High Penetration of PV Systems in Low Voltage Distribution

Networks: Investigation of Power Quality Challenges and

Mitigation

Annapoorna Chidurala

Bachelor of Technology in Electrical and Electronics Engineering

Master of Technology in Electrical Power Engineering

A thesis submitted for the degree of Doctor of Philosophy at

The University of Queensland in 2016

School of Information Technology and Electrical Engineering

Abstract

World electricity demand is continuously increasing and fossil fuel supplies are not sustainable.
Solar Photovoltaic (PV) energy is one of the emerging resources around the world, which
produces emission free electricity. Nowadays, the advancements in rooftop solar PV
technology, government subsidies, decreasing capital cost and feed-in-tariffs have promoted
installation in residential and commercial applications. The exponential uptake in widespread
integration of PV systems in existing low voltage (LV) distribution networks is raising
additional new challenges in terms of power quality, stability and protection. In LV distribution
networks, poor power quality (PQ) is the most serious concern.

Characteristically, LV distribution networks are not designed for significant back-feed of

power generation to the main grid. Also, these networks are unbalanced in nature due to
asymmetry in system impedances and single-phase loads. This together with a large number of
small-scale PV system integrations in LV networks can cause poor PQ challenges in terms of
voltage quality and harmonics. PV systems can themselves generate harmonics, due to the
usage of power electronic inverters. In addition, the augmentation of power electronics based
appliances; the loads are becoming voltage sensitive and nonlinear in nature. The proliferation
of widespread PV penetrations and a multitude of nonlinear load characteristics can have a
stringent impact on the network harmonic levels. Therefore, the main objective of this research
is to investigate and understand the impacts of high PV systems penetration on PQ of the
distribution network and aim to alleviate them.

In the first part of this thesis, the investigation of voltage quality challenges in the LV
distribution network with high PV penetration are discussed. In this research, various voltage
quality issues such as voltage rise, unbalance, fluctuations/flicker and sag/swell issues have
been explored. Primarily, the analysis of results has been carried out through PSCAD
simulations in various case studies. For this purpose, an IEEE-13 bus unbalanced distribution
network is considered as a test system. Furthermore, to evaluate the severity of voltage quality
issues in real-time grid connected PV systems, field measurement based investigations have
been performed. Practical field tests have been conducted at two different sizes of 1.5 MW and
3.3MW PV systems located at the University of Queensland (UQ), St Lucia and Gatton
campuses respectively. The impact of dynamic variations in solar irradiation has also been
taken into account for the analysis. Further, a data clustering technique is also applied to
estimate the probability of voltage flicker severity in the networks. Measurement results show

ii
that voltage quality concerns in the 1.5MW PV system are insignificant compared to the
3.3MW system.

In the second part of this thesis, the characteristics of harmonic emissions from PV inverters
and their aggregations during various operating conditions are assessed. The simulation results
are validated with the field measurement data collected by various PQ analysers connected at
the UQ PV site. Analysis revealed that individual voltage and current harmonic magnitudes are
additive in nature due to increased PV system penetration. In addition, a comprehensive
analysis has been performed in several different cases studies with high penetration of different
PV inverter technologies to evaluate the severity of harmonic propagation and resonance issues
on the distribution network. This analysis has also considered the harmonic distortions
associated with various power electronic based nonlinear loads. Further, comparative studies
have been performed with real-time harmonic measurements, which are obtained using online
JAVA programs. The study has highlighted the PV system harmonic contributions on
distribution transformer K-factor. Results confirm that the total harmonic distortions (THD) of
voltage and current are exceeding the IEEE limits when the number of PV systems increases
in the network. Moreover, the impacts of PV controller performance due to solar irradiation
variations on the incidence of grid harmonic resonance have been presented.

Furthermore, this research has suggested a novel solution to overcome the above PQ issues.
The concept of adopting the PV inverter as a virtual DSTATCOM named as Solar-
DSTATCOM has been proposed. Also, a new control strategy for the PV inverter has been
developed to provide independent phase voltage regulation and load reactive power and
harmonic compensation, which could eliminate issues in the unbalanced distribution network.
Initially, the Solar-DSTATCOM controller has been verified in a PSCAD simulation
environment. Further, different case studies have been performed on the IEEE-13 bus network
for PQ issues compensation. In addition, the proposed Solar-DSTATCOM control system has
been tested and verified in controller hardware-in-the-loop simulation environment, which
combines the real-time digital simulator and dSPACE DS1103 hardware board. Detailed
investigations are carried out for various different case studies, which include daytime, night-
time operations, the impact of dynamic load profiles and finally harmonic analysis. The
analysis has revealed that Solar-DSTATCOM exceptional performance in the hardware
environment has enhanced the grid PQ by providing voltage regulation, reactive power
compensation and power factor correction. The harmonic emissions are well within the limits.

iii
Declaration by author

This thesis is composed of my original work, and contains no material previously published or
written by another person except where due reference has been made in the text. I have clearly
stated the contribution by others to jointly-authored works that I have included in my thesis.

I have clearly stated the contribution of others to my thesis as a whole, including statistical
assistance, survey design, data analysis, significant technical procedures, professional editorial
advice, and any other original research work used or reported in my thesis. The content of my
thesis is the result of work I have carried out since the commencement of my research higher
degree candidature and does not include a substantial part of work that has been submitted to
qualify for the award of any other degree or diploma in any university or other tertiary
institution. I have clearly stated which parts of my thesis, if any, have been submitted to qualify
for another award.

I acknowledge that an electronic copy of my thesis must be lodged with the University Library
and, subject to the policy and procedures of The University of Queensland, the thesis be made
available for research and study in accordance with the Copyright Act 1968 unless a period of
embargo has been approved by the Dean of the Graduate School.

I acknowledge that copyright of all material contained in my thesis resides with the copyright
holder(s) of that material. Where appropriate I have obtained copyright permission from the
copyright holder to reproduce material in this thesis.

iv
Publications during candidature

Peer-reviewed International Journal Article (1):

1. Chidurala, Annapoorna, Kumar Saha, Tapan, Mithulananthan, N.: 'Harmonic impact

of high penetration photovoltaic system on unbalanced distribution networks – learning
from an urban photovoltaic network', IET Renewable Power Generation, Volume 10,
Issue 4, April 2016, pp. 485 – 494 DOI: 10.1049/iet-rpg.2015.0188.

International Journal Articles to be Submitted (2):

1. Chidurala, Annapoorna, Tapan Kumar Saha, and Mithulananthan N., “Harmonic

Resonance on the Distribution Network with PV Penetration and Field Measurements
Validation”, in preparation, to be submitted to International Transactions on Electrical
Energy Systems Journal by September 2016.
2. Chidurala, Annapoorna, Tapan Kumar Saha, and Mithulananthan N., “Real-Time
Controller Hardware-in-the-Loop Simulations of Solar-DSTATCOM” in preparation,
to be submitted to Electric Power Systems Research Journal by September 2016.
Peer-reviewed Conference Papers (5):

1. Chidurala, Annapoorna, Tapan Kumar Saha, and N. Mithulananthan, "Power quality

enhancement in unbalanced distribution network using Solar-DSTATCOM,"
Australasian Universities Power Engineering Conference (AUPEC), Hobart,
TASMANIA, Australia, Sept. 29-Oct. 3, 2013, pp. 1-6.

2. Chidurala, Annapoorna, Tapan Kumar Saha, N. Mithulananthan, and R. C. Bansal,

"Harmonic emissions in grid connected PV systems: A case study on a large scale
rooftop PV site," in PES General Meeting Conference & Exposition, 2014 IEEE,
National Harbor, MD, 27-31 July 2014, pp. 1-5. USA.
3. Chidurala, Annapoorna, Tapan Kumar Saha, and N. Mithulananthan, "Harmonic
characterization of grid connected PV systems & validation with field
measurements," in Power & Energy Society General Meeting, 2015 IEEE, 2015, pp. 1-
5, Denver, USA.
4. Chidurala, Annapoorna, Tapan Kumar Saha, and Mithulananthan N., “Field
Investigation of Voltage Quality Issues in Distribution Network with PV Penetration”,
IEEE PES Asia-Pacific Power and Energy Engineering Conference, Brisbane, QLD,
Australia, 15-18 Nov. 2015, pp. 1-5.

v
5. M J E ALAM, R YAN, Tapan Kumar Saha, Chidurala, Annapoorna, D EGHBAL,
“Learning from a 3.275 MW Utility Scale PV Plant Project”, in CIGRE PARIS Session
2016, 21, rue d’Artois, F-75008 PARIS 2016.

vi
Publications included in this thesis

1. Chidurala, Annapoorna; Kumar Saha, Tapan; Mithulananthan, N.: 'Harmonic impact of

high penetration photovoltaic system on unbalanced distribution networks – learning from
an urban photovoltaic network', IET Renewable Power Generation, Volume 10, Issue 4,
April 2016, pp. 485 – 494 DOI: 10.1049/iet-rpg.2015.0188.
This paper is incorporated in Chapter 4.

Contributor Statement of Contribution

Annapoorna Chidurala Simulation and Modelling (100%)

Practical Measurement data collection (100%)
Results interpretation and Discussion (75%)
Paper Writing (75%)
Tapan K. Saha Results interpretation and Discussion (20%)
Paper Writing and Review (20%)

Mithulananthan N. Results interpretation and Discussion (5%)

Paper Writing and Review (5%)

2. Chidurala, Annapoorna, T. K. Saha, and N. Mithulananthan, "Power quality

enhancement in unbalanced distribution network using Solar-DSTATCOM," Australasian
Universities Power Engineering Conference (AUPEC), Hobart, TASMANIA, Sept. 29-
Oct. 3, 2013, pp. 1-6.

This paper partially incorporated in Chapter 4 and Chapter 7.

Contributor Statement of Contribution

Annapoorna Chidurala Simulation and Modelling (100%)

Results interpretation and Discussion (75%)
Paper Writing (75%)

Tapan K. Saha Results interpretation and Discussion (15%)

Paper Writing and Review (15%)

Mithulananthan N. Results interpretation and Discussion (10%)

Paper Writing and Review (10%)

vii
3. Chidurala, Annapoorna, T. K. Saha, N. Mithulananthan, and R. C. Bansal, "Harmonic
emissions in grid connected PV systems: A case study on a large scale rooftop PV site,"
in PES General Meeting Conference & Exposition, 2014 IEEE, National Harbor, MD,
USA, 27-31 July 2014, pp. 1-5.
This paper partially incorporated in Chapter 7.
Contributor Statement of Contribution

Annapoorna Chidurala Simulation and Modelling (100%)

Results interpretation and Discussion (75%)
Paper Writing (75%)

Tapan K. Saha Results interpretation and Discussion (15%)

Paper Writing and Review (15%)

Mithulananthan N. Results interpretation and Discussion (10%)

Paper Writing and Review (5%)

R.C Bansal Paper Writing and Review (5%)

4. Chidurala, Annapoorna, T. K. Saha, and N. Mithulananthan, "Harmonic characterization

of grid connected PV systems & validation with field measurements," in Power &
Energy Society General Meeting, 2015 IEEE, 2015, pp. 1-5, Denver, USA.
This paper is incorporated in Chapter 5.
Contributor Statement of Contribution

Annapoorna Chidurala Simulation and Modelling (100%)

Field Experiments and Data Collection (100%)
Results interpretation and Discussion (75%)
Paper Writing (75%)
Tapan K. Saha Results interpretation and Discussion (20%)
Paper Writing and Review (20%)

Mithulananthan N. Results interpretation and Discussion (5%)

Paper Writing and Review (5%)

viii
5. Chidurala, Annapoorna, T. K. Saha, and N. Mithulananthan, “Field Investigation of
Voltage Quality Issues in Distribution Network with PV Penetration”, IEEE PES Asia-
Pacific Power and Energy Engineering Conference, Brisbane, QLD, Australia, 15-18 Nov.
2015, pp. 1-5.
This paper is incorporated in Chapter 4.
Contributor Statement of Contribution

Annapoorna Chidurala Simulation and Modelling (100%)

Field Experiments and Data Collection (100%)
Results interpretation and Discussion (75%)
Paper Writing (75%)
Tapan K. Saha Results interpretation and Discussion (20%)
Paper Writing and Review (20%)

Mithulananthan N. Results interpretation and Discussion (5%)

Paper Writing and Review (5%)

6. M J E ALAM, R YAN, T K SAHA, Chidurala, Annapoorna, D EGHBAL, “Learning

from a 3.275 MW Utility Scale PV Plant Project”, in CIGRE PARIS Session 2016, 21, rue
d’Artois, F-75008 PARIS 2016.
This paper partially incorporated in Chapter 4.
Contributor Statement of Contribution

M J E ALAM Field Experiments and Data Collection (100%)

Results interpretation and Discussion (75%)
Paper Writing (50%)
R YAN Simulation and Modelling (100%)
Results interpretation and Discussion (75%)
Paper Writing and Review (15%)
Tapan K. Saha Results interpretation and Discussion (25%)
Paper Writing and Review (20%)

Annapoorna Chidurala Data Collection (100%)

Results interpretation and Discussion (75%)
Paper Writing and Review (15%)

ix
Contributions by others to the thesis

“No contributions by others”

Statement of parts of the thesis submitted to qualify for the award of another degree

“None”

x
Acknowledgement:

First and foremost, I would like to take this opportunity to thank my God “Sri Saibaba” for his
presence in my life. His abundance grace, kindness and blessings gave me strength to complete
this research successfully.

I would like to offer my sincere thanks and gratitude to all those who helped me in many ways
in completing this research work.

I would like to express my gratitude to my principal advisor Prof. Tapan Kumar Saha and Co-
advisor Dr. Mithulan Nadarajah for their valuable suggestions, continuous support in every
aspect to accomplish my PhD study. Especially, Prof. Tapan Saha for novel association of
ideas, encouragement, appreciation and his intellectual zeal have motivated me to endeavor
this research successfully. He has been a continuous source of inspiration to me throughout the
research. His immense knowledge, creativity and passion always inspired me during the study.

I would like to acknowledge the financial support from the University of Queensland through
UQI Scholarship. I am also grateful to Australian Research Council and Industry partner,
Ingenero for their financial support towards my living expenses and attending conferences.

I am grateful to all the staff and colleagues of Power and Energy Systems group, who made
my stay pleasant and profitable. I would like to thank Dr. Olav Krause and Mr. Shane Goodwin
for their assistance in obtaining the field Measurement data for my research work. I would like
to thank all my friends from the research group: Mr. Kapila Bandara, Dr. Hung Duong and Mr.
Jalil Yaghoobi, who provided me with their help and assistance.

Furthermore, I am grateful to all my family members, especially my husband Aravind Kumar,

for his continuous support and encouragement to strive towards my goals. This thesis would
not have been possible without his constant love and support. I would like to dedicate this thesis
to him.

xi
Keywords

Solar PV system, Power quality, Unbalanced distribution system, Voltage quality, Harmonic
impact, Resonance, Solar-DSTATCOM, Controller hardware-in-the-loop simulation, RTDS,
dSPACE.

Australian and New Zealand Standard Research Classifications (ANZSRC)

ANZSRC code: 090607, Power and Energy Systems Engineering (excl. Renewable Power),
70%

ANZSRC code: 090608, Renewable Power and Energy Systems Engineering (excl. Solar
Cells), 30%

Fields of Research (FoR) Classification

FoR code: 0906, Electrical and Electronic Engineering, 100%

xii
Table of Contents
ABSTRACT…………………………………………………………………………… ii

TABLE OF CONTENTS……………………………………………………………….. xiii

LIST OF FIGURES…………………………………………………………………….. xviii

LIST OF TABLES…………………………………………………………………....... xxiv

ABBREVIATIONS AND ACRONYMS ………………………………………………….. xxvi

1. INTRODUCTION…………………………………………………………………. 1

1.1 BACKGROUND…………………………………………………………….. 1

1.2 RESEARCH OBJECTIVES…………………………………………………… 3

1.3 THESIS OUTLINE…………………………………………………………... 4

2. LITERATURE REVIEW…………………………………………………………... 6

2.1 INTRODUCTION……………………………………………………………. 6

2.2 GROWTH OF SOLAR PV TECHNOLOGY AND ITS STATISTICS………………. 6

2.3 POWER QUALITY ISSUES WITH SOLAR PV SYSTEM INTEGRATIONS……….. 8

2.3.1 Voltage Rise……………………………………………………...... 8

2.3.2 Voltage Fluctuations and Flicker………………………………...…. 9

2.3.3 Voltage Unbalance Factor………………………………………….. 9

2.3.4 Voltage Sag/Swell………………………………………………….. 10

2.3.5 Harmonic Emission Issues from PV Systems and their Impacts………. 10

2.3.6 Possibility of Harmonic Resonance Incidences………………………. 12

2.3.7 Other Issues……………………………………………………….. 13

2.4 SOLUTIONS FOR POWER QUALITY IMPROVEMENT…………………………. 13

2.5 OVERVIEW OF STANDARDS ………………………………………………. 15

2.5.1 Voltage Quality Standards…………………………………………. 16

2.5.2 Harmonic Standards……………………………………………….. 18

2.6 SUMMARY………………………………………………………………… 19

xiii
3. MODELLING AND SIMULATION TOOLS……………………………………..….. 20

3.1 INTRODUCTION……………………………………………………………. 20

3.2 GRID CONNECTED SOLAR PV SYSTEM……………………………………. 20

3.2.1 Modelling of Solar PV Source………………………………………. 21

3.2.2 DC/DC Converter………………………………………………….. 25

3.2.3 Maximum Power Point Tracking Algorithms………………………… 26

3.2.3.1 Perturb and Observe (P&O) Algorithm…………………….. 26

3.2.3.2 Incremental Conductance Algorithm……………………….. 28

3.2.4 PV Inverter (DC/AC) Types and Control Topologies………………… 30

3.2.4.1 Current Regulated Current Source Inverter…………………. 31

3.2.4.2 Current Controlled Voltage Source Inverter………………... 33

3.3 IEEE-13 BUS UNBALANCED DISTRIBUTION NETWORK………………….... 34

3.4 LOAD MODELLING…………………………...…………………………… 36

3.5 SOFTWARE AND HARDWARE TOOLS………………………………………. 40

3.5.1 PSCAD……………………………………………………………. 40

3.5.2 Real Time Digital Simulator (RTDS)………………………………… 40

3.5.3 dSPACE Control Board Platform…………………………………… 41

3.6 SUMMARY………………………………………………………………… 42

4. INVESTIGATION OF VOLTAGE QUALITY ISSUES………………………………. 43

4.1 INTRODUCTION…………………………………………………………… 43

4.2 ASSESSMENT OF VOLTAGE QUALITY ISSUES IN IEEE-13 BUS UNBALANCED 44

DISTRIBUTION NETWORK WITH HIGH PV PENETRATION…………………..
4.2.1 Case Study – I: Voltage Rise and Unbalance Assessment…………….. 44

4.2.2 Case study-II: Voltage Sag/Swell Assessment by Fault Conditions……. 48

4.3 FIELD INVESTIGATION OF VOLTAGE QUALITY ISSUES IN DISTRIBUTION 52

NETWORK WITH PV PENETRATION………………………………………..
4.3.1 Description of University of Queensland, St. Lucia Campus Network and 53
Measurement Setup…………………………………………………
4.3.1.1 Results and Analysis………………………………………….. 56

xiv
 4.3.1.1.1 Characteristics of Voltage Rise and Fluctuations at 56
1.5MW PV System………………………………..
4.3.1.1.2 Voltage Flicker Analysis at 1.5MW PV System…. 58

4.3.2 Field Measurements at the University of Queensland 3.3 MW PV Plant, 65

Gatton Campus…………………………………………………….
4.3.2.1 Results and Analysis……………………………………… 66

4.3.2.1.1 Voltage Rise and Fluctuations Assessment at 66

3.3MW PV System…………………………….
4.3.2.1.2 Voltage Flicker Analysis at 3.3MW PV System…. 68

4.4 SUMMARY………………………………………………………………… 70

5. ANALYSIS OF HARMONIC ISSUES WITH PV SYSTEMS……………………….… 72

5.1 INTRODUCTION……………………………………………………………. 72

5 5.2 TEST SYSTEM APPROACH FOR THE HARMONIC ANALYSIS………………… 73

5.3 HARMONIC CHARACTERISATION OF GRID CONNECTED PV SYSTEMS & 75

VALIDATION WITH FIELD MEASUREMENTS………………………………..
5.3.1 Simulation Approach and Analysis of Results………………………... 75

5.3.2 Field Measurement Approach and Analysis of Results……………….. 79

5.4 HARMONIC IMPACT OF HIGH PENETRATION PV SYSTEM ON UNBALANCED 84

DISTRIBUTION NETWORKS - LEARNING FROM AN URBAN PV NETWORK….
5.4.1 Simulation Approach and Analysis of Results……………………….. 85

5.4.1.1 Case Study I: Without Distortion in the Supply with PV 86

Penetration at the Same Bus…………………………………..
5.4.1.2 Case Study II: With Distortion in the Supply and Multiple PV 89
Penetration at the Same Bus…………………………………..
5.4.1.3 Case Study III: With Distortion in the Supply and Multiple PV 92
Penetrations at Multiple Locations……………………………
5.4.2 Field Measurements at the University of Queensland - Analysis of 93
Results……………………………………………………………..
5.5 SUMMARY………………………………………………………………… 100

6. HARMONIC RESONANCE ISSUES WITH PV PENETRATION…………………….. 102

6.1 INTRODUCTION……………………………………………………………. 102

xv
 6.2 HARMONIC RESONANCE MECHANISM ASSOCIATED WITH PV SYSTEMS….. 103

6.2.1 Impact of Grid Connected PV Inverter Control Strategies……………… 105

6.3 TEST SYSTEM APPROACH AND SIMULATION STUDY………………………. 107

6.4 RESULTS AND DISCUSSION……………………………………………… 108

6.4.1 Simulation Study on IEEE-13 Bus Distribution Network………………. 108

6.4.1.1 Case Study-I: Without Background Distortions in the 109

Network and with PV Systems………………………………
6.4.1.2 Case Study-II: With Background Distortions in the Network 113
and with PV Systems………………………………………
6.4.2 Analysis of Harmonic Measurements at The University of Queensland 117
PV Site……………………………………………………………..
6.5 SUMMARY………………………………………………………………… 123

7. SOLAR-DSTATCOM: A NEW SOLUTION TO POOR POWER QUALITY ……….. 124

7.1 INTRODUCTION……………………………………………………………. 124

7.2 MOTIVATION…………………………………………………………….... 125

7.3 SOLAR-DSTATCOM CONTROL STRATEGY……………………………… 127

7.3.1 DC Voltage Controller....................................................................... 127

7.3.2 PCC Voltage Controller…………………………………………..... 128

7.3.3 Load Reactive Power and Current Harmonic Compensation Controller.. 128

7.3.4 Hysteresis Controller…………………………………………………… 132

7.4 POWER QUALITY ENHANCEMENT IN IEEE-13 BUS UNBALANCED 132

NETWORK USING SOLAR-DSTATCOM…………………………………...
7.4.1 Case Study-I: Voltage Rise and Voltage Unbalance Assessment……… 133

7.4.2 Case Study-II: Voltage Sag/Swell Assessment by Fault Conditions…… 135

7.4.3 Case Study-III: Harmonic Distortion Analysis………………………… 139

7.5 REAL-TIME CONTROLLER HARDWARE-IN-THE-LOOP SIMULATIONS OF 140

SOLAR-DSTATCOM……………………………………………………...
7.5.1 Controller Hardware-in-the-loop (HIL) Simulation Environment……… 141

7.5.2 Test System and Solar-DSTATCOM Controller Setup in HIL Platform… 142

7.5.2.1 Solar-DSTATCOM Control Setup in MATLAB/Simulink… 144

xvi
 7.5.3 Results and Discussion……………………………………………. 147

7.5.3.1 Case study–I: Performance of Solar-DSTATCOM for Various 147

Daytime PV Profile Variations………………………………
7.5.3.2 Case study–II: Performance of Solar-DSTATCOM during 153
Night-time Operation………………………………………...
7.5.3.3 Case Study-III: Performance of Solar-DSTATCOM at Various 154
PV Power Levels and Load Profile Variations………………
7.5.3.4 Case Study-IV: Harmonic Distortion Analysis……………… 162

7.6 SUMMARY………………………………………………………………… 163

8. CONCLUSIONS AND FUTURE WORKS…………………………………………... 165

8.1 CONCLUSIONS…………………………………………………………….. 165

8.2 SUMMARY OF CONTRIBUTIONS…………………………………………… 172

8.3 FUTURE WORKS…………………………………………………………... 174

REFERENCES………………………………………………………………………… 175

APPENDIX A: IEEE-13 BUS TEST SYSTEM CONFIGURATION DATA………………. 186

xvii
List of Figures
Figure 1.1 Solar PV Global Capacity from 2004 to 2014 [2]……………………... 2
Figure 2.1 Grid Connected PV Uptake in Australia……………………………….. 7
Figure 3.1. Block Diagram of Grid Connected PV System………………………... 21
Figure 3.2. Solar PV Cell Equivalent Circuit………………………………………. 21
Figure 3.3. I-V Characteristic of Solar PV Cell [111] ……………………………... 21
Figure 3.4. Equivalent Circuit of a PV Module/Array [111]..……………………... 22
Figure 3.5. Simulation Model of PV Module/Array……………………………….. 24
Figure 3.6. I-V Characteristics of PV Array at Various Solar Irradiation
Conditions……………………………………………………………… 24
Figure 3.7. P-V Characteristics of PV Array at Various Solar Irradiation
Conditions……………………………………………………………… 25
Figure 3.8. Boost Converter Circuit……………………………………………….... 25
Figure 3.9. PV Panel P-V Characteristic Curve…………………………………….. 26
Figure 3.10. The Flow Chart of the P&O Algorithm………………………………... 27
Figure 3.11. PSCAD Model for P&O Algorithm……………………………………. 28
Figure 3.12. Flow Chart of the Incremental Conductance Algorithm……………….. 29
Figure 3.13. PSCAD Model for Incremental Conductance Algorithm……………… 30
Figure 3.14. (a) Single-phase and (b) Three-phase PV Inverter Systems with Filter.. 31
Figure 3.15. Control Circuit of Current Regulated Current Source Inverter………... 32
Figure 3.16. Control Circuit of Current Controlled Voltage Source Inverter……….. 33
Figure 3.17. IEEE-13 Bus Test System Detailed Line Diagram…………………….. 35
Figure 3.18. Simulation Model of IEEE-13 Bus Distribution System………………. 35
Figure 3.19. Simulation Design of Constant Power (PQ) Load……………………... 36
Figure 3.20. Simulation Design of Constant Current Load………………………….. 37
Figure 3.21. Single-phase Power Electronic based Nonlinear Loads……………….. 38
Figure 3.22. Three-Phase Power Electronic based Nonlinear Loads………………... 39
Figure 3.23. Simulation Model of Single-phase Induction Motor Load…………….. 40
Figure 3.24. Block Diagram of dSPACE Hardware Setup………………………….. 41
Figure 4.1. IEEE 13 Bus Test System……………………………………………… 44
Figure 4.2. Solar Irradiation Pattern and PV Power Output………………………... 45
Figure 4.3. %Voltage Unbalance Analysis at Various Buses……………………….. 48
Figure 4.4. RMS Voltages at Prefault and during SLG Fault without PV Inverter… 50

xviii
Figure 4.5. RMS Voltages at Prefault and during SLG Fault with PV Inverter……. 50
Figure 4.6. RMS Voltages at Prefault and during DLG Fault without PV Inverter… 51
Figure 4.7. RMS Voltages at Prefault and during DLG Fault with PV Inverter……. 51
Figure 4.8. RMS Voltages at Prefault and during TPG Fault without PV Inverter… 52
Figure 4.9. RMS Voltages at Prefault and during TPG Fault with PV Inverter……. 52
Figure 4.10. Overview of the University of Queensland PV Network, St. Lucia
Campus………………………………………………………………….. 54
Figure 4.11. Measurement Test Setup……………………………………………….. 55
Figure 4.12. Time versus Solar Irradiation and PV Output Power………………….. 56
Figure 4.13. Voltage Magnitudes at the PV Inverter Output………………………... 57
Figure 4.14. Voltage Magnitudes at Substation 11 Transformer TX1………………. 57
Figure 4.15. Voltage Magnitudes at Substation 17………………………………….. 57
Figure 4.16. Measured PV Inverter Output Currents………………………………... 58
Figure 4.17. Measured Voltage Flicker Indices at the PV Inverter Output………….. 59
Figure 4.18. Measured Flicker Indices at 11kV Line………………………………... 60
Figure 4.19. Time versus Solar Irradiation and PV Output Power for Typical Days.. 61
Figure 4.20. Measured Voltage Flicker Indices at the PV Array……………………. 61
Figure 4.21. Clustered based Probability Count of Flicker Indices at PV Array…….. 63
Figure 4.22. Clustered based Probability Count of Flicker indices at 11kV Line…... 64
Figure 4.23. Overview of UQ Gatton PV Plant……………………………………... 65
Figure 4.24. Time versus PV Power Output during Various Operating Days………. 67
Figure 4.25. Measured Voltage Magnitude Profile at the LV Side of the PV Inverter
Output………………………………………………………………….. 67
Figure 4.26. Measured Voltage Magnitude Profile at the HV Side of the PV
Integration Point………………………………………………………... 68
Figure 4.27. Probability Count of Short-term Flicker at the PV Interconnection
Node……………………………………………………………………. 69
Figure 4.28. Probability Count of Long-term Flicker at the PV Interconnection
Node……………………………………………………………………. 69
Figure 5.1. IEEE-13 Bus Test System……………………………………………… 73
Figure 5.2. Time versus PV Output Active and Reactive Powers and Solar
Irradiation……………………………………………………………….. 76
Figure 5.3. Voltage Harmonics at the Output of PV Inverter in Each Scenario…… 77

xix
Figure 5.4. Current Harmonics at the Output of PV inverter in Each Scenario……. 78
Figure 5.5. Voltage Harmonics at the Output of PV Inverter in Each scenario with
Distortion in the Supply………………………………………………... 78
Figure 5.6. Percentage of THDs of PV Inverter Output Currents in Two Case
Studies………………………………………………………………….. 79
Figure 5.7. Grid Connected Three-phase PV Inverters at UQ PV Site…………….. 80
Figure 5.8. Measurement Test Setup for Harmonic Analysis……………………… 80
Figure 5.9. Time versus PV Output Power and Sunlight…………………………… 81
Figure 5.10. Individual Voltage Harmonics of Phase-B at the PV Systems Output… 82
Figure 5.11. Individual Current Harmonics of Phase-B at the PV Systems Output… 83
Figure 5.12. Percentage of Current THDs of PV Inverter System…………………... 84
Figure 5.13. Overview of UQ Sub Station 11……………………………………….. 94
Figure 5.14. Solar PV System Measurements at UQ for Typical Operating Days….. 96
Figure 5.15. Harmonic Measurements at the Distribution Transformer TX2 for the
Selected Days…………………………………………………………… 98
Figure 5.16. K-Factor Measurements for the Corresponding Days at the
Transformer…………………………………………………………...… 99
Figure 5.17. Harmonic Measurements at the HV Line……………………………… 100
Figure 6.1. Single-phase Equivalent of Grid with PV System Connected at PCC… 104
Figure 6.2. Detailed Three-phase PV System Configuration………………………. 105
Figure 6.3. Control Circuit of Current Controlled Voltage Source Inverter………... 106
Figure 6.4. Control Circuit of Current Regulated Current Source Inverter………... 107
Figure 6.5. Time versus Solar Irradiation and PV Output Pattern…………………. 108
Figure 6.6. Percentage of Voltage THDs at Various Nodes for Phases A and B with
CRCSI………………………………………………………………….. 109
Figure 6.7. Percentage of Current THDs at Various Nodes for Phases A and B with
CRCSI………………………………………………………………….. 110
Figure 6.8. Impedance versus Frequency Scan with and without CRCSI PV
Systems at PCC………………………………………………………… 112
Figure 6.9. Percentage of Voltage THDs at Various Nodes for Phases A and B with
CCVSI………………………………………………………………….. 115
Figure 6.10. Percentage of Current THDs at Various Nodes for Phases A and B with
CCVSI………………………………………………………………….. 115

xx
Figure 6.11. Impedance versus Frequency Scan with and without CCVSI PV 116
Systems at PCC…………………………………………………………
Figure 6.12. Time versus PV Output Power and Sunlight…………………………... 117
Figure 6.13. Percentage of Voltage THD at HV Side for Parallel Resonance
Condition……………………………………………………………... 118
Figure 6.14. Percentage of Voltage THD at the Distribution Transformer TX2 for
Parallel Resonance Condition………………………………………….. 119
Figure 6.15. Percentage of Current THD at the Distribution Transformer TX2 for
Parallel Resonance Condition…………………………………………... 119
Figure 6.16. Percentage of Voltage THD at the PV Array Output for Parallel
Resonance Condition………………………………………………….... 120
Figure 6.17. Percentage of Current THD at the PV Array Output for Parallel
Resonance Condition…………………………………………………... 120
Figure 6.18. Time versus PV Output Power and Sunlight…………………………... 121
Figure 6.19. Percentage of Current THD at HV Side for Series Resonance
Condition……………………………………………………………….. 121
Figure 6.20. Percentage of Current THD at the Distribution Transformer for Series
Resonance Condition…………………………………………………... 122
Figure 6.21. Percentage of Voltage THD at the Distribution Transformer for Series
Resonance Condition…………………………………………………... 122
Figure 7.1. Block Diagram of Three-phase Grid Connected Solar-DSTATCOM
System…………………………………………………………….……. 127
Figure 7.2. PSCAD Model of DC Bus Voltage Regulation Block………………… 128
Figure 7.3. PSCAD Model of PCC Voltage Regulation Block…………………….. 128
Figure 7.4. PSCAD Model of Reactive Power and Current Harmonic Compensation
Controller………………………………………………………………. 129
Figure 7.5. Single-phase PLL Structure……………………………………………. 131
Figure 7.6. PSCAD Model of Hysteresis Controller……………………………….. 132
Figure 7.7. Static Solar Irradiation Pattern…………………………………………. 133
Figure 7.8. Active and Reactive Powers of a 900kW Solar-DSTATCOM………… 133
Figure 7.9. %Voltage Unbalance Analysis at Various Buses……………………… 135
Figure 7.10. Solar-DSTATCOM (a) Inverter Currents and (b) Compensated PCC
Voltages during SLG Fault………………………..……………………. 136

xxi
Figure 7.11. The RMS Voltages at Prefault and during SLG Fault with Solar-
DSTATCOM…………………………………………………………... 136
Figure 7.12. Solar-DSTATCOM (a) Inverter Currents (b) Compensated PCC
Voltages during DLG Fault………………………………..………….. 137
Figure 7.13. The RMS Voltages at Prefault and during DLG Fault with Solar-
DSTATCOM………………………………………….……………… 137
Figure 7.14. Solar-DSTATCOM (a) Inverter Currents (b) Compensated PCC
Voltages during TPG Fault…………………………………………….. 138
Figure 7.15. The RMS Voltages at Prefault and during TPG Fault with Solar-
DSTATCOM…………………………………………………………… 138
Figure 7.16. Percentage of THDs of Solar-DSTATCOM Output Currents…………. 139
Figure 7.17. Percentage of THDs of Solar-DSTATCOM Output Voltages………… 139
Figure 7.18. Individual Harmonic Distortions of Solar-DSTATCOM Output
Currents and PCC Voltages……………………………………………. 140
Figure 7.19. Experimental Setup of Hardware-in-the-Loop using RTDS and
dSPACE Control Board………………………………………………... 142
Figure 7.20. Test System and Solar-DSTATCOM Modelling in RSCAD
Environment……………………………………………………………. 143
Figure 7.21. (a) DC Bus Voltage Controller and (b) PCC Voltage Controllers……... 145
Figure 7.22. Load Reactive Power and Current Harmonic Compensation Block…... 146
Figure 7.23. Time versus Solar Irradiation of a Sunny Day…………………………. 148
Figure 7.24. Measured Active and Reactive Power Profiles at Solar-DSTATCOM
Output and Grid Supply during a Sunny Day………………………….. 148
Figure 7.25. Measured PCC RMS Voltages and Solar-DSTATCOM Output Currents
during a Sunny Day………………………………………….................. 149
Figure 7.26. DC Bus Voltage Profile during a Sunny Day………………………….. 149
Figure 7.27. Solar-DSTATCOM Controller Various Output Parameters during a
Sunny Day Operation…………………………………………………... 150
Figure 7.28. Time versus Solar Irradiation of a Cloudy Day………………………... 151
Figure 7.29. Measured Active and Reactive Power Profiles at Solar-DSTATCOM
Output and Grid Supply during a Cloudy Day………………………… 151
Figure 7.30. Measured PCC RMS Voltages and Solar-DSTATCOM Output Currents
during a Cloudy Day…………………………………………………… 152

xxii
Figure 7.31. DC bus Voltage Profile during a Cloudy Day…………………………. 152
Figure 7.32. Solar-DSTATCOM Controller Various Output Parameters during a
Cloudy Day Operation…………………………………………………. 153
Figure 7.33. Various Operating Parameters of the Solar-DSTATCOM and Grid
during the Night-time…………………………………………………... 154
Figure 7.34. Load Active and Reactive Power and PV Power Input Profiles for First
Scenario……………………………………………………………… 155
Figure 7.35. Active and Reactive Powers at Solar-DSTATCOM Output, Grid and
Load during the First Load Scenario…………………………………... 156
Figure 7.36. Calculated P.F of Solar-DSTATCOM, Grid, and Load……………….. 156
Figure 7.37. Measured PCC RMS Voltages during the First Load Scenario……….. 157
Figure 7.38. Solar-DSTATCOM Controller Various Output Parameters during the
First Load Scenario……………………………………………………. 158
Figure 7.39. Load Active and Reactive Power and PV Power Input Profiles for
Second Scenario………………………………………………………... 159
Figure 7.40. Active and Reactive Powers at Solar-DSTATCOM Output, Grid and
Load during the Second Load Scenario……………………………….. 160
Figure 7.41. Calculated P.F of Solar-DSTATCOM, Grid, and Load……………….. 160
Figure 7.42. Measured PCC RMS Voltages during the Second Load Scenario…….. 161
Figure 7.43. Solar-DSTATCOM Controller Various Output Parameters during the
Second Load Scenario…………………………………………………. 162
Figure 7.44. Voltage and Current THDs of Solar-DSTATCOM…………………….
Figure 7.45. Solar-DSTATCOM Individual Harmonic Distortion of Voltage and 163
Current…………………………………………………………………. 163

xxiii
List of Tables

Table 2.1 Harmonic Limits of Grid Connected PV Inverter Requirement……….……

19
Table 4.1 Scenario-1: RMS Voltage Magnitude Variation due to PV System
Integration at Bus 632 during Daytime and Night-time…………………. 45
Table 4.2 Scenario-2: RMS Voltage Magnitude Variation due to PV Systems
Integration at Buses 634, 671 and 675 during Daytime and Night-time…… 46
Table 4.3 Scenario-3: RMS Voltage Magnitude Variation due to the Combination of
Three-phase and Single-phase PV Systems Integration at Multiple
Locations during Daytime and Night-time………………………………… 47
Table 5.1 Percentage of Voltage THDs at Various Buses for Phases A and B with
CRCSI…………………………………………………………………….. 87
Table 5.2 Percentage of Current THDs at Various Buses for Phases A and B with
CRCSI…………………………………………………………………….. 87
Table 5.3 Percentage of Voltage THDs at Various Buses for Phases A and B with
CCVSI…………………………………………………………………….. 88
Table 5.4 Percentage of Current THDs at Various Buses for Phases A and B with
CCVSI…………………………………………………………………….. 88
Table 5.5 Percentage of Voltage THDs at Various Buses for Phases A and B with
CRCSI and Distortion in the Network…………………………………….. 89
Table 5.6 Percentage of Current THDs at Various Buses for Phases A and B with
CRCSI and Distortion in the Network…………………………………….. 90
Table 5.7 Percentage of Voltage THDs at Various Buses for Phases A and B with
CCVSI and Distortions in the Network…………………………………… 91
Table 5.8 Percentage of Current THDs at Various Buses for Phases A and B with
CCVSI and Distortion in the Network…………………………………….. 92
Table 5.9 Percentage of Voltage and Current THDs with 5 PV Systems (CRCSI) at
Multiple Locations…………………………………………………...…… 92
Table 5.10 Percentage of Voltage and Current THDs with 5 PV Systems (CCVSI) at
Multiple Locations………………………………………………………... 93
Table 7.1 Voltage Variations during Daytime and Night-time with Solar-
DSTATCOM……………………………………………………..………. 134

xxiv
Table A.1 Overhead Line Configuration Data……………………………………….. 186
Table A.2 Underground Line Configuration Data……………………..…………….. 186
Table A.3 Line Impedance Matrices…………………………………………………. 186
Table A.4 Line Segment Data………………………………………………………... 187
Table A.5 Transformer Data…………………………………………………………. 187
Table A.6 Capacitor Bank Data……………………………………………………… 187
Table A.7 Regulator Data…………………………………………………..………… 188
Table A.8 Regulator Data……………………………………………………………. 188
Table A.9 Distributed Load Data……………………………………………..……… 188

xxv
Abbreviations and Acronyms

PV Photovoltaic
DG Distributed Generation
REN Renewable Energy
AEMO Australian Energy Market Operator
IEEFA Institute for Energy Economics and Financial Analysis
NEM National Electricity Market
LV Low Voltage
HV High Voltage
OLTC On-Load Tap Changer
SVR Step Voltage Regulator
MPPT Maximum Power Point Tracking
THD Total Harmonic Distortion
TDD Total Demand Distortion
ESS Energy Storage System
CPD Custom Power Devices
DSTATCOM Distributed Static Compensator
DVR Dynamic Voltage Restorer
IEEE Institute of Electrical and Electronics Engineering
ANSI American National Standards Institute
AS Australian Standard
AS/NZS Australian/New Zealand Standard
IEC International Electrotechnical Commission
EN European
STC Standard Test Condition
P&O Perturb and Observe
PWM Pulse Width Modulation
PLL Phase Locked Loop
CRCSI Current Regulated Current Source Inverter
CCVSI Current Controlled Voltage Source Inverter
PC Personal Computer
CFL Compact Fluorescent Lamp

xxvi
ASD Adjustable Speed Drive
TV Television
PSCAD Power Systems Computer Aided Design
dSPACE Digital Signal Processing and Control Engineering
RTDS Real Time Digital Simulator
GTAI Gigabit Transceiver Analog Input
GTAO Gigabit Transceiver Analog Output
ADC Analog to Digital
DAC Digital to Analog
VU Voltage Unbalance
PCC Point of Common Coupling
P.F Power Factor
SLG Single Line to Ground
DLG Double Line to Ground
TPG Three Phase to Ground
PQ Power Quality
PI Proportional Integrator
UQ University of Queensland
QLD Queensland

xxvii
 Chapter 1. Introduction

CHAPTER 1
INTRODUCTION

1.1. Background

The world’s population is increasing, economies are developing and global energy
consumption is increasing faster than ever before, particularly in developing countries. Experts
predict that the demand for electricity will have increased by 45% in 2035 [1]. The present
energy requirement is mainly supplied by combustion of fossil fuels such as crude oil, coal and
natural gas. However, conventional energy sources are inadequate in meeting the rising
electricity demand. Moreover, use of fossil fuels in generating electricity creates environmental
problems mainly due to emissions of greenhouse gases. Global climate change, substantial
rising electricity demand and increasing electricity prices have increased concerns all over the
world. Therefore, many nations around the world are pushing towards the deployment of
renewable energy sources to ensure a secure, affordable, environmentally clean and sustainable
power supply. These resources can help in creating new economic opportunities and providing
access to people living without electricity. Renewable energy sources such as solar, wind,
hydro and geothermal are expected to supply a considerable percentage of future world
electricity requirements. As per REN21-2015 global status report [2], renewable energy
provided approximately 19.1% of global energy consumption in 2013. By the end of 2014, the
estimated renewable power production has increased to 22.8% of the global power generating
capacity and the capacity of total renewable power installation has reached 657GW (excluding
hydropower).

1
 Chapter 1. Introduction

During recent decades, renewable energy based distributed generation (DG) has been
gaining momentum rapidly due to their feasibility. Among several DG technologies, Solar
Photovoltaic (PV) systems are one of the fastest growing resources worldwide due to
decreasing capital costs and increasing technological advancements. Out of total renewable
power capacity (657GW), the capacity of solar PV installations worldwide had exceeded
177GW by the end of 2014, as shown in Fig. 1.1 [2, 3]. Australia is the ideal location for solar
power production because of the high insolation level. The forecasts from Australian Energy
Market Operator (AEMO) [4] show that Australia’s energy dependence heavily on rooftop
solar PV systems in the coming years. In 2030, the cumulative solar PV installed capacity could
contribute 21% of the Australia’s total electricity needs.

Figure 1.1. Solar PV Global Capacity from 2004 to 2014 [2].

Solar PV systems can be connected to both transmission and distribution networks. The
application of solar PV systems in power systems are categorized as grid connected and off-
grid. Usually, off-grid PV systems can be used to provide power to remote loads, which do not
have access to the power network. Grid-connected solar PV systems can be installed at a
location near to the load for power generation, which eliminates the transmission and
distribution costs. Recent expansion in rooftop solar PV technology and the potential benefits
from government subsidies have promoted installation of systems with capacities ranging from
a few kW to 10MW in residential, commercial and industrial applications. In addition, PV
system installed prices have gradually been decreasing over recent years, so the average size
of installations has been increasing. When compared to other forms of DG generation, solar
PV systems are physically small and have lower voltages, which are therefore appropriate to
interface with the low voltage distribution network. However, there are several concerns in
integrating this technology in low voltage (LV) distribution networks.

2
 Chapter 1. Introduction

Solar PV system integration can have positive aspects on the low voltage distribution
network to feed peak energy demands, to reduce the energy losses of distribution feeders and
to provide voltage support. However, the exponential uptake in widespread installations of
solar PV systems can have negative impacts on the existing distribution network, which can
inflict various power quality, stability and protection challenges. Power quality is the major
issue of concern in distribution networks, which covers various operational parameters and
plays a key role in industrial, commercial and residential power needs. The most frequent
power quality issues in existing distribution networks are voltage quality problems such as
unbalance, sags and swells.

In fact, existing power dispatch systems are passive networks and are built to distribute
power from large generating plants to the end-user in a single direction. Traditionally, low
voltage distribution networks possess unbalanced characteristics due to the unbalanced line
configurations and phase loadings. Also, these networks are not designed for two-way power
flow. Nevertheless, with the distributed integration of solar PV systems into the grid, the
networks become active and the customers can export and import power to the grid, which
could impact the network voltage profile in terms of voltage rise and unbalance. PV generators
are often owned and operated by customers, therefore the location and size of PV system
installations is dependent on their financial status and interest. Moreover, this type of resource
is intermittent in nature due to seasonal effects and limited predictability results in power
imbalance on the utility grid leads to power quality degradation, which has implications on
voltage regulation issues. Rapid variations in the solar irradiation due to cloud transients can
cause voltage fluctuations/flicker issues and the incidences may exacerbate in the network if
PV penetration increases. Furthermore, the solar PV system uses a power electronic based
inverter to convert DC electricity from PV panels to AC electricity, which can inject harmonics
into the network. Consequently, utility operators with current knowledge will face difficulties
in getting control over distribution network performance. Therefore, it is significantly
important to investigate and analyse the potential issues with high PV system penetration in
the existing distribution networks and to provide a potential solution for improving grid power
quality.

1.2. Research Objectives

The main objective of this thesis is to investigate various power quality issues imposed by
high penetration of solar PV systems in the utility grid and aiming to alleviate them. Following

3
 Chapter 1. Introduction

are the research objectives, which will be accomplished in this research for enabling high
penetration of rooftop PV in low voltage distribution network.

