess-
ment that how much of the service has been degraded and how quickly the service
has been restored. Moreover, it is important to assess how completely the service
has been restored. Resiliency also describes the degree of disruption across mul-
tiple dimensions, which could include type, quality, time and geography of service
provision.
It is important to assess the state of the system from the viewpoint of its design
and redundancy. For example, an electricity grid system that is designed with more
redundancy, operated with more contingencies for back up and designed with recov-
ery might experience a lesser and briefer disruption; and, if so, would be more resil-
ient than a system that has less redundancy has fewer backups and is more diffcult
to rebuild.
Moreover, it may be noted that different responses lead to different resiliency at
different costs. For example, with additional resources, it is possible to restore the
electrical system while it is subjected to outages following disruptions. It is also
important to utilise more effcient equipment, and, consequently, the quality of ser-
vice provided after recovery could exceed the original level of service provided.
It is interesting to note that resiliency also depends on the timescale. With effcient
maintenance an electrical network would suffer lesser disruption and even if there
are disruptions, the restoration will be more prompt and quick. Indirectly, this would
increase the resiliency of the power network.
Because resiliency is a subjective concept, defnition of resiliency includes some
additional relevant characteristics of the network. These characteristics are clarifed
as redundancy of the network, reliability, sustainability, vulnerability, fault tolerance
capability and fast clearing circuit breakers, recoverability, etc.

7.2.2 FRAMEWORK OF RESILIENCY

The literature on resiliency shows that there are many different opinions about the
“phenomenology” of resilience – which characterises resilient performance [24].
Therefore, instead of considering what resilient performance is, the proposed work
considers what enables resilient performance, what makes it possible, and conversely
what would make it impossible, if it was missing. This explains the philosophy of
the motivation of the proposed framework. The proposed framework considered the
diverse aspect of power systems’ disruption and functionality. The frst block dia-
gram events set the base for the framework. Other blocks of the framework explain
the characteristics of the grid on the basis of chosen events. Because power grid
has diverse and continuous challenges as explained previously, it need an organised
structure to deal with it. This is what motivated the authors to prepare a framework.
In the framework, each block is based on some rational questions on the basis of
which required action is to be taken to fulfl the resilient grid objective.
There are two ways of preparing a framework; one is strategic in which resources
are managed for desire outcome and another is operational based on available
Resilience in Power Transmission Network 155

resources outcome. In this chapter, mostly strategic perspective is considered because

economies like United States, India and China not only have wide resources but also
have the capability to manage the required resources.

A. Events: Background: What are the events (primary causes of outage like
tropical storm) for which the utility has a prepared response? How these
events were selected (experience, expertise, tradition, regulatory require-
ments, design basis, risk assessment, industry standard, etc.)
B. Resource allocation: In this section, the resource allocated for the event
response is to be studied. Budgets, number of generators available, man-
power, smart grid tech and questions related to monitoring system can be
asked in this section. This section utilities special arrangement for a par-
ticular event is of major attention
C. Capacity and capability: Capacity defnes the organised structure of
the allocated resource. The type and intensity of the damage to the grid
depend upon kinds of event. The capability of allocated resource to handle
the event specifc damage to grid is desirable. For the event of storm, the
required capabilities for response is different than in the event of cyber
attacks.
D. Threshold: When is a response activated? What is the triggering criterion
or threshold? Is the criterion absolute or does it depends on internal/external
factors? Is there a trade-off between, for example, safety and productivity?
Are some important that is associated with this section.
E. Performance and verifcation: This block is dedicated to the analysis of
the produce result. After an event recovery does not imply perfect restora-
tion of the system functionality rather it implies that system has returned to
a state where it is considered functional. Effciency, reliability, fault toler-
ance and robustness are some performance characteristic measures, which
are helpful for data verifcation.
F. Outcome: Whether the performance through prepared response is accept-
able. How much system generated the outcome that was it seeking to
achieve. Reduced damage from disasters and increased economic activity
are some measures of outcome.
G. Relevance: This is the beauty of proposed framework which allow utilities
to raise question upon its preparation. The most important question is that
when the response was prepared, and when was it last updated? And under
which circumstances?

A single event has the capability to create major disturbances in the grid but black-
outs are not the results of single event and defciency but the combination of sev-
eral defciencies. The following preconditions are the basis for a high power outage
risk:

• High loadability limit or high grid utilisation

• For better unit commitment and optimal power fow
• Defects due to ageing infrastructure
156 Power Transmission System Analysis Against Faults and Attacks

Under these abovementioned conditions grid vulnerability increases. Now the likeli-
hood of a power blackout for the following events is very high:

• Power plant shutdown due to capital maintenance or preventive maintenance

or due to supply failures (e.g. cooling water shortage during heat waves)
• Operating personnel failure during maintenance work or switching
operations
• Simultaneous grid interruption, for example, short circuit caused by tree
contact, excavation work, balloons drifting into power lines, cars hitting
utility poles, provisional shutdown due to electrical overloading risk
• Sudden demand of load growth
• Power line and its related electrical equipment may undergo outage
• Under breakdown due to natural calamities (e.g. storm, earthquake, snow or
ice load, food, lightning, extreme temperatures)
• Lack of communication between transmission/distribution system person-
nel and power generation company
• Intentional cyber attacks

7.3 RELATIONSHIP BETWEEN RESILIENCY

AND PERCOLATION THRESHOLD
The ability of the power system to resist, mitigate and overcome stresses, failures
and their consequences is receiving increasing attention from managers and sys-
tem engineers with robustness, and resilience is gradually becoming very important
considerations along with the more typical considerations of low cost and reliable
operation. Moreover, proper understanding of the vulnerable and critical locations
in transmission systems provide invaluable information that may be used to inform
power grid management practices and programmes, leading to more realistic risk
assessments and the development of defensive strategies to ensure network survival
in case of extreme events and natural or man-made disasters.
Modern power grid is highly interconnected and plays a vital role in fuelling
infrastructure and electric markets. Any disruptions in any components of power
system, which results in cascading failures due to natural disasters, such as acts of
war and terrorism, can ripple through the entire power system magnifying the orig-
inal damage. Even relatively minor disturbances, such as congestion of transmis-
sion lines, can result in disproportionately severe disruptions to the electric power
industry. Large centralised power generation sources, transmission lines, substations,
especially those with high-voltage transformers, control centres are potential targets
for terrorists and are vulnerable.
Network structure, network dynamics and failure mechanism play important roles
in determining the resilience of realistic critical infrastructures, which may also shed
light on the study of the resilience of the power grid. The characterisation of resil-
ience is the main motivation for the studies involving complex network analysis and
power grid [10]. The behaviour in terms of connectivity of the network when nodes
or edges are removed is the primary question in many works considering failures
that happen in a random fashion or following an attack. In general, the reliability is
Resilience in Power Transmission Network 157

assessed by evaluating the connectivity or the ability to effciently guarantee paths

between nodes when nodes or edges in the network are removed [27]. Consequently,
for failures related to nodes, all the samples show a good resilience to random break-
downs. In fact, the network is always able to guarantee certain connectivity until the
numbers of nodes removed are the biggest part of it. On the other hand, the grids
are extremely vulnerable to targeted attacks, that is, failures that focus on key nodes
for the entire network such as high degree nodes or nodes with high betweenness or
nodes or lines that manage the highest amount of load or electricity fow.
Intentional attacks strongly deviate from random failures: even a small fraction
of removed nodes having large degrees has dramatic consequences. To predict the
effects of such directed attacks on network structure, the critical probability associ-
ated to network breakdown can be computed.
The probability of achieving connectivity from the source nodes to the terminal
nodes gives the reliability of the network. When terminal connectivity is concerned,
the identifcation of operational limits of a network goes missing where a critical
fraction of functional components to sustain the network is considered instead of
studying paths in the terminal reliability. This gap is overcome by percolation theory
which provides an opportunity by referring network failure to the situation whereby
a critical fraction of network components have failed.
In percolation theory, the failure of a node/edge of network is modelled by
removal. As the removal of nodes/edges increases, the network undergoes a tran-
sition from the phase of connectivity (functional network) to the phase of discon-
nectivity (nonfunctional network). The probability threshold signifying this phase
transition can be found theoretically or computed numerically by percolation theory.
The probability threshold can be used as a statistical indicator for the operational
limits of the network, which is not considered in traditional terminal reliability anal-
ysis. Thus, percolation theory, based on statistical physics, can help to understand the
macroscopic failure behaviour of networks in relation to the microscopic states of the
network components.

7.4 POWER SYSTEM BEHAVIOUR AT PERCOLATION THRESHOLD

In application of complex network theory for an electric grid network, the grid buses
are abstracted as nodes and transmission lines as links [28]. During operation at
steady-state condition, the nodes are ftted with a probability of ρ = 1, ρ being the
probability that a node is functional [27]; however, when the system encounters a
contingency or an attack that would affect the nodes of the power system, the prob-
ability of each node to remain functional would become a fraction <1 and would
have the probability (1 − ρ) to be nonfunctional where ρ < 1. The threshold value
of probability of each node being functional under attack is termed as percolation
threshold ρc [27]. It can be used to obtain the critical fraction of nodes that can afford
to be damaged from any contingency or attack. At ρ = ρc the connectivity reaches the
point of criticality following the damages sustained by the entire network. Whenever
ρ > ρc there would be connectivity for the respective nodes.
Once the percolation threshold is obtained for the grid network under attack, it
is then possible to determine whether at least one path remains connected with the
158 Power Transmission System Analysis Against Faults and Attacks

main power network when the network is subjected under contingency or attack con-
dition. Molloy–Reed criterion [29] being an effective tool to determine the percola-
tion threshold of the nodes in a complex network, it can be applied in assessing the
resiliency of the grid network under stress. Following this criterion, the percolation
threshold transition is observable when <c2> = 2<c>, where <c2> is the second
moment of degree distribution and <c> is the frst moment of degree distribution.
When <c2> → 2<c>, the power grid network transforms its nature from a nonresil-
ient system to a resilient one in the limiting case.
Once the threshold of the grid network is established, it would be feasible to assess
the resiliency of a specifc structure of the grid network so that at least one critical load
is catered (which may be used as a resource point to restore other loads affected by the
vulnerability of any attack). In case the fraction (r) of the affected nodes of the grid net-
work becomes lesser than the critical fraction rcri of the quantum of affected nodes of
the grid network, the grid system would have better resiliency. In simulating the power
network, encountering attack or contingency, it would be useful to determine rcri of the
nodes that can be afforded to get damaged while maintaining the resiliency.
While encountering an attack or facing a contingency, the network graph con-
fguration gets altered. This altered graph of the network is then termed as damaged
graph. The damaged graph is characterised by the following degree distribution P(c)
[30,31]:

˙ i ˘ i −c
P (c ) = ˜ˇˆ  r    (1− r ) P (c )
c 
c
(7.1)
i c

The standard generating function methodology employed to study the percolation

properties identifes the two frst generating functions of the damaged graph which
can be represented as:
ˆ

F0 ( x ) = ˜P (c) (1− r ) x c
(7.2)
c

ˆ
1
F1 ( x ) =  
c ˜c  P (c) (1− r ) x c−1
(7.3)
c

Here, F0 (1) is the fraction of nodes from the original graph belonging to the damaged
graph and F1 (1) is the relation among c and the average number of nodes from V
that can be reached after deleting a fraction r of nodes. The generating function of
other components which can be reached from a randomly chosen node is:

H1 ( x ) = r + xF1 ( H1 ( x )) s (7.4)

And the generating function for the size of the component to which a randomly cho-
sen node belongs to is [30]:

H 0 ( x ) = r + xF0 ( H1 ( x )) (7.5)
Resilience in Power Transmission Network 159

Thus, the average component size other than the giant component will be:

s = H 0/ (1) = 1 − r + F0/ (1) × H1/ (7.6)

After long algebra, we see that this leads to singularity when F1/ (1) = 1
To ensure the percolation of the damaged graph, the following inequality holds
good:

∑c (c − 2) P (c) > ∑c (c −1) r   P (c)

c c (7.7)
Or, c −  2 c > r   c − c
2
( 2
)
Hence, it is possible to establish [32] equations given in (7.8) and (7.9).

c2 = ∑c P (c) 2

c (7.8)
= (1− r )   c 2 + r   (1− r )   c
2

c = ∑cP (c) = (1− r ) c (7.9)

c

Application of Molloy–Reed criteria [30] in solving Equations (7.5) and (7.6) yields
the expression for critical fraction of node removal rcri, as shown in Equation (7.10),

1
 rcri = 1− (7.10)
 c2 
  −1
 c 
 
Hence, it is evident from Equation (7.10) that the critical ratio of damaged to
undamaged nodes in the given power network (rcri) is governed by the ratio of vari-
ance (c 2 ) and average degree distribution (c) for the network confguration under
consideration.
When the variance in the degree distribution gets maximised, the grid network
becomes more resilient. It may also be observed that even if average degree con-
nectivity increases without increase in variance, it would not improve the resiliency
of the grid network. On the other hand, when both variance and average degree
connectivity increases, the network resiliency will be more prominent. It is possible
to compute the critical probability associated with grid contingency to predict the
consequences of contingency or attacks on the grid network structure. It is possible
to translate a contingency or an attack into equivalent random failure [31]. This gives

C/

  ° = 
/
˜ c  Pc(c)  
C
(7.11)
160 Power Transmission System Analysis Against Faults and Attacks

Here, ˜ / represents the probability that a particular line leads to a deleted node (bus)
following the removal of a fraction r of nodes (buses). This becomes equivalent to the
random removal of those links (lines) interconnecting the remaining nodes to those
already deleted.
The probability of having a node linked to other nodes can be represented as

P (c ) = exp (−c / ) ˜ (7.12)

˜ is a constant that characterises the exponential distribution corresponding to the

average degree c of the undamaged graph. This results:
/
˙C/ ˘ −C
˜ / = ˇ + 1  e ° (7.13)
ˆ° 

where C / represents the average degree for the damaged graph (network under attack).

7.5 ASSESSMENT OF RESILIENCY FOR TRANSMISSION NETWORK

The electric power transmission system, which is identifed as a critical infrastruc-
ture across the world, possesses all the characteristics of a complex network. With
ongoing smart grid projects implementation and research activities, power systems
have to interact more actively with consumers, internet-based communication net-
works and transportation networks. This essentially transforms it into a “complex
network of complex networks” with failure in one network affecting the entire sys-
tem. Algorithms for fast optimal restoration of system loads are well known and
widely used in industry. Though such restoration strategies are indispensable for con-
tinuity of power supply, self-organising nature of systems may conceal design faws
in the transmission systems which may determine over the long run; for example,
certain transmission system reconfguration may make the transmission system more
vulnerable to cyber terrorism. Complex network analysis of transmission system
resiliency is not to be interpreted as an attempt to improve upon existing (and suc-
cessful) reconfguration strategies; instead, it offers a bigger picture which may aid
in understanding transmission and better operation and planning.
Power transmission systems are characterised by the presence of heavily con-
nected branching nodes, where several buses feed off a long secondary and tertiary
feeder lateral. A terror attack or weather-based damage, aimed at knocking down
those branching nodes, has consequences for the continuity of power supply to
consumers.
In this section, the focus is on the idea that a deletion of a very small fraction of
nodes is enough to disintegrate the network down into islands. The damage extends
from possible limitation of reconfguration algorithms to include these “islands” [33]
to cause a cascading outage leading to blackouts. A complex network analysis of
these blackouts is well documented in [34].
Percolation theory has been found to be useful for studying the behaviour of phys-
ical systems characterised by random confgurations [35]. In context of transmission
Resilience in Power Transmission Network 161

systems, it is appropriate to consider the fow of electrical energy through the net-
work. A statistical physics backed percolation model served as an abstraction of com-
plex networks and connectivity between the nodes [36]. Thus, this theory comes
across as suitable candidate for analysis of resiliency of transmission system. Some
of the defnitions of system resiliency derived from the statistical tools employing
complex network theory [37] are given below:

• A power network is resilient enough to maintain connectivity of one critical

load to the main grid if the probability of random damage of node being
functional in the event of disruptive event is higher than the percolation
threshold for the network.
• A power network is resilient enough to maintain connectivity of one criti-
cal load to the main grid if the second moment of degree distribution of
nodes in the distribution system is greater than twice the frst moment
of degree distribution of the network confguration.

