Lecture Notes in Electrical Engineering 824

Surender Reddy Salkuti

Papia Ray   Editors

Next Generation
Smart Grids:
Modeling,
Control and
Optimization
Lecture Notes in Electrical Engineering

Volume 824

Series Editors
Leopoldo Angrisani, Department of Electrical and Information Technologies Engineering, University of Napoli
Federico II, Naples, Italy
Marco Arteaga, Departament de Control y Robótica, Universidad Nacional Autónoma de México, Coyoacán,
Mexico
Bijaya Ketan Panigrahi, Electrical Engineering, Indian Institute of Technology Delhi, New Delhi, Delhi, India
Samarjit Chakraborty, Fakultät für Elektrotechnik und Informationstechnik, TU München, Munich, Germany
Jiming Chen, Zhejiang University, Hangzhou, Zhejiang, China
Shanben Chen, Materials Science and Engineering, Shanghai Jiao Tong University, Shanghai, China
Tan Kay Chen, Department of Electrical and Computer Engineering, National University of Singapore,
Singapore, Singapore
Rüdiger Dillmann, Humanoids and Intelligent Systems Laboratory, Karlsruhe Institute for Technology,
Karlsruhe, Germany
Haibin Duan, Beijing University of Aeronautics and Astronautics, Beijing, China
Gianluigi Ferrari, Università di Parma, Parma, Italy
Manuel Ferre, Centre for Automation and Robotics CAR (UPM-CSIC), Universidad Politécnica de Madrid,
Madrid, Spain
Sandra Hirche, Department of Electrical Engineering and Information Science, Technische Universität
München, Munich, Germany
Faryar Jabbari, Department of Mechanical and Aerospace Engineering, University of California, Irvine, CA,
USA
Limin Jia, State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong University, Beijing, China
Janusz Kacprzyk, Systems Research Institute, Polish Academy of Sciences, Warsaw, Poland
Alaa Khamis, German University in Egypt El Tagamoa El Khames, New Cairo City, Egypt
Torsten Kroeger, Stanford University, Stanford, CA, USA
Yong Li, Hunan University, Changsha, Hunan, China
Qilian Liang, Department of Electrical Engineering, University of Texas at Arlington, Arlington, TX, USA
Ferran Martín, Departament d’Enginyeria Electrònica, Universitat Autònoma de Barcelona, Bellaterra,
Barcelona, Spain
Tan Cher Ming, College of Engineering, Nanyang Technological University, Singapore, Singapore
Wolfgang Minker, Institute of Information Technology, University of Ulm, Ulm, Germany
Pradeep Misra, Department of Electrical Engineering, Wright State University, Dayton, OH, USA
Sebastian Möller, Quality and Usability Laboratory, TU Berlin, Berlin, Germany
Subhas Mukhopadhyay, School of Engineering & Advanced Technology, Massey University,
Palmerston North, Manawatu-Wanganui, New Zealand
Cun-Zheng Ning, Electrical Engineering, Arizona State University, Tempe, AZ, USA
Toyoaki Nishida, Graduate School of Informatics, Kyoto University, Kyoto, Japan
Federica Pascucci, Dipartimento di Ingegneria, Università degli Studi “Roma Tre”, Rome, Italy
Yong Qin, State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong University, Beijing, China
Gan Woon Seng, School of Electrical & Electronic Engineering, Nanyang Technological University,
Singapore, Singapore
Joachim Speidel, Institute of Telecommunications, Universität Stuttgart, Stuttgart, Germany
Germano Veiga, Campus da FEUP, INESC Porto, Porto, Portugal
Haitao Wu, Academy of Opto-electronics, Chinese Academy of Sciences, Beijing, China
Walter Zamboni, DIEM - Università degli studi di Salerno, Fisciano, Salerno, Italy
Junjie James Zhang, Charlotte, NC, USA
The book series Lecture Notes in Electrical Engineering (LNEE) publishes the
latest developments in Electrical Engineering - quickly, informally and in high
quality. While original research reported in proceedings and monographs has
traditionally formed the core of LNEE, we also encourage authors to submit books
devoted to supporting student education and professional training in the various
fields and applications areas of electrical engineering. The series cover classical and
emerging topics concerning:
• Communication Engineering, Information Theory and Networks
• Electronics Engineering and Microelectronics
• Signal, Image and Speech Processing
• Wireless and Mobile Communication
• Circuits and Systems
• Energy Systems, Power Electronics and Electrical Machines
• Electro-optical Engineering
• Instrumentation Engineering
• Avionics Engineering
• Control Systems
• Internet-of-Things and Cybersecurity
• Biomedical Devices, MEMS and NEMS

For general information about this book series, comments or suggestions, please
contact [email protected].
To submit a proposal or request further information, please contact the
Publishing Editor in your country:
China
Jasmine Dou, Editor ([email protected])
India, Japan, Rest of Asia
Swati Meherishi, Editorial Director ([email protected])
Southeast Asia, Australia, New Zealand
Ramesh Nath Premnath, Editor ([email protected])
USA, Canada:
Michael Luby, Senior Editor ([email protected])
All other Countries:
Leontina Di Cecco, Senior Editor ([email protected])
** This series is indexed by EI Compendex and Scopus databases. **

More information about this series at https://link.springer.com/bookseries/7818

Surender Reddy Salkuti · Papia Ray
Editors

Next Generation Smart

Grids: Modeling, Control
and Optimization
Editors
Surender Reddy Salkuti Papia Ray
Department of Railroad and Electrical Department of Electrical Engineering
Engineering Veer Surendra Sai University of Technology
Woosong University Burla, Odisha, India
Daejeon, Korea (Republic of)

ISSN 1876-1100 ISSN 1876-1119 (electronic)

Lecture Notes in Electrical Engineering
ISBN 978-981-16-7793-9 ISBN 978-981-16-7794-6 (eBook)
https://doi.org/10.1007/978-981-16-7794-6

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Singapore Pte Ltd. 2022
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse
of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Preface

Economically functioned power systems that can house a high penetration of renew-
able energy are of global interest. Recently, there has been a significant deployment
of smart grids and a consequent rise in renewable power generation, resulting in
the decarbonization of electrical power systems. However, due to the variability and
intermittency of the renewable sources, significant uncertainty in power systems
occurs, implying a more complex operation and control. Research activities are
needed to provide adequate modeling and optimization techniques that account for
variability and randomness in the proposed grid. The output of a renewable energy
system is affected by the intermittency of resources. The storage system plays a vital
role in overcoming the intermittency and provides a reliable and smooth supply to
the demand. In smart grid networks, sustainability requires using renewable energy
resources (RERs), communication technologies, and controls to improve environ-
mental security and minimizing the cost of the system’s capacity and planning. For
the next generation of smart grids, several computational technologies are required
to design and study the long- and short-term energy independence based on RERs.
Adequate modeling of RERs is needed to handle the variability of resources and
uncertainties of the loads to meet the generation and load demand. Optimal allocation
of different types of renewable energy for distributed generation (DG) penetration
into the distribution network minimizes losses and costs, improving system secu-
rity and reliability. Distribution grids are being transformed from passive to active
networks due to new smart grids with RERs. In this context, smart grids are an
emerging trend in electrical power systems to achieve both economic and envi-
ronmental objectives while contributing to the reliability of energy supply to the
consumers. Existing smart grids require technical expertise to be run, monitored,
and maintained and cannot be deployed as standalone modules. However, to address
this issue, an intelligent smart grid can be built using technologies like the Internet
of Things (IoT) and machine learning. The scope of this book is to bridge the gap
between traditional power consumption methods and modern consumptions methods
using smart computation methods. The goal is to develop an evidence-based, econom-
ically sound, and validated approach to enhancing the resilience of smart power grids,
accounting for their interdependence with other infrastructure systems.

v
vi Preface

This book aims to develop the next generation of smart grid-connected RERs and
load with variability and randomness. Renewable energy’s behavior becomes unpre-
dictable in a changing environment, and it becomes pretty challenging to optimize
the smart grid’s non-stationary distributed energy resources with static optimization
methods. Under this situation, reliable operation of the power system becomes crit-
ical. Optimization and control strategies are needed to be developed to manage the
stochastic nature of the resources while ensuring the stability of the RER-connected
system. The contents of this book include the introduction of various energy optimiza-
tion techniques, modeling and control in renewable and power domain, smart grid
architectures and cyber and physical security, the impact of smart grid in distributed
energy generation, integration of energy sources for smart grid efficiency improve-
ment, intelligent learning models for the monitoring of energy management systems,
electricity markets for power systems with high penetration of RERs, storage and
electric vehicles, implementation of demand response programs in smart grids, and
the intelligent computing applications for electrical power energy systems. This book
also presents the application of artificial intelligence, machine learning, and IoT in the
operation of smart grids, application of various optimization techniques in design
and operation of smart grids, optimization and protection of power systems with
RERs and energy storage, reliability, and resiliency analysis of smart grid, smart
electric vehicles: optimization, security and privacy issues, and optimal generation
scheduling using smart power computations techniques.
The book’s primary emphasis is to develop a computational platform for modeling,
control, performance analysis, and optimization of next generation hybrid energy
systems. Further, this book leads toward cost-effective and environmentally benign
utilization of a future energy system portfolio by providing a cyber-enabled sustain-
able pathway toward deep integration of intelligent decision-makers in the smart
distribution grid. The results of this book will significantly advance state of the art
in power systems operations under uncertainty and will also considerably extend the
methods in solving general multi-stage robust optimization problems. This book is
an effort to educate the next generation of academicians and researchers proficient in
advanced analytics and improve national energy sustainability. The book is organized
into 17 chapters. This book covers various fields to satisfy control and power engi-
neers and researchers. The major topics covered in the book are modeling, control,
and optimization of the smart grid and its various applications.
Chapter “Overview of Next Generation Smart Grids” provides a basic concept
of the smart grid and the different challenging issues that the power sector faces
in modeling and control and their solutions. It also focuses on the concept of
cyber-enabled intelligence network communication, big data, machine learning,
and blockchain technology in smart grids (SGs). In Chapter “Modeling of Various
Renewable Energy Resources for Smart Electrical Power Systems”, the modeling
of power output from different renewable energy resources and the uncertainty in
handling solar photovoltaic and wind energy systems is discussed. Chapter “Smart
Grid Communication: Recent Trends and Challenges” focuses on the evolution of
communication technology from wired networks to wireless networks and the smart
grid architecture; various standards and protocols related to smart grid communi-
Preface vii

cation are also presented. In Chapter “Comparison of Selected MPPT Techniques

Using Different Performance Features”, the investigation is carried out to extract
the highest power from the photovoltaic generator, even in general situations, with
the help of maximum power point tracking methods, namely perturb and observe,
incremental conductance, and temperature control methods. Chapter “Short Term
Active Power Load Forecasting Using Machine Learning with Feature Selection”
proposes an artificial neural network (ANN) architecture designed, implemented,
and tested in a Microsoft Azure Notebook environment to predict a 33/11 kV
electric power distribution substation load. In Chapter “Evaluation of Algorithms
for Fundamental and Harmonic Impacts of Integration of Renewable Energy Sources
in Smart Power Distribution Networks”, a new power flow algorithm for fundamental
and harmonic analysis of smart power distribution network with the integration of
renewable energy-based DG and the nonlinear loads is proposed. Different case
studies are also carried out in this chapter on IEEE-13 bus and IEEE-34 bus test
feeders for analyzing the impacts of multiple integrations of synchronous based-
renewable DG and power electronic based-renewable DG. Chapter “A Compre-
hensive Review of Active Islanding Detection Methods and Islanding Assessment
in a Grid Connected Solar Based Microgrid” gives an overview of different islanding
detection methods (IDMs) of the power system network consisting of distribution
systems along with Distributed Generator (DG). This chapter also gives a techno-
economical comparison of various IDMs based on recent trends related to monitoring
islanding events. Additionally, a simulation study for a grid-connected solar-based
microgrid is also presented in this chapter to analyze five different islanding detec-
tion methods under unintentional islanding. Chapter “A Comparative Analysis of PI
and Predictive Control Strategy for HESS Based Bi-directional DC-DC Converter
for DC Microgrid Applications” presents a comparative analysis of proportional–
integral (PI) and model predictive control (MPC) of multiple input bidirectional DC-
DC converter (MIPC) for hybrid energy storage systems (HESS) in DC microgrid
applications. In this chapter, the simulation results are developed for the compar-
ative performance of MPC over the PI control schemes, and it is concluded from
the simulation results that the proposed MPC is better DC grid voltage restoration to
step-change in PV generation and load demand over traditional PI control scheme. In
Chapter “Parameters Estimation of Solar PV Using Jaya Optimization Technique”, a
novel algorithm-specific parameter-free optimization named Jaya Algorithm is used
to predict the solar PV module parameters accurately for three different technolo-
gies, i.e., poly-crystalline (Shell S75), mono-crystalline (Shell SM55), and thin-film
(Shell ST36). The MATLAB simulation is carried out in this chapter for evaluating
the PV module parameters for various environmental conditions. Chapter “Inte-
gration of Photovoltaic Distributed Generation into Grid” focuses on extracting
maximum power from the photovoltaic solar plant at various working conditions
using a maximum power point tracking algorithm. An active shunt filter is designed
in this chapter to minimize the harmonics injected due to inverter operation. In
Chapter “Transient Stability Enhancement of Power System with Grid Connected
DFIG Based Wind Turbine”, the impact of a doubly-fed induction generator (DFIG)
viii Preface

on the transient stability (TS) of the power system connected to the grid is inves-
tigated, and the implementation of the rotor resistance control method to enhance
the TS of the system is presented. Chapter “Design and Analysis of BLDC Motor
Driver for Hybrid Electric Vehicles” presents the design and analysis of brushless
DC (BLDC) motor drivers for hybrid electric vehicles (EVs). This chapter presents
the relative evaluation of the EV powered by a brushless DC (BLDC) motor and the
corresponding impact on the state of ripples and charge in the DC voltage at the battery
power. Chapter “A Novel Approach for Power Quality Improvement in Microgrid”
proposes an approach to improve the power quality (PQ) of the three-phase system
by manipulating the grid-connected smart Photovoltaic Distribution Static Compen-
sator (PV D-STATCOM) system with the help of a proposed dynamic voltage restorer
(DVR) and a reweighted zero attracting (RZA) control technique containing adap-
tive features with P and O (perturb and observe) approach-based maximum power
point tracking (MPPT) method. In Chapter “Characterization of Bifacial Passivated
Emitter and Rear Contact Solar Cell”, the characterization of bifacial p-type Passi-
vated Emitter and Rear Contact Cell (PERC) solar cell with various proportions
of tallness and width, back silicon nitrate layer with various thickness are stream-
lined for better efficiency than the standard cell which are reaching their physical
limits. Chapter “Network Reconfiguration of Distribution System with Distributed
Generation, Shunt Capacitors and Electric Vehicle Charging Stations” presents a
heuristic-based technique for solving the optimal network reconfiguration (ONR) in a
radial distribution system (RDS) using the fuzzy-based multi-objective methodology.
Minimization of real power losses and deviation of node voltage is considered as the
multiple objectives in this work, and they are modeled with fuzzy sets. The devel-
oped algorithm determines the optimal reconfiguration of feeders with the minimum
number of tie-line switch operations. Chapter “Role of Advanced Control Technolo-
gies in the Evolution of Smart” presents a discussion on the various challenges in
the execution of smart grid and some of the recently proposed control techniques
designed and applied to handle those issues. Also, various control applications in
the smart grid are supplemented in this chapter with appropriate test bench prob-
lems and experimental or simulation results. Chapter “Application of Solar Energy
as Distributed Generation for Real Power Loss Reduction in Radial Distribution
Network” discusses the integration of solar energy as distribution generation in radial
distribution systems and improves the system reliability and integrity by minimizing
the power losses caused in the distribution lines.
Unlike many other books, in this book, individual chapters are unlike most tech-
nical publications as they are journal-type chapters and are not textbooks in nature
that focus on the in-depth coverage of the topic. This book is written for researchers,
utility engineers, and advanced teaching in power and control engineering.
Enjoy Reading!

Daejeon, South Korea Surender Reddy Salkuti

Sambalpur, India Papia Ray
Contents

Overview of Next Generation Smart Grids . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Surender Reddy Salkuti, Papia Ray, and Sravanthi Pagidipala
Modeling of Various Renewable Energy Resources for Smart
Electrical Power Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Surender Reddy Salkuti
Smart Grid Communication: Recent Trends and Challenges . . . . . . . . . . . 49
Ishan Srivastava, Sunil Bhat, and Arvind R. Singh
Comparison of Selected MPPT Techniques Using Different
Performance Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Salauddin Ansari and Om Hari Gupta
Short Term Active Power Load Forecasting Using Machine
Learning with Feature Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Venkataramana Veeramsetty, D. Rakesh Chandra,
and Surender Reddy Salkuti
Evaluation of Algorithms for Fundamental and Harmonic Impacts
of Integration of Renewable Energy Sources in Smart Power
Distribution Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
R. Satish, K. Vaisakh, and Almoataz Y. Abdelaziz
A Comprehensive Review of Active Islanding Detection Methods
and Islanding Assessment in a Grid Connected Solar Based
Microgrid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Kumari Namrata, Akshit Samadhiya, and Papia Ray
A Comparative Analysis of PI and Predictive Control Strategy
for HESS Based Bi-directional DC-DC Converter for DC
Microgrid Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Srinivas Punna, Udaya Bhasker Manthati, and C. R. Arunkumar

ix
x Contents

Parameters Estimation of Solar PV Using Jaya Optimization

Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
C. Srinivasarathnam, Gurappa Battapothula, Anil Annamraju,
and Chandrasekhar Yammani
Integration of Photovoltaic Distributed Generation into Grid . . . . . . . . . . 257
Neelakanteshwar Rao Battu, Perka Krishna, and Venu Yarlagadda
Transient Stability Enhancement of Power System with Grid
Connected DFIG Based Wind Turbine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
D. Rakesh Chandra, Surender Reddy Salkuti,
and Venkataramana Veeramsetty
Design and Analysis of BLDC Motor Driver for Hybrid Electric
Vehicles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
Seong-Cheol Kim, Narasimha Sangam, Sravanthi Pagidipala,
and Surender Reddy Salkuti
A Novel Approach for Power Quality Improvement in Microgrid . . . . . . 313
Arvind R. Singh, Papia Ray, and Surender Reddy Salkuti
Characterization of Bifacial Passivated Emitter and Rear Contact
Solar Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333
Suresh Kumar Tummala, Phaneendra Babu Bobba,
and Satyanarayana Kosaraju
Network Reconfiguration of Distribution System with Distributed
Generation, Shunt Capacitors and Electric Vehicle Charging
Stations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
Surender Reddy Salkuti
Role of Advanced Control Technologies in the Evolution of Smart . . . . . . 377
Raseswari Pradhan
Application of Solar Energy as Distributed Generation for Real
Power Loss Reduction in Radial Distribution Network . . . . . . . . . . . . . . . . 403
Aliva Routray, Khyati D. Mistry, and Sabha Raj Arya

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
Editors and Contributors

About the Editors

Dr. Surender Reddy Salkuti received the Ph.D. degree in electrical engineering
from Indian Institute of Technology, New Delhi, India, in 2013. He was a Post-
doctoral Researcher at Howard University, Washington, DC, USA, from 2013 to
2014. He is currently working as an Associate Professor in the Department of Rail-
road and Electrical Engineering, Woosong University, Daejeon, Republic of Korea.
His current research interests include power system restructuring issues, ancillary
service pricing, real and reactive power pricing, congestion management, and market
clearing, including renewable energy sources, demand response, smart grid devel-
opment with integration of wind and solar photovoltaic energy sources, battery
storage and electric vehicles, artificial intelligence applications in power systems,
and power system analysis and optimization. He received Distinguished Researcher
Award from Woosong University Educational Foundation, Republic of Korea in
2016, and POSOCO Power System Award (PPSA), India in 2013. He is a Member
of IEEE and IEEE Power and Energy Society.

Dr. Papia Ray received the Ph.D. degree in Electrical Engineering from the Indian
Institute of Technology, New Delhi, India, in 2013. She is currently working as
an Associate Professor in the Department of Electrical Engineering, Veer Surendra
Sai University of Technology, Burla, Odisha, India and is having more than 17
years of teaching experience. Her current research interests include power system
protection, power quality, wide-area measurement systems, artificial intelligence
applications in power system protection and microgrid protection. She is the recipient
of the Young Scientist Award from DST in 2015. She has published several research
papers in various journals and conferences and is also an active reviewer for several
reputed journals. She is a Senior Member of IEEE, Life Member of Indian Society
for Technical Education, and Member of Institution of Engineer’s India Ltd. She
has edited a book on Microgrid: Operation, Control, Monitoring and Protection
published by Springer.

xi
xii Editors and Contributors

Contributors

Almoataz Y. Abdelaziz Faculty of Engineering and Technology, Future University

in Egypt, Cairo, Egypt
Anil Annamraju Department of Electrical and Electronics Engineering, School of
Engineering, Malla Reddy University, Telangana, India
Salauddin Ansari Department of Electrical Engineering, National Institute of
Technology Jamshedpur, Jamshedpur, Jharkhand, India
C. R. Arunkumar Electrical Engineering Department, NIT Warangal, Warangal,
Telangana, India
Sabha Raj Arya Department of Electrical Engineering, Sardar Vallabhbhai
National Institute of Technology Surat, Gujarat, India
Gurappa Battapothula Department of Electrical and Electronics Engineering,
B.V. Raju Institute of Technology, Narsapur, Telangana, India
Neelakanteshwar Rao Battu Department of Electrical and Electronics Engi-
neering, VNR VJIET, Hyderabad, Telangana, India
Sunil Bhat Electrical Engineering Department, Visvesvaraya National Institute of
Technology, Nagpur, India
Phaneendra Babu Bobba Electrical and Electronics Engineering Department,
Gokaraju Rangaraju Institute of Engineering and Technology, Hyderabad, India
Om Hari Gupta Department of Electrical Engineering, National Institute of Tech-
nology Jamshedpur, Jamshedpur, Jharkhand, India
Seong-Cheol Kim Department of Railroad Electrical Systems, Woosong Univer-
sity, Daejeon, Republic of Korea
Satyanarayana Kosaraju Mechanical Engineering Department, Gokaraju
Rangaraju Institute of Engineering and Technology, Hyderabad, India
Perka Krishna Department of Electrical and Electronics Engineering, VNR
VJIET, Hyderabad, Telangana, India
Udaya Bhasker Manthati Electrical Engineering Department, NIT Warangal,
Warangal, Telangana, India
Khyati D. Mistry Department of Electrical Engineering, Sardar Vallabhbhai
National Institute of Technology Surat, Gujarat, India
Kumari Namrata Department of Electrical Engineering, National Institute of
Technology Jamshedpur, Jamshedpur, India
Sravanthi Pagidipala Department of Electrical Engineering, National Institute of
Technology Andhra Pradesh (NIT-AP), Andhra Pradesh, India
Editors and Contributors xiii

Raseswari Pradhan Department of Electrical Engineering, VSSUT, Burla,

Odisha, India
Srinivas Punna Electrical and Electronics Engineering Department, BVRIT
Hyderabad College of Engineering for Women, Hyderabad, Telangana, India
D. Rakesh Chandra Department of Electrical and Electronics Engineering,
Kakatiya Institute of Technology and Science, Warangal, India
Papia Ray Department of Electrical Engineering, Veer Surendra Sai University of
Technology (VSSUT), Burla, Odisha, India
Aliva Routray Department of Electrical Engineering, Sardar Vallabhbhai National
Institute of Technology Surat, Gujarat, India
Surender Reddy Salkuti Department of Railroad Electrical Systems, Woosong
University, Daejeon, Republic of Korea
Akshit Samadhiya Department of Electrical Engineering, National Institute of
Technology Jamshedpur, Jamshedpur, India
Narasimha Sangam Department of Electrical and Electronics Engineering, TKR
College of Engineering and Technology, Hyderabad, India
R. Satish Department of Electrical and Electronics Engineering, ANITS (A),
Visakhapatnam, Andhra Pradesh, India
Arvind R. Singh School of Electrical Engineering, Shandong University, Jinan,
China
C. Srinivasarathnam Department of Electrical and Electronics Engineering,
Vasavi College of Engineering, Ibrahimbagh, Hyderabad, Telangana, India
Ishan Srivastava Electrical Engineering Department, Visvesvaraya National Insti-
tute of Technology, Nagpur, India
Suresh Kumar Tummala Electrical and Electronics Engineering Department,
Gokaraju Rangaraju Institute of Engineering and Technology, Hyderabad, India
K. Vaisakh Department of Electrical Engineering, Andhra University, AUCE (A),
Visakhapatnam, Andhra Pradesh, India
Venkataramana Veeramsetty Center for Artificial Intelligence and Deep
Learning, SR Engineering College, Warangal, India
Chandrasekhar Yammani Department of Electrical Engineering, National Insti-
tute of Technology, Warangal, Telangana, India
Venu Yarlagadda Department of Electrical and Electronics Engineering, VNR
VJIET, Hyderabad, Telangana, India
Overview of Next Generation Smart
Grids

Surender Reddy Salkuti, Papia Ray, and Sravanthi Pagidipala

Abstract Due to the integration of renewable energy resources (RERs), the elec-
tric power system is undergoing a significant change in operation, maintenance,
and planning. The advancement into a smart grid (SG) using advanced automation
and control approaches produces new challenges and opportunities. In this work,
an overview of next generation smart grids has been presented by highlighting the
latest and recent innovations in the SG fields. Advantages, barriers, and opportuni-
ties of smart grids are provided in this work. Challenges of integrating the RERs
into the grid and various energy storage technologies are reviewed. The output of
RERs is affected by the uncertain nature of resources. The energy storage system
plays a vital role to handle this uncertain nature of RERs and provides a smooth and
reliable supply to the load demand. Smart energy systems present various challenges
and opportunities in designing, integrating, and implementing electrical grids with
network and communication technologies and essential privacy and security issues
of various components within the electrical grid. Therefore, this work presents the
concepts of cyber-enabled intelligence network communication, big data, machine
learning, and blockchain technology in smart grids (SGs). This work also demon-
strates the impact of SGs in distributed energy generation and the comparative study
on electric vehicles (EVs) along with the classification, i.e., battery, hybrid electric,
and plug-in electric vehicles, current issues and challenges on the EV technology.
Further, a discussion is also provided on SG protection issues and their remedy.

S. R. Salkuti (B)
Department of Railroad and Electrical Engineering, Woosong University, Daejeon 34606,
Republic of Korea
e-mail: [email protected]
P. Ray
Department of Electrical Engineering, Veer Surendra Sai University of Technology (VSSUT),
Burla, Odisha, India
e-mail: [email protected]
S. Pagidipala
Department of Electrical Engineering, National Institute of Technology Andhra Pradesh
(NIT-AP), Andhra Pradesh, India

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 1
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6_1
2 S. R. Salkuti et al.

Keywords Renewable energy · Machine learning · Demand response · Big data ·

Smart grid · Energy storage · Electrical vehicles · Optimal control

Nomenclature

BEV Battery electric vehicle

CHP Combined heat and power
DG Distributed generation
DR Demand response
DER Distributed energy resources
DMS Distribution management system
PV Photovoltaic
VVO Voltage/VAR optimization
AMI Advanced metering infrastructure
GIS Geographic information system
OMS Outage management system
WAMS Wide area measurement system
ML Machine learning
PMU Phasor measurement unit
EMS Energy management system
IED Intelligent electronic device
SG Smart grid
RERs Renewable energy resources
WT Wind turbine
ICE Internal combustion engine
PHEV Plug-in hybrid electric vehicle
IoE Internet of energy
IED Integrated electronic device

1 Introduction

A smart grid (SG) is an advanced distribution and transmission grid that uses infor-
mation, control, and communication technologies to enhance the electrical grid’s
security, efficiency, economy, and reliability. The conventional grid can be changed
into the SG by smart metering, advanced network operation and control, faster
fault identification, and self-healing capabilities using grid automation. SG can be
defined as an interconnected system of information, communication technologies,
and control systems used to interact with automation and business processes across
the entire power sector, encompassing electricity generation, transmission, distribu-
tion, and consumers [1]. Smart renewable energy control centres which can forecast
Overview of Next Generation Smart Grids 3

and monitor renewable energy availability and potentially use energy storage to
manage the dispatch of power to match grid conditions or manage demand through
demand response (DR) programs to match capacity availability are expected to
become critical to the future integration of renewable energy.
In recent days, the power generation from renewable energy resources (RERs)
is increasing significantly, and the power generated from renewable energy will
play a significant role soon. To enhance the integration of renewable power in the
conventional power network, it has to deliver the liabilities acted to the network
due to the irregular nature of these resources. Because of their effect on system
adjusting, reserves procurement, scheduling and commitment of generating units,
changeability, and ramp or slope events in power output are the fundamental diffi-
culties to the system administrators [2]. With the integration of RERs such as solar PV
and wind, significant uncertainty into the power system is developed. It is a tremen-
dous challenge for system operators to maintain reliable operation and efficient elec-
tricity markets with simultaneous maximum utilization of renewable energy. As the
electric market structures change to improve the management of renewable sources,
advances in the design and pricing aspects of energy and ancillary services markets
are needed [3].
Distributed generation (DG) is from many small energy sources connected directly
on the customer side of the meter or to the distribution network. DGs are also termed
as embedded generation or dispersed generation or decentralized generation, or on-
site generation. Optimal size and location of DGs can reduce the power losses in
the system, improve system voltage, maximise DG capacity, enhance system reli-
ability, and minimise investment. Power companies generally generate electricity
with a relatively fixed amount of energy over short periods, regardless of the daily
fluctuation of the energy demand. The energy demand fluctuation leads to energy
storage techniques to supply enough electricity to consumers during peak demand.
Electricity storage techniques allow power companies to provide electricity when
needed [4]. The electricity is first generated and then stored in storage devices to be
used during the peak load.
Oil, natural gas, uranium are all power sources that the world has been using
for a long time and they all have the potential to run out at any time. RERs are
the sources that can last longer than the human race. Most renewable energies are
environmentally friendly, fight against global warming by reducing carbon emission.
Renewable energies are sustainable, reliable, and cost less compared to fossil fuels
and other nonrenewable energies. Solar energy initial installation remains expensive,
but it is increasingly being used around the globe, especially in the rural world where
people don’t have access to electricity or simply can’t afford to pay for their monthly
electric bills. Biomass is one of the most renewable energies used today due to
its advantages [5]. In addition to the fact that biomass energy does not pollute the
air allows countries to protect their environments by burning the wastes. Renewable
energy is a resource used all over the world. Renewable energy produces from natural
resources such as wind, solar photovoltaic (PV), biomass, geothermal, etc., and is
known to be beneficial to the earth. Table 1 presents the various characteristics of
the traditional power grid and the SG [6].
4 S. R. Salkuti et al.

Table 1 Characteristics of the traditional power grid and smart grid

Characteristics Traditional power grid Smart grid (SG)
Technological aspect The traditional power grid is SG uses advanced digital
electromechanical. As a result, technology. It allows for
it has no means of communication between devices
communication and there is and promotes self-regulation
minimal internal regulation
Distribution system It involves the one-way It allows for the two-way
distribution of power. Power is distribution of power, i.e., power
only distributed from the main can also go back to the
power plant with the help of distribution lines from a
traditional infrastructure secondary power provider
Generation The generation infrastructure is The generation infrastructure is
highly centralized. All the scattered and distributed. It
power is generated from a allows the power to be
central system that minimises distributed from multiple power
incorporating alternative energy sources, which help in balancing
sources into the grid the load and limit power outages
Incorporation of sensors It involves limited sensors as SG infrastructure system allows
the infrastructure is unable to multiple sensors throughout the
handle sensors making line, which helps to pinpoint
detection of problems difficult problems
Monitoring of power plants Traditional infrastructure is The SG involves advanced
minimal. As a result, energy technology and can monitor
distribution has to be manually itself, which prevents
monitored troubleshoot outages
Restoration and repairing For restoration purposes, In the case of SG technology,
technicians must physically go the sensors can detect, analyze
to the failure location and make and rectify the required
corresponding repairs to prevent problems without the need for
excessive power outages physical intervention
Equipment Due to the ageing and limited In the case of an SG system,
capacity of the traditional alternative energy sources limit
power grid, it is prone to the area that is impacted by the
failures and blackouts power outage
Control of energy In the case of traditional power With the availability of sensors
infrastructure, energy and innovative infrastructure,
companies have no control over energy providers have pervasive
the distribution of energy control over the power
distribution
Choices of consumers The traditional power grid SG allows consumers to have
system does not provide more choices in the way they
consumers with any flexibility receive energy
in receiving their electricity
Overview of Next Generation Smart Grids 5

A smart grid (SG) network should posses

• Better communication: This deals with interaction with the network and within
the network itself. An illustration of this will be the automatic shutting off of
power supplied to the home when people living in a house. In light of that, it is
easy to see that a lot of work goes into sensing and feedback circuitry built into
the grid as it receives information about the power grid itself.
• Self-healing capability: SG employs technologies that assist in self-healing in
response to natural faults or breakdown without human intervention.
• Adaptiveness and flexibility: SG should be able to adapt to variations in condi-
tions within the grid, especially regarding ever-changing load and the direction
of flow of power.
• Resilience: This refers to the ability of the SG to withstand attack, i.e., both human
and natural.
• Real-time measurements and control: In monitoring the central generation,
distribution generation, and other plants in the system, the grid must be capable
of making real-time measurements and controlling the grid due to the sensitivity
of generated power even within just microseconds.
• Reliable technology: SG uses state estimation technologies, which improves fault
detection and allows self-healing of the corresponding system without the need of
the technicians to intervene [7]. This ensures a proper reliable supply of electricity
and reduces related vulnerabilities.
• Flexibility in networking: SG technology enables better handling of the bidi-
rectional flow of energy, allowing for distributed generation. This includes solar
panels, fuel cells, electric batteries, wind turbines, and other such sources.
• Increased efficiency: SG technology has resulted in numerous contributions
towards the overall improvement in the efficiency of the power grid system and
energy infrastructure. The demand-side management has highly benefitted from
the smart infrastructure. This includes the voltage/VAR optimization (VVO) tech-
nique which reduces the power usage through the distribution lines during idle
time and advanced metering infrastructure (AMI) systems which improve the
outage management of usage data [8]. All these have resulted in better efficiency
and greater utilization of energy resources.
• Peak curtailment-usage and pricing: Communication systems and metering
technologies have reduced demand during high cost and peak usage hours. Smart
devices allow users to track the energy demand and usage of electricity. More-
over, it also allows utility companies to reduce consumptions and avoid system
overloads. This is done by increasing the electricity cost during peak hours which
is termed peak curtailment.
• Sustainability: The SG system is supposed to be highly flexible as it allows the
correlation of various RERs such as solar and wind energy without any additional
storage.
• Demand response: Demand response allows loads and generators to interact
amongst themselves in an automated real-time manner [9]. This, in turn, eliminates
6 S. R. Salkuti et al.

the additional costs of reserve generators and increases the life expectancy of
power systems [10].
• Advanced services in the near future: The use of robust two-way communica-
tions combined with technologically advanced sensors and distributed technology
will drastically improve the efficiency, productivity, reliability, and safety of power
generation, transmission, and distribution.
SG combines various technologies and applications to provide efficient, clean,
reliable, easily accessible, and low-cost energy. The need of the hour is to perform
complex testing and improve applied techniques and improve threshold and opti-
mized operation of the SG, keeping in mind the sustainable development of the
environment [11]. Various applications of SGs include:
• Improve response and adaptivity of transmission and distribution systems.
• Enables quicker detection and recovery from faults, disturbances, and outages in
the distribution lines and feeders.
• Reduces the cost and peak demand for energy.
• Allows for integration and scalability of various alternative sources of energy such
as wind and solar power.
• Utilizes sharing of load, which in turn reduces the load on a large scale.
• Many other technologies have also helped to optimize the SG on a large
scale, such as advanced metering infrastructure (AMI), distribution manage-
ment system (DMS), geographic information system (GIS), outage management
system (OMS), wide area measurement system (WAMS), phasor measurement
unit (PMU) and energy management systems (EMS).
The international energy roadmap study ranks windmills, biomass, solar PV, and
tidal power as future renewable power sources to sustain the world’s economy.
Progress in the field of sustainable energy scenario over the previous period has been
exceptional. The two primary sustainable energy resources are the sun and the wind.
The growth in electricity generation from renewables was substantially increased
[12, 13]. The IEA ventures by 2050, around (15–18)% of worldwide power will be
created from the wind, with sun-powered PV contributing as high as 16% [14, 15]
regardless of its different favourable circumstances.

2 Smart Grids: Advantages, Barriers, and Opportunities

A smart grid (SG) is a system that utilizes digital processing and communications
technology to manage resources connected to a power grid. The heavy dependence
on technology and hence power supply that aid the operation of these technologies
across all facets of life from the auto industry to medical equipment, households,
and so on has increased the need for efficient management of the assets to maximize
Overview of Next Generation Smart Grids 7

productivity [16]. The ageing power grid, workforce also compound this, and infras-
tructure within the US SG will help ensure better communication between human
beings and the power grid.

2.1 Advantages and Barriers of Smart Grids

Advantages of SG include:
• Growth of economic sector: The manufacture, installation, maintenance, and
corresponding operation of the SG and its respective components will create
opportunities for skilled and unskilled job sectors [17]. Moreover, it will help in
growth in business along with providing advanced technological solutions.
• Cost efficiency: SG technologies tend to be cheaper than traditional energy
sources. Moreover, with advancements in technologies, energy expenses reduce
over time.
• Greater customer satisfaction: The improved reliability and lower costs have
resulted in greater customer satisfaction [18]. Additionally, SG provides better
customer control over energy distribution and usage.
• Better reliability: SG reduces power outages and, at the same time, improves
power efficiency.
• Environment: SG results in a positive impact on the environment as it helps
reduce greenhouse gas emissions by enabling alternate RERs and electric vehicles
(EVs) [19]. Moreover, it also reduces oil consumption because of efficient power
generation.
• Intelligence: SGs enhance the capacity and capability of the existing power grid.
Also, they have predictive maintenance and self-healing capabilities.
Barriers of SG include:
• Security and privacy concerns: As the SG is void of human interference, it is
highly prone to hacking and cyber-attacks. It may lead hackers to gain access to
numerous smart meters and manipulate the data [20]. Also, since the system is
fully automatic, privacy concerns are raised over data collection and the use of
smart meters.
• Lack of government norms and policies: As SG technology is still in its budding
stage, there is a lack of proper stringent government policies to avoid thrift, loss,
and misuse of the power system. Also, there are no standard regulations regarding
the safety and threshold limitations of SG systems.
• Higher installation costs: The present energy infrastructure scenario is inade-
quate and insufficient to develop SG technologies [21]. As a result, SG systems
require a higher installation cost as it involves the procurement of smart meters
and various other sensors for monitoring and data distribution.
• Congestion in network topology: Smooth operation of a SG requires a contin-
uous network channel. This may result in network traffic, especially during peak
8 S. R. Salkuti et al.

hours. Moreover, there is no guaranteed service during abnormal conditions such

as heavy rainfall, storm, and lightning conditions.

2.2 Challenges

There are bound to be vital challenges with the rapid growth in digital technologies
and their widespread applications. These include insecurity and privacy issues along
with a lack of government policies [22]. The most common challenges regarding SG
include:
• Privacy issues: With the involvement of technology, there are bound to be privacy
concerns. There is a possibility of some people hijacking the system to gain
sensitive data regarding the consumers.
• Security issues: SG technology is prone to attack by hackers who may gain access
to the power system and disrupt the underlying technology to gain an advantage
by hampering the net-metering system [23].
• Grid volatility: SG is considered to be a volatile network, being highly intelligent
at both the transmission end and distribution end but a bit vulnerable along the
way due to a lack of complete network intelligence.
Various processes and systems to overcome the above challenges include:
• Increasing the grid strength: The power grid should have enough tensile strength
to withstand the power requirement. The power grid is one of the most complex
inter-connected physical systems in the world. Hence, absolute care should be
taken to sustain the grid network.
• Communication systems: SG should be optimized to integrate all kinds of sensors
and digital devices to enhance and upgrade the level and performance of the grid.
Additionally, there should be proper coordination and communication among all
the components to ensure reliable operation. Fiber optics has enabled faster travel
of data signals, thus reducing propagation time.
• Advanced metering infrastructure (AMI): It involves a net-metering system
that includes a smart meter and the infrastructure to allow for two-way communi-
cation of energy. With the help of broadband and radio technology, it has become
possible for grids to communicate with other grids. Moreover, it is also a viable
option for energy sustenance and reduces consumer costs.
• Economic dispatch and integration of various sources: It involves cost opti-
mization of electrical power analysis and minimises production and distribu-
tion costs. Also, the integration of various energy sources like solar and wind
is achieved with the intermittent coordination of multiple generators.
• Advanced power system monitoring: Use of synchronised PWMs and SCADA
has enabled better monitoring of SGs and proper control over the utility. AMI has
helped real-time measurements of voltage and current phasors. GPS has allowed
us to pinpoint the fault location and reduce the outage time precisely.
Overview of Next Generation Smart Grids 9

3 Technologies Required for the Smart Grid Development

SG has a set of diverse necessities. To accomplish all the requirements, various

technologies have been enabled and utilized [24].

3.1 Sensing and Measurement Technologies

These technologies include:

• Phasor measurement units (PMU) and Wide area monitoring, protection and
control (WAMPAC): These help in maintaining the security and reliability of the
power grid [1].
• Intelligent electronic devices (IED): These devices are used for protection relays
and determining faults. These involve integrated sensors and automated systems
for rapid analysis and diagnosis and results in timely resolutions. They also func-
tion to minimize network congestion and reduce the risk of outages. As a result,
these devices enhance the efficiency and operation of the resultant SG.
• Smart appliances: These devices form the control and communication part of
the power grid system [25]. They monitor the operation of consumer appliances
and try to maximize their efficiency, reliability, and safety.
• Smart meter and its associated software: These form the building blocks of
the entire SG system. It provides better electricity control and allows consumers
greater control over their electricity consumption. They also make the consumers
responsible by providing real-time info and proper bills.

3.2 Communication Technologies

The various technologies include:

• Two-way communication: It enables a back and forth bridge to allow a two-way
flow of electrical energy across various loads in a power system.
• Advanced architecture: It allows for technologies regarding EVs, micro-
generation, and plug-play home appliances.
• Communication technology: Along with the corresponding software, communi-
cation technology aims to provide the energy consumers with better opportunities
and more meaningful information [26].
• Hardware and software: These formed the building blocks of the system and
were responsible for maintaining the security and reliability of the utility.
10 S. R. Salkuti et al.

3.3 Smart Meters and Advanced Metering Infrastructure

(AMI)

AMI is defined as the integration and interconnection of numerous smart meters

and communication equipment that work in cohesion to enable a two-way flow of
power in the power grid between utilities and consumers. It involves data manage-
ment systems and advanced sensors that make the SG capable of monitoring its
power requirements. Smart meters are equipped with power monitoring software
technologies that enable faster detection and diagnosis of faults along with ways to
resolve power problems [27]. Moreover, AMI prepares the grid with self-healing
capability by monitoring power outages, enabling faster detections of faults. Various
other features of AMI include:
• Power generation systems and storage components can be analyzed and controlled
at the consumer level. It provides data to improve the operation and management
of the utility network.
• Before AMI enables auto-detection of faults and outages, it also informs the repair
crew of the location of the fault.
• An AMI system provides accurate records and notifications regarding outages
and their corresponding restoration through the smart meter sensors. This helps
in quicker resolution of outages and also enables faster providing of power supply.
• AMI implementation reduces the daily reading of meters as the readings are taken
automatically at regular intervals [28].
• It implements built-in analysis tools and software which detect and prevent
possible energy theft and tampering.
• By combining with SCADA, the AMI systems automatically generate investiga-
tion reports for field managers making their work easier.

4 Renewable Energy and Energy Storage

Renewable energy is a resource needed to transform the electric power system by

enhancing the SG’s performance, sustainability, and reliability. Distributed energy
resources (DER) in a microgrid would include photovoltaic (PV), small wind turbines
(WT), heat or electricity storage, combined heat and power (CHP), and controllable
loads. Renewable energy sources are practically inexhaustible in that most of their
energy is from nature and sources that will continue to exist like the sun. Renewable
energy is an area that will continue to grow in research and development; more money
has gone into the exploration of both new and existing renewable sources. Solar, the
most commonly used, has improved its application in our world drastically with the
advents of solar-powered cars, light, and houses [29]. More work needs to be done to
eradicate non-renewable energy in our society to create a world where green power
is the dominant energy source used. Finally, for renewable energy to facilitate energy
shortage, extensive research on utilising the available RER to its maximum efficiency
Overview of Next Generation Smart Grids 11

is needed. Also, areas throughout the world should implement primary RER based
on their most vital natural resource. For example, areas with the strong wind will
implement more hydropower, and regions with constant sunlight will use solar.

4.1 Renewable Energy

Renewable energy resources (RERs) could be harnessed from technological sources

like solar, wind, biomass, biogas, geothermal, and hydroelectric. There has been an
improvement in technologies and advancement in the equipment and use of renew-
able energy. More companies, through research and government incentives, have
increased the operational experience of renewable energy technologies. The issue of
renewable energy and its application presents questions on the efficiency, reliability,
tracking, and accountability of these resources. However, most renewable energy
technologies struggle to compete economically with conventional fuel technologies
making the budget and funds for renewable energy smaller than its counterpart [30].
The regulatory commission has categorized the utility industry into three; genera-
tion companies cover utility and non-utility companies, transmission companies, and
distribution companies.
The intermittent nature of wind generation makes its operation and planning a
complex problem, and there is a need for the current analytical models to consider this
uncertainty appropriately. Current electricity market clearing schemes cannot fully
integrate the essential features of non-dispatchable generation such as wind power.
This limitation is becoming an issue for grid operators as there is more and more
public and political pressure to increase the penetration of renewable generation.
This is due to the uncertainty associated with wind forecasts. As wind cannot be
forecast to a high degree of accuracy, additional reserve capacity needs to be carried
by the system in addition to the reserves already allocated to cater for unit outages and
demand forecast errors. Renewable energy could be discussed regarding connectivity,
sustainability, cost, environmental impact, security, etc.
• Connectivity: Renewable energies usually require different levels before they
could be used as power for its consumers’ stages from generation to transmission
to distribution and the utilization levels in its connectivity. Based on a particular
renewable technology, this varies [31]. Some RERs could be stand-alone, grid-
connected, or hybrid, which combines both stand-alone and grid-connected. In
comparison to non-renewable energies, the connectivity for renewable energy
technologies is quite challenging.
• Sustainability: This is one of the strong positives of renewable energy. In compar-
ison to non-renewable energies, it’s readily available to be harnessed. Since most
RER are gotten from nature like the wind, sun, water, it’s always bound to exist,
and others RER like bio-mass will exist due to human waste.
• Cost: The amount necessary to have RE technologies is said to cost less than the
amount used in electricity production. Most massive amounts go into research on
12 S. R. Salkuti et al.

renewable technologies and how to utilize these sources rather than how they are
set up effectively.
• Environmental impact: Renewable energy is considered green power due to
its friendliness to the environment. Most RERs have a significantly low carbon
emission rate and insignificant waste. In regards to areas of scenery, wildlife, and
land utilization, it has a negative impact. Most facilities/instruments for RERs are
huge, land-consuming, and disquieting to the area.
• Security and regulations: An essential concern to RERs is how safe the facilities,
devices, and instruments used in the operations are? From the production to trans-
mission to distribution and then utilization, companies try to fight off vandalisms,
wastage and enforce rigid standards on how the procedure should be.
A summary of some of the advantages of renewable energy [32] could be the
following:
• The abundance of these sources used in renewable energy technologies.
• RE sources are considered environmentally friendly and green due to low carbon
emissions.
• The research and development in this field provide thousands of jobs and employ-
ment opportunities to the society, thereby helping the economy of its subsequent
country.
• Like in non-renewable sources where some resources are gotten from other coun-
tries and either imported or exported, all renewable sources are present everywhere
in all countries as it is abundant in nature.
• Renewable sources cost less than the current price of electricity. Crude oil is
projected to increase its price in the coming years.
• Various tax incentives in the form of credit deductions, tax waivers are available
for businesses and individuals who use renewable resources.

4.2 Energy Storage

RERs such as solar and wind are not available throughout the year, i.e., seasonal.
So, it is of utmost importance to store the maximum amount of energy possible
with minimum losses. The SG requires energy storage directly or indirectly. In the
present scenario, the pumped hydroelectric storage technology is considered bulk
energy storage technology. Proper storage units connected with OFF-grid systems
are further connected to the utility to form a hybrid system, improve robustness, and
prevent intermittent outages. Older storage units include batteries, flywheels, and
compressed air systems. Nowadays, a hydrogen network is connected in parallel to
the grid utility as a storage unit. Storage devices, including batteries, supercapacitors,
and flywheels, could be used to match generation with demand in SGs [33]. The
storage systems can supply generation deficiencies, reduce load surges by providing
a ride-through capability for short periods, reduce network losses, and improve the
Overview of Next Generation Smart Grids 13

protection system by contributing to fault currents. V2G and EV mobility can reduce
the SG reliance on the grid supply.

5 Big Data, Machine Learning, and Blockchain Technology

in Smart Grids

5.1 Big Data

Data identification and collection, data storage and data filtering, data classification
and extraction, data cleaning and data summarization, data analysis and processing,
and data visualization are the various life cycle of big data analysis. Apache Spark,
Apache Hadoop, Apache Storm, Apache Flink, Apache Kafka, and Apache Samza
are the big data processing tools that are being developed. But the main challenge
faced during the data collecting is a time delay in receiving data [34]. This problem
is resolved by using 5G technologies, which can reduce the time delay problem. Big
data applications in power and energy systems include transient power prediction,
distribution and utility systems, SG applications, fault detection and prevention,
weather data and wind speed prediction, etc. Table 2 presents the applications of big

Table 2 Applications of big data in various industries

Industries Applications of big data
Power and energy Distribution systems, generation systems, utility
systems, SG applications
Health care and life sciences Clinical trials data analysis, disease pattern analysis,
chain management, drug discovery and development
analysis, patient care quality, and program analysis
Telecommunications Customer churn prevention, call detail record analysis,
mobile user location analysis, network performance and
optimization, revenue assurance, and price optimization
Web and digital media Large scale clickstream analytics, abuse, click-fraud
prevention, social graph analysis and profile
segmentation, campaign management, and loyalty
programs
Retail/consumers Supply chain management and analytics, market and
consumer segmentation, event and behaviour-based
targeting, merchandising, and market-based analysis
Finance and fraud services Compliance and regulatory reporting, risk analysis and
management, fraud detection and security analytics,
credit risk, scoring, and analysis
E-commerce and customer service Cross-channel analytics, event analytics, right offer at
the right time, next best option, or following best action
14 S. R. Salkuti et al.

Table 3 Challenges, state of the art, current status, and recent developments in big data
Challenges State of the art, current status, and recent
developments
• Privacy • Big data technique to handle a large amount
• Data miming of information in a short time using meter
• Integration of data data management
• Cybersecurity • Big data requirements and enhancements
• Demand prediction through analytics throughout the entire power network
processing in SG applications dispatching and planning
• Data quality and cost balance • Fault diagnosis on bogies of the high-speed
• Industrial fault diagnosis using big data train with big databases on deep neural
• Quantum cryptography for data security in networks
smart grids • Schedulable capacity forecasting technique
for thermostatically controlled load by big
data analysis
• Concept of device electrocardiogram in fault
diagnosis using big data
• Tensor-based big data management scheme
in SG systems
• The artificial neural network approach is
used for efficient electricity generation
forecasting

data in various industries [35, 36]. Multiple challenges, state of the art, current status,
and recent developments in big data are presented in Table 3.

5.2 Machine Learning (ML)

A myth is that artificial intelligence (AI) is only one application that is robotics. Still,
AI has much application such as medical sector, education sector, agriculture sector,
pharmacy sector, research sector, energy sector, transport sector, etc. AI makes it
possible for the machine to work as a human, learn like a human, and resemble the
behaviour of humans. AI can be divided into artificial general intelligence (AGI) and
artificial narrow intelligence (ANI). ANI is a type of AI in which a machine would
perform only some specific task. AGI is a type of AI in which machines perform
any given task that humans can perform or many more. In the future, all the sectors
are upgraded with AI. Various AI applications in the SG are power load forecasting,
power generation forecast of renewable energy, fault diagnosis, and protection of
flexible equipment, electricity consumption behaviour, network security protection,
and energy supply. Machine learning can be classified into two groups one is based
on the learning model, and another one is based on learning methods. The learning
model can be classified into supervised learning, unsupervised learning, and rein-
forcement learning [37, 38]. The learning model can be classified into Deep learning
Overview of Next Generation Smart Grids 15

and traditional machine learning. The various application of ML is electrical power

and energy, transportation, education, healthcare, financial services, and marketing
and sales.

5.3 Blockchain Technology

Blockchain technology can overcome all the difficulties faced by power utility
sectors, and it will effectively help in the futuristic vision of economic herald in
the energy distribution sector. The distributed ledger can start experimenting on
blockchain to adopt new business strategies and goals. There are many applications
of blockchain. One of those is electrical energy distribution. By using blockchain,
one can effectively convert electrical energy into electronic money for transaction
purposes, and it can be converted to any form. Blockchain provides a crystal clear,
secure, and transparent form of transaction. Some of the technical challenges for
applying blockchain technology in electricity trading, checking, and certifications.
It gives a comprehensive view of the implementation of smart contracts under varied
blockchain technologies in power systems. At the same time, implementing this
technology [39]. The most important two items that must be handled with utmost
care in blockchain application are the time frame and data privacy in power systems
applications.
In recent time there is a massive demand for energy, primarily electrical energy.
We are surrounded by innumerous electrical equipment. So, it becomes an integral
part of the generating stations to fulfil this electricity demand and supply of electricity
stably. Prosumers are the producers of electricity, and consumers, on the other hand,
consume electricity. With the help of SG, one can fulfil the objective of consumers
and keep an eye on the transmission losses, which can be minimized to a large extent.
The transaction can be done in a peer-to-peer fashion among the prosumer and the
consumer. To have this type of transaction, blockchain comes into the role to initiate
a centralized transaction system which would be very costly in other means [40]. By
using blockchain, one can solve complex problems that may arise in a SG.
With the help of blockchain technology, the trade of electricity became easier
without any intervention from a third party. A trust is created between the prosumer
and consumer. It has few advantages as we have real-time market records, less trans-
action cost because of simple structure, more privacy within the grid. In a centralized
structure of energy systems where there is a large-scale implementation of industrial
and commercial loads and environmental hazards, the power generation center has
been located from the load centres at a far distance. High voltage (HV) transmission
with distribution is used to transfer the power generated from the source to the load
centres [41]. Currently, three technological trends provide a new dimension of the
transformation in energy fields:
16 S. R. Salkuti et al.

Table 4 Advantages and disadvantages of centralized and distributed systems

System Advantages Disadvantages
Centralized system This system is rapid, and it can make In this system, the central
effective decisions. No duplicity in government controls and handles all
any transactions. This system has an power and authority. Applications
efficient bureaucratic chronology run on a single process. The
component constitutes a single
failure point
Distributed system There are multiple processes for There may be privacy issues
cooperating and executing the Complexity is more with very little
application processes predictability
Workloads can be distributed to They require a lot of effort to
different machines in this system managing all the systems operating
The limit to tolerate faults is Synchronization problems may arise
relatively high
All data can be shared in different
networks, but all the
decision-making processes will only
be held at one particular node

• The energy sector is more and more interested in shifting away from the direct
energy consumption of non-renewable energy sources commercially toward the
integration of different sources or hybrid grids.
• The increasing use of RERs and grids facilitates the decentralization of the energy
field and increases the versatility in real-time operation in the power-system
applications.
• Large parts of the energy value chain are digitalized due to the increased digital
elements in a grid.
Blockchain technology works on a decentralized system having a peer-to-peer
network, which signifies a collective model of trust among the peers who are unknown
to each other and an immutable ledger of records of transactions (distributed). Decen-
tralization means a user-to-user basis network. A Distributed immutable ledger
means data that is stored in it cannot be further modified or deleted. Transactions in
blockchain are just the transfer of values in Bitcoin. UTXO, an unspent transaction
output, defines input and output in the transaction of bitcoin.
Once a block is verified algorithmically, and when a data miner agrees to it, it is
added to the blockchain chain. An unspent transaction output (UTXO) can be spent as
a currency unit in further transactions. The main operation in a blockchain is valida-
tion and block creation of transactions with the consensus of the users participating
in the transaction. There are also many underlying operations. A smart contract
gives the capability for code execution in business logic on a Blockchain. Miner
nodes in the network receive, verify, gather and execute the transaction. Significant
innovations like smart contracts have broadened the applications of blockchain tech-
nologies. Private and permitted blockchains to allow for access to the blockchain in
Overview of Next Generation Smart Grids 17

a controlled manner, enabling much more diverse business models. Table 4 presents
the advantages and disadvantages of centralized and distributed systems.
Blockchain technology has a significant impact on utility industries. It has three
levels of technologies benefitting both permissioned and private ledgers [42]. These
levels are classified based on different technology attributes for the adaptability of
the leading utility.

5.3.1 Level 1: Foundational Technology

Foundational technology consists of bill pay, electric vehicle charging transactions

etc. It enhances the reliability and security of transactions inside the grid frame-
work. The transaction is unified from disrupting generating stations in the energy
ecosystem. Some common application include
• Bill pays: there are a lot of startup companies that are facilitating effective
methodology for transactions. Such as prepaid energy meter, transaction through
cryptocurrency, bitcoins, etc.
• Solar renewable energy certificates: Blockchain technology is used in trading
and earning credit certificates related to renewable energy. One example is the
solar coin that can be made for the generation of solar energy individually.
• EV charging: In the US, most people use an EV and blockchain-based transaction
is there to charge EVs in certain charging stations. Some of the companies which
are using blockchain are bloch charge, innogy. This technology enables the user
for digital payments
• Customer switching: There is a UK-based company known as electron. It is a
blockchain-based company. It enables the user for faster switching of suppliers.
Because of blockchain security and tracking of transactions, this application is
quite fruitful without intermediaries.

5.3.2 Level 2: Medium for Sustaining Business

Utility technology should adopt blockchain as well as internal business units to

sustain the investment made in technology. There should be some regulatory board
to keep an eye on all the transactions [43]. Appropriate support model to ensure
guard of interest of utility business. Some examples of experimental permissioned
blockchain applications with relevance for the utility sector are:
• Microgrids: specific distributed energy systems such as microgrids function inde-
pendently. Distributed energy systems like microgrids mainly operate indepen-
dently from a centralized system due to the small size of the operation. Blockchain
is much more relevant for managing all transactions.
18 S. R. Salkuti et al.

Table 5 Current market shortcomings and the possible role for blockchain
Current market shortcoming Possible role for blockchain
The cost of production is much more than The combination of the Internet of things, records
electric markets and information about generation time/demand,
location, type of generation/demand, and other
aspects of predicting the price for electric market
customers
Integration renewable energy sources Blockchain can be used for de-centralized and
integration into the electric grid automatic grid management and control systems.
As there are many producers and consumers in the
distribution grid, blockchain technology balances
the supply of electricity and demand for auto
verification of grid components
Adequate power resource Cryptocurrencies can be used as a token for
creating new markets and business models. There
will co-ownership for such models
Increase in coordination in-between Blockchain can give promising results in
operators of system developing an automated decentralized and
completely efficient asset control and management
of such assets, including a better balance between
demand and supply at different networks
Integrating demand response for ancillary Smart contracts are designed to engage and reward
services willing customers for helping in DR activity for
maintaining the grid more economically
Much adequate cross-border data Cross-border data exchange incurs costs and data
exchange in power system flows through multiple intermediates, which add
time delays in decision making and thus a potential
loss of data. Blockchains can streamline this
process by removing intermediaries, removing
delays, and maintaining data integrity as data is not
transferred but immutably shared

• Grid settlements and wholesale market trading: independent power producers

integrate a small number of people in a microgrid grid settlement, and whole-
sale markets involve many users within industry regulations, so blockchain does
provide scalability and regulatory compliance to all those units. Table 5 presents
the current market shortcomings and the possible role of blockchain.

6 Electric Vehicles (EVs)

This section focuses on EVs, which are trending in the latest era for meeting the
energy needs replacing the internal combustion engine (ICE) vehicles. In recent
years, EV’s have much attention as compared to conventional internal combustion
engine vehicles. This contemplation is due to the economic and environmental trouble
Overview of Next Generation Smart Grids 19

concerns linked with the utilization of natural gas and petroleum fuels. These are
used as fuel in ICE vehicles. More advancement and researches on vehicles like
fuel cell and hybrid cars, pure battery-powered electric vehicles, etc., are pursued
actively [44]. Because these vehicles are reasoned to serve as an efficient means to
deal with global warming caused by the tail pipe-auto emissions. In eco vehicles,
the compatibility between safety and running performance has to be ensured, espe-
cially in EVs where electric devices like batteries, converters, and inverters are part
of their propulsion force-producing system. If such an electric device fails during
the motion of the EV, the EVs shall fail to run safely and comfortably. This kind of
failure in motion may constitute series of traffic accidents like rear-end collisions. As
an advancement for developing the next generation of EVs, fail-safe functions must
be interlinked within the propulsion force-generating system. The EV can persist in
running safely even if malfunctions occur during EV motion. The hybrid electric
vehicles (HEV), which mainly contains at a minimum of two sources of power, first
one is a primary power source, and the other one is named as a secondary power
source, which has the favours of both the EV and conventional ICE vehicles also
could vanquish their disadvantages. Considering the HEVs and PHEVs, more elec-
trical components are utilized in the making. Electric machines, power electronic
converters, batteries, ultra-capacitors, sensors, and microcontrollers are examples of
various such features. Besides these electrification components and their subsystems,
the traditional internal combustion engine or other mechanical and hydraulic systems
have also been present. Those difficulties generated due to the modern propulsion
systems demand the advanced design of powertrain components, namely power elec-
tronic converters, energy storage, and other electric machines. This power manage-
ment system includes modelling, simulation of the powertrain, hybrid control theory
and optimization control of vehicle [45].
Characteristics of EVs:
• The EV’s are designed to steer with high voltage electrical energy stored batteries.
In this way, the exhaust emissions can be excluded, which further reduces air
pollution since gasoline or other fossil fuels are not required for the propulsion.
• The engine’s noises and vibrations are negligibly small during the entire driving
range compared to traditional ICEs. The electric motor employed in EVs is
generally higher performance motors.
• During the downhill or decelerating motion of the said vehicle, the restoring of
brake energy by regeneration can be employed and stored in the high voltage
batteries. Thereby the driving range from single charging will be extended and
also reduce the loss during braking.
• If the battery’s existing charge is not adequate to drive the vehicle, various charging
methods like AC or DC charge are applicable. The trickle charge can also be
applied.
20 S. R. Salkuti et al.

6.1 Battery Electric Vehicles (BEVs)

A battery electric vehicle (BEV) runs entirely on a battery and electric drive train
without an internal combustion engine. To recharge their batteries, these vehicles
must be plugged into an external electricity source. A hybrid electric vehicle (HEV)
with an internal combustion engine and an electric motor together with an engine
or generator depends on one or more energy sources [46]. A plug-in hybrid electric
vehicle (PHEV) uses rechargeable batteries or other power storage systems that can
be recharged by connecting them to external electricity sources.
Instead of an ICE vehicles, EVs have an electrical motor. The electric battery
vehicle uses a large traction battery pack to power the electric motor and, since it
operates on electricity, it must be plugged into the charging station. The vehicle emits
no exhaust from the tailpipe and does not contain the fuel components such as fuel
pump, fuel line, and fuel tank. The battery pack used to store the electrical energy
that drives the motor is used in the car. By plugging the vehicle into an electric
power source, batteries are charged. BEVs are more costly than conventional hybrid
vehicles.
The battery stores the energy necessary for the vehicle to operate. The battery
supplies the motor with an electric current. And the engine, thus, drives the greater
the battery size, the greater the range of the vehicle. Lithium-ion type batteries are the
most widely used batteries. The higher energy density of these batteries is capable
of storing more energy [47]. Using an inverter, the battery energy in the DC is
converted to AC. The AC is transmitted to the induction motor from the inverter,
which generates a spinning magnetic field that allows the motor to turn.
The single-speed transmission transfers power to the wheels from the induction
motor. This electric motor converts electrical energy into mechanical energy that
turns the wheels to move forward. The car uses more energy during the acceleration,
but the energy transforms through the induction motor into electricity during the
stop. This power is stored in the battery pack and can be used when more energy is
required. An electric vehicle battery can be charged by plugging it into an outlet or
charging station. Advantages of BEV are less air pollution, less noise pollution, higher
efficiency, low fuel and operating cost, and low maintenance cost. The disadvantages
of BEVs include are high prices and low range.

6.2 Hybrid Electric Vehicles (HEVs)

The phrase hybrid vehicle makes reference to a vehicle with the least of two power
sources. Among the two sources of power for HEV, an electric motor acts as one
energy source. In contrast, the other source of motive power can be extracted within
several divergent technologies. But the most used secondary source of power will be
an ICE destined to run either on gasoline or petroleum fuel [48]. As put forward by
the technical committee of the international electrotechnical commission (electric
Overview of Next Generation Smart Grids 21

road vehicles), an HEV is a vehicle in which propulsion energy is available from two
or more types of energy sources, and at least one of them can deliver electrical energy.
Predicted on the definition by the commission, various kind of energy combinations
of HEVs are:
• The battery and gasoline IC engine combo
• Battery and ICE with diesel
• Fuel cell and battery combo
• Battery with ultra-capacitor
• Battery along with flywheel combo
• Hybrid battery and other batteries
Incorporating an ICE with an electric motor is the most widely used combo for
the generation of propulsion force in hybrid EVs. With this integration, energy effi-
ciency will be improved and vehicle emissions will be fall off due to the electric
motor. The driving range will be extended because of the ICE in contemplation of
advancements in the HEV. The stereotypical ICE vehicles bestow good performances
and an extensive operating range because of the high energy density yielded by the
petroleum products. But these standard ICE vehicles have the disbenefit of indigent
fuel economy and more significant risks of environmental contamination [49]. By
far, the reasons behind their indigent fuel economy are:
• The contrast between real operation requirements and the vehicles fuel expertise
• Operating in urban areas, the wastage of kinetic energy during braking
• Due to the stop-and-go driving pattern, the productivity of hydraulic transmission
in current day automobiles.
The EVs powered by battery packs have dominance over typical ICE vehicles by
high energy productivity and almost zero environmental contamination. The compar-
ison of these vehicles based on the performance of their driving range per battery
charge ICEs holds the upper hand. However, HEVs can outsmart both traditional
ICEs and EVs due to the use of two power sources.
An ICE and an electric motor that utilizes energy stored in the battery are powered
by HEVs. Instead of charging the batteries by regenerative braking and an ICE,
a hybrid car cannot be plugged in to charge the battery. A smaller engine could
theoretically allow for the extra power provided by the electric motor. When stopped,
the battery can also control auxiliary loads such as sound systems and headlights
and decrease engine idling. Together, these characteristics result in increased fuel
economy. During coasting or braking, the regenerative braking mechanism restores
energy normally lost. It uses the wheels’ forward motion to spin the engine. This
provides energy which makes the vehicle slow down.
22 S. R. Salkuti et al.

6.3 Plug-In Hybrid Electric Vehicles (PHEVs)

A battery and an electric motor are available for PHEVs. They also have an ICE and
a fuel tank. The PHEV works with electricity until the battery is almost empty. In
the engine, the fuel is then burned to create extra electricity for the electric motor
or power the wheels. With plug-in supplies and a regenerative braking system, the
PHEV batteries can be recharged.
As proposed in the name itself, it’s also a shade of hybrid electric vehicle (HEVs).
Also named as electric vehicles with an extended running range that utilizes battery
packs to deliver energy to an electric motors and an alternate source of energy like
fuel or gasoline to satisfy the energy demands of ICE employed. i.e., similar to HEVs,
PHEVs also make use of both the electrical and mechanical power path with the help
of conventional ICE and general electric machines. The batteries could be charged
alongside plugging through an electric power source or regenerative braking with the
ICE. The PHEV may have larger battery packs than common HEVs [50]. The PHEVs
gain almost all their power while running through urban areas with the already stored
electricity in the battery packs. Before even beginning, one must ensure the battery is
sufficiently charged. In case of distress or complete depletion of batteries, ICEs may
power the vehicle during situations where sudden acceleration, at high speeds, or
when rigorous heating or air conditioning is needed. The dissimilarities of PHEVs
and HEVs depend on the capacity of the battery and their recharging techniques.
PHEV’s are furnished with larger capacities of batteries that can operate only with
battery power within a reasonable driving range, also called an all-electric driving
range.
The utilization of fuel/gasoline in PHEVs depends upon the total distance travelled
within the complete discharging period. The tailpipe emissions of a PHEV are zero if
it’s driven only with battery power. Even if it’s operating with an ICE alone, gasoline
consumption will be practically lesser and produce lower emissions than similar
conventional vehicles do. Comparing with the traditional ICEs that blindly rely on
fossil fuels, PHEVs offer consumers to choose between alternate cheap and clean
energy, thereby diminishing the dependence on a single source of energy.
Two fuel sources, an electric motor and a combustion engine, drive PHEVs.
PHEVs use less fuel and emit fewer emissions than traditional ICE vehicles powered
only by combustion engines. In contrast to HEVs, PHEV uses a high-capacity battery
for energy storage. The greater battery capacity allows the vehicle to cover longer
electric-mode driving distances than hybrid vehicles. The electric motor alone or
the combustion engine powers the car, depending on the driving conditions [51].
The high voltage battery either absorbs energy or emits it. The high voltage battery
releases energy while the electric motor is running, via regenerative braking, the
battery is partially charged by kinetic energy while driving. You should plug the
vehicle in to recharge your high-voltage battery. Charging can be done at any loca-
tion with domestic electricity installed. From the public charging stations, they can
also charge. The car must be parked with the ignition switched off before being
Overview of Next Generation Smart Grids 23

charged. The charging indicator illuminates until the vehicle is fully charged when
the charging cable is attached.
The vehicle functions in two modes to encapsulate it. The first mode is called the
charge depleting mode, in which only the electric motor powers the vehicle before
the high voltage battery reaches a low energy level [52]. When the battery level
is low, the mode of the vehicle switches to the hybrid mode’s charge sustaining
mode. It’s the second mode of driving. The vehicle uses power from the electric
motor to drive in hybrid mode during mild acceleration, providing enough pickup to
help increase speed. The gasoline engine and the electric motor are used to get the
desired speed when the acceleration is higher or uphill motion. The machine goes
to power assist mode while cruising at a relatively steady rpm, where the gasoline
engine is the primary source of drive power [53]. The device captures energy through
a regenerative braking system as the vehicle’s speed decreases and stores it in the
battery for future use. As mentioned above, external charging is often used to optimize
high voltage levels. Table 6 presents different types of EVs and their characteristics.

7 Protection of Smart Grid

A grid would be more efficient when more resiliency is added into the existing electric
network and made ready for diffident unavoidable tragedies and natural calamities.
Therefore, if some extra powerful features are supplemented to the existing grid,
it becomes a SG. The extra powerful features can be advanced protection schemes
in grid operation. This will make the existing grid network more efficient, faster in
power transmission and self-repair after power disturbances, inexpensive, improved
security etc.
Protection in SG refers to the protection of devices connected to the SG against
unintentional failures like fault, overload etc. One of the power system protection
applications in the SG is quick to fault diagnosis to prevent fluctuations in voltage and
power outages. When some schemes protect the physical infrastructures of SG, it is
referred to as physical protection of SG. This type of protection refers to inadvertent
situations because of equipment failure, errors created by human’s, natural disasters
etc. In this type of protection, two essential things have to be looked into. One is the
reliability of the system and the other one is protection scheme failure. Reliability
refers to SG’s component’s reliability and the way they are placed and they are
achieved in four various ways:
• Reliability of distributed generation
• Reliability of measurement infrastructure
• Reliability of network before action
• Decision-making performance by sub-station
RERs are used as DG in a SG whose penetration into the system makes coor-
dination between them with other protection devices quite challenging. As a result
reliability and stability of the system becomes a concern for the protection engineer.
24 S. R. Salkuti et al.

Table 6 Different types of electric vehicles (EVs) and their characteristics

Type of Battery electric Hybrid electric Plug-in hybrid Fuel cell
electric vehicle vehicle electric vehicle hybrid electric
vehicle vehicle
Propulsion Electric motor International International Electric motor
combustion engine combustion
(ICE) and electric engine (ICE) and
motor electric motor
Energy system Battery ICE unit and battery ICE unit and Fuel cell
battery along with
battery and
ultra-capacitor
to enhance
power density
Energy source Electric grid Gasoline station Electric grid Hydrogen,
charging charging and Hydrogen
gasoline station production,
transportation,
and
infrastructure
Characteristics Zero-emission, high Low emission, high Low emission, Zero-emission
energy efficiency, fuel economy, large battery or ultra-low
and high initial cost, long-range of capacity, long emission, high
readily available driving, high cost, driving range in energy
commercially road trips, and efficiency,
available high cost high cost, less
dependent on
fossil fuels
Challenges Battery management, Control of multiple Control of High fuel cell
and issues high cost of charging energy sources, multiple sources, cost, fuel cell
facilities battery management maintenance cost durability, and
reliability,
Hydrogen
infrastructure

One SG operation requirement is the smart measurement infrastructure, which helps

monitor network reliability, stability, and healthiness, which can be met with a phasor
measurement unit (PMU). Suppose the decision making performance by sub-station
can be achieved. In that case, the time span can be reduced as it does not have to wait
for the control network’s decision and the system achieves better stability and reli-
ability. Failure prevention and prediction are two important aspects of SG physical
protection. Prediction means identifying the failure event in the SG, and preven-
tion means to inhibit the failure from an occurrence. Recovery is another impor-
tant aspect of the SG that can be achieved by developing efficient techniques. For
example, a smart meter is used to recover missing data from the system. Because of
the inverter-based DGs into the SG, the operating characteristics differ from standard
distribution systems, making the protection engineer challenging. Another challenge
Overview of Next Generation Smart Grids 25

before power system engineer’s is to attain reliability and efficiency in a wired and
wireless communication network for control and data transfer purpose. For proper
unit protection and to optimize the relay coordination in SG, internet of energy (IoE)
based communication technology is recently used, which mainly exchanges infor-
mation that is further collected by integrated electronic devices (IED). This type of
protection scheme is called i-protection, which precisely identifies the fault with the
help of IEDs and IoE through a wide-area wireless network.

8 Conclusion

Smart grid (SG) has been the most impactful technology in recent times and is
still rising because of its significant benefits. But like any other more substantial
change, transition to the SG system from a conventional power grid system is highly
tedious and time-consuming. At the same time, it is a great decision to meet the
growing electricity demand of modern civilization. The successful transition requires
meticulous preparation, thorough understanding, and proper knowledge, along with
their corresponding impacts on every part of the ecosystem. The distributed energy
resource (DER) applications would increase the efficiency of the energy supply and
reduce the electricity delivery cost and carbon footprint in the SG. The SG needs
to be well-designed, robust structure to withstand the tension, efficient and reliable,
sustainable with the environment, intelligent enough to discover outages and fault
locations, secured and safe from cyber-attacks, able to monitor real-time data and
act upon it, scalable and integrable with other sources of energy. Various features of
the SG need to be implemented cost-effectively to form a symbiotic relationship and
benefit grid operators, energy providers, consumers, and the environment. There are
many challenging issues of protection in SGs like maintaining reliability, stability,
coordination etc. All these can be solved by developing optimized algorithms, which
is also ongoing research.

Acknowledgements This research work was funded by “Woosong University’s Academic

Research Funding—(2021–2022)”.

References

1. Fan D, Ren Y, Feng Q, Liu Y, Wang Z, Lin J (2021) Restoration of smart grids: current status,
challenges, and opportunities. Renew Sustain Energy Rev 143. https://doi.org/10.1016/j.rser.
2021.110909
2. Patil H, Sharma S, Raja L (2020) Study of blockchain based smart grid for energy optimization.
Mater Today Proc 2020. https://doi.org/10.1016/j.matpr.2020.11.013
3. Feng C, Wang Y, Chen Q, Ding Y, Strbac G, Kang C (2021) Smart grid encounters edge
computing: opportunities and applications. Adv Appl Energy 1. https://doi.org/10.1016/j.ada
pen.2020.100006
26 S. R. Salkuti et al.

4. Neffati OS, Sengan S, Thangavelu KD, Kumar SD, Setiawan R, Elangovan M, Mani D,
Velayutham P (2021) Migrating from traditional grid to smart grid in smart cities promoted
in developing country. Sustain Energy Technol Assess 45. https://doi.org/10.1016/j.seta.2021.
101125
5. Alotaibi I, Abido MA, Khalid M, Savkin AV (2020) A comprehensive review of recent advances
in smart grids: a sustainable future with renewable energy resources. Energies 13. https://doi.
org/10.3390/en13236269
6. Zhou J, He L, Li C, Cao Y, Liu X, Geng Y (2013) What’s the difference between traditional
power grid and smart grid?—from dispatching perspective. In: IEEE PES Asia-Pacific power
and energy engineering conference (APPEEC), Hong Kong, China, pp 1–6. https://doi.org/10.
1109/APPEEC.2013.6837107
7. Butt OM, Zulqarnain M, Butt TM (2021) Recent advancement in smart grid technology: future
prospects in the electrical power network. Ain Shams Eng J 12(1):687–695. https://doi.org/10.
1016/j.asej.2020.05.004
8. Moulema P, Mallapuram S, Yu W, Griffith D, Golmie N (2017) Chapter 24—Integrating renew-
able energy resources in smart grid toward energy-based cyber-physical systems. In: Cyber-
physical systems - foundations, principles and applications, intelligent data-centric systems
377–398. https://doi.org/10.3390/en14092657
9. Du P, Lu N, Zhong H (2019) Demand response in smart grids. Energy Renew Green Energy.
Springer. ISBN: 978-3-030-19769-8
10. Gelazanskas L, Gamage KAA (2014) Demand side management in smart grid: a review and
proposals for future direction. Sustain Cities Soc 11:22–30. https://doi.org/10.1016/j.scs.2013.
11.001
11. Ekanayake JB, Jenkins N, Liyanage K, Wu J, Yokoyama A (2012) Smart grid: technology and
applications. Wiley. ISBN: 978-1-119-96909-9
12. GWEC: Global wind report 2016, annual market update, Technical Report, Global Wind Energy
Council (GWEC)
13. PVPS Annual report 2016, Technical Report, Photovoltaic Power Systems Technology
Collaboration Programme
14. IEA: Technology roadmap, wind energy, Technical Report, Energy Technology Perspectives,
2013 edition, International Energy Agency
15. IEA-PVPS: Technology roadmap, solar photovoltaic energy, Technical Report, Energy
Technology Perspectives, 2014 edition, International Energy Agency
16. Moura PS, Lopez GL, Moreno JI, Almeida ATD (2013) The role of smart grids to foster energy
efficiency. Energ Effi 6:621–639. https://doi.org/10.1007/s12053-013-9205-y
17. Acakpovi A, Abubakar R, Asabere NY, Majeed IB (2019) Barriers and prospects of smart grid
adoption in Ghana. Proc Manuf 35:1240–1249. https://doi.org/10.1016/j.promfg.2019.06.082
18. Matisoff DC, Beppler R, Chan G, Carley S (2020) A review of barriers in implementing
dynamic electricity pricing to achieve cost-causality. Environ Res Lett 15(9). https://doi.org/
10.1088/1748-9326/ab9a69
19. Mikalauskas I (2015) Economic, social and environmental benefits of smart grids. Eur J
Interdisc Stud 7(2):30–39
20. Smart Grid Economic and Environmental Benefits (2013) A review and synthesis of research
on smart grid benefits and costs. Technical Report, Smart Grid Consumer Collaborative
21. El-hawary ME (2014) The smart grid—state-of-the-art and future trends. Electr Power Compon
Syst 42(3–4):239–250
22. Kezunovic M, McCalley JD, Overbye TJ (2012) Smart grids and beyond: achieving the full
potential of electricity systems. Proc IEEE 100:1329–1341. https://doi.org/10.1109/JPROC.
2012.2187131
23. Kappagantu R, Daniel SA (2018) Challenges and issues of smart grid implementation: a case
of Indian scenario. J Electr Syst Inf Technol 5(3):453–467. https://doi.org/10.1016/j.jesit.2018.
01.002
24. Dileep G (2020) A survey on smart grid technologies and applications. Renew Energy
146:2589–2625. https://doi.org/10.1016/j.renene.2019.08.092
Overview of Next Generation Smart Grids 27

25. Rohde F, Hielscher S (2021) Smart grids and institutional change: emerging contestations
between organisations over smart energy transitions. Energy Res Soc Sci 74. https://doi.org/
10.1016/j.erss.2021.101974
26. Lalle Y, Fourati M, Fourati LC, Barraca JP (2021) Communication technologies for smart
water grid applications: overview, opportunities, and research directions. Comput Netw 190.
https://doi.org/10.1016/j.comnet.2021.107940
27. Yu K, Arifuzzaman M, Wen Z, Zhang D, Sato T (2015) A key management scheme for secure
communications of information centric advanced metering infrastructure in smart grid. IEEE
Trans Instrum Measure 64(8):2072–2085. https://doi.org/10.1109/TIM.2015.2444238
28. Ghosal A, Conti M (2019) Key management systems for smart grid advanced metering infras-
tructure: a survey. IEEE Commun Surv Tutorials 21(3):2831–2848. https://doi.org/10.1109/
COMST.2019.2907650
29. Iris Ç, Lam JSL (2021) Optimal energy management and operations planning in seaports with
smart grid while harnessing renewable energy under uncertainty. Omega. https://doi.org/10.
1016/j.omega.2021.102445
30. Shi Y, Tuan HD, Savkin AV, Lin CT, Zhu JG, Poor HV (2021) Distributed model predictive
control for joint coordination of demand response and optimal power flow with renewables in
smart grid. Appl Energy 290. https://doi.org/10.1016/j.apenergy.2021.116701
31. Alarifi A, AlZubi AA, Alfarraj O, Alwadain A (2021) Automated control scheduling to improve
the operative performance of smart renewable energy systems. Sustain Energy Technol Assess
45. https://doi.org/10.1016/j.seta.2021.101036
32. Carpejani P, de Jesus ET, Catapan BLSB, Gouvea da Costa SE, Pinheiro de Lima E, Tortato U,
Machado CG, Richter BK (2020) Affordable and clean energy: a study on the advantages and
disadvantages of the main modalities. In: Leal Filho W, Borges de Brito P, Frankenberger F
(eds) International business, trade and institutional sustainability. World sustainability series.
Springer, Cham. https://doi.org/10.1007/978-3-030-26759-9_35
33. Salkuti SR (2020) Energy storage technologies for smart grid: a comprehensive review. Majlesi
J Electr Eng 14(1):39–48
34. Tu C, He X, Shuai Z, Jiang F (2017) Big data issues in smart grid—a review. Renew Sustain
Energy Rev 79:1099–1107. https://doi.org/10.1016/j.rser.2017.05.134
35. Munshi AA, Mohamed YARI (2017) Big data framework for analytics in smart grids. Electr
Power Syst Res 151:369–380. https://doi.org/10.1016/j.epsr.2017.06.006
36. Lai CS, Lai LL (2015) Application of big data in smart grid. In: IEEE international conference
on systems, man, and cybernetics, Hong Kong, China, pp 665-670. https://doi.org/10.1109/
SMC.2015.126
37. Hossain E, Khan I, Un-Noor F, Sikander SS, Sunny MSH (2019) Application of big data
and machine learning in smart grid, and associated security concerns: a review. IEEE Access
7:13960–13988. https://doi.org/10.1109/ACCESS.2019.2894819
38. Günel K, Ekti AR (2019) Exploiting machine learning applications for smart grids. In: 16th
international multi-conference on systems, signals and devices (SSD), Istanbul, Turkey, pp
679–685. https://doi.org/10.1109/SSD.2019.8893252
39. Musleh AS, Yao G, Muyeen SM (2019) Blockchain applications in smart grid-review and
frameworks. IEEE Access 7:86746–86757. https://doi.org/10.1109/ACCESS.2019.2920682
40. Agung AAG, Handayani R (2020) Blockchain for smart grid. J King Saud Univ Comput Inf
Sci. https://doi.org/10.1016/j.jksuci.2020.01.002
41. Dutta SD, Prasad R (2020) Digitalization of global cities and the smart grid. Wireless Pers
Commun 113:1385–1395. https://doi.org/10.1007/s11277-020-07478-w
42. Hasankhani A, Hakimi SM, Niasar MB, Shafie-khah M, Asadolahi H (2021) Blockchain tech-
nology in the future smart grids: a comprehensive review and frameworks. Int J Electr Power
Energy Syst 129. https://doi.org/10.1016/j.ijepes.2021.106811
43. Patil H, Sharma S, Raja L (2020) Study of blockchain based smart grid for energy optimization.
Mater Today Proc. https://doi.org/10.1016/j.matpr.2020.11.013
44. El Bassam N (2021) Chapter Thirteen—Energy storage, smart grids, and electric vehicles.
In: EI Bassam N (ed) Distributed renewable energies for off-grid communities, 2nd edition.
Elsevier, pp 263–295. https://doi.org/10.1016/B978-0-12-821605-7.00008-8
28 S. R. Salkuti et al.

45. Pournaras E, Jung S, Yadhunathan S, Zhang H, Fang X (2019) Socio-technical smart grid
optimization via decentralized charge control of electric vehicles. Appl Soft Comput 82. https://
doi.org/10.1016/j.asoc.2019.105573
46. Brenna M, Foiadelli F, Zaninelli D, Graditi G, Somma MD (2021) Chapter 10—The integration
of electric vehicles in smart distribution grids with other distributed resources. In: Graditi G,
Somma MD (eds) Distributed energy resources in local integrated energy systems, Elsevier,
pp 315–345. https://doi.org/10.1016/B978-0-12-823899-8.00006-6
47. Ahmadian A, Sedghi M, Elkamel A, Fowler M, Golkar MA (2018) Plug-in electric vehicle
batteries degradation modeling for smart grid studies: review, assessment and conceptual
framework. Renew Sustain Energy Rev 81(2):2609–2624. https://doi.org/10.1016/j.rser.2017.
06.067
48. Salehpour MJ, Tafreshi SMM (2020) Contract-based utilization of plug-in electric vehicle
batteries for day-ahead optimal operation of a smart micro-grid. J Energy Storage 27. https://
doi.org/10.1016/j.est.2019.101157
49. Hariri AM, Hejazi MA, Dezaki HH (2021) Investigation of impacts of plug-in hybrid electric
vehicles’ stochastic characteristics modeling on smart grid reliability under different charging
scenarios. J Clean Prod 287. https://doi.org/10.1016/j.jclepro.2020.125500
50. Kong P, Karagiannidis GK (2016) Charging schemes for plug-in hybrid electric vehicles in
smart grid: a survey. IEEE Access 4:6846–6875. https://doi.org/10.1109/ACCESS.2016.261
4689
51. Karthikeyan A, Vijayarajan S (2020) A comprehensive review on electric vehicles. Int J Sci
Technol Res 9(2):2071–2079
52. Zheng Y, Niu S, Shang Y, Shao Z, Jian L (2019) Integrating plug-in electric vehicles into
power grids: a comprehensive review on power interaction mode, scheduling methodology and
mathematical foundation. Renew Sustain Energy Rev 112:424–439. https://doi.org/10.1016/j.
rser.2019.05.059
53. Mwasilu F, Justo JJ, Kim EK, Do TD, Jung JW (2014) Electric vehicles and smart grid inter-
action: a review on vehicle to grid and renewable energy sources integration. Renew Sustain
Energy Rev 34:501–516. https://doi.org/10.1016/j.rser.2014.03.031
Modeling of Various Renewable Energy
Resources for Smart Electrical Power
Systems

Surender Reddy Salkuti

Abstract This chapter presents the modeling of various renewable energy resources
(RERs) such as solar photovoltaic (PV), wind, small hydro, pumped hydro,
geothermal, biomass, and battery energy storage. RERs are an infinitely sustainable
and environmentally friendlier form of energy that will help reduce the dependency
on traditional fossil fuels-based electrical power generation. The abundance of wind
and solar irradiation in nature makes wind and solar PV energy the resources with
great potential if accuracy in planning and technology compiled together. Modeling
of power output from different RERs and the uncertainty handling of solar PV and
wind energy systems are described in this chapter. Battery storage systems provide
energy to the system when the supply from RERs cannot meet the required load
demand. Battery storage units are used to store the extra amount of power produced
by the RERs. Small hydropower generators work at variable speeds as the water flows
at various speeds. Geothermal thermal energy involves accessing underground hot
water or steam from wells several miles into the earth. Sustainable biomass energy
can be generated by burning several types of biomass fuels which are converted into
several forms of energy after combustion. The modeling of all these resources along
with their advantages and disadvantages are discussed in this chapter.

Keywords Battery storage · Biomass · Geothermal · Pumped hydro · Wind

energy · Solar PV · Smart grid · Small hydro

1 Introduction

The fluctuating/intermittent nature of the renewable energy resources (RERs) is an

important factor that limits their large-scale integration into the utility grid. The pene-
tration of RERs has been increased as they can reduce the environmental impact and
meet the increased power demand. All nations throughout the world are promoting

S. R. Salkuti (B)
Department of Railroad and Electrical Engineering, Woosong University, Daejeon 34606,
Republic of Korea
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 29
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6_2
30 S. R. Salkuti

the construction of renewable power plants. The European Union (EU) and the USA
have set targets of 100% and 80% of power generation by RERs by 2050. Recently
several countries have set targets for the reduction of emissions of greenhouse gases
by increasing the share of RERs. New installations of RERs have been dominated
by wind and solar photovoltaic (PV) power plants. However, one needs to develop
suitable models to characterize their intermittent and variable behavior to incorporate
in the planning and operation of electrical power systems [1]. The amount of renew-
able power generation from the solar PV and wind energy systems purely depends on
meteorological conditions such as solar irradiance, ambient temperature, and wind
speed which are directly associated with the site/location. Therefore, the analysis of
solar insolation and wind speed characteristics at a particular installed site plays a
vital role in utilizing solar and wind energy systems efficiently. Presently, the share
of hydro-power generation is almost 58% of the electricity generated by renewable
sources throughout the world. To meet the ever-increasing load demand and for effi-
cient power-sharing with the change in climate conditions and subsequent change in
load an intelligent and robust control techniques are needed [2].
In recent days the power generation from RERs is increasing in a significant way
and the power generated from renewable power sources is going to play a significant
role shortly. To increase the integration of RERs in the conventional power network,
it has to deliver the liabilities acted to the entire system as a consequence of the irreg-
ular nature of these RERs. Because of their effect on system adjusting, scheduling,
reserves management, and commitment of generating units, changeability, and a
ramp or slope events in the power output are the major difficulties to the system
administrators [3, 4].
The European power system with ambitious de-carbonization with the efficient
coupling of RERs to conventional energy systems has been presented reference [5].
Reference [6] proposes a combined energy system, i.e., an energy hub consisting of
electrical, heating, and cooling equipment, demand response, and the optimization
of RERs. The changes of optimization techniques brought by large-scale integration
of RERs are presented in [7]. Reference [8] presents the importance of long-term
uncertainty modeling approaches to evaluate various support policies of renewable
and de-carbonization. A simulation model based on a systems dynamics method-
ology for representing the complex behavior of the national electricity market under
different scenarios analysis has been presented in [9]. An approach of utilizing multi-
type demand-side reserve offers to smooth fluctuations in RERs forecast on various
time scales has been presented in [10]. An optimal energy management problem of
microgrids with RERs and energy storage systems has been presented in reference
[11].
The aim of this chapter is to presents the modeling of various renewable energy
resources (RERs) such as solar, wind, small hydro, pumped hydro, geothermal,
and biomass. The rest of this work is organized as follows: Sect. 2 describes the
modeling of the wind energy system and its uncertainty handling approach. Modeling
of output power from solar PV energy system and its uncertainty handling approach
is described in Sect. 3. Section 4 presents the modeling of a battery energy storage
system. Modeling of small hydro and pumped hydro energy systems is described in
Modeling of Various Renewable Energy Resources for Smart … 31

Sect. 5. Sections 6 and 7 describe the modeling of geothermal energy and biomass
energy. Section 8 summarizes the major contributions with concluding remarks.

2 Wind Energy System

Wind energy is the kinetic energy (KE) of the air in motion which is more economical,
reliable, and the operation of a wind turbine (WT) is quite simple. As the wind blows
through the blades of WT with the pressure exerted on the surface area of the blade.
Aerodynamic forces cause the blade to turn the rotor. The generator and gearbox
are kept together in a single unit behind the blades. The output of the generator is
passed through for conversion from DC to AC. The WT converts the KE from the
wind into mechanical energy. This energy could then be used to generate electricity.
It is one of the renewable energies that were discovered to transform electric power
by enhancing the sustainability, performance, and reliability of the smart grid [12].
After converting the wind into mechanical energy, that energy can be used to produce
electricity is by using a wind generator. WTs are manufactures in the range of vertical
and horizontal axis types and each type of wind turbine is used for a specific purpose.
For example, large wind turbines are used in the production of commercial electric
power whereas the smallest turbines are used to provide additional power on boat
sailing or battery charging.
Wind energy is environmentally friendly and produced naturally as there is no
need for burning fuels to generate wind power. Wind turbines don’t require a lot of
space because windmills cover only a few square meters and the rest of the land can
be used for farming or agriculture purposes. However, the wind turbine can make a
disturbing noise, for those who live a hundred meters from the wind power station
[13]. Wind turbines can be used for standalone or hybrid power supply. A power
regulator is usually used to regulate the power output between the turbines’ output
and the load. Transformers are used to step the voltages down or up depending on
the use of voltages.
A WT has an expected average life of 20–25 years, which makes it a great source
of alternate power. It is being recognized as a vital component of sustainable envi-
ronmental and economic well-being, long-term, and security of energy supply and
price. Although wind energy is reliable and inexpensive, it requires windmills to be
built in areas where the wind is not only strong but also steady, which may be a disad-
vantage for some countries. Offshore may be the best place for wind power because
winds are usually strong and steady over there. The height of the turbine is a huge
factor in its power generation as the impact of a taller is considered to give higher
winds in the WT also, and the surface winds can be affected easily by the roughness
or irregularities of the earth surface [14]. Wind turbine also produces no emission,
and their efficiency in its application area as green power is top-notch. The common
issue with the turbine is its most suited application, i.e., for peak shaving applications
and the power it produces is dependent on strong winds. The wind turbine is fairly
inexpensive compare to the PV system.
32 S. R. Salkuti

The working of wind turbines depends on several factors such as the shape and
size of the rotors, height of the blades, and location of the turbine. Advantages of the
wind turbine technology include: environmentally friendly, low cost in its generation
of energy, a big source of energy and large industrial base already exists, it could be
scaled down to small turbines used in companies to power small institutions, offshore
advantage allows the mounting of the huge turbine to produce a large amount of
energy [15]. However, there are some disadvantages such as the intermittent nature
of wind power, noise generated from the rotating blade and motor, loss of scenery,
requirement of large amounts of land.

2.1 Modeling of Wind Turbines (WTs)

The WT is connected to an inductive/asynchronous machine that consumes the reac-

tive power (Q) and generates the active power (Pm ) [16]. This active power output
is given by,

1
Pm = ρπr 2 v 3 C p (1)
2

where r is the radius of WT (in m), v is wind speed (in m/s), ρ is air density (kg/m3 ),
C p is turbine power coefficient. Then the electrical power (Pe ) is given by,

Pe = n 0 Pm (2)

where n 0 = n m n g . n g and n m are efficiencies of generator and turbine, respectively.

2.2 Modeling of Wind Speed

The stochastic behavior of wind speed can be analyzed statistically by using the
Weibull Probability Distribution Function (PDF). Weibull PDF of a certain wind
speed at a particular time (t) can be expressed as,
  (kt −1)    
kt vt vt (kt −1)
f (v) =
t
exp − (3)
ct ct ct

The scale factor (ct ) and shape factor (kt ) at a particular time (t) can be evaluated
by using [17],
Modeling of Various Renewable Energy Resources for Smart … 33

μt
ct =  v  (4)
 1+ 1
kt
 −1.086
σvt
kt = (5)
μtv

where () is the gamma function. σvt and μtv are the standard deviation and mean of
wind speed (vt ) at a particular time t.

2.3 Modeling of Wind Power Output

The amount of power output from wind energy generators (WEG) falls in three
different regions. When the wind speed (vt ) is less than the vcin or higher than the
vcout , then the WEG/WT will produce zero power output (i.e., Pwt = 0). When vt is
between vcin and vr , then the wind power will increase linearly (as represented in
Eq. (6) and it is called the continuous range. When vt is between vr and vcout , then
it is going to generate rated wind power (i.e., Pwr ), and it is a discrete PDF. In the
similar lines, below the vcin or above the vcout , the WT will not produce any real
power and it is also considered as a discrete range. Then the power output can be
expressed as [18],
⎧
⎪
⎨  0   vt < vcin andvt > vcout
Pwr v Pwr
Pwt = vr −vcin
vt − vrcin
−vcin
vcin ≤ vt ≤ vr (6)
⎪
⎩
Pwr vr ≤ vt ≤ vcout

Generally, the wind power output is dependent on wind speed. The (Pwt ) with
wind speed can also be expressed as a cubical [19]. The amount of wind power
generated from a WT at a particular time ‘t’ can be expressed as,
⎧
⎪
⎨  0  vt < vcin andvt > vcout
Pwr vcin
3
Pwt = vt3 + Pwr vcin ≤ vt ≤ vr (7)
⎪ vr3 −vcin vr3 −vcin
3 3
⎩
Pwr vr ≤ vt ≤ vcout

where vt is the average wind speed at a particular time t. vr , vcout and vcin are the rated,
cut-out, and cut-in wind speeds. Pwr is rated/maximum wind power. The amount of
wind power generated can also be expressed as,

1
Pwt = ρ K max A f vt3 (8)
2
34 S. R. Salkuti

where K max is power coefficient, ρ is the air density, and A f is swept area by the
rotor of WT. The wind technology option uses modeling techniques like Gaussian,
Weibull, Rayleigh, and Beta PDFs due to its intermittent and variable behavior.

2.3.1 Wind Power Output in Continuous Range

In continuous range (i.e., the (Pwt ) in the region between cut-in to rated wind speeds),
the random variable transformation is used to transform the wind speed random
variable (vt ) to the wind power output (Pwt ) random variable. In this continuous
range, the Weibull PDF can be expressed as [19],
⎡  k ⎤
 vi + p
k(vr − vi ) p (vr − vi )
vi + (vr − vi )k−1 ex p ⎣− ⎦
pr
f p ( p) = (9)
c ∗ pr
k pr c

where f p ( p) is the WEG/wind power PDF. The above PDF (i.e., Eq. (9)) is valid
only for the continuous range.

2.3.2 Wind Power Output in Discrete Range

From Eqs. (6) and (7), it can be seen that two different discrete probabilities occur
when there is no power output from the wind plant (Pwt = 0W ) and the rated power
output (Pwt = Pwr ) [20]. The probability of an event at (Pwt = 0) is given by,
       
vi k vo k
Pr (P = 0) = Pr (V < vi ) + Pr (V ≥ vo ) = 1 − e − +e −
c c
(10)

The probability of an event at (Pwt = Pwr ) is given by,

       
vr k vo k
Pr (P = Pr ) = Pr (vr < V < vo ) = e − +e − (11)
c c

From the above Eqs. (9), (10), and (11), it is clear that it is a mixed probability
function with continuous and discrete power outputs.

3 Solar PV Energy System

Solar energy is an inexhaustible energy source. It is one of the renewable energies

which transform the electric power system. Scientists have found ways to harness
Modeling of Various Renewable Energy Resources for Smart … 35

the power of the sun. The process consists of converting the sunlight into electricity,
using photovoltaic (PV) cells. PV converts light into electric current using the PV
effect. Solar energy works very well in areas where the sun shines almost every day.
A Photoelectric system converts light energy to electricity. It consists of multiple
components. There are three main models of solar PV systems, i.e., grid-connected,
independent/stands alone and integrated models. The grid-connected solar PV system
operates a DC power to a power conduction unit converter that converts the DC to
AC, whereas the standalone model is equipped with a battery energy storage and
generator for backup power. For instance, the power produced by panels can be
used to supply homes or buildings or a power grid. Since the energy production by
solar PV panels is DC, an inverter is used to convert it to AC and a regulator is
used to regulate the power. Transformers are used to step the voltages down or up
depending on the use of those voltages. Whereas in the integrated system, the solar
PV system is coupled directly to their loads without major power conducting device
or the storage battery. This PV system is friendlier to the environment due to its no
emission, pollution, and minimal health hazards [21].
Two main factors considered for the solar PV systems include insola-
tion/irradiance and emission. Insolation is the availability of solar PV energy conver-
sion to electricity. Various factors affecting solar irradiance include the operating
temperature of the solar cells, position of the solar panel, and intensity of light.
Different materials applied for solar PV cells are amorphous silicon and cadmium
telluride, these are environmentally friendly. The cost of power from solar technology
is expected to reduce in the nearest future. Due to its stationary parts and constant
availability of the sun increases the lifetime of the solar PV system.
Applications of solar PV energy systems include solar power, solar vehicles, agri-
culture, solar lighting, water heating, water cookers, water treatments, etc. Some of
the advantages of solar PV systems include abundant availability, quick installation,
low cost in the application and processing, low maintenance, and easy customization
[22]. However, the disadvantages include higher cost than fossil energy, dependent
on the sun, high capital investment, and finding suitable locations for building solar
PV plants. Solar technology is going to play a larger role in the future, as there are
new developments that could result in lower costs and improved efficiency.