1. To assess voltage related issues in the low voltage unbalanced distribution network
with high PV penetration through simulations.
2. To investigate voltage quality problems in real grid connected PV sites through
field measurements. Specifically, to study and quantify the impact of short-term
solar radiation variations on voltage flicker emissions.
3. To examine the harmonic emission characteristics of solar PV inverters and their
aggregation issues for different operating conditions.
4. To investigate and analyse the harmonic impact of high penetration PV systems on
an unbalanced distribution network comprehensively.
5. To investigate harmonic resonance issues associated with different PV inverter
technologies and their dynamic operational interactions in the network.
6. To propose, develop and validate a new control methodology for solar PV inverters
to enhance grid power quality in offline simulation and Real Time Digital
Simulation environment.

1.3. Thesis Outline

This thesis is organised as follows:

Chapter 2 reviews the power quality impacts of PV system integration in the literature.
Firstly, this chapter introduces the growth of solar PV technology in different countries and
provides current and future statistic impacts on the distribution networks. After that, a detailed
literature review of various power quality issues and identified research gaps are discussed.
Then, traditional solutions for power quality improvement are reviewed. Finally, various
regulatory standards and current practices related to power quality criteria are introduced.

Chapter 3 provides the background theory, models and tools used in this research. To
simulate real PV system dynamic characteristics, detailed modelling of solar PV sources using
a mathematical model, DC-to-DC converter, maximum power point tracking techniques and
inverter technologies are introduced first. Following that, the test system and the load
modelling requirements are presented. Finally, utilized various software and hardware tools for
this research are explained.

Chapter 4 presents a comprehensive investigation of voltage quality issues introduced by

PV system integration in the distribution network. Primarily, the assessment on the unbalanced

4
 Chapter 1. Introduction

test system through simulation scenarios is presented. Further, to identify the proliferation of
various voltage quality issues in real-time network, field investigations on the real grid
connected PV sites have been performed, which are presented at the end.

Chapter 5 assess the characteristics of harmonic emissions from PV inverters and their
aggregations during various operating conditions. The analysis of simulation results in different
case studies and the validation with field measurement data are presented. Apart from that, a
comprehensive analysis has been presented in several different cases studies with high
penetration of different PV inverter technologies to evaluate the severity of harmonic
propagation on the various nodes of the distribution network. Similarly, comparative studies
of simulations and real-time network harmonic measurements are discussed.

Chapter 6 investigate the possibilities of harmonic resonance issues on the distribution

network with PV systems. At first, the resonance phenomena associated with PV system is
explained. After that, the simulation approach and the results of different case studies are
presented. Further, harmonic resonance issues on a real grid connected PV site identified
through network harmonic distortion measurements are presented.

Chapter 7 introduces a new solution to prevail the above mentioned power quality issues.
The concept of adopting a PV inverter as a Distributed Static Compensator (DSTATCOM)
named as Solar-DSTATCOM has been proposed. At first, the proposed new control
methodology and the principle of operation are explained. The effect of Solar-DSTATCOM
on the unbalanced test network is presented. Finally, validation of the proposed Solar-
DSTATCOM control system in a controller hardware-in-the-loop simulation environment is
presented.

Chapter 8 concludes the work done for this research and highlights the main findings. The
possible future direction of further research is also presented in this chapter.

5
 Chapter 2. Literature Review

CHAPTER 2

LITERATURE REVIEW

2.1. Introduction

In this chapter, firstly, an overview of solar PV technology evolution in different countries

and their penetration statistics on the distribution networks are reported. After that, a
comprehensive literature review on the power quality impacts of solar PV penetration in the
distribution networks and the identified gaps are discussed. Further, some countermeasures for
power quality improvement in the current literature are provided. In the end, various standards
and grid codes of different countries related to the power quality criteria are referred. This
chapter partially incorporates some of the literature reviews from the PhD candidate’s
published and submitted papers [5-11].

2.2. Growth of Solar PV Technology and its Statistics

Solar Photovoltaic (PV) power generation has become a promising technology for
integration into distribution networks due to their feasibility and availability of sun compared
to other types of renewable energy resources. The installed capacity of small scale and utility
scale grid connected PV system installations has grown significantly over the past five years.
In the U.S., the solar PV installation capacity has reached a maximum of 7.3 GW in 2015. With
this, the total cumulative installation capacity reached over 25GW, which was supplying a total

6
 Chapter 2. Literature Review

of 29.5% of all new electricity generation capacity [12]. This trend will continue to grow in the
coming years. India has set a target to install 175GW of renewable power generation by 2021-
22 [13], which is the highest target in the world. As per the Institute for Energy Economics and
Financial Analysis (IEEFA) report, 100GW of power is expected from solar resources. In
which, a nearly 40GW target has been set for distributed rooftop solar systems as a way to
decrease the nation’s grid losses and unpredictable regular power outages [13].

In Australia the importance for solar PV research programs are increasing and are aiming to
increase renewable penetration to meet future energy demand. Recently, a new target of
renewable power generation of 45,000GWh in 2020 has been set, which represents 25% of
total electricity generation in 2020 [14]. This renewable energy target is now encouraging the
small-scale renewable energy scheme, which promotes installation of rooftop PV systems and
is expected to contribute not less than 12000GWh. According to the Australian PV Institute
report, the potential installed PV system capacity has exceeded nearly 4.7GW by the end of
2015, as shown in Fig 2.1 [15].

5000 Australian PV installations since April 2001:

total capacity (kW)
Pv Installed capacity (kW)

4000

3000 Reported installed capacity

2000 Estimated installed capacity

1000

Source: http://pv-map.apvi.org.au/analyses

Figure 2.1. Grid Connected PV Uptake in Australia [15].

The forecasts conducted by the Australian Energy Market Operator (AEMO) for each
National Electricity Market (NEM) region [4] shows that the rooftop PV system installed
capacity in 2031 would increase up to 12 times higher than the present capacity level.
Nonetheless, the estimated rooftop PV installed capacity can impose severe impacts on the
existing distribution network, which has led to a new set of challenges in power quality,
stability and protection. The Energex distribution Annual planning report stated that the current
level of rooftop PV penetration (2015) is already causing power quality problems on the
network and is expected to exaggerate the issues in the next five years [16]. It is reported that
many distribution transformers on low voltage (LV) networks are overloading with higher PV

7
 Chapter 2. Literature Review

installed capacity, which in the worst condition, might be at risk of failure and need
replacement. This remarkable growth is mainly due to the decreasing PV system costs,
incentive programs from federal and state governments and rising electricity prices. Therefore,
it is very important to investigate the power quality impacts of PV systems and to take
preventive measures to improve the existing network withstand capability to facilitate more
PV systems integration.

2.3. Power Quality Issues with Solar PV System Integrations

2.3.1. Voltage Rise

Traditionally, distribution networks are designed as uni-direction power flow systems with
a distribution transformer energised from a single source. During normal operation, the current
flows through line impedance and to the loads, which results in voltage drops along the feeder
line. To compensate voltages on the distribution network, different voltage regulation devices
are used, such as on-load tap changing transformers (OLTC), step voltage regulators, switched
capacitors, which are designed by considering the primary substation. The objective of the
regulators is to regulate the voltage at the output terminals and is then set to boost the voltage
along the feeder. Due to the incorporation of PV systems into the grid, the distribution networks
are becoming a bi-directional power flow system. The integration of a large number of PV
generators in an existing grid causes a significant rise in the local voltage levels due to reverse
power flow, which can be seen as a major technical issue. Most importantly, in the case of high
PV generation and low load conditions the power injection from PV has a severe impact at the
end of the feeder as well as on the neighbourhoods as presented in [17]. In addition, various
types of PV inverters until now are typically designed to produce only real power and do not
regulate the voltages at the integration point. As a result, there is great difficulty in regulating
the voltage due to negative coordination of the unidirectional regulatory equipment and the PV
systems [18]. Consequently, in the case of short-term voltage variations due to solar irradiation
change, there is a tendency to increase the number of undesirable operations of regulatory
devices and therefore, reduce the equipment lifetime.

Reference [19], studied the impact of distributed generation on steady state voltage
regulation and also provided the guidelines to mitigate arising issues. In [20], the impact of
different penetration levels and sizes of PV systems on steady state voltage has been studied
and simulated for Australian Single Wire Earth Return (SWER) networks. It has been shown
that this kind of network often experiences voltage rise in large penetrations. In [17, 21], the

8
 Chapter 2. Literature Review

voltage rise impacts in an unbalanced distribution network with high PV penetration related to
the unbalanced feeder impedance, feeder layout and phase loading have been studied.
Conversely, the PV systems are advantageous in reducing line losses and providing voltage
support only when the PV generation capacity is lower than the load requirement as presented
in [22, 23], however, this is not always possible. This reverse power flow is unpredictable and
the management of voltage limits by utility operators at the interconnection point can lead to
both voltage rise and voltage drop. Hence, the voltage rise and its regulation issue is likely to
be of more concern in LV networks and needs to be investigated in the case of high PV
penetration.

2.3.2. Voltage Fluctuations/Flicker

The major problem with solar PV generation is its variable nature due to seasonal effects
and limited predictability, which results in intermittent generation, which can lead to power
imbalance. It can be noticed [24] that the solar fluctuations can have erratic variations on
voltage and power flow. Literature [25, 26], investigated the impacts of voltage variations in
unbalanced LV distribution networks due to PV power fluctuations. The study considered a
constant PV power profile variation to reflect cloud transient effects but it does not simulate a
real-time scenario. The rapid variations in solar irradiation due to cloud transients can cause
substantial fluctuations in the PV output power, which in turn could cause voltage fluctuations
at the PV integration node [27, 28]. The sudden ramp up/down in PV power can increase the
number of undesirable operations of tap changers, which may lead to a hunting effect and can
reduce the lifespan of equipment. In addition, sometimes these regulatory devices may react
slower than the cloud passage. With high connection densities of PV systems, voltage
fluctuations and flicker issues could be the critical constraints in any network. In [29], the PV
variability and the extreme ramp rates were examined to analyse the flicker. For this, a realistic
solar irradiation profile of only 60sec was applied on the distribution network and simulated
for instantaneous power drops. However, this study may be insufficient and not an accurate
approach for evaluation of the flicker levels with PV systems. It is important to consider various
solar irradiation characteristics and their impacts on network voltage profiles.

2.3.3. Voltage Unbalance Factor

Another significant voltage quality issue in the distribution system with high penetration of
PV units is voltage unbalance. Normally, the electric utilities are supplying the loads evenly
along the distribution feeder to minimise the voltage unbalance factor. Despite this fact, the

9
 Chapter 2. Literature Review

voltage levels at the consumer side are unbalanced due to uneven power consumption by loads.
In addition, as the rooftop PV installations are randomly placed along the distribution feeders
since the installations depend on customer interest. Therefore, the increment in voltage levels
in each phase have different levels with PV power variation, which may further increase the
unbalance level [25]. Along with this, the distributed massive integrations of solar PV systems
will raise the uncertainties in three phase electric network. In reference [30, 31], the voltage
imbalance sensitivity analysis in a residential distribution grid with random rooftop PV
penetrations has been carried out, and a stochastic evaluation has been done. This voltage
unbalance factor is higher at the end of the feeder than at the beginning of the feeder [32]. In
addition, the increase/decrease in voltage unbalance depends on the PV system size and
location, and phase loading. An increase in voltage unbalance can result in malfunctioning of
regulatory equipment and reduces the lifetime of customer appliances.

2.3.4. Voltage Sag/Swell

Furthermore, voltage sag/swell is recognized as one of the most important power quality
issues and is a commonly unavoidable occurrence in utility distribution networks. These are
caused by unsymmetrical and symmetrical three phase faults on sensitive loads or starting of
large industrial motors, which will trip the three-phase circuit breaker resulting in loss of supply
[33]. In practice, faults on the distribution network cause the PV systems to disconnect from
the grid [34]. This will create unbalanced voltage sags on highly loaded phases and swells on
lightly loaded phases of the network. References [35, 36] highlighted the influence of altering
weather patterns and cloud transients on PV power generations. Therefore, depending on the
high connection densities of PV systems on the distribution feeder when subjected to sudden
deficits in solar irradiation can cause large fluctuations for a few seconds, leading to poor power
quality such as voltage sag/swell.

2.3.5. Harmonic Emission Issues from PV Systems and their Impacts

In addition, harmonic problems are common in distribution networks, mainly coming from
nonlinear loads, transformers and increased use of power electronic equipment [37-39].
Characteristically, low voltage distribution networks have high R/X ratios and high
impedances, which can result in high harmonic voltages. In general, PV systems consist of PV
panels, Maximum power point tracking (MPPT) controller, a boost converter, a DC/AC
inverter and an interfacing LC filter [40, 41]. With rapid developments in PV technology, a
wide range of PV inverter technologies using a variety of control strategies are commercially

10
 Chapter 2. Literature Review

available in the market. The harmonics generation from these inverters will be divergent, which
depends on the type of control strategy and size of PV systems and on the grid voltage
harmonics. The harmonics that are injected into the network should comply with the standard
requirements in order to maintain the power quality of the grid. Nevertheless, these PV
inverters are negatively interacting in the real network [42].

As mentioned earlier, solar power generation is variable in nature. The active power
generation has a linear relation with the solar insolation. However, the current harmonic
distortion can possess an inverse trend to varying solar conditions. In addition, a PV inverter
with lower switching frequency can generate high harmonic distortion. Therefore, it is
important to characterise the harmonics produced from the grid connected PV inverter system
for different operating conditions. In the literature [43], the experiments were performed on PV
plant to analyse the multifaceted distorted waveform for varied solar conditions. In [42, 44],
the investigation was performed to study the power quality characteristics of discrete PV
inverter models. In addition, the analysis of current harmonic distortion from different single-
phase inverter models was presented and suggested an expression to estimate the average total
harmonic distortion. However, these PV inverters need further research to characterise the
individual harmonic aggregations in the case of multiple installations. The reason that the
harmonic distortions from several identical systems could be different is due to their
independent operation. Therefore, the harmonic magnitude and phases are also different for
each inverter and the resultant harmonics may suppress or increase because of harmonic
superimposition. Furthermore, widespread penetration of several PV systems can affect the
harmonic distortion levels of the distribution network [45, 46]. Therefore, accurate PV inverter
modelling is essential to estimate the exact harmonic influences in the network.

In reference [47], an approach to model the PV output power generation from a prediction
of solar irradiation was proposed. This method considers the input random variables such as
clearness index and fraction of diffusion irradiation for prediction. Consequently, the harmonic
emissions from PV system are altered by varying solar irradiation conditions. In [48], a
probabilistic PV harmonic model has been developed based on the measurement data, which
can be used in harmonic load flow studies. The statistical characterization of PV harmonic
current emissions at different fundamental frequency output levels was presented. In literature
[49], the PV system is modelled as a current source with harmonic frequencies, which
eliminates the impact of control strategy. Multiple installations of PV systems in various
locations together with the proliferation of nonlinear loads can introduce an issue of increased

11
 Chapter 2. Literature Review

harmonics in the distribution system [50]. A harmonic distortion analysis performed on the
distribution network with distributed generation and nonlinear loads was given in [51]. Until
now, utility network providers maintain the harmonic distortion levels at the customer
connection point, which are under the limits according to IEEE 519 standards [52, 53].
However, excessive harmonic currents flowing through the grid can lead to increasing line
losses, transformer overheating and premature ageing [54, 55].

The severity of the harmonic issue will depend on the type and rating of the equipment and
the loading condition of distribution feeders [56]. In addition, PV systems are usually treated
as a negative load in some studies [57, 58]. In the literature, several load models were
developed to study the harmonic analysis in the networks [46, 59]. The common approach used
to model harmonic producing loads is to include a set of current sources for each frequency.
Such modelling ignores the switching harmonics of nonlinear loads and injects a constant
harmonic pattern into the grid, which will not give realistic conditions. Dynamic aggregation
of different types of domestic and commercial loads that inject harmonics, induce several
complexities into the network. In references [38, 39, 60], a few different nonlinear load models
namely Compact Fluorescent Lamps (CFL), televisions, computers, variable speed drives, DC
drives and other loads have gained attention for harmonic analysis in the distribution network.
Therefore, to perform an accurate analysis, the collective impacts of uncertainty in the loading
patterns and aggregation of harmonic components need to be considered. Utility companies are
familiar with estimating the harmonic levels and their potential effects on the network.
However, with the penetration of PV systems on the customer side, it is becoming difficult to
estimate the exact harmonic contributions from PV systems and there is a lack of study found
in the literature.

2.3.6. Possibility of Harmonic Resonance Incidences

Furthermore, harmonic emission issues related with PV inverters will strongly depend on
the type of control scheme and the output filter design, which can configure a resonance circuit
with the grid. As a result, the high integration of different varieties of PV systems in the
distribution networks, which is already stressed by nonlinear load proliferations, are raising
potential concerns of harmonic resonance [50, 61, 62]. In LV distribution networks, harmonic
resonance conditions are common, which can cause significant problems in the network.
Typically, resonance events occur when the network capacitive reactance and inductive
reactance become equal at a particular frequency, called a resonance frequency. In general, the
network can have multiple resonance frequencies depending on the system configuration and

12
 Chapter 2. Literature Review

characteristics. These resonance frequencies will be a major issue when they are close to any
of the individual harmonic frequencies [63]. The type of resonance may be parallel or series in
the network. In [45, 64], harmonic resonance problems concerning PV systems integration in
a distribution network have been presented. The reported resonance phenomena are based on
the assumption that the PV system is modelled as a simple harmonic current source, with
capacitor filter and is ignoring the inverter control functions. However, these assumptions are
not valid in reality due to the operational and design characteristics of a PV system that leads
to inaccurate harmonic analysis. Furthermore, most of the PV inverter filters are designed as
LC and LCL types that are also a concern in the grid [65]. In fact, the network impedance
changes over time due to operating conditions of loads, the interaction between passive
elements and harmonic sources, which can excite the harmonic resonances. Depending on the
type of resonance, the network voltage and current harmonics could be amplified to higher than
expected [66]. Such resonances can cause overloading of neutral conductors, transformer
overheating, nuisance tripping of circuit breakers and capacitor bank overstress [55, 67].
Although the issues associated with PV systems on the distribution grid are widely analysed,
the harmonic resonance problems related to PV system penetration and their controller
interactions are rarely assessed.

2.3.7. Other Issues

Conventional inverters are designed to operate at unity power factor and produce only
active power which may reduce power factor at the distribution transformer as presented in
[68]. The active current required by load supplied by PV systems locally while reactive current
is drawn from the upstream network results in power factor reduction and increases the reactive
power requirement.

2.4. Solutions for Power Quality Improvement

To improve the power quality of low voltage distribution systems, several solutions have
been proposed in the literature. Optimal allocation and sizing of solar PV systems can provide
voltage support and reduces line losses as presented in [69]. Authors in [70, 71] investigated
the voltage quality issues with high PV penetration. Studies revealed that dispersed PV
integration could improve the under voltage and voltage unbalance issues as well as sags in the
distribution network. Nonetheless, in a utility distribution network this is not always possible
because of randomness of PV system installations, which can increase the feeder impedance
and overloading of transformers. Consequently, voltage regulation in such a network will

13
 Chapter 2. Literature Review

become complicated. As mentioned earlier, traditional Volt-VAr regulatory equipment may

operate slowly in the case of power fluctuations with high PV integrations [72].

Authors in [73] proposed a network reconfiguration approach to solve voltage variation

problems in an unbalanced distribution network. The sensitivity analysis has been performed
by swapping phase sequences, loads, and PV interconnections. However, this network
reconfiguration process is complicated in the case of real-time complex distribution networks
due to increased random PV installations. Another network solution is upgrading of distribution
feeder’s cross-section [74], which is very expensive for the utilities to change all the lines.
Further in [74, 75], regulation of the distribution transformer between two pre-defined tap
settings is proposed, which is based on the feeder loading data. However, most of distribution
transformers in the network have fixed tap sets, which require manual adjustments, which is
not a reliable process.

Alternatively, Energy Storage System (ESS) plays a vital role in the adoption of renewable
energy systems as they can handle the variability issues with renewable energy sources. Battery
technology is the most widely used ESS for LV networks. ESSs are designed to store surplus
energy during peak generation and low demand periods for later use during high power
demand. It has been proposed in [76-78] that distributed ESS can be used for preventing
overvoltage and voltage unbalance issues due to high PV penetration in LV residential
networks. In references [77, 78], a technique to mitigate the voltage unbalance with single-
phase energy storage systems is proposed. Results indicated that the voltage unbalance factor
has reduced significantly to less than 2% with ESS operation, which was higher than about
3.5% with high PV penetration in an LV network. In [79], a coordinated control scheme has
been proposed for an energy storage system with an on-load tap changing transformer to
mitigate voltage rise, which limits the active power injections. The proposed controller sends
charging/discharging operation commands to ESS in such a way to minimise the system losses
and to shave the peak load. Although OLTC is a promising solution, it practically may not
provide continuous voltage support. ESS can also have the capability of reactive power
compensation for the grid.

Conversely, reactive power compensators such as shunt capacitors and Custom Power
Devices (CPD) have traditionally been used to overcome power quality issues. In [80], various
voltage quality issues raised by random rooftop PV installations have been investigated and a
comparative analysis has been presented with two custom power devices, namely Distributed
Static Compensator (DSTATCOM) and Dynamic Voltage Restorer (DVR). It has been

14
 Chapter 2. Literature Review

concluded that DSTATCOM is more efficient for voltage profile improvement and unbalance
reduction. In literature [81-83], various functionalities of the DSTATCOM operations in
compensating voltage sags and swells in the case of disturbances have been presented. This
device is also capable of providing harmonic and dynamic reactive power compensation for
the grid. Besides, [84] showed that DVR has the capability to mitigate voltage flickers,
unbalance and voltage sags/swells problems. However, the above mentioned solutions are very
expensive for the utility operators to install and manage.

Apart from conventional solutions, PV inverters themselves could be used as potential

solutions for power quality improvement. Various methods, such as voltage regulation and
reactive power injection from PV systems will better manage the LV network. In [85, 86], to
minimise the steady state voltage variation and power loss, an adaptive VAr control for PV
systems has been proposed. In [87] various VAr control methods for a PV inverter have been
discussed. In [88-90], the application of a large-scale solar farm (PV) inverter operating as
STATCOM has been proposed for power quality improvement in transmission networks. The
proposed control system for a PV inverter with an energy storage system can perform the four-
quadrant operation, which provides active and reactive power exchange with the grid.

Nowadays, PV systems with smart inverter technology are commercially available, which
have the capability of injecting/absorbing fixed reactive power to address voltage quality issues
to some extent [91, 92]. Nonetheless, in the distribution level, small-scale to medium-scale PV
inverters can also possess versatile capabilities similar to the DSTATCOM by modifying a few
functionalities to the inverter controller. A number of control methods have been proposed for
the three-phase inverter topologies in the literature [87, 88, 93] to produce compensation
currents for voltage regulation and reactive power compensation. All these methods are well
suited for steady state, sinusoidal and balanced conditions of supply. However, in the case of
unbalanced supply and non-linear unbalanced load conditions these control techniques are
inappropriate. Therefore, an appropriate control strategy should be put in place for the PV
inverters to enhance power quality in the unbalanced network. Such transformation of PV
inverters as a source of impediment to a viable solution could address several integration issues
and offer a cost-effective solution for both utilities and PV customers.

2.5. Overview of Standards

This section introduces several regulatory standards to define the power quality criteria
about the distribution networks and PV system interconnection requirements.

15
 Chapter 2. Literature Review

2.5.1. Voltage Quality Standards

Voltage magnitude variation is the important constraint in distribution networks with high
PV penetration. According to the American standard ANSI C84.1-2006 [94], the preferred
operating service voltage range is to be within ±5%, indicated as Range A in the standard. The
standard also suggested some countermeasures and required regulatory actions to maintain the
voltage range and to protect equipment against increased voltages. In comparison, the
equivalent Australian standard AS 60038-2000 [95], does not provide different voltage ranges
and regulatory actions. However, it has defined the service voltage limits of a single phase
(230V) and three phase supply voltage (400V) with a tolerance of -6% and +10%. The upper
voltage range is higher than compared to the ANSI limits. However, in current practice, Ergon
Energy and Energex utilities in Queensland, Australia consider the AS 61000.3.100: 2011 [96]
standard voltage range. As per the standard, the steady state voltage limits at the customer
integration points for the 11kV network should be within ±5% of nominal voltage and for the
LV network should be within 415V ±6%. It also indicated the single-phase voltage limit for a
240V range to be ±6%.

Besides, the IEEE 1159-2009 [97] standard describes the various power quality issues
caused by electromagnetic phenomena. According to the standard, the drop in RMS voltage by
10% to 90% from rated is defined as voltage Sag. Whereas, the swell is defined as an increase
in RMS voltage from 10% to 80% of nominal for the duration of half a cycle to 1 minute.
Similarly, according to the AS/NZS 61000.4.30 [98] standard, the thresholds for the voltage
dip and swell are 90% and 110% of nominal voltage respectively.

In addition, the various definitions of voltage unbalance (VU) and its calculations are
presented in [99]. According to IEEE definition, the percentage of voltage unbalance present
in the phase voltages is expressed as given in equation (2.1).

%VU =
{
max Vavg −Va , Vavg −Vb , Vavg −Vc } × 100
(2.1)
Vavg

Where, Vavg = (Va + Vb + Vc ) / 3

The desirable voltage unbalance limit is specified in IEEE 1159 as 1% for any voltage level
and the maximum limit is 2%. In contrast, the Australian standard AS/NZS 61000.4.30 defines
voltage unbalance by utilizing the phase to phase voltages as given in equation (2.2). Here Vab,
Vbc and Vca are the fundamental phase to phase voltages. According to the Australian standard,
the limit for the voltage unbalance should not be over 5%.

16
 Chapter 2. Literature Review

1− 3− 6 β
=
% VU × 100 (2.2)
1+ 3− 6 β

4 +V 4 +V 4
Vab
Where, β= bc ca
(Vab +Vbc +Vca )
2 2 2 2

Alternatively, the true definition of the voltage unbalance is defined as the ratio of negative
sequence voltage magnitude to the positive sequence voltage magnitude as given in equation
(2.3).
|𝑉𝑉− |
%V U = |𝑉𝑉+ |
× 100 (2.3)

Other key issues in the distribution network are voltage fluctuations and flicker caused due
to loads. Traditionally, flicker is defined as a rapid objectionable change in the light level often
produced by voltage fluctuations [100]. It is usually measured by the quantification of voltage
magnitude and the frequency of the fluctuations. Existing IEEE 519 [52] standards for flicker
curves are not preferable in the case of PV induced flicker. The ramp rate of voltage variation
may be faster or slower depending on the cloud pattern and speed. Therefore, AS 4377 [101]
and IEEE 1453 [102] recommended practice suggested two indices for measuring the flicker
on a network namely, short-term severity Pst (measured over 10 min) and long-term severity
Plt (measured over 2 hours). The short-term flicker is calculated every 10 minutes using the
cumulative weighted probability values of the flicker sensation levels at 0.1%, 1%, 3%, 10%
and 50% of the time interval given in equation (2.4).

𝑃𝑃𝑠𝑠𝑠𝑠 = �0.0314𝑃𝑃0.1 + 0.0525𝑃𝑃1 + 0.0657𝑃𝑃3 + 0.28𝑃𝑃10 + 0.08𝑃𝑃50 (2.4)

Where, 𝑃𝑃0.1 , 𝑃𝑃1 , 𝑃𝑃3 , . . , etc. indicate the ratio of relative voltage change (∆V=Vmax-Vmin) to
the nominal voltage (Vnom).

From the standards, the long-term flicker index is determined by an average of Pst levels
over a 2 hour time period expressed in equation (2.5).

3 1
𝑃𝑃𝑙𝑙𝑙𝑙 = � ∑ 𝑃𝑃𝑠𝑠𝑠𝑠3 (2.5)
12

AS/NZS 61000.3.3: 2012 [103] specifies the limit for voltage fluctuations and flicker
produced by devices of input current <16A/phase. According to the standard, the Pst and Plt
values should be less than 1 and 0.65 respectively. In addition, AS/NZS 61000.3.7:2001 [104]
represents the compatibility levels for flicker in both low and medium voltage systems, which
are 1 and 0.8 for Pst and Plt respectively. Further, the utilities also set the limits depending on

17
 Chapter 2. Literature Review

the agreement by the customer, which are indicated as short-term (Epsti) and long-term (Eplti)
flicker emissions in the standard. These levels are restricted to 0.35 (Epsti) and 0.25 (Eplti).

Moreover, IEEE 1547-2003 [105], IEEE 929-2000 [106] and AS 4777.2 [107] provide the
integration requirements of photovoltaic inverter systems in distribution networks. According
to IEEE 1547-2003 and 929-2000, currently the PV inverters are not allowed to provide voltage
regulation and are required to be disconnect from the grid during the outside operating voltage
limits (0.88p.u - 1.10p.u). As per AS 4777.2, these voltage limits are 0.94p.u -1.10p.u. In
addition, IEEE 929 specified the power factor requirement of the PV inverter to be higher than
0.85 leading or lagging. The standard also recommended that the grid-connected PV systems
are designed precisely to operate at close to unity power factor. It is to be mentioned that the
reactive power can be supplied to the grid, but the reactive power compensation and the voltage
regulation is not allowed by the IEEE standard. In comparison, the corresponding Australian
standard AS4777.2-2005 [107] indicated the PV inverter operating power factor range from
0.95 lagging to 0.8 leading and it allows voltage regulation within this power factor range.
However, recent standard AS 4777.2-2015 [108] proposed that the PV inverter is allowed to
provide constant reactive power with fixed power factor not exceeding 0.95 leading/lagging.
Besides, the standard also stated that if the PV inverter is capable of operating in reactive power
mode, the maximum ratio of reactive power to rated apparent power should be up to 100%
leading or lagging.

2.5.2. Harmonic Standards

Many standards around the world have presented the limits for harmonic distortion in
distribution networks. The most common measure of harmonics is the Total Harmonic
Distortion (THD), which is defined as the ratio of the root-sum-square of the harmonic quantity
to the fundamental quantity. In addition, the individual harmonic distortion is expressed as the
level of harmonics with respect to the percentage of fundamental. IEEE standards 519-1995
[52] and 519-2014 [53] focus on the allowable harmonic distortion levels of the network and
IEEE 1547-2003 [105] describes the grid interconnection requirement of the distributed
resources. According to the standards, the maximum limit of individual odd current harmonics
up to 11th order is 4.0%, followed by up to 17th order is about 2.0%, 1.5% is the limit for up to
23rd order and the orders above 23rd to 35th should be less than 0.6%. Moreover, the current
THD should be less than 5%. Recent standard IEEE 519-2014 indicated the individual voltage
harmonic should be less than 5% and the total harmonic distortion should be below 8%.

18
 Chapter 2. Literature Review

Also, similar limits are specified by AS 4777.2 [107], IEEE 929-2000 and IEC 61000-3-2:
2012 [109] as indicated in Table 2.1. However, the recent standard AS 4777.2-2015 has
introduced more strict limits compared to the previous AS4777.2 and IEEE standards.
Similarly, as per the EN61727 [110] standard, the voltage THD limit for the photovoltaic
interconnection is 2% and individual harmonics should be no more than 1%.

Table 2.1. Harmonic Limits of Grid Connected PV Inverter Requirement

Standards Individual Current Odd Harmonic Order

IEEE 929-200
�∑∞ 2
𝑛𝑛=2 𝐼𝐼𝑛𝑛
3rd - 9th 11th - 15th 17th - 21st 23rd - 33rd >33rd THD =
and AS4777.2-2005 𝐼𝐼1

(%Fundamental) <4.0% <2.0% <1.5% <0.6% <0.3%

<5%
AS4777.2-2015 3rd – 7th 9th - 13th 15th – 19th 21st-33rd >33rd

(%Fundamental) <4.0% <2.0% <1.0% <0.6% <0.3%

2.6. Summary

This chapter has provided a clear insight into the problems associated with solar PV system
integration into distribution networks. All the studies reviewed so far have demonstrated that
power quality is one of the key issues in distribution networks. It has been discussed that
voltage quality issues such as overvoltage, voltage fluctuations/flicker, voltage unbalance and
sag/swell are critical concerns in the low voltage networks with PV. In addition, due to the
involvement of power electronic based inverter technology, there is a potential concern for
harmonic injection issues. With the proliferation of a large number of PV inverter systems
combined with nonlinear loads, the network harmonic distortion levels could be influenced and
excite resonance problems. Therefore, it is important to investigate and quantify how the PV
contributes to the network harmonic levels and its operational impacts on network
performance. The next chapter presents the background modellings and software tools used for
this research.

19
 Chapter 3. Modelling and Simulation Tools

CHAPTER 3

MODELLING AND SIMULATION TOOLS

3.1. Introduction

To investigate the impact of PV systems on various power quality aspects of the distribution
network, an appropriate and accurate modelling of the PV systems, load units and test network
is necessary. This chapter presents the background theory and detailed modelling of various
components of the grid connected PV systems such as PV source, maximum power point
tracking techniques, DC/DC converter and DC/AC inverter control strategies used for this
research. Afterwards, the test system and load modelling requirements suitable for different
voltage quality and harmonic impact analyses to reflect a real-time scenario have been
presented. Finally, the selected various software and hardware tools utilised in this research are
explained.

3.2. Grid Connected Solar PV System

Grid connected solar PV systems convert the sunlight into AC electricity. The functional
block diagram of a grid connected PV system is shown in Figure 3.1. The system includes solar
PV panels connected as arrays, a boost converter, an inverter with interfacing filter and an
isolation transformer. In the following sub-sections, background theory and modelling of these
components are presented.

20
 Chapter 3. Modelling and Simulation Tools

Figure 3.1. Block Diagram of Grid Connected PV System.

3.2.1. Modelling of Solar PV Source

The building block of the PV array is the solar cell, which is basically a P-N semiconductor
junction. The solar energy is converted into electricity through photovoltaic effect. In [40], a
mathematical modelling of PV arrays was presented. The equivalent circuit of a solar PV cell
is shown in Figure 3.2. The current Id flows through anti-parallel diode producing nonlinear I-
V characteristics in the PV cell as shown in Figure 3.3.

Figure 3.2. Solar PV Cell Equivalent Circuit.

Figure 3.3. I-V Characteristic of Solar PV Cell [111].

The current and voltage relationship of an ideal solar PV cell can be expressed as given in
equation (3.1).

21
 Chapter 3. Modelling and Simulation Tools

𝐼𝐼𝑝𝑝𝑝𝑝,𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 = 𝐼𝐼𝑠𝑠𝑠𝑠,𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 − 𝐼𝐼𝑑𝑑

𝑞𝑞𝑉𝑉𝑝𝑝𝑝𝑝,𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
= 𝐼𝐼𝑠𝑠𝑠𝑠,𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 − 𝐼𝐼𝑟𝑟𝑟𝑟,𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 �𝑒𝑒𝑒𝑒𝑒𝑒 � � − 1� (3.1)
𝑛𝑛𝑛𝑛𝑛𝑛

Where, Isc,cell is the current generated by the incident of sunlight (it is directly proportional
to the solar irradiation), Id is the Shockley diode current, Irs,cell is the reverse saturation or
leakage current of the diode, q is the electron charge (1.60217646 × 10−19 C), k is the
Boltzmann constant (1.3806503 × 10−23 J/K), T (in Kelvin) is the temperature of the p–n
junction and n is the diode ideality factor.

A PV module primarily consists of solar cells connected in series and a PV array is

composed of series and parallel connected modules. The single PV cell circuit can ideally be
scaled up to represent a number of series (Ns) and parallel (Np) combinations as represented in
Figure 3.4.

Figure 3.4. Equivalent Circuit of a PV Module/Array [111].

The equation that characterises the current and voltage relationship of a PV array is derived
in (3.2). From the equation, it can be seen that the PV output power has a linear relationship
with solar irradiation but the temperature has an inverse relationship with PV output power.

𝑞𝑞 𝑉𝑉𝑝𝑝𝑝𝑝 𝐼𝐼𝑝𝑝𝑝𝑝 𝑅𝑅𝑠𝑠 𝑝𝑝 𝑝𝑝𝑝𝑝 𝑁𝑁 𝑉𝑉 𝐼𝐼𝑝𝑝𝑝𝑝 𝑅𝑅𝑠𝑠

𝐼𝐼𝑝𝑝𝑝𝑝 = 𝑁𝑁𝑝𝑝 𝐼𝐼𝑠𝑠𝑠𝑠 − 𝑁𝑁𝑝𝑝 𝐼𝐼𝑟𝑟𝑟𝑟 �𝑒𝑒𝑒𝑒𝑒𝑒 � � 𝑁𝑁 + �� − 1� − 𝑅𝑅 � 𝑁𝑁 + � (3.2)
𝑛𝑛𝑛𝑛𝑇𝑇𝑐𝑐 𝑠𝑠 𝑁𝑁𝑝𝑝 𝑝𝑝 𝑠𝑠 𝑁𝑁𝑝𝑝

Where Ipv: PV array output current (A), Vpv: PV array output voltage (V), Isc: Solar cell
photocurrent (A), Irs: Solar cell reverse saturation current (A), Tc: Solar cell absolute operating
temperature (K), RS: Cell intrinsic series resistance (Ω), RP: Cell intrinsic shunt resistance (Ω).

The photocurrent or short circuit current for a given operating condition of the PV array is
assumed to be related to the short circuit current at standard test condition (STC).
S
Isc(T) = �Isc(Tref) + K I (T-Tref )� (3.3)
Sref

22
 Chapter 3. Modelling and Simulation Tools

Where Isc(Tref) (A) is the light-generated current at standard test condition (usually at 25◦C
and 1000 W/m2), KI is the temperature coefficient of the current, T and Tref being the actual
and reference temperatures [in Kelvin] respectively, S (W/m2) is the irradiation on the device
surface and Sref is the nominal irradiation.

The diode saturation current Irs and its dependence on the temperature for PV model
improvement can be expressed as given in equation (3.4). This modified equation could match
the open-circuit voltages of the model with real-time data for a very large range of
temperatures.
𝐼𝐼𝑠𝑠𝑠𝑠(𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇) +𝐾𝐾𝐼𝐼 (𝑇𝑇−𝑇𝑇𝑟𝑟𝑟𝑟𝑟𝑟 )
𝐼𝐼𝑟𝑟𝑟𝑟 = 𝑞𝑞(𝑉𝑉𝑜𝑜𝑜𝑜,(𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇) +𝐾𝐾𝑣𝑣 (𝑇𝑇−𝑇𝑇𝑟𝑟𝑟𝑟𝑟𝑟 )
(3.4)
𝑒𝑒𝑒𝑒𝑒𝑒�� ��−1
𝑛𝑛𝑛𝑛𝑇𝑇𝑟𝑟𝑟𝑟𝑟𝑟

Where, Voc(Tref) is the open circuit voltage at STC, KV is the temperature coefficient of the
voltage. Therefore, the equation (3.4) includes both coefficients KI and KV. The saturation
current is strongly reliant on the temperature and this equation proposes a different approach
to express the dependence of Irs on the temperature.

To develop an absolute PV array system represented in Figure 3.2, the above derived
mathematical model has been considered and developed in PSCAD software, which reflects a
real-time system as shown in Figure 3.5. The corresponding I-V characteristics of a PV Array
at various solar irradiation conditions from 1000W/m2 to 200W/m2 is shown in Figure 3.6.
Also, the relation between PV output power and open circuit voltage at different solar
irradiation conditions is presented in Figure 3.7.

23
 Chapter 3. Modelling and Simulation Tools

Figure 3.5. Simulation Model of PV Module/Array.

IDCA1 vs BACKV IDCA2 vs BACKV IDCA3 vs BACKV IDCA4 vs BACKV IDCA5 vs BACKV IDCA6 vs BACKV
IDCA7 vs BACKV IDCA8 vs BACKV IDCA9 vs BACKV
8.89
PV Panel Short Circuit Current (A)

8.001
7.112
6.223
5.334
4.445
3.556
2.667
1.778
0.889
0
0 10 20 30 40

PV Panel Open Circuit Voltage (V)

Figure 3.6. I-V Characteristics of PV Array at Various Solar Irradiation Conditions.

24
 Chapter 3. Modelling and Simulation Tools

PDCW1 vs BACKV PDCW2 vs BACKV PDCW3 vs BACKV PDCW4 vs BACKV PDCW5 vs BACKV PDCW6 vs BACKV

PDCW7 vs BACKV PDCW8 vs BACKV PDCW9 vs BACKV

280
PV Panel Output DC Power (W)

252
224
196
168
140
112
84
56
28
0
0 10 20 30 40
PV Panel Open Circuit Voltage (V)

Figure 3.7. P-V Characteristics of PV Array at Various Solar Irradiation Conditions.

3.2.2. DC/DC Converter

Figure 3.8 shows simple DC/DC converter called a Boost converter. This boost converter is
used to boost the output voltage of a PV array to achieve the optimum grid voltage level. The
duty cycle for the gate switch is obtained from the Maximum Power Point Tracking (MPPT)
system. This MPPT controller is essential for increasing the efficiency of the PV arrays.

The equation for boost converter output voltage in terms of input voltage is

𝑉𝑉𝑝𝑝𝑝𝑝
𝑉𝑉𝐿𝐿 = (3.5)
1−𝐷𝐷

Where, D is duty cycle of IGBT switch.

Figure 3.8. Boost Converter Circuit.

25
 Chapter 3. Modelling and Simulation Tools

3.2.3. Maximum Power Point Tracking Algorithms

MPPT algorithms [41, 112] are necessary for PV applications because the maximum power
point of a solar panel varies with the solar irradiation and temperature. Therefore, an MPPT
control algorithm is required to extract the maximum power from a solar PV array. In the
literature, many methods to find the maximum power point have been developed. These
techniques vary in many aspects such as complexity, cost, range of effectiveness, convergence
speed and correct tracking when solar irradiation and/or temperature changes. Among them,
the Perturb and Observe Algorithm (P&O) and the Incremental Conductance (InCond)
algorithms are the most widely used techniques, which have been considered in this research.
In both P&O and InCond methods, the tracking of maximum power point depends on the
amount of increment in reference voltage.

3.2.3.1. Perturb and Observe (P&O) Algorithm

The P&O method operates periodically by perturbing the terminal voltage or the current of
the PV array and comparing the PV power with the preceding cycle. This algorithm
continuously examines the sign of the last perturbation and the sign of the last increment or
decrement in the PV power to determine and take the decision of what the next perturbation
should be. For instance, Figure 3.9 shows the P-V characteristics curve for varying solar
irradiation. Initially, starting from point A a perturbation of voltage to increment by ∆V will
shift the operating point to B. Then the perturbation will be reversed due to the decrement in
the power.

Figure 3.9. PV Panel P-V Characteristic Curve [113].

26
 Chapter 3. Modelling and Simulation Tools

Measure Vpv(k), Ipv(k)

P(k) = Vpv(k) . Ipv(k)

K=k+1

Measure Vpv(k), Ipv(k)

P(k) = Vpv(k) . Ipv(k)

Yes
P(k)-P(k-1) = 0

No Yes
P(k)-P(k-1) > 0

Yes No No Yes
P(k)-P(k-1) > 0 P(k)-P(k-1) > 0

Vpv = Vpv - ∆V Vpv = Vpv + ∆V

Figure 3.10. The Flow Chart of the P&O Algorithm.

Conversely, if the solar irradiation increases then the P-V power curve shifts from P1 to P2
and the operating point will move from point A to C. Subsequently, if the PV power increases,
the perturbation will be in the same direction and if the PV power decreases, then the resulting
perturbation should be in the opposite direction. This process is repeated until the MPP is
obtained and the operating point oscillates around the MPP. The control algorithm of the P&O
method has been implemented based on these facts [4] as shown in Figure 3.10 and the PSCAD
model is shown in Figure 3.11.

27
 Chapter 3. Modelling and Simulation Tools

Figure 3.11. PSCAD Model for P&O Algorithm.

3.2.3.2. Incremental Conductance Algorithm

The incremental conductance algorithm [112, 113] is based on the fact that the slope of the
curve power Versace voltage of the PV module is zero at the MPP. By comparing the increment
of PV power with the increment of voltage between two consecutive samples, the change in
MPP voltage can be determined. One millisecond is selected as sampling time. The deviations
Δ𝑉𝑉 and Δ𝐼𝐼 are calculated from the difference of each sample to the prior one. A flow chart of
the Incremental conductance algorithm is shown in Figure 3.12. The following equations (3.6)
and (3.7) determine the MPP in this algorithm.
𝑑𝑑𝑑𝑑 𝑑𝑑(𝑉𝑉𝑉𝑉) 𝑑𝑑𝐼𝐼𝑝𝑝𝑝𝑝 𝛥𝛥𝐼𝐼𝑝𝑝𝑝𝑝
= = 𝐼𝐼𝑝𝑝𝑝𝑝 + 𝑉𝑉𝑝𝑝𝑝𝑝 ≃ 𝐼𝐼𝑝𝑝𝑝𝑝 + 𝑉𝑉𝑝𝑝𝑝𝑝 (3.6)
𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑 𝑑𝑑𝑉𝑉𝑝𝑝𝑝𝑝 𝛥𝛥𝑉𝑉𝑝𝑝𝑝𝑝

𝑑𝑑𝑑𝑑 𝛥𝛥𝐼𝐼𝑝𝑝𝑝𝑝 𝐼𝐼𝑝𝑝𝑝𝑝

Further, =0↔ = − at MPP,
𝑑𝑑𝑑𝑑 𝛥𝛥𝑉𝑉𝑝𝑝𝑝𝑝 𝑉𝑉𝑝𝑝𝑝𝑝

𝑑𝑑𝑑𝑑 𝛥𝛥𝐼𝐼𝑝𝑝𝑝𝑝 𝐼𝐼𝑝𝑝𝑝𝑝

>0↔ > − left of MPP, (3.7)
𝑑𝑑𝑑𝑑 𝛥𝛥𝑉𝑉𝑝𝑝𝑝𝑝 𝑉𝑉𝑝𝑝𝑣𝑣

𝑑𝑑𝑑𝑑 𝛥𝛥𝐼𝐼𝑝𝑝𝑝𝑝 𝐼𝐼𝑝𝑝𝑝𝑝

<0↔ < − right of MPP.
𝑑𝑑𝑑𝑑 𝛥𝛥𝑉𝑉𝑝𝑝𝑝𝑝 𝑉𝑉𝑝𝑝𝑝𝑝

28
 Chapter 3. Modelling and Simulation Tools

Figure 3.12. Flow Chart of the Incremental Conductance Algorithm.

29
 Chapter 3. Modelling and Simulation Tools

Figure 3.13. PSCAD Model for Incremental Conductance Algorithm.

The PSCAD model of incremental Conductance based MPPT algorithm is shown in Figure
3.13. The command signal Dv is used to increase or decrease the PV voltage by changing the
duty cycle. From equation (3.5), the duty cycle is inversely proportional to the output voltage.
This PV voltage Dv signal is changed to duty cycle command by multiplying with -1. Then
using the PI controller, duty cycle command is applied to PWM generator with switching
frequency to create gate pulses.

3.2.4. PV Inverter (DC/AC) Types and Control Topologies

To integrate the Solar PV system into a distribution grid a DC/AC inverter is needed, which
converts the DC voltage into AC voltage. The PV inverter is a mature technology consisting
of power electronic switches, which are controlled by control topology and switching patterns.
Depending on the capacity of the PV system and the application, the inverter has single-phase
and three-phase switching configurations. Figure 3.14 shows the grid connected single-phase
and three-phase PV inverter configurations along with the filter.