7.6 SIMULATION
To investigate the resilience characteristics of a large transmission system against
any attack or outage, the developed concept has been validated in the part of the east-
ern grid system of India (203 bus, 267 lines and 24 generator electrical grid network)
[38]. The following algorithm illustrates the process of testing grid resilience against
possible attacks and outages.

Step-by-step methodology:

1. The power network has been modelled as an undirected graph. A generation

dispatch process is performed with generation-load balance using optimal
power fow programme employing Newton–Raphson method in grid sys-
tems under consideration. The output exhibits parameters of grid operation
including power fow in the lines.
2. The degree/connectivity, average degree distribution and variance (second
moment of average degree distribution) of the network are determined.
3. The electrical betweenness of the nodes and lines of the network are
determined.
4. With the highest betweenness generator buses being targeted for
attacks, their resilience is analysed by determining the critical fraction
(rcri) of  damaged nodes and probability (ρ c) of the adjacent load nodes
when the MW delivery of these generators are controlled from their full
MW generation till outage (in fve steps). Respective graphs of rcri and
ρ c are plotted showing the pattern of these parameters with decrease in
generation.
5. Next, the highest betweenness lines are analysed. The critical fraction (rcri)
of damaged nodes and probability (ρc) that a specifc link leads to a deleted
node are then calculated following removal of these lines. Again graphs of
rcri and ρc are plotted for the lines after their removal.
162 Power Transmission System Analysis Against Faults and Attacks

In this simulation, in the frst step, the electrical betweenness of the nodes (buses)
and links (lines) are determined. In the following section of the simulation, generator
buses are ranked in descending order of electrical betweenness (Table 7.1). Here the
generator full load ratings (in p.u.) are taken into account, and according to ranking
of the electrical betweenness of these generator buses, the p.u. power ratings are cur-
tailed from full load value till zero (complete outage). The corresponding magnitudes
of rcri and ρc are computed and the obtained values are tabulated in Table 7.1.
It may be observed from Figure 7.2 that following generation drop of the genera-
tors in the transmission system, the magnitude of rcri and ρc reduce gradually indi-
cating drop in resiliency of the grid. The grid becomes vulnerable to outage and
attacks. Interesting to note how the values of rcri and ρc reduce gradually, indicating
the system is falling under poor resiliency. It may be observed that the slope of pro-
fle of rcri and ρc corresponding to generators at Farakka are steepest for the major
portion of operation, indicating that any attack or outage on generators at Farakka
would make the transmission system least resilient. It is also evident that generators
at Bakreshwar follows next. Hence, it is possible to identify specifc generating sta-
tion that would adversely affect the resiliency of the transmission network the most
by observing rcri and ρc profles.

TABLE 7.1
Computational Values of rcri and ρc Following Gradual Generation Drop of
Highest Betweenness Generator Buses in Descending Order
Generator Bus Generation Highest Betweenness Load Bus
Number (p.u.) Number Adjacent to the Generator rcri ρc
1 (FARAKKA) 4 47 (BALURGHAT) 0.64 0.44
3 0.49 0.38
2 0.32 0.21
1 0.28 0.12
4 (BAKRESHWAR) 1.9 188 (SAINTHIA) 0.83 0.72
1.5 0.74 0.36
1 0.68 0.29
0.5 0.6 0.23
3 (KTPS) 1.8 71 (KOLAGHAT) 0.78 0.38
1.4 0.73 0.32
1 0.69 0.29
0.5 0.66 0.23
8 (DPL) 0.9 199 (BOLPUR) 0.64 0.79
0.75 0.60 0.43
0.5 0.58 0.38
0.25 0.55 0.22
13 (CHUKHA) 0.675 114 (COOCHBEHAR) 0.63 0.36
0.5 0.6 0.29
0.35 0.58 0.21
0.2 0.55 0.18
Resilience in Power Transmission Network 163

FIGURE 7.2 Variation of rcri with change in p.u. generation for the respective generating
plants.

Figures 7.3 and 7.4 represent the bar diagrams for rcri and ρc of generators at respec-
tive generating stations arranged from the value of highest to lowest betweenness. It
is evident from the simulation that generators at Farakka have the lowest values of rcri
and ρc indicating their susceptibility to attacks and outages followed by generators
at Bakreshwar.
Next, the resiliency of lines has been studied by ranking the lines according to
electrical betweenness in descending order and then removing those selected lines
the values of rcri and ρc are computed. It has been observed that line number 169, 241,
245, 205, 106, 148 and 42 are the frst seven high betweenness lines. Outage of any or
pair of these lines would make the drop in resiliency for the entire network. Table 7.2
exhibits the respective values of rcri and ρc on removal of each of high betweenness
lines in the given network.
The pattern of the values of rcri and ρc when arranged according to the electrical
betweenness of selected lines are shown in Figures 7.5 and 7.6, respectively. It is
observed that as the electrical betweenness lines of the lines decrease, the value of rcri
and ρc increases. It signifes that for higher values of electrical betweenness the dam-
age is much larger in proportion to those lines with lower electrical betweenness. It is
evident that rcri and ρc values of the lines indicate how much resilient is the network
against tripping of any particular line.
The parametersrcri and ρc depict the pattern of resiliency of a transmission network
taking into account the possible outage of the generators as well as transmission lines
164 Power Transmission System Analysis Against Faults and Attacks

FIGURE 7.3 Bar diagram of rcri for generators arranged from the highest value of between-
ness to the lowest value of betweenness.

FIGURE 7.4 Bar diagram of ρc for generators arranged from the highest value of between-
ness to the lowest value of betweenness.
Resilience in Power Transmission Network 165

TABLE 7.2
Results on Removal of Highest Betweenness Lines
On Removal

Line No. rcri ρc

169 (FKK400-FKK220) 0.23 0.01
241 (FKK400-220-FKK33) 0.28 0.03
245 (117KTPS220-132-KTPS133) 0.32 0.07
205 (BKR15.75-BAKR400) 0.35 0.14
106 (HALD-TAML) 0.42 0.18
148 (HALD-HPCL) 0.47 0.21
42 (CHUKHA-BIRPARA) 0.49 0.23

with descending order of betweenness. This clearly signifes the relation of electrical
betweenness in measuring the criticality of grid resilience. The part of the eastern
grid system of India under study is basically a medium-to-large transmission sys-
tem and the assessment of its resiliency gives an idea of how structural topological
parameters like electrical betweenness are essential to maintain resiliency of trans-
mission systems.

FIGURE 7.5 Bar chart showing the values of rcri for selected lines for higher to lower
betweenness.
166 Power Transmission System Analysis Against Faults and Attacks

FIGURE 7.6 Bar chart showing the values of ρc for selected lines for higher to lower
betweenness.

7.7 SUMMARY
The ability to resist and mitigate stresses, failures, outages or any attacks can be
assessed by assessing the critical fraction of nodes and percolation threshold. These
parameters are observed on removal of lines and buses; hence, resiliency is ascer-
tained in transmission systems. It is possible to identify any specifc generating sta-
tion or transmission line whose removal would amount to the lowest resiliency of the
transmission network.

REFERENCES
1. D.L. Alderson, G.G. Brown, and W.M. Carlyle, “Operational models of infrastructure
resilience”, Risk Analysis, vol. 35, pp. 562–586, 2015.
2. L. Xu, D. Marinova, and X. Guo, “Resilience thinking: A renewed system approach for
sustainability science”, Sustainability Science, vol. 10, no. 1, pp. 123–138, 2015.
3. D. Woods, “Four concepts for resilience and the implications for the future of resilience
engineering”, Resilience Engineering and System Safety, vol. 141, pp. 5–9, 2015.
4. D. Helbing, “Globally networked risks and how to respond”, Nature, vol. 497, no. 7447,
pp. 51–59, 2013.
5. L. Howell, “Global Risks”, Eighth Edition, Switzerland: World Economic Forum, 2013.
6. L. Mili, “Taxonomy of the characteristics of power system operating states”, In 2nd
NSF-VT Resilient Ans Sustainable Critical Infrastructures (RESIN) Workshop, Tucson,
AZ, pp. 13–15, 2011.
Resilience in Power Transmission Network 167

7. P. Kundur, J. Paserba, and V. Ajjarapu, “Defnition and classifcation of power sys-

tem stability IEEE/CIGRE joint task force on stability terms and defnitions”, IEEE
Transactions on Power Systems, vol. 19, no. 3, pp. 1387–1401, 2004.
8. R.J. Campbell, “Weather related power outages and electric system resiliency”,
Congressional Research Service, Library of Congress, USA, 2012.
9. C. Yijia, S. Xiaogang, and C. Ke, “Structural vulnerability analysis of large power grid
based on complex network theory”, Transactions of China Electro Technical Society,
vol. 10, p. 26, 2007.
10. G.A. Pagani and M. Aiello, “The power grid as a complex network: A survey”, Physica
A, vol. 392, pp. 2688–2700, 2013.
11. United States Presidential Policy Directive-21, “Critical infrastructure security and
resilience”, vol. 12, 2013.
12. R.H. Lasseter, “Microgrids”, In IEEE Power Engineering Society Winter Meeting
2002, vol. 1, pp. 305–308, New York, 2002.
13. US Department of Housing and Urban Development, “Hurricane Sandy rebuilding
strategy”, Washington DC, 2013.
14. R.C. Van Den Honert and J. McAneney, “The 2011 Brisbane foods: Causes, impacts
and implications”, Water, vol. 3, no. 4, pp. 1149–1173, 2011.
15. B. Zhou, et al., “The great 2008 Chinese ice storm: Its socioeconomic-ecological
impact and sustainability lessons learned”, Bulletin of the American Meteorological
Society, vol. 92, no. 1, pp. 47–60, 2011.
16. C.S. Holling, “Resilience and stability of ecological systems”, Annual Review of
Ecology and Systematic, vol. 4, pp. 1–23, 1973.
17. C.S. Holling, L.H. Gunderson, and S. Light, Barriers and Bridges to the Renewal of
Ecosystems, New York: Columbia University Press, 1995.
18. B. Walker, S. Carpenter, J. Anderies, N. Abel, G.S. Cumming, M. Janssen, L. Lebel, J.
Norberg, G.D. Peterson, and R. Pritchard, “Resilience management in social-ecological
systems: A working hypothesis for a participatory approach”, Conservation Ecology,
vol. 6, no. 1, p. 14, 2002.
19. J.M. Kendra and T. Wachtendorf, “Elements of resilience after the world trade center
disaster: Reconstituting New York City’s Emergency Operations Centre”, Disasters,
vol. 27, no. 1, pp. 37–53, 2003.
20. G. Brown, M. Carlyle, J. Salmerón, and K. Wood “Defending critical infrastructure,”
Interfaces, vol. 36, no. 6, pp. 530–544, 2006.
21. D.A. Reed, “Electric utility distribution analysis for extreme winds,” Journal of Wind
Engineering and Industrial Aerodynamics, vol. 96, no. 1, pp. 123–140, 2008.
22. G.P. Cimellaro, A.M. Reinhorn, and M. Bruneau, “Framework for analytical quantifcation
of disaster resilience,” Engineering Structures, vol. 32, no. 11, pp. 3639–3649, 2010.
23. T. Aven, “On some recent defnitions and analysis frameworks for risk, vulnerability,
and resilience,” Risk Analysis, vol. 31, no. 4, pp. 515–522, 2011.
24. D. Henry and J.E. Ramirez-Marquez, “Generic metrics and quantitative approaches for
system resilience as a function of time,” Reliability Engineering and System Safety, vol.
99, pp. 114–122, 2012.
25. E. Hollnagel, D.D. Woods, and N. Leveson, “Resilience Engineering: Concepts and
Percepts, Farnham: Ashgate Publishing Ltd., 2007.
26. E. Zio, “Reliability engineering: Old problems and new challenges”, Reliability
Engineering and System Safety, vol. 94, no. 2, pp. 125–141, 2009.
27. D. Li, Q. Zhang, E. Zio, S. Havlin, and R. Kang, “Network reliability analysis based on
percolation theory”, Reliability Engineering and System Safety, vol. 142, pp. 556–562, 2015.
28. H. Ahmadi and J.R. Marti. “Mathematical representation of radiality constraint in dis-
tribution system reconfguration problem”, International Journal of Electrical Power
and Energy Systems, vol. 64, pp. 293–299, 2015.
168 Power Transmission System Analysis Against Faults and Attacks

29. M. Molloy and B. Reed, “A critical point for random graphs with a given degree
sequence”, Random Structures and Algorithms, vol. 6, no. 2–3, pp. 161–180, 1995.
30. R. Cohen, K. Erez, D. Ben-Avraham, and S. Havlin, “Resilience of the Internet to ran-
dom breakdowns”, Physical Review Letters, vol. 85, no. 221, p. 4626, 2000.
31. R.V. Sole, M. Rosas-Calas, B. Corominas-Murtra, and S. Valverde, “Robustness
of the European power grids under intentional attack”, Physical Review E, vol. 77,
pp. 026102–26109, 2008.
32. J.H. Grisi-Filho, R. Ossada, F. Ferreira, and M. Amaku, “Scale-free networks with
the same degree distribution: Different structural properties”, Physics Research
International, vol. 2013, pp. 2090–2220, 2013.
33. C. Wang, S. Ruj, M. Stojmenovic, and A. Nayak, “Modelling cascading failures in
smart grid using interdependent complex networks and percolation theory”, IEEE
8th Conference on Industrial Electronics and Applications (ICIEA), pp. 1023–1028,
Melbourne, 2013.
34. P. Crucitti, V. Latora, and M. Marchiori, “Model for cascading failures in complex net-
works”, Physical Review E, vol. 69, p. 045104, 2004.
35. R. Pastor- Satorras and A. Vespignani, “Evolution and Structure of the Internet: A
Statistical Physics Approach”, Cambridge: Cambridge University Press, 2007.
36. Y. Zhang and L. Guo, “Network percolation based on complex network”, Journal of
Networks, vol. 8, no. 8, pp. 1874–1881, 2003.
37. S. Chanda and A.K. Srivastava, “Quantifying resiliency of smart power distribution
systems with distributed energy resources”, 2015 IEEE 24th International Symposium
on Industrial Electronics (ISIE), Rio de Janeiro, Brazil, 2015.
38. K. Chakrabarti, “Investigation on voltage security of power transmission systems using
soft computing techniques”, PhD Thesis, BESU Shibpur, 2012.
 8 Effect of Distributed
Energy Sources

8.1 INTRODUCTION
Distributed renewable energy sources (like wind turbine, photovoltaic, fuel cell, bio-
mass, smart house, etc.) and energy storage devices (like battery, electric double layer
capacitor, superconducting magnetic energy storage, etc.) are the future of energy
demand [1]. Distributed generation (DG) located close to the load, that is, on the
distribution network or on the customer side of the metre, have great potential to
improve distribution system performance and should be encouraged.
DG [2] refers to small electric power generators, typically ranging in capacity
from 1W to 300 MW, which can be located on the utility system, at the customer site
or at a location not connected to the grid. DG can be conventional, such as combined
cycle turbines, small diesel generators, combustion turbines or wind turbines, solar
generation and other renewable energies, which will be discussed in detail in this
chapter.
Rating of DGs: The maximum rating of the DG which can be connected to a
distributed generation depends on the capacity of the distribution system that is inter-
related with the voltage level of the distribution system. Hence, the capacity of DGs
can vary widely. There are four different categories of DGs [1]:

Micro: DG range: ~1W<5kW;

Small: DG range: 5 kW<5MW;
Medium: DG range: 5 MW<50MW;
Large: DG range: 50 MW<~300MW.

Because of the various types of DGs, the generation electric current can be either
direct current (DC) or alternating current (AC). Photovoltaic, fuel cell and batteries
generate DC which is appropriate for DC loads and DC SG (Smart Grid). On the
other hand, the DC can be converted to the AC using power electronics interface,
which can be connected to the AC loads and power grid. Other DGs such as wind
turbine, micro turbine and biomass deliver an AC, which for some applications must
be controlled using modern power electronic equipment to acquire the regulated
voltage [1].