3.1 Modeling of Power Output from Solar PV Plant

In setting up of solar PV plants and systems, data typically gotten over a year takes
to account hourly global direct insolation, temperature and wind speed data, array
geometry, and balance of the system. Generally, the solar PV panels are tilted to get
maximum solar irradiance on them. Beam and diffuse radiation incident on solar
PV panels surface govern the amount of electricity generation. Total hourly solar
irradiation on a titled surface (G t in W/m2 ) can be expressed as [22],

G t = G b Rb + G d Rd + (G b + G d )Rr (12)
36 S. R. Salkuti

Fig. 1 Equivalent circuit of

the solar PV cell

where G b and G d are the beam part and diffused part of solar irradiations. Rb , Rd
and Rr are the tilt factors for beam, diffused, and reflected part of solar radiations.
The active power output from solar PV power (Ppv ) can be represented by [23],

Ppv = η AG t (13)

where η is the photoelectric conversion efficiency and A is the surface area of solar
PV systems (in m2 ). The most commonly used model for PV technology and the
equivalent of solar PV cell circuit model has been depicted in Fig. 1.
The amount of current flowing through a load of the solar PV cell (I) is given by,
    
q(V +I Rs ) V + I Rs
I = Isc − Ios e nK T
−1 − (14)
Rsh

where K is Boltzmann’s constant, Isc is short circuit current, q is electron charge (1.6
× 10–19 C), Rs and Rsh are series and shunt resistances for cells. T is the absolute
temperature of the p–n junction; V is the voltage across solar cell, n is the ideality
factor, which is a number between 1 and 2. Saturation current (Ios ) is given by,
 
−E g
γ
Ios = AT e nK T
(15)

where A is temperature constant, 7 is temperature dependency exponent, E g is energy

gap. The current of an ideal solar cell (Id ) is given by,
 q 
Id = Isc − Ios e( nkT Ed ) − 1 (16)

However, for an array of (Ns × Nsh ) solar PV cells,

Idmod
Id = (17)
Nsh

E dmod = E d × Ns (18)
Modeling of Various Renewable Energy Resources for Smart … 37

 
Ns
Rsmod = Rs (19)
Nsh
 
Ns
mod
Rsh = Rsh (20)
Nsh

where Ns , Nsh are the number of series and parallel cell junctions of a PV module.
mod
Rsmod is series resistance, Rsh is shunt resistance, E dmod is voltage and Idmod is current
of the entire module. The power output of solar PV panel is given by [24],

Gt r  
Ppv (G) = FP V Ppv 1 + γ (Tt − Tstd ) (21)
G std

G t is solar irradiance (in W/m2 ) at time t, G std is standard solar irradiance

(1000 W/m2 ), FP V is the factor reflecting the shading and wiring losses, Ppv r
is
the rated power output, Tt is the temperature at time t, Tstd is standard test condi-
tions temperature (25 °C), γ is the maximum correction for temperature. Some of
the factors that determine the output given by solar technologies include the type of
material, the area and intensity of the sun in the region, the cross-sectional area of
the solar PV panel, and the type of connectivity it operates.
By neglecting the solar PV cell temperature, the amount of power output from
solar PV cell (Ppv (G)) can be represented by [25],
⎧  
⎨ Pr Gt
f or 0 < G t < Rc
Ppv (G) = 
pv G std Rc
 (22)
⎩ Pr G t
f or G t > Rc
pv G std

where Rc is a certain solar PV irradiation point (say 150 W/m2 ).

3.2 Uncertainty Modeling of Solar PV Unit

Solar power is free after the recovery of the initial investment, and it is also environ-
mentally friendly. Solar energy does not require expensive maintenance and finally,
it does not pollute our air and it operates silently. Solar energy can be harness asong
as the sunshine and the energy can be stored in batteries for the evening and overnight
use. Solar panels don’t require a lot of maintenance and 90% of the solar panels last
30 years or more. At a selected location, the solar irradiation follows a bi-modal PDF,
which can be modeled as two unimodal PDFs. The unimodal PDF can be modeled
by using the Beta PDF, Weibull PDF, and Log-normal PDF. The modeling of hourly
solar irradiation by using Weibull PDF can be represented by [25],
38 S. R. Salkuti

  k1 −1   k1    k2 −1   k2 

k1 Gt Gt k2 Gt Gt
f G (G) = ω exp − + (1 − ω) exp −
c1 c1 c1 c2 c2 c2
(23)

where ω is the weight parameter (0 < ω < 1). c1 , c2 , k1 and k2 are scale and shape
factors, respectively. As mentioned earlier, the solar irradiance follows a Beta PDF.
This Beta distribution of solar irradiance (G t ) at a particular time t can be expressed
as [26],
 
 αt + β t (αt −1)
(1 − G t )(β −1)
t
f GT (G) = Gt (24)
(α )(β )
t t

where α t and β t are the shape parameter at a particular time t, and they are calculated
by using,
   
  μtG 1 + μtG
β = 1 − μG
t t
 t 2 −1 (25)
σG
μtG β t
αt =   (26)
1 − μtG

where μtG and σGt are the mean and standard deviation of solar irradiance (G t ) at a
particular time t. The expectation (E) and variance (Var) of the random variables (X)
of the Beta distribution are expressed using,

αt
E(X ) = (27)
αt + β t
αt β t
V ar (X ) = (28)
(α t + β t )2 (1 + α t + β t )

4 Battery Storage Modeling

The battery storage system is employed as a backup to store the excess of power
when the supply from RERs exceeds the load demand. This battery will also provide
energy to the system when the supply from RERs cannot meet the required load
demand. Battery storage units are used to store the extra amount of power produced
by the RERs. Generally, Lithium-ion (Li-ion) batteries are used for storage. Li-ion
battery has a low-self discharge rate, higher better energy capacity, and the ability
to be recharged hundreds of thousands of times than other battery types. The Li-ion
Modeling of Various Renewable Energy Resources for Smart … 39

batteries have high power density, high efficiency, high energy density, and longer
life cycle [27]. For the longer life of the battery, the minimum SOC is considered
as 20% and the round trip efficiency is considered as 80%. The amount of energy
stored in a battery is expressed as,

t
Q B = Q 0B + ∫ VB I B dt (29)
0

The state of charge (SoC) of a battery is expressed as,

 
QB
SoC Bt = × 100 (30)
Q max
B

where Q B is the total energy of battery at time t, Q 0B is energy available at the

beginning of the scheduling interval. Storage systems play a vital role in changing
the generation to suit the requirement which leads to enhancing the consistency and
performance of the system against the uncertainties [28]. The SoC of the storage
system (in Wh) is expressed as,
   
Pct Pdt
SoCt = SoCt−1 + ηcB − (31)
t t

where ηcB is the charging efficiency of the battery, t is the scheduling interval (say
1 h). Pdt and Pct are the discharging and charging powers of the battery storage system.
The SoC for the battery is considered an important parameter to be controlled [29].
The variation of battery SoC is expressed by using,

1 t
SoCt = SoC0 + ∫ idt (32)
Q 0

where Q is maximum capacity and ∫ idt is the instantaneous capacity of the battery
bank. Maintaining the SoC within the limits is essential to prevent it from any over-
charging or under discharging, and for a longer lifetime of the battery, and this
constraint can be expressed as [30],

SoCmin ≤ SoCt ≤ SoCmax (33)

Batteries are degraded by the depth of discharge (DOD), temperature, and SoC,
which minimizes the charging and discharge sequence of the storage system. The
degradation cost of battery (Cdt ) is given by [31],

I C B PBt t
CB = (34)
Nct S B ηcB ηdB
40 S. R. Salkuti

Degradation cost during charging (Cd,ch

t
) can be expressed as,

α Pb,ch
t
ηc C p
t
Cd,ch = (35)
Tc N ηd τ

Degradation cost during discharging (Cd,disch

t
) can be expressed as,

(1 − α)Pb,disch
t
ηc C p
t
Cd,disch == (36)
Td N ηd τ

where Tc and Td are the average number of charging and discharging hours per day.
τ is the time period after which the decay is measured. N is charging and discharging
cycles of the storage system, α is the ratio of average charging time to total operating
time per day, C p is the capital cost of the storage system. The cost function of
the battery storage system in terms of DOD during the charging and discharging
operation can be represented by [32],

I C B PBt t
CB = (37)
Nct S B ηcB ηdB

I C B is the investment cost of battery storage, t is the time step (say 1 h), PBt
is power output from storage battery at time t, Nct is the number of cycles of energy
storage at a particular DOD at time t, S B is the total capacity of the battery storage
system in Wh. ηdB and ηcB are discharging and charging efficiencies of the battery.

5 Small Hydro and Pumped Hydro Energy Systems

Water is another form of renewable energy that creates hydroelectric power is usually
located on dams where there is an abundance of moving water. To produce water
energy the industrials that produce water energy push water through pipes to turn a
turbine which is hooked to a generator. Once the turbine is in motion it then turns
the generator which produces electricity. This process consists of releasing kinetic
energy, which is transformed into mechanical energy, and then the production of
electricity. The working principle of a pumped-hydro storage plant is similar to the
traditional hydropower plant. Small hydropower generators work at variable speeds
as the water flows at various speeds.
The main advantage of hydro energy is the fact that water is natural and it can be
retained in one location and to be used later for energy production. As hydropower is
fueled by water, it is a clean source of power as it does not pollute the air as fossil fuels
do. This power can be made readily available on-demand (very high ramp up and ramp
down limits) as the flow of water through the turbines can be controlled. Hydropower
facilities create reservoirs that provide some kind of recreational activities [33].
Modeling of Various Renewable Energy Resources for Smart … 41

The hydro-electric power dams are very expensive to build and require adequate
preparation to execute. These dams are believed to last 100 years and since we will
continue to have water as a resource, then this technology is quite sustainable. These
plants do not produce any emissions or create waste but the dams disturb the scenery,
aquatic ecosystem, and surrounding lands.

5.1 Modeling of Small Hydro Power

The power plants that create hydroelectric power are usually located on dams where
there is an abundance of moving water. To produce water energy the industrials that
produce water energy push water through pipes to turn a turbine which is hooked to a
generator and the generator generates electricity. Large-scale hydro energy dams are
used to power grids. Transformers are used to step the voltages down or up depending
on the use of those voltages. The small hydro generator allows the cases of the high
head with a low volumetric rate of flow or a high volumetric rate of flow with a
low head for operation. The amount of electrical power obtained from small hydro
generator (PH ) can be represented by using [34],

gηh h e f f ρw Q w
PH = (38)
1000W/kW

where ρw is the water density (kg/m3 ), g is acceleration due to gravity (9.81 m/s2 ),
h e f f is the effective height (m), Q w is the flow rate of water (m3 /s), ηh is the efficiency
of the small hydro generator (%). This potential energy (E) of hydropower can be
represented by,

E = mgh (39)

Relating power to mass flow rate can be expressed as,

E m
= gh (40)
t t

Hence, for the hydro-electric power (P) can be expressed as,

P = ρϕgh = hrgk (41)

where m is mass, k is efficiency coefficient, and ϕ is the rate of fluid flow.

42 S. R. Salkuti

5.2 Modeling of Pumped Hydro Storage Plant

The working principle of a pumped-hydro storage plant is similar to the traditional

hydropower plant. The amount of power generated from the pumped hydro plant
during the generating mode (Pgen ) can be expressed as [35],

Pgen = ηg Q w h e f f g (42)

The amount of power used by the pumped hydropower plant to pump the water
(Ppump ) is given by,

Qw he f f g
Ppump = (43)
ηp

where η p is the pumping efficiency.

6 Geothermal Energy

It is clean and sustainable energy. It saves about 80% cost of fossil fuels. Also, it
decreases our dependence on fossil fuels since geothermal energy does not use fuel
to generate power. Another important benefit of geothermal energy is the fact that
geothermal energy does not pollute the air and people use this resource to heat/cool
their homes, prepare their foods, which is a creation of jobs for the local people. The
installation of a power plant to get steam from deep within the earth requires high
investments. The process requires recruiting highly skilled people for the job, and
those workers are usually relocated since this type of energy power is not widely used.
Another disadvantage is the geothermal site can run out of steam after injecting heavy
investments on that site which will make companies lose a lot of money. This energy
is available only in the regions where the temperatures are low and can produce steam
for a long period of time [36]. However, geothermal energy is not environmentally
friendly and may contain poisonous gases that can escape from the holes dug by the
workers.
Due to the large electrical energy output, geothermal plants can only be connected
to a grid to supply power to a city, town, etc. Transformers are used to step the voltages
down or up depending on the use of those voltages. The cost of geothermal energy
varies based on different criteria, such as temperature, depth of area drilled, etc. They
are connected to a grid in regards to the power supply. Generally, there are three main
types of geothermal power units:
• Dry-steam power units: These plants/units use the steam from underground to
drive the turbines to produce electricity.
Modeling of Various Renewable Energy Resources for Smart … 43

• Flash-power units: These plants pull water at high pressure from underground.
As the water rises, the pressure drops, and hence, the water vaporizes into steam
which drives the turbine. Any leftover water or condensed steam is condensed
and renewed.
• Binary cycle power units: These use heat from hot water from underground to
boil a working fluid that vaporizes in a heat exchanger and then it turns to the
turbine.

6.1 Modeling of Geothermal Energy

As mentioned earlier, geothermal energy involves accessing the hot water and under-
ground steam from the wells to several miles into the earth. The main feature of this
technology is the steam turbine. Essentially, the steam that comes out of the earth’s
crust is used to drive a turbine which activates a generator and ultimately produces the
power required. Some power plants use steam obtained directly to power a turbine,
while some others use hot water obtained to vaporize a liquid which in turn turns a
turbine. The heat energy (H ) equation is given by [37],

H = m × C p × T (44)

where T is change in temperature, m is mass, Cp is specific heat. The maximum

heat power generated from a geothermal source is given by [38],
 
Pmax = kv T − Tg = kvT (45)

where k is thermal conductivity, v is water outflow capacity. T and Tg are water and
border temperatures, respectively.
Geothermal technologies are clean, sustainable, cost-effective, and reliable. It
requires no fuel and it is relatively environmental-friendly as it does not produce as
much pollution as fossil fuels. However, although the technological advancements
have led to the broadening of the range and scale of power obtained here, it has broadly
been limited to areas close to tectonic plate boundaries. Drilling and exploration of
wells/pipes for deep underground resources are very expensive. Geothermal wells
release greenhouse gases trapped underground. But, these gases are far less pollutant
than the gases emitted by fossil fuels.
A lot of gases including carbon dioxide, hydrogen sulfide, and ammonia are get
carried by water from underground. This eventually leads to the release of pollutants
into the atmosphere which could cause acid rain, global warming, or other conditions.
It is important to note that these pollutants do not have as much negative impact
as fossil fuels. Also, the drilling and construction of wells to provide access to
geothermal energy sometimes affects land adversely and could trigger earthquakes
as part of hydraulic faulting.
44 S. R. Salkuti

7 Biomass Energy

Biomass energy is sustainable. This energy is produced from biomass organic matter
like wheat, corn, soybeans, and raisins that can generate materials and chemicals
which we generally get from petroleum. Biomass energy is generated by burning
many types of biomass fuels such as wood, municipal solid waste, refused derived
fuel, etc. The biomass fuels are converted into many forms of energy after combus-
tion: hot water, hot air, electricity, etc. Bio-power can also be derived from the
gasification process. This is the conversion of gas to a gas turbine to produce elec-
tricity. The Biomass energy conversion process utilizes the concept of pyrolysis of oil.
Biomass can be converted directly into fluid fuels, most common fuels are bio-diesel
or ethanol. Organic material will be always available to be used as energy resources.
This means that biomass plants can last longer than the human race. Biomass power
can only be produced when sufficient bio-products are available and converted. Its
efficiency is still very low and is also not an ideal green/sustainable power due to
pollutants emission.
Biomass causes no increase in carbon dioxide emissions to the atmosphere.
Carbon is removed from the atmosphere as the trees and plants grow, and the use
of biomass energy does not contribute to global climate change, which is a huge
advantage to the world. Also, using biomass to produce energy is a way of disposing
of the waste material and preventing our environment from being deteriorated by
those wastes. For example, burning the waste from particular cities would prevent
those cities from having wasted all over the places or in the countryside. Another
advantage of this source is that biomass energy is available throughout the world.
Biomass plants generate AC power to supply a grid. This is due to the fact biomass
plants produce large output power. However, the burning of biomass causes pollution
and global warming [39].
The main advantages of bio-energy include the use of waste products, carbon–
neutral, continuous source unlike solar or wind; cost of resources is low, reduced level
of methane, and easy availability. However, burning the biomass causes pollution and
global warming. Other disadvantages include limited potentials, require top-notch
management, may create air pollution if not handled carefully and fear of health
hazards if close to residential homes.

7.1 Modeling of Biomass Energy

Biogas production is made from biomass feedstock, i.e., agriculture residue, house-
hold waste, and forest residue in this case study biomass is used to produce biogas.
The biomass feedstock potential estimation is made considering the efficiency of 75%
as collection rate and the average collected feedstock of 160 kg/ha/yr. 80% of the
total biomass feedstock potential is suggested to use as electricity conversion using
a gasified biomass-based system. The hourly power output from biomass generator
Modeling of Various Renewable Energy Resources for Smart … 45

(PBio in kW) can be calculated by using [40],

η Bio FA C VBio × 1000

PBio = (46)
365 × 860 × Oh day

η Bio is the efficiency of biomass generator (%), FA is total fuel availability (in
tons/year), C VBio is the caloric value of biomass, Oh day represents the number of
operating hours per day.

8 Conclusions

Renewable power generation from the solar PV and wind energy systems for a
particular region is almost complimentary as the wind speed is high during the night
and solar irradiance is high during the day. Therefore, obtaining an optimum mixture
will give maximum potential benefits to the system. This chapter has presented the
modeling of various renewable energy resources (RERs) such as solar PV, wind, small
hydro, pumped hydro, geothermal, and biomass. Biogas has stated benefits (cleaner
combustion, high efficiency, ameliorates control) over solid biomass though having
a low heating value in comparison to fossil fuels. The main advantage of geothermal
energy is that geothermal energy is relatively cheap compared to fossil fuel.

Acknowledgements This research work was funded by “Woosong University’s Academic

Research Funding—2021”.

References

1. Akintande OJ, Olubusoye OE, Adenikinju AF, Olanrewaju BT (2020) Modeling the deter-
minants of renewable energy consumption: evidence from the five most populous nations in
Africa. Energy 206. https://doi.org/10.1016/j.energy.2020.117992
2. Assi AF, Isiksal AZ, Tursoy T (2021) Renewable energy consumption, financial development,
environmental pollution, and innovations in the ASEAN + 3 group: evidence from (P-ARDL)
model. Renew Energy 165(1):689–700. https://doi.org/10.1016/j.renene.2020.11.052
3. Mohammed NA, Al-Bazi A (2021) Management of renewable energy production and distri-
bution planning using agent-based modelling. Renew Energy 164:509–520. https://doi.org/10.
1016/j.renene.2020.08.159
4. Bramstoft R, Pizarro-Alonso A, Jensen IG, Ravn H, Münster M (2020) Modelling of renewable
gas and renewable liquid fuels in future integrated energy systems. Appl Energy 268. https://
doi.org/10.1016/j.apenergy.2020.114869
5. Bernath C, Deac G, Sensfuß F (2021) Impact of sector coupling on the market value of renew-
able energies—a model-based scenario analysis. Appl Energy 281. https://doi.org/10.1016/j.
apenergy.2020.115985
6. Li L, Ling L, Yang Y, Poursoleiman R (2021) Modeling and optimal energy operation consid-
ering probabilistic and sustainable renewable energy models and demand side management.
Energy Build 231. https://doi.org/10.1016/j.enbuild.2020.110557
46 S. R. Salkuti

7. Deng X, Lv T (2020) Power system planning with increasing variable renewable energy:
a review of optimization models. J Clean Prod 246. https://doi.org/10.1016/j.jclepro.2019.
118962
8. Scott IJ, Botterud A, Carvalho PMS, Silva CAS (2020) Renewable energy support policy
evaluation: the role of long-term uncertainty in market modelling. Appl Energy 278. https://
doi.org/10.1016/j.apenergy.2020.115643
9. Valencia-Calvo J, Olivar-Tost G, García-Ortega M (2020) Model and simulation of a renew-
able energy market: integration of renewable energy sources with the conventional generation
system. In: 15th Iberian conference on information systems and technologies, Sevilla, Spain,
pp 1–7. https://doi.org/10.23919/CISTI49556.2020.9140987
10. Yi W, Zhang Y, Zhao Z, Huang Y (2018) Multi objective robust scheduling for smart distribution
grids: considering renewable energy and demand response uncertainty. IEEE Access 6:45715–
45724. https://doi.org/10.1109/ACCESS.2018.2865598
11. Rahbar K, Chai CC, Zhang R (2018) Energy cooperation optimization in micro-grids with
renewable energy integration. IEEE Trans Smart Grid 9(2):1482–1493. https://doi.org/10.1109/
TSG.2016.2600863
12. Rekioua D (2014) Wind power electric systems—modeling, simulation and control. In: Green
energy and technology, Springer-Verlag London. ISBN: 978-1-4471-6425-8
13. Soriano LA, Yu W, Rubio JJ (2013) Modeling and control of wind turbine. Math Probl Eng.
https://doi.org/10.1155/2013/982597
14. Yaakoubi AE, Amhaimar L, Asselman A (2020) Wind energy conversion system modeling
toward different approaches. In: Maalawi KY (ed) Design optimization of wind energy
conversion systems with applications, IntechOpen. https://doi.org/10.5772/intechopen.89066
15. Afanasyeva S, Saari J, Kalkofen M, Partanen J, Pyrhönen O (2016) Technical, economic and
uncertainty modelling of a wind farm project. Energy Convers Manage 107:22–33. https://doi.
org/10.1016/j.enconman.2015.09.048
16. Evangelopoulos VA, Georgilakis PA (2014) Optimal distributed generation placement under
uncertainties based on point estimate method embedded genetic algorithm. IET Gener Transm
Distrib 8(3):389–400. https://doi.org/10.1049/iet-gtd.2013.0442
17. Hetzer J, Yu DC, Bhattarai K (2008) An economic dispatch model incorporating wind power.
IEEE Trans Energy Convers 23(2):603–611. https://doi.org/10.1109/TEC.2007.914171
18. Liu X, Xu W (2010) Minimum emission dispatch constrained by stochastic wind power avail-
ability and cost. IEEE Trans Power Syst 25(3):1705–1713. https://doi.org/10.1109/TPWRS.
2010.2042085
19. Reddy SS, Bijwe PR, Abhyankar AR (2015) Joint energy and spinning reserve market clearing
incorporating wind power and load forecast uncertainties. IEEE Syst J 9(1):152–164. https://
doi.org/10.1109/JSYST.2013.2272236
20. Reddy SS, Bijwe PR, Abhyankar AR (2015) Optimal posturing in day-ahead market clearing
for uncertainties considering anticipated real-time adjustment costs. IEEE Syst J 9(1):177–190.
https://doi.org/10.1109/JSYST.2013.2265664
21. Vinod KR, Singh SK (2018) Solar photovoltaic modeling and simulation: as a renewable energy
solution. Energy Rep 4:701–712. https://doi.org/10.1016/j.egyr.2018.09.008
22. Mazidi M, Zakariazadeh A, Jadid S, Siano P (2014) Integrated scheduling of renewable gener-
ation and demand response programs in a micro-grid. Energy Convers Manage 86:1118–1127.
https://doi.org/10.1016/j.enconman.2014.06.078
23. Chauhan A, Saini RP (2014) A review on integrated renewable energy system based power
generation for stand-alone applications: configurations, storage options, sizing methodologies
and control. Renew Sustain Energy Rev 38:99–120. https://doi.org/10.1016/j.rser.2014.05.079
24. Rajanna S, Saini RP (2016) Modeling of integrated renewable energy system for electrification
of a remote area in India. Renew Energy 90:175–187. https://doi.org/10.1016/j.renene.2015.
12.067
25. Reddy SS, Momoh JA (2015) Realistic and transparent optimum scheduling strategy for hybrid
power system. IEEE Trans Smart Grid 6(6):3114–3125. https://doi.org/10.1109/TSG.2015.240
6879
Modeling of Various Renewable Energy Resources for Smart … 47

26. Atwa YM, El-Saadany EF, Salama MMA, Seethapathy R (2010) Optimal renewable resources
mix for distribution system energy loss minimization. IEEE Trans Power Syst 25(1):360–370.
https://doi.org/10.1109/TPWRS.2009.2030276
27. Ganesan S, Subramaniam U, Ghodke AA, Elavarasan RM, Raju K, Bhaskar MS (2020) Inves-
tigation on sizing of voltage source for a battery energy storage system in micro-grid with
renewable energy sources. IEEE Access 8:188861–188874. https://doi.org/10.1109/ACCESS.
2020.3030729
28. Rancilio G, Lucas A, Kotsakis E, Fulli G, Merlo M, Delfanti M, Masera M (2019) Modeling
a large-scale battery energy storage system for power grid application analysis. Energies 12.
https://doi.org/10.3390/en12173312
29. Bera A, Tian Y, Almasabi S, Mitra J, Borges C (2019) Modeling of battery energy storage
systems for system reliability studies. In: IEEE power and energy society general meeting,
Atlanta, GA, USA, pp 1–5. https://doi.org/10.1109/PESGM40551.2019.8973773
30. Sarasua AE, Molina MG, Mercado PE (2013) Dynamic modelling of advanced battery energy
storage system for grid-tied AC micro-grid applications. In: Zobaa AF (ed) Energy storage—
technologies and applications, IntechOpen. https://doi.org/10.5772/52219
31. Salkuti SR (2020) Optimal day-ahead renewable power generation scheduling of hybrid elec-
trical power system. In: Ray P, Biswal M (eds) Micro-grid: operation, control, monitoring and
protection, Lecture notes in electrical engineering, vol 625. Springer, Singapore. https://doi.
org/10.1007/978-981-15-1781-5_2
32. Dubarry M, Baure G, Fernández CP, Yu TF, Widanage WD, Marco J (2019) Battery energy
storage system modeling: a combined comprehensive approach. J Energy Storage 21:172–185.
https://doi.org/10.1016/j.est.2018.11.012
33. Gil-González W, Montoya OD, Garces A (2020) Modeling and control of a small hydro-power
plant for a DC micro-grid. Electr Power Syst Res 180. https://doi.org/10.1016/j.epsr.2019.
106104
34. Kusakana K (2016) Optimal scheduling for distributed hybrid system with pumped hydro
storage. Energy Convers Manage 111:253–260. https://doi.org/10.1016/j.enconman.2015.
12.081
35. Gupta A, Saini RP, Sharma MP (2010) Steady-state modelling of hybrid energy system for
off grid electrification of cluster of villages. Renew Energy 35(2):520–535. https://doi.org/10.
1016/j.renene.2009.06.014
36. Blank L, Rioseco EM, Caiazzo A, Wilbrandt U (2020) Modeling, simulation, and optimization
of geothermal energy production from hot sedimentary aquifers. Comput Geosci. https://doi.
org/10.1007/s10596-020-09989-8
37. Deo M, Roehner R, Allis R, Moore J (2014) Modeling of geothermal energy production from
stratigraphic reservoirs in the great basin. Geothermics 51:38–45. https://doi.org/10.1016/j.geo
thermics.2013.10.011
38. Sanchez E, Torres CF, Guillen P, Larrazabal G (2013) Modeling and simulation of the
production process of electrical energy in a geothermal power plant. Math Comput Model
57(9–10):2140–2148. https://doi.org/10.1016/j.mcm.2011.03.021
39. Baruah D, Baruah DC (2014) Modeling of biomass gasification: a review. Renew Sustain
Energy Rev 39:806–815. https://doi.org/10.1016/j.rser.2014.07.129
40. Upadhyay S, Sharma MP (2014) A review on configurations, control and sizing methodologies
of hybrid energy systems. Renew Sustain Energy Rev 38:47–63. https://doi.org/10.1016/j.rser.
2014.05.057
Smart Grid Communication: Recent
Trends and Challenges

Ishan Srivastava, Sunil Bhat, and Arvind R. Singh

Abstract The complexity of power distribution network is increasing day by day

with rapid increase in power demand. For monitoring and control of this complex
power network, application of smart grid technologies is essential. These technolo-
gies can also improve the overall system reliability and system security. A two-way
communication network is the main requirement for optimal operation of a smart
grid. In modern grid, different sensors and Intelligent Electronic Devices (IED) are
installed at various locations. All these devices share different information related
to the state of the network like voltage, current, power flow etc., through a two-
way communication channel. Maximum utilization of the network communication
system resources like sensors, IEDs and smart meters can ensure most economi-
cal operation of the grid. However, this communication channel can be subjected
to various challenges like cyber attacks, interference and data loss which can be a
major hurdle in deployment gird automation strategies. In this chapter, the evolution
of communication technology from wired network to wireless network along with
architecture of the smart grid is discussed; various standards and protocols related
to smart grid communication are presented. The authors also explained different
challenges and their possible solutions for two-way communication infrastructure of
smart grid.

Keywords Smart grid · Wireless communication · Smart gird standards ·

Cyber-security · Compressive sensing

I. Srivastava · S. Bhat
Electrical Engineering Department, Visvesvaraya National Institute of Technology,
Nagpur 440010, India
e-mail: [email protected]
S. Bhat
e-mail: [email protected]
A. R. Singh (B)
School of Electrical Engineering, Shandong University, Jinan, P.R. China

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 49
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6_3
50 I. Srivastava et al.

Nomenclature

SG Smart Grid
RES Renewable Energy Sources
IED Intelligent Electronic Device
AMI Advanced Metering Infrastructure
AMR Automatic Meter Reading
NIST National Institute of Standards and Technology
DG Distributed Generation
ISO Independent System Operator
RTO Regional Transmission Operators
SCR Silicon Controlled Rectifier
IoT Internet-of-Things
LAN Local Area Network
NAN Neighborhood Area Network
QoS Quality of Service
HTC Human Type Communication
DMS Distribution Management System
PLC Power Line Communication
WMN Wireless Mesh Network
TVE Total Vector Error
FE Frequency Error
X Synchrophasor value
X̂ Reconstructed value
f Frequency of original signal
fˆ Frequency of reconstructed signal
FDIR Fault Detection, Identification, Isolation and Power Restoration

1 Introduction

The energy demand is increasing across the globe with the advancement of new
technologies. In the modern era, most of this energy is supplied in the form of
electricity. Due to the rise in number of customers and the loads, the modern grid’s
complexity is also increasing. The concept of Smart Grid (SG) is introduced to
manage this highly complex power network efficiently. Advancement in the field of
material science, communication engineering and power electronics is the basis of
the emergence of SG. The modern grid must be designed, expanded and operated to
meet the following demands:
1. The energy transmission and distribution must be economical.
2. The energy cost for the consumers must be less.
3. System security must be maintained.
4. Service restoration must be fast in the case of a fault event.
Smart Grid Communication: Recent Trends and Challenges 51

5. The environmental impacts must be appropriately monitored, and more number

of Renewable Energy Sources (RES) must be connected in the grid.
6. Better management and integration of RES.
The key feature of a modern grid is the presence of RES. The presence of RES
can help in providing clean energy and also it can effective in supplying different
loads during peak demand. Conventionally, the energy flows from generating station
to consumers in a vertical power structure. But, due to the presence of RES in the
modern gird, the power can flow in the reverse direction as well, i.e. the power can
be generated at customer end and can be fed to the grid. This reverse power flow can
further increase the complexity in the operation of modern gird. Therefore, a two-way
communication channel is used in the SG for proper monitoring and control.
SG is equipped with a large number of different sensors at various grid nodes
compared to a limited number of sensors in the conventional grid. The overall infras-
tructure of an SG consists of smart power generation, transmission and distribution
elements. Additionally, smart information system, including smart meter, sensors,
and phasor measurement units, is also present. Data exchange among various com-
ponents of the grid is of high importance for the grid’s smooth operation. This data
helps in monitoring, analysis, optimization and control of various parameters of the
power grid. This data exchange is possible through inter-operability of measuring,
monitoring and controlling devices present in the grid. For example, the smart meter
component present in the customer domain must be able to communicate with any
component present in the generation domain, market domain or service domain. This
kind of information exchange needs proper protocols and standards. Furthermore,
this kind of information transfer must be done on a secured server as it is prone to
cyber-attacks. To make the gird smarter, more number of sensors, intelligent elec-
tronic devices (IEDs) or PMUs are getting connected to the grid on a regular basis.
But with more number of such devices, the size of real-time data generated is also
increasing. Therefore, SG must be equipped with resources for data management and
storage. To manage this large data, different data pre-processing, data cleaning, data
analysis and data prediction agents must be incorporated in the smart information
system domain of the SG. This chapter’s subsequent sections present SG architec-
ture’s concept, standards related to SG, wired and wireless mode of communication
in SG and cyber-security aspects associated with SG.
The main differences in the conventional gird and the SG is reported in [1–3].
The existing or conventional grid is having a limited number of sensors compared to
an SG infrastructure. These sensors are responsible for providing sufficient data for
real-time monitoring and control of the grid. Since there is the provision of reverse
power flow in the modern grid, the communication network for the SG must be
designed for two way communication. This two-way communication is also essen-
tial for implementing Advanced Metering Infrastructure (AMI) in SG. The metering
system is electro-mechanical in the conventional power grid; this arrangement can’t
be used for real-time data transfer. Consequently, for the deployment of automation
strategies like Automatic Meter Reading (AMR) for AMI, the digital metering sys-
tem is an essential SG requirement. These automation strategies are responsible for
52 I. Srivastava et al.

collection of data from smart meters, and secure data transmission through the com-
munication network. This data can be utilized for billing, troubleshooting issues in
the grid and analyzing various patterns. Another important duty of SG components
is to restore the power to customers in case of power interruption. The service must
be restored to the customers in a timely manner. The amount of time taken by the
utility for restoring the power supply is directly affecting the reliability of the sys-
tem. In the conventional grid, the process of fault detection, identification, isolation
and power restoration (FDIR) is done manually by the utility. As a result, the time
requirement for FDIR is more. In SG, with the concept of distribution automation
and self-healing network, the FDIR time can be drastically reduced, thus improving
the system’s overall reliability.

2 Architecture of SG

According to the National Institute of Standards and Technology (NIST) [4], an SG

architecture is the model that describes different domains or entities present in the
system and various interactions within the system. This architecture covers different
design aspects of the system along with the protocols and standards, defined for the
proper operation of the grid. According to NIST, the SG consists of seven major
domains:
1. Bulk generation
2. Transmission
3. Distribution
4. Customers
5. Markets
6. Service provider
7. Operations.

2.1 Bulk Generation

The significant part of energy demand is fulfilled by non-renewable sources of energy

like coal and nuclear. This approach is mainly adopted in the power sector because the
related technology is matured, the system’s reliability is high, and the management
of power flow is smooth. On the other hand, the emission of greenhouse gases that
can be very harmful to the environment is the major drawback of this approach.
Additionally, the availability of non-renewable resources is limited, and the global
increase in energy demand will create an imbalance in supply and demand, in the
near future. Therefore, the use of RES becomes essential in the current scenario.
The fundamental difference in terms of generation in SG environment compared to
conventional gird is the use of these RES. As the gird is becoming more complex with
Smart Grid Communication: Recent Trends and Challenges 53

more customers, there is a need to generate electricity on a small scale at different

locations in the gird. This concept is known as Distributed Generation (DG). These
DGs and energy storage elements like battery are collectively known as DERs. It
is desired that most of these DGs present in the SG must be of renewable nature.
The power generation capacity of these DGs can range from 3 to 10,000 kW [3].
In SG having DG penetration, with proper monitoring and control, the reliability
of the system can also be achieved equivalent to the reliability of the conventional
gird [5]. The burden from the main grid can be reduced by using DGs in the power
network. The load demands or peak demands can be fulfilled with the help of DGs in
the system, and this approach is more economical compared to traditional grid. This
kind of cost-effective operation is the main reason for reduced transmission losses
in the SG, equipped with DG.
There are a lot of challenges associated with the use of DER in SG. The main chal-
lenge is to design a proper control system for the microgrid, as there are fluctuations
at the DG’s input side. Also, with increased DG penetration, the inertia of the main
gird can be reduced, which creates stability issues for the gird. Therefore, SG must be
equipped with components for monitoring and controlling. Power electronics plays
a crucial role in the integration and controlling of DGs.
The major components of bulk generation domain in SG architecture are:
• Control: It is responsible for overall control over the generation which is based on
the data available through WAMS or SCADA system. Load generation balance is
the crucial aspect for the stability of the gird. Bulk generation domain is responsible
for maintaining this balance, and therefore, power generation control becomes very
important.
• Measurement: All the data, which can be relevant for power generation is collected
through the SCADA system.
• Protection: This component ensures protection of all the equipment incorporated
for a bulk generation.
• Record: The various records related to system status are stored for analysis and
forecasting purposes.
• Asset management: It deals with the management of equipment. It involves regular
maintenance, repair and replacement of faulty equipment.

2.2 Transmission Domain

The transfer of power from generating station to distribution level is done through
a transmission network. In the SG, through various substations, the power can be
transferred from the bulk generating station to the distribution network via the power
transmission system. The system operator’s main responsibility is to maintain the
grid’s stability by managing the load generation balance. Additionally, these Inde-
pendent System Operators (ISOs) or Regional Transmission Operators (RTOs) per-
form various tasks to ensure the smooth operation of the network [6]. Various tasks
54 I. Srivastava et al.

like planning, data acquisition and data analysis also come under top priorities for
smooth operation. With the increase in load demand, the transmission network can
become congested if proper planning and expansion strategy is not implemented. The
data acquired from smart devices like phasor measurements units, smart protection
relays, and power quality monitors becomes essential for monitoring and controlling
the modern electric grid. The authors in [6] described the following relevant features
related to the present smart transmission grid:
• Digitalization: Two-way communications throughout the grid requires digital sig-
nals for transfer of data. Therefore, the transmission substations must be equipped
with the devices which are compatible with digital communication. Additionally,
the devices present in the transmission domain must communicate with the devices
present in other domains. Therefore, a standard digital communication channel is
required for interoperability.
• Flexibility: The transmission domain must have the feature for up-gradation and
expansion, in order to meet the demands of dynamic load patterns, dynamic market
structure, and increased DER penetration in the grid. Also, decentralized control
schemes must be incorporated for better grid control.
• Intelligence: The system must be equipped with smart technologies capable of
doing predictions and making decisions which is necessary in an automated envi-
ronment. For example, in case of service restoration, IEDs can play a crucial role
in automatically restoring the power to the customers of the healthy section of
the network. This self-healing capability of the system is essential to maintain
reliability.
Smart transmission network must have smart substations which must possess some
basic characteristics. For example, each substation must be connected to all the
other substations and the control centres present in the gird. Additionally, these
substations must have enough resources to monitor and control the gird’s different
zone independently. These substations must process the historical and real-time data
quickly to predict the probable events, which helps prepare contingency plans for
problems like blackouts, instability etc. The data obtained from PMUs contains a
time-stamping; this allows the system operator to deploy necessary strategies to
control the grid on time.

2.3 Distribution Domain

In the conventional gird with the limited number of sensors, complete gird data is
not available. This will delay service restoration in the case of a fault condition or
in case of disturbances created by extreme events. The reliability of the system is
inferior for such scenario. Also, the increase of DG penetration in the distribution
side is making the monitoring and control procedure more complicated. Therefore,
to meet the demands of modern gird, the smart distribution system is necessary.
Smart Grid Communication: Recent Trends and Challenges 55

Smart meters, various sensors and IEDs are used to collect the data in the smart
distribution system. This data can help plan and deploy various strategies like outage
management, fault detection, isolation and service restoration procedure, demand-
side management etc. The combination of these schemes is known as the Distribution
Management System (DMS). Smart meters are installed at customer premises and
are responsible for collecting different types of data like the voltage, real power
flow, and reactive power flow at the customer node. This data can be transferred to
the electric utility in every 15 min. But, generally, this time interval depends upon
the various factors and applications. The data transfer is done using a Local Area
Network (LAN) from the customer node to some customer utility interface like Meter
Data Management System (MDMS) which uses Zigbee technology [7]. The data
collected from smart meters are used for billing, load modelling, load forecasting,
state estimation, reactive power compensation management.
The other major distribution automation strategy is service restoration. Distribu-
tion network is generally designed to operate in a radial manner. Therefore, whenever
a fault occurs in the system, some healthy part of the system gets disconnected. In
that situation, the utility is to take some step to re-route the power to the healthy
section of the network. The strategy mentioned above can be deployed by changing
the topology of the network, which can be done by operating the switches present
in the network. The use of remotely controlled switches instead of manually con-
trolled switches can reduce the overall restoration process time and thus improve
system’s reliability. Modern distribution network must be equipped with these power
electronic-based remotely controlled switches. The modern distribution network is
becoming complicated day-by-day, so decentralized control strategy is more suit-
able for better monitoring and control, which may further reduce the overall time of
operation for various distribution automation strategies.

2.4 Customer Domain

The customer in the power sector can be broadly classified into three categories,
home customers (energy range ≤ 20 kW), commercial customers (energy range
20–200 kW) and industrial customer (energy range ≥ 200 kW). The link which
connects these customers to the utility is called as Energy Service Interface (ESI).
This interface plays a critical role in customers, utility interaction process. Every
type of customer is provided with the metering facility and a gateway for the proper
transfer of data.
AMI is needed for two-way communication through ESI. LAN or Home Area
Network (HAN) can be used for communication between the customer domain and
ESI. Distribution network also contains micro-generation units which can be of the
renewable type or non-renewable type. Connection and disconnection of these units
is very import aspect in overall control of the grid. The customers who feed the
56 I. Srivastava et al.

power back to the grid must be adequately compensated. Therefore, smart meters
at customer premises must be required when there are micro-generation units like
rooftop PV present in the system.

2.5 Market Domain

The market domain provides a platform for buying and selling of energy or grid
assets. This domain’s main responsibility is to decide various electricity market
prices with proper judgment and planning. Furthermore, these prices must be known
to the customers, independent system operators and other concerned authorities, so
this information transfer is managed by market domain. The market-related details,
like pricing and other relevant information must also be communicated to all the other
domains for planning and management purposes. This exchange of information is
possible with interoperability among different domains. Interoperability is the ability
of communication among various devices present in all the domains and sub-domains
of the SG.
The other significant role of the market domain is to promote aggregators by
expanding their capabilities. The contribution of small utilities in the power sector is
of high relevance. Therefore, these small players must be provided with a transparent
platform for participation in the big electricity market. The market domain is also
responsible for making regulations for retailing and wholesaling of electrical energy.
In the future grid, it can be expected that the complete information will be readily
available to the customers like the price of electricity at different times of the day, the
tariff information corresponding to different players present in distribution domain
for power supply. It will help demand-side management and peak reduction, by
providing a more economical energy choice to the customers.

2.6 Service Provider Domain

Service provider domain has a wide range of responsibilities, ranging from tradi-
tional utility services like billing and customer account management to managing
the SG’s overall ecosystem. Service provider personnel are mainly responsible for
the installation and maintenance of various components at the customer end. Further-
more, home management, building management, automation, account management,
billing and customer management also comes under the umbrella of utility services.
Smart Grid Communication: Recent Trends and Challenges 57

2.7 Operation Domain

This domain is responsible for maintaining, monitoring and controlling all the other
domains of the grid so that all the operations run smoothly in the system. The main
duties of the operation domain include:
1. Maintenance and construction
2. Manage finances
3. Manage supply chain logistics
4. Records and assets management
5. Ensure security
6. Communication infrastructure management
7. Expansion planning
8. Smooth network operations which include smooth power flow in the network
and high reliability of the network.

3 Wired and Wireless Technology for Smart Grid

Communication

A common understanding is that a two-way communication channel is required for

information transfer in the SG. But, there is a difference in opinion on the kind of
technology which should be used for communication [8]. The main reason for this
debate is the variety of network types. SG infrastructure has enterprise bus, wide
area networks, field area networks and premises networks [3].
The communication network either wired or wireless, must possess certain quali-
ties like reliability, the network must have provision for fast data transfer with privacy
and security.

3.1 Wired Communication Technology

There are mainly two technologies used in wired communication system namely
fibre-optic communication and Power Line Communication (PLC).

3.1.1 Fiber-Optic Communication

The main features related to this kind of communication channel are high bandwidth,
immunity towards electromagnetic/radio interference, suitable dielectric character-
istic for high voltage application. The main disadvantage of deploying fibre-optic
communication is the cost of installation, which is very high relative to other forms
of communication methods [9]. Therefore, despite having the advantages mentioned
58 I. Srivastava et al.

above, this type of communication channel is used when high bandwidth is required
in the network. As the data size increases day by day with more sensors in the system,
this communication method is becoming relevant.

3.1.2 Power Line Communication

PLC technology uses the infrastructure which was mainly deployed for electrical
power transfer between two nodes. These lines deliver AC power (50 Hz or 60 Hz)
or DC power. Additionally, the communication signal is transmitted, which is having
a frequency range from a few hundred Hz to a few thousand MHz [10]. There are
various terminology used for PLC technology [10] which are listed in Table 1. The
main advantage of using PLC over other communication medium is that the infras-
tructure required for PLC is already present. For SG applications, the utilities can
control the information transfer directly [11]. PLC has “through the-grid” property
[12], which enables it to send information about lines, switches, loads etc. present
in the grid. Additionally, real-time diagnostics and monitoring can be performed by
utilities using the information collected through PLC. With the use of power elec-
tronics, it is possible to inject data signals or any other type signal into a power
distribution line which carries electrical power [13].

Table 1 Various terminology used for PLC technology

Terminology Description
Carrier-current systems Transmit carrier-modulated data signals
Can be used to transmit narrowband frequency
signal (500 kHz)
Power line carrier Transmit carrier modulated signal
Used for the systems with operating frequency
of less than 500 kHz
Distribution line carrier Technology used n distribution network
Relatively complicated to deploy compared to
transmission network
Used for system with frequency less than
500 kHz
Broadband over power line communication Used in distribution domain
(BPL)
Signal of frequency range from 2 to 30 MHz
can be transmitted
Data transfer rate up to hundreds of Mbps
Power line tele-communication (PLT) Similar advantages that of broadband over
power line communication
More popular in European countries
Smart Grid Communication: Recent Trends and Challenges 59

Some of the methods for PLC schemes in medium-voltage lines are as follows:
a. Zero-crossing shift method: In this technique, with the help of the power elec-
tronic circuit, a slight shift is created at the point of zero-crossing of the power
signal. This zero-crossing point is selected based on the time at which the addi-
tional signal needs to be injected. Consequently, after injection of communication
signal, there is a shift in the power signal. The injected signal can be detected
at the receiver end by comparing the duration of sending end and receiving end
waveform, as shown in Fig. 1. An example of the power electronics circuit is
shown in Fig. 2, which can create the required shift in the waveform of phase A
[13].
As shown in the circuit (Fig. 2), when there is no communication signal injec-
tion, the series arm remains inactive, and thus no shift is observed in phase
A waveform. When a signal is injected in the power waveform, the shunt arm
gets de-energized, and series arm injects signal in waveform corresponding to
phase A.

V
Original Power Signal
Communication Signal
Shifted Power Signal

O π π π π ωt

-1

Fig. 1 Waveform showing zero-crossing shift in power signal

A B C

Coupling Transformer
Modulation Transformer

Series
Arm
Shunt
Bypass

Arm
Arm

Fig. 2 Circuit for creating zero-crossing shift in power signal

60 I. Srivastava et al.

Fig. 3 Scheme to create

voltage sag at zero-crossing Substation
point for outbound signal
generation Bus
To Feeder

Communication
Signal Generator

b. Zero-crossing distortion technique: This technology is useful and simple for

inbound [14] and outbound [15] communication through PLC. The outbound
signal is created with the circuit shown in Fig. 3 [13]. In this method, the sub-
station bus is connected to a power electronics switch like SCR and whenever
an information signal needs to be injected, or perturbation needs to be done in
the voltage waveform, this power electronic switch is turned on in a controlled
manner near the zero-crossing point of the voltage waveform. Consequently,
voltage sag is created near the zero-crossing point of the voltage waveform. This
injection is done in only one of the two cycles of the voltage waveform. The
signal insertion is done in this manner because the original signal can be easily
retrieved by subtracting two consecutive voltage waveform cycles. Additionally,
this procedure ensures the absence of voltage harmonics or noise in the obtained
information signal.
Whenever a piece of information needs to be transferred from a point where this
power electronics circuit is present to the load points located in the downstream
region, this technique can be incorporated. Alternatively, for the applications
such as automatic meter reading where the information is transferred from load
end to the utility, a power electronic circuit shown in Fig. 4a can be used to
generate a current pulse near the zero-crossing point of the voltage waveform
as shown in Fig. 4b. This current pulse is created by switching the SCR in a
controlled manner; thus, a temporary short circuit is obtained near load end.
Smart Grid Communication: Recent Trends and Challenges 61

Fig. 4 a Power electronic circuit to create current pulse near zero crossing point of voltage wave-
form. b Resultant waveform

Due to this short-circuit, a current pulse is drawn from the power supply, which
the utilities can detect as an information signal.
c. Ripple signalling technique: A ripple signal is a high-frequency signal super-
imposed in the carrier signal of frequency 50 Hz or 60 Hz. This signalling
technique based on ripple control is commonly used in PLC. The ripple con-
trol can be carried out using electronic or power electronics-based circuit [13].
The ripple signal produces a distortion in the voltage waveform. By analyzing
this distortion, the original communication signal can be retrieved. Some of the
ripple based technique is proposed in [16–18].
d. Summary: For smart metering infrastructure or AMI, narrowband PLC can be
used [19]. For data transfer from SG controllers to a home network, broadband
PLC can be used [3]. For SG ecosystems, technologies like Internet-of-Things
(IoT) can be used along with PLC technology to improve the gird’s overall per-
formance. This kind of integration helps to deploy automation strategies more
smoothly [2]. Powerline channel modelling is challenging since the commu-
nication channel can be affected by external factors such as noise, emissions,
impedance alterations etc. [3]. Therefore, a lot more focus is required in channel
modelling for PLC. Some of the proposed research work in this area is reported
in [20–22].

3.2 Wireless Communication in Smart Grid

There are many advantages of using wireless technology over wired communication,
such as less installation cost, mobility, and fast deployment. One of the benefits of
using wireless technology is the remote end application [3]. In the earlier stages of
SG development, wireless technology is not considered for two way communication
62 I. Srivastava et al.

because of low data rates and limited bandwidth. But, for the recent scenario, with
more mature wireless communication technology, it can be used effectively for SG
application. Following wireless methods of communication can be used for two way
communication in SG:
a. Wireless Mesh Network (WMN): The critical components of WMN are mesh
routers and mesh clients. These routers and clients are the nodes for the commu-
nication network, and are connected through a wireless network. This network
can dynamically configure and organize the required ad-hoc networks responsi-
ble for connectivity with all other nodes [23]. Some of the reported benefits [3]
of using WMN technology are mentioned below:
I. Improved communication channel reliability: The WMN network has some
redundant paths for communication. These paths become useful when some
problems like path failure or node failure arise in the system. Consequently,
the reliability of the communication network will improve.
II. Aids automation in SG: As already mentioned, WMN mesh routers can self-
configure and self-organize the required ad-hoc network which connects all
the nodes. Therefore, the automation process needed for the SG ecosystem
becomes smooth with WMN.
III. High data transfer rate: Wireless broadband communication like WiMAX
based on IEEE:802.16 standards, can be used for transferring data at a high
rate and over long distances. The basic architecture of WMN is shown in
Fig. 5.
b. Cellular communication system: This technology is widely used in the telecom
sector. The main reason for using a cellular network is its mature technology
which is improving further with time. For SG paradigm, cellular technology can
be useful. A survey is presented in [24], which discusses the role of cellular
network in Neighborhood Area Network (NAN). In the case of the power distri-
bution network, an LTE-based communication standard can be beneficial. The
advantages of using LTE-based communication network are Quality of Service
(QoS), broad coverage, and low latency [24]. The LTE network is originally
designed for Human Type Communication (HTC) which is different from the
type of communication carried out in the SG paradigm. Also, there are limited
numbers of users connected to the base station in the case of HTC. On the other
hand, the number of nodes that generates and sends the data to the base station
is relatively large for SG application. Consequently, the LTE technology needs
modification for future SG [24]. Since the deployment of 3G (3rd generation)
cellular technology in 2001, the data transfer rate has significantly improved in
a cellular network. The data used in 3G technology is digital broadband packet
data which operates on the spectrum range of 824–894 MHz [25]. The data trans-
mission rate can go up to 2 Mbps. The 3G technology is based on WCDMA,
UMTS, CDMA 2000, while the subsequent 4G (4th generation) cellular tech-
nology uses MiMAX or LTE. The frequency range is from 850 to 2600 MHz,
and the maximum speed up to 100 Mbps is attainable. In recent advancement,
Smart Grid Communication: Recent Trends and Challenges 63

Wired Network

Wired Network
Internet

Wired
Clients

Mesh Router with Mesh Router with Sensors with

Mesh Router Gateway Bridge Wireless Data
Transfer
Capability

Cellular Network
Home/Industrial/ Commercial WiMAX Network
Clients with Wi-Fi Network

Fig. 5 Architecture of WMN network

5G (5th generation) cellular technology is ready for SG application deployment.

This 5G network has added advantage of low latency, less call dropout rates,
and speed up to 10 Gbps with a frequency range of 3–90 GHz [26]. The main
drawbacks of using the cellular network for SG application are the call drop out
the problem, which can affect the data transferred. Also, the monthly charges
for cellular service may be costly for data transfer on a regular basis [25]. WMN
and cellular networks are used for NAN or Field Area Network (FAN).
For Home Area Network (HAN) or Building Area Network (BAN) or Industrial
Area Network (IAN), the following wireless technologies can be used.
1. ZigBee: ZigBee Alliance designed this technology for radio-frequency appli-
cations. The main features of this technology are reliability, security and cost-
effectiveness. It is compatible with the IEEE 802.15.4 standard and is mostly
used in the customer domain for communication. Additionally, this technology
can be used for information transfer through smart meters [3]. ZigBee uses DSSS
modulation technique to operate in the frequency range of 868 MHz, 915 MHz
and 2.4 GHz [25]. The data rate through ZigBee technology is of range from
20–50 kbps.
2. Wireless Local Area Network (WLAN): This network follows the IEEE 802.11
standard. The main features of this technology are high data rate, which can be
64 I. Srivastava et al.

up to 150 Mbps, low latency, and a coverage range of 250 m [27]. Different
data rates are obtained according to the applications based on standards IEEE
802.11a, IEEE 802.11g and IEEE 802.11i. WLAN technology can be used for
substation automation applications that use IEC 61850 protocol for two-way
communication with respect to SG paradigm. WLAN technology in the distri-
bution network in the presence of DER can play a crucial role in monitoring
and controlling the grid, which is equipped with smart sensors and IEDs. The
main challenge in deploying WLAN technology for smart gird is the interference
problem with other communication networks and electromagnetic interference
in high voltage applications.
Other modes of communications:
a. Satellite communication: Satellite communication can be cost-effective in sce-
narios where the monitoring and control need to be done through remote sub-
station [3]. In [28], the authors showed a cost-effective communication method
for remote generation deployment application. This cost-effective solution can
be obtained by minimizing data transmission through satellite and maximizing
telemetry data [28].
Although satellite communication may provide economical solutions for some
scenarios, it can’t act as a primary data transmission network for the smart gird.
The primary reason for the same is the delay in data transmission when com-
pared to terrestrial communication. So, the more feasible prospect is to keep the
satellite communication channel as a backup network. In case of faults in the
terrestrial network, this backup network takes over, ensuring good reliability of
the communication system.
b. Microwave communication: This kind of communication technology is used in
SG as a backhaul network between power generation and power distribution con-
trol centres. For the SG application, a medium-haul to long haul communication
microwave band can be used, which has a frequency range of 3–15 GHz. For
point-to-point communication, microwave technology has the advantage of high
bandwidth, and the antenna size requirement is also feasible. This technology is
useful for communication in remote areas [3]. Microwave communication is a
line of sight communication technology which means obstacles in the path can
distort the communication signal.

4 Smart Grid Standards

Standards are written agreements which are prepared by experts of respective fields.
Standards contain technical specifications or other criteria that can be used to operate
a system or perform specific tasks. There are rules, guidelines and definitions asso-
ciated with each standard which are essential for better understanding and smooth
operation of the system. Standards ensure the delivery of any product or service in a
consistent and harmonized way.
Smart Grid Communication: Recent Trends and Challenges 65

Fig. 6 Major standards development organizations

In the SG paradigm, the communication elements, monitoring elements, protec-

tion elements and control elements must have some rules, guidelines and specifica-
tions for their smooth operation. The presence of DER in the smart gird also need
some standard aspects such as rules for its connection and disconnection, rules for
maintaining its characteristics, guidelines for power supply through DER, guidelines
for maintaining electrical constraints at the point of common coupling. Similarly,
the communication system, which is the backbone of the SG’s ecosystem must have
proper rules, guidelines and definitions for data transfer among various components
of the power system.
There are various Standards Development Organizations (SDOs) present, which
are responsible for drafting, maintaining, and updating the standards with experts’
help. Some major organizations are shown in Fig. 6.
In the SG paradigm, based on the field of interest, various standards have been
approved by SDOs which are shown in Fig. 7.
Various standards for a different type of communication network are listed in
Table 2.
Some of the standards related to smart gird are discussed below:
1. IEEE 1547: The criteria and requirements for interconnection of DER with the
electricity grid are documented in this standard. This standard also provides the
guidelines for interoperability and smooth operation, testing, safety, and main-
tenance requirements. IEEE 1547 has amended over the years since it was first
published in 2003. The IEEE 1547 series of standards provides information
about the test procedure (IEEE 1547.1), application guide (IEEE 1547.2), mon-
itoring, exchange of information and control (IEEE 1547.3), a guide for design,
operation and integration of DER (IEEE 1547.4), recommended practice for
interconnecting DER with secondary distribution network (IEEE 1547.6).
 66
I. Srivastava et al.

Fig. 7 Various approved standards for smart grid applications

Smart Grid Communication: Recent Trends and Challenges 67

Table 2 Standards related to different type of communications in smart grid

Network type Range Data rate Standards
Wireless Wired
communication communication
IEEE 802.1
IEEE 802.3
IEEE 802.11
HAN/BAN/IAN 1–100 m 1–100 kbps IEEE 1901
IEEE 802.15.4
IEEE 1901.2
IEEEP 1905.1
Smart meters – – IEEE SC31 (1377,1701,1703,P1704)
IEEE 802.11 IEEE 802.1
NAN/FAN 100 m–10 km 100 kbps– IEEE 802.15.4 IEEE 802.3
10 Mbps IEEE 802.16 IEEE 1901
IEEE 802.16d/e
IEEE 802.1
WAN 10–100 km 10 Mbps–1 Gbps IEEE 802.20
IEEE 802.3
IEEE 802.22
LAN IEEE 1815
Substation – –
IEC 61850

2. IEEE C37.118-2005: This standard defines the method for evaluation of PMU
data. The error in the synchrophasor data is calculated in terms of Total Vector
Error (TVE), and this standard provides the guidelines for maximum permissi-
ble TVE in steady-state scenarios. IEEE C37.118 is spitted into two parts; IEEE
C37.118.1 defines synchrophasor measurements under steady-state as well as
under dynamic conditions. The frequency measurement and rate of change of
frequency measurement techniques and guidelines are also defined in this stan-
dard. IEEE C37.118.2 defines the method for exchanging real-time synchronized
PMU data among various types of power system equipment.
3. IEEE 802 series of standards: IEEE 802 consists of a series of standards that
cover the system’s networking aspects. In SG, where wireless communication
is important for data exchange, these set of standards play a crucial role. IEEE
802 series covers specifications related to LAN and Metropolitan Area Network
(MAN). The protocols for the data-link layer and physical layer are defined in
IEEE 802 series of standards. Protocols for LAN (IEEE 802.1), Ethernet (IEEE
802.3), wireless PAN (IEEE 802.15), vehicular mobility (IEEE 802.20-2008),
wireless MAN (IEEE 802.16.1-2012), along with some other essential protocols
are defined in IEEE 802 series.
68 I. Srivastava et al.

5 Challenges in Smart Gird Communication

The major challenges for SG communication are as follows:

5.1 Active Control Challenges

Modern grid is getting more complex with time as more loads with different char-
acteristics are added to the network. DERs are deployed in large numbers across the
distribution grid to manage these loads and maintain the system’s reliability. The
DERs or the micro-girds need proper controlling as the two way power flow can be
dangerous if not handled properly. Intelligent controllers are required for manag-
ing the power flow through micro-grids. These controllers need to perform tasks in
full coordination with other controllers and de-centralized control stations present
in the network. It is a more practical approach to control the electricity network in a
decentralized manner compared to a centralized one. To demonstrate the suitability
of decentralized control, consider an example of fault event between bus A and bus B
in the system shown in Fig. 8. In case of the centralized control scheme, for the fault
event, to re-energize the load at bus B, any of the tie-switches T1 or T2 must be oper-
ated. The centralized control station will decide to operate the switch, and the overall
procedure from decision making to deployment can be very time-consuming. On the
other hand, with a decentralized control scheme, the intelligent controller associated
with the DER will make a quick decision to close the switch S1 and feed the load
with DER if sufficient power is available. The main drawback using a DG unit is the
variable output characteristics and complicated grid integration procedure.
From the communication perspective, the connectivity among these intelligent
controllers is essential for proper coordination [29]. Most of the controlling is
required for real-time decision making. Therefore, the data transfer rate and the
bandwidth must be optimum. The connectivity is also needed between intelligent
controllers and the sensors which are deployed in the network. The health of the sys-

Tie-Switch
T1
Generator A B DER
Fault

Switch
S1

Tie-Switch T2

Fig. 8 Test system to demonstrate effectiveness of de-centralized control scheme

Smart Grid Communication: Recent Trends and Challenges 69

Sensor
Two Way Generator
Communication
IC Intelligent Controller

IC

Substation
IC

Communication
Channel

IC

Generator IC

Fig. 9 Communication network for active controlling in distribution grid

tem, as well as the corrective action required in case of fault event, can be predicted
from the data collected through sensors. Therefore, more number of sensors with
reliable data transfer network is also needed for the SG. Figure 9 shows a typical
communication network for active controlling in the distribution grid.

5.2 Cyber-Security

Data exchange in SG paradigm is very frequent. The data generated by sensors,

smart meters and PMUs may contain critical information about the customer and the
grid. This data can be easily misused by hackers, resulting in the breach of customer
privacy and compromise grid security. There are four basic criteria for information
security in the SG [30].
1. Confidentiality: The customer data or the propriety information must remain
confidential. This means only authorized personnel can access this data. All the
information collected by the sensors or other measuring devices must be used
by only the utility in the grid. It must not be disclosed to any other unauthorized
70 I. Srivastava et al.

person or organization. The data exchange or the information transfer must be

done through the proper channel in a lawful manner. The critical data related to
the customer or the gird can be used for malicious purpose.
2. Availability: The data in the grid must be available in a timely manner since
this information exchange is necessary for most of the control operations in the
grid. If due to some reason if any critical information is missing, then smooth
operation of the grid can be obstructed. Therefore, the operator needs to ensure
a consistent flow of data in the grid’s communication network.
3. Integrity: The data obtained from the measurement devices must be transferred
entirely without any modifications. The data must be genuine and it must not
be altered at any stage by any unauthorized individual or entity. For instance,
in the conventional grid, electromechanical meters are used for registering the
amount of electricity consumed by a household or an industry. In this setup, it
is a common practice in developing countries to alter the reading of the energy
meters to pay a reduced amount of bill to the utility. This is a typical example
when the integrity of information is compromised. In SG paradigm, with smart
meters, the hackers are regularly trying to take control over the network server
so that they can manipulate data for their own personal gains. Also, this activity
is not limited to the customer domain but also other domains with vulnerabilities
which are prone to attacks and can lead to serious issues in the grid’s operation.
4. Accountability: The activities of authorized person responsible for collecting
or monitoring data exchange is recorded, so that any unauthorized exchange of
information can be easily tracked. This kind of practice in the SG is necessary
so that the person-in-charge dealing with critical data can be held accountable
for his/her unauthorized actions.
There are four steps in which any hacker tries to attack and control the SG [31]. The
first step is called reconnaissance, in which the attacker attempts to gather all the
information related to the target. In this step, the attackers will do traffic analysis and
try different authorized and unauthorized sources for collection of information about
the system [30]. The next step is the scanning; in this move, the attacker attempts to
find and analyze the system’s vulnerabilities. The attacker tries to collect information
related to IPs, ports and services of the system. The third step is the exploitation
stage. The attackers try to take control of the system using the vulnerabilities and
weaknesses of the system. The type of activities the attacker tries to do is to violate
integrity and privacy in the system, try to various channels, and try to inject viruses
and Trojan in the system [30]. The final step is to maintain the access of the system; the
attacker tries to gain permanent control of the system once the security is breached.

5.2.1 Countermeasures to Cyberattacks

In [30], the authors have proposed a three-step cyber security process.

1. Pre-attack: In this phase, the authorized persons and organizations take the appro-
priate step to upgrade the existing system’s security. It can be achieved by iden-
Smart Grid Communication: Recent Trends and Challenges 71

tifying the vulnerabilities of the system. Different components and resources of

the system prone to attacks are identified and recommended security measures
are implemented to further enhance safety. The pre-attack countermeasures are
divided into three parts:
I. Network security: For network security, firewalls can provide a primary layer
of protection, but this layer is relatively weak and can be easily bypassed
by many attackers. Therefore, the firewall that allows or denies the connec-
tion in a network must be associated with other security features. Intrusion
Detection System (IDS), Network Data Loss Prevention (DLP), Security
Information and Event Management Systems (SIEM) are some technolo-
gies which can be used with firewall [29].
II. Cryptography technique: These techniques can ensure the exchanged infor-
mation’s confidentiality and integrity by encryption methods. The encryp-
tion can be of a single-key type, where a single key is used to encrypt
and de-encrypt the information or it can be two-key encryption where sep-
arate private and public keys are used for encryption and de-encryption.
For power system applications, various key management frameworks are
proposed. For SCADA system, single-key, key establishment scheme, key
management architecture, and advanced key management architecture are
proposed.
III. Device security: The SG components are interconnected; therefore, any
weak link in the system may compromise the security of the whole sys-
tem. So, performing regular configuration checks and other security checks
becomes utmost important. An additional system can be created to regularly
check the devices and equipments for weakness regarding cyber-security to
perform such a task.
2. Under-attack: This step comprises two parts, (1) detect the attack: in this phase,
the concerned authority will try to detect the identity of the attacker, determine the
severity of the attack and identify part of the system or network where actually
the attack has been done. Part (2): deployment of countermeasure strategies
against the attack, which include pushback and reconfiguration methods. In the
pushback method, the network administrator will block all the incoming traffic
so that attacker loses all the unauthorized control. In the reconfiguration method,
the network topology is changed which is required to isolate the attacker. Also,
some anti-jamming methods can be employed as countermeasures.
3. Post-attack: In this step, the attacks are analyzed to protect the grid from any
malicious activity in future. An investigation is carried out to find the system’s
weakness, the virus and Trojan details used in the attack, cyber-security loop-
holes, the leaked information and the attacker details. Furthermore, based on the
collected information, the current security database is updated to prevent future
attacks.
72 I. Srivastava et al.

5.3 Data-Loss Problem and Mitigation Techniques

In the SG, the data can be lost or distorted during transmission due to various rea-
sons, such as radio interference, hardware or software malfunctioning, less band-
width availability for data transmission. This data loss problem can compromise the
integrity and availability aspects of information in the grid. Therefore, the data loss
problem must be resolved promptly. Some of the data loss mitigation techniques
involve improving the communication network’s reliability by performing regular
inspection of hardware and software involved in information exchange by increasing
the bandwidth of the wireless communication channel so that data congestion can
be avoided and some mathematical techniques can also be used for reconstruction
of lost signal in SG.
In [32], the authors proposed a compressive sensing based technique to reconstruct
the synchrophasor data obtained through the PMUs. The compressive sensing based
technique can also be used to compress the data without compromising its integrity.
Consequently, the bandwidth requirement also gets reduced if the data is appropri-
ately compressed without the loss of any information [33]. In the compressive sensing
based method, the signal from PMU or the synchrophasor data is made sparse by
applying discrete wavelet transform and thresholding techniques. The obtained sam-
ples are encoded at the sending end before their transfer over the communication
channel in the next step. When the samples are received at the receiving end, then
l1 optimization method followed by inverse wavelet transform is applied to get the
reconstructed signal.
To find the error in the reconstruction process, TVE and Frequency Error (FE)
are calculated. TVE is given as:


 (X (n) − X̃ (n))2 + (X (n) − X̃ (n))2
T V E(n) = 
r eal r eal img img
(1)
(X r eal (n))2 + (X img (n))2
 
 
FE =  f − fˆ (2)

where n = 1, 2, 3 … N, X and X̂ represents synchrophasor value and reconstructed

value respectively. f is the original frequency signal and fˆ is reconstructed frequency
signal.
The TVE and FE obtained by the compressive sensing method is less than 1%
[33], which is within limits prescribed in the IEEE C37.118-2005 standard.
Smart Grid Communication: Recent Trends and Challenges 73

6 Conclusion

The electricity grid is becoming more complex with time; the primary reasons for the
same are increased DG penetration, variable load characteristics and dynamic elec-
tricity market structure. The need for proper coordination among various components
of the grid is of utmost importance. Interoperability is the key for the SG’s smooth
operation, which requires reliable and secures two-way communication channel for
smooth information flow among the grid components. For better management, the
SG is divided into different domains: generation, transmission, distribution, market,
operations, service and customer domain. Each domain is guided by specific rules
and is assigned with different responsibilities. Various standards and protocols are
managing this SG system. These standards are essential for the proper operation of
the gird. For example, there are standards defined for running various grid activities
like connecting DERs in the grid, operating the grid’s communication system, main-
taining security aspects of the grid, etc. These standards ensure the proper delivery
of the service along with the economical operation of various components of the
gird. The communication channel is prone to multiple cyber-attacks. Therefore, var-
ious provisions are being made and updated in a timely manner to tackle security
problems.

References

1. Nafi NS, Ahmed K, Gregory MA, Datta M (2016) A survey of smart grid architectures, appli-
cations, benefits and standardization. J Netw Comput Appl 76:23–36. https://doi.org/10.1016/
j.jnca.2016.10.003
2. Kabalci E, Kabalci Y (eds) (2019) Smart grids and their communication systems. Springer, pp
18–19. https://doi.org/10.1007/978-981-13-1768-2
3. Fang X, Misra S, Xue G, Yang D (2011) Smart grid—the new and improved power grid: a survey.
IEEE Commun Surv Tutor 14(4):944–980. https://doi.org/10.1109/SURV.2011.101911.00087
4. Greer C, Wollman DA, Prochaska DE, Boynton PA, Mazer JA, Nguyen CT et al (2014) NIST
framework and roadmap for smart grid interoperability standards, release 3.0. https://doi.org/
10.6028/NIST.SP.1108r3
5. Agency IE (2002) Distributed generation in liberalised electricity markets. https://doi.org/10.
1787/9789264175976-en
6. Li F, Qiao W, Sun H, Wan H, Wang J, Xia Y, Zhang P (2010) Smart transmission grid: vision and
framework. IEEE Trans Smart Grid 1(2):168–177. https://doi.org/10.1109/TSG.2010.2053726
7. Jiang Y, Liu CC, Xu Y (2016) Smart distribution systems. Energies 9(4):297. https://doi.org/
10.3390/en9040297
8. Souryal M, Gentile C, Griffith D, Cypher D, Golmie N (2010) A methodology to evaluate
wireless technologies for the smart grid. In: 2010 first IEEE international conference on smart
grid communications, Oct 2010. IEEE, pp 356–361. https://doi.org/10.1109/SMARTGRID.
2010.5622067
9. Gore RN, Valsan SP (2018) Wireless communication technologies for smart grid (WAMS)
deployment. In: 2018 IEEE international conference on industrial technology (ICIT), Feb
2018. IEEE, pp 1326–1331. https://doi.org/10.1109/ICIT.2018.8352370
74 I. Srivastava et al.

10. Lampe L, Tonello AM, Swart TG (eds) (2016) Power line communications: principles,
standards and applications from multimedia to smart grid. Wiley. https://doi.org/10.1002/
9781118676684
11. Ahmed MO, Lampe L (2013) Power line communications for low-voltage power grid tomogra-
phy. IEEE Trans Commun 61(12):5163–5175. https://doi.org/10.1109/TCOMM.2013.111613.
130238
12. Galli S, Scaglione A, Wang Z (2011) For the grid and through the grid: the role of power
line communications in the smart grid. Proc IEEE 99(6):998–1027. https://doi.org/10.1109/
JPROC.2011.2109670
13. Xu W, Wang W (2010) Power electronic signaling technology—a new class of power elec-
tronics applications. IEEE Trans Smart Grid 1(3):332–339. https://doi.org/10.1109/TSG.2010.
2066293
14. Mak ST, Moore TG (1984) TWACSTM , a new viable two-way automatic communication system
for distribution networks. Part II: inbound communication. IEEE Trans Power Appar Syst
8:2141–2147. https://doi.org/10.1109/TPAS.1984.318524
15. Mak ST (1984) A new method of generating TWACS® type outbound signals for communica-
tion on power distribution networks. IEEE Trans Power Appar Syst 8:2134–2140. https://doi.
org/10.1109/TPAS.1984.318523
16. Foord PM (1989) U.S. Patent No. 4,868,539. U.S. Patent and Trademark Office, Washington,
DC. https://patents.google.com/patent/US4868539/en
17. Galloway JH, Berman AM (1980) U.S. Patent No. 4,215,394. U.S. Patent and Trademark
Office, Washington, DC. https://patents.google.com/patent/US4215394
18. Hunt PC (2000) Low frequency bilateral communication over distributed power lines. United
States Patent No. US6154488A. https://patents.google.com/patent/US6154488A/en
19. Rieken DW, Walker MR II (2011) Ultra low frequency power-line communications using
a resonator circuit. IEEE Trans Smart Grid 2(1):41–50. https://doi.org/10.1109/TSG.2010.
2099674
20. Konate C, Kosonen A, Ahola J, Machmoum M, Diouris JF (2008) Power line channel modelling
for industrial application. In: 2008 IEEE international symposium on power line communica-
tions and its applications, Apr 2008. IEEE, pp 76–81. https://doi.org/10.1109/ISPLC.2008.
4510402
21. Meng H, Chen S, Guan YL, Law CL, So PL, Gunawan E, Lie TT (2004) Modeling of transfer
characteristics for the broadband power line communication channel. IEEE Trans Power Deliv
19(3):1057–1064. https://doi.org/10.1109/TPWRD.2004.824430
22. Zimmermann M, Dostert K (2002) Analysis and modeling of impulsive noise in broad-band
powerline communications. IEEE Trans Electromagn Compat 44(1):249–258. https://doi.org/
10.1109/15.990732
23. Akyildiz IF, Wang X (2005) A survey on wireless mesh networks. IEEE Commun Mag
43(9):S23–S30. https://doi.org/10.1109/MCOM.2005.1509968
24. Kalalas C, Thrybom L, Alonso-Zarate J (2016) Cellular communications for smart grid
neighborhood area networks: a survey. IEEE Access 4:1469–1493. https://doi.org/10.1109/
ACCESS.2016.2551978
25. Parikh PP, Kanabar MG, Sidhu TS (2010) Opportunities and challenges of wireless commu-
nication technologies for smart grid applications. In: IEEE PES general meeting, July 2010.
IEEE, pp 1–7. https://doi.org/10.1109/PES.2010.5589988
26. Sofana RS, Dragičević T, Siano P, Prabaharan S (2019) Future generation 5G wireless net-
works for smart grid: a comprehensive review. Energies 12(11):2140. https://doi.org/10.3390/
en12112140
27. Mahmood A, Javaid N, Razzaq S (2015) A review of wireless communications for smart grid.
Renew Sustain Energy Rev 41:248–260. https://doi.org/10.1016/j.rser.2014.08.036
28. Deep UD, Petersen BR, Meng J (2009) A smart microcontroller-based iridium satellite-
communication architecture for a remote renewable energy source. IEEE Trans Power Deliv
24(4):1869–1875. https://doi.org/10.1109/TPWRD.2009.2028803
Smart Grid Communication: Recent Trends and Challenges 75

29. Bouhafs F, Mackay M, Merabti M (2014) Communication challenges and solutions in the smart
grid. Springer. https://doi.org/10.1007/978-1-4939-2184-3
30. El Mrabet Z, Kaabouch N, El Ghazi H, El Ghazi H (2018) Cyber-security in smart grid: survey
and challenges. Comput Electr Eng 67:469–482. https://doi.org/10.1016/j.compeleceng.2018.
01.015
31. Knapp ED, Samani R (2013) Applied cyber security and the smart grid: implementing secu-
rity controls into the modern power infrastructure. Newnes. https://doi.org/10.1016/C2012-0-
01113-1
32. Srivastava I, Bhat S, Singh AR (2020) Fault diagnosis, service restoration, and data loss miti-
gation through multi-agent system in a smart power distribution grid. Energy Sources Part A
Recov Util Environ Eff 1–26. https://doi.org/10.1080/15567036.2020.1817190
33. Thadikemalla VSG, Srivastava I, Bhat SS, Gandhi AS (2020) Data loss mitigation mechanism
using compressive sensing for smart grids. In: 2020 IEEE international conference on power
electronics, smart grid and renewable energy (PESGRE2020), Jan 2020. IEEE, pp 1–6. https://
doi.org/10.1109/PESGRE45664.2020.9070414
Comparison of Selected MPPT
Techniques Using Different Performance
Features

Salauddin Ansari and Om Hari Gupta

Abstract In the last few decades, the tendency and necessity to drift towards renew-
able sources of energy for power generation have captivated scientists moving to
photovoltaic (PV) systems. Since the characteristics of the PV generator is non-linear
and its output power changes with solar irradiance and atmospheric temperature, the
need for advanced maximum power point tracking (MPPT) technique has always
been fascinated. To withdraw the highest power from the PV generators, MPPT
is essential. So, to extract the highest power from the photovoltaic generator, even
in general situations, maximum power point tracking methods namely perturb and
observe, incremental conductance, and temperature control methods are simulated
in MATLAB/Simulink. Different parameters of solar energy conversion system—
obtained using different algorithms—are compared for different temperatures and
irradiances.

Keywords Maximum power point tracking · Perturb and observe · Incremental

conductance · Temperature control · Duty ratio

Nomenclature

PV Photovoltaic
MPPT Maximum Power Point Tracking
SECS Solar Energy Conversion System
PO Perturb and Observe
IC Incremental Conductance
TC Temperature Control
MSE Mean Square Error
FF Fill Factor
IR Solar Irradiance

S. Ansari · O. H. Gupta (B)

Department of Electrical Engineering, National Institute of Technology Jamshedpur, Jamshedpur,
Jharkhand, India

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 77
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6_4
78 S. Ansari and O. H. Gupta

T Temperature
d Duty ratio
MPP Maximum Power-Point
d Incremental duty ratio
V Ripple in PV Voltage
P Ripple in PV Output Power
tdpo Delay time for PO algorithm
tdic Delay time for IC algorithm
tdtc Delay time for TC algorithm
trpo Rise time for PO algorithm
tric Rise time for IC algorithm
trtc Rise time for TC algorithm
tppo Peak time for PO algorithm
tpic Peak time for IC algorithm
tptc Peak time for TC algorithm
Eppo MSE of PV output power for PO algorithm
Epic MSE of PV output power for IC algorithm
Eptc MSE of PV output power for TC algorithm
Evpo MSE of PV voltage for PO algorithm
Evic MSE of PV voltage for IC algorithm
Evtc MSE of PV voltage for TC algorithm
Eipo MSE of PV current for PO algorithm
Eiic MSE of PV current for IC algorithm
Eitc MSE of PV current for TC algorithm
FFpo Fill factor for PO algorithm
FFic Fill factor for IC algorithm
FFtc Fill factor for TC algorithm
ηipo , ηopo Input and Overall efficiency for PO algorithm
ηiic , ηoic Input and Overall efficiency for IC algorithm
ηitc , ηotc Input and Overall efficiency for TC algorithm

1 Introduction

One of the vast threats of the 21st era is the increasing risk of pollution and drop
of energy surety for the upcoming generation. However, since the need for energy
continues to increase day by day, we have to take measures and promote new technolo-
gies based on yielding energy from sources that are ample and do not endanger the
environment. Due to the fullness and sustainability of the sun, solar energy is taking
into consideration as one of the sustainable renewable sources at present and future
[1–3]. With the advancement in material science, photovoltaic (PV) energy system
efficiency continues increasing and the price constantly continues decreasing [4–7].
Comparison of Selected MPPT Techniques Using Different … 79

In addition to many benefits, such as easy installation, no pollution, and easy mainte-
nance, the life of these systems is also very long [8, 9]. The technical improvements
in the photovoltaic system have decreased the price and increased the solar energy
conversion system (SECS) efficiency. There is still a long way to reach the desired
level of energy generation. Some of the important studies to harvest a maximum
of available power out of the PV system i.e., MPPT algorithms are included in this
chapter.
The MPPT plays a very important role in solar power generation since it extracts
maximum power output from its PV system for different atmospheric situations and
thereby, maximizing the system’s efficiency. In this way, MPPT can reduce the total
cost of the entire system. The MPPT technique continuously searching the operating
point corresponding to maximum power output and keep it at that point. Several
MPPT techniques are introduced in the literature [10–12]. However, the perturb
and observe (PO) and incremental conductance (IC) are the two specific methods
and remain the most extensively used methods in commercial photovoltaic MPPT
systems. Apart from these two, a temperature control (TC)-based MPPT method
is also discussed in this chapter. The main emphasis of this study is the compar-
ison of the aforementioned MPPT techniques for SECS by considering various
features/parameters. These features/parameters are as follows:
• PV voltage and its ripple
• PV current and its ripple
• PV power and its ripple
• delay time, rise time, and peak time
• PV internal and overall efficiency
• mean square error (MSE)
• fill factor (FF).
The remaining chapter is formulated according to the following. Section 2 intro-
duced the different parameters/terms related to SECS. Section 3 presents the working
principle of SECS. Section 4 describes different MPPT techniques, Sect. 5 describes
heuristic approaches of the MPPT Technique and then, in Sect. 6, results and
discussions are included. Finally, the Conclusion is added to Sect. 7.

2 Different Parameters Related to SECS

• Fill factor: The fill factor (FF) is a figure of merit of PV cell. It tells how good
or how bad a PV cell is. Figure 1 depicts the concept of fill factor. The FF is
nothing but a ratio of maximum accessible power (V mpp × I mpp ) to the maximum
theoretical power (V oc × I sc ). Mathematically, it can be written as given in (1).
 
Vmpp · Impp
FF = (1)
(Voc · Ioc )
80 S. Ansari and O. H. Gupta

Fig. 1 Fill factor representation

• MSE: It is the mean of the squares of the errors—that is, the mean squared
changes between the predicted value and the true value. Mathematically, MSE
can be written as

1  
n
MSE = Xi − X̂ (2)
n i=1

where n is the no. of samples of data points

X is the vector of true value
X̂ is the vector of the predicted value.
• Delay time: The delay time is termed as the amount of time needed for the response
of a system to reach 50% of its final value within the first cycle of oscillation.
• Rise time: The rise time is termed as the amount of time needed for the response
of a system to reach from 10% of its final value to 90% of its final value within
the first cycle of oscillation.
• Peak time: The peak time is termed as the amount of time needed for the response
of a system to reach its first peak.
Different times i.e., delay time (t d ), rise (t r ) and peak time (t p ) are depicted in
Fig. 2.

Fig. 2 Delay time (t d ), rise time (t r ), and peak time (t p )

Comparison of Selected MPPT Techniques Using Different … 81

3 Working Principle of Solar Energy Conversion System

The Photovoltaic panels are made up of semiconductor material designed to transform

solar irradiation into usable electrical energy. Since the solar irradiance (IR) and
temperature (T) randomly changes throughout the day, the power output also changes.
Figure 3 explains the concept of a solar energy conversion system that represents the
block diagram of SECS which comprises of PV panel, MPPT controller, DC–DC
boost converter connected through DC load RL . When sun light falls on the PV panel,
it generates PV voltage V pv and current I pv which has nonlinear relation as illustrated
in Fig. 4. To withdraw the maximum available electrical power from the PV panel,
an MPPT controller is employed. MPPT controller senses the PV voltage “V sense ”
and current “I sense ” and next computes the optimum duty ratio (d) corresponding to
the MPP which modifies the input impedance of boost converter.
Finally, the boost converter is cascaded with DC load, RL to transfer power to
the load RL . The variations in PV current, power, and impedance with PV voltage
are illustrated in Fig. 4, where slopes S 1 and S 2 represent the load lines—one
corresponding to the fixed load impedance (S 1 ) and another corresponding to load
impedance modulated by the boost converter (S 2 ). The slope S 1 will be equal to
1/RL while slope S 2 will be 1/RL (1 − d)2 . If load “RL ” is directly coupled to the PV
panel it may extract the power P1 which is not the maximum available power. To the
withdraw maximum power from the PV panel, the DC–DC boost converter is intro-
duced between the PV panel and the load RL which modulates the load impedance to
RL (1 − d)2 . Alternatively stated, the converter is shifting the operating point along
the locus of maximum power point (MPP) and fixing the operating point at MPP.
Hence, the maximum power “Pmpp ” is extracted from the PV panel. From Fig. 4,
it is clear that if we have to shift the operating point at MPP, the slope of the line
should be decreased and the input impedance of converter Rin = RL (1 − d)2 should
be increased. To increase Rin , d must be decreased, until the MPP is reached. After
the measurement of voltage and current, the MPPT technique computes d so as to
transfer maximum power to the load. The converter chosen can also be a boost, buck,
or buck-boost converter, but in this chapter, a boost converter is employed, due to its
greater efficiency and higher range control [13].

Fig. 3 Equivalent diagram of a SECS with MPPT

82 S. Ansari and O. H. Gupta

Fig. 4 Intersection between

load line, current–voltage,
and power–voltage curve

4 MPPT Techniques

Three MPPT techniques are included for comparative analysis and discussed next.

4.1 Perturb and Observe

The perturb and observe (PO) [14–19] technique is the most frequently used and
discussed MPPT technique. In this method, initially, the PV output voltage V (k) and
current I(k) are captured, and then PV output power P(k) = V (k) × I(k) is calculated.
From calculated power and voltage, the change in PV output power “P” and voltage
“V ” are obtained as follows:

P = P(K ) − P(K − 1) (3)

V = V(K ) − V(K − 1) (4)

Comparison of Selected MPPT Techniques Using Different … 83

Now, if the changes in PV output power and voltage are zero i.e., P = 0 and
V = 0, the algorithm returns to the start position. If the change in PV output power
is negative i.e., P < 0, the change in voltage will be checked. Now, if V < 0,
the operating voltage is increased else it will be decreased. While if the change in
PV output power is not negative i.e., P ≥ 0, the change in the voltage (V ) is
checked, and if V < 0, the operating voltage is decreased else it will be increased.
After shifting the operating voltage, the algorithm returns to the start position. The
corresponding flow diagram is given in Fig. 5. The PO technique has two primary
drawbacks. Initially, the value of perturbation given to the system is crucial and
determines the amount of oscillations under steady-state at MPP. Besides, the speed
of convergence of output power towards MPP is slower. The greater the perturbations,
quickly the algorithm finds the MPP. However, greater perturbation results in a high
value of oscillation at MPP.
At the same time, if the value of perturbation is too small, the oscillation about
MPP will decrease, though the speed of convergence towards MPP will also decrease.
This means that there will be a compromise among the speed of convergence and
steady-state oscillations.
To resolve this drawback, the amount of perturbation may be decreased as it
approaches towards the MPP—the same is proposed in [20, 21]. When the output
power is far-off from the MPP, larger perturbations are used while smaller pertur-
bations are used when the output power is closed to MPP. Overall, the amount of
perturbations is selected based on the power–current curve slope. However, it results
in complexity and a costly technique. Further, whenever the operating point of the
system changes rapidly, the method can settle to an erroneous operating point. To
resolve such drawbacks, different methods have been proposed [22, 23]. The key
facts related to PO algorithm are:
(a) Two sensors are required, one, for voltage and the other for current.
(b) Slow rate of convergence to MPP.
(c) It fails to follow MPP if the environmental conditions vary quickly.
(d) Under steady-state, response will be oscillatory about the MPP.

4.2 Incremental Conductance

The IC [24–28] technique uses the slope of the power–voltage curve of the PV
generator to track MPP. The slope of the power–voltage curve (Fig. 4) of the PV
panel is zero at MPP. The slope will be positive if the PV output power is less than
MPP and negative if the PV output power is more than MPP. Mathematically, can
write the aforementioned discussion as given below in Eq. (5).
⎫
dP
dV
= 0; at M P P ⎬
dP
> 0; le f t o f M P P (5)
dV ⎭
dP
dV
< 0; right o f M P P
84 S. Ansari and O. H. Gupta

Fig. 5 Flow diagram of PO algorithm

For maximum output power, the PV output power differentiation with respect to
voltage should be zero, i.e.:

dP d(VI) dI ∼ I
= =I+V =I+V =0 (6)
dV dV dV V
Equation (6) can be modified as given in (7) below.
Comparison of Selected MPPT Techniques Using Different … 85

I ∼ −I
= (7)
V V
Now, Eq. (5) can be expressed as:

I −I
⎫
V
= V
; at M P P ⎬
I −I
> ; le f t o f M P P (8)
V
I
V
−I ⎭
V
< V
; right o f M P P

Thus, it is possible to detect whether the PV generator is working at MPP, left

of MPP, or right of MPP. So, as per Eq. (8), the MPP can be tracked by comparing
instantaneous conductance (I/V ) with the incremental conductance (I/V ). The
flow diagram of the IC method is given in Fig. 6. Initially, PV output voltage V (k)
and current I(k) are captured. Then the initial and old values of voltage and current
are updated. From the captured voltage and current, V is determined as given in
(4) and I is determined as given in Eq. (9) below.

 I = I(K ) − I(K − 1) (9)

Now, if the change in PV output voltage is zero i.e., V = 0, the change in

the current I is to be checked, and if I = 0, the algorithm returns to the start
position. If I > 0, the operating voltage is to be decreased else it is to be increased.
While if PV output voltage is not zero i.e., V = 0, the change in (I/V ) and
instantaneous conductance (I(k)/V (k)) are checked and if I/V = −I(k)/V (k), the
algorithm returns to the start position. If I/V > −I(k)/V (k), the operating voltage
is increased else it will be decreased. The speed of tracking the MPP depends upon the
amount of change in the magnitude of operating voltage. If increment or decrement
in operating voltage is large, the algorithm tracks the MPP faster but the system
cannot remain at MPP but oscillates around it. If the size of increment or decrement
in operating voltage is too small, the response becomes slower but the oscillations
close to MPP are lesser. So, again, there is a compromise between the speed of
convergence and oscillations. There are mainly two advantages of this algorithm,
first, it gives effective results even during the rapid change in atmospheric conditions.
Second, it has low oscillations about MPP as compared to the PO algorithm. The
main drawback of this technique is that it requires complex control circuitry. The
key facts related to this algorithm are:
(a) Two sensors are required—one for voltage another for current.
(b) More complicated than the PO algorithm.
(c) Fast dynamic tracking regardless the atmospheric conditions.
(d) Under steady-state, lesser oscillations about the MPP.
86 S. Ansari and O. H. Gupta

Fig. 6 Flow diagram of the IC algorithm

4.3 Temperature Control

The TC [29–31] technique is originated on the concept that the output voltage of
the PV module depends on the surface temperature of the PV panel. The relation is
described by Eq. (10) below [32].

Vmpp (T ) = Vmpp (To ) + αV mpp (T − To ) (10)


d = V − Vmpp (T ) Sd (11)
Comparison of Selected MPPT Techniques Using Different … 87

d(k) = d + d(k − 1) (12)

where

V mpp (T ) MPP voltage at the measured temperature

V mpp (T o ) MPP voltage at a measured temperature from the datasheet
α Vmpp Temperature coefficient from the datasheet
T Measured temperature
To Reference temperature from the datasheet.

From Eq. (10), it is obvious that MPP voltage V mpp relies on the measured temper-
ature T because V mpp (To ), α Vmpp , and To are the datasheet information. The flowchart
used to implement the TC technique is described in Fig. 7. First, the system starts
to sense the PV output voltage V and PV module surface temperature T. With the
help of measured temperature T, the optimized value of MPP voltage V mpp (T) is
calculated from Eq. (10). Also, the incremental duty ratio (d) is proportional to
the change in the actual PV module output voltage V and the desired MPP voltage
V mpp (T) according to Eq. (11). Where constant S d is the step size of d and the
value of S d decides the convergence rate of the technique. For a higher value of S d ,
the algorithm tracks the MPP faster and for a smaller value of S d , it tracks the MPP
at a slower rate. Finally, to ensure the PV output voltage to be as near as possible
to MPP voltage V mpp , d(k) is upgraded according to Eq. (12) which is the sum of
the d and the previous value of duty ratio d(k − 1). Now, d(k) and minimum duty
ratio (d min ) are compared and if the duty ratio is less than the minimum duty ratio
i.e., d(k) < d min , it is updated as d(k) = d min and then the algorithm returns to the
start position. While if the duty ratio is greater than the maximum limit i.e., d(k) >
d max , it is again limited to d max and then the algorithm returns to the start position.
Else if the duty ratio is within the upper and lower limits, the algorithm returns to
the start position.
Since the temperature on the PV panel changes gradually because of the thermal
inertia, the voltage of the PV terminal will also change smoothly. At the same time,
the PV output current is directly in proportion to the surface IR; therefore, its dynamic
is faster than the dynamics of temperature. Thus, it is obvious that the PV output
power includes two separate dynamics due to the slow change in temperature and the
fast fluctuations of the IR. However, when the TC technique is used, only temperature
measurements are taken into account. Therefore, fast dynamics is eliminated from
the technique and makes tracking more smooth, stable, and faster.
The key facts related to this algorithm are:
(a) Two sensors are required one for voltage other for temperature.
(b) It is constructed on the concept that the PV output voltage is directly related
to the temperature.
(c) Here, only the slow dynamic is taken into consideration and the fast dynamic is
eliminated and therefore, the tracking trend is more smooth, stable, and faster.
88 S. Ansari and O. H. Gupta

Fig. 7 Flow chart of TC algorithm

5 Implementation of Heuristic Approaches of the MPPT

Technique

Figure 8 shows the implementation of heuristic approaches of the MPPT Technique

which is composed of a PV panel, MPPT controller, Boost converter, load, and filter
connected to the utility grid. Initially, solar energy is converted to electrical energy
with the help of a solar panel. Then, output voltage and current of the solar panel
are measured—which are utilized by the MPPT controller to produce a duty cycle
for the Boost converter corresponding to its maximum power. Since the output of
the Boost converter is DC in nature so to convert it into AC, an inverter is used.
An inverter control block generates switching pulses for the inverter by utilizing the
output voltage and current of the inverter. Therefore, the SECS generate AC active
Comparison of Selected MPPT Techniques Using Different … 89

Fig. 8 Implementation of heuristics approaches of the technique

power (Pdg ) and reactive power (Qdg ). A local load is connected to the system then
finally, SECS is attached to the utility grid through a step-up transformer. A filter is
used to eliminate any lower order harmonics from the supply. If the load is less than
the generated power of SECS then surplus power Pug and Pug are supplied by
the utility grid.

6 Results and Discussion

For the verification of all three algorithms i.e., PO, IC, and TC, MATLAB/Simulink
simulations were executed considering a readily available PV module ‘Sun Earth
Solar Power TPB156*156-60-P225W’. The DC–DC boost converter is employed
for MPPT. The electrical specifications of the PV panel are presented in Table 1
and the PV Power, voltage, and current, obtained from the power–voltage curve,
are given in Table 2. To obtain a comparative analysis, two aspects are taken into
consideration i.e., changes in IR and temperature. The influence of change in IR
and temperature on different parameters such as voltage, current, power, ripples in
voltage, current, and power are recorded in the present analysis. Moreover, Mean
Squared Error (MSE), Fill factor (FF), delay time, rise time, peak time, internal
(or PV) efficiency, and overall efficiency are also obtained. Different results of the
mentioned cases are included next.

6.1 Voltage and Voltage Ripples in Varying Atmospheric

Conditions

Figure 9 represents the PV output voltages obtained for PO, IC, and TC algorithms
(i.e., V po , V ic , and V tc respectively) at a fixed temperature (25 °C) and varying IR. It
can be observed from Fig. 9 that as the IR rises from 400 to 1200 W/m2 in the steps of
200, PV output voltages by PO and IC algorithms (i.e., V po and V ic , respectively) are
decreased slightly while that of TC algorithm (V tc ) has increased. It is also found that
the TC algorithm gives slightly lesser PV output voltages for IR of 400–600 W/m2
while for 800–1200 W/m2 , all the three algorithms gave approximately the same
90 S. Ansari and O. H. Gupta

Table 1 Specifications of ‘Sun Earth Solar Power TPB156*156-60-P225W’ PV module

Array data
Parallel string 2
Series connected module per string 2
Module data
Maximum power Pmax 225.04 W
Voltage at MPP V mpp 29 V
Current at MPP I mpp 7.76 A
Voltage in open circuit condition V oc 36.5 V
Temperature coefficient of V oc −0.34/°C
Short circuit current I sc 8.26 A
Temperature coefficient of I sc 0.07/°C
Cells per module 60
Module parameters
Current due to light Ilt 8.2671 A
Saturation current through diode Io 1.3914 * 10−10 A
Diode ideality factor 0.95481
Resistance in shunt Rsh 476.1198 
Resistance in series Rse 0.41169 

Table 2 PV power, voltage and current at different temperature T and irradiance IR obtained from
PV curve
T (°C) IR (W/m2 ) Ppv (W) V pv (V) PV current (A)
25 400 367.3 58.9 6.236
25 600 550 58.78 9.357
25 800 727.9 58.54 12.43
25 1000 900.2 58 15.52
25 1200 1066 57.4 18.58
30 1000 881.6 56.67 15.56
40 1000 864.1 54.18 15.58
50 1000 806.3 51.69 15.6

PV output voltages. Figure 10 shows the PV output voltages for the variations in
temperature (30, 40, and 50 °C) with IR kept constant at 1000 W/m2 . It is found
that as the temperature increases, PV output voltages obtained from PO, IC and TC
algorithms decrease. It is also observed that the TC algorithm gives slightly less PV
output voltage as well as slightly fewer ripples in PV output voltage as compared to
PO and IC algorithms.
For better comparative analysis, Fig. 11 includes the comparison bars of PV
voltages obtained from different MPPT techniques and the PV voltage, corresponding
Comparison of Selected MPPT Techniques Using Different … 91

Fig. 9 PV output voltage with PO, IC, and TC algorithm when irradiance varied

to MPP i.e., V pv —obtained from the PV characteristicsfor a given temperature and

IR. It can be concluded that all the methods give quite accurate results. In some of the
cases, particularly at 25 °C and 1200 W/m2 , the voltages obtained from PO, IC, and
TC methods are equal and slightly more than V pv . Similarly, Fig. 12 represents the
ripple in PV voltage V at different temperatures and IR for different methods. It is
observed that minimum ripples are there in the case of the TC method, followed by
IC and PO methods. It is also observed that as the temperature and IR are increased,
the value of ripples in PV voltage decreases. For example, at 25 °C and 400 W/m2 ,
the ripples in PV voltage (V ) is about 7 V but at 40 °C and 1000 W/m2 , its value
is about 2 V.
92 S. Ansari and O. H. Gupta

Fig. 10 PV output voltage with PO, IC, and TC algorithm when temperature varied

Fig. 11 Comparison of PV output voltage with PO, IC, and TC algorithm when both temperature
and irradiance are varied
Comparison of Selected MPPT Techniques Using Different … 93

Fig. 12 Ripple in PV output voltage with PO, IC, and TC algorithm when both temperature and
irradiance are varied

6.2 Power and Power Ripples in Varying Atmospheric

Conditions

Figure 13 represents the PV output powers Ppo , Pic and Ptc obtained for PO, IC, and
TC algorithms, respectively at 25 °C and different IR (400–1200 W/m2 ). It can be
noticed from this figure that all three techniques successfully track the MPP and as
the IR increases, PV output power also increases. For example, for time 0.75 s to 1 s,
the PV output power obtained from PV characteristics (Ppv ) is 900.2 W—given in
Table 2—while PO, IC, and TC algorithms (i.e., Ppo , Pic , Ptc , respectively) provide
891.25 W, 890.95 W and 892 W of power, respectively. The obtained powers are
very close to Ppv . Initially, due to the presence of transients from 0 to 0.25 s, the
TC algorithm provides lesser PV output power as compared to that of IC and PO
algorithms. Figure 14 presents PV output powers for different temperatures (30, 40,
and 50 °C) with a fixed IR of 1000 W/m2 . It can be noted from the figure that, even
in transient-period i.e., from 0 to 0.25 s, power by TC algorithm is almost equal to
Ppv —providing the most accurate result in comparison to other methods. Moreover,
for better understanding, PV output powers obtained from different algorithms at
different temperatures and IR are compared magnitude wise in Fig. 15. Finally,
Fig. 16 represents the ripples in PV output powers for different algorithms under the
same environmental conditions. It can be noticed that ripples in output power P
are maximum at 25 °C and 800 W/m2 which is about 27 W while it is minimum at
50 °C and 1000 W/m2 which is about 2 W. That means upon increasing the IR and
temperature, ripples in PV output power (P) decreases.
94 S. Ansari and O. H. Gupta

Fig. 13 PV output power with PO, IC and TC algorithm when irradiance is varied

6.3 Current and Current Ripples in Varying Atmospheric

Conditions

Figure 17 shows the comparison of PV currents obtained from PO, IC, and TC
methods (I po , I ic , I tc , respectively) with ideal PV current obtained from character-
istic I pv for different temperatures and IR. It is found that all the three algorithms
work similarly and the currents are almost the same as obtained from PV character-
istic (I pv ). It is also noticed that as the IR increases, the PV output current goes on
increasing while it is almost constant if IR is kept constant and only the temperature
is varied. Further, Fig. 18 represents the ripples in the PV output currents for changed
IR and temperature. It can be noticed in this figure that ripples in PV output current
Comparison of Selected MPPT Techniques Using Different … 95

Fig. 14 PV output power with PO, IC and TC algorithm when the temperature is varied

Fig. 15 Comparison of PV output power with PO, IC and TC algorithm when both temperature
and irradiance are varied
96 S. Ansari and O. H. Gupta

Fig. 16 Ripple in PV output power with PO, IC and TC algorithm when both temperature and
irradiance are varied

Fig. 17 Comparison of PV output current with PO, IC, and TC algorithm when both temperature
and irradiance are varied

Fig. 18 Ripple in PV output current with PO, IC, and TC algorithm when both temperature and
irradiance are varied
Comparison of Selected MPPT Techniques Using Different … 97

by the TC algorithm are minimum as compared to that of PO and IC algorithms. The

ripples in PV output currents obtained from PO and IC algorithms are almost equal.

