Application of ANN to voltage stability assessment and enhancement.

(1)

APPLICATION OF ANN TO VOLTAGE STABILITY ASSESSMENT AND ENHANCEMENT

by

SUTHAR BHAVIK N.

CENTRE FOR ENERGY STUDIES

Submitted in fulfillment of the requirements of the degree of

DOCTOR OF PHILOSOPHY

to the

INDIAN INSTITUTE OF TECHNOLOGY DELHI

MAY 2008

(2)

DEDICATED TO

MY FATHER

(3)

CERTIFICATE

This is to certify that the thesis entitled, "APPLICATION OF ANN TO VOLTAGE STABILITY ASSESSMENT AND ENHANCEMENT" being submitted by Mr. Bhavik N. Suthar to the Indian Institute of Technology, Delhi for the award of Doctor of Philosophy is a record of bonafide research work carried out by him under my guidance and supervision in conformity with the rules and regulations of Indian Institute of Technology Delhi.

The research report and results presented in this thesis have not been submitted, in part or full, to any other university or institute for the award of any degree or diploma.

R.,_• ok 1 ot

s

^clA

(Prof. R. Balasubramanian)

Professor,

Centre for Energy Studies,

Indian Institute of Technology, Delhi New Delhi-110016

(4)

ACKNOWLEDGEMENTS

I express my deep gratitude to my research advisot Prof. R.

Balasubramanian, for motivating, inspiring me and for extending keen interest and kind support throughout my research work. I have no words for his unbelievably quick and perfect evaluation of all my results and papers. He has nurtured research in me in such a manner that research and work have become a passion for me and I never felt my research work as a burden. His high technical skill and systematic approach have made this endeavor possible. Without his unconditional support, this task couldn't have been so easy for me. It was a fortunate and unforgettable experience to work under his reflective and revered guidance. His friendly nature and endless kindness can't be thanked adequately here.

I am profoundly thankful to the Head, CES, for providing me with all the necessary facilities during the course of my work. I wish to convey my sincere thanks to Prof. T. S. Bhatti, for motivating me in the research work and for helping me in many administrative works. I also wish to thank the other faculty members and office staff of the department.

I am grateful and obliged to the Directorate of Technical Education, Gujarat State for permitting me to pursue Ph.D. kinds of support provided by Government Engineering College, Modasa is greatly acknowledged. I also wish to convey my sincere thanks to Prof. L. D.

Arya and Prof. R. S .Tare of SGSITS, Indore for having been a constant motivation for my research.

I was fortunate to have an excellent work environment in the laboratory, which facilitated my work to a great deal. For this, I am highly thankful to Mr. Dalel Singh, lab-in-charge, for his constant help in every possible way to carry forward my research work. I am grateful to my friends and colleagues Gauri Shankar, Prince, Rajkumar and Manish, for keeping a cordial environment in the lab and for providing fruitful suggestions. Very special thanks to all friends of Gujju group: Sanjay, Saurabh, Bhavesh, Dr. Parag, Dr. Rajan, Jain Sir, Rajesh, Vadodaria, K.M., Kolte, Dipak, Dr. Momaya sir, Manishabhabhi, and all the Gujju

ii

(5)

indirectly contributed to the realization of this thesis. 4.4 11, ^.

.411/

B ivik Suthar team. We have shared a lot of activities and developed special relations with the families of Sanjay, Parag, Nimit and Vadodaria during my research work. I am also thankful to Alpesh, R. K Patel, S. P. Patel and all colleagues at GEC Modasa for their support and for looking after my administrative work at GEC, Modasa during my absence in Modasa during this research.

