andrew keane - integration of distributed generation

8/11/2019 Andrew Keane - Integration of Distributed Generation

1/133

Integration of Distributed Generation

by

Andrew Keane

A thesis presented to

The National University of Ireland

in fulfilment of the requirements

for the degree of

Philosophiae Doctor

in the

School of Electrical, Electronic & Mechanical Engineering

University College Dublin

Supervisor of Research & Nominating Professor: Prof. Mark OMalleyHead of School: Dr. David FitzPatrick

January 2007


2/133

Contents

Abstract IV

Acknowledgments VI

Publications Arising VIII

Acronyms IX

List of Figures XI

List of Tables XIII

1 Introduction 11.1 Electric Power Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.1.1 Transmission & Distribution of Electricity . . . . . . . . . . . . . 5

1.2 Distributed Generation. . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

1.3 Focus of Research. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2 Distributed Generation 102.1 Distributed Generation Network Issues . . . . . . . . . . . . . . . . . . . 13

2.1.1 General Planning & Operational Issues . . . . . . . . . . . . . . 132.1.2 Irish Distribution Network. . . . . . . . . . . . . . . . . . . . . . 17

2.2 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

2.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

3 Optimal Capacity 223.1 Optimal Capacity Allocation Methodology. . . . . . . . . . . . . . . . . 24

3.1.1 Objective Function . . . . . . . . . . . . . . . . . . . . . . . . . . 24

3.1.2 Equipment Ratings. . . . . . . . . . . . . . . . . . . . . . . . . . 24

3.1.3 Short Circuit Level . . . . . . . . . . . . . . . . . . . . . . . . . . 25

3.1.4 Short Circuit Ratio. . . . . . . . . . . . . . . . . . . . . . . . . . 26

I


3/133

3.1.5 Voltage Rise Effect . . . . . . . . . . . . . . . . . . . . . . . . . . 27

3.1.6 Energy Resource & Customer Initiatives . . . . . . . . . . . . . . 29

3.2 Test System & Constraint Characteristics . . . . . . . . . . . . . . . . . 31

3.3 Optimal Capacity Results . . . . . . . . . . . . . . . . . . . . . . . . . . 34

3.3.1 Network Sterilisation. . . . . . . . . . . . . . . . . . . . . . . . . 36

3.4 Impact of DG Capacity on Losses. . . . . . . . . . . . . . . . . . . . . . 39

3.4.1 Losses in a Radial System . . . . . . . . . . . . . . . . . . . . . . 40

3.4.2 Losses & Capacity Methodology . . . . . . . . . . . . . . . . . . 40

3.4.3 Test System and Loss Characteristics . . . . . . . . . . . . . . . 42

3.4.4 Optimal Capacity Allocation with Losses . . . . . . . . . . . . . 44

3.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

3.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

4 Energy Harvesting Networks 514.1 Energy Harvesting Networks . . . . . . . . . . . . . . . . . . . . . . . . 52

4.1.1 Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

4.1.2 Siting Restrictions . . . . . . . . . . . . . . . . . . . . . . . . . . 53

4.1.3 Losses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

4.1.4 Distributed Generation Load Factors . . . . . . . . . . . . . . . . 54

4.1.5 Firm & Non Firm Access . . . . . . . . . . . . . . . . . . . . . . 55

4.2 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

4.3 Test System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

4.3.1 System Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

4.3.2 Voltage Sensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . 61

4.3.3 Loss Adjustment Factors . . . . . . . . . . . . . . . . . . . . . . 624.3.4 Effective Load Factors . . . . . . . . . . . . . . . . . . . . . . . . 62

4.3.5 Available Energy Resources . . . . . . . . . . . . . . . . . . . . . 63

4.4 Results & Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

4.4.1 Energy Allocations . . . . . . . . . . . . . . . . . . . . . . . . . . 64

4.4.2 Annual Simulations . . . . . . . . . . . . . . . . . . . . . . . . . 66

4.4.3 Power Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

4.4.4 Allocation Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

4.4.5 Active Control & Dispatch . . . . . . . . . . . . . . . . . . . . . 71

4.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

5 Impact of Connection Policy on Distributed Generation 735.1 Network Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

5.1.1 Annual Simulations . . . . . . . . . . . . . . . . . . . . . . . . . 77

5.1.2 Costs and Benefits . . . . . . . . . . . . . . . . . . . . . . . . . . 78

5.2 Test Systems and Energy Resource . . . . . . . . . . . . . . . . . . . . . 80

5.2.1 Test Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

5.2.2 Energy Resource . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

5.3 System Results & Discussion . . . . . . . . . . . . . . . . . . . . . . . . 82

5.3.1 Energy Allocations . . . . . . . . . . . . . . . . . . . . . . . . . . 82

5.3.2 Wind Generation Development Costs . . . . . . . . . . . . . . . 84

5.3.3 System Costs & Benefits. . . . . . . . . . . . . . . . . . . . . . . 85

II


4/133

5.3.4 Net Benefits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

5.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

6 Minimum Cost Curtailment for Distributed Generation 896.1 Minimum Cost Curtailment . . . . . . . . . . . . . . . . . . . . . . . . . 90

6.1.1 Voltage Sensitivities . . . . . . . . . . . . . . . . . . . . . . . . . 91

6.1.2 Cost of Curtailment . . . . . . . . . . . . . . . . . . . . . . . . . 91

6.1.3 Curtailment Method . . . . . . . . . . . . . . . . . . . . . . . . . 93

6.1.4 Operating Constraints of Plant . . . . . . . . . . . . . . . . . . . 95

6.2 Existing Curtailment Methods . . . . . . . . . . . . . . . . . . . . . . . 96

6.2.1 Minimum Energy Curtailment . . . . . . . . . . . . . . . . . . . 97

6.2.2 Minimum Distance Curtailment. . . . . . . . . . . . . . . . . . . 97

6.2.3 Proportional Curtailment . . . . . . . . . . . . . . . . . . . . . . 98

6.3 Test System & Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

6.4 Results & Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

6.4.1 Fixed Voltage Sensitivity & Market Price . . . . . . . . . . . . . 100

6.4.2 Variable Voltage Sensitivities . . . . . . . . . . . . . . . . . . . . 102

6.4.3 Variable Market Price . . . . . . . . . . . . . . . . . . . . . . . . 103

6.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

7 Conclusion 1067.1 Synopsis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

7.2 Scope for Further Work . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

References 110

Appendix: Publications Arising 117

III


5/133

Abstract

Dwindling fossil fuel resources, increasing fuel prices and the liberalisation of electricity

systems are changing the context in which electric power systems are operated and

regulated. In particular, concerns about security of supply and reliability along with

the integration of new energy resources are presenting a number of new challenges

to system operators. One of the major changes that is being seen is the connection

of significant levels of generation to the distribution system for the first time. This

Distributed Generation (DG) is forcing a reexamination of the manner in which the

distribution system is planned and operated. This thesis presents a number of new

methodologies to facilitate the large scale integration of DG onto distribution systems.

To accommodate this new type of generation the existing distribution network

should be utilised and developed in an optimal manner. A new methodology is devel-

oped to determine the optimal allocation of DG capacity with respect to the technicalconstraints on DG. The methodology is implemented and tested on a section of distri-

bution network. Results are presented demonstrating that the proper placement and

sizing of DG is crucial to the accommodation of increasing levels of DG on distribu-

tion networks. The impact of DG on losses is also analysed and incorporated into the

optimal capacity methodology.

The introduction of DG is also leading to a fundamental change in how distribution

IV


6/133

V

networks are utilised and viewed. Distribution networks are now used as a means

to connect geographically dispersed energy sources to the electricity system, thereby

converting what were originally energy delivery networks, to networks used both for the

delivery and harvesting of energy. A further methodology is presented which maximises

the amount of energy that may be reaped from a given area, while taking account of

the available energy resources, connection costs, losses, voltage management and the

technical constraints. The optimal energy allocation is determined for a sample section

of network, illustrating the implementation of the methodology and the scope for non

firm access to the distribution network.