6.4 Tracking Performance and Efficiency

Figure 19 presents the delay times, rise times, and peak times for PO, IC, and TC
algorithms which give an idea about the tracking performances of the MPPT methods.
It can be observed from this figure that the delay time is minimum for the TC algorithm
(i.e., t dtc )—equal to 0.415 ms. The rise time is minimum for the IC algorithm (i.e.,
t ric )—equal to 2.465 ms whereas the peak time is minimum for PO method (i.e.,
t ppo )—equal to 2.94 ms. Figure 20 shows the mean squared errors (MSEs) for PV
output power, voltage, and current for PO, IC, and TC algorithms (considering 1000
samples for each case). The MSE of PV output powers for all methods (i.e., E ppo ,

Fig. 19 Delay time, rise time, and peak time with PO, IC, and TC algorithm

Fig. 20 MSE of PV output power, voltage, and current with PO, IC, and TC algorithm
98 S. Ansari and O. H. Gupta

E pic , and E ptc for PO, IC, and TC algorithms, respectively) are calculated and it is
found that, out of them, MSE is minimum for the PO algorithm.
Similarly, MSEs for the PV output voltages (i.e., E vpo , E vic , and E vtc for PO, IC,
and TC algorithms, respectively) are calculated and it is found that all the methods
have almost the same voltage MSE. Finally, MSEs for the PV output currents (i.e.,
E ipo , E iic , and E itc for PO, IC, and TC algorithms, respectively) are calculated and it
is observed that the value of MSE is minimum for the PO algorithm. Figure 21 repre-
sents the fill factors, FF po , FF ic , and FF tc —obtained for PO, IC, and TC algorithms,
respectively—at different temperatures and IR. It is observed from this figure that as
the IR and temperature are increased, fill factors obtained by all the three algorithms
decrease slightly. It is also clear from the figure that, out of the three algorithms,
the TC algorithm gives the highest fill factor. Figure 22 presents the internal and
overall efficiencies ïipo , ïopo , ïiic , ïoic , ïitc and ïotc for PO, IC and TC algorithms,
respectively. It is observed from this figure that both internal and overall efficiencies
for the IC algorithm have the highest values out of all algorithms. For the internal

Fig. 21 Fill Factor with PO, IC and TC algorithm when both temperature and irradiance are varied

Fig. 22 Internal and overall efficiency with PO, IC and TC algorithm

Comparison of Selected MPPT Techniques Using Different … 99

efficiency calculation, PV output power is considered while for the calculation of

overall efficiency, the load output power is considered. That means, for internal effi-
ciency calculation, output power equal to the product of V pv and I pv while for overall
efficiency calculation, the output power is equal to the product of V L and I L as given
in Fig. 3.

7 Conclusion

Three different MPPT techniques (i.e., PO, IC and TC) are tested on PV systems by
using MATLAB/Simulink offline simulations. The techniques have been compared
in terms of delay time, rise time, peak time, MSE, ripple content in voltage, current,
and power, internal efficiency, overall efficiency and fill factor for different IRs and
temperatures. It is found that all three MPPT techniques are successfully tracking
the MPP and obtained results are comparable. However, for a particular parameter,
one technique is found to be superior to another viz., TC method provides a better fill
factor, the PO method gives better MSE and the IC method provides better internal
and overall efficiencies.

References

1. Qazi A, Hussain F, Rahim NABD, Hardaker G, Alghazzawi D, Shaban K, Haruna K (2019)

Towards sustainable energy: a systematic review of renewable energy sources, technologies,
and public opinions. IEEE Access 7:63837–63851. https://doi.org/10.1109/ACCESS.2019.290
6402
2. Akhtar I, Kirmani S, Jamil M (2019) Analysis and design of a sustainable microgrid primarily
powered by renewable energy sources with dynamic performance improvement. IET Renew
Power Gener 13:1024–1036. https://doi.org/10.1049/iet-rpg.2018.5117
3. Koraki D, Strunz K (2018) Wind and solar power integration in electricity markets and distribu-
tion networks through service-centric virtual power plants. IEEE Trans Power Syst 33:473–485.
https://doi.org/10.1109/TPWRS.2017.2710481
4. Hegedus S, Luque A (2011) Achievements and challenges of solar electricity from photo-
voltaics. In: Handbook of photovoltaic science and engineering. Wiley, pp 1–38. https://doi.
org/10.1002/9780470974704.ch1
5. Branker K, Pathak MJM, Pearce JM (2011) A review of solar photovoltaic levelized cost of
electricity. https://doi.org/10.1016/j.rser.2011.07.104
6. Singh PP, Singh S (2010) Realistic generation cost of solar photovoltaic electricity. Renew
Energy 35:563–569. https://doi.org/10.1016/j.renene.2009.07.020
7. Yang CJ (2010) Reconsidering solar grid parity. Energy Policy 38:3270–3273. https://doi.org/
10.1016/j.enpol.2010.03.013
8. Ansari S, Gupta OH, Tripathy M (2021) An islanding detection methodology for SOFC-based
static DG using DWT. In: Lecture notes in electrical engineering. Springer Science and Business
Media Deutschland GmbH, pp 95–107. https://doi.org/10.1007/978-981-15-7994-3_9
9. Ansari S, Gupta OH (2021) Voltage ripple-based islanding technique on modified IEEE-13
bus test feeder for photovoltaic inverter. In: Lecture notes in electrical engineering. Springer
Science and Business Media Deutschland GmbH, pp 139–155. https://doi.org/10.1007/978-
981-15-7994-3_13
100 S. Ansari and O. H. Gupta

10. Logeswaran T, SenthilKumar A (2014) A review of maximum power point tracking algorithms
for photovoltaic systems under uniform and non-uniform irradiances. Energy Procedia 228–
235. https://doi.org/10.1016/j.egypro.2014.07.266
11. Soualmia A, Chenni R (2017) A survey of maximum peak power tracking techniques used in
photovoltaic power systems. In: FTC 2016—proceedings of future technologies conference.
Institute of Electrical and Electronics Engineers Inc., pp 430–443. https://doi.org/10.1109/FTC.
2016.7821645
12. Husain MA, Tariq A, Hameed S, Arif MSB, Jain A (2017) Comparative assessment of
maximum power point tracking procedures for photovoltaic systems. https://doi.org/10.1016/
j.gee.2016.11.001
13. Balasubrahmanyam CHS, Gupta OH (2020) Detailed study of solar energy conversion system
using boost converter—a new MPPT technique. J Inst Eng India Ser B 101:631–639. https://
doi.org/10.1007/s40031-020-00478-1
14. Wasynczuk O (1983) Dynamic behavior of a class of photovoltaic power systems. IEEE Power
Eng Rev PER-3:36–37. https://doi.org/10.1109/MPER.1983.5519293
15. Hua C, Lin JR (1996) DSP-based controller application in battery storage of photovoltaic
system. In: IECON proceedings (industrial electronics conference). IEEE, pp.1705–1710.
https://doi.org/10.1109/iecon.1996.570673
16. Hua C, Lin J, Shen C (1998) Implementation of a DSP-controlled photovoltaic system with
peak power tracking. IEEE Trans Ind Electron 45:99–107. https://doi.org/10.1109/41.661310
17. Abdelsalam AK, Massoud AM, Ahmed S, Enjeti PN (2011) High-performance adaptive perturb
and observe MPPT technique for photovoltaic-based microgrids. IEEE Trans Power Electron
26:1010–1021. https://doi.org/10.1109/TPEL.2011.2106221
18. Bianconi E, Calvente J, Giral R, Mamarelis E, Petrone G, Ramos-Paja CA, Spagnuolo G, Vitelli
M (2013) Perturb and observe MPPT algorithm with a current controller based on the sliding
mode. Int J Electr Power Energy Syst 44:346–356. https://doi.org/10.1016/j.ijepes.2012.07.046
19. Kumar N, Hussain I, Singh B, Panigrahi BK (2018) Framework of maximum power extraction
from solar PV panel using self predictive perturb and observe algorithm. IEEE Trans Sustain
Energy 9:895–903. https://doi.org/10.1109/TSTE.2017.2764266
20. Huynh P, Cho BH (1996) Design and analysis of a microprocessor-controlled peak-power-
tracking system. IEEE Trans Aerosp Electron Syst 32:182–190. https://doi.org/10.1109/7.
481260
21. John R, Mohammed SS, Zachariah R (2018) Variable step size perturb and observe MPPT
algorithm for standalone solar photovoltaic system. In: Proceedings of the 2017 IEEE inter-
national conference on intelligent techniques in control, optimization and signal processing,
INCOS 2017. Institute of Electrical and Electronics Engineers Inc., pp 1–6. https://doi.org/10.
1109/ITCOSP.2017.8303163
22. Yu GJ, Jung YS, Choi JY, Choy I, Song JH, Kim GS (2002) A novel two-mode MPPT control
algorithm based on comparative study of existing algorithms. In: Conference record of the
IEEE photovoltaic specialists conference, pp 1531–1534. https://doi.org/10.1109/pvsc.2002.
1190903
23. Hua C, Lin J (2003) An on-line MPPT algorithm for rapidly changing illuminations of solar
arrays. Renew Energy 28:1129–1142. https://doi.org/10.1016/S0960-1481(02)00214-8
24. Hussein KH, Muta I, Hoshino T, Osakada M (1995) Maximum photovoltaic power tracking:
an algorithm for rapidly changing atmospheric conditions. IEE Proc Gener Transm Distrib
142:59–64. https://doi.org/10.1049/ip-gtd:19951577
25. Liu F, Duan S, Liu F, Liu B, Kang Y (2008) A variable step size INC MPPT method for PV
systems. IEEE Trans Ind Electron 55:2622–2628. https://doi.org/10.1109/TIE.2008.920550
26. Mei Q, Shan M, Liu L, Guerrero JM (2011) A novel improved variable step-size incremental-
resistance MPPT method for PV systems. IEEE Trans Ind Electron 58:2427–2434. https://doi.
org/10.1109/TIE.2010.2064275
27. Zakzouk NE, Elsaharty MA, Abdelsalam AK, Helal AA, Williams BW (2016) Improved
performance low-cost incremental conductance PV MPPT technique. IET Renew Power Gener
10:561–574. https://doi.org/10.1049/iet-rpg.2015.0203
Comparison of Selected MPPT Techniques Using Different … 101

28. Alsumiri M (2019) Residual incremental conductance based nonparametric MPPT control for
solar photovoltaic energy conversion system. IEEE Access 7:87901–87906. https://doi.org/10.
1109/ACCESS.2019.2925687
29. Mutoh N, Matuo T, Okada K, Sakai M (2002) Prediction-data-based maximum-power-point-
tracking method for photovoltaic power generation systems. In: PESC record—IEEE annual
power electronics specialists conference, pp 1489–1494. https://doi.org/10.1109/psec.2002.
1022386
30. Park M, Yu IK (2004) A study on the optimal voltage for MPPT obtained by surface temperature
of solar cell. In: IECON proceedings (industrial electronics conference), pp 2040–2045. https://
doi.org/10.1109/iecon.2004.1432110
31. Swaraj K, Mohapatra A, Sahoo SS (2017) Combining PV MPPT algorithm based on tempera-
ture measurement with a PV cooling system. In: International conference on signal processing,
communication, power and embedded system, SCOPES 2016—proceedings. Institute of Elec-
trical and Electronics Engineers Inc., pp 482–489. https://doi.org/10.1109/SCOPES.2016.795
5876
32. Coelho RF, Concer FM, Martins DC (2010) A MPPT approach based on temperature measure-
ments applied in PV systems. In: 2010 IEEE international conference on sustainable energy
technologies, ICSET 2010. https://doi.org/10.1109/ICSET.2010.5684440
Short Term Active Power Load
Forecasting Using Machine Learning
with Feature Selection

Venkataramana Veeramsetty, D. Rakesh Chandra,

and Surender Reddy Salkuti

Abstract Active power load forecasting is an essential task for electric utilities in
order to operate network more reliable and also to participate successfully in energy
trading. In this chapter a new artificial neural network (ANN) topology is designed
to predict the load. The most efficient input features to predict the load have been
selected using correlation factors. The active power load data has been taken from
33/11 kV electric power distribution substation. The proposed ANN architecture is
designed, implemented and tested in Microsoft Azure Notebook environment. Based
on the simulation results, it has been observed that the proposed ANN model predicts
the load with good accuracy.

Keywords Active power · Artificial neural network · Feature selection · Load

forecasting · Machine learning

1 Introduction

In order to preserve the equilibrium of the load requirement And supply, electrical
power companies forecasts the load. Electrical load estimation is becoming one of
the core processes for periodic planning and expansion of power company facilities

V. Veeramsetty
Center for Artificial Intelligence and Deep Learning, SR Engineering College,
Warangal, Telangana, India
D. Rakesh Chandra
Department of Electrical and Electronics Engineering, KITS, Warangal, Telangana, India
e-mail: [email protected]
S. R. Salkuti (B)
Department of Railroad and Electrical Engineering, Woosong University, Daejeon,
Republic of Korea
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 103
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6_5
104 V. Veeramsetty et al.

[1]. Electric power load estimation is categorized into three categories based on time
horizons such as short-term, medium-term and long-term load forecasting [2, 3].
Short term power prediction is used to schedule the generation and transmission.
Time horizon for short term prediction is from few hours to days. Medium term
electric power prediction is used to purchase the fuel. Time horizon for medium term
prediction is from few weeks to months. Long term prediction is used to develop
the generation, transmission and distribution systems. Time horizon for long term
prediction is from one to ten years.
The estimate of short-term energy demand is a really an important problem for
proper management and delivery of electricity to avoid repeated power failures. It’s
just an essential prerequisite for the economic delivery of generation units in power
plants [4]. Strong control in the short term Load forecasting methodology must be
more precise in order to be more reliable It lets consumers pick a more efficient
method of energy use, minimizes production costs and optimizes output costs [5].
Machine learning (ML) has become an integral part in many fields like auto-
mobiles, international business, educational institution [6], civil engineering appli-
cations [7, 8], image processing [9] and also, electrical power system applications
[10]. ML analyses the system and makes forecasts based on previous existing data.
Many scholars have focused on active power Load prediction with various estima-
tion methods. Multiple linear regression model was used in [11] to predict the active
power load using hourly load demand and temperatures of South Sulawesi electrical
system. Active power load was estimated in [12] using Artificial neural networks and
multiple linear regression model. Active power load was predicted in [13] based on
load data of last four hours. Active power load was estimated in [10, 14, 15] based
load data for last four hours and load data at same hour for the last two days. Active
power load was predicted using support vector regression model in [16]. In this
methodology, the authors did not consider the load on weekends. A hybrid approach
based on decision tree and expert system was developed in [17] to predict the active
power load southeast of Brazil. Fourier-series forecast model with least-squares was
proposed in [18] to predict the hourly load data. A novel approach was developed
in [19] to forecast the active power load based on auto encoder and gated recurrent
unit (GRU). Auto encoder is used to compress the input features and GRU used to
forecast the load. An ANN model was developed in [20] to forecast the half-hourly
electric load demand in Tunisia. Authors have used Levenberg–Marquardt learning
algorithm to train the ANN model. Prediction of electricity demand and price one day
ahead using functional models was discussed in [21]. Estimation of electric power
consumption in Shanghai using grey forecasting model was discussed in [22]. All of
these approaches make useful contributions to load estimation, but lack certain essen-
tial functionality, such as dimensional reduction, to boost the efficiency of the model.
In this chapter active power load on a 33/11 kV substation was predicted using
artificial neural network (ANN). A new approach is used to predict the load at par-
ticular hour of the day based on load at previous two hours and at same hour but in
previous three days. Stochastic gradient descent optimizer [23] was used to update the
weight and bias parameters in the ANN. Dimensionality reduction using correlation
concept has been used to find more suitable input features among L(T−1), L(T−2),
Short Term Active Power Load Forecasting Using Machine … 105

L(T−24), L(T−48) and L(T−72). Lightweight ANN model is designed with more
suitable input features to predict the load. The proposed ANN topology with learning
algorithm is validated by comparing it with models developed in [13, 14].
The main contributions of this chapter are as follows:
• Complex and compact ANN models were developed to predict the active power
load.
• A new approach i.e. predict the active power load based on load at last two hours
and load at same time but in last three days was used.
• Dimensionality reduction with correlation concept was used to reduce the com-
plexity of the model so that over fitting problem was removed.
• Data analytic tools were used to process the data before feed it to the model.
The remaining of the chapter was organized as follows. Section 2 deliberates on
the ANN models to forecast the load on substation. Section 3 presents result analysis,
and conclusions of this chapter are described in Sect. 4.

2 Methodology

2.1 ANN Topology Design

The architecture of ANN that has been used in this chapter to predict the load L(T) at
particular hour (T) of the day based on load at last previous two hours [L(T−1) and

Fig. 1 Complex ANN architecture to predict load on 33/11 kV substation

106 V. Veeramsetty et al.

Table 1 ANN model parameters

Variable Parameters
Wi h 30
Who 6
bh 6
bo 1
Total parameters: 43
Hidden layer activation: softmax
Output layer activation: softmax

Fig. 2 Lightweight ANN architecture to predict load on 33/11 kV substation

L(T−2)] and load at same time but last three days [L(T−24), L(T−48), L(T−72)]
is shown in Fig. 1. Complete information of trainable parameters of the ANN model
are presented in Table 1.
The architecture of lightweight ANN that has been developed based on most
important features that have been selected using correlation concept is shown in
Fig. 2. Complete information of trainable parameters of the light weight ANN
model are presented in Table 2. Based on correlation concept L(T−1), L(T−24)
and L(T−48) have been considered as most important features.
Short Term Active Power Load Forecasting Using Machine … 107

Table 2 Lightweight ANN model parameters

Variable Parameters
Wi h 21
Who 7
bh 7
bo 1
Total parameters: 36
Hidden layer activation: softmax
Output layer activation: softmax

2.2 Dimentionality Reduction—Correlation

Correlation between the any two input features has been computed using Eq. 1. If
correlation between any of two input features is greater than 0.7 then one of them will
be removed from dataset. In this chapter 0.7 is considered as a threshold correlation
value to identify most important features.

1
corr elation(L(T − 1), L(T − 2)) =
  n − 1
L(T −1) L(T −2) (L(T − 1) ∗ mean(L(T − 1)))(L(T − 2) ∗ mean(L(T − 2)))
Std(L(T − 1))Std(L(T − 2))
(1)

2.3 Stochastic Gradient Descent Optimizer

In this chapter back propagation based on stochastic gradient descent optimizer has
been used to train the ANN models. Back propagation Algorithm that has been used
to train the ANN based on stochastic gradient descent optimizer is presented in
Algorithm 1.

3 Result Analysis

Historical load data at each hour of the day for the period from September-2018 to
November-2018 was taken from 33/11 kV substation near Kakathiya University in
Warangal [24]. Date processing techniques for observing the data distribution and
outliers, and for data normalization have been used before using this data to train
and test the ANN models. The proposed ANN models have been implemented and
tested in cloud computing environment using Microsoft Azure Notebook [25].
108 V. Veeramsetty et al.

Algorithm 1 Back Propagation Algorithm Using Stochastic Gradient Descent Opti-

mizer
Inputs
1: Read Dataset and initialize weight matrices Wi h and Who , and bias parameters bh and bo
2: set η=0.01, epochs=100, batch size=1

Steps
1: Set iteration t=1
2: Set Sample i=1
3: while t ≤ iter max do
4: while i ≤ n s do
5: Compute net input (neth ) to hidden layer for input ’X[i]’ using Eq. 2.

neth = [Wi h ]T ∗ X [i] + bh (2)

6: Compute output (oh ) of hidden layer using Eq. 3.
1
oh = (3)
1 + e−neth
7: Compute net input (neto ) to output layer using Eq. 4.

neto = [Who ]T ∗ oh + bo (4)

8: Compute output (oo ) of output layer using Eq. 5.
1
oo = (5)
1 + e−neto
9: Update weights between hidden and output layer using Eq. 6 and bias using 7.

Who = Who + η ∗ (d − O) ∗ O ∗ (1 − O) ∗ Oh (6)

bo = bo + η ∗ (d − O) ∗ O ∗ (1 − O) (7)
10: Update weights between input and hidden layer using Eq. 8 and bias at hidden layer using
9.
Wi h = Wi h + η ∗ (d − O) ∗ O ∗ (1 − O) ∗ Who ∗ Oh ∗ (1 − Oh ) ∗ X [i] (8)

bh = bh + η ∗ (d − O) ∗ O ∗ (1 − O) ∗ Who ∗ Oh ∗ (1 − Oh ) (9)
11: i=i+1
12: end while
13: t=t+1
14: end while
15: Print final weights Wi h and Who
16: Compute error MSE and MAE for both training and testing dataset using Eqs. 10 and 11
respectively
1 
ns
MSE = (di − Oi )2 (10)
ns
i=1

1 
ns
M AE = |di − Oi | (11)
ns
i=1
Short Term Active Power Load Forecasting Using Machine … 109

Fig. 3 Box plot for active

power load dataset

3.1 Data Analysis

Active power load data is analyzed for normal distribution and also for outliers using
box plot as shown in Fig. 3. In the box plot seen in Fig. 3, Median shows that there is
no left or right skew in the data at the centre of the inter-quartile spectrum and also
indicates that the data meets the ideal normal distribution. Box plot was also used to
identify the outliers in the data set. Outlier is identified as data points with a value
less than the minimum represented by the left most vertical line and more than the
maximum represented by the right most vertical line in the box plot. The box plot
shown in Fig. 3 confirms that there is no outliers in the active power load dataset.
Complex ANN model performance The ANN model with five input neurons corre-
sponding to five input features L(T−1), L(T−2), L(T−24), L(T−48) and L(T−72)
and one output neuron corresponding to output variable L(T) is simulated with differ-
ent number of hidden layers and neurons. The performance of various architectures
are presented in Table 3 in terms of MSE and MAE. From Table 3, it has been
observed that ANN model with one hidden layer with 6 hidden neurons is perform-
ing well in terms of training MSE and MAE i.e. 0.0108 and 0.0736 and testing MSE
i.e. 0.0121. Finally the complex model with 5 input neurons, 1 output neuron and 6
hidden neurons with one hidden layer is considered for active power load prediction.
Comparison between actual load and load that predicted using complex ANN
having testing MSE 0.0121, training MSE 0.0108 and training MAE 0.0736 is pre-
sented in Fig. 4. From Fig. 4, it has been observed that load predicted with complex
ANN is almost follows the actual load.
110 V. Veeramsetty et al.

Table 3 Complex ANN model training and testing performance

Layers Stage Error Hidden neurons
6 7 8
1 Training MSE 0.0108 0.0113 0.0116
MAE 0.0736 0.0766 0.0784
Testing MSE 0.0121 0.0128 0.0257
2 Training MSE 0.0113 0.0113 0.0115
MAE 0.0768 0.0767 0.0775
Testing MSE 0.0129 0.0129 0.0134
3 Training MSE 0.0219 0.0112 0.0122
MAE 0.1185 0.0761 0.0812
Testing MSE 0.0230 0.0127 0.0128

9000

8000
Active power load in kW

7000

6000

5000

4000
Predicted load using complex ANN
Actual Load
3000
0 50 100 150 200 250 300 350 400 450

Testing Data Sample Number

Fig. 4 Comparison between actual and predicted load using complex ANN

3.2 Lightweight ANN Model Performance

The light ANN model with three input neurons corresponding to three most important
input features L(T−1), L(T−24) and L(T−72) and one output neuron corresponding
to output variable L(T) is simulated with different number of hidden layers and
neurons. The performance of various architectures are presented in Table 4 in terms
of MSE and MAE. From Table 4, it has been observed that ANN model with one
hidden layer with 7 hidden neurons is performing well in terms of training MSE and
MAE i.e. 0.0112 and 0.0758 and testing MSE i.e. 0.0128. Finally the lightweight
model with 3 input neurons, 1 output neuron and 7 hidden neurons with one hidden
layer is considered for active power load prediction.
Comparison between actual load and load that predicted using lightweight ANN
having testing MSE 0.0128, training MSE 0.0112 and training MAE 0.0758 is pre-
Short Term Active Power Load Forecasting Using Machine … 111

Table 4 Lightweight ANN model training and testing performance

Layers Stage Error Hidden neurons
5 6 7 8
1 Training MSE 0.0114 0.0113 0.0112 0.0113
MAE 0.0763 0.0762 0.0758 0.0758
Testing MSE 0.0130 0.0128 0.0128 0.0128
2 Training MSE 0.0114 0.0115 0.0115 0.0115
MAE 0.0767 0.0775 0.0769 0.0771
Testing MSE 0.0132 0.0132 0.0131 0.0131
3 Training MSE 0.0349 0.0367 0.0116 0.0262
MAE 0.1523 0.1562 0.0777 0.1311
Testing MSE 0.0394 0.0134 0.0132 0.02713

9000

8000
Active power load in kW

7000

6000

5000

4000

3000

2000

1000 Predicted load using Lightweight ANN model

Actual load
0
0 50 100 150 200 250 300 350 400

Testing data sample number

Fig. 5 Comparison between actual and predicted load using complex ANN

sented in Fig. 5. From Fig. 5, it has been observed that lightweight ANN model
predicts the load almost equal the actual load.
Converging characteristics representing the variation in both training and testing
MSE during the training of complex ANN model is shown in Fig. 6a and lightweight
ANN model is shown in Fig. 6b. From Fig. 6 it has been observed that as training is
progress, the weights are updated in such a way that the difference between actual
and predicted load is decreasing. Validation loss is high comparing to training loss
but not have much difference that shows model is well trained and does not have
under fitting problem.
112 V. Veeramsetty et al.

(a) Converging characteristics of complex (b) Converging characteristics of light weight

ANN model ANN model

Fig. 6 Converging characteristics

Table 5 Comparison between ANN models

Model Trainable Training Testing
parameters
MSE MAE MSE
Complex ANN 43 0.0108 0.0736 0.0121
Lightweight 36 0.0112 0.0758 0.0128
ANN
% of absolute 16.28 3.7 3.0 5.8
change

3.3 Comparative Analysis

Comparison of the complex ANN model and the lightweight ANN model is shown in
Table 5. The size of the lightweight ANN model is small with 36 parameters compared
to the complex ANN model with 43 parameters. Owing to the distortion of the
model with reduced dimensionality, the lightweight ANN model had good training
and testing losses compared to the complex ANN model. Even though “Lightweight
ANN model” was compressed by 16.28% in terms of training parameters, losses i.e.
MSE, MAE and testing MSE are just increased by 3.7%, 3.0% and 5.8% respectively.
The proposed ANN models were validated by comparison with the load forecast-
ing models available in literature such as [10, 13, 14] based on training and testing
of MSE as described in Table 6. For Table 6, it has been observed that the proposed
ANN models predict loads with good accuracy compared to the existing models.
Short Term Active Power Load Forecasting Using Machine … 113

Table 6 Comparisons in terms of testing MSE

Model MSE
Training Testing
[13] 0.29 1.59
[14] 0.23 0.44
[10] 0.2 0.32
Complex ANN 0.0108 0.0121
Lightweight 0.0112 0.0128

4 Conclusions

In this chapter short term load forecasting was implemented using ANN based on
load at last two hours and load at same time but previous three days. Feature selection
based on correlation concepts was utilized to reduce the complexity of the model by
reducing number of input features. Light weight ANN model was also predicted load
like a complex model with less number of trainable parameters with loss of small
amount of accuracy. Practical historical load to train and test the ANN models was
collected from 33/11 kV substation near Kakathiya University in Warangal city.
The proposed model can be helpful to utilities to trade energy in hourly ahead
markets successfully by estimation the load with good accuracy. The discussed short
term load forecasting problem can be further extended using sequence models like
RNN, LSTM and GRU, also by using various dimensionality reduction techniques
like PCA and FA.

Acknowledgements This research work was supported by “Woosong University’s Academic

Research Funding - (2021–2022)”.

Appendix

This section demonstrates the mathematical derivation for the expressions presented
in Eqs. 6, 7, 8 and 9. Also, explains how weights and bias parameters of ANN
model update using stochastic gradient descent optimizer with help of numerical
calculations.
Mathematical Derivation
Forward pass
During forward pass stage of back propagation algorithm ANN model will estimate
the output as shown in Eqs. 2, 3, 4 and 5.
114 V. Veeramsetty et al.

Backward pass
Backward stage of back propagation algorithm will help to update weights and bias
parameters using specified optimizer. In this chapter stochastic gradient descent opti-
mizer has been used to updated the model parameters.
Change in weights between hidden and output layer

∂0.5 ∗ M S E ∂0.5 ∗ (d − oo )2
ΔWho = −η ∗ = −η ∗ (12)
∂ Who ∂ Who

where MSE = (d − oo )2 and oo = 1

1+e−neto
and neto = Who
T
+ Oh + bo

∂0.5 ∗ M S E ∂ M S E ∂oo ∂neto

ΔWho = −η ∗ = −η ∗ 0.5 ∗ (13)
∂ Wi h ∂oo ∂neto ∂ Who
∂oo
where ∂ M SE
∂oo
= −2 ∗ (d − oo ) and ∂net o
= oo ∗ (1 − oo ) and ∂neto
∂ Who
= oh . Replace all
these mathematical derivatives in Eq. 13 then will get the mathematical expression
for ΔWho as shown in Eq. 14.

ΔWho = η ∗ (d − oo ) ∗ oo ∗ (1 − oo ) ∗ oh (14)

Change in bias parameters at output layer

∂0.5 ∗ M S E ∂0.5 ∗ (d − oo )2
Δbo = −η ∗ = −η ∗ (15)
∂bo ∂bo

∂0.5 ∗ M S E ∂ M S E ∂oo ∂neto

Δbo = −η ∗ = −η ∗ 0.5 ∗ (16)
∂bo ∂oo ∂neto ∂bo

where ∂net
∂bo
o
=1. Replace it in Eq. 16 then will get the mathematical expression for
Δbo as shown in Eq. 17.

Δbo = η ∗ (d − oo ) ∗ oo ∗ (1 − oo ) (17)

Change in weights between input and hidden layer

∂0.5 ∗ M S E ∂0.5 ∗ (d − oo )2
ΔWi h = −η ∗ = −η ∗ (18)
∂ Wi h ∂ Wi h

∂0.5 ∗ M S E ∂ M S E ∂neth
ΔWi h = −η ∗ = −η ∗ 0.5 ∗ (19)
∂ Wi h ∂neth ∂ Wi h

where ∂net h
∂ Wi h
= X and ∂∂net
MSE
h
= ∂M SE
∂oh
∂oh
* ∂net h
. Replace all these mathematical derivatives
in Eq. 19 then will get the mathematical expression as shown in Eq. 20.
Short Term Active Power Load Forecasting Using Machine … 115

∂MSE ∂oh
ΔWi h = η ∗ 0.5 ∗ X ∗ ∗ (20)
∂oh ∂neth
∂oh M S E ∂neto
where ∂net h
= oh ∗ (1 − oh ) and ∂ M SE
∂oh
= ∂∂net o
* ∂oh . Replace all these mathematical
derivatives in Eq. 20 then will get the mathematical expression as shown in Eq. 21.

∂ M S E ∂neto
ΔWi h = η ∗ 0.5 ∗ X ∗ oh ∗ (1 − oh ) ∗ ∗ (21)
∂neto ∂oh

where ∂net
∂oh
o
= Who and ∂∂net
MSE
o
= ∂M SE
∂oo
∂oo
* ∂net o
. Replace all these mathematical deriva-
tives in Eq. 21 then will get the mathematical expression as shown in Eq. 22.

∂MSE ∂oo
ΔWi h = η ∗ 0.5 ∗ X ∗ oh ∗ (1 − oh ) ∗ Who ∗ ∗ (22)
∂oo ∂neto
∂oo
where ∂ M SE
∂oo
= (−2 ∗ (d − oo ) and ∂net o
= oo ∗ (1 − oo ). Replace all these mathe-
matical derivatives in Eq. 22 then will get the mathematical expression for ΔWi h as
shown in Eq. 23.

ΔWi h = η ∗ X ∗ oh ∗ (1 − oh ) ∗ Who ∗ (d − oo ) ∗ oo ∗ (1 − oo ) (23)

Change in bias parameters at hidden layer

∂0.5 ∗ M S E ∂0.5 ∗ (d − oo )2
Δbh = −η ∗ = −η ∗ (24)
∂bh ∂bh

∂0.5 ∗ M S E ∂ M S E ∂neth
Δbh = −η ∗ = −η ∗ 0.5 ∗ (25)
∂bi h ∂neth ∂bh

where ∂net
∂bi h
h
= 1 and ∂∂net
MSE
h
= ∂M SE
∂oh
∂oh
* ∂net h
. Replace all these mathematical derivatives
in Eq. 25 then will get the mathematical expression as shown in Eq. 26.

∂MSE ∂oh
Δbh = η ∗ 0.5 ∗ ∗ (26)
∂oh ∂neth
∂oh M S E ∂neto
where ∂net h
= oh ∗ (1 − oh ) and ∂ M SE
∂oh
= ∂∂net o
* ∂oh . Replace all these mathematical
derivatives in Eq. 26 then will get the mathematical expression as shown in Eq. 27.

∂ M S E ∂neto
Δbh = η ∗ 0.5 ∗ oh ∗ (1 − oh ) ∗ ∗ (27)
∂neto ∂oh

where ∂net
∂oh
o
= Who and ∂∂net
MSE
o
= ∂M SE
∂oo
∂oo
* ∂net o
. Replace all these mathematical deriva-
tives in Eq. 27 then will get the mathematical expression as shown in Eq. 28.
116 V. Veeramsetty et al.

Table 7 Sample dataset in MW

Dataset L(T−1) L(T−24) L(T−72) L(T)
Training 0.5 0.4 0.7 0.6
Testing 0.4 0.3 0.6 0.5

Table 8 Weights between input and hidden layer (Wi h )

Input Hidden neuron
neuron
1 2 3 4 5 6 7
1 −0.1 0.1 0.2 −0.2 0.3 −0.3 0.4
2 0.1 0.2 −0.2 0.3 −0.3 0.4 −0.1
3 0.2 −0.2 0.3 −0.3 0.4 0.1 −0.1

∂MSE ∂oo
Δbh = η ∗ 0.5 ∗ oh ∗ (1 − oh ) ∗ Who ∗ ∗ (28)
∂oo ∂neto
∂oo
where ∂ M SE
∂oo
= −2 ∗ (d − oo ) and ∂net o
= oo ∗ (1 − oo ). Replace all these mathemat-
ical derivatives in Eq. 22 then will get the mathematical expression for Δbh as shown
in Eq. 29.

Δbh = η ∗ oh ∗ (1 − oh ) ∗ Who ∗ (d − oo ) ∗ oo ∗ (1 − oo ) (29)

Numerical Calculations

Problem Update weights and bias parameters of an ANN model to estimate the load
at particular hour of the day L(T) based on load available at one hour, one day and
two days before i.e. L(T−1), L(T−24) and L(T−72). Architecture of ANN model
is presented in Fig. 2. Assume dataset as shown in Table 7.
Initialize weights and bias parameters
Initial random weights Wi h and Who are shown in Tables 8 and 9 respectively. Simi-
larly Initial random bias parameters bh and bo are shown in Table 10.
Forward pass calculations
Calculate net input (neth ) to hidden layer using Eq. 2 as presented below
Short Term Active Power Load Forecasting Using Machine … 117

Table 9 Weights between hidden layer and output layer (Who )

Hidden neuron Output neuron
1
1 −0.1
2 0.1
3 0.2
4 −0.2
5 0.3
6 −0.3
7 0.4

Table 10 Bias parameters (bh ) and (bo )

Hidden neurons Output
neuron
bh bo
1 2 3 4 5 6 7 1
−0.1 0.1 0.2 −0.2 0.3 −0.3 0.4 −0.4

⎡ ⎤
−0.1 0.1 0.2
⎢ 0.1 0.2 −0.2⎥
⎢ ⎥
⎢ 0.2 −0.2 0.3 ⎥
⎢ ⎥
neth = WiTh ∗ X + bh = ⎢⎢ 0.2 0.3 −0.3⎥
⎥
⎢ 0.3 −0.3 0.4 ⎥
⎢ ⎥
⎣−0.3 0.4 0.1 ⎦
0.4 −0.1 −0.1
⎡ ⎤ ⎡ ⎤
−0.1 0.03
⎢ ⎥ ⎢ ⎥
⎡ ⎤ ⎢ 0.1 ⎥ ⎢ 0.09 ⎥
0.5 ⎢ 0.2 ⎥ ⎢ 0.43 ⎥
⎢ ⎥ ⎢ ⎥
∗ ⎣0.4⎦ + ⎢ ⎥ ⎢
⎢−0.2⎥ = ⎢−0.39⎥
⎥
0.7 ⎢ 0.3 ⎥ ⎢ 0.61 ⎥
⎢ ⎥ ⎢ ⎥
⎣−0.3⎦ ⎣−0.22⎦
0.4 0.49

Calculate output (oh ) for hidden layer using Eq. 3 as presented below
118 V. Veeramsetty et al.
⎡ ⎤
1.9704
⎢1.9139⎥
⎢ ⎥
⎢1.6505⎥
1 ⎢ ⎥
oh = =⎢
⎢2.4770⎥
⎥
1 + e−neth ⎢1.5434⎥
⎢ ⎥
⎣2.2461⎦
1.6126

Calculate net input (neto ) to output layer using Eq. 4 as presented below


neto = Who
T
∗ oh + bo = −0.1 0.1 0.2 −0.2 0.3 −0.3 0.4
⎡ ⎤
1.9704
⎢1.9139⎥
⎢ ⎥
⎢1.6505⎥
⎢ ⎥
∗⎢ ⎥
⎢2.4770⎥ − 0.1 = 0.1633
⎢1.5434⎥
⎢ ⎥
⎣2.2461⎦
1.6126

Calculate output (oo ) for output layer using Eq. 5 as presented below
1 1
oo = −net
= = 0.5407
1+e o 1 + e−0.1633

Backward pass calculations

Calculate change in weights (ΔWho ) using Eq. 14. Assume learning rate η = 0.1.

ΔWho = η ∗ (d − oo ) ∗ oo ∗ (1 − oo ) ∗ oh = 0.1 ∗ (0.6 − 0.5407) ∗ 0.5407

⎡ ⎤ ⎡ ⎤
1.9704 0.0029
⎢1.9139⎥ ⎢0.0028⎥
⎢ ⎥ ⎢ ⎥
⎢1.6505⎥ ⎢0.0024⎥
⎢ ⎥ ⎢ ⎥
∗ (1 − 0.5407) ∗ ⎢ ⎥ ⎢
⎢2.4770⎥ = ⎢0.0036⎥
⎥
⎢1.5434⎥ ⎢0.0023⎥
⎢ ⎥ ⎢ ⎥
⎣2.2461⎦ ⎣0.0033⎦
1.6126 0.0024

Update weights between hidden and output layers using Eq. 6 as shown below
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
−0.1 0.0029 −0.0971
⎢ 0.1 ⎥ ⎢0.0028⎥ ⎢ 0.1028 ⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢ 0.2 ⎥ ⎢0.0024⎥ ⎢ 0.2024 ⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
Who = Who + ΔWho = ⎢ ⎥ ⎢ ⎥ ⎢
⎢−0.2⎥ + ⎢0.0036⎥ = ⎢−0.1964⎥
⎥
⎢ 0.3 ⎥ ⎢0.0023⎥ ⎢ 0.3023 ⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎣−0.3⎦ ⎣0.0033⎦ ⎣−0.2967⎦
0.4 0.0024 0.4024
Short Term Active Power Load Forecasting Using Machine … 119

Calculate change in bias parameter (Δbo ) at output layer using Eq. 17 as shown
below.

Δbo = η ∗ (d − oo ) ∗ oo ∗ (1 − oo ) = 0.1 ∗ (0.6 − 0.5407)

∗ 0.5407 ∗ (1 − 0.5407) = 0.0015

Update bias parameter at output layer as shown below

bo = bo + Δbo = −0.1 + 0.0015 = −0.0985

Calculate change in weights (ΔWi h ) between input and hidden layer using Eq. 23
as shown below.
Change in weights connected to first hidden neuron (h = 1) i.e. ΔWi1

ΔWi1 = 0.1 ∗ 1.9704 ∗ (1 − 1.9704) ∗ (−0.0971) ∗ (0.6 − 0.5407)

⎡ ⎤
0.5
∗ 0.5407 ∗ (1 − 0.5407) ∗ ⎣0.4⎦
0.7
⎡ ⎤
0.00013
ΔWi1 = ⎣0.00011⎦
0.00019

Change in weights connected to second hidden neuron (h = 2) i.e. ΔWi2

ΔWi2 = 0.1 ∗ 1.9139 ∗ (1 − 1.9139) ∗ (0.1028) ∗ (0.6 − 0.5407)

⎡ ⎤
0.5
∗ 0.5407 ∗ (1 − 0.5407) ∗ ⎣0.4⎦
0.7
⎡ ⎤
−0.00013
ΔWi2 = ⎣−0.00011⎦
−0.00019

Change in weights connected to third hidden neuron (h = 3) i.e. ΔWi3

ΔWi3 = 0.1 ∗ 1.6505 ∗ (1 − 1.6505) ∗ (0.2024) ∗ (0.6 − 0.5407)

⎡ ⎤
0.5
∗ 0.5407 ∗ (1 − 0.5407) ∗ ⎣0.4⎦
0.7
120 V. Veeramsetty et al.
⎡ ⎤
−0.00016
ΔWi3 = ⎣−0.00013⎦
−0.00022

Change in weights connected to fourth hidden neuron (h = 4) i.e. ΔWi4

ΔWi4 = 0.1 ∗ 2.477 ∗ (1 − 2.477) ∗ (−0.1964) ∗ (0.6 − 0.5407)

⎡ ⎤
0.5
∗ 0.5407 ∗ (1 − 0.5407) ∗ ⎣0.4⎦
0.7
⎡ ⎤
0.00053
ΔWi4 = ⎣0.00042⎦
0.00074

Change in weights connected to fifth hidden neuron (h = 5) i.e. ΔWi5

ΔWi5 = 0.1 ∗ 1.5434 ∗ (1 − 1.5434) ∗ (0.3023) ∗ (0.6 − 0.5407)

⎡ ⎤
0.5
∗ 0.5407 ∗ (1 − 0.5407) ∗ ⎣0.4⎦
0.7
⎡ ⎤
−0.00019
ΔWi5 = ⎣−0.00015⎦
−0.00026

Change in weights connected to sixth hidden neuron (h = 6) i.e. ΔWi6

ΔWi6 = 0.1 ∗ 2.2461 ∗ (1 − 2.2461) ∗ (−0.2967) ∗ (0.6 − 0.5407)

⎡ ⎤
0.5
∗ 0.5407 ∗ (1 − 0.5407) ∗ ⎣0.4⎦
0.7
⎡ ⎤
0.00061
ΔWi6 = ⎣0.00049⎦
0.00086

Change in weights connected to seventh hidden neuron (h = 7) i.e. ΔWi7

ΔWi7 = 0.1 ∗ 1.6126 ∗ (1 − 1.6126) ∗ (0.4024) ∗ (0.6 − 0.5407)

⎡ ⎤
0.5
∗ 0.5407 ∗ (1 − 0.5407) ∗ ⎣0.4⎦
0.7
Short Term Active Power Load Forecasting Using Machine … 121

⎡ ⎤
−0.00029
ΔWi7 = ⎣−0.00023⎦
−0.00041

Update the weights between input and hidden layers as shown below

Wi h = Wi h + ΔWi h = Wi h + [ΔWi1 ΔWi2 ΔWi3 ΔWi4 ΔWi5 ΔWi6 ΔWi7 ]

⎡ ⎤
−0.1 0.1 0.2 −0.2 0.3 −0.3 0.4
Wi h = ⎣ 0.1 0.2 −0.2 0.3 −0.3 0.4 −0.1⎦
0.2 −0.2 0.3 −0.3 0.4 0.1 −0.1
⎡ ⎤
0.000137 −0.000132 −0.000160 0.000529 −0.000187 0.000611 −0.000293
+ ⎣0.000109 −0.000106 −0.000128 0.000423 −0.000149 0.000489 −0.000234⎦
0.000191 −0.000185 −0.000224 0.000741 −0.000261 0.000856 −0.000410
⎡ ⎤
−0.0999 0.0999 0.1998 −0.1995 0.2998 −0.2994 0.3997
= ⎣ 0.1001 0.1999 −0.2001 0.3004 −0.3001 0.4005 −0.1002⎦
0.2002 −0.2002 0.2998 −0.2993 0.3997 0.1009 −0.1004

Calculate change in bias parameters (Δbh ) at hidden layer using Eq. 29 as shown
below.
Change in bias parameters connected to first hidden neuron (h = 1) i.e. Δb1

Δb1 = 0.1 ∗ 1.9704 ∗ (1 − 1.9704) ∗ (−0.0971) ∗ (0.6 − 0.5407) ∗ 0.5407

∗ (1 − 0.5407) = 0.00027

Change in bias parameter connected to second hidden neuron (h = 2) i.e. Δb2

Δb2 = 0.1 ∗ 1.9139 ∗ (1 − 1.9139) ∗ (0.1028) ∗ (0.6 − 0.5407)

∗ 0.5407 ∗ (1 − 0.5407) = −0.00026

Change in bias parameter connected to third hidden neuron (h = 3) i.e. Δb3

Δb3 = 0.1 ∗ 1.6505 ∗ (1 − 1.6505) ∗ (0.2024) ∗ (0.6 − 0.5407)

∗ 0.5407 ∗ (1 − 0.5407) = −0.00032

Change in bias parameter connected to fourth hidden neuron (h = 4) i.e. Δb4

Δb4 = 0.1 ∗ 2.477 ∗ (1 − 2.477) ∗ (−0.1964) ∗ (0.6 − 0.5407)

∗ 0.5407 ∗ (1 − 0.5407) = 0.0011

Change in bias parameter connected to fifth hidden neuron (h = 5) i.e. Δb5

122 V. Veeramsetty et al.

Δb5 = 0.1 ∗ 1.5434 ∗ (1 − 1.5434) ∗ (0.3023) ∗ (0.6 − 0.5407)

∗ 0.5407 ∗ (1 − 0.5407) = −0.00037

Change in bias parameter connected to sixth hidden neuron (h = 6) i.e. Δb6

Δb6 = 0.1 ∗ 2.2461 ∗ (1 − 2.2461) ∗ (−0.2967) ∗ (0.6 − 0.5407)

∗ 0.5407 ∗ (1 − 0.5407) = 0.0012

Change in weights connected to seventh hidden neuron (h = 7) i.e. Δb7

Δb7 = 0.1 ∗ 1.6126 ∗ (1 − 1.6126) ∗ (0.4024) ∗ (0.6 − 0.5407)

∗ 0.5407 ∗ (1 − 0.5407) = −0.00059

Update the bias parameters at hidden layer as shown below

bh = bh + Δbh = bh + [Δb1 ; Δb2 ; Δb3 ; Δb4 ; Δb5 ; Δb6 ; Δb7 ]

⎡ ⎤ ⎡ ⎤ ⎡ ⎤
−0.1 0.0003 −0.0997
⎢ 0.1 ⎥ ⎢−0.0003⎥ ⎢ 0.0997 ⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢ 0.2 ⎥ ⎢−0.0003⎥ ⎢ 0.1997 ⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
bh = ⎢ ⎥ ⎢ ⎥ ⎢
⎢−0.2⎥ + ⎢ 0.0011 ⎥ = ⎢−0.1989⎥
⎥
⎢ 0.3 ⎥ ⎢−0.0004⎥ ⎢ 0.2996 ⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎣−0.3⎦ ⎣ 0.0012 ⎦ ⎣−0.2988⎦
0.4 −0.0006 0.3994

Forward pass on testing data

Calculate net input (neth ) to hidden layer using Eq. 2 as presented below
⎡ ⎤
−0.0999 0.1001 0.2002
⎢ 0.0999 0.1999 −0.2002⎥
⎢ ⎥ ⎡ ⎤
⎢ 0.1998 −0.2001 0.2998 ⎥ 0.4
⎢ ⎥
neth = Wi h ∗ X + bh = ⎢
T
⎢−0.1995 0.3004 −0.2993⎥ ⎣ ⎦
⎥ ∗ 0.3
⎢ 0.2998 −0.3001 0.3997 ⎥ 0.6
⎢ ⎥
⎣−0.2994 0.4005 0.1009 ⎦
0.3997 −0.1002 −0.1004
⎡ ⎤ ⎡ ⎤
−0.0997 0.0105
⎢ 0.0997 ⎥ ⎢ 0.0795 ⎥
⎢ ⎥ ⎢ ⎥
⎢ 0.1997 ⎥ ⎢ 0.3994 ⎥
⎢ ⎥ ⎢ ⎥
+⎢ ⎥ ⎢
⎢−0.1989⎥ = ⎢−0.3682⎥
⎥
⎢ 0.2996 ⎥ ⎢ 0.5694 ⎥
⎢ ⎥ ⎢ ⎥
⎣−0.2988⎦ ⎣−0.2379⎦
0.3994 0.4690
Short Term Active Power Load Forecasting Using Machine … 123

Calculate output (oh ) for hidden layer using Eq. 3 as presented below
⎡ ⎤
0.5026
⎢0.5199⎥
⎢ ⎥
⎢0.5986⎥
1 ⎢ ⎥
oh = = ⎢0.4090⎥
1 + e−neth ⎢ ⎥
⎢0.6386⎥
⎢ ⎥
⎣0.4408⎦
0.6151

Calculate net input (neto ) to output layer using Eq. 4 as presented below


neto = Who
T
∗ oh + bo = −0.0971 0.1028 0.2024 −0.1964 0.3023 −0.2967 0.4024
⎡ ⎤
0.5026
⎢0.5199⎥
⎢ ⎥
⎢0.5986⎥
⎢ ⎥
∗⎢ ⎥
⎢0.4090⎥ − 0.0985 = 0.2567
⎢0.6386⎥
⎢ ⎥
⎣0.4408⎦
0.6151

Calculate output (oo ) for output layer using Eq. 5 as presented below
1 1
oo = −net
= = 0.5638
1+e o 1 + e−0.2567

Error computation
In this example considered only one sample in testing dataset with target load L(T) =
0.5. So MSE and MAE values are computed using Eqs. 10 and 11 respectively with
n s = 1 and presented below.

MSE = (0.5 − 0.5638)2 = 0.0041

MAE = |0.5 − 0.5638| = 0.0638

References

1. Almeshaiei E, Soltan H (2011) Alex Eng J 50(2):137. https://doi.org/10.1016/j.aej.2011.01.

015
2. Alfares HK, Nazeeruddin M (2002) Int J Syst Sci 33(1):23. https://doi.org/10.1080/
00207720110067421
3. Su P, Tian X, Wang Y, Deng S, Zhao J, An Q, Wang Y (2017) Energies 10(9):1303. https://
doi.org/10.3390/en10091303
124 V. Veeramsetty et al.

4. Mi J, Fan L, Duan X, Qiu Y (2018) Math Probl Eng 2018. https://doi.org/10.1155/2018/

3894723
5. Hu R, Wen S, Zeng Z, Huang T (2017) Neurocomputing 221:24. https://doi.org/10.1016/j.
neucom.2016.09.027
6. Lau E, Sun L, Yang Q (2019) SN Appl Sci 1(9):982. https://doi.org/10.1007/s42452-019-
0884-7
7. Kumar BA, Sangeetha G, Srinivas A, Awoyera P, Gobinath R, Ramana VV (2020) Soft com-
puting for problem solving. Springer, pp 677–690. https://doi.org/10.1007/978-981-15-0184-
5_58
8. Awoyera P, Akinmusuru J, Krishna AS, Gobinath R, Arunkumar B, Sangeetha G (2020) Soft
computing for problem solving. Springer, pp 197–207. https://doi.org/10.1007/978-981-15-
0035-0_15
9. Veeramsetty V, Singal G, Badal T (2020) Multimed Tools Appl 79:22569. https://doi.org/10.
1007/s11042-020-09031-0
10. Veeramsetty V, Deshmukh R (2020) SN Appl Sci 2(5):1. https://doi.org/10.1007/s42452-020-
2601-y
11. Amral N, Ozveren C, King D (2007) In: 2007 42nd international universities power engineering
conference. IEEE, pp 1192–1198. https://doi.org/10.1109/UPEC.2007.4469121
12. Kumar S, Mishra S, Gupta S (2016) In: 2016 second international conference on computational
intelligence & communication technology (CICT). IEEE, pp 184–186. https://doi.org/10.1109/
CICT.2016.44
13. Shaloudegi K, Madinehi N, Hosseinian S, Abyaneh HA (2012) IEEE Trans Power Syst
27(2):811. https://doi.org/10.1109/TPWRS.2011.2175254
14. Veeramsetty V, Chintham V, Vinod Kumar D (2018) Int Trans Electr Energy Syst 28(8):e2573.
https://doi.org/10.1002/etep.2573
15. Veeramsetty V, Chintham V, Vinod Kumar DM (2018) Int J Energy Sect Manag 12(3). https://
doi.org/10.1108/IJESM-03-2017-0002
16. Chen Y, Xu P, Chu Y, Li W, Wu Y, Ni L, Bao Y, Wang K (2017) Appl Energy 195:659. https://
doi.org/10.1016/j.apenergy.2017.03.034
17. Salgado RM, Lemes RR (2013) J Control Autom Electr Syst 24(6):854. https://doi.org/10.
1007/s40313-013-0067-5
18. Yang ZC (2015) J Control Autom Electr Syst 26(4):430. https://doi.org/10.1007/s40313-015-
0186-2
19. Ke K, Hongbin S, Chengkang Z, Brown C (2019) Evol Intell 12(3):385. https://doi.org/10.
1007/s12065-018-00196-0
20. Houimli R, Zmami M, Ben-Salha O (2019) Energy Syst 1–19. https://doi.org/10.1007/s12667-
019-00324-4
21. Lisi F, Shah I (2019) Energy Syst 1–33. https://doi.org/10.1007/s12667-019-00356-w
22. Li K, Zhang T (2019) Energy Syst 1–16. https://doi.org/10.1007/s12667-019-00344-0
23. Shamir O, Zhang T (2013) In: International conference on machine learning, pp 71–79. http://
arxiv.org/abs/1212.1824
24. Veeramsetty V (2020) Mendeley data. https://doi.org/10.17632/ycfwwyyx7d.2
25. Barga R, Fontama V, Tok WH, Cabrera-Cordon L (2015) Predictive analytics with Microsoft
Azure machine learning. Springer. https://doi.org/10.1007/978-1-4842-0445-0
Evaluation of Algorithms
for Fundamental and Harmonic Impacts
of Integration of Renewable Energy
Sources in Smart Power Distribution
Networks

R. Satish, K. Vaisakh, and Almoataz Y. Abdelaziz

Abstract With the development of renewable energy sources (RES), distributed

generation (DG), microgrid, electric vehicles etc., in traditional distribution network
which was earlier supply-side driven, single power supply, one-way service and
basic reliance on manual management of business model turns into the direction of
user interaction, the trend of two-way flow and highly automated and these changes
gradually formed the smart power distribution network (SPDN). The rapid advance-
ment in power electronic devices increases the fast growth of RES into SPDN and
also increases the non-linear portion of loads on the network. Due to this, harmonic
currents may inject into distribution networks. This chapter proposes new power
flow algorithms (PFAs) for fundamental and harmonic analysis of SPDN with the
integration of renewable energy based DG and presence of non-linear loads. The
proposed methodology completely exploits the radial feature in distribution networks
and developed BUS_NUM and BRANCH_NUM matrices. These matrices make the
implementation of PFAs simple. The accuracy of fundamental and harmonic PFAs is
tested on IEEE-13 bus test feeder and they are found to be accurate with the literature.
Different case studies are carried on IEEE-13 bus and IEEE-34 bus test feeders for
analyzing the impacts of multiple integration of synchronous based-renewable DG
and power electronic based-renewable DG.

Keywords Power flow · Unbalanced networks · Linear loads · Non-linear loads ·

Distributed generation · Renewable energy sources · Harmonic distortion · Smart
distribution networks

R. Satish (B)
Department of Electrical and Electronics Engineering, ANITS (A), Visakhapatnam, Andhra
Pradesh, India
K. Vaisakh
Department of Electrical Engineering, Andhra University, AUCE (A), Visakhapatnam, Andhra
Pradesh, India
A. Y. Abdelaziz
Faculty of Engineering and Technology, Future University in Egypt, Cairo, Egypt

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 125
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6_6
126 R. Satish et al.

Nomenclature

RES Renewable Energy Sources

DG Distributed Generation
RDN Radial Distribution Network
SPDN Smart Power Distribution Network
PFA Power Flow Algorithm
PEC Power Electronic Converter
FPFA Fundamental Power Flow Algorithm
HPFA Harmonic Power Flow Algorithm
THD Total Harmonic Distortion
FSIG Fixed Speed Induction Generator
DFIG Doubly Fed Induction Generator
SVCM Static Voltage Characteristic Model

1 Introduction

RES such as photovoltaic, wind, fuel cells, etc., uses synchronous/induction genera-
tors or induction generators combined with PECs or only PECs to transfer AC power
to the grids. Therefore RES can be a synchronous-based or a power electronic-based.
Renewable energy can make a contribution to improve power quality, minimize peak
loads and reduce the greenhouse gas emission. The connection of DG with the utility
grid is specified in [1]. In [2], proposed a three-phase PFA for real-time distribution
networks. It gives some primary discussions on PV node concept. DG modeling and
their integration to the distribution networks is presented in [3, 4]. Harmonics may
cause localized overheating, winding insulation stresses and torque pulsations with
torsional vibrations on electric machines, additional power losses on devices, light
flicker, communication interference and equipment loss of life or even damage if
resonance conditions occur. Therefore HPFAs are essential for finding the harmonic
distortion level in the presence of non-linear loads and renewable DG. The authors
of [5, 6] propose a HPFA based on backward/forward sweep method. The improve-
ment of power quality using fuel cell and fuzzy based controller is discussed in
[7]. By using this controller, the quality of power in the grid system especially in
micro grid connected with non-linear and unbalanced load is enhanced. For the fair
allocation of active power losses [8, 9], develops a new active power loss alloca-
tion technique which eradicates the influence of cross-term analytically from loss
formulation without any assumptions and approximations. A new active power loss
allocation scheme is developed by eliminating the influence of cross-term mathemati-
cally from loss equation for allocating losses to the network participants with/without
DGs [10]. This method assigns losses to the consumers/DG units with due consid-
eration to their power consumptions/injections and physical locations in the radial
distribution network.
Evaluation of Algorithms for Fundamental and Harmonic Impacts … 127

This chapter proposes a new FPFA and HPFA for three-phase SPDN using
BUS_NUM and BRANCH_NUM matrices. These algorithms use basic principles of
circuit theory and hence they can be easily understood. In this chapter, a new scheme
is developed to divide the distribution network in to new sections. The bus numbers
and branch numbers of these newly created sections are stored in row entries of
BUS_NUM and BRANCH_NUM matrices respectively. These matrices make the
PFA simple. The impacts of multiple installations of synchronous based-DG, power
electronic based-DG and presence of non-linear loads on fundamental and harmonic
voltage profile, fundamental and harmonic power loss and THD % are analyzed
through the proposed algorithms. The compensation-based method [2] is used to
handle the DGs in the proposed PFAs. This chapter is organized in the following
order. The modeling of network components is discussed in Sect. 2. Methodology
for developing BUS_NUM and BRANCH_NUM matrices is discussed in Sect. 3.
FPFA for SPDN is presented in Sect. 4. Integration of renewable DGs into FPFA is
presented in Sect. 5. HPFA for SPDN with integration of renewable DGs and pres-
ence of non-linear loads is presented in Sect. 6. Case studies and discussions with
case studies on IEEE-13 node test feeder are presented in Sect. 7. Case studies and
discussions with case studies on IEEE-34 node test feeder are presented in Sect. 8.
The conclusions are presented in Sect. 9.

2 Modeling of Network Components

The modeling of lines, loads, capacitor banks, voltage regulators and distribution
transformers for three-phase analysis is presented in [11]. The liner loads can be
modeled in several ways [12] for harmonic analysis. Each model will show different
impact on the network. In this chapter, the linear loads are modeled as impedances
with series combination of resistance and reactance. The non-linear loads are modeled
as constant current sources [13] with magnitude obtained from the typical harmonic
spectrum and rated load current obtained from the fundamental power flow. The
phase angle of the current source is obtained from the phase angle of rated current
at fundamental frequency and phase angle of the harmonic source current spectrum.
The modeling of other network components for harmonic analysis is presented in
[14].

2.1 Renewable DG

The type of device used to interface the renewable DG to grid depends on its applica-
tion and the type of energy source utilized. For example a wind turbine may employ
a squirrel cage induction generator, called a FSIG, a DFIG or a PEC connected
generator. DC sources or those generating at higher frequency, such as photovoltaic
systems, fuel cells require a PEC to interface them into SPDN. Figure 1 gives the
128 R. Satish et al.

Solar Wind Hydro Geothermal Biofuel

Fuel Cell
Power Power Power Power Power

Power
Induction Synchronous
Electronic
Generator Generator
Converter

Power system interface, Communication

interface, Measurement and Protection devices

Distribution Network

Fig. 1 Possible combination of renewable DG

possible interfacing of renewable DG with SPDN. The modeling of renewable DG

for PFA requires the knowledge of operation and control on the characteristics of the
interfacing devices. A brief modeling of interfacing devices is extracted as follows.

2.1.1 Synchronous Generator

Based on the excitation system the synchronous generators are of two types [3]. The
first one is fixed excitation voltage type and the second one is regulating excitation
voltage type. In regulating excitation voltage type the excitation can be controlled in
two ways. One is maintaining the constant terminal voltage (voltage control mode)
and the other is maintaining the constant power factor (power factor control mode).
In PFAs the synchronous generator which maintains constant terminal voltage is
considered as PV bus and which maintains constant power factor is treated as PQ
bus. The synchronous generator with fixed excitation voltage [15] is treated as a
SVCM.

2.1.2 Induction Generator

The active and reactive powers developed by the induction generator are functions of
both slip and bus voltage [16]. Assuming the active power developed as constant and
neglecting the very low dependency of reactive power with the slip, the induction
generator can be modeled as SVCM and further since the bus voltage is nearer to
Evaluation of Algorithms for Fundamental and Harmonic Impacts … 129

1 p.u in steady state cases, the reactive power is also constant. Hence the induction
generator is modeled as PQ bus in PFA.

2.1.3 Power Electronic Converter

The modeling of PECs in PFA depends on the control method employed in converter
control circuit. The PEC is modeled as PV bus if the control circuit in converter is
designed to control both P and V independently [14]. If it is designed to control P
and Q independently then it is modeled as PQ bus [17].
For HPFA, the power electronic based-renewable DG is considered as the sources
for harmonics. The current spectrum of the power electronic based-renewable DG
is taken from [18]. For synchronous based-renewable DG the harmonic current
injection is taken as zero.

3 Methodology for Developing BUS_NUM

and BRANC_NUM Matrices

The PFA of SPDN requires systematic numbering of buses and branches since it
saves the memory and it is also strengthen the numerical performance of the solution
methods. The numbering scheme adopted here is taken from [19]. In Fig. 2 starting
with the bus 1 (‘1’ is enclosed in circle) in the lateral-1 (‘1’ is in square brackets) and
then sequentially considering buses until it finds one that has more than one branch
leaving from it and is labeled as section-I. So that section-I is formed with buses

12 13 14 15 16
12 13 14 15 [2]
IV
11

1 2 3 4 5 6 7 8 9 10 11
1 2 3 4 5 6 7 8 9 10 [1]
I II III
16
17 18 19 20 21
17 18 19 20 [3]
V VI
21
22 23
22
[4]
VII

Fig. 2 Divided sections for the sample RDN

130 R. Satish et al.

Table 1 Sections information from Fig. 2

Section No. Bus numbers Head bus Tail bus Branch Head branch Tail branch
numbers which
are ahead of
buses
I 1, 2, 3 1 3 1, 2 1 2
II 4, 5, 6 4 6 3, 4, 5 3 5
III 7, 8, 9, 10, 11 7 11 6, 7, 8, 9, 10 6 10
IV 12, 13, 14, 15, 12 16 11, 12, 13, 14, 11 15
16 15
V 17, 18 17 18 16, 17 16 17
VI 19, 20, 21 19 21 18, 19, 20 18 20
VII 22, 23 22 23 21, 22 21 22

from 1 to 3. Then start proceeding from bus 4 until it find the one that has more
than one branch leaving from it and it is labeled as section-II. Therefore section-II
has buses from 4 to 6. Likewise the section-III created with buses from 7 to 11 in
it. Once all buses in lateral-1 are over then repeat the same procedure for all other
laterals After the entire distribution network is divided into sections we can have the
following information as in Table 1.
A MATLAB code is developed to divide the network in to sections and to store the
information of bus numbers and branch numbers of these sections in the following
matrix entries.
Evaluation of Algorithms for Fundamental and Harmonic Impacts … 131

4 Fundamental Power Flow Algorithm for SPDN

After the BUS_NUM and BRANCH_NUM matrices are developed for the SPDN,
the following steps illustrate the iterative procedure for PFA.
1. The voltages at all busses are assigned as substation bus voltage.
⎡ ⎤ ⎡ ⎤
Va 1∠0◦
⎣ Vb ⎦ = ⎣ 1∠ − 120◦ ⎦ p.u (1)
Vc 1∠120◦

2. Find the load current at all buses with the specified load power, type of load and
voltage at the bus.
3. The power flow program should start with collecting current at bus 23 (the tail
bus in section-VII as in BUS_NUM) thereby find the current in branch 22 (the
tail branch in section-VII as in BRANCH_NUM). Then proceed towards head
bus (i.e., bus 22) and head branch (i.e., branch 21) in finding the bus currents and
branch currents respectively. This is illustrated with a sample section consists
of three buses as shown in Fig. 3.

[Iabc ]k = [ILabc ]k + [Ishabc ]k + [ICabc ]k (2)

[Iabc ]jk = [Iabc ]k (3)

[Iabc ]j = [Iabc ]jk + [ILabc ]j + [Ishabc ]j + [ICabc ]j (4)

[Iabc ]ij = [Iabc ]j (5)

[Iabc ]i = [Iabc ]ij (6)

Fig. 3 A sample section consists of three busses

132 R. Satish et al.

where,
[Iabc ]k : Three-phase current at bus ‘k’.
[Iabc ]jk : Three-phase current in branch ‘jk’.
[ILabc ]k : Three-phase load current at bus ‘k’.
[Ishabc ]k : Three-phase current drawn by shunt admittance at bus ‘k’.
[ICabc ]k : Three-phase current drawn by capacitor bank at bus ‘k’ if any.
4. Then go to section-VI and repeat the above step in finding head bus current and
head branch current. Likewise, move up to section-I and find current at head
bus in section-I (bus 1) and head branch in section-I (branch-1).
5. Now start with bus 1 (head bus in section-I), and proceed towards the tail bus
in finding the bus voltages with Eq. 7.
⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤
Va Va Zaa Zab Zac Ia
⎣ Vb ⎦ = ⎣ Vb ⎦ − ⎣ Zba Zbb Zbc ⎦ • ⎣ Ib ⎦ (7)
Vc j Vc i Zca Zcb Zcc ij Ic ij

6. Then go to section-II and proceed towards the tail bus in finding the bus voltages.
Likewise move up to bus 23 (tail bus in section-VII).
7. Repeat the steps 3 to 6 until the voltage magnitude mismatches at all buses in
successive iterations is below the tolerance limit.
 
[Vabc ]r − [Vabc ]r−1  ≤ [εabc ] (8)
i i

where, ‘r’ is the iteration number.

8. Find the power losses in all the branches are obtained with Eq. 9.
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
S Lossa (Va )i · (Ia )∗ij (Va )j · (Ia )∗ji
⎣ S Lossb ⎦ = ⎢ ⎥ ⎢ ⎥
⎣ (Vb )i · (Ib )∗ij ⎦ − ⎣ (Vb )j · (Ib )∗ji ⎦ (9)
S Lossc ij (Vc )i · (Ic )∗ij (Vc )j · (Ic )∗ji

The flowchart of this algorithm is shown in Fig. 4.

5 Integration of Renewable DGs into FPFA of SPDN

The detailed iterative procedure for integration of renewable DG is illustrated with

the following steps:
1. Run the power flow as presented in section-4. After the power flow is converged,
select the locations of renewable DG.
Evaluation of Algorithms for Fundamental and Harmonic Impacts … 133

start

Read the system data

Brake the system in to sections and construct BUS_NUM

& BRANCH_NUM matrices

Set itr =1

Start with tail bus & branch in last section and move towards head bus & head
branch in finding the bus currents and branch currents using Eq. (2) to Eq. (6)

Move to the predecessor section and repeat above procedure to find the head
bus current and head branch current.

Repeat the above step until head bus current and head branch current of first
section is obtained.

Set itr = itr +1 Start with head bus of first section and find its voltage using Eq. (7), then
proceed towards the tail bus in finding the voltages

Move to the next section and repeat the above step until the bus voltage of tail
bus of last section is obtained.

No
Check for convergence
using Eq. (8)

Yes
Find the power loss using Eq. (9)

stop

Fig. 4 Flowchart of FPFA

2. Then for the outside γ th iteration, check the type of renewable DG available.
3. If the renewable DG is modeled as PQ at bus say ‘j’, then the current
injected

by the renewable DG is obtained with the specified power rating SG,abc j and
the bus voltage [Vabc ] j as,
⎡ ∗ ⎤γ

γ  SG,a /Va ∗
IG,abc j = ⎣ SG,b /Vb ⎦ (10)
 ∗
SG,c /Vc j

4. If the renewable DGs are modeled as PV buses. Then calculate the voltage
mismatches at all the PV buses as,
134 R. Satish et al.

⎡ ⎤γ  sp   cal γ
Va  Va   Va 
   
⎣ Vb ⎦ =  Vsp  −  Vcal  PV buses ∀ (11)
 b   b 
Vc  Vsp   Vcal 
c c

where,
[Vabc ]γ : Voltage mismatch matrix for outside γ th iteration. Its size is 3 · n × 1 for
‘n’ PV buses in the network.
 sp 
V : Magnitude of specified voltage at ith PV buses.
abc
 cal γ
V  : Magnitude of voltage calculated at ith PV buses for outside kth iteration.
abc

5. If the voltage mismatches are not less than the specified tolerance value, then
incremental current injections at PV buses in order to maintain the specified
voltages is calculated with Eq. 12.

−1
[I]γ = Zpv · [V]γ (12)

Zpv is the PV bus sensitivity matrix. The Zpv can be formed by observing the
following numerical properties of its entries [2]. The diagonal entry Zpp is equal to
the modulus of the sum positive sequence impedance of all line sections between PV
buses ‘p’ and the substation bus. Since all the lines are three phase line sections, the
size of Zpp is 3 × 3. If any two PV nodes, ‘p’ and ‘q’, have different paths from root
node, then the off diagonal entry Zpq (size 3×3) is zero. If ‘p’ and ‘q’ share a piece of
common path to the root node, then Zpq is equal to the modulus of the sum positive
sequence

impedance of all line sections on this common path. The dimension of
Zpv is equal to 3 · n × 3 · n.
6. If the reactive power generations of renewable DGs are unlimited for the spec-
ified real power generation, then incremental reactive current to be injected at
the jth PV bus is obtained with Eq. 13.
⎡ ⎤γ ⎡      ⎤γ
IG,a |Ia | · cos 90◦ + δv,a + j ∗ sin 90◦ + δv,a
  ◦   ◦ 
⎣ IG,b ⎦ = ⎣ |Ib | · cos 90 + δv,b + j ∗ sin 90 + δv,b ⎦ (13)
    
IG,c j |Ic | · cos 90◦ + δv,c + j ∗ sin 90◦ + δv,c j

δv,a , δv,b and δv,c are the voltage angles at the jth PV bus.
7. Now, find the current in branch ‘ij’ from Fig. 5.
⎡ ⎤γ ⎡ ⎤γ ⎡ ⎤γ
Ia IL,a IG,a
⎣ Ib ⎦ = ⎣ IL,b ⎦ − ⎣ IG,b ⎦ (14)
Ic ij IL,c j IG,c j

γ γ
γ
With [Vabc ]j and [Iabc ]ij the reactive power flow in the line Qabc ij is evaluated.
Then incremental reactive power injection is evaluated with Eq. 15.
Evaluation of Algorithms for Fundamental and Harmonic Impacts … 135

Fig. 5 A sample two nodes in the system with renewable DG placed at bus ‘j’

⎡ ⎤γ ⎡ ⎤γ ⎡ ⎤γ
QG,a QL,a Qa
⎣ QG,b ⎦ = ⎣ QL,b ⎦ − ⎣ Qb ⎦ (15)
QG,c j QL,c j Qc ij

Find the reactive power generation needed at jth PV bus using Eq. 16.
⎡ ⎤γ ⎡ ⎤γ−1 ⎡ ⎤γ
QG,a QG,a QG,a
⎣ QG,b ⎦ = ⎣ QG,b ⎦ + ⎣ QG,b ⎦ (16)
QG,c j QG,c j QG,c j

Find the complex power generation at PV bus using Eq. 17.

⎡ ⎤γ ⎡ ⎤ ⎡ ⎤γ
SG,a PG,a QG,a
⎣ SG,b ⎦ = ⎣ PG,b ⎦ + j ∗ ⎣ QG,b ⎦ (17)
SG,c j PG,c j QG,c j

where,

PG,abc j : Specified three-phase real power generation of the renewable DG at jth

bus.
8. With the complex power obtained in Eq. 18, the current injected by the renewable
DG is calculated with Eq. 10.
9. If the reactive power generations by the renewable DG were limited, then reac-
tive power generation limits must be checked first. The total three phase reactive
power needed at jth PV node is the sum of reactive power of three phases.
γ  γ  γ  γ
(QG )j = QG,a j + QG,b j + QG,c j (18)

γ
If Qj,min ≤ (QG )j ≤ Qj,max , then complex power generation is as Eq. 17.
136 R. Satish et al.

γ γ  γ  γ  γ
If (QG )j ≤ Qj,min then set (QG )j = Qj,min and QG,a j = QG,b j = QG,c j =
Qj,min /3.
γ γ  γ  γ  γ
If (QG )j ≥ Qj,max then set (QG )j = Qj,max and QG,a j = QG,b j = QG,c j =
Qj,max /3.
Where, Qj,min and Qj,max are the minimum and maximum reactive power generation
limits of renewable DG at jth PV bus.
These reactive power generations are combined with active power generations at
PV buses and find the current injections by renewable DG as in Eq. 10.
10. Now set γ = γ + 1 and run the power flow with the current injections at
renewable DG.
11. Power flow will be stopped when it attains convergence criterion.
The flowchart for this algorithm is shown in Fig. 6.

6 HPFA for SPDN with Integration of Renewable DGs

and Presence of Non-linear Loads

The detailed iterative procedure for HPFA with integration of renewable DGs is
illustrated with the following steps,
1. Run the FPFA with integrations of renewable DGs as presented in section-5.
2. With the converged bus voltages and specified loads, the impedance of linear
loads is calculated for the harmonic order-h of interest.
3. Find the harmonic current injections of non-linear loads and power electronic
based-renewable DGs for the selected harmonic order-h from the current spectra
of the non-linear loads and DGs respectively.
4. The harmonic voltage at the substation bus is taken as zero since the supply
voltage is assumed as a pure sinusoidal voltage waveform.
5. Find the net harmonic currents at all the buses with the harmonic current drawn
by linear loads and harmonic current injections of non-linear loads and renew-
able DGs. The current drawn by the linear loads at all the buses is zero for
the first iteration. The calculation of harmonic currents is illustrated with the
sample section as shown in Fig. 7. The net harmonic current at node ‘j’ is given
by Eq. 19 and the harmonic current in branch ‘ij’ is given by Eq. 20.

[Iabc ]hj = −[ISabc ]hj − [IGabc ]hj + [ILabc ]hj (19)

[Iabc ]hij = [Iabc ]hj (20)

Using the BUS_NUM and BRANCH_NUM matrices, the harmonic currents in

all branches are to be obtained by moving up to the substation.
Where,
Evaluation of Algorithms for Fundamental and Harmonic Impacts … 137

Start

Set γ=1
Run the FPFA as in Section-4

PQ bus Model the buses which

Inject current using
have DG & D-STATCOM
Eq. (10)
injections
PV bus
Find the ΔV at all PV buses using Eq. (11)

Yes
Run the FPFA as in Section-4 Is ΔV is less than ε

No Total FPL and VUF %

at all buses are
Create ZPV Matrix evaluated
Find current using Eq. (13) Stop

Find reactive power generations using Eq. (16)

Is there limits on reactive γ=γ+1

No
Yes power generations of DG
if and D-STATCOM
No Q j,min QG j Q j,max

Yes Inject current using

Set the complex power injections using Eq. (17) Eq. (10)

Set Q G j Q j,min if Q G j Q j,min Set Q G j Q j,max if Q G j Q j,max

obtain complex power injections obtain complex power injections
using Eq. (17) using Eq. (17)

Fig. 6 Flowchart of FPFA with multiple renewable DG integrations

Fig. 7 A sample section of two buses for HPFA

138 R. Satish et al.

[Iabc ]hj : Three-phase harmonic current at bus ‘j’ for harmonic order-h.
[Iabc ]hij : Three-phase harmonic current in branch ‘ij’ for harmonic order-h.
[ILabc ]hj : Three-phase harmonic current drawn by linear load at bus ‘j’ for harmonic
order-h.
[ISabc ]hj : Three-phase harmonic current injection by non-linear load at bus ‘j’ for
harmonic order-h.
[IGabc ]hj : Three-phase current injection by renewable DG at bus ‘j’ for harmonic
order-h.
6. Now, start with section-I and find the harmonic bus voltages by moving towards
the tail bus using Eq. 21.
⎡ ⎤h ⎡ ⎤h ⎡ ⎤h ⎡ ⎤h
Va Va Zaa Zab Zac Ia
⎣ Vb ⎦ = ⎣ Vb ⎦ − ⎣ Zba Zbb Zbc ⎦ • ⎣ Ib ⎦ (21)
Vc j Vc i Zca Zcb Zcc ij Ic ij

Move to the next section and repeat this procedure until harmonic voltage
at tail bus in last section is obtained.
7. Repeat the steps 5 and 6 until the magnitude mismatch of harmonic voltages
of h-order at all the busses is within the tolerance limit.
8. The complex power loss for h-order harmonic is obtained with Eq. 22.

⎡ ⎤h ⎡ ⎤h ⎡ ⎤h
SLossa (Va )i · (Ia )∗ij (Va )j · (Ia )∗ji
⎣ SLossb ⎦ = ⎢ ⎥ ⎢ ⎥
⎣ (Vb )i · (Ib )∗ij ⎦ − ⎣ (Vb )j · (Ib )∗ji ⎦ (22)
SLossc ij (Vc )i · (Ic )∗ij (Vc )j · (Ic )∗ji

9. Find the total real power loss due to harmonics frequency of order-h using
Eq. 23.

Nbr
 

[P_loss]h = real [SLoss]hbr (23)

br=1

10. Repeat the steps from 2 to 9 for all harmonics (order: 3, 5, 7, 9, 11, 13 and 15)
of the selected range of frequency.
11. The total power loss due to all harmonics is,

hm
Nbr
 

[Total_loss] = real [SLoss]hbr (24)

h=ho br=1

12. The r.m.s value of voltage at bus ‘i’, say phase ‘a’ is calculated as,
Evaluation of Algorithms for Fundamental and Harmonic Impacts … 139


 2 hm
 
(Va )i = (Va )1i  + (Va )h 2
i (25)
h=ho

13. The total harmonic distortion at every bus is defined as,


hm 
 
(Va )h 2
i
h=ho
(THD)ai =   (26)
(Va )1 
i

where
ho : Minimum harmonic order
hm : Maximum harmonic order
br: Branch number
Nbr: Total number of branches.
The flowchart for HPFA with multiple injections of renewable DGs is shown in
Fig. 8.

7 Case Studies and Discussions on IEEE-13 Bus Feeder

7.1 Case Study 1: Fundamental Power Flow Solution

The data for IEEE-13 bus feeder shown in Fig. 9 is taken from [20]. The FPFA
is developed on MATLAB software with convergence tolerance of 10–4 . The base
values selected for the system are 5000 kVA and 4.16 kV. This FPFA is converged
after 5 iterations. For testing the accuracy of FPFA, the voltage solution of IEEE-
13 bus test feeder is presented Table 2 in comparison with the IEEE results. The
maximum mismatches for voltage magnitudes and angle are found to be 0.0005 p.u
and 0.010° respectively and are of insignificant. Therefore the test results are almost
matches the IEEE results [20] in terms of accuracy. The power loss of the network
in comparison with IEEE results is presented in Table 3.

7.2 Case Study 2: Harmonic Power Flow Solution

The regulator between buses 650 and 632 is removed and the capacitor banks at bus
675 and 611 are removed from the network. The data for harmonic load composition
and current spectra of harmonic loads is taken from [21]. The convergence tolerance
140 R. Satish et al.

start

Read the system data

Run the FPFA as given in Section 5

Set h=ho

Find the impedance of linear loads and harmonic current

injections of non-linear loads and renewable DGs

Set the harmonic voltages at all buses equal to zero

Set itr =1

Start with tail bus & tail branch in last section and move towards head bus & head branch
in finding the harmonic bus currents and harmonic branch currents respectively.

Move to the predecessor section and repeat the above procedure

Repeat the above step until the harmonic current in head bus & head branch of
first section is obtained.

Start with head bus of first section and proceed towards the tail bus in finding
the harmonic voltages.
Set itr = itr +1
Move to the next section and repeat the above procedure Go to next ‘h’

Repeat the above step until harmonic voltage at tail bus of last section is obtained.

No
Check for convergence

Yes
Find the harmonic power loss

No
Is h=hm
Yes
Find total harmonic loss, total r.m.s voltage & THD %

stop

Fig. 8 Flowchart of HPFA with multiple integrations of renewable DGs

is taken as 10−4 . Table 4 presents the harmonic losses for harmonic frequencies of
order 3, 5, 7, 9, 11, 13 and 15, total harmonic loss of the network and total power loss
of the network including fundamental and harmonic loss. Total harmonic voltage
profile and THD % for the selected range of harmonics of order 3, 5, 7, 9, 11, 13
and 15 are presented in Table 5. From the Table 5 it is observed that, the maximum
THD % on the network is 5.2263 at bus 611 for c-phase and in [22] it was reported
that maximum THD % at bus 611 for c-phase is 5.23. Therefore the results of the
proposed HPFA are almost matches the literature in terms of accuracy. The THD%
Evaluation of Algorithms for Fundamental and Harmonic Impacts … 141

Fig. 9 IEEE 13 bus test

feeder

more than 5 is not desirable according to [23]. It is found that, the number of phases
in the network with THD% more than 5 is 2.

7.3 Case Study 3: Harmonic Power Flow Solution

with Multiple Integrations of Renewable DGs

One renewable DG is placed at bus 634 and other renewable DG placed at bus 675.
The renewable DG at bus 634 is taken as PQ model with per phase real and reactive
power generations are 300 kW and 197 kVAR respectively. Another renewable DG
placed at bus 675 is taken as PV model with its per phase real power generation is
260 kW and the range of three-phase reactive power generation is 100 kVAR to 650
kVAR. The phase voltages specified at this bus are 1 p.u. In this case, the harmonic
impacts are analyzed by treating both the renewable DGs as power electronic based
in case-A and synchronous based in case-B. The convergence tolerance is taken as
10–4 . Table 6 presents the harmonic power loss for harmonic frequencies of order 3,
5, 7, 9, 11, 13 and 15, total harmonic loss of the network and total power loss of the
network including fundamental and harmonic loss for case studies A and B. Table 7
presents the total r.m.s voltages and THD % for case studies A and B in case study
3.

7.4 Discussions

The results of case-2 such as voltage profile, power loss and THD % are used as
benchmarks to see the fundamental and harmonic impacts of renewable DGs. In
case-A of case-3, in which both DGs are power electronic based, the maximum
THD% is observed to be 5.7280 at bus 611 for c-phase which is more than THD%
in reported in case-2. The number of phases in buses with THD % more than 5 is
142 R. Satish et al.

Table 2 Fundamental voltage solution for case study 1

Bus Phase Calculated IEEE Results Error in voltage Error in voltage
results [20] Mag Ang
650 a 1 ∠ 0° 1 ∠ 0° 0.0000 0.00
b 1 ∠ −120° 1 ∠ −120° 0.0000 0.00
c 1 ∠ 120° 1 ∠ 120° 0.0000 0.00
RG a 1.0625 ∠ 0° 1.0625 ∠ 0° 0.0000 0.00
b 1.0500 ∠ − 1.0500 ∠ − 0.0000 0.00
120° 120°
c 1.0687 ∠ 120° 1.0687 ∠ 120° 0.0000 0.00
632 a 1.0210 ∠ − 1.0210 ∠ − 0.0000 0.00
2.49° 2.49°
b 1.0420 ∠ − 1.0420 ∠ − 0.0000 0.00
121.72° 121.72°
c 1.0175 ∠ 1.0170 ∠ −0.0005 0.00
117.83° 117.83°
671 a 0.9900 ∠ − 0.9900 ∠ − 0.0000 0.00
5.30° 5.30°
b 1.0529 ∠ − 1.0529 ∠ − 0.0000 0.00
122.34° 122.34°
c 0.9778 ∠ 0.977 ∠ 0.0001 −0.01
116.03° 116.02°
680 a 0.9900 ∠ − 0.9900 ∠ − 0.0000 0.00
5.30° 5.30°
b 1.0529 ∠ − 1.0529 ∠ − 0.0000 0.00
122.34° 122.34°
c 0.9778 ∠ 0.977 ∠ 0.0001 −0.01
116.03° 116.02°
633 a 1.0180 ∠ − 1.0180 ∠ − 0.0000 0.01
2.55° 2.56°
b 1.0401 ∠ − 1.0401 ∠ − 0.0000 0.00
121.77° 121.77°
c 1.0148 ∠ 1.0148 ∠ 0.0000 0.00
117.82° 117.82°
634 a 0.9940 ∠ − 0.9940 ∠ − 0.0000 0.00
3.23° 3.23°
b 1.0218 ∠ − 1.0218 ∠ − 0.0000 0.00
122.22° 122.22°
c 0.9960 ∠ 0.9960 ∠ 0.0000 −0.01
117.35° 117.34°
645 b 1.0328 ∠ − 1.0329 ∠ − 0.0001 0.00
121.90° 121.90°
(continued)
Evaluation of Algorithms for Fundamental and Harmonic Impacts … 143

Table 2 (continued)
Bus Phase Calculated IEEE Results Error in voltage Error in voltage
results [20] Mag Ang
c 1.0155 ∠ 1.0155 ∠ 0.0001 0.00
117.86° 117.86°
646 b 1.0311 ∠ − 1.0311 ∠ − 0.0000 0.00
121.98° 121.98°
c 1.0134 ∠ 1.0134 ∠ 0.0000 0.01
117.90° 117.90°
692 a 0.9900 ∠ − 0.9900 ∠ − 0.0000 0.01
5.30° 5.31°
b 1.0529 ∠ − 1.0529 ∠ − 0.0000 0.00
122.34° 122.34°
c 0.9778 ∠ 0.9777 ∠ −0.0001 −0.01
116.03° 116.02°
675 a 0.9835 ∠ − 0.9835 ∠ − 0.0000 0.01
5.55° 5.56°
b 1.0553 ∠ − 1.0553 ∠ − 0.0000 0.00
122.52° 122.52°
c 0.9759 ∠ 0.9758 ∠ −0.0001 −0.01
116.04° 116.03°
684 a 0.9881 ∠ − 0.9881 ∠ − 0.0000 0.00
5.32° 5.32°
c 0.9758 ∠ 0.9758 ∠ 0.0000 0.00
115.92° 115.92°
611 c 0.9738 ∠ 0.9738 ∠ 0.0000 0.00
115.78° 115.78°
652 a 0.9825 ∠ − 0.9825 ∠ − 0.0000 0.01
5.24° 5.25°

Table 3 Fundamental power loss for case study 1

Phase Calculated power loss IEEE results [20]
Active Reactive Active Reactive
(kW) (kVAR) (kW) (kVAR)
a 39.13 152.62 39.107 152.585
b −4.74 42.27 −4.697 42.217
c 76.59 129.69 76.653 129.850
Total 110.98 324.57 111.063 324.653
144 R. Satish et al.

Table 4 Harmonic power

Harmonic order Harmonic power loss
loss in case study 2
Active Reactive
(kW) (kVAR)
3 0.7958 6.5165
5 0.0856 1.1483
7 0.0072 0.1183
9 0.0043 0.0902
11 0.0008 0.0164
13 0.0008 0.0226
15 0.0010 0.0340
Total harmonic loss 0.8983 7.9464
Total power loss 148.23 441.49

found to be 6. In case-B of case-3, in which both DGs are synchronous based, the
maximum THD% is observed to be 5.2442 at bus 611 for c-phase which is more than
THD % reported in case-2. The number of phases in buses with THD % more than
5 is found to be 6. Figure 10 compares the fundamental voltage profile between case
2 and case 3. From Fig. 10, the improvement in voltage profiles is observed with
the integrations of renewable DGs as in case 2. Figure 11 compares the THD% of
case-A in case-3, case-B in case-3 with the THD % of case 2 and it is observed from
Fig. 11 that, the THD% is more for case-A, in which the renewable DGs are power
electronic based.

8 Case Studies and Discussions on IEEE-34 Bus Feeder

The date for IEEE-34 bus feeder is taken from [20]. The convergence tolerance is
10–4 . The base values selected for the system are 2500 kVA and 24.9 kV. The load
composition at spot loads for harmonic analysis is presented in Table 8. The current
spectra of the non-linear loads are taken from [21]. The description of different case
studies on IEEE-34 bus feeder is presented in Table 9. The test results of the case-1
are used as benchmark to see the fundamental and harmonic impacts of renewable
DGs on the network.
Figure 12 compares the fundamental voltage profile between case-1 and case-2.
From Fig. 12, improvement in voltage profile is absorbed with the integrations of
renewable DGs. Figure 13 compares the THD % between the case studies 1 and 2.
From Fig. 14, the maximum THD % is observed in case-A of case-2, which has
integrations of two power electronic based-DGs. From Fig. 15, it is observed that the
number of phases effected with THD% more than 5 is 19 for case-A of case-2 and
zero for case-B of case-2. From Figs. 13, 14 and 15 it is seen that, the integrations
of renewable synchronous based-renewable DGs reduces the THD%, reduces the
Evaluation of Algorithms for Fundamental and Harmonic Impacts … 145

Table 5 Harmonic power flow solution for case study 2

Bus Phase S. No. Fundamental r.m.s voltage Total r.m.s voltage THD %
650 a 1 1 ∠ 0° 0 0
b 2 1 ∠ −120° 0 0
c 3 1 ∠ 120° 0 0
632 a 4 0.9498 ∠ −2.7462° 0.9500 1.9173
b 5 0.9839 ∠ −121.6817° 0.9839 0.4974
c 6 0.9300 ∠ 117.8000° 0.9302 2.2737
671 a 7 0.9109 ∠ −5.8987° 0.9117 4.0623
b 8 0.9875 ∠ −122.2091° 0.9875 1.0363
c 9 0.8717 ∠ 115.9500° 0.8728 4.9409
680 a 10 0.9109 ∠ −5.8987° 0.9117 4.0623
b 11 0.9875 ∠ −122.2091° 0.9875 1.0363
c 12 0.8717 ∠ 115.9500° 0.8728 4.9409
633 a 13 0.9466 ∠ −2.8223° 0.9468 1.9098
b 14 0.9819 ∠ −121.7315° 0.9819 0.4919
c 15 0.9271 ∠ 117.7946° 0.9273 2.2648
634 a 16 0.9207 ∠ −3.6073° 0.9209 1.8801
b 17 0.9624 ∠ −122.2445° 0.9624 0.4873
c 18 0.9064 ∠ 117.2178° 0.9066 2.2406
645 b 19 0.9745 ∠ −121.8646° 0.9745 0.4991
c 20 0.9283 ∠ 117.8225° 0.9286 2.2769
646 b 21 0.9729 ∠ −121.9382° 0.9729 0.5000
c 22 0.9264 ∠ 117.8696° 0.9267 2.2815
692 a 23 0.9109 ∠ −5.8987° 0.9117 4.0623
b 24 0.9875 ∠ −122.2091° 0.9875 1.0363
c 25 0.8717 ∠ 115.9500° 0.8728 4.9409
675 a 26 0.9025 ∠ −6.0795° 0.9034 4.3128
b 27 0.9887 ∠ −122.3037° 0.9887 1.0491
c 28 0.8678 ∠ 116.0660° 0.8689 5.0687
684 a 29 0.9093 ∠ −5.9502° 0.9100 4.0765
c 30 0.8684 ∠ 115.9163° 0.8695 5.0741
611 c 31 0.8651 ∠ 115.8365° 0.8663 5.2263
652 a 32 0.9041 ∠ −5.8755° 0.9049 4.0900
146 R. Satish et al.

Table 6 Harmonic power loss for case-A and case-B in case study 3
Harmonic order Harmonic power loss
Case-A in Case-3 Case-B in Case-3
Active (kW) Reactive (kVAR) Active (kW) Reactive (kVAR)
3 0.7529 6.1520 0.8228 6.7138
5 0.2015 2.5941 0.0913 1.2239
7 0.0835 1.3267 0.0081 0.1341
9 0.0041 0.0857 0.0048 0.1018
11 0.0169 0.3981 0.0009 0.0186
13 0.0193 0.5764 0.0009 0.0259
15 0.0009 0.0322 0.0011 0.0404
Total harmonic loss 1.0791 11.1652 0.9299 8.2585
Total power loss 69.93 197.41 69.78 194.50

number of phases in buses with THD% less than 5 and reduces the maximum THD
% on the network and the integrations of power electronic based-renewable DGs
increases the THD %, increases the number of phases in buses with THD% less
than 5 and increases the maximum THD % on the network with respect to case-1.
Figure 16 compares the total power loss in different case studies.

9 Conclusion

The impacts of renewable DG on fundamental voltage, fundamental power loss and

combined harmonic impacts of renewable DG and non-linear loads on harmonic
voltage profile, harmonic power loss and THD % are analyzed with the proposed
FPFA and HPFA. The proposed algorithm is accurate and easy to understand and
effectively handling all types of renewable DG models. The use of BUS_NUM and
BRANCH_NUM matrices makes the PFAs simple. The integration of synchronous
based-renewable DG the network can reduce the harmonic pollution on the network
in terms of THD% at different buses of the network, number of phases in buses with
THD% more than 5 and maximum THD% on the network.. However the improve-
ment in fundamental voltage profile and reduction in power loss can be achieved
with any type of renewable DG. The location, type and size of renewable DGs are
the three main factors that can affect the power quality and this can limit the amount
of renewable DG penetration. In the case of a utility owned DG installation the utility
has to optimally plan the location and size of the DG in order to improve network
benefits and reliability. In practice, one is not always able to site DG at the loca-
tions determined by an optimization algorithm. In the case of customer owned DG
Evaluation of Algorithms for Fundamental and Harmonic Impacts … 147

Table 7 Harmonic power flow solution for case study 3

Bus Ph S. No. Fundamental r.m.s Total r.m.s voltage THD % Total r.m.s voltage THD%
voltage for case-A in for case-B in
case-3 case-3
650 a 1 1 ∠ 0° 0 0 0 0
b 2 1.0000 ∠ −120° 0 0 0 0
c 3 1.0000 ∠ 120° 0 0 0 0
632 a 4 0.9730 ∠ −1.68° 0.9732 2.1494 0.9732 1.9959
b 5 0.9973 ∠ − 0.9973 1.0734 0.9973 0.6219
120.48°
c 6 0.9540 ∠ 118.98° 0.9544 2.7960 0.9543 2.3845
671 a 7 0.9429 ∠ −4.10° 0.9437 4.1935 0.9437 4.0844
b 8 1.0045 ∠ − 1.0046 1.5876 1.0046 1.2144
120.42°
c 9 0.9056 ∠ 117.88° 0.9070 5.4857 0.9067 4.9782
680 a 10 0.9429 ∠ −4.10° 0.9437 4.1935 0.9437 4.0844
b 11 1.0045 ∠ − 1.0046 1.5876 1.0046 1.2144
120.42°
c 12 0.9056 ∠ 117.87° 0.9070 5.4857 0.9067 4.9782
633 a 13 0.9749 ∠ −1.64° 0.9752 2.2474 0.9751 2.0059
b 14 1.0006 ∠ − 1.0006 1.2622 1.0006 0.6309
120.40°
c 15 0.9574 ∠ 119.06° 0.9578 2.9268 0.9577 2.3982
634 a 16 0.9950 ∠ −1.00° 0.9957 3.7010 0.9952 2.0365
b 17 1.0249 ∠ − 1.0254 3.0848 1.0249 0.6456
119.60°
c 18 0.9828 ∠ 119.94° 0.9838 4.5109 0.9831 2.4403
645 b 19 0.9880 ∠ − 0.9880 1.0792 0.9880 0.6240
120.66°
c 20 0.9523 ∠ 119.00° 0.9527 2.8009 0.9526 2.3878
646 b 21 0.9863 ∠ − 0.9864 1.0810 0.9863 0.6250
120.73°
c 22 0.9503 ∠ 119.05° 0.9507 2.8067 0.9506 2.3927
692 a 23 0.9429 ∠ −4.10° 0.9437 4.1935 0.9437 4.0844
b 24 1.0045 ∠ − 1.0046 1.5876 1.0046 1.2144
120.42°
c 25 0.9056 ∠ 117.87° 0.9070 5.4857 0.9067 4.9782
675 a 26 0.9379 ∠ −4.19° 0.9388 4.4480 0.9388 4.3096
b 27 1.0080 ∠ − 1.0082 1.6510 1.0081 1.2267
120.44°
c 28 0.9047 ∠ 118.08° 0.9061 5.6530 0.9058 5.0896
(continued)
148 R. Satish et al.

Table 7 (continued)
Bus Ph S. No. Fundamental r.m.s Total r.m.s voltage THD % Total r.m.s voltage THD%
voltage for case-A in for case-B in
case-3 case-3
684 a 29 0.9411 ∠ −4.15° 0.9420 4.2097 0.9419 4.0965
c 30 0.9023 ∠ 117.84° 0.9037 5.5985 0.9035 5.1023
611 c 31 0.8990 ∠ 117.77° 0.9005 5.7280 0.9003 5.2442
652 a 32 0.9358 ∠ −4.07° 0.9366 4.2311 0.9366 4.1047

Case 2
Case 3
1
P.U Voltages

0.95

0.9

0.85
5 10 15 20 25 30
Serial numbers

Fig. 10 Comparison of fundamental voltage profile between case 2 and case 3 for IEEE-13 bus
feeder

Case 2
Case-A in Case 3
6 Case-B in Case 3

4
THD %

0
5 10 15 20 25 30
Serial numbers

Fig. 11 Comparison of THD % between case 2 and case 3 for IEEE-13 bus feeder

installation, the utility planner should conduct feasibility and assessment study to
evaluate any technical issues resulting from the new installation of customer owned
DG installation.
Evaluation of Algorithms for Fundamental and Harmonic Impacts … 149

Table 8 Load composition of spot loads on IEEE-34 bus feeder

Bus No. Load composition
Non-linear loads Linear loads
Fluorescent light Adjustable speed Composite residential
banks drives loads
830 None None 80 20
844 30 30 30 10
848 30 30 30 10
890 30 None 60 10
860 30 30 30 10
840 30 30 30 10

Table 9 Case studies on IEEE-34 bus feeder

Case study Description
Case-1 • Removed voltage regulators between buses 614 and 650 and 852 and 832
• Removed capacitor banks at buses 844 and 848
• Load composition of spot loads as in Table 9
Case-2 • 3-phase DG is placed at bus 848 is Case-A: Both DGs are Power electronic
modeled as PQ bus with capacity, P = based
150 kW and Q = 99 kVAR for each Case-B: Both DGs are Synchronous
phase based
• 3-phase DG is placed at bus 890 is
modeled as PV bus with P = 130 kW
per phase and three phase reactive
power limits: 50 ≤ Q ≤ 325 kVAR

Fig. 12 Comparison of 1
fundamental voltage profile
between case 1 and case 2
P.U Voltages

0.9
for IEEE-34 bus feeder
Case 2
0.8 Case 1

0.7
10 20 30 40 50 60 70 80 90
Serial numbers

Fig. 13 Comparison of 8
THD % between case 1 and Case-B in Case-2
case 2 for IEEE34 bus feeder 6 Case-A in Case-2
Case-1
THD %

0
10 20 30 40 50 60 70 80 90
Serial numbers
150 R. Satish et al.