I have no words to express the motivation and inspiration provided by my father to me, whose courage and descipline through out his life has been a spiritual inspiration to me and holds no bounds. His encouraging words, deeds and emotional time I have spent with him will always remain in my heart and soul. Angels escorted him into heaven in 1998.. His sudden demise put me in--a fast track research, without wasting time, I started my P.G and subsequently Ph. D through his spiritual inspiration. My mother has always taken care to avoid any discouragement to me in tough times and inspired me immensely thorough out my research. My wife, Archana has always been a source of motivation through her curiosity about my research and handling of all the problems I faced. She took all the responsibilities of taking care of Mudra and family. She has faced lots of difficulties during my research tenure, specifically related to our accommodation but never complained about them and instead she has always taken care that my research would not be affected due to problems we faced. My daughter, Mudra always made me fresh and energetic by her smile at the end of each day when I returned home tired.The word "Thanks" is not enough to say. I am also thankful to my sisters and brothers in-law for their constant encouragement and support during my studies.

I am grateful to the IIT Delhi, DST, AICTE, ISAP Society (Intelligent System Application to Power System), and iREP (International Research and Education in Power systems, USA) for providing me the travel support/grant to participate in the international conferences.

Finally, I would like to thank all of those, who have directly or

iii

(6)

ABSTRACT

The reason attributed to many blackouts and brownouts, experienced globally in various countries even in recent times, is the voltage collapse phenomenon. To prevent these happenings, the need for employing fast real-time voltage stability monitoring tools and get prior warning about the evolving voltage instability situations in a power system cannot be overemphasized. It is well known that the trained ANN's are eminently suitable tools for performing these functions.

In this thesis, a new approach to ANN-based voltage stability monitoring has been proposed. The novelty of this approach is that here the emphasis is shifted from a system-wide loading limit assessment methodology to a methodology of concentrating on the few vulnerable load buses of the system and get prior warning signals about the "Distance To Voltage Collapse" at these buses getting closer to the voltage collapse points. In this approach a separate ANN has been trained for each of the vulnerable load buses in the system from the voltage stability point of view. The vulnerable load buses of the system are identified by modal analysis of the system reduced Jacobian (of Q-V coupling) matrix. The output obtained from each of these ANN's is the "Distance to Voltage Collapse" in terms of the MVAR margin available at the specific vulnerable load bus corresponding to the current operating point (loading condition) of the system. The novel inputs proposed for these ANN's consist of the complex power contributions of each of the generators and other controllable reactive power sources provided in the system in meeting the load at the particular load bus at the current loading condition and the electrical distances between these sources and the load bus, in addition to the conventionally used inputs of reactive power margins at the sources and the complex power drawn and voltage magnitude at the particular load bus. Special care has been taken in

iv

(7)

generating a comprehensive set of realistic loading conditions in the system, including generating unit outage conditions and relevant "N-I" network element outage contingency conditions.

Having trained these ANN's, they can be utilized for evaluating the DTC's obtaining at the vulnerable load buses in real time. Some implementation aspects of the proposed ANN-based tool are discussed in Chapter-4.

Modern power systems are going through restructuring processes and in this context, with the many bilateral and multilateral transactions simultaneously taking place in these systems in addition to the pool operation transactions, the system operators have the daunting task of monitoring the stability of the system and taking appropriate corrective actions in real time. This issue has been addressed in Chapter-5 and the usefulness of the proposed ANN-based tool in this context has been demonstrated.

Modern power systems employ many compensating devices like SVC's and the normal control strategy used in these devices is to sense the voltages at the buses, where these are located in the system and control them to maintain constant voltages at these locations. With the proposed ANN-based tool available, it is feasible to use the DTC's obtained from this tool as an additional feedback signal for exercising a secondary level control and coordination of these SVC's.