8.1.1 SOLAR PHOTOVOLTAIC ENERGY

The basis of solar photovoltaic energy conversion lies in the concept of how light
energy is directly converted to electrical energy. Sunlight consists of packets of
energy, also known as photons. This photon inherently possesses energy, and when

169
170 Power Transmission System Analysis Against Faults and Attacks

it falls on PV cell it can liberate electrons. The energy of photons liberates electrons
from the valence band to the conduction band. and creates electron hole pair. One
photon is responsible for one electron-hole pair. The energy of photon depends on the
wavelength of the light.
hc
E=
˜
where h = Planck’s constant, c = speed of light and λ = wavelength.
The energy of photon must be equal to the energy gap of that material.
The effectiveness of PV conversion depends upon many factors:

• Surface properties like absorptivity and refectivity

• Rate of charge generation and charge recombination within semiconductor
material

In general, silicon (energy gap is 1.107 eV) is used as the material of PV cell but gal-
lium arsenide (GdAs), indium phosphide (InP), cadmium telluride (CdTe), etc. can
also be used.
A solar cell has limited open circuit voltage and current delivering capability.
To increase its power rating, they are connected in suitable series-parallel com-
bination in an enclosed chamber, known as a solar module or solar panel. In a
similar manner, solar modules are connected on load-specifc application-based
series-parallel combination to form a solar array. A solar panel is rated by its DC
output power at a certain voltage at SOCT(standard operating cell temperature).
At a higher temperature, the effciency of solar cell decreases; hence, it is recom-
mended to provide better ventilation in any solar module. During the formation of
a solar array, a designer should care about mismatch loss due to interconnection of
panels having different properties. It is an undesired phenomenon as it can create
a localised circulating power fow reducing the desired output of a solar cell to a
considerable extent.
Another problem regarding PV operation is shading. When one solar cell is cov-
ered or shaded by any means such as a leaf or a shadow, the covered portion heats
up to such an extent that the cell or module concerned may get damaged. This defect
is called a hot spot. A shaded cell does not produce electricity but consumes current
from the sound ones. Hence, the current from the string is converted into heat. To
protect a module from hot spot, bypass diodes are used. Their function is to pro-
vide an alternating path to the current concerned in shaded cell condition. To avoid
most of the unwanted situation and to maximise solar insolation, solar PV-based
systems are always installed in a place where the chances of shadow are a minimum.
Even in array, the distance between the rows of solar modules are chosen such that
the shadow of one module do not fall to any module on the next row. In Northern
Hemisphere, it is recommended that any solar installation should face south at an
elevation angle equal to the latitude of the concerned place.
For a reliable power supply through solar PV method, the following components
are used. However, depending on specifc application other components like inverter
might be used.
Effect of Distributed Energy Sources 171

i. PV module
PV module consists of solar cells arranged in a suitable fashion to suit for
the voltage or current demand. Several modules are connected in appropri-
ate series-parallel combination to increase the power range.
ii. Storage element (Battery):
In general, an off-grid system requires a storage system for utilising sun’s
energy in absence of sun light during the night. A battery is a good storage
choice in such a case. The capacity of the battery depends upon the esti-
mated load. The problem regarding battery storage system is that the life
period of a battery is much less than a solar module, so proper designing
involves choosing the proper capacity of battery from an economical view-
point. During operation, it has to be kept in mind that the battery should not
reach its threshold discharge level, otherwise its lifetime will be shortened.
For high-rating applications, several batteries are used in suitable series-
parallel connection known as battery bank.
When a battery is a part of balance of system for PV-based supply is
delivering current, it is not allowed to discharge to its full capacity. If this
should be the case (i.e. battery is allowed to discharge fully), then its life-
time will be much shortened, and due to the high cost of battery, it becomes
an uneconomical approach. To counter this problem, battery is allowed to
discharge to a certain level (say 40%) depending on the type and manu-
facturer of the battery. This is known as Depth of Discharge (DOD). To
maintain the DOD to the proper value, a charge monitoring unit, known as
charge controller is used.
iii. Charge controller:
Basically, charge controller is a protective measure taken to protect the
battery. The functioning of a charge controller is to protect the battery from
undercharge or overcharge protection. In either of the two cases, this pro-
tecting device cut outs the load relieving the battery. Charge controllers
might not be used in small-scale applications.
iv. Inverter:
The output of a solar photovoltaic system is always DC; however, most of
the electrical systems encountered today is AC in nature. To encounter this
problem an inverter is used at the end terminals of a PV system. In absence
of an inverter such as for lighting or to run a fan separate line is laid out with
different types of light or DC fan. Moreover, for a grid connected system,
an inverter is essential.

Responsibility of a designer is not only limited for proper load matching and estimat-
ing the capacity or rating of different components. The designer should equally focus
on the economics of the designed system. The target is to achieve a cost-beneft sys-
tem. To avoid any tussle between these two contradictory requirements, the designer
should choose an optimal design approach.
The cost of a PV system generally involves cost of it components such as
PV array and its size, battery bank size and other component costs, which are
application-specifc.
172 Power Transmission System Analysis Against Faults and Attacks

Solar PV system can be generalised as generator side which consists of mainly PV

array and accessories and balance of system, which includes load and other convert-
ing equipment (if any).Generator side consists of solar PV cell, modules or arrays and
the remaining things are popularly known as balance of system. The power of mod-
ule is generally given in Watt-peak or Wp, which is the amount of power generated by
solar PV cell at 1,000 W/m2 intensity at 25°C temperature at a air mass level of 1.5.
The designing process can be analysed as follows:

a. Estimation energy demand and numbers of module required: The frst

step involves calculating the energy required for a particular load profle.
This involves calculating the total energy required for any particular day.
Second, the total number of modules required is calculated based on the
power rating of the module concerned. When estimating the size of a mod-
ule, changes from the standard test condition to practical condition is done.
b. Inverter sizing (if required): The output of a solar PV-based system is DC
in nature. But nowadays generally all types of load are alternating in nature.
To counter this problem, an inverter may be installed depending on the situ-
ation. Generally, the inverter size should be 25%–30% higher than the total
power demand. In case any highly inductive machine operation is required
in the load profle (such as compressor or motor), the inverter size should be
minimum thrice the capacity of the total power demand to tackle the higher
starting current. The input rating of the inverter should be the same as the
PV array system for safer operation.
c. Battery sizing: Generally, the battery used for solar PV application is deep-
cycle type. The reason for using this special type of battery is that they can
discharge to a very low level. The capacity of the battery should be suf-
fcient to supply load during the absence of sunlight. Suffcient measures
are taken to consider battery losses (if any) and DOD. The battery voltage
should match with the system voltage. Hence, required number of batteries
are connected in series to provide the given voltage level. Similarly, to meet
any specifc demand, batteries are connected in parallel. This series-parallel
combination is known as a battery bank.
Apart from the above items, few other items may be required on spe-
cifc applications. For example, a charge controller may be required in few
applications. The size of the charge controller depends on the current to
be handled. But to provide better protection, that is, protection from short-
circuit, the capacity of charge controller is chosen according to a factor
(1.3–1.5) multiplied with the specifed short-circuit current capability of the
concerned system.

8.1.2 WIND ENERGY

Wind is the horizontal movement of air which arises due to uneven heating of the
earth and derives its energy from solar radiation. At daytime, air above land masses
heats up faster than air over the sea, oceans or any water bodies. Hot air expands and
rises while cool air cools more rapidly over landmasses than water off shore land,
Effect of Distributed Energy Sources 173

causing land and sea breezes. Due to the shape of the earth, Equator receives the
maximum heat causing winds to blow from subtropical belts towards the Equator. In
addition, the axial rotation of the earth induces a centrifugal force which has a thrust
on equatorial air masses to the upper atmosphere causing defection of the winds.
The concept of harnessing wind energy dates back to Egyptian civilisation where
they used wind power to sail their boats in the Nile River [3].Skilful technicians
from the Middle East introduced windmills to China. The technology to harness
wind energy reached Western Europe by the Arabs. Years have passed since then and
wind power has emerged as one of the most used renewable energy across the world.
After sudden price rise in fossil fuels, a number of countries were stimulated towards
generation of power using wind energy.
Wind turbines extract energy from wind stream by converting kinetic energy of
the wind to rotational motion required to operate an electric generator. By virtue
of kinetic energy, the velocity of the fowing wind decreases. It is assumed that air
masses which passes through rotor is only affected and remains separate from the
air which does not pass through the rotor. As the free wind (stream) interacts with
the turbine rotor, the wind transfers part of its energy into the rotor and the speed
of the wind decreases to a minimum leaving a trail of disturbed wind (also called
wake). The variation in velocity is considered to be smooth from far upstream to far
downstream. The wind leaving the rotor is below the atmospheric pressure (in wake
region) but at far downstream it regains its value to reach the atmospheric level. The
rise in static pressure is at the cost of kinetic energy, consequently, further decreas-
ing wind speed. Because wind fow is considered incompressible, air stream fow
diverges as it passes through the turbine. Also, the mass fow rate of wind assumed
constant at far upstream, at the rotor and at far downstream.
Basically, a wind turbine operates by slowing down the wind and extracting a part
of its energy in the process. In general, wind turbines have blades, sails or buckets
fxed to a central shaft. The extracted energy causes the shaft to rotate. This rotat-
ing shaft is used to drive a pump, to grind sees or to generate electric power. Power
extraction by wind turbine depends on variation of wind turbine power with rotor
diameter and wind speed. Mean wind speed, energy estimation and power density
duration curve also play equally important roles.
Wind turbines are classifed as horizontal-axis turbines or vertical-axis turbines
depending upon the orientation of the axis of rotation of their rotors. Wind turbines
are further classifed into “Lift” and “Drag” type.
Although wind energy is the frst among renewable energy to become an econom-
ically viable source, it is highly unpredictable depending on season, elevation of the
land and wind characteristics like frequency of distribution of wind speed. Hence,
assessment of wind energy potential has to be evaluated by recording meteorological
parameters such as wind direction, wind speed temperature and rainfall on hourly
basis. This data is collected from surveys and historical data are used for forecasting.
For correct forecasting, the wind energy at a particular site is decided by the wind
profle or regime (wind speed frequency distribution) of the site that can be math-
ematically described by Weibull Probability Distribution Function.
Many wind energy generators (WEGs) are commercially available in the capac-
ity of 1 kW to 3 MW. It is necessary to select the best-suited WEG for a particular
174 Power Transmission System Analysis Against Faults and Attacks

site for generating maximum energy. Production of electricity depends on mean

wind speed, hub height, cut-in speed, rated and furling wind speed of the machine.
Grid interfacing of a wind farm is also done. It can be integrated with state grid to
ensure a smooth supply of continuous power. There are limitations on the integra-
tion of wind turbines to the grid system. Pumping the generated electricity from
WEGs to state grid should have the minimum power quality impact on the grid
so that integration should contribute to the improvement in power quality. Grid
short-circuit power and grid short-circuit ratio have a great infuence on the power
quality of the grid.
A wind farm has several identical wind turbine generators. These are induction-
type generators which need reactive power for magnetising. With conventional energy
system, generators besides supplying active power, supply reactive power required by
consumers to operate their electrical equipment. However, WEGs (induction type)
require reactive power to start power generation. To have availability of reactive
power, each WEG is provided with shunt capacitors. These capacitors meet reactive
power requirement of WEG and maintain power factor at the rated value of 0.95.
Induction generators can be used in both stand-alone and grid-connected mode with
advantages such as low cost and robust construction. When connected with grid sys-
tems, WEG draws reactive volt-ampere from the grid. This VAR drains on the grid
system which is compensated by the use of terminal capacitors.

8.1.3 SMALL HYDRO POWER

Falling water as a source of energy is known from old days where it was used to
turn water wheels for grinding corn. Electric power is generated when water from a
high altitude is made to fow through hydraulic turbines. Hydropower projects essen-
tially harness energy from fowing or falling water in rivers, rivulets, artifcially cre-
ated storage dams or canals. Potential energy in water is converted into shaft work
utilising a hydraulic prime mover. The hydraulic turbine converts potential energy
of water or kinetic energy of fowing stream into mechanical energy by its rotating
shaft. Electrical energy is obtained from an electric generator coupled to the shaft
of the prime mover. It depends mainly on the head of the water and the discharge
through the turbine. Hydropower projects of ratings <10 MW are regarded as small
hydropower resources which are extremely favourable.

8.1.4 BIOMASS AND BIOGAS ENERGY

Biomass is organic or carbon-based matter (resources from forest, agriculture, aqua
culture and residue from industry and urban waste) that can either react with oxygen
for combustion or undergo metabolic process to release heat. Biomass can be used
in original form or can be transformed to convenient and useful form of solid, liquid
or gaseous. Biofuels can be fuel wood, charcoal, fuel pellets, bio-ethanol, biogas,
producer gas and biodiesel. Biomass conversion technologies include densifcation,
combustion, incineration, thermochemical (pyrolysis, gasifcation and liquefac-
tion) and biochemical conversion (anaerobic digestion and ethanol fermentation).
Biomass gasifcation can be done using gasifers like fuidised bed gasifer.
Effect of Distributed Energy Sources 175

Biogas is a gaseous fammable fuel (mainly composed of methane, carbon diox-

ide, hydrogen, hydrogen sulphide and other gases) which is obtained from biomass of
an agricultural digestion or fermentation of wet organic matter. Biogas is produced
through biological conversion process that involves anaerobic digestion proceeding
subsequently as hydrolysis, acid formation and methane formation.
Rapid urbanisation and industrialisation has resulted in enormous quantity of
waste in urban and industrial areas. Municipal solid waste, municipal liquid waste
and urban industrial wastes that are disposed are also taken for energy conversion.
Biomass and biogases are mainly used to power cooking and heating purposes pri-
marily in rural areas.

8.1.5 TIDAL ENERGY

Tidal energy is another method of harnessing hydropower. Tidal power possesses a
scope from power generation viewpoint in the coming future. Tide forecasting is more
reliable than that for wind or solar. High initial cost and limited availability of sites hav-
ing satisfactory tidal range has limited the scope of tidal energy extraction technique. At
present, continuous research on plant and turbine design may prove to be economical.
Tides are formed by moon’s gravitational pool on the oceans and at the estuar-
ies of large rivers on the earth. The sun’s pool in similar fashion to the moon plays
a minor role in this case. Because of the strong attraction of the celestial bodies to
the oceans, a bulge in the water level is created, causing a temporary increase in sea
level. Consequently, sea level rises causing water from the middle of the ocean to be
moved towards the shore areas, thereby creating a tide. This takes place in an unfail-
ing manner because of the consistent pattern of the changing positions of the moon
and sun relative to the earth.
During the rise and fall of sea level, the tidal generation system can be activated
with proper gearing mechanism where necessary. The potential energy created by
the pressure difference caused due to the water level difference across water passages
and barrages is converted to kinetic energy in the turbines. This turbine is coupled
mechanically with proper generator to produce electricity.
A tidal power plant consists of the following components:

a. Dam
b. Sluice gates from the basins to the sea
c. Power house

A dam is an artifcial barrier between the seawater and the site of interest or basin.
The primary function of a dam in tidal power plant is to absorb shock and pressure of
wave. The dam provides proper channelling for turbine operation. Tidal barrages are
constructed at suitable places where the head of water are satisfactory which is ~3 m.
Choice of proper site for constructing a dam is a very diffcult task because the energy
available is related to the size of the basin (the area from dam to plant, precisely to
turbine) and to the square of tidal range.
From the above discussion, it is clear that a dam should be built near to the estu-
ary or bay. But the problem will be that with larger size of basin, the effective tidal
176 Power Transmission System Analysis Against Faults and Attacks

range gets reduced. So, a balance must be maintained between these two contradic-
tory criteria. Gates are used in tidal plant to fll and empty the basin as and when
required. Due to the interaction with seawater, gates are prone to corrosion. This can
be tackled by cathodic protection while constructing the gate. Power house is housed
with the turbines, generators and other auxiliary instruments. Because of the small
head, turbines of large sizes are required in this type of plants. Alternator is directly
coupled to turbines. Generally, turbines used in hydro turbines with low head opera-
tion may be used in such a plant with proper modifcations wherever necessary.