30
 Chapter 3. Modelling and Simulation Tools

In this research, a bipolar pulse width modulation (PWM) technique is implemented for a
single-phase PV inverter switching. The firing pulses were generated by comparing the
reference signal with the triangular carrier waveform. For this, the inverter considers the grid
voltage as a reference signal used to synchronise the PV output in terms of magnitude and
frequency.

(a)

(b)

Figure 3.14. (a) Single-phase and (b) Three-phase PV Inverter Systems with Filter.

In order to control three-phase PV inverters, current control techniques are used in most
applications. In this thesis, two basic models of the PV inverter with current control capabilities
are considered namely Current Regulated Current Source Inverter and Current Controlled
Voltage Source Inverter [114]. These PV inverters are capable of controlling real and reactive
powers. The topology of the power inverters is modelled with a three phase two-level bridge
block as shown in Figure 3.14 (b). The inverter bridge contains six IGBT switches with their
respective anti-parallel diodes, which compensate the dead time impacts. Note that the power
inverter hardware for the current source inverter and the voltage source inverter are identical.

3.2.4.1. Current Regulated Current Source Inverter (CRCSI)

Figure 3.15 shows the schematic diagram of a control circuit for a PV current source
inverter. The current source inverter together with the controller operation synthesizes a current

31
 Chapter 3. Modelling and Simulation Tools

source connected to the grid. The control circuit utilises the grid voltages to determine the phase
angle that is obtained using a three phase PI controlled phase locked loop (PLL). Therefore,
the inverter output will be synchronised with the grid voltage in terms of frequency and phase
angle. The measured DC output voltage and MPPT controller (Vmpp) are compared and the
error is given to the PI controller to generate reference active power. In the controller, Pref and
Qref represent the reference real and reactive powers; Pt and Qt are the measured powers. These
real and reactive power reference signals are compared with measured values and the errors
are given to the PI controller. The generated real power error (Isd_ref) signifies the real current
component, which is in-phase with the voltage in the case of unity power factor (PF) operation
and Isq_ref is the reactive current component.

Figure 3.15. Control Circuit of Current Regulated Current Source Inverter.

The generated currents in a dq synchronous rotating frame can be transformed into

stationary abc frame (Ia_ref, Ib_ref and Ic_ref) currents through the inverse synchronous frame. The
input to the hysteresis controller is the error between the reference currents and the measured
PV output currents. If the error surpasses the hysteresis band (HB) limit, a switching instant
will be given to the PV inverter switches so that the measured current follows the reference
currents. A fixed band type hysteresis controller technique has been considered due to its
simplicity and faster dynamic response. Despite of it advantages, this current controller has an
issue of variable switching frequency, which may result in injection of high harmonic current
components and increased losses. Moreover, the variation in the PV output could change the

32
 Chapter 3. Modelling and Simulation Tools

inverter reference currents, which in turn could introduce disturbances in the current controller
response.

3.2.4.2. Current Controlled Voltage Source Inverter (CCVSI)

Grid connected voltage source inverter generally resembles a voltage source connection in
the grid. The detailed control scheme of the current controlled pulse width modulation
technique used to control the voltage source inverter is shown in Figure 3.16. This method has
several control loops to regulate the PV inverter output current indirectly by controlling the
output voltage.

Figure 3.16. Control Circuit of Current Controlled Voltage Source Inverter.

A DC link voltage controller is applied to regulate the DC voltage across the capacitor to a
constant value. It can be achieved by controlling the injected active current in the inverter. If
the PV output voltage is lower than the reference DC link voltage, the controller decreases the
active current so the DC voltage and the capacitor current increases and vice versa [114]. The
output of the PI controller is used in the current controller as a reference.
𝑃𝑃𝑝𝑝𝑝𝑝
𝐼𝐼𝑑𝑑_𝑟𝑟𝑟𝑟𝑟𝑟 = 𝐾𝐾𝑝𝑝𝑝𝑝 �𝑉𝑉𝑑𝑑𝑑𝑑_𝑟𝑟𝑟𝑟𝑟𝑟 − 𝑉𝑉𝑑𝑑𝑑𝑑 � + 𝐾𝐾𝑖𝑖𝑖𝑖 ∫�𝑉𝑉𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 − 𝑉𝑉𝑑𝑑𝑑𝑑 � 𝑑𝑑𝑑𝑑 + (3.7)
𝑉𝑉𝑑𝑑

Where, Kpr and Kin are the proportional and integral constants.

Further, the measured voltages and currents at the grid integration point are used in the
controller. A three-phase dq synchronous frame theory is applied to convert the voltage and
currents into dc quantities. Here, the phase angle is obtained using three phase PI controlled
PLL to synchronise the PV inverter with the grid. The PI controllers are used to track the

33
 Chapter 3. Modelling and Simulation Tools

reference current with zero steady state error. Then, the grid voltages and the DC voltage are
used in the feed-forward loop to get the reference voltages as expressed in the below equation
(3.8).

𝑀𝑀𝑑𝑑 = 0.5𝑉𝑉𝑑𝑑𝑑𝑑 (2𝐼𝐼𝑑𝑑_𝑟𝑟𝑟𝑟𝑟𝑟 − 2𝐼𝐼𝑑𝑑 − 𝐿𝐿𝐿𝐿𝐼𝐼𝑞𝑞 + 𝑉𝑉𝑑𝑑 )

𝑀𝑀𝑞𝑞 = 0.5𝑉𝑉𝑑𝑑𝑑𝑑 (−2𝐼𝐼𝑞𝑞 + 𝐿𝐿𝐿𝐿𝐼𝐼𝑑𝑑 + 𝑉𝑉𝑞𝑞 ) (3.8)

For unity PF operation, the q axis component is set to zero so the active and reactive powers
can be controlled. Subsequently, the reference signals are then transformed into the stationary
frame using the inverse dq transformation. These control signals are compared with the carrier
high-frequency signal, which generates the firing pulses for the PV inverter.

3.3. IEEE-13 Bus Unbalanced Distribution Network

As a part of this research, an IEEE-13 bus radial distribution network is selected as the test
system [115]. The IEEE-13 bus distribution feeders have interesting characteristics of a real
distribution network making it suitable for the analysis. This radial unbalanced industrial
distribution network encompasses the following characteristics:

• Three phase four-wire system with a nominal voltage of 4.16 kV.

• This system has three-phase, two-phase and single-phase line configurations.
• Overhead and unground lines with unsymmetrical line geometries possess unbalanced
line impedances.
• Substation voltage regulator has three single phase voltage regulators connected in
star.
• Single-phase and three-phase shunt capacitor banks.
• In line transformer
• Have balanced and unbalanced spot and distributed loads, which can be modelled as
constant kW and kVAr (PQ), constant impedance (Z) or constant current (I) load.

All the loads can be three-phase loads connected in wye or delta and the single-phase loads
connected line-to-ground or line-to-line. The test system configurations and the data are given
by IEEE PES Distribution Systems Analysis Subcommittee. The line diagram of the IEEE-13
bus test system is shown in Figure 3.17. A simulation model of the IEEE-13 bus network
modelled in PSCAD/EMTDC software is shown in Figure 3.18. Simulations have been
performed for power flow analysis and the results verified with the available benchmark data
sheet.

34
 Chapter 3. Modelling and Simulation Tools

Figure 3.17. IEEE-13 Bus Test System Detailed Line Diagram.

Figure 3.18. Simulation Model of IEEE-13 Bus Distribution System.

35
 Chapter 3. Modelling and Simulation Tools

3.4. Load Modelling

Various types of static and nonlinear load models have been considered for different power
quality analyses. The most widely used static load models are polynomial load model and the
exponential load model. In this research, the static load models including constant power,
constant impedance and constant current load are considered to study the impact of PV power
injection on the voltage quality of the grid.

The constant power loads are connected as a parallel combination of a variable resistor (R)
and inductor (L). The respective values are computed using equations (3.7) and (3.8), which
are dependent on the grid voltage.

𝑉𝑉𝑝𝑝ℎ 2
𝑅𝑅 = (3.7)
𝑃𝑃0

𝑉𝑉𝑝𝑝ℎ 2
𝐿𝐿 = (3.8)
2𝜋𝜋𝜋𝜋𝑄𝑄0

Where, Po and Qo are the constant real and reactive powers of load and Vph is measured
RMS voltage across R and L. Figure 3.19 shows the example simulation design of constant
power load in PSCAD software.

Figure 3.19. Simulation Design of Constant Power (PQ) Load.

Similarly, constant current load model is modelled as a parallel branch of variable resistor
and inductor as shown in Figure 3.20 and is calculated using equations (3.9) and (3.10).
However, the constant impedance load is not voltage dependent and is modelled as a fixed
resistor and inductor values which are measured using the rated grid RMS voltage.
𝑉𝑉𝑝𝑝ℎ
𝑅𝑅 = 𝑃𝑃𝑖𝑖 (3.9)
� �𝑉𝑉 �
𝑜𝑜

𝑉𝑉𝑝𝑝ℎ
𝐿𝐿 = 𝑄𝑄 (3.10)
2𝜋𝜋𝜋𝜋� 𝑖𝑖�𝑉𝑉 �
𝑜𝑜

36
 Chapter 3. Modelling and Simulation Tools

Figure 3.20. Simulation Design of Constant Current Load.

In addition, to study the harmonic impact of PV systems in real-time grid operating scenario,
a certain amount of harmonic distortion is injected into the grid through nonlinear load models.
These nonlinear load models include single-phase and three-phase power electronic based
devices and single-phase motor type loads. Most of the residential and commercial loads have
nonlinear loads such as Televisions (TV), Personal Computers (PC), Compact Fluorescent
Lamps (CFL), air conditioning appliances with Adjustable speed motor drives (ASD) and other
power conversion devices, which are distorting the grid currents and voltage waveforms. The
voltage distortion depends on both the system impedance and the amount of harmonic current
injected. Consequently, the aggregated impact of a wide variety and different configurations of
domestic and industrial loads can influence the network harmonic distortion levels.
Accompanied by the proliferation of PV systems and their interaction in the distribution
network are the potential issues of harmonics and occurrence of harmonic resonance.
Therefore, accurate modelling of these loads should be considered to simulate a real-time
network scenario. Regarding that, various types of nonlinear loads and their current harmonic
characteristics considered in this research have been discussed.

Loads such as CFL, TV, PC and DC motor drive [38, 39] utilise single-phase diode full
bridge converters for their operation, which generate primarily odd harmonics of 3rd, 5th, 7th,
11th,..... orders. Simulation models of each of the loads developed in PSCAD software and their
respective individual current harmonic spectrum are shown in Figure 3.21. It is seen that these
loads are rich in harmonics and generally produce a pulsed waveform, which is due to the
extraction of non-sinusoidal currents.

37
 Chapter 3. Modelling and Simulation Tools

(a) TV or Computer Monitor

(b) Compact Fluorescent Lamp

(c) Personal Computer

(d) DC Motor Drive

Figure 3.21. Single-phase Power Electronic based Nonlinear Loads.

Also, the PSCAD simulation models of three-phase loads namely lift motors and ASDs with
their current harmonic spectrums are shown in Figure 3.22. It is observed that the lift motor
utilises a three-phase diode bridge converter circuit with a motor on the DC side, which is
modelled as a DC voltage source. However, ASDs are represented as a three-phase thyristor

38
 Chapter 3. Modelling and Simulation Tools

bridge inverter with a motor on the DC side. This ASD motor is modelled as a direct current
source. Figure 3.22(b) also shows the speed control logic for ASD. These types of loads
generate current harmonics mainly 5th and 7th order and a small quantity of 11th, 13th and higher
orders based on the configuration. The resultant harmonics of several identical devices can be
added or subtracted and the collective effect of different kinds of loads could violate the
network harmonic distortion levels.

(a) Lift Motor

(b) Adjustable Speed Drive

Figure 3.22. Three-Phase Power Electronic based Nonlinear Loads.

Apart from that, in this analysis single-phase induction motor [116] type loads are
considered to generate dynamics in the network. Typically, these motors are used in various
household appliances such as washing machines, pumps, blowers, refrigerators and fans, which
range from 0.5hp to 15hp. Propagation of single-phase induction motor loads are problematic
in the network because they increase the unbalance between phase voltages. During the start-
up, these motors draw large currents and increase the reactive power requirement. Figure 3.23
shows the simulation model induction motor. It is seen that a three phase induction motor is
used to model single phase induction motor, which uses a running capacitor to generate the
phase difference required for one phase. Also, this model has a starting capacitor, which can

39
 Chapter 3. Modelling and Simulation Tools

be used during start-up and stopover to achieve nominal rotating speed. In this figure, tset_2
indicates the torque demand of the mechanical load and w is the rotating speed of the motor.

Figure 3.23. Simulation Model of Single-phase Induction Motor Load.

3.5. Software and Hardware Tools

3.5.1. PSCAD [117]

PSCAD stands for Power Systems Computer Aided Design. In this project, PSCAD
software is selected for simulation purposes. PSCAD software is a professional simulation tool
used for electromagnetic transient (EMT) analysis. PSCAD is also known as PSCAD/EMTDC
which is more suitable for time domain instantaneous simulations, also known as
electromagnetic transient solutions for both power and control systems. PSCAD with its
graphical user interface allows the user to design a schematic, run a simulation, analyse the
results and manage the data in an integrated graphical environment. Thus, PSCAD is an
efficient software for testing and simulation of power quality, especially transients and
harmonics. This software is the most preferred EMT program by manufacturers in industry and
a popular tool for research and educational purposes. PSCAD provides accurate, effective and
reliable solutions.

3.5.2. Real Time Digital Simulator (RTDS) [118]

In the final part of this research, Controller Hardware-in-the-Loop simulations have been
performed using an integrated platform of RSCAD/RTDS and dSPACE hardware control
board.

40
 Chapter 3. Modelling and Simulation Tools

RTDS is a digital electromagnetic transient power system simulator, which operates in real-
time. The simulator has high speed PB5 and Gigahertz processor cards (GPC) to solve the
power system equations and controller system components modelled in RTDS. Also, the
simulator contains various analog and digital Gigabit Transceiver Input/output cards namely
GTAI, GTAO, GTDI, and GTDO, which interfaces the low power signals exchange from/to
external devices [119]. These signals are generally within a range of +/-10V peak and can be
easily calibrated with high accuracy. The test system modelling and compilations are
performed in RSCAD draft file. The results of simulations and controlling are observed in
runtime mode.

3.5.3. dSPACE Control Board Platform [120]

The dSPACE stands for Digital Signal Processing and Control Engineering. The dSPACE
control board is specially designed for test and development of control system and for real-
time simulations. The hardware contains DS1103 controller card and link boards (DS817 and
DS814) to interconnect with the host PC, a PX4 expansion box and the connector panel
CLP1103 [120]. The block diagram of dSPACE hardware and the software system is shown in
Figure 3.24. This platform depends on the MATLAB/Simulink software, which provides
several Simulink libraries and real-time RTI library blocks to design the model. The RTI library
has various Analog to Digital (ADC) and DAC channels to interface with the external system.
These signals are usually in low power levels. Then, by using the Simulink coder the developed
model can be converted to custom C code, which generates a .sdf file suitable for dSPACE
controlDesk software. The generated code file can be loaded into the dSPACE controlDesk
software for online calibration. The input controlling, output monitoring and recording can be
performed during the run-time in controlDesk layout.

Figure 3.24. Block Diagram of dSPACE Hardware Setup.

41
 Chapter 3. Modelling and Simulation Tools

3.6. Summary
This chapter has provided background theory and modelling of various components used
throughout this research. The PV system components including PV arrays, converters, inverters
and their control technologies were explained, and their implementation in PSCAD software
were presented. Then, the test system with various types of load models required for the
assessment of voltage quality issues and to perform accurate harmonic analysis were simulated
and discussed. After that, various software and hardware tools used for this research were
introduced. Furthermore, the application of these models and their study approaches will be
explained in the contribution Chapters 4, 5, 6 and 7.

42
 Chapter 4. Investigation of Voltage Quality Issues

CHAPTER 4

INVESTIGATION OF VOLTAGE QUALITY ISSUES

4.1. Introduction1

This chapter provides a thorough investigation of voltage quality issues associated with high
PV penetration. The evaluation of voltage quality issues such as voltage rise, voltage
unbalance, sag/swell issues on the IEEE-13 bus unbalanced distribution network with PV
penetration performed through simulations are discussed in the following section. Furthermore,
field investigations have been performed on real-time grid connected PV systems located at
the University of Queensland St. Lucia and Gatton Campuses, which have installed capacities
of 1.5MW and 3.3MW respectively. Finally, the summary of this chapter is presented at the
end.

1
This chapter is incorporated with the following references:
Partially incorporated in Section 4.2:
Annapoorna Chidurala, T. K. Saha, and N. Mithulananthan, "Power quality enhancement in
unbalanced distribution network using Solar-DSTATCOM," in Power Engineering Conference
(AUPEC), 2013 Australasian Universities, Hobart, TASMANIA, Sept. 29-Oct. 3, 2013, pp. 1-6.
Section 4.3 through Subsection 4.3.1:
Annapoorna Chidurala, T. K. Saha, and N. Mithulananthan, "Field investigation of voltage quality
issues in distribution network with PV penetration," in Power and Energy Engineering Conference
(APPEEC), 2015 IEEE PES Asia-Pacific, Brisbane, QLD, Australia, 15-18 Nov. 2015, pp. 1-5.
Subsection 4.3.2:
M. J. E. Alam, R. Yan, T. K. Saha, Annapoorna Chidurala, D Eghbal, "Learning from a 3.275
MW Utility Scale PV Plant Project," in CIGRE PARIS Session 2016, 21, rue d’Artois, F-75008
PARIS 2016.

43
 Chapter 4. Investigation of Voltage Quality Issues

4.2. Assessment of Voltage Quality Issues in IEEE-13 Bus

Unbalanced Distribution Network with High PV Penetration

The IEEE 13-bus test system [115] has been considered in this work as shown in Figure 4.1.
This is quite a distinctive unbalanced radial distribution system that reflects the real distribution
system. In this study, the developed comprehensive PV model presented in Chapter 3 [40] has
been considered as a PV source in order to reflect the real-time system. To investigate the effect
of PV penetration on the voltage quality of an unbalanced distribution network, the study has
been carried out for two different case studies. In this investigation, emphasis has been given
to voltage rise, voltage unbalance and sag/swell issues.

Figure 4.1. IEEE 13 Bus Test System.

4.2.1. Case Study – I: Voltage Rise and Unbalance Assessment

In the first case study, the study has been performed in three different scenarios by
considering PV integration at the same bus and multiple buses to investigate their impact on
voltage rise and unbalance factor. In the first two scenarios, three-phase PV system integrations
are considered at the beginning of the feeder and dispersed PV penetration at the end of the
feeder. The total installed capacity of the PV system in these scenarios is 900kWp. Finally, in
the last scenario, the penetration capacity has been increased by 250kW with two single-phase
PV inverter systems to study the severity of voltage rise and unbalance issues. For this analysis,
a typical solar irradiation profile has been applied and the corresponding PV output power is

44
 Chapter 4. Investigation of Voltage Quality Issues

shown in Figure 4.2. It is shown that PV power generation is reliant on the solar irradiation and
has a linear relationship with it.

Figure 4.2. Solar Irradiation Pattern and PV Power Output.

In the first scenario, a 900kW PV system has been considered and is only connected to bus
632 in the IEEE 13 bus network. Simulations have been performed and the corresponding RMS
voltage magnitudes during day and night times are presented in Table 4.1.

Table 4.1 Scenario-1: RMS Voltage Magnitude Variation due to PV System Integration
at Bus 632 during Daytime and Night-time

Phase A Phase B Phase C

RMS Voltage (p.u) RMS Voltage (p.u) RMS Voltage (p.u)
Node
PV at PV at PV at PV at PV at PV at
∆VA ∆VB ∆VC
t=4sec t=10sec t=4sec t=10sec t=4sec t=10sec
632 1.02746 1.02057 0.00689 1.04296 1.04136 0.00159 1.02148 1.01716 0.00432
633 1.02433 1.01717 0.00716 1.04103 1.03942 0.00161 1.01890 1.01450 0.00440
634 1.00078 0.99347 0.00731 1.02282 1.02102 0.00180 1.00060 0.99563 0.00497
645 - - - 1.03388 1.03217 0.00171 1.01958 1.01518 0.00440
646 - - - 1.03219 1.03043 0.00177 1.01762 1.01311 0.00451
671 0.99681 0.98943 0.00738 1.05405 1.05225 0.00180 0.98260 0.97738 0.00522
675 0.99026 0.98281 0.00746 1.05645 1.05460 0.00185 0.98064 0.97538 0.00526
684 0.99486 0.98749 0.00737 - - - 0.98057 0.97536 0.00521
611 - - - - - - 0.97865 0.97335 0.00530
652 0.98942 0.98191 0.00751 - - - - - -

At time t= 4 sec the solar irradiation is 1000 W/m2 (assumed as daytime period or rated PV
condition) and at t=10 sec the solar irradiation is 0 W/m2 (assumed as without PV condition or
night-time) as indicated in Figure 4.2. It can be seen that the voltage magnitude due to PV
power injection in phase A has intensified over phases B and C. The voltage magnitude of

45
 Chapter 4. Investigation of Voltage Quality Issues

phase A at bus 632 has increased by nearly 0.7% and this increment value is increased towards
the end nodes of the feeder. Conversely, the voltage magnitudes of phase B at buses 671, 675
and 634 located at the end of the feeder, have attained higher values, nonetheless the increments
in voltages (∆VB) are lower. However, it can be observed that the values at buses 675 and 671
have reached 1.05645p.u and 1.05405p.u respectively, which exceeded the ANSI and AS/NZS
standard limit of 5%. It can also be seen that PV power injection has provided voltage support
in phases A and C of buses located at the end of the feeder.

Table 4.2 Scenario-2: RMS Voltage Magnitude Variation due to PV Systems Integration
at Buses 634, 671 and 675 during Daytime and Night-time

Phase A Phase B Phase C

RMS Voltage (p.u) RMS Voltage (p.u) RMS Voltage (p.u)
Node
PV at PV at PV at PV at PV at PV at
∆VA ∆VB ∆VC
t=4sec t=10sec t=4sec t=10sec t=4sec t=10sec
632 1.02770 1.02057 0.00713 1.04329 1.04136 0.00193 1.02306 1.01716 0.00590
633 1.02567 1.01717 0.00850 1.04263 1.03942 0.00321 1.02170 1.0145 0.00720
634 1.01032 0.99347 0.01685 1.03314 1.02102 0.01212 1.01117 0.99563 0.01554
645 - - - 1.03394 1.03217 0.00177 1.02090 1.01518 0.00572
646 - - - 1.03219 1.03043 0.00176 1.01886 1.01311 0.00575
671 1.00030 0.98943 0.01087 1.05450 1.05225 0.00225 0.98530 0.97738 0.00791
675 0.99389 0.98281 0.01108 1.05762 1.05462 0.00300 0.98511 0.97538 0.00973
684 0.99726 0.98749 0.00977 - - - 0.98401 0.97536 0.00865
611 - - - - - - 0.98280 0.97335 0.00945
652 0.99319 0.98191 0.01128 - - - - - -

Subsequently, in the second scenario, three-phase PV systems of each 300kWp rated power
capacity are considered to connect at buses 634, 671 and 675 in the IEEE 13 bus test system.
Simulations have been performed for the duration of 12sec with the solar irradiation profile as
indicated in Figure 4.2. The measured voltage magnitudes at various buses of the IEEE network
are tabulated in Table 4.2. It is observed that voltage rise due to individual PV systems at the
end of the feeders is larger compared to the first scenario. Similar to the first scenario, phase A
has been affected more followed by phase C and phase B. Significantly, bus 634 which is a
low voltage (480V) bus has experienced higher voltage increments due to PV penetration.
Phase B of remote buses 675 and 671 have voltage magnitudes of 1.05450p.u and 1.05762p.u
respectively, which are over the limits as per the standards. In comparison with Table 4.1, the
voltage variations due to reverse power flow have also increased the magnitudes in the

46
 Chapter 4. Investigation of Voltage Quality Issues

upstream network. On the other hand, neighbouring buses with higher loads have experienced
voltage support from PV power injection.

Table 4.3 Scenario-3: RMS Voltage Magnitude Variation due to the Combination of Three-phase
and Single-phase PV Systems Integration at Multiple Locations during Daytime and
Night-time

Phase A Phase B Phase C

RMS Voltage (p.u) RMS Voltage (p.u) RMS Voltage (p.u)
Node
PV at PV at PV at PV at PV at PV at
∆VA ∆VB ∆VC
t=4sec t=10sec t=4sec t=10sec t=4sec t=10sec
632 1.03167 1.02057 0.01110 1.04392 1.04136 0.00256 1.02404 1.01716 0.00688
633 1.02964 1.01717 0.01247 1.04324 1.03942 0.00382 1.02488 1.01450 0.01038
634 1.01330 0.99347 0.01983 1.03366 1.02102 0.01264 1.00876 0.99563 0.01313
645 - - - 1.03453 1.03217 0.00236 1.02242 1.01518 0.00724
646 - - - 1.03281 1.03043 0.00238 1.02033 1.01311 0.00722
671 1.00926 0.98943 0.01983 1.05595 1.05225 0.00370 0.99034 0.97738 0.01296
675 1.00494 0.98281 0.02213 1.06052 1.05462 0.00590 0.98849 0.97538 0.01311
684 1.01019 0.98749 0.02270 - - - 0.98708 0.97536 0.01172
611 - - - - - - 0.98550 0.97335 0.01215
652 1.00476 0.98191 0.02285 - - - -

Finally, in the third scenario, the penetration has been increased by 250kW in addition to
the individual PV connections at buses 634, 671 and 675. For this, two single-phase PV systems
of rated capacity 150kW and 100kW are connected to phase A of bus 684 and phase B of bus
675 respectively. Simulations have been performed similar to previous scenarios, and the load
flow results are presented in Table 4.3. It can be seen that the voltage difference ∆VA in phase
A is almost doubled compared to the second scenario. Even a small increment in PV penetration
has a significant effect on the voltage magnitudes. Similarly, the voltage magnitude at bus 675
is increased to above 1.06p.u, which is even higher than the limits specified by standards. Also,
the voltage variations with increased PV penetration are reflected on the adjacent buses.

Therefore, it is confirmed that dispersed PV penetration at the end of the feeders can cause
significant overvoltage’s than if provided at the beginning of the feeder. In addition, the voltage
amplitude is more exacerbated with single-phase PV inverter systems.

Furthermore, it is observed that the voltage variation (∆V) by PV systems in each phase is
different due to unbalanced loading on the network. The high density of PV system integration
into a distribution grid increases the VU at the PCC and has high values at the end of the feeder.

47
 Chapter 4. Investigation of Voltage Quality Issues

Therefore, the analysis is performed to verify the impact of PV penetration on network voltage
unbalance. Regarding this, the voltage unbalance factor is calculated based on the equation
specified by AS/NZS 61000.4.30 standard (see Chapter 2, equation (2.2)), which utilizes phase
to phase voltages. The percentage of voltage unbalance at various nodes of the IEEE network
for the above mentioned scenarios are plotted in Figure 4.3.

6
Percentage of Voltage Unbalance
5
% Voltage Unbalance

0
V632 V633 V634 V671 V692 V680 V675
Node No.
%VU in First Scenario %VU in Second Scenario %VU in Third Scenario

Figure 4.3. %Voltage Unbalance Analysis at Various Buses.

It can be seen that in all scenarios, the active power contribution from PV systems into the
grid differed the voltage magnitude and are much higher at buses located at the end of the
feeder. The values follow similar trends in all the scenarios. It is observed that the voltage
unbalance curve in the first scenario, in which the PV system is integrated at a single location
(at bus 632), has higher values compared to the dispersed PV integrations in the second and
third scenarios. It was found that the percentage of VU is above 5.1 in the first scenario, which
is over the limits according to AS/NZS 61000.4.30. Interestingly, the voltage unbalance is
increased in the third scenario compared to the second scenario. The voltage unbalance at bus
675 in the third scenario has reached 4.96%, which is close to the allowable limit of 5%. Further
increase in PV penetration can aggravate the voltage unbalance levels in the network. It is
estimated that distributed random installation of single phase PV systems can adversely affect
the network voltage unbalance level.

4.2.2. Case study-II: Voltage Sag/Swell Assessment by Fault Conditions

The voltage sag/swell issues are common in distribution networks, which are caused by
symmetric and unsymmetrical faults on the network. Moreover, PV systems as per the current
standards are not allowed to stay connected during fault conditions. However, sudden
connections/disconnections of PV systems in response to the faults can also impact the network

48
 Chapter 4. Investigation of Voltage Quality Issues

voltage levels. Therefore, in this analysis a 900kWp PV system is considered to be connected

at bus 632 of the IEEE 13 bus network similar to the first scenario of the above case study.

In this case study, simulations have been carried out for with and without PV inverter
penetration in an unbalanced distribution system for each of the three types of faults. In each
case the fault is initiated at t = 4.5 sec and the duration of the fault is 100ms. In figures below
Va, Vb, Vc represent the three phase RMS voltages at each bus at before fault condition and Vaf,
Vbf, Vcf denote the fault magnitudes during a fault. Also, LB and UB indicate the lower and
upper band of steady state voltage limits respectively. According to IEEE 1159-2009 [97] and
AS/NZS 61000.4.30 [98] standards, the typical voltage sag threshold is 90% of the nominal
and the swell threshold is equal to or greater than 110% of nominal.

4.2.2.1. Single Line to Ground Fault (SLG)

SLG fault is considered to occur on phase-A of bus 633. In reality, a fault on the grid causes
the PV system to be completely isolated from the grid as considered in the first scenario. At
the fault the affected phase voltage drops to zero. Also, due to an unsymmetrical fault
unbalanced voltage sag/swells would be created on the other adjacent bus lines and increase
system losses. The bar chart in Figure 4.4 shows the RMS voltage magnitude at various buses
before and during the fault. It is observed that the voltages of phase A and phase C are dropped
by around 15% and 11% of the rated voltages respectively, creating a voltage sag condition.
Conversely, a voltage swell of about 10% is observed in phase B.

In the second scenario, a conventional PV inverter (unity P.F) is considered to remain

connected during the fault. The PV system injects only active power into the grid that will
affect the operation of unidirectional voltage regulators and produce overvoltage on lightly
loaded phases. During the unbalanced fault, the PV inverter injects balanced currents into the
system due to the balanced controller operation. Figure 4.5 summarizes the simulated results
of the RMS voltages at remote buses before and during the fault. By comparing with the first
scenario, there is a little improvement in the voltages due to real power injection. As can be
observed, phase A voltage was dropped by 12% and there was a swell on phase B of about
11%, which is a lightly loaded phase in the test system. During the fault, it is noticed that the
voltage swell is further increased by PV power injection at the PCC, which is violating the
limits.

49
 Chapter 4. Investigation of Voltage Quality Issues

Figure 4.4. RMS Voltages at Prefault and during SLG Fault without PV Inverter.

Figure 4.5. RMS Voltages at Prefault and during SLG Fault with PV Inverter.

4.2.2.2. Double Line to Ground Fault (DLG)

In the same way, an unsymmetrical DLG fault is applied on phases A and B of bus 633. The
bar charts shown in Figures 4.6 and 4.7 depict the RMS voltages at remote buses for the first
and second scenarios as explained above. During the fault, the PV inverter system supplying
balanced currents into the grid caused voltage fluctuations at PCC as a result the other buses
are also affected. Results show that phase A voltage has experienced a notable sag of around
25% and a swell on phase C of about 10% of the rated voltage almost in both scenarios. It is
noticed that due to an unsymmetrical fault, phase C has been much more affected and phase B
was only slightly influenced by PV power injection, which is opposite to the case of SLG fault.

50
 Chapter 4. Investigation of Voltage Quality Issues

Figure 4.6. RMS Voltages at Prefault and during DLG Fault without PV Inverter.

Figure 4.7. RMS Voltages at Prefault and during DLG Fault with PV inverter.

4.2.2.3. Three-phase Symmetrical Fault (TPG)

Finally, the simulations have been carried out for the TPG fault to strike at bus 633, which
is often seen as the worst case. Due to the fault, all the bus voltages were affected and resulted
in sag on all the phases as presented in Figure 4.8 for the first scenario as mentioned earlier. It
is observed that the voltage of phase A plunged to 10% and phase C voltage collapsed to 12%
of nominal voltages.

Similarly, in the second scenario when the PV inverter remains connected during the fault,
it is noticed from Figure 4.9 that all the phases have experienced a sag in voltage. The phase A
and phase C voltages have dropped by around 8% and 7% respectively, but still this drop is
characterised as sag and the magnitudes are above the acceptable limits. It is shown that the
effect of PV power injection during the fault has slightly improved the voltage levels in
adjacent buses. Nevertheless, the voltage swells are negatively affected. Further investigation

51
 Chapter 4. Investigation of Voltage Quality Issues

will be performed on the real-time grid PV systems through field measurements in order to
evaluate the severity of various voltage quality issues.

Figure 4.8. RMS Voltages at Prefault and during TPG Fault without PV Inverter.

Figure 4.9. RMS Voltages at Prefault and during TPG Fault with PV inverter.

4.3. Field Investigation of Voltage Quality Issues in Distribution

Network with PV Penetration

The aim of this research was to investigate the voltage quality issues in a distribution
network with high PV system penetration. In this investigation, the power quality field
measurements for various PV operating conditions are obtained and analysed for voltage
quality issues. The study has focussed on two different scales of PV sites located at the
University of Queensland (UQ) St. Lucia Campus and Gatton campuses, which have an
installed capacity of 1.5MW and 3.25MW respectively. The description of UQ St. Lucia
network and an approach of the measurements are presented in Section 4.3.1. The results and
analysis of various voltage issues are discussed in Section 4.3.1.1. Next, the layout of the UQ

52
 Chapter 4. Investigation of Voltage Quality Issues

Gatton campus PV system is presented in Section 4.3.2. The corresponding results and analysis
are explained in subsection 4.3.2.1.

4.3.1. Description of University of Queensland, St. Lucia Campus Network and

Measurement Setup

(a) Aerial View of PV Site

53
 Chapter 4. Investigation of Voltage Quality Issues

(b) Distribution Feeder of the University Network

Figure 4.10. Overview of the University of Queensland PV Network, St. Lucia Campus.

The University of Queensland (UQ), St. Lucia campus has a large distribution network
encircled in a 2km radius. The network receives an 11kV line from the Energex network
connected to substation 7. From this substation, there are 8 parallel radial feeders used to feed
electricity to the individual buildings, which consist of nearly 40 substations. The university had
installed a 1.22MWp PV system in 2011, which is one of the largest rooftop integrated systems
in Australia. In 2013, the UQ St. Lucia campus expanded its PV system capacity to 1.5MWp
with around 6305 PV panels, which is deployed across eight buildings of the network shown in
Figure 4.10(a).

Figure 4.10 (b) shows one of the feeders, where two car park buildings have PV systems,
which has been considered for the study. Each car park building has an installed PV system
capacity of 339kWp connected to the low voltage side of substation 11. This substation includes
two distribution transformers of 750kVA 11kV/415V delta-wye connected as shown in Figure
4.10 (b). From substation 11 the rest of the 11kV line continues to substations 26, 30 and 17.
The UQ network is equipped with several online digital power meters and power quality
analysers interfaced to the internet for monitoring.

54
 Chapter 4. Investigation of Voltage Quality Issues

In this research, field measurements were conducted at car park 98A. This building has nearly
26 three-phase PV inverters of 12.5kW capacity each, which are connected directly to the grid.
Each inverter has two maximum power point tracking (MPPT) controllers and each MPPT is
organized with two arrays of around 30 PV panels of 250W rating. A Fluke power quality (PQ)
analyser of 430 series-II type was used for the measurement. This PQ instrument meets the
requirements of the standard AS/NZS 61000-4-15:2012 [121] and also certified by the IEC
61000-4-30 Class A standard for flicker and power quality. Figure 4.11 shows the test setup and
power board, where the PQ meter is connected at the output of one of those PV inverters.

Figure 4.11. Measurement Test Setup.

The PQ analyser recorded a number of parameters, which included power, voltage, current,
frequency and flicker data with 5sec resolution. To characterise the voltage variations for
various operating days, the measurements were continued for 15 days. The logged data was
extracted using Power log software. The PV output power and solar irradiation data were also

55
 Chapter 4. Investigation of Voltage Quality Issues

captured using weather sensors. In addition, the flicker data of one year at the PV array and the
11kV side was measured for further analysis.

4.3.1.1 Results and Analysis

4.3.1.1.1. Characteristics of Voltage Rise and Fluctuations at 1.5MW PV System

In distribution networks, voltage rise is the most prevalent limiting issue due to reverse
power flow with PV power injection. The PV power output strongly depends on the solar
irradiation and temperature which can impact the grid voltages. Figure 4.12 shows the time
versus PV output power and solar irradiation for the selected different PV operating days.

350 1200

Solar Irradiance (W/m2)

Time Vs Solar Irradiance and PV Array Output Power
300 1000
PV Output Power (kW)

250
800
200
600
150
400
100
50 200

0 0
5:00
6:00
7:00
8:00
9:00

5:00
6:00
7:00
8:00
9:00

5:00
6:00
7:00
8:00
9:00

5:00
6:00
7:00
8:00
9:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00

10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00

10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00

10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
Date and Time
24/04/2014 23/04/2014 25/04/2014 28/04/2014
Power Solar irradiance

Figure 4.12. Time versus Solar Irradiation and PV Output Power.

The measured per unit three-phase voltage magnitudes at the PV inverter output for the
corresponding days are plotted in Figure 4.13. The voltage magnitude at early morning and
night times is close to 0.99p.u. It can be seen that the voltage levels have increased with respect
to the PV power injection. During the rated PV output, the voltage level reached about 1.02p.u.,
which is within the limits as per AS/NZS 61000.3.100 [96] standard. The increment in the
voltage is appeared to be lower due to the load operation. Since the university peak load
demand is from 6 am to 6 pm which is similar to a commercial load pattern, which matches
with the PV operational output.

56
 Chapter 4. Investigation of Voltage Quality Issues

Voltage Magnitudes at the PV Inverter Output

1.02

Voltage (p.u.) 1.01

0.99

0.98

0.97
0:00
2:17
4:34
6:51
9:08

1:09
3:26
5:43
8:00

0:00
2:17
4:35
6:52
9:09

1:09
3:26
5:43
8:01
11:25
13:43
16:00
18:17
20:34
22:51

10:17
12:34
14:52
17:09
19:26
21:43

11:26
13:43
16:00
18:18
20:35
22:52

10:18
12:35
14:52
17:09
19:26
21:44
24/04/2014 23/04/2014 25/04/2014 28/04/2014
Date and Time
Va_PV Vb_PV Vc_PV

Figure 4.13. Voltage Magnitudes at the PV Inverter Output.

1.05 Voltage Magnitudes at the Sub 11 Transformer TX1

1.04
Voltage (p.u.)

1.03

1.02

1.01

1
0:00
2:20
4:40
7:00
9:20

1:40
4:00
6:20
8:40

1:00
3:20
5:40
8:00

0:20
2:40
5:00
7:20
9:40
11:40
14:00
16:20
18:40
21:00
23:20

11:00
13:20
15:40
18:00
20:20
22:40

10:20
12:40
15:00
17:20
19:40
22:00

12:00
14:20
16:40
19:00
21:20
23:40
24/04/2014 23/04/2014 25/04/2014 28/04/2014
Date and Time
VTx1_a VTx1_b VTx1_c

Figure 4.14. Voltage Magnitudes at Substation 11 Transformer TX1.

1.05 Voltage Magnitudes at the Sub 17 LV Side

1.04
Voltage (p.u.)

1.03

1.02

1.01

1
0:00
2:30
5:00
7:30

1:00
3:30
6:00
8:30

2:00
4:30
7:00
9:30

0:30
3:00
5:30
8:00
10:00
12:30
15:00
17:30
20:00
22:30

11:00
13:30
16:00
18:30
21:00
23:30

12:00
14:30
17:00
19:30
22:00

10:30
13:00
15:30
18:00
20:30
23:00

24/04/2014 23/04/2014 25/04/2014 28/04/2014

Date and Time
V17_a V17_b V17_c

Figure 4.15. Voltage Magnitudes at Substation 17.

In addition, to assess the PV impact on adjacent nodes of the network, the voltage
measurements were obtained at the upstream (Sub 11 at TX1) and downstream (Sub 17) nodes
as shown in Figures 4.14 and 4.15 respectively. From Figure 4.15, it is observed that the voltage
levels on the network have lower values during the day time due to the voltage drop across

57
 Chapter 4. Investigation of Voltage Quality Issues

loads. The active power injection from PV has actually improved the voltage level. The values
are within a range between 1.02-1.035p.u., which are below the upper limits.

From Figure 4.15, it can be seen that the voltages at the downstream nodes during low load
period (night-time) have values in the range of 1.03-1.035p.u. It is important to mention that
substation 17 comprises of a tap changing transformer, which was set to increase the voltage
magnitudes by 5%. As a result, substation 17 has higher voltage levels compared to the voltages
at the PV inverter output. During the PV operation period, the voltage levels are in the range
of 1.02- 1.03p.u., which are within the limits. However, the load demand pattern in residential
networks could be inverse compared to the PV operational profile. In this scenario, the network
voltage will rise with high PV penetration.

18 PV Inverter Output Current

PV Output Current (A)

13

3
11:46:07
13:43:47
15:41:27
17:39:07
19:36:47
21:34:27
23:32:07

11:18:17
13:15:57
15:13:37
17:11:17
19:08:57
21:06:37
23:04:17

10:50:17
12:47:57
14:45:37
16:43:17
18:40:57
20:38:37
22:36:17

10:22:17
12:19:57
14:17:37
16:15:17
18:12:57
20:10:37
22:08:17
0:00:07
1:57:47
3:55:27
5:53:07
7:50:47
9:48:27

1:29:57
3:27:37
5:25:17
7:22:57
9:20:37

1:01:57
2:59:37
4:57:17
6:54:57
8:52:37

0:33:57
2:31:37
4:29:17
6:26:57
8:24:37
-2

24/04/2014 23/04/2014 25/04/2014 28/04/2014

Date and Time
Ia_PV Ib_PV Ic_PV

Figure 4.16. Measured PV Inverter Output Currents.

Furthermore, it is noticed that the voltage levels are fluctuating with respect to solar
irradiation. Also, the PV inverter output currents have a strong correlation with the solar
irradiation transients and are influenced by the voltage magnitudes. Figure 4.16 shows the
measured PV inverter output currents for the corresponding days. These rapid and frequent
voltage variations due to cloud transients can exceed the standard steady state limits and can
cause voltage flicker.

4.3.1.1.2. Voltage Flicker Analysis at 1.5MW PV System

The flicker measurements at the output of the PV inverter for corresponding days have been
collected with a Fluke PQ meter and are plotted as shown in Figure 4.17. The severity of
voltage flicker is illustrated by the instantaneous flicker index Pinst, short-term index Pst (over
a 10 min period) and the long term index Plt (over a 2 hour period).

58
 Short-term Flicker (Pst)
Long-term Flicker (Plt) Instantaneous Flicker (Pinst)

0.15
0.25
0.35
0.45
0.55
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7

0.2
0.3
0.4

0.25
0.35
0.45
0:00:07
0:00:07 2:40:07 0:00:07
2:49:37 5:20:07 2:49:37
5:39:07 8:00:07 5:39:07

Figure 4.17.
8:28:37 10:40:07 8:28:37
11:18:07 13:20:07 11:18:07
14:07:37 14:07:37

24/04/2014
16:00:07

24/04/2014

24/04/2014
16:57:07 18:40:07 16:57:07
19:46:37 21:20:07 19:46:37
22:36:07 0:00:17 22:36:07
1:25:47 2:40:17 1:25:47
4:15:17 5:20:17 4:15:17

Plt AN
Pst AN
7:04:47 8:00:17 7:04:47
9:54:17 10:40:17 PInst AN 9:54:17
12:43:47 13:20:17 12:43:47
15:33:17

23/04/2014
15:33:17 16:00:17
23/04/2014

23/04/2014
18:22:47 18:40:17 18:22:47
21:12:17 21:20:17 21:12:17

59
0:01:47 0:00:17 0:01:47
PInst BN

Plt BN
2:51:17 2:40:17 2:51:17

Pst BN
5:40:47 5:20:17 5:40:47
Date and Time

Date and Time

Date and Time

8:30:17 8:00:17 8:30:17
11:19:47 10:40:17 11:19:47
14:09:17 13:20:17 14:09:17
25/04/2014

25/04/2014
25/04/2014
16:58:47 16:00:17 16:58:47
PInst CN

19:48:17 18:40:17 19:48:17

(b) Pst Index at the PV Inverter Output

(c) Plt Index at the PV Inverter Output

Plt CN
22:37:47 21:20:17 22:37:47
0:00:17

Pst CN
1:27:17 1:27:17
(a) Pinst Flicker Index at the PV Inverter Output

4:16:47 2:40:17 4:16:47

7:06:17 5:20:17 7:06:17
9:55:47 8:00:17 9:55:47
12:45:17 10:40:17 12:45:17
13:20:17 15:34:47
28/04/2014

15:34:47

28/04/2014
28/04/2014
18:24:17 16:00:17 18:24:17
21:13:47 18:40:17 21:13:47

Measured Voltage Flicker Indices at the PV Inverter Output.

21:20:17

attained higher values during a cloudy day. From Figure 4.17 (a), it is noticed that the
Results show that the voltage flicker levels during PV operating times are increased and
Chapter 4. Investigation of Voltage Quality Issues
 Chapter 4. Investigation of Voltage Quality Issues

instantaneous flicker index has reached 0.55 due to rapid voltage fluctuations. Nonetheless, the
short-term flicker index is within the 0.3 to 0.45 range while the Plt is below 0.43, which are
within the limits specified by standard AS/NZS 61000.3.3 [103]. It is also noticed that when
there is no PV power, the measured Pst and Plt have values lower than 0.3, which can be
represented as background flicker. Conversely, the measured flicker levels are violating the
utility prescribed flicker emission limits as per AS/NZS 61000.3.7 [104].

0.3
Short-term Flicker (Pst)

(a) Pst Index at 11kV Line

0.25

0.2

0.15

0.1

0.05
11:20
14:10
17:00
19:50
22:40

10:00
12:50
15:40
18:30
21:20

11:30
14:20
17:10
20:00
22:50

10:10
13:00
15:50
18:40
21:30
0:00
2:50
5:40
8:30

1:30
4:20
7:10

0:10
3:00
5:50
8:40

1:40
4:30
7:20
24/04/2014 23/04/2014 25/04/2014 28/04/2014
Date and Time
Pst AN Pst BN Pst CN

0.35
(b) Plt Index at 11kV Line
Long-term Flicker (Plt)

0.25

0.15

0.05
11:20
14:10
17:00
19:50
22:40

10:00
12:50
15:40
18:30
21:20

11:30
14:20
17:10
20:00
22:50

10:10
13:00
15:50
18:40
21:30
0:00
2:50
5:40
8:30

1:30
4:20
7:10

0:10
3:00
5:50
8:40

1:40
4:30
7:20

24/04/2014 23/04/2014 25/04/2014 28/04/2014

Date and Time
Plt AN Plt BN Plt CN

Figure 4.18. Measured Flicker Indices at 11kV Line.

Also, these voltage variations might be propagated towards the upstream network.
Therefore, to investigate the flicker level at the HV side of substation 11, the recorded short
term and long term flicker data have been extracted from the digital PQ meters. Figure 4.18
shows the measured Pst and Plt at the 11kV line for the above mentioned days. It can be seen
that both the Pst and Plt during the period from 6pm to 7am have values about 0.1. Similarly,
these indices are increasing with respect to the PV power variations and are satisfying the
limits. Also, the voltage variations in the network are observed to be significant and the Pst
values are fluctuating between 0.1 to and 0.25.