The increased applications for DG present a significant challenge to the existing

connection policies of distribution network operators. In particular, non firm access to

the network has been proposed as a method to increase the penetration of DG. The

impact of the connection policies arising from non firm access are investigated in detail

here. The Irish system is used as a case study, and with reference to the available energy

resource and network parameters, the costs and benefits of DG are determined under

a number of planning policies. The costs and benefits assessed include connection and

cycling costs along with emissions, capacity value and fuel bill saving. It is shown that

a significant increase in the net benefits of DG is gained if the appropriate connection

policy is utilised from the outset and conversely that significant costs are incurred if ad

hoc policies are employed. Furthermore, it is shown that non firm access has the scope

to facilitate a significant extra amount of DG capacity.

Voltage rise has been identified as a key barrier to further DG capacity. Active

management of the voltage constraint may be possible, leading to a form of constraint

management at distribution level for the first time. Here a novel method is proposed,

which minimises the cost of curtailment. It takes advantage of the dispatchable ca-

pability of certain forms of DG, such as biomass, hydro or landfill gas. There are a

number of well established methods for congestion management on the transmission

network. A number of these are applied to voltage management on the distribution

network and used for comparison with the new minimum cost method. The variability

of voltage sensitivities and market prices is also investigated, with their impact on the

cost of curtailment over the year quantified.


7/133

Acknowledgements

I would like to thank everyone whose help and support contributed to this thesis. There

are a number of people who deserve special mention.

To my supervisor, Prof. Mark OMalley, I must express my most sincere thanks. His

constant guidance, expertise and support was of tremendous help to me and made this

a very rewarding and enjoyable time. A better education in not just power engineering,

but also in the role of the university and academic could not be found.

I must also express my sincere thanks to ESB Networks for the time I spent there in

the early days of my research. In particular, I would like to thank Ivan Codd, Derek

Hynes, Donal Phelan and Tony Walsh for making it such an enjoyable and interesting

time

I would also like to thank Prof. Janusz Bialek and Dr. Qiong Zhou of the University

of Edinburgh for their hospitality during my time in Edinburgh and also for the subse-

quent fruitful collaboration.

I would like to acknowledge the valuable role of the sponsors of the Electricity Research

VI


8/133

VII

Centre; ESB Networks, Eirgrid, ESB Powergen, Airtricity, Viridian, Cylon and CER.

I would also like to thank Sustainable Energy Ireland who through the Irish Research

Council for Science Engineering and Technology have supported my research.

To all the members of Room 157 with whom I shared my time; Tim, Rebecca, Ronan,

Gill, Shane, Garth, Alan, Aidan, Ciara, Emma and Ronan. The tea breaks and general

craic made the days go by much easier. Many thanks to Rose Mary Logue for all the

encouragement and assistance right from day one. Also, many thanks to Daniel Burke

for proof reading this thesis. Finally, I must also pay tribute to Eleanor Denny, not

least for proof reading this thesis, but also for having the patience and good humour

to put up with an eejit such as myself sitting beside her for a couple of years.

To my friends and family, especially my parents whose absolute support of my educa-

tion and endeavours has been a constant for as long as I can remember.

Finally, to Orla, the endless encouragement and support, even from afar, was of tremen-

dous help to me. I owe a lot more than a PhD to you. Thank you for everything!


9/133

Publications Arising

Journal Publications:

1. Keane, A., Zhou, Q., Denny, E., Bialek, J., and OMalley, M., Minimum CostCurtailment for Distributed Generation Voltage ManagementIEEE Trans. PowerSyst. (in review), 2007.

2. Keane, A., Denny, E. and OMalley, M., Quantifying the Impact of ConnectionPolicy on Distributed Generation, IEEE Trans. Energy Conv., vol. 22, No. 1,pp. 189-196, 2007.

3. Keane, A. and OMalley, M., Optimal Utilisation of Distribution Networks forEnergy Harvesting, IEEE Trans. Power Syst., vol. 22, No. 1, pp. 467-475,2007.

4. Keane, A. and OMalley, M., Optimal Allocation of Embedded Generation onDistribution networks, IEEE Trans. Power Syst.,vol. 20, No. 3, pp. 1640-1646,2005.

Conference Publications:

1. Keane, A. and OMalley, M., Impact of Distributed Generation Capacity onLosses,IEEE PES General Meeting, Montreal, 2006.

2. Keane, A. and OMalley, M., Optimal Distributed Generation Plant Mix withNovel Loss Adjustment Factors, IEEE PES General Meeting, Montreal, 2006.

3. Keane, A. and OMalley, M., Impact of Distribution Network Constraints onDistributed Generation Capacity, 40th International Universities Power Engi-neering Conference, Cork, 2005.

4. Keane, A. and OMalley, M., Optimal Allocation of Embedded Generation on theIrish Distribution network,CIRED 18th International Conference on ElectricityDistribution, Turin, 2005.

VIII


10/133

Acronyms

Technical Acronyms

CHP Combined Heat and PowerDG Distributed GenerationELF Effective Load FactorGMC Generation Marginal CostLAF Loss Adjustment FactorLF Load FactorLFG Landfill GasLP Linear ProgrammingLV Low VoltageOPF Optimal Power FlowR ResistanceX ReactanceHz HertzHV High VoltagekA KiloamperekV KilovoltMV Medium Voltage

MVA Megavolt-ampereMVAr MegavarMW MegawattMWh Megawatt HourROCOF Rate of Change of FrequencySAIDI System Average Interruption Duration IndexSAIFI System Average Interruption Frequency IndexSCA Steel Core AluminiumSCL Short Circuit LevelT&D Transmission & DistributionXLPE Cross Linked Polyethylene

IX


11/133

X

Organisation Acronyms

AER Alternative Energy RequirementAIP All Island ProjectATSDR Agency for Toxic Substances and Disease RegistryBWEA British Wind Energy AssociationCENELEC European Committee for Electrotechnical StandardizationCER Commission for Energy RegulationCIAB Coal Industry Advisory BoardCIGRE International Council on Large Electric SystemsCPUC California Public Utilities Commission

DCMNR Department of Communications, Marine & Natural ResourcesDNO Distribution Network OperatorDTI Department of Trade and Industry UKEEI Edison Electric InstituteEIA Energy Information AdministrationESB Electricity Supply BoardESBI Electricity Supply Board InternationalESB NG Eirgrid (formerly known as Electricity Supply Board National Grid)ETSU Future Energy Solutions (formerly ETSU)EU European UnionEU ETS European Union Emissions Trading Scheme

EWEA European Wind Energy AssociationIEA International Energy AgencyIEEE Institute for Electrical and Electronic EngineersOFGEM The Office of Gas and Electricity MarketsNERC North American Reliability CouncilTSO Transmission System OperatorUCTE Union for the Co-ordination of Transmission of ElectricityU.S. DOE United States Department of EnergyU.S. EPA United States Environmental Protection AgencyUN United NationsUNFCCC United Nations Framework Convention on Climate Change


12/133

List of Figures

1.1 Share of total world energy consumption (IEA,2006b) . . . . . . . . . . 4

2.1 T&D levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

3.1 Voltage rise effect. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

3.2 Optimal allocation methodology . . . . . . . . . . . . . . . . . . . . . . 30

3.3 38kV 5 bus radial distribution network diagram. . . . . . . . . . . . . . 32

3.4 Individual bus voltage sensitivities to power injections at buses . . . . . 33

3.5 Dependence of bus voltages on power injections at bus C. . . . . . . . . 343.6 SCL at 38/110kV station vs. power injections at buses . . . . . . . . . . 35