8.1.6 OCEAN THERMAL ENERGY CONVERSION (OTEC)

OTEC is a process that can produce electricity using the temperature difference
between deep cold ocean water and warm tropical surface waters. OTEC plants
pump large quantities of deep cold seawater and surface seawater to run a power
cycle and produce electricity. OTEC is frm power (24/7), a clean energy source,
environmentally sustainable and capable of providing massive levels of energy.
Recently, higher electricity costs, increased concerns for global warming and a
political commitment to energy security have made initial OTEC commercialisation
economically attractive in tropical island communities, where a high percentage of
electricity production is oil based. Even within the United States, this island market
is very large; globally, it is many times larger.

8.1.7 GEOTHERMAL ENERGY

Geothermal energy is the energy present in the interior of the earth. Geothermal
energy can be extracted from earth’s interior in the form of heat. Volcanoes, geysers
and hot springs are visible signs of large amount of heat lying in earth’s interior. The
geothermal energy from earth’s interior is almost inexhaustible; although the amount
of thermal energy within the earth is very large, useful geothermal energy can be
extracted at only certain suitable sites [4].
Geothermal resources are of fve types: hydrothermal (hot water and wet steam),
vapour-dominated resource, hot dry rock resource, geo-pressured resource and
magma resource. The hydrothermal type and hot dry rock type are the most used.
Hydrothermal plants generally have common features consisting of (i) produc-
tion well to extract steam from the resource,(ii) a centrifugal separator to remove
solid matter from the steam,(iii) a turbine to convert thermal energy to mechanical
energy,(iv) a generator coupled to the turbine to generate electric power,(v) a con-
denser to condense wet steam exited from the turbine into water by direct contact
with cooling water and (vi) a cooling tower to cool warm water exited from the con-
denser and returning the cooled water to the condenser.
In case of hot dry rock type, injection for wells pumping inside and production
or extraction wells for hot water pumping out are drilled. A series of injection and
extraction wells can be drilled to tap a suffcient amount of geothermal energy. The
hot water extracted from the man-made reservoir is made to vaporise low boil-
ing point refrigerant which is used to run a turbine coupled with a generator. The
refrigerant vapour exiting the turbine is condensed in a condenser which is pumped
Effect of Distributed Energy Sources 177

into the heat exchanger again. The viability S of extracting energy from a dry feld
depends upon the degree to which the resource feld can be fractured to develop man-
made geothermal reservoirs.

8.1.8 MHD GENERATION

The magneto hydrodynamic (MHD) power generation technique is a direct method
of energy conversion which converts heat energy directly into electrical energy,
without any intermediate mechanical energy conversion, as is prominent in other
power plants. Therefore, in this process, substantial fuel economy can be achieved
due to the elimination of producing mechanical energy and then converting it to
electrical energy.
The concept of MHD power generation was introduced for the very frst time by
Michael Faraday in the year 1832 in his Bakerian lecture to the Royal Society. He
conducted an experiment at the Waterloo Bridge in Great Britain for measuring the
current from the fow of the river Thames in earth’s magnetic feld. This experiment
outlined the basic concept of MHD generation over the years then, and signifcant
research work has been conducted on this topic. On August 13, 1940 this concept
of MHD power generation was imbibed as the most widely accepted process for
the conversion of heat energy directly into electrical energy without a mechanical
sub-link.
The principal of MHD power generation is very simple and is based on Faraday’s
law of electromagnetic induction, which states that when a conductor and magnetic
feld moves relative to each other, voltage is induced in the conductor. As the name
implies, MHD generator is concerned with the fow of a conducting fuid in the
presence of magnetic and electric felds. In conventional generator or alternator, the
conductor consists of copper windings or strips, while in an MHD generator, the hot
ionised gas or conducting fuid replaces the solid conductor. A pressurised, electri-
cally conducting fuid fows through a transverse magnetic feld in a channel or duct.
A pair of electrodes are located on the channel walls at right angle to the magnetic
feld and connected through an external circuit to deliver power to a load connected
to it. Electrodes in the MHD generator perform the same function as brushes in a
conventional DC generator. The MHD generator develops DC power and the conver-
sion to AC is done using an inverter.

8.1.9 ENERGY STORAGE

At times the nature of the load is so fuctuating that there might be some instances
when there is a mismatch between the generation and requirement. This problem can
be met with the help of energy storage [5]. In other words, storage keeps energy in
any suitable form within itself until it is required in the same or any other form. The
storage element stores energy when generated amount is excess than the requirement
and delivers the same when the demand is more than the generated amount. In case
of electricity, it can be stated that the output in excess of load demand from any solar
PV plant or any windmill is fed to any grid system. Otherwise, a storage element is
required.
178 Power Transmission System Analysis Against Faults and Attacks

Basically, storage plays an important role in economic power generation. The

main aim for combined operation of power plant is to minimise the operating cost
with maximised output. This can be effectively done with the help of a storage sys-
tem. The plants are operated optimally. The excess energy generated from this opti-
mal schedule is stored in storage system which may be utilised during demanded
period.
An important point to note on energy storage is that the nature of energy stor-
age element is different in case of renewable energy resources than fossil fuel-based
resources. This is because extraction rate of energy from most of the renewable
energy-based resources (as an example we can think about a solar PV plant) is beyond
human control unlike fossil fuel-based energy extraction techniques. This diffculty
can be overcome by matching the load to the energy supply which is a very diffcult
task. Another option is to store energy for use in case of emergency.

8.2 ROLE OF DISTRIBUTED ENERGY SOURCES IN

POWER TRANSMISSION NETWORKS
Although frequent occurrences of blackouts have been common since past few
decades, in most of the studies, DG is not considered for vulnerability analysis of
power grids. Actually, in many existing power grids, some load nodes are far away
from the generators, so it is very diffcult for these nodes to acquire enough power
supply. In this case, the adoption of distribution generation technology will be a good
choice. DG can reduce long-distance transmission and balance the power fow, mak-
ing the power grid more reliable. In addition, DG can reduce the energy consumption
and cost, enhancing both reliability and fexibility of power grids. Of course, failure
of some part of the power grid with DG can also make the power grid become vul-
nerable, so the vulnerability assessment of power grid with DG is also equally related
to this concern.

8.2.1 DGS IN RESTRUCTURED ENVIRONMENT

The DG placement problem can be formulated as an optimisation, with similari-
ties to the optimal power fow problems. However, in practice, a network operator
may only have a limited infuence on the actual location of DG because this often
depends on factors such as site availability, as well as construction and planning
permission issues.
A lot of work has been done in the area of DG placement in the distribution
network [6–8]. Among the few studies on the placement of DGs in the transmis-
sion system, [9] discusses the impact on grid dynamics if the grid is powered using
DGs. They also assert that different dynamics can be seen on the grid by varying
the fractions and distributions of DGs in the transmission system. They also mention
that improper distribution and sizing of the DGs may lead to increased vulnerability
of the grid instead of increased robustness. The stability and reliability of the grid
improves with the use of conventional DGs close to the loads, but if an increas-
ing penetration of stochastic renewable energy sources is present in the grid, these
Effect of Distributed Energy Sources 179

energy sources introduce erratic power inputs into the grid, thereby causing it to fail
with a sharp transition [10].
The interconnection of several DGs in the transmission system causes a
structural change in the grid, and, therefore, the properties of the system would
change. Indices, such as characteristic path length, degree and degree distribu-
tion, clustering coeffcient, and betweenness, decide how the interconnection of
DGs aspects topological characteristics of the grid. They have also used weighted
graph indices and have suggested new indices based on structure and operational
conditions of the grid for the evaluation of structural properties of the grid with
incorporated DGs.

8.2.2 APPLICATIONS OF DGS

There are several applications of DG in the power system [10]:

• The DG can be scattered in different places, and can be utilised as a standby

power source. If the grid power cuts off the sensitive loads, for example,
process industries and hospitals, the DG can provide the emergency power
for these loads.
• The DG can supply power for geographically isolated communities which
are diffcult to connect to the main power grid. Therefore, the DG can
improve the economic condition for isolated communities.
• The electric power cost depends on the electric load. When the load demand
is high, the electric power price will be high and vice versa. The DG can sup-
ply the electric power to the load when the demand is high. Consequently,
the customer can reduce the electricity cost to pay time-of-use rates.
• The DG can supply power for the rural and remote applications which
include lighting, heating, cooling, communication and small industrial
processes.
• Individual DG owners are usually used as a base load to provide part of the
main required power and support the grid by enhancing the system volt-
age profle. The DG also helps to reduce the power losses and improve the
system power quality.

DG affects the operation of the distribution networks, including power fows and
voltages [1]. DG also has an effect on system losses. In many cases, the effect of DG
on losses is positive as DG is often located close to the demands where the energy
is consumed. This reduces the distance over which energy needs to be transported,
therefore reducing losses.
DG can also impact the reliability and quality of the power supply in the electri-
cal system. Previous studies have demonstrated some positive effects of DG on reli-
ability, where DG can reduce the loading on the network at critical times, or supply
part of the demand in the network during faults and shortages. However, DG can
also have a negative impact on reliability, particularly in cases where it has caused
problems for the coordination of the network protection systems [10].
180 Power Transmission System Analysis Against Faults and Attacks

8.3 MITIGATION STRATEGY EMPLOYING

DISTRIBUTED ENERGY SOURCES
The use of DG in the sub-transmission and distribution system can be used to miti-
gate cascading failures causing large-scale blackouts. Despite the fact that DGs have
their own arguments, their emphasis increases continuously because of their advan-
tages like providing local power, thus inhibiting power transfer for long distances,
which, on the other hand, improves grid reliability. This chapter focuses on alloca-
tion of DGs in the power networks. Renewable energy sources can be treated as a
separate area of study. The placement of DGs is effected on the basis of relative
criticality of nodes in the system employing the tool of electrical centrality which
gives electrical node signifcance. The allocation of DGs strengthens the robustness
of the grid network and obstructs cascading failures. Vulnerability analysis using
power fow model is done prior to the allocation and post allocation of the DGs. The
interaction of certain amount of DGs in a power network causes a topological change
in the network, and, therefore, to what extent the properties of the network would
change is also determined by assessing Grid Vulnerability Index (GVI).

8.3.1 NETWORK CRITICALITY ASSESSMENT USING

[ZBUS] CENTRALITY EMPLOYING DGS
The sub-transmission system has been receiving attention from complex network
theory researchers for some time. The metric of electrical centrality which can be
useful for installation of DG in distribution system and discussed in Chapter 4 dif-
ferentiate the electrical topology and physical topology of power grid. This measure
not only enables complex network analysis of power systems but is also more appro-
priate for the power grid than general topological analysis. This measure is based on
Zbus impedance matrix of a power system which fnds more electrically central nodes
in the system which is supposed to be highly connected to most other nodes for the
placement of distributed generators in the distribution system.

8.3.2 INCORPORATION OF DISTRIBUTED ENERGY SOURCES IN CRITICALITY

ASSESSMENT USING BETWEENNESS METRIC: SIMULATION
FOR DG INCORPORATION IN CRITICALITY ASSESSMENT OF
IEEE 57 BUS SYSTEM USING BETWEENNESS METRIC
The IEEE 57 bus system [11] has been considered to apply the concept presented here
to study electrical betweenness of buses and check the ranking of their criticality.
Figure 8.1 represents the topological network of IEEE 57 bus system.
Figure 8.2 represents the graphical plot of criticality magnitudes as obtained from
the measure of betweenness of load buses. The electrical betweenness of load bus 11
being the maximum; it is followed by the load bus numbers 13, 16, 10. Thus, the most
critical load bus is bus number 11 and it is vulnerable against any unplanned outage
and attack. Following incorporation of DGs in the most critical load bus (i.e. bus 11),
the criticality magnitude for this bus reduces. Thus, application of DGs reduces the
Effect of Distributed Energy Sources 181

FIGURE 8.1 Complex network topology of IEEE 57 bus.

FIGURE 8.2 Electrical betweenness of load buses in IEEE 57 bus system.

182 Power Transmission System Analysis Against Faults and Attacks

criticality and ensures lesser vulnerability of most critical bus following unplanned
outage or attack on the power network.
Figure 8.3 exhibits the decrement in magnitude of electrical betweenness of four
successive load buses (which have high level of betweenness) following incorpora-
tion of DG at most critical load bus (i.e. bus 11 which has the highest magnitude of
betweenness).
Load fow study was carried out after incorporation of DG to evaluate the losses.
DG incorporation resulted in improved effciency along with reduced losses and
raised the voltage profle. It has been observed that the steady state voltage profle
(Figure 8.4) of load buses improve with incorporation of DG while the power loss in
the network reduce from 28.12 to 24.32 MW.
Hence, employment of DG unit not only reduces betweenness (criticality) but
at the same time decreases system vulnerability, improves power transmission and
increases voltage profle making the grid more effcient. The role of DG penetration
in improvement of complex network theory parameters is an important tool which
can be employed to strengthen the infrastructure of power grid.

FIGURE 8.3 Magnitude of electrical betweenness of four successive load buses.

Effect of Distributed Energy Sources 183

FIGURE 8.4  Voltage magnitude enhancement with implementation of distributed generation.

8.3.3 Vulnerability Assessment Using GVI Metric Incorporating

DER: Simulation for Vulnerability Assessment of IEEE 57
Bus System Using GVI Metric on Incorporation of DG
To identify the effective locations to install DG, the power supplying efficiency of
each load node needs to be calculated. The determination of shortest path lengths
among all of the generators and load nodes is the prerequisite of such calculation.
Using the expression of Equation 5.3 (refer to Chapter 5) the power supplying effi-
ciencies of 35 load nodes in the IEEE 57 bus system have been obtained. It has been
observed that the ten buses (nodes) having the minimum power supplying efficien-
cies are the buses numbered as 31, 33, 30, 25, 32, 35, 47, 54, 53, 27 (the power sup-
plying efficiencies of these nodes are 0.1, 0.12, 0.32, 0.36, 0.55, 0.59, 0.63, 0.77, 0.79,
and 1.09 respectively). In this simulation, the five load nodes having the lowest power
supplying efficiencies have been identified. These load nodes have been numbered
as 31, 33, 25, 47, 54 for which the DG installation is recommended. The size of each
DG is assumed to be 10MW. With incorporation of new generating buses, the bus
numbers are modified as 58~62 and the entire network with (12 generation nodes,35
load nodes and 15 transmitting nodes) 62 nodes has been renumbered. In efficiency-
based vulnerability index calculation, e/ denotes the falling ratio of power supplying
efficiency. With larger value of e/the power supplying efficiency declines, hence the
related buses or lines will be more vulnerable (Tables 8.1 and 8.2).
The GVI values of the network prior to allocation and post allocation of DGs are
shown in Table 8.3.
It has been observed that the GVI values drastically reduce following installation of
DG in the designated load buses. Obviously, this reduces the vulnerability of the grid
network. In addition to assessment of vulnerability of the system following installation
184 Power Transmission System Analysis Against Faults and Attacks

TABLE 8.1
Vulnerability of Some Nodes of the IEEE 57 Nodes
Systems without DG under Attacks on Nodes
Node e
8(G) 0.3168
12(G) 0.2263
9(G) 0.1983
1(G) 0.1590
11 0.1222
13 0.1090
10 0.0953
43 0.0855
15 0.0830
41 0.0796

TABLE 8.2
Vulnerability of Some Nodes of the IEEE 57 Nodes
Systems with DG under Attacks on Nodes
Node e
8(G) 0.3090
12(G) 0.2165
9(G) 0.1933
1(G) 0.1545
11 0.1166
13 0.1028
10 0.0913
43 0.0831
15 0.0824
41 0.0775
58 (DG) 0.0060
59 (DG) 0.0050
60 (DG) 0.0033
61 0.0006
62 0.0003

of DG, the present simulation includes the effect of DG installed at designated load
buses on the system voltage and line loss parameter. Figure 8.5 exhibits the effect of
DGs on voltage profles of the 57 bus system while Table 8.4 indicates the line losses.
The black and grey lines in the Figure 8.5 represents the voltage profles of buses
prior to and post allocation of DGs to designated load buses. It is observed that this
methodology of GVI is backed up by improvement of voltage profle of the system.
Similarly, the line losses are estimated as shown in the Table 8.4.
Effect of Distributed Energy Sources 185

TABLE 8.3
GVI Values of IEEE 57 Bus System
Condition Value of GVI
Before use of DGs 36.5792
Identifed load buses(58-62) added with DGs 3.9941

FIGURE 8.5 Voltage profle of IEEE 57 bus system with and without DGs.