60
 Chapter 4. Investigation of Voltage Quality Issues

350 Time Vs Solar Irradiation and PV Array Output Power 1400

Solar Irradiation (W/m2)

PV Output Power (kW) 300 1200
250 1000
200 800
150 600
100 400
50 200
0 0
5:00
6:00
7:00
8:00
9:00

5:00
6:00
7:00
8:00
9:00

5:00
6:00
7:00
8:00
9:00

5:00
6:00
7:00
8:00
9:00

5:00
6:00
7:00
8:00
9:00

5:00
6:00
7:00
8:00
9:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00

10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00

10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00

10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00

10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00

10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
Date and Time
1/4/2014 9/4/2014 3/5/2014 5/8/2014 8/12/2014 21/03/2015
Power Solar Irradiance

Figure 4.19. Time versus Solar Irradiation and PV Output Power for Typical Days.

2
Short-term Flicker (Pst)

(a) Pst Index at the PV Array for Typical Cloudy Days

1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
12:30
16:40
20:50

13:30
17:40
21:50

10:20
14:30
18:40
22:50

11:20
15:30
19:40
23:50

12:20
16:30
20:40

13:20
17:30
21:40
0:00
4:10
8:20

1:00
5:10
9:20

2:00
6:10

3:00
7:10

4:00
8:10

0:50
5:00
9:10
01/4/2014 09/4/2014 03/5/2014 05/8/2014 08/12/2014 03/21/2015
Date and Time
Pst An Pst Bn Pst Cn

0.9
Long-term Flicker (Plt)

(b) Plt Index at the PV Array for Typical Cloudy Days

0.8
0.7
0.6
0.5
0.4
0.3
0.2
12:30
16:40
20:50

13:30
17:40
21:50

10:20
14:30
18:40
22:50

11:20
15:30
19:40
23:50

12:20
16:30
20:40

13:20
17:30
21:40
0:00
4:10
8:20

1:00
5:10
9:20

2:00
6:10

3:00
7:10

4:00
8:10

0:50
5:00
9:10

01/4/2014 09/4/2014 03/5/2014 05/8/2014 08/12/2014 03/21/2015

Date and Time
Plt AN Plt BN Plt CN

Figure 4.20. Measured Voltage Flicker Indices at the PV Array.

Although flicker levels during the measurement period are appeared to be lower, these
values might exceed the standard limits for critical operating conditions. The level of voltage
variation depends on the PV system capacity and rate of change in solar irradiation. For this

61
 Chapter 4. Investigation of Voltage Quality Issues

reason, the voltage flicker data of 1 year measured at the PV array has been analysed. Figure
4.19 shows the critical solar irradiation and PV output power patterns of cloudy days. The Pst
and Plt indices are plotted in Figure 4.20 for the respective days. Results show that the values
of Pst were exceeded the standard limit of 1 for several events. Respectively, the Plt was also
overlapped the limit of 0.65. Frequent incidence of voltage flicker on the network may lead to
hunting of tap changers and degradation of the PQ. Therefore, it is increasingly important to
quantify the probability of voltage flicker in the distribution network.

In this regard, a clustering algorithm has been adopted to verify the severity of voltage
flicker levels with the time of day. In this analysis, the measured daily flicker data of one year
has been utilized and is superimposed on a 24 hour time frame. The stacked data contains every
10 min and 2 hour time interval based groups. Then by using the clustering technique, the short
term values at that specific time have been divided into four clusters as follows

Cluster 1 – 0.05<Pst≤0.25; Cluster 2 – 0.25<Pst≤0.5;

Cluster 3 – 0.5<Pst≤1; and Cluster 4 – Pst>1.

Similarly, the long-term flicker data has been segregated into five clusters as follows

Cluster 1 – 0.05<Plt≤0.15; Cluster 2 – 0.15<Plt≤0.3;

Cluster 3 – 0.3<Plt≤0.45; Cluster 4 – 0.45<Plt≤0.65 and Cluster 5 – Plt>0.65.

Next, the probability counts for each cluster at that moment are calculated and this process
repeated for all time instants. The probability counts of Pst and Plt measured at the PV array are
plotted in Figures 4.21 (a) and (b) respectively.

(a) Probability Count of Short-term Flicker Index at PV

1000
Array
PROBABILITY COUNT

100

10

1
0:00 2:24 4:48 7:12 9:36 12:00 14:24 16:48 19:12 21:36 0:00
TIME (H:MM)
Cluster 1 (0.05<Pst<=0.25) Cluster 2 (0.25<Pst<=0.5)
Cluster 3 (0.5<Pst<=1) Cluster 4 (Pst>1)

62
 Chapter 4. Investigation of Voltage Quality Issues

(b) Probability Count of Long-term Flicker Index at PV Array

1000

PROBABILITY COUNT
100

10

0:00 2:24 4:48 7:12 9:36 12:00 14:24 16:48 19:12 21:36
TIME (H:MM)
Cluster 1 (0.05<Plt<=0.15) Cluster 2 (0.15<Plt<=0.3)
Cluster 3 (0.3<Plt<=0.45) Cluster 4 (0.45<Plt<=0.65)
Cluster 5 (Plt>0.65)

Figure 4.21. Clustered based Probability Count of Flicker Indices at PV Array.

The x-axis is the time period and the y-axis is the probability count shown in logarithmic
scale. From Figure 4.21 (a), it can be seen that during PV operation, the probability counts of
cluster 2 have increased while those of cluster 1 have decreased. A few number of times the
flicker level was within cluster 3. Nonetheless, there were a number of incidents where the
flicker level exceeded the limit of 1, which are a total of around 22 times in a year only during
6am to 6pm. In comparison with the flicker emission limits from AS/NZS 61000.3.7 standard,
the flicker levels in cluster 2 are violating the limits for nearly 70 to 95 times at every time
instant in a year. Similarly, results from Figure 4.21 (b) show that cluster 2 is the probable level
of flicker and has a maximum number of counts during PV operation time. In the worst
conditions, cluster 5 has accounted for a total of 17 events, which are negligible.

In addition, the inherent flicker levels at the high voltage side are analysed. The clustering
process has been repeated and the probability count of Pst and Plt are plotted as shown in Figure
4.22 (a) and (b) respectively. It can be observed that the short-term flicker index lies under 0.25
(i.e., cluster 1) for several events, which is within the limits specified by AS/NZS 61000.3.7
standard. Cluster 2 is the next intensified flicker level with less than 10 events per time period.
However, these statistics are minor compared with Figure 4.21 (a).

63
 Chapter 4. Investigation of Voltage Quality Issues

(a) Probability Count of Short-term Flicker Index at 11kV

1000
Line

PROBABILITY COUNT
100

10

0:00 4:48 9:36 14:24 19:12 0:00

TIME (H:MM)

Cluster 1 (0.05<Pst<=0.25) Cluster 2 (0.25<Pst<=0.5)

Cluster 3 (0.5<Pst<=1) Cluster 4 (Pst>1)

(b) Probability Count of Long-term Flicker Index at 11kV

1000 Line
PROBABILITY COUNT

100

10

1
0:00 2:24 4:48 7:12 9:36 12:00 14:24 16:48 19:12 21:36
TIME (H:MM)
Cluster 1 (0.05<Plt<=0.15) Cluster 2 (0.15<Plt<=0.3)
Cluster 3 (0.3<Plt<=0.45) Cluster 4 (0.45<Plt<=0.65)
Cluster 5 (Plt>0.65)

Figure 4.22. Clustered based Probability Count of Flicker Indices at 11kV Line.

From Figure 4.22 (b), it can be noticed that the flicker levels in cluster 2 are increasing with
respect to the PV operation time, which are below the limits as per the standards. In contrast to
Figure 4.21 (b), cluster 1 has the highest count recorded. So, the impact of PV fluctuations have
less influence on the upstream network flicker level. During no PV times, the flicker levels
cross over the 0.65 limit (cluster 5) and are calculated to be less than 8 events per time period.
This may be due to the voltage drop with load operation.

64
 Chapter 4. Investigation of Voltage Quality Issues

4.3.2. Field Measurements at the University of Queensland 3.3 MW PV Plant,

Gatton Campus

In 2015, the University of Queensland installed a 3.3MW PV system at the Gatton campus,
which is one of the largest university PV plants in the world. This PV plant is located in South
East Queensland and nearly 82km away from the St. Lucia campus (Brisbane) of UQ. The PV
system has advanced PV panel technologies, which include Fixed-Tilt type, Single-Axis, and
Dual-Axis tracking types as illustrated in Figure 4.23.

(a) Aerial View of Site

(b) Single-line Electrical Diagram

Figure 4.23. Overview of UQ Gatton PV Plant.

65
 Chapter 4. Investigation of Voltage Quality Issues

Each PV inverter is connected with 480 strings which contain a total of 7200 panels and
provides 684kW of DC power. Each inverter is rated at 725kVA and is allowed to provide
reactive power support at full load operation. Currently, the PV inverters are operated to
produce only 630kW active power and provide a lagging power factor. The plant contains three
distribution transformers of 1400kVA, 11kV/415V delta-wye connected. Each power
conversion system (PCS) is fed by two PV inverter systems except the inverter that contains
the Dual-Axis tracking system as shown in Figure 4.23 (b). These inverters are connected to
the low voltage side of the transformer. This Gatton PV plant is equipped with power quality
meters (PQM) at the distribution transformers. Various power quality parameters such as
power, RMS voltage, current and flicker data are recorded with second’s resolution. For this
analysis, the measurements are obtained at one of the PV inverter outputs and the 11kV side of
the grid integration point (i.e., at CB1152).

4.3.2.1. Results and Analysis

4.3.2.1.1. Voltage Rise and Fluctuations Assessment at 3.3MW PV System

Characteristically, the amount of voltage rise depends on the PV capacity, connection

density and the location. The UQ Gatton campus has lower load demand during the daytime
compared to the UQ St. Lucia campus load requirement, and the capacity of PV installation is
also higher. Hence, it is expected that there is a possibility of high reverse power flow into the
Energex distribution network. Consequently, the utility controls the voltage at the upstream
11kV network by using a step voltage regulator (SVR), which typically regulates the voltages
to be in a range of 0.9875p.u +/- 0.0125p.u. Further, according to the connection agreement by
the utility, the voltage should not exceed a maximum of 1.01p.u, so that around 3.5% of voltage
variation is allowed.

Therefore, to analyse the influence of PV penetration on the LV side of the network,

measurements were collected for different operating days with a resolution of 1 second by
using power quality analysers. Figure 4.24 shows the time versus output power of one of the
PV inverters for typical operating days, which are clear sunny, average cloudy, cloudy, heavy
cloudy and low power days. It is noticed that the PV power curve is wider during the daytime
compared to the power curves of rooftop PV (see Figure 4. 12), which is due to PV array
technology. The corresponding measured per-unit RMS voltage at the LV side of the PV
inverter output is plotted as shown in Figure 4.25. It is observed that the trend of voltage
variation has followed the trend of PV power injection. During the early morning and evening

66
 Chapter 4. Investigation of Voltage Quality Issues

periods, the voltages are between 0.98p.u - 0.99p.u, which are within the limits specified by
AS/NZS 61000.3.100: 2011 standard. During the daytime, the voltage magnitudes have
increased with respect to the PV power injection. The magnitudes are normally lower than the
value of 1.01p.u except during the high power production periods (i.e., when the inverter has
produced 600kW power, e.g., 5/07/2015), which are within the standard limits. It can also be
observed that due to heavy cloud transient periods the voltages are fluctuated between 0.97p.u-
1.018p.u, which is nearly 4.5% of the voltage difference.

Figure 4.24. Time versus PV Power Output during Various Operating Days.

Figure 4.25. Measured Voltage Magnitude Profile at the LV Side of the PV Inverter Output.

Further, the collective impact of all five inverters can affect the upstream HV side of the
network voltage levels. Therefore, field measurements at the 11kV connection point for the
above indicated days were collected with second’s resolution for the analysis as shown in
Figure 4.26. Measurements show that the voltages during the early morning and night times
are fluctuating between 1.01p.u - 1.02p.u. During the daytime, the voltage rise due to reverse
power flow however, is somehow reduced by the SVR operation and the values are lower than
1.02p.u. Nevertheless, the voltage fluctuations are recorded higher, which may be due to the
number of regulatory operations. It is observed that these voltage fluctuations are higher during

67
 Chapter 4. Investigation of Voltage Quality Issues

the heavy cloudy days and the magnitudes oscillated between 0.98 to above 1.03p.u. Also,
variations in solar irradiation due to fast moving clouds could be faster than the response of
regulators, which may fail to regulate the voltage instantly. Furthermore, these voltage
fluctuations could trigger flicker incidence issues on the network with high PV installation
capacity, which will be a concern for the utility.

Phase A Voltage Magnitude at the 11kV Side of PCC

1.04

1.03
Voltage (p.u)

1.02

1.01

0.99

0.98
VA_HV

2:20:01
5:05:05
7:42:59
00:00:00
02:48:59
05:27:51
08:16:07
10:57:10
13:51:19
16:38:18
19:22:44
22:06:15
00:52:02
03:35:34
06:12:31
08:50:52
11:36:16
14:17:49
17:04:44
19:59:57
22:38:11
01:21:01
03:56:33
06:34:43
09:31:13
12:32:18
15:31:34
18:22:09
21:09:51
23:49:13

10:21:14
12:48:29
15:15:59
17:46:15
20:27:16
23:11:06
01:44:22
04:11:18
06:44:07
09:33:55
12:30:13
15:13:49
18:13:28
21:07:51
Time (H:Mm:SS)
3/07/2015 5/07/2015 1/07/2015 17/07/2015 11/07/2015

Figure 4.26. Measured Voltage Magnitude Profile at the HV Side of the PV Integration Point.

4.3.2.1.2. Voltage Flicker Analysis at 3.3MW PV System

Flicker is a serious constraint in the network which could impact the operation network
voltage regulating equipment and can reduce the equipment life span. The objective of this
section is to analyse the flicker sensation levels due to a 3.3 MW PV system at the high-voltage
side of the integration point.

As per the interconnection agreement with Energex for the UQ Gatton PV plant, the short
term and long term flicker emissions are restricted to 0.35 and 0.25 respectively. According to
AS/NZS 61000.3.7: 2001 [104], the short term and long term flicker levels in low and medium
voltage networks should not exceed 1 and 0.8 respectively.

To evaluate the probability of the flicker severity on the HV side of the PV integration node,
the short term and long term indices are obtained for 3 months which are calculated based on
equations (2.4) and (2.5) as presented in Chapter 2. For this, a clustering algorithm has been
considered to quantify the flicker sensations during the time of the day. The collected daily
data has been overlapped into a 24-hour time frame so that around 90 data points are allocated
for every time instant. The flicker data contains every 10 min and 2 hour time intervals. Further,
the short-term flicker levels at every time instant are classified into five clusters as indicated in
Figure 4.27. The probability counts for each cluster at the specific time instant are calculated
and this process is continued for all time instants. It can be observed that cluster 1 has accounted

68
 Chapter 4. Investigation of Voltage Quality Issues

for the highest number of counts in which the flicker levels are below 0.4 and satisfying the
agreed short-term emission limit. Cluster 2 is the next intensified flicker level with a total of
180 counts during the whole 3 months period. It can be noticed that during the no PV power
generation periods, the possible level of flicker is in cluster 2 and has a maximum number of
counts, which may be due to the load operation. Furthermore, the probability counts of cluster
3 have increased during the PV operation periods, which has accounted for a total of 27 events.
Significantly, the flicker levels are between 1 and 5 limits for nearly 41 times during the PV
operation time. It can also be seen that the inherent flicker levels also exceed the limit of 5
sometimes during the day, which is due to maximum PV power drop with heavy cloud cover.

Probility Count of Short-term Flicker Index at 11kV Side

100
PROBABILITY COUNT

10

0.1
0:00 4:48 9:36 14:24 19:12 0:00
TIME (H:MM)
Cluster 1 (Pst<0.4) Cluster 2 (0.4<Pst<=0.6)
Cluster3 (0.6<Pst<=1) Cluster 4 (1<Pst<=5)
Cluster 5 (Pst>5)

Figure 4.27. Probability Count of Short-term Flicker at the PV Interconnection Node.

Probabilty Count of Long-term Flicker Index at 11kV Side

100
PROBABILITY COUNT

10

0.1
0:00 4:48 9:36 14:24 19:12 0:00
TIME (H:MM)
Cluster 1 (Plt<=0.25) Cluster 2 (0.25<Plt<=0.4)
Cluster 3 (0.4<Plt<=1) Cluster 4 (Plt>1)

Figure 4.28. Probability Count of Long-term Flicker at the PV Interconnection Node.

69
 Chapter 4. Investigation of Voltage Quality Issues

Next, the long-term flicker values are segregated into four clusters and can be seen in Figure
4.28. It can be observed that the PV plant has flicker levels below 0.25 for nearly 50 to 65
events at every time instant of the day which satisfies the limits recommended by the plant
agreement. Further, cluster 2 is also accounted for about 25 to 30 events at every time instant
of the day, which is nearly 30% chance of increased flicker levels. Finally, clusters 3 and 4
have accounted for a lower number of events during the 3 months period. It is confirmed that
the voltage flickers are appeared to be under control for most of the time and satisfied the
AS/NZS standard limits.

4.4. Summary

This chapter deals with various power quality problems such as voltage rise, unbalance,
sag/swells and voltage flicker. Firstly, the investigation through PSCAD simulations in the
IEEE 13 bus unbalanced distribution network was presented. In this study, the PV penetration
influence on the network voltage performance was assessed in two case studies. In the first
case study, the voltage rise and unbalance assessment was carried out in three different
scenarios, which included lumped PV system penetration at near to the distribution transformer
(at the beginning of the feeder), distributed penetration at multiple locations near to the loads
(i.e., at end of the feeder) and also with increased penetration by residential single-phase PV
systems. Results show that the voltage rise due to distributed PV integration is higher than
compared to the lumped PV penetration at the beginning of the feeder. This impact is even
higher with increased PV penetration. In the second case study, fault analysis was carried out
for with and without PV system presence during each of the three types of faults on the IEEE
network to determine the voltage sag/ swell intensities. Results showed that during the fault
condition, the unbalanced voltage sags are slightly compensated by the contribution of PV
power. However, the voltage swell is attained higher values compared to without PV scenario.

Furthermore, the influence of PV system penetration and their impacts due to rapid
fluctuations in the solar irradiations, on the voltage quality of the distribution network was
analysed. The investigation was performed through field measurements on the different PV
systems with capacities of 1.5MW and 3.3MW, which are located at the UQ St. Lucia and
Gatton Campus respectively. Practical measurements were conducted through the installation
of power quality meters in the real PV site. Various power quality measurements at the PV
inverter output level to the high voltage and low voltage side of the distribution network level
were collected. Further, a clustering method has also been applied to estimate the probability
of flicker severity in the network. This analysis identified the long-term and short-term

70
 Chapter 4. Investigation of Voltage Quality Issues

influence of PV systems on the network voltage levels during different real operating scenarios.
Results confirm that the network voltage magnitudes are increasing with respect to the PV
power injection. Measurement results demonstrate that voltage quality concerns in the 1.5MW
PV system are insignificant compared to the 3.3MW system. The following chapter
investigates the harmonic emission issues from PV systems and their penetration impacts,
which is another significant power quality issue in the distribution network.

71
 Chapter 5. Analysis of Harmonic Issues with PV Systems

CHAPTER 5

ANALYSIS OF HARMONIC ISSUES WITH PV

SYSTEMS

5.1. Introduction1

This chapter presents the harmonic impact analysis of high PV systems integration on
distribution networks. In the first part of this research, the harmonic emission characteristics of
PV systems and their aggregation issues have been analysed. This analysis has been performed
on the IEEE 13 bus network through simulations in two different case studies and the results
are validated with the experimental measurement data collected at the University of
Queensland St. Lucia campus. For this purpose, a Fast Fourier Transform (FFT) algorithm has
been utilised for the harmonic analysis.

1
This chapter is incorporated with the following references:
Section 5.2 and 5.3:
Annapoorna Chidurala, T. K. Saha, and N. Mithulananthan, "Harmonic characterization of grid
connected PV systems and validation with field measurements," in Power & Energy Society General
Meeting, 2015 IEEE, 2015, pp. 1-5, Denver, USA.
Section 5.2 and 5.4:
Annapoorna Chidurala, T. K. Saha, and N. Mithulananthan, "Harmonic impact of high penetration
photovoltaic system on unbalanced distribution networks – learning from an urban photovoltaic
network," IET Renewable Power Generation, Volume 10, Issue 4, April 2016, p. 485 – 494
DOI: 10.1049/iet-rpg.2015.0188.

72
 Chapter 5. Analysis of Harmonic Issues with PV Systems

Subsequently, the influence of harmonic emissions from PV systems on unbalanced

distribution networks has been assessed. For this purpose, the investigation has been carried
out through simulations on three different case studies. Further, an evaluation study has been
performed with the online field measurements obtained at the various connection levels of the
real-time grid connected PV site, which is located at the UQ St. Lucia campus. The following
section presents the test system modelling and an approach for the harmonic analysis.

5.2. Test System Approach for the Harmonic Analysis

In this analysis, the IEEE-13 bus radial distribution network is selected as a test system. The
modelling of the IEEE-13 bus distribution system has been presented in Chapter 3, Figure 3.17.
This radial unbalanced distribution network is short and has a three-phase four wire system
with a nominal voltage of 4.16kV. It has overhead and underground lines of three-phase, two-
phase and single-phase line configurations with unbalanced line impedances. The network is
equipped with a substation transformer, in-line transformer, single-phase voltage regulators
and shunt capacitor banks. The test system contains balanced, unbalanced spot loads and
distributed loads, which can be modelled as constant power, impedance (Z) and current (I)
loads.

Figure 5.1. IEEE-13 Bus Test System.

Furthermore, in this study, to simulate a real-time residential distribution network the test
system has been modified. Regarding this, the total length (2000ft) of the feeder between nodes
632 to 671 is divided into five equal sections and the loads are equally distributed among nodes

73
 Chapter 5. Analysis of Harmonic Issues with PV Systems

load_1 to load_4 as shown in Figure 5.1. In the figure, PCC (point of common coupling)
indicates the PV connection node, which is at the middle of the feeder section (load_4 to 671).
The investigations have been performed by considering the affect with and without background
harmonics applied to the test network.

To simulate the without background distortion case, the network is modelled with constant
power, constant current and constant impedance loads considered as presented in Chapter 3.
Therefore, the harmonic distortion in the IEEE network is ideally zero. However, in real-time
conditions the distribution networks possess harmonic distortions due to residential and
commercial loads. Therefore, in order to investigate the impact of PV inverter systems in a real
distribution network scenario, some power electronic based nonlinear loads have been
introduced to inject necessary background distortion in the network according to the IEEE 519
standards [53]. To achieve a realistic scenario, various single-phase rectifier bridges with RL,
RC and motor type load models are connected to the feeder nodes load_1, load_2, load_3 and
load_4.

These nonlinear loads comprise numerous power electronic based residential appliances,
which include DC drives, computers (PC), television (TV), compact fluorescent lamps (CFL),
single-phase induction motors, adjustable speed drives (ASD) and lift motors. The detailed
load models were developed in PSCAD and verified for the harmonic levels with the IEEE
standard requirement as presented in Chapter 3, Section 3.4. The capacity and the current total
harmonic distortion of these loads are as follows: CFL (20W) - 125%; TV (50W) - 91.24%;
computer (100W) - 91%; DC drives (10hp) - 56.7% etc. Several combinations of these loads
are aggregated in each phase of the feeder nodes (load_1 to load_4) in order to acquire the
actual load capacities in the IEEE network. Single-phase loads mainly generate odd order
harmonics (i.e. 3rd, 5th, 7th, etc.). Moreover, ASDs and lift motors generate primarily 5th, 7th,
11th, 13th and higher order harmonics. Due to the dynamic operation of loads, the generated
harmonics from the nonlinear loads can be amplified or suppressed due to harmonic
superimposition and can cause distortion in the distribution network. Consequently, increased
dispersed PV penetration in existing distribution networks combined with numerous nonlinear
loads worsens the PQ of the grid. Therefore, in this chapter, the assessment has been carried
out systematically on this IEEE network with PV systems penetrations in different scenarios to
estimate the exact harmonic contributions from the PV systems.

74
 Chapter 5. Analysis of Harmonic Issues with PV Systems

5.3. Harmonic Characterisation of Grid Connected PV Systems &

Validation with Field Measurements

In the first part of the assessment, the characterisation of harmonics and their interaction
issues associated with multiple PV systems interconnection have been studied with simulations
and compared with field measurements. Due to the solar irradiation dependency, the power
generation is variable. Furthermore, the harmonic emissions from several identical systems
could be different due to their independent operation. Hence, the harmonic magnitude and
phases are also different for each inverter and the resultant harmonics may suppress or increase
because of harmonic superimposition. Therefore, in this research, the harmonic emissions from
PV inverter systems have been assessed for various real operating conditions. In addition,
harmonics interaction impacts between identical PV inverter integrations have been studied.
The simulation approach and the analysis of results are presented in section 5.3.1. Furthermore,
the field test approach and measurement results are explained in section 5.3.2. The results from
simulations are validated with experimental measurement results in subsection 5.3.2.2.

5.3.1. Simulation Approach and Analysis of Results

To investigate the harmonic characteristics of PV systems, a Current Controlled Current

Source Inverter (CRCSI) based PV system has been considered. The operation of a control
scheme and the modelling of the grid connected PV system components are discussed in
Chapter 3 (refer section 3.2). For this analysis, a dynamic PV system model has been developed
in PSCAD, which has all the characteristics similar to a real PV system. The rated capacity of
the PV system is about 330kWp and is considered to operate at unity power factor. The
switching frequency of the PV system is 2,460Hz (41x60Hz). The PV system is connected to
the grid through an LC filter. The value of the filter inductance is chosen as 3mH and the filter
corner frequency should be 1/10 of the switching frequency. Therefore, the filter capacitance
is calculated as about 90µF. In this study, a static solar irradiation pattern is applied on the PV
system as shown in Figure 5.2, realizing this condition as a half-day operation from noontime
to night-time. The corresponding active and reactive power outputs of the PV system are shown
in Figure 5.2.

75
 Chapter 5. Analysis of Harmonic Issues with PV Systems

0.40 Time Vs 1050

0.35 PV Output Powers and Solar Irradiation 900

Solar Irradiation (W/m2)

PV Power (MW)
0.30 750
0.25
600
0.20
450
0.15
0.10 300

0.05 150
0.00 0

10.8585
0
0.5715

1.7145

2.8575

4.0005

5.1435

6.2865

7.4295

8.5725

9.7155
10.287
1.143

2.286

3.429

4.572

5.715

6.858

8.001

9.144

11.43
Time (sec)
Ppv Qpv S

Figure 5.2. Time versus PV Output Active and Reactive Powers and Solar Irradiation.

The analysis has been carried out on two case studies by considering with and without the
effect of network background harmonics on the performance of PV systems as described in
section 5.2. In each case study, the simulations were performed in five scenarios by integrating
identical PV systems consecutively. In the IEEE-13 bus network, PV systems are connected at
the PCC node. In the first case study, the IEEE network is modelled with static loads, which
produce almost zero harmonic distortion in the network. In each scenario, identical PV systems
are connected at the same node for a better understanding. In this case, the harmonics produced
by PV inverters are only due to switching harmonics. Therefore, the harmonics that are
generated from the input side (also the PV side), due to solar irradiation level, temperature,
type and arrangement of PV modules and the inverter current controller are not determined by
the supply voltage. Ideally, a three-phase 2-level full bridge PV inverter can generate harmonic
orders mainly according to equation (5.1) and do not produce triplen harmonics.

ℎ = 6𝑛𝑛 ± 1 Where, n = 1, 2, 3, 4… (5.1)

The individual harmonics up to 31st order of both voltages and currents were obtained for
various solar power generation conditions. The results of voltage and current individual
harmonics in percentage of actual magnitudes with respect to the active power generation are
plotted as shown in Figures 5.3 and 5.4 respectively. In figures, P/Pn indicates the ratio of
actual PV power to the rated power. During simulation, the current and voltage amplitudes
change continuously for different power levels. For solar irradiation variations, the hysteresis
current controller can inject high frequency harmonics to the current due to its variable
switching frequency. From the results, it can be observed that the 5th order harmonic is the most
prominent. As expected, the 7th, 11th and 13th order harmonics and the higher order components

76
 Chapter 5. Analysis of Harmonic Issues with PV Systems

29th and 31st are presented in the voltage. It is noticed that the magnitudes of third order
harmonics are also exist in both voltage and current. In Figure 5.3, it can be seen that the values
of individual harmonics are well within the limits according to the IEEE standards.
1
Individual Voltage Harmonic Distortions without Supply Harmonics
0.9 H3
Harmonic Distortion

0.8 H5
(% fundamental)

0.7 H7
0.6 H9
H11
0.5
H13
0.4
H15
0.3 H17
0.2 H19
0.1 H21
H23
0
H25
100%

100%

100%

100%

100%
75%
50%
25%

75%
50%
25%

75%
50%
25%

75%
50%
25%

75%
50%
25%
H27
PV Output Power (%P/ Pn) H29
1PV 2PV 3PV 4PV 5PV
H31
Number of PV Systems

Figure 5.3. Voltage Harmonics at the Output of PV Inverter in Each Scenario.

However, in each scenario the individual harmonics from the identical inverters are
increasing with respect to the PV power injection. In addition, the PV power injection into the
grid causes an increase in voltage magnitude at the connection point. Furthermore, the
individual voltage magnitudes are added up from each inverter integration. As a result, the
percentage of distortion showed in the figure is a result of voltage rise and increased individual
voltage harmonic magnitudes. During all the power levels, the 5th harmonic is the most intense,
having values of about 0.7-0.9%.

In contrast, the current harmonics are following a reverse trend for lower PV power levels
and are increasing as shown in Figure 5.4. Besides, the harmonic distortions were appeared to
be reduced by each PV inverter penetration. During rated conditions, the generated harmonics
have satisfied the standard requirements. The current harmonic magnitudes mainly depend on
the active power output and the switching harmonics. However, during low power levels the
harmonic distortions have increased to over 5% (5th), 3% (3rd) and 2.8% (7th), which are high.
It is noted that the standards only recommend harmonic limits during rated operating condition.
The produced higher harmonics with less magnitude may not influence the distribution
network. Conversely, the current amplitudes from multiple inverters could be added and those
harmonics can affect the grid harmonic levels. In addition, PV inverters operating at low power
level, for instance 4 PV inverters operating at 25% PV power can produce a rated current of 1
PV system. Here, increased individual harmonic magnitudes from PV systems with the
produced fundamental current also affects the grid harmonic levels. The presence of triple

77
 Chapter 5. Analysis of Harmonic Issues with PV Systems

harmonics flowing through the neutral can cause the overheating and overloading of the neutral
conductor. Further, depending on the network configuration, load characteristics and the PV
inverter capacity and control strategy can excite a harmonic resonance issue.
5.5
Individual Current Harmonics without Supply Harmonics H3
5
4.5 H5
Harmonic Distortion

H7
(% fundamental)

4
H9
3.5
H11
3
H13
2.5
H15
2
H17
1.5
H19
1
H21
0.5 H23
0 H25
100%

100%

100%

100%

100%
75%
50%
25%

75%
50%
25%

75%
50%
25%

75%
50%
25%

75%
50%
25%
H27
H29
1PV 2PV 3PV 4PV 5PV H31
Number of PV Systems

Figure 5.4. Current Harmonics at the Output of PV Inverter in Each Scenario.

In the second case study, the static loads are replaced with nonlinear loads in the IEEE test
network. Therefore, due to harmonic producing loads, the voltage THD in the network is about
2-2.5% and the current THD is measured to be 3.5%. Similarly, simulations have been
performed for five scenarios. The measured voltage harmonic at the output of the PV system,
i.e. at the integration node, is plotted in Figure 5.5. Results show that the contribution of
harmonics from the PV inverter has increased the voltage distortion by 1% for the first PV
inverter integration. Further, a small increment in voltage distortion is observed with every PV
inverter integration. In this case study, the 7th harmonic was the highest, followed by the 5th
order harmonic. In comparison with the first case study, the harmonics were amplified. The 3rd
harmonics and its odd multiples have values within the 1% range.
H3
4
Individual Voltage Harmonic Distortions with Supply Harmonics H5
3.5
H7
Harmonic Distortion

3
(%fundamental)

H9
2.5 H11
H13
2
H15
1.5 H17
1 H19

0.5 H21
H23
0
H25
100%

100%

100%

100%

100%
75%
50%
25%

75%
50%
25%

75%
50%
25%

75%
50%
25%

75%
50%
25%

H27
PV Output Power (%P/ Pn) H29
1PV 2PV 3PV 4PV 5PV
Number of PV Systems H31

Figure 5.5. Voltage Harmonics at the Output of PV Inverter in Each Scenario with Distortion
in the Supply.

78
 Chapter 5. Analysis of Harmonic Issues with PV Systems

Correspondingly, the current harmonics also had the same tendency as with the previous
case and the values are almost equal. The emitted current harmonics in this case study contain
two causes of distortion; switching harmonics and the harmonics due to distortion in the supply.
A comparative plot of current THD during rated power condition in both case studies is shown
in Figure 5.6. From the definition, THD refers to the instantaneous measurement of harmonic
distortion at individual equipment with respect to the actual fundamental (I1). It can be seen
that the current THD appears to be decreasing due to an increase in fundamental current with
PV integration. Even though, background harmonics present in the network, the current
harmonics generated from the PV systems are within the limits. Further, the measured inverter
current THD with nonlinear loads is slightly reduced compared with the linear loads case,
which might be due to harmonic suppression with filters. However, the increased PV
penetration can influence the grid by both current and voltage harmonic levels.
5
Average Inverters Current THD with PV systems for
Inverter Current THD (%)

4.5
4 with and without nonlinear(NL) loads
3.5
3
2.5
2
1.5
1
0.5
0
1 PV 2PV 3PV 4PV 5PV
Number of PV systems
Iinv_case1_without NL loads Iinv_case2_with NL loads

Figure 5.6. Percentage of THDs of PV Inverter Output Currents in Two Case Studies.

Furthermore, field measurements have been conducted at the University of Queensland PV

system to evaluate the PV inverter harmonic behaviour in real operating scenarios.

5.3.2. Field Measurement Approach and Analysis of Results

5.3.2.1. Measurement Setup and Data Collection

The University of Queensland has one of the largest rooftop PV systems in Australia with
1.22MWp, which was installed in 2011. The system is scattered on four buildings with around
5000 PV panels. It is equipped with a special type of grid connected three phase PV inverter to
synchronize with grid voltages. In this study, field measurements were taken at the multilevel
car park (98A), which has an installed DC capacity of 339kWp with nearly 26 inverters. Figure
5.7 shows the 12.5kW transformer less PV inverters that are connected directly to the grid to
feed the load demand. Each PV inverter has two MPPT controllers and each MPPT is employed
with two strings of PV panels as explained in Chapter 4, section 4.3.1.

79
 Chapter 5. Analysis of Harmonic Issues with PV Systems

Figure 5.7. Grid Connected Three-phase PV Inverters at UQ PV Site.

To assess the harmonic distortion levels at the output of the PV inverter, field measurements
have been conducted at two of these PV inverters for different PV operating conditions. For
this, two FLUKE power quality analysers of types 434 series-II [122] and 1735 [123], which
are certified by IEC 61000-4-30 Class A standard, were used for the measurement. One power
quality meter was connected to the output of one PV inverter and the other was connected at
the combined output point of two inverters (Inverters 1 and 2). For each meter, four Amp
clamps are used to measure the three phase and neutral currents and voltage measurements are
conducted to each phase, neutral and ground. Figure 5.8 shows the test set up at the switch
board.

Figure 5.8. Measurement Test Setup for Harmonic Analysis.

The power quality parameters include harmonics, voltage, current, frequency and power
data, which has been recorded for 15 days. The resolution of each measurement was 5sec and

80
 Chapter 5. Analysis of Harmonic Issues with PV Systems

the sampling time was 50µs. The recorded data was transferred to a computer and extracted
using Power Log software. With this software, the measurement data can be exported to
spreadsheets for analysis.

5.3.2.2. Measurement Results and Discussion

The university also had installed weather sensors to collect solar irradiation and temperature
data. In this study, measurements were collected for various typical operating conditions of the
PV system. Figure 5.9 shows some selected days for the harmonic evaluation. The plot
illustrates the linear relationship between solar irradiation and PV power with respect to time.
350 1200
Time versus PV Output Power and Sunlight
PV Output Power (KW)

300 1000

Sunlight (W/m2)
250
800
200
600
150
400
100
50 200

0 0
5:00
6:00
7:00
8:00
9:00

5:00
6:00
7:00
8:00
9:00

5:00
6:00
7:00
8:00
9:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00

10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00

10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
11/04/2014 9/04/2014 13/04/2014
Date and Time
Power Sunlight

Figure 5.9. Time versus PV output Power and Sunlight.

With the use of two power quality meters, measurements for the corresponding days were
obtained. The measured Phase-B individual grid voltage harmonics of PV inverters for up to
31st order and up to 25th order from both Fluke meters are plotted in Figures 5.10 (a) and (b).
1.3 H3
(a) Phase-B Voltage Harmonics at Inverter 1
1.2 H5
1.1 H7
1
H9
Harmonic Distortion

0.9
(% fundamental)

0.8 H11
0.7 H13
0.6 H15
0.5 H17
0.4 H19
0.3
H21
0.2
H23
0.1
0 H25
H27
4:30:09
5:54:49
7:19:29
8:44:09

5:31:19
6:55:39
8:18:19
9:41:09

4:56:49
6:21:29
7:46:09
9:10:49
10:08:49
11:33:29
12:58:09
14:22:49
15:47:29
17:12:09
18:36:49

11:04:29
12:27:59
13:51:19
15:14:29
16:38:19
18:02:19

10:35:29
12:00:09
13:24:49
14:49:29
16:14:09
17:38:49

H29
H31
11/04/2014 9/04/2014 13/04/2014
Date and Time

81
 Chapter 5. Analysis of Harmonic Issues with PV Systems

1.3
1.2 H3
(b) Phase-B Voltage Harmonics at Combination of
1.1 H5

Harmonic Distortion
1 Inverters 1 & 2.

(% fundamental)
H7
0.9
0.8 H9
0.7 H11
0.6
H13
0.5
0.4 H15
0.3 H17
0.2
H19
0.1
0 H21
4:28:06
5:53:06
7:18:06
8:43:06

5:23:06
6:48:06
8:13:06
9:38:06

4:58:06
6:23:06
7:48:06
9:13:06
10:08:06
11:33:06
12:58:06
14:23:06
15:48:06
17:13:06
18:38:06

11:03:06
12:28:06
13:53:06
15:18:06
16:43:06
18:08:06

10:38:06
12:03:06
13:28:06
14:53:06
16:18:06
17:43:06
H23
H25

11/04/2014 9/04/2014 13/04/2014

Date and Time

Figure 5.10. Individual Voltage Harmonics of Phase-B at the PV Systems Output.

It can be observed that the voltage harmonics behaviour has a similar tendency compared
with the simulation outcomes. In both measurements, 5th and 3rd order harmonics were the
highest and were increasing with respect to the PV power output. During rated power level, the
5th order harmonics have values between 1-1.15%, which are within the limits according to the
standards. In addition, the 3rd order harmonic is constant throughout the day having values of
about 0.4-0.5%. The harmonics orders of 7th to 13th lie in the lower than 0.2% range. It can also
be estimated that the level of voltage harmonics was increased by the increased actual voltage
amplitude with the addition of PV inverters and the influence of augmentation of individual
voltage harmonic magnitudes. The characteristics and the levels of individual harmonics are
signifying the first case study of simulation (see Figure 5.3). It is important to mention that the
university network had filters and a potential transformer, which reduces the higher frequency
harmonics, so the results are appeared to be lower. However, depending on the type of load
and the network, voltage harmonic level might influence the individual harmonic magnitudes
and that scenario can be seen in the second case study of simulation (see Figure 5.5). Similarly,
the measured Phase-B current individual harmonics of one PV inverter (Inverter 1) and the
combination of two PV inverters were plotted in Figures 5.11 (a) and (b) respectively.

82
 Chapter 5. Analysis of Harmonic Issues with PV Systems

24 H3
(a) Phase-B Current Harmonics at Inverter 1
21 H5

Harmonic Distortion (% I1)

H7
18
H9
15
H11
12 H13
9 H15
H17
6
H19
3
H21
0 H23
4:30:09
5:54:49
7:19:29
8:44:09

5:30:29
6:54:49
8:17:29
9:40:19

4:55:59
6:20:39
7:45:19
9:09:59
10:08:49
11:33:29
12:58:09
14:22:49
15:47:29
17:12:09
18:36:49

11:03:39
12:27:09
13:50:29
15:13:46
16:37:29
18:01:29

10:34:39
11:59:19
13:23:59
14:48:39
16:13:19
17:37:59
H25
H27
11/04/2014 9/04/2014 13/04/2014 H29
Date and Time H31

14
(b) Phase-B Current Harmonics at Combination of Inverters 1 & 2.
H3
12
H5
Harmonic Distortion (% I1)

10 H7
H9
8
H11
6 H13

4 H15
H17
2 H19
0 H21
H23
4:33:06
5:58:06
7:23:06
8:48:06

5:33:06
6:58:06
8:23:06
9:48:06

5:08:06
6:33:06
7:58:06
9:23:06
10:13:06
11:38:06
13:03:06
14:28:06
15:53:06
17:18:06
18:43:06

11:13:06
12:38:06
14:03:06
15:28:06
16:53:06
18:18:06

10:48:06
12:13:06
13:38:06
15:03:06
16:28:06
17:53:06
H25

11/04/2014 9/04/2014 13/04/2014

Date and Time

Figure 5.11. Individual Current Harmonics of Phase-B at the PV Systems Output.

It can also be inferred that the measured current harmonics from the inverters have similar
trends as with simulated results. During the early hours (low irradiation) of inverter operation,
high levels of current harmonics are observed due to decrease in fundamental current. Also in
this period, the PV inverters consume a small amount of reactive power for their initial
operation depending on their inverter technology, which could also impact the current
harmonics. Note that the PV inverters are considered to operate at unity power factor at their
rated condition. The triple harmonics, especially 3rd and 9th order harmonics were shown to be
higher and were unable to be controlled adequately. These harmonics from each phase can be
added up instead of cancelling out, which may overload the neutral conductor.

However, during high power levels, it can be noticed that the 5th order harmonic was the
most affected and had values below the 2% limit, which are satisfying the AS 4777.2 standard
[107] requirement. Results show that the harmonic magnitudes from 3rd to 27th order were
decreased gradually. From Figure 5.11 (a) it can be seen that during rated conditions, the
inverter current is also producing higher frequency components (27th to 31st orders) that have

83
 Chapter 5. Analysis of Harmonic Issues with PV Systems

values below 1%, which has slightly exceeded the limits. Furthermore, fluctuation in the solar
irradiation can lead to misoperation of the inverter and the controller response may not
effectively suppress the harmonic injection. By comparing two current harmonic spectrums,
the percentage of harmonics have shown to be reduced, which is with respect to the increase in
current magnitudes. However, the individual harmonic magnitudes might be multiplied in
nature with identical PV inverters due to harmonic superimposition. Consequently, these
current harmonics with higher magnitudes can increase the harmonics in the grid.
42
39 PV Inverters Current THD(%)
36
Current THD (%)

33
30
27
24
21
18
15
12
9
6
3
0
10:31:49
11:44:09
12:56:29
14:08:49
15:21:09
16:33:29
17:45:49
18:58:09

10:29:39
11:41:59
12:54:19
14:06:39
15:18:59
16:31:19
17:43:39
18:55:59

10:27:29
11:39:49
12:52:09
14:04:29
15:16:49
16:29:09
17:41:29
18:53:49
4:30:09
5:42:29
6:54:49
8:07:09
9:19:29

5:40:19
6:52:39
8:04:59
9:17:19

5:38:09
6:50:29
8:02:49
9:15:09
11/04/2014 9/04/2014 13/04/2014
Date and Time
ITHD_B_Inv 1 ITHD_B_Inv's 1&2 ITHD_B_PV array

Figure 5.12. Percentage of Current THDs of PV Inverter System.

In addition, the current THD measurements for the above scenarios and at the output of
identical PV system arrays were measured and plotted as shown in Figure 5.12. From the
results, it can be noticed that the THD is dropping as the number of PV system integrations
increase. The current harmonic distortion levels during rated conditions have values below 4%,
which are within the limits specified by the data sheet of the inverter and the standards. The
measurement results have shown a similar tendency with the simulation results. Further
research will assess the harmonic impact issues with high PV system integration in the
distribution networks.

5.4. Harmonic Impact of High Penetration PV System on Unbalanced

Distribution Networks - Learning from an Urban PV Network

This research has focussed on the harmonic impact of PV system installations in a typical
unbalanced distribution network. Harmonic analysis has been performed through simulations
of three different case studies, which are considered by with and without the existence of
ambient harmonics in the network as mentioned earlier. To consider the effect of a PV inverter
control scheme on the network harmonic levels, two types of PV inverter models have been

84
 Chapter 5. Analysis of Harmonic Issues with PV Systems

considered. In this analysis, these two PV inverter system interconnections at a single node
and multiple nodes of the network have been taken into account for the characterization of
harmonic levels. Further, a harmonic evaluation study has been performed on the University
of Queensland (UQ) St. Lucia campus network, which has an installed PV system of 1.5MW.
The field measurement data was collected for two months at various levels of the UQ network
for different operating conditions of the PV system.

This section is organized as follows: Section 5.4.1 describes the simulation approach and
PV system connections for the comprehensive harmonic analysis. In addition, this section
presents the results and discussion of different case studies. The description of the university
network and an approach to the online field measurements are presented and the results are
validated in Section 5.4.2. Finally, the summary of this chapter is presented in Section 5.5.

5.4.1. Simulation Approach and Analysis of Results

To analyse the influence of harmonic distortion levels introduced by the PV systems into
the distribution network, two types of PV inverter control topologies have been considered.
Since the harmonic generations from PV inverters with different control technology are
divergent, their integration impacts may differ. Hence, in this research work two basic models
of the PV inverter with current control capability are considered namely Current Regulated
Current Source Inverter (CRCSI) and Current Controlled Voltage Source Inverters (CCVSI).
The details of these PV inverter control methods are presented in Chapter 3, Section 3.4.

The rated capacity of each PV system is about 330kW and they are operated at close to unity
PF. The CRCSI PV topology utilises a hysteresis based current controller for generating the
firing pulses as shown in Figure 3.15 (refer Chapter 3). This controller is designed with a
switching frequency of 2460Hz (41x60Hz) and HB is limited to 4% of the actual current. The
parameters of the PI controller are tuned by using a trial and error method. The DC input
voltage is 600V and the AC rated output line-line voltage is 400V. The inductance of the LC
filter is 3mH and capacitance is calculated to be about 90µF. A pulse width modulation
technique has also been applied to the CCVSI PV topology, which will generate firing pulses
for the inverter. The switching frequency is preferred to be about 2460Hz (41x60Hz). The DC
output voltage of the boost converter is 1050V and the rated AC output voltage is 400V. The
output filter inductance is calculated to be about 1.11mH by considering the 10% ripple at rated
output current and the capacitance is about 250µF.

85
 Chapter 5. Analysis of Harmonic Issues with PV Systems

In the simulations, the investigation has been performed on three different case studies. In
each case study, these two PV inverter models have been considered in the harmonic analysis
and verified to meet standards IEEE 929-2000 [106] and AS4777.2:2015 [108] requirements.
It is important to note that in all the simulation case studies, the PV systems are operating at
rated output conditions. The total current harmonic distortion from the PV systems is less than
3-3.5%. For simulations, the PSCAD software tool is utilised. These case studies will be helpful
for utilities to estimate the grid harmonic distortion levels in the case of high PV penetration.