3.7 SCL contributions of other buses to Bus B. . . . . . . . . . . . . . . . . 35

3.8 Network sterilisation effect due to generation at bus A . . . . . . . . . . 37

3.9 Network sterilisation effect due to generation at bus B . . . . . . . . . . 37

3.10 Network sterilisation effect due to generation at bus C . . . . . . . . . . 38

3.11 Network sterilisation effect due to generation at bus D . . . . . . . . . . 38

3.12 Network sterilisation effect due to generation at bus E . . . . . . . . . . 39


3.14 No load individual loss characteristics for generation at all buses . . . . 43

3.15 Loss characteristic for generation at buses B & C . . . . . . . . . . . . . 44

3.16 Loss & power import/export over a typical day for scenario 4 . . . . . . 47

4.1 Iterative method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59


5.1 System impact methodology. . . . . . . . . . . . . . . . . . . . . . . . . 76

5.2 Future per-county distribution of installed wind power capacity (%) . . 81

5.3 DG output as a % of their installed capacity on a sample day . . . . . . 85

6.1 Different curtailment methods for two generator case . . . . . . . . . . . 96

6.2 Comparison of biomass output profiles . . . . . . . . . . . . . . . . . . . 102

XI


13/133

LIST OF FIGURES XII

6.3 Annual cost of curtailment under all scenarios . . . . . . . . . . . . . . . 105


14/133

List of Tables

3.1 Voltage Interdependency (kV/MW) . . . . . . . . . . . . . . . . . . . . 32

3.2 SC LT x Dependency (kA/MW) . . . . . . . . . . . . . . . . . . . . . . . 33

3.3 SCL Interdependency (MVA/MW) . . . . . . . . . . . . . . . . . . . . . 34

3.4 Optimal allocation (MW) . . . . . . . . . . . . . . . . . . . . . . . . . . 36

3.5 Generation Loss Interdependencies ij (MWloss/MW) . . . . . . . . . . 45

3.6 Generation Allocations (MW) . . . . . . . . . . . . . . . . . . . . . . . . 45

3.7 Total Annual Losses (MWh). . . . . . . . . . . . . . . . . . . . . . . . . 48

3.8 Total AnnualET x (MWh) . . . . . . . . . . . . . . . . . . . . . . . . . . 48

4.1 Generation Load Factors. . . . . . . . . . . . . . . . . . . . . . . . . . . 54

4.2 Bus Voltage Sensitivitiesij(kV/MW) . . . . . . . . . . . . . . . . . . . 61

4.3 Generation Loss Adjustment Factors . . . . . . . . . . . . . . . . . . . . 62

4.4 Effective Load Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

4.5 Energy Resources (MW) & Line Routes to buses (km) . . . . . . . . . . 64

4.6 Optimal Energy Allocation (MW) . . . . . . . . . . . . . . . . . . . . . 65

4.7 Bus Allocations PDG (MW) . . . . . . . . . . . . . . . . . . . . . . . . . 66

4.8 Energy Output (MWh) . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

4.9 Voltage Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

4.10 Allocation Costs (e) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

5.1 Assumed Energy Resource for the Five Network Sections. . . . . . . . . 82

5.2 Allocations under the Conn. Policies for the Five Network Sections . . . 83

5.3 Scaled Allocations under the Conn. Policies for the System . . . . . . . 84

5.4 Connection Costs for Wind Generation. . . . . . . . . . . . . . . . . . . 85

5.5 Additional Cycling Costs with Wind Generation . . . . . . . . . . . . . 86

5.6 Emissions Savings with Wind Generation . . . . . . . . . . . . . . . . . 86

5.7 Fuel Savings with Wind Generation (PJ). . . . . . . . . . . . . . . . . . 87

5.8 Net Benefits of Wind Generation (inem) . . . . . . . . . . . . . . . . . 87

6.1 Curtailed Energy (MWh) under Different Curtailment Methods . . . . . 101

XIII


15/133

LIST OF TABLES XIV

6.2 Yearly Variance of Voltage Sensitivities2(kV/MW)2 . . . . . . . . . . 102

6.3 Curtailed Energy (MWh) with Variable Voltage Sensitivities . . . . . . 103

6.4 Curtailed Energy (MWh) with Variable Market Price . . . . . . . . . . 104


16/133

CHAPTER1

Introduction

THE CONSUMPTION of energy across the world has risen dramatically since the

start of the industrial revolution. It was at this point in history that the worlds

fossil fuel resource began to be used intensely for the first time. Since then worldwide

energy consumption has continued to grow rapidly. We now find ourselves at a turning

point in our history. We are fast approaching the peak of our consumption of fossil fuel

and are faced with dwindling fossil fuel resources, particularly oil and gas ( EIA,2006).

Much of the technological and social development of our world has been based on

cheap, readily available energy. It has to some extent been taken for granted. This

is now changing as we are faced with increasing prices for oil and gas and increasing

concerns about security of supply (DTI, 2006a). This has forced a reexamination of

how we both consume and harvest energy. Increasing demand for an increasingly scarce

resource inevitably results in a rise in price. There is rising global demand for energy

as India, China and other countries rapidly grow their economies. Global demand for

1


17/133

Chapter 1. Introduction 2

natural gas is projected to increase nearly twofold by 2030. Hopes of reducing the global

demand for energy are not realistic. The main reserves of oil and gas are concentrated

in a few regions of the world: Russia, Central Asia, the Middle East, and Africa. Two

countries, Russia and Iran account for nearly half the worlds proven gas reserves (DTI,

2006b). Increased efficiency in both the conversion and transport of energy is possible,

but is only worthwhile if the efficiencies gained outweigh the cost. Another option is to

use alternative sources of energy. A number of energy resources which had previously

been disregarded by governments are now coming back into vogue. Nuclear energy is

now firmly back on the agenda in the UK as can be seen from the UK energy review in

July 2006 (DTI,2006b). There are also plentiful resources of coal remaining and new

technologies have been developed to reduce its harmful emissions (CIAB,2006).

The upward trend in fuel prices is likely to continue as the global fossil fuel resources

become more depleted (EIA, 2006). The increasing concentration of remaining fossil

fuel resources in certain regions may also make fossil fuel prices more volatile. The

future fossil fuel price characteristics will have a significant impact on electricity indus-

tries and the societies that rely on them. In addition to looking at energy resources such

as nuclear or coal, the search is now on to harness new forms of energy, which have not

previously been utilised. Concerns about global warming are a significant driver for re-

newable energy sources. The international community is taking the first steps towards

addressing this problem with initiatives such as the Kyoto Protocol (UNFCCC,1997)

and the European emission trading scheme (EU ETS,2005). Increasing the proportion

of electricity served by renewable sources is cited as a means to reduce greenhouse gas

emissions and reduce the reliance on fossil fuels (Zervos,2003;Neuhoff,2005).

A number of renewable energy resources are being considered as part of the solution

to the worlds energy problem. There are various forms of renewable energy, each at

varying stages of development. The most well developed form of renewable energy is

hydro power, which harnesses the energy of moving water in rivers. This is a well estab-

lished technology which has been employed since early in the 20th century. Indeed, the

majority of renewable energy produced worldwide is from hydro (IEA,2006a). With

reference to Ireland, the first electricity generator on the Irish system was a hydro


18/133


power station at Ardnacrusha, built in 1926 (Duffy, 1990). However, in Ireland and

many other countries the available hydro resource has been almost completely utilised.

Of the other renewable energy resources, wind energy is the most advanced with increas-

ing penetrations of wind energy being seen in power systems across the world (EWEA,

2006). Two other forms of renewable energy are tidal and wave generation. Both

are still in the developmental stage, but do have the potential to make a significant

contribution to the worlds energy needs (Bryans et al., 2005; Polinder and Scuotto,

2005). Landfill gas (LFG) is another energy resource, if somewhat limited, that is be-

ing utilised. It involves the burning of methane at landfill sites that results from the

decomposition of organic matter. Landfill gas energy facilities capture the methane

(the principal component of natural gas) and combust it for energy (EIA,2005). It is

considered a green form of generation as the methane gas burned, which is a greenhouse

gas, would have to be released into the atmosphere anyway. Biomass is, in its various

forms, now being seriously considered as an energy source. Biomass energy is derived

from two main energy sources: wood and alcohol fuels. It has previously been utilised

for industry residue, such as in sawmills (EIA,2005). Photovoltaics are another tech-

nology still largely in the developmental stage, although there are some installations

internationally (CPUC, 2006). All of these alternative energy resources are likely to

increase in utilisation and contribute to the worlds energy requirements in the coming

years.