TABLE 8.4
Total Active and Reactive Losses in the IEEE 57 Bus
System by Addition of DGs into Designated Load Buses
Bus No. Total Loss in MW Total Loss in MVar
2 27.0318 150.2279
5 25.9881 146.2857
6 22.7915 133.4965
9 20.1473 122.2515
10 19.8262 120.8614
13 18.7594 113.6647
14 18.1836 114.4084
15 17.4407 111.2176
16 15.7057 103.9075
17 15.4488 102.8686
18 15.1619 96.8877

(Continued)
186 Power Transmission System Analysis Against Faults and Attacks

TABLE 8.4 (Continued)

Total Active and Reactive Losses in the IEEE 57 Bus
System by Addition of DGs into Designated Load Buses
Bus No. Total Loss in MW Total Loss in MVar
19 14.9832 96.3194
20 14.8252 95.9033
23 14.4265 94.4963
25 13.9946 91.7036
27 13.8717 90.9784
28 13.9364 90.8887
29 14.5174 91.4546
30 14.3285 90.3546
31 13.9263 88.8416
32 13.7985 88.4834
33 13.5484 87.8374
35 13.0686 86.2027
38 12.6964 84.7547
41 12.7202 84.0709
42 12.5631 83.2418
43 12.5774 83.2283
44 12.4536 82.6353
47 12.2727 81.8934
49 11.8786 81.0033
50 10.9075 77.8336
51 11.2105 78.8697
52 11.3415 79.6156
53 12.1473 83.2988
54 12.3915 84.1744
55 12.7765 85.6174
56 12.9126 86.2513
57 13.0727 87.0634

The results indicate the gradual declining of line losses with repetitive addition
of distribution generation units to load buses. Reduction in the losses is because
of the proximity of DG to the load. The employment of large-scale DG units not
only decreases the system vulnerability but also achieves better power transmission,
increases voltage profle and makes a grid more effcient. In this method, a small
change in infrastructure can go a long way in strengthening the power grid.

REFERENCES
1. T. Funabashi, “Integration of Distributed Energy Resources in Power Systems”,
Cambridge, MA: Academic Press, 2016.
2. hdl.handle.net.
3. D.P. Kothari, K.C. Singal, and R. Ranjan, “Renewable Energy Sources and Emerging
Technologies”, Second Edition, New Delhi: PHI Learning Private Limited, 2012.
Effect of Distributed Energy Sources 187

4. G.S. Sawhney, “Non Conventional Energy Resources”, New Delhi: PHI Learning
Private Limited, 2012.
5. J.P. Barton and D.G. Infeld, “Energy storage and its use with intermittent renewable
energy,” IEEE Transactions on Energy Conversion, vol. 19, no. 2, pp. 441–448, 2004.
doi: 10.1109/TEC.2003.822305.
6. A.H. Ali, A. Youssef, T. George, and S. Kamel, “Optimal DG allocation in distribu-
tion systems using ant lion optimizer,”2018 International Conference on Innovative
Trends in Computer Engineering (ITCE), Aswan, pp. 324–331, 2018. doi: 10.1109/
ITCE.2018.8316645.
7. K. Mahmoud, N. Yorino, and A. Ahmed, “Optimal distributed generation allocation in
distribution systems for loss minimization,” IEEE Transactions on Power Systems, vol.
31, no. 2, pp. 960–969, 2016. doi: 10.1109/TPWRS.2015.2418333.
8. F.S. Abu-Mouti and M.E. El-Hawary, “Optimal distributed generation alloca-
tion and sizing in distribution systems via Artifcial Bee Colony algorithm,” IEEE
Transactions on Power Delivery, vol. 26, no. 4, pp. 2090–2101, 2011. doi: 10.1109/
TPWRD.2011.2158246.
9. P.J. Varghese, A. Sas, and K. Sunderamoorthy, Optimal siting and sizing of DGs for
congestion relief in transmission lines, 2017 IEEE PES Asia-Pacifc Power and Energy
Engineering Conference (APPEEC), Bangalore, 2017.
10. B. Hayes, “Distributed Generation Systems”, Amsterdam: Elsevier BV, 2017.
11. T. Chowdhury, A. Chakrabarti, and C.K. Chanda, “Analysis of vulnerability indices of
power grid integrated DG units based on Complex Network theory,” Annual IEEE India
Conference (INDICON), New Delhi, pp. 1–5, 2015.doi: 10.1109/INDICON.2015.7443605.
Appendix A
IEEE 57-BUS TEST SYSTEM
(American Electric Power)

Bus Code 1 = Slack Bus, 2 = PV Bus and 0 = PQ Bus

Load Generator Static Mvar

Bus Bus Voltage Angle
No. Code Mag. Degree MW Mvar MW MVAr Qmin Qmax +Qc/−Ql
Busdata=[ 1 1 1.040 0 478 128.9 55 17 0 0 0
2 2 1.010 −1.18 0 −0.8 3 88 −17 50 0
3 2 0.985 −5.97 40 −1 41 21 −10 60 0
4 0 0.981 −7.3 0 0 0 0 0 0 0
5 0 0.976 −8.52 0 0 13 4 0 0 0
6 2 0.980 −8.65 0 0.8 75 2 −8 25 0
7 0 0.984 −7.58 0 0 0 0 0 0 0
8 2 1.005 −4.45 450 62.1 150 22 −140 200 0
9 2 0.980 −9.56 0 2.2 121 26 −3 9 0
10 0 0.986 −11.43 0 0 5 2 0 0 0
11 0 0.974 −10.17 0 0 0 0 0 0 0
12 2 1.015 −10.46 310 128.5 337 24 −50 155 0
13 0 0.979 −9.79 0 0 18 2.3 0 0 0
14 0 0.970 −9.93 0 0 10.5 5.3 0 0 0
15 0 0.988 −7.18 0 0 22 5 0 0 0
16 0 1.013 −8.85 0 0 43 3 0 0 0
17 0 1.017 −5.39 0 0 42 8 0 0 0
18 0 1.001 −11.71 0 0 27.2 9.8 0 0 10
19 0 0.970 −13.20 0 0 3.3 0.6 0 0 0
20 0 0.964 −13.41 0 0 2.3 1 0 0 0
21 0 1.008 −12.89 0 0 0 0 0 0 0
22 0 1.010 −12.84 0 0 0 0 0 0 0
23 0 1.008 −12.91 0 0 6.3 2.1 0 0 0
24 0 0.999 −13.25 0 0 0 0 0 0 0
25 0 0.982 −18.13 0 0 6.3 3.2 0 0 5.9
26 0 0.959 −12.95 0 0 0 0 0 0 0
27 0 0.982 −11.48 0 0 9.3 0.5 0 0 0
(Continued)

189
190 Appendix A

Load Generator Static Mvar

Bus Bus Voltage Angle
No. Code Mag. Degree MW Mvar MW MVAr Qmin Qmax +Qc/−Ql
28 0 0.997 −10.45 0 0 4.6 2.3 0 0 0
29 0 1.010 −9.75 0 0 17 2.6 0 0 0
30 0 0.962 −18.68 0 0 3.6 1.8 0 0 0
31 0 0.936 −19.34 0 0 5.8 2.9 0 0 0
32 0 0.949 −18.46 0 0 1.6 0.8 0 0 0
33 0 0.947 −18.50 0 0 3.8 1.9 0 0 0
34 0 0.959 −14.10 0 0 0 0 0 0 0
35 0 0.966 −13.86 0 0 6 3 0 0 0
36 0 0.976 −13.59 0 0 0 0 0 0 0
37 0 0.985 −13.41 0 0 0 0 0 0 0
38 0 1.013 −12.71 0 0 14 7 0 0 0
39 0 0.983 −13.46 0 0 0 0 0 0 0
40 0 0.973 −13.62 0 0 0 0 0 0 0
41 0 0.996 −14.05 0 0 6.3 3 0 0 0
42 0 0.966 −15.50 0 0 7.1 4 0 0 0
43 0 1.010 −11.33 0 0 2 1 0 0 0
44 0 1.017 −11.83 0 0 12 1.8 0 0 0
45 0 1.036 −9.25 0 0 0 0 0 0 0
46 0 1.060 −11.09 0 0 0 0 0 0 0
47 0 1.033 −12.49 0 0 29.7 11.6 0 0 0
48 0 1.027 −12.57 0 0 0 0 0 0 0
49 0 1.036 −12.92 0 0 18 8.5 0 0 0
50 0 1.023 −13.39 0 0 21 10.5 0 0 0
51 0 1.052 −12.52 0 0 18 5.3 0 0 0
52 0 0.980 −11.47 0 0 4.9 2.2 0 0 0
53 0 0.971 −12.23 0 0 20 10 0 0 6.3
54 0 0.996 −11.69 0 0 4.1 1.4 0 0 0
55 0 1.031 −10.78 0 0 6.8 3.4 0 0 0
56 0 0.968 −16.04 0 0 7.6 2.2 0 0 0
57 0 0.965 −16.56 0 0 6.7 2.0 0 0 0];
Appendix A 191

Bus Bus R X 1/2 B

Line Code = 1 for Lines
nl nr p.u. p.u. p.u. > 1 or < 1 tr. Tap at Bus nl
Linedata=[ 1 2 0.0083 0.0280 0.0645 1
2 3 0.0298 0.0850 0.0409 1
3 4 0.0112 0.0366 0.0190 1
4 5 0.0625 0.1320 0.0129 1
4 6 0.0430 0.1480 0.0174 1
6 7 0.0200 0.1020 0.0138 1
6 8 0.0339 0.1730 0.0235 1
8 9 0.0099 0.0505 0.0274 1
9 10 0.0369 0.1679 0.0220 1
9 11 0.0258 0.0848 0.0109 1
9 12 0.0648 0.2950 0.0386 1
9 13 0.0481 0.1580 0.0202 1
13 14 0.0132 0.0434 0.0055 1
13 15 0.0269 0.0869 0.0115 1
1 15 0.0178 0.0910 0.0494 1
1 16 0.0454 0.2060 0.0273 1
1 17 0.0238 0.1080 0.0143 1
3 15 0.0162 0.0530 0.0272 1
4 18 0 0.5550 0 0.97
4 18 0 0.4300 0 0.978
5 6 0.0302 0.0641 0.0062 1
7 8 0.0139 0.0712 0.0097 1
10 12 0.0277 0.1262 0.0164 1
11 13 0.0223 0.0732 0.0094 1
12 13 0.0178 0.0580 0.0302 1
12 16 0.0180 0.0813 0.0108 1
12 17 0.0397 0.1790 0.0238 1
14 15 0.0171 0.0547 0.0074 1
18 19 0.4610 0.6850 0 1
19 20 0.2830 0.4340 0 1
20 21 0 0.7767 0 1.043
21 22 0.0736 0.1170 0 1
22 23 0.0099 0.0152 0 1
23 24 0.1660 0.2560 0.0042 1
24 25 0 1.1820 0 1
24 25 0 1.2300 0 1
24 26 0 0.0473 0 1.043
26 27 0.1650 0.2540 0 1
27 28 0.0618 0.0954 0 1
28 29 0.0418 0.0587 0 1
(Continued)
192 Appendix A

Bus Bus R X 1/2 B

Line Code = 1 for Lines
nl nr p.u. p.u. p.u. > 1 or < 1 tr. Tap at Bus nl
7 29 0 0.0648 0 0.967
25 30 0.1350 0.2020 0 1
30 31 0.3260 0.4970 0 1
31 32 0.507 0.7550 0 1
32 33 0.0392 0.0360 0 1
32 34 0 0.9530 0 0.975
34 35 0.0520 0.0780 0.0016 1
35 36 0.0430 0.0537 0.0008 1
36 37 0.0290 0.0366 0 1
37 38 0.0651 0.1009 0.0010 1
37 39 0.0239 0.0379 0 1
36 40 0.0300 0.0466 0 1
22 38 0.0192 0.0295 0 1
11 41 0 0.7490 0 0.955
41 42 0.2070 0.3520 0 1
41 43 0 0.4120 0 1
38 44 0.0289 0.0585 0.0010 1
15 45 0 0.1042 0 0.955
14 46 0 0.0735 0 0.9
46 47 0.0230 0.0680 0.0016 1
47 48 0.0182 0.0233 0 1
48 49 0.0834 0.1290 0.0024 1
49 50 0.0801 0.1280 0 1
50 51 0.1386 0.2200 0 1
10 51 0 0.0712 0 0.93
13 49 0 0.1910 0 0.895
29 52 0.1442 0.1870 0 1
52 53 0.0762 0.0984 0 1
53 54 0.1878 0.2320 0 1
54 55 0.1732 0.2265 0 1
11 43 0 0.1530 0 0.958
44 45 0.0624 0.1242 0.0020 1
40 56 0 1.1950 0 0.985
56 41 0.5530 0.4590 0 1
56 42 0.2125 0.3540 0 1
39 57 0 1.3550 0 0.98
57 56 0.1740 0.2600 0 1
38 49 0.1150 0.1770 0.0030 1
38 48 0.0312 0.0482 0 1
9 55 0 0.1205 0 0.94];
 Appendix A

FIGURE A.1 IEEE 57-bus system.

193
Appendix B
West Bengal State Electricity
Board’s (WBSEB) 203-Bus System
(A Semi-Government Power
Utility in Eastern Part of India)

Bus Code 0 = Slack Bus, 2 = PV Bus and 1 = PQ Bus.

195
 Generator Load
196

Voltage Voltage
Bus No. MW MVAr MW MVAr Mag. Angle Bus Code ysnt Qmin Qmax
Busdata=[ 1 4 0 0 0 1.05 0 0 0 −1 2; %FKK21
2 1.8 0 0 0 1.04 0 2 0 −0.5 1; %KT115.75
3 1.8 0 0 0 1.04 0 2 0 −0.5 1; %KT215.75
4 1.9 0 0 0 1.04 0 2 0 −0.5 1; %BKR15.75
5 1.8 0 0 0 1.05 0 2 0 −0.5 1; %FKK15.75
6 1.9 0 0 0 1.0433 0 2 0 −0.5 1; %BAKR15.7
7 0.6 0 0 0 1.02 0 2 0 −0.3 0.6; %STP13.8
8 0.9 0 0 0 1 0 2 0 −0.22 0.45; %DPL13.8
9 0.115 0 0 0 1.02 0 2 0 −0.03 0.06;%RMM211
10 1.8 0 0 0 1 0 2 0 −0.5 1; %BTP11
11 0.5 0 0 0 1 0 2 0 −0.2 0.4; %BTPS11
12 0.6 0 0 0 1 0 2 0 −0.15 0.3; %DPL11
13 0.675 0 0 0 1.01 0 2 0 −0.2 0.4; %CHU11
14 0.181 0 0 0 1.02 0 2 0 −0.02 0.04; %JAL111
15 0.136 0 0 0 1.02 0 2 0 −0.009 0.018; %JAL211
16 0.18 0 0 0 1.02 0 2 0 −0.045 0.09; %RANGIT11
17 1.5 0 0 0 1.04 0 2 0 −0.45 0.9; %MEJIA11
18 0.5 0 0 0 1.03 0 2 0 −0.3 0.6; %WARIA11
19 1.5 0 0 0 1.03 0 2 0 −0.5 1; %WARIA211
20 1.5 0 0 0 1.04 0 2 0 −0.4 0.8; %TALA11
21 0.068 0 0 0 1.02 0 2 0 −0.019 0.038; %TCF16.6
22 0.075 0 0 0 1.02 0 2 0 −0.019 0.038; %TCF26.6
23 0.075 0 0 0 1.02 0 2 0 −0.019 0.038; %TCF36.6
Appendix B

(Continued)
 Generator Load
Voltage Voltage
Bus No. MW MVAr MW MVAr Mag. Angle Bus Code ysnt Qmin Qmax
24 0.2 0 0 0 1 0 2 0 −0.05 0.1; %DPL6.3
Appendix B