(8)

CONTENTS

CERTIFICATE i

ACKNOWLEDGEMENTS ii

ABSTRACT iv

CONTENTS vi

LIST OF FIGURES x

LIST OF TABLES xiv

LIST OF SYMBOLS AND ABBREVIATIONS xvi

CHAPTER 1 INTRODUCTION 1

1.1 INTRODUCTION 1

1.2 VOLTAGE STABILITY AND VOLTAGE COLLAPSE PHENOMENON 3

1.3 COMPLEXITY OF THE PROBLEM 4

1.4 MOTIVATION 5

1.5 ORGANIZATION OF THE THESIS 6

CHAPTER 2 LITERATURE SURVEY 9

2.2 STATIC AND DYNAMIC VOLTAGE STABILITY 9 2.3 STATIC VOLTAGE STABILITY ANALYSIS METHODS 12

2.3.1 Sensitivity Factors 12

2.3.2 Singular Value Decomposition /Least Eigen Value Computation 14

2.3.3 Continuation Power Flow 16

2.3.4 Energy Function 17

vi

(9)

2.3.5 L-Index 18

2.3.6 Voltage Stability Index 20

2.3.7 Centroid Method 21

2.4 THE MODAL ANALYSIS METHOD 22 2.5 AI BASED METHODS FOR REACTIVE POWER-VOLTAGE CONTROL 23 2.6 ANN BASED METHODS FOR VOLTAGE STABILITY ANALYSIS 26 2.7 IMPORTANT ASPECTS OF THE PRESENT WORK 28

CHAPTER 3 VOLTAGE STABILITY ASSESSMENT METHOD USING ANN 31

3.2 GENERATION OF THE LOADING PATTERNS 32 3.3 IDENTIFICATION OF MOST VULNERABLE LOAD BUSES 33

3.3.1 The Modal Analysis 35

3.4 SELECTION OF INPUT/OUTPUT PATTERNS PRESENTED TO ANN 42

3.4.1 Input Pattern Generation 44

3.4.2 Output Pattern Generation 49

3.4.3 Flow Chart for ANN Patterns Generation 50 3.5 DESIGN AND TRAINING OF THE ANN'S 51

3.6 SUMMARY 51

CHAPTER 4 APPLICATION-1: ONLINE VOLTAGE STABILITY

ASSESSMENT 53

4.2 APPLICATION TO IEEE 30 BUS TEST SYSTEM 53 4.2.1 Identification of Most Vulnerable Load Buses 55

4.2.2 Input Pattern Generation 57

4.2.3 Output Pattern Generation 58

4.2.4 Effect of System Loading on DTC 60

4.3 DESIGN OF ANN ARCHITECHTURE AND ANN TRAINING 65

vii

(10)

4.4 TESTING OF THE TRAINED ANN'S 67 4.5 APPLICATION OF TRAINED ANN'S FOR REAL TIME VOLTAGE

STABILITY MONITORING 73

4.6 SUMMARY 76

CHAPTER 5 APPLICATION-2: VOLTAGE STABILITY ANALYSIS OF

RESTRUCTURED POWER SYSTEMS 79

5.2 APPLICATION TO CIGRE 32 BUS SYSTEM 80

5.2.1 Generation of Loading Patterns 82

5.2.2 Identification of Most Vulnerable Load Buses 86 5.2.3 Generation of Input Patterns for ANN 88 5.2.4 Generation of Output Patterns for ANN 89

53 DESIGN AND TRAINING OF ANN'S 89 5.4 TESTING OF THE TRAINED ANN'S 91 5.5 EFFECT OF BILATERAL TRANSACTIONS ON DTC'S 107

5.6 SUMMARY 111

CHAPTER 6 APPLICATION-3: ON-LINE REACTIVE POWER

MANAGEMENT 113

6.1 INTRODUCTION 113

6.2 APPLICATION TO CIGRE 32 BUS TEST SYSTEM 114

6.3 RESULTS AND DISCUSSIONS 114

6.4 SUMMARY 126

CHAPTER 7 CONCLUSION 127

7.1 SUMMARY OF THE RESEARCH CONTRIBUTION 127 7.2 SUGGESTIONS FOR FURTHER RESEARCH 129

viii

(11)

REFERENCES 131

APPENDIX I 139

APPENDIX H 142

APPENDIX III 145

Bio data 149

ix