Fig. 14 Comparison of
maximum THD % in Maximum THD %
different case studies for
IEEE-34 bus feeder 6.7656
5.5791
4.2619

Case-1 Case-A in Case-2 Case-B in Case-2

Fig. 15 Comparison of
number of phases in buses Number of phases of buses
with THD % more than 5 in 19
different case studies for
IEEE-34 bus feeder

4
0

Case-1 Case-A in Case-2 Case-B in Case-2

Fig. 16 Comparison of total

power loss in different case Active power loss (kW) Reactive power loss (kVAR)
studies for IEEE-34 bus 265.56
feeder 188.15

63.06 49.12 62.39 46.04

Case-1 Case-A in Case-2 Case-B in Case-2

References

1. IEEE Standard 1547-2003, IEEE Standard for Interconnecting Distributed Resources with
Electric Power Systems, pp 1–16. https://doi.org/10.1109/IEEESTD.2003.94285.
2. Cheng CS, Shirmohammadi D (1995) A three-phase power flow method for real-time distri-
bution system analysis. IEEE Trans Power Syst 10:671–679. https://doi.org/10.1109/59.
387902
3. Losi A, Russo M (2005) Dispersed generation modeling for object-oriented distribution load
flow. IEEE Trans Power Deliv 20:1532–1540. https://doi.org/10.1109/TPWRD.2004.838634
4. Moghaddas-Tafreshi SM, Mashhour E (2009) Distributed generation modeling for power flow
studies and a three-phase unbalanced power flow solution for radial distribution systems consid-
ering distributed generation. Electr Power Syst Res 79:680–686. https://doi.org/10.1016/j.epsr.
2008.10.003
Evaluation of Algorithms for Fundamental and Harmonic Impacts … 151

5. Teng J-H, Chang C-Y (2007) Backward/forward sweep-based harmonic analysis method
for distribution systems. IEEE Trans Power Deliv 22:1665–1672. https://doi.org/10.1109/
TPWRD.2007.899523
6. Amini MA, Jalilian A, Behbahani MRP (2019) Fast network reconfiguration in harmonic
polluted distribution network based on developed backward/forward sweep harmonic load
flow. Electr Power Syst Res 168:295–304. https://doi.org/10.1016/j.epsr.2018.12.006
7. Das SR, Mishra DP, Ray PK, Salkuti, A. K. Sahoo (2020) Power quality improvement using
fuzzy logic based compensation in a hybrid power system. Int J Power Electron Drive Syst 11.
https://doi.org/10.11591/ijpeds.v12.i1.pp576-584
8. Hota P, Mishra S, Mishra DP, Salkuti SR (2020) Allocating active power loss with network
reconfiguration in electrical power distribution systems, Int J Power Electron Drive Syst 11.
https://doi.org/10.11591/ijpeds.v12.i1.pp130-138
9. Hota P, Mishra SK, Mishra DP (2020) Loss allocation strategies in active power distribution
networks: a review. Adv Electr Control Signal Syst Sel Proc AECSS 665:889–902. https://doi.
org/10.1007/978-981-15-5262-5_68
10. Hota P, Mishra SK, Mishra DP (2020) A new active power loss allocation method for radial
distribution networks with DGs. Adv Electr Control Signal Syst Sel Proc AECSS 665:501–514.
https://doi.org/10.1007/978-981-15-5262-5_37
11. Kersting WH (2017) Distribution system modeling and analysis, 4th edn. CRC Press
12. Burch R, Cheng G, Hatziadoniu C et al (2003) Impact of aggregate linear load modeling on
harmonic analysis: a common practice and analytical models. IEEE Trans Power Delivery
18:625–630. https://doi.org/10.1109/TPWRD.2003.810492
13. Task Force on Harmonic Modeling and Simulation (1996) Modeling and simulation of the
propagation of harmonics in electric power networks Part I: concepts, models and simulation
techniques. IEEE Trans Power Deliv 11:452–465. https://doi.org/10.1109/61.484130
14. Arrillaga J, Watson NR (2003) Power system harmonics, 2nd edn. Wiley, New York
15. Chen H, Chen J, Shi D, Duan X (2006) Power flow study and voltage stability analysis for
distribution systems with distributed generation. In: IEEE power engineering society general
meeting, pp 1–8. https://doi.org/10.1109/PES.2006.1709198
16. Naka S, Genji T, Fukuyama Y (2001) Practical equipment models for fast distribution power
flow considering interconnection of distributed generators. In: Power engineering society
summer meeting, pp 1007–1012. https://doi.org/10.1109/PESS.2001.970196
17. Nehrir H, Wang C, Shaw SR (2006) Fuel cells: promising devices for distributed generation.
IEEE Power Energy Mag 4:47–53. https://doi.org/10.1109/MPAE.2006.1578531
18. RavikumarPandi V, Zeineldin HH, Xiao W (2013) Determining optimal location and size of
distributed generation resources considering harmonic and protection coordination limits. IEEE
Trans Power Syst 28:1245–1254. https://doi.org/10.1109/TPWRS.2012.2209687
19. Das D, Nagi HS, Kothari DP (1994) Novel method for solving radial distribution networks.
IEE Proc Gen Transm Distrib 141:291–298. https://doi.org/10.1049/ip-gtd:19949966
20. Radial Distribution Test Feeders. http://sites.ieee.org/pes-testfeeders/resources
21. Abu-Hashim R, Burch R, Chang G et al (1999) Test systems for harmonic modeling and
simulation. IEEE Trans Power Deliv 14:579–587. https://doi.org/10.1109/61.754106
22. Yang NC, Le MD (2016) Three-phase harmonic power flow by direct ZBUS method for
unbalanced radial distribution systems with passive power filters. IET Gen Transm Distrib
10:3211–3219. https://doi.org/10.1049/iet-gtd.2015.1368
23. IEEE Standard 519-2014, IEEE recommended practice and requirements for harmonic control
in electric power systems, pp 1–17. https://doi.org/10.1109/IEEESTD.2014.6826459
A Comprehensive Review of Active
Islanding Detection Methods
and Islanding Assessment in a Grid
Connected Solar Based Microgrid

Kumari Namrata, Akshit Samadhiya, and Papia Ray

Abstract In this literature, different islanding and their detection techniques are
overviewed for the power system network consisting of distribution system along
with Distributed Generator (DG). As penetration of renewable energy utilization and
DG are increasing continuously, it is important for a power system engineer to detect
and mitigate any possible occurrence of islanding event. Islanding Detection Methods
(IDMs) are divided as remote and local methods. Remote methods requires commu-
nication system for efficient operation while local methods are again categorised into
three types active, passive and hybrid methods. Active methods are based on direct
interaction with the power system operation via perturbation and passive methods
are based on utilization of local parameters while hybrid methods are the amalga-
mation of active and passive. These IDMs for DG are described and analysed as
per detection time, advantage, disadvantage, Non Detection Zone (NDZ) and power
quality issues. This work will give a broad idea for selecting the better one of IDMs.
Selection of IDMs are based on four relevant performance indices. A better IDM has
a minimum NDZ with lower detection time without degrading the power quality.
Integration of Renewable energy sources can pose technical challenges. Uninten-
tional Islanding may result in issues such as system instability, degradation in power
quality and malfunctioning of protection system. Fast and efficient methods needs to
be developed to prevent unintentional islanding. To achieve better reliability and high
accuracy, different islanding detection methods (IDMs) have been discussed in this
research work. An extensive analysis of IDMs based on different technical aspects is
presented. A techno-economical comparison of IDMs is presented based on recent
trends related to monitoring islanding events. Additionally, a simulation study for a
grid connected Solar based Microgrid is presented to analyze five different islanding

K. Namrata (B) · A. Samadhiya

Department of Electrical Engineering, National Institute of Technology Jamshedpur, Jamshedpur,
India
e-mail: [email protected]
P. Ray
Department of Electrical Engineering, Veer Surendra Sai University of Technology, Burla,
Odisha, India
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 153
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6_7
154 K. Namrata et al.

detection methods under unintentional islanding. Comparison of IDMs is based on

trip signal generation under two different loading scenarios during an islanding event.

Keywords Islanding · Distributed generator · Local parameters · Non detection

zone · Power quality · Photovoltaic · Utility grid

Nomenclature

IDMs Islanding Detection Methods

MG Micro-grid
DG Distributed Generator
NDZ Non Detection Zone
PQ Power quality
P Active Power
Q Reactive Power
CB Circuit Breaker
AI Anti-Islanding
ID Islanding Detection
PCC Point of common coupling
P Active Power mismatch
Q Reactive Power mismatch
t Detection time
Ed Error detection ratio
RID Remote islanding detection
LID Local islanding detection
PID Passive islanding detection
AID Active islanding detection
HID Hybrid islanding detection
TT Transfer trip
SCADA Supervisory control and data acquisition
OUV Over/Under voltage
OUF Over/Under frequency
PJD Phase jump detection
PLL Phase Locked Loop
VU Voltage unbalance
PSV Positive Sequence Voltage
THD Total harmonic distortion
CBSS Circuit breaker switching strategy
AFD Active frequency drift
Cf Chopping fraction
Tz Zero or dead time
SFS Sandia frequency shift
SVS Sandia voltage shift
A Comprehensive Review of Active Islanding Detection … 155

APS Automatic phase drift

IM Impedance measurement
NCCI Negative component of current injection

1 Introduction

Global consumption pattern has seen continual annual growth over the last two
decades and is expected that the demand may even double in coming 10 years [1].
A shift towards renewable based low carbon technologies in power generation has
been observed. Key factors such as technological advancement in semiconductors,
Solar cells and wind turbines, decarbonisation of power sector, efficient power elec-
tronics converters and advanced control strategies allows accelerated deployment of
renewable energy sources. Renewable based Microgrids ensures less transmission
losses and can even improve the power quality and voltage profile of the Grid [2].
However, increasing penetration of renewable based energy system can pose certain
challenges to the power quality, stability and safety of the Grid [3].
Renewable based generation systems or Distributed generations (DGs) as shown
in Fig. 1 can either operate in grid connected mode or isolate itself to provide
power to local loads. However, events related to unintentional islanding of DGs
may occur which needs prompt attention. These events leads to instability, transient

Fig. 1 Power islanding condition

156 K. Namrata et al.

Table 1 Technical requirements and guidelines defined by various for monitoring islanding events
in grid connected DGs [4, 5]
Standards Quality factor tdetection (ms) Allowable frequency Allowable voltage
range (Hz) range (%)
IEEE:1547 1 <2000 49.3–50.5 88–110
IEC:62116 1 <2000 f0 − 1.5 ≤ f ≤ f0 + 88–110
1.5
IEEE:929-2000 2.5 <2000 49.3–50.5 88–110

Islanding Detection Methods

Remote Methods Local Methods

Passive Active Hybrid

Fig. 2 Taxonomy of islanding detection methods

overvoltage, frequency deviation, malfunction of protection system that degrades

the power quality of the system. Certain standards are adopted widely to address
unintentional islanding issues for grid connection of DG units. Some standards are
listed in Table 1 that provides technical requirement for Anti islanding capability of
DG units before integration in the power system network.
For detection of the islanding phenomena different methods are shown in Fig. 2
remote islanding and local islanding methods. Again local islanding methods are
categorised as passive, active and hybrid methods [6]. A detailed assessment and
comparison of various Islanding Detection (ID) methods along with benefits and
limitations are presented for monitoring islanding based activities.
• Considering the future outlook of the electric power industry, wide scale pene-
tration of renewable technologies such as solar, wind, fuel cell etc. in the form of
distributed generation requires efficient converters, adoption of advance control
strategies and latest technological advancements.
• Detection of islanding condition is also an important issue to deal with.
• Due to challenges related to the power quality of the system, progress in the field
of islanding detection have gained momentum in recent years.
• However, very few publications have discussed about the fundamental problem,
comparison between contemporary and advance detection techniques, classifica-
tion of IDMs, performance indices for assessing IDMs, advantages and disadvan-
tages of each IDMs in terms of Non detection zones (NDZ), detection time and
power quality. Hence, a comprehensive analysis on various aspects of islanding
detection is worth to review and compile.
A Comprehensive Review of Active Islanding Detection … 157

The work is segregated into following sections. Section 2 deals with challenges
related to unintentional islanding and its effect on the power system. Section 3
discusses about various performance indices for selecting a better IDM. Section 4
broadly discusses about different types of local, remote, active and Hybrid islanding
detection methods. Section 5 compares different types of IDM based on NDZ,
detection time and power quality.

2 Certain Challenges with Islanding

Unintentional islanding into the system leads a big concern since integrated
power distribution networks was implemented. Such integrated system together
with inverter or non-inverter-based DG systems, different load conditions, and
compounded control strategy give rise to many challenge to power system network
reliability [7]. There are numerous IDMs presented in research paper, which uses
different tolls for effective capacity building, tools name as complex computing and
signal analysis [8]. Though, still some problems remain that need to be examined
in the field of islanding detection, example as performance of IDMs for concur-
rent operations, interaction between different DG systems which are fitted with AI
devices, when islanding occurs. Tackling mentioned challenges and highlighting the
responsiveness of the AI devices, IDMs are analytically analysed in this review. If an
island of the power system is created unintentionally, it causes trouble and serious
harms. Some of the important harms are discussed as.
• The voltage and frequency may get affected and not remain in their appropriate
standard level.
• Line staff safety may be vulnerable by DGs feeding into a system after opening
and tagging of primary sources.
• Instant reclosing may cause phase difference. This causes huge torques mechan-
ically and huge difference in current (I) /voltage (V), might shatter the physical
systems [9]. Sharp rise leads to devastation of the appliances also.

3 Indices / Parameters of Islanding Detection Methods

The objective of the islanding detection techniques is to monitor certain system

parameters such as rate of change of frequency, Voltage, current, P, Q and accord-
ingly determine any possible occurrence of grid disconnection based on the threshold
value of the parameters. The salient characteristics of an effective and reliable IDM
are listed below:
• Minimum NDZ
• Minimum detection time
158 K. Namrata et al.

• High Power quality (lower value of THD)

• Error detection ratio close to 1 i.e. minimum events of false detection.
Hence, IDMs can be assessed on the basis of four relevant performance indices
namely Non detection Zone (NDZ). Detection time (t), power quality and error
detection ratio (Ed ).

3.1 Non Detection Zone (NDZ)

NDZ became the primary cause for failing of IDMs. NDZ methods is build up by
observing of V, I and f deviation that lead to power mismatch. Power mismatch
between output of DG and load utilization when distributed generator operates in
islanding condition are due to fluctuation of V and f at a point of common coupling
(PCC). When P and Q are almost to zero, at that time V and f fluctuation are
not enough to detect the islanding. The range of power mismatch (P, Q), which
can’t cause V and f crossing the threshold set value to detect islanding is known non
detection zone (NDZ) [6, 10]. NDZ in power mismatch space is shown by [11].
 2  2
V P V
−1≤ ≤ −1 (1)
Vmax P Vmin
  2    2 
f Q f
Qf 1− ≤ ≤ Qf 1− (2)
f min P f max

where [Vmin , Vmax ] is the allowable voltage range, fmin is minimum frequency and
fmax is maximum frequency, V and P are rated voltage and active power, and Qf is
the quality factor. NDZ region is shown in Fig. 3. NDZ stands “non-detection zone”
the range of conditions where a real grid failure will be filtered out.
NDZ is a very useful feature used for evaluating and examining the effectiveness
of the islanding detection technique. Figure 3, shows P versus Q plane and this
power imbalance happen due to V and f variations. Out of the shaded region AI
device sense the islanding condition and thus disconnect the DG but this would not
be possible in case of power difference came within power mismatch space and thus
it became threat for the safety unit.

3.2 Detection Time

Time interval between the disconnecting of the main grid and detecting islanding b
IDMs that is expressed as
A Comprehensive Review of Active Islanding Detection … 159

ΔQ (Reactive Power Imbalance)

OF

Shaded region is NDZ

UV OV
ΔP (Active Power Imbalance)

UF

Fig. 3 Power mismatch space or NDZ region

t = t I d M − ttri p (3)

where t is the detection time, t I d M is the instant at which islanding detected and
ttri p is the instant at which circuit breaker of the main grid get trip and isolate the
DGs.

3.3 Power Quality

For the power quality, several aspects have to be taken in consideration like voltage,
current and frequency stability as well as continuous supply of power and there
waveform. Due to all these concerns it doesn’t have the particular definition and
thus it is also known as service quality, voltage and current quality but judging the
all parameters it can defined as the calculation, analysis, and betterment of the line
parameters to maintain a reference waveform.

3.4 Error Detection Ratio

It is related to the false detection of islanding by IDMs whenever utility grid and micro
grid (MG) are interconnected. Error measurement caused primarily by disturbance,
which results in measured units exceeding the set points [12]. It is described as

Nerr or detected
Ed = (4)
Nerr or detected + Ncorr ect detected
160 K. Namrata et al.

where E d is error detection ratio, Nerr or detected is count of time false detection and
Ncorr ect detected is count of time correct detection.

4 Islanding Detection Methods

Various detection methods are being identified the researchers across the world. Typi-
cally, these can be classified into remote islanding detection (RID) and local islanding
detection (LID) methods. RID methods are based on measurement and calculation
of technical parameters at the central side [13–18], while LID methods are based
realization at DG side. LID methods can be further divided into passive islanding
detection (PID) [19–53], active islanding detection (AID) and hybrid islanding detec-
tion (HID) methods as shown in Fig. 2. In this work, Active Islanding [54–64] and
Hybrid Islanding techniques [57, 65, 66] are extensively reviewed. Passive detection
techniques are discussed in our recent review article [67].

4.1 Active Islanding Detection Methods

In these AID methods, intentionally injecting small perturbation at the output end of
the DGs, that injecting signal do some significant changes in the system parameters
and if that changes exceed the predefined value then trip signal get activated of
the islanding condition and thus DG get disconnected. Remarkably AID methods
have advantage as their low or zero NDZ, but major demerit is that these methods
deteriorate the power quality. The basic flow chart of AID methods is shown in Fig. 4,
and then different AID methods discussed briefly.

4.1.1 Active Frequency Drift (AFD)

Electrical parameters voltage and current are associated with the magnitude,
frequency and phase. AFD uses the frequency/phase related parameters. In this
method a small disruption current is injected into the main system, and whenever
the islanding phenomena occurs, frequency drift take place by adding a zero time
in inverter current and make the inverter current waveform distorted with respect to
the original waveform i.e., shown in Fig. 5, however it is not affected during the grid
connected system. Distortion meaning is phase difference between inverter output
waveform (current) and original waveform of utility voltage at PCC. Due to distor-
tion, inverter detect phase error, and participate in drifting the DG/inverter current to
make the phase error zero. To make the drift occurs, DG connected inverter should
work at unity power factor, and frequency drift is compared with OUF relay for
islanding to be detected. For this a parameter is defined name as chopping fraction
(cf ) shown in Eq. (5).
A Comprehensive Review of Active Islanding Detection … 161

Start

Inject perturbation at PCC

Measure V,I,Hz,THD,P,Q etc….At PCC(A)

No
A > Threshold

Yes

Trip Signal
(islanding detected)

To Breaker (End)

Fig. 4 Flow chart of AID methods

Iout(A)
Original Waveform
Distorted Current Waveform

t(s)

TI/2
TZ
TV/2

Fig. 5 Operational waveform of AFD method [54]

TZ
cf = (5)
TV /2

where, TZ = zero or dead time and TV = time period of utility voltage. It is easy in
implementation and having small NDZ compare to novel PID methods. Drawbacks
162 K. Namrata et al.

are, effective for purely resistive load, as NDZ became large during large values
of capacitance (C) and inductance (L), also wouldn’t suitable for parallel inverter
systems, and have issue with the power quality [54].

4.1.2 Active Frequency Drift with Positive Feedback (AFDPF)

For supressing drawbacks of AFD (multi-inverter and NDZ issue), a positive feed-
back is used along with AFD. This feedback amplify the chopping fraction (cf ) or
dead time and assist the frequency drift at higher rate. Due to increase in the rate
of frequency detection, it leads to detect the islanding condition fastly compared to
AFD.

c f K = c f K −1 + G(ωk ) (6)

where c f K and c f K −1 are kth chopping fraction and (k − 1)th chopping fraction,
ωk = frequency difference = ω K −1 − ω0 , G is positive gain constant. No matter cf
is positive or negative, AFDFP reinforced the frequency drift for any load, but the
power quality degradation is still continued due to injected disturbance [55].

4.1.3 Sandia Frequency Shift (SFS)

This method is also an add-on of AFD, using a positive feedback for the frequency
of the output voltage at the inverter end, whose chopping fraction is expressed as

c f = c f 0 + K ( f PCC − fline ) (7)

where, c f 0 is chopping fraction without frequency variation, K is accelerating gain,

f PCC is frequency of voltage at PCC and fline is the frequency of the original line.
During normal operation DG sets try to change the frequency but system stability
maintain the same frequency, but during utility grid disconnected mode, there is rise
in the frequency attempted at PCC and thus increase the chopping fraction. Whenever
the frequency at the inverter end crosses the set value then islanding is detected. SFS
method has very small NDZ, as well as detecting efficiency and power quality is
improved [56].

4.1.4 Sandia Voltage Shift (SVS)

This method uses the same principle as SFS, just change is that it uses a positive
feedback of magnitude of voltage at PCC and changes the current and power. When
grid is linked there is no effect shown of this positive feedback, but during open linked
voltage is reduced at PCC and thus current and power reduced. For that reduction
A Comprehensive Review of Active Islanding Detection … 163

of voltage a UV relay is used for the protection. Possibility is also of rising of

voltage, and hence OUV relay is used, if the PCC voltage reaches OUV standards
then islanding is detected [57]. Among all AID methods this method is quite easy in
implementation, having small NDZ but facing the problem of power quality issue.

4.1.5 Slip Mode Frequency Shift (SMFS)

Practice is carried out by applying positive feedback to the phase of the PCC voltage.
A small perturbation in the phase will cause frequency deviation during the grid
disconnected, but a small disturbance in phase during grid connected mode will not
lead to frequency deviation, hence phase difference in normal operation is almost
zero as unity power factor maintain by the inverter. A SMFS curve Fig. 9, using
Eq. (8).
  
π f (k−1) − f 0
θ S M F S (k) = θm sin (8)
2 ( fm − f0 )

where θm is maximum phase deviation which occur at maximum frequency f m and

f (k−1) is previous cycle frequency and f 0 is rated frequency. Equation (8) is showing
the relation between frequency and phase angle. Working principle also validate that
slope of the SMFS line is greater than the load line during unstable region. When
grid get disconnected, operation will move towards the new stable point (shown by
dotted as f 1 and f 2), during going through unstable to stable a set threshold value of
inverter frequency crossed in either direction and thus islanding get detected by OUF
relay. Problem with this method is having large NDZ and quality of power issue [58].

4.1.6 Automatic Phase Drift (APS)

This method is extension of SMFS, used to rectify the problem of NDZ of AFD and
SMFS methods by utilizing positive feedback to the phase angle of output current
of inverter. Hence only starting angle of inverter current alter consistently by using
previous frequency of voltage, expression is shown as
 
1 f (k−1) − f o
θ A(k)P S = 360◦ + θ0(k) (9)
α fo

where θ A(k)P S is beginning angle of inverter I, α is constant and θ0(k) is additional phase
shift induced each time in the terminal voltage.

θ0(k) − θ0(k−1) = θ ∗ sgn( f s ) (10)

164 K. Namrata et al.

where θ = constant,  f s = change in steady state frequency, sgn = sign or signum

function i.e., defined as:
⎧
⎨ 1 i f fs > 0
sgn( f s ) = 0 i f fs = 0 (11)
⎩
−1 i f f s < 0

Due to additional phase shift, frequency deviate of the voltage to stabilizing to

new operating point, and during islanding this deviation of the frequency leads the
OUF standards, thus islanding get detected. Disadvantages of this method are speed
is slow and not suited for nonlinear load [59].

4.1.7 Variation in Active and Reactive Power

Method is based on the capacity of inverter to produce both active power and reactive
power independently. During the utility disconnected mode, DG has to supply active
power, and depending upon the load condition, voltage will rise or fall as voltage is
related directly to the active power. When the change in the voltage exceeds the set
threshold value, then islanding condition is detected by the help of OUV relay.
Similarly reactive power cause frequency variation, and it’s detected by the OUF
relay. The detection time is around 0.3–0.75 s, easy in implementation and has small
NDZ. Weaknesses are it can’t support multi-inverter as false detection sometime
raised, power quality issue continued [60].

4.1.8 Negative Component of Current Injection (NCCI)

This method is involving the use of small injection of negative component of current
(less than 3%) to the voltage source inverter (VSI) at the PCC terminal, leads to
disturbing the PCC voltage as unbalancing of the voltage occurs during the grid
disconnected. In normal mode NCC flow into grid without affecting the PCC voltage.
Thus islanding get detected if negative sequence voltage exceeds the threshold value
within 60–70 ms. Method is beneficial for any load, zero NDZ and detecting fast
compare to positive sequence voltage [61, 62]. This method can also use to detect
islanding by negative sequence impedance as it’s the ratio of negative component of
voltage and current.

4.1.9 Impedance Measurement (IM)

Ratio of rate of change of voltage (ROCOV) to the rate of change of current (ROCOC)
of the inverter gives the impedance. This (dv/di) technique is used for the detection
as during normal operation variation in the voltage is very less and hence impedance is
A Comprehensive Review of Active Islanding Detection … 165

low, but during grid disconnected mode voltage varies significantly by perturbation
in the current and thus impedance crosses the threshold value and islanding get
detected. Detection time is around 700–900 ms [63]. It’s advantage is that NDZ is
low and work effectively in case of single inverter and synchronously connected multi
inverter, but all time synchronously would not be possible and thus multi inverter is
became demerit for it.

4.1.10 Impedance Detection at Specific Frequency

This technique is exceptional case of harmonic detection method. A perturbation of

the special frequency harmonics of the current are injected in the inverter. Due to this
abruptly change occurs in the voltage of inverter placed at PCC terminal when the
utility is disconnected, but not significant change during utility connected to the DGs.
Due to production of harmonic voltage in the system, impedance can be measure
and if it crosses more than the set point, then islanding detected. It is not appropriate
for the multi inverter system [64].

4.2 Hybrid Islanding Detection Methods

As passive and active methods are suffering from large NDZ and power quality issue
respectively. To eradicate both the problems, a HID methods are used, that combine
both PID and AID methods. In this case AID methods apply after the use of the PID
methods as shown in the flow chart of Fig. 6. Some of the HID methods are discussed
in brief as the following.

4.2.1 Voltage Unbalance (VU) and Frequency Set Point

As VU (passive) is more sensitive to the load fluctuation, so instead of VU/THD

(passive), this method uses VU as a passive parameter. Other active parameter should
be needed that is completed by positive feedback to the voltage or current that leads to
deviation in the frequency by using SFS/SVS. Whenever some changes in the system
occur the voltage spike rises suddenly, and from precaution of the false detection due
to load switching or transient, a maximum VU is set for their threshold i.e., around
35 times of the VU average as per the Menon and Nehrir proposal [65]. Voltage
spike also leads to frequency deviation, if the frequency change in a specified time is
greater than the threshold value then islanding condition get detected. Technique have
advantages as very small NDZ, very low power quality issue and clearly discriminate
islanding and non-islanding conditions.
166 K. Namrata et al.

Start

Measure Passive Parameters (A)

A>
Max. Threshold

YES
NO

Max. Threshold NO
>A>
Min. Threshold

YES
Inject Perturbance to PCC

A> NO
Threshold

YES
Islanding Detected

Send signals to DG and local loads

Stop

Fig. 6 Flow chart of HID methods

4.2.2 Voltage Fluctuation Injection

This method consist of ROCOF/ROCOV (passive) and chopping fraction, CF (active)

method. As shown in the HID flow chart, if the rate of change of F/V is greater than the
large (maximum) setting value of threshold then islanding is confirmed directly, and
when the ROCOF/V is greater than the small (minimum) setting value of threshold
then immediately AID method get activated [57]. In AID chopping frequency (CF)
is checked, and if CF value is also more to their corresponding threshold value then
islanding detected, otherwise it would be considered as non-islanding case.

4.2.3 SFS and Q-f (ROCOF) Technique

This method is based on the integration of the Sandia frequency shift, SFS (active)
and ROCOF (passive) technique [66]. For large power variation, islanding confirmed
directly by ROCOF method only, but if the variation is in between the large and small
A Comprehensive Review of Active Islanding Detection … 167

then active method (SFS) will come into the action and by perturbation the islanding
can be detected. Similarly many HID methods (like ROCOV and P, ROCOF and
IM, ROCOF and frequency injection) can be formed by implementing the integrated
passive and active IDMs.

5 Comparison of IDMs

Key technical factors such as (NDZ), detection time and power quality were consid-
ered for comparing RID, PID, AID and HID methods. A fair comparison of widely
adopted IDMs has been presented in in Table 2. Each method has its own pros and
cons which are listed in Table 3. Selection of IDMs are dependent on number of
factors such as system power rating, network loading conditions, cost effectiveness,
protection devices and integration guidelines and allowable limits. In terms of cost,
PID has advantages over the others. RID requires high initial and maintenance costs.

6 System Architecture of a Grid Connected Solar Based DG

Figure 7 shows a Solar photovoltaic based DG connected to the 110 kV Utility

grid. A DC-DC boost converter stabilizes the output of the PV arrays at a constant
DC voltage of 500 V. A 3 level DC/AC power converter is used to process the
power to the local loads and the main grid. CB1 and CB2 are the breakers for 20 and
110 kV bus respectively. Two local loads are connected to 20 kV feeder. Five different
islanding detections methods are modelled to provide a trip signal during islanding
events. They are Over/Under Voltage IDM, Over/Under Current IDM, Over/Under
Frequency IDM, modified ROCOF IDM and Vector Shift IDM.

7 Results and Discussion

Five different IDMs are modelled and tested on a grid connected Solar photovoltaic
based Distributed Generation system in MATLAB/SIMULINK environment. The
PV inverter is connected to a 20 kV Bus through a 3- 100 kVA /Y transformer.
Circuit breaker CB1 connects the two local loads to the 20 kV Bus. The DG units and
local loads are connected to 110 kV, 2500 MVA, 50 Hz Utility Grid through a circuit
breaker CB2. Islanding event is achieved by opening CB2 at t = 0.1 s. Unintentional
islanding usually pose power quality issues when the local generation capacity is
adequately not enough to supply the local consumer demands. Hence, two loading
scenarios have been considered for the simulation study. Trip Signal is generated on
successful detection of islanding event in two cases described below:
168 K. Namrata et al.

Table 2 Technical comparison of different IDMs based on three key performance metrics
Method’s name (RID) NDZ t (ms) PQ
PLC Nil 200 No effect
TT Nil – No effect
SCADA Nil Slow speed in busy grid No effect
Method’s name (PID) NDZ t(ms) PQ
OUV/OUF Large 4 to 2000 No effect
ROCOF Large 24 No effect
ROCOP <than OUV/OUF >One cycle No effect
ROCOFOP <ROCOF 100 No effect
ROCOFORP Very small – No effect
PJD Large 10–20 No effect
VU Large 53 No effect
THD (V/I) Large, (Q high) 45 No effect
VU AND THD Small 25 to 2000 No effect
ROCOV and CPF Small 35 No effect
ROCONSV (ac) Nil 80 No effect
ROCOPSV (ac) Nil 10 No effect
ROCO (PSV and PSC) Nil 10 No effect
ROCOEVORP Nil – No effect
ROCOEV with CBSS Nil 100 to 300 No effect
PANSVNSC Nil Within 4.16 No effect
FHO Very small Upto 440 No effect
Method’s name (AID) NDZ t (ms) PQ
AFD Very small Within 2000 Degrade
AFDPF Very small <AFD Slight degrade
SFS Very small 500 Degrade
SVS Small – Slight degrade
SMFS <AFD 400 Spike in system
APS Small Small Degrade
NCCI Zero 60 –
IM Small 700–900 Slight degrade
Method’s name (HID) NDZ t (ms) PQ
VU and SFS/SVS Very small – Slight degrade
ROCOV and P Small Within 2000 Slight degrade
ROCOF and IM Small 216
A Comprehensive Review of Active Islanding Detection … 169

Table 3 Merits and Demerits of Different Islanding Detection Methods

Islanding detection methods Merits Demerits
Remote islanding methods • Highly reliable • Quiet expensive especially for
(RID) low-medium power Microgrid
Passive islanding methods • Lower detection time • Less effective in low loading
(PID) • Does not disturb the system conditions
• Reliable in extreme loading • Requires attention while
conditions deciding the threshold levels
Active islanding methods • Reliable even when • Perturbs the system
(AID) generation matches the • Slow response
demand • System stability may disturb
• Small NDZ under perturbations
Hybrid islanding methods • Injects disturbances only • Detection time is more
(HID) when islanding is suspected
• Smaller NDZ
• Power quality better than
active islanding case

Fig. 7 Grid connected solar photovoltaic based DG under study

• Local generating capacity is either equal to the local load requirement

• Local generating capacity is less than the local load requirement
In this study, Over/Under voltage, Over/Under current, Over/Under Frequency,
modified ROCOF and vector shift methods are examined.
170 K. Namrata et al.

7.1 Case 1: Local Load is Equal to Local Generation

This test is performed at the 20 kV feeder where local loads are connected. Load 1
and Load 2 demands 50 kW each.
For the performing the islanding test, circuit breaker (CB2) of the main grid is
opened at the instant of 0.1 s and reclose at instant of 0.25 s. Within the time interval
of CB2 open condition (islanding condition) many parameters are changed, and if
any parameters are crossed the corresponding threshold value then trip signal is
generated for the removal of DG.
As seen in Fig. 8, when the islanding events occurred, suddenly voltage and
current parameters are affected. In Fig. 9 change of frequency and modified ROCOF
are observed. RMS Value of Phase voltages can be observed in Fig. 10. Trip signal is
generated when the selected parameters crosses the set threshold values. Islanding is
detected by OUV, OUC, OUF, m-ROCOF and VS within 10 m, 15 ms, 65 ms, 25 ms
and 25 ms respectively as shown in Fig. 11. During reclosing of the load at 0.25 s, a
transient is seen which is suppressed within few cycles.

20
Voltage (kV)
Three Phase

-20

-40
0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Time (Sec)

20
Three Phase
Current (A)

-20

0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40

Time (Sec)

150
Active Power
Three Phase

Islanding Condition
100
(kW)

50
0
CB2 diconnects the DG from
-50 CB2 reconnects the DG to the
the main Grid
main Grid
0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Time (Sec)

Fig. 8 Variation of three phase voltage, current and active power in case local load equals PV
generation
A Comprehensive Review of Active Islanding Detection … 171

52
Frequency (Hz)
Islanding Condition
51.5
51 CB2 diconnects the DG from
the main Grid CB2 reconnects the DG to the
50.5
main Grid
50

0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40

Time (Sec)
15
Rate of change

10
of frequency
(Hz/Sec)

5
0
-5
-10
-15
0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Time (Sec)

Fig. 9 Variation of frequency (50 Hz), and ROCOF in case local load equals PV generation

20
Phase A RMS
Voltage ( kV)

15
10 CB2 diconnects the
DG from the main Grid CB 2 reconnects the DG to the
05 main Grid
00
0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Time (Sec)

20
Voltage ( kV)
Phase B RMS

15
10
05
00
0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Time (Sec)

20
Voltage (kV)
Phase C RMS

15
10
05
00
0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Time (Sec)

Fig. 10 Variation of Variation in RMS value of voltage Va, Vb and Vc in case local load equals
PV generation

7.2 Case 2: Local Load Exceeds Local Generation

This test is also performed at the 20 kV feeder where local load are connected as
per the specification: load 1 requires 2 MW and load 2 requires 30 MW and 2
MVAR (lagging) respectively. For performing the islanding test CB2of the main
172 K. Namrata et al.

Trip Signal Trip Signal for Over/Under Voltage based islanding detection

1.0
0.5
0
0.1 0.15 0.20 0.25 0.30 0.35 0.40
Time (Sec)
Trip Signal for Over/Under Current based Islanding detection
Trip Signal

1.0
0.5
0
0.1 0.15 0.20 0.25 0.30 0.35 0.40
Time (Sec)
Trip Signal for Over/Under Frequency based Islanding detection
Trip Signal

1.0
0.5
0
0.1 0.15 0.20 0.25 0.30 0.35 0.40
Time (Sec)
Trip Signal for modified ROCOF based Islanding detection
Trip Signal

1.0 Islanding Condition

0.5
0
0.1 0.15 0.20 0.25 0.30 0.35 0.40
Time (Sec)
Trip Signal for Vector Shift based Islanding detection
Trip Signal

1.0 CB2 diconnects the

0.5 DG from the main Grid CB2 reconnects the DG to the
0 main Grid
0.1 0.15 0.20 0.25 0.30 0.35 0.40
Time (Sec)

Fig. 11 Islanding detection (trip signal) in case local load equals PV generation

grid is opened for some time i.e., from 0.1 to 0.25 s. Within this period, electrical
parameters crosses the set threshold value then islanding get realized. In Fig. 12,
during islanding events, sudden change of voltage and current magnitude is seen, and
the change crosses the threshold value, hence islanding get detected. The detection
time of the OUV is around 7 ms, while detection time of over/under current is around
12 ms, i.e., the fast detection corresponding to the previous case discussed. Variation
in active power is also shown in Fig. 12, i.e., sudden dip of power during islanding
event. PV based DG is unable to supply the load resulting in load shedding.
In Fig. 13 change of frequency and modified ROCOF are seen and islanding is
detected with the corresponding detection time as 54 ms and 65 ms respectively. RMS
value of Phase voltages can be observed in Fig. 14. A vector shift is also detected
A Comprehensive Review of Active Islanding Detection … 173

20
Voltage (kV)
Three Phase
0

-20

0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40

Time (Sec)
Three Phase
Current (A)

20
0
-20

0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40

Time (Sec)
Active Power

150 Islanding Condition

Three Phase

100
(kW)

50
CB2 reconnects the DG to the
0 main Grid
CB2 diconnects the DG from
-50 the main Grid
0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Time (Sec)

Fig. 12 Variation of three phase voltage, current and active power in case local load exceeds PV
generation

50 Islanding Condition
Frequency (Hz)

49.8
49.6 CB2 diconnects the DG from CB2 reconnects the DG to the
49.4 the main Grid main Grid
49.2
49
48.8
0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Time (Sec)
15
10
Rate of change
of frequency

5
(Hz/Sec)

0
-5
-10
-15
0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Time (Sec)

Fig. 13 Variation of frequency (50 Hz), and modified ROCOF in case local load exceeds PV
generation
174 K. Namrata et al.

Fig. 14 Variation in RMS value of voltage Va, Vb and Vc in case local load exceeds PV generation

having the detection time is equal to 64 ms. Trip signals are generated on successful
islanding detection using OUV, OUC, OUF, m-ROCOF and VS as shown in Fig. 15.
Out of these five, three methods OUV, over/under current and OUF are based
on the magnitude comparison of voltage, current and frequency. It is observed that
the fast detection is done by OUV and followed by over/under current but problem
with these are having large false detection. So, further methods modified ROCOF
and voltage vector shift are observed and found that, it eliminate the false detection
condition to somehow. Hence, accuracy is improved in both methods as two passive
parameters are checked there and hence detection time increased. Detection time and
non-islanding conditions are discussed in Table 4.

8 Conclusion

In this paper, active islanding detection techniques have been described for the utility
connected micro grid system. The IDMs comparison is based on standards listed in
Table 1, mainly IEEE 1547. Remote methods requires noise free, efficient and faster
communication infrastructure, and hence the cost is high but is most reliable, as NDZs
are minimized and prevents degradation of power quality. An active method has very
small NDZ but degrade the power quality and that is why hybrid methods come into
A Comprehensive Review of Active Islanding Detection … 175

Fig. 15 Islanding detection (trip signal) in case local load exceeds PV generation

Table 4 Islanding detection

Technique’s name tdetection (ms) tdetection (ms)
time for each IDM
Case 1 Case 2
OUV 10 1
Over/under current 15 12
OUF 62 54
Modified ROCOF 150 65
Voltage vector shift 150 54
176 K. Namrata et al.

the picture to eradicate the high NDZ of passive and power quality issue of active, by
combining the both methods (passive and active), detection time became prolonged.
Generally passive detection time is low compared to both active (for seeing the
response after perturbation, needed somehow more time) and hybrid. After being
easy installation and lesser cost of PID methods, still it is not chosen because of
their limitation with multi inverter systems and NDZ during balance islanding. In
most of the cases AID are being used, but future trend will be hybrid and RID as
both giving the good result over islanding and non-islanding, and capable of use of
multi inverter systems with good reliability. Hence, this literature is providing a good
concept over islanding detection, and it will be helpful for the future research. A grid
connected architecture of a Solar based Distributed generation system is examined
under islanding events using five IDMs. Modified ROCOF and voltage vector shift
are observed to be more effective than conventional under/over voltage, current and
frequency detections as they eliminates false detection conditions.

References

1. Dutta S, Sadhu PK, Reddy MJB, Mohanta DK (2018) Shifting of research trends in islanding
detection method—a comprehensive survey. Protect Control Modern Power Syst 3(1). https://
doi.org/10.1186/s41601-017-0075-8
2. Acharya N, Mahat P, Mithulananthan N (2006) An analytical approach for DG allocation in
primary distribution network. Int J Electr Power Energy Syst 28:669–678. https://doi.org/10.
1016/j.ijepes.2006.02.013
3. Chiang WJ, Jou HL, Wu JC (2012) Active islanding detection method for inverter-based distri-
bution generation power system. Int J Electr Power Energy Syst 42(1):158–166. https://doi.
org/10.1016/j.ijepes.2012.03.038
4. Basso TS (2014) IEEE 1547 and 2030 standards for distributed energy resources intercon-
nection and interoperability with the electricity grid. National Renewable Energy Laboratory,
Golden, CO, USA, vol 15013. https://doi.org/10.2172/1166677
5. Reddy CR, Reddy KH (2019) Islanding detection techniques for grid integrated DG—a review.
Int J Renew Energy Res 9(2):960–977
6. Li C, Cao C, Cao Y, Kuang Y, Fang B (2014) A review of islanding detection methods for
microgrid. Renew Sustain Energy Rev 35:211–220. https://doi.org/10.1016/j.rser.2014.04.026
7. Vahedi H, Noroozian R, Jalilvand A, Gharehpetian GB (2011) A new method for islanding
detection of inverter-based distributed generation using DC-link voltage control. IEEE Trans
Power Deliv 26:1176–1186. https://doi.org/10.1109/TPWRD.2010.2093543
8. Khamis A, Xu Y, Dong ZY, Zhang R (2018) Faster detection of microgrid islanding events
using an adaptive ensemble classifier. IEEE Trans Smart Grid 9:1889–1899. https://doi.org/10.
1109/TSG.2016.2601656
9. Walling RA, Miller NW (2002) Distributed generation islanding implications on power system
dynamic performance. IEEE Power Eng Soc Summer Meeting 1:92–96. https://doi.org/10.
1109/PESS.2002.1043183
10. Bahrani B, Karimi H, Iravani R (2011) Non-detection zone assessment of an active islanding
detection method and its experimental evaluation. IEEE Trans Power Deliv 26(2):517–525.
https://doi.org/10.1109/TPWRD.2009.2036016
11. Ye Z, Kolwalkar A, Zhang Y, Du P, Walling R (2004) Evaluation of anti-islanding schemes
based on nondetection zone concept. IEEE Trans Power Electron 19(5):1171–1176. https://
doi.org/10.1109/TPEL.2004.833436
A Comprehensive Review of Active Islanding Detection … 177

12. Vieira JCM, Salles D, Freitas W (2011) Power imbalance application region method for
distributed synchronous generator anti-islanding protection design and evaluation. Electr Power
Syst Res 81(10):1925–1960. https://doi.org/10.1016/j.epsr.2011.06.009
13. Mahat P, Chen Z, Bak-Jensen B (2008) Review of islanding detection methods for distributed
generation. In: Third international conference on electric utility deregulation and restructuring
and power technologies, pp 2743–2748. https://doi.org/10.1109/DRPT.2008.4523877
14. Rumbayan M (2017) Development of power system infrastructure model for the island commu-
nities: a case study in a remote island of Indonesia. In: International conference on advanced
mechatronic systems, pp 515–518. https://doi.org/10.1109/ICAMechS.2017.8316470
15. Wang W, Kliber J, Zhang G, Xu W, Howell B, Palladino T (2007) A power line signaling based
scheme for anti-islanding protection of distributed generators: Part II: field test results. IEEE
Trans Power Deliv 22:1767–1772. https://doi.org/10.1109/TPWRD.2007.899620
16. Raza S, Mokhlis H, Arof H, Laghari JA, Wang L (2015) Application of signal processing
techniques for islanding detection of distributed generation in distribution network: a review.
Energy Convers Manage 96:613–624. https://doi.org/10.1016/j.enconman.2015.03.029
17. Kim MS, Haider R, Cho GJ, Kim CH, Won CY, Chai JS (2019) Comprehensive review of
islanding detection methods for distributed generation systems. Energies 12(5):837. https://
doi.org/10.3390/en12050837
18. Xiao H, Fang Z, Li C, Xu D, Venkatesh B, Singh BN (2017) Anti islanding protection relay for
medium voltage feeder with distributed generator. IEEE Trans Ind Electron 64(10):7874–7885.
https://doi.org/10.1109/TIE.2017.2694394
19. Guha B, Haddad RJ, Kalaani Y (2015) A novel passive islanding detection technique for
converter based distributed generation systems. In: IEEE power & energy society innovative
smart grid technologies conference, pp 1–5. https://doi.org/10.1109/ISGT.2015.7131907
20. Zhao J, Zhang D, He J (2017) A passive islanding detection method based on interharmonic
impedance. In: IEEE conference on energy internet and energy system integration, pp 1–6.
https://doi.org/10.1109/EI2.2017.8245574
21. IEEE 929 (2000) IEEE recommended practice for grid interface of photovoltaic (PV) systems.
Institute of Electrical and Electronics Engineers, New York. https://doi.org/10.1109/IEEESTD.
2000.91304
22. Timbus A, Oudalov A, Ho CN (2010). Islanding detection in smart grids. In: 2010 IEEE energy
conversion congress and exposition, pp 3631–3637. https://doi.org/10.1109/ECCE.2010.561
8306
23. Jones RA, Sims TR, Imece AF (1990) Investigation of potential islanding of a self-commutated
static power converter in photovoltaic systems. IEEE Trans Energy Convers 5(4):624–631.
https://doi.org/10.1109/60.63131
24. Gupta P, Bhatia RS, Jain DK (2016) Active ROCOF relay for islanding detection. IEEE Trans
Power Delivery 32(1):420–429. https://doi.org/10.1109/TPWRD.2016.2540723
25. Warin J, Allen WH (1990) Loss of mains protection. In: ERA conference on circuit protection
for industrial and commercial installations, pp 4–3
26. Guha B, Haddad RJ, Kalaani Y (2015) A passive islanding detection approach for inverter-
based distributed generation using rate of change of frequency analysis. In: IEEE South east
conference, pp 1–6. https://doi.org/10.1109/SECON.2015.7133024
27. Ahmad KNEK, Selvaraj J, Rahim NA (2013) A review of the islanding detection methods in
grid-connected PV inverters. Renew Sustain Energy Rev 21:756–766. https://doi.org/10.1016/
j.rser.2013.01.018
28. Aghdam HN, Ghadimi N, Farhadi P, Hashemi F, Ghadimi R (2011) Detecting the anti-islanding
protection based on combined changes of active and reactive output powers of distributed
generations. In: 3rd international conference on computer research and development, Shanghai,
pp 285–289. https://doi.org/10.1109/ICCRD.2011.5764196
29. Reddy CR, Reddy KH (2017) Islanding detection method for inverter based distributed
generation based on combined changes of ROCOAP and ROCORP. Int J Pure Appl Math
117(19):433–440
178 K. Namrata et al.

30. Menon D, Antony A (2016) Islanding detection technique of distribution generation system.
In: International conference on circuit, power and computing technologies, pp 1–5. https://doi.
org/10.1109/ICCPCT.2016.7530126
31. Nikolovski S, Baghaee HR, Mlakic D (2019) Islanding detection of synchronous generator-
based DGs using rate of change of reactive power. IEEE Syst J 13(4):4344–4354. https://doi.
org/10.1109/JSYST.2018.2889981
32. Raza S, Mokhlis H, Arof H, Mohamad H, Laghari JA (2016) Prioritization of different passive
parameters for islanding detection on the basis of response analysis. In: IEEE international
conference on power and energy, pp 615–619. https://doi.org/10.1109/PECON.2016.7951634
33. Raza S, Mokhlis H, Arof H, Laghari JA, Mohamad H (2016) A sensitivity analysis of different
power system parameters on islanding detection. IEEE Trans Sustain Energy 7(2):461–470.
https://doi.org/10.1109/TSTE.2015.2499781
34. Hu W, Sun Y (2009) A compound scheme of islanding detection according to inverter. In:
Asia-Pacific power and energy engineering conference, pp 1–4. https://doi.org/10.1109/APP
EEC.2009.4918040
35. Khichar S, Lalwani M (2018) An analytical survey of the islanding detection techniques of
distributed generation systems. Technol Econ Smart Grids Sustain Energy 3(10). https://doi.
org/10.1007/s40866-018-0041-1
36. Singam B, Hui LY (2006) Assessing SMS and PJD schemes of anti-islanding with varying
quality factor. In: IEEE international power and energy conference, pp 196–201. https://doi.
org/10.1109/PECON.2006.346645
37. Mahat P, Chen Z, Bak-Jensen B (2011) Review on islanding operation of distribution system
with distributed generation. In: IEEE power and energy society general meeting, pp 1–8. https://
doi.org/10.1109/PES.2011.6039299
38. Jang S, Kim KH (2004) An islanding detection method for distributed generations using voltage
unbalance and total harmonic distortion of current. IEEE Trans Power Deliv 19(2):745–752.
https://doi.org/10.1109/TPWRD.2003.822964
39. Tzelepis D, Dyśko A, Booth C (2016) Performance of loss-of-mains detection in multi-
generator power islands. In: 13th International conference on development in power system
protection, pp 1–6. https://doi.org/10.1049/cp.2016.0066
40. Jose BK, Vincent G (2017) Harmonic current based islanding detection for grid connected PV
systems. In: 2017 IEEE international conference on circuits and systems, pp 191–195. https://
doi.org/10.1109/ICCS1.2017.8325988
41. Zeineldin HH, Kirtley JH (2009) A simple technique for islanding detection with negligible
nondetection zone. IEEE Trans Power Deliv 24(2):779–786. https://doi.org/10.1109/TPWRD.
2009.2013382
42. Laaksonen H (2012) New multi-criteria-based algorithm for islanding detection in smart grids.
In: 3rd IEEE PES innovative smart grid technologies, pp 1–8. https://doi.org/10.1109/ISGTEu
rope.2012.6465661
43. Danandeh A, Seyedi H, Babaei E (2012) Islanding detection using combined algorithm based
on rate of change of reactive power and current THD techniques. In: Asia-Pacific power and
energy engineering conference, pp 1–4. https://doi.org/10.1109/APPEEC.2012.630746
44. Jun L, Liang HX, Xiao-Hu C, Miao X, Wen X (2010) Two islanding detection circuits based on
the impedance variation for the micro-grid. In: The 2nd international symposium on power elec-
tronics for distributed generation systems, pp 859–863. https://doi.org/10.1109/PEDG.2010.
5545813
45. Salman SK, King DJ, Weller G (2001) New loss of mains detection algorithm for embedded
generation using rate of change of voltage and changes in power factors. In: International
conference on developments in power system protection, pp 82–85. https://doi.org/10.1049/
cp:20010105
46. Kumar KM, Naresh M, Singh NK, Singh AK (2016) A passive islanding detection approach
for distributed generation using rate of change of negative sequence voltage and current. In:
IEEE international conference on electrical, computer and electronics engineering, pp 356–360.
https://doi.org/10.1109/UPCON.2016.7894679
A Comprehensive Review of Active Islanding Detection … 179

47. Rostami A, Jalilian A, Naderi SB, Negnevitsky M, Davari P, Blaabjerg F (2017) A novel passive
islanding detection scheme for distributed generations based on rate of change of positive
sequence component of voltage and current. In: Australasian universities power engineering
conference, pp 1–5. https://doi.org/10.1109/AUPEC.2017.8282394
48. Reddy CR, Reddy KH (2018) An efficient passive islanding detection method for integrated
DG system with zero NDZ. Int J Renew Energy Res 8:1994–2002
49. Rostami A, Bagheri M, Naderi SB, Negnevitsky M, Jalilian A, Blaabjerg F (2017) A novel
islanding detection scheme for synchronous distributed generation using rate of change of
exciter voltage over reactive power at DG-Side. In: Australasian universities power engineering
conference, pp 1–5. https://doi.org/10.1109/AUPEC.2017.8282417
50. Naveen G, Reddy KH, Reddy CR, Ramakrishna B, Bramaramba P, Reddy LB (2018) Passive
islanding detection method for integrated DG system with balanced islanding. Int J Pure Appl
Math 120(06):4041–4058
51. Rostami A, Jalilian A, Hagh MT, Muttaqi KM, Olamaei J (2018) Islanding detection of
distributed generation based on rate of change of exciter voltage with circuit breaker switching
strategy. IEEE Trans Ind Appl 55(1):954–963. https://doi.org/10.1109/IAS.2017.8101799
52. Bashir J, Jena P, Pradhan AK (2014) Islanding detection of a distributed generation system
using angle between negative sequence voltage and current. In: 18th National power systems
conference, pp 1–5. https://doi.org/10.1109/NPSC.2014.7103855
53. Bakhshi M, Noroozian R, Gharehpetian GB (2017) Novel islanding detection method for
multiple DGs based on forced helmholtz oscillator. IEEE Trans Smart Grid 9(6):6448–6460.
https://doi.org/10.1109/TSG.2017.2712768
54. Sundar DJ, Kumaran MS (2015) A comparative review of islanding detection schemes in
distributed generation systems. Int J Renew Energy Res 5(4):1016–1023
55. Reis MV, Villalva MG, Barros TA, Moreira AB, Ruppert E (2015) Active frequency drift with
positive feedback anti-islanding method for a single phase two-stage grid-tied photovoltaic
system. In: 13th Brazilian power electronics conference and 1st southern power electronics
conference, pp 1–6. https://doi.org/10.1109/COBEP.2015.7420121
56. Zeineldin HH, Kennedy S (2009) Sandia frequency-shift parameter selection to eliminate
nondetection zones. IEEE Trans Power Deliv 24(1):486–487. https://doi.org/10.1109/TPWRD.
2008.2005362
57. Khamis A, Shareef H, Bizkevelci E, Khatib T (2013) A review of islanding detection techniques
for renewable distributed generation systems. Renew Sustain Energy Rev 28:483–493. https://
doi.org/10.1016/j.rser.2013.08.025
58. Akhlaghi S, Akhlaghi A, Ghadimi AA (2016) Performance analysis of the Slip mode frequency
shift islanding detection method under different inverter interface control strategies. In: IEEE
power and energy conference, pp 1–7. https://doi.org/10.1109/PECI.2016.7459250
59. Hung GK, Chang CC, Chen CL (2003) Automatic phase-shift method for islanding detection
of grid-connected photovoltaic inverters. IEEE Trans Energy Convers 18(1):169–173. https://
doi.org/10.1109/TEC.2002.808412
60. De Mango F, Liserre M, Dell’ Aquila A (2006) Overview of anti-islanding algorithms for pv
systems. Part II: active methods. In: 12th international power electronics and motion control
conference, pp 1884–1889. https://doi.org/10.1109/EPEPEMC.2006.4778680
61. Alsharidah M, Ahmed NA, Alothman AK (2014) Negative sequence injection for islanding
detection of grid interconnected distributed generators. In: IEEE electrical power and energy
conference, pp 267–274. https://doi.org/10.1109/EPEC.2014.38
62. Karimi H, Yazdani A, Iravani R (2008) Negative-sequence current injection for fast islanding
detection of a distributed resource unit. IEEE Trans Power Electron 23(1):298–307. https://
doi.org/10.1109/TPEL.2007.911774
63. Nale R, Biswal M (2017) Comparative assessment of passive islanding detection techniques
for microgrid. In: Innovations in information, embedded and communication systems, pp 1–5.
https://doi.org/10.1109/ICIIECS.2017.8275935
64. PVPS I (2002) Evaluation of islanding detection methods for photovoltaic utility-interactive
power systems. Report IEA PVPS T5-09
180 K. Namrata et al.

65. Menon V, Nehrir MH (2007) A hybrid islanding detection technique using voltage unbal-
ance and frequency set point. IEEE Trans Power Syst 22(1):442–448. https://doi.org/10.1109/
TPWRS.2006.887892
66. Khodaparastan M, Vahedi H, Khazaeli F, Oraee H (2017) A novel hybrid islanding detection
method for inverter-based DGs using SFS and ROCOF. IEEE Trans Power Deliv 32(5):2162–
2170. https://doi.org/10.1109/TPWRD.2015.2406577
67. Shastri R, Samadhiya A, Namrata K (2021) A comprehensive review of remote and passive
IDMs of utility grid integrated MG system—Part I. In: Lecture Notes in Electrical Engineering,
vol 699, pp 485–496. https://doi.org/10.1007/978-981-15-7994-3_45
A Comparative Analysis of PI
and Predictive Control Strategy
for HESS Based Bi-directional DC-DC
Converter for DC Microgrid Applications

Srinivas Punna, Udaya Bhasker Manthati, and C. R. Arunkumar

Abstract In modern electric grid system easy integration with renewable energy
sources (RES), energy storage systems (ESSs) and electric load is an attractive
technology of DC microgrid. This chapter presents with a comparative analysis of
proportional integral (PI) and model predictive control (MPC) of multiple input bidi-
rectional DC-DC converter (MIPC) for hybrid energy storage systems (HESS) in DC
microgrid applications. The contrasting characteristics of battery and supercapcitor
(SC) is the perfect combination of HESS. This MIPC, provides decoupled control
of battery and SC power along with energy exchange between storage devices. A
controller is designed for DC microgrid applications, modified operation of controller
made it possible for both HESS charging and discharging operation thus make it a
unified controller. When batteries are used as energy storage, their rate of charge and
discharge is very low. So there are current stresses on the battery and reduced life
cycle, SC have high power density and can react immediately to sudden load varia-
tions. SC alone can’t be used as energy storage device, because it can’t supply power
to the load for longer duration. The simulation results are developed for comparative
performance of MPC over the PI control schemes. The proposed MPC was better
DC grid voltage restoration to step change in PV generation and load demand over
traditional PI control scheme.

Keywords Battery · Supercapacitor · HESS · Bi-directional converter · PI

controller · MPC · Power quality

S. Punna (B)
Electrical and Electronics Engineering Department, BVRIT Hyderabad College of Engineering
for Women, Hyderabad, Telangana 500090, India
e-mail: [email protected]
U. B. Manthati · C. R. Arunkumar
Electrical Engineering Department, NIT Warangal, Warangal, Telangana 506004, India
e-mail: [email protected]
C. R. Arunkumar
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 181
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6_8
182 S. Punna et al.

Nomenclature

HESS Hybrid Energy Storage Systems

MIPC Multiple input Bidirectional DC-DC Converter
SC Supercapacitor
ESSs Energy Storage Systems
PV Photovoltaic
MPPT Maximum Peak Power Point
P&O Perturb and Observe
ZVS Zero Voltage Switching
SOC State of Charge
PI Proportional Integral
MPC Model Predictive Control
ESR Equivalent Series Resistance
RPS Regulated Power Supply
VDC DC grid Voltage
dS Supercapacitor duty cycle
dB Battery duty cycle
VS Supercapacitor Voltage
VB Battery Voltage
iSC Supercapacitor current
iB Battery current
GiSC dSC Control to supercapacitor current transfer function
GvDC iSC Current to voltage transfer function
GiB dB Control to battery current transfer function
GPI PI Control transfer function
QSC Supercapacitor charge
iCH Supercapacitor charge current
LPF Low Pass Filter
iref Total Reference current
iB,ref Battery Reference current
iSC,ref Supercapacitor Reference current
PWM Pulse Width Modulation
ib (k + 1) Predicted battery current
isc (k + 1) Predicted supercapacitor current

1 Introduction

Due to easy integration with renewable energy sources as well as proliferation of

dc-compatible loads, DC microgrids are gaining popularity [1, 2]. High penetration
of these intermittent renewable energy sources into DC microgrids cause power
mismatch between generation and load sides culminating in dc bus voltage variations.
A Comparative Analysis of PI and Predictive Control Strategy … 183

Energy Storage Systems (ESS) of diverse variety and characteristics are used to tackle
this problem. The most commonly used ESS is battery. The battery is preferred
for high energy density and supercapacitor is preferred for high power density in
ESS applications [3, 4]. Intermittent renewable energy sources demand ESS with
high energy density whereas loads with high-pulse requirement demand ESS with
high power density. Hence a Hybrid Energy Storage System (HESS) was formed by
integrating battery and supercapacitor which fulfill the aforementioned requirements
of a DC grid [5–9]. In addition to this, HESS improves the stability of power converter
based microgrids which otherwise would be low due to low rotational inertia [10].
The advantages of both battery and supercapacitor are exploited by different HESS
topologies as shown in [11]. The most widely employed HESS configuration, that has
the advantage of individual battery and supercapacitor control is as shown in Fig. 1a.
It is possible to exchange energy between the components ESSs in this active parallel
configuration of HESS that is the battery ESS can charge the supercapacitor ESS or
vice versa. However, the energy exchange process is through DC microgrid and this
may push the grid operating limits beyond desired range.
Multiple input bidirectional (MIBD) converters, as in Fig. 1b, have better energy
exchange performance between input sources compared to multiple single-input
bidirectional dc-dc (MSIBD) converters in active parallel configuration. This makes
the former for a better choice for applications like hybrid electric vehicle (HEV)
and microgrids. The main advantage of MIBD converter over the MSIBD converter
are (i) Better energy exchange between input sources (ii) reduces the system size
(iii) reduces the cost of the converter [12]. The salient features of the two HESS
configurations are summarized in Table 1. In this study, Fig. 1b is selected, since it
offers lesser components and smaller size.
The DC microgrid setup by PV source supported by battery-supercapacitor HESS
is shown in Fig. 2.
Several multiple input topologies have been proposed which can interface multiple
sources with contrasting characteristics to a common load. In [13], multiple ESSs

Fig. 1 Different configurations of HESS, comprising of battery and SC, interfaced to DC microgrid.
a Two separate bidirectional converter modules used. b Single double-input bidirectional converter
184 S. Punna et al.

Table 1 Comparisons of HESS configurations

Configuration Figure 1a Figure 1b
Number of converters 2 1
Controlled DC link voltage Yes Yes
Controlled battery current Yes Yes
Controlled SC current Yes Yes
DC link voltage swing High High

Fig. 2 DC microgrid architecture setup powered by PV-HESS

are interfaced to a three-winding high frequency transformer wherein each source is

connected through a full-bridge circuit. For battery-supercapacitor HESS connected
number of switches required is eight which may affect the overall efficiency as far as
DC microgrid application is concerned. An Isolated multiport dc-dc converter was
presented in [14, 15] which can manage power from multiple energy sources to a
common load. But power flow between the sources, if they were replaced by ESSs,
was not explored. Even though isolated converters support wide range of voltage
levels as well as safety in the form of isolation, energy management of multiple
sources is difficult compared to that of non-isolated converters.
A number of multiple input non-isolated dc-dc bidirectional converters for inter-
facing multiple sources can be found in literature which offers far more flexibility in
implementation and power management than isolated ones [16–28]. In [16] a proce-
dure for developing all possible double-input single-output dc-dc converters was
described and was experimentally verified for a system with battery as one input.
But bidirectional power flow between the two input ports was not considered which
is necessary for HESS in microgrid applications. A multi-input dc-dc converter for
various energy sources of diverse characteristics was proposed in [17] but with no
power sharing option between the input sources.
A Comparative Analysis of PI and Predictive Control Strategy … 185

Wang et al. [18] has presented single inductor based multi-port converter for
HESS. The computation and controller effort are much more since the paper uses
model predictive control. More over SOC based analysis is required for the validity
of control strategy. Faraji et al. [19] proposes a multiport hybrid energy system with
PV and battery. The independent use of battery alone will increase the stress on the
battery. Also the proposed converter is not suitable for SC applications. Sato et al.
[20] also proposing multi-port converter for PV and battery. The proposed circuit able
to integrate battery with PV, however during various disturbances, the stress on the
battery will be high. A unified control of the PV-Battery-based grid energy manage-
ment system for both grid-connected and island operation was introduced by Yi et al.
[21]. In the proposed control method, battery balances the power flow inside the AC
micro grid and DC micro grid in all different operating conditions. This increases
the battery stress, system cost and life cycle of battery. A high efficiency dual input
interleaved converter was proposed in [26] which is especially suitable for energy
storage sources. However, Dual input interleaved converter circuit requires eight
switches makes the control strategy to be complicated. Few multi-input converters
for grid-tied and/or solar applications are reported in [27–29]. Similarly, multi-input
converters are also widely used in hybrid electric vehicle (HEV) applications as in
[30–33]. However, in all above multiple input converter topologies does not maintain
SOCSC in the predefined limits and protecting SC from over charge and discharge is
difficult.
Even though a large number of works about multi-input bidirectional converters
are available in literature, less or little importance has been given to their controller
development methods. A multi-input bidirectional dc-dc converter, proposed in [34],
is used to interface the HESS to microgrid. The converter in [34] has been modi-
fied in operation to suit for HESS operation in DC microgrid applications. One
attractive attribute of this converter is that all the switches are turned on with zero
voltage switching (ZVS) which can improve its efficiency. Apart from controlling
the charging/discharging of HESS, the unified controller can also share current
between battery and supercapacitor in such a way that the former is subjected
to less current stress which helps in extending its lifetime. Total HESS current
(charging/discharging) bifurcates such that high frequency component of the current
is being supplied by power density element supercapacitor and average or low
frequency current is being fed by energy density unit battery. The main contributions
of the chapter are as follows.
• A multi input converter based DC microgrid voltage stabilization is proposed in
this chapter.
• Detailed controller design and analysis is presented for HESS based Multiple-
Input Bidirectional converter. The presented small signal model based controller
for multi input converter can ensure stability in all operating regions.
• Implementation of energy management system for Multiple-Input Bidirectional
converter with HESS is presented for different PV and load conditions. The EMS
can easily track the SC SOC and enable different mode to ensure safe operation.
186 S. Punna et al.

Fig. 3 Two-input bidirectional converter topology for HESS [34]

• The main advantage of designed double-input Bidirectional converter is its energy

exchange mode where we can charge the SC independently from battery. Other
than that, there are several advantages to the double-input bidirectional converter,
such as efficient power distribution between the various ESSs, rapid regulation of
DC link voltage to PV power fluctuations and load variations.
• The performance evaluation of MPC over PI control scheme with step change in
PV generation and load demand for peak overshoot and settling time to restore
DC grid voltage are presented.
Two-input bidirectional DC-DC converter for HESS is shown in Fig. 3.
The chapter is organized as follows: The modeling of PV panel and P and O MPPT
algorithm is discussed in Sect. 2. Section 3 reviews the multiple input bidirectional
converter proposed in [34] and detailed modified operation as two-input bidirectional
dc-dc converter for HESS operation is elaborated. Small signal linear averaged model
of the two-input converter is developed in Sect. 4, which is crucial for the development
of unified controller for HESS charging/discharging with current sharing between
ESSs. Study of MPC for HESS is discussed in Sect. 5. Section 6 deals with the
development for state of charge (SOC) controller for supercapacitor. Simulation and
experimental validation is presented in Sects. 7 and 8. Results and inferences are
concluded in Sect. 9.
A Comparative Analysis of PI and Predictive Control Strategy … 187

2 Modelling of PV Panel

The PV panel consist of photodiodes that converts light energy in to electrical energy.
The basic unit of PV panel is known as PV cell. The PV cells are connected in series
and parallel to form a PV panel to meet the necessary power ratings. The series and
parallel combination of PV panels are utilized to make PV array. Further, the PV
array is designed based on the power, voltage and current rating required by the DC
microgrid. The electrical equivalent of PV array is shown in Fig. 4a [35].
The characteristic equation defined here is based on the PV array single diode
model. In general, the characteristics of the current and voltage (I-V) of the PV cell
shows non-linear, as shown in Fig. 4b. The characteristics equation of the PV array
can be given in Eq. (1),
    N p V pv
+ i pv Rs
V pv i pv
q Ns + N pv
Ns
i pv = N p I ph − N p Ir s e AK T −1 − (1)
Rsh

where Np: Total number of Parallel connected PV cells, Ns: Total number of series
connected PV cells. Iph : total photo-current generated from PV array. Irs represents
reverse saturation current of PV cell. Rs and Rsh represent the series and parallel
resistances. A: diode ideal factor, T: PV cell’s working temperature, K: Boltzmann’s
constant (1.38 × 1023 J/k). Vpv and Ipv represent PV array voltage and current. The
current equation of PV panel is non-linear and dependent on weather conditions
like insolation and temperature. To find the maximum power point on non-linear PV
curve maximum power point tracking algorithm is engaged.

Fig. 4 a Equivalent circuit representation of PV array, b I-V and P–V Characteristics of PV

188 S. Punna et al.

Fig. 5 a P&O MPPT flowchart, b P-V characteristics

2.1 P&O MPPT Algorithm

The maximum power from a PV panel is extracted by using MPPT algorithms.

Perturb and observe (P&O) is the conventional MPPT algorithm for maximum power
extraction. In this method extracted power is observed by varying the duty based on
comparing the power output. The change in power (ΔPPV ) and voltage (ΔV PV )
between two samples are recoded and compared to generate the duty for the next
sample. As, it can also be observed from the flowchart of P&O algorithm shown in
Fig. 5a, when both ΔPPV and ΔV PV are positive or ΔPPV and ΔV PV are negative
duty factor increases while, alternation of positive and negative sign in ΔPPV and
ΔV PV leads to reduces of duty factor.

3 Two-Input Bidirectional DC-DC Converter (TIBC)

Operation

The two input bidirectional converter, various modes of operation are elaborated in
detail in [34]. Modified operation of the converter is explained here. It consists of
three switch-legs. Battery (VB ) and supercapacitor (VS ) modules are connected to
legs 2 and 3 respectively. DC microgrid (VDC ) is connected to leg 1. In this topology,
the operating voltages are selected such a way that the DC grid voltage is greater
than ESSs voltages and SC voltage is selected as less than the battery voltage. The
inductor LB and LS are connected across legs 1, 2 and 1, 3 respectively. The operating
modes of presented TIBC is explained as follows:

3.1 HESS Discharging Mode

Whenever there exists a mismatch between PV output and load power, DC microgrid
voltage varies from steady state value. When the load exceeds PV generation capacity
A Comparative Analysis of PI and Predictive Control Strategy … 189

Table 2 Switching state in

Time scale T1 T2 T3
different time intervals
Operating switches S2 , S3 , S5 S2 , S4 , S5 S1 , S4 , S6

or when the power produced by PV falls due to decreased solar irradiation, the DC
microgrid voltage decreases. The deficient power must be supplied by HESS during
this period. So power flows in this mode from HESS to DC microgrid, by regulating
the bidirectional converter properly. The proposed converter switching sequence for
discharging mode is given in Table 2.
The converter operation using equivalent circuits and switching model can be
divided into three intervals as shown in Figs. 6 and 7 respectively. Switch pairs S1 /S2 ,
S3 /S4 and S5 /S6 always work in complimentary fashion. Switch pair S2 /S5 and S1 /S6
always switch together in this mode and their gating pulses are complimentary to
each other. At time instant t0 , switches S2 , S3 and S5 are turned on which will increase
inductor currents iB and iS linearly with slopes VB /LB and VS /LS respectively. At t1 ,

Battery Supercapacitor Battery Supercapacitor

VB VS VB VS

+ +
S1 S3 S5 S1 S3 S5
LB LB
iB
VDC iS VDC
LS LS
S2 S4 S6 S2 S4 S6
- -
(a) (b)
Battery Supercapacitor

VB VS

+
S1 S3 S5
LB

VDC
LS

S2 S4 S6
-

(c)

Fig. 6 Switching equivalent circuit of two-input bi-directional converter, a S2 , S3 , S5 operates, b

S2 , S4 , S5 operates, c S1 , S4 , S6 operates
190 S. Punna et al.

Fig. 7 Operational
waveforms for HESS
discharging mode

switch S3 is turned off to provide a free-wheeling path for current iB through body
diode of S4 . After dead time interval for switch pair S3 /S4 , switch S4 is turned on.
Since body diode of S4 is already conducting when gating signal was given, S4 turns
on with ZVS. At t2 , switches S2 and S5 are turned off causing inductor current iL2
to flow through body diodes of switches S1 and S6 with a negative slope of VDC /LS .
Inductor current iB also flows body diode of S1 with a negative slope of VDC /LB .
Switches S1 and S6 are gated on after dead time interval of switch pairs S2 /S1 and
S5 /S6 . Similar to the turn on of S4 , switches S1 and S6 are also turned on with ZVS
since body diodes of respective switches are already in conduction. At t3 , switches
S1 , S4 and S6 are turned off. Consequently, body diodes of switches S2 , S3 and S5
will conduct in order to maintain the flow of inductor currents. Currents, iB and iS ,
flow with positive slopes of VB /LB , and VS /LS respectively. After dead time interval,
gating pulses are given to switches S2 , S3 and S5 which turn on with ZVS. If dB is
duty cycle of gating pulse given to switch S3 and dS is duty cycle of gating pulse
given to switches S2 and S5 , applying volt-second balance to inductors LS and LB
gives

dS
VDC = . VS (2)
1 − dS
dB
VDC = . VB (3)
1 − dS
A Comparative Analysis of PI and Predictive Control Strategy … 191

Since VB is greater than VS , dB will always be lesser than dS . Therefore, by

controlling dB and dS , power flow from battery and supercapacitor to dc grid can be
controlled independently.

3.2 HESS Charging Mode

Whenever PV generation exceeds the load power demand or reduction in load power,
there exist an excess power in the DC microgrid that leads to increase in DC microgrid
voltage. This excess power utilized to charge the HESS and hence the direction of
power flow is from DC microgrid to HESS. The converter operation using equivalent
circuits and switching model can be divided into three intervals as shown in Figs. 8
and 9. Switch pair S1 and S6 are switched simultaneously. Switch pair S2 and S5
are also gated simultaneously but with complimentary pulses to that of S1 and S6 .
Similarly switches S4 and S3 are also gated in complimentary fashion. At t0 , switches
S1 , S4 and S6 are gated on to increase the inductor currents, iB and iS , in negative
direction linearly with slopes VDC /LB and VDC /LS respectively. Operating switches
in each time interval is tabulated in Table 3.
Inductors, LB and LS , store energy in this time interval until instant t1 . At t1 , switch
pair S1 /S6 are turned off. In order to maintain inductor current iS , body diodes of
switches S2 and S5 turn on. Energy stored in inductor LS is now used up to charge

Fig. 8 Operational
waveforms for HESS
charging mode
192 S. Punna et al.

Battery Supercapacitor Battery Supercapacitor

VB VS VB VS

+ +
S1 S3 S5 S1 S3 S5
LB LB

VDC VDC
LS LS
S2 S4 S6 S2 S6
- -
(a) (b)
Battery Supercapacitor
VB VS

+
S1 S3 S5
LB

VDC
LS

S2 S4 S6
-

(c)

Fig. 9 Switching equivalent circuit of two-input bi-directional converters, a S1 , S4 , S6 operates, b

S2 , S4 , S5 operates, c S2 , S3 , S5 operates

Table 3 Switching states in

Time scale T1 T2 T3
different time intervals
Operating switches S1 , S4 , S6 S2 , S4 , S5 S2 , S3 , S5

supercapacitor. Inductor current iB free-wheels through body diode of switch S2 .

After dead time interval for switch pair S1 /S6 and S2 /S5 , gating pulses are given to S2
and S5 which turns them on with ZVS since body diodes of respective switches are
already in conduction. At time instant t2 , switch S4 is turned off. Inductor current iS
increase linearly with slope VS /LS through body diode of switch S3 . After dead time
interval of switch pair S4 /S3 , gating pulse to S3 is applied to turn it on with ZVS.
Energy stored in inductor LB is now used to charge battery. At instant t3 , switches
S2 , S3 and S5 are tuned off, consequently body diodes of switches S1 , S4 and S6
turns on to maintain flow of inductor currents. Gating pulses are given to S1 , S4 and
S6 after dead time interval to turn them on with ZVS. If dS is duty cycle of gating
pulse to S1 /S6 switch pair and dB is duty cycle of gating pulse to switch S4 , applying
A Comparative Analysis of PI and Predictive Control Strategy … 193

volt-second balance to inductors LS and LB gives

dS
VS = . VDC (4)
1 − dS
dS
VB = . VDC (5)
1 − dB

Since VB is greater than VS which in turn is smaller than VDC , dB is always greater
than dS . By controlling dB and dS , power flow from DC microgrid to HESS can be
controlled.

3.3 HESS Energy Exchange Mode

Supercapacitor in HESS is a power density unit. It cannot supply energy for long
duration continuously like battery due to its high self-discharging effect. For proper
functioning of HESS in DC microgrid applications it is necessary that component
ESSs of HESS should have enough energy stored in it. To maintain supercapacitor
charge within desired limits, it has to be charged whenever needed from the energy
density unit battery for proper functioning of HESS. This mode constitutes power
flow from battery to supercapacitor. Circuit topology in this mode is given in Fig. 10a.
Steady state waveforms are as shown in Fig. 10b.
In this mode, first switch leg is inactive (switches S1 and S2 are off) thus effec-
tively isolating the DC microgrid from HESS during supercapacitor charging process.
Switch pairs S5 /S6 and S3 /S4 work in complimentary manner. Switch S5 is always
on in this mode which cause switch S6 to be always off. Switch S3 is operated upon
with duty cycle d. Power flow from battery to supercapacitor involves buck operation
of switch S3 . Inductors, LB and LS , are in series and this considerably reduce the

Fig. 10 HESS energy exchange mode of operation. a Active circuit through which current flows,
b steady-state waveforms
194 S. Punna et al.

current ripple. By controlling d, power flow from battery to supercapacitor can be

controlled. Applying volt-second balance to equivalent series inductor L (L = LB +
LS ) gives

VS = d . V B (6)

Similarly, power flow from supercapacitor to battery is possible by complimentary

operation of switch S3 . This indicates boost operation of switch S4 . Operation is
similar to that explained above.

3.4 Mode Transitions

The mode of operation is determined from present condition of DC microgrid and

continuous monitoring of supercapacitor SOC. Battery SOC is not considered in
this work since it is assumed that battery energy is not depleted fast compared to
supercapacitor. Flow chart for simple mode transition logic is shown in Fig. 11.
When DC microgrid voltage exceeds the set reference, HESS is put into charging
mode, provided supercapacitor SOC is within prescribed limits. When DC microgrid
voltage is less than set reference value, HESS is put into discharging mode, provided
supercapacitor SOC is in safe limits. If supercapacitor SOC violates the prescribed
limits, HESS is put into energy exchange mode, thus isolating the DC microgrid
electrically from HESS. Supercapacitor SOC working range and its mathematical
expression using coulomb’s counting method are

S OC M I N ≤ S OC ≤ S OC M AX (7)

Fig. 11 Transitions to
different modes represented
using a flow chart
A Comparative Analysis of PI and Predictive Control Strategy … 195

   
1
% S OC = S OCi + i C H dt × 100 (8)
Q SC

where SOCi is initial SOC of supercapacitor, QSC is rated supercapacitor charge and
iCH is supercapacitor charging current.

3.5 Mathematical Design Calculations of Filter Parameters

The design of filter inductor assumes that the bidirectional DC-DC converter always
operates in continuous conduction mode (CCM). Mathematical modeling of filter
inductor and capacitor for various converters is given in [36–38]. Equivalent circuit
of ON time period and OFF time period of boost converter are shown in Fig. 12.
Calculation of peak-to-peak inductor (i L ) current during switch ON duration is
calculated as

i L = i L (DT ) − i L (0) (9)

V P V .DT V P V .D VDC .D.(1 − D)

i L (O N ) = = = (10)
L fs L fs L

Similarly i L during switch OFF duration of time

i L = i L (DT ) − i L (T ) (11)

(VDC − V P V )(1 − D)T VDC .D.(1 − D)

i L (O F F) = = (12)
L fs L

In this work, the rating of the PV panel are: MPP voltage (V P V ): 32 V, MPP
current (iPV ): 3 A and maximum power (Pmax ): 96 W. For designing DC microgrid
voltage (VDC ) to 48 V, load resistance can be selected as follows,

L L
Switch ‘S’
closed
+ +
-
VPV + S C VDC -
VPV + Switch ‘S’
open C VDC

- -
(a) (b)

Fig. 12 Equivalent circuit of boost converter, a turn ON, b turn OFF

196 S. Punna et al.

2
VDC 482
Rload = = = 24  (13)
Pmax 96

For designing filter inductor, the peak-to-peak inductor current (i L ) is limited
to 5% of total current in this work, inductor (L) is defined as follows,

V P V .D
L= (14)
f s .i L

Similarly, the output filter capacitor (C) reduces the voltage ripple, thus can be
calculated as
D.VDC
C= (15)
f s .Rload .VDC

4 Small Signal Linear Averaged Model of 2-Input

Bidirectional Converter for PI Controller Design

Small signal averaged model is the basis for development of unified controller for
charging/discharging of HESS. An effective model will contribute a controller which
will not only provide good closed loop performance but also provide stable operation
amidst converter dynamics and external disturbances.
As per circuit operation as explained in previous sections, independent control
of power flow is possible from battery and supercapacitor to DC microgrid. Hence,
decoupled small signal model can be developed for battery stage as well as for
supercapacitor stage which is similar to developing dynamic model of two single-
input bidirectional converter. This decoupled averaged small signal model of battery
and supercapacitor power flow stage aids in designing simpler controllers which
otherwise would be complicated due to presence of multi-input multi-output control
systems [22]. Small signal model of battery to DC microgrid as well as supercapacitor
to DC microgrid are developed as follows.
A unified controller for battery/supercapacitor charging and discharging will
suffice since both the circuit operations share same small signal transfer function
due to complimentary action of switches. For developing linear models, HESS
discharging mode of operation is considered. Control system description for bifur-
cation of battery and supercapacitor power is shown in Fig. 13. The outer voltage
loop is designed based on supercapacitor stage because of its inherent fast dynamic
response.
Thus control system description of supercapacitor stage consists of an inner
current loop and an outer voltage loop as depicted in Fig. 14a. Similarly, battery
current reference is regulated by a current controller as in Fig. 14b. Inner current
loop of supercapacitor is designed for a higher bandwidth than that of battery for
A Comparative Analysis of PI and Predictive Control Strategy … 197

Fig. 13 Overall control system description for current bifurcation between SC and battery units

Fig. 14 Supercapacitor and battery control logic. a Control system description for supercapacitor
control. b Control system description for battery control

fast response. Inner current loop operate at a faster rate compared to that of outer
voltage loop. Hence voltage loop bandwidth is maintained lesser than inner current
loop. Switching frequency considered is 10 kHz.

4.1 Small Signal Linear Averaged Model

of Supercapacitor-DC Microgrid Stage

From Fig. 7 and as per operation explained in Sect. 3, during time interval (t0 ~ t2 )
the state equations are as follows.

di Ls
Ls = vsc (16)
dt
dv DC v DC
C =− (17)
dt R
During time interval (t2 ~ t3 ), the state equations are as follows.
198 S. Punna et al.

di Ls
Ls = −v DC (18)
dt
dv DC v DC
C = i Ls − (19)
dt R
After perturbation and linearization of above state equations, control-to-inductor
current and inductor current-to-voltage small signal transfer functions are obtained
as follows.
i SC (s)
G i SC dSC =
d SC (s)

VDC (1 + D SC ) 1 + s 1+D
RC
= SC
(20)
R(1 − D SC )3 s 2 L SC C 2 + s L SC 2 + 1
(1−D SC ) R(1−D SC )

R(1 − D SC ) 1 − s R(1−DSC )2
D SC L SC
v DC (s)
G v DC i SC = = (21)
i SC (s) (1 + D SC ) 1 + s 1+D RC
SC

The inner current loop PI controller is built based on the above model with a
bandwidth of 1.6 kHz and a phase margin of 60°. The Bode plot is shown in Fig. 15
for compensated and uncompensated current loop transfer functions. The PI transfer
function is given by

2291
G pi_i S = 0.4124 + (22)
s

Fig. 15 Bode plot of inner current controller of SC Control logic, with and without compensation
A Comparative Analysis of PI and Predictive Control Strategy … 199

Fig. 16 Bode plot of outer voltage control loop for with and without compensation

The outer voltage loop’s PI controller is also constructed based on the average
model. The bandwidth must be smaller than the right half plane zero. The designed
voltage controller has a bandwidth of 200 Hz and phase margin of 60°. Bode plot
for the same is as shown in Fig. 16. The transfer function is given by

266
G pi_v = 0.5054 + (23)
s

4.2 Small Signal Linear Averaged Model of Battery-DC

Micro Grid Stage

Discharge operation of battery is completed in three time intervals as shown in Fig. 7.

As explained in Sect. 3, during time interval (t0 ~ t1 ), the state equations are as follows.

di B
LB = vB (24)
dt
dv DC v DC
C =− (25)
dt R
The above system of state equations can be represented in matrix form after
converting to frequency domain which takes the following form

dX
= AX + BU (26)
dt
200 S. Punna et al.

Y = C X + EU (27)

where X is a matrix containing all state variables (iB , vDC ), U is a matrix containing
inputs and Y is a matrix containing all the system outputs. Equations (24) and (25)
can be rewritten in matrix form as
 di B   
iB
dt = [A1 ] + [B1 ][v DC ] (28)
dv DC
dt
v DC
 
i
[i B ] = [C1 ] B + [E 1 ][v DC ] (29)
v DC

 

0 0 1  
where A1 = , B 1 = L B , C = 1 0 and E = [0].
1 1
0 − RC1
0
Similarly, for second time interval (t1 ~ t2 ), the state equations and its matrix form
can be represented as follows.

di B
LB =0 (30)
dt
dv DC v DC
C =− (31)
dt R
 di B   
iB
dt = [A 2 ] + [B2 ][v DC ] (32)
dv DC
dt
v DC
 
iB
[i B ] = [C2 ] + [E 2 ][v DC ] (33)
v DC

   
0 0 0  
where A2 = , B2 = , C2 = 1 0 and E 2 = [0].
0 − RC1
0
In third time interval (t2 ~ t3 ), the state equations and state model in matrix form
are expressed as follows.

di B
LB = −v DC (34)
dt
dv DC v DC
C = iB − (35)
dt R
 di B
  
i
dt = [A3 ] B + [B3 ][v DC ] (36)
dv DC
dt
v DC
A Comparative Analysis of PI and Predictive Control Strategy … 201
 
i
[i B ] = [C3 ] B + [E 3 ][v DC ] (37)
v DC

 
0 − L1B 0  
where A3 = 1 , B3 = , C3 = 1 0 and E 3 = [0].
C
− RC1
0
From Fig. 7 it is evident that the bidirectional converter will be subjected to state
model represented by Eqs. (28) and (29) for a time duration of (dB ·TS ) after which
it will undergo transition to the model represented by Eqs. (32) and (33) where it
stays for a duration of [(dS − dB )·TS ]. The duration of third time interval is [(1 −
dS )·TS ] for which the model is Eqs. (36) and (37). Thus, averaged state model of the
converter is obtained as follows.

A = A1 · d B + A2 · (d S − d B ) + A3 · (1 − d S ) (38)

B = B1 · d B + B2 · (d S − d B ) + B3 · (1 − d S ) (39)

C = C1 · d B + C2 · (d S − d B ) + C3 · (1 − d S ) (40)

E = E 1 · d B + E 2 · (d S − d B ) + E 3 · (1 − d S ) (41)

dx
= Ax + Bu (42)
dt

y = C x + Eu (43)

Assuming perturbations in state variables x̂ around the steady state DC value of

X, then

x = X + x̂ (44)

Similarly, assuming perturbations in duty cycles (dB and dS ) and input (u),

d S = D S + d̂ S (45)

d B = D B + d̂ B (46)

u = U + û (47)

For obtaining small signal model, the perturbations are assumed to be very small
compared to steady state DC values, i.e., d̂ S  D S , d̂ B  D B and û  U .
Also, these perturbations are assumed to be constant during one switching period.
202 S. Punna et al.

Substituting Eqs. (44)–(47) into Eqs. (42) and (43) and linearizing by neglecting
second-order terms, the state space model of the converter is obtained as below.

d x̂
x̂˙ = = A x̂ + B û + F d̂ B + G d̂ S (48)
dt

ŷ = C x̂ + (C1 − C2 )X d̂ B + (C2 − C3 )X d̂ S (49)

where F = (A1 − A2 )X + (B1 − B2 )U, G = (A2 − A3 )X + (B2 − B3 )U and X

= −A−1 BU. Applying Laplace transform to Eqs. (48) and (49) and rearranging,
control-to-output transfer function of battery stage is obtained as

ŷ(s)
= C(s I − A)−1 F + (C1 − C2 )X (50)
d̂ B (s)

where I is identity matrix. Substituting the values for C, A, F, C1 , C2 and X from

Eqs. (42), (43), (48), (49), (28), (29), (32) and (33), the control-to-inductor current
transfer function of battery loop is given as:

î B (s) VDC 1 + s RC
G i B dB = = · (51)
d̂ B (s) D B R(1 − D S ) LBC
s 2 (1−D 2 + s R(1−D )2 + 1
LB
S) S

The PI controller is designed to regulate battery current based on the above transfer
function in Eq. (51). The bandwidth of the designed controller is selected such a way
that it is less than that of supercapacitor stage. The controller is designed with a
bandwidth of 1 kHz and phase margin of 60°. Bode plot for battery current loop with
compensated and uncompensated plant is shown in Fig. 18. Transfer function of PI
controller is given by

7300
G pi_i B = 1.971 + (52)
s

5 Control Scheme Using Model Predictive Controller

The block diagram of MPC based HESS control scheme is represented in Fig. 19.
The outer voltage control loop generates the total current reference (iref ) to be handled
by the HESS. The battery reference current (ibref ) and SC reference current (iscref )
used as current reference to the battery and SC systems. The MPC generates optimal
switching states to maintain battery and SC current according to their reference
currents.
A Comparative Analysis of PI and Predictive Control Strategy … 203

5.1 Outer Voltage Control Loop

The output of the outer voltage control loop is to calculate the total current demand
handled by the HESS to ensure the constant DC grid voltage. Expression for the
prediction of total reference current is as follows.
The output capacitor current is given by the equation:

dVdc
ic = C (53)
dt
V(k + 1) − V(k)
ic = C × (54)
Ts

where, v(k + 1) is the control target and the v(k) is the DC bus voltage to be regulated.
In other words, ‘(K + 1)th’ state represents the reference input and kth value indicates
the measured value.
Total reference current can be written as sum of output capacitor current (ic ) and
output DC bus current (idc )

iref = ic + idc (55)

[Vref − V(dc)]
ic (k + 1) = C × (56)
N × Ts

Total reference current can be written as:

iref (k + 1) = ic (k + 1) + idc (k) (57)

5.2 Inner Current Control

In HESS, battery supplies the average power required by the load and SC supports
the transient power requirement. Hence total current reference can be written as:
∗ ∗ ∗
i ref = i b,ref + i sc,ref (58)

To allocate the predicted HESS current properly between battery and SC, a LPF
with cutoff frequency of ωr is used. The corresponding battery reference current
(ibref ) and SC current reference (iscref ) are as follows:

∗ wr
i b,r ef = .i ∗ (59)
(S + wr ) r e f
204 S. Punna et al.

 
∗ wr
i SC,r = 1− .i ∗ (60)
ef
(S + wr ) r e f

It is important to calculate the battery current and SC current for the regulation
of DC bus voltage. The battery and SC currents are predicted from inductors current
equations. Thus, db and dsc can be calculate to regulate the DC bus voltage by the
following prediction equations.

di b
Lb = Vb db − Vdc (1 − db ) (61)
dt
di sc
L sc = Vsc dsc − Vdc (1 − db ) (62)
dt
i b (k + 1) − i b (k)
Lb = Vb (k)db (i) − Vdc (k)(1 − db (i)) (63)
Ts
Ts
i b (k + 1) = × [Vb (k)db (i) − Vdc (k)(1 − db (i))] + i b (k) (64)
Lb
Ts
i sc (k + 1) = × [Vsc (k)dsc (i) − Vdc (k)(1 − db (i))] + i sc (k) (65)
L sc

6 Supercapacitor State-of-Charge Controller

Unlike batteries, supercapacitors have low ESR. So it cannot retain energy for long
duration. In order to prevent supercapacitor energy from getting depleted beyond a
minimum permissible point, a control logic is devised to maintain the SOC within
required limits of Eq. (7). Whenever this desired range of SOC is violated, the
converter enters into HESS Energy exchange mode explained in Sect. 3. When SOC
of supercapacitor falls below prescribed minimum limit, it is charged with a constant
current, ICH , with battery power. When SOC exceeds the maximum safe limit, super-
capacitor is allowed to discharge its excess energy to battery with a constant current-
ICH . Thus SOC of supercapacitor is maintained within the limits by either buck (SC
charging) or boost (SC discharging) operation as explained in Sect. 3. The control
logic for SOC controller is shown in Fig. 17. A PI controller is used to regulate the
supercapacitor current. The controller is designed based on buck mode control-to-
output transfer function. Since its design is not within the scope, it is not shown here.
Same controller is sufficient for buck and boost operation in HESS Energy exchange
mode explained above since transfer function is same for both the operations. It is
to be noted that whenever HESS Energy exchange mode is active, either due to low
SOC or high SOC, DC microgrid is isolated from HESS electrically.
A Comparative Analysis of PI and Predictive Control Strategy … 205

Fig. 17 SOC charge controller for supercapacitor. a Logic to select between charging and
discharging. b Control logic diagram

Fig. 18 Bode plot for battery inner current control logic

7 Simulation Study and Discussion

In this segment, the results of the PI and MPC control schemes are displayed for
four test cases. The nominal parameters for simulation study are presented in Table
4. The entire model is implemented using MATLAB. The model consists of two
bidirectional converters—one for battery and other for SC. The PV array is a unidi-
rectional, which is connected to the boost converter. The different operating cases
are presented in the following sections for step change in PV generation and load
demand.
206 S. Punna et al.

Fig. 19 Block diagram representation of MPC based control scheme

Table 4 System parameters

S. No. Parameters Value
for simulation study
1 MPPT Voltage (Vmppt ) 32 V
2 MPPT Current (Imppt ) 3A
3 MPPT Power (Pmppt ) 96 W
3 SC Voltage (VSC ) 32 V
4 SC inductance (LS ) 0.355 mH
5 Battery voltage (VB ) 24 V
6 Battery inductance (LB ) 0.3 mH
7 Boost inductance (L) 4.1 mH
8 Resistance (R) 24 
9 DC grid voltage (VDC ) 48 V
10 Capacitance (C) 300

7.1 Step Change in PV Generation Using PI Control Scheme

The simulation results for step change in PV generation using PI control scheme as
shown in Fig. 20. Due to atmospheric variation power produced by the PV panel
changed from 96 to 192 W at t = 0.2 s and bring back to 96 W at t = 0.4 s. For
that PV current increases from 3 to 6 A at 0.2 s and bring back to 3 A at 0.4 s. In
this case load power requirement constant at 96 W. At t = 0.2 s PV power is more
than load power requirement by 96 W, DC grid voltage increases more than 48 V.
Immediately SC absorbs excess power of 96 W in short duration until battery can
regulate the DC grid voltage to 48 V. The battery and SC charge according to energy
A Comparative Analysis of PI and Predictive Control Strategy … 207

Fig. 20 Simulation results for step change in PV generation using PI control scheme, a PV power,
b load power, c battery power, d SC power, e PV current, f DC grid voltage g battery current, h SC
current

management scheme to maintain DC grid voltage constant at 48 V. From obtained

results shows that settling time, step increase in PV generation at t = 0.2 s is 110 ms
and step decrease in PV generation at t = 0.4 s is 120 ms.

7.2 Step Change in Load Demand Using PI Control Scheme

The simulation results for step change in load demand as shown in Fig. 21. At a time
instant t = 0.2 s load power demand increases from 96 to 192 W. This increases the
load current from 2 to 4 A. During this state PV current is constant at 3 A. At t =
0.2 s, when load power increases to 192 W which is the beyond range of PV power
generation. Due to this, power imbalance between source and load power. Due to
sudden load fluctuations will affect the DC grid voltage. HESS respond immediately
208 S. Punna et al.

Fig. 21 Simulation results for step change in load demand using PI Control scheme, a DC grid
power, b battery power, c SC power, d SC SOC, e DC grid voltage, f DC grid current, g battery
current, h SC current

to these sudden load fluctuations to handle the excess power in DC microgrid. The
SC handles the high frequency power transients and study state component of power
handled by the battery. The time taken to restore the voltage, at t = 0.2 s is 60 ms
and t = 0.4 s is 110 ms.

7.3 Step Change in PV Generation Using MPC

The simulation results for step change in PV generation using MPC scheme are
presented in Fig. 22. Due to the atmospheric variations power produced by the PV
panel changes from 96 to 192 W, bring back to 96 W at 0.2 s and 0.4 s respectively.
The step change in PV generation causes, step change in PV current at 0.2 s and 0.4 s.
A Comparative Analysis of PI and Predictive Control Strategy … 209

Fig. 22 Simulation results for Step change in PV generation using MPC control, a PV power, b
load power, c battery power, d SC power, e PV current, f DC grid voltage, g battery current, h SC
current

At time instant 0.2 s and 0.4 s the DC grid voltage is almost constant at 48 V with
negligible settling time of 2 ms and 5 ms. From the results MPC control scheme is
faster with less peak overshoot of DC grid voltage compared to PI control scheme.

7.4 Step Change in Load Demand Using MPC

The simulation results for step change in load demand for MPC control scheme are
shown in Fig. 23. The maximum power produced by the PV panel is limited to 96 W.
The load power requirement increased to 192 W at t = 0.2 s. Which is more than
the PV power generation. Due to this power imbalance between PV generation and
load power demand, DC grid voltage falls below 48 V. The proposed control scheme
210 S. Punna et al.

Fig. 23 Simulation results for step change in load demand using MPC control, a DC grid power,
b battery power, c SC power, d SC SOC, e DC grid voltage, f DC grid current, g battery current, h
SC current

controls the bidirectional converter, so that HESS supply excess power of 96 W to

maintain the DC grid voltage constant at 48 V. The voltage restored to set reference
value by using MPC schme is about 3 ms and 10 ms which 10 times lesser than PI
control scheme.

7.5 Comparative Performance Evaluation

The performance evaluation of MPC scheme over the PI control scheme with step
change in PV generation and load demand for the peak overshoot and settling time
to restore DC grid voltage. The maximum peak overshoot can be calculated during
step change in PV generation and load demand as follows.
A Comparative Analysis of PI and Predictive Control Strategy … 211

VDC,r e f − Vmax
%M P = × 100 (66)
VDC,r e f

The comparative performance of PI and MPC schemes are shown in Fig. 24. By
graphical results represents, MPC control scheme is faster and reduced peak over
shoot compared to PI control scheme. The MPC control scheme is designed such
that, SC supports the HESS up to battery steady state operation. The MPC control
scheme is faster DC grid regulation and robust control.

Fig. 24 Graphical comparative performance of PI controller over MPC, a settling time b peak
overshoot
212 S. Punna et al.

8 Experimental Results

The experimental setup developed for validation of proposed MPC controller for
two-input bidirectional converter as shown in Fig. 25. The dSPACE-DS1104 digital
controller used in this work. The current sensor LA 55-P and voltage sensor LV 25-P
are used for current and voltage measurements in the experimental validation. The
regulated power supply (RPS) acts like PV emulator whose current controlled by
boost converter. HESS is wired up using one unit of Exide Chloride 12 V, 7 Ah lead-
acid battery and Maxwell BMOD0058 16 V, 58F supercapacitor. The bidirectional
converter is connected using six MOSFET switches IRFP460.
The performance of DC microgrid supported by HESS is verified for different
cases. (1) Step variation of DC grid reference voltage (2) Step variation of PV
generation (3) Step variation of load demand (4) Energy exchange mode. System
parameters for DC grid are given Table 5. DC microgrid is designed at a nominal
voltage of 20 V.

8.1 Step Variation of DC Grid Reference Voltage

The experimental results for step change in DC grid voltage reference as shown in
Fig. 26. At time instant t1 , VDC,r e f is step changed from 18 to 20 V as shown in
Fig. 26a. In this case load resistance and PV generation kept constant at 25  and
1.3 A respectively. The smooth transition of DC grid voltage is observed. The HESS

Fig. 25 Hardware prototype developed for HESS

A Comparative Analysis of PI and Predictive Control Strategy … 213

Table 5 Nominal parameters

S. No. Parameter Value
of DC microgrid
1 SC Voltage (VSC ) 10 V
2 SC inductance (LS ) 1.43 mH
3 Battery voltage (VB ) 12 V
4 Battery inductance (LB ) 4.8 mH
5 Boost inductance (L) 4.1 mH
6 Resistance (R) 25 
7 DC grid voltage (VDC ) 20 V
8 Capacitance (C) 150

supplies required power to maintain the DC grid stability. SC supports transient

current demand and battery supports steady state demand to maintain the DC grid
stability. Similarly, VDC,r e f step changed from 20 to 18 V at time t2 as shown in
Fig. 26b. Figure 26c shows the experimental results of step increase and decrease of
DC grid reference (VDC,r e f ). At time t3 , VDC,r e f changed from 18 to 21 V immediately
HESS supplies required power to maintain the DC grid stability. At time t4 , VDC,r e f
changed from 21 to 18 V immediately HESS responds transient component of current
charging SC and steady state component of current charging battery to maintain DC
grid voltage stability.

8.2 Step Variation of PV Generation

The experimental results for step change in PV generation waveforms are given in
Fig. 27. At time t1 , PV generation suddenly increases to 40 W from 30 W this can
be emulated by increasing PV current to 2 A from 1.5 A as shown in Fig. 27a. Than
immediately HESS responds to store excess power of 10 W in short duration of
time, transient component of current charge SC until battery can regulate the DC
grid voltage to 20 V. At time t2 , PV generation brought back to original condition of
30 W from 40 W power as shown in Fig. 27b.Than immediately HESS responds such
way that supplies deficient 10 W power in short duration of time. At time instants t1
and t2 , DC grid voltage is almost stable at 20 V as shown in Fig. 27a, b. Figure 27c
shows the experimental results for sudden increase and decrease for PV generation
at time t3 and t4 respectively. With the implementation of proposed control strategy
better DC grid voltage regulation is achieved.
214 S. Punna et al.

5v/div
Vdc 2volt

iB
0.5A/div

iSC
0.5A/div

(a)

Vdc 5v/div
2volt
iB
0.5A/div

iSC 0.2A/div

(b)

21 V
Vdc 3volt 5v/div
18 V
iB 2A/div

iSC 0.5A/div

(c)

Fig. 26 Experimental results for step change in DC grid voltage reference, a step increase in
reference, b step decrease in reference, c step increase and decrease reference voltage
A Comparative Analysis of PI and Predictive Control Strategy … 215

Fig. 27 Experimental results for HESS, a step increase in PV generation, b step decrease in PV
generation, c step increase and decrease PV generation
216 S. Punna et al.

Fig. 28 Experimental results for HESS, a step decrease in load demand, b step increase in load
demand, c step increase and decrease load demand
A Comparative Analysis of PI and Predictive Control Strategy … 217

8.3 Step Variation of Load

Experimental results for step change in load demand as shown in Fig. 28. In this case
iPV is kept constant at 1.3 A and RPS connected to the boost converter maintaining
input voltage to 12 V. Initially load resistance is 25  up to instant t1 . At time t1 ,
load resistance increased to 35  as result decrease in load demand as shown in
Fig. 28a. The surge in grid voltage due to excess energy at DC microgrid. At that
instant HESS stores excess energy at DC microgrid to maintain grid voltage constant
at 20 V. Transient component of current is charging SC and battery charging current
is slowly increasing to steady state at up to the instant t2 .
The load resistance is brought back to original condition R = 25  at instant t3
as shown in Fig. 28b. As a result increase in load demand is analyzed. At instant t3 ,
small dip in DC grid voltage due to power mismatch between PV generation and load
demand. The HESS supplies deficient power to maintain grid voltage constant by
discharging battery and SC. Transient component of current met by SC and steady
state component of current supplies battery can be shown by using current waveforms.
The experimental study for both decrease in load and increase in load as shown in
Fig. 28c. DC grid voltage is almost constant at 20 V, small spikes due to fast dynamics
of power semiconductor devices.

8.4 Energy Exchange Mode

The power flow from battery to SC is illustrated in this mode. Experimental results for
energy exchange mode as shown in Fig. 29. Before time instant t1 HESS disconnected
from DC microgrid, in this period grid voltage 18 V due to reduced PV generation.