25 0 0 0 0 1 0 1 0 0 0; %PSP16.5
26 0 0 1.2 1 1 0 1 0 0 0; %JRT132
27 0 0 0.68 0.921 1 0 1 0 0 0; %ARAM132
28 0 0 0 0 1 0 1 0 0 0; %KTPS132
29 0 0 0 0 1 0 1 0 0 0; %MALPG132
30 0 0 0 0 1 0 1 0 0 0; %DRG132
31 0 0 0.55 0.34 1 0 1 0 0 0; %DOMJ132
32 0 0 0.51 0.316 1 0 1 0.067 0 0; %GOK132
33 0 0 0.5 1.32 1 0 1 0.067 0 0; %HOW132
34 0 0 0.6 1.363 1 0 1 0 0 0; %KASBA132
35 0 0 0.51 0.316 1 0 1 0 0 0; %LXP132
36 0 0 0 0 1 0 1 0 0 0; %MIDNA132
37 0 0 0.35 0.217 1 0 1 0 0 0; %NJP132
38 0 0 0.2 0.124 1 0 1 0 0 0; %NHAL132
39 0 0 0 0 1 0 1 0 0 0; %RISH132
40 0 0 0.45 0.527 1 0 1 0.067 0 0; %SATG132
41 0 0 1 1.239 1 0 1 0 0 0; %DPL132
42 0 0 0 0 1 0 1 0 0 0; %BRPPG132
43 0 0 0 0 1 0 1 0 0 0; %SLGPG132
44 0 0 0.72 0.446 1 0 1 0.033 0 0; %ADI132
45 0 0 0.15 0.193 1 0 1 0 0 0; %ALIP132
(Continued)
197
 Generator Load
198

Voltage Voltage
Bus No. MW MVAr MW MVAr Mag. Angle Bus Code ysnt Qmin Qmax
46 0 0 0.41 0.178 1 0 1 0 0 0; %ASOK132
47 0 0 0.32 0.198 1 0 1 0 0 0; %BALU132
48 0 0 0.58 0.259 1 0 1 0.033 0 0; %BNK132
49 0 0 0.37 0.477 1 0 1 0.033 0 0; %BARAS132
50 0 0 0.35 0.2169 1 0 1 0 0 0; %BASIR132
51 0 0 0.4 0.2558 1 0 1 0.0333 0 0; %JOKA132
52 0 0 0.46 0.533 1 0 1 0.0333 0 0; %BERH132
53 0 0 0.35 0.2169 1 0 1 0 0 0; %BRP132
54 0 0 0.4 0.2958 1 0 1 0 0 0; %VISH132
55 0 0 0 0 1 0 1 0 0 0; %BOLP132
56 0 0 0.33 0.2045 1 0 1 0 0 0; %BONG132
57 0 0 0.4 0.2479 1 0 1 0 0 0; %CLC132
58 0 0 0.53 0.3285 1 0 1 0.333 0 0; %CHK132
59 0 0 0.14 0.0868 1 0 1 0 0 0; %DALK132
60 0 0 0.24 0.1487 1 0 1 0 0 0; %DARJ132
61 0 0 0.31 0.1921 1 0 1 0.167 0 0; %DEBO132
62 0 0 0.57 0.3533 1 0 1 0.15 0 0; %DHRAM132
63 0 0 0.25 0.1549 1 0 1 0 0 0; %DHUL132
64 0 0 0 0 1 0 1 0 0 0; %EGRA132
65 0 0 0.43 0.2665 1 0 1 0.1667 0 0; %FALTA132
66 0 0 0.39 0.2417 1 0 1 0 0 0; %HALD132
67 0 0 0.49 0.4276 1 0 1 0 0 0; %HIZLI132
(Continued)
Appendix B
 Generator Load
Voltage Voltage
Bus No. MW MVAr MW MVAr Mag. Angle Bus Code ysnt Qmin Qmax
68 0 0 0.66 0.409 1 0 1 0.1333 0 0; %KLY132
Appendix B

69 0 0 0.74 0.4586 1 0 1 0.1667 0 0; %KATWA132

70 0 0 0.41 0.2541 1 0 1 0 0 0; %KHAN132
71 0 0 0.46 0.5702 1 0 1 0.1333 0 0; %KOLAG132
72 0 0 0.52 0.3223 1 0 1 0.196 0 0; %KRISH132
73 0 0 0.56 0.3325 1 0 1 0.1667 0 0; %LILO132
74 0 0 0.68 0.4214 1 0 1 0 0 0; %MALDA132
75 0 0 0.33 0.2045 1 0 1 0 0 0; %MANK132
76 0 0 0.4 0.2479 1 0 1 0 0 0; %MOINA132
77 0 0 0 0 1 0 1 0 0 0; %NBU132
78 0 0 0.25 0.3409 1 0 1 0 0 0; %PING132
79 0 0 0.45 0.4648 1 0 1 0 0 0; %PURU132
80 0 0 0.46 0.2851 1 0 1 0.3333 0 0; %RGJ132
81 0 0 0.23 0.1425 1 0 1 0 0 0; %RAIG132
82 0 0 0.56 0.3471 1 0 1 0 0 0; %RAINA132
83 0 0 0.4 0.2479 1 0 1 0 0 0; %RAMP132
84 0 0 0.7 0.6197 1 0 1 0 0 0; %RANA132
85 0 0 0.5 0.3099 1 0 1 0.1667 0 0; %SAIN132
86 0 0 0 0 1 0 1 0 0 0; %SL132
87 0 0 0.32 0.1983 1 0 1 0.0333 0 0; %SAMSI132
88 0 0 0.42 0.2603 1 0 1 0 0 0; %SLG132
89 0 0 0.25 0.2048 1 0 1 0 0 0; %SONA132
(Continued)
199
 Generator Load
200

Voltage Voltage
Bus No. MW MVAr MW MVAr Mag. Angle Bus Code ysnt Qmin Qmax
90 0 0 0.29 0.1797 1 0 1 0 0 0; %TAML132
91 0 0 0.54 0.3347 1 0 1 0 0 0; %TARAK132
92 0 0 0 0 1 0 1 0 0 0; %TITA132
93 0 0 0.5 0.3099 1 0 1 0 0 0; %UKHRA132
94 0 0 0.57 0.3533 1 0 1 0 0 0; %ULU132
95 0 0 0 0 1 0 1 0 0 0; %STPS132
96 0 0 0 0 1 0 1 0 0 0; %BTPS132
97 0 0 0.19 0.1178 1 0 1 0 0 0; %BALIC132
98 0 0 0 0 1 0 1 0 0 0; %TCFI132
99 0 0 0 0 1 0 1 0 0 0; %TCFII132
100 0 0 0 0 1 0 1 0 0 0; %TCF3132
101 0 0 0 0 1 0 1 0 0 0; %RMM2132
102 0 0 0.13 0.0806 1 0 1 0 0 0; %HM132
103 0 0 0.05 0.031 1 0 1 0 0 0; %DANK132
104 0 0 0 0 1 0 1 0 0 0; %RANGT132
105 0 0 0.1 0.062 1 0 1 0 0 0; %HPCL132
106 0 0 0.006 0.0037 1 0 1 0 0 0; %FCI132
107 0 0 0.23 0.1425 1 0 1 0 0 0; %JANGI132
108 0 0 0.31 0.1921 1 0 1 0 0 0; %CTOLA132
109 0 0 0.12 0.0744 1 0 1 0.3096 0 0; %GRAM132
110 0 0 0.35 0.2169 1 0 1 0 0 0; %GTOK132
111 0 0 0.28 0.1735 1 0 1 0 0 0; %MELLI132
(Continued)
Appendix B
 Generator Load
Voltage Voltage
Bus No. MW MVAr MW MVAr Mag. Angle Bus Code ysnt Qmin Qmax
112 0 0 1 1.2395 1 0 1 0 0 0; %WARIA132
Appendix B

113 0 0 0.38 0.2355 1 0 1 0 0 0; %BIRSG132

114 0 0 0.35 0.2169 1 0 1 0 0 0; %COOCH132
115 0 0 0 0 1 0 1 0 0 0; %JRT220
116 0 0 0 0 1 0 1 0 0 0; %ARAM220
117 0 0 0 0 1 0 1 0 0 0; %KTPS220
118 0 0 0 0 1 0 1 0 0 0; %BAKR220
119 0 0 0 0 1 0 1 0 0 0; %PARU220
120 0 0 0 0 1 0 1 0 0 0; %MALPG220
121 0 0 0 0 1 0 1 0 0 0; %BGURI220
122 0 0 0 0 1 0 1 0 0 0; %DRG220
123 0 0 0 0 1 0 1 0 0 0; %DOMJ220
124 0 0 0 0 1 0 1 0 0 0; %GOK220
125 0 0 0 0 1 0 1 0 0 0; %HOW220
126 0 0 0 0 1 0 1 0 0 0; %KASBA220
127 0 0 0 0 1 0 1 0 0 0; %LXP220
128 0 0 0 0 1 0 1 0 0 0; %MIDNA220
129 0 0 0 0 1 0 1 0 0 0; %NJP220
130 0 0 0 0 1 0 1 0 0 0; %NHAL220
131 0 0 0 0 1 0 1 0 0 0; %RISH220
132 0 0 0 0 1 0 1 0 0 0; %SATG220
133 0 0 0 0 1 0 1 0 0 0; %DPL220
(Continued)
201
 Generator Load
202

Voltage Voltage
Bus No. MW MVAr MW MVAr Mag. Angle Bus Code ysnt Qmin Qmax
134 0 0 0 0 1 0 1 0 0 0; %WARIA220
135 0 0 0 0 1 0 1 0 0 0; %CHUK220
136 0 0 0.3 0.1859 1 0 1 0 0 0 0; %BRPPG220
137 0 0 0 0 1 0 1 0 0 0; %SLGPG220
138 0 0 0.5 0.5099 1 0 1 0 0 0 0; %DALPG220
139 0 0 0 0 1 0 1 0 0 0; %PURN220
140 0 0 0.5 0.6197 1 0 1 0 0 0 0; %STPS220
141 0 0 1 2.479 1 0 1 0 0 0; %MEJIA220
142 0 0 0.45 0.4789 1 0 1 0 0 0 0; %FKK220
143 0 0 0 0 1 0 1 0 0 0; %SGPG220
144 0 0 0 0 1 0 1 0 0 0; %KRIS220
145 0 0 1 0.35 1 0 1 −0.5333 0 0; %JRT400 load
added
146 0 0 1 1 1 0 1 −0.5667 0 0; %ARAM400load
added
147 0 0 0 0 1 0 1 −0.5867 0 0; %KTPS400
148 0 0 0 0 1 0 1 −0.5 0 0; %BAKR400
149 0 0 2 3.479 1 0 1 −0.5333 0 0; %FKK400
150 0 0 3.5 3.8592 1 0 1 −0.5 0 0; %PARU400
151 0 0 0 0 1 0 1 −0.42 0 0; %MALPG400
152 0 0 0 0 1 0 1 −0.41 0 0; %BGURI400
153 0 0 0.5 0.6197 1 0 1 −0.525 0 0; %BARIP400
154 0 0 0.75 1.8592 1 0 1 −0.5867 0 0; %PURN400
(Continued)
Appendix B
 Generator Load
Voltage Voltage
Bus No. MW MVAr MW MVAr Mag. Angle Bus Code ysnt Qmin Qmax
155 0 0 0 0 1 0 1 −0.5366 0 0; %SGPG400
Appendix B

156 0 0 0 0 1 0 1 −0.5366 0 0; %TALA400

157 0 0 0 0 1 0 1 −0.5366 0 0; %PPSP2
158 0 0 0 0 1 0 1 0 0 0; %JALI66
159 0 0 0 0 1 0 1 0 0 0; %JALII66
160 0 0 0.12 0.0744 1 0 1 0.383 0 0; %CHALSA66
cap added
161 0 0 0.16 0.0992 1 0 1 0.433 0 0; %BANAR66
162 0 0 0 0 1 0 1 0.583 0 0; %BRP66
163 0 0 0.15 0.093 1 0 1 0.583 0 0; %KPONG66
164 0 0 0.23 0.018 1 0 1 0.613 0 0; %MELLI66
165 0 0 0.2 0.32 1 0 1 0 0 0; %JRT33
166 0 0 0.1 0.21 1 0 1 0 0 0; %ARAM33
167 0 0 0.23 0.15 1 0 1 0 0 0; %KTPS33
168 0 0 0.31 0.21 1 0 1 0 0 0; %FKK33
169 0 0 0.1 0.21 1 0 1 0 0 0; %PARU33
170 0 0 0.1 0.21 1 0 1 0 0 0; %BAKR33
171 0 0 0.1 0.2 1 0 1 0 0 0; %MLDAPG33
172 0 0 0.1 0.35 1 0 1 0 0 0; %BGURI33
173 0 0 0.23 0.41 1 0 1 0 0 0; %PURN33
174 0 0 0.1 0.25 1 0 1 0.05 0 0; %JRT133
175 0 0 0.1 0.21 1 0 1 0 0 0; %ARAM133
(Continued)
203
 Generator Load
204

Voltage Voltage
Bus No. MW MVAr MW MVAr Mag. Angle Bus Code ysnt Qmin Qmax
176 0 0 0.1 0.21 1 0 1 0 0 0; %KTPS133
177 0 0 0.1 0.45 1 0 1 0 0 0; %MLDPG133
178 0 0 0.1 0.21 1 0 1 0 0 0; %DRG33
179 0 0 0.1 0.21 1 0 1 0 0 0; %DOMJ33
180 0 0 0.1 0.21 1 0 1 0 0 0; %GOK33
181 0 0 0.1 0.21 1 0 1 0 0 0; %HOW33
182 0 0 0.1 0.21 1 0 1 0 0 0; %KASBA33
183 0 0 0.1 0.21 1 0 1 0 0 0; %LXP33
184 0 0 0.1 0.21 1 0 1 0 0 0; %MIDNA33
185 0 0 0.1 0.51 1 0 1 0 0 0; %NJP33
186 0 0 0.1 0.21 1 0 1 0 0 0; %NHAL33
187 0 0 0.1 0.21 1 0 1 0 0 0; %RISH33
188 0 0 0.1 0.1 1 0 1 0 0 0; %SATG33
189 0 0 0.1 0.1 1 0 1 0 0 0; %DPL33
190 0 0 0.1 0.4 1 0 1 0 0 0; %BRPPG33
191 0 0 0.1 0.4 1 0 1 0 0 0; %SLGPG33
192 0 0 0.1 0.1 1 0 1 0 0 0; %STPS33
193 0 0 0.1 0.1 1 0 1 0 0 0; %WARIA33
194 0 0 0.37 0.2176 1 0 1 0.38 0 0; %RISHRA33
195 0 0 0.36 0.018 1 0 1 0.5233 0 0; %BISHNU33
196 0 0 0.17 0.1413 1 0 1 0.5233 0 0; %EGRA33
197 0 0 0.41 0.2161 1 0 1 0.03 0 0; %NBU33
(Continued)
Appendix B
 Appendix B

Generator Load
Voltage Voltage
Bus No. MW MVAr MW MVAr Mag. Angle Bus Code ysnt Qmin Qmax
198 0 0 0.21 0.12 1 0 1 0.5367 0 0; %SLAKE33
199 0 0 0.23 0.1599 1 0 1 0.372 0 0; %BOLPUR33
200 0 0 0.25 0.30 1 0 1 0.5883 0 0; %TITAG33
201 0 0 0.32 0.2 1 0 1 0.5833 0 0; %MIDNAP33
202 0 0 0.1 0.25 1 0 1 0.15 0 0; %SGPG33
203 0 0 0 0 1 0 1 0 0 0;] %KRIS33

Actual tap-position and tap of trf. is included.