5.4.1.1. Case Study I: Without Distortion in the Supply with PV Penetration at

the Same Bus

In the first case study, the IEEE-13 bus system has been considered and modelled with static
loads. Simulations have been performed for two different scenarios with two PV inverter
models. In each scenario, identical PV inverter systems are integrated at the same node, which
is upstream of bus 671 and treated as a Point of Common Coupling (PCC). With an intention
to examine the harmonic effects of each PV system in distribution networks, simulations were
carried out for five situations in which PV systems are integrated gradually into the IEEE
network.

In the first scenario referring to CRCSI (see Chapter 3.5.1), a 330kW PV system is
connected at the PCC. Usually, harmonic distortions are specified in terms of Total Harmonic
Distortion (THD) and Total Demand Distortion (TDD), which are used for power quality
measurements in the network. THD is the ratio between root-sum-square of harmonic content
to the fundamental component whereas TDD is calculated with respect to the load current.
According to IEEE 519-2014 standard [53], the voltage THD should be less than 8% and the
current TDD should be less than 5%. However, in reality THD is sufficient for analysis because
the load current will always be greater than or equal to fundamental current. Due to the
technical limitations in PSCAD software, the current and voltage THDs are calculated only up
to 31st order, which is lower than the switching frequency considered by the PV systems. In
this research work, all the simulations have been carried out at full load condition to eliminate
the complexity in calculating the harmonic distortion levels.

Therefore, for every situation simulations have been performed and the measured THD of
voltages and currents at different buses are presented in Tables 5.1 and 5.2 respectively. In
Tables 5.1 - 5.10, 0PV indicates the scenario without PV systems in the network and the
numbers 1PV, 2PV, 3PV, etc., represent the number of PV systems connected to the grid. From

86
 Chapter 5. Analysis of Harmonic Issues with PV Systems

the simulated results it can be seen that the voltage THDs are higher at bus 671 and the PCC
compared to the before integration point, which is due to voltage reduction along the feeder.
However, the supply side current THDs increased due to a decrease in the current magnitude.
Moreover, the THDs are different in each phase and are dependent on the loading level. In the
IEEE-13 bus unbalanced system, phase A is an averagely loaded phase, phase B is a lightly
loaded phase and phase C is a highly loaded phase. Results show that both voltage and current
THDs in phase B have slightly higher harmonic levels compared with the other phases due to
the fundamental current differences. Harmonic levels of phase C are almost identical to that of
phase A results.

Table 5.1 Percentage of Voltage THDs at Various Buses for Phases A and B with CRCSI
Phase Info Node No. Number of PV systems integration at PCC
0PV 1PV 2PV 3PV 4PV 5PV
Phase-A V632,a 0.00 0.0511 0.1019 0.1528 0.2104 0.2829
Voltage Vload_4,a 0.00 0.0743 0.1188 0.1911 0.2634 0.3458
THD (%) Vpcc,a 0.00 0.0873 0.2102 0.3041 0.4068 0.4916
V671,a 0.00 0.1102 0.2401 0.3276 0.4263 0.5257
Phase-B V632,b 0.00 0.0894 0.1594 0.2154 0.3012 0.3524
Voltage Vload_1,b 0.00 0.1230 0.2011 0.2802 0.3516 0.38412
THD (%) Vpcc,b 0.00 0.1964 0.3225 0.4313 0.4936 0.5423
V671,b 0.00 0.2314 0.3304 0.4341 0.5299 0.6859

Table 5.2 Percentage of Current THDs at Various Buses for Phases A and B with CRCSI
Phase Info Node Number of PV systems integration at PCC
No. 0PV 1PV 2PV 3PV 4PV 5PV
Phase-A I632,a 0.00 0.2621 0.5022 0.8462 1.3389 1.6995
Current Iload_4,a 0.00 0.2527 0.6061 1.1793 1.6261 1.9678
THD (%) Ipcc,a 0.00 0.2979 0.4064 0.6236 0.8109 0.8476
I671,a 0.00 0.3160 0.4885 0.7271 0.9042 0.9815
Phase-B I632,b 0.00 0.5177 1.0125 1.4318 2.3640 2.8905
Current Iload_4,b 0.00 1.2604 2.5168 7.4040 12.9230 14.8460
THD (%) Ipcc,c 0.00 0.7887 1.2266 1.9197 2.3261 2.4654
I671,c 0.00 0.6193 1.4699 1.9423 2.1127 2.2800

Table 5.1 shows that with zero harmonic distortion in the network, the penetration of 1PV
system contributes a voltage distortion of 0.1% and 0.2% in phases A and B at bus 671
respectively. It can be noticed that the variation in voltage THD at different nodes for each PV

87
 Chapter 5. Analysis of Harmonic Issues with PV Systems

inverter penetration is almost multiplied due to an increase in the individual voltage harmonic
magnitudes. It may be observed that the increase in net voltage THD is 'n' times the THD with
1PV system, where ‘n’ is the number of PV systems. Results show that the supply current THD
has an increased trend in each case and reaches a maximum of around 14.84% in phase B with
5 PV systems, which is violating the limit according to IEEE 519 standard.

Table 5.3 Percentage of Voltage THDs at Various Buses for Phases A and B with CCVSI
Phase Info Node No. Number of PV systems integration at PCC
0PV 1PV 2PV 3PV 4PV 5PV
Phase-A V632,a 0.00 0.0622 0.1275 0.1993 0.2592 0.3003
Voltage Vload_4,a 0.00 0.0792 0.1545 0.2472 0.3301 0.4024
THD (%) Vpcc,a 0.00 0.1289 0.2648 0.4012 0.5339 0.6952
V671,a 0.00 0.1343 0.2717 0.4103 0.5409 0.7153
Phase-B V632,b 0.00 0.0927 0.1826 0.2601 0.3781 0.4305
Voltage Vload_4,b 0.00 0.1132 0.1983 0.3721 0.4206 0.5808
THD (%) Vpcc,b 0.00 0.2026 0.3886 0.5978 0.7527 0.8503
V671,b 0.00 0.2104 0.4080 0.6122 0.7625 0.9229

Table 5.4 Percentage of Current THDs at Various Buses for Phases A and B with CCVSI
Phase Info Node No. Number of PV systems integration at PCC
0PV 1PV 2PV 3PV 4PV 5PV
Phase-A I632,a 0.00 0.4468 0.9920 1.6623 2.6869 3.3257
Current Iload_4,a 0.00 0.5305 1.1906 2.6748 3.3337 4.6934
THD (%) Ipcc,a 0.00 0.1906 0.3890 0.6115 0.7710 1.0058
I671,a 0.00 0.1715 0.3417 0.6698 0.8605 1.0586
Phase-B I632,b 0.00 0.6331 1.1906 2.8712 4.0111 5.5531
Current Iload_4,b 0.00 1.7388 4.5953 11.689 21.9732 26.7457
THD (%) Ipcc,b 0.00 0.5249 1.1033 1.8677 2.3928 2.9176
I671,b 0.00 0.4409 1.1544 1.7285 2.1147 2.7717

Similarly, in the second scenario, simulations have been carried out with CCVSI systems
and the measured voltage and current THDs for both phases A and B are shown in Tables 5.3
and 5.4 respectively. It can be seen that the voltage and current distortion levels follow similar
trends as mentioned earlier. The percentage of voltage and current THDs are slightly greater
compared to the first scenario with CRCSI systems. The voltage THDs are increasing from
upstream to downstream nodes, which is due to voltage drop along the feeder. Conversely, the
current THDs are exhibiting an inverse trend and have increased values at the upstream level

88
 Chapter 5. Analysis of Harmonic Issues with PV Systems

because of PV power injection causing the supply current reduction. Even though PV systems
inject balanced harmonics in the network, due to unbalanced loading the current THDs in
phases A and B with respect to the fundamental are different. Significantly, phase B is more
affected and has a current THD of nearly 27%with respect to the fundamental on the supply
side, which is very high.

5.4.1.2. Case Study II: With Distortion in the Supply and Multiple PV Penetration
at the Same Bus

In reality, distribution networks are polluted with harmonics, so in order to characterise the
harmonic behaviour with PV systems integration in a real power system network, the IEEE-13
bus network is modelled with nonlinear loads in the PSCAD software environment to simulate
a realistic scenario as mentioned earlier. In this case study, different power electronic circuit
based nonlinear loads are aggregated and connected to the feeder nodes load_1 to load_4 as
explained above, which inject harmonics into the distribution network. It is to be mentioned
that the load on the IEEE network was increased by 10% and this was considered as a worst
case condition. In the same way, simulations have been performed on two scenarios with two
PV inverter models. In Tables 5.5 - 5.8, 0PV refers to having no PV and with distortion in the
network in which the average voltage THD is around 3.5% and the current THD is about 3-
4%.

Table 5.5 Percentage of Voltage THDs at Various Buses for Phases A and B with CRCSI
and Distortion in the Network
Phase Node No. Number of PV systems integration at PCC
Info 0PV 1PV 2PV 3PV 4PV 5PV
Phase-A V632,a 1.6405 2.0643 2.1171 2.1809 2.2519 2.3013
Voltage Vload_1,a 1.9993 2.5163 2.5449 2.6228 2.6347 2.6478
THD (%) Vload_4,a 3.1243 3.8408 3.8847 3.9270 3.9647 4.1915
Vpcc,a 3.1249 4.3427 4.3768 4.4210 4.5375 4.6969
V671,a 3.5241 4.5294 4.5944 4.6525 4.75316 4.9868

Phase-B V632,b 1.4764 1.9810 2.0400 2.0777 2.1384 2.2761

Voltage Vload_1,b 1.7999 2.4131 2.4363 2.4925 2.5718 2.6577
THD (%) Vload_4,b 2.7604 3.6220 3.6798 3.7139 3.8994 4.1378
Vpcc,b 2.7612 4.1084 4.1199 4.1652 4.3809 4.5877
V671,b 3.0963 4.2556 4.3783 4.4393 4.6377 4.8519

In the first scenario, a CRCSI system is considered in the simulation and connected only at
the PCC. The resultant percentage of voltage THD at various buses in the network is presented

89
 Chapter 5. Analysis of Harmonic Issues with PV Systems

in Table 5.5. It can be seen that for the first situation only 1PV integration in the network causes
the voltage THD at before PV integration node to increase by 0.5% in phase A and 0.9% in
phase B. Subsequently, the voltage THD is amplified by over 1% at node 671 after the PCC
node. Results show that the voltage THD has an increased trend as it goes to the end of the
feeder compared to the before PV connection point. In addition, the increment in voltage
distortion at PCC is about 1.3%. It can be seen that the voltage THD is high in the case of
penetration of 5PV systems, the value reaching 5%. If the number of PV systems increases in
the distribution network, severe voltage distortion may be caused.

Table 5.6 displays the current THDs in phases A and B from the supply side to the load side.
It is observed that the current THD is higher on the supply side due to fundamental current
reduction with PV power injection. It is to be noted that the fundamental currents at the supply
side nodes will be different with PV systems in this case study compared to the first case study,
due to the increased load level of the network. With the integration of 1PV system, the current
distortions are raised by around 2% in phase A and 3% in phase B. Moreover, the current THD
at the supply side is rising around 1% in each step of PV system integration. Similar to the
previous case study, phase B is much more affected compared to other phases and has reached
11% and 7.8% at before and after PV integration node respectively, which is beyond the limit
as per IEEE 519 standards.

Table 5.6 Percentage of Current THDs at Various Buses for Phases A and B with CRCSI
and Distortion in the Network
Phase Info Node No. Number of PV systems integration at PCC
0PV 1PV 2PV 3PV 4PV 5PV

Phase-A I632,a 2.8213 4.4357 4.7178 5.0265 5.4602 5.9655

Current Iload_1,a 3.0796 4.8323 5.1884 5.4444 5.8683 6.1089
THD (%) Iload_4,a 3.1092 4.9461 5.2794 5.4458 5.8958 6.1901
Ipcc,a 3.1091 4.6945 4.7475 4.7501 4.7565 4.8180
I671,a 3.3046 5.0801 5.1475 5.2026 5.2571 5.3414
Phase-B I632,b 3.1349 5.3518 5.6918 6.0944 6.6003 7.0231
Current Iload_1,b 4.1897 7.8139 8.4828 9.2361 10.0334 10.5048
THD (%) Iload_4,b 4.3456 8.0171 8.8282 9.4000 10.4277 10.9493
Ipcc,b 4.3483 7.5153 7.6064 7.6055 7.6943 7.8467
I671,b 4.3495 7.5752 7.6117 7.6738 7.7819 7.8445

90
 Chapter 5. Analysis of Harmonic Issues with PV Systems

In the second scenario, a CCVSI type system has been considered and simulations have
been carried out similar to the first scenario. Measured voltage and current THDs for different
nodes from the supply side to the load side are shown in Tables 5.7 and 5.8 respectively. It can
be noticed that the voltage and current THDs have lower values with the CCVSI system
compared to the previous scenario, this is because of harmonic suppressions between PV
systems and nonlinear loads. It can be seen that the voltage THDs at the downstream nodes
have high values of over 4% for phase A and about 4.8% for phase B, which are close to the
limit. If the penetration in the network goes high, there will be further increases in harmonic
distortion levels, which will violate the limits.

Table 5.7 Percentage of Voltage THDs at Various Buses for Phases A and B with CCVSI
and Distortions in the Network
Phase Info Node No. Number of PV systems integration at PCC
0PV 1PV 2PV 3PV 4PV 5PV
Phase-A V632,a 1.6405 2.0154 2.0562 2.1275 2.1769 2.3245
Voltage Vload_1,a 1.9993 2.4581 2.5206 2.6101 2.7627 2.8470
THD (%) Vload_4,a 3.1243 3.8012 3.9079 3.9991 4.0271 4.1296
Vpcc,a 3.1249 4.2744 4.3798 4.4120 4.4792 4.5487
V671,a 3.5241 4.2524 4.3195 4.3623 4.4245 4.5525
Phase-B V632,b 1.4764 1.8817 1.9487 2.0294 2.0856 2.2056
Voltage Vload_1,b 1.7999 2.2526 2.2943 2.3256 2.5262 2.6547
THD (%) Vload_4,b 2.7604 3.4769 3.5442 3.6484 3.7584 3.8411
Vpcc,b 2.7612 3.8123 3.9088 4.0838 4.2285 4.5358
V671,b 3.0963 3.8971 4.0188 4.2127 4.4651 4.7685

Similarly, the current THDs are following the same trend as in the previous scenario. Results
show that phase B has a high current THD value of 9.36% with respect to the fundamental at
node load_4 with 5PV systems integration as shown in Table 5.8. Also, it can be observed that
there is an increase in current THD of 0.4% for every PV system addition. Moreover, the
current THDs in phase B at the PCC and 671 nodes have reached higher values of above 7%.
The generated harmonic THDs are exceeding the IEEE standard limits, which will severely
affect the network along with connected neighbouring sensitive equipment and loads.

91
 Chapter 5. Analysis of Harmonic Issues with PV Systems

Table 5.8 Percentage of Current THDs at Various Buses for Phases A and B with CCVSI
and Distortion in the Network
Phase Info Node No. Number of PV systems integration at PCC
0PV 1PV 2PV 3PV 4PV 5PV
Phase-A I632,a 2.8213 4.2981 4.5061 4.8029 4.9837 5.0643
Current Iload_1,a 3.0796 4.8106 4.9862 5.2701 5.5887 5.8708
THD (%) Iload_4,a 3.1092 4.8421 5.0988 5.2787 5.6438 6.0260
Ipcc,a 3.1091 4.7157 4.7496 4.7608 4.8325 4.8139
I671,a 3.3046 5.1023 5.1382 5.1629 5.2086 5.2238
Phase-B I632,b 3.1349 5.1389 5.3664 5.6289 5.9077 6.1701
Current Iload_1,b 4.1897 7.5365 7.9001 8.3363 8.9774 9.1763
THD (%) Iload_4,b 4.3456 7.8239 8.2055 8.7053 9.1386 9.3661
Ipcc,b 4.3483 7.4443 7.4508 7.4758 7.4938 7.5197
I671,b 4.3495 7.3798 7.4557 7.4406 7.4353 7.4256

5.4.1.3. Case Study III: With Distortion in the Supply and Multiple PV
Penetrations at Multiple Locations

Finally, in the last case analysis has been carried out to study the impact of PV system
penetration at different nodes. In this regard, the IEEE-13 bus system is considered with
distortions as mentioned above. As for the first scenario, five PV inverter systems of CRCSI
have been considered and are integrated along the feeder start from load_1 to the PCC in the
test system. Simulations have been performed and the corresponding node voltage and current
THDs were measured in each phase as presented in Table 5.9.

Table 5.9 Percentage of Voltage and Current THDs with 5 PV Systems (CRCSI) at
Multiple Locations

Voltage THD (%) Current THD (%)

Node No.
Phase-A Phase-B Phase-C Phase-A Phase-B Phase-C
632 2.0871 2.0410 1.8765 5.3846 6.9656 7.5568
load_1 2.5311 2.4614 2.2894 5.7285 9.6245 8.8493
load_2 2.9746 2.8813 2.7036 5.4545 9.1788 8.3642
load_3 3.4253 3.3018 3.1316 5.1854 8.6758 7.8109
load_4 3.8935 3.7628 3.5758 4.9275 8.1494 7.4010
PCC 4.3521 4.1524 4.0363 4.7133 7.6377 6.9158
671 4.3505 4.1519 4.0304 5.0980 7.5822 6.5983

92
 Chapter 5. Analysis of Harmonic Issues with PV Systems

From the results, it is noticed that the harmonic distortion in each phase is different due to
unbalance in the network and is higher in phase A, which is an averagely loaded phase. The
voltage and current distortions are slightly lower compared to the previous case study, which
is due to the harmonic superimposition. The voltage THDs of each node along the feeder line
(i.e., from 632 to 671) are increasing and has higher values of nearly 4.4% at bus 671, which
are within the limits. Conversely, the current THD profile followed an opposite trend and has
values higher at the supply side nodes due to a decrease in current magnitude with PV
penetration. Nonetheless, current THDs reached 7-7.6% at bus 671, which are over the limits
according to IEEE 519 standards. In addition, these distortions will rise not only dependent on
the number and capacity of PV systems but also on the type of control strategy.

Likewise, simulations were carried out with 5PV systems of CCVSI and the measured
voltage and current THDs for various nodes are displayed in Table 5.10. In this scenario, the
voltage THD in phase A is higher compared to other phases. It is observed that with the
harmonic interactions between nonlinear loads and the PV systems, there will be a harmonic
cancellation or addition taking place in the network. Results show that the voltage and current
THDs have similar values as in the previous case study.

Table 5.10 Percentage of Voltage and Current THDs with 5PV Systems (CCVSI) at
Multiple Locations

Voltage THD (%) Current THD (%)

Node No.
Phase-A Phase-B Phase-C Phase-A Phase-B Phase-C
632 2.3149 2.1519 2.1778 5.2077 6.1717 6.4699
load_1 2.8289 2.6612 2.5220 5.6163 8.8829 8.2067
load_2 3.2642 3.1232 2.9914 5.3043 8.5414 7.9292
load_3 3.7075 3.4204 3.3462 5.0652 8.2420 7.6872
load_4 4.1586 3.8586 3.7643 4.9822 7.8055 7.3686
PCC 4.6399 4.2410 4.1982 4.8205 7.4066 6.9575
671 4.7679 4.3284 4.1381 5.1520 7.4804 6.6358

It can be concluded that the distortion levels increase with multiple installations at the same
node. The voltage distortion depends on both the system impedance and the aggregate of
harmonic current injected. The situation may be worse in a real distribution network with a
higher number of loads and with more PV inverter systems, which can further increase
harmonic levels. Hence, this harmonic analysis has been extended to justify and compare the
simulation results with real-time distribution network measurements.

93
 Chapter 5. Analysis of Harmonic Issues with PV Systems

5.4.2. Field Measurements at the University of Queensland - Analysis of Results

The University of Queensland, St Lucia campus is a large university of 2km radius. The
university has an extensive power network with a maximum load demand of about 23MW and
receives an 11kV supply from the Energex distribution network. The high voltage (HV) lines
are supplied around the university and nearly forty substations are employed to step down the
voltage to the 415V supply level as shown in Figure 5.13.

Figure 5.13. Overview of UQ Sub Station 11.

In 2015, the university installed a 1.5MW rooftop solar PV system, which is one of the
biggest rooftop PV systems in Australia. This PV system is distributed across eight buildings
with more than 6305 PV panels. The network consists of the latest electricity infrastructure
including digital power meters and power quality (PQ) analysers connected to the internet for
monitoring.

In this study, to assess the effect of PV integration on the harmonic distortion levels of the
distribution network, measurements were taken at the multilevel car park buildings. Each
building has an installed capacity of 339kWp and uses 26 three-phase inverters. Special types
of grid integrated PV inverters are used to synchronise with the grid voltages. The generated

94
 Chapter 5. Analysis of Harmonic Issues with PV Systems

PV power is supplied to other buildings to feed the peak load demand. The car park buildings
are interconnected to one of the substations through two 750kVA 11kV/415V distribution
transformers. PQ analysers connected at the HV side, transformers and at the output of the PV
arrays were used to monitor various PQ parameters.

For this analysis, the harmonic data from various PQ meters in different integration levels
were collected for two months. Using JAVA programs, the harmonic measurements were
extracted from an online data monitoring system for every 5ms time resolution. The solar
irradiation and the temperature data were obtained using weather sensors.

(a) Time versus PV Output Power and Sunlight.

1600 2000

1400
1600
Power Output (KW)

Power
1200

Sunlight (W/m2)
Sunlight
1000 1200
800

600 800

400
400
200

0 0
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00

10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00

10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00

10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
5:00
6:00
7:00
8:00
9:00

5:00
6:00
7:00
8:00
9:00

5:00
6:00
7:00
8:00
9:00

5:00
6:00
7:00
8:00
9:00

24/04/2014 11/04/2014 9/04/2014 13/04/2014

Date and Time

(b) PV Systems Voltage THD (%) at Carpark-98A

1.6

1.4

1.2
Voltage THD (%)

0.8

0.6

0.4

0.2

0
00:00:04
02:10:37
04:21:20
06:32:07
08:43:13
10:54:51
13:06:15
15:17:37
17:28:46
19:39:27
21:50:14
00:00:59
02:11:22
04:22:11
06:33:06
08:44:23
10:55:37
13:07:25
15:19:00
17:30:12
19:40:54
21:51:11
00:01:19
02:11:53
04:23:02
06:34:13
08:45:26
10:57:09
13:08:50
15:20:03
17:30:48
19:41:15
21:51:51
00:01:57
02:12:23
04:22:34
06:33:01
08:44:18
10:55:35
13:06:44
15:18:01
17:29:35
19:39:59
21:50:22

24/04/2014 11/04/2014 9/04/2014 13/04/2014

Date and Time
Va_98A Vb_98A Vc_98A

95
 Chapter 5. Analysis of Harmonic Issues with PV Systems

(c) PV Systems Current THD (%) at Carpark-98A

70

60

50
Current THD (%)

40

30

20

10

0
00:00:04
02:07:48
04:15:39
06:23:37
08:31:52
10:40:36
12:49:16
14:57:48
17:06:08
19:14:15
21:22:10
23:29:58
01:37:40
03:45:37
05:53:32
08:01:48
10:10:19
12:19:11
14:28:03
16:36:33
18:44:37
20:52:02
22:59:41
01:06:52
03:15:04
05:23:27
07:31:50
09:40:37
11:49:20
13:57:55
16:06:16
18:14:15
20:21:42
22:29:24
00:36:50
02:44:23
04:51:48
06:59:37
09:08:04
11:16:29
13:24:51
15:33:21
17:41:57
19:49:43
21:57:09
24/04/2014 11/04/2014 9/04/2014 13/04/2014
Date and Time
Ia_98A Ib_98A Ic_98A

Figure 5.14. Solar PV System Measurements at UQ for Typical Operating Days.

The nature of solar power generation is intermittent because of variation in solar irradiation
and it depends on the weather conditions. PV power has a linear relationship with the solar
irradiation. Hence, harmonic emissions are varied for generated PV power. Figure 5.14 (a)
shows the various operating conditions of the solar PV system and the characteristics of PV
power and sunlight with respect to the time. The PV power outputs during a sunny day, average
cloudy day, cloudy day and the worst day were shown in Figure 5.14 (a). The corresponding
measured THD of PV output voltages and currents at Carpark 98A are shown in Figures 5.14
(b) and (c) respectively. As can be observed, the percentage of voltage THD has higher values
during sunny and average mid cloudy days. It can be seen that during the day when PV power
is low (especially early morning or evening times) the simulation scenario of without PV and
with distortion in supply is replicated. In this period, the voltage THD levels were in the range
of about 0.2-0.4%. However, during peak power generation periods (sunny and average cloudy
days) the PV system produces voltage harmonics of nearly 1.4%, which satisfies the IEEE 929
[106] standard limit. The increase in voltage THD will depend on the capacity and the number
of inverters connected to the network.

Conversely, the current THD has an inverse relationship with generated PV power due to
the fundamental current reduction. The measured results show that the current THDs have high
values at low output power levels; and lower values around 3% during rated output periods,
which are with respect to the fundamental current. However, during rated conditions, the
harmonic injections from PV systems will have a negative impact on the distribution

96
 Chapter 5. Analysis of Harmonic Issues with PV Systems

transformer and adjacent distribution network. In addition, the distribution network contains
different linear and nonlinear loads, power electronic based loads and heating loads. The
proliferation of these types of loads means the distribution networks are already very sensitive
in nature and with the integration of PV systems can cause severe impacts. As the generated
PV harmonics may be added or subtracted with the network harmonics, they can have a subtle
effect on the network.

In addition, field measurements were recorded at the distribution transformer for the same
days to analyse the impact of these PV systems at the integration point. The data from the PQ
meters provides a K-factor of 1.2, where the K-factor is the sum of the squares of the individual
harmonic currents, times the harmonic squared, divided by the rated current as expressed in
equation (5.2).

𝐼𝐼𝑛𝑛 2
𝐾𝐾 − 𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 = ∑∞
𝑛𝑛=1 � � 𝑛𝑛2 (5.2)
𝐼𝐼𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟

From the standard IEEE C57.110-1998 [124], the distribution transformer is de-rated to 80%
of its maximum capacity. The measured voltage and current THDs at the transformer are
plotted in Figure 5.15.

(a) Percentage of Voltage THD at the Distribution Transformer TX2

1.8

1.6

1.4
Voltage THD (%)

1.2

0.8

0.6

0.4

0.2

0
00:00:04
02:11:24
04:21:56
06:33:02
08:44:04
10:55:26
13:06:55
15:18:25
17:30:02
19:41:07
21:51:49
00:02:29
02:13:29
04:24:17
06:35:12
08:46:33
10:57:50
13:08:57
15:20:09
17:31:22
19:42:23
21:53:38
00:04:25
02:15:31
04:26:23
06:37:12
08:48:34
10:59:46
13:11:13
15:22:23
17:33:20
19:44:09
21:54:54
00:05:57
02:16:41
04:27:31
06:38:31
08:49:34
11:00:51
13:11:55
15:23:08
17:34:23
19:45:35
21:56:18

24/04/2014 11/04/2014 9/04/2014 13/04/2014

Date and Time

VTx2_a VTX2_b VTX2_c

97
 Chapter 5. Analysis of Harmonic Issues with PV Systems

(b) Percentage of Current THD at the Distribution Transformer TX2

55
50
45
Current THD (%)
40
35
30
25
20
15
10
5
0
00:00:04
02:08:23
04:15:53
06:23:58
08:32:00
10:40:18
12:48:43
14:57:13
17:05:46
19:13:59
21:21:57
23:29:11
01:37:22
03:45:07
05:52:57
08:01:12
10:09:34
12:17:48
14:25:42
16:33:59
18:42:01
20:50:07
22:58:17
01:06:11
03:14:11
05:21:55
07:29:53
09:38:16
11:46:28
13:54:54
16:02:56
18:10:39
20:18:35
22:26:33
00:34:14
02:42:05
04:50:01
06:58:03
09:06:02
11:14:19
13:22:13
15:30:31
17:38:42
19:46:55
21:54:36
24/04/2014 11/04/2014 9/04/2014 13/04/2014

Date and Time

ITx2_a ITX2_b ITX2_c

Figure 5.15. Harmonic Measurements at the Distribution Transformer TX2 for the
Selected Days.

Results show that the voltage harmonics increased from 0.6% to almost 1.6% during the
daytime, which is almost 1% of the variation. Noticeably on the supply side, the impact on the
network voltage is less compared to the feeder end node due to voltage reduction. The measured
voltage THD levels are shown to be within the limits; however, the network has filters and
potential transformers, which eliminate higher harmonic components. It can be estimated that
the voltage THDs will be more severe in the network without filters. In addition, the measured
current THDs increased from 7% to almost 53% in the case of rated conditions, which is an
aggregation of harmonics from the loads and PV systems. However, considering the worst day,
the network current THD with a value of about 25% (with only load operation) can be seen as
no PV condition in the simulated scenario. Hence, during midday and with PV, the THD is
increased to almost 50%. It is observed that the field measurement results follow a similar trend
as that of the simulated results.

To study the impact of current harmonics flowing in the transformer, K-factor

measurements for the corresponding days have been taken and are shown in Figure 5.16. K-
factor is a measure to estimate the influence of harmonics on transformer temperature and to
overrate or de-rate the transformer to handle the worst conditions. It is noticed that during low
PV power days, the K-factor values are in the range of less than 2. Results show that during
the daytime, the K-factor values increase due to PV power injection. During cloudy days, the
K-factor value is very high, about 10 when the PV generated maximum power, which

98
 Chapter 5. Analysis of Harmonic Issues with PV Systems

designates the proportion of harmonic content flowing in the network. If the K-factor value is
higher, the greater are the heating effects, which indicate the existence of current harmonics in
the network.

12

10

8
K-factor

0
12:00
15:00
18:00
21:00

11:00
14:00
17:00
20:00
23:00

10:00
13:00
16:00
19:00
22:00

12:00
15:00
18:00
21:00
0:00
3:00
6:00
9:00

0:00
2:00
5:00
8:00

1:00
4:00
7:00

0:00
3:00
6:00
9:00

0:00
24/04/2014 4/11/2014 4/09/2014 13/04/2014

Date and Time

K Fac-A K Fac-B K Fac-C

Figure 5.16. K-Factor Measurements for the Corresponding Days at the Transformer.

Highly distorted currents flowing in the grid can cause overheating of distribution
transformers, which in turn overheat neutral conductors, tripping the circuit breaker,
overstressing of power factor correction capacitors and skin effect [55]. In addition, subtle
effects can lead to increased costs due to unnecessary maintenance and equipment
replacements. So, focus on this issue and the need to take preventive measures is required.
Finally, analysis has been carried out on the 11kV side (HV) and THD measurements for the
voltage and currents are plotted in Figure 5.17.

(a) Percentage of Voltage THD at the HV Line

1.6

1.4

1.2
Voltage THD (%)

0.8

0.6

0.4

0.2

0
00:00:02
02:07:34
04:15:48
06:23:30
08:31:14
10:39:09
12:47:26
14:55:24
17:03:31
19:11:28
21:19:14
23:27:26
01:36:00
03:44:00
05:52:00
08:00:00
10:07:00
12:16:00
14:23:00
16:36:15
18:44:34
20:52:29
23:00:15
01:09:00
03:17:00
05:25:00
07:33:00
09:41:00
11:49:00
13:57:00
16:05:00
18:13:00
20:21:00
22:30:00
00:38:04
02:46:08
04:54:03
07:02:08
09:10:05
11:17:57
13:25:35
15:33:57
17:42:04
19:50:01
21:57:49

24/04/2014 11/04/2014 9/04/2014 13/04/2014

Date and Time

Va_11KV Vb_11KV Vc_11KV

99
 Chapter 5. Analysis of Harmonic Issues with PV Systems

(b) Percentage of Current THD at the HV Line

7

5
Current THD (%)

0
00:00:02
02:04:45
04:10:05
06:15:01
08:19:54
10:24:54
12:30:21
14:35:27
16:40:40
18:45:50
20:50:52
22:56:02
01:02:00
03:07:00
05:12:00
07:17:00
09:22:00
11:27:00
13:32:00
15:37:00
17:47:40
19:52:57
21:57:48
00:03:00
02:08:00
04:14:00
06:19:00
08:24:00
10:29:00
12:35:00
14:40:00
16:45:00
18:50:00
20:56:00
23:01:00
01:06:35
03:11:45
05:16:48
07:22:02
09:27:07
11:32:08
13:36:55
15:42:34
17:47:43
19:52:52
21:57:49
24/04/2014 11/04/2014 9/04/2014 13/04/2014

Date and Time

Ia_11KV Ib_11KV Ic_11KV

Figure 5.17. Harmonic Measurements at the HV Line.

Results show that the voltage THD has a similar tendency as the transformer voltage THDs.
The voltage distortion varies with PV output power and drops during evening times. However,
the current harmonic distortions have an inverse trend, for instance during the afternoon, the
current THD reduces considerably. It can be seen that the THD values are oscillating from 1%
to about 5% during a sunny day. It implies that there is a drastic rise in current THD by 4%
that can cause adverse effects in the neighbouring network. In addition, the high integration of
PV systems and the network configurations can excite harmonic resonance, which will have
severe harmonic distortion levels in both voltages and currents.

5.5. Summary

This chapter has analysed the harmonic characteristics of grid connected PV systems for
variations in solar irradiation. In addition, this research has highlighted the individual harmonic
emissions and their relative interference associated with PV inverter penetration in various
scenarios. The harmonic analysis has been performed through simulations in two case studies
by considering with and without the presence of harmonics in the supply network. For this
study, the IEEE-13 bus distribution system has been considered. Furthermore, to simulate a
real-time network scenario, several types of probabilistic nonlinear loads were incorporated in
the network for harmonics injection. Moreover, the analysis has been carried out through field
tests conducted at the University of Queensland PV site under real operating conditions. For
this purpose, Fluke power quality analysers were connected at the inverter level for
measurement purposes. The measurement results were captured and compared with simulation

100
 Chapter 5. Analysis of Harmonic Issues with PV Systems

results. Results show that the harmonic behaviours of voltage and current waveforms are
comparable with simulation trends. The individual voltage harmonics had a linear relation to
the solar power generation but the current harmonics had an inverse relationship. The
individual voltage and current harmonics during rated power levels are within the limits
prescribed by the IEEE and AS/NZ standards. Moreover, with multiple PV integrations the
current harmonic levels appeared to be minimized due to increase in fundament current,
however the harmonic magnitudes incremented with every parallel PV penetration.

Further, a potential issue of increased harmonic distortion levels on a low voltage

distribution system with high penetration of PV systems has been assessed. To characterise the
effect of PV inverter control strategy on the network harmonic contribution level, two modern
PV inverter models were considered. A thorough harmonic impact analysis was carried out
through simulations for different case studies with the integration of PV systems in the
unbalanced distribution network. This analysis has also considered harmonic distortions
associated with various power electronic based nonlinear loads. Furthermore, a field
investigation was performed at the University of Queensland PV system. The harmonic field
measurements, recorded by various PQ analysers, at the LV and HV sides of the network for
different operating conditions and scenarios were collected through online JAVA programs.
Moreover, the simulation results were validated with real-time harmonic measurements.
Simulation results show that the THDs of voltage and current are exceeding the IEEE limits
when the number of PV systems increases in the network. The harmonics generated from PV
systems are multiplied in nature due to an increase in individual voltage magnitudes.
Simulation case studies revealed that the network harmonic levels increased with multiple
installations at the same node compared to the dispersed integration. Measurement results show
that the THD trends follow similar trends as with simulation results. Results shows that the
current THD has high values at low power levels and lower values at rated condition, which
are dependent to the fundamental component. However, during rated condition, the harmonic
injections from PV systems have negative impact on the distribution transformer and adjacent
distribution network. Further investigation will be continued to understand the impact of PV
systems on network harmonic resonance that is presented in Chapter 6.

101
 Chapter 6. Harmonic Resonance Issues with PV Penetration

CHAPTER 6

HARMONIC RESONANCE ISSUES WITH PV

PENETRATION

6.1. Introduction1

The main objective of this chapter is to investigate the harmonic resonance issues, which
can be exacerbated by PV system integration into the LV network. Although issues associated
with the PV systems on the distribution grid are widely analysed, the harmonic resonance
problems related to PV system penetration and their controller interactions have rarely been
assessed and reported. For this purpose, a harmonic analysis has first been performed through
simulations in two case studies. Also in this investigation, two types of PV inverter control
topologies are considered to study the effect of PV systems dynamic interaction in exciting
harmonic resonance conditions in the network. Furthermore, the field measurement data
obtained from a real network has been analysed and verified to validate harmonic resonance
issues.

1
This chapter is organised with the following reference:

Annapoorna Chidurala, T. K. Saha, and N. Mithulananthan, "Harmonic Resonance on the

Distribution Network with PV Penetration and Field Measurements Validation," in preparation, to
be submitted to International Transactions on Electrical Energy Systems Journal by September 2016.

102
 Chapter 6. Harmonic Resonance Issues with PV Penetration

This chapter is organized as follows; Section 6.2 represents the harmonic resonance
mechanism accompanied with PV system penetration. Section 6.3 describes the test system
and an approach to the simulation analysis. Section 6.4 represents results and discussion with
simulations as well as validation by field measurements. The approach for field measurements
at the University of Queensland (UQ) PV network is also presented in Section 6.4.2. Finally,
the summary of this chapter is presented in Section 6.5.

6.2. Harmonic Resonance Mechanism Associated with PV Systems

Resonance conditions exist in the network when network capacitive reactance and inductive
reactance become equal at a particular frequency, known as resonance frequency. In general,
the network can have multiple resonance frequencies depending on the system configuration
and characteristics. These resonance frequencies will be a major issue when they coincide with
any of the induced individual harmonic frequencies in the network. The type of resonance may
be series or parallel, which are caused by the series and parallel arrangement of the distribution
network passive elements.

Typically, harmonic resonance incidence depends on network configuration, operating

conditions, type and capacity of load and PV systems. Series resonance causes the total
reactance to be virtually zero at the resonance frequency. Therefore, the network only consists
of resistance and provides a low system impedance path, which will result in amplification of
harmonic currents for a small distortion in voltage. Parallel resonance makes the equivalent
impedance high, which creates to the amplification of harmonic voltages [63], [45]. In a critical
case at the resonance frequency, a relatively smaller distortion in the current can create a higher
voltage distortion.

The equivalent single-phase harmonic model of a PV connected distribution grid is shown

in Figure 6.1. This configuration ignores the impact of a current controller circuit and the PV
inverter bridge output voltage (Vx) is represented as a harmonic source, which contains a range
of voltage harmonics. In the figure, Zs represents the total impedance of the transformer and
line; ZL denotes the net impedance of loads. Therefore, the total grid impedance is simplified
as Zg (Zs//ZL). The injected harmonic current and corresponding voltage at the point of common
coupling (PCC) can be expressed as given in equations (6.1) and (6.2). These equations include
the effect of Vx, grid voltage and are also influenced by the combined effect of grid and filter
impedances.

103
 Chapter 6. Harmonic Resonance Issues with PV Penetration

Figure 6.1. Single-phase Equivalent of Grid with PV System Connected at PCC.

𝑉𝑉𝑥𝑥 (ℎ𝑛𝑛 )𝑍𝑍𝐶𝐶𝐶𝐶 (ℎ𝑛𝑛 )−𝑉𝑉𝑔𝑔 (ℎ𝑛𝑛 )𝑍𝑍𝐿𝐿𝐿𝐿 (ℎ𝑛𝑛 )−𝑉𝑉𝑔𝑔 (ℎ𝑛𝑛 )𝑍𝑍𝐶𝐶𝐶𝐶 (ℎ𝑛𝑛 )
𝐼𝐼𝑝𝑝𝑝𝑝𝑝𝑝 (ℎ𝑛𝑛 ) = (6.1)
𝑍𝑍𝑔𝑔 (ℎ𝑛𝑛 )𝑍𝑍𝐶𝐶𝐶𝐶 (ℎ𝑛𝑛 )+𝑍𝑍𝐿𝐿𝐿𝐿 (ℎ𝑛𝑛 )𝑍𝑍𝐶𝐶𝐶𝐶 (ℎ𝑛𝑛 )+𝑍𝑍𝐿𝐿𝐿𝐿 (ℎ𝑛𝑛 )𝑍𝑍𝑔𝑔 (ℎ𝑛𝑛 )

𝑉𝑉𝑥𝑥 (ℎ𝑛𝑛 )𝑍𝑍𝑔𝑔 (ℎ𝑛𝑛 )𝑍𝑍𝐶𝐶𝐶𝐶 (ℎ𝑛𝑛 )−𝑉𝑉𝑔𝑔 (ℎ𝑛𝑛 )𝑍𝑍𝐿𝐿𝐿𝐿 (ℎ𝑛𝑛 )𝑍𝑍𝐶𝐶𝐶𝐶 (ℎ𝑛𝑛 )
𝑉𝑉𝑝𝑝𝑝𝑝𝑝𝑝 (ℎ𝑛𝑛 ) = (6.2)
𝑍𝑍𝑔𝑔 (ℎ𝑛𝑛 )𝑍𝑍𝐶𝐶𝐶𝐶 (ℎ𝑛𝑛 )+𝑍𝑍𝐿𝐿𝐿𝐿 (ℎ𝑛𝑛 )𝑍𝑍𝐶𝐶𝐶𝐶 (ℎ𝑛𝑛 )+𝑍𝑍𝐿𝐿𝐿𝐿 (ℎ𝑛𝑛 )𝑍𝑍𝑔𝑔 (ℎ𝑛𝑛 )

Where, the symbols hn represent the nth harmonic order.

The resonance phenomenon associated with the PV system in the distribution network is a
complex combination of parallel and series resonance. The inverter output impedance for an
LC type filter is Zinv= ZLf//ZCf. Therefore, for the given circuit, the net impedance of the network
seen at the PCC is as in equation (6.3). For instance, if the network includes shunt capacitor
banks then the ZCf can be a combination of filter and capacitor banks.
𝑍𝑍𝐿𝐿𝐿𝐿 𝑍𝑍𝐶𝐶𝐶𝐶 𝑍𝑍𝑔𝑔
𝑍𝑍ℎ𝑎𝑎𝑎𝑎_𝐴𝐴𝐴𝐴 = (6.3)
𝑍𝑍𝑔𝑔 𝑍𝑍𝐶𝐶𝐶𝐶 +𝑍𝑍𝑔𝑔 𝑍𝑍𝐿𝐿𝐿𝐿 +𝑍𝑍𝐿𝐿𝐿𝐿 𝑍𝑍𝐶𝐶𝐶𝐶

For a parallel resonance condition, the impedance reaches its maximum value when the
denominator approaches the minimum value that can be represented as given in (6.4).

1 𝐿𝐿𝑓𝑓 +𝐿𝐿𝑔𝑔
𝑓𝑓𝑟𝑟𝑟𝑟𝑟𝑟 = � (6.4)
2𝜋𝜋 𝐶𝐶 𝑓𝑓 𝐿𝐿𝑓𝑓 𝐿𝐿𝑔𝑔

In contrast, the inverter output impedance can also form a series circuit with the associated
network impedance. Therefore, in this case the impedance seen at the source side is given in
(6.5). For a series resonance condition the impedance becomes minimum when the numerator
value approaches the minimum value, which is also expressed as given in equation (6.4).
𝑍𝑍𝐿𝐿𝐿𝐿 𝑍𝑍𝐶𝐶𝐶𝐶 +𝑍𝑍𝑔𝑔 𝑍𝑍𝐶𝐶𝐶𝐶 +𝑍𝑍𝑔𝑔 𝑍𝑍𝐿𝐿𝐿𝐿
𝑍𝑍ℎ𝑎𝑎𝑎𝑎_𝐶𝐶𝐶𝐶 = (6.5)
𝑍𝑍𝐶𝐶𝐶𝐶 +𝑍𝑍𝐿𝐿𝐿𝐿

As the number of parallel PV inverters increases in the distribution network, the resonance
frequency might be altered. For identical PV systems, the filter parameters are equivalent to

104
 Chapter 6. Harmonic Resonance Issues with PV Penetration

Lf/n and nCf. Therefore, the modified resonance frequency can be obtained as given in equation
(6.6).

1 𝐿𝐿𝑓𝑓 +𝑛𝑛𝑛𝑛𝑔𝑔
𝑓𝑓𝑟𝑟𝑟𝑟𝑟𝑟 = � (6.6)
2𝜋𝜋 𝑛𝑛𝑛𝑛𝑓𝑓 𝐿𝐿𝑓𝑓 𝐿𝐿𝑔𝑔

In practice, PV systems consist of PV panels, Maximum Power Point Tracking (MPPT), a

boost converter, a DC/AC inverter and an interfacing LC filter as shown in Figure 6.2. The
harmonic emission issues related with PV inverters strongly depend on the type of control
scheme and the output filter design which can configure a resonance circuit with the grid.
Frequent dynamic interactions between multiple PV inverter system control loops and the
supply network can also excite the potential resonance conditions in the grid. Moreover, during
the resonance, the amplification of voltage and current harmonics might exceed IEEE 519
standard limits [53], even if each PV system satisfies the standards in the network.

Figure 6.2. Detailed Three-phase PV System Configuration.

6.2.1. Impact of Grid Connected PV Inverter Control Strategies

To analyse the impact of a PV system controller interface on the grid harmonic resonance
occurrences, two types of control topologies have been considered. Figures 6.3 and 6.4 show
the control block diagrams of the Current Controlled Voltage Source Inverter (CCVSI) and
Current Regulated Current Source Inverter (CRCSI) type PV systems respectively [114, 125].

105
 Chapter 6. Harmonic Resonance Issues with PV Penetration

Figure 6.3. Control Circuit of Current Controlled Voltage Source Inverter.

The current regulated PWM control method is applied to operate the voltage source PV
inverter. This control topology has two controllers’, namely outer DC voltage controller and
current controller as shown in Figure 6.3. The DC voltage controller controls the dc capacitor
voltage by varying the active current injections. The current controller utilizes the measured
inverter output currents and output from the DC controller to generate references for the
switching. It then regulates the output current indirectly by regulating the voltage at the PV
inverter output. According to the control scheme, the VSI output voltage (Vx) is determined by
the control parameters and PV output power. The power flow variations or disturbance caused
by PV inverters due to transients in solar irradiation can have an inevitable impact on the
controller functions, which can influence the inverter output impedance. The dynamic
behaviour of the closed-loop control system can affect the distribution network harmonic
levels. The detailed PV inverter model has been considered to study the influence of PV power
transients. The resultant PV inverter output impedance involving control system parameters in
s-domain is expressed as given in equation (6.7).
𝑠𝑠𝐿𝐿𝑓𝑓 +𝐺𝐺𝑑𝑑𝑑𝑑 (𝑠𝑠)𝐺𝐺𝑖𝑖 (𝑠𝑠)𝐺𝐺𝑝𝑝𝑝𝑝𝑝𝑝 (𝑠𝑠)
𝑍𝑍𝑖𝑖𝑖𝑖𝑖𝑖 (𝑠𝑠) = (6.7)
𝐿𝐿𝑓𝑓 𝐶𝐶𝑓𝑓 𝑠𝑠 2 +𝐶𝐶𝑓𝑓 𝐺𝐺𝑖𝑖 (𝑠𝑠)𝐺𝐺𝑝𝑝𝑝𝑝𝑝𝑝 (𝑠𝑠)𝑠𝑠+𝐺𝐺𝑑𝑑𝑑𝑑 (𝑠𝑠)𝐺𝐺𝑖𝑖 (𝑠𝑠)𝐺𝐺𝑝𝑝𝑝𝑝𝑝𝑝 (𝑠𝑠)

It is noted that impedance Zinv depends on the control loop gain factors and PWM switching.
The interaction of PV output impedance with the grid can modify the total network impedance
(i.e., as in equations (6.3) or (6.5)), which in turn shifts the harmonic resonance frequencies in
the distribution network. The current and voltage controllers are tuned by a trial and error
method to achieve optimum performance with faster response time and minimum overshoot.