1.1 Electric Power Systems

Electricity demand is growing the fastest of all the energy consumed worldwide. From

Figure 1.1, it can be seen that in 1973 electricity accounted for 9% of global energy

consumption. This had grown to 16% by 2003, reflecting the growth of modern indus-

trial society. This rise in the use of electricity is closely linked to the development of

countries. It is a trend which is almost certain to continue into the future.

It is predicted that the worlds net electricity consumption will double from 14,781

billion kWH in 2003 to 30,116 billion kWh in 2030. Non-OECD countries account


19/133


Figure 1.1: Share of total world energy consumption (IEA,2006b)

for 71% of the projected growth and OECD countries 29% (EIA, 2006). This trend

has increased the importance of electricity networks and how they are operated and

planned. It has focused the attention of policymakers and politicians on how this

asset is being managed. The evolution of the electricity industry over the last 120

years has mainly been a technical one. Traditionally, electricity systems have been

run by vertically integrated utilities with the focus on the engineering aspects rather

than economic issues. This usually led to a high quality of supply for the consumer

but perhaps at the expense of cost effectiveness. A number of recent drivers have

altered this situation. For example, in accordance with EU Directive 03/54/EC all EU

countries are in the process of opening up their electricity sector to competition ( EU,

2003), in particular for generation.

It is the responsibility of the system operator to plan and develop the electric power

system, schedule and dispatch generation, operate the electricity market and ensure

system security. System operators are now faced with the challenge of maintaining

the high quality of supply that consumers expect while faced with volatile fuel prices,

dwindling and increasingly insecure energy resources and the integration of new forms

of energy. These issues when dealt with in a liberalised market situation are not merely

technical but also economic. As such, due consideration must be given to costs at every


20/133


stage of planning and operating the system. The electric power system is an energy

transport network and with the predicted large growth in electricity demand, it is likely

to become an even more important energy transport system.

1.1.1 Transmission & Distribution of Electricity

In most typical power systems, power is generated at a number of large centralised

power stations. The transmission and distribution (T&D) system is used to transport

this centrally generated power to the consumer, be they industrial or domestic. As a

result, the primary mission of the T&D systems is to deliver energy to consumers at

their place of consumption in a ready to use form on a continuous basis (Willis,2004).

The requirement of continuity leads to very high standards of reliability. Reliability

of supply is defined as the ability to supply adequate electric service on a nearly con-

tinuous basis with few interruptions over an extended period of time (IEEE/CIGRE,

2004). With this in mind it is worth noting that providing supply for 99.90% of the

time is an unacceptable level of reliability in any first world country. As a result of

this onerous task, the electric power system is extremely complex. It is essentially one

machine, in many cases reaching across national borders, consisting of multifarious in-

teracting components (UCTE, 2006). The task of running the power system remains

as complex as ever and is constantly facing new challenges. This thesis deals with one

of the major new challenges facing system operators - Distributed generation (DG).

1.2 Distributed Generation

The drivers detailed above regarding the liberalisation of electricity systems, the push

for renewable energy and dwindling fossil fuel resources are leading to changes in the

planning and operation of electric power systems. As new forms of energy connect,

new methodologies for planning and operating the power system are required, dealing

with issues ranging from system dynamics to longer term considerations, such as plant

mix (Doherty and OMalley,2005;Laloret al., 2005). One of the main changes being


21/133


seen in power systems across the world is the increased proliferation of what is known

as distributed generation (Jenkinset al., 2000). It is so called given its geographical

distribution throughout the system and its connection to the distribution system. It is

also known as embedded generation and dispersed generation, but for the purposes of

this thesis, the term distributed generation will be used.

There are a number of definitions of DG across the world (Ackermann et al.,2000).

Put simply however, DG can be defined as small scale generation, which is connected

to the distribution network. Edisons Pearl St. generator consisted of a steam-engine-

driven dc generator and supplied 59 customers in an area roughly 1.5km in radius (EEI,

2005). It was not only the first example of an electric power system but perhaps also the

first example of a distributed generator. DG can take the form of renewable generation

or small scale conventional units. There is a difference in emphasis between Europe and

North America. In the US, DG has been proposed as a sustainable and competitive

form of generation, not just restricted to renewable energy. In light of the recent

blackout in New York, it is proposed as a significant part of the solution to improve

both security and reliability of supply, with islanded operation of DG on microgrids a

real prospect for the future (U.S. DOE, 2000). In Europe however, it is viewed more

as renewable generation, which is given priority dispatch and should be accommodated

where possible. Nonetheless, across both continents increasing penetrations of DG are

being seen.

In recent years, there have been increasing volumes of applications for access to the

distribution system. In particular, in Ireland at the end of 2004, applications received

by the system operators concerned the connection of approximately 2,500MW of wind

generation. A significant amount of this capacity is to be distribution connected. Wind

farms with a rating less than 30MW are suitable for connection to the distribution

system, given the lower cost of connection for the generator and their relatively small

capacity. This new wind capacity is in addition to approximately 500MW of previously

contracted wind farm capacity and approximately 100MW of other forms of DG such

as landfill gas (LFG) and hydro. In a country with a peak load of approximately

4,945MW this is an extremely large amount of DG to integrate into the distribution


22/133


system (CER,2004a;Eirgrid,2007).

The increasing penetration of DG is changing the role of distribution systems. Whatwere originally passive networks purely for the delivery of electricity to the consumer,

are now networks that are being utilised for the harvesting of energy from a myriad

of distributed energy resources. The increased proliferation of these distributed gen-

erators has lead to changes in the characteristics of the network, with more variable

and bidirectional active and reactive power flows. These generators are altering the

technical characteristics of the networks, and pushing them to operate closer to their

limits of safe and reliable operation. As a result, the need for distribution networks to

operate at their maximum capacity is being felt particularly with DG. In many coun-

tries the generator must pay for any distribution network reinforcements required for

connection. This can present a considerable cost to the generator and may be a barrier

to further DG penetration. As a result, the available capacity of the existing network

should be utilised fully.

1.3 Focus of Research

The focus of the research in this thesis is on the integration of DG, in particular,

the planning and operation of distribution systems with significant penetrations of

DG. It is focused on the technical issues that arise when connecting generation to the

distribution system. The overall approach adopted attempts to maximise the amount

of DG with the least cost to society. It takes account of the technical constraints which

present a barrier to further DG penetration. A number of methodologies are developedwhich challenge the existing methods used by network operators when processing DG

applications. In particular, it assesses the optimal way to plan for increased integration

of DG, through maximisation of energy, while taking due account of capacity, losses

and costs. The operation of DG is also considered with regard to active constraint

management.

The first step in any research project is to undertake extensive reading in the topic


23/133


of interest and in so doing determine the main issues relating to the topic, in this

case the integration of DG. In Chapter 2these issues are discussed along with a brief

description of the Irish distribution network. The chapter also includes a literature

review of publications dealing with the challenges posed by DG.

One of the main issues arising with regard to DG is the available capacity on the

existing network. In Chapter3 the optimisation of DG capacity is addressed. All rel-

evant technical constraints are modelled and a methodology for maximising the DG

capacity that may be connected to the existing distribution network is developed. It

is shown that a significant increase in DG capacity is possible and that the effect of

network sterilisation is potentially severe. The voltage constraint is identified as the

main barrier to further DG penetration in rural distribution systems. The issue of

losses and DG is also addressed. The initial aim of including losses was to extend the

scope of the methodology to beyond that of just capacity. It was felt that while there

were clear benefits to pursuing a capacity maximising strategy, there were other consid-

erations that should also be considered. In Section3.4 the methodology is extended to

include the minimisation of losses when allocating DG capacity. One of the main issues

arising out of the losses analysis was that to capture fully the impact of DG on losses,

consideration of energy rather than just capacity was necessary. As a supplement to

the capacity analysis of losses, an annual simulation of the network is employed to

demonstrate the impact of losses over the whole year. Further to this, it became clear

that issues such as plant mix had a considerable bearing on not just losses but also on

maximising the benefit from DG.

