09/12/2013

UNIT-I
A FRAME-WORK FOR APPLICATION OF COMPLEX NETWORK THEORY
TO
VULNERABILITY ANALYSIS OF INDIAN POWER GRID SYSTEM

Abstract: Complex Network Systems have been extensively used to analyze network
based Infrastructures like Power-Grids, Communication Networks, Internet and Transport
Networks covering Road, Rail, Air and Waterways. This is different from analysis of other
Information networks where links represent some kind on conceptual relation or
interaction. Considerable research has been done on how to adapt these metrics to
Power-Grid domain specific attributes. In this paper, we examine how the same can be
extended/ customized to the scenario of Indian Power Grid System
1. Introduction
1.1 Overview: In present times, when internal terrorist attacks, short and intense fights
dominate the conflict landscapes, protection of key Critical Infrastructure will become
very much important. Most of the Infrastructure elements by definition will be
structured in Network Topology, of Nodes and Connections. Here edges between the
nodes are hard Power/Information/Fuel carrying lines rather than conceptual relations
and interactions. Properties of such Networks can be derived using Complex Network
Theory. Much work has been done already in analyzing such Networks especially “Power
Grids” using Complex Network System at International Level. In this assignment, a
Framework is developed for doing a similar analysis in Indian Context.
In this section, basic background material from both Complex Network Theory and
Power System Domain are introduced. Also, a brief Survey of Applications of Complex
Network Theory in analyzing Power-Grid is presented. Section 2, gives a high level-
overview of current Indian Power-Grid Scenario in detail. In Section 3, A framework is
developed about what should be the key questions that need to be answered, what data
needs to be collected from where and what is the method for arriving at the result.
Section 4, presents an illustrative example of the approach. Finally, in section 6, we
conclude, stating the summary of the framework, its advantages and limitations with
possible improvements.

1.2 Complex Network System: A network is a set of items, which we will call vertices or
sometimes nodes, with connections between them, called edges (Figure 1.0). Systems
taking the form of networks (also called “graphs” in much of the mathematical
literature) abound in the world. Examples include the Internet, the World Wide Web,
social networks of acquaintance or other connections between individuals,
 organizational networks and networks of business relations between companies, neural
networks, metabolic networks, food webs, distribution networks such as blood vessels
or postal delivery routes, networks of citations between papers, and many others (Figure
1.0).

A Network is said to be complex when the nodes and edges are so numerous that computation of its
properties through conventional methods will be prohibitively expensive. In such cases, we need to
resolve to statistical measures to derive its properties.
Following Terminologies are used in rest of the discussion,

Erdos Renyi Random Graphs: We can construct a specific type of random graph simply by taking N
nodes, also called vertices and by drawing Nz/2 lines, the edges, between randomly chosen pairs of
nodes. This type of random graph is called an ‘Erd¨os-R´enyi random graph’ after two mathematicians
who studied this type of graphs extensively

Generalized Random Graphs: Random Graphs in which the degree-distribution is non-poisson are called
generalized random graphs.

Small-World Networks:

Small World Network is one which exhibits ‘small world effect’. The ‘small world effect’ denotes the
result that the average distance linking two nodes belonging to the same network can be orders of
magnitude smaller than the number of nodes making-up the network. Two famous Small World Network
Models are prevalent.
The Watts and Strogatz model

Watts and Strogatz have proposed a small world model which interpolates smoothly in between a
regular lattice and an Erd¨os-R´enyi random graph. The construction starts with an one-dimensional
lattice. One goes through all links of the lattice and rewires the link with some probability p.

The Newman and Watts model

A variation of the Watts–Strogatz model has been suggested by Newman and Watts. Instead of rewiring
links between sites i.e., extra links, also called ‘shortcuts’, are added between pairs of sites chosen at
random, but no links are removed from the underlying lattice. This model is some-what easier to analyze
than the original Watts and Strogatz model, because it is not possible for any region of the graph to
become disconnected from the rest, whereas this can happen in the original model.

Scale-Free Networks
Networks with power-law degree distributions have been the focus of a great deal of attention in the
literature. They are also referred to as scale-free networks, although it is only their degree
distributions that are scale-free; one can and usually does have scales present in other network
properties.

Scale Free Networks fare better for Random Failures of Vertices. However, in Biased-Failures (planned
attacks), scale free network collapses very rapidly than random.

1.2 Power Systems And Power Grids: Electrical engineers are concerned with every step in the
process of generation, transmission, distribution, and utilization of electrical energy. The electric utility
industry is probably the largest and most complex industry in the world. The electrical engineer who
works in that industry will encounter challenging problems in designing future power systems to deliver
increasing amounts of electrical energy in a safe, clean, and economical manner.

