Advances in Renewable Energy & Electric Vehicles: Select Proceedings of AREEV 2022 (Lecture Notes in Electrical Engineering, 1083) [1st ed. 2024] 9819961505, 9789819961504

Citation preview

Lecture Notes in Electrical Engineering 1083

Suryanarayana Kajampady Shripad T. Revankar   Editors

Advances in Renewable Energy & Electric Vehicles Select Proceedings of AREEV 2022

Lecture Notes in Electrical Engineering Volume 1083

Series Editors Leopoldo Angrisani, Department of Electrical and Information Technologies Engineering, University of Napoli Federico II, Napoli, Italy Marco Arteaga, Departament de Control y Robótica, Universidad Nacional Autónoma de México, Coyoacán, Mexico Samarjit Chakraborty, Fakultät für Elektrotechnik und Informationstechnik, TU München, München, Germany Jiming Chen, Zhejiang University, Hangzhou, Zhejiang, China Shanben Chen, School of 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, University of Karlsruhe (TH) IAIM, Karlsruhe, Baden-Württemberg, Germany Haibin Duan, Beijing University of Aeronautics and Astronautics, Beijing, China Gianluigi Ferrari, Dipartimento di Ingegneria dell’Informazione, Sede Scientifica Università degli Studi di Parma, Parma, Italy Manuel Ferre, Centre for Automation and Robotics CAR (UPM-CSIC), Universidad Politécnica de Madrid, Madrid, Spain 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, Intelligent Systems Laboratory, Systems Research Institute, Polish Academy of Sciences, Warsaw, Poland Alaa Khamis, Department of Mechatronics Engineering, German University in Egypt El Tagamoa El Khames, New Cairo City, Egypt Torsten Kroeger, Intrinsic Innovation, Mountain View, CA, USA Yong Li, College of Electrical and Information Engineering, 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 Subhas Mukhopadhyay, School of Engineering, Macquarie University, NSW, Australia Cun-Zheng Ning, Department of Electrical Engineering, Arizona State University, Tempe, AZ, USA Toyoaki Nishida, Department of Intelligence Science and Technology, Kyoto University, Kyoto, Japan Luca Oneto, Department of Informatics, Bioengineering, Robotics and Systems Engineering, University of Genova, Genova, Genova, Italy Bijaya Ketan Panigrahi, Department of Electrical Engineering, Indian Institute of Technology Delhi, New Delhi, Delhi, India Federica Pascucci, Department di Ingegneria, Università degli Studi Roma Tre, Roma, Italy Yong Qin, State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong University, Beijing, China Gan Woon Seng, School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore, Singapore Joachim Speidel, Institute of Telecommunications, University of Stuttgart, Stuttgart, Germany Germano Veiga, FEUP Campus, INESC Porto, Porto, Portugal Haitao Wu, Academy of Opto-electronics, Chinese Academy of Sciences, Haidian District Beijing, China Walter Zamboni, Department of Computer Engineering, Electrical Engineering and Applied Mathematics, DIEM—Università degli studi di Salerno, Fisciano, Salerno, Italy Junjie James Zhang, Charlotte, NC, USA Kay Chen Tan, Department of Computing, Hong Kong Polytechnic University, Kowloon Tong, Hong Kong

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. **

Suryanarayana Kajampady · Shripad T. Revankar Editors

Advances in Renewable Energy & Electric Vehicles Select Proceedings of AREEV 2022

Editors Suryanarayana Kajampady Department of Electrical and Electronics Engineering NMAM Institute of Technology Karkala, Karnataka, India

Shripad T. Revankar School of Nuclear Engineering Purdue University West Lafayette United States Minor Outlying Islands

ISSN 1876-1100 ISSN 1876-1119 (electronic) Lecture Notes in Electrical Engineering ISBN 978-981-99-6150-4 ISBN 978-981-99-6151-1 (eBook) https://doi.org/10.1007/978-981-99-6151-1 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 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 Paper in this product is recyclable.

Contents

Review and Study of Solar String Inverters for a PV System . . . . . . . . . . . G. B. Praveen, M. Satyendra Kumar Shet, and K. Latha Shenoy An Improved Strategic Analysis on Fault Diagnosis in Modular Multilevel Converter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sujo Oommen, Burri Ankaiah, Mahesh Kumar, M. H. Ananda, K. Narayan Swamy, and M. C. Rashmi Mathematical Model Approach to Study and Analyze FOC-Based PMSM Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . H. Swathi Hatwar, Anup Shetty, and K. Suryanarayana Review on Power Converter Topology for EV Fast Charger . . . . . . . . . . . . Saptadipa Das, V. Joshi Manohar, A. V. Sunil Kumar, and R. Rajatha Mathematical Modeling and Analysis of Interleaved Two-Phase Boost PFC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dhanush Acharya, Suryanarayana K, Krishna Prasad, and L. V. Prabhu Design of a Solar Battery Charger with Maximum Power Point Tracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adrian Dsilva, Samith Suvarna, Laxmisha G. Ballal, and H. Swathi Hatwar Implementation of Edge Computing Model for the Processing of Data in Mines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . K. Aneesha Acharya, Akshit Gaurav, and Aman Srivastava

1

13

25 47

65

83

99

Adaptive Protection of Solar PV Microgrid Without ESS . . . . . . . . . . . . . . 111 L. Poireiton Meitei, K. P. Vittal, and James Antony Pinto Comparative Study of Sensor and Sensor Less Speed Control of Permanent Magnet Synchronous Machines . . . . . . . . . . . . . . . . . . . . . . . . 129 P. Sandhya, B. Rajalakshmi Samaga, and Ramzeena

v

vi

Contents

An Exposition of Digital Taylor-Fourier Transform . . . . . . . . . . . . . . . . . . . 139 Krishna Rao and K. N. Shubhanga Autonomous Microgrid Using New Perspective on Droop Control in AC Microgrid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 Siddaraj, Udaykumar R. Yaragatti, and H. Nagendrappa Solar Photovoltaic Charging of Electric Vehicle and V2G for Indian Electricity Demand Scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 S. Suraj, Narayan S. Manjarekar, and Soumyabrata Barik Review of Voltage Sag\Swell Mitigation Control Techniques with Dynamic Voltage Restorer in a Grid Integrated Distribution System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187 N. Sowmyashree, Hrushikesh Kulkarni, M.S. Shashikala, and K. T. Veeramanju Parametric Sensitivity Analysis of STATCOM Supplementary Modulation Controller Incorporated in SMIB System . . . . . . . . . . . . . . . . . 201 Dinesh Shetty and Nagesh Prabhu Review on the Addition of Antioxidants and Nanoparticles to Natural Ester as an Alternative to Transformer Oil . . . . . . . . . . . . . . . . . 217 S. N. Deepa, A. D. Srinivasan, M. Anusha, K. T. Veeramanju, and M. R. Chaitra

About the Editors

Suryanarayana Kajampady is currently a Professor and Head of the Department of Electrical and Electronics Engineering at NMAMIT, India. He obtained his B.E. in Electrical and Electronics Engineering from Mangalore University, M.Tech. in System Analysis and Computer Applications from Karnataka Regional Engineering College, India, and Ph.D. in the area of Power Electronics from VTU, India. His major areas of research interest include power electronics, control systems, EVs, and signal processing. He has published papers in journals and conferences of national and international repute. He is a Member of IEEE and ISSI. Shripad T. Revankar is a Professor of Nuclear Engineering and Director of the Multiphase and Fuel Cell Research Laboratory in the School of Nuclear Engineering at Purdue University, India. He received his B.Sc., M.Sc., and Ph.D. in physics from Karnataka University, India, and his M.Eng. in Nuclear Engineering from McMaster University, Canada. He also serves as BK21 Plus Visiting Professor in the Division of Advanced Nuclear Engineering at Pohang University of Science and Technology (POSTECH), South Korea. He has done pioneering work in niche technical areas including advanced reactor systems, reactor safety, reactor thermal hydraulics, composite fuel for advanced nuclear reactors, instrumentation, multiphase flow and heat transfer, microgravity multiphase flow, direct energy conversion, hybrid power systems, nuclear hydrogen generation, solar energy storage, packed bed reactor, renewable energy, and fuel cell technology. He has published more than 400 peerreviewed technical articles in archival scientific journals and conference proceedings and is the Author/Co-author of 7 books. He is a Life Member of the American Nuclear Society (ANS), American Society of Mechanical Engineer (ASME), American Institute of Chemical Engineers (AIChE), Korean Nuclear Society (KNS), and Indian Society for Heat and Mass Transfer (ISHMT). He received Technical Achievement Award from the American Nuclear Society Thermal Hydraulics Division in 2019 for his significant contributions to reactor thermal hydraulics through experiments and for the modeling of phenomena important in the analysis of nuclear reactor safety and applications.

vii

Review and Study of Solar String Inverters for a PV System G. B. Praveen, M. Satyendra Kumar Shet, and K. Latha Shenoy

Abstract A solar system is a setup that generates electricity by utilizing solar energy. Grid tied PV plants have the advantage of more effective utilization of generated power. Grid interconnection of PV systems is accomplished through the inverter, which converts dc power generated from PV modules to ac power used for power supply to electric equipments. Solar inverter system is therefore very important for grid connected PV systems. String inverter topologies have an important role in this system with complying to the code and standards. Keywords Solar energy · String inverter · Single phase ad three phase inverter

1 Introduction Energy security plays an important role in the economic growth for any country. India is one of the fastest growing economy country in the world. For economic growth, increasing prosperity, a growing rate of urbanization and rising per capita energy consumption has led to increased demand for energy in the country. Therefore in order to meet the challenge of the fastest growing economy renewable energy is a viable alternative source to meet the growing energy demand of our country. Realizing this fact, the Indian government has recently expressed an intention towards achieving 100 GW of solar capacity by 2022; out of which 40% is being expected through decentralized and roof top scale solar projects. In this context PV plants are required which produce electricity and reduce GHG emissions. Presently these PV plants are installed with govt. subsidy and the subsidy G. B. Praveen (B) Department of Electrical and Electronics Engineering, Yenepoya Institute of Technology, Moddbidri, India e-mail: [email protected] M. Satyendra Kumar Shet · K. Latha Shenoy Department of Electrical and Electronics Engineering, N.M.A.M. Institute of Technology, Nitte, Karnataka, India © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 S. Kajampady and S. T. Revankar (eds.), Advances in Renewable Energy & Electric Vehicles, Lecture Notes in Electrical Engineering 1083, https://doi.org/10.1007/978-981-99-6151-1_1

1

2

G. B. Praveen et al.

amount makes it lucrative because of reducing the pay back duration years. But even without subsidy, the pay back duration (more no. of years) and IRR equals will be less which make it financially viable. Therefore if the govt. continued to provide subsidy up to 30% to such projects for the next 2–3 years, then it will help to attract the domestic sector to install more PV plants on their rooftop. India being a tropical country lies between 8° and 37° north latitude and has an average annual temperature ranging between 25 and 27.5 °C, with about 300 clear sunny days in a year and the daily average solar energy incident over India varies from 4 to 7 kWh/m2 offering great potential for utilizing solar energy and being a long coastal line providing abundant wind energy throughout the year. In fact, the Indian govt. is concentrating on increased solar capacity within coming five years by promoting roof top scale solar and decentralized projects. For more than three decades our country has made considerable efforts to develop and evolve solar PV technology. The Indian Government is committed to clean energy and has undertaken the world’s largest program for producing 100 GW of grid connected solar power by 2022. The total solar energy absorbed by Earth’s atmosphere, oceans and land masses is app. 3,850,000 EJ per year. Wind is produced naturally when the sun heats the atmosphere, from the planet’s rotation, and from variations in the surface of the earth. As a result of the influence of bodies of water, forests, and meadows and other vegetation, and elevation changes, wind can be increased or decreased. Wind patterns and speeds vary significantly across terrain, as well as seasonally, but some of those patterns are predictable enough to plan around. Wind power or wind energy describes the process by which the wind is used to produce mechanical power or electricity. Wind turbines convert the kinetic energy in the wind into mechanical power. When sunlight falls on the solar panels it gets absorbed by the solar cells and the silicon semiconductors in the cells convert the solar energy into electric energy. This electric energy is in the form of DC power which can directly charge the battery. This DC power is sent to an inverter which converts it into AC power. This AC power is now sent to the mains in the home which in turn can power all necessary applications. A solar system is a setup that generates electricity by utilizing the solar energy system. A typical solar system consists of an inverter, mounting structure, batteries, grid box and balance of systems. A solar system comes in various sizes like 3, 5, 7.5, and 10 kW etc. The panels are connected in parallel to increase the available current flow while keeping the voltage output constant between each module. In the grid tie system, panels are wired in series which increases the voltage but does not increase the current. The string inverter converts 1–6 strings with an inverter. Realizing high power capacity that can be insulated in modular design & has MPPT for few strings. It continues to deliver industry leading reliability and technical innovation that lower the cost of the PV system installation. The main features of these inverters are endurance tested to 20 years operating life of reliability and easy installation. PV converters are semiconductor devices that convert part of incident solar radiations directly into electrical energy and solar cells are of crystalline silicon. Based on the working operation, PV systems are classified as follows

Review and Study of Solar String Inverters for a PV System

3

a. Stand alone systems; (off-grid PV systems) b. Hybrid systems c. Grid tied with battery backup system. Applications of these systems are as follows a. Rural household power supply b. Rural central power plants and c. Power supply for communication and lighting.

The performance of grid-connected PV systems can be evaluated by investigating the performance ratio, which is defined by the ratio of the system efficiency and the nominal efficiency of PV modules under STC.

2 Solar Inverter Inverters are of (a) Central (b) String (c) Micro and (d) multi-string types as shown in Fig. 1. The main solar inverters are • Central inverter—For high power application—100–500 KW. • String Inverter—For medium power application—3–20 KW—Residential application. • Module Inverter—For low power application—50–500 W.

Fig. 1 Types of string inverters

4

G. B. Praveen et al.

• Multistring inverter—High power common inverter—different DC—DC converter usage. String inverter: Each solar panel is connected in series to the string inverters. The inverter combines all the direct current received from each individual solar panel and, at once, converts it into alternating current. The number of solar panels that can be connected to a string inverter depends upon the input voltage rating of the inverter. String Inverters are of medium power type of 3–20 kW. It is made up of maximum six strings and requires one maximum power point tracker for few strings. Power capacity is depending upon number of strings. Design is of modular type and in case of faults, only partial power feeding capacity is lost. The challenging requirements are design and operations are high Efficiency, cost effective, with compliance to code and follow standards of practice. A PV inverter has to fulfil three main functions in order to free energy from a PV array into utility grid: • To separate the current into a sinusoidal waveform • To invert the current into an AC current • If the PV array voltage is lower than grid voltage, the PV array voltage has to be boosted with a further element. In PV systems using string inverters a number of PV modules are connected in series to form a string of up to 2−3 KW. In this power range the PV array voltage is usually between 150 and 450 V. Steps for providing SPV power system The main steps for providing the SPV power system are as follows Step 1: Measurement of Solar Radiation Step 2: Understanding Functional Requirements of SPV Power System Step 3: Considerations of Safety Measures Step 4: Designing of SPV Power System Step 5: Developing Site Layout Step 6: Carrying out Techno-Economic Analysis Step 7: Analysing limitations and scope Step 8: Developing Maintenance Schedule and its Compliance Step 9: Considering Future Extensions.

Review and Study of Solar String Inverters for a PV System

5

3 Survey Walker et al. [1] have discussed about the cascaded DC–DC converter connection of the PV Module. PV array is connected to the grid with the single DC–AC inverter and then connected to PV panels of string to the AC grid and proposed non-isolated per panel DC-DC converters connected in series to generate high voltage. This can be used in flexible voltage ranges in system. Cúk converter and buck-boost converter are more effective for this system. The main disadvantage is difficult to arranging a DC–DC converter on the panel substring and connecting the same arrangements to converters. Zhang et al. [2] proposed single phase grid system connected with the PV plant by using string inverters and discussed about its characteristics. It includes gridconnected single phase inverters, intelligent cluster controllers and user-friendly monitoring software. This system will be monitoring the inverters of PV modules and simultaneously gives high efficiency and maintenance of intelligent management. Rahim et al [3] proposed single phase multistring five level PV inverter topology which is used for the grid connected PV system by using the pulse width modulation switching scheme. These results are compared with the conventional 1–φ gridconnected PWM inverter multistring three-level and Savitha et al. [4] have discussed the design and simulation of the five level multi string inverter, which is applicable to the fuel cell by using the Proton Exchange Membrane Fuel Cell electrical model. It requires only six switches a to design an inverter and these methods, for determining the total harmonic distortion factor are less compared to the previous methods. The single phase multistring multilevel inverter system is the proposed system [5],where the single phase inverter based distributed energy resource is selected in the micro grid. To stabilize the DC output voltage and improve the efficiency, a high step-up converter is established and it acts as a front-end stage. The main drawbacks are THD are high compared to SVPWM and voltage fluctuations. Raju et al. [6] have discussed about harmonic reduction in Sinusoidal Pulse Width Modulation scheme using different types of Operational-Amplifier circuits or analog circuits. It will supply the power to standalone loads if the power is not sufficient to supply the grid. The filtering arrangement is far good if the carrier frequency increases much enough and less loss. But response can be made better by using the closed loop control feedback system. The feedback loop system is used for future work.. To develop a 50 KW grid tied solar inverter, Gen2 family of Silicon Carbide (SiC) power MOSFET devices are required [7]. The efficient new devices allowed the designers to develop a high power to weight ratio 1 KW/kg for an air cooled 50 kW 3-ph photovoltaic inverter with an MPPT boost function. The 50 KW interleaved boost circuit used was only 7 kg and peak efficiency of 99.4% while switching at 75 kHz. And it is limited to a certain power to weight ratio with the improvement of efficiency method being required. Keyhani et al. [8] introduced inverter with grid-tied multistring PV which is linked of high-frequency ac deals with soft switching, and high-frequency galvanic isolation.

6

G. B. Praveen et al.

An arbitrary number of PV strings is handled by single-stage topology in different parameters and locations. For each string dedicated MPPT is used as there may be dissimilarity in the levels of irradiance and working temperatures. The disadvantages are • • • •

Offers poor energy harvesting. The overall system performance has minimal impact on inverter failure. Ensures energy harvesting of optimum level and cost high-frequency. High frequency transformer is used.

The dual DC-DC converter structure is proposed to implement asymmetrical string inverter to use variously arranged PV strings and the converters are operated by independently running MPPT algorithms [9]. The hybrid power techniques can be used for these types of arrangement. Current and voltage sensors measure the output powers of each converter, where the obtained values are evaluated by P&O algorithm on MCU and the PWM signals are generated. At the same time, these values are transmitted to the computer by a microcontroller on optically isolated UART to USB unit. The received values are saved in a database for analysis. By using non-isolated DC-DC converter design, multilevel string inverter was modelled and implemented [10]. In distributed energy generation applications, the DC-DC converter and multilevel inverter are capable to be used. At the input a string consisting of two PV arrays can be managed by using a DC converter. To regulate PV string voltages, the dual buck converter is used at the input side, where converters are operating concurrently. DC converter output voltage is set to 120 V and coupling is done between the DC bus bar and supply. The unity power factor is obtained by the inverter. For future studies it is to be implemented in the grid system. Dogga et al. [11] discussed on different inverter topologies and their attributes such as Grid-connected/Stand-Alone Operation Capability, Isolation, Power Decoupling, Number of processing Stages, Dual Grounding Capability, Power Handling Capability, Components Count, Size, Wide Range of Operation capability, Cooling Requirement, Symmetrical operation in both half cycles, filter requirement on AC Side, and the Complexity Level of control strategy for solar applications were investigated and discussed in these topologies. As a final point, the rudimentary needs defined by the user have also been discussed, such as low cost, long lifetime, and high efficiency. In case of string configuration, two-stage inverters or single stage inverters with medium power handling capability are the best choice. Last, if several strings are to be connected to the grid, the multi-string concept seems to be the apparent choice.

4 Analysis of Single Phase and Three Phase Inverter The single phase single-phase full-bridge PWM inverter is simulated by using MATLab and analyses the performance of inverter. Let us consider the PWM inverter, the input DC voltage varies in the range of 295–325 V. Because of the low distortion

Review and Study of Solar String Inverters for a PV System

7

required in the output V 0 , ma ≤ 1.0. Its nameplate volt-ampere rating is specified as 2000 VA (1700 KWp), that is, V o1max I o1max = 2000 VA, where i, is assumed to be sinusoidal. If the output current is assumed to be purely sinusoidal, the inverter r.m.s voltampere output at the fundamental frequency equals Vo1max lo1max at the maximum rated output. With VT and IT as the peak voltage and current ratings of a switch, the combined utilization of the switches in an inverter is 16%. A full bridge single phase inverter is a switching device that generates a AC output voltage on the application of DC input by adjusting the switch turning ON and OFF based on the appropriate switching sequence, where the output voltage is generated. The simulation result of single phase inverter is verified and current and voltage output values are verified with theoretical values as shown in Fig. 2. The three phases of the full bridge inverter with the DC input voltage 313.97 V are connected to a balanced three phase star connected load consisting of the following values. f 1 = 47.619 Hz, triangular voltage, V tri = 1.0 V, Modulation index, ma = 0.95. The Resistance and inductance values per phase are 6y and L = 30 mH respectively. V(control) A = 0.95 cos (2π f 1t − 90◦ )V V(th A)1 = 74.76a−12.36◦ V (rms) Switching frequency, f s = 1 kHz, I A1 = 10a-30ºA (rms) (see Fig. 3). The simulation result of the three phase inverter is verified and current and voltage output values are as shown in Fig. 4.

5 Conclusion Various types of solar inverter are studied for a PV system. From the simulation studies carried out it analyses the basic single phase and three phase inverter and its performances. But the challenging task is that power can flow from the DC source to the grid and vice-versa by generating a sinusoidal voltage and frequency equivalent to that of the grid and also lagging or leading or in phase w.r.t grid. Also it generates the reference signal of (a) same frequency as grid (b) with desired phase angle and (c) without have harmonics. In addition, the key challenging task of design of this type of inverter is high efficiency, low cost, compliance to codes and standards which can be validated by studying few topologies of this type of inverter. With the above study approaches, the challenging task of design complying to the code and standards of the various topologies can be developed and validated.

8

Fig. 2 a Inverter—output current. b Inverter—PWM voltage

G. B. Praveen et al.

Review and Study of Solar String Inverters for a PV System

9

Id S1

S5

S3

R TH

L TH IR

R

VTH, R Y

Vd

n

B

S4

S2

S6 N

Fig. 3 Three phase inverter

Fig. 4 a Three phase current waveform. b Three phase voltage waveform—PWM voltage

10

G. B. Praveen et al.

Fig. 4 (continued)

References 1. Walker GR, Sernia PC (Jul 2004) Cascaded Dc–Dc converter connection of photovoltaic module. IEEE PESC, 1–7 2. Zhang X, Ni H, Yao D, Cao RX (May 2006) Design of single phase grid-connected photovoltaic power plant based on string inverters. In: IEEE industrial electronics and applications, pp 1–5 3. Rahim NA, Selvaraj J (Jun 2010) Multistring five-level inverter with novel PWM control scheme for PV application. IEEE Trans Ind Electron, 2111–2123 4. Savitha B, PriyaaGomathi A,Seyezhai R (Mar 2012) Design and simulation of five level multistring inverter for fuel cell applications. In: IEEE ICCEET, pp 472–476 5. Liao Y-H, Lai C-M (Sep 2011) Newly-constructed simplified single-phase multistring multilevel inverter topology for distributed energy resources. IEEE Trans Power Electron, 2386–2392 6. Raju NI, Islam MS, Ali T, Karim SA (May 2013) Sinusoidal PWM signal generation technique for three phase voltage source inverter with analog circuit & simulation of PWM inverter for standalone load & micro-grid system. In: IEEE ICIEV, pp 1–6 7. Mookken J, Agrawal B, Liu J (May 2014) Efficient and compact 50 KW Gen2 SiC device based PV string inverter. In: VDE PCIM, pp 1–7 8. Keyhani H, Toliyat HA (Aug 2014) Single-stage multi string PV inverter with an isolated high-frequency link and soft-switching operation. IEEE Trans Power Electron, 3919–3929 9. Kabalci E, Bayindir R, Gokkus G, Kabalci Y (Nov 2015) Dual DC-DC converter and monitoring interface for asymmetrical string inverters. In: IEEE ICRERA, pp 1580–1585 10. KabalcI E, Kabalci Y, Gokkus G (Nov 2015) Dual DC-DC converter design for string inverters used in solar plants. In: 4th international conference on renewable energy research and applications, Palenno, Italy, pp 22–25 11. Dogga R, Pathak MK (2019) Recent trends in solar PV inverter topologies. Sol Energy-J Int Sol Energy Soc 183:57–73

Review and Study of Solar String Inverters for a PV System

11

12. Kabalci E (2020) Review on novel single-phase grid-connected solar inverters: circuits and control methods. Sol Energy-J Int Sol Energy Soc 198:247–274 13. Lakhimsetty S, Shaik KM (2021) A comparative analysis of current control strategies for a solar based single-phase grid connected inverter. IEEE. 978-1-7281-5681-1/21

An Improved Strategic Analysis on Fault Diagnosis in Modular Multilevel Converter Sujo Oommen, Burri Ankaiah, Mahesh Kumar, M. H. Ananda, K. Narayan Swamy, and M. C. Rashmi

Abstract Modular multilevel converters are most popular among all converters due to not only their modular structure but also the power handling capacity. To ensure continuous availability and energy security, these converters are designed with fault diagnosis techniques. This paper presents a strategic review on key theories, models, and techniques utilized in the fault diagnosis strategies proposed by previous researchers till recently. In this paper, a detailed critical, comparative analysis of fault diagnosis strategies, and the major research gaps has been presented. A new classification of fault diagnosis methods based on system parameters which reveal fault characteristics and thus the presence of faults is also suggested. An experimental study was conducted to determine, understand the causes, sources, and effects of short circuit, and open circuit faults developed by the MOSFETS/IGBTs during their operation. Finally, the paper suggests a possible automated, fault diagnosis system that can not only be used in modular multilevel converters but also in any converters or systems. Index Terms Modular multilevel converters (MMCs) · Fault · Diagnosis · Modulation · Control · Automated · Strategic · Analysis · Classification

1 Introduction Centralized power generation system has been the back-bone of industrialization in the country. The centralized generation system has inherited problems such as increased power outages, service down time, and huge electricity losses. In addition, the depletion of fossil fuel sources has paved the way for a decentralized power generation system. The decentralized power generation system not only ensures S. Oommen (B) · B. Ankaiah · M. Kumar · M. H. Ananda · K. Narayan Swamy School of EEE, REVA University, Bengaluru, India e-mail: [email protected] M. C. Rashmi Electrical and Electronics, Government Polytechnic, Bantwal, India © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 S. Kajampady and S. T. Revankar (eds.), Advances in Renewable Energy & Electric Vehicles, Lecture Notes in Electrical Engineering 1083, https://doi.org/10.1007/978-981-99-6151-1_2

13

14

S. Oommen et al.

the reduction in carbon foot print but also ensures the availability but also energy security. Large-scale decentralized power generation systems have been integrated with renewable energy resources like solar, wind, small turbine/engines, and other power systems. Decentralized power generation has been the key tool to ensure stable and sustainable energy for remote regions and islands. Hence, there has been a great demand for advanced power electronic systems such as power converters, drive systems to meet the medium, and large power requirements of various applications. These applications include high voltage direct current (HVDC), electric vehicles (EVs), Tractions, and many more. For these applications, a class of converters, namely, modular multilevel converters (MMCs) are being preferred over other converters due to their modular structure [1].

2 Modular Multilevel Converter-MMC The concept of Modular Multilevel Converter-MMC has been introduced by Lesnicar and Marquardt [2] to the industry. The MMC technology can be easily configured as per the power requirement by integrating the modules known as sub-modules (SM). Each sub-module consists of standard power electronic devices, insulated gate bipolar transistors (IGBTs), and a capacitor across [2]. These converters are modular in structure, exhibit high quality output, clean waveform, and scalability. Figure 1 depicts the structure of the single phase MMC. It consists of two arms called ‘upper arm’ and ‘lower arm’ each having an equal number of submodules. The structure of the sub-module used in MMC construction is depicted in Fig. 2 [2]. The sub-module voltage Vsm is treated as a voltage-controlled source that ranges from 0 to Vc. The sub-modules have two switching devices with two antiparallel diodes as shown in Fig. 2. Therefore, the arm voltage (upper and lower) with ‘n’ number of submodules can be regulated by switching the appropriate number of sub-modules. The switching states and the corresponding sub-module voltage are summarized in Table 1. Thus, for the operation of MMC, the output voltage regulation is simple and straightforward [2]. Only one switch will be on at a time. The capacitor either charges or discharge depending on the direction of the current Ism [3]. It is essential to maintain the voltage across all sub-module capacitors to be at the same level such that the stress on all the power switching devices is equal [4]. To achieve this requirement, an output state controller has been employed with a controlled voltage balancing algorithm. This specialized algorithm is used to identify low voltage capacitors by measuring their voltages periodically (typically in the range of milli seconds) and switching ON the corresponding switches to charge them to equalize their voltage levels. Thus, the capacitor voltages of all the sub-modules in any arm can be balanced there by maintaining equal stress on all the switches. As described in Sect. 1.1, the sub-module is a controlled voltage source of magnitude Vsm varying from 0 to Vc. Hence, the output voltage level determines the required number of sub-modules for any MMC [5, 6].

An Improved Strategic Analysis on Fault Diagnosis in Modular …

15

Fig. 1 1Ø MMC structure [2]

Fig. 2 SM structure [2]

Table 1 Submodule controlled switching states Mode

S1

S2

Vsm

Ism

dVc/dt

1

ON

OFF

Vc

>0

>0

2

ON

OFF

Vc

i 2 p,min

3 Overview of AC/DC Topologies See Table 2.

4 Overview of DC/DC Topologies See Table 3.

Table 2 Compares and summarizes the PFC topologies

THD of output current

LCL active rectifier

Vienna rectifier

NPC converter

Totem pole PFC

Low

Very low

Very low

Low

Power density

High

High

Higher

High

Operation

Bidirectional

Unidirectional

Bidirectional

Bidirectional

High

High

Lower

Conduction loss Switching loss Efficiency

High

Mid

Low

Lower

High

Very high—(at high frequency)

High

Cost

Low

Mid

High

Low

Control

Easy

Mid

Mid

Easy

Mild

Difficult due to unsymmetrical loss distribution

Mild

Thermal management

60

S. Das et al.

Table 3 Compares and summarizes DC/DC topologies LLC converter

Phase-shifted full Dual active bridge (PSFB) bridge (DAB)

DAB in CLLC mode

Peak device High stress on both the side (primary and secondary side)

Very low

Low

High

KVA rating of transformer

High

Medium

Low

High

Power output to transformer

Low

Medium

High

Medium

Capacitor RMS currents (input and output)

High

Medium

Low

High

Operation

Unidirectional

Unidirectional

Bidirectional

Bidirectional

Conduction High losses of switches

Medium

Lowest

Medium

Turn ON-switching loss

ZVS

ZVS

ZVS

ZVS

Turn OFF-switching loss

Low (ZCS)

High

Device turn off at peak leakage inductor current value-High

Primary side turn off decided by g inductor current, secondary side turn off is zero due to ZCS-Low

Total losses

Very low

Higher

Medium

Low

Control complexity

Moderate

Very simple

Simple to complex

Moderate

Wide battery Needs additional Yes-with reduced With reduced voltage, fixed bus DC/DC efficiency efficiency-Yes voltage Stage-No

Limited range

Paralleling modules

Intensive

Easy

Easy

Intensive

Switching frequency

Si/SiC-fixed/ high

High

High

Very high

5 Conclusion In this paper, an overview of various topologies of the battery charger has been highlighted for the purpose of choosing better topology for the design of the EV DC fast charger. Normally the classification of the EV battery charger is distributed as the off-board and on-board charger. In this study, different schemes are presented for each level of the EV charger. Tables 1 and 2 described the comparison of four types of AC-DC and DC-DC converter topologies.

Review on Power Converter Topology for EV Fast Charger

61

Among all topologies, Totem Pole PFC of AC-DC stage and Dual Active Bridge (DAB) converter topology from DC-DC converter stage are more suitable for the DC fast charger based on the performance parameters of charger converter topologies mainly THD of output current, power direction, conduction loss, switching loss, power density and efficiency.

References 1. Harika S, Seyezhai R, Jawahar A (2019) Investigation of DC fast charging topologies for electric vehicle charging station (EVCS). In: TENCON 2019-2019 IEEE region 10 conference (TENCON) 2. Habib S, Khan MM, Abbas F, Sang L, Shahid MU, Tang H (2018) A comprehensive study of implemented international standards, technical challenges, impacts and prospects for electric vehicles. IEEE Access 6 3. Ronanki D, Kelkar A, Williamson SS (2019) Extreme fast charging technology—prospects to enhance sustainable electric transportation. Energies 12:3721. https://doi.org/10.3390/en1219 3721 4. Simorgh H, Doagou-Mojarrad H, Razmi H, Gharehpetian GB (2018) Cost-based optimal siting and sizing of electric vehicle charging stations considering demand response programmes. IET Gener Transm Distrib 12(8):1712–1720 5. Yilmaz M, Krein PT (2013) Review of battery charger topologies, charging power levels, and infrastructure for plug-in electric and hybrid vehicles. IEEE Trans Power Electron 28(5):2151– 2169 6. Prasad R, Namuduri C, Kollmeyer P (2015) Onboard unidirectional automotive G2V battery charger using sine charging and its effect on li-ion batteries. In: Energy conversion congress and exposition (ECCE), Montreal, Canada, pp 6299–6305 7. Tashakor N, Farjah E, Ghanbar T (2017) A bidirectional battery charger with modular integrated charge equalization circuit. IEEE Trans Power Electron 32(3):2133–2145 8. Tan KM, Ramachandaramurthy VK, Yong JY (2016) Integration of electric vehicles in smart grid: a review on vehicle to grid technologies and optimization techniques. Renew Sustain Energy Rev 53:720–732 9. Lozano JG, Milanés-Montero MI, Guerrero-Martínez MA, Romero-Cadaval E (2012) Electric vehicle battery charger for smart grids. Electr Power Syst Res 90:18–29 10. Anderson J, Peng FZ (Jun 2008) Four quasi-Z-source inverters. In: IEEE power electronics specialists conference, pp 2743–2749 11. Li Y, Peng F-Z (2011) Constant capacitor voltage control strategy for Z-source/quasi-Z-source inverter in grid-connected photovoltaic systems. Trans China Electrotech Soc 26(5):62–69 12. Zhang D, Lin H, Zhang Q, Kang S, Lu Z (2018) Analysis, design, and implementation of a single stage multi-pulse flexible-topology thyristor rectifier for battery charging in electric vehicles. In: 2018 IEEE transactions on energy conversion 13. Singh B, Singh BN, Chandra A, Al-Haddad K, Pandey A, Kothari DP (2004) A review of threephase improved power quality AC-DC converters. IEEE Trans Ind Electron 51(3):641–660 14. Nayak SK (2019) Electric vehicle charging topologies, control schemes for smart city application. In: 2019 IEEE transportation electrification conference (ITEC-India) 15. Shi C, Khaligh A, Wang H (2016) Interleaved SEPIC power factor preregulator using coupled inductors in discontinuous conduction mode with wide output voltage. IEEE Trans Ind Appl 52(4):3461–3471 16. Chen H, Wang X, Khaligh A (2011) A single stage integrated bidirectional AC/DC and DC/ DC converter for plug-in hybrid electric vehicles. In: IEEE vehicle power and propulsion conference, Chicago, USA, pp 1–6

62

S. Das et al.

17. Shi C, Tang Y, Khaligh A (2017) A single phase integrated onboard battery charger using propulsion system for plug-in electric vehicles. IEEE Trans Veh Technol 66(12):10899–10910 18. Aggeler D, Canales F, Zelaya-De La Parra H, Coccia A, Butcher N, Apeldoorn O (2010) Ultrafast DC-charge infrastructures for EV-mobility and future smart grids. In: Presented at the IEEE PES innovative smart grid technologies conference Europe (ISGT Europe), Gothenberg, pp 1–8. https://doi.org/10.1109/ISGTEUROPE.2010.5638899 19. Tu H, Feng H, Srdic S, Lukic S (2019) Extreme fast charging of electric vehicles: a technology overview. IEEE Trans Transp Electrific 5(4):861–878. https://doi.org/10.1109/TTE.2019.295 8709 20. Verma A, Singh B (2017) Three phase off-board bi-directional charger for EV with V2G functionality. In: 2017 7th international conference on power systems (ICPS), Pune, pp 145– 150. https://doi.org/10.1109/ICPES.2017.8387283 21. Liserre M, Blaabjerg F, Hansen S (Sep/Oct 2005) Design and control of an LCL-Filter-based three phase active rectifier. IEEE Trans Ind Appl 41(5) 22. Fang Y, Cao S, Xie Y, Wheeler P (2016) Study on bidirectional charger for electric vehicle applied to power dispatching in smart grid. In: Presented at the IEEE 8th international power electronics and motion control conference (IPEMC-ECCE Asia), Hefei, pp 2709–2713 23. Mortezaei A, Abdul-Hak M, Simoes MG (2018) A Bidirectional NPC-based level 3 EV charging system with added active filter functionality in smart grid applications. In: 2018 IEEE transportation electrification conference and expo (ITEC), Long Beach, CA, pp 201–206. https://doi.org/10.1109/ITEC.2018.8450196 24. Wu Q, Liu M, Zhou J, Ma G (2016) A 3000W super high efficiency rectifier for communications power supply. 2016 IEEE international telecommunications energy conference (INTELEC) 25. Tang Y, Ding W, Khaligh A (2016) A bridgeless totem-pole interleaved PFC converter for plug-in electric vehicles. In: 2016 IEEE applied power electronics conference and exposition (APEC) 26. Kim J, Lee J, Eom T, Bae K, Shin M, Won C (2018) Design and control method of 25 kW high efficient EV fast charger. In: Presented at the 21st international conference on electrical machines and systems (ICEMS), Jeju, pp 2603–2607 27. Rajendran G, Vaithilingam C, Prakash O (2019) Modeling of Vienna rectifier with PFC controller for electric vehicle charging stations. In: AIP conference proceedings, vol 2137(1). https://doi.org/10.1063/1.5120996 28. Chen S, Yu W, Meyer D (Mar 2019) Design and implementation of forced air-cooled, 140 kHz, 20 kW SiC MOSFET based Vienna PFC. In: Proceedings of the IEEE applied power electronics conference and exposition (APEC), pp 1196–1203 29. Taghizadeh S, Hossain MJ, Lu J (2015) Bidirectional isolated vehice to grid (V2G) system: an optimized implementation and approach. In: 2015 IEEE PES Asia-Pacific power and energy engineering conference (APPEEC), Brisbane, QLD, pp 1–5. https://doi.org/10.1109/APPEEC. 2015.7380912 30. Xue L, Shen Z, Boroyevich D, Mattavelli P, Diaz D (2015) Dual active bridge-based battery charger for plug-in hybrid electric vehicle with charging current containing low frequency ripple. IEEE Trans Power Electron 30(12):7299–7307. https://doi.org/10.1109/TPEL.2015. 2413815 31. Zahid ZU, Dalala ZM, Chen R, Chen B, Lai J (2015) Design of bidirectional DC–DC resonant converter for vehicle-to-grid (V2G) applications. IEEE Trans Transp Electr 1(3):232–244. https://doi.org/10.1109/TTE.2015.2476035 32. Moonem MA, Krishnaswami H (2012) Analysis and control of multi-level dual active bridge DC–DC converter. In: 2012 IEEE energy conversion congress and exposition (ECCE), Raleigh, NC, pp 1556–1561. https://doi.org/10.1109/ECCE.2012.6342628 33. Akagi H, Yamagishi T, Tan NML, Kinouchi S, Miyazaki Y, Koyama M (2015) Power-loss breakdown of a 750-V 100-kW 20-kHz bidirectional isolated DC–DC converter using SiCMOSFET/SBD dual modules. IEEE Trans Indus Appl 51(1):420–428. https://doi.org/10.1109/ TIA.2014.2331426

Review on Power Converter Topology for EV Fast Charger

63

34. Jung J, Kim H, Ryu M, Baek J (2013) Design methodology of bidirectional CLLC resonant converter for high-frequency isolation of DC distribution systems. IEEE Trans Power Electron 28(4):1741–1755 35. Tan L, Wu B, Rivera S (2015) A bipolar-DC-bus EV fast charging station with intrinsic DC-bus voltages equalization and minimized voltage ripples. In: IECON 2015–41st annual conference of the IEEE industrial electronics society, Yokohama, pp 002190–002195. https://doi.org/10. 1109/IECON.2015.7392426 36. Haritha AS, Jithin KJ (2019) An efficient resonant converter based charging scheme for electric vehicle. In: 2019 5th international conference on advanced computing and communication systems (ICACCS), Coimbatore, India, pp 599–603 37. Zhao H, Wang L, Chen Z, He X (2019) Challenges of fast charging for electric vehicles and the role of red phosphorus as anode material: review. Energies 12:3897. https://doi.org/10.3390/ en12203897 38. Junming Zeng, Guidong Zhang, Samson Shenglong Yu, Bo Zhang3 and Yun Zhang, “LLC Resonant Converter Topologies and Industrial Applications - A Review”, Chinese Journal of Electrical Engineering, Vol.6, No.3, September 2020 39. Dusmez S, Cook A, Khaligh A (2011) Comprehensive analysis of high quality power converters for level 3 off-board chargers. In: 2011 IEEE vehicle power and propulsion conference, Chicago, IL, pp 1–10. https://doi.org/10.1109/VPPC.2011.6043096 40. Feizi M, Beiranvand R (2020) An improved phase-shifted full bridge converter with extended ZVS operation range for EV battery charger applications. In: 2020 11th power electronics, drive systems, and technologies conference (PEDSTC), Tehran, Iran, pp 1–6. https://doi.org/ 10.1109/PEDSTC49159.2020.9088444 41. Feizi M, Beiranvand R (2020) Simulation of a high power self equalized battery charger using voltage multiplier and phase shifted full bridge converter for lithium-ion batteries. In: 2020 11th power electronics, drive systems, and technologies conference (PEDSTC), Tehran, Iran, pp 1–6. https://doi.org/10.1109/PEDSTC49159.2020.9088454

Mathematical Modeling and Analysis of Interleaved Two-Phase Boost PFC Dhanush Acharya, Suryanarayana K, Krishna Prasad, and L. V. Prabhu

Abstract Power factor correction (PFC) is a technique used to shape the input current of the power supply to be in synchronization with the mains voltage, in order to maximize the real power drawn from the mains. The paper deals with active PFC to shape the input current, thus harmonics are moved to much higher frequencies, making them much easier to filter out. Here an active interleaved two-phase boost converter topology is analyzed. The paper gives an in-depth study of current control and voltage control strategy of PFC. The developed boost converter converts rectified DC input voltage to regulated 400 V DC output voltage. The proposed converter is simulated in MATLAB/Simulink. Keywords PFC · Interleaved boost converter · Mathematical modeling · Current control · Voltage control

1 Introduction Power factor is defined as the ratio of real power to the apparent power provided by AC source. Real power is the energy consumed by the load to do the work and measured in watts. Reactive power is the results of reactive nature of the load. Here the power flows from back and forth between the source and load [1]. The power D. Acharya (B) · K. Prasad · L. V. Prabhu Hexomoto Controls Pvt. Ltd., Mysuru, India e-mail: [email protected] K. Prasad e-mail: [email protected] L. V. Prabhu e-mail: [email protected] S. K Department of Electrical and Electronics, NMAM Institute of Technology Affiliated to NITTE (Deemed to Be University), Nitte, Mangaluru, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 S. Kajampady and S. T. Revankar (eds.), Advances in Renewable Energy & Electric Vehicles, Lecture Notes in Electrical Engineering 1083, https://doi.org/10.1007/978-981-99-6151-1_5

65

66

D. Acharya et al.

factor takes values from zero to one. It plays a significant role in the design of electrical devices due to the regulations defined which define the minimum power factor or maximum level of harmonics a device must have. The two main causes of poor power factors are: 1. Displacement: Occurs when a circuit’s voltage and current waves are out of phase, due to the presence of inductors or capacitors. Displacement leads to circulation of excess RMS current than what is required to perform the work. 2. Distortion: Defined as the alteration of the wave’s original shape, caused by nonlinear circuits, such as rectifiers. These nonlinear waves have a lot of harmonic content, which distorts the current in the grid. AC/DC Converters are needed to power up DC load/ DC-DC converter from the utility supply. A commonly used cost-effective AC/DC converter is a diode rectifier circuit. Rectification is the process of converting the AC input voltage to DC output voltage. This can be achieved using a half-wave rectifier or a full wave rectifier with a filter capacitor. The full wave rectifier circuit consists of four diodes. Based on the polarity of input voltage, respective diodes in the bridge will conduct to get a rectified voltage. Further, this rectified voltage can be smoothened using an output filter capacitor to obtain DC power supply. This capacitor gets charged for a very short time, from the point where the voltage at the input of the capacitor is greater than the capacitor’s voltage. This creates a series of short current spikes in the input current. The waveshape of input current contains a large number of harmonics, consisting of integral multiples of the fundamental frequency. This distortion will lead to poor power factor which is undesired. Power factor correction (PFC) is a technique used to shape the input current of the power supply to be in synchronization with the mains voltage, to maximize the real power drawn from the mains. An ideal PFC circuit is one where the input current follows the input voltage without any input harmonics. To address displacement issues, external reactive components are commonly used to compensate for the circuit’s total reactive power. The distortion issues can be addressed using the following two options: 1. Passive power factor correction: It uses passive filters to remove harmonics and improves power factor and is used in low-power applications. 2. Active power factor correction: In this technique switching converters are used to shape the input current. The input current consists of harmonics in the order of switching frequency, which can be easily filtered out. This is considered the best PFC method but adds complexity to the design. This method alleviates some of drawbacks of passive power factor correction such as 1. Reduction of weight and size of the bulk inductor. 2. Output voltage of rectifier will change as the input supply changes. 3. Reduction of harmonic content but increases phase lag between current and voltages.

