1

CHAPTER 1

INTRODUCTION

1.1 GENERAL

Power System industry is now increasingly becoming privatized

and hence the system data is becoming increasingly distributed with complex
operational and control requirements. Secure operation of power systems
requires close monitoring of the system operating conditions. This is
traditionally accomplished by the state estimator which resides in the control
centre and has access to the measurements received from numerous
substations in the power system network. Real time monitoring of power
systems has become possible with the advent of Phasor Measurement Units
(PMU). It gives the real time status of the power system operating conditions,
which is required for power system analysis and control. PMU measures
voltage and current phasors in power systems, which has higher accuracy than
conventional measurements.

Synchronism among phasor measurements is achieved by sampling

of voltage and current waveforms using a common synchronizing signal from
the Global Positioning Satellite (GPS). Phasor measurements reduce the
effects of time-skew and are useful for many other applications such as
system protection, control and stability assessment, aid topology error
identification, parameter error detection and correction and improve the
accuracy of state estimation. PMUs facilitate innovative solutions to
traditional utility problems and offer power system engineers a whole range
 2

of potential benefits, including precise estimates of the power system state

that can be obtained at frequent intervals, enabling dynamic phenomena to be
observed from a central location and appropriate control actions taken. It is
neither economical nor necessary to install a PMU at every node of a wide-
area interconnected network. This is because the cost of a PMU depends on a
number of factors, including the number of measuring terminals (channels),
current transformer and potential transformer connections, power connection,
station ground connection, and GPS antenna connection. However, what
really distinguishes among different PMU costs is the number of channels,
since the remaining items are common to all PMU installations.

As the size of the power system is becoming larger and larger due
to the increase in demand, the interconnections between large power systems
may vary time to time due to addition of new generating units and due to
geographic conditions and hence it becomes difficult to find the optimal
location for installing PMUs for the real time electric power system network.
State estimation results become more accurate with the PMU data received
from the optimally located PMUs in a power system network. The
deregulation policy allows the geographically apart private power system
networks to join together or leave based on the demand requirements. This
dynamically changing environment initiated the need for interoperable,
scalable and secure state estimation services. In this thesis, an innovative
distributed service model is introduced for optimal placement of PMUs using
power system reduction technique and cost effective cloud services are
proposed for power system state estimation and optimal PMU placement.

1.2 STATE OF THE ART

State estimation is the main composition of the modern Energy

Management Systems (EMS). Since the concept of state estimation was
introduced by Schweppe and Wildes (1970), many methods have been
 3

proposed to calculate the state vector of the power system. The state of an
electric power system is defined as the vector of voltage magnitudes and
angles at all network buses. The state estimator is a data processing algorithm
for converting redundant meter readings and other available information into
an estimate of the state vector. The general nature of the problem,
mathematical modelling, an interactive technique for calculating the state
estimate, and concepts underlying the detection and identification of
modelling errors are discussed. Schweppe and Rom (1970) has developed an
approximate mathematical model for solving state estimation problem. The
proposed approximate mathematical model related to the dc load-flow model
yields non-iterative state estimation equations, simplified prediction of effects
of network and generation-load pattern changes on network flow and
simplified detection and identification of modelling errors.

Schweppe (1970) had discussed the implementation problems

related to the computational speed and storage, dimensionality of the state
vector and the time varying nature of the power systems. The actual
implementation of the state estimation, detection and identification algorithms
depends on the type of power system, the type of metering and
communications, the available human, hardware and software resources and
the uses to be made of the static estimate. In order to know the state of the
power system at any time, the state vector has to be calculated at regular
intervals.

Monticelli (2000) explains that the state estimation is a key

function for building a network real time model within Energy Management
Systems. He has reviewed the principal developments in state estimation and
related areas such as observability analysis, erroneous data processing,
network topology processing, topology estimation, and parameter estimation.
Jun Zhu (2007) revisits the state estimation problem formulation by assuming
 4

availability of at least one phasor measurement unit in the system. The author
investigates on the requirements to ensure robust state estimation in the
presence of single PMU errors. These requirements are verified by
implementing a GPS referenced state estimator using test systems containing
one or more PMU measurements.