POWER SYSTEM OVERVIEW

Electric power systems are real-time energy delivery systems. Real time means that power is generated,
transported, and supplied the moment you turn on the light switch. Electric power systems are not
storage systems like water systems and gas systems. Instead, generators produce the energy as the
demand calls for it.
Figure below shows the basic building blocks of an electric power system. The system starts with
generation, by which electrical energy is produced in the power plant and then transformed in the
power station to high-voltage electrical energy that is more suitable for efficient long-distance
transportation. The power plants transform other sources of energy in the process of producing
electrical energy. For example, heat, mechanical, hydraulic, chemical, solar, wind, geothermal, nuclear,
and other energy sources are used in the production of electrical energy. High-voltage (HV) power lines
in the transmission portion of the electric power system efficiently transport electrical energy over
long distances to the consumption locations. Finally, substations transform this HV electrical energy into
lower-voltage energy that is transmitted over distribution power lines that are more suitable for the
distribution of electrical energy to its destination, where it is again transformed for residential,
commercial, and industrial consumption. A full-scale actual interconnected electric power system is
much more complex than that shown in Figure below;
The simple point to point transmissions between Generation and Consumption Stations shown above
will in fact be a complex Power-Grid which will be dense collection of Station, High-Voltage Power Lines,
Sub-station and Distribution Lines. These will be dealt in detail in subsequent sections.

1. Power Generators: Major components associated with electric power generations are Motors
and Generators. Prime Movers in various ways Hydraulic, Nuclear, Wind etc., finally move these
rotors for Power Generation. When they are combined with appropriate Turbines and primary
movers, complete Power Plant design will emerge.
2. High-Voltage Transmission Lines: High Voltage power lines are always preferred to Low-Voltage
ones. The fundamental components of transmission lines consist of conductors, insulators, air
gaps and shielding.
3. Substations: Substations transform very high voltage electrical energy into a more useable for
distribution and consumption. Different equipment that will be used is Transformers, Circuit
Breakers, Disconnect Switches, Regulators, etc.)
4. Primary Distribution System: These will consist of both overhead and underground, are
designed, operated and used to serve residential, commercial and, industrial consumers. The
distribution system between substation and the consumer’s demarcation point (i.e., service
entrance equipment) will come under the purview of Primary Distribution System
5. Secondary Distribution System: The distribution will consist from Consumer’s demarcation
point to all the actual loads.
1.3 Application of CNS In Analysis of Power Grids

Different components of Power System will play part of Nodes and Edges while modeling them as
Network. Essentially Transmission Lines will be treated as edges and all other entities like Station, Sub-
Station, Generator and Consumer End-points will be considered as Vertices.

Application of Complex Network Theory to Power Grid, especially to US and Chinese has been
extensively studied and published in literature. There are three distinct types of analysis can be seen.

[a.] Study of Network Properties

Much study has been done to characterize Power Grid in terms of Complex Network Properties. Very
commonly used properties are

1. Degree of a Vertex which maps to Hubs/Bus with maximum number of Transmission Links
2. Degree Distribution: Mapping of Number of Vertices Pk as a function of degree k
3. Cliquishness coefficient: How “lumped” is the Network?
4. What is diameter of the Network?
5. What is Average shortest path length between two vertices in the graph?
Naturally, all these questions lead to say whether the grid exhibits “Small World” effect or not? If it
exhibits, the grid is treated to be robust otherwise not.

[b.] New Metrics on Centrality and Between-ness

Jawad, Muhammad Gou, Bel Applications of Complex network theory on power grids. A new set of
metrics based on centrality are introduced. They address concepts like which are the central points of
the grid, from geodesic point of view, from load point of view, what is the in-between mid-point for given
two points etc.,

[c.] Vulnerability Analysis

Power Grids are also analyzed from the aspect of Vulnerability. How Robust is the Network? How
susceptible it is to either to Random failures or biased failures (planned attack)? What is the resilience?
How blackouts are distributed or propagated? Since blackout of one node increases load on another,
how to model cascade failures? These are some of the aspects that are studied.

Here one step more is advanced than normal parameters. A Concept called “efficiency” which in effect
depends inversely on dij or shortest distance between two vertices is introduced. This efficiency will be
in terms of Total Network rather than at a local component of it. When an attack happens, efficiency
naturally decreases due to damages. Now, given G0 efficiency before attack and G f after attack,
we can write

G0 – G f
D=
G0

Where D quantifies Damage. In recent US blackouts it is shown that damage was around 40%. In a recent
paper by Arios *, practical considerations power grid are being considered. Instead of treating them as
mere graphs/networks, their physical properties and their impact are given importance. 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 flows through the lines of the path. As a result, we define the
Electrical distance as,
f lk Z
l
¿
d ij =∑ ¿
k

l ∈k

l
Where f K is the power transmission distribution factor of line l path k and Z l its impedance.