Mathematical Modeling and Analysis of Interleaved Two-Phase Boost PFC

67

The active PFC method changes the shape of current waveform such that current follows the voltage. In this way the harmonics are moved to much higher frequencies, making them much easier to filter out. Switching converters like buck and boost can be used in order to improve the power factor. Discontinuity in buck input current makes it less desirable to achieve unity power factor. In boost topology, input current will remain continuous when the converter is operating in CCM. Hence boost converter is the most commonly used topology in active PFC [2]. The conventional singlestage PFC controller uses a large inductor and requires substantial filtering to reduce high-frequency ripple. To address this, interleaved boost converter can be used [3]. The advantages of interleaved converters include: 1. Reduction of input ripple current with increase in ripple frequency which reduces differential mode EMI filter. 2. Significant reduction in output current ripple resulting in a reduced equivalentseries-resistance (ESR) loss of the output capacitor and reduction in capacitor volume. 3. Efficiency at lighter loads can be increased by employing phase shedding, i.e., by progressively turning off converters as the load is decreased. 4. Reduction of the inductor volume, and the current rating of the semiconductors.

2 System Overview The overall system consists of a full bridge rectifier which converts AC supply voltage to rectified DC voltage. This dc voltage is fed to an interleaved two-phase boost converter comprising of 2 inductors, diodes, and switches. The boost converter converts rectified DC input voltage to regulated 400 V DC output voltage. The output voltage, input voltage, and input current is fed to the controller. Based on the feedback signals controller generates the control signals for the switches. The DC output voltage from boost converter is fed to the load. The system block diagram is shown in Fig. 1.

3 Modeling of interleaved boost pfc As in Fig. 2, 2-Phase Interleaved PFC Boost converter consists of two inductors L1 and L2 , two diodes D1 and D2 , and two switches S1 and S2 . Both interleaved boost converters will have the same capacitor C at the output, Load is represented as R. Input voltage is represented as Vg (t), input current as ig (t), output voltage as v(t), output current as i(t). Both switches S1 and S2 will have the same switching frequency and phase shift of 180 degrees. The input current is the sum of two inductor currents iL1 and iL2 . Since the input ripple currents were 180 Degrees out of phase each other they cancel each other, thus reducing input ripple current. The best input inductor ripple current cancelation

68

D. Acharya et al.

Fig. 1 Overall circuit

Fig. 2 Equivalent model of interleaved PFC

occurs at 50% of duty cycle. The output ripple current iC is the sum of two diode currents minus the output current, which reduces the output capacitor ripple, as a function of duty cycle. As the duty cycle approaches 0, 50, and 100 percent duty cycle, the sum of two diode currents approaches DC. The capacitor has to filter only inductor ripple current. The interleaved boost converter can be operated in three different modes: Continuous Conduction Mode (CCM), Discontinuous Conduction Mode (DCM), and Critical Conduction Mode (CrCM) [4]. The switching waveforms and inductor currents in the four states of operation considering CCM are as in Fig. 3. Mode 1: 0 to T1 In this case, both switches S1 and S2 turn on. The inductor currents increase. Figure 4 illustrates the equivalent circuit model of mode 1.

Mathematical Modeling and Analysis of Interleaved Two-Phase Boost PFC

69

Fig. 3 Voltage and current waveforms

Fig. 4 Equivalent model of mode 1 operation

By applying Kirchhoff’s voltage and current law for mode 1 operation L 1 di L1 (t) = Vg dt

(1)

L 2 di L2 (t) = Vg dt

(2)

The input current ig (t) in mode 1 is represented as. i g (t) = i L1 (t) + i L2 (t)

(3)

70

D. Acharya et al.

The output current i(t) in mode 1 is expressed as. i (t) = i C (t) +

v(t) R

(4)

Rewriting Eq. 4: i (t) =

Cdvc (t) v(t) + dt R

(5)

where v(t) = vC (t), i(t) = 0. Thus, v(t) dv(t) =− dt RC

(6)

The differential equations are written in the form of a state-space representation technique. The inductor current iL1 (t), iL2 (t), and voltage across the capacitor vc (t) are taken as state variables. By setting L1 = L2 = L and defining new state variable iL = iL1 + iL2 , state equation is written as L 1 di L1 (t) = Vg dt

(7)

L 2 di L2 (t) = Vg dt

(8)

Ldi L = 2Vg dt

(9)

4 Mode 2: T1 to T2 In this case, switch S1 is ON and S2 is off. Figure 5 illustrates the equivalent circuit model of mode 2. By applying Kirchhoff’s voltage and current law for mode 2 operation L 1 di L1 (t) = Vg dt

(10)

L 2 di L2 (t) = Vg − v(t) dt

(11)

The input current ig (t) in mode 2 is represented as i g (t) = iL1 (t) + iL2 (t)

(12)

Mathematical Modeling and Analysis of Interleaved Two-Phase Boost PFC

71

Fig. 5 Equivalent model of mode 2 operation

The output current i(t) in mode 2 is expressed as. i (t) = i C (t) +

v(t) R

(13)

Rewriting Eq. 13: i (t) =

Cdvc (t) v(t) + dt R

(14)

where v(t) = vC (t), i(t) = iL2 (t) = iL (t) v(t) i(t) Cdvc (t) = + dt R 2

(15)

i L (t) vc (t) Cdvc (t) = − dt 2 R

(16)

Thus,

Arranging the above equations, Ldi L = 2Vg − vc (t) dt

5 Mode 3: T2 to T3 In this case, both switches S1 and S2 turn on. This mode is same as mode 1.

(17)

72

D. Acharya et al.

6 Mode 4: T3 to T4 In this case, switch S1 is off and S2 is on. Figure 6 illustrates the equivalent circuit model of mode 4. By applying Kirchhoff’s voltage and current law for mode 4 operation L 1 di L1 (t) = Vg − v(t) dt

(18)

L 2 di L2 (t) = Vg dt

(19)

The input current ig (t) in mode 4 is represented as. ig (t) = iL1 (t) + iL2 (t)

(20)

The output current i(t) in mode 2 is expressed as. i (t) = i C (t) +

v(t) R

(21)

Rewriting Eq. 21: i (t) =

Cdvc (t) v(t) + dt R

(22)

where vC (t) = v(t), i(t) = iL1 (t). Thus, i (t) vc (t) Cdvc (t) = − dt 2 R Arranging the above equations,

Fig. 6 Equivalent model of mode 4 operation

(23)

Mathematical Modeling and Analysis of Interleaved Two-Phase Boost PFC

Ldi L = 2Vg − vc (t) dt

73

(24)

The state variations will provide information about the system at every instant. The states of four modes are averaged over a switching cycle to obtain an averaged state-space model to study the converter behavior. For the converter under consideration, state, input, and output vectors are defined as: | | i L (t) x(t) = (25) vc (t) | | u(t) = vg (t) (26) y(t) = [v(t)]

(27)

The standard form of writing the differential equation of the linear circuit is K

d x(t) = Ax(t) + Bu(t) dt

(28)

y(t) = C x(t) + Eu(t)

(29)

where K is a constant matrix given by | K =

L 0 0C

| (30)

The linear first-order differential equation of mode 1 operation can be represented in state-space form as ⎡o⎤ || | | | | i iL 2 ⎣ L ⎦ = 0 01 + Vg o vc 0 0 −R v

(31)

c

where | A1 =

| | | 0 0 2 ; B = 2 0 − R1 0

Similarly mode 2 equations can be written as

(32)

74

D. Acharya et al.

⎡o⎤ || | | | | i iL 2 ⎣ L ⎦ = 0 01 + Vg o v 0 0 − c v R

(33)

c

where | A2 =

| | | 2 0 −1 ; B = 2 1 1 − 0 2 R

Since mode 3 is same as mode 1: | | | | 2 0 0 ; B = A3 = 3 0 0 − R1

(34)

(35)

Mode 4 equations in state-space form can be written as ⎡o⎤ || | | | | i iL 2 ⎣ L ⎦ = 01 −11 + Vg o −R vc 0 v 2

(36)

c

where | A4 =

| | | 0 −1 2 ; B = 4 1 1 − 0 2 R

(37)

The output equation can be written as | [v(t)] = [01]

| iL + [0]Vg vc

(38)

The presence of nonlinearities in the model due to switching harmonics can be minimized using averaging. The averaged state-space equations can be obtained by weighing state matrices as ) ( 1 1 − d ' A1 + d ' A2 + 2 2 ) ( ( 1 1 − d ' B1 + d ' B2 + B= 2 2 ) ( ( 1 1 ' ' − d C1 + d C2 + C= 2 2 ) ( ( 1 1 − d ' E1 + d ' E2 + E= 2 2 (

A=

) − d ' A3 + d ' A4

(39)

) − d ' B3 + d ' B4

(40)

) − d C3 + d ' C4

(41)

) − d ' E3 + d ' E4

(42)

'

Mathematical Modeling and Analysis of Interleaved Two-Phase Boost PFC

75

Since A1 = A3 and A2 = A4, the equations can be simplified and written as: ) ( A = 1 − 2d ' A1 + 2d ' A2

(43)

) ( B = 1 − 2d ' B1 + 2d ' B2

(44)

) ( C = 1 − 2d ' C1 + 2d ' C2

(45)

) ( E = 1 − 2d ' E 1 + 2d ' E 2

(46)

By substituting mode 1 and mode 2 equations in the above equation, the averaged model can be written as | | | | 2 0 −2d ' ;B = A= d ' −1 0 R | | 01 C= ; E = [0] (47) 10 At equilibrium, the state vectors and output vectors can be obtained as X = A−1 BU /\

(48)

} {( )−1 s I − K −1 A K −1 B + E u (s) {( } )−1 + C s I − K −1 A K −1 F + G d (s)

y (s) =C

/\

/\

(49)

where F = ( A1 − A2 )X + (B1 − B2 )U G = (C1 − C2 )X + (E 1 − E 2 )U By using above equations, control to output transfer functions can be written as G vd =

2Rd '2

L Vg s RVg − '2 2 '2 + R LCs + Ls 2d (2Rd + R LCs 2 + Ls)

(50)

The control to inductor current transfer function can be written as G id =

Vg (RCs + 1) Vg + d'(2Rd '2 + R LCs 2 + Ls) d'(2Rd '2 + R LCs 2 + Ls)

(51)

76

D. Acharya et al.

Fig. 7 PFC control loop

The voltage to current transfer function can be written as G vi L =

G vd G id

(52)

The control loop implemented in the system consists of an inner current control loop and outer voltage control loop. The inner loop is inductor current control loop and outer loop is voltage control loop [5]. The output voltage of boost circuit is compared with a reference voltage and error signal is fed to PI controller. The output of PI controller is multiplied with the rectified voltage to generate reference inductor current, which has the same phase as rectified voltage. The purpose of inner control loop is to force the inductor current to follow the voltage phase [6]. The feedforward technique is used to make output voltage independent of input voltage variations [7]. The overall structure of control loop design is shown in Fig. 7.

7 Simulation and results The boost converter is designed for switching frequency of 200 kHz. A switching pulse of 180 degrees of operation is obtained using MATLAB/Simulink. The ratings of the converter is given in Table 1. The PFC boost circuit is analyzed using MATLAB Simulink as in Fig. 8. The circuit consists of AC input source voltage followed by a bridge rectifier circuit to converter AC input voltage to DC. The DC output of the rectifier is fed to the

Mathematical Modeling and Analysis of Interleaved Two-Phase Boost PFC Table 1 System ratings

77

Description

Symbol

Value

Input voltage

Vg

100−260 V

Input frequency

f

47−63 Hz

Output voltage

V

390 V

Inductor

L

150uH

Capacitor

C

300uF

Fig. 8 MATLAB simulink model

interleaved boost converter. The gate control pulses for two switches are derived using PFC control subsystem derived using control loop. The bode plot and step response of current control loop is shown in Figs. 9 and 10 respectively. The bode plot and step response of voltage control loop is shown in Figs. 11 and 12, respectively. The sinusoidal input current and voltage waveform are as in Fig. 13. It is seen that input current and voltage are in phase with each other. The inductor current iL1 and iL2 are observed and found that they are 180-degree phase shift. The inductor current has a frequency twice that of switching frequency and having current ripple smaller. This is illustrated in Fig. 14. To analyze the system at different loading conditions at t = 0.01 s a step voltage of 400 V is applied at a light load of 800Y. It is seen that output voltage and current increase to the rated value without any overshoot and settles at t = 0.1 s. At t = 0.5 s, the system load is varied from 800Y to 400Y. It is observed from the output voltage and current waveforms in Fig. 16, the output voltage drops by nearly 30 V and then increases to reach the stable value of 400 V. At the same time current shoots up and then drops; and then increases to the rated value within 0.3 s. The results were shown in Fig. 15.

78

Fig. 9 Bode plot of current control loop Fig. 10 Step response of current control loop

D. Acharya et al.

Mathematical Modeling and Analysis of Interleaved Two-Phase Boost PFC

Fig. 11 Bode plot of voltage control loop Fig. 12 Step response of voltage control loop

79

80

Fig. 13 Input voltage and current waveform

Fig. 14 Inductor current waveforms

D. Acharya et al.

Mathematical Modeling and Analysis of Interleaved Two-Phase Boost PFC

81

Fig. 15 Output voltage and current waveform

8 Conclusion In this paper, the operation modes of 2 phase interleaved boost PFC are analyzed. Interleaving boost PFC mathematical modeling and MATLAB Simulink simulation results were presented. The inclusion of control loop makes the system more stable and improved setting time for step response. As future work, it is intended to implement the similar control loop to 3 kW system. Acknowledgements The author would like to thank M/s Hexmoto controls Pvt Ltd., Mysuru for giving encouragement to all these activities. The author would like to thank Nitte Institution for providing the opportunity to this work. The author would express sincere gratitude to Mr. Anup Shetty and Mrs. Swathi Hatwar for guidance and constructive feedback after reviewing this paper.

References 1. Active power factor correction-basics, application note AN-53. Power Integrations 2. Power factor correction (PFC handbook), choosing right power factor controller solution, on semiconductor 3. Choudhury S, Noon JP (2005) A DSP based digitally controlled interleaved PFC converter. In: Twentieth annual IEEE applied power electronics conference and exposition. APEC 2005, vol 1. pp 648–654. https://doi.org/10.1109/APEC.2005.1453016 4. Nussbaumer T, Raggl K, Kolar JW (2009) Design guidelines for interleaved single-phase boost PFC circuits. IEEE Trans Industr Electron 56(7):2559–2573. https://doi.org/10.1109/TIE.2009. 2020073 5. Digital PFC CCM boost converter 300 W design example using XMC1400 microcontroller, Application Note, Infineon Technologies 6. Chen M, Sun J (2006) Feedforward current control of boost single-phase PFC converters. IEEE Trans Power Electron 21(2):338–345. https://doi.org/10.1109/TPEL.2005.869746

82

D. Acharya et al.

7. Huang Y-T, Yang C-C, Li T-S, Chen Y-M (2021) A feedforward voltage control strategy for reducing the output voltage double-line-frequency ripple in single-phase AC–DC converters. IEEE J Emerg Sel Top Power Electron 9(6):6605–6612. https://doi.org/10.1109/JESTPE.2021. 3083258 8. Two-phase interleaved PFC converter w/power metering test results. Texas Instruments 9. The 5KW, two-channel, interleaved CCM PFC EVAL board. Infineon Technologies 10. PFC boost converter design guide. Application note, Infineon technologies 11. Interleaved power factor correction (IPFC) using the dsPIC DSC. Application note AN1278, Microchip Technology

Design of a Solar Battery Charger with Maximum Power Point Tracking Adrian Dsilva, Samith Suvarna, Laxmisha G. Ballal, and H. Swathi Hatwar

Abstract In recent times, an immense amount of effort is put into moving from conventional sources of energy to renewable sources of energy, such as solar, wind, hydroelectricity, and so on. Solar energy is more advantageous in comparison to other sources due to lower maintenance costs and reduced plant size. However, due to the large variation in irradiance, generating power efficiently from the solar panel is a challenge. A suitable MPPT algorithm can be implemented to increase the power obtained from the solar panel. This paper describes the design of a solar battery charger that utilizes a buck converter with the duty ratio controlled based on the Perturb and Observe MPPT algorithm.

1 Introduction As the global resources of conventional sources such as coal and fossil fuels are rapidly depleting and the global energy demand is ever-increasing, the search for alternate, renewable sources of energy has been promoted and expanded. However, an efficient and cost-effective method of generating electricity using renewable energy sources is still in development. One of the most promising methods of generating renewable energy is by using solar panels. Solar panels convert the solar radiation incident on it into electric energy through the photovoltaic effect. Since solar PV systems have little to no moving parts, the maintenance cost is less and the system life time is longer. However, solar PV systems have a few hindrances that have to be overcome. While the maintenance cost is low, the initial investment cost is high. Solar panels are also not very efficient, as the average efficiency is about 15–20%, while the maximum recorded efficiency is 47.1% [1]. The output power is also dependent on the irradiance and weather conditions, thus the power generated during cloudy weather is less in comparison to power generated during sunny weather. In order to A. Dsilva (B) · S. Suvarna · L. G. Ballal · H. Swathi Hatwar Department of Electronics and Electrical Engineering, N M A M Institute of Technology (Affiliated to Nitte DTU), Nitte, Udupi, India © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 S. Kajampady and S. T. Revankar (eds.), Advances in Renewable Energy & Electric Vehicles, Lecture Notes in Electrical Engineering 1083, https://doi.org/10.1007/978-981-99-6151-1_6

83

84

A. Dsilva et al.

improve the efficiency of the solar panel and the power generated, Maximum Power Point Tracking (MPPT) algorithms are implemented. As the name suggests, the Maximum Power Point Tracking (MPPT) algorithm changes the operating point of the solar panel in a manner that will maximize the power output, thus increasing the efficiency of the solar panel. The graph shown in Fig. 1, simulated in MATLAB Simulink, shows the relation between the panel voltage and panel current of a 17 V, 75W solar panel at 25 °C and for irradiances of 1000, 700, and 100W/m2 . The variation of panel output power with respect to the panel voltage for the same parameters is indicated in the graph below (Fig. 2). This paper explains the design and use of a buck converter to step down the panel voltage and charge a 12 V lead-acid battery, and the implementation of Perturb and Observe MPPT algorithm to obtain maximum output power from the panel. The circuit designed in this paper constitutes the battery charging circuit for a Solar Street Light project.

Fig. 1 Plot of panel current versus panel voltage

Fig. 2 Plot of panel power versus panel voltage

Design of a Solar Battery Charger with Maximum Power Point Tracking

85

2 System Overview The solar panel is selected such that it should provide sufficient energy during the day so as to fully charge a lead-acid battery with a nominal voltage of 12 V, and a battery capacity of 40Ah. In this project, a 75W solar panel is selected, assuming that due to irradiance variation the average power of about 60W, and assuming a charging current of about 4A at 14 V, the time taken to charge the battery would be about 10 h. The specifications of the panel are given below (Table 1). Based on the above specifications, it is required to step down the voltage from the solar panel to the charging voltage of the battery efficiently. A linear regulator would not be able to provide the required current without significant conduction losses. Hence, in this project, a buck converter circuit is to be designed with the following system specifications shown in Table 2. The switching frequency must be above the audible frequency range (above 20 kHz). As the switching frequency increases, the size of the inductor and capacitor required decreases. In this project, the switching frequency is taken as 100 kHz, and the ripple voltage is assumed to be 1%, so that the charging voltage remains almost constant. As the ripple current and the size of the inductor vary inversely, a ripple Table 1 Solar panel specifications Description

Abbreviation

Specification

Open circuit voltage

Voc

21.8 V

Short circuit current

Isc

4.9 A

Maximum power point voltage

Vpm

17 V

Maximum power point current

Ipm

4.4 A

Maximum power

Pmax

75 W

Table 2 Buck converter design specifications Description

Abbreviation

Specification

Nominal voltage of solar panel(input)

Vg

17 V

Nominal voltage of battery

V

12 V

Switching frequency

fs

100 kHz

Input power from panel

P

75W

Ripple voltage

/V

1%

Ripple current

/I

30% IL

Efficiency of circuit

η

90%

Input current

Ii

Ii =

P Vg

=

75 17

= 4.4 A

Required duty cycle

D

D=

V0 Vi

=

12 17

= 0.7D

Load current

Io

I0 = I L = η ×

P V

= 5.625A

86

A. Dsilva et al.

current in the range of 20–40% of the inductor current is generally considered for calculations. In order to implement a Maximum Power Point Tracking algorithm, it is necessary to measure the panel voltage(VPV ) and current (IPV ). Monitoring the battery voltage and ensuring that it does not over-charge is also an essential feature of the battery charger. Hence, the voltage to be measured is stepped down using a potential divider and is fed to the ADC of the microcontroller after signal processing. The current is obtained by measuring the voltage across a series sense resistor of known value. This voltage is fed to the ADC of the microcontroller after signal processing, and the current is calculated by dividing the measured voltage by the resistance of the sense resistor.

3 Circuit Design The overall system mainly consists of the buck converter circuit, Dual MOSFET driver, measurement and control circuit, solar panel, and the battery. The block diagram of the battery charger circuit is as shown in Fig. 3.

3.1 Selection of Buck Converter Components The value of the inductor depends on the ripple current and the switching frequency. As the switching frequency increases the value of inductance required decreases. For the given circuit the inductance value is given by L=

V (1 − D) = 21.3 μH = 22 μH /I L × f s

The capacitor in the buck converter maintains the output voltage when the top switch is in open condition. The value of capacitance is selected based on the maximum allowed ripple voltage and is calculated using the formula below. C=

/I L = 17.57 μF 8 × f s × /V

Hence we take two 10μF capacitors in parallel for an equivalent capacitance of 20μF. It can be noted that the inductor and capacitor values should be selected higher than the calculated value to obtain lower ripples in current and voltage, respectively.

87

Fig. 3 Block diagram of solar battery charger

$

(

$

(

Design of a Solar Battery Charger with Maximum Power Point Tracking

88

A. Dsilva et al.

3.2 Selection of MOSFETs The voltage rating of the MOSFETs must be at least twice greater than the maximum voltage appearing across the MOSFET, in order to prevent any voltage spikes from damaging it. The current rating must also be at least twice greater than the load current. The conduction losses of the switch can be reduced by using a switch with low ON-state resistance(RDS(on) ). The gate charge (Qg ) influences the rise time and fall time of the MOSFET, as well as the selection of the MOSFET driver. As the gate charge(Qg ) increases, the driver is required to source/sink more current in order to keep up with the switching frequency. Thus, the gate charge(Qg ) must be minimum. As electrons have higher mobility than holes, n-channel MOSFETs are preferred for switching applications. The temperature range of the MOSFET is also an important consideration in the selection of the MOSFET. Keeping these factors in mind, NTMFD5C680NLT1G [2], a Dual N-channel MOSFET is selected. The specifications are given in the table below.

3.3 MOSFET Gate Driver Selection For this configuration, a synchronous buck driver is preferred, with a high output source/sink current to quickly charge/discharge the MOSFET capacitance. The source/sink current must be high enough in order to ensure that the rise time and fall time of the MOSFETs are much lower than the switching time period. As the input PWM signals from the microcontroller [3] have a maximum voltage level of 3.3 V, the driver must have an input rising threshold less than 3.3 V. The NCP81075 [4] is a high-performance dual MOSFET gate driver optimized to drive the gates of both high and low-side power MOSFETs in a synchronous buck converter. It can operate with a switching frequency of up to 1 MHz and it can source/ sink current up to 4A. The low-side and high-side can be independently controlled, thus asynchronous mode is also possible by turning off the low-side switch. When using half-bridge configurations, it is necessary to generate high-side bias to drive the gate of the high-side MOSFET which is referenced to the switch node. A simple and cost-effective method to achieve this is by using a bootstrap circuit, which consists of a capacitor, a diode, a resistor, and a bypass capacitor. The calculations for the bootstrap circuit are given below (Table 3). A Zener diode is connected in parallel to the bootstrap capacitor to limit the voltage across it.

Design of a Solar Battery Charger with Maximum Power Point Tracking

89

Table 3 Bootstrap circuit calculations Bootstrap capacitance Bootstrap resistance

CBoot ≥ 10 × C g =>CBoot = 5.94nF = 10nF τ =R Boot ×C Boot D R Boot ×10n 1 500k = 0.7

R Boot = 140Y = 150Y

3.4 Reverse Polarity Protection For reverse polarity protection on the solar panel side, a P-channel MOSFET is used in the configuration shown in the schematic diagram. The MOSFET is to be selected with a voltage rating at least twice the maximum voltage across the switch. As the MOSFET will be in the ON-state for long periods of time, it should have a higher continuous current rating, and low RDS (on) so that the conduction losses are less. For this application, SQM110P06-8m9L [5] P-channel MOSFET is selected. A Zener diode is connected across it to limit the gate-to-source voltage.

3.5 Schematic Diagram of the Battery Charger The schematic of the battery charger circuit is illustrated in Fig. 5. It can be observed that there are a few additional components used in the circuit. The capacitors C13, C17, and C19 are decoupling capacitors for the gate driver IC. R4 and R8 are gateto-source resistors connected so as to discharge the gate charge of the MOSFET at a faster rate. R30 and R36 are pull-down resistors, which ensure that the signals reaching the MOSFET driver do not go into an unknown/invalid state, which may cause false triggering of the MOSFETs. A 10A fuse is provided at the output for over-current protection. R1 and R2 are sense resistors used while measuring the current.

4 Voltage and Current Measurement In this project, the microcontroller ADC operating voltage is in the range of 0–3.3 V [3]; therefore, the measured data should be scaled down to within the above range.

Fig. 4 Schematic diagram

90 A. Dsilva et al.

Design of a Solar Battery Charger with Maximum Power Point Tracking

91

Fig. 5 Circuit for battery voltage measurement

4.1 Battery Voltage Measurement A potential divider circuit is used to scale the voltage to below the microcontroller’s operating voltage (3.3 V). A voltage buffer is used for impedance matching and improving the accuracy of measurement. A low pass filter circuit with upper cutoff frequency of 10 kHz is used to remove any noise due to the switching circuit. Considering the maximum battery voltage of 14 V, Vbat is assumed to be 16 V(worst case) for calculations (Fig. 5).

4.2 Panel Voltage Measurement Similar to the previous circuit, a potential divider is used to bring the voltage below the microcontroller’s operating voltage (3.3 V). A voltage buffer is used for impedance matching and in order to improve the accuracy, and a low pass filter circuit with an upper cut-off frequency of 10 kHz is used to remove any noise due to the switching circuit. Considering the maximum panel voltage and switching transients, VPV is assumed to be 30 V (worst case) for calculations (Fig. 6).

4.3 Current Measurement To measure current, the voltage is measured across a sense resistor which is placed in series with the branch where current is to be measured. The terminals of the sense resistor are connected to INA2180 [6] current sense amplifier, which is basically a differential amplifier with fixed gain. It must be ensured that the common mode voltage should be within the range of the current sense amplifier. The output signal is then fed to the ADC of the microcontroller (Fig. 7).

92

A. Dsilva et al.

Fig. 6 Circuit for panel voltage measurement

5 Main Algorithm and MPPT Algorithm In the microcontroller program, the voltage and current data obtained from the signal processing circuits have to be read using the ADC and scaled back to its original value. Before running the MPPT algorithm, it is necessary to check if the battery is already charged, in order to prevent over-charging. In order to increase the lifespan of the solar panel, the minimum panel voltage must be above 60% of the open circuit voltage, in this case 0.6VOC = 0.6*21 = 15 V. As the name implies, the Perturb and Observe Algorithm creates a minor perturbation by shifting the operating point, which causes variation in the output power of the panel. The perturbation is created by varying the duty cycle of the buck converter in small steps. The panel voltage and current are measured periodically, from which the panel power is calculated and compared with the previous value. If the panel power has increased, the perturbation is continued in the same direction. If the panel power has decreased, the perturbation is reversed. This process is repeated until the operating point reaches the MPP. In this method, the MPP can be reached without any prior knowledge of the panel parameters, however as there is no indication when the MPP is reached, the operating point tends to oscillate about the MPP. The flowchart for the main program is illustrated in Fig. 8 and the MPPT algorithm is illustrated in Fig. 9.

Design of a Solar Battery Charger with Maximum Power Point Tracking

Fig. 7 Circuit for panel and battery current measurement

93

Fig. 8 Main algorithm

94 A. Dsilva et al.

Design of a Solar Battery Charger with Maximum Power Point Tracking

95

Fig. 9 P and O MPPT algorithm

Layout The PCB is a 2 layer board with signal and power ground planes at the bottom layer. The size of the board is 110 mm X 75 mm. The board also contains connectors for interfacing the LED Driver Board and I2C LCD display, which are the other parts of the Solar Street Light project (Figs. 10 and 11).

96

A. Dsilva et al.

Fig. 10 Board layout

Fig. 11 3D view of board

Acknowledgements We are grateful to Dr. Suryanarayana K., for the opportunity to work on this project, as well as his motivation and guidance. Special thanks to Mr. Anup Shetty and Mr. Ravikiran Rao for their guidance and constant support.

Design of a Solar Battery Charger with Maximum Power Point Tracking

97

References 1. Geisz JF, France RM, Schulte KL, Steiner MA, Norman AG, Guthrey HL, Young MR, Song T, Moriarty T Six-junction III–V solar cells with 47.1% conversion efficiency under 143 Suns concentration. Nat Energy 2. ON Semiconductor, NTMFD5C680NL MOSFET–power, dual, N-Channel. Datasheet 3. Microchip, PIC24FJ64GP205/GU205 16-Bit eXtreme low-power microcontrollers with USB in low pin count packages. Datasheet 4. ON Semiconductor, NCP81075 dual MOSFET gate driver, High performance. Datasheet 5. Siliconix V Automotive P-Channel 60V(D-S) 175°C MOSFET. Datasheet 6. Texas Instruments, INAx180 low-and high-side voltage output, Current-Sense Amplifiers . Datasheet

Implementation of Edge Computing Model for the Processing of Data in Mines K. Aneesha Acharya , Akshit Gaurav, and Aman Srivastava

Abstract Mining is one of the most dangerous industries. The risk factor involved in underground mining is exponentially higher than in other forms of mining, with the risk of mine collapse, hazardous gases, ventilation problems, mine inundation, etc. Internet of Things (IoT) can be a viable solution to monitor the working conditions of miners as well as in evacuation and rescue procedures in case of a mishap. This paper proposes a module (Rakshak module) that senses conditions inside a coal mine and sends data to managers above ground to monitor working conditions inside mines. The sensor module is attached to the suit’s collar neck, which workers wear in mines. This sensor data consists of carbon level, heart rate, oxygen saturation, temperature, and humidity are recorded and used for future trends and risk analysis. In order to tackle the problem of the slow data transmission rate, an edge computing model is used wherein multiple localized servers are placed across the mines and process the data on the network’s periphery. The proposed communication method reduces the time lag and logistical hassle involved in data transfer in a harsh environment. This will increase the overall safety of the miner. Also, the simulation results of the developed Rakshak module are predicted to reduce the number of accidents and mishaps inside mines. Keywords Coal mining · Sensor · Heart rate · Edge computing · Internet of things

K. Aneesha Acharya (B) · A. Gaurav Department of Instrumentation and Control Engineering, Manipal Institute of Technology, Manipal Academy of Higher Education, Manipal, India e-mail: [email protected] A. Gaurav e-mail: [email protected] A. Srivastava Department of Electrical and Electronics Engineering, Manipal Institute of Technology Manipal Academy of Higher Education, Manipal, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 S. Kajampady and S. T. Revankar (eds.), Advances in Renewable Energy & Electric Vehicles, Lecture Notes in Electrical Engineering 1083, https://doi.org/10.1007/978-981-99-6151-1_7

99

100

K. Aneesha Acharya et al.

1 Introduction Mining is one of the most dangerous industries. The risk factor involved in underground mining is exponentially higher than in other forms of mining, with the risk of mine collapse, hazardous gases, ventilation problems, mine inundation, etc. India is the world’s second-largest coal producer, producing around 716 million tons in 2021 [1]. Coal thermal power plants fulfill around 75% of India’s electricity needs [2]. The scale, size, and importance of the coal industry in India mean that the number of workers and miners employed in dangerous underground mines is very high. Around 485,000 coal miners produce more than 700 million tons annually [3]. These miners’ health is compromised as they work in harsh conditions with increased chances of pulmonary diseases and lung problems. The leading harmful gases in coal mining are Carbon Monoxide (CO) and Carbon Dioxide (CO2 ). While CO is flammable, CO2 is a suffocating gas [4]. Coal miners are also subjected to high temperatures and humidity in some cases. Mishaps such as mine collapse or inundation are also dangerous in underground coal mines [5]. The proposed Rakshak module features sensors incorporated in a robust module powered by swappable 9 V batteries to measure ambient temperature, humidity, and concentration of carbon monoxide and flammable gases around the miner, along with the physical conditions such as the pulse and oxygen saturation (SPO2 ) levels of the worker. These sensors are incorporated using a microcontroller, and the data obtained is processed using an edge computing model at nodes placed across the mine. As the data is processed at the network’s periphery (at the edge) and not on the cloud, the data transmission happens quickly, and the delay is less than in other communication models.

2 Design and Components The main Rakshak module is an assemblage of sensors incorporated in a polycarbonate casing to provide the module with adequate protection from damage from physical factors. The module can sense and measure ambient temperature, humidity, the concentration of carbon monoxide, and flammable gases around the miner and the worker’s physical conditions. The microcontroller (ESP32) is responsible for storing and wirelessly sending the Rakshak module data to the nodes. Sensors used in this module are DHT22 (temperature and humidity), MQ-9 (Carbon monoxide and flammable gases), MAX30102 (SpO2 and heart rate), and ESP32 MCU. Here, the data is processed in the mine site itself before being sent to the mine manager above ground. Its advantages include lower latency, decentralized nature, and saved bandwidth compared to cloud computing [6]. In this system, the microcontrollers present in the miner suit will communicate with stationary nodes present inside the mine using many-to-one communication. The nodes will be responsible for receiving data from the different modules, processing it, and taking action against the anomalies before sending it forward to the mine manager.

Implementation of Edge Computing Model for the Processing of Data …

101

2.1 Components ESP32 is a low-cost, low-power microcontroller with integrated Wi-Fi and dualmode Bluetooth [7]. It can perform either as a standalone application or as a slave to a host MCU. ESP32 module is shown in Fig. 1. DHT22 Temperature and Humidity Sensor (DHT22) as shown in Fig. 2 have a basic, low-cost capacitive humidity sensor and a thermistor to measure the temperature [8]. MQ-9 Gas Sensor is, as shown in Fig. 3, used to detect carbon monoxide and flammable gases such as methane and has high sensitivity due to its sensing material SnO2 [9].

Fig. 1 ESP32 used in the Rakshak module Fig. 2 DHT22 temperature and humidity sensor

Fig. 3 MQ-9 gas sensor

102

K. Aneesha Acharya et al.

Fig. 4 MAX30102 body vitals sensor

MAX30102 Body Vitals Sensor, as shown in Fig. 4, is an infrared pulse oximeter and heart rate sensor integrated into one chip which will be placed on the finger [10].

2.2 Module Design The microcontroller, gas sensor, humidity sensor, and temperature sensor will be placed inside a casing made of polycarbonate with small holes to provide accurate measurements of the sensors and wiring. The module’s dimensions are 6.8 × 6.2 × 4 cm, and the main module will weigh around 100 g. Polycarbonate was chosen as the perfect material for the casing because of its following properties [11]. . . . .

Intrinsic insulator. Lightweight. Cheap and cost-effective. High impact resistance.

The Rakshak module, as shown in Fig. 5b, will be placed on the upper back below the collar. This placement is chosen because it is farther away from the face of the person, and thus the carbon dioxide produced while exhaling will not interfere with the readings of the gas sensor. The module will also be embedded in extra layers of fabric so that factors like sweat and heat radiated from the worker’s body do not interfere with readings. The MAX30102 sensor is placed inside the glove on the index finger. This includes an IR sensor facing the finger through a small cut-out in the padded glove.