106
 Chapter 6. Harmonic Resonance Issues with PV Penetration

Conversely, current regulated CSI PV topology utilizes grid voltages to track the phase
angle, which is used to synchronize the PV inverter output with the network voltage. The
preferred real and reactive powers are compared with the actual generated powers and the error
is processed through PI controllers to produce direct and quadrature components of real and
reactive currents. Then by using inverse synchronous reference frame, the stationary reference
currents will be generated, which mainly depend on the transients in the PV system output.
This control technique uses a hysteresis controller to generate firing pulses. This type of current
controller is simple and faster in dynamic response [125]. Similar to the CCVSI control
scheme, the inverter output impedance also comprises the gain factors of control loops.

Figure 6.4. Control Circuit of Current Regulated Current Source Inverter.

The analysis has been carried out to investigate the influence of different PV inverter system
integrations and their operational interference on the harmonic resonance phenomena. This is
developed and tested in PSCAD software to assess the PV inverter output harmonic limits
according to the standards IEEE 929-2000 [106] and AS 4777.2:2005 [107].

6.3. Test System Approach and Simulation Study

The IEEE-13 bus distribution network is used for this study, which was shown in Figure 5.1
(refer Chapter 5). The investigation has been carried out on two case studies by considering
with and without the presence of background harmonic distortions in the supply network.
Loads are modelled as constant power, constant impedance and constant current loads for the
first case study as explained in the previous chapter. The simulations have been performed to
verify the load flow and harmonic distortions in the network. The parallel connection of these
loads can have less influence on the grid impedance, therefore the impedance Zg (refer Figure
6.1) is approximately equal to Zs.

107
 Chapter 6. Harmonic Resonance Issues with PV Penetration

In the second case study, the test system loads have been replaced with nonlinear loads for
harmonic injection. For this purpose, different types of nonlinear load such as power electronic
based single-phase diode bridges with RL, RC loads and motor and heating loads have been
considered [39]. These loads include CFL, computer, TV, adjustable speed drives (ASD), DC
drives and single-phase induction motor and lift motors as mentioned in Chapter 5. Therefore,
the network impedance includes the effect of these loads, which might shift the system
resonance frequencies. In the IEEE-13 bus network, the detailed load models are aggregated in
each phase and distributed at the feeder nodes from load_1 to load_4. Simulations have been
performed to verify the injected harmonic levels in the distribution network to meet IEEE 519
standard. According to IEEE 519:2014 standard, the voltage THD for <1kV voltage level
should be less than 8% and for 1kV<V<69kV level the THD should be less than 5%. In
addition, the current Total Demand Distortion (TDD) limit should be less than 5%. In this
paper, all the simulations have been performed at full load condition to eliminate the
complexity in determining the harmonic distortion levels, so TDD is the same as THD.

6.4. Results and Discussion

In this research, harmonic resonance phenomena on the network associated with the PV
systems have been assessed through simulations on the IEEE 13 bus network as well as with
harmonic field measurements. Here, the harmonic resonance conditions are identified by the
amplification of THDs and frequency versus impedance scan plots.

6.4.1. Simulation Study on IEEE-13 Bus Distribution Network

0.40 Time Vs 1050

0.35 PV output powers and Sunlight 900
PV Power (MW)

0.30
Sunlight(W/m2)

750
0.25 Ppv
600
0.20 Qpv
450
0.15
0.10 300
0.05 150
0.00 0
0.0
0.7
1.3
2.0
2.7
3.3
4.0
4.7
5.3
6.0
6.7
7.3
8.0
8.7
9.3
10.0
10.7
11.3

Time (Sec)

Figure 6.5. Time versus Solar Irradiation and PV Output Pattern.

In this research, the analysis has been performed in two case studies as mentioned earlier.
In each case study, CRCSI and CCVSI based PV systems have been analysed. In each case
study, the dynamic operation of two PV inverter control topologies for variations in solar
irradiation have been considered to characterise the harmonic resonance modes in the network.

108
 Chapter 6. Harmonic Resonance Issues with PV Penetration

Regarding this, a comprehensive model of PV source has been considered. This model has all
the dynamic characteristics similar to a real PV system. In this study, a typical solar irradiation
profile is applied to the PV system and the corresponding PV active and reactive powers are
shown in Figure 6.5. It is noticed that the output of the PV system has a linear correlation with
solar irradiation.

6.4.1.1. Case Study-I: Without Background Distortions in the Network and with
PV Systems
In the first scenario, a CRCSI PV inverter system of 330kWp capacity is considered, which
operates at close to unity power factor. Simulations have been carried out to analyse the
harmonic distortion levels in the grid with PV operation at both rated and decreased output
power conditions. In every simulation, 5 PV systems of the same capacity are connected
consecutively at the PCC to estimate the severity of harmonic contributions in the network.

0.8
(a) Voltage THD of Phase-A during Rated and Resonance with CRCSI
0.7 Type PV Systems
V632_A_res
0.6
Vload_1A_res
Voltage THD (%)

0.5 Vpcc_A_res
0.4 V671_A_res
0.3 V632_A

0.2
Vload_1A
Vpcc_A
0.1
V671_A
0
0PV 1PV 2PV 3PV 4PV 5PV
Number of PV Systems

1
(b) Voltage THD of Phase-B during Rated and Resonance with CRCSI
0.9
V632_B_res Type PV Systems
Voltage THD (%)

0.8
Vload_1B_res
0.7

0.6
Vpcc_B_res
0.5 V671_B_res
0.4 V632_B
0.3 Vload_1B
0.2
Vpcc_B
0.1
V671_B
0
0PV 1PV 2PV 3PV 4PV 5PV
Numberof PV Systems

Figure 6.6. Percentage of Voltage THDs at Various Nodes for Phases A and B with CRCSI.

109
 Chapter 6. Harmonic Resonance Issues with PV Penetration

During rated conditions, the measured voltage and current THDs at various buses for both
phases A and B are plotted as dark coloured bars shown in Figures 6.6 and 6.7 respectively.
The subscripts, for instance, 632_A, load_1A, PCC_A, 671_A represent the corresponding
node voltage/current harmonic distortion in the respective phase. Also, 0PV indicates with no
PV condition and the numbers 1PV, 2PV, 3PV, etc. denote the number of PV systems
integrated into the network. The harmonics produced from the PV inverter systems not only
distorts the system voltage but also affects the neighbouring devices due to the harmonic
interface. It can be seen that for every PV inverter integration, the voltage THDs at each node
have increased and the effect is higher at bus 671 and the PCC compared to the supply side
nodes. However, the current THDs at the before integration point is increased due to a reduction
in the current magnitude with PV power injection. It is noticed that the harmonics generated
by the PV systems are mainly due to active power injection and switching harmonics.

4
(a) Current THD of Phase-A during Rated and Resonance with
I632_A_res CRCSI Type PV Systems
3 Iload_1A_res
Current THD (%)

Ipcc_A_res
I671_A_res
2
I632_A
Iload_1A
1 Ipcc_A
I671_A
0
0PV 1PV 2PV 3PV 4PV 5PV
Number of PV systems

36
(b) Current THD of Phase-B during Rated and Resonance with
32
CRCSI Type PV Systems
I632_B_res
Current THD (%)

28
24
Iload_1B_res
Ipcc_B_res
20
I671_B_res
16
I632_B
12
Iload_1B
8
Ipcc_B
4
I671_B
0
0PV 1PV 2Pv 3PV 4PV 5PV
Number of PV systems

Figure 6.7. Percentage of Current THDs at Various Nodes for Phases A and B with CRCSI.

Subsequently, simulations have been continued for decreasing PV power trends (rest of the
time from 5-12 sec) as indicated in Figure 6.5. Remarkably, in all the simulations the THD of
voltage and currents in the grid are elevated abruptly during the period of time 5-6.5 sec. At

110
 Chapter 6. Harmonic Resonance Issues with PV Penetration

this instant, the values of voltage and current THDs were also measured and presented as
pattern bars in Figures 6.6 and 6.7 respectively. It can be seen that the network current THD is
more exaggerated than the voltage THD. In Figure 6.6, the net induced harmonic voltage is
increased by nearly 0.2% for each PV system integration. Figure 6.7 shows that the current
THDs have remarkably raised at supply side nodes in each step of PV system integration during
this period. It can be seen that the current THD at the before connection point was calculated
to be high in phase B of around 35% compared to the current THD during rated conditions with
5 PV systems. This value is outside the THD limits according to IEEE 519 standards.
Conversely, phases A and C are less affected and have THD values of less than 5%. The level
of THD in each phase is different due to unbalanced loading levels and line configurations of
the test network. Usually, a decrease in PV power injection causes an increase in supply current
magnitude. Therefore, the grid current THD reduces with respect to the increase in fundamental
current. Despite this fact, during this period the harmonic THD profile follows the inverse
trend.

Even though the PV inverter systems are individually satisfying the harmonic limits during
the rated conditions, these PV systems are negatively interacting for variations in PV power
generation. Furthermore, it is noted that the harmonic emissions from several identical PV
systems connected at the same node could be different due to their independent operation. Also,
the harmonic magnitude and phases are different for each inverter and the resultant harmonics
may result in harmonic superimposition. Here, the amount of increase in harmonic distortion
level is varying with the number of PV inverter systems. As the number of paralleled PV
systems increase, then the output current lower order harmonic magnitudes increase [5], [9].
Moreover, the harmonic injection depends on the current controller performance and filtering.

This PV inverter control topology includes a fixed band type hysteresis current controller
due to its simplicity and response. Despite its advantages, the hysteresis current controller has
an issue of variable switching frequency, which can lead to pollution of the inverter output
current. Moreover, the variation in the PV output power changes the inverter reference currents,
which in turn introduce disturbances in the current controller response. Therefore, the inverter
output currents contain significant harmonics, which are unable to be suppressed by the
interfacing filter, injected into the grid lead to the occurrence of harmonic resonance. The level
of distortions in the current is found to be higher realizing a series resonance condition at this
period of time (i.e., at 5.5 to 6.5sec). Further decrease of PV power injection caused THDs to
reduce in the grid that is due to sustain in the controller response. Severe harmonic currents

111
 Chapter 6. Harmonic Resonance Issues with PV Penetration

flowing through the grid can cause transformer overheating and overloading of the neutral
conductor.

Consequently, to confirm the system impedance variation due to multiple PV systems

integration a frequency scan has been conducted. The positive and zero sequence impedance
variations with respect to the frequency at the PCC for before and after PV systems integration
are plotted in Figure 6.8. Usually, the positive and negative sequence impedances are similar
to each other but the zero sequence may differentiate them. Note that the network shunt
capacitor banks are in operation during simulations. It can be seen that the grid positive
sequence impedance without PV units has resonance frequencies of 480Hz (8th) and 1440Hz
(24th) at peak and lower peak points respectively, which are not harmful harmonics.

8
Positive and Zero Sequence Impedances at PCC Vs Frequency Plot
7
with and without CRCSI Type PV Systems
Grid Impedance (ohm)

6
|Z+_without PV|(ohms)
5
|Z+_with 5PVs|(ohms)
4
|Z0_without PV|(ohms)
3
|Z0_with 5PVs|(ohms)
2
1
0
0.0 500.0 1000.0 1500.0 2000.0 2500.0 3000.0 3500.0 4000.0
Frequency (Hz)

Figure 6.8. Impedance versus Frequency Scan with and without CRCSI PV Systems at PCC.

However, the system impedance with parallel PV systems integration is changed

significantly and a frequency shift has occurred. It is to be mentioned that the PV inverter
output impedance is a result of the inverter current control and filter components (refer to
equation (6.7)). The frequency of the peak impedance moves nearly at 7th harmonic; however,
the impact is negligible and the harmonic magnitude of the voltage and current are well within
the limits. Conversely, with PV system integration the network lower impedance resonance
frequency is shifted to the 27th harmonic, which is close to the harmonic frequency injected by
the PV systems. Note that the PV system was the only harmonic source connected to the
network. On the other hand, zero sequence impedances with and without PV systems have peak
resonance frequencies at fairly close to 240Hz shown in Figure 6.8. However, similar to the
positive sequence there is a frequency shift that occurred at the lower impedance peak from
18th to 21st harmonic order, which is close to the harmonic frequency. At these resonance
frequencies, the harmonic current emissions from the PV systems are amplified and injected
into the grid current, which leads to an excitation of series resonance.

112
 Chapter 6. Harmonic Resonance Issues with PV Penetration

In the second scenario, a CCVSI type PV inverter has been considered for the harmonic
analysis. The PV system capacity is 330kWp and the switching frequency is chosen to be
2460Hz (41X60Hz). Similar to the previous scenario, the PV systems were integrated gradually
in the test network at node PCC. Simulations have been carried out for rated and reduced PV
output power levels. During rated conditions, the THD values possessed similar trends as in
the first scenario, which can be observed in Tables 5.3 and 5.4 (refer Chapter 5). The increase
in voltage THD with PV system is observed to be less than 1%. Moreover, due to supply current
magnitude reduction the current THD is increased to about 25% on the supply side, which is
at the before PV integration node. However, this current THD value is quite severe according
to IEEE 519 standard. Further, the assessment has been performed for varying solar irradiation
conditions. It is observed that the generated harmonics depend on the active power produced
from the PV systems. From the results, the net voltage and current harmonic profile at after
and before the PCC can be proposed as given in equations (6.8) and (6.9) respectively.
𝑃𝑃𝑎𝑎𝑎𝑎𝑎𝑎
𝑇𝑇𝑇𝑇𝑇𝑇𝑉𝑉 𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 𝑃𝑃𝑃𝑃𝑃𝑃 = 𝑇𝑇𝑇𝑇𝑇𝑇𝑉𝑉_𝑤𝑤𝑤𝑤𝑤𝑤ℎ𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 + 𝑖𝑖 ∗ (∆𝑇𝑇𝑇𝑇𝑇𝑇) ∗ �% � (6.8)
𝑃𝑃𝑟𝑟

𝑃𝑃𝑎𝑎𝑎𝑎𝑎𝑎
𝑇𝑇𝑇𝑇𝑇𝑇𝐼𝐼 𝑎𝑎𝑎𝑎 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑃𝑃𝑃𝑃𝑃𝑃 = 𝑇𝑇𝑇𝑇𝑇𝑇𝐼𝐼_𝑤𝑤𝑤𝑤𝑤𝑤ℎ𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 + 𝑖𝑖 ∗ (∆𝑇𝑇𝑇𝑇𝑇𝑇) ∗ �% � (6.9)
𝑃𝑃𝑟𝑟

Where, ∆THD is the voltage or current THD contribution with 1PV system (i.e., THDwith2PVs
- THDwith1PV); i = number of PV inverters; Pact and Pr refer to the actual and rated PV powers
respectively.

Conversely, the network has not been affected by resonance phenomenon as compared with
the first scenario. However, the harmonic resonance conditions are likely to change in the
network and are influenced by the connected load, type of PV system penetration and their
controller interactions.

6.4.1.2. Case Study-II: With Background Distortions in the Network and With
PV Systems

In this regard, modern power electronic based nonlinear loads have been connected, which
inject harmonics into the test system. Simulations have been performed and the measured
voltage THD is in the range 2.5 - 3% and the current THD is calculated to be 4%, which are
within the limits as per IEEE 519 standard. In the first scenario, a CRCSI PV system has been
considered and simulations have been performed in the same way as the previous case study.
Here, the harmonic emissions from the PV system contain switching harmonics and the
harmonics due to distortion in the supply. Simulated results show that the voltage THD with

113
 Chapter 6. Harmonic Resonance Issues with PV Penetration

5PV system penetration is increased to 4.5-5.2% at the feeder end node 671 during rated
conditions. Also, the current THD values have attained a high value of 9.5% and 7.6% at before
and after the PV integration point, these exceed IEEE 519 limits. It is estimated that the
harmonic levels in the network are observed as given in equations (6.8) and (6.9), which depend
on the number of PV systems and the percentage of produced PV power. The harmonic
distortion levels for rated PV operation can also be observed in Tables 5.5 and 5.6 (refer
Chapter 5).

As in the second scenario, simulations have been carried out with the integration of CCVSI
type PV systems in the same way. During the rated PV power level, the measured voltage and
current THDs for phases A and B at various nodes are plotted (coloured bars) as shown in
Figures 6.9 and 6.10 respectively. It can be observed that the network voltage and current THDs
due to PV systems are increased to nearly 4.8% and 9.5% in phase-B respectively.
Subsequently, simulations have been performed for decreasing solar irradiation levels. It can
be seen that during the period from 5 to 6.5 sec, the voltage THD is increased from 4.8% to
5.4% with 5 PV systems at 671 node, which is violating the IEEE limits. It can be noticed that
the rapid rise in voltage harmonic distortion is simulating a parallel resonance phenomenon.
During the resonance, amplification of the 5th and 7th harmonics are more noticeable compared
to the 3rd order harmonic.
6
(a) Voltage THD of Phase-A during Rated and Resonance V632_A_res
with CCVSI Type PV Systems
Vload_1A_res
Voltage THD (%)

Vpcc_A_res
4
V671_A_res
3
V632_A

2 Vload_1A

Vpcc_A
1
0PV 1PV 2PV 3PV 4PV 5PV V671_A
Number of PV Systems

114
 Chapter 6. Harmonic Resonance Issues with PV Penetration

6
(b) Voltage THD of Phase-B during Rated and Resonance with V632_B_res
CCVSI Type PV Systems
Voltage THD (%) 5
Vload_1B_res

4 Vpcc_B_res

V671_B_res
3
V632_B
2 Vload_1B

Vpcc_B
1
0PV 1PV 2PV 3PV 4PV 5PV
V671_B
Number of PV Systems

Figure 6.9. Percentage of Voltage THDs at Various Nodes for Phases A and B with CCVSI.

(a) Current THD of Phase-A during Rated and Resonance with

6
CCVSI Type PV Systems I632_A_res
5 Iload_1A_res
Current THD (%)

Ipcc_A_res
4
I671_A_res
3 I632_A

2
Iload_1A
Ipcc_A
1
0PV 1PV 2Pv 3PV 4PV 5PV I671_A
Number of PV Systems

11
(b) Current THD of Phase-B during Rated and Resonance with I632_B_res
10
CCVSI Type PV Systems
9 Iload_1B_res
Current THD (%)

8
Ipcc_B_res
7
6 I671_B_res
5 I632_B
4
Iload_1B
3
2 Ipcc_B
1
I671_B
0PV 1PV 2PV 3PV 4PV 5PV
Number of PV Systems

Figure 6.10. Percentage of Current THDs at Various Nodes for Phases A and B with CCVSI.

The current reference for each inverter is produced independently and is updated by the
MPPT controller. The change in the inverter current reference due to solar irradiation variations
is utilized in the current controller loop as shown in Figure 6.3, which can introduce a resonance
condition in a few fundamental cycles. In addition, this type of PV inverter uses grid voltages

115
 Chapter 6. Harmonic Resonance Issues with PV Penetration

as a reference source for synchronization. Therefore, the variation in grid voltage harmonics
also affects the inverter current controller performance. Hence, the dynamic operation of the
current controller due to transient disturbances has increased the oscillations and ripples at the
output of the PV inverter. As a result, the PV inverter injected high harmonic components into
the network and was unable to be adequately suppressed by filters. These harmonic currents
from the PV systems combining with network background harmonics result in amplification
of voltage distortion at the PCC. The increase in voltage distortion depends on both the system
impedance variation and the aggregate of harmonic currents injected by the PV system.
Excessive voltage distortions can reduce the lifetime of the equipment connected to the network
and increases the losses.

The system equivalent impedance will be different for the penetration of different PV
inverter technologies. Similar to the previous case study, to verify the system impedance
variation with the PV system integration, frequency scan analysis has been conducted. The
frequency versus positive and zero sequence impedances are plotted in Figure 6.11 in the case
of with and without 5PV systems integration. It is to be observed that the system with nonlinear
loads has increased the system impedance and also affected the system resonance frequency
points. The system impedance has both series and parallel resonance frequencies as in the
previous case. In comparison, the lower peak resonance frequency of positive sequence
impedance has been shifted from the 24th (refer Figure 6.8) to almost the 22nd harmonic due to
the presence of nonlinear loads. However, the peak impedance has a resonance frequency close
to 480Hz (8th), which remains the same. Moreover, the zero sequence peak and lower
impedance resonant frequency points were not altered, which have frequencies at the 4th and
18th orders respectively.
8 Positive and Zero Sequence Impedance at PCC Vs Frequency Plot with
Grid Impedance (ohm)

7 and without CCVSI Type PV Systems and with Nonlinear Loads

6
|Z+_withoutPV|(ohms)
5
|Z+_with5PVs|(ohms)
4 |Zo_withoutPV|(ohms)
3 |Zo_with5PVs|(ohms)
2
1
0
0.0 500.0 1000.0 1500.0 2000.0 2500.0 3000.0 3500.0 4000.0 4500.0 5000.0
Frequency (Hz)

Figure 6.11. Impedance versus Frequency Scan with and without CCVSI PV Systems at
PCC.

116
 Chapter 6. Harmonic Resonance Issues with PV Penetration

In addition, PV systems penetration at the PCC causes the sequence impedances to increase
their magnitude at the peak resonant points and a frequency shift has also occurred. It can be
seen that the peak resonance frequency is close to 7th order i.e., a shift in the frequency from
480Hz to 420Hz. In addition, the zero sequence impedance moves the peak resonance
frequency to 300Hz. It is observed that these frequencies are close to the harmonic frequencies
injected by both PV systems and nonlinear loads, which lead to higher voltage distortion in the
network, achieving a parallel resonance condition.

6.4.2. Analysis of Harmonic Measurements at The University of Queensland PV

Site
The UQ St Lucia campus has a large power network supplied by an 11kV line from the local
distribution utility. The network is highly loaded with a broad range of machine loads, modern
electronic devices and air conditioners. The total maximum load demand is about 23MW and
has peak power demand between 6am to 6pm. The network is employed with nearly 40
substations to step-down the voltage from 11kV to 415V as mentioned in the previous chapter.
The university has an installed solar PV system of 1.5 MW which is distributed across nine
buildings to feed the peak demand. To evaluate the PV inverters interaction in the real
distribution network, the harmonic study has been carried out at substation 11. At which, the
PV array of 340kW capacity is installed individually on car parks 98A and 98B. Special three-
phase grid-tied PV inverters are used to synchronize the PV output with the grid voltage and
frequency. Substation 11 has two transformers of 750kVA, 11kV/415V as shown in Figure
5.13 (refer Chapter 5). It is important to note that the university network does not contain shunt
capacitor banks; however, it has underground cables that can cause mutual capacitance between
phases and ground.
1600 1400
Time versus PV Output Power and Sunlight
Solar Irradiation( W/m2)

1400 1200
Power Sunlight 11am
PV Power (kW)

1200
P=1011.7kW, 1000
1000 11am
P=869.23kW 800
800
600
600
2pm 1pm
400
400 P=937.53kW P= 1154kW
200 200

0 0

24/09/2013 8/11/2013 24/10/2013 11/04/2014

Date and Time

Figure 6.12. Time versus PV Output Power and Sunlight.

117
 Chapter 6. Harmonic Resonance Issues with PV Penetration

For the analysis, the harmonic field measurement data of 1-year was been collected through
various power quality (PQ) meters connected at high voltage (HV) and low voltage (LV) levels
of the network. These PQ meters recorded the data for every 5ms time resolution and then the
data aggregation system recorded the data on an hourly basis. Figure 6.12 shows the time versus
PV output power and sunlight for sunny and cloudy days, which were obtained using weather
sensors.
7
Voltage THD (%) of HV Line during Sunny and Cloudy Days
6
Voltage THD (%)

5
4
3
2
1
0
0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00
12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00
24/09/2013 8/11/2013 24/10/2013 11/04/2014
Date and Time
Va_11kVline Vb_11kVline Vc_11kVline

Figure 6.13. Percentage of Voltage THD at HV Side for Parallel Resonance Condition.

The high penetration PV systems and their associated operational behaviours affect the grid
harmonic levels, which may excite resonance conditions in the network as discussed in section
6.2. In this study, the network resonance conditions are identified and analysed through
amplification factors of voltage and current THDs. During the selected days, the corresponding
voltage THDs of the HV line are plotted in Figure 6.13. Results indicate that significant peak
voltage distortions are found on the HV side. It is noticed that the variation in PV output power
due to fluctuations in solar irradiation has a considerable impact on network harmonic levels
at different time intervals. The results show that the voltage THD is notably increased to about
6.25% during PV operation, which is violating the 5% limit as per EEE 519 standards. It can
be seen that the voltage THD is more prominent in the network and might appear to be realizing
a parallel resonance phenomenon similar to the simulation case study. The exact time and the
corresponding PV power are indicated in Figure 6.12 at which the harmonic resonance has
been noticed. Besides, the network THD levels other than these time periods are below 1.5%.

118
 Chapter 6. Harmonic Resonance Issues with PV Penetration

3
Transformer Voltage THD (%) during Sunny and Cloudy Days
2.5

Voltage THD (%)

2

1.5

0.5

0
0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00
12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00
24/09/2013 8/11/2013 24/10/2013 11/04/2014
Date and Time
Va_Tx2 Vb_Tx2 Vc_Tx2

Figure 6.14. Percentage of Voltage THD at the Distribution Transformer TX2 for Parallel
Resonance Condition.
Transformer Current THD (%) during Sunny and Cloudy Days
25
Current THD (%)

20

15

10

0
0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00
12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00
24/09/2013 8/11/2013 24/10/2013 11/04/2014
Date and Time
Ia_Tx2 Ib_Tx2 Ic_Tx2

Figure 6.15. Percentage of Current THD at the Distribution Transformer TX2 for Parallel
Resonance Condition.

In addition, the voltage and current THDs at one of the transformers (LV side TX2) were
collected for the corresponding days as shown in Figures 6.14 and 6.15 respectively. It can be
seen that the voltage distortion has peak values during resonance, which lies between 2.5% to
2.8%. It is essential to specify that the university has already been equipped with a capacitive
voltage transformer which eliminates the high harmonic components. As a result, the harmonic
levels are shown to be damped and are well within the limits. However, these distortions could
probably be higher in the network without filtering equipment. Further, the current THDs are
increased from 8% to 20% due to an increase in PV power injection, which are over the limits
as specified by the standards.

119
 Chapter 6. Harmonic Resonance Issues with PV Penetration

3
PV Array Voltage THD (%) during Sunny and Cloudy Days
2.5

Voltage THD (%)

2
1.5
1
0.5
0
0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00
12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00
24/09/2013 8/11/2013 24/10/2013 11/04/2014
Date and Time
Va_PV Vb_PV Vc_PV

Figure 6.16. Percentage of Voltage THD at the PV Array Output for Parallel Resonance
Condition.

80 PV array Current THD (%) during Sunny and Cloudy Days

70
Current THD (%)

60
50
40
30
20
10
0
0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00
12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00
24/09/2013 8/11/2013 24/10/2013 11/04/2014
Date and Time
Ia_PV Ib_PV Ic_PV

Figure 6.17. Percentage of Current THD at the PV Array Output for Parallel Resonance
Condition.

Moreover, the measured voltage and current THDs at car park 98A PV array are plotted as
shown in Figures 6.16 and 6.17 respectively. This PV system utilizes grid voltages as a
reference source similar to the CCVSI type system. Although the PV system controller was
initially designed to avoid resonance frequency during installations, the PV system switching
and filter components can cause negative interactions with the grid. Due to instant variation in
solar irradiation the PV inverter current controller can be caused to malfunction, which can
inject distorted current into the grid. The injection of harmonic currents from several parallel
connected PV inverters and their interaction can cause the PCC voltage to distort, which can
lead to triggering of the resonance condition. Figure 6.17 shows the current THD is measured
with respect to the actual current magnitude. It is observed that the current THD has an inverse
correlation with the solar irradiation [9]. Results show that the voltage THDs at the PV array
system integration node are more noticeable and have values nearly 2.8% during resonance.

Consequently, the network can also be subjected to series resonance. Therefore, to verify
the harmonic interaction issues related to the PV inverter system performance on the network,

120
 Chapter 6. Harmonic Resonance Issues with PV Penetration

the PQ data has been analysed. The time versus PV output power and solar irradiation for the
selected days, of sunny and average cloudy days, are plotted in Figure 6.18.

1400
Time versus PV Output Power and Sunlight 1400

Solar Irradiation(W/m2)
1pm
1200 P=1018kW 1200
PV Power (kW)

Power Sunlight
1000 1000
800 12pm 800
2pm P=991.15kW
600 P=1017kW 600
12pm
400 400
P=835.15kW
200 200
0 0

3/08/2013 8/11/2013 24/10/2013 11/04/2014

Date and Time

Figure 6.18. Time versus PV Output Power and Sunlight.

15
Current THD (%) of HV Line during Sunny and Cloudy Days
Current THD (%)

10

0
0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00
12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00
3/08/2013 8/11/2013 24/10/2013 11/04/2014
Date and Time
Ia_11kVline Ib_11kVline Ic_11kVline

Figure 6.19. Percentage of Current THD at HV Side for Series Resonance Condition.

In the same way, the harmonic measurement data has been collected at both the HV and LV
side of the network for the respective days. The current THD of the HV line for the
corresponding sunny and cloudy days is plotted in Figure 6.19. Due to unbalance present in the
network, each phase possessed different THD values. It can be seen that the harmonic levels
are shown to be increased due to PV power injection. Remarkably, it can be observed that the
current THDs were significantly raised to high values of about 12-13% for sunny days and
around 9-11% in case of cloudy days, which are violating IEEE 519 standard limits.

121
 Chapter 6. Harmonic Resonance Issues with PV Penetration

35
Transformer Current THD (%) during Sunny and Cloudy days
30

Current THD (%)

25
20
15
10
5
0
0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00
12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00
3/08/2013 8/11/2013 24/10/2013 11/04/2014
Date and Time
Ia_Tx2 Ib_Tx2 Ic_Tx2

Figure 6.20. Percentage of Current THD at the Distribution Transformer for Series
Resonance Condition.
In addition, the corresponding measured current and voltage THDs for both days at the
distribution transformer TX2 are plotted as shown in Figures 6.20 and 6.21 respectively to
analyse the impact of PV systems. It is noticed that the current THD is increasing from morning
to noon time following the PV power trend. During the resonance periods, the current THD
has values in the range of 20-32%, which are over the limits as per IEEE 519 standards.
However, a small variation is observed in voltage THD during the resonance and are within
the limits.
5
Transformer Voltage THD (%) during Sunny and Cloudy days
Voltage THD (%)

0
0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00

0:00
3:00
6:00
9:00
12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00

12:00
15:00
18:00
21:00

3/08/2013 8/11/2013 24/10/2013 11/04/2014

Date and Time
Va_Tx2 Vb_Tx2 Vc_Tx2

Figure 6.21. Percentage of Voltage THD at the Distribution Transformer for Series
Resonance Condition.

Likewise, the PV output voltage and current harmonics have followed similar patterns
during the selected days as shown in Figures 6.16 and 6.17 respectively. The current harmonics
generation from the PV inverter mainly relies on the type of inverter, control topology and
filter. The negative interactions in PV current controllers increase the oscillations and ripples
at the output of the PV inverter and are dependent on the controller gains. The oscillations tend
to be more severe with an increase in PV system penetration, which in turn affects the grid
inductance and shifts the low impedance resonance point. It is noted that the harmonics from
several parallel PV inverters connected at the PCC are probably increased the oscillations in

122
 Chapter 6. Harmonic Resonance Issues with PV Penetration

the voltage waveform. In turn, the oscillations in the grid reference voltage affect the current
controller performance. The cumulative effect of PV inverter harmonics and the harmonics
present in the network due to loads can cause the occurrence of a series resonance condition.
Frequent occurrences of harmonic resonances can possess serious power quality issues in the
distribution grid such as damaging equipment, transformer heating and mal-operation of
protection devices.

6.5. Summary

This chapter has examined an important issue of harmonic resonance phenomena associated
with PV system dynamic operations and their penetrations on the distribution network. The
analysis was performed on an IEEE 13 bus unbalanced distribution network in two different
case studies, specifically with and without background harmonic distortions in the supply
system. In the simulation case studies, the dynamic operations of two different PV inverter
systems and the solar irradiation variations were taken into account to study the impacts on the
grid harmonic resonance incidence. In addition, a frequency versus impedance scan was
presented with respect to each resonance condition. Moreover, to verify and validate the
simulation results in the case of a real distribution network, the harmonic measurement data of
1-year collected from the UQ PV system were analysed. The measurements at both LV and
HV sides of the network were obtained for analysis. Simulation results show that the harmonic
resonance occurrence is not only reliant on the type of inverter control topology, filters, and
number of PV inverter systems but also on the network configuration and type of customers
load. Even though the PV inverters individually satisfy the harmonic limits during rated
conditions, these PV systems are negatively interacting for variation in solar irradiation levels.
As a result, the PV inverter controller dynamic operation due to transient disturbances injected
high harmonic components into the grid, which excited the resonance conditions in the
network. During the resonance, the amplification of voltage and current harmonics exceeded
the IEEE 519 standard limits. Measurement results indicate that significant peak voltage and
current distortions are found on the HV and LV side of the UQ network and are comparable to
the simulation results. The following chapter presents a new potential solution to mitigate these
power quality impacts on the distribution network with high PV penetration.

123
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

CHAPTER 7

SOLAR-DSTATCOM: A NEW SOLUTION TO

POOR POWER QUALITY

7.1. Introduction1

This chapter proposes a new solution to overcome the power quality challenges influenced
by PV system integration in an unbalanced distribution network. In this regard, a novel
conception of operating a PV inverter as a virtual DSTATCOM, which is one of the custom
power devices, named as Solar-DSTATCOM has been proposed. To mitigate power quality
issues in the case of both balanced and unbalanced network conditions, a new control algorithm
is proposed for the Solar-DSTATCOM.

1
This chapter is organised with the following references:
Partially incorporated in Section 7.3 and 7.4:
Annapoorna Chidurala, T. K. Saha, and N. Mithulananthan, "Power quality enhancement in
unbalanced distribution network using Solar-DSTATCOM," in Power Engineering Conference
(AUPEC), 2013 Australasian Universities, Hobart, TAS, Sept. 29-Oct. 3, 2013, pp. 1-6.
Partially incorporated in Section 7.4:
Annapoorna Chidurala, T. K. Saha, N. Mithulananthan, and R. C. Bansal, "Harmonic emissions in
grid connected PV systems: A case study on a large scale rooftop PV site," in PES General Meeting
Conference & Exposition, 2014 IEEE, National Harbor, MD, 27-31 July 2014, pp. 1-5.
Incorporated as Section 7.5:
Annapoorna Chidurala, T. K. Saha, and N. Mithulananthan, "Real-Time Controller Hardware-in-
the-Loop Simulations of Solar-DSTATCOM," in preparation, to be submitted to Elsevier Electric
Power Systems Research Journal by September 2016.

124
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

The main objective of the Solar-DSTATCOM control strategy is to provide individual phase
voltage regulation, harmonic and reactive power compensation at the integration point. Solar-
DSTATCOM has the ability to provide dynamic reactive power support during the whole 24
hour period along with real power generation during the daytime. In the first part of this
research, the proposed Solar-DSTATCOM controller has been verified in the PSCAD
simulation environment. Subsequently, the assessment has been performed on the IEEE-13 bus
unbalanced distribution network to compensate voltage rise, voltage unbalance and sag/swell
issues. Also, the harmonic emission characteristics for varying solar irradiation conditions has
been presented. Furthermore, the Solar-DSTATCOM controller has been verified in a real-time
Hardware-in-the-Loop simulation platform, which combines the Real-Time Digital Simulator
(RTDS) and dSPACE hardware board.

The following section presents the motivation of this research work. The proposed control
scheme of Solar-DSTATCOM is explained in Section 7.3. The power quality investigation
performed through PSCAD simulation on the IEEE-13 bus unbalanced distribution network
using Solar-DSTATCOM in different case studies is presented in Section 7.4. In Section 7.5,
the proposed Solar-DSTATCOM control system has been tested in a Controller Hardware-in-
the-Loop simulation environment by considering various realistic case studies. Finally, the
summary of this chapter is presented in Section 7.6.

7.2. Motivation

Traditionally, a solar PV system is considered as a real power source which generates active
power during the daytime and remains inactive overnight. Also, the peak power generation
period is during the midday and over the remaining period, the inverter operates at lower than
rated condition due to low solar irradiation levels. Consequently, the inverter capacity and the
investment of this equipment is unutilised most of the time. The injection of only active power
into the grid at near to the load could cause a reduction in power factor and increase the reactive
power requirement. Hence, grid operators have the advantage of putting obligations on the PV
power output, those limitations can be a cost for the PV owners. In addition, PV system owners
may incur further charges for reactive power purchase from the utility.

On the other hand, traditional distribution networks contain fixed/on-load tap changing
transformers at low voltage (e.g. 11 kV in Queensland State of Australia), step voltage
regulators and switched capacitors for providing voltage regulation with respect to the load
variations [126]. However, PV systems are sensitive to the solar irradiation variations. Frequent

125
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

variation in solar irradiation due to cloud cover can cause significant voltage fluctuations [27,
127], which can lead to an increase in the number of operations of regulatory devices and
therefore reduce equipment lifetime [23]. Also, sometimes the response of the equipment may
be slower than the cloud passage. Consequently, it is necessary to upgrade the regulatory
equipment, which would be relatively expensive for the utility.

To overcome these issues, devices with advanced design and faster control functionalities
are necessary. Traditionally, standard droop control strategies are installed in distribution
networks at centralised, decentralised and localised levels for the voltage quality management.
These control schemes follow four droop-settings which include fixed power factor, fixed
reactive power, reactive power control dependent on the voltage Q-V control and power factor
control dependent on the active power injection PF (P). In the recent past, the concept of
implementing Volt-VAr control schemes in PV inverters are adopted to achieve localised
control. Nowadays, PV systems with smart inverter technology are commercially available,
which have the capability of injecting/absorbing fixed reactive power to address voltage quality
issues to some point [91, 92].

According to the previous standards IEEE 1547 [105] and IEEE 929-2000 [106], the PV
inverters are not allowed to provide voltage regulation and are required to operate at unity
power factor. Nevertheless, recent AS 4777.2-2015 standard [108] proposed that the PV
inverter is allowed to provide constant reactive power with a fixed power factor not exceeding
0.95 leading/lagging. The standard also stated that if the PV inverter is capable of operating in
reactive power mode, the maximum ratio of reactive power to rated apparent power should be
up to 100% leading or lagging. In addition, PV inverters have versatile capabilities similar to
the DSTATCOM by modifying a few functionalities to the inverter controller. In order to
utilize the PV inverter system as a Solar-DSTATCOM, an appropriate control strategy should
be put in place. Numerous types of control methods have been proposed in the literature [87,
88, 93] to produce compensation currents. All these control methods are well suited for steady
state, sinusoidal and balanced conditions of the supply voltage. However, in the case of an
unbalanced supply and nonlinear unbalanced load conditions these control techniques are
unsuccessful. In order to compensate unbalanced network voltages, the active and reactive
power injection from PV systems in each phase has to be individually controlled. Therefore,
this chapter proposes a new control scheme, which independently compensates each of the
three phases to correct any voltage anomalies.

126
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

Establishing this type of inverter can address several integration issues and therefore can
accommodate more PV systems into the network. These PV inverter capabilities eliminate the
buying costs incurred for reactive power needs and by supplying reactive power to the grid
during the night-time can also offer additional profits for the PV owner.

7.3. Solar-DSTATCOM Control Strategy

The novel application of the PV inverter as a Solar-DSTATCOM provides grid voltage

regulation and reactive power support during the daytime and night-time. Figure 7.1 illustrates
the functional block diagram of the grid connected three-phase novel solar PV system
configuration. The proposed new control scheme includes different control blocks to achieve
the desired control aspects.

Figure 7.1. Block Diagram of Three-phase Grid Connected Solar-DSTATCOM System.

7.3.1. DC Voltage Controller

Figure 7.2 shows the control block used to regulate the dc bus voltage consisting of a
proportional-integral (PI) controller, which is a generic feedback controller widely used in
industrial applications. Usually, the DC bus capacitor is charged by solar power during the
daytime. However, during low power times (early morning and evening) and at night-time, as
there is lower or no power generation, the DC bus voltage regulation is applied to provide a
self-supporting DC link voltage. The DC bus voltage controller consumes a small amount of

127
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

active power from the grid and charges the DC capacitor in few cycles to retain the preferred
reference voltage level. The input to this controller is the difference between the PV reference
(Vpv_ref) and the measured voltage across the DC capacitor. The difference is given to the
proportional-integral (PI) regulator by selecting the appropriate proportional gain and integral
time constant to generate the reference active current Id_dc required to maintain the DC-bus
voltage constant. This current represents the losses due to inverter switching devices.

Figure 7.2. PSCAD Model of DC Bus Voltage Regulation Block.

7.3.2. PCC Voltage Controller

This block as shown in Figure 7.3 provides the voltage regulation for each phase at the point
of common coupling (PCC). This new control block uses three individual regulators that will
compare the sensed three-phase grid voltages with the reference voltages of 1p.u. The errors
are processed through PI controllers and generate the reference quadrature currents (Ipcc_a, Ipcc_b,
and Ipcc_c), which will be used to regulate the PCC voltages independently.

Figure 7.3. PSCAD Model of PCC Voltage Regulation Block.

7.3.3. Load Reactive Power and Current Harmonic Compensation Controller

The block diagram of reactive power and the current harmonic compensation controller is
shown in Figure 7.4. A single-phase synchronous reference frame (SRF) based algorithm [128]

128
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

is applied to provide reactive power compensation for each individual phase. This controller
utilizes the individual measured three-phase load currents (IL_a, IL_b, and IL_c) at the PCC which
are treated as the α-axis currents. The application of delay of π/2 radians in each of the currents
produces the quadrature β-axis currents of the imaginary two-phase stationary reference frame
(αβ-frame). Therefore, the equations for three phases in αβ-frame can be expressed as given in
(7.1).

I L _α _ a   I L _ a (ω t) 
=
 I L _ β _ a   I L _ a (ω t − π / 2) 

 I L _ α _b   I L _b (ω t) 
=
 I L _ β _b   I L _b (ω t − π / 2) 
(7.1)
   
 I L _ α _c   I L _c (ω t) 
=
 I L _ β _c   I L _c (ω t − π / 2) 


Figure 7.4. PSCAD Model of Reactive Power and Current Harmonic Compensation
Controller.

129
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

After that, the load current transformations from αβ frame to the dq synchronous reference
frame are attained through a transformation matrix as expressed in equation (7.2). In the
equation, 𝜃𝜃𝑎𝑎 , 𝜃𝜃𝑏𝑏 , and 𝜃𝜃𝑐𝑐 represent the utility phase angles, which are obtained from three single
phase locked loops (PLL). The description of the single-phase PLL circuit is provided in the
following subsection.

 I da  =
 cos θ a sin θ a  I L _α _ a 
 I qa   − sin θ a   I L _ β _ a 
cos θ a 

 I db   cos θ b sin θ b   I L _ α _b 
=
  I L _ β _b 
 I qb   − sin θ b cos θ b 
(7.2)

 I dc  =
 cos θ c sin θ c   I L _ α _c 
 I qc   − sin θ c   I L _ β _c 
cos θ c 

Accordingly, for each phase the fundamental active and reactive components of load

currents are extracted by using Low Pass Filters (LPF), symbolized as Id and Iq respectively.

At this instant, the filter eradicates the load higher harmonic current components from the dq
currents. Subsequently, the total direct current component Id_total is calculated by adding the
generated current from the DC voltage controller to the filtered direct current component Id .

This Id_total current is used to compensate the losses correlated with Solar-DSTATCOM
operation in each phase. Similarly, the subtraction of the PCC controller output quadrature

currents from the filtered quadrature current Iq produces the total quadrature current Iq_total ,

which refers to the total reactive component of the injected reference current used for voltage
regulation at the PCC.
Then by using the inverse dq transformations the controllers will generate the reference
compensation currents for each individual phase which are equivalent to the PCC currents.
This is expressed in the following equation (7.3). In the case of an unbalanced system, the
controller will generate the unbalanced reference currents for compensation. Hence, the PV
inverter currents are produced from the reference currents. In this case, the summation of the
three phase currents cannot be zero, due to the presence of zero sequence components.

130
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

]  da
+ I dc _ loss 
[cos θ a
* I
I= − sin θ a
sfa _ ref
 I qa − I pcc _a


]  db
+ I dc _ loss 
[cos θ b
* I
I= − sin θ b
sfb _ ref
qb − I pcc _b
 (7.3)
 I


]  dc
+ I dc _ loss 
[cos θ c
* I
I= − sin θ b
sfc _ ref
 I qc − I pcc _c



a. Single-phase PLL Block

The single-phase PLL structure shown in Figure 7.5 comprises phase detector (PD), filter
and voltage controlled oscillator (VCO). This block is used for synchronizing the injected
current with the voltages of each individual phase at the PCC. A few methods to create the
orthogonal components are presented in [129]. Here, the delay with a quarter of the
fundamental period is chosen for implementation. The PD tracks the phase angle error of the
relevant PCC voltage by comparing the estimated angle (θ) by PLL irrespective of the other
phases. The error is then given to the loop filter and it generates the estimated frequency (ω)
by adding it to grid frequency (ωff). The integral of ω is the estimated angle (θ) and is fed back
to the PD. In the case of three-phase grid converters, three such PLL structures are needed.
Each PLL circuit closely tracks the phase variation of its corresponding voltage disregarding
the other two grid voltages. Any fault results in an unbalanced phase angle leading to produce
an unbalanced current reference for the synchronisation. Therefore, the injected currents will
synchronize to the respective grid voltages in terms of phase angle and frequency.

Figure 7.5. Single-phase PLL Structure.

131
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

7.3.4. Hysteresis Controller

This controller shown in Figure 7.6 consists of three hysteresis controllers, one for each
phase. The inputs to these controllers are the differences between the reference currents and
the actual sensed PV inverter currents (Iinv_a, Iinv_b, and Iinv_c). When the difference surpasses the
hysteresis band limit, the controller will generate gating signals for the inverter. For example,
a switching instant is given to switch T1 and an opposite signal is applied to switch T4 in order
to avoid a short circuit for the solar farm inverter.

Figure 7.6. PSCAD Model of Hysteresis Controller.

7.4. Power Quality Enhancement in IEEE-13 Bus Unbalanced

Network using Solar-DSTATCOM

The proposed Solar-DSTATCOM has been examined in order to overcome the power
quality issues associated with conventional PV system penetration on the unbalanced
distribution network as presented in Chapter 4, Section 4.2. Similar to the work presented in
Chapter 4, the IEEE-13 bus unbalanced distribution network has been considered as a test
system. For this study, the Solar-DSTATCOM has been developed with a dynamic PV source
to reflect daytime and night-time operation. In the test system, the Solar-DSTATCOM has been
integrated at bus 632. The rating of the PV inverter is 950 kVA, with 900 kWp PV power. The
complete system is developed in PSCAD/EMTDC software. Simulations have been performed
for three different case studies to verify the reliability and robustness of the new control
strategy. In this investigation, the initial two case studies emphasize the compensation of
voltage quality issues, specifically voltage rise, voltage unbalance and sag/swells, raised by PV

132
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

systems penetration. In addition to that, in the final case study the harmonic emission
characteristics of Solar-DSTATCOM are captured during daytime and night-time operations.