Vdc 12.5v/div

HESS disconnected HESS Energy Exchange

HESS Discharhing Mode
from DC Microgrid Mode

iSC 0.5A/div

t1 t2 iB 1A/div

Fig. 29 Experimental results for HESS energy exchange mode. DC microgrid voltage (VDC ),
battery current (iB ), SC current (iS ). Transition from discharging mode to energy exchange mode
218 S. Punna et al.

At time instant t1 , HESS is connected for discharging of battery and SC to regulate

DC grid voltage to 20 V. At time instant t2 , supercapacitor SOC reduced to below
50%, which is taken as prescribed lower limit of SOC range than HESS shifts from
discharging mode to energy exchange mode. Now SC will charge from battery with a
constant current of iCH = 0.8 A. During energy exchange mode, HESS disconnected
from DC microgrid which is now evident from voltage waveform not being at nominal
voltage of 20 V.

9 Conclusion

In this chapter, two different control schemes for the two-input bidirectional converter
for HESS control. The performance of designed controllers are analyzed in various
test cases for the application of voltage regulation in DC microgrid. Both the
controllers are effectively stabilize DC microgrid against disturbances from step
change in PV generation as well as load variations. It could utilize inherent fast
dynamics of supercapacitor for absorbing incoming transients to the microgrid. This
unified controller proved to be enough for both charging and discharging operation
of HESS. Moreover, decoupled, separate and independent control of supercapacitor
and battery power was achieved as well as power flow between them. It can also be
used in the case of hybrid electric vehicle applications where two or more sources
supply power. The performance evaluation of MPC over PI control scheme with step
change in PV generation and load demand for peak overshoot and settling time to
restore DC grid voltage are tested.

References

1. Xiao J, Wang P, Setyawan L (2016) Implementation of multiple-slack-terminal DC microgrids

for smooth transitions between grid-tied and Islanded States. IEEE Trans Smart Grid 7(1):273–
281. https://doi.org/10.1109/TSG.2015.2445834
2. Karbalaye Zadeh M, Gavagsaz-Ghoachani R, Pierfederici S, Nahid-Mobarakeh B, Molinas M
(2016) Stability analysis and dynamic performance evaluation of a power electronics-based DC
distribution system with active stabilizer. IEEE J Emerg Sel Top Power Electron 4(1):93–102.
https://doi.org/10.1109/JESTPE.2015.2484218
3. Graditi G, Ippolito MG, Telaretti E, Zizzo G (2015) An innovative conversion device to the
grid interface of combined RES-based generators and electric storage systems. IEEE Trans Ind
Electron 62(4):2540–2550. https://doi.org/10.1109/TIE.2014.2336620
4. Choi ME, Kim SW, Seo SW (2012) Energy management optimization in a
battery/supercapacitor hybrid energy storage system. IEEE Trans Smart Grid 3(1):463–472.
https://doi.org/10.1109/TSG.2011.2164816
5. Feng X, Gooi HB, Chen SX (2014) Hybrid energy storage with multimode fuzzy power allo-
cator for PV systems. IEEE Trans Sustain Energy 5(2):389–397. https://doi.org/10.1109/TSTE.
2013.2290543
A Comparative Analysis of PI and Predictive Control Strategy … 219

6. Gao L, Dougal RA, Liu S (2005) Power enhancement of an actively controlled

battery/ultracapacitor hybrid. IEEE Trans Power Electron 20(1):236–243. https://doi.org/10.
1109/TPEL.2004.839784
7. Gee AM, Robinson FVP, Dunn RW (2013) Analysis of battery lifetime extension in a small-
scale wind-energy system using supercapacitors. IEEE Trans Energy Convers 28(1):24–33.
https://doi.org/10.1109/TEC.2012.2228195
8. Zheng JP, Jow TR, Ding MS (2001) Hybrid power sources for pulsed current applications.
IEEE Trans Aerosp Electron Syst 37(1):288–292. https://doi.org/10.1109/7.913688
9. Punna S, Manthati UB (2020) Optimum design and analysis of a dynamic energy management
scheme for HESS in renewable power generation applications. SN Appl Sci 2:495. https://doi.
org/10.1007/s42452-020-2313-3
10. Inthamoussou FA, Pegueroles-Queralt J, Bianchi FD (2013) Control of a supercapacitor energy
storage system for microgrid applications. IEEE Trans Energy Convers 28(3):690–697. https://
doi.org/10.1109/TEC.2013.2260752
11. Laldin O, Moshirvaziri M, Trescases O (2013) Predictive algorithm for optimizing power flow
in hybrid ultracapacitor/battery storage systems for light electric vehicles. IEEE Trans Power
Electron 28(8):3882–3895. https://doi.org/10.1109/TPEL.2012.2226474
12. Wang B, Xian L, Manandhar U, Ye J, Ukil A, Beng Gooi H (2017) A stand-alone hybrid pv/fuel
cell power systems using single-inductor dual-input single-output boost converter with model
predictive control. In: 2017 Asian conference on energy, power and transportation eletrification
(ACEPT), Oct 2017, pp 1–5. https://doi.org/10.1109/ACEPT.2017.8168569
13. Zhao C, Round SD, Kolar JW (2008) An isolated three-port bidirectional DC-DC converter
with decoupled power flow management. IEEE Trans Power Electron 23(5):2443–2453. https://
doi.org/10.1109/TPEL.2008.2002056
14. Zeng J, Qiao W, Qu L, Jiao Y (2014) An isolated multiport DC–DC converter for simultaneous
power management of multiple different renewable energy sources. IEEE J Emerg Sel Top
Power Electron 2(1):70–78. https://doi.org/10.1109/JESTPE.2013.2293331
15. Wang Y, Han F, Yang L, Xu R, Liu R (2018) A three-port bidirectional multi-element resonant
converter with decoupled power flow management for hybrid energy storage systems. IEEE
Access 6. https://doi.org/10.1109/ACCESS.2018.2872683
16. Yang P, Tse CK, Xu J, Zhou G (2015) Synthesis and analysis of double-input single-output
DC/DC converter. IEEE Trans Ind Electron 62(10):6284–6295. https://doi.org/10.1109/TIE.
2015.2424396
17. Khaligh A, Cao J, Lee YJ (2009) A multiple-input DC–DC converter topology. IEEE Trans
Power Electron 24(3):862–868. https://doi.org/10.1109/TPEL.2008.2009308
18. Wang B, Xian L, Manandhar U, Ye J, Zhang X, Gooi HB, Ukil A (2019) Hybrid energy storage
system using bidirectional single-inductor multipleport converter with model predictive control
in DC microgrids. Electr Power Syst Res 173(April-2019):38–47. https://doi.org/10.1016/j.
epsr.2019.03.015
19. Faraji R, Farzanehfard H (2018) Soft-switching nonisolated high step-up three-port DC-DC
converter for hybrid energy systems. IEEE Trans Power Electron 33. https://doi.org/10.1109/
TPEL.2018.2791840
20. Sato Y, Uno M, Nagata H (2020) Nonisolated multiport converters based on integration of
PWM converter and phase-shift-switched capacitor converter. IEEE Trans Power Electron
35(1). https://doi.org/10.1109/TPEL.2019.2912550
21. Yi Z, Dong W, Etemadi AH (2018) A unified control and power management scheme for
PV-battery-based hybrid microgrids for both grid-connected and islanded modes. IEEE Trans
Smart Grid 9(6). https://doi.org/10.1109/TSG.2017.2700332
22. Behjati H, Davoudi A (2013) Power budgeting between diversified energy sources and loads
using a multiple-input multiple-output DC-DC converter. IEEE Trans Ind Appl 49(6):2761–
2772. https://doi.org/10.1109/TIA.2013.2265251
23. Filsoof K, Lehn PW (2016) A bidirectional multiple-input multiple-output modular multilevel
DC-DC converter and its control design. IEEE Trans Power Electron 31(4):2767–2779. https://
doi.org/10.1109/TPEL.2015.2448112
220 S. Punna et al.

24. Wai RJ, Lin CY, Chen BH (2012) High-efficiency DC-DC converter with two input power
sources. IEEE Trans Power Electron 27(4):1862–1875. https://doi.org/10.1109/TPEL.2011.
2170222
25. Nejabatkhah F, Danyali S, Hosseini SH, Sabahi M, Niapour SM (2012) Modeling and control
of a new three-input DC–DC boost converter for hybrid PV/FC/battery power system. IEEE
Trans Power Electron 27(5):2309–2324. https://doi.org/10.1109/TPEL.2011.2172465
26. Wai RJ, Chen BH (2014) High-efficiency dual-input interleaved DC-DC converter for reversible
power sources. IEEE Trans Power Electron 29(6):2903–2921. https://doi.org/10.1109/TPEL.
2013.2275663
27. Chen CW, Liao CY, Chen KH, Chen YM (2015) Modeling and controller design of a semi-
isolated multiinput converter for a hybrid PV/wind power charger system. IEEE Trans Power
Electron 30(9):4843–4853. https://doi.org/10.1109/TPEL.2014.2367594
28. Danyali S, Hosseini SH, Gharehpetian GB (2014) New extendable single-stage multi-input
DC–DC/AC boost converter. IEEE Trans Power Electron 29(2):775–788. https://doi.org/10.
1109/TPEL.2013.2256468
29. Guo F, Fu L, Zhang X, Yao C, Li H, Wang J (2016) A family of quasi-switched-capacitor
circuit-based dual-input DC/DC converters for photovoltaic systems integrated with battery
energy storage. IEEE Trans Power Electron 31(12):8237–8246. https://doi.org/10.1109/TPEL.
2016.2519394
30. Gummi K, Ferdowsi M (2010) Double-input DC–DC power electronic converters for electric-
drive vehicles—topology exploration and synthesis using a single-pole triple-throw switch.
IEEE Trans Ind Electron 57(2):617–623. https://doi.org/10.1109/TIE.2009.2032762
31. Ahrabi RR, Ardi H, Elmi M, Ajami A (2017) A novel step-up multiinput DC–DC converter
for hybrid electric vehicles application. IEEE Trans Power Electron 32(5):3549–3561. https://
doi.org/10.1109/TPEL.2016.2585044
32. Akar F, Tavlasoglu Y, Ugur E, Vural B, Aksoy I (2016) A bidirectional nonisolated multi-
input DC–DC converter for hybrid energy storage systems in electric vehicles. IEEE Trans Veh
Technol 65(10):7944–7955. https://doi.org/10.1109/TVT.2015.2500683
33. Nahavandi A, Hagh MT, Sharifian MBB, Danyali S (2015) A nonisolated multiinput multi-
output DC–DC boost converter for electric vehicle applications. IEEE Trans Power Electron
30(4):1818–1835. https://doi.org/10.1109/TPEL.2014.2325830
34. Hintz A, Prasanna UR, Rajashekara K (2015) Novel modular multiple-input bidirectional
DC–DC power converter (MIPC) for HEV/FCV application. IEEE Trans Industr Electron
62(5):3163–3172. https://doi.org/10.1109/TIE.2014.2371778
35. Kollimalla SK, Mishra MK (2014) A novel adaptive p&o mppt algorithm considering sudden
changes in the irradiance. IEEE Trans Energy Convers 29(3):602–610. https://doi.org/10.1109/
TEC.2014.2320930
36. Erickson RW, Maksimovic D (2007) Fundamentals of power electronics. Springer Science and
Business Media. https://doi.org/10.1007/978-3-030-43881-4
37. Prasanth S, Manthati UBVSP, Punna S (2019) Efficient interleaved buck converter driver for
LED applications. In: 2019 IEEE international conference on sustainable energy technologies
and systems (ICSETS), Bhubaneswar, India, 2019, pp 206–212. https://doi.org/10.1109/ICS
ETS.2019.8744943
38. Arunkumar CR, Manthati UB (2019) Design and small signal modelling of battery-
supercapacitor HESS for DC microgrid. In: TENCON 2019—2019 IEEE region 10 conference
(TENCON), Kochi, India, pp 2216–2221. https://doi.org/10.1109/TENCON.2019.8929544
Parameters Estimation of Solar PV Using
Jaya Optimization Technique

C. Srinivasarathnam, Gurappa Battapothula, Anil Annamraju,

and Chandrasekhar Yammani

Abstract Integration of Renewable Energy Sources into the Power Systems is

increasing day-by-day because of greater importance to environmental protection,
in addition to gradual depletion of fossil fuels. Among all the Renewable Energy
Sources, Photo-Voltaic (PV) System has become most popular due to its abundant
availability, clean and environmentally pollution-free, noise-free and low mainte-
nance cost. For efficient scheduling of PV system, to meet the load demand and also
to reduce the uncertainty, accurate estimation of PV power generation is essential.
Thus, estimating PV system parameters accurately has become a research topic of
great interest. The mathematical model of the PV system is non-linear in nature. The
data provided by the PV manufacturer is inadequate for accurate system analysis.
In view of the above, in this chapter, a novel algorithm-specific parameter-free opti-
mization, named Jaya Algorithm is used to predict the Solar PV modules parameters
accurately for three different technologies i.e., Poly-crystalline (Shell S75), Mono-
crystalline (Shell SM55) and Thin film (Shell ST36). The MATLAB simulation is

C. Srinivasarathnam (B)
Department of Electrical and Electronics Engineering, Vasavi College of Engineering,
Ibrahimbagh, Hyderabad, Telangana, India
e-mail: [email protected]
G. Battapothula
Department of Electrical and Electronics Engineering, B.V. Raju Institute of Technology,
Narsapur, Telangana, India
e-mail: [email protected]
A. Annamraju
Department of Electrical and Electronics Engineering, School of Engineering, Malla Reddy
University, Telangana, India
e-mail: [email protected]
C. Yammani
Department of Electrical Engineering, National Institute of Technology, Warangal, Telangana,
India
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 221
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6_9
222 C. Srinivasarathnam et al.

carried out for evaluating the PV module parameters for various environmental condi-
tions. The effectiveness of the proposed method has been validated upon comparison
with other optimization techniques available in the literature and with experimental
data.

Keywords Renewable energy · Solar PV power · Parameters estimation ·

Single-diode model · Optimization

Nomenclature

RES Renewable Energy Sources

GDP Gross Domestic Product
GW Giga Watts
kWh Kilo Watt Hour
PV Photo-Voltaic
RMSE Root Mean Square Error
AE Absolute Error
a Diode ideality factor
Rs Series resistance
RP Shunt Resistance
I0 Diode reverse saturation current
I ph Photo current generated by the module
ω Inertia weight
Vitj Particle ‘i’ velocity towards dimension ‘j’ at iteration ‘t’
X it j Position vector of particle ‘i’ in dimension ‘j’ at iteration t
t
Pbest,i Personal best position of particle ‘i’ in the iteration t
G best Global best position of all particle through all iterations
C1 and C2 Positive acceleration constants which are used to level the contribution
of the cognitive and social components respectively
t t
r1j and r2j Random numbers from uniform distribution at iteration t

1 Introduction

Human life is very much dependent on energy. Industries, commercial, and day-to-
day activities cannot be progressed without energy. However, energy is available in
different forms such as Chemical energy, Thermal energy, Radiant energy, Nuclear
energy, Electrical energy, Motion energy, etc., as shown in Fig. 1 [1]. According to
the “Law of Conservation of energy,” energy can be transformed from one form to
another, but neither can it be created nor destroyed. Out of various forms of energy,
electrical energy is the most important as it can efficiently be generated (converted
Parameters Estimation of Solar PV … 223

Fig. 1 Various forms of energy

from other forms of energy), easily transmitted, and for a reasonable cost, it can
be utilized. The ease of transmission of electrical energy gives rise to a possibility
of generating electrical energy in bulk at centralized places and transmit it over a
long distance, to be used ultimately by a large number of users [2]. Because of ease
in generation, transmission, and utilization, throughout the World, the demand for
electrical energy is increasing day-by-day.
As stated in [3], electrical power systems is a technical wonder, and according
to the National Academy of Engineering, electricity and its accessibility are the
greatest engineering achievements of the twentieth century, ahead of computers and
airplanes. The authors in [4, 5] stated that adult literacy rate, life expectancy at birth,
GDP per capita (the level of economic development), consumption expenditure per
capita, urbanization rate are the five indices that reflect the human development.
The per capita electricity consumption should be strengthened to enhance the level
of welfare of society and human development. Therefore, special attention is being
paid to the generation of electricity. India has installed a capacity of 1.36 GW during
Independence whereas it has raised to 365 GW as on October 31, 2019 [6]. As a
reflection, the per capita energy consumption of electricity in India has grown from
16 kWh during 1946–47 to 1181 kWh in the year 2018–19, as shown in Fig. 2. The
per capita energy consumption of electricity by various countries is shown in Fig. 3.
It is clear from Fig. 3 that the per capita energy consumption of electricity in India
is much lower than the developed countries and is lesser than the world’s average
consumption. Further, the worlds’ average electricity consumption is much lower
than the developed countries’ electricity consumption. Thus, there is a need to think
of supplying reliable power to the consumer at an affordable price.
224 C. Srinivasarathnam et al.

Fig. 2 Growth in per capita consumption of electrical energy in India

Fig. 3 Per capita energy consumption of electrical energy in various countries

In addition to the above problem, for developing countries with spurring social
and economic progress, it is pivotal to have affordable and reliable electricity in rural
areas. It is proclaimed by Ministry of Power, Government of India [7] that still 13.90
lakhs households in India are yet to be electrified as on 31.10.2019, and as per the
World Bank report, the percentage of the World’s population having access of elec-
tricity up to 2018 is 89.59% [8] as depicted in Fig. 4. The Indian Government, by grid
expansion, has made remarkable progress on rural electrification. However, copious
householders have no electricity access. Thus, it is a challenge to the power systems
engineers to supply electrical power to the remote locations where geographically
grid expansion is not feasible.
The power generation from the conventional energy sources encounters challenges
such as run out of fossil fuels such as coal and oil, change in climate conditions,
continuous increase in energy demand [9]. Further, combustion of fossil fuels releases
harmful gases which include Carbon dioxide, Sulphur dioxide, Nitrogen oxide, which
causes global warming and acid rains. To address these challenges faced by the power
Parameters Estimation of Solar PV … 225

Fig. 4 Access of electricity to percentage of population in the world

systems engineers, researchers are focusing on alternate energy sources. Considering

the capital investment and environment condition, Renewable Energy Sources (RES)
are found to be the best alternate [10]. Among the RES (Solar power and Wind
power), Solar Photo-Voltaic (PV) power generation is deemed to be popular around
the world being ubiquitous in urban and rural areas, its abundant availability in nature,
long lifecycle time, clean and environmentally pollution-free, noise-free, negligible
maintenance as it does not rely on moving parts and reliable [11]. Further, modularity
is the salient feature of PV technology, wherein the existing PV system can be easily
upgradable within a short period of time, without replacing the entire existing system.
The PV system is found to be ideal for earth applications as well as space applications
due to its flexibility and benefits [12, 13].
Optimal utilization of solar energy is essential due to its high initial investment
cost. Thus, the proper design of the PV system is crucial for increasing system
efficiency. The accuracy of predicting the solar panel characteristics is decided by
the determination of PV model parameters exactly. An accurate estimation of the
PV system is necessary for (a) Demand side management in power systems, (b)
Design of Maximum Power tracking of Solar system (c) Efficiency estimation, at
different weather conditions [14]. In view of this, parameter estimation of the PV
cell accurately has become an interesting topic for the researchers today.
To determine the performance and non-linear behavior of the PV system, several
mathematical models are available in the literature. Among the available methods,
single diode modeling and two diode modeling are popular [15]. Due to simplicity
in modeling and a reduced number of model components, the single diode model is
mostly preferred. Furthermore, the parameters extraction methods of any model have
been categorized into (a) Deterministic algorithms and (b) Stochastic algorithms.
Deterministic methods normally use algebraic manipulation which involves
numerical iterations. The traditional optimization algorithms, such as Dynamic
Programming, Nonlinear Programming, Geometric Programming, Sequential
226 C. Srinivasarathnam et al.

Programming, etc., have certain limitations in their search mechanism. The search
mechanism of these algorithms depends on the type of objective and constraint func-
tions, modeling of variables type. The efficiency of these algorithms relies on the
solution space size, the structure of solution space, the number of variables. The
above problems associated with traditional techniques have been surmounted with
population-based optimization techniques over the past two decades.
The authors in reference [16] addressed the PV parameters estimation for three
different PV modules, which are multi-crystalline (Kyocera KC200GT), monocrys-
talline (Shell SQ80), and thin-film (Shell ST40). An Iterative method based on the
Gauss-Siedel iterative resolution algorithm has been attempted on a Single diode
model and an analytical method tested on two diode model to evaluate the PV param-
eters accurately. The author has considered both Absolute Error (AE) and Root Mean
Square Error (RMSE) to validate the reliability of the methods. In reference [17],
an iterative and the Lambert W function has been attempted by the authors for PV
parameters estimation of single diode model. The authors have tested the perfor-
mance of the proposed model on multi-crystalline, mono-crystalline and thin-film
models. RMSE has been considered to address the accuracy of the proposed model
results with experimental results. A non-linear least-squares optimization method
which depends on the modification of Newton method with Levenberg parameters
has been attempted by the author in reference [18].
Though computationally deterministic methods are fast, certain approximations
are required, in order to ensure the model is analytically manageable, which leads
to compromise in the accuracy of the solution. Moreover, the analytical methods
are tedious, computationally complex and time consuming [11]. Thus, now-a-days,
most of the researchers are opting stochastic algorithms.
Stochastic algorithms such as Genetic Algorithm(GA) [19], Particle Swarm Opti-
mization(PSO) [20], Enhanced Learner Particle Swarm Optimization [21] Teaching
Learning Based Optimization(TLBO) [22], Simulation Annealing (SA), Jaya algo-
rithm [23] etc., generate population randomly in the search space and these population
get update their position randomly to identify the best solution in the search space.
The author in [24] attempted Shuffled Frog Leaping Algorithm (SFLA) technique
to evaluate the PV parameters considering Single diode model. The models examined
by the author are Kyocera KC200GT and SolarexMSX-60. Power error function has
been treated as an objective function such that the power output at MPP by the
proposed method and the experimental method are to be closure. It was proved
that simulation results by SFLA coincide with the experimental details. The author
has proclaimed that SFLA results are superior that Genetic Algorithm and Iterative
modeling algorithm.
In reference [11], the authors have attempted the Bacterial Foraging Algorithm to
estimate the PV parameters of single diode model. The results obtained are compared
with the Genetic Algorithm and Artificial Immune System. The authors have claimed
that the proposed algorithm is superior in terms of accuracy, consistency, speed of
convergence and absolute error.
Differential evolutionary (DE) algorithm has been addressed by the authors in [25]
to evaluate the PV parameters. The parameters Diode ideality factor (a) and series
Parameters Estimation of Solar PV … 227

resistance (Rs ) are evaluated using DE and other parameters Shunt Resistance (R P ),
Diode reverse saturation current (I0 ) and Photocurrent generated by the module (I ph )
are evaluated analytically. The authors proclaimed that by using DE and analytical
method, the computation burden had reduced drastically and substantial improvement
had been found in the results.
Based on the above literature, considering the advantages of the meta-heuristic
techniques and to reduce the computational burden in evaluating the PV parameters,
in this chapter, both meta-heuristic methods and analytical methods are considered
together to evaluate the PV parameters of a single diode model. To validate the results,
different technologies of PV panels are considered, which are (1) Poly-crystalline
(Shell S75), (2) Mono-crystalline (Shell SM55) and (3) Thin-film (Shell ST36).
The remainder of this chapter is organized as follows: Sect. 2 describes Mathe-
matical modeling of PV model. The problem formulation of parameters estimation of
single diode model is presented in Sect. 3. Meta-heuristic techniques applied to solve
the parameter estimation problem are PSO and Jaya algorithm, which are described
in Sects. 4 and 5 respectively. Section 6 presents the algorithm and flow chart respec-
tively using PSO and Jaya algorithms for the proposed problem. Simulation results
and analysis are presented in Sect. 7. Section 8 summarizes the major contributions
with concluding remarks.

2 Mathematical Modeling of PV Model

Due to average complexity and accuracy in results, the single diode model is mostly
used electrical model to represent the characteristics of a solar PV module. The
ideal model of a Solar PV module is a Current source (IPV ) in parallel with a diode.
However, due to leakage current in the semiconductor device, a shunt resistance (RP )
is added in parallel to the device. Further, due to resistance offered for the current flow
by the semiconductor material, metal grid in the module and leads contact resistance,
a resistance in series (RS ) is added to the model. Thus, the practical Solar PV model
is represented in Fig. 5. From the Fig. 5, by applying KCL, we get the current (I ) as
per Eq. (1).

Fig. 5 Single-diode PV
equivalent circuit
228 C. Srinivasarathnam et al.

V
I = IPV − ID − (1)
RP

where V is the terminal voltage/voltage across the diode, ID is the diode current, Ipv
is the Photo current generated by the module.
The general form of a diode current equation is given by Eq. (2).
 V 
I D = Io e aVt − 1 (2)

where IO is the diode reverse saturation current, a is diode ideality factor and Vt is
the thermal voltage across the diode, which is given by Eq. (3).

KT
Vt = (3)
q

where q is charge of electron which is given by −1.602x10−19 C, T is cell temperature

measured in Kelvin, K is Boltzmann constant given by 1.380649x10−23 J.K−1 . By
applying KVL in Fig. 5, we get

VD = V + I R S (4)

where VD is the voltage across the diode. Thus, the parameters to be estimated in
this single diode model are Diode ideality factor (a), Series resistance (Rs ), Shunt
Resistance (RP ), Diode reverse saturation current (I0 ) and Photo current generated
by the module (Ipv ).

3 Problem Formulation

As illustrated above, in the single diode model, five parameters a, Rs , RP , I0 and Iph
need to be evaluated for modelling the solar PV characteristics. In this chapter, both
analytical method and optimization techniques are used to evaluate these parame-
ters. Shunt resistance (RP ), diode reverse saturation current (I0 ) and Photocurrent
generated by the module (Iph ) are determined analytically to reduce the computa-
tional complexity and diode ideality factor (a), Series resistance (Rs ) are evaluated
using an optimization technique. These parameters are assessed such that the error
between the calculated value and experimental values is maintained as minimum as
possible. Also, due to the change in environmental conditions particularly tempera-
ture and irradiance, these parameter values get varied. The parameters of the Solar PV
model need to be evaluated such that the error is minimum under different operating
conditions.
Parameters Estimation of Solar PV … 229

From the manufacturer’s data sheet, the known parameters of the model are short
circuit current (ISC ), open circuit voltage (VOC ), Maximum power point operating
current (IM ) and operating voltage (VM ) and current temperature coefficient (KI ).
The photocurrent of a PV panel is determined using the Eq. (5) which is defined as
follows.
G  
IPV = I SC,ST C + K I (T − TST C ) (5)
G ST C

where ISC,STC is short circuit current at STC, G is irradiance at which surface of the
panel is exposed to environment, GSTC is irradiance at STC i.e., 1000 W/m2 , TSTC
is temperature at STC i.e., 25 °C. During open circuit condition, the output voltage
and current values are

V = VOC (6)

I =0 (7)

Substituting Eqs. (6) and (7) in Eq. (1), which gives

VOC
0 = IPV − ID − (8)
RP

On substituting Eqs. (2) in (8), we get Eq. (9) as below

 Voc  V
oc
0 = I P V − Io e aVt − 1 − (9)
RP

From the above Eq. (9), IO can be evaluated as follows,

I P V − VRocP
Io =  Voc  (10)
e aVt − 1

The IO in Eq. (10) depends on RP , which is an unknown parameter. Thus, to make

IO independent of RP , the following procedure is followed. We know, during MPP
condition,

V = VM (11)

I = IM (12)
230 C. Srinivasarathnam et al.

On substituting the following conditions in Eq. (1), we get

 VM +I M RS  V +I R
M M S
I M = I P V − Io e aVt − 1 − (13)
Rp

From the Eq. (13), RP is evaluated as

VM + I M R S
RP =  VM +I M RS  (14)
IPV − I0 e aVt − 1 − I M

On substituting Eqs. (14) in (10), we get an equation for IO which only depends
on known parameters and independent of unknown parameters. Thus, IO is defined
as per Eq. (15).

I P V + VM +I
Voc
R (I M
− IPV )
I O =  Voc   VMM+I MSRS   (15)
e aVt − 1 − e aVt − 1 VM +I Voc
M RS

3.1 Evaluation of Fitness Function

From the P-V characteristics of a solar panel, at Maximum Power Point (MPP),
the slope is equals to zero, irrespective of irradiance and temperature condition.
Therefore, the slope at MPP can be written as
 
dP
=0 (16)
dV MPP

We know from the characteristics that power (P) is given as product of voltage
(V) and current (I). Thus,

P = VI (17)

Upon derivating the Eq. (17) with respect to V, we get

 
dP dI
= V +I =0 (18)
dV MPP dV

Considering the Eq. (18) as a fitness function (J ), the objective is to minimize J ,

which is defined as per Eq. (19).
   
dP  d IM I M 
J= =  V+ =0 (19)
dI MPP d VM VM 
Parameters Estimation of Solar PV … 231

dIM
The term dVM
in Eq. (19) can be evaluated as follows. From the Eq. (13),

 VM +I M RS  V +I R
M M S
I M = I P V − Io e aVt − 1 − (20)
Rp

Upon derivating the Eq. (20) w.r.t VM , we get

V M +I M R S

d IM − aVIO
e aVt − R1P
= t
V M +I M R S (21)
d VM Io
R e aVt − RRPs
aVt s

From the Eq. (21), fitness function can be evaluated such that, objective function
value must be minimum. This can be achieved by applying stochastic algorithms as
discusses above. To analyze the performance of the proposed algorithm, Root Mean
Square Error (RMSE), an index, which is defined in Eq. (22) has been utilized, which
depends on experimental current (Iexp ) and the current obtained using optimization
technique, denoted as (Imeas ) at a given voltage. Minimum the value of RMSE, the
closure is the measured value of the current to the experimental value.


N  2
i=1 Iexp,i − Imeas,i
RMSE = (22)
N

4 Particle Swarm Optimization

The Particle Swarm Optimization (PSO) algorithm [26, 27] is a multi-agent parallel
search technique. In this optimization technique, a swarm of particles are generated
in the search space randomly and each particle represents a potential solution in
the swarm. These particles move in a multi-dimensional search space, such that the
particles adjust their position to find the global best solution, based on their own
experience and that of their neighbors.
In this optimization technique, each individual solution i = (1, 2, 3, . . . .n), n > 1
is called as particle. Each particle will have a current position in search space
Xi . These particles will move to evaluate the global best solution with a current
velocityVi . Each particle throughout its journey will have a best solution, identi-
fied as personal best, Pbest,i . This personal best Pbest,i can be treated as a minimum
objective function value or maximum objective function value in the search space
by this particle ‘i’ alone, for minimization function or for maximization function
respectively. A swarm of such particles i = (1, 2, 3, . . . .n), n > 1, will have each
individual best positions and the particle having lowest objective function value or
232 C. Srinivasarathnam et al.

highest objective function value can be treated as Global best solution for mini-
mization function or for maximization function respectively, which is denoted as
Gbest .
During the iterative process of the optimization technique, the personal best Pbest,i
and the Global best Gbest values are updated. The mathematical representation of
evaluating the personal best position of a particle ‘i’ for the next iteration (t + 1) is
defined as per Eq. (23) for objective function of minimization type.
  
t+1
Pbest,i = Pbest,i
t
if f xit+1 > Pbest,i
t
(23)
t+1
xi if f (xi ) ≤ Pbest,i
t+1 t

where f is the objective function.

Similarly, the Global best position of the solution set for the next iteration is
evaluated as per Eq. (24), which is defined as follows.
 t+1 
G best = min Pbest,i (24)

where i ∈ [1,2,…n] and n > 1.

From the above, it is important to understand that the personal best Pbest,i is the
best position so far achieved by the particle i’, during its journey, on the other hand,
global best position Gbest is the best position of any particle is the entire swarm during
their journey. The velocity and position of particle are updated as per Eqs. (25) and
(26) respectively, which are defined as follows.

Vit+1
j = ωV tij + c1 rt1j [Pbest,i
t
− X it j ] + c2 rt2j [G best − X it j ] (25)

j = X i j + Vi j
X it+1 t t+1
(26)

4.1 PSO Algorithm Parameters

It is a well-known fact that for any optimization technique, there will be a few
algorithm-specific parameters that affect the performance of the algorithm in terms
of finding the global best solution, convergence speed etc. Among all the algorithm-
specific parameters, some of the parameters have a significant impact and the other
parameters have a very small impact on its performance. The algorithm-specific
parameters for PSO algorithm that are noticeable are population size/swarm size,
maximum number of iterations, velocity and acceleration coefficients, inertia weight
etc.
Parameters Estimation of Solar PV … 233

4.1.1 Swarm Size

The number of particles in the swarm is called as population size or swarm size.
However, each particle represents a solution to the objective function. With the
consideration of larger swarm size, the particles cover the larger part of the search
space for identifying the global solution. Thus, the number of iterations required for
identifying global solution shall be reduced with the higher number of swarm size.
However, computational complexity per iteration and time to complete an iteration
will be more by the consideration of huge swarm size. In view of this, in the literature
based on a number of empirical studies, it is recommended to consider a swarm size
of n ∈ [20, 80].

4.1.2 Iteration Numbers

Based on the nature of the objective function/problem, the number of iterations may
vary to identify the global solution. Lower the number of iterations, the optimization
technique may be struck at a local minima/premature convergence. On contrary, a
too large number of iterations makes lead to unnecessary computational complexity
to the system and it requires more time.

4.1.3 Velocity Components

Updation of the velocity of the particle is an important step in the PSO Algorithm. It
is evident from the Eq. (24), the updation equation has three different contributions
to the particle movement.
(a) The term ω is called inertia weight. The contribution of ωVijt term in the
velocity updating Eq. (25) is that it decides the current flight direction of a
particle, based on the memory of the previous flight direction. If ω = 1, then
the current particle velocity is fully influenced by the previous velocity. On the
other hand, if 0 < ω < 1, then the particle velocity is less influenced by the
velocity of the previous flight, thus provide change for exploring more search
area.
(b) The second term in the Eq. (25) is c1 r1j t t
[Pbest,i − Xijt ], which is called the
cognitive component. The basic idea behind this term is that as the particle
Xijt move away from its Pbest
t
position, the difference term indicated in braces
increases and thus the particle is attracted towards its own best position.
(c) The third and the final term in the Eq. (25) is c2 r2jt
[Gbest − Xijt ], which is called
social component or social learning rate. The idea behind this term is to share
the information of best solution achieved by the swarm so far in their flight,
regardless of which particle has reached the best. This component pulls the
particles towards the best solution achieved so far by the swarm.
234 C. Srinivasarathnam et al.

4.1.4 Acceleration Coefficients

The variables C1 and C2 are called acceleration coefficients and are positive values.
The terms C1r1 and C2 r2 represents the stochastic influence of cognitive and social
components of the particle’s velocity respectively. The random number r1 and r2
varying in the region [0,1] randomly. C1r1 weighs the importance of particle’s own
previous experiences whereas C2 r2 weighs the importance of the global learning of
the swarm. In general, the values of C1 and C2 are static in nature throughout the
iterative process. The values of these coefficients are found empirically. Initializing
these values wrongly may lead to divergence of the problem. Based on the literature,
the empirical values for these coefficients are found to be C1 = C2 = 2.

5 Jaya Algorithm

There are different types of population-based algorithms such as Evolutionary Algo-

rithms (EA) and Swarm Intelligence (SI) algorithms. These probabilistic algorithms
require tuning of control parameters, some of which are common parameters of
all algorithms such as population size, number of iterations etc. and the rest are
algorithm-specific parameters such as crossover probability, mutation probability,
elitism probability in the Genetic Algorithm (GA), inertia weight, maximum velocity,
and cognitive parameters in Particle Swarm Optimization (PSO) algorithm, harmony
memory in Harmon Search (HS) algorithm. These parameters play a vital role
in finding the global optimum in the search space and improper tuning of these
parameters leads to premature convergence.
Considering the stated facts, a novel meta-heuristic algorithm which is indepen-
dent of algorithm-specific parameters rather needs only common control parameters
such as population size and the maximum number of iterations i.e., Jaya Algorithm
has been considered in this chapter. Jaya Algorithm was proposed by Prof. R. Venkata
Rao in 2016 and it is a simple and powerful global optimization algorithm. The Jaya
Algorithm procedure is described as follows.
Let ‘f’be the objective function to be minimized. Assuming that there are k number
of design variables of the objective function f . In this algorithm, in the beginning,
generate Psize (Population size) number of initial solutions (X) randomly. The random
solutions must be within the search space, bounded by the lower and upper limits of
the design variables. Now, evaluate the objective function value for all the Psize initial
solutions. Upon evaluation of objective function value, the three common phases of
Jaya Algorithm are to be followed to produce the next generation population.
Parameters Estimation of Solar PV … 235

5.1 Evaluation of the Best Solution Candidate and the Worst

Solution Candidate Phase

In each iteration, the candidate solution which provides minimum objective function
value and maximum objective function value are treated as the Best solution candidate
(Xi,best,k ) and the Worst solution candidate (Xi,worst,k ) among the total population
(Xi,j,k ) respectively. Here ‘i’ stands for iteration number, varies from zero to the
maximum number of iterations/generations, ‘j’ indicates the candidate number in
the population, which varies from one to population size (Psize ) and ‘k’ points out the
design variable, varies from one to the maximum number of design variables of the
objective function. On identification of Best solution candidate and Worst solution
candidate, the candidates have to be modified as specified in the ‘Updation Phase’
which is as follows.

5.2 Updation Phase

On obtaining the Best solution candidate and the Worst candidate solution of the
population, each candidate solution will be updated using the Eq. (27).
     
X i,1 j,k = X i, j,k + ri, j,1 X i,best,k −  X i, j,k  − ri, j,2 X i,wor st,k −  X i, j,k  (27)

where Xi,j,k indicate the value of the kth variable for the jth candidate during the
ith iteration, Xi,best,k and Xi,worst,k represents the Best candidate solution and the
Worst candidate solution in ith iteration respectively, ri,j,1 and ri,j,2 represents the
two random numbers of the kth variable during ith iteration in the range [0,1] and
1
Xi,j,k represents corresponding updated candidate.
 
The term “ri,j,1 Xi,best,k − |Xi,j,k | ” implies that the candidate
 tendency to
 move
towards the Best candidate solution and the term “−ri,j,2 Xi,worst,k − |Xi,j,k | ” indi-
cates the tendency of the candidate to avoid the worst solution. The updated candi-
1
dates (Xi,j,k ) and the previous candidate (Xi,j,k ) are given as inputs to the ‘Comparison
Phase’.

5.3 Comparison Phase

Upon completion of Updation phase, compute the fitness values of X i, j,k and the
updated candidate X i,1 j,k . The fitness value of the candidate solution is evaluated as
per Eq. (28),

1
f itness(X ) = (28)
1 + objective f unction value
236 C. Srinivasarathnam et al.

If the fitness of the updated candidate X i,1 j,k is better than the fitness of X i, j,k ,
then replace the candidate X i, j,k with X i,1 j,k for the next generation, otherwise retain
X i, j,k for the next generation and discard the updated candidate. This comparison
has to be performed for all the candidates of the population.
With the above, one iteration of the Jaya algorithm gets completed. All the fittest
candidates in the ‘Comparison phase’ at the end of the iteration are saved in the
memory and these values become the input to the next iteration. Repeat the above
process until the convergence criterion is satisfied. Thus, the algorithm tries to get
closer to the global optimal.

6 Algorithm for Estimation of Solar PV Parameters

The algorithm for the Solar PV parameters estimation using Jaya Algorithm and
Particle Swarm Optimization Algorithms are described in the following sections.

6.1 Algorithm for Estimation of Solar PV Parameters Using

Jaya Algorithm

1. Read PV module specifications, i.e., short circuit current (I SC ), open circuit

voltage (VOC ), maximum power point operating current (I M ) and operating
voltage (VM ) and current temperature coefficient (K I ).
2. Read irradiance values and temperature values at which parameters to be
estimated.
3. Read Jaya algorithm parameters: population size (Psize ) and maximum number
of iterations (iter_max).
4. Initialize the population randomly for Diode ideality factor (a), Series
resistance (Rs ) in the search space.
5. Set iteration count (iter) = 1.
6. Evaluate the Shunt Resistance (RP ), Diode reverse saturation current (I0 ) and
Photo current generated by the module (Iph ) analytically.
7. Calculate the fitness using Eq. (28).
8. Evaluate the Best fitness candidate and the Worst fitness candidate in the
population.
9. Update the candidate using updation phase as indicated in Eq. (27).
10. Go to Comparison phase and based on updated candidate fitness value either
replace the candidate or retain the candidate.
11. Increment iteration count (iter = iter + 1) and repeat above steps (Step 6–Step
10) until a convergence criterion is met.
12. Evaluate RMSE value using Eq. (22) at different points.
13. Stop the program and print the results.
Parameters Estimation of Solar PV … 237

6.2 Algorithm for Estimation of Solar PV Parameters Using

Particle Swarm Optimization

1. Read PV module specifications, i.e., short circuit current (I SC ), open circuit

voltage (VOC ), maximum power point operating current (I M ) and operating
voltage (VM ) and current temperature coefficient (K I ).
2. Read irradiance values and temperature values at which parameters to be
estimated.
3. Read PSO algorithm parameters: population size (Psize ), maximum number
of iterations (iter_max), Boundaries of velocities (V ), Inertia weight (w),
Cognitive and Social parameters.
4. Initialize the population randomly for Diode ideality factor (a), Series
resistance (Rs ) in the search space.
5. Set iteration count (iter) = 1.
6. Evaluate the Shunt Resistance (RP ), Diode reverse saturation current (I0 ) and
Photo current generated by the module (Iph ) analytically for all the particles.
7. Calculate the fitness using Eq. (28) for all the particles.
t+1
8. Identify Personal best (Pbest,i ) and the Global best (G best ) for the population
using Eqs. (23) and (24).
9. Update the particle velocity using Eq. (25).
10. Check for velocity limits.
11. Update the particle using Eq. (26).
12. Check whether the updated particle is violating their limits. If violated, keep
within their limits
13. Increment iteration count (iter = iter + 1) and repeat above steps (Step 6–Step
12) until a convergence criterion is met.
14. Evaluate RMSE value using Eq. (22) at different points.
15. Stop the program and print the results.

6.3 Flowchart for PV Parameters Estimation

The flowchart for parameters estimation of solar PV module is presented in the Figs. 6
and 7 using Jaya algorithm and PSO algorithm respectively.

7 Results and Analysis

In this section, the results attained for PV parameters using the proposed Jaya Algo-
rithm are demonstrated and are validated by comparison with the well-proven algo-
rithm in the literature named Particle Swarm Optimization (PSO) and the experi-
mental results. The simulation work has been carried out on MATLAB Simulink
238 C. Srinivasarathnam et al.

Fig. 6 Flowchart of parameter estimation of solar PV model using Jaya algorithm

Parameters Estimation of Solar PV … 239

Fig. 7 Flowchart of parameter estimation of solar PV model using PSO algorithm

240 C. Srinivasarathnam et al.

environment 2017a. The simulation work has been carried out on three different
technologies of PV modules namely (1) Poly-crystalline (Shell S75), (2) Mono-
crystalline (Shell SM55) and (3) Thin-film (Shell ST36) and the results are demon-
strated. The optimization parameters considered in this work are (a) Jaya algorithm:
Population size = 80, maximum iterations = 200. (b) PSO algorithm: Population
size = 80, maximum iterations = 200, C1 = 2, C2 = 2, inertia weight w = linearly
decreases from 0.9 to 0.4.
As deliberated, the parameters provided by the manufacturer’s datasheet for a PV
module are short circuit current (ISC ), open-circuit voltage (VOC ), Maximum power
point operating current (IM ) and operating voltage (VM ) and current temperature coef-
ficient (KI ). However, for accurate modeling of PV model characteristics at various
irradiance (G) and operating temperatures (T), these parameters are inadequate and
need to estimate the other parameters of the system, which are Shunt Resistance (RP ),
Series resistance (Rs ), Diode reverse saturation current (I0 ), Photo current generated
by the module (Iph ) and diode ideality factor (a). These parameters can be obtained
using meta-heuristic techniques. However, to reduce the computational complexity
of the system, in this chapter, both meta-heuristic technique and analytical method
have been attempted to address the problem of parameter estimation.
The PV parameters Diode ideality factor (a), Series resistance (Rs ) are generated
randomly in the search space within the limits using optimization techniques. From
the literature, it is noticed that the series resistance value is very small in comparison
to shunt resistance. Thus, the range of initial guess of Rs is taken as [0, 2]. The
range for diode ideality factor is preferred as [1, 2]. Upon initialization of variables,
the other parameters R P , I0 and I ph are evaluated analytically using the equation
described in mathematical modelling Sect. 3. Based on these values, fitness function
value has arrived. Several runs of optimization technique have been performed to get
the global solution in the search space.
From the simulation results of Poly-crystalline (Shell S75) model, the plots of
I-V characteristics and P-V characteristics for constant irradiance of 1000 W/m2
and varying temperature (T = 20 °C , T = 40 °C and T = 60 °C ) are illustrated in
Figs. 8 and 9 respectively. Similarly, Figs. 10 and 11 depict the I-V characteristics
and P-V characteristics for constant temperature of 25 °C and varying irradiance (G
= 1000, G = 800, G = 600, G = 400 and G = 200 W/m2 ). From these plots, it is
evident that the proposed Jaya Algorithm is exhibiting characteristics very near to the
experimental results obtained. Table 1 presents the experimental current values (Iexp )
and current measured (Imeas ) by Proposed Jaya Algorithm and PSO Algorithm for
different voltage values. From the Table 1, it is clear that the RMSE value obtained
by the proposed Jaya Algorithm is 0.0008259, which is much lower than the RMSE
value of 0.02784 attained by the PSO Algorithm.
Figures 12 and 13 illustrates the I-V characteristics and P-V characteristics respec-
tively for constant irradiance of 1000 W/m2 and varying temperature (T = 20, T =
40 and T = 60 °C) of Mono-crystalline (Shell SM55) model. Similarly, Figs. 14 and
15 depict the I-V characteristics and P-V characteristics for constant temperature of
25 °C and varying irradiance (G = 1000, G = 800, G = 600, G = 400 and G =
200 W/m2 ). Thus, from these plots, it is evident that the proposed algorithm works
Parameters Estimation of Solar PV … 241

Fig. 8 V-I characteristics at varying temperature and constant irradiance for S75 model

Fig. 9 P-V characteristics at varying temperature and constant irradiance for S75 model

Fig. 10 V-I characteristics at varying irradiance and constant temperature for S75 model
242 C. Srinivasarathnam et al.

Fig. 11 P-V characteristics at varying Irradiance and constant temperature for S75 model

better for the Mono-crystalline shells. Table 2 presents the Iex p and Imeas by Proposed
Jaya Algorithm and PSO Algorithm for different voltage values. From the Table 2, it
is clear that the RMSE value obtained by the proposed Jaya Algorithm is 0.0062505
which is lower than the RMSE value of 0.0070137 attained by the PSO Algorithm
by 10.88%.
From the simulation results of Thin-film (Shell ST36) model, the plots of I-V
characteristics and P-V characteristics for constant irradiance of 1000 W/m2 and
varying temperature (T = 20, T = 40 and T = 60 °C) are illustrated in Figs. 16 and
17 respectively. On the other hand, Figs. 18 and 19 depicts the I-V characteristics and
P-V characteristics for constant temperature of 25 °C and varying irradiance (G =
1000, G = 800, G = 600, G = 400 and G = 200 W/m2 ). By a close inspection of these
characteristics infers that the proposed Jaya algorithm yields most accurate results.
Table 3 illustrates the measured current and calculated current values at different
voltage levels. It can be noticed that the proposed Jaya Algorithm is producing an
RMSE index of 0.003242 lower than that of the PSO Algorithm. Table 4 illustrates
the parameter values obtained for different PV technologies S75, SM55 and ST36
using PSO algorithm and the proposed Jaya algorithm at STC. The variation of model
parameters under different irradiation and temperature conditions using PSO algo-
rithm and proposed Jaya algorithm for S75, SM55 and ST36 models are presented
in Tables 5, 6 and 7 respectively.
It is proclaimed in the literature that the values of Rs and R P must increase
gradually with decreasing irradiance level and the same is observed in our results for
various PV modules. Thus, from the above results, it is concluded that the proposed
algorithm works more accurately in tracing the experimental results in comparison
to the PSO Algorithm.
Parameters Estimation of Solar PV … 243

Table 1 RMSE values of PSO algorithm and the proposed Jaya algorithm for S75 PV module
Measured data PSO algorithm Proposed Jaya algorithm
V P V (V ) I P V (A) IC AL (A) I IC AL (A) I
0 4.696904 4.694773 0.002132 4.69679 0.000114
0.431 4.694183 4.69195 0.002232 4.694039 0.000144
0.864 4.691448 4.689115 0.002333 4.691275 0.000173
1.297 4.688713 4.686279 0.002434 4.688511 0.000202
1.73 4.685978 4.683443 0.002535 4.685746 0.000232
2.163 4.683243 4.680607 0.002636 4.682982 0.000261
2.596 4.680508 4.677772 0.002736 4.680217 0.000291
3.029 4.677772 4.674935 0.002836 4.677452 0.00032
3.462 4.675035 4.672099 0.002936 4.674685 0.000349
3.895 4.672297 4.669262 0.003035 4.671918 0.000379
4.328 4.669558 4.666424 0.003133 4.66915 0.000408
4.761 4.666817 4.663586 0.003231 4.66638 0.000437
5.194 4.664073 4.660746 0.003327 4.663608 0.000466
5.627 4.661327 4.657905 0.003422 4.660832 0.000495
6.06 4.658575 4.655061 0.003514 4.658051 0.000524
6.493 4.655816 4.652213 0.003603 4.655264 0.000552
6.926 4.653049 4.64936 0.003688 4.652468 0.00058
7.359 4.650268 4.6465 0.003768 4.64966 0.000608
7.792 4.64747 4.64363 0.003841 4.646835 0.000636
8.225 4.644649 4.640744 0.003904 4.643986 0.000662
8.658 4.641794 4.637838 0.003955 4.641105 0.000688
9.091 4.638893 4.634902 0.00399 4.63818 0.000713
9.524 4.635929 4.631925 0.004004 4.635192 0.000736
9.957 4.632878 4.628887 0.003991 4.63212 0.000758
10.39 4.629707 4.625764 0.003943 4.62893 0.000777
10.823 4.626371 4.622521 0.00385 4.625578 0.000793
11.256 4.622806 4.619106 0.0037 4.622002 0.000804
11.689 4.618928 4.615448 0.00348 4.618117 0.000811
12.122 4.614616 4.611445 0.003172 4.613805 0.000812
12.555 4.609708 4.606952 0.002757 4.608904 0.000805
12.988 4.603978 4.601764 0.002214 4.60319 0.000788
13.421 4.597115 4.59559 0.001525 4.596355 0.00076
13.854 4.588693 4.588019 0.000674 4.587975 0.000718
14.287 4.578125 4.578469 0.000344 4.577464 0.000661
14.72 4.564605 4.566121 0.001515 4.564019 0.000586
(continued)
244 C. Srinivasarathnam et al.

Table 1 (continued)
Measured data PSO algorithm Proposed Jaya algorithm
V P V (V ) I P V (A) IC AL (A) I IC AL (A) I
15.153 4.547031 4.549818 0.002787 4.546537 0.000494
15.586 4.523891 4.527943 0.004052 4.523506 0.000384
16.019 4.493127 4.49824 0.005112 4.492866 0.000261
16.452 4.451946 4.457586 0.00564 4.451813 0.000133
16.885 4.396574 4.401707 0.005133 4.39656 0.000014
17.318 4.321957 4.324837 0.00288 4.322029 0.000072
17.751 4.221396 4.21935 0.002046 4.22149 0.000094
18.184 4.086137 4.075442 0.010695 4.086146 0.000009
18.617 3.904952 3.880961 0.023991 3.904724 0.000228
19.05 3.663804 3.621578 0.042227 3.663153 0.000652
19.483 3.34571 3.281448 0.064262 3.344454 0.001256
19.916 2.930979 2.844461 0.086518 2.929025 0.001954
20.349 2.398001 2.295944 0.102057 2.395479 0.002522
20.782 1.724644 1.624425 0.100219 1.7221 0.002544
21.215 0.89017 0.822952 0.067218 0.888777 0.001393
21.599 0.001315 0.001781 0.000465 0.002575 0.00126
RMSE 0.02784 0.0008259

Fig. 12 V-I characteristics at varying temperature and constant irradiance for SM55 model

8 Conclusion

In this chapter, accurate modeling of the PV parameters using the single diode model
has been analyzed. Out of the five parameters, Shunt Resistance (RP ), Diode reverse
Parameters Estimation of Solar PV … 245

Fig. 13 P-V characteristics at varying temperature and constant irradiance for SM55 model

Fig. 14 V-I characteristics at varying irradiance and constant temperature for SM55 model

saturation current (I0 ) and Photocurrent generated by the module (Iph ) are analytically
computed whereas the other parameters Diode ideality factor (a), Series resistance
(Rs ) are optimized using different meta-heuristic techniques. To validate the results of
the proposed methodology, three different types of PV modules, i.e., Poly-crystalline
(Shell S75), Mono-crystalline (Shell SM55) and Thin-film (Shell ST36) are consid-
ered. From the results, it is noticed that the proposed Jaya algorithm is accurately
matching the experimental data at varying irradiance condition keeping temperature
constant and varying temperature condition maintaining irradiance constant. Further,
an index RMSE value of the proposed algorithm is found to be lower than that of PSO
method. Thus, it is envisaged that the proposed method is useful tool for accurate
modeling of PV parameters.
246 C. Srinivasarathnam et al.

Fig. 15 P-V characteristics at varying Irradiance and constant temperature for SM55 model

Table 2 RMSE values of PSO algorithm and the proposed Jaya algorithm for SM55 PV module
Measured data PSO algorithm Proposed Jaya algorithm
V P V (V ) I P V (A) IC AL (A) I IC AL (A) I
0.177 3.446021 3.446329 0.000308 3.446299 0.000278
0.91 3.443657 3.444013 0.000357 3.443973 0.000317
2.091 3.439847 3.440282 0.000435 3.440226 0.000379
3.391 3.435653 3.436173 0.00052 3.436099 0.000447
4.832 3.430997 3.431611 0.000614 3.431518 0.000521
5.777 3.427936 3.42861 0.000674 3.428504 0.000568
6.203 3.426552 3.427252 0.0007 3.427141 0.000589
7.266 3.423076 3.423839 0.000763 3.423714 0.000638
8.093 3.420332 3.421137 0.000806 3.421002 0.000671
9.298 3.416188 3.41704 0.000851 3.416891 0.000703
10.34 3.412292 3.413153 0.000861 3.412995 0.000703
11.73 3.405898 3.406683 0.000784 3.40652 0.000622
12.51 3.400939 3.401604 0.000665 3.401445 0.000506
13.08 3.396081 3.396607 0.000526 3.396454 0.000373
14.31 3.378468 3.378511 0.000043 3.378388 0.00008
15.11 3.356667 3.356277 0.00039 3.356185 0.000482
16.21 3.297256 3.296392 0.000864 3.296337 0.000919
16.95 3.220708 3.220067 0.000641 3.22 0.000708
(continued)
Parameters Estimation of Solar PV … 247

Table 2 (continued)
Measured data PSO algorithm Proposed Jaya algorithm
V P V (V ) I P V (A) IC AL (A) I IC AL (A) I
17.56 3.119123 3.11948 0.000358 3.119344 0.000221
18.16 2.968063 2.970556 0.002493 2.970262 0.002199
18.62 2.805351 2.810413 0.005062 2.80992 0.004569
18.96 2.652341 2.659772 0.007431 2.659086 0.006745
19.29 2.472833 2.48279 0.009957 2.481888 0.009055
19.64 2.245118 2.257716 0.012598 2.256568 0.01145
19.97 1.992057 2.00673 0.014674 2.005359 0.013302
20.27 1.727708 1.743533 0.015826 1.74199 0.014282
20.55 1.450478 1.466402 0.015924 1.464749 0.014271
20.82 1.154836 1.169661 0.014825 1.167968 0.013133
21.11 0.806348 0.818377 0.012029 0.816738 0.01039
21.34 0.507462 0.515893 0.00843 0.514382 0.006919
21.54 0.231722 0.235923 0.004202 0.234593 0.002871
RMSE 0.0070137 0.0062505

Fig. 16 V-I characteristics at varying temperature and constant irradiance for ST36 model
248 C. Srinivasarathnam et al.

Fig. 17 P-V characteristics at varying temperature and constant irradiance for ST36 model

Fig. 18 V-I characteristics at varying irradiance and constant temperature for ST36 model
Parameters Estimation of Solar PV … 249

Fig. 19 P-V characteristics at varying irradiance and constant temperature for ST36 model

Table 3 RMSE values of PSO algorithm and the proposed Jaya algorithm for ST36 PV module
Measured data PSO algorithm Proposed Jaya algorithm
V P V (V ) I P V (A) IC AL (A) I IC AL (A) I
0.106 2.671883 2.674943 0.003061 2.674039 0.002157
0.957 2.669945 2.673643 0.003699 2.672547 0.002602
2.091 2.667272 2.671794 0.004522 2.670448 0.003176
3.579 2.663484 2.669012 0.005528 2.667364 0.00388
4.714 2.660199 2.666404 0.006205 2.664553 0.004355
6.368 2.654177 2.661133 0.006956 2.659061 0.004883
7.480 2.648583 2.655801 0.007218 2.653654 0.005071
8.823 2.638642 2.645786 0.007144 2.643667 0.005025
10.45 2.617586 2.623852 0.006265 2.622005 0.004419
11.45 2.595709 2.600928 0.005219 2.5994 0.003691
12.57 2.557374 2.561014 0.00364 2.559963 0.002589
13.80 2.488798 2.490518 0.00172 2.490049 0.00125
15.03 2.378127 2.378395 0.000268 2.378376 0.00025
16.18 2.221675 2.221705 0.00003 2.22179 0.000115
17.15 2.040681 2.041566 0.000885 2.041436 0.000755
17.81 1.889276 1.89118 0.001904 1.890778 0.001501
18.47 1.714362 1.71744 0.003078 1.71672 0.002358
19.03 1.547697 1.551729 0.004031 1.55075 0.003053
(continued)
250 C. Srinivasarathnam et al.

Table 3 (continued)
Measured data PSO algorithm Proposed Jaya algorithm
V P V (V ) I P V (A) IC AL (A) I IC AL (A) I
19.59 1.364802 1.369596 0.004794 1.368415 0.003612
20.20 1.147986 1.15324 0.005254 1.151947 0.00396
20.64 0.980844 0.986111 0.005267 0.984831 0.003987
21.15 0.776545 0.781431 0.004886 0.780283 0.003738
21.54 0.613116 0.617394 0.004278 0.616442 0.003326
21.87 0.470239 0.473775 0.003536 0.473057 0.002818
22.22 0.314347 0.316861 0.002514 0.316461 0.002114
22.55 0.163465 0.164788 0.001323 0.164757 0.001293
RMSE 0.0044638 0.003242

Table 4 Variation of model parameter values obtained using PSO and proposed Jaya Algorithm at
STC conditions for various PV models
Parameters S75 SM55 ST36
PSO Proposed PSO Proposed PSO Proposed
algorithm Jaya algorithm Jaya algorithm Jaya
algorithm algorithm algorithm
At STC (G = 1000 W/m2 , T = 25 °C)
IPV 4.700 4.700 3.450 3.450 2.680 2.680
I0 4.62E−07 1.13E−07 1.67E−07 1.14E−07 4.89E−05 5.69E−05
a 1.451 1.334 1.395 1.356 1.948 1.975
Rs 0.107 0.170 0.286 0.306 1.302 1.286
RP 156.57 152.54 316.31 314.62 624.94 736.89
Table 5 Variation of model parameter values obtained using PSO and proposed Jaya algorithm under different irradiation and temperature conditions for S75
model
PSO algorithm Proposed Jaya algorithm
IPV I0 a Rs RP IPV I0 a Rs RP
Parameters Estimation of Solar PV …

G Varying irradiance
Constant T (= 25 °C) 1000 4.700 4.62E−07 1.451 0.107 156.57 4.700 1.13E−07 1.334 0.170 152.54
varying G (W/m2 ) 800 3.760 1.47E−07 1.355 0.142 165.99 3.760 1.25E−07 1.342 0.150 164.28
600 2.820 4.93E−08 1.274 0.191 196.10 2.820 9.42E−08 1.321 0.150 207.23
400 1.880 3.46E−08 1.250 0.200 280.00 1.880 3.46E−08 1.250 0.197 277.54
200 0.940 1.43E−08 1.194 0.200 500.00 0.940 1.42E−08 1.194 0.200 498.63
T Varying temperature
Constant G varying T (°C) 20 4.690 5.97E−08 1.333 0.157 131.75 4.690 7.42E−08 1.349 0.150 134.13
40 4.730 6.53E−07 1.337 0.146 133.23 4.730 1.19E−06 1.388 0.150 200.32
60 4.770 1.95E−05 1.479 0.100 300.00 4.770 8.82E−06 1.390 0.150 302.30
251
 252

Table 6 Variation of model parameter values obtained using PSO and proposed Jaya algorithm under different irradiation for SM55 model
PSO algorithm Proposed Jaya algorithm
IPV I0 a Rs RP IPV I0 a Rs RP
G Varying irradiance
Constant T (= 25 °C) varying G (W/m2 ) 1000 3.450 1.67E−07 1.395 0.286 316.31 3.45 1.14E−07 1.356 0.306 314.62
800 2.760 8.22E−07 1.541 0.194 628.37 2.760 7.94E−07 1.537 0.196 621.62
600 2.070 4.60E−07 1.484 0.201 653.26 2.070 4.70E−07 1.486 0.198 657.38
400 1.380 2.13E−07 1.416 0.203 769.90 1.380 1.18E−07 1.367 0.304 690.77
200 0.690 2.36E−08 1.250 0.500 1000.00 0.690 2.82E−08 1.262 0.479 1089.70
T Varying temperature
Constant G (= 1000 W/m2 ) varying T (°C ) 20 3.443 3.31E−07 1.505 0.232 418.24 3.443 3.45E−07 1.509 0.230 423.35
40 3.471 1.54E−06 1.449 0.252 486.15 3.471 1.57E−06 1.451 0.251 490.75
60 3.499 5.10E−06 1.372 0.292 650.00 3.499 7.06E−06 1.406 0.276 986.08
C. Srinivasarathnam et al.
Table 7 Variation of model parameter values obtained using PSO and proposed Jaya algorithm under different irradiation for ST36 model
PSO algorithm Proposed Jaya algorithm
IPV I0 a Rs RP IPV I0 a Rs RP
G Varying irradiance
Parameters Estimation of Solar PV …

Constant T (= 25 °C ) varying G (W/m2 ) 1000 2.680 4.89E−05 1.948 1.302 624.94 2.680 5.69E−05 1.975 1.286 736.89
800 2.144 6.42E−05 1.998 1.407 841.91 2.144 6.49E−05 2.000 1.404 841.74
600 1.608 6.47E−05 1.998 1.552 851.71 1.608 6.47E−05 1.999 1.552 851.64
400 1.072 6.43E−05 1.989 1.650 934.21 1.072 6.39E−05 1.999 1.653 932.68
200 0.536 5.03E−05 1.957 1.308 1161.40 0.536 6.37E−05 1.999 0.984 1228.10
T Varying temperature
Constant G (= 1000 W/m2 ) varying T (°C) 20 2.678 2.90E−05 1.934 1.292 337.72 2.678 4.23E−05 2.000 1.253 411.10
40 2.685 1.04E−05 1.518 1.621 606.34 2.685 8.38E−06 1.492 1.637 512.17
60 2.691 7.98E−07 1.071 1.999 1425.70 2.691 8.14E−07 1.073 1.999 1479.90
253
254 C. Srinivasarathnam et al.

References

1. Solar Schools, Types of Energy (2019). https://www.solarschools.net/knowledge-bank/energy/

types
2. Kothari DP, Nagrath IJ (2008) Power systems engineering. 2nd edn. TMC
3. Mohan N (2012) Elelectric power systems: a first course. Wiley
4. Niu S, Jia Y, Wang W, He R, Hu L, Liu Y (2013) Electricity consumption and human devel-
opment level: a comparative analysis based on panel data for 50 countries. Int J Electr Power
Energy Syst 53:338–347. https://doi.org/10.1016/j.ijepes.2013.05.024
5. Assadzadeh A, Nategh H (2015) The relationship between per capita electricity consumption
and human development indices. In: 6th IASTEM international conference, Berlin, Germany,
2015, November, pp 51–57. https://www.worldresearchlibrary.org/up_proc/pdf/117-144870
270051-57.pdf
6. Ministry of New and Renewable Energy. https://mnre.gov.in/
7. Press Information Bureau Government of India Ministry of Power. https://pib.gov.in/Pressrele
aseshare.aspx?PRID=1592833
8. Access to Electricity (% of Population). World Bank Open Data. http://data.worldbank.org/ind
icator/EG.ELC.ACCS.ZS
9. Chowdhury AH, Asaduz-Zaman M (2015) Load frequency control of multi-microgrid using
energy storage system. In: 8th international conference on electrical computer engineering
advanced technology a better tomorrow, ICECE 2014, pp 548–551.https://doi.org/10.1109/
ICECE.2014.7026975
10. Srinivasarathnam C, Yammani C, Maheswarapu S (2019) Frequency control of autonomous
hybrid multi-microgrid system. In: National power electronics conference, NPEC 2019, pp
1–6.https://doi.org/10.1109/NPEC47332.2019.9034751
11. Rajasekar RV, Neeraja Krishna Kumar N (2013) Bacterial foraging algorithm based solar PV
parameter estimation. Sol Energy 97:255–265https://doi.org/10.1016/j.solener.2013.08.019
12. Jacobus H, Lin B, Jimmy DH, Ansumana R, Malanoski AP, Stenger D (2011) Evaluating the
impact of adding energy storage on the performance of a hybrid power system. Energy Convers
Manage 52(7):2604–2610. https://doi.org/10.1016/j.enconman.2011.01.015
13. Bernal-Agustín JL, Dufo-López R, Rivas-Ascaso DM (2006) Design of isolated hybrid systems
minimizing costs and pollutant emissions. Renew Energy 31(14):2227–2244. https://doi.org/
10.1016/j.renene.2005.11.002
14. Humada AM, Darweesh SY, Mohammed KG, Kamil M, Mohammed SF, Kasim NK, Tahseen
TA, Mekhilef S (2020) Modeling of PV system and parameter extraction based on experimental
data: review and investigation. Sol Energy 199:742–760.https://doi.org/10.1016/j.solener.2020.
02.068
15. Alam MBEDF, Yousri DA (2015) Flower pollination algorithm based solar PV parameter
estimation. Energy Convers Manage 101:410–422. https://doi.org/10.1016/j.enconman.2015.
05.074
16. Et-torabi K, Nassar-eddine I, Obbadi A, Errami Y, Rmaily R, Sahnoun S, El fajri A, Agunaou
M (2017) Parameters estimation of the single and double diode photovoltaic models using a
Gauss–Seidel algorithm and analytical method: a comparative study. Energy Convers Manage
148:1041–1054.https://doi.org/10.1016/j.enconman.2017.06.064
17. Nassar-eddine MAI, Obbadi A, Errami Y, El fajri A (2016) Parameter estimation of photovoltaic
modules using iterative method and the Lambert W function—a comparative study. Energy
Convers Manage 119:37–48.https://doi.org/10.1016/j.enconman.2016.04.030
18. Easwarakhanthan T, Bottin J, Bouhouch I, Boutrit C (2007) Nonlinear minimization algorithm
for determining the solar cell parameters with microcomputers. Int J Sol Energy 4(1):1–12.
https://doi.org/10.1080/01425918608909835
19. Ismail MS, Moghavvemi M, Mahlia TMI (2013) Characterization of PV panel and global
optimization of its model parameters using genetic algorithm. Energy Convers Manage 73:10–
25. https://doi.org/10.1016/j.enconman.2013.03.033
Parameters Estimation of Solar PV … 255

20. Russell Eberhart JK (1999) A new optimizer using particle swarm theory. In: Sixth international
symposium on micro machine and human science, 1999, pp 39–43.https://doi.org/10.1109/
MHS.1995.494215
21. Rezaee Jordehi A (2018) Enhanced leader particle swarm optimisation (ELPSO): an efficient
algorithm for parameter estimation of photovoltaic (PV) cells and modules. Sol Energy 159:78–
87. https://doi.org/10.1016/j.solener.2017.10.063
22. Rao RV, Savsani VJ, Vakharia DP (2011) Teaching–learning-based optimization: a novel
method for constrained mechanical design optimization problems. Comput Des 43:303–315.
https://doi.org/10.1016/j.cad.2010.12.015
23. Venkata Rao R (2016) Jaya: a simple and new optimization algorithm for solving constrained
and unconstrained optimization problems. Int J Ind Eng Comput 7:19–34. https://doi.org/10.
5267/j.ijiec.2015.8.004
24. Hasanien HM (2015) Shuffled frog leaping algorithm for photovoltaic model identification
6(2):509–515.https://doi.org/10.1109/TSTE.2015.2389858
25. Jack V, Salam Z (2015) An improved method to estimate the parameters of the single diode
model of photovoltaic module using differential evolution. In: 4th international conference on
electric power and energy conversion systems (EPECS), Sharjah, pp 1–6. https://doi.org/10.
1109/EPECS.2015.7368514
26. Reddy SS, Rathnam CS (2016) Optimal power flow using glowworm swarm optimization. Int
J Electr Power Energy Syst 80:128–139. https://doi.org/10.1016/j.ijepes.2016.01.036
27. Eberhart R, Kennedy J (1995) A new optimizer using particle swarm theory. In: Sixth inter-
national symposium on micro machine and human science, pp 39–43. https://doi.org/10.1109/
MHS.1995.494215
Integration of Photovoltaic Distributed
Generation into Grid

Neelakanteshwar Rao Battu, Perka Krishna, and Venu Yarlagadda

Abstract Present trend of electricity generation is from clean and green renewable
energy sources; one of the prominent sources being photovoltaics (solar energy).
The objective of the work presented here includes extraction of maximum power
from the photovoltaic solar plant at various working conditions using maximum
power point tracking algorithm. The output produced from the solar plant is a low
magnitude d.c output, which is required to be boosted. The solar output is fed to a
boost converter to enhance its magnitude. The enhanced d.c output is fed to a 3-ϕ, five-
level voltage source inverter to convert d.c voltage to a.c equivalent of it. Conversion
process through the inverter introduces harmonics into the system. Shunt active filter
is designed in this work to minimize the harmonics injected due to inverter operation.
Two cases are studied for the application of proposed methodology; stand alone
operation of the designed photo-voltaic solar system and photovoltaics integrated to
the grid. The above cases are implemented and analyzed using MATLAB software.
Fast Fourier analysis is used to determine the total harmonic distortion for both the
cases. The results show considerable improvement in quality of power produced due
to mitigation of harmonics.

Keywords Solar PV · Irradiance · Distributed generation · Shunt active power

filter · Fast Fourier analysis · Total harmonic distortion · Boost converter ·
Maximum power point tracking · Perturb and observe

Nomenclature

PV Photovoltaic
MPPT Maximum Power Point Tracking

N. R. Battu (B) · P. Krishna · V. Yarlagadda

Department of Electrical and Electronics Engineering, VNR VJIET, Hyderabad, Telangana, India
e-mail: [email protected]
V. Yarlagadda
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 257
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6_10
258 N. R. Battu et al.

PO Perturb and Observe

VSI Voltage Source Inverter
FFA Fast Fourier analysis
THD Total Harmonic Distortion
PCC Point of common coupling
APF Active power filter
SAPF Shunt active power filter
PWM Pulse width modulation
SPWM Sinusoidal pulse width modulation
SMPS Switched-mode power supply
V Supply voltage
Vout Output voltage
Ns Number of turns on primary side of the transformer
Np Number of turns on secondary side of the transformer
IL Photo generated current
ID Diode current
Ish Shunt current
I Output current
Vsh Voltage across the shunt resistor and diode
I0 Reverse bias saturation current
VD Voltage across diode
VT Terminal voltage
n Ideality or quality factor

1 Introduction

Modern Power Systems are facing with everlasting power demand. To meet this
increasing power demand, various means of power generation is required. The gener-
ation of power from fossil fuels leads to hazardous gases. In contrast, renewable
power generation is quite useful in reducing these effects [1]. Solar plants find to be
one of the promising means of generating renewable power [2]. Solar panels used in
respective plants consist of solar cells. Solar cell converts the beam radiation into d.c
supply. Boost converters are used in solar power plants to increase the d.c output. The
d.c output provided by converter is fed to an inverter of suitable rating to convert d.c
supply to a.c supply. The output voltage waveform of the inverter is not a pure sine
wave. It contains harmonics which are injected by the converters. The total harmonic
distortion (THD) is usually not in the acceptable limits and it can be minimized by
proper design of shunt active Power filter at point of common coupling (PCC) [3].
Li Bin et al., have proposed a digital based approach for 3-Ø, 4-wire active power
filter (APF) to reduce the harmonics [4]. Nabae et al., have discussed about control
strategy used in active power filters using voltage source converters [5]. Ko et al.,
have discussed the impact of shunt active filters on distortions caused by harmonic
Integration of Photovoltaic Distributed Generation into Grid 259

currents [6]. Singh et al., have reviewed various types of active filters and their impact
on power quality improvement. [7]. Akagi et al., have discussed the new trends used
in active filters for power conditioning [8]. Villalva et al., have discussed modeling
of PV arrays [9]. Salmeron et al., have discussed about series and shunt active filters
to improve power quality [10]. Vodyakho et al., have proposed three level inverter
based shunt active power filter [11]. Viswanth et al., have proposed fuel cell based
SAPF to deal with power quality issues [12]. Chandra et al., have reviewed the active
filters used in power quality improvement [13].
The power produced by the solar cell/photovoltaic (PV) cell is not constant due to
variations in solar radiation [14]. However, maximum power can be extracted from
the available solar radiation using Maximum power point tracking (MPPT) technique
[15]. Coordinated control of solar and battery management is proposed by Hareesh
et al., to achieve power injection and MPPT simultaneously [16]. Amjad et al., have
summarized various MPPT techniques which are used in online and offline studies
under uniform and non-uniform irradiance conditions [17]. Ali et al., have proposed
an MPPT technique using the estimation of climatic parameters [18]. Nishant et al.,
have proposed an MPPT technique based on normal harmonic search algorithm [19].
Vardan et al., have proposed circle center-line based perturb and observe algorithm for
MPPT [20]. Nishant et al., have proposed leaky least logarithmic absolute difference-
based control algorithm and learning based incremental conductance MPPT algo-
rithm for PV system integrated with grid [21]. Singh et al., have proposed power
normalized kernel least mean fourth algorithm based neural network and learning
based hill climbing technique for MPPT [22]. Ikhlaq et al., have proposed a hybrid
Cauchy and Gaussian sine cosine optimization technique for MPPT [23]. Neeraj
et al., have proposed an adaptive neuro-fuzzy inference system based particle swarm
optimization technique for MPPT [24]. Bijaya et al., have proposed leaky least mean
logarithmic fourth based control technique and learning based perturb and observe
algorithm for MPPT in grid connected PV system [25]. Though, various techniques
are available in literature for tracking or extracting maximum power from the solar
radiation, among these approaches Perturb and Observe (PO) algorithm is found to
be simpler and advantageous over several others [26–28].
This paper proposes design of converter and inverter for conversion of solar energy
into electrical energy. It focuses on the modeling of SAPF to minimize harmonics
produced during conversion of electrical energy from solar part of it. Perturb and
observe based MPPT algorithm is used to control the inverter output. Fast Fourier
series is used to analyze the harmonic distortion in the system. The work presented
in this paper demonstrates analysis of solar power plant operating in standalone and
grid connected modes.
The remaining part of this chapter is organized as follows: Sect. 2 describes
about various types of non-conventional energy sources. Section 3 describes DC-DC
converters. Section 4 explains various MPPT algorithms. Section 5 describes about
inverters. Section 6 discusses different types of filters. Analysis of simulation results
in presented in Sect. 7. Section 8 summarizes the overall work presented in this
chapter.
260 N. R. Battu et al.

2 Non-conventional Energy Sources

Non conventional sources are freely available in nature and generate electricity with
release of less pollution as compared to conventional sources like thermal and nuclear
plants. The examples of these sources are solar, wind, tidal, geo thermal and wave
power plants.

2.1 Tidal Energy Source

Tidal energy is in the form of hydro power which is used to convert energy collected
from tides into electricity and tides are occurred due to gravitational force on the
earth. Generating electricity from these sources was first followed by Europe and
America in nineteenth century. First tidal plant was established at France in 1960
and capacity of the plant is 240 MW and it is the largest power station in world. In
2011, Sinha tidal station became world’s largest station in terms of output and its
output is 254 MW.

2.2 Wind Energy Source

Wind energy is freely available in nature and it converts wind energy into electrical
supply with help of wind turbines and generators. It generates less pollution as
compared to conventional sources. Highest usage of wind power is in Denmark and
it is 43.4%.

2.3 Wave Energy Source

Wave energy is in the form of waves on the surface of the water, which is converted
into useful electrical supply. First wave was established in U.K in the year 2008.

2.4 Geo Thermal Energy Source

The heat produced in the crust of the earth is used to generate electricity from it in
this type of sources. This source is used for industrial and agricultural purpose and
usage of this source is increasing gradually.
Integration of Photovoltaic Distributed Generation into Grid 261

2.5 Hydro Power

In hydro power plants, potential energy of the water is converted into electrical energy
with use of turbines and generators. It is free from environmental pollution. China
generates the highest output in the world and the output power generated by it is
22,500 MW. Almost 20% of total electricity generation is from this source. Hydro
power plants can be classified into four types: run of river, storage, pumped storage
hydro, and hydro kinetic technology.

2.6 Photovoltaic Power

Solar energy, in the form of radiation, is converted into electricity. This type of source
is classified into two types; active and passive solar. In active solar type techniques
like mirrors and solar heating are used. In passive solar type techniques like surface
coatings, fins, micro particles, nano particles are used. Advantages of solar energy
sources compared to other type of sources are as follows:
• Solar system directly converts the solar radiation into electrical supply and it
doesn’t need rotating devices like turbines.
• Maintenance cost is low.
• PV systems do not release sound, air and water pollution into atmosphere.
• Day-by-day its manufacturing design is enhanced to improve the efficiency.
• PV source is clean source of energy.
Disadvantages of solar energy sources are as follows:
• Manufacturing cost is high.
• Solar radiation is not available at latitude and if clouds are present.
• Fluctuation problems occur.
• Requirement of area to collect the supply is large.
• It has less efficiency.

2.7 Types of PV Cells

Based on the type of installation solar cells are classified into the following types:
monocrystalline cells, polycrystalline cells and thin film cells. Monocrystalline cells
are made by using silicon wafers, connected in rows and columns, by covering a
glass sheet on the top view of the cell. It is also known as single crystalline panel.
It offers high performance and its cost is high. Polycrystalline cells are made by
using only one silicon crystal and melt together to form a wafer. It is also called as
multi-crystalline silicon. It has lower efficiency as compared to mono crystalline cell
and it is available in low cost. Thin film cells are made by using amorphous silicon
262 N. R. Battu et al.

and cadmium and on the surface plastic is placed. It is thinner than other cells. It has
low efficiency and weighs less. Also, it is portable and flexible.

3 DC-DC Converters

DC-DC converter is a device which increases one level of d.c voltage to other
level. These converters are classified into two types; isolated and no-isolated type
converters. Isolated converters are used for high voltage applications and have sepa-
rate input and output terminals. Non-isolated converters are used for low voltage
applications. Input and output terminals of such converters have same ground [29].

3.1 Isolated Converters

Isolated converters are used for high voltage applications. These converters reduce
the noise and interference. These types of converters are classified into two types.
They are fly back converter and forward converter. The fly back converter is used for
dc-to-dc and ac-to-ac conversion. The fly back converter can act as step up or step
down converter in association with a transformer, which has split inductor. It has two
schemes for controlling; voltage control mode and current control mode. Forward
converter is used to increase or decrease the output voltage of the transformer and
provides isolation.

3.2 Non- Isolated Converters

Non-isolated converters are used in applications where low voltage is required. Both
the terminals in the circuit have same ground. These types of converters produce high
noise. Protection and performance of this type of converters is poor as compared to
the other type. These converters are classified into three types. They are buck, boost
and buck-boost converters. Buck converter is also known as step down converter. It
reduces the level of output voltage. This converter is used in SMPS, which consists of
two or more semi-conductor devices. To overcome the voltage variations capacitors
are employed. It also improves the voltage quality. Inductor is used to resist the sudden
change in current and they are connected in both supply and load side. Boost converter
is called as step up converter. It increases the input voltage level. It consists of semi-
conductors. Passive elements like capacitors are used to reduce the harmonics. The
output voltage is higher than the source voltage. Input current is greater than load
current. Buck boost converter is combination of both buck and boost converter.
Depending upon the application it may be used to increase or decrease the voltage
level.
Integration of Photovoltaic Distributed Generation into Grid 263

4 Solar Energy

4.1 MPPT Algorithm

Maximum power point tracking algorithm is used to extract maximum power from
the available irradiance. This algorithm is used for renewable energy sources. Non-
conventional energy sources generally offer less efficiency and this algorithm is used
to improve the efficiency. The commonly used MPPT algorithms are as follows:
constant voltage method, current sweep method, perturb and observe algorithm,
incremental conductance method and temperature method.
Constant voltage method is the basic type but has some demerits. This algorithm
maintains a constant voltage and does not allow tracking the voltage. Maximum
power cannot be extracted using this algorithm. In current sweep algorithm, current
and voltage characteristics of the solar cell and waveform of the sweep is used to
control the power. Perturb and Observe approach is also called as hill climbing
method and it is one of the mostly used methods and the reason for it is, it is easy
to implement. First, voltage and current are measured and if power tends to vary,
the voltage is adjusted to extract maximum power. The flow chart for this algorithm
is given in Fig. 1. In incremental conductance method array current and voltage is
measured by controller to predict the effect of a voltage change. Controller has high
computation. In temperature method, maximum voltage is estimated by measuring
the temperature of solar cell and comparing it with the reference value [30].

4.2 PV Cell Modeling

Equivalent circuit of PV cell is shown in Fig. 2 and explained in detail in references

[31, 32]. Output current flowing through the resistor ‘Rs ’ can be expressed as

I = I L − I D − Ish (1)

where, I L is the photo generated current, I D is the diode current, I Sh is the shunt
current in ampere and I is the output current.
Voltage across the current source is given by

Vsh = V + I Rs (2)

where, Vsh is the voltage across the shunt resistor and diode and V is the voltage at
output terminals. By using Shockley’s equation we get,
 VD 
I D = I0 e nVT − 1 (3)
264 N. R. Battu et al.

Fig. 1 Flow chart of perturb

Start
and observe method

Measure voltage
V(t) and current I(t)

Calculate power P(t) from

the voltage and current

Is
Yes
P(t)=P(t-1)
?
No
Is
No Yes
P(t)>P(t-1)
?
Is Is
V(t)<V(t-1) Yes Yes V(t)<V(t-1)
? ?
No No
V ref =V+ Δ V

V ref =V- Δ V

Fig. 2 Equivalent circuit of Rs I

solar cell
ID Ish +
L
o
IL R sh V
a
d
-

where, I0 is the reverse bias saturation current, VD is the voltage across diode, VT is
the terminal voltage, ‘n’ is the ideality or quality factor. By applying ohms law we
get,
 VD  V + IR
s
I = I0 e nVT − 1 − (4)
Rs

If Rsh is infinite we get,

 
IL − I
V = nVT ln − 1 − I Rs (5)
I0
Integration of Photovoltaic Distributed Generation into Grid 265

Fig. 3 Characteristics of Maximum

solar cell power point
Imax = I sc I-V curve P max
Imp
P = VxI

Current (A)
e

Power (W)
rv
cu
V
P-

0 Voltage (V) V mp V max = V oc

Finally current equation is

 
I = I L − I0 e A − 1 − B (6)

(V +I Rs ) V +I Rs
where, A = nVT
and B = Rsh
.

4.3 Characteristics of PV Cell

PV cell offers electrical characteristics. The solar radiation affects the output of the
cell. The output power of the PV cell is inversely proportional to temperature and
directly proportional to beam radiation. Minimum current and maximum voltage
is occurred at open circuit in the cell. If positive and negative terminals are short
circuited voltage of the cell is minimum and current is maximum. Current versus
voltage plot is shown in Fig. 3.

5 Inverters

Inverter is device which converts direct current to alternating current supply. Based
on some characteristics they are classified and mentioned below.

5.1 Based on Output

Based on output the inverters are classified as sine wave inverters, square wave
inverters and modified sine wave inverters. Sine wave inverter is a device which
transforms the d.c voltage into an output voltage similar to supply voltage. It is one
of the merits to improve the efficiency. Square wave is output of this device. But
this device is rarely used. The reason is losses are very high and efficiency is very
266 N. R. Battu et al.

low. Output of modified sine wave is combination of sine and square wave. Shape of
output waveform nearly looks like a sine wave.

5.2 Based on Source

Based on type of source used inverters are classified as voltage source inverters and
current source inverters. Input to voltage source inverters is voltage. They are used
for all applications, because they have more reliability and high efficiency. Input to
current source inverters is current and is used in industrial applications and is rarely
used in common applications.

5.3 Based on Load

Based on type of load connected inverters are classified as single phase inverters
and three phase inverters. Single phase inverters are used in household applications
and commercial purpose. Again they are classified in to two categories; full and
half bridge inverters. Half bridge inverter consists of two diodes and thyristors and
voltage is divided into two equal parts as shown in Fig. 4.
Full bridge inverter consists of four diodes and four thyristors and we get step
output and it is converted into sine wave by using filters. Diodes are used for feedback
mechanism. As compared to half bridge its voltage is doubled and power is four times
(Fig. 5).
Three phase inverters are used in industrial purpose and it consists of six diodes
and six thyristors as shown in Fig. 6.
Operation of inverters can be divided into two modes: 180° mode and 120° mode.
Thyristor conduction time is 180 in 180° mode and three thyristors act as three phases
and these are in conduction mode at any time. Short circuit problems are occurred
in this mode. Phase voltage looks like stepped voltage and line voltage looks like
quasi-square wave. Thyristor conduction time is 120 in 120° mode and two thyristors
are in conduction mode. Shapes of the both voltages are opposite to 180° mode.

Fig. 4 Half bridge inverter

V s/2 T1 D1

Load
+ V0 -
V s/2 T2 D2
Integration of Photovoltaic Distributed Generation into Grid 267

Fig. 5 Full bridge inverter

T1 D1 T3 D3

Vs
Load
+ V0 -

T4 D4 T2 D2

T1 D1 T3 D3 T3 D3

Vs

T4 D4 T2 D2 T2 D2

Vca
Vab Vbc

Load

Fig. 6 Three phase inverter

5.4 Based on Technique

Based on the PWM technique used inverters are classified as single PWM inverters,
multiple PWM inverters, sinusoidal PWM inverters and modified SPWM inverters.
Output voltage in single PWM technique can be easily controlled using PWM tech-
nique. This technique has two signals. Square wave is reference signal and trian-
gular wave signal act as carrier signal. By comparing both the signals gate pulse is
generated. It has only one signal. Reference signal controls the output voltage. This
technique is rarely used; the reason is it injects more harmonics. Multiple PWM
technique has more pulses for every half cycle of voltage. Sinusoidal PWM tech-
nique is frequently used in industrial applications. Reference signal is sine wave
signal. Output frequency is equal to the carrier signal frequency. Sine wave is the
output. Using above techniques variation of pulse width modulation is not possible.
To overcome this problem modified SPWM technique is used. At conduction period
268 N. R. Battu et al.

of last and first 60° carrier signal is applied. This technique exhibits advantages like
less switching losses and less switching devices.

5.5 Based on Level

Based on the output level inverters are classified as two level inverters and multiple
output inverters. Two level inverter has two output levels. They are positive and
negative voltages [9]. Multiple output inverters are again divided into four types.
They are flying capacitor inverters, diode clamped inverters, hybrid inverters and
cascaded H-type inverter.
Flying capacitor inverters have only capacitors and they allow only limited voltage.
Active and reactive power is controlled by this device. It suffers with disadvantages
of high switching frequency and switching losses. Diode clamped inverter consists of
diodes with different voltage levels and series capacitors. The output of this inverter
is less than its input voltage. By increasing the capacitors and switches the output
voltage can be increased. Hybrid inverter converts beam radiation into electrical
supply and it consists of battery and it stores the energy and it is very useful for
back to back power consumption. Cascaded H-type inverter consists of switches
like IGBT, MOSFET and capacitors. Numbers of components get reduced using this
type of inverter. Based on level of inverter, the number of cells depends and each
cell acts as a voltage source [5–8]. Applications of Diode clamped inverter are Static
var compensation, HV lines abd drives for variable speed. Applications of capacitor
bank inverter are Static var, IM control and sine wave rectifiers. Applications of H-
type inverter are interfacing of non-conventional source, PF compensators and active
filters.

6 Filters

Power consumption is exponentially increasing and also generation of electricity is

less. By using fossil fuels air pollution is occurred. To meet the increased consumption
of electricity it is required to generate electricity by using green sources. Grid operates
with AC supply and renewable source generates DC supply hence conversion is
required. Power electronic devices are used to in DC-to-DC conversion and DC-
to-AC conversion. These devices inject harmonics into the system. To eliminate
these harmonics and improve power quality, filters are used. Filters are classified as
follows:
Integration of Photovoltaic Distributed Generation into Grid 269

6.1 Based on Response

Based on response the filters are classified as high pass filters, low pass filters, pass
band filters and band stop filters. High pass filter allows only high frequency signals
to pass through it and opposes low frequency signals. Low pass filter allows only
low frequency signals to pass through it and opposes high frequency signals. Pass
band filter allows a particular frequency of band and opposes other frequency signals.
Band stop filter allows low and high frequency signals and opposes other frequency
signals.

6.2 Based on Their Construction

Based on response the filters are classified as active filters and passive filters. Active
filter is used to eliminate the harmonics by injections of real power with equal
frequency in reverse phase to cancel those harmonics. It consists of operational
amplifier or switches like transistors. Resistors and capacitors are also placed to
compensate the harmonics. These filters are categorized as series and shunt filters.
Series active filters are connected in series with utility grid and are used for mitigate
voltage harmonics and it are operated as voltage regulators. These types of filters
are frequently used for compensating voltage sag and swell problems. These filters
inject fundamental voltage to compensate those problems. Shunt active filters are
connected in parallel with grid and inject the fundamental current component to elim-
inate harmonics. These types of filters are used for mitigation of current harmonics
at load side. The injected mitigation current has magnitude which is equal and phase
which is opposite to that of harmonic component [13].
Passive filters consist of passive elements like resistors, capacitors and induc-
tors. They do not require external dc source. These filters are used to eliminate
the current harmonics and are mainly categorized as series, shunt and hybrid. Series
filters are connected between input source and output load in series. The combination
of inductor and capacitor does not allow the sudden changes in currents and volt-
ages. Shunt filters are connected in parallel. They consist of passive elements. These
filters are used to mitigate the harmonics and to balance the unbalanced voltages.
Hybrid filter is a combination of both series and shunt passive filters. It mitigates both
current and voltage harmonics. Advantages of active filter are: offers better perfor-
mance compared to passive filter, less in weight and it has high Q factor. Advantages
of passive filter are: low cost, less circuit complexity, and it does not require external
supply.
270 N. R. Battu et al.

7 Simulation Results

The work presented here deals with mitigation of harmonics, improvement of volt-
ages and currents and reducing total harmonic distortion (THD) at different condi-
tions. Shunt active power filter is used to eliminate the unwanted signal. Four case
studies are performed in this work. They are (i) RL load without grid, without filter,
(ii) RL load without grid with filter (iii) RL load with grid without filter and (iv) RL
load with grid, with filter.
The PV system is fed to boost converter to increase the voltage level. Inverter
output is fed to RL load. By conversion of dc to dc and dc to ac, current is in phase
with the voltage. Harmonics are injected into system. Due to this harmonics, heat is
increased. Increased harmonics results in reduced efficiency. RL load without filter
is as shown in Fig. 7 and THD in this case is found to be 12.67% as shown in Fig. 8.
Due to harmonics, system performance and power quality are poor. To improve
this a shunt active filter is designed and connected in parallel to load as shown in
Fig. 9. It eliminates the harmonics and THD in this case is decreased to 1.70% as
shown in Fig. 10.
PV system is integrated to the grid, the presence of RL load and conversion devices
inject harmonics into the system. Due to these harmonics power quality problems
occur. RL load with grid and without filter is as shown in Fig. 11 and THD% in this
case is found to be 14.55% as shown in Fig. 12.
Shunt active filter is connected in parallel with grid. Reactive power and current
harmonics are compensated by the filter. Also power factor is corrected in this case.
Simulink model of grid with filter is as shown in Fig. 13. THD % in this case is
decreased to 6.30 is as given in Fig. 14.
Each phase has a sub system with H-bridge cascaded model. R phase sub system
is as shown in Fig. 15. The output of this model is given to the grid. Inverter output
is controlled using pulse width modulation technique. Switching signal is as shown
in Fig. 16.

Fig. 7 RL load without grid without filter

Integration of Photovoltaic Distributed Generation into Grid 271

Fig. 8 THD% of RL load without filter

Fig. 9 RL load without grid with filter

Three phase inverter output is controlled using PWM technique. Inverter output
is not sinusoidal but is nearly sinusoidal. Five level inverter offers low harmonics
compared to that of three level. Output of five level inverter is as shown in Fig. 17.
Electric power distribution system is that part of utility grid which takes electrical
supply from the generators and transmits it to the consumers connected to distribution
system through the transmission lines. To integrate distributed generators with the
grid, both must have same frequency and same voltage. The grid voltages are as
272 N. R. Battu et al.

Fig. 10 THD% of RL load with filter

Fig. 11 RL load without filter with grid

shown in Fig. 18. If PV system is connected to the grid, total harmonic distortion is
found to be quite high, that is 14.55%. Shunt active filter is connected between grid
and load, to reduce the THD to 6.30%. Total harmonic distortion without and with
filter is tabulated in Table 1.
Integration of Photovoltaic Distributed Generation into Grid 273

Fig. 12 THD% of grid without filter

Fig. 13 RL load with grid and filter

8 Conclusion

The analysis has been made for two cases; case1 focuses on harmonic analysis for the
standalone photovoltaic solar system comprising of dc-dc boost converter and three
phase voltage source five level H-bridge inverter and case II deals with harmonic
analysis of grid connected solar system. Maximum power point tracking algorithm
is implemented to extract optimal power from the solar system in both the cases.
274 N. R. Battu et al.

Fig. 14 THD% of grid with filter

Fig. 15 R phase sub system

Integration of Photovoltaic Distributed Generation into Grid 275

Fig. 16 PWM signals

Fig. 17 Inverter output

276 N. R. Battu et al.

Fig. 18 Grid voltages

Table 1 THD % without and

Integration Without SAPF With SAPF (THD %)
filter
(THD %)
RL load without 12.67 1.70
grid
RL load with grid 14.55 6.30

DC-to-DC converter is used to increase the dc output voltage which is fed to the
cascaded H-bridge five level inverter. This inverter converts dc into ac supply which
contains harmonics. Shunt active filter is designed to minimize the harmonics, which
are injected by the boost converter and the inverter circuits. In case II, PV system
is integrated to the grid and Fast Fourier analysis is performed to obtain its THD.
Shunt active filter is designed to minimize the harmonics. Fast Fourier analysis has
been done and the total harmonic distortion is obtained in both the cases. The total
harmonic distortion of standalone system is observed as 12.67% without filter. SAPF
is designed and implemented in the system and total harmonic distortion is reduced
to 1.70%. In case 2, the solar system is integrated to grid at PCC. The total harmonic
distortion is observed as 14.55% without filter. Shunt active filter is designed and
connected at PCC after which the distortion was decreased to 6.30%. Thus, the
shunt active power filter may be treated as the most effective filter to minimize the
harmonics in both single and grid connected systems.
Integration of Photovoltaic Distributed Generation into Grid 277

References

1. Salkuti SR (2019) Day-ahead thermal and renewable power generation scheduling considering
uncertainty. Renew Energy 131: 956–65. https://doi.org/10.1016/j.renene.2018.07.106
2. Prasad KK, Myneni H, Kumar GS (2019) Power quality improvement and pv power injection
by dstatcom with variable dc link voltage control from rsc-mlc. IEEE Trans Sustain Energy
10(2):876–885. https://doi.org/10.1109/TSTE.2018.2853192
3. Schinke C, Vogt MR, Bothe K (2018) Optical modeling of photovoltaic modules with ray
tracing simulations, Chap 3. Wiley, pp 27–91. https://doi.org/10.1002/9781119364214.ch3
4. Bin L, Minyong T (2012) Control method of the three-phase four-leg shunt active power filter.
Energy Proc 14:1825–1830. https://doi.org/10.1016/j.egypro.2011.12.1174
5. Akagi H, Nabae A, Atoh S (1986) Control strategy of active power filters using multiple
voltage-source PWM converters. IEEE Trans Ind Appl IA-22(3):460–465. https://doi.org/10.
1109/TIA.1986.4504743
6. Ko WH, Gu JC (2016) Impact of shunt active harmonic filter on harmonic current distortion
of voltage source inverter-fed drives. IEEE Trans Ind Appl 52(4):2816–2825. https://doi.org/
10.1109/TIA.2016.2552138
7. Singh B, Al-Haddad K, Chandra A (1999) A review of active filters for power quality
improvement. IEEE Trans Ind Electron 46(5):960–971. https://doi.org/10.1109/41.793345
8. Akagi H (1996) New trends in active filters for power conditioning. IEEE Trans Ind Appl
32(6):1312–1322. https://doi.org/10.1109/28.556633
9. Villalva MG, Gazoli JR, Filho ER (2009) Comprehensive approach to modeling and simulation
of photovoltaic arrays. IEEE Trans Power Electron 24(5):1198–1208. https://doi.org/10.1109/
TPEL.2009.2013862
10. Salmeron P, Litran SP (2010) Improvement of the electric power quality using series active and
shunt passive filters. IEEE Trans Power Delivery 25(2):1058–1067. https://doi.org/10.1109/
TPWRD.2009.2034902
11. Vodyakho O, Mi CC (2009) Three-level inverter-based shunt active power filter in three-phase
three-wire and four-wire systems. IEEE Trans Power Electron 24(5):1350–1363. https://doi.
org/10.1109/TPEL.2009.2016663
12. Krishna VK, Dash SK, Geshma KR (2020) Development and analysis of power quality by using
fuel cell based shunt active power filter. In: 2020 2nd international conference on innovative
mechanisms for industry applications (ICIMIA), pp 448–452. https://doi.org/10.1109/ICIMIA
48430.2020.9074886
13. Singh B, Al-Haddad K, Chandra A (1999) A review of active filters for power quality
improvement. IEEE Trans Industrial Electronics 46(5):960–971. https://doi.org/10.1109/41.
793345
14. Rao B, Krishna P, Yarlagadda V (2020) Mitigation of photovoltaic solar system harmonics using
shunt active power filter. In: 2020 Fourth international conference on computing methodolo-
gies and communication (ICCMC), pp 566–571. https://doi.org/10.1109/ICCMC48092.2020.
ICCMC-000105
15. Esram T, Chapman PL (2007) Comparison of photovoltaic array maximum power point tracking
techniques. IEEE Trans Energy Conversion 22(2):439–449. https://doi.org/10.1109/TEC.2006.
874230
16. Myneni H, Ganjikunta SK (2020) Energy management and control of single-stage grid-
connected solar pv and bes system. IEEE Trans Sustain Energy 11(3):1739–1749. https://
doi.org/10.1109/TSTE.2019.2938864
17. Ali A, Almutairi K, Padmanaban S, Tirth V, Algarni S, Irshad K, Islam S, Zahir MH, Shafi-
ullah M, Malik MZ (2020) Investigation of mppt techniques under uniform and non-uniform
solar irradiation condition–a retrospection. IEEE Access 8:127,368–127,392. https://doi.org/
10.1109/ACCESS.2020.3007710
18. Chikh A, Chandra A (2015) An optimal maximum power point tracking algorithm for pv
systems with climatic parameters estimation. IEEE Trans Sustain Energy 6(2):644–652. https://
doi.org/10.1109/TSTE.2015.2403845
278 N. R. Battu et al.