3 Methodology One of the biggest issues mines face is real-time data processing. The setting up of large computing devices is a major problem due to factors such as the lack of energy and space and the surrounding dusty environment [12]. Besides, due to the complex environment, transmitting the collected data to the remote cloud

Implementation of Edge Computing Model for the Processing of Data …

103

(a)

(b)

(c)

Fig. 5 Description of the proposed suite consists of a Rakshak module b Module dimension c Internal structure of sensor module

can suffer from the inevitable high latency under the current communication technology. In order to tackle this problem, this paper proposes a design that uses the edge computing communication model. Edge computing is a promising distributed computation architecture to seize the opportunity of enabling low-cost, low-latency collection, and processing of data.

3.1 Working Principle For communication between the end-users (i.e., the modules) and the edge servers (i.e., the nodes), one can make use of the Wireless Body Area Network (WBAN), which is already installed in most of the existing mines. WBAN can also be installed in temporary mines as it is portable, unlike the wired networks, and cheaper than other alternatives, such as the setting up of a mobile 5G base station [13]. Figure 6

104

K. Aneesha Acharya et al.

Fig. 6 Hierarchical model of data transmission

provides a graphical representation of the hierarchical model. In this edge computing model, there are three main layers of data transmission and processing. . The End-Users (Modules)—These are the modules responsible for collecting all the data, such as environmental and miners’ physical conditions. . The Edge Server (Nodes)—These would be the multiple nodes placed across the mines. Each node is responsible for processing the data of a particular set of modules. In case of an anomaly in the data, these nodes will be responsible for sounding an alarm and forwarding the information to the control point. . The Access Point (Manager)—This is the main display point, which could be the manager/supervisor’s computer. The processed data transmitted from the nodes will be displayed here.

3.2 Simulation Edge computing can massively reduce the latency, which can help save miners’ lives in case of an emergency and can also help to reduce the network usage of the system. In order to test and simulate our computing model, this work uses iFogSim, a Java-based open-source simulation tool that can be used for simulating distributed architecture computing systems such as our edge computing model. iFogSim is used to evaluate our system’s latency and network usage and compare it with a cloud computing-based system [14].

3.3 Setup In order to set up the edge computing model, first created variables of the nodes and modules [15]. As shown in Fig. 7, 2 nodes and 6 modules were tested for the first test case and then gradually expanded up to 16 modules while keeping the number of nodes constant. This research work also set up a cloud-based computing model and compared it with the edge computing-based model. In order to do that, one can set up a scenario

Implementation of Edge Computing Model for the Processing of Data …

105

Fig. 7 Edge computing model on iFogSim

Fig. 8 Cloud computing model on iFogSim

in which a router was responsible for collecting the data from the various modules and then sending it onto the cloud server for processing. Figure 8 shows the topology created in iFogSim for the cloud-based model. Configuration parameters of the nodes, router, and cloud server that have been used in the above models are shown in Table 1.

4 Result The data obtained from the simulations have been tabulated in this section. This allows us to estimate how much faster the data computation and transmission will be if an edge computing model is used over a cloud computing model. The main parameters that have been measured are the latency and the network usage for differing numbers of modules. A graphical representation of the data has also been given in order to provide more perspective.

106 Table 1 Configuration parameters of the nodes, router, and cloud server

K. Aneesha Acharya et al.

Parameter

Cloud

Router

Nodes

CPU length (MIPS)

44,800

2800

120

RAM (MB)

40,000

4000

2

Uplink bandwidth (MB)

100

10,000

4

Downlink bandwidth (MB)

10,000

10,000

4

Level

0

1

1

Rate/MIPS

0.1

0.0

0.0

Busy power (Watt)

16*103

107.339

8.2

Idle power (Watt)

16*83.25

83.43

4

4.1 Latency Comparision Latency is a measure of the time taken from collecting the data to displaying the processed data. Figure 9 shows the latency comparison between edge and cloud computing. As one can see from Table 2, the latency in the edge computing model is considerably less than that of the cloud computing model [16]. This is because the edge computing model evades the requirement of frequent access to the cloud and instead just processes the data on edge. As one can see from the data, the rise in the number of modules significantly increases the latency in the cloud computing model, whereas only a gradual rise is observed in the edge computing model. Thus, the edge computing model provides a more cohesive and faster method of computing, despite less processing power compared to cloud computing due to the faster transmission [17].

Fig. 9 Latency comparison between edge and cloud computing

Implementation of Edge Computing Model for the Processing of Data …

107

Table 2 Comparison of latency between cloud and edge computing models Modules

Modules per node

Edge latency (ms)

6

2

7.2

Cloud latency (ms) 7.9

8

4

7.8

10.3

10

5

8.3

28.4

12

6

8.7

76.5

14

7

9.2

292.6

16

8

9.8

623.5

4.2 Network Usage Comparison Network usage is greatly reduced in the edge computing model because there is no need to transmit unprocessed data to the cloud. The network usage comparison between edge and cloud computing is shown in Fig. 10. As most of the data processing is done by the edge server, the data transmission bandwidth is much less (Table 3).

Fig. 10 Network usage comparison between edge and cloud computing

Table 3 Comparision of network usage between cloud and edge computing models Modules

Modules per node

Edge network usage (kb)

Cloud network usage (kb)

6

2

3203.2

27,654.9

8

4

3903.7

35,098.7

10

5

4709.1

46,808.4

12

6

5409.0

54,239.7

14

7

6345.8

59,087.8

16

8

7204.7

63,327.4

108

K. Aneesha Acharya et al.

5 Conclusion The increasing energy demand has led to the intensification of resource exploitation in underground coal mines [18]. The work on the deeper excavation has followed the increase in production. Greater depths worsen working conditions, which specifically refers to the effective ventilation of roadways and maintenance within tolerable limits of ventilation, gas, and fire parameters, which are the main indicators of the quality of the working environment in the pit, and alarm or normal state of working conditions. With the ever-increasing hazards and accidents in mining, it’s evident that the lives of the workers working in coal mines are in danger. Small steps to ensure their safety will go a long way. Technology in the form of IoT might be the most relevant way to monitor the state of the miners and mines. The proposed Rakshak module will sense abnormalities in the physical and environmental conditions of the mines and is an easy way to eliminate threats to the lives of miners that until recently have been overlooked by mine owners and managers.

References 1. Bhattacharya S, Singh AK, Choudhury A (2013) Coal resources. Production and use in India. https://doi.org/10.1533/9781782421177.2.169 2. Shanmugam K, Kulshreshtha P (2005) Efficiency analysis of coal-based thermal power generation in India during post-reform era. Int J Global Energy Issues 23:15–28. https://doi.org/10. 1504/IJGEI.2005.006408 3. Santra S, Bagaria N (2014) Labour productivity in coal mining sector in India: with special to major coal mining states. Researchjournali’s J Hum Resource 2:1–14 4. Özmen ˙I, Aksoy E (2015) Respiratory emergencies and management of mining accidents. Turk Thorac J 16(Suppl 1):S18-S20. https://doi.org/10.5152/ttd.2015.005. Epub 2015 Apr 1. PMID: 29404110; PMCID: PMC5783100 5. Stewart AG (2020) Mining is bad for health: a voyage of discovery. Environ Geochem Health 42:1153–1165. https://doi.org/10.1007/s10653-019-00367-7 6. Sulieman NA, Ricciardi Celsi L, Li W, Zomaya A, Villari M (2022) Edge-oriented computing: a survey on research and use cases. Energies 15:452. https://doi.org/10.3390/en15020452 7. S. V, S. R, A. B, V. S. V, Vigneswari P (2022) IoT based healthcare monitoring and tracking system for soldiers using ESP32. In: 2022 6th international conference on computing methodologies and communication (ICCMC). pp 377–381. https://doi.org/10.1109/ICCMC53470. 2022.9754076 8. Ahmad YA, Surya Gunawan T, Mansor H, Hamida BA., Fikri Hishamudin A, Arifin F (2021) On the evaluation of DHT22 temperature sensor for IoT application. In: 2021 8th international conference on computer and communication engineering (ICCCE). pp 131–134. https://doi. org/10.1109/ICCCE50029.2021.9467147 9. Tamizharasan V, Ravichandran T, Sowndariya M, Sandeep R, Saravanavel K (2019) Gas level detection and automatic booking using IoT. In: 2019 5th international conference on advanced computing & communication systems (ICACCS). pp 922–925. https://doi.org/10.1109/ICA CCS.2019.8728532 10. Jincheng Z, Yanfei L, Boyuan Z, Kai W (2021) Design and implementation of wearable oxygen saturation monitoring system. In: IEEE international conference on electrical engineering and mechatronics technology (ICEEMT). pp 71−74. https://doi.org/10.1109/ICEEMT52412.2021. 9601533

Implementation of Edge Computing Model for the Processing of Data …

109

11. Du BX, Xiao J (2004) Discharge characteristics on modified polycarbonate by adding flame retardant [electric insulation. In: The 17th annual meeting of the IEEE lasers and electro-optics society. LEOS 2004. pp 559–562. https://doi.org/10.1109/CEIDP.2004.1364311 12. Ranjan A, Sahu HB (2014) Communication challenges in underground mines. Search & Res 5(2):23–29 13. Zhang Y, Zhao L, Yang K, Xu L (2020) Mobile edge computing for intelligent mining safety: a case study of ventilator. In: 2020 IEEE intl conf on parallel & distributed processing with applications, big data & cloud computing, sustainable computing & communications, social computing & networking (ISPA/BDCloud/SocialCom/SustainCom). pp 1300–1305. https:// doi.org/10.1109/ISPA-BDCloud-SocialCom-SustainCom51426.2020.00192 14. Gupta H et al (2017) iFogSim: a toolkit for modeling and simulation of resource management techniques in the internet of things, edge and fog computing environments. Softw: Pract Exp 47:1275–1296 15. Buyya R, Srirama SN (2019) Modeling and simulation of fog and edge computing environments using iFogSim toolkit. In: Fog and edge computing: principles and paradigms. pp 433–465. https://doi.org/10.1002/9781119525080.ch17. Wiley 16. Elbamby MS et al (2019) Wireless edge computing with latency and reliability guarantees. Proc IEEE 107(8):1717–1737. https://doi.org/10.1109/JPROC.2019.2917084 17. Awaisi KS et al (2019) Towards a fog enabled efficient car parking architecture. IEEE Access 7:159100–159111. https://doi.org/10.1109/ACCESS.2019.2950950 18. Ma K, Zhang Y, Ruan M, Guo J, Chai T (2019) Land subsidence in a coal mining area reduced soil fertility and led to soil degradation in arid and semi-arid regions. Int J Environ Res Public Health 16(20):3929. https://doi.org/10.3390/ijerph16203929

Adaptive Protection of Solar PV Microgrid Without ESS L. Poireiton Meitei, K. P. Vittal, and James Antony Pinto

Abstract Conventional sources play a major role in today’s power scenario. However, the trend has now been shifted toward non-conventional sources, viz., Solar, wind, tidal, etc. considering the depletion rate of fossil fuels, transmission losses and environmental impacts. As a result, Microgrids are getting popular nowadays due to its high reliability, nature of operation, and low operational cost. In this paper, an adaptive protection scheme is developed in MATLAB for a solar PV Microgrid without Energy Storage Systems (ESS) which can operate in both Grid connected mode (GCM) and Islanded Mode (IM). Keywords Microgrid · Grid connected mode · Islanded mode · MATLAB · Adaptive protection scheme

1 Introduction A microgrid may consist of several Distributed Generations (DGs) located at various locations of the distribution network which offers on-site generation at load points. Due to the inclusion of DGs at various load sites, a part or whole of the microgrid network may be provided with uninterrupted power during grid outage depending on the priority of loads to be supplied. Despite the numerous advantages of a microgrid network, yet a microgrid faces an extreme challenge in providing reliable protection L. P. Meitei (B) · K. P. Vittal · J. A. Pinto Department of Electrical and Electronics Engineering, National Institute of Technology, Surathkal, Karnataka, India e-mail: [email protected] K. P. Vittal e-mail: [email protected] J. A. Pinto e-mail: [email protected] J. A. Pinto Manipal Academy of Higher Education, Manipal, India © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 S. Kajampady and S. T. Revankar (eds.), Advances in Renewable Energy & Electric Vehicles, Lecture Notes in Electrical Engineering 1083, https://doi.org/10.1007/978-981-99-6151-1_8

111

112

L. P. Meitei et al.

due to its nature of bidirectional power flow and difference in fault current magnitudes during GCM and IM. In addition to this, the nature of fault current from Inverter Based Distributed Generations (IBDGs) may differ from that contributed by Synchronous generators in a utility grid. In view of this, it is highly essential to develop an adaptive protection scheme which addresses the issues mentioned above as the conventional protection scheme will no longer be applicable in a microgrid since the traditional protection techniques in distribution system rely on the radial topology with the utility grid located at one end to supply power. In this traditional configuration, the fault current is contributed by utility grid alone with protective device (PD) settings adjusted accordingly to restrict the effect of faults whereas in a microgrid, Distributed Energy Resource (DER) units can increase fault current levels, result in bidirectional power flows, change fault current flow paths, and impact PD operations. Research has been going on to analyze the impacts of IBDGs on the overall protection of the distribution network during fault conditions for both modes of operation, i.e., GCM and IM. To achieve this feat, fault analysis has to be conducted analytically to estimate the fault current magnitude. Here, the main concern is to estimate the fault current supplied by the IBDG. IBDGs have lower fault current contribution compared to Non-IBDGs due to its overload handling characteristics [1]. Further, fault current contribution varies with control schemes employed for the IBDGs [2]. In [3], a protection strategy based on microprocessor-based relays for low-voltage microgrids is proposed which does not require any communication links or adaptive protective devices. The strategy proposed in [4] uses local information to overcome the challenges of overcurrent protection. A dynamic adaptive overcurrent relaying (AOCR) scheme has been presented for ADNs with high DER penetration to estimate relay pickup that ensures significantly less communication overhead [5]. In [6], an adaptive protection scheme is proposed where a Microgrid Central Protection Unit (CPU) monitors the microgrid continuously using a communication medium. The protection scheme proposed in [7] uses pre-calculated information to update relay settings. A hybrid protection system which combines traditional differential protection scheme with adaptive scheme is proposed in [8]. A rule-based adaptive protection scheme combined with machine learning methodology is proposed in [9]. In [10], a protection system that contains differential relays to monitor current parameters at specific points within the microgrid is presented. An adaptive overcurrent protection scheme which automatically amends the protection settings of all overcurrent relays in response to the impact of DG, Active Network Management (ANM) and islanding operation is proposed in [11]. The numerical relay design using the fast recursive discrete Fourier transform (FRDFT) algorithm embedded with a fuzzy-logic decision-making module can be used for obtaining optimal protection settings in case of changing system conditions [12]. The main objective of this paper is to model and simulate a microgrid with a Solar PV without ESS, analyze the magnitude of fault currents and develop a simple adaptive protection scheme which can effectively work under GCM and IM.

Adaptive Protection of Solar PV Microgrid Without ESS

113

2 Modeling of Microgrid The microgrid model being considered here in this paper is a typical type of microgrid where a solar PV is the sole supplier of power during IM. As a result, the PV system does not operate in Maximum Power Point Tracking (MPPT) during IM while it operates in MPPT during GCM operation. The DC output of the Solar PV is adjusted automatically based on the connected load during IM so as to provide constant output AC voltage and frequency. The Single line Diagram (SLD) of the microgrid model is shown in Fig. 1. The rating of the PV array considered in this paper is of 250 kW where it is connected to the utility grid through an Static transfer switch (STS) and a Distribution transformer of 250 kVA. A diode-clamped three-level inverter also known as Neutral Point clamped (NPC) inverter is considered in this paper to convert DC output into AC. A three-level inverter has reduced dv/dt and THD [13] in its ac output voltages in comparison to the two-level inverter. The inverter is controlled using Voltage Oriented Control (VOC) during GCM and V-f control during IM.

2.1 Inverter Control During Grid Connected Mode VOC uses dq frame of reference aligned with the grid voltage vector to make ease of the use of PI controllers in control system design. The grid current in abc frame measured at the secondary side of the distribution transformer is transformed into dq frame and it is decomposed into two parts, one along d-axis (id), which is oriented with the grid voltage vector and proportional to the active power and the other along q-axis (iq), which is orthogonal to the grid voltage vector and proportional to the reactive power. Hence, the VOC scheme is similar to P-Q control scheme except the fact that the Active power is controlled by the DC-link voltage. The block diagram for Inverter control during GCM is shown in Fig. 2.

Fig. 1 SLD of microgrid model

114

L. P. Meitei et al.

Fig. 2 Block diagram for inverter control in grid connected mode

2.1.1

MPPT Controller

The MPPT algorithm employed in this paper is the Perturb and Observe (P&O) Algorithm as it is easier to implement. In this algorithm, the reference voltage is perturbed and the direction of next perturbation is determined by observing the system response. The block diagram for P&O Algorithm is shown in the flowchart given in Fig. 3. From the algorithm, it is found out that the algorithm outputs a voltage reference (.Vdcref in Fig. 2) which generates maximum power at the particular value of Irradiation and Temperature.

2.1.2

Vdc Regulator

.

VOC is based on a cascaded voltage and current control loop [14]. The DC-link voltage .Vdc is controlled by the outer voltage loop also known as .Vdc regulator using a PI controller. The regulator output is proportional to the active power and hence

Fig. 3 Flowchart for P&O algorithm

Adaptive Protection of Solar PV Microgrid Without ESS

115

Fig. 4 .Vdc regulator block diagram

it provides the desired value of .id . Hence, the main role of this .Vdc regulator is to generate reference current component along d-axis, .Idref , which dictates the Active power to be generated by the PV source. The internal block diagram for the .Vdc regulator is shown in Fig. 4. where Vdc_meas .= Output DC voltage of the PV array Vdc_ ref .= Reference voltage corresponding to MPP (generated by MPPT Block) Vnom_ dc.= Nominal DC voltage.= 800 V (Voltage at MPP at 25 °C and 1000.W/m2 ) PI(z) .= PI Controller for .Vdc regulator, .kp = 2, .ki = 400 Id_ ref .= Reference current component along d-axis corresponding to maximum Active power generated by PV Array.

2.1.3

PLL and Measurements

In this block shown in Fig. 5, the three-phase voltage and current at the PCC, i.e., at the secondary side of the transformer (refer Fig. 1) are first converted into its corresponding pu values with V_base taken as the rated secondary voltage of the transformer. The PLL block in Fig. 5 tracks the phase of the reference voltage which is the grid voltage in the secondary side of the transformer. The PLL block generates the .ωt value for referencing. The abc to dq0 block performs Park transformation in a rotating reference frame by using Eq. 1. This block outputs the transformed voltages and currents and passed on to the next stage. ⎤⎡ ⎤ ⎡ ⎤ ⎡ ) cos(ωt + 2π ) cos(ωt) cos(ωt − 2π va vd 3 3 2 ⎣− sin(ωt) − sin(ωt − 2π ) − sin(ωt + 2π )⎦ ⎣vb ⎦ . ⎣ vq ⎦ = 3 3 3 1 1 1 vc v0 2 2 2

2.1.4

(1)

Current Regulator

The inputs to the current regulator block are the measured values .Id and .Iq , the reference values of .Id and .Iq and the measured values of .Vd and .Vq as shown in Fig. 6.

116

L. P. Meitei et al.

Fig. 5 PLL and measurements block diagram

Fig. 6 Current regulator block diagram

This block is the inner current control loop as described in VOC scheme. In GCM, the Iq_ref value is taken to be 0 (zero) as the reactive power is supposed to be injected by the utility grid. Note that, feedforward crossed compensation components are included in the dq current control loop for decoupling between the dq current axes as shown in Fig. 6.

Adaptive Protection of Solar PV Microgrid Without ESS

2.1.5

117

Reference Signal Generation

In this block,.Vdconv and.Vqconv generated by the Current regulator block is transformed into its corresponding three-phase sinusoidal signal each shifted by 120° which serves as the reference (modulating) signal for Sinusoidal PWM Generator.

2.1.6

PWM Generator

The PWM Generator (3-Level) block generates pulses for carrier-based pulse-width modulation (PWM) converters using three-level topology. The reference signal (Uref input), also called the modulating signal, is naturally sampled and compared with two symmetrical level-shifted triangle carriers as shown in Fig. 7. One reference signal is required to generate the four pulses of an arm. So, for a three-phase bridge, three reference signals are required to generate the 12 pulses as described above.

Fig. 7 Pulse generation for phase A arm (in phase disposition)

118

L. P. Meitei et al.

2.2 Inverter Control During Islanded Mode During GCM, the voltage and frequency references to the converter are provided by the utility grid. The converter is synchronized with the grid through PLL and a certain amount of active and reactive power is fed to or from the grid based on the load requirement. However, such support from the grid does not exist when the microgrid is operated in IM. Whenever the STS in Fig. 1 is opened either due to intentional islanding or due to faults, the control scheme for the microgrid switches to V-f Control mode. The primary objective of the microgrid during IM is to maintain nominal voltage and frequency by controlling the active and reactive power injected to the network [15]. Figure 8 shows the control structure of the three-phase inverter that provides voltage and frequency regulation in IM. Unlike GCM, the control structure consists of an outer voltage control loop in addition to the inner current control loop. The references for the voltage control loop are generated by a voltage formation loop which can be designed in different ways. The voltage formation loop provides the reference ∗∗ ∗∗ ∗ .v d and .vq for voltage control by providing the reference values of voltage (.V ) ∗ and frequency (.f ). Grid forming voltage formulation block utilized in this paper for providing and voltage and frequency references is shown in Fig. 9.

Fig. 8 Block diagram for inverter control in islanded mode

Adaptive Protection of Solar PV Microgrid Without ESS

119

Fig. 9 Grid forming voltage formulation block

3 Overcurrent Adaptive Protection Scheme During GCM, the fault current is contributed by both the utility grid and the DG whereas in IM of operation, the fault is fed by DG alone which is why the fault current level decreases in this mode. It is to be noted that IBDGs don’t have a rotating inertia to carry fault current based on an electro-magnetic (EM) characteristic. Also, the fault current of IBDGs have smaller decay time as compared to rotating machines. The block diagram for implementation of the proposed Adaptive protection scheme is shown in Fig. 10. As seen in the block diagram, the RMS current is sensed for the branch where the CB is installed which is then stepped down using a Current transformer (CT) for relaying purposes. The pickup current for the relay is set differently for GCM and IM which is lower in the latter case. Another input to the relaying scheme is the Mode of operation select signal. If the mode of operation signal is 1, the PCC breaker is ON and the network remains connected to the Utility grid. As a result, the fault behavior will be similar to the conventional radial power system networks and hence Normal Inverse Curve Time Overcurrent (TOC) Characteristics is chosen. The Operating time calculation for a Normal Inverse Curve is given below in Eq. 2. After calculation of the operating time, Trip command is issued to the corresponding breaker depending on the Trip time. t = TMS ×

. o

0.14 −1

( IsI )0.02

Fig. 10 Block diagram of adaptive overcurrent relaying scheme

(2)

120

L. P. Meitei et al.

where I is the current setting. .I is the actual current. .TMS is the IEC Time Multiplier setting. . s

If the mode of operation signal is 0, the PCC breaker is open or OFF and the network remains isolated from the Utility grid. In IM, the TOC characteristic curve of relays will be shifted to instantaneous OC settings to adapt to lower fault currents [16]. Moreover, adaptive protection schemes would automatically adjust the relay settings according to the network operating state. In this mode, Trip command is issued to the breaker in the next cycle of detection of fault.

4 Fault Analysis of Microgrid The network model shown in Fig. 1 is simulated for various types of faults at different locations indicated in Fig. 11 as denoted by locations 1, 2, and 3. The Circuit breaker (CB) must be able to trip whenever the OC relay issues a trip command during fault conditions. The CB should not undergo any false tripping when the operation changes from GCM to IM or vice versa. The fault analysis is conducted for both modes of operation, i.e., GCM and IM for fault resistances 0.001 .Ω and 1 .Ω at locations 1, 2, and 3 under the following operating conditions. Operating conditions: 1. Irradiation .= 1000 .W/m2 , 2. Temperature .= 25◦ , 3. Load .= 50 + 30 + 150 = 230 kW

Fig. 11 Fault locations indication for simulation

Adaptive Protection of Solar PV Microgrid Without ESS

121

Table 1 Fault current values for different fault types when fault resistance .= 0.001 .Ω Mode of Type of fault Minimum fault current (A) operation Near PV source At location (2) Near utility grid (1) (3) Grid connected mode

Islanded mode

3L

1239

1887

5655

SLG LL LLG 3L SLG LL LLG

800 1175 1281 542 380 464 494

820 1799 1910 530 349 450 485

894 4247 4323 514 323 464 480

Table 2 Fault current values for different fault types when fault resistance .= 1 .Ω Mode of Type of fault Minimum fault current (A) operation Near PV source At location (2) Near utility grid (1) (3) Grid connected mode

Islanded mode

3L

301

262

433

SLG LL LLG 3L SLG LL LLG

254 276 292 164 198 192 207

216 233 250 136 164 157 135

353 397 410 214 262 266 298

4. Impedance values, .R = 0.55/10 .Ω and .L = 0.239/10 mH 5. Fault initiation time .= 0.2 s The fault current magnitudes at all locations 1, 2, and 3 during both modes of operation for fault resistances values of 0.001 .Ω and 1 .Ω are shown in Tables 1 and 2 respectively.

5 Simulation Results for the Proposed Protection Scheme The proposed protection scheme is implemented using MATLAB Simulink. Simulation results for a symmetrical fault, i.e., 3L fault and an asymmetrical fault, i.e., SLG fault for fault resistance 0.001 .Ω at locations (1) and (3) during GCM and IM

122

L. P. Meitei et al.

Fig. 12 3L fault occurs at location (1) in GCM

Fig. 13 SLG fault occurs at location (1) in GCM

are shown in the following Figs. 12, 13, 14, 15, 16, 17, 18, and 19. When the breaker opens after CB operating time, current reduces to zero which can be observed from the simulation results.

Adaptive Protection of Solar PV Microgrid Without ESS

Fig. 14 3L fault occurs at location (3) in GCM

Fig. 15 SLG fault occurs at location (3) in GCM

123

124

Fig. 16 3L fault occurs at location (1) in IM

Fig. 17 SLG fault occurs at location (1) in IM

L. P. Meitei et al.

Adaptive Protection of Solar PV Microgrid Without ESS

Fig. 18 3L fault occurs at location (3) in IM

Fig. 19 SLG fault occurs at location (3) in IM

125

126

L. P. Meitei et al.

6 Inferences Fault analysis for the microgrid is conducted to determine the minimum fault current flowing into faulted section of the network when a fault occurs in GCM and IM. It is easily visible from Tables 1 and 2 that there is a significant difference in the fault current magnitude when the mode of operation and fault resistance changes. Hence, this fault analysis is helpful in deciding the pickup current for relay operation as higher values of pickup current may lead to mal-operation of Protective Devices leading to catastrophic damage to the network. The fault current magnitude decreases a. when mode of operation is shifted to Islanded mode b. when fault resistance increases. Also, it is found out that the fault current magnitude is higher when the fault occurs near the Utility grid source in GCM and the fault current magnitude does not vary much during Islanded mode of operation. From the simulation results, it is observed that whenever a fault occurs in GCM, the corresponding CB did not operate suddenly rather it waits for completion of Trip time calculated using Eq. 2. Further, it is also observed that higher the fault current, faster is the relay operation which agrees with the Inverse characteristics. Further, in IM, it is observed that the CB operates in the next cycle of fault detection irrespective of the magnitude of the fault current which agrees with its instantaneous characteristics. Whenever the Breaker operates, the faulted section is isolated and the remaining section of the microgrid resumes its operation in normal mode. As it can be seen from the simulation results, the current in the faulty section reduces to zero when the breaker operates and the bus voltage returns to its nominal value after isolation of faulty part. Hence, the protection scheme developed is verified.

7 Conclusion In this paper, a solar PV microgrid without ESS is modeled using MATLAB Simulink and an adaptive protective scheme is developed for the microgrid which adaptively changes its relay settings based on the mode of operation. The control schemes employed for Grid connected mode and Islanded mode are different as the primary aim of the Inverter is to supply maximum Active power in case of Grid connected mode and to maintain voltage and frequency at each bus in case of Islanded mode. The control scheme gets shifted from Voltage Oriented control (VOC) mode to V-f control mode when the mode of operation changes from GCM to IM and vice versa. The relay setting is set to a higher value for GCM based on the fault current analysis conducted in Sect. 4 while it is set to a lower value, i.e., .2Iload for IM. The relay operating time for the proposed protection scheme is being decided by Normal Inverse Curve for GCM and Instantaneous curve for IM. The protective action and the operating time of the CBs is verified from the simulation results obtained.

Adaptive Protection of Solar PV Microgrid Without ESS

127

References 1. Masaud TM, Mistry RD (2016) Fault current contribution of renewable distributed generation: an overview and key issues. In: 2016 IEEE conference on technologies for sustainability (SusTech). IEEE, pp 229–234 2. Shuai Z, Shen C, Yin X, Liu X, Shen ZJ (2017) Fault analysis of inverter interfaced distributed generators with different control schemes. IEEE Trans Power Deliv 33(3):1223–1235 3. Zamani MA, Sidhu TS, Yazdani A (2011) A protection strategy and microprocessor-based relay for low-voltage microgrids. IEEE Trans Power Deliv 26(3):1873–1883 4. Mahat P, Chen Z, Bak-Jensen B, Bak CL (2011) A simple adaptive overcurrent protection of distribution systems with distributed generation. IEEE Trans Smart Grid 2(3):428–437 5. Jain R, Lubkeman DL, Lukic SM (2018) Dynamic adaptive protection for distribution systems in grid-connected and islanded modes. IEEE Trans Power Deliv 34(1):281–289 6. Peiris W, Eranga W, Hemapala K, Prasad W (2018) An adaptive protection scheme for small scale microgrids based on fault current level. In: 2018 2nd international conference on electrical engineering (EECon). IEEE, pp 64–70 7. Oudalov A, Fidigatti A (2009) Adaptive network protection in microgrids. Int J Distrib Energy Resour 5(3):201–226 8. Ustun TS, Khan RH (2015) Multiterminal hybrid protection of microgrids over wireless communications network. IEEE Trans Smart Grid 6(5):2493–2500 9. Lin H, Sun K, Tan Z-H, Liu C, Guerrero JM, Vasquez JC (2019) Adaptive protection combined with machine learning for microgrids. IET Gener Transm Distrib 13(6):770–779 10. Louw C, Buque C, Chowdhury S (2014) Modelling and simulation of an adaptive differential current protection scheme for a solar PV microgrid 11. Coffele F, Booth C, Dy´sko A (2014) An adaptive overcurrent protection scheme for distribution networks. IEEE Trans Power Deliv 30(2):561–568 12. Kumar DS, Srinivasan D, Reindl T (2015) A fast and scalable protection scheme for distribution networks with distributed generation. IEEE Trans Power Deliv 31(1):67–75 13. Wu B, Narimani M (2017) High-power converters and AC drives. Wiley 14. Abu-Rub H, Malinowski M, Al-Haddad K (2014) Power electronics for renewable energy systems, transportation and industrial applications. Wiley 15. Fathima H, Prabaharan N, Palanisamy K, Kalam A, Mekhilef S, Justo J (2018) Hybridrenewable energy systems in microgrids: integration, developments and control. Woodhead Publishing Series in Energy. Elsevier Science 16. Che L, Khodayar ME, Shahidehpour M (2014) Adaptive protection system for microgrids: protection practices of a functional microgrid system. IEEE Electr Mag 2(1):66–80

Comparative Study of Sensor and Sensor Less Speed Control of Permanent Magnet Synchronous Machines P. Sandhya, B. Rajalakshmi Samaga, and Ramzeena

Abstract Permanent Magnet Synchronous Motors (PMSM) used in controllers demand precise and accurate estimation of shaft positions for high-level controlling actions. Shaft positions can be obtained by sensors or can be estimated without sensors. This study does a comparative study of obtaining the shaft position of PMSM with and without sensors. Sensor less control is based on closed-loop control system using Model Reference Adaptive System (MRAS). MATLAB/Simulink is used to implement the developed model and simulated to obtain the results. This paper highlights the significance of sensorless control of shaft position. Sensor less method eliminates the need of costly and clumsy sensors thereby making the system more compact and Economical. Keywords Adaptive model · Permanent magnet synchronous motor (PMSM) · Model reference adaptive system (MRAS) · Reference model · Sensor less control · Sensor control · Speed control

1 Introduction Recently, conventional induction motors are replaced by the permanent magnet synchronous motors in the wind energy conversion systems, industrial applications, electrical drive systems, and household appliances. This may be attributed to the fact that they have several advantages over the induction motors such as higher efficiency, P. Sandhya (B) Department of Electrical and Electronics Engineering, RIT, Hasana, Karnataka, India e-mail: [email protected] B. R. Samaga Department of Electrical and Electronics Engineering, NMAM Institute of Technology, Nitte, India Ramzeena KVGCE, Sullia, Karnataka, India © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 S. Kajampady and S. T. Revankar (eds.), Advances in Renewable Energy & Electric Vehicles, Lecture Notes in Electrical Engineering 1083, https://doi.org/10.1007/978-981-99-6151-1_9

129

130

P. Sandhya et al.

Fig. 1 Block diagram of MRAS

smaller size, ability to maintain high torque at low speed, higher power factor, and high power density [1]. Precise speed control of PMSM can be obtained only by accurate measurement of its shaft position. Sensors like shaft encoders and resolvers give the shaft position but make the drive costlier and less compact [2]. Sensor less control approach estimates the shaft position thereby eliminating the usage of sensors and making the system more economical and compact [2]. Shaft position can be estimated using different methods and algorithms and they are broadly classified as signal injection method, observer method, artificial intelligence-based non-adaptive, and adaptive methods [9]. In this paper, a comparative study on estimating shaft position of a permanent magnet synchronous machine with and without the speed sensors is presented. Model reference adaptive system (MRAS) is used to estimate the shaft position of the motor for sensorless control. In the proposed MRAS control scheme, actual motor itself is used as the reference model and current equations include the rotor speed ω; hence, the current model of PMSM is used as an adjustable model. Direct axis current, id , Quadrature axis current, iq are outputs from both models. The adaptive model will give the estimated output and it is compared with the output of the reference model, and error between these two outputs is fed back to the adaptive mechanism using which the predictable value of rotor speed is obtained. By integrating speed, rotor position will be obtained. The basic block diagram of MRAS is indicated in Fig. 1.

2 Mathematical Modeling of PMSM To obtain the satisfactory simulation results, there is a need to develop a precise and a robust mathematical model of PMSM. dq model of PMSM is developed using a rotor reference frame and implemented on MATLAB/Simulink. The rotor reference frame is chosen because the position of the rotor magnets determines independently

Comparative Study of Sensor and Sensor Less Speed Control …

131

of the stator voltages and currents and the instantaneous induced emfs. Angle θr is the angle between the fixed stator phase axis and the rotor d-axis while the angle α is between the rotating stator mmf and the rotor d-axis. Stator magneto motive force is rotating at the same velocity as the rotor. Following assumptions are made while developing the proposed model of PMSM in the rotor reference frame. • • • •

Negligible saturation. Sinusoidal induced EMF. Negligible hysteresis and Eddy current losses. Field current dynamics are ignored. The equations of voltages along dq axis are given by   Vq = Rq i q + p λq + ωr λd

(1)

Vd = Rd i d + p(λd ) − ωr λq

(2)

where Rq and Rd are the quadrature and direct axis winding resistances which are equal. Hence let us define Rs = Rd = Rq . id and iq are the direct axis and quadrature axis currents. Therefore, Eqs. (1) and (2) can be rewritten as   Vq = Rs i q + p λq + ωr λd

(3)

Vd = Rs i d + p(λd ) − ωr λq

(4)

It is assumed that there is no flux along the q-axis. Therefore, q-axis current in the rotor is zero. The q and d axes stator flux linkages λq and λd in the rotor reference frame are given by λq = L q i q

(5)

λd = L d i d + L m i f

(6)

where L m is the mutual inductance between the stator winding and rotor magnets. L q and L d are the quadrature and direct axis inductances, respectively. ωr is the speed of the motor in rad/sec. The permanent magnet excitation is modeled as constant current source if . The rotor flux is along the d-axis. Hence d-axis rotor current is if . Electromagnetic torque is given by Te =

 3 P λd i q − λq i d 22

(7)

132

P. Sandhya et al.

Fig. 2 Equivalent circuit of d and q axes

Mechanical torque is given by TM = TL + BM ωrm + J pωrm ωr = ωrm ∗

(8)

P 2

where P is the number of poles. The equivalent circuits based on Eqs. 3 and 4 are as shown in Fig. 2. The state space representation of the Eqs. (3) and (4) is as follows:        Rs ωr L q Vd d λf ωr L d Rs λaf Rs λf Vq id + id + iq = − − − iq + + dt Ld Ld Ld Lq Lq Ld Ld Ld Lq (9) and is in the form x˙ = Ax + Bu  where x = [x1 x2 ] = i d + λLafd i q  ωL A = − LRds Lr d q − ωLr Lq d − LRqs  V Bu = LVdd + RLs λd f Lqq where B = 1. The adaptive model is developed similarly  

d

Rs ωr L q ωr L d Rs

x1 x2 = − − x1 x2 + u 1 u 2 − dt Ld Ld Lq Lq 









where ωr is the evaluated speed in rad/Sec And is in the form x˙ = Ax + Bu









(10)

Comparative Study of Sensor and Sensor Less Speed Control …

133

Estimated speed is taken as the actual speed. State variable error is given by 

e=x−x

(11)

To nullify the error, an adaptive mechanism is developed which is explained in the next section.

3 Adaptive Mechanism The developed control block scheme of MRAS is as shown in Fig. 3. The estimated speed ωr is given by 

t 

ωr =

    λf λf (i q − i q ) dτ + k2 i d i q − i q i d (i q − i q ) + ω(0) k1 i d i q − i q i d − Ld Ld 













0

(12) where k 1 , k 2 ≥ 0 are constants.

Fig. 3 Block diagram of Adaptive Scheme

134

P. Sandhya et al. 



Using the adjustable model i d , i q are evaluated. Considering λf as a constant, input id and iq are derived by transformation of the stator current. The voltage and current of the actual motor are inputs to the speed and position estimation blocks and the expected speed and position are the outputs.

4 Development of SIMULINK Model PMSM model is developed in SIMULINK and is as shown in Fig. 4. Park transformation is used for transformation from abc to dqo variable. Inverse park transformation is used in rder to transform dqo to abc variable. Voltages are taken as the input and currents, speed, and torque are considered as outputs during the simulation of the model. The Simulink model of speed control of PMSM with sensor using PI controller and PWM inverter is as shown in Fig. 5. There are two control loops. External control loop is speed control loop and internal loop is current control loop. Error detector block compares the reference rotor speed and actual rotor speed and outputs the speed error signal. The error amplifier which is PI controller is the speed controller, which amplifies speed error. Output of speed controller is torque reference, which is reference for the quadrature axis component of stator current.

Fig. 4 PMSM model developed in SIMULINK/MATLAB

Comparative Study of Sensor and Sensor Less Speed Control …

135

Fig. 5 Speed control block of PMSM with Sensor

The closed-loop speed control without sensor is also developed on the platform of SIMULINK using MRAS estimator and is as shown in Fig. 6. The adaptive uses the current model of PMSM and the reference model of the motor. The adaptive model generates estimated output i d , i q using Eqs. (13) and (14) 



  1 −Rs i d + ωr L q i q + Vd dτ Ld

  1 iq = −Rs i q − ωr L d i d − ωr λf + Vq dτ Ld id =

(13) (14)

Rotor speed is estimated by calculating the difference between the outputs of the reference model and adaptive model. Integrating the speed gives the position.

5 Results and Discussions The proposed models are implemented in SIMULINK/MATLAB software. Simulations on the speed control model of PMSM with sensor and without sensor are performed.

136

P. Sandhya et al.

Fig. 6 Speed control block of PMSM without sensor

Figure 7 shows the variation of rotor speed with time in open loop control of PMSM. Figures 8 and 9 show the performance of the closed-loop speed control system with and without a sensor. The above graph shows the performance of controlling of speed of PMSM with the sensor. Speed reached the reference speed of 1000 rpm at 0.0115 s. The speed reached the reference speed of 1000 rpm at 0.04 s without the sensor. A comparison table of the performance indices of the control schemes developed is as shown in Table 1.