Ke Sun [*] summarizes following possible direction in future study in following five points

1. Complex Network Theory and Real World Power System need to be combined more compactly.
Since they are two parallelly developed field shared or even corresponding vocabulary is not
very directly available. It has to be built.
2. Identification and Characterization of Network Patterns in general and in field of Power Grids in
particular is far from level of maturity. It needs to be done.
3. Sophisticated Models describing structure (robustness etc.,) properties of Power Grid should be
developed. As of now, Small-World model is only the well known model.
4. Fourth, some processes taking place in network, such as percolation, cascading process, phase
transition etc, need to be further explained and analyzed in power grid.
5. Fifth, the work to quantify the interaction between power grid and other networks is to be done
as soon as possible.
2.0 Indian Power Grid System

Indian Power Grid spans 1,02,604 ckt kms of transmission lines at 800/765kV, 400kV, 220kV & 132kV
EHVAC & +500kV HVDC levels and 173 sub-stations.

Also the transformation capacity of about 1,79,323 MVA as on 30th November 2013. This gigantic
transmission network, spread over length and breadth of the country, is consistently maintained at an
availability of over 99%.

 ±800Kv,6000 MW Multi-terminal HVDC System (under construction) :

 Bi-pole of length around 2000 kM from Biswanath Chariali in Assam (North-eastern

Region) to Agra in Uttar Pradesh (Northern Region)on its completion shall be one of the
longest HVDC lines in the world.

 Bi-pole of length around 1500 kM from IPPs in Champa Complex in Chattisgarh to

Kurushetra in Northern Region.

 1200kV UHVAC Transmission System:

 Facilitated Indigenous development of 1200kV UHVAC through establishment of

1200kV National Test Station at Bina under PPP model in collaboration with 35 Indian
manufacturers.

 Switchyard (Bay-01) equipment’s, Single and Double circuit line in Test station were
successfully charged at 1200kV level in 2012.
Power Grid Corporation of India limited is an Indian state owned Electric utilities
company, whose head quarter is located at Gurgaon, Punjab.

It is responsible for management of about 50% of the total power generated in India on
its transmission network. It has presence all-over India and around 95,329 circuit- km
of transmission network, along with 156 EHVAC & HDVC sub-stations with an enormous
power transmission capacity of 138,673 MVA.

Power grid has branched itself into telecom business with a network of 25000km across
the country.

India is divided into 5 regions. The eastern region(ER), the western region (WR), the
northern region (NR) and north-east region (NER) are under the new grid, whereas the
Southern region (SR) is not synchronized like the rest.
Power Outage Crisis In India

Indian Power Grid was also victim of cascading power failures. Here is a quote from
MBA brief report (http://www.mbarendezvous.com/postinner.php?id=550)

“On 30 and 31st July 2012, India faced the largest power outage in the history. It
blackout affected about half of India’s population / 9% of the world population, which
spread across 22 states in India. 32 gig watts of generating capacity were taken offline
due to this outage; about 300 million people had no electricity at all for at least 8
hours.

The busiest airports and railways in the northern region were completely shut down.
Traffic signals were non-operational which caused chaos in the minds of people. Major
hospitals across the states, faced interruptions in health services, most of them relied
on back-up generators. Water treatment plans were totally off due to which millions
were unable to draw water by the help of electric pumps. Over 200 miners were
trapped underground due to lift failure. More than 300 intercity trains were called off.
Delhi, Kolkata’s metro were shut down. Half of India was completely paralyzed.

The failure was caused due to late monsoon arrival and extreme heat in Delhi. Haryana
and Punjab, the two major agricultural states drew more power for running irrigation
pumps for paddy fields, due to late monsoon arrival, the Hydro-power plans were
completely inactive.

Firstly Bina- Gwalior line tripped, as this line was connected to Agra-Bareilly
transmission section, it also tripped and the chain continued to the havoc power loss.
All major power stations were shut down. It took 15 hours to restore 80% of the
services.”

An Analysis of this Outage using Complex Network Theory is done by [*]

3.0 Framework for Applying CNS Model to Indian Power Grid System
 Earlier Work done
Surprisingly very little analysis work is available on Indian Power Grid. Following
three were particularly interesting:

Fairly detailed account and analysis of happening of 30th and 31st Power failure
by Randhir is available on Slide share (http://www.slideshare.net/randhir2k/analysis-
of-indian-power-grid-failure). However, it is more of specific to the event and Post-
Mortem nature, than overall architecture analysis.
 A detailed analysis of the event using Complex Network System by Guidong Zhang
et. al., in [*]
http://www.sciencedirect.com/science/article/pii/S0378437113002069