Kun Zhu et al (2009) has analyzed the optimal PMU placement

methods with application to state estimation accuracy. He has studied the
impact of input uncertainties on the robustness of power system state
estimation techniques. The estimation accuracy improvement contributed by
the proposed optimum placement algorithm is compared with that obtained by
the adopted PMU placement method in the Svenska Kraftnat (Swedish
Transmission System Operator) based on the assumed SCADA measurement
configuration. By applying the proposed strategic placement method, it can
be observed that there is a major angle estimation improvement when more
PMUs are deployed until the PMU measured buses reaches a certain number
and further introduction of PMU is not as effective. Extended Nordic 32
model has totally 41 buses and if 12 PMUs are placed, it provides 30 percent
converge of the network. The simulation shows that accuracy of the
estimation can be improved by strategically placing PMU at up to 30 percent
of the network buses however adding more PMUs will bring insignificant
improvement, given the current measurement configuration and accuracy
assumption. As for the angle estimates, it can be predicted that there is also
one threshold for magnitude estimates beyond which the further introduction
of PMUs has limited impact. From the analysis made, it has been found that
when the PMU monitored buses reaches a certain number in the entire
network, the further introduction of phasor measurements is not effective in
estimation accuracy improvement.
 5

An effective state estimation solution algorithm using phasor

measurement data has been proposed by Phadke (1993). Xue et al (2007)
have proposed a dynamic algorithm for solving power system state estimation
problem using both SCADA and PMU data. With the analysis using a test
case, the authors indicate that the method not only shortens the computer time
but also increases the precision of the estimates obtained.

A unified approach has been proposed by Nabil and Hanafy

Mahmoud (2009) to determine the optimal number of PMUs and their
locations in order to make the system completely observable and hence can be
used for power system state estimation. The PMU placement problem is
formulated as a binary integer linear programming problem, in which the
binary decision variables (0, 1) determine whether to install a PMU at each
bus, while preserving the system observability and lowest system metering
economy.

Chen and Lu (2002) demonstrated the potential advantages of the

Web as the platform for developing and deploying complex power system
simulations by using the distributed technologies and model-view-controller
concepts. Nithiyananthan and Ramachandran (2004) have developed an
effective RMI based distributed model to monitor the load flow and economic
dispatch for multi area power systems. RMI based client-server architecture
overcomes the difficulties associated with sequential computation of power
system operations. Santo et al (2004) have discussed that the phenomena that
can compromise power systems operation need to be carefully analyzed in
order to evaluate their impact on the security and reliability levels of the
electrical networks. The real time assessment of the system’s security and
reliability levels, especially under unforeseen contingencies, is known as on-
line power system security analysis. For complex networks this process
 6

requires large computational efforts whereas computation time should be less

than a few minutes for the information to be useful. To address this problem a
distributed architecture based on the Web is proposed for power system
security analysis. The architecture integrates a network of remotely controlled
units distributed in the most critical sections of the electrical network for field
data acquisition and safety check violations, a distributed solution engine for
the on-line analysis of the system security, and a Web-based interface for
graphical synoptic and reporting had been developed. The results obtained
from an intensive experimentation demonstrate the validity of the architecture
and stimulate the enhancement of the solution engine through the use of a
computational grid able to dynamically acquire the needed resources.

Deregulation and competitive markets for electricity have changed

the organizational structures of the electricity supply industry as well as the
operation of power systems. Chen et al (2006) have identified new
information integration requirements related to operation, marketing and
maintenance of a power system. A Web services infrastructure for
information integration in power systems has been developed that provides an
open, flexible and scalable framework with higher co-operation and
integration capability. XML is used for standardizing the data formats and
exchanging data among different platforms. Cross platform inter-application
communication support among different power system services is provided by
Simple Object Access Protocol (SOAP). A Web Services based SCADA
(WS-SCADA) system has been considered for the proposed infrastructure for
information integration in power system as a case study. At present, the
power system operations are highly complex and the power system services
are widely distributed. The power system services are not only developed
using different platforms but also deployed in various distributed
environments for their execution.
 7

With all the developments, it becomes necessary to develop a

model that supports interoperability between services on different platforms
in a reliable, scalable and inherently adaptable manner. The computing world
is rapidly transforming towards developing software for millions to consume
as a service rather than creating software for millions to run on their PCs
(Buyya et al 2008). Cloud computing can be classified as a new paradigm for
integrating with electric systems in which the services are readily available
and universally visible and the clients can access the services anytime and
anywhere in a distributed environment. Only a distributed environment
provides an opportunity for making the power delivery more efficient,
whether by minimizing the personnel visits to transmission and distribution
locations or by enabling better decisions through more timely and best
estimate of the power system network. Scalability, seamless data sharing and
interoperability stay as key factors for choosing distributed environment for
providing power system solutions.