Fig. 7 Waveform of rotor speed versus time during open loop control of PMSM

Comparative Study of Sensor and Sensor Less Speed Control …

137

Fig. 8 Waveform of rotor speed versus time with closed-loop control with sensor

Fig. 9 Waveform of rotor speed versus time with closed-loop control without sensor Table 1 Comparison of various speed control schemes of PMSM Performance indices

Settling time (s)

Maximum overshoot

Open loop PMSM

0.1

0.5

PMSM with Sensor

0.0115

0

Sensorless control of PMSM

0.04

0.2

138

P. Sandhya et al.

6 Conclusion The following three dynamic models for controlling the speed of the PMSM drive system are developed on SIMULINK. (i) Open loop. (ii) Closed loop with sensor. (iii) Closed loop without sensor by using MRAS estimator technique. All three models are simulated and obtained; waveforms of rotor speed versus time are presented. Performance analysis of the three models in terms of settling time and overshoot is given in Table 1. It is observed that all the three models gave satisfactory results. This study can be extended to determine. (i) The model inaccuracy and measurement of noise by using an extended Kalman Filter. (ii) Variation of the motor parameters in sensorless motor control. (iii) Hardware implementation with DSP or FPGA can be carried out. (iv) Models can be used for condition motoring.

References 1. Krishnan R, Electric motor drives: modeling, analysis and control, P 1st edn. Pearson Education 2. Review of position estimation methods for IPMSM drives without position sensors Part II adaptive methods XIX international conference on electrical machines—ICEM 2010, Rome 3. Yongdong L, Hao Z (2008) Sensorless control of permanent magnet synchronous motor, a survey. In: IEEE vehicle power and propulsion conference, Sep 3–5, 2008 4. Jabbar MA, Haque MA (1997) Sensorless permanent magnet synchronous drives. IEEE 5. Yongdong L, Hao Z (2008) Sensorless control of permanent magnet synchronous motor, a survey. In: IEEE vehicle power and propulsion conference, Sep 3–5, 2008 6. Gopinath GR, Das S (2014) An extended Kalman filter based speed and position estimator for permanent magnet synchronous motor. In: 2014 IEEE conference on power electronics, drives and energy system 7. Liu Q, Hameyer K (2015) A fast online parameter estimation of PMSM with signal injection. 978-1-4673-7151-3/15/$31.00 © 2015 IEEE 8. Kundo M (2007) Parameter estimation of permanent magnet synchronous machines. Trans Electr Electron Engg. IEEJ Trans 9. Review and comparison of sensorless techniques to estimate the position and speed of PMSM

An Exposition of Digital Taylor-Fourier Transform Krishna Rao and K. N. Shubhanga

Abstract Digital Taylor-Fourier Transform (DTFT) is a Taylor series-based extension to Discrete Fourier Transform. It is becoming popular due to its ability to estimate off-nominal frequency phasors accurately, which is lacking in DFT, the conventional technique used in Phasor Measurement Units (PMUs). In light of this, a lucid presentation of the mathematical formulation of DTFT is considered instructive and hence the same is attempted here. Further, the ability of DTFT filter to accurately estimate off-nominal frequency phasors, off-nominal frequency harmonics and exponentially varying sinusoids is illustrated and reasoned out. Keywords Digital Taylor-Fourier transform (DTFT) · Phasor measurement unit (PMU) · Harmonic filter · Electromechanical mode identification

1 Introduction Discrete Fourier Transform (DFT) and its variants such as Fast Fourier Transform (FFT) are used extensively to assess frequency components in a given signal [1]. In power systems, this is the basic technique employed in Phasor Measurement Units (PMU) [2]. While DFT-based phasor measurement works very well with pure sinusoidal signals of nominal frequency, it turns out to be less accurate under off-nominal frequency conditions [3]. Therefore, in the literature, there has been a continued effort to develop techniques for better phasor estimation, particularly at off-nominal frequency conditions. Some of the post-DFT processing solutions proposed are [2]: (i) passage of phasor estimates through a moving-average filter (ii) reconstruction of the K. Rao (B) Department of Electrical and Electronics Engineering, NMAMIT, NITTE (Deemed to be University), Nitte 574110, India e-mail: [email protected] K. N. Shubhanga Department of Electrical Engineering, National Institute of Technology Karnataka (NITK), Surathkal 575025, India © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 S. Kajampady and S. T. Revankar (eds.), Advances in Renewable Energy & Electric Vehicles, Lecture Notes in Electrical Engineering 1083, https://doi.org/10.1007/978-981-99-6151-1_10

139

140

K. Rao and K. N. Shubhanga

signal and resampling at a rate proportional to actual frequency. However, the method gaining greater credence of late is Digital Taylor-Fourier Transform (DTFT), which has been shown to estimate off-nominal frequency phasors accurately [4–8], thus solving an important drawback of DFT. This ability of DTFT can also be harnessed for estimation of multiple harmonics of off-nominal frequency as demonstrated in [9, 10]. DTFT fits a Taylor series-based polynomial to Fourier coefficients. Thus it treats the Fourier coefficients as time-dependent variables unlike conventional DFT which views them as constants. A lucid presentation of the mathematical formulation of DTFT is generally not available in the literature and hence the same is attempted in Sect. 2. Further, the working methodology of DTFT filter, its major attributes and main applications are illustrated through a few case studies in Sect. 3. The contribution of the paper is largely tutorial.

2 Evolution of DTFT from DFT The DFT of a signal over one fundamental cycle is [1]: ⎡

⎤

⎡

1 1 1 ⎢ ⎢1 ω N ⎥ ω2N ⎢ ⎥ 1 ⎢ ⎢ ⎢1 ω2N ⎥ ω4N .⎢ ⎢ ⎥= ⎢ ⎥ .. N ⎢ .. .. ⎣ ⎣. . | . N −1 2(N −1) y(N − 1) 1 ωN ωN .. . . .. . y(0) y(1) y(2) .. . yN

⎤ ⎤⎡ ··· 1 c0 · · · ω NN −1 ⎥ ⎢ c1 ⎥ ⎢ ⎥ −1) ⎥ ⎥ ⎢ c2 ⎥ · · · ω2(N N ⎥ ⎥⎢ .. ⎥ ⎢ .. ⎥ .. ⎣ | . . . | (N −1)2 c N −1 · · · ωN . . .. .

WN

(1)

cN

where N : is the total number of samples per fundamental cycle of duration .Tb , . yN : is the vector of signal samples, . W N : is the Fourier matrix, and . cN : is the vector of estimates of the Fourier coefficients. j2π Further, .ω N = e j2π fb Ts = e N where base (or nominal) frequency . f b = and .Ts is the sampling period. .

1 Tb

=

1 N Ts

Note: In all the mathematical expressions used in this paper, boldface letters in lower case represent vectors and boldface letters in upper case represent square matrices. From (1), one can, in general, represent a sample as

.

y(x) =

N −1 1 ∑ cl ωlx N N l=0

(2)

An Exposition of Digital Taylor-Fourier Transform

141

Note: (1) The Fourier coefficient .c0 represents the dc offset of the signal whereas .c1 and .c N −1 together represent the fundamental component of the signal while .c2 and .c N −2 together represent the second harmonic and so on. (2) If the signal is an undamped sinusoid of nominal frequency . f b Hz, only the Fourier coefficients .c1 and .c N −1 would be non-zero. The phasor of such a signal can be obtained from .c1 . Now consider an undamped sinusoid of frequency . f , which is drifted from the nominal frequency . f b by .Δ f . y (t) = A cos(2π f t) = A cos[2π( f b + Δ f )t] ( ) ( ) j2π f b t A j2πΔ f t − j2π f b t A − j2πΔ f t +e =e e e 2 2

. 1

(3)

Under this condition, from (1), .c1 becomes c =

. 1

N A j2πΔ f t e 2

(4)

This time-varying phasor is called dynamic phasor [4]. Next consider an exponentially varying sinusoid of off-nominal frequency: y (t) = Aeσ t cos(2π f t)

. 2

= Aeσ t cos[2π( f b + Δ f )t] ) ( ) ( − j2π f b t A σ t − j2πΔ f t j2π f b t A σ t j2πΔ f t e e e e +e =e 2 2

(5)

Once again, from (1), if the Fourier coefficient .c1 is to represent the dynamic phasor of the given exponentially varying sinusoid, it would have to be time-varying as detailed below. N A (σ t+ j2πΔ f t) e (6) .c1 = 2 With the conventional DFT, the Fourier coefficients.cl are constants and hence cannot represent the dynamic phasor of the off-nominal frequency sinusoid or the exponentially varying sinusoid. In such cases, the Fourier coefficients.c0 , c2 , c3 , . . . , c N −2 turn out to be non-zero even when the signal has neither a dc offset nor harmonics. In this context, it is ideal to have the Fourier coefficients as time-varying exponents. Then, one would have to handle transcendental equations, which is tedious. To circumvent this problem, the time-varying exponent representing the Fourier coefficient can be approximated by a Taylor series of order . K as follows [4]:

142

K. Rao and K. N. Shubhanga

c (t) = cle f f (t) = cl + (t − t0 )c˙l +

. l

(t − t0 ) K K (t − t0 )2 c¨l + · · · + cl 2! K!

(7)

K

where .cl is the value of .cl (t) at the reference time .t0 and .c˙l , c¨l . . . cl are respectively first-, second- … and . K th-derivative of .cl (t) evaluated at .t0 . Substituting (7) in (2), y(x) = .

[ N −1 N −1 N −1 ∑ 1 ∑ (t − t0 )2 ∑ lx cl ωlx + (t − t ) c ˙ ω + c¨l ωlx 0 l N N N+ N l=0 2! l=0 l=0 ] N −1 (t − t0 ) K ∑ K lx ··· + cl ω N K! l=0

(8)

where .t = x Ts . The recommended practice is to take the window centre as time reference .t0 [11]. Hence, .t0 = Nh Ts where . Nh = (N 2−1) . Over a cycle, .t varies from .0 to .(N − 1)Ts . Therefore .(t − t0 ) varies from .−(Nh )Ts to .(Nh )Ts . That is, .−(Nh )Ts ≤ (t − t0 ) ≤ (Nh )Ts . To obtain the equation for the first sample, one has to substitute .x = 0, and hence .t = 0 and .(t − t0 ) = −(N h )Ts in (8). This yields y(0) = .

[ N −1 N −1 N −1 ∑ (−Nh Ts )2 ∑ 1 ∑ l l l cl (ω0N ) + (−Nh Ts ) c˙l (ω0N ) + c¨l (ω0N ) + N l=0 2! l=0 l=0 ] (9) N −1 (−Nh Ts ) K ∑ K 0 l ··· + cl (ω N ) K! l=0

Similarly, for the second sample: .x = 1, t = 1(Ts ), (t − t0 ) = (−Nh + 1)Ts and hence y(1) = .

1 N

[ N∑ −1 l=0

l

cl (ω1N ) + (−Nh + 1)Ts

N −1 ∑ l=0

l

c˙l (ω1N ) +

N −1 ((−Nh + 1)Ts )2 ∑ l c¨l (ω1N ) 2!

l=0 ] N −1 K ∑ ((−Nh + 1)Ts ) K l +... + cl (ω1N ) K! l=0

(10)

.

This can be extended up to the last sample for which: x = N − 1, t = (N − 1)Ts , (t − t0 ) = Nh Ts and hence

An Exposition of Digital Taylor-Fourier Transform

143

[ N −1 N −1 N −1 ∑ (Nh Ts )2 ∑ 1 ∑ l l l cl (ω NN −1 ) + Nh Ts c˙l (ω NN −1 ) + c¨l (ω NN −1 ) + N l=0 2! l=0 l=0 ] N −1 (Nh Ts ) K ∑ K N −1 l ... + cl (ω N ) K! l=0 (11) Equation (9) can be rearranged as

y(N − 1) = .

[ ] [ ] 1 [ 11 . . . 1 1×N (−Nh Ts ) 11 . . . 1 1×N y(0) = N ⎡ ⎤ cN ⎢ c ] ⎢ ˙N ⎥ . ⎥ ] (−Nh Ts ) K [ ⎢ c¨ ⎥ 11 . . . 1 1×N ⎢ N ⎥ ··· ⎢ .. ⎥ K! ⎣ . |

(12)

K

cN where c = [c0 c1 . . . c N −1 ]T

(13)

c˙ = [c˙0 c˙1 . . . c˙ N −1 ]T

(14)

c¨ = [c¨0 c¨1 . . . c¨ N −1 ]T

(15)

. N

. N

. N

Similarly, K

.

∗

K

K

K

cN = [c0 c1 . . . c N −1 ]T

(16)

∗

Further, .c1 = c N −1 , c2 = c N −2 and so on. ∗

∗

Similarly, .c˙1 = c˙ N −1 , c˙2 = c˙ N −2 and so on. ∗

K

K

∗

K

K

Finally, .c1 = c N −1 , c2 = c N −2 and so on. On the same lines as (12), (10) can be rearranged as y(1) =

.

[ ] [ ] 1 [ 1ω N . . . ω NN −1 (−Nh + 1)Ts 1ω N . . . ω NN −1 N ⎡ ⎤ cN ⎢ c ] ˙N ⎥ ⎥ ] ⎢ ((−Nh + 1)Ts ) K [ ⎢ c¨N ⎥ N −1 ··· 1ω N . . . ω N ⎢ . ⎥ ⎢ . ⎥ K! ⎣ . | K

cN

(17)

144

K. Rao and K. N. Shubhanga

Finally, (11) can be rearranged as [ ] [ ] 1 [ −1)2 (N −1)2 N −1 N y(N − 1) = T 1 ω N N −1 . . . ω(N 1 ω . . . ω h s N N N N ⎡ ⎤ cN ˙⎥ ⎢ c . ] ] ⎢ N⎥ (Nh Ts ) K [ 2 ¨⎥ ⎢ c (N −1) ··· ⎢ .N ⎥ 1 ω N N −1 . . . ω N ⎢ . ⎥ K! ⎣ . |

(18)

K

cN Equations (12), (17) .. . . (18) can be combined into a single matrix equation as follows [11]: ⎡ .

⎢ ⎢ ⎢ ⎣

y(0) y(1) .. .

⎡

⎤ ⎥ 1 [ ⎥ IN W N TN W N ⎥= | N

y(N − 1)

⎤ cN ⎢ c˙N ⎥ ⎢ ⎥ ] ⎢ c¨ ⎥ 1 1 T 2 W N . . . K ! TN K W N ⎢ N ⎥ 2! N ⎢ .. ⎥ ⎣ . |

(19)

K

cN

where TN = diag[−Nh Ts , (−Nh + 1)Ts , . . . Nh Ts ] and . W N : the Fourier matrix . I N : identity matrix of order . N . .

K

T

Since (19) contains .(K + 1)N unknowns in .[cN c˙N c¨N . . . c N ] , one would have to extend DTFT to .(K + 1) = D cycles to get a minimum of .(K + 1)N equations as follows [11] (This may be referred to as the Taylor expansion of DFT): ⎡

y(0) y(1) .. .

⎤

⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ 1 [ . ⎢ y(N − 1) ⎥ = I D N W D N TD N W D N . . . ⎥ ⎢ N ⎢ y(N ) ⎥ ⎥ ⎢ ⎥ ⎢ .. | ⎣ . y(D N − 1)

⎤ cN ⎢ c˙N ⎥ ⎥ ]⎢ ⎢ c¨ ⎥ 1 K T W DN ⎢ N ⎥ K! DN ⎢ .. ⎥ ⎣ . | ⎡

K

cN

[ ] ] [ ] [ In short, . y D N (D N ×1) = B D N (D N ×D N ) c D N (D N ×1) . In (20), . I D N is identity matrix of order . D N ; . W D N and . TD N are given by

(20)

An Exposition of Digital Taylor-Fourier Transform

145

⎡

[W D N ](D N ×N )

.

⎤ WN ⎢WN ⎥ ⎢ ⎥ =⎢ . ⎥ ⎣ .. |

(21)

WN TD N = [diag(−L h Ts , (−L h + 1)Ts , . . . , L h Ts )](D N ×D N )

.

where . L h =

D N −1 . 2

Note: The reference instant .t0 has now been shifted from the centre of the first cycle to the centre of first . D cycles. Now . TD N = diag[T1 T2 . . . TD ] where T = diag[−L h Ts , (−L h + 1)Ts , . . . , (−L h + N − 1)Ts ] which indicates the Taylor expansion time values for the samples of first cycle. Similarly, . T2 , . . . , TD are diagonal matrices indicating the time values of samples of respective cycles. Hence it can be shown that ⎡ ⎤ T1 W N ⎢ T2 W N ⎥ ⎢ ⎥ . TD N W D N = ⎢ (22) .. ⎥ ⎣ . | TD W N

. 1

Therefore matrix . B D N in (20) can be rewritten as [11]: ⎡

.

BD N

I 1 ⎢ ⎢I = ⎢. N ⎣ .. I

T1 . . . T2 . . . .. . . . . TD . . .

⎤

⎡

WN 0 . . . ⎥ ⎢ 0 WN . . . ⎥ ⎢ ⎢ .. .. . . .. ⎥ ⎣ . . . . | 1 K 0 0 . . . T K! D (D N ×D N ) 1 TK K! 1 1 TK K! 2

0 0 .. . WN

⎤ ⎥ ⎥ ⎥ |

(23) (D N ×D N )

where all the sub-matrices are square matrices of order . N . The solution of (20) yields c

. DN

= B D N −1 y D N

(24)

With . c D N , the signal can be reconstructed as . y D N = BD N cD N

.

(25)

Thus, (24) and (25) are known as ‘DTFT analysis equation’ and ‘DTFT synthesis equation’ respectively.

146

K. Rao and K. N. Shubhanga

3 Spectral Estimation Performance of DTFT A few case studies involving off-nominal frequency sinusoids and an exponentially varying sinusoid are considered to bring out the efficacy of DTFT as a spectral estimator. The nominal frequency is taken as .50 Hz in Sects. 3.1 and 3.2 and as .1 Hz in Sect. 3.3.

3.1 Performance of DTFT as a Phasor Measurer The following signal is taken up for DTFT analysis with the Taylor series order K = 3 (i.e. a third-order DTFT filter).

.

y (t) = 1.0 cos[2π(49)t]

. 3

(26)

The above signal can be regarded as a typical current or voltage signal of any power system. For the sake of simplicity, the peak value of the signal is taken as unity. In fact, this can be regarded as the per unit measure of the actual signal. The signal frequency of .49 Hz deviates from the nominal frequency of .50 Hz by .2% and hence is clearly off-nominal. Sampling frequency . f s is .750 samples/s and the base frequency . f b is .50 Hz, which is same as nominal frequency. Therefore, the number of samples per cycle of nominal frequency, . N = f s / f b = 15. The following inferences are drawn from the results: (1) The amplitude of the fundamental component of the signal at any instant is . N2 times the absolute value of .c1e f f at that instant [see (7)]. In case of pure sinusoids, the amplitude is constant. Hence it should be sufficient to estimate .c1e f f at any one instant within the window, the natural choice being the time reference. From (7), .c1e f f (t0 ) = c1 . Hence the fundamental amplitude is estimated conventionally as . N2 times .|c1 |. In general, the amplitude of .lth harmonic is taken as . N2 times (N −1) .|cl | for .l = 1, 2, . . . , [11]. 2 (2) Although the signal frequency is off-nominal, the DTFT filter is able to estimate its amplitude accurately as .1.0000. (3) On other hand, a DFT filter estimates the amplitude of the above signal as .0.9904, which implies an error of .0.96%. As already mentioned, DFT is the common technique used for phasor estimation in Phasor Measurement Units (PMUs). However, if the phasor amplitude of a line working at .400 kV having a frequency of .49 Hz instead of the nominal frequency of .50 Hz is estimated with a DFT filter, it would be estimated as .396.16 kV resulting in an error of .3.84 kV. (That is, .0.96% of .400 kV is .3.84 kV.) (4) Figure 1 compares the amplitude response of a DFT filter against a DTFT filter of third order. It can be clearly seen that the DFT filter has a very sharp response

An Exposition of Digital Taylor-Fourier Transform

147

Fig. 1 Comparison of fundamental amplitude response of DTFT and DFT filters

(5) (6)

(7)

(8)

whereas the DTFT filter has a wide maximally flat passband. Throughout this band, the amplitude estimation is accurate [5]. Simultaneously, the amplitudes of all other harmonics are estimated correctly as zero in case of the DTFT filter. On the other hand, a DFT filter estimates the amplitude of second harmonic as .0.0141 or .1.41% although there is no second harmonic. Figure 2 compares the second harmonic estimate of the third-order DTFT filter and DFT filter. This clearly shows the ability of a DTFT filter to offer a wider stopband with maximally zero amplitude response around the fundamental frequency of .50 Hz. [5]. In fact, the DFT filter is the DTFT filter of zeroth order (. K = 0). The length of the signal window required for a DTFT filter is.(K + 1) cycles of base frequency. Thus, the signal window length required by a DFT filter equals one cycle of base frequency whereas the same for a third-order DTFT filter spans four cycles of base frequency. For the DTFT filter of third order (i.e. . K = 3) with a base frequency of .50 Hz, the maximally flat passband is found to extend from .48.1 to .51.9 Hz. If . K is increased to .5, this flat passband is found to extend further from .46 to .54 Hz. Thus, by increasing the DTFT filter order from .3 to .5, the flat passband width can be enhanced by roughly .110% although this entails greater data requirement and longer computation time. This is because the approximation of .c1e f f gets better as . K is increased, as can be inferred from (7).

148

K. Rao and K. N. Shubhanga

Fig. 2 Comparison of second harmonic response of DTFT and DFT filters

3.2 Performance of DTFT as a Harmonic Filter Bank Consider the following signal made up of a fundamental component and a second harmonic, both having off-nominal frequencies: y (t) = 1.0 cos[2π(49)t] + 0.2 cos[2π(99)t]

. 4

(27)

As in the previous subsection, sampling frequency, . f s is .750 samples/s and the base frequency, . f b is .50 Hz, and hence . N = f s / f b = 15 samples/cycle. Therefore the nominal frequency of the fundamental filter is .50 Hz whereas the same for the second harmonic filter is .100 Hz. The following inferences emerge from the comparative study of the amplitude estimation carried out using third-order DTFT filter (. K = 3) and DFT filter: (1) The DFT filter estimates the fundamental amplitude as .0.9854 and the second harmonic amplitude as .0.2133 respectively, indicating an error of .1.46% and .6.65%. As already alluded to, this is due to the sharp amplitude response of the DFT filter bank depicted in Fig. 3. Note that even the zero response of the DFT filter is sharp. Thus, the output of the second harmonic DFT filter is zero only at .0 Hz, .50 Hz and .150 Hz, etc. (That is, at exact multiples of .50 Hz). Thus, if the frequency of the fundamental component deviates slightly from the nominal (.50 Hz in this case), there is leakage into the harmonic bins. Similarly, when the frequency of the second harmonic deviates from its nominal value (.100 Hz in this case), there would be leakage into fundamental and other harmonic bins. In fact, the output of the third harmonic filter is indicated as .0.0095 in this case although the third harmonic is totally absent. Even the DC component is wrongly estimated as .0.0433. (2) The third-order DTFT filter estimates the fundamental and the second harmonic amplitudes accurately as .1 and .0.2 respectively. This is due to the fact that the fundamental filter and all the harmonic filters in the DTFT filter bank have fairly

An Exposition of Digital Taylor-Fourier Transform

149

Fig. 3 Combined fundamental and second harmonic response of a DFT filter

wide maximally flat passband and maximally zero passband [9] as shown in Fig. 4. For example, the maximally flat passband of the second harmonic filter extends from .98.1 to .101.9 Hz whereas its maximally zero stopband extends from .47.4 to .52.4 Hz. Thus, if the second harmonic frequency lies anywhere between .98.1 to .101.9 Hz, its amplitude would be estimated accurately. On the other hand, if the frequency of the fundamental component is anywhere between .47.4 and .52.4 Hz, it will not spill into second harmonic bin, i.e. it will not be mistaken for a second harmonic. Further, in this case, all other harmonics and DC component are rightly estimated by the DTFT bank as zero. (3) Normally, when the fundamental signal frequency is off-nominal, the harmonic frequencies would also be off-nominal. DTFT filter bank performs excellently in estimating all these.

3.3 Performance of DTFT in Estimating Exponentially Varying Sinusoids As evident from (4) and (6), the structures of the dynamic phasor in case of pure off-nominal frequency sinusoid and exponentially varying sinusoid are similar. This indicates that a DTFT filter should be capable of estimating exponentially varying sinusoids as well. To examine the same, an exponentially varying sinusoidal signal of nominal frequency is considered here, which is represented by the following equation: (−0.05t) . y4 (t) = e cos[2π(1)t] (28)

150

K. Rao and K. N. Shubhanga

Fig. 4 Combined fundamental and second harmonic response of a DTFT filter

Base frequency of the filter, . f b = 1 Hz, signal sampling frequency, . f s = 15 samples/s, and hence . N = f s / f b = 15 samples/cycle. The spectral estimation results with third-order DTFT lead to the following observations: (1) Among the effective Taylor-Fourier coefficients, only .c1e f f and .c N −1e f f are nonzero. This implies that there is no leakage of the fundamental component into harmonic bins, which would otherwise happen with the DFT. (2) Further, .the estimated fundamental amplitude = N2 |c1 | = 0.9063 matches with the actual amplitude of the signal at the centre of the four-cycle window. Thus, the amplitude estimation of the fundamental is accurate. (3) The fundamental component of the signal is reconstructed using (25) and is compared with the original signal in Fig. 5. To measure the extent of matching between the reconstructed signal and the original signal, one requires a fitness metric [12]. The commonly used metric is ‘Signal to Estimation-error Ratio (SER)’, which is defined in (29). SER = 20 log10

.

rms(y) rms(y − . y)

(29)

where . y: original sampled value, .. y: estimated value. If the value of SER (which is conventionally called Signal-to-Noise Ratio, SNR) is above .30 dB, matching between the original- and reconstructed-signal is considered to be good [13]. In this case, the value of SER is .124.11 dB, which is far higher than the threshold of .30 dB, denoting an excellent matching between the compared signals. (4) Successful reconstruction of the signal implies that the exponentially varying amplitude of the signal can be estimated satisfactorily using DTFT. Thus, the damping ratio of the signal can also be calculated. This has applications in

An Exposition of Digital Taylor-Fourier Transform

151

Fig. 5 Comparison of poorly damped sinusoid and estimated fundamental

identification of low frequency oscillations of power systems and also in other areas of system identification. An algorithm for the same is presented in [11]. (A detailed discussion of such an algorithm would be long-winded and is difficult to accommodate in a conference paper like this.)

4 Conclusions In this paper, a lucid presentation of the mathematical formulation of DTFT is offered. Besides, the superior filtering performance of DTFT vis-a-vis conventional DFT is illustrated through a few case studies and a cogent explanation is offered for the same.

References 1. Proakis JG, Manolakis DG (2000) Digital signal processing. Pearson Prentice Hall, Upper saddle river, New Jersey, USA 2. Phadke AG, Thorp JS (2010) Synchronized phasor measurements and their applications. Springer, US 3. Barchi G, Macci D, Petri D (2013) Synchrophasor estimators accuracy: a comparative analysis. IEEE Trans Instrum Meas 62(5):963–973 4. de la O Serna JA (2007) Dynamic phasor estimates for power system oscillations. IEEE Trans Instrum Meas 56(5):1648–1659 5. de la O Serna JA (2010) Dynamic phasor and frequency estimates through maximally flat differentiators. IEEE Trans Instrum Meas 59(7):1803–1811

152

K. Rao and K. N. Shubhanga

6. de la O Serna JA, Platas-Garza MA (2011) Maximally flat differentiators through WLS Taylor decomposition. Digital Signal Process 21:183–194 7. de la O Serna JA (2013) Synchrophasor estimation using Prony’s method. IEEE Trans Instrum Meas 62(8):2119–2128 8. Castello P, Liu J, Muscas C, Pegoraro PA, Ponci F, Monti A (2014) A fast and accurate PMU algorithm for P+M class measurement of synchrophasor and frequency. IEEE Trans Instrum Meas 63(12):2837–2845 9. Platas-Garza MA, de la O Serna JA (2011) Dynamic harmonic analysis through Taylor-Fourier transform. IEEE Trans Instrum Meas 60(3):804–813 10. de la O Serna JA (2013) Taylor-Fourier analysis of blood pressure oscillometric waveforms. IEEE Trans Instrum Meas 62(9):2511–2518 11. de la O Serna JA, Ramirez JM, Mendez AZ, Paternina MRA (2016) Identification of electromechanical modes based on digital Taylor-Fourier transform. IEEE Trans Power Syst 31(1):206– 215 12. Rao K, Shubhanga KN (2018) MAPE-an alternative fitness metric for Prony analysis of power system signals. Int J Emerging Electr Power Syst 19(6). https://doi.org/10.1515/ijeeps-20180091 13. Hauer JF (1991) Application of Prony analysis to the determination of modal content and equivalent models for measured power system response. IEEE Trans Power Syst 6(3)

Autonomous Microgrid Using New Perspective on Droop Control in AC Microgrid Siddaraj, Udaykumar R. Yaragatti, and H. Nagendrappa

Abstract Providing higher quality power to consumers through the existent microgrid is now a problem for the renewable energy source. Designing a droop controller for the microgrid is a necessity to construct a dependable and effective microgrid. In this paper, a P–F/Q–V droop method is used to connect several VSIs in parallel. Their parallels and differences are amply discussed in this study. A frequency droop control method and a virtual impedance approach are combined in the suggested method, which is coupled to two distributed generation (DG) local controllers and has each unit having a droop control and a voltage-current controller. An islanded system’s load flexibility and microgrid reliability can both be enhanced by adding this controller. The microgrid was subjected to this idea without any real-time communication. Simulink/MATLAB was used to simulate this. The results obtained show that the microgrid (MG) model is effective in autonomous operation. Keywords Droop control · Inverter · Microgrid · Virtual impedance

1 Introduction As the environmental problems are increasing day by day and with the cost of electricity being increased, research in the field of renewable energy resources is being boosted. Microgrid comprises distributed generation sources (DGs) with different types of loads. Renewable energy may be integrated with a distributed network with the help of a key component known as distributed generation. The use of DGs is increasing which motivates the need for better strategies for interconnection. With Siddaraj (B) Department of Electrical and Electronics Engineering, Manipal Institute of Technology, Manipal Academy of Higher Education, Manipal 576104, India e-mail: [email protected] U. R. Yaragatti · H. Nagendrappa Department of Electrical and Electronics Engineering, National Institute of Technology Karnataka, Surathkal 575025, India © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 S. Kajampady and S. T. Revankar (eds.), Advances in Renewable Energy & Electric Vehicles, Lecture Notes in Electrical Engineering 1083, https://doi.org/10.1007/978-981-99-6151-1_11

153

154

Siddaraj et al.

the help of the microgrid concept, smaller networks of sources and loads can continue to operate dependably even when they are cut-off from the main grid. A variety of DGs such as fuel cells, and photovoltaic cells are interfaced with the distributed network through power electronic converters [1]. These sources provide output in the form of DC, but it can be used both for AC and DC loads using conversion. For interfacing, the AC loads voltage source inverters (VSIs) may be used in between the sources which form the basis of the microgrids [2]. The existence of highly inertial synchronous machines makes conventional systems more stable. On the other hand, inverter-based microgrids use power electronic components, which have a much lower moment of inertia and are therefore more prone to oscillations. When the microgrids are working under isolated mode as in the case of this project, they will be responsible for maintaining the system stability. For the autonomous operation of microgrids particular oscillatory modes and poor damping arise because of circuit and control features [3]. The electrical industry has seen a rise in the usage of distributed generation (DG) and storage to encourage the adoption of green technology. The development of power electronics has improved the dependability and controllability of these renewable energy sources. After conversion, the majority of the DC output produced by these sources can be used for both DC and AC loads [4]. Such sources require inverters in between to interface with the AC loads, which serves as the foundation for inverter-based microgrids. To share the fundamental real and reactive powers with the other DG inverter, each DG inverter has a power loop based on droop control, as explained in this study with the model with the two inverters considered [5]. Voltage and current controllers are included in the internal controls of inverters to control voltage and set current to a reference value. Microgrids can work in two modes which are the autonomous and the interconnected mode [6]. In the Islanding/Autonomous condition, the loads are supported by DGs and without connecting to the main upstream utility grids. The droop control strategy is one of the best strategies which has its own advantages and disadvantages. Droop control is the best-accepted strategy for controlling parallel multiple inverters working under the autonomous mode [7]. Droop-based control has many advantages such as great flexibility, high reliability, and no communication needed. The designed system succeeds to maintain the adequate frequency level, voltage level, and load sharing with a smooth transition. To meet the increasing electricity demand, coordination of different distributed generation (DG) units is possible with the help of a droop control strategy. The entire Microgrid system is modeled in Simulink and the value of parameters is provided via MATLAB environment. The well-known shortcomings of conventional droop control, particularly in distribution lines, include slow response and oscillatory transient behavior. Since real and reactive power in the distribution network is not completely disconnected due to the high R/X ratio, the power distribution between the inverters is inaccurate [8]. Numerous authors have proposed various controllers to address the issue at the contrary. For instance, in [9], by adding a power derivative-integral term to the traditional droop controller, authors increased the dynamic response of parallel inverters,

Autonomous Microgrid Using New Perspective on Droop Control in AC …

155

although proper power sharing is lagging. A decentralized droop controller is developed in [10] and [11] using the concept of resistive and virtual output impedance for better power sharing not stable in voltage and frequency. Furthermore, there is a trade-off between the improvement in power sharing accuracy and the rise in the deviation of frequency/voltage with a bigger value of droop gains [12]. The main issue identified in nearly all of the work in this field is the trade-off between enhanced power sharing and system stability, both of which depend on the droop coefficient. The virtual impedance and frequency droop controller technique for inverterbased microgrids is presented in this work. The recommended controller not only improves system stability but also the active power sharing between the inverters without changing the droop coefficient. The results of simulations are displayed for the controller’s validation. Section 2 shows the microgrid architecture with the control method commonly used for such systems. Section III proposed the virtual impedance and frequency droop control. Section IV results and discussions are presented. Section V concludes the work provided in this paper and is followed by the references.

2 Microgrid Architecture with Controller A three-phase parallel inverter-based AC microgrid system modeled in MATLAB/ Simulink is used to operate and control the autonomous inverter-based microgrid depicted in Fig. 1. To confirm its robustness, a simulation study was conducted. With the help of the improved droop control scheme, the load share between the two DG inverters is managed. The PWM switching signal required for the controlled operation of both inverters is provided by a voltage control and current control. A distribution line and a nonlinear load are included in the stability study simulation of a two-inverter-based 3-phase test microgrid. Since all of the inverters’ ratings are the same, the droop coefficient is assumed to be the same for all of them [13]. Inverter-connected DG sources with the system variables are given in Table 1. The DG system as. The controller of the DG VSI is split into four sections as shown in Fig. 2. The power controller that, in accordance with the droop attributes defined for the actual and reactive power, provides the fundamental component’s voltage and frequency values from the VSI output. Other control system parts include voltage and current controllers. They are meant to dampen the LC filter sufficiently and cancel out highfrequency noise [14]. A. Power Controller The droop controller’s job is to behave as the synchronous alternator’s governor. When the demand on the synchronous alternator increases, a power system’s frequency and voltage will normally drop to the governor’s droop characteristics, as shown in Fig. 3 [15]. By decreasing reference frequency and voltage as the load grows, this droop principle is applied in VSI. From the inverter output terminal,

156

Siddaraj et al.

Fig. 1 Test Microgrid

Table 1 Microgrid parameters Parameters

Value

Parameters

Value

Inverter 1

12KVA

Inverter 2

12KVA

Switching frequency ( f s )

8 kHz

Current controller proportional gain (K pc )

10.5

Filter inductance (L f )

1.35 mH

Current controller integral gain (K ic ) 16e3

Filter capacitance (C f )

50 µF

Filter cutoff frequency (ωc )

31.4 r/s

Coupling inductance (L c )

0.35 mH

Nominal voltage (V n )

380

Line reactance (X line )

0.10 Ω

Nominal frequency (ωn )

314 r/s

Line resistance (Rline )

0.45 Ω

Droop gain (m)

9.4e−5

Voltage controller proportional gain (K pv )

0.05

Droop gain (n)

1.3e−3

Voltage controller integral gain (K iv ) 390

Filter resistance (r LC )

0.03 Ω

Feedforward term (F)

Inverter input resistance (r f )

0.1 Ω

0.75

compute the power indicated by (1) and (2) as follows (Fig. 4): p = vod i od + voq i oq ; q = −vod i oq + voq i od

(1)

The quantity of complex power generated by the filter with a cutoff frequency ωc is given as

Autonomous Microgrid Using New Perspective on Droop Control in AC …

157

Fig. 2 DGs Inverter controller

Fig. 3 Conventional droop characteristics,

P=

ωc ωc p; Q = q s + ωc s + ωc

(2)

The droop controller is expressed in (3) and (4). ωr = ω∗ − m P

(3)

vr = v∗ − n Q

(4)

Droop values m & n are expressed as: m=

ωmax − ωmin vod max − vod min , n= Pmax Q max

(5)

158

Siddaraj et al.

Fig. 4 Power controller

B. Voltage Controller The inverter voltage controller as shown in Fig. 5, gives reference current to a current controller. The reference voltage is established in the power control loop to control the inverter output. The objective of a conventional PI controller is to accomplish steady-state error zero between the set and nominal voltage. A feed-forward gain was included in the controller to speed up controller action, stated equations are (6), (7), and (8). dϕq dϕd ∗ ∗ = vod = voq − vod ; − voq dt dt

(6)

∗ i ld∗ = Fi od − wn Cf voq + K pv (vod − vod ) + K iv ϕd

(7)

∗ i lq∗ = Fi oq − wn Cf vod + K pv (voq − voq ) + K iv ϕq

(8)

C. Current Controller The block diagram of the inner current controller is shown in Fig. 6 it is a quickresponse control loop. An integral and proportional controller was utilized to make sure the VSI current does not go beyond the established set value. The reference output for this set by voltage controller, also expressed the governing equations. by (10), (11), and (12) [16]. dγq dγd = i ld∗ − i ld ; = i lq∗ − i lq dt dt

(9)

∗ vld = −wn L f i lq + K pc (i ld∗ − i ld ) + K ic γd

(10)

∗ vlq = wn L f i ld + K pc (i lq∗ − i lq ) + K ic γq

(11)

Autonomous Microgrid Using New Perspective on Droop Control in AC …

159

Fig. 5 Voltage controller

D. Virtual-Impedance Control As demonstrated in Fig. 7 the output impedance of a VSI is shaped using the virtual impedance method [17]. According to the current output, the reference voltage output decreases given as. vo = vr − Z v · i o ;

(12)

where Z V = RV + jX V is the virtual impedance, vo = V o ∠δ o is the output voltage and io is the current, respectively. vr = V r ∠δ r is the reference voltage of the voltage-current loop. According to Fig. 7, we have ( Vo ∠δo

Fig. 6 Current controller

Vr ∠δr − Vo ∠δo Rv + j X v

)∗

= P + jQ

(13)

160

Siddaraj et al.

By changing the output current for output power in (12), The corresponding voltage drop ΔV results from power passing through virtual impedance and angle difference δ V . The following equations are obtained by simplifying (13) Rv P + X v Q ΔV = Vr − Vo ∼ = Vo

(14)

X v P − Rv Q δv = δr − δo ∼ = Vo Vr

(15)

where V r is the voltage reference and δ r is the angle reference respectively. V o is the voltage output and δ o is the angle output, respectively. For ease, V r and V o are substituted by V * because the magnitude of their voltage falls within the permitted range of the nominal voltage deviation. Furthermore, when the virtual impedance is pure inductance, (14) and (15) are given as δo = δr − m d P

(16)

Vo = Vr − n d Q

(17)

where md =

Xv Xv ; nd = ∗ V ∗2 V

(18)

From (16) to (17) is considered as P–δ and Q–V feedback controller. Specifically, the type of (16) is corresponding to the angle droop in [18], and the type of (17) is the Q–V droop controller. Reference [18] has proved that greater coefficients md and nd can significantly enhance power sharing. As a result, the similarity offers a physical-based perspective to fine-tune the angle droop control parameters. The derivative of (16), the corresponding virtual inductance is given by ω0 = ωr − m d

dP dt

(19)

where ωr is the voltage reference angular frequency of. The derivative term of active power is changed by a high-pass filter to curb interference. Thus, the transitory droop function (19) is ω0 = ωr −

mds P s + ωc

(20)

where ωc is the high-pass filter cutoff frequency. From (20), virtual inductance control can be considered as a unique P–ω frequency droop controller, whose droop gain is a washout high-pass filter [19]. The washout filter-based active power sharing doesn’t contribute to the frequency deviation, unlike

Autonomous Microgrid Using New Perspective on Droop Control in AC …

161

Fig. 7 Virtual impedance-equivalent output voltage source

the constant feedback of (3). It should also be highlighted that the reactive power sharing cannot be made better using the washout filter-based approach suggested in [19, 20].

3 Proposed Virtual Impedance Method and Frequency Droop Virtual impedance and frequency droop control are typically implemented together [21]. As a result, the following adjusted droop control is given by (3), (4) into (21) and (22) ωo = ω∗ − m P − m d

dP dt

Vo = V ∗ − (n + n d )Q

(21) (22)

The P–ω droop is converted to a PD-type frequency droop control in (12). According to (13), to enhance reactive power sharing, a virtual impedance-derived equivalent Q–V droop gain nd is introduced.