7.4.1. Case Study-I: Voltage Rise and Voltage Unbalance Assessment

In this case, simulations have been carried out to study the impact of Solar-DSTATCOM
performance on the test network during the daytime and night-time operations. A typical static
solar irradiation pattern is applied on the Solar-DSTATCOM as shown in Figure 7.7.

Figure 7.7. Static Solar Irradiation Pattern.

The corresponding Solar-DSTATCOM output active power and reactive power with respect
to time is shown in Figure 7.8. Characteristically, the PV active power generation is reliant on
the solar irradiation and has a linear relationship with it. For a unity P.F inverter, the generated
reactive power remains at zero for the whole period. However, to enable the reactive power
capability of the PV inverter, the size of the PV system has been increased by 5% so that around
33% of total capacity is available for reactive power production. It can be observed that during
the daytime, the Solar-DSTATCOM has generated reactive power of nearly 285kVAr as can
be seen in Figure 7.8. The proposed solar-DSTATCOM operates as DSTATCOM during low
PV power levels and at night-time it has provided reactive power support to the grid.

Figure 7.8. Active and Reactive Powers of a 900kW Solar-DSTATCOM.

133
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

Subsequently, the measured p.u RMS voltage magnitudes at various buses of the IEEE -13
bus distribution network with Solar-DSTATCOM during daytime and night-time are presented
in Table 7.1. In the table, at time t= 4 sec the radiation is 1000 W/m2 (assumed as daytime)
and at t=10 sec the radiation is 0 W/m2 (night-time operation). It can be seen that voltage
variation differences (∆V) are very small due to the controller operation and reactive power
exchange and the voltage magnitudes are within the preferred operating limits [-2% to +6%]
according to AS/NZS 61000.3.100 [96]. It can also be noticed the voltage rise magnitudes due
to the PV power injection are lower when compared to the magnitudes with conventional PV
system penetration as presented in Chapter 4, Tables 4.1 to 4.3. It is observed that the Solar-
DSTATCOM has regulated the PCC and thereby the adjacent bus voltages to be within the
limit by applying the PCC voltage controller.

Table 7.1 Voltage Variations during Daytime and Night-time with Solar-DSTATCOM
Phase-A Phase-B Phase-C
RMS Voltage (p.u) RMS Voltage (p.u) RMS Voltage (p.u)
Bus No.
T= 10
T=4 sec T=10 sec ∆VA T= 4 sec ∆VB T= 4 sec T= 10 sec ∆VC
sec
V632 1.01890 1.01851 0.00039 1.03429 1.03057 0.0037 1.02708 1.02440 0.0026
V633 1.01601 1.01550 0.00039 1.03239 1.02863 0.0037 1.02448 1.02185 0.0026
V634 0.99189 0.99163 0.00026 1.01382 1.01009 0.0037 1.00576 1.00299 0.0027
V645 - - - 1.02506 1.02137 0.0036 1.02510 1.02235 0.0027
V646 - - - 1.02333 1.01952 0.0038 1.02303 1.02029 0.0027
V671 0.98776 0.98731 0.00045 1.04510 1.04123 0.0038 0.98772 0.98488 0.0028
V675 0.98125 0.98076 0.00049 1.0475 1.04355 0.0039 0.98584 0.98294 0.0029
V684 0.98583 0.98536 0.00047 - - - 0.98570 0.98281 0.0028

Furthermore, simulations have been performed to verify the voltage unbalance (VU) level
with Solar-DSTATCOM and compared it with conventional PV inverter operation as shown
in Figure 7.9. Results show that the percentage of VU with Solar-DSTATCOM during the
daytime and night-time are much lower compared to the VU with a conventional PV system,
which is less than the 4% limit according to AS/NZS 61000.4.30 [16] standard. The high PV
penetration and conventional PV inverter operation (unity P.F) has increased the VU as
presented in Chapter 4, Section 4.2.1. It can also be noticed that PV system integration into the
distribution grid has increased the VU at the PCC and has a high value at the end of the feeder.
The voltage unbalance with Solar-DSTATCOM followed a similar trend as with conventional
PV operation.

134
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

5.5
5
4.5

Percentage of Voltage
4
3.5

Unbalance
%VU due to Solar-
3
2.5 DSTATCOM during
2 daytime
%VU due to Solar-
1.5 DSTATCOM during
1 nighttime
0.5 %VU due to
0 conventional PV
V632 V633 V634 V671 V692 inverter
V680 V675
Bus No.

Figure 7.9. %Voltage Unbalance Analysis at Various Buses.

7.4.2. Case Study-II: Voltage Sag/Swell Assessment by Fault Conditions

In this case study, simulations have been performed on a test system to verify the
effectiveness of the Solar-DSTATCOM operations in each of the three types of faults. In each
case the fault is initiated at time t = 4.5 sec and the duration of the fault is 100ms. In the figures
below Va, Vb, and Vc represent the three phase RMS bus voltages at before fault and Vaf, Vbf,
and Vcf denote the voltage magnitudes during the fault. Also, LB (0.94p.u) and UB (1.06p.u)
indicate the lower and upper operating band limits from the relevant Australian Standard.

7.4.2.1. Single Line to Ground Fault (SLG):

SLG fault is considered to occur on phase A of bus 633. During the fault, the PV source is
isolated before the inverter operates as a Solar-DSTATCOM. In order to conquer the
unbalanced voltage sags and swells, the Solar-DSTATCOM injects the unbalanced currents
shown in Figure 7.10 (a) based on grid reactive power demand. Respectively, the PCC voltage
regulators measure the RMS voltages of each phase for the control. Here, the phase B voltage
controller operated in inductive mode (absorbed the reactive power) to decrease the voltage at
the PCC. On the other hand, phase A and phase C individual controllers operated in capacitive
mode to alleviate the voltage sag at the PCC. Figure 7.10 (b) depicts the compensated PCC
voltages.

135
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

Figure 7.10. Solar-DSTATCOM (a) Inverter Currents and (b) Compensated PCC Voltages
during SLG Fault.

The RMS voltage magnitudes at different buses before and during the fault are shown in
Figure 7.11. Results show that the magnitude of sag on phase A and phase B have only declined
to about 2% and the swell on phase C was about 2%. Therefore, the Solar-DSTATCOM with
novel control strategy has significantly improved the system voltages to be within the
acceptable limits.

Figure 7.11. The RMS Voltages at Prefault and during SLG Fault with Solar-DSTATCOM.

7.4.2.2. Double Line to Ground Fault (DLG):

In the same way, an unsymmetrical DLG fault is applied on phases A and B of bus 633. Due
to the DLG fault, all the bus voltages resulted in very small sag in all three phases. In contrast,
with the conventional PV inverter operation shown in Figure 4.4 (refer Chapter 4), the voltage

136
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

magnitudes of only phases A and B resulted in a sag condition during the fault. However, in
this scenario, the Solar-DSTATCOM controller injected the unbalance reference currents into
the grid to counterbalance the reactive power as shown in Figure 7.12 (a). Consequently, the
compensated PCC voltages during the fault are shown in Figure 7.12 (b). Similarly, the
mitigated RMS voltages for specific buses were shown in Figure 7.13. As regards to the fault,
the proposed control scheme has mitigated the voltage sags and swells remarkably.

(
b)

Figure 7.12. Solar-DSTATCOM (a) Inverter Currents (b) Compensated PCC Voltages
during DLG Fault.

Figure 7.13. The RMS Voltages at Prefault and during DLG Fault with Solar-DSTATCOM.

137
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

7.4.2.3. Three-phase Symmetrical Fault (TPG):

Simulations were carried out for the TPG fault which is often seen as the worst case, this
was at bus 633. During the fault, the Solar-DSTATCOM injects the appropriate currents for
improving the voltage sags. The injected inverter currents, three phase PCC voltages and the
simulated RMS voltage levels at various buses before and during the fault are shown in Figures
7.14 and 7.15 respectively. Here the magnitudes of injected currents are followed by the grid
reactive power demand. It can be observed that the voltage sag in phase A was nearly 4% and
the voltage magnitudes at phases B and C were improved significantly to have only 3% drop.

Figure 7.14. Solar-DSTATCOM (a) Inverter Currents (b) Compensated PCC Voltages
during TPG Fault.

Figure 7.15. The RMS Voltages at Prefault and during TPG Fault with Solar-DSTATCOM.

138
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

7.4.3. Case Study-III: Harmonic Distortion Analysis

Finally, the harmonic analysis was carried out using Fast Fourier Transformation (FFT) to
obtain harmonic distortions. A similar solar irradiation pattern as shown in Figure 7.7 is applied
to verify the effectiveness of the control scheme in eliminating harmonics generation at the PV
inverter output. The simulated results for the PV inverter current and voltage harmonic
distortions are shown in Figures 7.16 and 7.17 respectively. Figure 7.18 shows the
corresponding individual harmonic spectrums of Solar-DSTATCOM inverter currents and
PCC voltages during the rated condition. According to IEEE 519-2014 [53], the maximum
individual voltage distortion should be less than 5% of the fundamental component and the
total harmonic distortion (THD) should be less than 5% at the PCC. As per IEEE 929-2000
[106], the current THD should not exceed a 5% limit. The injected PV currents should have
low harmonic distortion levels at the PCC to meet the standards.

Figure 7.16. Percentage of THDs of Solar-DSTATCOM Output Currents.

Results show that the proposed harmonic compensation controller has eliminated the
harmonic distortion during the whole period. It can be seen that THD of the Solar-DSTATCOM
current is below 3%, which is lower than the limit prescribed by the standards. This limit
maintained constant even for variable solar irradiation conditions. Similarly, the THD of
voltage is almost below 0.9%.

Figure 7.17. Percentage of THDs of Solar-DSTATCOM Output Voltages.

139
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

Figure 7.18. Individual Harmonic Distortions of Solar-DSTATCOM Output Currents and

PCC Voltages.

Furthermore, the performance of Solar-DSTATCOM has been investigated and verified

comprehensively in a real-time hardware environment, which is presented in the following
section.

7.5. Real-Time Controller Hardware-in-the-Loop Simulations of

Solar-DSTATCOM

The objective of this study is to test and evaluate the Solar-DSTATCOM controller unique
capabilities in the real-time Hardware-in-the-loop simulation environment. Generally, any new
product in industry will be developed through different designing, testing and implementation
phases. In the early phases of a design, initial offline simulations are useful to analyse different
component responses. Nonetheless, these offline simulations are not able to replicate the
runtime behaviour and real-time events of the electrical network, which may not provide the
high fidelity results particularly when designing the control systems and its interfaces. Further,
the amount of real-time required to solve the mathematical equations and control functions
during a given time-step may be smaller or longer than the simulation time step, which may
not influence the accuracy of the results [130]. The solving speed usually depends on the
computational power of the PC and the system complexity. Therefore, potential design errors
will not be identified, which need to be critically investigated in a real-time simulation
environment in addition to the hardware.

Importantly, controller hardware-in-the-loop (HIL) simulations require several

considerations due to the involvement of real hardware and software, which need to take care.

140
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

Since the control system is developed in software and implemented on hardware, which
depends on the execution time, discreteness, errors due to fixed time steps, nonlinearities and
communication delay are essential to be considered. The dynamic performance evaluation and
testing of digital controller prototypes and its corresponding power electronic equipment are
highly recommended before the physical implementation of the product. This real-time HIL
testing can be advantageous to conduct various systematic and random tests under normal and
worst case operating conditions, which can eliminate potential hazards and the required
equipment costs [131]. In references [132-134], the feasibility and effectiveness of the
proposed controllers for grid connected PV inverters and wind energy conversion system were
evaluated in real-time controller HIL simulations, which utilize the integrated platform of
RTDS and dSPACE board. In [135], a unified controller for multiple DG inverter systems and
its interaction with a microgrid system has been verified in real-time hardware simulations.

Therefore, in this research real-time controller HIL simulations have been performed to test
and validate the potential of a Solar-DSTATCOM controller in achieving the preferred control
concepts. This real-time digital simulation involves industry standard simulation tools, RTDS
[118, 119], dSPACE DS1103 hardware control board [120] and MATLAB/Simulink software
[136]. The assessment has been carried out to verify the performance of the Solar-DSTATCOM
controller during the daytime for different solar irradiation profiles and also during the night-
time. Conversely, the load profile variation interactions on the Solar-DSTATCOM operation,
which is operating at various power levels have been evaluated. Finally, the harmonic analysis
of Solar-DSTATCOM has been performed through Fast Fourier Transform.

The following Section 7.5.1 presents the concept of controller hardware-in-the-loop

simulation environment. Section 7.5.2 provides the design of the Solar-DSTATCOM controller
and test system setup. The results of the various case studies and discussions are presented in
Section 7.5.3.

7.5.1. Controller Hardware-in-the-loop Simulation Environment

Controller hardware-in-the-loop simulation is a technique used for testing the digital

controllers, which combines real-time hardware with software simulations [135]. This
experimental setup consists of dSPACE and RTDS simulators, which are controlled by the
software ControlDesk and RSCAD respectively. The block diagram of the experimental setup
of hardware-in-the-loop using RTDS and the dSPACE control board is shown in Figure 7.19.

141
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

Initial tests have been conducted to realize the time delay and noise content between inputs
and outputs of the RSCAD/RTDS and dSPACE hardware system. The issue of signal latency
has been resolved with the use of a suitable solver in MATLAB/Simulink and with the optimum
simulation time step.

Figure 7.19. Experimental Setup of Hardware-in-the-Loop using RTDS and dSPACE

Control Board.

7.5.2. Test System and Solar-DSTATCOM Controller Setup in HIL Platform

In this research, to test and validate the Solar-DSTATCOM controller, a test system with a
dynamic load has been considered. The PV system is integrated near to the load referred to as
a point of common coupling (PCC). Figure 7.20 shows the complete design of the test system
and the Solar-DSTATCOM inverter system in RSCAD software. The test system is developed
in a large time step of 50µs simulations, and the PV inverter system is developed in a small
time step of less than 2µs simulations. The test system contains a three phase 11kV line-to-line
supply and a distribution transformer to step down the voltage to 400V. The capacity of the
dynamic load can be varied in runtime using a scheduler or slider component. Depending on

142
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

the requirement, various load profile settings have been considered, which will be explained in
the results section.

Figure 7.20. Test System and Solar-DSTATCOM Modelling in RSCAD Environment.

The PV system consists of PV panels connected as arrays, a boost converter and a DC/AC
inverter with an interfacing LCL filter. Also, to integrate a small time step circuit with a large
time step network, an interfacing transformer has been connected. The input solar irradiation
to the PV array can be varied by using slider/scheduler components in the simulations to
perform several tests. The PV output voltage has been given to the boost converter to get the
preferred DC voltage for conversion. The boost converter is operated with a Maximum Power
Point Tracking (MPPT) system, which generates a duty cycle for the gate switch. An
incremental conductance based MPPT algorithm has been utilized to get the maximum power.
The controller for the boost converter is developed in RSCAD software. The control system
for the DC/AC inverter is developed in Simulink software for controller testing.

143
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

7.5.2.1 Solar-DSTATCOM Control Setup in MATLAB/Simulink:

The concept of utilizing the PV system as a virtual DSTATCOM can eliminate several
power quality issues. The detailed description of the Solar-DSTATCOM controller can be
found in Section 7.3. The proposed new control consists of a DC bus voltage controller, PCC
voltage controller, load reactive power and harmonic compensation controller and hysteresis
controller as mentioned earlier. These controllers require a measured DC bus voltage, PCC
voltages, load currents and the inverter output currents. The inputs from RSCAD are transferred
to dSPACE hardware using GTAO card, which converts the digital signals into analog signals
and are scaled down with a scaling factor to produce lower voltage levels below +/-10V. By
using BNC cables, the inputs from the GTAI card are connected to the CLP1103 connector
panel. To simplify the analog scaling, the signals are converted to per unit values. A scaling
factor of 5 is applied for the DC voltages of 0.75p.u (or 0.75kV) to get 0.75V at the output
channels of the GTAI card. Similarly, a scaling factor of 5 is selected for the PCC per unit
voltages to obtain a 1V peak sinusoidal waveform. A scaling factor of 300 is applied to the
load currents, which are in Amps.

The block diagram of the Solar-DSTATCOM controller in the MATLAB software

environment with Simulink/dSPACE block sets is shown in Figures. 7.21 and 7.22. The DC
bus voltage controller is used to regulate the DC voltage constant across the DC link capacitor.
In the circuit, two ADC channels are used to convert analog signals to digital and by default,
the signals need to be multiplied by a factor of 10. Then, the signals are scaled up by a scaling
factor to get them back to the DC bus voltage range.

Figure 7.21(b) shows the PCC voltage controller, which is used to maintain the PCC
voltages within the steady state limits. The sensed PCC voltage signals are obtained through
MUXADC channel. In the figure, Ipc_A, Ipc_B and Ipc_C indicate the generated reference
quadrature currents for independent phase voltage regulation.

144
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

Figure 7.21. (a) DC Bus Voltage Controller and (b) PCC Voltage Controllers.

The block diagram for load reactive power and current harmonic compensation is shown in
Figure 7.22. These controllers require measured three-phase load currents, which will be
obtained using the MUXADC channel. As mentioned earlier, the currents will be scaled up to
the actual current magnitudes. Finally, the reference currents will be scaled down to achieve a
low power range and then converted to analog signals by using 3 DAC channels. Note that
before the DAC block, the signal is divided by a factor of 10 by default. Further, the analog
output signals from the CLP1103 connector panel are connected to the GTAI card with the help
of BNC cables. Further, the GTAI card converts the analog signals to digital as indicated in
Figure 7.19.

145
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

Figure 7.22. Load Reactive Power and Current Harmonic Compensation Block.

The developed control system is converted to custom C code and is loaded into ControlDesk
software for the real-time simulation. Finally, the reference currents are compared with the

146
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

inverter output currents, the differences are given to the hysteresis controllers to obtain the
firing pulses for the inverter switching. The total height of the hysteresis band is chosen to be
5% of the actual PV current. From RSCAD software library, the inverter with a built-in
hysteresis current controller functionality has been considered for the simulation.

7.5.3. Results and Discussions

The performance of the Solar-DSTATCOM controller has been investigated in three

different case studies with regards to both PV profile and load profile variations. For this, a
100kWp PV inverter system has been considered for the simulation. To enable the reactive
power capability, the size of the PV systems has been increased by 5% so that around 33% of
the reactive power can be produced during rated conditions. The analysis of results from real-
time HIL simulations in both RSCAD/RTDS and dSPACE controlDesk are presented.

7.5.3.1. Case study – I: Performance of Solar-DSTATCOM for Various Daytime

PV Profile Variations

The assessment has been performed for a clear sunny day and a typical cloudy day. The
real-time solar irradiation data from weather sensors with a resolution of 1min have been
collected and are assumed as seconds’ data. For daytime operation, the measured solar
irradiation levels from 6am to 6pm have been considered and are further reduced to get a total
of 360s data points. By using a scheduler component in RTDS, the real-time simulation has
been automated for a continuous 360s time. For every second, the input to the PV array changes
and produces new output values. It is to be noted that the load on the test system is fixed and
is operated at 0.24MW with 0.9 P.F (lagging).

A Sunny Day Operation:

Figure 7.23 shows the time versus solar irradiation profile of a sunny day. Simulations have
been performed and the corresponding measured active (P) and reactive (Q) powers at the
Solar-DSTATCOM output and grid interconnection point are captured as shown in Figure 7.24.
It can be seen that solar irradiation has a linear relationship with the PV output power and has
an inverse relationship with the reactive power. With the PV power injection, a reduction in
the grid supply P and Q trends can be observed. During the rated condition, the Solar-
DSTATCOM has produced an active power (PInv) of 100kW, reactive power (QInv) of 33kVAr
which corresponds to a P.F of 0.95. During low solar irradiation levels, the generated PInv has
a decreased trend. In contrast, the QInv is increased to its maximum and reached a value of about
94kVAr.

147
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

1100 Time vs Solar Irradiation of Sunny Day

Solar Irradiation (W/m2)

900

700

500

300

100
15
30
45
60
75
90
0

105
120
135
150
165
180
195
210
225
240
255
270
285
300
315
330
345
360
Time (Sec)

Figure 7.23. Time versus Solar Irradiation of a Sunny Day.

Figure 7.24. Measured Active and Reactive Power Profiles at Solar-DSTATCOM Output and
Grid Supply during a Sunny Day.

148
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

Figure 7.25. Measured PCC RMS Voltages and Solar-DSTATCOM Output Currents during
a Sunny Day.

Due to the plot data point limitations in RSCAD software, the waveforms of the Solar-
DSTATCOM inverter output currents and PCC voltages were recorded in dSPACE
ControlDesk software, which are illustrated in Figure 7.25. It is to be noticed that the PV system
capacity is completely utilized for the whole time period and it injected rated current into the
grid. The PV inverter acts as a DSTATCOM and has compensated the load reactive power
requirement and enhanced the voltage magnitudes at the PCC node. It is to be seen that with
reactive power injection, the voltage magnitudes are increased from 0.9p.u to 1.01p.u, which
are within the limits specified by the ANSI standard [94]. However, the voltage rise due to PV
power injection is observed to be minimal during midday due to the operation of the PCC
voltage controller.

Figure 7.26. DC Bus Voltage Profile during a Sunny Day.

The measured DC bus voltage is shown in Figure 7.26. It can be seen that the DC bus voltage
is almost constant and dropped due to solar irradiation levels below 100W/m2. During early
morning and evening periods, the DC bus voltage controller has been applied to maintain the
DC voltage constant.

Furthermore, due to the involvement of hardware and several signal conversions, the
controller inputs/outputs might be influenced by noise and the execution speed, which can

149
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

affect the results of simulations. To verify the controller operations, the output of various
controller blocks have been captured for 360sec time using the dSPACE ControlDesk software.
Figure 7.27 shows the controller trends during a sunny day operation.

Figure 7.27. Solar-DSTATCOM Controller Various Output Parameters during a Sunny Day
Operation.

Note that the PV inverter breaker was closed a few seconds after the RSCAD simulation
started. Immediately, the data recording process in dSPACE has been started, this delay can be
seen in the waveforms. In the figure, Idc_loss represents the loss current required to maintain the
DC bus voltage constant. Then, the PCC voltage regulators generate the currents required to
regulate the PCC voltages to be within the limits. The currents IdA_total and IqA_total denote the
total real and reactive components of injected inverter currents. These generated signals are
following the input dynamic trends and are noise free. It is noticed that the generated reference
currents are changed with respect to the load currents.

B Cloudy Day Operation:

Similarly, a typical cloudy day solar irradiation profile has been considered. Figure 7.28
shows the corresponding time versus solar irradiation plot. It can be seen that the solar

150
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

irradiation pattern contains high, average and low spikes, which could represent a worst case
condition. The Solar-DSTATCOM output and grid supply active and reactive power profiles
are illustrated in Figure 7.29. It can be seen that the response of Solar-DSTATCOM is faster
for sudden variations in the solar irradiation due to cloud transients. The trends of measured
PV and grid powers have a similar tendency as with the above scenario.

1100 Time Vs Solar Irradiation of Cloudy day

Solar Irradiation (W/m2)

900
700
500
300
100
15
30
45
60
75
90
0

105
120
135
150
165
180
195
210
225
240
255
270
285
300
315
330
345
360
Time (Sec)

Figure 7.28. Time versus Solar Irradiation of a Cloudy Day.

Figure 7.29. Measured Active and Reactive Power Profiles at Solar-DSTATCOM Output and
Grid Supply during a Cloudy Day.

Figure 7.30 shows the measured PCC voltage and Solar-DSTATCOM output currents. It
can be seen that the voltage fluctuations due to cloud transients are appeared to be lower. The
voltage magnitudes are observed to be between 0.98 and 1.0p.u, which are within the steady
state voltage limits.

151
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

Figure 7.30. Measured PCC RMS Voltages and Solar-DSTATCOM Output Currents during
a Cloudy Day.

Accordingly, the measured voltage across the DC link capacitor is shown in Figure 7.31. It
can be observed that the reduction in voltage is due to maximum PV power drop by heavy
cloud transients. At this period, the DC bus voltage controller extracts a small amount of active
power from the grid and charges the DC capacitor in few cycles to retain the preferred operating
voltage as can be seen in Figure 7.31. It is noticed that the Solar-DSTATCOM output currents
are following the same trend as the DC bus voltage.

Figure 7.31. DC bus Voltage Profile during a Cloudy Day.

Similarly, the performance of the Solar-DSTATCOM controller with regards to input

variations has been investigated. The resultant controller output currents are captured and
illustrated in Figure 7.32. It can be seen that the dynamic response of inverter controllers during
cloud transient periods are faster and ripple free to perform the preferred control action.

152
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

Figure 7.32. Solar-DSTATCOM Controller Various Output Parameters during a Cloudy

Day Operation.

7.5.3.2. Case study – II: Performance of Solar-DSTATCOM during Night-time

Operation

In addition, the operation of the Solar-DSTATCOM controller is demonstrated during the

night-time. Similar to the first case study, the load on the test system is considered constant
and it draws 0.24MW power with 0.9 P.F (lagging). It is noted that the proposed Solar-
DSTATCOM operates as a DSTATCOM during the night-time, and provides reactive power
support to the grid. Figure 7.33 shows the various output parameters of the Solar-DSTATCOM
and the results of grid active and reactive power variations. As explained above, when there is
no solar power the solar-DSTATCOM consumes a small amount of active power to regulate
the DC bus voltage. It is observed that when the transient period is over, the DC link voltage
is reached 0.75kV. As a result, the grid real power injection has been increased to 0.2462 MW
to compensate for the inverter losses. At this time, the Solar-DSTATCOM controller has
injected a reactive power of about 93.3kVAr. Hence, the power factor of the grid has improved
and maintained close to unity. It is confirmed that the proposed Solar-DSTATCOM controller
could be of benefit to both PV customers and utilities.

153
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

Figure 7.33. Various Operating Parameters of the Solar-DSTATCOM and Grid during the
Night-time.

7.5.3.3. Case Study - III: Performance of Solar-DSTATCOM at Various PV

Power Levels and Load Profile Variations

In this case study, the influence of load active and reactive power variations on the operation
of the Solar-DSTATCOM controller has been verified with controller HIL simulation. The
study has been carried out for two load scenarios by changing the active and reactive power
demands on the grid to obtain various P.F levels. Also, the variation in power output from the
Solar-DSTATCOM is considered for a comprehensive assessment.

A. PV Operational Impact during Various Grid P.F’s- If Only Qload Changes:

In the first scenario, the study was performed with variable load reactive power and fixed
active power. Consequently, the grid P.F’s are varied from 0.7 to 0.9 in steps to achieve realistic
conditions. In real time, this condition generally occurs while the large induction machines start
up, which increases the reactive power requirement. Moreover, the Solar-DSTATCOM output
powers are considered to be changing. The dynamic load scenarios are created for a total of
74sec, in which the load reactive power and PV power/solar irradiation magnitudes are varied
for every 3sec and 15sec respectively. Figure 7.34 shows the active and reactive power inputs
to the test system, which are scheduled in the RTDS software using scheduler components.

154
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

0.3 Only Qload Profile Variation for Different Load P.F

Active and Reactive Power

(MWand MVAr)
0.2

0.1

0
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73
Time (Sec)
Ppv Pload Q load

Figure 7.34. Load Active and Reactive Power and PV Power Input Profiles for First
Scenario.

A simulation was performed and the active and reactive powers at the Solar-DSTATCOM
output, grid supply and load were measured and are as shown in Figure 7.35. The
corresponding calculated P.Fs of PV systems, grid and load are indicated in Figure 7.36. It is
noticed that when the load reactive power demand increased, in particular when grid P.F
decreased from 0.9 to 0.7, the Solar-DSTATCOM controller was operated in P.F correction
mode to regulate the grid P.F. It can be seen that during rated conditions, the PV system has
sacrificed only 5% of the active power to compensate the grid reactive power demand. Hence,
at worst condition, the grid P.F has been increased from 0.7 to 0.92. Further, during low PV
power periods, the inverter has supplied maximum reactive power and the P.F was close to 1.
Figure 7.37 shows the measured grid voltage at the PCC node. It is noticed that when the
reactive power demand increases the system voltage will decreases and vice versa. In low P.F
periods, the PCC voltage is regulated to lie above 0.96p.u, and during 0.9 P.F periods the
voltage is regulated to be about 1.01p.u, these are within the limits specified by the relevant
ANSI standard.

155
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

Figure 7.35. Active and Reactive Powers at Solar-DSTATCOM Output, Grid and Load
during the First Load Scenario.

Figure 7.36. Calculated P.F of Solar-DSTATCOM, Grid, and Load.

156
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

Figure 7.37. Measured PCC RMS Voltages during the First Load Scenario.

It is observed that during the reduced PV power levels and at various load reactive power
levels, the Solar-DSTATCOM is operated as a DSTATCOM and injected both real and reactive
powers to achieve grid operational necessities. The network operating in lower P.F may result
in increased maintenance costs for the utilities, therefore these costs will usually be transferred
to the customers in the form of higher tariff charges. To overcome such problems, the Solar-
DSTATCOM can be a potential solution for both PV owners and utilities.

Similar to the previous case study, to verify the impact of load variations on the Solar-
DSTATCOM controller operation, various outputs of the controller blocks are captured using
RTDS and dSPACE software. Figure 7.38 illustrates the DC bus voltage, three-phase inverter
output currents and various controller output parameters. It can be seen that the trend of total
reactive current IqA_total has an inverse relationship with the load reactive power variations for
compensation. The output of the PCC regulator currents are increased during low P.F periods
to improve the grid voltage and are decreased at high P.F periods to compensate for the voltage
rise issue. It can be seen that the dynamics of controllers for rapid load changes are stable and
provided a quick response.

157
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

Figure 7.38. Solar-DSTATCOM Controller Various Output Parameters during the First
Load Scenario.

158
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

B. PV Operational Impact during Various Grid P.F’s- If Both Pload and Qload Changes:

In this scenario, the evaluation was carried out for both load active and reactive power
variation impacts on the Solar-DSTATCOM controller performance. The active and reactive
powers were varied such that the load P.F on the network varied between 0.7 and 0.9. Similar
to the previous scenario, the simulation scenarios were created for 74sec time and also
considered the variations in the PV output powers as displayed in Figure 7.39.

Both Pload and Qload Profiles Variation for Different Load P.F's
0.3
Active & Reactive Power
(MW& MVAr)

0.2

0.1

0
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73
Time (Sec)
Ppv Pload Q load

Figure 7.39. Load Active and Reactive Power and PV Power Input Profiles for Second
Scenario.
Similar to the first scenario, real-time simulations were performed and the active and
reactive powers at Solar-DSTATCOM, grid and load are presented in Figure 7.40. Also, the
corresponding calculated P.F’s are plotted in Figure 7.41. In comparison with the above
scenario, the Solar-DSTATCOM has not reduced its active power injection to the grid during
the worst case conditions. However, the controller supplied real and reactive powers in
proportion during rated conditions to maintain the grid P.F over 0.82. This kind of control
functionality enabled the priority for active power injection, which will be the primary concern
of PV system owners. Besides, the PV inverter operates at below its maximum rated current
capacity over 95% of the time due to the unavailability of peak solar irradiation. During these
times, the excess PV inverter capacity has been utilized for reactive power compensation,
which eventually is profitable for the utilities. Further, the grid P.F has been improved for
reduced PV output power periods, which is due to the reactive power support from the Solar-
DSTATCOM. The grid P.F during this period lies between 0.92 and unity.

159
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

Figure 7.40. Active and Reactive Powers at Solar-DSTATCOM Output, Grid and Load
during the Second Load Scenario.

Figure 7.41. Calculated P.F of Solar-DSTATCOM, Grid, and Load.

160
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

Figure 7.42. Measured PCC RMS Voltages during the Second Load Scenario.

Moreover, the grid PCC voltages are captured and presented in Figure 7.42. It is observed
that the voltage magnitudes due to load active and reactive power variations are within the
1.015p.u to 0.96p.u range. The Solar-DSTATCOM voltage controller is applied to improve the
voltage to above 0.96p.u in peak load reactive power demand periods. Subsequently, various
parameters of the Solar-DSTATCOM controllers are plotted in Figure 7.43 for analysis.
Similar to the first scenario, the controller dynamics subjected to input load and PV power
variations are faster in response and stable.

161
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

Figure 7.43. Solar-DSTATCOM Controller Various Output Parameters during the Second
Load Scenario.

7.5.3.4. Case Study-IV: Harmonic Distortion Analysis

In addition, Harmonic analysis has been done using Fast Fourier Transformation (FFT) to
obtain harmonic distortions. The inverter current total harmonic distortion (THD) is measured
to be very low about 1.7% and voltage THD is about 0.73% as indicated in Figure 7.44, which
are satisfying the limits specified by IEEE 929 [106] and IEEE 519 [53] PV interconnection
standards. Moreover, the measured percentage of individual harmonic distortions of voltage
and current with respect to their fundamental during rated conditions are shown Figure 7.45.
Results show that 5th order harmonic is intensified both in voltage and current, which is
satisfying the individual harmonic limits. The next prominent current harmonic is 3rd order,
which fluctuated between 0.4 and 1.1%. It is observed that the current magnitudes of 7th order
harmonic is below 0.4% and the harmonic orders from 9th to 31st order harmonics laid lower
than 0.3%, which are very low according to the IEEE standards. Correspondingly, individual
voltage harmonic orders have a similar tendency with the current harmonic orders. Results
show that the measured harmonic orders from 3rd to 31st are lower than 0.28% and are satisfying
the IEEE limits. Furthermore, for the reduced power levels the Solar-DSTATCOM can produce

162
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

a constant THD profile similar to the PSCAD simulation results as indicated in Figures 7.16
and 7.17.

Figure 7.44. Voltage and Current THDs of Solar-DSTATCOM.

Figure 7.45. Solar-DSTATCOM Individual Harmonic Distortion of Voltage and Current.

7.6. Summary

This chapter presented a novel solution to the poor power qualities caused by PV systems
penetration in the distribution network. Regarding that, a new concept of utilizing the PV
inverter as a DSTATCOM, which is a custom power device was proposed. A new control
strategy to alleviate problems in the unbalanced distribution network, which provides

163
 Chapter 7. Solar-DSTATCOM: A New Solution to Poor Power Quality

unbalanced voltage regulation, reactive power and harmonic compensations was introduced
and is known as Solar-DSTATCOM. In this chapter, the operation and modelling of the Solar-
DSTATCOM control strategy in PSCAD software environment was first presented. Primary
assessment was performed on the IEEE bus unbalanced distribution network. The power
quality analysis was presented in three case studies. In the first two case studies, the voltage
quality issues such as voltage rise, unbalance and sag/swell discussed in Chapter 4 were
compensated for with the Solar-DSTATCOM operation. The voltage quality impact of Solar-
DSTATCOM daytime and night-time operations were also presented. Finally, the harmonic
analysis was performed and the voltage and current harmonic distortion trends were captured
for various solar irradiation conditions.

Furthermore in this research, to test and validate the Solar-DSTATCOM controller

performance in a real-time environment, controller Hardware-in-the-loop simulations were
performed. The controller HIL platform consists of various hardware and software tools which
involve RSCAD/RTDS, dSPACE DS1103 control hardware along with associated
ControlDesk and MATLAB/Simulink software. The implementation of the proposed Solar-
DSTATCOM control algorithm in MATLAB/Simulink with dSPACE RTI library components
was presented. The input/output signals from/to RSCAD were converted and transferred
through GTAI and GTAO cards. Also, the test system along with a dynamic load considered
for this study was modelled in RSCAD. In this research, a comprehensive testing of Solar-
DSTATCOM was performed in four different case studies with various realistic grid and PV
operating scenarios. The dynamic performance of the Solar-DSTATCOM controller during
sunny and cloudy days was investigated. In addition, the night-time operation of the Solar-
DSTATCOM was verified. Moreover, various load profile variation interactions on the Solar-
DSTATCOM performance was presented. This analysis also considered the Solar-
DSTATCOM influence at various output power levels. The trends of various controller outputs
and grid operating parameters were captured.

Results from both studies show that the Solar-DSTATCOM performance in unbalanced and
balanced network condition has enhanced the grid power quality. Finally, the harmonic
emission analysis during rated conditions was presented. The following chapter provides the
summary, conclusions and contributions of this thesis.

164
 Chapter 8. Conclusions and Future Works

CHAPTER 8

CONCLUSIONS AND FUTURE WORKS

8.1. Conclusions

Solar PV systems are believed to be one of the most reliable alternative resources to produce
electrical energy as it relies on one of the earth natural resources, sunlight. With the
proliferation of distributed solar PV systems penetration in existing low voltage (LV)
distribution networks, power quality becomes a major area of interest. Although PV systems
can have positive impacts on the distribution network to meet local energy demand, they could
also have negative impacts. Solar PV systems are sensitive to variations in solar irradiation,
which can cause potential challenges in the LV distribution networks. The literature review
provided in Chapter 2 revealed that these PV systems could initiate a new set of power quality
concerns in terms of voltage quality and harmonics. In addition, traditional distribution
networks are unbalanced in nature due to unbalanced loading and line configurations.
Consequently, it is very important and interesting to examine the penetration impacts of PV
systems in unbalanced network conditions. Therefore, this research has been focussed on the
comprehensive investigation of various power quality issues influenced by high PV penetration
and also proposes a new compensation method for their mitigation.

In the first part of the thesis, various voltage quality issues such as voltage rise, unbalance,
sag/swell and voltage fluctuation/flicker emissions were thoroughly investigated in the
unbalanced distribution network with PV systems penetration. Secondly, harmonic emission

165
 Chapter 8. Conclusions and Future Works

issues from PV inverter technologies and their impact on the distribution network were
systematically analysed. Furthermore, excitation of harmonic resonance issues associated with
PV system dynamic operations were emphasized. After exploring the potential issues, a novel
solution was proposed to overcome these power quality issues in the perspective of high level
of localised PV penetrations.

In order to study the power quality impacts of PV systems, it is essential that accurate
simulation models of PV systems, various kinds of load units and distribution networks are
required for accomplishing the dedicated studies, which were presented in Chapter 3. This
chapter also provided the background theory of each component of the grid connected PV
systems and their models in PSCAD software. In addition, software and hardware tools used
throughout this research were presented. The contributions of the thesis along with the main
conclusions drawn from individual research works [5-11, 137] are highlighted as below.

In Chapter 4, the assessment of voltage quality issues with PV penetration were provided.
Firstly, the investigation was performed through PSCAD simulations on an IEEE-13 bus
unbalanced distribution network. For this analysis, a developed dynamic model of a PV system
was considered to reflect a realistic PV operating scenario. The analysis was carried out in two
case studies to investigate the severity of voltage quality issues.

 In the first case study, simulations were performed to assess voltage rise and unbalance in
three scenarios, which included lumped PV penetration at the beginning of the feeder,
dispersed PV penetration at multiple buses close to the loads, and with increased
penetration by single phase PV systems respectively. In the first two scenarios, the network
was considered to have 25% PV penetration and this penetration is increased by 5% in the
third scenario. Results show that the voltage rise due to distributed PV integration is higher
than compared to the lumped PV penetration at the beginning of the feeder. This impact is
even higher with increased PV penetration as in the third scenario. Results also
demonstrate that the PV power injection provided the voltage support in the highly loaded
phases. Furthermore, the rise in voltages even with the three-phase PV systems are
different in each phase due to unbalanced loading and has higher values on the lightly
loaded phase. In contrast, voltage unbalance factor is higher in the case of the first scenario
as compared to the second scenario. Furthermore, this voltage unbalance has increased in
the third scenario with random distributed single-phase PV systems integration.

166
 Chapter 8. Conclusions and Future Works

 In the second case study, fault analysis was carried out to estimate the influence of PV
presence and absence in the network during three types of faults. Results showed that
during the fault condition, the unbalanced voltage sags are slightly compensated by the
contribution of PV power. However, the voltage swell is attained higher values compared
to without PV scenario.

Furthermore in Chapter 4, to analyse the voltage quality issues caused by PV system

penetration in a real-time distribution network, field measurement based investigations were
performed on the University of Queensland (UQ), St Lucia and Gatton Campus PV sites. The
PV power generation strongly depends on the solar irradiation and temperature, which impacts
the grid voltages. In this research, rapid variation in solar irradiation levels were considered to
determine the effect on voltage variation issues.

 Firstly, the investigation was performed on the UQ St Lucia PV site, which had a grid
connected PV system with an installed capacity of 1.5 MW. Field measurements were
conducted through the installation of Fluke Power Quality Analysers at one of the UQ
buildings, which had a 339 kW PV system. For the analysis, the power quality
measurements at the PV inverter output level, low voltage and high voltage (HV) sides of
the corresponding distribution feeder were recorded for various operating days. Results
show that the network voltage magnitudes are increased with respect to the PV power
injection. In the UQ St Lucia network, the voltage rise at the integration point is actually
compensated by the voltage drop caused by loads and the voltage levels are well within
the limits. Furthermore, the voltage fluctuations and flicker caused by variations in solar
irradiation were analysed. Results show that during the measurement period, the voltage
flickers were lower and satisfied the AS/NZS standard. Nevertheless, the severity of
voltage flickers may vary depending on the ramp rate of cloud transients and the connected
PV systems size. For this reason, the measurement data of 1-year at the LV and HV sides
of the PCC was analysed. Measurements of some typical days exhibited higher flicker
emission levels. In order to quantify the probability of the flicker levels in a year, a data
clustering technique was applied for further assessment. It is confirmed that the voltage
flickers are insignificant in the UQ 339kW PV site and are well within the limits according
to the AS/NZS standard.

 Secondly, the measurement based investigation was performed on a large scale of a

3.3MW PV system, which is located at the UQ Gatton campus. Note that this distribution
network had a Step Voltage Regulator (SVR) for voltage regulation, which was connected

167
 Chapter 8. Conclusions and Future Works

at the 11kV side. The power quality measurements were recorded at the PV inverter output
and the grid interconnection point of the 11kV side for a three months’ period with 1
second resolution. Similar voltage variation trends with respect to PV power generation
are observed and the voltage magnitudes are within the limits specified by the AS/NZS
standard and the Energex agreement. However, the voltage rise impact on the 11kV
network with PV Power injection is slightly reduced by the SVR operation. However, the
voltage fluctuations are higher and exceeded the limits most of the time, which might be
due to the false operation of the SVR. In addition, these voltage fluctuations are increased
during heavy cloud periods. Furthermore, results of the flicker emissions at the 11kV side
of the PCC through clustering technique showed that the voltage flickers are appeared to
be under control for 70% of the time and satisfying the AS/NZS standard limits. From the
results, it is perceived that the flicker levels are more noticeable in the 3.3MW PV system
as compared with the 1.5MW PV system, which depend on the size, location of the PV
system and connected network operating condition.

In Chapter 5, the harmonic emission issues from PV systems and their penetration impacts
on the distribution network were analysed. In this chapter, firstly the characterisation of
harmonics and their interaction issues associated with multiple PV systems interconnection
was studied with simulations and compared with field measurements. This analysis considered
the variations in solar irradiation conditions. The simulation assessment was supplemented
with simulations in two case studies, which reveals the PV system operation in the case of with
and without the presence of background harmonic distortions in the supply network. To inject
harmonic distortions into the network, developed detailed power electronic based nonlinear
load models were connected to the IEEE-13 bus network. Moreover, a validation study was
performed through field investigations conducted at the UQ 1.5 MW PV site. The measurement
data for various operating days at the inverter level were recorded through Fluke power quality
analysers.

 Measurement results indicate that the characteristics of harmonics have identical trends
with simulations. The individual voltage harmonics had a linear relationship with the solar
power generation. The percentage of voltage distortion is increased due to the increased
voltage amplitude with the addition of PV inverters and also the influence of augmentation
of individual voltage harmonic magnitudes. The individual voltage and current harmonics
during rated power levels are within the limits prescribed by the IEEE and AS/NZ
standards. However, the current THD has increased due to lower fundamental component

168
 Chapter 8. Conclusions and Future Works

of current during low power levels. Nonetheless, the standards only provide limiting values
of the harmonic limits during rated conditions. Besides, in the case of multiple PV
integrations the current harmonic levels appeared to be minimized due to increase in
fundament current, however the individual harmonic magnitudes increased with every
parallel PV penetration.

Moreover in Chapter 5, the impact of increased harmonic distortion levels with high PV
penetration on a LV distribution network was assessed. A thorough harmonic behaviour
analysis was first performed on the IEEE-13 bus network with high PV penetration.
Furthermore, an evaluation study was conducted at the UQ, St Lucia PV network to validate
simulation results.

 In the simulation study, to simulate a real-time network scenario, several types of

probabilistic nonlinear loads were incorporated in the network for harmonics injection. In
addition, to consider the effect of a PV inverter control scheme on the network harmonic
levels, two types of PV inverter models were considered. The investigation was carried out
in three different case studies, namely PV systems integration at a single node in particular
with and without background harmonic distortions in the supply and finally PV penetration
at multiple nodes with supply distortions. Simulation results show that the harmonics
generated from PV systems are multiplied in nature due to an increase in individual voltage
magnitudes. These magnitudes are also dependent on the concentration of harmonics from
nonlinear loads. The voltage THDs are higher at the PCC and after the PCC node or end
feeder node. Results confirmed that the distortion levels increased with multiple
installations at the same node compared to dispersed integration.

 The current harmonic distortions are higher at upstream nodes of the network because of
PV power injection causing a supply current reduction. However, the current THDs are
exceeding the IEEE limit at the PCC node and the nodes after PCC with increased PV
penetration, where the magnitude of network fundamental current is same. Further, even
though three-phase PV systems inject balanced currents into the IEEE 13 bus network, the
harmonic distortion levels are different in each phase due to an unbalanced customer
loading level. Furthermore, depending on the type of control strategy, the harmonic
generations from PV systems and their interaction impacts in the network are dissimilar.

 In addition, the harmonic field measurements were taken by PQ analysers at the LV and
HV sides of the UQ network for different operating conditions. Measured results show that

169
 Chapter 8. Conclusions and Future Works

the THD trends follow similar trends as with simulation results. The THD of voltage and
currents increases with PV power injection. The levels of current harmonics are
significantly increased at the distribution transformer during rated PV power output and
full load condition. The corresponding k-factor measurements show the effect of excessive
harmonic flow in the transformer, which could lead to overloading and heating.

 The harmonic effect is shown to be under control, however increased PV system

penetration in the distribution network can cause adverse effects. The capacity of the
existing feeder lines and the transformers are not able to withstand such conditions. So, the
utilities are required to pay attention to preventive measures.

In addition, the analysis of harmonic resonances accompanied with the PV inverter systems
was presented in Chapter 6. Harmonic analysis was carried out through simulations on the
IEEE-13 bus unbalanced distribution network in two different case studies, namely with and
without harmonic distortions in the supply network. In each case study, the dynamic operations
of two PV inverter control topologies for variations in solar irradiation were considered to
characterise the harmonic resonance modes in the network. Also, a frequency versus
impedance scan has been presented with respect to each resonance condition. Furthermore, a
validation study has been carried out with the field measurement data obtained at the University
of Queensland PV system.

 Simulation results show that the harmonic resonance occurrence is not only reliant on the
type of inverter control topology, filters and the number of PV inverter systems but also on
the network configuration and type of load. The integration of PV systems in the
distribution network modified the magnitude of grid impedance and shifted its resonance
frequency. Analysis revealed that the generated harmonics PV systems during rated
condition have satisfied the standard limits. However, the dynamic controller operation of
PV systems due to cloud transients injected high harmonic components into the network,
which may have excited the resonance conditions in the network. The amplification of
voltage and current distortions during harmonic resonance exceeded the limits according
to the IEEE 519 standards.

 Moreover, to verify and validate the simulation results in the case of a real distribution
network, the harmonic measurement data of 1-year collected from the UQ PV system has
been analysed. Results show that the probability of harmonic resonance events are higher
in the real network and possess severe voltage and current THDs on the 11kV line and are

170
 Chapter 8. Conclusions and Future Works

confirming the simulation results. At present, during rated PV output power levels, the
distortions in the network are found to be under control. However, the regular incidence of
resonance conditions can lead to malfunction of sensitive devices, overheating and
premature ageing of the distribution transformer. Therefore, the utilities are required to take
measures of harmonic compensation devices to overcome these issues.