1.3 OBJECTIVES

Phasor Measurement Units have opened a new perspective for

power system state estimation and offers new opportunities in power system
monitoring, protection, analysis and control. Post-disturbance analyses are
much improved due to precise snapshots of the system states, which are
obtained through GPS synchronization. Advanced protection could be
implemented based upon synchronized phasor measurements with options for
improving overall system response to catastrophic events. Advancements in
computer technology and availability of GPS signals have steadily reduced
the time frame of synchronized information from minutes, to seconds,
milliseconds and now microseconds.
 8

The optimal placement of PMUs plays a vital role for proper power
system planning, operation and control. Utility of the PMU data can be
viewed in three perspectives namely, they take automatic actions in case of
real time applications, provide guidance for near real time applications and
they help in proper maintenance, planning and event analysis for off-line
applications. There is a substantial contribution of PMU data for the
enhancement of power system state estimation. Usually, once the PMU could
make the power system completely observable, the power system state
estimation would become very simple and when used along with SCADA
data, it gives more optimum results.

The optimal placement of phasor measurement units is an off-line

problem to be solved during the planning stage and the results obtained such
as number of PMUs to be installed and their locations are considered as
planning data. The optimal placement of phasor measurement units is a vital
problem that needs to be solved before finding the solution for power system
state estimation problem. The main objective of this research work is to
develop enhanced anytime anywhere accessible services for finding optimal
locations for placement of phasor measurement units and to solve power
system state estimation problem. As PMU technology is fully deployed and
as communication latencies are reduced, accurate system state models can be
determined very rapidly. The PMU derived state of the power system allows
detection of unfeasible operating conditions and leads to the development of
control algorithms for isolation and islanding of the network components,
thus preserving the integrity of the overall power grid.

The research is focused on the development of generalized service

oriented model, which is flexible and customizable for solving exclusively
power system problems. It is proposed to overcome the deployment
overheads for the power system service providers in the deregulated
 9

environment. The proposed model completely decouples the power system

clients from the development environment and the power system service
providers from the deployment environment.

In order to represent the power system data in a common format, it

is proposed to introduce a generalized XML schema to represent power
system data. XML schemas are introduced to acquire complete power system
data and to extract the data from the stored information as per the service
requirements. To achieve this, a dynamic XML data generation model is
proposed. It is aimed to implement the proposed innovative, scalable and
interoperable distributed model for finding the optimal number of phasor
measurement units and their locations in a power system network using power
system reduction technique. A non-linear tree structure is proposed by ranking
the nodes of the power system network using the connectivity strength and the
importance of the nodes with respect to other nodes. A height balanced AVL
(Adelson, Velskii and Landis) tree is established and rules have been framed
to decide the optimal location of phasor measurement units.

The proposed distributed services have been extended to cost

effective cloud environment for solving multi-area power system state
estimation and optimal PMU placement problems and hence to achieve
modularity and reusability in the power system operation logic. The power
system services such as optimal placement of phasor measurement units and
state estimation are uploaded to public cloud environments and their
performance indices are assessed using the QoS parameters namely the Round
Trip Time and Throughput.

1.4 SCOPE OF THE THESIS

The thesis entitled “Distributed Service Models for Power System

State Estimation” delineates the development of effective distributed and
 10

platform independent models for solving multi-area power system state

estimation and optimal placement of phasor measurement units. This
dissertation is divided into five chapters including the introduction.

Before any security assessment can be made or control actions

taken, a reliable estimate of the existing state of the system must be
determined. For this purpose the number of physical measurements cannot be
restricted to support conventional power-flow calculations. The inputs to the
conventional power flow are confined to the P, Q injections at load buses and
P, |V| values at voltage controlled buses. If even one of these inputs is
unavailable, the conventional power flow solution cannot be obtained.
Moreover, gross errors in one or more of the input quantities can cause the
power flow results to become useless. In practice, other conveniently
measured quantities such as P, Q line flows are available, but they cannot be
used in conventional power flow calculations. These limitations can be
overcome by state estimation. Real time monitoring of power systems has
become a reality with the advent of phasor measurement units.

The traditional state estimation utilizes the measurements from the

Supervisory Control and Data Acquisition system. State estimation gives
quasi real time status for power system analysis and control. PMU measures
synchronized voltage and current phasors in a power system. Improved
monitoring and remedial action capabilities allow network operators to utilize
the existing power system in a more efficient way. Improved information
allows fast and reliable emergency actions, which reduces the need for
relatively high transmission margins required by potential power system
disturbances. Hence, phasor measurement opens a wide range of new
applications such as monitoring and recording of power system dynamics,
improved state estimation, system wide power oscillation mitigation,
transmission line fault location, emergency control during large disturbances,
voltage control in a power system and recording of synchronised events.
 11

PMUs are promising tool for monitoring, protection and control of

power systems. Due to unprecedented change in the size and interconnections
of the power system networks, the location of the phasor measurement units
have to be changed accordingly in order to maintain complete observability.
As the size of the power system is becoming large and large due to increase in
demand and the interconnections between large power systems may vary time
to time due to addition of new generating units and due to geographic
conditions, it becomes difficult to find the optimal location of PMUs for the
real time electric power system network. The location for phasor
measurement unit is to be found out preserving the complete observability of
the network. Chapter 2 explains the basic concepts of state estimation and the
recent trends in power system state estimation and reviews the importance of
phasor measurement units.