4 Results and Discussions This paper assesses the performance of the suggested droop controller for nonlinear equal loads. Tested on MATLAB/Simulink according to Fig. 1. The system that is used as a standalone DG possesses two parallel VSI and the controller along with network impedance. The VSIs and their parameters are considered as given in Table 1. The load at t = 0s total load PLoad = 16,000 W, QLoad = 8000 VAR, and each inverter can share half the load proportionally, so, the active power of P1 = 7570 W,

162

Siddaraj et al.

P2 = 7560 W, and reactive power of Q1 = 3823.5VAR, Q2 = 3823 VAR. However, the power sharing initially is not satisfactory transient progress with the conventional droop controller as shown in Figs. 8 and 10, whereas in a modified droop controller the response is of damped oscillatory. Hence the dynamic response is enhanced comparatively as shown in Figs. 9, 10 and 11 of DG1 and DG2, respectively. Inverters’ frequency variation falls within the range of 49.8869 Hz to 49.8868 Hz, when the load was 16 kW, as shown in Figs. 12 and 13. The DG1 and DG2 the point of common coupling voltages amplitude per phase (rms), as shown in Figs. 14. and Fig. 8 Active and reactive power of DG1

Fig. 9 Active and reactive power of DG1

Fig. 10 Active and reactive power of DG2

Autonomous Microgrid Using New Perspective on Droop Control in AC …

163

Fig. 11 Active and reactive power of DG2

Fig. 12 Frequency with conventional droop control

Fig. 13 Frequency with modified droop control

15, respectively. The droop controller is therefore making sure that the tolerances are not greater than 5% and 1% of voltage and frequency, respectively.

164

Siddaraj et al.

Fig. 14 RMS voltage of DGs with conventional droop control

Fig. 15 RMS voltage of DGs with modified droop control

5 Conclusion This research proposes an improved droop-based controller for an independent parallel VSI microgrid system. The proposed controller’s objectives include maintaining the VSIs’ steady frequency and voltage magnitude as well as achieving proportionate power sharing of active and reactive power, with damped oscillations with an improved transient response. The provided controller precisely distributes power by avoiding an island microgrid’s communication link. The proposed controller was tested on an isolated DG that included two 3-phase VSIs and a supply fed by an LC filter to the loads. The results demonstrated that the controller may effectively improve the dynamic stability of distributed generations.

References 1. Strasser T, Andrén F, Kathan J, Cecati C, Buccella C, Siano P, Maˇrík V (2014) A review of architectures and concepts for intelligence in future electric energy systems. IEEE Trans Industr Electron 62(4):2424–2438 2. Díaz NL, Vasquez JC, Guerrero JM (2017) A communication-less distributed control architecture for islanded microgrids with renewable generation and storage. IEEE Trans Power Electron 33(3):1922–1939

Autonomous Microgrid Using New Perspective on Droop Control in AC …

165

3. Yazdani A, Dash PP (2009) A control methodology and characterization of dynamics for a photovoltaic (PV) system interfaced with a distribution network. IEEE Trans Power Deliv 24(3):1538–1551 4. Teodorescu R, Blaabjerg F, Liserre M, Loh PC (2006) Proportional-resonant controllers and filters for grid-connected voltage-source converters. IEE Proc-Electr Power Appl 153(5):750– 762 5. Vasquez JC, Guerrero JM, Miret J, Castilla M, De Vicuna LG (2010) Hierarchical control of intelligent microgrids. IEEE Ind Electron Mag 4(4):23–29 6. Guerrero JM, Chandorkar M, Lee TL, Loh PC (2012) Advanced control architectures for intelligent microgrids—Part I: decentralized and hierarchical control. IEEE Trans Industr Electron 60(4):1254–1262 7. Zhao X, Guerrero JM, Savaghebi M, Vasquez JC, Wu X, Sun K (2016) Low-voltage ridethrough operation of power converters in grid-interactive microgrids by using negativesequence droop control. IEEE Trans Power Electron 32(4):3128–3142 8. De Brabandere K, Bolsens B, Van den Keybus J, Woyte A, Driesen J, Belmans R (2007) A voltage and frequency droop control method for parallel inverters. IEEE Trans Power Electron 22(4):1107–1115 9. Guerrero JM, De Vicuna LG, Matas J, Castilla M, Miret J (2004) A wireless controller to enhance dynamic performance of parallel inverters in distributed generation systems. IEEE Trans Power Electron 19(5):1205–1213 10. Guerrero JM, Matas J, de Vicuna LG, Castilla M, Miret J (2007) Decentralized control for parallel operation of distributed generation inverters using resistive output impedance. IEEE Trans Industr Electron 54(2):994–1004 11. Kim J, Guerrero JM, Rodriguez P, Teodorescu R, Nam K (2010) Mode adaptive droop control with virtual output impedances for an inverter-based flexible AC microgrid. IEEE Trans Power Electron 26(3):689–701 12. Li Y, Li YW (2011) Power management of inverter interfaced autonomous microgrid based on virtual frequency-voltage frame. IEEE Trans Smart Grid 2(1):30–40 13. Chandorkar MC, Divan DM, Adapa R (1993) Control of parallel connected inverters in standalone AC supply systems. IEEE Trans Ind Appl 29(1):136–143 14. Bahrani B, Karimi A, Rey B, Rufer A (2012) Decoupled dq-current control of grid-tied voltage source converters using nonparametric models. IEEE Trans Industr Electron 60(4):1356–1366 15. Pogaku N, Prodanovic M, Green TC (2007) Modeling, analysis and testing of autonomous operation of an inverter-based microgrid. IEEE Trans Power Electron 22(2):613–625 16. Marwali MN, Keyhani A (2004) Control of distributed generation systems-Part I: Voltages and currents control. IEEE Trans Power Electron 19(6):1541–1550 17. Guerrero JM, De Vicuna LG, Matas J, Castilla M, Miret J (2005) Output impedance design of parallel-connected UPS inverters with wireless load-sharing control. IEEE Trans Industr Electron 52(4):1126–1135 18. Majumder R, Ledwich G, Ghosh A, Chakrabarti S, Zare F (2010) Droop control of converter-interfaced microsources in rural distributed generation. IEEE Trans Power Delivery 25(4):2768–2778 19. Yazdanian M, Mehrizi-Sani A (2015) Washout filter-based power sharing. IEEE Trans Smart Grid 7(2):967–968 20. Firdaus A, Mishra S (2018, March) A double derivative based droop controller for improved power sharing in inverter based autonomous microgrid. In: 2018 IEEMA engineer infinite conference (eTechNxT). IEEE, pp 1–6 21. Sun Y, Hou X, Yang J, Han H, Su M, Guerrero JM (2017) New perspectives on droop control in AC microgrid. IEEE Trans Industr Electron 64(7):5741–5745

Solar Photovoltaic Charging of Electric Vehicle and V2G for Indian Electricity Demand Scenario S. Suraj, Narayan S. Manjarekar, and Soumyabrata Barik

Abstract The increase in the price of fossil fuels and environmental issues such as global warming due to combustion engines has led people to switch to Electric Vehicles (EVs). The rise in the electricity demand from the Indian grid due to the charging of the large number of EVs is a major concern. The rise in Solar Photovoltaic (SPV) installation and its peak power generation during sun hours should be effectively utilized as the electricity demand during the same period is not much. Thus comes the option of charging the EVs with SPV which will not cause a burden on the Indian grid. The EVs battery is of high-capacity kWh rating and remains idle with 80% of State of Charge (SoC) for more than 16 h in a day. This unused charge with the help of Vehicle to Grid (V2G) technology can be used to inject power into the grid during the peak demand time which will give economic returns to the EV owners. This paper proposes SPV charging of EV and V2G technology to meet the 24 h electricity demand of India in a day with the aid of Bidirectional DC-DC (BiDCDC) and Bidirectional AC-DC (BiAC-DC) power converters and their controlling strategies. This proposed work also supports the charging of EVs via the grid (G2V) in case of the unavailability of solar power. The simulation of the proposed work is done for a 2-kW rated single-stage SPV and single-phase grid-tied inverter. Keywords Electric Vehicle · Solar Photovoltaic (SPV) · Bidirectional AC-DC (BiAC-DC) Converter · Bidirectional AC-DC (BiDC-DC) Converter · Vehicle to Grid (V2G)

S. Suraj (B) · N. S. Manjarekar · S. Barik Department of Electrical and Electronics Engineering, BITS Pilani, K. K. Birla Goa Campus, Goa, India e-mail: [email protected] N. S. Manjarekar e-mail: [email protected] S. Barik e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 S. Kajampady and S. T. Revankar (eds.), Advances in Renewable Energy & Electric Vehicles, Lecture Notes in Electrical Engineering 1083, https://doi.org/10.1007/978-981-99-6151-1_12

167

168

S. Suraj et al.

1 Introduction India holds the fifth position in the global automobile industry and is expected to become third by 2030. The crisis of fossil fuels and their economic impacts on the prices of these fuels has made people think about alternatives. The pollution made by internal combustion engines, especially air pollution along with noise pollution has paved the way to bring back a decade-old EV technology. The number of EVs in India stands at 13, 92,265 as on August 2022 as per the data provided by the Ministry of Road Transport and Highways, India. By 2030, this will likely increase by 45–50 million EVs on the road. India being the third largest producer of electricity has an installed capacity of 404.13 GW as of 31 July 2022 as per the Central Electricity Authority (CEA) [1]. Solar energy’s contribution to this is estimated to be 57.97 GW. Currently in India, as of September 2022 fossil fuel-based power generation plants produce 57.9% of electricity. of the country is being powered by fossil fuel-based (coal, lignite, and gas diesel) plants. The term demand depicts the amount of consumed electricity by the load. The all-India electricity demand met for the day 7 September 2022 with a peak demand of about 200 GW is shown in Fig. 1. The plot of the electricity generated by various installed sources to meet the demand for the same day is shown in Fig. 2. The Indian Power System Operation Corporation (POSOCO) provides these plots of electricity demand and generation. The typical daily load curve for each month of every financial year is shared by POSOCO in the form of a report [2]. The average daily load curve for the entire financial year 2019–2020 shown in Fig. 3 is obtained by averaging the daily load

Fig. 1 Indian electricity demand met

Solar Photovoltaic Charging of Electric Vehicle and V2G for Indian …

169

Fig. 2 Indian electricity generation

curve of all months of the financial year using MATLAB. The financial year 2019– 2020 is considered for best results since the electricity demand of India is yet to reach back to normal after the pandemic due to Covid-19. The typical all-India daily load curve, shown in Fig. 3, has a pair of demand levels low and high, respectively, during daytime and night time, respectively. Morning Peak, Day Lean, and Evening Peak or Night Peak. India is aiming to install 500 GW of renewable capacity by 2030 of which solar generation will be

Fig. 3 Average daily Indian electricity demand met during FY 2019–2020

170

S. Suraj et al.

Fig. 4 California duck curve

having a share of 280 GW. The increased solar generation without considering the demand during peak sun hours may lead to Duck’s belly [3] which happened in California due to the overgeneration during the non-peak hours by renewable energy sources as shown in Fig. 4. The solution for dealing with the net load crisis with an increase in solar generation and the impact the penetration of EVs [4] will have on the demand for electricity as its charging is done by plugging the EVs into the grid by charging the EVs with the solar power generated [5]. The typical solar charging station [6] for EVs is shown in Fig. 5. The boost converter is used to extract maximum power from the SPV and provides DC bus link voltage. The DC bus link voltage will charge the EV battery with a bidirectional DC-DC converter [7] or can be converted to AC with a bidirectional AC-DC converter [8] for powering AC loads. The bidirectional DC-DC converter acting as a boost converter allows the EV battery to inject power into the grid with the bidirectional AC-DC converter connected to the DC link acting as inverter, i.e., Vehicle to Grid (V2G) technology [9–11]. These bidirectional converters also allow the charging of EVs from the grid, i.e., Grid to Vehicle (G2V) technology. In this paper, the related work is discussed next in Sect. 2, followed by the proposed work in Sect. 3 and the simulation results of the proposed work in Sect. 4. Finally, Sect. 5 has the conclusion and mentions the future scope which can be considered as an extension of the work.

Solar Photovoltaic Charging of Electric Vehicle and V2G for Indian …

171

Fig. 5 Typical solar charging of EV[6]

2 Related Works Preetham and Shireen in [12] proposed a unique control strategy based on DC bus voltage sensing for a charging station consisting of a photovoltaic system, DC/DC boost converter, DC/AC bidirectional converter, and DC/DC buck converter. The different modes of operation in the proposed charging station were grid-connected rectification, PV charging and grid-connected rectification, PV charging, and gridconnected inversion. Fathabadi in [13] framed a V2G supported grid-connected solarpowered electric vehicle (EV) charging station capable of charging 15 EVs at a time with dedicated DC-DC converters for MPPT of PV followed by 15 bidirectional DC/ DC converters and a single bidirectional DC/AC inverter for V2G application. Tran et al. in [14] presented an SPV system for homes that charges EV batteries by the SPV or by the grid and also allows improvement in the stability of the grid at its peak load by V2G implementation. Shahrukh Adnan Khan et al. in [15] investigated vehicle-togrid (V2G), sun-to-vehicle (S2V), and vehicle-to-infrastructure (V2I). Yesheswini

172

S. Suraj et al.

et al. in [16] developed an SPV-based multi-port charging station with constant current/constant voltage charging from a constant voltage DC bus. A boost converter was employed for MPPT implementation and multiple buck converters for multi-port charging from the common DC bus. He et al. in [17] presented a standalone solar energy-fed EV charging station with a fuel cell stack used for support instead of batteries. Singh et al. in [18] developed an EV charging station that is powered by SPV and battery with the option of charging the EV from the grid or diesel generator when SPV power is unavailable, as well as the battery, is exhausted. Prem et al. in [19] proposed a DC-DC converter and its control strategy for battery-backed SPV charging station and its performance were validated with Real-Time Digital Simulation (RTDS) in OPAL-M Premchand et al. in [20] presented the two-stage single-phase SPV-based charging of EV with incremental conductance algorithm for Maximum Power Point Tracking (MPPT) and can charge the EV via grid during the absence of power from SPV. Varun et al. in [21] proposed a two-stage three-phase system in which EVs can be charged with SPV and V2G can provide power back into the grid.

3 Proposed Work The block diagram of the proposed work is shown in Fig. 6. The single-stage single PV system is connected to the common DC Link via the capacitor Cpv . The DC Link is interfaced to a single-phase grid via a Bidirectional AC-DC converter followed by an LCL filter. The same DC-link is interfaced with the EV battery with a bidirectional DC-DC converter. The switch SW1 is used to control the EV plug and SW2 is to connect and disconnect the PV power from the DC link. The proposed work has three power sources PV, EV battery, and Grid. The PV and EV batteries are of DC type while the Grid is of single-phase AC power of 50 Hz, 230 V RMS. The PV specification is chosen in such a manner that PVs voltage at maximum power point at any irradiance is greater than the minimum DClink voltage required for the proper inverter operation. In the case of a single-phase inverter system tied to the grid of 230 V RMS, 50 Hz the minimum DC link voltage

Fig. 6 Block diagram of proposed work

Solar Photovoltaic Charging of Electric Vehicle and V2G for Indian …

173

Fig. 7 Bidirectional converter a DC-DC. b DC-AC

should be 390 V. The incremental conductance method is used in this proposed work to obtain maximum power point tracking. The circuit diagram of bidirectional converters is shown in Fig. 7. The bidirectional DC-DC converter used in this framework is a conventional buck-boost converter as shown in Fig. 7a which converts the high DC voltage VHigh to low DC voltage VLow and vice versa. The DC-DC converter acts as buck topology when the EV battery is getting charged from the DC-link voltage VHigh by turning on switch SBuck only. During the discharging of the battery by turning on the switch SBoost alone bidirectional DC-DC converter acts as boost topology to form DC link voltage VHigh from the battery voltage VLow and injects the power into grid. The bidirectional AC-DC converter used in this work is a single-phase full-bridge converter with four switches S1 , S2 , S3, and S4 act as an inverter or as a rectifier as shown in Fig. 7b. When power is injected into the grid from PV or EV battery this bidirectional converter acts as an inverter with DC-link voltage as input DC voltage Vdc and produces output AC voltage Vac . The bidirectional AC-DC converter works as a rectifier to provide the DC-link voltage Vdc as output during the charging of the EV from the AC voltage Vac . In order to meet the Indian electricity demand effectively the direction of power flow from and between these sources is decided as per the time of the day as the demand varies with respect to the time. The timeline of power flow between these sources in a day of 24 h is shown in Fig. 11. During the daytime, the Indian electricity peak demand is from 9 to 11 am, and with the V2G technology power stored in EV battery can be injected into the Grid during the same period as shown in Fig. 18. The PV power generated will be maximum during the time period of 11 a.m. to 4 p.m. of the day as the sun hours is achieved in that time. During this time PV generated power is supplied to the EV battery charging and is called PV2EV (PV to Electric Vehicle) and at the same time power is injected into the grid from PV called PV2G (PV to Grid) as shown in Fig. 19. In

174

S. Suraj et al.

Fig. 8 Control structure of bidirectional converter during a charging. b Discharging

case of cloudy days or rainy season the PV cannot generate enough power to charge the EV battery and during those days Grid can charge the EV battery during 11 p.m. to 5 a.m. and is called as G2V (Grid to Vehicle) as shown in Fig. 20.

3.1 Control Strategy of EV Battery Charging and Discharging The buck or boost topology of the BiDC-DC converter is decided by generating the appropriate PWM signal of the switch SBuck during the charging and discharging process. The switch SBoost signal is an inverted signal of SBuck . The control structure for charging the EV is shown in Fig. 8a. The term Kb indicates the amount of current with which the EV battery has to be charged. The battery current is negative during the charging process so it has to be negated before generating the error signal for the PI controller. The charging of the EV battery can be from PV or can be from the grid. The control structure for discharging the EV is shown in Fig. 8b. The term Vdcref is the DC link voltage (Vdc) to be maintained for proper inverter operation of the bidirectional AC-DC converter while discharging the EV battery during V2G operation. The discharging control has two loop operations. The outer loop is for voltage regulation of DC Link voltage and the inner loop is for current regulation of battery current.

3.2 Generation of Alpha, Beta, and Active Components The generation of alpha–beta components from the grid voltage is shown in Fig. 9. The alpha–beta components generated in Fig. 9a are given to the PLL block shown in Fig. 9b for the generation of active component for injecting active power into the grid.

Solar Photovoltaic Charging of Electric Vehicle and V2G for Indian …

175

Fig. 9 Generation of a alpha–beta components. b Active component

Fig. 10 Control structure of power extraction from PV

Fig. 11 Control structure of V2G

3.3 Control Strategy of Extracting Power from SPV The control structure for extracting power from SPV is shown in Fig. 10. This control structure generates the reference signal Refpv2G , i.e., modulating signal for SPWM Inverter during power extraction from PV as the bidirectional AC-DC converter operates in inverter mode. The MPPT makes sure that PV is operated at the maximum power point. This control structure along with the battery charging control structure makes sure that the PV power generated is extracted and injected power to the grid and charges the EV battery at the same time.

3.4 Control Strategy of V2G Technology The control structure for V2G technology is shown in Fig. 11. The modulating signal for SPWM Inverter RefV2G during V2G operation is generated by this control structure where Amp is the amount of current to be injected into the grid. The bidirectional AC-DC converter operates in inverter mode in V2G operation. This control structure

176

S. Suraj et al.

Fig. 12 Control structure of G2V

along with the battery discharging control structure makes sure that the power from the battery is injected into the grid.

3.5 Control Strategy of G2V Technology The G2V technology means conventional charging of EVs from the grid. The control structure for G2V technology is shown in Fig. 12. This control structure generates the required RefG2V during G2V operation. The bidirectional AC-DC converter works in rectifier mode during this G2V mode. This control structure and battery charging control ensure that the EV battery is getting charged from the Grid.

3.6 Generation of PWM Signals for the Bidirectional AC-DC Converter The PWM signals generation method for the operation of a bidirectional AC-DC converter using the Unipolar Sinusoidal Pulse Width Modulation (SPWM) technique is shown in Fig. 13. The sinusoidal modulating wave can be any of Refpv2G or RefV2G or RefG2V generated by the control structure depending on the direction of power flow, i.e., SPV to Grid, Battery to Grid, and Grid to Vehicle, respectively.

Fig. 13 Generation of PWM signals for bidirectional AC-DC converter

Solar Photovoltaic Charging of Electric Vehicle and V2G for Indian …

177

4 Simulation and Results The MATLAB Simulink model of the proposed work is shown in Fig. 14.The solar panel used in implementing the proposed work has a maximum power of 2.1 kW. The SPV panel has one parallel module containing 11 strings in series with current– voltage and power-voltage characteristics for different levels of irradiations at a fixed temperature of 25 °C are shown in Fig. 15. The specifications of the components used for the simulation of the proposed work in MATLAB Simulink is shown in Table 1. The simulation is carried out for 10 s. The simulation can be divided into two cases: the first is charging of EV battery of 24 V,110 Ah with SPV, and the second is charging of EV from the grid when enough SPV power is unavailable due to cloudy or rainy days. In both cases, a V2G is implemented before the charging process starts. The simulation is done for a duration of 10 s for both cases, with the first 5 s is dedicated for V2G operation. The first 5 s is equivalent to 120 min, i.e., 9 a.m. to 11 a.m. sampled into 5 with each sample equivalent to 24 min. The last 5 s of the simulation is equivalent to 5 h, i.e., 11 a.m. to 4 p.m. which shows the charging from SPV or charging from the grid. The switching pulses of switches SW1 and SW2 that connect EV battery and SPV are shown in Fig. 16. The SW1 is always ON for the entire duration as the EV battery is always connected to the system. The SW2 is ON only when SPV power is available in the last 5 s of the simulation. The proportional (K p ) and integral constants (K i ) of the PI controller in charging control of Fig. 8a are found to be 0.2 and 110, respectively. In discharging control Fig. 8b, for the inner current loop PI controller the K p and K i are 0.05 and 10, respectively, while for outer voltage loop control the K p and K i are 0.5 and 50, respectively. For the PI controller in PLL of Fig. 9b the K p and K i are 10 and 500, respectively. In Figs. 10 and 11 for PI voltage controller K p and Ki constants are 1 and 10 respectively. The K p and resonant constant K r are 10 and 400 respectively for the PR controller of Figs. 10, 11, 12. The value of these controller constants of this proposed work is found out by trial-and-error method. The simulation starts with battery having 80% of SoC. The two parts of simulation are explained below. The SPV charges the battery in the last 5 s with a current amplitude of 1A, i.e., K b = 1A of Fig. 8a. Figure 18 shows the SPV voltage, current, and power output for the duration of 10 s as per the irradiance profile shown in Fig. 17. It can be seen that SPV operates at MPP for the given specific irradiance. The battery voltage, current, and SoC are shown in Fig. 19. The SoC and voltage drops during V2G operation while the battery current is positive implying that the battery discharges at 60A to inject 10A into the grid. The increase in voltage and SoC as the battery is charging from the SPV. The battery charging current is negative and of amplitude 1A. Figure 20 shows the inverter voltage, grid current, DC Link voltage, and irradiance. The inverter voltage is having amplitude of 325 V. The DC link voltage is maintained at 400 V throughout the V2G operation and during SPV operation the SPV voltage itself is the DC link voltage as MPPT is implemented.

178

Fig. 14 Simulink model of proposed work

Fig. 15 Photovoltaic I–V and P–V characteristics

S. Suraj et al.

Solar Photovoltaic Charging of Electric Vehicle and V2G for Indian … Table 1 Specification of simulation parameters

Parameter

Value

Capacitor of SPV

Cpv = 500 µF

Li-ion battery

24 V, 110 Ah, Initial SoC = 80%

Inductor for BiDC-DC

L = 6 mH

Capacitor of BiDC-DC

C = 5000 µF

Frequency of BiDC-DC

20 kHz

Frequency of BiAC-DC

10 kHz

Grid side filter, LCL

Lac = 4.06 mH, Cac = 6.01 µF, Lg = 4.35 mH

179

Fig. 16 Switching pulse of SW1, SW2

Fig. 17 Solar irradiance profile

4.1 V2G and Charging from SPV The SPV charges the battery in the last 5 s with a current amplitude of 1A, i.e., K b = 1A of Fig. 8a. The Fig. 17 shows the irradiance profile for the duration of 10 s and the Fig. 18 shows the corresponding SPV voltage, current, and power output. It can be seen that SPV operates at MPP for the given specific irradiance. The battery voltage, current, and SoC are shown in Fig. 19. The SoC and voltage drops during V2G operation while the battery current is positive implying that the battery discharges

180

S. Suraj et al.

Fig. 18 SPV output during V2G and SPV2G

Fig. 19 Battery output during V2G and SPV2G

Fig. 20 Inverter voltage, grid current, and DC link voltage during V2G and SPV2G

at 60A to inject 10A into the grid. The increase in voltage and SoC as the battery is charging from the SPV. The battery charging current is negative and of amplitude 1A. Figure 20 shows the inverter voltage, grid current, DC Link voltage, and irradiance. The inverter voltage is having amplitude of 325 V. The DC link voltage is maintained

Solar Photovoltaic Charging of Electric Vehicle and V2G for Indian …

181

at 400 V throughout the V2G operation and during SPV operation the SPV voltage itself is the DC link voltage as MPPT is implemented. The grid current is maintained at 10A during V2G and varies with irradiance during SPV operation as SPV gives preference to battery charging of 1A. The grid current increases and decreases with the respective increase and decrease of the irradiance. The zoomed in time version of Fig. 20 is shown in Figs. 21, 22. Figure 23 shows the inverter voltage, grid voltage, grid current, and DC link voltage. The grid voltage and current are in phase thus providing a unity power factor.

Fig. 21 Zoomed version of inverter voltage, grid current, and DC link voltage during V2G

Fig. 22 Zoomed version of inverter voltage, grid current, and DC link voltage during SPV2G

182

S. Suraj et al.

Fig. 23 Inverter voltage, grid voltage, grid current, and DC link voltage during V2G and SPV2G

4.2 V2G and Charging from Grid (G2V) The V2G operates for the first 5 s with 10A of current injecting to grid. The unavailability of SPV and the necessity of EV charging are solved by charging the EV via grid with a charging current of 5A, i.e., K b = 5A in G2V mode. The irradiance is maintained at 200 w/m2 throughout the simulation of 10 s to depict the unavailability of SPV. Figure 24 shows the SPV voltage, current, and power output for the same duration It can be seen that the SPV is disconnected and the voltage of the SPV is around 450 V, i.e., the open circuit voltage at the irradiance of 200 w/m2 . The current and power output of SPV remains at zero as it is open-circuited. The SoC and voltage drops during V2G operation to inject 10A into the grid by discharging at 60A as shown in Fig. 25. During the EV charging from the grid, the increase in voltage and SoC is steeper compared to charging from the SPV as the charging current is of 5A. Figure 26 shows the inverter voltage of 325 V, grid current, constant DC Link voltage of 400 V, and irradiance of 200 w/m2 . The inverter voltage is having amplitude of 325 V. The grid current is about 5A while charging the EV battery with 5A from the grid and during V2G operation the grid current is 10A as shown in Fig. 27. The grid as well as the inverter voltage and grid current are in phase for V2G and G2V operation as shown in Fig. 28 and its zoomed version in Fig. 29. The power factor becomes unity as the grid current and grid voltage are in phase as shown in Fig. 30.

Solar Photovoltaic Charging of Electric Vehicle and V2G for Indian …

183

Fig. 24 Zoomed inverter voltage, grid voltage grid current, and DC link voltage during V2G and SPV2G

Fig. 25 SPV output during V2G and G2V

Fig. 26 Battery output during V2G and G2V

184

S. Suraj et al.

Fig. 27 Inverter voltage, grid current, and DC link voltage during V2G and G2V

Fig. 28 Zoomed inverter voltage, grid current, DC link voltage and irradiance level during V2G and G2V

Fig. 29 Inverter voltage, grid voltage, grid current, and DC link voltage during V2G and G2V

Solar Photovoltaic Charging of Electric Vehicle and V2G for Indian …

185

Fig. 30 Zoomed Inverter Voltage, Grid Voltage, Grid current, and DC link voltage during V2G & G2V

5 Conclusion and Future Scope In this paper, to meet the Indian Electricity demand effectively the charging of EV battery with SPV power generation along with V2G operation is proposed. The control strategy for SPV system, V2G, and G2V is validated with simulation results. The grid voltage and current are maintained at unity power factor while charging the EV battery as well as during the current injection into the grid. The charging of EV battery during the unavailability of SPV is also addressed in this proposed work. The addition of a stationary energy storage system for SPV with the Second Life of EV battery can be utilized for further peak shaving the demand can be taken up as the future scope of the proposed work. Acknowledgements The authors would like to express their appreciation to the Power System Operation Corporation (POSOCO) of India for providing the monthly generation and demand plot of Indian Electricity demand which helped in the formulation, analysis, and to come up with a solution for this proposed work.

References 1. Central Electricity Authority of India (2022) Installed capacity. [Online]. Available: https:// www.cea.nic.in/installed-capacity-report/ 2. Power System Operation Corporation Limited (2022) Operational performance report for the month of September-2022 Electricity. [Online]. Available: https://posoco.in/download/mon thly_report_sep_2022/?wpdmdl=48348 3. Hafiz F, de Queiroz AR, Husain I (2018) Solar generation, storage, and electric vehicles in power grids: challenges and solutions with coordinated control at the residential level. IEEE Electrification Mag 6(4):83–90 4. IEA (2021) India energy outlook 2021. IEA, Paris

186

S. Suraj et al.

5. Akbari H, Browne MC, Ortega A, Huang MJ, Hewitt NJ, Norton B, McCormack SJ (2019) Efficient energy storage technologies for photovoltaic systems. Solar Energy 192:144–168 6. Premchand M, Gudey SK (2020) Solar based electric vehicle charging circuit in G2V and V2G modes of operation. In: 2020 IEEE students conference on engineering & systems (SCES), pp 1–6 7. Liu S, Xie X, Yang L (2020) Analysis, modeling and implementation of a switching bidirectional buck-boost converter based on electric vehicle hybrid energy storage for V2G system. IEEE Access 8:65868–65879 8. Yuan J, Dorn-Gomba L, Callegaro AD, Reimers J, Emadi A (2021) A review of bidirectional on-board chargers for electric vehicles. IEEE Access 9:51501–51518 9. de Melo HN, Trovão JPF, Pereirinha PG, Jorge HM, Antunes CH (2018) A controllable bidirectional battery charger for electric vehicles with vehicle-to-grid capability. IEEE Trans Veh Technol 67(1):114–123 10. Lee HS, Yun JJ (2019) High efficiency bidirectional buck boost converter for photovoltaic and energy storage systems in smart grid. IEEE Trans Power Electron 34(5):4316–4328 11. Wang Y-F, Xue L-K, Wang C-S, Wang P, Li W (2016) Interleaved high-conversion-ratio bidirectional DC–DC converter for distributed energy-storage systems—circuit generation, analysis, and design. IEEE Trans Power Electron 31(8):5547–5561 12. Preetham G, Shireen W (2012) Photovoltaic charging station for Plug-In Hybrid Electric Vehicles in a smart grid environment. In: 2012 IEEE PES innovative smart grid technologies (ISGT), pp 1–8 13. Fathabadi H (2017) Novel solar powered electric vehicle charging station with the capability of vehicle-to-grid. Sol Energy 142:136–143 14. Tran VT, Islam MR, Muttaqi KM, Sutanto D (2019) An efficient energy management approach for a solar-powered EV battery charging facility to support distribution grids. IEEE Trans Ind Appl 55(6):6517–6526 15. Khan MDSA, Kadir KM, Mahmood KS, Alam MII, Kamal A, Al Bashir MM (2019) Technical investigation on V2G, S2V, and V2I for next generation smart city planning. J Electron Sci Technol 17(4):1–11 16. Preetha Yesheswini B, Jai Iswarya S, Amani B, Prakash P, Sindhu MR (2020) Solar PV charging station for electric vehicles. In: 2020 international conference for emerging technology (INCET), pp 1–7 17. He F, Fathabadi H (2020) Novel standalone plug-in hybrid electric vehicle charging station fed by solar energy in presence of a fuel cell system used as supporting power source. Renew Energy 156:964–974 18. Singh B, Verma A, Chandra A, Al-Haddad K (2020) Implementation of solar PV-battery and diesel generator based electric vehicle charging station. IEEE Trans Ind Appl 56(4):4007–4016 19. Prem P, Sivaraman P, Sakthi Suriya Raj JS, Jagabar Sathik M, Almakhles D (2020) Fast charging converter and control algorithm for solar PV battery and electrical grid integrated electric vehicle charging station. Automatika 61(4):614–625 20. Premchand M, Gudey SK (2020) Solar based electric vehicle charging circuit in G2V and V2G modes of operation. In: 2020 IEEE students conference on engineering & systems (SCES), pp 1–6 21. Varun S, Bhongade S (2022) Solar power charging of electric vehicle along with the implementation of vehicle-to-grid (V2G) technology. In: Bohre AK, Chaturvedi P, Kolhe ML, Singh SN (eds) Planning of hybrid renewable energy systems electric vehicles and microgrid. Energy systems in electrical engineering. Springer, Singapore

Review of Voltage Sag\Swell Mitigation Control Techniques with Dynamic Voltage Restorer in a Grid Integrated Distribution System N. Sowmyashree, Hrushikesh Kulkarni, M.S. Shashikala, and K. T. Veeramanju

Abstract Voltage disturbances in power distribution systems caused by the integration of renewable energy sources can be efficiently minimized with the use of the Dynamic voltage restorer (DVR). DVR is an efficient custom power device that is used in power systems to mitigate voltage disturbances. The DVR’s performance is determined by its maximum voltage injection capability and the amount of energy stored in the restorer. Enhancement in the control technique of this non-linear device (DVR) is required to improve its efficiency. To efficiently regulate DVR, Synchronous reference frame Theory (SRF), Artificial Neural Networks (ANN), Integrated Sliding Mode Controller (ISMC), Fuzzy logic, and a variety of other control approaches have been developed. This paper gives a comprehensive review of voltage quality challenges, as well as the different DVR-based control strategies used in reducing voltage sag and swell in power networks along with their advantages and disadvantages. Keywords DVR · Solar PV · SRF · ANN

N. Sowmyashree (B) · H. Kulkarni · M.S. Shashikala · K. T. Veeramanju Department of Electrical and Electronics Engineering, SJCE (JSS Science and Technology University), Mysuru, India e-mail: [email protected] H. Kulkarni e-mail: [email protected] M.S. Shashikala e-mail: [email protected] K. T. Veeramanju e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 S. Kajampady and S. T. Revankar (eds.), Advances in Renewable Energy & Electric Vehicles, Lecture Notes in Electrical Engineering 1083, https://doi.org/10.1007/978-981-99-6151-1_13

187

188

N. Sowmyashree et al.

1 Introduction The dependence of the world on non-renewable resources for the production of energy has already started to prove itself as in-economical (costly), non-environmental friendly, and above all its likelihood of sustainability for a long time is unachievable. Thus, the world has changed its view on the production of energy toward renewable resources. Conventional types of energy production like nuclear power plants, hydroelectric power plants, and wind farms are affected by a shortage in fossil fuels but are still not completely extinct. Thus, the life span of these production units can be stretched and taken to high efficiency in usage with the help of technique of distributed generation. Major production of these clean energies in DG comes through PV and Windmills. In Fig. 1 (intro reference), it can be seen that 52.4% of its total energy production in 2015 came from renewable resources, so in Europe and North America with 34.2% and 27.7% respectively. Aspects such as clean energy and availability in different sizes and power have led to large-scale production and R&D in the field of production of renewable resources, especially the PV grid integrated systems [1]. Despite all these advantages, there are major drawbacks to PV systems. PV systems majorly depend on irradiance and ambient temperature, these give the power production in PV systems a lot of undesirable power quality issues and make its integration with the grid less efficient. Thus, the study of these power quality issues responsible for the serious damage by PV systems is essential in order to enhance the power quality and its reliability to sustain. Several techniques to generate power from renewable resources have been discovered and several power quality problems arise with them. Conventional devices for solving voltage sag/swell problems including DVR and UPQC are used. Rising technology mainly introduces modern methods of artificial intelligence techniques like Artificial Neural Networks (ANN) into conventional power quality mitigation devices to improve their efficiency. In [2], a DVR structure is proposed where it is used to mitigate voltage sag at the load terminals due to fault occurrence at the bus as well as along the line of a distribution system. The ability of the DVR through steadystate analysis is demonstrated in this paper. Different methodologies are discussed to mitigate deep and long-duration voltage sag. The harmonic activity was described and the mathematical model was tested and made to best suit for the Indian power quality issues scenarios. Many topologies and control methods have been presented

Fig. 1 Block diagram of dynamic voltage restorer (DVR)

Review of Voltage Sag\Swell Mitigation Control Techniques …

189

for DVRs in the literature. The presented topologies are categorized into two main groups. The first group of the presented topologies uses AC/DC/AC conversion, i.e., first rectifier, then capacitor and inverter. The second group of presented topology is DC/AC conversion, i.e., inverter operation. The second group of topologies is rarely used, because for large voltage sag, only capacitor bank is not capable of supplying the required reactive power at the instant of very deep voltage sag, i.e., < 0.5 pu. To mitigate the problems of power quality, distributed generation can also be a solution. In [3], the discussion on distributed generation as a solution to power quality issues concludes with a model of DG behavior for networks with the synchronous and asynchronous generators. The behavior of converter-based DG, synchronous generators, and asynchronous generators during voltage dips was verified. In opposition to the reported effects of synchronous and asynchronous generators on voltage dips in high to medium voltage networks, their influence on voltage dips in low-voltage networks was rather minimal. Converter-based DG was found to have a similar effect on voltage dips in low-voltage networks, in opposition to high-voltage networks. Different topologies with arrangements of DVR and ANN-based DVRs for power quality improvements are discussed in [2]. A DVR with an ANN controller and with the help of a PWM controller are made in series with the voltage bus and faults are avoided from damaging the other parts of the network. Voltage sag/swell is rectified with Fourier mathematical techniques used in building the ANN model. This paper concludes with the fact that an excellent total harmonic distortion (THD) mitigation is achieved with 13.5% when compared with other algorithms like Fuzzy logic and PID with 24.4% and 19.7%, respectively. Thus, the performance of the ANN-based DVR proves to be excellent. In [4], two theories, Synchronous Reference Frame (SRF) theory controlled DVR and Novel control theory, which is proposed are being compared. Novel theory proposes a mathematical model wherein an error signal from each phase is taken, fed to a common bus and PWM gives out appropriate Vsc. The Total Harmonic Distortion (THD) at the supply voltage and the load voltage are measured by the Fast Fourier Transform (FFT) analysis tool of the Simulink model at different conditions of single phase, two phases, three-phase sag and swell conditions. Each condition’s THD results are shown and examined under each case. Applications of DVR with various novel strategies in different scenarios can often be seen. In [5], voltage sag effects on motors running on ASD are observed and solution to Type A, Type B, and Type F voltage sag mitigation are provided using the ESRF method. Symmetrical, Unsymmetrical, and Phase angle jump voltage sag are very common, and Symmetrical voltage sag can prove vital to the functioning of motors. This affects the power quality and damage to motors. In order to solve this, ERF proposes to insert DVRs in each phase sequence of the power supply, aiding in the mitigation of symmetrical and unsymmetrical voltage sag. This has been proven effective to an extent of bringing down the sag due to 0.5 p.u. In accordance with [5, 6] discusses An Optimal Zero-Sequence Voltage Injection Strategy for DVR under Asymmetric Sag. This has a unique solution in compensating for the voltage injection strategy whenever the sag is detected by DVR on the supply line.

190

N. Sowmyashree et al.

In [7], discussion on a neural network-based control strategy of three phases four legs dynamic voltage restorer for voltage sag/swell and harmonic compensation. A designed algorithm and neural network learn the power quality features with backpropagation technique to resolve voltage sag/swell issues associated with the supply. The combination of several the traditional and novel strategies brings significant changes in the mitigation of voltage sag and swell. In [8], an adaptive neuro-fuzzy inference system-based control strategy to monitor DVR and mitigate voltage sag/ swell. Thus, mitigation of power quality issues like sag/swell using conventional custom power devices is improvised with recent technologies like Artificial Intelligence and with advanced mathematical models using fast Fourier transforms and other algorithms. Discussion and comparison of various methods to adopt a particular method is essential and with that the prevention of the power quality issues that arise with them need to be mitigated with such adaptive solutions.