In Chapter 7, to address these power quality issues a new approach of adopting the PV
inverter as a Solar-DSTATCOM with the help of a new control strategy was proposed. The
proposed control topology can independently regulate the phase voltage and provide harmonic
and reactive power compensation in each phase during the whole day. At first, the investigation
was performed through PSCAD simulations and further testing and a validation study was
performed in the real-time controller Hardware-in-the-loop platform.

 For the simulation study, an IEEE-13 bus unbalanced distribution network was considered.
In this work, a static variation of solar insolation condition was applied to verify the
effectiveness and robustness of the proposed control scheme in improving power quality of
the unbalanced distribution network. This study emphasized the compensation of various
power quality issues, namely voltage rise, unbalance, sag/swell and harmonics. Regarding
this, simulations were performed for three different case studies. Simulation results showed
that the voltage rise and voltage unbalance levels due to PV power injection were
compensated. Also, results from fault analysis demonstrated that the unbalanced voltage
sags and swells during unsymmetrical and symmetrical grid faults were greatly
ameliorated. Results from the harmonic analysis showed that individual voltage and current
harmonics were lower and were well within the Australian/New Zealand and IEEE
standards. The current THD has possessed a constant profile for varied solar irradiation
levels from daytime to night-time operation. Therefore, this proposed novel control
topology enabled independent phase reactive power compensation and voltage regulation
at the PCC and showed superior performance in an unbalanced system.

 In addition, this chapter presented a real-time controller Hardware-in-the-loop simulation

of the proposed Solar-DSTATCOM using RSCAD/RTDS, dSPACE DS1103 control
hardware and the MATLAB/Simulink software platform. For this, a test system along with
PV inverter systems were modelled in RSCAD and the proposed controller was
implemented in Simulink using dSPACE RTI library components. A comprehensive
testing of the Solar-DSTATCOM controller was performed in various case studies with
various realistic operating scenarios. The impacts of solar irradiation variations of sunny

171
 Chapter 8. Conclusions and Future Works

and cloudy days on the Solar-DSTATCOM daytime operation were investigated. Also, the
night-time operation of Solar-DSTATCOM has been presented. Furthermore, the Solar-
DSTATCOM performance has been verified for the load profile variations. The trends of
various controller outputs and grid operating parameters were captured. Results indicated
that the Solar-DSTATCOM has provided the reactive power support during both day and
night-time along with active power injection during the daytime. Moreover, the Solar-
DSTATCOM has improved grid performance during load operations by providing reactive
power compensation thereby improving the grid power factor and PCC voltage regulation
similar to the DSTATCOM. The controller responses with respect to the rapid input
variations are faster. The harmonics emissions are very low and satisfy the IEEE standard
limits.

Implementing this type of advanced functionality in the PV inverter systems will enable
additional revenue for PV owners and grid operators in addition to the feed-in-tariffs for active
power generation. Multiple installations of this type of PV inverter will better manage the LV
network by controlling the grid in a distributed manner and the response would be faster
compared to a single point control.

8.2. Summary of Contributions

This research has investigated power quality challenges associated with high penetration of
PV systems in LV distribution network and proposed a practical solution to overcome these
issues. The main contributions of this thesis are summarised as follows:

Investigation of Voltage Quality Impacts:

• Various voltage quality issues in a LV unbalanced distribution network with high PV

penetration were investigated thoroughly in PSCAD simulations.

• For that purpose, a comprehensive dynamic PV source, different PV inverter models and
IEEE-13 bus test systems was developed.

• Practical field investigation of voltage quality challenges in the real grid connected PV sites
located at the University of Queensland St. Lucia and Gatton Campuses, were performed.

• The impact of short-term solar irradiation variations on voltage flicker emissions were
evaluated.

• Using the actual measurement data of several months, the probability of voltage flicker
severity was estimated.

172
 Chapter 8. Conclusions and Future Works

Assessment of Harmonic Impact of PV Systems:

• The harmonic emission characteristics of PV systems and their aggregation issues during
real operating conditions were assessed through PSCAD simulations and validated with
actual measurements obtained through practical connection of power quality analysers.

• Harmonic impact of high penetration PV systems on the unbalanced distribution network

was comprehensively analysed through PSCAD simulations and validated with field
measurements.

• For the simulation study, IEEE-13 bus network with comprehensive linear, nonlinear
harmonic producing load models and different PV inverter systems were developed to
demonstrate real network operating scenarios.

• Developed online JAVA programs to extract the real-time harmonic measurements at

various voltage levels of UQ PV site.

• The phenomenon of harmonic resonance on the distribution network with PV system

penetrations and their dynamic operational interactions were also analysed through
simulations and compared with the actual harmonic measurements.

Practical Solution to Mitigate Power Quality Issues:

• A novel practical solution of adopting PV inverter as a virtual DSTATCOM was proposed

named as Solar-DSTATCOM.

• A new control algorithm for Solar-DSTATCOM to compensate power quality issues in

unbalanced and balanced distribution networks was designed and implemented.

• The performance of developed new control algorithm was verified initially in PSCAD
software environment and detailed investigations were performed on IEEE-13 bus
unbalanced network to compensate poor power quality issues during the daytime and night-
time.

• Comprehensive testing and validation of Solar-DSTATCOM controller was performed in

Real-time Hardware-in-the-Loop simulation platform. This practical setup combines Real-
time Digital simulator and dSPACE control hardware board.

• Various studies were conducted to demonstrate the performance of Solar-DSTATCOM

controller for daytime, night-time and dynamic load demand scenarios.

173
 Chapter 8. Conclusions and Future Works

8.3. Future Works

Recommendations for future research are as follows.

• Further research to investigate the Solar-DSTATCOM operation in mitigating the voltage

quality issues in various scenarios through RSCAD real-time simulations is recommended.

• Analysis of field measurement data monitoring at the distribution transformer levels are
required to better manage the LV network and to maximize PV penetration.

• More research is required to understand and mitigate the harmonic resonance issues
accompanied with PV systems penetration.

• In future, Energy storage systems will play a vital role in the LV networks. The collective
impact of PV and energy storage technology need to be studied. Also, the proposals like
community energy storage and distributed energy storage systems installation in the
network could be interesting.

174
 References

REFERENCES:
[1] "IEA World Energy Outlook," International Energy Agency, 2015.
[2] "Renewables 2015 Global Status Report, [Online] Available on
http://www.ren21.net/status-of-renewables/global-status-
report/http://www.ren21.net/wp-
content/uploads/2015/07/GSR2015_KeyFindings_lowres.pdf," 2015.
[3] "EPIA: Global Market Outlook- For Photovoltaics 2014-2018," Eur. Photovolt. Ind.
Assoc, 2014.
[4] "AEMO 2012b rooftop PV information; available online at
http://www.aemo.com.au/Electricity/~/.../Rooftop_PV_Information_Paper.ashx."
[5] A. Chidurala, T. K. Saha, and N. Mithulananthan, "Harmonic characterization of grid
connected PV systems and validation with field measurements," in Power & Energy
Society General Meeting, 2015 IEEE, Denver, CO, USA, 2015, pp. 1-5.
[6] A. Chidurala, T. Saha, and N. Mithulananthan, "Field investigation of voltage quality
issues in distribution network with PV penetration," in Power and Energy Engineering
Conference (APPEEC), 2015 IEEE PES Asia-Pacific, 2015, pp. 1-5, Brisbane,
Australia.
[7] A. Chidurala, T. K. Saha, and N. Mithulananthan, "Power quality enhancement in
unbalanced distribution network using Solar-DSTATCOM," in Power Engineering
Conference (AUPEC), 2013 Australasian Universities, 2013, pp. 1-6, Hobart,
Tasmania.
[8] A. Chidurala, T. K. Saha, and N. Mithulananthan, "Harmonic impact of high
penetration photovoltaic system on unbalanced distribution networks – learning from
an urban photovoltaic network," 'Harmonic impact of high penetration photovoltaic
system on unbalanced distribution networks – learning from an urban photovoltaic
network', IET Renewable Power Generation, Volume 10, Issue 4, April 2016, pp. 485
– 494 DOI: 10.1049/iet-rpg.2015.0188.
[9] A. Chidurala, T. K. Saha, N. Mithulananthan, and R. C. Bansal, "Harmonic emissions
in grid connected PV systems: A case study on a large scale rooftop PV site," in PES
General Meeting | Conference & Exposition, 2014 IEEE, 2014, pp. 1-5, National
Harbor, MD, USA.
[10] A. Chidurala, Tapan Kumar Saha, and Mithulananthan N., "Real-Time Controller
Hardware-in-the-Loop Simulations of Solar-DSTATCOM," Submitted to IEEE
transaction on Industrial Electronics, Submitted on 15 March, 2016.
[11] A. Chidurala, Tapan Kumar Saha, and Mithulananthan N., "Harmonic Resonance on
the Distribution Network with PV Penetration and Field Measurements Validation,"
Elsevier International Journal of Electrical Power & Energy Systems journal,
Submitted on 3 July 2015.

175
 References

[12] M. Munsell. (2016) U.S. Solar Market Insight: Q3 2015 available on

www.greentechmedia.com/research/subscription/u.s.-solar-market-insight.
[13] T. Buckley, "India's Electricity-Sector Transformation: available online on
http://ieefa.org/wp-content/uploads/2015/08/IEEFA-Indian-Electricity-Sector-
Transformation-August-2015.pdf," Institute for Energy Economics and Financial
Analysis (IEEFA), August 2015.
[14] "The Renewable Energy Target (RET) scheme | Department of the Environment:
available online on https://www.environment.gov.au/climate-change/renewable-
energy-target-scheme," Accessed on Feb 2016.
[15] Australian PV Institute live report on market analyses, available online at http://pv-
map.apvi.org.au/analyses.
[16] "Energex Distribution Annual Planning Report," August 2015.
[17] R. Tonkoski, D. Turcotte, and T. H. M. El-Fouly, "Impact of High PV Penetration on
Voltage Profiles in Residential Neighborhoods," Sustainable Energy, IEEE
Transactions on, vol. 3, pp. 518-527, 2012.
[18] "PVupscale. Publication review on the impacts of PV Distributed Generation and
Electricity networks [cited 03/11/2011]: available from
http://www.pvupscale.org/IMG/pdf/D41_final.pdf," 2007.
[19] G. M. Comfort. R, Mansoor. A, Barker. P, Short. T, Sundaram. A, "Power Quality
Impact of Distributed Generation Effect on Steady State Voltage Regulation," PQA
conference, 2001.
[20] M. E. Baran, H. Hooshyar, Z. Shen, J. Gajda, and K. M. M. Huq, "Impact of high
penetration residential PV systems on distribution systems," in Power and Energy
Society General Meeting, 2011 IEEE, 2011, pp. 1-5, San Diego, CA.
[21] Y. Ruifeng and T. K. Saha, "Voltage Variation Sensitivity Analysis for Unbalanced
Distribution Networks Due to Photovoltaic Power Fluctuations," Power Systems, IEEE
Transactions on, vol. 27, pp. 1078-1089, 2012.
[22] "Addressing Grid interconnection issues with variable renewable energy sources," Asia
Pacific Economic corporation, IT Power Australia, http://www.itpau.com.au/wp-
content/uploads/2012/08/ITP-APEC-Web-Summary.pdfDec 2010 2010.
[23] M. S. ElNozahy and M. M. A. Salama, "Technical impacts of grid-connected
photovoltaic systems on electrical networks—A review," Journal of Renewable and
Sustainable Energy, vol. 5, p. 032702, 2013.
[24] Patsalides, M., Evagorou, D., Makrides, G., Achillides, Z., Georghiou, G.E., Stavrou,
A., Efthymiou, V., Zinsser, B., Schmitt, W. and Werner, J.H. (2007) The effect of solar
irradiance on the power quality behaviour of grid connected photovoltaic systems. In,
International Conference on Renewable energies and Power Quality, ICREPQ, La
Coruña, ES, 25 - 27 Mar 2007.

176
 References

[25] Y. Ruifeng and T. K. Saha, "Investigation of voltage sensitivities to photovoltaic power

fluctuations in unbalanced distribution networks," in Power and Energy Society
General Meeting, 2012 IEEE, 2012, pp. 1-7, San Diego, CA.
[26] Y. Ruifeng and T. K. Saha, "Investigation of voltage variations in unbalanced
distribution systems due to high photovoltaic penetrations," in Power and Energy
Society General Meeting, 2011 IEEE, 2011, pp. 1-8, San Diego, CA.
[27] A. Woyte, V. Van Thong, R. Belmans, and J. Nijs, "Voltage fluctuations on distribution
level introduced by photovoltaic systems," Energy Conversion, IEEE Transactions on,
vol. 21, pp. 202-209, 2006.
[28] J. Marcos, L. Marroyo, E. Lorenzo, D. Alvira, and E. Izco, "Power output fluctuations
in large scale pv plants: One year observations with one second resolution and a derived
analytic model," Progress in Photovoltaics: Research and Applications, vol. 19, pp.
218-227, 2011.
[29] E. M. Stewart, T. P. Aukai, S. D. J. MacPherson, B. P. Quach, D. Nakafuji, and R.
Davis, "A realistic irradiance-based voltage flicker analysis of PV applied to Hawaii
distribution feeders," in Power and Energy Society General Meeting, 2012 IEEE, 2012,
pp. 1-7, San Diego, CA.
[30] F. Shahnia, A. Ghosh, G. Ledwich, and F. Zare, "Voltage Unbalance reduction in low
voltage distribution networks with rooftop PVs," in Universities Power Engineering
Conference (AUPEC), 2010 20th Australasian, 2010, pp. 1-5, Christchurch.
[31] F. Shahnia, Majumder, R., Ghosh, A.,Ledwich, G., Zare, F., "Voltage unbalance
analysis in residential low voltage distribution networks with rooftop PVs," ELECTRIC
POWER SYSTEMS RESEARCH, vol. 81, pp. 1805-1814, 2011.
[32] A. von Jouanne and B. Banerjee, "Assessment of voltage unbalance," Power Delivery,
IEEE Transactions on, vol. 16, pp. 782-790, 2001.
[33] A. Guinane, G. M. Shafiullah, A. M. T. Oo, and B. E. Harvey, "Voltage fluctuations in
PV penetration on SWER networks -A case study for regional Australia," in Power and
Energy Society General Meeting, 2012 IEEE, 2012, pp. 1-6, San Diego, CA.
[34] J. D. Mondol, Y. Yohanis, M. Smyth, and B. Norton, "Long term performance analysis
of a grid connected photovoltaic system in Northern Ireland," Energy Conversion and
Management, vol. 47, pp. 2925-2947, 11// 2006.
[35] D. L. Garrett and S. M. Jeter, "A photovoltaic voltage regulation impact investigation
technique. I. Model development," Energy Conversion, IEEE Transactions on, vol. 4,
pp. 47-53, 1989.
[36] "IEEE Std 1250-1995: IEEE guide for service to equipment sensitive to momentary
voltage disturbances," ed, 1995.
[37] M. I. A. Bakar, "Assessments for the impact of harmonic current distortion of non linear
load in power system harmonics," in Transmission and Distribution Conference and
Exposition: Latin America, 2008 IEEE/PES, 2008, pp. 1-6.

177
 References

[38] S. S. T. R.Kameswara Rao, "Harmonic modelling of residential and commercial loads

with unified power quality conditioner," International Journal of Scientific and
Engineering Research, pp. 770-743, 2012.
[39] Y. Jing, C. Liang, and C. Shuangyan, "Modeling of Home Appliances for Power
Distribution System Harmonic Analysis," Power Delivery, IEEE Transactions on, vol.
25, pp. 3147-3155, 2010.
[40] M. G. Villalva, J. R. Gazoli, and E. R. Filho, "Comprehensive Approach to Modeling
and Simulation of Photovoltaic Arrays," Power Electronics, IEEE Transactions on, vol.
24, pp. 1198-1208, 2009.
[41] T. Esram and P. L. Chapman, "Comparison of Photovoltaic Array Maximum Power
Point Tracking Techniques," Energy Conversion, IEEE Transactions on, vol. 22, pp.
439-449, 2007.
[42] J. Schlabbach, "Harmonic current emission of photovoltaic installations under system
conditions," in Electricity Market, 2008. EEM 2008. 5th International Conference on
European, 2008, pp. 1-5, Lisboa.
[43] G. Chicco, J. Schlabbach, and F. Spertino, "Experimental assessment of the waveform
distortion in grid-connected photovoltaic installations," Solar Energy, vol. 83, pp. 1026-
1039, 7// 2009.
[44] M. Sidrach-de-Cardona and J. Carretero, "Analysis of the current total harmonic
distortion for different single-phase inverters for grid-connected pv-systems," Solar
Energy Materials and Solar Cells, vol. 87, pp. 529-540, 5// 2005.
[45] J. H. R. Enslin and P. J. M. Heskes, "Harmonic interaction between a large number of
distributed power inverters and the distribution network," Power Electronics, IEEE
Transactions on, vol. 19, pp. 1586-1593, 2004.
[46] A. Bhowmik, A. Maitra, S. M. Halpin, and J. E. Schatz, "Determination of allowable
penetration levels of distributed generation resources based on harmonic limit
considerations," Power Delivery, IEEE Transactions on, vol. 18, pp. 619-624, 2003.
[47] F. J. F.J. Ruiz-Rodriguez Modeling of photovoltaic generator as a random variable:
Nova Publisher, 2014 in blue, : ‘Photovoltaics: synthesis, applications and emerging
technologies’ (Nova Publisher, 2014), pp. 91–120.
[48] F.-J. Ruiz-Rodriguez, Hernandez, J.-C., Jurado, F., "Harmonic modelling of PV
systems for probabilistic harmonic load flow studies," Int. J. Circuit Theory
Application, pp. 1541–1565 2015.
[49] R. Torquato, F. C. L. Trindade, and W. Freitas, "Analysis of the harmonic distortion
impact of photovoltaic generation in Brazilian residential networks," in Harmonics and
Quality of Power (ICHQP), 2014 IEEE 16th International Conference on, 2014, pp.
239-243, Bucharest.
[50] A. A. Latheef, "Harmonic impact of photovoltaic inverter systems on low and medium
voltage distribution systems," MEng, University of Wollongong, 2006.

178
 References

[51] H. Farooq, Z. Chengke, M. E. Farrag, and M. Ejaz, "Investigating the Impacts of

Distributed Generation on an Electrical Distribution System Already Stressed by Non-
Linear Domestic Loads," in Power and Energy Engineering Conference (APPEEC),
2012 Asia-Pacific, 2012, pp. 1-4, Shanghai.
[52] "IEEE 519-1995: IEEE recommended practices and requirements for harmonic control
in electrical power system," ed IEEE, 1995.
[53] "IEEE Standard 519-2014: EEE recommended practices and requirements for
harmonic control in electrical power systems," ed: IEEE, 2014.
[54] A. Elmoudi, M. Lehtonen, and H. Nordman, "Effect of harmonics on transformers loss
of life," in Electrical Insulation, 2006. Conference Record of the 2006 IEEE
International Symposium on, 2006, pp. 408-411, toronto, Ont.
[55] D. Martin, S. Goodwin, O. Krause, and T. Saha, "The effect of PV on transformer
ageing: University of Queensland's experience," in Power Engineering Conference
(AUPEC), 2014 Australasian Universities, 2014, pp. 1-6, Perth, WA, Australia.
[56] J. Desmet, C. Debruyne, J. Vanalme, and L. Vandevelde, "Power injection by
distributed generation and the influence of harmonic load conditions," in Power and
Energy Society General Meeting, 2010 IEEE, 2010, pp. 1-6, Minneapolis, MN.
[57] Modeling and Control of Sustainable Power Systems. Springer: Green Energy and
Technology, 2012.
[58] S. J. Lewis, "Analysis and management of the impacts of a high penetration of
photovoltaic systems in an electricity distribution network," in Innovative Smart Grid
Technologies Asia (ISGT), 2011 IEEE PES, 2011, pp. 1-7, Perth, WA, Australia.
[59] T. F. o. H. M. a. Simulation, "Modelling and simulation of the propagation of harmonics
in electric power networks. I. concepts, models, and simulation techniques," IEEE
Trans. Power Deliv.ery, vol. 11, pp. 452–465 1996.
[60] S. A. Ali, "Impact of Personal Computers and Compact Fluorescent Lamps on the
Power Quality," American Academic & Scholarly Research Journal, vol. 5, March
2013.
[61] J. T. E. Caamano, H. D. Moor, et al., "State-of-the-art on dispersed PV power
generation: Publications review on the impacts of PV distributed generation and
electricity networks," EIE/05/171/SI2.420208 23rd June 2007.
[62] B. G. S. Cobben, and H. Laukamp, "WP4 – Deliverable 4.3 - Impact of Photovoltaic
Generation on Power Quality in Urban Areas with High PV Population,"
EIE/05/171/SI2.4202082008.
[63] N. R. W. Jos Arrillaga, Power Systems harmonics: John Wiley & Sons Ltd, 2003.
[64] J. H. R. Enslin, W. T. J. Hulshorst, A. M. S. Atmadji, P. J. M. Heskes, A. Kotsopoulos,
J. F. G. Cobben, et al., "Harmonic interaction between large numbers of photovoltaic
inverters and the distribution network," in Power Tech Conference Proceedings, 2003
IEEE Bologna, 23-26 June 2003, pp 6, Vol.3.

179
 References

[65] Z. H. Y. C. Y. Wang, G. Sinha, and X. M. Yuan, "Output filter design for a grid
connected three-phase filter," in 34th IEEE Annual Power Electron. Spec. Conf.
(PESC’03), 15-19 June 2003, pp. 779-784.
[66] X. Wilsun, H. Zhenyu, C. Yu, and W. Haizhen, "Harmonic resonance mode analysis,"
Power Delivery, IEEE Transactions on, vol. 20, pp. 1182-1190, 2005.
[67] M. F. M. R. C. Dugan, S. Santoso, H. W. Beaty, Electrical Power Systems Quality,
Second Edition ed.
[68] V. Smith, "Small Scale Domestic Rooftop Solar Photovoltaic Systems," Endeavour
Energy Power Quality & Reliability Centre, University of Wollongong: NSW,
Australia2011.
[69] A. Al-Sabounchi, J. Gow, M. Al-Akaidi, and H. Al-Thani, "Minimizing line energy
loss of radial distribution feeder with a PV Distributed Generation unit avoiding reverse
power flow," in Electric Power and Energy Conversion Systems (EPECS), 2011 2nd
International Conference on, 2011, pp. 1-6, Sharjah.
[70] A. Saidian, M. Heidari, and D. Mirabbasi, "Improvement of voltage unbalance and
voltage sag in radial distribution systems using DG," in Industrial Electronics and
Applications (ICIEA), 2010 the 5th IEEE Conference on, 2010, pp. 835-839, Taichung.
[71] C. Gonzalez, J. Geuns, S. Weckx, T. Wijnhoven, P. Vingerhoets, T. De Rybel, et al.,
"LV distribution network feeders in Belgium and power quality issues due to increasing
PV penetration levels," in Innovative Smart Grid Technologies (ISGT Europe), 2012
3rd IEEE PES International Conference and Exhibition on, 2012, pp. 1-8, Berlin.
[72] K. Turitsyn, x030C, P. ulc, S. Backhaus, and M. Chertkov, "Distributed control of
reactive power flow in a radial distribution circuit with high photovoltaic penetration,"
in Power and Energy Society General Meeting, 2010 IEEE, 2010, pp. 1-6, Minneapolis,
MN.
[73] R. Yan and T. K. Saha, "Voltage Variation Sensitivity Analysis for Unbalanced
Distribution Networks Due to Photovoltaic Power Fluctuations," IEEE Transactions on
Power Systems, vol. 27, pp. 1078-1089, 2012.
[74] C. L. Masters, "Voltage rise: the big issue when connecting embedded generation to
long 11kV overhead lines," Power Engineering Journal, vol. 16, pp. 5-12, 2002.
[75] A. B. a. V. L. C. Oates, "Tap changer for distributed power," presented at the European
Conference on Power Electronics and Apllications, Aalborg, 2007.
[76] M. J. E. Alam, K. M. Muttaqi, and D. Sutanto, "Distributed energy storage for
mitigation of voltage-rise impact caused by rooftop solar PV," in Power and Energy
Society General Meeting, 2012 IEEE, 2012, pp. 1-8, San Diego, CA.
[77] K. H. Chua, L. Yun Seng, P. Taylor, S. Morris, and W. Jianhui, "Energy Storage System
for Mitigating Voltage Unbalance on Low-Voltage Networks With Photovoltaic
Systems," Power Delivery, IEEE Transactions on, vol. 27, pp. 1783-1790, 2012.

180
 References

[78] K. H. Chua, J. Wong, Y. S. Lim, P. Taylor, E. Morris, and S. Morris, "Mitigation of

Voltage Unbalance in Low Voltage Distribution Network with High Level of
Photovoltaic System," Energy Procedia, vol. 12, pp. 495-501, // 2011.
[79] L. Xiaohu, A. Aichhorn, L. Liming, and L. Hui, "Coordinated Control of Distributed
Energy Storage System With Tap Changer Transformers for Voltage Rise Mitigation
Under High Photovoltaic Penetration," Smart Grid, IEEE Transactions on, vol. 3, pp.
897-906, 2012.
[80] F. Shahnia, A. Ghosh, G. Ledwich, and F. Zare, "Voltage correction in low voltage
distribution networks with rooftop PVs using custom power devices," in IECON 2011
- 37th Annual Conference on IEEE Industrial Electronics Society, 2011, pp. 991-996.
[81] G. R. F. Q. Mafra, W. Uturbey, and B. J. C. Filho, "Analysis of the operation of a D-
STATCOM in unbalanced distribution systems under voltage disturbances," in
Transmission and Distribution Conference and Exposition: Latin America (T&D-LA),
2010 IEEE/PES, 2010, pp. 629-634, Sao Paulo.
[82] A. R. V. B. S.Siva Kesava Reddy, S. Suman, K.Divya Manasa, ", International Journal
of Engineering Research and Applications (IJERA) ISSN: 2248-9622, 2011., "Reactive
Power Compensation Using DSTATCOM," International Journal of Engineering
Research and Applications (IJERA) 2011.
[83] P. D. R. K. K.Madhale, "Design And Simulation Studies of D-STATCOM For Voltage
Sag, Swell Mitigation," International Journal of Power System Operation and Energy
Management, vol. 2, 2008.
[84] D. N. S.F. Torabi, Y. Shayestehfard, "Compensation of Sags And Swells Voltage Using
Dynamic Voltage Restorer (DVR) During Single Line to Ground and Three-Phase
Faults," International Journal on Technical and Physical Problems of Engineering
(IJTPE), vol. 4, pp. 126-132, Sep 2012 2012.
[85] Y. Hen-Geul, D. F. Gayme, and S. H. Low, "Adaptive VAR Control for Distribution
Circuits With Photovoltaic Generators," Power Systems, IEEE Transactions on, vol.
27, pp. 1656-1663, 2012.
[86] P. J. a. D. C. Aliprantis, "Distributed Volt/Var control by PV inverters," IEEE
Transaction on Power System, vol. 28, pp. 3429-3439, 2013.
[87] J. W. Smith, W. Sunderman, R. Dugan, and B. Seal, "Smart inverter volt/var control
functions for high penetration of PV on distribution systems," in Power Systems
Conference and Exposition (PSCE), 2011 IEEE/PES, 2011, pp. 1-6, Phoenix, AZ.
[88] R. K. Varma, S. A. Rahman, and T. Vanderheide, "New Control of PV Solar Farm as
STATCOM (PV-STATCOM) for Increasing Grid Power Transmission Limits During
Night and Day," Power Delivery, IEEE Transactions on, vol. 30, pp. 755-763, 2015.
[89] R. K. Varma, V. Khadkikar, and R. Seethapathy, "Nighttime Application of PV Solar
Farm as STATCOM to Regulate Grid Voltage," Energy Conversion, IEEE
Transactions on, vol. 24, pp. 983-985, 2009.

181
 References

[90] R. K. Varma, B. Das, I. Axente, and T. Vanderheide, "Optimal 24-hr utilization of a PV

solar system as STATCOM (PV-STATCOM) in a distribution network," in Power and
Energy Society General Meeting, 2011 IEEE, 2011, pp. 1-8, San Diego, CA.
[91] "Australia report on smart inverters: Good news for utilities.
http://ecogeneration.com.au/news/smart_inverters_good_news_for_utilities/55950."
[92] NREL, "Advanced inverter functions to support high levels of distributed solar- policy
and regulatory considerations. http://www.nrel.gov/docs/fy15osti/62612.pdf",
NREL/BR-6A20-62612, November 2014.
[93] S. K. B. Ambarnath Banerji , Bhim Singh, "DSTATCOM Control Algorithms: A
Review," International Journal of Power Electronics and Drive System (IJPEDS), vol.
2, pp. 285-296, September 2012 2012.
[94] A. N. S. Institute, "ANSI C84.1-2006, 2006," in American National Standard for
Electric Power Systems and Equipment- Voltage Ratings (60Hertz), ed: National
Electrical Manufacturers Association (NEMA).
[95] " AS 60038-2000: Australian Standard voltages," ed, 2000.
[96] "Australian Standard AS/NZS 61000.3.100:2011, Electromagnetic compatibility
(EMC) – Limits – Steady state voltage limits in public electricity systems," ed, 2011.
[97] "IEEE Std 1159-2009 (Revision of IEEE Std 1159-1995): IEEE Recommended
Practice for Monitoring Electric Power Quality," ed, 2009, pp. 1-81.
[98] "Australian Standard AS/NZS 61000.4.30:2012, Electromagnetic compatibility (EMC)
– Testing and measurement techniques-Power quality measurement methods
(IEC61000-4-30:Ed.2.0(2008))," ed, 2012.
[99] A. Singh, G. Singh, and R. Mitra, "Some Observations on Definitions of Voltage
Unbalance," in Power Symposium, 2007. NAPS '07. 39th North American, 2007, pp.
473-479, Las Cruces, NM.
[100] G. C. Cornfield, "Definition and measurement of voltage flicker [in lighting]," in
Electronics in Power Systems Measurement, IEE Colloquium on, 1988, pp. 4/1-4/4,
London.
[101] "Australian Standards, AS/NZS 4377:1996, Flicker meter – Evaluation of flicker
severity," ed, 1996.
[102] "IEEE Std 1453-2004: IEEE Recommended Practice for Measurement and Limits of
Voltage Fluctuations and Associated Light Flicker on AC Power Systems," (Adoption
of CEI/IEC 61000-4-15:1997+A1:2003), pp. 1-0, 2005.
[103] "AS/NZS 61000.3.3:2012, Electromagnetic compatibility (EMC) - Limits - Limitation
of voltage changes, voltage fluctuations and flicker in public low-voltage supply
systems, for equipment with rated current =16 A per phase and not subject to
conditional connection," ed, 2012.

182
 References

[104] "AS/NZS 61000.3.7:2001, Electromagnetic compatibility (EMC) Part 3.7: Limits -

Assessment of emission limits for fluctuating loads in MV and HV power systems," ed,
2001.
[105] "IEEE Std 1547-2003: IEEE Standard for Interconnecting Distributed Resources with
Electric Power Systems," ed, 2003.
[106] "IEEE 929-2000: IEEE recommended practice for utility interface of residential and
intermediate photovoltaic systems," in IEEE Stand. Coordinating Committee 21,
Photovoltaic, ed, 2000.
[107] "Australian Standard AS4777.2, Grid connection of energy systems via inverters -
Inverter requirements.," 2005.
[108] "AS/NZS 4777.2:2015, Grid connection of energy systems via inverters Part 2: Inverter
requirements ", ed, 2015.
[109] "IEC 61000-3-2: Electromagnetic compatibility (EMC) Part 3-2: Limits for harmonic
current emission. (equipment input current < 16 A per phase)," ed: International
Electrotechnical Commission (IEC), 2012.
[110] S. J. B. Ong and C. Yeong Jia, "An overview of international harmonics standards and
guidelines (IEEE, IEC, EN, ER and STC) for low voltage system," in Power
Engineering Conference, 2007. IPEC 2007. International, 2007, pp. 602-607,
Singapore.
[111] K. H. Matter, "Comparison for Maximum Power Point Tracing Algorithms for
Photovoltaic System in Purpose of Developing an Efficient System," Master of
Engineering, Faculty of Graduate Studies Department of Electrical Engineering The
Islamic University of Gaza, 2014.
[112] B. Subudhi and R. Pradhan, "A Comparative Study on Maximum Power Point Tracking
Techniques for Photovoltaic Power Systems," IEEE Transactions on Sustainable
Energy, vol. 4, pp. 89-98, 2013.
[113] I. V. Banu and M. Istrate, "Modeling of maximum power point tracking algorithm for
photovoltaic systems," in Electrical and Power Engineering (EPE), 2012 International
Conference and Exposition on, 2012, pp. 953-957, Iasi.
[114] M. S. E. Muljadi, and V. Gevorgian, "PSCAD Modules Representing PV Generator,"
NREL/TP-5500-5818, August 2013.
[115] "IEEE 13 Node Test Feeder," ed: IEEE Power Engineering Society, 2009.
[116] A. Hughes, Electric Motors and Drives: Fundamentals, Types and Applications, 2005.
[117] Power Systems Computer Aided Design (PSCAD/EMTDC) online available at
https://hvdc.ca/pscad/, Acessed in 2016.
[118] "RTDS Technologies, RTDS Manual Set, accessed in 2015.."
[119] RTDS Technologies website: https://www.rtds.com/applications/control-systems-
testing/, accessed in 2015.

183
 References

[120] dSPACE DS1103 PPC control hardware board manual:

https://www.dspace.com/en/inc/home/products/hw/singbord/ppcconbo.cfm, accessed
in 2015.
[121] "AS/NZS 61000. 4.15: 2012, Electromagnetic compatibility (EMC) - Part 4.15: Testing
and measurement techniques – Flicker meter - Functional and design specifications,"
ed, 2012.
[122] Fluke Power Analyzer 434 data sheet, available online at
http://www.transcat.com/media/pdf/Fluke_434.pdf.
[123] Fluke Power Analyzer 1735 data sheet, available online at
http://www.electrorent.com/products/search/pdf/FLU-1735.pdf.
[124] "IEEE C57.110–1998: IEEE recommended practice for establishing transformer
capability when supplying non-sinusoidal load currents," ed, 1986.
[125] C. N. M. Ho, V. S. P. Cheung, and H. S. H. Chung, "Constant-frequency hysteresis
current control of grid-connected VSI without bandwidth control," in Energy
Conversion Congress and Exposition, 2009. ECCE 2009. IEEE, 2009, pp. 2949-2956,
San Jose, CA.
[126] S. Civanlar and J. J. Grainger, "Volt/Var Control on Distribution Systems with Lateral
Branches Using Shunt Capacitors and Voltage Regulators Part II: The Solution
Method," Power Apparatus and Systems, IEEE Transactions on, vol. PAS-104, pp.
3284-3290, 1985.
[127] M. L. J. Marcos, Lorenzo E, Garcia M, , "Power output fluctuations in large-scale PV
plants," in Proceedings of International conference on renewable energies and power
quality, Spain, 2012.
[128] V. Khadkikar and A. Chandra, "Control of single-phase UPQC in synchronous d-q
reference frame," in Harmonics and Quality of Power (ICHQP), 2012 IEEE 15th
International Conference on, 2012, pp. 378-383, Hong Kong.
[129] A. V. Timbus, R. Teodorescu, F. Blaabjerg, M. Liserre, and A. Dell'Aquila,
"Independent synchronization and control of three phase grid converters," in Power
Electronics, Electrical Drives, Automation and Motion, 2006. SPEEDAM 2006.
International Symposium on, 2006, pp. 1246-1251, Taormina.
[130] J. Mahseredjian, V. Dinavahi, and J. A. Martinez, "Simulation Tools for
Electromagnetic Transients in Power Systems: Overview and Challenges," Power
Delivery, IEEE Transactions on, vol. 24, pp. 1657-1669, 2009.
[131] B. Lu, X. Wu, H. Figueroa, and A. Monti, "A Low-Cost Real-Time Hardware-in-the-
Loop Testing Approach of Power Electronics Controls," IEEE Transactions on
Industrial Electronics, vol. 54, pp. 919-931, 2007.
[132] Z. Qing, Y. N. , W. K. , F. L. , L. G. , and C. K. , "Hardware-in-the-loop Simulation
Platform of Photovoltaic Grid-Connected System " TELKOMNIKA Indonesian Journal
of Electrical Engineering vol. 12, pp. 2465 - 2473 April 2014.

184
 References

[133] S. Yang, " Novel sensorless generator control and grid fault ride-through strategies for
variable-speed wind turbines and implementation on a new real-time simulation
platform," Graduate Theses and Dissertations, 2010.
[134] H. Li, M. Steurer, K. L. Shi, S. Woodruff, and D. Zhang, "Development of a Unified
Design, Test, and Research Platform for Wind Energy Systems Based on Hardware-in-
the-Loop Real-Time Simulation," IEEE Transactions on Industrial Electronics, vol. 53,
pp. 1144-1151, 2006.
[135] L. Yunwei, D. M. Vilathgamuwa, and L. Poh Chiang, "Design, analysis, and real-time
testing of a controller for multibus microgrid system," Power Electronics, IEEE
Transactions on, vol. 19, pp. 1195-1204, 2004.
[136] MATLAB/Simulink:http://au.mathworks.com/products/simulink, accessed in 2015.
[137] J. E. Alam, R. Yan, T. K. Saha, Annapoorna Chidurala, D Eghbal, "Learning from a 3.275
MW Utility Scale PV Plant Project," in CIGRE, 21, rue d’Artois, F-75008 PARIS 2016.

185
 Appendix

APPENDIX A: IEEE-13 Bus Test System Configuration Data

The IEEE 13 Bus distribution network data sheet can be obtained from [115]. The test
system has seven line configurations which are provided in Table A.1 and A.2. The
corresponding impedance matrices are presented Table A.3. Also, the line length information
is presented in Table A.4.

Table A.1 Overhead Line Configuration Data

Config. Phasing Phase Neutral Spacing

ACSR ACSR ID
601 BACN 556,500 26/7 4/0 6/1 500
602 CABN 4/0 6/1 4/0 6/1 500
603 CBN 1/0 1/0 505
604 ACN 1/0 1/0 505
605 CN 1/0 1/0 510

Table A.2 Underground Line Configuration Data

Config. Phasing Cable Neutral Space ID

606 A B C N 250,000 AA, CN None 515
607 AN 1/0 AA, TS 1/0 Cu 520

Table A.3 Line Impedance Matrices

Config B in micro Siemens per

Impedance Z (R +jX) (ohm/mile)
. mile

𝟎𝟎. 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑 + j𝟏𝟏. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟏𝟏𝟏𝟏𝟏𝟏𝟏𝟏 + j𝟎𝟎. 𝟓𝟓𝟓𝟓𝟓𝟓𝟓𝟓 𝟎𝟎. 𝟏𝟏𝟏𝟏𝟏𝟏𝟏𝟏 + j𝟎𝟎. 𝟒𝟒𝟒𝟒𝟒𝟒𝟒𝟒 𝟔𝟔. 𝟐𝟐𝟐𝟐𝟐𝟐𝟐𝟐 −𝟏𝟏. 𝟗𝟗𝟗𝟗𝟗𝟗𝟗𝟗 −𝟏𝟏. 𝟐𝟐𝟐𝟐𝟐𝟐𝟐𝟐
601 � 𝟎𝟎. 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑 + j 𝟏𝟏. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟏𝟏𝟏𝟏𝟏𝟏𝟏𝟏 + j𝟎𝟎. 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑� 𝑗𝑗 � 𝟓𝟓. 𝟗𝟗𝟗𝟗𝟗𝟗𝟗𝟗 −𝟎𝟎. 𝟕𝟕𝟕𝟕𝟕𝟕𝟕𝟕�
𝟎𝟎. 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑 + j𝟏𝟏. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟓𝟓. 𝟔𝟔𝟔𝟔𝟔𝟔𝟔𝟔

𝟎𝟎. 𝟕𝟕𝟕𝟕𝟕𝟕𝟕𝟕 + j𝟏𝟏. 𝟏𝟏𝟏𝟏𝟏𝟏𝟏𝟏 𝟎𝟎. 𝟏𝟏𝟏𝟏𝟏𝟏𝟏𝟏 + j𝟎𝟎. 𝟒𝟒𝟒𝟒𝟒𝟒𝟒𝟒 𝟎𝟎. 𝟏𝟏𝟏𝟏𝟏𝟏𝟏𝟏 + j𝟎𝟎. 𝟓𝟓𝟓𝟓𝟓𝟓𝟓𝟓 𝟓𝟓. 𝟔𝟔𝟔𝟔𝟔𝟔𝟔𝟔 −𝟏𝟏. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 −𝟏𝟏. 𝟔𝟔𝟔𝟔𝟔𝟔𝟔𝟔
602 � 𝟎𝟎. 𝟕𝟕𝟕𝟕𝟕𝟕𝟕𝟕 + j𝟏𝟏. 𝟏𝟏𝟏𝟏𝟏𝟏𝟏𝟏 𝟎𝟎. 𝟏𝟏𝟏𝟏𝟏𝟏𝟏𝟏 + j𝟎𝟎. 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑 𝑗𝑗 � 𝟓𝟓. 𝟏𝟏𝟏𝟏𝟏𝟏𝟏𝟏 −𝟎𝟎. 𝟔𝟔𝟔𝟔𝟔𝟔𝟔𝟔�
𝟎𝟎. 𝟕𝟕𝟕𝟕𝟕𝟕𝟕𝟕 + j𝟏𝟏. 𝟐𝟐𝟐𝟐𝟐𝟐𝟐𝟐 𝟓𝟓. 𝟒𝟒𝟒𝟒𝟒𝟒𝟒𝟒

𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 + j𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 + j𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 + j𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎
603 � 𝟏𝟏. 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑 + j𝟏𝟏. 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑 𝟎𝟎. 𝟐𝟐𝟐𝟐𝟐𝟐𝟐𝟐 + j𝟎𝟎. 𝟒𝟒𝟒𝟒𝟒𝟒𝟒𝟒� j� 𝟒𝟒. 𝟕𝟕𝟕𝟕𝟕𝟕𝟕𝟕 −𝟎𝟎. 𝟖𝟖𝟖𝟖𝟖𝟖𝟖𝟖�
𝟏𝟏. 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑 + j𝟏𝟏. 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑 𝟒𝟒. 𝟔𝟔𝟔𝟔𝟔𝟔𝟔𝟔

𝟏𝟏. 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑 + j𝟏𝟏. 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 + j𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟐𝟐𝟐𝟐𝟐𝟐𝟐𝟐 + j𝟎𝟎. 𝟒𝟒𝟒𝟒𝟒𝟒𝟒𝟒 𝟒𝟒. 𝟔𝟔𝟔𝟔𝟔𝟔𝟔𝟔 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 −𝟎𝟎. 𝟖𝟖𝟖𝟖𝟖𝟖𝟖𝟖
604 � 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 + j𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 + j𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎� 𝑗𝑗 � 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 �
𝟏𝟏. 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑 + j𝟏𝟏. 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑 𝟒𝟒. 𝟕𝟕𝟕𝟕𝟕𝟕𝟕𝟕

𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 + j𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 + j𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 + j𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎
605 � 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 + j𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 + j𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎� 𝑗𝑗 � 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎�
𝟏𝟏. 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑 + j𝟏𝟏. 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑 𝟒𝟒. 𝟓𝟓𝟓𝟓𝟓𝟓𝟓𝟓

186
 Appendix

𝟎𝟎. 𝟕𝟕𝟕𝟕𝟕𝟕𝟕𝟕 + j𝟎𝟎. 𝟒𝟒𝟒𝟒𝟒𝟒𝟒𝟒 𝟎𝟎. 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑 + j𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟐𝟐𝟐𝟐𝟐𝟐𝟐𝟐 − 𝒋𝒋𝒋𝒋. 𝟎𝟎𝟎𝟎𝟎𝟎𝟑𝟑 𝟗𝟗𝟗𝟗. 𝟖𝟖𝟖𝟖𝟖𝟖𝟖𝟖 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎
606 � 𝟎𝟎. 𝟕𝟕𝟕𝟕𝟕𝟕𝟕𝟕 + j𝟎𝟎. 𝟒𝟒𝟒𝟒𝟒𝟒𝟒𝟒 𝟎𝟎. 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑 + j𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝑗𝑗 � 𝟗𝟗𝟗𝟗. 𝟖𝟖𝟖𝟖𝟖𝟖𝟖𝟖 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 �
𝟎𝟎. 𝟕𝟕𝟕𝟕𝟕𝟕𝟕𝟕 + j𝟎𝟎. 𝟒𝟒𝟒𝟒𝟒𝟒𝟒𝟒 𝟗𝟗𝟗𝟗. 𝟖𝟖𝟖𝟖𝟖𝟖𝟖𝟖

𝟏𝟏. 𝟑𝟑𝟑𝟑𝟑𝟑𝟑𝟑 + j𝟎𝟎. 𝟓𝟓𝟓𝟓𝟓𝟓𝟓𝟓 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 + j𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 + j𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟖𝟖𝟖𝟖. 𝟗𝟗𝟗𝟗𝟗𝟗𝟗𝟗 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎
607 � 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 + j𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 + j𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝑗𝑗 � 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 �
𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 + j𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 𝟎𝟎. . 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎

Table A.4 Line Segment Data

Node A Node B Length(ft.) Config.

632 645 500 603
632 633 500 602
633 634 0 XFM-1
645 646 300 603
650 632 2000 601
684 652 800 607
632 671 2000 601
671 684 300 604
671 680 1000 601
671 692 0 Switch
684 611 300 605
692 675 500 606

The following Table A.5 provides the transformer data. The capacitor bank information and
single phase voltage regulator data are presented in the Tables A.6 and A.7 respectively. The
data of different load types at various nodes are presented in Tables A.8 and A.9.

Table A.5 Transformer Data

kVA kV-high kV-low R-% X-%

Substation: 5,000 115 - D 4.16 Gr. Y 1 8
XFM -1 500 4.16 – Gr.W 0.48 – Gr.W 1.1 2

Table A.6 Capacitor Bank Data

Node Ph-A Ph-B Ph-C

kVAr kVAr kVAr
675 200 200 200
611 100
Total 200 200 300

187
 Appendix

Table A.7 Regulator Data

Regulator ID: 1
Line Segment: 650 - 632
Location: 50
Phases: A - B -C
Connection: 3-Ph,LG
Monitoring Phase: A-B-C
Bandwidth: 2.0 volts
PT Ratio: 20
Primary CT Rating: 700
Compensator Settings: Ph-A Ph-B Ph-C
R - Setting: 3 3 3
X - Setting: 9 9 9
Voltage Level: 122 122 122

Table A.8 Spot Load Data

Node Load Ph-1 Ph-1 Ph-2 Ph-2 Ph-3 Ph-3

Model kW kVAr kW kVAr kW kVAr
634 Y-PQ 160 110 120 90 120 90
645 Y-PQ 0 0 170 125 0 0
646 D-Z 0 0 230 132 0 0
652 Y-Z 128 86 0 0 0 0
671 D-PQ 385 220 385 220 385 220
675 Y-PQ 485 190 68 60 290 212
692 D-I 0 0 0 0 170 151
611 Y-I 0 0 0 0 170 80
TOTAL 1158 606 973 627 1135 753

Table A.9 Distributed Load Data

Node A Node B Load Ph-1 Ph-1 Ph-2 Ph-2 Ph-3 Ph-3

Model kW kVAr kW kVAr kW kVAr
632 671 Y-PQ 17 10 66 38 117 68

188