It is evident that considering losses and capacity on their own is not sufficient

to maximise the benefit from DG and that there are other significant factors to be

considered. The idea of maximising energy rather than just capacity presents itself as

one way of increasing the benefits of DG. One specific aspect of maximising energy is

the issue of non firm access. In Chapter4,the issue of firm vs. non firm access for DG

is discussed. The DG plant mix is considered and a new methodology which maximises

the energy harvested from DG is developed. In addition, the resulting connection costs

are considered and account is taken of the temporal variations of DG. In particular,


24/133


voltage sensitivities are employed to minimise the voltage rise on the network and hence

reduce the instances of overvoltage over the year. In addition, novel loss adjustment

factors are developed to take account of the average impact of DG on losses. Annual

simulations are utilised once more to test the effectiveness of the methodology.

The consideration of a non firm connection policy on a single section of network

in Chapter4gives rise to the further question of the impact of DG connection policy

on the whole system. In Chapter5 an assessment of the impact of connection policy

on the system net benefits is undertaken. Costs and benefits such as cycling, capital

and connection costs along with fuel and emissions savings are examined. Through

the analysis of representative sections of network and with reference to network and

energy resource data, the local results are scaled to a national level with system costs

and benefits of DG under a number of connection policies determined.

The final part of the research for this thesis resulted from collaboration with the

Institute for Energy Systems at the University of Edinburgh. Whereas the previous

chapters in the thesis detail aspects of DG planning, this chapter is focused on the

operation of DG. In Chapter 6, DG operational issues are addressed, specifically, theoperation of a voltage management scheme for DG is investigated. A number of existing

congestion management methods used on the transmission system are applied to the

distribution system and a novel approach is developed taking account of the cost of

curtailment. The voltage sensitivity factors used in the analysis vary, depending on

the operating condition. The variability of these sensitivities is investigated and the

impact of the variability on the curtailment method determined.

General conclusions regarding the entire work are given in Chapter 7 along with

suggestions for further work.


25/133

CHAPTER2

Distributed Generation

THERE are a number of concepts which govern the transmission and distribution

of power, some of which are given here (Willis,2004).

1. Power is more economical to produce in very large amounts

2. It is more economical to move power at high voltage

3. The higher the voltage, the greater the capacity and the greater the cost ofotherwise similar equipment

These concepts lead to the structure of the power systems that exists today. The

structure is a hierarchical system of voltage levels used to transport the power from bulk

generation sources to the consumer. The power is transported from the power plants

at high voltages, which are gradually stepped down as the power nears the consumer.

10


26/133

Chapter 2. Distributed Generation 11

These voltage levels lead to different parts of the system being classed as shown in

Figure2.1. The voltage levels shown are the voltage level used on the Irish system.

Figure 2.1: T&D levels

The transmission system interconnects all the major generating stations and load

centres. It is operated at high voltage, (typically 110kV). The distribution system

comprises everything below this level (


27/133


has implications for the three guiding concepts of T&D systems given above. The first

concept was that it is more economical to produce electrical power in large amounts.

This is also being somewhat challenged, in that it is now proving economical to produce

power from distributed energy resource. Secondly, it is more economical to move power

at higher voltage and this has lead to a relatively small number of large centrally

operated generation plants in systems. The increased proliferation of DG means that

there is less need for these high voltage networks as the generation may be connected

more cheaply to the lower voltage network, which is in turn closer to the load. Secondly,

it is still true that the higher the voltage the higher the capacity and cost, but it is

because of this very fact that smaller generators are now connecting to the distribution

system.

There are three basics principles of distribution planning (Willis, 2004). Firstly,

the system must reach the end consumer, i.e. there must be a line connecting the

consumer to the network. This factor shapes the structure of the system and leads to

many of the constraints and challenges faced by power distribution planners. Secondly,

the system must have enough capacity to meet peak demand, i.e. the system must be

planned to meet all extreme operating points rather than the average. Thirdly, the

reliability of supply must be kept at very high levels. Reliability of 99.975% is typically

required in most distribution systems, i.e. approximately 2 hours of service interruption

is acceptable each year. To achieve this level of reliability requires protection, control

equipment and a level of redundancy. All of this leads to reliability accounting for 50%

of the cost of most T&D systems (Willis,2004).

The combined result of these three principles was that distribution systems and

the whole power system was highly developed with a larger degree of redundancy built

into it. However, with the liberalisation of the electricity industry and the introduc-

tion of electricity markets, these principles are being influenced by the need for cost

effectiveness. DG is now connecting and it has now become necessary to operate the

distribution system closer to its maximum capacity (Kersting,2002). This pressure to

get more out of the existing assets can pose problems in terms of security of supply

and reliability. Operating a system closer to its limits is not inherently detrimental,


28/133


but it does inevitably lead to a number of network issues or constraints which should

be considered.

2.1 Distributed Generation Network Issues

2.1.1 General Planning & Operational Issues

The introduction of DG alters the characteristics of the distribution system. A number

of technical constraints and factors arise which are impacted by the amount of DG that

is connected. These issues are:

Equipment Ratings

Short Circuit Level

Short Circuit Ratio

Voltage Rise

Losses

Power Quality

Reliability

Protection

2.1.1.1 Equipment Ratings

The rated current of the lines must not be exceeded. Under standard voltage levels and

power factor conditions the rated current of the line can be translated directly into a

rated active power for that line. There is also the constraint of the transformer rating,

where the amount of generation connected minus the minimum load must not exceed

the transformer rating.


29/133


2.1.1.2 Short Circuit Level (SCL)

The magnitude of the transient voltage drop experienced at the buses in a network is

an indication of the strength of the system. In this manner the SCL is a measure of

the strength or robustness of a system (Elgerd,1971). The SCL of a system refers to

the current that results when there is a fault on the system. Generators, depending

on the type of electrical machine employed, may contribute to the SCL. An increase in

the SCL is generally favourable as it will increase the strength of the system. Although

it must be ensured that the short circuit level does not exceed the rating of breakers

and other equipment and this poses a hazard to the safe operation of the network.A maximum short circuit rating for all equipment is laid down in distribution codes

(ESB Networks, 2002). Previous work has shown it to be a significant factor to be

considered with DG (Vovos et al.,2005).

2.1.1.3 Short Circuit Ratio (SCR)

The short circuit ratio is the ratio of generator power to the short circuit level ( Taylor,1994). It gives an indication of the voltage dip experienced near the generation in the

event of a feeder outage. The connection of induction generators to high impedance

circuits may lead to voltage instability problems if the SCR is not kept within acceptable

limits. The dip in wind farm terminal voltage that results from a fault leads to an

acceleration of the induction generator, leading to overspeed (Holdsworthet al.,2001).

If the speed is increased to a level above the critical value, the generator will accelerate

out of control. This may lead to voltage collapse as the induction generator absorbs

more reactive power (Akhmatovet al.,2000). If the short circuit level is large enough,

the transient voltage dip will be limited and the system will remain stable.

2.1.1.4 Voltage Rise

If DG is connected to a network section, it will alter the active and reactive flows and

hence change the voltage dropped along the lines. It has been shown that DG leads


30/133


to a significant voltage rise at the end of the long, high impedance lines (Dinic et al.,

2006;Masters,2002). A rise in voltage occurs if there is low demand and high genera-

tion, which leads to a large amount of power flow along lightly loaded lines with high

impedance. This problem is particularly acute in rural areas, where demand tends to

be low. In addition, the resistive element of the lines on distribution systems is higher

than other lines. This leads to an X/R ratio of 5 rather than a more typical value of

between 10 and 20 on transmission systems. This only serves to exacerbate the voltage

rise problem.