2 Power Quality Good power quality concurs with several parameters which bunch up to good results. Consumers receiving a steady supply voltage, frequency that stays within the prescribed range, and smooth voltage curve waveform resembling a sine wave makes it up to a good power quality supply (G). Thus, power quality can be defined as a measure of the standard of power delivered. When power with low quality is delivered, it affects the utilities at consumer side, malfunctioning in the operation of protective systems at home, stations, and others. PV systems as DG when supplied to grid make a significant contribution to the grid and any disturbance with PV systems would adversely affect the power quality. Major issues to be addressed based on their severity are voltage fluctuation, Flickering, Harmonics, Voltage Sag, Swell, etc. Fluctuation in voltage occurs when the output power is unstable. Fluctuations in power being considered as a potential threat to PV systems is mainly due to its dependence on the surrounding environmental conditions, i.e., temperature and irradiance. Irradiance is mainly dependent on the moving clouds which are often at high speeds during noon time which are experimentally proven in [9–11] and block the solar radiation onto the PV systems. From this, it is clear that the size and topology of PV system arrangement also have significance in power fluctuation (concentrated collectors). Since a PV system with low capacity is relatively small and hence shaded area is vast (PREPA standards mentionable). Thus, in PV systems, significant drops in power happen during power fluctuation phenomena. Voltage flicker is also one of the noticeable problems at the PV grid—Tied system. Voltage fluctuations are a major cause of voltage flickers. Since irradiation in PV systems is highly dependent on cloud transient, it varies the voltage level, and as a result output power fluctuates [12, 13]. Thus, variations in voltage levels cause voltage flickers which can cause damage to equipment at the load side. The other fluctuation issue is frequency fluctuation which occurs due to an imbalance in the

Review of Voltage Sag\Swell Mitigation Control Techniques …

191

supply and demand of the power system. Irradiance has a significant role here too. Fluctuations in irradiance lead to frequency fluctuation often and worsen when high penetration of PV systems is in use, since they create a significant amount of drop in power. Another power quality issue is to deal with is harmonics. Harmonics are usually introduced by the power electronic devices used in renewable energy generation (xiaondong) or converters, inverters for integration of PV systems to grid. Since output of a PV system is DC, an inverter at high switching frequency helps in converting a DC Voltage and thus injects more harmonics [10, 14–16]. Harmonics are also caused by a disturbance in irradiation and also due to non-linear loads on the consumer side. Some research shows that irradiation is inversely proportional to Total Harmonic Distortion (TDHI) [10, 16, 17] and thus irradiation magnifies the TDHI. This is because, during fluctuations in irradiation, voltage levels drop down and its major effect is seen significantly on the fundamental component of the harmonic rather than its multiples. Maintaining a good power factor needs loads that are linear and efficient. If customer loads are non-linear and consume non-sinusoidal voltage from the sinusoidal voltage source, harmonics are caused. Harmonic currents, generated by non-linear loads interact with the power system impedance to give rise to harmonic voltage distortion. If this distortion exceeds the recommended limit it can cause severe damage to equipments, computers may exhibit data error or loss of data and electronic process control may zoperate out of sequence on the load side. In [18, 19], it clearly shows that 80% of the power quality complaints reported are of Voltage Sag. Voltage Sag or Voltage Dip (IEC term) is defined by the IEEE 1159 as the decrease in the RMS voltage level to 10%–90% (1%–90% for EN 50,160) of nominal, at the power frequency for durations of ½ cycle to one (1) minute. Also, voltage sag is classified as a short-duration voltage variation phenomenon, which is one of the general categories of power quality problems. This common power quality problem is usually caused by weather and utility equipment problems, which normally lead to system faults on the transmission or distribution system. For example, a fault on a parallel feeder circuit will result in a voltage drop at the substation bus that affects all the other feeders until the fault is cleared. The same concept would apply for a fault somewhere on the transmission system. Most of the faults on the utility transmission and distribution system are single-line-to-ground (SLG) faults [20]. An improvement in the voltage profile of distributed systems can improved with PV grid-tied systems even with low penetration systems as observed in [21]. Along with voltage sag, voltage swell occurs in PV grid-tied systems. As per IEEE standards voltage swell is an increase to between 1.1 pu and 1.8 pu in rms voltage or current at the power frequency durations from 0.5 to 1 min [22]. Voltage swell and over-voltages are the phenomenon that occurs when large load switching or power line switching occurs. If these swells have high peaks then the insulation (protection mechanism) would not be quick enough to withstand voltage swells or sags.

192

N. Sowmyashree et al.

3 Dynamic Voltage Restorer Dynamic voltage regulator (DVR) is static devices used to provide series compensation to mitigate voltage dip and restore the appropriate amplitude of voltage in the stable system by injecting or removing required power. Symmetrical component estimation (calculation) at the point of common coupling is done and voltage sag is mitigated by DVR. Power quality issues and hence voltage dips can be calculated as shown in the Eq. (1) Voltagesag/swell = Vs(Zf /Zf + ZS )

(1)

where Zs = The source impedance including the transformer impedance. Zf = The impedance between the PCC and the fault including fault and line impedances. This method of estimation ensures correction in dip and protects the sensitive loads during sag/swell. Figure 1 shows the block diagram of the DVR with injection transformer which is coupled with the system to inject power. There are mainly 2 types of DVR, one with storage unit and without storage unit. The DVRs without storage unit compensate by drawing power from voltage supply and the other uses the storage to compensate for the dip. Thus, DVRs only supply the part of waveform that is faulty (Sag or Swell) and not the entire waveform. This is a major difference between an UPS and DVR. As Fig. 1 shows, DVR consists of an injection transformer coupled with the transmission system, a filter circuit, a voltage source converter, and a control system for fast response and efficiency. DVR integrated in series with distributed generation and transmission system [23, 24]. Another key aspect of DVR systems is that they can be used for harmonic mitigation, fault current limiting, power factor correction, and reduction of transients, in addition to voltage sag mitigation [25]. A. Grid-integrated solar PV system with Dynamic voltage restorer Grid-integrated solar PV system with DVR is shown in Fig. 2. Power generated at solar PV system is stored in a battery as DC voltage. A boost converter is used in order to boost the battery voltage up to the required average DC grid voltage. Later this DC voltage is inverted to AC for Grid Voltage levels. At point of common coupling (PCC) power generated at the Solar PV distributed generation is integrated with the Grid [26]. Any disturbance in the grid like voltage sag/swell is later taken care by DVR which is placed along the grid. DVR with storage units try to inject real power into the grid in case of voltage sag and draw reactive power in case of voltage swell. These voltages come and go through inverters to battery storage units where power from distributed generation gets stored.

Review of Voltage Sag\Swell Mitigation Control Techniques …

193

Fig. 2 Grid integrated solar PV system with DVR

4 Control Techniques There are mainly three voltage sag compensation techniques that are used by dynamic voltage restorer in order to cater the power quality problems: pre-sag compensation, in-phase compensation, and phase advanced compensation technique [25]. Use of power electronics devices in a network makes the whole system non-linear. DVR proves to be a pure non-linear device due to the presence of power electronics switches in the inverter. Non-linear controllers used are ANN, Fuzzy logic, and Space Vector Pulse Width Modulation (SVPWM). ANN control method has adaptive and self-organization capacity. It also has an inherent learning capability that can provide improved precision by interpolation. Because of these reasons, ANN is chosen in the proposed system to overcome the shortcomings of the linear PID method [25]. A.

Artificial Neural Networks

ANN model proposed has an algorithmic procedure wherein it takes the data of the power quality, i.e., it analyses the issues with the power quality, randomizes the data, normalization of the data around a chosen point, and assigns particular weights and biases by backpropagation. Iterates through its neural network with suitable parameters. With the output values, it calculates the Mean Squared Error (MSE) and checks whether it is below the desired or agreeable error value. MSE is greater than the agreeable error then it stops iteration and updates the weights and biases and moves to the next layer of the ANN, else it continues to iterate until MSE is less than the desirable error value. With these steps in the algorithm, the ANN model is accurately able to calculate the sag/swell by iterating up to appropriate number of iterations and calculate the error with which the sag/swell occurs in comparison with the normal voltage level (voltage of the sine wave). This ensures that the ANN-based

194

N. Sowmyashree et al.

DVR maintains a stable network by injecting or removing the power from the line that has the power quality issue [27]. This algorithm when compared with other non-linear methods such as Fuzzy Logic or SVPWM gives more accurate results and reduced levels of THD (Total Harmonic Distortion). The mathematical model and the MATLAB Simulink model output match precisely and have been discussed thoroughly in [25]. This algorithm also overcomes the drawbacks of the PID control wherein it applies only to linear networks. B. Integrated Sliding Mode Control (ISMC) For controlling the DVR there are different methods as discussed above, another method that can be employed is Integrated Sliding Mode control, ISMC-DVR control strategy. In [4] they have implemented an in-phase compensation method. The DVR operation modes can be subdivided into 2 modes: Protection mode: To protect the DVR from the overcurrent on the load side due to large inrush currents and short circuit on load it is first isolated from the system whenever system parameters exceed the predetermined limits primarily current on the load side. So that the control system detects faults and manages bypass switches to remove DVR from the system. Injection mode: It is known that the primary function of DVR is compensating voltage disturbances on the distribution system; hence, to achieve these three single-phase AC voltages are injected in series with desired magnitude, phase, and wave shape. Among the different methods of DVR voltage injection, namely, presag compensation method, in-phase compensation method, and in-phase advanced compensation method; ISMC uses in-phase compensation method as mentioned before. Sliding mode controller (SMC): In this control method the system has to change its state to get the desired results. The main part of this method is to define the sliding line so that states of the system easily slide on these lines to attain stability at one point. x = −1Qx The above equation defines a sliding line. Another requirement is that since it is a two-stage process it is required to converge the sliding line and also stay on a sliding line hence this gives an exponential convergence with time constantQ. The overall system that is employing SMC with DVR includes a reference voltage calculator, SMC, inverter, filters, and injecting transformers. The output of the filter has to be connected to the system via an injection transformer; the function of DVR is to inject a reference voltage which is calculated by subtracting reference source voltage and actual source voltage. This generated reference voltage is subtracted from the actual injected DVR voltage the error obtained is processed by the sliding mode controller. Hence the output obtained from SMC is given to the inverter for switching purposes.

Review of Voltage Sag\Swell Mitigation Control Techniques …

195

C. Synchronous Reference Frame Another method for controlling the DVR is using Synchronous reference frame (SRF) theory. This theory involves numerous mathematical and arithmetic operations to generate the reference current signals. In this theory, first the current signals are transformed into rotating coordinates that is from a-b-c frame to d-q frame which is referred to as Park transform, and hence the loss component can be calculated from the PI controller and thus the d-q frame is transformed back to a-b-c frame using inverse Park transform. Now the obtained a-b-c coordinate system is considered as the generated reference current signals and is compared with 3-phase actual current waveform which is used to set up signals to power the switches of dynamic voltage restorer [4]. It is also necessary to use filters particularly low pass filters to remove the higher order components and also DC link capacitors are used to control the DC link voltages. [4]. D. Improvised Synchronous Reference Frame (Novel Theory) Since controlling using SRF method is complex due to the mathematical equations involved, [4] proposes a novel control methodology the transformation blocks are not utilized because the line voltage of the system is split into different phases and accordingly their error signals, also which can be given to PWM generator hence the trigonometric functions also reduce. By using simple mathematical operations like squaring each phase voltage later adding a delay and applying a square root to the obtained resultant signal which is considered to be the maximum value is then compared to a base value such as 1 p.u [28]. The resultant signal after processing it in a phase locked loop is the reference for the first phase which was divided. Similarly, for the other two phases this was repeated at a phase difference of 120° between any two phases. Thereafter the three-phase error signals are fed through the PWM generator which generates a gate signal thereby controlling the power switches of DVR. E. Comparison of SRF and Novel Control Theory for Voltage Sag Table 1 gives the comparison between SRF and Novel control theory approach with regard to voltage sag. F. Comparison of SRF and Novel Control Theory of Voltage Swell Table 2 gives the comparison between SRF and Novel control theory approach with regard to voltage swell. From the research findings it is observed that both conventional and intelligent control techniques have been proposed for the DVR to mitigate voltage sag and swell issues. In comparison with intelligent techniques, conventional control techniques are not reliable and fail to give efficient results in compensation with fast response, efficiency in mitigation of harmonics (THD) and increased energy consumption for voltage sag/swell compensation. It is proved that the conventional method fails to mitigate the voltage issues within a short period compared with their proposed method.

196

N. Sowmyashree et al.

Table 1 Comparison of SRF and novel control theory for voltage sag SRF control method

Novel control theory

Observations

One phase

Analysis carried out considering sag in one phase. The THD results for source voltage is 26.27% and for load voltage is 4.52%

The total harmonic distortion in source voltage and load voltage is 22.51% and 1.08% respectively

It is observed that using novel control method the reduction in load voltage is around 3.44%

Two phase

Considering the voltage across the load with sags in any two phases of supply voltage, the total harmonic distortion in supply voltage is 11.08 + I3%, whereas load voltage is 4.53%

Similarly, the harmonic distortion in source voltage is 11.08% while in load voltage it is 3.01%

There is a change in THD in load voltage for sag in two phases by decrease in 1.51%

Three phase

When there is voltage sag in 3 phases of the supply voltage the source voltage in this method has harmonic distortion of 11.01% while the load voltage is distorted by 4.15%

Whereas, in this method the distortion in source voltage is around 11.08% and load voltage has around 2.20% distortion

THD in load voltage for sag in three phases using novel control theory is 1.95% lesser than SRF method

Methods Number of phases

Hence, the table summarizes the different control techniques adapted with DVR along with their merits and demerits (Table 3).

It is observed that THD in load voltage for swell in three phases using the novel control theory is 2.14% lesser than SRF control method

When there is voltage swell in 3 phases of the supply voltage the source voltage in this method has harmonic distortion of 14.79% while the load voltage is distorted by 4.17% load voltage is distorted by 4.15%

Three phase

The total harmonic distortion for voltage swells in source voltage and load voltage is 11.67% and 2.03% respectively

The THD in load voltage for swell in two phases using the novel control theory is 1.09% lesser than SRF method

If there is swell in two phases of supply voltage following the DVR Whereas, in DVR controller using controller using SRF method the total harmonic distortion in novel control technique, the total source voltage and load voltage is 14.79% and 4.28% respectively harmonic distortion in supply voltage and load voltage are 11.67% and 3.09% respectively

THD in load voltage for swell in one phase using novel control theory is 0.13% less than that of SRF method

Observations

Two phase

Similarly, after injecting compensating signals by DVR using this method, the total harmonic distortion in supply voltage and load voltage are 11.25% and 4.44% respectively

Novel control theory

In case of swell in one phase of supply voltage crossing the limit p.u unit, by this method after compensating signals are injected by DVR the voltage magnitude across the load remains constant. The distortion in source voltage is around 14.79% and that in load voltage is 4.57%

SRF control method

One phase

Number of phases

Methods

Table 2 Comparison of SRF and novel control theory of voltage swell

Review of Voltage Sag\Swell Mitigation Control Techniques … 197

198

N. Sowmyashree et al.

Table 3 Control techniques adapted with DVR Control strategy

Merits and demerits

Remarks

PID

Conventional method used with DVR to compensate voltage in linear networks. Cannot be applied to non-linear networks

Load voltage restoration and THD mitigation capability is acceptable

Fuzzy logic

Easy to apply without any complications. Compensating performance of a transformer-less DVR is enhanced. THD mitigation is around 10.22% which is not acceptable when compared to other control techniques

Results show that fuzzy logic DVR control is very effective in damping system oscillations and improving the compensating performance compared to the traditional PI controller

Integrated sliding mode controller (ISMC)

SMC is robust and has good Load voltage restoration and dynamic response with stable THD mitigation capability is condition for supply and load moderate variations. Obtaining the convergence angle for sliding line and staying on this sliding line is difficult

Synchronous reference frame theory (SRF) and improved SRF method (novel)

Compared to previous control Use of ANN controller with the strategies it efficiently PWM increases its THD mitigates balanced and mitigation capability imbalanced sag to protect sensitive loads in power distribution line. In novel control method the limitation of using transformation blocks in SRF theory is eliminated

Artificial neural network (ANN)

It uses a technique of backpropagation increasing the value of every node with its weights thus providing accurate results with an increase in iterations. Algorithm is fast and efficient in terms of normalizing the data. Involves complex mathematical models.

Has the highest capability in THD mitigation.Widely used technique because of its accuracy.

5 Conclusion An overview of voltage quality challenges, like voltage sag and swell is discussed in this paper. A comparison of the different non-linear techniques employed with DVR for the mitigation of voltage sag and swell along with their pros and cons is presented. Based on the findings, it is concluded that a reduced THD of 2.83% is

Review of Voltage Sag\Swell Mitigation Control Techniques …

199

obtained using ANN [29] compared to other techniques. Fuzzy logic can provide the THD within the acceptable limits and hence it can also be considered. ANN is proven to be efficient in terms of fast response and reduced energy consumption. Training an AI model necessitates a large quantity of data and takes time, but with faster and dedicated GPUs, the ANN’s drawbacks can be overcome. With this consideration, ANN can be used as a better choice with DVR in the power system to foresee and mitigate a power quality issue before its occurrence.

References 1. Kow KW, Wong YW, Rajkumar RK, Rajku-mar RK (2016) A review on performance of artificial intelligence and conventional method in mitigating PV grid-tied related power quality events. Renew Sustain Energy Rev 56:334–346, 2016, ISSN 1364–0321 2. Haque Sunny MS, Hossain E, Ahmed M, Un-Noor F (2081) Artificial neural network based dynamic voltage restorer for improvement of power quality. In: 2018 IEEE energy conversion congress and exposition (ECCE), pp 5565–5572. https://doi.org/10.1109/ECCE.2018.8558470 3. Ipinnimo O, Chowdhury S, Chowdhury SP, Mitra J (2013) A review of voltage dip mitigation techniques with distributed generation in electricity networks 4. Suraya S, Irshad SM, Azeem MF, Al-Gahtani SF, Mahammad MH (2020) Multiple voltage disturbance compensation in distribution systems using DVR. Eng, Technol & Appl Sci Res 10(3):5732–5741 5. B. C., S. C. K., Sharma J, S. N. S., Guerrero JM (2022) Effect of fault ride through capability on electric vehicle charging station under critical voltage conditions. IEEE Trans Transp Electrification 8(2):2469–2478. https://doi.org/10.1109/TTE.2022.3145864 6. Kumar P, Arya SR, Mistry KD (2022) Performance enhancement of DVR using adaptive neural fuzzy and extreme learning machine-based control strategy. Int J Fuzzy Syst 24:3416–3430. https://doi.org/10.1007/s40815-022-01265-4 7. Akbari E (2020) Adaptive neuro-fuzzy inference system-based control strategy for dynamic voltage restorer (DVR) for both voltage sag/swell and unbalance compensation. Majlesi J Energy Manag 4(4) 8. Li Z, Guo X, Wang Z, Yang R, Zhao J, Chen G (2022) An optimal zero-sequence voltage injection strategy for DVR under asymmetric sag. IEEE J Emerg Sel Top Power Electron 10(2):2595–2607. https://doi.org/10.1109/JESTPE.2022.3149769 9. Noro Y, Naoi S, Toba K, Kimura M, Minegishi T, Shimizu M, Okuda Y (2012) Power fluctuation suppression system for large scale PV. In: PES T&D. pp 1–6 10. Farhoodnea M, Mohamed A, Shareef H, Zayandehroodi H (2013) Power quality analysis of grid-connected photovoltaic systems in distribution networks. Przeglad Elektrotechniczny (Electrical Review), 208–213 11. Marcos J, Marroyo L, Lorenzo E, García M (2012) Smoothing of PV power fluctuations by geographical dispersion. Prog Photovolt Res 20(2):226−37 12. Saidan A, Mirabbasi D, Heidari M (2010) The effect of DG on voltage flicker and voltage sag in closed loop distributed system. In: Industrial electronics and applications (ICIEA). pp 68–72 13. Angelino R, Carpineili G, Proto D, Bracale A (2010) Dispersed generation and storage systems of providing ancillary services in distributed systems. In: Power electronics and electrical drives automation and motion (SPEEDAM). pp 343–351 14. González P, Romero-Cadaval E, González E, Guerrero MA (2011) Impact of grid connected photovoltaic system in the power quality of a distribution network. In: Doctoral conference on computing, electrical and industrial systems. Springer, Berlin, Heidelberg, pp 466–473 15. Anwari M, Hamid MI, Rashid MIM (2009) Power quality analysis of grid-connected photovoltaic system with adjustable speed drives. In: 2009 IEEE PES/IAS conference on sustainable alternative energy (SAE). pp 1–5

200

N. Sowmyashree et al.

16. Golovanov N, Lazaroiu GC, Roscia M, Zaninelli D (2013) Power quality assessment in small scale renewable energy sources supplying distribution systems. Energies 6(2):634–645 17. Kontogiannis K, Vokas G, Nanou S, Papathanassiou S (2013) Power quality field measurements on PV inverters. Power 2(11) 18. Al-Mathnani AO, Shareef H, Mohamed A, Ali MAM, Hannan MA (2010) Power quality improvement using DVR with two fast vector control. In: 2010 4th international power engineering and optimization conference (PEOCO). pp 376–381 19. Gao T, Cao J, Xu Y, Zhang H, Yu P, Yao S (2013) From power quality to power experience. In: 2013 fourth international conference on networking and distributed computing, pp 116–120 20. http://www.powerqualityworld.com/2011/03/voltage-sags-dips-power-quality-basics.html 21. Shen W, Zhu Y (2011) Impacts of small photovoltaic power station on voltage sag in low-voltage distribution network. In: 2011 international conference on electrical and control engineering. pp 1585–1588 22. https://www.google.com/searchq=why+voltage+swell+occurs&rlz=1C1CHBF_enIN972 IN972&oq=why+voltage+swell+occurs&aqs=chrome.0.69i59.1861j0j9&sourceid=chrome& ie=UTF-8 23. Musarrat MN, Fekih A, Islam MR (2021) An improved fault ride through scheme and control strategy for DFIG-based wind energy systems. IEEE Trans Appl Supercond. 31(8):1–6. Art no. 5401906. https://doi.org/10.1109/TASC.2021.3096181 24. Mishra SK, Bhuyan SK, Rathod PV (2022) Performance analysis of a hybrid renewable generation system connected to grid in the presence of DVR. Ain Shams Eng J 13(4):101700 25. Sunny MSH, Hossain E, Ahmed M, Un-Noor F (2018) Artificial neural network based dynamic voltage restorer for improvement of power quality. In: 2018 IEEE energy conversion congress and exposition (ECCE). pp 5565–5572 26. Parsaeifard AH, .Manbachi M, Kopayi MBA, Haghifam MR (2010) A market-based generation expansion planning in deregulated environment based on distributed generations development. In: Probabilistic methods applied to power systems (PMAPS). pp 677–684 27. Liasi SG, Bina MT (2019) A neural network-based control strategy for three-phase four-leg dynamic voltage restorer for both voltage sag/swell and harmonic compensation. Int Power Syst Conf (PSC) 2019:478–484. https://doi.org/10.1109/PSC49016.2019.9081537 28. Akuskar CS, Kamble SS (2022) Simulation of dynamic voltage restorer using SRF theory to mitigate voltage sag and swell. In: 2022 2nd Asian conference on innovation in technology (ASIANCON). pp 1–6. https://doi.org/10.1109/ASIANCON55314.2022.9909137 29. A. M J, S. N, Shashikala MS (2019) Power quality enhancement using dynamic voltage restorer (DVR) by artificial neural network and hysteresis voltage control techniques. In: 2019 global conference for advancement in technology (GCAT). pp 1–6. https://doi.org/10.1109/GCAT47 503.2019.8978333

Parametric Sensitivity Analysis of STATCOM Supplementary Modulation Controller Incorporated in SMIB System Dinesh Shetty and Nagesh Prabhu

Abstract The low-frequency oscillations (LFOs) in power system occur due to inadequate damping torque of the system and LFOs can be damped using a Static Synchronous Compensator (STATCOM) incorporated with Supplementary Modulation Controller (SMC). However, variation in the parameters of SMC can cause oscillatory modes to collide both in eigenvector and eigenvalues. The major focus of this paper is to discuss a parametric variation of STATCOM SMC and its effect on the stability of the system. The investigation also identifies the limitations on damping of the system for parametric variations of the controller without and with energy sources (ES) at the DC bus of STATCOM. The sensitivity of eigenmodes for system parametric changes is analyzed extensively for the various operating conditions of STATCOM to obtain the safe operating range of controller parameters. The illustrative case studies on SMIB system equipped with STATCOM with and without an energy source shows that identification of a safe range of STATCOM SMC parameters and operating points is highly essential to ensure good damping of the swing mode of generators. Keywords STATCOM · Reactive current · Energy sources · Eigenvalues

1 Introduction The power system undergoes LFO because of the increasing power transfer via the transmission network. Typically, the LFO oscillates between 0.2 and 3 Hz. In general, if the system has positive damping at these frequencies, the oscillations often diminish quickly, and the system stays stable. Depending on particular operational conditions, D. Shetty · N. Prabhu (B) NITTE (Deemed to be University), Dept. of Electrical and Electronics Engineering, NMAM Institute of Technology, Nitte 574110, Karnataka, India e-mail: [email protected] D. Shetty e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 S. Kajampady and S. T. Revankar (eds.), Advances in Renewable Energy & Electric Vehicles, Lecture Notes in Electrical Engineering 1083, https://doi.org/10.1007/978-981-99-6151-1_14

201

202

D. Shetty and N. Prabhu

the LFO may intensify or continue to expand to the point where it throws synchronous machines out of sync. The negative impact of these oscillations enhances the voltage, power flow, torque, and speed fluctuations for an extended period of time. These oscillations’ damping depends on the state of the system. Further, the eigenvectors congruity between the two oscillatory modes for the system parametric variation can cause mode coupling and the interaction of two oscillatory modes [1–5]. It is observed that the change in damping of one mode is in the direction opposite to that of change in damping of other mode and the system matrix is found to be non-diagonalizable under mode coupling. When two complex pairs of eigenvalues coincide both in frequency and damping, these modes move close to each other and collide due to the variation of two independent parameters [3]. The post-collision effect can result in one of the modes to move to an unstable region. Generally, the exact interaction between two complex eigenvalues will not happen when only one parameter is varied, however, the complex eigenvalues can pass through near collision. In a power system, the phenomenon of exact interaction may not occur but certainly, two eigenvalues come very close. Eigenvectors and eigenvalues are very much sensitive to the parametric changes when the system is in the vicinity of collision and eigenvalues move very quickly and take approximately 90° turns in S-plane [3, 4]. It is also observed that one of the modes might incline and move toward the right half of S-plane after the collision. In the cases where modes collide only in eigenvalues and the system matrix is diagonalizable, then it is termed as weak resonance. Padiyar and Saikumar [6] have considered a simple fourth-order SMIB system with one axis generator model with Static exciter to study the interaction between the Swing mode and Exciter mode. The collision is observed between these modes due to variations in generator power dispatch and terminal voltage. Very interesting and noticeable study by Padiyar and Saikumar [6, 7] substantiates that the effect of parametric variations is an essential requirement while designing a power oscillation damping controller. The interaction of exciter mode [EM] and swing mode [SM] for the multimachine system has been carried out in [8] for variation in the parameters of STATCOM SMC. The SM that interacts with the EM is isolated by the application of the multimodal decomposition concept, and the EM is identified by computing participation factors. The asymptotic behavior of the modes in a reduced system is compared with the root loci of the full system [8]. SMC is desired for damping LFO and while designing SMC, care to be taken not to cause mode coupling to ensure the stable operation of the power system. In the present scenario, renewable energy (RE) generation is promising and gathering importance. STATCOM can be effectively used to supply active power when RE is integrated at the DC bus of STATCOM and can be denoted as STATCOM-ES [9, 10]. The contribution of the work is as follows: I. In this paper, the investigation on behavior of complex eigenvalues in SMIB system with STATCOM-ES connected to the load bus is studied for various operating conditions.

Parametric Sensitivity Analysis of STATCOM Supplementary …

203

II. The present work also emphasizes on the asymptotic behavior of the system oscillatory modes due to reactive current injection and reactive power absorption at load bus by STATCOM. The investigation on the sensitivity of oscillatory modes for change in controller parameters associated with SMC is very essential. The range of safe operating values for STATCOM SMC controller gain for the stable SM and EM is observed by means of the root loci. III. The prediction of the stability of oscillatory modes causing limits on damping ratio is investigated for different operating points of STATCOM (reactive current injection/absorption) is investigated to identify the region of mode coupling condition in the vicinity of the operating point. The paper is organized as follows: Sect. 2 describes the system descriptions, results analysis in the Sect. 3, here all 5 cases are studies for the occurrence of collision for SMIB system incorporated with STATCOM with reactive power and active power supply. Section 3 discussion on investigation and conclusion is drawn in Sect. 4.

1.1 System Description The Single Machine Infinite Bus (SMIB) System [11, 12] with STATCOM-ES consists of a synchronous generator represented by the 1.0 model is considered for the analysis and is shown in Fig. 1. Intermediate bus 3 is a load bus where STATCOM-ES is connected for active and reactive current injection. The data for the study system of Fig. 1 is given below. System data: Data given is in pu on a common base of 100 MVA with Frequency = 60 Hz. Machine data:

Fig. 1 SMIB system with STATCOM-ES

204

D. Shetty and N. Prabhu

Ra = 0; xd = 2.442, xq = 2.421, xd' = 0.83, xq' = 2.421, ' Td0 = 5.33s, H = 2.832, Dg = 0

Exciter: K A = 450, TA = 0.6 Line 1–3: R = 0, X = 0.168, B = 0.02 Line 2–3: R = 0, X = 0.126, B = 0.016 Statcom-ES: I p + jI r is the current injected by the STATCOM represented in pu. Load: PL = 1.0, Q L = 0.3 Infinite bus voltage: E b = 1.0 The linearized model of the SMIB system with STATCOMSMC is represented by the state space form as ˙ = AS ΔX + BS ΔU ΔX

(1)

ΔY = CS ΔX + DS ΔU

(2)

where X is the state variable vector for the full system under consideration ΔX = [Δδ, ΔS, ΔE'q , ΔE'fd ] Δδ = angle of swing mode ΔS = slip of swing mode ΔE 'q = quadrature voltage of the generator

(3)

Parametric Sensitivity Analysis of STATCOM Supplementary …

205

ΔE 'fd = field voltage of the generator ΔU = ΔIr = the reactive current by STATCOM ΔY = ΔVl = voltage at the load bus ˙ = The eigenvalues are computed for the SMIB system without SMC, i.e., ΔX AS ΔX at Δ U = 0 is shown in Table 1. It is noticed that the system under study is unstable for the operating point considered. The oscillatory modes of the unstable system are identified by eigenvalues as SM and EM using the participation factor [13–18]. It is observed that the real part of the swing mode is positive and is unstable. This necessitates the requirement of a damping controller to stabilize the system to damp low-frequency oscillations associated with swing mode. LFO due to unstable swing mode can be effectively damped by SMC incorporated with STATCOM to modulate reactive and/or active current reference. The block diagram of the supplemental modulation controller (SMCQ) for STATCOM-ES, which controls the reactive current I r , is shown in Fig. 2. Thevenin Voltage Vlth , which is synthesized using STATCOM current I r and STATCOM bus voltage V l , serves as the control signal under consideration. Rl The control law is given by dΔi = K r ΔVlth [5] Ir is the reactive current that dt the STATCOM injected into the bus, K r is the gain of the reactive current modulation controller, and X 1 is the value of the adjustable parameter used to synthesize Thevenin’s voltage. Where the STATCOM is connected, Vl is the measurement of the load bus voltage. The first-order plant transfer function with the time constant TP (0.02 s) describes the dynamics of the STATCOM. TC is the derivative circuit’s s and it is considered as 10 ms. T1r and T2r are the time constants time constant 1+sT c correspond to the compensator and m r denote the number of stages associated with compensator block of STATCOM SMCQ . The transfer function of the controller is given by Table 1 Eigenvalues of the SMIB system without a controller

Eigenvalues

Modes

0.52793 ± j6.4073

Swing mode

−1.5903 ± j5.9705

Exciter mode

Fig. 2 STATCOM reactive current supplementary modulation controller (SMCQ )

206

D. Shetty and N. Prabhu Ir (s) Vl (s)

=

(

−s K r 1+sTc

)(

(1+sT 1r ) (1+sT2r )

)m r (

1 1+sTp

) (4)

The work presented in this paper majorly concentrates on the analysis of damping of swing mode (SM) and exciter modes (EM) for the variation in tunable parameters K r and X 1 while observing the existence of the mode interaction phenomenon. In the sensitivity analysis, the effectiveness of the phase compensator while tuning the controller parameter of STATCOM SMCQ for various operating currents of STATCOM is investigated for the first time.

2 Result and Analysis Case 1: Injection of STATCOM reactive current Ir0 The following observations are made for the system under study with the controller parameters for quiescent values of Ir0 for STATCOM operating in capacitive mode, i.e., injection of reactive current. For K r = 2.0307 and X 1 = 0.013718, the interaction is observed. By varying K r and keeping X 1 constant, the two modes of interest EM and SM vary both in damping as well as frequency. The system is found to be stable for the value of K r > 1.878 with the real part of eigenvalues of swing mode at the left half of S-plane as shown in Fig. 3. The asymptotic behavior of the eigenvalues is such that, the damping of swing mode increases initially and exciter mode decreases as K r is increased. At K r = 2.0307, the exciter mode and swing mode becomes equal both in damping and frequency as depicted in Fig. 4.

Fig. 3 Rootloci of the system for varying K r

Parametric Sensitivity Analysis of STATCOM Supplementary …

207

Fig. 4 Variation of damping ratio as a function of K r

At this point, even a slight increase in K r results in swing mode frequency to drop and the exciter mode frequency to raise by limiting the damping ratio. The sensitivity of eigenvalues of SM and EM for the variation in reactance value X 1 in the proximity of collision is also investigated. At ΔX 1 > 0, with the increase in K r , the damping of the SM increases and frequency reduces and damping of EM increases while the frequency remains almost constant as depicted in Fig. 5, after nearing to common interaction, the frequency of SM reduces and at exciter mode frequency increases and damping of the modes remains almost constant.

Fig. 5 Rootloci of the system for varying K r with ΔX 1 > 0

208

D. Shetty and N. Prabhu

Fig. 6 Rootloci of the system for varying K r with ΔX 1 < 0

The shift in the asymptotes of modes of about 90° is observed as shown in Fig. 6 for ΔX 1 < 0, i.e., at |ΔX 1 |= 0.001. As K r is increased the frequency of swing mode decreases and damping increases. The frequency and damping of the exciter mode decrease. However, after passing through near Collision point, the frequency of swing mode increases and that of exciter mode decreases sharply but there is no significant change in the damping of both modes, and they remain nearly constant. EM and SM do not come closer. The sensitivity of eigenmodes for the large variation in X 1 is shown in Fig. 7. The trajectories of two oscillatory modes have been plotted and observed that, as X 1 increased, the unstable swing mode becomes stable and the exciter mode becomes unstable as shown in Fig. 7. At X 1 =0.013718 the modes approach very close in frequency and damping and also collides as shown in Fig. 8. The observations from the trajectories in Fig. 7 ascertain the fact that the increase in the value of X1 causes exciter mode to become unstable which further necessitates the design of controllers to ensure a stable system. The investigation is further extended to analyze the stability of the system for different operating values of STATCOM reactive current Ir0 and the trajectories of eigenvalues are plotted for variation of K r as given in the Fig. 9. It is observed that increasing the reactive current of I r in the capacitive operation of STATCOM, the swing mode of the system shifts toward the right half of S-plane indicating a reduction in stability margin. The trajectory of the eigenvalues also interprets that the increase in operating values of Ir0 causes the trajectories to move away from the near collision point and the damping of the modes remains almost constant in all the trajectories after crossing the approximately near collision points for increasing values of K r .

Parametric Sensitivity Analysis of STATCOM Supplementary …

209

Fig. 7 Rootloci of the system for varying X 1 with K r constant

Fig. 8 Variation in damping ratio as a function of X 1

The real part of the eigenvalues remains approximately the same after crossing near the resonance point for all the operating points of Ir0 . Table 2 shows the eigenvalues of the system for different operating values of Ir0 in the capacitive mode of operation for K r = 2.0307 and X 1 = 0.013718. It can be noted from Table 2 that the damping of swing mode decreases as Ir0 is increased and becomes unstable at fixed values of K r and X 1 . Hence, it is essential to identify the safe range of the controller parameters for any operating point reactive current Ir0 of STATCOM to ensure the stability and adequate damping of swing mode oscillations. The operating value of

210

D. Shetty and N. Prabhu

Fig. 9 Root loci for various operating values of Ir0

reactive current Ir0 = 0.1 is considered for the investigation of a range of safe operating values for controller parameters for the system. X 1 = 0.013718 and K r = 2.0307 are considered as a base case for further analysis in the paper. For increasing values of K r , and change in X 1 the different eigenvalues trajectories are plotted as shown in Fig. 10. It is interesting to observe that, as X 1 is decreased from the base case, the frequency of swing mode decreases while the damping increases, and the frequency of exciter mode decreases and damping keeps reducing. After passing the near collision, the frequency of exciter mode increases while that of swing mode reduces. However, the damping of both modes remains almost constant. Further reduction in X 1 causes both the modes to shift their asymptotes to approximately 90z. The frequency of swing mode increases and exciter mode decreases and damping will also find sluggish variation. An increase in X 1 beyond the base case causes the trajectories to move away from the near collision. The larger values of K r are required to ensure the stability of swing mode as shown in Fig. 10. It should be noted that any further increase in X 1 beyond (0.031718) makes the swing mode unstable irrespective of the value of K r . The eigenvalues of swing modes for different values of X 1 and K r are tabulated in Table 3. It is interesting to note that, as K r increases, the damping of swing mode increases and SM moves to a stable region without appreciable change in the frequency. However, there are limitations to the increase in K r , beyond which the swing mode eigenvalues become unstable. It is found that, at K r = 2, an increase in X 1 causes a decrease in the damping of swing mode and becomes unstable, and a properly designed compensator as in Eq. (4) is desired to ensure stability. Therefore the study suggests that collision will also help in choosing the safe ranges of operating values of controller parameters for STATCOM for effective damping of swing mode oscillations. To increase the

Parametric Sensitivity Analysis of STATCOM Supplementary …

211

Table 2 Eigenvalues of the swing modes of the system for different operating points and SMCQ parameter Eigen Ir0 = 0 modes

Ir0 = 0.1

Ir0 = 0.2

Ir0 = 0.3

SMC Q parameters K r = 1 and X 1 = 0.013718 SM EM

0.377 ± j6.08 −0.98 ± j6.07

0.44124 ± j 5.9937 −1.0532 ± j 6.0324

0.51232 ± j 5.9037 −1.1305 ± j 5.9891

0.58924 ± j 5.8163 −1.2148 ± j 5.9418

SMC Q parameters K r = 2.0307 and X 1 = 0.013718 SM

−0.230 ± j 5.92

0.076 ± j 5.7934

EM

−0.230 ± j 5.92

−0.54859 ± j 5.922

0.21456 ± j 5.7026

0.33268 ± j 5.6182

−0.69056 ± j5.8839

−0.81288 ± j 5.8388

SMC Q parameters K r = 4 and X 1 = 0.013718 SM

−0.122 ± j6.428 −0.13908 ± j 6.292

−0.15837 ± j 6.1401 −0.18132 ± j 5.956

EM

−0.377 ± j 5.05

−0.34219 ± j 5.0919 −0.31939 ± j 5.1479

−0.3615 ± j 5.0677

Fig. 10 Root loci for change in X 1 values for Ir0 = 0.1

stability margin of the system, the range of safe operating values of X 1 can be increased using a two-stage compensator with mr = 2 in Eq. (4) in the design of SMCQ . In this paper, the compensator design [18] is carried out for the unstable case of K r = 2 and X 1 = 0.023718. The designed parameters of the compensator are T 1r = 0.49 and T 2r = 0.63 to ensure the enhanced stability margin of the system, the system with compensator increases in damping of swing mode for an increase in K r . It is observed from the trajectory that, after crossing the near collision there is no appreciable change in the damping ratio of Swing mode.