A recent paper applies CNS to Power Grid of Orissa State, and analyzes its topology. This
is limited to Odessa region and tries to adapt Grid Computing Techniques more than Complex Network
Techniques. [*] [http://www.ijmca.org/index.php/ojs/article/view/14]
a. Key Questions to be answered…
a. What are the properties of the GRID?
b. What are the crucial Nodes(bus/station) from different perspectives
c. How robust is electrical connectivity? How resilient it is to Random as well
as Biased fairlures?
d. How to explain/understand cascading failures/blackouts?
e. How to evaluate/compare two alternate proposals?
f. How to Optimize Power Generation and/or Utilization?
b. Method
a. Collect details of Generators, Stations, Substations and Busses
http://en.wikipedia.org/wiki/List_of_power_stations_in_India
http://powermin.nic.in/transmission/major_grid_substations.htm
b. Collect details about Transmission Lines
http://www.cea.nic.in/reports/powersystems/0910/prog_11_12.pdf
c. With Points mentioned in “a” as Nodes and “Transmission Lines”
mentioned in b as edges construct a Graph or Network.
d. Computer Network Parameters of Indian Power Grid
i. Degree, Degree Distribution, Cluster Index
ii. Diameter, Average Path Length
iii. Centrality, Betweenness
iv. Efficiency, Net-Ability
 Interpret the GRID’s characteristics in terms of these of metrics with the help of
Complex Network Theory
4.0 Conclusion:
Sufficient groundwork has been done in the area of Complexity Network Theory and its
application in domain of Electrical Power Grids. However, as Ke Sun has pointed out
most of the work has been in ready application of CNT to PNG by making simplifying
assumption that Power Grid is just another Network. Only since very recently a serious
attempt is being made to build Power Grid Specific Abstractions to apply in CNS,
efficiency and net-ability can be said two prominent examples. Coupling these two
domains more strongly with stronger shared vocabulary is an important work that needs
to be taken up.
 Coming to Indian context, very little has been investigated in Indian Power Grids from
point of view of Complexity Network Theory. This problem needs to be studied in depth

5.0 Bibliography

1. Jawad, Muhammad, Gou, Bel “Applications of complex network theory on power grids”
–Institute of Electrical and Electronics Engineers May 9, 2013
2. Guo, Yifei; Cao, Jin; Duan, Rong; Li, Sheng “Power Grid Vulnerability Identifying Based on
Complex Network Theory” Institute of Electrical and Electronics Engineers— Dec 8, 2012
3. Boccaletti, S.; Latora, V.; Moreno, Y.; Chavez, M. Hwang, D.-U. Complex Networks:
Structure and dynamics, Physics Reports , Volume 424 (4) – Feb 1, 2006
4. Sun Ke “Complex Networks Theory: A New Method of Research in Power Grid” IEEE/PES
Transmission and Distribution Conference & Exhibition: Asia and Pacific Dalian, China
2005
5. Arianos E. Bompard et.al.,Power grid vulnerability: a complex network approach Chaos 19,
013119 (2009)

6. Das, Mishra, Roy “The Topological Structure of the Odisha Power Grid: A Complex
Network Analysis”.
7. Guidong Zhanga, b, Zhong Lib, Bo Zhanga, Wolfgang A. Halang Understanding the
cascading failures in Indian power grids with Complex Networks Theory” Physica A:
Statistical Mechanics and its Applications
Some Common Terminologies Used in Complex Networks Theory

Vertex (pl. vertices): The fundamental unit of a network, also called a site (physics), a node (computer
science), or an actor (sociology).

Edge: The line connecting two vertices. Also called a bond (physics), a link (computer science), or a tie
(sociology).

Directed/undirected: An edge is directed if it runs in only one direction (such as a one-way road between
two points), and undirected if it runs in both directions. Directed edges, which are sometimes called
arcs, can be thought of as sporting arrows indicating their orientation. A graph is directed if all of its
edges are directed. An undirected graph can be represented by a directed one having two edges
between each pair of connected vertices, one in each direction.

Degree: The number of edges connected to a vertex. Note that the degree is not necessarily equal to the
number of vertices adjacent to a vertex, since there may be more than one edge between any two
vertices. In a few recent articles, the degree is referred to as the “connectivity” of a vertex, but we avoid
this usage because the word connectivity already has another meaning in graph theory. A directed graph
has both an in-degree and an out-degree for each vertex, which are the numbers of incoming and
outgoing edges respectively.

Component: The component to which a vertex belongs is that set of vertices that can be reached from it
by paths running along edges of the graph. In a directed graph a vertex has both an in-component and
an out-component, which are the sets of vertices from which the vertex can be reached and which can
be reached from it. Geodesic path: A geodesic path is the shortest path through the network from one
vertex to another. Note that there may be and often is more than one geodesic path between two
vertices.

Diameter: The diameter of a network is the length (in number of edges) of the longest geodesic path
between any two vertices. A few authors have also used this term to mean the average geodesic
distance in a graph, although strictly the two quantities are quite distinct.

Also, following Network Models will be generally used in Complex Network Systems. *