2.1.1.5 Losses

Losses are an important consideration when designing and planning the distribution

system. Losses are inevitable on any network, however, the amount can vary consid-

erably depending on the design of the network. With the introduction of distributed

generation, the network is being utilised in a different way with more variable and bidi-

rectional power flows. The level of losses is closely linked to the power flows. Losses are

a function of the square of the current, i.e. a doubling in current results in losses beingquadrupled. Therefore the allocation of DG and the altered power flows that result

may have a significant impact on losses and may provide an opportunity to ameliorate

them.

2.1.1.6 Power Quality

DG can have a considerable impact on power quality within the distribution system.Voltage flicker refers to dynamic variations in the system voltage. It can be an issue with

wind power, given the variable nature of its energy source. However, the development

of more sophisticated turbines has reduced the severity of flicker, due to the increased

filtering effect of the generators. Some forms of DG may employ power electronic

converters to interface with the system. This can alter the harmonic impedance of

the system and care must be taken at the design and planning stage. In particular,

there is the potential for resonances between capacitors or cables, which may have a


31/133


detrimental effect on the operation of the generator. There are standards which dictate

the acceptable levels of each of these quantities, which must be adhered to (CENELEC,

1994). DG also has the potential to improve the power quality, through its contribution

to the short circuit level.

2.1.1.7 Reliability

Reliability has always been an important issue for system planners and operators

(Jenkinset al., 2000). Reliability can be measured by the indices of SAIFI (System

Average Interruption Frequency Index) and SAIDI (System Average Interruption Dura-

tion Index), which measure the average frequency and duration of supply interruptions

respectively. With the advent of DG, new reliability models and methods of assessing

distribution reliability have been proposed. In particular, a number of probabilistic

techniques have been proposed as opposed to the traditional deterministic approaches

used (Allanet al., 1999;Billinton et al., 2001). In addition, the possibility of islanded

operation of DG sources has also been discussed, with particular regard to dispatchable

forms of DG (IEEE,2003).

2.1.1.8 Protection

The majority of protection systems for distribution network operate based on unidi-

rectional power flows from the transmission network down through the lower voltage

networks (Jenkinset al.,2000). As mentioned previously DG is changing the flows on

the network. This can lead to problems with the operation of the protection system suchas false tripping (Wallace, 1999) another issue encountered with DG is loss of mains

detection or islanding. Two techniques for loss of mains are normally used. These are

vector shift and rate of change of frequency (ROCOF). ROCOF is the most common

form of loss of mains protection used as it may be used with induction generators.

Induction generators are the most common machine type used with wind generation.

ROCOF operates based on the resulting change in frequency from the loss of grid sup-

ply. There are however a number of performance issues associated with ROCOF relays


32/133


as detailed inBeddoes et al.(2005). New methods of islanding detection have also been

proposed which are designed to work effectively with DG (Jang and Kim,2004).

2.1.2 Irish Distribution Network

In Ireland, DG is typically connected at 38kV or MV (10kV or 20kV)(PB Power,2004).

The vast majority of applications for connections are wind farm projects. These wind

farms are typically located in remote areas where there is very little existing network

and also very little demand for electricity.

In Ireland and other countries, voltage rise tends to be the dominant constraint and

quite often the critical factor when offering capacity on the network. This is due to

the existing network in Ireland, which outside Dublin is typically a weak rural network

with a large amount of conductor. A weak network means a network with a low SCL

or fault level. There is approximately 159,000km of distribution network in Ireland.

These long stringy networks are not equipped to support a large amount of generation.

Due to these high impedance lines, the sending voltage at the transmission station is

usually set to a higher than nominal value (i.e. 41kV at 38kV) to ensure the voltage is

within standard at the end of the line when no generation is present, thus maximising

the load that can be supplied from these networks. Within each voltage level there are

a number of options for connecting generation into the system. These are:

Direct feed to 110kV/38kV or 110kV/MV station

Direct feed to 38kV/MV station

Tee connection onto existing line

If the capacity of a project is less than 5MW, connection to the MV network may

be possible and is typically cheaper. Tee connections are not generally permitted as

they can lead to a decrease in the security of the system.


33/133


2.2 Literature Review

The promise of a new dawn of widespread DG sources proliferated throughout distribu-

tion systems has lead to a certain degree of hype in the electricity industry. Nonetheless,

DG does offer many benefits as well as presenting numerous challenges. A large amount

of research has been carried out investigating the impact of DG. A selection of the more

relevant publications is given here.

A number of publications have looked at optimising the placement and sizing of

DG based on various criteria. InWallace and Harrison(2003), the authors employ an

optimal power flow (OPF) technique to maximise DG capacity with respect to voltage

and thermal constraints. Short circuit levels, short circuit ratio, equipment ratings and

losses are not considered. The effect of network sterilisation is clearly demonstrated

by comparison between allocating generation to buses individually rather than as a

group. In Vovos et al. (2005), a method is presented utilising OPF for the allocation

of generation capacity, which includes a detailed fault level constraint. InKuri et al.

(2004) genetic algorithms were used to place generation such that losses, costs and

network disruption were minimised and the rating of the generator maximised. The

constraints considered were voltage, thermal, short circuit and generator active and

reactive power capabilities. Generation is placed in single units at individual buses,

while ignoring the interdependence of the buses and the network sterilisation that can

result from improper DG placement. In El-Khattamet al. (2004) the authors use a

heuristic approach to determine the optimal DG size and site from an investment point

of view. Once again short circuit constraints are not considered and the focus of the

objective function is on optimal investment rather than maximising renewable energy.

It uses a cost benefit analysis to evaluate various placements of DG. In Duganet al.

(2001), a planning process that considers DG as well as more conventional options is

presented. DG investment in considered under a number of load growth scenarios,

with its benefit found to vary between each scenario. However, it is still found to be

a useful factor to be considered in the distribution planning process along with other

more conventional options such as network reinforcement.


34/133


Other publications have focused on reliability aspects of DG. InCelli et al. (2005),

a multiobjective planning strategy is presented using a genetic algorithm to identify the

best compromise DG sizing and siting. InChowdhury et al.(2003), a probabilistic relia-

bility model is presented to determine the impact of DG for use in distribution planning

studies. InMcDermott and Dugan(2003), the impact of DG on reliability and power

quality is measured. Reliability and power quality indices were applied to a sample

feeder to assess the impact. Work has also been done evaluating the contribution of wind

generation, in particular, to reliability (Clark and Miller,2006;Karki et al.,2006). The

issues of load growth and load patterns in distribution planning are discussed in Willis

(2004) and a multi stage approach to planning is described in Kuwubara and Nara

(1997). InQuezada et al. (2006), the amount of losses incurred with increasing pene-

trations of various DG sources is examined. InWang and Nehrir (2004), the authors

propose a method which places DG at the optimal place along feeders and within net-

worked systems with respect to losses. The allocation of losses to distributed generators

in the network has been addressed inCosta and Matos(2004). Previous work has at-

tempted to quantify the net benefits of DG (Chiradeja and Ramakumar,2004), where

a number of benefits such as reduced losses and voltage profile were assessed.

A number of the operational issues surrounding DG have also been addressed in

the literature. Congestion management is well established on transmission systems,

with existing schemes in place (NERC, 2006; Christie et al., 2000). However, given

their traditionally passive nature, constraint management on distribution networks

is much more unusual. Increasing penetrations of wind generation are also leading

to the necessity of constraining wind generation at critical times (CER, 2004b). In

Sveca and Soder(2003), methods to estimate the amount of wind power that could be

installed in areas with congestion problems are presented. The methods are applied

to the Swedish transmission network and the cost of the spilled energy is determined.

Other work has focused on the coordination of wind and hydro plant, with the hydro

plant being used as storage. The objective was to maximise the generators profit and

to smooth out the combined output of the generators (Castronuovo and Lopes,2004).