0.0915 ± j 5.833

−0.32913 ± j 5.347

−0.47766 ± j4.942

Kr = 3

Kr = 5

X 1 = 0.010718

Kr = 2

Controller gain

−0.4634 ± j4.934

−0.312 ± j 5.341

0.0911 ± j 5.822

X 1 = 0.011718

−0.4488 ± j 4.923

−0.2953 ± j 5.335

0.09125 ± j5.81

X 1 = 0.012718

Table 3 Eigenvalues of swing mode for change in X 1 and K r parameters of SMCQ

−0.4342 ± j4.9134

−0.278 ± j 5.329

0.09197 ± j 5.798

X 1 = 0.013718

−0.4036 ± j 4.895

−0.2441 ± j5.315

0.095197 ± j5.777

X 1 = 0.015718

−0.2795 ± j4.815

−0.11712 ± j5.256

0.12941 ± j 5.681

X 1 = 0.023718

212 D. Shetty and N. Prabhu

Parametric Sensitivity Analysis of STATCOM Supplementary …

213

Table 4 compares the eigenvalues of the swing mode without and with the compensator. It is observed that the compensator improves the damping of the swing mode. Case 2: Absorption of STATCOM reactive current Ir0 : The Investigation is also carried out with STATCOM, operating in inductive mode, i.e., absorbing the reactive power. Figure 11 shows the trajectory of the SM and EM for variation in K r . The trajectories of eigenvalues of SM and EM show that for X 1 is at base case (X 1 = 0.013718) and at any operating values of Ir0 , the swing mode moves from unstable to stable region for variation in K r as shown in Table 5. The eigenvalues of SM and EM are shown in Table 5 for different values of K r and Ir0 at the base case of X 1 . The safe operating ranges of K r to maintain the stable swing modes reduce as Ir0 increases. Initially, the damping and frequency of SM increase with an increase in K r , on crossing near the collision point, frequency Table 4 Eigenvalues of swing mode with and without compensator for SMCq at Ir0 = 0.1 Controller gain

X 1 = 0.023718 Without compensator

With compensator

Kr = 2

0.12941 ± j 5.6814

Kr = 3

−0.11712 ± j 5.2563

−0.21 ± j5.648

0.0118 ± j 5.72

Kr = 4

−0.2000 ± j 4.98

−0.25 ± j5.5148

Kr = 5

−0.27954 ± j4.815i

−0.29585 ± j 5.439

Fig. 11 Root loci of the system for X 1 = 0.013718 with K r varying, Ir0 absorption

214

D. Shetty and N. Prabhu

Table 5 Eigenvalues for various operating values of Ir0 during the inductive mode of STATCOM Eigen modes

Ir0 = 0

Ir0 = −0.1

Ir0 = −0.2

Ir0 = −0.3

SMC Q parameters K r = 1 and X 1 = 0.013718 SM EM

0.377 ± j6.08 −0.98 ± j6.07

0.306 ± j6.18

0.234 ± j6.29

0.16 ± j6.41

−0.907 ± j 6.18

−0.833 ± j6.12

−0.75 ± j6.13

SMC Q parameters K r = 2.0307 and X 1 = 0.013718 SM

−0.230 ± j 5.92

−0.17 ± j6.29

−0.144 ± j6.48

−0.12 ± j6.64

EM

−0.230 ± j 5.92

−0.296 ± j 5.68

−0.32 ± j5.62

−0.33 ± j5.58

SMC Q parameters K r = 4 and X 1 = 0.013718 SM

−0.122 ± j6.428

−0.10755 ± j6.558

−0.094 ± j6.68

−0.083 ± j6.8

EM

−0.377 ± j 5.05

−0.395 ± j5.058

−0.406 ± j5.06

−0.418 ± j5.068

increases damping reduces and the system becomes unstable. However, the exciter mode remains stable. Therefore, to enhance the stability of swing mode, with a higher value of gain K r inclusion of the compensator is suggested.

3 Discussion Based on the case studies, it is observed that the stability of the system is dependent on the operating values of the reactive/active current of the STATCOM. The connection of STATCOM at the load bus emulates shunt admittance can change the net impedance seen at the generator terminals and can greatly influence the stability of the system. The net impedance of the system seen from the generator without STATCOM is found to be Z GT = 0.0157 + j0.287 p.u. It is observed that, while supplying reactive power, the STATCOM emulates negative susceptance and reduces the net resistance of the system as shown in Table 6. This causes the increased sensitivity of eigenvalues to parametric variation and aggravates the modal interactions to cause collision conditions. However, while absorbing reactive power the emulated susceptance is positive and the net resistance seen at the generator terminals increases which results in stable swing mode. Table 6 Changes in net impedance for different operations of STATCOM Case no.

Different operations of STATCOM

1.

Supplying −j0.3006 reactive power

0.0135 + j0.283

0.3327 ± j5.6182

2.

Absorbing j0.3147 reactive power

0.0173 + j0.291

−0.12 ± j6.64

Emulated admittance Net impedance of the from STATCOM Y s = system seen from generator terminal Z t Gs + jBs

Eigenvalues of swing mode

Parametric Sensitivity Analysis of STATCOM Supplementary …

215

4 Conclusion The detailed investigations on the swing mode of LFO of SMIB system equipped STATCOM-ES are carried out. The sensitivity of SMC gain K r and tunable parameter X 1 is investigated. Based on the observations of asymptotes of oscillatory modes, the collision of modes and the system stability are analyzed. The asymptotic behavior of the oscillatory modes facilitates to observe the transition of eigenvalues for variation in control parameters. The following inferences are drawn based on the observation of the results of case studies. When STATCOM is supplying reactive power (Capacitive mode), the exact collision is observed for the low value of reactive current injection. As the reactive current injection increases the eigenvalues trajectories move away from each other and swing mode becomes unstable for very large values of capacitive reactive current. It is interesting to note that STATCOM in the capacitive mode of operation emulates negative susceptance and reduces the net resistance of the system seen from generator terminals while supplying reactive power which is one of the primary reasons for the occurrence of instability and collision. This condition limits the variation in the damping ratio as well. The tunable parameter X 1 to synthesize Thevenin voltage and SMCQ gain K r is varied to maintain the stability of the swing mode. However, the sensitivity analysis for parameter variations shows that mode interaction can also occur for the larger values of current with a reduced value of X 1 . Identification of a safe operating region for tunable parameters is found to be very essential to ensure system stability while designing SMC. When STATCOM is absorbing reactive power, the asymptotic behavior of the oscillatory modes does not exhibit exactly collision and is to be noted that STATCOM emulates positive susceptance and increases the net resistance of the system seen from generator terminals. However, the sensitivity analysis for the variation of controller parameters shows that the calculation of a safe operating range for tunable parameters is essential to ensure system stability. The stability margin of the system is increased by using a two-stage phase compensator which increases the range of safe operating values of parameters in the design of supplementary modulation controller SMCQ which modulates reactive current reference. It is concluded parametric sensitivity may lead to system instability with an indication of the system damping getting limited.

References 1. Seiranyan AP (1994) Collision of eigenvalues in linear oscillatory systems. J Appl Math Mech 58(5):805–813 2. Seiranian P (1993) Sensitivity analysis of multiple eigenvalues. Mech Struct Mach 21(2):261– 284 3. Zhang DJ, Greene S, Engdahl H, Sauer PW (2001) Is collision a precursor to power system oscillations? IEEE Trans Circuits Syst I, Fundam Theory Appl 48(3):340–349

216

D. Shetty and N. Prabhu

4. Wenjuan D, Zen Z, Wang H (2019) The subsynchronous oscillations caused by an LCC HVDC line in a power system under the condition of near strong modal resonance. IEEE Trans Power Delivery 34(1):231–240 5. Padiyar KR, SaiKumar HV (2003) Analysis of Strong resonance in power systems with STATCOM supplementary modulation controllers. In: Proc. IEEE TENCON, Bangalore, India, Oct. 15–17, 2003 6. Jyothsna TR, Vaisakh K (2011) Effects of collision in tuning of multiple power system stabilisers. IET Gener Transm Distrib 5(11):1155–1164 7. Padiyar KR, SaiKumar HV (2004) Investigations on strong resonance in multimachine power systems with STATCOM supplementary modulation controller. IEEE Trans Power Syst 2:754– 776 8. Mala RC, Prabhu N, Gururaja Rao HV (2017) Performance of STATCOM-ES in mitigating SSR. Int J Powerlectronics Drives Syst 9. Tariq Masood and D.P.Kothari, “Centralised / Decentralised FACTS Controllers in Electric grid”, New Age Publishers New Delhi, 1st edition, 2019. 10. Padiyar KR (2002) Power system dynamics – stability and control. BS Publications, 2nd edn 11. Kothari DP, Nagrath IJ (2019) Power systems engineering. Tata McGraw Hill, New Delhi, 3rd edn 12. Dinesh Shetty and Nagesh Prabhu (2019) Low frequency oscillation detection in smart power system using refined Prony analysis for optimal allocation of supplementary modulation controller. In: 3rd international conference on trends in electronics and informatics (ICOEI), IEEE conference proceeding, October 2019 13. Setiadi et al (2018) Modal interaction of power systems with high penetration of renewable energy and BES systems. Electr Power Energy Syst 97:385–395 14. Krismanto AU, Mithulananthan N, Krause O (2016) Microgrid impact on low frequency oscillation and resonance in power system. In: 2016 IEEE innovative smart grid technologies - Asia (ISGT-Asia). pp 424–429 15. Lastomo D, Setiadi H, Faisal M, Ashfahani A, Bangga G, Hutomo G, et al (2017) The effects of energy storage on small signal stability of a power system. In: 2017 international seminar on technology and its application (ISITIA), Surabaya, Indonesia 16. Taufik M, Lastomo D, Setiadi H (2017) Small-disturbance angle stability enhancement using intelligent redox flow batteries. In: 2017 4th international conference on Electrical Engineering, Computer Science and Informatics (EECSI 2017), Yogyakarta, Indonesia 17. SaiKumar HV (2006) Investigations on small signal stability of power systems affected by FACTS supplementary modulation controllers,” Ph.D. Dissertation, Dept. Elect. Eng., Indian Inst. Science, Bangalore, India 18. Frankline GF, David Powell J, Abbas Emami N (2014) Feedback control of dynamic systems, 7 edition (Fifth Edition). Pearson Prentice Hall

Review on the Addition of Antioxidants and Nanoparticles to Natural Ester as an Alternative to Transformer Oil S. N. Deepa, A. D. Srinivasan, M. Anusha, K. T. Veeramanju, and M. R. Chaitra

Abstract The power transformer is a critical equipment in the transmission and distribution network that must be managed to ensure uninterrupted power service. Progress of reliable, ecologically harmless insulating oil for the transformer is a boundless exertion of the electrical industry. Current investigation is based on natural ester fluid, which displays exceptional dielectric presentation and environment friendly characteristics. To overcome the disadvantages of natural esters addition of antioxidants and nanoparticles to natural esters was introduced which showed an improved insulating properties in natural esters. Natural esters with addition of antioxidant and nanoparticles are the emerging trend as the potential replacement for the conventional mineral oil. This study details the overview of the effect of adding antioxidants and nanoparticles to the natural esters on the critical properties of natural esters. Natural esters studied in this work are sunflower oil (SFO), rice bran oil (RBO), neem oil (NO), castor oil (CO), mahua oil (MO), and coconut oil (CCO). Keywords Natural esters · Antioxidants · Nanoparticles · Transformer oil · Insulating properties

S. N. Deepa (B) · A. D. Srinivasan · M. Anusha · K. T. Veeramanju · M. R. Chaitra Department of Electrical and Electronics Engineering, JSS Science and Technology University, Mysuru, Karnataka 560007, India e-mail: [email protected] A. D. Srinivasan e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 S. Kajampady and S. T. Revankar (eds.), Advances in Renewable Energy & Electric Vehicles, Lecture Notes in Electrical Engineering 1083, https://doi.org/10.1007/978-981-99-6151-1_15

217

218

S. N. Deepa et al.

1 Introduction The transformers are huge energy equipment that transforms power from power generation source to different distribution channels and are installed in different places throughout the electrical supply distribution system. High voltage transformers are used to step up and step down the generated voltages as per the requirements of different distribution channels. These high-voltage transformers dissipate enormous amount of heat during the operation due to the rise in temperatures caused by the energy losses, which may lead to equipment failure and creates a hazardous environment. Thus, cooling of high-voltage transformers plays a vital role in the safe operation and proper maintenance of the equipment. Cooling of high-voltage transformers can be achieved by proper insulation of the transformer. Insulation of high-voltage transformers is done using solid, liquid, or gas mediums. Liquid insulation is assumed to be an indispensable part of keeping the transformer in great condition like blood in the human body. Specifically, they are answerable for the utilitarian workableness of the dielectric (protection) framework, the state of which can be a conclusive factor in deciding the life expectancy of the transformers [1]. In contrary to dry type transformers, several customers preferred liquid insulated coolant transformers (oil filled transformers) [2]. In liquid-filled transformers, the protective framework should provide acceptable mechanical strength for the windings to resist operational complexities, suitable dielectric strength for the operating and test voltage stresses, and substantial cooling channels to allow the liquid to dissipate heat released in the windings [3]. The technical and financial losses are directly dependent on the percentage of the transformer failures. Transformer insulation oil failure is the most severe type of failure among all other transformer failures [4]. However, it is crucial that the insulating oil be compatible with all of the other components of the transformer, especially the cellulose insulation, in addition to its various physicochemical, thermal, and dielectric characteristics [5]. Although the typical average life expectancy of a transformer ranges from 35 to 40 years, the oil ageing process must be purposeful and independent of any adverse impacts on the other components that constitute the transformer [6, 7]. A compound insulating medium comprised of oil and paper operates as the dielectric medium in a transformer. The transformer’s insulating oil is likely to serve as an insulant, coolant, barrier to protect the core, and diagnostic tool [8]. There are several critical properties which include electrical, physical, and chemical properties among them breakdown voltage, viscosity, flash, and fire point perform a vital role in determining the efficiency of the transformer oil [9]. Therefore, a careful selection of the insulating medium is crucial, together with knowledge of its characteristics, compliance with several other transformer components, and environmental consequences [10, 11]. Petroleum biproduct used to satisfy this condition is mineral oil but the only concern that it is non-biodegradable as it’s a petroleum product, use of silicon oil did satisfy the insulation condition but the problem of biodegradability remained as of mineral oil. Current trends and research are based on using natural esters and natural esters with additives as a substitute to mineral oil. This paper

Review on the Addition of Antioxidants and Nanoparticles to Natural …

219

details about the effect of adding antioxidants and nanoparticles to natural esters on its critical insulating properties.

2 Insulating Oils The following insulating oils are available for transformers: mineral oil, silicon oil, synthetic ester, and natural ester.

2.1 Mineral Oil When Eliu Thompson’s work on chemically stable mineral oil as transformer oil was patented, general electricals used it as transformer oil for the first time in 1892 [12]. Different forms of mineral oil have been used extensively for insulation since the 1970s as an alternative to harmful polychlorinated biphenyls (PCBs). They are obtained by the refinement of gasoline and then are treated in sulfuric acid refineries. This kind of oil is known as a petroleum-based product, basically made out of hydrogen and carbon atoms. Hydrocarbon molecules can be broadly classified into three classes, as illustrated in Fig. 1. From a chemical perspective, conventional mineral oil is a blend of naphthenic, paraffin, and aromatic hydrocarbons with diverse bonds. Figure 1 shows a cyclic structure for naphthenic hydrocarbons, a linear branching structure for paraffinic hydrocarbons, and a cyclic structure with alternate single and double carbon–carbon bonds for aromatic hydrocarbons. The oil’s aromatic content is maintained low throughout processing to increase oxidation dependability [13]. Fig. 1 Compounds of hydrocarbons in mineral oil

(a) Naphthenic

(b) Paraffin

(c) Aromatic

220

S. N. Deepa et al.

Mineral oil has high dielectric strength, high thermal stability, and low dielectric loss, which are requirements for fine insulating oil [14]. Between the 1870s and the 1990s, transformers employed liquid dielectrics made of petroleum-based oils [15]. They offered good insulation and a great medium for transferring heat produced by electrical losses. The drawback is that because paraffinic-based oils include a lot of paraffin wax, they often have high pour points. Additionally, these oils’ viscosity is increased since the oxidation-related sludge build-up is insoluble in them. Later, transformer oil based on naphthenic was developed, although oils with a naphthenic base are more susceptible to oxidation than those with a paraffin base, because the oxidized products are soluble and viscosity is reduced [16]. Oils with naphthenic bases have a lower pour point and will stay liquid in the transformer system at low temperatures. However, the flammability of naphthenic-based transformer oil is a factor to consider. Because of their low viscosity, widespread availability, and low cost as compared to other liquids, mineral oils have been utilized as an insulating and cooling liquid in power system equipment for more than 100 years [17]. Therefore, due to its costeffective solution, chemical and physical characteristics, it is still the transformer choice of material in the industry. Mineral oil becomes dissolved with gases when arcing or discharge faults happen in transformers used with mineral oils. Mineral oils’ ability to insulate is also significantly affected by the presence of moisture. Mineral oil also exhibits safety and environmental risks due to its low fire point, high toxicity and environmental harm, poor biodegradability, and highest cost in terms of carbon emissions, in addition to the damaging effects of moisture. With the growth of ultrahigh voltage electric power transmission systems, it is getting harder for standard mineral oil’s insulating characteristics to match the need of ultrahigh voltage transformers. Over the past 10 years, several studies have been conducted to increase the insulating system’s dielectric strength and antiaging capabilities.

2.2 Silicone Oil Since its introduction in the 1970s, silicone transformer liquid (oil) has gained widespread acceptance for use in transformers when a fire risk exists that necessitates an environmentally friendly alternative to traditional transformer oils. These liquids are also used in transformers built to function at temperatures exceeding rise transformers, which run between 55 and 65 °C. For almost 20 years [18], silicone transformer liquid was in use. Silicone fluids were developed subsequently as a replacement for chlorine-based lubricants due to their superior electrical insulating, antioxidative, higher fire point, reduced flammability, and great thermal stability [15]. Figure 2 illustrates the monomer structure of silicone oil, where the number of monomer units is indicated by m. A polymer chain can be made up of anywhere from 1001 to more than 1000 monomer units. A well-chosen m value is necessary for silicone oil to have low viscosity and strong fire resistance. Long chain molecules are required for higher fire resistance whereas low m values are required for reduced

Review on the Addition of Antioxidants and Nanoparticles to Natural …

221

Fig. 2 Structure of silicone oil

viscosity and therefore better cooling. In order to meet both the thermal characteristics and the flammability qualities, an ideal value of m is chosen [19] Silicone oil has the advantages of good thermal conductivity, high flash and fire points, reduced rate of heat discharge, smoke evolution, and toxicity, which implies minor damage from fire as it extinguishes itself, lower thermal evolution during a fire, less volatile at higher temperatures, and much less viscosity variability with temperature. Due to its high flash point, silicone oil is used in areas where safety is a top priority despite its significant expense [1, 20]. In comparison to mineral oil, silicone oil absorbs moisture more readily, which may have a negative effect on its electrical characteristics. The need for alternative products has grown as a result of silicone oil’s rising price and non-biodegradable nature.

2.3 Synthetic Esters Ester liquids, which include components that have settled naturally or artificially, can be used as an alternative to mineral oils. Esters may generally be divided into five categories: • • • • •

Monoesters. Dicarboxylic esters (Diesters). Glycerine esters. Polyol esters. Complex esters.

Esters of various types must be synthesized from an alcohol and an acid. For the high strain conditions required for use in a high-voltage transformer, only polyoland complex esters are acceptable [21]. Compared to vegetable oils, synthetic esters offer a remarkable combination of eco-friendliness and exceptional heat stability. The esters also behave well at low temperatures [22]. As seen in Fig. 3, the esterification of fatty acids and polyvalent alcohols produces synthetic esters, an example of an organic substance [23]. The core polyol backbone of the synthetic ester is connected to four saturated fatty acid groups by a molecular assembly. Synthetic ester has exceptional oxidation stability thanks to the saturated fatty acids groups in its molecular structure [24]. In 1976, the first transformer unit with a synthetic ester went into operation. The

222

S. N. Deepa et al.

Fig. 3 Esterification process, in which the hydrocarbon chains R and R' are saturated

research on synthetic ester liquids grew in depth throughout the ensuing years. The synthetic esters of today are mostly making a name for themselves in distribution and specialty transformers with somewhat modest powers and voltages. This situation results from the synthetic esters’ good environmental qualities (biodegradability and high fire point), which have been allowed to simply see the protective environmental requirements for certain projects like high-speed railroads, wind farms, and retail centres [24–26]. Despite this, the synthetic ester-based transformer promises to have a continuous increase in rated power and nominal voltage. Currently, transformer units’ maximum nominal voltage has reached 400 kV while using synthetic esters. The criteria pertaining to the requirements for renewed synthetic esters have also been created [27] with the advancement of synthetic ester-based claims. The IEC 61099 Standard, “Insulating Liquids—Specifications for Unused Synthetic Organic Esters for Electrical Purposes,” has seen several revisions, and its current version dates back to 2011. In contrast, the ASTM (American Society for Testing and Materials) is now painstakingly working on a paper titled “New standard specification for less-flammable synthetic ester liquids used in electrical equipment” [28].

2.4 Natural Esters Natural esters are commonly available sources of food, making them ideal starting points for completely biodegradable insulating solutions. They mostly consist of triglycerides, which are created naturally when three fatty acids are esterified with tri-alcohol glycerol. Figure 4 [29] depicts fatty acid ester triglycerides in detail. Early in the 1990s, services grew interested in fully biodegradable insulating liquids, primarily for use in transformers located near waterways where oil spills might contaminate the water [30]. Although historically rapeseed oil has been submitted for use in capacitor applications, seed oil esters have been painstakingly inappropriate for use in transformers. Their oxidation susceptibility has been a major hindrance to their use as a dielectric fluid. Although contemporary transformer design follows, coupled with proper fluid additions and other design tweaks, compensation for this unique [31]. Breakdown Voltage (BDV) is significantly reduced when moisture is present, whereas natural esters may store more moisture without having their BDV affected

Review on the Addition of Antioxidants and Nanoparticles to Natural …

223

Fig. 4 Triglyceride ester molecule

Table 1 Composition of fatty acids in natural esters Sl. no.

Insulating natural ester

Saturated fatty acids (%)

Monounsaturated fatty acids (%)

Poly unsaturated References fatty acids (%)

1

Sunflower oil

10.5

19.6

65.7

[2]

2

Rice bran oil

25

38

37

[36]

3

Neem oil

40

20.9

39.1

[37]

4

Castor oil

1.1

97.8

1.1

[37]

5

Mahua oil

32.7

46.3

19.6

[38]

6

Coconut oil

90

6

4

[39]

[32]. Vegetable oil extracted from seeds is a dark tint. They are made up of substances that are solid, such as proteins, fibers, and liquid fats and oils. Alkaline refining, bleaching, and deodorization are the final three processes that the crude vegetable oil goes through before becoming refined, bleached, and deodorized (RBD) oil. An ester-founded insulating fluid is created from this RBD oil. To eliminate the free fatty acids in vegetable oil, alkaline refining, discovered via neutralization, is carried out. Bleaching is done using a clay filter press to remove coloring agents. The deodorization process is then carried out using a high temperature, high vacuum method [33, 34]. Different triple unsaturated fatty acids have low viscosity but are more prone to oxidation than saturated fatty acids, which have a high viscosity but are chemically unchanging. For insulating reasons, oils with a high percentage of single unsaturated fatty acids are suitable. Table 1 lists the fatty acid makeup of natural esters, and Fig. 5 displays several typical fatty acids [35–37].

3 AntiOxidants The word “antioxidant” is used in a broad sense to refer to any chemical substance that prevents the release of oxygen or ozone. Antioxidants are substances that, like vegetable oils, obstruct the oxidation process by specifically reacting with the fat radical to generate a stable radical that does not quickly react with oxygen [38– 40]. Antioxidants often provide a self-protective mechanism in the compounds they

224

S. N. Deepa et al.

Fig. 5 Chemical structure of some common fatty acids

are utilized in. It possesses crucial traits like electron delay and electron trap (free radicals) that are classified as free electron scavengers [41]. Primary antioxidants, secondary antioxidants, metal deactivators, oxygen scavengers, reducing agents, and synergists are the five categories into which antioxidants often get categorized. The major antioxidant may also be divided into natural and synthetic antioxidants, which include butylated hydroxy toluene (BHT), butylated hydroxytoluene, phenolic acid, flavonoids, and carotenoids (BHT). Due to their great efficacy, low cost, and wide availability, synthetic antioxidants are frequently employed as food additives to prevent rancidification. Since then, synthetic antioxidants have been added to edible vegetable oils, including butylated hydroxytoluene (BHT), tertiary butyl hydroquinone (TBHQ), 2, 4, and 5-trihydroxybutyrophenone (THBP), octyl gallate (OG), nordihydroguaiaretic acid (NDGA), propyl gallate (PG), and butylated hydroxy anisole (BHA). Secondary antioxidants, also known as class 11 or defence antioxidants, slow down the oxidation reaction. It does not convert prooxidant or catalyst metal ions’ free radicals into stable radical chelators. Citric acid, ascorbic acid, and carotenoids are only a few of the several substances employed as secondary antioxidants. Chelators, also known as chelating agents, are used to deactivate metal. The main goals of metal chelators are to increase induction time and decrease maximal oxidation rate. As a reducing agent and oxygen scavenger, ascorbic acid. When exposed to light, carotenoids produce photoactivated sensitizers (lipid oxidation), which cause them to get excited and release energy into the surrounding region. Synergists do not really act as antioxidants; rather, they help main antioxidants work more effectively [42, 43]. Antioxidants react with fatty acid alkyl, alkoxy, and peroxyl radicals to produce fatty acid RH, ROH, and ROOH as well as an antioxidant radical (A). The response Eqs. 1–3 demonstrate this. R + AH → RH + A

(1)

RO + AH → ROH + A

(2)

Review on the Addition of Antioxidants and Nanoparticles to Natural …

225

ROO + AH → ROOH + A

(3)

A non-radical product like RA, ROA, and ROOA is produced when the antioxidant radical reacts with the alkyl, alkoxy, and peroxyl radicals. The following response Eqs. 4–6 make this clear. The antioxidants grow into oxidized antioxidants (given in Eq. 7) in the process of scavenging fatty acid-free radicals [44, 45]. R + A → RA

(4)

RO + A → ROA

(5)

ROO + A → ROOA

(6)

Antioxidant + O2 → Oxidized Antioxidant

(7)

The various antioxidant forms are described in Table 2. When temperatures are normal, antioxidant concentrations are high, and one of the major concerns with employing antioxidants at high temperatures is that their percentage decreases from what it previously was [46, 47]. Table 2 Types of antioxidants Synthetic antioxidant

Natural antioxidant

Synergists

Butylated hydroxy toluene

VITAMIN-E (TOCOPHEROL) Alpha (α)

Ascorbic acid

Butylated hydroxy anisole

VITAMIN-E (TOCOPHEROL) Beta (β)

Carsonic acid

Tert butyl hydro quinone

VITAMIN-E (TOCOPHEROL) Gamma (γ)

Citric acid

Propyl gallate

VITAMIN-E (TOCOPHEROL) Delta (δ)

Phosphoric acid

Pyrogallol

VITAMIN-E (TOCOTRIENO) Alpha (α)

Ethylenediamene Tetra-acetic acid

Lauryl tert butyl hydro quinone

VITAMIN-E (TOCOTRIENO) Beta (β)

Carotenes

2,4,5-trihydroxy butyrophenone

VITAMIN-E (TOCOTRIENOL) Gamma (γ)

Oryzanol

–

VITAMIN-E (TOCOTRIENOL) Delta (δ)

Rosemary extracts

226

S. N. Deepa et al.

4 Nanoliquid Dielectrics T. J. Lewis proposed the concept of micro dielectrics in 1994 with the intention of developing insulating materials capable of withstanding high voltage [48–50]. A multi-core model of the dispersed particle in a nanocomposite was put out by Tanaka et al. in 2005. According to it, the particle has three distinct layers, the innermost of which is connected to a nanoparticle, the next contains tightly bonded molecules, and the outermost has loosely attached particles [51, 52]. The emphasis shifted to liquid dielectrics and nanofluids as strong nanocomposites with superior insulating qualities became more prevalent [53, 54]. In 1995, Choi introduced the idea of a nanoliquid at the Argonne National Laboratory in the USA [55]. The term “nanofluid” or “nanoliquids” refers to a nano-size particle that is consistently disseminated at a few weight percentages (wt.%) into a base liquid [56]. The enlarged surface space of the suspended metallic nanoparticle per unit volume of the base fluid indicates that it is a liquid with better characteristics [57]. As the amount of nanoparticles rises, the fluid’s conductivity rises [58]. The nanoparticles are typically grouped into three types: semi-conductive, conductive, and magnetic nanoparticles. Under the influence of electrical pressure, the nanoparticles exhibit an electron scavenging mechanism that converts quick electrons into moderately negative charged particles in insulating liquid [59–61]. The idea to use nanoparticles in high-voltage applications was sparked by studies on the development of liquids used to cool various types of equipment, such as electronic components, cooling systems for vehicle engines and heat-generating components, nuclear cooling frameworks, and aircrafts. This interest in nanoliquids is related to the significant volume of interfaces present in most fluids and the subsequent interactions between the charged surfaces of nanoparticles and the molecular structure of supporting materials, which have the potential to enhance the dielectric properties of nanoliquids. The volume division and characteristics of the extra nanoparticles, as well as their shape and size, would determine the amount of advancement of nanofluids with regard to cooling applications [62, 63]. These days, the terms “nanoliquid” and “dielectric liquid” are respected and prominent, and they have been the focus of much research over the past few years. A fluid containing nanoparticles, or manometer-sized particles, is referred to as a nanofluid. a base liquid. These Nanofluids are colloidal nanoparticle solutions that have been manufactured. Nanoparticles are employed in nanofluids, and they are frequently formed of metals, oxides, carbides, or carbon nanotubes. Dielectrics used as base liquids frequently include transformer oil, ethylene glycol, and water. For liquid dielectric, the similar technique was adopted to improve its insulating qualities [64]. The use of nanoparticles in the insulating oil in transformers can result in an improvement in the insulating qualities of the insulation, which can lead to an increase in power capacity, a reduction in the size and insulation of the transformer, and ultimately a reduction in cost [53]. Because oils combined with nanoparticles have a high shallow trap density of thermally induced current, it has been shown that expanding nanoparticles is the optimum method for increasing the strength of oils [65]. On the surface of nanoparticles, polarized charges or surface charges create

Review on the Addition of Antioxidants and Nanoparticles to Natural …

227

potential wells where free electrons are tapped and converted into sluggish, negatively charged particles. Thus, this affects how quickly streamers spread. The nanoparticle radii determine the well’s maximum practicable depth. Therefore, without considering stability, a large radius nanoparticle should be chosen, with a typical size of 10 nm [55, 66]. A feature that was at the core of the development of nanofluids, nanoparticles greatly enhance heat transfer and the cooling performance of transformers [67]. The characteristics of nanofluids are further influenced by the volume component of the particles, their shape and size, and the surface area where the particles and fluid come into contact [68].

4.1 Synthesis of Nanoparticles To increase the qualities and lower the cost of manufacturing, several synthesis techniques are either being developed or improved. To boost the optical, mechanical, physical, electrical, and chemical characteristics of NPs, a few techniques are modified [69]. The many techniques used to create nanoparticles can be categorized as bottom-up or top-down techniques. Figure 6 is a depiction of a process. The term “bottom-up” or “constructive technique” describes a strategy in which a substance is constructed from the bottom up, atom by atom, molecule by molecule, or cluster by cluster. This course is mostly utilized for creating materials at the nanoscale that are consistent in size, shape, and distribution. It certainly regulated the process to stop additional particle formation and successfully covers chemical synthesis. Despite being nothing new, the bottom-up strategy is crucial for creating and processing nanostructures and nanomaterials. Depletion of a bulk substance to nanometric size particles is referred to as a topdown approach. This course starts with bulk material and reduces it, thereby utilizing physical processes like crushing, milling, or grinding to break up bigger particles. Notably, this technique is not suitable for producing materials with consistent shapes, and producing extremely small particles is difficult and requires a lot of energy.

Fig. 6 Scheme of top-down and bottom-up synthesis of nanoparticles

228

S. N. Deepa et al.

The flaw in nanostructures and nanomaterials is the main issue with this method. It is noteworthy that the top-down technique procedure might seriously impair the processed pattern’s crystal structure.

4.2 Synthesis of Nanoliquids Preparation of nanofluids is having two primary methods named as one-step preparation process and two-step preparation process [66]. • One-step preparation process The cycles of drying, storing, and transporting nanoparticles are avoided in this preparation procedure, which limits agglomeration and enhances the stability of the nanoparticle dispersion in the oil [70]. The main drawback of a one-step preparation process is that it is difficult to explain the impact of nanoparticles without removing the impurity impact, which may limit its applicability. Other drawbacks include the fact that the solution contains residual reactants from an imperfect reaction or stabilization. The requirement to use low vapor pressure liquids and its restricted manufacturing are two factors that contribute to various problems with this preparation method [71]. Intensive magnetic force agitation, ultrasonic agitation, high-shear mixing, homogenizing, and ball milling are used in this preparation method to distribute the nano-sized nanopractical powder into a base fluid in the processing stage [70, 72]. Due to their inexpensive cost and compatibility with a large variety of nanoparticles, these preparation methods are more widely used than the one-step method. The disadvantage of this production method is the nanoparticles’ inevitable clumping and aggregation as a result of their enlarged surface area and massive surface mobility [73]. Therefore, the nanoparticles have a tendency to collect residue and block the container’s bottom. Figures 7 and 8 illustrates the standard two-step preparation procedure. Fig. 7 Two-step preparation process

Review on the Addition of Antioxidants and Nanoparticles to Natural …

229

Fig. 8 Preparation process of nanoliquid by two-step method

5 Effect of Antioxidants and Nanoparticles on Natural Esters Table 3 describes the work done so far on the effect of adding antioxidants and nanoparticles to the sunflower oil which is a natural ester that can be used as an alternative to mineral oil in transformer applications. The effects of adding these additives to sunflower oil on critical insulating properties of transformer are focused. References [74, 75] describe the critical insulating properties of sunflower oil without any additives. Reference [5] is a work done describing the effect of adding different antioxidants to the sunflower oil on the insulating properties of the oil. Antioxidants are added at different concentrations to the sunflower oil and each concentration of antioxidant had a different effect on the insulating properties of sunflower oil. Among all the antioxidants studied in this work Propyl Gallate (C10 H12 O5 ) and αT (α-Tocopherol) + CA (C31 H52 O3 ), when added to the sunflower oil showed the prominent improvement in breakdown voltage (BDV) at Rtp. Addition of antioxidants to the sunflower oil had made the property of viscosity to decrease and the increase in the value of viscosity compared to the base sunflower oil was observed, and all the antioxidants added had almost the same values of viscosity with very less difference between them. The bulbous improvement of flash and fire points was observed by the addition of 5 g of L-Ascorbic acid (AA) (C6 H8 O6 ), 1 g of Butylated Hydroxy Tolciene (C15 H24 O2 ), and αT (α-Tocopherol) + CA (C31 H52 O3 ) to the base sunflower oil. References [76–78] are the works done on the addition of nanoparticles to the base sunflower oil to serve as an alternative to existing transformer oil. Among all the nanoparticles studied in these works Fe3 O4 and ZnO added to the base oil showed a significant improvement in BDV. All the nanoparticles added to base oil had a prominent decrement in viscosity values compared to base oil. Fe3 O4 and ZnO had the significant contribution in the viscosity decrement. Almost all the nanoparticles showed the betterment in flash and fire points than that of base sunflower oil and in the same range as of the antioxidants. Addition of SiO2 nanoparticle to the base sunflower oil resulted in very high increase of both flash and fire points. From Figs. 9, 10, 11, and 12, it can be observed that addition of nanoparticles to the sunflower oil provides the improvement in critical properties of sunflower oil to be used as insulating fluid in transformer applications. Best values for BDV and

–

–

L-Ascorbic acid (AA) (C6 H8 O6 )

Butylated hydroxy anisole (C11 H16 O2 )

Butylated hydroxy tolciene (C15 H24 O2 )

Propyl Gallate (C10 H12 O5 )

Sunflower oil

Sunflower oil

Sunflower oil

Sunflower oil

Sunflower oil

–

–

–

–

Nanoparticles

Antioxidants

Base oil

500 ml SFO + 5 g PG

500 ml SFO + 1 g PG

500 ml SFO + 5 g BHT

500 ml SFO + 1 g BHT

500 ml SFO + 5 g BHA

500 ml SFO + 1 g BHA

500 ml SFO + 5 g AA

IEC 60156 ASTM D445 ASTM D93 ASTM D93

IEC 60156 ASTM D445 ASTMD93 ASTMD93

IEC 60156 ASTM D445 ASTM D93 ASTM D93

@RTP–54 @70°C–60

@RTP–51 @70°C–60

@RTP–44 @70°C–47

@RTP–34 @70°C–52

@RTP–42 @70°C–52

@RTP–33 @70°C–48

@RTP–20 @70°C–40

@RTP–42 @70°C–45

IEC 60156 ASTM D445 ASTM D93 ASTM D93

500 ml SFO + 1 g AA

BDV (KV)

ASTM D1816 38–45 ASTM D445 ASTM D92 ASTM D92

Standards

–

Concentration of additives

Table 3 Addition of AntiOxidants and NPs on SFO

@30°C–119

@30°C–112

@30°C–127

@30°C–109

@30°C–125

@30°C–128

@30°C–127

@30°C–121

@20°C–78 @40°C–41.8–45 @100°C–8

Viscosity (Cst)

270

275

270

280

280

275

280

270

260

Flash point (°C)

280

285

285

300

290

280

320

295

270

Fire point (°C)

(continued)

[5]

[5]

[5]

[5]

[5]

[5]

[5]

[5]

[74] [75]

References

230 S. N. Deepa et al.

αT + CA α-Tocopherol (C31 H52 O3 )

–

Sunflower oil

Sunflower oil

SiO2

–

BHT + citricacid – (C6 H8 O7 )

Sunflower oil

Nanoparticles

Antioxidants

Base oil

Table 3 (continued)

IEC 60156 ASTM D445 ASTM D93 ASTM D93

Standards

1 mm–26 2.5 mm–54 4 mm–68 5 mm–76

0.1 wt.%

1 mm–22 2.5 mm–50 4 mm–64 5 mm–72

@30°C–43 @40°C–30 @50°C–23 @60°C–18

@30°C–41 @40°C–29.5 @50°C–22.5 @60°C–17

@30°C–40.5 @40°C–29 @50°C–22 @60°C–16

@30°C–115

@RTP–49 @70°C–60

1 mm–24 2.5 mm–52 4 mm–66 5 mm–74

IEC60156 IEC60156 IEC60156 IEC60156

@30°C–110

@30°C–119

@RTP–45 @70°C–48

@RTP–54 @70°C–60

@30°C–109

Viscosity (Cst)

@RTP–38 @70°C–44

BDV (KV)

0.05 wt.%

0.01 wt.%

IEC 60156 ASTM D445 ASTM D93 500 ml SFO + 5 g ASTM D93 [αT + CA (1:1 ratio)]

500 ml SFO + 1 g [ αT + CA (0.5:0.5 ratio)]

500 ml SFO + 5 g [BHT + CA (1:1ratio)]

500 ml SFO + 1 g [BHT + CA (0.5:0.5 ratio)]

Concentration of additives

320

326

328

290

280

270

260

Flash point (°C)

355

352

350

300

295

295

300

Fire point (°C)

(continued)

[76]

[76]

[76]

[5]

[5]

[5]

[5]

References

Review on the Addition of Antioxidants and Nanoparticles to Natural … 231

Fe3 O4

–

–

–

–

Sunflower oil

Sunflower oil

Sunflower oil

Sunflower oil

TiO2

Al2 O3

ZnO

Nanoparticles

Antioxidants

Base oil

Table 3 (continued)

0.5 g

0.25 g

0.05 g

0.5 g

0.25 g

IEC 60156 ASTM D445 ASTM D93 ASTM D93

@RTP–27 @90°C–62

@RTP–27 @90°C–44

@RTP–27 @90°C–24

@RTP–32 @90°C–46

@RTP–20 @90°C–34

@RTP–16 @90°C–32

2.5 mm–68 IEC 60156 ASTM D445 ASTM D93 ASTM D93

0.25% 0.05 g

2.5 mm–65

2.5 mm–63

2.5 mm–62.5

2.5 mm–59

0.2%

0.15%

0.1%

IEC 60156 ASTM D445 ASTM D93 ASTM D93

2.5 mm–71

0.05%

0.25%

2.5 mm–66

2.5 mm–65

2.5 mm–62

BDV (KV)

2.5 mm–67

IEC 60156 ASTM D445 ASTM D93 ASTM D93

Standards

0.2%

0.15%

0.1%

0.05%

Concentration of additives

@RTP–46.33 @90°C–21.20

@RTP–45.85 @90°C–21.20

@RTP–45.29 @90°C–21.23

@RTP–45.95 @90°C–21.88

@RTP–38.33 @90°C–22.88

@RTP–45 @90°C–23.23

29

28

27

27

25

28

26.6

26

25.5

24

Viscosity (Cst)

263

260

263

289

295

295

285

282

278

273

272.5

286

283

281

278

273

Flash point (°C)

283

275

276

312

305

310

298

295

292.5

292

285

303

298

294

292

286

Fire point (°C)

(continued)

[78]

[78]

[77]

[77]

References

232 S. N. Deepa et al.

Antioxidants

–

–

Base oil

Sunflower oil

Sunflower oil

Table 3 (continued)

Fe2 O3

Cds

Nanoparticles

0.5 g

0.25 g

0.05 g

0.5 g

0.25 g

0.05 g

Concentration of additives

IEC 60156 ASTM D445 ASTM D93 ASTM D93

IEC 60156 ASTM D445 ASTM D93 ASTM D93

Standards

@RTP–27 @90°C–48

@RTP–27 @90°C–45

@RTP–26 @90°C–42

@RTP–28 @90°C–52

@RTP–28 @90°C–46

@RTP–28 @90°C–46

BDV (KV)

@RTP–48.45 @90°C–25

@RTP–47.3 @90°C–24

@RTP–46.3 @90°C–22.24

@RTP–42.13 @90°C–27.40

@RTP–42.15 @90°C–27.42

@RTP–42.12 @90°C–21.58

Viscosity (Cst)

260

258

258

263

265

262

Flash point (°C)

276

276

269

285

283

276

Fire point (°C)

[78]

[78]

References

Review on the Addition of Antioxidants and Nanoparticles to Natural … 233

234

S. N. Deepa et al.

Fig. 9 BDV comparison for SFO with and without additives

viscosity of SFO observed with the addition of nanoparticles, flash and fire points were improved by adding either antioxidants or nanoparticle and the improvement is similar in both cases.