Slack bus should be from bus for line 206,
Data with shunt capacitor only.
tap_pos=1 for tr at SE and tap_pos=2 for tr at RE and zero for tr. Line.
205
206 Appendix B

Bus tap
No. fb eb r x hlch Tap posl lmt
Linedata=[ 52 44 96 0.0036 0.0071 0.0108 1.0 0 10; %ADI132-BTPS132
53 44 91 0.0095 0.0189 0.0287 1.0 0 10; %ADI132-
TARAK132
54 45 53 0.0385 0.077 0.0292 1.0 0 10; %ALIP132-BRP132
55 27 82 0.0116 0.0232 0.0352 1.0 0 10; %ARAM132-
RAINA132
56 27 91 0.0071 0.0143 0.0217 1.0 0 10; %ARAM132-
TARAK132
57 46 57 0.031 0.0621 0.0235 1.0 0 10; %ASOK132-CLC132
58 46 26 0.0053 0.0107 0.0162 1.0 0 10; %ASOK132-JRT132
59 36 97 0.0178 0.0357 0.0135 1.0 0 10; %MIDNA132-
BALIC132
60 78 97 0.0185 0.0371 0.0141 1.0 0 10; %PING132-
BALIC132
61 47 109 0.0407 0.0813 0.0309 1.0 0 0 10; %BALU132-
GRAM132
62 48 54 0.0134 0.0268 0.0407 1.0 0 10; %BNK132-VISH132
63 49 50 0.0125 0.025 0.0379 1.0 0 10; %BARAS132-
BASIR132
64 49 26 0.0078 0.0157 0.0238 1.0 0 10; %BARAS132-
JRT132
65 52 32 0.0064 0.0128 0.0195 1.0 0 10; %BERH132-GOK132
66 53 42 0.0006 0.0011 0.0004 1.0 0 10; %BRP132-
BRPPG132
67 53 76 0.0132 0.0264 0.0401 1.0 0 10; %BRP132-
MOINA132
68 55 30 0.0481 0.0961 0.0365 1.0 0 10; %BOLP132-DRG132
69 55 85 0.026 0.0519 0.0197 1.0 0 10; %BOLP132-SAIN132
70 56 72 0.0189 0.0377 0.0573 1.0 0 10; %BONG132-
KRISH132
71 96 62 0.0073 0.0146 0.0222 1.0 0 10; %BTPS132-
DHRAM132
72 96 62 0.0188 0.0377 0.0143 1.0 0 10; %BTPS132-
DHRAM132
73 96 70 0.0285 0.0571 0.0217 1.0 0 10; %BTPS132-
KHAN132
74 96 68 0.0107 0.0214 0.0081 1.0 0 10; %BTPS132-KLY132
75 96 73 0.0328 0.0656 0.0249 1.0 0 10; %BTPS132-LILO132
76 96 39 0.0111 0.0221 0.0336 1.0 0 10; %BTPS132-RISH132
77 96 40 0.0378 0.0756 0.0287 1.0 0 10; %BTPS132-SATG132
78 96 108 0.0189 0.0378 0.0143 1.0 0 0 10; %BTPS132-
CTOLA132
79 58 36 0.0357 0.0713 0.0271 1.0 0 10; %CHK132-
MIDNA132
(Continued)
Appendix B 207

Bus tap
No. fb eb r x hlch Tap posl lmt
80 58 54 0.0321 0.0642 0.0244 1.0 0 10; %CHK132-VISH132
81 59 81 0.0182 0.0364 0.0552 1.0 0 10; %DALK132-
RAIG132
82 59 100 0.0388 0.0776 0.1178 1.0 0 0 10; %DALK132-
TCF3132
83 60 77 0.0467 0.0934 0.0355 1.0 0 10; %DARJ132-NBU132
84 60 101 0.0146 0.0292 0.0111 1.0 0 0 10; %DARJ132-
RMM2132
85 61 69 0.009 0.018 0.0274 1.0 0 10; %DEBO132-
KATWA132
86 62 26 0.005 0.01 0.0152 1.0 0 10; %DHRAM132-
JRT132
87 62 68 0.0093 0.0185 0.007 1.0 0 10; %DHRAM132-
KLY132
88 62 84 0.0132 0.0264 0.0401 1.0 0 10; %DHRAM132-
RANA132
89 62 92 0.0103 0.0207 0.0314 1.0 0 10; %DHRAM132-
TITA132
90 63 74 0.0541 0.1081 0.041 1.0 0 10; %DHUL132-
MALDA132
91 63 32 0.0499 0.0999 0.0379 1.0 0 10; %DHUL132-
GOK132
92 31 94 0.0096 0.0193 0.0292 1.0 0 10; %DOMJ132-ULU132
93 30 41 0.0036 0.0071 0.0108 1.0 0 10; %DRG132-DPL132
94 30 75 0.0139 0.0278 0.0422 1.0 0 10; %DRG132-
MANK132
95 30 85 0.0421 0.0842 0.0319 1.0 0 10; %DRG132-SAIN132
96 30 93 0.0065 0.013 0.0197 1.0 0 10; %DRG132-
UKHRA132
97 30 54 0.0185 0.0371 0.0563 1.0 0 10; %DRG132-VISH132
98 64 67 0.0435 0.087 0.033 1.0 0 10; %EGRA132-
HIZLI132
99 65 51 0.0206 0.0412 0.0156 1.0 0 10; %FALTA132-
JOKA132
100 65 35 0.0121 0.0243 0.0368 1.0 0 10; %FALTA132-LXP132
101 51 89 0.0128 0.0257 0.0097 1.0 0 10; %JOKA132-
SONA132
102 32 69 0.0203 0.0407 0.0617 1.0 0 10; %GOK132-
KATWA132
103 32 80 0.0378 0.0756 0.0287 1.0 0 10; %GOK132-RGJ132
104 32 83 0.0414 0.0827 0.0314 1.0 0 10; %GOK132-
RAMP132
105 32 85 0.0385 0.077 0.0292 1.0 0 10; %GOK132-SAIN132
106 66 90 0.0143 0.0285 0.0433 1.0 0 10; %HALD132-
TAML132
(Continued)
208 Appendix B

Bus tap
No. fb eb r x hlch Tap posl lmt
107 102 103 0.0043 0.0086 0.0032 1.0 0 10; %HM132-DANK132
108 103 73 0.0043 0.0086 0.0032 1.0 0 10; %DANK132-
LILO132
109 102 39 0.0078 0.0157 0.006 1.0 0 10; %HM132-RISH132
110 67 36 0.0053 0.0107 0.0162 1.0 0 10; %HIZLI132-
MIDNA132
111 33 73 0.0043 0.0086 0.013 1.0 0 10; %HOW132-LILO132
112 33 73 0.0046 0.0093 0.0141 1.0 0 10; %HOW132-LILO132
113 26 56 0.0126 0.0251 0.0381 1.0 0 10; %JRT132-BONG132
114 51 34 0.0208 0.0417 0.0158 1.0 0 10; %JOKA132-
KASBA132
115 51 35 0.0272 0.0545 0.0207 1.0 0 10; %JOKA132-LXP132
116 34 86 0.009 0.0179 0.0272 1.0 0 10; %KASBA132-SL132
117 34 89 0.0087 0.0174 0.0066 1.0 0 10; %KASBA132-
SONA132
118 69 40 0.0155 0.031 0.0471 1.0 0 10; %KATWA132-
SATG132
119 70 40 0.01 0.02 0.0076 1.0 0 10; %KHAN132-
SATG132
120 28 71 0.0012 0.0025 0.0038 1.0 0 10; %KTPS132-
KOLAG132
121 28 90 0.0089 0.0178 0.0271 1.0 0 10; %KTPS132-
TAML132
122 28 94 0.0104 0.0208 0.0315 1.0 0 10; %KTPS132-ULU132
123 73 39 0.0136 0.0271 0.0103 1.0 0 10; %LILO132-RISH132
124 73 39 0.0136 0.0271 0.0103 1.0 0 10; %LILO132-RISH132
125 74 29 0.0021 0.0041 0.0063 1.0 0 10; %MALDA132-
MALPG132
126 74 81 0.0556 0.1113 0.0422 1.0 0 10; %MALDA132-
1RAIG132
127 74 87 0.0349 0.0699 0.0265 1.0 0 10; %MALDA132-
SAMSI132
128 36 78 0.0285 0.0571 0.0217 1.0 0 10; %MIDNA132-
PING132
129 36 54 0.0678 0.1355 0.0514 1.0 0 10; %MIDNA132-
VISH132
130 37 76 0.02 0.0399 0.0606 1.0 0 10; %NJP132-
MOINA132
131 77 101 0.0492 0.0984 0.0374 1.0 0 10; %NBU132-
RMM2132
132 77 43 0.0071 0.0143 0.0054 1.0 0 10; %NBU132-
SLGPG132
133 77 98 0.0142 0.0284 0.0108 1.0 0 10; %NBU132-TCFI132
134 37 98 0.0143 0.0285 0.0108 1.0 0 10; %NJP132-TCFI132
(Continued)
Appendix B 209

Bus tap
No. fb eb r x hlch Tap posl lmt
135 37 88 0.0046 0.0093 0.0141 1.0 0 10; %NJP132-SLG132
136 95 79 0.0125 0.025 0.0379 1.0 0 10; %STPS132-
PURU132
137 87 81 0.0506 0.1013 0.0384 1.0 0 10; %SAMSI132-
RAIG132
138 85 83 0.0314 0.0628 0.0238 1.0 0 10; %SAIN132-
RAMP132
139 40 75 0.0257 0.0514 0.078 1.0 0 10; %SATG132-
MANK132
140 98 99 0.0056 0.0113 0.0043 1.0 0 10; %TCFI132-TCFII132
141 98 100 0.0178 0.0357 0.0135 1.0 0 10; %TCFI132-TCF3132
142 99 100 0.0121 0.0243 0.0092 1.0 0 10; %TCFII132-TCF3132
143 77 37 0.0071 0.0143 0.0054 1.0 0 10; %NBU132-NJP132
144 43 104 0.0317 0.0635 0.0964 1.0 0 10; %SLGPG132-
RANGT132
145 104 101 0.0193 0.0385 0.0146 1.0 0 10; %RANGT132-
RMM2132
146 38 66 0.0014 0.0029 0.0011 1.0 0 10; %NHAL132-
HALD132
147 38 105 0.0007 0.0014 0.0005 1.0 0 10; %NHAL132-
HPCL132
148 66 105 0.0014 0.0029 0.0011 1.0 0 10; %HALD132-
HPCL132
149 39 108 0.0039 0.0078 0.003 1.0 0 10; %RISH132-
CTOLA132
150 31 107 0.0061 0.0121 0.0184 1.0 0 10; %DOMJ132-
JANGI132
151 66 106 0.0018 0.0036 0.0054 1.0 0 10; %HALD132-FCI132
152 81 109 0.0407 0.0813 0.0309 1.0 0 10; %RAIG132-
GRAM132
153 27 113 0.0082 0.0164 0.0249 1.0 0 10; %ARAM132-
BIRSG132
154 104 110 0.0521 0.1041 0.0395 1.0 0 10; %RANGT132-
GTOK132
155 104 111 0.0214 0.0428 0.0162 1.0 0 10; %RANGT132-
MELLI132
156 110 111 0.0428 0.0856 0.0325 1.0 0 10; %GTOK132-
MELLI132
157 111 43 0.0642 0.1284 0.0487 1.0 0 10; %MELLI132-
SLGPG132
158 104 43 0.0663 0.1327 0.0503 1.0 0 10; %RANGT132-
SLGPG132
159 114 45 0.0143 0.0285 0.0108 1.0 0 10; %COOCH132-
ALIP132
(Continued)
210 Appendix B

Bus tap
No. fb eb r x hlch Tap posl lmt
160 114 53 0.0514 0.1027 0.039 1.0 0 10; %COOCH132-
BRP132
22 116 122 0.0166 0.058 0.2425 1.0 0 10; %ARAM220-
DRG220
23 122 119 0.0014 0.0049 0.0207 1.0 0 10; %DRG220-PARU220
24 116 123 0.0037 0.013 0.2169 1.0 0 10; %ARAM220-
DOMJ220
25 116 128 0.0046 0.016 0.2673 1.0 0 10; %ARAM220-
MIDNA220
26 116 131 0.0094 0.0328 0.1372 1.0 0 10; %ARAM220-
RISH220
27 116 140 0.0125 0.0438 0.7335 1.0 0 10; %ARAM220-
STPS220
28 118 122 0.0025 0.0088 0.1477 1.0 0 10; %BAKR220-
2DRG220
29 118 124 0.0053 0.0184 0.3083 1.0 0 10; %BAKR220-
GOK220
30 118 132 0.009 0.0315 0.5263 1.0 0 10; %BAKR220-
SATG220
31 123 125 0.0012 0.0041 0.0684 1.0 0 10; %DOMJ220-
HOW220
32 122 133 0.0006 0.0022 0.0376 1.0 0 10; %DRG220-DPL220
33 122 140 0.0064 0.0225 0.3759 1.0 0 10; %DRG220-STPS220
34 125 117 0.0046 0.016 0.2669 1.0 0 10; %HOW220-KTPS220
35 115 126 0.0038 0.0133 0.2233 1.0 0 10; %JRT220-
KASBA220
36 115 126 0.0076 0.0267 0.1118 1.0 0 10; %JRT220-
KASBA220
37 115 132 0.0051 0.0177 0.297 1.0 0 10; %JRT220-SATG220
38 126 127 0.0082 0.0287 0.1201 1.0 0 10; %KASBA220-
LXP220
39 117 130 0.0036 0.0126 0.2109 1.0 0 10; %KTPS220-
NHAL220
40 119 134 0.001 0.0036 0.0602 1.0 0 10; %PARU220-
WARIA220
41 122 134 0.0011 0.0038 0.0639 1.0 0 10; %DRG220-
WARIA220
42 136 135 0.0055 0.0192 0.7218 1.0 0 10; %BRPPG220-
CHUK220
43 138 120 0.0069 0.0243 0.406 1.0 0 10; %DALPG220-
MALPG220
44 138 137 0.0076 0.0265 0.4436 1.0 0 10; %DALPG220-
SLGPG220
45 121 136 0.0053 0.0186 0.312 1.0 0 10; %BGURI220-
BRPPG220
(Continued)
Appendix B 211

Bus tap
No. fb eb r x hlch Tap posl lmt
46 121 137 0.0001 0.0004 0.0075 1.0 0 10; %BGURI220-
SLGPG220
48 141 134 0.0022 0.0076 0.1278 1.0 0 10; %141MEJIA220-
WARIA220
49 132 144 0.003 0.0106 0.1767 1.0 0 10; %SATG220-KRIS220
50 143 115 0.0091 0.0319 0.1335 1.0 0 10; %SGPG220-JRT220
51 127 143 0.0055 0.0193 0.0808 1.0 0 10; %LXP220-SGPG220
7 146 148 0.004 0.0151 0.9204 1.0 0 10; %ARAM400-
8BAKR400
8 145 148 0.005 0.0189 1.154 1.0 0 10; %JRT400-BAKR400
9 146 147 0.002 0.0075 0.4545 1.0 0 10; %ARAM400-
KTPS400
10 145 147 0.0042 0.0156 0.9516 1.0 0 10; %JRT400-KTPS400
11 147 153 0.0056 0.021 1.2783 1.0 0 10; %KTPS400-
BARIP400
12 145 149 0.0073 0.0273 1.6618 1.0 0 10; %JRT400-FKK400
13 149 150 0.0023 0.0087 2.1163 1.0 0 10; %FKK400-PARU400
14 149 151 0.0006 0.0022 0.5255 1.0 0 10; %FKK400-
MALPG400
15 151 154 0.0026 0.0099 2.4145 1.0 0 10; %MALPG400-
PURN400
16 152 154 0.0025 0.0093 2.2725 1.0 0 10; %BGURI400-
PURN400
17 150 157 0.0061 0.0228 1.3919 1.0 0 10; %PARU400-PPSP2
18 146 157 0.0032 0.0122 2.9685 1.0 0 10; %ARAM400-PPSP2
19 155 149 0.0095 0.0355 2.166 1.0 0 10; %SGPG400-FKK400
20 155 145 0.0022 0.0083 0.5042 1.0 0 10; %SGPG400-JRT400
21 152 156 0.0034 0.0128 3.1247 1.0 0 10; %BGURI400-
TALA400
1 158 159 0.0114 0.0128 0.0009 1.0 0 10; %JALI66-JALII66
2 159 160 0.1027 0.1155 0.0081 1.0 0 10; %JALII66-
CHALSA66
3 160 161 0.1912 0.215 0.0038 1.0 0 10; %CHALSA66-
BANAR66
4 161 162 0.1096 0.1233 0.0022 1.0 0 10; %BANAR66-BRP66
5 160 163 0.131 0.1723 0.0065 1.0 0 10; %CHALSA66-
KPONG66 Zlinechng
6 164 163 0.0571 0.0642 0.0011 1.0 0 10; %MELLI66-
KPONG66
161 145 115 0 0.0132 0 0.95 1 10; %JRT400-JRT220 tap
chngori 1.05
162 146 116 0 0.0132 0 0.95 1 10; %ARAM400-
ARAM220 tap chng
ori 1.05
(Continued)
212 Appendix B