The various roles of PMU in state estimation and the need for
optimal placement of PMUs have been explained. The models and techniques
for power system state estimation with PMU and FACTS (Flexible AC
Transmission System) devices have been reviewed. The traditional Binary
Integer Linear Programming (BILP) method for optimal PMU placement has
been revisited. Several test cases including real time networks have been
considered and solved using BILP. The obtained results are analyzed for
maintaining the complete observability of the power systems. Using a test
case, a study in the improvement of the voltage profile when an Interline
Power-Flow Controller (IPFC) is employed in a power system network has
been analyzed.

State estimation is very essential for the monitoring of power

system operations and can be obtained using Weighted Least Square (WLS)
algorithm. The result gives the current operating state of the power system
which primarily helps in maintaining the security of the power system.
 12

The Web provides an environment for invoking the power system

services from anywhere through the Internet thus providing an opportunity for
the power system engineers to make effective, efficient and timely decisions
by understanding the current operating state of the power system without the
need for personnel visits to transmission and distribution locations.

In the proposed model for finding the optimal locations for the
placement of phasor measurements units, the main aim is to reduce the size of
the considered network by eliminating the inner nodes and retaining the
boundary nodes. The influence of elimination of a node with respect to the
retained node is represented by an importance factor that is based on the rank
of the eliminated node. The rank of a node depends on the connectivity strength
and line admittance value. The PMUs are placed in the inner nodes by
analyzing the importance factors of the eliminated nodes with respect to the
retained nodes. This stays as one of the valid assumptions because only by
placing PMUs in the inner buses the system can become completely observable.

Representation of power systems using service oriented models and

their implementation in distributed environment are becoming inherent due to
exponential growth in power industries and flexible adoption to new IT
technologies. Many research works have been reported for representing and
solving power system problems in distributed environments which includes
RMI, Component based, SOA and Grid computing. A generalized
deployment independent service oriented model has been proposed for
representing and solving power system problems. The proposed distributed
service model has four components namely information, interaction, behaviour
and service description which have been detailed in Chapter 3.

The information component stores all the structured XMLised

power system data along with its semantics. The behaviour component
decouples the implementation from the service description. Based on the
 13

service requested by the power system client and the corresponding data
extracted, SOAP request is formulated by the interaction component. The
different power system services namely the XMLised data generation service,
optimal PMU placement service and state estimation service are completely
described in the service description component. The optimal PMU placement
problem with power system reduction technique is considered as a case study
to implement and to test the functionality of the proposed service oriented
model. Nodes are prioritized based on the connectivity strength and AVL tree
is constructed to validate the optimal location of phasor measurement units.

In chapter 4, the proposed componentized service oriented model

for the solution of power system problems is extended to cloud services. The
information component is ported into the ‘Storage as a Service’ layer. The
services are described and the behaviour component is implemented in the
‘Software as a Service’ layer. Cloud service utilizes the SOAP communication to
interact between these two layers. Cloud is on-demand and service oriented.
Cloud Computing is based on the use of distributed computing resources that
are easily allocated, de-allocated, migrated and possibly reallocated on user
request. As it heavily relies on virtualization technologies it offers almost
unlimited amount of virtual computing resources. Virtualization plays a vital
role for realizing interoperability for heterogeneous power system
applications.

The design and development of flexible, standard and distributed

models using two well known PaaS (Platform as a Service) providers namely
Google App Engine and Azure Cloud for solving power system problems
such as state estimation and optimal PMU placement are proposed in chapter
4 to achieve modularity and reusability of the systems. The effectiveness of
the proposed cloud based architectural models is assessed by estimating the
performance measures such as Round Trip Time (RTT) and throughput for
 14

the different power system problems. The cloud environments are compared
considering various characteristics like the developed language, cost for
deploying in the PaaS providers, size of the file and the deployment time.

In chapter 5, a review of the work done, major conclusions arrived

and the contributions made are dealt with. Recommendations for further
research are also stated.