Fig. 10 Viscosity comparison for SFO with and without additives

Fig. 11 Flash point comparison for SFO with and without additives

Review on the Addition of Antioxidants and Nanoparticles to Natural …

235

Fig. 12 Fire point comparison for SFO with and without additives

Table 4 describes the work done so far on the effect of adding antioxidants and nanoparticles to the rice bran oil which is a natural ester that can be used as a substitute to mineral oil in transformer applications. The effects of adding these additives to rice bran oil on critical insulating properties of transformer are focused. Reference [36] describes the critical insulating properties of sunflower oil without any additives. References [5, 79] describe the effect of adding different antioxidants to the rice bran oil on the insulating properties of the oil. Antioxidants are added at different concentrations to the Rice bran oil and each concentration of antioxidant had a different effect on the insulating properties of Rice bran oil. Among all the antioxidants studied in this work Butylated Hydroxy Anisole (C11 H16 O2 ), L-Ascorbic acid (C6 H8 O6 ), and Butylated Hydroxy Tolciene (C15 H24 O2 ), when added to the rice bran oil showed the prominent improvement in break down voltage (BDV) at Rtp. The addition of antioxidants to the rice bran oil had made the property of viscosity to decrease and the increase in the value of viscosity compared to the base rice bran oil was observed. The bulbous improvement of flash and fire points was observed by the addition of 5 g of Gallic acid (C7 H6 O5 ), 1 g of L-Ascorbic acid (C6 H8 O6 ), and 5 g of L-Ascorbic acid (C6 H8 O6 ) to the base rice bran oil. [80] describes the work done on the addition of nanoparticles to the base rice bran oil to serve as an alternative to existing transformer oil. The nanoparticles studied in these works ZnO + Al2 O3 added to the base oil showed a significant improvement in BDV. The nanoparticles added to base oil had a prominent decrement in viscosity values compared to the base oil. ZnO + Al2 O3 had a significant contribution to the viscosity decrement. From Figs. 13, 14, 15, and 16 it can be observed that the addition of antioxidants to the rice bran oil provides the improvement in critical properties of rice bran oil to implement in transformers as a liquid insulating medium. Best values for BDV, flash and fire point of rice bran oil are observed with the addition of antioxidant, best value for viscosity of rice bran oil is observed with the addition of nanoparticles. Table 5 describes the work done so far on the effect of adding antioxidants and nanoparticles to the neem oil which is a natural ester that can be used as an alternative to mineral oil in transformer applications. The effects of adding these additives to neem oil on critical insulating properties of transformer are focused. Reference

–

–

Gallic acid (C7 H6 O5 )

L-Ascorbic acid (AA) (C6 H8 O6 )

Butylated hydroxy anisole (C11 H16 O2 )

Butylated hydroxy tolciene (C15 H24 O2 )

Rice bran oil

Rice bran oil

Rice bran oil

Rice bran oil

Rice bran oil

–

–

–

–

Nanoparticle

Antioxidants

Base oil

500 ml RBO + IEC 60156 1 g BHT ASTM D445 500 ml RBO + ASTMD93 ASTMD93 5 g BHT

500 ml RBO + IEC 60156 1 g BHA ASTM D445 500 ml RBO + ASTM D93 ASTM D93 5 g BHA

500 ml RBO + IEC 60156 1 g AA ASTM D445 500 ml RBO + ASTM D93 ASTM D93 5 g AA

@30°C–136 @30°C–140

@RTP–46 @70°C–50

@30°C–143

@RTP–55 @70°C–40 @RTP–44 @70°C–40

@30°C–143

@30°C–147

@RTP–40 @70°C–46 @RTP–44 @70°C–39

@30°C–145

@RTP–46 @70°C–60

@RTP–119 @60°C–30 @90°C–18

41.3

5 g GA

@RTP–122 @60°C–35 @90°C–20

36.1

@RTP–119 @40°C–84 @60°C–39 @80°C–30

Viscosity (Cst)

500 ml RBO + IEC 60156 1 g GA ASTM D445 ASTM D93 500 ml RBO + ASTM D93

BDV (KV) 39.8

Standards IEC 60156 ASTM D445 ASTM D93 ASTM D93

–

Concentration of additives

Table 4 Addition of AntiOxidants and NPs on RBO

250

265

265

270

287

290

285

260

274

Flash point (°C)

260

275

275

280

295

300

310

280

282

Fire point (°C)

(continued)

[5]

[5]

[5]

[79]

[36]

References

236 S. N. Deepa et al.

Antioxidants

Propyl gallate (C10 H12 O5 )

–

Base oil

Rice bran oil

Rice bran oil

Table 4 (continued)

500 ml RBO + 0.02% ZnO + 0.01% Al2O3

0.015% ZnO + 0.015% Al2O3

500 ml RBO + IEC 60156 0.01% ZnO + ASTM D445 0.02% Al2O3 ASTM D93 500 ml RBO + ASTM D93

ZnO + Al2 O3

Standards

500 ml RBO + IEC 60156 1 g PG ASTM D445 500 ml RBO + ASTM D93 ASTM D93 5 g PG

Concentration of additives

–

Nanoparticle

@RTP–19.23 @80°C–5.22

@RTP–19.25 @80°C–5.24

48.2

45.4

@RTP–19.24 @80°C–5.23

@30°C–154

@RTP–25 @70°C–35 47.6

@30°C–150

Viscosity (Cst)

@RTP–32 @70°C–41

BDV (KV)

238

242

241

275

270

Flash point (°C)

266

268

265

280

290

Fire point (°C)

[80]

[5]

References

Review on the Addition of Antioxidants and Nanoparticles to Natural … 237

238

Fig. 13 BDV comparison for RBO with and without additives

Fig. 14 Viscosity comparison for RBO with and without additives

Fig. 15 Flashpoint comparison for RBO with and without additives

S. N. Deepa et al.

Review on the Addition of Antioxidants and Nanoparticles to Natural …

239

Fig. 16 Fire point comparison for RBO with and without additives

[81] describes the critical insulating properties of neem oil without any additives. References [37, 82–84] are the works done describing the effect of adding different antioxidants to the neem oil on the insulating properties of oil. Antioxidants are added at different concentrations to the neem oil and each concentration of antioxidant had a different effect on the insulating properties of neem oil. Among all the antioxidants studied in this work Butylated Hydroxy Tolciene (C15 H24 O2 ), Butylated Hydroxy Anisole (C11 H16 O2 ), and Gallic Acid (C7 H6 O5 ), when added to the neem oil showed the prominent improvement in break down voltage (BDV). The addition of antioxidants to the neem oil had made the property of viscosity to increase and the decrement in the value of viscosity compared to the base sunflower oil was observed, and all the antioxidants added had almost decreased values of viscosity. The bulbous improvement of flash and fire points was observed by the addition of 0.75 of Butylated Hydroxy Tolciene (C15 H24 O2 ), 0.75 of Butylated Hydroxy Anisole (C11 H16 O2 ), and 0.5 of Gallic Acid (C7 H6 O5 ) to the base neem oil. References [5, 85] are the works done on the addition of nanoparticles to the base neem oil to serve as an alternative to existing transformer oil. Among all the nanoparticles studied in these works Magnetite (Fe3 O4 ) Iron Oxide added to the base oil showed a great improvement in BDV. All the nanoparticles added to base neem oil had an excellent decrement in viscosity values compared to base neem oil. Magnetite (Fe3 O4 ) Iron Oxide, Boron nitride (BN), and Alumina (Al2 O3 ) had the significant contribution in the viscosity decrement. All the nanoparticles showed the great increments in flash and fire points than that of base neem oil. The addition of Magnetite (Fe3 O4 ) Iron Oxide nanoparticle to the base neem oil resulted in a very high increase of both flash and fire points. From Figs. 17, 18, 19, and 20 it can be observed that the addition of antioxidant to the neem oil provides the improvement in critical properties of sunflower oil to implement in transformers as a liquid insulating medium. Best values for BDV and flash

–

–

Butylated hydroxy tolciene (C15 H24 O2 )

Butylated hydroxy anisole (C11 H16 O2 )

Gallic acid (C7 H6 O5 )

Neem oil

Neem oil

Neem oil

Neem oil

–

–

–

Nanoparticle

Antioxidants

Base oil

44

500 ml NO + 0.25 GA

48

38

500 ml NO + 1 BHT

500 ml NO + 0.5 GA

47

500 ml NO + 0.75 BHT

IEC 60156 ASTM D445 ASTM D93 ASTM D93

43

500 ml NO + 0.25 BHT 45

45

500 ml NO + 1 BHT

500 ml NO + 0.5 BHT

49.5

46

43

28.6

BDV (KV)

500 ml NO + 0.75 BHT

IEC 60156 ASTM D445 ASTM D93 ASTM D93

IEC 60156 ASTM D445 ASTM D93 ASTM D93

500 ml NO + 0.25 BHT 500 ml NO + 0.5 BHT

IEC 60156 ASTMD445 ASTM D93 ASTMD93

Standards

–

Concentration of additives

Table 5 Addition of AntiOxidants and NPs on NO

110.88

109.65

124.44

103.57

93.12

85.28

127.57

129.66

147.90

145.81

176.54

Viscosity (Cst)

312

300

280

305

298

288

298

308

306

299

156

Flash point (°C)

325

316

298

325

315

320

318

338

320

318

172

Fire point (°C)

(continued)

[37]

[37]

[37]

[81]

References

240 S. N. Deepa et al.

–

–

TBHQ (tert-butyl hydroquinone) (C10 H14 O2 )

Beta carotene β-carotene (C40 H56 )

Selenium (Se) H2 Se

–

Neem oil

Neem oil

Neem oil

Neem oil

Alumina (Al2 O3 )

–

Nanoparticle

Antioxidants

Base oil

Table 5 (continued)

22.1

IEC 60156 ASTM D445 ASTMD93 ASTMD93 IEC 60156 ASTM D445 ASTM D93 ASTM D93 IEC 60156 ASTM D445 ASTMD93 ASTMD93 IEC 60156 ASTM D445 ASTMD93 ASTMD93

500 ml NO + 1 g TBHQ

500 ml NO + 1 g β-carotene

500 ml NO + 1 g β-carotene

500 ml NO + 0.025% Al2O3 44

23.8

22.3

36

500 ml NO + 1 GA

BDV (KV) 40

Standards

500 ml NO + 0.75 GA

Concentration of additives

62.9

@40°C–94.95 @90°C–34.10

@40°C–110.1 @90°C–34.90

@40°C–113.2 @90°C–34.10

112.70

110.36

Viscosity (Cst)

310

238

243

260

280

291

Flash point (°C)

305

257

259

283

296

303

Fire point (°C)

(continued)

[85]

[84]

[83]

[82]

References

Review on the Addition of Antioxidants and Nanoparticles to Natural … 241

Boron nitride (BN)

–

–

Neem oil

Neem oil

Magnetite (Fe3 O4 ) Iron oxide

Nanoparticle

Antioxidants

Base oil

Table 5 (continued)

IEC 60156 ASTM D445 ASTM D93 ASTM D93

IEC 60156 ASTM D445 ASTM D93 ASTM D93

500 ml NO + 0.025% BN

–

Standards

Concentration of additives

45

44

BDV (KV)

62.5

62.8

Viscosity (Cst)

314

321

Flash point (°C)

316

308

Fire point (°C)

[5]

[85]

References

242 S. N. Deepa et al.

Review on the Addition of Antioxidants and Nanoparticles to Natural …

243

and fire point of neem oil are observed with the addition of antioxidants, viscosity was decreased by adding either antioxidants or nanoparticle and the decrement is similar in both cases. Table 6 describes the work done so far on the effect of adding antioxidants and nanoparticles to the castor oil which is a natural ester that can be used as an alternative to mineral oil in transformer applications. The effects of adding these additives to castor oil on critical insulating properties of transformer are focused. References [86, 87] describe the critical insulating properties of castor oil without any additives. Reference [37] is a work done describing the effect of adding different antioxidants to the castor oil on the insulating properties of oil. Antioxidants are added at different concentrations to the castor oil and each concentration of antioxidant had a different effect on the insulating properties of sunflower oil. Among all the antioxidants studied in this work Butylated Hydroxy Tolciene (C15 H24 O2 ), Butylated Hydroxy Anisole (C11 H16 O2 ), and Gallic Acid (C7 H6 O5 ), when added to the castor oil showed the prominent improvement in break down voltage (BDV). Addition of antioxidants to

Fig. 17 BDV comparison for NO with and without additives

Fig. 18 Viscosity comparison for NO with and without additives

244

S. N. Deepa et al.

Fig. 19 Flashpoint comparison for NO with and without additives

Fig. 20 Fire point comparison for NO with and without additives

the castor oil had made the property of viscosity to increase in the value of viscosity compared to the base castor oil, and all the antioxidants added had almost the same values of viscosity with very less difference between them. The bulbous improvement of flash and fire points was observed by the addition of 0.75 of Butylated Hydroxy Tolciene (C15 H24 O2 ), 0.75 of Butylated Hydroxy Anisole (C11 H16 O2 ), and 0.25 of Gallic Acid (C7 H6 O5 ) to the base castor oil. References [88, 89] are the works done on the addition of nanoparticles to the base castor oil to serve as an alternative to existing transformer oil. The nanoparticles studied in these works SiO2 added to the base castor oil showed a significant improvement in BDV. The ZnO nanoparticle added to base oil had a huge decrement in viscosity values compared to the base oil. ZnO had the significant contribution in the viscosity decrement. The addition of ZnO nanoparticle to the base castor oil is in the acceptable range of both flash and fire points.

–

–

Butylated hydroxy tolciene (C15 H24 O2 )

Butylated hydroxy anisole (C11 H16 O2 )

Gallic acid (C7 H6 O5 )

Castor oil

Castor oil

Castor oil

Castor oil

–

–

–

Nanoparticle

Antioxidants

Base oil

47

500 ml CO + 0.25 GA

40 37

500 ml CO + 0.75 GA

500 ml CO + 1 GA

51

37

500 ml CO + 1 BHA

500 ml CO + 0.5 GA

48

500 ml CO + 0.75 BHA

IEC 60156 ASTM D445 ASTM D93 ASTM D93

39

500 ml CO + 0.25 BHA 46

42

500 ml CO + 1 BHT

500 ml CO + 0.5 BHA

45

43

41

31

BDV (KV)

500 ml CO + 0.75 BHT

IEC 60156 ASTM D445 ASTM D93 ASTM D93

IEC 60156 ASTM D445 ASTM D93 ASTM D93

500 ml CO + 0.25 BHT

500 ml CO + 0.5 BHT

IEC 60156 ASTM D445 ASTM D93 ASTM D93

Standards

–

Concentration of additives

Table 6 Addition of antioxidants and NPs on CO

212.47

210.91

207.79

210.39

218.20

206.74

202.58

218.20

184.36

205.18

202.58

210.39

155.71

Viscosity (Cst)

287

296

318

304

294

298

292

289

290

298

290

288

280

Flash point (°C)

301

312

332

320

306

312

303

301

301

318

302

298

290

Fire point (°C)

(continued)

[37]

[37]

[37]

[86, 87]

References

Review on the Addition of Antioxidants and Nanoparticles to Natural … 245

Antioxidants

–

–

Base oil

Castor oil

Castor oil

Table 6 (continued)

SiO2

ZnO

Nanoparticle IEC 60156 ASTMD445 ASTMD93 ASTM D93 IEC 60156 ASTM D445 ASTM D93 ASTM D93

500 ml CO + 0.5 wt.% ZnO

500 ml CO + 1.0 wt.% ZnO

500 ml CO + 2.0 wt.% ZnO

500 ml CO + 0.01 wt.% SiO2

500 ml CO + 0.1 wt.% SiO2

500 ml CO + 0.05 wt.% SiO2

Standards

Concentration of additives

68

65

54

–

–

–

BDV (KV)

–

–

–

19.07

18.75

18.51

Viscosity (Cst)

–

–

–

260

260

260

Flash point (°C)

–

–

–

270

270

270

Fire point (°C)

[89]

[88]

References

246 S. N. Deepa et al.

Review on the Addition of Antioxidants and Nanoparticles to Natural …

247

From Figs. 21, 22, 23 and 24 it can be observed that addition of antioxidants and nanoparticles to the castor oil provides the improvement in critical properties of castor oil to implement in transformers as a liquid insulating medium. Best values for BDV and viscosity of castor oil are observed with the addition of nanoparticles, flash and fire points were improved by adding antioxidants. Table 7 describes the work done so far on the effect of adding antioxidants and nanoparticles to the mahua oil which is a natural ester that can be used as a substitute to mineral oil in transformer applications. The effects of adding these additives to mahua oil on critical insulating properties of transformer is focused. Reference [90] describes the critical insulating properties of mahua oil without any additives. Reference [82] is a work done describing the effect of adding different antioxidants to the mahua oil on the insulating properties of oil. Antioxidants are added at different concentrations to the mahua oil and each concentration of antioxidant had a different effect on the insulating properties of

Fig. 21 BDV comparison for CO with and without additives

Fig. 22 Viscosity comparison for CO with and without additives

248

S. N. Deepa et al.

Fig. 23 Flash point comparison for CO with and without additives

Fig. 24 Fire point comparison for CO with and without additives

mahua oil. Among all the antioxidants studied in this work Beta Ceratene, when added to the mahua oil showed a prominent improvement in break down voltage (BDV). Addition of antioxidants to the mahua oil increases the value of viscosity compared to the base mahua oil. Reference [82] are the works done on the addition of nanoparticles to the base mahua oil to serve as an alternative to existing transformer oil. The nanoparticles studied in these works Al2 No3 and Carbon added to the base oil showed an improvement in BDV. The addition of nanoparticle showed the betterment in flash and fire points than that of base mahua oil. Addition of Al2 No3 nanoparticle to the base mahua oil resulted in a very high increase of both flash and fire points. From Figs. 25, 26, 27, and 28 it can be observed that the addition of nanoparticles to the mahua oil provides the improvement in critical properties of mahua oil to implement in transformers as a liquid insulating medium. Best values for BDV, viscosity,

–

–

TBHQ

Selenium

Beta-ceratene

–

–

Mahua oil

Mahua oil

Mahua oil

Mahua oil

Mahua oil

Nanoparticle

Carbon

Al2 No3

–

–

Antioxidants

–

Base oil

Mahua oil

IEC-60156 ASTM-D445 ASTM-D93 ASTM-D93

IEC-60156 ASTM-D445 ASTM-D93 ASTM-D93

IEC-60156 ASTM-D445 ASTM-D93 ASTM-D93

IEC-60156 ASTM-D445 ASTM-D93 ASTM-D93

IEC-60156 ASTM-D445 ASTM-D93 ASTM-D93

IEC-60156 ASTM-D445 ASTM-D93 ASTM-D93

Standards

Table 7 Addition of AntiOxidants and NPs on MO BDV (KV)

27.9

29.4

26.6

23.1

15.4

24.9

Viscosity (Cst)

@RTP–108.2

@RTP–109.3

@RTP–96.87

@RTP–81.88

@RTP–80.83

@40°C–37.19

Flash point (°C)

268

290

258

250

240

270

Fire point (°C)

291

313

273

268

260

290

References

[82]

[82]

[82]

[82]

[82]

[90]

Review on the Addition of Antioxidants and Nanoparticles to Natural … 249

250

S. N. Deepa et al.

Fig. 25 BDV comparison for MO with and without additives

Fig. 26 Viscosity comparison for MO with and without additives

flash and fire point of mahua oil are observed with the addition of nanoparticles. BDV and viscosity were improved by adding antioxidants to mahua oil. Table 8 describes the work done so far on the effect of adding antioxidants and nanoparticle to the coconut oil which is a natural ester that can be used as an alternative to mineral oil in transformer applications. The effects of adding these additives to coconut oil on critical insulating properties of transformer are focused. Reference [91] describes the critical insulating properties of coconut oil without any additives. Reference [92] is a work done describing the effect of adding antioxidants to the coconut oil on the insulating properties of oil. Antioxidants are added at particular concentrations to the coconut oil and the concentration of antioxidant had a particular effect on the insulating properties of Coconut oil. The addition of antioxidants to the coconut oil decreases the value of viscosity compared to the base coconut oil. The bulbous improvement of flash and fire points was observed by the addition of 0.5 Wt.% of TBHQ to the base coconut oil. Reference [93] is the work done on the addition of nanoparticles to the base coconut oil to serve as an alternative to existing transformer oil. The nanoparticles studied in this work. Silica (SiO2 ) added to the base oil showed the same range of BDV as base oil. The nanoparticles added to

Review on the Addition of Antioxidants and Nanoparticles to Natural …

251

Fig. 27 Flash point comparison for MO with and without additives

Fig. 28 Fire point comparison for MO with and without additives

base oil had a prominent increment in viscosity values compared to the base oil. The nanoparticles showed betterment in flash and fire point than base coconut oil. From Figs. 29, 30, 31, and 32 it can be observed that the addition of antioxidant to the coconut oil provides an improvement in the physical properties of coconut oil and the addition of nanoparticles to the coconut oil provided the improvement in the electrical properties that to implement in transformers as a liquid insulating medium. Best values for viscosity and flash and fire point of coconut oil are observed with the addition of antioxidants, flash and fire points were improved by adding either antioxidants or nanoparticles.

–

–

TBHQ

–

Coconut oil

Coconut oil

Coconut oil

Silica (SiO2)

–

Nanoparticle

Antioxidants

Base oil IEC 60156 ASTM D445 ASTM D93 ASTM D93 IEC 60156 ASTM D445 ASTM D93 ASTM D93 IEC 60156 ASTM D445 ASTMD93 ASTMD93

500 ml CO + 0.5 wt.%

500 ml CO + 0.5 wt.%

Standards

–

Concentration of additives

Table 8 Addition of AntiOxidants and NPs on CO

60

36.5

60

BDV (KV)

@60°C–90

@40°C–27.2

@40°C–29

Viscosity (Cst)

280

301

225

Flash point (°C)

310

320

300

Fire point (°C)

[93]

[92]

[91]

References

252 S. N. Deepa et al.

Review on the Addition of Antioxidants and Nanoparticles to Natural …

Fig. 29 BDV comparison for CO with and without additives

Fig. 30 Viscosity comparison for CO with and without additives

Fig. 31 Flash point comparison for CO with and without additives

253

254

S. N. Deepa et al.

Fig. 32 Fire point comparison for CO with and without additives

6 Conclusion This work revived the evolution of alternating insulation fluid used in power transformers. The suitable environmentally friendly alternatives found were natural esters and synthetic esters. Many researchers have worked on the comparison of physical, electrical, and chemical properties of different natural esters. Some studies are based on the improvement of insulating properties of natural esters with the use of additives. This work revives the addition of additives like antioxidants and nanoparticles have incredible improvement on critical insulation properties of natural esters. In this review, SFO, RBO, NO, CO, MO, CO along with the different additives added are summarized. Among SFO, RBO, NO, CO, MO, and CO revived NO has tremendous improvement with the additives added to natural esters, However, it is concluded that antioxidants have the disadvantage of higher viscosity value when added to a natural ester thus giving the scope of improvement for the further studies. Acknowledgements The preferred spelling of the word “acknowledgment” in America is without an “e” after the “g”. Avoid the stilted expression “one of us (R. B. G.) thanks...”. Instead, try “R. B. G. thanks...”. Put sponsor acknowledgments in the unnumbered footnote on the first page.

References 1. Fofana I (2013) 50 years in the development of insulating liquids. IEEE Electr Insul Mag 29(5):13–25 2. Martins MAG (2010) Vegetable oils, an alternative to mineral oil for power transformersexperimental study of paper aging in vegetable oil versus mineral oil. IEEE Electr Insul Mag 26(6):7–13

Review on the Addition of Antioxidants and Nanoparticles to Natural …

255

3. Fernández I, Ortiz A, Delgado F, Renedo C, Perez S (2013) Comparative evaluation of alternative fluids for power transformers. Electr Power Syst Res 98:58–69 4. Rao UM, Sood YR, Jarial RK Jul 2016) Transformer insulation technology with esters based dielectric fluids-bibliography survey. In: 2016 IEEE 1st international conference on power electronics, intelligent control and energy systems (ICPEICES). IEEE, pp 1–6 5. Raymon A, Pakianathan PS, Rajamani MPE, Karthik R (2013) Enhancing the critical characteristics of natural esters with antioxidants for power transformer applications. IEEE Trans Dielectr Electr Insul 20(3):899–912 6. Abdi S, Boubakeur A, Haddad A, Harid N (2011) Influence of artificial thermal aging on transformer oil properties. Electr Power Compon Syst 39(15):1701–1711 7. Fofana I, Bouaicha A, Farzaneh M (2011) Characterization of aging transformer oil–pressboard insulation using some modern diagnostic techniques. Eur Trans Electr Power 21(1):1110–1127 8. Rao UM, Fofana I, Jaya T, Rodriguez-Celis EM, Jalbert J, Picher P (2019) Alternative dielectric fluids for transformer insulation system: progress, challenges, and future prospects. IEEE Access 7:184552–184571 9. Oommen TV (1988) Static electrification properties of transformer oil. IEEE Trans Electr Insul 23(1):123–128 10. Karthik R, Sree Renga Raja T (2012). Investigations of transformer oil characteristics. IEEJ Trans Electr Electron Eng 7(4):369–374 11. Gouda OE, El Dein AZ (2019) Prediction of aged transformer oil and paper insulation. Electr Power Compon Syst 47(4–5):406–419 12. Mehta DM, Kundu P, Chowdhury A et al (2016) A review of critical evaluation of natural ester visavis mineral oil insulating liquid for use in transformers: Part II. IEEE Trans Dielectr Electr Insul 23(3):1705–1712 13. Bandara K, “Condition monitoring of biodegradable oil filled transformer,” PhD Thesis, University of Queensland,2016. 14. Jamail N, Ishak M, Muhamad N (2017) Insulation characteristic analysis of coconut oil and palm oil as liquid insulating material. Int J Simul Syst Sci Technol 41(49):1–6 15. Sheppard H (1986) A century of progress in electrical insulation. IEEE Electr Insul Mag 2(5):20–30 16. Mercier G WEMCOL capacitor fluid development. In: Electrical/electronics insulation conference 17. Xu Y, Qian S, Liu Q, Wang ZD (2014) Oxidation stability assessment of a vegetable transformer oil under thermal aging. IEEE Trans Dielectr Electr Insul 21(2):683–692 18. Oommen TV, Claiborne CC, Mullen JT (1997) Biodegradable electrical insulation fluids. In: Electrical insulation conference 19. Crine J (1986) Silicone oil as replacement fluid for PCBs in transformers. Can Electr Eng J 11(3):110–113 20. Fritsche R, Rimmele U, Trautmann F et al (2014) Prototype 420 kV power transformer using natural ester dielectric fluid. In: 19th annual high voltage electrical equipment & management conference, TechCon North America, Phoenix, USA 21. Boris H, Gockenbach E, Dolata B (2008) Ester fluids as alternative for mineral based transformer oil. In: 2008 IEEE international conference on dielectric liquids 22. Raof NA, Yunus R, Rashid U, Azis N, Yaakub Z (2019) Effect of molecular structure on oxidative degradation of ester-based transformer oil. Tribol Int, 105852 23. Haramija V, Vrsaljko D, Durina V (2014) Thermal properties of synthetic ester-based transformer oil during ageing in laboratory conditions. In: 2014 IEEE 18th international conference on dielectric liquids (ICDL) 24. Bandara K, Ekanayake C, Saha TK (2015) Compare the performance of natural ester with synthetic ester as transformer insulating oil. In: 2015 IEEE 11th international conference on the properties and applications of dielectric materials (ICPADM) 25. Martin R, Athanassatou H, Duart JC, Perrier C, Sitar I, Walker J, Claiborne C, Boche T, Cherry D, Darwin A et al (2010) Experiences in service with new insulating liquids. Cigré Technical Brochure 436; International Council on Large Electric Systems (CIGRE): Paris, France

256

S. N. Deepa et al.

26. Russel M, Lashbrook M, Satija N (2013) Synthetic esters for power transformers AT > 100 kV. ITMA J 27. Rozga P, Beroual A, Przybylek P, Jaroszewski M, Strzelecki K (2020) A review on synthetic ester liquids for transformer applications 28. IEC 61099 (2010) Insulating liquids—specifications for unused synthetic organic esters for electrical purposes; International Electrotechnical Commission (IEC): Geneva, Switzerland 29. Stockton DP, Bland Jr JR, McClanahan T, Wilson M, Harris DL, McShane P (Jan/Feb 2009) Seed-oil-based coolants for transformers. IEEE Ind Appl Mag, pp 68–74 30. McShane CP (May/Jun 2000) New safety dielectric coolants for distribution and power transformers. IEEE Ind Appl Mag, 24–32 31. Bashi SM, Abdullahi UU, Yunus Rm, Nordin A (2006) Use of natural vegetable oils as alternative dielectric transformer coolants 32. Sarfi V, Mohajeryami S, Majzoobi A (2017) Estimation of water content in a power transformer using moisture dynamic measurement of its oil. IET High Volt 2(1):11–16 33. Fernández I, Ortiz A, Delgado F et al (2013) Comparative evaluation of alternative fluids for power transformers. Electr Power Sys Res 98:58–69 34. Prevost TA (2006) Dielectric properties of natural esters and their influence on transformer insulation system design and performance. In: Transmission and distribution conference and exhibition 35. Hosier IL, Guushaa A, Westenbrink EW, Rogers C, Vaughan AS, Swingler SG (2011) Aging of biodegradable oils and assessment of their suitability for high voltage applications 36. Kumar SS, Iruthayarajan MW, Bakrutheen M (2016) Investigations on the suitability of rice bran oil and corn oil as alternative insulating liquids for transformers. IEEJ Trans Electr Electron Eng 11(1):10–14 37. Kumar SS, Iruthayarajan MW, Bakrutheen M, Kannan SG (2016) Effect of antioxidants on critical properties of natural esters for liquid insulations. IEEE Trans Dielectr Electr Insul 23(4):2068–2078 38. Bhatnagar AS, Prasanth Kumar PK, Hemavathy J, Gopala Krishna AG (2009) Fattyacid composition, oxidative stability, and radical scavenging activity of vegetable oil blends with coconut oil. J Am Oil Chem Soc 86(10):991–999 39. Singh A, Singh IS (1991) Chemical evaluation of mahua (Madhuca indica) seed. Food Chem 40(2):221–228 40. Aluyor EO , Ori-Jesu M (2008) The use of antioxidants in vegetable oils–a review. Afr J Biotechnol 7(25):4836–4842 41. Raymon A, Karthik R (2013) Enhancement of critical parameters of used transformer oil with naturally Activated Bentonite and investigation of vegetable oil performance with antioxidants. In: 2013 international conference on circuits, power and computing technologies (ICCPCT) 42. RubalyaValantina S, Neelamega P (2012) Antioxidant potential in vegetable oil. Res J 0/Chem Environ, India 16(2) 43. Du Y, Lv Y, Li C, Chen M, Zhong Y, Zhou J, Li X, Zhou Y (2012) Effect of semiconductive nanoparticles on insulating performances of transformer oil. IEEE Trans Dielectries Electr Insul 19(3) 44. Domingos AK, Saad EB, Vechiatto WWD, Wilhelm HM, Ramos L (2007) The influence of BHA, BHT and TBHQ on the oxidation stability of soybean oil ethyl esters (Biodiesel). J Brazilian Chem Soc 18:416–423 45. Emmanuel OA, Mudiakeoghene OJ (2008) The use of antioxidants in vegetable oils–a review. Afr J Biotechnol 7:4836–4842 46. Denisov ET, Afanas’ev IB (2005) Oxidation and antioxidants in organic chemistry and biology. Taylor & Francis Group 47. Krishnaiah D, Sarbatly R, Rajesh N (2011) Report-a review of the antioxidant potential of medicinal plant species. Food Bioprod Process 89(3):217–233 48. Smith SA, King RE, Min DB (2006) Oxidative and thermal stabilities of genetically modified high oleic sunflower oil. Food Chem 102(4):1208–1213 49. Lewis T (1994) Nanometric dielectrics. IEEE Trans Dielectr Electr Insul 1(5):812–825

Review on the Addition of Antioxidants and Nanoparticles to Natural …

257

50. Asokan AN, Anagha AP, Preetha P (2019) Effect of interphase on dielectric relaxation mechanisms in nanocomposites. IEEE Trans Dielectr Electr Insul 26(4):1211–1219 51. Vaughan A (2016) W(h)ither dielectrics? In: IEEE international conference on dielectrics (ICD), vol 1, pp 1–9 52. Lau KY, Vaughan A, Chen G (2015) Nanodielectrics: opportunities and challenges. IEEE Electr Insul Mag 31(4):45–54 53. Contreras J, Rodriguez E, Taha-Tijerina J (2017) Nanotechnology applications for electrical transformers, a review. Electr Power Syst Res 143:573–584 54. Choi SU, Eastman JA (1995) Enhancing thermal conductivity of fluids with nanoparticles. Argonne National Lab, IL (United States), Technical report 55. Sima W, Shi J, Yang Q et al (2015) Effects of conductivity and permittivity of nanoparticle on transformer oil insulation performance: experiment and theory. IEEE Trans Dielectr Electr Insul 22(1):380–390 56. Choi S (1995) Enhancing thermal conductivity of fluids with nanoparticles. In: Siginer DA, Wang HP (eds) Developments applications of non-Newtonian flows, FED-vol 231/MD-vol 66, ASME, New York, NY, USA, pp 99–105 57. Hwang JG, O’Sullivan F, Zahn M et al (2008) Modeling of streamer propagation in transformer oil-based nanofluids. In: Annual report conference electrical insulation and dielectric phenomena, CEIDP, pp 361–366 58. Ibrahim M, Abd-Elhady A, Izzularab M (2016) Effect of nanoparticles on transformer oil breakdown strength: experiment and theory. IET Sci Meas Technol 10:839–845 59. Du Y, Lv Y, Li C, Chen M, Zhong Y, Zhou J, Li X, Zhou Y (2012) Effect of Semiconductive Nanoparticles on Insulating Performances of Transformer Oil. IEEE Trans Dielectr Electr Insul 19(3):770–776 60. Karthik R, Raja TSR, Madavan S (2012) Enhancement of critical characteristics of transformer oil using nano materials. Arab J Sci Eng 3(20):369–374 61. Segal V, Rabinovich A, Nattrass S, Raj K, Nunes A (2000) Experimental study of magnetic colloidal fluids behavior in power transformers. J Magn Magn Mater 215:513–515 62. Xuan Y, Li Q (2000) Heat transfer enhancement of nanofluids. Int J Heat Fluid Flow, 21–58 63. Godson L, Raja B, Mohan D, Wongwises S (2010) Enhancement of heat transfer using nanofluids-an overview. Renew Sustain Energy Rev 14(2):629–641 64. Jacob J, Preetha P, Thiruthi Krishnan S (2020) Review on natural ester and nanofluids as an environmental friendly alternative to transformer mineral oil. IET Res J. ISSN 1751-8644 65. Ab Ghani S, Muhamad NA, Noorden ZA, Zainuddin H, Bakar NA, Talib MA (Oct 2018) Methods for improving the workability of natural ester insulating oils in power transformer applications: a review. Artic Electr Power Syst Res. https://doi.org/10.1016/j.epsr.2017.10.008 66. Hwang JG, Zahn M, O’sullivan FM et al (2010) Effects of nanoparticle charging on streamer development in transformer oilbased nanofluids. J Appl Phys 107(1):014310 67. Mehta D, Kundu P, Chowdhury A (2015) Indian transformer industry gearing up for next-gen green liquids. In: 5th Nirma university international conference on engineering (NUiCONE), pp 1–4 68. Thomas P et al (Feb 2019) Synthetic ester oil based high permittivity CaCu3 Ti4 O12 (CCTO) nanofluids an alternative insulating medium for power transformers. IEEE Trans Dielectr Electr Insul 26(1) 69. Cho EJ, Holback H, Liu KC, Abouelmagd SA, Park J, Yeo Y (2013) Nanoparticle characterization: state of the art, challenges, and emerging technologies 70. Li Y, Zhou J, Tung S, Schneider E, Xi S (2009) A review on development of nanofluid preparation and characterization. Powder Technol 196(2):89–101 71. Yu W, Xie H (Jul 2012) A review on nanofluids: preparation, stability mechanisms, and applications. J Nanomater 2012. Art. no. 435873 72. Segal V, Raj K (1998) An investigation of power transformer cooling with magnetic fluids. Indian J Eng Mater Sci 5(6):416–422 73. Lv Y, Zhou Y, Li C, Wang Q, Qi B (2014) Recent progress in nanofluids based on transformer oil: preparation and electrical insulation properties. IEEE Electr Insul Mag 30(5):23–32

258

S. N. Deepa et al.

74. Kolcunova I, Pavlik M, Lison L (2016) Assessment of long thermal ageing on the oil-paper insulation. Adv Electr Electron Eng 14(5):506–511 75. Rafiq M, Lv YZ, Zhou Y, Ma KB, Wang W, Li CR, Wang Q (2015) Use of vegetable oils as transformer oils–a review. Renew Sustain Energy Rev 52:308–324 76. Nagendran S, Chandrasekar S (2018) Investigations on partial discharge, dielectric and thermal characteristics of nano SiO2 modified sunflower oil for power transformer applications. J Electr Eng Technol 13(3):1337–1345 77. Madavan R, Balaraman S (2017) Investigation on effects of different types of nanoparticles on critical parameters of nano-liquid insulation systems. J Mol Liq 230:437–444 78. Raymon A, Sakthibalan S, Cinthal C, Subramaniaraja R, Yuvaraj M (2016) Enhancement and comparison of nano-ester insulating fluids. IEEE Trans Dielectr Electr Insul 23(2):892–900 79. Karthik M, Iruthayarajan MW, Bakrutheen M (Mar 2015) Investigation of vegetable oil blended with antioxidant. In: 2015 IEEE international conference on electrical, computer and communication technologies (ICECCT). IEEE, pp 1–7 80. Thirugnanam ARPR, Siluvairaj WIM, Karthik R (2017) Performance studies on dielectric and physical properties of eco-friendly based natural ester oils using semi-conductive nanocomposites for power transformer application. IET Sci Meas Technol 12(3):323–327 81. Mohammed LS, Bakrutheen M, Willjuice M, Karthik M (Dec 2015) Studies on critical properties of vegetable oil-based insulating fluids. In: 2015 annual IEEE India conference (INDICON). IEEE, pp 1–4 82. Sankarasubbu UD, Suburaj SK (2019) Examining the properties of neem oil, sunflower oil and mahua oil with antioxidants and nano powders for power transformer. Am J Electr Comput Eng 3(1):20–29 83. Karthik R, Raja TSR, Shanmugam SS, Sudhakar T (2012) Performance evaluation of ester oil and mixed insulating fluids. J Inst Eng, India, Series B 93(3):173–178 84. Karthik R, Sree Renga Raja T (2012) Investigations of transformer oil characteristics. IEEJTrans Electr Electron Eng 7:369–374 85. Bakrutheen M, Karthik R, Madavan R (2013) Investigation of critical parameters of insulating mineral oil using semiconductive nanoparticles. In: IEEE international conference circuits, power and computing technologies-ICCPCT, pp 294–299 86. Alternative to transformer insulation oil. Int J Sci Eng Res (IJSER) 2(4):31–37 87. Egbuna SO, Ude OC, Ude CN International journal of engineering sciences & research technology. Castor seed bio-transformer oil as an alternative to conventional transformer oil 88. Bhaumik S, Maggirwar R, Datta S, Pathak SD (2018) Analyses of anti-wear and extreme pressure properties of castor oil with zinc oxide nano friction modifiers. Appl Surf Sci 449:277– 286 89. Oktaviani WA, Sidik MAB (Oct 2019) Breakdown voltage of biodegradable oil RBDPO olein and oleum maydis by nanoparticles addition. In: 2019 international conference on electrical engineering and computer science (ICECOS). IEEE, pp 382–385 90. Rathore Y, Ramchandani D (2016) Mahua biodiesel: an alternative to conventional diesel fuel. IOSR J Mech Civ Eng (IOSR-JMCE), 42–51 91. Reddy BK (2019) Latest trends in use of transformer oils. Int J Eng Trends Technol 67(7):37–39 92. Dung N, Trung NH (2019) A study on breakdown properties of coconut oil. Vietnam J Sci Technol 57:320 93. Jacob J, Prabhu P, Sindhu K (2020) A review on natural ester and nanofluid as alternative to transformer mineral oil. IET Nanodielectrics, 3. https://doi.org/10.1049/iet-nde.2019.0038