Bus tap
No. fb eb r x hlch Tap posl lmt
163 147 117 0 0.0198 0 1.0 1 10; %KTPS400-KTPS220
164 148 118 0 0.0198 0 1.0 1 10; %BAKR400-
BAKR220
165 150 119 0 0.0198 0 1.05 1 10; %PARU400-
PARU220 tap chng
ori 1.05
166 151 120 0 0.0198 0 1.0 1 10; %MALPG400-
MALPG220
167 152 121 0 0.0198 0 1 1 10; %BGURI400-
BGURI220 tap chng
ori −1
168 154 139 0 0.0198 0 1 1 10; %PURN400-
PURN220
169 149 142 0 0.0397 0 1 1 10; %FKK400-FKK220
170 155 143 0 0.0198 0 1.05 1 10; %SGPG400-
SGPG220
171 115 26 0 0.0208 0 1 2 10; %JRT220-JRT132 tap
chng ori 0.95
172 116 27 0 0.0313 0 0.95 2 10; %ARAM220-
ARAM132
173 117 28 0 0.0625 0 1 2 10; %KTPS220-KTPS132
174 120 29 0 0.0667 0 1 2 10; %MALPG220-
MALPG132
175 122 30 0 0.0208 0 1 2 10; %DRG220-DRG132
176 123 31 0 0.0313 0 1 2 10; %DOMJ220-
DOMJ132
177 124 32 0 0.0313 0 1 2 10; %GOK220-GOK132
tap chng
178 125 33 0 0.0625 0 1 2 10; %HOW220-HOW132
179 125 33 0 0.0333 0 1 2 10; %HOW220-HOW132
180 126 34 0 0.0333 0 1 2 10; %KASBA220-
KASBA132
181 126 34 0 0.0625 0 1 2 10; %KASBA220-
KASBA132
182 127 35 0 0.0313 0 1 2 10; %LXP220-LXP132
183 128 36 0 0.0313 0 0.95 2 10; %MIDNA220-
MIDNA132
184 129 37 0 0.0313 0 1 2 10; %NJP220-NJP132
185 130 38 0 0.0313 0 1 2 10; %NHAL220-
NHAL132
186 131 39 0 0.0313 0 1 2 10; %RISH220-RISH132
187 132 40 0 0.0313 0 1 2 10; %SATG220-
SATG132
188 133 41 0 0.1 0 1 2 10; %DPL220-DPL132
(Continued)
Appendix B 213

Bus tap
No. fb eb r x hlch Tap posl lmt
189 136 42 0 0.1 0 1 2 10; %BRPPG220-
BRPPG132
190 137 43 0 0.1 0 1 2 10; %SLGPG220-
SLGPG132
191 140 95 0 0.1 0 1 2 10; %STPS220-STPS132
192 117 28 0 0.0333 0 1 2 10; %KTPS220-KTPS132
193 136 42 0 0.2 0 1 2 10; %BRPPG220-
BRPPG132
194 134 112 0 0.0313 0 1 2 10; %WARIA220-
WARIA132
195 144 72 0 0.0313 0 1 2 10; %KRIS220-
KRISH132
196 111 164 0 0.2 0 1.05 2 10; %MELLI132-
MELLI66 tap chng
orig. -1.05
197 10 96 0 0.0324 0 1 2 10; %BTP11-BTPS132
tap chng original -1
198 11 96 0 0.02 0 1 2 10; %BTPS11-BTPS132
tap chng original -1
199 7 140 0 0.0179 0 1 2 10; %STP13.8-STPS220
tap chng original -1
200 8 133 0 0.0769 0 1 2 10; %DPL13.8-DPL220
tap chng original -1
201 12 41 0 0.0314 0 1 2 10; %DPL11-DPL132 tap
chng original -1
202 24 41 0 0.1143 0 1 2 10; %DPL6.3-DPL132
tap chng original -1
203 2 117 0 0.0135 0 1 2 10; %KT115.75-
KTPS220 tap chng
original1
204 3 147 0 0.0169 0 1 2 10; %KT215.75-
KTPS400 tap chng
originl-1
205 4 148 0 0.0253 0 1 2 10; %BKR15.75-
BAKR400 tap chng
originl-1
206 1 149 0 0.0112 0 0.95 2 10; %FKK21-FKK400
tap chng original-1
207 5 149 0 0.0177 0 0.95 2 10; %FKK15.75-FKK400
tap chng original-1
208 13 135 0 0.0268 0 0.95 2 10; %CHU11-CHUK220
tap chng original-1
209 9 101 0 0.1333 0 1 2 10; %RMM211-
RMM2132 tap chng
original-1
(Continued)
214 Appendix B

Bus tap
No. fb eb r x hlch Tap posl lmt
210 21 98 0 0.25 0 1 2 10; %TCF16.6-TCFI132
tap chng original-1
211 22 99 0 0.25 0 1 2 10; %TCF26.6TCFII132
tap chng original-1
212 23 100 0 0.25 0 1 2 10; %TCF36.6-TCF3132
tap chng original-1
213 14 158 0 0.1571 0 1 2 10; %JAL111-JALI66 tap
chng original-1
214 15 159 0 0.3125 0 1 2 10; %JAL211-JALII66
tap chng original-1
215 16 104 0 0.0889 0 1 2 10; %RANGIT11-
RANGT132 tap chng
orig.-1
216 53 162 0 0.1333 0 1.05 2 10; %BRP66-BRP132 tap
chng
217 17 141 0 0.0101 0 1 2 10; %MEJIA11-
141MEJIA220
218 6 118 0 0.0405 0 1 2 10; %BAKR15.7-
BAKR220
219 18 112 0 0.0485 0 1 2 10; %WARIA11-
WARIA132
220 19 112 0 0.028 0 1 2 10; %WARIA21-
WARIA132
221 25 157 0 0.0112 0 0.9875 2 10; %PPSP16.5-PPSP2
tap chng orig.-0.9875
222 39 194 0 0.0533 0 1.05 1 10; %RISH132-
RISHRA33 tap chng
orig.-1.05
223 36 201 0 0.127 0 1.05 1 10; %MIDNA132-
MIDNAP33 tap chng
orig.-1.05
224 54 195 0 0.08 0 1.05 1 10; %VISH132-
BISHNU33 tap chng
orig.-1.05
225 64 196 0 0.254 0 1.05 1 10; %EGRA132-
EGRA33 tap chng
original-1.05
226 64 196 0 0.4 0 1.05 1 10; %EGRA132-
EGRA33 tap chng
original-1.05
227 77 197 0 0.127 0 1.05 1 10; %NBU132-NBU33
tap chng original-1.05
228 86 198 0 0.0533 0 1.05 1 10; %SL132-SLAKE33
tap chng original-1.05
(Continued)
Appendix B 215

Bus tap
No. fb eb r x hlch Tap posl lmt
229 55 199 0 0.16 0 1.05 1 10; %BOLP132-
BOLPUR33 tap chng
orig.-1.05
230 55 199 0 0.2 0 1.05 1 10; %BOLP132-
BOLPUR33 tap chng
orig.-1.05
231 92 200 0 0.08 0 1.05 1 10; %TITA132-TITAG33
Tap chng
original-1.05
232 20 156 0 0.0156 0 0.9875 2 10; %TALA11-TALA400
tap chng orig.-0.9875
233 145 165 0 0.3333 0 1 1 10; %JRT400-220-JRT33
tap chng original-1
234 146 166 0 0.3333 0 1 1 10; %ARAM400-220-
ARAM33 tap chng
original-1
235 147 167 0 0.5 0 1 1 10; %KTPS400-220-
KTPS33 tap chng
original-1
236 148 170 0 0.5 0 1 1 10; %BAKR400-220-
BAKR33 Tap chng
original-1
237 150 169 0 0.5 0 1 1 10; %PARU400-220-
PARU33
238 151 171 0 0.5 0 1 1 10; %MALPG400-220-
MLDAPG33
239 152 172 0 0.5 0 1 1 10; %BGURI400-220-
BGURI33 tap chng
ori-1
240 154 173 0 0.5 0 1 1 10; %PURN400-220-
PURN33
241 149 168 0 0.5 0 1 1 10; %FKK400-220-
FKK33*trf tap has
been chng
242 155 202 0 0.5 0 1 1 10; %SGPG400-220-
SGPG33
243 115 174 0 0.0741 0 1 1 10; %115JRT220-132-
JRT133
244 116 175 0 0.1111 0 1 1 10; %116ARAM220-132-
ARAM133
245 117 176 0 0.2222 0 1 1 10; %117KTPS220-132-
KTPS133
246 120 177 0 0.2222 0 1 1 10; %120MALPG220-
132-MLDPG133
(Continued)
216 Appendix B

Bus tap
No. fb eb r x hlch Tap posl lmt
247 122 178 0 0.0741 0 1 1 10; %122DRG220-132-
DRG33
248 123 179 0 0.1111 0 1 1 10; %123DOMJ220-132-
DOMJ33
249 124 180 0 0.1111 0 1 1 10; %124GOK220-132-
GOK33
250 125 181 0 0.2222 0 1 1 10; %125HOW220-132-
HOW33
251 125 181 0 0.1111 0 1 1 10; %125HOW220-132-
HOW33
252 126 182 0 0.1111 0 1 1 10; %126KASBA220-
132-KASBA33
253 126 182 0 0.2222 0 1 1 10; %126KASBA220-
132-KASBA33
254 127 183 0 0.1111 0 1 1 10; %127LXP220-LXP33
255 128 184 0 0.1111 0 1 1 10; %128MIDNA220-
132-MIDNA33
256 129 185 0 0.1111 0 1 1 10; %129NJP220-132-
NJP33
257 130 186 0 0.1111 0 1 1 10; %130NHAL220-132-
NHAL33
258 131 187 0 0.1111 0 1 1 10; %131RISH220-132-
RISH33
259 132 188 0 0.1111 0 1 1 10; %132SATG220-132-
SATG33
260 133 189 0 0.3333 0 1 1 10; %133DPL220-132-
DPL33
261 136 190 0 0.3333 0 1 1 10; %136BRPPG220-
132-BRPPG33
262 137 191 0 0.3333 0 1 1 10; %137SLGPG220-
132-SLGPG33
263 140 192 0 0.3333 0 1 1 10; %140STPS220-132-
STPS33
264 117 176 0 0.1111 0 1 1 10; %117KTPS220-132-
KTPS133
265 136 190 0 0.6667 0 1 1 10; %136BRPPG220-
132-BRPPG33
266 134 193 0 0.1111 0 1 1 10; %134WARIA220-
132-WARIA33
267 144 203 0 0.1111 0 1 1 10;] %144KRIS220-132-
KRIS33
%47 129 121 0 0 0 0.0004 1 10;

%%total line = 160 + 35(line trf-3 wdg) + 37 (load trf-2wdg) + 35 (for tertiary dg) = 267 %but here total
line 266, line 47 has removed due to it zero imp %Lines are arrange acc to bus voltage level.
 Appendix B
217

FIGURE B.1 WBSEB’s 203-bus system.

Index
anaerobic digestion 175 deterministic complexity 6
acid formation 175 diagonal entry 87
acyclic 12 digraphs 12
adjacency matrix 13 directed graphs 12
aggravated complexity 6 distributed energy resource (DER) 126
algorithmic complexity 6 distributed generation (DG) `169, 178, 179, 180,
artifcial life 6 182, 184, 186
Automatic Generation Control 3 distributed renewable energy sources 169
Automatic Voltage Control 3 double line to ground (L-L-G) 9, 25, 64
average shortest path length 110
edges 11
Barabasi-Albert network 18 effciency 109, 117
betweenness 109, 110, 116, 139 electrical betweenness 139, 140, 141
betweenness distribution 8 electrical centrality 112, 119, 120, 180
biochemical conversion 174 electrical distance 109, 112
bus current injection 107 electrical line betweenness 140
bus impedance matrix 85 energy estimation 173
exponential cumulative degree function 18
cascading failure 4, 135, 144 extended electrical betweenness 131
cellular automata 6
centrality 110 FACTS 3
chaos theory 6 False Data Injection Attack (FIA) 11, 21
charge controller 171 fault 9
clustering 7
column vector of initial bus currents 85 Gaussian distribution 8
column vector of initial bus voltages 85 genetic algorithm 6
combined cycle turbines 169 geodesic distance 109
combustion 174 geodesic length 110
combustion turbines 169 geodesic path 110, 139
complex networks 8 Global Average Power Supplying Effciency
complex network theory 1, 8 Index 126
complex power 45 global effciency 117, 124
complex transformation ratio 30 grid short circuit power 174
complexity 5, 6, 7, 11, 109 grid short circuit ratio 174
connectivity 111 Grid Vulnerability Index (GVI) 125, 180, 183, 184
contingencies 4, 11
contingency analysis 10, 11, 85, 100, 107, 123 HVDC 3
contingency procedure 10 hydrolysis 175
Contingency Vulnerability Index (CVI) 125
criticality 6 incineration 174
current injection 96, 107 inverter 171
current injection distribution factor 96 islanding 9, 135, 139
cyber attacks 11, 20, 21
cyber security 20 Kron’s reduction 54, 105

DC power fow 103, 104, 135 line betweenness 114

degree 109, 110, 111 line outage factor 104
degree distribution 8, 18, 159, 161 line to ground (L-G) fault 9, 25, 64
densifcation 174 line to line (L-L) fault 9, 25, 64, 68
Depth of Discharge (DOD) 171, 172 link 11

219
220 Index

magneto hydrodynamic (MHD) 177 security monitoring 9

mean degree 18 self-organization 6
mean wind speed 173 semi-random networks 19
methane formation 175 sequence impedance 42
Molloy-Reed criterion 158 sequence reactance 42
municipal liquid waste 175 single phase to ground fault 66
municipal solid waste (MSW) 175 small diesel generators 169
small world networks 7, 8, 18, 108
negative phase sequence 62 standard operating cell temperature
negative sequence 37 (SOCT) 170
negative sequence current 64 structural vulnerability 2, 109
negative sequence impedance 42, 45 Structural Vulnerability Index (SVI) 125
negative sequence network 64 sub-transient reactance (Xd’’) 42
netability 109, 119, 124, 129 Supervisory Control and Data Acquisition
node betweenness 114 (SCADA) 21
North American Electric Reliability Corporation symmetrical component 25, 37
(NERC) 138
thermochemical conversion 174
Ocean Thermal Energy Conversion (OTEC) 176 Thevenin’s equivalent circuit 97
open circuit voltage 92 Thevenin’s equivalent impedance 85, 88
open conductor fault 81 three phase balanced fault (L-L-L-G) 25,
Operational Vulnerability Index (OVI) 125 64, 65
Overload Cascade model 136, 137 topological structure 13
transfer admittances 27
percolation theory 157, 160 transient reactance (Xd’) 42
percolation threshold 157
positive phase sequence 62 undirected graph 13, 161
positive sequence 37, 43 undirected networks 12
positive sequence impedance 42, 45 Union for the Coordination of Transport of
positive sequence network 42 Electricity (UCTE) 18
positive sequence reactance 81 unweighted graph 13
power density duration curve 173 urban industrial waste 175
power law degree distribution 8
power law distribution function 18 Von Bertalanffy’s system theory 6
power law probability distribution 20 vulnerability 4, 10, 139, 182
power system reliability indices 18 vulnerability analysis 8
power transfer capacity 137
power transfer distribution factor (PTDF) 115, Weibull probability distribution function 173
116, 118, 140 weighted graph 15
power transmission capability 118 weighted networks 12
preferential attachment 7 wide area monitoring systems 11
principle of superposition 85 wind energy generators (WEG) 174

random networks 7, 8, 18, 109 [Ybus] 50

real time system status 11
regular networks 7, 8 zero phase sequence 62
resilience 150, 161 zero sequence component 41
resiliency 4, 151 zero sequence current 42, 64
robustness 19, 150 zero sequence impedance 42, 44, 45
zero sequence network 42, 64
scale free network 7, 8, 18, 19, 109 zero sequence reactance 81
security assessment 10 [Zbus] 50