19. Kumar N, Hussain I, Singh B, Panigrahi BK (2018) Normal harmonic search algorithm-based
mppt for solar pv system and integrated with grid using reduced sensor approach and pnklms
algorithm. IEEE Trans Ind Appl 54(6):6343–6352. https://doi.org/10.1109/TIA.2018.2853744
20. Saxena V, Kumar N, Singh B, Panigrahi BK (2021) A rapid circle centre-line concept-based
mppt algorithm for solar photovoltaic energy conversion systems. IEEE Trans Circuits Syst I
Regul Pap 68(2):940–949. https://doi.org/10.1109/TCSI.2020.3038114
21. Kumar N, Singh B, Panigrahi BK, Xu L (2019) Leaky-least-logarithmic-absolute-difference-
based control algorithm and learning-based inc mppt technique for grid-integrated pv system.
IEEE Trans Ind Electron 66(11):9003–9012. https://doi.org/10.1109/TIE.2018.2890497
22. Kumar N, Singh B, Panigrahi BK (2019) Pnklmf-based neural network control and learning-
based hc mppt technique for multiobjective grid integrated solar pv based distributed generating
system. IEEE Trans Ind Inform 15(6):3732–3742. https://doi.org/10.1109/TII.2019.2901516
23. Kumar N, Hussain I, Singh B, Panigrahi BK (2017) Single sensor-based mppt of partially
shaded pv system for battery charging by using cauchy and gaussian sine cosine optimization.
IEEE Trans Energy Convers 32(3):983–992. https://doi.org/10.1109/TEC.2017.2669518
24. Priyadarshi N, Padmanaban S, Holm-Nielsen JB, Blaabjerg F, Bhaskar MS (2020) An experi-
mental estimation of hybrid anfis–pso-based mppt for pv grid integration under fluctuating sun
irradiance. IEEE Syst J 14(1):1218–1229. https://doi.org/10.1109/JSYST.2019.2949083
25. Kumar N, Singh B, Panigrahi BK (2020) Llmlf-based control approach and lpo mppt tech-
nique for improving performance of a multifunctional three-phase two-stage grid integrated pv
system. IEEE Trans Sustain Energy 11(1):371–380. https://doi.org/10.1109/TSTE.2019.289
1558
26. Abdelsalam AK, Massoud AM, Ahmed S, Enjeti PN (2011) High-performance adaptive perturb
and observe mppt technique for photovoltaic based microgrids. IEEE Trans Power Electron
26(4):1010–1021. https://doi.org/10.1109/TPEL.2011.2106221
27. Kollimalla SK, Mishra MK (2014) Variable perturbation size adaptive P & OMPPT algorithm
for sudden changes in irradiance. IEEE Trans Sustain Energy 5(3):718–728. https://doi.org/10.
1109/TSTE.2014.2300162
28. Pai FS, Chao RM (2010) A new algorithm to photovoltaic power point tracking problems with
quadratic maximization. IEEE Trans Energy Convers 25(1):262–264. https://doi.org/10.1109/
TEC.2009.2032575
29. Harfman Todorovic M, Palma L, Enjeti PN (2008) Design of a wide input range dc–dc converter
with a robust power control scheme suitable for fuel cell power conversion. IEEE Trans Industr
Electron 55(3):1247–1255. https://doi.org/10.1109/TIE.2007.911200
30. Hohm D, Ropp M (2000) Comparative study of maximum power point tracking algo-
rithms using an experimental, programmable, maximum power point tracking test bed. In:
Conference record of the twenty-eighth IEEE photovoltaic specialists conference—2000 (Cat.
No.00CH37036), pp 1699–1702. https://doi.org/10.1109/PVSC.2000.916230
31. Sadanand DDK (2021) Modeling of photovoltaic solar cell based on cusbs absorber for the
enhancement of performance. IEEE Trans Electron Dev 68(3):1121–1128. https://doi.org/10.
1109/TED.2020.3048326
32. Bhuvaneswari G, Annamalai R (2011) Development of a solar cell model in matlab for pv
based generation system. In: 2011 annual IEEE India conference, pp 1–5. https://doi.org/10.
1109/INDCON.2011.6139509
Transient Stability Enhancement
of Power System with Grid Connected
DFIG Based Wind Turbine

D. Rakesh Chandra, Surender Reddy Salkuti,

and Venkataramana Veeramsetty

Abstract Transient stability (TS) is one of the major issues of wind power grid
integration. The reasons for transients are sudden severe disturbances such as faults.
To save the grid from severe damage we must concentrate on transient stability
management. A fixed-speed wind generator is easy to operate and reliable but it may
not operate effectively under critical conditions such as transient stability. Coming
to variable speed wind energy generators (WEGs) they can improve the TS of the
system, and the Doubly Fed Induction Generator (DFIG) is one among the variable
speed WEGs. In this work, the impact of DFIG on the TS of the power system
connected to the grid has been investigated, and also rotor resistance control method
is utilized to enhance the TS of the system. Here the proposed TS analysis has been
implemented on a Reliability Test System (RTS)-24 bus system. All the simulation
studies have been carried out in a MATLAB-compatible-based Power System Anal-
ysis Toolbox (PSAT) toolbox. From the obtained simulation studies it can be observed
that in DFIG wind turbine addition of external resistance has improved system tran-
sient stability and the system could reach steady-state quickly as damping ratio is
enhanced due to proportional relationship with the external resistance being added.

Keywords Renewable energy · Transient stability · Wind power · Induction

generator · Power system

D. Rakesh Chandra
Department of Electrical and Electronics Engineering, Kakatiya Institute of Technology and
Science, Warangal, India
e-mail: [email protected]
S. R. Salkuti (B)
Department of Railroad Electrical Systems, Woosong University, Daejeon 34606, Republic of
Korea
e-mail: [email protected]
V. Veeramsetty
Center for Artificial Intelligence and Deep Learning, SR Engineering College, Warangal, India
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 279
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6_11
280 D. Rakesh Chandra et al.

Nomenclature

TS Transient stability
WEGs Wind energy generators
DFIG Doubly Fed Induction Generator
RTS Reliability Test System
PSAT Power System Analysis Toolbox
NN Neural network
PLL Phase-locked loop
MG Microgrid
DDSG Direct-drive synchronous generator
CSWT Constant speed wind turbine
SCIGs Squirrel Cage Induction Generators
GB Gearbox
VSC Voltage source converter

1 Introduction

Recently, power generation using renewable energy resources (RERs) is on an

increase due to the considerations of green energy. At the beginning of 2020, the
global installed RERs utility-scale capacity stood at 2537 GW with wind and solar
PV combined contributing about 46% of the total global RERs share. In the modern
world, energy has a strong relationship with the socio-economic development of a
country. RERs are an adequate substitute for a conventional power system that can
provide greener, consistent, quality power with low network congestion and losses.
Wind energy availability is huge and it is eco-friendly. Nowadays, wind power gener-
ation has been increasing continuously throughout the world. Distributed generation
in the form of RERs produces dependency on conventional energy generation and
reduces the carbon footprint. Wind energy conversion system has several economic,
technical, social, and environmental challenges. Economic challenges include high
installation cost, technology challenges include power electronic converters (i.e.,
high power density converters and advanced cooling methods), controllers (pitch
angle controllers, inverter-side, and grid-side converters). At present fixed and vari-
able speeds, induction generators (IGs) are utilized for the generation of electrical
power by using the wind source. Thus the integration of wind energy into the utility
grid arise the issues such as rotor angle stability and transient stability (TS) problems.
In reference [1], a multilayered neural network (NN) is used to find the TS of
the power system by using the PSAT software. A simple DFIG wind turbine along
with its power converter model is simulated in [2] using DIgSILENT software to
meet the power demand. Enhancement of both voltage stability and TS has been
performed in reference [3] by using the wind energy converter with the help of P/Q
control and modern voltage source converter (VSC). Comparisons among various
Transient Stability Enhancement of Power … 281

types of wind turbines according to their actual market share and cost-performance
have been proposed in references [4, 5]. An approach to improve voltage stability
by using a static compensator is proposed in references [6, 7]. The impacts of WEG
on the voltage and TS of power grids are examined in reference [8] by determining
the effect of variable and fixed speed grid-connected IGs on power systems voltage
stability. The voltage stability of the system with grid-connected fixed and variable
speed wind turbines has been investigated in reference [9]. The stability of grid-
connected DFIG by taking into account the phase-locked loop (PLL) dynamics is
proposed in reference [10].
The coordinated control of DFIG based WEG and the microturbine in a DC
microgrid (MG) with the constant power load has been proposed in [11]. In refer-
ence [12], the performance of grid-connected DFIG based wind turbine system with
gear-train backlash is analyzed. A nonlinear control method to coordinate DFIG
based WEGs and static compensators in multi-machine power systems are proposed
in [13]. Reference [14] presents various issues related to the stability of wind energy
systems with variable speed in weak and strong grids. Reference [15] investigates the
modeling and TS analysis of the integrated WEG system. Reference [16] presents
a comprehensive review among three main types of wind turbines, i.e., direct-drive
synchronous generator (DDSG), constant speed wind turbine (CSWT), and DFIG.
The small-signal stability of a grid-connected power system with squirrel cage induc-
tion generator using a PSAT analysis with localization of Eigenvalues is proposed in
reference [17]. The behavior of DFIG for grid disturbances has been simulated and
validated in reference [18].
As wind energy exhibits both uncertainty and variability, wind generation cannot
be dispatched on demand. It is also highly complex to forecast wind energy with
precision. When it is used in combination with conventional sources, it poses a lot
of challenges to grid operators in maintaining grid stability. Wind energy integration
also poses challenges like power quality, effective load management in the power
system. A few more issues such as power balance, wind power reserve management,
and voltage control of the power system will also pose challenges to the system
operator. Power quality is the deviation from original sinusoidal voltage and current
waveforms such as flickers, harmonics, voltage sag, and swell. The objective of
this work is to enhance transient stability (TS) of the DFIG connected system with
the optimum rotor resistance control method. This work mainly concentrates on TS
analysis and enhancement where it is a real challenging issue for grid-connected
wind turbine systems. Time-domain simulation analysis will give the overall impact
of transient stability on the system. The rotor resistance control method is used for
the first time in this chapter to improve the TS of DFIG connected to the RTS-
24 bus system. This chapter investigates the transient stability of the RTS 24 bus
system when DFIG (at bus number 14) is connected to it. The detailed mathematical
modeling of DFIG is also presented in this chapter. All these simulation studies are
implemented in PSAT software of version 2.1.8.
The chapter is organized as follows: Sect. 2 describes detailed mathematical
modeling of DFIG wind turbine and the implementation details of wind turbine
282 D. Rakesh Chandra et al.

integrated models for TS analysis using PSAT software. The description of the Reli-
ability Test System (RTS) 24 bus system is presented in Sect. 3. Section 4 presents
the results and discussion. Lastly, the chapter is summarized in Sect. 5.

2 Modeling of Doubly Fed Induction Generator (DFIG)

Integration of wind power into utility grid via grid-connected wind turbines creates
stability problems. These wind turbine generators are usually either Squirrel Cage
Induction Generators (SCIGs) or Doubly Fed Induction Generators (DFIG) that use
constant speed and variable speed wind turbines, respectively. The requirement of
reactive power by the wind turbines causes voltage stability issues. This stability
problem is also based on the amount of power integrated into the grid. The stability
problem is mainly divided into small signal stability and TS. Small signal stability
is due to the power system being subjected to a small disturbance that leads to
small incremental change in system variables. TS is due to the power system being
subjected to severe disturbances such as breaker and faults interruptions etc. In recent
times the majority of wind energy generation is done by DFIG based wind turbines.
The basic configuration and schematic view of DFIG are depicted in Fig. 1. The
DFIG is mainly combined with a gearbox (GB), induction generator, and converter.
In this work, a partial scale converter with the rating of 20% to 30% of DFIG is used
so that the speed can be varied within ± 30% of the synchronous speed. A partial
scale converter makes DFIG more attractive in terms of economical point of view
[19]. Generally wound rotor (slip ring) induction generators are used as DFIG.
In Fig. 1, GB is the gearbox, GSC is the grid side converter and RSC is the rotor
side converter. GB is having more turns to improve wind speed to connect DFIG
[20]. RSC and GSC are used to regulate reactive power and slip power recovery.

WIND TURBINE

DFIG

GB GRID

RSC GSC

Fig. 1 The basic configuration of DFIG

Transient Stability Enhancement of Power … 283

2.1 Mathematical Modeling of DFIG

Wind turbines with DFIG are used more often due to lower capital investment.
DFIG is dominating the wind turbine market with a total share of 84.5% to 86%.
Based on the operational characteristics, grid integration of wind turbines will lead
to small signal or transient stability issues. For the same wind turbine rating, DFIG
wind turbines have better reactive power regulation capability and transient voltage
stability characteristics when compared to traditional induction generators. This
section presents the mathematical modeling of DFIG. The steady-state expressions
of DFIG are assumed and they are represented by,
 
vds = −rs i ds + (xs + xm )i qs + xm i qr (1)

vqs = −rs i qs − [(xs + xm )i ds + xm i dr ] (2)

 
vdr = −r R i dr + (1 − ωm ) (x R + xm )i qr + xm i qs (3)

vqr = −r R i qr − (1 − ωm )[(x R + xm )i dr + xm i ds ] (4)

In the above equations, xm and x S are magnetizing reactance and stator reactance
respectively. Similarly x R and r R are rotor reactance and resistance respectively [21,
22]. vds , vqs are the stator voltages of d and q axes. Similarly vdr , vqr are the rotor
voltages of d and q axes [23]. It may be noted that stator voltage mostly depends on
the magnitude and phase angle, and they are expressed as,

vds = V sin(−θ ) (5)

vqs = V cos(θ ) (6)

Here θ is the angle between V and vqs . The active power (P) and the reactive
power (Q) that are injected into the utility grid depends on the stator and grid-side
currents of the converter. These are expressed as [24],

P = vds i ds + vqs i qs + vdc i dc + vqc i qc (7)

Q = vqs i ds − vds i qs + vqc i qc − vdc i qc (8)

These can be rewritten by considering the grid-side converter and converter power
equations and they are expressed as,

Pc = vdc i dc + vqc i qc (9)

284 D. Rakesh Chandra et al.

Q c = vqc i dc − vdc i qc (10)

Rotor side active power (Pr ) and reactive power (Q r ) are

Pr = vdr i dr + vqr i qr (11)

Q r = vqr i dr − vdr i qr (12)

Further, if the converter is assumed to be lossless then the active power (Pc ) and
reactive power (Q c ) on the converter side are Pc = Pr and Q c = 0. Finally, the
injected power into the grid is given by [25],

P = vds i ds + vqs i qs + vdr i dr + vqr i qr (13)

Q = vqs i ds − vds i qs (14)

where i ds , i qs are the stator currents of the d and q axes, respectively. i dr and i qr
are the rotor currents of the d and q axes, respectively. In the generator, the motion
equation of a single shaft model is utilized. In this work, it can be considered that
converter controls can filter the dynamics of the shaft. Therefore,

ω̇m = (Tm − Te )/2Hm (15)

Te = ψds i qs − ψqs i ds (16)

The stator flux and generator currents relationship can be expressed as,

ψds = −[(xs + xm )i ds + xm i dr ] (17)

 
ψqs = − (xs + xm )i qs + xm i qr (18)

ψds and ψqs are stator fluxes on the d-axis and q-axis respectively. Then, electrical
torque (Te ) is expressed as,
 
Te = xm i qr i ds − i dr i qs (19)

The mechanical torque (Tm ) is given by [26],

Pw
Tm = (20)
ωm
Transient Stability Enhancement of Power … 285

2.2 Implementation of Wind Turbine Integrated Models

for TS Analysis Using PSAT Software

Wind power grid integration creates stability problems. The world is facing a severe
energy security problem with demand for energy going to be triple by 2050. Also,
increasing restrictions on conventional energy sources to resist environmental change
is giving rise to increased use of renewable resources for electricity production.
Wind energy being uncertain and dynamically generates the issue of the storage
facility when integrated into the grid. TS studies are addressed using PSAT which
is MATLAB-compatible software. TS studies have been done by applying a 3 phase
fault at time t = 3 s and the duration of fault is 150 ms. The step-by-step procedure to
implement the wind turbine integrated models for the TS analysis using PSAT [27,
28] is presented next:
• Step 1: Install PSAT software in MATLAB R2009a version.
• Step 2: Create the new data file (SIMULINK model) in PSAT (MATLAB
compatible toolbox) by clicking File/Open/New data file.
• Step 3: Construct SCIG wind turbine integrated to IEEE 14 bus system model
with 3 phase fault (at time t = 3 s and fault clearing time is 50 ms) in MATLAB
SIMULINK using PSAT library and save this file with a suitable name.
• Step 4: Give practical wind speed data as an input file to the wind turbine model
by editing the PSAT data file (.m model).
• Step 5: Load the data file by clicking “File/Open/ datafile/select data file (from
saved file)/load”.
• Step 6: After loading the data file, run power flow by clicking on the option “Power
Flow” (for this select Newton–Raphson method which is available in settings).
• Step 7: After executing power flow, carry out time-domain simulation (for this
select Trapezoidal method which is available in settings) by selecting the option
as “Time Domain” and simulate for 20 s of time (for this keep ending time = 20 s).
The system may reach a steady-state within 3–4 s. But to have further insight into
system behavior the study period is extended to 20 s.
• Step 8: After time-domain simulation, plot the curves such as rotor angles, bus
angles, and voltages, etc. for the time by clicking on the option “Plot” which is
on the right side of the PSAT tool-bar.

3 Reliability Test System (RTS) 24 Bus Test System

In this chapter, an RTS-24 bus test system has been used for the TS analysis and the
enhancement of grid-connected DFIG based wind turbine. This test system includes
10 generating units, one synchronous condenser, 4 transformers, and 38 transmission
lines. These transmission lines have two different voltage levels, i.e., 230 and 138 kV.
The 138 kV is in the lower portion whereas 230 kV is in the upper portion connected
by 230/138 kV transformers at buses 11, 12, and 24. The wind turbine is connected
286 D. Rakesh Chandra et al.

to RTS 24 bus system at bus number 14 as shown in Fig. 2. The total generation of
all conventional generators is 2340 MW in the case of the RTS-24 bus system. As
mentioned earlier, at bus number 14 wind turbine generator (i.e., DFIG) with the
capacity of 240 MW (all wind turbines together) is integrated, which is more than
10% of total conventional generation. A three-phase fault was applied at the 11th bus
at time t = 3 s, and fault clearing time is set to 50 ms. In RTS 24 bus system, the
wind turbines are connected at bus number 14 which has been chosen randomly. A
3-phase fault is applied at the 11th bus as shown in Fig. 2. The fault is applied at t =
3 s and the fault clearing time is 50 ms.

Fig. 2 PSAT Simulated RTS 24 bus system with DFIG wind turbine at bus 14 and fault at bus
number 11
Transient Stability Enhancement of Power … 287

4 Results and Discussion

The analysis of bus angles, rotational speeds of generators, generator rotor angles, and
voltages at all buses have been discussed in this section. The test system used for the
analysis is RTS 24 bus system with the integration DFIG wind turbine. Comparisons
of all the results of transient stability study under severe fault conditions are also
presented. DFIG wind turbines can supply the active power at constant frequency
and voltage despite rotor speed variation.
After execution of power flow, time-domain simulation has been carried out for
20 s. The fault is applied at t equal to 3 s to evaluate system transient stability and
fault is cleared at 50 ms. Figures 3, 4, 5 and 6 show the variation of rotor angles,
angular speed, bus angles, and voltage magnitudes of DFIG wind turbine integrated
RTS 24 bus system which is subjected to three-phase fault at bus number 11.
Figures 3, 4, 5 and 6 show that due to the application of the 3-phase fault at bus 11,
the rotor angles have varied from 1 to 1.4 radians, and finally reached to steady-state
as shown in Fig. 3. Figure 4 describes the variation of angular speeds in the range of
0.97–1.1 p.u and reached steady on the removal of a fault. Figures 5 and 6 indicate
that bus angles reached 0.15 radians, the voltage dropped to 0.3 p.u during the fault
and all these values have reached a steady-state after removing fault within simulated
time. From the above analysis, it can be concluded that the system is stable.
In the DFIG wind generator, the maximum peak overshoot is less, and hence
the steady-state is achieved in less time after clearing the fault. Further, with DFIG
transient stability can be improved using the rotor resistance control method and is
discussed in Sect. 4.1.

1.5

δ
Syn 1

δ
1 Syn 2

δ
Syn 3

δ
Syn 4
0.5
Rotor angle in radians

δ
Syn 5

δ
Syn 6

0 δ
Syn 7

δ
Syn 8

δ
-0.5 Syn 9

δ
Syn 10

-1

-1.5
0 2 4 6 8 10 12 14 16 18 20

Time in seconds

Fig. 3 Rotor angles of different generators for 24 bus system with DFIG wind turbine, fault at bus
11
288 D. Rakesh Chandra et al.

1.1 ω
Syn 1
ω
Syn 2
1.08
ω
Syn 3
ω
Angular speed in pu

1.06 Syn 4
ω
Syn 5
1.04 ω
Syn 6
ω
Syn 7
1.02
ω
Syn 8
ω
1 Syn 9
ω
Syn 10
0.98

0.96
0 2 4 6 8 10 12 14 16 18 20

Time in seconds

Fig. 4 Angular speeds at different synchronous generators for 24 bus systems with DFIG wind
turbine, fault at bus 11

θBus 1
0.15
θBus 2
θBus 3
0.1
θBus 4
θBus 5
0.05 θBus 6
θBus 7
0 θBus 8
Bus angle in radians

θBus 9
-0.05 θBus 10
θBus 11
-0.1 θBus 12
θBus 13
-0.15 θBus 14
θBus 15
-0.2 θBus 16
θBus 17

-0.25 θBus18
θBus 19

-0.3 θBus 20
θBus 21
θBus 22
-0.35
0 2 4 6 8 10 12 14 16 18 20
θBus 23
Time in seconds θBus 24

Fig. 5 Bus angles at various busses for 24 bus system with DFIG, fault at bus 11

4.1 Transient Stability Enhancement of DFIG Integrated

RTS 24 Bus System

TS enhancement is the process of improving the system damping and reducing the
maximum peak overshoot and thereby helping the system to reach a steady-state
quickly. For stability improvement, external rotor resistance (variable resistance) is
Transient Stability Enhancement of Power … 289

VBus01
1.1 VBus02
VBus03
VBus04
1
VBus05
VBus06
VBus07
0.9
VBus08
VBus09

0.8 VBus10
Voltage in pu

VBus11
VBus12
0.7 VBus13
VBus14
VBus15
0.6
VBus16
VBus17

0.5 VBus18
VBus19
VBus20
0.4 VBus21
VBus22
VBus23
0 2 4 6 8 10 12 14 16 18
VBus24 20

Time in seconds

Fig. 6 The voltage at various busses for 24 bus systems with DFIG, fault at bus 11

added on the rotor side which is possible in DFIG (slip ring) based wind turbines and
this method is known as rotor resistance control. In this section rotor, the resistance
control method is used to enhance the TS of the RTS 24 bus system. This is the usual
practice in the case of a wound rotor (slip ring) induction generator to improve torque-
slip controllability. In this method, an external resistor is added to the rotor windings
through a power electronic converter. Adding external resistance on the rotor side
causes more losses. So, resistance should be selected optimally to improve TS as
well as considering losses. In the present work, external resistance is changed with
different values and finally fixed at 50% of the original value. Studies are conducted
on DFIG integrated RTS 24 bus system and results are discussed below. The rotor
resistance has been modified from 0.01 to 0.015 p.u, and all other parameters remain
unchanged in the waveforms shown below.
Figure 7 depicts the rotor angle enhancements with rotor resistance control of RTS
24 bus system, fault at bus 11. Figure 7a presents the rotor angle enhancements with
Rs is 0.01 p.u, Xs is 0.10 p.u, Rr is 0.01 p.u and Xr is 0.08 p.u. Figure 7b presents the
rotor angle enhancements with Rs is 0.01 p.u, Xs is 0.10 p.u, Rr is 0.015 p.u and Xr
is 0.08 p.u. In Fig. 7, the maximum rotor angle is 1.4 radians (approximately) before
adjustment of rotor resistance (from Fig. 7a) and after rotor resistance modification
maximum rotor angle is limited to 1.3 radians (from Fig. 7b).
Figure 8 depicts the angular speed enhancements with rotor resistance control of
RTS 24 bus system, fault at bus 11. Figure 8a presents the angular speed enhance-
ments with Rs is 0.01 p.u, Xs is 0.10 p.u, Rr is 0.01 p.u and Xr is 0.08 p.u. Figure 8b
presents the angular speed enhancements with Rs is 0.01 p.u, Xs is 0.10 p.u, Rr is
290 D. Rakesh Chandra et al.

1.5

δ
Syn 1

δ
1 Syn 2

δ
Syn 3

δ
Syn 4
0.5
δ
Rotor angle in radians

Syn 5

δ
Syn 6

0 δ
Syn 7

δ
Syn 8

δ
Syn 9
-0.5

δ
Syn 10

-1

-1.5
0 2 4 6 8 10 12 14 16 18 20

Time in seconds

(a)

1.5

δ
Syn 1

1 δ
Syn 2

δ
Syn 3

δ
Syn 4
0.5
δ
Rotor angle in radians

Syn 5

δ
Syn 6

0 δ
Syn 7

δ
Syn 8

δ
Syn 9
-0.5
δ
Syn 10

-1

-1.5
0 2 4 6 8 10 12 14 16 18 20

Time in seconds

(b)

Fig. 7 Rotor angle enhancements with rotor resistance control of RTS 24 bus system, fault at bus
11
Transient Stability Enhancement of Power … 291

1.1 ω
Syn 1
ωSyn 2
1.08
ωSyn 3
Angular speed in pu

ωSyn 4
1.06
ωSyn 5
1.04 ω
Syn 6
ωSyn 7
1.02
ω
Syn 8
ωSyn 9
1
ω
Syn 10
0.98

0.96
0 2 4 6 8 10 12 14 16 18 20

Time in seconds
(a)

1.1 ω
Syn 1

1.08
ω
Syn 2
ωSyn 3
Angular speed in pu

1.06
ωSyn 4
1.04 ω
Syn 5

1.02
ωSyn 6
ωSyn 7
1
ωSyn 8
0.98 ωSyn 9
ω
0.96 Syn 10

0.94

0.92
0 2 4 6 8 10 12 14 16 18 20

Time in seconds
(b)
Fig. 8 Angular speed enhancements with rotor resistance control of RTS 24 bus system, fault at
bus11

0.015 p.u and Xr is 0.08 p.u. In Fig. 8, the synchronous machine angular speed is
1.1 p.u before adjustment of rotor resistance (from Fig. 8a) and after modification of
rotor resistance, it is less than 1.06 p.u (from Fig. 8b).
Figure 9 depicts the bus angle enhancements with rotor resistance control of RTS
24 bus system, fault at bus 11. Figure 9a presents the bus angle enhancements with
Rs is 0.01 p.u, Xs is 0.10 p.u, Rr is 0.01 p.u and Xr is 0.08 p.u. Figure 9b presents the
bus angle enhancements with Rs is 0.01 p.u, Xs is 0.10 p.u, Rr is 0.015 p.u and Xr
292 D. Rakesh Chandra et al.

θBus 1
0.15
θBus 2
θBus 3
0.1
θBus 4
θBus 5
0.05 θBus 6
θBus 7
0 θBus 8
Bus angle in radians

θBus 9
-0.05 θBus 10
θBus 11
-0.1 θBus 12
θBus 13
-0.15 θBus 14
θBus 15
-0.2 θBus 16
θBus 17

-0.25 θBus18
θBus 19

-0.3 θBus 20
θBus 21
θBus 22
-0.35
0 2 4 6 8 10 12 14 16 18 20
θBus 23
Time in seconds θBus 24

(a)
θBus 1

0.15 θBus 2
θBus 3

0.1 θBus 4
θBus 5

0.05
θBus 6
θBus 7
θBus 8
0
θBus 9
Bus angle in radians

θBus 10
-0.05
θBus 11
θBus 12
-0.1
θBus 13
θBus 14
-0.15
θBus 15
θBus 16
-0.2
θBus 17
θBus 18
-0.25
θBus 19
θBus 20
-0.3
θBus 21
θBus 22
-0.35 θBus 23
0 2 4 6 8 10 12 14 16 18 20
θBus 24
Time in seconds

(b)

Fig. 9 Bus angle enhancements with rotor resistance control of RTS 24 bus system, fault at bus 11

is 0.08 p.u. In Fig. 9, bus angles peak values are near to 0.15 radians with Rr is 0.01
p.u (from Fig. 9a) and 0.1 radians after adding external rotor resistance to 0.015 p.u
(from Fig. 9b).
The above case studies on test systems conclude that an increase in rotor resistance
improves the TS of the system. This is because of the proportionality of the damping
ratio with the resistance. As the damping ratio of the system increases the maximum
Transient Stability Enhancement of Power … 293

peak overshoot and the oscillations of the system reduce thereby improving the TS
of the system. After execution of power flow, time-domain simulation has been
carried out for the 20 s. A fault is applied at t = 3 s to evaluate system transient
stability and fault is cleared at 50 ms. In DFIG wind turbine addition of external
resistance has improved system transient stability and the system could reach steady-
state quickly as damping ratio is enhanced due to proportional relationship with the
external resistance being added.

5 Conclusions

This chapter presents the transient stability (TS) analysis of power systems by incor-
porating the wind energy generators by Power System Analysis Toolbox (PSAT)
software. The detailed mathematical modeling of DFIG is also presented in this work.
In DFIG, the control action is performed by using a power electronic-based converter
that can be controlled by using rotor frequency, and hence the rotor speed. With the
help of this converter, the DFIG can import or export the reactive power support from
the main grid, and it has rotor-side control as well as power electronics-based control.
These controlling methods make DFIG wind turbines efficient. Further, in DFIG wind
turbine addition of external resistance has improved system transient stability and
the system could reach steady-state quickly as damping ratio is enhanced due to
proportional relationship with the external resistance being added. In the proposed
TS analysis, rotor angle variation, rotational speeds of individual generators, bus
voltages, and angles at all buses are analyzed. Simulation results are performed on
the RTS-24 bus system. All the obtained results are visualized with graphical repre-
sentation. It can be observed from the results that the deviations in the rotor angle
are low in magnitude and can show better damping by using the DFIG. As time
progresses, for the generator rotational speeds and voltage at individual busses, the
oscillations, and maximum peak values have reduced in the case of DFIG. Wind
power integration studies including the energy storage systems are the scope for
future research work.

Acknowledgements This research work was funded by “Woosong University’s Academic

Research Funding—2021”.

References

1. Karami ESZ (2013) Transient stability assessment of power systems described with detailed
models using neural networks. Int J Electr Power Energy Syst 45(1):279–292. https://doi.org/
10.1016/j.ijepes.2012.08.071
2. Lei Y, Mullane A, Lightbody G, Yacamini R (2006) Modeling of the wind turbine with a doubly
fed induction generator for grid integration studies. IEEE Trans Energy Convers 21(1):257–264.
https://doi.org/10.1109/TEC.2005.847958
294 D. Rakesh Chandra et al.

3. Ullah NR, Thiringer T (2007) Variable speed wind turbines for power system stability
enhancement. IEEE Trans Energy Convers 22(1):52–60. https://doi.org/10.1109/TEC.2006.
889625
4. Goudarzi N, Zhu WD (2013) A review on the development of wind turbine generators across
the world. Int J Dyn Control 1:192–202. https://doi.org/10.1007/s40435-013-0016-y
5. Li H, Chen Z (2008) Overview of different wind generator systems and their comparisons. IET
Renew Power Gener 2(2):123–138. https://doi.org/10.1049/iet-rpg:20070044
6. Thomas PC, Sreerenjini K, Pillai AG, Cherian VI, Joseph T, Sreedharan S (2013) Placement
of STATCOM in a wind integrated power system for improving the load ability. In: IEEE
international conference on power, energy and control, Dindigul, India, pp 110–114.https://
doi.org/10.1109/ICPEC.2013.6527633
7. Buch H, Bhagiya RD, Shah BA, Suthar B (2011) Voltage profile management in restructured
power system using STATCOM. In: Nirma University international conference on engineering,
Ahmedabad, India, pp 1–7. https://doi.org/10.1109/NUiConE.2011.6153279
8. Muljadi E, Nguyen TB, Pai MA (2008) Impact of wind power plants on voltage and transient
stability of power systems. In: IEEE energy 2030 conference, Atlanta, GA, USA, pp 1–7.
https://doi.org/10.1109/ENERGY.2008.4781039
9. Devaraj D, Jeevajyothi R (2011) Impact of wind turbine systems on power system voltage
stability. In: IEEE international conference on computer, communication and electrical
technology, Tirunelveli, India, pp 411–416. https://doi.org/10.1109/ICCCET.2011.5762510
10. Azizi AH, Rahimi M (2018) Dynamic performance analysis, stability margin improvement and
transfer power capability enhancement in DFIG based wind turbines at weak ac grid conditions.
Int J Electr Power Energy Syst 99:434–446. https://doi.org/10.1016/j.ijepes.2018.01.040
11. Akhbari A, Rahimi M (2020) Control and stability analysis of DFIG wind system at the load
following mode in a DC microgrid comprising wind and microturbine sources and constant
power loads. Int J Electr Power Energy Syst 117.https://doi.org/10.1016/j.ijepes.2019.105622
12. Prajapat GP, Senroy N, Kar IN (2018) Modeling and impact of gear train backlash on perfor-
mance of DFIG wind turbine system. Electr Power Syst Res 163(A):356–364. https://doi.org/
10.1016/j.epsr.2018.07.006
13. Morshed MJ, Sardoueinasab Z, Fekih A (2019) A coordinated control for voltage and transient
stability of multi-machine power grids relying on wind energy. Int J Electr Power Energy Syst
109:95–109. https://doi.org/10.1016/j.ijepes.2019.02.009
14. Bhukya J, Mahajan V (2019) Optimization of damping controller for PSS and SSSC to improve
stability of interconnected system with DFIG based wind farm. Int J Electr Power Energy Syst
108:314–335. https://doi.org/10.1016/j.ijepes.2019.01.017
15. Pillai AG, Thomas PC, Sreerenjini K, Baby S, Joseph T, Sreedharan S (2013) Transient stability
analysis of wind integrated power systems with storage using central area controller. In: IEEE
international conference on microelectronics, communications and renewable energy, pp 1–
5.https://doi.org/10.1109/AICERA-ICMiCR.2013.6575991
16. Ouyang H, Li P, Zhu L, Hao Y, Xu C, He C (2012) Impact of large-scale wind power integration
on power system transient stability. In: EEE international conference on innovative smart grid
technologies, pp 1–6.https://doi.org/10.1109/ISGT-Asia.2012.6303130
17. Lopez YU, Navarro JAD (2008) Small signal stability analysis of wind turbines with squirrel
cage induction generators. In: IEEE international conference on transmission and distribu-
tion conference and exposition, Bogota, Colombia, pp 1–10. https://doi.org/10.1109/TDC-LA.
2008.4641779
18. Petersson TT, Harnefors L, Petru T (2005) Modeling and experimental verification of grid
interaction of a DFIG wind turbine. IEEE Trans Energy Convers 20(4):878–886. https://doi.
org/10.1109/TEC.2005.853750
19. Kadri A, Marzougui H, Aouiti A, Bacha F (2020) Energy management and control strategy
for a DFIG wind turbine/fuel cell hybrid system with super capacitor storage system. Energy
192.https://doi.org/10.1016/j.energy.2019.116518
20. Xiong L, Li P, Wu F, Ma M, Khan MW, Wang J (2019) A coordinated high-order sliding mode
control of DFIG wind turbine for power optimization and grid synchronization. Int J Electr
Power Energy Syst 105:679–689. https://doi.org/10.1016/j.ijepes.2018.09.008
Transient Stability Enhancement of Power … 295

21. Zhang T, Yao J, Pei J, Luo Y, Zhang H, Liu K, Wang J (2020) Coordinated control of HVDC
sending system with large-scale DFIG-based wind farm under mono-polar blocking fault. Int
J Electr Power Energy Syst 119.https://doi.org/10.1016/j.ijepes.2020.105943
22. Jerin ARA, Prabaharan N, Palanisamy K, Umashankar S (2017) FRT capability in DFIG based
wind turbines using DVR with combined feed-forward and feed-back control. Energy Proc
138:1184–1189. https://doi.org/10.1016/j.egypro.2017.10.233
23. Gebru FM, Khan B, Alhelou HH (2020) Analyzing low voltage ride through capability of
doubly fed induction generator based wind turbine. Comput Electr Eng 86.https://doi.org/10.
1016/j.compeleceng.2020.106727
24. Yao J, Pei J, Xu D, Liu R, Wang X, Wang C, Li Y (2018) Coordinated control of a hybrid wind
farm with DFIG-based and PMSG-based wind power generation systems under asymmetrical
grid faults. Renew Energy 127:613–629. https://doi.org/10.1016/j.renene.2018.04.080
25. Adetokun BB, Muriithi CM, Ojo JO (2020) Voltage stability assessment and enhancement of
power grid with increasing wind energy penetration. Int J Electr Power Energy Syst 120.https://
doi.org/10.1016/j.ijepes.2020.105988
26. Hossain MdE (2017) Performance analysis of diode-bridge-type non-superconducting fault
current limiter in improving transient stability of DFIG based variable speed wind generator.
Electr Power Syst Res 143:782–793. https://doi.org/10.1016/j.epsr.2016.09.020
27. Okedu KE (2017) Augmenting DFIG wind turbine transient performance using alternative
voltage source T-type grid side converter. Renew Energy Focus 18:1–10. https://doi.org/10.
1016/j.ref.2017.02.004
28. Milano F (2005) Power system analysis toolbox. Version 1.3.4, software and documentation,
July 14, 2005
Design and Analysis of BLDC Motor
Driver for Hybrid Electric Vehicles

Seong-Cheol Kim, Narasimha Sangam, Sravanthi Pagidipala,

and Surender Reddy Salkuti

Abstract This chapter presents the design and analysis of brushless DC (BLDC)
motor driver for hybrid electric vehicles. Due to the combustion of petroleum, it is
going to create a pollution problem. Therefore, it is needed that alternating energy
resources for petroleum reserves which are exhaustible in potential are required to
be determined. The majority of the automobiles right now are reliant on internal
combustion engines (ICEs) for the operation and it the root cause of worry since they
are accountable for smog. Therefore, the automobile companies are searching for
alternate energy sources which could decrease pollution. As a result of arising pollu-
tion issue, the hybrid electric-powered vehicles are important for long term planning.
The brushed DC machine works on a setup of wound wire coils and armature which
acts as a two-pole electromagnet. The brushless DC machine, by comparison, uses
a lasting magnet as it has an outside rotor. Additionally, it runs on 3 phases of coils
along with a special sensor that monitors the rotor position. This chapter presents
the relative evaluation of the electric vehicle (EV) that is powered by brushless DC
(BLDC) motor and the corresponding impact on the state of ripples and charge in the
DC voltage at the battery power. This comparison has been carried in the MATLAB
application and the outcomes are explained clearly in the results section.

Keywords Brushless DC motor · Hybrid electric vehicles · DC-DC converters ·

Pulse width modulation · Regenerative energy · Voltage source converter

S.-C. Kim · S. R. Salkuti (B)

Department of Railroad Electrical Systems, Woosong University, Daejeon 34606, Republic of
Korea
e-mail: [email protected]
N. Sangam
Department of Electrical and Electronics Engineering, TKR College of Engineering and
Technology, Hyderabad, India
e-mail: [email protected]
S. Pagidipala
Department of Electrical Engineering, National Institute of Technology Andhra Pradesh
(NIT-AP), Andhra Pradesh, India

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 297
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6_12
298 S.-C. Kim et al.

Nomenclature

BEV Battery electric vehicle

BLDC Brushless direct current
CO Carbon monoxide
CO2 Carbon dioxide
EVs Electric vehicles
ICE Internal combustion engine
HEV Hybrid electric vehicles
RBS Regenerative braking system
PID Proportional integral derivative
FLC Fuzzy logic control
Li-ion Lithium-ion
SOC State of charge
PHEVs Plug-in hybrid electric vehicles
NiMH Nickel–Metal Hydride
VSI Voltage Source Inverter
PWM Pulse Width Modulation
SPWM Sinusoidal pulse width modulation
VSC Voltage Source Converters
IGBTs Insulated gate bipolar transistors

1 Introduction

The transportation sector is growing rapidly throughout the world and it mostly
depends on oils. The fossil fuels utilized in mainstream automobile will release
various greenhouse gases such as carbon monoxide (CO), carbon dioxide (CO2 ), and
methane. The exorbitant of gases causes pollution associated with the environmental
surroundings, climate modification, and warming. To reduce these affects, automo-
bile industry is now certainly moving towards the electric vehicle (EV) technology
[1, 2]. The electric vehicles (EVs) have reduced fuel price in conformity with fossil-
fueled automobiles since they are primarily built with battery pack systems. The
battery power coordination in an EV is the large amount of element this is certainly
important in determining the distance. The electrical automobiles are trusted for
pollution free transportation environment as they operate with an electrical battery
[3]. In order to increase the distance travelled by the EVs and to increase the recap-
turing of the regenerative energy, one must enhance the performance of all compo-
nents utilized in the EVs such as electric engine, energy converter, and power storage
system, i.e., battery power or ultracapacitor.
In recent years, EVs have drawn much attention as compared to conventional
internal combustion engine (ICE) vehicles. This contemplation is due to the economic
and environmental troubles concerns linked with the utilization of natural gas and
Design and Analysis of BLDC Motor Driver for Hybrid … 299

petroleum fuels, and these are used as fuel in ICE vehicles. More advancement and
researches on vehicles like fuel cell and hybrid cars, pure battery powered EVs, etc.
are pursued actively [4]. Owing to the facts that these vehicles are reasoned to serve as
an efficient means to deal with global warming caused by the tail pipe-auto emissions.
In the case of eco vehicles, the compatibility between safety and running performance
has to be ensured especially in the case of EVs where electric devices like batteries,
converters and inverters are part of their propulsion force producing system. If such
an electric device fails during the motion of the EV, the EVs shall fails to run safely
and comfortably. This kind of failure in motion may constitute series traffic accidents
like rear end collisions. As an advancement for developing the next generation of
EVs, fail safe functions must be interlinked within the propulsion force generating
system, there by the EV can persist to run safely even if malfunctions occurs during
EV motion [5]. The hybrid electric vehicles (HEV), which mainly contains at a
minimum of two sources of power, first one is a primary power source and the other
one is named as secondary power source, which have the favors of both the EV and
conventional ICE vehicles also had the ability to vanquish their disadvantages.

1.1 Related Work

Unlike the conventional vehicles, the running capability of battery electric vehicle
(BEV) depends completely on the electric motor and battery unit; as there is no
involvement of traditional ICEs. In order to recharge the battery to a sufficient amount,
it should be plugged with an external electricity source [6, 7]. It can also be recharged
by regenerative braking process like all EVs. These vehicles have the merits like
no emissions, gases, free from compulsory oil changes, comparatively low cost of
running, fast and real smooth acceleration, and also have the capacity to conveniently
charge at home. But it also have the demerits like shorter range than gasoline vehicles
and moderately expensive than its gasoline equivalent despite the fact that within the
span of 2–3 years the gasoline savings will pay off.
Reference [8] reviews the technology and applications of electrical vehicles. It
critically appraises the level of state-of-the-art research improvements in this industry
developed in the last decade. It also highlights the technical issues of automobile tech-
nology. Reference [9] presents which converter or drive is suitable for the electric car.
Regenerative stopping can improve power usage efficiency which in turn improves
the driving distance of electric automobiles. The importance and characteristics of
regenerative braking system (RBS) has been presented in the reference [10]. The
RBS is adjusted to DC that is brushless BLDC engine, and it emphasizes in the
circulation with this particular force that is powerful in stopping alongside BLDC
motor control. In this work, BLDC motor control uses proportional integral derivative
(PID) control; consequently, the flow of stopping force uses the fuzzy logic control
(FLC). The braking torque could be managed in real-time by PID control because it
is certainly slower than PID control.
300 S.-C. Kim et al.

Reference [11] has designed the effectiveness of DC engine which is highly inex-
pensive. The controller takes input that is individual manage the rate and drives the
engine at that rate regardless of load. This permits fast, smooth motor response. A
converter drives the motor by utilizing the power through the battery pack. Refer-
ence [12] is targeted at the optimization of core size power inductors in bidirectional
DC-DC converters. It describes a characterization that is certainly experimental are
large-signal for energy inductors in EV applications and an answer getting an induc-
tance guide value when it comes to ability inductor in DC-DC converters simula-
tion studies. A prototype of one kilowatt inductive wireless power transfer power
converters with high frequency supply is modeled in reference [13]. Reference [14]
proposes the modeling and evaluation of sizing of battery pack required by using the
Simpson in R Studio.

1.2 Motivation and Contributions

This chapter provides the comparative analysis of DC machine powered electric

vehicle and BLDC powered electric vehicle, and the corresponding effect on state of
charge and ripples in the dc voltage at the battery. This comparison is carried out in the
MATLAB software and the comparative results are given individually. The objective
of this work is to provide a better percentage of the state of charge (SOC) to make
the design much better and efficient. This work discusses the concerns related to the
battery storage systems, their types along with their benefits and drawbacks. Also
describes the types of converters and primarilyDC-DC converters, their advantages,
and just how they are found in this task is certainly proposed. Simulation results
and analysis linked to the proposed work are also discussed. In short, the main
objectives are to provide better SOC, much better design, improved efficiency, and
less maintenance.
Organization of this chapter: In Sect. 2, the description and classification of battery
storage system is presented. Section 3 describes the modeling of DC-DC converter.
In Sect. 4, simulation results are described. In the last section, the whole chapter is
concluded.

2 Battery Storage System

Generally, there are two types of batteries, primary and secondary. Normally, the
standards for the secondary/rechargeable batteries rate the capacity over a 4, 8 h, or
even longer discharge times. The lithium-ion (Li-ion) battery is used in the design of
this hybrid EV due to its low cost and good performance and light weight [15]. It has
a nominal voltage of 350 V and an initial state of charge (SOC) of 88%. A power kit is
some sort of computer product comprising a genuine degree of electrochemical cells
with outside junctions for powering devices which might be electric as flashlights,
Design and Analysis of BLDC Motor Driver for Hybrid … 301

cellular phones, and automobiles that happen to be electric. These cells are wet
electrolytes fluids that are vulnerable to leakage and spillage [16]. Batteries convert
chemical power directly to provide power that is electric.
The various energy sources like batteries, ultracapacitors and flywheels of EVs
can retrieve and store regenerative energy effectively. Energy management subsystem
unit along with vehicle controller unit processes the energy recovery and there by the
regeneration process is controlled. In order to monitor the energy source usability and
regulate the refueling, the energy refueling unit is utilized. For every EV auxiliaries
including climate control and power steering unit, and the sufficient power with
various voltage levels are delivered by the auxiliary power supply subsystem.
In plug-in hybrid electric vehicles (PHEVs), batteries reflect as the greatest means
of source of energy. To accomplish the “all electric driving range” insistence the
powertrains of PHEV’s should carry large number of high rated battery packs with
abundant energy and superior power density. Energy supplied from the high rated
battery pack should be adequate enough for the impulses to sustain the continuous
speed range [17]. Through within the all electric driving range the batteries also need
to deliver power to the auxiliaries like air conditioning unit or power steering unit etc.
But the use of large quantities of onboard high rated batteries bring up the questions
about safety regarding with chances of fire threat, accidents or even the possibilities
of short circuit during either operation cannot be avoided. It creates another critical
situation; there by the selection of battery technology is highly crucial in considera-
tion with their efficiency, price, and authenticity. The mostly acknowledged batteries
in PHEVs are currently Lithium-ion (Li-ion) batteries. They can dispense sufficient-
energy density and highly demanded-power density for the batteries of equal weight,
and ensures wider all electric drive range with best ever performance of vehicle. One
of the other kinds of commercialized battery for PHEVs is Nickel–Metal Hydride
(NiMH) batteries.
The EVs are designed to steer with high voltage electrical energy stored batteries.
By this way the exhaust emissions can be excluded which further reduces the air
pollution since gasoline or other fossil fuels are not required for the propulsion.
During the entire driving range, the noises and vibrations from engine is negligibly
small in comparison with traditional ICEs. The electric motor employed in EVs
is generally higher performance motors [18]. During the downhill or decelerating
motion of the said vehicle the restoring of brake energy by regeneration can be
employed and will be stored to the high voltage batteries. There by the driving range
from single charging will be extended and also reduce the loss during the braking. In
case existing charge of battery is not adequate to drive the vehicle, various charging
methods like AC or DC charge is applied. The trickle charge can also be applied.
The incorporation of an ICE with electric motor is the most widely used combo
for the generation of propulsion force in hybrid EVs. With this integration, energy
efficiency will be improved and vehicle emissions will be fall off due to the electric
motor and driving range will extended because of the ICE. The stereotypical ICE
vehicles show good performances along with large operating range clearly because
of the high energy-density yielded by the petroleum products. But these common
ICE vehicles have the disadvantage of indigent fuel economy and greater risks of
302 S.-C. Kim et al.

environmental contamination [19]. By far the reasons behind their indigent fuel
economy are the contrast between real operation requirements and the vehicles fuel
expertise, operating in urban areas, the wastage of kinetic energy during braking,
and due to the stop and go driving pattern the productivity of hydraulic transmission
in current day automobiles.
The existent EVs powered by battery packs have dominance over typical ICE
vehicles by high energy productivity and almost zero environmental contamination.
The comparison of these vehicles based on the performance of their driving range
per battery charge ICEs hold the upper hand. However, due to the use of two power
sources, hybrid electric vehicles can outsmart both traditional ICEs and EVs.

3 DC-DC Converter

The power electronic devices in PHEVs comprises of inverters, converters can be

bidirectional or unidirectional, chargers etc. The inverter serves to transfer the battery
DC power in to the required AC power for propulsion of electric motor(s). Another
important fact is that, with the motor drive power electronic components, the retrieval
of regenerative energy and their storage in batteries are possible. Diverse numbers of
DC-DC converters are beneficial for applications that require multiple voltage levels
[20]. In contemplation of regenerative energy storage in to battery bidirectional power
flow has to be ensured regarding with the converter. Apart from these AC to DC
converters are essential for battery chargers, as they transfigure the AC power flow
of grid to suitable DC power to charge the battery unit. To achieve high efficiency
batteries programmable digital controllers having proper voltage and current profiles
are needed along with proper power factor correction. Figure 1 depicts the buck-boost
converter connected with DC machine. The fuzzy logic control (FLC) finds the duty
cycle of S1 and S2 to ensure the charge and discharge of battery. Buck-boost converter
is a kind of DC-to-DC converter which contains a manufacturing voltage magnitude
that is often more than or perhaps much less in comparison with the voltage magnitude
of input, and this is equivalent to a flyback converter working with an inductor [21,
22].

Fig. 1 Buck-boost converter

connected with DC machine
S1 Ra
D1

C2 La

VB C1 S2 D2 Ea
Design and Analysis of BLDC Motor Driver for Hybrid … 303

Fig. 2 The schematic diagram of VSI

3.1 Voltage Source Inverter (VSI)

The main purpose of VSI is to convert a constant DC voltage into an AC voltage

with variable magnitude and frequency. The VSI maintains a fixed polarity of DC
voltage, and the reversal of power can be obtained without reversing the direction of
the current. The schematic view of VSI is depicted in Fig. 2.

3.2 Pulse Width Modulation (PWM)

It is a modulation technique which is used to encode a pulsating signal. The switching

frequency of PWM is much higher than what would affect the load. By doing that the
resultant waveform is as smooth as possible [23, 24]. The main advantages of using
PWM are cheap to create, consumes less energy, use of high frequencies (around
40 to 100 kHz), extremely efficient in energy-smart while dimming the radiant light
or perhaps in voltage conversions, high control ability of power and high efficiency,
i.e., around 90%. A modulation strategy is helpful to encode a communication as an
illustration that is pulsing. This modulation strategy assists to encode information
for transmission, and it allows the control related to the power supplied which could
be electrical products and services. Figures 3 and 4 depict the block diagram and

Fig. 3 SPWM block

Sine wave
diagram
generator
Comparator
Ramp generator
304 S.-C. Kim et al.

Fig. 4 SPWM simulation diagram

simulation diagram of sinusoidal pulse width modulation (SPWM). The (ON and
OFF time frames) period of pulse is continuous. The (ON time and OFF period)
about the pulse is called period with this pulse, and the ratio attached with the time
that is full punctually that is from the period is named as duty cycle.

3.3 Voltage Source Converters (VSC)

The polarity of a DC voltage is typically fixed and it is smoothed by using a capaci-

tance [14, 25]. Because of this, an HVDC converter utilizing insulated gate bipolar
transistors (IGBTs) is often known as a voltage source converter (VSC). The control-
lability introduces several benefits; particularly the ability to change the ON/OFF
status of IGBT’s a couple of times per duration to have the ability to enhance
harmonic performance. A two-level converter can function as the type that is most
basic of three-phase VSC and will undoubtedly be seen as a six pulse bridge [26,
27]. Figure 5 depicts the two-level VSC for HVDC.
Some HVDC methods were constructed with three-level converters to improve
the negative performance are harmonic the converter is unquestionably two-level.
The three-level converters can synthesize 3 (instead of just two) discrete voltage
quantities while in the AC terminal of the phase [28]: + 21 Ud , zero, − 21 Ud . Here,
a common type of the converter is used [29], i.e., diode clamped converter with 4
IGBT valves in each phase.
Design and Analysis of BLDC Motor Driver for Hybrid … 305

Fig. 5 Two-level VSC for HVDC

4 Results and Discussion

Batteries of PHEVs are charged with the assistance of EV supply equipment (EVSE),
i.e., charging stations. The charging stations communicate along with vehicles to
guarantee that electricity flows in to the system will be shielded and pertinent. The
classifications of EVSE in plug-in vehicles are elicited from the rate at which batteries
will charge. The two major kinds of EVSE are ‘AC level 1’ and ‘AC level 2’, which
will be able to supply the vehicle with alternating current. The vehicle is equipped
with an on board electrical equipment/charger which transfer AC to DC for charging
the vehicle battery [30, 31]. DC fast charging might also appropriate for direct DC
supply to automobile, simply named as DC level 2.
Converter circuit configuration of hybrid electric vehicle (HEV) using DC
machine has been depicted in Fig. 6. The duty ratio is produced by fuzzy logic
control (FLC) with ramp revolution of changing and hence the pulse created for

Fig. 6 Converter circuit configuration of hybrid electric vehicle using DC machine

306 S.-C. Kim et al.

State of Charge of Battery

87.54

87.52

87.5

87.48
Battery SOC(in %)

87.46

87.44

87.42

87.4
0 5 10 15 20 25 30 35 40 45 50

Time (in Sec)

Fig. 7 Battery SOC (in %)

semiconductor switch to the group of zero and 25 moments. DC unit voltage is

elevated on the worthiness that is needed.
The battery energy state of charge (SOC) is improved based on the simulation
outcome. In the EV, the regenerative braking has happened in this state. The charge
and also the discharge states of this particular battery energy are required for deter-
mining the distance of travel. Figure 7 depicts the plot of battery SOC (in percentage)
versus time for the converter system with DC machine. The output waveforms of
battery voltage and current are depicted in Fig. 8.
The proposed converter system with BLDC motor is simulated and it is shown
in Fig. 9. Figure 10 depicts the plot of battery SOC (in percentage) versus time for
the converter system with BLDC motor. The output waveforms of battery voltage
and current are depicted in Fig. 11. The voltage waveform of BLDC motor has been
depicted in Fig. 12.
The DC machines operate in generating mode in the case of battery is charged
and also the bi-directional DC-DC converter operates in the buck mode. In the EV,
regenerative braking has occurred in this state. In the BLDC machine, when the
battery is charged, the machine is operated in generator mode. The battery SOC has
been increased from 87.6 to 87.8%. Results being exemplary analysis are acquired
straight into the MATLAB and all the simulation results are analyzed for different
conditions alongside the SOC, the voltage quantities are also improved.

5 Conclusions

The charging and discharging states of battery plays an important role in determining
the distance to travel the electric vehicle. This chapter has presented the design and
Design and Analysis of BLDC Motor Driver for Hybrid … 307

1400
Battery Voltage(in V)

1200

1000

800

0 5 10 15 20 25 30 35 40 45 50

Time (in Sec)

20000

15000
Battery Current (in A)

10000

5000

-5000
0 5 10 15 20 25 30 35 40 45 50

Time (in Sec)

Fig. 8 Battery voltage and current versus time

Fig. 9 Proposed converter system with BLDC motor

analysis of brushless DC (BLDC) motor driver for the hybrid electric vehicles. The
relative evaluation of the electric vehicle (EV) that is powered by BLDC motor and
the corresponding impact on the state of ripples and charge in the DC voltage at the
battery power are also presented. This comparison has been carried in the MATLAB
application and the outcomes are explained clearly in the results section. Battery
pack energy has been enhanced from 87.47 to 87.55% according to the simulation
effect. In the all-electric automobile, regenerative braking has occurred. The charge
308 S.-C. Kim et al.

State of Charge of Battery

88.05

88

87.95

87.9
Battery SOC(in %)

87.85

87.8

87.75

87.7

87.65

87.6
0 5 10 15 20 25 30 35 40 45 50

Time (in Sec)

Fig. 10 SOC of battery (in %)

400
Battery Voltage(in V)

300

200

100

0
0 5 10 15 20 25 30 35 40 45 50

Time (in Sec)

150
Battery Current (in A)

100

50

-50
0 5 10 15 20 25 30 35 40 45 50

Time (in Sec)

Fig. 11 Battery voltage and current of converter system with BLDC motor
Design and Analysis of BLDC Motor Driver for Hybrid … 309

1600

1550

1500
Voltage of BLDC(in V)

1450

1400

1350

1300

1250

1200
0 5 10 15 20 25 30 35 40 45 50

Time (in Sec)

Fig. 12 Voltage of converter system with BLDC machine

and also discharge states of the electric battery may be many needed for distance
to determining. If the battery pack is charged, the unit which is dc operated turbine
setting and the bi-directional DC-DC converter will be operated in the buck mode.
Adjustable torque which is unfavorable is placed about the BLDC device as well
as influence the battery pack. Based upon the simulation results, the battery state of
charge (SOC) has been improved from 87.6 to 87.8%.

Acknowledgements This research work was funded by “Woosong University’s Academic

Research Funding-2021”.

References

1. Metin NA, Boyar A, Kabalci E (2019) Design and analysis of bi-directional DC-DC driver for
electric vehicles. In: 1st global power, energy and communication conference (GPECOM), pp
227–232. https://doi.org/10.1109/GPECOM.2019.8778532
2. Yong JY, Ramachandaramurthy VK, Tan KM, Mithulananthan N (2015) A review on the state-
of-the-art technologies of electric vehicle, its impacts and prospects. Renew Sustain Energy
Rev 49:365–385. https://doi.org/10.1016/j.rser.2015.04.130
3. Zhang J, Shen T (2018) Receding horizon optimal control of PHEV with demanded torque
estimation model. IFAC-PapersOnLine 51(31):183–187. https://doi.org/10.1016/j.ifacol.2018.
10.034
310 S.-C. Kim et al.

4. İnci M, Büyük M, Demir MH, İlbey G (2021) A review and research on fuel cell electric vehi-
cles: topologies, power electronic converters, energy management methods, technical chal-
lenges, marketing and future aspects. Renew Sustain Energy Rev 137. https://doi.org/10.1016/
j.rser.2020.110648
5. Kahveci O, Kaya MF (2021) Hydrogen production from Al–Cu alloy using electric vehicle’s
waste DC motor coils. Int J Hydrogen Energy. https://doi.org/10.1016/j.ijhydene.2021.06.201
6. Bayardo RG, Loukianov AG, Fuentes-Aguilar RQ, Utkin VI (2020) Adaptive speed tracking
controller for a brush-less DC motor using singular perturbation. IFAC-PapersOnLine
53(2):3880–3885. https://doi.org/10.1016/j.ifacol.2020.12.2100
7. Aliasand AE, Josh FT (2020) Selection of motor foran electric vehicle: a review. Mater Today:
Proc 24(3):1804–1815. https://doi.org/10.1016/j.matpr.2020.03.605
8. Zhang F, Zhang X, Zhang M, Edmonds ASE (2016) Literature review of electric vehicle
technology and its applications. In: 5th international conference on computer science and
network technology, pp 832–837. https://doi.org/10.1109/ICCSNT.2016.8070276
9. Tiwari A, Jaga OP (2017) Component selection for an electric vehicle: a review. In: International
conference on computation of power, energy information and communication (ICCPEIC), pp
492–499. https://doi.org/10.1109/ICCPEIC.2017.8290416
10. Nian X, Peng F, Zhang H (2014) Regenerative braking system of electric vehicle driven by
brushless DC motor. IEEE Trans Ind Electron 61(10):5798–5808. https://doi.org/10.1109/TIE.
2014.2300059
11. Haishui Z, Dahu W, Tong Z, Keming H (2010) Design on a DC motor speed control. In:
International conference on intelligent computation technology and automation, pp 59–63.
https://doi.org/10.1109/ICICTA.2010.795
12. Perdigão MS, Trovão JPF, Alonso JM, Saraiva ES (2015) Large-signal characterization of
power inductors in EV bidirectional DC–DC converters focused on core size optimization.
IEEE Trans Ind Electron 62(5):3042–3051. https://doi.org/10.1109/TIE.2015.2402632
13. Sandeep V, Shivanand MN, Reddy SS (2021) Design of a one kilowatt wireless charging
system for electric vehicle in line with Bharath EV standards. Int J Emerg Electr Power Syst
22(3):255–267. https://doi.org/10.1515/ijeeps-2020-0178
14. Sandeep V, Shastri S, Sardar A, Salkuti SR (2020) Modeling of battery pack sizing for electric
vehicles. Int J Power Electron Drive Syst 11(4):1987–1994. https://doi.org/10.11591/ijpeds.
v11.i4.pp1987-1994
15. Cipin R, Mach M, Toman M, Knobloch J (2017) Measurement and evaluation of DC motor
starting torque. In: IEEE international conference on environment and electrical engineering
and IEEE industrial and commercial power systems Europe, Milan, pp 1–5. https://doi.org/10.
1109/EEEIC.2017.7977475
16. Suresh K, Arulmozhiyal R (2016) Design and implementation of bi-directional DC-DC
converter for wind energy system. Circuits Syst 17:3705–3722. https://doi.org/10.4236/cs.
2016.711311
17. Sandeep V, Shivanand MN, Reddy SS (2021) Design of a one kilowatt wireless charging
system for electric vehicle in line with Bharath EV standards. Int J Emerg Electr Power Syst
22(3):255–267. https://doi.org/10.1515/ijeeps-2020-0178
18. Hur J, Sung HG, Lee BK, Won CY, Lee BH (2006) Development of high-efficiency 42V cooling
fan motor for hybrid electric vehicle applications. In: IEEE vehicle power and propulsion
conference, pp 1–6. https://doi.org/10.1109/VPPC.2006.364307
19. Pothirasan N, Rajasekaran MP, Muneeswaran V (2018) Real time reactive power compensation
for battery/photovoltaic hybrid power source for internet of hybrid electric vehicle system. Cogn
Syst Res 52:473–488. https://doi.org/10.1016/j.cogsys.2018.07.030
20. Gupte S (2014) Experimental analysis and feasibility study of 1400 CC diesel engine car
converted into hybrid electric vehicle by using BLDC hub motors. Energy Procedia 54:177–184.
https://doi.org/10.1016/j.egypro.2014.07.261
21. Salkuti SR (2021) Energy storage and electric vehicles: technology, operation, challenges, and
cost-benefit analysis. Int J Adv Comput Sci Appl 12(4):40–45. https://doi.org/10.14569/IJA
CSA.2021.0120406
Design and Analysis of BLDC Motor Driver for Hybrid … 311

22. Salehifar M, Moreno-Eguilaz M, Putrus G, Barras P (2016) Simplified fault tolerant finite
control set model predictive control of a five-phase inverter supplying BLDC motor in electric
vehicle drive. Electr Power Syst Res 132:56–66. https://doi.org/10.1016/j.epsr.2015.10.030
23. Lee BC, Song CH, Kim DH, Kim KC (2020) Study on process derivation and characteristic
analysis for BLDC motor design using dual rotor structure with high torque density. Energies
13(24):6745. https://doi.org/10.3390/en13246745
24. Kim H-J, Park H-S, Kim J-M (2020) Expansion of operating speed range of high-speed BLDC
motor using hybrid PWM switching method considering dead time. Energies 13(19):5212.
https://doi.org/10.3390/en13195212
25. Karki A, Phuyal S, Tuladhar D, Basnet S, Shrestha BP (2020) Status of pure electric vehicle
power train technology and future prospects. Appl Syst Innov 3(35):1–28. https://doi.org/10.
3390/asi3030035
26. Hassanin MA, Abdel-Kader FE, Amer SI, Abu-Moubarka AE (2018) Operation of brushless
DC motor to drive the electric vehicle. In: Twentieth international Middle East power systems
conference (MEPCON), pp 500–503. https://doi.org/10.1109/MEPCON.2018.8635158
27. Prasetyo HF, Rohman AS, Hariadi FI, Hindersah H (2016) Controls of BLDC motors in electric
vehicle testing simulator. In: 6th international conference on system engineering and technology
(ICSET), pp 173–178. https://doi.org/10.1109/ICSEngT.2016.7849645
28. Salkuti SR (2019) Optimal operation of microgrid considering renewable energy sources,
electric vehicles and demand response. E3S Web Conf 87:1–6. https://doi.org/10.1051/e3s
conf/20198701007
29. Purwadi A, Dozeno J, Heryana N (2013) Testing performance of 10 kW BLDC motor and
LiFePO4 battery on ITB-1 electric car prototype. Procedia Technol 11:1074–1082. https://doi.
org/10.1016/j.protcy.2013.12.296
30. Hwang MH, Lee HS, Yang SH, Cha HR, Park SJ (2019) Electromagnetic field analysis and
design of an efficient outer rotor inductor in the low-speed section for driving electric vehicles.
Energies 12(24):4615. https://doi.org/10.3390/en12244615
31. Carev V, Roháč J, Šipoš M, Schmirler M (2021) A multilayer brushless DC motor for heavy
lift drones. Energies 14(9):2504. https://doi.org/10.3390/en14092504
A Novel Approach for Power Quality
Improvement in Microgrid

Arvind R. Singh, Papia Ray, and Surender Reddy Salkuti

Abstract This chapter proposes an approach to improve the power quality (PQ) of
the three-phase system by manipulating the grid-connected smart Photovoltaic Distri-
bution Static Compensator (PV D-STATCOM) system with the help of a proposed
dynamic voltage restorer (DVR) and a reweighted zero attracting (RZA) control tech-
nique containing adaptive features with P and O (perturb and observe) approach based
maximum power point tracking (MPPT) method. The PV D-STATCOM is having a
PV array alongside the quality of active filtering converter is essential to transform
DC voltage into its equivalent AC voltage quantity with necessary PQ improvement
measures such as harmonic distortion elimination, reactive power compensation in
all 3 unbalanced phases of the concerned distribution system and with some other
remedial measures. The implementation of the RZA strategy, which is discussed in
this chapter, is showing reasonable steady-state behavior and transient responses.
The proposed work on grid-connected PV systems can efficiently perform around
the clock for constant and variable solar irradiation connected to a comparatively
linear load. This approach is smart enough to operate in both modes, such as iden-
tifying solar PV power and the capability of bidirectional or multidirectional power
flow without manual interventions. The proposed approach is designated as smart
as it can automatically accomplish both modes, sensing the PV power and gifted of
multidirectional power flow. The proposed system accomplishes the twin role of a
PQ refining device by operating as D-STATCOM ad by transferring solar power to
the consumer and grid acquired from the PV array.

A. R. Singh
School of Electrical Engineering, Shandong University, Jinan, China
P. Ray
Department of Electrical Engineering, Veer Surendra Sai University of Technology (VSSUT),
Burla, Odisha, India
e-mail: [email protected]
S. R. Salkuti (B)
Department of Railroad and Electrical Engineering, Woosong University, Daejeon 34606,
Republic of Korea
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 313
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6_13
314 A. R. Singh et al.

Keywords Power quality · PV D-STATCOM · Dynamic voltage restorer · Voltage

source inverter · Reactive power · Harmonics

Nomenclature

DG Distributed Generation
MG Microgrid
ESS Energy Storage Systems
PCC Point of Common Coupling
PV Photovoltaic
PQ Power Quality
RESs Renewable Energy Resources
VSI Voltage Source Inverter
DERs Distributed Energy Resources
DVR Dynamic Voltage Restorer
PV-DSTATCOM Photovoltaic Distribution Static Compensator
SSC Static Series Compensator
SVB Series Voltage Booster
RZA Reweighted Zero Attracting
AVR Automatic Voltage Regulator
MPPT Maximum Power Point Tracking
FFT Fast Fourier Transform
THD Total Harmonic Distortion
UPQC Unified Power Quality Conditioner

1 Introduction

The issues related to power quality are a significant challenge in today’s electricity
distribution and transmission scenario. The objective of power quality improvement
has enforced the introduction of susceptible and complex devices in the vicinity of
electrical infrastructure. Nonstandard voltage, disturbing power, and frequency are
a few of the power quality-related problems. This power quality problem results in
damage or failure of end-user equipment. A microgrid (MG) is a coordinated combi-
nation of a group of unpredicted loads, energy storage systems (ESS), and distributed
generation (DG) units that operate at the same level of distribution voltage to obtain a
harmonious operation [1]. The point of common coupling (PCC) is a concept which
indicates the node at which the MG is synchronized with the primary grid. Conven-
tionally the MGs prefer to operate in grid-connected mode to acquire the benefits of a
high inertial grid. Despite the choice of having grid connectivity, these MGs hold the
potential to feed a sufficient amount of emergency power to the trailing loads even
A Novel Approach for Power Quality Improvement in Microgrid 315

when the grid is getting disconnected. The progress of traditional microgrids (MGs)
has been followed with versatile objectives such as resolving the problem of growing
power system efficiency, beneficial exploitation of the unconventional renewable
energy resources (RESs) with minimum environmental impact, optimal manage-
ment of power fluctuation along with other power quality improvement measures,
and electrical network extension facilities deprived of violating the classical oper-
ating philosophies of conventional distribution systems [2]. For a reliable MG smooth
operation, it is necessary to handle some of the most critical problems, such as power
quality issues, energy balance, and safety. A unique and operative power quality
improvement technique for the MG operation has been proposed in this chapter with
necessary explanations.
Voltage Source Inverter (VSI) based MGs are being developed and designed with
various control strategies concentrating on supplying sinusoidal current and voltage
free of harmonics to the concerned system even when the non-linear or disturbed
loads are connected to the grid [2]. The improvement of power quality uncertainties
in the MGs is in demand to eradicate harmonics-related noises and deliver unpol-
luted sinusoidal waveform at rated frequency and voltage. Typically, MGs are more
vulnerable to the power quality troubles due to comparatively lesser inertia level
(because of the presence of intensive non-linear electronic devices in the concerned
distributed energy resources (DERs) [3]), perturbation of the electricity consump-
tions and power generation, lower short circuit level, the existence of numerous
vendors and decision-makers, the requirement of action mode changeover from a
grid-allied to sophisticated island mode and vice-versa.
Power quality is expressed as maintaining the magnitude of concerned voltage
waveform in its pure sinusoidal nature with prescribed magnitude, frequency within
a recommended limit without any abnormality in shape. Power quality-related issues
to the end-user equipment and consumers occur in the electrical system when there is
any abnormality in the waveform’s frequency and magnitude related to power beyond
a specified range [4]. The diverse kinds of power quality disturbances comprise
the unbalanced voltage, sags and swells in voltage waveform, harmonic distortion,
transients, short term disturbances like the under-voltages and over-voltages, power
outage, perceptible flicker, electrical noise, and voltage notching. Due to the high
intermittent characteristics of the existing distributed energy generators like wind
power, solar power, and fuel cells, the voltage generated from these resources cannot
be connected directly to the primary grids. However, this can be achieved with the help
of an interfacing converter between the outputs of these renewable energy resources
and the ac power distribution systems. Harmonics are generally referred to as the
distortion of the sinusoidal waveform by waveforms of different frequencies [5].
A developed MG model integrating solar generation, wind power systems, aqua
electrolysers, fuel cells, battery energy storage systems, and diesel generators are
proposed in the reference [6]. A comprehensive survey on MG to improve the
power quality parameters is taken as the main objective is presented in reference
[7]. A decision-making approach based on the MATLAB-Simulink model has been
suggested in [8] to evaluate the power quality features of the single-phase AC MG
with necessary explanations. Reference [9] addresses the power quality improvement
316 A. R. Singh et al.

of multi-area MGs. A new mathematical morphology-based shunt active filter for

improving MG’s power quality is described [10].
Similarly, the fuzzy logic control approach for the UPQC is discussed in refer-
ence [11] for unbalanced power networks like the low voltage distribution systems.
This fuzzy logic control scheme proposed here is used for the current control in-
place of proportional-integral controllers, which have a tolerable performance over a
wide variety of operating nodes. A vigorous voltage control technique with an angle
droop method to improve power quality is proposed in reference [12]. A robust
control approach for multifunctional grid-connected inverters under an unbalanced
loading condition and primary voltage environments is proposed [13]. Reference
[14] suggested an integrated approach to alleviating power quality difficulties in an
MG by optimally controlling the operational schedule of generation sources and load
demands.
It has been observed from the literature survey that proper research is still needed
for the development of an accurate scheme that can limit the harmonic disturbances
without compromising the reactive power compensation and power balance in all
three phases of the distribution system. This research gap is fulfilled in this chapter
by developing two methods that compensate for various power quality disturbances
like harmonic distortion. In the first method, a smart PV-DSTATCOM is suggested
in this chapter to mitigate power quality problems. It acts as an active filter that
nullifies the effect of non-linear load currents and improves the grid’s power factor,
regulating the DC-link voltage. The multidirectional flow of power has also been
achieved in this model [15]. The proposed system works in both the modes of constant
irradiation and variable irradiation with a constant load. The second method can
safeguard the sensitive consumer loads from numerous perturbations in the power
supply infrastructure by mitigating short-term power quality issues such as voltage
sags and swells with a dynamic voltage restorer (DVR). PV-DSTATCOM and DVR’s
main objective is to reduce the harmonics, voltage sags, and swells and improve
power quality. This chapter focuses on mitigating the short-term voltage disturbances
like the voltage sag and swell in the low voltage unbalanced distribution networks
by effectively implementing a DVR. The primary objectives of this chapter are as
follows:
• The total harmonic distortion (THD) initiated by non-linear consumer loads
should be reduced below 4% by implementing PV-DSTATCOM with maximum
power point tracking (MPPT) controller. The reweighted zero attracting (RZA)
control scheme with adaptive features should be installed at the point of the
common coupling (PCC) node where the MG is connected with the primary grid
to accomplish developed PQ performance in the MG. The PV-DSTATCOM also
transforms the DC voltage into AC.
• The enhancement of the PQ by suppressing the harmonic distortions is made using
a hysteresis controller.
• To increase the efficiency of PV cells, MPPT is implemented.
• Dynamic voltage restorer is used to mitigating voltage swells and voltage sags
required by the customer and the reduced harmonics.
A Novel Approach for Power Quality Improvement in Microgrid 317

The rest part of this chapter is arranged as follows. Section 2 presents power
quality improvement using a PV-DSTATCOM into the MG to eliminate the system’s
harmonics. Section 3 proposes an approach to eliminate any nonlinearities and miti-
gate voltage sag and swell with DVR for obtaining good performance of the system.
Section 4 presents simulation results and discussion. Conclusions are drawn in Sect. 5
by comparing proposed approaches and drawing the contributions.

2 Power Quality Improvement by Developing

PV-DSTATCOM

The RERs, such as solar PV, are manifested from free energy resources freely obtain-
able from solar radiation and have gained widespread acceptance amongst electricity
consumers. The single-stage algorithm or topology is preferred as a DC-DC boost
converter requirement is eradicated, reducing budget and improving the efficiency
of the concerned system. The concept of PV-distribution static compensator (PV-
DSTATCOM), which is used to transform the DC output voltage from the solar
cell to equivalent AC, is regarded as a converter with a solar photovoltaic array
with an operational amplifier based active filtering feature to improve the system
performance by improving power quality and suppressing harmonic distortions with
reactive power compensation in the unbalanced distribution system network [16].
The rising demand for energy, an increase of unwanted power outages, and high
penetration of electronic circuit-based non-linear loads with the depreciating power
infrastructure create a formidable challenge in front of researchers. A smart PV-
DSTATCOM system acting as an active filter to counterbalance the aforementioned
problems has been proposed in this work. It can nullify the effect of the harmful
effects of non-linear load currents, standardize the voltage of DC-link, and develop
the grid’s power factor. It can also achieve power flow in multidirectional nodes. The
discussed system can perform into two modes DSTATCOM and PV-DSTATCOM,
and achieve two ways of sensing the PV power automatically. In this chapter, an
adaptive reweighted zero attracting (RZA) control approach extracts the reference
from measured grid currents [17]. The traditional control strategy with the LMS
algorithm can provide an appropriate steady-state behaviour and transient response,
but it has a wide range of mean square errors. The RZA control scheme proposed in
this chapter uses a zero attractor generator in the adaptive iteration by incorporating
parallel relaxation and comprehensive sensing to achieve improved performance
of the control scheme. The RZA control provides accurate reference signals and
improves the filtering process.
318 A. R. Singh et al.

2.1 Power System Model

Let’s assume a grid equivalent model comprising of 3 phase generating source model
associated with loading in series and corresponding self-impedances [18, 19]. The
current and voltage are measured at the node defined by the PCC. THD of the
complete grid system is measured over FFT analysis. Figure 1. Depicts the simple
arrangement of a grid with non-linear loads leading to the production of harmonics.
Total distribution in both 3-phase current and voltage waveforms and the total
THD of the circuit with the help of the FFT study are shown in Figs. 2 and 3.
Harmonics are introduced due to nonlinear consumer loads and generating units,
which try to distort the current and voltage waveforms, as shown in Figs. 2 and 3.
These unwanted harmonics can be reduced by installing a smart and flexible branch
at PCC [20]. The harmonic load is only connected to the B phase; hence R and Y
phases are pure sinusoidal, but the B phase is distorted. In Fig. 4, introduced THD due
to non-linear load and generating units that distort current and voltage waveforms is
portrayed. Hence, THD found here by utilizing the FFT analysis was 58.10%.

Fig. 1 The simple arrangement of a grid with non-linear loads leading to the production of
harmonics
A Novel Approach for Power Quality Improvement in Microgrid 319

Fig. 2 Distortion in a current waveform due to 5th and 11th order harmonics existing in the R
phase

Fig. 3 Distortion in voltage waveform due to 5th and 11th order harmonics existing in the R phase

Fig. 4 THD without smart branch and controller

2.2 Grid Design and Its THD with PV-DSTATCOM

A smart PV-DSTATCOM system with an active filtering scheme that can nullify
the unwanted non-linear load currents with power factor compensation at PCC and
DC voltage connection regulation has been explained and proposed in this chapter
with necessary illustrations [21]. The proposed system can automatically perform
320 A. R. Singh et al.

two types of PV power sensing modes: the DSTATCOM mode and PV-DSTATCOM
mode. An adaptive RZA scheme of control has been used in this work to iden-
tify the required grid reference current. A precise value of current and voltage can
extract maximum power from a PV array with constant insolation; hence, the MPPT
approaches have been presented, such as adaptive perturb and observe technique
[22, 23]. Let’s consider a conventional grid containing 3 phase generating sources
coupled to a linear or nonlinear load connected in series with their corresponding
self-impedances. Currents and voltages are measured at the traditional PCC. The
harmonic distortion term defined by THD of the whole grid is measured during
FFT analysis. Schematic diagram of proposed PV-DSTATCOM scheme topology
has been depicted in Fig. 5.
Parameters utilized for the simulation of PV-DSTATCOM installed in the grid are
presented in Table 1. It has been estimated that by inserting a suitable voltage at the
sampling time period of (TS ), the current at PCC can be made harmonic-free. An
adaptive RZA control technique was implemented to extract grid currents in the form
of reference [24]. There present one particular significance of current and voltage at
which extreme power can be found from PV array; hence the MPPT practices have
been testified, such as adaptive P&O (perturb and observe) procedure.

Fig. 5 Schematic diagram of proposed PV-DSTATCOM scheme topology

A Novel Approach for Power Quality Improvement in Microgrid 321

Table 1 Parameters used for

Parameters Value
the simulation of PV
DSTATCOM installed in the OC voltage of PV simulator 394.80 V
grid array (VOC N )
SC current array (I SC N ) 16.62A
Array power (PM P P ) 5.35 kW
DC bus voltage (V Re f ) 380 V
Interfacing inductor (L f ) 2.9mH
DC bus capacitor (Cdc ) 2350uF
Sampling time (TS ) 30us
Grid voltage (VL L ) 220Vrms
Ripple filter (R f and C f ) 5,10uF
DC PI controller (K pd and 0.1,0.01
K id )
Non-linear load 1.1 kW load with a 3-phase
diode bridge with
μ 0.054
ε 10.2
ρ 0.004

2.3 System Topology and Control Architecture

The suggested scheme comprises 5350 W solar PV array connected to the grid’s 3-
phase AC system to feed linear and non-linear load up through ripple filters [25]. The
constraints of the default solar PV array at nominal working environments have been
utilized in this work. The flowchart which depicts the operation of PV-DSTATCOM
is presented in Fig. 6.
The voltage of 2-lines (vab and vbc ) that are sensed at PCC are applied to determine
phase to ground voltages on PCC [26],

va = (2vab + vbc )/3 (1)

vb = ( − vab + vbc )/3 (2)

The unit of in-phase and templates of quadrature is calculated as [27],

u pa = va /vt (3)

u pb = vb /vt (4)

u pc = vc /vt (5)
322 A. R. Singh et al.

Fig. 6 Flowchart depicting the process of PV-DSTATCOM scheme

√ √
u qa = ( − u pb / 3) + ( − u pc / 3) (6)

√  √ 
u qb = ( 3u pa /2)+ (u pb − u pc )/2 3 (7)

√  √ 
u qc = ( 3u pa /2)+ (u pb − u pc )/2 3 (8)

The error of DC voltage (Vde ) is determined as the variance among the reference
∗
voltage of DC bus (Vdc ) achieved from the perturb and observe (P&O) MPPT and DC
voltage that is sensed (Vdc ). P&O’s hill-climbing method works on the perturbation
of solar output voltage and observing power sign, and consequently chooses the
condition for the later perturbation to reach analogical hill-top or the extreme practical
PowerPoint [28]. DC voltage fault at r th sampling instantaneous is expressed as,
∗
Vde (r ) = Vdc − Vde (9)

The component of active power loss (wcp ) which is adaptable DC bus voltage is
approximate by using Vde and a Proportional Integral (PI) controller gains as,

wcp (r + 1) = wcp (r ) + K pd [Vde (r + 1) − Vde (r )] + Kid Vde (r + 1) (10)

A Novel Approach for Power Quality Improvement in Microgrid 323

where K pd and K id are proportional and integral gains, G p (s) is a transfer function
of the concerned plant and G c (s) is a transfer function for the concerned PI voltage
controller. Similarly, to regulate the PCCvoltages observed voltage error (Vte (n)) is
fed to another controller of PI nature [29]. Calculated fault and failure component of
reactive power (wcq ) at r th the sampling period is evaluated as,

Vte (r ) = Vt∗ − Vt (11)

wcq (r + 1) = wcq (r ) + K pt [Vte (r + 1) − Vte (r )] + Kit Vte (r + 1) (12)

where K pt and K it Are proportional and integral gains used in this controller, which
is having PI characteristics. For an upgraded error performance as the solar irradi-
ance stage varies, the feed-forwards term of PV from the solar photovoltaic array is
calculated by using,

w pv (r ) = 2 p pv (r )/3vt (13)

where Ppv is obtained PV power. Weight of the fundamental active element of the
load current of phase ‘a’ under steady-state and dynamic conditions is calculated by
using [30]

ρsgn(wpa(r ))
wpa(r + 1) = wpa(r ) − + μepa(r )upa (14)
1 + ε|wpa(r )|

where,

e pa (r ) =i La (r )u pa (r )w pa (r ) (15)

e pa is a load component of the concerned active error, ρ is an invariable constant to

switch the strength of zero attractors, K a is the reweighted term for the zero attractors,
u is the factor of convergence,sgn is a signum function, s is an independent constant,
w pa (r ), i La (r ) and u pa (r ) are the active reference component of the weight, load
current, and in-phase unit template of phase ‘a’ at r th instant [31].
Similarly, weights of primary effective terms (w pb (r ), w pc (r )) of non-linear
load currents and adaptive instrument faults of supplementary 2 phases ‘b’ and ‘c’
with steady-state and dynamic situations (when load unbalancing is undertaken) are
conveyed as,

ρsgn(wpb(r ))
wpb(r + 1) = wpb(r ) − + μepb(r )upb (16)
1 + ε|wpb(r )|

where,
324 A. R. Singh et al.

e pb (r ) =i Lb (r )u pb (r )w pb (r ) (17)

ρsgn(wpc(r ))
wpc(r + 1) = wpc(r ) − + μepc(r )upc (18)
1 + ε|wpc(r )|

where,

e pc (r ) =i Lc (r )u pc (r )w pc (r ) (19)

Similarly, the weight of primary reactive share of the nonlinear current drawl by
phase ‘a’ under steady-state and dynamic situations is represented by [32],

ρsgn(wqa(r ))
wqa(r + 1) = wqa(r ) − + μeqa(r )uqa (20)
1 + ε|wqa(r )|

where the reactive error eqa (r) is expressed as,

eqa (r ) =i La (r ) − u qa (r )wqa (r ) (21)

ρsgn(wqb(r ))
wqb(r + 1) = wqb(r ) − + μeqb(r )uqb (22)
1 + ε|wqb(r )|

where

eqb (r ) =i Lb (r ) − u qb (r )wqb (r ) (23)

ρsgn(wqc(r ))
wqc(r + 1) = wqc(r ) − + μeqc(r )uqc (24)
1 + ε|wqc(r )|

eqc (r ) =i Lc (r ) − u qc (r )wqc (r ) (25)

The active weight component (wsp ) of reference grid currents under dynamic and
steady-state situations is represented by,

wsp =w L pa +wcp − w pv (26)

The reference active grid currents under dynamic and steady-state conditions are
calculated by using,

i ∗pa = wsp u pa , i ∗pb = wsp u pb , i ∗pc = wsp u pc (27)

Similarly, the total reactive weight component (wsq ) of reference grid currents
under steady-state and dynamic conditions is represented by,
A Novel Approach for Power Quality Improvement in Microgrid 325

wsq = − w Lqa +wcq (28)

where,

w Lqa = (wqa +wqb + wqc )/3 (29)

The reference reactive grid currents under dynamic and steady-state situations are
calculated by using,
∗ ∗ ∗
i qa = wsq u qa , i qb = wsq u qb , i qc = wsq u pc (30)

Therefore, the total reference 3-phase grid currents under dynamic and steady-
state situations are calculated by using,
∗
i qa = i ∗pa + i qa
∗ ∗
, i sa = i ∗pa + i qb
∗ ∗
, i sc = i ∗pc + i qc
∗
(31)

The difference between the reference grid and sensed current output is delivered
through the hysteresis regulator to produce a pulse signal for the regulated switching
by gate pulses of concerned VSC [13].

3 Power Quality Improvement by Developing Dynamic

Voltage Restorer (DVR)

The nonstandard behaviour of concerned power system parameters like the current,
voltage, and frequency that may damage the end-use equipment is regarded as the
power quality disturbance [15]. Short-term power fluctuations like the power system
swell or sag are some of the substantial concerns that occur frequently. A power
quality enhancement device for the unbalanced distribution system known as the
DVR is implemented effectively in this chapter to mitigate these short-term voltage
fluctuations like voltage swell and voltage sags. DVR can deliver the optimum cost-
effective solution to minimize the short-term voltage fluctuations such as voltage
swells and sags required by end-use customers [16]. A DVR can be a resourceful
solution to the required power quality problems with enormous rapid and flexible
features.

3.1 Circuit Model of DVR

In an unbalanced distribution network, the solid-state (or static) power electronic

component-based DVR can also be referred to as a static series compensator (SSC) or
series voltage booster (SVB), which is preferably coupled in series with the concerned
326 A. R. Singh et al.

primary distribution system with necessary arrangement [17]. The DVR refurbishes
the load side voltage to a preferred pre-sag voltage amount by providing a series
voltage of the preferred amplitude and frequency even when the source voltage is
not balanced [18].
The DVR is usually connected with the primary unbalanced distribution system
with the help of a series A voltage injection with the incoming 3 phase network. This
helps to enhance the power quality of the disturbed system by necessary adjustment
of the PCC voltage magnitude, phase shift, or wave shape. The short-term voltage
sag’s compensation can be achieved, and reactive and active power addition into an
unbalanced power distribution network. Although the reactive component of power
cannot be generated from any source, it is compensated by the help of a VSI inside
DVR. This can protect the sensitive loads or end-user equipment in the parallel
feeder with the necessary breaker action by disconnecting faulted feeder and limiting
the fault current with necessary parameter modification [33]. DVR also improves
harmonics present in the line voltage along with the mitigation of voltage swell and
sag. Figure 7 depicts the primary DVR circuit and its components. A DVR coupled
to the grid with a non-linear load is shown in this figure.
DVR comprises the energy storage unit, VSI, series injection transformer, filter,
control unit, and bypass switch/circuit breaker. The energy storage section’s func-
tion is to cater to VSI, which feeds to a series injection transformer. The VSI is a
combination of power electronic devices that can switch the battery source’s power
with the necessary frequency. This works as a source of generating 3-phase voltage
with the required magnitude phase and frequency to compensate for the unbalanced
value of the concerned load voltage [19]. The transformer presented in Fig. 7 injects
the generated voltage from VSI in sequence with the unstable distribution line. The
HV part is coupled to a network, while the LV side is connected to VSI. The filter
and control unit is used to sense the fluctuations related to power quality and filter
the harmonics in the output of VSI [20]. Bypass switch/circuit breaker is used for
isolation of DVR from the grid.

Fig. 7 Basic circuit model of DVR

A Novel Approach for Power Quality Improvement in Microgrid 327

Fig. 8 Equivalent single line diagram of DVR

3.2 Proposed Technique

The DVR re-establishes the load side voltage to a preferred pre-sag voltage amount
by delivering a voltage of the preferred frequency and magnitude even if it has an
unbalanced source voltage [34, 35]. Following proposed objectives achieve this:
• A dc source is converted to a harmonic free ac source with AVR help connected
with filter.
• A Series injection transformer is implemented to alleviate the voltage sag by
inserting voltage at PCC.
Bypass switch further enhances the convenience of implementing the circuit. The
basic diagram of a grid-connected DVR is depicted in Fig. 8, making it very conve-
nient to understand its operation principle by considering a 3-phase fault. Hence, a
DVR coupled to the grid with a non-linear load is depicted in Fig. 7.

4 Results and Discussion

In this portion, the proposed PV-DSTATCOM and DVR are implemented to improve
the system’s power quality. The simulation results of these two approaches are
presented next:

4.1 Improvement of Power Quality by Developing

PV-DSTATCOM

Figure 9 shows the three-phase waveform obtained as the output of the simulating
PV-DSTATCOM model. Also, the final THD found by applying the FFT analysis on
illustrated in Fig. 10.
328 A. R. Singh et al.

Fig. 9 The 3 phase waveform was obtained by simulating the PV DSTATCOM model

Fig. 10 The final THD was obtained by using the FFT analysis

From Figs. 9 and 10, it has been observed that the waveform is almost sinusoidal
after incorporating the PV-DSTATCOM in the model, and the THD has also reduced
to 6.64%, which was previously 48%. Further, the use of automated solar PV helps
in uninterrupted supply. As a result of the simulation, the harmonics caused due to
consumer loads which are non-linear and devices remain practically mitigated. THD
was condensed from 58.7% instead of PV-DSTATCOM to 5% with PV-DSTATCOM.
By increasing the generation voltage, the waveform gets smoother further.
A Novel Approach for Power Quality Improvement in Microgrid 329

4.2 Improvement of Power Quality by Developing DVR

The voltage at PCC due to 3 phase fault and the injected voltage and finally the
recovered sag are depicted in Fig. 11. Figure 12 illustrates the THD during the
3-phase sudden short-circuit.
Now the proposed DVR is implemented to improve the PQ of the concerned
system. Figure 13 depicts the THD during the 3-phase sudden short-circuit after
using the proposed DVR. From this figure, it can be observed that the THD has
also decreased by 8%, along with the waveform’s improvement. The voltage sag is
mitigated and restored to the normal value. Hence in this way, it can be seen that
the 3-phase short-circuit fault can be overcome without causing equipment failure or
interrupting power supply and is one of the most convenient methods for low voltage
applications.

Fig. 11 Voltage sag injected voltage and recovered voltage waveforms

Fig. 12 THD during 3-phase sudden short circuit

330 A. R. Singh et al.

Fig. 13 THD during 3-phase sudden short-circuits after using DVR

DVR is used in various applications as it is less expensive. Compared to

DSTATCOM and UPS, DVR is a lot cheaper, and the UPS requires a lot of main-
tenance due to battery leakage and replacement problems. DVR is tiny and is a
better power-effective device than UPS, Superconducting Magnetic Energy Storage
(SMES), and DSTATCOM. DVR also compensates for line voltage harmonics along
with mitigation of voltage sag and swell.

5 Conclusions

This chapter presented a new power quality improvement approaches in microgrids

(MGs) using a grid-connected smart photovoltaic distribution static compensator
(PV-DSTATCOM) system implementing an adaptive RZA control technique through
perturbing and observe (P&O) MPPT approach and the use of DVR for a three-phase
system. These two techniques are verified by performing various recreations. The
advantages of PV-DSTATCOM with MPPT and RZA controller were analyzed with
sensitivity, and the DVR can be practicality taken into consideration. The proposed
technique removes the necessity for non-linear current contribution amongst inverters
by not negotiating voltage quality. The THD diminishes expressively by using these
efficient techniques. Without using any of these methods, the THD was perceived to
be roughly 58% in the output voltage, but after implementing the proposed strategies,
the THD has reduced to about 21%. Thus, minimizing the disturbances in practice.
Secondly, the IGBT-based controllers were introduced by PV-DSTATCOM, which
includes an MPPT controller and an RZA controller. This resulted in a reduced THD
of about (4 – 6) %. Later, DVR use in the smart grid was analyzed to mitigate voltage
sag due to power quality disturbances and improved harmonics. The approaches
proposed in this chapter provides the system with enough prediction ability and
smartness for holding the required current and voltage quality at the PCC, and high
power quality is achieved with significantly fewer expenses.
A Novel Approach for Power Quality Improvement in Microgrid 331

Acknowledgements This research work was supported by “Woosong University’s Academic

Research Funding-2021”.

References

1. Yazdi F, Hosseinian SH (2019) A novel “Smart Branch” for power quality improvement in
microgrids. Int J Electr Power Energy Syst 110:161–170. https://doi.org/10.1016/j.ijepes.2019.
02.026
2. Lavanya V, Kumar NS (2018) Control strategies for power quality improvement in microgrid.
Int J Renew Energy Res 8(1):151–154
3. Zhong QC (2013) Harmonic droop controller to reduce the voltage harmonics of inverters.
IEEE Trans Ind Electron 60(3):936–945. https://doi.org/10.1109/TIE.2012.2189542
4. Wang R, Hsu SC, Zheng S, Chen JH, Li XI (2020) Renewable energy microgrids: Economic
evaluation and decision making for government policies to contribute to affordable and clean
energy. Appl Energy 274. https://doi.org/10.1016/j.apenergy.2020.115287
5. Singh B, Kandpal M, Hussain I (2018) Control of grid-tied smart PV-DSTATCOM system
using an adaptive technique. IEEE Trans Smart Grid 9(5):3986–3993. https://doi.org/10.1109/
TSG.2016.2645600
6. Veronica AJ, Kumar NS (2017) Development of hybrid microgrid model for frequency
stabilization. Wind Eng 41(5):343–352. https://doi.org/10.1177/0309524X17723203
7. Bacha S, Picault D, Burger B, Otadui IE, Martins J (2015) Photovoltaics in microgrids: an
overview of grid integration and energy management aspects. IEEE Ind Electron Mag 9(1):33–
46. https://doi.org/10.1109/MIE.2014.2366499
8. Chitra N, Kumar AS, Priyadharshini P, Prabaakaran K (2014) Survey on microgrid: power
quality improvement techniques. ISRN Renew Energy 2014:1–7. https://doi.org/10.1155/2014/
342019
9. Kaushal J, Basak P (2018) A novel approach for determination of power quality monitoring
index of an AC microgrid using fuzzy inference system. Iran J Sci Technol Trans Electr Eng
42:429–450. https://doi.org/10.1007/S40998-018-0087-Z
10. Hashempour MM, Lee T, Savaghebi M, Guerrero JM (2019) Real-time supervisory control for
power quality improvement of multi-area microgrids. IEEE Syst J 13(1):864–874. https://doi.
org/10.1109/JSYST.2018.2823899
11. Sahoo B, Routray SK, Rout PK (2020) Application of mathematical morphology for power
quality improvement in microgrid. Int Trans Electr Energy Syst 30(5). https://doi.org/10.1002/
2050-7038.12329
12. Eldin AAH, Mansour AA, Gamal ME, Youssef KH (2019) Power quality improvement of
smart microgrids using EMS-based fuzzy controlled UPQC. Turk J Electr Eng Comput Sci
27:1181–1197. https://doi.org/10.3906/elk-1807-166
13. Qian H, Xu Q, Zhao J, Yuan X (2020) A robust GPS-based control scheme for power-sharing
and quality improvement in microgrid. Int J Electr Power Energy Syst 123. https://doi.org/10.
1016/j.ijepes.2020.106324
14. Safa A, Berkouk ELM, Messlem Y, Gouichiche A (2018) A robust control algorithm for a
multifunctional grid tied inverter to enhance the power quality of a microgrid under unbalanced
conditions. Int J Electr Power Energy Syst 100:253–264. https://doi.org/10.1016/j.ijepes.2018.
02.042
15. Thomas D, Hoop GD, Deblecker O, Genikomsakis KN, Ioakimidis CS (2020) An integrated
tool for optimal energy scheduling and power quality improvement of a microgrid under
multiple demand response schemes. Appl Energy 260. https://doi.org/10.1016/j.apenergy.2019.
114314
332 A. R. Singh et al.

16. Kumar YVP, Bhimasingu R (2017) Electrical machines based DC/AC energy conversion
schemes for the improvement of power quality and resiliency in renewable energy microgrids.
Int J Electr Power Energy Syst 90:10–26. https://doi.org/10.1016/j.ijepes.2017.01.015
17. Kumar YVP, Ravikumar B (2016) A simple modular multilevel inverter topology for the power
quality improvement in renewable energy based green building microgrids. Electr Power Syst
Res 140:147–161. https://doi.org/10.1016/j.epsr.2016.06.027
18. Srivatchan NS, Rangarajan P, Rajalakshmi S (2015) Control scheme for power quality improve-
ment in islanded microgrid operation. Procedia Technol 21:212–215. https://doi.org/10.1016/
j.protcy.2015.10.090
19. Igual R, Medrano C (2020) Research challenges in real-time classification of power quality
disturbances applicable to microgrids: a systematic review. Renew Sustain Energy Rev 132.
https://doi.org/10.1016/j.rser.2020.110050
20. Shafiullah GM, Amanullah MT, Ali ABMS, Wolfs P, Arif MT (2013) Smart grids: opportunities,
developments and trends. Springer, London. https://doi.org/10.1007/978-1-4471-5210-1
21. Steffel SJ, Caroselli PR, Dinkel AM, Liu JQ, Sackey RN, Vadhar NR (2012) Integrating solar
generation on the electric distribution grid. IEEE Trans Smart Grid 3(2):878–886. https://doi.
org/10.1109/TSG.2012.2191985
22. Wang T, O’Neill D, Kamath H (2015) Dynamic control and optimization of distributed energy
resources in a microgrid. IEEE Trans Smart Grid 6(6):2884–2894. https://doi.org/10.1109/
TSG.2015.2430286
23. Jager K, Isabella O, Smets AHM, Van Swaaij RACMM, Zeman M (2014) Solar energy
fundamentals, technology and systems. UIT Cambridge, Cambridge
24. Wu TF, Chang CH, Lin LC, Kuo CL (2011) Power loss comparison of single- and two-stage
grid-connected photovoltaic systems. IEEE Trans Energy Convers 26(2):707–715. https://doi.
org/10.1109/TEC.2011.2123897
25. Das SR, Mishra DP, Ray PK, Salkuti SR, Sahoo AK (2021) Power quality improvement using
fuzzy logic-based compensation in a hybrid power system. Int J Power Electron Drive Syst
12(1):576–584. https://doi.org/10.11591/ijpeds.v12.i1.pp576-584
26. Agarwal RK, Hussain I, Singh B (2016) LMF based control algorithm for single-stage three-
phase grid integrated solar PV system. IEEE Trans Sustain Energy 7(4):1379–1387. https://
doi.org/10.1109/TSTE.2016.2553181
27. Subudhi B, Pradhan R (2013) A comparative study on maximum power point tracking tech-
niques for photovoltaic power systems. IEEE Trans Sustain Energy 4(1):89–98. https://doi.org/
10.1109/TSTE.2012.2202294
28. Martinez JA, Arnedo JM (2006) Voltage sag studies in distribution networks—part I: system
modelling. IEEE Trans Power Delivery 21(3):1670–1678. https://doi.org/10.1109/TPWRD.
2006.874113
29. Das B, Panigrahi PK, Das SR, Mishra DP, Salkuti SR (2021) Power quality improvement in a
photovoltaic based microgrid integrated network using multilevel inverter. Int J Emerg Electr
Power Syst. https://doi.org/10.1515/ijeeps-2021-0040
30. Choi SS, Li BH, Vilathgamuwa DM (2011) Dynamic voltage restoration with minimum energy
injection. IEEE Trans Power Syst 15(1):51–57. https://doi.org/10.1109/59.852100
31. Vilathgamuwa D, Wijekoon HM, Choi SS (2012) A novel technique to compensate voltage
sags in multiline distribution system the interline dynamic voltage restorer. IEEE Trans Ind
Electron 53(5):1603–1611. https://doi.org/10.1109/TIE.2006.882017
32. Ray P, Salkuti SR (2020) Smart branch and droop controller based power quality improvement
in microgrids. Int J Emerg Electr Power Syst 21(6). https://doi.org/10.1515/ijeeps-2020-0094
33. Marei MI, EI-Saadany EF, Salama MMA (2012) A new approach to control DVR based on
symmetrical components estimation. IEEE Trans Power Deliv 22(4):2017–2024. https://doi.
org/10.1109/TPWRD.2007.905537
34. Mathur RM, Varma RK (2013) Thyristor-based FACTS controllers for electrical transmission
systems. Wiley, New York
35. Kaushal J, Basak P (2020) Power quality control based on voltage sag/swell, unbalancing,
frequency, THD and power factor using artificial neural network in PV integrated AC microgrid.
Sustain Energy Grids Netw 23. https://doi.org/10.1016/j.segan.2020.100365
Characterization of Bifacial Passivated
Emitter and Rear Contact Solar Cell

Suresh Kumar Tummala, Phaneendra Babu Bobba,

and Satyanarayana Kosaraju

Abstract A Passivated Emitter and Rear Contact Cell (PERC) defines a new archi-
tecture of solar cell which differs from standard cell architecture. PERC enhances
light capture at the rear surface and to optimize electron capture. This leads to an
increase (or) achieve better efficiency than when compared with standard cell which
are reaching their physical limits. In this paper, the characterization of bifacial p-type
PERC solar cell with various proportions of tallness and width, back Silicon Nitrate
layer with various thickness are streamlined. Most elevated back productivity of the
bifacial solar cell (~22%) was obtained at AM 1.5 (air mass), 1000 W/m2 , 26 °C
standard test condition.

Keywords PERC · Bifacial · Finger · Generation profile · Reflection · Absorbtion

Nomenclature

PERC Passivated Emitter Rear Contact

LCOE Levelized Cost of Energy
c-Si Crystalline Silicon
PV Photovoltaic
EQE External Quantum Efficiency
PMP Maximum Power
JSC Short Circuit Current
VOC Open Circuit Voltage
RS Series Resistance
RSH Shunt Resistance

S. K. Tummala (B) · P. B. Bobba

Electrical and Electronics Engineering Department, Gokaraju Rangaraju Institute of Engineering
and Technology, Hyderabad, India
S. Kosaraju
Mechanical Engineering Department, Gokaraju Rangaraju Institute of Engineering and
Technology, Hyderabad, India

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 333
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6_14
334 S. K. Tummala et al.