In University of Strathclyde (2004), the operation of a proposed active management


35/133


scheme is investigated on the Orkney islands in Scotland. The feasibility and benefits

of non firm access are highlighted, as are the potentially complex operational issues

surrounding the implementation of such a scheme. InZhou and Bialek(2007), the is-

sue of the operation of curtailment on the distribution network is addressed, with a

number of curtailment methods compared and the calculation of voltage sensitivities

examined. The influence of losses on the curtailment rules is also taken account of.

Previous work has examined the impact of distributed resources on congestion man-

agement on the transmission network in terms of contribution factors (Liuet al.,2005).

In Furusawa et al. (2006), the use of customer side generation to provide congestion

relief on the transmission system is examined. Other work has focused on the reliability

worth of DG (Bae et al.,2004) and the consideration of an optimal operating strategy

for DG on an hourly basis. Other operational issues have been investigated such as in

El-Khattamet al. (2006), where a Monte Carlo simulation is employed to assess the

impact of all possible DG operation conditions on the system. In Kashem and Ledwich

(2005), the operational issues of using multiple DG sources for voltage support are

examined with a number of recommendations made.

The voltage constraint has been demonstrated to be the dominant constraint in a

number of circumstances and hence limit to further DG capacity (Masters, 2002). It

has traditionally been assessed at an infrequent N-1 peak condition. While infrequent,

it is important for the operation of the system that the voltage stays within its limits.

Existing generators of all types have the ability to ramp down their output if required

to do so and this capability has been proposed as method of actively managing the

constraint breaches. In addition a number of innovative voltage control techniques

have been proposed to solve the voltage rise effect (Hirdet al., 2004;Hill et al., 2005;

Shafiuet al., 2004;Dinic et al.,2006), along with solutions to the other technical con-

straints such as the short circuit level (Collinsonet al.,2003).


36/133


2.3 Conclusion

It is evident from the selection of publications mentioned above that a considerable

body of work has been carried out into the integration of DG. It is also apparent that

the integration of DG has a number of significant impacts on planning and operation

of distribution systems. As mentioned in Section1.3,one the big issues for DG is the

utilisation of the available network capacity. It is this topic that is the focus of Chapter

3.


37/133

CHAPTER3

Optimal Capacity

THIS chapter addresses the issue of the available generation capacity for DG on the

existing distribution system. A new methodology is developed which allocates

the capacity optimally. There is the prospect of the distribution network developing

from a passive network to an active one in order to facilitate increasing levels of dis-

tributed generation (Liew and Strbac,2003). If this is to be the case a number of steps

should be followed. Firstly, best use should be made of the existing distribution network

by optimal allocation of DG. Secondly, the development from passive to active shouldbe planned optimally with all relevant factors taken into consideration. Traditionally,

DNOs have allocated available generation capacity on a first come first served basis.

However, the placement of generation on a first come first served basis invariably limits

the overall capacity of distributed generation (Wallace and Harrison, 2003). Here, a

new methodology is developed to determine the suitable locations and ratings for DG

on distribution networks with respect to the technical constraints, which will enable

22


38/133

Chapter 3. Optimal Capacity 23

a high penetration of generation on the network and avoid network sterilisation. Net-

work sterilisation results when capacity is allocated to the bus/buses whose voltage,

short circuit levels and/or short circuit ratio are most sensitive to power injections.

Thus, no more generation can be connected as the buses are constrained. Further, the

methodology is developed to include losses. The minimisation of losses is included and

the issues surrounding DG plant mix are initially investigated.

The basis for the new methodology is in exploiting the interdependence, if any, of

the buses with regard to the technical constraints. The constraints all have either linear

or approximately linear characteristics with respect to increasing power injections or

they place a fixed and independent limit on the power injection. The methodology

described here determines the optimal allocation with respect to all of the relevant

technical constraints. Optimal allocation ensures better use of the existing assets and

achieves a higher penetration of DG in a cost effective manner. In Section 3.1, the

technical constraints on DG are outlined and the development of the optimal capacity

methodology is explained. A section of the Irish distribution network is modelled in

DIgSILENT Powerfactory using network data obtained from the distribution network

operator (DNO). The methodology is tested on this section and results and discussion

are shown in Section 3.3, which illustrate how network sterilisation is avoided if DG

is optimally allocated. In Section 3.4 losses are described. In Section 3.4.2 the for-

mulation of an objective function with losses included and the extended optimisation

methodology are outlined. Results are shown in Section 3.4.4 illustrating the effects

that DG can have on losses. The results show that the absolute maximisation of DG

capacity does not always lead to the maximisation of energy serving load. Discussion

of the results is in Section 3.5and conclusions are given in Section3.6.


39/133


3.1 Optimal Capacity Allocation Methodology

3.1.1 Objective Function

The objective is to maximise generation capacity subject to the constraints outlined

below. This methodology ensures optimal use of the existing network assets, thus

helping to meet the renewable energy targets in a cost effective manner. Generation

capacity should be allocated across the buses such that none of the technical constraints

are breached and the capacity maximised. Therefore the proposed objective function

is as shown in Equation (3.1).

J=N

i=1

PDG i (3.1)

WherePDG i is the DG capacity at the ith bus andNis the number of buses. Without

loss of generality, it is assumed that there is one generator connected at each bus.

The objective function J (MW), given in Equation (3.1) is maximised subject to the

constraints, which are formalised below. These constraints were discussed previouslyin Section2.1.

3.1.2 Equipment Ratings

3.1.2.1 Transformer Capacity

The amount of generation connected minus the summer valley load must not exceed the

rating of the transformer at the higher voltage. If there is some existing generation then

this must be subtracted from the total. The result is the remaining capacity available

below that station. In the case of two parallel transformers, the capacity is taken as

the rating of the smaller transformer plus the summer valley load. The constraint is

expressed formally as in Equation (3.2).

PT x PTrafoCap. (3.2)


40/133


Where PT x refers to power flow through the transmission substation transformer and

PTrafoCap refers to the rating of that transformer.

3.1.2.2 Thermal Constraint

This is a stand alone constraint, simply put the rated current of the lines must not be

exceeded. It is given by Equation (3.3).

Ii < IRatedi i N. (3.3)

Where Ii is the current flowing from generator i to bus i and IRatedi is the maximum

rated current for the line between each generator and its corresponding bus. Under

standard voltage and power factor conditions the rated current of the line can be

translated directly into a rated active power for that line.

3.1.3 Short Circuit Level

A maximum short circuit rating for all equipment is laid down in distribution codes

(ESB Networks, 2002; OFGEM, 2006). A short circuit calculation is carried out to

ensure that this constraint is not exceeded as the level of installed capacity increases.

The short circuit level (SCL) is highest at the transmission system bus. Buses close to

this bus may find their capacity limited as a result. The constraint is given by Equation

(3.4).

SC LT x< SCLRated. (3.4)

Where SC LT x is the short circuit level at the transmission substation busbar and

SC LRatedis the highest current that switchgear can safely break under fault conditions.

The contribution of increasing levels of generation at each bus to SCLT x is determined

by short circuit analysis. The SCL contributions of generation at each bus are combined


41/133


and formalised into an algebraic equation as shown in Equation (3.5).

Nj=1

jT xPDG j+ T x SC LRated. (3.5)

WherejT xis the dependency of the SCL at the transmission station to power injections

at bus j , i.e. the slope of the SCL vs. power injection characteristic of the j th bus as

shown later in Figure3.6. PDG j is the power injection at thejth bus, T xis the initial

SCL at the transmission bus with no generation present.

Equation (3.5) is a very accurate model of the short circuit characteristics when an

intelligent choice of the range for calculation of the slopes is made. The optimisation

is initially calculated with respect to the constraints as outlined. The initial range for

calculation of the slopes is selected such that the fixed limit constraints are satisfied.

The allocation calculated by the optimisation algorithm for each bus is then used as the

new range over which to recalculate the slopes for the SCL constraint, thus improving

the accuracy of the dependencies. When the optimisation is recalculated with the

new dependencies it has been found that it gives a very similar answer to the initial

allocation. It is found that the linear approximation is adequate, but that the iteration

process is worthwhile as it does not add considerably to the calculation time and ensures

an accurate determination of the optimal allocation. The short circuit level used at the

transmission station is the winter peak level given in the transmission system operators

(TSO) forecast statement (ESB NG,2005).