JL Current density generated inside the solar cell

Jph (λ) Incident Photon Current
A(λ) Fraction of the incident light
η(λ) Collection Efficiency
Tnom Nominal Temperature
T Operating Temperature
Eg Band gap of the semiconductor
q Charge on electron
k Boltzmann’s constant
NF Number of Finger
WF Width of Finger
PF Finger Pitch
SF Finger Spacing

1 Introduction

The mission for grid equality is the steady mission of the photovoltaic business
and lessening the expense of power with high effective advances is a basic method-
ology for the business to achieve this mission. As PV modules are the center part
of a PV framework, the consistent improvement of module productivity is the most
immediate and great innovation course to accomplish lattice equality. Among high-
effectiveness cell innovations [1], PERC (passivated emitter and rear cells) is without
a doubt the most cost-effective. Contrasted and regular cell, the creation of PERC
cell needs just to add back passivation [2] and laser opening areas, combined with
metallization methods. This will adequately improve the change effectiveness of
the cells [3]. In the previous a half year, the world record for PERC cell change
productivity has been continually re-energized, exhibiting the solid likely potential
gain of PERC. The exhibition of mechanical kind screen-printed 6-inch PERC sun
powered cells has been fundamentally expanded during recent years. Crystalline
silicon (c-Si) is the crystalline form of silicon, either polycrystalline silicon (poly-
Si, consisting of small crystals) or monocrystalline silicon (mono-Si, a continuous
crystal). Crystalline silicon is the dominant semiconducting material used in photo-
voltaic technology to produce solar cells. These cells are assembled into solar panels
as part of a photovoltaic system to generate solar power from sunlight. The main
comparison of the p-type PERC and standard PV cell are:
Standard PV cells are typically Crystalline Silicon cells with a high energy trans-
formation effectiveness rate. This implies that the cells, framed into boards, can
produce more power even though the boards are sufficiently enormous to find a way
into restricted spaces like roofs, making them space-proficient and over the long haul,
cost-effective. Whereas, PERC cell innovation is more direct as they just contrast
marginally from the standard PV cells. It is likewise savvier as it permits energy
yield to be amplified because of more productive energy transformation.
Characterization of Bifacial Passivated Emitter and Rear … 335

Bifacial PERC is additionally turning into a significant course in the advancement

of the innovation [4]. The usually utilized assembling innovation for bifacial PERC
cell is to change the printing cycle of the PERC single-sided cell and changing the
back surface from entire aluminum layer to a neighborhood aluminum layer. This
permits accidental light on the back surface to enter the cell from the uncovered
layer [5, 6] and make photoelectric change on both the front and back sides. with
an irrelevant expansion in assembling cost, bifacial PERC can accomplish a force
age gain of 10%-25% at the framework level, which will essentially diminish the
Levelized Cost of Energy [7] of the PV framework while enormously improving the
intensity and advancement capability of PERC innovation.
Bifacial silicon cell is an exceptional sort of gem silicon solar cell structure, which
can get daylight and produce electrical force from both the front and posteriors. These
cells are planned with uncommon structure and finger contact, so they can ingest
more daylight and produce a lot of intensity from the two sides. The effectiveness
gains [8] of bifacial cells comparative with single-sided cells are hard to gauge
by a solitary cell. Along these lines, much of the time, they should be made into
modules. Bifacial modules can be introduced deftly [9, 10] in a wide scope of uses.
The type of customary single-sided cell modules establishment applications, for
example, ground photovoltaic force stations and rooftop photovoltaic frameworks,
is similarly reasonable for bifacial modules. The bifacial modules are particularly
useful for ground establishment since they exploit the ground’s mirrored light [11],
produce greater power under similar conditions, and are additionally appropriate for
establishment on the rooftop. Ground and rooftop-mounted bifacial modules should
be mounted inclined on the scene. The bifacial cell and module can likewise be
liberated [12, 13] from the course of the establishment. The back is as wonderful
as the front, particularly appropriate for the vertical establishment of the scene, for
example, vertical establishment as a fence, soundproof divider, thruway guardrail
[14], and so on Thusly, the bifacial modules are reasonable for establishment in
different situations, for example, ground power station, uniaxial global positioning
framework, water surface force station, daylight room, photovoltaic shed, thruway
soundproof divider, drape divider, vehicle shed, fabricating coordinated PV, etc. In
the interim, in some ground power plant or housetop power plant [15], bifacial cells
and modules additionally were introduced and have been run consistently for quite
a while.
To additionally understand the benefits of high force transformation of high-
effectiveness cells, joining PERC cells with a creative module [16, 17] development
procedure is a significant overhaul course for makers of PERC cells and modules.
From an innovation viewpoint to improve module yield power, the half-cut cell
is without a doubt an elite and ease item that is anything but difficult to execute
in enormous scope creation [18]. The half-cut cell is the procedure that slices a
traditional cell down the middle and afterward interfacing the two parts together
[19]. Contrasted the regular cells, the lone activity required is to cut the entire cell
into two sections by a laser prior to string welding them. The PV cell parting and
transmission are completely programmed [20] and the string welder needs just slight
alteration to accomplish enormous scope creation.
336 S. K. Tummala et al.

1.1 Methodology

The principle preferred position of utilizing bifacial solar modules to construct photo-
voltaic plants is high force age. In the most part, the bifacial cells are made into bifa-
cial modules. At the point when the direction, tendency, and stature of the bifacial
modules are fixed, the force age gain of the bifacial modules predominantly returns
from the side, which improves the current and greatest yield intensity of the get
together by accepting the environmental dissipated light and the mirrored light of
the ground. Moreover, the force age gain of the back can be improved by raising the
stature of the part and expanding the tendency point of the segment. In this chapter,
characterization of bifacial PERC solar cell is presented with a focus on Reflection,
External Quantum Efficiency, C-V measurements, Generation profiles, and photon
currents.

1.2 Bifacial PERC

Presently, the worldwide market is overwhelmed by mono-facial PV cells and

modules. In any case, the International Technology Roadmap for Photovoltaics 2019
predicts that bifacial cells will pick up 60% of the worldwide market in 10 years, and
they will be utilized in both bifacial and mono-facial modules. This is fundamentally
because of the assumption that bifacial cells would create more force. Nonetheless,
genuine bifacial modules with bifacial cells and straightforward back covers are
required to make up about half of worldwide piece of the overall industry by 2029.
Bifacial modules cannot be rated the same way as mono-facial modules; therefore,
further discussion into standard testing conditions are required. Nevertheless, the
possible gain from this technology attracts attention from the photovoltaic market.
The electrical outputs of the solar cell or module are calculated by solving an
equivalent-circuit model. The solution to that model shown [21] in Fig. 1 is an
IV curve (i.e., the relationship between current and voltage), from which standard
electrical outputs [22] like maximum power PMP , short-circuit current JSC , and open-
circuit voltage VOC are determined. The inputs are defined there are two possible
exceptions: the series resistance Rs (optional) and the light-generated current JL .
JL represents the current density generated inside the solar cell and collected by
the p–n junction. It’s called the light-generated current density.
JL is determined from the equation:

∞
JL = Jph (λ)A(λ)η(λ)dλ (1)
0
Characterization of Bifacial Passivated Emitter and Rear … 337

Fig. 1 Electrical equivalent circuit of bifacial cell

where Jph (λ) is the incident photon current, A(λ) is the fraction of the incident light
that becomes absorbed by the active region of the solar cell, and η(λ) is the collection
efficiency within that active region [23].
Inputs for equivalent-circuit models are commonly quoted at ‘nominal’ tempera-
tures such as 25 °C or 300 K. When installed in the field, however, solar cells tend
to operate at higher temperatures. The option of loading circuit inputs at a nominal
temperature Tnom , setting the operating temperature T, and allowing SunSolve [24]
to calculate the circuit inputs at T using the same approach as PV Syst, which applies
these equations:

JL = JLnom [1 + μIL (T − Tnom )] (2)

m1 = m1nom [1 + μm1 (T − Tnom )] (3)

where μIL and μm1 are coefficients for the light-generated current and ideality factor
of the primary diode m1 ; and
 3   
T qEg 1 1
J01 = J01nom exp − (4)
Tnom m1 k T Tnom

where Eg is the band gap of the semiconductor, q is the charge on an electron, and k
is Boltzmann’s constant, giving the recombination current of the primary diode J01
at temperature T.
The Monte-Carlo algorithm (a Monte Carlo algorithm is a randomized algo-
rithm whose output may be incorrect with a certain (typically small) probability)
necessarily leads to uncertainty in the outputs. These uncertainties are calculated by
dividing the number of rays traced into multiple sub-simulations, and then applying
statistical analysis to determine a mean value and a 95% confidence interval for each
338 S. K. Tummala et al.

output. If, for example, the current absorbed within the solar cell is determined to be
J cell = 36.23 ± 0.45 mA/cm2 after 100,000 rays are traced, this means that we can be
95% confident that J cell would be between 35.78 and 36.68 mA/cm2 after infinitely
many rays are traced.

1.3 Algorithm

Step 1: Several rays are created. Each ray is assigned a wavelength, an intensity, a
direction, and a location.
Step 2: Each light beam continues along a straight line until it converges with an
aspect of the module’s surface.
Step 3: At this interaction, and at all future interactions with an interface: reflectance,
transmittance, and absorptance are calculated; the intensity of the ray is reduced by
the value of the absorptance; the magnitudes of reflectance and transmittance are
translated into probabilities.
Step 4: If the ray passes through an absorbing layer, its intensity is reduced by
applying Beer’s law.
Step 5: Steps 2–4 is repeated for each ray until the ray is lost from the module,
the ray’s intensity decreases below a threshold and the ray has intersected with the
maximum allowable number of interfaces.
Step 6: The gains and losses are recorded for each ray and summed and averaged to
give the results of the ray packet.
Step 7: Steps 1–6 is repeated for as many ray packets that are required for the total
number of rays to equal the value entered.

2 Characterization of PERC

Passive layer of PERC will be acting as reflector for the unabsorbed light back to
the cell for second absorption leading to increase the overall cell efficiency. Figure 2
shown is a typical PERC cell considered for the analysis which is having 4 rectangular
interconnection, 5 busbars and 120 fingers on both the sides.
The geometry of the cell has the configuration of grid contact layout both in
front and rear sides. Finger orientation is along X and bus bar orientation is along Y.
Number of fingers (NF ) on both the sides is 120 with round rectangular cross section.
The height (H) of the fingers is 15 μm and width of each finger (WF ) is 45 μm on both
sides of the cell. Finger pitch (PF ) and finger spacing (SF ) are considered as 0.1306 cm
and 0.1261 cm respectively. Similarly, bus bars are considered as 5 numbers (NB with
Characterization of Bifacial Passivated Emitter and Rear … 339

Fig. 2 c-Si bifacial PERC Cell

rectangular cross section having 15 μm height (H) and 1100 μm width (WB ). To
calculate RS , front and rear electrodes of the cell are designed with Ag, Dupont PV
19 conductor having resistivity 2.6 * 10−6 cm (Fig. 3).
The shape of the cell is pesudosqaure with length (XC ) 15.68 cm, diameter (DC )
21 cm and cell area (AC ) 244.3 cm2 . Entire analysis is carried out using sunlight spec-
trum with AM 1.5 g [Gue95] and scaling factor 1 covered for full area. Minimum
wavelength of the spectrum is considered as 300 nm whereas the maximum wave-
length is 1200 nm with wavelength interval of 20 nm. The Grid anode speaks to the
standard H-bar plan for a cell cathode. It comprises of a bunch of fingers master-
minded at the right points to a progression of bus bars. The two fingers and bus
bars run the full degree of the substrate (for example to the edge of the wafer). This
arrangement would regularly be utilized for the front cathode of a norm or PERC
cell and for the back terminal of bi-facial cells. Simulation permits to set either the
number of fingers as well as bus bars to zero.
The Coating terminal speaks to a surface covered totally by an anode. It is
comprised of a progression of bus bars with a constant layer of metal between

Fig. 3 a Cell image—front display, b cell image—back display

340 S. K. Tummala et al.

everyone. This setup is commonly be utilized to reproduce the backside of a standard

screen print or PERC cell. Note that the coatings are applied right to the edge of the
wafer and it is preposterous to expect to characterize a hole at the edge. All setups
permit the meaning of cell interconnection strips. Strips are adjusted to a similar
pivot as the bus bars. The length of the strip is equivalent to the degree of the unit cell
for example the wafer size + any encompassing zone. Strips were applied to both
the front and back of the cell.
Solar Cell Equivalent Circuit Parameters [21]:
The cell considered has the following equivalent circuit parameters

Light-collected current (JL ) 39.36 mA/cm2

Saturation current 1 (J01 ) 0.2 pA/cm2
Ideality factor 1 (m1 ) 1
Saturation current 2 (J02 ) 10 nA/cm2
Ideality factor 2 (m2 ) 2
Shunt resistance (Rsh ) 20 kcm2
Non-grid resistance (Rng ) 0.5 cm2
Grid resistance (Rsg ) 0.09807 cm2
Series resistance (Rs ) 0.5909 cm2
Nominal temperature Tnom 26.85 °C
Operating temperature T 26.85 °C

3 Results

Generation Profile
A generation profile describes the generation rate of electron–hole pairs G as a
function of depth in the solar cell. Its units are cm−3 s−1 . G is calculated as a function
of distance from both the uppermost point of the solar cell (z), and the nearest point
on the front surface (ζ ). Figure 4 explains the difference between z and ζ. The first
figure in Fig. 4 shows how z and ζ are defined, giving the example of generation
occurring at point 1 and point 2 (inside a pyramid); the second figure plots contours
of ζ within a textured cell [24].
Additionally, is known that G is characterized regarding the cell territory. That is
the reason G(z) diminishes with diminishing z when z is not exactly the height of the
pyramids. Although the age rate may be high inside the silicon part of a pyramids,
G(z) is here characterized over the whole cell territory instead of simply the zone
contained inside the pyramids. The explanation we characterize G(z) this way is so
the necessary of G(z) and G(ζ) are the equivalent and liken to the current thickness
consumed in the cell JA,cell :
Characterization of Bifacial Passivated Emitter and Rear … 341

Fig. 4 Difference between z and ζ

Fig. 5 Generation rate versus depth

w w1
G(z)dz = G(ζ)dζ = JA,cell (5)
0 0

W Thickness of the cell including the height of the front and rear texture, and
W Thickness of the cell including the height of the rear texture but excluding the
height of the front texture.

Figures 5 and 6 shows the plot of the generation rate with respect to depth and
zeta of the cell respectively. The depth and zeta of the cell is represented in μm, and
the generation rate is defined as cm−3 s−1 .
The intensity (W/m2 ) of an illuminating source is determined by incorporating the
spectral (W/m2 /nm) intensity of the chosen range over the chosen frequency scope
and multiplying by the scaling factor. The spectrum of the chose range is thought
342 S. K. Tummala et al.

Fig. 6 Generation rate versus zeta

to be the incident power on an ar plane. Thusly must scale any custom spectra that
have estimated by instruments not perpendicular to the light source.
RAT (Reflection, Absorption and Transmission) analysis is performed using
Sunsolve simulation tool. There are two ways to deal with figuring the division
of incident rays that are scattered. The main methodology is to set a consistent dissi-
pating portion . This implies that for each beam that collaborates with the surface,
 of these will be dissipated by the picked dispersing model, and the rest of not
be dispersed. In this manner, setting  = 0 will imply that there is no dispersing,
regardless of what the dissipating model is chosen. The subsequent methodology
is to apply a scalar dispersing model. With this model, the dissipating part  isn’t
consistent yet diminishes with expanding frequency. This division additionally relies
upon the episode point, the refractive list of the materials, and whether the beam is
reflected or consumed.
For Reflection
 
−( 2π
λ σrms ∗2∗ni ∗cos θi )
2
=1−e (6)

For Transmission
 
 = 1 − e ( λ rms i
− 2π σ ∗|n ∗cos θi −nt ∗cos θt |)
2
(7)

where σ rms is the root mean square roughness of the surface.

Reflection of sunlight in the front panel, Esc in both front and rear side are
computed and plotted on a graph as shown in the Fig. 7 without considering spec-
tral intensity. Figure 8 shows the Reflection, Absorption when considering spectral
intensity.
Characterization of Bifacial Passivated Emitter and Rear … 343

Quantum efficiency is given as a function of wavelength. No sunlight is absorbed

below the bandgap, so the quantum efficiency is zero in larger wavelengths. From
Fig. 9 after 1150 nm the QE is zero. QE is not measured below 300 nm as the power
from AM1.5 g is low and the response is reduced due to front surface recombination.
The quantum efficiency for the cell considered is maintained constantly 94% (±2%)
between wavelengths of 460–970 nm, this efficiency is good for a better power

Fig. 7 Refl, Abs, Esc versus wavelength without considering spectral intensity

Fig. 8 Refl, Abs, Esc versus wavelength considering spectral intensity

344 S. K. Tummala et al.

generation in the solar cell as the number of carriers collected are more. Equivalent
current densities in solar cell are plotted in Fig. 10.
Figure 11 shows main losses versus λ, where the stacked columns are the current
densities measured in mA/cm2 (Tables 1 and 2).
Table 3 represents equivalent parameters of the cell like temperature, maximum
output power, maximum power point voltage, maximum power point current, fill
factor of the solar cell, open circuit voltage and short circuit current.

Fig. 9 External quantum efficiency in solar cell

Fig. 10 Equivalent current densities in solar cell

Characterization of Bifacial Passivated Emitter and Rear … 345

Figure 12 shows the IV graph of the cell which is producing an output power of
5.068 W with a fill factor of 0.795. Maximum Point (MP) is shown in the graph at
which the voltage generated is 563.2 mV and the current is 8.998 A. The equivalent
data of the IV plot is recorded in Table 3 and grid resistance of the cell is recorded
in Table 4.
Total amount of carriers that have been reflected from the cell with respect to
spectrum or wavelength is plotted and represented in Fig. 13. As discussed in Fig. 9,
reflection will be more below 460 nm and above 970 nm as the power is less. During
460-970 nm the reflection will be minimum. Similarly, the total reflection from the

Fig. 11 Main losses with respect to spectral intensity

Table 1 Equivalent photon

Mean 95% CI Fraction of
current densities (mA/cm2 )
(mA/cm2 ) (mA/cm2 ) J inc (%)
Incident 46.32 – 100.0
Lost to environment
Reflected front 3.901 0.03894 8.423
Escaped front 0.9209 0.01120 1.988
Escaped rear 1.712 0.01575 3.696
Absorbed in components
Cell 39.00 0.03879 84.20
absorption
Parasitic 0.7792 – 1.682
absorption
Remainder 0.006600 0.001614 0.01425
346 S. K. Tummala et al.

Table 2 Equivalent photon

Mean 95% CI Fraction of
current densities (A)
(mA/cm2 ) (mA/cm2 ) J inc (%)
Incident 11.32 – 100.0
Lost to environment
Reflected front 0.9531 0.009515 8.423
Escaped front 0.2250 0.002737 1.988
Escaped rear 0.4183 0.003847 3.696
Absorbed in components
Cell 9.528 0.009476 84.20
absorption
Parasitic 0.1904 – 1.682
absorption
Remainder 0.001612 3.944 × 10−4 0.01425

Table 3 Current voltage (IV)

Parameter Abb Units Output
data
Temperature T K 300.0
Maximum power PMP W 5.068
Maximum-power-point voltage VMP mV 563.2
Maximum-power-point current IMP A 8.998
Fill Factor FF 0.795
Open-circuit voltage VOC mV 669.1
Short-circuit current ISC A 9.528

rear side of the cell is plotted in Fig. 14. Up to 950 nm the reflection from rear side is
zero. As the spectrum increases beyond 950 nm capturing of carriers will decrease
at the rear side which is shown in the plot of Fig. 14.
Total amount of carriers that have been absorbed from the cell with respect to
spectrum or wavelength is plotted and represented in Fig. 15. As discussed in Fig. 9,
reflection will be more below 460 nm and above 970 nm as the power is less. During
460–970 nm the absorption will be maximum. Similarly, the total absorption from
the rear side of the cell is plotted in Fig. 16. Up to 500 nm the absorption from rear
side is zero. As the spectrum increases beyond 950 nm capturing of carriers will
decrease at the rear side which is shown in the plot of Figs. 16 and 17.
Figure 18 shows the plot of the total amount of sunlight escaped in the front side of
the cell with respect to wavelength. Front surface escape can be read as the sunlight
(photons) that has entered the cell but was unable to absorb by the cell and escaped
in the front surface. Front surface escape is referred as advance loss in the cell, and
this is due to surface texture and light trapping properties. Further to the analysis
presented an attempt is made to show the effect of haze on the cell. Figures 19 and
20 show respectively the effect of total haze on reflection, absorption, Esc and EQE
for different spectrums and the same are plotted.
Characterization of Bifacial Passivated Emitter and Rear … 347

Fig. 12 I-V Plot showing maximum point

Table 4 Grid resistance of solar cell

Rs grid (·cm2 ) Rs contact (·cm2 ) Rs total (·cm2 ) Metal A V (cm3 )
coverage (cm2 )
(%)
Front elements
Busbars 0.01248 0 0.01248 8.621 0.01293
Fingers 0.04611 0 0.04611 8.120 0.01044
Front 0.05859 0 0.05859 6.852 16.74 0.02337
metal
total
Rear elements
Busbars 0.01245 0 0.01245 8.621 0.01293
Fingers 0.01523 0 0.01523 21.11 0.02940
Rear 0.02769 0 0.02769 12.17 29.73 0.04233
metal
total
Total for 0.08628 0 0.08628 19.02 46.47 0.06570
cell
348 S. K. Tummala et al.

Fig. 13 Total reflection from front side of cell versus wavelength

Fig. 14 Total reflection from rear side of cell versus wavelength

Color outputs are calculated from the upward hemispherical spectrum (front
color), the downwards hemispherical spectrum (rear color), and, if a detector is
included, the spectrum incident to the detector (detector color). The program effec-
tively ‘averages’ the color of the module components that reflect light. If a module
contains a dark blue cell surrounded by a white back sheet, its ‘averaged’ color might
appear light blue; and if the cell also had fingers, then the ‘averaged’ color might be
a bluey grey (Fig. 21).
Characterization of Bifacial Passivated Emitter and Rear … 349

Fig. 15 Total absorption from front side of cell versus wavelength

Fig. 16 Total absorption from rear side of cell versus wavelength

The color of the cell has been determined using CIE 1964, 10°, Wyman 2013_Eq
3 [24] color matching function and the chromatic display type is sRGB, 6504 K,
standard. The shading coordinating capacities were dictated by estimating the mean
shading view of an example of human eyewitnesses over the visual reach from.360 to
970 nm. XYZ is a device-independent representation of colour. It does not depend on
the display monitor. The precision of XYZ increases as the number of rays increases
and as the wavelength interval decreases (down to 1 nm). XYZ is scaled such that
Y = luminosity. XYZ depends on the incident spectrum as well as the reflection,
350 S. K. Tummala et al.

Fig. 17 Cell absorption versus wavelength

Fig. 18 Front surface escape of light versus wavelength

absorption, and transmission of the simulated structure. XYZ will only be calculated
if the inputs for the minimum and maximum wavelengths are ≤360 nm and ≥970 nm,
respectively.

970 nm
X = λ xλ P(λ) (8)
λ=360
Characterization of Bifacial Passivated Emitter and Rear … 351

Fig. 19 Haze effect on Refl, Abs, Esc versus wavelength

Fig. 20 Haze effect versus EQE on Refl, Abs, Esc and spectrum

970 nm
Y = λ yλ P(λ) (9)
λ=360

970 nm
Z = λ z λ P(λ) (10)
λ=360

The X and Z parts give the shading or chromaticity of the range. Because the
apparent tone depends just upon the general extents of X, Y, and Z, we characterize
352 S. K. Tummala et al.

Fig. 21 Cell Color using color matching function

its chromaticity coordinates as:

X
x= (11)
X +Y + Z
Y
y= (12)
X +Y + Z
Z
z= (13)
X +Y + Z

Now
X +Y + Z
x+y+z= =1 (14)
X +Y + Z

z may be obtained from x and y

z = 1 − (x + y) (15)

RGB, LAB and LUV are all calculated from XYZ. These outputs depend on the
white point of the chromatic display type. Most computer monitors have a sRGB
display type where the white point is based on the 6504 K spectrum. RGB also
depends on the chromaticity coordinates and gamma curves of the chromatic display
type.

4 Conclusion

Characterization of bifacial p-type PERC solar cell with various proportions of size
and width are presented in this paper. Back Silicon Nitrate layer with various thick-
ness are streamlined and most elevated back productivity of bifacial solar cell (~22%)
Characterization of Bifacial Passivated Emitter and Rear … 353

was obtained at AM1.5 g standard test condition. Absorption, Reflection and Escape
factors are presented to verify the advance losses in the PERC cell. Total reflected
sunlight on the front side of the cell is recorded as 8.4%, escaped front side and
rear side are noted as 1.9% and 3.5% respectively. The Cell absorption is found
to be 86.215% which included parasitic absorption of 1.925%. p-type PERC when
compared to normal c-Si cell has 10.79% more absorption rate.

Acknowledgements Authors would like to thank FIST (Fund for Improvement of S&T Infras-
tructure), Department of Science & Technology, Ministry of Science & Technology, Govt. of
INDIA and GRIET Management for their sanction of grants to the institution. Grant Number:
SR/FST/College-29/2017 dt: 16 Jan 2018.

References

1. Sporleder K, Bauer J, Richter S, Hähnel A, Großer S, Turek M Hagendorf C (2019) Local

corrosion of silicon as root cause for potential-induced degradation at the rear side of bifacial
PERC solar cells. Phys Status Solidi RRL Rapid Lett. https://doi.org/10.1002/pssr.201900163
2. Cai X, Ni Z, Chen C, Ke P, Chen H, Cao H, Zhang Q (2019) The study on anti-PID performance
of high efficiency bifacial cell module. IOP Conf Ser: Mater Sci Eng. https://doi.org/10.1088/
1757-899X/556/1/012033
3. Ye F, Deng W, Guo W, Liu R, Chen D, Chen Y, Yang Y, Yuan N, Ding J, Feng Z (2017) 22.13%
efficient industrial p-type mono PERC solar cell. IEEE PVSC, Washington. https://doi.org/10.
1109/PVSC.2016.7750289
4. Zhang S, Hu D, Qian H, Jie J, Ni Z, Zhang X, Xie L (2019) Application of silicon oxide on
high efficiency mono-crystalline silicon PERC solar cells. Energies. https://doi.org/10.3390/
en12061168
5. Renshaw JS, Upadhyaya A, Upadhyaya V, Rohatgi A (2012) Crystalline silicon solar cells with
segmented selective emitter by ultraviolet laser doping. 38th IEEE PVSC, TX. https://doi.org/
10.1109/PVSC.2012.6317794
6. Lee E, Lee H, Choi J, Oh D, Shim J, Cho K, Kim J, Lee S, Hallam B, Wenham SR, Lee H
(2011) Improved LDSE processing for the avoidance of overplanting yielding 19.2% efficiency
on commercial grade crystalline Si solar cell. Sol Energy Mater Sol Cells. https://doi.org/10.
3390/en13061388
7. Dahlinger M, Eisele SJ, Lill PC, Kohler JR, Werner JH (2012) Full area laser doped boron
emitter silicon solar cells. 38th IEEE PVSC, Austin. https://doi.org/10.1109/PVSC.2012.631
7778
8. Kim C, Lee J, Lim S, Jeong C (2015) Enhanced absorption and short circuit current density
of selective emitter solar cell using double textured structure. Solar Energy. https://doi.org/10.
1016/j.solener.2015.04.010
9. Hirata K, Saitoh T, Ogane A, Sugimura E, Fuyuki T (2011) Selective emitter formation by
laser doping for phosphorous-doped n-type silicon solar cells. Appl Phys Exp. https://doi.org/
10.1143/APEX.5.016501
10. Wang KS, Lin D, An XR, Mai L, Mitchell E, Wenham SR (2012) 18.8% Efficient laser-doped
semi-conductor fingers screen-printed silicon solar cell with light-induced plating. 38th IEEE
PVSC, Austin. https://doi.org/10.1109/PVSC.2012.6317805
11. Zhang QZ, Shu BF, Chen MB, Zhong NB, Luo JJ, Wu XL (2017) Numerical investigation on
selective emitter formation by laser doping for phosphorous-doped silicon solar cells. J Laser
Appl 02-2003. https://doi.org/10.2351/1.4979303
354 S. K. Tummala et al.

12. Lopez G, Ortega P, Colina M, Voz C, Martin I, Morales-Vilches, Orpella A, Alcubilla R (2015)
Emitter formation using laser doping technique on n- and p-type c-Si substrates. Appl Surface
Sci 336:182–187. https://doi.org/10.1016/j.apsusc.2014.10.140
13. Takahashi K, Kurosawa M, Ikenoue H, Sakashita M, Takeuchi W, Nakatsuka O, Zaima S
(2016) Low thermal budget n-type doping into Ge surface using ultraviolet laser irradiation in
phosphoric acid solution. Appl Phys Lett 108:052104. https://doi.org/10.1063/1.4941236
14. Dullweber T, Wehmeier N, Nowack A, Brendemühl T, Kajari-Schröder S, Brendel R (2016)
Industrial bifacial n-type silicon solar cells applying a boron co-diffused rear emitter and an
aluminum rear finger grid. Phys Status Solid 213(11):3046–3052. https://doi.org/10.1002/pssa.
201600346
15. Jager U, Wolf A, Steinhauser B, Benick J, Nekarda J, Preu R (2012) Laser doping for high
efficiency silicon solar cells. Laser Mater Proc 847309. https://doi.org/10.1063/1.4915326
16. Sun X, Khan MR, Deline C, Alam MA (2018) Optimization and performance of bifacial
solar modules: a global perspective. Appl Energy 212:1601–1610. https://doi.org/10.1016/j.
apenergy.2017.12.041
17. Castillo-Aguilella JE, Hauser PS, Castillo J (2016) Multi-variable bifacial photovoltaic module
test results and best-fit annual bifacial energy yield model. IEEE Access 4:498–506. https://
doi.org/10.1109/ACCESS.2016.2518399
18. Park H, Chang S, Park S, Kim WK (2019) Outdoor performance test of bifacial n-type silicon
photovoltaic modules. Sustainability 11(22):6234. https://doi.org/10.3390/su11226234
19. Liu J, Janssen GJM, Koppes M, Kossen EJ, Tool K, Komatsu Y, Anker J, Gutjahr A, Vlooswijk
A, Luchies JM (2015) Selective emitter in n-type c-Si solar cells. 31st European photovoltaic
solar energy conference, Hamburg, pp 633–636. https://doi.org/10.1016/j.egypro.2016.07.122
20. Richter A, Benick J, Feldmann F, Fell A, Hermle M, Glunz SW (2017) n-Type Si solar cells with
passivating electron contact: identifying sources for efficiency limitations by wafer thickness
and resistivity variation. Solar Energy Mater Cells 173:96–105. https://doi.org/10.1016/j.sol
mat.2017.05.042
21. Bhang BG, Lee W, Kim GG, Choi JH, Park SY, Ahn H-K (2019) Power performance of bifacial
c-Si PV modules with different shading ratios. IEEE J Photovoltaics 9(5):1413–1420. https://
doi.org/10.1109/JPHOTOV.2019.2928461
22. Simayi S, Kida Y, Shirasawa K, Suzuki T, Takato H (2017) Method of removing single-side
doped layer while maintaining pyramid textured surface of n-type bifacial solar cells. IEEE J
Photovoltaics 7:458–462. https://doi.org/10.1109/JPHOTOV.2016.2646060
23. Lu G, Zheng F, Wang J, Shen W (2017) Thin Al2 O3 passivated boron emitter of n-type bifacial
c-Si solar cells with industrial process. Prog Photovoltaics: Res Appl 25(4):280–290. https://
doi.org/10.1002/pip.2859
24. PV Lighthouse: SunSolve™, accessed 12 Dec 2020, https://www.pvlighthouse.com.au
Network Reconfiguration of Distribution
System with Distributed Generation,
Shunt Capacitors and Electric Vehicle
Charging Stations

Surender Reddy Salkuti

Abstract This chapter presents a heuristic-based technique for solving the optimal
network reconfiguration (ONR) in a radial distribution system (RDS) using the
fuzzy-based multi-objective methodology. Minimization of real power losses and
deviation of nodes voltage is considered as the multiple objectives in this work and
they are modeled with fuzzy sets. The developed algorithm determines the optimal
reconfiguration of feeders with the minimum number of tie-line switch operations.
This work focuses on different combinations of ONR along with renewable-based
distributed generation (DG) units, shunt capacitors, and electric vehicle charging
stations (EVCSs). The load flow analysis implemented in this chapter is based on
an iterative approach of the receiving end voltage of RDS. The effectiveness of the
proposed heuristic-based methodology has been implemented on the IEEE 69 bus
RDS.

Keywords Network reconfiguration · Distributed generation · Power loss ·

Voltage stability · Distribution system load flow · Radial distribution system ·
Shunt capacitors · Electric vehicles

Nomenclature

NR Network reconfiguration
DG Distributed generation
DSLF Distribution system load flow
RDS Radial distribution system
EVs Electric vehicles
DERs Distributed energy resources
PEVs Plugin electric vehicles

S. R. Salkuti (B)
Department of Railroad and Electrical Engineering, Woosong University, 171 Dongdaejeon-ro
(155-3 Jayang-dong), Dong-gu, Daejeon 34606, Republic of Korea
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 355
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6_15
356 S. R. Salkuti

BFS Backward forward sweep

LIM Load impedance matrix
OFR Optimal feeder reconfiguration
ONR Optimal network reconfiguration
N Ec V , N Ed V Number of EVs charging and discharging
PG2V , PV 2G Active powers from vehicle-to-grid and grid-to-vehicle
ηc , ηd Charging and discharging efficiencies
Rc , Rd Charging and discharging rates
Ntie Number of tie-line switches

1 Introduction

The distribution system plays a crucial role among the other components of the
electrical power system that is generation and transmission. The power system is
becoming more and more complex with the increasing demand. In developing coun-
tries, power generation is usually insufficient to meet the increasing load demand.
Therefore, it is necessary to reduce the total losses in the network of which the
major part is contributed by the distribution system. The power industry is adopting
deregulation to obtain the economic efficiency of power system operation. In the
deregulated scenario, both generation and distribution companies are dedicated to
their function [1], which avoids the monopoly and creates a competitive market envi-
ronment between them. This forces the power utilities to fulfill the energy demand
of the consumers at a reasonable cost.
Recently the electric vehicles (EVs) have gained importance due to the increasing
air pollution, climate change, and increased oil prices. Distributed energy resources
(DERs) such as EVs and distributed generation (DG) are growing as an opportunity
to decarbonize the energy system. The necessity of EVs is very clear with their
great potential to electrify the transportation sector [2]. The renewable-based DG
sources create uncertainty in the power distribution system. At the same time, they
also pose new technical challenges to the power system, which can be addressed
with increased flexibility. For better utilization of electrical energy, the optimization
of both distribution system operation and control becomes necessary. This can be
achieved through the automation of the distribution system. One of the methods
adopted is the remote control of the configuration by which losses in the branches
of the entire system can be minimized. The distribution system reconfiguration is
carried out by modifying the topological structure of the network by changing the
status of the sectionalizing and tie-line switches [3, 4]. And also, the optimal switch
operations may reduce losses in the system. Both of these are met by reconfiguration.
Hence, both the ONR and less number of switching operations will reduce the power
losses and they are met by the proposed ONR approach.
Most of the distribution systems generally operated in radial topology which
enables suitable voltage and power flow control, reduced fault current, and easier
Network Reconfiguration of Distribution System with Distributed … 357

protection coordination schemes over the meshed system. Typically, the radial distri-
bution systems (RDSs) have two types of switches, namely, sectionalizing switches
which are usually open, and tie-line switches which are usually closed [5]. When the
fault occurs either on distributors or feeders, the tie-line switch allows some portion
of the faulted part to be restored promptly, thereby enhancing the reliability of the
system. According to Ref. [2], the power loss in the distribution network constitutes
70% of the total power loss. Therefore, the major cause of power interruption is due
to problems in the distribution system.

1.1 Related Work

There has been considerable interest in the recent past to develop algorithms for
feeder reconfiguration (FR) of the distribution system under various operating contin-
gencies. Usually, distribution companies try to keep active power losses below the
standard ones to gain profit rather than paying penalties. Thus the active power loss
minimization is a major concern of the distribution system researchers, which has
a significant impact on the maximum loadability of the network and hence, on the
power system stability particularly in overburdened networks. Some of the estab-
lished techniques to handle distribution systems under such a competitive scenario
include network reconfiguration (NR), DG allocation, shunt capacitor placement,
and simultaneous NR and DG allocation [6]. Hence, the existing distribution system
requires to be optimized to satisfy the demand in the most reliable, economical,
and environmentally friendlier way, while meeting the associated geographical or
operational constraints.
A high-performance nonlinear sliding mode controller has been proposed in Ref.
[7] for an EV charging system to improve the power factor (pf) to handle the unbal-
anced EV chargers and to compensate for voltage distortions. The super sense genetic
algorithm (SSGA) is applied in Ref. [8] to solve the problem of complex combina-
torial NR problem of RDSs. An approach for the optimal network rearrangement by
incorporating the plugin electric vehicles (PEVs) proposed in Ref. [9] is based on the
random programming model of the Monte Carlo simulation method. An approach
for optimal placement and sizing of electric vehicle charging stations (EVCSs) on
a distribution network is proposed in [10]. A fuzzy approach-based multi-objective
heuristic technique for ONR in distribution systems considering the DGs is proposed
in [11]. A single-phase (1-ϕ) EV charging coordination approach with the three-phase
(3-ϕ) supply and chargers connected to the EVs with the less loaded phase of the
feeder at the starting of charging has been proposed in [12]. The optimal planning
approach of EVCSs and shunt capacitors is proposed in [13] and it is solved by using
the dragonfly algorithm (DA).
An equilibrium optimizer algorithm has been applied to the ONR problem in Ref.
[14] with loss reduction, voltage magnitude enhancement, and reliability indices
improvement objectives. An efficient technique for balanced and unbalanced RDSs
optimization by ONR and optimal capacitor placement has been proposed in [15].
358 S. R. Salkuti

The ONR allows better penetration of renewable energy sources (RESs) in the RDS
and it is solved in Ref. [16] using the mixed particle swarm optimization (PSO) for
loss minimization and voltage profile enhancement improvement. Reference [17]
proposes optimal battery energy storage systems and allocation of PV-based DG
have been solved by the PSO algorithm.

1.2 Scope and Contributions

From the literature on ONR with loss minimization objective in the distribution
system, the research gap has been identified and explored the work area with current
research performance and its limitations. From the literature, it has been identified
that there is a requirement for solving the ONR problem by simultaneously installing
the renewable-based DG units, shunt capacitors, and EVCSs. Renewable-based DG
units, i.e., wind plants and solar PV farms have wind speed and solar insolation as
input parameters and they are highly intermittent. The distribution load flow (DLF)
used in this chapter is based on the iterative approach. The potential of this approach
has made the ONR approach is very powerful and can be applied to any size of the
distribution network. The ONR and DG allocation to strengthen the efficiency of
distribution systems based on power loss minimization and voltage deviation mini-
mization, as these are two major issues in the recent competitive power scenario, and
they are considered as the objective functions with the presence of shunt capacitors
and EVCSs. The simulation has occurred to both balanced as well as unbalanced
radial distribution systems (RDSs).
This chapter is organized as follows: The description of RDS, ONR, and the
summary of the literature work has been presented in Sect. 1. A brief description
of distribution load flow (DLF) analysis has been presented in Sect. 2. Section 3
describes the modeling of shunt capacitors and EVCSs in the distribution system.
Problem formulation is presented in Sect. 4. Section 5 describes the solution method-
ology. Section 6 describes the results and discussion on the 69 bus test system. The
conclusions of this chapter have been summarized in Sect. 7.

2 Distribution Load Flow (DLF) Analysis

The analysis of DLF is basic but it is an essential mathematical tool for the analysis
of distribution systems in both the planning and operational stages. The primary aim
of the power flow analysis is to determine the magnitude and phase of steady-state
voltage at all buses, active and reactive power flows in each line, for a specified
loading. There are numerous power flow methods like Newton Raphson, Gauss-
Seidel, fast decoupled methods, and many more methods with a modification in
conventional ones. Due to the different properties of the distribution systems, these
methods are not suitable for load flow analysis [18]. Certain applications including
Network Reconfiguration of Distribution System with Distributed … 359

distribution automation and power system optimization require efficient and robust
load flow solutions. Over the last few decades, these load flow methods have been
evolved in several different dimensions to handle both static and dynamic power
distribution system problems. Traditionally, the Backward Forward Sweep (BFS)
load flow techniques are applied to distribution systems and it has two-step analyses.
Other load flow methods include little modification into existing techniques for their
advantage over the older ones. In literature, several conventional methods have been
utilized to solve distribution system problems [19]. The open branches are electrically
represented with very high impedance. Whenever a branch is connected, then its
parameters are replaced with actual values (i.e., resistance and reactance values) and
vice versa when a branch is removed or disconnected. In other words, when a branch
exchange takes place, only those parameters will be modified accordingly for further
processing. This saves a lot of computation burden.
The proposed load flow solution presented in this chapter depends on an itera-
tive approach of the receiving end voltage of the RDS. It is successfully applied on
ill-conditioned RDS with consideration of realistic load [20]. In the first step, the
effective power at each bus is determined after forming adjacent branches and adja-
cent node matrices. A sparse technique is used to determine the branches and nodes
beyond a particular node. The detailed mathematical formulation is given below
considering the electrical equivalent of a branch connected between the nodes a and
b of RDS, and it is shown in Fig. 1.
The amount of current flowing from node a to node b can be expressed as [20],

|Va |∠δa − |Vb |∠δb Pb − j Q b

Iab = = (1)
Rab + j X ab (|Vb |∠δb )∗

where active power (Pb ) can be expressed in terms of active power load at a bus/node
i (PDi ) and active power loss of line k (Ploss,k ). Mathematically, it can be expressed
as,


Nb 
Nbr
Pb = PDi + Ploss,k (2)
i=1 k=1

where Nb shows all the buses beyond the bus b. Nbr shows all the branches beyond
the bus b. From Eq. (1), Pb can be expressed as [21],

Fig. 1 Electrical equivalent Va ∠δ a Vb ∠δb

of a typical distribution
system branch Iab
Rab+jXab
Pb+jQb
a b
PDa+jQDa PDb+jQDb
360 S. R. Salkuti

|Va ||Vb | sin(δa − δb ) + Rab Q b

Pb = (3)
X ab

From the above equation, the voltage magnitude (|Vb |) and angle (δb ) at the end
of receiving node can be calculated by using,
  
Rab
|Vb | = − |Va | sin δ − cos δ
X ab
 2  2 1/2
Rab Rab
+ |Va | sin δ − cos δ − 4Q b + X ab (4)
X ab X ab

where δ = δa − δb .
 
−1 Pb X ab − Q b Rab
δb = δa − tan (5)
|Vb |2 + Pb Rab + Q b X ab

The active and reactive power losses are calculated by using Eq. (3), and they are
expressed as,

Pb2 + Q 2b Rab
Ploss,ab = (6)
|Vb |2

Pb2 + Q 2b X ab
Q loss,ab = (7)
|Vb |2

All such techniques work well with static systems where there is no change in
the topology of the network [22]. Again under critical loading conditions, there
is no guarantee of their convergence. Even in converged cases, these methods are
very inefficient in respect of storage requirements and solution speed. Moreover,
for dynamic systems, it is a challenge to arrange the line data as per the load flow
requirement and to maintain the radiality, and ensure connectivity. This necessitates
the utilization of improved data structure-based techniques.

3 Modeling of Shunt Capacitor and EVCS

in the Distribution System

The ever-growing population leads to a significant increment in customer load

demand. It leads to the placement of DGs in RDS being nearer to the load demand.
Among the various renewable-based DGs, solar PV and wind energy are widely used
as they are abundantly available. As there is a rapid growth in load demand, the line
losses in the distribution network are quite high and need to be taken care of [23].
Network Reconfiguration of Distribution System with Distributed … 361

Various techniques have been implemented to RDSs apart from the DG penetration
to optimize the power losses in the RDS. This section presents the modeling of shunt
capacitors and EVCSs in the RDS.

3.1 Modeling of Shunt Capacitor

Shunt capacitors supply the amount of reactive power to the RDS at the bus where
they are connected. This in turn causes a reduction in reactive power flowing in the
line. If the reactive power and the system voltage are assumed to be constant, then
the losses are inversely proportional to the power factor, and hence improvement in
the power factor causes a reduction in system losses. The other benefits of installing
the shunt capacitors are voltage profile improvement, decrease in kVA loading, and
reduces system improvement cost/kVA of load supplied [24]. And also, to overcome
the compensation during the light load conditions, the automatic switching units can
be provided but this switching equipment is costly and this, in turn, will limit the
number of capacitors and thus the minimum capacity of the capacitor bank that has
to be provided on the feeder.
The placement of shunt capacitors in the DS reduces the system losses, enhances
the voltage profile, and also corrects the power factor. Figure 2 depicts the represen-
tation of the shunt capacitor in the DS. This capacitor injects reactive power (Q c )
into the system.
Amount of reactive power injected at bus b (Q in j,b ) can be expressed by [25],

Q in j,b = Q Db − Q c (8)

c
Now the active power loss with shunt capacitor (Ploss,ab ) can be expressed as [25],
 

Pb2 + Q in
2
j,b Rab Pb2 + (Q Db − Q c )2 Rab
c
Ploss,ab = = (9)
|Vb |2 |Vb |2

2

Pb2 + Q 2b Rab Q c − 2Q Db Q c Rab

c
Ploss,ab = + = Ploss,ab + Ploss,ab
c
(10)
|Vb |2 |Vb |2

Fig. 2 Representation of Va ∠δ a Vb ∠δ b
shunt capacitor in the
distribution system Iab
Rab+jXab
Pb+jQb
a b
PDa+jQDa Shunt jQC PDb+jQDb
Capacitor
362 S. R. Salkuti

where Ploss,ab
c
is the reduction in power loss, i.e., active power loss before and after
placing the shunt capacitor [26], and it can be expressed from Eq. (10) as,

Q 2c − 2Q Db Q c Rab
Ploss,ab
c
= (11)
|Vb |2

3.2 Modeling of EVCS in the Distribution System

Figure 3 depicts the representation of EVCS in the distribution system.

EVCS
The power demand of EVCS at bus b (PDb ) can be calculated by using [27,
28],
EVCS
PDb = N Ec V PG2V ηc Rc − N Ed V PV 2G ηd Rd (12)

4 Problem Formulation

This section presents the general description of distribution systems and the mathe-
matical modeling of optimal feeder reconfiguration (OFR) or optimal network recon-
figuration (ONR). Factors that are affecting the increase in the power losses in the
distribution network are feeder length, low voltage, low power factor, poor workman-
ship in fittings, and reduction of line losses. Various methods used for the reduction
of distribution system losses are the construction of a new substation, reinforcement
of feeder, reactive power compensation, HV distribution system, grading of conduc-
tors, and feeder reconfiguration [29]. In this work, two objectives, i.e., real power
loss and voltage deviations are modeled with fuzzy sets [29, 30]. Some heuristics
are developed to reduce the number of tie-line switching operations.

Fig. 3 Representation of Va ∠δ a Vb ∠δ b
EVCS in the RDS Iab
Rab+jXab
Pb+jQb
a b IEV
PDa+jQDa

EVCS
Network Reconfiguration of Distribution System with Distributed … 363

4.1 Fuzzy Membership Function for Active Power Loss

Reduction (µL i )

The basic purpose for membership function, i.e., objective in the fuzzy domain is to
minimize the active power loss of the system. The variable αi can be defined as

Ploss (i)
αi = 0
f or i = 1, 2, . . . , Nk (13)
Ploss

where Nk is total number of lines in loop including the tie-line, when the kth tie-
switch is closed, Ploss (i) is total active power loss when ith line in the loop is opened,
0
and Ploss is total real power loss before the NR. Membership/objective function for
active power loss reduction (μL i ) can be written as [30],
⎧
⎪ αmax −αi
⎨ αmax −αmin
f or αmin < αi < αmax
μL i = 1 f or αi ≤ αmin (14)
⎪
⎩0 f or αi ≥ αmax

4.2 Fuzzy Membership Function for Maximum Node Voltage

Deviation (µVi )

The main aim of this function is to minimize the deviation of nodes’ voltage. The
variable βi can be expressed as
 

βi = max Vi, j − Vs  f or i = 1, 2, . . . , Nk ; j = 1, 2, . . . , N B (15)

where N B is total number of buses in RDS, Vs is substation voltage, and Vi, j is jth bus
voltage corresponding to the opening of the ith line [29, 30]. The fuzzy membership
function for maximum bus voltage deviation (μVi ) can be expressed as,
⎧
⎪ βmax −βi
⎨ βmax −βmin f or βmin < βi < βmax
μVi = 1 f or βi ≤ βmin (16)
⎪
⎩0 f or βi ≥ βmax
364 S. R. Salkuti

4.3 Constraints

The active and reactive power balances of the RDS system including the DG units,
shunt capacitors, and EVCSs are expressed as [31, 32],


N DG
PD = PGGrid + EVCS
PDb + PDG,i (17)
i=1


N DG 
Nc
Q D = Q Grid
G + Q DG,i + Q c, j (18)
i=1 j=1

Voltages at each bus can be expressed as,

0.95 ≤ Vi ≤ 1.05 (19)

Active and reactive powers of DG units can be expressed as [33, 34],

min
PDG,i ≤ PDG,i ≤ PDG,i
max
(20)

DG,i ≤ Q DG,i ≤ Q DG,i

Q min max
(21)

4.4 Selection of Best-Compromised Solution

When optimizing two or more objectives simultaneously, a best-compromised

solution needs to be determined [35]. The procedure for determining the best-
compromised solution using the min-max principle is determined next:
The membership function values of the two objectives are determined. When the
kth tie-line switch of RDS is closed, a loop is formed with number of lines in the
loop Nk . After opening the ith line in the loop, run the DLF to determine μL i and
μVi for i = 1, 2, …, Nk . Determine the fuzzy decision for overall satisfaction [36,
37] by using,

Dk,i = min(μL i , μVi ) f or 1, 2, . . . , Nk (22)

The optimal solution is the maximum of overall degrees of satisfaction, and it is

expressed as [38],

O Sk = max Dk,i f or 1, 2, . . . , Nk (23)

Network Reconfiguration of Distribution System with Distributed … 365

5 Solution Methodology

This section presents heuristics for minimizing the number of operations of tie-line
switches. Here, heuristic rules are developed to minimize the number of tie-line
switch operations [39, 40]. The flow chart of the proposed solution methodology has
been depicted in Fig. 4, and the step-by-step approach is presented next:
• Step 1: Read the RDS test system data.

Read the input data related To the distribution system and convert all the line
parameters to per unit values. Read the number of tie-line switches ( N tie ) for
the test system under consideration.

Run the load flow described in section 2, and calculate the line
0
losses. The base case losses is designated as Ploss

Determine the voltage difference (VD) across open tie-line

switches and identify the tie-line switches that are open
across which the VD is maximum.

Yes
Is max(VD)<ε? STOP

No
Compute the membership values and then the overall satisfaction.
Dk ,i = min ( μ Li and μVi )

When the selected tie-line switch is closed then identify

the number of branches ( N k ) in the loop.

Run the load flow with the branch k open.

Calculate the line losses.

Compute the membership values and then the overall satisfaction.

Dk ,i = min ( μ Li and μVi )

Yes
k = k+1 Is k < N k ?
No
Determine the optimal solution for the operation of tie-line switch k.
OSk = max( Dk ,i ) Where i=1,2,3, , N k

N tie = N tie − 1

Yes
Is N tie > 0 ?
No
STOP

Fig. 4 Flow chart of the proposed ONR/OFR algorithm

366 S. R. Salkuti

• Step 2: Execute the load flow solution as described in Sect. 2.

• Step 3: Determine voltage difference (Vtie ) across the open tie-line switches.
• Step 4: Identify the open tie-line switch across which Vtie is maximum, and it
can be represented as (Vmaxtie ).
• Step 5: If Vmax
tie > specified value (ε), then go to Step 6 else go to Step 11.
• Step 6: When the selected tie-line switch is closed then identify the number of
branches (Nk ) in the loop.
• Step 7: Open one line at a time in the loop, and determine the membership value
for each objective function. Compute μL i and μVi using the Eqs. (14) and (16),
respectively.
• Step 8: Compute the overall degree of satisfaction using Eq. (22).
• Step 9: Determine the optimal solution for the operation of the kth tie-line switch
using Eq. (23).
• Step 10: Make the number of tie-line switches (Ntie ) equal to Ntie − 1, and
rearrange the coding of the rest of the tie-line switches, and go to Step 2.
• Step 11: Display the output results.

6 Results and Discussion

The proposed ONR methodology has been implemented on IEEE 69 bus test system
which has a single feeder with a single substation [41]. Figure 5 depicts the single-
line diagram of 69 bus RDS. System load demand, line, and tie-line data have been
taken from Ref. [42]. This system has 68 lines, i.e., sectionalizing switches, they
are 1–68 and they are normally closed. Five tie-line switches (which form 5 loops)
considered in this work are 69, 70, 71, 72, and 73, they open tie switches. The base
voltage and kVA are 12.66 kV and 1000 kVA, respectively. The real and reactive
power load of 3,802 kW and 2,694 kVAr, respectively. In this test system, the DG
units are placed at buses 5, 28, 45, and 60; shunt capacitors are placed at buses 22,
36, and 64; EVCSs are placed at buses 18 and 59.
In the present work, the convergence criterion (ε) is considered as 0.01, and it has
been assumed that αmin is 0.5, αmax is 1, βmin is 0.05 and βmax is 0.10. The active
power loss obtained in the base case condition is 224.96 kW, and all the tie-line
switches, i.e., 69, 70, 71, 72, and 73. The minimum voltage obtained in this base
case is 0.9066 p.u. at bus 54.

6.1 Case 1: Tie-Line Switch Operation 1

In this case, the voltage difference across each tie-line switch is determined. The
voltage differences across tie-line switches 69, 70, 71, 72 and 73 are 0.0031 p.u.,
0.0008 p.u., 0.0416 p.u., 0.0742 p.u. and 0.0471 p.u., respectively. From these volt-
ages, it can be observed that voltage difference across line number 72 is maximum,
Network Reconfiguration of Distribution System with Distributed … 367

36 37 38 39 40 41 42 43 44 45 46
35 36 37 38 39 40 41 42 43 44 45 1 to 69 : Buses /
Nodes
1 to 68 :
69 DG Lines/Branches
SC
71 69 to 73 : Tie-line
51 52 68 69 Switches
50 51 67 68
EVCS

1 2 4 5 6 7 8 9 10 11 13 15 17 19 21 23 25 26 27
1 3 4 5 7 9 10 13 17 19 21 22 24 26
SS
2 6 8 11 12 14 15 16 18 20 23 25
3 12 14 16 18 20 22 24
DG SC

46 70
66 67
65 73
66 EVCS
47 49 50
47 48 49
53 54 55 56 57 58 59 60 61 62 63 64 65
48 52 53 54 55 56 57 58 60 61 62 63 64
59
72
DG SC
28 29 30 31 32 33 34 35
27 28 29 30 31 32 33 34

DG

Fig. 5 SLD of IEEE 69 bus RDS before ONR

i.e., 0.0742 p.u. Table 1 presents the membership values of power loss and voltage
deviation, and the overall satisfaction for tie-line switch operation 1. In this case,
line number 72 is closed and the membership values for opened lines are presented
in Table 1. The overall satisfaction has been determined by using Eq. (23), and they
are presented in the table. From the results obtained, it is observed that by using the
fuzzy set intersection, the fuzzy decision for overall satisfaction is obtained when
line 46 is open and line 72 is closed. The obtained value of overall satisfaction is
0.7362, which is the maximum of Dk,i .

6.2 Case 2: Tie-Line Switch Operation 2

The voltage differences across tie-line switches 69, 70, 71, 72 and 73 are 0.00312 p.u.,
0.0008 p.u., 0.0416 p.u., 0.0742 p.u. and 0.0471 p.u., respectively. From these volt-
ages, after the tie-line switch operation 1, it can be observed that voltage difference
across line number 73 is maximum, i.e., 0.0471 p.u. Table 2 presents the membership
values of power loss and voltage deviation, and the overall satisfaction for tie-line
switch operation 2. In this case, line number 73 is closed and the membership values
for opened lines are presented in Table 2. The overall satisfaction has been determined
368 S. R. Salkuti

Table 1 Membership values for tie-line switch operation 1

Closed line Opened line Membership values of Dk,i = min(μL i , μVi )
power loss and voltage
deviation
μL i μVi
Base case 1 0.5654 0.5654
72 38 0 0.4378 0
72 37 0 0.3937 0
72 36 0 0.3515 0
72 35 0 0.3518 0
72 47 0.7353 0.8742 0.7353
72 46 0.7362 0.8745 0.7362
72 45 0.7352 0.8734 0.7352
72 44 0.7350 0.8734 0.7350
72 43 0.7224 0.8642 0.7224
72 42 0.7082 0.8540 0.7082
72 41 0.7052 0.8540 0.7052
72 8 0 0.1457 0
72 7 0 0 0
72 6 0 0 0
72 5 0 0 0
72 4 0 0 0

by using Eq. (23), and they are presented in the table. From the results obtained, it
is observed that by using the fuzzy set intersection, the fuzzy decision for overall
satisfaction is obtained when line 53 is open and line 73 is closed. The obtained value
of overall satisfaction is 0.7575 which is the maximum of Dk,i .

6.3 Case 3: Tie-Line Switch Operation 3

The voltage differences across tie-line switches 69, 70, 71, 72 and 73 are 0.00312
p.u., 0.0008 p.u., 0.0416 p.u., 0.0742 p.u. and 0.0471 p.u., respectively. From these
voltages, after the tie-line switch operations 1 and 2, it can be observed that voltage
difference across line number 71 is maximum, i.e., 0.0416 p.u. Table 3 presents the
membership values of power loss and voltage deviation, and the overall satisfaction
for tie-line switch operation 3. In this case, line number 71 is closed and the member-
ship values for opened lines are presented in Table 3. The overall satisfaction has been
determined by using Eq. (23), and they are presented in the table. From the results
obtained, it is observed that by using the fuzzy set intersection, the fuzzy decision
Network Reconfiguration of Distribution System with Distributed … 369

Table 2 Membership values for tie-line switch operation 2

Closed line Opened line Membership values of Dk,i = min(μL i , μVi )
power loss and voltage
deviation
μL i μVi
After tie-line switch operation 1 0.7386 0.8632 0.7386
73 26 0.7203 0.8621 0.7203
73 25 0.7225 0.8520 0.7225
73 24 0.7265 0.8505 0.7265
73 23 0.7071 0.8248 0.7071
73 22 0.7078 0.8246 0.7078
73 21 0.7046 0.7051 0.7046
73 20 0.6199 0.7056 0.6199
73 19 0.6175 0.7062 0.6175
73 18 0.6163 0.7120 0.6163
73 17 0.5656 0.6485 0.5656
73 16 0.5152 0.5836 0.5152
73 15 0.4420 0.5352 0.4420
73 14 0.4447 0.5360 0.4447
73 13 0.4338 0.5225 0.4338
73 12 0.4232 0.5122 0.4232
73 11 0.0774 0.1995 0.0774
73 10 0 0 0
73 9 0 0 0
73 8 0 0 0
73 7 0 0 0
73 6 0 0 0
73 5 0 0 0
73 4 0 0 0
73 53 0.7575 0.9129 0.7575
73 52 0.7548 1 0.7548
73 51 0.7542 1 0.7542
73 50 0.7435 0.9792 0.7435
73 49 0 0 0
73 48 0 0 0
73 72 0 0 0
73 38 0 0 0
73 37 0 0 0
73 36 0 0 0
(continued)
370 S. R. Salkuti

Table 2 (continued)
Closed line Opened line Membership values of Dk,i = min(μL i , μVi )
power loss and voltage
deviation
μL i μVi
73 35 0 0 0
The bold values in the table represent the fuzzy decision for overall satisfaction by using the fuzzy
set intersection

Table 3 Membership values for tie-line switch operation 3

Closed line Opened line Membership values of Dk,i = min(μL i , μVi )
power loss and voltage
deviation
μL i μVi
After tie-line switch operation 2 0.7252 0.8625 0.7252
71 14 0.8618 0.9321 0.8618
71 13 0.8712 0.9652 0.8712
71 12 0.8706 0.9025 0.8706
71 11 0.8568 0.9158 0.8568
71 10 0.7990 0.9198 0.7990
71 9 0.7887 0.9138 0.7887
71 8 0.7458 0.9166 0.7458
71 7 0.3452 0.8252 0.3452
71 6 0.2898 0.7879 0.2898
71 5 0.2898 0.7842 0.2898
71 4 0.2858 0.7840 0.2858
71 68 0.7365 0.9198 0.7365
71 67 0.7165 0.9174 0.7165
71 66 0.7138 0.9133 0.7138
71 65 0.7086 0.9165 0.7086
71 64 0.7152 0.9114 0.7152
71 63 0.7098 0.9144 0.7098
71 62 0.6954 0.9100 0.6954
71 61 0.6788 0.9152 0.6788
71 60 0.6763 0.9126 0.6763
71 59 0.6552 0.9126 0.6552
71 58 0.6466 0.9126 0.6466
The bold values in the table represent the fuzzy decision for overall satisfaction by using the fuzzy
set intersection
Network Reconfiguration of Distribution System with Distributed … 371

36 37 38 39 40 41 42 43 44 45 46
35 36 37 38 39 40 41 42 43 44 45 1 to 69 : Buses /
Nodes
1 to 68 :
69 DG Lines/Branches
SC
71 69 to 73 : Tie-line
51 52 68 69 Switches
50 51 67 68
EVCS

1 2 4 5 6 7 8 9 10 11 13 15 17 19 21 23 25 26 27
1 3 4 5 7 9 10 17 19 21 22 24 26
SS
2 6 8 11 12 14 15 16 18 20 23 25
3 12 14 16 18 20 22 24
DG SC
70
66 67
65 73
66 EVCS
47 49 50
47 48 49
53 54 55 56 57 58 59 60 61 62 63 64 65
48 52 54 55 56 57 58 60 61 62 63 64
59
72
DG SC
28 29 30 31 32 33 34 35
27 28 29 30 31 32 33 34

DG

Fig. 6 SLD of IEEE 69 bus RDS after ONR

for overall satisfaction is obtained when line 13 is open and line 71 is closed. The
obtained value of overall satisfaction is 0.8712 which is the maximum of Dk,i .

6.4 Case 4: Tie-Line Switch Operation 4

The voltage differences across tie-line switches 69, 70, 71, 72 and 73 are 0.0031
p.u., 0.0008 p.u., 0.0416 p.u., 0.0742 p.u. and 0.0471 p.u., respectively. From these
voltages, after the tie-line switch operations 1, 2, and 3, it can be observed that voltage
difference across line number 69 is maximum, i.e., 0.0031 p.u. However, this voltage
difference is less than ε (0.01). Therefore, there is no further network reconfiguration
is required. Figure 6 depicts the final topology of IEEE 69 bus RDS after ONR. Bus
voltages before and after the ONR are presented in Table 4. From this table, it can be
observed the voltage profile has been improved after the proposed ONR approach.

7 Conclusions

This paper proposes an optimal network/feeder reconfiguration (ONR/OFR) problem

of the radial distribution system (RDS), and it is solved by simultaneously allo-
cating the distributed generation (DG), shunt capacitors, and electric vehicle charging
372 S. R. Salkuti

Table 4 Bus voltages before and after the ONR

Bus number Before ONR After ONR Bus Number Before ONR After ONR
1 1.0000 1.0000 36 0.9998 0.9997
2 1.0000 1.0000 37 0.9990 0.9965
3 0.9999 0.9999 38 0.9967 0.9855
4 0.9998 0.9998 39 0.9962 0.9828
5 0.9989 0.9996 40 0.9756 0.9921
6 0.9889 0.9968 41 0.9756 0.9932
7 0.9784 0.9941 42 0.9720 0.9921
8 0.9759 0.9934 43 0.9687 0.9920
9 0.9748 0.9932 44 0.9643 0.9919
10 0.9698 0.9896 45 0.9599 0.9920
11 0.9687 0.9890 46 0.9374 0.9922
12 0.9655 0.9879 47 0.9264 0.9919
13 0.9626 0.9874 48 0.9221 0.9458
14 0.9597 0.9875 49 0.9171 0.9411
15 0.9568 0.9872 50 0.9097 0.9342
16 0.9563 0.9871 51 0.9094 0.9339
17 0.9554 0.9856 52 0.9090 0.9336
18 0.9554 0.9856 53 0.9071 0.9327
19 0.9549 0.9850 54 0.9066 0.9820
20 0.9546 0.9846 55 0.9686 0.9889
21 0.9542 0.9840 56 0.9686 0.9889
22 0.9541 0.9844 57 0.9652 0.9876
23 0.9541 0.9838 58 0.9652 0.9876
24 0.9539 0.9835 59 0.9999 0.9999
25 0.9538 0.9837 60 0.9997 0.9993
26 0.9537 0.9828 61 0.9996 0.9986
27 0.9537 0.9827 62 0.9995 0.9984
28 0.9999 0.9999 63 0.9995 0.9984
29 0.9999 0.9999 64 0.9988 0.9947
30 0.9997 0.9997 65 0.9986 0.9925
31 0.9997 0.9997 66 0.9985 0.9924
32 0.9996 0.9996 67 0.9985 0.9923
33 0.9993 0.9993 68 0.9984 0.9916
34 0.9990 0.9990 69 0.9984 0.9919
35 0.9989 0.9989
Network Reconfiguration of Distribution System with Distributed … 373

stations (EVCSs). In the proposed ONR problem, the objectives, i.e., active power
loss and voltage deviation minimizations are solved by using the fuzzy-based multi-
objective methodology. An iterative approach-based distribution load flow (DLF)
has been used in this work. The proposed algorithm identifies the ONR of feeders
with the minimum number of tie-line switch operations. Simulation studies have
been performed on 69 bus RDS.

Acknowledgements This research work was funded by “Woosong University’s Academic

Research Funding-2021”.

References

1. Uniyal A, Sarangi S (2021) Optimal network reconfiguration and DG allocation using adaptive
modified whale optimization algorithm considering probabilistic load flow. Electr Power Syst
Res 192. https://doi.org/10.1016/j.epsr.2020.106909
2. Aman MM, Jasmon GB, Bakar AHA, Mokhlis H, Karimi M (2014) Optimum shunt capacitor
placement in distribution system—a review and comparative study. Renew Sustain Energy Rev
30:429–439. https://doi.org/10.1016/j.rser.2013.10.002
3. Das S, Das D, Patra A (2017) Reconfiguration of distribution networks with optimal placement
of distributed generations in the presence of remote voltage controlled bus. Renew Sustain
Energy Rev 73:772–781. https://doi.org/10.1016/j.rser.2017.01.055
4. Srinivasan G, Visalakshi S (2017) Application of AGPSO for power loss minimization in radial
distribution network via DG units, capacitors and NR. Energy Procedia 117:190–200. https://
doi.org/10.1016/j.egypro.2017.05.122
5. Biswas PP, Mallipeddi R, Suganthan PN, Amaratunga GAJ (2017) A multiobjective approach
for optimal placement and sizing of distributed generators and capacitors in distribution
network. Appl Soft Comput 60:268–280. https://doi.org/10.1016/j.asoc.2017.07.004
6. Manbachi M, Sadu A, Farhangi H, Monti A, Palizban A, Ponci F, Arzanpour S (2016) Impact
of EV penetration on volt–VAR optimization of distribution networks using real-time co-
simulation monitoring platform. Appl Energy 169:28–39. https://doi.org/10.1016/j.apenergy.
2016.01.084
7. Farkoush SG, Kim CH, Jung HC, Lee S, Umpon NT, Rhee SB (2017) Power factor improvement
of distribution system with EV chargers based on SMC Method for SVC. J Electr Eng Technol
12(4):1340–1347. https://doi.org/10.5370/JEET.2017.12.4.1340
8. Agrawal P, Kanwar N, Gupta N, Niazi KR, Swarnkar A (2020) Network reconfiguration of
radial active distribution systems in uncertain environment using super sense genetic algorithm.
Int J Emerg Electr Power Syst 21(2). https://doi.org/10.1515/ijeeps-2019-0051
9. Tabatabaei J, Moghaddam MS, Baigi JM (2020) Rearrangement of electrical distribution
networks with optimal coordination of grid-connected hybrid electric vehicles and wind power
generation sources. IEEE Access 8:219513–219524. https://doi.org/10.1109/ACCESS.2020.
3042763
10. Chen L, Xu C, Song H, Jermsittiparsert K (2021) Optimal sizing and sitting of EVCS in the
distribution system using metaheuristics: a case study. Energy Rep 7:208–217. https://doi.org/
10.1016/j.egyr.2020.12.032
11. Gampa SR, Das D (2017) Multi-objective approach for reconfiguration of distribution systems
with distributed generations. Electr Power Comp Syst 45(15):1678–1690. https://doi.org/10.
1080/15325008.2017.1378944
12. Vega-Fuentes E, Denai M (2019) Enhanced electric vehicle integration in the UK low-voltage
networks with distributed phase shifting control. IEEE Access 7:46796–46807. https://doi.org/
10.1109/ACCESS.2019.2909990
374 S. R. Salkuti

13. Rajesh P, Shajin FH (2021) Optimal allocation of EV charging spots and capacitors in distribu-
tion network improving voltage and power loss by quantum-behaved and Gaussian mutational
dragonfly algorithm (QGDA). Electr Power Syst Res 194. https://doi.org/10.1016/j.epsr.2021.
107049
14. Cikan M, Kekezoglu B (2021) Comparison of metaheuristic optimization techniques including
equilibrium optimizer algorithm in power distribution network reconfiguration. Alex Eng J.
https://doi.org/10.1016/j.aej.2021.06.079
15. Sedighizadeh M, Bakhtiary R (2016) Optimal multi-objective reconfiguration and capacitor
placement of distribution systems with the Hybrid Big Bang-Big Crunch algorithm in the fuzzy
framework. Ain Shams Eng J 7(1):113–129. https://doi.org/10.1016/j.asej.2015.11.018
16. Essallah S, Khedher A (2020) Optimization of distribution system operation by network recon-
figuration and DG integration using MPSO algorithm. Renew Energy Focus 34:37–46. https://
doi.org/10.1016/j.ref.2020.04.002
17. Mukhopadhyay B, Das D (2020) Multi-objective dynamic and static reconfiguration with opti-
mized allocation of PV-DG and battery energy storage system. Renew Sustain Energy Rev 124.
https://doi.org/10.1016/j.rser.2020.109777
18. Wang Y, Zhang N, Chen Q, Yang J, Kang C, Huang J (2017) Dependent discrete convolution
based probabilistic load flow for the active distribution system. IEEE Trans Sustain Energy
8(3):1000–1009. https://doi.org/10.1109/TSTE.2016.2640340
19. Murari K, Padhy NP (2019) A network-topology-based approach for the load-flow solution of
AC–DC distribution system with distributed generations. IEEE Trans Industr Inf 15(3):1508–
1520. https://doi.org/10.1109/TII.2018.2852714
20. Nagaraju K, Sivanagaraju S, Ramana T, Prasad PV (2011) A novel load flow method for radial
distribution systems for realistic loads. Electr Power Comp Syst 39(2):128–141. https://doi.
org/10.1080/15325008.2010.526984
21. Liu J, Wang X, Fang W, Cheng L, Niu S, Huo C, Wang J (2016) A novel load flow model
for distribution systems based on current injections. In: China international conference on
electricity distribution (CICED), pp 1–6. https://doi.org/10.1109/CICED.2016.7576094
22. Sereeter B, Markensteijn AS, Kootte ME, Vuik C (2021) A novel linearized power flow
approach for transmission and distribution networks. J Comput Appl Math 394. https://doi.
org/10.1016/j.cam.2021.113572
23. Muthukumar K, Jayalalitha S (2017) Integrated approach of network reconfiguration with
distributed generation and shunt capacitors placement for power loss minimization in radial
distribution networks. Appl Soft Comput 52:1262–1284. https://doi.org/10.1016/j.asoc.2016.
07.031
24. Tolabi HB, Lashkar Ara A, Hosseini R (2020) A new thief and police algorithm and its appli-
cation in simultaneous reconfiguration with optimal allocation of capacitor and distributed
generation units. Energy 203. https://doi.org/10.1016/j.energy.2020.117911
25. Available. [Online]: https://assets.researchsquare.com/files/rs-281250/v1_stamped.pdf?c=162
0244094
26. Esmaeilian HR, Fadaeinedjad R (2015) Distribution system efficiency improvement using
network reconfiguration and capacitor allocation. Int J Electr Power Energy Syst 64:457–468.
https://doi.org/10.1016/j.ijepes.2014.06.051
27. Salkuti SR (2020) Optimal network reconfiguration with distributed generation and electric
vehicle charging stations. Int J Math Eng Manage Sci 6(4):1174–1185. https://doi.org/10.
33889/IJMEMS.2021.6.4.070
28. Manan WIABWA, Bin Saedi A, Peeie MHB, Abu Hanifah MSB (2021) Modeling of the
network reconfiguration considering electric vehicle charging load. In: 8th international confer-
ence on computer and communication engineering (ICCCE), pp 82–86. https://doi.org/10.1109/
ICCCE50029.2021.9467143
29. Das D (2006) A fuzzy multiobjective approach for network reconfiguration of distribu-
tion systems. IEEE Trans Power Deliv 21(1):202–209. https://doi.org/10.1109/TPWRD.2005.
852335
Network Reconfiguration of Distribution System with Distributed … 375

30. Das D (2006) Reconfiguration of distribution system using fuzzy multi-objective approach. Int
J Electr Power Energy Syst 28(5):331–338. https://doi.org/10.1016/j.ijepes.2005.08.018
31. Syahputra R, Robandi I, Ashari M (2012) Reconfiguration of distribution network with DG
using fuzzy multi-objective method. In: International conference on innovation management
and technology research, pp 316–321. https://doi.org/10.1109/ICIMTR.2012.6236410
32. Stojanović B, Rajić T (2017) Novel approach to reconfiguration power loss reduction problem
by simulated annealing technique. Int Trans Electr Energy Syst 27(12). https://doi.org/10.1002/
etep.2464
33. Sun Q, Yu Y, Li D, Hu X (2021) A distribution network reconstruction method with DG and
EV based on improved gravitation algorithm. Syst Sci Control Eng 9(2):6–13. https://doi.org/
10.1080/21642583.2020.1833781
34. Amin A, Tareen WUK, Usman M, Memon KA, Horan B, Mahmood A, Mekhilef S (2020)
An integrated approach to optimal charging scheduling of electric vehicles integrated with
improved medium-voltage network reconfiguration for power loss minimization. Sustainability
12(21). https://doi.org/10.3390/su12219211
35. Hota AP, Mishra S, Mishra DP, Salkuti SR (2021) Allocating active power loss with network
reconfiguration in electrical power distribution systems. Int J Power Electron Drive Syst
12(1):130–138. https://doi.org/10.11591/ijpeds.v12.i1.pp130-138
36. Salkuti SR (2021) Feeder reconfiguration in unbalanced distribution system with wind and
solar generation using ant lion optimization. Int J Adv Comput Sci Appl 12(3):31–39. https://
doi.org/10.14569/IJACSA.2021.0120304
37. Shi Q, Li F, Olama M, Dong J, Xue Y, Starke M, Winstead C, Kuruganti T (2021) Network
reconfiguration and distributed energy resource scheduling for improved distribution system
resilience. Int J Electr Power Energy Syst 124. https://doi.org/10.1016/j.ijepes.2020.106355
38. Agrawal P, Kanwar N, Gupta N, Niazi KR, Swarnkar A (2021) Resiliency in active distribution
systems via network reconfiguration. Sustain Energy Grids Netw 26. https://doi.org/10.1016/
j.segan.2021.100434
39. Kanwar N, Gupta N, Niazi KR, Swarnkar A (2015) Improved meta-heuristic techniques for
simultaneous capacitor and DG allocation in radial distribution networks. Int J Electr Power
Energy Syst 73:653–664. https://doi.org/10.1016/j.ijepes.2015.05.049
40. Salkuti SR (2021) Feeder reconfiguration in unbalanced distribution system with wind and
solar generation using ant lion optimization. Int J Adv Comput Sci Appl 12(3):86–95. https://
doi.org/10.14569/IJACSA.2021.0120304
41. Banarjee S, Chanda CK, Das D (2013) Reconfiguration of distribution networks based on
fuzzy multiobjective approach by considering loads of different types. J Inst Eng (India): Ser
B 94:29–42. https://doi.org/10.1007/s40031-013-0043-2
42. Savier JS, Das D (2007) Impact of network reconfiguration on loss allocation of radial distri-
bution systems. IEEE Trans Power Deliv 22(4):2473–2480. https://doi.org/10.1109/TPWRD.
2007.905370
Role of Advanced Control Technologies
in the Evolution of Smart

Raseswari Pradhan

Abstract In power system, only electric grid means to a vast and variable setup of
electrical components that distribute electrical energy from the generating locations
to the user end. The components are transmission buses, substations, feeder buses
etc. The electric grid is deliberated to be a technological prodigy in handling so
many generating units, high megawatts of generating capacity and several miles
of transmission lines. However, in recent scenario the electricity disruption like a
blackout is very common not in India but in advanced countries like United State
of America. A grid would be more efficient when more resiliency is added into the
existing electric network and made ready for diffident unavoidable tragedies and
natural calamities. Therefore, if some extra powerful features are supplemented to
the existing grid, then it becomes a smart grid. The extra powerful features can
be advanced control techniques application in grid operation. This will make the
existing grid network more efficient, faster in power transmission and self-repair
after power disturbances, inexpensive, improved security etc. This chapter presents
a discussion on the various challenges in execution of smart grid. After that, some
of recent proposed control techniques designed and applied to handle those issues.
Various control applications in smart grid are supplemented with appropriate test
bench problem and experimental or simulation results.

Keywords Smart grid · Definition · Evolution · Challenges · Control

technologies · Communication technologies · Commercial applications

Nomenclature

IC Integrated-Circuit
IoT Internet of Things
PHEV Plug-in Hybrid Electric Vehicles

R. Pradhan (B)
Department of Electrical Engineering, VSSUT, Burla, Odisha 768018, India
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 377
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6_16
378 R. Pradhan

CPS Cyber Physical Systems

RFID Radio Frequency Identification
AMI Advanced Metering Infrastructure
MPPT Maximum Power Point Tracking
V2G Vehicle-to-Grid
PMU Phasor Measurement Units
GPS Global Positioning System Satellite
IT Information Technology
GIS Geographic Information System
ERP Enterprise Resource Planning
EAM Enterprise Asset Management
CRM Customer Relationship Management
SCADA Supervisory Control and Data Acquisition
DMS Distribution Management System
OMS Outage Management Systems
DSM Demand Side Management
S/S Sub-Station
WAMS Wide Area Measurement System
GPP Generator Protection Panels
DGT Directorate General of Training
UPS Uninterruptible Power Supply
ATS Automatic Transfer Switch
TVSS Transient Voltage Surge Suppression
PSG Peer-to-Peer Semantic Grid
GE General Electricals
NIST National Institute of Standard and Technology

1 Introduction

Electricity is one of the greatest findings of the nineteenth century as all other scientif-
ically, socially and economically based revolutionary advancement depends on elec-
tricity only. Therefore, in recent scenario, electricity became a fundamental require-
ment for human being. This is the energy that can be generated and transmitted easily
and economically over long distance. It is also easier to transform from this form to
other form of energy. However, the entire process of utilization of electricity from
generation to used end need a very complex structure of network which is known as
electric grid [1].
An electric grid is well-known as an electrical-network of multiple number of
well-synchronized generating stations and feeder stations. In between them, a large
number of transmission and distribution buses are constructed for efficient trans-
mission from generation to consumers. Again, the whole complicated structure is
Role of Advanced Control Technologies in the Evolution of Smart 379

controlled by one or more control centres. These are definitely some massive inter-
connected physical networks. Also, electrical grid is the backbone infrastructure for
electricity industry [2]. According to various report, the existing grid is not up-to-
date to meet demand of current twenty-first century [3]. Still, the structure of the
existing grid is so large that it is a very matter of concern to manage, to maintain and
to look after too so that it efficiently works [4]. Also, it is difficult and expensive to
coordinate between generation and user ends so as less energy wastage will be there.
When intelligence is incorporated into the existing massive network of electrical
grid, it becomes smart grid [5].
Smart grid provides the opportunity for small scale distributed generation fulling
the local electricity demands. This helps in establishing partial or full independence
from centralized grid network. So, these types of grids are easy to control and main-
tain. Further, in case of occurrence of any kind fault in the system, it can be isolated
from the main grid by intentional islanding process. Since the size is small, its
components can be placed in proper locations to have efficient working environment.

2 Generalized Definition of Smart Grids

It is a highly adopted research topic in current days. Hence, numerous types of defini-
tion are available in literature. In some papers, smart grids are defined as the electrical
systems with state-of-the-art observers, controllers and communiqué machineries to
convey trustworthy and secure form of electricity. This technology has enhanced
the efficacies of both generators and feeders. They too provide flexible choices for
consumers. In other words, smart grid is an amalgamation of the existing complex
physical network system and new advanced cyber network systems [6]. Here, phys-
ical network system serves as the skeleton whereas cyber network system is its brain
and nervous system to deal the technological challenges of the grid.
In different parts of world, smart grid has different meanings. In USA, smart
grid is usually means to the renovation of the electric sectors from a centralized and
service-provider-controlling system to that of largely consumer collaborative system
[7]. In Europe, it means the public participation in broad ways and also amalgamation
of all its belonging nations [8]. Smart grid in China refers a safer, reliable, fast
responding, sustainable and inexpensive physical network [9]. Recently, the Chinese
Government is paying more attention to building new structures and improving the
existing infrastructure in order make the system more consumer-centric and market-
oriented [10].
Power network is very vast and hence its grid structure is also very large having
many sections like generation, transmission, distribution etc. Smart grids are usually
enabled with recent advanced technologies like Integrated-Circuit (IC) techniques in
all of these sectors. Therefore, smart grid is continuously becoming more inexpensive,
competent and sustainable. For further improvement in it, there is requirement of
following a universal standard. For grid, that bench-mark level is delivered by U.S.
National Institute of Standards and Technology. According to it, there are seven
380 R. Pradhan

number of vital sections such as customers, functioning, amenity supplier, generation,

conduction and feeder.
Figure 1 presents a generalized layout of a smart grid. In this figure, it is shown
as the amalgamation of a large number of conventional power plants, nonconven-
tional plants, micro-grids, large demand sectors, small residential consumers, energy
storage technologies and smart grid managing them [1]. According to IEEE Grid
Vision 2050 [11], massive emphasis is required from smart grids on bidirectional
flow of management in operations and control.
A Smart grid is becoming stronger and powerful by employing different kind of
advanced inventive technologies and facilities. Some of them are listed as follows.
• Advanced generation technologies for all sizes of generators
• More involvement of consumers in optimized operation of the system
• Making the system more consumer-centric
• Enhancing self-recovery capability of the whole electricity supply system
• Increase in reliability and security of electricity supply
• Taking care of the standard, government regulation and market as well commercial
requirement.

Solar firm Fuel cells

Communication

Renewable power generating

plants

Microgrid
DG plant
Wind firm

Hydel plant Residential loads

Capacitor
banks
Smart Grid Industrial loads

Battery banks Consummers

Storage

Commercial loads
Coal power
plant

Conventional power plants

Nuclear plant Electric Vehicles

Transmission Unit

Fig. 1 Layout of a smart grid [1]

Role of Advanced Control Technologies in the Evolution of Smart 381

Table 1 Comparison of
Parameters Conventional grid Smart grid
conventional grid with smart
grid [8] Generation Centralized Distributed
Operation Electromechanical Digital
Grid fault Failures and Adaptive and
conditions blackouts islanding
Real-time Nil Extensive
observing
Response Slow Very fast
Data lost Manual restoration Self-healing
Communication One way Two ways
Back ups No storage of energy Present
Control Utility based A large part is
consumer based

Table 2 Per hour revenue

Industry Amount (in thousands of
generated in various
dollars)
industries of USA
Cellular communications 41
Telephone ticket sales 72
Airline’s reservation system 90
Semiconductor manufacturer 2000
Credit card operation 2580
Brokage operation 6480
Source US Department of Energy

Table 1 expresses a comparative study of the conventional electric grid and that
of the smart grid. According to department of Energy of USA, the average revenue
wasted if there is power interrupt in the country for one hour as shown in Table 2.
This justifies smart grid introduction in power sectors.

3 Evolution of Smart Grids

Conventional grids are basically designed with exhaustible sources like coal, gas,
petroleum product etc. Due to massive use of these commodities, their quantities are
decaying and for next stock, thousands of years will be required. Again, they envi-
ronmental impact too. Therefore, trends have been inclined towards use of cleaner
and inexhaustible energy sources like wind, solar etc. However, there is need of
making changes to the existing grid for this purpose and hence evolution of smart
grid happened.
382 R. Pradhan

Rendering to a report of World-Energy-Outlook [12] in the year 2014, the energy

requirement globally will be increased up to nearly 40% by 2040. In that situation,
it is difficult to maintain the standard of energy without any harm to environments if
conventional grids are continued to be used. Even though renewable energy sources
are abundantly available but their harvesting technologies is complicated. Hence,
there is requirement of advanced technologies as well as government support for
efficient, reliable and secured harvesting of these sources. Also, there is need of
proper methods, support and policies for their integration with grid. In this regard,
globally many countries including India have their non-renewable ministries to make
new energy policies and incentives for adapting those policies.
In the USA, these kinds of initiation have been started in the tenure of President
Obama and became doubled by 2008. Report is there that enhancement in imple-
mentation of sustainable energy sources will be around 20% by 2020 in both USA
and Europe [13, 14]. In this regards China is also not far behind. It has set target
of around 15% by 2020 and 86% by 2050 [15, 16]. However, for these types of
implementation smart grids are required.
Smart grids are designed and commissioned in such a way that both supply as
well as usage would be more reliable, proficient, secured, cost-effective and viable
to environment. Their primary target is to integrate intelligence to the working of all
of its components from supply side to consumer side including the transmission and
distribution lines. The better they integrate intelligence to physical world, the more
will be their success. For intelligence integration, usually cyber systems are used.
They primary jobs are sensing, processing and collecting information for intelligent
management and control of the grid. Therefore, a new technology has been immerged
that is the cyber–physical system (CPS). This system is capable of better handle of
integration and interaction issues of physical systems and cyber systems in smart
grids [17].

4 Challenges in Implementation of Smart Grids

Implementation of CPS technologies in smart grids will no doubt strengthen them.

However, new challenges arise due to the use of CPSs handling the systems unusual
characteristics.
The acceptation of a smart grid in future will be decided by certain special features
like compact, flexible, portable, safe and secure etc. For this there is always a require-
ment of good interaction between its various stakeholders in physical worlds (power
network and consumers) and cyber worlds as shown in Fig. 2. However, it has many
technical, social and environmental challenging factors while its implementation [9].
Role of Advanced Control Technologies in the Evolution of Smart 383

Fig. 2 Interaction between

physical world and cyber Cyber

Communication
system Actuators
World

Network
Actuating signals

Sensing signals
Communication
Network
Real
Physical Sensors
World

4.1 Technical Challenges

In this kind of grid, so many types of technical challenges are there to be attended. One
of such condition is irregularity in some generating sources like the renewable energy
generations such as solar plant, wind farm, etc. In this case, it supplies substandard
quality of electricity. Further, in some of the huge power networks including smaller
dispersed energy sources, such as, solar plant, battery banks, wind-firms, plug-in
hybrid electric vehicles (PHEVs) suffer unpredictable operation because of various
conflicting energy market policies. Another important technical factor is the variance
of peak and average electricity demands of an area. It may be some cases present
where only slight reduction in peak demand may accommodate the present power
generation capacity. Therefore, there is no need of building new power plants. But,
for some cases, the difference between peak and average electricity demand is so
high that introduction of some new power plants is unnecessary for feeding these
peak loads. A smart grid needs to handle these kinds of situations in multi area power
networks. In these kind of girds, wastage of electricity in transmission can be reduced
by implementing distributed generation concept. In consumer sides also efficiency
can be enhanced by introducing various advanced technologies such as usage of
smart measurement, sensors, transmitters and processors. There is requirement of
several vital techniques for the above matters. Some are listed as follows.
• Necessity in scattered control machineries with high grade of communication
network. Here, sources, loads, and storage units can enjoy self-control or local-
control not always the centralized control.
• Need of approximately correct prediction of load side demand well before in
advance.
• Requirement of accurate estimation of renewable energy sources generation taking
account the weather forecasting. So that the grid operation would be smooth even
if the intermittent natures of these energy sources.
• Enhancement of grid efficiency by reducing the peak demand by implementing
smart load management, active tariff and power-cut if required.
384 R. Pradhan

• Prerequisite of efficient and smart energy storage equipment that helps in cutting
down the peak demand to some extent, shave peaks in energy demands [18].
To deal with all the above discussed factors, it needs completely systematic and
dedicated steps of involvement while doing modification to the existing grid to make
it smarter. CPS are capable of such kind of paradigm. Figure 2 is showing such type
network that relate physical world with cyber world.

4.2 Environmental Challenges

Although CPS has great help in maintaining smartness of a smart grid is, there are
some environmental issues exist in implementation of smart grids due to this CPS.
However, even if these issues can’t be isolated from each other. They are interrelated
to one and another. These issues are now tried to be discussed precisely as follows.

4.2.1 Eco System

Eco system is means to be a natural community [19]. Development of a smart grid

should not hamper the eco system of its surrounding areas. Decisions in these devel-
opments need to be taken considering the area’s ecological, social and financial
prospects. These decisions must be very future-perspectives without neglecting the
present scenario. The approach needs to be always a cross-functional type approach.
Taking an example; as in a grid with thermal power plant, the details of systematic
functioning need to be considered like expenditures, profits, influence coal mining,
transportation, burning, waste dumping, pollution control and other side-effects to
surrounding environment. Eco system of a smart grid has many numbers of connected
and interconnected networks starting from raw resources to demand side consumers
involving many physical, cyber and social–economic systems.
Individual life cycles of each element of such a large-scale system are to be consid-
ered during the designing stage itself. The ecosystem of a smart grid always requires
an internal feedback mechanism that would monitor its self-regulation, regaining
after disturbances and controlling power. Smart grid ecosystems must be structured
as interaction between ecological, societal and commercial systems as shown in
Fig. 3.
In this figure, power system is the physical world that is required to be is the
components whose physical behaviours need to be supervised and controlled. Here,
the cyber system denotes to intelligence involvement by implementing the innovative
software and hardware for better data-exchange between various physical compo-
nents of the systems. The society and economy systems mean the involvement of
consumers in grid operation, tariff and service provide. Eventually, the term nature
and environments mean to surrounding nature of a grid. Grids as well as nature
Role of Advanced Control Technologies in the Evolution of Smart 385

Fig. 3 Layout showing

energy ecosystem Cyber

k on

Co
or ti
World

m
w ca

m two
et u n i

Ne
un rk
m

ic
om

at
N

io
C

n
Environment
Consumers

Co
m

n
m two

or tio
Ne
un rk

tw ica
ic

k
Ne u n
Power

at
io

m
Grids

m
Co
have great impact on each other. The arrows demarcate the communication networks
enabling data flows between each unit of this grid eco system.

4.2.2 Big Data

This term is usually applied in the data assembly and its analysis. It has five significant
features like bulk size, speed, authenticity, alteration, and cost [20]. Data can be
captured by using smart meters as smart meters have these features. In smart system,
it needs to efficient manage a huge bulk of data related to consumers. Data analysis
needs to handle real time scrutiny of such massive data collected by smart meters
from both utilities and users. It must be fast but authenticate. It also capable of
handle variation in data but must not be very costly. This is more appropriate in
case of sensors and any such kind of devices generated information. For analysis
of these data some advanced analytical algorithms are required. However, these
analytical algorithms need large sets of available information. In smart grids, data
can be collected either conventionally from industries or with new technologies like
web-cams, climate predicting systems, atlases, charts and images [21]. The issues in
big data are related with smart architecture and design of a smart grid having faster
information technologies and specialized computation [20].

4.2.3 Cloud Computing

Real-time, fast and smooth management of big data is necessary in smart grids. For
this work, there is a need of efficient tool for storage, computation and complex
webbing of computing resources. Cloud computing is such a very latest paradigm
service provider these packages [21]. Some of its remarkable benefits are self-service,
flexibility in usage, resource sharing etc. Some other advantages are easy supervision,
386 R. Pradhan

enhance in revenue by lessening cost, continuous services, tragedy controlling and

intelligence in analysis of data [22].
The first cloud-based measurement device was introduced in Denmark in small-
scale. Next, some other countries have also implemented it. In England, a Smart
Energy Cloud system is use that is supporting the countries’ cloud-based measure-
ment device program. Report is there that the number of units will cross beyond
100 million by 2020. Again, this technology is showing tremendous results like
more correct billing, smarter functioning and visibility. Therefore, cloud computing
platforms has a great future in power sector. There are a number of cloud computing
platforms established and in service in power systems. IBM Core metrics and Google
Big Query are two big players in this field.
However, there are some risks in involving cloud computing in service like issues
of security and privacy. This is because the cloud computing service providers are
exposed to much confidential information of the consumers and may misuse them.
Therefore, there is requirement of bound the service provider with some regulatory
constraints [23]. But these issues should not affect the operation of the grid.

4.2.4 Internet of Things (IoT)

In IoT, the term “things” involves some entities like physical uniqueness, physical
characteristics, virtual behaviours and intelligence of users. IoT is the extension of
the Internet amenities to physical world like smart grid. In this recent service, the
conventional Internet service is made enriched by applying some new expertise like
RFID, sensors, smart devices [24]. IoT is growing very fast. Smart grid consists of
a large communication network that link all power apparatus from the transmission
and distribution sectors. This includes meters handling electricity, water, gas, heating
systems and building automation systems. Efficient operations of the smart grid can
be accomplished by using the global standards for IoT. Using IoT, smart grid becomes
more linked, profitable and intelligent. Recently, there is beginning of a new era of
power grid that is Internet of Energy. It consists of intelligent transmission and
distribution of electricity smarter substation and well-organized feeder. The future
of IoT lies in designing a secure and closer to real-time responses. Because, current
existing structure of IoT is prone to cyber-attacks and poor qualities of responses. For
effective architecture and design of smart grid, there is need of effective and optimal
interaction and communication between all stakeholders of a smart grid [25].

4.2.5 Networking

Networking is becoming very important in recent era in every sectors so in smart

grid too. In smart grid many complex networks are involved like communication
networks, living systems, societal networks, cyber networks and physical electrical
networks [26]. This gives diversity to the grid. But, due to this many problems too
arises as follows.
Role of Advanced Control Technologies in the Evolution of Smart 387

• The dynamics of these networks are simplified to make them linearized for better
analysis. However, they are physically far away from the real network.
• The links are assumed to be static where as in actual case they may be dynamic
too.
• Smart grids have heterogeneous combinations of many dynamic, static, nonlinear
models interconnected to each other whereas the networking systems represented
as having homogeneous components.
• Power networks may have different functional layers involving telecommuni-
cation, smart measuring units and feeder management systems etc. There are
always issues involved in prioritizing the components of each functional layers
and dealing with their problems while functioning.

4.2.6 Legislation and Regulation

As discussed before, to secure the personal data of consumers’ and prevention from
misuse of them, there is a high demand of development some standard legislations and
regulations across the globe. Currently, the practice is that different countries have
set their own rules and regulation that different from others. So, there is no uniformity
in implementation of constraints and barriers on numerous analyses [27]. The vital
issues in application of smart grid are mostly cyber related. Extend of involvement
of the stakeholders is one of the major issues. Similarly, incentive related problem is
another issue. Legitimate barriers in developing and regulating smoothly are some
other important issues too.
Since, the power sectors and organizations are different country to country. There-
fore, legal and regulatory systems too are different. Therefore, cyber issues are
required to be addressed accordingly following the respective system of each country.
Smart grid standardization is necessary in order to make it uniform worldwide.
IEEE has a significant part in this field. IEEE has over 100 Smart Grid-related
approved standards. It also has several Smart Grid-related pending standards [3].
Some of the standards in Communication Protocols are as follows [28].
• From one Regulatory body to other, i.e., IEC 60870-6/TASE.2
• From one Regulatory body to site apparatus, i.e., IEEE 1815 (DNP3) for USA
and IEC 60870-5 for UK
• Site apparatus, i.e., IEC 61850 and IEEE 1815 (DNP3).

5 Control Issues in Implementation of Smart Grids

The vital control issues in the operation of a smart grids are involved in strengthening
the grid-structure, improvement in its intelligence, communications, integration of
various intermittent generating units, moving offshore, harvesting the goodness of
388 R. Pradhan

Renewable power
generating plants

Solar Wind
firm firm
Residential loads
Hydel Fuel
plant cells
DG
plant

Power Substations

Industrial loads
Nuclear plant

Controlling
Transmission Centre
Coal power Commercial loads
plant
Unit

Power Substations

Generation Unit
Consummers

Fig. 4 Control structure in smart electrical grid and its interfaces

distributed generation and storage units. The control structure of a smart grid is
shown in Fig. 4.

5.1 Reinforcement of the Grid

The existing conventional grid is very old. Further, they are very huge and compli-
cated. Again, the grid intimately related to social and economic growth of an area.
Therefore, it is required to ensure enhancement in grid power by integration of
renewable resources, power handling capabilities and control.

5.2 Improvement in Intelligence

FREEDM substations are established to improve the stake of sustainable energy.

They involve large number of advanced and prototype components. The prospective
grids are going to be different from that of the existing grids in terms of organization
and application. Figure 4 shows such a futuristic set-up of grid. It comprises mostly
distributed types of generation and storage systems. This trademarks the grid more
flexible and consistent in operation. This system uses a large scale of power elec-
tronics devices in order to make them automatic, adaptive and superiority in energy
management and power quality improvement [29].
However, in these kinds of set-up, the current flow during fault condition is
very small. Hence, fault can’t be detected using that of conventional fault detection
Role of Advanced Control Technologies in the Evolution of Smart 389

methods. Further, source and load sides are not symmetrical in nature. Therefore,
normal symmetrical based analysis is not applicable to these systems. The traditional
protection systems are also not appropriate for these systems [30].

5.3 Betterment in Communications

In a smart grid, usually almost all of its components are linked by communication
network for the effective exchange of information. Correct and fast exchange of infor-
mation is necessary for quick response of the devices in certain situations. Feeder
systems are basically constructed taking care of a fixed toleration level in voltage and
current in normal operating conditions. Beyond that the protection systems become
active. In a smart grid, the exchange of information about any contingency situation
and required command after decision making to the respective protection devices
must be faster and smoother. This can be possible by using two-way digital commu-
nication like wired phones and mobile phones, radios, Fiber-optical-devices, Power
line transporter transmission, Satellite and web [31].

5.4 Participation of Intermittent Sources

Cost is one of the main factors in all commercial sectors including the power sectors.
Therefore, optimal economic power dispatch of various available generating units
is necessary [32]. To solve this problem, calculation of optimal load sharing of the
generating station is required. Previously, only conventional sources were present.
Therefore, this problem was simpler dealing with only stock of fossil fuels and their
cost. Recently, more and more intermittent and distributed sources are getting added
to the system. So, the problem becomes more complicated. The challenges lie in
location of power plants away from load centers. So, it will be not economical to
install communication networks too [31].
Different distributed generation technologies are available such as wind-plant,
solar arrays etc. It is always necessary to select right one among these technologies for
a particular location for better performance. In addition to that some supplementary
technologies need to be implemented to the existing systems like use of Maximum
Power Point Tracking (MPPT) in wind and solar plants to minimize energy losses
[33].

5.5 Offshore Movement

This term is mostly applied in case of wind farms. Now-a-days, it is applied in

smart grids as wind farms are part of them. Therefore, smart grid technology is
390 R. Pradhan

now capable of controlling and operating remote plants efficiently and optimally.
Such an innovative technology is proposed in paper [34]. It can enable a grid to
tackle uncertainties in monitoring and maintenance equipment of a power plant. An
adaptive condition-based maintenance scheme is discussed here.

5.6 Capturing the Benefits of Back-Up Devices

As discussed in the previous sections, some of the generating systems of a smart grids
are intermittent in nature basically the renewable energy-based systems. This may
inject unwanted disturbances to the stable grid and through it to other components of
the grid. Therefore, a disturbance created in a small part of grid may imbalance the
whole grid causing black-out like situations. Therefore, there is need of some back-up
devices to be installed with the other generating systems such as battery banks. These
devices are usually installed at the distribution level. Now-a-days, back-up devices
are very essential elements in off-grid application. In grid integrated application
also, back-up devices are installed in order to enhance the system robustness against
intermittent supply of energy. Apart from batter banks, diesel generators, flywheels,
fuel cells, ultra-capacitors etc. are also very popular back-up systems in power sectors
[32]. The combination of generating systems with back-up systems are known as
hybrid systems.
However, the back-up systems must be selected, designed and commissioned in
such a way that the quality of power and voltage level will not be hampered both in
short and long terms. In addition to it, stability of the whole system will be retained
even with bulk injection of back-up powers. A vast research is going in this area.
However, those are not sufficient and competent enough to use commercially [33,
34]. Different hybrid systems have different issues and hence they are required to be
addressed differently. For an example, in wind farm-based hybrid systems, back-up
devices are designed to handle intermittence generation as well as ramp rates that
limits its generation level. Similarly, in solar based hybrid systems, both intermittence
and power angle correction factors are vital issues [35].

5.7 Plug-In Hybrid Vehicles Usage

Smart grids are characterized with distributed generations and dispersed consumers.
A smarter grid needs to address them all proficiently. Electric vehicles have a great
future and their injection volumes are increasing. However, they are highly movable
and discrete in nature. The new technology to it is the plug-in hybrid vehicles and
known as vehicle-to-grid (V2G). V2G does not means only its utilization as means
of transportation but also its usage of electric-rechargeable batteries. For batteries
recharging, grid support is necessary. Therefore, in near future, they will emerge as
a major challenge for grid network, the future electricity networks. V2G is still not
Role of Advanced Control Technologies in the Evolution of Smart 391

commercially famous. To make it more reliable, more research is required. In some

of the researches [36], V2G systems are developed with in a confined area of grids
called micro grids.