3.1.4 Short Circuit Ratio

The short circuit ratio (SCR) is the ratio of generator power PDG(MW) at each bus to

the short circuit level at each busSCLBus (MVA) (Taylor,1994). It gives an indication

of the voltage dip experienced near the generation in the event of a feeder outage. The

connection of induction generators to high impedance circuits may lead to voltage

instability problems if the SCR is not kept within acceptable limits (Holdsworthet al.,

2001). If the short circuit ratio is small enough, a transient voltage dip occurring will be


42/133


limited and the system will remain stable. The contribution of other generators to the

SCL is considered as it may be significant depending on the proximity of the bus. The

phase angle at the busbar is omitted as an extra margin of safety. It could be included

and the allowable ratio set to a lower value such as 6%. A value of 10% is largely in

line with values used by other utilities and is in line with the value recommended in the

European standard EN50160 (CENELEC,1994). The ratio itself is shown in Equation

(3.6).PDG i

SC Li. cos() 100 10% i N. (3.6)

WhereSC Lirefers to the short circuit level at the ith bus and cos() is the power factor

at the generator. A base value for the short circuit level at the ith bus is calculated

with no generation present on the network, i. The contribution, if any, of generation

at other buses to this level is then calculated allowing the short circuit characteristic of

each bus to be formulated into an algebraic equation. The SCL at the ith bus is given

by Equation (3.7).

SC Li = i+N

j=1

ji PDG j i =j, i N. (3.7)

Subbing (3.7) into (3.6).

PDG i 0.1cos()N

j=1

ji PDG j 0.1cos()i. (3.8)

For this constraint the short circuit level used at the transmission station is the

summer night valley level, as this is the time when the system short circuit level will

be lowest.

3.1.5 Voltage Rise Effect

From (Jenkinset al., 2000) and with reference to the circuit shown in Figure 3.1, the

voltage at the generator is given by Equation (3.9).


43/133


Figure 3.1: Voltage rise effect

VG= VL+RPL+ XQL

VL+j

XPL RQLVL

(3.9)

Where Z=R+jX is the impedance of the line, PL andQL are active and reactive power

at the bus andVG andVL are the voltages at the generator and bus respectively. Thus

it can be seen that the generator voltage will be the load/bus voltage plus some value

related to the impedance of the line and the power flows along that line. It is evident

that the larger the impedance and power flow the larger the voltage rise. The increased

active power flows on the distribution network have a large impact on the voltage level

because the resistive element of the lines on distribution networks are higher than other

lines. The voltage must be kept within standard limits at each bus as given by Equation

(3.10).

Vmin i < Vi < Vmax i i N. (3.10)

Where Vmin i & Vmaxi refer to the minimum and maximum voltage limits at the ith

bus. The relationship between voltage and power injections at each bus is determined.

As megawatts are added at each bus the voltage rises. Increasing levels of generation

are added incrementally at each bus in turn and load flow analysis is carried out to

determine a voltage vs. active power characteristic for each bus. Next the interdepen-

dence of the bus voltage levels is examined. Once again increasing levels of generation

are added incrementally at each bus, but now the voltage level at every other bus is

examined. Thus characteristics are determined for voltage levels at each bus due to

generation at all other buses. By combination of these characteristics the voltage con-


44/133


straint may be formalised into algebraic equations for each bus as shown in Equation

(3.11).

Nj=1

ji PDG j+i Vmax j i N . (3.11)

Where i is the dependency of the voltage level at bus i on power injections at bus i,

i.e. the slope of the voltage vs. power injection characteristic of theith bus. irefers to

the initial voltage level at the ith bus with no generation,ji refers to the dependency

of the voltage level at bus i on power injections at bus j. This analysis is carried out

under minimum load conditions as this is the worst case scenario for voltage rise. Both

the standby and normal forward feed conditions are considered. There is usually morethan one possible standby feeding arrangement, but the most severe feeding condition

is usually readily identifiable.

3.1.6 Energy Resource & Customer Initiatives

All of the network buses may not have a substantial energy resource available, if any.

This can be factored in by adding an inequality constraint for each bus, limiting itsallocation to the available energy resource at that bus. On the other hand, a customer

may wish to install a generator to serve its load and export any surplus generation onto

the network. This constraint allows the effect of proposed generation at individual buses

on the overall DG capacity to be assessed. These constraints are formalised in Equation

(3.12), with the allocation to each bus being at least equal to any existing generation

and at most equal to the available energy resource at each bus.

PInstalled i PDG i PAvail i i N. (3.12)

WherePAvaili and PInstalled i are the available energy resource and any existing gener-

ation at the ith bus respectively.

The voltage levels and short circuit levels at the buses are interdependent, thus

some buses will have more influence on the voltage and short circuit levels than others

as illustrated later in Figures3.4 and 3.6. A flowchart of the methodology is shown in


45/133


Figure3.2.

START

LOAD FLOW ANALYSIS:CALCULATE BUS VOLTAGE

CHARACTERISTICS FORPOWER INJECTIONS

SHORT CIRCUIT ANALYSIS:CALCULATE SCL

CHARACTERISTICS FORTRANSMISSION BUS

CHECK EQUIPMENTRATINGS

INPUT MIN.LOAD DATA

INPUT MAX.SCL DATA

INPUT MIN.SCL DATA

SHORT CIRCUIT ANALYSIS:CALCULATE SCL CHARACTERISTICS

FOR ALL OTHER BUSES

FORMULATE LINEAR EQNS. FORCONSTRAINTS AND OBJECTIVE

FUNCTION

EXECUTE LINEAR PROGRAMMINGALGORITHM TO DETERMINE

INITIAL ALLOCATION

END

EXECUTE LINEAR PROGRAMMINGALGORITHM TO DETERMINE

OPTIMAL ALLOCATION

SHORT CIRCUIT ANALYSIS:RECALCULATE SCL

CHARACTERISTICS FORTRANSMISSION BUS

Figure 3.2: Optimal allocation methodology

There is a varying level of interdependence between the buses, which is dependent

on a number of factors:

Proximity to transmission station

Relative distance between buses

Type of conductor

The proximity of the bus to the transmission station has a large bearing on the

sensitivity of that bus to changes in voltage. As would be expected the closer a bus

is to the transmission in-feed the higher the voltage at that bus. In addition, these


46/133


buses tend to be less sensitive to generation at other buses and also less sensitive to

generation at the bus itself. The converse is also true. The distance between the buses

affects their level of interdependence and perhaps more importantly from an allocation

point of view, the relative distance will define each of their relationships. The type of

conductor also has a bearing on these relationships. For example a 630XLPE cable has

a lower impedance than a 100SCA overhead line and therefore the voltage rise effect

due to generation at other buses will be much reduced with the cable.

In terms of short circuit levels, the closer a bus is to the transmission station, the

higher the SCL. Once again the relative distance between the buses will determine

their interdependence as will the position of normally open points on the network.

The impedance of the conductors will affect the SCL and the typically high impedance

distribution networks will result in a low SCL.

3.2 Test System & Constraint Characteristics

The methodology was tested on a number of sample sections from the Irish distribution

network. Results are given here for a 38/110kV station with 5 buses shown in Figure

3.3.

The methodology has been tested on larger networks, however, this section is chosen

to illustrate the results as it demonstrates the potential for network sterilisation very

well. The most severe case for voltage rise is the standby feeding condition shown in

Figure3.3 with the 3.5km out of service line in grey. The individual voltage sensitivity

characteristics were calculated and are shown in Figure3.4.

From Figure3.4 it can be seen that bus A is the least sensitive to power injections

and therefore based on this constraint only, bus A will have the largest allocation of

generation. It is the least sensitive to power injections due to its proximity to the 110kV

station. It can also be seen that bus B and bus C have very similar profiles. They are

separated from the rest

andrew keane - integration of distributed generation

Documents