5.8 Advanced Technologies Implementation

Recently, a number of advanced technologies are designed and formulated to be

applied in different section of a power network like monitoring, sensing, measure-
ment, data acquisition, control etc. Phasor measurement units (PMUs) are such an
advanced technology. These are well-synchronized with the real-time power grid
network and widely used as measuring devices. Their advanced characters are very
beneficial in system observation, protection and controlling. Here, synchronization
is accomplished by maintaining equal sampling frequency of voltage and current
waveforms using Global Positioning System Satellite (GPS) [37].

6 Digital Control of Smart Grids

Digital control is usually a combination of energy and information technology

(IT). Introduction of digital control in power sector has enabled it with two-way-
communication between consumer and energy provider. This technology is expected
to come to picture in 1980s. However, 2015 onwards its utilization paradigm has
covered smart grid, smart homes, digital billing, distributed generation, maintenance
services, smart interconnection etc. The utilization of digital control in power sector
is very vast. Some of its distinct components are shown in Fig. 5. It consists of smart
grid, smart homes, renewable energy sources, consumer engagement, distribution
intelligence and plugin-vehicle.
Recently, digital control technology offers smart grid solutions from generation
sites to the utility sides of electrical energy. Its services include protection and control
of all kind of generator sets of the grid. It also ensures safety and smooth operation
of a hybrid generation system with renewable energy sources. It facilitates contin-
uous transfer to power backup units in order to provide uninterrupted power supply.
Therefore, using digital control technologies, power grid reliability is enhanced. In
future, digital control technologies will play big roles in smooth functioning of a
smart city. Figure 6 shows how it can help in the operation of smart city.
392 R. Pradhan

Renewable
Energy
Source Smart
Residential
Intelligent Projects
Distribution

Digital
Technology
Electrical Vehicles
Smart Charging Stations
Grids
Consumer
Participation

Fig. 5 Components of digital control in power sectors

Miscellaneous
Service
Management

E-Governance Waste
and Human Management
Management

Smart
City

Traffic
Efficient Management
Power Supply
Management
Water
Management

Fig. 6 Future of digital control in smart city

Role of Advanced Control Technologies in the Evolution of Smart 393

7 Adaptive Control Technologies for Smart Grid

With adaptive control technology, an advanced distribution system can be managed.

Advanced distribution system can manage both demand and efficiency more profi-
ciently. This can be possible due to use of adaptive optimization techniques and grid
improvement tasks. The usage of adaptive enable the grid to analyze and manage
distributed energy resources. It also makes them automatic, self-healing and self-
sustaining. Advanced distribution system further supports closed-loop type control
structure. Here, the operator sects the adaptive schemes studying the forecasted grid
conditions data collected using advanced data acquisition devices.
A layout of an advanced distribution system based adaptive control system is
shown in Fig. 7. Its advanced management system takes care of real-time services
like sub-station (S/S) automation, feeder automation and functioning of intelligent
devices. It also looks after enterprise services like GIS, ERP, CRM and EAM. In this

ERP CRM GIS EAM

Smart Technology Service Bus

OMS

Digital DSM
DMS Technology

SCADA

Real World Application Service Bus

Intelligent Intelligent Smart

Substation Feedder Devices

Fig. 7 Paradigm of adaptive control technologies in smart grid

394 R. Pradhan

management system, various advanced technologies have been used like SCADA,
DMS, OMS and DSM [38].
Adaptive control technology has future in smart grid in the following fields.

7.1 Advanced Metering Infrastructure (AMI)

It serves as an interface between the grid and consumers. The information exchange
is bidirectional in nature. It is enabled with accurate load data acquiring, advanced
control, real-time tariff and outage detection. In AMIs, data acquisition interval is
programmed to be at 5, 15, 30 or 60 min of time intervals depending upon the types
of loads. It has a Home Automation Network (HAN) gateway. It also includes remote
connect/disconnect switch.

7.2 Phasor Measurement Unit (PMUs)

PMUs are controlled by GPS. They are specifically synchronized with the real-time
system voltage and current waveforms. By this they can remotely measures data of
a grid by catching its electrical waves. It is very effective in fault detection without
even physically attached to the network.

7.3 Distributed Weather Sensing

It helps in gathering weather data like cell temperature, solar irradiance, wind speed,
etc. in case of renewable energy-based hybrid systems. There by the power outage
can be forecasted in advanced and required modification in grid are carried out. It is
an effective tool for efficient power management [39].

7.4 Wide Area Measurement System (WAMS)

In this system, a chain of PMUS is made in service. This network provides continuous
real-time monitoring and data acquisition for a wide range of areas.
Role of Advanced Control Technologies in the Evolution of Smart 395

8 Adaptive Control Technologies for Smart Grid

As discussed earlier, communication between different sections of a smart grid very

vital. Therefore, there is requirement of innovative communication skills. Classifi-
cation of these communication technologies has been done depending upon their
utilization. They are as follows.

8.1 Integrated Communications

Any interference in this network may severely affect its quality of wireless sensing
capabilities. Various features of integrated communications are fast, fully unified
and bidirectional data exchange. So that the smart grid becomes vibrant and
collaborative [40].

8.2 Secure Web Mechanism

The advanced technologies must be protected, secured and reliable [41].

8.3 Wireless Sensor System

These systems must handle severe environments, packet faults, changing link
capability, constrained in resources and maintain required standard.

8.4 Communication Infrastructure

In this system, a chain of PMUS is made in service. This network provides continuous
real-time monitoring and data acquisition for a wide range of areas.
These are used as the back-bone of a communication network. Some of the
components are given as follows.
• System Integrators
• WAN Network
• LAN Network
• Data Storage Management and Analytics
• CRM.
396 R. Pradhan

Hydel plant DG plant Solar firm Wind firm Fuel cells

Local Data
Concerntrator

Local Data
WAN Concerntrator

Local Data
Concerntrator

HAN HAN HAN HAN

Energy transmission lines

Residential Commercial Microgrid Industrial
Loads Loads Control Loads Data communication lines
Units

Fig. 8 Advanced communication technologies in smart grid

The future advanced communication technologies in smart grid are shown in Fig. 8
[42].

9 Some Commercially Available Smart Grid Technologies

Some leading companies like GE are involved in this field. Some of its products are
shown in Table 3.

Table 3 Various commercial smart grid based products by GE

Protection and control Distributed grid management Micro grid control
Generator protection: Wireless trip control: Optimized generation and grid
G60 DGT wireless transfer trip interconnect relays:
G30 F60
489 C30
Generator protection panels: Advanced load shedding and Load management/load
GPP restoration: shedding relays:
C90Plus C90Plus
F35
Plant system optimization: Quick standby devices: Wireless LAN extension:
PMCS UPS Mercury
C90Plus ATS iNET-II
TVSS entra NET
PSG
Source GE Technologies
Role of Advanced Control Technologies in the Evolution of Smart 397

Hydel plant DG plant Solar firm Wind firm Fuel cells

Residential Commercial Microgrid Industrial

Loads Loads Control Loads
Units

Fig. 9 Layout of GE Microgrid Control System. Source GE Technologies

The GE Microgrid Control System offers an advanced technique in which different

power generation and feeder data are accumulated efficiently. Then that data is
utilized to formulate the optimal cost-effective scheme so as to provide uninterrupted
supply. It also rich with the presence of controllable and optimized generating units,
storage units and delivery of clean power. A layout of GE Microgrid Control System
is shown in Fig. 9.
GE is also providing solution for advanced digital communication from generation
to utility sectors to the user ends. This is possible by a fast and secured two-way
wireless network. It can send quick trip signals to dispersed generation units. Then,
acknowledge report after work done is also sent back for confirmation. This kind of
wireless arrangement cuts costs and ensured safety (Fig. 10).
Many other smart grids related commercial products are available across the
world. In USA, Plug-in Electric Vehicles are in high demand and are being sold
very fast. It is expected that by 2020, the sold unit will cross beyond one Million.
Two big companies have been ventured in this field such as Nissan Leaf and Chevrolet
Volt.
Since, smart grid is growing day by day in USA, Major government and local
agencies are involved in its smooth operation and control like National Institute of
Standard and Technology (NIST), LOS Alamos, Edison, Oak Ridge, NSF, Ercot etc.
Back up technologies also very vital in case of smart grid operation to support
during peak load period. For back up batter bank is an important component. There-
fore, many companies are serving in sectors like Ktech Corp, California’s Central
Valley etc.
For transmission systems, Midwest Energy Inc. Project, Hays, KS are providing
solution. They are supplying nine relay-based PMUs and synchro phasor commu-
nications network for watching power-angle and signal-frequency. It also analyses
disturbances and resulting faults, monitors signals, their stability status etc. Similarly,
SGIG Distribution Automation Project, Atlantic City, NJ; Golden Spread Electric
Cooperative Project in Amarilo, TX; Whirlpool Corporation Smart Grid Project,
398 R. Pradhan

Cyber

5.
id
World

ili G
of to d

Tr by
gr
er ted an
ut D

i p re
ty
i tt a m

co ce
m nic m
ns u co

m ive
gr al

co se
id
ili ign

m r
7. irm to s
tra mm Trip

an of
nf nt
Tr ati ub
ut s

d DG
ty

ip on s
by firm
co 4.

re
pi

ce
ng co atio
ed n
iv co

iv
ac mm n

ed
ce ip

tio an
by

re Tr

n d
t
8.
3. Fault DG
Power Grid Circuit
occurrence site
substation confirmed 6. Ac breaker
9. Tripping confirmed tuate
to d
by utility grid open
circu

Power supply
2. Fault occurrence it
informed to substation

Transmission Lines

1. Electrical Fault occurrence

Communication Network
Power Transmission Network

Fig. 10 Wireless infrastructure solution by GE. Source GE smart grid https://www.gegridsoluti

ons.com/multilin/smartgrid_innovations.htm

Benton Harbor, MI; Peak Energy Agriculture Rewards (PEAR) in Fresno, CA and
AMI Project in Lubbock, TX are some other famous companies that have ventured
to provide smart grid solution commercially.

10 Conclusion

This chapter gives an overview about different aspects of a smart grid. It has presented
that smart grid technology has eradicated various age-old problems like incorporated
controlling, vast capacity sources, dependency on unsustainable sources etc. It also
problems in automation increasing responsiveness from consumers. With the help
of better communication systems, now smart grid is being powered with two-ways
smooth and authenticate data flows. A comparison has made with that of the conven-
tional to justify its superiority. This chapter also has a brief discussion on different
challenges and issues faced by smart grid like technical, ecological, social, control,
protocols etc. It has also discussion on on-going expansion and future technolo-
gies of smart grid like smart city applications. Further this chapter has elaborated
usage of adaptive control techniques and advanced communication technologies to
make smart grid more intelligent and user-end friendly. It is also discussed with
some IEEE and other universal standards, how this technology is made universally
Role of Advanced Control Technologies in the Evolution of Smart 399

uniform, secured and protected. It is seen that smart grid offers a very vast area of
research. This chapter can be a very good reference for the aspirant researchers and
engineers working in various field of smart grid.

References

1. Meng W, Ma R, Chen HH (2014) Smart grid neighborhood area networks: a survey. IEEE
Netw 28:24–32. https://doi.org/10.1109/MNET.2014.6724103
2. Gungor VC, Sahin D, Kocak T, Ergut S, Buccella C, Cecati C, Hancke GP (2011) Smart
grid technologies: communication technologies and standards. IEEE Trans Ind Inf 7:529–539.
https://doi.org/10.1109/TII.2011.2166794
3. Wang W, Xu Y, Khanna M (2011) A survey on the communication architectures in smart grid.
Comput Netw 55:3604–3629. https://doi.org/10.1016/j.comnet.2011.07.010
4. Wu FF, Moslehi K, Bose A (2005) Power system control centers: past, present, and future.
Proc IEEE 93:1890–1908. https://doi.org/10.1109/JPROC.2005.857499
5. Tushar W, Yuen C, Saha TK, Morstyn T, Chapman AC, Alam MJE, Poor HV (2021) Peer-to-
peer energy systems for connected communities: a review of recent advances and emerging
challenges. Appl Energy 282:116131. https://doi.org/10.1016/j.apenergy.2020.116131
6. Cecati C, Mokryani G, Piccolo A, Siano P (2010) An overview on the smart grid concept.
In: IECON 2010—36th annual conference on IEEE industrial electronics society, Nov 2010.
IEEE, pp 3322–3327. https://doi.org/10.1016/j.asej.2020.05.004
7. Gunduz MZ, Das R (2020) Cyber-security on smart grid: threats and potential solutions.
Comput Netw 169:107094. https://doi.org/10.1016/j.comnet.2019.107094
8. Narayanan SN, Khanna K, Panigrahi BK, Joshi A (2019) Security in smart cyber-physical
systems: a case study on smart grids and smart cars. In: Smart cities cybersecurity and privacy.
Elsevier, pp 147–163. https://doi.org/10.1016/B978-0-12-815032-0.00011-1
9. Jha AV, Appasani B, Ghazali AN, Pattanayak P, Gurjar DS, Kabalci E, Mohanta DK (2021)
Smart grid cyber-physical systems: communication technologies, standards and challenges.
Wireless Netw 27:2595–2613. https://doi.org/10.1007/s11276-021-02579-1
10. Xin C, Chen X, Chen H, Chen S, Zhang M (2020) Green product supply chain coordination
under demand uncertainty. IEEE Access 8:25877–25891. https://doi.org/10.1109/ACCESS.
2020.2963944
11. Stoustrup J, Annaswamy A, Chakrabortty A, Qu Z (2019) Smart grid control, vol 10. Springer
International Publishing, pp 978–983. https://doi.org/10.1007/978-3-319-98310-3
12. US Energy Information Administration (ed) (2011) Annual energy outlook 2011: with
projections to 2035. DOI-DOE/EIA-0383. Government Printing Office
13. Lins C (2010) Renewable energy technology roadmap: 20% RES by 2020. In: Local govern-
ments and climate change: sustainable energy planning and implementation in small and
medium sized communities, vol 39, p 131. https://doi.org/10.1007/978-1-4020-9531-3
14. Koronen C, Åhman M, Nilsson LJ (2020) Data centres in future European energy systems—
energy efficiency, integration and policy. Energy Effic 13:129–144. https://doi.org/10.1007/s12
053-019-09833-8
15. Tan Q, Yu S (2021) Dynamics of renewable energy in China: drivers and challenges. In:
Renewable energy transition in Asia: policies, markets and emerging issues, vol 1, pp 113–135.
https://doi.org/10.1007/978-981-15-8905-8_6
16. Hai T, Sharafati A, Mohammed A, Salih SQ, Deo RC, Al-Ansari N, Yaseen ZM (2020) Global
solar radiation estimation and climatic variability analysis using extreme learning machine
based predictive model. IEEE Access 8:12026–12042. https://doi.org/10.1109/ACCESS.2020.
2965303
400 R. Pradhan

17. Sun D, Li W, Yao X, Liu H, Chai J, Xie K, Feng L (2021) Research on IoT architecture
and application scheme for smart grid. In: Proceedings of the 9th international conference on
computer engineering and networks, vol 1. Springer, Singapore, pp 921–928. https://doi.org/
10.1007/978-981-15-3753-0_90
18. Bullich-Massagué E, Cifuentes-García FJ, Glenny-Crende I, Cheah-Mañé M, Aragüés-Peñalba
M, Díaz-González F, Gomis-Bellmunt O (2020) A review of energy storage technologies for
large scale photovoltaic power plants. Appl Energy 274:115213. https://doi.org/10.1016/j.ape
nergy.2020.115213
19. Marrone M, Linnenluecke MK, Richardson G, Smith T (2020) Trends in environmental
accounting research within and outside of the accounting discipline. Account Audit Account J
33:2167–2193. https://doi.org/10.1108/AAAJ-03-2020-4457
20. Lytras M, Visvizi A, Zhang X, Aljohani NR (2020) Cognitive computing, big data analytics
and data driven industrial marketing. Ind Mark Manag 90:663–666. https://doi.org/10.1016/j.
indmarman.2020.03.024
21. Talaat M, Alsayyari AS, Alblawi A, Hatata AY (2020) Hybrid-cloud-based data processing
for power system monitoring in smart grids. Sustain Cities Soc 55:102049. https://doi.org/10.
1016/j.scs.2020.102049
22. Priyanka EB, Thangavel S, Gao XZ (2021) Review analysis on cloud computing based smart
grid technology in the oil pipeline sensor network system. Pet Res 6:77–90. https://doi.org/10.
1016/j.ptlrs.2020.10.001
23. Hashmi SA, Ali CF, Zafar S (2021) Internet of things and cloud computing-based energy
management system for demand side management in smart grid. Int J Energy Res 45:1007–
1022. https://doi.org/10.1002/er.6141
24. Rath M, Tomar A (2021) Smart grid modernization using Internet of Things technology. In:
Advances in smart grid power system, vol 1. Academic Press, pp 191–212. https://doi.org/10.
1016/B978-0-12-824337-4.00007-2
25. Babu RG, Glorindal G (2021) Big data security using IoT-based smart grid communications.
In: Cognitive informatics and soft computing. Springer, Singapore, pp 391–401. https://doi.
org/10.1007/978-981-16-1056-1_32
26. Ahmadzadeh S, Parr G, Zhao W (2021) A review on communication aspects of demand
response management for future 5G IoT-based smart grids. IEEE Access. https://doi.org/10.
1109/ACCESS.2021.3082430
27. Lopes JAP, Madureira AG, Matos M, Bessa RJ, Monteiro V, Afonso JL, Magalhães P (2020)
The future of power systems: challenges, trends, and upcoming paradigms. Wiley Interdiscip
Rev Energy Environ 9:e368. https://doi.org/10.1002/wene.368
28. Gunduz MZ, Das R (2020) Cyber-security on smart grid: threats and potential solutions.
Computer Netw 169:107094. https://doi.org/10.1016/j.comnet.2019.107094
29. Nazir M, Enslin JH, Burkes K (2021) Hybrid bypass protection of hybrid smart transformers
for advanced grid support. In: 2021 IEEE 12th international symposium on power electronics
for distributed generation systems (PEDG). IEEE, pp 1–5. https://doi.org/10.1109/PEDG51
384.2021.9494212
30. Wu Y, Wu Y, Guerrero JM, Vasquez JC (2021) Digitalization and decentralization driving
transactive energy Internet: key technologies and infrastructures. Int J Electr Power Energy
Syst 126:106593. https://doi.org/10.1016/j.ijepes.2020.106593
31. Li Y, Wang T, Wang S (2020) Cost-efficient approximation algorithm for aggregation points
planning in smart grid communications. Wireless Netw 26:521–530. https://doi.org/10.1007/
s11276-019-02152-x
32. Arya A, Sisodia S, Mehroliya S, Rajeshwari CS (2020) A novel approach to optimize bidding
strategy for restructured power market using game theory. In: 2020 IEEE international sympo-
sium on sustainable energy, signal processing and cyber security (iSSSC). IEEE, pp 1–6. https://
doi.org/10.1109/iSSSC50941.2020.9358902
33. Das L, Munikoti S, Natarajan B, Srinivasan B (2020) Measuring smart grid resilience: methods,
challenges and opportunities. Renew Sustain Energy Rev 130:109918. https://doi.org/10.1016/
j.rser.2020.109918
Role of Advanced Control Technologies in the Evolution of Smart 401

34. Rathor SK, Saxena D (2020) Energy management system for smart grid: an overview and key
issues. Int J Energy Res 44:4067–4109. https://doi.org/10.1002/er.4883
35. Wang Y, Das R, Putrus G, Kotter R (2020) Economic evaluation of photovoltaic and energy
storage technologies for future domestic energy systems—a case study of the UK. Energy
203:117826. https://doi.org/10.1016/j.energy.2020.117826
36. Das S, Acharjee P, Bhattacharya A (2020) Charging scheduling of electric vehicle incorporating
grid-to-vehicle and vehicle-to-grid technology considering in smart grid. IEEE Trans Ind Appl
57:1688–1702. https://doi.org/10.1109/TIA.2020.3041808
37. Siamak S, Dehghani M, Mohammadi M (2020) Dynamic GPS spoofing attack detection, local-
ization, and measurement correction exploiting PMU and SCADA. IEEE Syst J 15:2531–2540.
https://doi.org/10.1109/JSYST.2020.3001016
38. Duan P, Soleimani H, Ghazanfari A, Dehghani M (2020) Distributed energy management in
smart grids based on cloud-fog layer architecture considering PHEVs. IEEE Trans Ind Appl.
https://doi.org/10.1109/TIA.2020.3010899
39. Al Kadhimi BM, Mohammed AB (2020) The aspect of power quality in a wind generation:
overview. J Phys Conf Ser 1530:12104. IOP Publishing
40. Shi L, Crow ML (2008) Power and energy society general meeting-conversion and delivery of
electrical energy in the 21st century. IEEE, Pittsburgh
41. Ari AAA, Ngangmo OK, Titouna C, Thiare O, Mohamadou A, Gueroui AM (2020) Enabling
privacy and security in Cloud of Things: Architecture, applications, security & privacy
challenges. Appl Comput Inform. https://doi.org/10.1016/j.aci.2019.11.005
42. Li C, Wang Q, Jing Z, Zhu H, Wang P, Yu C (2021) A method of data transmission technology
for line condition monitoring in smart transmission grid. E3S Web Conf 252:01002. EDP
Sciences. https://doi.org/10.1051/e3sconf/202125201002
Application of Solar Energy
as Distributed Generation for Real Power
Loss Reduction in Radial Distribution
Network

Aliva Routray, Khyati D. Mistry, and Sabha Raj Arya

Abstract This work discusses the integration of solar energy as Distribution Gener-
ation in radial distribution systems. It improves the system reliability and integrity by
minimizing the losses. To generate solar power, daily solar radiation data are gener-
ated using obstructed astronomical model to get accurate radiations as solar panel
input. The global radiation is modelled with obstructed astronomical technique, self-
shadowing effect, and photovoltaic temperature. The hourly data distribution pattern
of solar radiation is non-parametric. This issue proceeds to the implementation of
Kernel Density Estimation techniques in this work. The solar energy as DG is imple-
mented to IEEE-15 and IEEE-28 bus distribution networks separately. The optimum
location and capacity of DG in the distribution system are determined by employing
Grey Wolf Optimization (GWO) technique.

Keywords Distributed Generation · Grey Wolf Optimization · Radial distribution

system · Self-shadowing effect · Temperature model

Nomenclature

RDS Radial Distribution System

DG Distribution Generation
GMA Greenwich Meridian Angle
AST Apparent Solar Time
STC Standard Test Condition
NOCT Nominal Operating Cell Temperature
KDE Kernel Density Estimation
MCS Monte Carlo Simulation
PDF Probability Density Function
RTS Reliability Test System

A. Routray · K. D. Mistry · S. R. Arya (B)

Department of Electrical Engineering, Sardar Vallabhbhai National Institute of Technology Surat,
Gujarat 395007, India

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 403
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6_17
404 A. Routray et al.

GWO Grey Wolf Optimization

PV Photovoltaic
Po Power Generated from PV Module
Yo Solar module’s rated capacity (kW) at STC
IGeff Solar Irradiance (kWh/m2 )
IGeff,STC Solar Irradiance at STC (kWh/m2 )
K Temperature coefficient
To PV Cell Temperature (°C)
To,STC PV Cell temperature at STC (°C)
Tamb Ambient temperature (°C)
To,NOCT PV Cell temperature at NOCT (°C)
Tamb,NOCT Nominal atmospheric temperature (°C) at NOCT (°C)
ïo,STC Panel Efficiency at STC
τo Solar transmittance
Ao Solar absorptance
IGeff,NOCT Solar radiation at NOCT (kWh/m2 )
dS Distance between consecutive panels
a Azimuth angle of the shadow
LSh Length of the shadow
Ang Solar altitude angle

1 Introduction

The transmission and distribution systems are operated with heavy loading condi-
tions. The distribution system is the bigger point of concern as it stays close to the
consumer side. Enhanced Efficiency and power quality are the parameters those need
to provide a favorable economic environment for the minimization of power losses
in the system [1]. Several approaches for the loss minimization problem have been
identified in many existing works of literatures. These methodologies are different
from each other in terms of employed method, problem formulation, loss mini-
mizing tool and the obtained solution. Capacitor placements, network reconfigura-
tion, Distributed Generation (DG) allocation, feeder grading, high voltage distribu-
tion, usage of power electronics devices etc. are some of the methods for power loss
minimization. However, few techniques like capacitor placement, network reconfigu-
ration and DG allocation are more often used for reducing the line losses by improving
the efficiency of distribution system [2]. There are many benefits of DG integra-
tion such as exploitation of renewable energies, decreased environmental impacts,
improved energy efficiency, good congestion management, bus voltage increment
and power loss reduction in the distribution system [3]. The DGs are also econom-
ical because of less maintenance cost. Moreover, this methodology of DG penetration
is comparatively more effective and very much beneficial than the other methods of
loss minimization. The main reason for the extensive popularity of DGs is, their
Application of Solar Energy as Distributed Generation for Real … 405

green technology and environmental concern for the world. Apart from it, the pene-
tration of DGs into the distribution system is widely increased as it can be used
as a small-scale power generation source and can be easily placed to the consumer
side [4]. DGs can use renewable as well as non-renewable power sources for power
generation.
Solar energy as DG has the potential to meet the energy needs and to achieve the
environmental goal. Much research has been carried out on the penetration of solar
energy as DG into the distribution system because of clean energy [5]. Photovoltaic
incorporation is the medium for solar energy to serve commercial and residential
customers. The construction of solar panels is taken place from the fundamental
unit of PV cell which is doped with silicon semiconductor. The PV cell forms the
module and further the modules form the PV array. The PV cell absorbs the solar
radiation and the electricity is produced in DC form [6]. The solar energy can be
stored for future use as solar cell only operates during the day time. Apart from the
solar power generation, PV-based energy also takes part in space programs as it sends
power to the satellite equipment, direct application for home lighting, water pumping,
road lighting etc. in remote areas and supplies excessive load demands during the
peak hour of the summer days. However, the penetration level potentially creates
challenges in fields of reliability, power quality, over-voltage issues and protection
due to the intermittent energy produced during the daytime.
The PV application area is the vast area of study which can be the current research
areas of many researchers. Solar energy can be integrated into the grid with an accept-
able level of penetration else it can cause an adverse effect on the entire system. The
adverse impact can be a reverse power flow that damages the transformer operation.
When properly sized and placed, the PV system can help the distribution system in
minimizing the feeder loss. In literature [7], solar panels are utilized on the rooftops to
provide additional energy for domestic application. Optical allocation of solar energy
as DG using any optimization technique is widely implemented in the distribution
system. In [8], solar-based DGs are used in the distribution system and the optimal
allocation is found by bat algorithm. In [9], the PV-DG placement is accomplished
for assessment of PV potential in a distribution system using optimization technique.
At different load levels, the performance of the distribution system is studied using
PV-based DG [10]. A novel optimization technique is implemented in [11], to a
distribution system for optimal solar energy as DG allocation. The results with DG
connection in the distribution system are obtained to have low power losses in the
lines and bus voltage improvement. Resiliency analysis of a distribution system with
solar energy as DG has been carried out in literature [12]. Demand-side management
using solar energy as DG for a micro-grid is studied in [13]. Active power, reactive
power and the bus voltages values are considered after solar integration onto the
distribution system and the stability of the system is studied [14].
Solar energy is used as DG for active power loss reduction which is the primary
objective of this work. In this work, hourly solar radiation is formed from the Kernel
Density Estimation followed by Monte Carlo Simulation. The obtained radiation
is implemented through several astronomical models using obstruction and scaling
factors. The self-shadowing effect and ambient temperature effect are added to the
406 A. Routray et al.

solar farm to get the accurate solar radiations. The hourly load demand is considered
during the Distribution System Load Flow analysis. The position and capacity of
solar farm as DG are decided by applying Grey Wolf Optimization technique. The
IEEE-15 and IEEE-28 bus standard distribution systems are studied for the loss
minimization analysis using solar energy. The power losses as well as the bus voltages
are determined with DG and the results are analyzed.

2 Effect of External Parameters on Solar Radiation

A few challenges take place in the process of solar power generation such as solar
radiation data generation using measuring tools or data prediction with algorithms,
the effect of temperature and other climatic factors on the PV panel efficiency, grid
integration and its impact on power electronics components etc. Extensive active
researches are carried out on these challenging areas.

2.1 Irradiance Astronomical Model

The available global radiation is termed instantaneous solar radiation (IGsol ). An

irradiance astronomical model is used for converting the instantaneous radiation
into effective radiation (IGeff ) falling on PV panels [15]. The effective solar radiation
depends on many parameters such as solar declination angle (), latitude of installed
PV panel’s location, tilt angle of the panel, day of the year, hour of the day etc. [6,
16, 17]. The sinusoidal relationship of solar declination angle for spring equinox day
(D = 81) can be expressed as given in Eq. (1).

(D) = 23.45 sin(F(D)) (1)

where (D) is the solar declination angle, (D) p [−23.45°, 23.45°] and F(D) =
360(D − 81)/365. The equation of time (T(D)) is expressed in Eq. (2) which gives
the time difference between clock time and solar time.

T (D) = (9.87 sin 2F(D) − 7.53 cos F(D) − 1.5 sin F(D)) ∗ (1/60) (2)

Again the time difference (TM ) between the installed solar panel time zone and
the reference Greenwich Meridian Angle (GMA) can be given as in Eq. (3).
◦
TM = (L o −G M A)/15 (3)

where Lo is the longitude of the location, GMA = UTCoff . 15° and UTCoff is the
time offset between Greenwich and time zone of installed solar panel location. Now
Application of Solar Energy as Distributed Generation for Real … 407

the Apparent Solar Time (AST) is calculated as given in Eq. (4) and its time in hour
range belongs to [0, 24].

AST (t, D) = t + T (D) (4)

The hour angle is the Hr(t, D) is determined as,

Hr (t, D) = 15( AST (t, D)−12) (5)

Here, t is the clock time of the day in hour. Now the angle of incidence (Ai )
between the solar panel and the solar radiation can be generalized as given in Eq. (6).

CosAi (t, D) = sin (t, D) × sin L A × cos φ − sin (t, D) × cos L A

× sin ϕ × cos η + cos (t, D) × cos L A × cos ϕ
× cos Hr (t, D) + cos (t, D) × sin L A × sin φ × cos η
× cos Hr (t, D) + cos (t, D) × sin φ × sin η × sin Hr (t, D) (6)

where LA is the latitude of the location, φ is the inclination angle of the solar panel
towards the sun, and η is the solar azimuthal displacement. The incidence angle Ai (t,
D) determines the effective solar irradiance as given by Eq. (7).

IGe f f (t, D) = IGsol (t, D) × cos Ai (t, D) (7)

The solar declination angle and incidence angle for the month of June is deter-
mined using the above equations and can be observed from Fig. 1. However, the
cosine of incidence angle is considered for 1st June and implemented in this chapter.
The astronomical model helps in tracking the incidence angle being a function of
time. Similarly, the incidence angle can be calculated for any day with all the solar
hours in a year. The solar declination angle is required to determine the effective solar

Fig. 1 Solar declination angle and cosine of incidence angle for the month of June
408 A. Routray et al.

radiation from the global radiation fall on the earth’s surface. This accurate radiation
on the PV panel gives the real power output obtained from it helps to realize the
system practically.

2.2 Obstructed Solar Astronomical Model

The solar panel efficiency is practically hindered with objects like buildings, trees,
birds, weather etc. These are known as the obstructions on the way of solar power
production. In the solar astronomical model these factors are not taken into consider-
ation. Hence, the calculations lack precision in solar radiation with large deviation.
This astronomical model is always an advantage for actual radiation fetching. It uses
multiplying factors to determine the closest solar radiation that falls on the PV panels
[18]. In this work, the radiation values are fetched from the installed solar farm in
Bagalkot, Karnataka, India during the month of June, 2017. The global solar radia-
tion can be obtained for any location from the NASA web application [19]. These
radiation values are modelled using obstruction factor and the effective radiation is
used in further work.
Here, G1(t, D) is the radiation obtained from the basic irradiance model, G2(t,
D) is the radiation obtained after using the scaling factor in astronomical model,
G3(t, D) is an effective radiation obtained after using obstruction factor, M(t, D)
is the measured value of radiation from the farm and O(t, D) be the obstruction
factor. Scaling factor is the ratio of measured solar radiation data to the radiation
formed from the solar astronomical model. Mathematically, the scaling factor (SF)
is expressed as,

M(t, D)
SF = (8)
G1(t, D)

Now, the obstruction factor can be defined as the ratio of the difference between
radiations obtained by using scaling factor and the global solar radiation with respect
to radiation value obtained after using scaling factor. Mathematically, the expression
is given in Eq. (9).

G2(t, D) − M(t, D)
O(t, D) = (9)
G2(t, D)

The relation between scaled radiation and obstructed radiation is given in Eqs. (10)
and (11).

G2(t, D) = S F × G1(t, D) (10)

G3(t, D) = G2(t, D) × (1 − O(t, D)) (11)

Application of Solar Energy as Distributed Generation for Real … 409

Fig. 2 Irradiance data at Bagalkot, Karnataka

The solar insolation is calculated from the astronomical model and is compared
with the measured radiation values from the respective solar farm. From Fig. 2, it can
be observed that, there is a significant difference between both the radiation values.
The solar insolation is obtained for all 30 days of June 2017. However, 1st June is
further considered for the purpose of analysis.
The obstruction factor is useful for obtaining the closest solar radiation value
on PV panels for any random day. The scaling and obstruction values are 0.9227
and 0.0773 respectively. Using these factors, the analysis is extended for effective
radiation calculation in June 2019.
The final solar irradiance value is obtained after executing the radiation from
NASA with the several astronomical models. The radiations that are determined using
these models are given in Table 1. The relative difference between the radiations in
each astronomical modelling can be observed in Fig. 3.
The obstructed model helps in finding the practical insolation data for the respec-
tive solar farm location. After using the scaling and the obstruction factors with the
astronomical model, from Table 1, the radiations for 1st June with basic, scaled, and
obstructed models are found to be 4.858 kW/m2 , 4.483 kW/m2 and 4.136 kW/m2
respectively. The daily average solar insolation is calculated from the above values
and it is found to be decreased by 14.863% after implementing the obstructed astro-
nomical model. The solar power is obtained from the radiation obtained using the
obstructed astronomical model and employed for further study.
410 A. Routray et al.

Table 1 Effective solar radiation from different models in June 2019

Days in From Cosine of Using Using scaled Using
June NASA(kW/m2 ) incidence astronomical astronomical model obstructed
angle model (kW/m2 ) (kW/m2 ) astronomical
model
(kW/m2 )
1 6.23 0.7799 4.858777 4.483194 4.1366
2 6.33 0.7795 4.934235 4.552819 4.2009
3 6.05 0.7777 4.705085 4.341382 4.0058
4 5.56 0.7773 4.321788 3.987714 3.6795
5 6.38 0.7756 4.948328 4.565822 4.2129
6 6.13 0.7752 4.751976 4.384648 4.0457
7 4.77 0.7736 3.690072 3.404829 3.1416
8 5.02 0.7731 3.880962 3.580964 3.3042
9 6.31 0.7716 4.868796 4.492438 4.1452
10 6.32 0.7696 4.863872 4.487895 4.141
11 6.38 0.7692 4.907496 4.528147 4.1781
12 4.46 0.7678 3.424388 3.159683 2.9154
13 4.06 0.7673 3.115238 2.87443 2.6522
14 5.93 0.7569 4.488417 4.141462 3.8213
15 6.09 0.7548 4.596732 4.241405 3.9135
16 5.87 0.7528 4.418936 4.077352 3.7622
17 5.53 0.7509 4.152477 3.831491 3.5353
18 6.16 0.749 4.61384 4.25719 3.9281
19 5.7 0.7472 4.25904 3.929816 3.626
20 5.18 0.7523 3.896914 3.595683 3.3177
21 5.49 0.7502 4.118598 3.80023 3.5065
22 2.48 0.7483 1.855784 1.712332 1.58
23 4.66 0.7469 3.480554 3.211507 2.9633
24 4.96 0.7463 3.701648 3.415511 3.1515
25 5.21 0.7447 3.879887 3.579972 3.3032
26 5.16 0.7445 3.84162 3.544663 3.2707
27 4.88 0.7427 3.624376 3.344212 3.0857
28 3.12 0.7425 2.3166 2.137527 1.9723
29 3.46 0.741 2.56386 2.365674 2.1828
30 2.96 0.7404 2.191584 2.022175 1.8659
Application of Solar Energy as Distributed Generation for Real … 411

Fig. 3 Solar irradiance values with different models in June-2019

2.3 PV Cell Temperature Model

Ambient temperature significantly affects the solar power generation from PV

modules [6, 20]. The increase in the ambient temperature, decreases the efficiency
of solar panel output. The analysis is carried out in this subsection for the same
location and month. A solar panel is designed to give maximum efficiency when it
is operated in Standard Test Condition (STC). The PV generated solar panel output
can be expressed as,

IGe f f  
Po = Yo × × 1 + κ × (To − To,ST C ) (12)
IGe f f,ST C

Mathematically, the solar cell temperature is calculated as given in Eq. (13),

  
IGe f f ηo,ST C (1−κ×To,ST C )
IGe f f,ST C
To,N OC T − Tamb,N OC T 1 − τo ×Ao
To = Tamb +   IGe f f  κ×ηo,ST C  (13)
1 + To,N OC T − Tamb,N OC T IGe f f,N OC T τo ×Ao

The solar panel which is considered in this work, has a peak power of 335 W.
The electrical datasheet for the solar panel is collected from Ref. [21]. The electrical
data for PV solar cell at Standard Test Condition (STC) and Nominal Operating Cell
Temperature (NOCT) is mentioned.
The temperature of the atmosphere for June 2019, is extracted from the recorded
past weather data [22]. The PV cell temperature is determined using the above equa-
tions with available solar irradiance data. Here the solar absorptance and solar trans-
412 A. Routray et al.

mittance values are taken as 0.6 and 0.9 as per the best classification [20]. The
ambient temperature is less than the solar module temperature and can be observed
from Fig. 4.
The temperature model is implemented to the astronomically generated solar
radiations. The effective radiations is reduced after considering PV temperature and
astronomical model as shown in Fig. 5. To meet the customer demand requirement
in any distribution system, a number of solar panels can be installed in a solar farm.
As per the daily average solar radiation value and PV cell temperature value, the
effective solar power is generated and fed to the distribution system as DG.

Fig. 4 Daily average ambient temperature and solar module temperature

Fig. 5 Hourly solar irradiance with astronomical model and temperature effect
Application of Solar Energy as Distributed Generation for Real … 413

2.4 Panel Size and Location

PV panels always work in stand-alone mode for any grid operation. The number of
solar panels or the size of the farm depends on the degree of maximum electrical load
connected to the system. A radial distribution system consists of load buses except
for one source node as generating node. The solar farm power is injected into any
of these load buses depending on the total consumer demand and DG constraint. In
this work, the solar farm location is considered to be Bagalkot, Karnataka, India.
The solar power penetration is assumed to be 30% of the total load demand of the
distribution system [23]. Hence, the number of solar panels are calculated as per the
load demand requirement.

2.5 Shadowing Effect in Solar Panels

Solar panels are installed in a certain location forming a solar farm. The inter-
row disctance between two consecutive solar panel rows is maintained to avoid the
shading of solar panels. This shading of solar panels is known as the self-shadowing
effect. This effect prevents unshaded solar cells from operating at maximum power.
In this section, a preventive measure has been taken to eliminate the power loss
caused by the shadowing effect. An illustration of the self-shadowing effect is given
in Fig. 6. The installation of solar panels has specific mounting guidelines as the
length of the uppermost tip and the lower most tip of the solar panel which are 3 m
and 2.168 m respectively [24]. The distance between the consecutive solar panel
rows is calculated using Eqs. (14)–(17) [25].

d S = L Sh × cos(a) (14)

3
L Sh = (15)
tan(ang)

Fig. 6 Solar panel dimension and self shadowing of consecutive panels

414 A. Routray et al.

Table 2 Number of solar panels for different test systems

Test system Total load 30% of load Solar farm area (m2 ) No. of
demand (kW) demand (kW) Without With required
self-shadowing self-shadowing solar
panels
IEEE-15 1266.4 380 242 941 350
IEEE-28 761.05 230 403 1568 210

sin(ang) = cos(L A ) × cos() × cos(t) + sin(L A ) × sin() (16)

cos() × sin(t)
a= (17)
cos(ang)

The panel elevation angle is assumed to be 30° with respect to the horizontal
surface [25]. Hence the horizontal component of the panel becomes 0.866 times
the original panel length. Now the effective length becomes 1.44 m. The effective
spacing is calculated using the above equations and found to be 4.14 m [25]. The test
systems with total load demand and number of solar panels needed after considering
the self-shadowing effect are given in Table 2.

2.6 Stochastic Modelling of Solar Irradiance

The solar radiation data are needed to find the solar power output. With the high uncer-
tainty and non-parametric behaviour of hourly solar radiation pattern, the stochastic
method is implemented to generate the hourly radiation dataset. The process needs
very large amount of data to come out with better results. The solar database resources
are found in NREL website which provides vast information regarding solar radiation
for several locations in the United States [26]. In this work, hourly solar radiation data
is considered for over 10 years that is from 1991 to 2010. Kernel Density Estimation
is implemented to the uncertain radiation data as the Probability Density Function
of the data cannot be identified using available standard distribution functions [15].
After finding the Probability Distribution Function for solar radiation data, Cumu-
lative Probability Distribution is determined and therefore generating the random
radiation samples for application in Monte Carlo Simulation. The hourly histogram
plots for all the solar hoursare obtained and given in Figs. 7, 8, 9, 10, 11, 12, 13, 14,
15, 16, 17 and 18.
The Kernel Density Function uses the following equations. Here, PDFHr is the
hourly probability function, RRn is the random variable, KDE(.) is the Kernel
Function Estimator, ss is the sample size and sp is the smoothing parameter.
Application of Solar Energy as Distributed Generation for Real … 415

Fig. 7 Histogram and probability density function of solar irradiance data at 7.00 a.m.

Fig. 8 Histogram and probability density function of solar irradiance data at 8.00 a.m.

ss  
1 R Rn − R Rni
P D FHr (R Rn ) = K DE (18)
ss × sp i=1 sp

Solar radiation is generated using the above method for location Surat, Gujarat,
India on an hourly basis and given in Table 3. In this chapter, the solar radiation is
considered as a daily average radiation and is fetched from NASA for the specified
location as mentioned before.
The radiation data of those having more than the threshold radiation value (ITH =
50 Wh/m2 ) are considered for power calculation. The radiations below the threshold
value are assumed to be zero.
416 A. Routray et al.

Fig. 9 Histogram and probability density function of solar irradiance data at 9.00 a.m.

Fig. 10 Histogram and probability density function of solar irradiance data at 10.00 a.m.

3 Load Modelling Profile

The hourly analysis of active power loss in distribution lines is carried out in this
chapter. In this work, a constant load model is taken into consideration for hourly
power flow calculation in the distribution lines. The percentage of loads for every
weekday or weekend and every hour is different. This percentage load profile data
is considered from IEEE-Reliability Test System (RTS) [27]. The hourly average
load percentage values for a summer season are proposed with the time-varying
solar power. The load values are mentioned in percent for weekday hourly load and
Application of Solar Energy as Distributed Generation for Real … 417

Fig. 11 Histogram and probability density function of solar irradiance data at 11.00 a.m.

Fig. 12 Histogram and probability density function of solar irradiance data at 12.00 p.m.

weekend hourly load as given in Table 4. In this chapter, peak of load demand is
considered for the study that is at 12.00 p.m. of the day.
The tabulation gives the information about the hourly peak load in the percentage
of daily peak load. Since for weekdays all the peak percentage of loads are equal,
the power or energy consumption graph for all the five days is similar. Saturday
and Sunday are the weekend days. The hourly peak values are different from the
weekdays load values. The PV generation is independent and the power generation
values are obtained from the solar irradiance values which is distinct at the different
times of the day and position of the sun as well as the rotation of the earth.
418 A. Routray et al.

Fig. 13 Histogram and probability density function of solar irradiance data at 1.00 p.m.

Fig. 14 Histogram and probability density function of solar irradiance data at 2.00 p.m.

4 Problem Formulation

The objective function of the work is to determine the active power losses in the
distribution system using DG. Here bus number varies from i to nn. Few operational
constraints are to be satisfied to ensure system integrity. The power balance equation
at bus ‘i’ after DG is connected to the grid is given in Eq. (19).

Px DGi = Pxi − Px Di (19)

Application of Solar Energy as Distributed Generation for Real … 419

Fig. 15 Histogram and probability density function of solar irradiance data at 3.00 p.m.

Fig. 16 Histogram and probability density function of solar irradiance data at 4.00 p.m.

Here, PxDGi is the power generation from DG, Pxi is the bus power generation and
PDi is the power supplied to the loads connected at the nodes at bus ‘i’ respectively.
Mathematically, the objective function in this work can be expressed as,

nn
Total Power Loss = Pxi Loss (20)
i=1

The constraints to the objective function are described below.

420 A. Routray et al.

Fig. 17 Histogram and probability density function of solar irradiance data at 5.00 p.m.

Fig. 18 Histogram and probability density function of solar irradiance data at 6.00 p.m.

Constraints of Equality
The power losses in the lines and the power consumed by the consumers as loads
must be equal to the sum of power generation from the grid as well as from the DG.
The power balance equation is given by,

Px G + Px DG = Px D + Px L O SS (21)

Here, PxG is the power generated from the grid, PxD is the consumed power by
the consumers, PxDG is the power generation from the DG source and PxLOSS is the
losses in the distribution lines.
Application of Solar Energy as Distributed Generation for Real … 421

Table 3 Hourly generated

Sl. No. Time Solar radiation (Wh/m2 )
solar irradiance
Using Monte Carlo Using astronomical
simulation modeling
1 1.00 a.m. 0 0
2 2.00 a.m. 0 0
3 3.00 a.m. 0 0
4 4.00 a.m. 0 0
5 5.00 a.m. 0 0
6 6.00 a.m. 64 5.6576
7 7.00 a.m. 251 76.6554
8 8.00 a.m. 483 238.5537
9 9.00 a.m. 700 448.77
10 10.00 a.m. 861 634.6431
11 11.00 a.m. 974 755.0448
12 12.00 p.m. 1026 772.3728
13 1.00 p.m. 1010 749.925
14 2.00 p.m. 939 728.2884
15 3.00 p.m. 805 602.3815
16 4.00 p.m. 619 409.9637
17 5.00 p.m. 405 212.058
18 6.00 p.m. 188 64.2208
19 7.00 p.m. 33 4.2471
20 8.00 p.m. 0 0
21 9.00 p.m. 0 0
22 10.00 p.m. 0 0
23 11.00 p.m. 0 0
24 12.00 a.m. 0 0

Constraints of Inequality
In every distribution line, there is some limit to the current flow in the line due to the
current-carrying capacity. If Ixmax is the maximum current limit for the respective
line and Ix is the flowing current, then the equation is given as in Eq. (22).

Ix ≤ Ix max (22)

Similarly, the nodal voltage of the system must be in its limitation or tolerance
range. If Vx is the actual voltage at any node and Vxmin is the minimum and Vxmax is
the maximum voltage limits, then the equation can be given as in (23).

|Vx min | ≤ |Vx | ≤ |Vx max | (23)

422 A. Routray et al.

Table 4 Hourly peak load in percentage of daily peak load [27]

Day/time (h) 01:00 02:00 03:00 04:00 05:00 06:00
Weekday 64 60 58 56 56 58
07:00 08:00 09:00 10:00 11:00 12:00
64 76 87 95 99 100
13:00 14:00 15:00 16:00 17:00 18:00
99 100 100 97 96 96
19:00 20:00 21:00 22:00 23:00 00:00
93 92 92 93 87 72
Day/time 01:00 02:00 03:00 04:00 05:00 06:00
Weekend 74 70 66 65 64 62
07:00 08:00 09:00 10:00 11:00 12:00
62 66 81 86 91 91
13:00 14:00 15:00 16:00 17:00 18:00
93 92 91 91 92 94
19:00 20:00 21:00 22:00 23:00 00:00
95 95 100 93 88 80

Constarints of DG
DG is connected as an additional part of the existing system. Hence it also follows
some limitations as, the maximum capacity of DG should be less than its maximum
limits. If PxDG is the DG power generation at any time and PxDGmax is the maximum
power that can be generated using the DG, then the constraint is given in Eq. (24).

|Px DG | ≤ Px DG max (24)

This helps the system to stay stable and maximum benefits can be obtained to the
system with consideration of DG constraints.

5 Optimization Technique

Optimum placement and capacity of solar energy as DG is a very important segment

of the work. To minimize the active power losses in the distribution lines of the
test systems, Grey Wolf Optimization (GWO) technique is implemented in this
work. GWO is a metaphor-based algorithm that imitates the hunting process of
grey wolves to determine the global optimum of complex optimization problems.
Its multiguide and multi-stage mechanism enables it to explore the search space
Application of Solar Energy as Distributed Generation for Real … 423

thoroughly. However, complex computation mechanism and presence of multiple

algorithm-specific parameter (which needs to be tuned) makes its implementation
difficult [28].
The size of a grey wolf pack generally varies from 5 to 12 and every grey wolf
pack follows a very strict social hierarchy. The leader of the pack is termed as L1-
wolf and it takes a decision regarding hunting, which is followed by other wolves
in the pack. In this social hierarchy of grey wolves, the position of L2-wolf lies
just below L1-wolf. L2-wolf plays the role of an advisor for L1-wolf for decision
making. Responsibilities of L2-wolf also include carrying out the command of L1-
wolf throughout the pack and providing feedback to L1-wolf. The lowest rank wolf
in the hierarchy is known as L3-wolf. L3-wolf follows the order of other dominant
wolves. Apart from L1-wolf, L2-wolf and L3-wolf, other wolves in the pack are
called L4-wolf. They obey the orders of L1-wolf and L2-wolf but dominate L3-wolf.
L4-wolves generally carry out the task of watching the boundaries of the pack and
warning the pack of impending danger. It should be observed that every wolf in the
pack contributes to the hunting process, even though the hunting is mainly guided
by L1 and L2-wolves.

5.1 Mathematical Model of GWO

Stage-1: Initialization and Hierarchy Assignment

In this stage, a set of feasible random feasible solutions are generated and their fitness
is evaluated. The solution having the best fitness value is designated as L1-wolf while
solutions having second and third best fitness values are termed as L2 and L3 wolf
respectively. Other solutions in the solution pool are called L4-wolves. In the GWO
technique, generally, search process is guided by L1, L2 and L3-wolves.
Stage-2: Encirclement of Prey
In this stage, grey wolves encircle prey before their hunting. This phenomenon math-
ematically is modeled by positioning solutions (grey wolves) around prey which is
the optimum solution. Mathematical operations are carried in this stage and they are
given in the below equations.
−
→ − →
→ − → −
C 1 = C 2 · PP (t) − P (t) (25)

−
→ −
→ −
→ − →
P (t + 1) = PP (t + 1) − C 3 · C 1 (26)

In these equations, ‘t’ denotes current iteration, PP (t) and P(t) represent the position
vector of the prey and grey wolf respectively. C2 and C3 are coefficient vectors whose
values are determined as given below.
424 A. Routray et al.

−
→
C 3 = 2−
→
c3 · −
→
rr 1 − −
→
c3 (27)

−
→
C2=2·−
→
rr 2 (28)

In the above equations, rr1 and rr2 are random vectors in the range [0, 1], while
components of vector c3 linearly decrease from 2 to 0 with an increase in iteration.
Random vector rr1 and rr2 play a crucial role in positioning the wolves around the
prey.
Stage-3: Hunting
Generally, the hunting process is guided by L1, L2 and L3 wolves. Mathematically,
the hunting process is modeled by adjusting the position of L4 wolves according
to the positions of L1, L2 and L3 wolves. Following mathematical operations are
carried out in this stage.
−
→ − →
→ −−→ −
C 1 L1 = C 21 · PPL1 − P (29)

−
→ − →
→ −−→ −
C 1 L2 = C 22 · PPL2 − P (30)

−
→ − →
→ −−→ −
C 1 L3 = C 23 · PPL3 − P (31)

−
→ −−→ − → −
→
P1 = PPL1 − C 31 · C 1 L1 (32)

−
→ −−→ − → −
→
P2 = PPL2 − C 32 · C 1 L2 (33)

−
→ −−→ − → −
→
P3 = PPL3 − C 33 · C 1 L3 (34)

−
→ − → − →
−
→ P1 + P2 + P3
P (t + 1) = (35)
3
In the above equations, P1, P2 and P3 are the updated positions of L4 wolf P
according to the L1, L2 and L3 wolves. The final position of 4 wolf P is represented
by P(t + 1).
Stage-3(a): Attacking Prey (Exploitation)
When the prey stops moving, the grey wolf attacks the prey and finishes the hunting
process. This is mathematically modeled by linearly decreasing ‘c’ from 2 to 0
with increasing iteration which in turn decreases the range of C3. From the above
equations, it is clear that, C3 lies in the range [−2c, 2c] and ‘c’ generally decreases
from 2 to 0. When the values of C3 lies in the range of [−1, 1], then the updated
Application of Solar Energy as Distributed Generation for Real … 425

position of the search agent will be in between the current position of prey and the
search agent. Hence the value of |C3| < 1 makes the wolves move towards the prey.
Stage-3(b): Searching for Prey (Exploration)
This operator brings diversity to the search process. When C3 lies in the range [1, 2]
and [−2, −1], L1, L2 and L3 wolves diverge in the search space in search of prey.
Hence, |C3| > 1 causes through exploration of search space.

6 Test System Under Study

IEEE-15 [29] and IEEE-28 [30] standard radial distribution systems are examined in
this work. These test systems with single line diagram are mentioned below. These
standard systems are balanced distribution systems.

6.1 IEEE-15 Bus System

IEEE-15 [29] standard radial distribution systems are examined in this work. The
base voltage and base power values for IEEE-15 bus system are 10 kV and 11 MVA.
The single line diagram for these test bus systems is given in Fig. 19.

Fig. 19 IEEE-15 node radial distribution system

426 A. Routray et al.

Fig. 20 IEEE-28 node radial distribution system

6.2 IEEE-28 Bus System

IEEE standard 28 bus [30] radial distribution system is studied and is shown in Fig. 20.
The system base voltage and base MVA of the system are 11 kV and 100 MVA
respectively.

7 Result and Discussion

The test systems are studied with the obstructed astronomical model, temperature
model, self-shadowing effect, load model and optimization algorithm. The solar radi-
ation obtained from the meteorological website is implemented with the obstructed
astronomical model. The power generated from the solar farm is fed to the test
systems as DG. The generated power output is fed to the test distribution system
after considering all the factors. Then the system is run with a power flow algorithm
to find the current flow in the branch, power losses and bus voltage at the nodes.
The proposed power flow technique is executed in MATLAB 2016b. The system
parameter is examined without DG and with DG. The system performance in terms
of power losses, bus voltages and optimization characteristics are given under this
result section.

7.1 Real Power Loss Minimization

The power flow method is implemented and the line losses are found as active
and reactive power losses. With DG penetration, the system losses are remarkably
minimized. After considering the daily peak load on 1st June (Saturday) to be 96%,
the active and reactive power losses are minimized from 56.71 to 34.24 kW and
52.58 to 30.71 kVAR with DG in case of IEEE-15 bus system. Similarly, the active
Application of Solar Energy as Distributed Generation for Real … 427

Table 5 Reduction in power loss minimization with DG in percentage

Date IEEE-15 bus RDS IEEE-28 bus RDS
%age reduction in %age reduction in %age reduction in %age reduction in
Ploss Qloss Ploss Qloss
01-Jun-19 65.45 71.21 143.23 150.34
02-Jun-19 65.37 71.12 142.86 149.96
03-Jun-19 64.67 70.39 139.85 146.86
04-Jun-19 64.59 70.30 139.51 146.51
05-Jun-19 65.3 71.08 142.67 149.77
06-Jun-19 65.62 71.39 144 151.12
07-Jun-19 65.54 71.30 143.6 150.73

and reactive power losses are minimized from 62.98 to 25.89 kW and 42.13 to
16.83 kVAR with DG in case of IEEE-28 bus system. The power losses in the
distribution lines can be calculated for other days with DG penetration. A noticeable
decrement in power losses with DG connection has been observed from the results.
The solar power is extracted from the solar farm and is injected into the test system
from 1st June to 7th June 2019. The percentage reduction in power losses for both
the test systems is given in Table 5.

7.2 Bus Voltage Improvement

The DG injection into the distribution system increases the nodal voltage, being
an additional advantage of DG penetration. The daily bus voltages are found for
the complete week in June for distribution test systems. The bus voltages after DG
penetration have a remarkable effect on bus voltage improvement at the system
nodes. The bus voltage before DG penetration and after DG penetration for 1st June
(Saturday) is given in Tables 6 and 7 for IEEE-15 and IEEE-28 bus respectively.
Likewise, the bus voltages for other days of the week can also be found with DG and
the results are given in Tables 6 and 7 for IEEE-15 and IEEE-28 bus respectively.
The bus voltages are observed to be significantly improved for both the system with
DG allocation.
The systems are realized with peak load and available solar power at the peak hour
of the day. From the availability of hourly DG power, the bus voltages are found at
12.00 p.m. The bus voltages for both the test systems are given in Figs. 21 and 22,
respectively. The voltage profiles are improved with DG connection.
428 A. Routray et al.

Table 6 Nodal voltage with DG connection in IEEE-15 bus system

Bus\days Saturday With DG (p.u)
No DG (p.u) With DG 2nd Jun 3rd Jun 4th Jun 5th Jun 6th Jun 7th Jun
(p.u)
Bus-1 1 1 1 1 1 1 1 1
Bus-2 0.9725 0.9787 0.9791 0.983 0.9835 0.9794 0.9778 0.9782
Bus-3 0.9585 0.97 0.9706 0.9761 0.9767 0.971 0.9687 0.9693
Bus-4 0.953 0.9682 0.9689 0.9747 0.9754 0.9692 0.9668 0.9675
Bus-5 0.952 0.9673 0.968 0.9739 0.9746 0.9683 0.9659 0.9666
Bus-6 0.96 0.9663 0.967 0.9731 0.9738 0.9674 0.9648 0.9655
Bus-7 0.9579 0.9641 0.9649 0.9714 0.9722 0.9653 0.9626 0.9634
Bus-8 0.9588 0.965 0.9658 0.9722 0.9729 0.9662 0.9635 0.9643
Bus-9 0.9693 0.9755 0.9761 0.9805 0.981 0.9763 0.9745 0.975
Bus-10 0.9683 0.9745 0.9751 0.9797 0.9802 0.9753 0.9734 0.974
Bus-11 0.9521 0.9636 0.9644 0.971 0.9718 0.9648 0.962 0.9628
Bus-12 0.9481 0.9597 0.9606 0.9679 0.9688 0.961 0.958 0.9589
Bus-13 0.9469 0.9585 0.9594 0.967 0.9678 0.9598 0.9567 0.9576
Bus-14 0.9508 0.966 0.9668 0.973 0.9737 0.9671 0.9646 0.9653
Bus-15 0.9506 0.9703 0.9709 0.9763 0.977 0.9713 0.969 0.9696

7.3 Optimum DG Location and Capacity

The optimization technique is used to find the optimum DG capacity and its location
at the test system node. Grey Wolf Optimization algorithm is used for minimizing
the line losses in the distribution system. The DG capacity is considered as 30% of
entire load demand of both the distribution systems. The DG location is found to
be at bus number-15 bus number-24 in the case of IEEE-15 and IEEE-28 radial test
distribution systems. The system after combined optimization technique and load
flow technique, the active power losses are minimized and bus voltage is improved
with DG injection.

8 Conclusion

In this work, solar energy is used as a DG to minimize the line losses in the distribution
system. The solar radiation from the meteorological website is modelled with an
astronomical model and obstructed astronomical model. The scaling factor and the
obstruction factor from the obstructed astronomical model are calculated to find the
effective solar radiation fall on the solar panels. This causes the decrease in daily
average solar radiation value to 14.867%. The radiation is further modelled with the
Application of Solar Energy as Distributed Generation for Real … 429

Table 7 Nodal voltage with DG connection in IEEE-28 bus RDS (p.u)

Bus\days Saturday With DG
No DG (p.u) With DG 2nd Jun 3rd Jun 4th Jun 5th Jun 6th Jun 7th Jun
(p.u)
Bus-1 1 1 1 1 1 1 1 1
Bus-2 0.9868 0.9911 0.9913 0.9929 0.9931 0.9914 0.9907 0.9909
Bus-3 0.9679 0.9785 0.979 0.9829 0.9834 0.9792 0.9776 0.9781
Bus-4 0.9544 0.9697 0.9703 0.9759 0.9765 0.9706 0.9683 0.969
Bus-5 0.9409 0.9626 0.9634 0.9703 0.9711 0.9638 0.9609 0.9617
Bus-6 0.9308 0.9578 0.9587 0.9665 0.9674 0.9592 0.9559 0.9569
Bus-7 0.922 0.9555 0.9565 0.9647 0.9656 0.957 0.9536 0.9545
Bus-8 0.9197 0.9532 0.9543 0.9629 0.9639 0.9548 0.9512 0.9522
Bus-9 0.9194 0.953 0.954 0.9626 0.9637 0.9545 0.9509 0.952
Bus-10 0.9192 0.9528 0.9538 0.9625 0.9635 0.9543 0.9507 0.9517
Bus-11 0.9485 0.9638 0.9646 0.9712 0.972 0.965 0.9622 0.963
Bus-12 0.9468 0.9622 0.963 0.9699 0.9707 0.9634 0.9605 0.9613
Bus-13 0.9458 0.9612 0.962 0.9691 0.97 0.9624 0.9595 0.9603
Bus-14 0.9455 0.9609 0.9617 0.9689 0.9697 0.9622 0.9592 0.96
Bus-15 0.9453 0.9607 0.9615 0.9687 0.9696 0.962 0.959 0.9598
Bus-16 0.9398 0.9615 0.9623 0.9694 0.9702 0.9628 0.9598 0.9607
Bus-17 0.9291 0.9561 0.9571 0.9651 0.9661 0.9576 0.9542 0.9552
Bus-18 0.9282 0.9552 0.9562 0.9644 0.9654 0.9567 0.9533 0.9542
Bus-19 0.9266 0.9537 0.9547 0.9632 0.9642 0.9552 0.9516 0.9526
Bus-20 0.9257 0.9529 0.9539 0.9626 0.9636 0.9544 0.9508 0.9518
Bus-21 0.9251 0.9523 0.9533 0.9621 0.9631 0.9538 0.9502 0.9512
Bus-22 0.9193 0.9571 0.958 0.9659 0.9668 0.9585 0.9552 0.9561
Bus-23 0.9178 0.9586 0.9595 0.9671 0.968 0.96 0.9568 0.9577
Bus-24 0.9167 0.96 0.9609 0.9682 0.9691 0.9613 0.9582 0.9591
Bus-25 0.9164 0.9598 0.9606 0.968 0.9689 0.9611 0.958 0.9589
Bus-26 0.9163 0.9596 0.9605 0.9679 0.9688 0.9609 0.9578 0.9587
Bus-27 0.9192 0.9528 0.9538 0.9625 0.9635 0.9543 0.9507 0.9518
Bus-28 0.9191 0.9527 0.9537 0.9624 0.9634 0.9542 0.9506 0.9516

temperature effect to get the actual radiation on PV panels. A Solar farm is formed
after considering the self-shadowing effect and the solar power generated from it is
penetrated to the distribution system as DG. The solar insolation values are considered
for June 2019 in Bagalkot, Karnataka, India. The best location and capacity of DG
are found using the GWO optimization technique to get the most minimum power
losses in the distribution lines. The optimum size of DG is assumed to be 30% of the
total load demand of the test system that is produced by the solar farm. The number
430 A. Routray et al.

Fig. 21 Bus voltage improvement in IEEE-15 RDS

Fig. 22 Bus voltage improvement in IEEE-28 RDS

of solar panels is determined for each test systems under study to match the necessary
load demand. For system analysis, the LIM load flow technique is implemented to
the distribution systems. IEEE-15 and IEEE-28 bus RDS are used as system under
study for the loss minimization using DG. The system parameters like active power
Application of Solar Energy as Distributed Generation for Real … 431

loss and bus voltages with DG are compared with system performance without DG.
From the result, the power losses in both the test systems are significantly decreased
and the bus voltages are improved remarkably with DG placements.

References

1. Nguyen TT (2021) Enhanced sunflower optimization for placement distributed generation in

distribution system. Int J Electr Comput Eng 11(1):10. https://doi.org/10.11591/ijece.v11i1.
pp107-113
2. Kalambe S, Agnihotri G (2014) Loss minimization techniques used in distribution network:
bibliographical survey. Renew Sustain Energy Rev 29:184–200. https://doi.org/10.1016/j.rser.
2013.08.075
3. Rathore A, Patidar NP (2021) Optimal sizing and allocation of renewable based distribution
generation with gravity energy storage considering stochastic nature using particle swarm
optimization in radial distribution network. J Energy Storage 35:102–282. https://doi.org/10.
1016/j.est.2021.102282
4. Oda ES, Abd El Hamed AM, Ali A, Elbaset AA, Abd El Sattar M, Ebeed M (2021) Stochastic
optimal planning of distribution system considering integrated photovoltaic-based DG and
DSTATCOM under uncertainties of loads and solar irradiance. IEEE Access 9:26541–26555.
https://doi.org/10.1109/access.2021.3058589
5. Goswami DY, Kreith F (2007) Handbook of energy efficiency and renewable energy. CRC
Press. https://doi.org/10.1201/9781420003482
6. Solanki CS (2015) Solar photovoltaics: fundamentals, technologies and applications. PHI
Learning Pvt. Ltd., India. ISBN 987-81-203-5111-0
7. Uzum B, Onen A, Hasanien HM, Muyeen SM (2021) Rooftop solar PV penetration impacts
on distribution network and further growth factors—a comprehensive. Rev Electron 10(1):55.
https://doi.org/10.3390/electronics10010055
8. Sudabattula SK, Kowsalya M (2016) Optimal allocation of solar based distributed generators
in distribution system using Bat algorithm. Perspect Sci 8:270–272. https://doi.org/10.1016/j.
pisc.2016.04.048
9. Bawazir RO, Cetin NS (2020) Comprehensive overview of optimizing PV-DG allocation in
power system and solar energy resource potential assessments. Energy Rep 6:173–208. https://
doi.org/10.1016/j.egyr.2019.12.010
10. Ahmed IM, Kamel S, Abdel-Mawgoud H, Yu J (2018) Optimal allocation of PV based DG in
distribution networks at different load levels. In: Twentieth international Middle East power
systems conference, pp 649–654. https://doi.org/10.1109/mepcon.2018.8635122
11. Saha S, Mukherjee V (2021) A novel multi-objective modified symbiotic organisms search
algorithm for optimal allocation of distributed generation in radial distribution system. Neural
Comput Appl 33(6):1751–1771. https://doi.org/10.1007/s00521-020-05080-6
12. Janamala V (2021) A new meta-heuristic pathfinder algorithm for solving optimal allocation
of solar photovoltaic system in multi-lateral distribution system for improving resilience. Appl
Sci 3(1):1–17. https://doi.org/10.1007/s42452-020-04044-8
13. Panda S, Rout PK, Sahu BK (2021) Demand side management by PV integration to micro-grid
power distribution system: a review and case study analysis. In: Green technology for smart
city and society, pp 417–432. https://doi.org/10.1007/978-981-15-8218-9_35
14. Salih ZH, Hasan GT, Mohammed MA, Klib MAS, Ali AH, Ibrahim RA (2019) Study the
effect of integrating the solar energy source on stability of electrical distribution system. In:
22nd international conference on control systems and computer science (CSCS), pp 443–447.
https://doi.org/10.1109/CSCS.2019.00081
432 A. Routray et al.

15. Marco M, Zordan D, Dini P, Rossi M (2014) Solar stat: modeling photovoltaic sources through
stochastic Markov processes. In: IEEE international energy conference (ENERGYCON), pp
688–695. https://doi.org/10.1109/ENERGYCON.2014.6850501
16. Zobaa AF, Ramesh CB (2011) Handbook of renewable energy technology. World Scientific
17. Dave JV, Halpern P, Myers HJ (1975) Computation of incident solar energy. IBM J Res Dev
19(6):539–549. https://doi.org/10.1147/rd.196.0539
18. Jeong J, Culler D (2012) A practical theory of micro-solar power sensor networks. ACM Trans
Sens Netw 9(1):1–36. https://doi.org/10.1145/2379799.2379808
19. https://power.larc.nasa.gov/data-access-viewer/
20. HOMER (2019) The hybrid optimization model for electric renewable. [Online]. Available:
http://www.homerenergy.com
21. “TRINASOLAR: homepage”. [Online]. Available: http://www.trinasolar.com/us/product/duo
max72/duomax-peg14
22. “TIMENDATE: homepage”. [Online]. Available: https://www.timeanddate.com/weather/
india/bagalkot/historic?month=6&year=2019
23. Duong MQ, Nam Tran NT, Sava GN, Scripcariu M (2017) The impacts of distributed generation
penetration into the power system. In: International conference on electromechanical and power
systems (SIELMEN), pp 295–301. https://doi.org/10.1109/SIELMEN.2017.8123336
24. Weinstock D, Appelbaum J (2009) Optimization of solar photovoltaic fields. J Sol Energy Eng
131(3):031003-1–9. https://doi.org/10.1115/1.3142705
25. Gurupira T, Rix AJ (2017) Constrained optimization of photovoltaic (PV) module layouts.
In: IEEE AFRICON, Cape Town, pp 1179–1184. https://doi.org/10.1109/AFRCON.2017.809
5649
26. “NREL: renewable resource data center home page”. [Online]. Available: http://www.nrel.gov/
rredc/
27. Probability Subcommittee (1979) IEEE reliability test system. IEEE Trans Power Appar Syst
98(6):2047–2054. https://doi.org/10.1109/TPAS.1979.319398
28. Mirjalili SM, Lewis A (2014) Grey wolf optimizer. J Adv Eng Softw 69:46–61. https://doi.org/
10.1016/j.advengsoft.2013.12.007
29. Das D, Kothari DP, Kalam A (1995) Simple and efficient method for load flow solution of
radial distribution networks. Int J Electr Power Energy Syst 17(5):335–346. https://doi.org/10.
1016/0142-0615(95)00050-0
30. Das D, Negi HS, Kothari DP (1994) Novel method for solving radial distribution networks.
Proc IEE Gener Transm Distrib 141(4):291–298. https://doi.org/10.1049/ip-gtd:19949966
Author Index

A Pradhan, Raseswari, 377

Abdelaziz, Almoataz Y., 125 Punna, Srinivas, 181
Annamraju, Anil, 221
Ansari, Salauddin, 77
Arunkumar, C. R., 181 R
Arya, Sabha Raj, 403 Rakesh Chandra, D., 103, 279
Ray, Papia, 1, 153, 313
Routray, Aliva, 403
B
Battapothula, Gurappa, 221
Battu, Neelakanteshwar Rao, 257
Bhat, Sunil, 49 S
Bobba, Phaneendra Babu, 333 Salkuti, Surender Reddy, 1, 29, 103, 279,
297, 313, 355
Samadhiya, Akshit, 153
G Sangam, Narasimha, 297
Gupta, Om Hari, 77 Satish, Raavi, 125
Singh, Arvind R., 49, 313
Srinivasarathnam, C., 221
K Srivastava, Ishan, 49
Kim, Seong-Cheol, 297
Kosaraju, Satyanarayana, 333
Krishna, Perka, 257
T
Tummala, Suresh Kumar, 333
M
Manthati, Udaya Bhasker, 181
Mistry, Khyati D, 403 V
Vaisakh, K., 125
Veeramsetty, Venkataramana, 103, 279
N
Namrata, Kumari, 153

Y
P Yammani, Chandrasekhar, 221
Pagidipala, Sravanthi, 1, 297 Yarlagadda, Venu, 257

© The Editor(s) (if applicable) and The Author(s), under exclusive license 433
to Springer Nature Singapore Pte Ltd. 2022
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6
Subject Index

A B
Absolute error, 226 Back propagation algorithm, 107, 108, 113,
Acceleration coefficients, 232, 234 114
Access, 3, 7, 8, 16, 43, 69, 70, 224, 225 Backward/forward sweep method, 126
Accuracy, 11, 29, 103, 112, 113, 125, 139, Backward pass, 114, 118
140, 153, 174, 225–227 Bacterial foraging algorithm, 226
Active filters, 259, 268, 269, 316, 317 Band stop filter, 269
Bandwidth, 57, 58, 62, 64, 68, 72, 196–199,
Active frequency drift, 160, 162
202
Active islanding, 153, 160, 169, 174
Battery, 1, 5, 12, 19–24, 29, 31, 35, 37–40,
Active power load, 103–105, 109, 110, 359 53, 181, 183–186, 188, 191–194,
Active power loss, 126, 322, 357, 359, 196, 197, 199, 202–209, 211–213,
361–363, 366, 373, 405, 416, 418, 217, 218, 259, 268, 297–302,
422, 428, 431 305–309, 326, 330, 390
Adaptive control technologies, 393–395 Battery bank, 39, 383, 390
Adjustable speed drives, 149 Battery electric vehicles, 20, 24, 299
Advanced communication technologies, Battery energy storage, 29, 30, 35, 315, 358
396, 398 Best solution, 226, 231, 233, 235
Advanced metering infrastructure, 5, 6, 8, Beta distribution, 38
10, 51, 394 Bias parameters, 104, 108, 113–117, 119,
Algorithm-specific parameters, 221, 232, 121, 122
234, 423 Bidirectional DC-DC converters, 181, 185,
Alternate energy sources, 225, 297 186, 188, 195, 300
Analytical method, 226–228, 240 Bifacial PERC, 335, 336, 339
Angular speed, 287–289, 291 Biomass energy, 3, 29, 31, 44
Anti-islanding (AI), 157, 158 Blockchain technology, 1, 13, 15–18
Bode plot, 198, 199, 202, 205
Apparent solar time, 407
Boltzmann constant, 228
Artificial immune system, 226
Boost, 81, 104, 194, 204, 206, 213, 262
Artificial neural network, 14, 103, 104 Boost converter, 81, 88, 89, 167, 195, 205,
Astronomical model, 405–410, 412, 428 212, 217, 257, 258, 262, 270, 273,
Automatic meter reading, 51, 60 276, 302, 317
Automatic phase drift, 163 Box plot, 109
Automatic Transfer Switch (ATS), 396 Brushless DC (BLDC) motor, 297, 299,
Azimuthal displacement, 407 306–308
© The Editor(s) (if applicable) and The Author(s), under exclusive license 435
to Springer Nature Singapore Pte Ltd. 2022
S. R. Salkuti and P. Ray (eds.), Next Generation Smart Grids: Modeling, Control
and Optimization, Lecture Notes in Electrical Engineering 824,
https://doi.org/10.1007/978-981-16-7794-6
436 Subject Index

Buck, 81, 193, 204, 262, 302, 306, 309 Constant voltage method, 263
Buck boost converter, 262 Contact resistance, 227
Buck converter, 262 Control strategy, 55
Building area network, 63 Control issues, 387
Bulk generation, 52, 53 Controller design, 185, 196
Bus, 57, 60, 68, 125, 127–150, 167, 203, Control parameter, 234
279, 281, 282, 285–293, 321, 322, Control strategy, 155–157, 181, 185, 213,
338–340, 355, 358, 359, 361–364, 258, 315, 317
366, 367, 371–373, 377, 378, 403, Control target, 203
406, 413, 418, 419, 425–430 Conventional energy, 30, 224, 259, 260,
Bus voltage, 128, 132, 133, 136, 138, 182, 263, 280, 285
203, 204, 293, 321, 322, 363, 371, Convergence criterion, 136, 236, 237, 366
372, 405, 406, 426–428, 430, 431 Converter control circuit, 129
Converters, 19, 35, 81, 129, 155, 156, 181,
183–186, 188, 189, 191, 195, 196,
C 201, 202, 204, 205, 210, 212, 218,
Capacitor banks, 127, 132, 139, 149, 268, 258, 259, 262, 276, 280, 282–284,
361 293, 298–300, 302, 304–309, 313,
Capital investment, 35, 225, 283 315, 317
Carbon-dioxide, 43, 44, 224, 298 Correlation, 5, 103–107, 113
Crossover probability, 234
Cascaded H-type inverter, 268
Crystalline silicon, 261, 334
Current stress, 181, 185
Current sensor, 212
Centralized places, 223
Current source, 127, 227, 263
Characteristics, 3, 4, 19, 21, 23, 24, 30, 54,
Current source inverter, 266
65, 68, 73, 77, 91, 93, 94, 111, 112,
Current sweep method, 263
128, 157, 181, 183, 184, 187, 188,
Current temperature coefficient, 229, 236,
225, 227, 228, 230, 240–242,
237, 240
244–249, 263, 265, 283, 299, 315,
Customer Relationship Management
323, 382, 386, 426
(CRM), 393, 395
Characterization, 300, 333, 336, 338, 352
Cyber physical systems, 382, 384
Charge of electron, 228
Charging, 17, 19–24, 31, 39, 40, 181, 185,
186, 191, 193–196, 204, 205, 213, D
217, 218, 301, 305, 306, 355, 357, DC-DC converters, 181, 184–186, 188,
371 195, 259, 262, 300, 302, 306, 309
Chemical energy, 222 DC grid voltage, 181, 186, 188, 203, 206,
Chopping fraction, 160, 162, 166 207, 209, 210, 212–214, 217, 218
Circuit breaker switching strategy, 168 DC microgrid, 181–185, 187–189, 191,
Coal, 52, 224, 381, 384 193–197, 204, 208, 212, 213, 217,
Cognitive component, 233 218, 281
Collection efficiency, 337 Dead time, 161, 162, 190, 192
Communication systems, 5, 8, 49, 57, 62, Degradation cost, 39, 40
64, 65, 73, 153, 398 Demand, 3, 5, 6, 11, 12, 14, 15, 18, 19, 22,
Comparison phase, 235, 236 25, 29, 30, 40, 49–56, 104, 155, 167,
Compensation-based method, 127 169, 170, 183, 191, 203, 207, 209,
Complex power generation, 135 213, 223, 224, 258, 280, 281, 285,
Composite residential loads, 149 315, 317, 356, 357, 362, 379, 380,
Computer, 223, 300, 352 383, 384, 387, 393, 397, 405,
Conclusions, 79, 105, 127, 317, 358 412–414
Congestion, 7, 9, 72, 280, 404 Demand side management, 225
Consistency, 39, 226 Depth of discharge, 39
Constant load model, 416 Detection time, 153, 156–159, 164, 165,
Constant speed wind turbine, 281 167, 169, 172, 174, 176
Subject Index 437

Deterministic algorithms, 225 Electric vehicles, 1, 7, 17–20, 22, 24, 125,

DG capacity, 3, 428 297, 298, 300, 306, 307, 355–357,
DG constraints, 413, 422 371, 390, 397
DG location, 428 Elitism probability, 234
Digital control of smart grid, 391 Encirclement of prey, 423
Diode clamped inverters, 268 Energy density, 20, 21, 39, 183, 185, 193,
Diode ideality factor, 90, 226, 228, 236, 301
237, 240, 245 Energy management systems, 6, 185
Diode reverse saturation current, 227, 228, Energy service interface, 55
236, 237, 240, 245 Energy storage, 1, 3, 10, 12, 19, 22, 40, 53,
Direct-drive synchronous generator, 281 181, 185, 326, 380, 384
Directorate General of Training (DGT), 396 Energy storage systems, 1, 30, 181, 183,
Discharging, 22, 39, 40, 181, 185, 186, 293, 314
188–190, 193, 194, 196, 204, 205, Enterprise Asset Management (EAM), 393
217, 218, 306 Enterprise resource planning, 393
Distributed energy resources, 10, 25, 315, Environmental challenges, 280, 384
356, 393 Environmental impact, 11, 12, 29, 51, 315,
Distributed generation, 3, 5, 23, 53, 125, 381, 404
155, 156, 167, 176, 257, 280, 314, Equation of time, 406
355, 356, 371, 379, 383, 388–391, Equivalent Series Resistance (ESR), 204
403, 404 Error detection ratio, 158–160
Distributed generator, 153, 158, 271 Evolutionary algorithm, 234
Distribution management system, 6, 55 Evolution of smart grid, 381
Distribution networks, 3, 49, 53, 55, 58, 62, EV supply equipment, 305
64, 65, 125–127, 130, 157, 316, 325, Exploitation, 70, 315, 404, 424
326, 357, 358, 360, 362, 403 Exploration, 10, 43, 425
Distribution system, 4, 6, 13, 24, 54, 55, External quantum efficiency, 336, 344
104, 153, 271, 313, 315–317, 325,
326, 355–362, 393, 403–406, 412,
413, 418, 425–430 F
Doubly fed induction generator, 279, 282 Fast Fourier Transform (FFT), 318
dSPACE DS1104, 212 Fault detection, 5, 13, 52, 55, 388, 394
Duty cycle, 88, 190, 192, 193, 201, 302, Feature selection, 103, 113
304 FFT analysis, 318, 320, 327, 328
Dynamic programming, 225 Filter inductor, 195, 196
Dynamic voltage restorer, 313, 316, 325 Filters, 88, 89, 195, 196, 258, 259, 266,
268–274, 276, 284, 321, 326, 327
Finger contact, 335
E Finger pitch, 338
Economical, 31, 49, 50, 53, 56, 64, 73, 153, Finger spacing, 338
282, 357, 389, 404 Fitness, 235–237, 423
Effective solar radiation, 406, 408, 410, 428 Fitness function, 230, 231, 240
Effective spacing, 414 Fixed excitation voltage type, 128
Efficiency, 2, 5–7, 9–11, 20, 21, 24, 25, 31, Fixed Speed Induction Generator (FSIG),
32, 35, 36, 39–42, 44, 45, 78, 79, 81, 127
89, 97–99, 104, 162, 184, 185, 225, Fixed-speed wind generator, 279
226, 261–263, 265, 266, 270, Flow chart, 88, 160, 161, 165, 166, 194,
299–303, 315–317, 333, 338, 343, 227, 263, 264, 365
356, 358, 383, 393, 404, 406, 411 Fluorescent light banks, 149
Electrical energy, 9, 15, 19–21, 42, 56, 81, Flying capacitor inverters, 268
88, 187, 222–224, 259, 261, 301, Flywheels, 12, 21, 301, 390
356, 377, 391 Forecasting, 14, 53, 55, 103, 104, 112, 113,
Electric grid, 18, 24, 54, 181, 377, 378, 381 383
438 Subject Index

Forward pass, 113, 116, 122 Grid resistance, 340, 345, 347
Fossil fuels, 3, 19, 22, 24, 29, 40, 42, 43, 45, Grid volatility, 8
221, 224, 258, 268, 298, 301, 389 Gross Domestic Product (GDP), 223
Frequency drift, 160, 162
Fuel cell hybrid electric vehicle, 24
Fuel cells, 5, 19, 21, 24, 126, 127, 156, 259, H
299, 315, 390 Half bridge inverter, 266
Full bridge inverter, 266, 267 Harmful gases, 224
Fundamental power flow algorithm, 131 Harmonic current injections, 129, 136, 138
Fundamental voltage, 142, 144, 146, 148, Harmonic detection, 165
149, 269 Harmonic distortion elimination, 313
Fuzzy logic control, 299, 302, 305, 316 Harmonic load composition, 139
Harmonic pollution, 146
Harmonic Power Flow Algorithm (HPFA),
126, 127, 129, 136, 137, 139, 140,
G
146
Gauss-Seidel, 358
Harmonics, 60, 89, 125–127, 129, 136,
Gearbox, 31, 282 138–141, 144–147, 165, 257–259,
Gem silicon solar cell structure, 335 262, 267–271, 273, 276, 281, 304,
General Electricals (GE), 396–398 315–320, 326–328, 330
Generation, 1–7, 9–15, 17–19, 21, 29–32, Harmonic source current spectrum, 127
35, 39, 45, 51–53, 55, 56, 60, 62–64, Harmonic spectrum, 127
73, 77–79, 104, 136, 154, 155, 169, Harmony memory, 234
182, 207, 209, 221, 223–225, Head branch, 130–132
234–236, 257, 258, 261, 268, Head bus, 130–132
280–282, 286, 299, 301, 315, 316, High pass filter, 269
328, 340–342, 344, 356, 378–381, Home area network, 55, 63
383, 388, 390, 391, 396, 397, 403, Hour angle, 407
405, 406, 411, 417, 419, 420, 422 Hybrid electric vehicle, 19, 20, 22, 24, 183,
Generation profile, 336, 340 185, 218, 297, 299, 301, 302, 305,
Generator protection panels, 396 307, 383
Genetic algorithm, 226, 234, 357 Hybrid energy storage systems, 181, 183
Geographic information system, 6 Hybrid inverters, 268
Geometric programming, 225 Hybrid islanding detection, 157, 160, 165
Geothermal energy, 31, 42, 43, 45 Hydro Power, 41, 260, 261
Global best solution, 231, 232
Global positioning system satellite, 391
Global radiation, 403, 406, 408 I
Global warming, 3, 19, 43, 44, 224, 299 Impedance measurement, 164
Greenhouse gas emission, 7, 126 Incidence angle, 407
Greenwich meridian angle, 406 Incident photon current, 337
Grey wolf optimization, 403, 406, 422, 428 Incremental conductance method, 263
Grid, 1–18, 23, 25, 31, 35, 41, 42, 44, Independent system operators, 53, 56
49–58, 61, 64–73, 126, 127, Indian Government, 224
155–159, 162–165, 167, 168, 170, Induction generator, 126, 128, 129, 280,
172, 174, 183, 185, 191, 199, 282, 283, 289
211–213, 217, 224, 227, 257, 259, Inequality constraints, 421
268–274, 276, 279–285, 293, 302, Inertia weight, 232–234, 237, 240
313–321, 324–327, 330, 334, 338, Information technology, 385, 391
339, 347, 377, 379–391, 393–399, Insulated gate bipolar transistors, 304
405, 406, 412, 418, 420 Integrated-circuit, 379
Grid connected, 126, 153, 155, 156, 160, Intelligent electronic devices, 9, 49, 51
163, 167, 169, 176, 259, 273, 276, Internal combustion engines, 18–20, 297,
279 298
Subject Index 439

Internet of things, 18, 386 Matlab, 77, 89, 99, 130, 139, 167, 205, 221,
Inverter, 19, 20, 24, 35, 88, 157, 160, 237, 257, 279, 285, 297, 300, 306,
162–165, 167, 176, 257–259, 307, 315, 426
265–268, 270, 271, 273, 275, 276, Maximum power point, 81, 187, 229, 230,
280, 299, 302, 316, 330 236, 237, 240, 344
Irradiance, 30, 35, 37, 38, 45, 77, 81, 90–96, Maximum power point tracking, 77, 187,
98, 228–230, 236, 237, 240–242, 257, 259, 263, 273, 313, 316, 389
244–249, 251–253, 259, 263, 323, Mean square error, 79, 317
394, 406–409, 412, 414–421 Meta-heuristic techniques, 227, 240, 245
Islanding detection, 153, 154, 156–158, Metaphor based algorithm, 422
160, 167, 169, 172, 174–176 Meter data management system, 55
Isolated converters, 184, 262 Metropolitan area network, 67
Iterative procedure, 131, 132, 136 Microgrid, 10, 17, 18, 30, 53, 125, 153,
155, 169, 183–185, 189, 218,
313–315, 330, 397
J Model predictive control, 181, 185
Jaya algorithm, 221, 227, 234, 236–238, Modified ROCOF, 167, 169, 170, 172–176
240, 242–247, 249–253 Modified sine wave inverter, 265
Module, 37, 86, 87, 89, 90, 183, 188, 221,
222, 226–228, 236, 237, 240, 242,
K 243, 245, 246, 249, 334–336, 338,
Kelvin, 228 348, 405, 411, 412
Kernel density estimation, 403, 405, 414 Mono-crystalline, 221, 226, 227, 240, 242,
245
Monocrystalline cells, 261
L Monte carlo simulation, 357, 405, 414
Lambert W function, 226 MOSFET switch IRFP460, 212
Laterals, 129, 130 Motion energy, 222
Lead-acid battery, 212 Multi-directional power flow, 313, 316
Leakage current, 227 Multi input bidirectional DC-DC converter,
Levelized cost of energy, 335 181, 185
Levenberg parameters, 226 Multiple PWM inverter, 267
Life expectancy, 6, 223 Mutation probability, 234
Light-generated current, 336, 337
Linear loads, 127, 136, 138, 149, 313
Literacy rate, 223 N
Lithium-ion (Li-ion) battery, 20, 38, 300, NASA, 408–410, 415
301 National institute of standard and
Load demand, 1, 29, 30, 38, 53, 54, 104, technology, 397
181, 186, 205, 207, 209, 210, 212, Negative component of current injection,
216–218, 221, 316, 356, 360, 366, 164
405, 406, 413, 414, 417, 428–430 Neighborhood area network, 62
Load flow, 355, 358–360, 366, 373, 406, Neural network, 14, 259, 280
428, 430 Newton method, 226
Load modelling, 55, 416 Nickel–metal hydride batteries, 301
Local generation, 167, 170, 171 Nitrogen-oxide, 224
Local islanding detection, 160 Nominal operating cell temperature, 411
Local load, 89, 155, 167, 169–175 Nominal temperature, 337, 340
Low pass filter, 269 Non-conventional energy sources, 259, 260
Non detection zone, 153, 156, 158
Non-isolated converters, 184, 262
M Non-linear, 77, 125–127, 136, 144, 146,
Machine learning, 1, 13–15, 103, 104 149, 187, 221, 225, 226, 315–319,
Mathematical modelling, 240 321, 323, 326–328, 330
440 Subject Index

Non-linear loads, 125–127, 136, 138, 144, Photovoltaic, 10, 35, 77–79, 81, 126, 127,
146, 149, 316–319, 321, 323, 326, 257, 259, 273, 334–336, 403, 405
327 Photovoltaic array, 317, 323
Nuclear energy, 222 Photovoltaic cell, 35, 259
Numbering scheme, 129 Photovoltaic distribution static
compensator, 313, 330
Photovoltaic power, 261
O Plug-in hybrid electric vehicles, 20, 22, 24,
Objective function, 226, 231–235, 358, 301, 383
363, 366, 418, 419 Point of common coupling, 65, 158, 258,
Obstructed astronomical model, 403, 409, 314
426, 428
Polycrystalline cells, 261
Obstruction factor, 408, 409, 428
Positive feedback, 162, 163, 165
Open circuit voltage, 229, 236, 237, 344
Operating temperature, 35, 240, 337, 340 Positive sequence impedance, 134
Operating voltage, 83, 85, 188, 229, 236, Positive sequence voltage, 164
237, 240 Power flow, 49, 51, 52, 55, 57, 68
Operational constraints, 357, 418 Power line communication, 57, 58
Optimization, 5, 8, 13, 19, 30, 51, 72, 221, Power converter, 167, 183, 280, 300
226, 231, 234, 237, 240, 300, Power density, 24, 39, 181, 183, 185, 193,
356–359, 396, 422, 426 280, 301
Optimization algorithm, 146, 225, 234, Power electronic converter, 19, 129, 280,
236, 426 289
Optimization technique, 30, 221, 222, 226, Power factor, 128, 160, 163, 270, 302, 316,
228, 231–233, 240, 259, 393, 405, 317, 319, 357, 361, 362
422, 428, 429 Power factor control mode, 128
Outage management systems, 6 Power flow, 127, 132, 136, 139, 141, 145,
Output of solar, 37 147, 184, 185, 189, 191, 193, 194,
Over/under current, 167, 169, 172, 174, 175 196, 217, 218, 285, 287, 293, 302,
Over/under frequency, 167, 169 313, 317, 356, 358, 405, 416, 426
Over/under voltage, 167, 169 Power flow algorithms, 125, 426
Power flow program, 131
Power loss minimization, 357, 358, 404,
P 426
Pass band filters, 269 Power quality, 54, 126, 146, 153, 155–160,
Passivated emitter, 333, 334 162–165, 167, 169, 174, 176, 259,
Passive filters, 269 268, 270, 281, 313–317, 325–327,
Passive islanding detection, 160 329, 330, 388, 404, 405
Peak loads, 3, 126, 383, 397, 417, 422, 426, Power system analysis toolbox, 279, 293
427
Probability density function, 414–420
Peak overshoot, 186, 209–211, 218, 287,
Problem formulation, 227, 228, 358, 362,
288, 293
404, 418
Peak-peak inductor current, 196
Peer-to-peer semantic grid, 396 Proportional integral, 181, 322
Perturb and observe, 77, 79, 82, 188, 259, Proportional integral derivative, 299
263, 264, 313, 320, 322 Pulse width modulation, 267, 270, 303
Perturb and observe algorithm, 259, 263 Pumped hydro energy systems, 30, 40
Phase Jump Detection (PJD), 168 PV array, 167, 187, 205, 259, 313, 320,
Phase Locked Loop (PLL), 281 321, 405
Phasor measurement units, 6, 9, 24, 51, PV bus sensitivity matrix, 134
391, 394 PV cell, 35–37, 79, 187, 225, 261, 263,
Photoelectric, 35, 36, 335 265, 316, 334–336, 405
Photon current densities, 345, 346 PV cell modelling, 263
Photon currents, 336 PV cell temperature, 37, 411, 412
Subject Index 441

PV generation, 170–175, 181, 186, 188, S

191, 205–210, 212, 213, 215, 217, Sandia frequency shift, 162, 166
218, 417 Sandia voltage shift, 162
PV node, 126, 134, 135 SCADA system, 53, 71
PV panel, 35, 37, 81, 83, 86–89, 186–188, Scaled astronomical Model, 410
195, 206, 208, 209, 227, 229, 406, Scaling factor, 339, 341, 405, 408, 428
408, 409, 413, 429 Search space, 226, 231, 233, 234, 236, 237,
240, 422, 425
Self shadowing effect, 413
Q Semi-conductor device, 262
Quality factor, 158, 264 Sequential programming, 226
Series filter, 269
Series resistance, 37, 227, 228, 236, 237,
R 240, 245, 336, 340
Radial distribution network, 126, 403 Settling time, 186, 207, 209–211, 218
Radial distribution system, 355, 357, 358,
Shading of solar panels, 413
371, 403, 413, 425, 426
Shape parameter, 38
Radiant energy, 222
Short circuit current, 36, 90, 229, 236, 237,
Radio frequency identification, 386
240, 344
Random number, 234, 235
Short-circuit fault, 329
Rate of change of frequency, 67, 157
Shuffle frog leap algorithm, 226
Reactive power, 55
Shunt active filter, 257–259, 269, 270, 272,
Reactive power compensation, 55, 313,
276, 316
316, 317, 362
Shunt filter, 269
Reactive power generation, 134–136, 141
Shunt resistance, 36, 37, 227, 228, 236,
Reactive power loss, 360, 426, 427
237, 240, 244, 340
Real power generation, 134, 135, 141
Real power loss, 138, 355, 362, 363, 403, Silicon nitrate, 333, 352
426 Simulation annealing, 226
Real-time distribution networks, 126 Sine wave inverter, 265
Rechargeable batteries, 20, 300, 390 Single diode modelling, 225
Reflection, 223, 336, 342, 345, 346, 348, Single phase inverter, 266
349, 353 Single PWM inverter, 267
Regenerative energy, 298, 301, 302 Sinusoidal pulse width modulation, 304
Regional transmission operators, 53 Sinusoidal PWM inverter, 267
Regulated power supply, 212 Sinusoidal voltage waveform, 136
Regulating excitation voltage type, 128 Slip, 128, 282, 289
Reliability test system, 279, 282, 285, 416 Slip mode frequency shift, 163
Reliable, 1, 3, 5, 6, 8, 25, 31, 43, 69, 73, Slope, 3, 30, 81, 83, 163, 189–192, 230
103, 104, 157, 169, 174, 223–225, Small hydro energy, 30, 40
279, 315, 357, 379, 382, 391 Small signal linear averaged model, 186,
Remote islanding detection, 160 196, 197, 199
Renewable energy resources, 1, 3, 11, 29, Smart appliances, 9
30, 45, 280, 315 Smart grid technologies, 49, 389, 396, 398
Renewable energy sources, 10, 16, 18, 51, Smart meter, 7–10, 24, 49, 51, 52, 55, 56,
125, 153, 155, 181–183, 221, 225, 63, 67, 69, 70, 385
257, 263, 358, 382, 383, 391 Smart power distribution network, 125
Reverse saturation current, 187 Social, 13, 224, 237, 280, 382, 384, 388,
Reweighted zero attracting, 313, 316, 317 398, 423
R.M.S value of voltage, 138 Social component, 233, 234
Root mean square error, 226, 231 Social learning rate, 233
Rotor resistance, 288, 289, 291, 292 Society, 10, 12, 223, 384
Rotor resistance control, 279, 281, 287, Solar absorptance, 411
289–292 Solar based DGs, 167
442 Subject Index

Solar declination angle, 406, 407 T

Solar energy, 3, 17, 34, 35, 37, 77–79, 81, Technical challenges, 15, 153, 356, 383
88, 225, 257, 259, 261, 263, 403, Tail branch, 130, 131
405, 406, 422, 428 Tail bus, 130–132, 138
Solar farm, 406, 408, 409, 412–414, 426, Teaching learning based optimization, 226
427, 429 Temperature, 30, 35–37, 39, 42, 43, 77, 79,
Solar insolation, 30, 358, 409, 429 81, 86, 87, 89–96, 98, 99, 104, 187,
Solar irradiation, 29, 35–37, 81, 189, 313 228–230, 236, 237, 240–242,
Solar panel efficiency, 408 244–249, 251–253, 263, 265, 337,
Solar photovoltaic, 3, 29, 30, 167, 169, 317, 344, 346, 394, 403, 405, 406, 411,
323 412, 429
Solar power, 6, 35, 37, 79, 89, 90, 225, 258, Temperature coefficient, 87, 90
259, 313, 315, 334, 403, 405, 406, Temperature method, 263
408, 409, 411–414, 416, 427, 429 Temperature model, 412, 426
Solar transmittance, 412 Test system, 68, 285, 287, 292, 358, 365,
Spot loads, 144, 149 366, 414, 422, 425–431
Square wave inverter, 265 Thermal energy, 29, 222, 260
Squirrel cage induction generator, 127, 281, Thermal voltage, 228
282 Thin-film, 226, 227, 240, 242, 245
Stability, 23–25, 53, 155, 159, 162, 169, Thin film cells, 261
183, 185, 213, 280–283, 285, 288, Three phase inverter, 266, 267, 271
357, 390, 397, 405 Three-phase PFA, 126
Three-phase system, 313, 330
Standard Test Condition (STC), 37, 229,
250, 333, 353, 411 Threshold radiation, 415
Tidal energy source, 260
State of charge, 39, 186, 300, 306, 309
Tolerance limit, 132, 138
Static compensator, 281, 317
Total harmonic distortion, 139, 257, 258,
Static Voltage Characteristic Model
270, 272, 276, 316
(SVCM), 128
Transfer function, 196, 198, 199, 202, 204,
Steady-state behaviour, 317
323
Stochastic algorithm, 225, 226, 231 Transformers, 31, 35, 41, 42, 89, 127, 167,
Stochastic gradient descent optimizer, 104, 184, 262, 285, 326, 327, 405
107, 108, 113, 114 Transient response, 313, 317
Stochastic modelling, 414 Transient stability, 279–281, 283, 287, 288,
Strengthened, 223 293
Sub-station, 23, 24, 393 Transient Voltage Surge Suppression
Substation bus voltage, 131 (TVSS), 396
Sulphur dioxide, 224 Transmitted, 20, 58, 223, 378
Supercapacitor, 12, 183–186, 188, Transportation sector, 298, 356
191–197, 202, 204, 205, 212, 218 Two-diode model, 226
Superconducting magnetic energy storage, Two-way communication, 6, 8, 9, 49, 51,
330 54, 55, 57, 64, 73
Supervisory Control And Data Acquisition
(SCADA), 8, 10, 71, 168, 393, 394
Sustainability, 5, 10, 11, 31, 78 U
Sustainable energy, 6, 42, 382, 388 Ultracapacitors, 298, 301
Swarm intelligence algorithm, 234 Unbalanced load, 126
Swarm size, 232, 233 Uncertainty, 3, 11, 29, 30, 39, 221, 281,
Switching frequency, 197, 268, 303 315, 337, 356, 390, 414
Synchronous based-renewable DG, 125, Uncertainty modeling, 30, 37
129, 144, 146 Unified controller, 181, 185, 186, 196, 218
Synchronous generators, 128, 281, 288 Unintentional islanding, 153–157, 167
Synchronous machine, 291 Uninterruptible Power Supply (UPS), 396
Subject Index 443

Updation phase, 235, 236 WAMS, 6, 53, 394

Urbanization rate, 223 Wave energy source, 260
Utility grid, 29, 88, 89, 126, 159, 162, 167, Weekend hourly load, 417
269, 271, 280, 282, 283 Weibull PDF, 32, 34, 37
Utilization, 3, 5, 11, 12, 19, 22, 49, 153, Weights, 38, 104–108, 110–114, 116–122,
158, 223, 225, 298, 356, 360, 378, 269, 300, 301, 323, 324
390, 391, 395 Wide area measurement system, 6, 394
Width of finger, 338
Wind energy, 5, 29–31, 45, 260, 280–282,
V 285, 360
Vector Shift IDM, 167 Wind energy generators, 33, 279, 293
Vehicle-to-grid, 390 Wind energy source, 260
Velocity, 231–234, 237 Wind power, 11, 31–34, 225, 260, 279,
Voltage control mode, 128, 262 281, 282, 285, 293, 315
Voltage fluctuation injection, 166 Wind power generation, 280
Voltage mismatches, 133, 134
Wind speed prediction, 13
Voltage mismatch matrix, 134
Wind turbines, 5, 10, 31, 32, 127, 155, 260,
Voltage regulator, 127, 149, 269
279–283, 285–289, 293
Voltage sag, 60, 269, 281, 316, 317,
Wireless mesh network, 62
325–327, 329, 330
World, 3, 6, 8, 10, 11, 29, 30, 44, 160,
Voltage sensor, 212
223–225, 260, 261, 280, 285, 298,
Voltage source converter, 258, 280, 304
334, 379, 382–384, 386, 397, 405
Voltage source inverter, 164, 257, 266, 303,
Worst solution, 235
315
Voltage stabilization, 185
Z
W Zero time, 160
Wafer, 261, 339, 340 Zero voltage switching, 185